EPA/600/R-14/006
C-PPA Environmertal Protection Second External Review Draft
\r1Lmt JmAgency January 2015
www. epa. gov/ncea/isa
Integrated Science Assessment
for Oxides of Nitrogen-
Health Criteria
(Second External Review Draft)
January 2015
National Center for Environmental Assessment-RTF Division
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, NC
-------
DISCLAIMER
1 This document is the second external review draft, for review purposes only. This information is
2 distributed solely for the purpose of pre-dissemination peer review under applicable information quality
3 guidelines. It has not been formally disseminated by EPA. It does not represent and should not be
4 construed to represent any Agency determination or policy. Mention of trade names or commercial
5 products does not constitute endorsement or recommendation for use.
January 2015 ii DRAFT: Do Not Cite or Quote
-------
CONTENTS
INTEGRATED SCIENCE ASSESSMENT TEAM FOR OXIDES OF NITROGEN XIV
AUTHORS, CONTRIBUTORS, AND REVIEWERS XVII
CLEAN AIR SCIENTIFIC ADVISORY COMMITTEE OXIDES OF NITROGEN NAAQS REVIEW
PANEL XXI
ACRONYMS AND ABBREVIATIONS XXII
PREAMBLE XXXIII
1. Process of Integrated Science Assessment Development xxxiii
Figure I Schematic of the key steps in the process of the review of National
Ambient Air Quality Standards. xxxiv
2.
3.
4.
5.
6.
7.
Figure II Characterization of the general process of Integrated Science
Assessment (ISA) development.
Literature Search
Figure III Illustration of processes for literature search and study selection
used for development of Integrated Science Assessments.
Study Selection
Evaluation of Individual Study Quality
a. Atmospheric Science and Exposure Assessment
b. Epidemiology
c. Controlled Human Exposure and Animal Toxicology
d. Ecological and Other Welfare Effects
Evaluation, Synthesis, and Integration of Evidence across Disciplines and Development of
Scientific Conclusions and Causal Determinations
a. Evaluation, Synthesis, and Integration of Evidence across Disciplines
b. Considerations in Developing Scientific Conclusions and Causal Determinations
Table 1 Aspects to aid in judging causality.
c. Framework for Causal Determinations
Table II Weight of evidence for causal determination.
Public Health Impact
a. Approach to Identifying, Evaluating, and Characterizing At-Risk Factors
Table III Characterization of evidence for potential at-risk factors.
b. Evaluating Adversity of Human Health Effects
c. Concentration-Response Relationships
Public Welfare Impact
a. Evaluating Adversity of Ecological and Other Welfare Effects
b. Quantitative Relationships: Effects on Welfare
References for Preamble
XXXV
xxxvi
xxxviii
xxxviii
xxxix
xl
xli
xliii
xliv
xlv
xlv
xlix
Hi
liv
Iv
Ivii
Iviii
Ix
Ix
Ixi
Ixii
Ixiii
Ixv
Ixvi
PREFACE LXVIII
Legislative Requirements for the Review of the National Ambient Air Quality Standards Ixviii
Introduction to the Primary National Ambient Air Quality Standard for Nitrogen Dioxide Ixx
History of the Review of Air Quality Criteria for the Oxides of Nitrogen and the Primary National
Ambient Air Quality Standards for Nitrogen Dioxide Ixx
Table I History of the primary National Ambient Air Quality Standards for
nitrogen dioxide (NOz) since 1971. Ixxi
References for Preface Ixxv
EXECUTIVE SUMMARY LXXVII
Purpose and Scope of the Integrated Science Assessment Ixxvii
Sources and Human Exposure to Nitrogen Dioxide Ixxviii
Health Effects of Nitrogen Dioxide Exposure Ixxx
January 2015 iii DRAFT: Do Not Cite or Quote
-------
CONTENTS (Continued)
Table ES-1 Causal determinations for relationships between nitrogen dioxide
(NO2) exposure and health effects from the 2008 and current
Integrated Science Assessment (ISA) for Oxides of Nitrogen. Ixxxi
Short-term Nitrogen Dioxide Exposure and Respiratory Effects Ixxxii
Figure ES-1. Biological pathways for relationships of short-term and long-term
nitrogen dioxide (NO2) exposure with asthma. Ixxxiii
Long-term Nitrogen Dioxide Exposure and Respiratory Effects Ixxxiii
Nitrogen Dioxide Exposure and Other Health Effects Ixxxiv
Policy-Relevant Considerations for Health Effects Associated with Nitrogen Dioxide Exposure Ixxxv
Summary of Major Findings Ixxxvi
References for Executive Summary Ixxxviii
CHAPTER 1 INTEGRATED SUMMARY 1-1
1.1 Purpose and Overview of the Integrated Science Assessment 1-1
1.2 Process for Developing Integrated Science Assessments 1-3
1.3 Content of the Integrated Science Assessment 1-5
1.4 From Emissions Sources to Exposure to Nitrogen Dioxide 1-6
1.4.1 Emission Sources and Distribution of Ambient Concentrations 1-6
Figure 1-1 Reactions of oxides of nitrogen species in the ambient air. 1-8
1.4.2 Assessment of Human Exposure 1-9
1.4.3 Factors Potentially Correlated with Nitrogen Dioxide Exposure to Consider in Evaluating
Relationships with Health Effects 1-12
1.5 Health Effects of Nitrogen Dioxide Exposure 1-15
1.5.1 Respiratory Effects 1-16
Figure 1-2 Characterization of potential modes of action for health effects
related to exposure to nitrogen dioxide (NO2). 1-17
1.5.2 Health Effects beyond the Respiratory System 1-22
Table 1-1 Key evidence contributing to causal determinations for nitrogen
dioxide (NO2) exposure and health effects evaluated in the current
draft Integrated Science Assessment (ISA) for Oxides of Nitrogen. 1-31
1.6 Policy-Relevant Considerations 1-37
1.6.1 Durations of Nitrogen Dioxide Exposure Associated with Health Effects
1.6.2 Lag Structure of Relationships between Nitrogen Dioxide Exposure and Health Effects
1.6.3 Concentration-Response Relationships and Thresholds
1.6.4 Regional Heterogeneity in Effect Estimates
1.6.5 Public Health Significance
1.7 Conclusions
References for Chapter 1
1-37
1-39
1-40
1-42
1-44
1-48
1-51
CHAPTER 2 ATMOSPHERIC CHEMISTRY AND AMBIENT CONCENTRATIONS OF
OXIDES OF NITROGEN 2-1
2.1 Introduction 2-1
2.2 Atmospheric Chemistry and Fate 2-1
Figure 2-1 Schematic diagram of the cycle of reactive, oxidized nitrogen
species in the atmosphere. 2-3
2.3 Sources 2-10
2.3.1 Overview 2-10
Figure 2-2 U.S. national average A/Ox (sum of nitrogen dioxide and nitric oxide)
emissions from 1990 to 2013. 2-11
Figure 2-3 Major sources of A/Ox (sum of nitrogen dioxide and nitric oxide)
emissions averaged over the U. S. from the 2008 and 2011 National
Emissions Inventories. 2-12
Figure 2-4 Percentage contributions from major sources of the annual A/Ox
(sum of nitrogen dioxide and nitric oxide) emissions averaged over
the 21 largest U.S. Core-Based Statistical Areas with populations
greater than 2.5 million (blue—urban) compared to the national
average (red—national). 2-14
January 2015 ii DRAFT: Do Not Cite or Quote
-------
CONTENTS (Continued)
Table 2-1 Source distribution of the annual A/Ox (sum of nitrogen dioxide and
nitric oxide) emissions in the 21 largest U.S. Core-Based Statistical
Areas with populations greater than 2.5 million—2011 National
Emissions Inventory.
2.3.2 Highway Vehicles
2.3.3 Off-Hiqhway
2.3.4 Fuel Combustion — Utilities and Other
2-15
2-16
2-17
2-19
Figure 2-5 Fuel Combustion-Other emissions vs. average ambient January
temperature for the 21 largest U.S. Core-Based Statistical Areas
>2.5 million population. 2-20
2.3.5 Other Anthropogenic Sources 2-20
Table 2-2 Relative contributions to Other Anthropogenic A/Ox (sum of nitrogen
dioxide and nitric oxide) sources in selected cities3. 2-22
2.3.6 Biogenics and Wildfires 2-23
2.3.7 Emissions Summary 2-23
2.4 Measurement Methods 2-24
2.4.1 Federal Reference and Equivalent Methods 2-24
2.4.2 Other Methods for Measuring Nitrogen Dioxide 2-27
Figure 2-6 Comparison of nitrogen dioxide (NO2) measured by cavity attenuated
phase shift (CAPS) spectroscopy to NO2 measured by
chemiluminescence/MoOx catalytic converter (MC) for 4 days in
October 2007 in Billerica, MA. 2-28
Figure 2-7 Comparison of nitrogen dioxide (NO2) measured by quantum
cascade-tunable infrared differential absorption spectroscopy
(QC-TILDAS) to NO2 measured by chemiluminescence with
photolytic converter during April and May 2009 in Houston, TX. 2-29
2.4.3 Satellite Measurements of Nitrogen Dioxide 2-30
Figure 2-8 Seasonal average tropospheric column abundances for nitrogen
dioxide (NO2: 1015 molecules/cm2) derived by ozone monitoring
instrument (OMI) for winter (upper panel) and summer (lower panel)
for 2005 to 2007. 2-31
Figure 2-9 Seasonal average tropospheric column abundances for nitrogen
dioxide (NO2:1015 molecules/cm2) derived by ozone monitoring
instrument (OMI) for winter (upper panel) and summer (lower panel)
for 2010 to 2012. 2-32
2.4.4 Measurements of Total Oxides of Nitrogen in the Atmosphere 2-33
2.4.5 Ambient Sampling Network Design 2-34
Figure 2-10 Map of monitoring sites for oxides of nitrogen in the U.S. from four
networks. 2-35
2.5 Ambient Concentrations of Oxides of Nitrogen 2-36
2.5.1 National Scale Spatial Variability 2-36
Figure 2-11 98th percentiles of U.S. 1-hour daily maximum nitrogen dioxide
(NO2) concentrations (ppb) for 2011-2013. 2-38
Figure 2-12 U.S. annual average nitrogen dioxide (NO2) concentrations (ppb) for
2013. 2-39
Table 2-3 Summary statistics for 1-hour daily maximum nitrogen dioxide (NO2)
concentrations based on state and local air monitoring stations
(ppb). 2-40
Table 2-4 Summary statistics for nitrogen dioxide (NO2), nitric oxide (NO), and
A/Ox (sum of NO2 and NO) annual average concentrations based on
state and local air monitoring stations (ppb). 2-47
Figure 2-13 Seasonal average surface nitrogen dioxide (NO2) concentrations
in ppb for winter (upper panel) and summer (lower panel) derived by
ozone monitoring instrument (OMI)/Goddard Earth Observing
System (GEOS)-Chem for 2009-2011. 2-43
2.5.2 Urban-Scale Spatial Variability 2-44
Figure 2-14 Coefficient of divergence between monitor pairs in four U.S. cities. 2-45
Figure 2-15 Coefficient of divergence among a subset of five Los Angeles, CA
monitors. 2-46
Table 2-5A Percent difference in annual average nitrogen dioxide concentration
between monitors in Boston. 2-47
January 2015 iii DRAFT: Do Not Cite or Quote
-------
CONTENTS (Continued)
Table 2-56
Percent difference in annual average nitrogen dioxide concentration
between monitors in Los Angeles 2011.
2.5.3 Microscale- to Neighborhood-Scale Spatial Variability, Including near Roads
Table 2-6 Summary of near-road nitrogen dioxide concentration gradients from
different studies.
Figure 2-16 Influence of nitrogen dioxide concentration magnitude on the ratio of
NO2 concentrations at <1 m from the road (Cnear) to concentrations
at 200-350 m (Cfar) in rural Wales.
Figure 2-17 Diurnal variation of near-road (red: within 15m of major interstate)
and downwind (gray: within 100 m of major interstate) nitrogen
dioxide (NO2) concentrations observed during year-long field
campaigns in Los Angeles, CA and Detroit, Ml.
Figure 2-18 Absolute difference between nitrogen dioxide (NO2) concentrations
at near-road sites during year-long field campaigns in Los Angeles,
CA and Detroit, Ml.
Table 2-7
Table 2-8
Table 2-9A
Table 2-96
Comparison of near road and area wide 1-hour daily maximum
concentrations for monitors with year round data (ppb).
Near-road network 1-hour daily maximum nitrogen dioxide
concentration summary for first quarter 2014 (ppb)._
Roadside and urban background nitrogen dioxide concentrations in
London, U.K. 2010-2012.
Roadside and urban background nitrogen dioxide concentrations in
London, U.K. 2004-2006.
2.5.4
Seasonal, Weekday/Weekend, and Diurnal Trends
Figure 2-19 January and July hourly profiles of nitric oxide (NO) and nitrogen
dioxide (NO2) (ppb) for Atlanta, GA (site in Atlanta with maximum
1-hour NO2 concentrations).
2.5.5
Figure 2-20 Weekend/weekday hourly profiles of nitric oxide (NO) and nitrogen
dioxide (NO2) (ppb) for Atlanta, GA (site in Atlanta with maximum
NO2 concentrations).
Multiyear Trends in Ambient Measurements of Oxides of Nitrogen
Figure 2-21 U.S. national annual average ambient nitrogen dioxide concentration
trends, 1990-2012.
2.5.6 Background Concentrations
2.6 Conclusions
References for Chapter 2
2-48
^2-49
2-50
2-54
2-58
2-59
2-60
2-67
2-63
2-64
2-68
2-69
2-70
.2-70
2-71
2-72
'2-73
2-75
CHAPTER 3
EXPOSURE TO OXIDES OF NITROGEN
3.1 Introduction
3.2 Methodological Considerations for Use of Exposure Data
3.2.1 Measurement
3.2.2 Modeling
3.2.3
Choice of Exposure Metrics in Epidemiologic Studies
Table 3-1 Summary of sampling methods, their typical use in epidemiologic
Figure 3-1
studies, and related errors and uncertainties.
Average nitrogen dioxide concentrations measured in studies using
different monitor siting.
3.3 Characterization of Nitrogen Dioxide Exposures
3.3.1 Nitrogen Dioxide Concentration as an Indicator of Source-based Mixtures
3.3.2
Figure 3-2
Table 3-2
Table 3-3
Indoor Dynamics
Table 3-4
Spatial variability in concentrations of near-road pollutants, including
NO2, NO, A/Ox, CO, PM2.s, PMw, EC, benzene, VOCs, and UF
Particles. Concentrations are normalized by measurements at the
edge of the road.
Near- and on-road measurements of nitrogen dioxide (NO2), nitric
oxide (NO), and the sum of NO and NO2 (NOx).
Summary (mean, range) within 300 m of monitoring sites, by site
type, in a spatially dense monitoring campaign in New York City,
based on 2-week integrated samples per season.
-3-1
_3-2
_3-2
_3-5
.3-20
3-21
3-22
3-24
3-24
3-26
3-28
Indoor nitrogen dioxide (NO2) and nitrous acid (MONO)
concentrations in the presence and absence of combustion.
3-31
.3-33
3-34
January 2015
IV
DRAFT: Do Not Cite or Quote
-------
CONTENTS (Continued)
3.4 Exposure Assessment and Epidemiologic Inference 3-37
3.4.1 Conceptual Model of Total Personal Exposure 3-37
3.4.2 Personal-Ambient Relationships and Nonambient Exposures 3-40
Table 3-5 Ambient, outdoor, transport, indoor, and personal nitrogen dioxide
measurements (ppb) across studies. 3-42
Table 3-6 Correlations between measured nitrogen dioxide (NO2)
concentrations from personal, outdoor, indoor, and ambient
monitors. 3-49
Table 3-7 Meta regression results from 15 studies examining the relationship
between personal nitrogen dioxide exposure measurements and
ambient concentrations. 3-52
3.4.3 Factors Contributing to Error in Estimating Exposure to Ambient Nitrogen Dioxide 3-52
Figure 3-3 Distribution of time sample population spends in various
environments, from the U.S. National Human Activity Pattern Survey
(all ages). 3-53
Figure 3-4 Regional-scale variability in nitrogen dioxide for urban and rural area
data across the U.K. 3-55
Figure 3-5 Urban-scale variability in nitrogen dioxide (NO2) and the sum of nitric
oxide and NO2 (NOx) in Atlanta, GA. On the y-axis, y' denotes the
semivariogram, i.e., a unitless function that describes the ratio
between spatial and temporal variance of the differences between
two observations. 3-55
3.4.4 Confounding 3-57
Table 3-8 Synthesis of nitrogen dioxide ambient-ambient copollutant
correlations from measurements reported in the literature. 3-60
Figure 3-6 Summary of temporal nitrogen dioxide-copollutant correlation
coefficients from measurements reported in studies listed in Table
3-8, sorted by temporal averaging period. 3-70
Table 3-9 Pearson correlation coefficients between ambient nitrogen dioxide
and personal copollutants. 3-75
Table 3-10 Pearson correlation coefficients between personal nitrogen dioxide
and ambient copollutants. 3-75
Table 3-11 Pearson correlation coefficients between personal nitrogen dioxide
and personal copollutants. 3-76
Table 3-12 Correlation coefficients between indoor nitrogen dioxide and indoor
copollutants. 3-77
3.4.5 Implications for Epidemiologic Studies of Different Designs 3-82
Table 3-13 The influence of exposure metrics on error in health effect estimates. _ 3-86
3.5 Conclusions 3-94
References for Chapter 3 3-98
CHAPTER 4 DOSIMETRY AND MODES OF ACTION FOR INHALED OXIDES OF
NITROGEN 4-1
4.1 Introduction 4-1
4.2 Dosimetry of Inhaled Oxides of Nitrogen 4-2
4.2.1 Introduction 4-2
4.2.2 Dosimetry of Nitrogen Dioxide 4-3
Table 4-1 Small molecular weight antioxidantconcentrations in epithelial lining
fluid and predicted penetration distances for nitrogen dioxide. 4-10
4.2.3 Dosimetry of Nitric Oxide 4-22
4.2.4 Summary of Dosimetry 4-26
4.3 Modes of Action for Inhaled Oxides of Nitrogen 4-28
4.3.1 Introduction 4-28
Table 4-2 Chemical properties of nitrogen dioxide (NO2) and nitric oxide (NO)
4.3.2
4.3.3
4.3.4
4.3.5
that contribute to modes of action.
Nitrogen Dioxide
Nitric Oxide
Metabolites of Nitric Oxide and Nitrogen Dioxide
Mode of Action Framework
Figure 4-1 Mode of action of inhaled nitrogen dioxide (NO2): short-term
exposure and respiratory effects.
4-29
4-30
4-53
4-56
4-59
4-60
January 2015 v DRAFT: Do Not Cite or Quote
-------
CONTENTS (Continued)
Figure 4-2 Mode of action of inhaled nitrogen dioxide (NO2): long-term exposure
and respiratory effects. 4-62
Figure 4-3
Figure 4-4
4.4 Summary
References for Chapter 4
Mode of action of inhaled nitrogen dioxide (NO2): short-term and
long-term exposure and extrapulmonary effects.
Mode of action of inhaled nitric oxide (NO).
4-63
4-65
4-65
4-66
CHAPTER 5 INTEGRATED HEALTH EFFECTS OF SHORT-TERM EXPOSURE TO
OXIDES OF NITROGEN
5.1 Introduction
5.1.1 Scope of Chapter
5.1.2 Evidence Evaluation and Integration to Form Causal Determinations
Table 5-1 Summary and description of scientific considerations for evaluating
the quality of studies on the health effects of oxides of nitrogen.
5.2 Respiratory Effects
5.2.1 Introduction
5.2.2 Asthma Exacerbation
Table 5-2
Table 5-3
Table 5-4
Table 5-5
Table 5-6
Figure 5-1
Figure 5-2
Table 5-7
Table 5-8
Figure 5-3
Table 5-9
Table 5-10
Table 5-11
Figure 5-4
Table 5-12
Table 5-13
Table 5-14
Resting exposures to nitrogen dioxide (NO2) and airway
responsiveness in individuals with asthma._
Exercising exposures to nitrogen dioxide (NO2) and
airwayresponsiveness in individuals with asthma.
Fraction of individuals with asthma having nitrogen dioxide
(NO2)-induced increase in airway responsiveness to a non-specific
challenge.
Fraction of individuals with asthma having nitrogen dioxide
(NO2)-induced increase in specific airway responsiveness to an
allergen challenge.
Fraction of individuals with asthma having nitrogen dioxide
(NO2)-induced increase in airway responsiveness regardless of
challenge types.
Change in provocative dose (dPD) due to exposure to nitrogen
dioxide (NO2) in resting individuals with asthma.
Log-normal distribution of change in provocative dose (dPD) due to
exposure to nitrogen dioxide in resting individuals with asthma.
_ 5-1
_5-1
_5-2
_ 5-3
5-13
5-13
^5-15
5-21
5-22
5-26
5-27
5-28
5-29
5-31
Sensitivity analysis for distribution of responses and nitrogen dioxide
(NO2)-induced increase in responsiveness to a nonspecific
challenge. 5-33
Mean and upper percentile concentrations of nitrogen dioxide (NO2)
in epidemiologic studies of lung function in populations with asthma. 5-44
Associations of nitrogen dioxide (NO2) ambient concentrations or
personal exposure with percentage change in forced expiratory
volume (FEVi) (top plot) and change in percent predicted FEVi
(bottom plot) in children and adults with asthma. 5-47
Epidemiologic studies of lung function in children and adults with
asthma.
Controlled human exposure studies of individuals with asthma.
Mean and upper percentile concentrations of nitrogen dioxide (NO2)
in epidemiologic studies of respiratory symptoms in populations with
asthma.
Associations of ambient nitrogen dioxide (NO2) concentrations with
respiratory symptoms and asthma medication use in children with
asthma.
Epidemiologic studies of respiratory symptoms and asthma
medication use in children with asthma.
Controlled human exposure studies of respiratory symptoms.
Mean and upper percentile concentrations of oxides of nitrogen in
studies of asthma hospital admissions and emergency department
(ED) visits.
5-48
5-61
5-63
5-67
5-68
5-79
5-81
January 2015
VI
DRAFT: Do Not Cite or Quote
-------
CONTENTS (Continued)
Table 5-15
Figure 5-5
Figure 5-6
Figure 5-7
Table 5-16
Table 5-17
Table 5-18
Table 5-19
Figure 5-8
Copollutant model results from Iskandar et al. (2012) for a 20-ppb
increase in 24-h avg nitrogen dioxide (NO2) concentrations and a
40-ppb increase in 24-h avg A/Ox (sum of NO and A/Cy
concentrations.
Locally weighted scatterplot smoothing concentration-response
estimates (solid line) and twice-standard error estimates (dashed
lines) from generalized additive models for associations between
3-day avg (lag 0-2) nitrogen dioxide (NO2) concentrations and
emergency department visits for pediatric asthma at the 5th to 95th
percentile of NO2 concentrations in the Atlanta, GA area.
5-88
Rate ratio and 95% confidence intervals for single-pollutant and joint
effect models for each pollutant combination in warm and cold
season analyses for an interquartile range (IQR) increase in each
pollutant at lag 0-2 days. IQR for 1-h max nitrogen dioxide (NO2)
concentrations = 12.87 ppb.
5-94
5-97
Percentage increase in asthma hospital admissions and emergency
department (ED) visits from U. S. and Canadian studies evaluated in
the 2008 Integrated Science Assessment for Oxides of Nitrogen and
recent studies in all-year and seasonal analyses._
Corresponding risk estimates for studies presented in Figure 5-7.
Controlled human exposure studies of pulmonary inflammation in
populations with asthma.
Animal toxicological studies of pulmonary inflammation.
Mean and upper percentile concentrations of nitrogen dioxide (NO2)
in epidemiologic studies of pulmonary inflammation and oxidative
stress in populations with asthma.
5-100
5-101
5-104
5-109
Associations of personal or ambient nitrogen dioxide (NO2) with
exhaled nitric oxide (eNO) in populations with asthma.
Table 5-20 Epidemiologic studies of pulmonary inflammation and oxidative
stress in children and adults with asthma.
5.2.3 Allergy Exacerbation
Table 5-21 Epidemiologic studies of allergy exacerbation.
5.2.4 Exacerbation of Chronic Obstructive Pulmonary Disease
Table 5-22 Epidemiologic panel studies of adults with chronic obstructive
pulmonary disease (COPD).
Controlled human exposure studies of respiratory symptoms in
adults with chronic obstructive pulmonary disease (COPD)._
Table 5-23
Table 5-24
Figure 5-9
Mean and upper percentile concentrations of nitrogen dioxide (NO2)
in studies of hospital admission and emergency department visits for
chronic obstructive pulmonary disease.
Percentage increase in chronic obstructive pulmonary disease
hospital admissions and emergency department (ED) visits in
relation to nitrogen dioxide concentrations from U.S. and Canadian
studies evaluated in the 2008 Integrated Science Assessment for
Oxides of Nitrogen and recent studies.
5-110
5-113
5-114
5-128
5-729
5-131
5-133
5-137
5-139
5.2.5
Table 5-25 Corresponding risk estimate for studies presented in Figure 5-9. _
Respiratory Infection
Table 5-26 Animal toxicological studies of susceptibility to infection.
Table 5-27
Table 5-28
Table 5-29
Controlled human exposure studies of susceptibility to infection. _
Epidemiologic studies of respiratory infections reported or diagnosed
in children.
Mean and upper percentile concentrations of nitrogen dioxide (NO2)
in studies of hospital admissions and emergency department visits
for respiratory infection.
5-144
5-145
5-146
5-148
5-152
5-154
Figure 5-10 Percentage increase in respiratory infection-related hospital
admissions and Emergency Department (ED) visits in relation to
nitrogen dioxide concentrations from U.S. and Canadian studies
evaluated in the 2008 Integrated Science Assessment for Oxides of
Nitrogen and recent studies. _
Table 5-30
Table 5-31
Corresponding risk estimate for studies presented in Figure 5-10.
Animal toxicological studies of subclinical lung host defense effects. _
5-157
5-164
5-165
5-168
January 2015
VII
DRAFT: Do Not Cite or Quote
-------
CONTENTS (Continued)
5.2.6
Table 5-32 Controlled human exposure studies of subclinical lung host defense
effects.
Aggregated Respiratory Conditions
Table 5-33 Mean and upper percentile concentrations of nitrogen dioxide in
studies of hospital admissions and emergency department visits for
aggregrated respiratory conditions.
5-770
5-171
Figure 5-11
Risk ratio and 95% confidence intervals for associations between
various lag 1 day nitrogen dioxide (NO2) metrics and respiratory
emergency department visits. _
Figure 5-12 Spatial correlations for nitrogen dioxide (NO2) metrics in the Atlanta,
GA area.
Figure 5-13 Percentage increase in all respiratory disease hospital admissions
and emergency department (ED) visits in relation to nitrogen dioxide
concentrations from U. S. and Canadian studies evaluated in the
2008 Integrated Science Assessment for Oxides of Nitrogen and
recent studies.
5-174
5-181
5-182
5.2.7
Table 5-34 Corresponding risk estimate for studies presented in Figure 5-13.
Respiratory Effects in Healthy Populations
Table 5-35 Mean and upper percentile oxides of nitrogen concentrations in
epidemiologic studies of lung function in the general population.
Epidemiologic studies of lung function in children and adults in the
general population.
Controlled human exposure studies of lung function and respiratory
symptoms in healthy adults.
Table 5-36
Table 5-37
Table 5-38
Table 5-39
Table 5-40
Mean and upper percentile concentrations of nitrogen dioxide (NO2)
in epidemiologic studies of respiratory symptoms in children in the
general population.
Epidemiologic studies of respiratory symptoms in children in the
general population.
Mean and upper percentile concentrations of oxides of nitrogen in
epidemiologic studies of pulmonary inflammation and oxidative
stress in the general population. _
Figure 5-14 Associations between ambient nitrogen dioxide (NO2) concentrations
and exhaled nitric oxide (eNO) among children and adults in the
general population.
Table 5-41
Table 5-42
Epidemiologic studies of pulmonary inflammation, injury, and
oxidative stress in children and adults in the general population. _
Controlled human exposure studies of pulmonary inflammation,
injury, and oxidative stress in healthy adults.
Table 5-43 Animal toxicological studies of pulmonary inflammation, injury, and
oxidative stress.
5.2.8 Respiratory Mortality
Figure 5-15 City-specific concentration-response curves of nitrogen dioxide and
daily chronic obstructive pulmonary disease (COPD) mortality in four
Chinese cities.
5.2.9
Summary and Causal Determination
Figure 5-16 Associations of ambient or personal NO2 with respiratory effects
adjusted for PM2.5, EC/BC, or PNC/UFP.
Figure 5-17 Associations of ambient nitrogen dioxide (NO2) with respiratory
effects adjusted for VOCs or CO.
Table 5-44
Table 5-45
Corresponding effect estimates for nitrogen dioxide (NO2)-
associated respiratory effects in single- and co-pollutant models
presented in Figures 5-16 and 5-17. _
Summary of evidence for a causal relationship between short-term
nitrogen dioxide (NO2) exposure and respiratory effects.
5.3 Cardiovascular and Related Metabolic Effects
5.3.1 Introduction
5.3.2 Myocardial Infarction
Figure 5-18 Results of studies of short-term exposure to oxides of nitrogen and
hospital admissions forischemic heart disease.
Table 5-46 Corresponding risk estimates for hospital admissions for ischemic
heart disease for studies presented in Figure 5-18.
5-183
5-184
5-185
5-787
5-191
5-202
5-205
5-208
5-212
5-215
5-216
5-222
5-225
5-232
5-235
5-235
5-245
5-246
5-247
5-250
5-255
5-255
\ 5-256
5-258
5-259
January 2015
Vlll
DRAFT: Do Not Cite or Quote
-------
CONTENTS (Continued)
Table 5-47 Epidemiologic studies of ST-segment amplitude. 5-265
5.3.3 Arrhythmia and Cardiac Arrest 5-266
Table 5-48 Epidemiologic studies of arrhythmia and cardiac arrest. 5-268
Table 5-49 Epidemiologic studies of out-of-hospital cardiac arrest. 5-270
5.3.4 Cerebrovascular Disease and Stroke 5-271
Figure 5-19 Results of studies of short-term exposure to oxides of nitrogen and
hospital admissions for cerebrovascular disease and stroke. 5-273
Table 5-50 Corresponding risk estimates for hospital admissions for
cerebrovascular disease and stroke for studies presented in
Figure 5-19.
5.3.5 Decompensation of Heart Failure
5.3.6 Increased Blood Pressure and Hypertension
Table 5-51 Epidemiologic studies of blood pressure.
5.3.7 Venous Thromboembolism
5.3.8 Cardiometabolic Effects
5.3.9 Aggregated Cardiovascular Effects
5-274
5-278
5-279
5-280
5-283
5-284
5-284
Figure 5-20 Results of studies of short-term exposure to oxides of nitrogen and
hospital admissions for all cardiovascular disease. 5-286
Table 5-52 Corresponding effect estimates for hospital admissions for all
cardiovascular disease studies presented in Figure 5-20. 5-287
5.3.10 Cardiovascular Mortality 5-289
Figure 5-21 Pooled concentration-response curve for nitrogen dioxide (NO2) and
daily stroke mortality in eight Chinese cities for lag 0-1 day. 5-292
5.3.11 Subclinical Effects Underlying Cardiovascular Effects 5-292
Table 5-53 Epidemiologic studies of heart rate/heart rate variability. 5-294
Table 5-54 Epidemiologic studies of QT-interval duration. 5-303
Table 5-55 Epidemiologic studies ofbiomarkers of cardiovascular effects. 5-305
Table 5-56 Controlled human exposure studies of short-term nitrogen dioxide
(NO2) exposure and cardiovascular effects. 5-316
Table 5-57 Animal toxicological studies of short-term nitrogen dioxide (NO2)
exposure and cardiovascular effects. 5-320
5.3.12 Summary and Causal Determination 5-322
Table 5-58 Summary of evidence, which is suggestive, but not sufficient, to infer
a causal relationship between short-term nitrogen dioxide (NO2)
exposure and cardiovascular and related metabolic effects. 5-328
5.4 Total Mortality 5-337
5.4.1 Introduction and Summary of 2008 Integrated Science Assessment for Oxides of
Nitrogen 5-331
5.4.2 Associations between Short-Term Nitrogen Dioxide Exposure and Mortality 5-332
5.4.3 Associations between Short-term Nitrogen Dioxide Exposure and Mortality in All-Year
Analyses 5-333
Table 5-59 Air quality characteristics of studies evaluated in the 2008 Integrated
Science Assessment for Oxides of Nitrogen and recently published
multicity and select single-city studies. 5-334
Figure 5-22 Summary of multicity studies evaluated in the 2008 Integrated
Science Assessment for Oxides of Nitrogen and recently published
studies that examined the association between short-term nitrogen
dioxide exposure and total mortality. 5-337
Table 5-60 Corresponding percentage increase in total mortality (95% Cl) for
Figure 5-22 5-338
Figure 5-23 Percentage increase in total, cardiovascular, and respiratory
mortality from multicity studies for a 20-ppb increase in
24-hour average or 30-ppb increase in 1-hour maximum nitrogen
dioxide concentrations. 5-339
Table 5-61 Corresponding percentage increase (95% Cl) for Figure 5-23. 5-340
5.4.4 Potential Confounding of the Nitrogen Dioxide-Mortality Relationship 5-341
Table 5-62 Percentage increase in total and cause-specific mortality for a
20-ppb increase in 24-hour average NO2 concentrations in single-
and co-pollutant models with PMio in all-year analyses or Os in
summer season analyses. 5-343
5.4.5 Modification of the Nitrogen Dioxide-Mortality Relationship 5-345
5.4.6 Potential Seasonal Differences in the Nitrogen Dioxide-Mortality Relationship 5-346
January 2015 ix DRAFT: Do Not Cite or Quote
-------
CONTENTS (Continued)
5.4.7 Nitrogen Dioxide-Mortality Concentration-Response Relationship and Related Issues 5-348
Figure 5-24 Percentage increase in total and cause-specific mortality due to
short-term nitrogen dioxide exposure at single day lags, individual
lag days of a constrained polynomial distributed lag model, and
multiday lags of an unconstrained distributed lag model. 5-349
Figure 5-25 Percentage increase in total and cause-specific mortality due to
short-term nitrogen dioxide exposure in single- and multi-day lag
models. 5-350
Figure 5-26 Flexible ambient concentration-response relationship between short-
term nitrogen dioxide (NO2, in ppb) exposure and mortality at lag
day 1. Pointwise means and 95% CIs adjusted for size of the
bootstrap sample. 5-352
Figure 5-27 CAPES concentration-response curve for the association between
total and cause-specific mortality and 24-hour average nitrogen
dioxide (NO2) concentrations at lag 0-1 days. 5-354
Figure 5-28 Concentration-response curve for association between total mortality
and 24-hour average nitrogen dioxide (NO2) concentrations at lag
0-1 days in the four cities of the Public Health and Air Pollution in
Asia study. 5-356
5.4.8 Summary and Causal Determination 5-357
Table 5-63 Summary of evidence, which is suggestive, but not sufficient to infer,
a causal relationship between short-term nitrogen dioxide (NO2)
exposure and total mortality. 5-359
References for Chapter 5 5-367
CHAPTER 6 INTEGRATED HEALTH EFFECTS OF LONG-TERM EXPOSURE TO
OXIDES OF NITROGEN 6-1
6.1 Scope and Issues Considered in Health Effects Assessment 6-1
6.1.1 Scope of Chapter 6-1
6.1.2 Evidence Evaluation and Integration to Form Causal Determinations 6-2
6.2 Respiratory Effects 6-4
6.2.1 Introduction 6-4
6.2.2 Development of Asthma or Chronic Bronchitis 6-5
Table 6-1 Prospective cohort studies of long-term exposure to nitrogen dioxide
(NO2) or sum of NO2 and nitric oxide (NOx) and asthma incidence in
children. 6-7
Figure 6-1 Associations of long-term exposure to nitrogen dioxide (NO2), nitric
oxide (NO), and the sum of NO and NO2 (NOx) with asthma
incidence from prospective studies of children. 6-16
Figure 6-2 Adjusted overall and age-specific association between annual
average levels of air pollution at the birth address and asthma during
the first 8 years of life. 6-18
Table 6-2 Animal toxicological studies of the respiratory effects of long-term
nitrogen dioxide (NO2) exposure. 6-27
6.2.3 Severity of Asthma, Chronic Bronchitis, and Chronic Obstructive Pulmonary Disease:
Respiratory Symptoms and Hospital Admissions 6-31
Table 6-3 Prospective studies of long-term nitrogen dioxide exposure and
respiratory symptoms in children. 6-33
Figure 6-3 Concentration-response relationships between respiratory effects
and indoor nitrogen dioxide (NO2) illustrated with constrained,
natural spline functions (solid lines) with 95% confidence limits
(small dashed lines) and threshold function (bold dashed line) from
hierarchical ordered logistic regression models. 6-37
Figure 6-4 Odds ratios for within-community bronchitis symptoms associated
with nitrogen dioxide (NO2), adjusted for other pollutants in
copollutant models for the 12 communities of the Children's Health
Study. 6-39
6.2.4 Development of Allergic Disease 6-42
6.2.5 Lung Function and Lung Development 6-45
Table 6-4 Prospective studies of long-term nitogen dioxide exposure and lung
function and lung development in children. 6-46
January 2015 x DRAFT: Do Not Cite or Quote
-------
CONTENTS (Continued)
Figure 6-5 Community-specific average growth in forced expiratory volume in
1 second (FEVi; mL) among girls and boys during the 8-year period
from 1993 to 2001, plotted against average nitrogen dioxide (NO2)
concentrations from 1994 through 2000. 6-54
Figure 6-6 Community-specific proportion of 18-year-olds with a forced
expiratory volume in 1 second (FEVi) below 80% of the predicted
value, plotted against the average concentrations of nitrogen dioxide
(NO2) from 1994 through 2000. 6-56
Figure 6-7 Associations of nitrogen dioxide (NO2) or the sum of nitric oxide and
NO2 (NOx) with lung function indices from prospective studies. 6-59
6.2.6 Changes in Lung Morphology 6-62
6.2.7 Respiratory Infection 6-64
6.2.8 Chronic Obstructive Pulmonary Disease 6-67
6.2.9 Summary and Causal Determination 6-69
Table 6-5 Summary of key evidence for a likely to be a causal relationship
between long-term nitrogen dioxide (NO2) exposure and respiratory
effecfs.
6.3 Cardiovascular and Related Metabolic Effects
6.3.1 Introduction
6.3.2 Heart Disease
6-75
6-80
6-80
6-82
Table 6-6 Epidemiologic studies of the association of long-term exposure to
nitrogen dioxide (NO2) or the sum of nitric oxide and NO2 (NOx) with
heart disease. 6-82
6.3.3 Diabetes 6-88
Table 6-7 Epidemiologic studies of the association of long-term exposure to
nitrogen dioxide (NO2) or the sum of nitric oxide and NO2 (NOx) with
cardiometabolic disorders. 6-90
6.3.4 Cerebrovascular Disease and Stroke 6-92
Table 6-8 Epidemiologic studies of the association of long-term exposure to
nitrogen dioxide (NO2) or the sum of nitric oxide and NO2 (NOx) with
cerebrovascular disease or stroke. 6-95
6.3.5 Hypertension 6-97
Table 6-9 Epidemiologic studies of the association of long-term exposure to
nitrogen dioxide (NO2) or the sum of nitric oxide and NO2 (NOx) with
hypertension and blood pressure. 6-98
6.3.6 Markers of Cardiovascular Disease 6-102
Table 6-10 Study details for toxicological studies examining cardiovascular
effects from long-term nitrogen dioxide (NO2) exposure. 6-104
6.3.7 Inflammation and Oxidative Stress 6-105
6.3.8 Cardiovascular Mortality 6-107
6.3.9 Summary and Causal Determination 6-107
Table 6-11 Summary of evidence, which is suggestive, but not sufficient, to infer
a causal relationship between long-term nitrogen dioxide (NO2)
exposure and cardiovascular and related metabolic effects.
6.4 Reproductive and Developmental Effects
6.4.1 Introduction
6.4.2 Fertility, Reproduction, and Pregnancy
Table 6-12 Key reproductive and developmental epidemiologic studies for
nitrogen dioxide (NO2).
6.4.3 Birth Outcomes
6.4.4 Postnatal Development
6-110
6-115
6-115
6-118
6-122
6-130
6-140
Table 6-13 Reproductive and developmental toxicological studies for nitrogen
dioxide (NO2). 6-147
6.4.5 Summary and Causal Determination 6-148
Table 6-14 Summary of evidence supporting the causal determinations for
relationships between long-term nitrogen dioxide (NO2) exposure
and reproductive and developmental effects. 6-151
6.5 Total Mortality 6-154
6.5.1 Review of Mortality Evidence from 2008 Integrated Science Assessment for Oxides of
Nitrogen 6-154
6.5.2 Recent Evidence for Mortality from Long-term Exposure to Oxides of Nitrogen 6-157
January 2015 xi DRAFT: Do Not Cite or Quote
-------
CONTENTS (Continued)
Figure 6-8 Results of studies of long-term exposure to nitrogen dioxide (NO2) or
the sum of nitric oxide and NO2 (NOx) and total mortality. 6-161
Table 6-15 Corresponding risk estimates for Figure 6-8. 6-762
Figure 6-9 Results of studies of long-term exposure to nitrogen dioxide (NO2),
nitric oxide (NO), or the sum of NO and NO2 (NOx) and
cardiovascular mortality. 6-164
Table 6-16 Corresponding risk estimates for Figure 6-9. 6-165
Figure 6-10 Results of studies of long-term exposure to nitrogen dioxide (NO2),
nitric oxide (NO), or the sum of NO and NO2 (NOx), and respiratory
6.5.3
mortality.
Table 6-1 7 Corresponding risk estimates for Figure 6-10.
Summary and Causal Determination
Table 6-1 8 Summary of evidence, which is suggestive, but not sufficient, to infer
a causal relationship between long-term nitrogen dioxide (NO2)
exposure and total mortality.
6.6 Cancer
6.6.1
6.6.2
6.6.3
6.6.4
6.6.5
6.6.6
6.6.7
6.6.8
6.6.9
Lunq Cancer
Table 6-19 Animal toxicological studies of carcinogenicity and genotoxicity with
exposure to nitrocien dioxide (NO2).
Leukemia Incidence and Mortality
Bladder Cancer Incidence and Mortality
Breast Cancer Incidence
Prostate Cancer Incidence
Other Cancers Incidence and Mortality
Production of N-Nitroso Compounds and other Nitro Derivatives
Genotoxicity
Summary and Causal Determination
Table 6-20 Summary of evidence, which is suggestive, but not sufficient, to infer
6-168
6-169
6-171
6-773
6-775
6-175
6-183
6-185
6-186
6-186
6-187
6-188
6-188
6-189
6-190
a causal relationship between long-term nitrogen dioxide (NO2)
exposure and cancer. 6-192
References for Chapter 6 6-194
CHAPTER 7 POPULATIONS AND LIFESTAGES POTENTIALLY AT RISK FOR HEALTH
EFFECTS RELATED TO NITROGEN DIOXIDE EXPOSURE 7-1
7.1 Introduction 7-1
7.2 Approach to Evaluating and Characterizing the Evidence forAt-Risk Factors 7-2
Table 7-1 Characterization of evidence for factors potentially increasing the
risk for nitrogen dioxide-related health effects. 7-3
7.3 Pre-Existing Disease/Conditions 7-3
Table 7-2 Prevalence of respiratory diseases, cardiovascular diseases,
diabetes, and obesity among adults by age and region in the U.S. in
2010. 7-4
7.3.1 Asthma 7-5
Table 7-3 Controlled human exposure studies evaluating pre-existing asthma. 7-7
Table 7-4 Epidemiologic studies evaluating pre-existing asthma. 7-8
7.3.2 Chronic Obstructive Pulmonary Disease 7-8
Table 7-5 Controlled human exposure studies evaluating pre-existing COPD. 7-10
Table 7-6 Epidemiologic studies evaluating pre-existing COPD. 7-11
7.3.3 Cardiovascular Disease 7-11
Table 7-7 Epidemiologic studies evaluating pre-existing cardiovascular
disease. 7-13
Table 7-8 Controlled human exposure and toxicological studies informing
pre-existing cardiovascular disease. 7-15
7.3.4 Diabetes 7-15
Table 7-9 Epidemiologic studies evaluating pre-existing diabetes. 7-76
7.3.5 Obesity 7-17
Table 7-10 Toxicological study evaluating pre-existing obesity. 7-78
Table 7-11 Epidemiologic studies evaluating pre-existing obesity. 7-79
7.4 Genetic Factors 7-20
Table 7-12 Epidemiologic studies evaluating genetic factors. 7-23
7.5 Sociodemographic Factors 7-26
January 2015 xii DRAFT: Do Not Cite or Quote
-------
CONTENTS (Continued)
7.5.1 Lifestage 7-26
Table 7-13 Epidemiologic studies evaluating childhood lifestage. 7-29
Table 7-14 Toxicological studies informing childhood lifestage. 7-30
Table 7-15 Epidemiologic studies evaluating older adult lifestage. 7-32
Table 7-16 Controlled human exposure studies informing older adult lifestage. 7-35
7.5.2 Socioeconomic Status 7-35
Table 7-17 Epidemiologic studies evaluating socioeconomic status. 7-38
7.5.3 Race/Ethnicity 7-42
Table 7-18 Epidemiologic studies evaluating race/ethnicity. 7-43
7.5.4 Sex 7-44
Table 7-19 Epidemiologic studies evaluating sex. 7-46
7.5.5 Residence in Urban Areas 7-50
Table 7-20 Epidemiologic studies evaluating urban residence. 7-52
7.5.6 Proximity to Roadways 7-52
Figure 7-1 Map of population density in Los Angeles, CA in relation to primary
and secondary roads. 7-54
Table 7-21 Epidemiologic studies evaluating proximity to roadways (all
long-term exposure). 7-57
7.6 Behavioral and Other Factors 7-58
7.6.1 Diet 7-58
Table 7-22 Controlled human exposure and toxicological studies evaluating diet _ 7-60
Table 7-23 Epidemiologic studies evaluating diet (all long-term exposure). 7-67
7.6.2 Smoking 7-61
Table 7-24 Epidemiologic studies evaluating smoking status. 7-62
7.6.3 Physical Activity 7-64
Table 7-25 Epidemiologic studies evaluating physical activity (all long-term
exposure). 7-66
7.7 Conclusions 7-66
Table 7-26 Summary of evidence for potential increased NO2 exposure and
increased risk of NO2-related health effects. 7-68
References for Chapter 7 7-70
January 2015 xiii DRAFT: Do Not Cite or Quote
-------
INTEGRATED SCIENCE ASSESSMENT TEAM FOR
OXIDES OF NITROGEN
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
Dr. Reeder Sams II (Acting Deputy Director)—National Center for Environmental
Assessment-RTP 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
Dr. Steven J. Dutton (Acting Branch Chief)—National Center for Environmental
Assessment, Office of Research and Development, U.S. Environmental Protection
Agency, Research Triangle Park, NC
Scientific Staff
Dr. Molini M. Patel (Team Leader, Integrated Science Assessment for Oxides of
Nitrogen)—National Center for Environmental Assessment, Office of Research and
Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
Ms. Breanna Alman—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. James Brown—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
Mr. Evan Coffman—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. Laura Datko-Williams—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. Erin Hines—National Center for Environmental Assessment, Office of Research and
Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Ellen Kirrane—National Center for Environmental Assessment, Office of Research and
Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Dennis Kotchmar—National Center for Environmental Assessment, Office of Research
and Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
January 2015 xiv DRAFT: Do Not Cite or Quote
-------
Dr. Thomas Luben—National Center for Environmental Assessment, Office of Research and
Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Stephen McDow—National Center for Environmental Assessment, Office of Research
and Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
Ms. Connie Meacham—National Center for Environmental Assessment, Office of Research
and Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
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
Dr. Elizabeth Oesterling Owens—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
Dr. Kristen Rappazzo—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. Jennifer Richmond-Bryant—National Center for Environmental Assessment, Office of
Research and Development, 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. Tina Stevens—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. David Svendsgaard—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
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
Technical Support Staff
Ms. Marieka Boyd—National Center for Environmental Assessment, Office of Research and
Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
January 2015 xv DRAFT: Do Not Cite or Quote
-------
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. Nathan Ellenfield—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. Gerald Gurevich—National Center for Environmental Assessment, Office of Research
and Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
Ms. Katie Jelen—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. Ryan Jones—National Center for Environmental Assessment, Office of Research and
Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
Ms. Diane LeBlond—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. Danielle Moore—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. Candis O'Neal—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. Sandy Pham—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. 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
January 2015 xvi DRAFT: Do Not Cite or Quote
-------
AUTHORS, CONTRIBUTORS, AND REVIEWERS
Authors
Dr. Molini M. Patel (Team Leader, Integrated Science Assessment for Oxides of
Nitrogen)—National Center for Environmental Assessment, Office of Research and
Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
Ms. Breanna Alman—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. James Brown—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
Mr. Evan Coffman—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. Rachelle Duvall—National Exposure Research Laboratory, 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
Ms. Laura Datko-Williams—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. Ellen Kirrane—National Center for Environmental Assessment, Office of Research and
Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Dennis Kotchmar—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. Stephen McDow—National Center for Environmental Assessment, Office of Research
and Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
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
Dr. Elizabeth Oesterling Owens—National Center for Environmental Assessment, Office of
Research and Development, U.S. Environmental Protection Agency, Research Triangle
Park, NC
January 2015 xvii DRAFT: Do Not Cite or Quote
-------
Dr. Jennifer Peel—Department of Environmental and Radiological Health Sciences,
Colorado School of Public Health, Colorado State University, Fort Collins, CO
Dr Joseph P. 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, Ryals School for
Public Health, University of Alabama at Birmingham, Birmingham, AL
Dr. Kristen Rappazzo—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. Jennifer Richmond-Bryant—National Center for Environmental Assessment, Office of
Research and Development, 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. Tina Stevens—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. Giuseppe Squadrito—Department of Environmental Health Sciences, Ryals School for
Public Health, University of Alabama at Birmingham, Birmingham, AL
Dr. David Svendsgaard—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
Dr. Lisa Vinikoor-Imler—National Center for Environmental Assessment, Office of
Research and Development, U.S. Environmental Protection Agency, Research Triangle
Park, NC
Dr. Gregory Wellenius—Department of Community Health (Epidemiology Section), Brown
University, Providence, RI
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
Contributors
Ms. Rachel Housego—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. Meagan Madden—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. April Maxwell—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
January 2015 xviii DRAFT: Do Not Cite or Quote
-------
Dr. Lok Namsal—Universities Space Research Association, NASA Goddard Space Flight
Center, Greenbelt, MD
Dr. Havala Pye—National Exposure Research Laboratory, 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
Reviewers
Mr. Chad Bailey—Office of Transportation and Air Quality, Office of Air and Radiation,
U.S. Environmental Protection Agency, Ann Arbor, MI
Dr. Lisa Baxter—National Exposure Research Laboratory, Office of Research and
Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Melinda Beaver—National Exposure Research Laboratory, Office of Research and
Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
Mr. Roger Erode—Office of Air Quality Planning and Standards, Office of Air and
Radiation, U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Philip Bromberg—School of Medicine, University of North Carolina, Chapel Hill, NC
Mr. Matthew Davis—Office of Children's Health Protection, U.S. Environmental Protection
Agency, Washington, DC
Dr. Steven J. Dutton—National Center for Environmental Assessment, Office of Research
and Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Mark Frampton—Pulmonary and Critical Care Division, Department of Medicine,
University of Rochester School of Medicine, Rochester, NY
Dr. Terry Gordon—Department of Environmental Medicine, New York University School
of Medicine, Tuxedo, NY
Dr. Gary Hatch—National Health and Environmental Effects Research Laboratory, Office of
Research and Development, U.S. Environmental Protection Agency, Research Triangle
Park, NC
Ms. Beth Hassett-Sipple—Office of Air Quality Planning and Standards, Office of Air and
Radiation, U.S. Environmental Protection Agency, Research Triangle Park, NC
Mr. Marc Houyoux—Office of Air Quality Planning and Standards, Office of Air and
Radiation, 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. Thomas Long—National Center for Environmental Assessment, Office of Research and
Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Deborah Luecken—National Exposure Research Laboratory, Office of Research and
Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
Ms. Connie Meacham—National Center for Environmental Assessment, Office of Research
and Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
January 2015 xix DRAFT: Do Not Cite or Quote
-------
Dr. Deirdre Murphy—Office of Air Quality Planning and Standards, Office of Air and
Radiation, U.S. Environmental Protection Agency, Research Triangle Park, NC
Mr. David Orlin—Air and Radiation Law Office, Office of General Counsel, U.S.
Environmental Protection Agency, Washington, DC
Dr. Jennifer Peel—Department of Environmental and Radiological Health Sciences,
Colorado School of Public Health, Colorado State University, Fort Collins, CO
Dr. Edward Postlethwait—Department of Environmental Health Sciences, Ryals School for
Public Health, University of Alabama at Birmingham, Birmingham, AL
Dr. Havala Pye—National Exposure Research Laboratory, U.S. Environmental Protection
Agency, Research Triangle Park, NC
Mr. Venkatesh Rao—Office of Air Quality Planning and Standards, Office of Air and
Radiation, U.S. Environmental Protection Agency, Research Triangle Park, NC
Ms. Elise Richman—Office of Children's Health Protection, U.S. Environmental Protection
Agency, Washington, DC
Dr. Mary Ross—National Center for Environmental Assessment, Office of Research and
Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Reeder Sams II—National Center for Environmental Assessment-RTP Division, Office
of Research and Development, U.S. Environmental Protection Agency, Research
Triangle Park, NC
Dr. Stephanie Sarnat—Department of Environmental Health, Rollins School of Public
Health, Emory University, Atlanta, GA
Dr. Matthew Strickland—Department of Environmental Health, Rollins School of Public
Health, Emory University, Atlanta, GA
Dr. John Vandenberg—National Center for Environmental Assessment-RTP Division,
Office of Research and Development, U.S. Environmental Protection Agency, Research
Triangle Park, NC
Ms. Debra Walsh—National Center for Environmental Assessment-RTP Division, Office of
Research and Development, U.S. Environmental Protection Agency, Research Triangle
Park, NC
Mr. Nealson Watkins—Office of Air Quality Planning and Standards, Office of Air and
Radiation, U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Margaret Zawacki—Office of Transportation and Air Quality, Office of Air and
Radiation, U.S. Environmental Protection Agency, Ann Arbor, MI
January 2015 xx DRAFT: Do Not Cite or Quote
-------
CLEAN AIR SCIENTIFIC ADVISORY COMMITTEE
OXIDES OF NITROGEN NAAQS REVIEW PANEL
Chair of the Oxides of Nitrogen Review Panel
Dr. H. Christopher Frey*, North Carolina State University, Raleigh, NC
Oxides of Nitrogen Review Panel Members
Mr. George A. Allen**, Northeast States for Coordinated Air Use Management
(NESCAUM), Boston, MA
Dr. Matthew Campen, University of New Mexico, Albuquerque, NM
Dr. Ronald Cohen, University of California, Berkeley, Berkeley, CA
Dr. Douglas Dockery, Harvard University, Boston, MA
Dr. Philip Fine, South Coast Air Quality Management District, Diamond Bar, CA
Dr. Panos Georgopoulos, University of Medicine and Dentistry of New Jersey—Robert
Wood Johnson Medical School, Piscataway, NJ
Dr. Jack Harkema**, Michigan State University, East Lansing, MI
Dr. Michael Jerrett, University of California, Berkeley, Berkeley, CA
Dr. Joel Kaufman, University of Washington, Seattle, WA
Dr. Michael T. Kleinman, University of California, Irvine, Irvine, CA
Dr. Timothy V. Larson, University of Washington, Seattle, WA
Dr. Jeremy Sarnat, Emory University, Atlanta, GA
Dr. Richard Schlesinger, Pace University, New York, NY
Dr. Elizabeth A. (Lianne) Sheppard, University of Washington, Seattle, WA
Dr. Helen Suh**, Northeastern University, Boston, MA
Dr. Ronald Wyzga**, Electric Power Research Institute, Palo Alto, CA
Dr. Junfeng (Jim) Zhang, Duke University, Durham, NC
* Chair of the statutory Clean Air Scientific Advisory Committee (CASAC) appointed by the EPA
Administrator
** Members of the statutory Clean Air Scientific Advisory Committee (CASAC) appointed by the EPA
Administrator
Science Advisory Board Staff
Mr. Aaron Yeow, Designated Federal Officer, U.S. Environmental Protection Agency,
Science Advisory Board (1400R), 1200 Pennsylvania Avenue, NW, Washington, D.C.
20460-0001, Phone: 202-564-2050, Fax: 202-565-2098, (yeow.aaron@epa.gov)
January 2015 xxi DRAFT: Do Not Cite or Quote
-------
ACRONYMS AND ABBREVIATIONS
Acronym/Abbreviation
a
a-ATD
A4
AADT
ABI
Abs
ABTS*-
ABTS2~
ACS
ADRB2
AERMOD
AHR
AHSMOG
AIRES
ARIES
AK
AKR/J
AL
ALKP
ALRI
a.m.
AM
AMF
ANN
ANPR
APEX
Meaning
alpha, exposure factor
alpha 1 -antitrypsin deficiency
not classifiable for humans or
animals
annual average daily traffic
ankle brachial index
absorbance coefficient
2,2'-azino-bis
(3 - ethylbenzothiazoline - 6-
sulfonic acid) radical
2,2'-azino-bis
(3-ethylbenzothiazoline-6-
sulphonic acid)
American Cancer Society
beta-2-adrenergic receptor
American Meteorological
Society/Environmental
Protection Agency Regulatory
Model
airway hyperresponsiveness
California Seventh-Day
Adventists cohort
air exchange rate
Aerosol Research Inhalation
Epidemiology Study
Atlanta Aerosol Research
Inhalation Epidemiology Study
Alaska
mice strain with short life-span;
often used as model for aging
Alabama; alpine
alkaline phosphatase
acute lower respiratory infection
ante meridiem (before noon)
alveolar macrophages
air mass factor
artificial neural networks
advanced notice of public
rulemaking
Air Pollution Exposure model
Acronym/Abbreviation Meaning
APHEA
APHENA
AQCD
AQI
AQM
AQS
AR
ARIMA
AT
ATS
AUSTAL2000
avg
AW
AZ
P
Po
Pi
Pz
BAL
BALF
BAMSE
BC
BD
BEIS
BHPN
BIR
Air Pollution and Health: A
European Approach study
Air Pollution and Health: A
European and North American
Approach study
apolipoprotein E knockout
air quality criteria document
air quality index
air quality model
air quality system
Arkansas; airway responsiveness
autoregressive integrated moving
average
atascadero
American Thoracic Society
Ausbreitungsmodell gemaB der
Technischen Anleitung zur
Reinhaltung der Luft
average
area wide
Arizona
beta
model intercept
effect estimate for the ambient
exposure
vector of slope related to each
covariate
bronchoalveolar lavage
bronchoalveolar lavage fluid
Children, Allergy, Milieu,
Stockholm, Epidemiology
Survey
black carbon
bronchodilator
Biogenic Emission Inventory
System
N-bis (2-hydroxy-propyl)
nitrosamine
birch
January 2015
xxn
DRAFT: Do Not Cite or Quote
-------
Acronym/ Abbreviation
BL
BMI
BP
BR-
BS
BSA
BTEX
BW
BWHS
C'
C&RT
Ca2+
CA
Ca
Ca,csm
CAA
CAMP
CAP
CAPES
CAPS
CARB
CASAC
CASNET
Cb
CBSA
CBVD
CC16
CDC
Meaning
bronchial lavage
body mass index
blood pressure
bromide
black smoke
body surface area
sum of the VOCs benzene,
toluene, ethybenzene, xylene
body weight
Black Women's Health Study
benzene
degrees Celsius; the product of
microenvironmental
concentration; carbon;
average concentration
normalized by the standard
deviation of concentration
classification and regression tree
calcium
California; cat allergen
ambient NCh concentration
ambient concentration at a
central site monitor
Clean Air Act
Childhood Asthma Management
Program
concentrated ambient particle
China Air Pollution and Health
Effects Study
cavity attenuated phase shift
carachol
Clean Air Scientific Advisory
Committee
Clean Air Status and Trends
Network
NO2 concentration contribution
away from the influence of the
road
Core Based Statistical Area
cerebrovascular disease
club cell protein
Centers for Disease Control and
Prevention
Acronym/Abbreviation
Cfar
CFD
CFR
cGMP
CH4
CHD
CHF
CHS
CHIMERE
C,
CI(s)
cIMT
Cj
CJ-A
CJ-B
Cf
CL/MC
CL/PC
C1NO
C1NO2
CMAQ
Cnear
CO
Coj
CO2
COD
COLD
COMET
Cong.
COPD
C-R
Meaning
farthest concentration
computational fluid dynamics
Code of Federal Regulations
cyclic guano sine monophosphate
methane
coronary heart disease
congestive heart failure
Children's Health Study
regional chemistry transport
model
average NO2 concentration in the
rth microenvironment; substrate
concentrations
confidence interval(s)
carotid intima-media thickness
average NO2 concentration in the
jth microenvironment
Ciudad Juarez-site A
Ciudad Juarez-site B
chloride
chemiluminescence analyzer
with a MoOx catalytic converter
chemiluminescence analyzer
with measurements from a
photolytic converter
nitrosyl chloride
nitryl chloride
Community Multiscale Air
Quality
nearest concentration
carbon monoxide; Colorado
ambient exposure to NO2
outdoor concentration
carbon dioxide
coefficient of divergence
cold-dry air
single cell gel electrophoresis
congress
chronic obstructive pulmonary
disease
concentration-response
(relationship)
January 2015
XXlll
DRAFT: Do Not Cite or Quote
-------
Acronym/Abbreviation
CRDS
CRP
CT
CTM
CIS
Cu
Cv
cv
CVD
Cx
Cys'
D
d
D
B.C. Cir
DBF
DC
DE
DEARS
DEP
df
DHA
DL
DM
DNA
DOAS
DOCs
dPD
DPF
DPPC
Dt/dx
DVT
Meaning
cavity ring down spectroscopy
C-reactive protein
Connecticut
chemical transport models
California Teachers Study
copper
NO2 concentration contribution
from vehicles on a roadway
coefficient of variation
cardiovascular disease
NO2 concentration at a distance
x from a road
cysteine radical
molecular diffusion coefficient
ofNO2
distance
distance in kilometers
Distric of Columbia Circuit
diastolic blood pressure
District of Columbia
Deleware
Detroit Exposure and Aerosol
Research Study
diesel exhaust particles
degrees of freedom
dehydroascorbate
distributed lag
diabetes mellitus
deoxyribonucleic acid
differential optical absorption
spectroscopy
diesel oxidation catalysts
change in provocative dose
diesel particulate filter
dipalmitoyl phosphatidylcholine
change in x with respect to time
as the limit of x approaches zero
deep vein thrombosis
exempli gratia (for example)
Acronym/Abbreviation Meaning
Eanj
EEC
EC
ECG
ECP
ECRHS
ED
EGR
EGU
Eij
ELF
Ena
eNO
eNOS
Eo
Eoj
EP
EPA
EP-A
EP-B
ESCAPE
ESR
ET
ET-1
ETS
EXPOLIS
f
FEF
FEF25-75%
the sum of an individual's
ambient NO2 exposure
indoor exposures from
nonambient sources
expired (exhaled) breath
condensate
elemental carbon
electrocardiographic
eosinophil cationic protein
European Community
Respiratory Health Survey
emergency department
exhaust gas recirculation
electric power generating unit
indoor NO2 exposures in the jth
microenvironment
epithelial lining fluid
the sum of an individual's
nonambient NO2 exposure
exhaled nitric oxide
endothelial nitric oxide synthase
outdoor microenvironmental
NO2 exposures
outdoor NO2 exposures in the jth
micro environment
entire pregnancy
U.S. Environmental Protection
Agency
El Paso-site A
El Paso-site B
European Study of Cohorts for
Air Pollution Effects
erythrocyte sedimentation rate
total personal exposure
vasoconstrictor endothelin-1
environmental tobacco smoke
exposure in polis or cities
female
forced epiratory flow
forced expiratory flow at
25-75% of exhaled volume
January 2015
xxiv
DRAFT: Do Not Cite or Quote
-------
Acronym/Abbreviation Meaning
FEF5o%
FEM
FEV
FEVi
FL
FR
FRM
FVC
T
Y'
g
g/bhp-h
GA
GAM
GASPII
GCLC
GCLM
GD
GEE
GEOS
GINI
GINI SOUTH
GINIplus
GIS
GLM
GLMM
GM-CSF
forced expiratory flow at 50% of
forced vital capacity
federal equivalent method
forced expiratory volume
forced expiratory volume in
1 second
Florida
Federal Register
federal reference method
forced vital capacity
gamma; uptake coefficients
semivariogram
gram
grams per brake horsepower-
hour
Georgia
generalized additive models
Gene and Environmental
Prospective Study in Italy
gene that encodes the catalytic
subunit for the human enzyme
glutamate-cysteine ligase
gene that encodes the regulatory
subunit for the human enzyme
glutamate-cysteine ligase
gestation day
generalized estimating equations
Goddard Earth Observing
System
German Infant Nutritional
Intervention
German Infant Nutritional
Intervention covers the urban
city of Munich, Germany, and its
surrounding areas
(approximately 28,000 km2)
German Infant Nutritional
Intervention plus environmental
and genetic influences
geographic information systems
generalized linear model
generalized linear mixed model
granulocyte macrophage-colony
stimulating factor
Acronym/Abbreviation
GP
GPS
GPx
GS*
GSD
GSH
GSNOR
GSR
GSS
GST
GSTM1
GSTP1
GSTT1
GW
h
H+
H2S04
HC
hCAEC
HC1
HDL
HDM
HERO
HEV
HF
HFE
HFn
HGF
HI
HIST
HMOX
HN02
HN03
HN04
HO-1
Meaning
general practice
global positioning system
glutathione peroxidase
glutathione radical
geometric standard deviation
glutathione
nitrosoglutathione reductase
glutathione reductase
glutathione synthetase
glutathione S-transferase
glutathione S-transferase Mu 1
glutathione S-transferase Pi 1
glutathione S-transferase theta 1
gestational week
hour(s)
hydrogen ion
sulfuric acid
hydrocarbon(s)
human coronary artery
endothelial cell
hydrochloric acid
high-density lipoprotein
house dust mite; house dust mite
allergen
Health and Environmental
Research Online
hold-out evaluation
high frequency; high frequency
component of HRV
human hemochromatosis protein
high frequency domain
normalized for heart rate
hepatocyte growth factor
Hawaii
histamine
heme oxygenase
nitrous acid
nitric acid
peroxynitric acid
heme oxygenase-1
January 2015
XXV
DRAFT: Do Not Cite or Quote
-------
Acronym/ Abbreviation
H02
H02N02
HONO
HOONO
HR
HRV
HS
HSC
hs-CRP
IA
I ACID
i.e.
I.V.
ICAM-1
ICAS
ICD
ICS
ID
IDW
IFN-y
IgE
IGM
IHD
IL
IL-6
IL-8
He
IM
IMSI
IMT6seg
IMTcca
IN
Meaning
hydroperoxyl radical
peroxynitric acid
nitrous acid
pemitrous acid
hazard ratio(s); heart rate
heart rate variability
hemorrhagic stroke
Harvard Six Cities
high sensitivity C-reactive
protein
Iowa
inorganic acid
id est (that is)
intravenously
intercellular adhesion molecule 1
Inner-City Asthma Study
International Classification of
Diseases; implantable
cardioverter defibrillators
inhaled corticosteroids
Idaho
inverse distance weighting
interferon gamma
immunoglobulin E
impaired glucose metabolism
ischemic heart disease
interleukin; Illinois
interleukin-6
interleukin-8
isoleucine
immediately after exposure
Integrated Mobile Source
Indicator
intima-media thickness of the
left and rifht common carotid
arteries, internal carotid arteries,
and carotid bulbs
intima-media thickness of the
common carotid artery
Indiana; isoprene nitrate
Acroni
INDAI
INFj
iNOS
IOM
IQR
IRP
IRR
IS
ISA
IT
IUGR
IVF
j
JE
k
kcal
kg
h
k
km
kPa
KS
KY
L
LA
LAT
LB
LEW
LDH
LE
LETO
LF
LF/HF
LIE
LIF
Meaning
probabilistic model for indoor
pollution exposures
infiltration of outdoor NCh
inducible nitric oxide synthase
Institute of Medicine
interquartile range
Integrated Review Plan
incidence rate ratios
ischemic stroke
Integrated Science Assessment
intratracheal
intrauterine growth restriction
in vitro fertilization
microenvironment
joint model estimate
reaction rate; decay constant
derived from empirical data
kilocalorie(s)
kilogram(s)
second-order rate constants)
decay rate
kilometer(s)
kilopascal(s)
Kansas
Kentucky
liter(s)
Louisiana; Los Angeles; Lake
Arrowhead
L-type amino acid transporter
Long Beach
low birth weight
lactate dehydrogenase
Lake Elsinore
Long-Evans Tokushima
low-frequency component of
HRV
ratio of LF and HF components
ofHRV
Long Island Expressway
laser induced fluorescence
January 2015
xxvi
DRAFT: Do Not Cite or Quote
-------
Acronym/Abbreviation Meaning
LISA
Lifestyle-Related factors on the
Immune System and the
Development of Allergies in
Childhood
Acronym/Abbreviation
METS
MI
LISAplus
LM
LN
LOESS
LOOCV
LOPAP
LOX-1
Lp-PLA2
LRTI
LUR
u
ug/m3
m
M
MA
Ml
M2
M3
M4
MAAS
max
MCP-1
MD
MDA
ME
MESA-Air
MET
METH
Lifestyle-Related factors on the
Immune System and the
Development of Allergies in
Childhood plus the influence of
traffic emissions and genetics
Lompoc
Lancaster
locally weighted scatterplot
smoothing
Leave-one-out cross-validation
long path absorption photometer
lectin-like oxLDL receptor
lipoprotein-associated
phospholipase A2
lower respiratory tract infection
land use regression
mu; micro
micrograms per cubic meter
meter
male
Massachusetts
Month 1
Month 2
Months
Month 4
Manchester Asthma and Allergy
Study
maximum
monocyte chemoattractant
protein-1
Maryland
malondialdehyde
Maine
Multi-Ethnic Study of
Atherosclerosis and Air
Pollution
MET receptor tyro sine kinase
gene
methacholine
min
ML
mL
MLI
MMEF
mmHg
MMP
MMP-3
MMP-7
MMP-9
MN
mo
MO
MOA
mol
MoOx
MPO
mRNA
MS
MT
n
N
N2O3
N204
N2O5
NA
Na+
NAAQS
NAB
NaCl
NADPH
Meaning
metabolic equivalents
myocardial infarction ("heart
attack"); myocardial ischemia;
Michigan
minimum
Mira Loma
milliliter(s)
mean linear intercept
maximum (or maximal)
midexpiratory flow
millimeters of mercury
matrix metalloproteinase
matrix metalloproteinase-3
matrix metalloproteinase-7
matrix metalloproteinase-9
Minnesota
month(s)
Missouri
mode(s) of action
mole
molybdenum oxide
myeloperoxidase
messenger ribonucleic acid
Mississippi
Montana
sample size; total number of
microenvironments that the
individual has encountered
nitrogen; population number
dinitrogen trioxide
dinitrogen tetroxide
dinitrogen pentoxide
not available
sodium ion
National Ambient Air Quality
Standards
North American Background
sodium chloride
reduced nicotinamide adenine
dinucleotide phosphate
January 2015
xxvii
DRAFT: Do Not Cite or Quote
-------
Acronym/ Abbreviation
NAL
NAMS
NAS
NC
NCEA
NCICAS
NCore
ND
NDMA
NE
NEI
NFKB
NH
NH3
(NH4)2SO4
NHAPS
NHS
NJ
NLCS
nm
NM
NMMAPS
NMOR
NO
N02
N03
non-HS
NOS
NOx
NOy
NOz
Meaning
nasal lavage
National Air Monitoring Stations
National Academy of Sciences
North Carolina
National Center for
Environmental Assessment
National Cooperative Inner-City
Asthma Study
National Core network
North Dakota
N-nitrosodimethylamine
Nebraska
National Emissions Inventory
nuclear factor kappa-light-chain-
enhancer of activated B cells
New Hampshire
ammonia
ammonium sulfate
National Human Activity Pattern
Survey
Nurses Health Study
New Jersey
Netherlands Cohort Study on
Diet and Cancer
nanometer
New Mexico
The National Morbidity
Mortality Air Pollution Study
N-nitrosomorpholine
nitric oxide
nitrogen dioxide
nitrite
nitrate
nitrate radical
non-hemhorragic stroke
nitric oxide synthase
the sum of NO and NO2
oxides of nitrogen
reactive oxides of nitrogen (e.g.,
HNO3, HONO, PAN, particulate
nitrates)
Acronym/Abbreviation Meaning
NQO1
NR
NS
NV
NY
OACID
Os
OAQPS
OC
OH
8-OHdG
OK
OLETF
OLM
OMI
OR
OVA
P
P
Pa
PA
PAARC
PAH(s)
PAMS
PAN
PAPA
Pb
PEL
PC
PCA
PCO
PD
NADPH-quinone oxidoreductase
(genotype)
not reported; no quantitative
results reported; near road
not statistically significant
Nevada
New York
organic acid
ozone
Office of Air Quality Planning &
Standards
organic carbon
hydroxide; Ohio
8-hydroxy-29-deoxyguanosine
Oklahoma
Otsuka Long-evans Tokushima
Fatty
ozone limiting method
Ozone Monitoring Instrument
odds ratio(s); Oregon
ovalbumin
probability
Pearson correlation
pascal(s)
policy assessment; Pennsylvania
Air Pollution and Chronic
Respiratory Diseases
polycyclic aromatic
hydrocarbon(s)
Photochemical Monitoring
Stations
peroxyacetyl nitrate; peroxyacl
nitrate
Public Health and Air Pollution
in Asia
lead
planetary boundary layer
provocative concentration
principal component analysis
protein carbonyl
provocative dose
January 2015
XXVlll
DRAFT: Do Not Cite or Quote
-------
Acronym/Abbreviation Meaning
Acronym/Abbreviation Meaning
PE
PEF
PFK
PIAMA
PiZZ
PJ
PK
p.m.
PM
PMio
pulmonary embolism
peak expiratory flow
phosphofructokinase
Prevention and Incidence of
Asthma and Mite Allergy
severe alpha-1 antitrypsin
deficiency
air pollutant penetration
pyruvate kinase
post meridiem (after noon)
particulate matter
In general terms, particulate
matter with a nominal
aerodynamic diameter less than
or equal to 10 |im; 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 (im 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.
PMlO-2.5
PM2.5
PMA
PMN(s)
PNC
PND
In general terms, particulate
matter with a nominal
aerodynamic diameter less than
or equal to 10 |im and greater
than a nominal 2.5 |im; a
measurement of thoracic coarse
particulate matter or the coarse
fraction of PMio. In regulatory
terms, particles with an upper
50% cut-point of 10 |im
aerodynamic diameter and a
lower 50% cut-point of 2.5 |im
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.
In general terms, particulate
matter with a nominal
aerodynamic diameter less than
or equal to 2.5 |im; a
measurement of fine particles. In
regulatory terms, particles with
an upper 50% cut-point of
2.5 |im 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.
phorbol myristate acetate
polymorphonuclear cell(s),
polymorphonuclear leukocyte
particle number concentration
postnatal day
January 2015
xxix
DRAFT: Do Not Cite or Quote
-------
Acronym/Abbreviation
pNN50
pNO
pNOs
PPARy
ppb
ppm
PROtEuS
PTB
PVMRM
Ql
Q2
Q3
Q4
Q5
QC-TILDAS
QT interval
QTc
QTVI
QUIC
R
R2
RAG
RANCH
RBC
RC(=O)
RC(=O)OONO2
REA
Meaning
Proportion of pairs of successive
normal simus intervals exceeds
50 milliseconds divided by the
total number of successive pairs
of normal simus intervals
particulate nitrogen species
particulate nitrate
peroxisome proliferator activated
receptor gamma
parts per billion
parts per million
Prostate Cancer and
Environment Study
preterm birth
plume volume molar ratio
method
1 st quartile or quintile
2nd quartile or quintile
3rd quartile or quintile
4th quartile or quintile
5th quintile
quantum cascade - tunable
infrared laser differential
absorption spectrometer
time between start of Q wave
and end of T wave in ECG
corrected QT interval
QT variable index
Quick Urban and Industrial
Commplex
Pearson correlation coefficient;
Spearman correlation coefficient
Pearson correlation coefficient
square of the correlation
coefficient
ragweed
Road Traffic and Aircraft Noise
Exposure and Children's
Cognition and Health
red blood cells
acyl group
peroxyacylnitrates
Risk and Exposure Assessment
Acronym/ Abbreviation
REGICOR
RH
RI
RIVM
rMSSD
RNS
RONCh
ROS
RR
RSNO
RSV
RV
s
S. Rep
s/L
S/N
SALIA
SA-LUR
SAPALDIA
SAT
SBP
SC
SCR
SD
SDNN
SE
Se
SEARCH
sec
SEI
Se-L
Meaning
Registre Gironi del Cor
relative humidity
Rhode Island
National Air Quality Monitoring
Network of the National Institute
of Public Health and the
Environment
root mean square of successive
differences; a measure of HRV
reactive nitrogen species
organic nitrates
reactive oxygen species
risk ratio(s), relative risk
S-nitrosothiols
respiratory syncytial virus
Riverside
second(s)
Senate Report
seconds per liter
Signal-to-noise ratio
Study on the Influence of Air
Pollution on Lung,
Inflammation, and Aging
source-area land use regression
Swiss Study on Air Pollution
and Lung Disease in Adults
switching attention test
systolic blood pressure
South Carolina
selective catalytic reduction
standard deviation; South
Dakota; San Dimas
standard deviation of all normal-
to-normal intervals, an index of
total HRV
standard error
selenium
Southeast Aerosol Research
Characterization
second(s)
socio-economic index
low selenium
January 2015
XXX
DRAFT: Do Not Cite or Quote
-------
Acronym/Abbreviation
SES
Se-S
Sess.
SF6
SGA
sGaw
SHARP
SHEDS
SHEEP
sIC AM-1
SLAMS
SM
SNP
S02
S04
SOA
SOD
SP-D
SPE
sRaw
SRTT
ST segment
sVCAM-1
TEARS
Tl
T2
T3
TEARS
Meaning
socioeconomic status
supplemented selenium
session
sulfur hexafluoride
small for gestational age
specific airway conductance
Study of Houston Atmospheric
Radical Precursors
Stochastic Human Exposure and
Dose Simulation
Stockholm Heart Epidemiology
Program
soluble intercellular adhesion
molecule-1
state and local air monitoring
stations
Santa Maria
single nucleotide polymorphism
sulfur dioxide
sulfate
secondary organic aerosols
superoxide dismutase
surfactant protein D
single-pollutant model estimate
specific airway resistance
simple reaction time test
segment of the
electrocardiograph between the
end of the S wave and beginning
of the T wave
soluble vascular adhesion
molecule-1
fraction of time spent in a
microenvironment across an
individual's microenvironmental
exposures, time
thiobabituric acid reactive
substances (species)
first trimester
second trimester
third trimester
thiobarbituric acid reactive
substances
Acronym/Abbreviation
TCHS
TEA
Thl7
Th2
TIA
TIM
TIMP-2
tj
TLR
TN
TNT
TNF-a
TSP
TWA
TX
U.S.C.
UCD
UF1
UF2
UFP
UK
U.K.
ULTRA
UP
URI
U.S.
UT
VA
Val
VCAM-1
VEGF
VOC
Meaning
Taiwan Children Health Study
triethanolamine
T helper cell 17
T-derived lymphocyte helper 2
transient ischemic attack
timothy
tissue inhibitor of matrix
metalloproteinase-2
fraction of total time spent in the
jth microenvironment
Toll-like receptor
Tennessee
tumor necrosis factor
tumor necrosis factor alpha
total suspended solids
time-weighted average
Texas
U.S. Code
University of California, Davis
ultrafine particle number
beginning at 3 nanometers
ultrafine particle number
beginning at 15 nanometers
ultrafine particle(s)
universal kriging
United Kingdom
The Exposure and Risk
Assessment for Fine and
Ultrafine Particles in Ambient
Air Study conducted in Europe
Upland
upper respiratory infection
United States of America
Utah
Virginia
valine
vascular adhesion molecule-1
minute volume
vascular endothelial growth
factor
volatile organic compound
January 2015
xxxi
DRAFT: Do Not Cite or Quote
-------
Acronym/Abbreviation Meaning
Acronym/Abbreviation Meaning
VPTB
VT
VT
vWF
WBC
WHI
WHO
WI
WV
WY
X
Y
very preterm birth
tital volume
ventricular tachyarrhythmias;
Vermont
von Willbrand factor
white blood cell
Women's Health Initiative
World Health Organization
Wisconsin
West Virginia
Wyoming
distance from the road
health effect of interest
fraction of time spent indoors
fraction of a day spent in each
indoor microenvironment
yr
Z
Z*
Zn
8
T
fraction of all time spent
outdoors
fraction of a day spent in each
outdoor microenvironment
year(s)
covariate vector; the measured
concentration; standard normal
deviate
the true concentration
zinc
random error
half-time
radical species
January 2015
xxxu
DRAFT: Do Not Cite or Quote
-------
PREAMBLE
1. Process of Integrated Science Assessment Development
1 This Preamble outlines the general process for developing an Integrated Science
2 Assessment (ISA) including the framework for evaluating weight of evidence and
3 drawing scientific conclusions and causal judgments. The ISA provides a concise review,
4 synthesis, and evaluation of the most policy-relevant science to serve as a scientific
5 foundation for the review of the National Ambient Air Quality Standards (NAAQS).: As
6 described in the Preface, the NAAQS are established based on consideration of the air
7 quality criteria (represented by the ISA) for the pollutants identified by the Administrator
8 using Section 108 of the Clean Air Act (CAA). The pollutants currently identified are
9 carbon monoxide (CO), lead (Pb), oxides of nitrogen, photochemical oxidants, particulate
10 matter (PM), and sulfur oxides (CAA. 1990a. b). Figure I depicts the general NAAQS
11 review process. Information for individual NAAQS reviews is available online.2
12 In the process for each review, the development of the ISA is preceded by the release of
13 an Integrated Review Plan (IRP) that discusses the planned scope of the NAAQS review;
14 the planned approaches for developing the key assessment documents [e.g., ISA, Risk
15 and Exposure Assessment (if warranted), Policy Assessment]; and the schedule for
16 release and review of the documents and subsequent rulemaking notices. The key
17 policy-relevant questions included in the IRP serve to clarify and focus the NAAQS
18 review on the critical scientific and policy issues, including addressing uncertainties
19 discussed during the previous review and newly emerging literature. The IRP is informed
20 by a U.S. Environmental Protection Agency (EPA)-hosted public "kick-off workshop"
21 that seeks input on the current state of the science and engages stakeholders and experts
22 in discussion of the policy-relevant questions that will frame the review.
1 The general process for NAAQS reviews is described at http://www.epa.gov/ttn/naaqs/review.html.
2 Information for individual NAAQS reviews is available at www.epa.gov/ttn/naaqs.
January 2015 xxxiii DRAFT: Do Not Cite or Quote
-------
Workshop on
science-policy issues
EPA
proposed
decisions on
Inte
poicy
•grated Review Plan (IRP): timeline and key
nicy-relevant issues and scientific questions
Integrated Science Assessment (ISA): evaluation and
synthesis of most policy-relevanl studies
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
Figure I
Public hearings
and comments
on proposal
Agency decision
making and draft
final notice
Interacjency
review
Clean Air Scientific
Advisory Committee
(CASAC) review
Public comment
Schematic of the key steps in the process of the review of
National Ambient Air Quality Standards.
i
2
3
4
5
6
7
10
11
12
13
This Preamble is a general discussion of the basic steps and criteria used in developing an
ISA. Details and considerations specific to an individual ISA are included in the Preface
and introductory section for that assessment. The general process for ISA development is
illustrated in Figure II An initial step (not shown) 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 welfare effects of the
pollutant, or from atmospheric and exposure sciences fields.
The fundamental process for developing an ISA includes:
• literature searches;
• study selection;
• evaluation of individual study quality;
• evaluation, synthesis, and integration of the evidence; and
• development of scientific conclusions and causal determinations.
January 2015
xxxiv
DRAFT: Do Not Cite or Quote
-------
Literature Search and
Study Selection
(See Figure III)
Evaluation of Individual Study Quality
After study selection, the quality of individual studies is evaluated by EPA or outside experts in the fields of
atmospheric science, exposure assessment, dosimetry, animal toxicology, controlled human exposure studies,
epidemiology, ecology, and other welfare effects, considering the design, methods, conduct, and documentation of
each study. Strengths and limitations of individual studies that may affect the interpretation of the study are
considered.
Develop Initial Sections
Review and summarize new study results as well
as findings and conclusions from previous
assessments by category of outcome/effect and
by discipline, e.g., toxicological studies of lung
function.
Peer Input Consultation
Review of initial draft materials by scientists
from both outside and within EPA in public
meeting or public teleconference.
Evaluation, Synthesis, and Integration of Evidence
Integrate evidence from scientific disciplines -for example, toxicological, controlled human exposure, and
epidemiologic study findings fora particular health outcome. Evaluate evidence for related groups of endpoints or
outcomes to draw conclusions regarding health or welfare effect categories, integrating health or welfare effects
evidence with information on mode of action and exposure assessment.
Development of Scientific Conclusions and Causal Determinations
Characterize weight of evidence and develop judgments regarding causality for health or welfare effect categories.
Develop conclusions regarding concentration- or dose-response relationships, potentially at-nsk populations,
lifestages, or ecosystems.
Draft Integrated Science Assessment
Evaluation and integration of newly published studies
after each draft.
Final Integrated Science Assessment
Clean Air Scientific Advisory Committee
Independent review of draft documents for scientific
quality and sound implementation of causal
framework; anticipated review of two drafts of ISA in
public meetings.
Public Comments
Comments on draft ISA solicited by EPA
Figure II
Characterization of the general process of Integrated Science
Assessment (ISA) development.
January 2015
XXXV
DRAFT: Do Not Cite or Quote
-------
1 In developing an ISA, EPA reviews and summarizes the evidence from studies on
2 atmospheric sciences, human exposure, animal toxicology, controlled human exposure,
3 epidemiology, and/or ecology and other welfare1 effects. In the process of developing the
4 first draft ISA, EPA may convene a peer input meeting in which the scientific content of
5 preliminary draft materials is reviewed to ensure that the ISA is up-to-date and is focused
6 on the most policy-relevant findings, and to assist EPA with integration of evidence
7 within and across disciplines.
8 EPA integrates the evidence from across scientific disciplines or study types and
9 characterizes the weight of evidence for relationships between the pollutant and various
10 outcomes. The integration of evidence on human health or welfare effects involves
11 collaboration between scientists from various disciplines. As an example, an evaluation
12 of human health effects evidence would generally include the integration of the results
13 from epidemiologic, controlled human exposure, and toxicological studies, consideration
14 of exposure assessment, and application of the causal framework (described below) to
15 draw conclusions.
16 Integration of results on human health or welfare effects that are logically or
17 mechanistically connected (e.g., respiratory symptoms; asthma exacerbations) informs
18 judgments of causality on the broader health effect category (e.g., effects on the
19 respiratory system). Using the causal framework described in this Preamble. EPA
20 scientists consider aspects such as strength, consistency, coherence, and biological
21 plausibility of the evidence and develop causality determinations on the nature of the
22 relationships. Causality determinations often entail an iterative process of review and
23 evaluation of the evidence. Two drafts of the ISA are typically released for review by the
24 Clean Air Scientific Advisory Committee (CASAC) and the public, and comments
25 received on the characterization of the science as well as the implementation of the causal
26 framework are carefully considered in revising the draft ISA and completing the final
27 ISA.
2. Literature Search
28 In addition to the call for information in the Federal Register referenced above, EPA
29 maintains an ongoing literature search process to identify relevant scientific studies
30 published since the last ISA for a given criteria pollutant. Search strategies are designed a
31 priori for pollutants and scientific disciplines and iteratively modified to optimize
1 Welfare effects as defined in 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."
January 2015 xxxvi DRAFT: Do Not Cite or Quote
-------
1 identification of pertinent publications. In addition, papers are identified for inclusion in
2 several ways: specialized searches on specific topics, identification of new publications
3 by relational searches conducted using citations from previous assessments, review of
4 tables of contents for journals in which relevant papers may be published, identification
5 of relevant literature by expert scientists, review of citations in previous assessments, and
6 recommendations by the public and CASAC during the call for information and external
7 review processes. References identified through the multipronged search strategy are
8 screened by title and abstract. Those references that are potentially relevant after reading
9 the title are "considered" for inclusion in the ISA and are added to the Health and
10 Environmental Research Online (HERO) database developed by EPA
11 (http://hero.epa.gov/).' Additional review steps (described in Section 3 below) precede a
12 decision to include a study in the ISA. The references cited in the ISA include a hyperlink
13 to the HERO database. This literature search and study selection process is depicted in
14 Figure III.
15 Studies and reports that have undergone scientific peer review and been published (or
16 accepted for publication) are considered for inclusion in the ISA. This includes only
17 studies that have been ethically conducted (e.g., approval by an Institutional Review
18 Board or Institutional Animal Care and Use Committee). All relevant epidemiologic,
19 controlled human exposure, toxicological, ecological, and welfare effects studies
20 published since the last review are considered, including those related to
21 exposure-response relationships, mode(s) of action (MOA), and populations and
22 lifestages at increased risk of air pollution-related health effects. Studies and data
23 analyses on atmospheric chemistry, air quality and emissions, environmental fate and
24 transport, dosimetry, toxicokinetics, and exposure are also considered for inclusion in the
25 document. References considered for inclusion in a specific ISA can be found using the
26 HERO website (http://hero.epa.gov).
1 The list of considered references and bibliographic information is accessible to the public through HERO.
January 2015 xxxvii DRAFT: Do Not Cite or Quote
-------
Figure
Caltfor
Information
and
Literature
Search
Citations from
Past Assessments
Peer Review
Recommendations
Illustration of processes for literature search and study selection
used for development of Integrated Science Assessments.
i
2
3
4
5
6
7
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 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.
10
11
12
13
14
15
Study Selection
References considered for inclusion in the ISA undergo abstract and full-text review to
determine whether they will be included in the ISA. The selection process is based on the
extent to which the study is informative, pertinent, and policy relevant. Informative,
pertinent, and policy-relevant 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.
Emphasis is placed on studies that examine effects associated with pollutant
January 2015
xxxvin
DRAFT: Do Not Cite or Quote
-------
1 concentrations and exposure conditions relevant to current human population and
2 ecosystem exposures, and particularly those pertaining to concentrations currently found
3 in ambient air. Other studies are included if they contain unique data, such as a
4 previously unreported effect or MOA for an observed effect, or examine multiple
5 concentrations to elucidate exposure-response relationships.
4. Evaluation of Individual Study Quality
6 After selecting studies for inclusion, the individual study quality is evaluated by
7 considering the design, methods, conduct, and documentation of each study, but not the
8 study results. This uniform approach aims to consider the strengths, limitations, and
9 possible roles of chance, confounding, and other biases that may affect the interpretation
10 of individual studies and the strength of inference from their results.
11 These criteria provide standards for evaluating various studies and for focusing on the
12 policy-relevant studies in assessing the body of human health, ecological, and other
13 welfare effects evidence. Particular aspects or the absence of some of these features in a
14 study do not necessarily define a less informative study or exclude a study from
15 consideration in an ISA. As stated initially, the intent of the ISA is to provide a concise
16 review, synthesis, and evaluation of the most policy-relevant science to serve as a
17 scientific foundation for the review of the NAAQS, not extensive summaries of all
18 human health, ecological, and other welfare effects studies for a pollutant. Of most
19 relevance for inclusion of studies is whether they provide useful qualitative or
20 quantitative information on exposure-response relationships for effects associated with
21 pollutant exposures at doses or concentrations relevant to ambient conditions that can
22 inform decisions on whether to retain or revise the standards.
23 In general, in assessing the scientific quality of studies on health and welfare effects, the
24 following considerations have been taken into account.
25 • Were study design, study groups, methods, data, and results clearly presented
26 in relation to the study objectives to allow for study evaluation? Were
27 limitations and any underlying assumptions of the design and other aspects
28 stated?
29 • Were the ecosystems, study site(s), study populations, subjects, or organism
30 models adequately selected, and are they sufficiently well defined to allow for
31 meaningful comparisons between study or exposure groups?
32 • Are the air quality data, exposure, or dose metrics of adequate quality and
33 sufficiently representative of information regarding ambient conditions?
34 • Are the health, ecological, or other welfare effect measurements meaningful,
35 valid, and reliable?
January 2015 xxxix DRAFT: Do Not Cite or Quote
-------
1 • Were likely covariates or modifying factors adequately controlled or taken
2 into account in the study design and statistical analysis?
3 "Do the analytical methods provide adequate sensitivity and precision to
4 support conclusions?
5 • Were the statistical analyses appropriate, properly performed, and properly
6 interpreted?
7 Additional considerations specific to particular disciplines are discussed below.
a. Atmospheric Science and Exposure Assessment
8 Atmospheric science and exposure assessment studies that are considered for inclusion in
9 the ISA focus on measurement of, behavior of, and exposure to ambient air pollution
10 using quality-assured field, experimental, and/or modeling techniques. The most
11 informative measurement-based studies will include detailed descriptive statistics for
12 measurements taken at varying spatial and temporal scales. These studies will also
13 include a clear and comprehensive description of measurement techniques and
14 quality-control procedures used. Quality-control metrics (e.g., method detection limits)
15 and quantitative relationships between and within pollutant measurements (e.g.,
16 regression slopes, intercepts, and fit statistics) should be provided when appropriate.
17 Measurements that include contrasting conditions for various time periods (e.g.,
18 weekday/weekend, season), populations, regions, and categories (e.g., urban/rural) are
19 particularly useful. The most informative modeling-based studies will incorporate
20 appropriate chemistry, transport, dispersion, and/or exposure modeling techniques with a
21 clear and comprehensive description of model evaluation procedures, metrics, and
22 technique strengths and limitations. The ISA also may include analyses of data pertinent
23 to characterizing air quality or exposure such as emissions sources and ambient air
24 pollutant concentrations. Sources of monitoring and modeling data should be clearly
25 referenced and described to foster transparency and reproducibility of any analyses. In
26 general, atmospheric science studies and data analyses focusing on locations pertinent to
27 the U.S. will have maximum value in informing review of the NAAQS.
28 Exposure measurement error, which refers to inaccuracies in the characterization of the
29 exposures of study participants, can be an important contributor to uncertainty in air
30 pollution epidemiologic study results. Exposure measurement error can influence
31 observed epidemiologic associations between ambient pollutant concentrations and health
32 outcomes by biasing effect estimates toward or away from the null and widening
33 confidence intervals around those estimates (Zeger et al.. 2000). Factors that could
34 influence exposure estimates include, but are not limited to: choice of exposure metric,
35 spatial variability of the pollutant concentration, nonambient sources of exposure,
January 2015 xl DRAFT: Do Not Cite or Quote
-------
1 topography of the natural and built environment, meteorology, instrument errors,
2 time-activity patterns, and differential infiltration of air pollutants into indoor
3 environments. The influence of these factors on effect estimates also depends on
4 epidemiologic study design. For example, when longitudinal studies depend on spatial
5 contrasts in exposure estimates, it is important that the exposure estimates correspond in
6 space to the population of interest. Likewise for time-series studies, the temporal
7 variability of the exposure estimate must correspond temporally to the true exposures of
8 the study population.
b. Epidemiology
9 In evaluating individual study quality for inference about health effects in epidemiologic
10 studies, EPA considers, in addition to the general quality considerations discussed
11 previously, whether a given study: (1) presents information on associations with short- or
12 long-term pollutant exposures at or near conditions relevant to ambient exposures;
13 (2) addresses potential confounding, particularly by other pollutants; (3) assesses
14 potential effect modifiers; (4) evaluates health endpoints and populations not previously
15 extensively researched; and (5) evaluates important methodological issues related to
16 interpretation of the health evidence (e.g., lag or time period between exposure and
17 effects, model specifications, thresholds).
18 In the evaluation of epidemiologic evidence, one important consideration is potential
19 confounding. Confounding is "... a confusion of effects. Specifically, the apparent effect
20 of the exposure of interest is distorted because the effect of an extraneous factor is
21 mistaken for or mixed with the actual exposure effect (which may be null)" (Rothman
22 and Greenland. 1998). A confounder is associated with both the exposure and the effect;
23 for example, confounding can occur between correlated pollutants that are associated
24 with the same effect. One approach to remove spurious associations due to possible
25 confounders is to control for characteristics that may differ between exposed and
26 unexposed persons; this is frequently termed "adjustment." Scientific judgment is needed
27 to evaluate likely sources and extent of confounding, together with consideration of how
28 well the existing constellation of study designs, results, and analyses address the potential
29 for erroneous inferences.
30 Several statistical methods are available to detect and control for potential confounders;
31 however, none of these methods is completely satisfactory. Multivariable regression
32 models constitute one tool for estimating the association between exposure and outcome
33 after adjusting for characteristics of participants that might confound the results. As much
34 of the uncertainty in inferring causality may be due to potential confounding by
January 2015 xli DRAFT: Do Not Cite or Quote
-------
1 copollutants, this topic is of particular importance when evaluating individual studies.
2 The use of copollutant regression models has been the prevailing approach for controlling
3 potential confounding by copollutants in air pollution health effects studies. Trying to
4 determine whether an individual pollutant is independently associated with the health
5 outcome of interest from copollutant regression models is made difficult by the
6 possibility that one or more air pollutants is acting as a surrogate for an unmeasured or
7 poorly measured pollutant or for a particular mixture of pollutants. In addition, pollutants
8 may independently exert effects on the same system; for example, several pollutants may
9 be associated with a respiratory effect through either the same or different MOAs.
10 Despite these limitations, the use of copollutant models is still the prevailing approach
11 employed in most air pollution epidemiologic studies and provides some insight into the
12 potential for confounding or interaction among pollutants.
13 Confidence that unmeasured confounders are not producing the findings is increased
14 when multiple studies are conducted in various settings using different subjects or
15 exposures, each of which might eliminate another source of confounding from
16 consideration. For example, multicity studies can provide insight on potential
17 confounding through the use of a consistent method to analyze data from across locations
18 with different concentrations of copollutants and other covariates. Intervention studies,
19 because of their quasi-experimental nature, can be particularly useful in characterizing
20 causation.
21 Another important consideration in the evaluation of epidemiologic studies is
22 effect-measure modification, which occurs when the effect differs between subgroups or
23 strata; for example, effect estimates that vary by age group or a potential risk factor. As
24 stated by Rothman and Greenland (1998):
25 "Effect-measure modification differs from confounding in several ways.
26 The main difference is that, whereas confounding is a bias that the
27 investigator hopes to prevent or remove from the effect estimate,
28 effect-measure modification is a property of the effect under study ... In
29 epidemiologic analysis one tries to eliminate confounding but one tries to
30 detect and estimate effect-measure modification."
31 When a risk factor is a confounder, it is the true cause of the association observed
32 between the exposure and the outcome; when a risk factor is an effect modifier, it
33 changes the magnitude of the association between the exposure and the outcome in
34 stratified analyses. For example, the presence of a pre-existing disease or indicator of low
35 socioeconomic status (SES) may act as effect modifiers if they are associated with
36 increased risk of effects related to air pollution exposure. It is often possible to stratify the
37 relationship between health outcome and exposure by one or more of these potential
January 2015 xlii DRAFT: Do Not Cite or Quote
-------
1 effect modifiers. For variables that modify the association, effect estimates in each
2 stratum will be different from one another and different from the overall estimate,
3 indicating a different exposure-response relationship may exist in populations represented
4 by these variables.
c. Controlled Human Exposure and Animal
Toxicology
5 Controlled human exposure and animal toxicological studies experimentally evaluate the
6 health effects of administered exposures in human volunteers and animal models under
7 highly controlled laboratory conditions. Controlled human exposure studies are also
8 referred to as human clinical studies. In these experiments, investigators expose subjects
9 to known concentrations of air pollutants under carefully regulated environmental
10 conditions and activity levels. In addition to the general quality considerations discussed
11 previously, evaluation of controlled human exposure and animal toxicological studies
12 includes assessing the design and methodology of each study with focus on
13 (1) characterization of the intake dose, dosing regimen, and exposure route;
14 (2) characterization of the pollutant(s); (3) sample size and statistical power to detect
15 differences; and (4) control of other variables that could influence the occurrence of
16 effects. The evaluation of study design generally includes consideration of factors that
17 minimize bias in results such as randomization, blinding, and allocation concealment of
18 study subjects, investigators, and research staff, and unexplained loss of animals or
19 withdrawal/exclusion of subjects. Additionally, studies must include appropriate control
20 groups and exposures to allow for accurate interpretation of results relative to exposure.
21 Emphasis is placed on studies that address concentration-dependent responses or
22 time-course of responses and studies that investigate potentially at-risk populations (e.g.,
23 age or pre-existing disease).
24 Controlled human exposure or animal toxicological studies that approximate expected
25 human exposures in terms of concentration, duration, and route of exposure are of
26 particular interest. Relevant pollutant exposures are considered to be those generally
27 within two orders of magnitude of recent ambient concentrations. This range in relevant
28 exposures is to account for differences in dosimetry, toxicokinetics, and biological
29 sensitivity of various species, strains, or potentially at-risk populations. Studies using
30 higher concentration exposures or doses will be considered to the extent that they provide
31 information relevant to understanding MOA or mechanisms, interspecies variation, or
32 at-risk human populations. In vitro studies may provide mechanistic insight for effects
33 examined in vivo or in epidemiologic studies.
January 2015 xliii DRAFT: Do Not Cite or Quote
-------
d. Ecological and Other Welfare Effects
1 Ecological effects considered in the IS As typically include several of the topics given as
2 examples by the Clean Air Act definition of welfare including soils, water, vegetation,
3 animals, and wildlife. Additional topic areas, often referred to as "other welfare effects"
4 that may be evaluated by an ISA include visibility, weather, and climate, as well as
5 materials damage, economic values, and impacts to personal comfort and well-being. In
6 evaluating studies that consider welfare effects, in addition to assessing the general
7 quality considerations discussed previously, emphasis is placed on studies that evaluate
8 effects at or near concentrations of the ambient air pollutants. Studies conducted in any
9 country that contribute significantly to the general understanding of air pollutant effects
10 may be evaluated for relevancy to U.S. air quality considerations and inclusion in the
11 ISA.
12 For ecological effects, studies at higher concentrations are used to evaluate ecological
13 effects only when they are part of a range of concentrations that also included more
14 typical values, or when they inform understanding of MOAs and illustrate the wide range
15 of sensitivity to air pollutants across taxa or across biomes and ecoregions. In evaluating
16 quantitative exposure-response relationships, emphasis is placed on findings from studies
17 conducted in the U.S. and Canada as having ecological and climatic conditions most
18 relevant for review of the NAAQS. The type of experimental approach used in the study
19 (e.g., controlled laboratory exposure, growth chamber, open-top chamber, mesocosm,
20 gradient, field study, etc.) is also evaluated when considering the applicability of the
21 results to the review of criteria air pollutant effects.
22 In evaluating studies on climate and visibility, emphasis is placed on studies that use
23 well-established measurement and modeling techniques, especially those that report
24 uncertainty or compare results from an ensemble of techniques. Novel methods may also
25 be informative in addressing knowledge gaps not well characterized by existing
26 techniques. Relevant climate studies include those evaluating direct and indirect climate
27 impacts of criteria air pollutants at a global scale, while for visibility, studies conducted
28 in the U.S. and Canada provide information more applicable for review of the NAAQS.
29 In both cases, studies that evaluate effects by source sector or region, such as regional
30 climate modeling studies, are particularly informative. Studies that report impacts of
31 multiple PM components and, for climate, multiple criteria pollutants are useful in
32 evaluating interactions and the relative contributions of atmospheric constituents. For
33 example, ozone (Os) and hence its climate forcing depends on atmospheric chemistry
34 involving CO and NOx (the sum of nitric oxide and nitrogen dioxide). Visibility
35 preference and valuation studies that explicitly separate preferences for visibility from
January 2015 xliv DRAFT: Do Not Cite or Quote
-------
1 concerns about health risks of air pollution are particularly relevant in considering a
2 welfare-based secondary NAAQS.
5. Evaluation, Synthesis, and Integration of Evidence across
Disciplines and Development of Scientific Conclusions and
Causal Determinations
3 EPA has developed an approach for integrating the scientific evidence gained from the
4 array of studies discussed above in order to draw conclusions regarding the causal nature
5 of ambient air pollutant-related health or welfare effects. Evidence from all disciplines is
6 integrated to evaluate consistency and inconsistency in the pattern of effects as well as
7 strengths and limitations of the evidence across disciplines. Part of this approach includes
8 a framework for making determinations with regard to the existence of a causal
9 relationship between the pollutant in ambient air and health or welfare effects (described
10 in Section 5b). This framework establishes uniform language concerning causality and
11 brings specificity to the conclusions.
a. Evaluation, Synthesis, and Integration of Evidence
across Disciplines
12 The ISA focuses on evaluation of the findings from the body of evidence across
13 disciplines, drawing upon the results of all studies judged of adequate quality and
14 relevance per the criteria described previously. Evidence from across scientific
15 disciplines for related and similar health or welfare effects is evaluated, synthesized, and
16 integrated to develop conclusions and causality determinations. This includes the
17 evaluation of strengths and weaknesses in the overall collection of studies across
18 disciplines. Confidence in the collective body of evidence is based on evaluation of study
19 design and quality. The roles of different types of evidence in drawing the conclusions
20 varies by pollutant or assessment, as does the availability of different types of evidence
21 for causality determination. Consideration of human health effects are informed by
22 controlled human exposure, epidemiologic, and toxicological studies. Evidence on
23 ecological and other welfare effects may be drawn from a variety of experimental
24 approaches (e.g., greenhouse, laboratory, field) and numerous disciplines
25 (e-g-, community ecology, biogeochemistry, paleontological/historical reconstructions).
26 Other evidence including mechanistic, toxicokinetics, and exposure assessment may be
27 highlighted if it is relevant to the evaluation of health and welfare effects and is of
28 sufficient importance to affect the overall evaluation. Causal inference can be
29 strengthened by the integration of evidence across disciplines. A weak inference from
January 2015 xlv DRAFT: Do Not Cite or Quote
-------
1 one line of evidence can be addressed by other lines of evidence, and coherence of these
2 lines of evidence can add support to a cause-effect interpretation of the association.
3 Interpretation of the body of epidemiologic associations as evidence of causal
4 relationships involves assessment of the full evidence base with regard to elimination of
5 alternative explanations for the association.
6 Evaluation and integration of evidence must also include consideration of uncertainty,
7 which is inherent in scientific findings. "Uncertainty" can be defined as a deficit of
8 knowledge to describe the existing state or future outcome with accuracy and precision
9 (e-g-, the lack of knowledge about the correct value for a specific measure or estimate).
10 Uncertainty analysis may be qualitative or quantitative in nature. In many cases, the
11 analysis is qualitative and can include professional judgment or inferences based on
12 analogy with similar situations. Quantitative uncertainty analysis may include use of
13 simple measures (e.g., ranges) and analytical techniques. Quantitative uncertainty
14 analysis might progress to more complex measures and techniques, if needed for decision
15 support. Various approaches to evaluating uncertainty include classical statistical
16 methods, sensitivity analysis, or probabilistic uncertainty analysis, in order of increasing
17 complexity and data requirements. However, data may not be available for all aspects of
18 an assessment, and those data that are available may be of questionable or unknown
19 quality. Ultimately, the assessment is based on a number of assumptions with varying
20 degrees of uncertainty. While the ISA may include quantitative analysis approaches such
21 as meta-regression in some situations, generally qualitative evaluation of uncertainties is
22 used in assessing the evidence from across studies.
23 Publication bias is another source of uncertainty that can impact the magnitude of health
24 risk estimates. It is well understood that studies reporting non-null findings are more
25 likely to be published than reports of null findings. Publication bias can result in
26 overestimation of effect estimate sizes (loannidis, 2008). For example, effect estimates
27 from single-city epidemiologic studies have been found to be generally larger than those
28 from multicity studies. This is an indication of publication bias because null or negative
29 single-city results may be reported in multicity analyses but might not be published
30 independently (Bell etal., 2005).
31 Potential strengths and limitations of the body of studies can vary across disciplines and
32 are evaluated during data synthesis and integration. Direct evidence of a relationship
33 between pollutant exposures and human health effects may come from controlled human
34 exposure studies. These studies can also provide important information on the biological
35 plausibility of associations observed in epidemiologic studies and inform determinations
36 of factors that may increase or decrease the risk of health effects in certain populations. In
37 some instances, controlled human exposure studies can be used to characterize
January 2015 xlvi DRAFT: Do Not Cite or Quote
-------
1 concentration-response relationships at pollutant concentrations relevant to ambient
2 conditions. Controlled human exposures are typically conducted using a randomized
3 crossover design, with subjects exposed both to the pollutant and a clean air control. In
4 this way, subjects serve as their own experimental controls, effectively limiting the
5 variance associated with potential inter-individual confounders. Limitations that must be
6 considered in evaluating controlled human study findings include the generally small
7 sample size and short exposure time used in experimental studies, and that severe health
8 outcomes are not assessed. By experimental design, controlled human exposure studies
9 are structured to evaluate physiological or biomolecular outcomes in response to
10 exposure to a specific air pollutant and/or combination of pollutants. In addition, the
11 study design generally precludes inclusion of subjects with serious health conditions, and
12 therefore the results often cannot be generalized to an entire population, which includes
13 populations or lifestages at potentially increased risk of air pollutant-induced effects.
14 Although some controlled human exposure studies have included health-compromised
15 individuals such as those with mild or moderate respiratory or cardiovascular disease,
16 these individuals may also be relatively healthy and may not represent the most sensitive
17 individuals in the population. Thus, observed effects in these studies may underestimate
18 the response in certain populations. In addition, the study design is limited to exposures
19 and endpoints that are not expected to result in severe health outcomes.
20 Epidemiologic studies provide important information on the associations between health
21 effects and exposure of human populations to ambient air pollution. In epidemiologic or
22 observational studies of humans, the investigator does not control exposures or intervene
23 with the study population. Broadly, observational studies can describe associations
24 between exposures and effects. These studies fall into several categories; for
25 example, cross-sectional, prospective cohort, time-series, and panel studies. Each type of
26 study has various strengths and limitations. Cross-sectional ecologic studies use health
27 outcome, exposure, and covariate data available at the community level (e.g., annual
28 mortality rates and pollutant concentrations), but do not have individual-level data.
29 Prospective cohort studies include some data collected at the individual level, typically
30 health outcome data, and in some cases, individual-level data on exposure and covariates
31 are collected. Time-series and case-crossover studies are often used to evaluate the
32 relationship between day-to-day changes in air pollution exposures and a specific health
33 outcome at the population-level (i.e., mortality, hospital admissions or emergency
34 department visits). Panel studies include repeated measurements of health outcomes, such
35 as respiratory symptoms or heart rate variability, at the individual level. "Natural
36 experiments" offer the opportunity to investigate changes in health related to a change in
37 exposure, such as closure of a pollution source.
January 2015 xlvii DRAFT: Do Not Cite or Quote
-------
1 When evaluating the collective body of epidemiologic studies, consideration of many
2 study design factors and limitations must be taken into account to properly inform their
3 interpretation. One key consideration is the evaluation of the potential independent
4 contribution of the pollutant to a health outcome when it is a component of a complex air
5 pollutant mixture. Reported effect estimates in epidemiologic studies may reflect
6 (1) independent effects on health outcomes; (2) effects of the pollutant acting as an
7 indicator of a copollutant or a complex ambient air pollution mixture; and (3) effects
8 resulting from interactions between that pollutant and copollutants.
9 The third main type of health effects evidence, animal toxicological studies, provides
10 information on the pollutant's biological action under controlled and monitored exposure
11 circumstances. Taking into account biological differences among species, these studies
12 contribute to our understanding of potential health effects, exposure-response
13 relationships, and MOAs. Further, animal models can inform determinations of factors
14 that may increase or decrease the risk of health effects in certain populations. These
15 studies evaluate the effects of exposures to a variety of pollutants in a highly controlled
16 laboratory setting and allow exploration of toxicological pathways or mechanisms by
17 which a pollutant may cause effects. Understanding the biological mechanisms
18 underlying various health outcomes can prove crucial in establishing or negating
19 causality. In the absence of human studies data, extensive, well-conducted animal
20 toxicological studies can support determinations of causality, if the evidence base
21 indicates that similar responses are expected in humans under ambient exposure
22 conditions.
23 Interpretations of animal toxicological studies are affected by limitations associated with
24 extrapolation between animal and human responses. The differences between humans
25 and other species have to be taken into consideration, including metabolism, hormonal
26 regulation, breathing pattern, and differences in lung structure and anatomy. Also, in spite
27 of a high degree of homology and the existence of a high percentage of orthologous
28 genes across humans and rodents (particularly mice), extrapolation of molecular
29 alterations at the gene or protein level is complicated by species-specific differences in
30 transcriptional regulation and/or signaling. Given these differences, there are
31 uncertainties associated with quantitative extrapolations of observed pollutant-induced
32 pathophysiological alterations between laboratory animals and humans, as those
33 alterations are under the control of widely varying biochemical, endocrine, and neuronal
34 factors.
35 For ecological effects assessment, both laboratory and field studies (including field
36 experiments and observational studies) can provide useful data for causal determination.
37 Because conditions can be controlled in laboratory studies, responses may be less
January 2015 xlviii DRAFT: Do Not Cite or Quote
-------
1 variable and smaller effects may be easier to detect. However, the control conditions may
2 limit the range of responses (e.g., animals may not be able to seek alternative food
3 sources) or incompletely reflect pollutant bioavailability, so the responses under
4 controlled conditions may not reflect responses that would occur in the natural
5 environment. In addition, larger-scale processes are difficult to reproduce in the
6 laboratory.
7 Field observational studies measure biological changes in uncontrolled situations with
8 high natural variability (in organismal genetics or in abiotic seasonal, climatic, or
9 soil-related factors) and describe an association between a disturbance and an ecological
10 effect. Field data can provide important information to assess multiple stressors or
11 circumstances where site-specific factors significantly influence exposure. They are also
12 often useful for analyses of pollutant effects at larger geographic scales and higher levels
13 of biological organization. However, because conditions are not controlled, variability of
14 the response is expected to be higher and may mask effects. Field surveys are most useful
15 for linking stressors with effects when stressor and effect levels are measured
16 concurrently. The presence of confounding factors can make it difficult to attribute
17 observed effects to specific stressors.
18 Ecological impacts of pollutants are also evaluated in studies "intermediate" between the
19 lower variability typically associated with laboratory exposures and high natural
20 variability usually found in field studies. Some use environmental media collected from
21 the field to examine the biological responses under controlled laboratory conditions.
22 Others are experiments that are performed in the natural environment while controlling
23 for some, but not all, of the environmental or genetic variability (e.g., mesocosm studies).
24 This type of study in manipulated natural environments can be considered a hybrid
25 between a field experiment and laboratory study because some sources of response
26 variation are removed through use of control conditions, while others are included to
27 mimic natural variation. Such studies make it possible to observe community and/or
28 ecosystem dynamics and provide strong evidence for causality when combined with
29 findings of studies that have been made under more controlled conditions.
b. Considerations in Developing Scientific
Conclusions and Causal Determinations
30 In its evaluation and integration of the scientific evidence on health or welfare effects of
31 criteria pollutants, EPA determines the weight of evidence in support of causation and
32 characterizes the strength of any resulting causal classification. EPA also evaluates the
33 quantitative evidence and draws scientific conclusions, to the extent possible, regarding
January 2015 xlix DRAFT: Do Not Cite or Quote
-------
1 the concentration-response relationships and the loads to ecosystems, exposures, doses or
2 concentrations, exposure duration, and pattern of exposures at which effects are observed.
3 Approaches to assessing the separate and combined lines of human health evidence
4 (e.g., epidemiologic, controlled human exposure, and animal toxicological studies) have
5 been formulated by a number of regulatory and science agencies, including the National
6 Academy of Sciences (NAS) Institute of Medicine [IOM; (IOM. 2008)1. the International
7 Agency for Research on Cancer (IARC. 2006). the U.S. EPA (2005). and the Centers for
8 Disease Control and Prevention [CDC; (CDC. 2004)1. Causal inference criteria have also
9 been described for ecological effects evidence (U.S. EPA. 1998; Fox. 1991). These
10 formalized approaches offer guidance for assessing causality. The frameworks are similar
11 in nature, although adapted to different purposes, and have proven effective in providing
12 a uniform structure and language for causal determinations.
13 The 1964 Surgeon General's report on tobacco smoking defined "cause" as a
14 "significant, effectual relationship between an agent and an associated disorder or disease
15 in the host" (HEW. 1964). More generally, a cause is defined as an agent that brings
16 about an effect or a result. An association is the statistical relationship among variables;
17 alone, however, it is insufficient proof of a causal relationship between an exposure and a
18 health outcome. Unlike an association, a causal claim supports the creation of
19 counterfactual claims; that is, a claim about what the world would have been like under
20 different or changed circumstances (IOM. 2008).
21 Many of the health and environmental outcomes reported in these studies have complex
22 etiologies. Diseases such as asthma, coronary heart disease, or cancer are typically
23 initiated by multiple agents. Outcomes depend on a variety of factors, such as age,
24 genetic background, nutritional status, immune competence, and social factors (IOM.
25 2008; Gee and Payne-Sturges. 2004). Effects on ecosystems are also often multifactorial
26 with a complex web of causation. Further, exposure to a combination of agents could
27 cause synergistic or antagonistic effects. Thus, the observed risk may represent the net
28 effect of many actions and counteractions.
29 To aid judgment, various "aspects"1 of causality have been discussed by many
30 philosophers and scientists. The 1964 Surgeon General's report on tobacco smoking
31 discussed criteria for the evaluation of epidemiologic studies, focusing on consistency,
32 strength, specificity, temporal relationship, and coherence (HEW. 1964). Sir Austin
33 Bradford Hill (Hill. 1965) articulated aspects of causality in epidemiology and public
34 health that have been widely used (IOM. 2008; IARC. 2006; U.S. EPA. 2005; CDC.
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.
January 2015 1 DRAFT: Do Not Cite or Quote
-------
1 2004). These aspects (Hill. 1965) have been modified (Table I) for use in causal
2 determinations specific to health and welfare effects for pollutant exposures (U.S. EPA.
3 2009).' Although these aspects provide a framework for assessing the evidence, they do
4 not lend themselves to being considered in terms of simple formulas or fixed rules of
5 evidence leading to conclusions about causality (Hill. 1965). For example, one cannot
6 simply count the number of studies reporting statistically significant results or
7 statistically nonsignificant results and reach credible conclusions about the relative
8 weight of evidence and the likelihood of causality. Rather, these aspects provide a
9 framework for systematic appraisal of the body of evidence, informed by peer and public
10 comment and advice, which includes weighing alternative views on controversial issues.
11 In addition, it is important to note that the aspects in Table I cannot be used as a strict
12 checklist, but rather to determine the weight of evidence for inferring causality. In
13 particular, not meeting one or more of the principles does not automatically preclude a
14 determination of causality [see discussion in (CDC. 2004)].
1 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 lexicological
studies, as well as in vitro data, and to be more consistent with the EPA Guidelines for Carcinogen Risk Assessment.
January 2015 li DRAFT: Do Not Cite or Quote
-------
Table I
Aspects to aid in judging causality.
Aspect
Description
Consistency
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. Elevated risks are not defined by statistical
significance. 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, animal, or welfare studies) may be strengthened by other lines of
evidence that support a cause-and-effect interpretation of the association. There may
be coherence in demonstrating effects from evidence across various fields and/or
across multiple study designs or related health endpoints within one scientific line of
evidence. For example, evidence on 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).
Biological plausibility
Biological gradient
(exposure-response
relationship)
An inference of causality is strengthened by results from experimental studies or other
sources demonstrating biologically plausible mechanisms. A proposed mechanism,
which is based on experimental evidence and which links exposure to an agent to a
given effect, is an important source of support for causality.
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).
Strength of the
observed association
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.
Experimental evidence 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.
Temporality of the
observed association
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
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.
Analogy 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.
Consistency of findings across studies is informed by the repeated observation of effects
or associations across multiple independent studies. Further strength is provided by
January 2015
lii
DRAFT: Do Not Cite or Quote
-------
1 reproducibility of findings in different populations under different circumstances.
2 However, discordant results among independent investigations may be explained by
3 differences in study methods, random errors, exposure, confounding factors, or study
4 power, and thus may not be used to rule out a causal connection.
5 In evaluating the consistency of findings across studies, EPA emphasizes the importance
6 of examining the pattern of results across various studies, and has not focused solely on
7 statistical significance or the magnitude of the direction of the association as criteria of
8 study reliability. Statistical significance is influenced by a variety of factors including,
9 but not limited to, the size of the study, exposure and outcome measurement error, and
10 statistical model specifications. Statistical significance is just one of the means of
11 evaluating confidence in the observed relationship and assessing the probability of
12 chance as an explanation. Other indicia of reliability such as the consistency and
13 coherence of a body of studies as well as other confirming data may be used to justify
14 reliance on the results of a body of epidemiologic studies, even if individual studies may
15 lack statistical significance. Traditionally, statistical significance is used to a larger extent
16 to evaluate the findings of controlled human exposure and animal toxicology studies.
17 Understanding that statistical inferences may result in both false positives and false
18 negatives, consideration is given to both trends in data and reproducibility of results.
19 Thus, in drawing judgments regarding causality, EPA emphasizes statistically significant
20 findings from experimental studies, but does not limit its focus or consideration to
21 statistically significant results in epidemiologic studies.
22 In evaluating the strength of the observed association, EPA considers both the magnitude
23 and statistical precision (i.e., width of confidence interval) of the association in
24 epidemiologic studies. In a large study that accounts for several potential confounding
25 factors, a strong association can serve to increase confidence that a finding is not due to a
26 weak unmeasured confounder, chance, or other biases. However, in a study that accounts
27 for several potential confounding factors and other sources of bias, a weak association
28 does not rule out a causal connection. The health effects evaluated in the ISAs tend to
29 have multiple risk factors that likely vary in strength of effect, and the magnitude of
30 effect of air pollution exposure will depend on the prevalence of other risk factors in the
31 study population. Further, a small effect size can be important from a public health
32 impact perspective. The air pollution-related change in a health effect observed in a study
33 represents a shift in the distribution of responses in the study population and potentially
34 an increase in the proportion of individuals with clinically important effects.
35 In making judgments regarding causality, the biological plausibility of effects resulting
36 from air pollutant exposure is considered. Experimental results from in vivo studies
37 involving animal models and humans, as well as from in vitro studies when appropriate,
January 2015 liii DRAFT: Do Not Cite or Quote
-------
1 may be used to establish biological plausibility and to interpret other lines of evidence
2 (e.g., health effects from epidemiologic studies). Biological plausibility is often provided
3 from understanding the MOA by which exposure to a pollutant leads to health effects.
4 This understanding may encompass several different levels of biological organization
5 including, but not limited to, molecular and cellular events in the pathways leading to
6 disease. While a complete understanding of the MOA is not considered necessary for
7 making causal determinations within the ISA, biological plausibility plays a key role.
c. Framework for Causal Determinations
8 In the ISA, EPA assesses the body of relevant literature, building upon evidence available
9 during previous NAAQS reviews, to draw conclusions on the causal relationships
10 between relevant pollutant exposures and health or environmental effects. ISAs use a
11 five-level hierarchy that classifies the weight of evidence for causation.: This
12 weight-of-evidence evaluation is based on integration of findings from various lines of
13 evidence from across health and environmental effect disciplines that are integrated into a
14 qualitative statement about the overall weight of the evidence and causality. The five
15 descriptors for causal determination are described in Table II.
1 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.
January 2015 liv DRAFT: Do Not Cite or Quote
-------
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
(e.g., doses or exposures generally within one to two
orders of magnitude of recent concentrations). That is,
the pollutant has been shown to result in health effects
in studies in which chance, confounding, and other
biases could be ruled out with reasonable confidence.
For example: (1) controlled human exposure studies
that demonstrate consistent effects, or
(2) observational studies that cannot be explained by
plausible alternatives or that are supported by other
lines of evidence (e.g., animal studies or mode of
action information). Generally, the determination is
based on multiple high-quality studies conducted by
multiple research groups.
Evidence is sufficient to conclude that there is a
causal relationship with relevant pollutant exposures
(e.g., 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, confounding, and other biases 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, the 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. That is, the pollutant has been shown to
result in health effects in studies where results are not
explained by chance, confounding, and other biases,
but uncertainties remain in the evidence overall. For
example: (1) 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 (2) animal
toxicological evidence from multiple studies from
different laboratories demonstrate effects, but limited or
no human data are available. Generally, the
determination is based on 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, confounding, and other biases 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, the determination is
based on multiple studies by multiple research groups.
Suggestive,
but not
sufficient, to
infer a causal
relationship
Evidence is suggestive of a causal relationship with
relevant pollutant exposures but is limited, and chance,
confounding, and other biases cannot be ruled out. For
example: (1) when the body of evidence is relatively
small, at least one high-quality epidemiologic study
shows an association with a given health outcome
and/or at least one high-quality toxicological study
shows effects relevant to humans in animal species, or
(2) when the body of evidence is relatively large,
evidence from studies of varying quality is generally
supportive but not entirely consistent, and there may be
coherence across lines of evidence (e.g., animal
studies or mode of action information) to support the
determination.
Evidence is suggestive of a causal relationship with
relevant pollutant exposures, but chance,
confounding, and other biases 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 indicates there is 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 and lifestages, 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.
January 2015
Iv
DRAFT: Do Not Cite or Quote
-------
1 This standardized language was drawn from sources across the federal government and
2 wider scientific community, especially the EPA Guidelines for Carcinogen Risk
3 Assessment (U.S. EPA. 2005). U.S. Surgeon General's report, The Health Consequences
4 of Smoking (CDC. 2004). and NAS IOM document, Improving the Presumptive
5 Disability Decision-Making Process for Veterans (IOM. 2008). a comprehensive report
6 on evaluating causality.
7 This framework:
8 • describes the kinds of scientific evidence used in making determinations on
9 causal relationships between exposure and health or welfare effects,
10 • summarizes the key aspects of the evaluation of evidence necessary to reach a
11 conclusion about the existence of a causal relationship,
12 • identifies issues and approaches related to uncertainty, and
13 • classifies and characterizes the weight of evidence in support of a general
14 causal relationship.
15 Determination of causality involves the evaluation and integration of evidence for
16 different types of health, ecological, or welfare effects associated with short- and
17 long-term exposure periods. In making determinations of causality, evidence is evaluated
18 for major outcome categories or groups of related endpoints (e.g., respiratory effects,
19 vegetation growth), integrating evidence from across disciplines, and evaluating the
20 coherence of evidence across a spectrum of related endpoints to draw conclusions
21 regarding causality. In discussing the causal determination, EPA characterizes the
22 evidence on which the judgment is based, including strength of evidence for individual
23 endpoints within the outcome category or group of related endpoints.
24 In drawing judgments regarding causality for the criteria air pollutants, the ISA focuses
25 on evidence of effects in the range of relevant pollutant exposures or doses and not on
26 determination of causality at any dose. Emphasis is placed on evidence of effects at doses
27 (e-g-, blood Pb concentration) or exposures (e.g., air concentrations) that are relevant to,
28 or somewhat above, those currently experienced by the population. The extent to which
29 studies of higher concentrations are considered varies by pollutant and major outcome
30 category, but generally includes those with doses or exposures in the range of one to two
31 orders of magnitude above current or ambient conditions. Studies that use higher doses or
32 exposures may also be considered to the extent that they provide useful information to
33 inform understanding of MOA, inter-species differences, or factors that may increase risk
34 of effects for a population and if biological mechanisms have not been demonstrated to
35 differ based on exposure concentration. Thus, a causality determination is based on
36 weight-of-evidence evaluation for health or welfare effects, focusing on the evidence
January 2015 Ivi DRAFT: Do Not Cite or Quote
-------
1 from exposures or doses generally ranging from recent ambient concentrations to one or
2 two orders of magnitude above recent ambient concentrations.
3 In addition, EPA evaluates evidence relevant to understanding the quantitative
4 relationships between pollutant exposures and health or welfare effects. This includes
5 evaluating the form of concentration-response or dose-response relationships and, to the
6 extent possible, drawing conclusions on the concentrations at which effects are observed.
7 The ISA also draws scientific conclusions regarding important exposure conditions for
8 effects and populations and lifestages that may be at greater risk for effects, as described
9 in the following section.
6. Public Health Impact
10 Once a determination is made regarding the causality of relationship between the
11 pollutant and outcome category, the public health impact of exposure to the pollutant is
12 evaluated. Important questions regarding the public health impact include:
13 • What populations and lifestages appear to be differentially affected (i.e., at
14 greater or less risk of experiencing effects)?
15 • What exposure conditions (dose or exposure, duration, and pattern) are
16 important?
17 • What is the severity of the effect (e.g., clinical relevance)?
18 • What is the concentration-response, exposure-response, or dose-response
19 relationship in the human population?
20 • What is the interrelationship between incidence and severity of effect?
21 To address these questions, the entirety of quantitative evidence is evaluated to
22 characterize pollutant concentrations and exposure durations at which effects were
23 observed for exposed populations, including populations and lifestages potentially at
24 increased risk. To accomplish this, evidence is considered from multiple and diverse
25 types of studies, and a study or set of studies that best approximates the
26 concentration-response relationships between health outcomes and the pollutant may be
27 identified. Controlled human exposure studies provide the most direct and quantifiable
28 exposure-response data on the human health effects of pollutant exposures. To the extent
29 available, the ISA evaluates results from epidemiologic studies that characterize the
30 shape of the relationship between a pollutant and a health outcome. Animal data may also
31 inform evaluation of concentration-response relationships, particularly relative to MOAs
32 and characteristics of at-risk populations.
January 2015 Mi DRAFT: Do Not Cite or Quote
-------
a. Approach to Identifying, Evaluating, and
Characterizing At-Risk Factors
1 A critical part of assessing the public health impact of an air pollutant is the
2 identification, evaluation, and characterization of populations potentially at greater risk of
3 an air pollutant-related health effect. Under the Clean Air Act, the NAAQS are intended
4 to protect public health with an adequate margin of safety. In doing so, protection is
5 provided for both the population as a whole and those groups potentially at increased risk
6 for health effects from exposure to a criteria air pollutant. To inform decisions under the
7 NAAQS, the ISA evaluates the currently available information regarding those factors
8 (e.g., lifestage, pre-existing disease, etc.) that contribute to portions of the population
9 being at greater risk for an air pollutant-related health effect.
10 Studies often use a variety of terms to classify factors and subsequently populations that
11 may be at increased risk of an air pollutant-related health effect, including "susceptible,"
12 "vulnerable," "sensitive," and "at-risk," with recent literature introducing the term
13 "response-modifying factor" (Vinikoor-Imler et al.. 2014; Sacks et al.. 2011; U.S. EPA.
14 2010. 2009). The inconsistency in the definitions for each of these terms across the
15 scientific literature has shifted the focus away from answering the key questions: Which
16 populations are at increased risk and what evidence forms the basis of this conclusion?
17 (Vinikoor-Imler et al.. 2014). Due to the lack of a consensus on terminology in the
18 scientific community, the term "susceptible populations" was used in reviews and
19 previous ISAs (Sacks et al.. 2011; U.S. EPA. 2010. 2009) to encompass these various
20 factors. However, it was recognized that even using the term "susceptible populations"
21 was problematic because it often refers to populations at increased risk specifically due to
22 biological or intrinsic factors such as pre-existing disease or lifestage. As such, starting
23 with the ISA for Ozone and Related Photochemical Oxidants (U.S. EPA. 2013). the
24 terminology "at-risk" was introduced to define populations and lifestages potentially at
25 increased risk of an air pollutant-related health effect. In assessing the overall public
26 health impact of an air pollutant, the ISA focuses on identifying, evaluating, and
27 characterizing "at-risk" factors to address the main question of what populations and
28 lifestages are at increased risk of an air pollutant-related health effect. Each "at-risk"
29 factor is evaluated with a focus on identifying whether the factor contributes to a
30 population at increased risk of an air pollutant-related health effect. It is recognized that
31 some factors may lead to a reduction in risk, and these are recognized during the
32 evaluation process, but for the purposes of identifying those populations or lifestages at
33 increased risk to inform decisions on the NAAQS, the focus of this ISA is on
34 characterizing those factors that may increase risk.
January 2015 Iviii DRAFT: Do Not Cite or Quote
-------
1 It is recognized that factors may be intrinsic and increase risk for an effect through a
2 biological mechanism. Intrinsic factors include genetic or developmental factors, race,
3 sex, lifestage, or the presence of pre-existing diseases. In general, people in this category
4 would have a steeper concentration-risk relationship compared to those not in the
5 category. Additionally, increased risk may be attributable to an extrinsic, nonbiological
6 factor, such as SES (e.g., educational attainment, income, access to healthcare), activity
7 pattern, and exercise level. Some groups may be at increased risk because of increased
8 internal dose at a given exposure concentration. This category includes individuals that
9 have a greater dose of delivered pollutant because of breathing patterns such as children
10 who are typically more active outdoors. In addition, some groups could have greater
11 exposure (concentration x time) regardless of the delivered dose, such as outdoor
12 workers. Finally, there are those who might be at increased risk for experiencing a greater
13 exposure by being exposed at a higher concentration, such as populations that live near
14 roadways. Some factors described above are multifaceted and may influence the risk of
15 an air pollutant-related health effect through a combination of ways (e.g., SES).
16 Additionally, it is recognized that some portions of the population may be at increased
17 risk of an air pollutant-related health effect because they experience insults from a
18 combination of factors. The emphasis is to identify and understand the factors that
19 potentially increase the risk of air pollutant-related health effects, regardless of whether
20 the increased risk is due to intrinsic factors, extrinsic factors, increased dose/exposure, or
21 a combination due to the often interconnectedness of factors.
22 To identify at-risk factors that potentially lead to some portions of the population being at
23 increased risk of air pollution-related health effects, the evidence is systematically
24 evaluated across relevant scientific disciplines (i.e., exposure sciences, dosimetry,
25 toxicology, and epidemiology). The evaluation process first consists of evaluating studies
26 that conduct stratified analyses (i.e., epidemiologic, controlled human exposure) to
27 compare populations or lifestages exposed to similar air pollutant concentrations within
28 the same study design. Experimental studies also provide an important line of evidence in
29 evaluating factors that can lead to increased risk of an air pollutant-related health effect.
30 Specifically, toxicological studies conducted using animal models of disease and
31 controlled human exposure studies that examine individuals with underlying disease or
32 genetic polymorphisms can provide support for coherence with the health effects
33 observed in epidemiologic studies as well as an understanding of biological plausibility.
34 The potential increased risk of an air pollutant-related health effect may also be
35 determined from studies that examined factors that result in differential air pollutant
36 exposures. Building on the causal framework discussed in detail in Table II. the
37 characterization of each at-risk factor consists of evaluating the strength of evidence
38 across scientific disciplines and assessing whether the factor can contribute to increased
39 risk of an air pollutant-related health effect. The conclusions drawn consider the "aspects
January 2015 lix DRAFT: Do Not Cite or Quote
-------
1
2
o
6
4
to aid in judging causality" discussed in Table I. 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 III.
Table
Characterization of evidence for potential at-risk factors.
Classification
Health Effects
Adequate evidence
There is substantial, consistent evidence within a discipline to conclude that a factor results in
a population or lifestage being at increased risk of air pollutant-related health effect(s) relative
to some reference population or lifestage. Where applicable, this evidence includes coherence
across disciplines. Evidence includes multiple high-quality studies.
Suggestive The collective evidence suggests that a factor results in a population or lifestage being at
evidence increased risk of air pollutant-related health effect(s) 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 The collective evidence is inadequate to determine whether a factor results in a population or
evidence lifestage being at increased risk of air pollutant-related health effect(s) 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.
Evidence of There is substantial, consistent evidence within a discipline to conclude that a factor does not
no effect result in a population or lifestage being at increased risk of air pollutant-related health effect(s)
relative to some reference population or lifestage. Where applicable, the evidence includes
coherence across disciplines. Evidence includes multiple high-quality studies.
b.
Evaluating Adversity of Human Health Effects
5
6
7
8
9
10
11
12
13
14
In evaluating health evidence, a number of factors can be considered in delineating
between adverse and nonadverse health effects resulting from exposure to air 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 more subtle health effects
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
January 2015
Ix
DRAFT: Do Not Cite or Quote
-------
1 between populations. Some changes that may not be considered adverse in healthy
2 individuals would be potentially adverse in more at-risk individuals.
3 Professional scientific societies may evaluate the magnitude of change in an outcome or
4 event that is considered adverse. For example, the extent to which changes in lung
5 function are adverse has been discussed by the American Thoracic Society in an official
6 statement titled What Constitutes an Adverse Health Effect of Air Pollution? (ATS.
7 2000). An air pollution-induced shift in the population distribution of a given risk factor
8 for a health outcome was viewed as adverse, even though it may not increase the risk of
9 any one individual to an unacceptable level. For example, a population with asthma could
10 have a distribution of lung function such that no identifiable individual has a level
11 associated with significant impairment. Exposure to air pollution could shift the
12 distribution such that no identifiable individual experiences clinically relevant effects.
13 This shift toward decreased lung function, however, would be considered adverse
14 because individuals within the population would have diminished reserve function and
15 therefore would be at increased risk to further environmental insult. The committee also
16 observed that elevations of biomarkers, such as cell number and types, cytokines, and
17 reactive oxygen species, may signal risk for ongoing injury and clinical effects or may
18 simply indicate transient responses that can provide insights into mechanisms of injury,
19 thus illustrating the lack of clear boundaries that separate adverse from nonadverse
20 effects.
21 The more subtle health outcomes may be connected mechanistically to health events that
22 are clearly adverse. For example, air pollution may affect markers of transient myocardial
23 ischemia such as ST-segment abnormalities or onset of exertional angina. These effects
24 may not be apparent to the individual, yet may still increase the risk of a number of
25 cardiac events, including myocardial infarction and sudden death. Thus, small changes in
26 physiological measures may not appear to be clearly adverse when considered alone, but
27 may be a part of a coherent and biologically plausible chain of related health outcomes
28 that range up to responses that are very clearly adverse, such as hospitalization or
29 mortality.
c. Concentration-Response Relationships
30 An important consideration in characterizing the public health impacts associated with
31 exposure to a pollutant is whether the concentration-response relationship is linear across
32 the range of concentrations or if nonlinear relationships exist along any part of this range.
33 The shape of the concentration-response curve at and below the level of the current
34 standards is of particular interest. Various sources of variability and uncertainty, such as
January 2015 Ixi DRAFT: Do Not Cite or Quote
-------
1 low data density in the lower concentration range, possible influence of exposure
2 measurement error, and variability among individuals with respect to air pollution health
3 effects, tend to smooth and "linearize" the concentration-response function and thus can
4 obscure the existence of a threshold or nonlinear relationship. Because individual
5 thresholds vary from person-to-person due to individual differences such as genetic
6 differences or pre-existing disease conditions (and even can vary from one time to
7 another for a given person), it can be difficult to demonstrate that a threshold exists in a
8 population study. These sources of variability and uncertainty may explain why the
9 available human data at ambient concentrations for some environmental pollutants
10 (e.g., PM, Os, Pb, environmental tobacco smoke, radiation) do not exhibit
11 population-level thresholds for cancer or noncancer health effects, even though likely
12 mechanisms include nonlinear processes for some key events.
7. Public Welfare Impact
13 Once a determination is made regarding the causality of relationships between the
14 pollutant and outcome category, important questions regarding the public welfare impact
15 include:
16 • What endpoints or services appear to be differentially affected (i.e., at greater
17 or less risk of experiencing effects)? What elements of the ecosystem
18 (e.g., types, regions, taxonomic groups, populations, functions, etc.) appear to
19 be affected, or are more sensitive to effects? Are there differences between
20 locations or materials in welfare effects responses, such as impaired visibility
21 or materials damage?
22 • What is concluded from the evidence with regard to other types of welfare
23 effects?
24 • Under what exposure conditions (amount deposited or concentration,
25 duration, and pattern) are effects seen?
26 • What is the shape of the concentration-response, exposure-response, or
27 dose-response relationship?
28 To address these questions, the entirety of quantitative evidence is evaluated to
29 characterize pollutant concentrations and exposure durations at which effects were
30 observed. To accomplish this, evidence is considered from multiple and diverse types of
31 studies, and a study or set of studies that best approximates the concentration-response
32 relationships between welfare outcomes and the pollutant may be identified. Controlled
33 experimental studies provide the most direct and quantifiable exposure-response data on
34 the effects of pollutant exposures. To the extent available, the ISA also evaluates results
35 from less controlled field studies that characterize the shape of the relationship between a
36 pollutant and an outcome. Other types of data may also inform evaluation of
January 2015 Mi DRAFT: Do Not Cite or Quote
-------
1 concentration-response relationships, particularly relative to MOAs and characteristics of
2 at-risk ecosystems.
a. Evaluating Adversity of Ecological and Other
Welfare Effects
3 The final step in assessing the public welfare impact of an air pollutant is the evaluation
4 of the level considered to be adverse. A secondary standard, as defined in
5 Section 109(b)(2) of the CAA must "specify a level of air quality the attainment and
6 maintenance of which, in the judgment of the Administrator, based on such criteria, is
7 requisite to protect the public welfare from any known or anticipated adverse effects
8 associated with the presence of such air pollutant in the ambient air." In setting standards
9 that are "requisite" to protect public health and welfare, as provided in Section 109(b),
10 EPA's task is to establish standards that are neither more nor less stringent than necessary
11 for these purposes.
12 Adversity of ecological effects can be understood in terms ranging in biological level of
13 organization from the cellular level to the individual organism and to the population,
14 community, and ecosystem levels. In the context of ecology, a population is a group of
15 individuals of the same species, and a community is an assemblage of populations of
16 different species that inhabit an area and interact with one another. An ecosystem is the
17 interactive system formed from all living organisms and their abiotic (physical and
18 chemical) environment within a given area (IPCC. 2007). The boundaries of what could
19 be called an ecosystem are somewhat arbitrary, depending on the focus of interest or
20 study. Thus, the extent of an ecosystem may range from very small spatial scales to,
21 ultimately, the entire Earth (IPCC. 2007).
22 Effects on an individual organism are generally not considered to be adverse to public
23 welfare. However if effects occur to enough individuals within a population, then
24 communities and ecosystems may be disrupted. Changes to populations, communities,
25 and ecosystems can in turn result in an alteration of ecosystem processes. Ecosystem
26 processes are defined as the metabolic functions of ecosystems including energy flow,
27 elemental cycling, and the production, consumption, and decomposition of organic matter
28 (U.S. EPA. 2002). Growth, reproduction, and mortality are species-level endpoints that
29 may be clearly linked to community and ecosystem effects and are considered to be
30 adverse when negatively affected. Other endpoints such as changes in behavior and
31 physiological stress can decrease ecological fitness of an organism, but are harder to link
32 unequivocally to effects at the population, community, and ecosystem level. Support for
33 consideration of adversity beyond the species level by making explicit the linkages
January 2015 Ixiii DRAFT: Do Not Cite or Quote
-------
1 between stress-related effects at the species and effects at the ecosystem level is found in
2 A Framework for Assessing and Reporting on Ecological Condition: an SAB report
3 (U.S. EPA. 2002). Additionally, the National Acid Precipitation Assessment Program
4 (NAPAP. 1991) uses the following working definition of "adverse ecological effects" in
5 the preparation of reports to Congress mandated by the Clean Air Act: "any injury
6 (i.e., loss of chemical or physical quality or viability) to any ecological or ecosystem
7 component, up to and including the regional level, over both long and short terms."
8 Beyond species-level impacts, consideration of ecosystem services allows for evaluation
9 of how pollutant exposure may adversely impact species or processes of particular
10 economic or cultural importance to humans. On a broader scale, ecosystem services may
11 provide indicators for ecological impacts. Ecosystem services are the benefits that people
12 obtain from ecosystems (UNEP. 2003). According to the Millennium Ecosystem
13 Assessment, ecosystem services include: "provisioning services such as food and water;
14 regulating services such as regulation of floods, drought, land degradation, and disease;
15 supporting services such as soil formation and nutrient cycling; and cultural services such
16 as recreational, spiritual, religious, and other nonmaterial benefits" (UNEP. 2003). For
17 example, a more subtle ecological effect of pollution exposure may result in a clearly
18 adverse impact on ecosystem services if it results in a population decline in a species that
19 is recreationally or culturally important.
20 A consideration in evaluating adversity of climate-related effects is that criteria air
21 pollutants have both direct and indirect effects on radiative forcing. For example, CO has
22 a relatively small direct forcing effect, but it influences the concentrations of other
23 atmospheric species, such as Os and methane (CFLj), which are important contributors to
24 climate forcing. PM has both direct and indirect effects. For example, black carbon and
25 sulfate contribute directly to warming and cooling, while aerosols are involved in cloud
26 formation. Thus, it is crucial to consider the role of multiple pollutants together in
27 evaluating the climate impact of criteria pollutants. Although climate effects of criteria
28 air pollutants impact terrestrial and aquatic environments in diverse ways over multiple
29 time scales, their effect on temperature is the main metric of adversity, with some
30 consideration of proximate effects such as precipitation and relatively rapid feedbacks
31 impacting the composition of the troposphere. Downstream effects such as land use
32 changes are more difficult to link back to changes in concentrations that would be
33 produced by the NAAQS. The relative adversity of U.S. versus global emissions and
34 concentrations is informed by regional climate modeling studies, including consideration
35 of uncertainty and spatial and temporal variability.
36 The adversity of visibility impacts may be expressed in terms of psychological stress,
37 such as impairment of aesthetic quality or enjoyment of the environment, or in monetary
January 2015 Ixiv DRAFT: Do Not Cite or Quote
-------
1 terms, such as willingness to pay to improve air quality. Understanding the relationship
2 between pollutant concentration and perception of visibility, including distinguishing
3 between concerns about health risks due to air pollution and perceived visibility
4 impairment, are crucial in evaluating the level of protection provided by a welfare-based
5 secondary NAAQS.
6 Adversity of materials damage is evaluated considering the impact to human and
7 economic well being. Physical damage and soiling impair aesthetic qualities and function
8 of materials. Additionally, damage to property and cultural heritage sites due to pollutant
9 deposition may be considered adverse.
b. Quantitative Relationships: Effects on Welfare
10 Evaluations of causality generally consider the probability of quantitative changes in
11 welfare effects in response to exposure. A challenge to the quantification of
12 exposure-response relationships for ecological effects is the great regional and local
13 spatial variability, as well as temporal variability, in ecosystems. Thus,
14 exposure-response relationships are often determined for a specific ecological system and
15 scale, rather than at the national or even regional scale. Quantitative relationships,
16 therefore, are estimated site by site and may differ greatly between ecosystems.
January 2015 Ixv DRAFT: Do Not Cite or Quote
-------
References for Preamble
ATS (American Thoracic Society). (2000). What constitutes an adverse health effect of air pollution? Am
J Respir Crit Care Med 161: 665-673. http://dx.doi.Org/10.1164/ajrccm.161.2.ats4-00
Bell. ML: Dominici F: Samet JM. (2005). Ameta-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
CAA (Clean Air Act). (1990a). Clean Air Act, as amended by Pub. L. No. 101-549, section 108: Air
quality criteria and control techniques, http://www.law.cornell.edu/uscode/text/42/7408
CAA (Clean Air Act). (1990b). Clean Air Act, as amended by Pub. L. No. 101-549, section 109: National
primary and secondary ambient air quality standards, 42 USC 7408.
http://www.epa.gov/air/caa/titlel.html#ia
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. J Toxicol 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 [Review]. 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
IOM (Institute of Medicine). (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
IPCC (Intergovernmental Panel on Climate Change). (2007). Climate change 2007: Impacts, adaptation
and vulnerability. Cambridge, UK: Cambridge University Press, http://www.ipcc.ch/ipccreports/ar4-
wg2.htm
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.
Sacks. JD: Stanek. LW: Luben. TJ: Johns. DO: Buckley. BJ: Brown. JS: Ross. M. (2011). Particulate-
matter induced health effects: Who is susceptible? [Review]. Environ Health Perspect 119: 446-454.
http://dx.doi.org/10.1289/ehp.1002255
January 2015 Ixvi DRAFT: Do Not Cite or Quote
-------
U.S. EPA (U.S. Environmental Protection Agency). (1998). Guidelines for ecological risk assessment
[EPA Report]. (EPA/630/R-95/002F). Washington, DC: U.S. Environmental Protection Agency, Risk
Assessment Forum, 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://vosemite.epa.gov/sab%5CSABPRODUCT.NSF/7700D7673673CE83852570CA0075458A/$Fil
e/epec02009.pdf
U.S. EPA (U.S. Environmental Protection Agency). (2005). Guidelines for carcinogen risk assessment.
(EPA/63O/P-03/00IF). Washington, DC: U.S. Environmental Protection Agency, Risk Assessment
Forum, http://www.epa. gov/cancerguidelines/
U.S. EPA (U.S. Environmental Protection Agency). (2009). Integrated science assessment for particulate
matter [EPA Report]. (EPA/600/R-08/139F). Research Triangle Park, NC: U.S. Environmental
Protection Agency, National Center for Environmental Assessment.
http://cfpub.epa.gov/ncea/cfm/recordisplav.cfm7de id=216546
U.S. EPA (U.S. Environmental Protection Agency). (2010). Integrated science assessment for carbon
monoxide [EPA Report]. (EPA/600/R-09/019F). Research Triangle Park, NC: U.S. Environmental
Protection Agency, National Center for Environmental Assessment.
http://cfpub.epa.gov/ncea/cfm/recordisplav.cfm7de id=218686
U.S. EPA (U.S. Environmental Protection Agency). (2013). Integrated science assessment for ozone and
related photochemical oxidants. (EPA/600/R-10/076F). Research Triangle Park, NC: U.S.
Environmental Protection Agency, National Center for Environmental Assessment.
http://cfpub.epa.gov/ncea/isa/recordisplav.cfm?deid=247492
UNEP (United Nations Environment Programme). (2003). Ecosystems and human well-being: A
framework for assessment. Washington, DC: Island Press.
Vinikoor-Imler. LC: Owens. EO: Nichols. JL: Ross. M: Brown. JS: Sacks. JD. (2014). Evaluating
potential response-modifying factors for associations between ozone and health outcomes: a weight-
of-evidence approach [Review]. Environ Health Perspect 122: 1166-1176.
http://dx.doi.org/10.1289/ehp.1307541
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. http://dx.doi.org/10.1289/ehp.00108419
January 2015 Ixvii DRAFT: Do Not Cite or Quote
-------
PREFACE
Legislative Requirements for the Review of the National Ambient
Air Quality Standards
1 Two sections of the Clean Air Act (CAA) govern the establishment, review, and revision
2 of the National Ambient Air Quality Standards (NAAQS). Section 108 [42 U.S. Code
3 (U.S.C.) 7408] directs the Administrator to identify and list certain air pollutants and then
4 to issue air quality criteria for those pollutants. The Administrator is to list those air
5 pollutants that in her "judgment, cause or contribute to air pollution which may
6 reasonably be anticipated to endanger public health or welfare;" "the presence of which
7 in the ambient air results from numerous or diverse mobile or stationary sources;" and
8 "for which ... [the Administrator] plans to issue air quality criteria ..." [42 U.S.C.
9 7408(a)(l); (CAA. 1990a)1. Air quality criteria are intended to "accurately reflect the
10 latest scientific knowledge useful in indicating the kind and extent of all identifiable
11 effects on public health or welfare, which may be expected from the presence of [a]
12 pollutant in the ambient air ..." [42 U.S.C. 7408(b)]. Section 109 [42 U.S.C. 7409;
13 (CAA. 1990b)] directs the Administrator to propose and promulgate "primary" and
14 "secondary" NAAQS for pollutants for which air quality criteria are issued.
15 Section 109(b)(l) defines a primary standard as one "the attainment and maintenance of
16 which in the judgment of the Administrator, based on such criteria and allowing an
17 adequate margin of safety, are requisite to protect the public health."1 A secondary
18 standard, as defined in Section 109(b)(2), must "specify a level of air quality the
19 attainment and maintenance of which, in the judgment of the Administrator, based on
20 such criteria, is requisite to protect the public welfare from any known or anticipated
21 adverse effects associated with the presence of [the] air pollutant in the ambient air."2
22 The requirement that primary standards provide an adequate margin of safety was
23 intended to address uncertainties associated with inconclusive scientific and technical
24 information available at the time of standard setting. It was also intended to provide a
25 reasonable degree of protection against hazards that research has not yet identified.3 Both
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).
2 Welfare effects as defined in 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"(CAA. 2005).
3 See Lead Industries Association v. EPA, 647 F.2d 1130, 1154 [(District of Columbia Circuit (D.C. Cir) 1980];
American Petroleum Institute v. Costle, 665 F.2d 1176, 1186 (D.C. Cir. \9%l)', American Farm Bureau Federation
January 2015 Ixviii DRAFT: Do Not Cite or Quote
-------
1 kinds of uncertainty are components of the risk associated with pollution at levels below
2 those at which human health effects can be said to occur with reasonable scientific
3 certainty. Thus, in selecting primary standards that provide an adequate margin of safety,
4 the Administrator is seeking not only to prevent pollution levels that have been
5 demonstrated to be harmful but also to prevent lower pollutant levels that may pose an
6 unacceptable risk of harm, even if the risk is not precisely identified as to nature or
7 degree. The CAA does not require the Administrator to establish a primary NAAQS at a
8 zero-risk level or at background concentration levels, but rather at a level that reduces
9 risk sufficiently so as to protect public health with an adequate margin of safety.: In so
10 doing, protection is provided for both the population as a whole and those groups
11 potentially at increased risk for health effects from exposure to the air pollutant for which
12 each NAAQS is set.
13 In addressing the requirement for an adequate margin of safety, the U.S. Environmental
14 Protection Agency (EPA) considers such factors as the nature and severity of the health
15 effects involved, the size of the sensitive group(s), and the kind and degree of the
16 uncertainties. The selection of any particular approach to providing an adequate margin
17 of safety is a policy choice left specifically to the Administrator's judgment.2
18 In setting standards that are "requisite" to protect public health and welfare as provided in
19 Section 109(b), EPA's task is to establish standards that are neither more nor less
20 stringent than necessary for these purposes. In so doing, EPA may not consider the costs
21 of implementing the standards.3 Likewise, "[Attainability and technological feasibility
22 are not relevant considerations in the promulgation of national ambient air quality
23 standards."4
24 Section 109(d)(l) requires that "not later than December 31, 1980, and at 5-year intervals
25 thereafter, the Administrator shall complete a thorough review of the criteria published
26 under Section 108 and the national ambient air quality standards... and shall make such
27 revisions in such criteria and standards and promulgate such new standards as may be
28 appropriate..." Section 109(d)(2) requires that an independent scientific review
29 committee "shall complete a review of the criteria... and the national primary and
30 secondary ambient air quality standards... and shall recommend to the Administrator any
31 new... standards and revisions of existing criteria and standards as may be
v. EPA, 559 F. 3d 512, 533 (D.C. Cir. 2009); Association of Battery Recydersv. EPA, 604 F. 3d 613, 617-18 (D.C.
Cir. 2010).
1 See Lead Industries v. EPA, 647 F.2d at 1156 n.51; Mississippi v. EPA, 111, F. 3d 246, 255, 262-63 (D.C. Cir.
2013).
2 See Lead Industries Association v. EPA, 647 F.2d at 1161-62; Mississippi v. EPA, 723 F. 3d at 265.
3 See generally, Whitman v. American Trucking Associations, 531 U.S. 457, 465-472, 475-476 (2001).
4 See American Petroleum Institute v. Costle, 665 F. 2d at 1185.
January 2015 Ixix DRAFT: Do Not Cite or Quote
-------
1 appropriate...." Since the early 1980s, this independent review function has been
2 performed by the Clean Air Scientific Advisory Committee (CASAC).:
Introduction to the Primary National Ambient Air Quality Standard
for Nitrogen Dioxide
3 Nitrogen dioxide (NCh) is the indicator for gaseous oxides of nitrogen [e.g., NC>2, nitric
4 oxide (NO)]. Consistent with Section 108(c) of the CAA (42 U.S.C.21 7408), EPA
5 considers the term oxides of nitrogen to refer to all forms of oxidized nitrogen including
6 multiple gaseous species (e.g., NO2, NO) and particulate species (e.g., nitrates). The
7 review of the primary NO2 NAAQS focuses on evaluating the health effects associated
8 with exposure to the gaseous oxides of nitrogen. The atmospheric chemistry, exposure,
9 and health effects associated with nitrogen compounds present in particulate matter (PM)
10 were most recently considered in the EPA's review of the NAAQS for PM. The welfare
11 effects associated with oxides of nitrogen are being considered in a separate assessment
12 as part of the review of the secondary NO2 NAAQS and sulfur dioxide [SO2; (U.S. EPA.
13 2013)1.
14 NAAQS are defined by four basic elements: indicator, averaging time, level, and form.
15 The indicator defines the pollutant to be measured in the ambient air for the purpose of
16 determining compliance with the standard. The averaging time defines the time period
17 over which air quality measurements are to be obtained and averaged or cumulated,
18 considering evidence of effects associated with various time periods of exposure. The
19 level of a standard defines the air quality concentration used (i.e., an ambient
20 concentration of the indicator pollutant) in determining whether the standard is achieved.
21 The form of the standard defines the air quality statistic that is compared to the level of
22 the standard in determining whether an area attains the standard. For example, the form
23 of the current primary 1-hour NO2 standard is the 3-year average of the 98th percentile of
24 the annual distribution of 1-hour daily maximum NO2 concentrations. The Administrator
25 considers these four elements collectively in evaluating the protection to public health
26 provided by the primary NAAQS.
History of the Review of Air Quality Criteria for the Oxides of
Nitrogen and the Primary National Ambient Air Quality Standards
for Nitrogen Dioxide
27 In 1971, the EPA added nitrogen oxides to the list of criteria pollutants under
28 Section 108(a)(l) of the CAA and issued the initial air quality criteria [36 Federal
1 Lists of CAS AC members and of members of the CAS AC Oxides of Nitrogen Primary NAAQS Review Panel are
available at: http://yosemite.epa.gov/sab/sabproduct.nsf/WebCASAC/CommitteesandMembership7OpenDocument.
January 2015 Ixx DRAFT: Do Not Cite or Quote
-------
1
2
o
6
4
5
6
7
Register (FR) 1515, January 30, 1971]. Based on these air quality criteria, the EPA
promulgated NAAQS for nitrogen oxides using NC>2 as the indicator (36 FR 8186, April
30, 1971). Both primary and secondary standards were set at 100 ug/m3 [equal to
0.053 parts per million (ppm)], annual average. The standards were based on scientific
information contained in the 1971 Air Quality Criteria Document for Nitrogen Oxides
(U.S. EPA. 1971). Since then, the Agency has completed multiple reviews of the air
quality criteria upon which the NAAQS are set and the primary standards themselves.
Table I provides a brief summary of these reviews.
Table I History of the primary National Ambient Air Quality Standards for
nitrogen dioxide (NO2) since 1971.
Final
Rule/Decisions
1971
36 FR 81 86
Apr 30, 1971
1985
50 FR 25532
Jun 19, 1985
1996
61 FR 52852
Oct8, 1996
2010
75 FR 6474
Feb9, 2010
Indicator Averaging Time Level Form
NO2
Primary
Primary
NO2
Primary
1 year 53 ppba Annual arithmetic average
NO2 standard retained, without revision.
NO2 standard retained, without revision.
1 hour 100ppb 3-year average of the 98th
annual distribution of daily
concentrations
annual NO2 standard retained, without revision.
percentile of the
maximum 1-hour
aThe initial standard level of the annual NO2 standard was 100 |jg/m3 which is equal to 0.053 ppm or 53 parts per billion (ppb).
The units for the standard level were officially changed to ppb in the final rule issued in 2010 (75 FR 6531, February 9, 2010).
9
10
11
12
13
14
15
16
17
18
The EPA retained the primary and secondary NC>2 standards, without revision, in reviews
completed in 1985 and 1996 (50 FR 25532, June 19, 1985; 61 FR 52852,
October 8, 1996). These decisions were informed, respectively, by scientific information
contained in the 1982 Air Quality Criteria Document for Oxides of Nitrogen [(U.S. EPA.
1982) which updated the scientific criteria upon which the initial NO2 standards were
based] and the 1993 Air Quality Criteria Document for the Oxides of Nitrogen
(U.S. EPA. 1993). In the latter of the two decisions, the EPA concluded that "the existing
annual primary standard appears to be both adequate and necessary to protect human
health against both long- and short-term NO2 exposures" and that retaining the existing
annual standard is consistent with the scientific data assessed in the 1993 Air Quality
January 2015
Ixxi
DRAFT: Do Not Cite or Quote
-------
1 Criteria Document (U.S. EPA. 1993) and the Staff Paper (U.S. EPA. 1995) and with the
2 advice and recommendations of CASAC" (61 FR 52854, October 8, 1996).1
3 The last review of the air quality criteria for oxides of nitrogen (health criteria) and the
4 primary NO2 standard was initiated in December 2005 (70 FR 73236,
5 December 9, 2005).2> 3 The Agency's plans for conducting the review were presented in
6 the Integrated Review Plan (IRP) for the Primary National Ambient Air Quality Standard
7 for NO2 (U.S. EPA. 2007a). which included consideration of comments received during a
8 CASAC consultation as well as public comment on a draft IRP. The science assessment
9 for the review was described in the 2008 Integrated Science Assessment for Oxides of
10 Nitrogen—Health Criteria (U.S. EPA. 2008a). multiple drafts of which received review
11 by CASAC and the public. The EPA also conducted quantitative human risk and
12 exposure assessments, after consultation with CASAC and receiving public comment on
13 a draft analysis plan (U.S. EPA. 2007b). These technical analyses were presented in the
14 Risk and Exposure Assessment (REA) to Support the Review of the NO2 Primary
15 National Ambient Air Quality Standard (U.S. EPA. 2008b). multiple drafts of which
16 received CASAC and public review.
17 Over the course of the last review, the EPA made several changes to the NAAQS review
18 process. An important change was the discontinuation of the Staff Paper, which
19 traditionally contained staff evaluations to bridge the gap between the Agency's science
20 assessments and the judgments required of the EPA Administrator in determining
21 whether it was appropriate to retain or revise the NAAQS.4 In the course of reviewing the
22 second draft REA, however, CASAC expressed the view that the document would be
23 incomplete without the addition of a policy assessment chapter presenting an integration
24 of evidence-based considerations and risk and exposure assessment results. CASAC
25 stated that such a chapter would be "critical for considering options for the NAAQS for
26 NO2" (Samet 2008). In addition, within the period of CASAC's review of the second
27 draft REA, the EPA's Deputy Administrator indicated in a letter to the CASAC chair,
28 addressing earlier CASAC comments on the NAAQS review process, that the risk and
1 In presenting rationale for the final decision, the EPA noted that "a 0.053 ppm annual standard would keep annual
NO2 concentrations considerably below the long-term levels for which serious chronic effects have been observed in
animals" and that "[retaining the existing standard would also provide protection against short-term peak NCh
concentrations at the levels associated with mild changes in pulmonary function and airway responsiveness observed
in controlled human [exposure] studies" (61 FR 52854, October 8, 1996; 60 FR 52874, 52880, October 11, 1995).
2 Documents related to reviews completed in 2010 and 1996 are available at:
http ://www. epa. gov/ttn/naaqs/standards/nox/s noxindex, html.
3 The EPA conducted a separate review of the secondary NO2 NAAQS jointly with a review of the secondary SO2
NAAQS. The Agency retained those secondary standards, without revision, to address the direct effects on
vegetation of exposure to oxides of nitrogen and sulfur (77 FR 20218, April 3, 2012).
4 Initial changes to the NAAQS review process included a policy assessment document reflecting Agency (rather
than staff) views published as an advanced notice of public rulemaking (ANPR). Under this process, the ANPR
would have been reviewed by CASAC (Peacock. 2006).
January 2015 Ixxii DRAFT: Do Not Cite or Quote
-------
1 exposure assessment will include "a broader discussion of the science and how
2 uncertainties may effect decisions on the standard" and "all analyses and approaches for
3 considering the level of the standard under review, including risk assessment and weight
4 of evidence methodologies" (Peacock. 2008). Accordingly, the final 2008 REA included
5 a policy assessment chapter that considered the scientific evidence in the 2008 ISA and
6 the exposure and risk results presented in other chapters of the 2008 REA as they related
7 to the adequacy of the then current primary NC>2 standard and potential alternative
8 standards for consideration (U.S. EPA. 2008^.' CASAC discussed the final version of
9 the 2008 REA, with an emphasis on the policy assessment chapter during a public
10 teleconference on December 5, 2008 (73 FR 66895, November 12, 2008). Following that
11 teleconference, CASAC offered comments and advice on the primary NO2 standard in a
12 letter to the Administrator (Samet. 2008).
13 After considering an integrative synthesis of the body of evidence on human health
14 effects associated with the presence of NO2 in the air and the exposure and risk
15 information, the Administrator determined that the then-existing primary NC>2 NAAQS,
16 based on an annual arithmetic average, was not sufficient to protect the public health
17 from the array of effects that could occur following short-term exposures to ambient NC>2.
18 In so doing, the Administrator particularly noted the potential for adverse health effects to
19 occur following exposures to elevated NO2 concentrations that can occur around major
20 roads (75 FR 6482). In a notice published in the Federal Register on July 15, 2009, the
21 EPA proposed to supplement the existing primary annual NC>2 standard by establishing a
22 new short-term standard (74 FR 34404). In a notice published in the Federal Register on
23 February 9, 2010, the EPA finalized a new short-term standard with a level of 100 ppb,
24 based on the 3-year average of the 98th percentile of the annual distribution of daily
25 maximum 1-hour concentrations. The EPA also retained the existing primary annual NC>2
26 standard with a level of 53 ppb, annual average (75 FR 6474). The EPA's final decision
27 included consideration of CASAC (2009) and public comments on the proposed rule. The
28 EPA's final rule was upheld against challenges in a decision issued by the U.S. Court of
29 Appeals for the District of Columbia Circuit on July 17, 2012.2
30 Revisions to the NAAQS were accompanied by revisions to the data handling
31 procedures, the ambient air monitoring and reporting requirements, and the Air Quality
1 Subsequent to the completion of the 2008 REA, EPA Administrator Jackson called for additional key changes to
the NAAQS review process including reinstating a policy assessment document that contains staff analysis of the
scientific bases for alternative policy options for consideration by senior Agency management prior to rulemaking
(Jackson. 2009). A Policy Assessment will be developed for the current review as discussed in Chapter 7 of the 2014
Integrated Review Plan for the Primary National Ambient Air Quality Standards for Nitrogen Dioxide (U.S. EPA.
2014).
2 See American Petroleum Institute v. EPA, 684 F. 3d 1342 (D.C. Cir. 2012).
January 2015 Ixxiii DRAFT: Do Not Cite or Quote
-------
1 Index (AQI).: One aspect of the new monitoring network requirements included
2 requirements for states to locate monitors near heavily trafficked roadways in large urban
3 areas and in other locations where maximum NO2 concentrations can occur. Subsequent
4 to the 2010 rulemaking, the EPA revised the deadlines by which the near-road monitors
5 are to be operational in order to implement a phased deployment approach (78 FR 16184,
6 March 14, 2013). The near-road NO2 monitors will become operational between
7 January 1, 2014 and January 1, 2017.
1 The current federal regulatory measurement methods for NCh are specified in 40 Code of Federal Regulations
(CFR) part 50, Appendix F and 40 CFR part 53. Consideration of ambient air measurements with regard to judging
attainment of the standards is specified in 40 CFR part 50, Appendix S. The NCh monitoring network requirements
are specified in 40 CFR part 58, Appendix D, Section 4.3. The EPA revised the AQI for NCh to be consistent with
the revised primary NCh NAAQS as specified in 40 CFR part 58, Appendix G. Guidance on the approach for
implementation of the new standards was described in the Federal Register notices for the proposed and final rules
(74 FR 34404; 75 FR 6474).
January 2015 Ixxiv DRAFT: Do Not Cite or Quote
-------
References for Preface
CAA (Clean Air Act). (1990a). Clean Air Act, as amended by Pub. L. No. 101-549, section 108: Air
quality criteria and control techniques, http://www.law.cornell.edu/uscode/text/42/7408
CAA (Clean Air Act). (1990b). Clean Air Act, as amended by Pub. L. No. 101-549, section 109: National
primary and secondary ambient air quality standards, 42 USC 7408.
http://www.epa.gov/air/caa/titlel. html#ia
CAA (Clean Air Act). (2005). Clean Air Act, section 302: Definitions.
http://www.gpo.gov/fdsvs/pkg/USCODE-2005-title42/pdf/USCODE-2005-title42-chap85-subchapIII-
sec7602.pdf
CASAC (Clean Air Scientific Advisory Committee). (2009). Comments and recommendations
concerning EPA's proposed rule for the revision of the National Ambient Air Quality Standards
(NAAQS) for nitrogen dioxide (unsigned) [EPAReport]. (EPA-CASAC-09-014). Washington, D.C.:
U.S. Environmental Protection Agency.
Jackson. LP (2009). [Letter from EPA Administrator Lisa P. Jackson to Dr. Jonathan M. Samet. Subject:
Changes to the NAAQS review process]. Available online at
http://vosemite.epa.gov/sab/sabproduct.nsf/WebCASAC/Jackson%2005-21-
09/$File/NAAQS%20Letter%20to%20CASAC%20Chair-Mav%202009.pdf
Peacock. M. (2006). Memorandum from Marcus Peacock, Deputy Administrator of the U.S. EPA, to Dr.
George Gray, Assistant Administrator of ORD, and Bill Wehrum, Acting Assistant Administrator of
OAR: Process for reviewing National Ambient Air Quality Standards. Washington, D.C.: U.S.
Environmental Protection Agency.
Peacock. MC. (2008). Letter from Marcus C. Peacock to Rogene Henderson (regarding CASAC
comments on application of new NAAQS process). Peacock, MC.
Samet JM. (2008). Letter from Dr. Jonathan M. Samet to The Honorable Stephen L. Johnson. Subject:
Clean Air Scientific Advisory Committees (CASAC) peer review of draft chapter 8 of EPAs risk and
exposure assessment to support the review of the NO2 primary National Ambient Air Quality
Standard. Samet, JM.
U.S. EPA (U.S. Environmental Protection Agency). (1971). Air quality criteria for nitrogen oxides [EPA
Report]. (AP-84). Washington DC: U.S. Environmental Protection Agency, Air Pollution Control
Office. http://nepis.epa.gov/Exe/ZyPURL.cgi?Dockev=20013K3B.txt
U.S. EPA (U.S. Environmental Protection Agency). (1982). Air quality criteria for oxides of nitrogen
[EPA Report]. (EPA/600/8-82-026). Research Triangle Park, NC.
U.S. EPA (U.S. Environmental Protection Agency). (1993). Air quality criteria for oxides of nitrogen, vol.
1-3 [EPAReport]. (EPA/600/8-91/049aF-cF). Research Triangle Park, NC: U.S. Environmental
Protection Agency, Environmental Criteria and Assessment Office.
http://cfpub.epa. gov/ncea/cfm/recordisplav.cfm?deid=40179
U.S. EPA (U.S. Environmental Protection Agency). (1995). Review of the national ambient air quality
standards for nitrogen dioxide: Assessment of scientific and technical information [EPA Report].
(EPA/452/R-95/005). Research Triangle Park, NC: U.S. Environmental Protection Agency, Office of
Air Quality, Planning and Standards. http://nepis.epa.gov/exe/ZyPURL.cgi?Dockev=00002UBE.txt
U.S. EPA (U.S. Environmental Protection Agency). (2007a). Integrated review plan for the primary
national ambient air quality standard for nitrogen dioxide. Research Triangle Park, NC: U.S.
Environmental Protection Agency, National Center for Environmental Assessment.
http://www.epa.gov/ttn/naaqs/standards/nox/data/20070823 nox review plan final.pdf
January 2015 Ixxv DRAFT: Do Not Cite or Quote
-------
U.S. EPA (U.S. Environmental Protection Agency). (2007b). Nitrogen dioxide health assessment plan:
Scope and methods for exposure and risk assessment. Draft [EPA Report]. Research Triangle Park,
NC: U.S. Environmental Protection Agency, Office of Air Quality Planning and Standards.
http://www.epa.gov/ttn/naaqs/standards/nox/data/20070927 risk exposure scope.pdf
U.S. EPA (U.S. Environmental Protection Agency). (2008a). Integrated science assessment for oxides of
nitrogen Health criteria [EPA Report]. (EPA/600/R-08/071). Research Triangle Park, NC: U.S.
Environmental Protection Agency, National Center for Environmental Assessment.
http://cfpub.epa.gov/ncea/cfm/recordisplav.cfm7de id=194645
U.S. EPA (U.S. Environmental Protection Agency). (2008b). Risk and exposure assessment to support the
review of the NO2 primary national ambient air quality standard [EPA Report]. (EPA-452/R-08-008a).
Research Triangle Park, NC: U.S. Environmental Protection Agency, Office of Air Quality Planning
and Standards, http://www.epa.gov/ttn/naaqs/standards/nox/data/20081121 NO2REA final.pdf
U.S. EPA (U.S. Environmental Protection Agency). (2013). Notice of workshop and call for information
on integrated science assessment for oxides of nitrogen and oxides of sulfur. Fed Reg 78: 53452-
53454.
U.S. EPA (U.S. Environmental Protection Agency). (2014). Integrated review plan for the primary
national ambient air quality standards for nitrogen dioxide [EPA Report]. (EPA-452/R-14/003).
Research Triangle Park, NC: U.S. Environmental Protection Agency, National Center for
Environmental Assessment.
http://www.epa.gov/ttn/naaqs/standards/nox/data/201406finalirpprimarvno2.pdf
January 2015 Ixxvi DRAFT: Do Not Cite or Quote
-------
EXECUTIVE SUMMARY
Purpose and Scope of the Integrated Science Assessment
1 This Integrated Science Assessment (ISA) is a comprehensive evaluation and synthesis of
2 policy-relevant science aimed at characterizing exposures to ambient oxides of nitrogen
3 and their relationships with health effects. Thus, this ISA serves as the scientific
4 foundation for the review of the primary (health-based) National Ambient Air Quality
5 Standards (NAAQS) for nitrogen dioxide (NCh).1 NO2 is the indicator for gaseous oxides
6 of nitrogen (i.e., oxidized nitrogen compounds), which also include nitric oxide and gases
7 produced from reactions involving NCh and nitric oxide (Figure 2-1, Section 2.2).2-3 In
8 2010, the U.S. Environmental Protection Agency (EPA) retained the NAAQS of 53 parts
9 per billion (ppb) annual average concentration to protect against health effects potentially
10 related to long-term NCh exposures. In addition, EPA set a new 1-hour NAAQS at a level
11 of 100 ppb, based on the 3-year average of each year's 98th percentile of the highest
12 daily 1-hour concentration. The 1-hour NAAQS was set to protect against respiratory
13 effects related to short-term NCh exposures in populations potentially at increased risk,
14 such as people with asthma or people who spend time on or near high-traffic roads. EPA
15 also set requirements for a network of monitors to measure NC>2 near high-traffic roads,
16 one of the places where the highest concentrations are expected to occur.
17 This ISA updates the 2008 ISA for Oxides of Nitrogen (U.S. EPA. 2008) with studies and
18 reports published from January 2008 through August 2014. EPA conducted searches to
19 identify peer-reviewed literature on relevant topics such as health effects, atmospheric
20 chemistry, ambient concentrations, and exposure. The Clean Air Scientific Advisory
21 Committee (a formal independent panel of scientific experts) and the public also
22 recommended studies and reports. To fully describe the state of available science, EPA
23 also identified relevant studies from previous assessments to include in this ISA.
24 As in the 2008 ISA, this ISA determines the causality of relationships with health effects
25 only for NCh (Chapter 5 and Chapter 6). Key to interpreting the health effects evidence is
26 understanding the sources, chemistry, and distribution of NC>2 in the ambient air
27 (Chapter 2) that influence exposure (Chapter 3). the uptake of inhaled NC>2 in the
28 respiratory tract, and subsequent biological mechanisms that may be affected (Chapter 4).
1 The ecological effects of oxides of nitrogen are being considered in a separate assessment as part of the review of
the secondary (welfare-based) NAAQS for NO2 and sulfur dioxide (U.S. EPA. 2013).
2 Total oxides of nitrogen also include several paniculate species such as nitrites. Section 108(c) of the Clean Air
Act, 42 U. S.C. § 7408(c) specifies that criteria for oxides of nitrogen include consideration of nitric and nitrous
acids, nitrites, nitrates, nitrosamines, and other derivatives of oxides of nitrogen. Health effects associated with the
paniculate species are addressed in the review of the NAAQS for paniculate matter (U.S. EPA. 2009).
3 The blue electronic links can be used to navigate to other parts of this ISA and to information on cited references .
January 2015 Ixxvii DRAFT: Do Not Cite or Quote
-------
1 Further, the ISA aims to characterize the independent effect of NCh exposure on health
2 rather than its role as just a marker for other air pollutants. The ISA also informs
3 policy-relevant issues (Section 1.6), such as (1) exposure durations and patterns
4 associated with health effects; (2) concentration-response relationship(s), including
5 evidence of potential thresholds for effects; and (3) populations or lifestages at increased
6 risk for health effects related to NCh exposure (Chapter 7).
Sources and Human Exposure to Nitrogen Dioxide
7 A main objective of the ISA is to characterize health effects related to ambient NCh
8 exposure. This requires understanding what factors affect exposure to ambient NCh and
9 the ability to estimate exposure accurately and accounting for the influence of factors that
10 are related to NCh exposure, such as other pollutants and demographic characteristics.
11 For the U.S. as a whole and for major cities, motor vehicle emissions are the largest
12 single source of NCh in the ambient air (Section 2.3.1. Figure 2-3). Electric power plants,
13 industrial facilities, other forms of transportation, soil, and wildfires also can contribute
14 considerably to ambient NCh concentrations on a national scale and to differences in
15 concentrations and population exposures among locations.
16 Because many sources of NCh are ubiquitous, there is widespread potential for exposure
17 to NCh. However, given that motor vehicles are a major source, air concentrations of NCh
18 can be highly variable across neighborhoods (Section 2.5.2). depending on distance to
19 roads. NCh concentrations tend to decrease over a distance of 200-500 meters from the
20 road and can be 30 to 100% higher within 10-20 meters of a road than at locations
21 farther away (Section 2.5.3). The first year of data from a new near-road monitoring
22 network for a small group of U.S. cities show that 1-hour NCh concentrations tend to be
23 higher near roads than at most other sites within a city (Table 2-8. Section 2.5.3.2). But,
24 peak NCh concentrations are not always higher near roads, indicating that in addition to
25 distance from road, factors such as other local sources, season, wind direction, chemical
26 reactions with ozone in the air (Figure 2-1). and physical features of the environment
27 (Sections 2.2 and 2.5.3) also affect the distribution of ambient NCh concentrations.
28 Because ambient NCh concentrations show variability among geographic regions, within
29 communities, and over time, there is potential for large variation in ambient NCh
30 exposure among people. Also contributing to variation in ambient NCh exposure are the
31 differences in the outdoor and indoor locations where people spend time and the amount
32 of time spent in those locations (Sections 3.4.1 and 3.4.3; Figure 3-3). NCh
33 concentrations vary by the type of location, including inside vehicles and buildings
34 (Figure 3-1). and the ventilation of buildings can affect the amount of ambient NCh that
35 penetrates indoors (Section 3.4.3.3). Therefore, understanding the extent to which the
January 2015 Ixxviii DRAFT: Do Not Cite or Quote
-------
1 methods used to estimate exposure adequately account for variation in ambient
2 concentrations across locations and people's activity patterns is essential to characterize
3 relationships between ambient NC>2 exposure and health effects.
4 In this ISA, health effects are examined largely in relation to ambient NC>2 concentrations
5 measured at community monitoring sites. These monitors do not cover all locations
6 where people live or spend their time and are not sited to capture the variability in NO2
7 concentrations observed within cities, including near roads. Thus, NC>2 measured at these
8 sites have some error in representing people's actual exposures. This error may be
9 reflected in the wide range of relationships observed between total personal NO2
10 exposure and ambient concentrations averaged over periods up to 1 week (Section 3.4.2).
11 Such relationships are not well characterized for exposure periods of months to years.
12 These uncertainties do not necessarily mean that ambient NC>2 concentrations are poor
13 measures of personal exposure because variation among people in indoor or in-vehicle
14 exposures and activity patterns may obscure relationships between ambient exposure and
15 concentrations. Compared with NC>2 concentrations at community monitors, there may be
16 more confidence in exposure metrics that account for local variability in NC>2
17 concentrations and people's activity patterns. These metrics can include short-term
18 personal, home, and school NCh measurements and long-term NC>2 exposure estimated at
19 people's homes with models and are examined in some recent health effect studies.
20 Error in estimating exposure can impact associations observed between ambient NC>2
21 concentrations and health effects in various ways. In studies of short-term exposure that
22 examine changes in NC>2 over time (e.g., day to day), NCh from community monitors has
23 shown lower magnitude associations with health effects (Section 3.4.5) compared with
24 NO2 measured at people's locations and/or more uncertainty in the association. In studies
25 of long-term exposure that compare people in locations that vary in ambient NC>2
26 concentrations, NCh from community monitors has shown both smaller and larger
27 associations with health effects compared with NC>2 concentrations estimated for people's
28 locations. The impact on health effect associations of using NC>2 concentrations at
29 community sites to represent near-road exposures is not clear. Given the impact of
30 exposure error, this ISA draws conclusions about health effects related to NO2 exposure
31 by considering the availability of results for NC>2 measured at community monitoring
32 sites versus other locations where people live or spend time and considering how well the
33 various methods represent differences in exposure over time or across locations.
34 The important contribution of motor vehicles to ambient NC>2 concentrations not only has
35 implications for estimating NCh exposure but also indicates the need to consider other
36 traffic-related pollutants. NC>2 concentrations often are moderately to highly correlated
January 2015 Ixxix DRAFT: Do Not Cite or Quote
-------
1 with pollutants also emitted by motor vehicles, such as PlVfc s,1 UFP (see footnote 1),
2 elemental or black carbon (EC/BC), and carbon monoxide (Figure 3-6. Section 3.4.4.1).
3 Traffic-related pollutants show effects on many of the same biological processes and
4 health outcomes (Table 5-1). Thus, in characterizing relationships of NCh with health
5 effects, the ISA evaluates the extent to which an effect of NO2 can be distinguished from
6 that of other traffic-related pollutants. Important lines of evidence include epidemiologic
7 studies that statistically adjust the NO2 association for another pollutant and experimental
8 studies that inform the direct effect of NC>2 exposure on health outcomes and biological
9 mechanisms.
Health Effects of Nitrogen Dioxide Exposure
10 In this ISA, information on NC>2 exposure, the potential influence of other traffic-related
11 pollutants, and health effects from epidemiologic, controlled human exposure, and
12 toxicological studies is integrated to form conclusions about the causal nature of
13 relationships between NC>2 exposure and health effects. Health effects examined in
14 relation to the full range of NC>2 concentrations relevant to ambient conditions are
15 considered. Based on peak concentrations (Section 2.5) and the ISA definition that
16 ambient-relevant exposures be within one to two orders of magnitude of current
17 conditions (Preamble. Section 5.c). NC>2 concentrations up to 5,000 ppb2 are defined to be
18 ambient relevant. A consistent and transparent framework (Preamble. Table II) is applied
19 to classify the health effects evidence according to a five-level hierarchy:
1) Causal relationship
2) Likely to be a causal relationship
3) Suggestive, but not sufficient, to infer a causal relationship
4) Inadequate to infer a causal relationship
5) Not likely to be a causal relationship
20 The conclusions presented in Table ES-1 are informed by recent findings and whether
21 recent findings integrated with information from the 2008 ISA for Oxides of Nitrogen
22 (U.S. EPA. 2008) support a change in conclusion. Important considerations include
23 judgments of error and uncertainty in the collective body of available studies; the
24 consistency of findings integrated across epidemiologic, controlled human exposure, and
25 toxicological studies to inform an independent effect of NO2 exposure and potential
1 PM25: In general terms, paniculate matter with an aerodynamic diameter less than or equal to a nominal 2.5 ^m, a
measure of fine particles. UFP: Definitions vary but often refer to particles with an aerodynamic diameter less or
equal to a nominal 0.1 ^m, a measure of ultrafme particles.
2 The 5,000-ppb upper limit applies mostly to animal toxicological studies and also a few controlled human
exposure studies. Experimental studies examining NCh exposures greater than 5,000 ppb were included if they
provided information on the uptake of NCh in the respiratory tract or on potential biological mechanisms.
January 2015 Ixxx DRAFT: Do Not Cite or Quote
-------
1
2
o
6
4
underlying biological mechanisms; consistency in epidemiologic evidence across various
methods used to estimate NO2 exposure; and examination in epidemiologic studies of the
potential influence of other traffic-related pollutants and other factors that could bias
associations observed with NO2 exposure (Section 5.1.2).
Table ES-1 Causal determinations for relationships between nitrogen dioxide
(NO2) exposure and health effects from the 2008 and current
Integrated Science Assessment (ISA) for Oxides of Nitrogen.
Exposure Duration and
Health Effect Category3
Causal Determination13
2008 ISA
Current Draft ISA
Short-term NO2 Exposure (minutes up to 1 month)
Respiratory effects
Section 5.2. Table 5-45
Sufficient to infer a likely causal
relationship
Causal relationship
Cardiovascular and related
metabolic effects0
Section 5.3, Table 5-58
Inadequate to infer the presence or
absence of a causal relationship
Suggestive, but not sufficient, to infer a
causal relationship
Total mortality
Section 5.4. Table 5-63
Suggestive, but not sufficient, to
infer a causal relationship
Suggestive, but not sufficient, to infer a
causal relationship
Long-term NO2 Exposure (more than 1 month to years)
Respiratory effects
Section 6.2, Table 6-5
Suggestive, but not sufficient, to
infer a causal relationship
Likely to be a causal relationship
Cardiovascular and related
metabolic effects0
Section 6.3. Table 6-11
Inadequate to infer the presence or
absence of a causal relationship
Suggestive, but not sufficient, to infer a
causal relationship
Reproductive and
developmental effects0
Sections 6.4.2. 6.4.3. and 6.4.4.
Table 6-14
Inadequate to infer the presence or
absence of a causal relationship
Fertility, Reproduction, and Pregnancy:
Inadequate to infer a causal relationship
Birth Outcomes:
Suggestive, but not sufficient, to infer a
causal relationship
Postnatal Development:
Inadequate to infer a causal relationship
Total mortality
Section 6.5. Table 6-18
Inadequate to infer the presence or
absence of a causal relationship
Suggestive, but not sufficient, to infer a
causal relationship
Cancer
Section 6.6, Table 6-20
Inadequate to infer the presence or
absence of a causal relationship
Suggestive, but not sufficient, to infer a
causal relationship
aAn array of outcomes is evaluated as part of a broad health effect category: physiological measures (e.g., airway
responsiveness), clinical outcomes (e.g., hospital admissions), and cause-specific mortality. Total mortality includes all
nonaccidental causes of mortality and is informed by findings for the spectrum of morbidity effects (e.g., respiratory,
cardiovascular) that can lead to mortality. The sections and tables referenced include a detailed discussion of the evidence that
supports the causal determinations and the NO2 concentrations with which health effects have been associated.
bSince the 2008 ISA for Oxides of Nitrogen, the phrasing of causal determinations has changed slightly, and the weight of
evidence that describes each level in the hierarchy of the causal framework has been more explicitly characterized.
°ln the current ISA, the cardiovascular effects category is expanded to include related metabolic effects. Reproductive and
developmental effects are separated into smaller subcategories of outcomes based on varied underlying biological processes
and exposure patterns over different lifestages.
January 2015
Ixxxi
DRAFT: Do Not Cite or Quote
-------
Short-term Nitrogen Dioxide Exposure and Respiratory Effects
1 A causal relationship is determined for short-term NC>2 exposure and respiratory effects.
2 The conclusion is strengthened from the 2008 ISA for Oxides of Nitrogen from likely to
3 be a causal relationship (Table ES-1) based on the evidence demonstrating that NO2
4 exposure can trigger asthma attacks. There is some evidence relating short-term NO2
5 exposure to chronic obstructive pulmonary disease, respiratory infection, respiratory
6 effects in healthy populations, and respiratory mortality but uncertainty as to whether the
7 effects of NO2 exposure are independent of other traffic-related pollutants (Table 5-45).
8 Supporting a relationship with asthma attacks, epidemiologic studies in diverse locations
9 consistently show that short-term increases in ambient NO2 concentration are associated
10 with increases in hospital admissions and emergency department visits for asthma,
11 increases in respiratory symptoms and airway inflammation in people with asthma, and
12 decreases in lung function in children with asthma (Section 5.2.9). These associations are
13 found not only with community-average ambient NO2 concentrations but also with
14 personal NO2 and NO2 measured outside children's schools and inside their homes
15 (Sections 5.2.9.3 and 5.2.9.6). Correlations between NO2 and other traffic-related
16 pollutants are weaker for total personal exposures than for ambient concentrations, and
17 the same may be true for indoor exposures. So, associations with personal and indoor
18 NO2 may be less influenced by pollutants that are related to outdoor NO2. Further, studies
19 that measured pollutants at peoples' locations tend to show that NO2 remains associated
20 with respiratory effects after accounting for the effect of another traffic-related pollutant
21 such as PM2 5 or (examined in fewer studies) EC/BC, metals, or UFP (Figure 5-16 and
22 Figure 5-17). The key evidence for an independent effect of NO2 are the controlled
23 human exposure findings for NO2-induced increases in airway responsiveness and
24 allergic inflammation, which are hallmarks of asthma attacks (Figure ES-1). An effect of
25 short-term NO2 exposure on asthma attacks also is plausible given that inhaled NO2
26 reacts with substances such as antioxidants in the fluid lining the lung (Section 4.2.2) to
27 form reactive species. The production of such reactive species is an early event involved
28 in increasing airway responsiveness and allergic responses (Figure ES-1).
29 The 2008 ISA described much of the same evidence and determined a likely to be causal
30 relationship, citing uncertainty as to whether results for NO2 in epidemiologic studies
31 reflected the effects of other traffic-related pollutants. The 2008 ISA did not explicitly
32 evaluate the extent to which various lines of evidence supported effects on asthma
33 attacks. In this ISA, the determination of a causal relationship is not just based on new
34 evidence but on an evaluation and integration of findings related to asthma attacks. The
35 epidemiologic evidence for asthma attacks and controlled human exposure study findings
36 for increased airway responsiveness and allergic inflammation together are sufficient to
January 2015 Ixxxii DRAFT: Do Not Cite or Quote
-------
rule out the influence of other traffic-related pollutants with reasonable confidence and
demonstrate an independent effect of short-term NO2 exposure on respiratory effects.
* 1 Airway
responsiveness
1triggerJAsthma
attack
f Airway
responsiveness
Note: Adapted from Figure 4-1 and Figure 4-2 (Section 4.3.5). White boxes and solid arrows describe pathways well supported
by available evidence. Gray boxes and dotted arrows describe potential pathways for which evidence is limited or inconsistent.
Figure ES-1.
Biological pathways for relationships of short-term and long-term
nitrogen dioxide (NO2) exposure with asthma.
3
4
5
6
7
8
9
10
11
12
13
14
15
Long-term Nitrogen Dioxide Exposure and Respiratory Effects
There is likely to be a causal relationship between long-term NO2 exposure and
respiratory effects (Section 6.2.9) based on the evidence for development of asthma. The
conclusion is strengthened from the 2008 ISA (Table ES-1) because where previous
findings were inconsistent, recent epidemiologic studies consistently observe NC>2-related
increases in asthma development in children who are followed over time. There is more
uncertainty about relationships with respiratory effects such as development of allergy
and respiratory infection (Table 6-5). Asthma development is associated not only with
community-level ambient NC>2 concentrations but also with ambient NC>2 exposure
estimated at children's homes with models that well capture the spatial variability in
communities. Associations between NO2 and asthma development are independent of
factors such as socioeconomic status and exposure to smoking, but the potential influence
of other traffic-related pollutants is not well studied. This uncertainty also applies to the
findings for NC^-related decreases in lung function and lung development in children.
January 2015
Ixxxiii
DRAFT: Do Not Cite or Quote
-------
1 There is some support for an independent effect of NO2 on asthma development provided
2 by findings of increased airway responsiveness in rodents (Figure ES-1). Also, evidence
3 relating short-term NCh exposure to airway inflammation in epidemiologic studies of
4 healthy people and allergic responses in experimental studies of rodents and healthy
5 people indicates that repeated short-term NO2 exposure could lead to the development of
6 asthma. Together, the epidemiologic and experimental evidence for asthma development
7 supports a relationship between long-term NC>2 exposure and respiratory effects, but
8 because experimental evidence is limited, there remains some uncertainty about the
9 potential influence of other traffic-related pollutants in the epidemiologic evidence.
Nitrogen Dioxide Exposure and Other Health Effects
10 There is more uncertainty about relationships of NC>2 exposure with health effects outside
11 of the respiratory system. NC>2 itself is unlikely to enter the bloodstream, and reactions
12 caused by ambient-relevant concentrations of NCh in the airways do not clearly affect
13 concentrations of reactive compounds, such as nitrite, in the blood. Only a few results
14 suggest that compounds that can cause inflammation or oxidative stress may enter the
15 blood from the respiratory tract in response to NO2 exposure (Section 4.3.2.9). This
16 uncertainty about the effects of NO2 exposure on underlying biological mechanisms is
17 common to nonrespiratory health effects.
18 For short-term and long-term NO2 exposure, evidence is suggestive, but not sufficient, to
19 infer a causal relationship with cardiovascular and related metabolic effects, total
20 mortality, birth outcomes, and cancer (Table ES-1). For short-term NO2 exposure, recent
21 epidemiologic studies continue to show associations with total mortality and add support
22 for cardiovascular and related metabolic effects by indicating a possible effect on
23 triggering heart attacks. Where there was little previous support, increases in recent
24 epidemiologic evidence result in strengthening conclusions for total mortality and cancer
25 related to long-term NO2 exposure. New epidemiologic findings for heart disease and
26 diabetes and reduced fetal growth point to possible relationships of long-term NO2
27 exposure with health effect categories new to this ISA: cardiovascular and related
28 metabolic effects and birth outcomes. For fertility, reproduction, and pregnancy, as well
29 as postnatal development, evidence is inadequate to infer a causal relationship with
30 long-term NO2 exposure (Table ES-1) because neither epidemiologic nor toxicological
31 studies consistently show effects. For all nonrespiratory effects, epidemiologic studies do
32 not account for the potential influence of other traffic-related pollutants, which combined
33 with the few or inconclusive results from controlled human exposure or toxicological
34 studies, produces large uncertainty as to whether short-term or long-term NO2 exposure
35 has independent relationships with health effects outside of the respiratory system.
January 2015 Ixxxiv DRAFT: Do Not Cite or Quote
-------
Policy-Relevant Considerations for Health Effects Associated
with Nitrogen Dioxide Exposure
1 Multiple durations of short-term and long-term NC>2 exposure are observed to be
2 associated with health effects (Section 1.6.1). For short-term exposure, asthma-related
3 effects are associated with total personal NO2 exposure and NO2 measured at children's
4 schools or community monitors averaged over 1 to 5 days. These associations are
5 observed with both daily average and the daily highest 1-hour NCh concentration. No
6 particular duration of exposure shows a stronger effect. Controlled human studies
7 demonstrate increased airway inflammation and airway responsiveness in adults with
8 asthma following NC>2 exposures of 15 to 60 minutes. These results support the
9 epidemiologic evidence indicating that NC>2 exposures of 2 or 5 hours near high-traffic
10 roads can lead to respiratory effects in adults with and without asthma.
11 For long-term exposure, asthma development in children is associated with estimates of
12 residential ambient NC>2 exposure averaged over 1 to 10 years, representing various
13 developmental periods, such as infancy, childhood, and lifetime. The pattern of NO2
14 exposure underlying these associations is not well characterized, but some evidence from
15 experimental studies in humans and rodents suggests that repeated exposure over many
16 days or weeks can induce allergic responses, which are involved in asthma development.
17 Information on the shape of the NC>2 concentration-health effect relationship is limited
18 mostly to epidemiologic studies. A few results show that asthma emergency department
19 visits and diagnosis of asthma, respectively, increase with increasing short-term and long-
20 term average ambient NC>2 concentrations (Section 1.6.3). As analyzed for asthma
21 emergency department visits, the association with the daily highest 1-hour NC>2
22 concentrations is present at low concentrations. But, uncertainty in the relationship is
23 noted at concentrations well below the level of the current 1-hour NAAQS.
24 Health effects related to NO2 exposure potentially have a large public health impact.
25 Many people in the U.S. live, work, or spend time near roads and may have high
26 exposures to NCh. Higher NCh exposure also is indicated for urban, low socioeconomic
27 status, and nonwhite populations. Further, people with asthma, children (especially ages
28 0-14 years), and older adults (especially ages 65 years and older) are identified as being
29 at increased risk of NCh-related health effects (Chapter 7). Evidence does not clearly
30 identify other at-risk populations in terms of other diseases or behavioral, genetic, or
31 sociodemographic factors. Short-term and long-term NC>2 exposure is linked to clinically
32 relevant increases in airway responsiveness, emergency department visits and hospital
33 admissions for asthma, and development of asthma, which can have a large impact on
34 public health. Given that asthma is the leading chronic illness and the leading cause of
January 2015 Ixxxv DRAFT: Do Not Cite or Quote
-------
1 missed school days and hospital admissions among U.S. children, NO2-related asthma
2 attacks and asthma development have the potential to affect children's overall well-being.
Summary of Major Findings
3 Expanding on findings from the 2008 ISA for Oxides of Nitrogen (U.S. EPA. 2008).
4 recent epidemiologic studies show associations of short-term and long-term NO2
5 exposure with an array of health effects. However, except for respiratory effects, there
6 remains large uncertainty about whether NO2 exposure has an effect that is independent
7 of other traffic-related pollutants. As in the 2008 ISA, recent information shows that
8 motor vehicle emissions are the largest single source of NO2 in the air and that NO2
9 concentrations tend to be variable across locations, decreasing with increasing distance
10 from roads. Information to assess whether NO2 exposure estimates adequately represent
11 the variability in ambient NO2 concentrations and people's activity patterns varies among
12 the health effects evaluated in this ISA. The major findings from this ISA about NO2
13 exposure and health effects and related uncertainties are summarized below.
14 • Evidence for asthma attacks supports a causal relationship between short-term
15 NO2 exposure and respiratory effects, and evidence for development of
16 asthma supports a likely to be causal relationship between long-term NO2
17 exposure and respiratory effects. These are stronger conclusions than those
18 determined in the 2008 ISA for Oxides of Nitrogen.
19 • Evidence is suggestive, but not sufficient, to infer a causal relationship for
20 short-term or long-term NO2 exposure with cardiovascular and related
21 metabolic effects, birth outcomes, total mortality, and cancer. These also are
22 stronger conclusions than those determined in the 2008 ISA for Oxides of
23 Nitrogen.
24 • Recent and previous findings combined indicate that people with asthma,
25 children, and older adults are at increased risk for NO2-related health effects.
26 • There is continued evidence for increased NO2 exposure among people living
27 or spending time near or on roads, low socioeconomic status populations, and
28 nonwhite populations.
29 • Data from the U.S. near-road network are starting to become available and
30 may help address gaps in understanding of the:
31 o variability in NO2 concentrations within cities and NO2 exposure in the
32 population,
33 o health effects associated with NO2 exposures occurring near roads, and
34 o potential for other traffic-related pollutants to influence associations
35 observed between NO2 exposure and health effects.
36 • Epidemiologic studies continue to examine ambient NO2 concentrations at
37 community monitoring sites, which represent exposure with error.
January 2015 Ixxxvi DRAFT: Do Not Cite or Quote
-------
1 o Error in representing changes in short-term exposure over time
2 (e.g., day-to-day) tend to decrease the magnitude of association with
3 health effects and/or increase the uncertainty of the estimates.
4 o Error in representing differences among people in long-term NO2
5 exposure can decrease health effect associations and/or increase the
6 uncertainty of the estimates but sometimes can overestimate associations.
7 o NO2 concentrations measured at people's locations may better represent
8 exposure and improve understanding of relationships with health effects.
9 "An independent effect of short-term NO2 exposure on respiratory effects is
10 supported by additional epidemiologic evidence for ambient NO2 measured at
11 people's locations and personal NO2 exposure and associations that are
12 independent of another traffic-related pollutant integrated with results from
13 previous controlled human exposure studies describing biological pathways.
14 • For long-term NO2 exposure and respiratory effects, there is new supporting
15 epidemiologic evidence for NO2 exposure estimated at or near children's
16 homes. Epidemiologic studies did not examine the influence of other traffic-
17 related pollutants, but limited findings from previous experimental studies
18 provide some support for an independent effect of NO2 exposure.
19 • For nonrespiratory effects, there is continued or new epidemiologic evidence,
20 but epidemiologic, controlled human exposure, and toxicological studies
21 have not sufficiently addressed the uncertainty as to whether NO2 exposure
22 has independent effects.
23 • In limited investigation, the relationship between the daily highest 1-hour
24 NO2 concentration and emergency department visits for asthma is present at
25 concentrations well below the current 1-hour NAAQS.
January 2015 Ixxxvii DRAFT: Do Not Cite or Quote
-------
References for Executive Summary
U.S. EPA (U.S. Environmental Protection Agency). (2008). Integrated science assessment for oxides of
nitrogen Health criteria [EPA Report]. (EPA/600/R-08/071). Research Triangle Park, NC: U.S.
Environmental Protection Agency, National Center for Environmental Assessment.
http://cfpub.epa.gov/ncea/cfm/recordisplav.cfm7de id=194645
U.S. EPA (U.S. Environmental Protection Agency). (2009). Integrated science assessment for particulate
matter [EPA Report]. (EPA/600/R-08/139F). Research Triangle Park, NC: U.S. Environmental
Protection Agency, National Center for Environmental Assessment.
http://cfpub.epa.gov/ncea/cfm/recordisplav.cfm7de id=216546
U.S. EPA (U.S. Environmental Protection Agency). (2013). Notice of workshop and call for information
on integrated science assessment for oxides of nitrogen and oxides of sulfur. Fed Reg 78: 53452-
53454.
January 2015 Ixxxviii DRAFT: Do Not Cite or Quote
-------
CHAPTER 1 INTEGRATED SUMMARY
1.1 Purpose and Overview of the Integrated Science Assessment
1 The Integrated Science Assessment (ISA) is a comprehensive evaluation and synthesis of
2 the policy-relevant science "useful in indicating the kind and extent of identifiable effects
3 on public health or welfare which may be expected from the presence of [a] pollutant in
4 ambient air" (CAA. 1990). This ISA communicates critical science judgments of the
5 health criteria for a broad category of gaseous oxides of nitrogen (i.e., oxidized nitrogen
6 compounds) for which nitrogen dioxide (NO2) is the indicator. As such, this ISA serves
7 as the scientific foundation for the review of the current primary (health-based) National
8 Ambient Air Quality Standards (NAAQS) for NO2. Gaseous oxides of nitrogen include
9 NO2, nitric oxide (NO), and their various reaction products (Figure 1-1, Section 2.2):.
10 There also are particulate oxides of nitrogen (e.g., nitrates, nitro-polycyclic aromatic
11 hydrocarbons)2, which were considered in the most recent review of the NAAQS for
12 particulate matter (PM) and evaluated in the 2009 ISA for PM (U.S. EPA. 2009). The
13 welfare effects of oxides of nitrogen are being evaluated in a separate assessment
14 conducted as part of the review of the secondary (welfare-based) NAAQS for NO2 and
15 sulfur dioxide [SO2; (U.S. EPA.2013c)1.
16 This ISA evaluates relevant scientific literature published since the 2008 ISA for Oxides
17 of Nitrogen (U.S. EPA. 2008). integrating key information and judgments contained in
18 the 2008 ISA and the 1993 Air Quality Criteria Document for Oxides of Nitrogen
19 (U.S. EPA. 1993). Thus, this ISA updates the state of the science that was available for
20 the 2008 ISA, which informed decisions on the primary NO2 NAAQS in the review
21 completed in 2010. In 2010, the U.S. Environmental Protection Agency (EPA) retained
22 the existing annual average (avg) NO2 NAAQS with a level of 53 parts per billion (ppb)
23 to protect against health effects potentially associated with long-term exposure. EPA
24 established a new 1-hour (h) NAAQS at a level of 100 ppb NO2 based on the 3-year (yr)
25 avg of each year's 98th percentile of 1-h daily maximum (max) concentrations.3 The
26 1-h standard was established to protect against a broad range of respiratory effects
27 associated with short-term exposures in potential at-risk populations such as people with
28 asthma and people who spend time on or near high-traffic roads. In 2010, EPA also set
29 requirements for a monitoring network in urban areas that includes monitors near (within
1 The blue electronic links can be used to navigate to other parts of this ISA and to information on cited references.
2 Section 108(c) of the Clean Air Act, 42 U.S.C. § 7408(c) specifies that criteria for oxides of nitrogen include
consideration of nitric and nitrous acids, nitrites, nitrates, nitrosamines, and other derivatives of oxides of nitrogen,
including multiple gaseous and particulate species.
3 The legislative requirements and history of the NO2 NAAQS are described in detail in the Preface to this ISA.
January 2015 1-1 DRAFT: Do Not Cite or Quote
-------
1 50 meters [m]) of high-traffic roads, one of the locations where the highest NO2
2 concentrations are expected to occur (U.S. EPA. 2010).
3 This review of the primary NO2 NAAQS is guided by several policy-relevant questions
4 that are identified in The Integrated Review Plan for the Primary National Ambient Air
5 Quality Standard for Nitrogen Dioxide (U.S. EPA. 2014). To address these questions and
6 update the scientific judgments in the 2008 ISA, this ISA aims to:
7 • Characterize the evidence for health effects associated with short-term
8 (minutes up to 1 month) and long-term (more than 1 month to years)
9 exposure to oxides of nitrogen by integrating findings across scientific
10 disciplines and across related health outcomes and by considering important
11 uncertainties identified in the interpretation of the scientific evidence,
12 including the role of NO2 within the broader ambient mixture of pollutants.
13 • Inform policy-relevant issues related to quantifying health risks, such as
14 exposure concentrations, durations, and patterns associated with health
15 effects; concentration-response relationships and evidence of thresholds
16 below which effects do not occur; and populations and lifestages potentially
17 with increased risk of health effects related to exposure to oxides of nitrogen.
18 Although the scope of the ISA includes all gaseous oxides of nitrogen, much of the
19 information on the distribution of oxides of nitrogen in the air, human exposure and dose,
20 impact of errors associated with exposure assessment methods, and health effects is for
21 NO2. There is limited information for NO and the sum of NO and NO2 (NOx) as well as
22 large uncertainty in relating health effects to NO or NOx exposure. In the body, NO is
23 produced from nitrates and nitrites derived from diet and enzymatic pathways that are
24 enhanced during inflammation. Ambient NO concentrations generally are in the range of
25 endogenous NO concentrations exhaled from the respiratory tract. It is not clear whether
26 ambient-relevant NO exposures substantially alter endogenous NO production in the
27 respiratory tract or pathways affected by endogenous NO (Section 4.2.3). Thus, the
28 potential for detrimental health effects occurring from ambient-relevant NO exposure is
29 unclear. This lack of evidence leaves NO2 as the component of NOx to consider in
30 evaluating health effects in relation to NOx exposure. Because of the varying ratio of
31 NO2 to NOx across locations, time of day, and season (Section 2.5). NOx may not
32 represent NO2 exposure consistently. The lack of support for ambient-relevant NO
33 exposure to result in potentially detrimental health effects and the exposure measurement
34 error related to using NOx to represent NO2 exposure are the rationale for determining the
35 causal nature of health effects only for NO2 exposure.
36 In addressing policy-relevant questions, this ISA aims to characterize the independent
37 health effects of NO2 exposure, not the role of NO2 as just a marker for a broader mixture
38 of pollutants in the ambient air. The potential influence of other traffic-related pollutants
39 was the main uncertainty in the conclusions drawn in the 2008 ISA Oxides of Nitrogen
January 2015 1-2 DRAFT: Do Not Cite or Quote
-------
1 (U.S. EPA. 2008). As described in this ISA, recent epidemiologic findings together with
2 evidence from previous controlled human exposure studies sufficiently describe a
3 coherent, biologically plausible relationship between short-term NCh exposure and
4 respiratory effects ranging from increased airway responsiveness to increased symptoms,
5 emergency department (ED) visits, and hospital admissions related to asthma
6 exacerbation. Recent epidemiologic studies provide new evidence supporting a
7 relationship of long-term NC>2 exposure with respiratory effects, specifically, the
8 development of asthma in children, and a small body of previous experimental studies
9 provide some indication that NC>2 exposure may have an independent effect. Recent
10 epidemiologic studies continue to suggest that short-term NC>2 exposure may be
11 associated with cardiovascular and related metabolic effects and mortality, and new
12 findings potentially link long-term NC>2 exposure to cardiovascular and related metabolic
13 effects, poorer birth outcomes, mortality, and cancer. However, for nonrespiratory
14 effects, epidemiologic studies have not adequately accounted for effects of other
15 traffic-related pollutants, and findings from experimental studies continue to be limited.
16 The information in the ISA contributing to these findings will serve as the scientific
17 foundation for the review of the current primary 1-hour and annual NC>2 NAAQS.
1.2 Process for Developing Integrated Science Assessments
18 EPA uses a structured and transparent process for evaluating scientific information and
19 determining the causality of relationships between air pollution exposures and health
20 effects (see Preamble). The ISA development process describes approaches for literature
21 searches, criteria for selecting and evaluating relevant studies, and a framework for
22 evaluating the weight of evidence and forming causal determinations. As part of this
23 process, the ISA is reviewed by the Clean Air Scientific Advisory Committee (CASAC),
24 a formal independent panel of scientific experts, and the public. As this ISA informs the
25 review of the primary NC>2 NAAQS, it assesses information relevant to characterizing
26 exposure to gaseous oxides of nitrogen and potential relationships with health effects.
27 Relevant studies include those examining atmospheric chemistry, spatial and temporal
28 trends, and exposure assessment, as well as EPA analyses of air quality and emissions
29 data. Also relevant are epidemiologic, controlled human exposure, and toxicological
30 studies on health effects as well as studies on dosimetry and modes of action.
31 EPA initiated the current review of the primary NAAQS for NC>2 in February 2012 with a
32 call for information from the public (U.S. EPA. 2012). Thereafter, EPA routinely
33 conducted literature searches to identify relevant peer-reviewed studies published since
34 the previous ISA (i.e., from January 2008 through August 2014). Multiple search
35 methods were used (Preamble, Section 2) including searches in databases such as
January 2015 1-3 DRAFT: Do Not Cite or Quote
-------
1 PubMed and Web of Science. Also, recommendations were made by CASAC and the
2 public. EPA identified additional studies considered to be the definitive work on
3 particular topics from previous assessments to include in this ISA. Some studies were
4 judged to be irrelevant (i.e., did not address a topic described in the preceding paragraph)
5 based on title and were excluded. Studies that were judged to be potentially relevant
6 based on review of the abstract or full text and "considered" for inclusion in the ISA are
7 documented and can be found at the Health and Environmental Research Online (HERO)
8 website. The HERO project page for this ISA (http://hero.epa.gov/oxides-of-nitrogen)
9 contains the references that are cited in the ISA, the references that were "considered" for
10 inclusion but not cited, and electronic links to bibliographic information and abstracts.
11 Health effects were considered for evaluation in this ISA if they were examined in
12 previous EPA assessments for oxides of nitrogen or multiple recent studies
13 (e-g-, neurodevelopment). Literature searches identified one or two recent epidemiologic
14 studies each on outcomes such as gastrointestinal effects, bone density, headache, and
15 depression [Supplemental Table Sl-1; (U.S. EPA. 2013d)]. A review of these studies
16 indicated they are similar in design and conducted in areas and populations for which
17 associations between ambient NO2 concentrations and other health effects have been
18 documented. These few studies were excluded from this ISA because they do not provide
19 new information on particular geographic locations, potential at-risk populations or
20 lifestages, or range of ambient NO2 concentrations and because these studies more likely
21 are subject to publication bias.
22 The Preamble describes the general framework for evaluating scientific information,
23 including criteria for assessing study quality and developing scientific conclusions.
24 Aspects specific to evaluating studies of oxides of nitrogen are described in Table 5-1.
25 For epidemiologic studies, emphasis is placed on studies that characterize quantitative
26 relationships between oxides of nitrogen and health effects, examine exposure metrics
27 that well represent the variability in concentrations in the study area, consider the
28 potential influence of other air pollutants and factors correlated with oxides of nitrogen,
29 examine potential at-risk populations and lifestages, or combine information across
30 multiple cities. With respect to the evaluation of controlled human exposure and
31 toxicological studies, emphasis is placed on studies that examine effects relevant to
32 humans and NO2 concentrations that are defined in this ISA to be relevant to ambient
33 exposures. Based on peak ambient concentrations (Section 2.5) and the ISA definition
34 that ambient-relevant exposures be within one to two orders of magnitude of current
35 levels, NO2 concentrations 5,000 ppb1 or less are defined to be ambient relevant.
36 Experimental studies with higher exposure concentrations were included if they informed
1 The 5,000-ppb upper limit applies largely to animal toxicological studies but also a few controlled human exposure
studies.
January 2015 1-4 DRAFT: Do Not Cite or Quote
-------
1 dosimetry or potential modes of action. For the evaluation of human exposure to ambient
2 NO2, emphasis is placed on studies that examine the quality of data sources used to assess
3 exposures, such as central site monitors, land use regression (LUR) models, and personal
4 exposure monitors. The ISA also emphasizes studies that examine factors that influence
5 exposure such as time-activity patterns and building ventilation characteristics.
6 Integrating information across scientific disciplines and related health outcomes and
7 synthesizing evidence from previous and recent studies, the ISA draws conclusions about
8 relationships between NC>2 exposure and health effects. Determinations are made about
9 causation not just association and are based on judgments of aspects such as the
10 consistency, coherence, and biological plausibility of observed effects (i.e., evidence for
11 effects on key events in the mode of action) as well as related uncertainties. The ISA uses
12 a formal causal framework (Table II of the Preamble) to classify the weight of evidence
13 according to the five-level hierarchy summarized below.
14 • Causal relationship: the consistency and coherence of evidence integrated
15 across scientific disciplines and related health outcomes are sufficient to rule
16 out chance, confounding, and other biases with reasonable confidence.
17 • Likely to be a causal relationship: there are studies where results are not
18 explained by chance, confounding, or other biases, but uncertainties remain
19 in the evidence overall. For example, the influence of other pollutants are
20 difficult to address, or evidence among scientific disciplines may be limited
21 or inconsistent.
22 • Suggestive, but not sufficient, to infer a causal relationship: evidence is
23 generally supportive but not entirely consistent or overall is limited. Chance,
24 confounding, and other biases cannot be ruled out.
25 • Inadequate to infer a causal relationship: there is insufficient quantity,
26 quality, consistency, or statistical power of results from studies.
27 • Not likely to be a causal relationship: several adequate studies, examining
28 the full range of human exposure concentrations and potential at-risk
29 populations and lifestages, consistently show no effect.
1.3 Content of the Integrated Science Assessment
30 The ISA consists of the Preamble. Preface (legislative requirements and history of the
31 primary NCh NAAQS), Executive Summary, and seven chapters. Chapter 1 synthesizes
32 the scientific evidence that best informs policy-relevant questions that frame this review
33 of the primary NCh NAAQS. Chapter 2 characterizes the sources, atmospheric processes
34 of oxides of nitrogen, and trends in ambient concentrations. Chapter 3 describes methods
35 to estimate human exposure to oxides of nitrogen and the impact of error in estimating
36 exposure on associations with health effects. Chapter 4 describes the dosimetry and
January 2015 1-5 DRAFT: Do Not Cite or Quote
-------
1 modes of action for NO2 and NO. Chapters 5 and 6 evaluate and integrate epidemiologic,
2 controlled human exposure, and toxicological evidence for health effects related to
3 short-term and long-term exposure to oxides of nitrogen, respectively. Chapter 7
4 evaluates information on potential at-risk populations and lifestages.
5 The purpose of this chapter is not to summarize each of the aforementioned chapters but
6 to synthesize the key findings for each topic that informs the characterization of NC>2
7 exposure and relationships with health effects. This chapter also integrates information
8 across the ISA to inform policy-relevant issues such as NC>2 exposure durations and
9 patterns associated with health effects, concentration-response relationships, and the
10 public health impact of NC>2-related health effects (Section 1.6). A key consideration in
11 the health effects assessment is the extent to which evidence indicates that NO2 exposure
12 independently causes health effects versus indicating that NO2 may be serving as a
13 marker for a broader mixture of air pollutants, especially those related to traffic. To that
14 end, this chapter draws upon information about the sources, distribution, and exposure to
15 ambient NC>2 (Section 3.4.5) and identifies pollutants and other factors related to the
16 distribution of or exposure to ambient NCh that can potentially influence epidemiologic
17 associations observed between health effects and NO2 exposure (Section 1.4.3). The
18 discussions of the health effects evidence and causal determinations (Section 1.5)
19 describe the extent to which epidemiologic studies accounted for these factors and the
20 extent to which findings from controlled human exposure and animal toxicological
21 studies support independent relationships between NO2 exposure and health effects.
1.4 From Emissions Sources to Exposure to Nitrogen Dioxide
22 Characterizing human exposure is key to understanding the relationships between
23 ambient NC>2 exposure and health effects. The sources of oxides of nitrogen and the
24 transformations that occur in ambient air influence the spatial and temporal pattern of
25 NO2 concentrations in the air. These patterns have implications for variation in exposure
26 in the population, the adequacy of methods used to estimate exposure and, in turn, the
27 strength of inferences that can be drawn from associations observed between NCh
28 exposure and health effects.
1.4.1 Emission Sources and Distribution of Ambient Concentrations
29 The strength and distribution of emissions sources are important determinants of the
30 distribution of NC>2 in the ambient air and, in turn, human exposure. Information on
31 emissions is available for NOx, which is emitted primarily as NO. NO rapidly reacts with
January 2015 1-6 DRAFT: Do Not Cite or Quote
-------
1 radicals and ozone (Os) to form NO2 in the air. Based on the 2011 National Emissions
2 Inventory, the largest single source of NOx emissions in the U.S. overall and in major
3 population centers (city and surrounding communities) is highway vehicles (40-67%;
4 Section 2.3. Table 2-1). Sources such as electric utilities, commercial and residential
5 boilers, and industrial facilities are more variable across locations but can be important
6 contributors to ambient NO2 concentrations for the U.S. as a whole and in certain
7 populated areas. Some of these sources can affect local air quality with large, transient
8 emissions of NOx. Natural sources such as microbial processes in soil and wildfires make
9 small (2% of the inventory) contributions to emissions in U.S. population centers, and
10 emissions from natural and anthropogenic sources from continents other than North
11 America (i.e., North American Background) account for less than 1% (typically 0.3 ppb)
12 of ambient concentrations (Section 2.5.6). Although highway vehicles are a large,
13 ubiquitous source of NOx, the varying presence and mix of specific emissions sources
14 across locations can contribute to heterogeneity in ambient NO2 concentrations regionally
15 and locally, which has implications for variation in exposure to ambient NO2 within the
16 population.
17 In addition to emissions sources, factors that influence NO2 ambient concentrations
18 include chemical transformations, transport to other locations, meteorology, and
19 deposition to surfaces (Figure 1-1 and in more detail, Figure 2-1). NO and NO2 react with
20 gas phase radicals and Os to form other oxides of nitrogen such as peroxyacetyl nitrate
21 (PAN) and nitric acid (HNOs; Section 2.2). NO and NO2 also are involved in reaction
22 cycles with radicals produced from volatile organic compounds (VOCs) to form Os. The
23 reactions of NO and NO2 into other oxides of nitrogen typically occur more slowly than
24 the interconversion between NO2 and NO does, and NO and NO2 are the most prevalent
25 oxides of nitrogen in populated areas. Compounds such as HNOs and PAN can make up a
26 large fraction of ambient oxides of nitrogen downwind of major emission sources.
27 Sources, atmospheric transformations, and meteorology contribute to the temporal trends
28 observed in ambient NO2 concentrations. As a result of pollution control technologies on
29 vehicles and electric utilities (Section 2.3.2). NOx emissions from highway vehicles and
30 fuel combustion decreased by 49% in the U.S. from 1990 to 2013 (Figure 2-2). During
31 that time (1990-2012), U.S.-wide annual average NO2 concentrations decreased by 48%
32 (Figure 2-20). In addition to long-term trends, ambient NO2 concentrations show seasonal
33 trends, with higher concentrations measured in the winter than summer. Reflecting trends
34 in traffic, ambient concentrations at most urban sites are higher on weekdays than
35 weekends, and within a day, concentrations peak in early mornings, decrease until late
36 afternoon, then increase again in early evening corresponding with morning and evening
37 commutes. Diurnal trends in ambient NO2 also are affected by meteorology, with
January 2015 1-7 DRAFT: Do Not Cite or Quote
-------
concentrations rising during the night when atmospheric mixing is reduced because of
low wind speeds and low mixing layer heights.
r " NOY
i
1 NOZ
14.
*
INORGANIC
PRODUCTS
(e.g.. nitric
oxide)
! i
Long range transport to remote
regions at low temperatures w
/^>N02[^.
U ozone ozone \\
radical j
\ species sunlight l\
V ^ 1 M/-\ / ,<
...::^._N.0..v^ NC
1
1
> 1
ORGANIC I
PRODUCTS 1
(e.g.. peroxyacetyl I
nitrate) '
^ i 1
deposition
c^.
deposition
emissions
4
5
6
7
8
9
10
Note: The inner shaded box contains NOX (sum of nitric oxide [NO] and nitrogen dioxide [NO2]). The outer box contains oxides of
nitrogen (NOZ) formed from reactions of NOX.(NOZ). Oxides of nitrogen in the outer and inner boxes (NOX + NOZ) are collectively
referred to as NOY by the atmospheric sciences community.
Source: National Center for Environmental Assessment. For more details on the various reactions, see Figure 2-1.
Figure 1-1 Reactions of oxides of nitrogen species in the ambient air.
The spatial variation in emissions sources and chemical transformation of oxides of
nitrogen likely contribute to the variability in ambient NC>2 concentrations observed at
regional, urban, neighborhood, and near-road scales (Section 2.5). Measurements from
U.S. air monitoring networks1 of several hundred sites (Section 2.5.1) show wide
variation in ambient NC>2 concentrations across the U.S. Across sites, the mean 1-h daily
maximum ambient NC>2 concentration for 2011-2013 was 19 ppb, and the 5th to 99th
percentile range was 2-55 ppb (Table 2-3). The mean annual average NO2 concentration
was 8.6 ppb, and the 5th to 99th percentile range was 1.4-22.5 ppb (Table 2-4). Ambient
1 The air monitoring networks serve many objectives: determining compliance with the NAAQS, providing the
public with air pollution data in a timely manner, and providing estimates of ambient exposure for research studies.
January 2015
1-8
DRAFT: Do Not Cite or Quote
-------
1 NC>2 concentrations are higher in large cities than in less populated locations
2 (Figures 2-11 and 2-12). Ambient NC>2 concentrations also can vary widely across sites
3 within cities where vehicle emissions are the major source (Figure 2-14. Table 2-5).
4 Some city sites agree well in terms of temporal correlations or magnitude of
5 concentration; however, the siting of most monitors away from sources likely does not
6 capture the extent of variability in ambient NC>2 in a city. Preliminary data from the
7 near-road network for 1 year for a small group of U.S. cities tend to show higher NC>2
8 concentrations near roads than at many other sites in the same city (Table 2-8). Across
9 near-road sites in five cities, mean 1-hour NC>2 concentrations in winter were 21-45 ppb,
10 and maximums were 52-97 ppb. Studies have measured 1-h to 2-week avg NCh
11 concentrations of 29 to 65 ppb within 20 m of a road, which are about 20 ppb higher than
12 concentrations 80 to 500 m from the same road (Section 2.5.3. Table 2-6). The wide
13 variation in ambient NC>2 concentrations across spatial and temporal scales, largely
14 influenced by motor vehicle emissions, can contribute to variation in NC>2 exposure
15 within the population and has important implications for accurately characterizing
16 exposure.
1.4.2 Assessment of Human Exposure
17 Characterizing the adequacy of various exposure assessment methods to represent the
18 variability in ambient concentrations in a location is key in drawing inferences from
19 epidemiologic associations with health effects. Exposure is determined by concentrations
20 in specific ambient, indoor, and in-vehicle locations and time spent in those locations
21 (Section 3.4.1). People vary in the locations where they spend time and time spent in
22 those locations (Section 3.4.3.1). and NCh concentrations can vary widely across outdoor,
23 indoor, and in-vehicle locations (Figure 3-1). Measures of NC>2 exposure that do not fully
24 account for the variability in ambient concentrations and people's activity patterns have
25 some amount of error, and this error can alter associations observed with health effects.
26 The extent and impact of error can differ by exposure assessment method and between
27 epidemiologic studies of short-term exposure, which tend to examine temporal (e.g., day
28 to day) changes in NC>2, and studies of long-term exposure, which tend to compare people
29 living in locations that differ in ambient NC>2 concentrations (Section 3.4.5).
30 Ambient NC>2 concentrations at central site monitors represent both short-term and
31 long-term exposure with some amount of error. Central site monitors do not cover all
32 locations where people live or spend their time and also are not likely to capture the
33 temporal or spatial variability in ambient NC>2 concentrations in a given area. Long-term
34 personal NC>2 exposures and their relationships with ambient NC>2 concentrations are not
35 well characterized. A wide range of correlations is observed between short-term total
January 2015 1-9 DRAFT: Do Not Cite or Quote
-------
1 personal and ambient NO2 concentrations (0.12 to 0.43; Table 3-6) and in ambient NO2
2 concentrations across sites within some cities (Section 2.5.2). On one hand, poor
3 correlations may not necessarily mean that concentrations at central sites are inadequate
4 exposure metrics because the data may not reflect relationships between ambient NO2
5 concentrations and the ambient component of personal exposure (Section 3.4.2). On the
6 other hand, the correlations could indicate heterogeneity among individuals in how well
7 short-term temporal changes in NO2 concentrations at central site monitors represent
8 temporal changes in ambient exposure. The adequacy of long-term average ambient NO2
9 concentrations from central site monitors (from one monitor or combined across multiple
10 monitors) to represent variability in ambient exposures will depend on how close the
11 monitors are to the population or how well local emission sources are dispersed
12 (Section 3.4.5.2). In locations with few or well-dispersed sources and similar ambient
13 NO2 concentrations across sites, concentrations at central site monitors may adequately
14 capture the spatial variation in long-term ambient NO2 exposures of the population.
15 Proximity to roads may contribute substantially to short-term and long-term ambient NO2
16 exposure among people living or working near roads or commuting on roads, and the
17 2008 ISA for Oxides of Nitrogen cited the potential for in-vehicle exposures to dominate
18 short-term personal exposure (U.S. EPA. 2008). Data from the U.S. near-road monitoring
19 network are too preliminary to allow for meaningful comparisons of the temporal or
20 spatial patterns in NO2 near and away from roads. Data from London, U.K. show that
21 24-h avg NO2 concentrations often are higher (26-170%; Table 2-9) at roadside than at
22 background sites but tend to be moderately to highly correlated between sites (correlation
23 coefficient [r] = 0.63-0.86; Table 2-9). These data indicate that central site monitors may
24 capture the short-term temporal variability near roads but may not represent the
25 magnitude of long-term average NO2 concentrations near roads. Another consideration in
26 estimating exposure from central site monitors is that the chemiluminescence
27 measurement method tends to overestimate ambient NO2 concentrations because of
28 interference from other oxides of nitrogen. However, interference generally is less than
29 10% in urban locations (Section 2.4.1) and may not vary widely day to day
30 (Section 3.4.3.4) to produce substantial error in characterizing daily changes in NO2
31 concentration. It is not clear how interference compares among locations and what impact
32 interference may have on comparisons of long-term average NO2 concentrations among
33 locations.
34 In addition to concentrations at central site monitors, epidemiologic studies examined
35 short-term NO2 exposures at people's locations, including personal ambient and total
36 NO2 exposure as well as measurements taken outdoors at schools and indoors at homes
37 and schools. A time-weighted average of NO2 concentrations in people's locations
38 correlated well with total personal short-term NO2 exposures (Section 3.4.3.1). indicating
January 2015 1-10 DRAFT: Do Not Cite or Quote
-------
1 that NC>2 concentrations in locations where a person spends time can represent a
2 component of personal exposure and aid in inference about NC>2-related health effects.
3 Epidemiologic studies examining total personal or indoor NCh concentrations also have
4 important roles in characterizing health effects of NC>2 exposure because they can help
5 distinguish NC>2-related health effects from the potential influence of other traffic-related
6 pollutants. Correlations between NCh and some copollutants are lower for total personal
7 or indoor metrics than ambient metrics (Section 3.4.4.3. Table 3-10). Results for total
8 personal and indoor NC>2 concentrations also can inform health effects related to ambient
9 exposure for populations with high total personal-ambient NC>2 correlations and
10 populations for whom indoor concentrations are affected by the penetration of ambient
11 NO2 from open windows or other factors that increase building air exchange rate
12 (Section 3.4.3.3). There is an increase among recent studies in the use of LUR and
13 dispersion models to estimate long-term NC>2 exposures at the neighborhood scale or at
14 an individual's residence. Dispersion models have uncertainties related to systematic bias
15 in representing ambient NC>2 concentrations (Section 3.5). LUR models may account for
16 differences among individuals in residential distance to sources and have been
17 demonstrated to represent well the variability in long-term average ambient NO2
18 concentrations in many locations (Section 3.5).
19 Errors in representing the temporal and spatial variability in short-term and long-term
20 averages, respectively, of ambient NO2 concentrations in a given area and exposures of
21 the population can impact the characterization of relationships between NO2 exposure
22 and health effects. For short-term exposure, if ambient NCh concentrations and people's
23 ambient exposures are temporally correlated, there may be little error in health effect
24 estimates. NC>2 measurements at people's locations may include error because the metrics
25 may not represent potentially important exposures across the range of locations where
26 people spend time. However, larger magnitude health effects have been estimated for
27 more spatially resolved measures of short-term NC>2 exposure than for NC>2 measured at a
28 single central site monitor or averaged over multiple monitors in a city (Section 3.4.5.1).
29 Such findings indicate that not accounting for the varying population distribution around
30 a central site monitor or varying correlations in NC>2 across monitors can decrease
31 associations with NCh (Section 3.4.5.1). Compared with NC>2 estimated by LUR,
32 long-term average NC>2 concentrations at central site monitors often show smaller
33 associations with health effect but larger associations in some cases (Section 3.4.5.2).
34 Thus, spatial misalignment of long-term NC>2 exposure metrics can alter health effect
35 estimates in either direction. Exposure error also can impact the precision [i.e., 95%
36 confidence interval (CI)] of health effect estimates due to variable relationships between
37 personal and ambient NC>2 across people and time and differences in nonambient
38 exposures. It is unclear how error produced from using ambient NC>2 concentrations at
39 central site monitors to represent near-road exposures impacts health effect associations
January 2015 1-11 DRAFT: Do Not Cite or Quote
-------
1 because differences in temporal or spatial patterns for near-road NO2 concentrations or
2 correlations with personal NO2 exposure are not widely characterized. For short-term and
3 long-term exposure, evaluating how well personal, central site, location-specific, or
4 modeled estimates of NO2 exposure capture the variability in ambient concentrations or
5 exposure and the potential impact of exposure error is a key consideration in drawing
6 inferences from epidemiologic studies about NO2-related health effects.
1.4.3 Factors Potentially Correlated with Nitrogen Dioxide Exposure to Consider
in Evaluating Relationships with Health Effects
7 The large influence of motor vehicle emissions on the distribution of ambient NC>2
8 concentrations not only affects the assessment of NC>2 exposure but also has implications
9 for co-exposure to other traffic-related pollutants. NO2 concentrations are higher near
10 roads as are concentrations of elemental or black carbon (EC/BC), ultrafine particles
11 (UFP), carbon monoxide (CO), and VOCs (Section 3.3.1). The exact nature of gradients
12 varies among pollutants, but concentrations of traffic-related pollutants, including NCh,
13 decrease with increasing distance from the road. PM2 s1 and organic carbon (OC) do not
14 show clear gradients; however, a portion of PIVb 5 and OC comes from vehicle emissions.
15 These correlations and evidence that the copollutants show relationships with many of the
16 same health effects as NO2 and have similar modes of action (Table 5-1. Section 5.1.2.1)
17 point to the importance of evaluating the potential for NO2-related health effects to be
18 confounded (i.e., biased) by other traffic-related pollutants or for NO2 to represent a
19 mixture of traffic-related pollution. Common sources, atmospheric reactions, or similar
20 trends due to meteorologic conditions extend the potential for co-exposures to pollutants
21 beyond those emitted from vehicles. Factors such as socioeconomic status (SES), season,
22 and temperature also show correlations with NO2 concentrations and relationships with
23 similar health effects. The potential for a particular factor to confound NO2-health effect
24 associations varies depending on the extent of correlation with NO2 concentrations, the
25 nature of the relationship with the health effect, and study design (i.e., whether temporal
26 variation in short-term exposure or spatial variation in long-term exposure is examined).
27 Short-term average NO2 concentrations show a range of correlations with traffic-related
28 copollutants, but high correlations often are observed (Figure 3-6. Table 3-8). For
29 example, for averaging times of 1 to 24 hours, the 25th to 75th percentile ranges of
30 correlation coefficients are 0.59-0.96 for CO, 0.41-0.61 for PM25, and 0.58-0.67 for
31 EC. Limited data indicate similar correlations with short-term averages of VOCs, and
32 lower correlations with OC. Long-term average ambient NO2 concentrations show
1 In general terms, paniculate matter with an aerodynamic diameter less than or equal to a nominal 2.5 \im, a
measure of fine particles.
January 2015 1-12 DRAFT: Do Not Cite or Quote
-------
1 similar correlations with PIVb 5 and CO as short-term averages, but the distribution of
2 correlations is shifted to higher values. Correlations of long-term averages of NC>2 with
3 EC/BC, VOCs, OC, and UFP are not well characterized (Figure 3-6). Information on
4 seasonal correlations between ambient concentrations of NO2 and traffic-related
5 copollutants is sparse, but there is some indication of lower NC^-PNfe.s correlations for
6 short-term averages in the warm season (Section 3.4.4.1). These data point to potentially
7 lower confounding by PM2 5 in the warm season. Although traffic-related copollutants
8 have been associated with many of the same health effects as NO2 (Table 5-1). the wide
9 range of correlations with short-term and long-term average NC>2 concentrations indicates
10 variation among locations in confounding potential.
11 Much of the data characterizing correlations of NO2 with traffic-related copollutants are
12 based on measurements at central site monitors. The varying spatial patterns among
13 pollutants may obscure true correlations across study areas or correlations in personal
14 exposure. Except for UFP, the few available data do not indicate systematically higher
15 correlations near roads (Figure 3-6). However, correlations of short-term averages of NC>2
16 with PM25 (r = 0.06 to 0.38), EC (r = 0.22 to 0.49), and VOCs (r = -0.42 to 0.14) can be
17 weaker for personal exposures than ambient concentrations (Table 3-11. Section 3.4.4.3).
18 Correlations of short-term averages of NO2 with PIVbs and BC sometimes can be lower
19 indoors than outdoors (Table 3-8 and Table 3-12). These limited data indicate that
20 associations of short-term personal or indoor NO2 exposures with health effects may be
21 less subject to confounding by certain traffic-related copollutants. In some locations,
22 short-term average ambient NO2 concentrations are related more strongly to personal PM
23 than personal NO2 exposure. However, recent data show negative to moderate
24 correlations between ambient NO2 concentrations and personal PlVfc 5 or EC (r = -0.19 to
25 0.44; Table 3-9). suggesting that ambient NO2 concentrations are not necessarily just a
26 surrogate for personal PM exposure. The varying correlations between short-term
27 average concentrations of NO2 and other traffic-related pollutants across various
28 microenvironments indicate that the potential for confounding by traffic-related
29 copollutants varies by the exposure assessment method. Similar information to compare
30 copollutant correlations among microenvironments is not available for long-term average
31 NO2 concentrations.
32 Other potential confounding factors to consider for long-term NO2 exposure are measures
33 of traffic proximity or intensity, which could represent exposure to other pollutants that
34 display gradients with distance to road. Although NO2 is not unique to vehicle emissions
35 and can indicate sources such as off-highway vehicles and electric utilities (Section 2.3).
36 distance to roads, the length of nearby roads, and vehicle counts are predictors of ambient
37 NO2 concentrations in LUR models (Section 3.2.2.1). Given recent findings linking
38 residential proximity to roads with respiratory effects and possibly with cardiovascular
January 2015 1-13 DRAFT: Do Not Cite or Quote
-------
1 effects and mortality (HEI. 2010). roadway proximity could confound NCh-health effect
2 associations by indicating exposure to traffic pollution. Studies considering the potential
3 influence of exposure to traffic, including residential proximity to roads, are another line
4 of evidence to inform the independent effects of long-term NC>2 exposure on health.
5 Short-term and long-term averages of NC>2 also show a range of correlations with the
6 copollutants PMio,1 SC>2, and Os. Short-term and long-term average NC>2 concentrations
7 tend to be moderately correlated with PMio (r for 25th-75th percentiles = 0.40-0.66 for
8 short-term averages, 0.44-0.75 for long-term averages) and 862 (Figure 3-6. Table 3-8).
9 Short-term averages of Os often are inversely correlated with NC>2, and peak correlations
10 are moderate (r for 25th-75th percentile = -0.51 to 0.32) even in the summer, when Os
11 concentrations are higher (Table 3-8). Higher correlations are observed between
12 long-term averages of NC>2 and Os (r for 25th-75th percentiles = 0.26-0.63). The wide
13 range of correlations observed for short-term and long-term average concentrations of
14 NC>2 with PMio, SCh, and Os indicates the variable potential for these pollutants to
15 confound health effect associations for NC>2. For short-term average NO2 concentrations,
16 the distributions of correlations with PMio and 862 are shifted to lower values compared
17 to correlations with most traffic-related pollutants, indicating the lower potential for
18 confounding. Specific to long-term exposure, relationships of long-term 862 and Os
19 exposure with many of the health effects evaluated in this ISA are uncertain (Table 5-1)
20 as is their potential to confound NCh-health effect associations.
21 Residence near traffic has been linked to higher noise or stress levels, but information on
22 whether noise or stress confounds health effect associations with short-term or long-term
23 NO2 exposure is limited. Weak to moderate correlations tend to be reported between
24 noise and short-term (r = 0.14-0.62) and long-term (r = 0.22-0.46) average ambient NCh
25 concentrations, but high correlations have been observed for short-term NC>2 averages
26 (r = 0.83; Section 3.4.4.4). The impact of short-term changes in noise or stress on health
27 effects is not well characterized, but some data link long-term noise exposure and stress
28 to cardiovascular effects (Section 6.3.2) and decreases in cognitive function
29 (Section 6.4.4). Thus, the potential for stress or noise to confound NCh-health effect
30 associations is uncertain for short-term exposure but may exist for long-term exposure.
31 Other potential confounding factors to consider include temperature and humidity for
32 associations of health effects with short-term NC>2 exposure because of similar
33 time-varying patterns as ambient NC>2 concentrations and health effects. Also, similar to
34 many health effects, short-term averages of ambient NC>2 concentration vary by day of
35 the week and season and exhibit long-term time trends. For studies of long-term NCh
1 In general terms, paniculate matter with an aerodynamic diameter less than or equal to a nominal 10 ^m, a
measure of thoracic particles.
January 2015 1-14 DRAFT: Do Not Cite or Quote
-------
1 exposure that compare individuals living in different locations, it is important to evaluate
2 confounding by factors such as SES, race (Sections 7.5.2 and 7.5.3). and age, all of which
3 can covary with long-term NCh exposures among individuals and spatially with
4 long-term ambient NC>2 concentrations among communities.
5 For studies reviewed in this ISA, the main method to account for potential confounding is
6 multivariable models that include NC>2 concentrations and the putative confounder. The
7 NO2 effect estimate represents the effect of NO2, keeping the level of the covariate
8 constant. Confounding is assessed by examining the change in the magnitude of the effect
9 estimate and width of the 95% CI, not a change in statistical significance. There are
10 limitations to multivariable models, and correlations between variables and the exposure
11 assessment method are important considerations in drawing inferences about
12 confounding (Section 5.1.2.1). High correlations between NC>2 concentrations and the
13 potential confounder can misleadingly decrease or increase the magnitude or precision of
14 the effect estimate for NO2 or the covariate and particularly are a concern for models that
15 include a traffic-related copollutant or include three or more pollutants in the same
16 model. Potential differences in exposure measurement error between NO2 and the
17 copollutant also limit inferences about an independent NCh association from copollutant
18 models. Inference from copollutant models may be stronger for pollutants measured at
19 people's locations and for personal exposure than for pollutants measured at central site
20 monitors. As in the 2008 ISA for Oxides of Nitrogen (U.S. EPA. 2008). a key issue in
21 this ISA is the extent to which epidemiologic studies examined potential confounding by
22 traffic-related copollutants and the extent to which other lines of evidence support
23 independent relationships between NO2 exposure and health effects.
1.5 Health Effects of Nitrogen Dioxide Exposure
24 This ISA evaluates relationships between an array of health effects and short-term
25 (Chapter 5) and long-term (Chapter 6) exposures to NC>2 as examined in epidemiologic,
26 controlled human exposure, and animal toxicological studies. Short-term exposures are
27 defined as those with durations of minutes up to 1 month, with most studies examining
28 effects related to exposures in the range of 1 hour to 1 week. Long-term exposures are
29 defined as those with durations of more than 1 month to years. Drawing from the health
30 effects evidence described in detail in Chapter 5 and Chapter 6. information on dosimetry
31 and modes of action presented in Chapter 4. as well as issues regarding exposure
32 assessment and potential confounding described in Chapter 3 and Section 1.4. the
33 subsequent sections and Table 1-1 present the key evidence that informs the causal
34 determinations for relationships between NC>2 exposure and health effects.
January 2015 1-15 DRAFT: Do Not Cite or Quote
-------
1.5.1 Respiratory Effects
1 Relationships of short-term and long-term NC>2 exposure with respiratory effects are
2 supported by the dosimetry and modes of action characterized for inhaled NC>2. Although
3 it is not clear how ambient-relevant NC>2 exposures compare with NC>2 produced
4 endogenously in the lung during inflammation and other immune responses
5 (Section 4.2.2.4). ambient-relevant concentrations of inhaled NC>2 are absorbed
6 throughout the respiratory tract. Dosimetry models predict that total NC>2 dose is
7 relatively constant across the tracheobronchial region and rapidly decreases in the gas
8 exchange (i.e., alveolar) region (Section 4.2.2.3). NCh is a reactive gas that reacts with
9 antioxidants and other constituents of the epithelial lining fluid of the respiratory tract to
10 form secondary oxidation products (Section 4.2.2.1). Although these reactions are rapid,
11 antioxidant levels vary across regions of the lung, and untransformed NCh that reaches
12 the alveolar region of the airways may penetrate the thin layer of the epithelial lining
13 fluid to reach underlying tissue. Thus, the variable physical and chemical nature of the
14 respiratory tract may influence the site in the respiratory tract of NC>2 uptake and
15 NC>2-induced respiratory effects. The formation of secondary oxidation products likely is
16 the initiating event in the sequence of events comprising the mode of action for NC>2
17 (Section 4.3.2.1). These products can induce oxidative stress, inflammation, allergic
18 responses, and altered immune function, which are part of the mode of action for
19 respiratory effects (Figures 1-2 and 4-1) as described in the sections that follow.
January 2015 1-16 DRAFT: Do Not Cite or Quote
-------
II-
Inhalation
of NO; "*•
Activation!
neural reflexes
*
..••'
ff
r*
•
Reactions in
respiratory tract
Formation of
oxidation/nitration
products
\
\
T
ActivatiorV
Sensitizationof
neural reflexes
jr
^
S^
r
".
s
de granulation
Irrp aired epithelial
f Allergic
responses
| Pulmonary
oxidatwe stress
T
S pi 1 love r of i nfla rrmat ory
mediators to blood
Spillover of reactive
intermediates to blood
1
:
:
:
'
: f Airway
± •
Allergic
* sensitization
^ Airway
remodeling
Oxidative stress
Vascular activation
triqqer
t,»
_..*'
*
*•*
•..
\
•
,.•**
.-•*'
Asthma
exacerbation
n i +
of asthma
Other organ
Cardiovascular
effects
Note: Adapted from Figures 4-1. 4-2. and 4-3 in Section 4.3.5. Solid arrows and white boxes represent pathways for which there is
consistent evidence. Dotted lines and gray boxes represent potential pathways for which the evidence is limited or inconsistent.
Figure 1-2 Characterization of potential modes of action for health effects
related to exposure to nitrogen dioxide (NO2).
i
2
o
J
4
5
6
7
8
9
10
11
12
13
14
Respiratory Effects Associated with Short-term Exposure to Nitrogen
Dioxide
A causal relationship exists between short-term NO2 exposure and respiratory effects
based on evidence for asthma exacerbation. The conclusion is strengthened from the
likely to be a causal relationship determined in the 2008 ISA for Oxides of Nitrogen
because epidemiologic, controlled human exposure, and animal toxicological evidence
together describe a coherent and biologically plausible pathway by which NO2 exposure
can trigger an asthma exacerbation (Table 1-1). There is some evidence indicating that
short-term NC>2 exposure may be related to other respiratory effects, such as exacerbation
of allergy or chronic obstructive pulmonary disease (COPD), respiratory infection,
respiratory mortality, and respiratory effects in healthy people. However, because of
inconsistency across disciplines and/or limited information to support biological
plausibility, there is uncertainty whether short-term NCh exposure has independent
relationships with nonasthma respiratory effects (Section 5.2.9. Table 5-45).
Supporting a causal relationship, recent epidemiologic studies continue to provide
consistent and coherent evidence for NCh-related exacerbation of asthma. Across diverse
January 2015
1-17
DRAFT: Do Not Cite or Quote
-------
1 geographic locations, short-term increases in ambient NO2 concentration are associated
2 with increases in hospital admissions and ED visits for asthma. As uncontrolled
3 symptoms are the major reason for seeking medical treatment, coherence is demonstrated
4 by observations of NO2-related increases in respiratory symptoms and decreases in lung
5 function in children with asthma. Associations are observed in studies with maximum
6 concentrations of 48 to 106 ppb for 24-h avg NO2 and 59 to 306 ppb for daily 1-h max
7 NO2. Epidemiologic evidence is consistent across the various methods used to estimate
8 NO2 exposure: personal ambient and total NO2 measurements, NO2 measured outside
9 children's schools, NO2 measured inside children's schools and homes, and ambient NO2
10 concentrations averaged across central site monitors in a city. NO2 measured at people's
11 locations may represent exposure better than NO2 measured at central site monitors, and
12 along with findings for personal and indoor exposure, provide a strong basis for inferring
13 a causal relationship between short-term NO2 exposure and asthma exacerbation.
14 NO2 associations with asthma-related effects persist with adjustment for temperature,
15 humidity, season, long-term time trends, as well as a copollutant such as PMio, SO2, or
16 Os. In a few studies, NO2 associations are attenuated with adjustment for EC/BC, UFP, or
17 a VOC. However, for the most part, recent studies add findings for NO2 associations that
18 persist with adjustment for a traffic-related copollutant such as PM2 5, or as examined in
19 fewer studies, EC/BC, UFP, or CO. Potential confounding by OC, PM metal species, or
20 VOCs is poorly studied, but NO2 associations with asthma exacerbation tend to persist in
21 the few available copollutant models. In some cases, single-pollutant models indicate
22 asthma-related effects in association with NO2 but not PM2 5 or EC/BC, which were
23 moderately correlated with NO2 (r = 0.22-0.57). Recent epidemiologic results also
24 suggest asthma exacerbation in relation to exposure indices that combine NO2 with EC,
25 PM2 5, Os, and/or SO2 concentrations, but neither epidemiologic nor experimental studies
26 strongly indicate synergistic effects between NO2 and copollutants. Inference about
27 associations for NO2 that are independent of PM2 5, EC/BC, OC, or UFP is strong
28 particularly in studies that measure pollutants at people's locations and for personal
29 exposure because of comparable measurement error among pollutants. Associations with
30 personal total and indoor NO2 measurements also support an independent effect of NO2
31 exposure because the lower (e.g., r = -0.37 to 0.31) correlations observed with many
32 traffic-related copollutants compared to ambient NO2 concentrations indicate that the
33 findings for personal and indoor NO2 may be less prone to confounding by the same
34 traffic-related copollutants than findings for ambient NO2 concentrations (Section 1.4.3).
35 In the indoor studies, the relative contribution of indoor and outdoor sources to indoor
36 NO2 concentrations are unknown. And, while associations of outdoor school NO2 with
37 asthma-related effects persist with adjustment for indoor NO2 in one group of children, it
38 is unclear whether indoor exposure alters responses of people to outdoor NO2 exposure.
January 2015 1-18 DRAFT: Do Not Cite or Quote
-------
1 Because copollutant models have limitations (Section 1.4.3) and are not analyzed for
2 every correlated copollutant or study, evidence from experimental studies is critical. The
3 key evidence for an independent effect of NO2 exposure in exacerbating asthma are the
4 findings from previous controlled human exposure studies for increases in airway
5 responsiveness in adults with asthma following NC>2 exposures of 200 to 300 ppb for
6 30 minutes and 100 ppb for 1 hour. Airway hyperresponsiveness is a key feature of
7 asthma and can lead to poorer control of symptoms. A recent meta-analysis shows that
8 NC>2 exposure cut in half the dose of the challenge agent required to increase airway
9 responsiveness, which is a measure of a clinically relevant change. Such evidence for
10 clinically relevant increases in airway responsiveness induced by NC>2 exposures that are
11 not much higher than peak ambient concentrations (Section 2.5) provide plausibility for
12 short-term ambient NC>2 exposures inducing asthma exacerbation. Biological plausibility
13 also is supported by experimental studies of adults with asthma showing that NO2
14 exposures of 260 ppb for 15-30 minutes enhanced allergic inflammation, which is a key
15 event in the mode of action for asthma exacerbation via its role in increasing airway
16 responsiveness (Figure 1-2). These results support the NCh-related respiratory effects
17 observed in populations with asthma that also had high prevalence of allergy. The results
18 for increases in airway responsiveness and allergic inflammation occurring after NC>2
19 exposures of 100-300 ppb for up to 1 hour also support the few findings of increased
20 respiratory effects in adults with asthma and healthy adults associated with NC>2 exposure
21 (range: 5.7-154 ppb) occurring over 2 or 5 hours at locations near roads.
22 Not all evidence supports NC>2-related respiratory effects. NO2 exposure has variable
23 effects on oxidative stress in experimental studies. NC>2-related decreases in lung function
24 are observed in epidemiologic but not controlled human exposure studies. Neural reflexes
25 do not appear to be involved (Figure 1-2, Section 4.3.2.2). but NC>2-induced (500 ppb)
26 mast cell degranulation in rats suggests airway obstruction, which could lead to decreases
27 in lung function.
28 Much of the evidence from epidemiologic and experimental studies was available in the
29 2008 ISA. However, compared to the 2008 ISA, this ISA more explicitly evaluates the
30 coherence and biological plausibility for specific respiratory outcome groups. Rather than
31 new evidence, the integration of epidemiologic and experimental evidence for asthma
32 exacerbation—uptake of NCh in the respiratory tract and reactions to form reactive
33 oxidation products, allergic inflammation, airway responsiveness, asthma symptoms, and
34 hospital admissions and ED visits for asthma, associations with NC>2 measured in
35 people's locations (which may better represent exposure), associations with adjustment
36 for another traffic-related pollutant—describes a coherent, biologically plausible pathway
37 to support a causal relationship between short-term NCh exposure and respiratory effects.
January 2015 1-19 DRAFT: Do Not Cite or Quote
-------
Respiratory Effects Associated with Long-Term Exposure to Nitrogen
Dioxide
1 There is likely to be a causal relationship between long-term NO2 exposure and
2 respiratory effects based on evidence for the development of asthma. The conclusion is
3 strengthened from that determined in the 2008 ISA for Oxides of Nitrogen because
4 whereas previous epidemiologic evidence was limited and inconsistent, recent evidence
5 consistently indicates associations between ambient NO2 concentrations and asthma
6 incidence in children (Table 1-1). As with findings for short-term NC>2 exposure, the
7 evidence base varies across respiratory outcome groups, and there is more uncertainty as
8 to whether long-term ambient NC>2 exposure is related to decreased lung function or
9 development or to increases in COPD, respiratory infection, or respiratory mortality.
10 Epidemiologic studies are noteworthy for isolating the development of asthma from the
11 exacerbation of pre-existing asthma. Studies followed children overtime, in several cases
12 from birth, and examined NCh exposure for periods preceding asthma diagnosis. Recent
13 epidemiologic studies also associate long-term NO2 exposure with asthma in adults, but
14 asthma diagnosis or symptoms ascertained only during follow-up could represent
15 recurrence of childhood asthma. Asthma incidence is associated with NO2 exposures
16 estimated at children's homes based on outdoor measurements or LUR models that often
17 well predicted measured concentrations in study locations (R2 = 0.42 to 0.69;
18 Section 6.2.9. Table 6-5). Such findings strengthen inference about NO2-related asthma
19 development. Results also are consistent for less spatially resolved ambient NO2
20 concentrations at central site monitors. Asthma incidence is associated with the average
21 NO2 concentration for the first year of life and NC>2 averaged over multiple years (study
22 means: 14 to 28 ppb), and no single critical exposure period is identified. Associations
23 with asthma are found with adjustment for SES, smoking exposure, gas stove use,
24 community of residence, and in one study, psychosocial stress. However, potential
25 confounding by traffic-related pollutants or proximity to roads is not examined.
26 A small body of previous experimental studies supports the biological plausibility for a
27 relationship between long-term NC>2 exposure and asthma development by demonstrating
28 effects on key events in the underlying mode of action. As illustrated in Figure 1-2. the
29 information on mode of action is coherent between studies of short-term and long-term
30 NC>2 exposure. NC>2 exposure (1,000 to 4,000 ppb) for 6-12 weeks increased airway
31 responsiveness in a study of rodents. A simultaneous NCh-induced increase in airway
32 resistance suggests airway obstruction and airway remodeling, which also are linked to
33 asthma development. Findings across disciplines indicate increased oxidative stress and
34 inflammation in relation to long-term NC>2 exposure but not consistently across studies.
35 The temporal pattern of NC>2 exposure underlying the epidemiologic associations with
36 asthma is not well delineated. However, a few experimental studies show that repeated
January 2015 1-20 DRAFT: Do Not Cite or Quote
-------
1 short-term NC>2 exposures over 4 to 14 days led to the development of allergic responses
2 in healthy adults and healthy rodent models (2,000-4,000 ppb) and to increased airway
3 responsiveness in rodents (4,000 ppb). This evidence for short-term NC>2 exposure
4 supports a relationship between long-term NC>2 exposure and asthma development
5 because it demonstrates the development of asthma-related effects in healthy humans and
6 animal models and indicates that repeated increases in exposure may be important. NCh
7 exposures that induce effects related to asthma development are higher than those that
8 induce effects related to asthma exacerbation as described in the preceding section but are
9 within the range of exposures considered to be ambient relevant (Section 1.2).
10 Epidemiologic studies continue to support associations of long-term NC>2 exposure with
11 decreases in lung function and development and increased respiratory disease severity in
12 children. These outcomes are associated with similar NO2 concentrations and durations as
13 asthma development and similar exposure assessment methods (Table 6-5) Some studies
14 of lung function observed associations with long-term NC>2 after adjustment for
15 short-term NC>2 exposure, but associations for asthma symptoms do not exclude the
16 potential influence of short-term NC>2 exposure. Compared with asthma development,
17 there is more uncertainty whether long-term NC>2 exposure has an independent effect on
18 decreasing lung function or development or increasing respiratory disease severity. NC>2
19 exposure does not alter lung function in experimental animals, and the hyperproliferation
20 of lung epithelial cells and fibrosis in adult animals are not directly related to the lung
21 function changes described in children. Associations of long-term NC>2 exposure with
22 bronchitic symptoms or lung function persisted when adjusted for PM2 5, EC, OC, or
23 distance to freeway, but such findings are limited in number and not entirely consistent.
24 Together, evidence from recent epidemiologic studies and previous experimental studies
25 supporting effects on the development of asthma indicates there is likely to be a causal
26 relationship between long-term NC>2 exposure and respiratory effects. Epidemiologic
27 studies observe associations with NC>2 exposure estimated at children's homes with LUR
28 models, which may better represent differences in ambient NC>2 exposure among subjects
29 compared with less spatially resolved NC>2 measurements from central site monitors.
30 Potential confounding by traffic-related copollutants largely is unexamined for asthma
31 development. However, findings from experimental studies for increased airway
32 responsiveness and allergic responses, which are part of the mode of action for asthma
33 development, are considered to provide some support for an independent effect of
34 long-term NC>2 exposure. Because such evidence is limited, some uncertainty remains in
35 attributing epidemiologic associations between long-term NC>2 exposure and asthma
36 development specifically to NC>2 among the array of traffic-related pollutants.
January 2015 1-21 DRAFT: Do Not Cite or Quote
-------
1.5.2 Health Effects beyond the Respiratory System
1 Epidemiologic studies show associations between NO2 exposure and health effects in
2 various organ systems, and associations are observed with a similar range of short-term
3 and long-term NC>2 concentrations as respiratory effects (Table 1-1). However, compared
4 to respiratory effects, there is more uncertainty in relationships with NC>2 exposure,
5 largely in identifying an independent effect from other traffic-related pollutants. For some
6 health effects, epidemiologic findings also are inconsistent. A common source of
7 uncertainty across nonrespiratory health effects is the limited availability of controlled
8 human exposure and toxicological studies to inform understanding of how
9 ambient-relevant exposures to NCh may affect biological processes that underlie the
10 health effects observed beyond the respiratory system. NO2 itself is not likely to enter the
11 blood (Section 4.2.2). Among the various products of NO2 reactions that occur in the
12 epithelial lining fluid of the respiratory tract, nitrite has been identified in the blood.
13 However, nitrite produced from inhaled NC>2 may not appreciably alter levels derived
14 from diet or induce potentially detrimental health effects (Section 4.2.3). Nitrite can react
15 with red blood cell hemoglobin to form methemoglobin. Methemoglobin has been linked
16 with health effects but has not been found with ambient-relevant NO2 exposure
17 concentrations (Section 4.3.4.1). A recent controlled human exposure study suggests that
18 mediators from the respiratory tract may spillover into the blood. This spillover could
19 lead to systemic inflammation and oxidative stress (Figure 1-2. Section 4.3.5). providing
20 a potential mechanism by which NC>2 exposure could lead to health effects beyond the
21 respiratory system.
Cardiovascular and Related Metabolic Effects
22 Although it is not clear how inhaled NC>2 affects underlying biological pathways,
23 epidemiologic evidence indicates associations of short-term and long-term NC>2 exposure
24 with cardiovascular and related metabolic effects. For both short-term and long-term NCh
25 exposure, the 2008 ISA for Oxides of Nitrogen concluded that evidence was inadequate
26 to infer a causal relationship with cardiovascular effects. There was a lack of supporting
27 evidence for long-term NCh exposure. There was supporting evidence for short-term NC>2
28 exposure but uncertainty about potential confounding by traffic-related copollutants. In
29 this ISA, the health effect category is expanded to include recent studies of metabolic
30 effects, most of which are for long-term NC>2 exposure. New findings relating long-term
31 NC>2 exposure to the development of diabetes and heart disease and additional findings
32 relating short-term NCh exposure to the triggering of myocardial infarction support a
33 suggestive, but not sufficient, to infer a causal relationship with cardiovascular and
34 related metabolic effects (Table 1-1). Evidence is inconsistent for the effects of
January 2015 1-22 DRAFT: Do Not Cite or Quote
-------
1 short-term and long-term NC>2 exposure on cardiovascular effects, such as arrhythmia,
2 cerebrovascular diseases, and hypertension. There still is uncertainty whether NC>2
3 exposure has effects that are independent of other traffic-related pollutants.
4 Recent epidemiologic studies continue to indicate that short-term NCh exposure may
5 trigger a myocardial infarction. There are consistent findings for associations between
6 short-term increases in ambient NC>2 concentration and hospital admissions or ED visits
7 for myocardial infarction, angina, and their underlying cause, ischemic heart disease
8 (Section 5.3.12. Table 5-58). Coherence is found with epidemiologic evidence for
9 NO2-related ST segment changes, a nonspecific marker of myocardial ischemia, and
10 increases in cardiovascular mortality, of which ischemic heart disease is the leading cause
11 (Finegold et al.. 2013). The robustness of epidemiologic findings is demonstrated as
12 associations consistently observed in studies conducted over several years, in diverse
13 geographic locations, and with data pooled from multiple cities. Also, as with findings for
14 asthma exacerbation (Section 1.5.1). associations of short-term NC>2 exposure with
15 effects related to myocardial infarction persist with adjustment for meteorology,
16 long-term time trends, and a copollutant such as PMio, 862, or Os (Section 5.3.12.1).
17 Most of the epidemiologic evidence is based on NCh exposures assigned as the average
18 ambient concentration across multiple monitors within a city; however, ST segment
19 changes are associated with outdoor residential NO2, which may better represent
20 temporal changes in subjects' personal exposures.
21 New epidemiologic evidence for increases in diabetes and heart disease in relation to
22 long-term NC>2 exposure is suggestive, but not sufficient, to infer a causal relationship
23 (Section 6.3.9. Table 6-11). The study reviewed in the 2008 ISA observed a weak
24 association with cardiovascular events. The most consistent recent findings are for
25 diabetes. Similar to asthma development, diabetes is associated with ambient NO2
26 concentrations estimated at subjects' homes using LUR, which may capture differences
27 in ambient NO2 exposure among subjects. Most studies examine concurrent 1-yr avg NCh
28 concentrations, but some aim to represent longer exposures more relevant to disease
29 development by examining people who did not change residence. There is also some
30 support for heart disease and mortality from ischemic heart disease related to long-term
31 NO2 exposure. Heart disease is associated with 1- or 2-yr avg NO2 concentrations
32 estimated at a neighborhood scale from central site monitors or dispersion models or at
33 subjects' homes with LUR. Most studies assess heart disease by acute cardiovascular
34 events such as myocardial infarction or hospital admissions without considering the
35 potential influence of short-term NC>2 exposure. In some cases, exposures are assessed for
36 periods after the cardiovascular event, and it is uncertain the extent to which these
37 periods represent exposures during disease development. In addition to assessing
38 residential NC>2 exposures, many studies of heart disease and diabetes are noteworthy for
January 2015 1-23 DRAFT: Do Not Cite or Quote
-------
1 their large sample sizes, prospective follow-up of subjects (up to 20 years), and
2 adjustment for potential confounding by age, sex, SES, and comorbid conditions.
3 Despite the epidemiologic evidence relating short-term and long-term NCh exposure to
4 cardiovascular and related metabolic effects, studies do not adequately account for
5 potential confounding by other traffic-related pollutants, as was the case in the 2008 ISA
6 (U.S. EPA. 2008). In limited examination of copollutant models with PM2 5, UFP, or CO,
7 associations of short-term NO2 exposure with effects related to myocardial infarction are
8 not consistently observed. Confounding by other traffic-related pollutants has not been
9 examined. Also in contrast with findings for asthma exacerbation (Section 1.5.1).
10 copollutant model results are based on NO2 and copollutant concentrations measured at
11 central site monitors. Differential exposure measurement error may limit the reliability of
12 copollutant model results. Studies of long-term NO2 exposure and heart disease and
13 diabetes do not examine potential confounding by stress or traffic-related copollutants.
14 Evidence for NO2 associations that are independent of noise also is limited.
15 New findings from experimental studies point to the potential for NO2 exposure to induce
16 cardiovascular effects and diabetes but are not sufficient to address the uncertainties in
17 the epidemiologic evidence. Consistent with findings that reactive products of inhaled
18 NO2 or mediators of inflammation may spill over from the respiratory tract to the blood
19 (Figure 1-2). some recent experimental studies find increases in mediators of
20 inflammation and oxidative stress in the blood or heart tissue of healthy humans and
21 rodent models in response to short-term NO2 exposure (Section 5.3.12.1). Evidence does
22 not strongly support the involvement of neural reflexes as examined by changes in
23 respiratory rate or decreases in heart rate variability (Figure 1-2. Sections 4.3.2.2 and
24 5.3.12.2). Findings for increases in inflammation and oxidative stress describe early,
25 nonspecific changes induced by NO2 exposure that have the potential to lead to
26 myocardial infarction. Although the findings are mostly for single-day exposures, they
27 also may describe a possible way for recurrent NO2 exposures to lead to the development
28 of heart disease or diabetes. Limited findings of dyslipidemia in rats and epidemiologic
29 findings of vascular damage in adults in relation to long-term NO2 exposure also describe
30 potential pathways for NO2 exposure to lead to heart disease. The limited extent and
31 consistency of findings from experimental studies and nonspecific nature of most of the
32 evidence is not sufficient to demonstrate an independent effect of NO2 exposure.
33 In conclusion, evidence is suggestive but not sufficient to infer causal relationships for
34 cardiovascular and related metabolic effects with both short-term and long-term NO2
35 exposure. Conclusions were changed from the 2008 ISA based on more epidemiologic
36 evidence linking myocardial infarction to short-term exposure and new evidence linking
37 heart disease and diabetes to long-term exposure. However, an independent effect of NO2
January 2015 1-24 DRAFT: Do Not Cite or Quote
-------
1 exposure is not clearly demonstrated. Examination of confounding by other traffic-related
2 pollutants is absent for long-term NCh exposure and gives inconsistent results for
3 short-term NC>2 exposure. Some but not all recent experimental studies show that
4 short-term NC>2 exposure increases inflammation and oxidative stress in the blood or
5 heart tissue. Increases in inflammation and oxidative stress describe a potential way for
6 short-term or long-term NCh exposure to lead to cardiovascular and related metabolic
7 effects, but because the findings are not linked to any specific health effect, unlike the
8 mode of action evidence for asthma exacerbation or development (Section 1.5.1). they do
9 not rule out chance, confounding, and other biases in the epidemiologic evidence.
Total Mortality
10 Similar to the evidence described above for cardiovascular and related metabolic effects,
11 epidemiologic evidence supports associations of both short-term and long-term NCh
12 exposure with total mortality from all nonaccidental causes. However, potential
13 confounding by other traffic-related pollutants remains largely unresolved, and it is not
14 clear what biological processes NO2 exposure may affect to lead to mortality. This
15 uncertainty weighed with the supporting epidemiologic evidence is the basis for
16 concluding that evidence is suggestive, but not sufficient, to infer a causal relationship for
17 both short-term and long-term NC>2 exposure with total mortality (Table 1-1). For
18 short-term exposure, the nature of the evidence has not changed substantively, resulting
19 in the same conclusion as the 2008 ISA. For long-term NCh exposure, whereas evidence
20 in the 2008 ISA was limited, inconsistent, and inadequate to infer a causal relationship,
21 several recent epidemiologic studies report associations with total mortality, supporting a
22 stronger causal determination.
23 Evidence is suggestive, but not sufficient, to infer a causal relationship between
24 short-term NC>2 exposure and total mortality based on consistent epidemiologic findings
25 across geographic locations, including several studies pooling data across cities
26 (Section 5.4.8. Table 5-63). Ambient NC>2 exposures were assessed as the average
27 concentration across central site monitors within a city, which has uncertainty in
28 adequately representing the temporal pattern in personal NC>2 exposures. Similar to
29 findings for asthma exacerbation (Section 1.5.1). associations with mortality persist with
30 adjustment for meteorological factors, long-term time trends, and a copollutant among
31 PMio, SC>2, or Os. A multicontinent study suggests interaction between NC>2 and PMio,
32 with higher PMio-mortality associations observed for periods of higher ambient NC>2
33 concentrations. However, in contrast with asthma exacerbation, potential confounding of
34 associations between short-term NCh exposure and total mortality by traffic-related
35 copollutants remains unexamined.
January 2015 1-25 DRAFT: Do Not Cite or Quote
-------
1 The generally supportive evidence from the large number of recent epidemiologic studies
2 is suggestive, but not sufficient, to infer a causal relationship between long-term NC>2
3 exposure and total mortality (Section 6.5.3. Table 6-18). Epidemiologic associations are
4 observed in large cohorts in diverse locations followed for long durations up to 26 years.
5 Increases in total mortality are found in association with NC>2 concentrations averaged
6 over 1 to 16 years and assessed for the year of death and for periods up to 20 years before
7 death. Not all studies observe associations, but the inconsistency does not appear to be
8 due to differences among studies in long-term average ambient NO2 concentrations or the
9 exposure period examined. Total mortality is associated with long-term NCh exposure
10 assigned from central site monitors and exposures estimated at people's homes by LUR
11 models that well represented the spatial variability in ambient NO2 concentrations in the
12 study areas (R2 = 0.61 and 0.71). NC>2 associations persist with adjustment for potential
13 confounding by age, sex, smoking, education, and comorbid conditions. In a few studies,
14 associations between long-term NC>2 exposure and mortality persist with adjustment for
15 traffic density or proximity, but confounding by traffic-related copollutants remains a
16 concern because NO2 associations are inconsistently observed with adjustment for PM2 5
17 or BC exposures estimated from central site monitors or LUR models.
18 Evidence relating NC>2 exposure to cardiovascular and respiratory effects can inform the
19 uncertainty as to whether NC>2 exposure has an independent effect on mortality by
20 indicating whether NC>2 exposure affects the underlying causes of mortality. In the U.S.,
21 cardiovascular disease, namely ischemic heart disease, is the leading cause of death [35%
22 as cited in (Hoyert and Xu. 2012)]. Respiratory causes comprise a smaller fraction of
23 mortality (9% in the U.S.), but COPD and respiratory infections are among the leading
24 causes of all mortality in the world. As described in the preceding sections, independent
25 effects of short-term and long-term ambient NO2 exposure on myocardial infarction,
26 heart disease, diabetes, COPD, and respiratory infection are uncertain. Strong evidence
27 demonstrates NO2-related asthma exacerbation, but asthma is not a leading cause of
28 mortality. Thus, it is not clear what spectrum of cardiovascular and respiratory effects
29 NO2 exposure may induce to lead to mortality and by what biological processes
30 short-term or long-term NO2 exposure may lead to mortality.
31 In conclusion, evidence is suggestive, but not sufficient, to infer a causal relationship for
32 total mortality with both short-term and long-term NO2 exposure based on supporting
33 epidemiologic evidence. The evidence bases for total mortality related to short-term and
34 long-term NO2 exposure share many characteristics. Although there is supporting
35 epidemiologic evidence, studies do not adequately account for potential confounding by
36 other traffic-related pollutants. Thus, it is uncertain the extent to which epidemiologic
37 findings for total mortality can be attributed specifically to short-term or long-term NO2
38 exposure. Also uncertain are the independent effects of NO2 exposure on the
January 2015 1-26 DRAFT: Do Not Cite or Quote
-------
1 cardiovascular and respiratory morbidity conditions that make up the leading causes of
2 mortality. Because potential confounding by traffic-related copollutants is largely
3 unaddressed and the biological processes underlying the effects of NO2 exposure on
4 mortality are unclear, the epidemiologic associations of short-term and long-term NO2
5 exposure with total mortality do not rule out chance, confounding, and other biases.
Reproductive and Developmental Effects
6 The 2008 ISA for Oxides of Nitrogen concluded that evidence was inadequate to infer a
7 causal relationship between NO2 exposure and a heterogeneous group of reproductive and
8 developmental effects based on limited and inconsistent epidemiologic and animal
9 toxicological evidence for effects on birth outcomes. This ISA presents separate
10 conclusions for more defined categories of outcomes that are likely to occur by different
11 biological processes and exposure patterns over different stages of development:
12 (1) fertility, reproduction, and pregnancy (Section 6.4.2); (2) birth outcomes
13 (Section 6.4.3); and (3) postnatal development (Section 6.4.4). For all three categories,
14 there is a large increase in recent epidemiologic studies. However, only for birth
15 outcomes is there reasonable consistency in the findings to support strengthening the
16 causal determination to suggestive, but not sufficient, to infer a causal relationship with
17 long-term NO2 exposure (Table 1-1). For all three categories of reproductive and
18 developmental effects, there is large uncertainty in identifying an independent effect of
19 NO2 exposure. In particular, animal toxicological evidence to support biological
20 plausibility remains limited and inconclusive.
Fertility, Reproduction, and Pregnancy
21 Evidence is inadequate to infer a causal relationship between long-term NO2 exposure
22 and effects on fertility, reproduction, and pregnancy (Section 6.4.5. Table 6-14). This
23 conclusion is based heavily on findings from the epidemiologic studies of pre-eclampsia,
24 a pregnancy complication related to hypertension and protein in the urine (Table 1-1).
25 Associations are inconsistently observed with for ambient NO2 exposures estimated at
26 homes by LUR models that well predicted ambient NO2 concentrations in the study areas
27 (R2 = 0.59 to 0.86). Studies that observe associations considered confounding by
28 maternal age, smoking, SES, diabetes, and parity, but few examine other traffic-related
29 pollutants to assess the potential for confounding. Other lines of evidence to inform
30 biological plausibility are not available. Toxicological studies have not examined effects
31 related to pre-eclampsia, and there is a lack of coherence with epidemiologic findings for
32 conditions that contribute to pre-eclampsia, such as gestational hypertension and
33 placental function. Inconsistent and limited findings from animal toxicological and/or
34 epidemiologic studies for detrimental effects on sperm quantity and quality, fertility,
January 2015 1-27 DRAFT: Do Not Cite or Quote
-------
1 maternal weight gain in pregnancy, and litter size add to the uncertainty regarding a
2 relationship of NC>2 exposure with fertility, reproduction, and pregnancy.
Birth Outcomes
3 Evidence is suggestive, but not sufficient, to infer a causal relationship between NO2
4 exposure and effects on birth outcomes based primarily on recent epidemiologic
5 associations with fetal growth restriction (Section 6.4.5. Table 6-14). The combined
6 epidemiologic and toxicological findings for effects on birth weight and infant mortality
7 are inconsistent as are epidemiologic findings for preterm birth and birth defects.
8 Evidence for NC^-related decreases in fetal growth is not entirely consistent, but many
9 studies observe associations with ambient NCh concentrations at homes estimated by
10 LUR models that well predict NO2 concentrations in the study areas (R2 = 0.68 to 0.91;
11 Table 1-1). A few studies observe stronger associations for children whose mothers spent
12 more time at home and less time outdoors in locations other than home, which may be the
13 result of improved relationships of residential ambient NC>2 with personal exposures.
14 Other strengths of recent studies include fetal or neonatal physical measurements and
15 analysis of confounding by season of conception, maternal age, smoking, SES, and in one
16 study, noise. However, epidemiologic studies do not examine potential confounding by
17 traffic-related copollutants. Further, toxicological studies have not examined fetal growth,
18 and a mode of action for NO2 is not clearly described (Figure 1-2). Prenatal ambient NO2
19 exposure is associated with a marker of inflammation in fetal cord blood but not maternal
20 blood. The role of inflammation in affecting birth outcomes is not clearly established, and
21 epidemiologic findings do not rule out effects of other pollutants. Thus, despite the
22 supporting evidence for fetal growth restriction, there is considerable uncertainty in
23 attributing epidemiologic findings specifically to NO2 exposure.
Postnatal De velopment
24 Evidence is inadequate to infer a causal relationship between NO2 exposure and effects
25 on postnatal development based largely on the inconclusive findings across several recent
26 epidemiologic studies of cognitive function in children (Section 6.4.5. Table 6-14).
27 Associations are inconsistently found for concurrent, infancy, or prenatal NO2 exposure
28 estimated at children's homes or schools with LUR models that well represent the
29 variability in ambient NO2 concentrations in the study areas (R2 = 0.64 to 0.85;
30 Table 1-1). Further, potential confounding by traffic-related copollutants or stress is not
31 examined, although one study shows an association with decreases in memory, adjusting
32 for noise. The recent study indicating that short-term NO2 exposure of adult rats induced
33 oxidative stress and neuronal degeneration, which potentially could lead to impaired
34 cognitive function, is not sufficient to address the uncertainties in epidemiologic findings.
35 Findings for other effects on postnatal development are both limited and inconsistent.
January 2015 1-28 DRAFT: Do Not Cite or Quote
-------
1 Specifically, evidence integrated from epidemiologic and toxicological studies is
2 inconclusive for motor function and psychological or emotional distress. Evidence is
3 inconsistent for decrements in attention and limited for autism as examined in
4 epidemiologic studies and for physical development as examined in toxicological studies.
Cancer
5 The largest evidence base in support of a relationship between NO2 exposure and cancer
6 is that for lung cancer (Table 1-1). A few recent epidemiologic studies indicate
7 associations between NO2 exposure and leukemia, bladder cancer, prostate cancer, and
8 breast cancer, but findings for NC>2 exposure inducing carcinogenicity or mutagenicity in
9 bone marrow, spermatocytes, and lymphocytes is inconsistent and based on higher than
10 ambient-relevant NC>2 exposures. The findings from some recent epidemiologic studies
11 for associations of NO2 exposure with lung cancer and mortality combined with previous
12 findings in rodents that NC>2 exposure may be involved in lung tumor promotion is the
13 basis for strengthening the causal determination from inadequate to infer a causal
14 relationship in the 2008 ISA for Oxides of Nitrogen to suggestive, but not sufficient, to
15 infer a causal relationship (Section 6.6.9. Table 6-20).
16 Among the many recent epidemiologic studies, some report associations for NC>2 with
17 lung cancer incidence or mortality; others do not. Findings are inconsistent for NC>2
18 exposure assessed from central site monitors and estimated at subjects' homes with
19 well-validated LUR models. In studies observing associations, NCh concentrations were
20 averaged over 1 year at the beginning of the study up to 30 years before the outcome.
21 Thus, in many cases, the exposure period is considered to be relevant for cancer.
22 However, it is not clear whether exposures assessed by LUR or dispersion models for
23 periods of 1 or 5 years before cancer or mortality adequately account for decreases in
24 ambient NC>2 concentration over years or represent longer duration exposures. Studies not
25 finding associations do not differ in mean NC>2 concentrations or exposure duration
26 examined. Many studies examined large numbers of cancer cases, followed adults for
27 7-30 years, and adjusted for potential confounding by SES, smoking, diet, and
28 occupational exposures. One study observes an association of residential NO2 exposure
29 with lung cancer mortality that persists with adjustment for PM2 5, but examination of
30 confounding by traffic-related copollutants is largely absent.
31 NO2 exposure does not independently induce lung tumor formation in various animal
32 models or transform other chemicals in the body into carcinogens at ambient-relevant
33 concentrations. However, a potential role for high-concentration exposures in tumor
34 promotion is indicated by findings of NO2 exposures of 4,000 to 10,000 ppb increasing
35 lung tumors in mice with spontaneously high tumor rates or with co-exposure to diesel
January 2015 1-29 DRAFT: Do Not Cite or Quote
-------
1 exhaust particles or known carcinogens. The formation of secondary oxidation products
2 in the respiratory tract (Section 1.5.1) and limited evidence for NC>2-induced increases in
3 hyperplasia of the lung epithelium of rodents are early events that have the potential to
4 mediate NCh-related lung cancer. While NO2 exposure impairs host defense in animal
5 models (Section 5.2.9). parameters more directly linked to antitumor immunity such as
6 cytotoxic or regulatory T cells and interferon-gamma have not been studied.
7 In conclusion, evidence is suggestive, but not sufficient, to infer a causal relationship
8 between long-term NC>2 exposure and cancer based primarily on lung cancer.
9 Associations between ambient NO2 concentrations and lung cancer incidence and
10 mortality are found in some but not all epidemiologic studies. NC>2 exposures in rodents,
11 some at higher than ambient-relevant concentrations, show an effect on lung tumor
12 promotion, but experimental studies do not support direct effects of NCh exposure on
13 carcinogenesis. Because potential confounding by traffic-related copollutants is largely
14 unaddressed and information to support biological plausibility is limited, the associations
15 for long-term NC>2 exposure with lung cancer incidence and mortality observed in some
16 epidemiologic studies do not rule out chance, confounding, and other biases.
January 2015 1-30 DRAFT: Do Not Cite or Quote
-------
Table 1-1 Key evidence contributing to causal determinations for nitrogen dioxide (NO2) exposure and health
effects evaluated in the current draft Integrated Science Assessment (ISA) for Oxides of Nitrogen.
Health Effect Category3 and Causal Determination13
NO2 Concentrations
Associated with Effects
Respiratory Effects and Short-term Exposure (Section 5.2)
Current Draft ISA—Causal relationship. 2008 ISA—Sufficient to infer a likely causal relationship.
Key evidence
(Table 5-45)
Reason for change in
causal determination
Uncertainty remaining
Strongest evidence is for effects on asthma exacerbation. Consistent epidemiologic evidence for
decreases in lung function and increases in respiratory symptoms in children with asthma and
increases in asthma hospital admissions and ED visits. Associations observed with NO2 measured at
central site monitors and at subjects' locations (i.e., personal ambient, outdoor school). Copollutant
models show NO2 associations that are independent of PlVh.s or as examined in fewer studies, EC/BC,
OC, UFP, VOCs, PM metals with pollutants measured at subjects' locations, or CO measured at
central site monitors. NO2 associations persist with adjustment for meteorology, medication use, PM-io,
SO2, or Os. Coherent findings available for total personal and indoor NO2 with lower potential for
copollutant confounding.
Independent effect of NO2 demonstrated in controlled human exposure studies. In adults with asthma,
NO2 exposures not much higher than peak ambient concentrations induce clinically-relevant increases
in airway responsiveness and increases in allergic responses, which are part of the mode of action for
asthma exacerbation. Inconsistent experimental results for effects on lung function and respiratory
symptoms in absence of challenge agent.
Uncertainty in independent effect of NO2 on other respiratory effects (i.e, allergy exacerbation, COPD
exacerbation, respiratory infection, respiratory effects in healthy populations) due to limited coherence
among findings from epidemiologic and experimental studies.
Epidemiologic evidence for NO2 exposures assessed for subject's locations and in copollutant models
with a traffic-related copollutant plus evidence from experimental studies describing mode of action
demonstrate consistency, coherence, and biological plausibility for effect of NO2 exposure on asthma
exacerbation to rule out chance, confounding, and other biases with reasonable confidence
Strength of inference from copollutant models about independent associations of NO2, especially with
pollutants measured at central site monitors. Potential for NO2-copollutant mixture effects.
Overall study ambient
maximums
Central site monitors:
24-h avg: 55 to 80 ppb
1-h max: 59 to 306 ppb
Outdoor school:
24-h avg: 7.5, 16.2 ppb
Personal ambient:
2-havg: 77.7, 154 ppb
Total personal:
24-h avg: 48, 106 ppb
Airway responsiveness:
200 to 300 ppb for 30 min,
100 ppb for 1 h
Allergic inflammation:
260 for 15 min and
581 ppb for 30 min
January 2015
1-31
DRAFT: Do Not Cite or Quote
-------
Table 1-1 (Continued): Key evidence contributing to causal determinations for nitrogen dioxide (NC>2) exposure and health
effects evaluated in the current draft Integrated Science Assessment (ISA) for Oxides of Nitrogen.
Health Effect Category3 and Causal Determination13
NO2 Concentrations
Associated with Effects
Respiratory Effects and Long-term Exposure (Section 6.2)
Current draft ISA—Likely to be a causal relationship. 2008 ISA—Suggestive but not sufficient to infer a causal relationship.
Key evidence
(Table 6-5)
Reason for change in
causal determination
Strongest evidence is for effects on asthma development. Consistent epidemiologic evidence from
recent cohort studies for associations of ambient NO2 averaged over 1-10 years (early childhood,
lifetime) with asthma incidence in children. Associations found with NO2 estimated at homes and
measured at central site monitors. NO2 associations persist with adjustment for SES and smoking
exposure. Potential confounding by traffic-related copollutants or proximity to roads not examined.
Small body of experimental studies show effects on hallmarks of asthma. Long-term NO2 exposure
increases allergic responses and airway responsiveness in rodents. Short-term NO2 exposure induces
development of allergic responses in humans and rodents. Inconsistent epidemiologic associations
between long-term NO2 exposure and development of allergic responses in children.
More uncertainty in relationships with other respiratory effects because of limited coherence among
disciplines. Epidemiologic evidence for increased severity of respiratory disease and decreased lung
function and lung development in children. Animal toxicological evidence for respiratory infection.
New epidemiologic evidence for associations between estimates of residential ambient NO2 exposure
and asthma development and biological plausibility from a small body of experimental studies.
Uncertainty remaining
Some uncertainty remains in identifying an independent effect of NO2 exposure from traffic-related
copollutants because evidence from experimental studies for effects related to asthma development is
limited, and epidemiologic analysis of confounding is lacking.
Overall study ambient
means: 14 to 28 ppb for
residential annual avg
estimates
Individual city ambient
means: 9.6-51.3 ppb for
annual avg; 7.3-31.4 ppb
for 10-yr avg
Allergic responses: 2,000
ppb for 4 days in humans;
3,000 ppb for 2 weeks and
4,000 ppb for 12 weeks in
rodents
Airway responsiveness:
1,000 to 4,000 ppb in
rodents for 6 or 12 weeks
Cardiovascular and Related Metabolic Effects and Short-term Exposure (Section 5.3)
Current Draft ISA - Suggestive, but not sufficient, to infer a causal relationship. 2008 ISA - Inadequate to infer a causal relationship.
Key evidence
(Table 5-58)
Reason for change in
causal determination
Strongest evidence is for effects related to triggering myocardial infarction. Consistent epidemiologic
evidence for ST segment changes, increases in hospital admissions and ED visits for myocardial
infarction and ischemic heart disease, and cardiovascular mortality. Most evidence is based on NO2
averaged across central site monitors in a city. Associations persist with adjustment for meteorology,
PM-io, SO2, or 03. NO2 associations inconsistent in copollutant models with PM2.5 or CO.
Some, but not entirely consistent findings, from experimental studies for early, nonspecific effects with
the potential to lead to myocardial infarction: increases in markers of inflammation and oxidative stress
in plasma of humans and heart tissue of rats. Inconsistent epidemiologic findings for inflammation.
Inconsistent evidence for effects on cerebrovascular effects, arrhythmia, and hypertension.
Additional epidemiologic evidence for array of effects related to the triggering of myocardial infarction.
Uncertainty remaining
Effect of NO2 independent from traffic-related copollutants is uncertain because experimental evidence
is limited and not specific to myocardial infarction, and epidemiologic analysis of confounding is limited.
Potential exposure error associated with NO2 measured at central site monitors not well characterized.
Individual city ambient
24-h avg: 90th: 22 to 53
ppb, Maximums: 58 to 135
ppb
Overall study ambient
1-h max: 90th: 68 ppb
Oxidative stress in rats:
5,320 ppb, for 6 h
inflammation in rats: 2,660
and 5,320 ppb for 6 h
Inflammation in human
cells exposed to human
plasma, oxidative stress in
human plasma: 500 ppb for
2h
January 2015
1-32
DRAFT: Do Not Cite or Quote
-------
Table 1-1 (Continued): Key evidence contributing to causal determinations for nitrogen dioxide (NC>2) exposure and health
effects evaluated in the current draft Integrated Science Assessment (ISA) for Oxides of Nitrogen.
Health Effect Category3 and Causal Determination13
NO2 Concentrations
Associated with Effects
Cardiovascular and Related Metabolic Effects and Long-term Exposure (Section 6.3)
Current Draft ISA—Suggestive, but not sufficient, to infer a causal relationship. 2008 ISA—Inadequate to infer a causal relationship.
Key evidence:
(Table 6-11)
Reason for change in
causal determination
Strongest evidence is for development of diabetes and heart disease. Generally supportive but not
entirely consistent epidemiologic evidence from recent cohort studies for associations of diabetes,
myocardial infarction, and heart failure with ambient NO2 averaged over 1-2 year periods around time
of outcome assessment. Coherence with evidence for cardiovascular mortality. Associations found
with NO2 estimated at homes and measured at central site monitors. NO2 associations persist with
adjustment for age, sex, SES, comorbid conditions, and in a few cases, noise. Potential confounding
by traffic-related copollutants, proximity to roads, or stress not examined.
Some but not entirely consistent findings from experimental studies for early, nonspecific effects with
the potential to lead to heart disease or diabetes: dyslipidemia in rats with long-term NO2 exposure,
increases in markers of inflammation and oxidative stress in plasma of humans and heart tissue of rats
with short-term NO2 exposure. Inconsistent epidemiologic associations between long-term NO2
exposure and inflammation.
Large increase in recent epidemiologic studies of heart disease and diabetes, with generally supportive
but not entirely consistent evidence. New evidence for estimates of residential NO2 exposure.
Uncertainty remaining
Effect of NO2 independent from traffic-related copollutants is uncertain because experimental evidence
is limited and not specific to heart disease or diabetes, and epidemiologic analysis of confounding is
lacking.
Overall study ambient
means: 4.2 to 31.9 ppb for
residential annual avg
estimates; 34 ppb for 9.5-yr
avg at central site monitors
Dyslipidemia in rats:
160 ppb for 32 weeks
Oxidative stress in rats:
5,320 ppb for 6 h,
inflammation in rats: 2,660
and 5,320 ppb for 6 h
Inflammation in human
cells exposed to human
plasma, oxidative stress in
human plasma: 500 ppb for
2h
Total Mortality and Short-term Exposure (Section 5.4)
Current Draft ISA and 2008 ISA—Suggestive, but not sufficient, to infer a causal relationship.
Key evidence:
(Table 5-63)
Reason for no change
in causal determination
Uncertainty remaining
Consistent epidemiologic evidence for increases in total mortality in association with NO2 averaged
across central site monitors in a city. Associations persist with adjustment for meteorology, long-term
time trends, PM-io, SO2, or Os. Potential confounding by traffic-related copollutants not examined.
Evidence does not clearly describe independent NO2 effects on biological processes leading to
mortality. Large percentage of mortality is due to cardiovascular causes, for which independent effect
of NO2 is uncertain. The strongest evidence for respiratory morbidity is for asthma and is more limited
or inconsistent for COPD and respiratory infection, which are larger causes of mortality in adults.
Effect of NO2 independent from traffic-related copollutants is uncertain because epidemiologic analysis
of confounding is lacking, and independent effect of NO2 on biological processes (i.e., effects on
morbidity) that lead to mortality not clearly demonstrated. Potential exposure error associated with NO2
measured at central site monitors not well characterized.
Individual city ambient
24-h avg maximums: 55 to
135 ppb
Individual city ambient
1-h max:
90th: 33 to 133 ppb
Maximums: 96 to 147 ppb
January 2015
1-33
DRAFT: Do Not Cite or Quote
-------
Table 1-1 (Continued): Key evidence contributing to causal determinations for nitrogen dioxide (NC>2) exposure and health
effects evaluated in the current draft Integrated Science Assessment (ISA) for Oxides of Nitrogen.
Health Effect Category3 and Causal Determination13
NO2 Concentrations
Associated with Effects
Total Mortality and Long-term Exposure (Section 6.5)
Current Draft ISA—Suggestive, but not sufficient, to infer a causal relationship. 2008 ISA—Inadequate to infer a causal relationship
Key evidence:
(Table 6-18)
Reason for change in
causal determination
Generally supportive but not entirely consistent epidemiologic evidence from recent cohort studies,
including those with extended follow-up (up to 26 years) of existing cohorts. Associations found with
NO2 averaged over 1 to 16 years for periods 0 to 20 years before death. Most evidence is based on
NO2 measured at central site monitors, but associations also observed with NO2 estimated at homes.
Associations found with adjustment for age, sex, smoking, education, comorbid conditions and in some
cases, neighborhood-level SES. In limited analysis, NO2 associations persist with adjustment for traffic
proximity or density but mostly are attenuated in copollutant models with PIVh.sor BC.
Evidence does not clearly describe independent NO2 effects on biological processes leading to
mortality. Large percentage of mortality is due to cardiovascular causes, for which independent effect
of NO2 is uncertain. The strongest evidence for respiratory morbidity is for asthma and is more limited
or inconsistent for COPD and respiratory infection, which are larger causes of mortality in adults.
Large increase in recent epidemiologic studies, with generally supportive but not entirely consistent
evidence. New evidence for estimates of residential NO2 exposure in some but not all recent studies.
Overall study ambient
means:
12.1 to21.7ppbfor
residential annual avg
estimates
13.9 to 33.6 ppb for 1-yrto
15-yr avg at central site
monitors
Uncertainty remaining
Effect of NO2 independent from traffic-related copollutants is uncertain because epidemiologic analysis
of confounding is limited and inconclusive, and independent effect of NO2 on biological processes (i.e.,
effects on morbidity) that lead to mortality not clearly demonstrated. Potential exposure error
associated with NO2 measured at central site monitors not well characterized.
Reproductive and Developmental Effects Long-term Exposure0
2008 ISA—Inadequate to infer a causal relationship for broad category.
Fertility, Reproduction, and Pregnancy (Section 6.4.2)
Current Draft ISA—Inadequate to infer a causal relationship.
Key evidence
(Table 6-14)
Reason for no change
in causal determination
Uncertainty remaining
Heterogeneous group of outcomes related to a successful pregnancy with little support for relationship
with NO2 exposure. Inconsistent epidemiologic evidence among several recent studies for associations
of pre-eclampsia, increases in blood pressure, and systemic inflammation in pregnancy with NO2
estimated at homes with LUR or measured at central site monitors. Studies adjust for maternal age,
smoking, SES, diabetes, and parity. Lack of toxicological studies to inform a potential effect of NO2.
More limited and inconsistent epidemiologic evidence for effects on fertility. No effect on fertility in
rodents, but change in estrous cyclicity found. No epidemiologic or toxicological evidence for effects on
sperm count or quality. Limited, inconclusive evidence in rodents for changes in pregnancy weight.
Increase in recent epidemiologic studies, but results lack sufficient consistency, including those for
residential estimates of NO2 exposure. Limited and inconclusive toxicological evidence does not
adequately inform a potential effect of NO2
Overall study ambient
mean for pre-eclampsia:
31 ppb for residential 3rd
trimester avg estimate
January 2015
1-34
DRAFT: Do Not Cite or Quote
-------
Table 1-1 (Continued): Key evidence contributing to causal determinations for nitrogen dioxide (NC>2) exposure and health
effects evaluated in the current draft Integrated Science Assessment (ISA) for Oxides of Nitrogen.
Health Effect Category3 and Causal Determination13
NO2 Concentrations
Associated with Effects
Birth outcomes (Section 6.4.3)
Current Draft ISA—Suggestive, but not sufficient, to infer a causal relationship.
Key evidence
(Table 6-14)
Reason for change in
causal determination
Strongest evidence is for fetal growth restriction. Generally supportive but not entirely consistent recent
epidemiologic evidence for decreased head circumference and fetal or birth length, particularly as
assessed with fetal or neonatal physical measurements. Associations found with NO2 estimated at
homes and measured at central site monitors. NO2 associations persist with adjustment for maternal
age, SES, smoking, alcohol use, and season of conception. Potential confounding by traffic-related
copollutants not examined, and no available toxicological studies to inform a potential effect of NO2.
Evidence for decreased birth weight in a study of rats, but large epidemiologic evidence base is
inconsistent. Inconsistent epidemiologic evidence for associations with preterm birth, birth defects,
early life mortality, and no or inconclusive toxicological evidence to inform a potential effect of NO2.
Large increase in epidemiologic studies, with generally supportive but not entirely consistent evidence
for associations between residential ambient NO2 exposure and fetal growth restriction.
Overall study ambient
means:
Entire pregnancy: 15.5 to
20 ppb
Specific trimesters: 7.8 to
36 ppb
Decreased birth weight in
rats: 1,300 ppb for 3 mo
Uncertainty remaining
Effect of NO2 independent from traffic-related copollutants is uncertain because evidence from
experimental studies and epidemiologic analysis of confounding are lacking.
Postnatal development (Section 6.4.4)
Current Draft ISA—Inadequate to infer a causal relationship
Key evidence
(Table 6-14)
Reason for no change
in causal determination
Uncertainty remaining
Inconsistent epidemiologic evidence from recent studies for associations with neurodevelopmental
effects such as cognitive function, attention, motor function, and emotional responses. Association
found with indoor NO2, but not consistently with ambient NO2 exposure estimated at home or school by
LUR. Associations found with adjustment for SES and in one study, noise. Potential confounding
inconsistently examined for smoking and not examined for stress or traffic-related copollutants.
Limited and inconclusive toxicological evidence for effects on motor function and emotional responses.
In a study of adult rats, short-term NO2 exposure induced neurodegeneration and oxidative stress,
which have the potential to lead to neurodevelopmental effects.
Limited and inconclusive toxicological evidence for impaired physical development in rats, and no
analogous epidemiologic investigation.
Large increase in epidemiologic studies of cognitive function, but results lack sufficient consistency,
including those for residential or school estimates of NO2 exposure. Limited and inconclusive
toxicological evidence does not adequately inform a potential effect of NO2
Overall study ambient
means for cognitive
function:
16.5 ppb for concurrent
school annual avg estimate
15.4 ppb for prenatal home
annual avg estimate
Neurodegeneration in rat
brains: 2,500 ppb for
7 days
Oxidative stress in rat
brains: 5,320 ppb for
7 days
January 2015
1-35
DRAFT: Do Not Cite or Quote
-------
Table 1-1 (Continued): Key evidence contributing to causal determinations for nitrogen dioxide (NC>2) exposure and health
effects evaluated in the current draft Integrated Science Assessment (ISA) for Oxides of Nitrogen.
Health Effect Category3 and Causal Determination13
NO2 Concentrations
Associated with Effects
Cancer and Long-term Exposure (Section 6.6)
Current Draft ISA—Suggestive, but not sufficient, to infer a causal relationship. 2008 ISA—Inadequate to infer a causal relationship.
Key evidence
(Table 6-20)
Reason for change in
causal determination
Uncertainty remaining
Strongest evidence is for lung cancer. Inconsistent epidemiologic evidence from several recent cohort
studies followed for 7-30 years for associations of lung cancer incidence and mortality with NO2
exposures averaged over 1 to 30 years. Inconsistency observed for NO2 estimated at homes and
measured at central site monitors. Associations persist with adjustment for smoking, diet, SES, and
occupational exposures, but confounding by traffic-related copollutants not widely examined.
Lack of toxicological evidence for direct effect of NO2 in lung tumor induction, but findings for high NO2
exposures with co-exposure to carcinogens suggest possible role for NO2 in lung tumor promotion.
Evidence for formation of secondary oxidation products in the respiratory tract and limited evidence for
hyperplasia of lung epithelium, which have the potential to lead to carcinogenicity.
Limited epidemiologic evidence for associations with cancers of other sites, but inconsistent findings
for mutagenic and genotoxic effects in experimental animals to support independent effect of NO2.
Evidence in some but not all epidemiologic studies for lung cancer incidence and mortality, including
associations with residential estimates of NO2 exposure. Some toxicological evidence for role of NO2 in
lung tumor promotion
Effect of NO2 independent from traffic-related copollutants is uncertain because epidemiologic analysis
of confounding and results from experimental studies that NO2 acts as a direct carcinogen are lacking.
Overall study ambient
means:
12.1 to 23.2 ppb for
residential annual avg
estimates
Individual city ambient
means:
1.2 to 33.7 ppb for 10-yr
avg at central site monitors
6.4 to 32.4 ppb for 3-yr avg
at central site monitors
Lung tumor promotion in
rodents: 4,000 to
10,000 ppb for 6 to
17 months
BC = black carbon; CO = carbon monoxide; COPD = chronic obstructive pulmonary disease; EC = elemental carbon; ED = emergency department; ISA = Integrated Science
Assessment; NO2 = nitrogen dioxide; O3 = ozone; OC = organic carbon; PM2.5 = particulate matter with an aerodynamic diameter less than or equal to a nominal 2.5 |jm;
PM10 = particulate matter with an aerodynamic diameter less than or equal to a nominal 10 |jm; SES = socioeconomic status; SO2 = sulfur dioxide; UFP = ultrafine particles;
VOC = volatile organic compound.
aA large spectrum of outcomes is evaluated as part of a broad health effect category including physiological measures (e.g., airway responsiveness, lung function), clinical
outcomes (e.g., respiratory symptoms, hospital admissions), and cause-specific mortality. Total mortality includes all nonaccidental causes of mortality and is informed by the nature
of the evidence for the spectrum of morbidity effects (e.g., respiratory, cardiovascular) that can lead to mortality. The sections and tables referenced include a detailed discussion of
the available evidence that informed the causal determinations.
bSince the completion of the 2008 ISA for Oxides of Nitrogen, the phrasing of causal determinations has changed slightly, and the weight of evidence that describes each level in
the hierarchy of the causal framework has been more explicitly characterized.
°ln the 2008 ISA, a single causal determination was made for the broad category of reproductive and developmental effects. In the current draft ISA, separate causal determinations
are made for smaller subcategories of reproductive and developmental effects based on varying underlying biological processes and exposure patterns over different lifestages.
January 2015
1-36
DRAFT: Do Not Cite or Quote
-------
1.6 Policy-Relevant Considerations
1 As described in the Preamble and Section 1.1. this ISA informs policy-relevant issues
2 that are aimed at characterizing quantitative aspects of relationships between ambient
3 NO2 exposure and health effects and the impact of these relationships on public health.
4 To that end, this section integrates information from the ISA to describe NCh exposure
5 durations and patterns related to health effects, the shape of the concentration-response
6 relationship, regional heterogeneity in relationships, the adverse nature of health effects,
7 and at-risk populations and lifestages. In addressing these policy-relevant issues, this
8 section focuses on respiratory effects, for which the evidence indicates there is a causal
9 and likely to be a causal relationship, respectively, with short-term and long-term NCh
10 exposure. Because of uncertainty in the independent effects of NO2 exposure, other
11 health effects are discussed if they potentially provide new insight on a particular issue.
1.6.1 Durations of Nitrogen Dioxide Exposure Associated with Health Effects
12 The primary NC>2 NAAQS are based on 1-h daily max concentrations (3-yr avg of each
13 year's 98th percentile) and annual average concentrations and were set to protect against
14 a broad range of respiratory effects associated with short-term NC>2 exposures and various
15 health effects potentially associated with long-term exposure, respectively (Section 1.1).
16 Thus, an important consideration in the review of the primary NCh NAAQS is whether
17 the nature of the health effects evidence varies among NO2 exposure durations.
18 For short-term exposure, the majority of previous and recent evidence associates health
19 effects with 24-h avg ambient NC>2, but a smaller body of evidence is equally consistent
20 for subdaily averages such as 1-h or 8-h max NC>2 and NC>2 averaged over periods of 2 or
21 5 hours. The 24-h avg and 1-h max ambient NC>2 metrics, assessed primarily from
22 concentrations averaged across multiple monitors within a city, are associated with a
23 spectrum of effects related to asthma exacerbation. In the few within-study comparisons
24 and based on typical increases in 24-h avg and 1-h max ambient NC>2 concentrations (20
25 and 30 ppb, respectively; Section 5.1.2.3). the magnitude of association with respiratory
26 effects did not clearly differ between 24-h avg and 1-h max NC>2 (Sections 5.2.2 and
27 5.2.7). A study of asthma-related ED visits in Atlanta, GA observed similar associations
28 for 1-h max and 24-h avg NCh with a 1-day lag, and a slightly larger association for
29 6-h nighttime avg NC>2 (12:00 a.m.-6:00 a.m.; Section 5.2.2.4). Based on measurements
30 from central site monitors, the distribution of concentrations and spatial heterogeneity
31 varied among the array of NC>2 averaging times, which may account for differences in
32 associations with asthma ED visits. For example, nighttime avg NC>2 had a wider range of
January 2015 1-37 DRAFT: Do Not Cite or Quote
-------
1 concentrations than 24-h avg NC>2. Nighttime avg NC>2 was similar to 1-h max NO2 in
2 spatial heterogeneity but lower in concentration. The spatial heterogeneity in ambient
3 NO2 concentrations within urban areas and with distance to roads (Sections 2.5.2 and
4 2.5.3) and diurnal trends with higher concentrations measured during morning commute
5 hours (Section 2.5.4) are not unique to Atlanta, GA. This heterogeneity in ambient NCh
6 concentrations along with diurnal variation in people's time-activity patterns suggest that
7 the array of NC>2 averaging times vary in the extent to which they represent people's
8 exposures, which could obscure true differences in association with health effects
9 NC>2 measurements aligned with subjects' locations including total and ambient personal,
10 outdoor and indoor school, and indoor home NC>2 are associated with asthma-related
11 effects (Sections 5.2.2.2 and 5.2.2.5) and mostly are integrated over 1 or multiple days.
12 From these integrated exposure metrics, any diurnal pattern of NC>2 exposure that may
13 underlie associations with asthma-related effects cannot be discerned. The relative
14 importance of daily average exposures or acute peaks in exposure occurring as a result of
15 diurnal variation in ambient concentrations is not clear. Any contribution of acute peaks
16 in indoor NCh exposures (Table 3-4) to associations observed between 3-day or 4-week
17 avg indoor NC>2 and asthma-related effects also is not known. However, NO2 exposures
18 of 2 or 5 hours during time spent in outdoor traffic and nontraffic locations are related to
19 pulmonary inflammation and lung function decrements in adults (Section 5.2.9.3).
20 Inference from these results is strong because they are based on personal ambient NCh
21 measurements or NC>2 measured at the locations of outdoor exposures. Controlled human
22 exposure studies showing clinically relevant increases in airway responsiveness
23 (Section 5.2.2.2) and allergic inflammation (Section 5.2.2.5) in adults with asthma in
24 response to 100-400 ppb NC>2 exposures in the range of 30 minutes to 6 hours provides
25 biological plausibility for subdaily ambient NC>2 exposures inducing asthma exacerbation.
26 With respect to long-term exposure, asthma development in children is associated with
27 1-yr avg ambient NC>2 concentrations estimated at children's homes and 3-yr or 10-yr avg
28 NC>2 concentrations at central site monitors (Section 6.2.2.1). The NC>2 concentrations
29 averaged over 1 year during prenatal or infancy periods could represent critical time
30 windows of exposure for asthma development or represent longer durations of NC>2
31 exposure for subjects who remain in the same home or neighborhood. Experimental
32 studies do not provide direct insight into what the epidemiologic findings may be
33 indicating are important periods of long-term NC>2 exposure for asthma development
34 because experimental studies examined NO2 exposures of less than one year in
35 adulthood. However, findings for increased allergic responses and airway responsiveness
36 in humans or rodents indicate that repeated increases in NC>2 exposure over multiple days
37 or exposures over 1 to 3 months may play a role in asthma development (Section 6.2.2.3).
January 2015 1-38 DRAFT: Do Not Cite or Quote
-------
1 Overall, asthma exacerbation and asthma development are linked to a range of short-term
2 and long-term, respectively, durations of NC>2 exposure. There is no indication of a
3 stronger association for any particular short-term or long-term duration of NC>2 exposure.
1.6.2 Lag Structure of Relationships between Nitrogen Dioxide Exposure and
Health Effects
4 Characterizing the NC>2 exposure lags (i.e., time between exposure and effect) associated
5 with health effects can aid in understanding the nature of relationships between NO2
6 exposure and health effects. The lag structure for associations with NO2 exposure may
7 vary among health effects depending on differences in the time course by which
8 underlying biological processes occur. Identifying important lag structures can depend on
9 whether the lag structure varies within the population according to differences among
10 individuals in time-activity patterns, pre-existing disease, or other factors that influence
11 exposure and responses to exposure. Another consideration in drawing inferences about
12 important lag structures is that differences in associations among exposure lags,
13 particularly single-day and multiday averages NC>2 concentrations, may not only have a
14 biological basis but may be influenced by differences in the extent to which single-day
15 and multiday average ambient NO2 concentrations represent people's acutal exposures.
16 Epidemiologic panel studies of children with asthma observed increases in pulmonary
17 inflammation and respiratory symptoms and decreases in lung function in association
18 with increases in NC>2 concentration lagged 0 day (same day as outcome) or 1 day and
19 multiday averages of 2 to 7 days (Section 5.2.2). Consistent with these findings, increases
20 in asthma-related hospital admissions and ED visits were observed in association with
21 NO2 concentrations lagged 0 or 1 day or averaged over 2 to 5 days. Whereas no particular
22 lag of NO2 exposure was more strongly associated with decreases in lung function,
23 several studies indicate larger increases in pulmonary inflammation, respiratory
24 symptoms, and asthma-related hospital admissions and ED visits for increases in
25 multiday averages of NC>2 than single-day lags. For measures of personal ambient and
26 total NC>2, outdoor school NC>2, and indoor NC>2, which may better represent exposure
27 compared with measurements from central site monitors, asthma-related effects also were
28 associated with multiday average NO2 concentrations (i.e., 2 to 4 days).
29 Studies in which adults with asthma and healthy adults were exposed for 2 or 5 hours in
30 outdoor traffic and nontraffic locations indicate decreases in lung function and increases
31 in pulmonary inflammation immediately or 2 hours after exposures (Sections 5.2.2 and
32 5.2.7). In both populations, decreases in lung function also were found the day after
33 exposures. In healthy adults, increases in pulmonary inflammation did not persist the day
January 2015 1-39 DRAFT: Do Not Cite or Quote
-------
1 after outdoor exposure (Section 5.2.7.4). These data based on personal ambient exposure
2 assessment or NCh measured at the locations of people's outdoor exposures support other
3 epidemiologic findings showing increases in respiratory effects at lag 0 or 1 day of NC>2
4 exposure and also indicate a similar lag structure for respiratory effects in people with
5 and without asthma. Experimental studies show that NC>2 exposure affects the biological
6 processes underlying the asthma-related effects observed in epidemiologic studies on a
7 similar time frame. Controlled human exposure studies found airway responsiveness in
8 adults with asthma to increase immediately after or 20 minutes to 4 hours after a single
9 NC>2 exposure and over 4 days of repeated exposure (Section 5.2.2.1). In experimental
10 studies, NCh exposure enhanced allergic inflammation 30 minutes up to 19 hours after a
11 single or 2-day exposure in humans and 7 days after exposure in rats (Section 5.2.2.5).
12 Thus, the findings from experimental studies provide biological plausibility for the
13 asthma-related effects observed in epidemiologic studies in association with 2 or 5 hour
14 exposures, same-day NO2 exposures, as well as exposures averaged over multiple days.
1.6.3 Concentration-Response Relationships and Thresholds
15 Characterizing the shape of the concentration-response relationship aids in quantifying
16 the public health impact of NC>2 exposure. A key issue is whether the relationship is
17 linear across the full range of ambient concentrations or whether there are deviations
18 from linearity at and below the levels of the current 1-h NAAQS of 100 ppb and annual
19 NAAQS of 53 ppb. Also important for the review of the primary NC>2 NAAQS is
20 understanding at what ambient NC>2 concentrations there is uncertainty in the relationship
21 with health effects. Characterization of the concentration-response relationship in
22 epidemiologic studies is complicated by fewer observations in the low range of ambient
23 concentrations, the influence of other pollutants or risk factors for the health effects, and
24 heterogeneity among individuals in the population in their response to air pollution
25 exposures. The shape of the concentration-response relationship for health effects related
26 to short-term NC>2 exposure is examined in a limited number of epidemiologic studies
27 and is better characterized for respiratory hospital admissions and ED visits and total
28 mortality than for other health effects.
29 Recent U.S. studies support a linear relationship between short-term NCh exposure and
30 asthma ED visits (Section 5.2.2.4). For 1-h max NC>2 concentrations (lag 0-2 day avg)
31 combined across urban monitors by population-weighting, a linear association was
32 observed with pediatric asthma ED visits in Atlanta, GA during 1993-2004. Also, risk
33 estimates increased across quintiles of NC>2 between 28 and 181 ppb (with concentrations
34 less than 28 ppb as the reference). Results from nonparametric models provide evidence
35 of asthma ED visits in the warm season (May-October) increasing with increasing
January 2015 1-40 DRAFT: Do Not Cite or Quote
-------
1 1-h max NC>2 concentrations between 11 and 37 ppb (5th to 95th percentiles). There is
2 similar confidence in the relationship throughout this range of NC>2 concentrations, with a
3 relatively tight 95% CI even at 11 ppb NC>2. There is uncertainty in the relationship at
4 1-h max NC>2 concentrations less than 11 ppb, where effect estimates were reported to be
5 unstable. The distribution of 1-h max NCh varied across monitors in the Atlanta area,
6 with higher concentrations measured in downtown Atlanta (mean 42 ppb). However, the
7 population-weighted average may better represent concentrations where study subjects
8 live and spend time. For the relationship between 24-h avg NC>2 and pediatric asthma ED
9 visits in Detroit, MI during 2004-2006, evidence does not indicate that associations
10 differ below and above 23 ppb NO2 (between the 82nd and 85th percentiles), where the
11 investigators tested for a deviation from a linear relationship. The risk was not assumed
12 to be zero below 23 ppb, and the model assuming a deviation from linearity did not fit the
13 data better than the linear model did. NO2 concentrations were averaged between two
14 Detroit sites, and information on how ambient NCh concentrations compared between
15 sites was not reported. These limited findings from Atlanta and Detroit indicate that the
16 association between short-term NCh exposure and asthma ED visits in children is present
17 at NC>2 concentrations well below the level of the current 1-h primary NAAQS.
18 The concentration-response relationship for short-term NCh exposure and asthma-related
19 effects is not well examined in controlled human exposure or animal toxicological
20 studies. Combining data across multiple studies, a recent meta-analysis observed that
21 NC>2 exposure cut in half the dose of the challenge agent required to induce an increase in
22 airway responsiveness (i.e., provocative dose) in adults with asthma, but the provocative
23 dose did not change with increasing NC>2 concentration in the range of 100-500 ppb
24 ( Figure 5-1). Experimental studies do not strongly inform whether asthma-related effects
25 increase with increasing NC>2 concentration because few studies examined multiple NCh
26 exposure concentrations, and the range of these NC>2 concentrations (greater than
27 100 ppb) exceed those examined in epidemiologic studies of concentration-response.
28 Linear concentration-response relationships also are indicated for mortality associated
29 with short-term NC>2 exposure in the U.S., Canada, and Asia based on comparisons of
30 linear and various nonlinear models with natural and cubic splines or quadratic and cubic
31 terms for NC>2 (Section 5.4.7). A few previous results point to nonlinear associations but
32 for health effects for which the concentration-response relationship has not been widely
33 examined, including cough in children in the general population or cardiovascular
34 hospital admissions in adults. These studies tend to find NC^-related increases in effects
35 that are larger in magnitude per increment in NC>2 concentration in the lower range of
36 NO2 concentrations than in the upper range of concentrations. The implications of results
37 for these nonasthma health effects is less clear given the uncertainty as to whether NO2
38 exposure has independent relationships with nonasthma health effects.
January 2015 1-41 DRAFT: Do Not Cite or Quote
-------
1 For long-term NC>2 exposure, there is evidence for a linear concentration-response
2 relationship with respiratory effects, although from only one or two studies per specific
3 outcome. For asthma development, in analyses of tertiles or quartiles of residential
4 estimates of NO2 exposure, a linear concentration-response is indicated in one study but
5 not another (Section 6.2.2.2). In the study observing a linear relationship, annual average
6 NO2 concentrations ranged from 1.8 to 24 ppb, but because tertiles of NO2 concentration
7 were not reported, the range of NC>2 concentrations where there may be more or less
8 uncertainty in the relationship with asthma development cannot be assessed. Also based
9 on categories of NCh concentration or splines, linear relationships are indicated for
10 associations of long-term averages of NO2 with asthma symptoms in children, chronic
11 bronchitis in adults, and asthma hospital admissions in adults (Section 6.2.3). but some of
12 these findings could reflect associations with short-term NC>2 exposure. Analysis of the
13 concentration-response with categories of long-term average NC>2 concentrations does not
14 provide a strong basis for assessing whether there is a threshold for respiratory effects.
15 In summary, the shape of the concentration-response relationship is better characterized
16 in epidemiologic studies and for short-term NC>2 exposure than long-term exposure.
17 Previous and recent evidence indicates a linear relationship between short-term NCh
18 exposure and hospital admissions or ED visits for asthma and multiple respiratory
19 conditions combined using various methods, including analysis of splines, higher order
20 terms for NC>2 (e.g., quadratic, cubic), and categories of NC>2 concentration. Few
21 controlled human exposure or toxicological studies of asthma-related effects examined
22 multiple NC>2 exposure concentrations; therefore, that evidence lacks strong insight into
23 the concentration-response relationship. Associations with asthma ED visits are present at
24 ambient NC>2 concentrations well below the level of the 1-h NAAQS. In Atlanta, GA, a
25 linear relationship is indicated for 1-h max NC>2 concentrations averaged over 3 days,
26 with similar confidence in the relationship across the range of 11 to 37 ppb but
27 uncertainty in the relationship at concentrations less than 11 ppb. Another source of
28 uncertainty is that 24-h avg or 1-h max NCh concentrations were averaged across
29 multiple central site monitors within a city, which may not reflect varying distributions of
30 concentrations within the city or population exposures.
1.6.4 Regional Heterogeneity in Effect Estimates
31 In addition to examining the shape of the concentration-response relationship for
32 NO2-related health effects across the distribution of concentrations, studies have
33 examined whether associations vary across geographical regions. In one study,
34 heterogeneity was noted among Asian cities in the shape of the NC^-mortality
35 relationship. Information on regional heterogeneity is limited, particularly for the U.S.
January 2015 1-42 DRAFT: Do Not Cite or Quote
-------
1 and for relationships of NO2 exposure with asthma exacerbation or development. There is
2 no strong indication of heterogeneity in associations of short-term NO2 exposure with
3 respiratory symptoms in children in the general population among Korean cities
4 (Section 5.2.7.3). A few studies observe regional heterogeneity in associations between
5 short-term NO2 exposure and total mortality among European and Asian cities
6 (Section 5.4.7). A nonlinear concentration-response relationship observed in one of four
7 Asian cities was hypothesized to be due to differences among cities in mortality from
8 infection, air conditioning use, time spent by the population outdoors, or temperature. On
9 a smaller geographic scale, NO2-related respiratory effects do not clearly differ between
10 two cities in Ohio with similar ambient NO2 concentrations (Section 5.2.2.4) or
11 neighboring urban and suburban communities in Europe that differed in ambient NO2
12 concentrations (Section 7.5.5). Limited results point to potential within-city differences in
13 asthma exacerbation in relation to short-term NO2 exposure. NO2-related asthma ED
14 visits were larger in Bronx than Manhattan, NY (Section 5.2.2.4). and NO2-related lung
15 function and pulmonary inflammation among children with asthma differed between two
16 El Paso, TX schools (Sections 5.2.2.2 and 5.2.2.5). The reasons for the heterogeneity
17 were not explicitly analyzed. In the El Paso study the schools differed in proximity to
18 road, ambient NO2 concentrations, racial composition, and asthma medication use.
19 For long-term NO2 exposure, differences are observed among Chicago, Houston, San
20 Francisco, New York, and Puerto Rico in the association with asthma prevalence among
21 Latino and African American individuals ages 8-21 years (Section 6.2.2.1). A test for
22 heterogeneity was not statistically significant, but associations are observed only in the
23 San Francisco and New York cohorts. Odds ratios for the average ambient NO2
24 concentration for the first year or first 3 years of life are largest in the San Francisco
25 cohort, which comprised only African American individuals. Associations are not
26 observed in Chicago, Puerto Rico, or Houston. The reasons for heterogeneity among the
27 locations were not explicitly analyzed, but the locations differed in the distribution of
28 ambient NO2, SO2, and PM2 5 concentrations, which may indicate varying air pollution
29 mixtures among locations. San Francisco had lower ambient NO2 and SO2 concentrations
30 than New York. PM2 5 and SO2 were associated with asthma prevalence in Houston but
31 not in New York or San Francisco.
32 In summary, with limited available information, including one U.S. study of asthma
33 prevalence, it is not clear whether there is regional heterogeneity in the relationship
34 between short-term or long-term NO2 exposure and respiratory effects. There is some
35 evidence of heterogeneity in associations of short-term NO2 exposure with mortality
36 among cities in Europe and Asia. Given the uncertainty as to whether NO2 exposure has
37 an independent relationship with mortality, the extent to which the regional heterogeneity
38 in risk is applicable specifically to NO2 exposure is uncertain.
January 2015 1-43 DRAFT: Do Not Cite or Quote
-------
1.6.5 Public Health Significance
1 The public health significance of air pollution-related health effects is informed by the
2 adverse nature of the health effects that are observed, the size of the population exposed
3 to the air pollutant or affected by the health outcome, and the presence of populations or
4 lifestages with higher exposure or increased risk of air pollution-related health effects.
Characterizing Adversity of Health Effects
5 Both the World Health Organization (WHO) and the American Thoracic Society (ATS)
6 have provided guidance in describing what health effects may be considered adverse.
7 WHO defines health as "the state of complete physical, mental, and social well-being and
8 not merely the absence of disease or infirmity" (WHO. 1948). By this definition, changes
9 in health outcomes that are not severe enough to result in a diagnosis of a clinical effect
10 or condition can be considered adverse if they affect the well-being of an individual. ATS
11 also has considered a wide range of health outcomes in defining adverse effects.
12 Distinguishing between individual and population risk, ATS described its view that small
13 air pollution-related changes in an outcome observed in individuals might be considered
14 adverse on a population level. This is because a shift in the distribution of population
15 responses resulting from an increase in air pollution exposure might increase the
16 proportion of the population with clinically important effects or at those at increased risk
17 of a clinically important effect that could be caused by another risk factor (ATS. 2000).
18 Increases in ambient NO2 concentrations are associated with a broad spectrum of health
19 effects related to asthma, including those characterized as adverse by ATS such as ED
20 visits and hospital admissions (ATS. 2000). NO2 exposure also is associated with more
21 subtle effects such as increases in airway responsiveness and pulmonary inflammation
22 and decreases in lung function (Section 1.5.1). Increases in airway responsiveness and
23 pulmonary inflammation are part of the mode of action for both asthma exacerbation and
24 asthma development (Figure 1-2) and show a distribution within populations.
25 NO2-associated changes in airway responsiveness or pulmonary inflammation may be
26 considered adverse on a population level because they can increase the proportion of the
27 population with clinically important changes that can lead to exacerbation or
28 development of asthma. A meta-analysis of controlled human exposure studies
29 demonstrates that NO2 exposures of 140-200 ppb for 1-2 hours cuts in half the dose of a
30 challenge agent required to increase airway responsiveness in adults with asthma
31 (Section 5.2.2.1). Such observations that NO2 concentrations not much higher than peak
32 ambient concentrations can induce clinically relevant effects related to asthma
33 exacerbation further support a role for ambient NO2 exposures in inducing adverse health
34 effects.
January 2015 1-44 DRAFT: Do Not Cite or Quote
-------
At-Risk Populations and Lifestages for Health Effects Related to Nitrogen
Dioxide Exposure
1 The primary NAAQS are intended to protect public health with an adequate margin of
2 safety. In so doing, protection is provided for both the population as a whole and those
3 groups potentially at increased risk for health effects from exposure to the air pollutant
4 for which each NAAQS is set (Preface to the ISA). Hence, the public health significance
5 of health effects related to NO2 exposure also is informed by whether specific lifestages
6 or groups in the population are identified as being at increased risk of NCh-related health
7 effects. The large proportion of the U.S. population living near roads, where ambient NCh
8 concentrations are higher compared to many other locations (Section 2.5.3). indicates the
9 widespread potential for elevated ambient NO2 exposures. In 2009, 17% of U.S. homes
10 were estimated to be within 91 m of sources of ambient NO2 such as a four-lane highway,
11 railroad, or airport (Section 7.5.6). The percentage of the population with elevated NO2
12 exposures may be greater in cities. For example, 40% of the Los Angeles, CA population
13 was estimated to live within 100 m of a major road (Section 7.5.6). In addition to those
14 living near roads, people spending time near roads and commuting or working on roads
15 have the potential for elevated NO2 exposure, and in turn, potential for increased risk of
16 NO2-related health effects.
17 At-risk populations or lifestages also can be characterized by specific biological,
18 sociodemographic, or behavioral factors, among others. Since the 2008 ISA for Oxides of
19 Nitrogen and as used in the recent ISAs for O3 (U.S. EPA. 2013b) and lead (U.S. EPA.
20 2013a). EPA has developed a framework for drawing conclusions about the role of such
21 factors in modifying risk of health effects of air pollution exposure (Table III of the
22 Preamble). Similar to the causal framework, conclusions are based on judgments of the
23 consistency and coherence of evidence within and across disciplines (Chapter 7). Briefly,
24 the evaluation is based on studies that compare exposure or health effect relationships
25 among groups that differ according to a particular factor (e.g., people with and without
26 asthma) and experimental studies conducted in a population or animal model with a
27 particular factor or pathophysiological condition. Where available, information on
28 exposure, dosimetry, and modes of action is evaluated to assess coherence with health
29 effects evidence and inform how a particular factor may increase risk of NC>2-related
30 health effects (e.g., by increasing exposure, increasing biological effect for a given dose).
31 There is adequate evidence that people with asthma, children, and older adults are at
32 increased risk for NCh-related health effects (Table 7-26). These conclusions are
33 substantiated by the fact that evidence is for asthma exacerbation, for which an
34 independent effect of short-term NCh exposure is demonstrated (Section 1.5.1). Limited,
35 supporting evidence suggests that females, people of low SES, and people with low
36 antioxidant diets have increased risk for NCh-related health effects. The inconsistent
January 2015 1-45 DRAFT: Do Not Cite or Quote
-------
1 evidence is inadequate to determine whether genetic variants, COPD, cardiovascular
2 disease, diabetes, obesity, race/ethnicity, smoking, urban residence, or proximity to roads
3 increase NO2-related health effects. The uncertainty common to many of these factors is
4 that the evidence is for cardiovascular effects, diabetes, or mortality, which are not
5 clearly related to NO2 exposure.
6 A causal relationship between short-term NC>2 exposure and respiratory effects is based
7 on the evidence for asthma exacerbation (Section 1.5.1). The increased risk for people
8 with asthma is supported further by controlled human exposure studies demonstrating
9 increased airway responsiveness at lower NO2 concentrations in adults with asthma than
10 in healthy adults (Section 7.3.1). Differences in NC>2 dosimetry (Section 4.2.2) or
11 exposure among people with asthma are not well described. Epidemiologic evidence does
12 not consistently indicate differences in NO2-related respiratory effects between children
13 with asthma and without asthma. However, because asthma is a heterogeneous disease
14 and the populations examined varied in prevalence of asthma medication use and atopy,
15 the inconsistent epidemiologic results are not considered to be in conflict with controlled
16 human exposure studies, which examined primarily adults with mild, atopic asthma.
17 The increased risk of NC^-related asthma hospital admissions and ED visits for children
18 (Section 7.5.1.1) and older adults (Section 7.5.1.2) suggests that among people with
19 asthma the effects of NO2 exposure may vary by lifestage. Although not clearly
20 delineated for NC>2, several physiological and behavioral traits may contribute to the
21 increased risk for children. Compared with adults, children have developing respiratory
22 systems and increased oronasal breathing and ventilation rates (Section 4.2.2.3). Limited
23 data do not clearly indicate higher personal NC>2 exposures in children (Table 3-5) but do
24 indicate more time and vigorous activity outdoors (Section 7.5.1.1). Thus, children may
25 have greater NC>2 uptake in the respiratory tract or less exposure measurement error.
26 Many studies reported a higher proportion of asthma ED visits or hospital admissions
27 among children than other lifestages. Thus, higher incidence of asthma exacerbation in
28 children may be a reason for their increased risk.
29 Because the respiratory system continues to develop throughout childhood, it is possible
30 that critical time windows of exposure exist for NC^-related respiratory effects. However,
31 the evidence shows that asthma development in children is associated with several
32 different time windows of long-term NCh exposure: the prenatal or infancy period, the
33 first year of life, year of diagnosis, or lifetime exposure (Section 7.5.1.1). Studies do not
34 consistently identify a specific time window of long-term NC>2 exposure more strongly
35 associated with the development of asthma as ascertained in children ages 4-18 years.
36 Children not only comprise a large proportion of the U.S. population (24% in the 2010
37 U.S. census) but also have a higher rate of asthma health care encounters than adults
January 2015 1-46 DRAFT: Do Not Cite or Quote
-------
1 (e-g-, 10-7 vs. 7.0 per 100 persons with asthma).1 Further, asthma is the leading chronic
2 illness (9.5% prevalence) and reason for missed school days in children in the U.S. Many
3 U.S. schools are located near high-traffic roads (7% within 250 m; Section 7.5.6). NCh
4 concentrations outside schools are associated with asthma-related effects in children
5 (Sections 5.2.2.2 and 5.2.2.5). and school could be an important source of NC>2 exposure.
6 Based on the large number of children in the U.S., the high prevalence of asthma
7 morbidity among children, and potential for high NC>2 exposures, higher risks of asthma
8 exacerbation for children compared with adults can translate into large numbers of people
9 affected, magnifying the potential public health impact of NC>2 exposure.
10 The public health impact of NCh-related health effects also is magnified by the growing
11 proportion of older adults in the U.S. As with children, it is not well understood why
12 older adults have increased risk for NCh-related hospital admissions for asthma. Older
13 adults did not consistently have a higher proportion of asthma hospital admissions
14 compared with younger adults, so higher incidence of asthma exacerbation does not seem
15 to explain their higher NC>2-related risk estimates. Differences in NC>2 dosimetry also are
16 not described for older adults (Section 4.2.2.3). Time-activity patterns have been shown
17 to differ between older and younger adults, but there is not a clear difference in time
18 spent in a particular location that could explain differential exposure to NC>2 in older
19 adults (Section 7.5.1.2). Older adults have higher prevalence of many chronic diseases
20 compared to younger adults (Table 7-2). COPD, cardiovascular diseases, and diabetes did
21 not consistently modify NO2-related health effects, but studies have not examined
22 whether co-occurring morbidity contributes to the increased risk of NO2-related asthma
23 exacerbation among older adults or whether age alone influences risk.
24 Although evidence does not clearly identify increased NO2-related health effects in
25 populations of low SES or nonwhite race or populations living near roads or in urban
26 areas, there is an indication of higher NO2 exposure among these groups. In particular,
27 some communities are characterized as having both higher ambient NO2 concentrations
28 and higher proportions of nonwhite and low SES populations (Section 7.5.2). Further, a
29 few studies characterize schools located near high-traffic roads as having higher
30 nonwhite and low SES populations compared to schools located farther away from roads
31 (Section 7.5.6). Nonwhite and low SES populations also are recognized to have higher
32 risks of health effects, including asthma, though it is not clear whether higher NO2
33 exposure and higher risk of health effects interact to influence NO2-related health effects
34 in these groups. A recent study observed higher risk of NO2-related asthma hospital
35 admissions among Hispanic children compared with white children only in the low SES
36 group (Section 7.5.2). While these findings suggest that co-occurring risk factors in a
National Center for Health Care Statistics Data Brief. Available:
http ://www. cdc. gov/nchs/data/databriefs/db94. htm
January 2015 1-47 DRAFT: Do Not Cite or Quote
-------
1 population could influence the risk of NO2-related health effects, information at present is
2 too limited to draw firm conclusions.
3 In summary, the public health significance of NO2-related health effects is supported by
4 many lines of evidence. A large proportion of the U.S. population lives near roads or
5 spends time near or on roads, resulting in a large number of people potentially with
6 elevated ambient NO2 exposure. NO2 exposure is linked to health effects that are clearly
7 adverse such as ED visits and hospital admissions for asthma and development of asthma.
8 NO2-related effects such as increases in airway responsiveness can be considered adverse
9 on a population level because an increase in NO2 exposure can lead to an increase in the
10 number of people with clinically important effects. The public health significance of
11 NO2-related health effects also is supported by the increased risk for people with asthma,
12 children, and older adults. The roles of co-occurring risk factors or combined higher NO2
13 exposure and health risk within a population in influencing risk of NO2-related health
14 effects is not well understood. The large proportions of children and older adults in the
15 U.S. population and the high prevalence of asthma in children can translate into a large
16 number of people affected by NO2 and thus magnify the public health impact of ambient
17 NO2 exposure.
1.7 Conclusions
18 There is a causal relationship between short-term NO2 exposure and respiratory effects.
19 This conclusion is stronger than that determined in the 2008 ISA for Oxides of Nitrogen
20 and is supported by the evidence integrated from epidemiologic and controlled human
21 exposure studies for asthma exacerbation. Asthma-related effects continue to be
22 associated with NC>2 concentrations at central site monitors. Further, recent epidemiologic
23 studies add evidence for associations with personal ambient and total NO2 measurements
24 as well as NC>2 concentrations outside schools and inside homes. Epidemiologic evidence
25 continues to support independent associations of NC>2 exposure with asthma-related
26 effects in copollutant models with another traffic-related pollutant such as PIVbs, EC/BC,
27 OC, UFP, CO, or a VOC. The potential influence of the full array of traffic-related
28 pollutants or mixtures has not been examined. Thus, the key evidence for an independent
29 effect of NO2 are the findings from previous controlled human exposure studies that NO2
30 exposure not much higher than peak ambient concentrations enhances allergic
31 inflammation and induces clinically relevant increases in airway responsiveness. These
32 are hallmarks of the mode of action for asthma exacerbation. There is likely to be a
33 causal relationship between long-term NO2 exposure and respiratory effects. The
34 strengthening of the conclusion from the 2008 ISA is based on new epidemiologic
35 evidence for associations of asthma development in children with residential NO2
January 2015 1-48 DRAFT: Do Not Cite or Quote
-------
1 exposure estimated by LUR models that well represented the variability in ambient NO2
2 concentrations in study areas. Epidemiologic studies did not examine confounding by
3 traffic-related copollutants, but a small body of previous experimental studies showing
4 that long-term and short-term NC>2 exposure increase airway responsiveness and allergic
5 responses in healthy humans and rodent models provide some indication that long-term
6 NO2 exposure may have an independent effect on asthma development. For both
7 short-term and long-term exposure, results for NC>2 measured in subjects' locations,
8 which may better represent exposure, than concentrations at central site monitors provide
9 a stronger basis for inferring relationships with respiratory effects.
10 Evidence is suggestive, but not sufficient, to infer a causal relationship for short-term and
11 long-term NC>2 exposure with cardiovascular and related metabolic effects and total
12 mortality and for long-term NC>2 exposure with birth outcomes and cancer. While there is
13 continued or new supporting epidemiologic evidence for these health effects, there still is
14 large uncertainty as to whether NC>2 exposure has an effect independent of traffic-related
15 copollutants. Epidemiologic studies have not adequately accounted for confounding, and
16 there is a paucity of support from experimental studies. Some recent experimental studies
17 show NCh-induced increases in systemic inflammation or oxidative stress. Such changes
18 are not consistently observed or necessarily linked to any health effect, unlike the mode
19 of action information available for asthma.The insufficient consistency of epidemiologic
20 and toxicological evidence is inadequate to infer a causal relationship for long-term NC>2
21 exposure with fertility, reproduction, and pregnancy as well as postnatal development.
22 As described above, key considerations in drawing conclusions about relationships
23 between ambient NC>2 exposure and health effects include evaluating the adequacy of
24 NO2 exposure estimates to represent the temporal or spatial patterns in ambient NC>2
25 concentrations in a given area and separating the effect of NC>2 from that of other
26 traffic-related pollutants. In the U.S., although motor vehicle emissions have decreased
27 greatly over the last few decades, vehicles still are the largest single source of ambient
28 NO2 in U.S. population centers and can contribute to spatial and temporal heterogeneity
29 in ambient NC>2 concentrations. Recent information combined with that in the 2008 ISA
30 for Oxides of Nitrogen (U.S. EPA. 2008) shows that ambient NO2 concentrations can be
31 30 to 100% higher at locations within 10-20 m of a road compared with locations farther
32 away. Additionally, the first year of data from the U.S. near-road monitoring network
33 generally show higher NO2 concentrations at near-road monitoring sites than many other
34 sites within a city.
35 As in the 2008 ISA, many studies assess exposure with ambient NO2 concentrations
36 measured at monitors whose siting away from sources likely does not capture the
37 variability in ambient NO2 concentrations within an area. The resulting error in
January 2015 1-49 DRAFT: Do Not Cite or Quote
-------
1 representing temporal variation in short-term exposure and spatial variation in long-term
2 exposure can produce smaller magnitude or less precise associations with health effects.
3 Such findings are similar to those reported in the 2008 ISA (U.S. EPA. 2008). This ISA
4 additionally indicates that error that results from using NCh concentrations at central site
5 monitors to represent long-term exposure in some cases can produce larger health effect
6 estimates compared with residential NO2 exposure metrics from LUR models. Thus,
7 spatial misalignment of study subjects and ambient NC>2 concentrations potentially can
8 overestimate health effect associations with long-term NO2 exposure if the difference in
9 exposure between groups that differ in the health effect systematically is underestimated.
10 Given the potential impact of exposure measurement error, the variable relationships
11 observed between short-term personal and ambient NO2 concentrations, and largely
12 uncharacterized relationships between long-term personal and ambient NCh
13 concentrations, the increase in recent epidemiologic findings for exposures assessed for
14 people's locations (e.g., ambient or total personal, outdoor or indoor home or school)
15 may increase confidence in inferences about relationships between ambient NO2
16 exposure and health effects. Also, data from the near-road monitoring network may help
17 address gaps in the understanding of the variability in ambient NC>2 concentrations and
18 people's exposures within urban areas and the potential importance of the near-road
19 environment as a source of NCh exposure contributing to health effects.
20 In addition to characterizing causality, characterizing quantitative aspects of NCh-related
21 health effects is key to the review of the primary NC>2 NAAQS. Limited investigation
22 indicates a linear relationship for short-term ambient NC>2 exposure with asthma ED
23 visits. There is uncertainty about whether the relationship is present at 1-h max NC>2
24 concentrations far below the level of the current 1-hour NAAQS. Recent evidence
25 continues to indicate that people with asthma, children, and older adults are at increased
26 risk for NC>2-related health effects. While recent evidence points to higher NC>2 exposure
27 among people of low SES or nonwhite race or people living in urban areas or close to
28 roads, it is not clear whether this higher NCh exposure leads to increased health effects.
29 Large numbers of people in the U.S. live near (e.g., within 100 m) or travel on major
30 roads and potentially have elevated exposures to ambient NO2 compared with people
31 away from roads. The large numbers of children and older adults in the U.S. population
32 and the high prevalence of asthma in children can translate into a large number of people
33 potentially affected by NC>2 exposure and thus magnify the public health impact of
34 ambient NO2 exposure.
January 2015 1-50 DRAFT: Do Not Cite or Quote
-------
References for Chapter 1
ATS (American Thoracic Society). (2000). What constitutes an adverse health effect of air pollution? Am
J Respir Crit Care Med 161: 665-673. http://dx.doi.Org/10.1164/ajrccm.161.2.ats4-00
CAA (Clean Air Act). (1990). Clean Air Act, as amended by Pub. L. No. 101-549, section 108: Air quality
criteria and control techniques, http://www.law.cornell.edu/uscode/text/42/7408
Finegold. JA: Asaria. P: Francis. DP. (2013). Mortality from ischaemic heart disease by country, region,
and age: Statistics from World Health Organisation and United Nations. Int J Cardiol 168: 934-945.
http://dx.doi.0rg/10.1016/i.iicard.2012.10.046
HEI (Health Effects Institute). (2010). Traffic-related air pollution: A critical review of the literature on
emissions, exposure, and health effects. (Special Report 17). Boston, MA: Health Effects Institute
(HEI). http://pubs.healtheffects.org/view.php?id=334
Hovert. PL: Xu. J. (2012). Deaths: Preliminary data for 2011. National Vital Statistics Reports 61: 1-51.
U.S. EPA (U.S. Environmental Protection Agency). (1993). Air quality criteria for oxides of nitrogen, vol.
1-3 [EPAReport]. (EPA/600/8-9!/049aF-cF). Research Triangle Park, NC: U.S. Environmental
Protection Agency, Environmental Criteria and Assessment Office.
http://cfpub.epa. gov/ncea/cfm/recordisplav.cfm?deid=40179
U.S. EPA (U.S. Environmental Protection Agency). (2008). Integrated science assessment for oxides of
nitrogen Health criteria [EPAReport]. (EPA/600/R-08/071). Research Triangle Park, NC: U.S.
Environmental Protection Agency, National Center for Environmental Assessment.
http://cfpub.epa.gov/ncea/cfm/recordisplav.cfm7de id=194645
U.S. EPA (U.S. Environmental Protection Agency). (2009). Integrated science assessment for particulate
matter [EPAReport]. (EPA/600/R-08/139F). Research Triangle Park, NC: U.S. Environmental
Protection Agency, National Center for Environmental Assessment.
http://cfpub.epa.gov/ncea/cfm/recordisplav.cfm7de id=216546
U.S. EPA (U.S. Environmental Protection Agency). (2010). Primary national ambient air quality standards
for nitrogen dioxide. Fed Reg 75: 6474-6537.
U.S. EPA (U.S. Environmental Protection Agency). (2012). Notice of workshop and call for information
on integrated science assessment for oxides of nitrogen. Fed Reg 77: 7149-7151.
U.S. EPA (U.S. Environmental Protection Agency). (2013a). Integrated science assessment for lead [EPA
Report]. (EPA/600/R-10/075F). Research Triangle Park, NC: U.S. Environmental Protection Agency,
National Center for Environmental Assessment.
http://cfpub.epa.gov/ncea/cfm/recordisplav.cfm7de id=255721
U.S. EPA (U.S. Environmental Protection Agency). (2013b). Integrated science assessment for ozone and
related photochemical oxidants (final report) [EPAReport]. (EPA/600/R-10/076F). Washington, DC.
http://cfpub.epa.gov/ncea/isa/recordisplav.cfm?deid=247492
U.S. EPA (U.S. Environmental Protection Agency). (2013c). Notice of workshop and call for information
on integrated science assessment for oxides of nitrogen and oxides of sulfur. Fed Reg 78: 53452-
53454.
U.S. EPA (U.S. Environmental Protection Agency). (2013d). Supplemental table Sl-1. Epidemiologic
studies of health effects not evaluated in the IS A for oxides for nitrogen [EPAReport].
U.S. EPA (U.S. Environmental Protection Agency). (2014). Integrated review plan for the primary
national ambient air quality standards for nitrogen dioxide [EPAReport]. (EPA-452/R-14/003).
Research Triangle Park, NC: U.S. Environmental Protection Agency, National Center for
Environmental Assessment.
http://www.epa.gov/ttn/naaqs/standards/nox/data/201406finalirpprimarvno2.pdf
January 2015 1-51 DRAFT: Do Not Cite or Quote
-------
WHO (World Health Organization). (1948). Preamble to the Constitution of the World Health
Organization as adopted by the International Health Conference, New York, 19-22 June, 1946. In
Constitution of the World Health Organization (pp. 2). Geneva, Switzerland.
http ://whqlibdoc.who, int/hist/official records/constitution.pdf
January 2015 1-52 DRAFT: Do Not Cite or Quote
-------
CHAPTER 2 ATMOSPHERIC CHEMISTRY AND
AMBIENT CONCENTRATIONS OF OXIDES OF
NITROGEN
2.1 Introduction
1 This chapter presents concepts and findings relating to emissions sources, atmospheric
2 science, and spatial and temporal concentration patterns for oxides of nitrogen. It is
3 intended as a prologue for detailed discussions on the evidence for human exposure to
4 and health effects of oxides of nitrogen that follow in the subsequent chapters, and as a
5 source of information to help interpret those effects in the context of data about
6 atmospheric concentrations.
7 In the Integrated Science Assessment (ISA), the term "oxides of nitrogen" (NOy) refers
8 to all forms of oxidized nitrogen (N) compounds, including nitric oxide (NO), nitrogen
9 dioxide (NO2), and all other oxidized N-containing compounds formed from NO and
10 NO2. NO and NO2, along with volatile organic compounds (VOCs), are precursors in the
11 formation of ozone (Os) and photochemical smog. NO2 is an oxidant and can react to
12 form other photochemical oxidants such as peroxyacyl nitrates (PANs) and toxic
13 compounds such as nitro-substituted poly cyclic aromatic hydrocarbons (nitro-PAHs).
14 NO2 can also react with a variety of atmospheric species to produce organic and
15 inorganic nitrates, which make substantial contributions to the mass of atmospheric
16 particulate matter (PM) and the acidity of clouds, fog, and rainwater. The abbreviation
17 NOx refers specifically to the sum of NO and NO2. This chapter describes the origins,
18 distribution, and fate of gaseous oxides of nitrogen. Aspects of particulate nitrogen
19 species [such as particulate nitrate (pNOs)] were addressed in the 2009 ISA for
20 Particulate Matter (U.S. EPA. 2009) and will be addressed in the upcoming ISA for
21 Particulate Matter.
2.2 Atmospheric Chemistry and Fate
22 The chemistry of oxidized nitrogen compounds in the atmosphere was reviewed in the
23 2008 ISA for Oxides of Nitrogen—Health Criteria (U.S. EPA. 2008a). The role of NOx
24 in Os formation was reviewed in Chapter 3 of the 2013 ISA for Ozone and Related
25 Photochemical Oxidants (U.S. EPA. 2013b) and has been presented in numerous texts
26 (e-g-, Jacobson. 2002; Jacob. 1999; Seinfeld and Pandis. 1998). The main points from the
January 2015 2-1 DRAFT: Do Not Cite or Quote
-------
1 2008 ISA for Oxides of Nitrogen will be presented here along with updates based on
2 recent material.
3 The overall chemistry of reactive, oxidized nitrogen compounds in the atmosphere is
4 summarized in Figure 2-1. Sources include naturally occurring processes associated with
5 wildfires, lightning, and microbial activity in soils. Anthropogenic sources are dominated
6 by emissions from electricity generating units and motor vehicles. Oxidized nitrogen
7 compounds are emitted to the atmosphere mainly as NO, with only 10% or less emitted
8 as NO2. Further details about the composition of sources is given in Section 2.3. Freshly
9 emitted NO is converted to NO2 by reacting with Os, and NO is recycled during the day
10 by photolysis of NO2. Thus, NO and NO2 are often "lumped" together into their own
11 group or family, which the atmospheric sciences community refers to as NOx (shown in
12 the inner box in Figure 2-1). A large number of oxidized nitrogen species in the
13 atmosphere are formed from the oxidation of NO and NO2. These include nitrate radicals
14 (NOs), nitrous acid (HONO), nitric acid (FINOs), dinitrogen pentoxide (ISbOs), nitryl
15 chloride (C1NO2), peroxynitric acid (FINO/O, PAN and its homologues (PANs), other
16 organic nitrates, such as alkyl nitrates (including isoprene nitrates), and pNOs. These
17 reactive oxidation products are referred to collectively as NOz. All of the species shown
18 within the dashed lines of Figure 2-1 constitute NOy (NOy = NOx + NOz). The boxes
19 labeled "inorganic" and "organic" in Figure 1-1 (see Chapter 1) contain the species
20 shown in the left and right halves of Figure 2-1.
January 2015 2-2 DRAFT: Do Not Cite or Quote
-------
Long-range transport to remote
regions at low temperatures v
CINO, «L A.^
PANS
A
RC(0)00
I
Koprene nitrates
alkyl nltratei
NOZ
NO; - NOy - NOX
3
4
5
6
emissions
Note: The inner shaded box contains NOX (= NO + NO2). The outer box contains other species (NOZ) formed from reactions of NOX.
All species shown in the outer and inner boxes are collectively referred to as NOY by the atmospheric sciences community. MPP
refers to multiphase processes, hi/to a solar photon, M to a species transferring/removing enough energy to cause a molecule to
decompose/stabilize, and R to an organic radical.
Source: National Center for Environmental Assessment.
Figure 2-1 Schematic diagram of the cycle of reactive, oxidized nitrogen
species in the atmosphere.
High NO concentrations found near heavy traffic and in power plant plumes are typically
associated with Os concentrations much lower than in surrounding areas as Os can be
titrated away, or lost, by reacting with NO. In addition, the reaction of NO with OB can
produce appreciable amounts of NO2 rather quickly. For example, 10 ppb NO2 can be
formed in about 20 seconds (for an initial NO concentration of 30 ppb and initial
Os = 40 ppb at 298° K [250C])1. Higher temperatures and concentrations of reactants
1 Sample calculation based on solution to an equation for a second order reaction dx/dt = &([NO]0 - x)([O3]0 - x),
where x is the concentration of each species reacted; k is the rate coefficient for the reaction,
= 3 x l(T12 e(~1500/T)cm3/sec-molecule (Sander etaL 2011): T = emperature in Kelvin; [NO]0 and [O3]0 are initial
concentrations of NO and Os.
January 2015
2-3
DRAFT: Do Not Cite or Quote
-------
1 result in shorter times, while dispersion and depletion of reactants increase this time. A
2 rough estimate of the time for transport away from a broad boulevard is is about a minute
3 (During etal.. 2011). shorter for more open conditions and ranges up to the order of an
4 hour in Midtown Manhattan street canyons (Richmond-Bryant and Reff. 2012). The time
5 for reaction must be compared to the time for mixing away from the road and for
6 replenishment of Os, because it is the interplay between these factors that determines how
7 far NO will travel downwind before it is oxidized. These dependencies imply seasonal
8 variability and also geographic variability in the time scale for the reaction. In general,
9 cooler months present the most favorable conditions for NO to travel further before it is
10 oxidized (lower temperature, decreased vertical mixing of Os to the surface, generally
11 lower Os). At anytime of the year, if loss of Os has been extensive near the surface as
12 happens in many locations at night, then NO could travel a kilometer or more before it is
13 oxidized, resulting in a more uniform downwind distribution of NO2 than if NO were
14 being oxidized right at its source. The NO2 that is formed depletes hydroxyl radicals
15 (OH), so that they cannot oxidize hydrocarbons to continue the cycle of new Os
16 formation. During the day, NO2 photolyzes back to NO within a few minutes, setting up
17 the cycle shown in Figure 2-1. Although the assumption of a photostationary state to
18 describe the relations in the NO/NO2/Os triad might not be strictly valid, several studies
19 [eg., During et al. (2011) and Clapp and Jenkin (2001)] have shown the assumption of a
20 photostationary state can provide a useful approximation of the relationship among these
21 species. Once the sun sets, NO2 no longer photolyzes to reform NO. If very little or no Os
22 is present due to titration in a statically stable, near-surface boundary layer, then NO2
23 accumulates through the night solely from direct emissions.
24 Because of the interplay between dispersion and chemical reaction, the distribution of
25 NO2 downwind of roads would likely differ from that of a traffic pollutant that is present
26 in ambient air mainly as the result of direct emissions. In addition, day-night differences
27 in both transport and chemistry will also result in day-night differences in the patterns of
28 spatial and temporal variability of NO2. Examples of the behavior of NO2 and NOx
29 downwind of streets and highways are examined in Section 2.5.3. In summary, the major
30 influences on NO2 concentrations within and downwind of urban centers are the fraction
31 of emissions of NOx as NO2, dispersion, and the NO/NO2/Os equilibrium, which is
32 established on a time scale of a few minutes during daylight.
33 All the other species mentioned above in the definition of NOy (i.e., NOz) are products of
34 reactions of NO or NO2. Inorganic NOz species are shown on the left side of the outer
35 box and organic species are shown on the right side of the outer box in Figure 2-1.
36 Ammonium nitrate and other inorganic participate species (e.g., sodium [Na+], calcium
37 [Ca2+] nitrates) are formed from species shown on the left side of the figure; organic
38 nitrates are formed from species shown on the right side of Figure 2-1. The conversion of
January 2015 2-4 DRAFT: Do Not Cite or Quote
-------
1 NOx into the inorganic and organic species in the outer box (collectively referred to as
2 NOz) typically takes place on much longer time scales than for interconversions between
3 NO and NCh, e.g., about an hour for conditions in Houston, TX, in April-May of 2009
4 (Ren et al.. 2013) but likely longer in many other areas, especially those at higher
5 latitudes and generally during the cold season. As a result, NOx emitted during morning
6 rush hour by vehicles can be converted almost completely to products by late afternoon
7 during warm, sunny conditions. However, note the conversion of NO2 to HNOs and
8 hence the atmospheric lifetime of NOx depends on the concentration of OH radicals,
9 which in turn depends on the concentration of NO2 [eg., Valin et al. (2013) and Hameed
10 etal. (1979)1.
11 Inorganic NOz species shown on the left side of the outer box of Figure 2-1 include
12 HONO, HNO3, C1NO2, HNO4, and pNO3. Pernitric acid (HNO4) is unlikely to represent
13 an important reservoir for NOx except perhaps under extremely cold conditions. Mollner
14 et al. (2010) identified pernitrous acid (HOONO), an unstable isomer of nitric acid, as a
15 product of the major gas phase reaction forming HNOs. However, because HOONO is
16 unstable, it is also not a substantial reservoir for NOx. Considering the troposphere as a
17 whole, most of the mass of products shown in the outer box of Figure 2-1 is in the form
18 of PAN and HNOs. The stability of PAN at low temperatures allows its transport to
19 remote regions where it has been shown to exert strong influence on the local production
20 of Os [see Fischer et al. (2014) and references therein]. Other organic nitrates (e.g., alkyl
21 nitrates, isoprene nitrates) increase in importance in the planetary boundary layer (PEL),
22 particularly at locations closer to sources (Perring etal.. 2013; Horowitz et al.. 2007;
23 Singh et al.. 2007).
24 In addition to the above compounds, there is a broad range of gas-phase organic nitrogen
25 compounds that are not shown in Figure 2-1. They are emitted by combustion sources
26 and are also formed in the atmosphere from reactions of NO, NO2, and NOs. These
27 compounds include: nitro-aromatics (such as nitrotoluene), nitro-PAHs [such as
28 nitro-naphthalene, e.g., Nishino et al. (2008)1. nitrophenols [e.g., Harrison et al. (2005)1;
29 nitriles [such as ethane-nitrile; e.g., de Gouw et al. (2003)]: and isocyanic acid (Roberts
30 etal.. 2014).
31 Sources of NOx are distributed with height with some occurring at or near ground level
32 and others aloft as indicated in Figure 2-1. NOx emitted by elevated sources can be
33 oxidized to NOz products and/or be transported to the surface, depending on time of day,
34 abundance of oxidants, and strength of vertical mixing. During times of rapid convection,
35 typically in the afternoon on hot sunny days, vertical mixing through the PEL can take
36 place in aboutl hour [see e.g., Stull (2000)] and fresh emissions can be brought rapidly to
37 the surface. After sunset, turbulence subsides and emissions entrained into the nocturnal
January 2015 2-5 DRAFT: Do Not Cite or Quote
-------
1 residual boundary layer, are not mixed downward to the surface. Also, because the
2 prevailing winds aloft are generally stronger than those at the surface, emissions from
3 elevated sources (e.g., the stacks of electrical utilities) can be distributed over a wider
4 area than those emitted at the surface (e.g., motor vehicles). Emissions from elevated
5 sources entrained into the nocturnal residual boundary layer can be transported over long
6 distances, up to a few hundred km overnight depending on location [see e.g., Husar et al.
7 (1978)1. Oxidation of NOx can occur during the night and in the morning in the residual
8 layer prior to its breakup.Turbulence then mixes NOx and its oxidation products
9 downward. Emissions directly into the free troposphere are unlikely except in areas such
10 as the Intermountain West where PEL heights can be <200 m during winter, or even
11 <100 m in some locations. Because people live closer to surface sources such as motor
12 vehicles, they are more likely to be exposed to NO and NO2 from these sources. Thus,
13 atmospheric chemical reactions determine the partitioning of a person's exposure to NO2
14 and its reaction products from different sources, and sources of a person's exposure
15 cannot be judged solely by the source strengths given in the National Emissions
16 Inventory (NEI). Issues related to the transport and dispersion of NOx emitted by traffic
17 are discussed in depth in Section 2.5.3.
18 Oxidized nitrogen compounds are ultimately lost from the atmosphere by wet and dry
19 deposition to the Earth's surface. Soluble species are taken up by aqueous aerosols and
20 cloud droplets and are removed by wet deposition by rainout (i.e., incorporation into
21 cloud droplets that eventually coagulate into falling rain drops). Both soluble and
22 insoluble species are removed by washout (i.e., impaction with falling rain drops, another
23 component of wet deposition), and by dry deposition (i.e., impaction with the surface and
24 gas exchange with plants). NO and NO2 are not very soluble, and therefore wet
25 deposition is not a major removal process for them. However, a major NOx reservoir
26 species, HNOs, is extremely soluble, and its deposition (both wet and dry) represents a
27 major sink for NOy.
28 Many of the species shown in Figure 2-1, including particulate nitrate (pNOs) and gas
29 phase HONO, are formed by multiphase processes. Data collected in Houston as part of
30 TexAQS-II summarized by Olaguer et al. (2009) indicate that concentrations of HONO
31 are much higher than can be explained by gas-phase chemistry and by tailpipe emissions.
32 N2Os is the acid anhydride of HNOs, and its uptake on aqueous aerosol represents a major
33 sink for NOx. The uptake of ^Os by atmospheric aerosols or cloud droplets leads to the
34 loss of Os and NOx and the production of aqueous-phase nitric acid, aerosol nitrate, and
35 gaseous halogen nitrites. Maclntyre and Evans (2010) showed that the sensitivity of key
36 tropospheric species, such as Os, varies from very small to high over the range of uptake
37 coefficients (y) for ^Os obtained in laboratory studies. For example, global Os loss
January 2015 2-6 DRAFT: Do Not Cite or Quote
-------
1 ranges from 0 to over 10%, with large regional variability over the range of reported
2 N2Os uptake coefficients. However, uptake coefficients for ^Os, or yflSbOs), on
3 atmospheric particles are not well defined, largely due to uncertainty and variability in
4 aerosol composition. As noted by Brown and Stutz (2012). yflSbOs) is largest («0.02) for
5 aqueous inorganic aerosols and water droplets, except for nitrate in aerosol, which can
6 reduce y(N2Os) by up to an order of magnitude. The uptake of ^Os by mineral particles
7 could also represent an important removal process. For example, values of yCNbOs) for
8 calcite and Saharan dust (e.g.) of «0.03. However, as noted by Tang et al. (2014) not
9 enough is known to permit a global assessment of the importance of N2Os uptake on
10 mineral surfaces. Organic aerosol and soot can reduce yCN^Os) by two orders of
11 magnitude or more, further complicating the task of assessing the importance of uptake of
12 N2Os on aerosol surfaces.
13 The uptake of N2Os by aqueous aerosols containing chloride (Cl~) and bromide (Br") has
14 been associated with the release of gaseous nitryl chloride (C1NO2) from marine aerosol
15 (sea-spray) (Osthoff et al.. 2008). Nitryl chloride has been found not only in coastal and
16 marine environments, but also well inland. For example, Thornton et al. (2010) found
17 production rates of gaseous C1NO2 near Boulder, CO, from reaction of N2Os with
18 particulate Cl~ at levels similar to those found in coastal and marine environments. They
19 also found that substantial quantities of ^Os are recycled through C1NO2 back into NOx
20 instead of forming HNOs. Nitryl chloride (C1NO2) readily photolyzes to yield Cl and NO2
21 and can represent a significant source of reactive Cl, capable of initiating the oxidation of
22 hydrocarbons (generally with much higher rate coefficients than OH radicals). Riedel
23 etal. (2014) found increases in the production of radicals by 27% and of Os by 15%
24 during the 2010 CalNex [California Research at the Nexus of Air Quality and Climate
25 Change in May to June 2010 in Southern California (Ryerson et al.. 2013)1 field study.
26 However, C1NO2 was found to cause only modest Os increases (e.g., ~1 to 1.5 ppb for
27 nominal Os concentrations between 60 and 85 ppb) in a model study of the Houston, TX,
28 airshed (Simon et al.. 2009). Differences are likely related to differences in the NOx
29 sensitivity of the two airsheds. Therefore, caution is advised in extrapolating results
30 obtained in one airshed to another.
31 The lifetimes of PANs are strongly temperature dependent, but these compounds are
32 stable enough at low temperatures to be transported long distances before they
33 decompose to release NO2. Freshly released NO2 can then participate in Os formation in
34 regions remote from the original NOx source (cf Figure 2-1). Nitric acid (HNOs) acts
35 similarly as a reservoir to some extent, but its high solubility and high deposition rate
36 imply that it is removed from the gas phase faster than PAN; thus HNOs would not be as
37 important as a source of NOx in remote regions. The oxidation of many tropospheric
38 species is initiated by OH radicals during the day. During the night NOs radicals, formed
January 2015 2-7 DRAFT: Do Not Cite or Quote
-------
1 from the reaction of NC>2 and Os, assume the role of dominant oxidant for many species
2 such as biogenic and anthropogenic alkenes; for some species (e.g., dimethyl sulfide),
3 they are the overall dominant oxidant [see, e.g., Brown and Stutz (2012)1. The reaction of
4 NOs with alkenes results in the production of gas phase organic nitrates and secondary
5 organic aerosol formation. Many of the reactions shown in Figure 2-1 occur mainly
6 during the night when NOs radicals are most abundant. For example, the formation of
7 N2Os, which has a short lifetime with respect to photolysis and thermal decomposition, is
8 favored at low temperatures during the night. Many of the reactions of NOs, in addition
9 to those of Os, with alkenes also result in the production of OH and hydroperoxyl (HCh)
10 radicals during the night.
11 Isoprene nitrates (INs) and their reaction products could be important for controlling the
12 abundance of NOx and hence the abundance of Os over the eastern U.S. [e.g., Perring
13 et al. (2009)]. Isoprene nitrates (INs) and their reaction products could also be important
14 for exporting reactive nitrogen species to remote areas. Yields for IN formation from
15 isoprene oxidation have been estimated to range from 4% (Horowitz et al.. 2007) to 6 to
16 12% (Xie etal.. 2013) based on model simulations of data collected during the
17 International Consortium for Atmospheric Research on Transport and Transformation
18 (ICART) campaign in 2004 and from 7 to 12% in laboratory studies (Lockwood etal..
19 2010; Paulot et al., 2009; Perring et al., 2009; Horowitz et al., 2007; von Kuhlmann et al.,
20 2004). The initial step in the production of INs involves the reaction of isoprene with OH
21 radicals to produce isoprene peroxy radicals. Under low NOx conditions, isoprene peroxy
22 radicals will mainly react with HO2 radicals to produce organic peroxides, with smaller
23 amounts of methacrolein, methyl vinyl ketone, and formaldehyde. Under higher NOx
24 conditions, isoprene peroxy radicals can also react with NO resulting in the production of
25 many of the same or similar compounds such as methacrolein and methyl vinyl ketone as
26 well as "first generation" INs. Lifetimes on the order of one to a few hours can be
27 estimated for these first generation INs based on their reactions with OH radicals and Os
28 (Lockwood et al.. 2010; Paulot et al.. 2009). The reaction products can further react with
29 NO (after internal rearrangement) to form secondary organic nitrates such as ethanal
30 nitrate, methacrolein nitrate, propanone nitrate, and methylvinylketone nitrate. Lee et al.
31 (2014) found that the rate coefficient for reaction of the first generation INs with Os is a
32 factor of-100 lower than in earlier studies, leading to greater stability and higher levels
33 of first generation INs, particularly at night. Greater stability leads in turn to increased
34 probability of deposition and transport away from source regions. The second generation
35 organic nitrates are more stable than the first generation INs because they lack a double
36 carbon (C=C) bond. Paulot et al. (2009) estimated the yield of NOx from the destruction
37 of second-generation nitrates to be -55%. Obviously, the relative importance of pathways
38 forming nitrates or other products depends on the ambient concentrations of NO and
39 other oxides of nitrogen for which many key experimental details are still lacking.
January 2015 2-8 DRAFT: Do Not Cite or Quote
-------
1 In addition to oxidation initiated by OH radicals, isoprene is also oxidized by
2 radicals. Rollins et al. (2009) determined a 70% yield of first generation carbonyl nitrates
3 based on experiments in large reaction chambers. These first generation nitrates can be
4 further oxidized by NO, leading to the production of second generation organic (alkyl)
5 nitrates. Mao etal. (2013) estimated that the global mean lifetime is ~5 days for these
6 second generation organic nitrates. Mao etal. (2013) also suggested that the export of INs
7 and other organic nitrates followed by their decomposition is potentially a larger source
8 of NOx to the boundary layer of the western North Atlantic Ocean compared to the
9 export of PANs. It should also be noted that some isoprene nitrates (INs) are low enough
10 in volatility that they can partition to the aerosol phase and form PM [e.g., Rollins et al.
11 (2009)1.
12 Describing Os formation accurately requires detailed knowledge of the chemistry of INs.
13 Regional or global models that assume a lower yield for forming these nitrates and a
14 higher yield for recycling NOx tend to over-predict Os concentrations in areas with high
15 isoprene emissions, such as the Southeast, compared to those that have a higher yield for
16 the formation of these nitrates and/or a lower yield for their recycling back to NOx
17 (U.S. EPA. 2013b). The formation rates and the rates that are assumed to recycle INs and
18 other organic nitrates back to NOx also have implications for calculating the yield of Os
19 from isoprene emissions. For example, Fiore et al. (2005) found a negative dependence of
20 Os production on isoprene emissions in the eastern U.S. in summer, whereas Mao et al.
21 (2013) found a positive yield for Os from isoprene emissions. Xie etal. (2013)
22 determined that the uncertainties in the isoprene nitrates could affect Os production by
23 10% over the U.S. and that uncertainties in the NOx recycling efficiency had a larger
24 effect than the IN yield. These considerations underlie the importance of further
25 laboratory and field studies to more quantitatively determine the response of Os to
26 changes in isoprene emissions at different NOx levels.
27 As mentioned earlier, NO and NO2 are important precursors of Os formation. However,
28 because Os changes in a nonlinear way with changes in the concentrations of its
29 precursors (NOx and VOCs), Os is unlike many other atmospheric species with rates of
30 formation that vary directly with emissions of their precursors. At the low NOx
31 concentrations found in environments ranging from remote continental areas to rural and
32 suburban areas downwind of urban centers, the net production of Os typically increases
33 with increasing NOx. In this low-NOx regime, the overall effect of the oxidation of
34 VOCs is to generate (or at least not consume) radicals, and Os production varies directly
35 with NOx. In the high-NOx regime, NO2 reacts with OH radicals to form HNOs [e.g.,
36 Hameed et al. (1979)1. Otherwise, these OH radicals would oxidize VOCs to produce
37 peroxy radicals, which in turn would oxidize NO to NO2. In this regime, Os production is
38 limited by the availability of radicals (Tonnesen and Jeffries. 1994). and Os shows only a
January 2015 2-9 DRAFT: Do Not Cite or Quote
-------
1 weak dependence on NOx concentrations. Reaction of Os with NO in fresh motor vehicle
2 exhaust depletes Os in urban cores, but Os can be regenerated during transport downwind
3 of urban source areas, and additional chemical production of Os can occur, resulting in
4 higher Os concentrations than found upwind of the urban center. Similar depletion of Os
5 can occur in power plant plumes with subsequent Os regeneration downwind.
6 Brown et al. (2012) conducted a field study comparing nighttime chemistry in the plumes
7 of two power plants in Texas, one with selective catalytic reduction (SCR) NOx
8 emissions controls and the other without these controls. They noted that the plume from
9 the power plant with SCR controls did not have enough NOx to deplete all of the Os
10 present in background air. As a result, almost all of the NOx in the plume was oxidized to
11 NOz species, so the NOx that was oxidized was not available to participate in Os
12 production the next day. This situation contrasts with that in the plume from the power
13 plant without controls. In that plume, there was minimal formation of NOz species.
14 Instead, NOx was more nearly conserved and the NO2 that was formed from the reaction
15 of emitted NO with Os photolyzed the following morning, leading to higher Os formation
16 rates compared to plumes from the plant with controls.
2.3 Sources
2.3.1 Overview
17 Estimated total NOx emissions in the U.S. from all sources decreased by 49% over the
18 period from 1990 to 2013, as shown in Figure 2-2. The NEI is a national compilation of
19 emissions sources collected from state, local, and tribal air agencies as well as emission
20 estimates developed by EPA from collected or estimated data by source sector. Emissions
21 after 2011 for mobile sources and electric utilities are regularly added to the 2011 NEI,
22 but emissions for the other sectors are based on 2011 estimates. Through this process
23 some of the major sectors in the 2011 NEI have emission estimates more recent than
24 2011, while emissions from other source sectors are based on 2011 data. When emissions
25 for later years from these sources are added, the inventory is still referred to as a version
26 of the 2011 NEI.
January 2015 2-10 DRAFT: Do Not Cite or Quote
-------
25 -
-Ł• 20 -
c
o
15 -
Ul
g
LJJ
10 -
5 -
M
O
M
O
N>
O
O
N)
Total NOV Emissions
NJ
O
O
M
O
O
MW
OO
OO
Year
NJM
OO
OO
M
O
O
M
O
M
O
NJ
O
M
O
Source: National Center for Environmental Assessment Analysis of 201 1 National Emissions Inventory Data (U.S. EPA. 2013a)
Figure 2-2 U.S. national average NOx (sum of nitrogen dioxide and nitric
oxide) emissions from 1990 to 2013.
3
4
5
6
7
The major sources of NOX in the U.S. identified from the 2008 and 2011 NEI (U.S. EPA.
2013a. 2011) are described in Figure 2-3. The values shown are U.S. nationwide averages
and may not reflect the mix of sources relevant to individual exposure in populated areas.
For most sources, data are generally available for all 50 states and the District of
Columbia (in some cases, such as agricultural burning, data available in the NEI exclude
Alaska and Hawaii). Biogenic emissions were estimated using 2011 meteorology and
land-use information using the Biogenic Emission Inventory System, version 3.14 (BEIS
3.14) model. Although the BEIS domain includes Canada and Mexico, the NEI uses
BEIS estimates from counties that make up the contiguous 48 states.
January 2015
2-11
DRAFT: Do Not Cite or Quote
-------
Highway Vehicles
Off-Highway
Fuel Combustion-Utilities
Fuel Combustion-Other
Other Anthropogenic
Biogenics and Wildfires
-
0
D2008
• 2011
4 6
Millions of Tons
8
10
Source: National Center for Environmental Assessment Analysis of 2011 National Emissions Inventory Data (U.S. EPA. 2013a.
2011).
Figure 2-3 Major sources of NOx (sum of nitrogen dioxide and nitric oxide)
emissions averaged over the U.S. from the 2008 and 2011 National
Emissions Inventories.
i
2
o
J
4
5
6
7
The source categories used in Figure 2-3 represent groups of similar NEI source sectors.
Highway Vehicles include all on-road vehicles, including light duty as well as heavy duty
vehicles, both gasoline- and diesel-powered. Off-Highway vehicles and engines include
aircraft, commercial marine vessels, locomotives, and nonroad equipment. Fuel
Combustion-Utilities includes electric power generating units (EGUs), which derive their
power generation from all types of fuels, but is dominated by coal combustion, which
accounts for 85% of all NOx emissions from utilities in the 2011 NEI. Fuel
January 2015
2-12
DRAFT: Do Not Cite or Quote
-------
1 Combustion-Other includes commercial/institutional, industrial, and residential
2 combustion of biomass, coal, natural gas, oil, and other fuels. Other Anthropogenic
3 sources include field burning, prescribed fires, and various industrial processes (e.g.,
4 cement manufacturing and oil and gas production). On a national scale, field burning and
5 prescribed fires are the greatest contributors to the Other Anthropogenic sources
6 category. Biogenics and Wildfires include NEI emission estimates for biogenic (plant and
7 soil) emissions and wildfires. For NOx, biogenic emissions are dominated by soil
8 emissions, which are one to two orders of magnitude greater than vegetation emissions.
9 Highway Vehicles are the largest source in the 2011 NEI, contributing 37% of the total
10 NOx emissions. Off-Highway vehicles and engines account for 20% of emissions, Fuel
11 Combustion-Utilities (by EGUs) for 14%, Fuel Combustion-Other for 11%, Other
12 Anthropogenic sources for 10%, and Biogenics and Wildfires for 8% of 2011 NEI
13 national emissions of NOx. Nationwide estimates of total NOx emissions in the 2011 NEI
14 are 13% lower than 2008 NEI estimates, decreasing from 18.0 megatons to
15 15.6 megatons. This decrease reflects lower emission estimates in the 2011 NEI than in
16 the 2008 NEI for the four largest categories in Figure 2-3: 17% lower for Highway
17 Vehicles, 10% lower for Off-Highway vehicles and engines, 33% lower for Fuel
18 Combustion-Utilities, and 6% lower for Fuel Combustion-Other. However, estimated
19 emissions were 17% higher for Other Anthropogenic sources, with the greatest increases
20 observed for oil and gas production, agricultural field burning, prescribed fires, and
21 mining. Although Biogenics and Wildfire emissions have increased as a proportion of
22 total national emissions, anthropogenic sources (i.e., the other categories) still account for
23 more than 90% of emissions in the 2011 NEI.
24 A somewhat different source mixture than the national average occurs in the most
25 populated areas. Figure 2-4 compares contributions from different groups of sources to
26 the 21 Core-Based Statistical Areas (CBSAs) in the U.S. with populations greater than
27 2.5 million, where 39% of the U.S. population lives. Relative to the national average, the
28 urban areas have greater contributions to total NOx emissions from both Highway
29 Vehicle emissions and Off-Highway emissions, and smaller contributions from Fuel
30 Combustion-Utilities (EGUs), Other Anthropogenic emissions, and Biogenics and
31 Wildfires.
32 Table 2-1 provides details on source distributions from individual CBSAs.
January 2015 2-13 DRAFT: Do Not Cite or Quote
-------
Highway Vehicles
Off-Highway
Fuel Combustion-Utilities
Fuel Combustion-Other
Other Anthropogenic
Biogenics and Wildfires
Source: National Center for Environmental Assessment Analysis of 2011 National Emissions Inventory Data (U.S. EPA. 2013a).
Figure 2-4 Percentage contributions from major sources of the annual NOx
(sum of nitrogen dioxide and nitric oxide) emissions averaged
over the 21 largest U.S. Core-Based Statistical Areas with
populations greater than 2.5 million (blue—urban) compared to
the national average (red—national).
January 2015
2-14
DRAFT: Do Not Cite or Quote
-------
Table 2-1 Source distribution of the annual NOx (sum of nitrogen dioxide and
nitric oxide) emissions in the 21 largest U.S. Core-Based Statistical
Areas with populations greater than 2.5 million— 2011 National
Emissions Inventory.
New York, NY
Los Angeles, CA
Chicago, IL
Dallas, TX
Houston, TX
Philadelphia, PA
Washington, DC
Miami, FL
Atlanta, GA
Boston, MA
San Francisco, CA
Riverside, CA
Phoenix, AZ
Detroit, Ml
Seattle, WA
Minneapolis, MN
San Diego, CA
Tampa, FL
St. Louis, MO
Baltimore, MD
Denver, CO
Urban Average
Highway
%
44.3
59.6
40.1
53.9
46.0
51.2
55.2
50.5
57.7
46.4
52.8
53.0
64.0
52.5
66.5
50.2
64.3
52.7
56.9
48.5
55.6
53.4
Off-Highway
%
27.6
26.2
27.1
21.5
25.6
22.5
23.4
32.2
19.6
27.3
30.6
20.8
25.9
19.4
25.7
19.4
25.5
23.2
13.6
21.4
22.5
23.9
Utilities Other Fuel
% %
4.4
0.6
11.1
1.2
3.0
4.0
7.8
7.6
15.1
2.7
0.8
1.6
1.7
8.4
0.1
11.3
0.6
13.5
15.3
11.5
3.2
6.0
21.6
10.4
14.8
9.2
9.5
14.7
10.9
4.8
5.7
18.7
9.7
8.5
5.2
14.4
5.1
13.9
4.3
5.6
6.7
11.0
12.2
10.3
Other Anth. Bio/Wildfire
% %
1.8
2.5
5.4
9.5
11.4
6.7
1.5
2.7
1.1
4.6
4.9
12.4
0.6
4.1
2.3
3.3
0.8
2.1
5.2
6.6
2.9
4.4
0.4
0.7
1.4
4.7
4.5
0.9
1.3
2.2
0.9
0.4
1.2
3.7
2.7
1.2
0.4
2.0
4.6
2.9
2.4
1.0
3.6
2.0
Source: National Center for Environmental Assessment Analysis of 2011 National Emissions Inventory Data (U.S. EPA. 2013a).
January 2015 2-15 DRAFT: Do Not Cite or Quote
-------
2.3.2 Highway Vehicles
i
2
o
J
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
Nationally, Highway Vehicles account for about 37% of NOx emissions, according to the
2011 NEI. In the 21 largest CBSAs in the U.S. represented in Figure 2-4. more than half
of the urban NOx emissions are from Highway Vehicles, ranging from 40% in Chicago to
67% in Seattle, and together, Highway Vehicles and Off-Highway vehicles and engines
account for more than three quarters of total emissions. Other estimates of high
contributions from Highway Vehicles have also been reported. For example, on-road
vehicles were estimated to account for about 80% of anthropogenic NOx concentrations
in the Los Angeles area (Mcdonald et al.. 2012) and 72% in the Atlanta area (Pachon
et al.. 2012). Highway Vehicle NOx emissions nationwide are roughly equally split
between light duty gasoline engines (48%) and heavy duty diesel engines (46%),
according to the 2011 NEI. This is in spite of a national vehicle fleet distribution of more
than 230 million mostly gasoline-powered light duty vehicles compared to only
10 million mostly diesel-powered heavy duty vehicles1. Mcdonald et al. (2012) estimated
that diesel engines were the dominant on-road NOx sources in the San Joaquin Valley in
California, accounting for up to 70% of NOx emissions. In contrast in Fulton County, GA
it was estimated that 60% of on-road NOx emissions were from gasoline vehicles and
40% from diesel (Pachon etal.. 2012). Mcdonald et al. (2012) estimated that in
California, gasoline engine-related NOx emissions steadily decreased by 65% over the
period from 1990 to 2010. They also found that the ratio of NOx emission factors for
heavy duty diesel versus light duty gasoline engines grew from ~3 to ~8 between 1990
and 2010 due to improved effectiveness of catalytic converters on gasoline engines.
However, NOx emissions from on-road diesel engines in the U.S. have also decreased
substantially as the result of stricter emission standards, and emissions continue to
decline (Mcdonald et al.. 2012). Emission standards for heavy duty diesel trucks were
first established at 10.7 g/bhp-h in 1988 and the current standard of 0.20 g/bhp-h was
gradually phased in for model years 2007 through 2010 (66 FR 5002), so that emission
standards from heavy duty diesel trucks were reduced by more than a factor of 50
between 1988 and 2010. The current standard is achieved using a urea-based SCR
catalyst in engine exhaust placed downstream of a diesel oxidation catalyst (DOC) and a
catalyzed diesel particulate filter (DPF) used for PM emissions control. In extensive
testing of diesel engines substantial reductions in NOx were observed, averaging 61%
relative to the 2010 standard requirements and 97% relative to the 2004 standard
requirements (Southwest Research Institute. 2013). However, while total diesel NOx
https://www.fhwa.dot.gov/policvinformation/statistics/2010/vml .cfm.
January 2015
2-16
DRAFT: Do Not Cite or Quote
-------
1 emissions have substantially decreased because of urea-based SCR control, the NCh/NOx
2 ratio has increased. But these reductions for diesel emissions together with the recent
3 final Tier 3 rule for gasoline engine emissions and lower S 41 gasoline (79 FR 23414) are
4 likely to result in a substantial decline in NOx emissions as newer vehicles penetrate into
5 the on-road fleet over the next several years.
2.3.3 Off-Highway
6 Off-Highway engines constitute the next largest group of NOx emission sources after
7 Highway Vehicles, both on a nationwide basis and in large urban CBSAs as shown in
8 Figure 2-4 and Table 2-1. Emissions from the nonroad source sector can also
9 significantly contribute to local and national air quality. The 2011 NEI estimated that
10 approximately 20% of nationwide NOx was from Off-Highway engines. Zhu et al. (2011)
11 estimated that nonroad diesel engines contribute 12% of total nationwide NOx emissions
12 from mobile sources. Off-Highway sources include aviation, marine, and railroad
13 engines, as well as nonroad agricultural and industrial equipment, all of which emit NOx
14 through combustion processes.
15 Examples of nonroad equipment include farm tractors, excavators, bulldozers, and wheel
16 loaders. Nationally, agricultural and industrial equipment accounts for more than 60% of
17 Off-Highway NOx emissions, mostly from diesel-powered equipment (U.S. EPA.
18 2013a). EPA has set a series of standards to reduce NOx emissions from nonroad diesel,
19 referred to as Tier 1-4 standards. The most recent standard, Tier 4, was introduced in May
20 2004, and the phase-in is currently underway, covering a time period between 2008 and
21 2015. In most cases, advanced diesel engine design, exhaust gas recirculation (EGR),
22 and/or SCR have been used to comply with these standards, with DOC/DPFs used in
23 several engine categories.
24 Although Fuel Combustion-Utilities is generally a smaller contributor to total NOx in
25 urban areas than it is nationally, emergency generators are an emerging concern. In urban
26 areas emissions of NOx have been observed to increase substantially on days of near
27 peak electricity demand because of small natural gas and petroleum powered steam
28 turbines used to generate additional electrical power to meet demand. These generators
29 are classified in the NEI as nonroad equipment that fall into the category of Off-Highway
30 engines. They are typically operated in densely populated areas. They are usually older
31 units with higher emissions and lower stack heights than larger generators, and they are
32 often located close to residential neighborhoods. Because of these factors, emergency
33 generators can have substantial impacts on local air quality. For example, Gilbraith and
January 2015 2-17 DRAFT: Do Not Cite or Quote
-------
1 Powers (2013) estimated that reducing emissions from emergency generators could
2 decrease NOx emissions in New York City alone by 70 tons per year.
3 Aircraft, commercial marine transport, and locomotive emissions account for the
4 remaining 40% of Off-Highway emissions, nationally. Aircraft includes all aircraft types
5 used for public, private, and military purposes, classified into four types: commercial, air
6 taxis, general aviation, and military. Airport-related NOx emissions can significantly
7 impact local and regional air quality. In the U.K., within a 2-3-km radius of London
8 Heathrow Airport, Carslaw et al. (2006) reported that airport emissions can comprise up
9 to 15% of total ambient NOX. In Atlanta, GA, Unal et al. (2005) showed that roughly
10 2.6% of regional NOx concentrations can be attributed to emissions from activities at
11 Hartfield-Jackson International Airport. Compared to airport-related emissions of other
12 gaseous pollutants (e.g., ammonia [NH3], carbon monoxide [CO], sulfur dioxide [802],
13 VOC), airport NOx emissions had the largest contribution to decreased regional air
14 quality in Atlanta, GA.
15 Commercial marine vessels include boats and ships used either directly or indirectly in
16 the conduct of commerce or military activity. Globally, marine transport is a significant
17 source of NOx emissions, accounting for more than 14% of all global nitrogen emissions
18 from fossil fuel combustion (mostly NOx) (Corbett et al.. 1999). On a regional scale, the
19 contribution of shipping emissions to total NOx emissions is variable and can be a
20 substantial fraction near port cities (Kim etal., 2011; Williams et al., 2009; Vutukuru and
21 Dabdub. 2008). In Los Angeles, CA, Vutukuru and Dabdub (2008) estimated that
22 commercial shipping contributed 4.2% to total NOx emissions in 2002. Using the
23 NEI-05, Kimetal. (2011) estimated that roughly 50% of NOx concentration near the
24 Houston Ship Channel is associated with commercial shipping emissions. However, this
25 estimation is much higher than observed in satellite and aircraft measurements.
26 Locomotives powered by diesel engines are a source of NOx emissions. Using a
27 fuel-based approach to quantify emissions, Dallmann and Harley (2010) estimated that
28 diesel locomotives emitted on average 50% of total NOx from all nonroad mobile sources
29 and roughly 10% of total NOx from all mobile sources in the U.S. from 1996-2006
30 (Dallmann and Harley. 2010). Locomotives can comprise a much larger fraction of NOx
31 emissions for areas in or near large rail yard facilities (>90% of emissions), including
32 NO2 nonattainment areas (U.S. EPA. 2010). In a year-long study at the Rougemere Rail
33 Yard facility near Dearborn, MI, 98% of NOx emissions was attributed to locomotive
34 operation, with only minimal impacts from other sources such as on-road mobile sources
35 and stationary sources (U.S. EPA, 2009). Cahill etal. (2011) measured gaseous and PM
36 pollutants during a 2-week period near the Roseville Rail Yard in Placer County, CA.
37 They observed several transient NOx emission events, where NO levels of 100s of ppb
January 2015 2-18 DRAFT: Do Not Cite or Quote
-------
1 were observed downwind of the Roseville Rail Yard, which was roughly 7 times larger
2 than the observed urban background NO.
2.3.4 Fuel Combustion—Utilities and Other
3 Fuel combustion for electric power generation and for industrial, residential, commercial,
4 and institutional purposes (excluding motor vehicles and nonroad equipment) accounts
5 for about 25% of NOx emissions nationwide. As indicated in Figure 2-3. Fuel
6 Combustion-Utilities accounts for the majority of fuel combustion NOx emissions in the
7 U. S., and about 14% of total NOx emissions. About 85% of the fuel used for power
8 generation is coal. However, in urban areas, fuel combustion for purposes other than
9 electric power generation appears to be a greater source of emissions (as shown in
10 Figure 2-4).
11 In contrast to Fuel Combustion-Utilties, coal accounts for only about 1% of Fuel
12 Combustion-Other emissions. However, Fuel Combustion-Other is still dominated by
13 fossil fuels, with natural gas contributing about 68% and oil combustion contributing
14 about 14% of other fuel combustion emissions. Thus, even though biofuels are an
15 important NOx source globally (Jaegle et al.. 2005). only about 10% of Fuel
16 Combustion-Other emissions in the U.S. are due to biomass burning. For Fuel
17 Combustion-Utilities and Fuel Combustion-Other combined, fossil fuels account for
18 more than 90% of U.S. stationary source fuel combustion, and biomass only 4%.
19 Combustion of biofuels accounts for only about 1% of total NOx emissions nationwide.
20 Fuel Combustion-Other accounts for an additional 12% of urban NOx emissions, but
21 ranges as high as 22% in New York and 19% in Boston, as shown in Table 2-1.
22 Figure 2-5 shows that the contribution Fuel Combustion-Other to overall urban NOx
23 emissions varies with average January temperatures. This trend suggests that winter
24 heating is the driving factor for Fuel Combustion-Other emissions, and that in winter the
25 Fuel Combustion-Other contribution is likely to be considerably greater than the
26 contribution presented on an annual basis in Table 2-1. possibly rivaling Highway
27 Vehicle emissions in winter.
January 2015 2-19 DRAFT: Do Not Cite or Quote
-------
6000
5000
.3! 4000
Q.
o
01
Q.
f 3000
01
Q.
x
O
z
V)
,2 2000
1000
•- Minneapolis
Det
roit
Chicago
St Louis
Philadelphia
B
oston
Denver -*
Wash
/- Baltimore
•
ngton
Seattle
A1
Dallas
/
*
.r •
/
lanta ph
fHoustt
f
•'
oenix
•-
an
San Francisco
Los Angel
s- Tamp
•-
— San Dieg
2S
a
— Miami
D
10 20 30 40 50 60
Average January Temperature (°F)
70
80
Source: National Center for Environmental Assessment Analysis of 2011 National Emissions Inventory Data (U.S. EPA. 2013a).
Figure 2-5 Fuel Combustion-Other emissions vs. average ambient January
temperature for the 21 largest U.S. Core-Based Statistical Areas
>2.5 million population.
2.3.5 Other Anthropogenic Sources
i
2
o
J
4
5
Other Anthropogenic sources include prescribed and agricultural fires as well as
industrial operations such as oil and gas production and mining. As emissions estimates
from other major source categories have decreased between 2008 and 2011, emissions
from these sources have increased by 17%, from about 1.4 megatons in 2008 to more
than 1.6 megatons in 2011. On a national scale, agricultural burning and prescribed fires
January 2015
2-20
DRAFT: Do Not Cite or Quote
-------
1 are responsible for a large fraction of the Other Anthropogenic sources category and its
2 the increase in national emissions for Other Anthropogenic sources between 2008 and
3 2011. However, in urban areas, fires are less of a contributor, and Other Anthropogenic
4 sources are mainly industrial. Other Anthropogenic sources vary considerably among the
5 21 largest CBSAs with populations greater than 2.5 million. In three CBSAs, NOx
6 emissions from Other Anthropogenic sources exceed 10,000 tons per year. Emissions in
7 these three CBSAs are separated by industrial sector in Table 2-2.
8 In Chicago, emissions from several different sources make substantial contributions to
9 Other Anthropogenic emissions. In contrast, Other Anthropogenic emissions in Dallas are
10 dominated by oil and gas production, which is not an important source in Chicago. The
11 oil and gas production sector is an increasing source of NOx, with a 2011 emission
12 estimate of more than 600,000 tons, compared to slightly more than 400,000 tons in
13 2008. Pacsi etal. (2013) estimated that routine operating activities from the Barnett Shale
14 production facility near Dallas, TX can emit roughly 30 to 46 tons NOx/day, depending
15 on the demand for natural gas electricity generation. Nonroutine gas flares can also result
16 in episodic peaks of large NOx emissions, affecting local air quality (Olaguer. 2012).
17 Houston, TX, presents yet another variation, with anthropogenic emissions mainly from
18 petroleum refining and chemical manufacturing. These data demonstrate that sources
19 with relatively small nationwide or annual emissions may contribute substantially to
20 emissions on a local scale. For example, cement manufacturing, which is listed in
21 Table 2-2 as an important source in the local Dallas, TX, airshed, accounts for less than
22 1% of annual national emissions, but has been characterized by variable emissions with
23 high peaks (Walters et al.. 1999).
January 2015 2-21 DRAFT: Do Not Cite or Quote
-------
Table 2-2 Relative contributions to Other Anthropogenic NOx (sum of nitrogen
dioxide and nitric oxide) sources in selected cities3.
Bulk gasoline terminals
Fires — agricultural field burning
Fires — prescribed fires
Gas stations
Industrial processes — cement manufacturing
Industrial processes — chemical manufacturing
Industrial processes — ferrous metals
Industrial processes — not elsewhere classified
Industrial processes — nonferrous metals
Industrial processes — oil & gas production
Industrial processes — petroleum refineries
Industrial processes — pulp & paper
Industrial processes — storage and transfer
Miscellaneous nonindustrial not elsewhere classified
Solvent — degreasing
Solvent — graphic arts
Industrial surface coating & solvent use
Waste disposal
Total
Chicago
nrb
nr
1
nr
nr
8
9
35
21
0
13
nr
nr
nr
nr
nr
1
11
100
Dallas
nr
1
4
nr
19
nr
2
6
nr
66
nr
nr
nr
nr
nr
nr
nr
1
100
aNOx emissions as percent of "Other Anthropogenic sources" emissions in the Core-Based Statistical Area.
bnr indicates that no emissions were reported for this sector, i.e., there were no sources with emissions above the
threshold.
Source: National Center for Environmental Assessment Analysis of 201 1 National Emissions Inventory Data (U.S.
Houston
nr
3
1
nr
nr
43
nr
3
nr
9
37
nr
nr
nr
nr
nr
1
4
100
reporting
EPA, 201 3a).
January 2015
2-22
DRAFT: Do Not Cite or Quote
-------
2.3.6 Biogenics and Wildfires
1 The NEFs Biogenics sector includes emissions from plants and soil. In the case of NOx,
2 biogenic emissions are dominated by emissions from soil. Biogenic emissions account for
3 about 6% of total NOx emissions in the 2011 NEI. However, spatial and temporal
4 variability in NOx emissions from soil leads to considerable variability in emission
5 estimates. For example, satellite-based estimates that 15-40% of the total NO2 column in
6 various locations over the Great Plains region can be attributed to soil emissions in spring
7 and summer months Hudman et al. (2010). This is consistent with geographic differences
8 in soil contributions described in the 2011 NEI, in which soil contributions accounted for
9 13-34% of NOx emissions in Iowa, Kansas, Nebraska, North Dakota, and South Dakota.
10 About 60% of the total NOx emitted from soils is estimated to occur in the central corn
11 belt of the U.S. Because of low population density and the wide area over which
12 emissions are distributed, soil emissions are a less important concern for exposure than
13 more concentrated sources in more highly populated areas.
14 Biogenic emissions for the 2011 NEI were computed based on the BEIS model. The
15 BEIS modeling domain includes the contiguous 48 states in the U.S., parts of Mexico,
16 and Canada. The NEI uses the biogenic emissions from counties from the contiguous
17 48 states and DC. Both nitrifying and denitrifying organisms in the soil can produce
18 NOx, mainly in the form of NO. Emission rates depend mainly on the amount of applied
19 fertilizer, soil temperature, and soil moisture. As a result, a high degree of uncertainty is
20 associated with soil emissions, and estimates obtained from satellite observations can be
21 greater than source-based estimates (Jaegle et al., 2005).
22 Emissions from wildfires can produce enough NOx to cause local and regional
23 degradation of air quality in some regions (Pfister et al.. 2008). Roughly 15% of global
24 NOx emissions are from biomass burning (Penman et al.. 2007). Burling et al. (2010)
25 reported that NOx emissions from Southwest U.S. vegetation ranged from 2.3 to 5.1 g/kg,
26 with the majority of the NOx present as NO. Emissions vary considerably among
27 different species of biota, making it difficult to estimate emissions for key ecosystems,
28 such as extratropical forests (Mcmeeking et al.. 2009). Emissions from forest wildfires
29 can be more than double per amount of energy released than for shrubs (Mebust et al..
30 2011).
2.3.7 Emissions Summary
31 Major categories of NOx emissions in the U.S. are Highway Vehicles, Off-Highway
32 vehicles and engines, Fuel Combustion-Utilities, Fuel Combustion-Other, Other
January 2015 2-23 DRAFT: Do Not Cite or Quote
-------
1 Anthropogenic emissions, and Biogenics and Wildfire emissions. Of these,
2 Fuel-Combustion-Utilities and Biogenics and Wildfire emissions are less important in
3 populated urban areas with the highest NO2 concentrations and thus, potentially have less
4 impact on human exposure to NO2. Instead, in urban areas, emissions are generally
5 dominated by Highway Vehicles and Off-Highway vehicles and engines, which make up
6 more than three-quarters of emissions in the 21 largest CBSAs with populations greater
7 than 2.5 million. Other sources can make important contributions. For example, in cities
8 with average January temperatures below freezing, NOx emissions from Fuel
9 Combustion-Other can also be important, and episodic emissions from Other
10 Anthropogenic sources can be important locally. However, Highway Vehicles is
11 generally the greatest source of NOx emissions in urban areas.
2.4 Measurement Methods
2.4.1 Federal Reference and Equivalent Methods
12 This discussion focuses on current methods and on promising new technologies, but no
13 attempt is made here to cover in detail the development of these methods, or of methods
14 such as wet chemical techniques, which are no longer in use. More detailed discussions
15 of the histories of these methods can be found elsewhere (U.S. EPA. 1996. 1993).
16 NO is routinely measured using the chemiluminescence induced by its reaction with Os at
17 low pressure. The Federal Reference Method (FRM) for NO2 makes use of this technique
18 of NO detection with a prerequisite step that is meant to reduce NO2 to NO on the surface
19 of a molybdenum oxide (MoOx) substrate heated to between 300 and 400°C. On June 1,
20 2012, an automated Federal Equivalent Method (FEM) for measuring NO2 using a
21 photolytic converter to reduce NO2 to NO met the equivalency specifications outlined in
22 40 CFR Part 53 and was approved by the U.S. EPA (77FR 32632). Although photolytic
23 converters have lower conversion efficiencies than FRM-based analyzers, they have been
24 found to be stable over a period of at least two months (Pollack et al.. 2011). Also, two
25 monitors using cavity attenuated phase shift (CAPS) spectroscopy have been approved
26 more recently as FEMs (78 FR 67360, November 12, 2013; 79 FR 34734, June 18, 2014).
27 These techniques are described below.
28 Because the FRM monitor cannot detect NO2 specifically, the concentration of NO2 is
29 determined as the difference between the NO in the air stream passed over the heated
30 MoOx substrate and the NO in the air stream that has not passed over the substrate.
January 2015 2-24 DRAFT: Do Not Cite or Quote
-------
1 However, the reduction of NCh to NO on the MoOx catalyst substrate also reduces other
2 oxidized nitrogen compounds (i.e., NOz compounds shown in the outer box of
3 Figure 2-1) to NO. This interference by NOz compounds has long been recognized
4 following Winer et al. (1974) who found over 90% conversion of PAN, ethyl nitrate,
5 ethyl nitrite, and n-propyl nitrate; and 6-7% conversion of nitroethane to NO with a
6 MoOx converter. HNOs produced a response but its form could not be determined. As a
7 result of their experiments, Winer etal. (1974) concluded that "the NOx mode of
8 commercial chemiluminescent analyzers must be viewed to a good approximation as
9 measuring total gas phase 'oxides of nitrogen,' not simply the sum of NO and NO2."
10 Numerous later studies have confirmed these results (Dunleaetal.. 2007; Steinbacher
11 et al.. 2007; U.S. EPA. 2006; McClenny et al.. 2002; Parrish and Fehsenfeld. 2000;
12 Nunnermackeretal.. 1998: Croslev. 1996: U.S. EPA. 1993: Rodgers and Davis. 1989:
13 Fehsenfeld et al., 1987). The sensitivity of the FRM to potential interference by
14 individual NOz compounds was found to be variable, depending on characteristics of
15 individual monitors, such as the design of the instrument inlet, the temperature and
16 composition of the reducing substrate, and on the interactions of atmospheric species
17 with the reducing substrate.
18 Only recently have attempts been made to systematically quantify the magnitude and
19 variability of the interference by NOz species in ambient measurements of NO2. Dunlea
20 et al. (2007) found an average of about 22% of ambient NO2 (~9 to 50 ppb), measured in
21 Mexico City over a 5-week period during the spring of 2004, was due to interference
22 from NOz compounds. However, similar comparisons have not been carried out under
23 conditions typical for state and local air monitoring stations (SLAMS) monitoring sites in
24 the U.S. Dunlea et al. (2007) compared NO2 measured using the conventional
25 chemilumine scent instrument with other (optical) techniques. The main sources of
26 interference were HNOs and various organic nitrates. Efficiency of conversion was
27 estimated to be -38% for HNO3 and -95% for PAN and other organic nitrates. Peak
28 interference of up to 50% was found during afternoon hours and was associated with Os
29 and NOz compounds, such as HNOs and the alkyl and multifunctional alkyl nitrates.
30 Lamsal et al. (2008) used data for the efficiency of reduction of NOz species on the
31 MoOx catalytic converters to estimate seasonal correction factors for NO2 measurements
32 across the U.S. These factors range from <10% in winter to >80% with the highest values
33 found during summer in relatively unpopulated areas. In general, interference by NOz
34 species in the measurement of NO2 is expected to be larger downwind of urban source
35 areas and in relatively remote areas because of the conversion of NO2 to NOz during
36 transport downwind of source areas.
January 2015 2-25 DRAFT: Do Not Cite or Quote
-------
1 In a study in rural Switzerland, Steinbacher et al. (2007) compared continuous
2 measurements of NO2 from a chemiluminescence analyzer with a MoOx catalytic
3 converter (CL/MC) with measurements from a photolytic converter (CL/PC) that reduces
4 NO2 to NO. They found the conventional technique using catalytic reduction (as in the
5 FRM) overestimated the measured NO2 compared to the photolytic technique, on average
6 by 10% during winter and 50% during summer.
7 Villenaetal. (2012) and Kleffmann et al. (2013) suggested that negative interference in
8 the chemiluminescent method using the photolytic converter could occur by the
9 production of HO2 and RO2 radicals by the photolysis of VOCs (e.g., glyoxal) in the
10 photolytic converter. Subsequent to photolysis and prior to detection, these radicals react
11 with NO that is either produced by the photolytic converter or already in the sampling
12 stream. Because the chemiluminescent techniques rely on detection of NO, a negative
13 artifact results. The most direct evidence for this artifact was found at high concentrations
14 in a smog chamber containing 1 ppm glyoxal, a concentration more than a thousand times
15 higher than typically found in ambient air. Similar indications were also found by
16 Kleffmann et al. (2013) in a street canyon (at the University of Wuppertal, Germany) and
17 in an urban background environment (University of Santiago, Chile). However,
18 Kleffmann et al. (2013) also found that the magnitude of the negative artifact is smaller
19 when a light source with a smaller spectral range is used and that this artifact is expected
20 to be most apparent under high VOC conditions, such as in street canyons.
21 Within the urban core of metropolitan areas, where many of the ambient monitors are
22 sited close to strong NOx sources such as motor vehicles on busy streets and highways
23 (i.e., where NO2 concentrations are highest), the positive artifacts due to the NO2
24 oxidation products are much smaller on a relative basis. Conversely, the positive artifacts
25 are larger on a relative basis away from NOx sources. Data for PAN and HNOs were
26 collected in Houston, TX in April and May of 2009 during the Study of Houston
27 Atmospheric Radical Precursors (SHARP) campaign (Olaguer et al.. 2014). Median
28 concentrations of PAN and HNOs during the afternoon were 181 (interquartile range
29 [IQR] 94) ppt and 164 (IQR 158) ppt for NO2 <1 ppb measured by CL/PC during
30 SHARP; and 157 (IQR 54) ppt and 146 (IQR 402) ppt for NO2 >10 ppb. These results
31 suggest that potential interference in CL/MC caused by HNOs and PAN is estimated to
32 be <1 ppb using the conversion efficiencies obtained by Dunlea et al. (2007) and
33 concentrations of HNOs and PAN obtained during SHARP. However, the extent of
34 interference could be expected to be most problematic for NO2 <~1 ppb.
35 In summary, the current FRM for determining ambient NOx concentrations and then
36 reporting NO2 concentrations by subtraction of NO is subject to a consistently positive
37 interference by NOx oxidation products, including HNOs, PAN, and its analogues, and
January 2015 2-26 DRAFT: Do Not Cite or Quote
-------
1 total organic nitrates (RONCh). The magnitude of this positive bias is largely unknown as
2 measurements of these oxidation products in urban areas are sparse.
2.4.2 Other Methods for Measuring Nitrogen Dioxide
3 Optical methods such as those using differential optical absorption spectroscopy (DOAS)
4 or laser-induced fluorescence (LIF) are also available. However, these particular
5 methods, even those that have been commercialized (e.g., DOAS), can be more
6 expensive than either the FRM monitors or photolytic reduction technique and also
7 require specialized expertise to operate; moreover, the DOAS obtains a path-integrated
8 rather than a point measurement. Cavity attenuated phase shift (CAPS) monitors are an
9 alternative optical approach requiring much less user intervention and expense than either
10 DOAS or LIF (Kebabian et al.. 2008). At first glance, it might appear that this technique
11 is not highly specific to NO2, as it is subject to interference by species that absorb at
12 440 nm such as 1,2-dicarbonyl compounds. However, this source of interference is
13 expected to be small (~1%), and if necessary, the extent of this interference can be
14 limited by shifting the detection to longer wavelengths and adjusting the lower edge of
15 the detection band to 455 nm. In principle, detection limits could be <30 ppt for a
16 60 second time scale.
17 Lee etal. (2011) describe the development of a dual continuous wave mode quantum
18 cascade - tunable infrared laser differential absorption spectrometer or QC-TILDAS to
19 measure NO2 and HONO simultaneously. The one-second detection limit (signal-to-noise
20 ratio [S/N] = 3) is 30 ppt. A field comparison of measurements of NO2 by CAPS and by
21 chemiluminescence monitors with MoOx converters (CL/MC) is shown in Figure 2-6.
22 The CAPS—CL/MC (Thermo Electron 421) data were obtained over 4 days in a parking
23 lot located -200 m from a major arterial highway (Route 3 in Billerica, MA) in October
24 2007. Figure 2-7 shows the results of a comparison of NO2 measured by QC-TILDAS to
25 NO2 measured by chemiluminescence with a photolytic converter (CL/PC). The
26 QC-TILDAS-CL/PC data were collected in Houston, TX in April and May of 2009
27 during the SHARP campaign (Olaguer et al.. 2014). Both comparisons show very high R2
28 (>0.99) and close agreement over concentrations ranging from <1 ppb to >30 ppb, and
29 both comparisons are characterized by small nonzero intercepts. For the CAPS
30 instrument (see Figure 2-6). slightly higher values than those reported by the CL/MC
31 monitor are seen at concentrations <~2 ppb. Figure 2-7 shows that the QC-TILDAS
32 obtains slightly lower concentrations than reported by CL/PC for NO2 concentrations
33 <~1 ppb. Although CAPS presents a practical alternative to chemiluminescence for NO2
34 measurements, an important consideration in routine network deployment of CAPS or
35 any other method that only measures NO2 (e.g., does not measure NO) is the potential
January 2015 2-27 DRAFT: Do Not Cite or Quote
-------
loss of NOx data, which has been used as an indicator for traffic- or other
combustion-related pollution.
10
•u
1
CAPS vs CL/MC
= 1.007x + 0,0357
R1 = 0.9939
10
Chemiluminescence/Me NO: (ppb)
Source: National Center for Environmental Assessment, using data from Kebabian et al. (2008).
Figure 2-6 Comparison of nitrogen dioxide (NOa) measured by cavity
attenuated phase shift (CAPS) spectroscopy to NOa measured by
chemiluminescence/MoOx catalytic converter (MC) for 4 days in
October 2007 in Billerica, MA.
January 2015
2-28
DRAFT: Do Not Cite or Quote
-------
QC-TILDAS vs CL/PC
-g. 10 J
Q.
a *
1C
Chemiluminescence IMO2/PC (ppb)
Source: National Center for Environmental Assessment, using data from Lee et al. (2013)
Figure 2-7 Comparison of nitrogen dioxide (NO2) measured by quantum
cascade-tunable infrared differential absorption spectroscopy
(QC-TILDAS) to NO2 measured by Chemiluminescence with
photolytic converter during April and May 2009 in Houston, TX.
i
2
o
J
4
5
6
7
8
9
10
11
12
Villena et al. (2011) describe the development of a long path absorption photometer
(LOPAP) to measure NC>2. In this technique, NO2 is sampled in a stripping coil using a
modified Griess-Saltzman reagent with the production of an azodye whose visible
absorption is measured by long-path photometry. This reaction was the basis for a much
earlier manual method for measuring NC>2 (Saltzman. 1954). Interference, which can be
minimized by additional stripping coils, could be caused by HONO, Os, and PAN. In an
intercomparison with a CL/PC carried out over 4 days in March 2007 on the 5th floor
balcony of a building at the University of Wuppertal in Germany, very good agreement
(mean deviation of 2%) was obtained. Interestingly, in the entire range of measurements
(~0.5 ppb to ~40 ppb) the relation between LOPAP and CL/PC can be characterized by
LOPAP (ppb) = 0.984 x CL/PC - 0.42 (ppb); but if the range <6 ppb only is considered,
the relation becomes LOPAP (ppb) = 0.998 x CL/PC + 0.19 (ppb).
January 2015
2-29
DRAFT: Do Not Cite or Quote
-------
1 Diode laser-based cavity ring-down spectroscopy (CRDS) has also been used to detect
2 NO2. Fuchs et al. (2009) developed a portable instrument that relies on NCh absorption at
3 404 nm, with 22 ppt detection limit at 1 second (S/N = 2). As opposed to
4 chemiluminescence monitors that measure NC>2 indirectly based on direct measurement
5 of NO, NC>2 (formed by reaction of NO with excess Os) is directly measured in CRDS.
6 NO is then determined by subtracting NO2 measured in the first cavity from the sum of
7 NO2 and NO (i.e., NOx) measured in the second cavity. The Os is generated by
8 photolysis of O2 in the Schumann-Runge bands at 185 nm. This conversion should be
9 much more complete than relying on the reduction of NO2 and NOz species with variable
10 efficiency on a Molybdenum oxide converter. Note that the optical methods relying on
11 NO2 absorption at -400 nm described above (i.e., CAPS, CRDS) might be subject to
12 positive interference from absorption by trace components (e.g., glyoxal and methyl
13 glyoxal). However, absorption cross sections for these dicarbonyls are much lower than
14 for NO2 at this wavelength, and concentrations for these potentially interfering species
15 are generally lower than for NO2. Furthermore, it is possible that thermal decomposition
16 of NOz species, such as PAN, in inlets or their reduction on inlet surfaces or in optical
17 cavities can be a source of NO2 in these or other instruments requiring an inlet.
2.4.3 Satellite Measurements of Nitrogen Dioxide
18 Remote sensing by satellites is an approach that could be especially useful in areas where
19 surface monitors are sparse. Retrieving NO2 column abundances from satellite data
20 involves three steps: (1) determining the total NO2 integrated line-of-sight (slant)
21 abundance by spectral fitting of solar backscatter measurements, (2) removing the
22 stratospheric contribution by using data from remote regions where the tropospheric
23 column abundance1 is small, and (3) applying an air mass factor to convert tropospheric
24 slant columns into vertical columns. The retrieval uncertainty is largely determined by
25 Steps 1 and 2 over remote regions where there is little tropospheric NO2, and by Step 3,
26 over regions of elevated tropospheric NO2 (Boersma et al.. 2004; Martin et al.. 2002).
27 Satellite retrievals are largely limited to cloud fractions <20%. The algorithm used here to
28 derive the tropospheric column of NO2 is given in Bucselaetal. (2013). This algorithm
29 was used to generate the maps in Figure 2-8 for 2005 to 2007 and in Figure 2-9 for 2010
30 to 2012 showing seasonal average NO2 columns obtained by the Ozone Monitoring
31 Instrument (OMI) on the AURA satellite. Other algorithms, for example the Berkeley
32 High-Resolution product (Russell et al.. 2011). which is based on higher resolution input
1 Column refers to the integrated line-of-sight abundance in a unit cross section, such that its units are
molecules/cm2.
January 2015 2-30 DRAFT: Do Not Cite or Quote
-------
fields (topography, albedo, and NCh vertical profile shape) in the retrievals, can reduce
the uncertainty in the measurements.
OMI Troposphertc NO2 (10 molec. crn )
1O
8
O.S
O.1
Note: Images shown were constructed by Dr. Lok Lamsal of Universities Space Research Association from data obtained by the
OMI instrument on the AURA satellite (http://aura.asfc.nasa.gov/scinst/omi.html) using the algorithm described in Bucsela et al.
(2013). Top panel (winter; DJF: December, January, February). Lower panel (summer; JJA: June, July, August).
Figure 2-8 Seasonal average tropospheric column abundances for nitrogen
dioxide (NOa: 1015 molecules/cm2) derived by ozone monitoring
instrument (OMI) for winter (upper panel) and summer (lower
panel) for 2005 to 2007.
January 2015
2-31
DRAFT: Do Not Cite or Quote
-------
OMI Tropospherie NO? (10 molec. crn )
Note: Images shown were constructed by Dr. Lok Lamsal of Universities Space Research Association from data obtained by the
OMI instrument on the AURA satellite (http://aura.gsfc.nasa.gov/scinst/omi.html) using the algorithm described in Bucselaetal. (2013).
Top panel (winter; DJF: December, January, February). Lower panel (summer; JJA: June, July, August).
Figure 2-9 Seasonal average tropospheric column abundances for nitrogen
dioxide (NO2: 1015 molecules/cm2) derived by ozone monitoring
instrument (OMI) for winter (upper panel) and summer (lower
panel) for 2010 to 2012.
i
2
3
4
5
6
7
Areas of high column NC>2 abundance are found over major source areas during both
2-year periods shown in Figures 2-8 and 2-9. High column abundances are found over
many major urban areas, such as Los Angeles, CA; Houston, TX; Chicago, IL; and New
York City, NY; and over major power plant complexes such as the Four Corners
(Colorado, New Mexico, Arizona, and Utah) and the Ohio River Valley. A diffuse area
with column abundances above background is found over the Bakken Shale fields in
northwestern North Dakota in winter. However, in general, the area of very high column
abundance of NCh (shown in red) is smaller in the 2010 to 2012 composite than from
January 2015
2-32
DRAFT: Do Not Cite or Quote
-------
1 2005 to 2007. The photochemical lifetime of NCh is longer in winter than in summer
2 resulting in lower column abundances of NC>2 in summer than in winter during the 2-year
3 periods shown in Figures 2-8 and 2-9.
4 Because satellite instruments do not return surface concentrations directly, information
5 on NC>2 surface concentrations must be inferred from the column measurements. Lamsal
6 et al. (2008) and Lamsal et al. (2010) combined satellite data for column NO2 from OMI
7 with results from the GEOS-Chem global scale chemistry-transport model to derive
8 surface concentrations of NC>2 (see Figure 2-13 for an example of seasonally averaged
9 surface NC>2 concentrations derived by this method). This method accounts for the
10 feedback from the abundance of NO2 on the lifetime of NC>2. Note however that data are
11 collected only during the daily satellite overpass in early afternoon and this method has
12 only been applied for the time of satellite overpass. Some other means must be used to
13 extend the time period of applicability, for example by scaling the afternoon value by the
14 diel variation in a model, provided the model bias in simulating NC>2 has been
15 characterized over the times of interest in a 24-hour cycle (Stavrakou et al.. 2008; Kim
16 etal.. 2006).
2.4.4 Measurements of Total Oxides of Nitrogen in the Atmosphere
17 Commercially available NOx monitors have been converted to NOy monitors by moving
18 the MoOx converter to interface directly with the sample inlet. Because of losses on inlet
19 surfaces and differences in the efficiency of reduction of NOz compounds on the heated
20 MoOx substrate, NOx concentrations cannot be considered as a universal surrogate for
21 NOy. However, most of the NOy is present as NOx close to sources of fresh combustion
22 emissions, such as highways during rush hour. To the extent that all the major oxidized
23 nitrogen species can be reduced quantitatively to NO, measurements of NOy
24 concentrations should be more reliable than those for NOx concentrations, particularly at
25 typical ambient levels of NO2. Exceptions might apply in locations near NOx sources,
26 where NOx measurements are likely to be less biased and confidence in measurement
27 accuracy increases.
28 Alternatively, multiple methods for observing components of NOy have been developed
29 and evaluated in some detail. As a result of these methods, as applied in the field and the
30 laboratory, knowledge of the chemistry of odd-N species has evolved rapidly. Recent
31 evaluations of methods can be found in Arnold et al. (2007) for HNOs, (Wooldridge
32 etal.. 2010) for speciated PANs, and Pinto etal. (2014) for HONO. However, it is worth
33 reiterating that the direct measurements of NO are still the most reliable method. Reliable
34 measurements of NOy and NO2 concentrations, especially at the low concentrations
January 2015 2-33 DRAFT: Do Not Cite or Quote
-------
1 observed in many areas remote from sources, are also crucial for evaluating the
2 performance of three-dimensional, chemical transport models of oxidant and acid
3 production in the atmosphere.
2.4.5 Ambient Sampling Network Design
4 Figure 2-10 shows routinely operating monitoring sites for approximately 500 ambient
5 air oxidized nitrogen across the U.S. Four networks are highlighted: (1) regulatory-based
6 SLAMS designed to determine National Ambient Air Quality Standards (NAAQS)
7 compliance, (2) Clean Air Status and Trends Network (CASTNET) which provides
8 weekly averaged values of total nitrate (HNOs and pNOs) in rural locations, (3) National
9 Core (NCORE) network of approximately 70 stations designed to capture
10 area-representative multiple-pollutant concentrations that provides routinely measured
11 NOy, and (4) the Southeast Aerosol Research Characterization (SEARCH), a privately
12 funded network of 6-10 sites that provides direct measurements of true NC>2 as well as
13 NOy and other nitrogen species (oxidized and reduced forms).
January 2015 2-34 DRAFT: Do Not Cite or Quote
-------
• SLAMS
A CASTNET
D NCORE_
• SEARCH
Note: SLAMS = State and Local Air Monitoring Stations, CASTNET = Clean Air Status and Trends Network, NCORE = National
Core Network, SEARCH = Southeast Aerosol Research Characterization.
Source: U.S. Environmental Protection Agency.
Figure 2-10 Map of monitoring sites for oxides of nitrogen in the U.S. from
four networks.
i
2
o
6
4
5
6
7
With the exception of 4-6 sites in the SEARCH network, direct or true NC>2 is not
measured routinely (Hansen et al.. 2003). The regulatory networks rely on
chemiluminescence difference techniques that provide NO concentration directly and
report a calculated NO2 concentration as the difference between NOx concentration and
NO concentration as discussed above. Criteria for siting ambient monitors are given in
the State and Local Air Monitoring Stations/National Air Monitoring
Stations/Photochemical Monitoring Stations (SLAMS/NAMS/PAMS) Network Review
Guidance (U.S. EPA. 1998). NO2 monitors are meant to be representative of several
scales: microscale (in close proximity, up to 100 m from the source), middle (several city
January 2015
2-35
DRAFT: Do Not Cite or Quote
-------
1 blocks, 100 to 500 m), neighborhood (0.5 to 4 km), and urban (4 to 50 km)
2 (40 CFR Part 58, Appendix D). Microscale to neighborhood-scale monitors are used to
3 determine the highest concentrations and source impacts, while neighborhood- and
4 urban-scale monitors are used for monitoring population exposures.
5 EPA promulgated new minimum monitoring requirements in February of 2010,
6 mandating that state and local air monitoring agencies install near-road NO2 monitoring
7 stations within the near-road environment in larger urban areas. Under these new
8 requirements, state and local air agencies will operate one near-road NO2 monitor in any
9 CBSA with a population of 500,000 or more, and two near-road NO2 monitors in CBSAs
10 with 2,500,000 or more persons or roadway segments carrying traffic volumes of 250,000
11 or more vehicles. These monitoring data are intended to represent the highest population
12 exposures that may be occurring in the near-road environment throughout an urban area
13 over the averaging times of interest. The near-road NO2 network is intended to focus
14 monitoring resources on near-road locations where peak ambient NO2 concentrations are
15 expected to occur as a result of on-road mobile source emissions and to provide a clear
16 means to determine whether the NAAQS is being met within the near-road environment
17 throughout a particular urban area. The network is now being phased in, and the first
18 phase became operational in January of 2014.
2.5 Ambient Concentrations of Oxides of Nitrogen
19 This section provides a brief overview of ambient concentrations of NC>2 and associated
20 oxidized N compounds in the U.S.; it also provides estimates of background
21 concentrations used to inform risk and policy assessments for the review of the NAAQS.
22 In the 2008 ISA for Oxides of Nitrogen, NC>2 concentrations were summarized with an
23 explanation that the annual average NCh concentrations of ~15 ppb reported by the
24 regulatory monitoring networks were well below the level of the NAAQS (53 ppb), but
25 that the 1-hour daily maximum concentrations can be much greater in some locations,
26 especially in areas with heavy traffic (U.S. EPA. 2008a).
2.5.1 National Scale Spatial Variability
27 In the 2008 ISA for Oxides of Nitrogen, data were analyzed for NO2 measured at
28 monitoring sites located within urbanized areas in the U.S. (U.S. EPA. 2008a). NO2
29 concentrations were ~15 ppb for averaging periods ranging from a day to a year, and the
30 1-hour daily maximum NO2 concentration was -30 ppb, about twice as high as the
January 2015 2-36 DRAFT: Do Not Cite or Quote
-------
1 24-h avg. Data on NOz concentrations were very limited but indicated that HNOs and
2 HONO concentrations were considerably lower than NO2 concentrations. HNOs
3 concentrations ranged from <1 to >10 ppb and HONO concentrations were reported as
4 <1 ppb even under heavily polluted conditions. HNOs concentrations were highest
5 downwind of an urban center. HONO concentrations were present in areas with traffic, at
6 concentrations several percent of NO2 concentrations (U.S. EPA. 2008a'). Field study
7 results indicating much higher NOz concentrations than NOx concentrations in relatively
8 unpolluted rural air were also described (U.S. EPA. 2008a).
9 Figure 2-11 presents a national map of the 98th percentile of 1-hour daily maximum
10 concentrations based on 2011-2013 data, and Figure 2-12 presents annual average NO2
11 concentrations based on 2013 calendar year data. In both figures data are included only
12 for monitors with 75% of days reported for each calendar quarter over the 3-year period
13 and only for days with 75% of all hours reported. Because of the completeness
14 requirements, there are cases where sites have valid annual average data but not valid
15 1-hour daily maximum concentrations. The highest concentrations are in the Northeast
16 Corridor, California, and other urbanized regions, and the lowest concentrations are in
17 sparsely populated regions, most notably in the West. These observations are consistent
18 with those described in the 2008 ISA (U.S. EPA. 2008a). Tables 2-3 and 2-4 present
19 summary data on 1-hour daily maximum and annual average concentrations for the
20 period 2011-2013. Table 2-3 also includes summary data by individual years and by
21 quarters averaged over the 3 years, as well as summary data for selected urban areas that
22 are examined in recent U.S. epidemiologic studies on the health effects of NO2
23 (Chapters 5 and 6).
January 2015 2-37 DRAFT: Do Not Cite or Quote
-------
Legend
2011-2013 98th Percentile Daily 1-Hour
Maximum NC»2 Concentration (ppb)
I I 25-37
| 38-46
I 1*7-56
^H 57 - 73
Note: Concentrations indicated are the highest concentration in the county, and not intended to represent county-wide
concentrations.
Source: U.S. Environmental Protection Agency Analysis of data from state and local air monitoring stations in 2014.
Figure 2-11 98th percentiles of U.S. 1-hour daily maximum nitrogen dioxide
(NO2) concentrations (ppb) for 2011-2013.
January 2015
2-38
DRAFT: Do Not Cite or Quote
-------
Legend
Nth Concentration (ppb)
^B 0-5
| 6- 10
I I 11 -14
| | 15-19
^B 20 - 25
Note: Concentrations indicated are the highest concentration in the county, and not intended to represent county-wide
concentrations.
Source: U.S. Environmental Protection Agency 2014 analysis of data from state and local air monitoring stations.
Figure 2-12 U.S. annual average nitrogen dioxide (NO2) concentrations (ppb)
for 2013.
January 2015
2-39
DRAFT: Do Not Cite or Quote
-------
Table 2-3 Summary statistics for 1-hour daily maximum nitrogen dioxide (NO2)
concentrations based on state and local air monitoring stations
(ppb).
Percent! les
NO2
NO2
NO2
NO2
NO2
NO2
NO2
NO2
Atlanta3
Atlanta— allb
Boston3
Boston— allb
Denver3
Denver — allb
Houston3
Houston— allb
Los Angeles3
Los Angeles — allb
New York3
New York— allb
Seattle3
Seattle— allb
Year
2011-2013
2011
2012
2013
1st Quarter
2nd Quarter
3rd Quarter
4th Quarter
2011-2013
2011-2013
2011-2013
2011-2013
2011-2013
2011-2013
2011-2013
2011-2013
2011-2013
2011-2013
2011-2013
2011-2013
2011-2013
2011-2013
n
390,713
127,610
130,170
132,933
94,612
96,962
99,125
100,101
3,215
3,215
6,246
10,986
966
2,184
9,525
16,610
8,328
30,612
9,469
11,803
none
1,649
Mean
19
19
18
18
22
16
16
21
13
13
25
19
38
41
21
18
27
28
27
27
13
1
1
1
1
1
1
1
1
1
2
2
5
1
6
9
1
1
4
4
1
1
3
5
2
2
2
2
2
2
2
2
2
2
8
3
14
21
3
3
7
7
3
3
4
10
3
4
3
3
4
3
3
4
3
3
11
4
22
26
5
4
10
10
5
5
5
25
8
8
8
7
10
6
7
10
4
4
16
9
30
33
10
8
16
17
13
15
7
50
16
16
16
15
20
12
13
20
8
8
24
17
39
41
18
15
26
28
27
27
11
75
27
28
27
26
32
22
22
31
18
18
32
28
46
48
29
26
36
38
38
38
17
90
38
39
37
37
41
33
32
40
34
34
39
36
53
55
45
36
44
47
47
47
24
95
44
45
43
43
47
40
38
46
41
41
44
41
58
61
45
43
49
52
52
52
31
99
55
57
55
54
58
52
50
58
52
52
52
49
68
73
56
54
60
63
64
63
46
aCity name only rows contain hourly data that meet 75% completeness criteria.
bCity—all rows report data regardless of whether completeness criteria are met.
Source: Office of Air Quality Planning and Standards and National Center for Environmental Assessment Analysis of Air Quality
System Network Data 2011 -2013.
January 2015
2-40
DRAFT: Do Not Cite or Quote
-------
Table 2-4 Summary statistics for nitrogen dioxide (NO2), nitric oxide (NO), and NOx (sum of NO2 and NO)
annual average concentrations based on state and local air monitoring stations (ppb).
Percent! les
Pollutant
Year
n
Mean
Min
1
5
10
25
50
75
90
95
99
Max
NO2
NO2
NO2
NO2
NO2
2011-2013
2011
2012
2013
1,041
338
347
356
8.6
9.0
8.5
8.3
0.1
0.2
0.1
0.3
0.6
0.6
0.7
0.6
1.4
1.5
1.4
1.3
2.2
2.5
2.2
2.1
4.3
4.7
4.2
4.2
8.1
8.4
8.1
7.7
11.8
12.3
11.6
11.6
16.2
16.8
15.9
15.8
18.6
19.6
18.6
18.1
22.5
23.9
22.1
21.8
26.0
25.3
26.0
24.6
NO
NO
NO
NO
NO
2011-2013
2011
2012
2013
1,127
363
377
387
4.8
5.0
4.8
4.6
0.01
0.01
0.01
0.01
0.02
0.03
0.03
0.03
0.1
0.1
0.1
0.1
0.3
0.3
0.2
0.2
1.0
1.1
1.0
1.0
2.9
3.1
2.9
2.6
6.8
7.4
6.6
6.5
11.3
12.7
10.9
11.0
15.3
15.1
15.0
15.7
25.3
23.9
27.7
21.5
48.8
46.9
48.8
36.2
NOx
NOx
NOx
NOx
NOx
2011-2013
2011
2012
2013
1,011
320
342
349
13.4
13.7
13.2
13.3
0.1
0.1
0.1
0.3
0.6
0.6
0.7
0.7
1.5
1.5
1.3
1.7
2.6
2.6
2.6
2.6
5.4
5.8
5.2
5.4
11.3
11.8
11.2
10.9
18.6
19.3
18.5
18.3
28.1
28.9
26.8
28.0
31.8
31.7
31.3
32.7
45.4
44.8
48.9
44.1
68.4
68.4
61.0
61.7
Source: Office of Air Quality Planning and Standards and National Center for Environmental Assessment analysis of Air Quality System network data 2011 -2013.
January 2015
2-41
DRAFT: Do Not Cite or Quote
-------
1 The relatively short lifetime of NO2 with respect to conversion to NOz species results in
2 gradients and low concentrations away from major sources that are not adequately
3 captured by the existing monitoring networks (see Figure 2-10 for location of monitoring
4 sites). Satellite data coupled with model simulations might be more useful for showing
5 large-scale features in the distribution of NO2. Winter and summer seasonal average NO2
6 concentrations for 2009-2011 derived from the OMI instrument on the AURA satellite
7 and the GEOS-Chem global, three-dimensional chemistry-transport model are shown in
8 Figure 2-13. In this method, integrated vertical column abundances of NO2 derived from
9 the OMI instrument are scaled to surface mixing ratios using scaling factors derived from
10 GEOS-Chem [see (Lamsal et al., 2010; Lamsal et al., 2008); also see Section 2.4 for
11 more complete descriptions of the method]. A nested version of GEOS-Chem at
12 50 km x 50 km horizontal resolution is used in this method. A description of the
13 capabilities of GEOS-Chem and other three-dimensional chemistry transport models is
14 given in the ISA for Ozone and Related Photochemical Oxidants (U.S. EPA. 2013b).
15 Large variability in NO2 concentrations is apparent in Figure 2-13. As expected, the
16 highest NO2 concentrations are seen in large urban regions, such as in the Northeast
17 Corridor, and lowest values are found in sparsely populated regions located mainly in the
18 West. Minimum hourly values can be less than ~10 ppt, leading to a large range between
19 maximum and minimum concentrations. Although overall patterns of spatial variability
20 are consistent with the current understanding of the behavior of NO2, not much
21 confidence should be placed on values <~100 ppt due to limitations in the satellite
22 retrievals. NO2 concentrations tend to be higher in January than in July, largely reflecting
23 lower planetary boundary layer heights in winter. Such seasonal variability is also evident
24 on a local scale, as measured by surface monitors. For example, in Atlanta, GA, NOx
25 measurements also exhibited higher concentrations in winter and lower concentrations in
26 summer, when NOx is more rapidly removed by photochemical reactions (Pachon et al.,
27 2012).
January 2015 2-42 DRAFT: Do Not Cite or Quote
-------
O Ml-derived surface NO2 (ppb)
Note: Images shown were constructed by Dr. Lok Lamsal of Universities Space Research Association from data obtained by the
OMI instrument on the AURA satellite (http://aura.asfc.nasa.gov/scinst/omi.html) using the algorithm described in Bucsela et al.
(2013). Output from the GEOS-Chem, global-scale, three-dimensional, chemistry-transport model to derive surface concentration
fields from the satellite data as described in Lamsal et al. (2008) and Lamsal et al. (2010).
Top panel (winter; DJF: December, January, February). Lower panel (summer; JJA: June, July, August).
Figure 2-13 Seasonal average surface nitrogen dioxide (NCh) concentrations
in ppb for winter (upper panel) and summer (lower panel) derived
by ozone monitoring instrument (OMI)/Goddard Earth Observing
System (GEOS)-Chem for 2009-2011.
January 2015
2-43
DRAFT: Do Not Cite or Quote
-------
2.5.2 Urban-Scale Spatial Variability
1 Figure 2-14 describes 1-hour daily maximum concentration agreement between pairs of
2 SLAM monitors from 2011-2013 for selected U.S. CBSAs with more than one monitor.
3 Agreement is expressed as coefficient of divergence (COD), which has been widely used
4 to assess spatial variability of air pollutant concentrations (U.S. EPA, 2008a; Wilson
5 etal.. 2005; Pinto etal.. 2004). In practical terms, a COD = 0 indicates perfect agreement
6 and COD values increase as spatial variability increases. COD values in Figure 2-14
7 generally range from about 0.1 to 0.4, with a few higher values. This indicates a range of
8 variability across CBSAs from fairly uniform to a moderate degree of variability (Wilson
9 etal.. 2005). At first glance, distance between sites does not appear to be an important
10 factor for explaining variability between site pairs on an urban scale. However, for
11 extremely short distances, a trend with distance is observed, especially for data within the
12 same city. For example, for Boston, MA, the six observations with the shortest distances
13 between them exhibit a trend of increasing COD with distance, from about 3 km to about
14 10 km. These data are for the four sites closest to the city center. As indicated by the
15 COD values, there is a substantial degree of variability for all but the closest sites, with
16 CODs ranging above 0.4 even for comparison between site pairs near the city center.
January 2015 2-44 DRAFT: Do Not Cite or Quote
-------
0.6-
0.5-
8
5
E> n
a
•&
e
•
•r>
E 0.3-
8
o
0.2-
0.1-
A o Los Angeles. CA
+ Phoenix, AZ
A x Philadelphia. PA
A A Boston. MA
A
O - ^ O
A A °
o o o
A O
Ł
A A O + Q O o
o o o
o o
A Q On 6 o O ^ °
A X o X ^ o o °
A _y O O O
A 0 0 0 x ^ ^-n00 ° o 0 0
A 9, @ o ° ? °° °
+o ° ^?^fto a oP° ° < ° o
O O c^> O O *" O
oo S3 oo
Ao++>^9 0 0®
I I 1 t I I
0 20 40 60 80 100
Distance between Monitors {km)
Source: National Center for Environmental Assessment 2014 analysis of Air Quality System network data.
Figure 2-14 Coefficient of divergence between monitor pairs in four U.S.
cities.
A similar trend is observed for Los Angeles, CA but over a broader scale. The highest
COD increases somewhat regularly with distance, up to about 40 km. Also, for all sites
within 15 km of each other, a fairly high degree of agreement is observed. Another major
difference between Boston and Los Angeles is that in Los Angeles good agreement
(COD ~0.1) is often observed between sites up to 50 km or more apart, suggesting that
other factors besides distance are important. Five of the Los Angeles monitors (Main
Street Los Angeles, Burbank, Pasadena, Pomona, and Long Beach North) form a subset
of monitors with distinctly lower variability than the area as a whole, with low CODs for
each possible combination of monitors within this group, as shown in Figure 2-15. Other
sites near the ocean or mountains exhibit poorer agreement with these monitors, even if
the distances between them are shorter.
January 2015
2-45
DRAFT: Do Not Cite or Quote
-------
CD
U
|
OJ
u
o
u
Southeast Los Angeles (Excluding Orange Co.)
0.25
0.2
0.15
o
(
1
0 2
0 3
0 4
0 :
0 6
Distance between Monitors (km)
Source: National Center for Environmental Assessment 2014 analysis of Air Quality System network data.
Figure 2-15 Coefficient of divergence among a subset of five Los Angeles, CA
monitors.
i
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Yet another pattern is observed for Phoenix, AZ and Philadelphia, PA. For these cities,
low COD values are observed for all sites except rural locations outside of the urbanized
area. The Phoenix data in Figure 2-14 fall into two clusters, with urban site pairs ranging
up to 10 km distance from each other and urban-rural site pairs 40 to 60 km from each
other. All of the comparisons between urban sites exhibit a COD <0.2, but poorer
agreement is observed between urban and rural site pairs. Similarly, good agreement
(COD ~0.1) is observed between two monitoring sites operating within the city of
Philadelphia about 10 km apart, but poorer agreement is observed for more distant
suburban sites. This result is consistent with observations of Sarnat et al. (2010). who
observed that using monitors in rural areas of counties considered part of the Atlanta, GA
metropolitan area affected relative risk estimates for associations with health effects, but
that using different urban monitors within approximately 15 to 30 km of a study subject
did not.
To summarize urban scale spatial variability for NO2, good agreement between nearby
sites in city cores is not unusual, and was observed for 2011 to 2013 data for all sites in
January 2015
2-46
DRAFT: Do Not Cite or Quote
-------
1 Philadelphia and Phoenix. In Los Angeles, good agreement was also usually observed
2 over similar distances to those compared in Philadelphia and Phoenix (i.e., between sites
3 separated by less than 15 km). In contrast, agreement between monitors in Boston
4 became poorer over a shorter distance, but followed a trend of increasing variability with
5 distance between sites over 3 to 10 km, a smaller spatial scale than the other cities.
6 Similar results are observed for annual averages. Tables 2-5A and 2-5B present the
7 difference in annual average NO2 concentrations between pairs of sites divided by the
8 average between the two sites for that year to get a percent difference in concentration for
9 Boston and Los Angeles. The CODs of annual average NO2 concentrations show wide
10 ranges in agreement similar to those reported for 1-hour daily maximum NC>2
11 concentrations for the NC>2 concentrations measured in both Boston and Los Angeles.
12 The nearest site pairings in Boston agree within 3 to 20%, while the other two site
13 pairings exhibit poorer agreement ranging from 38 to 65% and 31 to 90%.
14 For Los Angeles, the 14 sites in Los Angeles County that reported data for 2011 are
15 shown in Table 2-5B. A number of site pairings agree within 10 to 15%. For example,
16 concentrations at the Pico Rivera, Pomona, and Long Beach Hudson sites all agree within
17 10% of the concentrations reported at the Los Angeles Main Street site.
Table 2-5A Percent difference in annual average nitrogen dioxide concentration
between monitors in Boston.
2011
2012
2013
0002 vs. 0040
%
41
65
38
0002 vs. 0042
%
10
20
3
0040 vs. 0042
%
31
47
90
Source: National Center for Environmental Assessment Analysis of Air Quality System Network Data from 2011 -2013.
January 2015 2-47 DRAFT: Do Not Cite or Quote
-------
Table 2-5B Percent difference in annual average nitrogen dioxide concentration
between monitors in Los Angeles 2011.
Site
0016
0113
1103
1201
1302
1602
1701
2005
4002
4006
5005
6012
9033
0002
%
7
17
21
16
3
22
26
7
7
12
32
35*
47
0016
%
20
57
22
35
58
62
44
31
49
5
3
10
0113
%
38
2
15
39
43
24
10
29
15
18
30
1103
%
36
23
1
5
14
28
9
52
55
66
1201 1302 1602 1701 2005 4002 4006 5005 6012
% % % % % % % % %
13
37 25
41 28 4
23 10 15 19
9 4 29 33 14
28 15 10 14 5 19
17 30 53 57 39 25 44
20 32 56 60 42 28 47 3
32 45 67 71 54 40 58 4 13
Source: National Center for Environmental Assessment Analysis of Air Quality System Network Data from 2011.
1 While these results indicate that relatively good agreement in 1-hour daily maximum and
2 annual average NO2 concentrations between pairs of nearby urban monitors in the same
3 metropolitan area occurs in some cases, it does not rule out the possibility of greater
4 variability on a smaller spatial scale. Vardoulakis et al. (2011) described a distinction
5 between "intra-urban" and "street scale" variability, explaining that long-term monitoring
6 sites tend to be situated away from sources and hot spots that can strongly influence
7 variability. They compared results from long-term monitoring sites to short-term
8 networks of passive samplers placed in areas between the long-term monitors at varying
9 distances from key roads and intersections, and found that "street level" variability of
10 passive sampler measurements exhibited greater variability than "intra-urban" variability
11 based on long-term monitors. Spatial variability near roads is described in detail in the
12 following section.
January 2015 2-48 DRAFT: Do Not Cite or Quote
-------
2.5.3 Microscale- to Neighborhood-Scale Spatial Variability, Including near
Roads
2.5.3.1 Near-Road Gradient Observations
1 The spatial trends described in this section provide a background for understanding traffic
2 related NO2 exposure on and near roads, described in Section 3.3.1.1. Numerous
3 observations have been summarized in several recent reviews, each of which concluded
4 that a zone of elevated NO2 concentration typically extends from 200 to 500 m from
5 roads with heavy traffic (HEI. 2010; Karner etal.. 2010; Zhou and Levy. 2007).
6 Table 2-6 describes observations from studies that were included in these reviews and/or
7 in the 2008 Risk and Exposure Assessment for Oxides of Nitrogen (U.S. EPA. 2008b) to
8 estimate on-road concentrations, as well as more recent observations. A direct
9 comparison of the observations included in Table 2-6 is not appropriate because different
10 experimental designs, measurement methods, averaging times, distances from the road,
11 time of year, and other important factors vary among studies. However, Table 2-6
12 provides a broad overview of the magnitudes of concentration differences observed and
13 the spatial extent over which differences in concentration have been observed.
January 2015 2-49 DRAFT: Do Not Cite or Quote
-------
Table 2-6 Summary of near-road nitrogen dioxide concentration gradients from different studies.
Author
Gilbert etal. (2003)
Monn(2001)
Pleiiel etal. (2004)
Roorda-Knape et al.
(1998)
Singer etal. (2004)
Smarqiassi et al. (2005)
Beckerman et al. (2008)
Zou et al. (2006)
Gonzales et al. (2005)
Cape et al. (2004)
Bell and Ashenden
(1997)
Biqnaletal. (2007)
Uemura et al. (2008)
Maruo et al. (2003)
Rodes and Holland
(1981)
Location
Montreal
Zurich
Rural
Sweden
Netherlands
Oakland
Montreal
Toronto
Shanghai
El Paso
Scotland
Rural Wales
Rural
England
Tokyo
Sapporo
LA— high Os
medium Os
Low 03
Method
Passive
Passive
Passive
Passive
Passive
Passive
Passive
Passive
Passive
Passive
Passive
Passive
Passive
Sensor
Chemilum.
Time of
Year
September
Summer
Winter
Fall/Spring
May-June
August
All year
Winter
April-May
All year
All year
July
July-August
Averaging
Time
1 week
1 week
1 month
2 weeks
1 week
1 day
1 week
2 weeks
1 week
1 week
1 week
11 -17 days
48 h
1/2 day
1 h
Nearest
Cone.
Cnear
29 ppb (0 m)
~20ppb(15m)
-25 ppb (15m)
8-18 ppb
(10m)
24-25 ppb
30 ppb (60 m)
33 ppb (<10 m)
(4-28 m)
-50-65 ppb
(Om)
-25 ppb
(0.25 m)
3-50 ppb (1 m)
8-28 ppb
(<1 m)
-25 ppb
35-47 ppb
(<50 m)
-28-43 ppb
-120 ppb (8m)
-80 ppb (8 m)
-70 ppb (8 m)
Farthest
Cone.
Cfar
18 ppb (200 m)
-12 ppb (80m)
-25 ppb (80 m)
4-10 ppb
(100m)
•\c -17 nnh
(-300 m)
20 ppb (>350 m)
20 ppb (>1 km)
OQ ylQ nnh
(350 m)
-15 ppb
3-40 ppb (10m)
2-14 ppb
(200-350 m)
5-15 ppb
(250 m)
oo 07 nnh
(>200 m)
— OH O°, nnh
(150m)
-40 ppb (388 m)
-40 ppb (388 m)
-40 ppb (388 m)
Spatial
Extent
200m
>80 m
None
500 m
>300 m
350 m
300 m
>350 m
>3.75 km
>10 m
~100m
Ann f^nn m
400-500 m
400-500 m
Difference
(Cnear ~~ Cfar)
/Cfar
%
60
>30
-0
80-100
-I n fin
60
70
on -1 nn
-30-40
-70
<0-70
20-660
70-400
-40
-200
-100
-80
Difference
Cnear ~~ Cfar
ppb
11 ppb
-8 ppb
-0 ppb
2-8 ppb
-7 ppb
11 ppb
13 ppb
-12-18 ppb
-10 ppb
0-11 ppb
3-20 ppb
-10-20 ppb
<10 ppb
-10 ppb
80 ppb
40 ppb
30 ppb
January 2015
2-50
DRAFT: Do Not Cite or Quote
-------
Table 2-6 (Continued): Summary of near-road nitrogen dioxide (NO2) concentration gradients from different
studies.
Massoli
Polidori
Author
etal. (2012)
and Fine (2012)
Kimbrouqh et al. (2013)
McAdam et al. (2011)
Durantetal. (2010)
Nitta et
al. (1993)
Nakaietal. (1995)
Location
New York
Southern
Calif
Raleigh
Downwind
only
Ontario
Somerville,
MA
Tokyo
Tokyo
Method
Chemilum.
Chemilum.
Chemilum.
Chemilum.
TILDAS
Colorimetry
Colorimetry
Time of
Year
July
Summer
Winter
All year
Summer
January
All year
All year
Averaging
Time
4:30-9:30
a.m.
1 h
5 min
1 h
Real time
(a.m.)
Continuous
Continuous
Nearest
Cone.
Cnear
25-40 ppb
(10m)
28 ppb (15m)
37 ppb (15m)
25 ppb (20 m)
28 ppb (20 m)
0-30 ppb
(10 m)
-15-35 ppb
(<50 m)
34-57 ppb
(Om)
46 ppb (0 m)
Farthest
Cone. Spatial
Cfar Extent
25-40 ppb None
(500 m)
18 ppb (80 m)
32 ppb (80 m)
20 ppb (300 m)
23 ppb (300 m)
0-17 ppb (60m) None
(400 m)
24-42 ppb
(150m)
35 ppb (150 m)
Difference
(Cnear ~ Cfar)
/Cfar
%
~0
56
15
30
20
~0
>0
10-50
30
Difference
Cnear ~~ Cfar
ppb
~0
10 ppb
5 ppb
5 ppb
5 ppb
<1 ppb
(avg)
<10 ppb
8-17 ppb
11 ppb
Cfar = concentrations measured at the farthest distance; Cnear = <
January 2015
2-51
DRAFT: Do Not Cite or Quote
-------
1 If sufficient detail is given on individual experiments, ranges for concentrations measured
2 at the nearest distance (Cnear), concentrations measured at the farthest distance (Gar), and
3 differences between them are provided in Table 2-6. Otherwise averages over the entire
4 study, or over various categories such as season, wind direction, location, or Os
5 concentration are given. Studies using passive sampling methods are listed first in
6 Table 2-6. With the exception of Rodes and Holland (1981). earlier studies were mainly
7 limited to passive samplers that required collection for 1 or 2 weeks, making it difficult to
8 explore effects of time of day or wind direction, which typically shifts on shorter time
9 scales. More recently, more studies have used chemiluminescence, QC-TILDAS, and
10 other methods. These methods not only provide greater time resolution, but also result in
11 the collection of larger numbers of samples, both of which are useful for better
12 understanding the factors influencing near-road concentration patterns. There are
13 essentially three types of experimental designs used in the studies listed in Table 2-6:
14 (1) samples are collected simultaneously at varying distance from the same road;
15 (2) samples are collected by a mobile laboratory with high time resolution, with samples
16 collected at different distances from a road, not simultaneously, but with minimal elapsed
17 time between sampling at different distances from the road; or (3) samples are collected
18 over a wider spatial scale at varying distances from a number of heavily trafficked roads
19 and distance parameters are not linked to the same road for all samples.
20 Most of the studies conclude that the spatial extent of elevated NO2 concentrations is
21 within the range of the 200 to 500 m described by HEK2010) or Zhou and Lew (2007).
22 However, some recent studies (not necessarily included in Table 2-6) concluded that the
23 influence of the road on NCh concentrations can still extend to several kilometers, but
24 with smaller differences in concentration (Gilbert et al.. 2007; Gonzales et al.. 2005). In
25 other studies, remarkably greater differences in NC>2 concentration with distance were
26 observed within 10 to 20 m of the road than at further distances from the road, suggesting
27 the possibility of an exponential relationship of decreasing concentration with distance
28 from the road with a steeper decrease right next to the road than further than 10 to 20 m
29 from the road (Cape et al.. 2004; Bell and Ashenden. 1997). Compared to NO, ultrafine
30 particles (UFP), and other traffic-related pollutants, NO2 concentrations decrease less
31 rapidly with distance from the road over a range of about 200 to 500 m, and exhibit a
32 somewhat greater spatial extent of elevated concentration (Gordon etal.. 2012; Chaney
33 etal.. 2011: McAdametal.. 2011: Beckerman et al.. 2008: Zhou and Lew. 2007). This
34 has been attributed to chemical production downwind of roads (cf Section 2.2) and other
35 nontraffic related sources of NO2 (Chaney et al.. 2011: Zhou and Levy. 2007: Rodes and
36 Holland. 1981). Because of the interplay between dispersion and chemical reaction
37 described in Section 2.2. the distribution of NO2 downwind of roads would likely differ
38 from that of a strictly primary traffic pollutant. For example, Massoli et al. (2012). in a
January 2015 2-52 DRAFT: Do Not Cite or Quote
-------
1 study examining the behavior of traffic-related pollutants upwind and downwind of the
2 Long Island Expressway (LIE) during morning rush hour(s), found that the
3 concentrations of carbon dioxide [CO2] and NOx are highest closest to the highway and
4 decreased by approximately 50% within 150 m downwind of the LIE, whereas the
5 concentration of NO2 was found to be nearly constant over this distance from the road.
6 Table 2-6 also compares differences in Cnear and Cfar distance reported in each study. The
7 nearest distance ranges from 0 to 60 m and the farthest distance from 80 m to several km.
8 In early studies using passive monitors with usual sampling periods of 1 to 2 weeks, NO2
9 Cnear ranged from 30 to 100% higher than Cfar in the majority of the observations listed.
10 A few observations of NO2 concentrations were more than 100% higher at the location
11 nearest the road than at the location farthest from the road. Most of these greater than
12 usual differences were observed when Cfar was much lower than usual. This is illustrated
13 in Figure 2-16. which shows that on a major road in a rural area of Great Britain (Bell
14 and Ashenden. 1997) NO2 Cnear ranged up to six times higher than Cfar, but the greatest
15 differences were observed only when Cfar was lower than usual. Differences were
16 consistently greater than 200% when Cfar was less than 4 ppb, but less than 100% when
17 Cfar exceeded 10 ppb. Because Cfar was so low, even for the greatest differences in
18 concentrations observed by Bell and Ashenden (1997). the absolute difference in
19 concentration between distances of <1 m and 200 m never exceeded 20 ppb. Differences
20 of similar magnitude were observed by Bignal et al. (2007) for a British rural area where
21 Cfar ranged from 5 to 10 ppb. Because data were collected in a rural area, the differences
22 observed by Bignal et al. (2007) would not necessarily be applicable for absolute
23 differences that might be observed in urban areas where NO2 concentrations are typically
24 higher. However, Figure 2-16 they clearly demonstrates that when concentration
25 differences are expressed in terms of a ratio of concentrations, the ratio observed at an
26 average concentration could be greater than that observed at higher concentrations.
27 Although a concentration dependence for (Cfar - Cnear)/Cfar was observed for rural areas in
28 Figure 2-16. low Cfar measurements do not explain all of the high ratios of
29 (Cfar - Cnear)/Cnear in Table 2-6. Rodes and Holland (1981). observed concentration ratios
30 for (Cfar ~ Cnear)/Cfar ranging from 100 to 200% along with Cfar concentrations averaging
31 about 40 ppb, which they attributed to rapid formation of NO2 between the road and the
32 nearest monitor 8 m from the road because of high Os concentrations. Most of the NOx
33 emitted from vehicles is emitted as NO, which can be rapidly converted into NO2 in the
34 presence of Os as described in Section 2.2. However, the results of Table 2-6 in general
35 indicate that concentrations nearest the road rarely appear to be more than 100% higher at
36 a distance of 10 to 20 m from the road than at 80 to 500 m away from the road.
January 2015 2-53 DRAFT: Do Not Cite or Quote
-------
J5
0
0 2 4 6 8 10 12 14 16
Cfar (ppb)
Source: National Center for Environmental Assessment analysis of data from Bell and Ashenden (1997).
Figure 2-16 Influence of nitrogen dioxide concentration magnitude on the
ratio of NO2 concentrations at <1 m from the road (Cnear) to
concentrations at 200-350 m (Cfar) in rural Wales.
1 As studies with better time resolution have been conducted, more observations of the lack
2 of any difference between concentrations nearest the road and farthest from the road
3 (Cnear ~ Cfar = ~0) have been reported. Monn (2001) observed little difference in winter
4 with passive samplers, and in more recent studies a lack of near-road NO2 concentration
5 gradient appears to be common in early morning measurements (Massoli etal.. 2012;
6 McAdametal.. 2011).
7 There are now thousands of individual chemiluminescence measurements from two new
8 studies (Kimbrough et al.. 2013; Polidori and Fine. 2012) that solidly support
9 observations from earlier passive sampling studies. Kimbrough et al. (2013) reported
10 average concentrations of more than seven thousand 5-minute measurements, and
11 showed that NO2 concentrations 20 m from the road were an average of 27% higher than
12 at 300 m from the road in Las Vegas, NV. In Southern California, Polidori and Fine
13 (2012) reported that NCh concentrations at 15 m were 56% higher in summer and 15%
14 higher in winter than at 80 m. This seasonal difference has been noted in other studies
15 (Monn. 2001) and is consistent with the difference in seasonal ratio distributions
16 estimated in the 2008 Risk and Exposure Assessment for Oxides of Nitrogen (U.S. EPA.
17 2008b). Averaging over the two seasons gives an NC>2 concentration 36% higher at 15 m
18 than at 80 m, which is remarkably similar to the observation of Kimbrough et al. (2013).
January 2015 2-54 DRAFT: Do Not Cite or Quote
-------
1 The absolute difference in measured NO2 concentrations between the nearest and farthest
2 locations (Cnear - Car) is also consistent across most studies, with concentration
3 differences rarely exceeding 20 ppb. The exception is the Rodes and Holland (1981)
4 study from Los Angeles in the early 1980s. Because this is an older study than the others,
5 the vehicle fleet was not strictly regulated for NOx emissions. As a result, the
6 concentrations observed may not be relevant to current conditions. Excluding this study,
7 the range in Cnear - Cfar is somewhat smaller than the range for Cfar across all of the
8 studies, which further implies that a ratio of concentrations at different distances from the
9 road could be more strongly influenced by the concentration away from the road (Car)
10 than by the concentration nearest the road (dear).
1 1 Several investigators have attempted to fit NO2 concentration data as a function of
12 distance from the road. NC>2 concentrations followed a logarithmic function with distance
13 from a road over a range of 100 m (Pleijel et al.. 2004). more than 300 m (Roorda-Knape
14 etal.. 1998). and more than 1,000 m (Gilbert et al.. 2003):
Cx = Cb + Cv — A"log x
Equation 2-1
15 where
16 x = distance from the road
17 k = decay constant derived from empirical data
18 Cx = NCh concentration at a distance x from a road
19 Cb = NC>2 concentration contribution away from the influence of the road
20 Cv = NCh concentration contribution from vehicles on a roadway
21 Cape et al. (2004) used an exponential decay function to fit measured NO2 concentrations
22 measured from 1 m to 10m from the road:
Equation 2-2
23 This approach was also used in the 2008 Risk and Exposure Assessment for Oxides of
24 Nitrogen (U.S. EPA. 2008b) to estimate on-road concentrations from the published
25 studies.
January 2015 2-55 DRAFT: Do Not Cite or Quote
-------
A shifted power law model has also been used (Zou et al.. 2006):
Equation 2-3
2 Table 2-6 also provides some preliminary insight into factors that affect near-road
3 concentrations. For example, the difference between NO2 concentrations at different
4 distances from a road is consistently greater in summer than in winter (Kimbrough et al..
5 2013: Monn. 2001: Bell and Ashenden. 1997). as also described in the 2008 Risk and
6 Exposure Assessment for Oxides of Nitrogen (U.S. EPA. 2008b).
7 There is less consistency concerning time of day. Near-road NO2 concentrations have
8 been observed consistently to decrease after sunrise (Gordon et al.. 2012: Massoli et al..
9 2012: Durant et al.. 2010). However, both increases (Gordon et al.. 2012) and decreases
10 (Durant et al.. 2010) in concentration gradient have been observed after sunrise. In the
11 morning, observations of no variation of NO2 concentration with distance for short time
12 intervals have been observed before sunrise (Gordon et al.. 2012). after sunrise (Durant
13 etal.. 2010). or both before and after sunrise (Massoli etal.. 2012).
14 A slight effect of wind conditions has been observed. Concentration varies with distance
15 from the road under all wind conditions, but is more pronounced downwind from the
16 road (Kimbrough et al.. 2013: McAdam etal.. 2011: Beckerman et al.. 2008: Roorda-
17 Knape etal.. 1998). When air is sampled both upwind and downwind of the road, more
18 gradual gradients are observed on the downwind side of the roadway (Durant et al.. 2010:
19 Clements et al.. 2009: Hu et al.. 2009: Beckerman et al.. 2008). Also, higher
20 concentrations are observed at low wind speeds, especially for winds blowing from the
21 road (Kimbrough et al.. 2013).
22 To summarize the results and conclusions of the studies in Table 2-6. a zone of elevated
23 NO2 concentration typically extends up to a distance of 200 to 500 m from roadways with
24 sufficient traffic. NO2 concentrations measured from 0 to 20 m from the road range up to
25 20 ppb higher, or up to 100% higher than concentrations measured between 80 and 500 m
26 from a road. More recent data from intensive studies suggest concentrations at 15 to 20 m
27 average 20-40% higher than concentrations 80 m from the road, with greater differences
28 during daylight hours and in the summer.
29 Figure 2-17 explores the near-road gradient in more detail for a single study, providing
30 observations of seasonal and diurnal variations based on near-road field measurements in
31 Los Angeles, CA (Polidori and Fine. 2012) and Detroit, MI (Vette etal.. 2013). In both
32 cities, hourly NO2 measurements at near-road (within 15m from a major interstate) and
January 2015 2-56 DRAFT: Do Not Cite or Quote
-------
1 downwind sites (within 100 m from major interstate) exhibit the evolving nature of NO2
2 concentrations and roadway gradients during different seasons and hours of the day. On a
3 seasonal basis, higher average NO2 concentrations generally occur during winter months,
4 when atmospheric inversions are more prevalent, although some of the highest
5 concentrations in Detroit are observed during the day in summer. Additionally, because
6 absolute NO2 concentrations tend to decrease more during the summer when
7 concentrations are lower compared to winter, larger ratios of concentrations at different
8 distances from the road are observed during the summer, as described earlier in this
9 section.
10 On a diurnal basis, NO2 roadway concentrations typically increase during morning rush
11 hour (6:00-10:00 a.m.) then gradually decrease from late morning to mid afternoon as
12 the atmospheric mixing layer expands. Roadway NO2 concentrations begin to increase
13 again during afternoon rush hour and nighttime, and are generally similar to or slightly
14 lower than NO2 concentrations during morning rush hour. As demonstrated in
15 Figure 2-17. this diurnal profile is more evident in the winter compared to the summer.
16 Notably, while maximum concentrations tend to occur during morning rush hour and
17 nighttime, especially during the winter, the NO2 roadway gradient is largest during
18 afternoon hours (10:00 a.m.-5:00 p.m.). This trend is further demonstrated in
19 Figure 2-18. which shows the absolute difference in NO2 concentrations between
20 near-road and downwind sites during winter and summer. In both cities, the absolute
21 difference between sites is below 15 ppb during morning rush hour and nighttime,
22 whereas a somewhat larger difference is observed during mid-afternoon hours
23 (12:00 p.m.-5:00 p.m.).
January 2015 2-57 DRAFT: Do Not Cite or Quote
-------
Detroit Winter
Detroit Summer
10m 100m
0246
10 12 14 16 18 20 22
Hour
8 -
_
CL
Q.
10m
100m
0 2 4 6 8 10 12 14 16 18 20 22
Hour
8 -I
Los Angeles Winter
0 2 4 6 8 10 12 14 16 18 20 22
-S <=• H
Q. to
Los Angeles Summer
• 15m
80m
0 2 4 6 8 10 12 14 16 18 20 22
Hour
Source: National Center for Environmental Assessment analysis of data obtained from Polidori and Fine (2012) and Vette et al.
(2013). The box represents the interquartile range of concentrations observed during a given hour, with the 10th and 90th
percentiles of concentrations shown by bottom and top whiskers, respectively.
Figure 2-17 Diurnal variation of near-road (red: within 15 m of major
interstate) and downwind (gray: within 100 m of major interstate)
nitrogen dioxide (NOa) concentrations observed during year-long
field campaigns in Los Angeles, CAand Detroit, Ml.
January 2015
2-58
DRAFT: Do Not Cite or Quote
-------
Detroit
Los Angeles
Q.
Q.
o «
z
Ł
o °
0
-------
1 of all 1-hour NC>2 measurements from all monitors in the city, but usually not the highest
2 98th percentile or maximum concentration.
Table 2-7 Comparison of near road and area wide 1-hour daily maximum
concentrations for monitors with year round data (ppb).
AW NR
City Period Sites Season Mean
Boston 6/2013-5/2014 7 Warm 26.2
Cold 35.5
Denver 6/2013-5/2014 2 Warm 39.0
Cold 45.0
Des Moines 2013 1 Warm 16.7
(all year)
Cold 21.3
Detroit 2012-2013 2 Warm 30.4
(2 years)
Cold 30.0
Minneapolis 4/2013-3/2014 3 Warm 22.2
Cold 31.1
NR
AW Mean 98%
7-27 43
13-34 58
32-37 54
41-47 71
15 30
17 35
22-23 50
25-26 51
9-15 42
15-24 51
AW 98%
16-49
33-63
62-77
71-83
41
35
40-42
42-45
20-38
31-59
NR
Max
51
64
62
97
34
42
78
71
54
52
AW Max
25-70
40-68
58-63
106-136
44
39
51-54
48-54
26-42
43-70
AW = area wide; NR = near road
Source: National Center for Environmental Assessment and Office of Air Quality Planning and Standards Analysis of Air Quality
System Network Data from 2012-2014.
3
4
5
6
7
8
9
10
11
Comparing near-road NO2 concentrations to those measured in another part of the city is
not the same as comparing them to concentrations a few hundred meters away from the
road, as illustrated in Figure 2-14. In many cases, the other monitors have been
intentionally sited in areas of high NC>2 concentrations, and considerable variability has
been observed on the same spatial scale as the distance between monitors (e.g., Boston in
Figure 2-14). A good example of this is the highest 1-hour daily maximum NCh
concentration of 136 ppb in Denver in Table 2-7 at a monitor located 3 km from the
near-road monitor, but one block from high rise buildings that form the edge of the
high-density central business district.
January 2015
2-60
DRAFT: Do Not Cite or Quote
-------
1
2
o
6
4
5
6
7
8
9
10
11
12
Additional monitors have not yet operated for a full year, but were operational in the first
quarter of 2014. Results from these monitors are presented in Table 2-8. 1-hour daily
maximum NC>2 concentrations ranged from 49 ppb in Houston, TX to 97 ppb in Denver,
CO, with 98th percentile concentrations for the quarter ranging from 48 to 71 ppb. More
than half of the monitors reported maximum and 98th percentile concentrations of 1-hour
daily maximum NC>2 concentrations exceeding 60 ppb. Seasonal average NO2
concentrations ranged from 23.5 to 46.0 ppb, with the highest average concentrations in
Denver, CO and Phoenix, AZ. These are two of only a few cities that fell into the highest
concentration classes for both 1-hour daily maximum and annual average concentrations
of NO2 during the previous 3 years, as indicated in Figures 2-11 and 2-12 in
Section 2.5.1. This suggests other influences besides roads with heavy traffic can also
influence NO2 concentrations, even at locations very close to the road (i.e., within 50 m).
Table 2-8 Near-road network 1-hour daily maximum nitrogen dioxide
concentration summary for first quarter 2014 (ppb).
City
Denver, CO
St. Louis, MO
Cincinnati, OH
Philadelphia, PA
Indianapolis, IN
Boston, MA
Phoenix, AZ
Detroit, Ml
S.F. -Oakland, CA
Richmond, VA
Birmingham, AL
Columbus, OH
Minneapolis, MN
San Antonio, TX
Houston, TX
Maximum
97
71
68
65
64
64
62
62
61
59
55
53
52
51
49
98th Percentile
71
66
67
60
64
60
62
61
55
55
54
53
51
51
48
Seasonal Mean
44.7
35.3
42.3
36.3
38.4
36.8
46.0
34.9
30.2
34.6
23.5
29.3
32.9
28.1
29.2
Source: Office of Air Quality Planning and Standards and National Center for Environmental Assessment analysis of Air Quality
System Network Data 2014.
January 2015
2-61
DRAFT: Do Not Cite or Quote
-------
1 These preliminary results from the U.S. near-road network are similar to data from the
2 London, U.K. network with several near-road monitors that have been in operation for a
3 longer period, despite potential differences from the U.S. in fleet mix, traffic mitigation
4 policies, and small geographic scope that may limit generalizability. London data were
5 analyzed because the city has a well-established system of roadside and urban
6 background monitors. Air quality data were obtained from the Airbase database
7 (EIONET. 2014) for 2004 to 2006 and 2010 to 2012 in the form of hourly NO2
8 measurements, and monitors of interest were those whose city was listed as London and
9 were within 10m of the roadway to capture NC>2 primarily derived from mobile sources.
10 Overall, there were large differences in NO2 concentrations between roadside and urban
11 background monitors, which ranged from 2.4 to 9.8 km apart as shown in Tables 2-9A
12 and 2-9B. The differences in 24-h avg NCh concentrations ranged from approximately
13 24% lower to 170% higher at the roadside than urban background site. The largest
14 relative differences in 24-h avg NC>2 concentrations were observed when the ambient
15 urban background concentrations were less than 20 ppb. All roadside monitors were
16 positively correlated with the overall urban background monitors, and the Pearson
17 correlations were greater than r = 0.70 for two out of three. Interquartile ranges were
18 generally similar between roadside monitor-urban background monitor pairs, indicating
19 that while in the majority of cases roadside monitors had higher NC>2 concentrations than
20 urban background monitors, temporal variability was similar between the two monitors.
21 As for the preliminary results from the U.S. near-road network, these results suggest that
22 while concentrations measured at roadside monitors were generally higher than those
23 measured at urban background monitors, there were still large ranges in mean
24 differences.
January 2015 2-62 DRAFT: Do Not Cite or Quote
-------
Table 2-9A Roadside and urban background nitrogen dioxide concentrations in London, U.K. 2010-2012.
Roadside
Urban Bkg
Roadside
Urban Bkg
Roadside
Urban Bkg
Distance
Between
Monitors
Monitor Pairs km
London Marylebone Rd 2.4
London Bloomsbury
Camden Kerbside 2.8
London N.Kensington
Haringey Roadside 9.8
London Bloomsbury
Mean
Concentration A Mean3
ppb %
52.34 68
31.17
43.76 124
19.58
23.78 -24
31.17
98th
Percentile of
1 -Hour Daily A 98th
Maxb Percentile 24-H Avg
ppb % IQR
140.53
69.68
132.02
63.56
64.22
69.68
102 25.7
12.01
108 16.89
12.61
-7.9 12.03
12.01
1-H Max
IQR
59.58
14.36
32.45
15.96
18.62
14.36
24-H Avg. Correlation
With Urban
Background Monitors
95% Cl
0.30
(0.25,
0.74
(0.71,
0.84
(0.83,
0.36)
0.77)
0.86)
Cl = confidence interval; IQR = interquartile range.
aDifference in mean NO2 between roadside and urban background monitors.
b3-year average.
Source: National Center for Environmental Assessment Analysis of European Air Quality Database Data from 2010-2012.
January 2015
2-63
DRAFT: Do Not Cite or Quote
-------
Table 2-9B Roadside and urban background nitrogen dioxide concentrations in London, U.K.
Roadside
Urban Bkg
Roadside
Urban Bkg
Roadside
Urban Bkg
Roadside
Urban Bkg
Roadside
Urban Bkg
Roadside
Urban Bkg
Roadside
Urban Bkg
Roadside
Urban Bkg
Distance
Between
Monitors
Monitor Pairs km
London Marylebone Rd 2.4
London Bloomsbury
Southwark Roadside 3.3
London Eltham
London Cromwell Rd 2 3.4
London Bexley
Camden Kerbside 3.8
London N.Kensington
Tower Hamlets Roadside 4.1
London Hackney
Haringey Roadside 4.5
London Hackney
London Bromley 5.2
London Lewisham
London A3 Roadside 6.2
London Teddington
98th Percentile
Mean of 1-H Daily A 98th
Concentration A Mean3 Maxb* Percentile 24-H Avg
ppb % ppb % IQR
58.34
31.06
32.98
16.84
42.69
18.47
37
21.2
32.34
25.74
24.4
25.74
24.73
26.03
35.33
13.09
87.84 163.52
70.55
95.9 80.69
53.7
131.15 97.68
64.5
74.51 116.28
71.77
25.65 80.05
91.2
-5.21 66.08
91.2
-4.98 80.36
87.44
169.95 93.01
53.47
131.77 28.62
13.1
50.25 11.13
10.66
51.44 12.1
11.86
62.01 15.66
12.65
-12.23 15.18
14.29
-27.54 12.02
14.29
-8.1 12.36
12.59
73.94 14.66
11.15
2004-2006.
1-H Max
IQR
45.49
15.96
13.3
16.49
21.01
15.96
26.87
17.02
19.15
19.68
16.49
19.68
17.56
17.02
19.15
19.21
24-H Avg.
Correlation
With Urban
Background
Monitor
95% Cl
0.24
(0.18,
0.86
(0.84,
0.63
(0.59,
0.78
(0.75,
0.81
(0.79,
0.80
(0.78,
0.70
(0.67,
0.64
(0.77,
0.29)
0.88)
0.66)
0.80)
0.83)
0.82)
0.73)
0.81)
Cl = confidence interval; IQR = interquartile range; Bkg = background.
aRoadside vs urban background comparison.
b3-year average.
Source: National Center for Environmental Assessment Analysis of European Air Quality Database Data from 2004-2006.
January 2015
2-64
DRAFT: Do Not Cite or Quote
-------
1 While NC>2 measurements are more widely used than NOx for exposure estimates and
2 epidemiologic studies, NC>2 accounts for only a fraction of NOx near roads with heavy
3 traffic. For example, Clements et al. (2009) measured concentrations of NO, NO2, and
4 NOx, 5 m downwind from a state road in Austin, TX, and observed NOx concentrations
5 of approximately 40-50 ppb, NO concentrations of approximately 15-40 ppb, and NO2
6 concentrations of approximately 5-15 ppb under downwind conditions. NO2 accounted
7 for 10-3 8% of the NOx.
8 It follows that NO is often a greater contributor to NOx near roads. Baldauf et al. (2008a)
9 presented time-series of pollutants measured 5 m from 1-40 in Raleigh, NC, and reported
10 that NO concentrations reached near 250 ppb between 8:00 a.m. and 9:00 a.m., with
11 minimum NO concentrations around 50 ppb during that time period. The predominance
12 of NO (rather than NO2) in the near-road environment contrasts with nationwide annual
13 average concentrations in Table 2-4. for which annual average NO2 accounts for more
14 than 60% of annual average NOx.
15 Wind speed and atmospheric stability also impact roadway NOx concentrations. Peak
16 roadway concentrations are often observed during presunrise hours when winds are weak
17 and atmospheric inversions are present (Gordon etal.. 2012; Durantetal.. 2010; Hu
18 etal.. 2009). During these presunrise hours, the NOx concentrations exhibit a more
19 gradual decay from the roadway than after sunrise. Hu et al. (2009) observed this effect
20 during a near-road field campaign in Santa Monica, CA. They observed elevated NO
21 concentrations (90-160 ppb) as far as 1,200 m downwind of the roadway during
22 pre-sunrise hours, which is much larger than the expected spatial extent of NO
23 (100-300 m) (Karner etal., 2010; Zhou and Levy. 2007). NOx concentration gradients
24 continue to change throughout the day as atmospheric stability evolves. After sunrise,
25 near-road NOx concentrations drop as vertical mixing increases (Gordon etal.. 2012;
26 Durant et al.. 2010) until concentrations reach a minimum during the late afternoon
27 (Gordon et al.. 2012). In some studies, no clear gradient is observed in NOx
28 concentrations (or other traffic-related species) during mid-morning or early evening
29 hours (Gordon etal.. 2012; DurantetaL. 2010). However, the exact response of the
30 horizontal concentration gradient to changes in boundary layer height is unresolved to
31 some extent.
32 Dispersion of NOx in the near-road environment is influenced by several factors:
33 atmospheric turbulence, vehicle-induced turbulence, and roadway-induced turbulence
34 (Baldauf etal.. 2009; Wang and Zhang. 2009). Atmospheric turbulence occurs as a result
35 of meteorological factors within the urban boundary layer. Vehicle-induced turbulence
36 results from the air disturbances caused by the direction and speed of vehicle motion.
37 Roadway-induced turbulence happens when wind-driven air masses undergo separation
January 2015 2-65 DRAFT: Do Not Cite or Quote
-------
1 following impact with a roadway structure in the built environment. These sources of
2 turbulence interact with each other to create complex, unique dispersion profiles at a
3 given road segment to influence NOx concentrations. This discussion addresses the
4 physical factors influencing dispersion of NOx.
5 Several atmospheric conditions affect regional or urban airflow profiles and potentially
6 can impact the dispersion profile of NOx even in the absence of adjacent buildings,
7 roadway structures, or traffic-related turbulence. In urban areas, effects of the built
8 environment can be seen at regional-, urban-, neighborhood-, and street-level scales
9 (Fernando. 2010; Britter and Hanna. 2003). Roughness created by upstream buildings
10 contributes to local turbulence levels, even in the absence of adjacent buildings. Land
11 forms such as slopes and valleys can also affect the atmospheric turbulence level because
12 they interact with atmospheric stability conditions to restrict air movement. Finn et al.
13 (2010) observed that tracer gas concentration increased with increasing atmospheric
14 stability. This finding is consistent with results with other studies (Gordon et al.. 2012;
15 Durantetal.. 2010; Hu et al.. 2009) that observed the highest concentrations of NO, NO2,
16 and NOx before sunrise when traffic levels and atmospheric stability are high. Hu et al.
17 (2009) also argued that atmospheric stability potentially extends the decay profile of
18 near-roadway pollutants. Additionally, the presence of slopes and valleys can cause spots
19 where airflow converges or diverges (Fernando. 2010). Heat flux can be sizeable in urban
20 areas where the "heat island" effect from roadways and buildings can raise local
21 temperatures by several degrees (Britter and Hanna. 2003); heat flux potentially
22 contributes to convection near roadways and other structures in the built environment.
23 Underscoring the dominant role of local turbulence on dispersion patterns, Venkatram
24 et al. (2007) measured meteorological factors potentially affecting NO concentrations
25 near a road segment in Raleigh, NC and found that among meteorological variables
26 vertical velocity fluctuations had the largest effect on NO concentration.
27 Vehicle motion creating high levels of turbulence on and near roads can contribute to the
28 dispersion of traffic-related air pollution in the vicinity of a roadway (Baldauf et al..
29 2008a). An early description of this was provided by Sedefian et al. (1981) for the
30 General Motors experiments, in which groups of vehicles were driven along a test track
31 while towers with mounted anemometers measured mean and fluctuating velocities. It
32 was observed that vehicle-induced turbulence dissipates slowly under low mean wind
33 conditions and vice versa. Vehicle-induced turbulence was found in that study to
34 contribute to vertical dispersion of emitted pollutants. Computational fluid dynamics
35 (CFD) simulations by Wang and Zhang (2009) also found that vehicle-induced
36 turbulence contributed to vertical dispersion. Rao et al. (2002) observed large
37 measurements of turbulence kinetic energy in the wake of a vehicle outfitted with a trailer
38 carrying sonic anemometers driving along a runway. Sedefian et al. (1981) found that
January 2015 2-66 DRAFT: Do Not Cite or Quote
-------
1 advection of vehicle-induced turbulence away from the roadway was related to the speed
2 and direction of mean winds, di Sabatino et al. (2003) showed that vehicle-induced
3 turbulence is related to traffic levels. In light traffic, the wake behind a vehicle is isolated,
4 but for increasing traffic, the wakes interact and turbulence is a function of the number of
5 vehicles and vehicle length scale. At congested traffic levels, the vehicle-induced
6 turbulence becomes independent of the number of vehicles. For street canyon simulations
7 and measurements, Kastner-Klein et al. (2003) observed that predictions of tracer
8 concentrations were overestimated when vehicle-induced turbulence was not considered;
9 this implies additional dispersion related to vehicle-induced turbulence. Traffic
10 directionality was investigated by He and Dhaniyala (2011) and Kastner-Klein et al.
11 (2001). He and Dhaniyala (2011) observed that turbulent kinetic energy from two-way
12 traffic was roughly 20% higher than for one-way traffic, and they found that the turbulent
13 kinetic energy increased with decreasing distance between the traffic lanes. Kastner-
14 Klein etal. (2001) observed that two-way traffic suppresses the mean flow of
15 vehicle-induced air motion along a street canyon, whereas one-way traffic produces a
16 piston-like effect [note that the Kastner-Klein et al. (2001) study was for the geometrical
17 case of a street canyon]. Substantially higher turbulence levels were produced with
18 two-way traffic compared with one-way traffic for the Kastner-Klein et al. (2001) study
19 as well.
20 The presence of near-road structures results in recirculating airflow regions that may trap
21 air pollutants on one side and disperse them on another side, depending on wind
22 conditions (Baldauf et al.. 2008b). Finn etal. (2010) simulated transport from a roadway
23 using a point source tracer gas with barrier and open terrain conditions. With airflow
24 from the simulated roadway and high atmospheric stability, high concentrations were
25 trapped in the roadway region with a negligible tracer gas in the wake downstream of the
26 barrier with considerable lateral and vertical plume dispersion. For open terrain, transport
27 of the tracer was characterized by a narrow plume. Hagler etal. (2011) used CFD to
28 model airflow and concentrations around barriers of different heights and similarly found
29 reductions in inert tracer concentration downwind of the barrier compared with the open
30 terrain case with trapping of air pollutants upstream of the barrier. With the barrier in
31 place, downwind tracer concentrations were observed at elevations of twice the barrier
32 height. Mean airflow vectors also illustrate a wind disturbance at elevations of twice the
33 barrier height. Even for the open terrain case, vertical dispersion occurs. In additional
34 simulations involving a service readjust downstream of the barrier, Hagler etal. (2011)
35 observed entrainment of tracer in the wake downstream of the barrier. Tokairin and
36 Kitada (2005) used CFD to investigate the effect of porous fences on contaminant
37 transport near roads and observed tracer gas retention and airflow recirculation when the
38 fences were designed with less than 40-50% porosity. Heist et al. (2009b) investigated
39 the effect of geometry of road cuts and noise barriers in wind tunnel tracer gas
January 2015 2-67 DRAFT: Do Not Cite or Quote
-------
1 experiments. They observed that elevated roadways, depressed roadways, and noise
2 barriers all resulted in lower downwind concentrations compared with the open terrain
3 case with elevated roadways producing the least reduction in concentration. As in Hagler
4 et al. (2011). Heist et al. (2009b) observed measurable concentrations at elevations that
5 resulted from Gaussian dispersion for all geometries of the road cut or barrier, but
6 vertical dispersion was enhanced or dampened depending on the specific geometry.
7 Similarly, for wind tunnel simulations of a single tower above a matrix of street canyons,
8 the tower was shown to induce both airflow and tracer concentration along the leeward
9 edge of the building to a height exceeding the tower height (Brixey et al., 2009; Heist
10 et al.. 2009a).
11 For the special case of street canyons, retention time for traffic-based pollution increases
12 on the roadway with increasing building height-to-road width ratio because recirculating
13 airflow forms closed streamlines within the canyon (Li et al.. 2005; Liu et al.. 2005). For
14 wind tunnel simulations of tracer emission at street level with and without traffic,
15 Kastner-Klein et al. (2001) observed measurable tracer concentrations near the top of the
16 street canyon but with some dispersion from maximum tracer levels at the canyon floor.
17 Dilution of NOx concentrations through these recirculating air structures leads to a steep
18 decrease in concentration with increasing distance from the ground (Lee et al.. 2012). For
19 low-aspect-ratio street canyons, secondary recirculating structures can arise; while
20 contaminant retention still occurs in this case, ventilation occurs more readily than for the
21 high-aspect-ratio case (Simoens and Wallace. 2008; Simoens et al., 2007). Cheng et al.
22 (2008) used CFD to evaluate factors leading to contaminant retention in street canyons
23 and observed that the exchange rate for air and a tracer gas was driven by the turbulent
24 component of airflow at the roof-level interface of the street canyon. Subsequent
25 simulations showed that exchange rate was also aided by unstable atmospheric conditions
26 (Cheng etal.. 2009). CFD simulations by Gu etal. (2010) of transport within a street
27 canyon with and without vegetation suggested that the recirculating flow is dampened by
28 the presence of vegetation.
2.5.4 Seasonal, Weekday/Weekend, and Diurnal Trends
29 Month-to-month variability in 24-h avg NC>2 concentrations was described in the 2008
30 ISA for Oxides of Nitrogen (U.S. EPA. 2008a). Strong seasonal variability in NO2 was
31 reported, with higher concentrations in winter and lower concentrations in summer.
32 Monthly maxima varied regionally. Day-to-day variability in NC>2 concentration was
33 generally larger during the winter.
January 2015 2-68 DRAFT: Do Not Cite or Quote
-------
Recent data presented in Table 2-3 continue to show similar seasonal trends for average
seasonal NO2 concentrations across 3 years. Mean and 99th percentile concentrations are
highest in the first and fourth quarters. Concentration patterns of NO and NC>2 are
affected strongly by emissions and meteorology, as concentrations peak during early
morning hours and in winter when PEL heights are lowest (Figure 2-19). NC>2 exhibits
flatter profiles relative to NO as secondary formation processes influence concentration
patterns.
70
- January 130890002 NO
•January 130890002 NCb
-July 130890002 HO
•Jutyl 30890002 N02
12345678 91011121*1415161718192021222324
Hour of the Day
Source: National Center for Environmental Assessment analysis of Air Quality System Network Data.
Figure 2-19 January and July hourly profiles of nitric oxide (NO) and nitrogen
dioxide (NOa) (ppb) for Atlanta, GA (site in Atlanta with maximum
1-hour NO2 concentrations).
10
11
12
13
14
15
16
17
18
Figure 2-19 shows a typical diurnal cycle for a nonnear-road site for NO and NO2. As
described in the 2008 ISA for Oxides of Nitrogen (U.S. EPA. 2008a). the NO2
concentration typically exhibits a daily maximum during morning rush hour, although the
concentration maximum can also occur at other times of day. This pattern is shown for
Atlanta, GA, in Figure 2-19. but it is also typical for other urban sites. Although the
concentration trends shown in Figure 2-19 are for a nonnear-road monitoring site, they
are similar to trends observed for the Las Vegas and Detroit near-road concentration
patterns in Figure 2-17. NO levels well above zero at night imply that Os has been
completely titrated.
Differences between weekdays and weekends are shown for the same monitor in
Figure 2-20. Typically, weekday concentrations of NOx, particularly NO, exceed
January 2015
2-69
DRAFT: Do Not Cite or Quote
-------
1
2
o
6
4
5
6
7
8
9
10
11
12
13
14
weekend concentrations, and diurnal cycles are more compressed on weekends. The
weekend effect for NO was first observed by Cleveland et al. (1974) and is a general
characteristic of urban NO and NOx concentrations observed in many locations (Tonse
et al.. 2008; Pun et al.. 2003; Marr and Harley. 2002). Differences between weekdays and
weekends were also noted in the 2008 ISA for Oxides of Nitrogen (U.S. EPA. 2008a).
with more pronounced differences at sites more influenced by traffic. Both empirical
observations and modeling simulations of weekly cycles of NOx based on summer
satellite column data converted to concentrations using a chemistry transport model of the
vertical NO2 distribution (see Section 2.4.5) also indicate higher concentrations on
weekdays than on weekends regardless of land coverage, for urban, forest, and other
regions (Choi etal.. 2012). In southern California, NOx concentrations were an average
of 46% lower in ground-based measurements, and 34% lower in airborne measurements
(Pollack et al., 2012). In Atlanta, NOx concentrations were 24% higher on weekdays than
on weekends (Pachon et al.. 2012).
5 10
•1J0890Q02 NO Weekday
•lJ0890QG2NOWeekend
-1J0890002N02 Weekday
• 1}0890002N02 Weekend
r—r—i—i—i—i—r—r—i—i—I— —I—r—:—i—r
12545678 9101112131415161718192021222324
Source: National Center for Environmental Assessment analysis of Air Quality System Network Data.
Figure 2-20 Weekend/weekday hourly profiles of nitric oxide (NO) and
nitrogen dioxide (NO2) (ppb) for Atlanta, GA (site in Atlanta with
maximum NO2 concentrations).
2.5.5 Multiyear Trends in Ambient Measurements of Oxides of Nitrogen
15 The annual average NO2 concentration across the U.S. based on concentrations from the
16 national air quality monitoring network decreased by 49% from 1990 to 2012, as shown
17 in Figure 2-21. The blue band shows the distribution of air pollution levels among the
January 2015
2-70
DRAFT: Do Not Cite or Quote
-------
4
5
6
7
trend sites, displaying the middle 80%. The white line represents the average among all
the trend sites. Ninety percent of sites have NO2 concentrations below the top line, while
10% of sites have concentrations below the bottom line.
Information on trends on a regional basis and at individual, local air monitoring sites can
be found at http://www.epa.gov/air/airtrends/nitrogen.html: National Trends in Nitrogen
Dioxide Levels. The steady decline in NO2 concentrations over the years can be
attributed mainly to reductions in emissions from mobile and stationary sources (see
Figure 2-2).
70
-60
Q.
g 40
ra
c 30
QJ
U
10
NO2 Air Quality, 1990-2012
(Annual Arithmetic Average)
National Trend Based on 135 Sites
National Standard
9999999999
9999999999
0123456789
0000000
0000000
0123456
1990 to 2012 : 48% decrease in national average
11111111112222222222222
000000
000111
78901 2
Source: http://www.epa.gov/airtrends/nitrogen.html.
Figure 2-21 U.S. national annual average ambient nitrogen dioxide
concentration trends, 1990-2012.
January 2015
2-71
DRAFT: Do Not Cite or Quote
-------
1 In Atlanta, GA, NOx concentrations decreased from 1999 to 2001, increased during 2002
2 and 2003, and decreased again until 2007. The decrease from 1999 to 2001 was attributed
3 to the implementation of EPA's acid rain program, and the decrease from 2002 to 2007 to
4 decreases in on-road NOx emissions (Pachon et al.. 2012).
2.5.6 Background Concentrations
5 In the context of a review of the NAAQS, EPA generally defines "background
6 concentrations" in a way that distinguishes among concentrations that result from
7 precursor emissions that are relatively less controllable from those that are relatively
8 more controllable through U.S. policies or through international agreements. The most
9 commonly used form in the past and in this document is North American Background
10 (NAB), which refers to simulated NC>2 concentrations that would exist in the absence of
11 anthropogenic emissions from the U.S., Canada, and Mexico. This definition of
12 background includes contributions resulting from emissions from natural sources
13 (e-g-, soils, wildfires, lightning) around the world. Other definitions can also be used. For
14 example, in the 2013 ISA for Ozone and Related Photochemical Oxidants (U.S. EPA,
15 2013b). a U.S. background, which includes emissions from Canada and Mexico in
16 addition to those in the definition of a North American background, and a natural
17 background, which includes only emissions from natural sources globally, were used.
18 Background is used to inform policy considerations regarding the current or potential
19 alternative standards.
20 As can be seen from Figure 2-13. maximum seasonally averaged concentrations of NO2
21 occur along the Northeast Corridor, the Ohio River Valley, and in the Los Angeles basin.
22 While NO2 concentrations are often above 5 ppb, NAB is less than 300 ppt over most of
23 the continental U.S., and less than 100 ppt in the eastern U.S., as shown in Figure 2-4
24 through Figure 2-18 in the 2008 ISA for Oxides of Nitrogen (U.S. EPA. 2008a). The
25 distribution of background concentrations in the 2008 ISA was shown to largely reflect
26 the distribution of soil NO emissions and lightning, with some local increases due to
27 biomass burning, mainly in the western U.S. In the northeastern U.S., where present-day
28 NO2 concentrations are highest, NAB contributes <1% to the total.
29 The only updates to the results given in the 2008 ISA (U.S. EPA. 2008a) are the
30 global-scale model calculations of Lin et al. (2012). In addition to U.S. and other North
31 American sources, various NOy species from sources outside North America have long
32 enough residence times in the atmosphere enabling them to be transported to the U.S.
33 (Lin et al.. 2012). As noted in the 2013 ISA for Ozone and Related Photochemical
34 Oxidants (U.S. EPA. 2013b). spring is the dominant season for effects of intercontinental
January 2015 2-72 DRAFT: Do Not Cite or Quote
-------
1 transport of pollution to be detected in the U.S. Lin etal. (2012) calculated that transport
2 of NOx from other continents contributes less than 10 ppt to the regional background in
3 the western U.S., but concentrations of PAN could range from 50 to 80 ppt.
4 The annual median NC>2 concentration of ~8 ppb reported by the SLAMS monitoring
5 network is well below the level of the current annual NAAQS (0.053 ppm) and the hourly
6 NAAQS (100 ppb). Background concentrations of NC>2 are much lower than average
7 ambient concentrations and are typically less than 0.1 ppb over most of the U.S., with the
8 highest values found in agricultural areas. All these values indicate that background
9 concentrations of NC>2 are well beneath the level of the current NO2 NAAQS.
2.6 Conclusions
10 A large number of oxidized nitrogen species occur in the atmosphere. They are emitted to
11 the atmosphere mainly as NO, which interconverts with NC>2. Thus, NO and NO2 are
12 often combined into their own group and referred to as NOx. NOx plays an important role
13 in the formation of atmospheric Os and particulate matter. The conversion of NOx into
14 other species, such as PAN, HNOs, or particulate nitrate typically takes place on much
15 longer time scales than does interconversions between NO and NO2. As a result, near
16 sources, such as in heavily populated areas or busy roads with heavy traffic, oxides of
17 nitrogen are mainly present as NOx. However, in remote areas downwind of major
18 sources, more oxidized species account for a greater fraction of oxides of nitrogen.
19 NOx emissions in the U.S. have been roughly cut in half since 1990. In most of the
20 largest urban areas in the U.S., motor vehicle traffic accounts for 40-67% of emissions
21 and Off-Highway diesel and gasoline engines contribute an additional 20-30%.
22 Off-Highway vehicles and engines, electric power generation, other stationary fuel
23 combustion, industrial and agricultural process, and fires are all important NOx sources
24 on a national scale, with highway vehicles, Off-Highway vehicles and engines, and
25 stationary fuel combustion especially important in urban areas. Urban stationary fuel
26 combustion emissions account for a greater fraction of NOx emissions in colder climates.
27 In some cities, specific industrial sources like oil and gas production, petroleum refining,
28 or cement manufacturing account for a greater fraction of NOx emissions locally than
29 they do nationally. However, traffic emissions are generally responsible for the greatest
30 share of NOx in the U.S., especially in populated areas.
31 NO and NO2 are most commonly measured by a Federal Reference Method based on
32 chemiluminescence of NO induced by its reaction with Os. NO2 is measured by first
33 reducing it to NO, and then measuring the chemiluminescence of NO. Recent
34 advancements in NO2 measurements include improved methods of conversion of NO2 to
January 2015 2-73 DRAFT: Do Not Cite or Quote
-------
1 NO, development of optical methods to measure NO2 directly, and development of
2 satellite measurement methods. NO2 is measured at hundreds of monitors in several
3 national monitoring networks. In 2014 the first phase of the new near-road monitoring
4 network was initiated in recognition that millions of people live within a few hundred
5 meters of a major roadway, and that concentrations of NO2 typically decrease with
6 increasing distance from a major road.
7 If annual average NO2 concentrations for individual monitoring sites are averaged over
8 all monitoring sites in the U.S., the overall average is about 15 ppb. Similarly, the
9 average daily 1-hour maximum NO2 concentration over all U.S. monitoring sites is about
10 30 ppb. Average NO2 concentrations are usually somewhat higher in winter than in
11 summer. Concentrations are highest in populated urban areas where sources are
12 dominated by vehicle emissions. Within urban areas there can be a high degree of spatial
13 variability, although good intra-urban agreement has also been frequently observed.
14 Concentrations within urban areas are usually highest near major roadways and major
15 stationary sources. Near roadways, an NO2 concentration gradient is often observed,
16 especially in the summer and during daylight hours. NO2 concentrations are typically up
17 to 20 ppb higher within 20 m of a major road than at a distance a few hundred meters
18 from the road, and the spatial extent of elevated concentration typically ranges from 200
19 to 500 m. Preliminary results from EPA's new near-road monitoring network indicate
20 that seasonal average NO2 concentrations are usually higher near roads with heavy traffic
21 than in other locations in the same city.
22 Much of the most recent research on atmospheric NO2 and NOx has focused on its role as
23 a traffic pollutant and its spatial variability, especially in proximity to major roads.
24 Because traffic is the largest source of NOx in the U.S., especially in populated areas, this
25 research is highly relevant to human exposure, and the results described in this chapter
26 provide a useful context for NO2 exposure and health assessment.
January 2015 2-74 DRAFT: Do Not Cite or Quote
-------
References for Chapter 2
Arnold. JR: Hartsell. BE: Luke. WT: Ullah. SMR: Dasgupta. PK: Huev. LG: Tate. P. (2007). Field test of
four methods for gas-phase ambient nitric acid. Atmos Environ 41: 4210-4226.
http://dx.doi.0rg/10.1016/j.atmosenv.2006.07.058
Baldauf. R: Thoma. E: Hays. M: Shores. R: Kinsev. J: Gullett. B: Kimbrough. S: Isakov. V: Long. T:
Snow. R: Khlystov. A: Weinstein. J: Chen. FL: Seila. R: Olson. D: Gilmour. I: Cho. SH: Watkins. N:
Rowley. P: Bang. J. (2008a). Traffic and meteorological impacts on near-road air quality: Summary of
methods and trends from the Raleigh near-road study. J Air Waste Manag Assoc 58: 865-878.
http://dx.doi.Org/10.3155/1047-3289.58.7.865
Baldauf. R: Thoma. E: Khlystov. A: Isakov. V: Bowker. G: Long. T: Snow. R. (2008b). Impacts of noise
barriers on near-road air quality. Atmos Environ 42: 7502-7507.
http://dx.doi.0rg/10.1016/j.atmosenv.2008.05.051
Baldauf. R: Watkins. N: Heist. D: Bailey. C: Rowley. P: Shores. R. (2009). Near-road air quality
monitoring: Factors affecting network design and interpretation of data. Air Qual Atmos Health 2: 1-9.
http://dx.doi.org/10.1007/s 11869-009-0028-0
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/j.atmosenv.2007.09.042
Bell. S: Ashenden. TW. (1997). Spatial and temporal variation in nitrogen dioxide pollution adjacent to
rural roads. Water Air Soil Pollut 95: 87-98.
Signal KL: Ashmore. MR: Headlev. AD: Stewart. K: Weigert K. (2007). Ecological impacts of air
pollution from road transport on local vegetation. Appl Geochem 22: 1265-1271.
http://dx.doi.0rg/10.1016/i.apgeochem.2007.03.017
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
Britter. RE: Hanna. SR. (2003). Flow and dispersion in urban areas. Annu Rev Fluid Mech 35: 469-496.
http://dx.doi.org/10.1146/annurev.fluid.35.101101.161147
Brixev. LA: Heist. DK: Richmond-Bryant. J: Bowker. GE: Perry. SG: Wiener. RW (2009). The effect of a
tall tower on flow and dispersion through a model urban neighborhood: Part 2: Pollutant dispersion. J
Environ Monit 11: 2171-2179. http://dx.doi.org/10.1039/B907137g
Brown. SS: Dube. WP: Karamchandani. P: Yarwood. G: Peischl. J: Ryerson. TB: Neuman. JA: Nowak.
JB: Hollowav. JS: Washenfelder. RA: Brock. CA: Frost GJ: Trainer. M: Parrish. DP: Fehsenfeld. FC:
Ravishankara. AR. (2012). Effects of NO x control and plume mixing on nighttime chemical
processing of plumes from coal-fired power plants. J Geophys Res 117: D07304.
http://dx.doi.org/10.1029/2011JD016954
Brown. SS: Stutz. J. (2012). Nighttime radical observations and chemistry. Chem Soc Rev 41: 6405-6447.
http://dx.doi.org/10.1039/c2cs35181a
Bucsela. EJ: Krotkov. NA: Celarier. EA: Lamsal LN: Swartz. WH: Bhartia. PK: Boersma. KF: Veefkind.
JP: Gleason. JF: Pickering. KE. (2013). Anew stratospheric and tropospheric NO2 retrieval algorithm
for nadir-viewing satellite instruments: Applications to OMI. Atmospheric Measurement Techniques
Discussions 6: 1361-1407. http://dx.doi.org/10.5194/amtd-6-1361-2013
January 2015 2-75 DRAFT: Do Not Cite or Quote
-------
Burling. IR: Yokelson. RJ: Griffith. DWT: Johnson. TJ: Veres. P: Roberts. JM: Warneke. C: Urbanski. SP:
Reardon. J: Weise. PR: Hao. WM: De Gouw. J. (2010). Laboratory measurements of trace gas
emissions from biomass burning of fuel types from the southeastern and southwestern United States.
Atmos Chem Phys 10: 11115-11130. http://dx.doi.org/10.5194/acp-10-11115-2010
Cahill. TA: Cahill. TM: Barnes. DE: Spada. NJ: Miller. R. (2011). Inorganic and organic aerosols
downwind of California's Roseville railyard. Aerosol Sci Technol 45: 1049-1059.
http://dx.doi.org/10.1080/02786826.2011.580796
Cape. JN: Tang. YS: van Dijk. N: Love. L: Sutton. MA: Palmer. SCF. (2004). Concentrations of ammonia
and nitrogen dioxide at roadside verges, and their contribution to nitrogen deposition. Environ Pollut
132: 469-478. http://dx.doi.Org/10.1016/i.envpol.2004.05.009
Carslaw. DC: Beevers. SD: Ropkins. K: Bell MC. (2006). Detecting and quantifying aircraft and other
on-airport contributions to ambient nitrogen oxides in the vicinity of a large international airport.
Atmos Environ 40: 5424-5434. http://dx.doi.Org/10.1016/j.atmosenv.2006.04.062
Chanev. AM: Cryer. DJ: Nicholl EJ: Seakins. PW. (2011). NO and NO(2) interconversion downwind of
two different line sources in suburban environments. Atmos Environ 45: 5863-5871.
http://dx.doi.0rg/10.1016/j.atmosenv.2011.06.070
Cheng. WC: Liu. CH: Leung. DYC. (2008). Computational formulation for the evaluation of street
canyon ventilation and pollutant removal performance. Atmos Environ 42: 9041-9051.
http://dx.doi.0rg/10.1016/j.atmosenv.2008.09.045
Cheng. WC: Liu. CH: Leung. DYC. (2009). On the correlation of air and pollutant exchange for street
canyons in combined wind-buoyancy-driven flow. Atmos Environ 43: 3682-3690.
http://dx.doi.0rg/10.1016/j.atmosenv.2009.04.054
Choi Y: Kim. H: Tong. D: Lee. P. (2012). Summertime weekly cycles of observed and modeled NOx and
O3 concentrations as a function of satellite-derived ozone production sensitivity and land use types
over the Continental United States. Atmos Chem Phys 12: 6291-6307. http://dx.doi.org/10.5194/acp-
12-6291-2012
Clapp. L: Jenkin. M. (2001). Analysis of the relationship between ambient levels of O3, NO2 and NO as a
function of NOx in the UK. Atmos Environ 35: 6391-6405. http://dx.doi.org/10.1016/S1352-
2310(01)00378-8
Clements. AL: Jia. Y: Denbleyker. A: Mcdonald-Buller. E: Eraser. MP: Allen. DT: Collins. PR: Michel.
E: Pudota. J: Sullivan. D: Zhu. Y. (2009). Air pollutant concentrations near three Texas roadways, part
II: Chemical characterization and transformation of pollutants. Atmos Environ 43: 4523-4534.
http://dx.doi.0rg/10.1016/j.atmosenv.2009.06.044
Cleveland. WS: Graedel TE: Kleiner. B: Warner. JL. (1974). Sunday and workday variations in
photochemical air pollutants in New Jersey and New York. Science 186: 1037-1038.
http://dx.doi.org/10.1126/science.186.4168.1037
Corbett JJ: Fischbeck. PS: Pandis. SN. (1999). Global nitrogen and sulfur inventories for oceangoing
ships. J Geophys Res Atmos 104: 3457-3470. http://dx.doi.org/10.1029/1998JD100040
Croslev. DR. (1996). NOy blue ribbon panel. J Geophys Res 101: 2049-2052.
http://dx.doi.org/10.1029/95JD02276
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: Warneke. C: Parrish. DP: Hollowav. JS: Trainer. M: Fehsenfield. FC. (2003). Emission
sources and ocean uptake of acetonitrile (CH 3 CN) in the atmosphere. J Geophys Res 108: 4329.
http://dx.doi.org/10.1029/2002JD002897
January 2015 2-76 DRAFT: Do Not Cite or Quote
-------
Penman. KL: Brasseur. G: Chidthaisong. A: Ciais. P: Cox. PM: Dickinson. RE: Hauglustaine. D: Heinze.
C: Holland. E: Jacob. D: Lohmann. U: Ramachandran. S: da Silva Dias. PL: Wofsv. SC: Zhang. X.
(2007). Couplings between changes in the climate system and biogeochemistry. In Climate change
2007: The physical science basis (pp. 499-587). New York, NY: Cambridge University Press.
http://www.ipcc.ch/publications and data/publications ipcc fourth assessment report wgl report t
he_physical science basis.htm
di Sabatino. S: Kastner-Klein. P: Berkowicz. R: Britter. RE: Fedorovich. (2003). The modelling of
turbulence from traffic in urban dispersion models Part I: Theoretical considerations. Environ Fluid
Mech 3: 129-143.
Dunlea. EJ: Herndon. SC: Nelson. DP: Volkamer. RM: San Martini F: Sheehy. PM: Zahniser. MS:
Shorter. JH: Wormhoudt. JC: Lamb. BK: Allwine. EJ: Gaffnev. JS: Marlev. NA: Grutter. M: Marquez.
C: Blanco. S: Cardenas. B: Retama. A: Ramon Villegas. CR: Kolb. CE: Molina. LT: Molina. MJ.
(2007). Evaluation of nitrogen dioxide chemiluminescence monitors in a polluted urban environment.
Atmos Chem Phys 7: 2691-2704.
Durant JL: Ash. CA: Wood. EC: Herndon. SC: Jayne. JT: Knighton. WB: Canagaratna. MR: Trull JB:
Brugge. D: Zamore. W: Kolb. CE. (2010). Short-term variation in near-highway air pollutant gradients
on a winter morning. Atmos Chem Phys 10: 8341-8352. http://dx.doi.org/10.5194/acp-10-8341-2010
During. I: Bachlin. W: Ketzel M: Baum. A: Friedrich. U: Wurzler. S. (2011). A new simplified NO/NO2
conversion model under consideration of direct NO2-emissions. Meteorologische Zeitschrift 20: 67-
73. http://dx.doi.org/10.1127/0941-2948/2011/0491
EIONET (European Environment Information and Observation Netork). (2014). European topic centre on
air pollution and climate change mitigation: AirBase- the European air quality database. Available
online at http://acm.eionet.europa.eu/databases/airbase
Fehsenfeld. FC: Dickerson. RR: Hubler. G: Luke. WT: Nunnermacker. LJ: Williams. EJ: Roberts. JM:
Calvert. JG: Curran. CM: Delanv. AC: Eubank. CS: Fahev. DW: Fried. A: Gandrud. BW: Langford.
AO: Murphy. PC: Norton. RB: Pickering. KE: Ridley. BA. (1987). A ground-based intercomparison of
NO, NOx, and NOy measurement techniques. J Geophys Res 92: 14,710-714,722.
http://dx.doi.org/10.1029/JD092iD12pl4710
Fernando. HJS. (2010). Fluid dynamics of urban atmospheres in complex terrain. Annu Rev Fluid Mech
42: 365-389. http://dx.doi.org/10.1146/annurev-fluid-121108-145459
Finn. D: Clawson. KL: Carter. RG: Rich. JD: Eckman. RM: Perry. SG: Isakov. V: Heist. DK. (2010).
Tracer studies to characterize the effects of roadside noise barriers on near-road pollutant dispersion
under varying atmospheric stability conditions. Atmos Environ 44: 204-214.
http://dx.doi.0rg/10.1016/j.atmosenv.2009.10.012
Fiore. AM: Horowitz. LW: Purves. DW: Lew. H: 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: Jacob. DJ: Yantosca. RM: Sulprizio. MP: Millet. DB: Mao. J: Paulot. F: Singh. HB: Roiger.
A: Ries. L: Talbot RW: Dzepina. K: Pandev Deolal S. (2014). Atmospheric peroxyacetyl nitrate
(PAN): a global budget and source attribution. Atmos Chem Phys 14: 2679-2698.
http://dx.doi.org/10.5194/acp-14-2679-2014
Fuchs. H: Dube. WP: Lerner. BM: Wagner. NL: Williams. EJ: Brown. SS. (2009). A sensitive and
versatile detector for atmospheric NO2 and NOx based on blue diode laser cavity ring-down
spectroscopy. Environ Sci Technol 43: 7831-7836. http://dx.doi.org/10.1021/es902067h
Gilbert. NL: Goldberg. MS: Brook. JR: Jerrett M. (2007). The influence of highway traffic on ambient
nitrogen dioxide concentrations beyond the immediate vicinity of highways. Atmos Environ 41: 2670-
2673. http://dx.doi.0rg/10.1016/j.atmosenv.2006.12.007
Gilbert. NL: Woodhouse. S: Stieb. DM: Brook. JR. (2003). Ambient nitrogen dioxide and distance from a
major highway. Sci Total Environ 312: 43-46. http://dx.doi.org/10.1016/80048-9697(03)00228-6
January 2015 2-77 DRAFT: Do Not Cite or Quote
-------
Gilbraith. N: Powers. SE. (2013). Residential demand response reduces air pollutant emissions on peak
electricity demand days in New York City. Energy Policy 59: 459-469.
http://dx.doi.0rg/10.1016/j.enpol.2013.03.056
Gonzales. M: Quails. C: Hudgens. E: Neas. L. (2005). Characterization of a spatial gradient of nitrogen
dioxide across a United States-Mexico border city during winter. Sci Total Environ 337: 163-173.
http://dx.doi.0rg/10.1016/j.scitotenv.2004.07.010
Gordon. M: Staebler. RM: Liggio. J: Li SM: Wentzell J: Lu. G: Lee. P: Brook. J. R. (2012). Measured
and modeled variation in pollutant concentration near roadways. Atmos Environ 57: 138-145.
http://dx.doi.0rg/10.1016/j.atmosenv.2012.04.022
Gu. Z: Zhang. Y: Lei. K. (2010). Large eddy simulation of flow in a street canyon with tree planting under
various atmospheric instability conditions. Science China Technological Sciences 53: 1928-1937.
http://dx.doi.org/10.1007/sll431-010-3243-x
Hagler. GSW: Tang. W: Freeman. MJ: Heist. DK: Perry. SG: Vette. AF. (2011). Model evaluation of
roadside barrier impact on near-road air pollution. Atmos Environ 45: 2522-2530.
http://dx.doi.0rg/10.1016/j.atmosenv.2011.02.030
Hameed. S: Pinto. JP: Stewart. RW. (1979). Sensitivity of the predicted CO-OH-CH4 perturbation to
tropospheric NOx concentrations. J Geophys Res Oceans 84: 763-768.
http://dx.doi.org/10.1029/JC084iC02pQ0763
Hansen. DA: Edgerton. ES: Hartsell BE: Jansen. JJ: Kandasamy. N: Hidv. GM: Blanchard. CL. (2003).
The southeastern aerorsol research and characterization study: Part I - overview. J Air Waste Manag
Assoc 53: 1460-1471. http://dx.doi.org/10.1080/10473289.2003.10466318
Harrison. MAJ: Barra. S: Borghesi D: Vione. D: Arsene. C: Olariu. RI. (2005). Nitrated phenols in the
atmosphere: a review [Review]. Atmos Environ 39: 231-248.
http://dx.doi.0rg/10.1016/j.atmosenv.2004.09.044
He. M: Dhaniyala. S. (2011). A dispersion model for traffic produced turbulence in a two-way traffic
scenario. Environmental Fluid Mechanics (Dordrecht, 2001) 11: 627-640.
http://dx.doi.org/10.1007/sl0652-011-9215-2
HEI (Health Effects Institute). (2010). Traffic-related air pollution: A critical review of the literature on
emissions, exposure, and health effects. (Special Report 17). Boston, MA: Health Effects Institute
(HEI). http://pubs.healtheffects.org/view.php?id=334
Heist. DK: Brixev. LA: Richmond-Bryant. J: Bowker. GE: Perry. SG: Wiener. RW. (2009a). The effect of
a tall tower on flow and dispersion through a model urban neighborhood: Part 1. Flow characteristics.
J Environ Monit 11: 2163-2170. http://dx.doi.org/10.1039/B907135k
Heist. DK: Perry. SG: Brixev. LA. (2009b). A wind tunnel study of the effect of roadway configurations
on the dispersion of traffic-related pollution. Atmos Environ 43: 5101-5111.
http://dx.doi.0rg/10.1016/j.atmosenv.2009.06.034
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
Hu. S: Fruin. S: Kozawa. K: Mara. S: Paulson. SE: Winer. AM. (2009). A wide area of air pollutant
impact downwind of a freeway during pre-sunrise hours. Atmos Environ 43: 2541-2549.
http://dx.doi.0rg/10.1016/j.atmosenv.2009.02.033
Hudman. RC: Russell. AR: Valin. LC: Cohen. RC. (2010). Interannual variability in soil nitric oxide
emissions over the United States as viewed from space. Atmos Chem Phys 10: 9943-9952.
http://dx.doi.org/10.5194/acp-10-9943-2010
Husar. RB: Patterson. DE: Husar. JD: Gillani. NV: Wilson. WE. Jr. (1978). Sulfur budget of a power plant
plume. Atmos Environ 12: 549-568. http://dx.doi.org/10.1016/0004-6981(78)90236-6
Jacob. DJ. (1999). Introduction to atmospheric chemistry. New Jersey: Princeton University Press.
January 2015 2-78 DRAFT: Do Not Cite or Quote
-------
Jacobson. MZ. (2002). Atmospheric pollution: History, science, and regulation. New York: Cambridge
University Press.
Jaegle. L: Steinberger. L: RV. M: Chance. K. (2005). Global partitioning of NOx sources using satellite
observations: Relative roles of fossil fuel combustion, biomass burning and soil emissions. Faraday
Discuss Discuss. 130: 407-423. http://dx.doi.org/10.1039/b502128f
Karner. AA: Eisinger. PS: 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
Kastner-Klein. P: Fedorovich. E: Ketzel. M: R. B: Britter. R. (2003). The modelling of turbulence from
traffic in urban dispersion models part II: Evaluation against laboratory and full-scale concentration
measurements in street canyons. Environ Fluid Mech 3: 145-172.
http://dx.doi.Org/10.1023/A:1022049224166
Kastner-Klein. P: Fedorovich. E: Rotach. MW. (2001). A wind tunnel study of organised and turbulent air
motions in urban street canyons. J Wind Eng Ind Aerod 89: 849-861. http://dx.doi.org/10.1016/S0167-
6105(01)00074-5
Kebabian. PL: Wood. EC: Herndon. SC: Freedman. A. (2008). A practical alternative to
chemiluminescence-based detection of nitrogen dioxide: Cavity attenuated phase shift spectroscopy.
Environ Sci Technol 42: 6040-6045. http://dx.doi.org/10.1021/es703204j
Kim. SW: Heckel A: McKeen. SA: Frost. GJ: Hsie. E. -Y: Trainer. MK: Richter. A: Burrows. JP:
Peckham. SE: Grell GA. (2006). Satellite-observed U.S. power plant NOx emission reductions and
their impact on air quality. Geophys Res Lett 33: L22812. http://dx.doi.org/10.1029/2006GL027749
Kim. SW: Mckeen. SA: Frost. GJ: Lee. SH: Trainer. M: Richter. A: Angevine. WM: Atlas. E: Bianco. L:
Boersma. KF: Brioude. J: Burrows. JP: de Gouw. J: Fried. A: Gleason. J: Hilboll A: Mellqvist J:
Peischl. J: Richter. D: Rivera. C: Ryerson. T: Te Lintel Hekkert. S: Walega. J: Warneke. C: Weibring.
P: Williams. E. (2011). Evaluations of NOx and highly reactive VOC emission inventories in Texas
and their implications for ozone plume simulations during the Texas Air Quality Study 2006. Atmos
Chem Phys 11: 11361-11386. http://dx.doi.org/10.5194/acp-ll-11361-2011
Kimbrough. ES: Baldauf. RW: Watkins. N. (2013). Seasonal and diurnal analysis of NO2 concentrations
from a long-duration study conducted in Las Vegas, Nevada. J Air Waste Manag Assoc 63: 934-942.
http://dx.doi.org/10.1080/10962247.2013.795919
Kleffmann. J: Tapia. GV: Bejan. I: Kurtenbach. R: Wiesen. P. (2013). NO2 measurement techniques:
Pitfalls and new developments. In I Barnes; KJ Rudzin'ski (Eds.), Disposal of dangerous chemicals in
urban areas and mega cities. New York, NY: Springer, http://dx.doi.org/10.1007/978-94-007-5034-0 2
Lamsal. LN: Martin. RV: Van Donkelaar. A: Celarier. EA: Bucsela. EJ: Boersma. KF: Dirksen. R: Luo. C:
Wang. Y. (2010). Indirect validation of tropospheric nitrogen dioxide retrieved from the OMI satellite
instrument: Insight into the seasonal variation of nitrogen oxides at northern midlatitudes. J Geophys
Res 115: D05302. http://dx.doi.org/10.1029/2009JD013351
Lamsal LN: Martin. RV: Van Donkelaar. A: Steinbacher. M: Celarier. EA: Bucsela. E: Dunlea. EJ: Pinto.
JP. (2008). Ground-level nitrogen dioxide concentrations inferred from the satellite-borne ozone
monitoring instrument. J Geophys Res 113: D16308. http://dx.doi.org/10.1029/2007JD009235
Lee. BH: Wood. EC: Zahniser. MS: Mcmanus. JB: Nelson. DP: Herndon. SC: Santoni. GW: Wofsv. SC:
Munger. JW. (2011). Simultaneous measurements of atmospheric HONO and NO2 viaabsorption
spectroscopy using tunable mid-infrared continuous-wave quantum cascade lasers. Appl Phys B 102:
417-423. http://dx.doi.org/10.1007/s00340-010-4266-5
Lee. H: Gu. M: Kim. YJ: Hwang. J: Jung. U. (2012). First-time remote sensing of NO2 vertical
distributions in an urban street canyon using Topographic Target Light scattering Differential Optical
Absorption Spectroscopy (ToTaL-DOAS). Atmos Environ 54: 519-528.
http://dx.doi.0rg/10.1016/i.atmosenv.2012.02.065
January 2015 2-79 DRAFT: Do Not Cite or Quote
-------
Lee. K: Yeom. J: Yoon. C: Yang. W: Son. BS: Jeon. JM: Yu. SD: Park. CH. (2013). Seasonal and
geographic effects on predicting personal exposure to nitrogen dioxide by time-weighted
microenvironmental model. Atmos Environ 67: 143-148.
http://dx.doi.0rg/10.1016/j.atmosenv.2012.ll.010
Lee. L: Teng. AP: Wennberg. PO: Crounse. JD: Cohen. RC. (2014). On Rates and Mechanisms of OH and
O3 Reactions with Isoprene-Derived Hydroxy Nitrates. J Phys Chem A 118: 1622-1637.
http://dx.doi.org/10.1021/ip4107603
Li. X: Liu. C: Leung. D. (2005). Development of a model for the determination of air exchange rates for
street canyons. Atmos Environ 39: 7285-7296. http://dx.doi.Org/10.1016/j.atmosenv.2005.09.007
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. CH: Leung. DYC: Barth. MC. (2005). On the prediction of air and pollutant exchange rates in street
canyons of different aspect ratios using large-eddy simulation. Atmos Environ 39: 1567-1574.
http://dx.doi.0rg/10.1016/j.atmosenv.2004.08.036
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
Maclntyre. 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
Mao. J: Paulot. F: Jacob. DJ: Cohen. RC: Crounse. JD: Wennberg. PO: Keller. CA: Hudman. RC:
Barklev. MP: Horowitz. LW. (2013). Ozone and organic nitrates over the eastern United States:
Sensitivity to isoprene chemistry. J Geophys Res Atmos 118: 1125611268.
http://dx.doi.org/10.1002/igrd.50817
Marr. LC: Harlev. RA. (2002). Modeling the effect of weekday-weekend differences in motor vehicle
emissions on photochemical air pollution in central California. Environ Sci Technol 36: 4099-4106.
http://dx.doi.org/10.1021/es020629x
Martin. RV: Chance. K: Jacob. DJ: Kurosu. TP: Spurr. RJD: Bucsela. E: Gleason. JF: Palmer. PI: Bey. I:
Fiore. AM: Li. Q: Yantosca. RM: Koelemeijer. RBA. (2002). An improved retrieval of tropospheric
nitrogen dioxide from GOME. J Geophys Res 107: 4437. http://dx.doi.org/10.1029/2001JD001027
Maruo. YY: Ogawa. S: Ichino. T: Murao. N: Uchiyama. M. (2003). Measurement of local variations in
atmospheric nitrogen dioxide levels in Sapporo, Japan, using a new method with high spatial and high
temporal resolution. Atmos Environ 37: 1065-1074. http://dx.doi.org/10.1016/81352-2310(02)00974-
3
Massoli. P: Former. EC: Canagaratna. MR: Williams. LR: Zhang. Q: Sun. Y: Schwab. JJ: Trimborn. A:
Onasch. TB: Demerjian. KL: Kolb. CE: Worsnop. PR: Jayne. JT. (2012). Pollution gradients and
chemical characterization ofparticulatematter from vehicular traffic near major roadways: Results
from the 2009 Queens College air quality study in NYC. Aerosol Sci Technol 46: 1201-1218.
http://dx.doi.org/10.1080/02786826.2012.701784
McAdam. K: Steer. P: Perrotta. K. (2011). Using continuous sampling to examine the distribution of
traffic related air pollution in proximity to a major road. Atmos Environ 45: 2080-2086.
http://dx.doi.0rg/10.1016/j.atmosenv.2011.01.050
McClenny. WA: Williams. EJ: Cohen. RC: Stutz. J. (2002). Preparing to measure the effects of the NOx
SIP call—methods for ambient air monitoring of NO, NO2, NOy, and individual NOz species. J Air
Waste Manag Assoc 52: 542-562. http://dx.doi.org/10.1080/10473289.2002.10470801
Mcdonald. BC: Dallmann. TR: Martin. EW: Harlev. RA. (2012). Long-term trends in nitrogen oxide
emissions from motor vehicles at national, state, and air basin scales. J Geophys Res 117: DOOV18.
http://dx.doi.org/10.1029/2012JD018304
January 2015 2-80 DRAFT: Do Not Cite or Quote
-------
Mcmeeking. GR: Kreidenweis. SM: Baker. S: Carrico. CM: Chow. JC: Collett JL. Jr: Hao. WM: Holden.
AS: Kirchstetter. TW: Malm. WC: Moosmiiller. H: Sullivan. AP: Wold. CE. (2009). Emissions of
trace gases and aerosols during the open combustion of biomass in the laboratory. J Geophys Res
Atmos 114: D19210. http://dx.doi.org/10.1029/2009id011836
Mebust AK: Russell AR: Hudman. RC: Valin. LC: Cohen. RC. (2011). Characterization of wildfire NOx
emissions using MODIS fire radiative power and OMI tropospheric NO2 columns. Atmos Chem Phys
11: 5839-5851. http://dx.doi.org/10.5194/acp-ll-5839-2011
Mollner. AK: Valluvadasan. S: Feng. L: Sprague. MK: Okumura. M: Milligan. DB: Bloss. WJ: Sander.
SP: Martien. PT: Harlev. RA: McCoy. AB: Carter. WRL. (2010). Rate of gas phase association of
hydroxyl radical and nitrogen dioxide. Science 330: 646-649.
http://dx.doi.org/10.1126/science. 1193030
Monn. C. (2001). Exposure assessment of air pollutants: Areview on spatial heterogeneity and
indoor/outdoor/personal exposure to suspended paniculate matter, nitrogen dioxide and ozone
[Review]. Atmos Environ 35: 1-32. http://dx.doi.org/10.1016/sl352-2310(00)00330-7
Nakai S: Nitta. H: Maeda. K. (1995). Respiratory health associated with exposure to automobile exhaust
II Personal NO2 exposure levels according to distance from the roadside. J Expo Sci Environ
Epidemiol 5: 125-136.
Nishino. N: Atkinson. R: Arev. J. (2008). Formation of nitro products from the gas-phase OH radical-
initiated reactions of toluene, naphthalene, and biphenyl: effect of NO2 concentration. Environ Sci
Technol 42: 9203-9209. http://dx.doi.org/10.1021/es802046m
Nitta. H: Sato. T: Nakai. S: Maeda. K: Aoki. S: Ono. M. (1993). Respiratory health associated with
exposure to automobile exhaust: I. Results of cross-sectional studies in 1979, 1982, and 1983. Arch
Environ Health 48: 53-58. http://dx.doi.org/10.1080/00039896.1993.9938393
Nunnermacker. LJ: Imre. D: Daum. PH: Kleinman. L: Lee. YN: Lee. JH: Springston. SR: Newman. L:
Weinstein-Llovd. J: Luke. WT: Banta. R: Alvarez. R: Senff. C: Sillman. S: Holdren. M: Keiglev. GW:
Zhou. X. (1998). Characterization of the Nashville urban plume on July 3 and July 18, 1995. J
Geophys Res 103: 28129-28148.
Olaguer. EP (2012). The potential near-source ozone impacts of upstream oil and gas industry emissions.
J Air Waste Manag Assoc 62: 966-977. http://dx.doi.org/10.1080/10962247.2012.688923
Olaguer. EP: Kolb. CE: Lefer. B: Rappengliick. B: Zhang. RY: Pinto. JP (2014). Overview of the SHARP
campaign: Motivation, design, and major outcomes. J Geophys Res 119: 2597-2610.
http://dx.doi.org/10.1002/2013JD019730
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-
3289.59.11.1258
Ostfaoff. HP: Roberts. JM: Ravishankara. AR: Williams. EJ: Lerner. BM: Sommariva. R: Bates. TS:
Coffman. D: Quinn. PK: Dibb. JE: Stark. H: Burkholder. JB: Talukdar. RK: Meagher. J: Fehsenfeld.
FC: Brown. SS. (2008). High levels of nitryl chloride in the polluted subtropical marine boundary
layer. Nat Geosci 1: 324-328. http://dx.doi.org/10.1038/ngeol77
Pachon. JE: Balachandran. S: Hu. Y: Mulholland. JA: Darrow. LA: Sarnat JA: Tolbert. PE: Russell AG.
(2012). Development of outcome-based, multipollutant mobile source indicators. J Air Waste Manag
Assoc 62: 431-442. http://dx.doi.org/10.1080/10473289.2012.656218
Pacsi. AP: Alhajeri. NS: Zavala-Araiza. D: Webster. MD: Allen. DT. (2013). Regional air quality impacts
of increased natural gas production and use in Texas. Environ Sci Technol 47: 3521-3527.
http://dx.doi.org/10.1021/es3044714
Parrish. DP: Fehsenfeld. FC. (2000). Methods for gas-phase measurements of ozone, ozone precursors
and aerosol precursors. Atmos Environ 34: 1921-1957. http://dx.doi.org/10.1016/S1352-
2310(99)00454-9
January 2015 2-81 DRAFT: Do Not Cite or Quote
-------
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
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. http://dx.doi.org/10.5194/acp-9-1451-2009
Perring. AE: Pusede. SE: Cohen. RC. (2013). An observational perspective on the atmospheric impacts of
alkyl and multifunctional nitrates on ozone and secondary organic aerosol. Chem Rev 113: 5848-5870.
http://dx.doi.org/10.1021/cr300520x
Pfister. GG: Wiedinmyer. C: Emmons. LK. (2008). Impacts of the fall 2007 California wildfires on
surface ozone: Integrating local observations with global model simulations. Geophys Res Lett 35:
L19814. http://dx.doi.org/10.1029/2008gl034747
Pinto. JP: Dibb. J: Lee. BH: Rappengliick. B: Wood. EC: Lew. M: Zhang. RY: Lefer. B: Ren. XR: Stutz.
J: Tsai C: Ackermann. L: Golovko. J: Herndon. SC: Oakes. M: Meng. QY: Munger. JW: Zahniser. M:
Zheng. J. (2014). Intercomparison of field measurements of nitrous acid (HONO) during the SHARP
campaign. J Geophys Res Atmos 119: 5583-5601. http://dx.doi.org/10.1002/2013JD020287
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.
Pleijel. H: Karlsson. GP: Gerdin. EB. (2004). On the logarithmic relationship between NO2 concentration
and the distance from a highroad. Sci Total Environ 332: 261-264.
http://dx.doi.0rg/10.1016/j.scitotenv.2004.03.020
Polidori. A: Fine. PM. (2012). Ambient concentrations of criteria and air toxic pollutants in close
proximity to a freeway with heavy-duty diesel traffic. Final report. South Coast Air Quality
Management District, http://www.aqmd.gov/docs/default-source/air-qualitv/air-qualitv-monitoring-
studies/near-roadwav-studv.pdf?sfvrsn=2
Pollack. IB: Lerner. BM: Ryerson. TB. (2011). Evaluation of ultraviolet light-emitting diodes for
detection of atmospheric NO2 by photolysis - chemiluminescence. J Atmos Chem 65: 111-125.
http://dx.doi.org/10.1007/sl0874-011-9184-3
Pollack. IB: Ryerson. TB: Trainer. M: Parrish. DP: Andrews. AE: Atlas. EL: Blake. PR: Brown. SS:
Commane. R: Daube. BC: De Gouw. JA: Dube. WP: Flynn. J: Frost GJ: Oilman. JB: Grossberg. N:
Hollowav. JS: Kofler. J: Kort. EA: Kuster. WC: Lang. PM: Lefer. B: Lueb. RA: Neuman. JA: Nowak.
JB: Novelli PC: Peischl J: Perring. AE: Roberts. JM: Santoni. G: Schwarz. JP: Spackman. J. R.:
Wagner. NL: Warneke. C: Washenfelder. RA: Wofsv. SC: Xiang. B. (2012). Airborne and ground-
based observations of a weekend effect in ozone, precursors, and oxidation products in the California
South Coast Air Basin. J Geophys Res 117: DOOV05. http://dx.doi.org/10.1029/2011JD016772
Pun. BK: Seigneur. C: White. W. (2003). Day-of-week behavior of atmospheric ozone in three US cities.
J Air Waste Manag Assoc 53: 789-801. http://dx.doi.org/10.1080/10473289.2003.10466231
Rao. KS: Gunter. RL: White. JR: Hosker. RP (2002). Turbulence and dispersion modeling near highways.
Atmos Environ 36: 4337-4346. http://dx.doi.org/10.1016/81352-2310(02)00353-9
Ren. X: van Duin. D: Cazorla. M: Chen. S: Mao. J: Zhang. L. i: Brune. WH: Flynn. JH: Grossberg. N:
Lefer. BL: Rappenglueck. B: Wong. K: Tsai. C: Stutz. J: Dibb. JE: Jobson. BT: Luke. WT: Kelley. P.
(2013). Atmospheric oxidation chemistry and ozone production: Results from SHARP 2009 in
Houston, Texas. J Geophys Res Atmos 118: 5770-5780. http://dx.doi.org/10.1002/jgrd.50342
Richmond-Bryant. J: Reff. A. (2012). Air pollution retention within a complex of urban street canyons: A
two-city comparison. Atmos Environ 49: 24-32. http://dx.doi.Org/10.1016/j.atmosenv.2011.12.036
January 2015 2-82 DRAFT: Do Not Cite or Quote
-------
Riedel. TP: Wolfe. GM: Danas. KT: Oilman. JB: Kuster. WC: Bon. DM: Vlasenko. A: Li. SM: Williams.
EJ: Lerner. BM: Veres. PR: Roberts. JM: Hollowav. JS: Lefer. B: Brown. SS: Thornton. JA. (2014).
An MCM modeling study of nitryl chloride (C1NO2) impacts on oxidation, ozone production and
nitrogen oxide partitioning in polluted continental outflow. Atmos Chem Phys 14: 3789-3800.
http://dx.doi.org/10.5194/acp-14-3789-2014
Roberts. JM: Veres. PR: Vandenboer. TC: Warneke. C: Graus. M: Williams. EJ: Lefer. B: Brock. CA:
Bahreinj R: Oztiirk. F: Middlebrook. AM: Wagner. NL: Dube. WP: De Gouw. JA. (2014). New
insights into atmospheric sources and sinks of isocyanic acid, HNCO, from recent urban and regional
observations. J Geophys Res Atmos 119: 1060-1072. http://dx.doi.org/10.1002/2013JD019931
Rodes. CE: Holland. DM. (1981). Variations of NO, NO2 and O3 concentrations downwind of a Los
Angeles freeway. Atmos Environ 15: 243-250.
Rodgers. MO: Davis. DP. (1989). A UV-photofragmentation/laser-induced fluorescence sensor for the
atmospheric detection of HONO. Environ Sci Technol 23: 1106-1112.
Rollins. AW: Kiendler-Scharr. A: Fry. JL: Brauers. T: Brown. SS: Porn. HP: Dube. WP: Fuchs. H:
Mensah. A: Mentel TF: Rohrer. F: Tillmann. R: Wegener. R: Wooldridge. PJ: Cohen. RC. (2009).
Isoprene oxidation by nitrate radical: Alkyl nitrate and secondary organic aerosol yields. Atmos Chem
Phys 9: 6685-6703. http://dx.doi.org/10.5194/acp-9-6685-2009
Roorda-Knape. MC: Janssen. NAH: De Hartog. JJ: Van Vliet. PHN: Harssema. H: Brunekreef. B. (1998).
Air pollution from traffic in city districts near major motorways. Atmos Environ 32: 1921-1930.
Russell AR: Perring. AE: Valin. LC: Bucsela. EJ: Browne. EC: Wooldridge. PJ: Cohen. RC. (2011). A
high spatial resolution retrieval of NO 2 column densities from OMI: method and evaluation. Atmos
Chem Phys 11: 8543-8554. http://dx.doi.org/10.5194/acp-ll-8543-2011
Rverson. TB: Andrews. AE: Angevine. WM: Bates. TS: Brock. CA: Cairns. B: Cohen. RC: Cooper. OR:
de Gouw. JA: Fehsenfeld. FC: Ferrare. RA: Fischer. ML: Flagan. RC: Goldstein. AH: Hair. JW:
Hardestv. RM: Hostetler. CA: Jimenez. JL: Langford. AO: Mccaulev. E: Mckeen. SA: Molina. LT:
Nenes. A: Oltmans. SJ: Parrish. DP: Pederson. J. R.: Pierce. RB: Prather. K: Quinn. PK: Seinfeld. JH:
Senff. CJ: Sorooshian. A: Stutz. J: Surratt. JD: Trainer. M: Volkamer. R: Williams. EJ: Wofsv. SC.
(2013). The 2010 California Research at the Nexus of Air Quality and Climate Change (CalNex) field
study. J Geophys Res Atmos 118: 5830-5866. http://dx.doi.org/10.1002/jgrd.50331
Saltzman. BE. (1954). Colorimetric microdetermination of nitrogen dioxide in the atmosphere. Anal
Chem 26: 1949-1955.
Sander. SP: Abbatt. JPD: Barker. JR: Burkholder. JB: Friedl RR: Golden. DM: Huie. RE: Kolb. CE:
Kurylo. MJ: Moortgat. GK: Orkin. VL: Wine. PH. (2011). Section 1. Biomolecular reactions. In
Chemical kinetics and photochemical data for use in atmospheric studies: Evaluation number 17 (JPL
Publication 10-6) (pp. 1-1 - 1-34). Pasadena, CA: Jet Propulsion Laboratory.
Sarnat SE: 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/jes.2009.10
Sedefian. L: Rao. ST: Czapskj U. (1981). Effects of traffic-generated turbulence on near-field dispersion.
Atmos Environ (1967) 15: 527-536. http://dx.doi.org/10.1016/0004-6981(81)90182-7
Seinfeld. JH: Pandis. SN. (1998). Atmospheric chemistry and physics: From air pollution to climate
change. New York: John Wiley & Sons. http://dx.doi.Org/10.1063/l.882420
Simoens. S: Ayrault M: Wallace. JM. (2007). The flow across a street canyon of variable width- Part 1:
Kinematic description. Atmos Environ 41: 9002-9017.
http://dx.doi.0rg/10.1016/j.atmosenv.2007.08.013
Simoens. S: Wallace. JM. (2008). The flow across a street canyon of variable widthPart 2: Scalar
dispersion from a street level line source. Atmos Environ 42: 2489-2503.
http://dx.doi.0rg/10.1016/i.atmosenv.2007.12.013
January 2015 2-83 DRAFT: Do Not Cite or Quote
-------
Simon. H: Kimura. Y: McGaughev. G: Allen. DT: Brown. SS: Osthoff. HP: Roberts. JM: Byun. D: Lee.
EX (2009). Modeling the impact of C1NO2 on ozone formation in the Houston area. J Geophys Res
Atmos 114: DOOF03. http://dx.doi.org/10.1029/2008id010732
Singer. BC: Hodgson. AT: Hotchi. T: Kim. JJ. (2004). Passive measurement of nitrogen oxides to assess
traffic-related pollutant exposure for the East Bay Children's Respiratory Health Study. Atmos Environ
38: 393-403. http://dx.doi.Org/10.1016/j.atmosenv.2003.10.005
Singh. HB: Salas. L: Herlth. D: Kolyer. R: Czech. E: Avery. M: Crawford. JH: Pierce. RB: Sachse. GW:
Blake. PR: Cohen. RC: Bertram. TH: Perring. A: Wooldridge. PJ: Dibb. J: Huev. G: Hudman. RC:
Turquetv. S: Emmons. LK: Flocke. F: Tang. Y: Carmichael GR: Horowitz. LW. (2007). Reactive
nitrogen distribution and partitioning in the North American troposphere and lowermost stratosphere. J
Geophys Res 112: D12S04. http://dx.doi.org/10.1029/2006JD007664
Smargiassi A: Baldwin. M: Pilger. C: Dugandzic. R: Brauer. M. (2005). Small-scale spatial variability of
particle concentrations and traffic levels in Montreal: a pilot study. Sci Total Environ 338: 243-251.
http://dx.doi.0rg/10.1016/j.scitotenv.2004.07.013
Southwest Research Institute. (2013). Phase 2 of the advanced collaborative emissions study. (SwRI Final
Report 03.17124). Alpharetta, GA: Coordinating Research Council, Inc.
http://crcao.org/reports/recentstudies2013/ACES%20Ph2/03-17124 CRC%20ACES%20Phase2-
%20FINAL%20Report Khalek-R6-SwRI.pdf
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: LI0801. http://dx.doi.org/10.1029/2008GL033521
Steinbacher. M: Zellweger. C: Schwarzenbach. B: Bugmann. S: Buchmann. B: Ordonez. C: Prevot ASH:
Hueglin. C. (2007). Nitrogen oxides measurements at rural sites in Switzerland: Bias of conventional
measurement techniques. J Geophys Res Atmos 112: D11307.
http://dx.doi.org/10.1029/2006JD007971
Stull. R. (2000). Meteorology for scientists and engineers: a technical companion book with Ahrens'
Meteorology Today.
Tang. MJ: Schuster. G: Crowlev. JN. (2014). Heterogeneous reaction of N2O5 with illite and Arizona test
dust particles. Atmos Chem Phys 14: 245-254. http://dx.doi.org/10.5194/acp-14-245-2014
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
Tokairin. T: Kitada. T. (2005). Study on the effect of porous fence on air quality and traffic noise level
around a double-decked road structure. Environ Monit Assess 105: 121-143.
http://dx.doi.org/10.1007/sl0661-005-3160-9
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
Tonse. SR: Brown. NJ: Harlev. RA: Jin. L. (2008). A process-analysis based study of the ozone weekend
effect. Atmos Environ 42: 7728-7736. http://dx.doi.Org/10.1016/j.atmosenv.2008.05.061
U.S. EPA (U.S. Environmental Protection Agency). (1993). Air quality criteria for oxides of nitrogen, vol.
1-3 [EPAReport]. (EPA/600/8-91/049aF-cF). Research Triangle Park, NC: U.S. Environmental
Protection Agency, Environmental Criteria and Assessment Office.
http://cfpub.epa. gov/ncea/cfm/recordisplav.cfm?deid=40179
U.S. EPA (U.S. Environmental Protection Agency). (1996). Air quality criteria for ozone and related
photochemical oxidants [EPAReport]. (EPA/600/P-93/004AF). Research Triangle Park, NC.
January 2015 2-84 DRAFT: Do Not Cite or Quote
-------
U.S. EPA (U.S. Environmental Protection Agency). (1998). SLAMS/NAMS/PAMS network review
guidance [EPAReport]. (EPA/454/R-98/003). Research Triangle Park, NC.
U.S. EPA (U.S. Environmental Protection Agency). (2006). Air quality criteria for ozone and related
photochemical oxidants [EPA Report]. (EPA/600/R-05/004AF). Research Triangle Park, NC: U.S.
Environmental Protection Agency, National Center for Environmental Assessment.
http://cfpub.epa.gov/ncea/cfm/recordisplav.cfm?deid=149923
U.S. EPA (U.S. Environmental Protection Agency). (2008a). Integrated science assessment for oxides of
nitrogen Health criteria [EPA Report]. (EPA/600/R-08/071). Research Triangle Park, NC: U.S.
Environmental Protection Agency, National Center for Environmental Assessment.
http://cfpub.epa.gov/ncea/cfm/recordisplav.cfm7de id=194645
U.S. EPA (U.S. Environmental Protection Agency). (2008b). Risk and exposure assessment to support the
review of the NO2 primary national ambient air quality standard [EPA Report]. (EPA-452/R-08-008a).
Research Triangle Park, NC: U.S. Environmental Protection Agency, Office of Air Quality Planning
and Standards, http://www.epa.gov/ttn/naaqs/standards/nox/data/20081121 NO2REA final.pdf
U.S. EPA (U.S. Environmental Protection Agency). (2009). Integrated science assessment for particulate
matter [EPA Report]. (EPA/600/R-08/139F). Research Triangle Park, NC: U.S. Environmental
Protection Agency, National Center for Environmental Assessment.
http://cfpub.epa.gov/ncea/cfm/recordisplav.cfm7de id=216546
U.S. EPA (U.S. Environmental Protection Agency). (2010). Final regulatory impact analysis (RIA) for the
NO2 National Ambient Air Quality Standards (NAAQS) [EPA Report]. Research Triangle Park, NC.
http://www.epa.gov/ttnecasl/regdata/RIAs/FinalNO2RIAfulldocument.pdf
U.S. EPA (U.S. Environmental Protection Agency). (2011). 2008 National Emissions Inventory data and
documentation. Available online at http://www.epa.gov/ttn/chief/net/2008inventorv.html (accessed
February 17,2011).
U.S. EPA (U.S. Environmental Protection Agency). (2013a). 2011 National Emissions Inventory data and
documentation. Available online at http://www.epa.gov/ttnchie l/net/2011 inventory.html (accessed
December 1, 2014).
U.S. EPA (U.S. Environmental Protection Agency). (2013b). Integrated science assessment for ozone and
related photochemical oxidants (final report) [EPA Report]. (EPA/600/R-10/076F). Washington, DC.
http://cfpub.epa.gov/ncea/isa/recordisplav.cfm?deid=247492
Uemura. H: Arisawa. K: Hiyoshi. M: Satoh. H: Sumiyoshi. Y: Morinaga. K: Kodama. K: Suzuki. T:
Nagai M: Suzuki. T. (2008). Associations of environmental exposure to dioxins with prevalent
diabetes among general inhabitants in Japan. Environ Res 108: 63-68.
http://dx.doi.0rg/10.1016/j.envres.2008.06.002
Unal A: Hu. Y: Chang. ME: Talat Odman. M: Russell. AG. (2005). Airport related emissions and impacts
on air quality: Application to the Atlanta International Airport. Atmos Environ 39: 5787-5798.
http://dx.doi.0rg/10.1016/j.atmosenv.2005.05.051
Valin. LC: Russell. AR: Cohen. RC. (2013). Variations of OH radical in an urban plume inferred from NO
2 column measurements. Geophys Res Lett 40: n/a-n/a. http://dx.doi.org/10.1002/grl.50267
Vardoulakis. S: Solazzo. E: Lumbreras. J. (2011). Intra-urban and street scale variability of BTEX, NO2
and O3 in Birmingham, UK: Implications for exposure assessment. Atmos Environ 45: 5069-5078.
http://dx.doi.0rg/10.1016/j.atmosenv.2011.06.038
Venkatram. A: Isakov. V: Thoma. E: Baldauf. R. (2007). Analysis of air quality data near roadways using
a dispersion model. Atmos Environ 41: 9481-9497. http://dx.doi.Org/10.1016/j.atmosenv.2007.08.045
Vette. A: Burke. J: Norris. G: Landis. M: Batterman. S: Breen. M: Isakov. V: Lewis. T: Gilmour. MI:
Kamal A: Hammond. D: Vedantham. R: Bereznicki. S: Tian. N: Croghan. C. (2013). The Near-Road
Exposures and Effects of Urban Air Pollutants Study (NEXUS): Study design and methods. Sci Total
Environ 448: 38-47. http://dx.doi.Org/10.1016/i.scitotenv.2012.10.072
January 2015 2-85 DRAFT: Do Not Cite or Quote
-------
Villena. G: Bejan. I: Kurtenbach. R: Wiesen. P: Kleffmann. J. (2011). Development of a new Long Path
Absorption Photometer (LOPAP) instrument for the sensitive detection of NO2 in the atmosphere.
Atmos Meas Tech 4: 1663-1676. http://dx.doi.org/10.5194/amt-4-1663-2011
Villena. G: Bejan. I: Kurtenbach. R: Wiesen. P: Kleffmann. J. (2012). Interferences of commercial NO2
instruments in the urban atmosphere and in a smog chamber. Atmos Meas Tech 5: 149-159.
http://dx.doi.org/10.5194/amt-5-149-2012
von Kuhlmann. R: Lawrence. MG: Poschl U: Crutzen. PJ. (2004). Sensitivities in global scale modeling
of isoprene. Atmos Chem Phys 4: 1-17.
Vutukuru. S: Dabdub. D. (2008). Modeling the effects of ship emissions on coastal air quality: A case
study of southern California. Atmos Environ 42: 3751-3764.
http://dx.doi.0rg/10.1016/j.atmosenv.2007.12.073
Walters. LJ: May. MS: Johnson. DE: MacMann. RS: Woodward. WA. (1999). Time-variability of NO x
emissions from Portland cement kilns. Environ Sci Technol 33: 700-704.
http://dx.doi.org/10.1021/es980203b
Wang. YJ: Zhang. KM. (2009). Modeling near-road air quality using a computational fluid dynamics
model, CFD-VIT-RIT. Environ Sci Technol 43: 7778-7783. http://dx.doi.org/10.1021/es9014844
Williams. EJ: Lerner. BM: Murphy. PC: Herndon. SC: Zahniser. MS. (2009). Emissions of NOx, SO2,
CO, and HCHO from commercial marine shipping during Texas Air Quality Study (TexAQS) 2006. J
Geophys Res 114: D21306. http://dx.doi.org/10.1029/2009JD012094
Wilson. JG: Kingham. S: Pearce. J: Sturman. AP (2005). A review of intraurban variations in paniculate
air pollution: implications for epidemiological research [Review]. Atmos Environ 39: 6444-6462.
http://dx.doi.0rg/10.1016/j.atmosenv.2005.07.030
Winer. AM: Peters. JW: Smith. JP: Pitts. JN. (1974). Response of commercial chemiluminescent nitric
oxide-nitrogen dioxide analyzers to other nitrogen-containing compounds. Environ Sci Technol 8:
1118-1121. http://dx.doi.org/10.1021/es60098a004
Wooldridge. PJ: Perring. AE: Bertram. TH: Flocke. FM: Roberts. JM: Singh. HB: Huev. LG: Thornton.
JA: Wolfe. GM: Murphy. JG: Fry. JL: Rollins. AW: Lafranchi. BW: Cohen. RC. (2010). Total Peroxy
Nitrates (Sigma PNs) in the atmosphere: the Thermal Dissociation-Laser Induced Fluorescence (TD-
LIF) technique and comparisons to speciated PAN measurements. Atmos Meas Tech 3: 593-607.
http://dx.doi.org/10.5194/amt-3-593-2010
Xie. Y: Paulot. F: Carter. WPL: Nolte. CG: Luecken. DJ: Hutzell. WT: Wennberg. PO: Cohen. RC:
Finder. RW. (2013). Understanding the impact of recent advances in isoprene photooxidation on
simulations of regional air quality. Atmos Chem Phys 13: 8439-8455. http://dx.doi.org/10.5194/acp-
13-8439-2013
Zhou. Y: Levy. JI. (2007). Factors influencing the spatial extent of mobile source air pollution impacts: A
meta-analysis [Review]. BMC Public Health 7: 89. http://dx.doi.org/10.1186/1471-2458-7-89
Zhu. DZ: Nussbaum. NJ: Kuhns. HP: Chang. MCO: Sodeman. D: Moosmuller. H: Watson. JG. (2011).
Real-world PM, NO(x), CO, and ultrafine particle emission factors for military non-road heavy duty
diesel vehicles. Atmos Environ 45: 2603-2609. http://dx.doi.Org/10.1016/j.atmosenv.2011.02.032
Zou. X: Shen. Z: Yuan. T: Yin. S: Cai J: Chen. L: Wang. W. (2006). Shifted power-law relationship
between NO2 concentration and the distance from a highway: Anew dispersion model based on the
wind profile model. Atmos Environ 40: 8068-8073. http://dx.doi.Org/10.1016/j.atmosenv.2006.09.012
January 2015 2-86 DRAFT: Do Not Cite or Quote
-------
CHAPTER 3 EXPOSURE TO OXIDES OF
NITROGEN
3.1 Introduction
1 Assessment of exposure to ambient oxides of nitrogen builds from the characterization of
2 concentrations and atmospheric chemistry presented in Chapter 2. The primary
3 conclusions from Chapter 2 were that NCh concentrations have declined over the past
4 20 years, but concentrations are still elevated near roads and in urban areas, with
5 vehicular traffic and off-highway vehicles contributing the majority of NC>2 emissions.
6 For this reason, NC>2 exposure assessment focuses predominantly on urban and near-road
7 settings.
8 True personal exposure to ambient oxides of nitrogen is given by the concentration of
9 oxides of nitrogen emitted from ambient sources and encountered by an individual over a
10 given time. Personal ambient exposure is influenced by a number of factors, including:
11 • time-activity in different microenvironments (e.g., vehicle, residence,
12 workplace, outdoor);
13 • climate (e.g., weather, season);
14 • characteristics of indoor microenvironments (e.g., window openings,
15 draftiness, air conditioning); and,
16 • microenvironmental emission sources (e.g., roadways, construction
17 equipment, indoor gas stoves) and concentrations.
18 Surrogates for personal exposure to ambient oxides of nitrogen include ambient NC>2
19 measured at a central site monitor or modeled using spatial techniques such as land use
20 regression (LUR), Gaussian dispersion models, or chemical transport models (CTM). All
21 exposure surrogates are subject to measurement errors related to spatial and temporal
22 variability of the ambient concentration field, quality of additional input data,
23 representativeness of predictor variables, and accuracy of the monitoring or modeling
24 methodology. The following sections describe methods to estimate personal exposure,
25 current data used to characterize exposure to ambient oxides of nitrogen, exposure-related
26 factors that influence interpretation of epidemiologic models of the health effects of
27 oxides of nitrogen, and considerations for use of exposure metrics in epidemiologic
28 studies of different design. This chapter focuses on the ambient component of personal
29 exposure to NCh, because the NAAQS regulates ambient oxides of nitrogen, for which
30 NO2 is the indicator. However, studies using total personal NC>2 exposure and indoor NC>2
31 concentrations as exposure metrics can also inform the understanding of exposure and
32 related health effects and so are included as supporting evidence where appropriate. This
January 2015 3-1 DRAFT: Do Not Cite or Quote
-------
1 chapter focuses on studies of exposure among the general population. Exposure of at-risk
2 groups, based for example on socioeconomic status, race, and proximity to roadways, is
3 addressed in Chapter 7; occupational exposures to ambient NO2 are discussed in
4 Chapter 7 within the subsections for socioeconomic status and proximity to roadways.
5 Intake of NO2 based on ventilation rate, and in relation to physical activity, is described
6 in Section 4.2. The information provided in this chapter will be used to help interpret the
7 health effects studies of NO2 exposure presented in Chapters 5, 6, and 7.
3.2 Methodological Considerations for Use of Exposure Data
8 The following sections outline various facets of characterizing NC>2 exposure, including
9 research-grade (i.e., central site) and personal NO2 exposure sampling techniques and
10 NC>2 exposure modeling. The section ends with a discussion of the application of
11 measurement and modeling techniques in epidemiogic studies of different designs.
3.2.1 Measurement
3.2.1.1 Central Site and Near-Road Monitoring
12 Monitoring of NCh concentrations by chemiluminescent sampling is described in detail in
13 Section 2.4.1 along with limitations of the monitoring methodology. In summary, NC>2
14 concentrations are calculated by FRM as the difference between NO measured in the air
15 stream that has passed over a heated MoOx substrate (measuring total oxides of nitrogen)
16 and NO in the air stream that was diverted away from the substrate. FRMs are subject to
17 positive bias because oxidized nitrogen compounds other than NO2 are often detected by
18 the MoOx substrate. A FEM is also available to measure NO2 directly using a photolytic
19 converter to reduce NO2 to NO. Evaluation of the chemilumine scent method is provided
20 in Section 2.4.1 along with a description of the measuring technique. Monitors set up by
21 state agencies as part of the SLAMS network that report to the AQS are typically
22 centrally sited, although the same monitors are used in select cases for near-road
23 monitoring. See Section 2.4.5 for more details.
24 In addition to judging compliance with the NAAQS, NO2 concentrations measured by
25 centrally sited or near-road FRMs and FEMs are frequently used by epidemiologic
26 researchers as exposure metrics for studies of the health effects of exposure to oxides of
27 nitrogen, as described further in Section 3.4. Central site monitoring data can be used in
28 epidemiologic studies of short-term exposure to NO2 when focused on the time series of
January 2015 3-2 DRAFT: Do Not Cite or Quote
-------
1 exposure or in epidemiologic studies of long-term exposure when comparing average
2 NO2 concentrations among different geographic areas. Section 3.4.3 explores the factors
3 causing errors associated with siting central site or near-road monitors at a single
4 location, and Section 3.4.5 considers the influence of those errors on health effect
5 estimates, respectively. Briefly, with respect to time-series exposure estimation for
6 epidemiologic studies of short-term exposure, correlation between measured
7 concentration and concentration and some distant point decreases with distance. For
8 epidemiologic studies of long-term exposure to NC>2, difference between the measured
9 concentration and the true exposure would result in exposure misclassification. These
10 issues are potentially exacerbated by the fact that there are a limited number of samplers
11 in the network.
3.2.1.2 Personal and Area Sampling
12 Personal sampling for NC>2 was described in detail in Annex 3.3 to the 2008 ISA for
13 Oxides of Nitrogen (U.S. EPA. 2008) and is briefly summarized here. Active sampling
14 systems typically involve air pumped past a chemiluminescent device; they enable
15 measurements of NC>2 over short time periods to produce near real-time data. Given the
16 weight of most active sampling systems, they are not used extensively for personal
17 sampling. Passive samplers based on Pick's first law of diffusion are more commonly
18 deployed for personal or area NO2 sampling in a badge, tube, or radial manifold. These
19 are typically deployed over periods ranging from a few days to several weeks. Passive
20 sampling results are integrated over the time period during which the sorbent material is
21 exposed, which is selected by the user and usually spans days to weeks. The 2008 ISA for
22 Oxides of Nitrogen (U.S. EPA. 2008) reported that, depending on the sorbent material,
23 personal NO2 samplers may be subject to biases related to interferences from HONO,
24 PAN, HNO3 (Gairetal.. 1991). and high RH (Centre di Ricerche Ambientali. 2006).
25 These biases depend on ambient temperature and atmospheric levels of the copollutants
26 and relative humidity. Personal sampling for NO2 exposure is most commonly used in
27 epidemiologic panel studies.
28 Recent work has been performed to evaluate passive sampling device performance.
29 Sather et al. (2007) compared Ogawa passive samplers with a collocated NO2 FRM
30 monitor over a 4-week field study in El Paso, TX and observed good agreement, with an
31 average absolute difference of 1.2 ppb with R2 = 0.95. For measurements in Umea,
32 Sweden, Hagenbjork-Gustafsson et al. (2009) observed that, when using the
33 manufacturer's recommended uptake rates to calculate concentration, passive NO2
34 measurements were negatively biased by 9.1%, and NOx concentration measurements
35 were positively biased by 15% compared with an FRM. When uptake rates were derived
January 2015 3-3 DRAFT: Do Not Cite or Quote
-------
1 in the field based on the chemiluminescent FRM, NO2 measurements were positively
2 biased by 2%, and NOx concentration measurements were unbiased compared with the
3 FRM. These results suggest that deviation from temperature conditions under which the
4 samplers were laboratory tested may lead to biased results. Jimenez et al. (2011) used
5 Palmes-type passive diffusion tubes to measure both NO2 and NOx concentrations and
6 investigated specific sources of biases in their measurements. They found that, within the
7 passive diffusion tubes, NO and Os were reacting to form NO2, causing NO
8 measurements to be negatively biased while NO2 measurements were positively biased.
9 Wind was also a source of positive bias in the NO2 and NOx concentration measurements
10 because increased airflow effectively reduced the diffusion lengths of the gas collection
11 tubes. In laboratory and field evaluation of NO2 passive diffusion tubes, Buzicaetal.
12 (2008) observed negligible difference between the diffusion tubes and FRM
13 measurements; however, uncertainty increased with decreasing concentration. When
14 comparing biases among samplers, note that the FRM is subject to positive biases related
15 to sensitivity to PAN, RONO2, and HNO3 (see Sections 2.4.1 and 3.2.1.1).
16 Triethanolamine (TEA) is often employed as a sorbent material in denuders used for
17 capturing NO2 during active sampling and in passive sampling because it can be applied
18 in an even coating. However, sampling efficiency is sensitive to sampler flow rate (Vichi
19 and De Santis. 2012). relative humidity (Poddubny and Yushketova. 2013; Sereviciene
20 and Paliulis. 2012; Vardoulakis et al.. 2009). averaging time (Vardoulakis et al.. 2009).
21 and ambient temperature (Poddubny and Yushketova. 2013). Heal (2008) found that NO2
22 bias was sensitive to method of application of the TEA to the substrate. Sekine et al.
23 (2008) and Nishikawa et al. (2009) experimented with size and number of filters,
24 respectively, in a passive sampler and found minimal effect on NO2 or NOx
25 concentration. However, Ozden and Dogeroglu (2008) observed that TEA-complexed
26 NO2 was sensitive to photodegradation if not stored in a dark glass tube, resulting in
27 underprediction of NO2 exposure.
28 Recent attention has been given to using passive or miniature active monitors for
29 saturation sampling, i.e., siting monitors over a dense grid. This is typically done in urban
30 areas. For example, Ross et al. (2013) sited roughly 150 passive NO2 monitors across the
31 five boroughs of New York City to create a dense concentration map for exposure
32 estimates and to provide training and validation data for LUR. Similarly, Shmool et al.
33 (2014) deployed Ogawa passive badges for NO2 sampling, along with PM2 5, BC, relative
34 humidity, and barometric pressure, across metropolitan Pittsburgh, PA. The monitoring
35 boxes were sited to capture air pollution gradients along the urban-to-suburban land use
36 gradient and included areas influenced by industrial sources and highways. Skouloudis
37 and Kassomenos (2014) deployed sensors for NO2, NOx, CO, Os, and CeRe to correspond
38 to the population distribution on the island of Malta. Active samplers were used in this
January 2015 3-4 DRAFT: Do Not Cite or Quote
-------
1 scheme, with a global positioning system (GPS) and data transmission capabilities for
2 near real-time analysis. Skouloudis and Kassomenos (2014) proposed that these dense
3 area samplers could also be assimilated with satellite measurements to improve the
4 accuracy of the exposure estimates.
3.2.2 Modeling
5 Computational models can be employed to provide estimates of exposure in
6 epidemiologic studies when measurements are not available at locations and/or times
7 needed to estimate spatial and temporal variability in concentration within communities
8 and the epidemiologic study design requires greater spatial variability than attainable
9 through ambient NC>2 measurements. These methods can sometimes account for complex
10 urban morphometry and meteorology, which can interact to cause turbulence that may
11 affect pollutant residence times (Fernando. 2010) or incorporate localized sources that
12 might not otherwise be detected by central site monitoring (Goldman et al.. 2012). Such
13 estimates can then be used as inputs to exposure models described in Section 3.4. These
14 modeling approaches produce data at times and/or locations where exposures are
15 uncharacterized, but each method carries its own uncertainty (Fuentes. 2009). Detailed
16 descriptions of computational models used for predicting spatially resolved concentration
17 profiles for exposure assessment have been provided in Section AX 3.6 of the 2008 ISA
18 for Oxides of Nitrogen Annex (U.S. EPA. 2008) and Section 3.8 of the 2009 ISA for PM
19 (U.S. EPA. 2009). Methods include LUR models, spatial interpolation through statistical
20 techniques, CTM, and dispersion models.
3.2.2.1 Statistical Modeling
Land Use Regression Models
21 LUR modeling has been applied extensively to estimate the spatial distribution of
22 ambient NC>2 or NO for exposure assessment on a neighborhood or urban scale for
23 application in epidemiologic studies of long-term exposure (Clougherty et al.. 2013;
24 Hatzopoulou et al.. 2013; Cesaroni etal.. 2012; Gonzales et al. 2012; Mukerjee et al..
25 2012a; Mukerjee et al.. 2012b; Oiamo etal.. 2012; Esplugues et al., 2011; Fernandez-
26 Somoano etal.. 2011: Hystad etal.. 2011: Oiamo etal.. 2011; Rose etal.. 2011; Smith
27 etal.. 2011; Szpiro etal.. 2011; Adamkiewicz et al.. 2010; Aguilera etal.. 2009; Cohen
28 et al.. 2009; Hart etal.. 2009; Iniguez et al.. 2009; Karr et al.. 2009; Mukeriee et al..
29 2009; Su etal.. 2009b; Aguilera et al.. 2008; Atari et al.. 2008; Cesaroni et al.. 2008;
30 Rosenlund et al.. 2008). As such, LUR was used in many of the long-term epidemiologic
January 2015 3-5 DRAFT: Do Not Cite or Quote
-------
1 studies described in Chapter 6. LUR fits a multiple linear regression model of
2 concentration based on land use data and then applies that model to locations without
3 monitors as an attempt to increase heterogeneity in the spatial resolution of the
4 concentration field compared with other methods, such as central site monitoring
5 (Marshall et al., 2008). Recently, LUR has been implemented to examine local-scale
6 concentration estimates across the U.S. (Novotny et al.. 2011; Hart et al.. 2009) and
7 Canada (Hvstad et al., 2011). Models are typically calibrated using data from NO2 or NO
8 from passive sampler measurements and several predictor variables, such as land use,
9 road length, population density, and proximity to areas of high concentrations (city
10 center, major road and/or highway, and point sources). Given that most passive
11 measurement methods are not designed for short-term sampling, LUR models are
12 typically based on several days, weeks, or years of data and hence do not account for
13 short-term temporal variability well. Hence, LUR is commonly used to estimate air
14 pollution exposure in long-term epidemiologic studies. Several methodological issues
15 must be considered when interpreting LUR model results; these issues include number of
16 measurement sites used to fit the statistical model, predictor variable selection, and
17 comparison of LUR performance among LUR model formulations and with other
18 models. These issues affect how well the spatial variability of NO2 or NOx concentration
19 in a city is represented by the LUR.
20 Finer spatial resolution of calibration points can improve goodness of fit of the model for
21 the city in which it was fit. Using 155 monitoring sites throughout New York City,
22 Clougherty et al. (2013) ran an LUR with resolutions down to 50 m (R2 = 0.671). At this
23 fine scale, roadways and localized sources can be better represented than at coarser
24 scales. Parenteau and Sawada (2012) examined LUR model performance when basing the
25 model on successively finer spatial resolution from 2 km down to 50 m, with the
26 geographic borders of the finely resolved regions tied to population groupings based on
27 population density mapping. The two finer resolution approaches yielded better
28 agreement with measured NO2 data (R2 = 0.80-0.81) than the less spatially resolved
29 approach (R2 = 0.70). Likewise, Dijkema et al. (2011) compared LUR based on spatial
30 resolution and observed better agreement with NO2 observations for neighborhood-level
31 simulations (R2 = 0.57) compared with whole-city simulations (R2 = 0.47). Janssen et al.
32 (2012) proposed using LUR to improve performance of a CTM by downscaling the CTM
33 to the LUR. Downscaling entails a redistribution of the CTM-modeled concentrations
34 through a statistical model to conform to measured concentrations at points in space
35 where measurements are available using the LUR-derived regression parameters. Janssen
36 et al. (2012) found that the spatial representativeness of the CTM for NO2 improved by
37 roughly 20% when incorporating the LUR downscaler, based on a comparison of the
1 Unless otherwise noted for the LUR studies, R2 refers to model fit.
January 2015 3-6 DRAFT: Do Not Cite or Quote
-------
1 CTM and downscaled CTM with central site monitor measurements. It is worth noting
2 that any errors and uncertainties associated with a particular LUR run would transfer to
3 the downscaled result if LUR were used as a basis for downscaling CTM results.
4 Studies have evaluated LUR model performance when the LUR was fit with different
5 numbers of NCh measurement sites and observed that the LUR model design is sensitive
6 to the number of measurement sites. Basagana et al. (2012) evaluated LUR models for
7 24-120 NO2 measurement sites in Girona, Spain and different numbers of predictor
8 variables, starting with 106 prediction variables related to land use and then reducing the
9 set to 18 components through principal component analysis (PCA). Johnson et al. (2010)
10 evaluated LUR performance in New Haven, CT when the LUR model was fit with NO2
11 data from 25-285 measurement sites. Wang et al. (2012) also evaluated LUR
12 performance when fit with 24-120 NCh monitors distributed across the Netherlands.
13 These studies (Basagana et al.. 2012; Wang et al.. 2012; Johnson et al.. 2010) observed
14 that, when a large number of prediction covariates were used, the model performed better
15 (higher adjusted R2 and R2 for cross-validation) for a smaller number of NC>2
16 measurement sites compared with the model using a larger number of NC>2 sites, but
17 when the number of prediction covariates was reduced through PCA, then a larger
18 number of NCh measurement sites was needed.
19 LUR results may not be generalizable between or across cities unless the model fit is
20 randomly distributed in space across the city, includes model training data covering the
21 complete ambient concentration distribution, and the cities are similar with respect to
22 source strength, source distribution, and topography. Allen et al. (2011) developed
23 separate LUR models for two Canadian cities (Winnipeg, Manitoba and Edmonton,
24 Alberta) with 50 calibration points each and then applied the models to the other city to
25 compare performance. As anticipated, locally generated model performance
26 (NO2: R2 = 0.81-0.84; NO: R2 = 0.55-0.56) was superior to performance of the model fit
27 for the other city (NO2: R2 = 0.37-0.52; NO: R2 = 0.24-0.41) and to bivariate local
28 models using only road proximity (R2 < 0.19). NCh models consistently performed better
29 than NO models. Wang et al. (2014) developed a LUR model for NO2 based on data from
30 23 European study areas (containing 20-40 sites within each study area) with NO2,
31 PM2 5, land use, and traffic data. Given the continental design of the study, a regional
32 background concentration variable was also imposed on the model. The LUR model fit
33 was R2 = 0.56 for all of the urban areas combined. After fitting the LUR model, Wang
34 et al. (2014) tested the LUR model's ability to predict concentrations for different
35 configurations of cities by leaving one out of different analyses and found comparable
36 results (R2 = 0.59). Generally, the R2 for NO2 were either comparable or lower than R2 for
37 single city studies. This would be expected given the smoothing effect of fitting a model
38 over a large geographic area.
January 2015 3-7 DRAFT: Do Not Cite or Quote
-------
1 Selection of predictor variables, such as meteorology, traffic, land use, and population
2 density, influences the ability of the LUR model to predict concentrations of oxides of
3 nitrogen and depends on the specific city for which the model is fit. Su et al. (2008a) and
4 Ainslie et al. (2008) developed the Source Area-LUR (SA-LUR) to incorporate the
5 effects of meteorology on the model results. The SA-LUR integrates data for wind speed,
6 wind direction, and cloud cover variables in estimates for NO and NO2 and was found to
7 perform better when seasonal variability in concentrations was high. Su et al. (2008b)
8 included a street canyon aspect ratio as a LUR predictor variable to account for retention
9 of pollutants in street canyons. They observed that, upon adding the aspect ratio to the
10 LUR model, R2 increased from 0.56 to 0.67 for NO2 and from 0.72 to 0.85 for NO.
11 Similarly, when Clougherty et al. (2013) added "built space within 1 km" to their LUR
12 model of NO2, R2 increased by 0.41. Franklin et al. (2012) explored bivariate correlations
13 between NO2, NO, and NOx concentrations and several predictors reflecting traffic,
14 population, elevation, and land use in twelve southern California communities. For NO2,
15 Pearson correlations of concentration with distance to road were R = -0.42 and -0.35 for
16 freeway and nonfreeway roads, respectively, and produced an -8.2% change in
17 concentration per IQR increase in distance in the LUR model. Correlations with traffic
18 volume within a 300-m buffer were R = 0.41, and traffic volume within a 300-m buffer
19 produced a 2.4% change in the LUR prediction per IQR. Correlation with neighborhood
20 elevation was R = -0.50, and neighborhood elevation produced a -6.7% change in
21 LUR-modeled concentration per IQR increase in elevation. Su et al. (2009a) developed a
22 method to optimize the SA-LUR variable selection process in which correlations between
23 several land use variables and NO2 concentrations were computed across a 3-km buffer of
24 the NO2 measurement (1.5-km buffer for traffic-related variables), and the data for
25 correlation versus distance were fit to a curve describing that relationship. The variable
26 with highest correlation at the optimum buffer distance was added to the model if its
27 addition produced a statistically significant change (p < 0.1) in the model R2. Su et al.
28 (2009a) found the important variables to be distance from monitor, 24-hour traffic levels,
29 expressway casement, open land use, railway, major road, land grade, population density,
30 and distance to coast. It can be anticipated that the important variables might be different
31 depending on city-specific factors.
32 Several studies of LUR have explored temporal treatments in the model. LUR has been
33 evaluated across seasons and spatial variability in the NO2 concentration profile has not
34 been found to change substantially with season, despite more temporal variability during
35 mild weather compared with either cold or warm weather (Dons et al.. 2014; Grouse
36 et al.. 2009). Therefore, the authors concluded that an annual average would be
37 acceptable for LUR simulations. LUR models applied several years after model
38 development have demonstrated moderate-to-good predictive ability in a few studies.
39 Eeftens etal. (2011) compared LUR obtained from NO2 measurements at 35 locations in
January 2015 3-8 DRAFT: Do Not Cite or Quote
-------
1 the Netherlands over the years 1999-2000 with LUR developed from NO2 measurements
2 at 144 locations in the Netherlands during 2007. Both the NO2 measurements and the
3 LUR models agreed well for the two time periods studied (/? = 0.9998; R2 = 0.89). Wang
4 et al. (2013b) tested stability of an LUR model for Vancouver, Canada between 2003
5 (based on 116 sites) and 2010 (based on 116 sites, with 73 from the 2003 study). Wang
6 etal. (2013b) evaluated the model by testing how much variability in the measurements
7 was predicted by models from the other year, with moderate results. Linear regression for
8 comparison of the 2003 model with 2010 measurements produced R2 = 0.58-0.60 for NO
9 and R2 = 0.52-0.61 for NO2, while comparison of the 2010 model with 2003
10 measurements produced R2 = 0.50-0.55 for NO and R2 = 0.44-0.49 for NO2. Wang et al.
11 (2013b) attributed the diminished performance for the 2003 model using 2010 data
12 (compared with using the 2010 model for 2003 data) to reductions in NO and NO2
13 concentrations over the 7-year time period. Visual inspection of the NO and NO2
14 concentration maps from the Wang etal. (2013b) study suggests that changes in spatial
15 correlation over time may have contributed to reduced model performance in comparison
16 with the Eeftens etal. (2011) study.
17 LUR evaluation depends on the validation algorithm, model conditions, and basis for
18 validation (i.e., to what the modeling results are compared when computing R2). In a
19 recent study of LUR application in 20 European study areas, Wang etal. (2013a) found
20 that leave-one-out cross-validation (LOOCV), typically used to validate LUR, produced
21 higher R2 for NO2 compared with hold-out evaluation (HEV) (LOOCV: R2 = 0.83;
22 HEV: R2 = 0.52). LOOCV involves repeatedly withholding a fraction of the monitoring
23 sites from the fitting process for performance evaluation and then computing an ensemble
24 R2, whereas HEV entails prediction with the LUR at locations not fit by the model.
25 Therefore, HEV may provide a more conservative estimate of model fit. Mercer et al.
26 (2011) compared 10-fold cross-validated LUR with universal kriging (UK), in which a
27 surface of concentrations was built based on measured values, for three seasons in Los
28 Angeles with roughly 150 measurement sites. UK performance was slightly better than
29 LUR for all seasons with little difference in model performance among the seasons
30 (UK: R2 = 0.75, 0.72, and 0.74; LUR: R2 = 0.74, 0.60, 0.67). Li et al. (2012) developed a
31 new formulation for LUR using generalized additive models (GAM) and cokriging to
32 boost the performance of LUR over other LUR model variations of models and evaluated
33 it for Los Angeles, CA. GAM enables incorporation of localized nonlinear effects among
34 the prediction covariates, while cokriging is intended to improve spatial smoothing. The
35 LUR using GAM and cokriging, had the highest cross-validation (R2 = 0.88-92),
36 compared with universal kriging (R2 = 0.68-0.75) and multiple linear LUR
37 (R2 = 0.42-0.64).
January 2015 3-9 DRAFT: Do Not Cite or Quote
-------
1 LUR comparison with other models has produced variable results, in part because the
2 comparison data does not always have the same spatial resolution or account for the same
3 physical phenomena. Beelen etal. (2010) compared LUR with a dispersion model
4 incorporating a near-road module for modeling NC>2 concentrations in a Rotterdam,
5 Netherlands neighborhood. The dispersion model agreed better (Pearson R = 0.77)
6 compared with LUR (R = 0.47) with NC>2 measurements from 18 evaluation sites.
7 Dijkemaetal. (2011) also compared LUR for the city of Amsterdam and neighborhoods
8 therein with a dispersion model and found better agreement of the dispersion models with
9 observations for the city-wide model than for LUR (dispersion: R2 = 0.74;
10 LUR: R2 = 0.47) although agreement was comparable for the neighborhood specific
11 model (R2 = 0.57 for both models). Marshall et al. (2008) compared LUR with inverse
12 distance-weighted (IDW) spatial interpolation of NO and NO2 measurements, nearest NO
13 and NO2 measurements, and a Community Multiscale Air Quality (CMAQ) model run
14 for Vancouver, Canada. The LUR location was matched to each CMAQ grid cell centroid
15 and compared with the grid cell concentration. LUR and CMAQ produced similar
16 average absolute difference in the concentration compared with measured central site
17 concentrations for NO (LUR: 42%, CMAQ: 47%) and NO2 (LUR: 17%, CMAQ: 17%),
18 while nearest monitor and spatial interpolation methods produced less than 5% difference
19 for both pollutants and methods. However, it is important to recognize that these methods
20 were compared to a central site monitor, which cannot capture the spatial variability of
21 the NO2 concentration distribution. Specifically, IDW, central site monitoring of NO2
22 concentration, and nearest monitor NO2 concentration estimation approaches cannot
23 account for localized sources unless the sources are close to the monitors. Therefore,
24 agreement among the models does not necessarily signify accurate depiction of the
25 spatial distribution of NO2 concentration.
26 Recent studies have explored hybrid application of LUR and dispersion models. For
27 example, Wilton etal. (2010) included concentrations computed with the CALINE3
28 dispersion model in their LUR to estimate NOx and NO2 concentrations in Los Angeles
29 and Seattle. They observed modest improvements in model R2 (Los Angeles,
30 NOX: R2 = 0.71-0.74 vs. R2 = 0.53-0.55; Los Angeles, NO2: R2 = 0.79 vs. R2 = 0.74;
31 Seattle, NO2: R2 = 0.81 vs. R2 = 0.72) when CALINE3-computed concentration was
32 included as one variable along with land use, roadway length, and traffic density
33 variables. However, Lindstrom et al. (2013) applied LUR with CALINE3-computed NOx
34 concentration for Los Angeles participants in the Multi-Ethnic Study of Atherosclerosis
35 and found no appreciable improvement (R2 within ± 0.04) in model performance for a
36 variety of averaging times (daily "snap shot," 2-week, 10-year). Molter et al. (2010) also
37 used dispersion modeling data in lieu of measurement data when fitting an LUR for
38 Greater Manchester, U.K. and found reasonable agreement of NO2 predictions with
39 monitoring data (R2 = 0.86) and with a separate data set where 25% of the data were set
January 2015 3-10 DRAFT: Do Not Cite or Quote
-------
1 aside for evaluation (R2 = 0.62). Note that the nature of the monitoring data (i.e., central
2 site or other) was not explicitly stated in the Molter etal. (2010) study.
3 Although not used specifically for prediction at alternate locations, multiple linear
4 regression has been used to predict point concentrations based on meteorological and
5 source characteristics, in a manner similar to LUR but not including land use variables.
6 This technique is not employed for exposure assessment in any of the epidemiologic
7 studies cited in Chapter 6. but it is noted as an emerging exposure assessment method.
8 Vlachogianni et al. (2011) used multiple regression to forecast concentrations at two
9 locations each in Helsinki, Finland and Athens, Greece. They noted that the model
10 including only measured meteorological parameters (temperature, relative humidity, wind
11 speed, and wind direction) did not capture the true variability of the NO2 concentration
12 time series as well as a second model that also included atmospheric turbulence
13 parameters (Monin-Obukhov length and mixing height). This was particularly true during
14 cold weather. Carslaw and Taylor (2009) modeled NOx concentrations as a function of
15 meteorological and temporal variables. They found that wind speed, temperature,
16 temporal parameters (a trend metric, hour of the day, Julian day), and wind direction
17 were most important in the NOx model. Carslaw and Taylor (2009) highlighted that the
18 boosted regression tree methodology enabled examination of the influence of interactions
19 among the variables in addition to the direct dependence of NOx on each individual
20 variable.
Spatiotemporal Interpolation Modeling
21 Spatiotemporal modeling can also be used to describe NO2 concentrations for application
22 in exposure assessment in epidemiologic studies of long-term exposure. These methods
23 are not typically used in epidemiologic studies but can be considered emerging methods.
24 Le and Zidek (2006) developed an approach to use autoregressive integrated moving
25 average (ARIMA) models to capture the temporal pattern of the concentration field in
26 conjunction with an empirical Bayes maximum likelihood model to estimate the spatial
27 pattern of the true concentration field. Pollice and Jona lasinio (2010) applied a first-order
28 autoregression model to estimate the NO2 concentration field in Taranto, Italy and
29 observed reasonable agreement between observations and the predictions. Similarly,
30 ARIMA models can be used on their own at the location of a single model to forecast the
31 concentration time series at a specific location. Kumar and Jain (2009) found that
32 observations generally agreed with an ARIMA model of a point concentration within the
33 model's ±95% confidence interval (CI). Kumar and Jain (2009) tested several orders of
34 autoregression, integration, and moving average and found that different goodness-of-fit
35 measures favored selection of different autoregression and moving average orders.
36 However, Chaudhuri and Dutta (2014) tested different model orders and found that a
January 2015 3-11 DRAFT: Do Not Cite or Quote
-------
1 model with no autoregression but with second-order integration and moving average best
2 fit the NO2 data, with R2 = 0.83 and negative bias of 16%.
3 Artificial neural networks (ANN) provide another technique for predicting NO2
4 concentrations in a manner similar to LUR. In ANN, data are input through "nodes" in
5 the system, which are weighted based on different criteria to represent the influence of
6 various predictor variables. Nodes can reflect influential parameters such as meteorology
7 and source presence, and it is flexible to analyze data over space and/or time. They can
8 also be included as latent variables that aggregate the effects of the input variables
9 (Arhami et al.. 2013). In generally, ANN has been shown to validate well with
10 observation data (Baawain and Al-Serihi. 2014; Arhami et al.. 2013; Vlachogianni et al.,
11 2011; Konovalov et al.. 2010; Moustris et al.. 2009). Singh et al. (2012) tested variations
12 of ANNs, which vary based on the number of layers of nodes and the interactions
13 between layers in the model. Based on the model output of Singh et al. (2012). it was
14 found that suspended particulate matter concentration was the variable that contributed
15 the most in explaining variability in NO2 concentrations (with other variables for SO2
16 concentration, temperature, relative humidity, and wind speed). The Singh etal. (2012)
17 study illustrates how ANN may be used to gain insight into mechanistic processes
18 influencing NO2 concentration.
3.2.2.2 Mechanistic Models
Chemical Transport Models
19 CTMs can be used to develop estimates of human exposure to NO, NO2, or NOx. CTMs,
20 such as CMAQ, are deterministic of chemical transport that account for physical
21 processes including advection, dispersion, diffusion, gas-phase reaction, and mixing
22 while following the constraint of mass conservation (Byun and Schere. 2006). These
23 models provide regional concentration estimates and are typically run with surface grid
24 resolutions of 4, 12, or 36 km. Temporal resolution of CTMs can be as fine as 1 hour,
25 although larger temporal aggregation often occurs for the purpose of maintaining
26 reasonable data file size.
27 CTMs can be applied in epidemiologic studies of either short- or long-term exposure to
28 NO2 or NOx but are more commonly used in long-term exposure studies. These models
29 are used to compute interactions among atmospheric pollutants and their transformation
30 products, the production of secondary aerosols, the evolution of particle size distribution,
31 and transport and deposition of pollutants. CTMs are driven by emissions inventories for
32 primary species such as NO2, SO2, NHs, VOCs, and primary PM, and by meteorological
January 2015 3-12 DRAFT: Do Not Cite or Quote
-------
1 fields produced by other numerical prediction models. Values for meteorological state
2 variables such as winds and temperatures are taken from operational analyses,
3 re-analyses, or weather circulation models. In most cases, these are off-line
4 meteorological analyses, meaning that they are not modified by radiatively active species
5 generated by the air quality model. Work to integrate meteorology and chemistry was
6 done in the mid-1990s by Luetal. (1997a) and Lu et al. (1997b) and references therein,
7 although limits to computing power prevented their wide-spread application. More
8 recently, new, integrated models of meteorology and chemistry are now available as well;
9 see, for example, Binkowski et al. (2007) and the Weather Research and Forecast model
10 with chemistry (WRF Chem) (http://ruc.noaa.gov/wrfAVGl I/). Given observed biases in
11 the CTMs [e.g., U.S. EPA (2008)1. much attention has been given to bias correction of
12 these models for application in exposure assessment, as detailed below under the Hybrid
13 Models section below.
Dispersion Models
14 Dispersion models, or Gaussian plume models, estimate the transport and dispersion of
15 ambient air pollutants emanating from a point or line source through solution of an
16 equation that estimates the spread of the pollutant to follow a Gaussian curve that is a
17 function of distance from the source. Given that dispersion models typically capture
18 average concentrations, they are most commonly used in epidemiologic studies of
19 long-term exposure. Several studies of health effects related to NOx exposure employ
20 dispersion models to estimate NOx concentrations [e.g., Gruzieva et al. (2013).
21 McConnell et al. (2010). and Oftedal etal. (2009)1 because NO2 has high local spatial
22 variability (Section 2.5.3). The grid spacing in regional CTMs, usually between 1 and
23 12 km2, is too coarse to resolve spatial variations on the neighborhood scale. More finely
24 resolved spatial scales that better represent human exposure scales are provided by
25 local-scale dispersion models. Several dispersion models are available to simulate
26 concentration fields near roads, and each has its own set of strengths and weaknesses.
27 Several line source Gaussian dispersion models are available to simulate the dispersion of
28 emissions from a roadway. The California Department of Transportation developed the
29 CALINE model (http://www.dot.ca.gov/hq/env/air/software/caline4/calinesw.htm) for
30 this purpose. The CALINE family of models is not supported by the California
31 Department of Transportation for modeling of highway source NO2 and does not include
32 NOx transformation chemistry. Benson (1992) validated the CALINE3 and CALINE4
33 model versions using data from field studies at U.S. Highway 99 in Sacramento, CA and
34 a General Motors test track in Michigan. Benson (1992) found that more than 85% of
35 model predictions fell within a factor of two of measured observations for SFe (an inert
36 tracer gas). Among those that fell outside the factor of two envelope, 85% were positively
January 2015 3-13 DRAFT: Do Not Cite or Quote
-------
1 biased and mostly occurred when wind speeds were below 1 m/s. Additionally, the NO2
2 module of CALINE4 was tested by Benson (1992) under a limited set of conditions, and
3 it was recommended that CALINE4 not be used to predict NO2 dispersion under parallel
4 wind conditions without ample data to calibrate the model predictions.
5 The University of California, Davis (UCD) 2001 model was designed to improve upon
6 the design of CALINE by using an array of point sources to represent a three-dimensional
7 highway source of emissions and by using power law functions for wind speed and
8 vertical eddy diffusivity (Held et al.. 2003). UCD 2001 exhibited improved performance
9 for parallel, low speed winds (<0.5 m/s), with 87% and 83% reduction in error compared
10 with CALINE3 and CALINE4, respectively, for the General Motors SFe evaluation data
11 set. Snyderetal. (2013) recently released a Research Line-source (RLINE) dispersion
12 model that incorporates improved formulations of horizontal and vertical dispersion and
13 found that the predictions were within a factor of two of the observations for neutral,
14 convective, and weakly stable atmospheric conditions, but negative bias was observed for
15 stable conditions based on a line source SF6 experiment in Idaho Falls, ID. During
16 comparison with the U.S. 99 data set, 81% of data were within a factor of two for
17 downwind measurements, but only 19% for upwind measurements when winds were
18 within 30° of perpendicular to the road; 75% of downwind predictions were within a
19 factor of two of observations when winds were less than 1.5 m/s, and 88% were within a
20 factor of two for wind speeds greater than 1.5 m/s. Only 51% were within a factor of two
21 when winds were within 30° of parallel to the road. Additionally, a hybrid optimization
22 model fitting CALINE3 line-dispersion calculations for concentration to observations of
23 NO2 was developed and applied in the greater Tel Aviv, Israel area (Yuval etal., 2013).
24 Cross-validation was reported to have negligible bias in the model predictions with 36%
25 error; note that the authors did not clearly distinguish bias and error in this manuscript.
26 The American Meteorological Society/Environmental Protection Agency Regulatory
27 Model (AERMOD; http://www.epa.gov/scramOOl/dispersion_prefrec.htm) is a steady
28 state point source plume model formulated as a replacement to the Industrial Source
29 Complex (ISC3) dispersion model (Cimorelli et al.. 2005). In the stable boundary layer, it
30 assumes the concentration distribution to be Gaussian in both the vertical and horizontal
31 dimensions. In the convective boundary layer, the horizontal distribution is also assumed
32 to be Gaussian, but the vertical distribution is described with a bi-Gaussian probability
33 density function. AERMOD has provisions that can be applied to flat and complex terrain
34 and multiple source types (including point, area, and volume sources) in both urban and
35 rural areas. It incorporates air dispersion based on the structure of turbulence in the
36 planetary boundary layer and scaling concepts and is meant to treat surface and elevated
37 sources, in both simple and complex terrain in rural and urban areas. The dispersion of
38 emissions from line sources like highways in AERMOD is handled as a source with
January 2015 3-14 DRAFT: Do Not Cite or Quote
-------
1 dimensions set using an area or volume source algorithm in the model; however, actual
2 emissions usually are not in a steady state.
3 Most simple dispersion models including AERMOD are designed without explicit
4 chemical mechanisms but have nondefault options to estimate conversion of NO to NO2
5 based on a NOx/Os titration model. Hendrick et al. (2013) evaluated two modules used
6 with AERMOD to compute NO2 concentrations: the plume volume molar ratio method
7 (PVMRM) and the ozone limiting method (OLM). Both methods assume ratios of
8 NO2:NOx that are based on the concentration of co-occuring Os. Hendrick et al. (2013)
9 validated the models against more than 12 months of hourly observations taken near a
10 small power plant in Wainwright, AK, and they observed that the PVMRM overpredicted
11 NO2 at low concentrations and underpredicted at high concentrations, although the
12 average bias was small; the OLM also overpredicted NO2 concentrations at high observed
13 NO2.
14 AERMOD results have been compared with measurements and other models to evaluate
15 relative performance. Gibson et al. (2013) found poor agreement with respect to
16 magnitude of NOx concentrations and correlations (R2 = 0.001-0.003) at hourly,
17 monthly, and annual timescales when comparing AERMOD results with observations in
18 Halifax, Canada where several industrial facilities emit NOx. Cohan et al. (2011)
19 compared AERMOD output with 24-hour central site monitoring observations averaged
20 over August 2005 from San Jose, CA, where there are combined emissions from a port,
21 rail yard, and roadways. They observed that the AERMOD model consistently
22 underpredicted the observations; negative bias was more pronounced for simulations
23 from January compared with August. Misraetal. (2013) compared AERMOD with the
24 Quick Urban and Industrial Complex (QUIC) model. QUIC approximates average
25 airflow around buildings in urban environments then models pollution parcels based on
26 Lagrangian particle dispersion. In this case, AERMOD underpredicted NOx
27 concentrations in an urban street canyon, while most QUIC predictions were within a
28 factor of two of the observed NOx concentrations.
29 There are also nonsteady state models for different types of sources. For example,
30 CALPUFF (http://www.src.com/calpuff/calpuffl.htm). which is EPA's recommended
31 dispersion model for transport in ranges >50 km, is a nonsteady-state puff dispersion
32 model that simulates the effects of time- and space-varying meteorological conditions on
33 pollution transport, transformation, and removal and has provisions for calculating
34 dispersion from surface sources (U.S. EPA. 1995). However, CALPUFF was not
35 designed to treat the dispersion of emissions from roads, and like AERMOD has some
36 limited chemistry options to estimate production of secondary pollutants. The distinction
37 between a steady-state and time-varying model may not be important for studying health
January 2015 3-15 DRAFT: Do Not Cite or Quote
-------
1 effects where long-exposure timescales are relevant; however, when short-exposure
2 timescales are of interest (e.g., 1-hour), it would be more important to approximate the
3 short-term variability in concentrations. CALPUFF was validated against SF6 data at two
4 military test sites in Nevada (Chang et al.. 2003). where it was shown that 53% of
5 CALPUFF predictions were within a factor of two under SFe observations for one site
6 and 29% of predictions were within a factor two under the observations at a second site;
7 the second test site had surrounding mountains acting to increase vertical dispersion for
8 which CALPUFF did not account well. Cui et al. (2011) evaluated CALPUFF by
9 releasing SFe from a weather tower at the bank of the Gan Jiang River in China, which
10 has a combination of open field, agricultural land, and forest. CALPUFF was found to be
11 negatively biased with only 25-27% of data within a factor of two of observations. The
12 authors concluded that CALPUFF did not predict hourly dispersion well. Similarly,
13 Ghannam and El-Fadel (2013) compared NCh concentrations calculated using CALPUFF
14 with NO2 measurements and observed that the model severely underpredicted the
15 measurements, sometimes by up to three orders of magnitude, but was stated to have
16 captured the temporal variability, although correlations were not reported. Ghannam and
17 El-Fadel (2013) attributed this underprediction to underestimation of the emissions input
18 to the model. The results of Cui et al. (2011) and Ghannam and El-Fadel (2013)
19 indicating negative bias are consistent with those of Chang et al. (2003) for the site where
20 vertical dispersion may have played a larger role in the airflow characteristics.
21 An example of where AERMOD has been used in better understanding the relationship
22 between ambient concentrations and health risks is found in Maantay et al. (2009). These
23 researchers coupled AERMOD with geographic information system proximity buffers
24 around a stationary point source in Bronx, NY. They observed that buffers based on the
25 predicted plume shape to set levels of human exposure to NOx, along with PMio, PIVb 5,
26 CO, and SO2, corresponded better with asthma hospitalization rates compared with
27 circular buffers centered around the emissions source.
Hybrid Models
28 Substantial uncertainties at the subgrid scale remain when using CTM to model
29 concentrations at resolutions of 4-36 km (U.S. EPA. 2008). In densely populated regions
30 of the country, monitor density may be finer than CTM surface grid resolution.
31 Moreover, CMAQ and other CTMs suffer from pollutant-specific concentration biases,
32 such as underestimation of total nitrate that require correction (Fuentes and Raftery,
33 2005) prior to interpretation for exposure assessment. Bayesian Maximum Entropy
34 models for merging CMAQ and concentration data (Fuentes and Raftery. 2005) and
35 downscaling (Berrocal et al.. 2010a. b) have recently been developed to improve spatial
36 resolution and provide bias correction for the modeled concentration used as an exposure
January 2015 3-16 DRAFT: Do Not Cite or Quote
-------
1 surrogate, but such methods must be used with caution. For instance, Chen etal. (2014)
2 ran a 36-km resolution CMAQ for NO2, NOx, and other copollutants, fused the CMAQ
3 results with monitor observations, and compared both the raw and fused model results
4 with monitor observation data. The raw CMAQ simulations overpredicted NO2 and NOx
5 concentrations, particularly in the winter. These overpredictions were substantially
6 reduced (and in some cases the model slightly underpredicted concentrations) for the
7 fused model. Isakov et al. (2009) modeled subgrid spatial variability within CMAQ using
8 the AERMOD dispersion model prior to linking the modeled results with stochastic
9 population exposure models to predict annual and seasonal variation in urban population
10 exposure within urban microenvironments. In each case, these papers have referred to
11 other air pollutants, but the methodology is still applicable for NO2 exposure prediction.
12 Berrocal et al. (201 Ob) proposed a downscaling approach to combine monitoring and
13 CMAQ modeling data to improve the accuracy of spatially resolved modeling ozone
14 data. Specifically, a Bayesian model is developed to regress CMAQ model estimates on
15 monitoring data, and then the regression model is used to predict concentrations using the
16 CMAQ model results as an input field. Berrocal et al. (2010a) extended the approach to
17 include two pollutants (ozone and PM2 5) in a single modeling framework, and Berrocal
18 etal. (2012) added smoothing processes that incorporate spatial autocorrelation and
19 correction for spatial misalignment between monitoring and modeled data. Although
20 these papers did not specifically use NO2 data, the methods can be applied for NO2 as
21 they have been for Os and PM2 5. Bentayeb et al. (2014) applied a similar data
22 assimilation method in which local measurements and elevation data were combined with
23 CTM output in a geostatistical forecasting model. This algorithm was applied for NO2 as
24 well as PMio, PM2 5, SO2, CeHe, and Os. Correlations between assimilated values and
25 measurements ranged between Pearson R = 0.75-0.90. Debry and Mallet (2014) also
26 employed data assimilation for forecasting but combines three CTMs in an ensemble
27 average to minimize the influence of their errors in conjunction with assimilation of
28 observation data. The method of Debry and Mallet (2014) reduced error in hourly, daily,
29 and peak NO2 concentrations by 19, 26, and 20%, respectively.
30 In a slightly different approach, Crooks and Isakov (2013) blended CMAQ, AERMOD,
31 and monitoring data for NOx, PM2 5, and CO using a Bayesian model based on a wavelet
32 basis series. In this method, the true exposure is represented by the B-spline wavelet
33 series, and then the CMAQ grid cell concentrations, AERMOD receptor concentrations,
34 and measurement points are represented by the wavelet field modified by some assumed
35 error. These components each comprise linear contributions to a Gaussian likelihood
36 model. For NOx, the model was found to favor CMAQ data when modeling background
37 and monitor data in dense urban areas where spatial variability is higher. The blended
38 model results had lower prediction error and bias compared with kriging when smaller
January 2015 3-17 DRAFT: Do Not Cite or Quote
-------
1 numbers of points were used for the kriging surface, although the blended model did not
2 perform as well as kriging when densely gridded data were available for that purpose.
3 Similarly, Robinson et al. (2013) used geographically weighted regression, which used a
4 combination of dispersion model results and monitoring data as input for a regression
5 model to compute concentrations in local population centers and then used kriging to fill
6 in gaps between those population centers. When compared with other kriging methods,
7 the geographically weighted regression approach produced the smallest residual mean
8 squared errors when modeling average NO2 concentrations across the U.K. for the year
9 2004. Beevers et al. (2012) also blended CMAQ with a near-road dispersion model and
10 applied the blended model for estimation of human exposure to NOx in London, U.K.
11 (Beevers etal., 2013). Predicted peak rush hour (0600-0900) NOx exposures exceeded
12 observed NOx concentrations by roughly 25% at a heavily trafficked road but performed
13 belter when averaged over multiple sites.
Nondimensional Scale Models
14 Although nondimensional scale models are not currently used for exposure assessment in
15 epidemiologic studies, they are described briefly here as emerging methods for potential
16 use in exposure assessment. Existing wind tunnel and observational data have been used
17 in nondimensional scale models of wind movement that support NOx fate and transport
18 modeling in the presence of built structures. For example, the Ausbreitungsmodell gemaB
19 der Technischen Anleitung zur Reinhaltung der Luft (AUSTAL2000) airflow model has
20 been developed using a combination of wind speed and direction, mixing layer height,
21 and stability classifications. Air pollutant transport is then modeled using a Lagrangian
22 dispersion model with a random walk to simulate the influence of turbulence on the air
23 pollution parcels' movement. Langner and Klemm (2011) compared AUSTAL2000 to
24 the AERMOD dispersion model using five test cases with varied topography and building
25 presence for which experimental field data existed. Although these runs simulated SO2
26 and SF6 transport, the model performance is instructive for analysis of NOx transport as
27 well. In every case, AERMOD performed better than AUSTAL2000 in capturing the
28 observed concentrations.
29 The Operational Street Pollution Model also uses nondimensional scale modeling but is
30 developed specifically to capture street canyon recirculation. Berkowicz et al. (2008)
31 developed a model that includes a turbulent mixing velocity in the street canyon and free
32 convection. Monthly and 6-mo avg NO2 concentrations were calculated using a turbulent
33 plume model. Modeled concentrations were compared with NO2 measurements from a
34 1995 panel study and found to agree reasonably well (6-12% negative bias;
35 R2 = 0.75-0.81).
January 2015 3-18 DRAFT: Do Not Cite or Quote
-------
3.2.2.3 Stochastic Exposure Models
1 Although they are not typically used for exposure assessment in epidemiologic modeling,
2 stochastic exposure models inform the risk assessment performed as part of the national
3 ambient air quality standard review process. The state of the science for stochastic
4 population exposure models has not changed substantially since the 2008 ISA for Oxides
5 of Nitrogen, as described in detail in 2008 Annex 3.6 (U.S. EPA. 2008). Examples of
6 stochastic population exposure models include Air Pollution Exposure (APEX),
7 Stochastic Human Exposure and Dose Simulation (SHEDS), and exposure in polis (or
8 cities) (EXPOLIS), which involve stochastic treatment of the model input factors (Kruize
9 et al.. 2003; Burke et al.. 2001). Advancement in exposure modeling has come from its
10 integration with chemical transport models of outdoor air quality through a hybrid
11 approach (Isakov et al.. 2009) and characterization of the uncertainty in these models
12 (Ozkavnak et al.. 2009; Zidek et al.. 2007).
13 Hybrid exposure modeling uses ambient air quality input from grid-based models rather
14 than from central site monitoring data, as is typically done (Isakov et al.. 2009). In the
15 hybrid version, the CMAQ model is used to simulate concentrations for a coarse discrete
16 grid, e.g., 12 km x 12 km. Next, local scale concentrations from point and mobile sources
17 are estimated using Gaussian dispersion modeling through AERMOD. In combination,
18 these models produce an ambient air quality estimate at the location of the receptor that is
19 then input into APEX or SHEDS to estimate total human exposure. Isakov et al. (2009)
20 observed that the omission of specific point and traffic sources led to an underestimate in
21 median concentration by up to a factor of two, depending on location; these simulations
22 were for benzene and PM2.5; NOx tends to be comparable in spatial variability compared
23 with benzene and more spatially variable compared with PM2 5 (Beckerman et al., 2008).
24 Recent studies have considered the variability and uncertainty associated with exposure
25 modeling. Ozkavnak et al. (2009) considered uncertainty and variability in simulations
26 involving estimation of concentration, exposure, and dose in separate compartments of a
27 model. They found that uncertainty and variability propagated from one compartment to
28 the next. Zidek et al. (2007) addressed uncertainty and variability in exposure modeling
29 by using distributions of input parameters in the exposure model framework rather than
30 point estimates. These models estimate time-weighted exposure for modeled individuals
31 by summing exposure in each microenvironment visited during the exposure period.
32 Zidek et al. (2007) found that use of distributions of input enabled examination of cases
33 for potential subpopulations with common characteristics. Note that both of these studies
34 model PM, but the findings are applicable to NOx.
35 Sarnatet al. (2013) recently compared risks of cardiovascular and respiratory morbidity
36 with 24-hour NOx and other primary and secondary air pollutants in Atlanta using
January 2015 3-19 DRAFT: Do Not Cite or Quote
-------
1 various exposure metrics. Epidemiologic results based on the mean, median, and 95th
2 percentile of the exposure distributions from APEX were compared with measures from a
3 central site monitor, regional background, AERMOD, and a hybrid model merging
4 AERMOD output with regional background data. NOx concentrations modeled with
5 APEX were generally higher than those obtained with the hybrid model, likely because
6 the APEX model incorporates road activity levels in their exposure estimates.
7 Epidemiologic analyses for asthma/wheeze produced statistically significantly higher risk
8 ratios for the APEX mean, median, and 95th percentile compared with the hybrid model
9 and central site and background metrics but a negligible difference among the APEX and
10 hybrid results for respiratory or cardiovascular diseases.
3.2.3 Choice of Exposure Metrics in Epidemiologic Studies
11 Appropriateness of the exposure metric for a given study depends in part on
12 epidemiologic study design and spatial variability of the pollutant. Table 3-1 summarizes
13 the methods described in Sections 3.2.1 and 3.2.2. Based on epidemiologic studies using
14 various methods for exposure assessment, Figure 3-1 illustrates the range of NC>2
15 concentrations to which people may be exposed in different locations (HEI. 2010).
16 Because this figure is the result of the HEI (2010) review, the data points included were
17 obtained based on varying temporal scales. The figure illustrates variability in exposures
18 across locations and also the variability measured within a type of location. Given the
19 natural variability of exposures over space and time and nuances of the specific exposure
20 assessment techniques, it is important to be cognizant of the specific applicability and
21 limitations of each approach, as summarized in Table 3-1.
January 2015 3-20 DRAFT: Do Not Cite or Quote
-------
Table 3-1 Summary of sampling methods, their typical use in epidemiologic
studies, and related errors and uncertainties.
Method
Epidemiologic Application Errors and Uncertainties in Exposure Estimates
Central site monitors
Short-term community
time-series exposure of a
population within a city
Correlation between exposure and measurement
decreases with increasing distance from the monitor
(Section 3.4.5)
Long-term exposure for
comparison of populations
among different cities
Potential for exposure misclassification if the monitor
site does not correspond to the exposed population
(Section 3.4.5)
Positive instrument bias (Section 3.2.1.1)
Passive monitors
Short-term panel
Positive instrument bias (Section 3.2.1.2)
Long-term exposure across a
city (or for LUR model fit)
Positive instrument bias (Section 3.2.1)
Potential for exposure misclassification if the monitors
are sited at fixed locations (Section 3.4.5)
LUR
Long-term exposure, usually
across a city but sometimes fit
among multiple cities
Potential for exposure misclassification if grid is not
finely resolved (Section 3.2.2.1)
Potential for bias if the model is misspecified or
applied to a location different from where the model
was fit (Section 3.4.5)
IDW
Long-term exposure across a
city
Potential for negative bias if sources are not captured
or overly smoothed (Section 3.2.2.1)
Spatiotemporal modeling Not reported
Not yet well understood (Section 3.2.2.1)
CTM
Long-term exposure, sometimes
within a city but more typically
across a larger region
Potential for exposure misclassification when grid
cells are too large to capture spatial variability of
exposures (Section 3.2.2.2)
Dispersion modeling
Long-term exposure within a
city
Potential for bias where the dispersion model does
not capture boundary conditions and resulting fluid
dynamics well, e.g., in large cities with urban
topography affecting dispersion (Section 3.2.2.2)
Parameterization
modeling
Not reported
Not yet well understood (Section 3.2.2.2)
LUR = land use regression; IDW = inverse distance-weighted; CTM = chemical transport model.
January 2015
3-21
DRAFT: Do Not Cite or Quote
-------
1 V/VV
%
B>
c
O 1 nn
._ I UU
2
1
c
O
O-in
i u
CO
0)
1
i I
- - 1- -
!*
•
i.
t
*
Sampling time
A < 24 hr
• > 24 hr
* #
A *
* *
* I •
* * I
, .« f. ,
9
Source: HEI (2010)
Figure 3-1
Nitrogen Dioxide
Average nitrogen dioxide concentrations measured in studies
using different monitor siting.
i
2
3
4
5
6
7
8
9
10
11
12
13
Concentrations measured by central site or near-road monitoring are commonly used as a
surrogate for human exposure, and they can be used in studies of both short-term and
long-term exposure to NC>2 (Section 3.2.1.1). Central site measurements are subject to
positive bias from instrument error. For epidemiologic studies of short-term exposure,
correlation between measured central site concentration and concentration at some distant
point decreases with distance. For epidemiologic studies of long-term exposure to NO2,
the difference between the measured concentration and the true exposure would result in
exposure misclassification. Passive sampling techniques such as Palmes tube
measurements are subject to positive instrumentation biases. Additionally, passive
monitors left in place for sampling durations of days to weeks may produce data having
errors and uncertainties that are similar to those associated with using a fixed-site monitor
to capture exposures for a population that is dispersed over space and moving in time.
The influence of exposure error in passive sampling methods is discussed in more detail
January 2015
3-22
DRAFT: Do Not Cite or Quote
-------
1 in Sections 3.4.5.2 and 3.4.5.3. Passive sampling can be used for panel studies, or when
2 samples are integrated over a month or averaged over several months, as input for
3 long-term studies (Section 3.2.1.2). The integrated nature of the passive samples limits
4 their application in time-resolved studies. Passively sampled concentrations are also used
5 commonly as input for LUR model fitting (Section 3.2.2.1).
6 LUR is generally thought to illustrate spatial variability of NC>2 exposures well for use in
7 long-term exposure studies. The quality of the exposure metric provided by the model
8 depends on several factors, including spatial resolution of the model, representativeness
9 of the model fit locations for the city and population under study, and inclusion of the
10 right variables in fitting the model. IDW is also used for exposure estimation between
11 spatially distributed passive NO2 measurements (Section 3.2.2.1). However, if the
12 monitors are not dense enough to capture the true spatial variability of NC>2 related to
13 localized sources, exposure is likely to be underestimated.
14 CTMs and dispersion models are based on physics of air flow and contaminant transport
15 (Section 3.2.2.2). Like central site monitors, CTM can be used to compare NC>2 exposures
16 among different cities for long-term exposure studies. However, coarse spatial resolution
17 of CTMs limits its applicability within cities. Dispersion models are frequently used for
18 within-city NC>2 exposure estimation in long-term exposure studies, but the simplifying
19 assumption of Gaussian dispersion can add error to the exposure estimate if meteorology
20 or topography of the built environment are complex. Given this complexity, the direction
21 of exposure error is not predictable. Biases in dispersion model output can occur in either
22 direction, and they depend strongly on the specific environment (i.e., topography,
23 meteorology, source representation) being modeled. Correction methods may sometimes
24 be applied to minimize such error for a given location, but the effectiveness of error
25 minimization must be determined on a case-by-case basis. Subsequent sections will
26 describe characterization of NC>2 exposures, a conceptual model of exposure,
27 relationships among exposure metrics, sources of exposure error, confounding, and
28 implications of exposure error for epidemiologic studies of different designs.
January 2015 3-23 DRAFT: Do Not Cite or Quote
-------
3.3 Characterization of Nitrogen Dioxide Exposures
3.3.1 Nitrogen Dioxide Concentration as an Indicator of Source-based Mixtures
3.3.1.1 Mobile Source Emissions
1 Seventeen percent of U.S. homes are located within 91m of a highway with four or more
2 lanes, a railroad, or an airport (U.S. Census Bureau, 2009). Moreover, 7% of U.S. schools
3 serving 3,152,000 school children are located within 100 m of a major roadway, and 15%
4 of U.S. schools serving 6,357,000 school children are located within 250 m of a major
5 roadway (not specifically defined in this study in terms of annual average daily traffic
6 [AADT], number of lanes, or other criteria) based on data from the National Center for
7 Education Statistics (Kingsley et al.. 2014). Average one-way commuting times for the
8 U.S. labor force working outside the home are 19.3 minutes for bicyclists, 11.5 minutes
9 for walkers, and 25.9 minutes for all other modes of transportation. Among the populace
10 working outside the home, 15.6% spend 45 minutes or more commuting to work each
11 day (U.S. Census Bureau. 2007). Based on Figure 2-4. the proportion of NOx emissions
12 from mobile sources in the 21 CBSAs with at least 2.5 million residents is 16% higher
13 than it is among the general population. Hence, a large share of the U.S. population is
14 exposed to the on- and near-road environment on a regular basis, and those exposures are
15 likely to be higher for the 38% of the population living in urban areas (U.S. Census
16 Bureau. 2013). This has implications for potential NO2 exposure. Section 2.5.3 describes
17 spatial patterns of NO2 concentrations near roads as a background for understanding
18 traffic-related NO2 exposure. This section builds on the observations of NO2
19 concentration gradients described in Chapter 2 to consider how near-road concentrations
20 influence traffic-related NO2 and NOx exposure.
21 Time spent in traffic can be an important determinant of personal NO2 exposure. Molter
22 et al. (2012) calculated associations with time spent in several home, transit, and school
23 microenvironments for a cohort of 12-13 year-old children from Greater Manchester,
24 U.K. based on 2-day sampling periods per season and observed that time spent in transit
25 was positively associated with both NO2 exposure and mean prediction error of a
26 microenvironmental model of personal NO2 exposure, where mean prediction error
27 compares the microenvironmental model with NO2 measurements. Together, these
28 findings suggest that exposures are higher on roads and consequently that time spent in
29 transit may comprise a larger share of daily NO2 exposure compared with the proportion
30 of time in a day that is spent in transit. Ragettli et al. (2014) estimated exposures among
31 commuters reporting to the 2010 Swiss Mobility and Transport Microcensus in Basel,
January 2015 3-24 DRAFT: Do Not Cite or Quote
-------
1 Switzerland based on the times and routes they reported for their commutes and modes of
2 transport and using concentration estimates from a combination of dispersion modeling
3 and LUR. Ragettli etal. (2014) found that travel in motor vehicles produced the highest
4 exposure (reported as the product of concentration and time), followed closely by
5 bicyclists and those taking public transit. Pedestrians had measurably lower exposures.
6 Health studies often focus on the independent effects of NO2 or use NO2 concentration as
7 a surrogate for exposure to traffic pollution mixtures when measurements of other
8 pollutants are not available. NO2 concentration is routinely measured at sampling sites
9 nationwide, and NO2 is a prevalent reaction product of NO, which is a component of
10 vehicle exhaust (see Section 2.2). Section 3.4.4 concludes that NO2 concentration
11 generally correlates spatially with other traffic-related pollutants in urban areas. NO2
12 concentration has also been observed in at least one study to correlate with
13 nonconcentration measures of traffic. With respect to exposure, these observations make
14 it hard to distinguish NO2 from other pollutants when considering the health impacts
15 potentially attributable to each.
16 As a surrogate for traffic-related exposure, NO2 concentration may do an adequate job of
17 capturing spatial and temporal trends of traffic pollution. Microscale spatial variability of
18 NO2 concentrations near roads has been studied extensively, and NO2 concentration
19 gradients from a number of studies are summarized and compared in Section 2.5.3. Based
20 on 1-2 weeks of passive sampling measurements for NO2, Wheeler et al. (2008) and
21 Beckerman et al. (2008) reported correlations among NO2 and several traffic-related air
22 pollutants, including benzene (Pearson R = 0.85) and toluene (R = 0.63). The near-road
23 air pollutant gradients displayed in the review by Karner etal. (2010) suggested that NO2
24 is correlated with traffic-related air pollutants across various distances from a roadway.
25 These studies concluded that gradients in NO2 concentrations were spatially correlated
26 with gradients in traffic-related pollution.
27 The size and shape of the near-road gradient for NO2 determines the spatial zone where
28 near-road exposures are most likely. Observations of the structure of the NO2 near-road
29 concentration gradient are summarized in Table 2-6 in Section 2.5.3. Although NO2 tends
30 to correlate with most roadway pollutants in a near-road environment, the NO2
31 concentration gradient tends to be shallower than gradients for other primary
32 traffic-related pollutants (e.g., CO, UFP). These gradients influence how exposure and
33 copollutant correlations change spatially across the near-road environment. For example,
34 near road NO2 concentration is typically 30-100% higher than the urban background
35 concentration, defined here as the lowest concentration measured upwind of the road. In
36 contrast, peak near-road UFP counts are approximately 5-6 times higher than the urban
37 background concentration (Karner etal.. 2010). These results suggest that, although NO2
January 2015 3-25 DRAFT: Do Not Cite or Quote
-------
1
2
o
6
4
5
6
7
8
9
10
11
may capture many aspects of pollutant gradients from the roadway, NC>2 concentration
used as a marker for traffic may underestimate the magnitude of the concentration
gradient for other near-road pollutants, such as UFP and CO. Figure 3-2 presents the
spatial variability of NO2 and copollutants at various gradients from the roadway reported
in Karner et al. (2010) to compare the spatial near-road gradient of NO2, NO, and NOx
concentrations with those of other traffic-related pollutants (Beckerman et al.. 2008). The
review of Karner et al. (2010) showed that the NO2 gradient was much less steep
compared with the gradients for NO and NOx, with decay to background levels within
400-500 m. In a later study of near-road concentrations in Medford, MA, Padro-Martinez
etal. (2012) used continuous instrumentation mounted on a mobile sampling unit
operated over the course of a year, to illustrate a similar gradient.
0 100 200 300 400
_i i i i i_
I 1
Ą
0 , n.
•*• 1.2 -
| 1Q-
I o.a -
B
o 0.6 -
'-a
% nj
1= 0.4 -
3
g 02-
(J
J ° ° "
Rapid: >50% drop by 150 m
,.ccKsyi
- EC<4B]
Uf\ t&7\
I NO* (30)
UF1 Particle no (76)
t\ UF2 Partitl* r». (93)
i\V VOC1 (80)
Vi'\\
VV^X --•---:
'. NO*- —
"" • r • " "
Less rapid v gradual decay
- NOjCiaS)
pi* HfiHi
rM2.5^o1J
\ *.
"* -X^- ^
No trend
•• i*-^ =^ "^ ^.-**^*
*••***
rniig(Dfj
. . vnf? iwi
100 200 300 400
100 200 300 400
Distance from edqfi (m)
Note: NO2, NO, and NOX concentration gradients are presented in the center panel. NO2 = nitrogen dioxide, NO = nitric oxide,
NOx = sum of NO2 and NO, CO = carbon monoxide, PM25 = in general terms, participate matter with an aerodynamic diameter less
than or equal to a nominal 2.5 |jm, a measure of fine particles, PMio = in general terms, particulate matter with an aerodynamic
diameter less than or equal to a nominal 10 |jm, a measure of thoracic particles, EC = elemental carbon, VOC = volatile organic
compound, UF = ultrafine.
Data presented from Karner et al. (2010) were synthesized from 41 peer-reviewed references, 11 of which reported data for NO2, 5
of which reported data for NO, and 6 of which reported data for NOX. The number in parentheses refers to regression sample size.
UF1 and UF2 are measures of ultrafine particle number.
Source: Reprinted with permission of the American Chemical Society, Karner et al. (2010).
Figure 3-2 Spatial variability in concentrations of near-road pollutants,
including NO2, NO, NOx, CO, PM2.5, PMio, EC, benzene, VOCs, and
UF Particles. Concentrations are normalized by measurements at
the edge of the road.
12
13
14
As also pointed out in Section 2.5.2. near road NO concentrations are typically much
higher than near road NO2 concentrations; Table 2-6 describes the near road
concentration gradient for NO2 only. Table 3-2 expands on these observations to consider
January 2015
3-26
DRAFT: Do Not Cite or Quote
-------
1 on-road exposure to NO2 and NO while in transit. Recent on-road and near-road
2 measurements of both NO and NO2 concentrations indicate that on-road NO exposures
3 can be much higher than on-road NO2 exposures immediately upon their emission. In
4 particular, the Los Angeles data for NOx and NO suggest that rush hour NO2
5 concentrations are roughly 50-60 ppb, but NO concentrations reach roughly
6 200-360 ppb in the morning and 95-260 ppb in the afternoon, based on 2-h avg of
7 1-minute data (Fujita et al.. 2011). Beckerman et al. (2008) measured 1-week integrated
8 NO and NO2 samples next to two highways in Toronto, Canada and also observed that
9 mean NO levels were 3-4 times higher than mean NO2 levels.
10 The relationship between NO2 concentration and traffic metrics informs exposure
11 assessment because it establishes potential for exposure among those commuting or
12 living in the near-road environment. In Minneapolis, MN, Pratt et al. (2014) compared
13 direct traffic metrics, such as traffic volume, with LUR-computed NO2 concentrations
14 (which were not estimated from traffic volume although road length was included in the
15 model). They observed a correlation (type unstated) of 0.58 between NO2 concentration
16 and traffic density (AADT per km2), with a slope of 0.103 on a log-log model of NO2
17 versus traffic density. Gauderman et al. (2005) measured the correlation between NO2
18 concentrations and various traffic metrics in 12 Southern California communities. On
19 average across the communities, the Spearman correlation between NO2 concentration
20 and increasing distance to freeway was r = -0.54, but the correlation between NO2
21 concentration and traffic volume within 150m of a freeway was r = 0.24. The
22 contribution of mobile source emissions to NO2 concentration varies with strength of
23 additional sources. For example, Ducret-Stich et al. (2013) modeled NO2 concentration as
24 a function of background NO2 levels; light duty and heavy duty traffic counts; and
25 meteorological, topographic, and temporal variability in the Swiss Alps with a model
26 R2 = 0.91. They observed that background NO2 concentration contributed 83% of the
27 variability in the model, while heavy duty and light duty traffic counts contributed 8 and
28 7%, respectively. Similarly, NOx has been found to have mixed correlation with traffic
29 density in a nationwide long-term exposure epidemiologic study of the U.S. Veterans
30 Cohort [ 1976-2001 (Lipfert et al.. 2009)]. In areas deemed high traffic density (higher
31 traffic than the average 1985 traffic density), Pearson R = 0.27, while for areas of low
32 traffic density (lower traffic than the average 1985 traffic density), R = 0.56.
January 2015 3-27 DRAFT: Do Not Cite or Quote
-------
Table 3-2 Near- and on-road measurements of nitrogen dioxide (NO2), nitric oxide (NO), and the sum of
NO2 (NOx).
Reference Location and Date
Beckerman et al. Toronto site 1 , August 2004
(2008)
Toronto site 2, August 2004
Zhu et al. (2008) Los Anqeles 1-710 (mostly diesel
trucks), NR
Los Angeles I-405 (mostly autos),
NR
Fuiita et al. (2011) Los Angeles 1-110 (mostly autos),
Sep-Dec 2004
Los Angeles I-405 (mostly autos),
Sep-Dec 2004
Los Angeles SR-60 (mostly autos),
Sep-Dec 2004
Los Angeles truck route, Sep-Dec
2004
Los Angeles 1-1 1 0 (mostly autos),
Sep-Dec 2004
Los Angeles I-405 (mostly autos),
Sep-Dec 2004
Distance to Road (m)a
28,47, 57, 107, 126, 194,
209, 382, 507, 742, 986
4,28,38, 56, 105, 114, 175,
246, 335, 346, 438, 742, 875
0
0
0
0
0
0
0
0
Averaging Time
1-week integrated
1-week integrated
2-h avg of 1-min data
unfiltered
2-h avg of 1-min data
unfiltered
2-h avg of 1-min data
(morning)
2-h avg of 1-min data
(morning)
2-h avg of 1-min data
(morning)
2-h avg of 1-min data
(morning)
2-h avg of 1-min data
(afternoon)
2-h avg of 1-min data
(afternoon)
NO (ppb)
44.2(19.9)b;
77.6C
70.5 (62.7)b;
239.3C
NR
NR
347 (235)b
198(94)b
329(114)b
361 (143)b
95 (49)b
98 (56)b
NO2 (ppb)
14.6(2.8)b;
18.6C
17.5(4.6)b;
28.2C
NR
NR
NR
NR
NR
NR
NR
NR
NO and
NOx (ppb)
NR
NR
432 (0.9)b
267(114)b
411 (250)b
245(100)b
388(120)b
426(154)b
148(62)b
140(64)b
January 2015
3-28
DRAFT: Do Not Cite or Quote
-------
Table 3-2 (Continued): Near- and on-road measurements of nitrogen dioxide (NO2), nitric oxide (NO), and the sum
of NO and NO2 (NOx).
Reference Location and Date Distance to Road (m)a
Los Angeles SR-60 (mostly autos), 0
Sep-Dec 2004
Los Angeles truck route, Sep-Dec 0
2004
Fruin et al. (2008) Los Anqeles 1-10 (mostly autos), 0
Feb-April 2003
Los Angeles 1-710 (mostly diesel 0
trucks), Feb-April 2003
MacNauqhton et al. Boston bike path separate from 0
(2014) vehicle traffic, NR
Boston bike lane adjacent to 0
vehicle traffic, NR
Boston designated bike lane 0
shared between bikes and buses,
NR
Averaging Time NO (ppb)
2-h avg of 1 -min data 1 1 2 (55)b
(afternoon)
2-h avg of 1-min data 258 (1 14)b
(afternoon)
2-to-4-havg of20-sdata 280d
2-to-4-h avg of 20-s data 390d
Average over 40 3-h NR
sampling periods with 1-min
data
Average over 40 3-h NR
sampling periods with 1-min
data
Average over 40 3-h NR
sampling periods with 1-min
data
N02 (ppb) NOx (ppb)
NR 170(65)b
NR 321 (125)b
NR NR
NR NR
14.7 NR
(0.582)b
19.5 NR
(0.343)b
24.2(1.72)b NR
avg = average; NR = not reported.
aDistance of 0 m indicates on-road measurements.
bAverage (standard deviation).
°Maximum.
dAverage of medians.
January 2015
3-29
DRAFT: Do Not Cite or Quote
-------
1 Several recent studies have evaluated the use of central site NO2 or NOx concentration as
2 a surrogate for personal exposure to traffic pollution mixtures. In a near-road
3 environment, NOx concentration can be correlated with pollutants that are also associated
4 with health effects, including UFP and water soluble metals (Sanchez Jimenez et al.,
5 2012): PAHs (Brook et al.. 2007): BTEX (Beckerman et al.. 2008): and EC (Minguillon
6 etal.. 2012) or BC (Clougherty et al., 2013). Correlations generally in the range of
7 0.6-0.8 of NO2 with CO, NOx, and EC (or BC) concentrations forms the basis for a
8 proposed multipollutant mobile source indicator that combines these three species into an
9 Integrated Mobile Source Indicator (IMSI) for traffic-related air pollution. The IMSI is a
10 weighted average of mobile source pollutant concentrations weighted by the ratio of
11 mobile source to total emissions for each pollutant, which Pachonetal. (2012) developed
12 using CO, NOx, and EC. Although the IMSI is not currently used in any epidemiologic
13 studies of the health effects of NO2 or NOx, the IMSI is an informative tool that may shed
14 light on the relationship between traffic-related sources and human exposures, as shown
15 in Equation 3-1.
EmissionCOimobile ^ g,
^""•^'•i' "-EC, total *" "•'"'•"•""'NOx total " Emi-ssi°nCO,total
16 IMSIEB —
EmissionECimobile EmissionNOxmobUe EmissionCOimobih
_ EmissionECitotal EC EmissionNOxfotal N0* EmissionCOitotal
EmissionECimobile EmissionNOxmobUe EmissionCOimobile
EmissionECitotal EmissionNOxtotal EmissionCOitotal
Equation 3-1
17 Note that C' = average concentration normalized by the standard deviation of
18 concentration. Urban street-side (mostly street canyon) NO and NO2 concentrations have
19 been measured and compared with downwind sites, including those located in parks and
20 reference sites (i.e., sites that are located away from or upwind from traffic-related
21 emissions). A study where criteria pollutant concentrations were sampled using high
22 density siting throughout the five boroughs of New York City with 2-week integrated
23 samples per season (Ross etal.. 2013). Consistent with Karner etal. (2010). the
24 street-side sites generally showed higher NO concentrations compared with NO2 (NO:
25 mean 31.82 ppb, max 151.76 ppb; NO2: mean 27.60 ppb, max 87.18 ppb) in Ross et al.
26 (2013) (see Table 3-3). The NO on average was lower than the NO2 away from the road,
27 for example at park sites (NO: mean 18.88 ppb, max 45.15 ppb;NO2: mean 22.13 ppb
28 max 36.94 ppb). The ranges for overall and truck traffic density, Census population, and
29 building areas were all higher for the street-side sites compared with the park sites. In a
30 mobile van study of street canyons in Helsinki, Finland operating continuous monitors
31 during rush hour (sampling interval: 1-minute), Pirjolaet al. (2012) found that the
32 topographical characteristics of the roadway influenced the concentration gradient. They
33 studied concentration profiles on the upwind and downwind sides within a street canyon
January 2015 3-30 DRAFT: Do Not Cite or Quote
-------
1 and observed downwind-to-upwind ratios of 0.28 and 0.70 for NO and NO2, respectively,
2 when the street canyon aspect ratio (building height-to-street width) was 0.55. When the
3 aspect ratio increased to 0.70, downwind-to-upwind ratios decreased to 0.18 and 0.65 for
4 NO and NO2, respectively.
Table 3-3 Summary (mean, range) within 300 m of monitoring sites, by site
type, in a spatially dense monitoring campaign in New York City,
based on 2-week integrated samples per season.
Street-Side Sites Reference Sites Regulatory Sites
(/7=138) Park Sites (/7=12) (n=5) (n=5)
NO2 concentration (ppb)
NO concentration (ppb)
Roadway length (km)
Traffic density (vehicle-km/h)
Truck density (vehicle-km/h)
2000 Census population (number)
Building area (m2)
Residential space area (m2)
Commercial space area (m2)
Industrial space area (m2)
27.6 (8.32-87.2)a
31.8(2.69-152)a
4.3-6.0
561-2,800
13.4-83.2
1,316-5,819
90.7-382
53.79-242
15.6-105
0-7.19
22.1 (8.10-36.9)a
18.9(4.93-45.2)a
2.1-3.7
302-2,560
0.910-24.4
117-3,455
0-163
0-124
0-29.0
0-4.65
20.2 (9.43-38.2)a
15.9(5.42-54.8)a
1.9-2.5
119-783
5.80-13.5
0-522.7
0-38.5
0-30.7
0-18.7
0-0
22.7(17.1-34.2)a
12.1 (3.30-40.0)a
NO2 = nitrogen dioxide; NO = nitric oxide.
Source: Matte et al. (2013).
aavg (range).
5 NO2 and NO emissions, concentrations, and therefore exposures, are also subject to
6 interventions in the built environment. After a tunnel was built in Sydney, Australia to
7 reduce urban pollution levels, Cowie et al. (2012) observed statistically significant
8 reductions in NO2 and NOx concentrations by 1.4 and 4.6 ppb, adjusted for meteorology,
9 based on 2-week passive sampler measurements taken at three periods during Fall
10 2006-2008. Beevers and Carslaw (2005) studied the impact on annual NOx emissions of
11 the London congestion pricing zone implemented in 2003 to reduce traffic in central
12 London. Overall, they reported a 12% decrease in NOx emissions within the congestion
13 pricing zone and a 1.5% increase in NOx emissions at the surrounding ring road, related
January 2015 3-31 DRAFT: Do Not Cite or Quote
-------
1 to some individuals re-routing their drives to the surrounding ring road where no payment
2 was required. Similarly, Panteliadis et al. (2014) studied the impact of congestion pricing
3 in Amsterdam, Netherlands and observed a 6.6% reduction in NCh concentrations at a
4 roadside measurement, with an 11% reduction in the traffic contribution to ambient NC>2
5 concentrations. However, Masiol et al. (2014) analyzed the effects of traffic-free Sundays
6 over 13 years on air quality in the Po Valley of Italy and saw no appreciable change in
7 NO2 levels. Rao etal. (2014) studied the influence of tree canopies on NO2 levels in
8 Portland, OR using LUR modeling and observed a 38% reduction in NC>2 related to
9 increasing the tree canopy at higher elevations in the city. MacNaughton et al. (2014)
10 measured NO2 exposures of bicyclists in Boston using real-time monitoring (3-h avg of
11 1-minute data) equipment and GPS and observed that riding in a shared bicycling/bus
12 lane, traffic density, background NC>2 concentration, and vegetation density were
13 associated with measured NO2 exposures. The city of Beijing, China restricted traffic
14 during the 2008 Olympics, thus creating a natural experiment in pollution reduction.
15 Zhang etal. (2013) reported that the average of 1-hour NO2 concentration measurements
16 dropped from 25.6 ± 3.66 ppb to 14.6 ± 3.76 ppb when comparing periods before (June
17 2-July 20) and during (July 21-September 19) the Olympic games. After the Olympics
18 (September 20-October 30), concentrations increased back up to 41.4 ± 3.81 ppb. Huang
19 etal. (2012) reported reductions of 21.6 and 12.9% for the periods before and during the
20 Olympics compared with the previous year. The reduced NO2 concentrations that
21 followed these interventions suggest that controls can lead to reduced NO2 exposures.
3.3.1.2 Other Outdoor Sources
22 As described in Section 2.3. other sources contributing to ambient NOx emissions include
23 nonroad mobile sources, electric generating units, industrial sources, and wildfires.
24 Nonroad mobile sources, such as airports, shipping ports, and rail yards, can contribute
25 substantially to local and regional ambient NOx concentrations (Kim et al., 2011;
26 Williams et al.. 2009: Vutukuru and Dabdub. 2008: Carslaw et al.. 2006: Unal et al..
27 2005). Carslaw et al. (2012a) took advantage of the natural experiment of the Icelandic
28 volcano eruption of 2010, when airports across Europe were shut down for 6 days, to
29 evaluate the local effect on airport NOx. Downwind of the airport, a 38% reduction in
30 average NOx concentrations (from 42 ppb down to 26 ppb) was observed. At shipping
31 ports and airports, traffic from ground-level support activities can also contribute a large
32 portion to NOx emissions from these sources (Klapmeyer and Marr. 2012: Kim etal..
33 2011). Outside of urban centers where traffic is not a dominant source, other sources of
34 NOx may include wildfires and residential wood-burning. As such, NOx concentration
January 2015 3-32 DRAFT: Do Not Cite or Quote
-------
1 may not always be a reliable proxy for traffic pollution. Section 2.3 discusses different
2 sources of NOx in more detail.
3.3.2 Indoor Dynamics
3.3.2.1 Sources, Sinks, and Penetration
3 The general understanding of oxide of nitrogen production indoors has not changed since
4 the 2008 ISA for Oxides of Nitrogen (U.S. EPA. 2008). Indoor sources of oxides of
5 nitrogen are combustion-based, including gas stoves, gas heating, oil furnaces, coal
6 stoves, wood burning stoves, kerosene heaters, smoking, candle burning, and to a lesser
7 extent, electric cooking. The magnitude of indoor oxides of nitrogen depends on
8 ventilation of the indoor space and appliances, source strength, and rate of pollutant
9 reaction. Recent studies show associations between indoor combustion and indoor NO2
10 levels (Vrijheidetal.. 2012; Kornartit etal.. 2010; Park et al.. 2008) or indoor NOx levels
11 (Cattaneo et al.. 2014). depending on what was measured during the study. HONO can
12 also be emitted directly during combustion or through surface reactions. Park et al.
13 (2008) measured HONO and NO2 during combustion and compared their results with
14 older studies in the peer-reviewed literature, as shown in Table 3-4. High peak-to-mean
15 ratios suggest high temporal variability of the exposures that might be differentiated from
16 exposures of outdoor origin through time-series analysis. This review also generally
17 found higher HONO concentrations in the presence of indoor combustion sources.
18 Oxides of nitrogen can be lost through indoor deposition and ventilation (U.S. EPA.
19 2008). Sarwar et al. (2002) reported deposition velocities of 6-7 x 10~5 m/sec for NO2,
20 HONO, HNO3, HO2NO2, NO3 ~, and N2O5. Much lower deposition velocities (not
21 detected -2 x 10~6 m/s) were reported for NO, PAN, and organic NOs species.
January 2015 3-33 DRAFT: Do Not Cite or Quote
-------
Table 3-4 Indoor nitrogen dioxide (NO2) and nitrous acid (MONO)
concentrations in the presence and absence of combustion.
Study
Braueretal. (1990)3
Braueretal. (1990)b
Braueretal. (1991 )c
Spenqleretal. (1993)d
Simon and Dasqupta
(1995)e
Leaderer et al. (1999)'
Khoder (2002)9
Lee et al. (2002)h
Jarvis et al. (2005)'
Hong et al. (2007V
Park et al. (2008)k
Combustion Source
No source (background)
Gas range3
Convective space
heater3
No source
Gas rangeb
Unknown
Gas range, stove,
furnace
Kerosene heater
No source'
Gas stoves'
Kerosene heaters'
No source'
Gas stoves'
Gas appliances
(summer)
Gas appliances (winter)
Gas range, etc.
Gas hob
Gas oven
Gas range
Gas range
Measurement
Frequency
15 min
15 min
15 min
15 min
15 min
15 min
24-h
8 min
24-h
24-h
24-h
24-h
24-h
24-h
24-h
6-day
4 min
4 min
N02 (ppb)
Peak
29
157
955
5.0
37
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
81.1
189.3
24-h avg
17
36
209
1.8
8
NR
60(24-115)
NR
NR
NR
NR
NR
NR
39 (20-73)
65(27-120)
28
(4.3-52.0)
12.8
12.8
NR
19.4
MONO
Peak
8
35
106
3.5
31
NR
NR
5-10
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
9.3
15.2
(PPb)
24-avg
5
13
42
3.4
9.6
4.7
NR
0.8
4.0
6.8
2.4
5.5
3.7
6.8
4.6
4.1
5.0
NR
2.1
12
/n n o o\
(0.0-11.3)
(0.2-35.9)
(0.1-20.1)
(0.4-20.1)
(1.3-7.3)
(1.6-12.5)
/n -i o-i -i \
avg = average; NR = not reported.
aLocation: Chicago, IL research home, unvented combustion condition; gas range operation hours: 1 h (with one burner and
2,320 kcal/h); convective space heater operation hours: 4 h (with one burner and 2,785 kcal/h).
""Location: Maryland research home, unvented combustion condition; gas range operation hours: 1 h (with one burner and
2,320 kcal/h).
°Location: 11 Boston, MA homes (winter).
location: 10 homes in Albuquerque, NM (winter).
location: Four different home environments with a small kerosene heater (2,270 kcal/h).
'Location: 58 homes (summer) and 223 homes (winter) in southwest Virginia and Connecticut; 39 inside homes without gas stoves
(summer); 19 inside homes with gas stoves (summer); 74 inside kerosene-heater homes (winter); 96 inside homes without kerosene
heaters and gas stoves (winter); 52 inside homes without kerosene heaters and with gas stoves (winter).
9Location: Four homes in suburban residential areas in Greater Cairo, Egypt.
location: 119 homes in southern California (spring).
'Location: Homes in European community.
'Location: Living room of an apartment in Gwangju, Korea (May 2006).
kLocation: Korean apartment (city and year unspecified, October).
Source: Reprinted with permission of Elsevier, Parket al. (2008).
January 2015
3-34
DRAFT: Do Not Cite or Quote
-------
3.3.2.2 Indoor Chemistry
1 The 2008 ISA for Oxides of Nitrogen (U.S. EPA. 2008) described well-established
2 reactions involving oxides of nitrogen and other indoor air pollutants for gas-phase and
3 surface chemistry that serves as both a source and sink for oxides of nitrogen. Knowledge
4 of indoor chemistry helps identify potential sources of uncertainty in estimates of indoor
5 exposure to ambient oxides of nitrogen. Moreover, epidemiologic studies of indoor
6 exposure may providing supporting evidence to the assessment of health effects from
7 ambient NC>2 exposure. Identification of the uncertainty in those exposure estimates can
8 aid interpretation of those studies.
9 For gas phase reactions, indoor NO can be oxidized to NO2 via reaction with Os or HO2
10 radicals generated by indoor Os chemistry or VOCs found in household products. NO2
11 can react with Os to form NOs radicals that may subsequently oxidize organic
12 compounds. NO2 also reacts with free radicals to produce PAN. NO2 removed through
13 surface reactions was known to contribute to NO levels indoors either by surface
14 reduction of NO2 or by reaction of NO2 with aqueous HONO on indoor surfaces (Spicer
15 et al., 1989). Conversion of NO2 to HONO occurs through a number of indoor surface
16 reactions, and the reaction increases with increased relative humidity (U.S. EPA. 2008).
17 A recent study has demonstrated the role of irradiance in humidity-driven surface
18 reaction of NO2 to HONO on paints (Bartolomei et al.. 2014). Surface reactions of NO
19 and OH radicals may also produce HONO, but the reaction rate is slower than for NO2.
20 Indoor combustion can lead to direct emission of NO and HONO, and conversion of NO
21 to NO2 can lead to secondary HONO production from heterogeneous reactions involving
22 NO2 on indoor surfaces. Park et al. (2008) observed HONO to be correlated with both
23 NO (Spearman r = 0.64) and NO2 (r = 0.68) during combustion. They noted that HONO
24 concentrations were 4-8% of NO2 concentrations during gas range operations but rose to
25 -25% of NO2 concentrations after combustion ceased, which underscores the role of
26 surface reaction as the major source of HONO production. In a model of combustion
27 products for oxides of nitrogen during candle and incense burning, Loupa and
28 Rapsomanikis (2008) observed simultaneous NO and HONO production, the latter of
29 which were in agreement with older test chamber results of HONO production during
30 combustion (De Santis etal. 1996). These studies on surface reactions of NO2 provide
31 insight into indoor NO2 sinks that may reduce NO2 exposures as well as exposures to
32 HONO, of which health effects are less well understood.
33 Recent gas-phase indoor chemistry work has shed light on processes involving organic
34 compounds and/or secondary organic aerosols (SOA). Carslaw et al. (2012b) modeled
35 indoor reactions forming SOA and observed that for their base case simulation, organic
36 nitrates constituted 64% of the overall SOA, while PANs constituted an additional 21%.
January 2015 3-35 DRAFT: Do Not Cite or Quote
-------
1 In sensitivity tests varying ambient concentrations and meteorological conditions, organic
2 nitrates varied from 23-76% of the SOA, and PAN varied from 6-42%. N0jgaard et al.
3 (2006) investigated the interference of NC>2 in ozonolysis of monoterpenes in a
4 simulation of indoor air chemistry and observed that NC>2 reacted with Os and hence
5 reduced SOA formation from ozonolysis of alkenes a-pinene and /?-pinene while
6 increasing the mode of the SOA size distribution. Intermediate NOs products may play a
7 role in this process, as described above. However, the presence of NO2 had less effect on
8 ozonolysis of J-limonene, and this is thought to occur because the ozonolysis reaction
9 rate is faster. In chamber experiments and computational chemistry models, Cao and Jang
10 (2008) and Cao and Jang (2010) tested toluene SOA formation in the presence of low
11 (<3 ppb), medium (90-135 ppb), and high (280-315 ppb) NOx concentrations and found
12 that the organic matter component of the toluene SOA yield generally decreased with
13 increasing NOx concentrations, especially when high NO levels (-222-242 ppb) were
14 present. Ji etal. (2012) explored rate constants of NO2 reactions with various low
15 molecular weight aldehydes found indoors and observed that the reaction rates, k,
16 increased in the following order: ^formaldehyde < fecetaidehyde < ŁProPanai < foutanai. Jietal. (2012)
17 concluded from this observation that NO2 reacts more with longer chain, low molecular
18 weight aldehydes compared with shorter chain, low molecular weight aldehydes.
19 RC(=O)- radicals and HONO were both observed to be products of these reactions. These
20 sinks may result in lower NO2 exposures, but little information is available regarding
21 organic nitrate reaction product exposures.
22 Reactions involving N2Os (formed by reaction of NO2 and NOs in the presence of another
23 molecule) in an indoor context have been studied in recent years. In an examination of
24 NOs and N2Os (measured as the sum of those two species) in an office building, N0j gaard
25 (2010) observed that alkenes remove more indoor NOs and N2Os than either ventilation
26 or surface deposition. Griffiths et al. (2009) studied ^Os uptake by organic aerosols in a
27 reaction cell and large chamber (260 m3) and observed little ^Os uptake by solid organic
28 aerosols, more efficient uptake by liquid aerosols, and uptake that increased with
29 increasing RH. N2Os uptake by dicarboxylic acids (oxalic acid, malonic acid, succinic
30 acid, and glutaric acid) was 30-90% of that by (NH4)2SO4 and (NH^SO^mixed
31 dicarboxylic acid aerosols at similar RH. ^Os uptake by malonic or azelaic acid in the
32 presence of higher RH is consistent with findings of Thornton et al. (2003) for
33 experiments conducted in a reaction cell. Raff etal. (2009) suggested that N2Os
34 autoionizes to NO2 + NOs and then reacts quickly with water to form HNOs; it is
35 possible that HNOs might then participate in the liquid aerosol reactions described by
36 Griffiths et al. (2009) and Thornton et al. (2003). Raff et al. (2009) also proposed
37 autoionization of N2Os as a likely mechanism for reaction with HC1, which would result
38 in C1NO and HNOs formation while NO2 and water vapor experienced an intermediate
39 surface reaction to form HONO, which would react with HC1. Complexity of reactions
January 2015 3-36 DRAFT: Do Not Cite or Quote
-------
1 involving N2Os in creating NC>2 as an intermediary reaction product also lends
2 uncertainty to NC>2 exposure assessment. This uncertainty may lead to variability in
3 personal or indoor NC>2 exposure measurements.
3.4 Exposure Assessment and Epidemiologic Inference
4 The 2008 ISA for Oxides of Nitrogen (U.S. EPA. 2008) examined several factors
5 influencing exposure to ambient oxides of nitrogen and measurements used to represent
6 exposures. These include high spatial and temporal variability of NO2 concentrations in
7 urban areas and near roads, location of NO2 samplers, and ventilation of indoor
8 microenvironments. The 2008 ISA for Oxides of Nitrogen (U.S. EPA. 2008) concluded
9 that errors associated with the use of NO2 concentrations measured at central site
10 monitors as exposure metrics for epidemiologic studies tended to bias the health effect
11 estimate towards the null for both short-term exposure and long-term exposure
12 epidemiologic studies. The following sections explore new evidence regarding a
13 conceptual exposure model, exposure metrics employed in epidemiologic studies,
14 personal-ambient relationships, factors that introduce exposure error, potential
15 confounding, and how the exposure errors may or may not introduce bias and uncertainty
16 into epidemiologic health effect estimates, depending on the epidemiologic study design.
3.4.1 Conceptual Model of Total Personal Exposure
17 Total personal exposure (Ł>) integrates the product of microenvironmental concentration
18 (Q and fraction of time spent in a microenvironment across an individual's
19 microenvironmental exposures, t:
n
ET =
Equation 3-2
20 where Q = average NO2 concentration in the/th microenvironment, tj = fraction of total
21 time spent in the/th microenvironment, and n = total number of microenvironments
22 which the individual has encountered (U.S. EPA. 2008) (Klepeis etal.. 2001). Hence,
23 both the microenvironmental NO2 concentration and time-activity aspects of total
24 exposure must be considered.
January 2015 3-37 DRAFT: Do Not Cite or Quote
-------
1 Alternatively, based on the principle of mass balance, an individual's total NO2 exposure
2 can be expressed as the sum of its ambient NO2 exposure (Ea) and nonambient NO2
3 exposure (Ena) components (U.S. EPA. 2008) (Wilson and Brauer. 2006):
= E +
Equation 3-3
4 Ea represents the amount of NC>2 exposure derived from outdoor sources, and Ena
5 represents the amount of NC>2 exposure from indoor sources. The microenvironmental
6 formulation presented in Equation 3-2 and the component formulation presented in
7 Equation 3-3 can be rectified by recognizing that Ea and Ena can both be expressed in
8 terms of microenvironmental concentrations and time spent in each outdoor and indoor
9 microenvironment. During the fraction of a day spent in each outdoor microenvironment
10 (yoj), ambient exposure to NC>2 having an outdoor concentration of C0j is:
Equation 3-4
1 1 Indoor NC>2 exposures in the/th microenvironment (Ey) are more complicated because
12 some part of indoor exposure may emanate from nonambient sources, and some part of
13 indoor exposure infiltrates from outdoors. Indoor exposures from nonambient sources are
14 given as Ena,j. Exposures in each indoor microenvironment from ambient sources are also
15 influenced by infiltration of outdoor NC>2 (INF}), time spent indoors (yy), and COJ:
Ei,j = ytjINFj • CoJ +
Equation 3-5
16 Infiltration is a function of the/th microenvironment' s air exchange rate (a/), air pollutant
17 penetration (/*,), and decay rate (&/):
January 2015 3-38 DRAFT: Do Not Cite or Quote
-------
Equation 3-6
1 Hence, indoor NCh exposure for microenvironment/ is the sum of the ambient and a
2 nonambient components:
""naj
Equation 3-7
3 Finally, Ea can be described as the sum of the outdoor NO2 exposure and the ambient
4 component of the indoor NC>2 exposure, summed over/ indoor microenvironments
5 (U.S. EPA. 2008) (Wilson and Brauer. 2006: Wilson et al.. 2000):
/ = 1 j = l
Equation 3-8
6 Ambient concentration of NCh is often used as a surrogate for human exposure. In
7 concert, a second simplifying assumption is often made that the exposed individual
8 resides in one indoor microenvironment, such that time-activity data are reduced to "time
9 indoors" and "time outdoors." Errors associated with this approach, which may vary
10 depending on the epidemiologic study design in which the exposure surrogate is used, are
11 described in detail in Section 3.4.3. In this case, outdoor microenvironmental NC>2
12 exposures (E0) are expressed simply as the product of the fraction of all time spent
13 outdoors (y0) and ambient NCh concentration (Ca): E0 = y0Ca. Furthermore, based on the
14 assumption that the individual occupies only one indoor and one outdoor
15 microenvironment, then the infiltration term can be simplified to yi\P-al(a + Ł)], and
16 because j0+j; = 1:
January 2015 3-39 DRAFT: Do Not Cite or Quote
-------
Ea=
Equation 3-9
1 Then, an exposure factor (a) can be defined to express the influence of time-weighting
2 and infiltration on NO2 exposure:
a =y0+ (l-y0)[Po/(o + fc)]
Equation 3-10
3 Last, an approximate expression for total personal exposure is obtained:
ET = aCa + Ena
Equation 3-11
4 Comparison of Equations 3-3. 3-9. and 3-11 reveals that a can also be approximated as
5 the ratio EJCa. Subsequent sections examine how Ea, a, and Ca are modeled or measured,
6 and how errors and uncertainties in the simplifying assumptions behind Equations 3-9.
7 3-10. and 3-11 may influence health effect estimates computed from epidemiologic
8 studies of varying design.
3.4.2 Personal-Ambient Relationships and Nonambient Exposures
9 Personal exposure measurements typically capture both ambient and nonambient
10 exposure contributions; for the purpose of this document, these are referred to as "total
11 personal exposure" measurements. The 2008 ISA for Oxides of Nitrogen (U.S. EPA.
12 2008) concluded that literature relating ambient NC>2 concentrations measured by a
13 central site monitor to personal NCh exposures was mixed for studies of both short-term
14 and long-term NC>2 exposure, with some studies finding associations between the
15 personal and central site monitors and other studies finding no association. These
16 inconsistencies reflected various factors that influence exposure in respective studies,
17 including proximity and strength of sources of ambient and nonambient NOx,
18 spatiotemporal variability of NC>2 concentrations, and time-activity behavior of the
19 exposed sample population. Recent studies have found that personal NO2 concentration
20 measurements taken for adults and children tend to be more highly correlated with indoor
January 2015 3-40 DRAFT: Do Not Cite or Quote
-------
1 concentrations compared with personal correlations with outdoor or ambient
2 concentrations, although wide variability in the correlations was observed (see Tables 3-5
3 and 3-6). Personal-outdoor (i.e., measurements taken outdoors but not at a central site
4 monitor) correlations also tended to be higher for summer compared with winter. This is
5 not surprising because open windows and greater time spent outdoors during summer
6 likely increase exposure to outdoor air (Brown et al.. 2009). The study results indicate
7 that, for epidemiologic studies of short-term exposure, indoor sources of NC>2 can add
8 noise to the ambient NCh exposure signal. As described further in Section 3.4.5.1.
9 uncertainty in the NC>2 exposure term can lead to negative bias and added uncertainty in
10 the epidemiologic health effect estimate for short-term exposure studies.
January 2015 3-41 DRAFT: Do Not Cite or Quote
-------
Table 3-5 Ambient, outdoor, transport, indoor, and personal nitrogen dioxide measurements (ppb) across
studies.
Study
Sarnat et al.
(2012)
Williams et al.
(2012b):
Mena et al.
(2012a)
Suh and Zanobetti
(2010)
Location
El Paso, TX
(large city)
Ciudad
Juarez,
Mexico
(large city)
Detroit, Ml
(large city)
Metropolitan
Atlanta, GA
(large city)
Time Period
January-
May, 2008
Summer,
2004-2007
Winter,
2004-2007
Fall,
1999-Spring,
2000
Ambient Outdoor
Sampling (Central (Outside
Interval Site) Residence) Transport Indoor
96-h 14.0-20.6C 4.5-14.2c NRd 4.0-8.1C
NR 18.7-27.2C NR 23.1-120.8
24-h Williams: NR NR NR
22.0e;
Meng:
22.0e; 22.7C
24.0e;23.9c NR NR NR
24-h 17.96e; NR NR NR
17.13C
Personal
NR
NR
Total:
Williams: 25.5C;
Meng: 25. 4C
Ambient: 16. Oe;
21. Oc
Total:
24.0e; 35.6C
Ambient:
18.0e;20.4c
8.08e; 11.60C
Personal-Ambient
Slopea'b
NR
NR
Meng: 0.24; 0.13f
Meng: 0.08; 0.07f
NR
January 2015
3-42
DRAFT: Do Not Cite or Quote
-------
Table 3-5 (Continued): Ambient, outdoor, transport, indoor, and personal nitrogen dioxide measurements (ppb)
across studies.
Study
Brown et al.
(2009)
Delfino et al.
(2008)
Delqado-Saborit
(2012)
Location
Metropolitan
Boston, MA
(large city)
Riverside,
CA; Whittier,
CA (SoCAB)
(large city)
Birmingham,
U.K. (large
city)
Ambient Outdoor
Sampling (Central (Outside
Time Period Interval Site) Residence) Transport Indoor
Nov 24-h 25.89; 26.8C NR NR NR
1999-Jan
2000
June-July 22.09; 22.8C NR NR NR
2000
July-Dec 24-h 25.3e; 25.0C NR NR NR
2003
(Riverside);
July-Dec
2004
(Whittier)
July-Oct 5-min 47C 64C Car: 40C Office: 14C
2011 Bus:71c Home: 17C
Bike: 125C
Train: 58C
Personal-Ambient
Personal Slopea>b
10.49; 12.9= All: 0.19
Windows closed:
0.09
Windows open:
0.31
LowAER: 0.21
HighAER: 0.15
13.99; 17.4C All: 0.23
Windows closed:
0.64
Windows open:
0.10
LowAER: 0.34
HighAER: 0.19
26.7e; 28.6C NR
AII:23C 1-havg: 0.044
Gas oven: 31C Sampling event:
Electric oven: 19C 0.14
January 2015
3-43
DRAFT: Do Not Cite or Quote
-------
Table 3-5 (Continued): Ambient, outdoor, transport, indoor, and personal nitrogen dioxide measurements (ppb)
across studies.
Ambient Outdoor
Sampling (Central (Outside
Study Location Time Period Interval Site) Residence) Transport Indoor
Kornartit et al. Hertfordshire Winter 2000 7-day NR NR NR Electric oven:
(2010) , U.K. Bedroom: 7.8C
(Greater Living room:
London 7.9C
Area) (large Kitchen: 7.1 c
city)
Gas oven:
Bedroom:
10. 8C
Living room:
13.7C
Kitchen: 20.6C
Summer 2001 NR NR NR Electric oven:
Bedroom:
12.7C
Living room:
13.1C
Kitchen: 11. Oc
Gas oven:
Bedroom:
14.3C
Living room:
14.7C
Kitchen: 14. 2C
Personal-Ambient
Personal Slopea>b
Electric oven: 8.1 c NR
Gas oven: 11. 2C
Electric oven: NR
13.3C
Gas oven: 14.6C
January 2015
3-44
DRAFT: Do Not Cite or Quote
-------
Table 3-5 (Continued): Ambient, outdoor, transport, indoor, and personal nitrogen dioxide measurements (ppb)
across studies.
Ambient
Sampling (Central
Study Location Time Period Interval Site)
Lee etal. (2013) Seoul, Korea July 2008 NR 29.5f; 30.7C
(large city)
Jan 2009 NR 29.59; 31. 1C
Daegu, July 2008 NR 19.99; 21. 1C
Korea (mid-
sized city)
Jan 2009 NR 23.09; 24.3C
Asan, Korea July 2008 NR 26.09; 27.9C
(small city)
Jan 2009 NR 21.69;23.9C
Outdoor
(Outside
Residence) Transport
NR NR
NR NR
NR NR
NR NR
NR NR
NR NR
Indoor Personal
Home: 25.39; 27C
24.49; 25.7C
Work: 19.29;
21. 5C
Home: 22.59; 24.2C
20.99; 24.9C
Work: 27.99;
29.9C
Home: 21.49;22.6C
19. 39; 20. 3C
Work: 21. 39;
22.8C
Home: 20.39; 21. 7C
23.39; 25.1C
Work: 20.39;
22.9C
Home: 22.69; 24.3C
23.89; 24.9C
Work: 21. 19;
25.6C
Home: 19.99; 22. 3C
20.39; 22. 9C
Work: 13.09;
18.6C
Personal-Ambient
Slopea'b
NR
NR
NR
NR
NR
NR
January 2015
3-45
DRAFT: Do Not Cite or Quote
-------
Table 3-5 (Continued): Ambient, outdoor, transport, indoor, and personal nitrogen dioxide measurements (ppb)
across studies.
Ambient
Sampling (Central
Study Location Time Period Interval Site)
Lee etal. (2013) Suncheon, July 2008 NR 15.09; 15.9C
(Continued) Korea (rural)
Jan 2009 NR 12.59; 15.2C
Total July 2008 NR 21.79;23.7C
Jan 2009 NR 20.69; 23.6C
Du etal. (2011) Beiiinq, Oct2006 Varied NR
China with transit
times
Outdoor
(Outside
Residence) Transport
NR NR
NR NR
NR NR
NR NR
NR Subway:
20C;
Nonsubway
:22C;
Taxi
drivers: 25C
Indoor Personal
Home: 14.09; 15.3=
13.09; 14. 30
Work: 12.09;
14.5C
Home: 12.99; 15.70
15.99; 20.4C
Work: 9.39;
12.9C
Home: 20.59; 22.6C
19.59; 21.2C
Work: 18.49;
21. 4C
Home: 18. 69; 21. Oc
19. 99; 23. 3C
Work: 16.49;
21. 1C
NR NR
Personal-Ambient
Slopea'b
NR
NR
NR
NR
NR
January 2015
3-46
DRAFT: Do Not Cite or Quote
-------
Table 3-5 (Continued): Ambient, outdoor, transport, indoor, and personal nitrogen dioxide measurements (ppb)
across studies.
Sampling
Study Location Time Period Interval
Physick et al. Melbourne, May 2006; Ambient:
(2011) Australia June 2006; 1 h; Perso
(large city) April 2007; nal:
May 2007 Participant
s wore two
sets of
passive
samplers.
One was
worn for
48 h.
One was
worn only
during the
hours
spent at
home, at
work, in
transit, or
while
performing
other
activities.
Sahsuvaroqlu Hamilton, Oct2003 72-h
et al. (2009) Canada
(mid-sized
Clty) May 2004
Aug 2004
Total
Ambient Outdoor
(Central (Outside
Site) Residence) Transport
6:00p.m.to NR NR
8:00 a.m.:
19.8e; 18.7C
8:00 a.m. to
6:00 p.m.:
20.3e;21.2c
NR All: 32.0C NR
Non-ETS:
31. 7C
NR All: 17.6C NR
Non-ETS:
16.8C
NR All: 9.7C NR
Non-ETS:
9.6C
NR All: 19.3C NR
Non-ETS:
18. 9C
Indoor
Home: 17.2e;
16.8C
Work: 21. 6e;
21. 7C
All: 22.4C
Non-ETS:
21. 9C
All: 13.5C
Non-ETS:
12.3C
All: 8.2C
Non-ETS: 7.4C
All: 14.4C
Non-ETS:
13.6C
Personal
Total: 12.2h
Home: 8.2h
Work: 14.7h
Transit: 23. 4h
Other: 17.4h
All: 23.3C
Non-ETS: 22.4C
All: 14.4C
Non-ETS: 14.0C
All: 8.8C
Non-ETS: 8.2C
All: 15.2C
Non-ETS: 14.6C
Personal-Ambient
Slopea'b
NR
All: 0.06
Nonsmoking: 0.13
All: 0.31
Nonsmoking: 0.28
All: 0.50
Nonsmoking: 0.15
All: 0.62
Nonsmoking: 0.62
January 2015
3-47
DRAFT: Do Not Cite or Quote
-------
Table 3-5 (Continued): Ambient, outdoor, transport, indoor, and personal nitrogen dioxide measurements (ppb)
across studies.
Ambient
Sampling (Central
Study Location Time Period Interval Site)
Schembari et al. Barcelona, Nov 2008 and 7-day NR
(2013) Spain (large Oct 2009
city)
Mollov et al. Melbourne, Auq 7-dav NR
(2012) Australia 2008-Dec
(large city) 2008; Jan
2009-April
2009
Peqas et al. Aveiro, April-June 7-dav NR
(2012) Portugal 2010
(small city
center,
suburb)
Chatzidiakou et al. Suburban Nov 2011 5-dav NR
(2014) London,
Enqland NR
NR
London, NR
England NR
NR
Rivas etal. (2014) Barcelona Jan-June 4-dav 22C; 20d
and Sant 2012
Cugat, Spain
AER = air exchange rate; avg = average; ETS = environmental tobacco smoke; NR
aUnadjusted models only.
Total personal NO2 exposure vs. ambient concentration unless noted otherwise.
°Average.
dMedian.
ePersonal exposure to ambient NO2 vs. ambient concentration.
'Geometric mean.
9Data provided by the authors for Figure 1 of Phvsick et al. (201 1).
hReported in |jg/m3 and converted to ppb assuming 25°C and 760 mmHg.
Averaged over 4 classrooms and 2 weeks.
'Indoor-outdoor ratio, rather than slope, is reported for Schembari et al. (2013).
Integrated measurement over 2 weeks.
'Estimated from reported indoor-outdoor ratio and outdoor NO2 concentration.
Outdoor
(Outside
Residence) Transport Indoor Personal
18. 7g''; NR 19. 2g''; 20. 6C'' 17. 7g''; 18. 6C''
9.5e;10.0c NR 7.9e; 8.4C NR
City center: NR City center: NR
10. 50'1; 7. 4c>i>i; Suburb:
Suburb: 6.9^
7.4 NR 3.71 NR
5.1 NR 2.9' NR
5.1 NR 2.7' NR
19 NR 13' NR
20 NR 16' NR
22 NR 18' NR
25C; 24d NR 16C; 16d NR
= not reported; SoCAB = South Coast Air Basin.
Personal-Ambient
Slopea'b
1.01k
0.9k
NR
NR
NR
NR
NR
NR
NR
NR
January 2015
3-48
DRAFT: Do Not Cite or Quote
-------
Table 3-6 Correlations between measured nitrogen dioxide (NO2)
concentrations from personal, outdoor, indoor, and ambient
monitors.
Study
Sarnatetal. (201 2)a
Williams et al. (2012a)a
Suh and Zanobetti (201 0)a
Brown et al. (2009)
Delfinoetal. (2008)
Kousaetal. (2001)
Delqado-Saborit (2012)
Lee etal. (2013)
Location
Ciudad Juarez, Mexico; El
Paso, TX
Wayne County, Ml
Atlanta, GA
Boston, MA
2 southern California cities
Helsinki, Finland; Basel,
Switzerland; Prague,
Czech Republic
Birmingham, U.K.
Seoul, South Korea
Daegu, South Korea
Asan, South Korea
Suncheon, South Korea
All 4 cities
Personal-
Ambient
NR
All Subjects:
0.11; Vest-
compliant
(>60%)c:0.14
0.12
Winter: 0.00;
Summer: 0.03
0.43
NR
1-h NO2: 0.024;
Sampling event
NO2: 0.15
NR
NR
NR
NR
NR
Personal-
Outdoor
NR
NR
NR
NR
NR
0.61
NR
Summer:
0.39;
Winter: 0.47
Summer:
0.43;
Winter: 0.47
Summer:
0.62;
Winter: 0.11
Summer:
0.46;
Winter: 0.56
Summer:
0.58;
Winter: 0.53
Personal-
Indoor
NR
NR
NR
NR
NR
0.73
NR
Summer:
0.50; Winter:
0.55
Summer:
0.32; Winter:
0.59
Summer:
0.63; Winter:
0.37
Summer:
0.46; Winter:
0.60
Summer:
0.60; Winter:
0.55
Outdoor-
Indoor
CJ-A: 0.36;
CJ-B: 0.92;
EP-A: 0.66;
EP-B: 0.01
NR
NR
NR
NR
0.66
NR
Summer:
0.71; Winter:
0.22
Summer:
0.65; Winter:
0.57
Summer:
0.67; Winter:
0.37
Summer:
0.77; Winter:
0.80
Summer:
0.78; Winter:
0.55
January 2015
3-49
DRAFT: Do Not Cite or Quote
-------
Table 3-6 (Continued): Correlations between measured nitrogen dioxide (NO2)
concentrations from personal, outdoor, indoor, and
ambient monitors.
Study
Sahsuvaroqlu et al.
(2009)b
Schembari et al. (201 3)b
Vieiraetal. (201 2)a
Van Roosbroeck et al.
(2008)a
Location
Lake Ontario, Canada
(winter)
Lake Ontario, Canada
(spring)
Lake Ontario, Canada
(summer)
Lake Ontario, Canada
(all seasons)
Barcelona, Spain
Sao Paolo, Brazil
Netherlands (3 schools)
Personal- Personal-
Ambient Outdoor
NR All
Subjects:
0.002; Non-
ETS: 0.020
NR All
Subjects:
0.233; Non-
ETS: 0.187
NR All
Subjects:
0.067;
Non-ETS:
0.011
NR All
Subjects:
0.517;
Non-ETS:
0.540
NR 0.58
NR <0.35
NR 0.35
Personal-
Indoor
All Subjects:
0.430;
Non-ETS:
0.283
All Subjects:
0.589;
Non-ETS:
0.599
All Subjects:
0.822;
Non-ETS:
0.783
All Subjects:
0.729;
Non-ETS:
0.693
0.78
NR
NR
Outdoor-
Indoor
NR
NR
NR
NR
0.53
All subjects:
0.13;
Non-ETS:
0.42
NR
CJ-A = Ciudad Juarez Site A; CJ-B = Ciudad Juarez Site B; EP-A = El Paso Site A; EP-B = El Paso Site B; ETS = Environmental
Tobacco Smoke; NR = not reported.
aSpearman coefficient.
bPearson coefficient.
°Subjects wore the sampling vests at least 60% of the sampling period.
4
5
6
7
8
9
10
11
12
13
14
Several studies have investigated factors that influence the relationship between
short-term personal exposure measurements and ambient concentrations. It was observed
that, even when the median or average total personal NO2 exposures and ambient
concentrations were comparable, the total personal exposure measurements and central
site monitor concentrations might not have always been correlated. For example,
Williams et al. (2012a) measured total personal NO2 exposures for the Detroit Exposure
and Aerosol Research Study (DEARS) population of nonsmoking adults in 24-hour
intervals and found a low association (Spearman r = 0.14 for participants complying with
study protocols; r = 0.11 for all participants) between total personal NO2 exposure with
NO2 measured at central site monitors. This result indicated the influence of nonambient
sources on the DEARS participants' total personal NO2 exposures, suggesting that total
personal and ambient NO2 exposures are not always well correlated. Likewise, Suh and
Zanobetti (2010) measured correlation of Spearman r = 0.12 between 24-hour total
January 2015
3-50
DRAFT: Do Not Cite or Quote
-------
1 personal exposure and central site NO2 measurements among an Atlanta panel of
2 30 adults. Vieiraetal. (2012) calculated Spearman correlations between 12-hour outdoor,
3 indoor, and personal NC>2 measurements. All correlations between personal and outdoor
4 NC>2 measurements were below r = 0.35. Indoor and outdoor NC>2 concentrations were
5 more correlated (r = 0.42), although when smokers were included, correlation between
6 indoor and outdoor NC>2 dropped (r = 0.13). Van Roosbroeck et al. (2008) compared
7 personal NC>2 measurements for children obtained over 1 to 4 weeks in a panel study with
8 measurements taken outside the children's schools, and they observed correlation of
9 Pearson R = 0.35. Outdoor school NC>2 underestimated personal NC>2 when used as a
10 surrogate, but when additional variables representing indoor exposures (such as exposure
11 to gas cooking and unvented water heaters) were added to the model, R increased to 0.77,
12 suggesting that indoor sources play a large role in NC>2 exposure among the study
13 participants. Bellander et al. (2012) measured personal NO2 exposure using 7-day
14 integrated diffusion samplers and modeled it as a function of NO2 concentrations
15 measured at an urban area, rural area, roadside, and outside of the participants' homes
16 and places of work in Stockholm County, Sweden. They observed slopes ranging from
17 0.25-0.37 (R2 = 0.01-0.20). Kousaet al. (2001) developed a time-weighted
18 microenvironmental model of NC>2 exposure based on time-activity data and 48-hour
19 microenvironmental measurements. The microenvironmental model corresponded with
20 personal exposure measurements (/? = 0.90; R2 = 0.74).
21 Meng et al. (2012b) performed a random effects meta-analysis of 15 studies that
22 calculated slopes and correlations between personal NO2 measurements of ET and central
23 site ambient NC>2 concentrations for 32 sample populations, of which 7 were from daily
24 average analyses, 8 were from longitudinal panel analyses, and 17 were from analyses
25 whose correlations were pooled over short time periods up to 1 week in length.
26 Metaregression results are shown in Table 3-7. Meng etal. (2012b) found that the
27 magnitude and correlation of associations between personal exposure and ambient
28 concentration depended on several factors, including study design (pooled data across
29 days, longitudinal panel, or daily average), season, meteorological conditions, ambient
30 PM2 5 level, and pre-existing cardiopulmonary disease of the exposure subjects. Together,
31 the low associations reported in these studies indicate that most of the total personal NO2
32 exposure measurements for these studies were influenced either by nonambient sources
33 or by spatially variable NC>2 not well detected by the central site monitor. However,
34 Meng etal. (2012b) also stated that the longitudinal panel studies included in their
35 meta-analysis had several measurements below detection limit that could have
36 erroneously reduced the correlations, which otherwise would be expected to be higher.
January 2015 3-51 DRAFT: Do Not Cite or Quote
-------
Table 3-7 Meta regression results from 15 studies examining the relationship
between personal nitrogen dioxide exposure measurements and
ambient concentrations.
Slope Correlation Slope Correlation
Study Design Based on Original Studies Corrected for Publication Bias
Pooled3 0.40 0.42 0.30 0.37
Longitudinal panelb 0.14 0.16 0.14 0.16
Daily average0 0.29 0.72 0.20 0.45
aPooled analyses: Piechocki-Minguv et al. (2006). Linnetal. (1996). Liard et al. (1999). Gauvinetal. (2001). Aim et al. (1998).
Brown et al. (2009). Sarnat et al. (2006). Delfino et al. (2008).
"Longitudinal analyses: Sarnat et al. (2005). Sarnat et al. (2001). Sarnat et al. (2000). Linaker et al. (2000). Kim et al. (2006).
Koutrakisetal. (2005).
°Daily average analyses: Mukala et al. (2000). Liard et al. (1999). and Aim etal. (1998)
Source: Meng et al. (2012b).
3.4.3 Factors Contributing to Error in Estimating Exposure to Ambient Nitrogen
Dioxide
1 Recent studies of factors influencing exposure error build from the existing literature
2 presented in the 2008 ISA for Oxides of Nitrogen U.S. EPA (2008). which have focused
3 on time-activity patterns, spatial variability, infiltration, nonambient exposures, and
4 instrument accuracy and precision, as described in the subsequent subsections. These
5 factors can influence epidemiologic results for studies of short-term and long-term NC>2
6 exposure, as detailed further in Section 3.4.5.
3.4.3.1 Time-Activity Patterns
7 The complex human activity patterns across the population (all ages) are illustrated in
8 Figure 3-3 (Klepeis et al.. 2001) for data from the National Human Activity Pattern
9 Survey (NHAPS). This figure is presented to illustrate the diversity of daily activities
10 among the entire population as well as a generalized proportion of time spent in each
11 microenvironment. Time-activity data become an important source of uncertainty when
12 considering that ambient exposures vary in different microenvironments (e.g., transit,
13 residential), and that exposure assignment is typically based on the assumption that study
14 participants are in one location (residential) for the study duration.
January 2015 3-52 DRAFT: Do Not Cite or Quote
-------
1
o
ad
OJ
sp
n
Near Vehicle
(Outdoors)
&,&,&<&&i&ifii&,&i(SH&&irt
ooooooooooooooooooooooooo
OOOOOOOOOOOOOOOOOOOOOOOOO
C-i •— i r4 CO Tf ij-j \b t^ 00 O\ O
(N rt c<
00
O --
Time of Day
Source: Reprinted with permission of Nature Publishing Group, Klepeis et al. (2001).
Figure 3-3 Distribution of time sample population spends in various
environments, from the U.S. National Human Activity Pattern
Survey (all ages).
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Different time-activity patterns have been found when analyzing activity patterns for
different populations or life stages. For example, Wuetal. (2010) observed activity
patterns for a panel of adults and children from Camden, NJ communities with larger
percentages of nonwhites (85%) and those below the poverty line (33%) compared with
NHAPS. The study participants spent more time outdoors compared with the nationwide
cohort (3.8 hours vs. 1.8 hours nationally); note that Wu et al. (2010) undersampled
participants ages 65+ years, and the median age of the population studied in Wu et al.
(2010) was 27 years compared with 35 years nationwide. Other recent time-activity panel
studies have included working adults (Isaacs et al.. 2013; Bellander et al., 2012; Kornartit
etal.. 2010). pregnant women (Iniguez et al.. 2009). adolescents (DeCastro et al.. 2007).
and children (Molter et al.. 2012; Xue etal.. 2004). In many cases, the time-activity data
were limited to residential, occupational, and outdoor location categories to simplify
assignment of concentrations to which the subjects were estimated to be exposed in each
microenvironment. The implication of these findings is that, given that time-activity data
vary among different populations, the one-location assumption used in many studies
January 2015
3-53
DRAFT: Do Not Cite or Quote
-------
1 varies in accuracy among those different populations. However, given that few studies
2 are as large as NHAPS, it would be premature to make conclusions about time-activity
3 data for smaller segments of the population.
4 Recently, Kornartit et al. (2010) tested the associations between time-weighted exposure
5 estimates from area samples with personal sampling measurements for a London, U.K.
6 panel study. Kornartit et al. (2010) measured NO2 concentration for 1 week with passive
7 Palmes tube samplers in several outdoor and indoor microenvironments for 55 subjects
8 aged 21-60 years and correlated a time-weighted average of those microenvironmental
9 NC>2 concentration measurements with personal NC>2 concentration measurements, also
10 measured with Palmes tubes. They observed a slope of 0.94 for the relationship between
11 time-weighted average and personal NO2 concentrations (R2 = 0.85) in winter and a slope
12 of 0.59 (R2 = 0.65) in summer. Higher levels of NC>2 were observed for both
13 time-weighted average and personal concentrations in summer compared with winter.
14 However, correlations between personal NO2 exposure and time-weighted
15 microenvironmental NCh concentrations were higher in winter, implying panel studies
16 using personal NC>2 exposure measurements may be more dominated by indoor sources
17 during cold-weather months. The authors concluded that the time-weighting approach
18 provided a reasonable approximation of personal exposure but sometimes underestimated
19 it.
3.4.3.2 Spatial Variability in Nitrogen Dioxide Concentrations
20 Data for spatial variability in ambient NO, NC>2, and NOx concentrations are provided in
21 Section 2.5 for national, urban, neighborhood, and micro scales. The data illustrate that
22 national variation in wintertime concentrations largely follows the degree of urbanization,
23 while variation at urban and smaller scales is influenced by source location, source
24 strength, meteorology, and natural and urban topography. Figure 3-4 illustrates
25 regional-scale variability in background levels of daily 1-h max NC>2 based on Pearson
26 correlation between monitor pairs for urban and rural monitors across the U.K. (Butland
27 etal.. 2013). Likewise, Figure 3-5 depicts urban-scale variability for NC>2 and NOx, based
28 on a semivariogram function (Goldman et al.. 2010). Gradients in near-road
29 concentrations of NO2 and NO indicate spatial variability at finer scales within 300 m of
30 the road (see Figures 2-16 and 2-17 in Section 2.5.3 and Figure 3-2 in Section 3.3.1.1).
31 All of these data indicate that the magnitude of the error in exposure estimation increases
32 with distance between the monitor and the subject. Hence, there is a potential for
33 exposure misclassification if the ambient NO2 concentration measured at a given site
34 differs from the concentration at the location of an epidemiologic study participant, and
35 this issue is present regardless of the spatial scale of the epidemiology study.
January 2015 3-54 DRAFT: Do Not Cite or Quote
-------
C CO
O '
O
o •<
o
S2 CM-
Urban background loge nitrogen dioxide
Fitted line
P = 071027-0 00073x0
R-stpO 71
200 400 600 800
Distance in km (D)
1000
Source: Butland et al. (2013)
Rural loge nitrogen dioxide
C CO
O '
o
o
o
Fitted line
P = 061761 -000073xD
R-sq=0 40
200 400 600 800
Distance in km (D)
1000
Figure 3-4 Regional-scale variability in nitrogen dioxide for urban and rural
area data across the U.K.
NO2
0 50 100
Distance (km)
NOx
0 50 100
Distance (km)
Source: Goldman et al. (2010)
Figure 3-5 Urban-scale variability in nitrogen dioxide (NOa) and the sum of
nitric oxide and NOa (NOx) in Atlanta, GA. On the y-axis, v'
denotes the semivariogram, i.e., a unitless function that describes
the ratio between spatial and temporal variance of the differences
between two observations.
January 2015
3-55
DRAFT: Do Not Cite or Quote
-------
3.4.3.3 Infiltration and Building Ventilation
1 Given that people spend the majority of their time indoors, building air exchange rates
2 influence exposure to ambient NC>2. In an analysis of daily average NO2 data from the
3 DEARS, Meng et al. (2012a) observed seasonal differences, with slopes of 0.24 ± 0.04
4 for ET versus the concentration measured at a central site monitor, Ca,csm, and of
5 0.13 ± 0.06 for Ea versus Ca,csm for summer measurements. For winter measurements, the
6 associations were lower (ET vs. Ca,COT!: slope = 0.08 ± 0.05; Ea vs. Ca,COT!:
7 slope = 0.07 ± 0.07). Mengetal. (2012a) found that high air exchange rate (>1.3 air
8 changes per hour), no central air conditioning, use and nonuse of window fans, and
9 presence of old carpeting were determinants of a, the exposure factor defined in
10 Equation 3-10 and approximated by the ratio of Ea to Ca, for NC>2 in summer; none of
11 these factors were determinants of a for NO2 in winter. In Molter et al. (2012). outdoor
12 exposures were calculated with LUR, while indoor exposures were calculated using the
13 Probabilistic Model for Indoor Pollution Exposures (INDAIR) model that accounts both
14 for infiltration due to home ventilation characteristics and indoor sources. Sensitivity to
15 air exchange rate of INDAIR predictions of indoor NC>2 in the absence of indoor sources
16 underscores potential for bias and uncertainty in a, which depends on air exchange rate,
17 penetration, and indoor deposition (Dimitroulopoulou et al.. 2006).
3.4.3.4 Instrument Accuracy and Precision
18 Instrument error occurs when the NC>2 measurements are subject to interferences that
19 cause positive or negative biases or noise. NO2 measurements are subject to positive bias
20 from detection of other oxidized nitrogen compounds on the measurement substrate, and
21 these errors are larger in warm seasons. See Section 2.4.1 for details on errors that affect
22 FRMs and FEMs used for central site monitoring, and see Section 3.2.1.2 for a
23 description of errors to which personal samplers are subject. Intermonitor comparison is
24 often used to estimate instrument precision. For example, Goldman et al. (2010)
25 investigated instrument precision error at locations where ambient monitors were
26 colocated. Instrument precision error increased with increasing concentration. When
27 instrument error and concentration are correlated, error in the exposure estimates will be
28 larger in locations where there are more prevalent or stronger sources or at times when
29 NO2 emissions are higher for a given location. For example, it would be anticipated that
30 the magnitude of the instrument error is largest at times of day when emissions are
31 highest, such as rush hour. Instrument error was also observed to exhibit some
32 autocorrelation at 1- and 2-day lags in the Goldman et al. (2010) study. Hence, the
33 diurnal variability in relative NC>2 instrument error does not change substantially from
34 day to day. For epidemiologic studies of short-term NC>2 exposure, the influence of
January 2015 3-56 DRAFT: Do Not Cite or Quote
-------
1 instrument error would not be expected to change if the health data were obtained on a
2 daily basis.
3.4.4 Confounding
3 To assess the independent effects of NO2 in an epidemiologic study of the health effects,
4 it is necessary to identify (Bateson et al.. 2007): (1) which copollutants (e.g., PM25, UFP,
5 BC) and additional exposures (e.g., noise, traffic levels) are potential confounders of the
6 health effect-NO2 relationship so that their correlation with NO2 can be tested and, if
7 needed, they are accounted for in the epidemiologic model; (2) the time period over
8 which correlations might exist so that potential confounders are considered appropriately
9 for the time period relevant for the epidemiologic study design (e.g., pollutants or other
10 factors that are correlated over the long term might not be important for a short-term
11 exposure epidemiologic study); and (3) the spatial correlation structure across multiple
12 pollutants, if the epidemiologic study design is for long-term exposure. This section
13 considers temporal copollutant correlations and how relationships among copollutants
14 may change in space. Temporal copollutant correlations are computed from the time
15 series of concentrations for two different collocated pollutants. Temporal correlations are
16 informative for epidemiologic studies of short-term exposure when the sampling interval
17 is a month or less for each of the copollutants. Temporal correlations are informative for
18 epidemiologic studies of long-term exposures when sampling intervals are months-to-
19 years. Spatial relationships are evaluated by comparing within-pollutant variation across
20 space for different pollutants. The following sections review coexposures that can
21 potentially confound the relationship between a health effect and NO2 over different
22 temporal and spatial resolutions.
3.4.4.1 Temporal Relationships among Ambient Nitrogen
Dioxide and Copollutant Exposures
23 Studies and analyses reported in the 2008 ISA for Oxides of Nitrogen (U.S. EPA. 2008)
24 demonstrated that ambient NO2 was correlated with several traffic-related pollutants in
25 urban and suburban areas generally in the range of Pearson R = 0.5 to 0.75 for PM2 5 and
26 CO and R = 0.8 to 0.9 for EC. These results suggest that in some cases NO2 can be a
27 surrogate for traffic pollution (U.S. EPA. 2008). In contrast, correlations between NO2
28 and Os were generally R = -0.71 to 0.1. Numerous air quality, exposure, and
29 epidemiologic studies have more recently evaluated associations between concentrations
30 of ambient NO2 and those of other pollutants. Many of these studies report Pearson or
31 Spearman correlations of ambient NO2 with other criteria pollutants, mainly focusing on
January 2015 3-57 DRAFT: Do Not Cite or Quote
-------
1 those related to traffic sources (PM2 5, CO, PMio). A few studies have explored
2 associations between NO2 and other traffic-related pollutants, such as EC, UFP, and
3 VOCs. Data for correlations between NO2 and other criteria pollutants are summarized in
4 Table 3-8. broken into short-term exposure and long-term exposure epidemiology studies.
5 Figure 3-6 plots data for correlations between NO2 and all copollutants for which data
6 were available, including PM2 5, PMio PMio-2.5, O3, CO, SO2, EC, OC, UFP, particle
7 number concentration (PNC), toluene, and benzene. Figure 3-6 separates the data by
8 averaging period. "Within-hourly" denotes averaging time ranging from 20 seconds to
9 1-h daily max. "Within daily" is noted for averaging time ranging from 3 to 24 hours.
10 Three hour averaging times are typically applied during rush hour measurement periods.
11 "Within monthly" refers to averaging times ranging from 84 hours to 1 month. "Annual
12 or longer-term correlations" are for studies that averaged the data over a period of 1 to
13 5 years. The studies presented in Table 3-8 only include monitored data and not
14 correlations computed from LUR studies. Some of these studies used personal or area
15 sampling in lieu of central site monitoring. Note that, while Table 3-8 and Figure 3-6 are
16 informative for considering the influence of averaging time on correlations, small sample
17 sizes for any given pollutant and averaging period preclude making definitive
18 conclusions about the observations. In particular, the number of near-road studies
19 reporting correlations between NO2 and copollutants was too small to make any
20 conclusions about differences in NO2-copollutant correlations between near-road and
21 central site or personal measures.
22 The higher the copollutant correlation, the more difficult it is to disentangle the health
23 effects from NO2 from those of the copollutants. This is particularly true of traffic-related
24 copollutants, and recent evidence indicates that copollutant confounding adds such
25 uncertainty. Figure 3-6 shows the range of temporal NO2 copollutant correlation
26 coefficients among the studies in Table 3-8 plus one additional measurement study that
27 did not include other criteria air pollutants (Williams et al., 2012a). Existing studies
28 indicate that NO2 has, in general, correlations over Pearson R = 0.5 with other NAAQS
29 and traffic-related pollutants. Similar to findings in the 2008 ISA for Oxides of Nitrogen
30 (U.S. EPA. 2008) the strongest temporal correlations are typically observed for NO2 with
31 primary traffic-related pollutants, such as benzene, CO, EC, and PNC. A wide range of
32 temporal correlations is observed forNO2 with PM2s, PMio, and SO2. Correlations of
33 NO2 with PM2 5 and PMio tend to be positive for the within-hourly, within-daily, and long
34 term metrics. For the within-monthly measures, median correlations are closer to zero.
35 The reason for this difference is unknown, but fewer data are available for the
36 within-monthly correlations. The lowest temporal correlations are typically observed for
37 NO2 with Os and PMio-2 5, with correlations having a wide range in magnitude (R = -0.71
38 to 0.66; median R = 0.15). These observations are not surprising given the nonlinear
39 relationship between NO2 concentration and instantaneous Os production rate observed
January 2015 3-58 DRAFT: Do Not Cite or Quote
-------
1 close to the location of emission (Pusede and Cohen. 2012; LaFranchi et al.. 2011;
2 Murphy et al.. 2007. 2006). Temporal correlations for near-road studies are highlighted in
3 red for Figure 3-6. It is notable that the near-road correlations did not appear to be
4 systematically different from the urban scale correlations. Statistical testing for near-road
5 versus urban scale interpollutant correlations was not performed given the small number
6 of near-road studies.
Short-Term Temporal Correlations
7 For the shorter time periods (within hourly and within daily), UFP, BC, CO, and EC
8 tended to have higher correlations with NO2, while Os had several negative correlations
9 with NO2. The within-daily category had the most data for PM2 5 and PMio, and a wide
10 range of correlations was observed with NO2 for each of those copollutants. Fewer data
11 were available for within-monthly correlations. Black carbon, benzene, and toluene were
12 observed to have the highest correlations with NO2 in this temporal category. Across
13 time-averaging periods, there is not a discernible pattern with respect to correlations of
14 near-road measurements.
January 2015 3-59 DRAFT: Do Not Cite or Quote
-------
Table 3-8 Synthesis of nitrogen dioxide ambient-ambient copollutant correlations from measurements
reported in the literature.
Study3 Averaging Time
Location Scale
Correlation
Measure CO Os
SO2 PM2.5 PMlO
Short-term exposure epidemiologic studies
Polidori and Fine (2012) 1-min
Lew et al. (2014) <2 min
Padre-Martinez et al. (2012) 2 min
Chuanq et al. (2008) Hourly
Los Angeles, CA Near-road
(15m downwind of
1-710) summer
Los Angeles, CA Near-road
(80 m downwind of
1-710) summer
Los Angeles, CA Urban
(background) summer
Los Angeles, CA Near-road
(15m downwind of
1-710) winter
Los Angeles, CA Near-road
(80 m downwind of
1-710) winter
Los Angeles, CA Urban
(background) winter
Montreal, Canada (all Urban
year)
Montreal, Canada Urban
(summer)
Montreal, Canada Urban
(fall)
Montreal, Canada Urban
(winter)
Boston, MA Urban
Boston, MA Urban
Pearson 0.65 NRb
Pearson 0.65 NR
Pearson 0.66 NR
Pearson 0.60 NR
Pearson 0.62 NR
Pearson 0.79 NR
Pearson 0.77 -0.74
Pearson 0.40 -0.33
Pearson 0.16 -0.36
Spearman 0.51 NR
Pearson NR NR
NR NR NR
NR NR NR
NR NR NR
NR NR NR
NR NR NR
NR NR NR
0.11 0.29 0.39
0.17 0.34 0.35
0.25 0.26 0.30
0.04 0.34 0.35
NR 0.21 NR
NR 0.38 0.33
January 2015
3-60
DRAFT: Do Not Cite or Quote
-------
Table 3-8 (Continued): Synthesis of nitrogen dioxide ambient-ambient correlations from measurements reported
in the literature.
Study3
Strickland et al. (2010)
Villeneuve et al. (2007)
Jalaludinetal. (2007)
Mortimer et al. (2002)
Burnett et al. (2000)
Mar et al. (2000)
Tolbert et al. (2007)
Darrowetal. (2011)
Moshammer et al. (2006)
Darrowetal. (2011)
Averaging Time
1-h daily max
1-h daily max
1-h daily max
1-h daily max
1-h daily max
1-h daily max
1-h daily max
1-h daily max
Morning commute
(7:00 a.m. -10:00 a.m.)
Daytime
(8:00 a.m. -7:00 p.m.)
Nighttime
(12:00 a.m.-6:00 a.m.)
8-h avg
24-h avg
Location
Atlanta, GA (cold
season)
Atlanta, GA (warm
season)
Edmonton, Canada
Sydney, Australia
8 U.S. cities
8 Canadian cities
Phoenix, AZ
Atlanta, GA
Atlanta, GA
Atlanta, GA
Atlanta, GA
Atlanta, GA
Linz, Austria
Atlanta, GA
Scale
Urban
Urban
Urban
Urban
Urban
Urban
Urban
Urban
Urban
Urban
Urban
Urban
Urban
Urban
Correlation
Measure
Spearman
Spearman
Pearson
NR
NR
NR
NR
Spearman
Partial
Spearman
Partial
Spearman
Partial
Spearman
Partial
Spearman
Pearson
Partial
Spearman
CO
0.59
0.54
0.74
0.6
NR
0.65
0.87
0.7
0.61
0.57
0.53
0.66
NR
0.66
03
0.11
0.42
NR
0.25
0.27
0.12
NR
0.44
0.40
-0.16
-0.07
-0.66
NR
0-1 C
SO2
0.36
0.37
NR
0.46
NR
0.49
0.57
0.36
NR
NR
NR
NR
NR
NR
PM2.5
0.37
0.36
NR
0.65
NR
0.53
0.77
0.47
0.50
0.46
0.41
0.52
0.54
0.20
PMio
0.46
0.44
NR
0.48
NR
0.53
0.53
0.53
NR
NR
NR
NR
0.62
NR
January 2015
3-61
DRAFT: Do Not Cite or Quote
-------
Table 3-8 (Continued): Synthesis of nitrogen dioxide ambient-ambient correlations from measurements reported
in the literature.
Study3
Faustini et al. (2011)
Samolietal. (2011)
Ko et al. (2007)
Mehta et al. (2013)
Andersen et al. (2008)
Mannes et al. (2005)
Averaging Time Location
24-h avg Milan
Mestre
Turin
Bologna
Florence
Pisa
Rome
Cagliari
Taranto
Palermo
24-h avg Athens, Greece
24-h avg Hong Kong
24-h avg Ho Chi Minh City,
Vietnam (dry season)
Ho Chi Minh City,
Vietnam (wet season)
24-h avg Copenhagen,
Denmark
24-h avg Sydney, Australia
Scale
Urban
Urban
Urban
Urban
Urban
Urban
Urban
Urban
Urban
Urban
Urban
Urban
Urban
Urban
Near-road
Urban
Correlation
Measure
Pearson
Pearson
Pearson
Pearson
Pearson
Pearson
Pearson
Pearson
Pearson
Pearson
NR
Pearson
NR
NR
Spearman
Pearson
CO
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
0.57
03
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
0.34
0.44
0.17
OCQ
0.29
SO2
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
0.55
0.66
0.29
0.01
NR
NR
PM2.5
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
0.44
NR
NR
0.41
0.66
PMio
0.79
0.66
0.72
0.66
0.65
0.57
0.5
0.23
0.19
0.22
NR
0.4
0.78
0.18
0.43
0.47
January 2015
3-62
DRAFT: Do Not Cite or Quote
-------
Table 3-8 (Continued): Synthesis of nitrogen dioxide ambient-ambient correlations from measurements reported
in the literature.
Study3
Schildcrout et al. (2006)
Liu et al. (2009)
Straketal. (2013)
O'Connor et al. (2008)
Timonen et al. (2006)
Quo et al. (2009)
Roias-Martinez et al. (2007)
Sarnatetal. (2001)
Sarnat et al. (2005)
Averaging Time
24-h avg
24-h avg
24-h avg
24-h avg
24-h avg
24-h avg
24-h avg
24-h avg
24-h avg
Location
Albuquerque, NM
Baltimore, MD
Boston, MA
Denver, CO
San Diego, CA
St. Louis, MO
Toronto, Canada
Ontario, Canada
Locations across the
Netherlands
Inner-cities across the
U.S.
Amsterdam,
the Netherlands
Erfurt, Germany
Helsinki, Finland
Beijing, China
Mexico City, Mexico
Baltimore, MD
(summer)
Baltimore, MD
(winter)
Boston, MA (summer)
Boston, MA (winter)
Scale
Urban
Urban
Urban
Urban
Urban
Urban
Urban
Urban
Urban
Urban
Urban
Urban
Urban
Urban
Urban
Urban
Urban
Near-road
Near-road
Correlation
Measure
NR
NR
NR
NR
NR
NR
NR
Spearman
Spearman
NR
Spearman
Spearman
Spearman
Pearson
Pearson
Spearman
Spearman
Spearman
Spearman
CO
0.76
0.69
0.8
0.85
0.92
0.71
0.63
NR
NR
0.54
0.76
0.86
0.32
NR
NR
0.75
0.76
NR
NR
03
0.04
0.44
0.47
0.24
0.39
0.42
0.4
-0.51
ORO
OO-l
NR
NR
NR
NR
0.17
0.02
-0.71
NR
NR
SO2
NR
0.49
0.68
0.56
0.23
0.58
0.63
0.18
NR
0.59
NR
NR
NR
0.53
NR
NR
-0.17
NR
NR
PM2.5
NR
NR
NR
NR
NR
NR
NR
0.71
0.45
0.59
0.49
0.82
0.35
0.67
NR
0.37
0.75
0.44
0.64
PMio
0.26
0.62
0.48
0.64
0.55
0.45
0.64
NR
0.49
NR
NR
NR
NR
NR
0.25
NR
NR
NR
NR
January 2015
3-63
DRAFT: Do Not Cite or Quote
-------
Table 3-8 (Continued): Synthesis of nitrogen dioxide ambient-ambient correlations from measurements reported
in the literature.
Study3 Averaging Time
Kim etal. (2006) 24-h avg
Roberts and Martin (2006) 24-h avg
Andersen et al. (2007) 24-h avg
Chen etal. (2008) 24-h avg
Arhami et al. (2009) 24-h avq
Delfino et al. (2009) 24-h avq
Baxter et al. (201 3) 24-h avq
Williams et al. (2012c) 24-h avg
Location
Toronto, Canada
Cleveland, OH
Nashville, TN
Copenhagen,
Denmark
Shanghai, China
San Gabriel Valley,
CA (summer/fall)
San Gabriel Valley,
CA (fall/winter)
Riverside, CA
(summer/fall)
Riverside, CA
(fall/winter)
San Gabriel Valley
and Riverside, CA
(aggregated)
Boston, MA
Pittsburgh, PA
Memphis, TN
Detroit, Ml
Milwaukee, Wl
San Diego, CA
Riverside, CA
Research Triangle
Park, NC
Scale
Near-road
Urban
Urban
Urban
Urban
Urban
Urban
Urban
Urban
Urban
Urban
Urban
Urban
Urban
Urban
Urban
Urban
Urban
Correlation
Measure
Spearman
NR-Pairwise
NR-Pairwise
Spearman
NR
Spearman
Spearman
Spearman
Spearman
NR
Spearman
Spearman
Spearman
Spearman
Spearman
Spearman
Spearman
Spearman
CO
0.72
0.67
0.36
0.74
NR
NR
NR
NR
NR
0.79
NR
NR
NR
NR
NR
NR
NR
NR
03
NR
0.36
0.26
NR
NR
NR
NR
NR
NR
0/1O
NR
NR
NR
NR
NR
NR
NR
0-1 O
SO2
NR
0.56
0.08
NR
0.73
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
PM2.5
0.44
NR
NR
NR
NR
0.1
0.44
0.07
0.56
0.19
0.41
0.46
0.27
0.59
0.55
0.57
0.37
0.03
PMio
NR
0.63
0.44
0.42
0.71
0.31
0.34
0.21
0.64
NR
NR
NR
NR
NR
NR
NR
NR
NR
January 2015
3-64
DRAFT: Do Not Cite or Quote
-------
Table 3-8 (Continued): Synthesis of nitrogen dioxide ambient-ambient correlations from measurements reported
in the literature.
Study3
Williams etal. (2012a)
Delfinoetal. (2008)
Suh and Zanobetti (2010)
Schembari etal. (2013)
Laurent etal. (2013)
Peters et al. (2009)
Sanchez Jimenez et al.
(2012)
Steinvil et al. (2009)
Steinvil et al. (2008)
Taoetal. (2012)
Wichmann et al. (2012)
Averaging Time
24-h avg
24-h avg
24-h avg
24-h avg
24-h avg
24-h avg
24-h avg
24-h avg
24-h avg
24-h avg
24-h avg
Location
Detroit, Ml
Los Angeles, CA
Atlanta, GA
Barcelona, Spain
Los Angeles and
Orange counties, CA
Erfurt, Germany
Glasgow, U.K.
Glasgow, U.K.
Glasgow, U.K.
London, U.K.
London, U.K.
Tel Aviv, Israel
Tel Aviv, Israel
Guangzhou, Foshan,
Zhongshan, and
Zhuhai, China
Copenhagen,
Denmark (warm
period)
Copenhagen,
Denmark (cold
period)
Scale
Near-road
Urban
Urban
Urban
Urban
Urban
Near-road
Background
Background
Near-road
Background
Urban
Urban
Urban-
regional
Urban
Urban
Correlation
Measure
Spearman
Spearman
Spearman
Spearman
Pearson
Spearman
Spearman
Spearman
Spearman
Spearman
Spearman
Partial
Pearson
Partial
Pearson
Pearson
Spearman
Spearman
CO
NR
NR
NR
NR
0.83
0.68
0.6
0.4
0.74
0.3
0.61
0.75
0.86
0.72
0.62
0.72
03
NR
NR
NR
NR
OQ-1
OCC
NR
NR
NR
NR
NR
OrtA
070
0.17
NR
NR
SO2
NR
NR
NR
NR
NR
0.54
NR
NR
NR
NR
NR
0.70
0.72
0.82
NR
NR
PM2.5
NR
0.36
0.47
0.41
0.77
0.63
NR
NR
NR
0.49
0.42
NR
NR
NR
NR
NR
PMio
NR
NR
NR
NR
0.70
0.64
0.83
0.69
NR
0.67
0.37
0.076
0.082
0.82
0.47
0.46
January 2015
3-65
DRAFT: Do Not Cite or Quote
-------
Table 3-8 (Continued): Synthesis of nitrogen dioxide ambient-ambient correlations from measurements reported
in the literature.
Study3 Averaging Time Location
Dimitriou and Kassomenos 24-h avg London, U.K. (cold
(2014) period)
London, U.K. (warm
period)
Paris, France (cold
period)
Paris, France (warm
period)
Copenhagen,
Denmark (cold
period)
Copenhagen,
Denmark (warm
period)
Hamburg, Germany
(cold period)
Hamburg, Germany
(warm period)
Scale
Urban
Near road
Urban
Near road
Urban
Near road
Urban
Near road
Urban
Near road
Urban
Near road
Urban
Near road
Urban
Near road
Correlation
Measure
Pearson
Pearson
Pearson
Pearson
Pearson
Pearson
Pearson
Pearson
Pearson
Pearson
Pearson
Pearson
Pearson
Pearson
Pearson
Pearson
CO
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
03
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
SO2
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
PM2.5
0.52
0.49
0.63
0.60
0.65
0.60
0.54
0.75
0.31
0.36
0.42
0.53
0.21
0.40
0.50
0.69
PMio
0.49
0.70
0.56
0.67
0.71
0.68
0.50
0.83
0.35
0.37
0.42
0.55
0.23
0.52
0.51
0.70
January 2015
3-66
DRAFT: Do Not Cite or Quote
-------
Table 3-8 (Continued): Synthesis of nitrogen dioxide ambient-ambient correlations from measurements reported
in the literature.
Study3 Averaging Time
Dimitriou and Kassomenos 24-h avg
(2014)
(Continued)
Clouqhertv et al. (2013) 84-h avq
Sarnat et al. (2012) 96-h avg
Greenwald et al. (2013) 96-h avg
Wheeler et al. (2008) 2-week
Trasande et al. (2013) 1 -month avq
Location
Stockholm, Sweden
(cold period)
Stockholm, Sweden
(warm period)
New York, NY
El Paso, TX (site A)
El Paso, TX (site B)
Ciudad Juarez,
Mexico (site A)
Ciudad Juarez,
Mexico (site B)
2 sites in El Paso, TX
Windsor, ON, Canada
(all year)
Windsor, ON, Canada
(winter)
Windsor, ON, Canada
(spring)
Windsor, ON, Canada
(summer)
Windsor, ON, Canada
(fall)
United States
Scale
Urban
Near road
Urban
Near road
Urban
Urban
Near-road
Urban
Near-road
Urban
Urban
Urban
Urban
Urban
Urban
Varies
Correlation
Measure
Pearson
Pearson
Pearson
Pearson
Pearson
Spearman
Spearman
Spearman
Spearman
Pearson
Spearman
Spearman
Spearman
Spearman
Spearman
Pearson
CO
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
0.12
03
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
Onoo
SO2
NR
NR
NR
NR
0.51
NR
NR
NR
NR
Ooo
0.85
0.84
0.61
0.51
0.66
O-i n
PM2.5
0.20
0.49
0.38
0.58
0.74
-0.39
-0.28
-0.28
0
0.2
NR
NR
NR
NR
NR
Onnn
PMio
0.24
0.45
0.45
0.52
NR
-0.3
-0.1
-0.1
0.11
0.31
NR
NR
NR
NR
NR
-0.011
Long-term exposure epidemiologic studies
Dadvand et al. (2014) 9-mo
Barcelona, Spain
Urban
Spearman
NR
NR
NR
0.48
0.33
January 2015
3-67
DRAFT: Do Not Cite or Quote
-------
Table 3-8 (Continued): Synthesis of nitrogen dioxide ambient-ambient correlations from measurements reported
in the literature.
Study3
Katanodaetal. (2011)
Donqetal. (2011)
Hwanq and Lee (2010)
Heinrichetal. (2013)
Ducret-Stich et al. (2013)
Eeftensetal. (2012)
Averaging Time Location
1-yr avg Japanese cities
1-yr avg 7 cities across China
1-yr avg 14 Taiwanese
communities
1-yr avg North Rhine-
Westphalia, Germany
1-yr avg Swiss Alps
1-yr avg Oslo, Norway
Stockholm County,
Sweeden
Helsinki/Turku,
Finland
Copenhagen,
Denmark
Kaunas, Lithuania
Manchester, England
London/Oxford,
England
Netherlands/Belgium
Ruhr Area, Germany
Munich/Augsberg,
Germany
Vorarlberg, Austria
Paris, France
Scale
Urban
Urban
Urban
Urban
Near-road
On highway
Urban
Urban
Urban
Urban
Urban
Urban
Urban
Urban
Urban
Urban
Urban
Urban
Correlation
Measure
Pearson
NR
NR
Spearman
Spearman
Spearman
Spearman
Spearman
Spearman
Spearman
Spearman
Spearman
Spearman
Spearman
Spearman
Spearman
Spearman
Spearman
CO
NR
0.23
0.86
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
03
NR
0.66
Or\~7
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
SO2
0.76
0.52
0.55
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
PM2.5
NR
NR
0.37
0.50
NR
NR
0.24
0.75
0.71
0.40
0.04
0.40
0.84
0.57
0.69
0.29
0.04
0.86
PMio
NR
0.7
NR
NR
0.51
0.04-0.63
0.34
0.80
0.80
0.60
0.17
0.59
0.82
0.74
0.65
0.67
0.35
0.91
January 2015
3-68
DRAFT: Do Not Cite or Quote
-------
Table 3-8 (Continued): Synthesis of nitrogen dioxide ambient-ambient correlations from measurements reported
in the literature.
Study3
Eeftensetal. (2012)
(Continued)
McConnell et al. (2003)
Ganetal. (2012a)
Averaging Time Location
1-yravg Gyor, Hungary
(Continued)
Lugano, Switzerland
Turin, Italy
Rome, Italy
Barcelona, Spain
Catalunya, Spain
Athens, Greece
Heraklion, Greece
4-yr avg 12 communities in
southern California
5-yr avg Vancouver, Canada
Scale
Urban
Urban
Urban
Urban
Urban
Urban
Urban
Urban
Urban
Urban
Correlation
Measure
Spearman
Spearman
Spearman
Spearman
Spearman
Spearman
Spearman
Spearman
Pearson
Spearman
CO
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
03
NR
NR
NR
NR
NR
NR
NR
NR
0.59
NR
SO2
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
PM2.5
0.02
0.66
0.65
0.73
0.90
0.72
0.49
0.18
0.54
0.47
PMio
0.12
0.83
0.67
0.75
0.69
0.63
0.70
0.37
0.2
NR
avg = average; CO = carbon monoxide; NR = not reported; O3 = ozone; PM2 5 = in general terms, particulate matter with an aerodynamic diameter less than or equal to a nominal 2.5 |jm, a
measure of fine particles; PMio = in general terms, particulate matter with an aerodynamic diameter less than or equal to a nominal 10 |jm, a measure of thoracic particles; SO2 = sulfur
dioxide.
"•Correlation data computed from LUR studies are not included here.
January 2015
3-69
DRAFT: Do Not Cite or Quote
-------
Within-Hourly Correlations
Within-Daily Correlations
sop] -
F-IC -
PWPL5] -
PM[10-2.5] -
PMI1D] -
oc -
0[3] -
EC -
CO -
Benzene -
BC -
PM|IO-J.51
PM|10|
Within-Monthly Correlations
Annual or Longer-Term Correlations
OC -
.;;,;,] _
i;i -
CO -
t
*
m
UFP -
Toluene -
SOP] -
PNC -
PMP-5] -
PW(10-2.5) -
PMI1D] -
OC -
0[3] -
EC -
CO -
Benzene —
BC -
Correlation
Correlation
BC = black carbon; CO = carbon monoxide; EC = elemental carbon; LUR = land use regression; O[3] = ozone; OC = organic
carbon; PM[2.5] = in general terms, participate matter with an aerodynamic diameter less than or equal to a nominal 2.5 |jm, a
measure of fine particles; PM[10] = in general terms, particulate matter with an aerodynamic diameter less than or equal to a
nominal 10 |jm, a measure of thoracic particles; PM[10-2.5] = 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 measure of thoracic coarse particles; PNC = particle
number concentration; SO2 = sulfur dioxide; UFP = ultrafine particulate matter.
Notes: Boxes represent the interquartile range of the data with the median line plotted, and 90th and 10th percentile of the data are
plotted as the whiskers. Correlation data computed from LUR studies are not included here. Correlations shown by closed red
circles come from near-road studies, and correlations shown by open black circles either come from urban-regional scale studies or
do not specify the study's spatial scale.
Source: National Center for Environmental Assessment analysis of data from studies referenced in Table 3-8.
Figure 3-6 Summary of temporal nitrogen dioxide-copollutant correlation
coefficients from measurements reported in studies listed in
Table 3-8. sorted by temporal averaging period.
January 2015
3-70
DRAFT: Do Not Cite or Quote
-------
1 Fewer studies have explored seasonal correlations between NO2 and copollutants. Among
2 these, a majority of studies report correlations of NC>2 with PM2s and PMio. In general,
3 studies show stronger correlations of NC>2 with PM2s and PMio during cooler seasons.
4 Connell et al. (2005) investigated associations between PIVb 5 and gaseous copollutants in
5 Steubenville, OH using linear regression. NC>2 was more strongly correlated with PIVb 5
6 during the fall (R2 = 0.53) and winter (R2 = 0.53) seasons compared with the spring
7 (R2 = 0.27) and summer (R2 = 0.086) seasons. Similarly, Sarnat et al. (2005) found
8 positive associations between PIVb 5 and NC>2 during both seasons (summer: /? = 0.44;
9 winter: /? = 0.64), with stronger associations in the winter in Baltimore, MD. Arhami
10 et al. (2009) evaluated relationships between ambient copollutants at two sites in southern
11 California (San Gabriel Valley and Riverside) for warmer and cooler seasons. During the
12 warm season, the Spearman correlation coefficient (average among sites) was r = 0.09
13 between NCh and PlVfc 5, whereas during the winter the correlation was r = 0.50.
14 However, they did not observe a consistent seasonal trend between NO2 and PMio. While
15 associations between NO2 and PMio were substantially lower during the summer
16 (r = 0.21) at the Riverside site, correlations were relatively similar during both seasons at
17 the San Gabriel Valley site (summer PMio: r = 0.31; winter PMio: r = 0.34). In contrast,
18 for a study of copollutant variation in Montreal, Canada, Levy et al. (2014) reported
19 higher magnitude Pearson correlations for several copollutants in summer (CO: R = 0.77;
20 O3: R = -0.74; SO2: R = 0.17; PM25: R = 0.34; UFP: R = 0.77; BC: R = 0.80;
21 PMio: R = 0.35) compared with winter (CO: R = 0.16; O3: R = -0.36; SO2: R = 0.04;
22 PM25: R = 0.34; UFP: R = 0.71; BC: R = 0.0.55; PMio: R = 0.35). The Levy et al. (2014)
23 study measured the pollutants using near-real-time instrumentation with recording
24 intervals ranging from 1 second to 2 minutes.
25 The relationship between NO2 and Os may also have seasonal patterns, although limited
26 seasonal data exist between these two pollutants. In the 2008 ISA for Oxides of Nitrogen
27 (U.S. EPA. 2008). ambient concentrations of NO2 and Os from several sites across Los
28 Angeles, CA were compared during a multiyear period. Slightly positive correlations
29 between these two pollutants were observed during the summer (Spearman r = 0.0 to
30 0.4), while negative correlations were observed during the winter (r = -0.5 to -0.8). The
31 slightly positive correlations during the summer can be attributed in part to increased
32 photochemical activity, resulting in enhanced Os formation. Higher Os concentrations
33 increase the ratio of NO2 to NO due to enhanced oxidation, thereby resulting in a stronger
34 correspondence between NO2 and Os during the summer. The magnitude of the
35 relationship between NO2 and Os may be dampened by the nonlinear relationship
36 between the two species (Pusede and Cohen. 2012). Only one study in Table 3-8 reported
37 seasonal differences in the correlation between NO2 and Os. Sarnat et al. (2001) measured
38 daily concentrations of gaseous and PM pollutants during different seasons in Baltimore,
39 MD. Similar to the trends reported in the 2008 ISA for Oxides of Nitrogen, they observed
January 2015 3-71 DRAFT: Do Not Cite or Quote
-------
1 a negative correlation between NO2 and Os during the winter (r = -0.71) and a near-zero
2 correlation during the summer (r = 0.02). However, because there is a lack of studies
3 reporting such correlations, it is uncertain whether or not this seasonal trend exists
4 between the two pollutants in different locations.
5 Recent studies have also compared NO2-copollutant temporal correlations across
6 different regions in the U.S., based on central site monitoring data. Baxter et al. (2013)
7 studied differences in air pollution for the Northeast (Boston, MA; Pittsburgh, PA), South
8 (Memphis, TN; Birmingham, AL), Midwest (Milwaukee, WI; Detroit, MI), and West
9 (San Diego, CA; Riverside, CA). Average Spearman correlation coefficients between
10 PM2 5 and NO2 for each region were different (Northeast: r = 0.44; South [data available
11 for Memphis only]: r = 0.27; Midwest: r = 0.57; West: r = 0.47). Schildcrout et al. (2006)
12 compared a number of gaseous and particulate pollutants in different cities across the
13 U.S., including Albuquerque, NM; Baltimore, MD; Boston, MA; and Denver, CO. While
14 correlations between ambient NO2 and CO were relatively similar in all four locations,
15 larger differences were observed between NO2 and PMio correlations, ranging from a
16 Spearman correlation of r = 0.64 in Denver to r = 0.26 in Albuquerque. Other multicity
17 studies conducted outside of the U.S. show that NO2 copollutant correlations are widely
18 variable across cities (Faustini etal.. 2011: Dales etal.. 2010. 2009: Stieb etal.. 2008:
19 Timonen etal.. 2006).
20 A small subset of studies investigated temporal correlations between NO2 and
21 traffic-related VOCs, such as BTEX. In these studies, correlations between NO2 and
22 VOCs are variable. Brook et al. (2007) demonstrated that benzo(e)pyrene and hopanes,
23 specific mobile source tracers, were more strongly correlated with NO2 (Spearman
24 r = 0.27-0.80) compared to PM2 5 (r = 0.26-0.62) at several urban sites in Canada.
25 Beckerman et al. (2008) observed correlations between NO2 and BTEX of Pearson
26 R = 0.46-0.85 in a near-road field campaign. In a panel study, Greenwald et al. (2013)
27 compared ambient concentrations of traffic pollutants monitored outside two schools in
28 El Paso, TX, including one school within close proximity to a major roadway with heavy
29 diesel truck traffic. A Spearman correlation of r = 0.77 was observed between NO2 and
30 BTEX, suggesting that both pollutants are related to traffic sources.
Long-Term Temporal Correlations
31 Long-term exposure epidemiology studies for which interpollutant correlations were
32 computed were substantially less numerous than short-term exposure epidemiology
33 studies (Atkinson et al.. 2013; Heinrich et al.. 2013; Gan etal.. 2012a; Darrow et al..
34 2011: Dong etal.. 2011: Katanoda etal.. 2011: Hwang and Lee. 2010: Delfino et al..
35 2009; Delfino et al.. 2008; McConnell et al.. 2003). For long-term averages, the most data
January 2015 3-72 DRAFT: Do Not Cite or Quote
-------
1 were for PM2 5 and PMio. In each case, the median correlations were near 0.5, and the
2 correlations were positive and ranging from near 0 to near 0.9. The sample size for other
3 copollutants was low in the long-term averages. Median correlations were comparable
4 between long-term exposure and short-term exposure epidemiology studies for CO, SO2,
5 PM25, BC, and PMio. The largest difference was for the correlation between NO2 and O3,
6 which was 0.59 over the long-term exposure epidemiology studies and 0.17 for all studies
7 pooled. However, given that only three long-term studies were available to compute
8 correlation between NO2 and O3 and one of those three studies reported a negative
9 correlation, there is insufficient information to make a conclusion regarding
10 independence of the effects of NO2 and O3. Long-term correlations were not computed
11 for UFP, EC, OC, PNC, PMio-2.5, benzene, and toluene, and the small relative number of
12 long-term exposure epidemiology studies compared with short-term exposure
13 epidemiology studies reporting temporal correlations add uncertainty to these numbers.
3.4.4.2 Spatial Variability among Ambient Nitrogen Dioxide and
Copollutants
14 When an epidemiologic study design relies on spatial contrasts to draw conclusions, such
15 as for a long-term exposure epidemiologic study, unmeasured spatial correlation between
16 copollutants can lead to positive bias in the health effect estimate for each of the
17 pollutants included in the model. Moreover, bias related to confounding can only be
18 addressed when the spatial scale of variability in the exposure metric is smaller than the
19 spatial scale of variability in the confounder (Paciorek, 2010). Dionisio et al. (2013)
20 compared the coefficient of variation (CV = o/u) of six air pollutants across space using a
21 hybrid AERMOD-background model of concentrations in the Atlanta, GA metropolitan
22 area. They observed the following ordinal relationship of the covariates' CVs:
23 NOX (0.88) > CO (0.58) > EC (0.50) > PM25 (0.13) > O3 (0.07) > SO4 (0.05). Dionisio
24 etal. (2013) did not report the CV of NO2, which would be expected to have a lower CV
25 than NOx. Likewise, Goldman et al. (2012) and Ivy et al. (2008) both used monitoring
26 data from the Atlanta, GA metropolitan area to estimate spatial correlation functions, and
27 they observed that NO2 and NOx, along with CO, SO2, and EC, had substantially steeper
28 spatial correlograms than O3, PMio, PM2.5, SO4, NO3, NH4, and OC. Saiani etal. (2011)
29 also observed that spatial correlation decreased more substantially with distance between
30 monitoring sites for NO2 compared with PMio and O3 when looking at six Italian cities.
31 Changes in correlations across space have been observed in a small number of studies.
32 For their long-term near-road study, Ducret-Stich et al. (2013) point out that the temporal
33 correlations of NO2 with EC and PNC were high close to the highway where they
34 obtained measurements and decreased with increasing distance from the road. This
January 2015 3-73 DRAFT: Do Not Cite or Quote
-------
1 suggests that the influence of NO2 on health effects might be belter detected in an
2 epidemiologic study of long-term exposure when the participants are further from the
3 road so that an independent effect can be detected. The next section examines spatial
4 distributions of those copollutants to shed additional light on how copollutant
5 relationships change in space. Atari et al. (2009) tested the relationship between NO2 and
6 SO2 across individual-level and Census tract-level spatial resolutions, which were
7 estimated by a LUR model developed for testing odor threshold in Sarnia, Canada. They
8 observed higher spatial correlation when averaging over a Census tract (R = 0.65)
9 compared with individual-level resolution (R = 0.49). These findings illustrate greater
10 spatial variability for NO2, NOx, CO, SO2, and EC compared with the other pollutants.
11 Based on the conclusions of Paciorek (2010). the observations noted in Dionisio etal.
12 (2013). Goldman etal. (2012). Ivy et al. (2008). Saiani etal. (2011). Atari et al. (2009).
13 and Sanchez Jimenez et al. (2012) suggest that differences in the spatial variability of
14 NO2 compared with copollutants having different spatial variation make it unlikely that
15 copollutant confounding will occur everywhere in space. This is consistent with the
16 findings of Ducret-Stich et al. (2013) regarding differences in copollutant correlations
17 over space.
3.4.4.3 Personal and Indoor Relationships between Nitrogen
Dioxide and Copollutant Exposures
18 Many studies have investigated the relationship between personal and ambient
19 measurements of NC>2 and other pollutants to evaluate the use of central site
20 measurements as a proxy for personal exposure to pollution. Other studies have explored
21 relationships between indoor NO2 and copollutants to understand sources and personal
22 exposure in an indoor environment. Tables 3-9. 3-10. 3-11. and 3-12 present correlations
23 of ambient, personal, or indoor NO2 with similar measurements of copollutants. A limited
24 number of studies reported in the 2008 ISA for Oxides of Nitrogen (U.S. EPA. 2008)
25 investigated the relationship between personal NO2 and personal or ambient
26 measurements of other pollutants (e.g., PIVb 5, EC, CO, volatile organic compounds, and
27 HONO). Short-term correlation of personal NO2 with these pollutants ranged from
28 Spearman r = 0.26 to r = 0.71. Similar to the results in the 2008 ISA for Oxides of
29 Nitrogen (U.S. EPA. 2008). correlations of r = -0.33 to r = 0.44 were observed between
30 personal NO2 and personal or ambient measurements of other regional (PIVb .5) and
31 traffic-related pollutants (e.g., EC, OC). Additionally, Os consistently showed a negative
32 or no correlation with NO2. More recent studies report indoor NO2 copollutant
33 correlations and observe a broader range of correlations between NO2 and EC of
34 r = -0.37 tor = 0.66.
January 2015 3-74 DRAFT: Do Not Cite or Quote
-------
Table 3-9 Pearson correlation coefficients between
Study
Delfinoetal. (2008)
Suh and Zanobetti (2010)
Williams et al. (2012a)
Schembari et al. (2013)
Location
Los Angeles, CA
Atlanta, GA
Chapel Hill, NC
Barcelona, Spain
n
<170
<277
<357
<65
ambient nitrogen dioxide and
Averaging Times
All: 24-h
All: 24-h
All: 24-h
NO2: 7-day;
PM2.5/EC: 2-day
PM2.5
0.32
0.25
0-1 Q
0.21
personal
EC
0.2
0.33
-0.17
0.44
copollutants.
oc
0.16
NR
NR
NR
03
NR
Onn
OfH
NR
EC = elemental carbon; NR = not reported; O3 = ozone; OC = organic carbon; PM2 5 = in general terms, participate matter with an aerodynamic diameter less than or equal to a
nominal 2.5 |jm, a measure of fine particles.
Table 3-10 Pearson correlation coefficients between personal nitrogen dioxide and ambient copollutants.
Study Location
Delfino etal. (2008) Los Angeles, CA
Suh and Zanobetti (2010) Atlanta, GA
Williams et al. (2012a) Chapel Hill, NC
Schembari et al. (2013) Barcelona, Spain
n Averaging Times
<170 All: 24-h
<277 All: 24-h
<326 All: 24-h
<65 NO2: 7-day;
PM2.5/EC: 2-day
PM2.5 EC OC O3
0.21 0.2 0.18 NR
0.2 0.22 NR NR
Ooo n o Nip n OP
0.28 0.22 NR NR
EC = elemental carbon; NR = not reported; O3 = ozone; OC = organic carbon; PM2 5 = in general terms, particulate matter with an aerodynamic diameter less than or equal to a
nominal 2.5 |jm, a measure of fine particles.
January 2015 3-75 DRAFT: Do Not Cite or Quote
-------
Table 3-11 Pearson
Study
Delfinoetal. (2008)
Suh and Zanobetti (2010)
Williams et al. (2012a)
Schembari et al. (2013)
correlation coefficients
Location
Los Angeles, CA
Atlanta, GA
Chapel Hill, NC
Barcelona, Spain
between personal nitrogen
n Averaging Times
<486 All: 24-h
<277 All: 24-h
<326 All: 24-h
<65 NO2: 7-day;
PM2.5/EC: 2-day
dioxide and
PM2.5
0.38
0.29
0.06
0.11
personal copollutants.
EC OC
0.22 0.2
0.49 NR
0.33 NR
0.3 NR
03
NR
OHQ
0-1 -1
NR
EC = elemental carbon; NR = not reported; O3 = ozone; OC = organic carbon; PM2 5 = in general terms, participate matter with an aerodynamic diameter less than or equal to a
nominal 2.5 |jm, a measure of fine particles.
January 2015 3-76 DRAFT: Do Not Cite or Quote
-------
Table 3-12 Correlation coefficients between indoor nitrogen dioxide
Study Location n Averaging Times
Sarnatetal. (201 2)a El Paso, TX (Site A) 15 NO2: 4-day;
PI\/l->c/FP'9 Hau
El Paso, TX (Site B) 15 NO2: 4-day;
PMWFC'9 rlav
Ciudad Juarez, Mexico 15 NCb: 4-day;
f /\\ PMWFC'9-dav
Ciudad Juarez, Mexico 15 NO2: 4-day;
(Sitp R1 PMWFC'9-dav
Greenwald et al. (201 3)b 2 sites in El Paso, TX 18-26 All: 4-day
and indoor copollutants.
PM EC OC
-0.35 (PM2.s) 0.58 NR
-0.26 (PMio-2.s)
-0.19 (PMio)
0.06 (PM2.s) -0.37 NR
0.28 (PMio-2.s)
0.12 (PMio)
-0.29 (PM2.s) 0.66 NR
-0.58 (PMio-2.s)
-0.5 (PMio)
-0.04 (PM2.s) 0.45 NR
-0.5 (PM-IO-2.5)
-0.34 (PMio)
0.76 (PM2.s) 0.45 NR
0.83 (PMio)
03
NR
NR
NR
NR
NR
EC = elemental carbon; NR = not reported; O3 = ozone; OC = organic carbon; PM = particulate matter; PM2 5 = PM2 5 = in general terms, particulate matter with an aerodynamic
diameter less than or equal to a nominal 2.5 |jm, a measure of fine particles; PMio in general terms, particulate matter with an aerodynamic diameter less than or equal to a nominal
10 |jm, a measure of thoracic particles; PMio-2.5 = 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 measure of thoracic coarse particles.
aSpearman correlation.
bPearson correlation.
January 2015
3-77
DRAFT: Do Not Cite or Quote
-------
1 In addition to these findings, higher correlations were typically observed between
2 ambient measurements of NO2 and other traffic-related pollutants (see Section 3.4.3.1)
3 compared to personal measurements [e.g., correlations among personal exposure
4 measurements in Table 3-11; (Schembari et al., 2013; Williams et al., 2012a; Suh and
5 Zanobetti. 2010; Delfino et al.. 2008)]. For example, Suh and Zanobetti (2010) observed
6 a stronger relationship between ambient NO2:EC (r = 0.61) and ambient NO2:PM2s
7 (r = 0.47) compared to personal NO2:EC (r = 0.49) and personal NO2:PM2 5 (r = 0.29).
8 Delfino et al. (2008) observed similar results in the NO2:EC relationship in a health study
9 investigating the relationship between traffic-related pollution and lung function
10 decrements in Los Angeles, CA. While the ambient NO2:EC correlation was r = 0.61,
11 lower correlations were observed for personal NO2:EC (r = 0.22). Additionally, a small
12 number of time-series studies have used NO2 in receptor models to relate health effects to
13 sources/factors (Baxter et al.. 2013; Cakmak et al.. 2009; Halonen et al.. 2009; Mar etal..
14 2000). Each of these studies used factor analysis, the EPA positive matrix factorization
15 method1, or PCA analysis and found high loadings of NO2 and traffic-related copollutants
16 (e.g., EC, OC, CO) on the same factor, which was attributed to traffic-related pollution.
17 Correlations between NO2 and VOCs also suggest different sources for personal
18 exposure. For example, Martins et al. (2012) estimated personal NO2 and BTEX
19 exposure during four 1-week periods using a microenvironmental approach that
20 combined outdoor and indoor concentrations with time-activity patterns. It consistently
21 observed correlations of r = -0.423 to r = 0.138 between NO2 and BTEX during different
22 seasons. The lack of correlation between these pollutants can be attributed in part to
23 differences in sources between indoor and outdoor microenvironments. While exposure
24 to VOCs, namely benzene, was attributed mainly to indoor sources, NO2 was largely
25 associated with traffic sources. These studies emphasize that proximity to roadways and
26 time spent in various indoor and outdoor microenvironments can impact the relationship
27 between NO2 and traffic-related VOCs.
28 Weaker correlations observed between personal measurements of NO2 and other
29 traffic-related pollutants (compared to ambient measurement correlations) suggest that
30 personal exposure to NO2 may include a number of outdoor and indoor sources
31 comprising traffic and nontraffic emissions (e.g., gas stoves, residential wood burning,
32 biomass burning). These observations provide further evidence that nonambient sources
33 of NO2 provide noise to the ambient NO2 signal. At the same time, the weaker
34 correlations between total personal NO2 exposures and copollutant exposures indicate
35 that for panel studies of total NO2 exposure, ambient copollutants would be unlikely to
36 confound health effect estimates for NO2 exposure. Titration conditions for NO, NO2, and
http://intranet.epa.gov/heasd/products/pmf/pmf.htm.
January 2015 3-78 DRAFT: Do Not Cite or Quote
-------
1 Os also likely differ from indoors to outdoors, given variation in solar radiation and other
2 atmospheric factors that influence atmospheric chemistry. Additionally, personal
3 exposures are influenced by building air exchange rate and time-activity patterns that
4 differ among study participants. This is in contrast to ambient NC>2 concentrations, which
5 appear to be largely driven by variability in traffic pollution in many areas. This type of
6 exposure error associated with ambient concentrations is discussed in more detail in
7 Section 3.4.3.3.
8 Few studies have reported indoor NO2 copollutant correlations for short-term averaging
9 times, focusing on correlations between NC>2 and PM in different size fractions as well as
10 NC>2 and BC. In these studies, correlations of Spearman r = -0.37 to 0.66 were observed
11 between indoor NO2 and EC; however, lower correlations are observed for indoor NO2
12 and PM compared with NC>2 and EC. Sarnatetal. (2012) measured indoor concentrations
13 of NC>2, EC, PM2 5, PMio-2.5, and PMio at four elementary schools in two cities near the
14 U.S.-Mexico border: El Paso, TX and Ciudad Juarez, Mexico. NO2 and PM showed
15 weaker and/or inverse correlations at all four elementary schools (r = -0.58 to 0.12).
16 Greenwald et al. (2013) later conducted a follow-up study to Sarnatetal. (2012) and
17 measured similar pollutants at the same schools in El Paso, TX. Although Greenwald
18 etal. (2013) reported similar NO2-EC correlations to those reported in Sarnat et al.
19 (2012). stronger correlations were observed between NO2 and PM2 5 (r = 0.76) and
20 between NCh and PMio (r = 0.83). Differences in the NCh and PM correlations between
21 these two studies reflect that NO2 and PM can have many different sources in indoor
22 environments, which impact their temporal and spatial patterns. Moreover, the results of
23 Greenwald et al. (2013) suggest the potential for confounding of NC>2 health effect
24 estimates by PM based on indoor measurements. Taken together, the existence and extent
25 of such confounding is uncertain.
26 In general, ambient NC>2 would not necessarily be expected to correlate well with
27 personal measures of copollutants. For example, in the case where the exposed
28 population spends time at residences or workplaces sufficiently far from the near-road
29 environment, personal NC>2 exposure would not be expected to correlate with ambient
30 copollutants of traffic-related origin. Low correlations between ambient NC>2 and
31 personal measures of copollutants could support inferences regarding the independent
32 effects of NO2.
3.4.4.4 Traffic and Noise as Confounders
33 For the purpose of inferring causality from the body of epidemiologic studies of
34 short-term and long-term exposure to traffic-related pollutants, the Health Effects
January 2015 3-79 DRAFT: Do Not Cite or Quote
-------
1 Institute Report on Traffic-Related Air Pollution (HEL 2010) raised the concern that
2 distance-to-road models are especially subject to confounding the associations between
3 health effects and exposures because traffic indicators may encompass additional
4 information, such as noise, other air pollutants, stress, and socioeconomic status, that may
5 also be associated with the health effects of interest. However, recent evidence is mixed
6 regarding the correlations of NO and NO2 with traffic and noise levels. Most of these
7 studies are for short-term exposure. Hence, the role of traffic and noise as confounders or
8 independent variables in the relationship between health effects and NO or NO2 is
9 unclear.
10 Several studies have examined the relationship of traffic-related noise with NO and NO2.
11 Kheirbek et al. (2014) added noise level meters to the dense New York, NY monitoring
12 project described in Ross etal. (2013) and observed that 1-week avg noise level, obtained
13 at 60 locations during Fall 2012, correlated with Pearson R = 0.59 for NO2 and R = 0.61
14 for NO. Davies et al. (2009) measured 2-week avg of NO2 and NOx concurrently with
15 5-minute noise samples at 103 sites and observed correlations of R = 0.53 for NO2 and
16 R = 0.64 for NOx. Gan etal. (2012b) calculated the correlations among air pollutants and
17 noise from road traffic and aircraft using 5-min data from 103 sites in Vancouver, BC,
18 Canada during 2003 (dates not stated). They observed lower correlations for NO2 with
19 road traffic noise (Spearman r = 0.33) and aircraft noise (r = 0.14) compared with the
20 correlation of NO with these two noise sources (road traffic: r = 0.41; aircraft: r = 0.26).
21 For both NO2 and NO, correlations were higher for road traffic noise than aircraft noise.
22 Over a 5-yr avg, Gan etal. (2012a) reported the correlation between NO2 and noise from
23 road traffic of Spearman r = 0.33 from Gan etal. (2012b) as well as a correlation between
24 NO and noise from road traffic of Spearman r = 0.39.
25 Ross etal. (2011) also examined relationships of different frequency noises with NO and
26 NO2 using continuous monitors collecting 48,000 samples per second for six 24-hour
27 periods in August 2009. Ross etal. (2011) measured the relationships between traffic
28 level, noise, and concentrations of NO2 and NO in New York, NY as part of the Ross
29 etal. (2013) study. Unweighted noise of all frequencies was uncorrelated with NO2
30 (Spearman r = -0.01) but correlation increased for NO (Spearman r = 0.43) for all times.
31 Correlations were higher for medium frequency noise (NO2: r = 0.22; NO: r = 0.57).
32 Correlations between noise and traffic counts segregated by fleet mix were generally
33 higher for cars (unweighted noise: r = 0.37; medium frequency: r = 0.33), trucks
34 (unweighted noise: r = 0.64; medium frequency: r = 0.71), and buses (unweighted noise:
35 r = 0.61; medium frequency: r = 0.60) compared with the correlations with
36 nonsegregated traffic data. Likewise, at night, high frequency noise was correlated with
37 NO2 (r = 0.83) and NO (r = 0.73).
January 2015 3-80 DRAFT: Do Not Cite or Quote
-------
1 Distance to road has also been observed to influence the relationship between noise and
2 NO2 for both long-term and short-term noise and NC>2 exposure studies. For the years
3 1987-1996, Beelen et al. (2009) estimated correlations among 1-yr avg NCh, traffic level,
4 and noise, and they observed correlations between traffic and noise depending on spatial
5 designation (R = 0.30-0.38) and for the correlation of NC>2 and noise (R = 0.46). When
6 segregating loud noise >65 dBA, correlation dropped (R = 0.22). Note that Beelen et al.
7 (2009) did not specify whether Pearson or Spearman correlations were computed. Ross
8 et al. (2011) noted within-day variability in these relationships, where truck and car
9 traffic are correlated (r = 0.81) during the morning rush hour but inversely correlated at
10 night (r = -0.67). Dadvand et al. (2014) measured 24-h avg noise, NOx, and NC>2 at
11 50-m, 200-m, 500-m, and beyond 500-m buffers from the road in Barcelona, Spain from
12 2001-2005 and observed that all three decreased with increasing distance from the road.
13 Measured temporal Spearman correlation of noise was r = 0.45 for NCh and r = 0.56 for
14 NOx. Allen et al. (2009) also studied the relationship between NO2, UFP, and 5-min avg
15 A-weighted equivalent noise for 105 locations in Chicago, IL and Riverside, CA using
16 measurements taken in December 2006 and April 2007. After adjustment for regional
17 unspecified air pollutant gradients, Pearson correlations with noise were R = 0.16-0.62
18 for NC>2 (winter Chicago: R = 0.16; spring Chicago: R = 0.41; spring Riverside: R = 0.62)
19 and 0.49-0.62 for NO. In Chicago, correlations of noise with NO and NO2 were higher
20 within a 100-m buffer of the road, while correlations of noise with NO and NO2 were
21 lower within a 100-m buffer in Riverside.
22 For short-term exposure studies, more evidence is available to consider the relationship
23 between traffic-related noise and NO2 compared with long-term exposure studies.
24 Collectively, these studies suggest that potential for confounding of NO2 effects by noise
25 may be influenced by temporal and spatial resolution of the data, noise frequency, and
26 fleet mix. Specifically, confounding is less probable as distance from the road increases.
27 However, overall noise may be unlikely to act as a confounder. It should be noted that
28 noise would also have to be etiologically related to the health outcome under
29 consideration to confound the relationship between the health effect and NO2 exposure.
30 When noise is decomposed by frequency, confounding is more likely.
January 2015 3-81 DRAFT: Do Not Cite or Quote
-------
3.4.5 Implications for Epidemiologic Studies of Different Designs
Human health effects related to ambient NO2 exposure can be modeled by:
Y = +E+]Z + e
Equation 3-12
2 where Y= health effect of interest, /?o = model intercept, /?; = health effect estimate for
3 the ambient exposure, Ea = ambient NC>2 exposure, /fe = vector of slope related to each
4 covariate, Z = covariate vector, and e = random error. Y and Ea can sometimes have
5 non-normal distributions, and hence normalization functions such as lognormal or logit
6 may sometimes be applied. For simplicity, Equation 3-12 is presented as a linear
7 function, which is appropriate because most epidemiologic studies assume normality of
8 the data.
9 Estimates of NC>2 exposures are subject to errors that can vary in nature, as described in
10 Section 3.4.3. Classical error is defined as error scattered around the true personal
11 exposure and independent of the measured exposure. Classical error results in bias of the
12 epidemiologic health effect estimate. Classical error can also cause inflation or reduction
13 of the standard error of the health effect estimate. Berkson error is defined as error
14 scattered around the exposure surrogate (in most cases, the central site monitor
15 measurement) and independent of the true value (Goldman et al.. 2011; Reeves et al..
16 1998). Recent studies demonstrate that exposure error is a combination of Berkson-like
17 and classical-like errors and depends on how exposure metrics are averaged across space.
18 Szpiro et al. (2011) defined Berkson-like and classical-like errors as errors sharing some
19 characteristics with Berkson and classical errors, respectively, but with some differences.
20 Specifically, Berkson-like errors occur when the measurement does not capture all of the
21 variability in the true exposure. Berkson-like errors increase the variability around the
22 health effect estimate in a manner similar to Berkson error, but Berkson-like errors are
23 spatially correlated and not independent of predicted exposures, unlike Berkson errors.
24 Classical-like errors can add variability to predicted exposures and can bias health effect
25 estimates in a manner similar to classical errors, but they differ from classical errors in
26 that the variability in estimated exposures is also not independent across space.
27 The results of Meng et al. (2012b). described in Section 3.4.2. illustrate that
28 epidemiologic study design can influence the relationship between personal exposure
29 measurements to NC>2 concentrations and ambient concentrations (see Table 3-7). This
30 meta-analysis found that correlations were highest for short-term exposure community
31 time-series epidemiology studies (designated as "daily average" in Table 3-7). and
January 2015 3-82 DRAFT: Do Not Cite or Quote
-------
1 correlations were lowest for longitudinal panel cohort studies. The following sections
2 consider how exposure assessment errors may influence interpretation of health effect
3 estimates for epidemiologic studies of different designs.
3.4.5.1 Community Time-Series Studies
4 In most short-term exposure epidemiologic studies of the health effects of NO2, the health
5 effect endpoint is modeled as a function of ambient exposure, Ea, which is defined as the
6 product of ambient concentration, Ca, and a, a term encompassing time-weighted
7 averaging and infiltration of NC>2 (Section 3.4.1). Community time-series epidemiologic
8 studies capturing the exposures and health outcomes of a large cohort frequently use the
9 concentration measured at a central site monitor (Ca,csm) as a surrogate for Ea in an
10 epidemiologic model (Wilson et al., 2000). At times, an average of central site monitored
11 concentrations is used for the Ea surrogate. For studies involving thousands of
12 participants, it is not feasible to measure personal exposures. Moreover, for community
13 time-series epidemiology studies of short-term exposure, the temporal variability in
14 concentration is of primary importance to relate to variability in the health effect estimate
15 (Zeger et al.. 2000). Ca,csm can be an acceptable surrogate if the central site monitor
16 captures the temporal variability of the true air pollutant exposure. When averaging
17 across large numbers of individuals, a, which varies between 0 and 1, may quantify the
18 bias introduced by substituting Ca,csm for Ea for the case where a is absorbed into the
19 health effect estimate. Spatial variability in NO2 concentrations across the study area
20 could attenuate an epidemiologic health effect estimate if the exposures are not correlated
21 in time with Ca,csm when central site monitoring is used to represent exposure. If exposure
22 assessment methods that more accurately capture spatial variability in the concentration
23 distribution over a study area are employed, then the confidence intervals around the
24 health effect estimate may decrease. Ca,Csm may be an acceptable surrogate for Ea if the
25 concentration time series at the central site monitor is correlated in time with the
26 exposures.
27 Goldman et al. (2011) simulated the effect of classical-like and Berkson-like errors due to
28 spatiotemporal variability among ambient or outdoor (i.e., a noncentral site monitor
29 situated outside the home) air pollutant concentrations over a large urban area on health
30 effect estimates of emergency department (ED) visits for a time-series study of
31 cardiovascular disease. The relative risk (RR) per ppm was negatively biased in the case
32 of classical-like error (1-h daily max NCh: -1.3%; 1-h daily max NOx: 1.1%) and
33 negligibly positively biased in the case of Berkson-like error (1-h daily max
34 NO2: 0.0042%; 1-h daily max NOX: 0.0030%). The 95% confidence interval range for
35 RR per ppm was wider for Berkson-like error (1-h daily max NCh: 0.028; 1-h daily max
January 2015 3-83 DRAFT: Do Not Cite or Quote
-------
1 NOX: 0.023) compared with classical-like error (1-h daily max NO2: 0.0025; 1-h daily
2 maxNOX: 0.0043).
3 Recent studies have explored the effect of spatial exposure measurement error on health
4 effect estimates to test the appropriateness of using central site monitoring for time-series
5 studies. Goldman et al. (2010) simulated spatial exposure measurement error based on a
6 semivariogram function across monitor sites with and without temporal autocorrelation at
7 1- and 2-day lags to analyze the influence of spatiotemporal variability among ambient or
8 outdoor concentrations over a large urban area on a time-series study of ED visits for
9 cardiovascular disease. A random term was calculated through Monte Carlo simulations
10 based on the data distribution from the semivariogram, which estimated the change in
11 spatial variability in exposure with distance from the monitoring site. The average of the
12 calculated random term was added to a central site monitoring time series (considered in
13 this study to be the base case) to estimate population exposure to NO2 subject to spatial
14 error. For the analysis with temporal autocorrelation considered, RR per ppm for
15 1-h daily max NO2 dropped slightly to 1.0046 (95% CI: 1.0026, 1.0065), and RR per ppm
16 for 1-h daily max NOx dropped to 1.0079 (95% CI: 1.0057, 1.0100) when both were
17 compared with the central site monitor RR per ppm = 1.013 9 (for all air pollutants):.
18 When temporal autocorrelation was not considered, RR per ppm dropped to 1.0044 for
19 1-h daily max NO2 and 1.0074 for 1-h daily max NOx. The results of Goldman et al.
20 (2010) suggest that spatial exposure measurement error from use of central site
21 monitoring data results in biasing the health effect estimate towards the null, but the
22 magnitude of the change in effect was small.
23 Goldman et al. (2012) also studied the effect of different types of spatial averaging on
24 bias in the health effect risk ratio and the effect of correlation between measured and
25 "true" ambient concentrations of NO2, NOx, and other air pollutant measures to analyze
26 the influence of spatiotemporal variability among ambient or outdoor concentrations over
27 a large urban area on health effect estimates. Concentrations were simulated at alternate
28 monitoring locations using the geostatistical approach described above for Goldman et al.
29 (2010) for the 20-county Atlanta metropolitan area for comparison with measurements
30 obtained directly from monitors at those sites. Geostatistical-simulated concentrations
31 were considered to be "true" in this study, and other exposure assessment methods were
32 assumed to have some error. Five different exposure assessment approaches were tested:
33 using a single central site monitor, averaging the simulated exposures across all
34 monitoring sites, performing a population-weighted average across all monitoring sites,
35 performing an area-weighted average across all monitoring sites, and
36 population-weighted averaging of the geostatistical simulation (see Table 3-13). Goldman
that 95% CIs were not reported for the central site monitor RR or for the cases where temporal autocorrelation
was not considered.
January 2015 3-84 DRAFT: Do Not Cite or Quote
-------
1 et al. (2012) observed that the exposure measurement error was somewhat correlated with
2 both the measured and true values, reflecting both Berkson-like and classical-like error
3 components. For the central site monitor, the exposure measurement errors were
4 somewhat inversely correlated with the true value but had relatively higher positive
5 correlation with the measured value. For the other exposure estimation methods, the
6 exposure measurement errors were inversely correlated with the true value, while they
7 had positive but lower magnitude correlation with the measured value. At the same time,
8 the exposure measurement bias, given by the ratio of the exposure measurement error to
9 the measured value, was much higher in magnitude at the central site monitor than for the
10 other estimation methods for NC>2 and for NOx concentrations with the exception of the
11 area-weighted average, which produced a large negative exposure measurement bias.
12 These findings suggest more Berkson-like error in the more spatially resolved exposure
13 metrics compared with the central site monitor and more classical-like error for the
14 converse (i.e., more classical-like error in the central site monitor estimate compared with
15 the other exposure assessment techniques). Hence, more bias would be anticipated for the
16 health effect estimate calculated from the central site monitor, and more variability would
17 be expected for the health effect estimate calculated with the more spatially resolved
18 methods. It was observed that the more spatially variable air pollutants studied in
19 Goldman et al. (2012) also had more bias in the health effect estimates. This was noted
20 across exposure assessment methods but was more pronounced for the central site
21 measurement data.
22 Butland et al. (2013) conducted a simulation study to test how spatial resolution
23 influences health effect estimates in a time-series epidemiologic model of mortality as a
24 function of NCh exposure for urban and rural areas. The test domain was subdivided into
25 squares ranging in area from 1 km2 to 25 km2. Health effect estimates simulated using the
26 1-km2 resolution area were considered to be "true," and mortality estimates were sampled
27 from a Poisson distribution of mortality data. Monitor data were simulated based on a
28 lognormal distribution using the correlelogram among pairs of NC>2 monitors to establish
29 the variability of the distribution as a function of distance. The error structure in the
30 model was constructed to include both Berkson-like and classical-like components.
31 Health effect estimates for mortality based on NO2 exposures were attenuated by 29 and
32 38% for urban and rural areas, respectively, when reducing the spatial resolution from
33 1 km2 to 25 km2 over a 3-year time-series analysis.
January 2015 3-85 DRAFT: Do Not Cite or Quote
-------
Table 3-13 The influence of exposure metrics on error in health effect estimates.
Exposure Estimation Approach
Bias[(Z- Z")I2
3» /?(2
', Z*)b R[(Z- Z
>),zr *[(z-z),zr
NO2
Central site monitor
Unweighted average
Population-weighted average
Area-weighted average
Geostatistical model —
population-weighted average
0.62
0.25
0.18
-0.07
N/A
0.24
0.38
0.38
0.38
0.45
-0.46
-0.73
-0.78
-0.87
-0.82
0.61
0.20
0.14
-0.04
0.0017
NOx
Central site monitor
Unweighted average
Population-weighted average
Area-weighted average
Geostatistical model —
population-weighted average
0.71
0.31
0.03
-0.88
N/A
0.33
0.45
0.46
0.47
0.52
-0.11
-0.63
-0.81
-0.96
-0.80
0.81
0.29
0.02
-0.31
-0.00042
N/A = not applicable; NO2 = nitrogen dioxide; NOX =the sum of nitric oxide and NO2.
Note: Model errors were based on comparisons between measured data and simulated data at several monitoring sites. Errors
were estimated for a single central site monitor, various monitor averages, and values computed from a geostatistical model. Z
denotes the measured concentration, and Z* denotes the true concentration, considered here to be from the CTM. Bias in the
exposure metric is given as the proportion of error between the measurement and true value to the measurement.
aData provided by the authors for Figure 5 of Goldman et al. (2012).
bData provided by the authors of Figure 4 of Goldman et al. (2012).
°Pearson correlation.
Source: Goldman et al. (2012).
1
2
3
4
5
6
7
Sarnatet al. (2010) studied the spatial variability of concentrations of NCh, along with
CO, Os, and PIVb 5, in the Atlanta, GA metropolitan area and how spatial variability
affects interpretation of epidemiologic results, using time-series data for circulatory
disease ED visits. Sensitivity to spatial variability was examined at slightly greater than
neighborhood scale (8 km) in this study. Interestingly, Sarnatet al. (2010) found that
relative risk varied with distance between the monitor and study population when
comparing urban to rural locations, but distance of the study population to the monitor
was not an important factor when comparing urban population groups. This suggests that,
January 2015
3-86
DRAFT: Do Not Cite or Quote
-------
1 even for spatially heterogeneous NO2, urban scale concentration measures may produce
2 results comparable to neighborhood-scale concentration measures if the sites were
3 comparable throughout the city, for example, as a result of similar traffic patterns.
4 However, Sarnat et al. (2010) caution that, because their study was limited to 8-km radii,
5 it is not possible to interpret this work with respect to near-road and on-road microscale
6 concentrations.
7 In a study of the effect of concentration metric choice (central site, arithmetic average
8 across space, or population-weighted average) used to represent exposure in a time-series
9 epidemiologic model, Strickland et al. (2011) found that choice of the concentration
10 metric resulted in large differences in the observed associations between ED visits for
11 pediatric asthma and exposure for spatially heterogeneous NO2 but not for spatially
12 homogeneous PM2 5 when using a unit standardization for computing the relative risk.
13 However, when Strickland et al. (2011) used IQR for standardization, there were little
14 differences among the relative risk estimates across the concentration metrics. The
15 differences observed between unit and IQR standardization are due to the fact that the
16 IQR reflects the spatial variability in the exposure metrics for the spatial and
17 population-weighted averages.
18 Error type also influences the health effect estimate from time-series studies. Dionisio
19 etal. (2014) decomposed the exposure measurement error into spatial and
20 population-based components. Spatial error was defined as the difference between
21 concentration simulated by an AERMOD dispersion model and concentration measured
22 at a CSM, and population error was defined as the difference between the SHEDS
23 exposure model (using only ambient sources) and the dispersion model. Errors were
24 computed for each ZIP code centroid. Three pollutants with high spatial variability (NOx,
25 CO, EC) termed "local" by the authors and three pollutants with low spatial variability
26 (PM2 5, Os, SO4) termed "regional" by the authors were included in the study. Although
27 NO2 was not included explicitly, the local results are relevant. Dionisio etal. (2014)
28 observed more variability in both the spatial and population components of the exposure
29 measurement error across the ZIP codes for the local pollutants compared with the
30 regional pollutants. Attenuation of the health effect estimate by the spatial error
31 component was much larger for the local pollutants compared with the regional
32 pollutants, and the amount of bias by the spatial error component was roughly the same
33 for NOx, CO, and EC. However, the population error component caused much more
34 attenuation of the health effect estimate for NOx compared with CO and EC. In fact, CO
35 had negligible bias of the health effect estimate due to the population error component.
36 This discrepancy is likely related to the deposition rate of NOx compared with the zero
37 deposition rate for CO modeled in SHEDS. Given that NO2 has a higher deposition rate
38 than NO, the results of Dionisio et al. (2014) suggest that health effect estimates modeled
January 2015 3-87 DRAFT: Do Not Cite or Quote
-------
1 in time-series studies of NOx exposure are likely extendable to NCh; see Section 3.3.2.1
2 for information related to deposition of indoor NC>2. Hence, it is likely that spatial
3 variability and indoor deposition both cause bias in the health effect estimate for NC>2.
4 Analysis of time-series epidemiologic studies have suggested that nonambient
5 contributions introduce Berkson error into the exposure term, where the error does not
6 bias health effect estimates for ambient NO2 assuming that nonambient NO2 sources are
7 independent of ambient sources, but it does cause the confidence intervals around the
8 health effect estimates to widen (Sheppard. 2005; Wilson etal.. 2000). No data from
9 cohort studies are available to test if this theory can be applied more broadly to all
10 epidemiologic studies. Sheppard et al. (2005) simulated the effect of nonambient sources
11 for a time-series study of the health effects of PM exposure and found that, as long as the
12 ambient and nonambient sources were uncorrelated, the nonambient exposures would
13 widen the confidence interval around the health effect estimates but would not bias the
14 health effect estimate. This result is generalizable to NO2 because it did not depend on the
15 particle size distribution.
16 Exposure measurement error related to instrument precision has a smaller effect on health
17 effect estimates in time-series studies compared with error related to spatial gradients in
18 the concentration because instrument precision would not be expected to modify the
19 ability of the instruments to respond to changes in concentration over time. Goldman
20 etal. (2010) investigated the influence of instrument error on health effect estimates in a
21 time-series epidemiology study by studying differences in exposure estimates and health
22 effect estimates obtained using colocated monitors. In this study, a random error term
23 based on observations from colocated monitors was added to a central site monitor's time
24 series to simulate population estimates for ambient air concentrations subject to
25 instrument precision error in 1,000 Monte Carlo simulations. Very little changes in the
26 risk ratios were observed for 1-h daily max NCh and 1-h daily max NOx concentrations.
27 For 1-h daily max NC>2 concentration, the RR per ppm of NC>2 concentration with
28 simulated instrument precision error was 1.0133 compared with RR per ppm = 1.0139 for
29 the central site monitor. For 1-h daily max NOx concentration with simulated instrument
30 precision error, RR per ppm = 1.0132 compared with the central site monitor's RRof
31 1.0139. The amount of bias in the health effect estimate related to instrument precision
32 was very small.
3.4.5.2 Longitudinal Cohort Studies
33 For cohort epidemiologic studies of long-term human exposure to NO2, where the
34 difference in the magnitude of the concentration is of most interest, ifCa,csm is used as a
January 2015 3-88 DRAFT: Do Not Cite or Quote
-------
1 surrogate for Ea, then a can be considered to encompass the exposure measurement error
2 related to uncertainties in the time-activity data and air exchange rate. Spatial variability
3 in NO2 concentrations across the study area could lead to bias in the health effect estimate
4 if Co, asm is systematically higher or lower than Ea. This could occur, for example, if the
5 study participants are clustered in a location where their NCh exposure is higher or lower
6 than the exposure estimated at a modeled or measurement site. Ca,csm may be an
7 acceptable surrogate for Ea if the central site monitor is located in close proximity to the
8 study participants (e.g., in a dense urban setting) and spatial variability of the NO2
9 concentration across the study area where the study participants are located is minimal in
10 the vicinity of each sample group. There is limited information regarding whether Ca,csm
11 is a biased exposure surrogate in the near-road environment for epidemiologic studies of
12 long-term exposure.
13 Sensitivity of the epidemiologic model to the temporal and spatial characteristics of
14 exposure data depends on the temporal characteristics of the disease process. Birth
15 outcome studies serve as an example where the exposure window becomes an important
16 consideration that helps to delineate short-term exposure from long-term exposure
17 epidemiologic study design. For example, Ross et al. (2013) studied the role of spatial
18 and temporal resolution of NCh estimates in the application of LUR to study birth
19 outcomes in New York City. Seasonal variability was more evident when averaging NC>2
20 estimates across the final 6 weeks of gestation compared with the entire gestation period,
21 but temporal variation had less influence on NC>2 predictions compared with PlVfc 5. This
22 finding reflects the fact that variability in NC>2 concentrations is more prominent in space
23 than in time compared with PM2 5. Additionally, Brauer et al. (2008) studied the influence
24 of NC>2 exposure models (IDW of central site monitoring data and LUR) on health effect
25 estimates for birth outcomes and observed higher adjusted odds ratios for IDW compared
26 with LUR (which produced health effect estimates closer to null, Section 6.4.3). Clark
27 etal. (2010) compared IDW with LUR for the analysis of asthma risk from in utero and
28 first-year-of-life exposure to NCh, NO, and other pollutants. They observed comparable
29 adjusted odds ratios for the first year of NO2 exposure and higher adjusted odds ratio
30 higher for IDW compared with LUR for in utero NO2 exposures (Section 6.4.3). The
31 biologically relevant time period in eliciting a birth outcome likely determines whether
32 spatial or temporal variation in concentration is more important to the epidemiologic
33 model. It is possible that, if the biologically relevant time period is short, then temporal
34 variability may play a larger role. In that case, the seasonal differences in NO2
35 concentration become more important for measuring an effect. If the biologically relevant
36 time period is longer, then the spatial contrasts evident in concentration maps become
37 more important so that exposure misclassification can lead to over- or under-estimation
38 of the effect.
January 2015 3-89 DRAFT: Do Not Cite or Quote
-------
1 Spatial resolution of the exposure estimates has been evaluated to examine the influence
2 of spatial exposure error in cohort studies. This has been considered with spatially-
3 resolved alternatives to central site monitoring data, such as data from a LUR, to describe
4 exposure of individuals within a cohort that is spatially dispersed within a study area (see
5 Section 3.2.2). Sellier et al. (2014) and Lepeule et al. (2010) evaluated various
6 approaches to estimate exposure (nearest central site monitor, geostatistical model, LUR
7 model, dispersion model) in a study of birth weight among a French mother-child cohort
8 in the French cities of Nancy and Poitiers. Correlations among the methods varied with
9 respect to methodology, distance, and land use type. For example, the correlation
10 between LUR and dispersion modeling had a minimum Pearson R = 0.58 (for urban
11 locations), while the correlation between central site monitoring and LUR had a
12 minimum R = 0.20 (also for urban locations). No effect of the method was observed on
13 change in birth weight, but confidence intervals around the health effect estimate
14 generally increased for dispersion models, which tended to be the most spatially
15 heterogeneous among the four methods studied.
16 The influence of spatial exposure misclassification on health effect estimate varies with
17 the particular study parameters, such as model selection and location. Madsen et al.
18 (2010) compared odds ratios for birth weight per quartiles of NO2 exposure estimated
19 from a near-road monitoring station and a dispersion model. Higher exposure variability
20 was captured by the dispersion model, but the adjusted odds ratio showed an effect only
21 for the near-road monitoring station exposure data, where time-averaged or residential
22 exposures were likely to be overestimated. Wu etal. (2011) compared health effect
23 estimates obtained using nearest monitors and LUR for birth outcomes for NO2 in Los
24 Angeles County and Orange County, CA. Odds ratios for NO2 were comparable for
25 nearest monitor and LUR for Los Angeles County, where the LUR was fit, but the odds
26 ratio decreased for Orange County in comparison with nearest monitor. This is consistent
27 with studies reporting higher exposure error when LUR models are fit in one city and
28 applied elsewhere, as described in Section 3.2.2.1. Ghosh etal. (2012) compared health
29 effect estimates for low birth weight using NO2 exposure estimates from LUR (scaled to
30 account for seasonal fluctuations in concentration) to nearest monitoring station in Los
31 Angeles County, CA and found negligible difference between the health effect estimates
32 obtained with each exposure metric.
33 Paciorek (2010) performed simulations to test the effect of spatial errors on health effect
34 estimates in long-term exposure epidemiologic studies. He identified unmeasured spatial
35 confounding as a key driver in biasing health effect estimates in a spatial regression.
36 Paciorek (2010) maintained that bias can be reduced when variation in the exposure
37 metric occurs at a smaller spatial scale than that of the unmeasured confounder. Szpiro
38 etal. (2011) also explored the effect of specification of spatial conditions on the exposure
January 2015 3-90 DRAFT: Do Not Cite or Quote
-------
1 metric used in a cohort epidemiologic study. They evaluated bias and uncertainty of the
2 health effect estimate obtained when using correctly specified and misspecified exposure
3 simulation conditions, where correct specification (i.e., predicting the "true"
4 concentration) was considered for comparison purposes to be the use of three spatial
5 prediction variables and misspecification implies unmeasured confounding in the model.
6 LUR calculations were used to simulate exposure, and correct specification was
7 considered when three spatial covariates were included in the model; the misspecified
8 model omitted a geographic covariate in the LUR. Each case was compared with the
9 reference case where the health effect estimate was obtained by using monitoring data for
10 the exposure metric. Szpiroet al. (2011) also reduced the amount of variability in the
11 third covariate of the correctly specified exposure model in an additional set of
12 simulations. Prediction accuracy of the exposure estimate was higher for the correctly
13 specified model compared with the misspecified model. However, bias in the health
14 effect estimate was also slightly more negative for the correctly specified model, and the
15 health effect estimate was more variable for the correctly specified model compared with
16 the misspecified model when the variability in the exposure covariate decreased. The
17 results of Szpiro et al. (2011) suggested that use of more accurately defined exposure
18 metrics in a cohort study does not necessarily improve health effect estimates. The Szpiro
19 et al. (2011) simulations were for a generic air pollutant. The results are relevant to NO2
20 exposure, but the spatial scale issue raised by Paciorek (2010) was not considered in the
21 Szpiro etal. (2011) analysis.
22 Basagana et al. (2013) also investigated the effect of differences in LUR model fitting on
23 error in the epidemiologic health effect estimates. In this study, Basagana et al. (2013) fit
24 three LUR models with 20, 40, or 80 measurement locations. For this simulation study,
25 the model considered correctly specified contained five covariates. As a comparison case,
26 Basagana et al. (2013) fit misspecified models containing 20 or 100 covariates (including
27 the five original covariates). The misspecification effectively added error to the model.
28 The simulated exposure error produced a combination of Berkson-like and classical-like
29 errors on the health effect estimate. Compared with the true health effect estimate, bias
30 towards the null was observed to increase with decreasing number of measurement
31 locations used to fit the LUR model. At the same time, the mean squared error of the
32 health effect estimate increased with decreasing number of measurement locations.
33 Moreover, bias towards the null and mean squared error also grew with increasing the
34 number of covariates from 5 to 20 to 100. Notably, R2 did not trend with the number of
35 variables, suggesting that R2 is not a sufficient measure of LUR model quality.
36 Under the assumption that the exposure estimates come from the same underlying
37 distribution as the true exposures associated with each health effect, Szpiro and Paciorek
38 (2013) performed several simulations to investigate the influence of the distribution and
January 2015 3-91 DRAFT: Do Not Cite or Quote
-------
1 variability of the exposure estimates on health effect estimates in longitudinal cohort
2 studies, building upon their previous studies of the influence of model misspecification
3 (Szpiro et al.. 2011) and unmeasured confounding (Paciorek. 2010) on longitudinal health
4 effect estimates. In one set of simulations, the distribution of the exposure was varied.
5 When the assigned exposure measurements were set to be uniform across space, the
6 health effect estimate was biased away from the null compared with the case when the
7 exposure subjects are colocated with the study participants. When an additional spatial
8 covariate was omitted, the health effect estimate was biased towards the null compared
9 with the correctly specified model. Additional simulations by Szpiro and Paciorek (2013)
10 based on the Multi-Ethnic Study of Atherosclerosis and Air Pollution (MESA-Air) NOx
11 concentration and health effects data set for Baltimore showed that bias in the health
12 effect estimate was subject to classical-like error and incorrectly estimated standard
13 errors. Both factors biased the health effect estimate towards the null. Bias in the health
14 effect estimate was minimized when a bias correction factor was introduced and a
15 bootstrapping technique was employed to recalculate standard error based on the
16 distribution of the exposures associated with the health effect data. Although this
17 approach was criticized as regression calibration (Spiegelman. 2013). Szpiro and
18 Paciorek (2013) illustrated the influence of classical-like and Berkson-like errors on
19 long-term exposure cohort study health effect estimates through these simulations.
20 Not considering time-activity patterns of study participants adds uncertainty to exposure
21 estimates obtained via spatial modeling such as LUR. Setton et al. (2011) investigated
22 how both spatial variability and unaccounted study participant mobility bias health effect
23 estimates in long-term exposure epidemiologic models of health effects from NC>2
24 exposure in a simulation study of cohorts in southern California and Vancouver, BC. In
25 this case, concentration at each participant's home was modeled (using the
26 Comprehensive Air Quality Model with Extensions (CAMx) for southern California and
27 using LUR and IDW interpolation of monitoring data for Vancouver). Populations were
28 simulated using human activity data for Vancouver and transportation survey data for
29 southern California. Bias in the health effect estimate increased in magnitude towards the
30 null with distance from home and time spent away from home. Moreover, when spatial
31 variability increased (through comparison of spatially variable LUR-derived NO2
32 concentrations with a smoother monitor-based approach for mapping NCh for the
33 Vancouver data), the health effect estimate obtained from the IDW-based approach was
34 closer to the null compared with the LUR-based health effect estimate. Setton et al.
35 (2011) interpreted this finding as evidence of the influence of smoothing spatially
36 heterogeneous concentration profiles on the health effect estimate.
37 Instrumentation bias could be anticipated to influence health effect estimates from
38 epidemiologic studies of long-term NCh exposures in some situations. Section 3.2.1.2
January 2015 3-92 DRAFT: Do Not Cite or Quote
-------
1 describes how passive monitors are likely to overestimate exposure given the influences
2 of ambient temperature, relative humidity, and presence of copollutants. Therefore, LUR
3 exposure may be overestimated when the LUR is fit using passive monitoring data.
4 Sections 2.4.1 and 3.2.1.1 describe how the presence of copollutants can also cause NO2
5 concentrations measured using central site monitors to be overestimated. Overestimating
6 exposure can drive health effect estimates towards the null. Ambient temperature and
7 relative humidity would not be expected to vary greatly within a city. However, local
8 copollutant concentrations may be spatially variable such that an LUR model fit, and
9 resulting health effect estimates, could have some differential bias across a city related to
10 instrument error. Because climate and ambient sources are more likely to differ among
11 cities, instrumentation error leading to overestimates of exposure could have a larger
12 influence on the comparison of health effect estimates among cities when LUR or central
13 site monitors are used to estimate exposures.
14 In the case of long-term exposure cohort studies, nonambient contributions to the total
15 personal exposure estimates would be expected to widen the confidence interval around
16 the health effect estimates by adding noise to the exposure signal, as is the case for
17 time-series studies of short-term exposure. Also, addition of any non-negative
18 nonambient component to the personal exposure measurement, such that the average total
19 personal NO2 exposure would necessarily be equal to or greater than the average personal
20 exposure to ambient NO2, would result in an underestimate of exposure. This associated
21 nondifferential exposure misclassification could bias the health effect estimate towards
22 the null.
3.4.5.3 Panel Studies
23 Consideration of errors in use of Ca,csm as a surrogate for Ea provides information on the
24 impact of this proxy measure on health effect estimates in panel studies. Van Roosbroeck
25 et al. (2008) evaluated health effect estimates among a panel of children for associations
26 of four respiratory outcomes with 48-hour NCh data from a single monitor located at the
27 children's school. These health effect estimates were compared with those obtained from
28 personal NC>2 monitoring to capture spatial variability in NC>2 concentrations and
29 time-activity data. Van Roosbroeck et al. (2008) observed that health effect estimates
30 were biased towards the null by roughly one-half to one-third when using a single
31 monitor outside the school in lieu of personal exposure monitors. In this case, bias in the
32 single-monitor health effect estimate was likely influenced by the spatial variability of the
33 NC>2 concentration profile, time-activity of the study participants, and infiltration of
34 ambient NC>2 indoors. The authors also adjusted the health effect estimate for nonambient
January 2015 3-93 DRAFT: Do Not Cite or Quote
-------
1 sources, including parental smoking, gas cooking, and presence of an unvented water
2 heater.
3 Sarnatet al. (2012) considered the influence of exposure metric on health effect estimates
4 obtained for a panel of school children. This study was conducted along the U.S.-Mexico
5 border in El Paso, TX, and Ciudad Juarez, Mexico, and 96-h avg concentrations
6 measured from central site chemiluminescent monitors, passive monitors outside the
7 children's schools, and passive monitors inside the children's schools were all used as
8 surrogates for exposure to NO2. The largest health effect estimate was observed for
9 measurements outside the school. In comparison, the health effect estimates for NO2
10 measured inside the schools and at central site monitors were several times smaller (see
11 Table 5-20). Based on the comparison between outdoor and central site monitoring
12 results, Sarnatet al. (2012) concluded that exposure misclassification from using central
13 site measurements, in lieu of measurements at the site of exposure, could lead to biasing
14 the health effect estimate towards the null. They proposed that this bias was related to the
15 failure of central site monitors to capture intra-urban spatial variability. The 2008 ISA for
16 Oxides of Nitrogen (U.S. EPA. 2008) also did not find conclusive evidence of the
17 influence of exposure measurement error on health effect estimates from panel
18 epidemiologic studies of NO2 exposure. In general, there is uncertainty regarding the
19 influence of NO2 monitor placement on the magnitude and directionality of bias of the
20 health effect estimate as related to use of central site monitors in lieu of localized
21 monitors in panel studies. As for epidemiologic studies of long-term NO2 exposure (see
22 Section 3.4.5.2). instrumentation error leading to overestimates of exposure could have a
23 differential influence on health effect estimates, especially for inter-city comparisons.
3.5 Conclusions
24 This chapter presents the current state of the science for assessment of human exposure to
25 NO2. It builds upon the 2008 ISA for Oxides of Nitrogen (U.S. EPA. 2008). which
26 concluded that errors associated with the use of NO2 concentrations measured at central
27 site monitors as exposure metrics for epidemiologic studies tended to bias the health
28 effect estimate towards the null for both short-term exposure and long-term exposure
29 epidemiologic studies. As detailed within this chapter, recent studies provide support for
30 the conclusions presented in the 2008 ISA for Oxides of Nitrogen (U.S. EPA. 2008) for
31 short-term exposure studies but differ in some cases for long-term exposure studies.
32 Commonly used exposure assessment methods include central site monitors, passive
33 monitors, LUR, CTM, and dispersion models (see Section 3.2). The influence of
34 measurement errors from each of these techniques varies with study design. These
January 2015 3-94 DRAFT: Do Not Cite or Quote
-------
1 methods are listed in Table 3-1, along with their application (i.e., the design of the study
2 in which they are used) and associated errors. Community time-series studies of
3 short-term NC>2 exposure typically use central site monitoring. Panel studies tend to
4 employ central site monitors or, in some cases, passive monitors. Studies of long-term
5 NC>2 exposure often use a variety of methods, including central site monitors, LUR,
6 dispersion models and spatial smoothing techniques. Errors associated with these
7 methods vary in importance based on their application. Dispersion modeling can be
8 subject to errors related to simplifying assumptions about the meteorology, urban or
9 natural topography, or photoreactivity of NO to form NO2. Additionally, NO2 exposure
10 estimates from inverse distance weighting or other spatial smoothing techniques can be
11 subject to error if the spatial scale of monitoring does not capture all sources. Studies
12 employing exposure estimates obtained using these methods often report R2, bias, and/or
13 mean squared error to describe the quality of the exposure estimates. Given that these
14 metrics do not always correlate, caution must be taken to interpret the quality of exposure
15 data from an individual study on the basis of one metric.
16 Factors contributing to error in NO2 exposure assessment include temporal activity of
17 epidemiologic study participants, spatial variability of NO2 concentrations across the
18 study area, infiltration of NO2 indoors, and instrument accuracy and precision (see
19 Section 3.4.3). With respect to time-activity data, variability within and among different
20 populations causes the limitation of having only one monitoring location in many studies
21 to have varying influence on exposure estimates within and among those different
22 populations. In general, spatial misalignment can occur when the time-activity patterns of
23 study participants are not factored into the study design or when the location where NO2
24 exposure is estimated does not coincide with the residential, school, or work location of
25 interest. Studies of spatial variability of human exposure indicate that the magnitude of
26 the error in exposure estimation increases with distance between the monitor and the
27 subject. As a result, there is a potential for exposure misclassification if the ambient NO2
28 concentration measured at a given site differs from that at the location of an
29 epidemiologic study participant, and this issue is present regardless of the spatial scale of
30 the epidemiology study. At the same time, the influence of spatial variability depends
31 strongly on the temporal design of the epidemiologic study, as described in the
32 paragraphs below. Infiltration and air exchange rate influence indoor levels of NO2 in the
33 absence of indoor sources and hence presents the potential for bias and uncertainty in a,
34 which depends on air exchange rate, penetration, and indoor deposition. NO2 monitors
35 are often subject to positive biases resulting from interference by NOy species.
36 Community time-series epidemiologic studies most commonly use central site monitors
37 to estimate human exposure to ambient NO2 (see Section 3.4.5.1). Temporal variability in
38 exposure is the relevant feature of the exposure data in a community time-series study.
January 2015 3-95 DRAFT: Do Not Cite or Quote
-------
1 Additionally, personal exposure measurements cannot feasibly be obtained for health
2 studies with large numbers of participants. There is some uncertainty associated with
3 using central site measurements of NC>2 concentrations to represent personal exposure
4 because the temporal variability of the central site exposure estimate may differ from the
5 temporal variability of the true exposure. Exposure estimates using NC>2 concentration
6 measurements from central site monitors do not capture the spatial variability of the
7 concentration field, which becomes a more important source of error for time-series
8 epidemiology studies if the NC>2 concentration at the site of the study participants is not
9 well correlated with measurements at the central site monitor. Nonambient contributions
10 and differential infiltration of NC>2 can also add error or uncertainty to a health effect
11 estimate. Instrument precision and accuracy are not thought to have a substantial
12 influence on health effect estimates in time-series studies. Simulation studies testing the
13 influence of exposure error in time-series studies suggest that exposure error may widen
14 the confidence intervals of the health effect estimate and bias the estimate towards the
15 null. This implies that reported health effect estimates for time-series studies of NO2
16 exposure are potentially lower than true health effect estimates or that the reported
17 confidence intervals around those health effect estimates are wider than the true
18 confidence intervals.
19 Long-term exposure epidemiology studies compare subjects or populations at different
20 locations (see Section 3.4.5.2). Therefore, spatial, rather than temporal, contrasts are
21 more important in long-term exposure studies. NCh concentrations measured at central
22 site monitors are often used to represent exposures when human health cohorts are
23 compared among cities. There is some uncertainty associated with using central site
24 measurements of NC>2 concentrations to represent personal exposure because the central
25 site exposure estimate likely varies in a positive or negative direction from the personal
26 exposure. This condition adds uncertainty to epidemiologic health effect estimates that
27 are derived from spatial contrasts, such as multicity or within-city studies. Moreover,
28 positive biases from measurement of NOy artifacts have the potential to enhance spatial
29 contrasts in exposure. LUR or dispersion models are often used to estimate exposure at
30 the residential locations of study participants in long-term exposure epidemiologic studies
31 because those models are designed to capture spatial variability within a geographic area
32 such as a city. LUR has been demonstrated to provide reasonable estimates of NO2
33 exposure if the model is trained and applied in the same general location such that the
34 exposure estimates and true exposures are assumed to come from the same data
35 distribution. Spatial misalignment of the participant locations with the participants'
36 exposure estimates can increase the uncertainty around the health effect estimate.
37 Moreover, if the misalignment is systematic, for example if the study participants are
38 clustered in a location where their NC>2 exposure is higher or lower than the exposure
39 estimated at a modeled or measurement site, the health effect estimate can be biased in
January 2015 3-96 DRAFT: Do Not Cite or Quote
-------
1 either direction depending on whether the modeled NCh exposure is higher or lower than
2 the true exposure. This issue adds uncertainty to health effect estimates derived from any
3 given individual long-term study.
4 Panel epidemiologic studies of NO2 exposure using central site monitors or centrally
5 located monitors are subject to exposure misclassification due to spatial misalignment
6 between the monitored ambient NC>2 concentration and the true personal exposure to
7 ambient NC>2 (see Section 3.4.5.3). Available panel studies that compare health effect
8 estimates among exposure assessment techniques have suggested that such spatial
9 misalignment leads to attenuating the health effect estimate. However, only a limited
10 number of panel studies have studied the influence of exposure measurement error on
11 health effect estimates. For this reason, it is difficult to reach a conclusion about the
12 magnitude and direction of error in the health effect estimates related to exposure error.
13 Copollutant confounding can occur when common sources emit multiple pollutants and
14 therefore can increase uncertainty in identifying whether the copollutants are
15 independently associated with a health effect (see Section 3.4.4). For traffic, NO (reacting
16 to NCh), CO, EC, UFP, and benzene are commonly coemitted and can be highly
17 correlated with NO2 in time and space. During winter, NO2 emitted from heating fuel
18 sources can also be highly correlated with PlVfc 5 and PMio. For both short-term exposure
19 and long-term exposure epidemiologic studies, it is difficult to distinguish the health
20 effect associated with NO2 exposure among health effects attributed to other highly
21 correlated pollutants. The temporal correlations may vary over space given that different
22 pollutants have different spatial scales over which they vary from peak to background
23 levels. For long-term exposure epidemiologic studies, bias related to copollutant
24 confounding can be reduced when the spatial scale of the NO2 exposure metric is smaller
25 than the spatial scale of the correlated copollutant. Bias related to copollutant
26 confounding may be more likely for unstable copollutants (e.g., UFP) or air pollutants
27 that disperse more quickly than NO2 (e.g., CO), compared with more spatially
28 homogeneous pollutants (e.g., PM2s).
January 2015 3-97 DRAFT: Do Not Cite or Quote
-------
References for Chapter 3
Adamkiewicz. G: Hsu. HH: Vallarino. J: Melly. SJ: Spengler. JD: Lew. JI. (2010). Nitrogen dioxide
concentrations in neighborhoods adjacent to a commercial airport: A land use regression modeling
study. Environ Health 9: 73. http://dx.doi.org/10.1186/1476-069X-9-73
Aguilera. I: Guxens. M: Garcia-Esteban. R: Corbella. T: Nieuwenhuijsen. MJ: Foradada. CM: Sunyer. J.
(2009). Association between GIS-based exposure to urban air pollution during pregnancy and birth
weight in the INMA Sabadell cohort. Environ Health Perspect 117: 1322-1327.
http://dx.doi.org/10.1289/ehp.0800256
Aguilera. I: Sunyer. J: Fernandez-Patier. R: Hoek. G: Aguirre-Alfaro. A: Meliefste. K: Bomboi-Mingarro.
MT: Nieuwenhuijsen. MJ: Herce-Garraleta. D: Brunekreef. B. (2008). Estimation of outdoor NOx,
NO2, and BTEX exposure in a cohort of pregnant women using land use regression modeling.
Environ SciTechnol 42: 815-821. http://dx.doi.org/10.1021/es0715492
Ainslie. B: Steyn. DG: Su. J: Buzzelli M: Brauer. M: Larson. T: Rucker. M. (2008). A source area model
incorporating simplified atmospheric dispersion and advection at fine scale for population air pollutant
exposure assessment. Atmos Environ 42: 2394-2404.
http://dx.doi.0rg/10.1016/j.atmosenv.2007.12.021
Allen. RW: Amram. O: Wheeler. AJ: Brauer. M. (2011). The transferability of NO and NO2 land use
regression models between cities and pollutants. Atmos Environ 45: 369-378.
http://dx.doi.0rg/10.1016/j.atmosenv.2010.10.002
Allen. RW: Davies. H: Cohen. MA: Mallach. G: Kaufman. JD: Adar. SD. (2009). The spatial relationship
between traffic-generated air pollution and noise in 2 US cities. Environ Res 109: 334-342.
http://dx.doi.0rg/10.1016/j.envres.2008.12.006
Aim. S: Mukala. K: Pasanen. P: Tiittanen. P: Ruuskanen. J: Tuomisto. J: Jantunen. MJ. (1998). Personal
NO2 exposures of preschool children in Helsinki. J Expo Anal Environ Epidemiol 8: 79-100.
Andersen. ZJ: Loft S: Ketzel M: Stage. M: Scheike. T: Hermansen. MN: Bisgaard. H. (2008). Ambient
air pollution triggers wheezing symptoms in infants. Thorax 63: 710-716.
http://dx.doi.org/10.1136/thx.2007.085480
Andersen. ZJ: Wahlin. P: Raaschou-Nielsen. O: Scheike. T: Loft S. (2007). Ambient particle source
apportionment and daily hospital admissions among children and elderly in Copenhagen. J Expo Sci
Environ Epidemiol 17: 625-636. http://dx.doi.org/10.1038/sj.jes.7500546
Arhami. M: Kamali. N: Rajabi. MM. (2013). Predicting hourly air pollutant levels using artificial neural
networks coupled with uncertainty analysis by Monte Carlo simulations. Environ Sci Pollut Res Int
20: 4777-4789. http://dx.doi.org/10.1007/sll356-012-1451-6
Arhami. M: Polidori. A: Delfino. RJ: Tjoa. T: Sioutas. C. (2009). Associations between personal, indoor,
and residential outdoor pollutant concentrations: Implications for exposure assessment to size-
fractionated particulate matter. J Air Waste Manag Assoc 59: 392-404. http://dx.doi.org/10.3155/1047-
3289.59.4.392
Atari. DO: Luginaah. I: Xu. X: Fung. K. (2008). Spatial variability of ambient nitrogen dioxide and sulfur
dioxide in Sarnia, "Chemical Valley," Ontario, Canada. J Toxicol Environ Health A 71:
1572-1581. http://dx.doi.org/10.1080/15287390802414158
Atari. DO: Luginaah. IN: Fung. K. (2009). The relationship between odour annoyance scores and
modelled ambient air pollution in Sarnia, "Chemical Valley", Ontario. Int J Environ Res
Public Health 6: 2655-2675. http://dx.doi.org/10.3390/iierph6102655
January 2015 3-98 DRAFT: Do Not Cite or Quote
-------
Atkinson. RW: Carey. IM: Kent. AJ: van Staa. TP: Anderson. HR: Cook. DG. (2013). Long-term
exposure to outdoor air pollution and incidence of cardiovascular diseases. Epidemiology 24: 44-53.
http://dx.doi.org/10.1097/EDE.Ob013e318276ccb8
Baawain. MS: Al-Serihj AS. (2014). Systematic approach for the prediction of ground-level air pollution
(around an industrial ort) using an artificial neural network. Aerosol Air Qual Res 14: 124-134.
http://dx.doi.org/10.4209/aaar.2013.06.0191
Bartolomei V: Sorgel M: Gligorovskj S: Alvarez. EG: Gandolfo. A: Strekowskj R: Quivet E: Held. A:
Zetzsch. C: Wortham. H. (2014). Formation of indoor nitrous acid (HONO) by light-induced NO2
heterogeneous reactions with white wall paint. Environ Sci Pollut Res Int 21: 9259-9269.
http://dx.doi.org/10.1007/sll356-014-2836-5
Basagana. X: Aguilera. I: Rivera. M: Agis. D: Foraster. M: Marrugat. J: Elosua. R: Kiinzli. N. (2013).
Measurement error in epidemiologic studies of air pollution based on land-use regression models. Am
J Epidemiol 178: 1342-1346. http://dx.doi.org/10.1093/aje/kwtl27
Basagana. X: Rivera. M: Aguilera. I: Agis. D: Bouso. L: Elosua. R: Foraster. M: de Nazelle. A:
Nieuwenhuijsen. M: Vila. J: Kiinzli. N. (2012). Effect of the number of measurement sites on land use
regression models in estimating local air pollution. Atmos Environ 54: 634-642.
http://dx.doi.0rg/10.1016/j.atmosenv.2012.01.064
Bateson. TF: Coull BA: Hubbell B: Ito. K: Jerrett M: Lumlev. T: Thomas. D: Vedal S: Ross. M. (2007).
Panel discussion review: session three - issues involved in interpretation of epidemiologic analyses -
statistical modeling. J Expo Sci Environ Epidemiol 17: S90-S96.
http://dx.doi.org/10.1038/sj.jes.7500631
Baxter. LK: Duvall RM: Sacks. J. (2013). Examining the effects of air pollution composition on within
region differences in PM(2.5) mortality risk estimates. J Expo Sci Environ Epidemiol 23: 457-465.
http://dx.doi.org/10.1038/ies.2012.114
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/j.atmosenv.2007.09.042
Beelen. R: Hoek. G: Houthuijs. D: van den Brandt. PA: Goldbohm. RA: Fischer. P: Schouten. LJ:
Armstrong. B: Brunekreef. B. (2009). The joint association of air pollution and noise from road traffic
with cardiovascular mortality in a cohort study. Occup Environ Med 66: 243-250.
http://dx.doi.org/10.1136/oem.2008.042358
Beelen. R: Voogt. M: Duyzer. J: Zandveld. P: Hoek. G. (2010). Comparison of the performances of land
use regression modelling and dispersion modelling in estimating small-scale variations in long-term
air pollution concentrations in a Dutch urban area. Atmos Environ 44: 4614-4621.
http://dx.doi.0rg/10.1016/i.atmosenv.2010.08.005
Beevers. SD: Carslaw. DC. (2005). The impact of congestion charging on vehicle emissions in London.
Atmos Environ 39: 1-5. http://dx.doi.Org/10.1016/j.atmosenv.2004.10.001
Beevers. SD: Kitwiroon. N: Williams. ML: Carslaw. DC. (2012). One way coupling of CMAQ and a road
source dispersion model for fine scale air pollution predictions. Atmos Environ 59: 47-58.
http://dx.doi.0rg/10.1016/i.atmosenv.2012.05.034
Beevers. SD: Kitwiroon. N: Williams. ML: Kelly. FJ: Ross Anderson. H: Carslaw. DC. (2013). Air
pollution dispersion models for human exposure predictions in London. J Expo Sci Environ
Epidemiol. http://dx.doi.Org/10.1038/jes.2013.6
Bellander. T: Wichmann. J: Lind. T. (2012). Individual exposure to NO2 in relation to spatial and
temporal exposure indices in Stockholm, Sweden: The INDEX study. PLoS ONE 7: e39536.
http://dx.doi.org/10.1371/journal.pone.0039536
Benson. PE. (1992). Areview of the development and application of the CALINE3 and CALINE4
models. Atmos Environ 26: 379-390.
January 2015 3-99 DRAFT: Do Not Cite or Quote
-------
Bentaveb. M: Stempfelet M: Wagner. V: Zins. M: Bonenfant S: Songeur. C: Sanchez. O: Rosso. A:
Brulfert G: Rios. I: Chaxel E: Virga. J: Armengaud. A: Rossello. P: Riviere. E: Bernard. M: Vasbien.
F: Deprost R. (2014). Retrospective modeling outdoor air pollution at a fine spatial scale in France,
1989-2008. Atmos Environ 92: 267-279. http://dx.doi.Org/10.1016/j.atmosenv.2014.04.019
Berkowicz. R: Ketzel M: Jensen. SS: Hvidberg. M: Raaschou-Nielsen. O. (2008). Evaluation and
application of OSPM for traffic pollution assessment for a large number of street locations. Environ
Modell Softw 23: 296-303. http://dx.doi.Org/10.1016/j.envsoft.2007.04.007
Berrocal VJ: Gelfand. AE: Holland. DM. (2010a). Abivariate space-time downscaler under space and
time misalignment. Ann Appl Stat 4: 1942-1975. http://dx.doi.org/10.1214/10-AOAS351
Berrocal VJ: Gelfand. AE: Holland. DM. (201 Ob). A spatio-temporal downscaler for output from
numerical models. J Agric Biol Environ Stat 15: 176-197. http://dx.doi.org/10.1007/sl3253-009-0004-
z
Berrocal. VJ: Gelfand. AE: Holland. DM. (2012). Space-time data fusion under error in computer model
output: An application to modeling air quality. Biometrics 68: 837-848.
http://dx.doi.0rg/10.llll/j.1541-0420.2011.01725.x
Binkowskj FS: Arunachalam. S: Adelman. Z: Pinto. JP (2007). Examining photolysis rates with a
prototype online photolysis module in CMAQ. J Appl Meteor Climatol 46: 1252-1256.
http://dx.doi.Org/10.1175/JAM2531.l
Brauer. M: Koutrakis. P: Keeler. GJ: Spengler. JD. (1991). Indoor and outdoor concentrations of
inorganic acidic aerosols and gases. J Air Waste Manag Assoc 41: 171-181.
http://dx.doi.org/10.1080/10473289.1991.10466834
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
Brauer. M: Ryan. PB: Suh. HH: Koutrakis. P: Spengler. JD: Leslie. NP: Billick. IH. (1990).
Measurements of nitrous acid inside two research houses. Environ Sci Technol 24: 1521-1527.
http://dx.doi.org/10.1021/es00080a011
Brook. JR: Burnett. RT: Dann. TF: Cakmak. S: Goldberg. MS: Fan. X: Wheeler. AJ. (2007). Further
interpretation of the acute effect of nitrogen dioxide observed in Canadian time-series studies. J Expo
Sci Environ Epidemiol 17: S36-S44. http://dx.doi.org/10.1038/sj.jes.7500626
Brown. KW: Sarnat JA: Suh. HH: Coull BA: Koutrakis. P. (2009). Factors influencing relationships
between personal and ambient concentrations of gaseous and particulate pollutants. Sci Total Environ
407: 3754-3765. http://dx.doi.Org/10.1016/i.scitotenv.2009.02.016
Burke. JM: Zufall MJ: Ozkaynak. H. (2001). A population exposure model for particulate matter: Case
study results for PM2.5 in Philadelphia, PA. J Expo Anal Environ Epidemiol 11: 470-489.
http://dx.doi.org/10.1038/sj.jea.7500188
Burnett. RT: Brook. J: Dann. T: Delocla. C: Philips. O: Cakmak. S: Vincent R: Goldberg. MS: Krewski.
EX (2000). Association between particulate- and gas-phase components of urban air pollution and daily
mortality in eight Canadian cities. Inhal Toxicol 12: 15-39.
http://dx.doi.org/10.1080/08958370050164851
Rutland. BK: Armstrong. B: Atkinson. RW: Wilkinson. P: Heal MR: Dohertv. RM: Vieno. M. (2013).
Measurement error in time-series analysis: a simulation study comparing modelled and monitored
data. BMC Med Res Methodol 13: 136. http://dx.doi.org/10.1186/1471-2288-13-136
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
January 2015 3-100 DRAFT: Do Not Cite or Quote
-------
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
Cakmak. S: Dales. RE: Vida. CB. (2009). Components of particulate air pollution and mortality in Chile.
Int J Occup Environ Health 15: 152-158. http://dx.doi.org/10.1179/107735209799195844
Cao. G: Jang. M. (2008). Secondary organic aerosol formation from toluene photooxidation under various
NOx conditions and particle acidity. Atmos Chem Phys Discuss 8: 14467-14495.
http://dx.doi.org/10.5194/acpd-8-14467-2008
Cao. G: Jang. M. (2010). An SOA model for toluene oxidation in the presence of inorganic aerosols.
Environ Sci Technol 44: 727-733. http://dx.doi.org/10.1021/es901682r
Carslaw. DC: Beevers. SD: Ropkins. K: Bell. MC. (2006). Detecting and quantifying aircraft and other
on-airport contributions to ambient nitrogen oxides in the vicinity of a large international airport.
Atmos Environ 40: 5424-5434. http://dx.doi.Org/10.1016/j.atmosenv.2006.04.062
Carslaw. DC: Taylor. PJ. (2009). Analysis of air pollution data at a mixed source location using boosted
regression trees. Atmos Environ 43: 3563-3570. http://dx.doi.Org/10.1016/j.atmosenv.2009.04.001
Carslaw. DC: Williams. ML: Barratt B. en. (2012a). A short-term intervention study - Impact of airport
closure due to the eruption of Eyjafjallajokull on near-field air quality. Atmos Environ 54: 328-336.
http://dx.doi.0rg/10.1016/j.atmosenv.2012.02.020
Carslaw. N: Mota. T: Jenkin. ME: Barley. MH: Mcfiggans. G. (2012b). A significant role for nitrate and
peroxide groups on indoor secondary organic aerosol. Environ Sci Technol 46: 9290-9298.
http://dx.doi.org/10.1021/es301350x
Cattaneo. A: Tecce. N: Derudi M: Gelosa. S: Nano. G: Cavallo. DM. (2014). Assessment of modeled
indoor air concentrations of particulate matter, gaseous pollutants, and volatile organic compounds
emitted from candles. Hum Ecol Risk Assess 20: 962-979.
http://dx.doi.org/10.1080/10807039.2013.821902
Centro di Ricerche Ambientali (Centra di Ricerche Ambientali di Padova). (2006). Radiello(R) manual.
Padova, Italy: Fondazione Salvatore Maugeri Clinica del Lavoro e Delia Riabilitazione I.R.C.C.S.,
Centro di Ricerche Ambientali. http://www.radiello.com/english/Radiello%27s%20manual%2001-
06.pdf
Cesaronj G: Badalonj C: Porta. D: Forastiere. F: Perucci CA. (2008). Comparison between various
indices of exposure to traffic-related air pollution and their impact on respiratory health in adults.
Occup Environ Med 65: 683-690. http://dx.doi.org/10.1136/oem.2007.037846
Cesaronj G: Porta. D: Badalonj C: Stafoggia. M: Eeftens. M: Meliefste. K: Forastiere. F. (2012).
Nitrogen dioxide levels estimated from land use regression models several years apart and association
with mortality in a large cohort study. Environ Health 11: 48. http://dx.doi.org/10.1186/1476-069X-ll-
48
Chang. JC: Franzese. P: Chavantrakom. K: Hanna. SR. (2003). Evaluations of CALPUFF, HPAC, and
VLSTRACK with two mesoscale field datasets. J Appl Meteorol 42: 453-466.
Chatzidiakou. L: Mumovic. D: Summerfield. AJ: Hong. SM: Altamirano-Medina. H. (2014). A Victorian
school and a low carbon designed school: Comparison of indoor air quality, energy performance, and
student health. Indoor Built Environ 23: 417-432. http://dx.doi.org/10.1177/1420326X14532388
Chaudhurj S: Dutta. D. (2014). Mann-Kendall trend of pollutants, temperature and humidity over an
urban station of India with forecast verification using different ARIMA models. Environ Monit Assess
186: 4719-4742. http://dx.doi.org/10.1007/sl0661-014-3733-6
Chen. G: Li. J: Ying. Q: Sherman. S: Perkins. N: Rajeshwarj S: Mendola. P. (2014). Evaluation of
observation-fused regional air quality model results for population air pollution exposure estimation.
Sci Total Environ 485-486: 563-574. http://dx.doi.Org/10.1016/i.scitotenv.2014.03.107
January 2015 3-101 DRAFT: Do Not Cite or Quote
-------
Chen. G: Song. G: Jiang. L: Zhang. Y: Zhao. N: Chen. B: Kan. H. (2008). Short-term effects of ambient
gaseous pollutants and particulate matter on daily mortality in Shanghai, China. J Occup Health 50:
41-47. http://dx.doi.org/10.1539/joh.50.41
Chuang. KJ: Coull BA: Zanobetti. A: Suh. H: Schwartz. J: Stone. PH: Litonjua. A: Speizer. FE: Gold.
DR. (2008). Particulate air pollution as a risk factor for ST-segment depression in patients with
coronary artery disease. Circulation 118: 1314-1320.
http://dx.doi.org/10.1161/CIRCULATIONAHA.108.765669
Cimorelli AJ: Perry. SG: Venkatram. A: Weil JC: Paine. R: Wilson. RB: Lee. RF: Peters. WD: Erode.
RW. (2005). AERMOD: A dispersion model for industrial source applications. Part I: General model
formulation and boundary layer characterization. J Appl Meteorol 44: 682-693.
http://dx.doi.Org/10.1175/JAM2227.l
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
Cloughertv. JE: Kheirbek. I: Eisl. HM: Ross. Z: Pezeshki. G: Gorczynski. JE: Johnson. S: Markowitz. S:
Kass. D: Matte. T. (2013). Intra-urban spatial variability in wintertime street-level concentrations of
multiple combustion-related air pollutants: The New York City Community Air Survey (NYCCAS). J
Expo Sci Environ Epidemiol 23: 232-240. http://dx.doi.org/10.1038/jes.2012.125
Cohan. A: Wu. J: Dabdub. D. (2011). High-resolution pollutant transport in the San Pedro Bay of
California. Atmos Pollut Res 2: 237-246. http://dx.doi.org/10.5094/APR.2011.030
Cohen. MA: Adar. SD: Allen. RW: Avol. E: Curl. CL: Gould. T: Hardie. D: Ho. A: Kinnev. P: Larson. TV:
Sampson. P: Sheppard. L: Stukovsky. KD: Swan. SS: Liu. LJ: Kaufman. JD. (2009). Approach to
estimating participant pollutant exposures in the multi-ethnic study of atherosclerosis and air pollution
(MESA Air). Environ SciTechnol 43: 4687-4693. http://dx.doi.org/10.1021/es8030837
Connell. DP: Withum. JA: Winter. SE: Statnick. RM: Bilonick. RA. (2005). The Steubenville
Comprehensive Air Monitoring Program (SCAMP): Associations among fine particulate matter, co-
pollutants, and meteorological conditions. J Air Waste Manag Assoc 55: 481-496.
http://dx.doi.org/10.1080/10473289.2005.10464631
Cowie. CT: Rose. N: Gillett. R: Walter. S: Marks. GB. (2012). Redistribution of traffic related air
pollution associated with a new road tunnel. Environ Sci Technol 46: 2918-2927.
http://dx.doi.org/10.1021/es202686r
Crooks. J: Isakov. V. (2013). A wavelet-based approach to blending observations with deterministic
computer models to resolve the intraurban air pollution field. J Air Waste Manag Assoc 63: 1369-
1385. http://dx.doi.org/10.1080/10962247.2012.758061
Grouse. PL: Goldberg. MS: Ross. NA. (2009). A prediction-based approach to modelling temporal and
spatial variability of traffic-related air pollution in Montreal, Canada. Atmos Environ 43: 5075-5084.
http://dx.doi.0rg/10.1016/j.atmosenv.2009.06.040
Cui. H: Yao. R: Xu. X: Xin. C: Yang. J. (2011). A tracer experiment study to evaluate the CALPUFF real
time application in a near-field complex terrain setting. Atmos Environ 45: 7525-7532.
http://dx.doi.0rg/10.1016/j.atmosenv.2011.08.041
Dadvand. P: Ostro. B: Figueras. F: Foraster. M: Basagana. X: Valentin. A: Martinez. D: Beelen. R:
Cirach. M: Hoek. G: Jerrett. M: Brunekreef. B: Nieuwenhuijsen. MJ. (2014). Residential proximity to
major roads and term low birth weight: the roles of air pollution, heat, noise, and road-adjacent trees.
Epidemiology 25: 518-525. http://dx.doi.org/10.1097/EDE.0000000000000107
Dales. RE: Cakmak. S: Vidal. CB. (2009). Air pollution and hospitalization for headache in Chile. Am J
Epidemiol 170: 1057-1066. http://dx.doi.org/10.1093/aje/kwp217
Dales. RE: Cakmak. S: Vidal. CB. (2010). Air Pollution and hospitalization for venous thromboembolic
disease in Chile. J Thromb Haemost 8: 669-674. http://dx.doi.Org/10.llll/i.1538-7836.2010.03760.x
January 2015 3-102 DRAFT: Do Not Cite or Quote
-------
Darrow. LA: Klein. M: Sarnat JA: Mulholland. JA: Strickland. MJ: Sarnat SE: Russell AG: Tolbert PE.
(2011). 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
Davies. HW: Vlaanderen. JJ: Henderson. SB: Brauer. M. (2009). Correlation between co-exposures to
noise and air pollution from traffic sources. Occup Environ Med 66: 347-350.
http://dx.doi.org/10.1136/oem.2008.041764
De Santis. F: Allegrinj I: Fazio. MC: Pasella. D. (1996). Characterization of indoor air quality in the
Church of San luigi Dei Francesi, Rome, Italy. Int J Environ Anal Chem 64: 71-81.
http://dx.doi.org/10.1080/03067319608028336
Debry. E: Mallet. V. (2014). Ensemble forecasting with machine learning algorithms for ozone, nitrogen
dioxide and PM10 on the Prev'Air platform. Atmos Environ 91: 71-84.
http://dx.doi.0rg/10.1016/j.atmosenv.2014.03.049
DeCastro. BR: Sax. AN: Chillrud. SN: Kinnev. PL: Spengler. JD. (2007). Modeling time-location patterns
of inner-city high school students in New York and Los Angeles using a longitudinal approach with
generalized estimating equations. J Expo Sci Environ Epidemiol 17: 233-247.
http://dx.doi.0rg/doi:10.1038/sj.jes.7500504
Delfino. RJ: Staimer. N: Tjoa. T: Gillen. D: Kleinman. MT: Sioutas. C: Cooper. D. (2008). Personal and
ambient air pollution exposures and lung function decrements in children with asthma. Environ Health
Perspect 116: 550-558. http://dx.doi.org/10.1289/ehp.10911
Delfino. RJ: Staimer. N: Tjoa. T: Gillen. PL: Polidori. A: Arhami. M: Kleinman. MT: Vaziri. ND:
Longhurst J: Sioutas. C. (2009). Air pollution exposures and circulating biomarkers of effect in a
susceptible population: Clues to potential causal component mixtures and mechanisms. Environ
Health Perspect 117: 1232-1238. http://dx.doi.org/10.1289/ehp.0800194
Delgado-Saborit JM. (2012). Use of real-time sensors to characterise human exposures to combustion
related pollutants. J Environ Monit 14: 1824-1837. http://dx.doi.org/10.1039/c2eml0996d
Dijkema. MB: Gehring. U: van Strien. RT: van der Zee. SC: Fischer. P: Hoek. G: Brunekreef. B. (2011).
A comparison of different approaches to estimate small-scale spatial variation in outdoor NO
concentrations. Environ Health Perspect 119: 670-675. http://dx.doi.org/10.1289/ehp.0901818
Dimitriou. K: Kassomenos. P. (2014). Local and regional sources of fine and coarse particulate matter
based on traffic and background monitoring. Theor Appl Climatol 116: 413-433.
http://dx.doi.org/10.1007/s00704-013-0961-6
Dimitroulopoulou. C: Ashmore. MR: Hill. MTR: Byrne. MA: Kinnerslev. R. (2006). INDAIR: a
probabilistic model of indoor air pollution in UK homes. Atmos Environ 40: 6362-6379.
http://dx.doi.0rg/10.1016/j.atmosenv.2006.05.047
Dionisio. KL: Baxter. LK: Chang. HH. (2014). An empirical assessment of exposure measurement error
and effect attenuation in bipollutant epidemiologic models. Environ Health Perspect 122: 1216-1224.
http://dx.doi.org/10.1289/ehp.1307772
Dioniskx KL: Isakov. V: Baxter. LK: Sarnat. JA: Sarnat. SE: Burke. J: Rosenbaum. A: Graham. SE:
Cook. R: Mulholland. J: Oezkaynak. H. (2013). Development and evaluation of alternative approaches
for exposure assessment of multiple air pollutants in Atlanta, Georgia. J Expo Sci Environ Epidemiol
23: 581-592. http://dx.doi.org/10.1038/jes.2013.59
Dong. GH: Chen. T: Liu. MM: Wang. D: Ma. YN: Ren. WH: Lee. YL: Zhao. YD: He. PC. (2011).
Gender differences and effect of air pollution on asthma in children with and without allergic
predisposition: Northeast Chinese children health study. PLoS ONE 6: e22470.
http://dx.doi.org/10.1371/journal.pone.0022470
Dons. E: Van Poppel M: Int Panis. L: De Prins. S: Berghmans. P: Koppen. G: Matheeussen. C. (2014).
Land use regression models as a tool for short, medium and long term exposure to traffic related air
pollution. Sci Total Environ 476-477: 378-386. http://dx.doi.Org/10.1016/j.scitotenv.2014.01.025
January 2015 3-103 DRAFT: Do Not Cite or Quote
-------
Du. X: Fu. L: Ge. W: Zhang. S: Wang. H. (2011). Exposure of taxi drivers and office workers to traffic-
related pollutants in Beijing: Anote. Transport Res Transport Environ 16: 78-81.
http://dx.doi.0rg/10.1016/i.trd.2010.08.002
Ducret-Stich. RE: Tsai. MY: Ragettli. MS: Ineichen. A: Kuenzlj N: Phuleria. HC. (2013). Role of
highway traffic on spatial and temporal distributions of air pollutants in a Swiss Alpine valley. Sci
Total Environ 456-457: 50-60. http://dx.doi.Org/10.1016/j.scitotenv.2013.03.065
Eeftens. M: Beelen. R: Fischer. P: Brunekreef. B: Meliefste. K: Hoek. G. (2011). Stability of measured
and modelled spatial contrasts in NO2 over time. Occup Environ Med 68: 765-770.
http://dx.doi.org/10.1136/oem.2010.061135
Eeftens. M: Tsai. MY: Ampe. C: Anwander. B: Beelen. R: Bellander. T: Cesaroni. G: Cirach. M: Cyrys. J:
de Hoogh. K: De Nazelle. A: de Vocht. F: Declercq. C: Dedele. A: Eriksen. K: Galassi C:
Grazuleviciene. R: Grivas. G: Heinrich. J: Hoffmann. B: Lakovides. M: Ineichen. A: Katsouvanni. K:
Korek. M: Kraemer. U: Kuhlbusch. T: Lanki. T: Madsen. C: Meliefste. K: Moelter. A: Mosler. G:
Nieuwenhuijsen. M: Oldenwening. M: Pennanen. A: Probst-Hensch. N: Quass. U: Raaschou-Nielsen.
O. le: Ranzi. A: Stephanou. E: Sugiri. D: Udvardv. O: Vaskoevi. E. va: Weinmayr. G: Brunekreef. B:
Hoek. G. (2012). Spatial variation of PM2.5, PM10, PM2.5 absorbance and PMcoarse concentrations
between and within 20 European study areas and the relationship with NO2 - Results of the ESCAPE
project. Atmos Environ 62: 303-317. http://dx.doi.Org/10.1016/j.atmosenv.2012.08.038
Esplugues. A: Ballester. F: Estarlich. M: Llop. S: Fuentes-Leonarte. V: Mantilla. E: Vioque. J: Iniguez. C.
(2011). Outdoor, but not indoor, nitrogen dioxide exposure is associated with persistent cough during
the first year of life. Sci Total Environ 409: 4667-4673.
http://dx.doi.0rg/10.1016/i.scitotenv.2011.08.007
Faustini. A: Stafoggia. M: Berti. G: Bisanti. L: Chiusolo. M: Cernigliaro. A: Mallone. S: Primerano. R:
Scarnato. C: Simonato. L: Vigotti. MA: Forastiere. F. (2011). The relationship between ambient
particulate matter and respiratory mortality: A multi-city study in Italy. Eur Respir J 38: 538-547.
http://dx.doi.org/10.1183/09031936.00093710
Fernandez-Somoano. A: Estarlich. M: Ballester. F: Fernandez-Patier. R: Aguirre-Alfaro. A: Herce-
Garraleta. MD: Tardona. A. (2011). Outdoor NO(2) and benzene exposure in the INMA (Environment
and Childhood) Asturias cohort (Spain). Atmos Environ 45: 5240-5246.
http://dx.doi.0rg/10.1016/i.atmosenv.2011.02.010
Fernando. HJS. (2010). Fluid dynamics of urban atmospheres in complex terrain. Annu Rev Fluid Mech
42: 365-389. http://dx.doi.org/10.1146/annurev-fluid-121108-145459
Franklin. M: Vora. H: Avol. E: Mcconnell. R: Lurmann. F: Liu. F: Penfold. B: Berhane. K: Gilliland. F:
Gauderman. WJ. (2012). Predictors of intra-community variation in air quality. J Expo Sci Environ
Epidemiol22: 135-147. http://dx.doi.org/10.1038/jes.2011.45
Fruin. S: Westerdahl. D: Sax. T: Sioutas. C: Fine. PM. (2008). Measurements and predictors of on-road
ultrafine particle concentrations and associated pollutants in Los Angeles. Atmos Environ 42: 207-
219. http://dx.doi.0rg/10.1016/i.atmosenv.2007.09.057
Fuentes. M. (2009). Statistical issues in health impact assessment at the state and local levels. Air Qual
Atmos Health 2: 47-55. http://dx.doi.org/10.1007/sll869-009-0033-3
Fuentes. M: Raftery. AE. (2005). Model evaluation and spatial interpolation by Bayesian combination of
observations with outputs from numerical models. Biometrics 61: 36-45.
http://dx.doi.Org/10.llll/i.0006-341X.2005.030821.x
Fujita. EM: Campbell DE: Zielinska. B: Arnott WP: Chow. JC. (2011). Concentrations of air toxics in
motor vehicle-dominated environments (pp. 3-77). (ISSN 1041-5505
156). Boston, MA: Health Effects Institute.
Gair. AJ: Penkett SA: Oyola. P. (1991). Development of a simple passive technique for the determination
of nitrogen-dioxide in remote continental locations. Atmos Environ 25 A: 1927-1939.
January 2015 3-104 DRAFT: Do Not Cite or Quote
-------
Gan. WQ: Davies. HW: Koehoorn. M: Brauer. M. (2012a). Association of long-term exposure to
community noise and traffic-related air pollution with coronary heart disease mortality. Am J
Epidemiol 175: 898-906. http://dx.doi.org/10.1093/aje/kwr424
Gan. WQ: Mclean. K: Brauer. M: Chiarello. SA: Davies. HW. (2012b). Modeling population exposure to
community noise and air pollution in a large metropolitan area. Environ Res 116: 11-16.
http://dx.doi.0rg/10.1016/j.envres.2012.04.001
Gauderman. WJ: Avol E: Lurmann. F: Kuenzli. N: Gilliland. F: Peters. J: McConnell. R. (2005).
Childhood asthma and exposure to traffic and nitrogen dioxide. Epidemiology 16: 737-743.
http://dx.doi.org/10.1097/01.ede.0000181308.51440.75
Gauvin. S: Le Moullec. Y: Bremont. F: Momas. I: Balducci. F: Ciognard. F: Poilve. MP: Zmirou. D:
Investigators. V. (2001). Relationships between nitrogen dioxide personal exposure and ambient air
monitoring measurements among children in three French metropolitan areas: VESTA study. Arch
Environ Occup Health 56: 336-341. http://dx.doi.org/10.1080/00039890109604465
Ghannam. K: El-Fadel. M. (2013). A framework for emissions source apportionment in industrial areas:
MM5/CALPUFF in a near-field application. J Air Waste Manag Assoc 63: 190-204.
http://dx.doi.org/10.1080/10962247.2012.739982
Ghosh. JKC: Wilhelm. M: Su. J: Goldberg. D: Cockburn. M: Jerrett M: Ritz. B. (2012). Assessing the
influence of traffic-related air pollution on risk of term low birth weight on the basis of land-use-based
regression models and measures of air toxics. Am J Epidemiol 175: 1262-1274.
http://dx.doi.org/10.1093/aje/kwr469
Gibson. MD: Kundu. S: Satish. M. (2013). Dispersion model evaluation of PM2.5, NOX and SO2 from
point and major line sources in Nova Scotia, Canada using AERMOD Gaussian plume air dispersion
model. Atmos Pollut Res 4: 157-167. http://dx.doi.org/10.5094/APR.2013.016
Goldman. GT: Mulholland. JA: Russell AG: Gass. K: Strickland. MJ: Tolbert PE. (2012).
Characterization of ambient air pollution measurement error in a time-series health study using a
geostatistical simulation approach. Atmos Environ 57: 101-108.
http://dx.doi.0rg/10.1016/j.atmosenv.2012.04.045
Goldman. GT: Mulholland. JA: Russell AG: Srivastava. A: Strickland. MJ: Klein. M: Waller. LA:
Tolbert. PE: Edgerton. ES. (2010). Ambient air pollutant measurement error: characterization and
impacts in a time-series epidemiologic study in Atlanta. Environ Sci Technol 44: 7692-7698.
http://dx.doi.org/10.1021/esl01386r
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 10: 61. http://dx.doi.org/10.1186/1476-069X-10-61
Gonzales. M: Myers. O: Smith. L: Olvera. HA: Mukerjee. S: Li. WW: Pingitore. N: Amava. M: Burchiel.
S: Berwick. M. (2012). Evaluation of land use regression models for NO2 in El Paso, Texas, USA. Sci
Total Environ 432: 135-142. http://dx.doi.Org/10.1016/j.scitotenv.2012.05.062
Greenwald. R: Sarnat SE: Ravsoni. AU: Li. WW: Johnson. BA: Stock. TH: Holguin. F: Sosa. T: Sarnat
JA. (2013). Associations between source-indicative pollution metrics and increases in pulmonary
inflammation and reduced lung function in a panel of asthmatic children. Air Qual Atmos Health 6:
487-499. http://dx.doi.org/10.1007/sll869-012-0186-3
Griffiths. PT: Badger. CL: Cox. RA: Folkers. M: Henk. HH: Mentel TF. (2009). Reactive uptake of
N2O5 by aerosols containing dicarboxylic acids. Effect of particle phase, composition, and nitrate
content. J Phys Chem A 113: 5082-5090. http://dx.doi.org/10.1021/jp8096814
Gruzieva. O: Bergstrom. A: Hulchiy. O: Kull. I: Lind. T: Melen. E: Moskalenko. V: Pershagen. G:
Bellander. T. (2013). Exposure to air pollution from traffic and childhood asthma until 12 years of age.
Epidemiology 24: 54-61. http://dx.doi.org/10.1097/EDE.Ob013e318276clea
January 2015 3-105 DRAFT: Do Not Cite or Quote
-------
Guo. Y: Jia. Y: Pan. X: Liu. L: Wichmann. HE. (2009). The association between fine participate air
pollution and hospital emergency room visits for cardiovascular diseases in Beijing, China. Sci Total
Environ 407: 4826-4830. http://dx.doi.Org/10.1016/j.scitotenv.2009.05.022
Hagenbjork-Gustafsson. A: Eriksson. K: Forsberg. B. (2009). Field evaluation of the Ogawa Diffusive
Sampler for NO(2)/NO(x) in a cold climate [Abstract]. Epidemiology 20: S161-S161.
http://dx.doi.org/10.1097/01.ede.0000362548.83420.d7
Halonen. JI: Lanki. T: Yli-Tuomi. T: Tiittanen. P: Kulmala. M: Pekkanen. (2009). Particulate air pollution
and acute cardiorespiratory hospital admissions and mortality among the elderly. Epidemiology 20:
143-153. http://dx.doi.org/10.1097/EDE.Ob013e31818c7237
Hart. J: Yanosky. J: Puett R: Ryan. L: Dockery. D: Smith. T: Garshick. E: Laden. F. (2009). Spatial
modeling of PM10 and NO2 in the continental United States, 1985-2000. Environ Health Perspect
117: 1690-1696. http://dx.doi.org/10.1289/ehp.0900840
Hatzopoulou. M: Weichenthal S: Barreau. G: Goldberg. M: Farrell. W: Grouse. D: Ross. N. (2013). A
web-based route planning tool to reduce cyclists' exposures to traffic pollution: A case study in
Montreal, Canada. Environ Res 123: 58-61. http://dx.doi.Org/10.1016/j.envres.2013.03.004
Heal MR. (2008). The effect of absorbent grid preparation method on precision and accuracy of ambient
nitrogen dioxide measurements using Palmes passive diffusion tubes. J Environ Monit 10: 1363-1369.
http://dx.doi.org/10.1039/b811230d
HEI (Health Effects Institute). (2010). Traffic-related air pollution: A critical review of the literature on
emissions, exposure, and health effects. (Special Report 17). Boston, MA: Health Effects Institute
(HEI). http://pubs.healtheffects.org/view.php?id=334
Heinrich. J: Thiering. E: Rzehak. P: Kramer. U: Hochadel. M: Rauchfuss. KM: Gehring. U: Wichmann.
HE. (2013). Long-term exposure to NO2 and PM10 and all-cause and cause-specific mortality in a
prospective cohort of women. Occup Environ Med 70: 179-186. http://dx.doi.org/10.1136/oemed-
2012-100876
Held. T: Chang. DPY: Niemeier. DA. (2003). UCD 2001: An improved model to simulate pollutant
dispersion from roadways. Atmos Environ 37: 5325-5336.
Hendrick. EM: Tino. VR: Hanna. SR: Egan. BA. (2013). Evaluation of NO2 predictions by the plume
volume molar ratio method (PVMRM) and ozone limiting method (OLM) in AERMOD using new
field observations. J Air Waste Manag Assoc 63: 844-854.
http://dx.doi.org/10.1080/10962247.2013.798599
Hong. JH: Lee. JH: Park. SS. (2007). Application of an in-situ measurement system to determine HONO
levels in an indoor environment. Journal of Korean Society for Atmospheric Environment 23: 191-
202. http://dx.doi.Org/10.5572/KOSAE.2007.23.2.191
Huang. W: Zhu. T: Pan. X: Hu. M: Lu. SE: Lin. Y: Wang. T: Zhang. Y: Tang. X. (2012). Air pollution and
autonomic and vascular dysfunction in patients with cardiovascular disease: Interactions of systemic
inflammation, overweight, and gender. Am J Epidemiol 176: 117-126.
http://dx.doi.org/10.1093/aje/kwr511
Hwang. BF: Lee. YL. (2010). Air pollution and prevalence of bronchitic symptoms among children in
Taiwan. Chest 138: 956-964. http://dx.doi.org/10.1378/chest.09-2600
Hystad. P: Setton. E: Cervantes. A: Poplawski. K: Deschenes. S: Brauer. M: van Donkelaar. A: Lamsal
L: Martin. R: Jerrett M: Demers. P. (2011). Creating national air pollution models for population
exposure assessment in Canada. Environ Health Perspect 119: 1123-1129.
http://dx.doi.org/10.1289/ehp.1002976
Iniguez. C: Ballester. F: Amoros. R: Estarlich. M: Murcia. M: Esplugues. A: Rebagliato. M. (2009). The
impact of outdoor NO2 exposure on fetal growth assessed by ultrasounds during pregnancy
[Abstract]. Epidemiology 20: S78. http://dx.doi.org/10.1097/01.ede.0000362944.07976.ba
January 2015 3-106 DRAFT: Do Not Cite or Quote
-------
Iniguez. C: Ballester. F: Estarlich. M: Llop. S: Fernandez-Patier. R: Aguirre-Alfaro. A: Esplugues. A:
Inma Study group. V. (2009). Estimation of personal NO2 exposure in a cohort of pregnant women.
Sci Total Environ 407: 6093-6099. http://dx.doi.Org/10.1016/j.scitotenv.2009.08.006
Isaacs. K: McCurdv. T: Glen. G: Nysewander. M: Errickson. A: Forbes. S: Graham. S: McCurdv. L:
Smith. L: Tulve. N: Vallero. D. (2013). Statistical properties of longitudinal time-activity data for use
in human exposure modeling. J Expo Sci Environ Epidemiol 23: 328-336.
http://dx.doi.org/10.1038/ies.2012.94
Isakov. V: Touma. JS: Burke. J: Lobdell DT: Palma. T: Rosenbaum. A: Ozkaynak. H. (2009). Combining
regional- and local-scale air quality models with exposure models for use in environmental health
studies. J Air Waste Manag Assoc 59: 461-472. http://dx.doi.Org/10.3155/1047-3289.59.4.461
Ivy. D: Mulholland. JA: Russell. AG. (2008). Development of ambient air quality population-weighted
metrics for use in time-series health studies. J Air Waste Manag Assoc 58: 711-720.
http://dx.doi.0rg/10.3155/1047-3289.58.5.711
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
Janssen. S: Dumont G: Fierens. F: Deutsch. F: Maiheu. B: Celis. D: Trimpeneers. E: Mensink. C. (2012).
Land use to characterize spatial representativeness of air quality monitoring stations and its relevance
for model validation. Atmos Environ 59: 492-500. http://dx.doi.Org/10.1016/j.atmosenv.2012.05.028
Jarvis. PL: Leaderer. BP: Chinn. S: Burnev. PG. (2005). Indoor nitrous acid and respiratory symptoms
and lung function in adults. Thorax 60: 474-479. http://dx.doi.org/10.1136/tfax.2004.032177
Ji Y: Gao. Y: Li. G: An. T. (2012). Theoretical study of the reaction mechanism and kinetics of low-
molecular-weight atmospheric aldehydes (C1C4) with NO2. Atmos Environ 54: 288-295.
http://dx.doi.0rg/10.1016/j.atmosenv.2012.02.040
Jimenez. AS: Heal MR: Beverland. IJ. (2011). Intercomparison study of NO(x) passive diffusion tubes
with chemiluminescence analysers and evaluation of bias factors. Atmos Environ 45: 3062-3068.
http://dx.doi.0rg/10.1016/j.atmosenv.2011.03.011
Johnson. M: Isakov. V: Touma. JS: Mukerjee. S: Ozkaynak. H. (2010). Evaluation of land-use regression
models used to predict air quality concentrations in an urban area. Atmos Environ 44: 3660-3668.
http://dx.doi.0rg/10.1016/j.atmosenv.2010.06.041
Karner. AA: Eisinger. PS: 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
Karr. CJ: Rudra. CB: Miller. KA: Gould. TR: Larson. T: Sathyanaravana. S: Koenig. JQ. (2009). Infant
exposure to fine paniculate matter and traffic and risk of hospitalization for RSV bronchiolitis in a
region with lower ambient air pollution. Environ Res 109: 321-327.
http://dx.doi.0rg/10.1016/j.envres.2008.ll.006
Katanoda. K: Sobue. T: Satoh. H: Tajima. K: Suzuki. T: Nakatsuka. H: Takezaki. T: Nakavama. T: Nitta.
H: Tanabe. K: Tominaga. S. (2011). An association between long-term exposure to ambient air
pollution and mortality from lung cancer and respiratory diseases in Japan. J Epidemiol 21: 132-143.
http://dx.doi.org/10.2188/iea.JE20100098
Kheirbek. I: Ito. K: Neitzel R: Kim. J: Johnson. S: Ross. Z: Eisl H: Matte. T. (2014). Spatial Variation in
Environmental Noise and Air Pollution in New York City. J Urban Health 91: 415-431.
http://dx.doi.org/10.1007/sll524-013-9857-0
Khoder. MI. (2002). Nitrous acid concentrations in homes and offices in residential areas in Greater
Cairo. J Environ Monit 4: 573-578. http://dx.doi.org/10.1039/b201818g
January 2015 3-107 DRAFT: Do Not Cite or Quote
-------
Kim. D: Sass-Kortsak. A: Purdham. JT: Dales. RE: Brook. JR. (2006). Associations between personal
exposures and fixed-site ambient measurements of fine paniculate matter, nitrogen dioxide, and
carbon monoxide in Toronto, Canada. J Expo Sci Environ Epidemiol 16: 172-183.
http://dx.doi.org/10.1038/sj.jea.7500446
Kim. SW: Mckeen. SA: Frost. GJ: Lee. SH: Trainer. M: Richter. A: Angevine. WM: Atlas. E: Bianco. L:
Boersma. KF: Brioude. J: Burrows. JP: de Gouw. J: Fried. A: Gleason. J: Hilboll A: Mellqvist J:
Peischl. J: Richter. D: Rivera. C: Ryerson. T: Te Lintel Hekkert. S: Walega. J: Warneke. C: Weibring.
P: Williams. E. (2011). Evaluations of NOx and highly reactive VOC emission inventories in Texas
and their implications for ozone plume simulations during the Texas Air Quality Study 2006. Atmos
Chem Phys 11: 11361-11386. http://dx.doi.org/10.5194/acp-ll-11361-2011
Kingslev. SL: Eliot. MN: Carlson. L: Finn. J: Macintosh. PL: Suh. HH: Wellenius. GA. (2014). Proximity
of US schools to major roadways: a nationwide assessment. J Expo Sci Environ Epidemiol 24: 253-
259. http://dx.doi.0rg/10.1038/ies.2014.5
Klapmever. ME: Marr. LC. (2012). CO2, NOx, and particle emissions from aircraft and support activities
at a regional airport. Environ Sci Technol 46: 10974-10981. http://dx.doi.org/10.1021/es302346x
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.
http://dx.doi.org/10.1038/sj.jea.7500165
Ko. FW: Tarn. W: Wong. TW: Chan. DP: Tung. AH: Lai. CKW: Hui. DSC. (2007). Temporal relationship
between air pollutants and hospital admissions for chronic obstructive pulmonary disease in Hong
Kong. Thorax 62: 780-785. http://dx.doi.org/10.1136/thx.2006.076166
Konovalov. IB: Beekmann. M: Richter. A: Burrows. JP: Hilboll A. (2010). Multi-annual changes of NOx
emissions in megacity regions: nonlinear trend analysis of satellite measurement based estimates.
Atmos Chem Phys 10: 8481-8498. http://dx.doi.org/10.5194/acp-10-8481-2010
Kornartit C: Sokhi. RS: Burton. MA: Ravindra. K. (2010). Activity pattern and personal exposure to
nitrogen dioxide in indoor and outdoor microenvironments. Environ Int 36: 36-45.
http://dx.doi.0rg/10.1016/j.envint.2009.09.004
Kousa. A: Monn. C: Rotko. T: Aim. S: Oblesbv. L: Jantunen. MJ. (2001). Personal exposures to NO2 in
the EXPOLIS-study: Relation to residential indoor, outdoor and workplace concentrations in Basel,
Helsinki and Prague. Atmos Environ 35: 3405-3412.
Koutrakis. P: Suh. HH: Sarnat. JA: Brown. KW: Coull BA: Schwartz. J. (2005). Characterization of
particulate and gas exposures of sensitive subpopulations living in Baltimore and Boston (pp. 1-65;
discussion 67-75). (ISSN 1041-5505
131). Boston, MA: Health Effects Institute, http://pubs.healtheffects.org/view.php?id=91
Kruize. H: Hanninen. O: Breugelmans. O: Lebret E: Jantunen. M. (2003). Description and demonstration
of the EXPOLIS simulation model: two examples of modeling population exposure to particulate
matter. J Expo Sci Environ Epidemiol 13: 87-99. http://dx.doi.org/10.1038/sj.jea.7500258
Kumar. U: Jain. VK. (2009). ARIMA forecasting of ambient air pollutants (O3, NO, NO2 and CO). Stoch
Environ Res Risk Assess 24: 751-760. http://dx.doi.org/10.1007/s00477-009-0361-8
LaFranchi. BW: Goldstein. AH: Cohen. RC. (2011). Observations of the temperature dependent response
of ozone to NOx reductions in the Sacramento, CA urban plume. Atmos Chem Phys 11: 6945-6960.
http://dx.doi.org/10.5194/acp-ll-6945-2011
Langner. C: Klemm. O. (2011). A comparison of model performance between AERMOD and
AUSTAL2000. J Air Waste Manag Assoc 61: 640-646.
Laurent O: Wu. J: Li. L: Chung. J: Bartell S. (2013). Investigating the association between birth weight
and complementary air pollution metrics: a cohort study. Environ Health 12: 18.
http://dx.doi.org/10.1186/1476-069X-12-18
January 2015 3-108 DRAFT: Do Not Cite or Quote
-------
Le. ND: Zidek. JV. (2006). Statistical analysis of environmental spacetime processes. New York, NY:
Springer, http://dx.doi.org/10.1007/0-387-35429-8
Leaderer. BP: Naeher. L: Jankun. T: Balenger. K: Holford. TR: Toth. C: Sullivan. J: Wolfson. JM:
Koutrakis. P. (1999). Indoor, outdoor, and regional summer and winter concentrations of PM10,
PM25, SO42-, H+, NH4+, NO3-, NH3, and nitrous acid in homes with and without kerosene space
heaters. Environ Health Perspect 107: 223-231.
Lee. K: Xue. J: Geyh. AS: Ozkaynak. H: Leaderer. BP: Weschler. CJ: Spengler. JD. (2002). Nitrous acid,
nitrogen dioxide, and ozone concentrations in residential environments. Environ Health Perspect 110:
145-150.
Lee. K: Yeom. J: Yoon. C: Yang. W: Son. BS: Jeon. JM: Yu. SD: Park. CH. (2013). Seasonal and
geographic effects on predicting personal exposure to nitrogen dioxide by time-weighted
microenvironmental model. Atmos Environ 67: 143-148.
http://dx.doi.0rg/10.1016/j.atmosenv.2012.ll.010
Lepeule. J: Caini. F: Bottagisi S: Galineau. J: Hulin. A: Marquis. N: Bohet. A: Siroux. V: Kaminski. M:
Charles. MA: Slama. R. (2010). Maternal exposure to nitrogen dioxide during pregnancy and
offspring birth weight: Comparison of two exposure models. Environ Health Perspect 118: 1483-1489.
http://dx.doi.org/10.1289/ehp.0901509
Lew. I: Mihele. C: Lu. G: Naravan. J: Hilker. N: Brook. JR. (2014). Elucidating multipollutant exposure
across a complex metropolitan area by systematic deployment of a mobile laboratory. Atmos Chem
Phys 14: 7173-7193. http://dx.doi.org/10.5194/acp-14-7173-2014
Li. L: Wu. J: Wilhelm. M: Ritz. B. (2012). Use of generalized additive models and cokriging of spatial
residuals to improve land-use regression estimates of nitrogen oxides in Southern California. Atmos
Environ 55: 220-228. http://dx.doi.Org/10.1016/j.atmosenv.2012.03.035
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. http://dx.doi.org/10.1006/enrs.1999.3993
Linaker. CH: Chauhan. AJ: Inskip. HM: Holgate. ST: Coggon. D. (2000). Personal exposures of children
to nitrogen dioxide relative to concentrations in outdoor air. Occup Environ Med 57: 472-476.
Lindstrom. J: Szpiro. AA: Sampson. PD: Oron. AP: Richards. M: Larson. TV: Sheppard. L. (2013). A
flexible spatio-temporal model for air pollution with spatial and spatio-temporal covariates. Environ
Ecol Stat 21: 411-433. http://dx.doi.org/10.1007/sl0651-013-0261-4
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.
Lipfert FW: Wyzga. RE: Baty. JD: Miller. JP (2009). Air pollution and survival within the Washington
University-EPRI veterans cohort: Risks based on modeled estimates of ambient levels of hazardous
and criteria air pollutants. J Air Waste Manag Assoc 59: 473-489. http://dx.doi.org/10.3155/1047-
3289.59.4.473
Liu. L: Poon. R: Chen. L: Frescura. AM: Montuschi. P: Ciabattoni. G: Wheeler. A: Dales. R. (2009).
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
Loupa. G: Rapsomanikis. S. (2008). Air pollutant emission rates and concentrations in medieval churches.
J Atmos Chem 60: 169-187. http://dx.doi.org/10.1007/sl0874-008-9116-z
Lu. R: Turco. RP: Jacobson. MZ. (1997a). An integrated air pollution modeling system for urban and
regional scales: 1. Structure and performance. J Geophys Res 102: 6063-6079.
http://dx.doi.org/10.1029/96JD03501
January 2015 3-109 DRAFT: Do Not Cite or Quote
-------
Lu. R: Turco. RP: Jacobson. MZ. (1997b). An integrated air pollution modeling system for urban and
regional scales: 2. Simulations for SCAQS 1987. J Geophys Res Atmos 102: 6081-6098.
http://dx.doi.org/10.1029/96JD03502
Maantav. JA: Tu. J: Maroko. AR. (2009). Loose-coupling an air dispersion model and a geographic
information system (GIS) for studying air pollution and asthma in the Bronx, New York City. Int J
Environ Health Res 19: 59-79. http://dx.doi.org/10.1080/09603120802392868
MacNaughton. P: Melly. S: Vallarino. J: Adamkiewicz. G: Spengler. JD. (2014). Impact of bicycle route
type on exposure to traffic-related air pollution. Sci Total Environ 490: 37-43.
http://dx.doi.0rg/10.1016/j.scitotenv.2014.04.lll
Madsen. C: Gehring. U: Walker. SE: Brunekreef. B: Stigum. H: Naess. O: Nafstad. P. (2010). Ambient air
pollution exposure, residential mobility and term birth weight in Oslo, Norway. Environ Res 110: 363-
371. http://dx.doi.0rg/10.1016/j.envres.2010.02.005
Mannes. 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
Mar. TF: Norris. GA: Koenig. JQ: Larson. TV. (2000). Associations between air pollution and mortality in
Phoenix, 1995-1997. Environ Health Perspect 108: 347-353.
Marshall. JD: Nethery. E: Brauer. M. (2008). Within-urban variability in ambient air pollution:
Comparison of estimation methods. Atmos Environ 42: 1359-1369.
http://dx.doi.0rg/10.1016/j.atmosenv.2007.08.012
Martins. PC: Valente. J: Papoila. AL: Caires. I: Araujo-Martins. J: Mata. P: Lopes. M: Torres. S: Rosado-
Pinto. J: Borrego. C: Annesi-Maesano. I: Neuparth. N. (2012). Airways changes related to air
pollution exposure in wheezing children. Eur Respir J 39: 246-253.
http://dx.doi.org/10.1183/09031936.00025111
Masiol. M: Agostinelli C: Formenton. G: Tarabotti. E: Pavoni. B. (2014). Thirteen years of air pollution
hourly monitoring in a large city: Potential sources, trends, cycles and effects of car-free days. Sci
Total Environ 494-495: 84-96. http://dx.doi.Org/10.1016/j.scitotenv.2014.06.122
Matte. TD: Ross. Z: Kheirbek. I: Eisl. H: Johnson. S: Gorczynski. JE: Kass. D: Markowitz. S: Pezeshki.
G: Cloughertv. JE. (2013). Monitoring intraurban spatial patterns of multiple combustion air pollutants
in New York City: Design and implementation. J Expo Sci Environ Epidemiol 23: 223-231.
http://dx.doi.org/10.1038/ies.2012.126
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. http://dx.doi.org/10.1164/rccm.200304-466OC
McConnell R: Islam. T: Shankardass. K: Jerrett M: Lurmann. F: Gilliland. F: Gauderman. J: Avol. E:
Kiinzli. 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.
http://dx.doi.org/10.1289/ehp.0901232
Mehta. S: Ngo. LH: Dzung. DV: Cohen. A: Thach. TO: Dan. VX: Tuan. ND: Giang. LT. (2013). Air
pollution and admissions for acute lower respiratory infections in young children of Ho Chi Minh
City. Air Qual Atmos Health 6: 167-179. http://dx.doi.org/10.1007/sll869-011-0158-z
Meng. Q: Williams. R: Pinto. JP (2012a). Determinants of the associations between ambient
concentrations and personal exposures to ambient PM2.5, NO2, and O3 during DEARS. Atmos
Environ 63: 109-116. http://dx.doi.Org/10.1016/j.atmosenv.2012.09.019
Meng. QY Svendsgaard. D: Kotchmar. DJ: Pinto. JP. (2012b). Associations between personal exposures
and ambient concentrations of nitrogen dioxide: A quantitative research synthesis. Atmos Environ 57:
322-329. http://dx.doi.0rg/10.1016/i.atmosenv.2012.04.035
January 2015 3-110 DRAFT: Do Not Cite or Quote
-------
Mercer. LD: Szpiro. AA: Sheppard. L: Lindstrom. J: Adar. SD: Allen. RW: Avol. EL: Oron. AP: Larson.
T: Liu. LJ: Kaufman. JD. (2011). Comparing universal kriging and land-use regression for predicting
concentrations of gaseous oxides of nitrogen (NOx) for the Multi-Ethnic Study of Atherosclerosis and
Air Pollution (MESA Air). Atmos Environ 45: 4412-4420.
http://dx.doi.0rg/10.1016/j.atmosenv.2011.05.043
Minguillon. MC: Rivas. I: Aguilera. I: Alastuev. A: Moreno. T: Amato. F: Sunyer. J: Querol X. (2012).
Within-city contrasts in PM composition and sources and their relationship with nitrogen oxides. J
Environ Monit 14: 2718-2728. http://dx.doi.org/10.1039/c2em30469d
Misra. A: Roorda. MJ: Maclean. HL. (2013). An integrated modelling approach to estimate urban traffic
emissions. Atmos Environ 73: 81-91. http://dx.doi.Org/10.1016/j.atmosenv.2013.03.013
Molloy. SB: Cheng. M: Galbally. IE: Kevwood. MD: Lawson. SJ: Powell JC: Gillett R: Dunne. E:
Selleck. PW. (2012). Indoor air quality in typical temperate zone Australian dwellings. Atmos Environ
54: 400-407. http://dx.doi.Org/10.1016/j.atmosenv.2012.02.031
Molter. A: Lindlev. S: de Vocht. F: Agius. R: Kerry. G: Johnson. K: Ashmore. M: Terry. A:
Dimitroulopoulou. S: Simpson. A. (2012). Performance of a microenviromental model for estimating
personal NO2 exposure in children. Atmos Environ 51: 225-233.
http://dx.doi.0rg/10.1016/j.atmosenv.2012.01.030
Molter. A: Lindlev. S: de Vocht. F: Simpson. A: Agius. R. (2010). Modelling air pollution for
epidemiologic research—Part I: A novel approach combining land use regression and air dispersion.
Sci Total Environ 408: 5862-5869. http://dx.doi.Org/10.1016/j.scitotenv.2010.08.027
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
Moshammer. H: Hutter. HP: Hauck. H: Neuberger. M. (2006). Low levels of air pollution induce changes
of lung function in a panel of schoolchildren. Eur Respir J 27: 1138-1143.
http://dx.doi.org/10.1183/09031936.06.00089605
Moustris. KP: Ziomas. 1C: Paliatsos. AG. (2009). 3-Day-ahead forecasting of regional pollution index for
the pollutants NO2, CO, SO2, and O3 using artificial neural networks in Athens, Greece. Water Air
Soil Pollut 209: 29-43. http://dx.doi.org/10.1007/sll270-009-0179-5
Mukala. K: Aim. S: Tiittanen. P: Salonen. RO: Jantunen. M: Pekkanen. J. (2000). Nitrogen dioxide
exposure assessment and cough among preschool children. Arch Environ Health 55: 431-438.
Mukerjee. S: Smith. L: Neas. L: Norris. G. (2012a). Evaluation of land use regression models for nitrogen
dioxide and benzene in Four US cities. Scientific World Journal 2012: 865150.
http://dx.doi.org/10.1100/2012/865150
Mukerjee. S: Smith. LA: Johnson. MM: Neas. LM: Stallings. CA. (2009). Spatial analysis and land use
regression of VOCs and NO(2) from school-based urban air monitoring in Detroit/Dearborn, USA. Sci
Total Environ 407: 4642-4651. http://dx.doi.Org/10.1016/j.scitotenv.2009.04.030
Mukerjee. S: Willis. RD: Walker. JT: Hammond. D: Norris. GA: Smith. LA: Welch. DP: Peters. TM.
(2012b). Seasonal effects in land use regression models for nitrogen dioxide, coarse paniculate matter,
and gaseous ammonia in Cleveland, Ohio. Atmos Pollut Res 3: 352-361.
http://dx.doi.org/10.5094/APR.2012.039
Murphy. JG: Day. DA: Clearv. PA: Wooldridge. PJ: Millet. DB: Goldstein. AH: Cohen. RC. (2006). The
weekend effect within and downwind of Sacramento: Part 2. Observational evidence for chemical and
dynamical contributions. Atmos Chem Phys Discuss 6: 11971-12019. http://dx.doi.org/10.5194/acpd-
6-11971-2006
Murphy. JG: Day. DA: Clearv. PA: Wooldridge. PJ: Millet. DB: Goldstein. AH: Cohen. RC. (2007). The
weekend effect within and downwind of Sacramento - Part 1: Observations of ozone, nitrogen oxides,
and VOC reactivity. Atmos Chem Phys 7: 5327-5339.
January 2015 3-111 DRAFT: Do Not Cite or Quote
-------
Nishikawa. Y: Murano. K: Mukai H. (2009). Comparison of sampling resistance for one to three sheets
of membrane type passive sampler. Water Air Soil Pollut 197: 241-247.
http://dx.doi.org/10.1007/sll270-008-9807-8
N0j gaard. JK. (2010). Indoor measurements of the sum of the nitrate radical, NO3, and nitrogen
pentoxide, N2O5 in Denmark. Chemosphere 79: 898-904.
http://dx.doi.0rg/10.1016/j.chemosphere.2010.02.025
N0j gaard. JK: Bilde. M: Stenbv. C: Nielsen. OJ: Wolkoff. P. (2006). The effect of nitrogen dioxide on
particle formation during ozonolysis of two abundant monoterpenes indoors. Atmos Environ 40: 1030-
1042. http://dx.doi.0rg/10.1016/j.atmosenv.2005.ll.029
Novotny. EV: Bechle. MJ: Millet. DB: Marshall. JD. (2011). National satellite-based land-use regression:
NO2 in the United States. Environ Sci Technol 45: 4407-4414. http://dx.doi.org/10.1021/esl03578x
O'Connor. GT: Neas. L: Vaughn. B: Kattan. M: Mitchell H: Grain. EF: Evans. R. Ill: 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/j.jaci.2008.02.020
Oftedal. B: Nystad. W: Brunekreef. B: Nafstad. P. (2009). Long-term traffic-related exposures and asthma
onset in schoolchildren in Oslo, Norway. Environ Health Perspect 117: 839-844.
http://dx.doi.org/10.1289/ehp.11491
Oiamo. TH: Luginaah. IN: Atari. DO: Gorev. KM. (2011). Air pollution and general practitioner access
and utilization: a population based study in Sarnia, 'Chemical Valley,' Ontario. Environ Health 10: 71.
http://dx.doi.org/10.1186/1476-069X-10-71
Oiamo. TH: Luginaah. IN: Buzzelli. M: Tang. K: Xu. X: Brook. JR: Johnson. M. (2012). Assessing the
spatial distribution of nitrogen dioxide in London, Ontario. J Air Waste Manag Assoc 62: 1335-1345.
http://dx.doi.org/10.1080/10962247.2012.715114
Ozden. O: Dogeroglu. T. (2008). Field evaluation of a tailor-made new passive sampler for the
determination of NO2 levels in ambient air. Environ Monit Assess 142: 243-253.
http://dx.doi.org/10.1007/sl0661-007-9921-x
Ozkaynak. H: Frev. HC: Burke. J: Finder. RW (2009). Analysis of coupled model uncertainties in source-
to-dose modeling of human exposures to ambient air pollution: A PM2.5 case study. Atmos Environ
43: 1641-1649. http://dx.doi.Org/10.1016/j.atmosenv.2008.12.008
Pachon. JE: Balachandran. S: Hu. Y: Mulholland. JA: Darrow. LA: Sarnat JA: Tolbert PE: Russell AG.
(2012). Development of outcome-based, multipollutant mobile source indicators. J Air Waste Manag
Assoc 62: 431-442. http://dx.doi.org/10.1080/10473289.2012.656218
Paciorek. CJ. (2010). The importance of scale for spatial-confounding bias and precision of spatial
regression estimators. Stat Sci 25: 107-125. http://dx.doi.org/10.1214/10-STS326
Padro-Martinez. LT: Patton. AP: Trull. JB: Zamore. W: Brugge. D: Durant JL. (2012). Mobile monitoring
of particle number concentration and other traffic-related air pollutants in a near-highway
neighborhood over the course of a year. Atmos Environ 61: 253-264.
http://dx.doi.0rg/10.1016/j.atmosenv.2012.06.088
Panteliadis. P: Strak. M: Hoek. G: Weijers. E: van Per Zee. S: Dijkema. M. (2014). Implementation of a
low emission zone and evaluation of effects on air quality by long-term monitoring. Atmos Environ
86: 113-119. http://dx.doi.0rg/10.1016/i.atmosenv.2013.12.035
Parenteau. MP: Sawada. MC. (2012). The role of spatial representation in the development of a LUR
model for Ottawa, Canada. Air Qual Atmos Health 5: 311-323. http://dx.doi.org/10.1007/sll869-010-
0094-3
Park. SS: Hong. JH: Lee. JH: Kim. YJ: Cho. SY: Kim. SJ. (2008). Investigation of nitrous acid
concentration in an indoor environment using an in-situ monitoring system. Atmos Environ 42: 6586-
6596. http://dx.doi.0rg/10.1016/i.atmosenv.2008.05.023
January 2015 3-112 DRAFT: Do Not Cite or Quote
-------
Pegas. PN: Nunes. T: Alves. CA: Silva. JR: Vieira. SLA: Caseiro. A: Pio. CA. (2012). Indoor and outdoor
characterisation of organic and inorganic compounds in city centre and suburban elementary schools
of Aveiro, Portugal. Atmos Environ 55: 80-89. http://dx.doi.Org/10.1016/j.atmosenv.2012.03.059
Peters. A: Breitner. S: Cyrys. J: Stolzel. M: Pitz. M: Wolke. G: Heinrich. J: Kreyling. W: Kiichenhoff. H:
Wichmann. H. (2009). The influence of improved air quality on mortality risks in Erfurt, Germany.
(137). Boston, MA: Health Effects Institute. http://pubs.healtheffects.org/view.php?id=299
Physick. W: Powell. J: Cope. M: Boast K: Lee. S. (2011). Measurements of personal exposure to NO2
and modelling using ambient concentrations and activity data. Atmos Environ 45: 2095-2102.
http://dx.doi.0rg/10.1016/j.atmosenv.2011.01.063
Piechocki-Minguy. A: Plaisance. H: Schadkowskj C: Sagnier. I: Saison. JY: Galloo. JC: Guillermo. R.
(2006). A case study of personal exposure to nitrogen dioxide using a new high sensitive diffusive
sampler. Sci Total Environ 366: 55-64. http://dx.doi.Org/10.1016/j.scitotenv.2005.08.009
Pirjola. L: Lahde. T: Niemi. JV: Kousa. A: Ronkko. T: Karjalainen. P: Keskinen. J: Frev. A: Hillamo. R.
(2012). Spatial and temporal characterization of traffic emissions in urban microenvironments with a
mobile laboratory. Atmos Environ 63: 156-167. http://dx.doi.Org/10.1016/j.atmosenv.2012.09.022
Poddubny. VA: Yushketova. NA. (2013). Aphysicochemical model of sorption processes in NO(2)
passive sampling with air humidity effects. Environ Monit Assess 185: 3819-3829.
http://dx.doi.org/10.1007/sl0661-012-2830-7
Polidorj A: Fine. PM. (2012). Ambient concentrations of criteria and air toxic pollutants in close
proximity to a freeway with heavy-duty diesel traffic. South Coast Air Quality Management District.
www.aqmd.gov/tao/AQ-Reports/I710Fwv Studv.pdf
Pollice. A: Jona lasinio. G. (2010). A multivariate approach to the analysis of air quality in a high
environmental risk area. Environmetrics 21: 741-754. http://dx.doi.org/10.1002/env.1059
Pratt. GC: Parson. K: Shinoda. N: Lindgren. P: Dunlap. S: Yawn. B: Wollan. P: Johnson. J. (2014).
Quantifying traffic exposure. J Expo Sci Environ Epidemiol 24: 290-296.
http://dx.doi.org/10.1038/ies.2013.51
Pusede. SE: Cohen. RC. (2012). On the observed response of ozone to NOx and VOC reactivity
reductions in San Joaquin Valley California 1995-present. Atmos Chem Phys 12: 8323-8339.
http://dx.doi.org/10.5194/acp-12-8323-2012
Raff. JD: Njegic. B: Chang. WL: Gordon. MS: Dabdub. D: Gerber. RB: Finlavson-Pitts. BJ. (2009).
Chlorine activation indoors and outdoors via surface-mediated reactions of nitrogen oxides with
hydrogen chloride. PNAS 106: 13647-13654. http://dx.doi.org/10.1073/pnas.0904195106
Ragettli MS: Tsai MY: Braun-Fahrlander. C: de Nazelle. A: Schindler. C: Ineichen. A: Ducret-Stich. RE:
Perez. L: Probst-Hensch. N: Kiinzlj N: Phuleria. HC. (2014). Simulation of population-based
commuter exposure to NO using different air pollution models. Int J Environ Res Public Health 11:
5049-5068. http://dx.doi.org/10.3390/ijerphll0505049
Rao. M: George. LA: Rosenstiel. TN: Shandas. V: Dinno. A. (2014). Assessing the relationship among
urban trees, nitrogen dioxide, and respiratory health. Environ Pollut.
http://dx.doi.0rg/10.1016/j.envpol.2014.07.011
Reeves. GK: Cox. PR: Darby. SC: Whitlev. E. (1998). Some aspects of measurement error in explanatory
variables for continuous and binary regression models. Stat Med 17: 2157-2177.
http://dx.doi.org/10.1002/(SICI)1097-0258(19981015)17:19<2157::AID-SIM916>3.0.CO:2-F
Rivas. I: Viana. M: Moreno. T: Pandolfi. M: Amato. F: Reche. C: Bouso. L: Alvarez-Pedrerol. M:
Alastuev. A: Sunyer. J: Querol X. (2014). Child exposure to indoor and outdoor air pollutants in
schools in Barcelona, Spain. Environ Int 69: 200-212. http://dx.doi.Org/10.1016/j.envint.2014.04.009
Roberts. S: Martin. MA. (2006). Using supervised principal components analysis to assess multiple
pollutant effects. Environ Health Perspect 114: 1877-1882. http://dx.doi.org/10.1289/ehp.9226
January 2015 3-113 DRAFT: Do Not Cite or Quote
-------
Robinson. DP: Lloyd. CD: Mckinlev. JM. (2013). Increasing the accuracy of nitrogen dioxide (NO2)
pollution mapping using geographically weighted regression (GWR) and geostatistics. Int J Appl
Earth Obs Geoinf 21: 374-383. http://dx.doi.Org/10.1016/j.jag.2011.ll.001
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
Rose. N: Cowie. C: Gillett R: Marks. GB. (2011). Validation of a spatiotemporal land use regression
model incorporating fixed site monitors. Environ Sci Technol 45: 294-299.
http://dx.doi.org/10.1021/esl00683t
Rosenlund. M: Forastiere. F: Stafoggia. M: Porta. D: Perucci M: Ranzi. A: Nussio. F: Perucci C. (2008).
Comparison of regression models with land-use and emissions data to predict the spatial distribution
of traffic-related air pollution in Rome. J Expo Sci Environ Epidemiol 18: 192-199.
http://dx.doi.org/10.1038/sj.jes.7500571
Ross. Z: Ito. K: Johnson. S: Yee. M: Pezeshki. G: Cloughertv. JE: Savitz. D: Matte. T. (2013). Spatial and
temporal estimation of air pollutants in New York City: exposure assignment for use in a birth
outcomes study. Environ Health 12: 51. http://dx.doi.org/10.1186/1476-069X-12-51
Ross. Z: Kheirbek. I: Cloughertv. JE: Ito. K: Matte. T: Markowitz. S: Eisl. H. (2011). Noise, air pollutants
and traffic: Continuous measurement and correlation at a high-traffic location in New York City.
Environ Res 111: 1054-1063. http://dx.doi.Org/10.1016/j.envres.2011.09.004
Sahsuvaroglu. T: Su. JG: Brook. J: Burnett. R: Loeb. M: Jerrett M. (2009). Predicting personal nitrogen
dioxide exposure in an elderly population: Integrating residential indoor and outdoor measurements,
fixed-site ambient pollution concentrations, modeled pollutant levels, and time-activity patterns. J
Toxicol Environ Health A 72: 1520-1533. http://dx.doi.org/10.1080/15287390903129408
Sajani. SZ: Hanninen. O: Marchesi S: Lauriola. R (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
Samoli. E: Nastos. PT: Paliatsos. AG: Katsouvanni. K: Priftis. KN. (2011). Acute effects of air pollution
on pediatric asthma exacerbation: Evidence of association and effect modification. Environ Res 111:
418-424. http://dx.doi.0rg/10.1016/i.envres.2011.01.014
Sanchez Jimenez. A: Heal MR: Beverland. IJ. (2012). Correlations of particle number concentrations and
metals with nitrogen oxides and other traffic-related air pollutants in Glasgow and London. Atmos
Environ 54: 667-678. http://dx.doi.Org/10.1016/j.atmosenv.2012.01.047
Sarnat JA: Brown. KW: Schwartz. J: Coull BA: Koutrakis. R (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. J Air Waste Manag Assoc 50: 1184-
1198. http://dx.doi.org/10.1080/10473289.2000.10464165
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: Ruiz. PA: Koutrakis. P: Suh. HH. (2006). The influences of ambient particle
composition and size on particle infiltration in Los Angeles, CA, residences. J Air Waste Manag Assoc
56: 186-196. http://dx.doi.org/10.1080/10473289.2006.10464449
Sarnat. SE: 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/jes.2009.10
January 2015 3-114 DRAFT: Do Not Cite or Quote
-------
Sarnat SE: Ravsonj AU: Li. WW: Holguin. F: Johnson. BA: Flores Luevano. S: Garcia. JH: Sarnat JA.
(2012). Air pollution and acute respiratory response in a panel of asthmatic children along the U.S.-
Mexico border. Environ Health Perspect 120: 437444. http://dx.doi.org/10.1289/ehp.1003169
Sarnat. SE: Sarnat. JA: Mulholland. J: Isakov. V: Ozkaynak. H: Chang. HH: Klein. M: Tolbert. PE.
(2013). Application of alternative spatiotemporal metrics of ambient air pollution exposure in a time-
series epidemiological study in Atlanta. J Expo Sci Environ Epidemiol 23: 593-605.
http://dx.doi.org/10.1038/ies.2013.41
Sarwar. G: Corsi R: Kumura. Y: Allen. D: Weschler. CJ. (2002). Hydroxyl radicals in indoor
environments. Atmos Environ 36: 3973-3988. http://dx.doi.org/10.1016/S1352-2310(02)00278-9
Sather. ME: Slonecker. ET: Mathew. J: Daughtrev. H: Williams. DP. (2007). Evaluation of Ogawa
Passive Sampling Devices as an alternative measurement method for the Nitrogen Dioxide annual
standard in El Paso, Texas. Environ Monit Assess 124: 211-221. http://dx.doi.org/10.1007/sl0661-
006-9219-4
Schembarj A: Triguero-Mas. M: De Nazelle. A: Dadvand. P: Vrijheid. M: Cirach. M: Martinez. D:
Figueras. F: Querol. X: Basagana. X: Eeftens. M: Meliefste. K: Nieuwenhuijsen. MJ. (2013).
Personal, indoor and outdoor air pollution levels among pregnant women. Atmos Environ 64: 287-
295. http://dx.doi.0rg/10.1016/j.atmosenv.2012.09.053
Schildcrout JS: Sheppard. L: Lumlev. T: Slaughter. JC: Koenig. JQ: 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/aje/kwj225
Sekine. Y: Watts. SF: Rendell. A: Butsugan. M. (2008). Development of highly Sensitive passive sampler
for nitrogen dioxide using porous polyethylene membrane filter as turbulence limiting diffuser. Atmos
Environ 42: 4079-4088. http://dx.doi.Org/10.1016/j.atmosenv.2008.01.029
Sellier. Y: Galineau. J: Hulin. A: Caini. F: Marquis. N: Navel. V: Bottagisi. S: Giorgis-Allemand. L:
Jacquier. C: Slama. R: Lepeule. J: group. tEmcs. (2014). Health effects of ambient air pollution: Do
different methods for estimating exposure lead to different results? Environ Int 66: 165-173.
http://dx.doi.0rg/10.1016/j.envint.2014.02.001
Sereviciene. V: Paliulis. D. (2012). Influence of temperature and relative humidity on the performance of
nitrogen dioxide diffusive sampler. Res J of Chem Sci 2: 89-92.
Setton. E: Marshall JD: Brauer. M: Lundquist KR: Hystad. P: Keller. P: Cloutier-Fisher. D. (2011). The
impact of daily mobility on exposure to traffic-related air pollution and health effect estimates. J Expo
Sci Environ Epidemiol 21: 42-48. http://dx.doi.org/10.1038/jes.2010.14
Sheppard. L. (2005). Acute air pollution effects: Consequences of exposure distribution and
measurements. J Toxicol Environ Health A 68: 1127-1135.
http://dx.doi.org/10.1080/15287390590935987
Sheppard. L: Slaughter. JC: Schildcrout. J: Liu. L. -JS: Lumlev. T. (2005). Exposure and measurement
contributions to estimates of acute air pollution effects. J Expo Sci Environ Epidemiol 15: 366-376.
http://dx.doi.org/10.1038/sj.jea.7500413
Shmool. JLC: Michanowicz. PR: Cambal. L: Tunno. B: Howell. J: Gillooly. S: Roper. C: Tripathy. S:
Chubb. LG: Eisl HM: Gorczynskj JE: Holguin. FE: Shields. KN: Cloughertv. JE. (2014). Saturation
sampling for spatial variation in multiple air pollutants across an inversion-prone metropolitan area of
complex terraing. Environ Health 13: 28. http://dx.doi.org/10.1186/1476-069X-13-28
Simon. PK: Dasgupta. PK. (1995). Continuous automated measurement of the soluble fraction of
atmospheric paniculate matter. Anal Chem 67: 71-78. http://dx.doi.org/10.1021/ac00097a012
Singh. KP: Gupta. S: Kumar. A: Shukla. SP (2012). Linear and nonlinear modeling approaches for urban
air quality prediction. Sci Total Environ 426: 244-255.
http://dx.doi.0rg/10.1016/i.scitotenv.2012.03.076
January 2015 3-115 DRAFT: Do Not Cite or Quote
-------
Skouloudis. AN: Kassomenos. P. (2014). Combining environment and health information systems for the
assessment of atmospheric pollution on human health. Sci Total Environ 488: 364-370.
http://dx.doi.0rg/10.1016/j.scitotenv.2014.02.116
Smith. LA: Mukerjee. S: Chung. KG: Afghani. J. (2011). Spatial analysis and land use regression of
VOCs and NO2 in Dallas, Texas during two seasons. J Environ Monit 13: 999-1007.
http://dx.doi.org/10.1039/cOem00724b
Snyder. MG: Venkatram. A: Heist. DK: Perry. SG: Petersen. WB: Isakov. V. (2013). RLINE: Aline source
dispersion model for near-surface releases. Atmos Environ 77: 748-756.
http://dx.doi.0rg/10.1016/j.atmosenv.2013.05.074
Spengler. JD: Brauer. M: Samet JM: Lambert. WE. (1993). Nitrous acid in Albuquerque, New Mexico,
homes. Environ SciTechnol 27: 841-845. http://dx.doi.org/10.1021/es00042a005
Spicer. CW: Coutant. RW: Ward. GF: Joseph. DW: Gaynor. AJ: Billick. IH. (1989). Rates and
mechanisms of NO2 removal from indoor air by residential materials. Environ Int 15: 643-654.
http://dx.doi.org/10.1016/0160-4120(89)90087-l
Spiegelman. D. (2013). Regression calibration in air pollution epidemiology with exposure estimated by
spatio-temporal modeling. Environmetrics 24: 521-524. http://dx.doi.org/10.1002/env.2249
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/j.envres.2007.08.006
Stieb. DM: Burnett. RT: Smith-Doiron. M: Brion. O: Shin. HH: Economou. V. (2008). Anew
multipollutant, no-threshold air quality health index based on short-term associations observed in daily
time-series analyses. J Air Waste Manag Assoc 58: 435-450. http://dx.doi.org/10.3155/1047-
3289.58.3.435
Strak. M: Hoek. G: Godri. KJ: Gosens. I: Mudwav. IS: van Oerle. R: Spronk. HM: Cassee. FR: Lebret E:
Kelly. FJ: Harrison. RM: Brunekreef. B: Steenhof. M: Janssen. NA. (2013). Composition of PM
affects acute vascular inflammatory and coagulative markers - the RAPTES project. PLoS ONE 8:
e58944. http://dx.doi.org/10.1371/journal.pone.0058944
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 10: 36.
http://dx.doi.org/10.1186/1476-069X-10-36
Su. JG: Brauer. M: Ainslie. B: Steyn. D: Larson. T: Buzzelli. M. (2008a). An innovative land use
regression model incorporating meteorology for exposure analysis. Sci Total Environ 390: 520-529.
http://dx.doi.0rg/10.1016/j.scitotenv.2007.10.032
Su. JG: Brauer. M: Buzzelli. M. (2008b). Estimating urban morphometry at the neighborhood scale for
improvement in modeling long-term average air pollution concentrations. Atmos Environ 42: 7884-
7893. http://dx.doi.0rg/10.1016/j.atmosenv.2008.07.023
Su. JG: Jerrett M: Beckerman. B. (2009a). A distance-decay variable selection strategy for land use
regression modeling of ambient air pollution exposures. Sci Total Environ 407: 3890-3898.
http://dx.doi.0rg/10.1016/i.scitotenv.2009.01.061
January 2015 3-116 DRAFT: Do Not Cite or Quote
-------
Su. JG: Jerrett M: Beckerman. B: Wilhelm. M: Ghosh. JK: Ritz. B. (2009b). Predicting traffic-related air
pollution in Los Angeles using a distance decay regression selection strategy. Environ Res 109: 657-
670. http://dx.doi.0rg/10.1016/j.envres.2009.06.001
Suh. HH: Zanobetti. A. (2010). Exposure error masks the relationship between traffic-related air pollution
and heart rate variability. J Occup Environ Med 52: 685-692.
http://dx.doi.org/10.1097/JOM.Ob013e3181e8071f
Szpiro. AA: Paciorek. CJ. (2013). Measurement error in two-stage analyses, with application to air
pollution epidemiology. Environmetrics 24: 501-517. http://dx.doi.org/10.1002/env.2233
Szpiro. AA: Paciorek. CJ: Sheppard. L. (2011). Does more accurate exposure prediction necessarily
improve health effect estimates? Epidemiology 22: 680-685.
http://dx.doi.org/10.1097/EDE.Ob013e3182254cc6
Tao. Y: Huang. W: Huang. X: Zhong. L: Lu. SE: Li. Y: Dai. L: Zhang. Y: Zhu. T. (2012). Estimated acute
effects of ambient ozone and nitrogen dioxide on mortality in the Pearl River Delta of Southern China.
Environ Health Perspect 120: 393-398. http://dx.doi.org/10.1289/ehp.1103715
Thornton. JA: Braban. CF: Abbatt JPD. (2003). N2O5 hydrolysis on sub-micron organic aerosol: the
effect of relative humidity, particle phase, and particle size. Phys Chem Chem Phys 5: 4593-4603.
Timonen. KL: Vanninen. E: De Hartog. J: Ibald-Mulli. A: Brunekreef B: Gold. PR: Henrich. J: Hoek. G:
Lankj T: Peters. A: Tarkiainen. T: Tiittanen. P: Kreyling. W: Pekkanen. J. (2006). Effects of ultrafine
and fine particulate and gaseous air pollution on cardiac autonomic control in subjects with coronary
artery disease: The ULTRA study. J Expo Sci Environ Epidemiol 16: 332-341.
http://dx.doi.org/10.1038/sj.jea.7500460
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.7500625
Trasande. L: Wong. K: Roy. A: Savitz. DA: Thurston. G. (2013). Exploring prenatal outdoor air pollution,
birth outcomes and neonatal health care utilization in a nationally representative sample. J Expo Sci
Environ Epidemiol 23: 315-321. http://dx.doi.org/10.1038/jes.2012.124
U.S. Census Bureau. (2007). 2007 American community survey [Database]. Washington, DC: U.S.
Department of Commerce, U.S. Census Bureau. Retrieved from
http://www.census.gov/acs/www/index.html
U.S. Census Bureau. (2009). American housing survey for the United States: 2009. Washington, DC: U.S.
Government Printing Office, http://www.census.gov/prod/2011pubs/hl50-09.pdf
U.S. Census Bureau. (2013). American community survey 2013 1-year estimate. Available online at
http://factfinder2.census.gov/faces/tableservices/jsf/pages/productview.xhtml?pid=ACS 13 1YRB01
003&prodType=table
U.S. EPA (U.S. Environmental Protection Agency). (1995). A user's guide for the CALPUFF dispersion
model [EPAReport]. (EPA/454/B-95/006). Research Triangle Park, NC: U.S. Environmental
Protection Agency, Office of Air Quality, Planning, and Standards.
http://nepis.epa. gov/Exe/ZyPURL.cgi?Dockev=2001206A.txt
U.S. EPA (U.S. Environmental Protection Agency). (2008). Integrated science assessment for oxides of
nitrogen - Health criteria: Annexes [EPA Report]. (EPA/600/R-08/072). Research Triangle Park, NC:
U.S. Environmental Protection Agency, National Center for Environmental Assessment.
http://cfpub.epa.gov/ncea/cfm/recordisplav.cfm7de id=194645
U.S. EPA (U.S. Environmental Protection Agency). (2009). Integrated science assessment for particulate
matter [EPA Report]. (EPA/600/R-08/139F). Research Triangle Park, NC: U.S. Environmental
Protection Agency, National Center for Environmental Assessment.
http://cfpub.epa.gov/ncea/cfm/recordisplav.cfm7de id=216546
January 2015 3-117 DRAFT: Do Not Cite or Quote
-------
Unal A: Hu. Y: Chang. ME: Talat Odman. M: Russell. AG. (2005). Airport related emissions and impacts
on air quality: Application to the Atlanta International Airport. Atmos Environ 39: 5787-5798.
http://dx.doi.0rg/10.1016/j.atmosenv.2005.05.051
Van Roosbroeck. S: Li. R: Hoek. G: Lebret. E: Brunekreef. B: Spiegelman. D. (2008). Traffic-related
outdoor air pollution and respiratory symptoms in children: The impact of adjustment for exposure
measurement error. Epidemiology 19: 409-416. http://dx.doi.org/10.1097/EDE.Ob013e3181673bab
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/j.atmosenv.2009.02.048
Vichj F: De Santis. F. (2012). The measurement of the sink properties of triethanolamine (TEA) as a
coating for collecting NO2 by using annular diffusion denuders. Environ Technol 33: 1065-1069.
http://dx.doi.org/10.1080/09593330.2011.606848
Vieira. SE: Stein. RT: Ferraro. AA: Pastro. LD: Pedro. SS: Lemos. M: da Suva. ER: Sly. PD: Saldiva. PH.
(2012). Urban air pollutants are significant risk factors for asthma and pneumonia in children: the
influence of location on the measurement of pollutants. Arch Bronconeumol 48: 389-395.
http://dx.doi.0rg/10.1016/j.arbres.2012.05.005
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 6: 40. http://dx.doi.org/10.1186/1476-069X-6-40
Vlachogianni. A: Kassomenos. P: Karppinen. A: Karakitsios. S: Kukkonen. J. (2011). Evaluation of a
multiple regression model for the forecasting of the concentrations of NOx and PM10 in Athens and
Helsinki. Sci Total Environ 409: 1559-1571. http://dx.doi.Org/10.1016/j.scitotenv.2010.12.040
Vrijheid. M: Martinez. D: Aguilera. I: Bustamante. M: Ballester. F: Estarlich. M: Fernandez-Somoano. A:
Guxens. M: Lertxundi N: Martinez. MD: Tardon. A: Sunyer. J. (2012). Indoor air pollution from gas
cooking and infant neurodevelopment. Epidemiology 23: 23-32.
http://dx.doi.org/10.1097/EDE.Ob013e31823a4023
Vutukuru. S: Dabdub. D. (2008). Modeling the effects of ship emissions on coastal air quality: A case
study of southern California. Atmos Environ 42: 3751-3764.
http://dx.doi.0rg/10.1016/j.atmosenv.2007.12.073
Wang. M: Beelen. R: Basagana. X: Becker. T: Cesaronj G: de Hoogh. K: Dedele. A: Declercq. C:
Dimakopoulou. K: Eeftens. M: Forastiere. F: Galassi. C: Grazuleviciene. R: Hoffmann. B: Heinrich. J:
lakovides. M: Kunzli. N: Korek. M: Lindlev. S: Molter. A: Mosler. G: Madsen. C: Nieuwenhuijsen.
MJ: Phuleria. H: Pedeli. X: Raaschou-Nielsen. O: Ranzi. A: Sugiri. D: Stempfelet M: Tsai. MY:
Lanki T: Udvardv. O: Varro. MJ: Wolf. K: Weinmayr. G: Yli-Tuomi. T: Hoek. G: Brunekreef. B.
(2013a). Evaluation of land use regression models for NO2 and particulate matter in 20 European
study areas: the ESCAPE project. Environ Sci Technol 47: 4357-4364.
http://dx.doi.org/10.1021/es305129t
Wang. M: Beelen. R: Bellander. T: Birk. M: Cesaroni. G: Cirach. M: Cyrys. J: de Hoogh. K: Declercq. C:
Dimakopoulou. K: Eeftens. M: Eriksen. KT: Forastiere. F: Galassi. C: Grivas. G: Heinrich. J:
Hoffmann. B: Ineichen. A: Korek. M: Lanki. T: Lindlev. S: Modig. L: Molter. A: Nafstad. P:
Nieuwenhuijsen. MJ: Nystad. W: Olsson. D: Raaschou-Nielsen. O: Ragettli M: Ranzi. A: Stempfelet.
M: Sugiri. D: Tsai. MY: Udvardv. O: Varro. MJ: Vienneau. D: Weinmayr. G: Wolf. K: Yli-Tuomj T:
Hoek. G: Brunekreef. B. (2014). Performance of multi-city land use regression models for nitrogen
dioxide and fine particles. Environ Health Perspect 122: 843-849.
http://dx.doi.org/10.1289/ehp.1307271
Wang. M: Beelen. R: Eeftens. M: Meliefste. K: Hoek. G: Brunekreef. B. (2012). Systematic evaluation of
land use regression models for NO. Environ Sci Technol 46: 4481-4489.
http://dx.doi.org/10.1021/es204183v
January 2015 3-118 DRAFT: Do Not Cite or Quote
-------
Wang. R: Henderson. SB: Sbihi. H: Allen. RW: Brauer. M. (2013b). Temporal stability of land use
regression models for traffic-related air pollution. Atmos Environ 64: 312-319.
http://dx.doi.0rg/10.1016/j.atmosenv.2012.09.056
Wheeler. AJ: Smith-Doiron. M: Xu. X: Gilbert. NL: Brook. JR. (2008). Intra-urban variability of air
pollution in Windsor, Ontario—measurement and modeling for human exposure assessment. Environ
Res 106: 7-16. http://dx.doi.Org/10.1016/j.envres.2007.09.004
Wichmann. J: Ketzel M: Ellermann. T: Loft. S. (2012). Apparent temperature and acute myocardial
infarction hospital admissions in Copenhagen, Denmark: A case-crossover study. Environ Health 11:
19. http://dx.doi.org/10.1186/1476-069X-ll-19
Williams. EJ: Lerner. BM: Murphy. PC: Herndon. SC: Zahniser. MS. (2009). Emissions of NOx, SO2,
CO, and HCHO from commercial marine shipping during Texas Air Quality Study (TexAQS) 2006. J
Geophys Res 114: D21306. http://dx.doi.org/10.1029/2009JD012094
Williams. R: Brook. R: Bard. R: Conner. T: Shin. H: Burnett. R. (2012a). Impact of personal and ambient-
level exposures to nitrogen dioxide and particulate matter on cardiovascular function. Int J Environ
Health Res 22: 71-91. http://dx.doi.org/10.1080/09603123.2011.588437
Williams. R: Jones. P: Croghan. C: Thornburg. J: Rodes. C. (2012b). The influence of human and
environmental exposure factors on personal NO(2) exposures. J Expo Sci Environ Epidemiol 22: 109-
115. http://dx.doi.org/10.1038/ies.2011.20
Williams. R: Rappold. A: Case. M: Schmitt M: Stone. S: Jones. P: Thornburg. J: Devlin. RB. (2012c).
Multi-pollutant exposures in an asthmatic cohort. Atmos Environ 61: 244-252.
http://dx.doi.0rg/10.1016/j.atmosenv.2012.07.049
Wilson. WE: Brauer. M. (2006). Estimation of ambient and non-ambient components of particulate matter
exposure from a personal monitoring panel study. J Expo Sci Environ Epidemiol 16: 264-274.
http://dx.doi.org/10.1038/si.ies.7500483
Wilson. WE: Mage. DT: Grant. LD. (2000). Estimating separately personal exposure to ambient and
nonambient particulate matter for epidemiology and risk assessment: Why and how. J Air Waste
ManagAssocSO: 1167-1183.
Wilton. D: Szpiro. A: Gould. T: Larson. T. (2010). Improving spatial concentration estimates for nitrogen
oxides using a hybrid meteorological dispersion/land use regression model in Los Angeles, CAand
Seattle, WA. Sci Total Environ 408: 1120-1130. http://dx.doi.Org/10.1016/j.scitotenv.2009.ll.033
Wu. J: Wilhelm. M: Chung. J: Ritz. B. (2011). Comparing exposure assessment methods for traffic-
related air pollution in an adverse pregnancy outcome study. Environ Res 111: 685-692.
http://dx.doi.0rg/10.1016/j.envres.2011.03.008
Wu. XM: Fan. ZT: Ohman-Strickland. P. (2010). Time-location patterns of a population living in an air
pollution hotspot. J Environ Public Health 2010: 625461. http://dx.doi.org/10.1155/2010/625461
Xue. J: McCurdv. T: Spengler. J: Ozkaynak. 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/sj.jea.7500319
Yuval: Bekhor. S: Brodav. DM. (2013). Data-driven nonlinear optimisation of a simple air pollution
dispersion model generating high resolution spatiotemporal exposure. Atmos Environ 79: 261-270.
http://dx.doi.0rg/10.1016/j.atmosenv.2013.06.005
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. http://dx.doi.org/10.1289/ehp.00108419
Zhang. J: Zhu. T: Kipen. H: Wang. G: Huang. W: Rich. D: Zhu. P: Wang. Y: Lu. SE: Ohman-Strickland.
P: Diehl S: Hu. M: Tong. J: Gong. J: Thomas. D: Committee. HHR. (2013). Cardiorespiratory
biomarker responses in healthy young adults to drastic air quality changes surrounding the 2008
Beijing Olympics. Res Rep Health Eff Inst Feb: 5-174.
January 2015 3-119 DRAFT: Do Not Cite or Quote
-------
Zhu. Y: Fung. DC: Kennedy. N: Hinds. WC: Eiguren-Fernandez. A. (2008). Measurements of ultrafine
particles and other vehicular pollutants inside a mobile exposure system on Los Angeles freeways. J
Air Waste ManagAssoc 58: 424-434.
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. http://dx.doi.org/10.1007/sl0651-
007-0028-x
January 2015 3-120 DRAFT: Do Not Cite or Quote
-------
CHAPTER 4 DOSIMETRY AND MODES OF
ACTION FOR INHALED OXIDES OF NITROGEN
4.1 Introduction
1 This chapter has two main purposes. The first is to describe the principles that underlie
2 the dosimetry of nitrogen dioxide (NCh) and nitric oxide (NO) and to discuss factors that
3 influence it. The second is to describe the modes of action that may lead to the health
4 effects that will be presented in Chapter 5 and Chapter 6. This chapter is not intended to
5 be a comprehensive overview, but rather, it updates the basic concepts derived from the
6 NC>2 and NO literature presented in the 1993 Air Quality Criteria for Oxides of Nitrogen
7 (AQCD) and the 2008 Integrated Science Assessment (ISA) for Oxides of
8 Nitrogen—Health Criteria (U.S. EPA. 2008) (U.S. EPA. 1993) and introduces the recent
9 relevant literature.
10 In Section 4.2. particular attention is given to chemical properties of inhaled NO2 and NO
11 that affect absorption, distribution, metabolism, and elimination. Inhaled NO2 and NO
12 and subsequent reaction products are discussed in relation to endogenous production of
13 these chemical species. Because few NO2 dosimetry studies have been published since
14 the 1993 AQCD (U.S. EPA. 1993). much of the information from that report has been
15 pulled forward into the current document and is discussed in the context of more recent
16 research. The topics of dosimetry and modes of action are bridged by reactions of NO2
17 with components of the epithelial lining fluid (ELF) and by reactions of NO with heme
18 proteins, processes that play roles in both uptake and biological responses.
19 Section 4.3 highlights findings of studies published since the 2008 ISA (U.S. EPA. 2008)
20 that provide insight into the biological pathways affected by exposure to NO2 and NO.
21 Earlier studies that represent the current state of the science are also discussed. Studies
22 conducted at more environmentally relevant concentrations of NO2 and NO (i.e.,
23 <5,000 ppb, see Section 1.1) are of greater interest because biological pathways
24 responsible for effects at higher concentrations may not be identical to those occurring at
25 lower concentrations. Some studies at higher concentrations are included if they were
26 early demonstrations of key biological pathways or if they are recent demonstrations of
27 potentially important new pathways. This information is used to develop a mode of action
28 framework for inhaled NO2 and NO, which serves as a guide to interpreting health effects
29 evidence presented in subsequent chapters (Chapters 5 and_6).
January 2015 4-1 DRAFT: Do Not Cite or Quote
-------
4.2 Dosimetry of Inhaled Oxides of Nitrogen
4.2.1 Introduction
1 This section provides a brief overview of NCh and NO dosimetry and updates
2 information provided in the 2008 ISA for Oxides of Nitrogen—Health Criteria
3 (U.S. EPA. 2008). Dosimetry refers to the measurement or estimation of the amount of a
4 compound, or its reaction products, absorbed and/or generated at specific sites in the
5 respiratory tract during an exposure. New to this ISA is the inclusion of basic information
6 regarding the endogenous production of NO2 and NO. It is important to consider inhaled
7 NO2 and NO and their subsequent reaction products in relation to endogenous production
8 of these chemical species. To establish an environmentally relevant context, ambient NO2
9 and NO concentrations are briefly discussed below; more detail is provided in Chapter 2.
10 Ambient NO2 concentrations are highest in the winter months, near major roadways,
11 during weekday morning hours, and decrease moderately during the afternoon (see
12 Atlanta, GA data in Figures 2-20 and 2-21). One-hour average, near-road (15 m) NO2
13 concentrations in Los Angeles, CA ranged from 3 to 80 ppb with median values of about
14 40 ppb in the winter and 30 ppb in the summer months of 2009 (Polidori and Fine. 2012).
15 Away from major roadways, 1-h avg NO2 concentrations may still reach 50 to 70 ppb
16 with median NO2 concentrations between roughly 10 to 30 ppb depending on the season
17 and distance from roadways (Polidori and Fine. 2012). As will be discussed, the uptake
18 of inhaled NO2 may potentially increase levels of NO2-derived reaction products beyond
19 levels endogenously occurring in the respiratory tract.
20 Similar to NO2, ambient NO concentrations are highest in the winter months near major
21 roadways during weekday morning hours, but decrease to very low levels during the
22 afternoon (see Atlanta, GA data in Figures 2-20 and 2-21). One-hour average, near-road
23 (15m) NO concentrations in Los Angeles, CA ranged from 0 ppb to over 400 ppb with
24 median values of about 50 ppb in the winter and 20 ppb in the summer months of 2009
25 (Polidori and Fine. 2012). Away from major roadways, 1-h avg NO concentrations may
26 still reach 250 ppb, but median NO concentrations are 5 ppb or less (Polidori and Fine.
27 2012). For the same roadway (Interstate 710), (Zhuetal.. 2008) reported on-road NOx
28 (i.e., the sum of NO and NO2) concentrations of around 400 ppb (average of eight 2-hour
29 samples collected between 10:00 a.m. to noon during the period from June 2006 to May
30 2007). As will be discussed, these ambient NO concentrations are generally in the range
31 of those occurring endogenously in the respiratory tract.
January 2015 4-2 DRAFT: Do Not Cite or Quote
-------
4.2.2 Dosimetry of Nitrogen Dioxide
1 NO2 is a highly reactive gas that occurs as a radical wherein, although technically a
2 resonance structure, the unpaired electron is more localized to the nitrogen atom than
3 either of the oxygen atoms. Once inhaled, NC>2 first encounters the aqueous phase of the
4 ELF, which is a contiguous but biologically complex aqueous fluid layer that covers all
5 of the respiratory tract surfaces (Bastacky et al.. 1995). The ELF constituent composition
6 shows appreciable heterogeneity with respect to anatomic site and species. The ELF of
7 alveolar surfaces and conducting airway surfaces has a monomolecular layer of surface
8 active lipids (Bernhard et al.. 2004; Hohlfeld. 2002; Mercer et al.. 1994). largely fully
9 saturated, which reduces surface tension and may provide a resistive barrier to the
10 interfacial transfer of NC>2 (see below). Upon dissolution into the ELF, NC>2 is converted
11 from a gas to a nonelectrolyte solute, and thus becomes subject to partitioning and
12 reaction/diffusion. Thus, the ELF represents the initial barrier between NCh contained
13 within the intra-respiratory tract gas phase and the underlying epithelia (Postlethwait and
14 Bidani. 1990). NCh chemically interacts with antioxidants, unsaturated lipids, and other
15 compounds in the ELF. It preferentially reacts with one electron donors (e.g., small
16 molecular weight antioxidants, protein thiols, etc.), undergoes radical-radical addition
17 reactions, may also abstract allylic hydrogen atoms from polyunsaturated fatty acids and,
18 through a complex series of reactions, can add to unsaturated fatty acids to generate
19 nitrolipids (Bonacci etal.. 2012: Rudolph etal.. 2010: O'Donnell et al.. 1999). The
20 compounds thought responsible, in large part, for the respiratory effects of inhaled NCh
21 are the reaction products themselves or the metabolites of these products in the ELF.
22 Quantifications of absolute NC>2 absorption reported in the 1993 AQCD and the 2008
23 ISA (U.S. EPA. 2008) (U.S. EPA. 1993) are briefly discussed below for thoroughness.
4.2.2.1 Mechanisms of Absorption of Nitrogen Dioxide
24 At the time of the 1993 AQCD (U.S. EPA. 1993). it was thought that inhaled NO2
25 probably reacted with the water molecules in the ELF to form nitrous acid (HNCh) and
26 nitric acid (HNOs). However, some limited data suggested that the absorption of NO2 was
27 linked to reactive substrates in the ELF and subsequent nitrite (N(V) production. By the
28 time of the 2008 ISA (U.S. EPA. 2008). chemical reactions between NO2 and ELF
29 constituents were more readily recognized as governing NC>2 absorption in the respiratory
30 tract.
January 2015 4-3 DRAFT: Do Not Cite or Quote
-------
4.2.2.1.1 Reaction with Epithelial Lining Fluid Water
1 Previous studies have demonstrated that it is not NO2 but instead the NO2 dimer,
2 dinitrogen tetroxide (^O/O, that reacts with water to yield NCV and nitrate [N(V
3 (Tinlavson-Pitts et al.. 2003: Schwartz and White. 1983: England and Corcoran. 1974)1.
4 However, in aqueous solutions, NC>2 rapidly reacts with many solutes (e.g., ascorbate and
5 urate), particularly those that are easily oxidized. Furthermore, at environmentally
6 relevant concentrations of NC>2 (e.g., around 100 ppb), the direct reactions of NC>2 with
7 dissolved substrates also become important because, at equilibrium, there is very little
8 N2O4 compared to NC>2. For example, using the delta Gibbs energies of formation of
9 gaseous NO2 and N2O4 (Chase. 1998). one can calculate that at equilibrium, when the
10 concentration of NO2 is 1,000 and 100 ppb, there are 1.48 x 105 and 1.48 x 106,
11 respectively, molecules of NO2 for each molecule of N2O.4. Thus, at environmental
12 exposure levels there are approximately 1.5 million NC>2 molecules for each N2O4
13 molecule. At these concentrations, it is far more likely for NC>2 (compared to ^64) to
14 penetrate into the aqueous milieu of the ELF. Ensuing reactions of NC>2 with dissolved
15 reactive substrates also become more likely than reaction with a second NO2 molecule (to
16 form N2O4). During reactive uptake by pure water, all reactions occur via N2O4 regardless
17 of the concentration of NC>2. However, in the presence of dissolved reactive substrates
18 and at low, environmentally relevant concentrations of NC>2, this process becomes
19 unlikely, and reactive uptake instead occurs via direct reactions of NO2. The latter
20 conditions resemble reactive uptake of NO2 by the ELF that would entail direct reactions
21 of NC>2 with, for example, dissolved small molecular weight antioxidants like glutathione
22 (GSH), ascorbate, or urate.
23 Enami et al. (2009) revisited the discussions regarding NC>2 reaction with water versus
24 ELF solutes. Because the authors postulate that NO2 effects are largely due to nitrate
25 formation and acidification via proton production, this issue warrants some discussion.
26 The claim by Enami et al. (2009) that "antioxidants catalyze the hydrolytic
27 decomposition of NC>2.. .but are not consumed in the process" is problematic in view of
28 the vast existing environmental health literature that regards NC>2 as an oxidant gas (Pryor
29 et al.. 2006: Augusto et al.. 2002: Ford et al.. 2002: Kirsch et al.. 2002: Wardman. 1998:
30 Postlethwait et al.. 1995: Huie. 1994: NetaetaL 1988: Finlayson-Pitts et al.. 1987:
31 Kikugawa and Kogi. 1987: Priitzetal.. 1985: Pryor and Lightsey. 1981). However,
32 Enami et al. (2009) measured nitrate without measuring nitrite and therefore their data do
33 not strongly support their contention, except to suggest that some hydrolysis of NC>2 may
34 be occurring because nitrate was detected. Nitrite data are important because any excess
35 nitrite found (reaction with water generally yields a 1:1 ratio of nitrite and nitrate; thus, a
36 yield of nitrite above 1 would be considered in excess) would indicate that it is the main
January 2015 4-4 DRAFT: Do Not Cite or Quote
-------
1 product formed as a result of one-electron oxidations by NC>2. Thus, by not measuring
2 nitrite, an important index to assess oxidation by NC>2 was missed.
3 Note that Enami et al. (2009) conducted their experiments in the absence of oxygen,
4 which makes their model of dubious applicability to the lung. At environmentally
5 relevant concentrations and physiologic temperatures, intra-pulmonary gas phase NC>2
6 will exist in its monomeric form. Furthermore, in the presence of aqueous-phase reactive
7 substrates, nitrite, but little or no nitrate, is formed during controlled in vitro exposures.
8 Thus, broad reactivity of NCh with a diversity of reactive substrates (solutes) within the
9 ELF facilitates chemical interactions with antioxidants, lipids, and
10 proteins/peptides/amino acids.
4.2.2.1.2 Governing Determinants of Nitrogen Dioxide Absorption within the
Respiratory Tract
11 The absorption of inhaled NCh into the ELF is governed by a process termed "reactive
12 absorption" that involves dissolution followed by chemical reaction with reactive
13 substrates in the ELF (Postlethwait and Bidani. 1990). as well as reactions within the
14 interfacial region. Due to the limited aqueous solubility of NC>2 and thus the rapid
15 saturation of the aqueous phase interfacial thin film (Bidani and Postlethwait. 1998). the
16 net flux of NC>2 into reactant-free water is constrained by the relatively slow direct
17 reaction of NC>2 with water (see above) compared with its radical reactions with
18 biological substrates (further discussion below). Thus, rapid reactions with ELF
19 substrates provide the net driving force for NC>2 mass transfer from the intra-pulmonary
20 gas phase into the ELF (Bidani and Postlethwait. 1998; Postlethwait and Bidani. 1994;
21 Postlethwait et al.. 1991a; Postlethwait and Bidani. 1990). Concentrations office" solute
22 NC>2 are likely negligible due to reaction-mediated removal. Empirical evidence suggests
23 that acute NC>2 uptake in the lower respiratory tract is rate governed by chemical
24 reactions of NC>2 with ELF constituents rather than solely by gas solubility in the ELF,
25 wherein the reaction between NC>2 and water does not significantly contribute to the
26 absorption of inhaled NC>2 (Postlethwait and Bidani. 1994. 1990). Absorption was also
27 observed to increase with increasing temperature, an indication of chemical reaction
28 rather than aqueous solubility, where solubility increases with temperature decrements
29 (Postlethwait and Bidani. 1990). Postlethwait et al. (1991b) proposed that inhaled NO2
30 (<10,000 ppb) did not penetrate the ELF to reach underlying sites and suggested that
31 cytotoxicity likely was initiated by products formed during NC>2 reactions with ELF
32 constituents. Subsequently, the reactive absorption of NC>2 was examined in a number of
33 studies that sought to identify the substrates that predominantly drive NC>2 reactive
34 absorption and to quantify the mass transfer kinetics of NC>2 in the respiratory tract.
35 Uptake was observed to be first-order with respect to NO2 at concentrations less than
January 2015 4-5 DRAFT: Do Not Cite or Quote
-------
1 10,000 ppb, to be aqueous substrate dependent, and to be saturable, meaning that the
2 absolute amount of NC>2 uptake would reach a maximum value even if reactive substrate
3 concentrations were in significant excess (Postlethwait et al.. 1991a. b).
4 The absorption of inhaled NCh is thought to be coupled with either radical-mediated
5 hydrogen abstraction to form HNCh (Postlethwait and Bidani. 1994. 1989) or electron
6 transfer from ELF anionic species that directly reduces NC>2 to nitrite (Adgent et al..
7 2012). Both mechanisms produce an organic radical from the initial ELF substrate. At
8 physiologic pH, any formed HNC>2 subsequently dissociates to hydrogen ion (H+) and
9 nitrite. The concentration of the resulting nitrite is likely insufficient to alter
10 physiological function because basal nitrite levels may not change appreciably in either
11 the respiratory tract or the circulation due to ambient NC>2 exposure. This is, in part,
12 because nitrite will diffuse into the underlying epithelial cells and vascular space where,
13 in the presence of red blood cells, it is oxidized to nitrate [(Postlethwait and Bidani. 1989;
14 Postlethwait and Mustafa. 1981) see also Section 4.2.2.41. Consequently, by default,
15 effects are probably attributable to the organic radical secondary oxidants formed
16 (Adgent etal. 2012; Velsoretal.. 2003; Velsor and Postlethwait 1997) and/or to the
17 proton load, although the ELF buffering capacity is anticipated to compensate for
18 environmentally relevant exposure-related proton generation.
19 Postlethwait et al. (1995) sought to determine the preferential absorption substrates for
20 NO2 in the ELF lavaged from male Sprague-Dawley rats. Because bronchoalveolar
21 lavage (BAL) fluid collected from rats may be diluted up to 100 times relative to the
22 native ELF (the dilution will be procedure specific), the effect of concentrating the BAL
23 fluid on NC>2 absorption was also investigated. A linear association was found between
24 the first-order rate constant for NCh absorption and the relative concentration of the BAL
25 fluid constituents. This suggested that concentration of the reactive substrates in the ELF
26 determines, in part, the rate of NC>2 absorption. The absorption due to specific ELF
27 constituents was also examined in chemically pure solutions. Albumin, reduced cysteine,
28 glutathione, ascorbate, and urate were the hydrophilic moieties found to be the most
29 active substrates for NC>2 absorption. Unsaturated fatty acids (such as oleic, linoleic, and
30 linolenic) were also identified as active absorption substrates and thought to account for
31 up to 20% of NO2 absorption. Vitamins A and E exhibited the greatest reactivity of the
32 substrates that were examined. However, the low concentrations of urate (the ELF of
33 rodents and some primates contains significantly less urate than the ELF of humans due
34 to differences in nitrogenous waste metabolism) and vitamins A and E were thought to
35 preclude them from being appreciable substrates in vivo. The authors concluded that
36 ascorbate and glutathione were the primary NC>2 absorption substrates in rat ELF.
37 Postlethwait et al. (1995) also found that the pulmonary surfactant component,
38 dipalmitoyl phosphatidylcholine (DPPC), was relatively unreactive towards NC>2, and
January 2015 4-6 DRAFT: Do Not Cite or Quote
-------
1 subsequent studies documented that compressed monomolecular interfacial films of
2 DPPC inhibit NC>2 absorption in vitro (Connor et al.. 2001). Documenting whether
3 surface active phospholipids (which comprise surfactant) inhibit NC>2 mass transfer
4 in vivo is extremely challenging because any in situ manipulations that disrupt the
5 surface tension-lowering actions of a surfactant lead to a plethora of pathophysiologic
6 sequelae. However, even though such potentially important influences on NC>2 mass
7 transfer have not been verified in vivo, modeling studies could estimate how such effects
8 would influence the intra-pulmonary distribution of inhaled NC>2, local mass transfer
9 rates, and thus dosimetry.
4.2.2.1.3 Reaction/Diffusion of Nitrogen Dioxide in the Epithelial Lining Fluid,
Potential for Penetration to Underlying Cells
10 Because the uptake of NO2 from inhaled air into the ELF is governed by reactive
11 absorption, it may be postulated that rapid ELF reactions prevent NCh from reaching
12 underlying respiratory tract tissues. To evaluate this supposition, consideration must be
13 given to the time required for NC>2 to diffuse through some thickness of the ELF versus
14 the rate of NC>2 reactions with substrates in that ELF.
15 The ELF varies in composition and thickness with distal progression into the lung. The
16 ELF of most of the tracheobronchial region may generally be described as consisting of
17 two layers: an upper mucus layer and a periciliary layer, which surrounds the cilia
18 (Button etal.. 2012: Widdicombe. 2002: Widdicombe and Widdicombe. 1995: Van As.
19 1977). The length of human cilia is about 7 (im in the trachea and bronchi and around
20 5 (im in the bronchioles (Song et al.. 2009: Clary-Meinesz et al.. 1997: Widdicombe and
21 Widdicombe. 1995). In the healthy lung, the thickness of the periciliary layer is roughly
22 the length of the cilia (Song et al.. 2009: Widdicombe and Widdicombe. 1995). This
23 periciliary layer forms a continuous liquid lining over the tracheobronchial airways;
24 whereas the upper mucus layer is discontinuous and diminishes or is absent in smaller
25 bronchioles (Widdicombe. 2002: Van As. 1977). The periciliary layer may be the only
26 ELF layer (i.e., there is little to no overlaying mucus) in the ciliated airways of infants
27 and healthy adults who are unaffected by disease, infection, etc. (Bhaskar et al.. 1985).
28 The ELF covering the alveolar surface is considerably thinner than the periciliary layer
29 found in the tracheobronchial region. The alveolar ELF consists of two layers: an upper
30 surfactant layer and a subphase fluid (Ng etal.. 2004). Bastacky et al. (1995) conducted a
31 low-temperature scanning electron microscopy analysis of rapidly frozen samples
32 (9 animals; 9,339 measurements) of rat lungs inflated to approximately 80% total lung
33 capacity. The alveolar ELF was found to be continuous, but of varied depth. Three
34 distinct ELF areas were described: (1) a thin layer [0.1 (im median depth, geometric
January 2015 4-7 DRAFT: Do Not Cite or Quote
-------
1 standard deviation (GSD) -2.16]: over relatively flat areas and comprising 80% of the
2 alveolar surface, (2) a slightly thinner layer (0.08 (im, GSD -1.79) over protruding
3 features and accounting for 10% of the surface, and (3) a thick layer (0.66 (im, GSD
4 -2.18) occurring at alveolar junctions and accounting for 10% of the surface. Based on
5 these distributions of thicknesses, 10% of the alveolar region is covered by an ELF layer
6 of 0.04 (im or less. Presuming that these depths would also occur in humans at 80% total
7 lung capacity and assuming isotropic expansion and contraction, depths should be
8 expected to be 20-40% greater during normal tidal breathing (rest and light exercise)
9 when the lung is inflated to between 50-60% total lung capacity averaged across the
10 respiratory cycle. During tidal breathing, a median ELF depth of 0.12-0.14 (im would be
11 expected over 80% of the alveolar surface with 10% of the alveolar surface having a
12 median depth of around 0.05 (im or less. Considering the entire distribution of depths
13 during tidal breathing, about 30, 60, and 90% of the alveolar surface would be estimated
14 to have alining layer thickness of less than or equal to 0.1, 0.2, and 0.5 (im, respectively.
15 The root mean square distance (d) that NC>2 can diffuse in some time (t) is given by the
16 Einstein-Smoluchowski equation:
d =
Equation 4-1
17 where D is the molecular diffusion coefficient of NC>2. A D value for NC>2 in water at
18 25°C of 1 .4 x 10~9 m2/sec has been reported and will be used in the calculations (Ford
19 et al.. 2002). In the lung, the D for NC>2 would be increased by temperature and decreased
20 by the higher viscosity of the ELF compared to water. The time available for diffusion
21 can be estimated based on the half-time for reactions between NO2 and reactive
22 substrates, assuming pseudo first-order kinetics apply. This half-time (T) has the form:
1 Although the authors stated that the distributions appeared to be log-normal, they did not report the (GSD) for the
three distinct areas they described. The GSD values were calculated from 25, 50, and 75th percentiles of the
distributions.
January 2015 4-8 DRAFT: Do Not Cite or Quote
-------
= ln(2)
T
Equation 4-2
..
1 where ^"i l l is the summation of the products of the second-order rate constants
2 and substrate concentrations (ci) for the primary reactive substances in the ELF.
3 Substituting x for t in Equation 4-1 yields:
2Dln(2)
vn;
Equation 4-3
4 and approximates the distance NO2 may diffuse before it chemically reacts with ELF
5 constituent molecules (e.g., antioxidants, proteins, lipids, etc.). A similar approach of
6 comparing the half-time in Equation 4-2 to the time for diffusion through the ELF or
7 other phase boundaries such as a membrane bilayer (see Equation 4-1 and solve for t)
8 was originally applied by Pryor (1992) and later by Ford et al. (2002).
9 In considering the classes of ELF biomolecules that react with NO2, one may focus on the
10 water-soluble small molecular weight antioxidants (e.g., ascorbate, urate, and
11 glutathione), which exist in the ELF in high concentrations and are very reactive toward
12 NC>2 and consequently have large kid terms. Lipids, on the other hand, would not be
13 expected to decrease considerably the transit time of NO2 because only those lipids
14 containing fatty acids with two or more double bonds have significant reactivity towards
15 NO2, and the lipids in the ELF are highly saturated.
16 The reaction rate constants of 3.5 x 107 M^sec"1, 2 x 107 M^sec"1, and 2 x 107 M^sec"1
17 were assumed for the small molecular weight antioxidants ascorbate, urate, and
18 glutathione, respectively (Ford et al.. 2002). These rates were determined in solution
19 using the pulse radiolysis fast kinetics technique. The kinetics of ascorbate and urate were
20 directly monitored, while in the case of glutathione, ABTS2 [2,2'-azino-bis
21 (3-ethylbenzothiazoline-6-sulfonic acid)] was used to produce the intense chromophore
22 ABTS* (note, here and elsewhere the superscript * designates a radical species) from its
23 reaction with the glutathiyl radical.
January 2015 4-9 DRAFT: Do Not Cite or Quote
-------
1 Species and anatomical loci must be considered when selecting appropriate
2 concentrations of reactive ELF biomolecules. Table 4-1 illustrates the small molecular
3 weight antioxidant composition differences between human and rat bronchoalveolar ELF
4 and the differences between human nasal and bronchoalveolar ELF (Squadrito etal..
5 2010; Van Per Vliet etal.. 1999). Predicted by Equation 4-3 and shown in Table 4-1.
6 NO2 is predicted to penetrate 0.2 to 0.6 um into the ELF and would not likely reach
7 airway tissues in the bronchi or bronchioles. Even extending the time for diffusion to 5i,
8 NC>2 would only be predicted to penetrate 0.5 to 1.3 um into the ELF, which does not
9 approach the 5 urn depth expected in the ciliated airways. However, minimal NCh
10 diffusion through the ELF in the bronchi and bronchioles does not preclude the potential
11 importance of reaction products reaching the underlying tissues in these regions.
Table 4-1 Small molecular weight antioxidantconcentrations in epithelial lining
fluid and predicted penetration distances for nitrogen dioxide.
Species — Site
Substrate Concentration, c.
Human — nasal
Human — bronchoalveolar
Rat — bronchoalveolar
Ascorbate
(MM)
28 ± 19
40 ± 18
1,004 ±325
Urate
225 ± 105
207 ± 167
81 ±27
Glutathione
<0.5
109 ±64
43 ± 15
n
Ł*ic,
(sec'1)
5.5 x 103
7.7 x 103
3.8 x 104
ELF Penetration
Distance
um
0.6
0.5
0.2
Rate constant, ki (M~1sec~1)
3.5 x 107
Substrate concentrations from Van Der Vliet et al.
constants from Ford et al. (2002).
2x 1Q7
2 x 107
(1999) for human and from Sauadrito et al.
(2010) for rat.
Reaction rate
12 In the alveolar region, the thickness of the ELF is sufficiently thin (<0.2 um over 60% of
13 the alveolar surface) for NO2 to diffuse through. There are some important differences
14 between the ELF of the alveolar region and the ELF of the tracheobronchial airways. In
15 studies modeling NC>2 and ozone (Os) uptake, a first-order rate constant has been
16 assumed for the alveolar ELF, which is 60-times slower than that of the tracheobronchial
17 ELF (Miller et al.. 1985; 1982). The slower reaction rate in the alveolar ELF would
18 increase the estimated potential diffusion distance to nearly 4 um, well beyond the depth
19 of the alveolar ELF. Additionally, the presence of DPPC , a principle component of
20 pulmonary surfactant, has been shown in vitro to reduce the uptake of NCh and Os by
21 inhibiting their ability to reach and react with the underlying subphase fluid containing
January 2015 4-10 DRAFT: Do Not Cite or Quote
-------
1 ascorbate, glutathione, and uric acid (Connor et al., 2004; Connor et al., 2001). The
2 physical properties of the interfacial saturated phospholipids may act to reduce the
3 diffusivity of NC>2. Both the DPPC and the overall slower reaction rate in the alveolar
4 ELF would increase diffusive resistance and increase the back diffusion of NO2 from the
5 surfactant into the gas phase. Nonetheless, the time for NO2 diffusion through a 0.2-(im
6 alveolar ELF is over two orders of magnitude faster than the NCh reaction rate half-time
7 in the alveolar ELF. Thus, of the inhaled NCh reaching the alveolar region and diffusing
8 into the ELF, an appreciable amount of NC>2 may reasonably be expected to diffuse
9 through the ELF to reach underlying tissues over much of the alveolar surface. Reaction
10 rates in these underlying tissues are expected to exceed those in the alveolar and
11 tracheobronchial ELF and would more rapidly consume NC>2 (Pryor. 1992; Miller et al..
12 1985).
4.2.2.2 Epithelial Lining Fluid Interactions with Nitrogen Dioxide
13 Small molecular weight antioxidants vary appreciably across species. For example, due
14 to the lack of urate oxidase, humans, primates, and select other species have increased
15 levels of urate. Conversely, rodent concentrations of urate are small compared to humans.
16 Such differences need to be recognized when considering preferential reactive absorption
17 substrates and the profile of products formed via reaction with NC>2. Glutathione and
18 ascorbate are the primary NCh-absorption substrates in rat ELF with near
19 1:1 stoichiometric yields of NC>2 uptake: nitrite formation, suggesting that one-electron
20 reduction of NCh is a predominant reaction pathway that also yields the corresponding
21 organic radical (Postlethwait et al.. 1995).
22 Beyond cell-specific differential susceptibility and the airway lumen concentration of
23 NC>2, site-specific injury was proposed to depend on rate of bioactive reaction product
24 formation relative to the extent of quenching (detoxification) of these products within the
25 ELF. Velsor and Postlethwait (1997) investigated the mechanisms of acute cellular injury
26 from NC>2 exposure. In an in vitro test system using red blood cells, the maximal levels of
27 membrane oxidation were observed at low antioxidant levels versus null (absent
28 antioxidants) or high antioxidant levels. Glutathione- and ascorbate-related membrane
29 oxidation was superoxide- and hydrogen peroxide-dependent, respectively. The authors
30 proposed that increased absorption of NC>2 occurred at the higher antioxidant
31 concentrations, but little secondary oxidation of the membrane occurred because the
32 reactive species (e.g., superoxide and hydrogen peroxide) generated during absorption
33 were quenched. A lower rate of NC>2 absorption occurred at the low antioxidant
34 concentrations, but oxidants were not quenched and so were available to interact with the
35 cell membrane. Further in vitro analyses also suggested that exposure-related responses
January 2015 4-11 DRAFT: Do Not Cite or Quote
-------
1 may not be strictly linear with respect to the inhaled NC>2 dose (concentration and/or
2 time) because the dependence of NO2 absorption and biologic target oxidation
3 demonstrated a bell-shaped function with respect to the initial antioxidant concentration
4 (Adgent et al.. 2012; Velsor et al.. 2003). Because the ELF varies throughout the
5 respiratory tract, the heterogeneous distribution of epithelial injury observed following
6 NO2 exposures may be explained, in part, by the ELF-dependent effects on local NCh
7 uptake and product formation. However, it should be noted that while these
8 dose-response relationships have been documented in vitro, in vivo validation has not yet
9 been accomplished due to the complexities in reproducibly modulating in situ ELF
10 compositions. Importantly, such in vitro results are difficult to directly extrapolate to the
11 in vivo situation, as precise rates of NO2 uptake, and thus product formation, are a
12 function of many factors including gas-phase NC>2 concentration, aqueous substrate
13 concentrations, surface area, gas flow, and pH of the ELF (Adgent et al., 2012; Bidani
14 and Postlethwait. 1998). However, an in vivo study of healthy male albino mice (5 weeks
15 old) suggested that a low dose of ascorbate (25 mg/kg) may exacerbate inflammatory
16 responses in terminal bronchial tissues following NC>2 exposure (20,000 ppb; 4 h/day,
17 10 days); whereas at a higher dose of ascorbate (100 mg/kg), NCh-exposed mice tissues
18 were similar to tissues from filter air-exposed controls (Zhang et al.. 201 Ob). These
19 in vivo responses seem parallel to those observed in vitro.
20 Antioxidant levels also vary spatially between lung regions and temporally with NC>2
21 exposure. While in vitro studies have clearly illustrated the role of antioxidants in
22 mediating NC>2 uptake and membrane oxidation, the temporal dynamics of biological
23 responses to NCh that occur in vivo are far more complex. Given the rapid reactions of
24 inhaled NC>2 with various biological substrates, the short half-life of some primary and
25 secondary reaction products as well as the continuous turnover of the ELF, specific
26 chemical species do not likely persist at any given anatomic locale for any appreciable
27 time. Kelly etal. (1996a) examined the effect of a 4-hour NCh (2,000 ppb) exposure on
28 antioxidant levels in bronchial lavage (BL) fluid and BAL fluid of 44 healthy
29 nonsmoking adults (19-45 years, median 24 years). The baseline concentrations of urate
30 and ascorbate were strongly correlated between the BL fluid and BAL fluid within
31 individuals (r = 0.88,/? < 0.001; r = 0.78,/? = 0.001; respectively); whereas the
32 concentrations of glutathione in the BL fluid and BAL fluid were not correlated. At
33 1.5 hours after the NC>2 exposure, urate and ascorbate were significantly reduced in both
34 lavage fractions, while glutathione levels were significantly increased but only in BL
35 fluid. By 6 hours post-exposure, ascorbate levels had returned to baseline in both lavage
36 fractions, but urate had become significantly increased in both lavage fractions and
37 glutathione levels remained elevated in BL fluid. By 24 hours post-exposure, all
38 antioxidant levels had returned to baseline. The levels of glutathione in BAL fluid did not
39 change from baseline at any time point in response to NC>2 exposure.
January 2015 4-12 DRAFT: Do Not Cite or Quote
-------
1 The depletion of urate and ascorbate, but not glutathione, has also been observed with
2 ex vivo exposure of human BAL fluid to NO2. Kelly et al. (1996b) collected BAL fluid
3 from male lung cancer patients (n = 16) and exposed the BAL fluid ex vivo at 37°C to
4 NO2 (50 to 2,000 ppb; 4 hours) or O3 (50 to 1,000 ppb; 4 hours). Kelly and Tetlev (1997)
5 also collected BAL fluid from lung cancer patients (n = 12; 54 ± 16 years) and exposed
6 the BAL fluid ex vivo to NO2 (50 to 1,000 ppb; 4 hours). Both studies found that NO2
7 depletes urate and ascorbate, but not glutathione, from BAL fluid. Kelly etal. (1996b)
8 noted a differential consumption of the antioxidants, with urate loss being greater than
9 that of ascorbate, which was lost at a much greater rate than glutathione. Kelly and Tetley
10 (1997) found that the rates of urate and ascorbate consumption were correlated with their
11 initial concentrations in the BAL fluid, such that higher initial antioxidant concentrations
12 were associated with a greater rate of antioxidant depletion. Illustrating the complex
13 interaction of antioxidants, these studies also suggest that glutathione oxidized by NO2
14 may be again reduced by urate and/or ascorbate.
15 Human and animal results stemming from samples obtained after exposure should be
16 viewed with appropriate caution. As detailed below, secondary reactions within the ELF,
17 sample handling and, importantly, the temporal sequence of exposure relative to sample
18 acquisition may all confound data interpretation. Because the ELF is a dynamic
19 compartment, samples obtained after exposure (>30 minutes) may not reflect biochemical
20 conditions that were present during exposure. This is a critical point, as while there is
21 some value in quantifying the net short-term effects on ELF composition due to exposure,
22 the biological consequences of exposure are largely a function of the ELF conditions
23 during exposure, which initiate a cascade of events leading to alterations in cell signaling,
24 cell injury, inflammation, and so forth. Thus, measurements of ELF components should
25 be interpreted in the context of ELF turnover time, clearance of "stable" reaction
26 products, and species generated/regenerated as a consequence of secondary redox
27 reactions. Reported measurements may reflect net effects on individual antioxidants but
28 lend limited insights into the initial reactions of NO2 within the ELF, and by extension,
29 into what bioactive products may be formed and how differences in ELF constituent
30 profiles govern biological outcomes. A clear example is evident in the work of Ford etal.
31 (2002). who characterized the reaction of the GSH radical (GS*) with urate (UH2 ) at a
32 pH (6.0) slightly below the recognized ELF pH (~6.8 to 7.0). NO2 more readily reacts
33 with glutathione than urate, producing GS* and NO2~. However, the subsequent reaction
34 GS* + UH2 -» GSH + UH* has a rate constant of ~3 x 107 JVT1 sec"1, which could
35 translate to an initial NO2 reaction with glutathione followed by reduction of the thiyl
36 radical by urate. This could result in an apparent, but potentially inaccurate, conclusion of
37 direct loss of urate during subsequent analyses. In addition, some reports have suggested
38 observations that include significant levels of the ascorbate oxidation product
39 dehydroascorbate (DHA). As with the example of secondary urate oxidation, such
January 2015 4-13 DRAFT: Do Not Cite or Quote
-------
1 observations need to be evaluated with caution as the half-life of DHA under biological
2 conditions is very short (minutes; the ascorbyl radical dismutation produces reduced
3 ascorbate and DHA; and DHA spontaneously decomposes to its keto acid). Furthermore,
4 because high redox couples are maintained in the ELF, and the ELF is constantly turning
5 over due to secretion and mucociliary clearance, it is unlikely that any appreciable
6 accumulation of DHA would occur. Therefore, care must be taken to avoid introducing
7 methodological artifacts (e.g., ascorbate oxidation during sample acquisition, handling,
8 and/or storage) that could significantly confound data interpretation. Consequently, an
9 understanding of the precise and preferential substrates is needed to discern the genesis of
10 species differences and the products formed that account for NO2 exposure-related
11 cellular perturbations.
12 Thus, variability in antioxidant concentrations and reactions among species may affect
13 NC>2 dose and health outcomes. Guinea pigs and mice have a lower basal activity of
14 glutathione transferase and glutathione peroxidase and lower a-tocopherol levels in the
15 lung compared to rats (Ichinose etal.. 1988; Sagai etal.. 1987). Human nasal lavage fluid
16 has a high proportion of urate and low levels of ascorbate; whereas mice, rats, or guinea
17 pigs have high levels of ascorbate and undetectable levels of urate. Glutathione is not
18 detected in the nasal lavage fluid of most of these species, except monkeys. Guinea pigs
19 and rats have a higher antioxidant to protein ratio in nasal lavage fluid and BAL fluid
20 than humans (Hatch. 1992). The BAL fluid profile differs from that of the nasal lavage
21 fluid. Humans have a higher proportion of glutathione and less ascorbate in their BAL
22 fluid compared to guinea pigs and rats (Slade etal.. 1993; Hatch. 1992). Rats have the
23 highest antioxidant-to-protein mass ratio in their BAL fluid (Slade etal.. 1993).
24 Antioxidant defenses also vary with age (Servais et al.. 2005) and exposure history (Duan
25 etal.. 1996). In the case of another reactive gas, Os, some studies have found that
26 differences in antioxidant levels among species and lung regions did not appear to be the
27 primary factor affecting Os-induced tissue injury (Duan etal.. 1996; 1993). However, a
28 close correlation between site-specific Os dose, the degree of epithelial injury, and
29 depletion of reduced glutathione was observed in monkeys (Plopper et al., 1998). For
30 both NO2 and Os, differences in reactive substrates among species and regions of the
31 respiratory tract are recognized, but the importance of these differences in relation to
32 tissue injury is not fully understood.
January 2015 4-14 DRAFT: Do Not Cite or Quote
-------
4.2.2.3 Regional and Total Respiratory Absorption of Nitrogen
Dioxide
1 Very limited work related to the quantification of NC>2 uptake has been published since
2 the 1993 AQCD (U.S. EPA. 1993) or the subsequent 2008 ISA (U.S. EPA. 2008).
3 Consequently, only an abbreviated discussion of this is included.
4.2.2.3.1 Experimental Studies of Nitrogen Dioxide Uptake
Upper Respiratory Tract Absorption
4 The nasal uptake of NO2 has been experimentally measured in dogs, rabbits, and rats
5 under conditions of unidirectional flow. Yokoyama (1968) reported 42.1 ± 14.9%
6 (mean ± SD) uptake of NCh in the isolated nasal passages of two dogs (3.5 L/min) and
7 three rabbits (0.75 L/min) exposed to 4,000 and 41,000 ppb NC>2. Uptake did not appear
8 to depend on the exposure concentration and was relatively constant over a 10- to
9 15-minute period. Cavanagh and Morris (1987) measured 28 and 25% uptake of NC>2
10 (40,400 ppb) in the noses of four naive and four previously exposed rats (0.10 L/min;
11 4-hours; 40,400 ppb), respectively, and uptake was constant over the 24-minute period
12 during which it was monitored.
13 Kleinman and Mautz (1991) measured the penetration of NCh through the upper airways
14 during inhalation in six tracheostomized dogs exposed to 1,000 or 5,000 ppb NC>2.
15 Uptake in the nasal passages was significantly greater at 1,000 ppb than at 5,000 ppb,
16 although the magnitude of this difference was not reported. The mean uptake of NO2
17 (1,000 ppb) in the nasal passages decreased from 80 to 70% as the ventilation rate
18 increased from about 3 to 7 L/min. During oral breathing, uptake was not dependent on
19 concentration. The mean oral uptake of NC>2 (1,000 and 5,000 ppb) decreased from 60 to
20 30% as the ventilation rate increased from 3 to 7 L/min. Although nasal uptake tended to
21 be greater than oral uptake, the difference was not statistically significant. The tendency
22 for greater nasal than oral uptake on NC>2 is consistent with that observed for Os as
23 described in Chapter 5 of the 2013 ISA for Ozone and Related Photochemical Oxidants
24 (U.S. EPA. 2013).
25 Overall, NO2 fractional absorption (uptake efficiency) in the upper respiratory tract is
26 greater in the nasal passage than in the oral passage and decreases with increasing
27 ventilation rates. As a result, a greater proportion of inhaled NO2 is delivered to the lower
28 respiratory tract at higher ventilation rates associated with exercise. In humans, exercise
29 causes a shift in the breathing pattern from nasal to oronasal relative to rest. Because the
30 nasal passages scrub gas-phase NO2 more efficiently than the mouth and uptake
January 2015 4-15 DRAFT: Do Not Cite or Quote
-------
1 efficiency decreases with increasing flow, exercise delivers a disproportionately greater
2 quantity of the inhaled mass to the lower respiratory tract, where the NC>2 is readily
3 absorbed.
4 Additionally, children tend to have a greater oral breathing contribution than adults at rest
5 and during exercise (Bennett et al.. 2008; Becquemin et al.. 1999). Chadha et al. (1987)
6 found that the majority (11 of 12) of patients with asthma or allergic rhinitis also breathe
7 oronasally at rest. Thus, compared to healthy adults, children and individuals with asthma
8 might be expected to have greater NO2 penetration into the lower respiratory tract.
9 Furthermore, normalized to body mass, median daily ventilation rates (m3/kg per day)
10 decrease over the course of life (Brochu et al., 2011). This decrease in ventilation relative
11 to body mass is rapid and nearly linear from infancy through early adulthood. Relative to
12 normal-weight adults (25-45 years of age), ventilation rates normalized to body mass are
13 increased 1.5-times in normal-weight children (7-10 years of age) and doubled in normal
14 weight infants (0.22-0.5 years of age). Relative to their body mass, children respire
15 greater amounts of air and associated pollutants than adults and have a greater portion of
16 respired pollutants reaching the lower respiratory tract than adults.
Lower Respiratory Tract Absorption
17 Postlethwait and Mustafa (1989) investigated the effect of exposure concentration and
18 breathing frequency on the uptake of NC>2 in isolated perfused rat lungs. To evaluate the
19 effect of exposure concentration, the lungs were exposed to NO2 (4,000 to 20,000 ppb)
20 while ventilated at 50 breaths/min with a tidal volume (Vr) of 2.0 mL. To examine the
21 effect of breathing frequency, the lungs were exposed to NO2 (5,000 ppb) while
22 ventilated at 30-90 breaths/min with a VT of 1.5 mL. All exposures were for 90 minutes.
23 The uptake of NC>2 ranged from 59 to 72% with an average of 65% and was not affected
24 by exposure concentration or breathing frequency. A combined regression analysis
25 showed a linear relationship between NC>2 dose to the lungs and total inhaled dose.
26 Illustrating variability in NC>2 uptake measurements, Postlethwait and Mustafa (1989)
27 observed 59% NC>2 uptake in lungs ventilated at 30 breaths/min with a VT of 1.5 mL;
28 whereas Postlethwait and Mustafa (1981) measured 35% NO2 uptake for the same
29 breathing condition. In another study, 73% uptake of NC>2 was reported for rat lungs
30 ventilated at 50 breaths/min with a VT of 2.3 mL (Postlethwait et al.. 1992). It should be
31 noted that typical breathing frequencies are around 80, 100, and 160 breaths/min for rats
32 during sleep, rest, and light exercise, respectively (de Winter-Sorkina and Cassee. 2002).
33 Hence, the breathing frequencies at which NC>2 uptake has been measured are lower than
34 for rats breathing normally. Furthermore, one must consider the potential impacts of the
35 methods used to measure NC>2 uptake (mass balance; wet chemical versus automated
36 analyzer which may or may not include a dilution component due to the sampling rate)
January 2015 4-16 DRAFT: Do Not Cite or Quote
-------
1 and the lack of perfusion of the bronchial circulation in isolated rat lungs (Postlethwait
2 etal.. 1990). In addition to measuring uptake in the upper respiratory tract, Kleinman and
3 Mautz (1991) also measured NO2 uptake in the lower respiratory tract of tracheostomized
4 dogs. In general, there was about 90% NO2 uptake in the lung that was independent of
5 ventilation rates from 3 to 16 L/min.
Total Respiratory Tract Absorption
6 Bauer et al. (1986) measured the uptake of NO2 (300 ppb) in 15 adults with asthma
7 exposed for 30 minutes (20 minutes at rest, then 10 minutes exercising on a bicycle
8 ergometer) via a mouthpiece during rest and exercise. There was a statistically significant
9 increase in uptake from 72% during rest to 87% during exercise. The minute ventilation
10 also increased from 8.1 L/min during rest to 30.4 L/min during exercise. Hence, exercise
11 increased the NO2 dose rate by 4.5-times in these subjects. In an earlier study by Wagner
12 (1970). seven healthy adults inhaled a NO2/NO mixture containing 290 to 7,200 ppb NO2
13 for brief (but unspecified) periods. The average NO2 uptake during 4,100 ppb and
14 7,200 ppb exposures was 82% during normal respiration (Vr, 0.4 L) and 92% during
15 maximal respiration (Vr, 2 to 4 L). Kleinman and Mautz (1991) also measured the total
16 respiratory tract uptake of NO2 (5,000 ppb) in nontracheostomized female beagle dogs
17 standing at rest or exercising on a treadmill. The dogs breathed through a small face
18 mask. Total respiratory tract uptake of NO2 was 78% during rest and increased to 94%
19 during exercise. This increase in uptake may, in large part, be due to the increase in VT
20 from 0.18 L during rest to 0.27 L during exercise. Coupled with an increase in minute
21 ventilation from 3.8 L/min during rest to 10.5 L/min during exercise, the dose rate of NO2
22 was 3.3-times greater for the dogs during exercise than rest.
4.2.2.3.2 Dosimetry Models of Nitrogen Dioxide Uptake
23 Few theoretical studies have investigated NO2 dosimetry. The original seminal dosimetry
24 model of Miller et al. (1982) were developed before much of the above information
25 regarding NO2 reaction/diffusion within the ELF had been obtained. In this model, there
26 was a strong distinction between uptake and dose. Uptake referred to the amount of NO2
27 being removed from gas phase per lung surface area (ug/cm2); whereas dose referred to
28 the amount of NO2 per lung surface area (ug/cm2) that diffused through the ELF and
29 reached the underlying tissues.
30 Miller et al. (1982). and subsequently Overton (1984). did not attempt to predict the
31 amount of reactants in the ELF or the transport of reaction products to the tissues. They
32 assumed that reactions of NO2 with constituents in the ELF were protective in that these
33 reactions reduced the flux of NO2 to the tissues. Others have postulated that NO2 reaction
January 2015 4-17 DRAFT: Do Not Cite or Quote
-------
1 products formed in the ELF, rather than NC>2 itself, could mediate responses (Velsor and
2 Postlethwait. 1997; Postlethwait and Bidani. 1994; Overton. 1984). Overall, these
3 modeling studies predict that the net NO2 uptake (NO2 flux to air-liquid interface) is
4 relatively constant from the trachea to the terminal bronchioles and then rapidly decreases
5 in the pulmonary region. The pattern of net NO2 uptake rate is expected to be similar
6 among species and unaffected by age in humans. However, the NO2 uptake per unit
7 surface area may be several times higher in infants compared to adults, because children
8 under age 5 have a much smaller surface area in the extrathoracic (nasal) and alveolar
9 regions (Sarangapani et al.. 2003).
10 The predicted tissue dose and dose rate of NO2 (NO2 flux to liquid-tissue interface) are
11 low in the trachea, increase to a maximum in the terminal bronchioles and the first
12 generation of the pulmonary region, and then decrease rapidly with distal progression.
13 The site of maximal NO2 tissue dose is predicted to be fairly similar among species,
14 ranging from the first generation of respiratory bronchioles in humans to the alveolar
15 ducts in rats. However, estimates of NO2 penetration in Table 4-1 showed that NO2 is not
16 expected to go deeper than 0.2 to 0.6 urn into the ELF of the ciliated airways before
17 reacting with substrates. The production of toxic NO2 reaction products in the ELF and
18 the movement of the reaction products to the tissues have not been modeled.
19 Contrary to what in vitro studies have shown (Velsor and Postlethwait. 1997). modeling
20 studies have generally considered NO2 reactions in the ELF to be protective. The
21 complex interactions among antioxidants, spatial differences in antioxidants across
22 respiratory tract regions, temporal changes in ELF constituent levels in response to NO2
23 exposure, and species differences in antioxidant defenses need to be considered in the
24 next generation of dosimetric models. Current NO2 dosimetry models are inadequate to
25 put response data collected from animals and humans on a comparative footing with each
26 other and with exposure conditions in epidemiologic studies. Total dose or liquid dose of
27 NO2 could be used as a first approximation for inter-species dosimetric comparisons
28 using currently available NO2 models.
29 As stated above, the total dose or uptake (|ig per cm2 surface area) of NO2 is predicted to
30 be relatively constant across the tracheobronchial airways with a rapid decrease in dose
31 with progression into the gas exchange region (Miller et al.. 1982). The model used by
32 Miller et al. (1982) for NO2 was generally the same as that subsequently used by Miller
33 etal. (1988) for O3. Miller etal. (1988) predicted that the total dose of O3 is relatively
34 similar among several mammalian species (namely, the rabbit, guinea pig, rat, and
35 human). The total dose of NO2 would also be expected to be relatively similar among
36 these mammalian species. Although it may not be strictly appropriate to apply identical
37 reaction rates for each of these species, varying the reaction rate from zero to that of O3
January 2015 4-18 DRAFT: Do Not Cite or Quote
-------
1 increased the predicted total dose of NO2 by less than 5 times in the trachea and bronchi.
2 This is small relative to the 400-time decrease in total dose from the first generation of
3 respiratory bronchioles to the alveolar sacs (Miller et al.. 1982).
4 Asgharian et al. (2011) recently developed a model for soluble and reactive gas uptake
5 that applied many of the basic concepts described by Miller et al. (1985). Unlike Miller
6 et al. (1985). who separately considered liquid and tissue layers, Asgharian et al. (2011)
7 lumped the liquid layer lining the airways and the tissue layer together with the same
8 diffusion and reaction rates. The model predicted that formaldehyde could penetrate to a
9 maximum of 200 (im tissue in the trachea during inhalation before being removed by
10 reactions. Because predictions were for a single breath, it is possible that deeper tissues
11 may be reached during continuous breathing. Applying the model to experimental Os
12 data, Asgharian et al. (2011) estimated a first-order reaction rate of 105 sec"1 (i.e.,
13 half-time of only 7 (isec). By comparison, the rate of 0.018 sec"1 (i.e., half-time of 39 sec)
14 was used for formaldehyde. Lumping the liquid and tissue layers may be appropriate for
15 the relatively slow-reacting formaldehyde, but it is perhaps less so for Os and NC>2, which
16 are expected to be removed by reactions within the liquid layer of ciliated airways (see
17 Table 4-1). For the rapidly reacting gases Os and NC>2, a distinction between liquid and
18 tissue compartments may be mechanistically important to discern whether the gas itself
19 or its reaction products are associated with health outcomes.
20 Existing dosimetric models can predict the total dose per surface area of distinct areas of
21 the lungs (e.g., individual generations of the tracheobronchial airways and alveolar
22 region). This total dose appears to be very similar among several mammalian species.
23 Similarly, the site of maximal NC>2 tissue dose, near the beginning of the gas exchange
24 region, is also predicted to be fairly similar among species. However, differences in
25 potential NC>2 reactive substrates and reaction products among species have not been
26 considered in modeling efforts. Thus, despite the predicted similarities in total NC>2 dose
27 and site of maximal tissue dose, there is uncertainty related to inter-species differences in
28 concentrations of reactive substrates and reaction products formed within the ELF and
29 tissues. The importance of specific reaction products in mediating health effects in
30 different species is similarly unclear. With regard to humans, individuals with asthma are
31 more likely to experience health effects from ambient NO2 exposures than healthy
32 individuals (Section 7.3.1). Specific aspects of asthma pathology that may affect NO2
33 uptake and disposition and that may be included in dosimetric models have not been
34 identified. Furthermore, most models have focused on the lungs and have not considered
35 the inter-species differences in the dose to nasal passages nor the potential importance of
36 neural or other pathways in affecting health outcomes. Although total dose in the
37 tracheobronchial airways and tissue dose in the alveolar region can be predicted,
January 2015 4-19 DRAFT: Do Not Cite or Quote
-------
1 modeling efforts do not sufficiently link these endpoints to subsequent downstream
2 events.
4.2.2.4 Endogenous Generation, Metabolism, Distribution, and
Elimination of Nitrogen Dioxide
3 Along with carbon monoxide (CO), NO2 is a criteria pollutant believed to be produced
4 endogenously in the lung. Evidence in support of a claim for endogenously produced Os
5 [e.g., Babior et al. (2003) has received serious criticism (Pryor et al., 2006; Kettle etal..
6 2004; Sies. 2004; Smith. 2004) and is here considered controversial. A useful discussion
7 of the issues can be found in Drahl (2009).
8 This endogenous production and function may have important implications for the
9 interpretation of health effects studies. NO2 may be produced endogenously by various
10 processes, including the acidification of nitrite
11 (2 H+ + 2 NO2~ -» 2 HNO2 -» H2O + N2O3 -» NO + NO2 + H2O) (as can transpire in
12 phagolysosomes), the decomposition of peroxynitrite and/or the nitrosoperoxylcarbonate
13 anion (ONOCT + CO2 -» ONOOCO2~ -» CO3*~ + NO2), and the action of peroxidases
14 when using nitrite and H2O2 as substrates. Nitrated proteins form when tyrosine residues
15 are first oxidized to a tyrosyl radical intermediately followed by radical-radical addition
16 of NO2 to produce 3-nitrotyrosine. NO2 is the terminal nitrating agent, and the presence
17 of nitrated proteins provides solid evidence for the endogenous production of NO2 per se.
18 Endogenous NO2 is expected to increase with dietary consumption of nitrite and nitrate
19 (which occurs in substantial concentrations in some leafy vegetables, e.g., spinach) as
20 well as during immune responses and inflammation. There is no known antioxidant
21 enzymatic process for the decomposition of NO2, but this is probably due to the
22 spontaneous reactions that NO2 undergoes with small molecular weight antioxidants,
23 such as glutathione and ascorbate, which result in formation of nitrite and antioxidant
24 radicals. These reactions are so fast that they only allow NO2 to diffuse small distances in
25 the submicrometer range before reacting (see above. Ford et al.. 2002). NO2 is slightly
26 hydrophobic (Squadrito and Postlethwait 2009) and faces no significant physical barriers
27 to prevent it from readily traversing biological membranes. But in light of its high
28 reactivity, NO2 is unlikely to become systemically distributed, and therefore, its
29 endogenous steady-state levels in distant tissues are unlikely to be affected, for example,
30 by inhaled NO2.
31 With regard to the lung, understanding the balance between endogenous products and
32 those derived from inhaled ambient NO2 is a complex and challenging issue. Because
33 inhaled NO2 predominantly undergoes univalent reduction to nitrite during reactive
January 2015 4-20 DRAFT: Do Not Cite or Quote
-------
1 absorption, changes in nitrite concentrations can be used as a surrogate for initial
2 considerations of how inhaled NC>2 compares with that produced endogenously. As an
3 example, rat lung ELF contains low (iM to nM levels of nitrite, with nitrate being
4 substantially more prevalent. Due to salivary and gut microflora nitrate reductase activity
5 and to reactions of nitrite, especially with heme proteins which yield nitrate, there is a
6 constant cyclic flux between nitrite and nitrate, with nitrate being the primary excretion
7 product in urine. In a rat with numerous simplifying conditions, assuming a gas phase
8 concentration of 200 ppb NC>2, a minute ventilation of 150 mL/min, an exposure time of
9 4 hours, quantitative conversion of NC>2 to nitrite, 100% uptake efficiency, an ELF
10 volume of 150 (iL, and no ELF clearance [even though nitrite has been shown to diffuse
11 out of the ELF quickly (Postlethwait and Bidani. 1989)1. there would be a net
12 accumulation of approximately 0.3 (imol of nitrite. If the NCh-derived nitrite were evenly
13 distributed throughout the ELF pool, this would equate to an additional 2 mM
14 concentration of nitrite. However, in vitro studies using isolated lungs have not reported
15 increases of this magnitude consequent to 10,000-20,000 ppb NC>2 exposures, well above
16 ambient concentrations, demonstrating that the ELF is a dynamic compartment and that
17 small molecular weight reaction products (though charged) move readily from the
18 respiratory tract surface to the vascular space.
19 Both nitrite and nitrate levels are very diet dependent, and diet represents the primary
20 source for both ions. Although environmental exposures at current ambient NC>2
21 concentrations would likely have a minimal effect on the overall balance of nitrite and
22 nitrate outside the respiratory tract, how inhaled NCh compares with endogenous
23 production rates or amounts within the respiratory tract remains essentially unknown.
24 However, the uptake of inhaled NCh may potentially increase levels of nitrite and/or
25 other reaction products beyond levels endogenously occurring in the respiratory tract.
4.2.2.5 Metabolism, Distribution, and Elimination of Products
Derived from Inhaled Nitrogen Dioxide
26 As stated earlier, NC>2 absorption may generate some HNCh, which subsequently
27 dissociates to H+ and nitrite. Nitrite enters the underlying epithelial cells and
28 subsequently the blood. In the presence of red blood cells and/or heme proteins, nitrite is
29 oxidized to nitrate (Postlethwait and Mustafa. 1981). Nitrate is the primary stable oxide
30 of nitrogen product and it is subsequently excreted in the urine. There has been concern
31 that inhaled NC>2 may lead to the production of N-nitrosamines, many of which are
32 carcinogenic, because NC>2 can produce nitrite and nitrate (in blood). Nitrate can be
33 converted to nitrite by bacterial reduction in saliva, the gastrointestinal tract, and the
34 urinary bladder. Nitrite has been found to react with secondary amines to form
January 2015 4-21 DRAFT: Do Not Cite or Quote
-------
1 N-nitrosamines. This remains speculative because nitrosamines are not detected in tissues
2 of animals exposed by inhalation to NC>2 unless precursors to nitrosamines and/or
3 inhibitors of nitrosamine metabolism are coadministered. Rubenchik et al. (1995) could
4 not detect N-nitrosodimethylamine (NDMA) in tissues of mice exposed to 4,000 to
5 4,500 ppb NC>2 for 1 hour. However, NDMA was found in tissues when mice were
6 simultaneously given oral doses of amidopyrine and 4-methylpyrazole, an inhibitor of
7 NDMA metabolism. Nevertheless, endogenous NO2 production and the cyclic
8 inter-conversion of nitrite and nitrate may provide the precursors that drive nitrosamine
9 formation. However, because ambient NO2 contributes only modest amounts of
10 nitrite/nitrate relative to dietary intake, any substantial contribution to systemic
11 nitrosamine formation is not likely. Thus, the relative importance of inhaled NC>2 in
12 endogenous N-nitrosamine formation has yet to be demonstrated. Metabolism of inhaled
13 NC>2 may also, in some cases, transform other chemicals potentially present in the body
14 into mutagens and carcinogens. Van Stee etal. (1983) reported N-nitrosomorpholine
15 (NMOR) production in mice gavaged with 1 g of morpholine/kg body weight per day and
16 then exposed (5-6 hours daily for 5 days) to 16,500-20,500 ppb NO2. NMOR is a
17 nitrosamine, which is a potent animal carcinogen. The single site containing the greatest
18 amount of NMOR was the gastrointestinal tract, as would be expected due to the pH-
19 dependent facilitation of N-nitrosation chemistry. Later, Van Stee et al. (1995) exposed
20 mice to approximately 20,000 ppb 15NO2 and 1 g/kg morpholine simultaneously.
21 N-nitrosomorpholine was found in the body of the exposed mice. Of the NMOR in the
22 body, 98.4% was labeled with 15N which was derived from the inhaled 15NO2, and 1.6%
23 was derived presumably from endogenous sources. Inhaled NO2 may also be involved in
24 the production of mutagenic (and carcinogenic) nitroderivatives of other coexposed
25 compounds, such as polycyclic aromatic hydrocarbon(s) (PAHs), via nitration reactions.
26 Mivanishi et al. (1996) coexposed rats, mice, guinea pigs, and hamsters to 20,000 ppb
27 NO2 and various PAHs (pyrene, fluoranthene, fluorene, anthracene, or chrysene). Nitro
28 derivatives of these PAHs, which were found to be highly mutagenic in the Ames/X
29 typhimurium assay, were excreted in the urine of these animals. Specifically, the nitrated
30 metabolites of pyrene (l-nitro-6/8-hydroxypyrene and l-nitro-3-hydroxypyrene) was
31 detected in the urine. Further studies indicated that these metabolites are nitrated by an
32 ionic reaction in vivo after the hydroxylation of pyrene in the liver.
4.2.3 Dosimetry of Nitric Oxide
33 NO occurs within the respiratory tract gas phase due to the following: (1) inhalation of
34 ambient NO and (2) off-gassing from its endogenous production within pulmonary
35 tissues, airspace surface inflammatory cells, and blood. The net uptake of NO within the
January 2015 4-22 DRAFT: Do Not Cite or Quote
-------
1 gas exchange regions depends on the balance between the intra-pulmonary gas phase
2 concentration (discussed below) and the inhaled ambient concentration.
3 While NO exists as a radical species, it is much less reactive than many other radical
4 species. However, it selectively participates in radical-radical reactions such as with
5 superoxide radical anions [62*~, which produces peroxynitrite (ONOO")], thiyl radicals
6 [e-g-, cysteine (Cys*), glutathione (GS*), which produce S-nitrosothiols (RSNO)], and
7 organic peroxyl radicals (Madej et al.. 2008; Goldstein et al.. 2004). In addition, NO
8 reacts with heme-containing proteins such as hemoglobin (Pacher et al.. 2007). Although
9 the radical-based reactions generally occur at near diffusion-controlled rates, the
10 prevalence of non-NO radical species at any given time is low. Thus, in terms of the
11 overall uptake and tissue diffusion of NO within the lung, interception due to reactions is
12 not expected to consume appreciable amounts of the total NO involved in mass transfer
13 from the alveolar to the vascular space. Inhaled NO uptake occurs against the background
14 of endogenous NO production, which is derived primarily from the catalytic activities of
15 the several isoforms of nitric oxide synthase [NOS (Forstermann and Sessa. 2012)].
16 Estimates of nitrite and/or nitrate stemming from NO production via NOS suggest that
17 endogenous NO production, even during inflammatory states, is at best modest compared
18 to dietary intake, although, under specific conditions, plasma levels have been shown to
19 transiently increase due to nondietary, endogenous biological activities. Additional
20 endogenously generated NO may also occur from the acidification of nitrite in the
21 presence of electron donors, such as within phagolysosomes, by dissociation of RSNO
22 and by complex interactions within red blood cells that likely lead to the release of NO
23 (Weitzberg et al.. 2010). In combination, these processes result in the appearance of NO
24 within the intra-pulmonary gas phase, which can be measured in expired breath and is
25 routinely labeled as either "eNO" or expressed as the fractional amount of expired gas
26 "FeNO."
27 Reported eNO concentrations from the lower respiratory tract span a broad range (~5 to
28 >300 ppb), with nasal/sinus concentrations generally accepted as being greater than what
29 is measured from the lower respiratory tract [e.g., See and Christiani (2013).
30 Alexanderson et al. (2012). Gelb et al. (2012). Nodaetal. (2012). Taylor (2012). Bautista
31 etal. (2011). Linhares et al. (2011). and Olinetal. (1998)1. eNO has been reported to be
32 affected by a variety of factors including disease state, diet, sex (or height), species,
33 smoking history, environmental exposures, and so forth. Although eNO from the lower
34 respiratory tract is increased by asthma, this is not the case for nasal NO (ATS/ERS.
35 2005).
36 For the general U.S. population, results of the 2007-2011 National Health and Nutrition
37 Examination Survey show a geometric mean eNO of 9.7 ppb in children (n = 1,855;
January 2015 4-23 DRAFT: Do Not Cite or Quote
-------
1 6-11 years of age; 10% with current asthma) and 13.3 ppb in teenagers and adults
2 [n = 11,420; 12-80 years of age; 8% with current asthma (See and Christiani. 2013)1. In
3 healthy, never-smokers [558 males (M), 573 females (F); 25-75 years of age], Olin et al.
4 (2007) reported a geometric mean eNO of 16.6 ppb (95% reference interval, 6 to 47 ppb).
5 The eNO levels increased with age and height of the individuals, but did not depend on
6 sex. In healthy children (23 M, 28 F, 1-5 years of age), a geometric mean eNO of 7 ppb
7 (95% CI: 3,12) has been reported (van der Heijden et al.. 2014). The eNO levels in these
8 children were unrelated to age, height, weight, or sex. These eNO levels correspond to
9 NO output rates of about 40-50 nL/min from the lower respiratory tract of healthy adults
10 and about 20-30 nL/min for healthy children.
11 Kharitonov et al. (2005) reported nasal NO concentrations of 750 ppb (95% CI: 700, 810)
12 in children [n = 20; 10 ± 3 (SD) years] and 900 ppb (95% CI: 870, 930) in adults (n = 29;
13 38 ± 11 years). Another study of healthy adults (n = 10; 18-35 years of age) found a
14 nasal NO concentration of 670 ppb. Higher NO concentrations (9,100 ± 3,800 ppb; n = 5)
15 have been reported for the paranasal sinuses of healthy adults (Lundberg et al.. 1995).
16 Asthma and current rhinitis do not appear to affect nasal NO concentrations
17 (Alexanderson et al.. 2012; Kharitonov et al.. 2005). Nasal NO is reduced by exercise
18 (ATS/ERS. 2005). The nasal NO concentrations described above correspond to NO
19 output rates of about 300 nL/min for the nasal airways of adults with or without asthma
20 and 230 nL/min for children with or without asthma. Nasal NO output rates of healthy
21 primates are in the range of 200 to 450 nL/min (ATS/ERS. 2005). With a NO output of
22 730 nL/min, a large contribution to nasal NO appears to derive from the paranasal
23 sinuses. Based on these NO output rates, the nasal passages may contribute, on average,
24 roughly 15-20 ppb NO to the lower respiratory tract during rest.
25 The other primary approach to noninvasive assessment of the respiratory tract surface is
26 expired breath condensate (EEC), which captures aerosolized materials contained in
27 exhaled air, including those directly related to reactive nitrogen chemistry (e.g., nitrite,
28 nitrate, 3-nitrotyrosine). Unfortunately, this relatively new field of analyzing exhaled
29 constituents has encountered numerous situations where concentrations of eNO and EEC
30 constituents are unrelated (Ravaet al., 2012; Dressel et al.. 2010; Malinovschi et al..
31 2009: Cardinale et al.. 2007: Vints et al.. 2005: Chambers and Avres. 2001: Olin et al..
32 2001; Zetterquist et al.. 1999; Olinetal.. 1998; Jilmaetal.. 1996). Given the endogenous
33 production of NO and the lack of a correlation between the two measurements, neither
34 eNO nor EEC can be employed as a metric of exposure history with any significant
35 degree of specificity for inhaled ambient NO.
36 The absorption of inhaled NO proceeds similarly to oxygen and carbon monoxide. In a
37 study of seven healthy adults, Wagner (1970) observed an average NO (5,000 ppb)
January 2015 4-24 DRAFT: Do Not Cite or Quote
-------
1 uptake of 88% during normal respiration (Vr, 0.4 L) and 92% during maximal respiration
2 (Vi, 2 to 4 L). Because blood acts as a near "infinite" sink for NO, it has been proposed
3 as an alternative to CO for measuring pulmonary diffusing capacity [e.g., Chakraborty
4 et al. (2004) and Heller etal. (2004)]. NO absorption follows Henry's law for dissolution
5 into the aqueous phase, followed by diffusion into the vascular space where it interacts
6 with red blood cell hemoglobin to ultimately form nitrate. Thus, due to its chemical
7 conversion, NO net flux from alveolar gas phase to the blood occurs when the alveolar
8 concentration exceeds that found in tissue and/or blood. Mass transfer resistances may be
9 encountered (Borland et al.. 2010; Chakraborty et al.. 2004). but their combined effects
10 are likely small due to the low (ppb) concentrations of NO.
11 The formation of RSNO within the ELF may contribute to the overall epithelial cell
12 uptake via an L-type amino acid transporter [LAT (Toroket al.. 2012; Brahmajothi et al..
13 2010)]. An in vitro study by Brahmajothi et al. (2010) showed that pre-incubation of
14 cultured alveolar epithelial cells with L-cysteine increased intracellular RSNO
15 concentrations by 3-times compared with diffusive transport. This increase involved
16 transport via the LAT. LAT transport was further augmented by addition of glutathione
17 and was independent of sodium transport. The authors concluded that NO gas uptake by
18 alveolar epithelium occurred predominantly by forming extracellular
19 S-nitroso-L-cysteine, which was then transported by LAT rather than by diffusion.
20 Subsequently, Torok etal. (2012) also showed that LAT transport exceeded diffusive
21 transport in isolated mice lungs. However, the precise extent of contribution of LAT
22 transport remains unclear because formation of RSNO requires several steps due to the
23 slow direct reactivity of NO with reduced thiols. In vivo, the time for these reactions may
24 exceed the time for diffusion into and through alveolar epithelial cells. Furthermore,
25 because blood acts as a sink for NO (i.e., a near zero boundary condition), lower
26 intracellular concentrations of NO would occur in vivo compared to the nonzero
27 boundary conditions in cell cultures and isolated lungs (Asgharian et al.. 2011). While
28 diffusive transport of NO is known and relatively well characterized, the importance of
29 LAT transport in vivo has not been determined.
30 Ambient NO levels are likely similar to those endogenously occurring within the lung
31 airspaces, except during morning commutes or near major roadways where they may
32 possibly exceed endogenous levels. It is not known whether periods of high ambient NO
33 exposure could alter endogenous NO production within the respiratory tract or pathways
34 affected by endogenous NO. Importantly, it should be noted that in the clinical setting,
35 therapeutic administration is a very different situation wherein > 10,000 ppb NO may be
36 administered continuously for prolonged periods.
January 2015 4-25 DRAFT: Do Not Cite or Quote
-------
4.2.4 Summary of Dosimetry
1 The uptake of inhaled NO2 in the respiratory tract is governed by "reactive absorption,"
2 which involves chemical reactions with antioxidants, unsaturated lipids, and other
3 compounds in the ELF. In vitro studies have clearly illustrated the role of antioxidants in
4 mediating NC>2 uptake. The rapid reactions of NC>2 with tracheobronchial ELF substrates
5 provides a net driving force for NCh mass transfer from the gas phase into the ELF.
6 Concentrations of "free" solute NC>2 are likely negligible due to its reaction-mediated
7 removal. Thus, it is not NC>2 itself, but rather its reaction products that are believed to
8 interact with the apical surfaces of the tracheobronchial epithelial. At high substrate
9 concentrations, oxidants/cytotoxic products are at least partially quenched due to
10 secondary antioxidant reactions. At low substrate concentration, ELF-derived
11 oxidants/cytotoxic products have a lower probability of being intercepted by unreacted
12 antioxidants and instead may reach underlying targets.
13 Within the alveolar region, much of the inhaled NC>2 entering the ELF will diffuse
14 through rapidly enough to avoid reactions and will reach underlying tissue surfaces. A
15 principle component of pulmonary surfactant, DPPC, may partially reduce the uptake of
16 NC>2 by slowing its diffusion and decreasing reaction with substrates in the subphase
17 fluid. Reducing the reactive absorption increases diffusive resistance and back diffusion
18 into the air phase, thereby reducing uptake from the gas phase. Nonetheless, rapid
19 reactions of NC>2 with tissues will maintain a concentration gradient for NC>2 through the
20 alveolar ELF to the underlying tissues.
21 Exercise, relative to rest, increases the dose rate of NO2 to the respiratory tract because of
22 greater NC>2 penetration through the extrathoracic airways and a greater intake rate of
23 NO2. The uptake of NC>2 by the upper respiratory tract decreases with increasing
24 ventilation rates occurring with activity. This causes a greater proportion of inhaled NC>2
25 to be delivered to the lower respiratory tract. In humans, exercise results in a shift in the
26 breathing pattern from nasal to oronasal relative to rest. Because the nasal passages scrub
27 gas-phase NC>2 more efficiently than the mouth and because uptake efficiency decreases
28 with increasing flow, exercise delivers a disproportionately greater quantity of the inhaled
29 mass to the lower respiratory tract, where the NC>2 is readily absorbed. Experimental
30 studies have shown that exercise increases the dose rate of NO2 to the respiratory tract by
31 3- to 5-times compared to resting exposures.
32 Compared to healthy adults, children and individuals with asthma might be expected to
33 have greater NC>2 penetration into the lower respiratory tract. Children tend to have a
34 greater oral breathing contribution than adults at rest and during exercise. Limited data
35 also suggest that patients with asthma or allergic rhinitis breathe oronasally at rest.
36 Because the nasal passages scrub gas-phase NO2 more efficiently, a greater quantity of
January 2015 4-26 DRAFT: Do Not Cite or Quote
-------
1 the inhaled NCh may reach the lower respiratory tract of oronasally breathing individuals.
2 The dose rate to the lower airways of children compared to adults is increased further
3 because children breathe at higher minute ventilations relative to their lung volumes.
4 Current dosimetry models for NC>2 do not adequately consider reactive absorption and
5 secondary reactions that affect the probability of oxidants and/or cytotoxic products
6 reaching target sites. Differences in potential NC>2 reactive substrates and reaction
7 products among species have not been considered in modeling efforts. Although the
8 models predict similar total NC>2 dose in the tracheobronchial airways and sites of
9 maximal NC>2 tissue dose (i.e., near the beginning of the gas exchange region) among
10 several mammalian species, the models do not sufficiently link these NO2 doses to
11 specific reaction products and downstream events. It is unclear to what extent
12 environmental exposures at current ambient NC>2 concentrations might affect the overall
13 balance of nitrite and nitrate or how ambient NC>2 uptake compares with endogenous
14 production rates/amounts in the respiratory tract. However, the uptake of inhaled NCh
15 could increase levels of nitrite and/or other reaction products beyond levels that are
16 endogenously occurring in the respiratory tract.
17 The uptake of inhaled NO occurs against the background of endogenous NO production
18 in the respiratory tract. In terms of the overall uptake and tissue diffusion of NO within
19 the lung, interception due to reactions is not expected to consume appreciable amounts of
20 the total NO involved in mass transfer from the alveolar to the vascular space. The
21 absorption of inhaled NO proceeds similarly to oxygen and CO. Blood acts as a near
22 "infinite" sink for NO. Absorption of NO follows Henry's law for dissolution into the
23 aqueous phase, and is followed by diffusion into the vascular space, where it interacts
24 with red blood cell hemoglobin to ultimately form nitrate. Ambient NO concentrations
25 are likely similar to those endogenously occurring within the lung airspaces, except
26 during morning commutes or near major roadways, where they may possibly exceed
27 endogenous levels. It is not known whether periods of high ambient NO exposure could
28 alter endogenous NO production within the respiratory tract or pathways affected by
29 endogenous NO.
January 2015 4-27 DRAFT: Do Not Cite or Quote
-------
4.3 Modes of Action for Inhaled Oxides of Nitrogen
4.3.1 Introduction
1 The purpose of this section is to describe the biological pathways that underlie health
2 effects resulting from short-term and long-term exposures to NO2 and NO. Extensive
3 research carried out over several decades in humans and in laboratory animals has
4 yielded much information on these pathways. This section will discuss some of the
5 representative studies with particular emphasis on studies published since the 2008 ISA
6 for Oxides of Nitrogen—Health Criteria (U.S. EPA. 2008) and on studies in humans.
7 This information will be used to develop a mode of action framework for inhaled NO2
8 and NO.
9 Mode of action refers to a sequence of key events, endpoints, and outcomes that result in
10 a given toxic effect (U.S. EPA. 2005). Elucidation of mechanism of action provides a
11 more detailed understanding of key events, usually at the molecular level (U.S. EPA.
12 2005). The framework developed in this chapter will include some mechanistic
13 information on initiating events at the molecular level, but will mainly focus on the
14 effects of NO2 and NO at the cellular, tissue, and organism level.
15 NO2 is a radical species and a highly reactive oxidant gas [(Tukuto et al.. 2012)
16 Table 4-21. It is well appreciated that oxidation and nitration reaction products, which are
17 formed as a result of NO2 exposure, initiate numerous responses at the cellular, tissue,
18 and whole organ level of the respiratory system. Exposure to NO2 may also have effects
19 outside the respiratory tract. NO is a radical species and a gas that is more selective in its
20 reactivity than NO2 RTukuto et al.. 2012) Table 4-21. Once inhaled, NO rapidly crosses
21 the alveolar capillary barrier into the vascular compartment and avidly binds to
22 hemoglobin. Subsequent reactions with hemoglobin lead to the generation of circulating
23 nitrate, nitrite, and methemoglobin.
January 2015 4-28 DRAFT: Do Not Cite or Quote
-------
Table 4-2 Chemical properties of nitrogen dioxide (NO2) and nitric oxide (NO)
that contribute to modes of action.
NO2 NO
Radical species Radical species
Somewhat hydrophobic Very hydrophobic
Very reactive Selectively reactive
Less diffusible More diffusible
Reactions with unsaturated fatty acids, thiols, and low Radical-radical reactions with
molecular weight antioxidants (1) superoxide to form peroxynitrite
(2) thiyl radicals to form RSNO
(3) organic peroxyl radicals
Reacts with amino acids, proteins and lipids to form Reacts with heme-containing proteins, transition metals and
nitrated species oxygen
Initiates radical reactions and lipid peroxidation Quenches radical reactions
Metabolites include nitrite and nitrate Metabolites include nitrite and nitrate
RSNO = S-nitrosothiols.
1 Both NO2 and NO are formed endogenously in cells and tissues (Sections 4.2.2.4 and
2 4.2.3). Formation of endogenous NO is catalyzed by NOS. Three pathways contribute to
3 the formation of endogenous NO2: (1) acidification of nitrite, usually occurring in the
4 phagolysosomes; (2) reaction of peroxynitrite with carbonate to form
5 nitrosoperoxylcarbonate anion, which decomposes to carbonate anion and NO2; and
6 (3) reaction of peroxidases using nitrite and hydrogen peroxide as substrates. These
7 enzymatic and nonenzymatic pathways are increased during immune responses and
8 inflammation, leading to higher endogenous levels of NO and NO2. Furthermore, dietary
9 consumption of nitrate leads to enhanced levels of NO in the stomach and to enhanced
10 circulating levels of nitrite due to activity of the enterosalivary cycle (Weitzberg and
11 Lundberg, 2013; Lundberg et al.. 2011). The contribution of environmentally relevant
12 concentrations of inhaled NO2 and NO to levels of circulating nitrite and nitrate is
13 thought to be minimal (Section 4.2.2.4). However, inhaled NO2 may act on the same
14 targets as endogenous NO2 produced during inflammation in the respiratory tract (Ckless
15 et al.. 2011). Because endogenous NO2 is thought to contribute to the development of
16 lung disease, inhaled NO2 may further this process.
17 The following subsections describe the current understanding of biological pathways that
18 may be responsible for the pulmonary and extrapulmonary effects of inhaled NO2 and
19 NO. For NO2, this includes the formation of oxidation and nitration reaction products
20 (Section 4.3.2.1), activation of neural reflexes (Section 4.3.2.2). initiation of
January 2015 4-29 DRAFT: Do Not Cite or Quote
-------
1 inflammation (Section 4.3.2.3). alteration of epithelial barrier function (Section 4.3.2.4).
2 enhancement of bronchial smooth muscle reactivity (Section 4.3.2.5). modification of
3 innate/adaptive immunity (Section 4.3.2.6). and remodeling of airways and alveoli
4 (Section 4.3.2.7). The potential induction of carcinogenesis is also briefly described
5 (Section 4.3.2.8). While NO2 exposure may result in effects occurring outside of the
6 respiratory tract, biological pathways underlying extrapulmonary effects of NO2 are not
7 well understood (Section 4.3.2.9). Activation of neural reflexes and release of NO2
8 metabolites or vasoactive mediators from the lung to the bloodstream are possibilities.
9 Inhaled NO impacts the pulmonary and systemic vasculature mainly through interaction
10 with heme proteins (Section 4.3.3). Other effects of NO may be due to circulating
11 metabolites, such as nitrite, nitrate, and methemoglobin, to interactions with redox-active
12 transition metals, and to reactions with thiyl and superoxide radicals. Because
13 endogenous NO is an important mediator of cell signaling, inhaled NO has the potential
14 to disrupt cell signaling.
4.3.2 Nitrogen Dioxide
4.3.2.1 Formation of Oxidation and Nitration Reaction Products
15 The 2008 ISA and the 1993 AQCD (U.S. EPA. 2008) (U.S. EPA. 1993) summarized
16 biochemical effects observed in the respiratory tract after NO2 exposure. These effects
17 have been attributed to the strong oxidizing potential of NO2, resulting in the formation
18 of reactive oxygen species (ROS). Key responses include oxidation of membrane
19 polyunsaturated fatty acids, thiol groups, and antioxidants. Chemical alterations of lipids,
20 amino acids, proteins, and enzymes can lead to functional changes in membranes,
21 enzymes, and oxidant/antioxidant status. For example, lipid peroxidation of unsaturated
22 fatty acids in membranes may alter membrane fluidity and permeability. As a result,
23 epithelial barrier functions may be impaired, and phospholipases may be activated
24 leading to the release of arachidonic acid. In addition, oxidation of protein thiols may
25 result in enzyme dysfunction. Further, consumption of low molecular weight antioxidants
26 by NO2 may result in decreased antioxidant defenses. Effects may occur directly through
27 the action of NO2 or secondarily due to its reaction products, such as organic radicals,
28 ROS, or reactive nitrogen species (RNS). Later effects may occur due to release of ROS
29 and/or RNS by leukocytes responding to cell damage.
30 As summarized in the 2008 ISA and the 1993 AQCD (U.S. EPA. 2008) (U.S. EPA.
31 1993). considerable attention has been paid to the effects of NO2 on the antioxidant
32 defense system in the ELF and in respiratory tract tissue. Studies in humans and animals
January 2015 4-30 DRAFT: Do Not Cite or Quote
-------
1 exposed to NCh have demonstrated changes in low molecular weight antioxidants such as
2 glutathione, ascorbate, and a-tocopherol, and in the activities of enzymes responsible for
3 glutathione synthesis or maintenance of redox status. For example, a controlled human
4 exposure study found depletion of urate and ascorbate, but not glutathione, in BAL fluid
5 1.5 hours following a 4-hour exposure to 2,000 ppb NC>2 (Kelly etaL 1996a). While
6 these results may be interpreted as evidence that NO2 prefers to react with urate or
7 ascorbate over glutathione, an alternative interpretation is that glutathione reacts with
8 NC>2 and that the product of the reaction is reduced by urate or ascorbate. Other studies
9 have found that antioxidant status modulates the effects of NO2 inhalation. For example,
10 in a controlled human exposure study, supplementation with ascorbate and a-tocopherol
11 decreased the levels of lipid peroxidation products found in BAL fluid following a 3-hour
12 exposure to 4,000 ppb NCh (Mohsenin. 1991). Additionally, changes in lung antioxidant
13 enzyme activity have been reported in animals exposed to NC>2 (U.S. EPA. 2008). For
14 example, long-term exposure to NCh resulted in decreased glutathione peroxidase activity
15 in weanling mice that were a-tocopherol deficient, while supplementation with
16 a-tocopherol resulted in an increase in glutathione peroxidase activity (Ayaz and
17 Csallany. 1978). Thus, NCh inhalation is capable of perturbing glutathione-dependent
18 reactions. These changes may reflect altered cell populations because injury induced by
19 NC>2 exposure may result in the influx of inflammatory cells or the proliferation of
20 resident epithelial or mesenchymal cells. Changes in cell populations due to proliferative
21 repair may also account for the upregulation of Phase II, Phase I, and glycolytic enzymes,
22 which have been observed following NC>2 exposure.
23 As discussed in Section 4.2.2.1.2. reactive absorption of NC>2 gas occurs by reactions
24 with antioxidants and other components of the ELF. Studies employing in vitro and
25 in vivo systems point to the ability of antioxidants to both react with NC>2 to form reactive
26 intermediates and to quench those reactive intermediate species. NO2 exposure in the
27 presence of ELF antioxidants resulted in the formation of superoxide and hydrogen
28 peroxide in an in vitro cell system fVelsor and Postlethwait. 1997). In this study,
29 quenching of NCh-derived secondary oxidants was dependent on antioxidant
30 concentration, with lower concentrations promoting and higher concentrations reducing
31 oxidative injury. A recent in vivo study provided additional support for this mechanism.
32 Supplementation of mice with ascorbate had a biphasic effect, with a lower dose of
33 ascorbate promoting and a higher dose of ascorbate reducing lung injury and
34 inflammation induced by exposure to 20,000 ppb NC>2 (Zhang et al.. 201 Ob). Thus,
35 toxicity resulting from NC>2 exposure may be due to a product derived from the initial
36 ELF substrate and/or to secondary reaction products formed. These reaction products
37 may not be long-lived due to short half-lives and/or continuous turnover of the ELF.
38 Further, quenching of reaction products by ELF antioxidants may limit damage to
39 respiratory epithelium. The heterogeneous distribution of epithelial injury due to reactive
January 2015 4-31 DRAFT: Do Not Cite or Quote
-------
1 intermediates formed from inhaled NC>2 may reflect ELF-dependent local effects because
2 the ELF is nonuniform in composition and quantity along the respiratory tract.
3 Furthermore, localized endogenous formation of NCh in the respiratory tract could
4 potentially overwhelm the antioxidant capacity and contribute to epithelial injury.
5 Nitrogen-based metabolites and RNS are also formed in the ELF as a result of NO2
6 exposure. Nitrite is the primary product of the chemical reactions of NC>2 in the
7 respiratory tract. As discussed in Section 4.2.2.1.2. nitrite formed in the ELF can diffuse
8 into respiratory tract epithelial cells and subsequently into the vascular space. While the
9 effects of nitrite on the epithelial cell are not well known, it is unlikely that nitrite is
10 responsible for the toxicity of NC>2. Interestingly, numerous studies have explored the
11 effects of increased systemic nitrite on various tissues and organs. Nitrite has been found
12 to protect against ischemia-reperfusion injury in the heart and other organs (Weitzberg
13 and Lundberg. 2013). In addition, systemic nitrite administration prevented airway and
14 epithelial injury due to exposure to chlorine gas in rats (Yadav et al.. 2011). Further,
15 nitrite is known to have a direct relaxing effect on smooth muscle [(Folinsbee. 1992) see
16 Section 4.3.4.11. suggesting that it may play a role in bronchodilation.
17 RNS, such as RSNO and nitrated proteins, fatty acids, and lipids, may be formed in the
18 respiratory tract following NC>2 exposure. Evidence for these reaction products is mainly
19 provided by in vitro cell systems and ex vivo systems (Section 4.3.4). However, Matalon
20 et al. (2009) recently demonstrated the nitration of surfactant protein D (SP-D) in mice
21 exposed to 20,000 ppb NC>2 for 4 hours. SP-D nitration was accompanied by protein
22 cross-linking and a decrease in SP-D aggregating activity, which could potentially impact
23 microbial clearance, immune regulation, and surfactant metabolism. In addition to
24 inhibiting protein function, nitration of proteins may induce antigenicity or trigger
25 immune reactions (Daiber and Muenzel. 2012). Further, the presence of nitrated amino
26 acids, such as 3-nitrotyrosine, in cells or tissues is an indicator of endogenous NC>2 and
27 peroxynitrite formation. Other potential RNS formed may have less deleterious effects.
28 For example, nitrated (or nitro) fatty acids have a direct relaxing effect on smooth
29 muscle, perhaps even on airway smooth muscle (Que et al.. 2009; Lima et al.. 2005). In
30 addition, RSNOs are known to be bronchodilators (Que et al.. 2009). Additional
31 discussion of the biological effects of these products of NO2 metabolism is found in
32 Section 4.3.4.
4.3.2.2 Activation of Neural Reflexes
33 NO2 is classified as a pulmonary irritant (Alarie. 1973). Pulmonary irritants stimulate
34 afferent nerve endings in the lung, resulting in increased respiratory rate, decreased VT,
January 2015 4-32 DRAFT: Do Not Cite or Quote
-------
1 and subsequent rapid shallow breathing. Sometimes pulmonary irritants also stimulate
2 mild bronchoconstriction, bradycardia, and hypotension (Alarie. 1973). All of these
3 pathways involve the vagus nerve.
4 In guinea pigs, NC>2 exposure (5,200-13,000 ppb; 2-4 hours) by nose-cone resulted in
5 statistically significant increases in respiratory rate and decreases in VT (Murphy et al..
6 1964). These responses were concentration and time dependent and were reversible when
7 animals were returned to clean air. In contrast, no changes in these respiratory parameters
8 were observed with 4-hour exposures to 16,000 and 50,000 ppb NO. In another study,
9 guinea pigs exposed to 7,000-146,000 ppb NC>2 for 1 hour demonstrated a
10 concentration-dependent increase in respiratory rate 10 minutes following exposure and a
11 concentration-dependent decrease in VT 10 minutes, 2 hours, and 19 hours following
12 exposure (Silbaugh et al., 1981). NO2 exposure-induced increases in respiratory rate have
13 also been reported in rats (Freeman et al.. 1966) and mice (McGrath and Smith. 1984). In
14 mice, statistically significant increases in respiratory rate and decreases in VT were found
15 in response to an 8-minute exposure to 100,000 ppb NCh, but not to 15,000 or 50,000 ppb
16 NO2 (McGrath and Smith. 1984). In this latter study, continuous pre-exposure to
17 5,000 ppb NC>2 for 3 days lessened the response to 100,000 ppb NO2, suggesting the
18 development of tolerance or an attenuated response to NO2 (U.S. EPA. 1993). In rats,
19 continuous exposure to 800 ppb and higher concentrations of NCh resulted in elevated
20 respiratory rates throughout life (Freeman et al.. 1966). However, no NC>2
21 exposure-induced increases in respiratory rate in human subjects have been reported. In
22 fact, respiratory rates tended to decrease in humans exposed to 0-480 ppb for 20 minutes
23 (Bvlinet al., 1985). The authors proposed that NC>2 in this range of concentrations did not
24 act as a pulmonary irritant in humans.
25 NC>2 has been shown to elicit a small increase in airway resistance, which is consistent
26 with mild bronchoconstriction, in humans but not in rabbits or guinea pigs [ Alarie (1973)
27 and studies cited below]. One study in human subjects at rest found a nonmonotonic
28 response to NO2 in terms of airway resistance (Bvlinet al.. 1985). In this study, specific
29 airway resistance was increased after 20 minutes of exposure to 250 ppb NO2 and was
30 decreased after 20 minutes of exposure to 480 ppb NO2. The authors suggested that reflex
31 bronchoconstriction occurred at the lower concentration and that other mechanisms
32 counteracted this effect at the higher concentration. Other controlled human exposure
33 studies found no change in airway resistance with acute exposures of 530-1,100 ppb NO2
34 and increases in airway resistance with acute exposures above 1,600-2,500 ppb in
35 healthy human subjects (U.S. EPA. 1993). Human subjects with chronic lung disease
36 exposed for 5 minutes to 2,100 ppb NO2 also exhibited increased airway resistance (von
37 Nieding and Wagner. 1979). In addition, both forced expiratory volume in 1 second
38 (FEVi) and forced vital capacity were decreased in healthy human subjects exposed to
January 2015 4-33 DRAFT: Do Not Cite or Quote
-------
1 2,000 ppb NC>2 for 4 hours (Blomberg et al.. 1999). These latter changes in pulmonary
2 function are consistent with reflex bronchoconstriction. Because the response was
3 lessened with each successive exposure on 4 consecutive days, the authors suggested the
4 development of tolerance or an attenuated response.
5 Some evidence points to NC>2 exposure-induced histamine release from mast cells, rather
6 than reflex bronchoconstriction, as the mechanism underlying changes in airway
7 resistance. A study in rats showed that mast cell degranulation occurred after acute
8 exposure to NCh [500 ppb for 4 hours; 1,000 ppb for 1 hour (Thomas et al.. 1967)1. In
9 addition, a histamine-suppressive agent, but not atropine or /^-agonists, blocked
10 NO2-mediated increases in airway resistance in healthy humans and in humans with
11 chronic lung disease exposed to 5,000-8,000 ppb NO2 for 5 minutes (von Nieding and
12 Wagner. 1979). Because atropine inhibits vagal responses, these findings indicate that
13 neural reflexes were not involved in NC>2-induced changes in pulmonary function in
14 human subjects. More recent studies in animals have provided experimental evidence for
15 a relationship between lipid peroxidation/oxidative stress and the release of histamine by
16 allergen-activated mast cells (Beaven. 2009; Gushchin et al.. 1990). Taken together, these
17 studies suggest that NCh exposure may lead to lipid peroxidation, which may promote
18 mast cell-mediated changes in pulmonary function, albeit at high concentrations.
19 There is some experimental support for NC>2 exposure-induced cardiovascular reflexes.
20 An acute exposure to NC>2 in an occupational setting resulted in tachycardia in one case
21 report (U.S. EPA. 1993; Bates etal.. 1971). while rats exposed acutely to 20,000 ppb or
22 higher concentrations of NC>2 exhibited bradycardia (U.S. EPA. 1993; Tsubone et al..
23 1982). This latter response was abolished by injection of atropine, which inhibits vagal
24 responses. Further, a decreased heart rate, which was not accompanied by an increase in
25 respiratory rate, was observed in mice exposed to 1,200 and 4,000 ppb NC>2 for 1 month
26 (Suzuki et al.. 1981). The lack of respiratory rate response suggests that the decreased
27 heart rate was due to a different mechanism than rapid stimulation of irritant receptors by
28 NO2. Controlled human exposure studies have also examined the effects of NO2 on heart
29 rate and heart rate variability (Section 5.3.11.1). Older studies and one recent study failed
30 to find statistically significant changes in heart rate at ambient-relevant concentrations of
31 NO2. A recent controlled human exposure study involving a 1 -hour exposure to 400 ppb
32 NO2 failed to find an effect on heart rate variability in subjects with coronary heart
33 disease (Scaife et al.. 2012). However, a second recent controlled human exposure study
34 reported an effect on heart rate variability, which is a measure of autonomic tone,
35 resulting from a 2-hour exposure to 500 ppb NO2 (Huang etal.. 2012). Altered heart rate
36 variability found in epidemiologic studies (Section 5.3.11.1) is consistent with a possible
37 effect of NO2 exposure on autonomic tone.
January 2015 4-34 DRAFT: Do Not Cite or Quote
-------
1 In summary, NC>2 is a pulmonary irritant that may activate reflexes through vagal
2 pathways to increase respiratory rate, decrease VT, stimulate reflex bronchoconstriction,
3 and induce bradycardia. Responses are rapid, concentration dependent, and variable
4 among species. Evidence that reflex responses occur in humans is weak because no
5 increases in respiratory rate have been reported as a result of NO2 exposure. Further,
6 modest increases in airway resistance in human subjects exposed to NO2 were not
7 blocked by atropine, which inhibits vagal responses. Findings attributed to reflex
8 bronchoconstriction in humans may be due to alternative pathways such as mast cell
9 degranulation. The recent demonstration that NCh exposure (500 ppb; 2 hours) results in
10 altered heart rate variability suggests the possible activation of a neural reflex in humans.
11 However, the clearest evidence for reflex responses mediated by the vagus nerve
12 involved exposures of experimental animals to NC>2 at concentrations of at least
13 5,000 ppb.
4.3.2.3 Initiation of Inflammation
14 As summarized in the 2008 ISA and the 1993 AQCD (U.S. EPA. 2008) (U.S. EPA.
15 1993). NO2 exposure-induced membrane perturbations resulted in the release of
16 arachidonic acid and the formation of eicosanoid products (Sections 5.2.2.5 and 5.2.7).
17 Animal toxicological and controlled human exposure studies have found increases in
18 concentrations of eicosanoids in BAL fluid immediately following exposure to NC>2
19 (Torres etal.. 1995; Schlesinger et al.. 1990). Eicosanoids play an important role in the
20 recruitment of neutrophils. Interestingly, higher concentrations and longer durations of
21 exposure to NCh employed in these studies resulted in inhibited eicosanoid production
22 (Robison and Forman. 1993; Schlesinger et al.. 1990).
23 Recently, acute exposure of mice to 10,000 ppb and higher concentrations of NC>2 was
24 shown to activate the transcription factor nuclear factor kappa-light-chain-enhancer of
25 activated B cells (NFKB) in airway epithelium (Ather et al.. 2010; Bevelander et al..
26 2007). NFKB activation resulted in the production of pro-inflammatory cytokines.
27 Inflammation and acute lung injury in this model were found to be dependent on an
28 active NFKB pathway. Controlled human exposure studies demonstrated increased levels
29 of cytokines IL-6 and IL-8 in BL fluid following NC>2 exposure. IL-8 levels were
30 increased at 1.5 and 16 hours following a 4-hour exposure to 2,000 ppb NC>2, while levels
31 of IL-6 were increased at 16 hours following the exposure [Section 5.2.7 (U.S. EPA.
32 2008) (Devlin et al.. 1999; Blomberg et al.. 1997)1. The cell signaling pathways
33 responsible for upregulating cytokines at this lower level of exposure to NC>2 are not
34 clear.
January 2015 4-35 DRAFT: Do Not Cite or Quote
-------
1 Airway inflammation often occurs following NC>2 exposure. Studies in rodents exposed
2 acutely (1 hour to 3 days) to NCh (500-30,000 ppb) have demonstrated airway
3 inflammation mainly consisting of neutrophils and macrophages, and sometimes of mast
4 cells and lymphocytes, by histological technique or sampling of BAL fluid [as
5 summarized in Sandstrom et al. (1990) (Poynter et al.. 2006: Pagani et al.. 1994)1.
6 Numerous studies in healthy human subjects exposed to NC>2 have documented airway
7 inflammation in endobronchial biopsy tissue and in sputum, BL fluid, and BAL fluid.
8 Many of these studies were conducted while subjects were exercising intermittently and
9 exposed to 1,500-4,000 ppb NCh for a few hours. Neutrophilia was a prominent feature
10 (U.S. EPA. 2008) (Frampton et al.. 2002: Devlin etal.. 1999: Azadniv et al.. 1998:
11 Blomberg et al.. 1997). In addition, other types of inflammatory cells, including
12 macrophages, lymphocytes, and mast cells, have been demonstrated (Frampton et al..
13 2002: Sandstrom et al.. 1991: Sandstrom et al.. 1990).
14 Controlled human exposure studies have also evaluated the effects of repeated NO2
15 exposure on airway inflammation in healthy adults. Persistent neutrophilic inflammation,
16 demonstrated by increased numbers of neutrophils and increased levels of
17 myeloperoxidase in the BL fluid, was observed following 4 consecutive days of 4-hour
18 exposure to 2,000 ppb NCh (Blomberg et al.. 1999). Repeated exposure also led to the
19 upregulation of cytokines characteristic of the T helper cell 2 (Th2) inflammatory
20 response and also intercellular adhesion molecule 1 (ICAM-1) in respiratory epithelium
21 (U.S. EPA. 2008) (Pathmanathan et al.. 2003). Upregulation of ICAM-1 suggests a
22 potential mechanism for the persistent neutrophil influx that was observed (Blomberg
23 etal.. 1999). A study of repeated exposure to 4,000 ppb (exposure every other day for a
24 total of six exposures) found inflammatory responses that differed from those observed
25 after a single exposure (Sandstrom et al.. 1992). In particular, numbers of mast cells and
26 lymphocytes in the lavage fluid, which were increased following a single exposure, were
27 not increased following repeated exposure. Furthermore, repeated exposure to 1,500 ppb
28 NC>2 (by the same protocol) resulted in smaller numbers of some lymphocyte
29 subpopulations in BAL obtained following exposure compared with numbers in BAL
30 obtained prior to exposure (Sandstrom et al.. 1992). In contrast, no changes in
31 lymphocyte subpopulations were reported following repeated exposure to 600 ppb NC>2
32 (4 exposures over 6 days), with the exception of a slight increase in natural killer cells
33 (Rubinstein et al.. 1991).
34 Recently, a controlled human exposure study investigated the effects of repeated NO2
35 exposure on eosinophilic airway inflammation in subjects with atopic asthma (Ezratty
36 etal.. 2014). Subjects were exposed to 200 or 600 ppb NO2 for 30 minutes on the first
37 day and twice for 30 minutes on the second day. Compared with baseline, the number
38 and percentage of eosinophils and the amount of eosinophil cationic protein (ECP) in
January 2015 4-36 DRAFT: Do Not Cite or Quote
-------
1 sputum were significantly increased after the three exposures to 600, but not 200 ppb
2 NC>2. Furthermore, ECP was highly correlated with eosinophil counts in sputum. No
3 increases in either of these parameters were observed 6 hours after the first exposure to
4 600 ppb NC>2. While the design of this study did not include an allergen challenge, several
5 other studies examined eosinophilic inflammation and other allergic responses to NC>2
6 and an allergen. These are discussed in Sections 4.3.2.6.2. 4.3.2.6.3. and 5.2.2.5.
7 Collectively, these studies suggest that exposure to NO2 may prime eosinophils for
8 subsequent activation by allergens in previously sensitized individuals (Davies et al..
9 1997; WangetaL 1995b).
4.3.2.4 Alteration of Epithelial Barrier Function
10 Lipid peroxidation and altered phospholipid composition in the respiratory tract
11 following NO2 exposure may affect membrane fluidity and airway epithelial barrier
12 function. NC>2 exposure-induced inflammation may further impair epithelial barrier
13 function. Increases in vascular permeability may occur, leading to the influx of plasma
14 proteins such as albumin into the airway lumen.
15 As summarized in the 2008 ISA and the 1993 AQCD (U.S. EPA. 2008) (U.S. EPA.
16 1993). numerous studies have demonstrated increases in biomarkers of increased
17 permeability, such as protein and albumin, as well as biomarkers of cellular injury, such
18 as lactate dehydrogenase (LDH) and shed epithelial cells, in BAL fluid following
19 exposure to NCh (Section 5.2.7). Because LDH can be oxidatively inactivated, use of this
20 indicator may underestimate the extent of injury during oxidative stress. Many, but not
21 all, of these effects were observed at NO2 concentrations that are higher than
22 ambient-relevant levels. Notably, one controlled human exposure study found increased
23 albumin levels in BL fluid following 4 consecutive days of 4-hour exposure to 2,000 ppb
24 NQ2 (Blomberg et al.. 1999).
25 Several studies in experimental animals found that antioxidant deficiency worsened the
26 cellular injury and/or impaired epithelial barrier function following NC>2 exposure.
27 Ascorbate deficiency enhanced protein levels in the BAL fluid of NC>2-exposed guinea
28 pigs, suggesting a role for BAL fluid ascorbate in preventing the deleterious effects of
29 NO2 (Hatchet al.. 1986). Similarly, a-tocopherol deficiency enhanced lipid peroxidation
30 in NCh-exposed rats (Sevanian et al.. 1982). Recently, selenium deficiency was found to
31 enhance the injury response in rats exposed to 1,000-50,000 ppb (acute, subacute, and
32 chronic exposures) NCh (de Burbure et al.. 2007). Levels of both BAL fluid total protein
33 and serum club cell secretory protein (CC16) were increased in selenium-deficient rats
34 exposed to NCh. Selenium supplementation diminished this response, which suggests that
January 2015 4-37 DRAFT: Do Not Cite or Quote
-------
1 the selenium-containing enzyme, glutathione peroxidase, may have played an important
2 mitigating role.
3 Increases in lung permeability due to high concentrations of NC>2 (100,000 ppb and
4 above) are known to cause death from pulmonary edema (Lehnert et al.. 1994; Gray
5 etal.. 1954). At lower concentrations, more subtle effects have been reported. Exposure
6 of rats to 5,000 and 10,000 ppb NC>2 for 3 or 25 days resulted in epithelial degeneration
7 and necrosis and in proteinaceous edema (Earth etal.. 1995). while exposure to
8 800-10,000 ppb NO2 for 1 and 3 days resulted in concentration-dependent increases in
9 BAL fluid protein (Muller etal.. 1994). BAL fluid protein was also elevated in guinea
10 pigs exposed for 1 week to 400 ppb NO2 (Sherwin and Carlson. 1973).
11 High concentrations of NC>2 (70,000 ppb, 30 minutes) were found to enhance
12 translocation of instilled antigen from the lung to the bloodstream of guinea pigs
13 (Matsumura. 1970). More subtle increases in lung permeability due to NC>2 exposure
14 could enhance the translocation of an antigen to local lymph nodes and subsequently to
15 the circulation (U.S. EPA. 2008) (Gilmour et al.. 1996) and/or to the immunocompetent
16 and inflammatory cells underlying the epithelium that are involved in allergic reactions
17 (Jenkins etal.. 1999). However, increased lung permeability following exposure to NC>2
18 does not always lead to allergic sensitization (Alberg et al.. 2011). Increased epithelial
19 permeability may alternatively contribute to the activation of neural reflexes and the
20 stimulation of smooth muscle receptors by allowing greater access of an agonist (Dimeo
21 etal.. 1981).
22 Susceptibility to NC>2 exposure-induced lung injury was investigated in several mice
23 strains with differing genetic backgrounds (Kleeberger et al.. 1997). Lavageable total
24 protein, a biomarker for increased lung permeability, was variable among mouse strains
25 following a 3-hour exposure to 15,000 ppb NCh. In addition, repeated exposure to NC>2
26 (10,000 ppb, 6 h/day, 5 consecutive days) resulted in adaptation of the permeability
27 response in one of the tested strains but not in the other. Although specific genes were not
28 identified, this study provided evidence that genetic components conferred susceptibility
29 to NO2, at least in terms of lung permeability.
4.3.2.5 Enhancement of Bronchial Smooth Muscle Reactivity
30 Exposure to NC>2 enhanced the inherent reactivity of airway smooth muscle in human
31 subjects with and without asthma [(Folinsbee. 1992) Section 5.2.2.1] and in animal
32 models (see below). This "airway responsiveness" is defined as the sensitivity of airways
33 to a variety of natural or pharmacological stimuli (O'Byrne et al.. 2009). Airway
34 hyperresponsiveness (AHR) is a key feature of asthma, which is a chronic inflammatory
January 2015 4-38 DRAFT: Do Not Cite or Quote
-------
1 disease of the airways. As summarized in the 2008 ISA (U.S. EPA. 2008) and in
2 Section 5.2.2.1. numerous studies found that human subjects who were exposed to NC>2
3 were more sensitive to the nonspecific stimuli methacholine than human subjects who
4 were exposed to air. Subjects with asthma exhibited greater sensitivity than subjects
5 without asthma when similarly exposed. In addition, several studies found that NC>2
6 exposure enhanced airway responsiveness to specific stimuli, such as allergens, in
7 subjects with mild allergic asthma.
8 Exercise during exposure to NC>2 appeared to modify airway responsiveness in subjects
9 with asthma [(Folinsbee. 1992) Section 5.2.2.1]. Mechanisms by which this occurs are
10 not understood, but two hypotheses have been postulated. First, exercise-induced
11 refractoriness, which has been demonstrated in some subjects with asthma, may alter
12 responsiveness to NC>2 (Magnussen et al.. 1986). A second hypothesis is that nitrite
13 formed by reactions of NC>2 in the ELF mediates compensatory relaxation of airway
14 smooth muscle (Folinsbee. 1992). Exercise would increase the total dose of NO2 to the
15 respiratory tract, thus increasing nitrite formation. Recent studies have shown that RNS
16 have bronchodilatory effects. For example, endogenous RSNOs are an important
17 modulator of airway responsiveness in subjects with asthma and in eosinophilic
18 inflammation (Lee etal.. 2011; Que et al.. 2009).
19 Animal toxicological studies have also demonstrated NCh-induced AHR to nonspecific
20 and specific challenges, as summarized in the 2008 ISA and the 1993 AQCD (U.S. EPA.
21 2008) (U.S. EPA. 1993) and in Sections 5.2.2.1 and 6.2.2.3. Exposures ranged from acute
22 to subchronic in these studies, and results suggest that more than one mechanism may
23 have contributed to the observed AHR. Acute exposure of guinea pigs to NO2
24 (10 minutes, 7,000 ppb and higher) resulted in concentration-dependent AHR to
25 histamine, which was administered immediately after exposure (Silbaugh et al.. 1981).
26 This response was short-lived because no enhanced responsiveness was seen at 2 and
27 19 hours post-exposure to NCh. The rapidity of the response and the concomitant change
28 in respiratory rate suggest enhanced vagally mediated reflex responses (Section 4.3.2.2)
29 as a possible underlying mechanism. A 7-day exposure to 4,000 ppb NCh also induced
30 AHR to histamine in guinea pigs (Kobayashi and Shinozaki. 1990). Eicosanoids were
31 proposed to play a role in this response. In addition, a study in mice sensitized and
32 challenged with ovalbumin found that short-term exposure to NC>2 (25,000 ppb, but not
33 5,000 ppb; 3 days) resulted in AHR to methacholine (Poynter et al.. 2006). This enhanced
34 sensitivity correlated with an increase in numbers of eosinophils, suggesting eosinophilic
35 inflammation as a possible underlying mechanism in this model of allergic airway
36 disease. A subchronic study demonstrated dose-dependent AHR to histamine in
37 NO2-exposed guinea pigs [1,000-4,000 ppb, 24 h/day, 6-12 weeks (U.S. EPA. 2008)
38 (Kobayashi and Miura. 1995)1. Specific airway resistance in the absence of a challenge
January 2015 4-39 DRAFT: Do Not Cite or Quote
-------
1 agent was also increased, which indicates the development of airway obstruction. This
2 finding suggests airway remodeling as a possible underlying mechanism for AHR.
3 Another subchronic exposure study (5,000 ppb NCh, 4 h/day, 5 days/week, 6 weeks)
4 found a delayed bronchial response, which was measured as increased respiratory rate
5 and was suggestive of AHR, in guinea pigs sensitized and challenged with C. albicans
6 and exposed to NO2 (Kitabatake et al.. 1995).
7 Mechanisms underlying the effects of NC>2 on airway responsiveness are not well
8 understood. Effects of NC>2 exposure on redox status in the respiratory tract should be
9 considered because asthma pathogenesis, including airway inflammation,
10 hyperresponsiveness, and remodeling, may be under redox control (Comhair and
11 Erzurum. 2010; Kloek et al.. 2010). In support of this mechanism, supplementation with
12 the antioxidant ascorbate was found to prevent nonspecific AHR in subjects with asthma
13 who were exposed to NO2 (Mohsenin. 1987).
14 Several different inflammatory pathways may underlie the increased airway
15 responsiveness following NC>2 exposure (Krishna and Holgate. 1999). First, mast cell
16 activation may contribute to NC>2 exposure-induced AHR. As discussed in
17 Section 4.3.2.2. acute exposure to NC>2 led to mast cell activation in rats and possibly in
18 human subjects. Histamine released by mast cells can directly bind to receptors on
19 smooth muscle cells and cause contraction. This response would have the appearance of
20 reflex bronchoconstriction but would not involve neural pathways. Secondly, neutrophilic
21 and eosinophilic inflammation, which have been demonstrated following single and
22 repeated exposures to NO2 (Section 4.3.2.3). may play a role. Neutrophils and other
23 inflammatory cell types release mediators, such as IL-13, IL-17A, and tumor necrosis
24 factor-a (TNF-a), which can alter the calcium sensitivity of the smooth muscle and
25 enhance a contractile response to a stimulus (Kudo et al.. 2013). Eosinophils can release
26 ECP and other mediators involved in allergen-induced asthmatic responses. This pathway
27 may contribute to the enhanced immune response to allergens demonstrated following
28 NO2 exposure (Section 4.3.2.6.2). Eosinophil release of ECP may also cause damage to
29 the airway epithelium in allergic airway disease (Ohashi et al.. 1994). This damage may
30 result in epithelial shedding and mucociliary dysfunction, which may allow greater access
31 of allergens to the airway epithelium and submucosa. In addition, epithelial shedding may
32 lead to greater exposure of sensory nerve endings on nerve fibers and to enhanced
33 activation of neural reflexes and airway smooth muscle contraction (Hesterberg et al..
34 2009; Cockcroft and Davis. 2006). These processes may explain the close relationship
35 that has been observed between epithelial shedding and AHR (Ohashi etal.. 1994). Thus,
36 neutrophilic and/or eosinophilic airway inflammation following NC>2 exposure may
37 contribute to AHR through the release of mediators or by impairing epithelial barrier
38 function (Section 4.3.2.4). Thirdly, chronic airway inflammation may cause structural
January 2015 4-40 DRAFT: Do Not Cite or Quote
-------
1 changes in the airway walls that enhance the contractile response of the smooth muscle to
2 a given stimuli (Cockcroft and Davis. 2006).
3 Evidence also supports a role for endogenous NO2 in mediating AHR. Increased
4 peroxynitrite formation occurs during inflammatory states, resulting from the reaction of
5 NO and superoxide. Peroxynitrite subsequently reacts with CO2 to form
6 nitrosoperoxylcarbonate anion, which decomposes to carbonate radical and NO2
7 (Section 4.2.2.4). Recent studies provided evidence that endogenous peroxynitrite
8 contributes to AHR in animal models of allergic airway disease (Section 4.3.2.6.2). These
9 studies demonstrate that NO metabolism is dysfunctional in inflamed lungs and enhances
10 peroxynitrite formation. Amelioration of the dysfunction resulted in less nitrative stress,
11 airway remodeling and airway responsiveness (Ahmad etal.. 2011; Mabalirajan et al..
12 2010b; Maarsingh etal..2009; Maarsingh et al.. 2008).These studies highlight the
13 possibility that inhaled NO2 can add to the lung burden of endogenous NO2, which
14 contributes to AHR and allergic airway disease in animal models (Section 4.3.2.6.2).
4.3.2.6 Modification of Innate/Adaptive Immunity
15 Host defense depends on effective barrier function and on innate and adaptive immunity
16 (Al-Hegelan et al.. 2011). The effects of NO2 on barrier function in the airways were
17 discussed above (Section 4.3.2.4). This section focuses on the mechanisms by which
18 exposure to NO2 impacts innate and adaptive immunity. Both tissue damage and foreign
19 pathogens are triggers for the activation of the innate immune system. Innate immune
20 system activation results in the influx of inflammatory cells, such as neutrophils, mast
21 cells, basophils, eosinophils, monocytes, and dendritic cells, and the generation of
22 cytokines, such as TNF-a, IL-1, IL-6, keratinocyte chemoattractant, and IL-17. Further,
23 innate immunity encompasses complement, collectins, and the phagocytic functions of
24 macrophages, neutrophils, and dendritic cells. In addition, airway epithelium contributes
25 to innate immune responses. Innate immunity is highly dependent on cell signaling
26 networks involving toll-like receptor (TLR) 4 in airway epithelium and other cell types.
27 Adaptive immunity provides immunologic memory through the actions of B and T
28 lymphocytes. Important links between the two systems are provided by dendritic cells
29 and antigen presentation.
4.3.2.6.1 Impairment of Host Defenses
30 As summarized in the 2008 ISA (U.S. EPA. 2008). potential mechanisms by which NO2
31 exposure may impair host defenses include ciliary dyskinesis, damage to ciliated
32 epithelial cells, and altered alveolar macrophage function, all of which may contribute to
January 2015 4-41 DRAFT: Do Not Cite or Quote
-------
1 altered mucociliary transport and/or clearing infectious and noninfectious particles from
2 the lung. Altered alveolar macrophage function and other potential mechanisms, such as
3 increases in pro-inflammatory mediators and cytokines, increased immunoglobulin E
4 (IgE) concentrations, interactions with allergens, and altered lymphocyte subsets, reflect
5 modification of innate and/or adaptive immunity. These changes may underlie
6 susceptibility to infection, which has been observed in animals exposed to NC>2
7 (Section 5.2.5.1).
8 Controlled human exposure studies have demonstrated reduced mucociliary activity due
9 to depressed ciliary function, depressed phagocytic activity, and superoxide production in
10 alveolar macrophages, and altered humoral- and cell-mediated immunity following
11 exposure to 1,500-4,000 ppb NC>2 for a few hours [(Frampton et al.. 2002; Devlin et al..
12 1999; Helledav et al.. 1995; Sandstrom et al.. 1992; Sandstrom et al.. 1992; Sandstrom
13 et al.. 1991) Section 5.2.5]. Studies involving repeated daily exposure to 1,500 ppb NCh
14 (but not 600 ppb NCh) found reductions in lymphocyte subpopulations (Sandstrom et al..
15 1992; Rubinstein et al.. 1991; Sandstrom et al.. 1990). Furthermore, repeated daily
16 exposure to 2,000 ppb NCh resulted in upregulation of ICAM-1 in bronchial biopsy
17 specimens (Pathmanathan et al.. 2003). These findings suggest a potential mechanism
18 underlying susceptibility to viral infection because ICAM-1 is a major receptor for rhino-
19 and respiratory-syncytial viruses. Finally, enhanced susceptibility of airway epithelium to
20 influenza viral infection was suggested in a study involving exposure to 1,000-3,000 ppb
21 NC>2 over 3 days, although statistical significance was not achieved (Goings etal.. 1989).
22 Humans exposed to 600 and 1,500 ppb NC>2 for 3 hours exhibited an increased injury
23 response, as measured in bronchial epithelial cells, resulting from influenza and
24 respiratory syncytial virus (Frampton et al.. 2002). Epidemiologic evidence for
25 associations between exposure to NO2 and increased respiratory infections is somewhat
26 inconsistent (Sections 5.2.5.2 and 5.2.5.3).
27 As summarized in the 2008 ISA (U.S. EPA. 2008) and 1993 AQCD (U.S. EPA. 1993).
28 studies in NCh-exposed animals (500-10,000 ppb) have demonstrated altered
29 mucociliary clearance and several changes in alveolar macrophages. These changes
30 include morphological evidence of damage to alveolar macrophages (membrane bleb
31 formation and mitochondrial damage), decreased viability, and decreased function
32 [decreased superoxide production, decreased phagocytic capacity, and decreased
33 migration towards a stimulus (Robison et al.. 1993; Davis etal.. 1992; Rose etal.. 1989;
34 Schlesinger et al.. 1987; Schlesinger and Gearhart. 1987; Suzuki et al.. 1986; Greene and
35 Schneider. 1978; Powell etal.. 1971)]. A recent study involving exposure to 20,000 ppb
36 NO2 demonstrated nitration of SP-D, a surfactant protein that functions as a collectin
37 (Matalon et al.. 2009). This was accompanied by cross-linking and a decrease in SP-D
38 aggregating activity, which could potentially impact the role of SP-D in microbial
January 2015 4-42 DRAFT: Do Not Cite or Quote
-------
1 clearance and surfactant metabolism. Infectivity models have shown increased mortality
2 and decreased bactericidal activity (U.S. EPA. 2008) (Jakab. 1987: Miller et al.. 1987:
3 Ehrlich. 1980: Ehrlich et al.. 1977). as a result of NCh exposure. Further discussion is
4 found in Sections 5.2.4 and 6.2.9.
4.3.2.6.2 Exacerbation of Allergic Airway Disease
5 Inhaled allergens activate an acute immune response in allergen-sensitive individuals.
6 This response is characterized by early and late phases. Key players in the early asthmatic
7 response are mast cells and basophils, which release mediators following allergen binding
8 to IgE receptors on their cell surfaces. These mediators include histamine and cysteinyl
9 leukotrienes, which bind airway smooth muscle receptors and induce contraction.
10 Mediators also activate T lymphocyte subsets (i.e., CD4+ T-cells), resulting in the release
11 of Th2 cytokines that can cause airway smooth muscle contraction and recruit mast cells.
12 Th2 cytokines also promote the influx and activation of eosinophils and neutrophils.
13 Airway mucosal eosinophilia is characteristic of asthma and rhinitis. Eosinophils exert
14 their effects via degranulation and/or cytolysis, resulting in release of ECP and other
15 mediators (Eriefalt et al., 1999). Th2 cytokines also activate B lymphocytes, resulting in
16 the production of allergen-specific IgE. These responses initiated by Th2 cytokines
17 contribute to the late asthmatic response, which is characterized by airway obstruction
18 generally occurring 3-8 hours following an antigen challenge (Cockcroft and Davis.
19 2006). and to other responses occurring greater than >8 hours following an antigen
20 challenge.
Exogenous Nitrogen Dioxide
21 As summarized in the 2008 ISA (U.S. EPA. 2008) and in Section 5.2.2.1. exposure to
22 NC>2 affects the acute immune response to inhaled allergens. Several controlled human
23 exposure studies found that NC>2 exposure enhanced airway responsiveness to specific
24 stimuli, such as house dust mite (HDM) allergen (Jenkins et al.. 1999: Tunnicliffe et al..
25 1994) in subjects with mild allergic asthma. Further, repeated exposure to NC>2 resulted in
26 an enhanced response to a dose of allergen that was asymptomatic when given alone
27 (Strand et al.. 1998). Airway responses were measured during the first 2 hours after
28 allergen challenge which falls within the timeline of the early phase asthmatic response.
29 These results provide evidence that NO2 exposure exacerbates the early phase asthmatic
30 response to allergen challenge, as measured by enhanced contraction of airway smooth
31 muscle cell.
32 Controlled human exposure studies also demonstrated that NCh exposure exacerbated the
33 late phase asthmatic response to allergen challenge in subjects with mild allergic asthma.
January 2015 4-43 DRAFT: Do Not Cite or Quote
-------
1 Airway obstruction, measured as a spontaneous fall in FEVi occurring after resolution of
2 the early asthmatic response (generally 3-8 hours after an antigen challenge), was
3 observed in subjects with asthma exposed to 400 ppb NCh for 1 hour (Tunnicliffe et al..
4 1994) and to 250 ppb NO2 for 30 minutes for 4 consecutive days (Strand et al.. 1998).
5 Other studies measured cell counts and mediators characteristic of the late phase
6 asthmatic response. Increased numbers of neutrophils and increased levels of ECP in
7 BAL fluid and/or BL fluid, both indicators of inflammatory response to allergen
8 challenge, were reported following exposure to 260 ppb NO2 for 15-30 minutes (Barck
9 etal.. 2005; Barck etal.. 2002). Furthermore, increased ECP levels were observed in
10 sputum and blood, and an increase in myeloperoxidase (indicator of neutrophil
11 activation) was seen in blood. In subjects with allergic rhinitis, NO2 exposure (400 ppb
12 for 6 hours) increased eosinophil activation, measured by ECP in nasal lavage, following
13 nasal allergen provocation (Wang et al., 1995a). These studies suggest that exposure to
14 NO2 may prime eosinophils for subsequent activation by an allergen in previously
15 sensitized individuals (Davies etal.. 1997; Wang etal.. 1995b). However, another study
16 found decreased sputum eosinophils 6 hours after HDM challenge in subjects with
17 HDM-sensitive allergic asthma exposed to 400 ppb NO2 for 3 hours (Witten et al., 2005).
18 Late phase allergic responses were also investigated in animal models of allergic airway
19 disease (see also Section 5.2.2.5). Increased specific immune response to HDM allergen,
20 including enhanced antigen-specific serum IgE, and increased lung inflammation were
21 demonstrated in Brown Norway rats sensitized to and challenged with HDM allergen
22 followed by 3-hour exposure to 5,000 ppb NO2 (Gilmour et al.. 1996). Similarly, a recent
23 study showed that NO2 exposure (25,000 ppb, 6 h/day for 3 days) increased the degree
24 and duration of the allergic inflammatory response in mice sensitized and challenged with
25 ovalbumin (Poynter et al., 2006). Both neutrophilic and eosinophilic airway inflammation
26 were found in these studies; exposure of mice to a lower concentration of NO2
27 (5,000 ppb) failed to induce this response. Two other studies in ovalbumin-sensitized and
28 ovalbumin-challenged mice found decreased eosinophilic inflammation in response to
29 5,000 ppb NO2; however, one of these studies found an increase in eosinophils following
30 exposure to 20,000 ppb NO2 (Hubbard et al.. 2002; Proust et al.. 2002). This increase in
31 eosinophils was accompanied by increased levels of eosinophil peroxidase, a marker of
32 activation. Both responses were observed 3 days, but not 1 day, after the 3-hour NO2
33 exposure. These results in animal models provide some evidence of NO2-mediated
34 enhancement of late phase allergic responses, albeit at higher concentrations than those
35 considered environmentally relevant or following repeated exposures. It is important to
36 note that eosinophil activation and eosinophil influx reflect different processes and that
37 only the study by Hubbard et al. (2002) measured markers of activation. The
38 ovalbumin-sensitized and ovalbumin-challenged mouse model may not mimic the
39 eosinophil degranulation and/or cytolysis that are characteristic of asthma and allergic
January 2015 4-44 DRAFT: Do Not Cite or Quote
-------
1 rhinitis in humans (Malm-Erjefalt et al., 2001). Hence species-related differences may
2 account for the differences in results of animal and controlled human exposure studies.
3 Collectively, these studies demonstrate that inhaled NO2 enhanced both early and late
4 phase responses to inhaled allergens in humans with asthma and allergy. Furthermore,
5 exposure to NO2 augmented allergic inflammation in some rodent models of allergic
6 airway disease. These results provide evidence for NO2-induced exacerbation of allergic
7 airway disease in the presence of an allergen challenge. Evidence for NO2-induced
8 airway eosinophilia in the absence of an allergen challenge was described in
9 Section 4.3.2.3. Hence, NO2 exposure may lead to asthma exacerbations by multiple
10 pathways.
Endogenous Nitrogen Dioxide
11 Several recent animal toxicological studies have explored the role of endogenous NO and
12 peroxynitrite, the latter of which decomposes to form NO2, on allergic airway disease in
13 animal models. In one study, upregulating the enzyme endothelial nitric oxide synthase
14 (eNOS; and presumably NO production) decreased airway inflammation, airway
15 remodeling, and airway responsiveness in a mouse model of asthma (Ahmad et al..
16 2011). Asthma phenotype-related features, such as cell infiltrates, mucus hypersecretion,
17 peribronchial collagen, and Th2 cytokines, were also diminished. Further, decreased
18 inducible nitric oxide synthase (iNOS) expression and 3-nitrotyrosine immunostaining in
19 airway epithelium were reported, as were diminished epithelial injury and apoptosis.
20 Because 3-nitrotyrosine is a marker of NO2/peroxynitrite formation, these findings
21 suggest that an increase in NO may have resulted in reduced peroxynitrite. While it is
22 known that NO rapidly reacts with superoxide to form peroxynitrite and that superoxide
23 levels are increased in inflammation, it is also known that an excess of NO will react with
24 peroxynitrite and quench its reactivity. In fact, Stengeretal. (2010) found that high
25 concentrations of inhaled NO (10,000 ppb) prevented the formation of 3-nitrotyrosine in
26 the lungs of neonatal mice exposed to hyperoxia.
27 In a second set of studies, increased levels of the NOS substrate, L-arginine, were found
28 to decrease airway inflammation and airway responsiveness in a guinea pig model of
29 asthma (Maarsingh et al., 2009). Similarly, increased L-arginine levels reduced
30 peroxynitrite formation and airway responsiveness in a mouse model of asthma
31 (Mabalirajan et al. 2010b). Markers of allergic inflammation, such as eosinophilia and
32 Th2 cytokines, markers of oxidative and nitrative stress, and markers of airway
33 remodeling, such as goblet cell metaplasia and subepithelial fibrosis, were also decreased.
34 Further, increased L-arginine levels reduced mitochondrial dysfunction and airway injury
35 (Mabalirajan et al., 2010a). Limitation of L-arginine is known to uncouple NOS enzyme
January 2015 4-45 DRAFT: Do Not Cite or Quote
-------
1 activity, resulting in the production of superoxide in addition to NO. This situation is
2 commonly found in disease models and leads to peroxynitrite formation. Increasing
3 L-arginine availability is a common strategy used to prevent enzyme uncoupling and
4 peroxynitrite formation. Another approach was employed in a study by North et al.
5 (2009) where inhibition of the enzyme arginase 1 (arginase 1 decreases arginine
6 availability) was found to decrease airway responsiveness in a mouse model of asthma.
7 Similar findings were reported using arginase inhibition in a guinea pig model of allergic
8 asthma where arginase was upregulated (Maarsingh et al.. 2008). Inhibition of arginase
9 resulted in amelioration of the asthma phenotype. These effects were attributed to
10 decreased enzyme uncoupling, thus promoting the formation of NO, diminishing the
11 generation of superoxide, and reducing the formation of peroxynitrite. In contrast, a
12 different study found that arginase inhibition resulted in increased S-nitrosylated and
13 nitrated proteins, increased inflammation, mucous cell metaplasia, NFxB activation, and
14 increased airway responsiveness in a mouse model of asthma (Ckless et al.. 2008).
15 However, antigen-specific IgE and IL-4 levels were reduced. Thus, only some features of
16 the asthma phenotype were ameliorated by arginase inhibition. The authors suggested
17 that peroxynitrite, whose presence was indicated by the increase in nitrated proteins in
18 mice treated with arginase, may have contributed to increased airway responsiveness in
19 this model.
20 Evidence for similar pathways in humans is provided by a study in which endogenous
21 markers of reactive nitrogen and oxygen chemistry were measured in individuals with
22 and without asthma (Anderson et al.. 2011). Levels of total nitrite and nitrate were higher
23 in the BAL fluid of subjects with asthma compared to healthy subjects. In subjects with
24 asthma, upregulation of iNOS was observed, and it was greater in distal airways
25 compared with more proximal airways. In addition, levels of dihydroethidium-positive
26 cells, which are capable of producing ROS (such as superoxide), were higher in both the
27 BL fluid and BAL fluid of subjects with asthma compared with healthy subjects. Levels
28 of arginase were also higher in BAL fluid of subjects with asthma compared with healthy
29 subjects. These results suggest that uncoupling of NOS and/or NOS dysfunction,
30 resulting in enhanced peroxynitrite/NO2 formation, may contribute to the asthma
31 phenotype in human subjects. They also provide biological plausibility for results of
32 another study demonstrating a correlation between increased airway responsiveness and
33 the induction of iNOS, the induction of arginase, and the production of superoxide in
34 subjects with asthma.
35 Collectively, these studies provide evidence that the balance between endogenous NO
36 and peroxynitrite influences features of the asthma phenotype in animal models of
37 allergic airway disease and possibly in adults with asthma. Enhanced levels of
38 superoxide, which are characteristic of asthma and other inflammatory states, favor the
January 2015 4-46 DRAFT: Do Not Cite or Quote
-------
1 formation of peroxynitrite at the expense of NO. Evidence from experimental studies
2 indicates that peroxynitrite and other RNS are found in and contribute to allergic airway
3 disease in animal models. Inhaled NO2 may exacerbate allergic airway disease by adding
4 to the lung burden of RNS in inflammatory states.
4.3.2.6.3 T-helper Cell 2 Skewing and Allergic Sensitization
5 A controlled human exposure study demonstrated that repeated daily exposures of
6 healthy adults to NO2 resulted in increased expression of IL-5, IL-10, IL-13, and ICAM-1
7 in respiratory epithelium following the last exposure [(Pathmanathan et al., 2003)
8 Section 5.2.7]. These interleukins are characteristic of a Th2 inflammatory response. IL-5
9 is known to promote eosinophilia, while IL-13 is known to promote mucus production
10 and AHR (Bevelander et al.. 2007). These findings suggest a potential mechanism
11 whereby repeated exposure to NO2 may exert a pro-allergic influence. Further,
12 upregulation of ICAM-1 suggests a potential mechanism for leukocyte influx. A separate
13 study by these same investigators found persistent neutrophilic inflammation following
14 the 4 days of repeated exposure (Blomberg et al.. 1999).
15 In addition, two studies in animals examined the effects of longer term exposures to NO2
16 on the development of allergic responses (Sections 5.2.7 and 6.2.5.2). In one study,
17 exposure of guinea pigs to 3,000 or 9,000 ppb NO2 increased the numbers of eosinophils
18 in nasal epithelium and mucosa after 2 weeks (Ohashi et al.. 1994). In the other, exposure
19 to 4,000 ppb NO2 for 12 weeks led to enhanced IgE-mediated release of histamine from
20 mast cells isolated from guinea pigs (Fujimaki and Nohara. 1994). This response was not
21 found in mast cells from rats similarly exposed. Both studies provide further evidence for
22 NO2 having a pro-allergic influence.
23 Furthermore, a recent study in mice provides evidence that NO2 may act as an adjuvant
24 promoting the development of allergic airway disease in response to a subsequent
25 inhalation exposure to ovalbumin (Bevelander et al., 2007). Findings included AHR,
26 mucous cell metaplasia, and eosinophilic inflammation, as well as ovalbumin-specific
27 IgE and IgGl, CD4+ T-cells biased toward Th2, and a T helper cell 17 (Thl7) phenotype
28 in the blood. These results are consistent with an allergic asthma phenotype in humans.
29 The eosinophilic inflammation, mucus gene upregulation, and ovalbumin-specific IgE
30 production were found to be dependent on TLR2 and myeloid differentiation primary
31 response gene (88) pathways. TLR2 is known to promote maturation of dendritic cells,
32 inflammation, and Th2 skewing. A subsequent study in the same model found that NO2
33 exposure had several effects on pulmonary CD1 lc+ dendritic cells, including increased
34 cytokine production, upregulation of maturation markers, increased antigen uptake,
35 migration to the lung-draining lymph node, and improved ability to stimulate naive CD4+
January 2015 4-47 DRAFT: Do Not Cite or Quote
-------
1 T-cells (Hodgkins et al., 2010). Dendritic cells are key players in adaptive immune
2 responses by regulating CD4+-mediated T cell responses through the presentation of
3 antigens in the draining lymph node. Further, dendritic cells can express a distinct pattern
4 of costimulatory molecules and produce cytokines that create an environment for T cell
5 polarization, thus skewing the T helper cell response. Changes reported in these two
6 studies are consistent with the promotion of allergic sensitization and suggest a role for
7 TLR2 in mediating this effect. A third study by these same investigators found that NO2
8 exposure resulted in antigen-specific IL-17A generation from Thl7 cells, which is
9 characteristic of the severe asthma phenotype that is unresponsive to glucocorticoid
10 treatment in humans (Martin et al.. 2013). Although all studies involved 1-hour exposures
11 to high concentrations of NO2 (10,000-15,000 ppb), they are included here because they
12 describe potentially new mechanisms by which inhaled NO2 may exert its effects. It
13 should additionally be noted that airway inflammation was seen in mice exposed to
14 15,000 ppb, but not to 10,000 ppb, NO2 for 1 hour and that pulmonary damage was
15 minimal in this model (Martin et al., 2013).
16 In contrast, a similar study failed to find that NO2 acted as an adjuvant in a mouse model
17 of allergic airway disease (Alberg et al.. 2011). The exposure consisted of 5,000 or
18 25,000 ppb NO2 for 4 hours and followed exposure to ovalbumin which was administered
19 intranasally. Adjuvant activity was measured as the production of allergen-specific IgE
20 antibodies. Methodological differences in study design with respect to the timing of
21 ovalbumin and NO2 exposures and the route of ovalbumin exposure may account for
22 differences in findings between this study and others. In fact, Bevelander et al. (2007)
23 found that NO2 promoted allergic sensitization when exposure occurred prior (but not
24 subsequent) to ovalbumin.
25 It has been hypothesized that both endogenous and exogenous ROS and/or RNS can alter
26 the balance between tolerance and allergic sensitization due to an inhaled agent (Ckless
27 etal.. 2011). Some activities of dendritic cells and T-cells, such as maturation of the
28 antigen presenting capacity of dendritic cells, dendritic cell stimulation of CD4+ T-cells,
29 and polarization of T-cells, are redox sensitive. Endogenous ROS and RNS are produced
30 by a variety of respiratory tract cells, including epithelial cells, dendritic cells, T
31 lymphocytes, macrophages, neutrophils, and eosinophils, especially during inflammation.
32 Peroxynitrite formation, myeloperoxidase activity and/or nitrite acidification may also be
33 enhanced during inflammation and may contribute to endogenous NO2 levels. ROS and
34 RNS are thought to promote the allergic phenotype. Air pollution-derived exogenous
35 ROS and RNS can potentially contribute to oxidative and/or nitrative stress in the
36 respiratory tract and influence the adaptive immune response that occurs once dendritic
37 cells are activated. Thus, recent studies suggest the possibility of an interaction between
38 inhaled NO2 and NO2 endogenously formed in the respiratory tract.
January 2015 4-48 DRAFT: Do Not Cite or Quote
-------
1 Collectively, these studies in humans and animals provide evidence that NO2 exposure
2 may lead to the development of allergic responses in nonallergic individuals or animals
3 via Th2 skewing and allergic sensitization. Furthermore, they are consistent with a link
4 between exposure to ambient NO2 and increased prevalence of allergic sensitization
5 found in a few epidemiologic studies discussed in Section 6.2.5.
4.3.2.7 Remodeling of Airways and Alveoli
6 As summarized in the 2008 ISA (U.S. EPA. 2008) and the 1993 AQCD (U.S. EPA.
7 1993). numerous studies have examined morphological changes in the respiratory tract
8 resulting from chronic NC>2 exposure. The sites and types of morphological lesions
9 produced by exposure to NC>2 were similar in all species when effective concentrations
10 were used (U.S. EPA. 1993). The centriacinar region, including the terminal conducting
11 airways, the alveolar ducts, and the alveoli, exhibited the greatest sensitivity to NO2
12 exposure, while the nasal cavity was minimally affected. The cells most injured in the
13 centriacinar region were the ciliated cells of the bronchiolar epithelium and the Type I
14 cells of the alveolar epithelium. These were replaced with nonciliated bronchiolar cells
15 and Type II cells, respectively, which were relatively resistant to continued NC>2
16 exposure. Some lesions rapidly resolved post-exposure. One study found that collagen
17 synthesis rates were increased in NC>2-exposed rats. Because collagen is an important
18 structural protein in the lung and because increased total lung collagen is characteristic of
19 pulmonary fibrosis, it was proposed that NC>2 exposure may cause fibrotic-like diseases.
20 Exposure to NC>2 also enhanced pre-existing emphysema-like conditions in animal
21 models (U.S. EPA. 2008). Other studies demonstrated that NO2 exposure induced air
22 space enlargements in the alveolar region and suggested that chronic exposures could
23 result in permanent alterations resembling emphysema-like diseases (U.S. EPA. 1993). A
24 recent study confirmed and extended these findings. NO2 exposure in rats (10,000 ppb for
25 21 days) caused increased apoptosis of alveolar epithelial cells and enlargement of air
26 spaces (Tehrenbach et al.. 2007). Further, alveolar septal cell turnover was increased, and
27 changes in extracellular matrix were reported. However, there was no loss of alveolar
28 walls (i.e., total alveolar wall volume or total alveolar surface area), indicating that the
29 lesions induced did not meet the 1985 National Heart Lung and Blood Institute definition
30 of human emphysema (U.S. EPA. 1993).
31 A chronic study in rats exposed to 9,500 ppb NO2 for 7 h/day, 5 days/week for 24 months
32 found an additional effect on morphology (Mauderly et al.. 1990). Bronchiolar epithelium
33 was observed in centriacinar alveoli, and this response progressed with increasing length
34 of exposure. This has been termed "alveolar bronchiolization" (Nettesheim et al.. 1970).
January 2015 4-49 DRAFT: Do Not Cite or Quote
-------
1 reflecting the replacement of one type of epithelium by another. Long-term consequences
2 of alveolar bronchiolization are not known.
3 The relationship between NC>2 exposure-induced morphologic changes in animal models
4 and impaired lung development seen in epidemiological studies is not clear. Effects of
5 NC>2 exposure on lung morphology in rats have been shown to be age-dependent
6 (U.S. EPA. 2008) (U.S. EPA. 1993). Six-week-old rats exposed to NO2 for 6 weeks were
7 more sensitive to the effects of NO2 exposure than 1-day-old rats exposed for 6 weeks
8 (Chang et al.. 1986). In humans, the respiratory and immune systems are immature in
9 newborns, and the respiratory system continues to develop until about 20 years of age.
10 This suggests the potential for NC>2 exposure-induced permanent morphological changes
11 in humans if exposure should occur during critical windows of development. However,
12 experimental evidence to substantiate this claim is currently lacking.
13 Evidence from animal models of allergic airway disease suggests a role for endogenous
14 NC>2 in airway remodeling. These studies, described above in Section 4.3.2.6.2. found
15 that decreased NO bioavailability during inflammation favored the formation of
16 peroxynitrite, which decomposes to NO2. Interventions that reduced peroxynitrite
17 formation, as evidenced by decreased 3-nitrotyrosine immunostaining, resulted in an
18 amelioration of airway remodeling, as measured by mucus hypersecretion, peribronchial
19 collagen, goblet cell metaplasia, subepithelial fibrosis, and epithelial apoptosis (Ahmad
20 etal.. 2011; Mabalirajan et al.. 2010b). Exposure to inhaled NO2 was found to enhance
21 allergic airway inflammation and airway responsiveness in experimental animals
22 previously sensitized and challenged with allergen (Poynter et al.. 2006). Airway
23 remodeling was not evaluated in this study which involved acute exposures to NO2.
24 Whether repeated or chronic exposures to NO2 lead to airway remodeling in the context
25 of allergic airway disease is not known. However, in nonallergic guinea pigs, subchronic
26 exposure to NO2 (60-4,000 ppb, 24 h/day, 6-12 weeks) enhanced both airway
27 responsiveness and specific airway resistance, suggesting that airway remodeling may
28 have contributed to the development of AHR (Kobavashi and Miura. 1995).
4.3.2.8 Potential Induction of Carcinogenesis
29 Some studies have explored the potential carcinogenicity of NC>2. There is no clear
30 evidence that NO2 acts as a carcinogen [(U.S. EPA. 2008) (U.S. EPA. 1993)
31 Section 6.6.91. However, NC>2 may act as a tumor promoter at the site of contact, possibly
32 due to its ability to produce cellular damage and promote regenerative cell proliferation.
33 In addition, it has been shown to be genotoxic and mutagenic in some systems, including
34 human nasal epithelial mucosa cells ex vivo exposed to urban-level concentrations
January 2015 4-50 DRAFT: Do Not Cite or Quote
-------
1 [100 ppb (Koehler etal.. 2011. 2010)1. Some studies demonstrated that inhaled NO2 at
2 high concentrations (e.g., 20,000 ppb) can contribute to the formation of mutagens and
3 carcinogens if other precursor chemicals are found in the body, e.g., N-nitrosomorpholine
4 from morpholine and nitro-pyrene from pyrene [(U.S. EPA. 2008) Section 4.2.2.5].
4.3.2.9 Transduction of Extrapulmonary Responses
5 While the respiratory tract has been viewed as the primary target of inhaled NC>2, effects
6 outside the respiratory tract have been demonstrated in numerous controlled human
7 exposure and toxicological studies (U.S. EPA. 2008) (U.S. EPA, 1993). These include
8 hematological effects and effects on the heart, central nervous system, liver, and kidneys
9 and on reproduction and development. Epidemiologic evidence of associations between
10 NC>2 exposure and some extrapulmonary effects has also been described (Sections 5.3.
11 6.3. and 6.4).
12 Some NO2-induced effects, which have been demonstrated, are briefly described here.
13 Two controlled human exposure studies involving NC>2 inhalation over several hours
14 found effects on circulating red blood cells, including reduced hemoglobin and
15 hematocrit levels; one of these also found reduced acetylcholinesterase activity
16 [(Frampton et al., 2002; Posin et al., 1978) Section 5.3.111. Changes in lymphocyte
17 numbers and subsets in the peripheral blood have been demonstrated in human subjects
18 following exposure to NC>2 (Frampton et al., 2002; Sandstrom et al., 1992). A recent
19 controlled human exposure study found altered blood lipids (Huang etal.. 2012). Studies
20 in experimental animals have demonstrated decreases in red blood cell number as well as
21 increases in diphosphoglycerate, sialic acid, and methemoglobin following several days
22 of NC>2 exposure (Section 5.3.11). However, changes in hematocrit and hemoglobin did
23 not occur following longer term exposure to NC>2. Increases in blood glutathione levels
24 and altered blood lipids resulting from NCh exposure have also been reported (U.S. EPA.
25 2008). More recent studies in rats exposed for 7 days to NCh (2,660 or 5,320 ppb NCh)
26 have shown mild pathology of brain and heart tissue, which was accompanied by markers
27 of inflammation and/or oxidative stress [(Li et al.. 2012; Li etal.. 2011) Sections 5.3.11
28 and 6.4.4]. In addition, animal studies demonstrated reproductive and developmental
29 effects resulting from exposure to NC>2 during gestation. These included decreased litter
30 size and neonatal weight, and effects on post-natal development (Section 6.4). Many, but
31 not all, of these extrapulmonary effects in animal models have been observed at
32 concentrations of NC>2 that are higher than ambient-relevant concentrations.
33 Given the reactivity of NC>2, extrapulmonary effects are likely due to NC>2-derived
34 reaction products rather than to NCh itself. One pathway by which a reaction product
January 2015 4-51 DRAFT: Do Not Cite or Quote
-------
1 could mediate extrapulmonary effects of NO2 would be the activation of pulmonary
2 irritant receptors that results in cardiovascular reflex responses (Section 4.3.2.2).
3 Evidence suggests that the reduction in heart rate observed after acute exposure of
4 experimental animals to high concentrations of NO2 may be due to stimulation of
5 pulmonary irritant receptors. However, much weaker evidence exists for activation of
6 pulmonary irritant receptors in humans because studies observed no increases in
7 respiratory rate or decreases in heart rate. A recent controlled human exposure study
8 found altered heart rate variability following exposure to an ambient-relevant
9 concentration of NO2 [500 ppb (Huang et al.. 2012)1; whether this effect was due to
10 pulmonary irritant receptor stimulation is unclear.
11 Alternatively, NO2-derived reaction products in the lung may "spillover" into the
12 circulation. One reaction product of inhaled NO2, nitrite, is known to gain access to the
13 circulation. In the presence of red blood cell hemoglobin, nitrite is oxidized to nitrate
14 (Postlethwait and Mustafa. 1981). and nitrosylhemoglobin and methemoglobin are
15 formed. Nitrite has known effects on blood cells, vascular cells, and other tissues. Much
16 recent attention has been paid to nitrite's systemic vasodilatory effects that occur under
17 hypoxic conditions. As discussed in the 2008 ISA and the 1993 AQCD (U.S. EPA. 2008)
18 (U.S. EPA. 1993). one controlled human exposure study demonstrated that NO2 exposure
19 (4,000 ppb, 75 minutes, intermittent exercise) resulted in a reduction in blood pressure
20 (Linnet al.. 1985). which is consistent with the systemic vasodilatory properties of nitrite
21 under conditions of low oxygen. However, studies from other laboratories did not see this
22 effect (Section 5.3.6.2). Furthermore, dosimetric considerations suggest that contributions
23 of nitrite derived from ambient NO2 to plasma levels of nitrite are small compared to
24 nitrite derived from dietary sources (Section 4.2.2.4).
25 Besides nitrite and nitrate, other NO2-derived reaction products may potentially
26 translocate to the circulation. The formation of fatty acid epoxides followed by transport
27 to the circulation and then to the liver was postulated to explain the effect of NO2
28 exposure (250 ppb, 3 hours) on pentobarbital-induced sleeping time in mice (Miller et al..
29 1980). Findings of lipid peroxidation and markers of oxidative stress in some animal
30 studies (Li et al.. 2012; Li et al.. 2011). which utilized much higher concentrations of
31 NO2 than the study by Miller etal. (1980). also suggest the presence of circulating ROS
32 or RNS. However, there is no experimental evidence to date for the translocation of
33 NO2-derived ROS and/or RNS to the circulation following NO2 exposure.
34 A third pathway by which a NO2-derived reaction product may transduce extrapulmonary
35 responses is the "spillover" of inflammatory or vasoactive mediators from the lung into
36 the circulation. This possibility is consistent with changes in peripheral blood
37 inflammatory cells and in tissue markers of inflammation that have been observed
January 2015 4-52 DRAFT: Do Not Cite or Quote
-------
1 following exposure to NO2. Confirmation that this mechanism occurs in human subjects
2 exposed to ambient-relevant concentrations of NO2 was provided by a recent study
3 [(Channell et al.. 2012) Section 5.3.111. Exposure of healthy human subjects to NO2
4 (500 ppb for 2 hours) resulted in circulating pro-inflammatory factors in the plasma.
5 Application of plasma to cultured endothelial cells resulted in upregulation of ICAM-1
6 and vascular cell adhesion molecule 1, as well as the release of IL-8 into the supernatant
7 of the cultured cells. Furthermore, the amount of soluble lectin-like receptor for oxidized
8 low-density lipoprotein was increased in plasma obtained 24 hours post-exposure.
9 Changes in plasma high density lipoprotein levels were observed in a separate study
10 employing the same exposure parameters (Huang etal.. 2012). These findings point to a
11 pathway by which inhaled NO2 leads to circulating soluble factors that promote
12 inflammatory signaling in the vasculature.
4.3.3 Nitric Oxide
13 As summarized in the 2008 ISA, the 1993 AQCD (U.S. EPA. 2008) (U.S. EPA, 1993).
14 and a recent review (Hill et al.. 2010). the synthesis of endogenous NO in cells is
15 catalyzed by three different isoforms of NOS (eNOS, iNOS, neuronal NOS). NO is
16 involved in intracellular signaling in virtually every cell and tissue. In general, low levels
17 of endogenous NO play important roles in cellular homeostasis, while higher levels are
18 important in cellular adaptation and still higher levels are cytotoxic. Further, signaling
19 functions of NO may be altered in the presence of acute inflammation (Hill et al.. 2010).
20 Like NO2, NO is a radical species (Fukuto et al.. 2012). However, it is more selectively
21 reactive than NO2 (Hill etal.. 2010). In addition, it is more hydrophobic and can more
22 easily cross cell membranes and diffuse much greater distances compared with NO2. As a
23 result, there may be overlap between endogenous and exogenous NO in terms of
24 biological targets and pathways. The following discussion focuses on common
25 mechanisms underlying the effects of both endogenous and exogenous NO.
26 Because NO has a high affinity for heme-bound iron, many of its actions are related to its
27 interactions with heme proteins (Hill etal.. 2010). For example, activation of the heme
28 protein guanylate cyclase is responsible for smooth muscle relaxation and vasodilation of
29 pulmonary and systemic vessels, and possibly for bronchodilator effects. Inhaled NO
30 rapidly reacts with soluble guanylate cyclase in the pulmonary arterial smooth muscle. At
31 the same time, inhaled NO rapidly diffuses into the circulation and reacts with red blood
32 cell hemoglobin to form nitrosylhemoglobin, which is subsequently oxidized to
33 methemoglobin and nitrate. Increased blood concentrations of nitrosylhemoglobin and
34 methemoglobin have been reported in mice exposed for 1 hour to 20,000-40,000 ppb
January 2015 4-53 DRAFT: Do Not Cite or Quote
-------
1 NO, as well as in mice exposed chronically to 2,400 and 10,000 ppb NO (U.S. EPA.
2 1993). Some S-nitrosohemoglobin may be formed in partially deoxygenated blood
3 (Wennmalm et al.. 1993). NO can also disrupt iron-sulfur centers in proteins (Hill et al..
4 2010). Furthermore, redox reactions of NO and transition metals, such as iron and
5 copper, facilitate S-nitrosylation of protein and nonprotein thiols. Binding of NO to iron-
6 and copper-containing proteins in the mitochondria may play an important role in
7 mitochondrial respiration. NO also rapidly reacts with superoxide, an oxygen-derived
8 radical species, to produce the potent oxidant peroxynitrite (Hill et al.. 2010).
9 Peroxynitrite subsequently reacts with CO2 to form the nitrosoperoxylcarbonate anion,
10 followed by decomposition to carbonate radical and NO2 (Section 4.2.2.4).
11 Endogenous NO is formed in the respiratory tract at high levels (Section 4.2.3). and it has
12 physiologic functions. The paranasal sinuses are a major source of NO in air derived
13 from the nasal airways, with average levels of 9,100 ppb NO (n = 5) measured in the
14 sinuses (Lundberg et al.. 1995). Expression of iNOS was found to be higher in epithelial
15 cells of the paranasal sinuses than in epithelial cells of the nasal cavity. This NO
16 produced by nasal airways is thought to play a role in sinus host defense through
17 bacteriostatic activity. In addition, NO produced by nasal airways was found to modulate
18 pulmonary function in humans through effects on pulmonary vascular tone and blood
19 flow (Lundberg et al.. 1996). In healthy subjects, a comparison of nasal and oral
20 breathing demonstrated that nasal airway NO enhanced transcutaneous oxygen tension.
21 In intubated patients, nasal airway NO increased arterial oxygenation and decreased
22 pulmonary vascular resistance. Additionally, endogenous NO has been shown to act as a
23 bronchodilator (Belvisi etal.. 1992). Endogenous NO produced at high concentrations by
24 phagocytic cells is also known to participate in the killing of bacteria and parasites; this
25 contributes to host defense (U.S. EPA. 2008). Another effect of endogenous NO on host
26 defense is modulation of ciliary beat frequency (Jain etal.. 1993). Specifically, NO
27 derived from more distal airways was found to increase ciliary beat frequency.
28 Furthermore, endogenous NO production can be upregulated during inflammation
29 (Anderson et al.. 2011). In fact, induction of iNOS in proximal or distal airways of
30 subjects with asthma results in levels of NO in exhaled breath as high as 20-50 ppb
31 (Alving et al.. 1993; Hamid et al.. 1993).
32 Endogenous NO has known pro- and anti-inflammatory effects; thus its role in
33 inflammatory lung disease is not clear. While both eNOS and iNOS contribute to NO
34 production in the lung, the relatively low levels of NO produced by eNOS are thought to
35 be more important in metabolic homeostasis (Ahmad etal.. 2011). Some evidence points
36 to a role of iNOS-derived NO in the pathogenesis of asthma because it has been
37 correlated with inflammation, epithelial injury, and clinical exacerbations of asthma
38 [(Anderson et al.. 2011) Section 4.3.2.6.21. Furthermore, preferential iNOS upregulation
January 2015 4-54 DRAFT: Do Not Cite or Quote
-------
1 was found in the distal airways compared with more proximal airways in subjects with
2 asthma. This is of interest because asthma is a disease of the small airways. As mentioned
3 above, signaling functions of NO may be altered in the presence of acute inflammation
4 (Hill et al.. 2010). which is characterized by enhanced levels of superoxide. Superoxide
5 reacts with NO to form peroxynitrite, which has been shown in animal models to play a
6 role in the pathogenesis of allergic airway disease (Section 4.3.2.6.2).
7 NO exposure has been shown to alter pulmonary function, morphology, and vascular
8 function (U.S. EPA. 2008) (U.S. EPA. 1993). Studies in animals have demonstrated that
9 inhaled NO reversed acute methacholine-induced bronchoconstriction (Hogman et al..
10 1993; Dupuyetal.. 1992). This was observed with exposures of 5,000 ppb NO in guinea
11 pigs and 80,000 ppb in rabbits. Chronic inhalation exposures have been found to alter the
12 morphology of the alveolar septal units in rats (Mercer et al.. 1995). This effect was not
13 seen with chronic inhalation exposures to NO2 at similar concentrations (500 ppb with
14 twice daily spikes of 1,500 ppb). In addition, inhaled NO has been shown to alter
15 transferrin and red blood cells in mice. Further, acute inhalation exposure of NO
16 decreased pulmonary vascular resistance in pigs and reduced pulmonary arterial pressure
17 in a rodent model of chronic pulmonary hypertension. A recent study also found that
18 inhaled NO (1,000, 5,000, 20,000, and 80,000 ppb) selectively dilated pulmonary blood
19 vessels, improved ventilation-perfusion mismatch, and reduced hypoxemia-induced
20 pulmonary vascular resistance in a pig model (Lovich et al.. 2011).
21 Inhaled NO is used clinically at concentrations higher than those that are environmentally
22 relevant. Although it can cause both pulmonary and systemic vasodilation, effects on
23 pulmonary vasculature occur at lower concentrations than those required for vasodilation
24 of systemic vessels. This selectivity for pulmonary vasculature is likely due to the rapid
25 scavenging of NO by hemoglobin in the blood. Hence, inhaled NO has been used to
26 mitigate pulmonary hypertension in newborns and adults. High concentrations of inhaled
27 NO are also known to alter ciliary beating and mucus secretion in the airways, to increase
28 renal output, to alter distribution of systemic blood flow, to alter coagulation, fibrinolysis,
29 and platelet functions, and to modulate the inflammatory response (U.S. EPA. 2008).
30 Endogenous NO is an important mediator of cardiovascular homeostasis. It has
31 anti-inflammatory and antithrombotic effects, is cytoprotective, and induces antioxidant
32 defenses (Wang and Widlansky. 2009). Two recent studies in animal models demonstrate
33 that high concentrations of inhaled NO may result in vascular toxicity. One of these
34 studies found rapid formation of plasma nitrites/nitrates in rats exposed for 1 hour to
35 3,000-10,000 ppb NO (Knuckles etal.. 2011). Plasma nitrites/nitrates doubled after an
36 hour of exposure to 3,000 ppb NO and tripled after an hour of exposure to 10,000 ppb
37 NO. These changes were accompanied by an enhanced constriction response to
January 2015 4-55 DRAFT: Do Not Cite or Quote
-------
1 endothelin-1 in coronary arterioles, which reflected altered vasomotor tone. Although this
2 latter effect appears to run counter to the vasodilator role of NO, it should be noted that
3 high concentrations of NO, as were used in this study, are known to inhibit eNOS activity
4 in other models (Griscavage etal.. 1995). The increase in aortic eNOS content reported is
5 consistent with enzyme inactivation and turnover. Another recent animal toxicological
6 study conducted in ApoE~'~ mice, a model of atherosclerosis, found that exposure to very
7 high concentrations of inhaled NO over the course of a week (17,000 ppb NO for 6 h/day
8 for 7 days) led to increases in messenger ribonucleic acid for aortic endothelin-1 and
9 matrix metalloproteinase (MMP)-9, as well as to enhanced vascular gelatinase activity
10 (Campen et al.. 2010). These effects, which are biomarkers of vascular remodeling and
11 plaque vulnerability, were not seen with 2,000 ppb NO2. The authors suggested that the
12 activity of eNOS was uncoupled, resulting in oxidative stress due to the production of
13 superoxide instead of or in addition to NO. Both of these studies suggest that inhaled NO
14 has the potential to disrupt normal signaling processes mediated by endogenous NO.
15 As mentioned above, endogenous NO plays key signaling roles in virtually every cell and
16 tissue (Hill etal.. 2010) and, as such, is an important mediator of homeostasis. Inhaled
17 NO at high concentrations has the potential to have beneficial or deleterious effects on
18 multiple organ systems. An important consideration is whether effects are mediated by an
19 NO metabolite, by the release of NO from a metabolite that serves as a storage pool of
20 NO, or through methemoglobin formation in the blood. Further discussion of the
21 biological functions of NO metabolites is found below.
4.3.4 Metabolites of Nitric Oxide and Nitrogen Dioxide
4.3.4.1 Nitrites/Nitrates
22 Rapid appearance of nitrite and nitrate in the blood was demonstrated in rats exposed for
23 1-2 hours to 5,000-40,000 ppb NO2 (OdaetaL 1981). Elevated levels of blood nitrite
24 and nitrate were maintained as long as the exposure to NO2 continued. A small increase
25 in levels of nitrosylhemoglobin, but not methemoglobin, was detected in blood. The lack
26 of accumulation of methemoglobin was likely due to reduction of methemoglobin to
27 hemoglobin catalyzed by methemoglobin reductase. Two other studies measured
28 methemoglobin in the blood of mice exposed to NO2, with conflicting results (U.S. EPA.
29 1993). Rapid formation of plasma nitrites/nitrates has also been demonstrated in rats
30 exposed for 1 hour to 3,000-10,000 ppb NO (Knuckles etal.. 2011).
31 Recently, it has been proposed that nitrite is a storage form of NO because it can be
32 reduced back to NO under conditions of low oxygen tension in a reaction catalyzed by
January 2015 4-56 DRAFT: Do Not Cite or Quote
-------
1 deoxyhemoglobin (Gladwin et al., 2005). In addition, nitrite is a signaling molecule in its
2 own right and does not require conversion to NO for this activity (Bryan. 2006). Nitrite
3 can increase cyclic guanosine monophosphate (cGMP) levels and heat shock protein 20
4 expression, decrease cytochrome P450 activity and alter heme oxygenase-1 expression
5 (Bryan et al., 2005). Nitrite is also bactericidal (Major et al., 2010). Furthermore, under
6 acidic conditions, nitrite can react with thiols to form RSNOs. Nitrite also reacts with
7 hemoglobin to form iron-nitrosyl-hemoglobin and with oxyhemoglobin to form nitrate.
8 Nitrite acts as a vasodilator under hypoxic conditions, through a reaction catalyzed by
9 deoxyhemoglobin (Cosby et al., 2003). The venous circulation may be more sensitive to
10 nitrite than the arterial circulation (Maher et al.. 2008).
11 A recent study found that inhaled nitrite decreased pulmonary blood pressure in newborn
12 lambs with hemolysis-induced pulmonary vasoconstriction (Blood etal.. 2011). Nitrite
13 was converted to NO in lung tissue by a mechanism that did not require reaction with
14 deoxyhemoglobin in the circulation. This mechanism resulted in increased exhaled NO
15 gas as well as the relaxation of vascular smooth muscle, which led to pulmonary
16 vasodilation. Although concentrations of inhaled nitrite employed were high (0.87 mol/L
17 sodium nitrite), this study is discussed here because it illustrates a novel biological
18 activity of lung nitrite that is normally formed by reactions of NO2 and NO in the ELF
19 and/or the blood.
4.3.4.2 S-Nitrosothiols
20 RSNOs are found endogenously in tissues and extracellular fluids. High concentrations of
21 RSNOs are found in the lung, and their levels may vary depending on disease status (Que
22 et al.. 2009). For example, levels of RSNOs in BAL fluid were higher in individuals with
23 asthma compared with healthy subjects (Que et al.. 2009). Transport of RSNOs from
24 extracellular compartments into isolated perfused lungs and cultured alveolar epithelial
25 cells occurs via a specific amino acid transport pathway (Torok et al.. 2012; Brahmajothi
26 etal.. 2010).
27 Some S-nitrosohemoglobin may be formed in partially deoxygenated blood following
28 inhalation of NO ("Wennmahn et al.. 1993). However, inhaled NO mainly reacts with red
29 blood cell hemoglobin to form nitrosylhemoglobin, which is subsequently oxidized to
30 methemoglobin and nitrate (Hill et al.. 2010). The exact mechanisms by which RSNO
31 formation occurs are not completely clear (Fukuto et al., 2012). NO does not react
32 directly with thiol groups, but it can form RSNOs via reactions with thiyl groups and
33 through intermediate formation of N2Os or metal nitrosyls, such as nitrosylhemoglobin
34 (Fukuto et al.. 2012; Hill etal.. 2010). Recent evidence suggests that NO may diffuse into
January 2015 4-57 DRAFT: Do Not Cite or Quote
-------
1 extracellular fluid and be transformed to RSNOs (Torok et al.. 2012; Brahmajothi et al..
2 2010). These experiments were conducted ex vivo in isolated perfused lungs and in vitro
3 in cultured lung epithelial cells, neither of which is a blood-perfused system. Hence it is
4 not clear whether this mechanism contributes to RSNO formation in vivo where the
5 majority of inhaled NO diffuses rapidly across the alveolar capillary barrier and binds to
6 hemoglobin.
7 RSNOs are thought to serve as a storage or delivery form of NO and to play a role in cell
8 signaling (Fukuto et al.. 2012; Hill etal.. 2010). They may mediate protein
9 S-glutathionylation and thiol oxidation reactions that can act as redox switches to initiate
10 cell signaling events or alter enzyme activity (Hill et al.. 2010).
11 In the lung, RSNOs act as endogenous bronchodilators (Que et al.. 2009) and suppress
12 inflammation by decreasing activation of the transcription factor NFxB (Marshall and
13 Stamler. 2001). Furthermore, augmentation of airway RNSOs by ethyl nitrite inhalation
14 protected against lipopolysaccharide-induced lung injury in an animal model (Marshall
15 etal.. 2009). Several findings suggest an inverse relationship between endogenous airway
16 RSNO levels and airway responsiveness. First, levels of airway S-nitrosoglutathione
17 levels were decreased in children with asthmatic respiratory failure and in adults with
18 asthma (Que et al.. 2009; Gaston et al.. 1998). Second, the enzyme nitrosoglutathione
19 reductase (GSNOR), which regulates airway S-nitrosoglutathione content, was expressed
20 at higher levels in BAL cell lysates in human subjects with asthma than in healthy
21 subjects (Que et al.. 2009). GSNOR expression was inversely correlated with
22 S-nitrosoglutathione content. In addition, GSNOR activity in BAL fluid was increased
23 and was inversely correlated with airway responsiveness in human asthma (Que et al..
24 2009). Third, levels of airway RSNOs were inversely correlated with airway
25 responsiveness in human subjects with eosinophilic inflammation (Lee et al.. 2011).
4.3.4.3 Nitrated Fatty Acids and Lipids
26 Nitration of unsaturated fatty acids and lipids can occur during inflammation and
27 ischemia/reperfusion by reactions with NO and nitrite-derived species. (Higdon et al..
28 2012; Khoo et al.. 2010). However, there is no firm evidence that these reactions occur
29 following exposure to inhaled NO2. Nitrated fatty acids (also known as nitro-fatty acids)
30 can release NO, which stimulates vascular smooth muscle relaxation through
31 cGMP-dependent pathways in vitro (Limaet al.. 2005). However, most of the cell
32 signaling effects of nitrated fatty acids in vivo are likely due to post-translational
33 modification of proteins (Khoo etal.. 2010). These electrophilic species react with
January 2015 4-58 DRAFT: Do Not Cite or Quote
-------
1 susceptible thiol groups in transcription factors (Higdon et al.. 2012; Bonacci et al..
2 2011).
3 Nitro-fatty acids, such as nitro-oleic acid and nitro-linoleic acid, are anti-inflammatory
4 (Bonacci etal.. 2011) and vasculoprotective (Khoo etal.. 2010). These effects are
5 mediated via activation of the peroxisome proliferator-activated receptor gamma
6 (PPARy) and antioxidant response element (ARE) pathways and suppression of NFxB
7 and signal transducer and activator of transcription 1 pathways (Bonacci etal.. 2011). In
8 a mouse model, nitro-oleic acid upregulated vascular eNOS and heme oxygenase-1 and
9 inhibited angiotensin Il-induced hypertension (Khoo etal.. 2010; Zhang et al.. 2010a).
10 Nitro-oleic acid protected against ischemia/reperfusion injury in a mouse model (Rudolph
11 etal.. 2010). Nitro-oleic acid also activated MMPs (a pro-inflammatory effect) through
12 thiol alkylation in vitro and inhibited MMP expression in macrophages through activation
13 of PPARy (Bonacci et al.. 2011). Expression of MMP was also suppressed in a mouse
14 model of atherosclerosis.
4.3.4.4 Nitrated Amino Acids and Proteins
15 Peroxynitrite and NO2 can react with amino acids to produce nitrated amino acids and
16 proteins (Hill etal.. 2010). These products can also be formed from nitrite and peroxide
17 in a reaction catalyzed by myeloperoxidase. Nitration of proteins may cause inhibition of
18 protein function and/or induce antigenicity. Specific antibodies formed against nitrated
19 proteins may potentially trigger immune reactions (Daiber and Muenzel. 2012). The
20 presence of nitrated amino acids, such as 3-nitrotyrosine, in cells or tissues is an indicator
21 of NO2 and/or peroxynitrite formation. A recent study reported formation of nitrated
22 SP-D resulting from in vivo exposure to 20,000 ppb NO2 (Matalon et al.. 2009). This
23 modification was accompanied by cross-linking and loss of aggregating activity.
4.3.5 Mode of Action Framework
24 This section describes the key events, endpoints, and outcomes that comprise the modes
25 of action of inhaled NO2 and NO. Biological pathways discussed above that may
26 contribute to health effects resulting from short-term and long-term exposures to NO2 and
27 NO (Chapters 5 and 6) are summarized as a part of this analysis.
28 Because inhalation of NO2 results in redox reactions in the respiratory tract, the initiating
29 event in the development of respiratory effects is the formation of oxidation and/or
30 nitration products in the ELF and possibly in airway or alveolar epithelium (Figure 4-1).
31 Reactive intermediates thus formed are responsible for a variety of downstream key
January 2015 4-59 DRAFT: Do Not Cite or Quote
-------
1
2
o
6
4
5
6
events, which may include respiratory tract inflammation and/or oxidative stress,
impaired epithelial barrier function, altered mucociliary clearance, activation and/or
sensitization of neural reflexes, mast cell degranulation, and increased allergic responses.
These key events may lead to several endpoints including bronchoconstriction, AHR, and
impaired host defenses. The resulting outcomes of short-term NC>2 exposure may thus be
asthma exacerbation (Section 5.2.2) and respiratory tract infections (Section 5.2.5).
Mast cell
degranulation
Redox reactions in
respiratory tract
ELF and tissue
Formation of
oxidation/nitration
products
Legend
Pollutant
• Key Events
| Endpoints
B Outcomes
— Pathway
••• Potential Pathway
Activation/
sensitization
neural reflexes
Impaired epithelial
barrier function
Alveolar macrophage
function
^Mucociliary clearance
Source: National Center for Environmental Assessment.
Bronchoconstriction
Inflammation/
oxidative stress
Airway
hyperresponsiveness
Impaired
host defense
Asthma
exacerbation
Respiratory
tract
infections
Figure 4-1 Mode of action of inhaled nitrogen dioxide (NO2): short-term
exposure and respiratory effects.
9
10
11
12
13
14
15
16
The strongest evidence for this mode of action comes from controlled human exposure
studies. NCh exposure resulted in enhanced inflammatory mediators (e.g., eicosanoids,
interleukins) and neutrophils in BL and/or BAL fluid of healthy subjects. Repeated
exposure of healthy subjects led to increased albumin levels in BL fluid, suggesting
impaired epithelial barrier function. In addition, repeated exposure of subjects with
asthma to NO2 enhanced eosinophils and a biomarker of eosinophil activation in sputum.
Increased airway resistance was demonstrated following NO2 exposure in healthy human
subjects; this response did not involve vagally mediated neural reflexes. However, there
appeared to be a role for mast cell degranulation. NCh exposure also enhanced airway
responsiveness to nonspecific challenges, especially in subjects with asthma. Antioxidant
January 2015
4-60
DRAFT: Do Not Cite or Quote
-------
1 supplementation dampened NO2 exposure-induced lipid peroxidation and airway
2 responsiveness, providing support for in vitro findings implicating redox reactions in the
3 ELF. In addition, both early and late asthmatic responses to an allergen challenge (e.g.,
4 AHR, neutrophil, and eosinophil activation) were enhanced by NC>2 exposure.
5 NC>2-induced impairment of ciliary function and alveolar macrophage phagocytic activity
6 suggested impairment of host defenses.
7 Experimental studies in animals suggest that vagally mediated neural reflexes, mast cell
8 degranulation, and production of eicosanoids, which may sensitize receptors on nerve
9 fibers and signal the influx of neutrophils, may contribute to NC>2 exposure-induced
10 AHR. Exposure to NC>2 also enhanced allergic responses (e.g., IgE, eosinophilic, and
11 neutrophilic inflammation). Nitration of the collectin protein SP-D and inhibition of its
12 aggregating activity were also observed in NO2-exposed animals. This may potentially
13 impact microbial clearance and surfactant metabolism. NO2 exposure-induced alteration
14 of mucociliary clearance and alveolar macrophages also suggested impairment of host
15 defenses.
16 Furthermore, there is some evidence for enhanced endogenous formation of peroxynitrite,
17 which decomposes to NC>2, in both human subjects with asthma and animal models of
18 allergic airway disease. In experimental animals, endogenous peroxynitrite/NCh
19 formation was associated with AHR and allergic inflammatory responses. Reduction of
20 peroxynitrite formation lessened airway responsiveness, allergic inflammation, and
21 airway remodeling. These findings raise the possibility that inhaled NC>2 can add to the
22 lung burden of endogenous NCh which is found in and contributes to AHR and allergic
23 airway disease.
24 The initiating events in the development of respiratory effects due to long-term NC>2
25 exposure are recurrent and/or chronic inflammation and oxidative stress (Figure 4-2).
26 These are the driving factors for potential downstream key events, allergic sensitization,
27 and airway remodeling, which may lead to the endpoint AHR. The resulting outcome
28 may thus be new asthma onset, which presents as an asthma exacerbation that leads to
29 physician-diagnosed asthma (Section 6.2.2.1).
January 2015 4-61 DRAFT: Do Not Cite or Quote
-------
Recurrent/chronic
Inflammation and
oxidative stress
Allergic
sensitization
Legend
H Pollutant
• Key Events
• Endpoints
• Outcomes
— Pathway
••• Potential Pathway
Source: National Center for Environmental Assessment.
Airway
hyperresponsiveness
trigger
New onset asthma/
Asthma exacerbation
Figure 4-2 Mode of action of inhaled nitrogen dioxide (NO2): long-term
exposure and respiratory effects.
i
2
o
J
4
5
6
7
10
11
12
13
14
15
16
17
18
19
20
21
22
The strongest evidence for this mode of action in humans comes from controlled human
exposure studies involving repeated exposures of healthy subjects to NO2 over several
days. Findings included upregulation of Th2 cytokines in respiratory epithelium, which is
characteristic of allergic skewing and part of the allergic sensitization pathway. In
addition, persistent airway neutrophilia and upregulation of 1C AM-1 in airway epithelium
were observed. Reductions in lymphocyte subpopulations, suggesting impaired host
defense, also occurred. Although these were short-term exposure studies, findings
suggest that cumulative effects may occur over time and the possibility that chronic or
recurrent exposure to NCh may lead to the development of asthma.
Studies in experimental animals exposed to NO2 for several weeks found nasal
eosinophilia and enhanced mast cell responses. Other evidence suggests that endogenous
NO2 acts as an adjuvant promoting the development of allergic airway disease in
response to an inhaled allergen. This is consistent with mechanistic studies that suggest
that allergic sensitization involves several redox-sensitive steps and that ROS and/or RNS
promote the development of an allergic phenotype.
Findings that reduction of endogenous peroxynitrite production decreased airway
remodeling in animal models of allergic airway disease suggest that endogenous NO2
may contribute to airway remodeling. Subchronic exposure to NCh enhanced both airway
responsiveness and specific airway resistance, suggesting that airway remodeling may
have contributed to the development of AHR in this nonallergic animal model. Thus,
evidence points to the possibility that inhaled NO2 can add to the lung burden of
endogenous NO that contributes to airway remodeling. Mechanistic studies indicate that
January 2015
4-62
DRAFT: Do Not Cite or Quote
-------
1
2
o
6
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
inflammatory mediators and structural changes occurring due to airway remodeling can
alter the contractility of airway smooth muscle. Thus, persistent inflammation, allergic
sensitization, and airway remodeling due to enhanced endogenous NO2 production or to
long-term NC>2 exposure may contribute to the development of AHR.
There is more uncertainty regarding the mode of action for extrapulmonary effects of
inhaled NO2 (Figure 4-3). However, evidence suggests the following. The initiating
events occur in the respiratory tract, where redox reactions lead to the formation of
oxidation and/or nitration products. Reactive intermediates thus formed are responsible
for downstream key events, which may include activation and/or sensitization of neural
reflexes, spillover of reactive intermediates into the circulation, and respiratory tract
inflammation and/or oxidative stress. This latter key event may lead to the spillover of
inflammatory mediators into the circulation. Circulating reactive intermediates or
inflammatory mediators may potentially result in systemic inflammation and/or oxidative
stress, which may mediate distal effects in other organs. Alternatively, circulating soluble
factors may result in vascular activation, which may lead to the endpoint vascular
dysfunction. Activation of neural pathways could lead to the endpoint altered autonomic
tone. The resulting outcomes might include cardiovascular effects or other organ effects
(Sections 5.3. 6.3. and 6.4).
Redox reactions in
respiratory tract
ELF and tissue:
formation of
oxidation/nitration
products
Legend
Pollutant
• Key Events
• Endpoints
• Outcomes
— Pathway
••• Potential Pathway
Respiratory tract
inflammation/
oxidative stress
Vascular
dysfunction
Cardiovascular
effects
Source: National Center for Environmental Assessment.
Figure 4-3 Mode of action of inhaled nitrogen dioxide (NO2): short-term and
long-term exposure and extrapulmonary effects.
January 2015
4-63
DRAFT: Do Not Cite or Quote
-------
1 The strongest evidence for this mode of action in humans comes from recent controlled
2 human exposure studies. One study found altered heart rate variability, which is a
3 measure of autonomic tone, and altered blood lipids. Whether altered heart rate
4 variability was due to stimulation of pulmonary irritant receptors is unclear because no
5 studies in humans exposed to NO2 have observed increases in respiratory rate or
6 decreases in heart rate. The other study found that plasma from human subjects exposed
7 to NO2 contained increased levels of a soluble factor (soluble lectin-like oxidized low
8 density lipoprotein receptor-1) compared with plasma from control subjects. In addition,
9 this plasma stimulated vascular activation in an in vitro assay. These results indicate that
10 spillover of a reactive intermediate or inflammatory mediator into the circulation
11 occurred, which may transduce a downstream effect in the vasculature or in other organs.
12 This possibility is consistent with changes in peripheral blood lymphocyte number and
13 subsets, as well as with altered blood lipids, which have been observed in humans
14 following exposure to NO2. These findings point to a pathway by which inhaled NO2
15 leads to circulating soluble factors that promote inflammatory signaling in the vasculature
16 and/or other organs.
17 In experimental animal studies, findings of altered blood glutathione levels and lipids,
18 decreased pentobarbital sleeping time, and mild pathology in brain and heart, which was
19 accompanied by tissue markers of oxidative stress and inflammation, are consistent with
20 the possibility that exposure to NO2 results in circulating soluble factors that promote
21 inflammatory signaling and/or oxidative stress. There is also support for NO2
22 exposure-induced cardiovascular reflexes because one study showed bradycardia in
23 experimental animals that were exposed to high concentrations of NO2. Evidence
24 indicated that this response in experimental animals was mediated by pulmonary irritant
25 receptors and the vagus nerve, which is consistent with NO2-induced changes in
26 respiratory rate demonstrated in several studies.
27 Because NO has a high affinity for heme proteins and because there is no barrier to its
28 diffusion across membranes, it rapidly crosses cell membranes and binds to heme
29 proteins (Figure 4-4). For inhaled NO, that involves diffusing across the alveolar
30 capillary barrier and binding to hemoglobin in red blood cells and, to a lesser extent,
31 diffusing across airway epithelium to react with soluble guanylate cyclase in airway
32 smooth muscle. These comprise the initiating events in the mode of action for inhaled
33 NO. The resulting key events include reactions with hemoglobin to form
34 nitrosylhemoglobin, methemoglobin, nitrate, and possibly S-nitrosohemoglobin, and
35 activation of soluble guanylate cyclase, which produces mediators that relax airway
36 smooth muscle. Because health effects of inhaled NO have not been identified in
37 Chapters 5 and 6, no endpoints or outcomes have been included in this analysis.
January 2015 4-64 DRAFT: Do Not Cite or Quote
-------
NO —
major JT
^ Diffusion across
cell membranes
minor^v
Legend in
Pollutant
• Key Events
— Pathway
Binding to
hemoglobin in
red blood cells
Binding to soluble
guanylate cyclase
airway smooth muscle
Reactions
hemoglobin
Enzyme
activation and —
mediator
production
Circulating
products
Relaxation of airway
smooth muscle
••• Potential Pathway
Source: National Center for Environmental Assessment.
Figure 4-4 Mode of action of inhaled nitric oxide (NO).
4.4 Summary
i
2
3
4
5
6
7
This chapter provides a foundation for understanding how exposure to the gaseous air
pollutants NC>2 and NO may lead to health effects. This encompasses the many steps
between uptake into the respiratory tract and the biological responses that ensue. While
NO2 reacts with components of the ELF and with respiratory epithelium, NO reacts with
heme proteins in the circulation. These chemical interactions are responsible for targeting
these oxides of nitrogen species to different tissues, that is, the NO2 to the respiratory
tract and NO to the circulation. Biologic responses to inhaled NO2 and NO were
organized into a mode of action framework that may be used to guide interpretation of
health effects evidence presented in subsequent chapters.
January 2015
4-65
DRAFT: Do Not Cite or Quote
-------
References for Chapter 4
Adgent MA: Squadrito. GL: Ballinger. CA: Krzywanskj DM: Lancaster. JR. Jr: Postlethwait EM.
(2012). Desferrioxamine inhibits protein tyro sine nitration: Mechanisms and implications. Free Radic
Biol Med 53: 951-961. http://dx.doi.Org/10.1016/i.freeradbiomed.2012.06.003
Ahmad. T: Mabalirajan. U: Sharma. A: Aich. J: Makhija. L: Ghosh. B: Agrawal. A. (2011). Simvastatin
improves epithelial dysfunction and airway hyperresponsiveness: From asymmetric dimethyl-arginine
to asthma. Am J Respir Cell Mol Biol 44: 531-539. http://dx.doi.org/10.1165/rcmb.2010-0041OC
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
Alarie. Y. (1973). Sensory irritation by airborne chemicals [Review]. CRC Crit Rev Toxicol 2: 299-363.
http://dx.doi.org/10.3109/10408447309082020
Alberg. T: Nilsen. A: Hansen. JS: Nygaard. UC: L0vik. M. (2011). Nitrogen dioxide: No influence on
allergic sensitization in an intranasal mouse model with ovalbumin and diesel exhaust particles. Inhal
Toxicol 23: 268-276. http://dx.doi.org/10.3109/08958378.2011.566898
Alexanderson. C: Olin. AC: Dahlman-Hoglund. A: Finizia. C: Toren. K. (2012). Nasal nitric oxide in a
random sample of adults and its relationship to sensitization, cat allergen, rhinitis, and ambient nitric
oxide. Am J Rhinol Allergy 26: e99-e!03. http://dx.doi.org/10.2500/ajra.2012.26.3777
Alving. K: Weitzberg. E: Lundberg. JM. (1993). Increased amount of nitric oxide in exhaled air of
asthmatics. Eur Respir J 6: 1368-1370.
Anderson. JT: Zeng. M: Li. Q: Staplev. R: Moore. PR: Chenna. B: Fineberg. N: Zmijewski. J: Eltoum.
IE: Siegal GP: Gaggar. A: Barnes. S: Velu. SE: Thannickal. VJ: Abraham. E: Patel RP: Lancaster. J.
R.: Chaplin. DP: Dransfield. MT: Deshane. JS. (2011). Elevated levels of NO are localized to distal
airways in asthma. Free Radic Biol Med 50: 1679-1688.
http://dx.doi.0rg/10.1016/j.freeradbiomed.2011.03.015
Asgharian. B: Price. OT: Schroeter. JD: Kimbell JS: Jones. L: Singal M. (2011). Derivation of mass
transfer coefficients for transient uptake and tissue disposition of soluble and reactive vapors in lung
airways. Ann Biomed Eng 39: 1788-1804. http://dx.doi.org/10.1007/sl0439-011-0274-9
Ather. JL: Alcorn. JF: Brown. AL: Guala. AS: Suratt BT: Janssen-Heininger. YM: Poynter. ME. (2010).
Distinct functions of airway epithelial nuclear factor-kappaB activity regulate nitrogen dioxide-
induced acute lung injury. Am J Respir Cell Mol Biol 43: 443-451.
http://dx.doi.org/10.1165/rcmb.2008-0416OC
ATS/ERS (American Thoracic Society and European Respiratory Society). (2005). ATS/ERS
recommendations for standardized procedures for the online and offline measurement of exhaled
lower respiratory nitric oxide and nasal nitric oxide, 2005. Am J Respir Crit Care Med 171: 912-930.
http://dx.doi.org/10.1164/rccm.200406-710ST
Augusto. O: Bonini. MG: Amanso. AM: Linares. E: Santos. CC: De Menezes. SL. (2002). Nitrogen
dioxide and carbonate radical anion: two emerging radicals in biology [Review]. Free Radic Biol Med
32: 841-859.
Ayaz. KL: Csallany. AS. (1978). Long-term NO2 exposure of mice in the presence and absence of
vitamin E. II. Effect of glutathione peroxidase. Arch Environ Occup Health 33: 292-296.
Azadniv. M: Utell MJ: Morrow. PE: Gibb. FR: Nichols. J: Roberts. NJ. Jr: Speers. DM: Torres. A: Tsai
Y: Abraham. MK: Voter. KZ: Frampton. MW. (1998). Effects of nitrogen dioxide exposure on human
host defense. Inhal Toxicol 10: 585-601. http://dx.doi.org/10.1080/089583798197556
January 2015 4-66 DRAFT: Do Not Cite or Quote
-------
Babior. BM: Takeuchj C: Ruedi J: Gutierrez. A: Wentworth. P. (2003). Investigating antibody-catalyzed
ozone generation by human neutrophils. PNAS 100: 3031-3034.
http://dx.doi.org/10.1073/pnas.0530251100
Barck. C: Lundahl. J: Hallden. G: Bylin. G. (2005). Brief exposures to NO2 augment the allergic
inflammation in asthmatics. Environ Res 97: 58-66. http://dx.doi.Org/10.1016/j.envres.2004.02.009
Barck. C: Sandstrom. T: Lundahl J: Hallden. G: Svartengren. M: Strand. V: Rak. S: Bylin. G. (2002).
Ambient level of NO2 augments the inflammatory response to inhaled allergen in asthmatics. Respir
Med 96: 907-917. http://dx.doi.org/10.1053/rmed.2002.1374
Barth. PJ: Muller. B: Wagner. U: Bittinger. A. (1995). Quantitative analysis of parenchymal and vascular
alterations in NO2-induced lung injury in rats. Eur Respir J 8: 1115-1121.
http://dx.doi.org/10.1183/09031936.95.08071115
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. J Appl Physiol 79: 1615-1628.
Bates. DV: Macklem. PT: Christie. RV. (1971). Respiratory function in disease: An introduction to the
integrated study of the lung (2nd ed.). Philadelphia, PA: W.B. Saunders Company.
Bauer. MA: Utell MJ: Morrow. PE: Speers. DM: Gibb. FR. (1986). Inhalation of 0.30 ppm nitrogen
dioxide potentiates exercise-induced bronchospasm in asthmatics. Am Rev Respir Dis 134: 1203-
1208.
Bautista. AP: Eisenlohr. CP: Lanz. MJ. (2011). Nasal nitric oxide and nasal eosinophils decrease with
levocetirizine in subjects with perennial allergic rhinitis. Am J Rhinol Allergy 25: 383-387.
http://dx.doi.org/10.2500/aira.2011.25.3668
Beaven. MA. (2009). Our perception of the mast cell from Paul Ehrlich to now. Eur J Immunol 39: 11-25.
http://dx.doi.org/10.1002/eii.200838899
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: Miura. M: Verleden. GM: Tadjkarimi. S: Yacoub. MH: Barnes. PJ. (1992).
Inhibitory NANC nerves in human tracheal smooth muscle: A quest for the neurotransmitter. J Appl
Physiol 73: 2505-2510.
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
Bernhard. W: Haslam. PL: Floros. J. (2004). From birds to humans: new concepts on airways relative to
alveolar surfactant [Review]. Am J Respir Cell Mol Biol 30: 6-11.
http://dx.doi.org/10.1165/rcmb.2003-0158TR
Bevelander. M: Mavette. J: Whittaker. LA: Paveglio. SA: Jones. CC: Robbins. J: Hemenwav. D: Akira. S:
Uematsu. S: Poynter. ME. (2007). Nitrogen dioxide promotes allergic sensitization to inhaled antigen.
J Immunol 179: 3680-3688.
Bhaskar. KR: O'Sullivan. DP: Seltzer. J: Rossing. TH: Drazen. JM: Reid. LM. (1985). Density gradient
study of bronchial mucus aspirates from healthy volunteers (smokers and nonsmokers) and from
patients with tracheostomy. Exp Lung Res 9: 289-308. http://dx.doi.org/10.3109/01902148509057529
Bidani. A: Postlethwait EM. (1998). Kinetic determinants of inhaled reactive gas absorption. In MP
Hlastala; HT Robertson (Eds.), Complexity in structure and function of the lung (pp. 243-296). New
York, NY: Marcell Dekker Inc.
Blomberg. A: Krishna. MT: Bocchino. V: Biscione. GL: Shute. JK: Kelly. FJ: Frew. AJ: Holgate. ST:
Sandstrom. T. (1997). The inflammatory effects of 2 ppm NO2 on the airways of healthy subjects. Am
J Respir Crit Care Med 156: 418-424.
January 2015 4-67 DRAFT: Do Not Cite or Quote
-------
Blomberg. A: Krishna. MT: Helledav. R: Soderberg. M: Ledin. MC: Kelly. FJ: Frew. AJ: Holgate. ST:
Sandstrom. T. (1999). Persistent airway inflammation but accomodated antioxidant and lung function
responses after repeated daily exposure to nitrogen dioxide. Am J Respir Crit Care Med 159: 536-543.
http://dx.doi.0rg/10.1164/airccm.159.2.9711068
Blood. AB: Schroeder. HJ: Terry. MH: Merrill-Henry. J: Bragg. SL: Vrancken. K: Liu. T: Herring. JL:
Sowers. LC: Wilson. SM: Power. GG. (2011). Inhaled nitrite reverses hemolysis-induced pulmonary
vasoconstriction in newborn lambs without blood participation. Circulation 123: 605-612.
http://dx.doi.org/10.1161/CIRCULATIONAHA.110.001073
Bonacci G: Baker. PR: Salvatore. S. R.: Shores. D: Khoo. NK: Koenitzer. J. R.: Vitturi. DA: Woodcock.
S. R.: Golin-Bisello. F: Cole. MP: Watkins. S: St Croix. C: Batthyany. CI: Freeman. BA: Schopfer. FJ.
(2012). Conjugated linoleic acid is a preferential substrate for fatty acid nitration. J Biol Chem 287:
44071-44082. http://dx.doi.org/10.1074/jbc.M112.401356
Bonacci. G: Schopfer. FJ: Batthyany. CI: Rudolph. TK: Rudolph. V: Khoo. NK: Kellev. EE: Freeman.
BA. (2011). Electrophilic fatty acids regulate matrix metalloproteinase activity and expression. J Biol
Chem 286: 16074-16081. http://dx.doi.org/10.1074/ibc.Mlll.225029
Borland. CDR: Dunningham. H: Bottrill. F: Vuylsteke. A: Yilmaz. C: Dane. DM: Hsia. CCW. (2010).
Significant blood resistance to nitric oxide transfer in the lung. J Appl Physiol 108: 1052-1060.
http://dx.doi.org/10.1152/japplphvsiol.00904.2009
Brahmajothi. MV: Mason. SN: Whorton. AR: Mcmahon. TJ: Auten. RL. (2010). Transport rather than
diffusion-dependent route for nitric oxide gas activity in alveolar epithelium. Free Radic Biol Med 49:
294-300. http://dx.doi.0rg/10.1016/j.freeradbiomed.2010.04.020
Brochu. P: Brodeur. J: Krishnan. K. (2011). Derivation of physiological inhalation rates in children,
adults, and elderly based on nighttime and daytime respiratory parameters. Inhal Toxicol 23: 74-94.
http://dx.doi.org/10.3109/08958378.2010.543439
Bryan. NS. (2006). Nitrite in nitric oxide biology: Cause or consequence? A systems-based review
[Review]. Free Radic Biol Med 41: 691-701. http://dx.doi.Org/10.1016/j.freeradbiomed.2006.05.019
Bryan. NS: Fernandez. BO: Bauer. SM: Garcia-Saura. MF: Milsom. AB: Rassaf. T: Malonev. RE: Bharti.
A: Rodriguez. J: Feelisch. M. (2005). Nitrite is a signaling molecule and regulator of gene expression
in mammalian tissues. Nat Chem Biol 1: 290-297. http://dx.doi.org/10.1038/nchembio734
Button. B: Cai. LH: Ehre. C: Kesimer. M: Hill. DB: Sheehan. JK: Boucher. RC: Rubinstein. M. (2012). A
periciliary brush promotes the lung health by separating the mucus layer from airway epithelia.
Science 337: 937-941. http://dx.doi.org/10.1126/science.1223012
Bylin. G: Lindvall. T: Rehn. T: Sundin. B. (1985). Effects of short-term exposure to ambient nitrogen
dioxide concentrations on human bronchial reactivity and lung function. Eur Respir J 66: 205-217.
Campen. MJ: Lund. AK: Doyle-Eisele. ML: Mcdonald. JD: Knuckles. TL: Rohr. AC: Knipping. EM:
Mauderly. JL. (2010). A comparison of vascular effects from complex and individual air pollutants
indicates a role for monoxide gases and volatile hydrocarbons. Environ Health Perspect 118: 921-927.
http://dx.doi.org/10.1289/ehp.0901207
Cardinale. F: Tesse. R: Fucilli. C: Loffredo. MS: lacoviello. G: Chinellato. I: Armenio. L. (2007).
Correlation between exhaled nitric oxide and dietary consumption of fats and antioxidants in children
with asthma [Letter]. J Allergy Clinlmmunol 119: 1268-1270.
http://dx.doi.0rg/10.1016/j.jaci.2007.01.028
Cavanagh. DG: Morris. JB. (1987). Mucus protection and airway peroxidation following nitrogen dioxide
exposure in the rat. J Toxicol Environ Health 22: 313-328.
http://dx.doi.org/10.1080/15287398709531074
Chadha. TS: Birch. S: Sackner. MA. (1987). Oronasal distribution of ventilation during exercise in
normal subjects and patients with asthma and rhinitis. Chest 92: 1037-1041.
http://dx.doi.0rg/10.1378/chest.92.6.1037
January 2015 4-68 DRAFT: Do Not Cite or Quote
-------
Chakrabortv. S: Balakotaiah. V: Bidani A. (2004). Diffusing capacity reexamined: Relative roles of
diffusion and chemical reaction in red cell uptake of O2, CO, CO2, and NO. J Appl Physiol 97: 2284-
2302. http://dx.doi.org/10.1152/japplphvsiol.00469.2004
Chambers. DC: Ayres. JG. (2001). Effects of nitrogen dioxide exposure and ascorbic acid
supplementation on exhaled nitric oxide in healthy human subjects. Thorax 56: 774-778.
http://dx.doi.org/10.1136/thorax.56.10.774
Chang. LY: Graham. JA: Miller. FJ: Ospital JJ: Crapo. JD. (1986). Effects of subchronic inhalation of
low concentrations of nitrogen dioxide: I. The proximal alveolar region of juvenile and adult rats.
Toxicol Appl Pharmacol 83: 46-61. http://dx.doi.org/10.1016/0041-008X(86)90321-2
Channel!. MM: Paffett ML: Devlin. RB: Madden. MC: Campen. MJ. (2012). Circulating factors induce
coronary endothelial cell activation following exposure to inhaled diesel exhaust and nitrogen dioxide
in humans: Evidence from a novel translational in vitro model. Toxicol Sci 127: 179-186.
http://dx.doi.org/10.1093/toxsci/kfs084
Chase. MW. Jr. (1998). NIST-JANAF thermochemical tables, monograph 9 (Fourth Edition ed.).
Washington DC: American Chemical Society, http://www.nist.gov/srd/reprints.cfm
Ckless. K: Hodgkins. SR: Ather. JL: Martin. R: Poynter. ME. (2011). Epithelial, dendritic, and CD4(+) T
cell regulation of and by reactive oxygen and nitrogen species in allergic sensitization [Review].
Biochim Biophys Acta 1810: 10251034. http://dx.doi.Org/10.1016/i.bbagen.2011.03.005
Ckless. K: Lampert A: Reiss. J: Kasahara. D: Poynter. ME: Irvin. CG: Lundblad. LKA: Norton. R: van
der Vliet. A: Janssen-Heininger. YMW. (2008). Inhibition of arginase activity enhances inflammation
in mice with allergic airway disease, in association with increases in protein S-nitrosylation and
tyrosine nitration. J Immunol 181: 4255-4264.
Clary-Meinesz. C: Mouroux. J: Huitorel P: Cosson. J: Schoevaert D: Blaive. B. (1997). Ciliary beat
frequency in human bronchi and bronchioles. Chest 111: 692-697.
http://dx.doi.0rg/10.1378/chest.lll.3.692
Cockcroft DW: Davis. BE. (2006). Mechanisms of airway hyperresponsiveness [Review]. J Allergy Clin
Immunol 118: 551-559. http://dx.doi.Org/10.1016/i.iaci.2006.07.012
Comhair. SAA: Erzurum. SC. (2010). Redox control of asthma: Molecular mechanisms and therapeutic
opportunities [Review]. Antioxid Redox Signal 12: 93-124. http://dx.doi.org/10.1089/ARS.2008.2425
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/aiplung.00397.2003
Connor. LM: Bidani. A: Goerke. J: Clements. JA: Postlethwait. EM. (2001). NO2 interfacial transfer is
reduced by phospholipid monolayers. J Appl Physiol 91: 2024-2034.
Cosby. K: Partovi. KS: Crawford. JH: Patel RP: Reiter. CD: Martyr. S: Yang. BK: Waclawiw. MA: Zalos.
G: Xu. X: Huang. KT: Shields. H: Kim-Shapiro. DB: Schechter. AN: Cannon. RO: Gladwin. MT.
(2003). Nitrite reduction to nitric oxide by deoxyhemoglobin vasodilates the human circulation. Nat
Med 9: 1498-1505. http://dx.doi.org/10.1038/nm954
Daiber. A: Muenzel. T. (2012). Increased circulating levels of 3-nitrotyrosine autoantibodies marker for or
maker of cardiovascular disease? [Editorial]. Circulation 126: 2371-2373.
http://dx.doi.org/10.1161/CIRCULATIONAHA.112.143214
Davies. RJ: Rusznak. C: Calderon. MA: Wang. JH: Abdelaziz. MM: Devalia. JL. (1997). Allergen-irritant
interaction and the role of corticosteroids [Review]. Allergy 52: 59-65.
http://dx.doi.0rg/10.llll/i.1398-9995.1997.tb04873.x
Davis. JK: Davidson. MK: Schoeb. TR: Lindsev. JR. (1992). Decreased intrapulmonary killing of
Mycoplasma pulmonis after short-term exposure to NO2 is associated with damaged alveolar
macrophages. Am Rev Respir Dis 145: 406-411. http://dx.doi.Org/10.1164/ajrccm/145.2 Pt 1.406
January 2015 4-69 DRAFT: Do Not Cite or Quote
-------
de Burbure. CY: Heilier. JF: Neve. J: Becker. A: Albrecht C: Borm. PJA: Gromadzinska. J: Wasowicz.
W: Rydzynski. K: Bernard. AM. (2007). Lung permeability, antioxidant status, and NO2 inhalation: A
selenium supplementation study in rats. J Toxicol Environ Health A 70: 284-294.
http://dx.doi.org/10.1080/15287390600884875
de Winter-Sorkina. R: Cassee. FR. (2002). From concentration to dose: Factors influencing airborne
particulate matter deposition in humans and rats. (RIVM Report 650010031). Bilthoven, Netherlands:
National Institute for Public Health and the Environment.
http://www.rivm.nl/en/Documents and_publications/Scientific/Reports/2003/februari/From concentra
tion to dose factors influencing airborne_particulate matter deposition in humans and rats?sp=c
ml2bXE9ZmFsc2U7c2VhcmNoYmFzZT01MDI5MDtvaXZtcTlmYWxzZTs=&pagenr=5030
Devlin. RB: Horstman. DP: Gerritv. TR: Becker. S: Madden. MC. (1999). Inflammatory response in
humans exposed to 2.0 PPM nitrogen dioxide. Inhal Toxicol 11: 89-109.
http://dx.doi.org/10.1080/089583799197195
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.
Powell. AR: Kilburn. KH: Pratt. PC. (1971). Short-term exposure to nitrogen dioxide: Effects on
pulmonary ultrastructure, compliance, and the surfactant system. Arch Intern Med 128: 74-80.
http://dx.doi.org/10.1001/archinte.1971.00310190078009
Drahl C. (2009). Probing for in-body ozone. Chem Eng News 87: 40-42. http://dx.doi.org/10.1021/cen-
v087n025.pQ40
Dressel H: Muller. F: Fischer. R: Rommelt H: Hohlfeld. JM: Behr. J: Huber. RM: Nowak. D: Torres. RA.
(2010). Independent information of nonspecific biomarkers in exhaled breath condensate. Respiration
80: 401-409. http://dx.doi.org/10.1159/000319945
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.
http://dx.doi.0rg/10.1165/ajrcmb.14.l.8534488
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.
http://dx.doi.org/10.1006/taap.1993.1223
Dupuv. PM: Shore. SA: Drazen. JM: Frostell C: Hill. WA: Zapol WM. (1992). Bronchodilator action of
inhaled nitric oxide in guinea pigs. J Clin Invest 90: 421-428. http://dx.doi.org/10.1172/JCI115877
Ehrlich. R. (1980). Interaction between environmental pollutants and respiratory infections. Environ
Health Perspect 35: 89-99.
Ehrlich. R: Findlav. JC: Fenters. JD: Gardner. DE. (1977). Health effects of short-term inhalation of
nitrogen dioxide and ozone mixtures. Environ Res 14: 223-231. http://dx.doi.org/10.1016/0013-
9351(77)90034-2
Enamj S: Hoffmann. MR: Colussi. AJ. (2009). Absorption of inhaled NO(2) [Letter]. J Phys Chem B
113: 7977-7981. http://dx.doi.org/10.1021/jp902667x
England. C: Corcoran. WH. (1974). Kinetics and mechanisms of the gas-phase reaction of water vapor
and nitrogen dioxide. Ind Eng Chem Fundam 13: 373-384. http://dx.doi.org/10.1021/il60052a014
Erjefalt. JS: Greiff. L: Andersson. M: Matsson. E: Petersen. H: Linden. M: Ansari. T: Jeffery. PK:
Persson. CG. (1999). Allergen-induced eosinophil cytolysis is a primary mechanism for granule
protein release in human upper airways. Am J Respir Crit Care Med 160: 304-312.
http://dx.doi.0rg/10.1164/airccm.160.l.9809048
January 2015 4-70 DRAFT: Do Not Cite or Quote
-------
Ezrattv. V: Guillossou. G: Neukirch. C: Dehoux. M: Koscielny. S: Bonav. M: Cabanes. PA: Samet JM:
Mure. P: Ropert. L: Tokarek. S: Lambrozo. J: Aubier. M. (2014). Repeated nitrogen dioxide exposures
and eosinophilic airway inflammation in asthmatics: a randomized crossover study. Environ Health
Perspect 122: 850-855. http://dx.doi.org/10.1289/ehp.1307240
Fehrenbach. H: Zimmermann. G: Starke. E: Bratu. VA: Conrad. D: Yildirim. AO: Fehrenbach. A. (2007).
Nitrogen dioxide induces apoptosis and proliferation but not emphysema in rat lungs. Thorax 62: 438-
446. http://dx.doi.org/10.1136/thx.2006.062364
Finlavson-Pitts. BJ: Sweetman. LL: Weissbart. B. (1987). A Fourier transform infrared spectrometry
study of the reactions of phosphatidylcholines with gaseous N2O5 and NO2. Toxicol Appl Pharmacol
89: 438-448. http://dx.doi.org/10.1016/0041-008X(87)90163-3
Finlavson-Pitts. BJ: Wingen. LM: Sumner. AL: Svomin. D: Ramazan. KA. (2003). The heterogeneous
hydrolysis of NO2 in laboratory systems and in outdoor and indoor atmospheres: An integrated
mechanism. Phys Chem Chem Phys 5: 223-242. http://dx.doi.org/10.1039/b208564j
Folinsbee. LJ. (1992). Does nitrogen dioxide exposure increase airways responsiveness? Toxicol Ind
Health 8: 273-283.
Ford. E: Hughes. MN: Wardman. P. (2002). Kinetics of the reactions of nitrogen dioxide with glutathione,
cysteine, and uric acid at physiological pH. Free Radic Biol Med 32: 1314-1323.
http://dx.doi.org/10.1016/S0891-5849(02)00850-X
Forstermann. U: Sessa. WC. (2012). Nitric oxide synthases: Regulation and function [Review]. Eur Heart
J 33: 829-837. http://dx.doi.org/10.1093/eurheartj/ehr304
Frampton. MW: Boscia. J: Roberts. NJ. Jr: Azadniv. M: Torres. A: Cox. C: Morrow. PE: Nichols. J:
Chalupa. D: Frasier. LM: Gibb. FR: Speers. DM: Tsai Y: Utell. MJ. (2002). Nitrogen dioxide
exposure: Effects on airway and blood cells. Am J Physiol 282: L155-L165.
Freeman. G: Furiosi. NJ: Havdon. GB. (1966). Effects of continuous exposure of 0.8 ppm NO2 on
respiration of rats. Arch Environ Health 13: 454-456.
http://dx.doi.org/10.1080/00039896.1966.10664595
Fujimaki. H: Nohara. O. (1994). Changes in the response of lung mast cells isolated from rats and guinea
pigs exposed to nitrogen dioxide. Inhal Toxicol 6: 515-520.
http://dx.doi.org/10.3109/08958379409040508
Fukuto. J: Carrington. SJ: Tantillo. DJ: Harrison. JG: Ignarro. LJ: Freeman. BA: Chen. A: Wink. DA.
(2012). Small molecule signaling agents: The integrated chemistry and biochemistry of nitrogen
oxides, oxides of carbon, dioxygen, hydrogen sulfide, and their derived species [Review]. Chem Res
Toxicol 25: 769-793. http://dx.doi.org/10.1021/tx2005234
Gaston. B: Sears. S: Woods. J: Hunt. J: Ponaman. M: McMahon. T: Stamler. J. (1998). Bronchodilator S-
nitrosothiol deficiency in asthmatic respiratory failure. Lancet 351: 1317-1319.
http://dx.doi.org/10.1016/S0140-6736(97)07485-0
Gelb. AF: Barnes. PJ: George. SC: Ricciardolo. FL: Dimaria. G: Zamel N. (2012). Review of exhaled
nitric oxide in chronic obstructive pulmonary disease [Review]. J Breath Res 6: 047101.
http://dx.doi.Org/10.1088/1752-7155/6/4/047101
Gilmour. MI: Park. P: Selgrade. MK. (1996). Increased immune and inflammatory responses to dust mite
antigen in rats exposed to 5 ppm NO2. Fundam Appl Toxicol 31: 65-70.
http://dx.doi.org/10.1006/faat.1996.0076
Gladwin. MT: Schechter. AN: Kim-Shapiro. DB: Patel RP: Hogg. N: Shiva. S: Cannon. RO. Ill: Kelm.
M: Wink. DA: Espev. MG: Oldfield. EH: Pluta. RM: Freeman. BA: Lancaster. JR: Feelisch. M:
Lundberg. JO. (2005). The emerging biology of the nitrite anion. Nat Chem Biol 1: 308-314.
http://dx.doi.org/10.1038/nchembioll05-308
January 2015 4-71 DRAFT: Do Not Cite or Quote
-------
Goings. SAJ: Kulle. TJ: Bascom. R: Sauder. LR: Green. DJ: Hebel. JR: Clements. ML. (1989). Effect of
nitrogen dioxide exposure on susceptibility to influenza A virus infection in healthy adults. Am Rev
Respir Dis 139: 1075-1081. http://dx.doi.Org/10.1164/ajrccm/139.5.1075
Goldstein. S: Lind. J: Merenyi. G. (2004). Reaction of organic peroxyl radicals with NO 2 and NO in
aqueous solution: Intermediacy of organic peroxynitrate and peroxynitrite species. J Phys Chem A
108: 1719-1725. http://dx.doi.org/10.1021/jp037431z
Gray. EL: Patton. FM: Goldberg. SB: Kaplan. E. (1954). Toxicity of the oxides of nitrogen. II. Acute
inhalation toxicity of nitrogen dioxide, red fuming nitric acid, and white fuming nitric acid. AMA
Arch Ind Health 10: 418-422.
Greene. ND: Schneider. SL. (1978). Effects of NO2 on the response of baboon alveolar macrophages to
migration inhibitory factor. J Toxicol Environ Health 4: 869-880.
http://dx.doi.org/10.1080/15287397809529708
Griscavage. JM: Hobbs. AJ: Ignarro. LJ. (1995). Negative modulation of nitric oxide synthase by nitric
oxide and nitroso compounds [Review]. In LJ Ignarro; F Murad (Eds.), Nitric oxide: Biochemistry,
molecular biology, and therapeutic implications (pp. 215-234). San Diego, CA: Academic Press.
http://dx.doi.org/10.1016/S1054-3589(08)61088-l
Gushchin. IS: Petvaev. IM: Tsinkalovsky. OR. (1990). Kinetics of oxygen metabolism indices in the
course of histamine secretion from rat mast cells. Agents and Actions 30: 85-88.
http://dx.doi.org/10.1007/BF01969005
Hamid. Q: Springall. PR: Riveros-Moreno. V: Chanez. P: Howarth. P: Redington. A: Bousquet. J:
Godard. P: Holgate. S: Polak. JM. (1993). Induction of nitric oxide synthase in asthma. Lancet 342:
1510-1513. http://dx.doi.org/10.1016/80140-6736(05)80083-2
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: Selgrade. MK: Stead. AG. (1986). Nitrogen dioxide exposure and lung antioxidants
in ascorbic acid-deficient guinea pigs. Toxicol Appl Pharmacol 82: 351-359.
http://dx.doi.org/10.1016/0041-008X(86)90212-7
Helledav. R: Huberman. D: Blomberg. A: Stjernberg. N: Sandstrom. T. (1995). Nitrogen dioxide exposure
impairs the frequency of the mucociliary activity in healthy subjects. Eur Respir J 8: 1664-1668.
http://dx.doi.org/10.1183/09031936.95.08101664
Heller. H: Korbmacher. N: Gabler. R: Brandt. S: Breitbach. T: Juergens. U: Grohe. C: Schuster. KD.
(2004). Pulmonary 15NO uptake in interstitial lung disease. Nitric Oxide 10: 229-232.
http://dx.doi.0rg/10.1016/j.niox.2004.06.001
Hesterberg. TW: Bunn. WB: Mcclellan. RO: Hamade. AK: Long. CM: Valberg. PA. (2009). Critical
review of the human data on short-term nitrogen dioxide (NO2) exposures: Evidence for NO2 no-
effect levels [Review]. Crit Rev Toxicol 39: 743-781. http://dx.doi.org/10.3109/10408440903294945
Higdon. A: Piers. AR: Oh. J: Landar. A: Parley-Usmar. VM. (2012). Cell signalling by reactive lipid
species: new concepts and molecular mechanisms [Review]. Biochem J 442: 453-464.
http://dx.doi.org/10.1042/BJ20111752
Hill BG: Dranka. BP: Bailey. SM: Lancaster. JR: Darlev-Usmar. VM. (2010). What part of NO don't you
understand? Some answers to the cardinal questions in nitric oxide biology [Review]. J Biol Chem
285: 19699-19704. http://dx.doi.org/10.1074/ibc.R110.101618
Hodgkins. SR: Ather. JL: Paveglio. SA: Allard. JL: Leclair. LA: Suratt BT: Boyson. JE: Poynter. ME.
(2010). NO2 inhalation induces maturation of pulmonary CDllc+ cells that promote antigenspecific
CD4+ T cell polarization. Respir Res 11: 102. http://dx.doi.org/10.1186/1465-9921-ll-102
Hogman. M: Frostell C: Arnberg. H: Hedenstierna. G. (1993). Inhalation of nitric oxide modulates
methacholine-induced bronchoconstriction in the rabbit. Eur Respir J 6: 177-180.
January 2015 4-72 DRAFT: Do Not Cite or Quote
-------
Hohlfeld JM. (2002). The role of surfactant in asthma. Respir Res 3: 4. http://dx.doi.org/10.1186/rrl76
Huang. YC: Rappold. AG: Graff. DW: Ohio. AJ: Devlin. RB. (2012). Synergistic effects of exposure to
concentrated ambient fine pollution particles and nitrogen dioxide in humans. Inhal Toxicol 24: 790-
797. http://dx.doi.org/10.3109/08958378.2012.718809
Hubbard. AK: Symanowicz. PT: Thibodeau. M: Thrall RS: Schramm. CM: Cloutier. MM: Morris. JB.
(2002). Effect of nitrogen dioxide on ovalbumin-induced allergic airway disease in a murine model. J
Toxicol Environ Health A 65: 1999-2005. http://dx.doi.org/10.1080/00984100290071397
Huie. RE. (1994). The reaction kinetics of NO2 [Review]. Toxicology 89: 193-216.
http://dx.doi.org/10.1016/0300-483X(94)90098-l
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.
Jain. B: Rubinstein. I: Robbins. RA: Leise. KL: Sisson. JH. (1993). Modulation of airway epithelial cell
ciliary beat frequency by nitric oxide. Biochem Biophys Res Commun 191: 83-88.
http://dx.doi.org/10.1006/bbrc.1993.1187
Jakab. GJ. (1987). Modulation of pulmonary defense mechanisms by acute exposures to nitrogen dioxide.
Experientia, Suppl 51: 235-242.
Jenkins. HS: Devalia. JL: Mister. RL: Sevan. AM: Rusznak. C: Davies. RJ. (1999). The effect of
exposure to ozone and nitrogen dioxide on the airway response of atopic asthmatics to inhaled
allergen: dose- and time-dependent effects. Am J Respir Crit Care Med 160: 33-39.
http://dx.doi.0rg/10.1164/airccm.160.l.9808119
Jilma. B: Kastner. J: Mensik. C: Vondrovec. B: Hildebrandt J: Krejcv. k: Wagner. OF: Eichler. H. g.
(1996). Sex differences in concentrations of exhaled nitric oxide and plasma nitrate. Life Sci 58: 469-
476. http://dx.doi.org/10.1016/0024-3205(95)02311-9
Torres. R: Nowak. D: Grimminger. F: Seeger. W: Oldigs. M: Magnussen. H. (1995). The effect of 1 ppm
nitrogen dioxide on bronchoalveolar lavage cells and inflammatory mediators in normal and asthmatic
subjects. Eur Respir J 8: 416-424. http://dx.doi.org/10.1183/09031936.95.08030416
Kelly. FJ: Blomberg. A: Frew. A: Holgate. ST: Sandstrom. T. (1996a). Antioxidant kinetics in lung lavage
fluid following exposure of humans to nitrogen dioxide. Am J Respir Crit Care Med 154: 1700-1705.
http://dx.doi.0rg/10.1164/airccm.154.6.8970358
Kelly. FJ: Cotgrove. M: Mudwav. IS. (1996b). Respiratory tract lining fluid antioxidants: the first line of
defence against gaseous pollutants. Cent Eur J Public Health 4: 11-14.
Kelly. FJ: Tetlev. TD. (1997). Nitrogen dioxide depletes uric acid and ascorbic acid but not glutathione
from lung lining fluid. Biochemistry 325: 95-99.
Kettle. AJ: Clark. BM: Winterbourn. CC. (2004). Superoxide converts indigo carmine to isatin sulfonic
acid: Implications for the hypothesis that neutrophils produce ozone. J Biol Chem 279: 18521-18525.
http://dx.doi.org/10.1074/ibc.M400334200
Kharitonov. SA: Walker. L: Barnes. PJ. (2005). Repeatability of standardised nasal nitric oxide
measurements in healthy and asthmatic adults and children. Respir Med 99: 1105-1114.
http://dx.doi.0rg/10.1016/j.rmed.2005.02.019
Khoo. NK: Rudolph. V: Cole. MP: Golin-Bisello. F: Schopfer. FJ: Woodcock. SR: Batthyany. C:
Freeman. BA. (2010). Activation of vascular endothelial nitric oxide synthase and herne oxygenase-1
expression by electrophilic nitro-fatty acids. Free Radic Biol Med 48: 230-239.
http://dx.doi.0rg/10.1016/j.freeradbiomed.2009.10.046
Kikugawa. K: Kogj K. (1987). Oxidation products of linoleate and linolenate exposed to nitrogen
dioxide. Chem Pharm Bull (Tokyo) 35: 344-349. http://dx.doi.org/10.1248/cpb.35.344
January 2015 4-73 DRAFT: Do Not Cite or Quote
-------
Kirsch. M: Korth. HG: Sustmann. R: De Groot H. (2002). The pathobiochemistry of nitrogen dioxide
[Review]. Biol Chem 383: 389-399. http://dx.doi.org/10.1515/BC.2002.043
Kitabatake. M: Yamamoto. H: Yuan. PF: Manjurul H: Murase. S: Yamauchj T. (1995). Effects of
exposure to NO2 or SO2 on bronchopulmonary reaction induced by Candida albicans in guinea pigs. J
Toxicol Environ Health 45: 75-82. http://dx.doi.org/10.1080/15287399509531981
Kleeberger. SR: Zhang. LY: Jakab. GJ. (1997). Differential susceptibility to oxidant exposure in inbred
strains of mice: Nitrogen dioxide versus ozone. Inhal Toxicol 9: 601-621.
http://dx.doi.org/10.1080/089583797198006
Kleinman. MT: Mautz. WJ. (1991). The effects of exercise on dose and dose distribution of inhaled
automotive pollutants. (Research Report no. 45). Cambridge, MA: Health Effects Institute.
http://pubs.healtheffects.org/view. php?id=45
Kloek. J: Mortaz. E: van Ark. I: Lilly. CM: Nijkamp. FP: Folkerts. G. (2010). Glutathione prevents the
early asthmatic reaction and airway hyperresponsiveness in guinea pigs. J Physiol Pharmacol 61: 67-
72.
Knuckles. TL: Buntz. JG: Paffett M: Channel!. M: Harmon. M: Cherng. T: Lucas. SN: Mcdonald. JD:
Kanagy. NL: Campen. MJ. (2011). Formation of vascular S-nitrosothiols and plasma nitrates/nitrites
following inhalation of diesel emissions. J Toxicol Environ Health A 74: 828-837.
http://dx.doi.org/10.1080/15287394.2011.570225
Kobavashj T: Miura. T. (1995). Concentration- and time-dependent increase in specific airway resistance
after induction of airway hyperresponsiveness by subchronic exposure of guinea pigs to nitrogen
dioxide. Fundam Appl Toxicol 25: 154-158. http://dx.doi.org/10.1006/faat.1995.1049
Kobavashj T: Shinozaki Y. (1990). Effect of subacute exposure to nitrogen dioxide on the airway
responsiveness of guinea pig. Agents Actions Suppl 31: 71-74.
Koehler. C: Ginzkev. C: Friehs. G: Hackenberg. S: Froelich. K: Scherzed. A: Burghartz. M: Kessler. M:
Kleinsasser. N. (2010). Aspects of nitrogen dioxide toxicity in environmental urban concentrations in
human nasal epithelium. Toxicol Appl Pharmacol 245: 219-225.
http://dx.doi.0rg/10.1016/i.taap.2010.03.003
Koehler. C: Ginzkev. C: Friehs. G: Hackenberg. S: Froelich. K: Scherzed. A: Burghartz. M: Kessler. M:
Kleinsasser. N. (2011). Ex vivo toxicity of nitrogen dioxide in human nasal epithelium at the WHO
defined 1-h limit value. Toxicol Lett 207: 89-95. http://dx.doi.Org/10.1016/j.toxlet.2011.08.004
Krishna. MT: Holgate. ST. (1999). Inflammatory mechanisms underlying potentiation of effects of
inhaled aeroallergens in response to nitrogen dioxide in allergic airways disease [Review]. Clin Exp
Allergy 29: 150-154. http://dx.doi.Org/10.1046/i.1365-2222.1999.00512.x
Kudo. M: Khalifeh Soltani. SM: Sakuma. SA: McKlerov. W: Lee. TH: Woodruff. PG: Lee. JW: Huang.
K: Chen. C: Arjomandi. M: Huang. X: Atabai. K. (2013). Mfge8 suppresses airway
hyperresponsiveness in asthma by regulating smooth muscle contraction. PNAS 110: 660-665.
http://dx.doi.org/10.1073/pnas.1216673110
Lee. BJ: Jeung. YJ: Lee. JY: Choi. DC. (2011). Increased S-nitrosothiol levels in nonasthmatic
eosinophilic bronchitis compared with cough variant asthma. Int Arch Allergy Immunol 156: 99-103.
http://dx.doi.org/10.1159/000321919
Lehnert BE: Archuleta. DC: Ellis. T: Session. WS: Lehnert. NM: Gurlev. LR: Stavert DM. (1994). Lung
injury following exposure of rats to relatively high mass concentrations of nitrogen dioxide.
Toxicology 89: 239-277. http://dx.doi.org/10.1016/0300-483X(94)90100-7
Li. H: Chen. L: Guo. Z: Sang. N: Li G. (2012). In vivo screening to determine neurological hazards of
nitrogen dioxide (NO2) using Wistar rats. J Hazard Mater 225-226: 46-53.
http://dx.doi.0rg/10.1016/i.ihazmat.2012.04.063
January 2015 4-74 DRAFT: Do Not Cite or Quote
-------
Li. H: Han. M: Guo. L: Li. G: Sang. N. (2011). Oxidative stress, endothelial dysfunction and
inflammatory response in rat heart to NO inhalation exposure. Chemosphere 82: 1589-1596.
http://dx.doi.0rg/10.1016/j.chemosphere.2010.ll.055
Lima. ES: Boninj MG: Augusto. O: Barbeiro. HV: Souza. HP: Abdalla. PS. (2005). Nitrated lipids
decompose to nitric oxide and lipid radicals and cause vasorelaxation. Free Radic Biol Med 39: 532-
539. http://dx.doi.0rg/10.1016/j.freeradbiomed.2005.04.005
Linhares. D: Jacinto. T: Pereira. AM: Fonseca. JA. (2011). Effects of atopy and rhinitis on exhaled nitric
oxide values - a systematic review. Clin Transl Allergy 1:8. http://dx.doi.org/10.1186/2045-7022-l-8
Linn. WS: Solomon. JC: Trim. SC: Spier. CE: Shamoo. DA: Venet TG: Avol EL: Hackney. JD. (1985).
Effects of exposure to 4 ppm nitrogen dioxide in healthy and asthmatic volunteers. Arch Environ
Occup Health 40: 234-239.
Lovich. MA: Bruno. NK: Plant CP: Wei. AE: Vasquez. GB: Johnson. BJ: Fine. DH: Gilbert. RJ. (2011).
Use of ultra pure nitric oxide generated by the reduction of nitrogen dioxide to reverse pulmonary
hypertension in hypoxemic swine. Nitric Oxide 24: 204-212.
http://dx.doi.0rg/10.1016/j.niox.2011.04.006
Lundberg. JO: Carlstrom. M: Larsen. FJ: Weitzberg. E. (2011). Roles of dietary inorganic nitrate in
cardiovascular health and disease [Review]. Cardiovasc Res 89: 525-532.
http://dx.doi.org/10.1093/cvr/cvq325
Lundberg. JON: Farkas-Szallasi. T: Weitzberg. E: Rinder. J: Lidholm. J: Anggaard. A: Hokfelt. T:
Lundberg. JM: Alving. K. (1995). High nitric oxide production in human paranasal sinuses. Nat Med
1: 370-373. http://dx.doi.org/10.1038/nm0495-370
Lundberg. JON: Settergren. G: Gelinder. S: Lundberg. JM: Alving. K: Weitzberg. E. (1996). Inhalation of
nasally derived nitric oxide modulates pulmonary function in humans. Acta Physiol Scand 158: 343-
347. http://dx.doi.Org/10.1046/j.1365-201X.1996.557321000.x
Maarsingh. H: Bossenga. BE: Bos. IS: Volders. HH: Zaagsma. J: Meurs. H. (2009). L-arginine deficiency
causes airway hyperresponsiveness after the late asthmatic reaction. Eur Respir J 34: 191-199.
http://dx.doi.org/10.1183/09031936.00105408
Maarsingh. H: Zuidhof. AB: Bos. IS: van Duin. M: Boucher. JL: Zaagsma. J: Meurs. H. (2008). Arginase
inhibition protects against allergen-induced airway obstruction, hyperresponsiveness, and
inflammation. Am J Respir Crit Care Med 178: 565-573. http://dx.doi.org/10.1164/rccm.200710-
1588OC
Mabalirajan. U: Ahmad. T: Leishangthem. GD: Dinda. AK: Agrawal. A: Ghosh. B. (2010a). L-arginine
reduces mitochondrial dysfunction and airway injury in murine allergic airway inflammation. Int
Immunopharmacol 10: 1514-1519. http://dx.doi.Org/10.1016/j.intimp.2010.08.025
Mabalirajan. U: Ahmad. T: Leishangthem. GD: Joseph. DA: Dinda. AK: Agrawal A: Ghosh. B. (201 Ob).
Beneficial effects of high dose of L-arginine on airway hyperresponsiveness and airway inflammation
in a murine model of asthma. J Allergy Clin Immunol 125: 626-635.
http://dx.doi.0rg/10.1016/j.jaci.2009.10.065
Madej. E: Folkes. LK: Wardman. P: Czapski. G: Goldstein. S. (2008). Thiyl radicals react with nitric
oxide to form S-nitrosothiols with rate constants near the diffusion-controlled limit. Free Radic Biol
Med 44: 2013-2018. http://dx.doi.Org/10.1016/j.freeradbiomed.2008.02.015
Magnussen. H: Reuss. G: Torres. R. (1986). Airway response to methacholine during exercise induced
refractoriness in asthma. Thorax 41: 667-670.
Maher. AR: Milsom. AB: Gunaruwan. P: Abozguia. K: Ahmed. I: Weaver. RA: Thomas. P: Ashrafian. H:
Born. GVR: James. PE: Frenneaux. MR (2008). Hypoxic modulation of exogenous nitrite-induced
vasodilation in humans. Circulation 117: 670-677.
http://dx.doi.org/10.1161/CIRCULATIONAHA.107.719591
January 2015 4-75 DRAFT: Do Not Cite or Quote
-------
Major. TA: Panmanee. W: Mortensen. JE: Gray. LD: Hoglen. N: Hassett DJ. (2010). Sodium nitrite-
mediated killing of the major cystic fibrosis pathogens Pseudomonas aeruginosa, Staphylococcus
aureus, and Burkholderia cepacia under anaerobic planktonic and biofilm conditions. Antimicrob
Agents Chemother 54: 4671-4677. http://dx.doi.org/10.1128/AAC.00379-10
Malinovschi. A: Janson. C: Holm. L: Nordvall. L: Arving. K. (2009). Basal and induced NO formation in
the pharyngo-oral tract influences estimates of alveolar NO levels. JApplPhysiol 106: 513-519.
http://dx.doi.org/10.1152/iapplphvsiol.91148.2008
Malm-ErjeFalt M: Persson. CG: ErjeFalt JS. (2001). Degranulation status of airway tissue eosinophils in
mouse models of allergic airway inflammation. Am J Respir Cell Mol Biol 24: 352-359.
http://dx.doi.0rg/10.1165/aircmb.24.3.4357
Marshall HE: Potts. EN: Kelleher. ZT: Stamler. JS: Foster. WM: Auten. RL. (2009). Protection from
lipopolysaccharide-induced lung injury by augmentation of airway S-nitrosothiols. Am J Respir Crit
Care Med 180: 11-18. http://dx.doi.org/10.1164/rccm.200807-1186OC
Marshall HE: Stamler. JS. (2001). Inhibition of NF-kappa B by S-nitrosylation. Biochemistry 40: 1688-
1693. http://dx.doi.org/10.1021/bi002239y
Martin. RA: Ather. JL: Lundblad. LKA: Suratt. BT: Boyson. JE: Budd. RC: Alcorn. JF: Flavell RA:
Eisenbarth. SC: Poynter. ME. (2013). Interleukin-1 receptor and caspase-1 are required for the Thl7
response in NO2-promoted allergic airway disease. Am J Respir Cell Mol Biol 48: 655-664.
http://dx.doi.org/10.1165/rcmb.2012-0423OC
Matalon. S: Shrestha. K: Kirk. M: Waldheuser. S: McDonald. B: Smith. K: Gao. Z: Belaaouaj. A: Crouch.
EC. (2009). Modification of surfactant protein D by reactive oxygen-nitrogen intermediates is
accompanied by loss of aggregating activity, in vitro and in vivo. FASEB J 23: 1415-1430.
http://dx.doi.org/10.1096/fi.08-120568
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.
Mauderlv. JL: Cheng. YS: Gillett. NA: Griffith. WC: Henderson. RF: Pickrell. JA: Ulblff. RK. (1990).
Influence of preexisting pulmonary emphysema on susceptibility of rats to chronic inhalation
exposure to nitrogen dioxide. Inhal Toxicol 2: 129-150.
http://dx.doi.org/10.3109/08958379009145250
McGrath. JJ: Smith. PL. (1984). Respiratory responses to nitrogen dioxide inhalation: Subtitle: Nitrogen
dioxide. J Environ Sci Health, Part A: Environ Sci Eng 19: 417-431.
http://dx.doi.org/10.1080/10934528409375168
Mercer. RR: Costa. PL: Crapo. JD. (1995). Effects of prolonged exposure to low doses of nitric oxide or
nitrogen dioxide on the alveolar septa of the adult rat lung. Lab Invest 73: 20-28.
Mercer. RR: Russell ML: Roggli. VL: Crapo. JD. (1994). Cell number and distribution in human and rat
airways. Am J Respir Cell Mol Biol 10: 613-624.
Miller. FJ: Graham. JA: Illing. JW: Gardner. DE. (1980). Extrapulmonary effects of NO2 as reflected by
pentobarbital-induced sleeping time in mice. Toxicol Lett 6: 267-274.
Miller. FJ: Graham. JA: Raub. JA: Illing. JW: Menache. MG: House. DE: Gardner. DE. (1987).
Evaluating the toxicity of urban patterns of oxidant gases. II. Effects in mice from chronic exposure to
nitrogen dioxide. J Toxicol Environ Health 21: 99-112. http://dx.doi.org/10.1080/15287398709531005
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.
January 2015 4-76 DRAFT: Do Not Cite or Quote
-------
Miller. FJ: Overtoil. JH. Jr: Jaskot RH: Menzel DB. (1985). A model 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. FJ: Overton. JH: Myers. ET: Graham. JA. (1982). Pulmonary dosimetry of nitrogen dioxide in
animals and man. In T Schneider; L Grant (Eds.), Air pollution by nitrogen oxides (pp. 377-386).
Amsterdam: Elsevier Scientific Publishing Company.
Miyanishi K: Kinouchj T: Kataoka. K: Kanoh. T: Ohnishj Y. (1996). In vivo formation of mutagens by
intraperitoneal administration of polycyclic aromatic hydrocarbons in animals during exposure to
nitrogen dioxide. Carcinogenesis 17: 1483-1490.
Mohsenin. V. (1987). Effect of vitamin C on NO2-induced airway hyperresponsiveness in normal
subjects: Arandomized double-blind experiment. Am Rev Respir Dis 136: 1408-1411.
http://dx.doi.0rg/10.1164/ajrccm/136.6.1408
Mohsenin. V. (1991). Lipid peroxidation and antielastase activity in the lung under oxidant stress: Role of
antioxidant defenses. J Appl Physiol 70: 1456-1462.
Miiller. B: Schafer. H: Barth. P; von Wichert. P. (1994). Lung surfactant components in bronchoalveolar
lavage after inhalation of NO2 as markers of altered surfactant metabolism. Lung 172: 61-72.
http://dx.doi.org/10.1007/BF00185077
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.
http://dx.doi.org/10.1080/00028896409342583
Neta. P: Huie. RE: Ross. AB. (1988). Rate constants for reactions of inorganic radicals in aqueous
solution. J Phys Chem Ref Data 17: 1027-1284. http://dx.doi.Org/10.1063/l.555808
Nettesheim. P: Hanna. MG. Jr: Dohertv. DG: Newell. RF: Hellman. A. (1970). Effect of chronic exposure
to air pollutants on the respiratory tracts of mice: Histopathological findings. In P Nettesheim; MG
Hanna, Jr.; JW Deatherage, Jr. (Eds.), Morphology of experimental respiratory carcinogenesis:
proceedings of a biology division, Oak Ridge National Laboratory conference; May; Gatlinburg, TN.
Oak Ridge, TN: U.S. Atomic Energy Commission.
Ng. AW; Bidani. A; Heming. TA. (2004). Innate host defense of the lung: effects of lung-lining fluid pH
[Review]. Lung 182: 297-317. http://dx.doi.org/10.1007/s00408-004-2511-6
Noda. N; Takeno. S; Fukuiri. T; Hirakawa. K. (2012). Monitoring of oral and nasal exhaled nitric oxide in
eosinophilic chronic rhinosinusitis: A prospective study. Am J Rhinol Allergy 26: 255-259.
http://dx.doi.org/10.2500/aira.2012.26.3772
North. ML; Khanna. N; Marsden. PA; Grasemann. H; Scott. JA. (2009). Functionally important role for
arginase 1 in the airway hyperresponsiveness of asthma. Am J Physiol Lung Cell Mol Physiol 296:
L911-L920. http://dx.doi.org/10.1152/ajplung.00025.2009
O'Byrne. PM; Gauvreau. GM; Brannan. JD. (2009). Provoked models of asthma: What have we learnt?
[Review]. Clin Exp Allergy 39: 181-192. http://dx.doi.Org/10.llll/j.1365-2222.2008.03172.x
O'Donnell VB; Eiserich. JP; Chumlev. PH; Jablonsky. MJ; Krishna. NR; Kirk. M; Barnes. S; Parley -
Usmar. VM; Freeman. BA. (1999). Nitration of unsaturated fatty acids by nitric oxide-derived reactive
nitrogen species peroxynitrite, nitrous acid, nitrogen dioxide, and nitronium ion. Chem Res Toxicol
12: 83-92. http://dx.doi.org/10.1021/tx980207u
Oda. H; Tsubone. H; Suzuki. A; Ichinose. T; Kubota. K. (1981). Alterations of nitrite and nitrate
concentrations in the blood of mice exposed to nitrogen dioxide. Environ Res 25: 294-301.
Ohashj Y; Nakai. Y; Sugiura. Y; Ohno. Y; Okamoto. H; Tanaka. A; Kakinoki. Y; Havashi. M. (1994).
Nitrogen dioxide-induced eosinophilia and mucosal injury in the nose of the guinea pig. Acta
Otolaryngol 114: 547-551.
January 2015 4-77 DRAFT: Do Not Cite or Quote
-------
Olin. AC: Aldenbratt A: Ekman. A: Ljungkvist. G: Jungersten. L: Alving. K: Toren. K. (2001). Increased
nitric oxide in exhaled air after intake of a nitrate-rich meal. Respir Med 95: 153-158.
http://dx.doi.org/10.1053/rmed.2000.1010
Olin. AC: Bake. B: Toren. K. (2007). Fraction of exhaled nitric oxide at 50 mL/s: Reference values for
adult lifelong never-smokers. Chest 131: 1852-1856. http://dx.doi.org/10.1378/chest.06-2928
Olin. AC: Hellgren. J: Karlsson. G: Ljungkvist. G: Nolkrantz. K: Toren. K. (1998). Nasal nitric oxide and
its relationship to nasal symptoms, smoking and nasal nitrate. Rhinology 36: 117-121.
Overton. JH. Jr. (1984). Physicochemical processes and the formulation of dosimetry models. In FJ
Miller; DB Menzel (Eds.), Fundamentals of extrapolation modeling of inhaled toxicants: Ozone and
nitrogen dioxide (pp. 93-114). Washington, DC: Hemisphere Publishing Corporation.
Pacher. P: Beckman. JS: Liaudet L. (2007). Nitric oxide and peroxynitrite in health and disease
[Review]. Physiol Rev 87: 315-424. http://dx.doi.org/10.1152/phvsrev.00029.2006
Paganj P: Romano. M: Erroi. A: Ferro. M: Salmona. M. (1994). Biochemical effects of acute and
subacute nitrogen dioxide exposure in rat lung and bronchoalveolar lavage fluid. Arch Environ
Contam Toxicol 27: 426-430. http://dx.doi.org/10.1007/BF00213182
Pathmanathan. S: Krishna. MT: Blomberg. A: Helledav. R: Kelly. FJ: Sandstrom. T: Holgate. ST: Wilson.
SJ: Frew. AJ. (2003). Repeated daily exposure to 2 ppm nitrogen dioxide upregulates the expression of
IL-5, IL-10, IL-13, and ICAM-1 in the bronchial epithelium of healthy human airways. Occup
Environ Med 60: 892-896. http://dx.doi.org/10.1136/oem.60.ll.892
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. http://dx.doi.0rg/10.1165/aircmb.19.3.3183
Polidori. A: Fine. PM. (2012). Ambient concentrations of criteria and air toxic pollutants in close
proximity to a freeway with heavy-duty diesel traffic. Final report. South Coast Air Quality
Management District, http://www.aqmd.gov/docs/default-source/air-qualitv/air-qualitv-monitoring-
studies/near-roadwav-studv.pdf?sfvrsn=2
Posin. C: Clark. K: Jones. MP: Patterson. JV: Buckley. RD: Hackney. JD. (1978). Nitrogen dioxide
inhalation and human blood biochemistry. Arch Environ Occup Health 33: 318-324.
Postlethwait EM: Bidani. A. (1989). Pulmonary disposition of inhaled NO2-nitrogen in isolated rat lungs.
Toxicol Appl Pharmacol 98: 303-312.
Postlethwait. EM: Bidani. A. (1990). Reactive uptake governs the pulmonary air space removal of inhaled
nitrogen dioxide. J Appl Physiol 68: 594-603.
Postlethwait. EM: Bidani. A. (1994). Mechanisms of pulmonary NO2 absorption [Review]. Toxicology
89: 217-237.
Postlethwait. EM: Bioani. A: Evans. MJ. (1990). The effect of NO2 exposure on perfusate distribution in
isolated rat lungs: pulmonary versus bronchial circulation. Toxicol Appl Pharmacol 106: 456-461.
http://dx.doi.org/10.1016/0041-008X(90)90340-Z
Postlethwait. EM: Langford. SD: Bidani. A. (1991a). Interfacial transfer kinetics of NO2 into pulmonary
epithelial lining fluid. J Appl Physiol 71: 1502-1510.
Postlethwait. EM: Langford. SD: Bidani. A. (1991b). Transfer of NO2 through pulmonary epithelial
lining fluid. Toxicol Appl Pharmacol 109: 464-471.
Postlethwait. EM: Langford. SD: Bidani. A. (1992). Kinetics of NO2 air space absorption in isolated rat
lungs. J Appl Physiol 73: 1939-1945.
Postlethwait. EM: Langford. SD: Jacobson. LM: Bidani. A. (1995). NO2 reactive absorption substrates in
rat pulmonary surface lining fluids. Free Radic Biol Med 19: 553-563. http://dx.doi.org/10.1016/0891-
5849(95)00058-6
January 2015 4-78 DRAFT: Do Not Cite or Quote
-------
Postlethwait EM: Mustafa. MG. (1981). Fate of inhaled nitrogen dioxide in isolated perfused rat lung. J
Toxicol Environ Health 7: 861-872. http://dx.doi.org/10.1080/15287398109530029
Postlethwait. EM: Mustafa. MG. (1989). Effect of altered dose rate on NO2 uptake and transformation in
isolated lungs. J Toxicol Environ Health 26: 497-507. http://dx.doi.org/10.1080/15287398909531272
Povnter. ME: Persinger. RL: Irvin. CG: Butnor. KJ: van Hirtum. H: Slav. W: Heintz. NH: Robbins. J:
Hemenwav. D: Taatjes. DJ: Janssen-Heininger. Y. (2006). Nitrogen dioxide enhances allergic airway
inflammation and hyperresponsiveness in the mouse. Am J Physiol Lung Cell Mol Physiol 290: L144-
L152. http://dx.doi.org/10.1152/aiplung.00131.2005
Proust B: Lacroix. G: Robidel. F: Marliere. M: Lecomte. A: Vargaftig. BB. (2002). Interference of a
short-term exposure to nitrogen dioxide with allergic airways responses to allergenic challenges in
BALB/c mice. Mediators Inflamm 11: 251-260. http://dx.doi.org/10.1080/096293502900000113
Priitz. WA: Monig. H: Butler. J: Land. EJ. (1985). Reactions of nitrogen dioxide in aqueous model
systems: Oxidation of tyrosine units in peptides and proteins. Arch Biochem Biophys 243: 125-134.
http://dx.doi.org/10.1016/0003-9861(85)90780-5
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: Houk. KN: Foote. CS: Fukuto. JM: Ignarro. LJ: Squadrito. GL: Davies. KJA. (2006). Free
radical biology and medicine: It's a gas, man! [Review]. Am J Physiol Regul Integr Comp Physiol
291: R491-R511. http://dx.doi.org/10.1152/ajpregu.00614.2005
Pryor. WA: Lightsev. JW. (1981). Mechanisms of nitrogen dioxide reactions: Initiation of lipid
peroxidation and the production of nitrous acid. Science 214: 435-437.
http://dx.doi.org/10.1126/science.214.4519.435
Que. LG: Yang. Z: Stamler. JS: Lugogo. NL: Kraft. M. (2009). S-nitrosoglutathione reductase: An
important regulator in human asthma. Am J Respir Crit Care Med 180: 226-231.
http://dx.doi.org/10.1164/rccm.200901-0158OC
Rava. M: Varraso. R: Decoster. B: Huyvaert H: Le Moual N: Jacquemin. B: Kiinzlj N: Kauffmann. F:
Zerimech. F: Matran. R: Nadif. R. (2012). Plasma and exhaled breath condensate nitrite-nitrate level
in relation to environmental exposures in adults in the EGEA study. Nitric Oxide 27: 169-175.
http://dx.doi.0rg/10.1016/j.niox.2012.06.007
Robison. TW: Forman. HJ. (1993). Dual effect of nitrogen dioxide on rat alveolar macrophage
arachidonate metabolism. Exp Lung Res 19: 21-36.
Robison. TW: Murphy. JK: Beyer. LL: Richters. A: Forman. HJ. (1993). Depression of stimulated
arachidonate metabolism and superoxide production in rat alveolar macrophages following in vivo
exposure to 0.5 ppm NO2. J Toxicol Environ Health 38: 273-292.
http://dx.doi.org/10.1080/15287399309531718
Rose. RM: Pinkston. P: Skornik. WA. (1989). Altered susceptibility to viral respiratory infection during
short-term exposure to nitrogen dioxide. Res Rep Health Eff Instl-24.
Rubenchik. BL: Glavin. AA: Galenko. PM: Kilkichko. AA: Olenick. IP: Artemov. KV. (1995). Gaseous
nitrogen dioxide increases the endogenous synthesis of carcinogenic N-nitrosodimethylamine in
animals. J Environ Pathol Toxicol Oncol 14: 111-115.
Rubinstein. I: Reiss. TF: Bigbv. BG: Stites. DP: Boushev. HA. Jr. (1991). Effects of 0.60 PPM nitrogen
dioxide on circulating and bronchoalveolar lavage lymphocyte phenotypes in healthy subjects.
Environ Res 55: 18-30.
Rudolph. V: Rudolph. TK: Schopfer. FJ: Bonacci G: Woodcock. SR: Cole. MP: Baker. PR: Ramani. R:
Freeman. BA. (2010). Endogenous generation and protective effects of nitro-fatty acids in a murine
model of focal cardiac ischaemia and reperfusion. Cardiovasc Res 85: 155-166.
http://dx.doi.org/10.1093/cvr/cvp275
January 2015 4-79 DRAFT: Do Not Cite or Quote
-------
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
Sandstrom. T: Andersson. MC: Kolmodin-Hedman. B: Stjernberg. N: Angstrom. T. (1990).
Bronchoalveolar mastocytosis and lymphocytosis after nitrogen dioxide exposure in man: a time-
kinetic study. Eur Respir J 3: 138-143.
Sandstrom. T: Helledav. R: Bjermer. L: Stjernberg. N. (1992). Effects of repeated exposure to 4 ppm
nitrogen dioxide on bronchoalveolar lymphocyte subsets and macrophages in healthy men. Eur Respir
J5: 1092-1096.
Sandstrom. T: Ledin. MC: Thomasson. L: Helledav. R: Stjernberg. N. (1992). Reductions in lymphocyte
subpopulations after repeated exposure to 1.5 ppm nitrogen dioxide. Br J Ind Med 49: 850-854.
Sandstrom. T: Stjernberg. N: Eklund. A: Ledin. MC: Bjermer. L: Kolmodin-Hedman. B: Lindstrom. K:
Rosenhall L: Angstrom. T. (1991). Inflammatory cell response in bronchoalveolar lavage fluid after
nitrogen dioxide exposure of healthy subjects: A dose-response study. Eur Respir J 4: 332-339.
Sarangapanj 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. InhalToxicol 15: 987-1016. http://dx.doi.org/10.1080/08958370390226350
Scaife. A: Barclay. J: Hillis. GS: Srinivasan. J: Macdonald. DW: Ross. JA: Ayres. JG. (2012). Lack of
effect of nitrogen dioxide exposure on heart rate variability in patients with stable coronary heart
disease and impaired left ventricular systolic function. Occup Environ Med 69: 587-591.
http://dx.doi.org/10.1136/oemed-2011-100126
Schlesinger. RB: Driscoll. KE: Gunnison. AF: Zelikoff. JT. (1990). Pulmonary arachidonic acid
metabolism following acute exposures to ozone and nitrogen dioxide. J Toxicol Environ Health 31:
275-290. http://dx.doi.org/10.1080/15287399009531456
Schlesinger. RB: Driscoll. KE: VoUmuth. TA. (1987). Effect of repeated exposures to nitrogen dioxide
and sulfuric acid mist alone or in combination on mucociliary clearance from the lungs of rabbits.
Environ Res 44: 294-301. http://dx.doi.org/10.1016/80013-9351(87)80238-4
Schlesinger. RB: Gearhart. JM. (1987). Intermittent exposures to mixed atmospheres of nitrogen dioxide
and sulfuric acid: effect on particle clearance from the respiratory region of rabbit lungs. Toxicology
44: 309-319. http://dx.doi.org/10.1016/0300-483X(87)90032-1
Schwartz. SE: White. WH. (1983). Kinetics of reactive dissolution of nitrogen oxides into aqueous
solution. In SE Schwartz (Ed.), Trace atmospheric constituents: Properties, transformations, and fates
(pp. 1-116). New York, NY: Wiley.
See. KG: Christianj DC. (2013). Normal values and thresholds for the clinical interpretation of exhaled
nitric oxide levels in the US general population: Results from the National Health and Nutrition
Examination Survey 2007-2010. Chest 143: 107-116. http://dx.doi.org/10.1378/chest.12-0416
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
Sevanian. A: Elsaved. N: Hacker. AD. (1982). Effects of vitamin E deficiency and nitrogen dioxide
exposure on lung lipid peroxidation: Use of lipid epoxides and malonaldehyde as measures of
peroxidation. J Toxicol Environ Health 10: 743-756. http://dx.doi.org/10.1080/15287398209530292
Sherwin. RP: Carlson. DA. (1973). Protein content of lung lavage fluid of guinea pigs exposed to 0.4 ppm
nitrogen dioxide. Arch Environ Health 27: 90-93.
Sies. H. (2004). Ozone in arteriosclerotic plaques: searching for the "smoking gun". Angew
Chem Int Ed Engl 43: 3514-3515. http://dx.doi.org/10.1002/anie.200460118
January 2015 4-80 DRAFT: Do Not Cite or Quote
-------
Silbaugh. SA: Mauderly. JL: Macken. CA. (1981). Effects of sulfuric acid and nitrogen dioxide on airway
responsiveness of the guinea pig. J Toxicol Environ Health 8: 31-45.
http://dx.doi.org/10.1080/15287398109530048
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.
http://dx.doi.org/10.3109/01902149309064358
Smith. LL. (2004). Oxygen, oxysterols, ouabain, and ozone: A cautionary tale [Review]. Free Radic Biol
Med 37: 318-324. http://dx.doi.Org/10.1016/i.freeradbiomed.2004.04.024
Song. Y: Namkung. W: Nielson. DW: Lee. JW: Finkbeiner. WE: Verkman. AS. (2009). Airway surface
liquid depth measured in ex vivo fragments of pig and human trachea: Dependence on Na+ and Cl-
channel function. Am J Physiol Lung Cell Mol Physiol 297: L1131-L1140.
http://dx.doi.org/10.1152/ajplung.00085.2009
Squadrito. GL: Postlethwait. EM. (2009). On the hydrophobicity of nitrogen dioxide: could there be a
"lens" effect for NO(2) reaction kinetics. Nitric Oxide 21: 104-109.
http://dx.doi.0rg/10.1016/j.niox.2009.05.008
Squadrito. GL: Postlethwait. EM: Matalon. S. (2010). Elucidating mechanisms of chlorine toxicity:
Reaction kinetics, thermodynamics, and physiological implications [Review]. Am J Physiol Lung Cell
Mol Physiol 299: L289-L300. http://dx.doi.org/10.1152/aiplung.00077.2010
Stenger. MR: Rose. MJ: Joshj MS: Rogers. LK: Chicoine. LG: Bauer. JA: Nelin. LD. (2010). Inhaled
nitric oxide prevents 3-nitrotyrosine formation in the lungs of neonatal mice exposed to >95%
oxygen. Lung 188: 217-227. http://dx.doi.org/10.1007/s00408-010-9235-6
Strand. V: Svartengren. M: Rak. S: Barck. C: Bylin. G. (1998). Repeated exposure to an ambient level of
NO2 enhances asthmatic response to a nonsymptomatic allergen dose. Eur Respir J 12: 6-12.
http://dx.doi.org/10.1183/09031936.98.12010006
Suzuki. AK: Tsubone. H: Ichinose. T: Oda. H: Kubota. K. (1981). [Effects of subchronic nitrogen dioxide
exposure on arterial blood pHa, PaCO2 and PaO2 in rats]. Nippon Eiseigaku Zasshi 36: 816-823.
Suzuki. T: Ikeda. S: Kanoh. T: Mizoguchi. I. (1986). Decreased phagocytosis and superoxide anion
production in alveolar macrophages of rats exposed to nitrogen dioxide. Arch Environ Contam Toxicol
15: 733-739. http://dx.doi.org/10.1007/BF01054920
Taylor. DR. (2012). Advances in the clinical applications of exhaled nitric oxide measurements [Review].
J Breath Res 6: 047102. http://dx.doi.Org/10.1088/1752-7155/6/4/047102
Thomas. HV: Mueller. PK: Wright. R. (1967). Response of rat lung mast cells to nitrogen dioxide
inhalation. J Air Pollut Control Assoc 17: 33-35. http://dx.doi.org/10.1080/00022470.1967.10468939
Torok. JA: Brahmajothi. MV: Zhu. H: Tinch. BT: Auten. RL: McMahon. TJ. (2012). Transpulmonary flux
of S-nitrosothiols and pulmonary vasodilation during nitric oxide inhalation: role of transport. Am J
Respir Cell Mol Biol 47: 37-43. http://dx.doi.org/10.1165/rcmb.2011-0439OC
Tsubone. H: Oda. H: Suzuki. AK: Kubota. K. (1982). Electrocardiographic abnormalities in rats by acute
exposure to nitrogen dioxide. Toxicol Lett 12: 125-129. http://dx.doi.org/10.1016/0378-
4274(82)90174-6
Tunnicliffe. WS: Surge. PS: Ayres. JG. (1994). Effect of domestic concentrations of nitrogen dioxide on
airway responses to inhaled allergen in asthmatic patients. Lancet 344: 1733-1736.
http://dx.doi.org/10.1016/s0140-6736(94)92886-x
U.S. EPA (U.S. Environmental Protection Agency). (1993). Air quality criteria for oxides of nitrogen, vol.
1-3 [EPAReport]. (EPA/600/8-91/049aF-cF). Research Triangle Park, NC: U.S. Environmental
Protection Agency, Environmental Criteria and Assessment Office.
http://cfpub.epa. gov/ncea/cfm/recordisplav.cfm?deid=40179
January 2015 4-81 DRAFT: Do Not Cite or Quote
-------
U.S. EPA (U.S. Environmental Protection Agency). (2005). Guidelines for carcinogen risk assessment.
(EPA/63O/P-03/00IF). Washington, DC: U.S. Environmental Protection Agency, Risk Assessment
Forum, http://www.epa.gov/cancerguidelines/
U.S. EPA (U.S. Environmental Protection Agency). (2008). Integrated science assessment for oxides of
nitrogen Health criteria [EPA Report]. (EPA/600/R-08/071). Research Triangle Park, NC: U.S.
Environmental Protection Agency, National Center for Environmental Assessment.
http://cfpub.epa.gov/ncea/cfm/recordisplav.cfm?deid=194645
U.S. EPA (U.S. Environmental Protection Agency). (2013). Integrated science assessment for ozone and
related photochemical oxidants (final report) [EPA Report]. (EPA/600/R-10/076F). Washington, DC.
http://cfpub.epa.gov/ncea/isa/recordisplav.cfm?deid=247492
Van As. A. (1977). Pulmonary airway clearance mechanisms: A reappraisal. Am Rev Respir Dis 115: 721-
726.
van der Heijden. HHA. CM: Brouwer. ML: Hoekstra. F: van der Pol P: Merkus. PJF. M. (2014).
Reference values of exhaled nitric oxide in healthy children 1-5 years using off-line tidal breathing.
Pediatr Pulmonol 49: 291-295. http://dx.doi.org/10.1002/ppul.22796
Van Der Vliet. A: O'Neill. CA: Cross. CE: Koostra. JM: Volz. WG: Halliwell. B: Louie. S. (1999).
Determination of low-molecular-mass antioxidant concentrations in human respiratory tract lining
fluids. Am J Physiol 276: L289-L296.
Van Stee. EW: Sloane. RA: Simmons. JE: Brunnemann. KD. (1983). In vivo formation of N-
nitrosomorpholine in CD-I mice exposed by inhalation to nitrogen dioxide and by gavage to
morpholine. J Natl Cancer Inst 70: 375-379.
Van Stee. EW: Sloane. RA: Simmons. JE: Moorman. MP: Brunnemann. KD. (1995). Endogenous
formation of N-nitrosomorpholine in mice from 15NO2 by inhalation and morpholine by gavage.
Carcinogenesis 16: 89-92.
Velsor. LW: Ballinger. CA: Patel J: Postlethwait EM. (2003). Influence of epithelial lining fluid lipids on
NO2-induced membrane oxidation and nitration. Free Radic Biol Med 34: 720-733.
http://dx.doi.org/10.1016/S0891-5849(02)01370-9
Velsor. LW: Postlethwait. EM. (1997). NO2-induced generation of extracellular reactive oxygen is
mediated by epithelial lining layer antioxidants. Am J Physiol 17: L1265-L1275.
Vints. AM: Oostveen. E: Eeckhaut G: Smolders. M: De Backer. WA. (2005). Time-dependent effect of
nitrate-rich meals on exhaled nitric oxide in healthy subjects. Chest 128: 2465-2470.
http://dx.doi.0rg/10.1378/chest.128.4.2465
von Nieding. G: Wagner. HM. (1979). Effects of NO2 on chronic bronchitics. Environ Health Perspect
29: 137-142.
Wagner. HM. (1970). Absorption of NO and NO2 in MIK- and MAK-concentrations during inhalation.
[Absorption von NO und NO2 in MIK- und MAK-konzentrationen bei der inhalation]. Staub,
Reinhaltung der Luft 30: 380-381.
Wang. J: Widlansky. ME. (2009). Lifestyle choices and endothelial function: Risk and relevance
[Review]. Curr Vase Pharmacol 7: 209-224. http://dx.doi.org/10.2174/157016109787455581
Wang. JH: Devalia. JL: Puddle. JM: Hamilton. SA: Davies. RJ. (1995a). Effect of six-hour exposure to
nitrogen dioxide on early-phase nasal response to allergen challenge in patients with a history of
seasonal allergic rhinitis. J Allergy Clin Immunol 96: 669-676.
Wang. JH: Puddle. J: Devalia. JL: Davies. RJ. (1995b). Nitrogen dioxide increases eosinophil activation
in the early-phase response to nasal allergen provocation. Int Arch Allergy Immunol 107: 103-105.
http://dx.doi.org/10.1159/000236944
Wardman. P. (1998). Nitrogen dioxide in biology: Correlating chemical kinetics with biological effects. In
ZB Alfassi (Ed.), N-centered radicals (pp. 155-179). Chichester, UK: Wiley.
January 2015 4-82 DRAFT: Do Not Cite or Quote
-------
Weitzberg. E: Hezel. M: Lundberg. JO. (2010). Nitrate-nitrite-nitric oxide pathway: Implications for
anesthesiology and intensive care [Review]. Anesthesiology 113: 1460-1475.
http://dx.doi.org/10.1097/ALN.Ob013e3181fcf3cc
Weitzberg. E: Lundberg. JO. (2013). Novel aspects of dietary nitrate and human health [Review]. Annu
Rev Nutr 33: 129-159. http://dx.doi.org/10.1146/annurev-nutr-071812-161159
Wennmalm. A: Benthin. G: Edlund. A: Jungersten. L: Kieler-Jensen. N: Lundin. S: Westfelt. UN:
Petersson. AS: Waagstein. F. (1993). Metabolism and excretion of nitric oxide in humans. An
experimental and clinical study. Circ Res 73: 1121-1127. http://dx.doi.Org/10.1161/01.RES.73.6.1121
Widdicombe. JH. (2002). Regulation of the depth and composition of airway surface liquid [Review]. J
Anat 201: 313-318. http://dx.doi.0rg/10.1046/i.1469-7580.2002.00098.x
Widdicombe. JH: Widdicombe. JG. (1995). Regulation of human airway surface liquid [Review]. Respir
Physiol 99: 3-12. http://dx.doi.org/10.1016/0034-5687(94)00095-H
Witten. A: Solomon. C: Abbritti. E: Arjomandi. M: Zhai. W: Kleinman. M: Balmes. J. (2005). Effects of
nitrogen dioxide on allergic airway responses in subjects with asthma. J Occup Environ Med 47:
1250-1259. http://dx.doi.org/10.1097/01.iom.0000177081.62204.8d
Yadav. AK: Doran. SF: Samal AA: Sharma. R: Vedagiri. K: Postlethwait. EM: Squadrito. GL: Fanucchi.
MV: Roberts. LJ: Patel RP: Matalon. S. (2011). Mitigation of chlorine gas lung injury in rats by
postexposure administration of sodium nitrite. Am J Physiol Lung Cell Mol Physiol 300: L362-L369.
http://dx.doi.org/10.1152/aiplung.00278.2010
Yokoyama. E. (1968). Uptake of SO2 and NO2 by the isolated upper airways. Koshu Eisei Kenkyu 17:
302-306.
Zetterquist W: Pedroletti. C: Lundberg. JO: Alving. K. (1999). Salivary contribution to exhaled nitric
oxide. Eur Respir J 13: 327-333. http://dx.doi.Org/10.1034/i.1399-3003.1999.13bl8.x
Zhang. J: Villacorta. L: Chang. L: Fan. Z: Hamblin. M: Zhu. T: Chen. CS: Cole. MP: Schopfer. FJ: Deng.
CX: Garcia-Barrio. MT: Feng. YH: Freeman. BA: Chen. YE. (2010a). Nitro-oleic acid inhibits
angiotensin II-induced hypertension. Circ Res 107: 540-548.
http://dx.doi.org/10.1161/CIRCRESAHA.110.218404
Zhang. Y: Ma. C: Xiao. Y: Liu. C: Xu. O: Wang. Y: Xu. X: Hao. L. (2010b). Dual role of vitamin C
utilization in NO2-induced oxidative stress in lung tissues of mice. Bull Environ Contam Toxicol 84:
662-666. http://dx.doi.org/10.1007/s00128-010-0021-l
Zhu. Y: Fung. DC: Kennedy. N: Hinds. WC: Eiguren-Fernandez. A. (2008). Measurements of ultrafine
particles and other vehicular pollutants inside a mobile exposure system on Los Angeles freeways. J
Air Waste ManagAssoc 58: 424-434.
January 2015 4-83 DRAFT: Do Not Cite or Quote
-------
CHAPTER 5 INTEGRATED HEALTH EFFECTS
OF SHORT-TERM EXPOSURE TO OXIDES OF
NITROGEN
5.1 Introduction
5.1.1 Scope of Chapter
1 The preceding chapters describe the widespread potential for human exposure to ambient
2 oxides of nitrogen (Chapters 2 and 1) and the capability for ambient-relevant
3 concentrations of inhaled NO2 to initiate a cascade of molecular and cellular responses,
4 particularly in the airways (Chapter 4). These lines of evidence point to the potential for
5 ambient exposure to oxides of nitrogen to induce health effects. However, the preceding
6 chapters also identify the importance of assessing exposure measurement error due to
7 heterogeneity in ambient concentrations of oxides of nitrogen, effects of other correlated
8 pollutants, and the extent to which mode-of-action information is available to support
9 biological plausibility. With consideration of these issues, this chapter summarizes,
10 integrates, and evaluates the evidence for relationships between various health effects and
11 short-term (i.e., minutes up to 1 month, Section 1.5) exposure to oxides of nitrogen. The
12 chapter sections comprise evaluations of the epidemiologic, controlled human exposure,
13 and animal toxicological evidence for the effects of short-term exposure to oxides of
14 nitrogen on health outcomes related to respiratory effects (Section 5.2). cardiovascular
15 and related metabolic effects (Section 5.3). and total mortality (Section 5.4).
16 Reproductive and developmental effects also have been examined in relation to
17 short-term exposure to oxides of nitrogen. However, this evidence is evaluated with
18 studies of long-term exposure in Chapter 6 because associations are often compared
19 among various short- and long-term exposure periods that are difficult to distinguish.
20 Individual sections for broad health categories (i.e., respiratory, cardiovascular, mortality)
21 begin with a summary of conclusions from the 2008 ISA for Oxides of Nitrogen followed
22 by an evaluation of recent (i.e., published since the completion of the 2008 ISA for
23 Oxides of Nitrogen) studies that builds upon evidence from previous reviews. Within
24 each of these sections, results are organized into smaller outcome groups [e.g., asthma
25 exacerbation, myocardial infarction (MI)] that comprise a continuum of subclinical to
26 clinical effects. The discussion of individual events and outcomes is then organized by
27 specific scientific discipline (i.e., epidemiology, controlled human exposure, toxicology).
28 This organization permits clear description of the extent of coherence and biological
January 2015 5-1 DRAFT: Do Not Cite or Quote
-------
1 plausibility for the effects of oxides of nitrogen on a group of related outcomes, and in
2 turn, transparent characterization of the weight of evidence in drawing conclusions.
3 Sections for each of the broad health categories conclude with an integrated assessment
4 of the evidence and a conclusion regarding causality. A determination of causality was
5 made for each broad health category by evaluating the evidence for each category
6 independently with the causal framework (described in the Preamble to the ISA).
7 Findings for mortality informed multiple causal determinations. Findings for
8 cause-specific mortality (i.e., respiratory, cardiovascular) were used to assess the
9 continuum of effects and inform the causal determinations for respiratory and
10 cardiovascular and related metabolic effects. A separate causal determination was made
11 for total mortality (Section 5.4) based on the evidence for non-accidental causes of
12 mortality combined and also informed by the extent to which evidence for the spectrum
13 of cardiovascular and respiratory effects provides biological plausibility for NO2-related
14 total mortality. Judgments regarding causality were made by evaluating the evidence over
15 the full range of exposures in animal toxicological, controlled human exposure, and
16 epidemiologic studies defined in this ISA to be relevant to ambient exposure [i.e., up to
17 5,000 ppb NO2 or NO] as described in Section 1.2. Experimental studies that examined
18 higher NO2 or NO concentrations were evaluated particularly to inform mode of action.
5.1.2 Evidence Evaluation and Integration to Form Causal Determinations
5.1.2.1 Evaluation of Individual Studies
19 As described in the Preamble to the ISA (Section 5.a), causal determinations were
20 informed by the integration of evidence across scientific disciplines (e.g., exposure,
21 animal toxicology, epidemiology) and related outcomes and judgments of the quality of
22 individual studies. Table 5-1 describes aspects considered in evaluating the quality of
23 controlled human exposure, animal toxicological, and epidemiologic studies. These
24 aspects aid in the assessment of various sources of bias and uncertainty within a study
25 and in turn, judgments about the strength of inference from their results. This evaluation
26 was applied to studies included in this ISA from previous assessments and those
27 published since the 2008 ISA for Oxides of Nitrogen. The study aspects are based on the
28 ARRIVE (Kilkenny etal.. 2010) and STROBE (von Elm et al.. 2007) guidelines for
29 animal experiments and epidemiologic studies, respectively. These guidelines were
30 developed to improve standards of reporting and ensure that the data from animal
31 experiments and epidemiologic studies can be fully evaluated. The aspects found in
32 Table 5-1 are consistent with the questions and criteria that have been proposed by
January 2015 5-2 DRAFT: Do Not Cite or Quote
-------
1
2
o
J
4
5
6
7
8
9
10
11
12
13
14
15
16
previous approaches for evaluating health science data.: Additionally, the aspects are
compatible with published EPA guidelines related to cancer, neurotoxicity, reproductive
toxicity, and developmental toxicity (U.S. EPA. 2005. 1998. 1996. 1991).
The study aspects described in Table 5-1 are not intended to be a complete list that
informs the evaluation of study quality but they comprise the major aspects of study
evaluation considered in this ISA. Where possible, study quality considerations, for
example, exposure assessment and confounding (i.e., bias due to a relationship with the
outcome and correlation with exposures to oxides of nitrogen), are framed to be specific
to oxides of nitrogen. Thus, judgments of the quality of a study can vary depending on
the specific pollutant being assessed. Importantly, these aspects were not used as a
checklist. Particular aspects or the absence of some of these features in a study did not
necessarily define a less informative study or exclude a study from consideration in the
ISA. Further, these aspects were not criteria for a particular determination of causality in
the five-level hierarchy. As described in the Preamble, causal determinations were based
on judgments of the overall strengths and limitations of the collective body of available
studies and the coherence of evidence across scientific disciplines and related outcomes.
Table 5-1 Summary and description of scientific considerations for evaluating
the quality of studies on the health effects of oxides of nitrogen.
NOTE: Study aspects of interest are reported in gray boxes (e.g., Study Design). Summary bullets are provided in the top sections
and are described in more detail in the text immediately following.
Study Design
Controlled Human Exposure
Animal Toxicology
Epidemiology
Clearly defined hypotheses/aims
Appropriately matched control
exposures
Randomization and allocation
concealment
Balanced crossover (repeated
measures) or parallel design
studies
Clearly defined hypotheses/aims
Appropriately matched control
exposures
Randomization and allocation
concealment
All groups handled and cared for
equally
Clearly defined hypotheses/aims
Key designs for short-term
exposure: time series, case
crossover, panel
Key designs for long-term
exposure: prospective cohort,
nested case-control
High power studies key: large
sample sizes, multiple years,
multicity studies
1 For example, NTP OHAT approach (Rooney etal. 2014). IRIS Preamble (U.S. EPA. 2013b). ToxRTool
(Klimisch et al.. 1997).
January 2015
5-3
DRAFT: Do Not Cite or Quote
-------
Table 5-1 (Continued): Summary and description of scientific considerations for
evaluating the quality of studies on the health effects from
oxides of nitrogen.
Controlled Human Exposure:
Studies should clearly describe the primary and any secondary objectives of the study, or specific hypotheses being
tested. Study subjects should be randomly exposed without knowledge of the exposure condition. Preference is
given to balanced crossover (repeated measures) or parallel design studies which include control exposures (e.g.,
to clean filtered air). In crossover studies, a sufficient and specified time between exposure days should be provided
to avoid carry over effects from prior exposure days. In parallel design studies, all arms should be matched for
individual characteristics such as age, sex, race, anthropometric properties, and health status. Similarly, in studies
evaluating effects of disease, appropriately matched healthy controls are desired for interpretative purposes.
Animal Toxicology:
Studies should clearly describe the primary and any secondary objectives of the study, or specific hypotheses being
tested. Studies should include appropriately matched control exposures (e.g., to clean filtered air, time matched).
Studies should use methods to limit differences in baseline characteristics of control and exposure groups. Studies
should randomize assignment to exposure groups and where possible conceal allocation to research personnel.
Groups should be subjected to identical experimental procedures and conditions and care of animals, including
housing, husbandry, etc. Blinding of research personnel to study group may not be possible due to animal welfare
and experimental considerations; however, differences in the monitoring or handling of animals in all groups by
research personnel should be minimized.
Epidemiology:
Studies should clearly describe the primary and any secondary aims of the study, or specific hypotheses being
tested.
For short-term exposure, time-series, case crossover, and panel studies are emphasized over cross-sectional
studies because they examine temporal correlations and are less prone to confounding by factors that differ
between individuals (e.g., SES, age). Studies with large sample sizes and conducted over multiple years are
considered to produce more reliable results. If other quality parameters are equal, multicity studies carry more
weight than single-city studies because they tend to have larger sample sizes and lower potential for publication
bias.
For long-term exposure, inference is considered to be stronger for prospective cohort studies and case-control
studies nested within a cohort (e.g., for rare diseases) than cross-sectional, other case-control, or ecologic studies.
Cohort studies can better inform the temporality of exposure and effect. Other designs can have uncertainty also
related to the appropriateness of the control group or validity of inference about individuals from group-level data.
Study design limitations can bias health effect associations in either direction.
Study Population/Test Model
Controlled Human Exposure
Animal Toxicology
Epidemiology
Similarly matched control and
exposed subjects
Subject characteristics reported
Clearly indicated inclusion and
exclusion criteria
Independent, clinical assessment
of the health condition
Loss or withdrawal of subjects
should be reported with rationale
Animal characteristics reported
Studies testing and reporting both
sexes and multiple life stages
preferred
Loss or exclusion of animals
should be reported with rationale
Representative of population of
interest
High participation and low
drop-out over time that is not
dependent on exposure or health
status
Clearly indicated inclusion and
exclusion criteria
Independent, clinical assessment
of health condition
Groups are compared if from
same source population
January 2015
5-4
DRAFT: Do Not Cite or Quote
-------
Table 5-1 (Continued): Summary and description of scientific considerations for
evaluating the quality of studies on the health effects from
oxides of nitrogen.
Controlled Human Exposure:
In general, the subjects recruited into study groups should be similarly matched for age, sex, race, anthropometric
properties, and health status. In studies evaluating effects of specific subject characteristics (e.g., disease, genetic
polymorphism, etc.), appropriately matched healthy controls are preferred. Relevant characteristics and health
status should be reported for each experimental group. Criteria for including and excluding subjects should be
clearly indicated. For the examination of populations with an underlying health condition (e.g., asthma),
independent, clinical assessment of the health condition is ideal, but self-report of physician diagnosis generally is
considered to be reliable for respiratory diseases and history of cardiovascular events.3 The loss or withdrawal of
recruited subjects during the course of a study should be reported. Specific rationale for excluding subject(s) from
any portion of a protocol should be explained.
Animal Toxicology:
Ideally, studies should report species, strain, sub-strain, genetic background, age, sex, and weight. However,
differences in these parameters across studies do not make the studies incomparable. Unless data indicate
otherwise, all animal species and strains are considered appropriate for evaluating effects of NO2 or NO exposure. It
is preferred that the authors test for effects in both sexes and multiple life stages, and report the result for each
group separately. All animals used in a study should be accounted for, and rationale for exclusion of animals or data
should be specified.
Epidemiology:
The ideal study population is recruited from and is representative of the target population. Studies with high
participation and low drop-out overtime that is not dependent on exposure or health status are considered to have
low potential for selection bias. Criteria for including and excluding subjects should be clearly indicated. For
populations with an underlying health condition, independent, clinical assessment of the health condition is ideal,
but self-report of physician diagnosis generally is considered to be reliable for respiratory diseases and history of
cardiovascular events.3 Groups with and without an underlying health condition should be compared if they are from
the same source population. Selection bias can influence results in either direction or may not affect the validity of
results but rather reduce the generalizability of findings to the target population.
Pollutant
Controlled Human Exposure
Animal Toxicology
Epidemiology
Studies of NO2 are emphasized
Studies of NO2 are emphasized
NO2 emphasized over NO, NOx
Comparisons of health effect
associations among gaseous
oxides of nitrogen species ideal
Controlled Human Exposure:
The focus is on studies testing NO2 exposure.
Animal Toxicology:
The focus is on studies testing NO2 exposure.
Epidemiology:
Health effects are evaluated mostly for NO2, and less so for NO or NOx. Studies that compare health effect
associations among these species are informative. Typically, one species is examined, and studies of NO2 are
emphasized. It is not clear that ambient-relevant NO exposures induce negative health effects (Section 4.2.3). The
relationship of NOx to NO2 varies with distance from roads, and thus, may vary among subjects. Hence, there is
uncertainty about the extent to which associations with NOx reflect those for NO2 vs. other pollutants from traffic.
January 2015
5-5
DRAFT: Do Not Cite or Quote
-------
Table 5-1 (Continued): Summary and description of scientific considerations for
evaluating the quality of studies on the health effects from
oxides of nitrogen.
Exposure Assessment or Assignment
Controlled Human Exposure
Animal Toxicology
Epidemiology
Well characterized and reported
exposure conditions
Limited to studies that utilize NO2
and/or NO concentrations less
than or equal to 5,000 ppb
Preference is given to studies that
include exposure control groups
Randomized exposure groups
Well characterized and reported
exposure conditions
Inhalation exposure
Limited to studies that utilize NO2
and/or NO concentrations less
than or equal to 5,000 ppb
All studies should include
exposure control groups
Randomized exposure groups
Exposure metrics that accurately
represent temporal or spatial
variability for study area
Comparisons of exposure
measurement methods
Indoor and total personal
exposures can inform
independent effects of NO2
Lag/duration of exposure metric
correspond with time course for
health effect
Controlled Human Exposure:
Studies should well characterize pollutant concentration, temperature, and relative humidity and/or have measures
in place to adequately control the exposure conditions for subject safety. For this assessment, the focus is on
studies that utilize NO2 and/or NO concentrations less than or equal to 5,000 ppb (Section 1.2). Studies that utilize
higher exposure concentrations may provide information relevant to MOA, dosimetry, or at-risk human populations.
Preference is given to balanced crossover or parallel design studies which include control exposures (e.g., to clean
filtered air). Study subjects should be randomly exposed without knowledge of the exposure condition. Method of
exposure (e.g., chamber, facemask, etc.) should be specified and activity level of subjects during exposures should
be well characterized.
Animal Toxicology:
Studies should characterize pollutant concentration, temperature, and relative humidity and/or have measures in
place to adequately control the exposure conditions. The focus is on inhalation exposure. Non-inhalation exposure
experiments may provide information relevant to MOA. In vitro studies may be included if they provide mechanistic
insight or examine similar effects as in vivo, but are generally not included. All studies should include exposure
control groups (e.g., clean filtered air). For this assessment, the focus will be on studies that utilize NO2 and/or NO
concentrations less than or equal to 5,000 ppb (Section 1.2). Studies that utilize higher exposure concentrations
may provide information relevant to MOA, dosimetry, interspecies variation, or at-risk human populations.
January 2015
5-6
DRAFT: Do Not Cite or Quote
-------
Table 5-1 (Continued): Summary and description of scientific considerations for
evaluating the quality of studies on the health effects from
oxides of nitrogen.
Epidemiology:
Of primary relevance are relationships of health effects with the ambient component of exposure to oxides of
nitrogen. However, information about ambient exposure rarely is available for individual subjects; most often,
inference is based on ambient concentrations. Studies that compare exposure assessment methods are considered
to be particularly informative. The duration or lag of the exposure metric should correspond with the time course for
physiological changes in the outcome (e.g., up to a few days for symptoms) or latency of disease (e.g., several
years for cancer).
Given the spatial heterogeneity in ambient oxides of nitrogen and variable relationships between personal
exposures and ambient concentrations (Section 3.4.3), validated methods that capture the extent of variability for
the particular study design (temporal vs. spatial contrasts) and location carry greater weight. Central site
measurements, whether averaged across multiple monitors or assigned from the nearest or single available
monitor, have well-recognized limitations in capturing spatial variation in oxides of nitrogen. Inference from central
site measurements can be adequate if correlated with personal exposures, closely located to study subjects, highly
correlated across monitors within a location, used in locations with well-distributed sources, or combined with
time-activity information.
In studies of short-term exposure, metrics that may capture variation in ambient oxides of nitrogen and strengthen
inference include concentrations in subjects' microenvironments (e.g., outdoor home, school, in-vehicle) and
individual-level outdoor concentrations combined with time-activity data. Results for total personal and indoor NO2
exposure are other lines of evidence that inform judgments about causality of NO2 because inference is based on
an individual's microenvironmental exposures and potential for copollutant confounding may be lower or different
than that for ambient concentrations. Results for total personal exposure can inform the effects of ambient exposure
when well correlated with ambient concentrations. For long-term exposures, LUR models that well represent spatial
variation in ambient NO2 can provide estimates of individual exposure. Less weight is placed on NOx from
dispersion models because of limitations in accurate estimation of within-community conditions (Section 3.2.1.2).
And because NOx from dispersion models often shows near perfect correlations (r= 0.94-0.99) with EC, PlVh.s, and
CO, the effects of NOx cannot be distinguished from traffic-related copollutants.
Exposure measurement error often attenuates health effect estimates or increases the precision of the association
(i.e., width of 95% CIs), particularly associations based on temporal variation in short-term exposure
(Section 3.4.5.1). However, exposure measurement error can bias estimates away from the null, particularly for
long-term exposures.
Outcome Assessment/Evaluation
Controlled Human Exposure Animal Toxicology Epidemiology
• Same manner of outcome • Same manner of outcome • Same manner of outcome
assessment for all groups assessment for all groups assessment for all groups
• Validated, reliable methods • Validated, reliable methods • Validated, reliable methods
• Reporting of outcome • Reporting of outcome • Assessment is blind to exposure
assessment details assessment details status
• Blinding of endpoint evaluators • Blinding of endpoint evaluators • Appropriate timing of endpoint
• Appropriate timing of endpoint • Appropriate timing of endpoint evaluation
evaluation evaluation
Controlled Human Exposure:
Endpoints should be assessed in the same manner for control and exposure groups (e.g., time after exposure,
methods, endpoint evaluator) using valid, reliable methods. Blinding of endpoint evaluators is ideal, especially for
qualitative endpoints (e.g., histopathology). For each experiment and each experimental group, including controls,
precise details of all procedures carried out should be provided including how, when, and where. Time of the
endpoint evaluations is a key consideration that will vary depending on the endpoint evaluated. Endpoints should be
assessed at time points that are appropriate for the research questions.
Animal Toxicology:
Endpoints should be assessed in the same manner for control and exposure groups (e.g., time after exposure,
methods, endpoint evaluator) using valid, reliable methods. Blinding of endpoint evaluators is ideal, especially for
qualitative endpoints (e.g., histopathology). For each experiment and each experimental group, including controls,
precise details of all procedures carried out should be provided including how, when, and where. Time of the
endpoint evaluations is a key consideration that will vary depending on the endpoint evaluated. Endpoints should be
assessed at time points that are appropriate for the research questions.
January 2015 5-7 DRAFT: Do Not Cite or Quote
-------
Table 5-1 (Continued): Summary and description of scientific considerations for
evaluating the quality of studies on the health effects from
oxides of nitrogen.
Epidemiology:
Outcomes should be assessed or reported without knowledge of exposure status. Such bias could produce
artifactual associations. Outcomes assessed by interview, self-report, clinical examination, or analysis of biological
indicators should be defined by consistent criteria and collected by validated, reliable methods. Independent, clinical
assessment is ideal for outcomes such as lung function or incidence of disease, but report of physician diagnosis
has shown good reliability.3 Outcomes should be assessed at time intervals that correspond with the time course for
physiological changes (e.g., up to a few days for symptoms). When health effects of long-term exposure are
assessed by acute events such as symptoms or hospital admissions, inference is strengthened when results are
adjusted for short-term exposure. Validated questionnaires for subjective outcomes such as symptoms are regarded
to be reliable,13 particularly when collected frequently and not subject to long recall. For biological samples, the
stability of the compound of interest and the sensitivity and precision of the analytical method is considered.
If not based on knowledge of exposure status, errors in outcome assessment tend to bias results toward the null.
Potential Copollutant Confounding
Controlled Human Exposure Animal Toxicology Epidemiology
• Well-characterized exposure • Well-characterized exposure • Traffic-related copollutants are
key: CO, PM2.s, BC/EC, OC, UFP,
metal PM components, VOCs
• Also considered: PM-io, SO2, Os
Controlled Human Exposure:
Exposure should be well characterized to evaluate independent effects of NO2.
Animal Toxicology:
Exposure should be well characterized to evaluate independent effects of NO2.
Epidemiology:
Not accounting forcopollutant confounding can produce artifactual associations; thus, studies that examine
copollutant confounding carry greater weight. The predominant method is copollutant modeling, which is especially
informative when measurement error is comparable for copollutants and correlations are not high. Interaction and
joint effect models are examined to a lesser extent. Copollutant confounding also can be informed by evaluating
correlations between oxides of nitrogen and copollutants and comparing health associations between gaseous
oxides of nitrogen and copollutants in single-pollutant models if exposure measurement error is comparable among
pollutants. Studies that examine only gaseous oxides of nitrogen are considered poorly to inform the potential for
copollutant confounding. Copollutant confounding is evaluated based on the extent of their correlations typically
observed with oxides of nitrogen and relationships observed with health effects.
Among copollutants, of primary concern are traffic-related pollutants, which include CO, PlVhs, BC/EC, OC, UFP,
metal PM components such as copper, zinc, and iron, as well as VOCs such as benzene, acetaldehyde, toluene,
ethylbenzene, and xylene. Short-term and long-term metrics for these pollutants consistently show moderate to high
correlations with oxides of nitrogen (Figure 3-6). Many traffic-related pollutants also are characterized to have
common modes of action.0 Common key events include formation of secondary oxidation products, inflammation,
and for respiratory effects, increases in airway responsiveness. They also show relationships with many of the
health effects evaluated in this ISAd except as follows. For long-term exposure, there is uncertainty regarding
confounding by UFP because of their short atmospheric lifetime. Also for long-term exposure, CO is not considered
to be an important confounding copollutant for mortality or lung cancer.d
Of less concern is confounding by PM-io, SO2, and Os because they show varying and often lower correlations with
NO2 (Figure 3-6). Os generally is negatively or weakly positively correlated with NO2 but may be a confounding
copollutant where moderate positive correlations are observed. Os and SO2 in particular show similarities with NO2
in mode of action. PM-io, SO2, and Os show relationships with the health effect evaluated in this ISAd except as
follows. For short-term exposure, SO2 is not considered to be a strong confounding copollutant for cardiovascular
effects. For long-term exposure, neither Os norSO2 is considered to be a strong confounding copollutant.
January 2015 5-8 DRAFT: Do Not Cite or Quote
-------
Table 5-1 (Continued): Summary and description of scientific considerations for
evaluating the quality of studies on the health effects from
oxides of nitrogen.
Other Potential Confounding Factors6
Controlled Human Exposure
Animal Toxicology
Epidemiology
Preference given to studies with
adequate control of factors
influencing health response
Preference given to studies with
adequate control of factors
influencing health response
Potential confounders related to
health effect and correlated with
oxides of nitrogen should be
examined
Potential confounders vary by
study design (temporally vs.
spatially correlated) and by health
effects
Controlled Human Exposure:
Preference is given to studies utilizing experimental and control groups that are matched for individual level
characteristics (e.g., body weight, smoking history, age) and time varying factors (e.g., seasonal and diurnal
patterns).
Animal Toxicology:
Preference is given to studies utilizing experimental and control groups that are matched for individual level
characteristics (e.g., body weight, litter size, food and water consumption) and time varying factors (e.g., seasonal
and diurnal patterns).
Epidemiology:
Factors are considered to be potential confounders if demonstrated in the scientific literature to be related to health
effects and correlated with oxides of nitrogen and/or traffic indicators. Not accounting for confounders can produce
artifactual associations; thus, studies that statistically adjust for multiple factors or control for them in the study
design are emphasized. Less weight is placed on studies that adjust for factors that mediate the relationship
between oxides of nitrogen and health effects, which can bias results toward the null. In the absence of information
linking health risk factors to oxides of nitrogen or traffic indicators, a factor may be evaluated as a potential effect
measure modifier, but uncertainty is noted as to its role as a confounder. Confounders vary according to study
design, exposure duration, and health effect and include the following:
For time-series and panel studies of short-term exposure:
• Respiratory Effects—meteorology, day of week, season, medication use, allergen exposure (potential effect
modifier)
• Cardiovascular Effects—meteorology, day of week, season, medication use
• Total Mortality—meteorology, day of week, season, long-term temporal trends
For studies of long-term exposure:
• Respiratory Effects—socioeconomic status, race, age, medication use, smoking, stress
• Cardiovascular, Reproductive, and Development Effects—socioeconomic status, race, age, medication use,
smoking, stress, noise
• Total Mortality—socioeconomic status, race, age, medication use, smoking, comorbid health conditions
• Cancer—socioeconomic status, race, age, occupational exposure
Statistical Methodology
Controlled Human Exposure
Animal Toxicology
Epidemiology
Clearly described and appropriate
statistical methods for the study
design and research question
Preference given to adequately
powered studies
Consideration given to trends in
data and reproducibility
Clearly described and appropriate
statistical methods for the study
design and research question
Preference given to adequately
powered studies
Consideration given to trends in
data and reproducibility
Multivariable regression adjusting
for potential confounders ideal
Exception is multipollutant
models. Multicollinearity can
produce unreliable results
Results based on small sample
sizes can be unreliable
January 2015
5-9
DRAFT: Do Not Cite or Quote
-------
Table 5-1 (Continued): Summary and description of scientific considerations for
evaluating the quality of studies on the health effects from
oxides of nitrogen.
Controlled Human Exposure:
Statistical methods should be clearly described and appropriate for the study design and research question (e.g.,
correction for multiple comparisons). Generally, statistical significance is used to evaluate the findings of controlled
human exposure studies. Detection of statistical significance is influenced by a variety of factors including, but not
limited to, the size of the study, exposure and outcome measurement error, and statistical model specifications.
Sample size is not a criterion for exclusion; ideally, the sample size should provide adequate power to detect
hypothesized effects (e.g., sample sizes less than three are considered less informative). Because statistical tests
have limitations, consideration is given to both trends in data and reproducibility of results.
Animal Toxicology:
Statistical methods should be clearly described and appropriate for the study design and research question (e.g.,
correction for multiple comparisons). Generally, statistical significance is used to evaluate the findings of animal
toxicology studies. Detection of statistical significance is influenced by a variety of factors including, but not limited
to, the size of the study, exposure and outcome measurement error, and statistical model specifications. Sample
size is not a criterion for exclusion; ideally, the sample size should provide adequate power to detect hypothesized
effects (e.g., sample sizes less than three are considered less informative). Because statistical tests have
limitations, consideration is given to both trends in data and reproducibility of results.
Epidemiology:
Multivariable regression models that include potential confounding factors are emphasized. However, multipollutant
models (more than two pollutants) are considered to produce too much uncertainty because of copollutant
collinearity to be informative. Models with interaction terms aid in the evaluation of potential confounding as well as
effect modification. Sensitivity analyses with alternate specifications for potential confounding inform the stability of
findings and aid in judgments of the strength of inference of results. In the case of multiple comparisons,
consistency in the pattern of association can increase confidence that associations were not found by chance alone.
Statistical methods should be appropriate for the power of the study. For example, categorical analyses with small
sample sizes can be prone to bias results toward or away from the null. Statistical tests such as t-tests and
Chi-squared tests are not considered sensitive enough for adequate inferences regarding pollutant-health effect
associations. For all methods, the effect estimate and precision of the estimate (i.e., width of 95% Cl) are important
considerations rather than statistical significance.
BC = black carbon, Cl = confidence interval, CO = carbon monoxide, EC = elemental carbon, ISA = Integrated Science
Assessment, LUR = land use regression, MOA = mode of action, NO = nitric oxide, NO2 = nitrogen dioxide, NOX = sum of NO and
NO2, O3 = ozone, OC = organic carbon, PM = particulate matter, SES = socioeconomic status, SO2 = sulfur dioxide,
UFP = ultrafine particles, VOC = volatile organic compound.
aTorenetal. (1993): (Murgia et al. (2014): Weaklev et al. (2013): Yang et al. (2011): Heckbert et al. (2004): Barr et al. (2002):
Muhaiarine et al. (1997))
"Burnev et al. (1989)
°lnformation on modes of action for NO2 is described in Section 4.3. The characterization of similar modes of action for many
traffic-related pollutants is based on information described in the most recently completed ISAs (U.S. EPA. 2013a. 2010. 2009.
2008b) and the Health Effects Institute's 2010 review of Traffic-related air pollution (HEI. 2010).
Judgments regarding potential confounding by other criteria pollutants are based on studies evaluated in this ISA, causal
determinations made in the most recently completed ISAs (U.S. EPA. 2013a. 2010. 2009. 2008b). as well as recent reviews
published by the Health Effects Institute. Judgments regarding potential confounding by the PM components EC/BC, OC, metals,
and UFP as well as VOCs should not be inferred as conclusions regarding causality. Their consideration is based on associations
with oxides of nitrogen and health effects observed in the studies examined in this ISA and reviews conducted by the Health
Effects Institute (HEI Review Panel on Ultrafine Particles. 2013: HEI. 2010). Judgments regarding potential confounding by PM10
should not be inferred as conclusions regarding causality specifically for that size fraction. The 2009 ISA for PM evaluated PM10
studies but did not form individual causal determinations for that size fraction because PM10 comprises both fine and thoracic
coarse particles.
eMany factors evaluated as potential confounders can be effect measure modifiers (e.g., season, comorbid health condition) or
mediators of health effects related to oxides of nitrogen (comorbid health condition).
January 2015 5-10 DRAFT: Do Not Cite or Quote
-------
5.1.2.2 Evaluation of Confounding in Epidemiologic Studies
1 Epidemiologic studies of short-term exposure to oxides of nitrogen relied primarily on
2 temporal variation in exposure (e.g., day-to-day changes in ambient NCh concentrations)
3 and health effects. Other risk factors for health effects also exhibit similar temporal trends
4 as oxides of nitrogen and include meteorological variables, season, long-term time trends,
5 medication use, and copollutant exposures. These factors and others specified in
6 Table 5-1 are important to evaluate as potential confounders of associations for oxides of
7 nitrogen, particularly given the small effect sizes typically observed. Epidemiologic
8 studies reviewed in this ISA varied in the extent to which they considered potential
9 confounding. Because no single study considered all potential confounders, and not all
10 factors were examined in the collective body of studies, residual confounding by
11 unmeasured factors is possible. Residual confounding also is possible by poorly
12 measured factors. In this ISA, potential confounding was assessed as the extent to which
13 the collection of studies examined factors well documented in the literature to be
14 associated with exposure to oxides of nitrogen and health outcomes.
15 In epidemiologic studies evaluated in this ISA, confounding was assessed primarily using
16 multivariable models that include NO2 concentrations and the putative confounder in the
17 same model. The NO2 effect estimate represents the effect of NO2 keeping the level of
18 the covariate constant. In the ISA, confounding is assessed by examining the change in
19 the magnitude of the effect estimate and width of the 95% confidence interval (CI) for
20 NO2 in multivariable models, not just a change in statistical significance. The limitations
21 of multivariable models are well recognized. If NO2 and the potential confounder are
22 highly correlated, the collinearity (i.e., covariates predict each other) introduced by
23 including them in the same model can misleadingly decrease or increase the magnitude or
24 precision of the effect estimates for NO2 or the potential confounder. Collinearity can
25 occur, for example, if pollutants are from the same sources or are derived from NO2
26 [e.g., ozone (Os)], or if meteorology affects formation of both pollutants. Adding
27 correlated but non-causal variables can produce models that fit the data poorly, and
28 residual confounding is possible if confounders are excluded or poorly measured.
29 For evaluation of copollutant confounding, the predominant method of studies reviewed
30 in this ISA was copollutant models (NO2 plus one copollutant). Inference about the
31 independent effects of NO2 from copollutant models can be limited because the varying
32 spatial distributions of NO2 and the copollutant may not satisfy the assumptions of equal
33 measurement error or constant correlations for NO2 and the copollutant (Gryparis et al..
34 2007). Further, copollutant models for NO2 assumed linear relationships with
35 copollutants, and nonlinear relationships are possible because of varying near road
January 2015 5-11 DRAFT: Do Not Cite or Quote
-------
1 gradients (Figure 3-2). Other methods for evaluating copollutant confounding do not
2 require the aforementioned assumptions, including a hierarchical Bayesian approach that
3 estimates single-pollutant effects in a particular location then combines these
4 single-pollutant effects across locations in a model as the predictor and outcome,
5 respectively (Gryparis et al., 2007; Schwartz and Coull. 2003). Such Bayesian models are
6 unavailable for NO2. Models examining joint effect or interaction terms for NO2 and a
7 copollutant also can inform potential confounding and synergistic effects. These are
8 available only to a limited extent, particularly for traffic-related copollutants. Because
9 examination of copollutant confounding is based largely on copollutant models, their
10 limitations are considered in drawing inferences about independent associations for NO2.
11 Emphasis is placed on results based on exposure assessment methods that likely produce
12 comparable measurement error for NO2 and copollutants such as ambient or total
13 personal and microenvironmental exposure assessment.
5.1.2.3 Additional Considerations for Epidemiologic Studies
14 The ISA presents epidemiologic effect estimates for associations with health outcomes
15 scaled to the same increment of oxide of nitrogen concentration. This standardization
16 increases comparability among studies that scale effect estimates to various changes in
17 concentrations, e.g., interquartile range (IQR) for the study period or an arbitrary unit
18 such as 10 ppb. The increments for standardization vary by averaging time (e.g.,
19 24-h avg, 1-h max) and oxide of nitrogen. For 24-h avg, effect estimates were scaled to a
20 20-ppb increase for NCh or NO and a 40-ppb increase for NOx. For 1-h max, effect
21 estimates were scaled to a 30-ppb increase for NO2, an 80-ppb increase for NO, and a
22 100-ppb increase for NOx. For 8-h max, the increments for standardization are 25 ppb for
23 NO2, 45 ppb for NO, and 65 ppb for NOx. These increments were derived by calculating
24 the U.S.-wide percentile distributions for various averaging times and then calculating the
25 approximate difference between the median (a typical pollution day) and the 95th
26 percentile (a more polluted day) for a given averaging time [see Table 2-1 for 1-h max
27 percentiles and Table S5-1 for 24-h avg and 8-h max percentiles; (U.S. EPA. 2014f)1.
28 There were common exceptions to this standardization method. Averaging times other
29 than 24-h avg or 1-h max were examined, for example, 2-h to 15-h avg. Effect estimates
30 based on these averaging times were not standardized but are presented in the ISA as
31 reported in their respective studies. Some studies reported effect estimates in terms of
32 Mg/m3 increases in oxides of nitrogen, which could be converted to ppb and standardized
33 for NO2 and NO but not NOx. Because the proportions of NO2 and NO are unknown for
34 the various NOx metrics, NOx concentrations could not be converted from (ig/m3 to ppb.
35 Also, data are not available to calculate the percentiles of NOx concentrations in (ig/m3 at
January 2015 5-12 DRAFT: Do Not Cite or Quote
-------
1 a national scale for the U.S. or other countries. Therefore, the ISA presents effect
2 estimates based on (ig/m3 of NOx as they are reported in their respective studies.
5.1.2.4 Integration of Scientific Evidence
3 In addition to quality and strength of inference from individual studies, causal
4 determinations were based on the integration of multiple lines of evidence, which
5 included evaluation of the consistency and coherence of findings across disciplines and
6 related outcomes and the extent to which chance, confounding, and other biases could be
7 ruled out with reasonable confidence. Aspects considered in evidence integration are
8 described in detail in the Preamble, and examples are summarized below. Controlled
9 human exposure and animal toxicological studies can provide direct evidence for health
10 effects related to NO2 or NO exposures. Coherence between experimental and
11 epidemiologic findings can address uncertainties within the collective body of evidence.
12 For example, experimental evidence for effects from a controlled exposure could address
13 whether epidemiologic associations with health outcomes plausibly reflect an
14 independent effect of ambient NO2 exposure or could be confounded by other factors.
15 Experimental studies additionally can provide biological plausibility for observed effects
16 by describing key events within the modes of action. Thus, the integration of evidence
17 across a spectrum of related outcomes and across disciplines was used to inform
18 uncertainties for any particular outcome or discipline due to factors such as chance,
19 publication bias, selection bias, and confounding by copollutant exposures or other
20 factors. The evaluation of health effects also drew upon information on potential error
21 associated with various exposure assessment methods and the uptake and distribution of
22 oxides of nitrogen in the body. The subsequent sections assess study quality and strength
23 of inference and integrate multiple lines of evidence to characterize relationships between
24 oxides of nitrogen and various health effects.
5.2 Respiratory Effects
5.2.1 Introduction
25 The 2008 ISA for Oxides of Nitrogen concluded that evidence was sufficient to infer a
26 likely to be causal relationship between short-term exposure to NO2 and respiratory
27 effects (U.S. EPA. 2008aX based heavily on a large body of epidemiologic evidence. In
28 studies that were not available until after the completion of the 1993 AQCD for Oxides of
29 Nitrogen (U.S. EPA. 1993). short-term increases in ambient NO2 concentrations were
January 2015 5-13 DRAFT: Do Not Cite or Quote
-------
1 consistently associated with increases in respiratory-related hospital admissions and
2 emergency department (ED) visits. The coherence of these findings with NO2-related
3 increases in respiratory symptoms in children with asthma supported an effect of NO2
4 exposure on asthma exacerbation. NO2 was not consistently related to lung function
5 decrements across epidemiologic and controlled human exposure studies and populations
6 with varying respiratory conditions such as asthma or chronic obstructive pulmonary
7 disease (COPD). However, epidemiologic studies of children and adults with asthma
8 observed associations with lung function measured by supervised spirometry (U.S. EPA.
9 2008a).
10 The 2008 ISA identified multiple lines of evidence as supporting an independent
11 relationship between short-term NO2 exposure and respiratory effects. Controlled human
12 exposure studies demonstrated NO2-induced increases in airway responsiveness in adults
13 with asthma. These findings for increased airway responsiveness, a characteristic feature
14 of asthma, provided biological plausibility for epidemiologic evidence for asthma
15 exacerbation. Further, airway responsiveness was increased following <1 to 6-hour
16 exposures to NO2 at concentrations in the range of 100 to 300 ppb, which are not much
17 higher than peak ambient concentrations (Section 2.5).
18 Previous epidemiologic studies also indicated independent associations for NO2. Personal
19 and indoor NO2 were associated with respiratory effects, and associations with both
20 personal and ambient NO2 were observed in copollutant models that adjusted for another
21 traffic-related pollutant such as carbon monoxide (CO) or fine particulate matter (PIVb 5).
22 In the few available results, NO2-related respiratory effects were observed with
23 adjustment for elemental carbon (EC), organic carbon (OC), or ultrafine particles (UFP);
24 other traffic-related copollutants were not examined for potential confounding.
25 Controlled human exposure and animal toxicological studies also demonstrated
26 NO2-induced impairments in host defense, changes in the oxidant/antioxidant balance,
27 and increases in pulmonary inflammation at concentrations of 1,500 to 5,000 ppb NO2,
28 higher than those demonstrated to increase airway responsiveness (U.S. EPA. 2008a).
29 The 2008 ISA did not explicitly link these NO2-induced biochemical and immunological
30 changes to lines of evidence for asthma exacerbation. Although there was coherence of
31 evidence across related outcomes and disciplines supporting a relationship between
32 short-term ambient NO2 exposure and respiratory effects, due to the high correlations of
33 NO2 with other traffic-related pollutants and limited analysis of potential confounding,
34 sufficient uncertainty was noted regarding the role of NO2 as an indicator for another
3 5 traffic-related pollutant or a mixture of such pollutants.
36 As will be described in the following sections, consistent with the body of evidence
37 presented in the 2008 ISA for Oxides of Nitrogen, recent studies continue to demonstrate
January 2015 5-14 DRAFT: Do Not Cite or Quote
-------
1 respiratory effects related to short-term NO2 exposure. The majority of the recent
2 evidence is from epidemiologic studies, which expand on findings for associations
3 between ambient NO2 and a broad array of respiratory effects from subclinical increases
4 in pulmonary inflammation to respiratory mortality, but particularly for effects related to
5 asthma exacerbation. Because there are few recent controlled human exposure and animal
6 toxicological studies, previous findings are a large basis of the characterization and
7 integration of evidence. Where available, results from recent studies are evaluated in the
8 context of results from previous studies. To clearly characterize differences in the weight
9 of evidence and the extent of coherence among disciplines and related outcomes, the
10 discussion of scientific information is organized by respiratory outcome group, e.g.,
11 asthma exacerbation, allergy exacerbation, respiratory infection.
5.2.2 Asthma Exacerbation
12 As detailed in the preceding section, previous studies provided several lines of evidence
13 in support of a relationship between short-term NC>2 exposure and asthma exacerbation,
14 represented as respiratory effects in populations with asthma. This evidence is
15 corroborated by recent studies. In characterizing the current state of the evidence, this
16 section begins with effects on increasing airway responsiveness and decreasing lung
17 function. These are indications of bronchoconstriction and airway obstruction, which can
18 lead to poorer control of asthma symptoms and potentially hospital admissions or ED
19 visits for asthma. The evaluation of clinical indicators of asthma exacerbation follows
20 with discussion of pulmonary inflammation and oxidative stress, which are part of the
21 mode of action for asthma exacerbation and mediate decreases in lung function and
22 increases in airway responsiveness (Figure 4-1).
5.2.2.1 Airway Responsiveness in Individuals with Asthma
Overview
23 Controlled human exposure studies evaluating the effect of inhaled NO2 on the inherent
24 responsiveness of the airways to challenge by bronchoconstricting agents have had mixed
25 results. In general, existing meta-analyses show statistically significant effects of NC>2 on
26 the airway responsiveness of individuals with asthma. However, no meta-analysis has
27 provided a comprehensive assessment of the clinical relevance of changes in airway
28 responsiveness, the potential for methodological biases in the original papers, and the
29 distribution of responses. This section provides analyses showing that a statistically
30 significant fraction (i.e., 70% of individuals with asthma exposed to NC>2 at rest)
January 2015 5-15 DRAFT: Do Not Cite or Quote
-------
1 experience increases in airway responsiveness following 30-minute exposures to NCh in
2 the range of 200 to 300 ppb and following 60-minute exposures to 100 ppb. The
3 distribution of changes in airway responsiveness is log-normally distributed with a
4 median change of 0.75 [provocative dose (PD) following NC>2 divided by PD following
5 filtered air exposure] and geometric standard deviation (GSD) of 1.88. About a quarter of
6 the exposed individuals experience a clinically relevant reduction in their provocative
7 dose due to NO2 relative to air exposure. The fraction experiencing an increase in
8 responsiveness was statistically significant and robust to exclusion of individual studies.
9 The results of the meta-analysis showed minimal change in airway responsiveness for
10 individuals exposed to NO2 during exercise. A variety of factors that may affect the
11 assessment of airway responsiveness and how those factors may directionally bias the
12 results of individual studies and the analyses in this current assessment are considered.
Background
13 Bronchial challenge agents can be classified as nonspecific [e.g., histamine, sulfur
14 dioxide (862), cold air] or specific (i.e., allergen). Nonspecific agents can be
15 differentiated between "direct" stimuli [e.g., histamine, carbachol, and methacholine]
16 which act on airway smooth muscle receptors and "indirect" stimuli (e.g., exercise, cold
17 air) which act on smooth muscle through intermediate pathways, especially via
18 inflammatory mediators (Cockcroft and Davis. 2006c). Specific allergen challenges [e.g.,
19 house dust mite, cat allergen] also act "indirectly" via inflammatory mediators to initiate
20 smooth muscle contraction and bronchoconstriction. This section focuses on changes in
21 airway responsiveness to bronchial challenge attributable to NO2 in individuals with
22 asthma. Discussed in Section 4.3.2.5. toxicological studies have demonstrated increased
23 airway responsiveness to nonspecific challenges following short-term exposure.
24 Described in Sections 5.2.2.5 and 4.3.2.6. altered responses to specific allergens
25 following NC>2 exposure have also been demonstrated in human and animal studies.
26 Responses to bronchial challenge are typically quantified in terms of the PD or
27 provocative concentration (PC) of an agent required to produce a 20% reduction in forced
28 expiratory volume in 1 second (FEVi) (PD20 or PC20, respectively) or a 100% increase in
29 specific airway resistance (sRAW) (PDioo or PCioo, respectively). There is a wide range
30 in airway responsiveness that is influenced by many factors, including medications,
31 cigarette smoke, air pollutants, respiratory infections, occupational exposures, disease
32 status, and respiratory irritants. In the general population, airway responsiveness is
33 log-normally distributed with individuals having airway hyperresponsiveness (AHR)
34 tending to be those with asthma (Postma and Boezen. 2004; Cockcroft et al.. 1983).
35 Along with symptoms, variable airway obstruction, and airway inflammation, AHR is a
36 primary feature in the clinical definition and characterization of asthma severity (Reddel
January 2015 5-16 DRAFT: Do Not Cite or Quote
-------
1 et al.. 2009). However, not all individuals with asthma experience airway
2 hyperresponsiveness. The range in airway responsiveness among individuals with asthma
3 extends into the range of healthy individuals without asthma (Cockcroft. 2010). In
4 asthma, there is a strong relationship between the degree of nonspecific airway
5 responsiveness and the intensity of the early airway response to specific allergens to
6 which individuals have become sensitized (Cockcroft and Davis. 2006a).
7 In studies investigating the effect of NC>2 exposure on airway responsiveness, individuals
8 with asthma generally have a lower PD of a bronchial challenge agent than healthy
9 individuals to produce a given reduction in lung function. In Morrow and Utell (1989a).
10 the average PD of carbachol producing a given change in lung function in individuals
11 with mild-to-moderate asthma was 16 times lower than in age-matched healthy controls.
12 Similarly, Hazuchaetal. (1983) reported a 10-12 times lower average baseline PDioo to
13 methacholine in individuals with mild asthma than healthy age-matched controls. The
14 PDs for asthma in Morrow and Utell (1989a) did not overlap with those of the healthy
15 controls, whereas Hazucha et al. (1983) observed an overlap with 2 of 15 subjects with
16 asthma being relatively unresponsive to bronchial challenge. The bronchoconstrictive
17 response to indirect acting agents (especially specific allergens) can be more difficult to
18 predict and control than the bronchoconstrictive response to non-specific agents that act
19 directly on airway smooth muscle receptors (O'Byrne et al.. 2009). Consequently, most of
20 the available literature relevant to the evaluation of the effects of NO2 on airway
21 responsiveness has focused primarily on the responses of individuals with asthma to
22 bronchial challenge with "nonspecific" bronchoconstricting agents (e.g., methacholine,
23 SO2, cold air).
24 In healthy adults without asthma or AHR, there is likely little or no clinical significance
25 of transient, small increases in airway responsiveness following low-level NC>2 inhalation
26 exposures. In individuals with asthma, however, transient changes in airway
27 responsiveness in response to inhaled pollutants may have clinical consequences.
28 Increased airway responsiveness is linked with airway inflammation and airway
29 remodeling (Chetta et al.. 1996). increased risk for exacerbation (Van Schayck et al..
30 1991). reduced lung function (Xuan et al.. 2000). and increased symptoms (Murray et al..
31 1981). A variety of environmental challenges can transiently increase AHR and worsen
32 asthma control, including allergen exposures (Strand etal.. 1997; Brusasco et al.. 1990).
33 viral infections (Cheung et al.. 1995; Fraenkel et al.. 1995). cigarette smoke (Tashkin et
34 al.. 1993). O3 (Kehrletal.. 1999). and other respiratory irritants (Kinsella et al.. 1991).
35 An exposure that worsens airway responsiveness to one agent in individuals with asthma
36 may also enhance airway responsiveness to other challenge agents. Transient increases in
37 airway responsiveness following NC>2 or other pollutant exposures have the potential to
January 2015 5-17 DRAFT: Do Not Cite or Quote
-------
1 increase symptoms and worsen asthma control, even if the pollutant exposure does not
2 cause acute decrements in lung function.
3 Three meta-analyses in the peer-reviewed literature have assessed the effects of NO2
4 exposure on airway responsiveness in individuals with asthma (Goodman et al., 2009;
5 Kjaergaard and Rasmussen. 1996; Folinsbee. 1992). Kjaergaard and Rasmussen (1996)
6 reported statistically significant effects of NO2 exposure on the airway responsiveness of
7 subjects with asthma exposed to less than or equal to 300 ppb NO2 but not for exposures
8 in excess of 300 ppb NO2. With consideration given to activity level during exposure,
9 Folinsbee (1992) found statistically significant increases in airway responsiveness of
10 subjects with asthma exposed to NO2 at rest across all concentration ranges (namely,
11 <200 ppb, 200 to 300 ppb, and >300 ppb). However, there was no statistically significant
12 effect of NO2 exposures on responsiveness during exercise. For instance, following
13 exposures between 200 and 300 ppb NO2, 76% of subjects exposed at rest had increased
14 responsiveness which was statistically significant, whereas only 52% of subjects exposed
15 while exercising had increased responsiveness, which was not a statistically significant
16 change. The analyses of Folinsbee (1992) and Kjaergaard and Rasmussen (1996) in effect
17 assessed nonspecific responsiveness because few studies of allergen responsiveness were
18 available.
19 The analyses conducted by Folinsbee (1992) were detailed in Chapter 15 of the 1993
20 AQCD for Oxides of Nitrogen (U.S. EPA. 1993). Results of these analyses appeared in
21 Table 15-10 of the 1993 AQCD and supported the conclusion that NO2 exposure
22 increases airway responsiveness in individuals with asthma. The results of a slightly
23 modified analysis focusing exclusively on non-specific responsiveness appeared in
24 Tables 3.1-3 of the 2008 ISA for Oxides of Nitrogen (U.S. EPA. 2008a). The overall
25 conclusion of that modified analysis was that NO2 exposures as low as 100 ppb (the
26 lowest concentration experimentally evaluated) conducted during rest, but not exercise,
27 increased non-specific responsiveness of individuals with asthma. Due to differences in
28 study protocols (e.g., rest vs. exercise) in the NO2-airway responsiveness literature, the
29 original (Folinsbee. 1992) and updated meta-analyses in the 2008 ISA for Oxides of
30 Nitrogen (U.S. EPA, 2008a) assessed only the fraction of individuals experiencing
31 increased or decreased airway responsiveness following NO2 exposure.
32 Goodman et al. (2009) provided meta-analyses and meta-regressions evaluating the
33 effects of NO2 exposure on airway responsiveness in subjects with asthma. By
34 considering studies of specific allergen and nonspecific responsiveness following NO2
35 exposure, Goodman et al. (2009) evaluated a larger number of studies than the analysis in
36 the 2008 ISA for Oxides of Nitrogen (U.S. EPA. 2008a). which was limited to
37 nonspecific responsiveness in subjects with asthma in an attempt to reduce the
January 2015 5-18 DRAFT: Do Not Cite or Quote
-------
1 heterogeneity among studies. Goodman et al. (2009) evaluated changes in three endpoints
2 following NO2 exposure relative to a control air exposure: (1) the fraction of subjects
3 with asthma experiencing increases in responsiveness, (2) the PD of the bronchial
4 challenge agent, and (3) the FEVi response to the challenge agent. Overall, statistically
5 significant effects of NO2 exposure on each of these three endpoints were observed.
6 Consistent with the meta-analysis provided in the 2008 ISA for Oxides of Nitrogen
7 (U.S. EPA. 2008a). Goodman et al. (2009) found 64% (95% CI: 58, 71%) of subjects
8 with asthma exposed at rest to NO2 experienced an increase in airway responsiveness,
9 whereas there was no effect of NO2 exposure during exercise with 52% (95% CI: 43,
10 60%) having an increase in responsiveness. Additionally, NO2 exposure resulted in
11 statistically significant reductions in PD as well as increases in the FEVi decrement
12 following bronchial challenge.
13 Goodman et al. (2009) concluded that, "NO2 is not associated with clinically relevant
14 effects on AHR at exposures up to 600 ppb based primarily on the small magnitude of
15 effects and the overall lack of exposure-response associations." Relative to therapeutic
16 agents used to treat airway responsiveness, which may be considered effective if they
17 more than double the PD for methacholine, the authors concluded that a -50% change in
18 the PD due to NO2 exposure would be considered adverse. Using the summary statistics
19 provided in individual studies, the effect of NO2 exposure was a -27% (95% CI: -37,
20 -18%) change in the PD. Stratifying by rest and exercise exposure, the NO2-induced
21 changes in PD were -30% (95% CI: -38, -22%) and -24% (95% CI: -40, -7%),
22 respectively. Thus, the authors concluded that the effects of NO2 exposure on airway
23 responsiveness were sufficiently small so as not to be considered adverse. The
24 appropriateness of weighing the deleterious effects of a generally unavoidable ambient
25 exposure using the criteria for judging the efficacy of beneficial therapeutic agents is not
26 clear. Based on the lack of a monotonic increase in responsiveness with exposure, the
27 authors also suggested that NO2 is not a causal factor. However, as airway responsiveness
28 data is log-normally distributed (Postma and Boezen. 2004; Cockcroft et al.. 1983). use
29 of arithmetic mean PD data may affect the validity of some analyses in the Goodman et
30 al. (2009) study. For example, in the study by Bylin etal. (1988) following exposure to
31 140 ppb NO2, there was an arithmetic mean increase of 17% in the PD relative to filtered
32 air, which was driven by a few individuals; whereas, the median and geometric mean
33 show a 24% and 16% decrease, respectively, in the PD following NO2 relative to filtered
34 air exposure.
35 None of the above described meta-analyses provided a comprehensive assessment of the
36 clinical relevance of changes in airway responsiveness, the potential for methodological
37 biases in the original papers, or the distribution of responses. This section provides such
38 analyses of airway responsiveness data and a discussion of factors that may have affected
January 2015 5-19 DRAFT: Do Not Cite or Quote
-------
1 the experimental determination of airway responsiveness as presented by Brown (2015).
2 Detailed descriptions of individual studies are provided in the 1993 AQCD for Oxides of
3 Nitrogen (U.S. EPA. 1993) and 2008 ISA for Oxides of Nitrogen (U.S. EPA. 2008a). As
4 done in the 2008 ISA for Oxides of Nitrogen (U.S. EPA. 2008a). the fraction of
5 individuals having an increase in airway responsiveness following NO2 exposure was
6 assessed. Due to considerable variability in exposure protocols and the potential for this
7 variability in protocols to affect estimates of PD (see Factors Affecting Airway
8 Hyperresponsiveness and Dose-response), the magnitude of NO2-induced changes in PD
9 was not evaluated in the original work by Folinsbee (1992) or in related EPA documents
10 (U.S. EPA. 2008a) (U.S. EPA. 1993). Herein, the magnitude of the PD change for
11 nonspecific agents is evaluated in studies that presented individual subject data for
12 persons with asthma exposed to NO2 at rest. The focus on resting exposures and
13 nonspecific challenges when assessing the magnitude of change in PD (dPD) was due to
14 the statistically significant effects of NO2 exposure on airway responsiveness for these
15 conditions as reported in the 2008 ISA for Oxides of Nitrogen [see Section 3.1.3.2
16 of (U.S. EPA. 2008a)]. In assessing the magnitude of PD change, additional
17 consideration was given to individuals experiencing a doubling-dose change in PD
18 following exposure to NO2 relative to filtered air. In a joint statement of the American
19 Thoracic Society (ATS) and European Respiratory Society, one doubling dose change in
20 PD is recognized as a potential indicator, although not a validated estimate, of clinically
21 relevant changes in airway responsiveness (Reddel et al.. 2009). Additional analyses also
22 evaluate the distribution of PD responses to NO2 and the concentration/dose-response
23 relationship.
Methods
Study and Data Selection
24 Studies included in the meta-analyses were identified from the meta-analysis by
25 Goodman et al. (2009). the 2008 ISA for Oxides of Nitrogen (U.S. EPA. 2008a). and
26 from a literature search for controlled human exposure studies of individuals with asthma
27 exposed to NO2 that were published since the 2008 ISA. For inclusion into the
28 meta-analyses, studies were required to provide data on the number of subjects whose
29 airway responsiveness increased or decreased following exposure to NO2 and filtered air.
30 Only one new controlled human exposure study (Riedl et al.. 2012) investigating the
31 effect of NO2 on airway responsiveness was published since the 2008 ISA for Oxides of
32 Nitrogen (U.S. EPA. 2008a). The location and type of airway responsiveness data
33 extracted from papers using both resting and exercising exposures is provided in
34 Supplemental Table 5-2 (U.S. EPA. 2014g).
January 2015 5-20 DRAFT: Do Not Cite or Quote
-------
1
2
o
6
4
5
6
7
8
9
10
11
As an update to Table 3.1-2 in the 2008 ISA for Oxides of Nitrogen (U.S. EPA. 2008a).
Tables 5-2 and 5-3 present studies selected for inclusion into the meta-analyses. Relative
to Table 3.1-2 in the 2008 ISA, Tables 5-2 and 5-3 include data that are either new or
previously excluded (namely, specific allergen challenges had been intentionally
excluded) for 155 subject exposures from nine studies (Riedl et al., 2012; Witten et al.,
2005: Barck et al.. 2002: Jenkins et al.. 1999: Strand et al.. 1998: Strand et al.. 1997:
Tunnicliffe et al., 1994: Morrow and Utell 1989a: Oreheketal., 1976). In general, the
subjects recruited for these studies ranged in age from 18 to 50 years with the exception
of Avol et al. (1989) who studied children aged 8-16 years. The disease status of subjects
was mild asthma in most studies, but ranged from inactive asthma up to severe asthma in
a few studies.
Table 5-2 Resting exposures to nitrogen dioxide (NO2) and airway
responsiveness in individuals with asthma.
Time Change
Chall- Post- inARa Average PD ± SEb
Reference
Ahmed et al.
(1983a)
Ahmed et al.
(1983b)
Hazucha et al.
(1983)
Oreheket al.
(1976)
Tunnicliffe et al.
(1994)
Bylinetal. (1988)
Orehek et al.
(1976)
Jorres and
Magnussen (1990)
Barck et al. (2002)
Strand et al. (1997)
N
20
20
15
20
8
20
4
14
13
18
NU2
ppb
100
100
100
100
100
140
200
250
260
260
txp.
(min)
60
60
60
60
60
30
60
30
30
30
enge
Type
CARB
RAG
METH
CARB
HDM
HIST
CARB
SO2
SIR,
TIM
SIR,
TIM
tna
Point
sGaw
sGaw
sRaw
sRaw
FEVi
sRaw
sRaw
sRaw
FEVi
sRaw
exp
min
NA
IM
20
IM
IM
25
IM
27
240
240
+
13
10
6
14
3
14
3
11
5
9
-
7
8
7
3
5
6
0
2
7
9
Air
6.0 ±2.4
9.0 ±5.7
1.9±0.4
0.56 ±0.08
-14.62
AFEVi
0.39 ±0.07
0.60 ±0.10
46.5 ±5.1
-5 ±2
AFEVi
860 ± 450
NO2
2.7 ±0.8
11.7±7.6
2.0 ± 1.0
0.36 ±0.05
-14.41
AFEVi
0.28 ±0.05
0.32 ± 0.02
37.7 ±3.5
-4 ±2
AFEVi
970 ± 450
p-valuec
NA
n.s.
n.s.
<0.01d
n.s.
n.s.
n.s.
<0.01
n.s.
n.s.
Strand et al. (1998) 16 260 30 SIR FEVi
240 11 4 -0.1 ±0.8 -2.5 ±1.0 0.03
AFEVi AFEVi
Bvlinetal. (1988)
20 270 30 HIST sRaw
25 14 6 0.39 ±0.07 0.24 ± 0.04 <0.01
January 2015
5-21
DRAFT: Do Not Cite or Quote
-------
Table 5-2 (Continued): Resting exposures to nitrogen dioxide (NO2) and airway
responsiveness in individuals with asthma
Reference
Tunnicliffe et al.
(1994)
Bvlin etal. (1985)
Mohsenin (1987a)
Bvlin etal. (1988)
N
8
8
10
20
NO2
ppb
400
480
500
530
Exp.
(min)
60
20
60
30
Chall-
enge
Type
HDM
HIST
METH
HIST
End
Point
FEVi
sRaw
pEF
sRaw
Time
Post-
exp
mm
IM
20
IM
25
Change
in ARa
+
8
5
7
12
—
0
0
2
7
Average
Air
-14.62
AFEVi
>30
9.2 ±4.7
0.39 ±0.07
PD±SEb
NO2
-18.64
AFEVi
>20
4.6 ±2.6
0.34 ±0.08
p-valuec
0.009
0.04
0.042
n.s.
AR = airway responsiveness; BIR = birch; CARB = carbachol; Exp. = exposure, FE\A = forced expiratory volume in 1 s;
HDM = house dust mite allergen; HIST = histamine; IM = immediately after exposure; METH = methacholine; NA = not available;
NO2 = nitrogen dioxide; n.s. = less than marginal statistical significance, p > 0.10; pEF = partial expiratory flow at 40% vital
capacity; RAG = ragweed; SO2 = sulfur dioxide; sGaw = specific airway conductance; sRaw = specific airway resistance;
TIM = timothy.
aChange in AR: number of individuals showing increased (+) or decreased (-) airway responsiveness after NO2 exposure
compared to air.
bPD ± SE: arithmetic or geometric mean provocative dose (PD) ± standard error (SE). See individual papers for PD calculation and
dosage units. AFENA, indicates the change in FENA, response at a constant challenge dose.
Statistical significance of increase in AR to bronchial challenge following NO2 exposure compared to filtered air as reported in the
original study unless otherwise specified. Statistical tests varied between studies, e.g., sign test, t-test, analysis of variance.
dStatistical significance for all individuals with asthma from analysis by Dawson and Schenker (1979). Oreheket al. (1976) only
tested for differences in sub-sets of individuals classified as "responders" and "non-responders."
Table 5-3 Exercising exposures to nitrogen dioxide (NO2) and
airwayresponsiveness in individuals with asthma.
Reference
Roqeretal. (1990)
Kleinman et al.
(1983)
Jenkins etal. (1999)
Jorres and
Maqnussen (1991)
Strand etal. (1996)
Avoletal. (1988)
Avoletal. (1989)
Bauer etal. (1986)
Morrow and Utell
(1989a)
Roqeretal. (1990)
n
19
31
11
11
19
37
34
15
20
19
NO2
ppb
150
200
200
250
260
300
300
300
300
300
Exp.
min
80
120
360
30
30
120
180
30
240
80
Challenge End
Type Point
METH
METH
HDM
METH
HIST
COLD
COLD
COLD
CARB
METH
sRaw
FEVi
FEVi
sRaw
sRaw
FEVi
FEVi
FEVi
FEVi
sRaw
Time
Post
-exp
min
120
IM
IM
60
30
60
60
60
30
120
Change
in ARa
+ -
10d 7d
20 7
6 5
6 5
13 5
11d 16d
12d 21d
9 3
ye 2e
8d 9d
Average
Air
3.3 ±0.7
8.6 ±2.9
2.94
0.41 ± 1.6
296 ± 76
-8.4 ± 1.8
AFEVi
-5 ±2
AFEVi
0.83 ±0.12
3.31 ± 8.64e
AFEVi
3.3 ±0.7
PD±SEb
NO2
3.1 ±0.7
3.0± 1.1
2.77
0.41 ± 1.6
229 ± 56
-10.7 ±2.0
AFEVi
-4 ±2
AFEVi
0.54 ±0.10
-6.98±3.35e
AFEVi
3.3 ±0.8
P-
value0
n.s.
<0.05
n.s.
n.s.
0.08
n.s.
n.s.
<0.05
n.s.
n.s.
January 2015
5-22
DRAFT: Do Not Cite or Quote
-------
Table 5-3 (Continued): Exercising exposures to nitrogen dioxide (NO2) and airway
responsiveness in individuals with asthma.
Reference
Rubinstein et al.
(1990)
Riedletal. (2012)
Riedletal. (2012)
Jenkins et al. (1999)
Witten et al. (2005)
Avoletal. (1988)
Roqeretal. (1990)
n
9
15
15
10
15
37
19
NO2
ppb
300
350
350
400
400
600
600
Exp
min
30
120
120
180
180
120
80
Challenge
Type
SO2
METH
CA
HDM
HDM
COLD
METH
End
Point
sRaw
FEVi
FEVi
FEVi
FEVi
FEVi
sRaw
Time
Post
-exo
min
60
90
90
IM
IM
60
120
Change
in
+
4
6
4
7
8
13e
11d
ARa
-
5
7
11
3
7
16e
8d
Average
Air
1.25 ±0.23
7.5 ±2.6
-6.9 ± 1.7
AFEVi
3.0
550 ± 240
-8.4 ± 1.8
AFEVi
3.3 ±0.7
PD±SEb
NO2
1.31 ±0.25
7.0 ±3.8
-0.5 ± 1.7
AFEVi
2.78
160 ±60
-10.4 ±2.2
AFEVi
3.7± 1.1
D-
valuec
n.s.
n.s.
<0.05f
0.018
n.s.
n.s.
n.s.
AR = airway responsiveness; CARB = carbachol; CA = cat allergen; COLD = cold-dry air; Exp. = exposure; FEV! = forced
expiratory volume in 1 s; HDM = house dust mite allergen; HIST = histamine; IM = immediately after exposure;
METH = methacholine; NO2 = nitrogen dioxide; n.s. = less than marginal statistical significance, p > 0.10; SO2 = sulfur dioxide;
sRaw = specific airway resistance.
aChange in AR: number of individuals showing increased (+) or decreased (-) airway responsiveness after NO2 exposure
compared to air.
bPD ± SE: arithmetic or geometric mean provocative dose (PD) ± standard error (SE). See individual papers for PD calculation and
dosage units. AFEV! indicates the change in FEV! response at a constant challenge dose.
°Statistical significance of increase in AR to bronchial challenge following NO2 exposure compared to filtered air as reported in the
original study. Statistical tests varied between studies, e.g., sign test, t-test, analysis of variance.
dNumber of individuals having an increase or decrease in airway responsiveness is from Folinsbee (1992).
eData for 0.25% carbachol challenge from Appendix H of Morrow and Utell (1989b).
'Significantly greater AFEV! in response to a constant challenge dose following exposure to filtered air than NO2, i.e., a protective
effect of NO2 exposure.
1
2
o
J
4
5
6
7
10
11
12
13
14
For studies that assessed airway responsiveness at multiple time points post-exposure or
over repeated days of exposure, the data from the first time point and first day of
exposure were selected for inclusion in Tables 5-2 and 5-3 to reduce the heterogeneity
among studies. Selection of the earliest time point assessing airway responsiveness was,
in part, due to late phase responses (3-8 hours post-allergen challenge) being
mechanistically different from early phase responses (<30 minutes post-allergen
challenge) (O'Byrne et al.. 2009; Cockcroft and Davis. 2006c). It should be noted that
Tables 5-2 and 5-3 are sorted by NO2 exposure concentration and, as such, studies that
evaluated multiple NO2 exposure concentrations appear in multiple rows.
Fraction of Individuals with Nitrogen Dioxide-Induced Increase in
Airway Responsiveness
Based on the summary data in Tables 5-2 and 5-3. the fraction of individuals
experiencing an NC>2-induced increase in airway responsiveness was assessed in a
manner consistent with the analysis conducted by Folinsbee (1992). Specifically, a
two-tailed sign test was used to assess the statistical significance of directional changes in
airway responsiveness between the NO2 and filter air exposure days. The nonparametric
January 2015
5-23
DRAFT: Do Not Cite or Quote
-------
1 sign test, which assumes only that the responses of each subject are independent and
2 makes no assumptions about the distribution of the response data, is appropriate to test
3 the null hypothesis that observed values have the same probability of being positive or
4 negative. This test allows estimation of whether a statistically significant fraction of
5 individuals experience an increase or decrease in airway responsiveness, but does not
6 provide information on the magnitude of the change in that endpoint. The significance of
7 a two-tailed sign test was calculated in Microsoft® Office Excel® 2013 (subsequently,
8 Excel®) as described by Currell and Dowman (2014).
Magnitude and Distribution of Nitrogen Dioxide-Induced Increase in
Airway Responsiveness
9 Individual subject airway responsiveness data for non-specific challenges following
10 resting exposures to filtered air and NO2 were available for extraction from five studies
11 (Torres and Magnussen. 1990: Bvlin et al.. 1988: Mohsenin. 1987a: Bylinetal.. 1985:
12 Oreheketal.. 1976). Data were obtained for 72 individuals and 116 NO2 exposures.
13 Twenty individuals in the Bylin etal. (1988) study were exposed to three NO2
14 concentrations and four individuals in the Oreheketal. (1976) study were exposed to two
15 NO2 concentrations. The dPD due to NO2 for each individual was assessed as:
dPD =
PDNQ2
Equation 5-1
16 where: PDNo2 and PDan are the PD following NO2 and air exposures, respectively. Given
17 that airway responsiveness is recognized as being log-normally distributed (Postma and
18 Boezen. 2004: Cockcroft et al.. 1983). this method of assessing dPD provides
19 non-negative values for log transformation and plotting.
20 The distribution of dPD data (median and GSD) was determined for each study and
21 overall for all subjects. To assess the distribution of dPD, the cumulative percentile for
22 each datum was determined using the Excel® PERCENTRANK function. The lowest and
23 highest dPD were assigned the cumulative probabilities of 0.1% and 99.9% rather than
24 the 0 and 1 assigned by Excel®. Next, the standard normal deviate (z) for each cumulative
25 percentile was determined using the Excel® NORMSINV function. The natural
26 logarithms of the dPD were subsequently regressed against their corresponding z-values
27 using the Excel® INTERCEPT and SLOPE functions. The median equals e (base of
28 natural logarithms, 2.71828) raised to the power of the intercept and the GSD equals e
29 raised to the power of the slope.
30 To assess the potential "adversity" or clinical relevance of changes in dPD, a sign test
31 was utilized to determine whether there were a statistically greater number of individuals
January 2015 5-24 DRAFT: Do Not Cite or Quote
-------
1 experiencing a doubling dose reduction in dPD (<0.5) versus those having a doubling
2 dose increase in dPD (>2). The equations used for this sign test are the same as those
3 described above for determining the fraction of individuals with an NO2-induced increase
4 in airway responsiveness. A sensitivity analysis was performed to ensure that no single
5 study or group of exposures affected the distribution of dPD and assessment of a
6 doubling dose change. The sensitivity analysis was performed by removing entire studies
7 or repeated subject exposures to multiple concentrations in two studies. Additionally,
8 broad ranges of NO2 exposure concentrations (e.g., the upper or lower half of the data)
9 were excluded for the sensitivity analyses to see if specific exposure concentrations were
10 driving results. Finally, dose-response was assessed by regressing the logarithms of dPD
11 against NO2 exposure concentration and against the product of NO2 exposure
12 concentration and duration using the Excel® Regression Data Analysis tool.
Results
Fraction of Individuals with Nitrogen Dioxide-Induced Increase in
Airway Responsiveness
13 Tables 5-2 and 5-3 provide all studies with data on the fraction of individuals
14 experiencing a change (increase or decrease) in airway responsiveness following both
15 NO2 and filtered air exposures. The statistical significance reported in the original
16 publications for changes in airway responsiveness following NO2 exposure compared to
17 filtered air is also provided in these tables. Based on all listed studies, the general
18 tendency of most studies is toward increased airway responsiveness following NO2
19 exposure with some studies reaching statistical significance. Fewer studies showed no
20 effect or a tendency for decreased airway responsiveness following NO2. Published since
21 the 2008 ISA for Oxides of Nitrogen (U.S. EPA. 2008a). one study reported a statistically
22 significant decrease in airway responsiveness following NO2, but the authors attributed
23 the protective effect of NO2 to chance (Riedl et al.. 2012).
24 Tables 5-4. 5-5. and 5-6 present the fraction of individuals experiencing an NO2-induced
25 increase in airway responsiveness to non-specific agents, specific allergens, and all
26 challenge types, respectively. Footnotes for these tables indicates the group from
27 Tables 5-2 and 5-3 that were included in the analyses. For example, in Table 5-4
28 Footnote C, the results for resting exposures (see Table 5-2) to 100 ppb NO2 are for the
29 33 individuals having an increase in nonspecific responsiveness and the 17 individuals
30 having a decrease in nonspecific responsiveness in the studies by Ahmed et al. (1983a).
31 Hazuchaetal. (1983). and Orehek et al. (1976). Table 5-4 updates Table 3.1 -3 of the
32 2008 ISA for Oxides of Nitrogen (U.S. EPA. 2008a) and is consistent with the prior
33 conclusion that statistically significant increases in nonspecific airway responsiveness
January 2015 5-25 DRAFT: Do Not Cite or Quote
-------
1 (following resting NC>2 exposures) occur in the range of 200 and 300 ppb NC>2 for
2 30-minute exposures and at 100 ppb NC>2 for 60-minute exposures in individuals with
3 asthma. Increases in airway responsiveness were not observed following the exercising
4 exposures to NCh. In general, Table 5-5 does not show statistically significant effects of
5 NC>2 exposure on airway responsiveness to allergen challenge, except at NO2
6 concentrations over 300 ppb. This may be, in part, due to the small number of individuals
7 in the analysis. Considering both specific and nonspecific challenges, Table 5-6 shows
8 statistically significant effects of NC>2 on airway responsiveness for resting but not
9 exercising exposures as was also shown for nonspecific challenges in Table 5-4.
10 However, given differing mechanisms of effect (see discussion of Bronchial Challenge
11 Agent later in this section), preference should be given to the analysis of nonspecific
12 responsiveness (Table 5-4) over the combined analysis of specific and nonspecific agents
13 (Table 5-6).
Table 5-4 Fraction of individuals with asthma having nitrogen dioxide
(NO2)-induced increase in airway responsiveness to a non-specific
challenge.
NO2 Concentration (ppb) All Exposures3'13 Exposure with Exercisea'b Exposure at Resta'b
[NO2] = 100 0.66 (50; p = 0.033) - 0.66 (50; p = 0.033)c
100<[NO2]<200 0.66 (87; p = 0.005) 0.59 (17; n.s.)d 0.67 (70; p = 0.006)e
200 < [NO2] < 300 0.59 (199; p = 0.011) 0.55 (163; n.s.)f 0.78 (36; p = 0.001)9
[NO2]>300 0.57(94; n.s.) 0.49 (61; n.s.)h 0.73 (33; p = 0.014)1
AII[NO2] 0.60 (380; p< 0.001) 0.54 (241; n.s.) 0.71 (139; p < 0.001)
n.s. = less than marginal statistical significance (p > 0.10).
aData are the fraction of subjects with asthma having an increase in airway responsiveness following NO2 vs. air exposure. Values
in parentheses are number of individuals with asthma having a change (±) in responsiveness and the p-value for a two-tailed sign
test.
bAnalysis is for the 380 subjects with asthma in Tables 5-2 and 5-3 having a change (+/-) in non-specific airway responsiveness.
°33 increases, 17 decreases; 100 ppb data from Ahmed et al. (1983a). Hazucha et al. (1983). and Oreheket al. (1976).
d10 increases, 7 decreases; 150 ppb data from Roger et al. (1990).
e47 increases, 23 decreases; 100 ppb data from Ahmed et al. (1983a). Hazucha et al. (1983). and Oreheket al. (1976): 140 ppb
data from.Bvlin etal. (1988).
'90 increases, 73 decreases; 200 ppb data from Kleinman et al. (1983): 250 ppb data from Jb'rres and Magnussen (1991): 260 ppb
data from Strand etal. (1996): 300 ppb data from Avol et al. (1988). Avol etal. (1989). Bauer et al. (1986). Morrow and Utell
(1989a). Roger etal. (1990). and Rubinstein etal. (1990).
928 increases, 8 decreases; 200 ppb data from Oreheketal. (1976): 250 ppb data from Jorres and Magnussen (1990): 270 ppb
data from Bvlinetal. (1988).
h30 increases, 31 decreases; 350 ppb data from Riedletal. (2012): 600 ppb data from Avol etal. (1988) and Roger etal. (1990).
'24 increases, 9 decreases; 480 ppb data from Bvlin et al. (1985): 500 ppb data from Mohsenin (1987a): 530 ppb data from Bvlin
etal. (1988).
January 2015 5-26 DRAFT: Do Not Cite or Quote
-------
Table 5-5 Fraction of individuals with asthma having nitrogen dioxide
(NO2)-induced increase in specific airway responsiveness to an
allergen challenge.
NO2 Concentration (ppb)
[NO2] = 100
All Exposures3'13
0.50(26; n.s.)
Exposure with
Exercisea'b
-
Exposure at Resta'b
0.50 (26; n.s.)c
100<[NO2]<200 0.50(26; n.s.) - 0.50 (26; n.s.f
200 < [NO2] < 300 0.55(56; n.s.) 0.55 (11; n.s.)d 0.56 (45; n.s.)e
[NO2]>300 0.56(48; n.s.) 0.48 (40; n.s.)f 1.00 (8; p = 0.008)9
AII[NO2] 0.55(130; n.s.) 0.49 (51; n.s.) 0.58 (79; n.s.)
n.s., less than marginal statistical significance (p > 0.10), NO2 = nitrogen dioxide.
aSee Footnote a of Table 5-4.
""Analysis is for the 130 subjects with asthma in Tables 5-2 and 5-3 having a change (+/-) in specific allergen airway
responsiveness.
°13 increases, 13 decreases; 100 ppb data from Ahmed et al. (1983b) and Tunnicliffe et al. (1994).
d6 increases, 5 decreases; 200 ppb data from Jenkins et al. (1999).
e25 increases, 20 decreases; 260 ppb data from Barck et al. (2002). Strand etal. (1997). and Strand et al. (1998).
'19 increases, 21 decreases; 350 ppb data from Riedl etal. (2012): 400 ppb data from Jenkins etal. (1999) and Witten et al.
(2005).
98 increases, 0 decreases; 400 ppb data from Tunnicliffe et al. (1994).
January 2015 5-27 DRAFT: Do Not Cite or Quote
-------
Table 5-6 Fraction of individuals with asthma having nitrogen dioxide
(NO2)-induced increase in airway responsiveness regardless of
challenge types.
NO2 Concentration (ppb)
[NO2] = 100
All Exposures3'13
0.61 (76; p = 0.08)
Exposure with
Exercise3'13
-
Exposure at Rest3'13
0.61 (76;p = 0.08)c
100<[NO2]<200 0.62 (1 13; p = 0.014) 0.59 (17; n.s.)d 0.63 (96; p = 0.018)8
200 < [NO2] < 300 0.58 (255; p = 0.008) 0.55 (174; n.s.)f 0.65 (81; p = 0.007)9
[NO2]>300 0.57(142; n.s.) 0.49 (101; n.s.)h 0.78 (41; p < 0.001)1
AII[NO2] 0.59 (510; p< 0.001) 0.53 (292; n.s.) 0.67 (218; p < 0.001)
n.s., less than marginal statistical significance (p > 0.10), NO2 = nitrogen dioxide.
aSee Footnote a of Table 5-4.
bAnalysis is for the 510 subjects with asthma in Tables 5-2 and Table 5-3 having a change (+/-) in airway responsiveness.
°46 increases, 30 decreases; 100 ppb data from Ahmed et al. (1983a). Ahmed etal. (1983b) Hazucha et al. (1983). Orehek et al.
(1976). and Tunnicliffe et al. (1994).
d10 increases, 7 decreases; 150 ppb data from Roger et al. (1990).
e60 increases, 36 decreases; 1 00 ppb data from Ahmed etal. (1983a), Hazucha etal. (1983), Orehek etal. (1976), Ahmed et al.
(1983b). and Tunnicliffe et al. (1994): 140 ppb data from.Bvlin et al. (1988).
'96 increases, 78 decreases; 200 ppb data from Kleinman et al. (1983) and Jenkins etal. (1999): 250 ppb data from Jb'rres and
Magnussen (1991): 260 ppb data from Strand etal. (1996): 300 ppb data from Avol et al. (1988). Avolet al. (1989). Bauer et al.
(1986), Morrow and Utell (1989a), Roger etal. (1990), and Rubinstein et al. (1990).
953 increases, 28 decreases; 200 ppb data from Orehek etal. (1976): 250 ppb data from Jorres and Magnussen (1990): 260 ppb
data from Barck et al. (2002). Strand et al. (1997). and Strand etal. (1998): 270 ppb data from Bvlinetal. (1988).
M9 increases, 52 decreases; 350 ppb data from Riedl etal. (2012): 400 ppb data from Jenkins etal. (1999) and Witten et al.
(2005): 600 ppb data from Avol etal. (1988) and Roger etal. (1990).
'32 increases, 9 decreases; 400 ppb data from Tunnicliffe et al. (1994): 480 ppb data from Bvlin etal. (1985): 500 ppb data from
Mohsenin (1987a): 530 ppb data from Bvlinetal. (1988).
Magnitude and Distribution of Nitrogen Dioxide-Induced Increase in
Airway Responsiveness
I The dPD for each individual was calculated as the PD following NO2 divided by the PD
2 following air exposure. Hence, a dPD greater than one suggests reduced responsiveness,
3 whereas a dPD less than one suggests increased responsiveness following NO2 exposure.
4 The dPD from the five studies providing individual PD data following resting exposures
5 to NO2 and filtered air are illustrated in Figure 5-1. All of the median responses
6 illustrated in Figure 5-1 show increased responsiveness following NC>2 exposure, i.e., an
7 NC>2-induced reduction in the PD. It should be noted in Figure 5-1 that the dPD are on a
8 log scale. The untransformed dPD data from Bvlin et al. (1988) and Mohsenin (1987a)
9 were positively skewed with a few individuals having large values of dPD. This results in
10 a large difference between the median dPD and arithmetic mean dPD. For example, at the
11 140 ppb concentration in the Bylin etal. (1988) study, the median dPD of 0.73 suggests
12 NC>2 increased responsiveness, which is consistent with 14 individuals having an increase
13 in responsiveness versus 6 having a decrease, whereas the arithmetic mean dPD of 1.15
14 suggests a reduction in responsiveness. The untransformed dPD data from Bvlin et al.
15 (1985). Jorres and Magnussen (1990). and Orehek et al. (1976) were more symmetrical
January 2015 5-28 DRAFT: Do Not Cite or Quote
-------
than Bylinetal. (1988) and Mohsenin (1987a). For the full data set in Figure 5-1,
un-transformed dPD had a skew of 3.0 (using Excel® SKEW function), whereas the
log-transformed data had a skew of only 0.2.
10.0 T
ID
1.0
0.1
00
oo
co
ID
O 00
CTl 00
:Q >•
— i co
LD Is* 00
00 00 00
>- O >•
CO ^ CO
0
200
400
600
N02 concentration (ppb)
Note: Points illustrate the responses of 72 individual subjects, and bars are median responses. Doubling dose changes are
illustrated by horizontal dotted lines. Data are from Or76 (Oreheket al.. 1976). By88 (Bvlinet al.. 1988). J690 (Jorres and
Magnussen. 1990). By85 (Bvlinet al.. 1985). and Mo87 (Mohsenin. 1987a).
Figure 5-1 Change in provocative dose (dPD) due to exposure to nitrogen
dioxide (NO2) in resting individuals with asthma.
5
6
7
8
9
10
11
12
13
A clinically relevant change in dPD is indicated by a doubling dose change, i.e., dPD >2
or <0.5. A clinically relevant, doubling dose, NC>2-induced increase in responsiveness
(dPD <0.5) was observed in 24% of the data, while 8% had a double dose decrease in
responsiveness (dPD >2). Of the 28 responses where dPD was <0.5, 17 were from Bvlin
et al. (1988). Of the nine responses where dPD was >2, eight were again from the Bylin
etal. (1988) study. Subject 1 in the Bylinetal. (1988) study had the three highest dPD
in Figure 5-1. which generally reflects the reproducibility of response. For all subjects in
the Bvlin etal. (1988) study, the Spearman's rank correlation between the 140 and
530 ppb exposures was 0.56 (p = 0.01) and was 0.48 (p = 0.03) between the 270 ppb
January 2015
5-29
DRAFT: Do Not Cite or Quote
-------
1 exposure and both the 140 and 530 ppb exposures. Clearly this study has the potential to
2 affect both the assessment of a doubling dose change in dPD as well as the distribution of
3 responses.
4 Figure 5-2 illustrates a log-probability plot of the dPD data. The data are log-normally
5 distributed with an estimated (from fitted line on plot) median of 0.75 and a GSD of 1.88.
6 The lowest and highest dPD were assigned the cumulative probabilities of 0.1% and
7 99.9%. Removing these two values did not affect the median and only slightly reduced
8 the geometric standard deviation from 1.88 to 1.87. Most of the data (namely 69%)
9 suggests an NCh-induced increase in responsiveness (dPD <1) due to NO2 exposure,
10 while 24% of the data suggests decrease responsiveness (dPD >1). Consistent with the
11 results in Table 5-4. a two-tailed sign test shows a statistically significant (p < 0.001)
12 reduction in the dPD in 74% of the 108 dPD responses not equal to one. Of the 37 dPD
13 having more than a doubling dose change, 76% show a clinically relevant NCh-induced
14 reduction in dPD (p = 0.003; two-tailed sign test).
January 2015 5-30 DRAFT: Do Not Cite or Quote
-------
10.0
5 10 20 30 50 70 80 90 95 99 99.9
Cumulative Percent Less Than Indicated dPD
Note: Data are for 72 individuals and 116 NO2 exposures illustrated in Figure 5-1. Line is log-normal fit (0.75, median dPD; 1.88,
geometric standard deviation). Table within figure is the number of observations within intervals of dPD. Doubling dose changes are
illustrated by horizontal dotted lines. The discontinuity between the 70th and 77th percentiles is due to 8 of the 116 dPD being equal
to one.
Figure 5-2 Log-normal distribution of change in provocative dose (dPD) due
to exposure to nitrogen dioxide in resting individuals with
asthma.
January 2015
5-31
DRAFT: Do Not Cite or Quote
-------
1 Table 5-7 provides a sensitivity analysis for the distribution of responses and
2 NO2-induced increases in responsiveness. The first row of the table provides the results
3 based on all dPD for all 72 individuals and 116 NCh exposures in five studies (Torres and
4 Magnussen. 1990: Bylinetal.. 1988: Mohsenin. 1987a: Bylinetal.. 1985: Orehek et al..
5 1976). Subsequent rows show results with specific studies excluded. Both Bylin et al.
6 (1988) and Orehek et al. (1976) included multiple exposure concentrations. For rows
7 examining results with exclusion of these two studies, the first row excludes the entire
8 study (all exposure concentrations) with subsequent rows excluding data for specific
9 exposure concentrations from these studies. The last row of Table 5-7 provides results
10 excluding all but the lowest exposure concentration from both Bylin et al. (1988) and
11 Orehek et al. (1976). The sensitivity analysis shows that the NCh-induced increase in
12 airway responsiveness overall and the clinically relevant, doubling dose increase in
13 responsiveness were robust to exclusion of individual studies and subparts of studies with
14 multiple exposures. Also evaluated in this sensitivity analysis, the concentration range of
15 the dataset was split into roughly halves and thirds to determine if effects were more
16 marked for a specific range of concentrations. Dividing the dataset in half, effects were
17 slightly stronger when concentrations >250 ppb were excluded than when concentrations
18 <250 ppb were excluded. When dividing the dataset in thirds, effects were least evident
19 when excluding concentrations <480 ppb and doubling dose changes were found only for
20 the lowest concentration range (i.e., > 140 ppb excluded), although those doubling dose
21 changes were only marginally significant (p = 0.057). These findings suggest more of an
22 NC>2 effect on airway responsiveness following lower concentration exposures.
23 Using the full dPD dataset of 116 exposures, linear regression did not show an
24 association between log-transformed dPD and either NC>2 concentration (p = 0.44) or
25 concentration x exposure duration (p = 0.89).
January 2015 5-32 DRAFT: Do Not Cite or Quote
-------
Table 5-7 Sensitivity analysis for distribution of responses and nitrogen
dioxide (NO2)-induced increase in responsiveness to a nonspecific
challenge.
Population
All 5 studiesd
Distribution3
0.75(1.88)
All dPD Responses13
0.74 (108; p< 0.001)
Doubling Dose dPD
Only11
0.76 (37; p = 0.003)
Studies excluded from analysis:
Bvlin etal. (1985)e
Bvlinetal. (1988)
NO2of 140, 270, and270ppbe
NO2of 140and270ppbf
NO2of 140and530ppbf
NO2of270and530ppbf
Jorres and Maqnussen (1990)e
Mohsenin (1987a)e
Oreheketal. (1976)
NO2of 100and200ppbe
NO2of 100ppbf
NO2of200ppbf
Bvlin etal. (1988)
NO2of270and530ppbf
and Oreheketal. (1976)
NO2of200ppbf
0.76(1.89)
0.70(1.64)
0.73(1.81)
0.71 (1.72)
0.73(1.78)
0.74(1.94)
0.76(1.83)
0.80(1.89)
0.79(1.89)
0.76(1.89)
0.74(1.78)
0.74 (102; p< 0.001)
0.82 (49; p< 0.001)
0.76 (68; p< 0.001)
0.78 (69; p< 0.001)
0.78 (69; p< 0.001)
0.73 (95; p< 0.001)
0.74 (98; p< 0.001)
0.72 (88; p< 0.001)
0.73 (91; p< 0.001)
0.73 (105; p< 0.001)
0.77 (66; p< 0.001)
0 74 (35- r> - 0 0061
0 Q9 l"\9' n ~ 0 0061
0.81 (21; p = 0.007)
0.85 (20; p = 0.003)
0.80 (20; p = 0.012)
0 76 H7' n - 0 0031
0 76 H4' n ~ 0 0031
0 70 HO' n ~ 0 0431
0.71 (31; p = 0.029)
0.75 (36; p = 0.004)
n 7Q MQ- n - n 01Q1
Concentrations excluded from analysis:
>140 ppb
<140 or>270 ppb
<480 ppb
>250 ppb
<250 ppb
0.71 (1.81)
0.77(1.56)
0.78(1.93)
0.73(1.71)
0.77(1.93)
0.77 (37; p = 0.003)
0.78 (36; p = 0.001)
0.69 (35; p = 0.041)
0.79 (53; p< 0.001)
0.69 (55; p = 0.006)
0.79 (14; p = 0.057)
0.78(9; n.s.)
0.71 (14; n.s.)
0.80 (15; p = 0.035)
0.73 (22; p = 0.052)
n.s. = less than marginal statistical significance (p > 0.10), NO2 = nitrogen dioxide, dPD = change in provocative dose.
daData are for 72 individuals and 116 NO2 exposures illustrated in Figures 5-1 and 5-2 and from Orehek et al. (1976). Bvlin et al.
(1988), Jorres and Maanussen (1990), Bvlin et al. (1985), and Mohsenin (1987a).
bMedian (geometric standard deviation) of dPD data.
°Data are the fraction of subjects with asthma having an increase in airway responsiveness following NO2 vs. air exposure. Values
in parentheses are number of individuals with asthma having a change (±) in non-specific airway responsiveness and the p-value
for a two-tailed sign test.
dData are the fraction of subjects with asthma having a doubling dose reduction in dPD due to NO2 exposure. Values in
parentheses are number of individuals with asthma having at least a doubling dose change (±) in non-specific airway
responsiveness and the p-value for a two-tailed sign test.
eEntire study deleted.
'Specific concentrations deleted from study with multiple exposure concentrations.
January 2015 5-33 DRAFT: Do Not Cite or Quote
-------
Discussion
1 The analyses conducted here show that the airway responsiveness of individuals with
2 asthma is increased by brief exposures to NC>2. There was a statistically significant
3 fraction of individuals with asthma exposed to NC>2 at rest, which experienced an increase
4 in responsiveness. About 70% had an increase in nonspecific airway responsiveness
5 following 30-minute exposures to NC>2 in the range of 200 to 300 ppb and following
6 60-minute exposures to 100 ppb. The median response of these individuals is an
7 NC>2-induced reduction in dPD to 0.75 (1.88, geometric standard deviation). About a
8 quarter of the exposed individuals experienced a clinically relevant, doubling dose
9 reduction in their dPD due to NC>2 exposure. The fraction experiencing a doubling dose
10 increase in responsiveness was also statistically significant and robust to exclusion of
11 individual studies. Results showed minimal change in airway responsiveness for
12 individuals exposed to NC>2 during exercise. The remainder of this discussion considers a
13 variety of factors that may affect the assessment of airway responsiveness and how those
14 factors may have directionally biased the results of individual studies and the analyses
15 conducted as part of this assessment.
Exercise
16 In considering why increases in airway responsiveness occurred only after resting
17 exposure to NCh, Folinsbee (1992) and Bylin (1993) suggested that exercise itself may
18 affect the mechanisms responsible for increased responsiveness. Based on the literature at
19 that time, both of these authors noted that exercise may cause a refractory period during
20 which airway responsiveness to challenge is diminished. Specifically, airway
21 responsiveness to methacholine had been observed to be reduced following exercise
22 (Inman et al.. 1990). A more rapid reversal of methacholine-induced bronchoconstriction
23 had also been observed following periods of exercise as compared to rest (Freedman et
24 al.. 1988). Additionally, the refractory period from exercise had been found to correlate
25 with the responsiveness to methacholine; i.e., individuals who experienced a smaller
26 bronchoconstrictive response following repeated bouts of exercise subsequently also had
27 a smaller response to methacholine challenge (Magnussen et al., 1986). Recent literature
28 continues to support the possibility that exercise may lead to a period of reduced airway
29 responsiveness. The review by O'Byrne et al. (2009) noted with repeated bouts of
30 exercise, the bronchoconstrictive response to exercise can be abolished in many
31 individuals with asthma. The most probable mechanism explaining this exercise
32 refractory period is the release of inhibitory prostaglandins that partially protect the
33 airways. There may also be changes in eicosanoids associated with NO2 exposure itself
34 (Sections 4.3.2.3 and 5.2.2.5). Refractory periods following exercise of 40 minutes to
35 3 hours has been reported (Dryden et al., 2010).
January 2015 5-34 DRAFT: Do Not Cite or Quote
-------
1 A comparison of two studies that utilized the same challenge agent following the same
2 duration of NCh exposure and nearly the same exposure concentration supports the
3 conclusion that exercise may diminish the subsequent responsiveness to bronchial
4 challenge. Torres and Magnussen (1990) found a statistically significant increase in
5 airway responsiveness to a SC>2 challenge in subjects with asthma following exposure to
6 250 ppb NO2 for 30 minutes at rest; whereas, Rubinstein et al. (1990) found no change in
7 responsiveness to a SCh challenge following exposure of subjects with asthma to 300 ppb
8 NC>2 for 30 minutes with 20 minutes of exercise.
9 Overall, the literature on airway responsiveness supports the development of a refractory
10 period following bouts of exercise. An effect of exercise refractoriness is consistent with
11 greater increases in airway responsiveness following resting than exercising exposures to
12 NO2 as shown in Table 5-4.
Bronchial Challenge Delivery and Assessment
13 Variations in methods for administering the bronchoconstricting agents may substantially
14 affect the results (Cockcroft and Davis. 2006b; Cockcroft et al.. 2005). A repeated
15 measures study of 55 subjects with asthma evaluating two ATS-recommended methods
16 of methacholine delivery found a highly statistically significant (p < 0.00001), twofold
17 difference in PC20 which was attributable to the delivery method (Cockcroft and Davis.
18 2006b). Even in the same subjects exposed by the same investigators in the same facility
19 to the same bronchial challenge agent, there can be a doubling dose difference due to the
20 delivery method. The difference observed by Cockcroft and Davis (2006b) may, in part,
21 be due to the use of full vital capacity inspirations with breath-hold as part of the delivery
22 technique that yielded the higher PC20. The maximal lung inflations are recognized to
23 induce bronchodilation.
24 The full vital capacity inspiration required for FEVi measurements when assessing
25 airway response to challenge may cause a partial reversal of bronchospasm versus the use
26 of other measures such as specific airway resistance or conductance (Jackson et al.. 2004;
27 Beaupre and Orehek. 1982; Orehek et al.. 1981). It is likely that the use of forced vital
28 capacity (FVC) maneuvers contributed to the lack of statistically significant effects in
29 NO2 studies employing exercising exposures and specific allergen challenges. For
30 nonspecific challenges (Table 5-4). responsiveness was assessed using FVC maneuvers
31 in only 6% of 139 individuals exposed at rest versus 62% of 241 individuals exposed
32 during exercise. For specific allergen challenges (Table 5-5). responsiveness was
33 assessed using FVC maneuvers in 54% of 79 individuals exposed at rest and 100% of
34 51 individuals exposed during exercise. Thus, the preferential use of FVC maneuvers in
35 studies exposing individuals to NO2 during exercise as well as in studies evaluating
36 responsiveness to specific allergens could have contributed to not finding statistically
January 2015 5-35 DRAFT: Do Not Cite or Quote
-------
1 significant effects of NO2 exposure on airway responsiveness. Where statistically
2 significant effects were observed, generally the studies using resting exposures and
3 nonspecific challenge agents, FVC maneuvers were seldom used to assess
4 responsiveness. Consistent with the results in Tables 5-4 and 5-5. the use of FVC
5 maneuvers may have biased NO2 studies using exercise and specific allergen challenges
6 to ward the null.
Bronchial Challenge Agent
7 Bronchial challenge agents differ in the mechanisms by which they cause
8 bronchoconstriction, acting either "directly" or "indirectly" on bronchial smooth muscle
9 receptors. Even similarly delivered nonspecific, direct acting agents may affect the lung
10 differently. In a comparison of responses to methacholine and histamine in healthy
11 volunteers not having AHR, Verbanck et al. (2001) reported that histamine caused an
12 overall narrowing of the airways (i.e., similar between parallel lung regions), whereas
13 methacholine caused a differential narrowing of parallel airways, which altered
14 ventilation distribution. The differential effects of these two direct acting agents may, in
15 part, be due to their differing target receptors and the distribution of these receptors in the
16 airways (O'Byrne et al.. 2009). Comparison of the airway responsiveness among
17 bronchial challenge agents is complicated by the differing mechanisms by which they
18 initiate bronchoconstriction.
19 The lack of statistical significance in Table 5-5 does not necessarily diminish the
20 potential importance of allergen exposures. First, as described above, use of FVC
21 maneuvers in NC>2 studies may have biased results toward not finding an effect on airway
22 responsiveness. Second, 80% of children with asthma are thought to be sensitized to
23 common household allergens (O'Byrne et al.. 2009). Third, individuals with asthma may
24 experience an early phase response to allergen challenge with declines in lung function
25 within 30 minutes, which primarily reflects release of histamine and other mediators by
26 airway mast cells. Approximately half of those individuals having an early phase
27 response also have a late phase response with a decline in lung function 3-8 hours after
28 the challenge, which reflects enhanced airway inflammation and mucous production
29 (O'Byrne et al.. 2009: Cockcroft and Davis. 2006c). The early response may be reversed
30 with bronchodilators, whereas the late response requires steroidal treatment. Studies have
31 reported NCh-induced effects on allergen responsiveness for both the early phase
32 (Jenkins et al.. 1999: Strand etal.. 1998: Tunnicliffe et al.. 1994) and late phase (Strand et
33 al.. 1998: Tunnicliffe et al.. 1994). These effects were observed following 30-minute
34 resting exposures to concentrations as low as 260 ppb NC>2. Finally, the response to an
35 allergen is not only a function of the concentration of inhaled allergen, but also the degree
36 of sensitization as measured by the level of allergen-specific IgE and responsiveness to
January 2015 5-36 DRAFT: Do Not Cite or Quote
-------
1 nonspecific agents (Cockcroft and Davis. 2006a). These factors make it difficult to
2 predict the level of responsiveness to an allergen, and although rare, severe
3 bronchoconstriction can occur with inhalation of very low allergen concentrations
4 (O'Byrne et al.. 2009). Given the ubiquity of allergens and potential severity of responses
5 to allergen inhalation, that NCh exposure might augment these responses is of concern.
6 The responsiveness to allergens in animals and humans is also addressed in
7 Sections 4.3.2.6 and 5.2.2.5.
Subject Inclusion/Exclusion
8 Exercise is a major trigger of asthma symptoms in between 60 and 90% of people with
9 asthma (Dryden et al.. 2010). In their study of NC>2 effects on airway responsiveness,
10 Roger etal. (1990) reported that all of their volunteers with asthma experienced either
11 cold air or exercise-induced bronchoconstriction. Morrow and Utell (1989a) reported
12 that, "Many of the asthmatic subjects were unable to undertake the carbachol challenge
13 after either NC>2 or air exposures, presumably because of pre-existing exercise-induced
14 bronchoconstriction." Consequently, in their study, data on changes in airway
15 responsiveness were only available for 9 of 20 subjects. Thus, the existence of
16 exercise-induced bronchospasm and symptoms may have caused an underlying
17 difference in the health status of subjects for which airway responsiveness was evaluated
18 between studies involving resting versus exercising exposures. Assessing those
19 individuals with less responsive airways could bias results toward not finding an effect of
20 NC>2 on airway responsiveness in studies utilizing exercising exposures.
Medication Usage
21 There was a wide range in restrictions on asthma medication usage among NC>2 studies. It
22 is recommended that short-acting bronchodilators be stopped 8 hours before and
23 long-acting bronchodilators 36 hours before the bronchial challenge (Reddel et al.. 2009).
24 Even after withholding salmeterol (a long-acting bronchodilator) for 24 hours, there is
25 still a greater than twofold reduction in airway responsiveness relative to an unmedicated
26 baseline (Reddel et al.. 2009). In their NO2 study, Hazuchaetal. (1983) required that
27 subjects not receive steroid therapy or daily bronchodilator therapy for a month prior to
28 bronchial challenge testing. Other NO2 study investigators recorded asthma medication
29 usage and asked subjects to refrain from usage for defined periods of time depending on
30 the medication, such as 8 hours for short-acting bronchodilators [e.g., (Witten etal..
31 2005; Avol etal.. 1988)1. Restrictions were far less in some studies, for example,
32 Kleinman et al. (1983) asked subjects to withhold bronchodilators for at least 4 hours
33 prior to exposure, but subjects were not excluded for non-compliance because medication
34 usage was generally balanced between filtered air and NO2 exposure days. Still other
35 studies provided no indication of asthma medications or prohibitions for study inclusion
January 2015 5-37 DRAFT: Do Not Cite or Quote
-------
1 [e-g-, (Bylin etal.. 1988)1. Pretreatment (500 mg, 4 times per day for 3 days) with
2 ascorbic acid was shown to prevent NC>2-induced increases in airway responsiveness of
3 healthy individuals (Mohsenin. 1987b). Thus, the use of asthma medications or dietary
4 supplements may have reduced the ability of studies to identify an effect of NO2 on
5 airway responsiveness and may have affected observed provocative doses.
Airway Caliber
6 Bylin (1993) suggested that NC>2 may have a direct effect on airway smooth muscle,
7 possibly relaxing and inducing mild bronchodilation at higher NC>2 doses. Consistent with
8 this supposition, Bylin etal. (1985) reported statistically significant decreases in sRaw
9 following exposure to 480 ppb NCh in healthy individuals, and a similar trend for sRaw
10 decreases in individuals with asthma. Bronchoconstriction shifts the deposition site of
11 challenge agents proximally, whereas bronchodilation shifts the deposition site more
12 distally. Decreasing the surface dose in the bronchi may in turn decrease the
13 responsiveness to the challenge.
14 The importance of particle dosimetry (which is affected by factors such as inhaled
15 particle size, airway dimensions, and breathing rates) on airway responsiveness has been
16 investigated in numerous studies. Some of the more conclusive findings are described
17 here. Moss and Oldham (2006) found that the dose of methacholine producing a 200%
18 increase in airway resistance in Balb/c mice and B6C3F1 mice was equivalent in terms of
19 the amount deposited within the first six generations of airways. Wanner etal. (1985)
20 found a strong correlation between the decrease in FEVi following histamine challenge
21 and the estimated histamine dose to the airways of 10 smokers (r = -0.82,/> < 0.005) and
22 10 nonsmokers (r = -0.83,/? < 0.005). In a study of 19 individuals with asthma, Casset et
23 al. (2007) found that the PD2o of house dust mite (RDM) allergen decreased with
24 increasing inhaled particle size from 1 um to 10 urn (mass median aerodynamic
25 diameter). As inhaled particle size was increased, the pattern of particle deposition would
26 be expected to move toward the larger more central airways. These studies demonstrate
27 lower airway responsiveness for distal versus proximal deposition of challenge agents;
28 and thus, are supportive of the supposition proposed by Bylin (1993).
29 Simply considering airway caliber may not adequately capture the complexity and
30 anatomical heterogeneity of lung disease from asthma. In a comparison of individuals
31 with asthma and healthy controls, Laube etal. (1992) reported that increasing
32 heterogeneity in particle deposition was significantly associated with decreasing PD2o to
33 methacholine. Heterogeneity in deposition is, in part, due to heterogeneity in ventilation
34 distribution. In another study of individuals with asthma, Downie et al. (2007) found
35 heterogeneity in ventilation distribution to be a predictor of airway responsiveness
36 independent of airway inflammation and airway caliber.
January 2015 5-38 DRAFT: Do Not Cite or Quote
-------
1 The literature supports an effect of the surface dose of challenge agents to the conducting
2 airways on airway responsiveness. The dose of bronchial challenge agents to the
3 conducting airways may have been affected by numerous factors within and among
4 studies evaluating the effect of NO2 on airway responsiveness. Although it is clear that
5 such factors could contribute to variability within and among studies, the available
6 information is insufficient to support an effect such as decreased airway responsiveness at
7 higher NC>2 concentrations due to bronchodilation.
Effect of Challenge Time Following Nitrogen Dioxide Exposure
8 With respect to the data in Tables 5-2 and 5-3. bronchial challenges were delivered an
9 average of 60 minutes post-exposure. For non-specific agents, on average, challenges
10 were delivered 16 minutes following resting exposures and 67 minutes following exercise
11 exposures (p < 0.01). Although challenges may take upwards of 40 minutes to complete
12 (Mohsenin. 1987a). the difference in the time when challenge agents were delivered
13 could plausibly affect differences in airway responsiveness among studies.
14 Strand et al. (1996) exposed exercising adults with asthma to 260 ppb NO2 for
15 30 minutes. Responsiveness to histamine was assessed at 30-minutes, 5-hours, 27-hours,
16 and 7-days post-exposure. The PDioo tended (p = 0.08) to decrease after 30 minutes,
17 became statistically significantly decreased by 5 hours (p = 0.03), and returned to
18 baseline by 27-hours post NC>2 exposure compared to filtered air. Although the PDioo
19 following NO2 exposure was fairly constant between 30 minutes and 5 hours, the PDioo
20 following filtered air was increased at the 5-hour time point, which may have contributed
21 to the statistically significant difference between NO2 and filtered air after 5 hours. This
22 5-hour time point is just beyond reported refractory periods following exercise of
23 40 minutes to 3 hours (Dryden et al.. 2010). A comparison across other NC>2 studies of
24 human subjects for an effect of challenge delivery timing is not possible due to
25 differences in NC>2 concentration and exposure duration. Silbaugh et al. (1981) found a
26 rapid return to baseline responsiveness in guinea pigs by 2 hours post exposure.
27 Although there is strong evidence for a refractory period following exercise and the
28 preferential use of full vital capacity maneuvers, which may relax constricted airways in
29 studies using exercise, the existing data on airway responsiveness following NC>2
30 exposure are insufficient to assess the influence of challenge delivery timing on airway
31 responsiveness in those studies.
Effect of Repeated Nitrogen Dioxide Exposures
32 To mimic a daily commute, Strand et al. (1998) exposed adults with asthma on
33 4 sequential days to either filtered air or 260 ppb NC>2 for 30 minutes during rest. The
34 early phase response to allergen challenge was statistically significantly increased by
January 2015 5-39 DRAFT: Do Not Cite or Quote
-------
1 NC>2 exposure; the 4-day mean change in FEVi was -2.5 after NO2 versus -0.4% after air
2 (p = 0.018). The late phase response to allergen challenge was also significantly greater
3 after NO2 with a 4-day avg change in FEVi after NO2 of-4.4 versus -1.9% after air
4 (p = 0.009). This study suggests that the effect of NO2 exposure on airway responsiveness
5 to allergen challenge is relatively constant over several contiguous days of repeated NO2
6 exposure. Recently, Ezratty etal. (2014) demonstrated increases in eosinophils and
7 eosinophil cationic protein (ECP) after repeated NC>2 exposures which could increase
8 airway responsiveness. Repeated ambient NO2 exposures could potentially attenuate or
9 augment responses observed in the controlled exposure studies.
Extraneous Factors
10 Although some early studies progressively increased NC>2 exposure concentrations for
11 safety purposes, the maj ority of controlled human exposure studies investigating the
12 effects of NO2 are of a randomized, controlled, crossover design in which subjects were
13 exposed, without knowledge of the exposure condition and in random order to clean
14 filtered air (the control) and, depending on the study, to one or more NCh concentrations.
15 The filtered air control exposure provides an unbiased estimate of the effects of the
16 experimental procedures on the outcome(s) of interest. Comparison of responses
17 following this filtered air exposure to those following NC>2 exposure allows for estimation
18 of the effects of NO2 itself on an outcome measurement while controlling for independent
19 effects of the experimental procedures. Furthermore, the studies by Hazucha et al. (1983)
20 and Strand etal. (1997) provided airway responsiveness data at the time of enrollment in
21 their study and airway responsiveness data following resting exposures to filtered air.
22 Little to no discernible change was observed between airway responsiveness at inclusion
23 and following the resting exposure which suggests that experimental procedures (other
24 than exposure to NCh) did not affect airway responsiveness.
Dose Response
25 Folinsbee (1992) noted that greater NC>2 doses occur with exercise due to both the
26 increased ventilation rates and a tendency for increased exposure duration. However, in
27 his meta-analyses, the effects of NO2 exposure on airway responsiveness were found
28 following resting, but not exercising, exposures to NC>2.
29 The dose-response of NC>2 on airway responsiveness may be modulated by a number of
30 factors that have been described in this section. The finding of greater airway
31 responsiveness following exposures at rest compared to exercise, despite a lower intake
32 dose of NCh during the resting exposures, is consistent with an effect of exercise
33 refractoriness. Greater airway responsiveness following exposures at rest compared to
34 exercise is also consistent with the preferential usage of full vital capacity maneuvers in
January 2015 5-40 DRAFT: Do Not Cite or Quote
-------
1 studies having exercise to assess airway responsiveness. Issues related to subject
2 selection and medication may have reduced observed effects of NO2 on airway
3 responsiveness and contributed to variability within and among studies. Both the choice
4 of bronchial challenge agent and method of delivery would have likely contributed to
5 variability among studies. Limited evidence also suggests airway dilation at higher intake
6 doses could reduce airway responsiveness. Overall, the effects of exercise refractoriness,
7 use of vital capacity maneuvers, and potential for some individuals with asthma with
8 exercise-induced bronchoconstriction to be excluded from the evaluation of airway
9 responsiveness appear to be the most likely contributors to not readily finding effects of
10 NO2 on airway responsiveness at higher intake doses occurring with exercise. Other
11 methodological differences, if randomly occurring, among studies such as the choice of
12 challenge agents, challenge delivery method, and asthma medication usage would likely
13 add variability to assessment of airway responsiveness and thereby bias data toward the
14 null of no discernible dose-response.
15 A few studies have investigated the effects of NO2 exposure on airway responsiveness at
16 more than one concentration. Intra-study evaluation of a potential dose-response reduces
17 the inherent variability and uncertainty occurring with inter-study comparisons.
18 Tunnicliffe et al. (1994) found a statistically significant and larger increase in airway
19 responsiveness at 400 ppb as compared to tendency for increased responsiveness at
20 100 ppb. Orehek et al. (1976) provided responsiveness data for four individuals following
21 exposure to both 100 and 200 ppb NO2. Of these four individuals, three had similar PDioo
22 between the two exposures, one individual had a doubling difference in the PDioo
23 (0.42 mg at 200 ppb vs. 0.94 mg at 100 ppb). Bylin etal. (1988) found statistically
24 significant effects of NO2 on airway responsiveness at 270 ppb, but not at 140 ppb or
25 530 ppb. These three studies (Tunnicliffe etal.. 1994; Bylin et al.. 1988; Orehek et al..
26 1976) for resting exposure to NO2 provide limited support for increasing airway
27 responsiveness with increasing NO2 concentration in individuals with asthma.
28 Additionally, conducted as part of this assessment, the regression of individual
29 log-transformed dPD data against dose in terms of both concentration and concentration
30 x exposure duration did not show a dose-response relationship ( Figure 5-1). The
31 dose-response evidence from studies that used exercising protocols is less compelling.
32 Roger etal. (1990) did not find a change in airway responsiveness at either 150 or
33 300 ppb NO2. Jenkins etal. (1999) found statistically significant increases in airway
34 responsiveness to allergens following a 3-hour exposure to 400 ppb NO2, but not
35 following a 6-hour exposure to 200 ppb NO2 despite equivalence in terms of the total
36 intake dose (concentration x exposure duration).
37 Several inter-study differences likely contribute to variability and uncertainty in cross
38 study comparisons of provocative dose and lung function response to bronchial challenge
January 2015 5-41 DRAFT: Do Not Cite or Quote
-------
1 agents. Evaluation of the proportional change in these outcomes following NO2 and
2 filtered air exposure as performed by Goodman et al. (2009) and herein should allow for
3 a valid comparison across studies because the air control would, theoretically, adjust for
4 many methodological differences among studies. However, even after this adjustment,
5 clear differences between resting and exercising exposures exist. Exercise itself, the
6 preferential use of full vital capacity maneuvers to assess responsiveness, and exclusion
7 of individuals with exercise-induced bronchospasm would all act to reduce the measured
8 NC>2 effect on airway responsiveness in the studies with exercise. Not using
9 log-transformed data may also affect the validity of statistical analysis requiring
10 homoscedasticity and normally distributed data. It may not be possible to adequately
11 remove the influence of some methodological factors that so substantially affect the
12 airways or the determination of airway responsiveness in individuals with asthma. Thus,
13 it is not clear to what extent inter-study assessments of the dose-response relationship
14 between NCh exposure and airway responsiveness are affected by methodological biases
15 of studies. The few studies having evaluated effects at multiple NO2 concentrations, using
16 resting exposure, are somewhat supportive of a dose-response relationship showing
17 increasing airway responsiveness with increasing NC>2 exposure concentration. However,
18 linear regression did not show an association between log-transformed dPD in Figure 5-1
19 and either NC>2 concentration (p = 0.44) or concentration x exposure duration (p = 0.89).
Summary and Conclusions
20 There is a wide range of airway responsiveness influenced by many factors, including
21 exercise, medications, cigarette smoke, air pollutants, respiratory infections, disease
22 status, and respiratory irritants. In the general population, airway responsiveness is
23 log-normally distributed with individuals having asthma generally being more responsive
24 than healthy age-matched controls. Nonspecific bronchial challenge agents causing
25 bronchoconstriction may act directly (i.e., histamine, carbachol, and methacholine) on
26 airway smooth muscle receptors or act indirectly (i.e., exercise, cold air) though
27 intermediate pathways, especially via inflammatory mediators. Specific challenge agents
28 (i.e., allergens) also act indirectly on smooth muscle to initiate bronchoconstriction.
29 Likely affecting the observed changes in airway responsiveness due to NC>2 exposure,
30 there are methodological differences among NO2 studies including subject activity level
31 (rest vs. exercise) during NCh exposure, asthma medication usage, choice of airway
32 challenge agent (e.g., direct and indirect non-specific stimuli), method of administering
33 the bronchoconstricting agents, and physiological endpoint used to assess airway
34 responsiveness. Most of these intra-study differences likely contributed to variability and
35 uncertainty in comparison among studies of provocative doses and lung function
36 responses to bronchial challenge agents. A few factors such as exercise, the use of full
January 2015 5-42 DRAFT: Do Not Cite or Quote
-------
1 vital capacity maneuvers to deliver challenge agents or measure airway responsiveness to
2 challenge, and exclusion of subjects with exercise-induced bronchospasm may have
3 preferentially biased studies to toward observing minimal NC>2 effect on airway
4 responsiveness.
5 The analyses provided here show that in individuals with asthma exposed to NO2 at rest,
6 statistically significant increases in nonspecific airway responsiveness occur in the range
7 of 200 and 300 ppb NC>2 for 30-minute exposures and at 100 ppb NC>2 for 60-minute
8 exposures. Following exposure to NC>2, relative to filtered air exposure, there was a
9 median decrease of 25% (1.88 geometric standard deviation) in the provocative dose. A
10 clinically relevant, doubling dose increase (halving of the provocative dose) due to NO2
11 occurred in a quarter of these individuals with asthma exposed to NO2 during rest. A
12 sensitivity analysis showed these findings to be robust and not driven by individual
13 studies. Consistent with the majority of studies which did not find statistically significant
14 changes in airway responsiveness when exposing individuals to NC>2 during exercise, the
15 meta-analyses of data also showed no effect. Effects of exercise refractoriness and
16 methodological aspects of these studies likely contributed to not readily finding effects of
17 NC>2 on airway responsiveness in these studies. Analyses of the available data show
18 clinically relevant and statistically significant effects of NC>2 on the airway
19 responsiveness of individuals with asthma exposed to NC>2 during rest but not exercise.
5.2.2.2 Lung Function Changes in Populations with Asthma
20 Compared with evidence for airway responsiveness, the 2008 ISA for Oxides of Nitrogen
21 reported weak evidence in controlled human exposure studies for the effects of NO2
22 exposure on changes in lung function in adults with asthma in the absence of a challenge
23 agent (U.S. EPA, 2008a). Epidemiologic evidence also was weak in adults with asthma.
24 In previous epidemiologic studies, the evidence in children with asthma was based on
25 unsupervised lung function measurements and was inconsistent. Most recent studies were
26 epidemiologic and support associations between ambient NC>2 concentrations and lung
27 function decrements in children with asthma.
Epidemiologic Studies
28 Collectively, previous and recent studies found associations between increases in ambient
29 NC>2 concentrations and decrements in supervised spirometry measures (primarily FEVi)
30 in children with asthma. Across the various populations examined, results are less
31 consistent for lung function measured under unsupervised conditions, primarily peak
32 expiratory flow (PEF) at home. Results also are inconsistent for NO and NOx. Ambient
January 2015 5-43 DRAFT: Do Not Cite or Quote
-------
concentrations of NC>2, locations, and time periods for epidemiologic studies of lung
function are presented in Table 5-8.
Table 5-8 Mean and upper percentile concentrations of nitrogen dioxide (NO2)
in epidemiologic studies of lung function in populations with asthma.
Study3
Delfino et al.
(2008a)
Smarqiassi et al.
(2014)
O'Connor et al.
(2008)
Gillespie-Bennett et
al. (2011)
Wiwatanadate and
Trakultivakorn
(2010)
Mortimer et al.
(2002)
Just et al. (2002)
Odajima et al.
(2008)
Delfino etal. (2003)
Holguinetal. (2007)
Barraza-Villarreal et
al. (2008)
Location
Riverside, CA
Whittier, CA
Montreal, Canada
Boston, MA;
Bronx, NY;
Chicago, IL;
Dallas, TX;
New York, NY;
Seattle, WA;
Tucson, AZ
Bluff, Christchurch,
Dunedin, Porirua, Hutt
Valley, New Zealand
Chiang Mai, Thailand
Bronx and East
Harlem, NY;
Chicago, IL;
Cleveland, OH;
Detroit, Ml;
St. Louis, MO;
Washington, DC
Paris, France
Fukuoka, Japan
Los Angeles, CA
(Huntington Park area)
Ciudad Juarez, Mexico
Mexico City, Mexico
Study Period
July-Dec 2003
July-Dec 2004
Oct 2009-
Apr2010
Aug 1998-
July2001
Sept 2006
Aug 2005-
June2006
June-Aug
1993
Apr-June
1996
Apr-Sept 2002
Oct 2002-
Mar2003
Nov 1999-
Jan. 2000
2001-2002
June 2003-
June2005
NO2 Metric
Analyzed
24-h avg total
personal
24-h avg
central site
24-h avg total
personal
24-h avg
4-week avg
24-h avg
4-h avg
(6 a.m.-
10 a.m.)
24-h avg
3-h avg
•(7p.m.-10
p.m.)
1-h max
8-h max
1-week avg
8-h max
Mean/Median
Concentration
(PPb)
28.6
25.0
6.3
NR
3.9
17.2
NR
28.6C
20.0
11.0
7.2
5.9
18.2
37.4
Upper
Percentile
Concentration
(PPb)
Max: 105.7
Max: 29.2
75th: 7.4
Max: 70.6
NR
NR
90th: 26.5
Max: 37.4
NR
Max: 59. Oc
Max: 51.3
Max: 49.0
90th: 9 Max:
90th: 8 Max:
NR
Max: 77.6
14
11
January 2015
5-44
DRAFT: Do Not Cite or Quote
-------
Table 5-8 (Continued): Mean and upper percentile concentrations of nitrogen
dioxide (NO2J in epidemiologic studies of lung function in
populations with asthma.
Study3
Location
Study Period
Upper
Mean/Median Percentile
NO2 Metric Concentration Concentration
Analyzed (ppb) (ppb)
Liu et al. (2009)
Dales et al. (2009)
Hernandez-Cadena
et al. (2009)
Martins et al. (2012)
Greenwald et al.
(2013)
Spira-Cohen et al.
(2011)
Yamazaki et al.
(2011)
Qian et al. (2009a)
Laqorio et al. (2006)
McCreanor et al.
(2007)
Maestrelli et al.
(2011)
Canova et al. (2010)
Hiltermann et al.
(1998)
Wiwatanadate and
Liwsrisakun (2011)
Parketal. (2005)
Windsor, ON, Canada
Mexico City, Mexico
Viseu, Portugal
El Paso, TX
Bronx, NY
Yotsukaido, Japan
Boston, MA;
New York, NY;
Madison, Wl;
Denver, CO;
Philadelphia, PA;
San Francisco, CA
Rome, Italy
London, U.K.
Padua, Italy
Padua, Italy
Bilthoven,
the Netherlands
Chiang Mai, Thailand
Incheon, South Korea
Oct-Dec 2005
May-Sept
2005
Jan and June
2006 and 2007
Mar-June
2010
Spring 2002,
Spring/Fall
2004, Spring
2005
Oct-Dec 2000
Feb 1997-Jan
1999
May-June,
Nov-Dec 1999
Nov-Mar
2003-2005
1999-2003
Summer/Fall
2004, Winter/
Summer/Fall
2005
July-Oct1995
Aug 2005-
June2006
Mar-June
2002
24-h avg
1-h max
1-week avgb
96-h avg
6-h avg
(9 a.m. -3 p.m.)
1-h avg
(6 p.m.-7 p.m.)
24-h avg
24-h avg
2-h avg
(10:30 a.m.-
12:30 p.m.)
24-h avg
24-h avg
24-h avg
24-h avg
24-h avg
19.8
57
Across 4 periods:
4.5, 3.5, 9.8, 8.2C
School A: 6.5
School B: 17.5
NR
32.6
20.8
37.6C
Oxford St: 75.5C
Hyde Park: 11. 5C
Across seasons
and years:
20.9-37.0C
27.2C
11. 2C
17.2
Control days:
31.6
Dust days: 20.7
95th: 29.5
75th: 69
Max: 116
Max across
4 periods: 4.6,
4.0, 10.9, 9.4C
NR
NR
NR
75th: 25.5
Max: 60.7
Max: 54. 3C
Max: 154C
Max: 77.7C
Range of 75th:
23.0-42.5C
48.1C
22.5C
90th: 26.5
Max: 37.4
NR
avg = average; NR = not reported, NO2 = nitrogen dioxide.
aStudies presented in order of first appearance in the text of this section.
""Subject-level exposure estimates calculated from outdoor NO2 at schools and other locations plus time-activity patterns.
°Concentrations converted from |jg/m3 to ppb using the conversion factor of 0.532 assuming standard temperature (25°C) and
pressure (1 atm).
January 2015
5-45
DRAFT: Do Not Cite or Quote
-------
Children with Asthma
1 In contrast with studies reviewed in the 2008 ISA for Oxides of Nitrogen (U.S. EPA.
2 2008a), several recent studies of children with asthma conducted spirometry under
3 supervised conditions, and most indicate a relationship with short-term NO2 exposure
4 (Figure 5-3 and Table 5-9). Studies of supervised spirometry measured lung function
5 daily, weekly, biweekly, or seasonally. Evidence for lung function measured daily by
6 subjects at home is less consistent. Among these studies, some reported an association
7 with NO2 (Gillespie-Bennett et al.. 2011; Delfino et al.. 2008a: O'Connor et al.. 2008).
8 whereas others did not (Wiwatanadate and Trakultivakorn. 2010; Odajima et al.. 2008;
9 Just et al.. 2002; Mortimer et al.. 2002). Results were inconsistent between U.S. multicity
10 studies [National Cooperative Inner-city Asthma Study (NCICAS), Inner City Asthma
11 Study (ICAS)] (O'Connor et al.. 2008; Mortimer et al.. 2002). Several studies that
12 reported no association with home lung function measurements did not provide
13 quantitative results, including NCICAS (Odajima et al.. 2008; Delfino etal.. 2003; Just et
14 al.. 2002; Mortimer et al.. 2002). Thus, the relative magnitude and precision of their
15 results cannot be assessed. However, a relationship between ambient NO2 and PEF is
16 indicated in children with asthma in a recent meta-analysis ("Weinmayr et al.. 2010) that
17 included mostly European studies as well as some studies reviewed in the 2008 ISA for
18 Oxides of Nitrogen.
January 2015 5-46 DRAFT: Do Not Cite or Quote
-------
Study
NO2 Metrics Analyzed Exposure Assessment
McCreanor et al. (2007)
Smargiassi etal. (2014)
Greenwald etal. (2013)
Holguinetal. (2007)
Martins etal. (2012)
Soira-Cohen etal.
(2011)
Liu etal. (2009)
Barraza-Villarreal et al.
2-h avg, lag 0 h
24-h avg, lag 0 day
24-h avg,
lag 0-3 day avg
24-h avg,
lag 0-6 day avg
24-h avg,
lag 0-4 day avg
6-h avg, lag 0 day
24-h avg, lag 0 day
1 -h max,
Personal outdoor ^
Personal total — •—
School •
School W
I
Central sites -0|-
Central site •
(2008)
lag 1 -4 day avg
I
Delfino et al. (2008)b 24-h avg, lag 0 day Total personal
Dales et al. (2009)b 12-h avg, lag 0 day Central sites
O'Connor et al. (2008)b 24-h avg, Central sites
lag 1-5 day avg
Maestrelli et al. (2011) 24-h avg, lag 0 day Central sites
Lagorio et al. (2009) 24-h avg, lag 0 day Central sites
-25 -20 -15 -10 -50 5 10 15 20 25 30
Percent change in FEV1 per increase in NO2 (95% Cl)a
All subjects 4
No bronchodilator 4
bronchodilator —
-25 -20 -15 -10 -50 5 10 15 20 25 30
Change in % predicted FEV1 per increase in NO2 (95% Cl)a
Note: Results are separated into two plots for the two most common indices of FEVi examined in studies. Results from more
informative studies in terms of the exposure assessment method and potential confounding considered are presented first in each
plot. Red = recent studies, black = previous studies. Study details and quantitative results are reported in Table 5-9. Table 5-9
presents results for an array of lung function indices; some of these did not have sufficient numbers to present in a figure.
aEffect estimates are standardized to a 20-ppb increase for 24-h avg NO2 and a 30-ppb increase in 1-h maxium NO2. Effect
estimates for 1-h average to 12-h average NO2 are not standardized but presented as they are reported in their respective studies
(Section 5.1.2.3).
"•Studies with home-based FENA, measurements. All other studies conducted supervised spirometry.
Figure 5-3 Associations of nitrogen dioxide (NO2) ambient concentrations or
personal exposure with percentage change in forced expiratory
volume (FEVi) (top plot) and change in percent predicted FEVi
(bottom plot) in children and adults with asthma.
January 2015
5-47
DRAFT: Do Not Cite or Quote
-------
Table 5-9 Epidemiologic studies of lung function in children and adults with asthma.
Study
Population Examined and Methodological Details
NO2 Metrics
Analyzed
Effect Estimate (95% Cl)
Lag Day Single-Pollutant Model3
Copollutant Examination
Children with asthma: studies with small spatial scale exposure assessment and/or examination of copollutant confounding
tDelfino et al. (2008a)
Riverside, Whittier, CA
n = 53, ages 9-18 yr, persistent asthma and exacerbation in
previous 12 mo
Repeated measures. Home spirometry. Examined daily for
10 days. 519 observations. Recruitment by referral from school
nurses. Parent report of physician-diagnosed asthma. Non-
smokers from nonsmoking homes. No information on
participation rate. Mixed effects model with random effect for
subject with pollutant concentrations centered on subject mean
and adjusted for personal relative humidity, personal
temperature, and follow-up period. Adjustment for city, beta
agonist use, weekend, gas stove use did not alter results.
NO2-total personal
24-h avg
Monitoring checked
daily
Central site and
personal NO2
moderately correlated.
r=0.43.
% predicted FEV-i
All subjects
0-1 avg -1.68 (-3.17,-0.19)
0 All subjects
-1.45 (-2.33,-0.57)
No bronchodilator, n = 37
-1.72 (-2.69,-0.75)
Bronchodilator use, n = 16
-0.70 (-2.90, 1.50)
NO2-central site
24-h avg
Site within 8 or 16 km
of homes.
All subjects
-1.30 (-2.44,-0.15)
with 1-h max PM2.s:
-1.27 (-2.77, 0.22)
Moderate correlation with NO2.
• Spearman r= 0.38 for
personal PM2.5, 0.36 for central
site.
• PM2snot altered by adjustment
for NO2.
EC, OC not associated with
• FEVi.
Central site NO2 with personal
PM2.5: -0.86 (-2.60, 0.89).
tSmarqiassi etal. (2014)
Montreal, Canada
n = 72, ages 7-12 yr, 29% with ED visit in previous 12 mo, 43%
using steroid medication during study
Repeated measures. Supervised spirometry. Examined daily
for 10 days. 700 observations. Residence near oil refineries &
high traffic areas. Recruitment from asthma clinic or schools.
Asthma ascertained by respirologist or parental report of
physician diagnosis. No information on participation rate. Linear
mixed effects models with random effect for subject, random
and fixed effect for study day and adjusted for age, sex, height,
month, day of week, asthma medication use, parental
education, ethnicity, personal temperature, personal humidity.
NO2-total personal
24-h avg
99% samples above
limit of detection
65% time spent
indoors
FEVi: 0.56% (-1.80, 2.92)
FVC: 0.36% (-1.44, 2.15)
FEF25-75%:
0.35% (-4.7,5.4)
No copollutant model.
No consistent associations
with personal PM2.5, benzene,
total polycyclic aromatic
hydrocarbons.
Correlations among
pollutants = 0.11 to 0.11.
January 2015
5-48
DRAFT: Do Not Cite or Quote
-------
Table 5-9 (Continued): Epidemiologic studies of lung function in children and adults with asthma.
Study
Population Examined and Methodological Details
NO2 Metrics
Analyzed
Effect Estimate (95% Cl)
Lag Day Single-Pollutant Model3
Copollutant Examination
tMartinsetal. (2012)
Viseu, Portugal
n = 51, mean age 7.3 (SD: 1.1)yr, 53% with atopy.
Repeated measures. Supervised spirometry. Four
measurements over two different seasons. Recruitment from
urban and suburban schools. -66% participation rate. Parental
report of wheeze in previous 12 mo. GEE adjusted forage, sex,
parental smoking, parental education, atopy, time of visit,
average temperature, relative humidity. Also included height,
weight, older siblings, mold or dampness in home, fireplace in
home, pets in home because changed at least one pollutant
effect estimate >10%.
NO2-modeled personal 0-4 avg
outdoor
24-h avg
Estimated from school
outdoor NO2, 20 city
locations,
MM5/CHIMERE
modeling, and daily
activity patterns.
20% time spent at
school, 65% at home.
FEVi:
-22% (-38,-1.49)
FEV-i/FVC:
-10% (-20, 0.83)
FEF25-75%:
-33% (-54, -2.57)
FEVi after bronchodilator:
19% (3.46, 37)
ForFEV-i:
with PMio: -27% (-67, 60)
with benzene:
-3.60% (-29, 31)
with ethylbenzene:
-17% (-41, 17)
Benzene unaltered by
adjustment for NO2.
Ethylbenzene & PMio
attenuated.
Correlations with NO2 negative
or weakly positive. Spearman
r=-0.72 to-0.55 for PMio,
-0.43 to 0.14 for VOCs.
tGreenwald et al. (2013)
El Paso, TX
n = 38, mean age 10 yr, 76% Mexican-American
Repeated measures. Supervised spirometry. Examined weekly
for 13 weeks. 413-441 observations. Recruitment from schools
in low and high traffic area. No information on participation rate.
School record of physician-diagnosed asthma. GLM with
subject as random effect and adjusted for school, temperature,
relative humidity, indoor NO.
NO2-school outdoor
School A: residential
area
School B: 91 m from
major road.
NO2-school indoor
All 24-h avg
FEVi:
0-3 avg School A: 25% (-15, 84)
School B:
-17% (-32, 0.12)
School A: 38% (-12, 116)
School B: -14% (-32, 7.2)
No copollutant model.
BC, SO2 (central site)
associated with FEVi.
Moderate correlation with NO2.
Pearson r= 0.62, -0.22.
. School BTEX associated with
FEVi, highly correlated with
NO2. r=0.77.
tHolguin et al. (2007)
Ciudad Juarez, Mexico
n = 95, ages 6-12 yr, 78% mild, intermittent asthma, 58% atopy
Repeated measures. Supervised spirometry. Examined
biweekly for 4 mo. 87% participation. Self-report of physician-
diagnosed asthma. Linear and non-linear mixed effects model
with random effect for subject and school adjusted for sex,
body mass index, day of week, season, maternal and paternal
education, passive smoking exposure.
NO2-school outdoor
24-h avg
Schools located
239-692 m from
homes.
0-6 avg FEVi: -2.4% (-5.09, 0.24)
No copollutant model.
No association with PM2.5, EC.
Weak to moderate correlations
with NO2. Spearman r= 0.30
for PM2.5, 0.49 for EC.
Road density at home not
school associated with lung
function.
January 2015
5-49
DRAFT: Do Not Cite or Quote
-------
Table 5-9 (Continued): Epidemiologic studies of lung function in children and adults with asthma.
Study
Population Examined and Methodological Details
NO2 Metrics
Analyzed
Lag Day
Effect Estimate (95% Cl)
Single-Pollutant Model3
Copollutant Examination
tSpira-Cohen (2013) Spira-Cohen et al. (2011)
Bronx, NY
n = 40, ages 10-12 yr, 100% non-white, 44% with asthma ED
visit or hospital admission in previous 12 mo
Repeated measures. Supervised spirometry. Examined daily
for 1 mo. 454 observations. No information on participation
rate. Recruitment from schools by referrals from school nurses.
Parental report of physician-diagnosed asthma. Mixed effects
model with subject as random effect adjusted for height, sex,
temperature. Adjustment for school (indicator of season) did not
alter results.
NO2-school outdoor
6-h avg
(9 a.m.-3 p.m.)
Schools 53-737 m
from highways with
varying traffic counts.
Most children walk to
school.
89% time spent
indoors.
0 FEVi: 0.56% (-3.93, 5.05)
PEF: 2.20% (-2.41, 6.81)
Per60-ppb increase NO2
(5th-95th percentile
change)
NO2 effect estimate adjusted
for personal EC not reported.
Personal EC associated with
lung function and not altered
by NO2 adjustment.
Personal EC-School NO2
correlation NR. School EC-
School NO2 moderately
correlated. r= 0.36.
tGillespie-Bennett et al. (2011)
Bluff, Dunedin, Christchurch, Porirua, Hutt Valley, New Zealand
n = 358, ages 6-1 Syr
Cross-sectional. Home spirometry. Multiple measures of lung
function, one NO2 measurement. Recruitment from a home
heating intervention. 77% participation. Mixed effects model
with log-transformed NO2 and random effect for subject.
Adjustment for age, sex, region, ethnicity, intervention, income,
temperature did not alter results.
NCb-home outdoor
4-week Per log increase NCb:
av9 Evening FEVi
-88 (-191, 15) mL
No copollutant model.
No other pollutants examined.
NO2-home indoor
Evening FEVi
-13 (-26, -0.38) mL
tliu et al. (2009)
Windsor, ON, Canada
n = 182, ages 9-14 yr
Repeated measures. Supervised spirometry. Examined weekly
for 4 weeks, same day of week. 672 observations. Recruitment
from schools. No information on participation rate. Parental
report of physician-diagnosed asthma. Mixed effect model with
random effect for subject and adjusted for testing period,
temperature, relative humidity, daily medication use.
NO2-central site
24-h avg
Average of 2 sites
99% subjects live
within 10 km of sites.
0 FEVi:-1.22% (-3.2, 0.84)
FEF25-75%:
-4.8% (-8.6,-0.94)
0-2 avg FEVi: -2.3% (-5.5, 0.92)
FEF25-75%:-8.0(-14, -1.6)
For lag 0-2 avg NO2 and FEVi
withPM2.5: 1.2% (-3.8, 6.4)
withSO2: -1.5% (-4.9, 2.2)
PM2 s association not altered
by NO2 adjustment, SO2
attenuated. NO2 highly
correlated with PM2.5.
Spearman r = 0.71 for PM2 5,
0.18forSO2.
tDalesetal. (2009)
Windsor, ON, Canada
n = 182, ages 9-14 yr
Repeated measures. Same cohort as above. Unsupervised
peak flow. Examined daily for 4 weeks, same day of week.
672 observations. Recruitment from schools. No information on
participation rate. Parental report of physician-diagnosed
asthma. Mixed effect model with random effect for subject and
adjusted for sex, testing period, day of week, daily mean
temperature, relative humidity, time spent outdoors.
NO2-central site
12-h avg
(8 a.m.-8 p.m.)
Average of 2 sites;
99% subjects live
within 10 km of sites.
Mean 1.6 and
2.2 h/day spent
outdoors.
Evening % predicted FEVi: Copollutant model results only
-0.10 (-0.31, 0.10)
Diurnal change FEVi:
-0.34% (-0.64, -0.04)
Per 9.8 ppb increase in
NO2 (interquartile range)
in figure.
Evening FEVi: NO2 becomes
positive with PM2.5 adjustment.
Diurnal change FEVi: NO2 not
altered by adjustment for PM2.5
or SO2. SO2 and PM2.5 not
altered by adjustment for NO2.
NO2 highly correlated with
PM2.5. Pearson r= 0.68.
January 2015
5-50
DRAFT: Do Not Cite or Quote
-------
Table 5-9 (Continued): Epidemiologic studies of lung function in children and adults with asthma.
Study
Population Examined and Methodological Details
NO2 Metrics
Analyzed
Effect Estimate (95% Cl)
Lag Day Single-Pollutant Model3
Copollutant Examination
Children with asthma: studies with central site exposure assessment and no examination of copollutant confounding
Delfino et al. (2003)
Los Angeles, CA (Huntington Park area)
n = 22, ages 10-16 yr, 100% Hispanic, 27% on anti-
inflammatory medication
Repeated measures. Home peak flow. Recruitment from
schools. 92% follow-up participation. Non-smoking children
from non-smoking homes. Self or parental report of physician-
diagnosed asthma. General linear mixed effects model with
autoregressive parameter and subject specific intercept and
adjusted for respiratory infections. Adjustment for weekend or
max temperature did not alter results.
NO2-central site 0
8-h max 1
1 site within 4.8 km of
home and school.
No quantitative data. Only
reported no statistically
significant association with
PEF.
No copollutant model.
Associations found with EC,
OC, PM-iobutnotVOCs.
Moderate to high correlations
with 8-h max NO2.
Spearman r = 0.38 (PM-io) to
0.62 (OC). For VOCs, r= 0.57
(benzene) to 0.72 (xylene).
tBarraza-Villarreal et al. (2008)
Mexico City, Mexico
n = 163-179, ages 6-14 yr, 54% persistent asthma, 89% atopy
Repeated measures. Supervised spirometry. Examined every
15 days for mean 22 weeks. 1,503 observations. No
information of participation rate. Recruitment from pediatric
clinic. Asthma severity assessed by pediatric allergist. Linear
mixed effects model with random effect for subject and
adjusted for minimum temperature, time, sex, body mass index,
ICS use. Adjustment for outdoor activity, smoking exposure,
allergy medication, season did not alter results.
NO2-central site
1-h max
Site within 5 km of
school or home.
Low correlation for
school vs. central site:
Spearman r= 0.2.1
1-4avg FEVi:0% (1.05, 1.04)
FVC:-0.11% (-1.17, 0.95) and FVC.
No copollutant model
PM2.5 associated with FEV-i
Moderate correlation with NO2.
Pearson r= 61.
tHernandez-Cadena et al. (2009)
Mexico City, Mexico
n = 85, ages 7-12 yr, 62% mild, intermittent asthma, 90% atopy
Cross-sectional. Supervised spirometry. Recruitment from
asthma and allergy clinic. Atopy and asthma severity assessed
at clinic. Linear regression adjusted for sex, pet ownership in
previous 12 mo, visible mold in home, lag 1 max temperature.
Adjustment for age and passive smoking exposure did not alter
results. Did not examine potential confounding by SES.
NO2-central site
1-h max
Site within 5 km of
home or school.
24-h avg and 8-h max
similar results but less
precise.
FEV-i response to
bronchodilator:
-39% (-64, 5.4)
No copollutant model.
Os, not PM2.5 associated with
FEVi response.
03 moderately correlated with
NO2. r=0.35.
January 2015
5-51
DRAFT: Do Not Cite or Quote
-------
Table 5-9 (Continued): Epidemiologic studies of lung function in children and adults with asthma.
Study
Population Examined and Methodological Details
NO2 Metrics
Analyzed
Lag Day
Effect Estimate (95% Cl)
Single-Pollutant Model3
Copollutant Examination
Mortimer etal. (2002) NCb-central site
Bronx and East Harlem, NY; Chicago, IL; Cleveland, OH; 4-h avg
(6 a.m.-10 a.m.)
Average of all city
Detroit, Ml; St. Louis, MO; Washington, DC (NCICAS)
n = 846, ages 4-9 yr
Repeated measures. Home peak flow. Examined daily for four monitors.
2-week periods. Recruitment from ED visits and clinics. Parent
report of physician-diagnosed asthma and symptoms in
previous 12 mo or asthma symptoms for >6 weeks, or family
history of asthma. Participation from 55% full cohort. Sample
representative of full cohort except for greater asthma
medication use. Mixed effects model adjusted for city, follow-up
period, day of study, 24-h rainfall, 12-h avg temperature.
Single-
day lags
1 to 6
1 -5 avg
1 -4 avg
0-4 avg
0-3 avg
No quantitative data. Only
reported no association
with PEF.
No copollutant model.
Os associated with PEF. Weak
correlation with NO2. r= 0.27.
tO'Connor et al. (2008)
Boston, MA; Bronx, NY; Chicago, IL; Dallas, TX; New York, NY;
Seattle, WA; Tucson, AZ (ICAS)
n = 861, ages 5-12 yr, persistent asthma and atopy, 82% black
or Hispanic
Repeated measures. Home spirometry. Examined for 2 weeks
every 6 mo for 2 yr. 70% of maximum data obtained. Recruited
from intervention study. Mixed effects model adjusted for site,
month, site*month interaction, temperature, intervention group.
NO2-central site
24-h avg
All monitors close to
home and not near
industry. Median
distance to
site = 2.3 km.
1-5 avg % predicted FEV-i:
-1.33 (-1.88, -0.78)
% predicted PEF:
-1.63 (-2.20, -1.06)
Only three-pollutant model
analyzed with PM2.sand Os.
Associations also found with
PM2.5, CO, SO2, and O3.
Moderate correlations with
NO2. r=0.59forPM25, 0.54
for CO, 0.59 for SO2. Weak
correlation with Os. r= -0.31.
tYamazaki etal. (2011)
Yotsukaido, Japan
n = 17, ages 8-15 yr, 100% atopy
Repeated measures. Supervised peak flow before medication
use. Examined daily during long-term hospital stay. No air
conditioning in hospital. Permitted to go outside if asthma
stable. Poor generalizability. 1,198 observations. GEE adjusted
for sex, age, height, temperature, day of week, temporal trends.
NO2-central site
1-h avg
(6 p.m.-7 p.m.)
Monitor adjacent to
hospital. No major
roads nearby.
No quantitative data. PEF
decreases with increasing
NO2 0 to 23 hours before
measurement.
Strongest associations at
0 hand 12 h.
Only three-pollutant model
analyzed with PM2.5 and Os.
PM2.5, not Os, also associated
with evening PEF.
PM2.5 moderately correlated
with NO2. r= 0.62.
Just etal. (2002)
Paris, France
n = 82, ages 7-15 yr, asthma attack in previous 12 mo and
daily asthma medication use, 90% atopy
Repeated measures. Home peak flow. Examined daily for
3 mo. 82% participation. Recruitment from hospital outpatients.
GEE adjusted for time trend, day of week, pollen, temperature,
humidity.
NO2-central site
24-h avg
Average of 14 sites
NR No quantitative data. Only
reported no relationship
with PEF.
No copollutant model.
03 associated with PEF. No
correlation with NO2. Pearson
r=0.09.
January 2015
5-52
DRAFT: Do Not Cite or Quote
-------
Table 5-9 (Continued): Epidemiologic studies of lung function in children and adults with asthma.
Study
Population Examined and Methodological Details
tOdajimaetal. (2008)
Fukuoka, Japan
n = 70, ages 4-1 1 yr, 66% with asthma exacerbation
Repeated measures. Home peak flow. Examined daily for 1 yr.
>15,000 observations. Participation rate not reported.
Recruitment from hospital where received treatment. GEE
adjusted forage, sex, height, growth of child, temperature.
NO2 Metrics
Analyzed
NO2-central site
3-h avg
24-h avg
1 site
Effect Estimate (95% Cl)
Lag Day Single-Pollutant Model3
No quantitative data. Only
reported no association
with PEF.
Copollutant Examination
Only three-pollutant model
analyzed with suspended PM
and 03.
Suspended PM associated
with PEF in warm season.
Weak correlation with NO2.
r= 0.30 for 24-h avg.
tWiwatanadate and Trakultivakorn (2010)b
Chang Mai, Thailand
n = 31, ages 4-11 yr, asthma plus symptoms in previous
12 mo, 52% mild intermittent
Repeated measures. Home peak flow. Examined daily for 1 yr.
97% participation. Recruitment from allergy clinic. GLM with
random effect for subject and adjusted for time trend, day of
week, height, weight, atmospheric pressure, temperature,
sunshine duration.
NO2-central site
24-h avg 0
1 site within 25 km of 1
homes.
PEF
-1.80 (-5.40, 1.80)L/min
2.60 (-1.20, 6.40)L/min
No copollutant model.
No consistent (across various
lags of exposure) associations
found for PM2.5, CO, PM-io,
SO2, or Os.
Adults with Asthma: studies with small spatial scale exposure assessment and/or examination of copollutant confounding
McCreanoretal. (2007)
London, U.K.
n = 31 mild asthma, 32 moderate asthma, ages 19-55 yr, all
with airway hyperresponsiveness, 84% with atopy
Randomized cross-over natural experiment. Supervised
spirometry. Exposure on busy road and park. 55 observations.
Participation rate not reported. Recruitment from
advertisements and volunteer databases. Mixed effects model
with random effect for subject and adjusted for temperature,
relative humidity.
NO2-personal outdoor 0-h
2-h avg
Measured next to
subjects during
outdoor exposures.
FEVi:
-0.22% (-0.40, -0.05)
FEF25-75%:
-0.78% (-1.4, -0.13)
22-h FEVi:
Post- -0.13% (-0.35, 0.10)
exposure FEF25-75%:
-0.75% (-1.6, 0.10)
per 5.3 ppb increase in
NO2
For FEF25-75%
with UFP: -0.47% (-1.3, 0.39)
with EC: -0.43% (-1.1, 0.26)
withPM25: -0.48% (-1.3,
•0.25)
Moderate correlations with
NO2. Spearman r= 0.58 for
UFP and EC, 0.60 for PM2.5.
tQian et al. (2009b)
Boston, MA; New York, NY; Denver, CO; Philadelphia, PA; San
Francisco, CA; Madison, Wl
n = 119, ages 12-65 yr, persistent asthma, nonsmokers
NO2-central site
24-h avg
Average of all
monitors within 32 km
of subject ZIP code
centroid.
PEF
All subjects
-0.68% (-1.30, -0.06)
Placebo
-0.29% (-.35, 0.80)
Beta-agonist
-1.08% (-2.18, -0.05)
ICS
with SO2: -0.11% (-0.87, 0.64)
withPMm -0.80% (-1.7, 0.10)
withOs: -0.68% (-1.3, -0.05)
SO2 slightly attenuated with
NO2 adjustment.
PM-io, 03 not associated with
PEF.
Correlations NR.
January 2015
5-53
DRAFT: Do Not Cite or Quote
-------
Table 5-9 (Continued): Epidemiologic studies of lung function in children and adults with asthma.
Study
Population Examined and Methodological Details
NO2 Metrics
Analyzed
Effect Estimate (95% Cl)
Lag Day Single-Pollutant Model3
Copollutant Examination
Repeated measures. Home PEF. No information on
participation rate. Study population representative of full cohort.
Examined daily for 16 weeks. >14,000 observations. Trial of
asthma medication, a priori comparison of medication
regimens. Linear mixed effects model adjusted forage, sex,
race/ethnicity, center, season, week, daily average
temperature, daily average humidity. Adjustment for viral
infections did not alter results.
-0.61% (-1.67, 0.39)
0-2 avg All subjects
-0.58% (-1.43, 0.27)
Adults with Asthma: studies with central site exposure assessment and no examination of copollutant confounding
tMaestrellietal. (2011)
Padua, Italy
n = 32, mean age 39.6 (SD: 7.5) yr, 81% persistent asthma
Repeated measures. Supervised spirometry. 6 measures over
2 yr. 166 observations. Selected from database of beta-agonist
users (>6/yr for 3 yr), diagnosis clinically confirmed. 76%
follow-up participation. Drop outs did not differ from
participants. GEE adjusted for daily average temperature,
humidity, atmospheric pressure, asthma medication use,
current smoking status.
NO2-central site
24-h avg
Average of 2 city sites
0
% predicted FEV-i:
1.07 (-6.57, 8.71)
No copollutant model.
CO associated with FEV-i. No
association with personal or
central site PlVh.s.
No association for central site
PM-io, SO2, Os.
Correlations with NO2 NR.
Lagorioetal. (2006)
Rome, Italy
n = 11, ages 18-64 yr, 100% mild, intermittent asthma
Repeated measures. Supervised spirometry. Examined 2/week
for two 1-mo periods. Mean 9, 15 observations/subject.
Participation rate not reported. Recruitment of non-smokers
from outpatient clinic. GEE adjusted for season, temperature,
humidity, beta-agonist use.
NO2-central site
24-h avg
Average of 5 city sites
% predicted FEVi:
0 -4.14 (-6.71,-1.56)
0-1 avg -4.81 (-7.54, -2.09)
No copollutant model.
CO at lag 0-2 avg associated
with FEVi. No association for
PIvh.s, PM-io, PM-io-2.5, Os.
Low to moderate correlations
with NO2. Spearman r= 0.05
for CO, 0.17 for O3,0.43-0.51
for PM.
tCanovaetal. (2010)
Padua, Italy
n = 19, ages 15-44 yr, 81% moderate/severe asthma
Repeated measures. Home PEF/FEV-i. Examined for five 30-
day periods for 2 yr. Recruitment from prescription database of
subjects with mean >6 prescription/yr for 3 yr. 50% subjects
provided fewer than 33% maximum observations. GEE
adjusted for temperature, humidity, atmospheric pressure, ICS
use, smoking status.
NO2-central site
24-h avg
2 city sites
0, 1, 2, 3 Results reported only in a
(single- figure
daV) NO2 shows null
0-1 avg associations with PEF and
0-3 avg FEVi.
No copollutant model.
CO associated with PEF.
Moderate correlation with NO2.
Spearman r= 0.48.
January 2015
5-54
DRAFT: Do Not Cite or Quote
-------
Table 5-9 (Continued): Epidemiologic studies of lung function in children and adults with asthma.
Study
Population Examined and Methodological Details
NO2 Metrics
Analyzed
Lag Day
Effect Estimate (95% Cl)
Single-Pollutant Model3
Copollutant Examination
Parketal. (2005) NCb-central site
Incheon, Korea 24-h avg
n = 64 with asthma, ages 16-75 yr, 31% with severe asthma 10 city sites
Repeated measures. Home PEF. Examined daily for 3-4 mo.
Recruited from medical center. GEE model, covariates NR.
0 PEF
0.45 (-1.01, 1.90)L/min
No copollutant model.
CO, PM-io, Os associated with
PEF.
No association for SO2.
tWiwatanadate and Liwsrisakun (2011)
Chiang Mai, Thailand
n = 121 with asthma and symptoms in previous 12 mo, ages
13-78 yr, 48% moderate/severe persistent asthma
Repeated measures. Home PEF. Examined daily for 10 mo.
Recruited from allergy clinic patients. GLM with random effect
for subject and adjusted for sex, age, asthma severity, day of
week, weight, pressure, temperature, sunshine duration, rain.
NO2-central site
24-h avg
1 city site
Evening PEF:
1.00(0.0,2.00)
Average PEF:
1.6(0.60,2.60)
Units of PEF not reported.
Only multipollutant models
analyzed. No associations with
PM-io, SO2, 03.
Interactions between NO2 and
copollutants or meteorological
variables reported not to be
statistically significant.
Hiltermann et al. (1998)
Bilthoven, the Netherlands
n = 60 with asthma, ages 18-55 yr, all with airway
hyperresponsiveness, 87% with atopy
Repeated measures. Home PEF. Examined daily for 4 mo.
Recruitment from outpatient clinic. Model adjusted for allergen
concentrations, smoking exposure, day of week, temperature,
linear and quadratic term for study day
NO2-central site
24-h avg
1 city site
Site within 20 km of
subjects' homes. 3 city
sites highly correlated.
r=0.88.
Diurnal change PEF
0 -0.75 (-8.12, 6.62)L/min
0-6 avg -3.01 (-16, 10) L/min
No copollutant model.
BS at lag 0 associated with
PEF. No association with PM-io
orOs.
Note: More informative studies in terms of the exposure assessment method and potential confounding considered are presented first.
avg = average; BTEX = benzene, toluene, ethylbenzene, xylene; GEE = generalized estimating equations; GLM = generalized linear model; ICAS = Inner City Asthma Study;
ICS = inhaled corticosteroid; NCICAS = National Cooperative Inner-city Asthma Study; NR = not reported; SES = socioeconomic status, BS = black smoke, Cl = confidence interval,
CO = carbon monoxide, EC = elemental carbon, ED = emergency department, FEF2^75% = forced expiratory flow from 25% to 75% of vital capacity, FENA, = forced expiratory volume
in 1 second, FVC = forced vital capacity, NO2 = nitrogen dioxide, O3 = ozone, OC = organic carbon, PEF = peak expiratory flow, PM = particulate matter, SD = standard deviation,
SO2 = sulfur dioxide, UFP = ultrafine particles, VOC = volatile organic compound.
aEffect estimates were standardized to a 20-ppb increase in 24-h avg NO2, a 25-ppb increase in 8-h max NO2, and a 30-ppb increase 1-h max NO2. Effect estimates for other
averaging times (1-h avg to 12-h avg) are not standardized but presented as they are reported in their respective studies (Section 5.1.2.3).
fStudies published since the 2008 ISA for Oxides of Nitrogen.
January 2015
5-55
DRAFT: Do Not Cite or Quote
-------
1 With respect to the populations examined, most studies assessed asthma as parental
2 report of physician-diagnosed asthma. Children were recruited mostly from schools,
3 supporting the likelihood that study populations were representative of the general
4 population of children with asthma. Study populations represented a range of asthma
5 severity, as ascertained by Global Initiative for Asthma guidelines or medication use, ED
6 visit, or hospital admission for asthma in the previous year. Based on a priori hypotheses,
7 results did not demonstrate larger NC^-associated decrements in lung function in children
8 with asthma than children without asthma (Barraza-Villarreal et al.. 2008; Holguin et al..
9 2007). Post-hoc analyses pointed to stronger associations among children with asthma not
10 taking inhaled corticosteroid (ICS) (Hernandez-Cadena et al.. 2009; Liu et al.. 2009) or
11 not taking controller bronchodilators (Delfino et al., 2008a). The limited results for larger
12 associations in ICS non-users together with observations for NCh-associated lung
13 function decrements in populations with high prevalence of atopy (53%, 58%) (Martins
14 et al.. 2012; Holguin et al.. 2007) are supported by findings for NCh-induced increases in
15 allergic inflammation (Section 5.2.2.5) and findings for mast cell degranulation (which
16 leads to histamine release) in mediating NCh-induced lung function decrements
17 (Section 4.3.2.2). Bronchodilator use has been shown to reduce airway responsiveness in
18 response to a challenge agent (Section 5.2.2.1).
19 For children with asthma, key evidence for NCh-associated lung function decrements was
20 provided by studies with exposure assessment in subjects' locations: total personal NO2
21 (Smargiassi et al., 2014; Delfino et al.. 2008a). personal outdoor NC>2 estimated from
22 measurements at school and other locations and time-activity data (Martins et al.. 2012)
23 or outdoor school NC>2 (Greenwald et al., 2013; Spira-Cohen et al., 2011; Holguin et al..
24 2007). Among these studies, few reported participation rates (66%, 87%; Table 5-9);
25 however, none reported issues with selective participation according to a specific subject
26 characteristic or NC>2 exposure. These studies examined limited lags of NO2 exposure but
27 were similar in finding associations with multiday (i.e., lag 0-1 avg, 0-4 avg) averages
28 of 24-h avg NCh. Studies that measured or estimated personal exposures provide
29 evidence of an effect of outdoor NO2 exposure on decreasing lung function. Among
30 children with wheeze in Portugal, indoor school and home NC>2 concentrations were
31 below the limit of detection (Martins et al., 2012). and time-weighted average of
32 microenvironmental NC>2 concentrations have shown agreement with personal NC>2
33 (Section 3.4.4.1).
34 Studies of total personal NC>2 produced contrasting results. Among children with asthma
35 in the Los Angeles, CA area, slightly larger decrements in percent predicted FEVi were
36 found with total personal NO2 (-1.5 [95% CI: -2.3, -0.57] per 20-ppb increase in lag 0
37 day NO2) than central site NO2 (-1.3 [95% CI: -2.4, -0.15]) (Delfino etal.. 2008a). A
38 Spearman correlation of 0.43 between personal and central site NC>2 indicated that
January 2015 5-56 DRAFT: Do Not Cite or Quote
-------
1 ambient NO2 had some influence on personal exposures. In contrast, total personal NO2
2 exposure was not associated with lung function among children with asthma living near
3 NO2 emissions sources (i.e., oil refineries, high traffic roads) in Montreal, Canada
4 (Smargiassi et al.. 2014). Both studies of total personal exposure examined children for
5 10 consecutive days, and the total number of observations was higher in Smargiassi et al.
6 (2014) than in Delfino (2006) (-700 vs. -500). However, it is uncertain whether the
7 Montreal study was sufficiently powered to detect associations with NO2 exposures,
8 which were far lower than in the Los Angeles study (mean: 6.3 ppb vs. 28.6 ppb) and
9 showed low variability among children and days (IQR: 2.9 ppb vs. 16.8 ppb). The
10 Montreal study did not provide strong evidence that lung function was associated with
11 total personal exposures to PM2 5, VOCs, or polycyclic aromatic hydrocarbons either
12 (Smargiassi et al.. 2014).
13 Among studies of outdoor school NCh, associations with FEVi were found in populations
14 in El Paso, TX, and Ciudad Juarez, Mexico, which are located along the U.S./Mexico
15 border (Greenwald etal.. 2013: Holguin et al.. 2007) (Figure 5-3 and Table 5-9).
16 Between two El Paso schools, associations were limited to the school located near a
17 major road and characterized by higher outdoor pollutant concentrations and a larger
18 percentage of Mexican-American children (Greenwald et al.. 2013). No association with
19 FEVi was found in children with asthma in Bronx, NY for school NC>2 averaged over the
20 6-h school day (Spira-Cohen et al.. 2011). An effect of outdoor NC>2 is supported by
21 similar FEVi decrements for outdoor and indoor NC>2 in an El Paso school (Greenwald et
22 al.. 2013) and larger lung function decrements for home outdoor than indoor NC>2 among
23 children in five New Zealand towns (Gillespie-Bennett et al.. 2011). The latter results
24 have weaker implications as multiple daily FEVi measures were related to a single
25 4-week average of NC>2.
26 Compared with NC>2 exposures estimated for subjects' locations, evidence for
27 associations with lung function decrements is more uncertain for NC>2 measured at central
28 sites. Among studies that measured ambient NC>2 at central sites, some found associations
29 with lung function decrements (Yamazaki et al.. 2011: Dales et al.. 2009: Hernandez-
30 Cadena et al.. 2009; Liu et al.. 2009: O'Connor et al.. 2008). Many studies reported lack
31 of association (Wiwatanadate and Trakultivakorn. 2010: Barraza-Villarreal et al.. 2008:
32 Odajima et al.. 2008: Just et al.. 2002: Mortimer et al.. 2002). Among children in Mexico
33 City, Mexico, and Thailand, various lung function parameters showed no or imprecise
34 associations with NCh (Wiwatanadate and Trakultivakorn. 2010: Barraza-Villarreal et al..
35 2008) (Table 5-9). Other studies did not report quantitative results, and it was not
36 possible to assess the extent to which their findings did or did not suggest associations.
37 Across the studies examining central site NO2, exposures were assigned as ambient
38 measurements from a site located within 5 or 10 km of subjects' homes or schools,
January 2015 5-57 DRAFT: Do Not Cite or Quote
-------
1 measurements averaged among city monitors, or measurements from one site. The central
2 site NO2 assessment method did not appear to influence results; however, in Mexico City,
3 a low correlation (r = 0.21) between central site and school NO2 suggests that the central
4 site may not have adequately represented the variability within the study area (Barraza-
5 Villarreal et al.. 2008).
6 Compared with NO2 exposures estimated for subjects' locations, evidence for
7 associations with lung function decrements is more uncertain for NO2 measured at central
8 sites. Among studies that measured ambient NO2 at central sites, some found associations
9 with lung function decrements (Yamazaki et al.. 2011; Dales et al.. 2009; Hernandez-
10 Cadena et al.. 2009; Liu et al.. 2009; O'Connor et al.. 2008). Many studies reported lack
11 of association (Wiwatanadate and Trakultivakorn. 2010; Barraza-Villarreal et al.. 2008;
12 Odajima et al.. 2008; Just et al.. 2002; Mortimer et al.. 2002). Among children in Mexico
13 City, Mexico, and Thailand, various lung function parameters showed no or imprecise
14 associations with NO2 (Wiwatanadate and Trakultivakorn. 2010; Barraza-Villarreal et al..
15 2008) (Table 5-9). Other studies did not report quantitative results, and it was not
16 possible to assess the extent to which their findings did or did not suggest associations.
17 Across the studies examining central site NO2, exposures were assigned as ambient
18 measurements from a site located within 5 or 10 km of subjects' homes or schools,
19 measurements averaged among city monitors, or measurements from one site. The central
20 site NO2 assessment method did not appear to influence results; however, in Mexico City,
21 a low correlation (r = 0.21) between central site and school NO2 suggests that the central
22 site may not have adequately represented the variability within the study area (Barraza-
23 Villarreal etal.. 2008).
24 Adjustment for potential confounding varied among studies but in most cases included
25 temperature. Several studies adjusted for (or considered in preliminary analyses) relative
26 humidity; a few studies adjusted for day of the week, smoking exposure, or asthma
27 medication use (Table 5-9). Few studies analyzed copollutant models, and while Holguin
28 et al. (2007) found that neither PM2 5 nor EC was associated with FEVi among children
29 with asthma in Ciudad Juarez, Mexico, most studies found associations with the
30 traffic-related pollutants PM2 5, BC/EC, or VOCs as well as with PMio, SO2, or O3. A
31 wide range of correlations with NO2 were reported for PM25 (r = 0.30-0.71). Negative or
32 weakly positive correlations were reported for other pollutants (e.g., -0.72 for PMio to
33 0.18 for 802). In copollutant models, NO2 effect estimates were attenuated in some cases
34 and unchanged in others. Copollutant effect estimates adjusted for NO2 generally were
35 not altered. Among children with wheeze in Portugal, the association of modeled outdoor
36 NO2 with FEVi was attenuated (-3.7% [95% CI: -33, 25%] per 20-ppb increase in
37 1-week avg NO2) with adjustment for benzene (Spearman r = -0.42 to 0.14). Among
38 children with asthma in Windsor, Ontario, Canada, associations of 12-h avg and 24-h avg
January 2015 5-58 DRAFT: Do Not Cite or Quote
-------
1 NO2 with FEVi became positive with adjustment for highly correlated (r = 0.71) PM2 5
2 (Dales et al.. 2009: Liu et al.. 2009) (Table 5-9). NO2 associations with FEVi diurnal
3 change were unchanged by PM2 5 or SO2 adjustment (Dales et al.. 2009). In a more
4 detailed copollutant analysis of personal and central site measures, Delfino et al. (2008a)
5 found the association of personal NO2 with FEVi to be robust (-1.3-point [95% CI: -2.8,
6 0.22] change in percent predicted FEVi per 20-ppb increase in NO2) to adjustment for
7 personal PM2 5, which was weakly correlated with personal NO2 (Spearman r = 0.38).
8 Adjustment for personal PM2 5 (r = 0.32) reduced the association of central site NO2 with
9 FEVi (-0.86-point [95% CI: -2.6, 0.89] change per 20-ppb increase in NO2). The
10 attenuation could indicate that ambient NO2 serves as an indicator of personal PM2 5 but
11 could also result from less exposure measurement error for personal PM2 5 than central
12 site NO2. Nonetheless, the moderate personal-ambient NO2 correlation (r = 0.43) and the
13 copollutant model results for personal NO2 provide evidence for independent effects on
14 FEVi of ambient NO2.
Adults with Asthma
15 Most previous and recent studies of lung function in adults with asthma were based on
16 PEF measured at home and indicated no association or inconsistent associations among
17 the various lung function parameters or NO2 exposure lags examined (Wiwatanadate and
18 Liwsrisakun. 2011: Canovaetal.. 2010: Park etal.. 2005: Hiltermann et al.. 1998). Most
19 of these studies recruited subjects from outpatient clinics or doctors' offices, and the
20 nonrandom selection of the general population may produce study populations less
21 representative of the asthma population. Ambient NO2-associated decreases in PEF were
22 found in a recent multicity U.S. study of adults with asthma (Qian et al.. 2009b). The few
23 studies with supervised spirometry also produced inconsistent evidence overall
24 (Maestrelli etal.. 2011: McCreanor et al.. 2007: Lagorio et al.. 2006) (Table 5-9). Most
25 studies examined 24-h NO2 that was assessed primarily from central site measurements,
26 and results were equally inconsistent for NO2 exposures assigned from one site or
27 averaged from multiple city sites.
28 The strongest study with personal outdoor NO2 measurements and examination of
29 traffic-related copollutants provides evidence for an independent association for NO2.
30 Among adults in London, U.K. with mild to moderate asthma, NO2 measured next to
31 subjects while walking next to a high-traffic road (allowing only diesel buses and taxis)
32 and while in a park was associated with decrements in FEVi and forced expiratory flow
33 from 25% to 75% of vital capacity (FEF25-75%) (McCreanor etal.. 2007). NO2-related
34 decrements occurred 2 to 22 hours after exposure. A 5.3-ppb increase in 2-h avg NO2 was
35 associated with a -0.22% (95% CI: -0.40, -0.05) change in FEVi and -0.78% (95% CI:
36 -1.4, -0.13) change in FEF25-?5%. In the London walking study and other studies, lung
January 2015 5-59 DRAFT: Do Not Cite or Quote
-------
1 function also was associated with the traffic-related pollutants EC, UFP, PM2 5, or CO,
2 which were moderately correlated with NC>2 (Spearman r = 0.43-0.60) (McCreanor et al..
3 2007; Lagorio et al., 2006). Associations also were observed with PMio and 862. In the
4 U.S. multicity study of adults, NCh-PEF effect estimates were attenuated with adjustment
5 for SC>2 (Qian et al., 2009b). Copollutant effect estimates were unaltered or less
6 attenuated with adjustment for NC>2. The London walking study, with pollutants
7 measured on site of outdoor exposures, provided some evidence for an independent
8 association for NCh. NCh-associated decrements in FEVi were attenuated to near null
9 with adjustment for UFP, EC, or PIVb 5 (McCreanor et al., 2007). Associations with
10 FEF25-75% decreased in magnitude and precision with copollutant adjustment but
11 remained negative (e.g., -0.45% [95% CI: -0.73, 0.17%] per 30-ppb increase in 2-h avg
12 NC>2 with adjustment for UFP, Spearman r = 0.58). These results indicate that the
13 decrements in some lung function parameters associated with near-road exposures of
14 relatively short duration (2 hours) were attributable to NC>2.
Controlled Human Exposure Studies
15 Most controlled human exposure studies examined adults and were reviewed in the 2008
16 ISA for Oxides of Nitrogen (U.S. EPA. 2008a). Consistent with epidemiologic findings,
17 numerous controlled human exposure studies generally did not report effects on lung
18 function in adults with asthma. As detailed in Table 5-10. exposures ranged from 200 to
19 4,000 ppb NO2 for 30 minutes to 6 hours, and most studies incorporated exercise in the
20 exposure period to assess lung function during various physiological conditions.
21 Whereas NO2 consistently induced increases in airway responsiveness in adults with
22 asthma (Section 5.2.2.1). direct changes in lung function or airway resistance were not
23 consistently found. Linnetal. (1985b) exposed adults with asthma and healthy adults to
24 4,000 ppb NO2 for 75 minutes and reported no changes in airway resistance after NO2
25 exposure in either group. Kleinman et al. (1983) found no statistically significant changes
26 in forced expiratory flows or airway resistance after exposure to 200 ppb NO2 for 2 hours
27 with light exercise; however, Bauer et al. (1986) reported statistically significant
28 decrements in forced expiratory flow rates in adults with asthma after exposure to
29 300 ppb NO2 for 30 minutes. Torres and Magnussen (1991) found no changes in lung
30 function in adults with asthma exposed to 250 ppb NO2 for 30 minutes; however,
31 exposure to 1,000 ppb NO2 for 3 hours with intermittent exercise (adjusted to individual
32 maximum workload) resulted in small reductions in FEVi in adults with asthma (Torres et
33 al.. 1995). Koenig etal. (1987) exposed healthy adolescents and those with asthma to 120
34 or 180 ppb NO2 and did not find any changes in lung function measured during or after
35 exposure relative to air exposures, nor were there differences between subjects with
36 asthma and healthy subjects.
January 2015 5-60 DRAFT: Do Not Cite or Quote
-------
1
2
o
6
4
5
There is no strong evidence for interactions between NC>2 and Os in controlled human
exposure studies. Jenkins etal. (1999) found no change in lung function in adults with
asthma following exposure to 200 ppb NC>2 for 6 hours (with or without 200 ppb Os) or
400 ppb NC>2 for 3 hours (without 400 ppb Os). Statistically significant decreases in FEVi
were found following the 3-hour exposure to Os and Os + NC>2.
Table 5-10 Controlled human exposure studies of individuals with asthma.
Study
Bauer et al.
(1986)
Jenkins et al.
(1999)
Disease Status3; n,
Sex; Age
(mean ± SD)
Asthma;
n = 15; 33 ± 7.8 yr
Asthma;
n = 9 M, 2 F;
31.2±6.6yr
Exposure Details (Concentration;
Duration)
300 ppb for 30 min (20 min at rest, 10 min of
exercise at VE > 3 times resting
(1)200ppbNO2for6h
(2) 200 ppb NO2 + 1 00 ppb O3 for 6 h
(3)400ppbNO2for3h
Outcomes Examined
Pulmonary function
before, during, and after
exposure.
Pulmonary function
tests before and after
exposure.
Jorres et al.
Healthy; n = 5 M, 3 F;
27 yr (range: 21-33)
Asthma; n = 8 M, 4 F;
27 ± 5 yr
(4) 400 ppb NO2 + 200 ppb O3 for 3 h
(1-4) Exercise 10 min on/40 min off at
VE = 32 L/min)
1,000 ppb for 3 h;
Exercise 10 min on/10 min off at individual's
maximum workload
Pulmonary function
tests before, during, and
after exposure.
Symptoms immediately,
6 h, and 24 h after
exposure.
BAL fluid analysis 1 h
after exposure (cell
counts, histamine,
prostaglandins).
Jorres and
Maqnussen
(1991)
Kleinman et al.
(1983)
Asthma; n = 9 M, 2 F;
29 yr (range: 17-55)
Asthma;
n = 12M, 19 F;
31 ± 11 yr
250 ppb for 1 h;
Rest for 20 min followed by 10 min of
exercise (VE = 30 L/min)
200 ppb for 2 h;
Exercise 15 min on/15 min off atVE = ~2
times resting
Airway resistance
measured before,
during, and after
exposure.
Pulmonary function
testing before and after
exposure.
Symptoms before,
immediately after, and
day after exposure.
January 2015
5-61
DRAFT: Do Not Cite or Quote
-------
Table 5-10 (Continued): Controlled human exposure studies of individuals with
asthma.
Study
Koeniq et al.
(1987)
Linn et al.
(1985b)
Disease Status3; n,
Sex; Age
(mean ± SD)
Healthy;
(1)n = 3M, 7F
(2 ) n = 4 M, 6 F;
Asthma;
(1)n = 4 M, 6 F
(2)n = 7M, 3F;
14.4 yr
(range: 12-19 yr)
Healthy; n = 16 M, 9 F;
(range: 20-36 yr)
Exposure Details (Concentration;
Duration)
(1) 120ppbNO2,
(2) 180ppbNO2;
(1-2) Exposures were 30 min at rest with
10 min of exercise at VE = 32.5 L/min
4,000 ppb for 75 min;
Outcomes Examined
Pulmonary function
tests before, during, and
after exposure.
Symptoms immediately
after and 1 day after.
Airway resistance
before, during, and after
Asthma; n = 12 M, 11 F;
(range: 18-34yr)
L/min and 50 L/min
exposure.
Symptoms before,
during, immediately
after, 1 day after and
1 week after exposure.
Riedl et al. Asthma
(2012) Phase 1: methacholine
challenge;
n = 10M, 5F;
37.3 ± 10.9 yr
Phase 2: cat allergen
challenge;
n=6M, 9F;
36.1 ± 12.Syr
350 ppb for 2 h;
Exercise 15 min on/15 min off at
VE = 15-20 L/min
Symptoms before,
during, 1-22 h after
exposure.
Vaqaqqini et al.
(1996)
Healthy; n
34 ± 5 yr
Asthma; n
29 ± 14 yr
= 7 M; 300 ppb for 1 h;
Exercise at VE = 25 L/min
= 4M, 4F;
Symptoms before and
2 h after exposure.
Cell counts in sputum 2-
h post-exposure.
COPD; n = 7 M;
58 ± 12 yr
BAL = bronchoalveolar lavage, COPD = chronic obstructive pulmonary disease, F = female, M = male, NO2 = nitrogen dioxide,
O3 = ozone, SD = standard deviation.
i
2
3
4
5
6
5.2.2.3 Respiratory Symptoms and Asthma Medication Use in
Populations with Asthma
The preceding epidemiologic evidence describing associations between short-term
increases in ambient NC>2 concentrations and decreases in lung function in children with
asthma, particularly those with atopy, supports evidence for NCh-related increases in
respiratory symptoms in children with asthma. Decreased lung function can indicate
airway obstruction (Section 4.3.2). which can cause symptoms. Further characterizing the
mode of action by which NCh exposure may induce respiratory symptoms is evidence for
January 2015
5-62
DRAFT: Do Not Cite or Quote
-------
1
2
o
6
4
5
6
7
8
9
10
NO2-induced increases in airway responsiveness (Section 5.2.2.1) and pulmonary
inflammation (Section 5.2.2.5) (Figure 4-1). Epidemiologic studies reviewed in the 2008
ISA for Oxides of Nitrogen consistently found increased respiratory symptoms in
children with asthma in association with increases in indoor, personal, and ambient NCh
concentrations (U.S. EPA, 2008a). There is weak support from a controlled human
exposure study of symptoms in adolescents with asthma. NC^-related increases in
respiratory symptoms in adults with asthma were found in previous epidemiologic studies
but not in controlled human exposure studies. Recent studies, most of which were
epidemiologic, continue to indicate associations between short-term increases in ambient
NO2 concentration and increases in respiratory symptoms in children with asthma.
11
12
13
14
15
16
17
18
Epidemiologic Studies
Epidemiologic studies examined respiratory symptoms in relation to ambient NC>2
concentrations rather than NO or NOx, and evidence is stronger for children with asthma
than adults with asthma. Across the various populations examined, symptom data were
collected by having subjects or their parents complete daily diaries for periods of 2 weeks
to several months. Heterogeneity in the number of consecutive days of follow-up and the
frequency of diary collection from study subjects do not appear to influence results.
Ambient NO2 concentrations, locations, and time periods for epidemiologic studies of
respiratory symptoms are presented in Table 5-11.
Table 5-11 Mean and upper percentile concentrations of nitrogen dioxide (NO2)
in epidemiologic studies of respiratory symptoms in populations
with asthma.
Study3
Schildcrout et al.
(2006)
Romieu et al.
(2006)
Location
Albuquerque, NM
Baltimore, MD
Boston, MA
Denver, CO
San Diego, CA
St. Louis, MO
Toronto, Canada
Mexico City,
Mexico
Study Period
Nov1993-
Sept1995
Oct 1998-Apr
2000
NO2 Metric
Analyzed
24-h avg NO2
1-h max NO2
Mean/Median
Concentration
ppb
Across cities:
17.8-26.0
66
Upper Percentile
Concentrations
ppb
90th: across cities
26.7-36.9 ppb
Max: 298
Seqala et al. (1998) Paris, France
Nov 1992- 24-h avg NO2 30.3b
May 1993
Pateletal. (2010) New York City and 2003-2005, 24-h avg NO2 NR
nearby suburb, NY months NR
Max: 64.9b
NR
January 2015
5-63
DRAFT: Do Not Cite or Quote
-------
Table 5-11 (Continued): Mean and upper percentile concentrations of nitrogen
dioxide (NO2) in epidemiologic studies of respiratory
symptoms in populations with asthma.
Study3
Barraza-Villarreal et
al. (2008),
Escamilla-Nufiez et
al. (2008)
Mannetal. (2010)
Zoraetal. (2013)
Jalaludin et al.
(2004)
Spira-Cohen et al.
(2011)
Sarnatetal. (2012)
Holquin et al.
(2007)
Gillespie-Bennett et
al. (2011)
Gent et al. (2009)
Delfinoetal. (2003)
Delfino et al. (2002)
O'Connor et al.
(2008)
Mortimer et al.
(2002)
Location
Mexico City,
Mexico
Fresno/Clovis, CA
El Paso, TX
Western and
Southwestern
Sydney, Australia
Bronx, NY
El Paso, TX and
Ciudad Suarez,
Mexico
Ciudad Juarez,
Mexico
Bluff, Dunedin,
Christchurch,
Porirua, Hutt
Valley, New
Zealand
New Haven
County, CT
Los Angeles, CA
(Huntington Park
area)
Alpine, CA
Boston, MA
Bronx, NY
Chicago, IL
Dallas, TX
New York, NY
Seattle, WA
Tucson, AZ
Bronx & East
Harlem, NY
Chicago, IL
Cleveland, OH
Study Period
June 2003-
June2005
Winter-
Summer,
2000-2005
Mar-June
2010
Feb-Dec 1994
Spring 2002,
Spring/Fall
2004, Spring
2005
Jan-Mar 2008
2001-2002
Sept 2006
Aug 2000-
Feb 2004
Nov 1999-Jan
2000
Mar-Apr 1996
Aug 1998-
July2001
June-Aug
1993
NO2 Metric
Analyzed
8-h max NO2
24-h avg NO2
96-h avg NO2
15-h avg NO2
(6 a.m.-9
p.m.)
6-h avg NO2
(9 a.m.-3
p.m.)
96-h avg NO2
1-week avg
NO2
4-week avg
NO2
NO2 — avg
time NR
1-h max NO2
8-h max NO2
1-h max NO2
8-h max NO2
24-h avg NO2
4-h avg NO2
(6 a.m.-
10 a.m.)
Mean/Median
Concentration
ppb
37.4
Median: 18.6
School 1: 9.3
School 2: 3.4
15.0
NR
El Paso schools:
4.5, 14.2; central
sites: 14.0, 18.5,
20.5
Ciudad Juarez
schools: 18.7,27.2;
central site: none
18.2
3.9
NR
7.2
5.9
24
15
NR
NR
Upper Percentile
Concentrations
ppb
Max: 77.6
75th: 24.7
Max: 52.4
Max: 16.2
Max: 7.5
Max: 47.0
NR
NR
NR
NR
NR
90th: 9.0; max:
90th: 7.9; max:
Max: 53
Max: 34
NR
NR
14
11
January 2015
5-64
DRAFT: Do Not Cite or Quote
-------
Table 5-11 (Continued): Mean and upper percentile concentrations of nitrogen
dioxide (NO2) in epidemiologic studies of respiratory
symptoms in populations with asthma.
Study3
Location
Study Period
NO2 Metric
Analyzed
Mean/Median
Concentration
ppb
Upper Percentile
Concentrations
ppb
Detroit, Ml
St. Louis, MO
Washington, DC
Just et al. (2002)
Ostroetal. (2001)
Boezen et al.
(1998)
Forsberq et al.
(1998)
von Klot et al.
(2002)
Maestrelli et al.
(2011)
Wiwatanadate and
Liwsrisakun (2011)
Hiltermann et al.
(1998)
Laurent et al.
(2009)
Carlsen et al.
(2012)
Kim etal. (2012)
Karakatsani et al.
(2012)
Feo Brito et al.
(2007)
Paris, France
Central Los
Angeles, CA
Amsterdam
Meppel,
the Netherlands
Landskrona,
Sweden
Erfurt, Germany
Padua, Italy
Chiang Mai,
Thailand
Bilthoven, the
Netherlands
Strasbourg, France
Reykjavik, Iceland
Seoul and Kyunggi
Province, South
Korea
Amsterdam,
the Netherlands
Athens, Greece
Birmingham, U.K.
Helsinki, Finland
Ciudad Real
Puertollano, Spain
Apr-June
1996
Aug-Oct 1993
Winter 1993-
1994
Jan-Mar, yr
NR
Sept 1996-
Nov1997
1999-2003
Aug 2005-
June2006
July-Oct1995
2004, all yr
Mar 2006-
Dec 2009
2005-2009
Oct 2002-
Mar 2004
May-June
2000-2001
24-h avg NO2
24-h avg NO2
24-h avg NO2
24-h avg NO2
24-h avg NO2
24-h avg NO2
24-h avg NO2
24-h avg NO2
24-h avg NO2
Dispersion
model
24-h avg NO2
1-h max
24-h avg NO2
24-h avg NO2
24-h avg NO2
28.6b
Los Angeles: 79.5
Pasadena: 68.1
24.5b
14.2b
16.2b
24.5b
Across seasons and
yr: 20. 9-37. Ob
17.2
11. 2b
18.6b
11. 7b
27.4b
Asthma
exacerbation 34.3
spring, 26.6
summer, 30.6 fall,
38.8 winter
No asthma
exacerbation: 32.7
spring, 26.0
summer, 30.6 fall,
37.7 winter
20.4b
21. 2b
18.3b
12.1b
17.4b
29.5b
Max: 59.0b
Max: 220
Max: 170
Max: 40. 4b
Max: 28.9b
Max: 38. 1b
Max: 63.3b
75th: 23.0-42.5b
90th: 26.5
Max: 37.4
Max: 22.5b
NR
Max: 52.9b
Max: 64. 4b
75th: asthma
exacerbation:
41.3 spring, 35.3
summer, 42.0 fall,
46.8 winter
No asthma
exacerbation:
41.4 spring, 35.2
summer, 41.6 fall,
48.9 winter
Max: 51. 8b
Max: 59.0b
Max: 44.2b
Max:41.4b
Max: 35.6b
Max: 100.5b
NR = not reported, NO2 = nitrogen dioxide.
aStudies presented in order of first appearance in the text of this section.
bConcentrations converted from |jg/m3 to ppb using the conversion factor of 0.532 assuming standard temperature (25°C) and
pressure (1 atm).
January 2015
5-65
DRAFT: Do Not Cite or Quote
-------
Children with Asthma
I Several recent studies add to the evidence for increases in respiratory symptoms in
2 children with asthma associated with short-term increases in ambient NC>2. Across
3 previous and recent studies, there is heterogeneity in the magnitude and precision (width
4 of 95% CIs) of the association. However, the results collectively indicate a pattern of
5 elevated risk of respiratory symptoms across the various symptoms and lags of NC>2
6 exposure examined (Figure 5-4 and Table 5-12). The consistency of findings also is
7 supported by a meta-analysis of 24 mostly European studies and some U.S. studies,
8 including several reviewed in the 2008 ISA for Oxides of Nitrogen. In the meta-analysis,
9 there was some evidence of publication bias with exclusion of the multicounty European
10 PEACE studies, but with adjustment for publication bias, an increase in 24-h avg NC>2
11 was associated with increased risk of asthma symptoms ("Weinmayr et al.. 2010). Across
12 individual studies reviewed in this ISA, the most consistent results were for total
13 respiratory or asthma symptoms, wheeze, and cough. Increases in ambient NC>2
14 concentrations were not consistently associated with increases in rescue inhaler or
15 beta-agonist use in children with asthma (Patel et al.. 2010; Romieu et al.. 2006;
16 Schildcrout et al.. 2006; Segalaetal.. 1998).
17 Study populations were recruited from schools, asthma or allergy clinics, or doctors'
18 offices. Asthma was assessed by parental report of physician-diagnosed asthma or
19 clinical examination. Neither of these methodological issues appeared to affect whether
20 an association was found. Many studies reported follow-up participation rates of
21 77-92%, and none reported selective drop-out among a particular group within the study
22 population. In a-priori-determined comparisons of children with and without asthma, one
23 study found stronger associations in children with asthma (Patel etal., 2010); another
24 found stronger associations in children without asthma, 72% of whom had atopy
25 (Barraza-Villarreal et al.. 2008; Escamilla-Nunez et al.. 2008).
26 Many asthma study populations had high prevalence of atopy (47-89%), and larger
27 NO2-associated increases in symptoms were found in children with asthma who also had
28 allergies (Zoraet al.. 2013; Mannetal.. 2010). These results were based on 16 to 47% of
29 the study populations but are coherent with experimental evidence for NCh-induced
30 allergic responses in adults with asthma and animal models of allergic disease
31 (Section 5.2.2.5). Study populations also varied in asthma severity; some studies
32 examined mostly children with mild, intermittent asthma and others examined children
33 with persistent asthma. Comparisons by asthma severity indicated larger NCh-related
34 increases in respiratory symptoms among children with mild, intermittent asthma than
35 severe or moderate asthma (Mann etal. 2010; Segalaetal.. 1998). but these results also
36 were based on small numbers of subjects. Jalaludin et al. (2004) found that elevated risk
January 2015 5-66 DRAFT: Do Not Cite or Quote
-------
was limited to children with more severe asthma (asthma plus airway
hyperresponsiveness). But, results were based on a 3-pollutant model that can produce
unreliable results because of potential multicollinearity.
Study
Symptom Composite
Spira-Cohen et al. (2011)
Delfino et al. (2003)
Delfino et al. (2002)
Schildcroutetal. (2006)
Mortimer etal. (2002)
Just et al. (2002)
Segalaetal. (1998)
O'Connor et al. (2008)
Wheeze
Spira-Cohen et al. (2011)
Gent et al. (2009)
Mann etal. (2010)
Jalaludin et al. (2004)
Escamilla-Nunezetal. (2008)
Barraza-Villarreal et al. (2008)
Ostroetal. (2001)
Pateletal. (2010)
Asthma Medication Use
Schildcroutetal. (2006)
Jalaludin et al. (2004)
Romieuetal. (2006)
Segalaetal. (1998)
NO2 Metrics Analyzed
6-h avg, lag 0 day
24-h avg, lag 0 day
24-h avg, 0-2 day sum
4-h avg, lag 1 -6 day avg
24-h avg, lag 1-19 day avg
6-h avg, lag 0 day
Averaging time NR, lag 0 day
24-h avg, lag 2 day
15-h avg, lag 0 day
1-h max, lag 1 day
1-h max, lag 0 day
1-h max, lag 3 day
24-h avg, lag 0 day
24-h avg, lag 0 day
15-h avg, lag 0 day
1 -h max, lag 1 -6 day avg
Exposure
Assessment
School
Central site
Central site
Central sites
Central sites
Central sites
Central sites
Central sites
School
Central site
Central site
Central site
Central sites
Central sites
Central sites
Central sites
Central sites
Central site
Central site
Central sites
Subgroup
All subjects
No medication use
Mild asthma
Moderate asthma
All subjects
Cat allergic
Boys, mild asthma
GSTM1 null
GSTM1 positive
Mild asthma
I
I.
I .
1 .*
- •
t
I .
I
1 . *
I
i a
1 •
1 * .
k
1
r
• .
i
0.0 0.5 1.0 1.5 2.0 2.5
Odds ratio per increase in NO2 (95% Cl)a
3.0
Note: Results from more informative studies in terms of the exposure assessment method and potential confounding considered are
presented first within an outcome group. Red = recent studies, black = previous studies. Study details and quantitative results are
reported in Table 5-12. The figure presents a subset of results included in Table 5-12 for which quantitative results were available
for NO2 examined as a linear variable and for specific outcomes examined by multiple studies.
aEffect estimates are standardized to a 20-ppb, 25-ppb, and 30-ppb increase for 24-h avg, 8-h max, and 1-h max NO2, respectively.
Effect estimates for other averaging times (4-h avg to 15-h avg) are not standardized and presented as reported in their respective
studies (Section 5.1.2.3).
Figure 5-4 Associations of ambient nitrogen dioxide (NO2) concentrations
with respiratory symptoms and asthma medication use in
children with asthma.
January 2015
5-67
DRAFT: Do Not Cite or Quote
-------
Table 5-12 Epidemiologic studies of respiratory symptoms and asthma medication use in children with asthma.
Study
Population Examined and Methodological Details
NO2 Metrics Odds Ratio (95% Cl)
Analyzed Lag Day Single-Pollutant Model3
Copollutant Examination
Studies with small spatial scale exposure assessment and/or examination of copollutant confounding
tZoraetal. (2013)
El Paso, TX
n = 36, mean age 9.3 (SD: 1.5) yr, 47% with atopy
Repeated measures. Asthma control questionnaire given weekly at
school for 13 weeks. Questionnaire ascertains symptoms, activity
limitations, asthma medication use. Recruitment from schools via school
nurses. Parent report of physician-diagnosed asthma. No information on
participation rate. Linear mixed effects model adjusted for random subject
effect and humidity, temperature, school.
NO2-school
outdoor
24-h avg
1 school 91 m
from major
road, 1 school
in residential
area.
0 4 avg change in asthma control
score (higher score
indicates poorer control):
Allergy, n = 17
0.56 (-0.10, 1.22)
No allergy, n = 19
-0.29 (-1.07, 0.49)
No copollutant models
analyzed for subgroups.
BC, benzene, toluene, also
associated with poorer
asthma control.
Correlations with NCbweak
' to high. Spearman r = 0.29
to 0.56 for BC, 0.37 to 0.71
for benzene, 0.16 to 0.71
for toluene.
tSarnatetal. (2012)
El Paso, TX and Ciudad Juarez, Mexico
n = 29 per city, ages, 6-12 yr, asthma and current symptoms
Repeated measures. Daily symptom diaries. Recruitment from schools
representing a gradient of traffic, subjects from non-smoking homes. No
information on participation rate. Self report of physician-diagnosed
asthma. GLM with subject as random effect and adjustment for school,
temperature, relative humidity. Adjustment for medication use, cold
symptoms did not alter results.
NO2-school 0-4 avg
outdoor
24-h avg
In each city,
1 school 91 m
from major
road, 1 in
residential area.
No quantitative results
reported; associations
reported to be consistent
with the null.
No copollutant model.
tHolguin et al. (2007)
Ciudad Juarez, Mexico
n = 95, ages 6-12 yr, 78% mild, intermittent asthma, 58% with atopy
Repeated measures. Daily symptom diaries given by parents for 4 mo,
checked biweekly. 87% participation. Parent report of physician-
diagnosed asthma. Linear and nonlinear mixed effects model with
random effect for subject and school adjusted for sex, body mass index,
day of week, season, maternal and paternal education, passive smoking
exposure.
NO2-school
outdoor
24-h avg
Schools located
239-692 m
from homes.
0-6 avg No quantitative results
reported. No air pollutants
reported to be associated
with respiratory symptoms.
No copollutant model.
Road density at home and
school reported not to be
associated with respiratory
symptoms.
January 2015
5-68
DRAFT: Do Not Cite or Quote
-------
Table 5-12 (Continued): Epidemiologic studies of respiratory symptoms and asthma medication use in children
with asthma.
Study
Population Examined and Methodological Details
NO2 Metrics Odds Ratio (95% Cl)
Analyzed Lag Day Single-Pollutant Model3
Copollutant Examination
tSpira-Cohen et al. (2011): Spira-Cohen (2013)
Bronx, NY
n = 40, ages 10-12 yr, 100% nonwhite, 44% with asthma ED visit or
hospital admission in previous 12 mo
Repeated measures. Daily symptom diaries for 1 mo, checked daily.
454 observations. Recruitment from schools by referrals from school
nurses. Parental report of physician-diagnosed asthma. 89% time
indoors. No information on participation rate. Mixed effects model with
subject as random effect adjusted for temperature, height, sex.
Adjustment for school (indicator of season) did not alter results.
NO2-school
outdoor
6-h avg
(9 a.m.-3 p.m.)
Schools
53-737 m from
highways with
varying traffic
counts. Most
children walk to
school.
0
Total symptoms:
1.10(0.84, 1.45)
Wheeze:
1.20(0.75, 1.9)
OR per60-ppb increase
NO2 (5th to 95th percentile)
Personal EC associated
with symptoms with NO2
adjustment.
No quantitative data
reported.
tGillespie-Bennett et al. (2011)
Bluff, Dunedin, Christchurch, Porirua, Hutt Valley, New Zealand
n = 358, ages 6-1 Syr
Cross-sectional. Daily symptom diaries for 112 days. Recruitment from a
home heating intervention. 77% participation. Mixed effects model with
log-transformed NO2 and random effect for subject. Adjustment for age,
sex, region, ethnicity, intervention, income, temperature did not alter
results.
NCb-home
outdoor
24-h avg
1 measure per
subject.
NO2-home
indoor
24-h avg
Up to 4
measures per
subject.
4-week Lower respiratory
av9 symptom:
1.09(0.78, 1.51)
Reliever inhaler:
1.47(0.96,2.26)
OR per log increase NO2
Lower respiratory
symptom:
1.14(1.12, 1.16)
Reliever inhaler:
1.14(1.11, 1.17)
No copollutant model.
No other pollutants
examined.
tGentetal. (2009)
New Haven county, CT
n = 149, ages 4-12 yr
Repeated measures. Daily symptom diaries reported monthly.
Recruitment from larger cohort, pediatric asthma clinic, and school.
Parent report of physician diagnosed asthma. No information on
participation rate. GEE adjusted for season, day of week, date, motor
vehicle factor obtained by source apportionment.
NO2-central site
Avg time not
reported
1 site 0.9-30
km of homes
(mean
10.2km).
NR
Wheeze with source
apportionment factor of EC,
zinc, lead, copper,
selenium: 1.08(0.99, 1.18).
Factor results not altered
by NO2 adjustment.
Moderate correlation with
NO2. Pearson r= 0.49.
January 2015
5-69
DRAFT: Do Not Cite or Quote
-------
Table 5-12 (Continued): Epidemiologic studies of respiratory symptoms and asthma medication use in children
with asthma.
Study
Population Examined and Methodological Details
NO2 Metrics Odds Ratio (95% Cl)
Analyzed Lag Day Single-Pollutant Model3
Copollutant Examination
Delfino et al. (2003)
Los Angeles, CA (Huntington Park region)
n = 16, ages 10-16 yr, 100% Hispanic, 27% on anti-inflammatory
medication
Repeated measures. Daily symptom diaries for 3 months, collected
weekly. Recruitment from schools of non-smoking children from
non-smoking homes. Self or parental report of physician diagnosed
asthma. 92% follow-up participation. GEE with autoregressive parameter
and adjusted for respiratory infections. Excluded potential confounding by
weekend, maximum temperature.
NO2-central site
4.8 km of home
& school
8-h max
1-h max
Asthma symptoms not
interfering with daily activity
1.27(1.05, 1.54)
1.18(0.96, 1.43)
8-h max
1-h max
Asthma symptoms
interfering with daily activity
1.40(1.02, 1.92)
1.27(0.81, 1.99)
ORs per 1.4 ppb increase
in 8-h max and 2.0 ppb
increase in 1-h max NO2
(interquartile ranges).
Copollutant model results in
figure. ORs for NO2 not
altered byxylene or toluene
adjustment. Smaller but
positive ORs for NO2, wider
95% Cl with adjustment for
benzene, ethylbenzene,
• acetylaldehyde,
formaldehyde. Moderate to
high correlations with 8-h
max NO2. Spearman
r = 0.57 (benzene) to
0.72 (xylene). No
interactions between NO2
andVOCs. ORsforVOCs
attenuated with NO2
adjustment.
No copollutant model with
PM25, EC, OC. r
= 0.38-0.62. No
association with CO.
Delfino et al. (2002)
Alpine, CA and adjacent areas
n = 22, ages 9-19 yr, 36% with mild persistent or more severe asthma,
77% with atopy
Repeated measures. 92% follow-up. Daily symptom diaries for 61 days,
collected weekly or biweekly. 1,248 observations (94% of expected).
Recruitment from schools. Asthma diagnosis based on referrals from
health maintenance organization and newspaper advertisements.
Subjects were nonsmokers from nonsmoking homes. GEE with
autoregressive lag 1 correlation matrix with no covariates. Adjustment for
day of week, linear trend, temperature, humidity did not alter results.
Adjustment for respiratory infection increased pollutant ORs.
NO2-central site
1-h max
8-h max
1 site 1-4.7 km
from subjects'
homes.
Symptoms interfering with
daily activity:
All subjects:
1.35(0.82,2.20)
No anti-inflammatory
medication, n = 12
1.80(0.89,3.64)
On anti-inflammatory
medication, n = 10
0.91 (0.21,3.97)
All subjects:
1.65(0.94,2.89)
No anti-inflammatory
medication, n = 12
2.25(1.09,4.63)
On anti-inflammatory
medication, n = 10
1.10(0.22,5.46)
Positive interaction for 8-h
max NO2 with 1-h max
PMio(p< 0.01) and 1-h
maxOs(p = 0.12).
Fungi and pollen allergen
associated with symptoms.
No NO2-allergen
interactions. No quantitative
results for NO2-allergen
copollutant models, but
ORs reported to decrease.
Moderate correlations with
NO2. Pearson r= 0.29 for
fungi, 0.27 for pollen, 0.55
for PMm
January 2015
5-70
DRAFT: Do Not Cite or Quote
-------
Table 5-12 (Continued): Epidemiologic studies of respiratory symptoms and asthma medication use in children
with asthma.
Study
Population Examined and Methodological Details
NO2 Metrics Odds Ratio (95% Cl)
Analyzed Lag Day Single-Pollutant Model3
Copollutant Examination
Schildcroutetal. (2006)
Albuquerque, NM; Baltimore, MD; Boston, MA; Denver, CO; San Diego,
CA; St. Louis, MO; Toronto, ON, Canada (CAMP cohort)
n = 990, ages, 5-12 yr, mild to moderate asthma
Repeated measures. Daily symptom diaries for 21-201 days. No
information on participation rate. GEE for individual cities combined for
study-wide estimates. City-specific models adjusted for day of week,
ethnicity, annual family income, response to methacholine, maximum
temperature, humidity, temperaturexhumidity, calendar date. Pollutant
analyzed as daily deviation from subject mean.
NO2-central site
24-h avg
Average of
multiple sites
within 80 km of
ZIP code.
0
0-2 sum
Asthma symptoms:
1.06(1.00, 1.12)
Rescue Inhaler use:
1.04(1.00, 1.08)
Asthma symptoms:
1.05(1.01, 1.09)
Joint effect models
NO2+CO: 1.07(1.0, 1.14)
NO2+SO2: 1.06(0.98, 1.15)
NO2+PMio:1.06(0.99, 1.13)
Moderate to high
correlations with NO2.
r = 0.23 to 0.58 for SO2,
0.26 to 0.64 for PMio, 0.63
to 0.92 for CO.
tMannetal. (2010)
Fresno, Clovis, CA
n = 280, ages 6-11 yr, 47% mild persistent asthma, 25% moderate to
severe asthma, 63% with atopy
Repeated measures. Daily symptom diaries for 14 days every 3 mo.
Recruitment from schools, advertisements, physician's offices, local
media. Imputed wheeze values for 7.6% days. Participation from 89% of
original cohort. Group examined representative of original cohort. GEE
adjusted for fitted daily mean wheeze, home ownership, race, sex,
asthma severity, panel group, 6-mo cohort, 1-h minimum temperature.
Adjustment for medication use did not alter results.
NO2-central site
24-h avg
1 site within
20 km of
homes.
Wheeze:
All subjects:
1.24(1.05, 1.48)
Fungi allergic, n = 85
1.61 (1.24,2.08)
Cat allergic, n = 49
1.73(1.14,2.62)
Boys, intermittent asthma,
n=47
2.58(1.61,4.13)
With PMio-2.s, all subjects:
1.14(0.95, 1.37).
PMio-2.s association not
• altered by NO2 adjustment.
Weak correlation with NO2.
r=0.12.
Mortimer etal. (2002)
Bronx and East Harlem, NY; Chicago, IL; Cleveland, OH; Detroit, Ml; St.
Louis, MO; Washington, DC (NCICAS cohort)
n = 846, ages 4-9 yr
Repeated measures. Daily symptom data collected for 2-week periods
every 3 mo. Recruitment from ED visits and clinics. Parent report of
physician-diagnosed asthma and symptoms in previous 12 mo or asthma
symptoms for >6 weeks, or family history of asthma. Participation from
55% full cohort. Sample representative of full cohort except for greater
asthma medication use. Mixed effects model adjusted for city, follow-up
period, day of study, 24-h rainfall, 12-h avg temperature.
NO2-central site
4-h avg
(6 a.m.-10
a.m.)
Average of all
city monitors.
Lag 1 6 Morning symptoms:
av9 1.48(1.02,2.16)
OR per 20 ppb increase in
NO2 (interquartile range).
With 03 (summer):
1.40(0.93,2.09)
Weak correlation with NO2.
r=0.27.
Os effect estimate also
slightly attenuated.
SO2 and PMio also
associated with symptoms.
Correlations with NO2 not
reported.
January 2015
5-71
DRAFT: Do Not Cite or Quote
-------
Table 5-12 (Continued): Epidemiologic studies of respiratory symptoms and asthma medication use in children
with asthma.
Study
Population Examined and Methodological Details
NO2 Metrics
Analyzed Lag Day
Odds Ratio (95% Cl)
Single-Pollutant Model3
Copollutant Examination
Studies with central site exposure assessment and no examination of copollutant confounding
Jalaludin et al. (2004)
Sydney, Australia
n = 125, mean age 9.6 yr, 45 with wheeze, asthma, and airway
hyperresponsiveness, 60 with wheeze and asthma, 20 with wheeze
Repeated measures. Daily symptom diary mailed in monthly for 11 mo.
Recruitment from schools. Parent-report of physician-diagnosed asthma.
84% follow-up participation. GEE adjusted for time trend, temperature,
humidity, number of hours spent outdoors, total pollen and alternaria,
season.
tEscamilla-Nufiez et al. (2008)
Mexico City, Mexico
n = 147, ages 6-14 yr, 43% with persistent asthma, 89% atopy
Repeated measures. Symptom data collected every 15 days for mean
22 weeks. Children with asthma recruited from pediatric clinic. Asthma
severity assessed by pediatric allergist. No information on participation
rate. Linear mixed effects model with random effect for subject and
adjusted for asthma severity, atopy, lag 1 minimum temperature, time,
sex. Adjustment for outdoor activities, smoking exposure, season did not
alter results.
tBarraza-Villarreal et al. (2008)
Mexico City, Mexico
n = 126, ages 6-14 yr, 44% persistent asthma, 89% with atopy
Part of same cohort as above. No information on participation rate. Linear
mixed effects model with random effect for subject and adjusted for sex,
body mass index, lag 1 minimum temperature, ICS use, time. Adjustment
for outdoor activities, smoking exposure, anti-allergy medication use,
season did not alter results.
Romieu et al. (2006)
Mexico City, Mexico
n = 151, mean age 9 yr, mild or moderate asthma
NO2-central site 0
15-h avg
(6 a.m. -9 p.m.)
Site within 2 km
of schools.
NO2-central site 1
1-h max
Site within 5 km
of school or
home.
NO2-central site 0
1-h max
Site within 5 km
of school or
home.
Low correlation
for central site
vs. school:
Spearman
r=0.21
NO2-central site 1-6 avg
1-h max
Site within 5 km
of home.
Wheeze:
1.01 (0.94, 1.09)
Wet cough:
1r\c / *i nn *i *ir\\
.05 (1.00, 1.10)
Beta agonist use:
1r\ *i /r\ o~7 *i r\c\
.01 (0.97, 1.05)
OR per 8.2 ppb increase in
NO2 (interquartile range)
Cough: 1.07(1.02, 1.12)
Wheeze: 1.08(1.02, 1.14)
Wheeze: 1.09(1.03, 1.15)
Cough: 1.09(1.04, 1.14)
Cough by genotype:
GSTM1 null
1.09(1.00, 1.19)
GSTM1 positive
1.19(1.11, 1.27)
GSTP1 lie/lie or Ile/Val
NO2 associations found in
children with asthma/airway
hyperresponsiveness but
examined only in
multipollutant model with
Osand PMm
Negative or weak
correlations with NO2.
r=-0.31 forOs, 0.26 for
PMio.
No copollutant model.
PM2.sand Os also
associated with symptoms.
No statistically significant
interaction between NO2
and PM25 or Os.
Quantitative results not
reported.
No copollutant model.
PM2.sand Os also
associated with symptoms.
Moderate correlations with
NO2. Pearson r= 0.61 for
PM2.5, 0.28 for Os.
No copollutant model.
Associations with Os found
with different variants than
NO2.
Moderate correlation with
NO2 Pearson r ~~ 0 57 for
Osand PMio.
1.19(1.11, 1.27)
January 2015
5-72
DRAFT: Do Not Cite or Quote
-------
Table 5-12 (Continued): Epidemiologic studies of respiratory symptoms and asthma medication use in children
with asthma.
Study
Population Examined and Methodological Details
NO2 Metrics Odds Ratio (95% Cl)
Analyzed Lag Day Single-Pollutant Model3
Copollutant Examination
Repeated measures. Daily symptom diaries 61-92 days per subject,
collected weekly. Recruitment from allergy clinic as part of a Vitamin C/E
supplementation trial. Diagnosis by clinical examination. 99% follow-up
participation. GEE adjusted for supplementation group, minimum
temperature, smoking exposure, asthma severity, time.
GSTP1 Val/Val
1.08(0.99, 1.18)
BD use by genotype:
GSTM1 null
0.94(0.87, 1.02)
GSTM1 positive
1.09(1.02, 1.17)
GSTP1 lie/lie or Ile/Val
1.08(1.02, 1.14)
Ostroetal.(2001)
Central Los Angeles and Pasadena, CA
n = 138 (83% LA), ages 8-13 yr, 85% mild or moderate asthma, 100%
African American
Repeated measures. 90% follow-up. Daily symptom diaries for 13 weeks,
mailed in weekly. Excluded subjects returning diaries after 2 weeks.
9,126 observations. Recruitment from hospitals, urgent care clinics,
medical practices, asthma camps in Los Angeles and school nurses in
Pasadena. GEE adjusted for day of study, age, income, town, lag 1
temperature, lag 1 humidity.
tPateletal. (2010)
New York City and nearby suburb, NY
n = 57, ages 14-20 yr
Repeated measures. Daily symptom diaries for 4-6 weeks, collected
weekly. Recruitment from schools. Self-report of physician-diagnosed
asthma. 75-90% participation across schools. GLMM with random effect
for subject and school and adjusted for weekend, daily 8-h max Os, urban
location. Adjustment for season, pollen counts did not alter results.
Just etal. (2002)
Paris, France
n = 82, ages 7-15 yr, asthma attack in previous 12 mo and daily asthma
medication use, 90% atopy
NO2-central site
1-h max
Los Angeles
site within
1 6 km of 90%
of subjects'
homes.
Pasadena site
within 8 km of
subjects'
homes.
NO2-central site
24-h avg
1 site 2.2-9.0
km from
schools, 1 site
40 km from
schools.
NO2-central site
24-h avg
Average of
1 1 sites
GSTP1 Val/Val
0.94(0.85, 1.04)
3 Onset of shortness of
breath:
1.08(0.99, 1.18)
Onset of wheeze:
1.08 (1.02, 1.13)
Onset of Cough:
1r\~7 i *i nn *i *i c \
.07 (1.00, 1.15)
No quantitative results for
extra medication use but
reported not to be
associated with NO2.
0 Wheeze:
1.16(0.93, 1.45)
Chest tightness:
1.26 (1.00, 1.58)
0 Asthma attack:
1.75(0.82,3.70)
Night cough:
2.11 (1.20,3.71)
No copollutant model.
Symptoms associated with
PM2.5, PM-io, fungi.
Weak to moderate
correlations with NO2.
r= 0.18 for pollen,
0.26-0.48 for fungi, 0.34 for
PM2.5, 0.63 for PM-io.
No copollutant model with
BC.
BC also associated with
symptoms.
Across locations, moderate
to high correlations with
NO2. Spearman
r= 0.56-0.90.
No copollutant model
BS associated with cough.
High correlation with NO2.
Pearson r= 0.92.
January 2015
5-73
DRAFT: Do Not Cite or Quote
-------
Table 5-12 (Continued): Epidemiologic studies of respiratory symptoms and asthma medication use in children
with asthma.
Study
Population Examined and Methodological Details
NO2 Metrics Odds Ratio (95% Cl)
Analyzed Lag Day Single-Pollutant Model3
Copollutant Examination
Repeated measures. Daily symptom diaries for 3 mo, collected weekly.
Recruitment from hospital outpatients. 82% participation. GEE adjusted
for time trend, day of week, pollen, temperature, humidity.
Segalaetal. (1998)
Greater Paris area, France
n = 43 mild asthma, 41 moderate asthma, 89% atopy, 69% ICS users,
ages 7-15 yr
Repeated measures. Daily symptom diary for 25 weeks, collected weekly.
Recruitment from outpatients of children's hospital. 84% follow-up
participation. GEE adjusted for day of week, time trend, temperature,
humidity, age, sex.
NO2-central site
24-h avg
Average of
8 sites
Incident asthma:
Mild asthma, n = 43
1.89(1.13,3.15)
Moderate asthma, n = 41
1.31 (0.85,2.03)
No copollutant model.
Associations also found
with BS, PMi3, & SO2.
. Moderate correlations with
NO2. Pearson r= 0.61 for
BS, 0.55 for PMis, 0.54 for
. SO2.
Beta agonist use:
Mild asthma, n = 43
1.27(0.82, 1.98)
Moderate asthma, n = 41
1.56(0.51,4.73)
tO'Connor et al. (2008)
Boston, MA; Bronx, NY; Chicago, IL; Dallas, TX; New York, NY; Seattle,
WA; Tucson, AZ (ICAS cohort)
n = 861, ages 5-12 yr, persistent asthma and atopy, 82% black or
Hispanic
Repeated measures. Symptom data collected for 2 week period every
2 mo for 2 yr. Recruitment from intervention of physician feedback. 89%
of maximum possible diaries obtained. Mixed effects model adjusted for
site, mo, sitexmo interaction, temperature, intervention group.
NO2-central site
24-h avg
All monitors
near home, not
near industry.
Median
distance to
site = 2.3 km.
1-19 avg wheeze-cough:
1.17(0.99, 1.37)
Slow Play:
1.25(1.04, 1.51)
Missed school in 2 week
period:
1.65(1.18,2.32)
Only 3-pollutant model
analyzed.
Associations also found
with PM2.5 and CO.
Moderate correlations with
NO2. r=0.59forPM25,
0.54 for CO.
Note: More informative studies in terms of the exposure assessment method and potential confounding considered are presented first.
BD = bronchodilator; CAMP = Childhood Asthma Management Program; GEE = generalized estimating equations; GLM = generalized linear model; GLMM = generalized linear
mixed model; GST = Glutathione-S-transferase; ICAS = Inner City Asthma Study; ICS = inhaled corticosteroid; NCICAS = National Cooperative Inner-city Asthma Study,
Cl = confidence interval, CO = carbon monoxide, EC = elemental carbon, ED = emergency department, NO2 = nitrogen dioxide, O3 = ozone, OC = organic carbon, OR = odds ratio,
PM = particulate matter, SD = standard deviation, SO2 = sulfur dioxide, VOC = volatile organic compound.
aEffect estimates are standardized to a 20 ppb for 24-h avg NO2, 25 ppb for 8-h max, and a 30-ppb increase for 1-h max NO2. Effect estimates for other averaging times are not
standardized but presented as they are reported in their respective studies (Section 5.1.2.3).
fStudies published since the 2008 ISA for Oxides of Nitrogen.
January 2015
5-74
DRAFT: Do Not Cite or Quote
-------
1 Several studies are noteworthy for NC>2 exposure assessment in subjects' locations or
2 analysis of the influence of other traffic-related pollutants on NO2 associations. Outdoor
3 school or home NC>2 concentrations are spatially aligned with a location of subjects and
4 may better represent ambient exposures for that location. In a group of 17 children with
5 asthma and allergy in El Paso, TX, a 20-ppb increase in outdoor school 4-day avg NC>2
6 was associated with a 0.56 (95% CI: -0.17, 1.28)-point poorer asthma control score
7 (composite of symptoms, activity limitation and asthma medication use) (Zoraetal..
8 2013). Among children in Bronx, NY, 6-h avg school-day NCh (9 a.m.-3 p.m.) was
9 associated with total symptoms (OR: 1.10 [95% CI: 0.84, 1.45] per 60-ppb NO2) and
10 wheeze (OR: 1.05 [95% CI: 0.87, 1.39]) but 95% CIs were wide (Spira-Cohen et al..
11 2011). Other studies did not indicate associations of school or home NO2 with respiratory
12 symptoms in children with asthma, but it is not clear whether their results represent
13 inconsistency in the evidence base. Studies conducted in El Paso, TX and Ciudad Juarez,
14 Mexico only reported that NO2 was not associated with respiratory symptoms in children
15 with asthma but did not report quantitative results (Sarnat et al.. 2012; Holguin et al..
16 2007). Outdoor home NO2 was associated with reliever inhaler use but not respiratory
17 symptoms among children with asthma in multiple New Zealand towns (Gillespie-
18 Bennett et al.. 2011). However, daily outcomes were analyzed with a single 4-week
19 sample of NO2, which cannot represent temporal variability in exposure. Home indoor
20 NO2, which was represented as up to four measurements per subject, showed stronger
21 associations with both outcomes.
22 Most studies observed NO2-related increases in respiratory symptoms with adjustment for
23 temperature, humidity, season, and day of week. A few studies additionally adjusted for
24 asthma medication use, colds, smoking exposure, and allergens (Table 5-12). Some
25 studies with central site exposure assessment are informative for their analysis of
26 copollutant confounding or interactions. Studies with school and central site exposure
27 assessment found symptoms associated with the traffic-related pollutants EC/BC, OC,
28 CO, and VOCs, which showed a wide range of correlations with NO2 (r = 0.16-0.92)
29 (Table 5-12). In New Haven County, CT, NO2 was associated with wheeze in children
30 with asthma with adjustment for a source apportionment factor comprising EC, zinc,
31 lead, copper, and selenium (OR: 1.08 [95% CI: 0.99, 1.18] per unspecified increase in lag
32 0 NO2) (Gent et al.. 2009). In El Paso, TX, neither NO2 nor BC was associated with
33 asthma control in a copollutant model (Zoraet al.. 2013). However, these results were
34 based on the whole study population. NO2 and BC were associated with asthma control
35 only in children with asthma and allergies; thus, the copollutant model results do not
36 clearly inform potential confounding. Among children in Los Angeles, CA, CO was not
37 associated with symptoms, and NO2-asthma symptom associations were relatively
38 unchanged with adjustment for various VOCs. No NO2-VOC interaction was found
39 (Delfino et al.. 2003). NO2 concentrations assigned from a central site within 4.8 km of
January 2015 5-75 DRAFT: Do Not Cite or Quote
-------
1 children's homes may well represent the broad variability in NO2 in the study area given
2 the high correlations observed among monitors in the area (Section 2.5.3). although it is
3 uncertain whether variability within neighborhoods is adequately represented.
4 In the multicity Childhood Asthma Management Program (CAMP) study, the joint effect
5 of NO2 and CO (OR: 1.07 [95% CI: 1.00, 1.14] for a 20-ppb increase in lag 0-2 day sum
6 of 24-h avg NO2) was similar to the NO2 (OR: 1.05 [95% CI: 1.01, 1.09]) and CO
7 single-pollutant ORs (Schildcrout et al.. 2006). These results indicate a lack of
8 multiplicative interaction between NO2 and CO but do not inform potential confounding.
9 Although analysis is limited in both the number of studies and array of traffic-related
10 copollutants, there is evidence that NO2 has an association with respiratory symptoms in
11 children with asthma independent of EC or VOCs, but interactions are not demonstrated.
12 Interactions with NO2 were not found consistently for PMio and not found at all for SO2,
13 Os, or allergens (Schildcrout et al.. 2006; Delfino et al.. 2002). These copollutants were
14 not examined as potential confounding factors. An NO2-wheeze association was observe
15 to decrease in magnitude and precision (i.e., wider 95% CI) with adjustment for PMio-2.5
16 (r = 0.12) (Mannet al., 2010). based on exposures assessed from a site up to 20 km of
17 children's homes.
18 Other studies largely corroborate the aforementioned evidence but do not provide a
19 strong basis for assessing an independent effect of NO2 exposure on respiratory
20 symptoms in children with asthma because of both central site exposure assessment and
21 no examination of potential confounding by other traffic-related pollutants (Table 5-11).
22 In these multi- and single-city studies, central sites were located 2-16 km from children's
23 homes or schools (Patel etal.. 2010; Barraza-Villarreal et al.. 2008; Escamilla-Nunez et
24 al.. 2008; Romieu et al.. 2006; Jalaludin et al.. 2004; Ostro etal.. 2001). averaged across
25 city sites (O'Connor etal.. 2008; Just et al.. 2002; Mortimer et al.. 2002). or an
26 unspecified method (Segala et al.. 1998). For most locations, information was not
27 reported to assess whether the temporal variation in these metrics represented the
28 variation across the study areas. Low within-city correlations in NO2 were reported for
29 Mexico City (Barraza-Villarreal et al.. 2008; Escamilla-Nunez et al.. 2008). and high
30 correlations for NO2 were reported for Los Angeles/Pasadena (Ostro etal.. 2001)
31 (Section 2.5.3). The recent multicity ICAS found increases in symptoms, slow play, and
32 missed school in association with a 19-day avg of 24-h avg NO2 (O'Connor et al.. 2008).
33 but there is potential for residual temporal confounding for associations with NO2
34 exposure on the order of weeks. ICAS could not examine shorter lags because symptom
35 data were collected with a time resolution of 2 weeks.
36 In addition to NO2, most studies with central site exposure assessment found associations
37 with the traffic-related copollutants BC, black smoke (BS), PM2 5, and CO, which were
January 2015 5-76 DRAFT: Do Not Cite or Quote
-------
1 moderately to highly correlated with NO2 (r = 0.34-0.92) (Patel etal.. 2010; Barraza-
2 Villarreal et al.. 2008; Just et al.. 2002; Ostro etal.. 2001). A potential confounding effect
3 of these traffic-related pollutants was examined only in multipollutant models, which can
4 produce unreliable results because of potential collinearity (Kim etal.. 2012; Escamilla-
5 Nunez et al.. 2008; O'Connor et al.. 2008). Copollutant confounding was not examined
6 for PMio, 862, or Os either. These pollutants were moderately correlated with NC>2
7 (r = 0.28-0.31) in most studies, although some reported higher correlations
8 (r = 0.54-0.68) (Ostro et al.. 2001; Segalaetal.. 1998V
9 In addition to the limited findings from copollutant models with traffic-related
10 copollutants, an independent effect of NO2 exposure is supported by numerous studies
11 that show increases in respiratory symptoms in association with increases in indoor NC>2
12 averaged over 3 to 7 days or a 4-week avg (Luetal. 2013; Gillespie-Bennett et al.. 2011;
13 Hansel et al.. 2008). Previous findings indicated reductions in respiratory symptoms after
14 an intervention to switch to flued gas heaters led to a reduction in indoor classroom NCh
15 concentrations (Pilotto etal.. 2004). Although potential differences in pollutant mixtures
16 between the indoor and outdoor environments have not been well characterized, a recent
17 study found that correlations between NC>2 and copollutants differed between the indoor
18 and outdoor environments for BC, PM, and 862 (Sarnat et al.. 2012). suggesting that NC>2
19 may exist as part of a different pollutant mixture in the indoor and outdoor environments.
Adults with Asthma
20 Previous and recent evidence indicates associations of ambient NO2 concentrations with
21 respiratory symptoms (Maestrelli et al.. 2011; Wiwatanadate and Liwsrisakun. 2011; von
22 Klot et al.. 2002; Boezen etal.. 1998; Forsberg et al.. 1998) and asthma medication use or
23 sales (Carlsenet al.. 2012; Laurent et al.. 2009; von Klot et al.. 2002; Forsberg et al..
24 1998; Hiltermann et al.. 1998) among adults with asthma or bronchial
25 hyperresponsiveness. Most studies were conducted in Europe and recruited subjects
26 primarily from clinics, doctors' offices, and administrative databases. Subjects
27 represented a mix of asthma severity and prevalence of ICS use and atopy. A few studies
28 did not find associations with symptoms, including Kim etal. (2012). which analyzed
29 only a multipollutant model with SC>2, PMio, Os, and CO, whose results can be unstable.
30 Null results also were reported in studies with more reliable statistical analysis that were
31 conducted in four European countries (Karakatsani et al.. 2012) and one with adults with
32 asthma and allergy (Feo Brito et al.. 2007). Results from the latter study contrast those
33 from experimental studies showing NO2-induced allergic inflammation in humans with
34 asthma and animal models of allergic disease (Section 5.2.2.5). Across studies,
35 respiratory symptoms were associated with lag day 0 NO2. Medication use or sales were
36 associated more strongly with multiday averages of NO2 (i.e., lag 3-5 avg, 0-5 avg,
January 2015 5-77 DRAFT: Do Not Cite or Quote
-------
1 0-6 avg) than with single-day lags (Carlsen et al.. 2012; von Klot et al.. 2002;
2 Hiltermann et al.. 1998). and Carlsen etal. (2012) found a stronger association for
3 beta-agonist sales with 1-h max than 24-h avg NC>2.
4 Despite the consistency of evidence, there is limited basis for inferring NC>2 effects on
5 respiratory symptoms in adults with asthma. Most studies assigned NC>2 exposure from a
6 single central site monitor located in the community and did not report information to
7 assess whether the NCh metrics were representative of the temporal variability across the
8 study areas and of subjects' ambient exposures. With exception of Boezen et al. (1998).
9 studies found associations with the traffic-related pollutants CO, BS, PIVb 5, and UFP, and
10 few analyzed potential confounding. Confounding is an uncertainty also for the
11 association between beta-agonist sales and block-level NCh estimated with a dispersion
12 model (Laurent et al.. 2009). Block-level NO2 was highly correlated with measured
13 concentrations (r = 0.87), but other traffic-related pollutants were not examined. Only
14 von Klot et al. (2002) conducted copollutant modeling and found an association between
15 lag 0-4 day avg NO2 and beta-agonist use with adjustment for PM2 5 or UFP (OR: 1.22
16 [95% CI: 1.05, 1.43] per 20-pbb increase in NO2, with adjustment for UFP, Pearson
17 r = 0.66). The NO2-wheeze association was attenuated with adjustment for UFP (OR:
18 1.02 [95% CI: 0.86, 1.21]). Copollutant effect estimates were attenuated with NO2
19 adjustment. Thus, an independent NO2 association was found for medication use, but an
20 independent association with wheeze was not discerned for either NO2 or UFP.
Controlled Human Exposure Studies
21 Controlled human exposure studies do not provide strong evidence for NO2-induced
22 increases in respiratory symptoms in adults or adolescents with asthma. The majority of
23 these controlled human exposure studies that assessed respiratory symptoms before,
24 during, or after exposure to NO2 did not find changes. Unlike studies of airway
25 responsiveness (Section 5.2.2.1). symptom studies did not include a challenge agent with
26 NO2 exposure. One recent study is available in addition to the studies reviewed in the
27 2008 ISA for Oxides of Nitrogen (U.S. EPA. 2008a). Study details are presented in
28 Table 5-13; overall, studies subjected participants to NO2 exposures of 120-4,000 ppb for
29 2-5 hours and then conducted an assessment of symptoms 24 hours later.
30 The majority of studies reported no change in symptoms, as measured by symptom score,
31 in healthy subjects or in adults with asthma (Torres etal.. 1995; Linn et al.. 1985b;
32 Kleinman et al.. 1983) or in adolescents (Koenig et al.. 1987). Vagaggini et al. (1996)
33 reported a small, but statistically significant increase in symptom score during NO2
34 exposures in healthy adults, but not those with asthma. Riedletal. (2012) recently
35 reported an increase in symptom score in adults with asthma during, but not after,
January 2015 5-78 DRAFT: Do Not Cite or Quote
-------
1
2
o
6
4
5
6
exposure to 350 ppb NO2 for 2 hours with alternating periods of exercise. The increase in
symptom score corresponded to a subject experiencing a mild increase in any two
symptoms or moderate elevation of any one symptom. Symptom scores were not
different between air and NCh-exposed subjects when categorically grouped as
respiratory, cardiovascular, or miscellaneous; nor were they different when subjects were
exposed to allergen after NO2 exposure Riedletal. (2012).
Table 5-13 Controlled human exposure studies of respiratory symptoms.
Disease Status; n, Sex; Age
Study (mean ± SD)
Exposure Details
(Concentration; Duration)
Time of Symptom
Assessment
Jorres et al.
(1995)
Kleinman et al.
(1983)
Koeniq et al.
(1987)
Asthma; n = 8 M, 4 F; 27 ± 5 yr
Healthy; n = 5 M, 3 F;
27 yr (range: 21-33)
Asthma; n = 12 M, 19 F;
31 ± 1 yr
Asthma;
(1)n =4 M, 6 F
(2) n = 7 M, 3 F;
Healthy;
(1)n = 3M, 7F
(2) n = 4 M, 6 F;
14. 4 yr (range: 12-19)
1,000 ppb for 3 h;
Exercise 10 min on/10 min off at
individual's maximum workload
200 ppb for 2 h;
Exercise 15 min on/15 min off at
VE = ~2 times resting
(1)120 ppb NO2
(2) 180ppbNO2;
(1-2) Exposures were 30 min at
rest with 10 min of moderate
exercise
Immediately and 6 and
24 h after exposure.
Before, immediately
after, and day after
exposure.
Immediately after, a day
after, and a week after
exposure.
Linn et al.
(1985b)
Riedletal.
(2012)
Asthma; n = 12 M, 11 F;
(range: 18-34 yr)
Healthy; n = 16 M, 9 F;
(range: 20-36 yr)
Asthma
Pha<;p 1 fmptharhnlinp rhallpr
4,000 ppb for 75 min;
Two 15 min periods of exercise
at VE = 25 L/min and 50 L/min
350 ppb for 2 h;
inpV Fyprrkp 1F> min nn/1F> min off at
Before, during,
immediately after, 1 day
after and 1 week after
exposure.
Before, during, 1-22 h
after exposure.
n = 10M, 5F; 37.3 ± 10.9 yr
Phase 2 (cat allergen challenge);
n=6M, 9F; 36.1 ± 12.5 yr
VE = 15-20 L/min
Vaqaqqini et al.
(1996)
Asthma; n = 4 M, 4 F; 29 ± 14 yr
Healthy; n = 7 M; 34 ± 5 yr
300 ppb for 1 h;
Exercise at VE = 25 L/min
Before and 2 h after
exposure.
F = female, M = male, NO2 = nitrogen dioxide, SD = standard deviation.
January 2015
5-79
DRAFT: Do Not Cite or Quote
-------
5.2.2.4 Hospital Admissions and Emergency Department Visits
for Asthma
1 The evidence for NCh-related effects on increasing airway responsiveness, decreasing
2 lung function, and increasing respiratory symptoms detailed in the preceding sections are
3 all indicators of asthma exacerbation that may lead people with asthma to seek medical
4 interventions. Thus, the preceding evidence is coherent with associations observed
5 between short-term increases in ambient NC>2 concentrations and hospital admissions and
6 ED visits for asthma. Since the completion of the 2008 ISA for Oxides of Nitrogen,
7 epidemiologic studies have continued to examine the association between short-term
8 exposure to ambient NOx or NC>2 and respiratory-related hospital admissions and ED
9 visits, but are primarily limited to single-city studies. The sections within this chapter that
10 detail the studies of respiratory-related hospital admissions and ED visits characterize
11 recent studies in the context of the collective body of evidence evaluated in the 2008 ISA
12 for Oxides of Nitrogen. As summarized in Section 5.2.6. the 2008 ISA for Oxides of
13 Nitrogen (U.S. EPA. 2008a) included the first thorough evaluation of respiratory
14 morbidity in the form of respiratory-related hospital admissions and ED visits, including
15 asthma. Previous studies of asthma hospital admissions and ED visits consistently
16 reported positive associations with short-term NO2 exposures (Figure 5-7. Table 5-16)
17 and observed associations that were generally robust and independent of the effects of
18 ambient particles or gaseous copollutants (e.g., Os, SO2, CO, benzene) (U.S. EPA.
19 2008a). The strongest evidence for associations between short-term NO2 exposures and
20 asthma were from studies of all ages and children.
21 Within this section focusing on asthma, as well as the rest of the chapter, studies of
22 respiratory-related hospital admissions and ED visits are evaluated separately because
23 often only a small percentage of respiratory-related ED visits will be admitted to the
24 hospital. Therefore, ED visits may represent potentially less serious, but more common,
25 outcomes. The air quality characteristics of the city, or across all cities, and the exposure
26 assignment approach used in each asthma hospital admission and ED visit study
27 evaluated in this section are presented in Table 5-14. Other recent studies of asthma
28 hospital admissions and ED visits are not the focus of this evaluation because they were
29 conducted in small individual cities, encompass a short study duration, had insufficient
30 sample size, and/or did not examine potential copolluant confounding. The full list of
31 these studies, as well as study specific details, can be found in Supplemental Table S5-3
32 (U.S. EPA. 2014h).
January 2015 5-80 DRAFT: Do Not Cite or Quote
-------
Table 5-14 Mean and upper percentile concentrations of oxides of nitrogen in studies of asthma hospital
admissions and emergency department (ED) visits.
Study
Location
Years
Exposure Assignment Metric
Mean/Median
Concentration
ppb
Upper Percentile of
Concentrations (ppb)
Copollutant Examination
Hospital Admissions
(Linnetal., 2000)
Los Angeles, Average of NO2
CA concentrations over all
(1992-1995) monitors.
24-h avg 3.4
NR
Correlations (r):
Range across seasons
CO: 0.84-0.94
PMio: 0.67-0.88
O3: -0.23 to 0.35
Copollutant models: none
(Burnett etal., 1999)
Toronto, Average of NO2
Canada concentrations from 4
(1980-1994) monitors.
24-h avg 25.2
NR
Correlations (r):
PIVh.s: 0.55
PMio-2.s: 0.38
PMio: 0.57
CO: 0.64
SO2: 0.54
Os: -0.08
Copollutant models: none
tSamolietal. (2011)
Athens, Average of NO2
Greece concentrations across 14
(2001-2004) monitors.
1-h max 44.4
75th: 53.1
Correlations (r):
SO2: 0.55
Copollutant models: PMio, SO2, Os
tlskandar etal. (2012) Copenhaqen,
Denmark
(2001-2008)
All hospitals located 24-h avg NO2: 11.3
within 15 km of a central
site monitor.
NOx: 14.5
75th:
NO2: 14.2
NOx: 17.7
Correlations (r):
NOx: 0.93
PMio: 0.43
PIvh.s: 0.33
UFP: 0.51
Copollutant models: NOx, PlVh.s,
PMio, UFP
tKo et al. (2007b)
Hong Kong
(2000-2005)
Average of NO2
concentrations across
14 monitors.
24-h avg 28.3
75th: 33.8
Max: 79.5
Correlations (r):
SO2: 0.57
PMio: 0.76
Os: 0.41
PIvh.s: 0.77
Copollutant models:
January 2015
5-81
DRAFT: Do Not Cite or Quote
-------
Table 5-14 (Continued): Mean and upper percentile concentrations of oxides of nitrogen in studies of asthma
hospital admissions and emergency department (ED) visits.
tSon
Study
etal. (2013)
Location
Years
8 South
Exposure Assignment
Average of hourly
Metric
24-h avg
Mean/Median
Concentration
ppb
11.5-36.9
Upper Percentile of
Concentrations (ppb)
NR
Copollutant
Correlations (r):
Examination
Korean cities ambient NO2
(2003-2008) concentrations from
monitors in each city.
PMio: 0.5
O3: -0.1
SO2: 0.6
CO: 0.7
Copollutant models: none
ED Visits
Peel et al. (2005)
Atlanta, GA Average of NO2
(1993-2000) concentrations from
monitors for several
monitoring networks.
1-h max 45.9
NR
Correlations (r):
PlVh.s: 0.46
PMio: 0.49
PMio-2.s: 0.46
UFP: 0.26
PM2.5 water soluble Metals: 0.32
PM2.5 sulfate: 0.17
PM2.5 acidity: 0.10
PM2.5 OC: 0.63
PIvh.s EC: 0.61
Oxygenated HCs: 0.30
O3: 0.42
CO: 0.68
SO2: 0.34
Copollutant models: none
Tolbert et al. (2000)
Atlanta, GA Average of NO2
(1993-2004) concentrations from
monitors for several
monitoring networks.
1-h max 81.7
Max: 306
Correlations (r):
PM2.s: 0.47
PMio: 0.53
PMio-2.s: 0.4
PM2.s sulfate: 0.14
PM2.s OC: 0.62
PM2.5 EC: 0.64
PM2.5 TC: 0.65
PM2.5 water soluble Metals: 0.32
Oxygenated HCs: 0.24
O3: 0.44
CO: 0.70
SO2: 0.36
Copollutant models: CO, PMio, Os
January 2015
5-82
DRAFT: Do Not Cite or Quote
-------
Table 5-14 (Continued): Mean and upper percentile concentrations of oxides of nitrogen in studies of asthma
hospital admissions and emergency department (ED) visits.
Study
Location
Years
Exposure Assignment Metric
Mean/Median
Concentration
ppb
Upper Percentile of
Concentrations (ppb)
Copollutant Examination
Jaffe et al. (2003)
2 Ohio cities
(Cincinnati
and
Cleveland)
(1991-1996)
When more than 1 NO2
monitor operating in a
day, monitor with highest
24-h avg concentration
used.
24-h avg
Cincinnati: 50
Cleveland: 48
NR
Correlations (r):
Cincinnati
PMio: 0.36
Os: 0.60
SO2: 0.07
Cleveland
PMio: 0.34
SO2: 0.28
Os: 0.42
Copollutant models: none
Ito et al. (2007b)
New York, Average of NO2
NY concentrations from
(1999-2002) 15 monitors.
24-h avg 31.1
NR
Correlations (r): NR
Copollutant models: PlVh.s, Os, SO2,
CO
ATSDR (2006)
Bronx and NO2 concentrations from
Manhattan, N 1 monitor in Bronx and 1
(1999-2000) in Manhattan.
24-h avg Bronx: 36 NR
Manhattan: 31
Correlations (r):
Bronx
Os: 0.03
SO2: 0.47
FRM PM2.5: 0.61
Max PM2.s: 0.55
Manhattan: NR
Copollutant models: Os, FRM and
Max PM2.5, SO2
tStrickland et al.
(2010)
tSarnatetal. (201 3a)
fVilleneuve et al.
(2007)
Atlanta, GA
(1993-2004)
Atlanta, GA
(1999-2002)
Edmonton,
Canada
(1992-2002)
Combined daily NO2
concentrations across
monitors using
population weighting.
NOx concentrations
predicted using fused
spatially interpolated
background pollutant
concentrations and
local-scale AERMOD
output for 186 zip codes.
Average of NO2
concentrations across 3
monitoring stations.
1-h max 23.3
24-h avg NOx: 30.1
24-h avg 50th: 17.5
Summer
50th: 28.5 Winter
NR
75th:
95th:
Max:
75th:
75th:
40.1
94.4
517.8
22
35,
.0 summer
.5 winter
Correlations
Copollutant
(r): NR
models:
Os
Correlation (r) with NOx:
CO: 0.93
Os: -0.03
PM2.s: 0.40
Copollutant models: none
Correlations
CO: 0.74
Copollutant
(r):
models:
CO
January 2015
5-83
DRAFT: Do Not Cite or Quote
-------
Table 5-14 (Continued): Mean and upper percentile concentrations of oxides of nitrogen in studies of asthma
hospital admissions and emergency department (ED) visits.
Study
fJalaludin et al.
(2008)
tStieb et al. (2009)
tOrazzo et al. (2009)
fStrickland et al.
(2011)
Location
Years
Sydney,
Australia
(1997-2001)
7 Canadian
cities
(1992-2003)
6 Italian
cities
(1996-2002)
Atlanta, GA
(1993-2004)
Exposure Assignment Metric
Average of NO2 1-h max
concentrations across
14 monitoring stations.
Average NO2 24-h avg
concentrations from all
monitors in each city.
Number of NO2 monitors
in each city ranged from
1-14.
Average of NO2 24-h avg
concentrations from all
monitors in each city.
NO2 concentrations 1-h max
obtained from 3 networks
of stationary monitors.
Each air pollutant
measured by at least 3
monitoring stations.
3 exposure metrics used:
(1) 1 downtown monitor
was selected to be the
central site monitor, (2) all
monitors used to calculate
unweighted average of
pollutant concentrations
for all monitors, and
(3) population-weighted
average concentration.
Mean/Median
Concentration
ppb
23.2
9.3-22.7
21.4-41.2
Central monitor:
42.0
Unweighted
average: 27.7
Population-
weighted average:
22.0
Upper Percentile of Copollutant Examination
Concentrations (ppb)
Max: 59.4 Correlations (r):
PMio: 0.67
PlVhs: 0.68
03: 0.21
CO: 0.71
SO2. 0.52
Copollutant models: PMio, PlVh.s, Os,
CO, SO2
75th: 12.3-27.6 Correlations (r) only reported by city
and season.
Copollutant models: none
NR Correlations (r): NR
Copollutant models: none
NR Correlations (r): NR
Copollutant models: none
January 2015
5-84
DRAFT: Do Not Cite or Quote
-------
Table 5-14 (Continued): Mean and upper percentile concentrations of oxides of nitrogen in studies of asthma
hospital admissions and emergency department (ED) visits.
Study Location Exposure Assignment Metric Mean/Median
Years Concentration
ppb
tLietal. (201 1b) Detroit, Ml Averaae of NO2 24-h ava 15.7
(2004-2006) concentrations from 2
monitors in Detroit
metropolitan area that
measure NO2.
tGassetal. (2014) Atlanta, GA Population-weiqhted 24-h avq NR
(1999-2009) average NO2
concentrations based on
same methods as
(Strickland etal., 2010).
tWinquist et al. (2014) Atlanta, GA Population-weighted 1-hr max Warm (May-Oct):
(1998-2004) average of NO2 22.3
concentrations. Cold (Nov-April):
25.6
Upper Percentile of Copollutant Examination
Concentrations (ppb)
75th: 21.2 Correlations (r), range across
Max: 55.2 monitors:
CO: 0.37-0.40
PM2.s: 0.56-0.66
SO2: 0.42-0.55
Copollutant models: none
NR Correlations (r): NR
Copollutant models: none
75th: Correlations (r):
Warm: 28.7 Warm:
Cold: 31. 7 O3: 0.54
CO: 0.75
SO2: 0.44
PM2.5: 0.52
EC: 0.68
Sulfate: 0.27
Secondary PM2.s: 0.31
Cold:
O3: 0.30
CO: 0.74
SO2: 0.41
PMzs: 0.49
EC: 0.57
Sulfate: 0.08
Secondary PM2.s: 0.12
January 2015
5-85
DRAFT: Do Not Cite or Quote
-------
Table 5-14 (Continued): Mean and upper percentile concentrations of oxides of nitrogen in studies of asthma
hospital admissions and emergency department (ED) visits.
Study
Location
Years
Exposure Assignment
Metric Mean/Median
Concentration
ppb
Upper Percentile of
Concentrations (ppb)
Copollutant Examination
Physician Visits
tBurra et al. (2009)
tSinclairetal. (2010)
Toronto,
Canada
(1992-2001)
Atlanta, GA
(1998-2002)
Average of NO2
concentrations from 6
monitors.
NO2 concentrations
collected as part of
AIRES at SEARCH
Jefferson street site.
1-h max 39.2
1-hmax 1998-2000:49.8
2000-2002: 35.5
1998-2002:41.7
95th: 60
Max: 105
NR
Correlations (r): NR
Copollutant models: none
Correlations (r): NR
Copollutant models: none
NR = not reported, AIRES = Aerosol Research Inhalation Epidemiology Study, CO = carbon monoxide, EC = elemental carbon, ED = emergency department, NO2 = nitrogen dioxide,
NOX = sum of NO and NO2,03 = ozone, OC = organic carbon, PM = particulate matter, SO2 = sulfur dioxide, UFP = ultrafine particles, SEARCH = Southeastern Aerosol Research
and Characterization.
fStudies published since the 2008 ISA for Oxides of Nitrogen.
January 2015
5-86
DRAFT: Do Not Cite or Quote
-------
Hospital Admissions
1 Generally, studies evaluated in the 2008 ISA for Oxides of Nitrogen that examined the
2 effect of short-term NO2 exposures on asthma hospital admissions were limited to single
3 cities. It is important to note the results of these studies should be viewed with caution
4 because they tended to include ages <5 years in the study population, which is
5 problematic considering the difficulty in reliably diagnosing asthma within this age range
6 [National Asthma Education and Prevention Program Expert (NAEPP. 2007)]. However,
7 it is unlikely the inclusion of these individuals in a study would introduce a systematic
8 positive bias. In contrast, the majority of studies on asthma ED visits (discussed in the
9 next section) have excluded ages <2 years in analyses to account for this difficulty.
10 In a time-series study conducted in Athens, Greece, Samoli et al. (2011) evaluated the
11 association of multiple ambient air pollutants and pediatric asthma hospital admissions
12 for ages 0-14 years. In an all-year analysis, the authors reported a positive association
13 with NO2 (6.4 % [95% CI: -3.8, 17.6]; lag 0 increase for a 30-ppb increase in 1-h max
14 NO2 concentrations). An examination of additional lags (quantitative results not
15 presented) revealed a similar pattern of associations at lag 2 and a 0-2 days distributed
16 lag. In copollutant analyses, NO2 risk estimates were robust when Os (7.6% [95% CI:
17 ~2.7, 19.0]) was included in the model, and were attenuated but remained positive with
18 wide confidence intervals when including PMio in the model (3.1% [95% CI: -7.3,
19 14.6]). There was evidence of confounding of the NO2 association when SO2 was
20 included in the model as demonstrated by an effect estimate and confidence interval for
21 NO2 reflective of a null association (-4.3% [95% CI: -16.9, 10.2]). Of the two
22 copollutants examined, SO2 was most highly correlated with NO2 (r = 0.55).
23 The association between short-term air pollution exposures and asthma hospital
24 admissions in children (0-18 years of age) was also examined in a study conducted by
25 Iskandar et al. (2012) in Copenhagen, Denmark. In a time-stratified case-crossover
26 analysis using an a priori lag of 0-4 days, the authors reported positive associations for
27 both NO2 (OR: 1.3 [95% CI: 1.1, 1.6] for a 20-ppb increase in 24-h avg NO2
28 concentrations) and NOx (OR: 1.6 [95% CI: 1.3,2.1] for a 40-ppb increase in 24-h avg
29 NOx concentrations), which are larger in magnitude than those observed in Samoli et al.
30 (2011). Within this study NOx and NO2 were highly correlated (r = 0.93). Correlations
31 for NOx and NO2 with PM2 5 and UFPs ranged from, r = 0.28-0.33 and r = 0.45-0.51,
32 respectively. The high correlation between NOx and NO2, and the fact NO2 is part of
33 NOx, suggests that these pollutants should not be included in the same model due to the
34 inability to clearly examine whether one pollutant has an independent effect compared to
January 2015 5-87 DRAFT: Do Not Cite or Quote
-------
1 the other. In additional copollutant models, NO2 and NOx associations remained
2 relatively unchanged in models with PM2 5 and UFP Table 5-15).
Table 5-15 Copollutant model results from Iskandar et al. (2012) for a 20-ppb
increase in 24-h avg nitrogen dioxide (NO2) concentrations and a
40-ppb increase in 24-h avg NOx (sum of NO and NO2)
concentrations.
Pollutant
NOx
NO2
Copollutant
NO2
PMio
PM2.5
UFP
NOx
PMio
PM2.5
UFP
Odds
1.6(1.3,2.1)
1.7(0.8, 3.5)
1.4(1.1, 1.8)
1.6(1.2,2.1)
1.6(1.2,2.2)
1.3(1.1, 1.6)
1.0(0.6, 1.6)
1.3(1.0, 1.5)
1.4(1.2, 1.7)
1.5(1.2, 1.8)
Ratio (95% Cl)
Cl = confidence interval, NO2 = nitrogen dioxide, NOx = sum of NO and NO2, PM = participate matter, UFP = ultrafine particles.
4 Ko et al. (2007b) examined the association between short-term air pollution exposures
5 and asthma hospital admissions for all ages at both single- and multiday lags in Hong
6 Kong. In a time-series analysis the authors reported positive associations at single-day
7 lags that were smaller in magnitude than those observed in Samolietal. (2011) [e.g.,
8 3.4% (95% Cl: 1.9, 5.4%); lag 0 for a 20-ppb increase in 24-h avg NO2 concentrations].
9 However, the results of Ko et al. (2007b) are consistent with those of Son et al. (2013) in
10 eight South Korean cities, which found the strongest association at lag 0 between
11 short-term NO2 exposures and asthma as well as allergic disease hospital admissions,
12 which encompasses asthma (3.6% [95% Cl: 0.5, 6.8] and 3.8% [95% Cl: 1.0, 6.6],
13 respectively for a 20-ppb increase in 24-h avg NO2 concentrations). However, unlike
14 Samolietal. (2011) and Son et al. (2013). Ko et al. (2007b) found the strongest evidence
15 of an association between short-term NO2 exposures and asthma hospital admissions at
16 multiday lags of 0-3 (10.9% [95% Cl: 8.1, 13.8] and 0-4 (10.9% [95% Cl: 8.1, 13.4])
17 days. In a copollutant analysis with Os, the authors reported evidence of a reduction in
18 NO2 risk estimates although they remained positive (2.3% [95% Cl: -0.8, 5.8]; lag
19 0-4 days), which is not consistent with the results of the copollutant analysis in Samoli et
20 al. (2011). This attenuation occurred even though NO2 and Os were not well correlated
21 (r = 0.41) in Hong Kong.
January 2015 5-88 DRAFT: Do Not Cite or Quote
-------
Emergency Department Visits
1 Similar to the asthma hospital admission studies evaluated in the 2008 ISA for Oxides of
2 Nitrogen, the majority of ED visit studies were limited to single-city studies. However,
3 these studies provided additional information with regard to potential seasonal
4 differences in risk estimates, indicating some evidence of larger associations during
5 warmer months.
6 Strickland et al. (2010) examined the association between NCh exposure and pediatric
7 asthma ED visits (ages 5-17 years) in Atlanta, GA, using air quality data over the same
8 years as two studies that focused on total respiratory ED visits, Darrow et al. (2011) and
9 Tolbert et al. (2007) (Section 5.2.6). However, unlike Darrow etal. (2011) and Tolbert et
10 al. (2007). which used a single-site centrally located monitor and the average of multiple
11 monitors to assign exposure, respectively, Strickland et al. (2010) used population
12 weighting to combine daily pollutant concentrations across monitors. In this study, the
13 authors developed a statistical model using hospital-specific time-series data that is
14 essentially equivalent to a time-stratified case-crossover analysis (i.e., using interaction
15 terms between year, month, and day-of-week to mimic the approach of selecting referent
16 days within the same month and year as the case day). Strickland et al. (2010) reported an
17 8.6% (95% CI: 4.2, 13.3) increase in ED visits for a 30-ppb increase in 1-h max NC>2
18 concentrations at lag 0-2 days in an all-year analysis. The potential confounding effects
19 of other pollutants on the NCh-asthma ED visit relationship was only examined in a
20 copollutant model with Os and correlations between pollutants were not presented. In the
21 copollutant model, NCh risk estimates were found to be relatively unchanged upon the
22 inclusion of Os (quantitative results not presented).
23 The magnitude of the association between short-term NCh concentrations and asthma ED
24 visits observed in Strickland et al. (2010) is larger than that observed in Sarnat et al.
25 (2013a) in a study also conducted in Atlanta, GA, which focused on the influence of air
26 exchange rates on air pollution-asthma ED visit associations detailed in Chapter 3 and
27 Chapter 7. Instead of using monitored NC>2 concentrations, the authors used NOx
28 concentrations estimated by "fus(ing) spatially interpolated background pollutant
29 concentrations and the local-scale air quality model AERMOD output for the 186 ZIP
30 code centroids" in the Atlanta metro area. Also, focusing on a lag of 0-2 days, Sarnat et
31 al. (2013a) reported a 1.3% increase in asthma ED visits (95% CI: 0.0, 2.4) for a 40-ppb
32 increase in 24-h avg NOx concentrations. The authors did not examine copollutant
33 models, but NOx was found to be highly correlated with CO (r = 0.93). The magnitude of
34 the association between Strickland et al. (2010) and Sarnat etal. (2013a) differs, which
35 could be a reflection of: (1) exposure measurement error and differences in exposure
36 assessment methods for NOx compared to NO2 and (2) the different age ranges included
January 2015 5-89 DRAFT: Do Not Cite or Quote
-------
1 in both studies, which is supported by the earlier studies focusing on all ages conducted
2 in Atlanta by Tolbert et al. (2000) and Peel et al. (2005) that report associations similar in
3 magnitude to that observed in Sarnatet al. (2013a) (Figure 5-7).
4 Additional evidence for an association between short-term increases in NO2
5 concentrations and asthma ED visits comes from studies conducted in Edmonton, Canada
6 (Villeneuve etal., 2007) and Sydney, Australia (Jalaludin et al., 2008). Villeneuve et al.
7 (2007) reported evidence of positive associations between short-term NO2 concentrations
8 and asthma ED visits for multiple lag structures (lag 1, lag 0-2, and lag 0-4 days) in the
9 population aged 2 years and older. The authors observed the strongest association for
10 lag 0-4 days (4.5% [95% CI: 0, 7.5] for a 20-ppb increase in 24-h avg NO2
11 concentrations). There was no evidence of an association at lag 0. In this study, NO2 and
12 CO were strongly correlated (r = 0.74), and as a result associations were examined in
13 copollutant models for each age group examined in the study, focusing on the warm
14 season (April-September). In copollutant models with CO, NO2 associations with asthma
15 ED visits were relatively similar to single-pollutant results except for one age group,
16 15-44 years, but in all instances NO2 associations were larger in magnitude than those
17 for CO (quantitative results not provided).
18 In a study focusing on children 1-14 years old, Jalaludin et al. (2008) examined air
19 pollution associations with asthma ED visits for single day lags up to 3 days as well as
20 the average of 0-1 day lags. Jalaludin et al. (2008) observed a similar magnitude of an
21 association for both lag 0 (7.5% [95% CI: 4.5, 10.5]) and lag 0-1 days (7.8% [95% CI:
22 4.5, 11.1] for a 30-ppb increase in 1-h max NO2 concentrations). An examination of the
23 potential confounding effects of other pollutants was assessed in copollutant models with
24 PMio, PM2 5, Os, CO, or SO2. NO2 was moderately to weakly correlated with each of these
25 pollutants (r ranging from 0.44-0.56). In copollutant models, the NO2-asthma ED visit
26 association remained positive, but was slightly attenuated with the magnitude of the
27 association ranging from a 4.2-6.1% increase in asthma ED visits. In addition to
28 analyzing ages 1-14 years, the authors examined whether risks varied among age ranges
29 within this study population (see Chapter 7).
30 In contrast with the majority of the evidence, short-term increases in NO2 concentrations
31 were not associated with asthma ED visits in a multicity study conducted in seven
32 Canadian cities (Stieb etal.. 2009). Compared to the other asthma ED visit studies
33 evaluated, mean NO2 concentrations across the cities included in this study were the
34 lowest with all cities having mean 24-h avg concentrations <23 ppb (Table 5-14). Stieb
35 et al. (2009) examined the association between short-term NO2 exposure and a number of
36 respiratory-related ED visits for all ages. There was no evidence that NO2 was associated
37 with asthma ED visits at single-day lags of 0 to 2 days (0.0% [95% CI: -2.6, 2.7]; lag 2
January 2015 5-90 DRAFT: Do Not Cite or Quote
-------
1 for a 20-ppb increase in 24-h avg NC>2 concentrations). Additionally, there was no
2 evidence of associations between respiratory-related ED visits, including asthma, and air
3 pollution averaged over-sub-daily time scales (i.e., 3-h avg of ED visits vs. 3-h avg
4 pollutant concentrations).
Emergency Department Visits for Wheeze
5 As stated previously [National Asthma Education and Prevention Program Expert
6 (NAEPP. 2007)1. asthma is difficult to diagnose in children less than 5 years of age;
7 however, asthma-like symptoms in children within this age range are often presented in
8 the form of transient wheeze. Although studies that examine ED visits for wheeze do not
9 directly inform upon the relationship between short-term NCh exposures and asthma, they
10 can add supporting evidence. Also, it should be noted that some studies that examine
11 asthma ED visits, as well as hospital admissions, often include International
12 Classification of Diseases (ICD) codes for wheeze in the definition of asthma (e.g.,
13 (Sarnat et al.. 2013a)). Orazzo et al. (2009) examined the association between NO2 and
14 wheeze ED visits in children, (ages 0-2 years) in six Italian cities. Daily counts of
15 wheeze were examined in relation to air pollution using a time-stratified case-crossover
16 approach in which control days were matched on day of week in the same month and
17 year as the case day. PMio, SCh, CO, and Os were also evaluated, but correlations with
18 NC>2 were not reported nor were copollutant analyses conducted. The authors reported
19 positive associations between short-term 24-h avg NC>2 exposures and wheeze ED visits
20 when examining various multiday lags (0-1 through 0-6 days) with risk estimates
21 ranging from 1.1% (95% CI: -1.2, 3.4) for lag 0-1 days to 2.5% (95% CI: -0.9, 6.0) for
22 lag 0-6 days.
Outpatient and Physician Visit Studies
23 Several recent studies examined the association between ambient NO2 concentrations and
24 less severe asthma exacerbation, which are often encountered through physician or
25 outpatient (non-hospital, non-ED) visits. Burra et al. (2009) examined asthma physician
26 visits among patients aged 1-17 and 18-64 years focusing on differences by sex and
27 income within age categories in Toronto, Canada. The authors reported evidence of
28 consistently positive associations between short-term increases in NCh concentrations
29 and asthma physician visits across the single- and multi-day lags examined (i.e., 0, 0-1,
30 0-2, 0-3, and 0-4 days). The magnitude of the effect estimates were found to be similar
31 between sexes, income quintiles, and both within and between ages. In a study conducted
32 in Atlanta, GA, Sinclair etal. (2010) examined the association between air pollution and
33 a number of respiratory-related outpatient visits from a managed care organization,
January 2015 5-91 DRAFT: Do Not Cite or Quote
-------
1 including asthma. The authors separated the analysis into two time periods to compare
2 the air pollutant concentrations and relationships for acute respiratory visits for the
3 25-month time-period examined in Sinclair and Tolsma (2004) (i.e., August
4 1998-August 2000) and an additional 28-month time-period of available data from the
5 Atlanta Aerosol Research Inhalation Epidemiology Study (AIRES) (i.e., September
6 2000-December 2002). Across the two time periods, mean 1-h max NC>2 concentrations
7 were lower in the 28-month versus the 25-month time period, 35.5 versus 49.8 ppb,
8 respectively (Table 5-14). A comparison of the two time periods indicated that risk
9 estimates across outcomes tended to be larger in the earlier 25-month period compared to
10 the later 28-month period, with evidence of consistently positive associations at lags of
11 0-2 and 3-5 days for asthma, but confidence intervals were relatively large.
Examination of Seasonal Differences
12 In addition to examining the association between short-term NC>2 concentrations and
13 asthma hospital admissions and ED visits in all-year analyses, some studies also
14 conducted seasonal analyses. Overall, these studies generally provide evidence of larger
15 associations in the warm or summer season compared to cooler months (Figure 5-7).
16 However, it should be noted that these studies did not examine potential copollutant
17 confounding by season, which could further explain the results reported across studies.
18 In the study of eight South Korean cities, Sonetal. (2013) examined potential seasonal
19 differences across respiratory hospital admission outcomes, including asthma and allergic
20 disease. For both outcomes, the association with NO2 was largest in magnitude during the
21 summer (asthma: 16.2% [95% CI: 5.1, 28.6], lag 0; allergic disease: 15.9 [95% CI: 4.6,
22 28.5], lag 0 for a 20-ppb increase in 24-h avg NO2 concentrations) despite the lowest NO2
23 concentrations during the summer season (<20 ppb compared to >24 ppb in the other
24 seasons) across the eight cities. However, when using the warm season as the referent in
25 Hong Kong, Ko et al. (2007b) reported evidence of larger effects in the winter (i.e.,
26 December to March), suggesting that differences in seasonal associations may vary by
27 geographic location. The difference in seasonal associations by geographic location is
28 further highlighted in a study by Samoli et al. (2011) conducted in Athens, Greece that
29 reported results consistent with Sonetal. (2013). Although risk estimates for asthma
30 hospital admissions were relatively consistent across winter, spring, and autumn, ranging
31 from a 13.1 to a 13.8% increase per 20-ppb increase in 24-h avg NO2, the largest
32 percentage increase was observed forthe summer (28.7% [95% CI: -3.4, 71.3]).
33 The asthma ED visit studies that conducted seasonal analyses also reported seasonal
34 patterns similar to those observed in the hospital-admission studies. Villeneuve et al.
35 (2007) reported associations to be generally stronger in the warm season (e.g., 21.4%
January 2015 5-92 DRAFT: Do Not Cite or Quote
-------
1 [95% CI: 13.6, 31.0] at lag 0-4 days for a 20-ppb increase in 24-h avg NO2
2 concentrations) than in the cold season (-2.9% [95% CI: -7.3, 1.5]) in Edmonton,
3 Canada. Additionally, Jalaludin et al. (2008) found evidence of larger effects during the
4 warm months (November-April) compared to the cold months (May-October) in
5 Sydney, Australia (Figure 5-7). These results are consistent with Strickland et al. (2010).
6 which reported stronger associations during the warm season (i.e., May-October) (16.0%
7 [95% CI: 9.1, 23.5]; lag 0-2 days) than the cold season (3.8% [95% CI: -1.9, 9.6]; lag
8 0-2 days) in a study of pediatric asthma ED visits in Atlanta, GA. Additional support for
9 these seasonal differences in associations was presented by Orazzo et al. (2009). who
10 focused on wheeze ED visits in six Italian cities, where associations were slightly larger
11 in the summer compared to the winter, but the confidence intervals were wide and
12 overlapping (quantitative results not provided). In the study of seven Canadian cities,
13 Stieb etal. (2009) also conducted seasonal analyses, but did not present detailed results.
14 However, the authors did state that there was no evidence of consistent associations
15 during the winter months (October-March) between any pollutant and respiratory
16 outcomes, including asthma.
17 Additional evidence for potential seasonal differences in NO2-associations with asthma
18 hospital admissions and ED visits comes from the analysis of asthma physician visits by
19 Sinclair et al. (2010). When focusing on asthma in children, the authors reported larger
20 risk estimates in the warm season at all lags for the 25-month period (e.g., warm: 9.6%
21 [95% CI: -7.4, 30.0]; cold: 1.2% [95% CI: -12.4, 16.8] at lag 0-2 days for a 30-ppb
22 increase in 1-h max NO2 concentrations), with less consistent evidence for seasonal
23 differences in the 28-month period.
Concentration-Response Relationship
24 To date, few studies have examined the concentration-response (C-R) relationship
25 between NO2 exposures and respiratory morbidity. In recent studies, Strickland et al.
26 (2010) and Li etal. (20 lib) examined the shape of the NO2-pediatric asthma ED visit
27 relationship using different analytical approaches. Strickland et al. (2010) examined the
28 C-R relationship by conducting quintile and locally weighted scatterplot smoothing
29 (LOESS) C-R analyses. In the quintile analysis, NO2 associations were positive and
30 stronger at quintiles representing higher concentrations, ranging from 28 ppb to
31 >181 ppb, relative to the first quintile (i.e., NO2 concentrations <28 ppb). Additionally,
32 the LOESS C-R relationship analysis provides evidence indicating elevated NO2
33 associations along the distribution of concentrations from the 5th to 95th percentile
34 (Figure 5-5). Collectively, these analyses do not provide evidence of a threshold.
January 2015 5-93 DRAFT: Do Not Cite or Quote
-------
Nitrogen Dioxide Warm Season
15 20 25 30
Concentration (ppb)
Source: Reprinted with permission of the American Thoracic Society (Strickland et al.. 2010).
Figure 5-5 Locally weighted scatterplot smoothing concentration-response
estimates (solid line) and twice-standard error estimates (dashed
lines) from generalized additive models for associations between
3-day avg (lag 0-2) nitrogen dioxide (NO2) concentrations and
emergency department visits for pediatric asthma at the 5th to
95th percentile of NO2 concentrations in the Atlanta, GAarea.
In a study conducted in Detroit, MI, Li et al. (20lib) focused on the C-R relationship by
examining if there is evidence of a deviation from linearity. Associations were examined
in both a time-series and time-stratified case-crossover study design assuming: (1) no
January 2015
5-94
DRAFT: Do Not Cite or Quote
-------
1 deviation from linearity and (2) a change in linearity at 23 ppb [i.e., the maximum
2 likelihood estimate within the 10th to 95th percentile concentration where a change in
3 linearity may occur (~80th percentile)]. The analysis assumed a deviation in linearity but
4 did not assume zero risk below the inflection point. The focus of the analysis was on
5 identifying whether risk increased above the risk observed in the linear models at NC>2
6 concentrations above 23 ppb. In the analyses assuming linearity, effect estimates varied
7 across models for a 0-4-day lag (time series: 2.9% [95% CI: -7.9, 15.1]; case-crossover:
8 9.1% [95% CI: -0.83, 20.2] for a 20-ppb increase in 24-h avg NO2 concentrations). In the
9 models that assumed a deviation from linearity, the authors did not observe evidence of
10 higher risk in either the time-series or case-crossover analyses at NO2 concentrations
11 greater than 23 ppb.
Exposure Assignment
12 Questions often arise in air pollution epidemiologic studies with regard to the method
13 used to assign exposure. Strickland et al. (2011) assessed this question in a study
14 conducted in Atlanta, GA focusing on pediatric asthma ED visits. Using data from the
15 warm season from a previous analysis (Strickland et al.. 2010). Strickland et al. (2011)
16 examined the relative influence of different exposure assignment approaches (i.e., central
17 monitor, unweighted average across available monitors, and population-weighted
18 average) on the magnitude and direction of associations between NO2 and pediatric
19 asthma hospital admission. Strickland et al. (2011) reported that effect estimates per IQR
20 increase in NC>2 were similar across the metrics; however, based on a standardized
21 increment, the magnitude of the association between NO2 and pediatric asthma ED visits
22 varied (central monitor: 7.9% [95 % CI: 4.2, 11.8] < unweighted average: 12.1 % [95 % CI:
23 6.7, 17.9] < population-weighted average: 16.2% [95% CI: 9.1, 23.7] for a 30-ppb
24 increase in 1-h max NCh concentrations at lag 0-2 days). Although Strickland et al.
25 (2011) represents one study in one location, the results suggest that the different
26 approaches used to assign exposure across the studies evaluated may alter the magnitude,
27 but not direction, of the associations observed.
Nitrogen Dioxide within the Multipollutant Mixture
28 An important question often encountered during the review of any criteria air pollutant, is
29 whether the pollutant has an independent effect on human health. In the case of NO2, this
30 is questioned because it is often found to be highly correlated with other traffic-related
31 pollutants. However, ambient exposures to criteria air pollutants are in the form of
32 mixtures, which make answering this question difficult and primarily limited to
January 2015 5-95 DRAFT: Do Not Cite or Quote
-------
1 examining copollutant models. Recent studies conducted by Gass etal. (2014) and
2 Winquist et al. (2014). both of which use pediatric asthma ED visit data from Atlanta, use
3 novel approaches to assess whether specific mixtures are more strongly associated with
4 health effects compared to others. Although the primary objective of these types of
5 studies is not to directly assess the independent effects of a pollutant, they can inform the
6 role of NO2 in the air pollution mixture.
7 Gass etal. (2014) used a classification and regression tree (C&RT) approach to examine
8 the association between short-term exposures to unique daily multipollutant mixtures of
9 NO2, CO, PM25, and Os, and pediatric (i.e., ages 2-18 years) asthma ED visits in Atlanta.
10 C&RT is a supervised learning approach that creates various groupings of pollutants
11 based on an outcome variable, which differs from similar techniques, such as principal
12 component analysis, that do not consider the outcome (Gass etal.. 2014). For this
13 approach, daily pollutant concentrations were divided into quartiles with the referent
14 group comprised of all days in which each pollutant was in the lowest quartile. The
15 C&RT analysis identified 13 different unique daily pollutant combinations or terminal
16 nodes. Similar to Strickland et al. (2010). Gass etal. (2014) examined the relationship
17 between each combination and pediatric asthma ED visits using a Poisson model in the
18 context of a time-referent case-crossover analysis. Of the 13 unique combinations, 5 of
19 the largest relative risks (RRs) (i.e., RR ranging from 1.05 to 1.08) were observed for
20 combinations where NO2 concentrations were in the 3rd or 4th quartile. Of note for three
21 of the five combinations with the largest RRs, PIVb 5 concentrations were also high, with
22 concentrations in the 4th quartile. However, the RR largest in magnitude was observed
23 for a combination where NO2 concentrations were low (1st and 2nd quartiles) and PM2 5
24 concentrations were high (4th quartile). Overall, these results suggest that high daily
25 concentrations of NO2 alone and in combination with high daily concentrations of PlVfcs
26 can impact respiratory morbidity.
27 Winquist et al. (2014) took a different approach to examining multipollutant mixtures by
28 focusing on the joint effect (i.e., the combined effect of multiple pollutants) of pollutants
29 often associated with specific air pollution sources. Associations between short-term NO2
30 exposures and pediatric asthma ED visits (i.e., ages 5-17) were examined in
31 single-pollutant models and also in a multipollutant context in joint models for pollutant
32 combinations representative of oxidant gases (i.e., Os, NO2, 802), traffic (i.e., CO, NO2,
33 EC), and criteria pollutants (i.e., Os, CO, NO2, SO2, PIVb 5). Using the model detailed in
34 Strickland et al. (2010). the authors reported results for an IQR increase for lag 0-2 days
35 in single-pollutant analyses as well as three types of joint effect models [i.e., no
36 interaction terms (primary), first-order multiplicative interactions between pollutants
37 (interactions), and nonlinear pollutant terms (nonlinear)] (Figure 5-6).
January 2015 5-96 DRAFT: Do Not Cite or Quote
-------
1.35
1.30
1.25
1.20
.2 1-15
n
06 1.10
01
4-t
n
oe i.os
1 ("1/1
i.Uv
0.95
0.90
0.85
i i
<
[ j
'
i _
! '
(
1j
* j
'
>
-
']
- ?"
i
'j
J
»
'!
-
>
§5 i
j
i
[ .
*
i
1
i S
' S
i
•t
'*!
j
i
h
i-
1
J
'i
«".
l
i
'
t
i;
1
'
i
'
•
i
''
*
[i
^'
ijilli . I tlj.i"
1 ttfj 't'"
8 1 a 1 1 i 1 5 1 i 1 8 8 i i i 1 1 1
f I • ** f I • E*B«
I* J* i*
Oxtdant
Gases
Cornbirt&tiOi
Tfatfic
Warm Season
Power Plant Criteria Po)i«Unts
Oxtddnt
Gases
SWE K
Secondary
SM J« Mf K
Power Plant Criteria Pollutants
Cold Season
Source: Winguist et al. (2014).
SPE = single-pollutant model estimate; JE = joint model estimate.
Figure 5-6 Rate ratio and 95% confidence intervals for single-pollutant and
joint effect models for each pollutant combination in warm and
cold season analyses for an interquartile range (IQR) increase in
each pollutant at lag 0-2 days. IQR for 1-h max nitrogen dioxide
(NO2) concentrations = 12.87 ppb.
i
2
o
J
4
5
6
7
8
9
10
11
12
13
14
Across pollutant combinations that contained NC>2, in the warm season, joint effect
models reported consistent positive associations with pediatric asthma ED visits. For each
pollutant combination the association observed was larger in magnitude than any
single-pollutant association, including NC>2, but not equivalent to the sum of each
individual pollutant association for a specific combination. Furthermore, in the warm
season analysis, associations across the different joint effects models were found to be
relatively similar. The results during the cold season were inconsistent; however, when
focusing on the traffic pollutant combination, results from the joint effects models were
relatively similar to the single-pollutant results. The results of Winquist et al. (2014)
suggest that NCh alone and in combination with other pollutants is associated with
asthma ED visits, but also highlight the difficulty in separating out the independent effect
of a pollutant that is part of a mixture where multiple pollutants are often highly
correlated.
January 2015
5-97
DRAFT: Do Not Cite or Quote
-------
Summary of Asthma Hospital Admissions and Emergency Department
Visits
1 Recent studies that examined the association between short-term NO2 exposure and
2 asthma hospital admissions and ED visits report relatively consistent positive associations
3 which supports the results of U.S. and Canadian studies evaluated in the 2008 ISA for
4 Oxides of Nitrogen (Figure 5-7, Table 5-16). Across asthma hospital admission and ED
5 visit studies, there was some evidence of a different pattern of associations for each
6 outcome, with more immediate effects (i.e., lag 0) for asthma hospital admissions and
7 evidence of prolonged effects for asthma ED visits, with a number of studies showing
8 effects at multiday lags ranging from 0-2 to 0-4 days. Of the studies that examined
9 potential copollutant confounding, evidence supported that associations between
10 short-term NO2 exposures and asthma hospital admissions and ED visits remained
11 relatively unchanged in copollutant models (i.e., similar in magnitude or attenuated
12 slightly, but remaining positive). Additionally it is important to note that NO2 is often
13 found to be highly correlated with other traffic-related pollutants (e.g., PM2 5, UFPs, CO);
14 therefore, limiting the ability to determine if short-term NO2 exposures are independently
15 associated with asthma hospital admissions and ED visits. Recent studies of
16 multipollutant exposures further inform upon the effect of short-term NO2 exposures on
17 respiratory morbidity, specifically asthma. These studies demonstrate that: high daily
18 concentrations of NO2 alone and in combination with high daily concentrations of other
19 pollutants, such as PM2 5, can impact respiratory morbidity; and associations are observed
20 between asthma ED visits and NO2 alone and in combination with other traffic-related
21 pollutants, oxidants, and criteria pollutants.
22 A number of recent studies also examined whether there was evidence that the
23 association between short-term NO2 exposures and asthma hospital admissions and ED
24 visits was modified by season or some other individual- or population-level factor
25 (Chapter 7). An examination of seasonal differences in NO2-asthma hospital admission
26 and ED visit associations provide some evidence of NO2 effects being larger in
27 magnitude in the summer or warm season, and that seasonal associations may vary by
28 geographic location. Studies of individual- and population-level factors, provide evidence
29 of differences in associations by lifestage, with larger NO2 effects for children and older
30 adults, and more limited evidence for differences by sex, race/ethnicity, and
31 socioeconomic status (SES), specifically insurance status (Chapter 7). Additionally, there
32 is evidence that exposure differences, specifically whether a population lives in housing
33 that has low or high AERs that may influence the association between short-term NOx
34 exposures and asthma ED visits.
35 Additionally some recent studies examined various study design issues, including model
36 specification and exposure assignment. An examination of model specification, as
January 2015 5-98 DRAFT: Do Not Cite or Quote
-------
1 detailed in Section 5.2.6. indicates that the relationship between short-term NO2
2 exposures and respiratory-related hospital admissions, including those for asthma and
3 allergic disease, are sensitive to using less than 6 degrees of freedom (df) per year to
4 account for temporal trends, but robust to alternative lags and df, ranging from 3 to 6, for
5 weather covariates (Son et al.. 2013). An examination of various exposure assignment
6 approaches including single central site, average of multiple monitors, and
7 population-weighted average, suggests that each approach can influence the magnitude,
8 but not direction, of the NO2-asthma ED-visit risk estimate (Strickland et al.. 2011).
9 Finally, a few recent studies examined whether the shape of the NCh-asthma ED visit
10 relationship is linear or provides evidence of a threshold. These studies provide evidence
11 of a linear, no-threshold relationship between short-term NC>2 exposures and asthma ED
12 visits (Lietal. (20 lib): Strickland et al. (2010)).
January 2015 5-99 DRAFT: Do Not Cite or Quote
-------
Study
Burnett etal. (1999)
Sonetal. (2013)a
Ko et al. (2007)
Samolietal. (2011)a
Iskandar et al. (2012)
Linn et al. (2000)
Sonetal. (2013)a,b
Tolbert et al. (2000)
Peel et al. (2005)
Sarnatetal. (2013)d
Ito et al. (2007)
ATSDR (2006)
ATSDR (2006)
Stieb et al. (2009)
Jalaludin et al. (2008)
Peel et al. (2005)
Li etal. (2011)
Villeneuve et al. (2007)
Strickland etal. (2010)
Jaffe et al. (2003)
Location Age
Toronto, Canada All
8 South Korean cities All
Hong Kong All
Athens, Greece 0-14
Copenhagen, Denmark 0-18
Los Angeles, CA <30
8 South Korean cities All
Atlanta, GA All
Atlanta, GA All
Atlanta, GA All
New York, NY All
Bronx, NY All
Manhattan, NY All
7 Canadian cities All
Sydney, Australia 1-14
Atlanta, GA 2-18
Detroit, MI 2-18
Edmonton, Canada 2+
Atlanta, GA 5-17
2 Ohio cities 5-34
Lag
0
0
0
0
0-4
0
0
0
0-4c
0-4d
0
0
0
0
1
0-2
0-2
0-1
0-4
0-4
2
0-1
0
0
0-2
0-4e
0-4f
0-4
0-2
Hospital Admissions
ED Visits
5.0 0.0 5.0 10.0 15.0 20.0 25.0 30.0 35.0 40.0
Note: Results are standardized to a 20-ppb increase in 24-h avg NO2, a 30-ppb increase in 1-h max NO2, and a 40-ppb increase in 24-h avg NOX. a = results were presented for four
seasons; however the summer and winter estimates represented the largest and smallest estimates across seasons; b = this estimate is for allergic disease, which includes asthma;
c = risk estimate for NO2; d = risk estimate for NOX; e = time-series results; f = case-crossover results. Black = U.S. and Canadian studies evaluated in the 2008 ISA for Oxides of
Nitrogen; red = recent asthma hospital admission and ED visit studies. Circle = all-year; diamond = warm/summer months; square = cool/winter months.
Figure 5-7 Percentage increase in asthma hospital admissions and emergency department (ED) visits from
U.S. and Canadian studies evaluated in the 2008 Integrated Science Assessment for Oxides of
Nitrogen and recent studies in all-year and seasonal analyses.
January 2015
5-100
DRAFT: Do Not Cite or Quote
-------
Table 5-16 Corresponding risk estimates for studies presented in
Study
Location
Age
Avg Time
Season
Lag
Figure 5-7.
% Increase
(95% Cl)
Hospital Admissions
Burnett et al. (1999)
tSon etal. (2013)
Ko et al. (2007b)
tSamoli et al. (2011)
tlskandar et al. (2012)
Linn etal. (2000)
tSon etal. (2013)c>d
Toronto, Canada
8 South Korean
cities
Hong Kong
Athens, Greece
Copenhagen,
Denmark
Los Angeles, CA
8 South Korean
cities
All
All
All
0-14
0-18
<30
All
24-h avg
24-h avg
24-h avg
1-h max
24-h avg
24-h avg
24-h avg
All
All
Summer
Winter
All
All
Summer
Winter
All
All
All
Summer
Winter
0
0
0-4
0
0-4
0
0
2.6(0.5,4.9)
3.6(0.5,6.8)
16.2(5.1,28.6)
-1.1 (-6.5,4.5)
10.9(8.1, 13.8)
6.4 (-3.8, 17.6)
28.7 (-3.4, 71.3)
12.9 (-6.6, 36.5)
34.0(13.0, 58.0)a
62.0(25.0, 107)b
2.8(0.8,4.9)
3.8(1.0,6.6)
15.9(4.6,28.4)
-0.3 (-5.4, 5.1)
ED Visits
Tolbert et al. (2000)
Peel et al. (2005)
tSarnatetal. (201 3a)
Ito et al. (2007b)
ATSDR (2006)
ATSDR (2006)
tStieb et al. (2009)
tJalaludin et al. (2008)
Peel et al. (2005)
tLietal. (201 1b)
tVilleneuve et al. (2007)
tStrickland etal. (2010)
Atlanta, GA
Atlanta, GA
Atlanta, GA
New York, NY
Bronx, NY
Manhattan, NY
7 Canadian cities
Sydney, Australia
Atlanta, GA
Detroit, Ml
Edmonton,
Canada
Atlanta, GA
All
All
All
All
All
All
All
1-14
2-18
2-18
2+
5-17
1-h max
1-h max
24-h avg
24-h avg
24-h avg
24-h avg
24-h avg
1-h max
1-h max
24-h avg
24-h avg
1-h max
All
All
All
All
All
All
All
All
Warm
Cold
All
All
All
All
Warm
Cold
All
1
0-2
0-2
0-1
0-4
0-4
2
0-1
0
0
0-2
0-4e
0-4f
0-4
0-2
0.7 (-0.8, 2.3)
2.1 (-0.4, 4.5)
1.3(0.0, 2.4)b
12.0(7.0, 15.0)
6.0(1.0, 10.0)
On ( OR o ~7\
7.8(4.5, 11.1)
8.4(4.2, 12.5)
4.4 (-1.7, 10.4)
4.1 (0.8, 7.6)
2.9 (-7.9, 15.1)
9.1 (-0.8,20.2)
4.5(0.0, 7.5)
21.4(13.6, 31.0)
4.4 (-1.7, 10.4)
8.6(4.2, 13.3)
January 2015
5-101
DRAFT: Do Not Cite or Quote
-------
Table 5-16 (Continued): Corresponding risk estimates for studies presented in
Figure 5-7.
Study Location Age Avg Time Season Lag
Warm
Cold
Jaffe et al. (2003) 2 Ohio cities 5-34 24-h avg Summer 1
% Increase
(95% Cl)
16.0(9.1,23.5)
3.8 (-1.9, 9.6)
6-1 / o n *\A n\
Cl = confidence interval, ED = emergency department.
aRisk estimate for NO2.
"Risk estimate for NOX.
°Results were presented for four seasons; the summer and winter estimates represented the largest and smallest estimates for
each season.
dEstimate for allergic disease, which includes asthma.
eTime-series analysis results.
'Case-crossover analysis results.
fStudies published since the 2008 ISA for Oxides of Nitrogen.
5.2.2.5 Subclinical Effects Underlying Asthma Exacerbation:
Pulmonary Inflammation and Oxidative Stress
1 The evidence described in the preceding sections for NCh-related increases in airway
2 responsiveness (Section 5.2.2.1). decreases in lung function and increases in respiratory
3 symptoms in children with asthma (Sections 5.2.2.2 and 5.2.2.3). and asthma hospital
4 admissions and ED visits (Section 5.2.2.4) is coherent and consistent with a sequence of
5 key events by which NCh can plausibly lead to asthma exacerbation. Adding to the
6 proposed mode of action is evidence indicating NC>2 exposure-mediated pulmonary
7 inflammation, a key early event in asthma exacerbation that can mediate increases in
8 airway responsiveness (Section 4.3.2.5). The initiation of inflammation by NC>2 exposure
9 is supported by observations of NC>2-induced increases in eicosanoids, which mediate
10 recruitment of neutrophils (Section 4.3.2.3). Further, NC>2-induced increases in reactive
11 oxygen species (ROS) and reactive nitrogen species may impair epithelial barrier
12 function and initiate inflammation (Section 4.3.2.1). as many transcription factors
13 regulating expression of pro-inflammatory cytokines are redox sensitive. Most
14 information on the effects of NC>2 on pulmonary oxidative stress and injury is in healthy
15 people and animal models, and findings are inconsistent at ambient-relevant
16 concentrations (Section 5.2.7.4).
17 The 2008 ISA for Oxides of Nitrogen described evidence for NCh-induced increases in
18 pulmonary inflammation in some controlled human exposure studies and animal
19 toxicological studies (U.S. EPA. 2008a). There was coherence with findings from the few
20 available epidemiologic studies in children with asthma, which found associations
21 between short-term increases in ambient NC>2 concentrations and increases in exhaled
22 nitric oxide (eNO). In particular, coherence is found among disciplines for
January 2015 5-102 DRAFT: Do Not Cite or Quote
-------
1 NO2-associated increases in allergic inflammation. Recent studies, most of which were
2 epidemiologic, continued to find NC>2-associated increases in pulmonary inflammation
3 and oxidative stress. Biological indicators of pulmonary inflammation and oxidative
4 stress included those measured in exhaled breath; bronchoalveolar, bronchial, and nasal
5 lavage fluid; and sputum. Indicators of systemic inflammation in blood are evaluated in
6 the context of cardiovascular effects in Section 5.3.
Experimental Studies
7 As described in Section 5.2.2.1. controlled human exposure studies in adults with asthma
8 and allergy demonstrated increases in airway responsiveness in response to NC>2 exposure
9 with or without allergen challenge. These observations are supported by findings in
10 controlled human exposure studies involving adults with asthma and allergy and in a rat
11 model of allergic airway disease that NO2 exposure with or without an allergen challenge
12 resulted in increased indicators of allergic inflammation. This includes increases in IgE
13 and the influx and/or activation of eosinophils and neutrophils. Results provide evidence
14 that NO2 exposure can lead to exacerbation of allergic airways disease (discussed below
15 and in Section 4.3.2.6). These results provide support for epidemiologic evidence of
16 NO2~associated increases in inflammation in children with asthma and allergy.
17 The 2008 ISA for Oxides of Nitrogen (U.S. EPA. 2008a) described several studies that
18 examined inflammatory responses in adults with mild allergic asthma who were exposed
19 to NO2 followed by a specific allergen challenge (Table 5-17). In a series of studies from
20 the Karolinska Institute in Sweden, adults at rest were exposed to air or 260 ppb NO2 for
21 15-30 minutes followed by an antigen (birch or timothy pollen) challenge 4 hours later.
22 Bronchoalveolar lavage (BAL) and bronchial wash fluids were collected 19 hours after
23 allergen challenge. NO2 exposure for 30 minutes increased polymorphonuclear cells
24 (PMN) in the BAL and bronchial wash fluids and increased ECP in the bronchial wash
25 fluid compared with air exposure (Barck et al.. 2002). Reduced cell viability of BAL cells
26 and reduced volume of BAL fluid were also reported. ECP is released by activated
27 eosinophils; it is toxic to respiratory epithelial cells and thought to play a role in the
28 pathogenesis of airway injury in asthma. In a subsequent study, Barck et al. (2005a)
29 exposed adults with mild allergic asthma to air or NO2 for 15 minutes on Day 1 and twice
30 on Day 2, and for 15 minutes with allergen challenges following all of the exposures.
31 NO2 exposure induced an increased level of ECP in both sputum and blood and increased
32 myeloperoxidase levels in blood. These results suggest that NO2 may prime circulating
33 eosinophils and enhance activation of airway eosinophils and neutrophils in response to
34 an inhaled allergen. Nasal responses to nasal allergen challenge were also examined
35 following a 30-minute exposure to NO2 (Barck et al.. 2005b). No enhancement of nasal
January 2015 5-103 DRAFT: Do Not Cite or Quote
-------
allergen responses was observed in adult subjects. As noted in the 2008 ISA for Oxides
of Nitrogen (U.S. EPA. 2008a). these studies indicate that brief exposures to 260 ppb
NO2 can enhance allergen responsiveness in individuals with asthma.
Table 5-17 Controlled human exposure studies of pulmonary inflammation in
populations with asthma.
Disease Status;
Study Age; n, Sex
Exposure Details
Endpoints Examined
Barck et al. (2002)
Adults with mild
asthma and
allergy to birch or
timothy pollen;
mean age: 29 yr;
n = 6M, 7F
Histamine inhalation test to confirm
airway hyperresponsiveness.
266 ppb NO2for30 min
Inhaled allergen challenge 4 h after
pollutant exposure.
Albumin in serum samples.
BW and BAL cell
parameters-volume recovered, cell
viability, total cell counts,
macrophage concentrations, % of
neutrophils, # eosinophils, # mast
cells (performed 19 h after allergen
challenge).
ECP, MPO, IL-5, IL-8, eotaxin,
ICAM-1.
Barck et al.
(2005a)
Adults with mild
asthma and
allergy to birch or
timothy pollen;
mean age: 32 yr;
n = 10 M, 8F
260 ppb NO2
Day 1: one 15 min exposure with
bronchial challenge 4 h after
exposure.
Day 2: two 15 min exposures with
bronchial challenge 3 h after 2nd
exposure.
Total and differential cells counts of
induced sputum and venous blood
(samples taken on morning of
Days 1-3).
ECP, MPO in sputum.
Barck et al.
(2005b)
Adults with rhinitis Seasonal allergy confirmed by
and mild asthma;
mean age = 31 yr;
n = 9M, 7F
positive nasal challenge of allergen.
AHR confirmed by histamine test.
260 ppb NO2
Nasal allergen challenge 4 h after
exposure.
Total and differential cell counts and
cell viability in NAL (performed
before exposure, before allergen
challenge, and 1 h, 4 h, and 18 h
after challenge).
ECP and MPO in NAL fluid and
blood.
Ezratty et al.
(2014)
Asthma; median
29 yr (range:
20-69 yr);
n = 14 M, 5F;
(1) Filtered air
(2)203ppbNO2± 1.5%
(3) 581 ppb NO2 ± 3.2%
Same design for each exposure,
30 min on Day 1, twice for 30 min of
Day 2 separated by 1 h.
Induced sputum at baseline, 6 h,
32 h, and 48 h after the end of the
first exposure. Cell counts, ECP.
Spirometry for flow volume curve at
baseline, and daily before and
immediately following exposure and
immediately before sputum
induction. Symptom questionnaire 0,
15, and 30 min into exposure. FEV1
and PEF by portable spirometer
twice during exposure and hourly for
6 h following exposure.
January 2015
5-104
DRAFT: Do Not Cite or Quote
-------
Table 5-17 (Continued): Controlled human exposure studies of pulmonary
inflammation in populations with asthma.
Study
Jorres et al.
(1995)
Disease Status;
Age; n, Sex
Asthma; 27 ±5 yr;
n = 8M, 4F;
Healthy; 27 yr
(range: 21-33 yr);
n = 5M, 3F;
Exposure Details
1,000 ppbforS h;
Exercise 10 min on/10 min off at
individual's maximum workload.
Endpoints Examined
BAL fluid analysis 1 h after
exposure (cell counts, histamine,
prostaglandins).
Riedletal. (2012)
Phase 1: adults
with mild asthma;
mean age: 37 yr;
n = 10 M, 5F
Phase 2: adults
with mild asthma
and cat allergy;
mean age: 36 yr;
n = 6M, 9F
Inhalation challenge to detect
bronchoconstrictive response
Phase 1: methacholine; Phase 2: cat
allergen).
(1) 100 ug/m3 DEP for 2 h with
intermittent exercise
(2) 350 ppb NO2 control for 2 h with
intermittent exercise
Total counts and differential cell
counts (alveolar macrophages,
lymphocytes, PMNs, eosinophils) in
induced sputum (taken 22 h after
exposure).
Induced sputum fluid
assay-RANTES, eotaxin, ECP, IgG,
lgG4, IgA, IgM, IgE.
Cat-specific IL-4, IL-5, IL-8, IL-12,
GM-CSF, IFN-Y, TNF-a, tryptase.
Vaqaqqini et al.
(1996)
Asthma;
29 ± 14 yr;
n = 4M, 4F;
Healthy; 34 ± 5 yr;
n = 7M;
300 ppb for 1 h;
Exercise at VE = 25 L/min
Cell counts in sputum 2-h
post-exposure.
Wang et al.
(1995a): Wang et
al. (1995b)
Adults with
seasonal rhinitis;
mean age: 26 yr;
n = 6M, 10F
Nasal provocation with grass pollen Nasal lavage for inflammatory
allergen to confirm increase in nasal mediators fluid-ECP, MCT, MPO,
airway resistance. IL-8 (30 min after allergen
(1)400ppbNO2for6h challenge).
(2) 400 ppb NO2 for 6 h + allergen
challenge
Wang et al. (1999) Adults with grass Nasal airway resistance tests at rest, NAL—total and differential cell
allergy;
mean age: 32 yr;
n = 8M, 8F
after saline, and after allergen
challenge to confirm reactivity for
inclusion in study.
(1) 200 ug Fluticasone propionate
(FP) + 400ppbNO2for6h
(2) Matched placebo + 400 ppb NO2
for6h
counts (30 min after allergen
challenge).
Immunoassay of NAL fluid-ECP,
RANTES.
Witten et al.
(2005)
Adults with
asthma and house
dust mite allergy;
mean age: 32 yr;
n = 6M, 9F
Inhaled allergen challenge to
determine predicted allergen PC20.
400 ppb NO2 for 3 h w/intermittent
exercise
2nd inhaled allergen challenge,
starting at 4 doubling doses less
than APC20 and doubling until 20%
decrease in FEV-i.
Total and differential cell counts in
induced sputum—macrophages,
lymphocytes, neutrophils, and
eosinophils (samples taken at 6 and
26 h after allergen challenge).
BW = bronchial wash, ECP = eosinophil cationic protein, F = female, FEV = forced expiratory volume, GM-CSF = granulocyte
macrophage-colony stimulating factor, HDM = house dust mite, ICAM-1 = inter-cellular adhesion molecule 1, IL = interleukin,
M = male, MPO = myeloperoxidase, NAL = nasal lavage, NO2 = nitrogen dioxide, PC = provocative concentration, PEF = peak
expiratory flow, PMN = polymorphonuclear cells.
January 2015
5-105
DRAFT: Do Not Cite or Quote
-------
1 Additional studies have been performed using longer NO2 exposures (Table 5-17). (Wang
2 etal. (1999); Wangetal. (1995a): Wangetal. (1995b)) found that exposure of adults to
3 400 ppb NO2 for 6 hours enhanced allergen responsiveness in the nasal mucosa in
4 subjects with allergic rhinitis. Mixed grass pollen was used as the challenge agent and
5 was administered immediately after the NO2 exposure. Responses included increased
6 numbers of eosinophils and increased levels of myeloperoxidase and ECP in nasal lavage
7 fluid collected 30 minutes after the allergen challenge. Witten et al. (2005) did not
8 observe enhanced airway inflammation with allergen challenge in adults with asthma and
9 allergy to HDM allergen who were exposed to 400 ppb NCh for 3 hours with intermittent
10 exercise. HDM allergen was administered immediately after the NO2 exposure and a
11 decrease in sputum eosinophils was found 6 hours later (Witten et al.. 2005). Sputum
12 ECP levels were increased although this change did not reach statistical significance. The
13 authors suggested that their findings may be explained by a decreased transit of
14 eosinophils across the bronchial mucosa occurring concomitantly with NCh-induced
15 eosinophilic activation. Other investigators have noted that numbers of eosinophils do not
16 always correlate with allergic disease activity (Erjefalt et al.. 1999). Airway mucosal
17 eosinophilia is a characteristic feature of asthma and rhinitis; eosinophils exert their
18 effects via degranulation or cytolysis resulting in release of ECP and other mediators.
19 However, under conditions favoring eosinophil cytolysis, ECP concentrations may be
20 high and numbers of eosinophils may be low.
21 As noted in the 2008 ISA for Oxides of Nitrogen (U.S. EPA. 2008a). differing findings
22 between the studies in allergic individuals could be due to differences in timing of the
23 allergen challenge, the use of multiple- or single-allergen challenges, the use of BAL
24 fluid versus sputum versus nasal lavage fluid, exercise versus rest during exposure, and
25 differences in subjects. Furthermore, study protocols varied in the timing of biological
26 sample collection post-exposure to NO2 or allergen.
27 A recent study of adults with mild allergic asthma also did not find enhanced airway
28 inflammatory responses following exposure to NO2 (350 ppb NO2,2 hours, intermittent
29 exercise) (Table 5-17) (Riedl et al., 2012). Subjects exposed to NO2 followed by
30 methacholine challenge 1.5 hours later had increased levels of blood IgM and decreased
31 levels of sputum IgG4, interleukin (IL)-4, eotaxin, RANTES, and fibrinogen measured
32 22 hours after exposure. Subjects exposed to NO2 followed by cat allergen 1.5 hours later
33 did not exhibit changes in sputum cell counts measured 22 hours after exposure. While
34 these results are not consistent with NO2 enhancing airway inflammatory responses, it
35 should be noted that markers of eosinophil activation were not measured.
36 Several other studies investigated allergic inflammation following NO2 exposure in the
37 absence of a challenge. Torres etal. (1995) exposed healthy adults and those with asthma
January 2015 5-106 DRAFT: Do Not Cite or Quote
-------
1 and allergy to 1,000 ppb NC>2 for 3 hours and performed bronchoscopy 1 hour later. The
2 macroscopic appearance of the bronchial epithelium was altered after exposure in adults
3 with asthma compared to healthy controls; however, no accompanying changes in cell
4 counts in the BAL fluid were observed. Eicosanoid levels were also measured;
5 thromboxane B2 was increased in healthy adults and those with asthma following NC>2
6 exposure while prostaglandin D2 was increased and 6-keto prostaglandin Fla was
7 decreased after exposure only in adults with asthma. Because eicosanoids are known
8 mediators of inflammation, these results suggest that exposure to NO2 resulted in
9 activation of cell signaling pathways associated with inflammation. Vagaggini et al.
10 (1996) observed a decrease in eosinophils in sputum collected from adults with asthma
11 following a 1-hour exposure to 300 ppb NCh, though this decrease was not statistically
12 significant. In contrast, a recent controlled human exposure study reported an increase in
13 eosinophils and ECP following repeated NC>2 exposure in adults with atopic asthma
14 Ezratty etal. (2014). Subjects were exposed to 203 or 581 ppb NC>2 for 30 minutes on
15 one day and twice for 30 minutes on the second day. Compared with baseline,
16 statistically significant increases in the amount of ECP and the number and percentage of
17 eosinophils in sputum were observed after the three exposures to 600, but not 200 ppb
18 NC>2. Furthermore, ECP was highly correlated with eosinophil count in sputum. No
19 increases in either of these parameters were observed 6 hours after the first exposure to
20 600 ppb NO2.
21 Allergic inflammatory responses were also investigated in animal models of allergic
22 airways disease (Table 5-18). These studies involved sensitization and challenge with an
23 antigen followed by exposure to NCh. In one study in rats, which were sensitized and
24 challenged with HDM allergen, exposure to NC>2 (5,000 ppb, 3 hours) enhanced specific
25 immune responses and increased the numbers of lymphocytes, neutrophils, and
26 eosinophils in the airways (Gilmour etal.. 1996). In this study, the most pronounced
27 responses occurred when rats were exposed to NC>2 immediately after sensitization and
28 immediately after challenge with HDM antigen. Rats exposed to NC>2 twice had increased
29 levels of antigen-specific IgG and IgA and increased levels of IgE in BAL fluid 7 days
30 post-exposure to NC>2. In addition, an increase in the ratio of inflammatory cells (i.e.,
31 lymphocytes, neutrophils, eosinophils) to alveolar macrophages was observed 7 days
32 post-exposure to NC>2, although the total number of lavagable cells did not change.
33 In several studies in mice, which were sensitized and challenged with ovalbumin, NC>2
34 exposure over several hours or days failed to increase allergic inflammatory responses.
35 Exposures to 700 or 5,000 ppb NCh for 3 hours on a single day, for 2 hours on
36 3 consecutive days or for 6 hours on 3 consecutive days either reduced or had no effect
37 on indicators of eosinophil inflammation such as eosinophil counts, eosinophil
38 peroxidase activity, and total cellularity (Poynter et al., 2006; Hubbard et al., 2002;
January 2015 5-107 DRAFT: Do Not Cite or Quote
-------
1 Proust et al., 2002). Other findings included decreases in IL 5 levels in the BAL fluid at
2 both 24 and 72 hours after exposure to 5,000 ppb NCh and reductions in perivascular and
3 peribronchial cellular infiltrates after exposure to 700 ppb NC>2. Others have noted that
4 the ovalbumin-induced airway inflammation in mice does not involve substantial
5 eosinophil degranulation or cytolysis, which is characteristic of asthma and allergic
6 rhinitis in humans (Tvlahn-Erjefalt et al.. 2001). This suggests that species-related
7 differences may account for NO2-induced decreases in eosinophilic inflammation seen in
8 mouse models. Mechanisms underlying the NO2~induced decrease in airways
9 eosinophilia are unknown.
10 In summary, several controlled human exposure studies of adults with asthma and allergy
11 found that exposures to 260 ppb NO2 for 15-30 minutes or 400 ppb NO2 for 6 hours
12 increased inflammatory responses to an allergen challenge. These responses included
13 increases in number and activation of eosinophils and neutrophils. In the absence of an
14 allergen challenge, repeated exposure to 600 ppb NO2 for 30 minutes also enhanced
15 allergic inflammation in subjects with asthma and allergy. Other studies involving a
16 single exposure to NO2 (300-350 ppb, 1-2 hours; 1,000 ppb, 3 hours) did not show these
17 responses. Allergic inflammation was also enhanced by a 3-hour exposure to 5,000 ppb
18 NO2 in a rat model of allergic airways disease, as demonstrated by increases in IgE levels
19 and numbers of eosinophils and neutrophils. These results provide evidence for
20 NO2-induced exacerbation of allergic airways disease both in the presence and absence of
21 an allergen challenge (Section 4.3.2.6).
January 2015 5-108 DRAFT: Do Not Cite or Quote
-------
Table 5-18 Animal toxicological studies of pulmonary inflammation.
Study
Disease Status;
Age; n; Sex
Exposure Details
Endpoints Examined
Gilmouret al. (1996)
Rats (Brown Immunization with 100 ug antigen
Norway); (D. farina and D.
6 weeks; F; pteronyssinuss) + killed Bordetella
n = 5/group pertussis in 0.3 ml_ saline
Challenge with 50 ug allergen
(2 weeks after immunization),
followed by:
5,OOOppbNO2for3h
Endpoints examined 7 days after
exposure:
Total and differential cell counts
from lung lavage
Antigen-specific IgG, IgA, IgE
antibodies in serum and lavage
fluid
Lymphocyte proliferation
responsiveness
Proust et al. (2002)
Mice (BALB/c); Immunization with injection of 10 ug
6-7 weeks; M; OVA (Day 0 and Day 7)
n = 5/group Challenge with either 10 ug OVA or
saline control (Day 14)
Exposure following OVA/saline
challenge:
(1)5,OOOppbNO2
(2) 20,000 ppb NO2
Challenge to 0.1 M aerosol of
methacholine for 20 sec
Endpoints examined 24 h after
exposure:
BAL fluid total and differential cell
counts
Eosinophil peroxidase activity
Immunoassay of IL-4, IL-5
Anti-OVA IgE and lgG1 in serum
Lung histology
Hubbard et al. (2002) Mice (CB57BI/6); Sensitization by weekly injections of
adult; M/F 25 ug OVA for 3 weeks
Challenge with 20 mg/m3 OVA
aerosol for 1 h for 3 days or 10 days
Exposure following OVA aerosol
challenge:
(1)700ppbNO2for2h
(2) 5,000 ppb NO2 for 2 h
Total and differential cell counts
from lung lavage (24 h after
exposure)
Histology analysis (24 h after
exposure)
Povnter et al. (2006) Mice (C57BL/6)
Sensitization by 20 ug of OVA via i.p.
injections on Days 0 and 7
Challenge with OVA aerosol (1% in
phosphate buffered saline) for
30 min on Days 14-16
Exposures subsequent to OVA
challenge:
(1) 5,000 ppb NO2 for 6 h/day for 1,
3, 5 days
(2) 25,000 ppb NO2 for 6 h/day for 1,
3, 5 days
*Select groups given 20-day
recovery period
Methacholine challenge (0, 3.125,
12.5, 50 mg/mL in aerosol)
Endpoints examined after last day
of exposure or after 20 day
recovery:
BAL fluid-total and differential cell
counts; LDH
Histopathology analysis
mRNA levels of Gob5, MucSAC,
Th2, dendritic cell chemokine
CCL20 and eotaxin-1
F = female, IL = interleukin, LDH = lactate dehydrogenase, M = male, mRNA = messenger RNA, NO2 = nitrogen dioxide, OVA =
ovalbumin, Th2 = T-derived lymphocyte helper 2.
January 2015
5-109
DRAFT: Do Not Cite or Quote
-------
Epidemiologic Studies of Populations with Asthma
1
2
3
4
5
6
7
8
9
10
The observations described in the preceding sections for NCMnduced increases in
allergic inflammation provide support for the epidemiologic associations observed for
ambient or personal NC>2 with increases in inflammation in children with asthma and
allergy. The limited evidence in adults with asthma is inconclusive. The number of these
epidemiologic studies has increased dramatically since the 2008 ISA for Oxides of
Nitrogen, and recent studies expand on previous studies with exposure assessment
conducted in subjects' locations (e.g., homes, schools) and additional examination of
potential confounding by traffic-related copollutants. Ambient NCh concentrations,
locations, and time periods for epidemiologic studies of pulmonary inflammation and
oxidative stress are presented in Table 5-19.
Table 5-19 Mean and upper percentile concentrations of nitrogen dioxide (NO2)
in epidemiologic studies of pulmonary inflammation and oxidative
stress in populations with asthma.
Study3
Liu et al. (2009)
Barraza-Villarreal
et al. (2008)
Delfino et al.
(2006)
Delfino et al.
(2013)
Martins et al.
(2012)
Sarnat et al.
(2012)
Greenwald et al.
(2013)
Holquin et al.
(2007)
Location
Windsor, ON,
Canada
Mexico City,
Mexico
Riverside, CA
Whittier, CA
Riverside, CA
Whittier, CA
2 sites combined
Viseu, Portugal
El Paso, TX and
Ciudad Juarez,
Mexico
El Paso, TX
Ciudad Juarez,
Mexico
Study Period
Oct-Dec 2005
June
2003-June
2005
Aug-Dec 2003
July-Nov 2004
Jan and June,
2006 and 2007
Jan-Mar 2008
Mar-June
2010
2001-2002
Mean
NO2 Metric Concentration
Analyzed (ppb)
24-havgNO2 19.8
8-h max NO2 37.4
24-h avg NO2 Personal: 24.3
Personal: 30.9
8-h max NO2 Central site: 39.3
Central site: 35.1
24-h avg NO2 Central site: 27.4
1-week avg NO2b Across 4 periods:
4.5, 3.5, 9.8, 8.2C
96-h avg NO2 El Paso school:
4.5, 14.2, central
sites: 14.0, 18.5,
20.5
Ciudad Juarez
school: 18.7,27.2,
central site: none
96-h avg NO2 School A: 6.5
School B: 17.5
1 -week avg NO2 18.2
Upper Percentile
Concentrations
(PPb)
95th: 29.5
Max: 77.6
Max: 47.6
Max: 106
Max: 72.4
Max: 96
Max: 73.8
Max across 4
periods: 4.6, 4.0,
10.9, 9.4C
NR
NR
NR
January 2015
5-110
DRAFT: Do Not Cite or Quote
-------
Table 5-19 (Continued): Mean and upper percentile concentrations of nitrogen
dioxide (NO2) in epidemiologic studies of pulmonary
inflammation and oxidative stress in populations with
asthma.
Study3
Flamant-Hulin et
al. (2010)
Linetal. (2011)
Liuetal. (2014a)
Berhane et al.
(2011)
Romieu et al.
(2008)
Qian et al.
(2009a)
Maestrelli et al.
(2011)
Location
Clermont-Ferrand,
France
Beijing, China
Munich and
Wesel, Germany
13 southern
California
communities
Mexico City,
Mexico
Boston, MA; New
York City, NY;
Philadelphia, PA;
Madison, Wl;
Denver, CO; San
Francisco, CA
Padua, Italy
NO2 Metric
Study Period Analyzed
NR 5-day avg NO2
June 2007 24-h avg NO2
Sept 2007
Dec 2007
June 2008
Sept 2008
NR 24-h avg NO2
Sept-June 24-h avg NO2
2004-2005
Jan-Oct 2004 8-h max NO2
Feb 1997-Jan 24-h avg NO2
1999
1999-2003 24-h avg NO2
Mean
Concentration
(PPb)
Schools <14 ppb:
10.1
Schools >14 ppb:
17.4
24.3
30.4
45.3
26.6
25.9
15.9C
NR
35.3
23.6
Range across
seasons and
years: 20.9-37. Oc
Upper Percentile
Concentrations
(PPb)
Across schools:
75th: 14.0C
Max: 19.7C
NR
NR
NR
NR
NR
95th: 29.7C
NR
Max: 73.5
75th: 28.8
Max: 48.1
Range of 75th:
23.0-42.5C
NR = not reported, NO2 = nitrogen dioxide.
aStudies presented in order of first appearance in the text of this section.
""Subject-level exposure estimates calculated from outdoor NO2 at schools and other locations plus time-activity patterns.
°Concentrations converted from |jg/m3 to ppb using the conversion factor of 0.532 for NO2 assuming standard temperature (25°C)
and pressure (1 atm).
1
2
3
4
5
6
7
8
9
10
11
As in previous studies, the majority of evidence is for eNO. Across studies, eNO was
collected with a similar protocol, following the guidelines established by the ATS (2000).
eNO assessment methods also accounted for NO in the collection room, although eNO
has not been shown to be a reliable indicator of NO exposure (Section 4.2.3). eNO has
not been examined in controlled human exposure or animal toxicological studies of NO2
exposure, but several observations support the epidemiologic findings. NO2 exposure has
been shown to increase some pro-inflammatory cytokines and increase neutrophils and
eosinophils (Sections 4.3.2.6), which can activate inducible nitric oxide synthase or
produce NO in the lung during an inflammatory response (Barnes and Liew. 1995).
Higher eNO has been associated with higher eosinophil counts (Brodyet al.. 2013).
Further, eNO commonly is higher in children and adults with asthma and increases
January 2015
5-111
DRAFT: Do Not Cite or Quote
-------
1 during acute exacerbation (Soto-Ramos et al.. 2013; Carraro et al.. 2007; Jones et al..
2 2001; Kharitonov and Barnes. 2000).
Children with Asthma
3 Several recent and previous studies found associations between short-term increases in
4 ambient NC>2 concentration and increases in pulmonary inflammation in children with
5 asthma. Children were recruited mostly from schools, supporting the likelihood that study
6 populations were representative of the general population of children with asthma.
7 Asthma was assessed as self- or parental report of physician-diagnosed asthma, but the
8 studies varied in whether they assessed asthma severity or required the presence of
9 current symptoms in subjects. Across studies, associations varied in magnitude and
10 statistical significance; however, the consistent pattern of increasing eNO with increasing
11 short-term NC>2 exposure provides evidence of an association (Figure 5-8 and
12 Table 5-20). Most studies analyzed multiple endpoints, pollutants, lags of exposure, or
13 subgroups; however, with a few exceptions (Liu et al.. 2009; Barraza-Villarreal et al..
14 2008). a pattern of association was found across the multiple comparisons, thus reducing
15 the likelihood of associations found by chance alone or from publication bias.
January 2015 5-112 DRAFT: Do Not Cite or Quote
-------
Study
NO2 Metrics
Analyzed
Children with Asthma
Delfino et al. (2006) 24-h avg
lag 0-1 day avg
Exposure Assessment Subgroup
Total personal
Delfino etal. (2013)
Martins et al. (2012)
Sarnat etal. (2012)
Greenwald et al. (2013)
Lin etal. (2011)
Liu et al. (2009)
Liu et al. (In press)
Barraza-Villarreal et al.
(2008)
Berhaneetal. (2011)
Adults with Asthma
Qian et al. (2009)
8-h max
lag 0-1 day avg
24-h avg
lag 0-1 day avg
24-h avg
lag 0-4 day avg
24-h avg
lag 0-4 day avg
24-h avg
lag 0-3 day avg
24-h avg
lag 0 day
24-h avg
lag 0 day
24-h avg
lag 0 day
8-h max
lag 0 day
24-h avg
lag 1-6 day avg
24-h avg
lag 0 day
Central sites
Central sites
Subject modeled
School
Central site
School
Central site
Central sites
Central site
Central site
Central site
Central sites
All subjects
No anti-inflamm med
Anti-inflamm med use
ICS use
Anti-LT and ICS use
All subjects
All subjects
No ICS use
ICS use
All subjects
School A
School B
All subjects
Placebo
Beta-agonist use
ICS use
L
e
-10 -5 0 5 10 15 20 25 30 35 40 45 50 55
Percent change in eNO per increase in NO2 (95% Cl)a
Note: Results from more informative studies in terms of exposure assessment method and potential confounding considered are
presented first. Red = recent studies, black = previous studies. Study details and quantitative results reported in Table 5-20.
Table 5-20 presents results for an array of indications of inflammation and oxidative stress for which there was not sufficient
numbers to present in a figure. For some studies, eNO results could not be presented in the figure because results were not
reported in terms of percentage change eNO.
aEffect estimates are standardized to a 20-ppb increase for 24-h avg NO2 and 30-ppb increase for 1-h max NO2.
Figure 5-8 Associations of personal or ambient nitrogen dioxide (NO2) with
exhaled nitric oxide (eNO) in populations with asthma.
January 2015
5-113
DRAFT: Do Not Cite or Quote
-------
Table 5-20 Epidemiologic studies of pulmonary inflammation and oxidative stress in children and adults with
asthma.
Study NO2 Metrics
Population Examined and Methodological Details Analyzed Lag Day
Effect Estimate (95% Cl)
Single-Pollutant Model3
Copollutant Examination
Children with asthma: studies with small spatial scale exposure assessment and/or examination of copollutant confounding
Delfino et al. (2006)
Riverside, Whittier, CA
n = 45, ages 9-18 yr, persistent asthma and exacerbation in
previous 12 mo
Repeated measures. Examined daily for 10 days,
372 observations. Recruitment in schools of non-smokers from
non-smoking homes. No information on participation rate. Self-
report of physician-diagnosed asthma. Mixed-effects model
with random effect for subject with pollutant concentrations
centered on subject mean and adjusted for personal measures
of relative humidity, measures of personal temperature, follow-
up period. Adjustment for city, daily beta agonist use, weekend
did not alter results.
NO2-total personal
24-h avg
Compliance
assessed with
motion detectors.
Monitoring checked
daily.
0
eNO:
All subjects: 1.2% (-2.0, 4.3)
0-1 avg All subjects: 7.5% (2.0, 13)
No anti-inflammatory medication,
n = 14
2.6% (-9.9, 15)
Anti-inflammatory medication,
n = 31
9.3% (3.1, 16)
ICSuse, n = 19: 7.0% (0.23, 14)
Anti-leukotrienes + ICS use, n = 12
9.1% (-3.7, 22)
NO2-central site 0 All subjects: 0.81% (-4.5, 6.1)
8-h max 0-1 avg All subjects: 11% (3.2, 19)
Copollutant model results
in figure only.
. With PM2.5, EC, or OC:
NO2 results robust but
. increase in 95% Cl.
Copollutant results robust
to NO2 adjustment.
Weak correlations for
. personal exposures.
Spearman r= 0.20-0.31.
Stronger correlations for
central site pollutants.
Pearson r = 0.25-0.70.
' Central site CO not
associated with eNO.
tDelfino et al. (2013)
Riverside, Whittier, CA
Same population and methodology as Delfino (2006) above.
Analysis also indicated lack of confounding by respiratory
infections.
NO2-central site
24-h avg
1 site Riverside
within 12 km of
subjects' homes
2 sites Whittier
averaged, distance
NR
eNO:
0 -0.12% (-3.8, 3.7)
1 5.0% (1.2, 9.1)
0-1 avg 9.0% (2.9, 15)
For lag 0-1 avg:
With oxidative potential of
PM2.5: 3.8% (-5.1, 14)
With in vitro ROS from
PM2.5: 5.8% (-1.9, 14)
Copollutant associations
attenuated with NO2
adjustment.
Moderate correlations with
NO2. Spearman r=0.43
for ROS, 0.49 for oxidative
potential.
January 2015
5-114
DRAFT: Do Not Cite or Quote
-------
Table 5-20 (Continued): Epidemiologic studies of pulmonary inflammation and oxidative stress in children and
adults with asthma.
Study NO2 Metrics
Population Examined and Methodological Details Analyzed Lag Day
Effect Estimate (95% Cl)
Single-Pollutant Model3
Copollutant Examination
tMartins (2013): Martins et al. (2012)
Viseu, Portugal
n = 51, mean age 7.3 (SD: 1.1)yr, 53% with atopy
Repeated measures. 4 measurements over 2 different
seasons. Recruitment from urban and suburban schools.
-66% participation rate. Parental report of wheeze in previous
12 mo. GEE adjusted forage, sex, parental smoking, parental
education, atopy, time of visit, average temperature, relative
humidity. Also included height, weight, older siblings,
mold/dampness in home, fireplace in home, pets in home
because their inclusion changed the effect estimate for at least
1 pollutant by >10%.
NCb-subject
modeled outdoor
24-h avg
Estimated from
school outdoor NO2,
20 city locations,
MM5/CHIMERE
modeling, and daily
activity patterns.
20% time spent at
school, 65% at
home.
0-4 avg eNO: 14% (-12, 40)
Exhaled breath condensate pH:
-2.6% (-3.9, -1.3)
For EEC pH only:
With PMio:
0.30 (-3.0, 3.6)
With benzene:
-1.7 (-3.6, 0.26)
With ethylbenzene:
-1.6 (-3.7, 0.49)
PMio robust to adjustment
for NO2. VOCs attenuated
to null. Negative or weakly
positive correlations with
NO2. Spearman r= -0.72
to -0.55 for PMio, -0.43 to
0.14 for various VOCs.
tSarnatetal. (2012)
El Paso, TX and Ciudad Suarez, Mexico
n = 29 per city, ages 6-1 yr, asthma and current symptoms
Repeated measures. Examined weekly for 16 weeks,
697 observations. Recruitment from schools representing a
gradient of traffic, subjects from non-smoking homes. No
information on participation rate. Self-report of physician-
diagnosed asthma. GLM with subject as random effect and
adjustment for school, temperature, relative humidity, indoor
NO. Adjustment for medication use, cold symptoms did not
alter results.
NO2-school outdoor
Each city: one
school 91 m from
major road, one in
residential area.
NO2-school indoor
NO2-central site
1 site in El Paso, TX
near major road.
All 24-h avg
eNO:
0-4 avg All subjects: 6.3% (2.5, 10)
No ICS use, n = 10: 6.6% (2.6, 11)
ICSuse, n = 19: 1.1% (-8.9, 12)
All subjects: 0.53% (0.11, 1.0)
All subjects: 1.7% (-1.0, 4.5)
With O3: 8.8% (4.6, 13)
No copollutant model with
. PM2.5 or PM-io-2.5, which
were associated with
. eNO. No association with
BC among all subjects.
. Weak to moderate
correlations with NO2.
Spearman r= -0.39 to
'0.32 for PM2.s; -0.24 to
0.04 for PM-io-2.5.
tGreenwald et al. (2013)
El Paso, TX
n = 38, mean age 10 yr, 76% Mexican-American
Repeated measures. Examined weekly for 13 weeks,
436 observations. Recruitment from schools in low- and high-
traffic area. No information on participation rate. School record
of physician-diagnosed asthma. GLM with subject as random
effect and adjusted for school, temperature, relative humidity,
indoor NO.
NO2-school outdoor
School A: residential 0-3 avg
area, School B:
91 m from major
road.
NO2-school indoor
All 24-h avg
eNO:
School A: -0.86% (-38, 58)
SchoolB: 30% (-3.1,73)
School A: -16% (-53, 47)
SchoolB: 5.6% (-19, 37%)
No copollutant model.
BC, VOCs (central site)
associated with eNO.
Moderate correlations with
NO2. Pearson
.r= 0.47-0.62.
BTEX associated with
eNO. Highly correlated
with NO2. r=0.77.
January 2015
5-115
DRAFT: Do Not Cite or Quote
-------
Table 5-20 (Continued): Epidemiologic studies of pulmonary inflammation and oxidative stress in children and
adults with asthma.
Study NO2 Metrics
Population Examined and Methodological Details Analyzed Lag Day
Effect Estimate (95% Cl)
Single-Pollutant Model3
Copollutant Examination
tHolquin et al. (2007)
Ciudad Juarez, Mexico
n = 95, ages 6-12 yr, 78% mild asthma, 58% with atopy
Repeated measures. Examined biweekly for 4 mo. 87%
participation. Self-report of physician-diagnosed asthma.
Linear and non-linear mixed effects model with random effect
for subject and school adjusted for sex, body mass index, day
of week, season, maternal and paternal education, passive
smoking exposure.
NO2-school outdoor
24-h avg
Schools located
239-692 m from
homes.
0-6 avg No quantitative results reported for
eNO. No association was reported.
No copollutant model.
Road density but not
PlVhsor EC associated
with eNO.
tZhu(2013): Lin etal. (2011)
Beijing, China
n = 36, ages 9-12 yr, 22% with asthma
Repeated measures before and after Olympics. Examined
daily for five 2-week periods. 1,581 observations. Recruitment
from school. Selection from 437 (60%) students who
responded to initial survey, 95% follow-up participation. GEE
adjusted for temperature, relative humidity, body mass index.
NO2-central site
24-h avg
Site 650 m from
school.
eNO:
0 All subjects: 22% (18, 26)
Asthma: 23% (16, 31)
1 Asthma: 12% (4.0, 20)
Among all subjects:
WithBC: 5.6% (0.38, 11)
With PlVh.s: 14% (9.5, 19)
No change in BC with NO2
adjustment. PlVh.5 reduced
but positive. NO2 highly
correlated with BC
(r= 0.68), moderately
correlated with PlVhs
(r=0.30).
tFlamant-Hulinetal. (2010)
Clermont-Ferrand, France
n = 34, mean age: 10.7 (SD: 0.7) yr, 44% with atopy
Cross-sectional. Recruitment from schools. 69% participation
rate. Self- or parental-report of lifetime asthma. For some
subjects, eNO measured up to 1 week before pollutants. GEE
adjusted for atopy, mother's birth region, parental education,
family history of allergy, prenatal and childhood smoking
exposure. Did not consider potential confounding by weather.
NO2-school outdoor
24-h avg
NO2-school indoor
24-h avg
0-4 avg log eNO comparing >14.3 vs.
<14.3ppbNO2:
0(-0.14, 0.14)
0(-0.13, 0.14)
No copollutant model.
PM2.5, acetylaldehyde
associated with eNO.
tLiu(2013): Liu etal. (2009)
Windsor, ON, Canada
n = 182, ages 9-14 yr
Repeated measures. Examined weekly for 4 weeks, same day
of week. 672 observations. Recruitment from schools. No
information on participation rate. Parental report of physician-
diagnosed asthma. Mixed effect model with random effect for
subject and adjusted for testing period, temperature, relative
humidity, daily medication use.
NO2-central site
24-h avg
Average of 2 sites.
99% subjects live
within 10 km of
sites.
0
1
0-2 avg
0
1
0-2 avg
eNO:
17% (-5.8, 47)
7.7% (-12, 32)
1.5% (-32, 50)
TEARS:
48% (3.9, 111)
22% (-11, 67)
131% (23, 334)
For TEARS only:
with PM2s:
31% (-30, 145)
withSO2:43%(-10, 126)
Small decrease in PM2.5
estimate with adjustment
for NO2. NO2 highly
correlated with PM2.5
(Spearman r= 0.71),
weakly with SO2 (r = 0.18).
January 2015
5-116
DRAFT: Do Not Cite or Quote
-------
Table 5-20 (Continued): Epidemiologic studies of pulmonary inflammation and oxidative stress in children and
adults with asthma.
Study NO2 Metrics
Population Examined and Methodological Details Analyzed Lag Day
Effect Estimate (95% Cl)
Single-Pollutant Model3
Copollutant Examination
tliuetal. (2014a)
Munich and Wesel, Germany
n = 192, age 10 yr
Cross-sectional. Recruitment from GINIplus, LISAplus birth
cohort studies. No information reported on participation rate or
ascertainment. Parental report of physician-diagnosed asthma.
GAM adjusted for cohort, city, sex, parental education,
parental history of atopy, indoor gas pollution, current pets,
maternal prenatal smoking, smoking exposure at age 10 yr,
temperature. Results not altered by adjustment for asthma
medication use or annual avg NO2 estimated from LUR.
NO2-central site
24-h avg
1 site per city in
suburban locations.
0 eNO:
Both cities: 51% (-11, 154)
Results in figure show association
only in Munich, null in Wesel.
tBarraza-Villarreal et al. (2008)
Mexico City, Mexico
n = 119-129, ages 6-14 yr, 54% persistent asthma, 89%
atopy
Repeated measures. Examined every 15 days for mean
22 weeks. 1,004 observations. Recruited from pediatric clinic.
Asthma severity assessed by pediatric allergist. No information
on participation rate. Linear mixed effects model with random
effect for subject and adjusted for sex, body mass index, lag
one minimum temperature, ICS use, time. Adjustment for
outdoor activities, smoking exposure, antiallergy medication
use, season did not alter results.
NO2-central site
8-h max
Monitors within 5 km
of school or home.
Low correlation for
school vs. central
site: Spearman
r=0.21
0
eNO: 8.4% (7.9, 9.0)
With PMio:
23% (-37, 137) among
children with asthma.
PMio results not altered
with NO2 adjustment.
34% (15, 56) among all
1,985 children. PMio
association attenuated
with NO2 adjustment.
Moderate correlated with
NO2. Spearman r = 0.59.
Children with asthma: studies with central site exposure assessment and no examination of copollutant confounding
lnterleukin-8: 1.2% (1.1, 1.3)
EBCpH: -0.5% (-1.5, 0.50)
No copollutant model.
• PM2.sand Os associated
with eNO and IL-8.
' Moderate or weak
correlation with NO2.
Pearson r = 0.61 for PM2 5,
0.28 for O3.
tRomieu et al. (2008)
Mexico City, Mexico
n = 107, mean age 9.5 yr. 48% persistent asthma, 90% atopy
Repeated measures. EEC collected every 2 weeks for
2-16 weeks. 480 observations. Recruitment from allergy clinic.
No information on participation rate. 25% EEC samples below
detection limit, assigned random value 0-4.1 nmol. Malondi-
aldehyde associated with wheeze and asthma medication use.
GEE model adjusted for sex, school shift, temperature,
chronological time. Adjustment for outdoor activities, parental
smoking did not alter results.
NO2-central site
8-h max
Similar results for
1-h max and 24-
h avg.
Monitors within 5 km
of school or home.
Log malondialdehyde:
0.13 (-0.10, 0.35)
No copollutant model.
PM2.5, distant to closest
avenue, 4.5-h traffic count,
and Os also associated
with malondialdehyde.
Moderate correlation with
NO2. Pearson r = 0.44 for
Os and 0.54 for PM2.5.
January 2015
5-117
DRAFT: Do Not Cite or Quote
-------
Table 5-20 (Continued): Epidemiologic studies of pulmonary inflammation and oxidative stress in children and
adults with asthma.
Study NO2 Metrics
Population Examined and Methodological Details Analyzed Lag Day
Effect Estimate (95% Cl)
Single-Pollutant Model3
Copollutant Examination
tBerhaneetal. (2011)
13 Southern CA towns
n = 169, ages 6-9 yr
Cross-sectional. Recruitment from schools. Parental report of
physician-diagnosed asthma and history of respiratory allergy.
Linear regression adjusted for community, race/ethnicity, age,
sex, asthma, asthma medication use, history of respiratory
allergy, eNO collection time, body mass index, smoking
exposure, parental education, questionnaire language,
season, multiple temperature metrics, eNO collected outdoors.
NO2-central site
24-h avg
Sites in each
community. # sites
in each community
NR.
1-6 avg eNO: -6.7% (-31 26%)
No copollutant model.
PM2.5, PMlO, Os
associated with eNO.
Moderate or weak
correlations with NO2.
Pearson r = 0.47 for PlVh 5,
0.49 for PMio, 0.15forO3.
Adults with Asthma: central site exposure assessment, no examination of potential confounding by traffic-related copollutants
tQian et al. (2009a)
Boston, MA; New York, NY; Denver, CO; Philadelphia, PA;
San Francisco, CA; Madison, Wl.
n = 119, ages 12-65 yr, persistent asthma, nonsmokers
Repeated measures. Examined every 2-4 weeks for
16 weeks. 480 person-days. No information on participation
rate. Study population representative of full cohort. Asthma
medication trial and a priori comparison of medication
regimens. Linear mixed effects model adjusted forage, sex,
race/ethnicity, center, season, week, daily average
temperature, daily average humidity. Adjustment for viral
infections did not alter results.
NO2-central site
24-h avg
Average of all
monitors within
51 km of subject
ZIP code centroid.
eNO:
0 All subjects: 1.1% (0.52, 1.7)
Placebo: 0.79% (-0.08, 1.7)
Beta-agonist use: 0.86% (0.08, 1.6)
ICSuse: 1.8% (0.62, 2.9)
0-3 All subjects: 0.94% (0.09, 1.8)
With PMio:
0.69% (-0.09, 1.5)
With O3: 0.94% (0.43, 1.5)
WithSO2: 1.2% (0.52, 1.9)
Copollutant effect
estimates attenuated with
1 adjustment for NO2.
Correlations NR.
tMaestrelli et al. (2011)
Padua, Italy
n = 32, mean age 39.6 (SD: 7.5) yr, 81% persistent asthma
Repeated measures. Examined 6 times over 2 yr. Selected
from database of beta-agonist users (>6/yr for 3 yr), diagnosis
clinically confirmed. 76% follow-up participation. Drop outs did
not differ from participants. GEE adjusted for daily average
temperature, humidity, atmospheric pressure, asthma
medication use, current smoking status.
NO2-central site
24-h avg
2 sites in city
eNO (ppb):
All subjects: 3.1 (-14,21)
Nonsmokers, n = 22: 2.9 (-20, 26)
EEC pH:
All subjects: 0(-0.19, 0.21)
Nonsmokers: -0.09 (-0.24, 0.05)
No copollutant model.
Personal and central site
PM2.5 and PMio not
associated with eNO. No
• associations with central
site CO. Association found
with Os and SO2.
Correlations NR.
Note: More informative studies in terms of the exposure assessment method and potential confounding considered are presented first.
ICS = inhaled corticosteroid, eNO = exhaled nitric oxide, NR = not reported, ROS = reactive oxygen species, GEE = generalized estimating equation, EEC = exhaled breath
condensate, GLM = generalized linear mixed effects model, TEARS = thiobarbituric acid reactive substances, GAM = generalized additive models, LUR = land-use regression,
Cl = confidence interval, CO = carbon monoxide, EC = elemental carbon, IL = interleukin, NO2 = nitrogen dioxide, O3 = ozone, OC = organic carbon, PM = particulate matter,
SD = standard deviation, VOC = volatile organic compound.
aEffect estimates are standardized to a 20 ppb increase for 24-h avg NO2 and 25 ppb increase for 8-h max NO2.
fStudies published since the 2008 ISA for Oxides of Nitrogen.
January 2015
5-118
DRAFT: Do Not Cite or Quote
-------
1 Key evidence was provided by studies with NO2 exposures assessed for subjects'
2 locations, comparison of various exposure metrics, and/or examination of confounding by
3 traffic-related copollutants. As reported by few studies, participation rates were high (87,
4 95%, Table 5-20). Selective participation by certain groups was not indicated. These
5 studies examined a limited number of exposure lags but specified them a priori. Across
6 studies, associations were found with multiday averages of NO2 (i.e., 0-1 avg to 0-6 avg)
7 (Figure 5-8 and Table 5-20). with Delfino et al. (2006) finding a stronger association of
8 eNO with lag 0-1 avg than lag 0 or 1 day NO2. Strong exposure assessment was
9 characterized as personal monitoring (Delfino et al.. 2006): estimation of individual
10 outdoor exposures based on monitoring, modeling, and daily activity patterns (Martins et
11 al.. 2012): monitoring at or near schools (Greenwald et al.. 2013: Sarnat et al.. 2012: Lin
12 et al.. 2011: Holguin et al.. 2007): or examination of central site ambient concentrations
13 that are temporally correlated with total personal NO2 measurements (Delfino et al..
14 2013).
15 In comparisons with central site NO2, associations with eNO were similar to personal
16 NO2 among children with asthma in Riverside and Whittier, CA. A 20-ppb increase in
17 24-h avg (lag 0-1 day avg) NO2 was associated with a 7.5% (95% CI: 2.0, 13%) increase
18 in eNO for personal NO2 exposure (Delfino et al.. 2006) and a 9.0% (95% CI: 2.9, 15%)
19 increase for ambient NO2 averaged between central sites in each community (Delfino et
20 al.. 2013). An increase in 8-h max NO2 assigned from each child's community central
21 site was associated with a similar increase in eNO as 24-h avg personal NO2 based on the
22 interquartile ranges of NO2 (1.4% [95% CI: 0.39, 2.3] per 12-ppb increase in 8-h max
23 central site NO2 and 1.6% [95% CI: 0.43, 2.8] per 17-ppb increase in 24-h avg personal
24 NO2). Personal and central site NO2 were moderately correlated (Spearman r = 0.43).
25 Thus, despite the potential for greater exposure measurement error due to
26 within-community variability in ambient NO2 concentrations and variation in
27 time-activity patterns (Section 3.4.4). daily variation in ambient NO2 to some extent is
28 represented in daily variation in personal NO2 exposures of these children that is
29 associated with eNO. Such results provide a rationale for drawing inferences about
30 ambient NO2 exposure from associations observed with total personal NO2 exposures.
31 Among children with wheeze in Portugal, a 20-ppb increase in 1-week avg individual
32 estimates of ambient NO2 exposure was associated with a 14% (95% CI: -12, 40%)
33 increase in eNO and a -2.6% (95% CI: -3.9, -1.3%) change in exhaled breath
34 condensate (EEC) pH (Martins et al.. 2012). School and home indoor NO2 concentrations
35 were nondetectable, providing support for an association with ambient NO2. Further,
36 time-weighted averages of microenvironmental NO2 have shown good agreement with
37 personal NO2 (Section 3.4.3.1). Children were reported to spend 85% of time at home or
January 2015 5-119 DRAFT: Do Not Cite or Quote
-------
1 school, underscoring the importance of the individual-level exposure estimation in this
2 study.
3 Evidence also points to associations of eNO in children with asthma with NO2
4 concentrations measured outside schools. Of the studies conducted in communities along
5 the Texas/Mexico border, most found NO2-associated increases in eNO. In comparisons
6 of NO2 exposure metrics, eNO was more strongly associated with outdoor school NO2
7 than central site NO2 (Sarnat et al.. 2012) or school indoor NO2 (Greenwald et al.. 2013;
8 Sarnat etal.. 2012) (Figure 5-8 and Table 5-20). In the Texas/Mexico study, a 20-ppb
9 increase in 96-h avg NO2 concentration was associated with increases in eNO of 6.3%
10 (95% CI: 2.5, 10.2%) for outdoor school, 0.5% (95% CI: 0.1, 1.0%) for indoor school,
11 and 1.7% (95% CI: -1.0, 4.5%) for central site. NO2 from the single central site in El
12 Paso was moderately to strongly correlated (Spearman r = 0.63-0.91) with school NO2
13 (Sarnat et al.. 2012). suggesting that for some schools, the central site measures captured
14 temporal variation in school-based measures. However, the variability in NO2 found
15 across schools (coefficient of variation = 59%) indicates that the stronger associations
16 with school NO2 may be attributable to school measurements better representing
17 variability in NO2 within the area. Misrepresenting variability has been shown to
18 influence exposure measurement error (Section 3.4.5.1). Holguin et al. (2007) did not
19 find an association with eNO in children with asthma in Ciudad Juarez schools. No
20 association was found in a study of children in France (Flamant-Hulin et al.. 2010).
21 However, this study had weaker methodology because of its cross-sectional design,
22 comparison of eNO between low and high NO2 (means 10.1 and 17.4 ppb), and for some
23 subjects, measurement of eNO 1 week before NO2. NO2 measured within 650 m of
24 subjects' schools (lag 0 day of 24-h avg) was associated with eNO among children in
25 Beijing, China examined before and after the 2008 Olympics (Lin etal.. 2011).
26 With regard to confounding, most studies with small spatial-scale exposure assessment
27 adjusted for temperature and humidity, with a few additionally adjusting for asthma
28 medication use (Sarnat et al.. 2012; Delfino et al.. 2006). An array of traffic-related
29 copollutants was examined, and most studies found associations with EC/BC, OC, PM2 5,
30 and VOCs. These copollutants showed a wide range of correlations with NO2 (Pearson or
31 Spearman r = -0.43 to 0.77). There is some evidence for NO2 effects that are independent
32 from these traffic-related copollutants. NO2-eNO associations were found with
33 adjustment for personal PM2 5, EC, or OC. Personal exposure measures were more
34 weakly correlated with NO2 (Spearman r = 0.20-0.33) than central site measures
35 (r = 0.20-0.70) (Delfino etal.. 2006). For central site NO2, associations with eNO
36 decreased but remained positive with adjustment for BC or the oxidative potential of
37 ambient PM2 5 extracts measured in vitro (Delfino et al.. 2013; Lin et al.. 2011)
38 (Table 5-20. Figures 5-16 and 5-17). The latter results support an independent association
January 2015 5-120 DRAFT: Do Not Cite or Quote
-------
1 with NCh because oxidative stress is a key event in the mode of action for NO2, PIVb 5,
2 and other traffic-related pollutants (Section 5.1.2). The studies conducted in El Paso, TX
3 and Ciudad Juarez, Mexico did not analyze copollutant models with EC/BC or PM2 5
4 However, NC>2 associations were less variable across schools than were PlVfc 5
5 associations, and in Ciudad Juarez, NC>2 but not EC or PM2 5, was associated with eNO
6 (Sarnatetal.. 2012: Holguin et al.. 2007V
7 Pulmonary inflammation also was associated with VOCs (Greenwald et al.. 2013;
8 Martins etal.. 2012). In the El Paso schools, because of the high correlation (Pearson
9 r = 0.77) between NC>2 and benzene, toluene, ethylbenzene, xylene (BTEX), an
10 independent association is not discernible for either pollutant. Reporting
11 copollutant-adjusted results only for EEC pH, Martins et al. (2012) found that
12 associations for individual estimates of outdoor NC>2 exposure were similar after
13 adjustment for VOCs, which showed no or negative correlations with NC>2 (range of
14 Spearman correlation coefficient across four visits: r = -0.42 to 0.03) (Table 5-20). VOC
15 estimates were attenuated to the null with adjustment for NCh; thus, NO2 may have
16 confounded associations for VOCs. Other pollutants, Os, SO2, PMio, and PMio-2.5, were
17 associated with pulmonary inflammation and oxidative stress but did not show strong
18 positive correlations with NO2 (r = -0.72 to 0.18). NO2 effect estimates increased with
19 adjustment for Os (Sarnatetal.. 2012) and became null with PMio (Martins et al.. 2012).
20 But, PMio and NO2 were strongly negatively correlated (r = -0.55 to -0.77).
21 Other studies have weaker implications for inferring an independent effect of NO2 on
22 pulmonary inflammation and oxidative stress in children with asthma. They all assigned
23 NO2 exposure as ambient concentrations from one city central site or sites 5 km or 10 km
24 from subjects' homes. While they adjusted for potential confounding by meteorological
25 factors and asthma medication use, most did not examine confounding by traffic-related
26 copollutants. Findings were variable for indicators of inflammation among eNO, IL-8,
27 and exhaled breath condensate pH as well as indicators of oxidative stress related to lipid
28 peroxidation. Some studies found associations with ambient NO2 (Liu etal.. 2014a;
29 Barraza-Villarreal et al.. 2008) or inconsistent associations across the lags of exposure or
30 specific endpoints examined (Liu et al., 2009). In others, effect estimates with wide 95%
31 CIs did not support associations (Berhane etal.. 2011; Romieu et al.. 2008) (Figure 5-8
32 and Table 5-20). Studies also found associations with traffic proximity and volume and
33 with PM2 5, which was moderately to highly correlated with NO2 (r = 0.49-0.71) (Liu et
34 al.. 2014a: Berhane etal.. 2011; Liu et al.. 2009; Romieu et al.. 2008):(Barraza-Villarreal
35 etal.. 2008). Copollutant modeling was conducted in a study of children in Windsor,
36 Canada, and effect estimates for NO2 were largely attenuated with adjustment for PM2 5
37 (Table 5-20) (Liu etal.. 2009). PM2 5 estimates were less altered with adjustment for NO2;
38 however, the reliability of the copollutant model is questionable because of the high
January 2015 5-121 DRAFT: Do Not Cite or Quote
-------
1 NO2-PM2.5 correlation (r = 0.71). In the limited analysis of copollutant models with PMio
2 (r = 0.59) or SCh (r = 0.18), NCh remained associated with pulmonary inflammation or
3 oxidative stress (Liu et al.. 2014a; Liu etal.. 2009).
4 Studies of children with asthma did not clearly identify potential factors that could
5 modify ambient NCh-associated increases in pulmonary inflammation but primarily
6 conducted post hoc analyses. Associations were not found to differ by sex (Sarnat et al..
7 2012; Liu et al.. 2009; Delfino et al.. 2006). Larger associations were found in children
8 not using ICS in some (Sarnat et al.. 2012; Liu et al., 2009). but not all, (Delfino et al..
9 2006) studies. Because of the heterogeneity in the definition of ICS use and lack of
/ O J
10 assessment of ICS compliance, it is not clear whether ICS use represents well-controlled
11 or more severe asthma across populations. Several studies specified comparisons between
12 children with and without asthma a priori. While children with asthma had higher eNO,
13 results indicated no difference in associations with NC>2 between groups (Patel et al..
14 2013; Lin etal.. 2011; Flamant-Hulin et al.. 2010: Holguin et al.. 2007) or larger
15 associations in children without asthma (Berhane etal.. 2011: Barraza-Villarreal et al..
16 2008).
Adults with Asthma
17 Recent epidemiologic studies of pulmonary inflammation in adults with asthma, which
18 were not available for the 2008 ISA for Oxides of Nitrogen (U.S. EPA. 2008a). showed
19 contrasting associations with ambient NC>2. Both studies examined adults predominately
20 with persistent asthma, assessed NO2 exposure from central site monitors, and adjusted
21 for temperature and humidity. In a U.S. multicity (Boston, MA; New York, NY;
22 Philadelphia, PA; San Francisco, CA; Madison, WI) study nested within an asthma
23 medication trial, a 20-ppb increase in lag 0 day of 24-h avg NCh (averaged from monitors
24 located within 32 km of subjects' homes) was associated with a 0.26-ppb (95% CI: 0.12,
25 0.40) increase in eNO (Qian et al.. 2009a). A similar increase in eNO was found for lag
26 0-3 day avg NO2 but not lags 1, 2 or 3. A larger effect was estimated in the daily ICS
27 group than the placebo or beta-agonist groups only for lag 0 day NO2. Among children
28 and adults with asthma in Padua, Italy, a large percentage of whom reported ICS use, lag
29 0 day of 24-h avg ambient NO2 was not associated with eNO or exhaled breath
30 condensate pH (Maestrelli et al.. 2011). The U.S. multicity study did not examine
31 whether the association for ambient NO2 was independent of other traffic-related
32 pollutants. Copollutant models were examined only for PMio, SO2, and Os, in which NO2
33 remained associated with eNO (Qian et al.. 2009a). Adjustment for NO2 attenuated the
34 effect estimates for PMio, SO2, and Os, indicating that the copollutant associations were
35 confounded by NO2.
January 2015 5-122 DRAFT: Do Not Cite or Quote
-------
5.2.2.6 Summary of Asthma Exacerbation
1 Evidence integrated across the array of health outcomes and disciplines strongly supports
2 a relationship between short-term NC>2 exposure and asthma exacerbation. The evidence
3 for allergic inflammation, increased airway responsiveness, and clinical events, such as
4 respiratory symptoms in populations with asthma as well as ED visits and hospital
5 admissions for asthma, is consistent with the sequence of key events within the mode of
6 action for asthma exacerbation (Figure 4-1) and provides biological plausibility for a
7 relationship with NC>2 exposure. Much of this evidence, especially from experimental
8 studies, was described in the 2008 ISA for Oxides of Nitrogen (U.S. EPA. 2008a). Recent
9 findings, primarily from epidemiologic studies, continue to indicate NC^-associated
10 increases in asthma exacerbation. Many recent epidemiologic studies contribute
11 additional exposure assessment in subjects' locations and examination of potential
12 confounding by or interactions with other traffic-related pollutants.
13 Epidemiologic studies consistently demonstrate associations between increases in
14 ambient NC>2 concentration and increases in asthma hospital admissions and ED visits
15 among subjects of all ages and children (Section 5.2.2.4). The robustness of evidence is
16 demonstrated by associations found in studies conducted in diverse locations in the U.S.,
17 Canada, and Asia, including several multicity studies. These observations are coherent
18 with evidence in children and adults with asthma for increases in respiratory symptoms
19 (Section 5.2.2.3). the major reason for seeking medical treatment. NC>2 was associated
20 with the use or sale of asthma medication in adults with asthma but not children with
21 asthma. Individual epidemiologic studies examined multiple outcomes and lags of
22 exposure; however, a pattern of association was consistently observed with NO2, which
23 does not point to a higher probability of findings due to chance alone.
24 Although controlled human exposure studies do not provide strong evidence for NC>2
25 exposure inducing respiratory symptoms in adults with asthma (Section 5.2.2.3).
26 biological plausibility for effects of NC>2 on asthma exacerbation is provided by evidence
27 for NO2-induced increases in airway responsiveness in adults with asthma, particularly in
28 response to nonspecific challenge agents and exposures in which subjects did not exercise
29 (Section 5.2.2.1). Of all the health outcomes examined in controlled human exposure
30 studies of NO2, increased airway responsiveness was induced by the lowest NO2
31 concentrations, 100 ppb for 1 hour and 200-300 ppb for 30 minutes. Further, a
32 meta-analysis indicates a clinically relevant doubling reduction in provocative dose in
33 adults with asthma in response to NO2 exposure relative to air exposure. Increased airway
34 responsiveness can lead to poorer asthma control. Thus, the evidence for relatively low
35 NO2 exposures inducing clinically relevant increases in airway responsiveness in adults
36 with asthma provides key support that ambient exposures to NO2 can exacerbate asthma.
January 2015 5-123 DRAFT: Do Not Cite or Quote
-------
1 T-derived lymphocyte helper 2 (Th2)-mediated airway obstruction can lead to both an
2 increase in respiratory symptoms and decrease in lung function. Lung function
3 decrements, as measured by supervised spirometry, were observed in epidemiologic
4 studies in association with ambient or personal NC>2 concentrations (Section 5.2.2.2).
5 Most controlled human exposure studies showed no effect of ambient-relevant NO2
6 exposures (200-4,000 ppb for 30 minutes to 6 hours) on lung function but did not include
7 a challenge agent. Information delineating mechanisms underlying NCh-related lung
8 function decrements is limited. Neural reflexes do not appear to be involved for
9 ambient-relevant exposures; however, there is some evidence for mast cell degranulation
10 mediating changes in lung function (Section 4.3.2.2). Mast cell degranulation leads to
11 histamine release, indicating a role for allergic inflammation in mediating NC>2-induced
12 lung function decrements. Consistent with this mechanistic evidence, NCh-associated
13 decreases in lung function are found in populations with asthma that have a high
14 prevalence of atopy (53-84%) and groups of children with asthma not using
15 anti-inflammatory ICS (Hernandez-Cadena et al.. 2009; Liu et al., 2009). In the few
16 NO2-controlled human exposure studies of adolescents or adults with atopic asthma,
17 NC>2-induced decreases in lung function were found at 1,000 ppb NC>2 (3 hours) but not
18 120-400 ppb (for 30 minutes to 6 hours) (Jenkins et al.. 1999; Torres etal. 1995; Koenig
19 etal.. 1987).
20 In addition to supporting NC>2-related decreases in lung function in populations with
21 asthma and allergy, evidence for NC>2-induced allergic inflammation demonstrates that
22 NO2 can affect key events within the mode of action for asthma exacerbation
23 (Section 5.2.2.5). Not all experimental studies of adults with asthma or animal models of
24 allergic disease found effects; however, several found NCh-induced increases in
25 eosinophil number and activation of eosinophils and/or neutrophils following exposures
26 with and without an allergen challenge. Similar to airway responsiveness, allergic
27 inflammation was enhanced by lower NC>2 exposures than many other health effects
28 examined in experimental studies: 260 ppb NC>2 for 15-30 minutes or 400 ppb NCh for
29 6 hours. In controlled human exposure studies, NC>2 did not consistently increase other
30 indicators of pulmonary inflammation in adults with asthma, including those with atopy,
31 in the absence of allergen challenge. Epidemiologic studies generally did not find
32 NO2-associated changes in inflammatory cell counts in populations with asthma;
33 however, they did consistently indicate ambient or personal NCh-associated increases in
34 eNO (Figure 5-8 and Table 5-20). These findings are coherent with experimental
35 evidence for allergic inflammation because increases in eosinophils and neutrophils are
36 linked with NO production during an inflammatory response. The limited studies in
37 adults with asthma produced conflicting results, but the large body of findings in children
38 with asthma shows a consistent pattern of association across the various lags of exposure
39 and outcomes examined. Collectively, the evidence for NO2-related increases in allergic
January 2015 5-124 DRAFT: Do Not Cite or Quote
-------
1 inflammation provide biological plausibility for NCh-associated increases in respiratory
2 symptoms in children with atopy (Barraza-Villarreal et al.. 2008; Escamilla-Nunez et al..
3 2008). Such associations were not observed in adults with asthma and pollen allergy (Feo
4 Brito et al.. 2007).
5 The evidence in children with asthma is substantiated by several studies with strong
6 exposure assessment characterized by spatially aligning NO2 concentrations with
7 subjects' location(s). Respiratory symptoms, lung function decrements, and pulmonary
8 inflammation were associated with total and ambient personal NO2 exposures (Martins et
9 al.. 2012; Delfino et al.. 2008a: McCreanor et al.. 2007) and NO2 measured outside
10 schools (Greenwald et al., 2013; Zoraet al.. 2013; Sarnat et al.. 2012; Spira-Cohen et al..
11 2011; Holguin et al.. 2007). Given the high variability in NCh concentrations, these
12 spatially-aligned measures may better represent temporal variation in subjects' ambient
13 NC>2 exposures than area-wide central site concentrations (Sections 2.5.3 and 3.4.4).
14 Among studies that compared various exposure assessment methods, Delfino et al.
15 (2008a) found similar associations with total personal and central site NC>2, and Sarnat et
16 al. (2012) found stronger associations for school than central site NC>2 concentrations.
17 Observations that daily variation in central site ambient NC>2 was related to variation in
18 total personal NC>2 (r = 0.43) (Delfino et al.. 2008a) and that indoor home or school NC>2
19 concentrations were negligible (Martins et al.. 2012) provide additional support for a
20 relationship of asthma exacerbation with ambient NC>2.
21 A key uncertainty noted in the 2008 ISA for Oxides of Nitrogen was whether NO2 had an
22 effect independent of other traffic-related pollutants (U.S. EPA. 2008a). Epidemiologic
23 studies of asthma-related respiratory effects found associations with NO2 as well as with
24 the traffic-related pollutants CO, BC/EC, UFP, PM2 5, other PM constituents, and VOCs
25 (Figures 5-16 and 5-17). Among the studies that examined copollutant confounding, most
26 indicate an independent association with NO2. The predominant method for evaluation
27 was copollutant models, which have well-recognized limitations for distinguishing
28 independent pollutant associations (Section 5.1.2.2). However, several studies with strong
29 exposure assessment provide a sound basis for inferring an independent NO2 association,
30 when integrated with experimental evidence.
31 In populations with asthma, total and ambient personal NO2, NO2 measured outside or
32 650 m from children's schools, and NO2 measured on location of outdoor exposures were
33 associated with respiratory symptoms, decreased lung function, and pulmonary
34 inflammation with adjustment for BC/EC, UFP, OC, or PM2 5 (Martins etal. 2012; Lin et
35 al.. 2011; Delfino et al.. 2008a; McCreanor et al.. 2007; Delfino et al.. 2006)
36 (Figures 5-16 and 5-17). In a few cases, adjustment for UFP or VOCs attenuated the NO2
37 association with one outcome in a study but not another (Martins etal.. 2012; McCreanor
January 2015 5-125 DRAFT: Do Not Cite or Quote
-------
1 et al.. 2007), indicating the potential for confounding to differ by outcome. Among
2 children with asthma in El Paso, TX, school NC>2 was not associated with symptoms after
3 adjusting for BC (Zoraetal.. 2013). However, a copollutant model was not examined in
4 the group also with atopy, to whom the association with NC>2 was limited. Copollutant
5 associations adjusted for NO2 were robust in some cases (Lin et al.. 2011; Delfino et al..
6 2006) and attenuated in other cases (Martins et al.. 2012; Delfino et al.. 2008a). Thus, in
7 some studies, NC>2 appeared to confound associations for traffic-related copollutants. The
8 spatial alignment of NC>2 with subjects' location(s) may have reduced differences in
9 exposure measurement error between NC>2 and copollutants, thereby improving the
10 reliability of copollutant model results. A wide range of correlations were reported
11 between NCh and traffic-related copollutants (r = -0.42-0.75). Also improving inference
12 from copollutant model results are the low NCh-copollutant correlations found for
13 personal measurements (r = 0.20-0.30 for EC, OC, PM2 5 and -0.42 to 0.08 for benzene
14 and ethylbenzene) (Martins et al.. 2012; Delfino etal.. 2006).
15 Consistent with findings for NC>2 exposures in subjects' locations, copollutant models
16 based on central site concentrations indicate ambient NCh remains associated with
17 asthma-related hospital admissions, ED visits, symptoms, medication use, and lung
18 function with adjustment for CO, UFP, a source apportionment factor comprising EC and
19 various metals, PM2 5, or oxidative potential of PM2 5 extracts (Delfino etal.. 2013;
20 Iskandaretal..2Q12; Dales et al.. 2009; Gent et al.. 2009; Jalaludin et al.. 2008;
21 Villeneuve et al.. 2007; Delfino etal.. 2003; von Klot et al.. 2002). Several traffic-related
22 PM constituents are shown to induce oxidative stress (Section 5.1.2.1). Copollutant
23 adjustment also had effects on central site NO2 associations that differed by outcomes
24 within studies (Liu et al.. 2009; Andersen et al.. 2008a; von Klot etal.. 2002). Central site
25 NC>2 tended to show moderate correlations with traffic-related copollutants
26 (r = 0.28-0.56, but 0.66 for UFP and 0.71 for PM2 5), but differences in spatial
27 distributions may result in differential exposure measurement error for central site
28 concentrations of NC>2 and traffic-related copollutants. Such differences may influence
29 findings from Delfino et al. (2008a). where the association between ambient NCh and
30 lung function remained positive but was reduced with adjustment for personal PM2.5.
31 Also supporting an independent association for NC>2, some studies found associations
32 with NCh but not EC, OC, or PIVb 5 for school or personal measurements (Sarnat et al..
33 2012; Delfino et al.. 2008a; Holguin et al.. 2007) or CO or PM2 5 for central site
34 measurements (Lagorio et al.. 2006; Delfino et al.. 2003).
35 Epidemiologic evidence also indicates that NO2 associations with asthma-related effects
36 are independent of other pollutants and temporally varying factors (Tables 5-9. 5-12.
37 5-20). In most cases, NO2 associations were found with adjustment for SO2, PMio,
38 PMio-2.5, or O3 (Liu etal.. 2014a; Sarnat etal.. 2012; Mann et al.. 2010; Patel et al.. 2010;
January 2015 5-126 DRAFT: Do Not Cite or Quote
-------
1 Strickland etal.. 2010; Dales et al.. 2009; Qian et al.. 2009a; Jalaludin et al.. 2008;
2 Mortimer et al.. 2002). In some copollutant models, associations for both NO2 and
3 copollutant were attenuated (Martins et al.. 2012; Liu et al.. 2009; Qian et al.. 2009a). and
4 an independent or confounding effect was not distinguished for either NO2 or copollutant.
5 In exception, Samoli etal. (2011) indicated that the NO2 association with asthma ED
6 visits was confounded by SO2 or PMio but not vice versa. Most epidemiologic studies
7 found associations between NO2 and asthma-related effects with adjustment for potential
8 confounding by temperature, humidity, and season. As examined in fewer studies, NO2
9 associations were found with adjustment for day of week, smoking, and asthma
10 medication use.
11 A few studies of asthma-related respiratory effects examined combined effects of central
12 site NO2 and traffic-related copollutants. Limited information indicates increases in
13 asthma ED visits when ambient concentrations of NO2 and PM2 5, CO, or EC are high
14 (Gass etal.. 2014; Winquist et al.. 2014). but does not provide evidence of interactions
15 between NO2 and VOCs or CO (Schildcrout et al.. 2006; Delfino et al.. 2003). Limited
16 epidemiologic information indicates joint effects of NO2, Os, and SO2 (Winquist et al..
17 2014). but interactions are not strongly demonstrated in controlled human exposure
18 studies (Jenkins et al.. 1999; Devalia et al.. 1994; Hazucha et al.. 1994; Adams etal..
19 1987).
20 Another line of evidence supporting an independent effect of short-term NO2 exposure on
21 asthma exacerbation is the coherence of evidence between indoor and outdoor NO2.
22 Except for Greenwald et al. (2013). indoor home or school NO2 concentrations were
23 associated with respiratory symptoms and pulmonary inflammation among children with
24 asthma (Lu etal.. 2013; Sarnat et al.. 2012; Gillespie-Bennett et al.. 2011; Hansel et al..
25 2008). Sarnat et al. (2012) found that correlations between NO2 and copollutants differed
26 between the indoor and outdoor environments for BC, PM, and SO2, suggesting that NO2
27 may exist as part of different pollutant mixtures in the indoor and outdoor environments.
28 In summary, epidemiologic studies consistently indicate associations of short-term
29 increases in NO2 concentrations with asthma-related hospital admissions, ED visits,
30 symptoms, and pulmonary inflammation. A majority of these studies examine central site
31 NO2 concentrations and do not examine potential confounding by traffic-related
32 copollutants. Therefore, evidence for increased airway responsiveness and allergic
33 inflammation in experimental studies is key in demonstrating the independent effects of
34 NO2. That these effects are found with exposures to 100-400 ppb NO2 for 15 minutes to
35 6 hours substantiates the supposition that ambient-relevant NO2 exposures can exacerbate
36 asthma. Additional support is provided by associations found with NO2 exposures
37 assessed in subjects' locations, in copollutant models with another traffic-related
January 2015 5-127 DRAFT: Do Not Cite or Quote
-------
1 pollutant, and indoor NC>2. Not all respiratory outcomes are associated with NO2 or show
2 coherence between epidemiologic and controlled human exposure studies, particularly
3 lung function. Potential confounding has not been assessed for all correlated
4 traffic-related pollutants, and reliable methods are not available for simultaneous control
5 for multiple pollutants. However, the evidence integrated across clinical events and key
6 events within the mode of action characterizes biological pathways by which short-term
7 NO2 exposure may induce asthma exacerbation.
5.2.3 Allergy Exacerbation
8 The evidence from experimental studies for the effects of short-term NO2 and allergen
9 co-exposure on increasing allergic inflammation in adults with asthma and animal models
10 of allergic disease (Section 5.2.2.5) not only supports NC>2-related asthma exacerbation
11 but also indicates that NCh-induced allergy exacerbation may be biologically plausible.
12 Support also is provided by in vitro findings that NC>2 can increase the allergencity of
13 pollen (Cuinica et al.. 2014; Sousaet al.. 2012). Studies examining clinical indications of
14 allergy exacerbation have become available since the 2008 ISA for Oxides of Nitrogen
15 (U.S. EPA. 2008a). In contrast with asthma exacerbation, short-term NC>2 exposure is not
16 clearly related with clinical indications of allergy exacerbation.
17 Equivocal epidemiologic evidence in adults with allergies is indicated by associations of
18 ambient NC>2 with physician visits, allergic rhinitis, or nonspecific respiratory symptoms
19 that are either inconsistent across the lags of exposure examined (Villeneuve et al.,
20 2006b), negative, or positive but with wide 95% CIs (Annesi-Maesano et al.. 2012; Feo
21 Brito etal.. 2007). The latter results are based on a multipollutant model, which can
22 produce unreliable results (Annesi-Maesano et al.. 2012) (Table 5-21). Studies only
23 reported that ambient NC>2 exposure was assigned as ambient concentrations from one
24 central site or the average of multiple sites. And, it is uncertain whether the temporal
25 variability in these NC>2 metrics adequately represents the variability in ambient
26 concentrations across the study area or in subjects' ambient exposures. Similarly
27 inconclusive, a recent controlled human exposure study of adults with allergic asthma did
28 not find increases in allergic inflammation after NC>2 exposure and found increases in
29 respiratory symptoms only during exposure (Riedl et al.. 2012).
January 2015 5-128 DRAFT: Do Not Cite or Quote
-------
Table 5-21 Epidemiologic studies of allergy exacerbation.
Study
Population Examined and Methodological
Details
NO2 Metrics
Analyzed
Effect Estimate
(95% Cl)
Single-Pollutant
Model3
Copollutant
Examination
tCorreia-Deur et al. (2012)
Sao Paolo, Brazil, Apr-Jul 2004
n = 36, one test positive (see below); n =
28, three tests positive, ages 9-11 yr
Repeated measures. Daily supervised
spirometry for 15 school days. Number of
observations not reported. Recruitment from
school. 86% participation. Allergic
sensitization ascertained by skin prick test,
blood eosinophils, and serum IgE. GEE with
autoregressive correlation matrix adjusted for
date, school absence, temperature, humidity.
NO2-outdoor school % change PEF:
24-h avg, lag 0 day
Mean: 69.9 ppbb
75th: 84.5 ppbb
90th: 102ppbb
Group with 1 positive
test
-0.87% (-1.7, -0.04)
Group with 3 positive
tests
-0.30% (-1.7, 1.1)
All subjects, lag 0
With CO (r= 0.51)
-1.5% (-3.0, 0)
With SO2(r= 0.60)
-1.9% (-3.3, -0.37)
With PMio(r=0.59)
-0.75% (-4.4, 3.1)
With O3(r= 0.40)
-1.5% (-3.3, 0.38)
Associations for CO &
Os not altered by NO2
adjustment. SO2 &
PM-io attenuated.
tBarraza-Villarreal et al. (2008)
Mexico City, Mexico, Jun 2003-Jun 2005
n = 50, ages 6-14 yr, 72% with atopy
Repeated measures. Examined every
15 days for mean 22 weeks. Participation rate
not reported. 1,503 observations. Recruitment
from friends or schoolmates of subjects with
asthma. Clinical assessment of allergy.
Supervised spirometry. Linear mixed effects
model with random effect for subject and
adjusted for sex, BMI, temperature, ICS use,
time. Adjustment for outdoor activities,
smoking exposure, anti-allergy medication
use, season did not alter results.
tEscamilla-Nufiez et al. (2008)
Mexico City, Mexico, Jun 2003-Jun 2005
NO2-central site
8-h max NO2
Closest site, within
5 km of homes or
schools.
R = 0.21 for central
site and school
NO2.
Mean: 37.4 ppb
Max: 77.6 ppb
NO2-central site
1-h max NO2, lag
OR for cough:
lag 0-1 day avg
1.28(1.04, 1.57)
% change FEV-i: lag
day 1-4 avg
-0.64% (-2.1, 0.82)
OR for cough:
1.23(1.03, 1.47)
No copollutant model.
PM2.5 associated with
lung function and
cough.
Moderate correlation
\A/!4-l-i KI/~\
with NO2.
Pearson r = 61 .
Only multipollutant
model with Osand
DlV/ln i- o no \\i-reiri
n = 50, ages 6-14 yr, 79% with atopy
Part of same cohort as above. Participation
rate not reported. Linear mixed effects model
with random effect for subject and adjusted
for atopy, temperature, time, sex. Adjustment
for outdoor activities, smoking exposure,
season did not alter results.
0-1 day avg
Closest site, within
5 km of homes or
schools.
Mean: 68.6 ppb
Upper percentile:
NR
Moderate correlation
between PM2.sand
NO2. Pearson
r=0.62.
Villeneuve et al. (2006b)
Toronto, Canada, 1995-2000
n = 52,971 physician visits for allergic rhinitis,
ages 65 yr or older
Time-series analysis. GLM adjusted for
temperature, relative humidity, daily number
visits for influenza, allergen levels natural
spline for time trend.
NO2-central site
24-h avg, lag 0 day
Average of 9 city
sites.
Mean: 25.4 ppb
Max: 71.7 ppb
Quantitative results NR No copollutant model.
NO2 associated with
physician visits for
allergic rhinitis at lag
0 day. Negative or null
associations at lag
1 day to 6 days.
Association also
observed for SO2 but
not PM2.5 or CO.
Correlations with NO2
NR.
January 2015
5-129
DRAFT: Do Not Cite or Quote
-------
Table 5-21 (Continued): Epidemiologic studies of allergy exacerbation.
Study
Population Examined and Methodological
Details
NO2 Metrics
Analyzed
Effect Estimate
(95% Cl)
Single-Pollutant
Model3
Copollutant
Examination
tFeo Brito et al. (2007)
Ciudad Real and Puertollano, Spain
May-June 2000 or 2001
n = 137, ages NR, mild/moderate asthma and
pollen allergy
Repeated measures, 90% follow-up
participation. Daily symptom diaries. Number
of observations not reported. Recruitment
from allergy clinics. Clinical assessment of
allergy. Poisson regression adjusted only for
linear and quadratic terms for season.
NO2-central site
24-h avg
4 sites in
Puertollano, 1
mobile site in
Ciudad Real
Mean and Max
Ciudad Real: 17.4b,
35.6b
Puertollano: 29.5b,
100b
% change in
symptoms:
Ciudad Real, lag day 4
4.75% (-5.75, 16.4)
Puertollano, lag day 3
-3.00% (-9.55, 4.03)
No copollutant model.
PM-io, SO2, Os
associated with
symptoms only in
Puertollano. Moderate
correlation with NO2.
r=0.67, 0.36, 0.36.
Pollen associated
with symptoms only in
Ciudad Real, rwith
NO2 = -0.10.
tAnnesi-Maesano et al. (2012)
Multiple metropolitan locations, France,
May-Aug 2004
n = 3,708 with severe allergic rhinitis, ages
6 yr and older, 82% adults
Cross-sectional. Recruitment from physicians'
offices. No information on participation rate.
Clinical assessment of allergy and symptom
severity. Multilevel model adjusted for age,
date of physician visit, asthma status, postal
code. Did not consider confounding by
meteorology or SES.
NO2-central site
24-h avg, lag day 1
Site in postal code
of home.
Mean: 9.9b
Max: 38.9b
NR
Only multipollutant
model analyzed with
SO2, 03, PM-io, pollen.
Correlation only
reported for pollen,
r=-0.12.
Note: More informative studies in terms of the exposure assessment method and potential confounding considered are presented
first.
GEE = generalized estimating equations, BMI = body mass index, ICS = inhaled corticosteroid, NR = not reported, OR = odds
ratio, Cl = confidence interval, CO = carbon monoxide, FEVi = forced expiratory volume in 1 second, GLM = generalized linear
model, NO2 = nitrogen dioxide, O3 = ozone, PEF = peak expiratory flow, PM = particulate matter, SES = socioeconomic status,
SO2 = sulfur dioxide.
aEffect estimates are standardized to a 20 ppb for 24-h avg NO2.
Concentrations converted from |jg/m3 to ppb using the conversion factor of 0.532 assuming standard temperature (25°C) and
pressure (1 atm).
fStudies published since the 2008 ISA for Oxides of Nitrogen.
1
2
3
4
5
6
7
8
9
10
11
In children with allergies, increases in ambient NC>2 were associated with decreases in
lung function (Correia-Deur et al.. 2012; Barraza-Villarreal et al.. 2008). with an
association with cough found in a cohort in Mexico City (Barraza-Villarreal et al., 2008;
Escamilla-Nufiez et al.. 2008). Strengths of these studies include the clinical assessment
of allergy and the supervised measurement of lung function. Although not specific to
allergy exacerbation, lung function can decrease during an allergy exacerbation due to
airway obstruction caused by Th2 cytokine-mediated inflammation. The studies in
children aimed to account for heterogeneity in ambient NC>2 concentrations. In one
cohort, exposures were assigned from sites within 5 km of children's home or school, but
a Pearson correlation of r = 0.21 between school and central site NO2 indicates the
variability at the central site may not represent the variability in the subjects' locations
January 2015
5-130
DRAFT: Do Not Cite or Quote
-------
1 (Barraza-Villarreal et al., 2008). Another study examined NO2 from a central site in the
2 backyard of the subjects' school (Correia-Deur et al.. 2012). providing a stronger basis
3 for inference of NC>2 effects. Increases in NO2 lagged 2 hours, from the same day, and
4 averaged over 3 days were associated with decreases in PEF. However, counter to
5 expectation, associations were observed for the 36 children identified as having atopy
6 with a less stringent definition (one positive test among skin prick test, serum IgE, or
7 blood eosinophils: -0.87% [95% CI: -1.7, -0.04] per 20-ppb increase in lag 0 day NO2),
8 not the 28 children with atopy defined more stringently (all three tests positive: -0.30%
9 [95% CI:-1.7, 1.1]).
10 Respiratory effects in populations with allergy were associated with other traffic-related
11 pollutants such as CO and PM2 5 as well as with PMio and 862, so it is unclear whether
12 the supporting epidemiologic evidence represents an independent effect of NC>2. Only
13 Correia-Deur et al. (2012) examined copollutant models. In analyses combining children
14 with and without allergies, NO2 remained associated with PEF with adjustment for CO,
15 which also was measured at school (Table 5-21). NO2 also remained associated with PEF
16 with adjustment for school SO2 concentrations. However, both NO2 and PMio (r = 0.60)
17 associations were attenuated when adjusted for each other, and a confounding or
18 independent effect cannot be distinguished for either pollutant. Os potentially could
19 confound NO2 effects in the warm season, and interactions between Os and allergens
20 have been reported (U.S. EPA. 2013a). However, NO2 was associated with PEF after
21 adjustment for O3 (r = 0.40) (Table 5-21) (Correia-Deur et al.. 2012).
22 In summary, the evidence does not clearly indicate whether NO2 exposure independently
23 induces allergy exacerbation. While there is evidence for effects on key events in the
24 mode of action, the limited evidence for effects on clinical events related to allergy
25 exacerbation is inconclusive. Further, in the limited analysis of copollutants, there is
26 evidence for an effect of NO2 on lung function decrements independent of CO or Os
27 measured at children's schools but uncertainty regarding confounding by PM2 5, PMio,
28 and the array of other traffic-related pollutants, which were not examined (Table 5-1).
5.2.4 Exacerbation of Chronic Obstructive Pulmonary Disease
29 COPD is characterized by deterioration of lung tissue and airflow limitation. In
30 exacerbation of COPD, episodes of reduced airflow, which can be indicated by decreases
31 in lung function, can lead to symptoms such as cough, sputum production, and shortness
32 of breath. Severe exacerbation can lead to hospital admissions or ED visits. This
33 spectrum of outcomes comprises the majority of investigations of the effects of
34 short-term NO2 exposure on COPD exacerbation, and as described in the sections that
January 2015 5-131 DRAFT: Do Not Cite or Quote
-------
1 follow, the consistency of findings from previous and recent studies varies among
2 outcomes. As described at the end of the section (Section 5.2.4.3). limited recent
3 information does not show NO2-related increases in pulmonary inflammation in adults
4 with COPD. Pulmonary inflammation is a key early event in COPD exacerbation,
5 mediating narrowing of the airways and reducing airflow.
5.2.4.1 Lung Function Changes and Respiratory Symptoms in
Adults with Chronic Obstructive Pulmonary Disease
6 Evidence does not clearly indicate a relationship for NC>2 exposure with changes in lung
7 function or respiratory symptoms in adults with COPD. Evidence is inconsistent in both
8 controlled human exposure and epidemiologic panel studies, many of which examine
9 both respiratory symptoms and lung function. Most of these studies were reviewed in the
10 2008 ISA for Oxides of Nitrogen (U.S. EPA. 2008a). and the only recent study, which is
11 epidemiologic, does not support associations for ambient NO2 concentrations with either
12 respiratory symptoms or lung function decrements in adults with COPD.
Lung Function Changes
13 Epidemiologic studies recruited adults with COPD from clinics, and the nonrandom
14 selection of the general population may produce study populations less representative of
15 the COPD population. NO2 exposures were assessed primarily from central site
16 measurements of 24-h avg NO2, and results are equally inconsistent for NO2 exposures
17 assigned from one site or averaged from multiple city sites (Table 5-22). Most previous
18 and recent epidemiologic studies in adults with COPD assessed lung function with
19 unsupervised home measurements, and associations with ambient NO2 concentrations
20 were inconsistent among the various lung function parameters (e.g., FEVi, PEF) or NO2
21 exposure lags (0-, 1-, 2-, or 2- to 7-day avg) examined (Peacock et al.. 2011; Silkoff et
22 al.. 2005; Higgins et al.. 1995). Lagorio et al. (2006) found an association between
23 ambient NO2 concentrations and FEVi (Table 5-22). with similar effects estimated for
24 adults with COPD and asthma.
January 2015 5-132 DRAFT: Do Not Cite or Quote
-------
Table 5-22 Epidemiologic panel studies of adults with chronic obstructive
pulmonary disease (COPD).
Study
Population Examined and
Methodological Details
Oxide of Nitrogen
Metrics Analyzed
Effect Estimate
(95% Cl)
Single-Pollutant
Model3
Copollutant
Examination
tPeacocketal. (2011)
London, U.K., Oct 1995-Oct 1997
n = 28-94, ages 40-83 yr
Repeated measures. Home PEF.
Examined daily for 21-709 days.
Recruitment from outpatient clinic.
75% follow-up. GEE adjusted for
temperature, season. Lung function
measures adjusted for indoor
temperature and time spent outdoors.
NO2-central site
1-h max NO2, lag 1 day
1 city site
Mean: 51.4 ppb
75th: 56 ppb
PEF:
0.17% (0.03, 0.32)
PEF >20% below
predicted:
OR: 1.0(0.86, 1.2)
Symptomatic fall in
PEF:
OR: 1.1 (0.97, 1.3)
Symptomatic fall in PEF
WithBS: 1.1 (0.84, 1.3)
With PMio:
0.97(0.81, 1.2)
Correlations NR. 95%
Cl forBS also
increased. No change in
OR for PMio with NO2
adjustment.
Silkoffetal. (2005)
Denver, CO, Winters 1999-2000 and
2000-2001
n = 34 with COPD, mean age 66 yr,
75% severe COPD
Repeated measures. Home PEF.
Recruitment from outpatient clinics,
research registries, advertisements.
93-96% diaries completed. Mixed
effects model with random effect for
subjects and adjusted for temperature,
relative humidity, barometric pressure.
NO2-central site
24-h avg, lag 0, 1, 2
days
1 city site
Means:
1999-2000: 16 ppb
2000-2001: 29 ppb
75th and Max:
1999-2000: 30, 54 ppb
2000-2001: 36, 54 ppb
No quantitative data.
Negative, positive, and
null associations with
symptoms across NO2
lags.
No copollutant model.
Mixed positive,
negative, null
associations for PM2.5,
PMio, Os.
Desqueyrouxetal. (2002)
Paris, France, Oct 1995-Mar 1996,
Apr-Sept 1996
n = 39, severe COPD, mean age 67 yr
Repeated measures. Recruitment from
physicians' offices. No information on
participation. GEE adjusted for FEV-i,
smoking, CO2 pressure, oxygen
treatment, dyspnea, temperature,
humidity, season, holiday.
NO2-central site
24-h avg, lag 1-5 day
avg
Average of 15 city sites
Means for Periods 1 & 2
31.4,26.1 ppbb
Max for Periods 1 & 2
68.1, 56.4 ppbb
Physician visits for
COPD exacerbation
OR: 0.76(0.28,2.10)
with Os:
0.47(0.02,9.45)
Os association robust to
NO2 adjustment.
Correlations not
reported.
SO2 and PMio not
associated with COPD.
Lagorioetal. (2006)
Rome, Italy, May-June, Nov-Dec
1999
n = 11, ages 40-64 yr, nonsmokers
Repeated measures. Supervised
spirometry. Examined 2/week for two
1-mo periods. Mean 9, 15 observations
per subject. Recruitment from
outpatient clinic. Participation rate NR.
GEE adjusted for season, temperature,
humidity, beta-agonist use.
NO2-central site
24-h avg, lag 0 day
Average of 5 city sites
within 2 km of subjects'
census tracts.
Mean: 37.6 ppbb
Max: 54.3 ppbb
% predicted FEV-i:
-2.3 (-3.6,-1.0)
No copollutant model.
Lung function
associated with PM2.5,
PMio. Moderate
correlation with NO2.
Spearman r= 0.43 for
PM2.5, 0.45 for PMio.
January 2015
5-133
DRAFT: Do Not Cite or Quote
-------
Table 5-22 (Continued): Epidemiologic panel studies of adults with chronic
obstructive pulmonary disease (COPD).
Study
Population Examined and
Methodological Details
Oxide of Nitrogen
Metrics Analyzed
Effect Estimate
(95% Cl)
Single-Pollutant
Model3
Copollutant
Examination
Harreetal. (1997)
Christchurch, New Zealand, June-Aug
1994
n = 40, ages 55-83 yr, nonsmokers
Repeated measures. Home PEF.
Recruitment from doctors' offices,
COPD support group, advertisements.
66% participation. Loglinear model
adjusted for day of study, temperature,
wind speed, CO, PM-io, SO2.
NO2-central site
24-h avg, lag 1 day
#siteNR.
Concentrations NR
PEF:
-0.72% (-1.5, 0.07)
Only multipollutant
model analyzed.
PM-io, CO, SO2 not
associated with PEF.
tBruske (2014); Bruske et al. (2010)
Erfurt, Germany, Oct 2001 -May 2002
n = 38, ages 35-78 yr, all male, 53%
also with asthma
Repeated measures. Examined 2/mo
for 6 mo. 381 observations after
excluding concurrent fever or infection.
Method of recruitment and COPD
assessment and participation rate NR.
Additive mixed models with random
intercept for subject and adjusted for
infection/antibiotic use in previous
2 weeks, long-term time trend,
temperature, humidity as linear terms
or penalized splines. Also evaluated
confounding by barometric pressure
and corticosteroid use.
NO2-central site
24-h avg, lag 0 0-23 h
before blood collection
1 site 3.5 km from
subjects' homes.
Mean: 13.5 ppbb
75th: 16.6 ppbb
NO-central site
Mean: 10.8 ppbb
75th: 14.0 ppbb
PMN:
-8.0% (-18, 3.1)
Lymphocytes:
8.4% (-5.0, 24)
PMN:
-0.80% (-10, 9.9)
Lymphocytes:
13% (-1.9, 23)
NO2 and NO reported
not to be associated
with eosinophils. No
quantitative results.
PMN with UFP:
7.3% (-14, 34).
Lymphocytes with UFP:
8.4% (-7.2, 27)
CO associated with
lymphocytes.
NO2 highly correlated
with UFP and CO.
Spearman r= 0.66,
0.78.
No copollutant model for
NO.
Note: More informative studies in terms of the outcome examined, exposure assessment method and potential confounding
considered are presented first.
PEF = peak expiratory flow, GEE = generalized estimating equations, Cl = confidence interval, CO = carbon monoxide,
CO2 = carbon dioxide, COPD = chronic obstructive pulmonary disease, FE\A = forced expiratory volume in 1 second, NO = nitric
oxide, NO2 = nitrogen dioxide, NR = not reported, O3 = ozone, OR = odds ratio, PM = particulate matter,
PMN = polymorphonuclear cells, SO2 = sulfur dioxide, UFP = ultrafine particles.
aEffect estimates were standardized to a 20-ppb increase in 24-h avg NO2 or a 30-ppb increase 1-h max NO2.
bConcentrations converted from |jg/m3 to ppb using the conversion factor of 0.532 assuming standard temperature (25°C) and
pressure (1 atm).
fStudies published since the 2008 ISA for Oxides of Nitrogen.
1
2
o
3
4
5
6
7
In addition to the inconsistent evidence for changes in lung function in adults with
COPD, there is uncertainty regarding an independent association of NCh from that of
copollutants. Studies did not examine a broad range of traffic-related copollutants.
Lagorio et al. (2006) found FEVi decrements in association with NC>2 but not PIVb 5,
which was moderately correlated with NC>2. Only Peacock et al. (2011) conducted
copollutant modeling, and the NO2-PEF effect estimate was attenuated with adjustment
for BS. In contrast, the effect estimate for BS was relatively unchanged with adjustment
for NO2. With respect to PMio, no association was found with FEVi (Lagorio et al..
January 2015
5-134
DRAFT: Do Not Cite or Quote
-------
1 2006). or the NO2 association with PEF was attenuated with adjustment for PMio
2 (Peacock etal.. 2011).
3 Similar to the epidemiologic studies, the controlled human exposure studies, which were
4 reviewed in the 2008 ISA for Oxides of Nitrogen (U.S. EPA. 2008a) and examined older
5 adults diagnosed with COPD, provide mixed results regarding decrements in lung
6 function with NO2 exposure. As presented in Table 5-23. exposures ranged from 200 to
7 4,000 ppb NO2 for 75 minutes to 6 hours, and most studies incorporated exercise in the
8 exposure period to assess lung function during various physiological conditions. Morrow
9 et al. (1992) exposed older adults diagnosed with COPD to 300 ppb NO2 for 3 hours and
10 reported consistent reductions in FVC that reached statistical significance at the end of
11 exposure, while Vagaggini et al. (1996) reported decreases in FEVi in subjects with
12 COPD exposed to 300 ppb NO2 for 1 hour. In contrast, Linn etal. (1985a) and Gong et
13 al. (2005) reported that exposure to 400-2,000 ppb for 1-2 hours had no effect on lung
14 function in adults with COPD. Furthermore, Gong et al. (2005) did not find NO2 to
15 enhance respiratory effects of PM exposures.
Respiratory Symptoms
16 The limitations and uncertainties described above for the evidence base relating NO2
17 exposure to lung function changes in adults with COPD largely apply to the evidence
18 base for respiratory symptoms. Epidemiologic panel studies examining adults with
19 COPD, many conducted in Europe, (Table 5-22) either found no association (Peacock et
20 al.. 2011; Desqueyroux et al.. 2002) between ambient NO2 and respiratory symptoms or
21 inconsistent associations across the lags of exposure or range of outcomes examined
22 (Silkoff etal.. 2005; Harre etal.. 1997). Most epidemiologic panel studies recruited
23 subjects from outpatient clinics or doctors' offices. Results were equally inconsistent for
24 symptoms such as cough, wheeze, dyspnea, total symptoms, and medication use
25 (Table 5-22). No pattern of association was found for either 24-h avg or 1-h max NO2 or
26 for a particular lag day of exposure examined (0, 1, or longer). Most of these studies
27 assigned exposures from a single central site, but associations with symptoms and
28 medication were inconsistent for NO2 assigned from the closest site (Desqueyroux et al..
29 2002) or site within 5 km (Harre et al.. 1997).
30 In the studies that found associations with specific symptoms or lags of NO2, associations
31 also were found with traffic-related pollutants such as PM2 5, BS, and CO (Peacock et al..
32 2011; Silkoff et al.. 2005; Harre etal.. 1997). Among adults in New Zealand, an increase
33 in 24-h avg NO2 was associated with an increase in inhaler use in a multipollutant model
34 with CO, PMio, and SO2 (Harre etal.. 1997). which has limited implications because of
35 multicollinearity. A recent study of adults in London, U.K. found that associations
January 2015 5-135 DRAFT: Do Not Cite or Quote
-------
1 between lag day 1 of 1-h max NCh and dyspnea were null with adjustment for BS or
2 PMio (Peacock et al.. 2011). Thus, in the few associations found between increases in
3 ambient NC>2 concentration and increases in symptoms or medication among adults with
4 COPD, there is uncertainty as to whether ambient NO2 has effects independent of other
5 traffic-related pollutants.
6 The equally inconsistent findings from controlled human exposure studies (Table 5-23)
7 do not address uncertainties in the epidemiologic evidence base. Some studies reported
8 no change in symptoms, as measured by symptom score, in adults with COPD (Gong et
9 al.. 2005; Morrow et al.. 1992). though some studies reported small, but statistically
10 significant increases in symptom scores during NO2 exposures of 300-2,000 ppb for
11 1 hour with exercise (Vagaggini et al.. 1996; Linn etal.. 1985a).
January 2015 5-136 DRAFT: Do Not Cite or Quote
-------
Table 5-23 Controlled human exposure studies of respiratory symptoms in
adults with chronic obstructive pulmonary disease (COPD).
Disease Status; n, Sex;
Study Age (mean ± SD)
Exposure Details
(Concentration; Duration)
Time of Outcome Assessment
Gong et al.
(2005)
Healthy: n = 2 M, 4 F;
68 ± 11 yr
COPD: n = 9 M, 9 F;
72 ± 7 yr
(1)400ppbNO2for2h
(2) 200 |jg/m3 CAPs for 2 h
(3) 400 ppb NO2 + 200 ug/m3 CAPs
for2h
(1-3) Exercise 15 min on/15 min off
at VE = ~2 times resting
Pulmonary function tests before
and immediately after exposure
and 4 h and 22-h post-exposure.
Symptoms before, during, and after
exposure.
Linn et al. COPD
(1985a) n = 13M, 9F
(1 never smoker,
13 former smokers, and
8 current smokers);
60.8 ± 6.9 yr
500, 1,000, and 2,000 ppb for 1 h;
Exercise 15 min on/15 min off
VE = 16L/min
Pulmonary function tests before,
during, and after exposure.
Symptoms before, during,
immediately after, 1 day after, and
1 week after exposure.
Morrow et
al. (1992)
Healthy: n = 10M, 10 F
(13 never smokers,
4 former smokers,
3 current smokers)
COPD: n = 13M, 7F
(14 current smokers,
6 former smokers);
59.9 ± 7.0 yr
300 ppb for 4 h;
Three 7-min periods of exercise at
VE = ~4 times resting
Pulmonary function tests before,
during, and after exposure and
24-h post-exposure.
Symptoms before, during, and after
exposure and 24-h post-exposure.
Vaqaqqini Healthy: n = 7 M; 34 ± 5 yr
etal. (1996) COPD: n = 7 M; 58 ± 12 yr
300 ppb for 1 h;
Exercise at VE = 25 L/min
Pulmonary function tests before
and 2 h after exposure.
Symptoms before and 2 h after
exposure.
CAPs = concentrated ambient particles, COPD = chronic obstructive pulmonary disease, F = female, M = male, NO2 = nitrogen
dioxide, SD = standard deviation.
5.2.4.2 Hospital Admissions and Emergency Department Visits
for Chronic Obstructive Pulmonary Disease
i
2
3
4
5
6
7
In contrast with the inconsistent evidence for the effects of short-term NC>2 exposure on
lung function changes and respiratory symptoms in adults with COPD (Section 5.2.4.1).
epidemiologic evidence is consistent for NCh-related increases in hospital admissions and
ED visits for COPD. The few studies of COPD hospital admissions or ED visits
evaluated in the 2008 ISA for Oxides of Nitrogen provided initial evidence of a positive
association between short-term NO2 exposures and COPD hospital admissions and ED
visits, with more studies focusing on hospital admissions (Figure 5-9 and Table 5-25).
However, these studies were more limited in their evaluation of potential confounders
January 2015
5-137
DRAFT: Do Not Cite or Quote
-------
1 and other factors that may modify the relationships of NO2 exposure with COPD hospital
2 admissions and ED visits. Consistent with the 2008 ISA for Oxides of Nitrogen, a few
3 recent studies have examined COPD hospital admissions and ED visits and generally add
4 to the initial evidence of a positive association observed in the 2008 ISA for Oxides of
5 Nitrogen. The air quality characteristics of the study cities and the exposure assignment
6 approach used in each study evaluated in this section are presented in Table 5-24. Other
7 recent studies of COPD hospital admissions and ED visits are not the focus of this
8 evaluation, as detailed in Section 5.2.2.4. but the full list of these studies and study
9 details, can be found in Supplemental Table S5-3 (U.S. EPA. 2014h).
January 2015 5-138 DRAFT: Do Not Cite or Quote
-------
Table 5-24 Mean and upper percentile concentrations of nitrogen dioxide (NO2) in studies of hospital admission
and emergency department visits for chronic obstructive pulmonary disease.
Study
Location
Years
Exposure Assignment Metric
Mean
Concentration Upper Percentile of
(ppb) Concentrations (ppb)
Copollutant Examination
Hospital Admissions
fFaustini et al.
(2013)
Ko et al. (2007a)
6 Italian cities
(2001-2005)
Hong Kong
(2000-2004)
Average of NO2 24-h avg
concentrations over all
monitors within each city.
Number of NO2 monitors in
each city ranged from 1-5a
Average of NO2 24-h avg
concentrations across
24.1-34.6 NR
27.2 75th: 34.0
Max: 83.8
Correlations (r), across
cities:
PMio: 0.22-0.79
Copollutant models: PMio
Correlations (r):
PlVh.s: 0.44
PMio: 0.40
SO2: 0.66
O3: 0.34
Copollutant models: none
tQiuetal. (201 3a)
tWonq et al. (2009)
Hong Kong
(1998-2007)
Hong Kong
(1996-2002)
Of 14 monitors, average NO2 24-h avg 30.9 NR
based on data from
10 monitors. 3 monitors sited
near roads and 1 monitor on a
remote island were excluded.
Average of NO2 24-h av^ 31.2 75th: 37.0
concentrations across ^ax- QQ 4
8 monitors.
Correlations (r): NR
Copollutant models: PMio
Correlations (r): NR
Copollutant models: none
January 2015
5-139
DRAFT: Do Not Cite or Quote
-------
Table 5-24 (Continued): Mean and upper percentile concentrations of nitrogen dioxide (NO2) in studies of hospital
admission and emergency department visits for chronic obstructive pulmonary disease.
Study
Location
Years
Exposure Assignment
Mean
Concentration
Metric (ppb)
Upper Percentile of
Concentrations (ppb)
Copollutant Examination
Emergency Department Visits
tStieb et al. (2009) 7 Canadian
cities
(1992-2003)
tArbex et al. (2009) Sao Paulo,
Brazil
(2001-2003)
Average NO2 concentrations 24-h avg 9.3-22.7
from all monitors in each city.
Number of NO2 monitors in
each city ranged from 1-14.
Average of NO2 1-h max 63.0
concentrations across
4 monitors.
75th: 12.3-27.6 Correlations (r) only reported
by city and season.
Copollutant models: none
75th: 78.6 Correlations (r):
Max: 204.6 PMm 0.60
SO2: 0.63
CO: 0.56
Copollutant models: none
CO = carbon monoxide, NO2 = nitrogen dioxide, NR = not reported, O3 = ozone, PM = particulate matter, SO2 = sulfur dioxide.
aMonitoring information obtained from Colais et al. (2012).
fStudies published since the 2008 ISA for Oxides of Nitrogen.
January 2015
5-140
DRAFT: Do Not Cite or Quote
-------
Hospital Admissions
1 Consistent with the 2008 ISA for Oxides of Nitrogen, relatively few recent studies have
2 focused on the outcome of COPD hospital admissions, but these studies build upon the
3 initial evidence of a positive association (Figure 5-9). Faustini et al. (2013) examined the
4 relationship between short-term air pollution exposures and respiratory-related hospital
5 admissions, including COPD, specifically on the adult population (i.e., individuals
6 35 years of age and older) in six Italian cities. In a time-series analysis, the authors
7 examined the lag structure of associations through single-day lags as well as cumulative
8 lags using cubic polynomial distributed lags to identify whether the NO2 effect on
9 respiratory-related hospital admissions was immediate (lag 0, lag 0-1 days), delayed (lag
10 2-5 days), or prolonged (lag 0-3, 0-5 days). For COPD hospital admissions, the authors
11 observed stronger evidence for immediate (lag 0: 4.6% [95% CI: 0.64, 8.6] for a 20-ppb
12 increase in 24-h avg NO2 concentrations) NO2 effects on COPD hospital admissions.
13 Smaller associations were observed when examining prolonged effects, (3.3% for lag
14 0-3 days and 3.1% for lag 0-5 days). There was no evidence for delayed effects (lag
15 2-5 days). In a copollutant model with PMio at lag 0, the association between NO2 and
16 COPD hospital admissions remained relatively unchanged compared to the
17 single-pollutant model results (3.9% [95% CI: -1.7, 9.8]).
18 In a study conducted in Hong Kong from 2000-2004, Ko et al. (2007a) also examined
19 the lag structure of associations between short-term air pollution exposures and COPD
20 hospital admissions. In analyses of both single-day lags and multiday averages, Ko et al.
21 (2007a) observed the largest magnitude of an association at lags ranging from 0-3 to
22 0-5 days (10.1% [95% CI: 8.5, 12.2] for a 20-ppb increase in 24-h avg NO2
23 concentrations at both 0-3 and 0-5 day lags). These associations are larger in magnitude
24 than those reported by Oiu et al. (2013a) at lag 0-3 (4.7% [95% CI: 3.3, 6.2] for a 20 ppb
25 increase in 24-h avg NO2 concentrations) for a study also conducted in Hong Kong, but
26 for a longer duration (1998-2007). Although Ko et al. (2007a) reported associations
27 larger in magnitude for multiday averages, the authors also observed a positive
28 association across single-day lags, with lag 0 having one of the stronger associations
29 (3.4% [95% CI: 1.9, 5.0]), which is of similar magnitude to the lag 0 effect observed in
30 Faustini et al. (2013). Ko et al. (2007a) only examined the potential confounding effects
31 of copollutants through the use of three- and four-pollutant models, which are difficult to
32 interpret. In comparisons of the single-pollutant results for NO2 and the other pollutants
33 examined (Os, PM2s, and PMio), similar patterns of associations were observed across
34 pollutants. Additionally, Ko et al. (2007a) examined whether there was evidence of
35 seasonal differences in NO2-COPD hospital admission associations. When using the
36 warm season as the referent, the authors reported evidence of larger associations in the
January 2015 5-141 DRAFT: Do Not Cite or Quote
-------
1 cold season (i.e., December to March). These results are consistent with the results of Ko
2 et al. (20071)) for asthma (Section 5.2.2.4) and support potential differences in seasonal
3 associations by geographic location.
4 In addition to examining the association between short-term air pollution exposures and
5 COPD hospital admissions, Qiu et al. (2013a) also examined whether air pollution
6 associations with COPD hospital admissions were modified by the interaction between
7 season and humidity. In models stratifying by both season (warm: May-October; cold:
8 November-April) and humidity (high or humid: >80%; low or dry: <80%) the authors
9 found larger NO2 associations in the cool season and high humidity days (5.6 and 6.3%,
10 respectively) compared to the warm season and low humidity days (3.8 and 4.6%,
11 respectively) for a 20-ppb increase in 24-h avg NO2 concentrations at lag 0-3 days.
12 When examining the j oint effect of season and humidity, Qiuetal. (2013 a) found that the
13 magnitude of the association was larger when season and humidity were considered
14 together. Specifically, the largest associations were observed for the combination of
15 warm season and humid days 7.3% [95% CI: 3.7, 11.1]; lag 0-3) and cool season and dry
16 days (9.3% [95% CI: 6.2, 12.5]; lag 0-3). When examining a copollutant model only
17 with PMio, across all combinations of models that examined the effect of season and
18 humidity on NO2-COPD hospital admissions, the associations were attenuated and in
19 some cases null, specifically for the combination of warm season and dry days and cool
20 season and humid days. These results further highlight the different seasonal patterns in
21 NO2 associations that have been reported across different geographic areas as well as the
22 potential influence of different weather conditions on NO2-related health effects.
Emergency Department Visits
23 As in the 2008 ISA for Oxides of Nitrogen, relatively few studies have examined the
24 relationship between short-term NO2 exposures and ED visits, compared to hospital
25 admissions. In the seven Canadian cities discussed previously, consistent with the asthma
26 ED visits results, Stieb et al. (2009) did not find evidence of associations between
27 24-h avg NO2 and COPD ED visits at individual lags ranging from 0 (0.1% [95% CI:
28 -6.1, 6.8] for a 20-ppb increase in 24-h ave NO2) to 2 (-5.2% [95% CI: -12.4, 2.7]) days.
29 Additionally, there was no evidence of consistent associations between any pollutant and
30 COPD ED visits at sub-daily time scales (i.e., 3-h avg of ED visits vs. 3-h avg pollutant
31 concentrations).
32 Arbex et al. (2009) also examined the association between COPD and several ambient air
33 pollutants, including NO2, in a single-city study conducted in Sao Paulo, Brazil, for
34 individuals over age 40 years. Associations between NO2 exposure and COPD ED visits
35 were examined in both single-day lags (0 to 6 days) and a polynomial distributed lag
January 2015 5-142 DRAFT: Do Not Cite or Quote
-------
1 model (0-6 days). However, for NO2, only those results that were statistically significant
2 were presented, that is, for individuals 65 years of age and older for lag 5 (4.3% [95% CI:
3 0.5, 8.3] for a 20-ppb increase in 24-h avg NO2 concentrations) and a distributed lag of
4 0-5 days (9.6% [95% CI: 0.2, 19.9]). The authors did not analyze copollutant models but
5 reported moderate correlations between NO2 and PMio (r = 0.60), SO2 (r = 0.63), and CO
6 (r = 0.56).
Summary of Chronic Obstructive Pulmonary Disease Hospital Admissions
and Emergency Department Visits
7 In combination with those studies evaluated in the 2008 ISA for Oxides of Nitrogen,
8 recent studies add to the growing body of literature that has examined the association
9 between short-term NO2 exposures and COPD hospital admissions and ED visits.
10 Overall, these studies have reported consistent positive associations with evidence of
11 NO2-COPD hospital admissions and ED visits occurring immediately (lag 0) as well as a
12 few days after exposure (average of lags up to 5 days) (Figure 5-9). However, caution
13 should be used in inferring the independent effects of NO2 exposure due to the relative
14 sparseness of copollutant model analyses as well as the high correlation often observed
15 between NO2 and other traffic-related pollutants (e.g., CO, PlVfc 5). Additionally, studies
16 that have focused on COPD hospital admissions and ED visits have not thoroughly
17 examined potential seasonal differences in associations; however, initial evidence
18 suggests that the combination of season and weather conditions, such as humidity, may
19 have a larger effect on NO2-COPD hospital admission associations than either
20 individually. Additionally, these studies have provided limited information on individual-
21 or population-level factors that could modify the NO2-hospital-admission or ED visit
22 relationship, or the shape of the C-R relationship.
January 2015 5-143 DRAFT: Do Not Cite or Quote
-------
Study
Ko et al. (2007)
Qiuetal. (2013)
Wong etal. (2009)
Yang et al. (2005)
Moolgavkar (2003)
Faustini etal. (2013)
Peel etal. (2005)
Stieb et al. (2009)
Arbexetal. (2009)
Atlanta, GA
7 Canadian cities
Sao Paulo, Brazil 65+
Age
All
All
All
65+
Lag
0-3
0-3
0-1
0
Hospital Admissions
— •
-•-
ED Visits
k
V
-4.0 -2.0 0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0 18.0 20.1
% Increase
Note: Black circles = U.S. and Canadian studies evaluated in the 2008 Integrated Science Assessment for Oxides of Nitrogen;
Red = recent studies. Effect estimates are standardized to a 20-ppb increase in 24-h avg nitrogen dioxide and 30-ppb increase in 1-
h max nitrogen dioxide.
Figure 5-9 Percentage increase in chronic obstructive pulmonary disease
hospital admissions and emergency department (ED) visits in
relation to nitrogen dioxide concentrations from U.S. and
Canadian studies evaluated in the 2008 Integrated Science
Assessment for Oxides of Nitrogen and recent studies.
January 2015
5-144
DRAFT: Do Not Cite or Quote
-------
Table 5-25 Corresponding risk estimate for studies presei
Study
Location
Age
Avg Time
nted in Figure 5-9.
Lag
% Increase
(95% Cl)
Hospital Admissions
Ko et al. (2007a)
tQiuetal. (201 3a)'
tWonq et al. (2009)a
tYanq et al. (2005)
tMoolqavkar (2003)
tFaustini et al. (2013)
Hong Kong
Hong Kong
Hong Kong
Vancouver, Canada
Cook County, IL
LA County, CA
6 Italian cities
All
All
All
65+
65+
65+
65+
35+
24-h avg
24-h avg
24-h avg
24-h avg
24-h avg
24-h avg
24-h avg
24-h avg
0-3
0-3
0-1
0-1
1
0
0
0-5 DL
10.1 (8.5, 12.2)
2.2(0.8,4.4)
7.1 (5.1, 9.1)
4.6(2.4,6.8)
19.0(4.0, 37.0)
4.9(1.6, 8.2)
3.6(2.8,4.5)
3.1 (-2.6, 9.2)
Emergency Department Visits
Peel et al. (2005)
tStieb et al. (2009)
tArbex et al. (2009)
Atlanta, GA
7 Canadian cities
Sao Paulo, Brazil
All
All
65+
1-h max
24-h avg
24-h avg
0-2
0
0-5 DL
5.3(0.9, 9.9)
0.1 (-6.1,6.8)
9.6(0.2, 19.9)
Cl = confidence interval, DL = distributed lag.
fStudies published since the 2008 ISA for Oxides of Nitrogen.
1
2
o
3
4
5
6
7
5.2.4.3 Subclinical Effects Underlying Chronic Obstructive
Pulmonary Disease—Pulmonary Inflammation
Exacerbation of COPD can be precipitated by increases in airway responsiveness and
pulmonary inflammation. While there is some supporting evidence for an effect of NC>2
exposure in initiation of inflammation (Sections 4.3.2.3 and 4.3.2.1). the effects of NC>2
on airway responsiveness and inflammation are not well characterized in adults with
COPD. Thus, little information on mode of action is available to support the associations
observed between ambient NO2 concentrations and hospital admissions and ED visits for
COPD (Section 5.2.4.1). In a recent epidemiologic study, neither ambient NO2 nor NO
was associated with indicators of inflammation such as increases in the numbers of blood
January 2015
5-145
DRAFT: Do Not Cite or Quote
-------
1 eosinophils or lymphocytes in adults with COPD consistently among lag days 0, 1, and 2
2 ambient concentrations (Table 5-22) (Bruske etal.. 2010). NO2 at lag day 0 was
3 associated with an increase in neutrophils only with adjustment for UFP. However, the
4 95% CI was wide, indicating an imprecise association, and the NO2-UFP correlation was
5 high (Spearman r = 0.68). UFP and OC were associated with decreases in neutrophils, of
6 which the relation to COPD exacerbation is not clear. Neither in adults with COPD nor
7 healthy adults did Vagaggini et al. (1996) find changes in inflammatory cell counts in
8 sputum following exposure to 300 ppb NO2 for 1 hour.
5.2.4.4 Summary of Exacerbation of Chronic Obstructive
Pulmonary Disease
9 Evidence for the effects of short-term NO2 exposure on COPD exacerbation is
10 inconsistent among the various outcomes examined and across scientific disciplines. In
11 epidemiologic studies, short-term increases in ambient NO2 concentration are
12 consistently associated with increases in hospital admissions and ED visits for COPD
13 (Section 5.2.4.2) but not with increases in respiratory symptoms or decreases in lung
14 function among adults with COPD (Section 5.2.4.1). In limited examination, an
15 epidemiologic and controlled human exposure study do not indicate NO2-related
16 increases in inflammation in adults with COPD (Section 5.2.4.3). Thus, a mode of action
17 for NO2 effects on COPD exacerbation is not clear. Epidemiologic studies assigned NO2
18 exposure as central site ambient concentrations (average of multiple monitors, nearest
19 site), and many found associations with the traffic-related pollutants CO, BS, UFP, and
20 PM2.5. In the one study that examined potential confounding by traffic-related
21 copollutants, the 95% CIs for associations for NO2 and BS with respiratory symptoms
22 increased, and an independent or confounding effect was not discerned for either
23 pollutant (Peacock et al., 2011). Also, controlled human exposure studies do not
24 consistently find NO2-induced (200-4,000 ppb for 30 minutes to 6 hours) increases in
25 respiratory symptoms or decreases in lung function in adults with COPD
26 (Section 5.2.4.1). Because of the inconsistent evidence across disciplines for effects on
27 clinical indications of COPD exacerbation and the lack of evidence for effects on
28 underlying mechanisms, there is uncertainty regarding a relationship between short-term
29 NO2 exposure and COPD exacerbation.
5.2.5 Respiratory Infection
30 The respiratory tract is protected from exogenous pathogens and particles through various
31 lung host defense mechanisms that include mucociliary clearance, particle transport and
January 2015 5-146 DRAFT: Do Not Cite or Quote
-------
1 detoxification by alveolar macrophages, and innate and adaptive immunity. The 2008
2 ISA for Oxides of Nitrogen reported clear evidence from animal toxicological studies for
3 NO2-induced susceptibility to bacterial or viral infection with some coherence with
4 results from controlled human exposure and epidemiologic studies (U.S. EPA. 2008a).
5 There is some mechanistic support for these observations, with NO2-induced impairments
6 in alveolar macrophage (AM) function found in some but not all animal toxicological
7 studies. Effects on mucociliary clearance and activity were not in a consistent direction,
8 but the exact mechanism by which mucociliary clearance may impair host defense is not
9 well characterized. Recent contributions to the evidence base are limited to epidemiologic
10 studies. These studies show associations between increases in ambient NO2
11 concentrations and increases in hospital admissions and ED visits for respiratory
12 infections but do not consistently show associations with respiratory infections reported
13 or diagnosed in children.
5.2.5.1 Susceptibility to Bacterial or Viral Infection in
Experimental Studies
Toxicological Studies
14 A large body of evidence, provided by studies reviewed in the 2008 ISA for Oxides of
15 Nitrogen (U.S. EPA. 2008a). demonstrates increased susceptibility of rodents to viral or
16 bacterial infection following short-term NO2 exposure. These studies used a variety of
17 experimental approaches but in most cases included an infectivity model of exposing
18 animals to NO2 or filtered air and then combining treatment groups for a brief exposure
19 to an aerosol of a viable agent, such as Streptococcus zooepidemicus, Streptococcus
20 pyogenesi, Staphylococcus aureus, and Klebsiella pneumoniae. The majority of studies
21 measured mortality over a specified number of days following the challenge, but several
22 studies also examined endpoints such as bacterial counts and clearance (Table 5-26).
23 While there are differences in sensitivity across species to various infectious organisms,
24 host defense mechanisms are shared, and the infectivity model is well accepted as an
25 indicator of impaired or weakened pulmonary defense.
January 2015 5-147 DRAFT: Do Not Cite or Quote
-------
Table 5-26 Animal toxicological studies of susceptibility to infection.
Species (Strain);
Study Age; Sex; n
Exposure Details
(Concentration; Duration)
Endpoints Examined
Amoruso et al.
(1981)
Davis et al.
(1991)
Dowell et al.
(1971)
Ehrlich et al.
(1977)
Rat (Sprague-
Dawley); F,
n = 4/group
Mice (C57BL/6N);
8~10 weeks;
n = 6/group
Dog (beagle);
n = 11
Mice (CF-1 );
5"8 weeks; F;
n = 5-88/group
1 ,300, 1 ,900, and 3,000 ppb NO2 for 3 h
5,OOOppbNO2for4h;
Mycoplasma pulmonis challenge immediately
after exposure.
3,OOOppbNO2for1 h
0, 1,500, 2,000, 3,500, and 5,000 ppb NO2 for
3h;
Streptococcus pyogenes challenge immediately
after exposure.
Analysis of BAL fluid and
superoxide production by
AMs (PMA stimulation).
Bacterial clearance,
bactericidal activity.
Histopathological
evaluation and lung
surfactant properties.
Mortality
Ehrlich (1980)
(1,2) Mice;
6"8 weeks;
n > 88/group
(3) Mice, hamsters,
and squirrel
monkeys
(1) 500 ppb NO2 continuously for 1 week~1 yr
(2) 1,500 ppb NO2 continuously for 2 rr3 mo
(3) 1,500-50,000 ppb NO2 for 2 h
(1~3) Klebsiella pneumoniae challenge
immediately after exposure.
(1-3) Mortality
Gardner et al. Mice (Swiss (1) 500 ppb NO2 continuously for 7 days~1 yr Mortality
(1979) albino); F; (2) 1,500 ppb NO2 continuously for 2 h~21 days
n = 20/group (3) 1 ]500 ppb NO2 7 h/day for 7 h~11 days
(3) 3,500 ppb NO2 continuously for
30 min~16 days
(4) 3,500 ppb NO2 7 h/day for 7 h~13 days
(1~4) Streptococcus pyogenes challenge
immediately after exposure.
Goldstein et al. Mice (Swiss (1) Staphylococcus aureus challenge
(1973) albino); M; immediately before exposure;
n = 30/group 0, 1,900, and 3,800 ppb NO2 for 4 h
(2) 0, 1,000, and 2,300 ppb NO2 for 17 h;
Staphylococcus aureus challenge immediately
after exposure.
(1) Bacterial counts and
bactericidal activity 5 h after
challenge (i.e., 1 h after
exposure).
(2) Bacterial counts and
bactericidal activity 0 h and
4 h after challenge.
Goldstein et al.
(1974)
Mice (Swiss
albino); M;
n = 30/group
(1) 1,740 ppb NO2+ 110 ppb O3
(2) 1,490 ppb NO2 + 200 ppb O3
(3) 2,300 ppb NO2 + 200 ppb O3
(4) 1,780 ppb NO2 + 270 ppb O3
(5) 4,180 ppb NO2 + 210 ppb Osfl'S) 17 h;
Staphylococcus aureus challenge immediately
after exposure.
Bacterial counts,
bactericidal activity, and
bacterial clearance 0 h and
4 h after challenge.
January 2015
5-148
DRAFT: Do Not Cite or Quote
-------
Table 5-26 (Continued): Animal toxicological studies of susceptibility to infection.
Species (Strain);
Study Age; Sex; n
Exposure Details
(Concentration; Duration)
Endpoints Examined
Goldstein et al. Rat (Sprague-
(1977) Dawley); F
500, 1,000, and 2,400 ppb NO2 for 1 and 2 h Agglutination of AMs
Graham et al.
(1987)
Mice (CD-1); 4-6
weeks;
n = 5"12/group
(1) 4,500 ppb NO2 for 1, 3.5, and 7 h
(2) 1,500 ppb NO2 continuously with a daily
spike of 4,500 ppb for 1, 3.5, and 7 h;
(1"2) Streptococcus zooepidemicus challenge
immediately and 18 h after exposure.
Mortality
Hooftman et al. Rats (Wistar); M;
(1988) n = 10/group
3,000 ppb NO2 for 6 h/day, 5 days/week up to
21 days
Histopathological
evaluation, analysis of BAL
fluid, and AM function and
morphology.
Illinq et al.
(1980)
Mice (CD-1);
5~6 weeks; F;
n = 16/group
1 ,000 ppb, 3,000 ppb NO2, and air for 3 h; Mortality after
With or without continuous exercise; Streptococcus pyogenes
Streptococcus pyogenes challenge immediately
after exposure.
Mochitate et al. Rats (Wistar); M;
(1986) 19~23 weeks;
n = 6/group
4,000 ppb NO2 continuously up to 10 days
BAL fluid cell counts and
MA function and
morphology.
Parker et al.
(1989)
Mice (C57BL/6N
and C3H/HeN);
6~10 weeks
0 and 5,000 ppb NO2 for 4 h;
Mycoplasma pulmonis challenge immediately
after exposure.
Histopathological
evaluation, bacterial
infection and clearance 4 h
up to 7 days post-
challenge, BAL fluid cell
counts.
Purvis and
Ehrlich(1963)
Mice (Swiss
Webster & albino);
n > 25/group
1,500, 2,500, 3,500, and 5,000 ppb NO2 for 2 h;
Klebsiella pneumoniae challenge 0~27 h post-
exposure.
Mortality
Robison etal. Rats (Sprague- 100, 500, and 1,000 ppb NCbfor 1 h; AMs
(1990) Dawley) exposed ex vivo
Viability, LTB4 production,
neutrophil chemotaxis, and
superoxide production.
Robison and Rats (Sprague- 100, 2,000, and 5,000 ppb NO2 for 1~4 h; AMs
Forman (1993) Dawley) exposed ex vivo
Arachidonate metabolite
production induced by
treatment with a calcium
ionophore.
Rombout et al.
Rats (Wistar); M,
6 weeks;
n = 3-6/group
500, 1,390, and 2,800 ppb NO2for
1, 2, 4, 8, 16, and 28 days
Histopathological
evaluation
Rose et al. Mice (CD-1); (1) 1,000, 2,500, and 5,000 ppb NO2 for 6 h/day
(1988) 4~6 weeks; for 2 days; intra-tracheal inoculation with murine
Rose et al n > 4/group Cytomegalovirus; 4 additional days (6 h/day) of
(1989) L exposure.
(2) re-inoculation 30 days (air) post-primary
inoculation.
Infection 5 and 10 days
post-inoculation,
histopathological
evaluation, and analysis of
BAL fluid (LDH, albumin,
macrophages).
January 2015
5-149
DRAFT: Do Not Cite or Quote
-------
Table 5-26 (Continued): Animal toxicological studies of susceptibility to infection.
Study
Schlesinqer
(1987b)
Sherwood et
al. (1981)
Suzuki et al.
(1986)
Species (Strain);
Age; Sex; n
Rabbits (New
Zealand white);
M, n = 5/group
Mice (Swiss
albino); M;
n = 8-24/group
Rats (Fischer 344);
M, 7 weeks;
n = 8/group
Exposure Details
(Concentration; Duration)
300 or 1 ,000 ppb NO2 for 2 h/day for 2, 6, and
13 days
1,000 ppb NO2 for 24 and 48 h;
Streptococcus (Group C) challenge immediately
after exposure.
4,000 NO2 ppb for 1, 3, 5, 7, and 10 days
Endpoints Examined
Viability and AM activity
(mobility, attachment, and
phagocytosis).
Bacterial counts 0~48-h
post-challenge, bacterial
clearance, histopathological
evaluation, mortality.
AM activity (phagocytosis
and superoxide
production), SOD and
G-6-PD activity.
F = female, LDH = lactate dehydrogenase, M = male, NO2 = nitrogen dioxide, O3 = ozone, SOD = superoxide dismutase.
1
2
o
J
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
In a series of studies, (Goldstein et al. (1974): 1973)) examined bactericidal activity and
clearance in mice challenged with radiolabeled Staphylococcus aureus either
immediately before or after NO2 exposure. The number of bacteria deposited in the lung
was not different in NCh-exposed animals compared to controls; however,
concentration-dependent decreases in bactericidal activity were observed in animals
exposed to NCh for 4 hours after challenge as well as those exposed to NCh for 17 hours
before challenge. While the 4-hour exposure did not yield statistically significant
differences compared to air controls at NO2 concentrations less than 7,000 ppb, the
17-hour exposure preceding challenge was statistically significant for concentrations
greater than 2,300 ppb. Parker etal. (1989) also used radiolabeled bacteria to determine
the effects of NO2 on susceptibility to infection. This study demonstrated that a 4-hour
exposure to 5,000 ppb NCh was sufficient to reduce bactericidal activity and increase the
number of bacteria in the lungs of C3H/HeN and C57BL/6N mice 3 days after challenge
compared to control mice, but did not result in an increase in incidence or severity of
lung lesions. These results were corroborated in a similar study by Davis etal. (1991).
It is also important to consider differences in response to NC>2 that are specific to the
infectious organism as Jakab (1988) has demonstrated. A 4-hour exposure to 5,000 ppb
NC>2 resulted in a decrease in bactericidal activity after challenge with Staphylococcus
aureus; however, bactericidal activity against Proteus mirabilis and Pasteurella
pneumotropica was not impaired with exposure to NC>2 at concentrations less than
20,000 ppb. Additionally, Sherwood et al. (1981) observed an increase in the propensity
of virulent group C Streptococci, but not Staphylococcus aureus following exposure to
1,000 ppb NO2 for 24 or 48 hours. In this study, Streptococci infection did not increase
the total mortality compared to controls, but NC>2-exposed mice died significantly earlier.
January 2015
5-150
DRAFT: Do Not Cite or Quote
-------
1 Several other studies reported that NO2 exposure increased mortality from bacterial
2 infection, tiling et al. (1980) exposed mice to 3,000 ppb NC>2 with continuous exercise for
3 3 hours while Ehrlich et al. (1977) exposed mice to 2,000 ppb NC>2 for 3 hours; both
4 studies subsequently exposed mice to an aerosol of Streptococcus pyogenes and reported
5 increased mortality rates compared to control animals exposed to clean air. Increases in
6 mortality from Streptococcus pyogenes infection following NCh exposure were also
7 reported by Ehrlich etal. (1979). In this study, the relationship between concentration and
8 time was examined, and these factors yielded very different results as the concentration
9 was more important than time in determining mortality. Results were consistent with
10 other studies, though mortality increased post-challenge following a 7-day exposure to
11 3,500 ppb NO2.
12 Ehrlich (1980) conducted similar studies to investigate the effects of NC>2 on Klebsiella
13 pneumoniae-mduced mortality. Challenge following exposure to 1,500 ppb NCh for more
14 than 8 hours resulted in increased mortality; however, a longer duration of exposure
15 (3 months) was required to increase infection mortality following 500 ppb NC>2 exposure.
16 This study also demonstrated species differences as increases in K. pneumoniae infection
17 mortality in mice were observed after a 2-hour exposure to 3,500 ppb NCh while
18 hamsters and squirrel monkeys did not experience increases in mortality at NC>2
19 concentrations less than 35,000 ppb and 50,000 ppb, respectively (Ehrlich. 1980).
20 Conversely, Purvis and Ehrlich (1963) did not observe increases in K. pneumonia
21 infection mortality in mice following a 2-hour exposure to NCh at concentrations less
22 than 5,000 ppb.
23 One study examined effects of NO2 peak exposures superimposed on a lower continuous
24 background level of NO2 on susceptibility to Streptococcus zooepidemicus infection
25 (Graham et al.. 1987). Mice were exposed to 4,500 ppb NO2 for 1, 3.5, and 7 hours or
26 exposed to these spikes with a continuous background exposure to 1,500 ppb NO2,
27 followed either immediately or 18 hours later with a Streptococcus zooepidemicus
28 challenge. Compared to control animals, the 4,500 ppb spikes alone or the spikes
29 superimposed on a background exposure did not result in differences in mortality from
30 infection; however, combined mortality rates (following the 1-hour exposure to
31 4,500 ppb and the 1-hour exposure to 4,500 ppb with 1,500 ppb background) were
32 significantly increased from immediate challenge after 4,500 ppb NO2 and were
33 proportional to duration of the 4,500 ppb exposure. In animals challenged 18 hours after
34 NO2 exposure, increases in mortality were only statistically significant with 3.5- and
35 7-hour exposures to 4,500 ppb NO2.
January 2015 5-151 DRAFT: Do Not Cite or Quote
-------
Controlled Human Exposure Studies
1
2
3
4
5
6
7
8
9
10
11
12
13
14
Compared with animal toxicological studies, controlled human exposure studies provided
less consistent evidence for NCh-induced infectivity (Table 5-27). Although
Pathmanathan et al. (2003) found increased expression of inter-cellular adhesion
molecule 1 (ICAM-1), an extracellular receptor for viruses, in airway biopsies following
exposure to 2,000 ppb NCh for 4 hours per day for 4 days, Frampton et al. (2002) did not
find evidence of increased susceptibility to ex vivo viral challenge in bronchial epithelial
cells collected from subjects exposed to 600 ppb or 1,500 ppb NO2 for 3 hours; however,
there was an increase in virus-induced cytotoxicity as measured by lactate dehydrogenase
(LDH) release. Consistent with Frampton et al. (2002). Goings etal. (1989) reported no
increase in infectivity of administered live, attenuated influenza virus in subjects exposed
to 1,000, 2,000, or 3,000 ppb NC>2 for 2 hours/day for 3 consecutive days. This study,
however, lacked a sham control. Another study (Frampton et al.. 1989) reported a trend
of decreased inactivation of influenza virus in AMs collected from subjects after a 3-hour
exposure to 600 ppb NO2, although results were not statistically significant.
Table 5-27 Controlled human exposure studies of susceptibility to infection.
Study
Frampton et al.
(1989)
Frampton et al.
(2002)
Goinqs et al.
(1989)
n, Sex; Age
(mean ± SD)
(1)n = 7M, 2F;
30 yr (range: 24'37)
(2) n = 1 1 M, 4 F;
25 yr (range: 19~37)
(1,2) n = 12 M, 9F;
F = 27.1 ±4.1 yr
M = 26.9 ± 4.5 yr
(1)n = 44
(2) n = 43
(3) n = 65; range:
18-35yr
Exposure Details
(Concentration; Duration)
(1 ) 600 ppb for 3 h,
(2) 1,500 ppb for 3 h;
(1,2) Exercise 10 min on/20 min off
at VE = ~4 times resting
(1 ) 600 ppb for 3 h,
(2) 1,500 ppb for 3 h;
(1,2) Exercise 10 min on/20 min off
at VE = 40 L/min
(1)2, 000 ppb for 2 h
(2) 3,000 ppb for 2 h
(3) 1,000 or 2,000 ppb for 2 h
Endpoints Examined
BAL fluid cell viability and differential
counts 3.5 h post-exposure,
inactivation of influenza virus by BAL
cells, IL-1 activity in BAL cells.
Bronchial and alveolar lavage fluid
cell viability and differential counts
3.5 h post-exposure, influenza and
RSV challenge in BAL cells,
peripheral blood characterization.
Nasal wash virus isolation and count
4 days after virus administration.
Serum and nasal wash antibody
response 4 weeks after virus
administration.
Pathmanathan n = 8 M, 4 F
et al. (2003) 26 yr (range:
21-32yr)
2,000 ppb for 4 h/day for 4 days;
Exercise 15 min on/15 min off at
75 watts
Biomarkers in bronchial epithelium-
exotoxin, GM-CSF, Gro-a, I-CAM 1,
IL-5, IL-6, IL-8, IL-10, IL-13, total and
active NF-K(3, and TNF-a (fiberoptic
bronchoscopy after end of last
exposure).
F = female, IL = interleukin, M = male, RSV = respiratory syncytial virus, SD = standard deviation.
January 2015
5-152
DRAFT: Do Not Cite or Quote
-------
5.2.5.2 Respiratory Infections Reported or Diagnosed in
Children
1 In contrast with findings in rodent models, epidemiologic evidence does not clearly
2 indicate a relationship between short-term NC>2 exposure and respiratory infection in
3 children (ages 0-15 years) as reported by self or parents or more objectively ascertained
4 as laboratory-confirmed or physician-diagnosed cases. Some studies found associations
5 (Espositoetal.. 2014; Luetal.. 2014; Stern etal.. 2013; Ghosh etal.. 2012; Just et al..
6 2002); others did not or found inconsistent associations among the outcomes examined
7 (Altug etal.. 2014; Stern etal.. 2013; Xu etal.. 2013) (Table 5-28). Xu etal. (2013) did
8 not provide strong evidence for an association of ambient NC>2 with laboratory-confirmed
9 cases of influenza (RR: 1.01 [95% CI: 0.97, 1.04] for an unreported increment inNO2);
10 however, study limitations preclude strong inferences from the results. There were a
11 mean of only two influenza cases per day, and potential collinearity in a multipollutant
12 model with PMio and Os (Spearman r for correlation with NC>2 = 0.62 and -0.42,
13 respectively) limits inference about NC>2 effects.
14 Results indicating associations between ambient NO2 or NOx and respiratory infections
15 also have weak implications (Esposito et al.. 2014; Lu etal.. 2014; Stern etal.. 2013;
16 Ghosh etal.. 2012; Just et al.. 2002). All of these studies assigned exposure from central
17 site concentrations (one city site or average of multiple sites) (Table 5-28). None reported
18 information on the spatial distribution of subjects around the monitoring site(s) or the
19 within-city variability in NC>2 or NOx concentrations to ascertain potential exposure
20 measurement error and its impact on effect estimates. A few studies examined more
21 spatially resolved exposure metrics but also have uncertain implications. Ghosh et al.
22 (2012) reported similar results in analyses restricted to homes for which central site NOx
23 belter represented exposure but did not report how these homes were selected. Further,
24 the adequacy of NOx to serve as an indicator of NO2 may vary among subjects because of
25 varying NO2 to NOx ratios across locations (Section 2.5.3). NO2 at the central site nearest
26 to schools was associated with pneumonia prevalence (Lu et al.. 2014). However,
27 pneumonia was ascertained as "ever having a diagnosis" and may not be temporally
28 matched to exposure assessed for a three-year period. The only study with
29 spatially-aligned measures of exposure did not observe associations between school NO2
30 and colds (Altug et al.. 2014). The importance of the microenvironmental measures is
31 underscored by the variability in traffic volume and road length reported within the study
32 area.
33 Also limiting strong inferences from results of these studies is uncertainty regarding
34 confounding by other traffic-related pollutants. In addition to NO2, respiratory infections
35 were associated with BS and PlVfc 5. Other copollutants such as PMio, SO2, and Os also
January 2015 5-153 DRAFT: Do Not Cite or Quote
-------
1
2
o
6
4
5
were associated with respiratory infection (Table 5-28). Studies did not examine other
traffic-related PM components, CO, copollutant models, or other methods to assess the
independent or mixture effects of NO2. Where reported, NO2 was moderately to highly
correlated with copollutants (r = 0.92 for BS; 0.6-0.8 for unspecified copollutants)
(Ghosh et al.. 2012; Just et al.. 2002).
Table 5-28 Epidemiologic studies of respiratory infections reported or
diagnosed in children.
Study
Population Examined and Methodological
Details
Oxide of Nitrogen
Metrics Analyzed
Effect Estimate
(95% Cl)
Single-Pollutant
Model3
Copollutant
Examination
tAltuq etal. (2014)
Eskisehir, Turkey, Feb-Mar, 2007
n = 605, ages 9-1 Syr
Cross-sectional. Self-reported respiratory
infections. Recruitment from schools of
participants of a larger study. Participation rate
not reported. Logistic regression adjusted for
sex, age, asthma, parental smoking, coal or
wood stove use, parental education, height,
weight, daily average temperature.
NO2-outdoor school Common cold last
24-h avg, lag 0'6
day avg
1 site at each of
16 schools
Means & max(ppb)
Suburban: 9.4, 13
Urban: 13, 18
Urban-traffic: 21, 28
7 days:
OR: 1.86(0.41,8.42)
Commom cold at the
moment:
OR: 4.59 (0.79, 26)
No copollutant model.
03 associated with
colds. Strong inverse
correlation with NO2.
Pearson r= -0.80.
NO2 and PM2.5 reported
to be highly correlated.
tEsposito et al. (2014)
Milan, Italy, Jan-Dec 2012
n = 718, ages 2-18 yr, 329 with wheeze or
asthma, 389 healthy children
Repeated measures. Daily symptom diaries for
12 mo. Diaries checked weekly, clinic visits
conducted every 2 mo. Recruited from
respiratory disease section (wheeze/asthma)
and outpatient surgery (healthy) sections of
pediatric clinic. 89% follow-up participation.
Followed cohort similar to cohort at baseline.
GEE adjusted for age, sex, siblings, parental
education, smokers in home, season, day of
week, temperature, humidity.
NO2-central site
1-h max, Lag 0~2
day avg
8 city sites,
7 surrounding area
Weighted avg at
municipality level
Tertiles (T) in ppb
1: <47.3b
2: 47.3-60.1b
3: >60.1b
RR for pneumonia
with T1 as reference
Children with asthma:
T2: 1.20(0.75, 1.90)
T3: 1.56(1.01,2.42)
Healthy children:
T2: 1.45(0.80, 0.63)
T3: 1.02(0.93, 1.12)
No copollutant model.
PM-io associated with
pneumonia.
Correlations NR.
Just etal. (2002)
Paris, France, Apr-Jun 1996
n = 82, ages 7~15 yr, children with asthma,
90% atopy
Repeated measures. Daily symptom diaries for
3 mo, collected weekly. Recruitment from
hospital outpatients. 82% follow-up
participation. GEE adjusted for time trend, day
of week, pollen, temperature, humidity.
NO2-central site
24-h avg, lag 0 day
Average of 11 sites
Mean: 28.6 ppbb
59.0
Respiratory infection:
OR: 7.19(2.53,20.4)
No copollutant model.
BS associated with
cough and infection.
High correlation with
NO2. Pearson r= 0.92.
tStern etal. (2013)
Bern, Basel, Switzerland, Apr 1999'Feb 2011
n = 366 ages 0~1 yr
NO2-central site
24-h avg, lag
5~11 day avg
2 sites
Rural mean:
8.1 ppbb
Incidence respiratory
tract infection:
RR: 1.20(0.82, 1.75)
Days with respiratory
tract infection:
No copollutant model.
PM-io lag 7 days
associated with
respiratory infection.
Correlation NR.
January 2015
5-154
DRAFT: Do Not Cite or Quote
-------
Table 5-28 (Continued): Epidemiologic studies of respiratory infections reported
or diagnosed in children.
Study
Population Examined and Methodological
Details
Repeated measures. Symptoms reported
weekly by telephone for 1 yr. Recruitment from
birth cohort. 95% follow-up participation. GAM
adjusted for study week, sex, siblings, nursery
care, prenatal maternal smoking, postnatal
maternal smoking, birth weight, maternal
atopy, parental education.
tXuetal. (2013)
Brisbane, Australia, 2001-2008, winters only
n = 2,922 influenza cases, ages 0~14 yr
Time-series. Laboratory-confirmed cases of
influenza referred by public or private health
sector. Only mean 2/day. No information
available on subjects' residential location.
Poisson regression adjusted for lag 0~9 day
avg temperature, lag day 0~9 avg PM-io, lag 0~9
day avg Os, PMio*temperature interaction.
tGhoshetal. (2012)
Teplice and Prachatice, Czech Republic, May
1994-June2003
n = 1,113 children, ages 0~4.5 yr
Repeated measures. Physician-diagnosed
infections between ages 0~4.5 yr ascertained
from medical records. Recruitment from birth
cohort. Participation rate not reported. GEE
with exchangeable correlation and adjusted for
city, year of birth, day of week, fuel used for
heating and cooking, season, 7-day avg
temperature. Restricted analyses to children
for whom central site may better represent
exposure (method not reported).
tLuetal. (201 4)b
Changsha, China, Sep 2011 'Jan 2012
n = 2706, ages 3'6 yr
Cross-sectional. Recruitment from schools.
59% participation. Potential temporal mismatch
of exposure (2008~201 1 ) and ever having
pneumonia diagnosis. Two-level model.
Pneumonia first adjusted for parental atopy,
antibiotic use, new furniture in home, coal,
wood or gas used in home, painted walls/air
conditioning in home, pets in home, visible
mold/dampness in home. Adjusted pneumonia
prevalence regressed with NO2. Confounding
by meteorological factors not examined.
Oxide of Nitrogen
Metrics Analyzed
Urban mean:
25.6 ppbb
Upper percentiles
NR
NO2-central site
24-h avg, lag
0~9 day avg
# sites NR
Mean: 5.9 ppbb
75th: 7.3 ppbb
Max: 13.3ppbb
N Ox-centra I site
24-h avg, lag
0~2 day avg
2 sites
Mean, 75th (ug/m3)
Teplice: 59.2, 73.3
Prachatice: 20.3,
i~) A A
24.4
NO2-central site
24-h avg
Nearest to school,
distance NR
Concentrations NR
2008~2011:
7 days >63.8 ppbb
standard before
2012
89 days >42.6 ppbb
e-*4-f* «rJ^i i-rJ Ol"\ *1 O
standard 2012
Effect Estimate
(95% Cl)
Single-Pollutant
Model3
NO2 < 26 ppb:
reference category
NO2> 26 ppb:
18%(0, 39)
Daily influenza
counts:
RR: 1.01 (0.97, 1.04)
increment of NO2 NR.
Results are presented
only for a
multipollutant model
that also includes
D l\ /I nnrt f~\
PMio and Os.
Acute bronchitis:
Birth to age 2 yr:
RR: 1.09(1.01, 1.16)
per 34 ug/m3 NOx
Age 2 yr4.5 yr:
RR: 1.05(0.94, 1.14)
per 32 ug/m3 NOx
OR for
NO2 >63.8 ppb
1 day/yr
1.04(1.02, 1.05)
Copollutant
Examination
Only multipollutant
model analyzed.
No copollutant model.
Association with PM2.5
reported in separate
paper.
Moderate to high
correlations reported
with unspecified
copollutants.
r= 0.6-0.8
No copollutant model
PM-ioand SO2
associated with
pneumonia.
Correlations NR.
Note: More informative studies in terms of the exposure assessment method and potential confounding considered are presented
first.
GEE = generalized estimating equations, GAM = generalized additive model, NR = not reported, RR = relative risk, Cl = confidence
interval, NO2 = nitrogen dioxide, NOX = sum of NO and NO2,03 = ozone, OR = odds ratio, PM = particulate matter, SO2 = sulfur
dioxide.
aEffect estimates are standardized to a 20 ppb for 24-h avg NO2. NOX effect estimates are presented as reported in the study
(Section 5.1.2.3).
""Concentrations converted from |jg/m3 to ppb using the conversion factor of 0.532 assuming standard temperature (25°C) and
pressure (1 atm).
fStudies published since the 2008 ISA for Oxides of Nitrogen.
January 2015
5-155
DRAFT: Do Not Cite or Quote
-------
5.2.5.3 Hospital Admissions and Emergency Department Visits
for Respiratory Infections
1 To date, relatively few studies have examined the association between short-term NO2
2 exposures and hospital admissions and ED visits due to respiratory infections. The 2008
3 ISA for Oxides of Nitrogen identified studies that evaluated a number of respiratory
4 infection outcomes, such as upper respiratory infections (URIs), pneumonia, bronchitis,
5 allergic rhinitis, and lower respiratory disease. Across these outcomes, studies have
6 generally not provided consistent evidence of an association between NO2 and hospital
7 admissions and ED visits due to respiratory infections (U.S. EPA. 2008a). Recent studies
8 add to the body of literature evaluated in the 2008 ISA for Oxides of Nitrogen, but
9 compared to other respiratory-related hospital admission and ED visit outcomes the total
10 body of literature remains limited. The air quality characteristics of the city, or across all
11 cities, and the exposure assignment approach used in each respiratory infection-related
12 hospital admission and ED visit study evaluated in this section are presented in
13 Table 5-29. As detailed in Section 5.2.2.4. other recent studies of respiratory
14 infection-related hospital admissions and ED visits are not the focus of this evaluation,
15 and the full list of these studies, as well as study details, can be found in Supplemental
16 Table S5-3 (U.S. EPA. 2014h).
January 2015 5-156 DRAFT: Do Not Cite or Quote
-------
Table 5-29 Mean and upper percentile concentrations of nitrogen dioxide (NO2) in studies of hospital
admissions and emergency department visits for respiratory infection.
Location Type of Visit
Study Years (ICD9/10)
Mean
Exposure Concentration
Assignment Metric ppb
Upper Percentile
Concentrations Copollutant
ppb Examination
Hospital Admissions
Burnett et al. (1999) Toronto, Canada Respiratory
(1980-1994) infection (464,
466, 480-7,
494)
Lin et al. (2005) Toronto, Canada Respiratory
(1998-2001) infection (464,
466, 480-487)
Karret al. (2006) Southern Los Acute
Angeles County, bronchiolitis
CA (466.1)
(1995-2000)
Average of NO2 24-h avg 25.2
concentrations
across 4 monitors.
Average of NO2 24-h avg 25.5
concentrations
across 7 monitors.
34 NO2 monitors, 1-h max 59
exposure assigned
based on nearest
monitor to residential
ZIP code.
NR Correlations (r):
PlVh.s: 0.55
PMio-2.s: 0.38
PMio: 0.57
CO: 0.64
SO2: 0.54
O3: -0.08
Copollutant models: none
75th: 29.3 Correlations (r):
PM2.5: 0.48
PMio-2.s: 0.40
PMio: 0.54
CO: 0.20
SO2: 0.61
O3: 0.0
Copollutant models: none
75th: 69 Correlations (r): NR
90th: 90 Copollutant models: none
January 2015
5-157
DRAFT: Do Not Cite or Quote
-------
Table 5-29 (Continued): Mean and upper percentile concentrations of nitrogen dioxide (NO2) in studies of hospital
admissions and emergency department visits for respiratory infection.
Study
Zanobetti and
Schwartz (2006)
tHEl (2012)
tMehtaetal. (2013)
tSeqala et al.
(2008)
tFaustini et al.
(2013)
Location
Years
Boston, MA
(1995-1999)
Ho Chi Minh City,
Vietnam
(2003-
2005)
Paris, France
(1997-2001)
6 Italian cities
(2001-2005)
Type of Visit
(ICD9/10)
Pneumonia
(480-487)
Acute lower
respiratory
infection
(J13-16, 18,
21)
Bronchiolitis
LRTI (466,
480-487)
Mean
Exposure Concentration
Assignment Metric ppb
Average of NO2 24-h avg NR
concentrations
across 5 monitors.
Average of NO2 24-h avg 1 1 .7
concentrations
across 9 monitors.
Average of NO2 24-h avg 27.0
concentrations from
21 monitors,
representative of
urban background.
Average of NO2 24-h avg 24.1-34.6
concentrations over
all monitors within
each city. Number of
NO2 monitors in each
city ranged from
1-5.a
Upper Percentile
Concentrations Copollutant
ppb Examination
50th: 23.2 Correlations (r):
PlVh.s: 0.55
BC: 0.70
CO: 0.67
O3: -0.14
PM non-traffic: 0.14
Copollutant models:
Max: 29.2 Correlations (r):
Dry season:
PMio: 0.78
O3: 0.44
SO2: 0.29
Rainy season:
PMio: 0.18
O3: 0.17
SO2: 0.01
Copollutant models:
PMio, Os
Max: 90.4 Correlations (r):
BS: 0.83
PMio: 0.74
SO2: 0.78
Copollutant models:
none
SO2,
none
NR Correlations (r), across
cities:
PMio: 0.22-0.79
Copollutant models: PMio
January 2015
5-158
DRAFT: Do Not Cite or Quote
-------
Table 5-29 (Continued): Mean and upper percentile concentrations of nitrogen dioxide (NO2) in studies of hospital
admissions and emergency department visits for respiratory infection.
Study
Location
Years
Type of Visit
(ICD9/10)
Exposure
Assignment
Metric
Mean
Concentration
ppb
Upper Percentile
Concentrations
ppb
Copollutant
Examination
ED Visits
Peel et al. (2005)
Atlanta, GA
(1993-2000)
Upper
respiratory
infection (460-
6,477)
Pneumonia
(480-486)
Average of NO2
concentrations from
monitors for several
monitoring networks.
1-h max 45.9
NR
Correlations (r):
PIvh.s: 0.46
PMio: 0.49
PMio-2.s: 0.46
UFP: 0.26
PIvh.s water-soluble
metals: 0.32
PIvh.s sulfate: 0.17
PIvh.s acidity: 0.10
PIvh.s OC: 0.63
PIvh.s EC: 0.61
Oxygenated HCs: 0.30
O3: 0.42
CO: 0.68
SO2: 0.34
Copollutant models: none
tStieb et al. (2009)
tSeqala et al.
(2008)
7 Canadian cities
(1992-2003)
Paris, France
(1997-2001)
Respiratory
infection (464,
466, 480-487)
Bronchiolitis
Average NO2 24-h avg 9.3-22.7
concentrations from
all monitors in each
city. Number of NO2
monitors in each city
ranged from 1-14.
Average of NO2 24-h avg 27.0
concentrations
across 21 monitors,
representative of
urban background.
reported by city and
season.
Copollutant models: none
Max: 90.4 Correlations (r):
BS: 0.83
PMio: 0.74
SO2: 0.78
Copollutant models: none
January 2015
5-159
DRAFT: Do Not Cite or Quote
-------
Table 5-29 (Continued): Mean and upper percentile concentrations of nitrogen dioxide (NO2) in studies of hospital
admissions and emergency department visits for respiratory infection.
Study
tZemek et al.
(2010)
Location
Years
Edmonton,
Canada
(1992-2002)
Mean Upper Percentile
Type of Visit Exposure Concentration Concentrations Copollutant
(ICD9/10)
Otitis media
(382.9)
Assignment Metric ppb ppb Examination
Average of NO2 24-h avg 21.9 75th: 27.6 Correlations (r): NR
concentrations Copollutant models: none
across 3 monitors.
Physician Visits
fSinclair et al.
(2010)
Atlanta, GA
(1998-2002)
Upper
respiratory
infection
Lower
respiratory
infection
NO2 concentrations 1-h max 1998-2000:49.8 NR Correlations (r): NR
collected as part of 2000-2002: 35.5 Copollutant models: none
AIRES at SEARCH „„„„ „„„„ ,„ _,
Jefferson Street site. 1998-2002:41.7
CO = carbon monoxide, ED = emergency department, HC = hydrocarbons, LRTI = lower respiratory tract infection, NO2 = nitrogen dioxide, NR = not reported, O3 = ozone,
OC = organic carbon, PM = particulate matter, SO2 = sulfur dioxide, UFP = ultrafine particles.
aMonitoring information obtained from Colais et al. (2012).
fStudies published since the 2008 ISA for Oxides of Nitrogen.
January 2015
5-160
DRAFT: Do Not Cite or Quote
-------
Hospital Admissions
1 Few recent studies have examined the association between short-term NCh exposures and
2 respiratory infection hospital admissions. A time-series study conducted in Ho Chi Minh
3 City, Vietnam (Mehtaet al.. 2013; HEI. 2012) examined the association between
4 short-term air pollution exposures and pediatric (ages 28 days-5 years) hospital
5 admissions for acute lower respiratory infections (ALRI, including bronchiolitis and
6 pneumonia). In a time-stratified case-crossover analysis focused only on the average of a
7 1-6 day lag, there was no evidence of an association between NO2 and ALRI hospital
8 admissions in the all-year analysis (-4.0% [95% CI: -18.0, 12.5] for a 20-ppb increase in
9 24-h avg NC>2 concentrations).
10 In an additional study that also examined respiratory infections (i.e., bronchiolitis) in
11 children, Segalaetal. (2008) focused on associations with winter (October-January) air
12 pollution because it is the time of year when respiratory syncytial virus (RSV) activity
13 peaks. It has been hypothesized that air pollution exposures may increase the risk of
14 respiratory infections, including bronchiolitis due to RSV (Segalaetal.. 2008). Focusing
15 on children <3 years of age in Paris, France, the authors conducted a bidirectional
16 case-crossover analysis along with a time-series analysis to examine air pollution (i.e.,
17 PMio, BS, NC>2, 802) associations with bronchiolitis ED visits and hospital admissions.
18 Although the authors specify the bidirectional case-crossover approach was used to
19 "avoid time-trend bias," it must be noted that the bidirectional approach has been shown
20 to bias results (Segala et al.. 2008; Levy et al.. 2001). In the case-crossover analysis NO2
21 was associated with bronchiolitis hospital admissions (15.9% [95% CI: 7.7, 29.0], lag
22 0-4 days for a 20-ppb increase in 24-h avg NC>2 concentrations); NCh was not examined
23 in the time-series analysis. Although a positive association was observed, the authors did
24 not analyze copollutant models. The lack of copollutant analyses complicates the
25 interpretation of these results because the pollutants were highly correlated, ranging from
26 r = 0.74-0.83.
27 Faustini et al. (2013). in the analysis of air pollution in six Italian cities, also examined
28 associations with lower respiratory tract infection (LRTI) hospital admissions. However,
29 the authors only focused on LRTIs in individuals with COPD over the age of 35. Unlike
30 the analyses focusing on only COPD hospital admissions where the strongest associations
31 were for immediate effects (i.e., lag 0 and 0-1 days) for the population of individuals
32 with COPD that had a hospital admission for a LRTI there was no evidence of an effect
33 at these shorter durations; the largest associations were observed at lag 2-5 days (10.0%
34 [95% CI: -2.7, 24.3]). The authors examined the NO2 association with LRTI hospital
35 admissions in copollutant models with PMio at lag 0-5 days and, consistent with the
January 2015 5-161 DRAFT: Do Not Cite or Quote
-------
1 other endpoints examined, reported that although results were attenuated they remained
2 positive (7.8% [95% CI: -6.5, 24.2]).
Emergency Department Visits
3 Studies that examined the effect of air pollution on ED visits attributed to respiratory
4 infections have focused on similar outcomes to those examined in the studies of hospital
5 admissions. Stieb et al. (2009). in their study of seven Canadian cities, also examined the
6 association between short-term NCh concentrations and respiratory infection ED visits.
7 The authors reported positive associations at lags of 1 and 2 days, but the confidence
8 intervals were wide, providing little evidence of an association. However, Segala et al.
9 (2008) in the study of winter (October-January) air pollution in Paris, France (discussed
10 above) reported evidence of an association between short-term NCh concentrations and
11 bronchiolitis ED visits (11.8% [95% CI: 7.7, 20.1]; lag 0-4 day avg for a 20-ppb increase
12 in 24-h avg NC>2 concentrations) in a bi-directional case-crossover analysis. As
13 mentioned previously the interpretation of these results is complicated by the lack of
14 copollutant analyses and the high correlation between pollutants examined (r = 0.74 to
15 0.83).
16 In an additional study conducted in Edmonton, Alberta, Canada, Zemek et al. (2010)
17 examined otitis media (i.e., ear infections) ED visits, for ages 1-3 years. Associations
18 were examined for single-day lags of 0 to 4 days in all-year as well as seasonal analyses.
19 The authors observed that NO2 was associated with increases in ED visits for otitis media
20 in the all-year analysis at lag 2 days (7.9% [95% CI: 1.6, 12.8] for a 20-ppb increase in
21 24-h avg NC>2 concentrations). When examining whether there was evidence of seasonal
22 patterns in associations, the authors found that the magnitude of the association was
23 larger in the warm months (April-September), 16.1% (95% CI: 3.1, 31.2), compared to
24 the cold months, (October-March), 4.7% (95% CI: 0, 11.2) at lag 2 days for a 20-ppb
25 increase in 24-h avg NO2 concentrations. Additionally, it is important to note that the
26 pattern of associations for CO were similar to that observed for NO2, but the authors did
27 not report correlations between pollutants or conduct copollutant analyses.
Outpatient and Physician Visits
28 In addition to examining severe occurrences of a respiratory infection that would require
29 a trip to a hospital, studies have also began to explore whether air pollution may lead to
30 less severe cases, which would be reflective of trips to an outpatient facility. In a study
31 conducted in Atlanta, GA, Sinclair et al. (2010) also examined the association between
32 air pollution and respiratory infection (e.g., upper respiratory infections, lower respiratory
33 infections) outpatient visits from a managed care organization. As detailed previously
January 2015 5-162 DRAFT: Do Not Cite or Quote
-------
1 (Section 5.2.2.4). the authors separated the analysis into two time periods to compare air
2 pollutant concentrations and relationships for acute respiratory visits for the 25-month
3 time period examined in Sinclair and Tolsma (2004). and an additional 28-month time
4 period of available data from AIRES. Across the two time periods, mean 1-h max NO2
5 concentrations were lower in the 28-month versus the 25-month time period, 49.8 ppb
6 versus 35.5 ppb, respectively (Table 5-29). For both outcomes, the daily number of
7 outpatient visit counts varied with LRI being rather small (i.e., 12 per day) compared to
8 that for URI (i.e., 263 per day). A comparison of the two time periods indicated that risk
9 estimates for LRI and URI tended to be larger in the earlier 25-month period compared to
10 the later 28-month period with relatively wide confidence intervals for both outcomes.
11 Additionally, the lag structure of associations varied between each time period. For LRI,
12 the largest magnitude of an association was for both lag 0-2 and 3-5 day avg in the
13 earlier time period, but only lag 3-5 day avg in the latter time period; whereas, for URI
14 the largest associations were for lag 0-2 and 3-5 days for the earlier time period, but a
15 positive association was only observed for lag 6-8 days in the latter time period. The
16 authors also examined potential seasonal differences in associations, but the inconsistent
17 results between the two time periods with respect to the lag structure of associations also
18 complicates the interpretation of seasonal results.
Summary of Respiratory Infection Hospital Admissions and Emergency
Department Visits
19 Recent studies that examined the association between short-term NO2 exposure and
20 hospital admissions and ED visits due to respiratory infections add to the body of
21 evidence detailed in the 2008 ISA for Oxides of Nitrogen, but studies have not
22 consistently examined similar respiratory infection outcomes (Figure 5-10 and
23 Table 5-30). Of the studies evaluated, the strongest associations are for studies that
24 focused on children, specifically less than 5 years of age. These studies demonstrate
25 associations with respiratory infection, bronchiolitis, and otitis media, specifically during
26 certain times of the year depending on geographic location. The relatively small number
27 of studies that have examined hospital admissions and ED visits due to respiratory
28 infections has resulted in an inadequate assessment of the lag structure of associations
29 and potential copollutant confounding.
January 2015 5-163 DRAFT: Do Not Cite or Quote
-------
Study
Mehta et al. (2013)
Lin et al. (2005)
Faustini et al. (2013)b
Burnett et al. (1999)
Karr et al. (2006)
Segala et al. (2008)
Zanobetti and Schwartz (2006)
Stieb et al. (2009)
Peel et al. (2005)
Segala et al. (2008)
Peel et al. (2005)
Zemek et al. (2010)
Location
Ho Chi Minh, Vietnam
Toronto, CAN
6 Italian cities
Toronto, CAN
LA County, CA
Paris, France
Boston, MA
7 Canadian cities
Atlanta, GA
Paris, France
Atlanta, GA
Edmonton, Canada
Age
28 days - 5
0-14
35+
All
0-1
— Pneumonia
• Otitis Media
-10.0 0.0 10.0 20.0 30.0 40.0 50.0
% Increase
Note: a = results are for the dry season (November-April); b = Lower Respiratory Infection in people with COPD. Black = U.S. and
Canadian studies evaluated in the 2008 Integrated Science Assessment for Oxides of Nitrogen; red = recent studies. Circle = all-
year; diamond = warm/summer months; square = cool/winter months. DL = distributed lag. Effect estimates are standardized to a
20-ppb increase in 24-h avg nitrogen dioxide and 30-ppb increase in 1-h max nitrogen dioxide.
Figure 5-10 Percentage increase in respiratory infection-related hospital
admissions and Emergency Department (ED) visits in relation to
nitrogen dioxide concentrations from U.S. and Canadian studies
evaluated in the 2008 Integrated Science Assessment for Oxides
of Nitrogen and recent studies.
January 2015
5-164
DRAFT: Do Not Cite or Quote
-------
Table 5-30 Corresponding risk e
Study
Location
estimate for studies presented in Figure 5-10.
Age Avg Time Lag Season
% Increase
(95% Cl)
Hospital Admissions
Respiratory Infection
tMehtaetal. (2013)
Lin et al. (2005)
tFaustini et al. (201 3)b
Burnett et al. (1999)
Ho Chi Minh,
Vietnam
Toronto, Canada
6 Italian cities
Toronto, Canada
28days-5 24-h avg 1-6 All
Dry*
0-14 24-h avg 0-5 All
35+ 24-h avg 0-5 DL All
All 24-h avg 2 All
-4.0 (-18.0, 12.5)
35.9(3.0,79.3)
41.1 (15.6,73.7)
6.9 (-4.3, 19.4)
5.4(3.5, 7.4)
Bronchiolitis
Karretal. (2006)
tSeqalaetal. (2008)
LA County, CA
Paris, France
0-1 1-h max 1 Winter
<3 24-h avg 0-4 Winter
O C / C Q -1 O\
15.9(7.7,29.0)
Pneumonia
Zanobetti and Schwartz
(2006)
Boston, MA
65+ 24-h avg 0-1 All
2.7 (-3.0, 8.4)
ED Visits
Respiratory Infection
tStieb et al. (2009)
Peel et al. (2005)
7 Canadian cities
Atlanta, GA
All 24-h avg 2 All
All 1-h max 0-2 All
0.7 (-1.5, 2.8)
2.9(0.9,4.7)
Bronchiolitis
tSeqalaetal. (2008)
Paris, France
<3 24-h avg 0-4 Winter
11.8(7.7,20.1)
Pneumonia
Peel et al. (2005)
Atlanta, GA
All 1-h max 0-2 All
0.0 (-2.5, 2.9)
Otitis Media
tZemeketal. (2010)
Edmonton,
Canada
1-3 24-h avg 2 All
Summer
Winter
7.9(1.6, 12.8)
16.1 (3.1, 31.2)
4.7(0.0, 11.2)
Cl = confidence interval, DL = distributed lag, ED = emergency department.
aDry season was defined as November-April.
"LRTI in people with COPD.
fStudies published since the 2008 ISA for Oxides of Nitrogen.
January 2015
5-165
DRAFT: Do Not Cite or Quote
-------
5.2.5.4 Subclinical Effects Underlying Respiratory Infections
1 Overall, NO2 exposure shows effects in varying directions on subclinical effects that may
2 characterize key events within the mode of action for respiratory infections (Figure 4-1).
3 Some support for the effects of NO2 on respiratory infection morbidity and mortality
4 observed in toxicological studies and some epidemiologic studies is provided by
5 toxicological findings for NO2-induced impairments in alveolar macrophage function.
6 There is uncertainty about the effects of NO2 on alveolar macrophages and
7 immunoglobulin antibody responses as examined in controlled human exposure and
8 epidemiologic studies, respectively.
Mucociliary and Alveolar Clearance
9 Airborne substances small enough to be respired may be trapped in the epithelial lining
10 fluid in the conducting airways and physically removed or cleared from the airway by
11 ciliated epithelial cells. Recent animal toxicological studies and studies included in the
12 2008 ISA for Oxides of Nitrogen (U.S. EPA. 2008a) demonstrated that exposure to high
13 concentrations of NO2, generally above 5,000 ppb, functionally impairs pulmonary
14 clearance and damage the ciliated epithelium of the airway. However, exposures to NO2
15 concentrations below 5,000 ppb have varying effects on pulmonary clearance in animal
16 toxicological and controlled human exposure studies. The examination of the effect of
17 NO2 on pulmonary clearance, which consists of mucociliary and alveolar clearance, is
18 limited to studies that were reviewed in the 2008 ISA.
19 Studies have been conducted in various animal models and provide evidence that NO2
20 exposure can potentially affect mucociliary clearance. Schlesinger (1987b) employed two
21 methods to measure ciliary clearance in rabbits exposed to 310 or 1,030 ppb NO2 for
22 2 hours per day for up to 14 days. Mean residence time of radioactive tracer microspheres
23 was not altered 24 hours following 2-, 6-, or 13-day exposures; however, patterns in
24 clearance, measured as the fraction of retained radioactive tracer microspheres, were
25 statistically significantly different from those in controls at both 310 or 1,030 ppb NO2
26 over 13 days of exposure. Vollmuth et al. (1986) studied mucociliary clearance in rabbits
27 exposed to 300 or 1,000 ppb NO2 for 2 hours while Ferin and Leach (1975) exposed rats
28 to 1,000 ppb NO2 in conjunction with 900 ppb NO for 7 hours per day, 5 days per week
29 for 11 or 22 days; both studies reported accelerated clearance of particles. A study
30 published by Ohashi et al. (1994) found different results, showing that guinea pigs
31 exposed to 3,000 or 9,000 ppb NO2 for 6 hours/day, 6 days/week for 2 weeks had
32 concentration-dependent reductions in ciliary activity. This study, however, reported
January 2015 5-166 DRAFT: Do Not Cite or Quote
-------
1 ciliary beat (measured by light refraction) in nasal tissues excised after animals were
2 exposed. This method could have affected the outcome and be less representative of
3 changes that occur from human ambient exposure.
4 In a controlled human exposure study of healthy adults, Helleday et al. (1995) used
5 fiberoptic bronchoscopy to measure ciliary activity and found a decrease 45 minutes after
6 a brief exposure to 1,500 or 3,500 ppb NC>2. In contrast, increases in ciliary activity were
7 reported 24 hours after a 4-hour exposure to 3,500 ppb NCh. It is important to note that
8 baseline measurements for each subject in this study were used as control values, and
9 therefore, the study lacked air controls and subject blinding.
Function and Morphology of Alveolar Macrophages
Toxicological Studies
10 Previous studies reported NO2 exposure to induce slight morphological differences and
11 increases in AM numbers in BAL fluid (Hooftman et al., 1988; Mochitate et al.. 1986;
12 Rombout et al.. 1986; Goldstein et al.. 1977; Powell et al.. 1971) and diminished
13 superoxide radical production (indicating reduced respiratory burst) at exposures as low
14 as 500 ppb (see Table 5-31 for study details). Robisonet al. (1990) and Robison et al.
15 (1993) exposed rat AMs to 100-20,000 ppb for 1 hour ex vivo and found a
16 concentration-dependent decrease in superoxide production, ranging from 81-55% of
17 control levels after phorbol mynstate acetate (PMA) stimulation. Similarly, Sprague
18 Dawley male rats exposed to 500 ppb NO2 for 8 hours/day for 0.5, 1, 5, or 10 days had
19 superoxide levels 63-75% of those in air-exposed animals after PMA stimulation
20 (Robison et al.. 1993). Suzuki et al. (1986) reported comparable observations in AMs
21 isolated from Fisher 344 rats exposed to 4,000 ppb NC>2 for 3, 5, or 10 days. Conversely,
22 PMA-stimulated AMs isolated from Sprague Dawley female rats exposed to NO2 below
23 6,100 ppb showed no change in superoxide production compared to controls (Amoruso et
24 al.. 1981). Overall, NC>2 exposure appears to decrease the ability of AMs to produce
25 superoxide anion, although inconsistencies are present across studies that could be the
26 result of strain or sex differences in response to NC>2.
27 Studies also found variable effects of ambient-relevant NO2 exposures on phagocytic
28 capacity of AMs. Rose et al. (1989) exposed CD-I mice to 1,000 and 5,000 ppb NO2 for
29 6 hours/day for 2 days and reported diminished phagocytosis of colloidal gold particles at
30 both concentrations of NC>2. In contrast, NC>2 exposure increased uptake of murine
31 Cytomegalovirus. Studies report both no change and decreased phagocytosis of latex
32 microspheres. Hooftman et al. (1988) exposed rats to 4,000, 10,000, or 25,000 ppb NO2
33 for 6 hour/day, 5 days/week and found no changes in phagocytosis of latex microspheres
34 below 10,000 ppb at 1, 2, or 3 weeks. Schlesinger (1987b). however, found decreased
January 2015 5-167 DRAFT: Do Not Cite or Quote
-------
1
2
o
6
4
phagocytosis of latex microsphere by AMs isolated from rabbits 24 hours after a 2 or
6-day exposure at 300 or 1,000 ppb [2 hours/day; all animals were co-exposed to
0.5 mg/m3 sulfuric acid (H2SO4)]. Suzuki et al. (1986) also reported decreased phagocytic
capacity of AMs isolated from rats exposed to 4,000 ppb NO2 for 7 days.
Table 5-31 Animal toxicological studies of subclinical lung host defense effects.
Species (Strain);
Study Age; Sex; n
Exposure Details
(Concentration; Duration)
Endpoints Examined
Amoruso et al.
(1981)
Rat (Sprague-
Dawley); F,
n = 4/group
1,300, 1,900, and 6,100 ppb
NO2 for 3 h
Analysis of BAL fluid and superoxide
production by AMs (PMA stimulation).
Powell et al.
(1971)
Dog (beagle);
n = 11
3,000 ppb NO2 for 1 h
Histopathological evaluation and lung
surfactant properties.
Ferin and Leach
(1975)
Rats (Long-Evans),
n = 5-10/group
1,000-24,000 ppb NO2 for 11 Retained TiC>2 particles at 8, 25, and
or 22 days (7 h/day, 130 days post-exposure.
5 days/week)
Goldstein et al.
(1977)
Rat
(Sprague-Dawley); F
500, 1,000, and 2,400 ppb NC>2 Agglutination of AMs.
for 1 and 2 h
Hooftman et al.
(1988)
Rats (Wistar); M;
n = 10/group
4,000, 10,000, 25,000 ppb NO2 Histopathological evaluation, analysis
for 6 h/day, 5 days/week for of BAL fluid, and AM function and
7-21 days morphology.
Mochitate et al.
(1986)
Rats (Wistar);
M; 19-23 weeks;
n = 6/group
4,000 ppb NO2 continuously up BAL fluid cell counts and AM function
to 10 days and morphology.
Ohashi et al.
(1994)
Guinea pigs (Hartley);
n = 10/group
3,000 or 9,000 ppb for 6 h/day, Ciliary beat in excised nasal tissue 24 h
6 days/week for 2 weeks after exposure.
Robison et al.
(1990)
Rats
(Sprague-Dawley)
100, 500, and 1,000 ppb NO2 Viability, LTB4 production, neutrophil
for 1 h; AMs exposed ex vivo chemotaxis, superoxide production.
Robison et al.
Rat (Sprague
Dawley);
n > 4/group
500 ppb NO2 for 8 h/day for
0.5, 1, 5, or 10 days
BAL fluid cell counts and arachidonate
metabolite levels, AM arachidonate
metabolism, respiratory burst activity,
and glutathione content.
Rombout et al.
Rats (Wistar); F, 6
weeks; n = 3-6/group
500, 1,390, and 2,800 ppb NO2 Histopathological evaluation.
for 1,2, 4, 8, 16, and 28 days
January 2015
5-168
DRAFT: Do Not Cite or Quote
-------
Table 5-31 (Continued): Animal toxicological studies of subclinical lung host
defense effects.
Species (Strain);
Study Age; Sex; n
Exposure Details
(Concentration; Duration)
Endpoints Examined
Rose et al.
(1988)
Rose et al.
(1989)
Schlesinqer
(1987b)
Suzuki et al.
(1986)
Vollmuth et al.
(1986)
Mice (CD-1 ); 4-6
weeks; n > 4/group
Rabbits (New
Zealand white);
M, n = 5/group
Rats (Fischer 344);
M, 7 weeks;
n = 8/group
Rabbit (New Zealand
white); M; n = 5/group
(1)1,000,2,500, and 5,000 ppb
NO2 for 6 h/day for 2 days;
intratracheal inoculation with
murine Cytomegalovirus;
4 additional days (6 h/day) of
exposure.
(2) re-inoculation 30 days (air)
post-primary inoculation.
310 or 1,030 ppb NO2 for
2 h/day for 2, 6, and 1 3 days
4,000 NO2 ppb for 1, 3, 5, 7,
and 10 days
300, 1,000, or 3,000 ppb for 2 h
Infection 5 and 10 days
post-inoculation, histopathological
evaluation, and analysis of BAL fluid
(LDH, albumin, macrophages).
Viability and AM activity (mobility,
attachment, and phagocytosis).
AM activity (phagocytosis and
superoxide production), SOD and
glucose-6-phosphate dehydrogenase
activity.
Retained tracer particles for 14 days
following exposure.
LDH = lactate dehydrogenase, M = male, NO2 = nitrogen dioxide, PMA = phorbol mynstate acetate, SOD = superoxide dismutase.
1
2
3
4
5
6
7
8
9
10
11
Controlled Human Exposure
Similar to animal toxicological studies, controlled human exposure studies did not
consistently demonstrate that ambient-relevant NO2 concentrations can alter AM
characteristics (see Table 5-32 for study details). Devlin et al. (1999) exposed healthy
subjects to 2,000 ppb NO2 for 4 hours with intermittent exercise and found that AMs
isolated from the BAL fluid had decreased phagocytic activity and superoxide production
in ex vivo experiments. Conversely, no change in ex vivo macrophage morphology or
function was reported after subjects were exposed to 2,000 ppb NO2 for 6 hours with
intermittent exercise (Azadniv et al.. 1998). In vitro exposure of human AMs for 3 hours
at 5,000 ppb NO2 did not result in statistically significant changes in cell viability or
neutrophil chemotactic factor (IL-8) or IL-1 release, markers of macrophage activity
(Pinkston et al.. 1988).
January 2015
5-169
DRAFT: Do Not Cite or Quote
-------
Table 5-32 Controlled human exposure studies of subclinical lung host defense
effects.
Study
n, Sex; Age
(mean ± SD)
Exposure Details
(Concentration; Duration)
Endpoints Examined
Azadnivet al.
(1998)
n = 11 M, 4F;
Early phase:
28.1 ±3.5yr
Late phase:
27.4 ± 4.2 yr
2,000 ppb for 6 h;
Exercise for approximately 10 of every
30 min at VE = 40 L/min
Alveolar macrophage
function 1 h (early phase)
and 18 h (late phase) after
exposure.
Devlin et al. (1999) n = 11 M; range
2,000 ppbfor4 h;
BAL fluid macrophage
Goings et al.
(1989)
Helledav et al.
(1995)
Pinkston et al.
(1988)
18-35yr
(1) n = 44
(2) n = 43
(3) n = 65; range:
18-35yr
n = 14 M, 10 F;
27 yrs (range:
23-30 yr)
Human AMs isolated
from 14 M and 1 F;
29±3.9yr
Exercise for 1 5 min on/1 5 min off at
VE = 50 L/min
(1)2,000ppbfor2h
(2)3,000ppbfor2h
(3) 1,000 or 2,000 ppb for 2 h
(1) 1,500 ppb for 45 min
(2) 3,500 ppb for 45 min
(3) 3,500 ppb for 4 h
Baseline obtained 2 weeks prior (each
subject served as own control).
5,000 ppb for 3 h (ex vivo)
superoxide production and
phagocytosis.
Nasal wash virus isolation
and count 4 days after virus
administration. Serum and
nasal wash antibody
response 4 weeks after virus
administration.
Fiberoptic bronchoscopy to
record mucociliary activity
frequency.
(1) and (2) 45 min following
exposure.
(3) 24 h following exposure.
Cell viability and release of
neutrophil chemotactic factor
and IL-1.
F = female, IL = interleukin, M = male, SD = standard deviation.
1
2
3
4
5
6
1
8
9
10
Immunoglobulin Antibody Response
Immunoglobulin M antibodies increase in response to infections, and a recent
epidemiologic study of adults infected with human immunodeficiency virus and
hospitalized for pneumocystis pneumonia found a 34% (95% CI: 6.5, -60) diminished
antibody response to pneumocystis proteins per 20-ppb increase in 24-h avg ambient NC>2
concentrations (lag 0-2 day avg) (Blount et al.. 2013). Potential confounding by other
traffic-related copollutants or factors such as meteorology, sex, and SES was not
examined. Further, because subjects were distributed at varying distances of the single
central site monitor, which was located within 1 km of major roads, the impact of
exposure measurement error on the results is uncertain. Thus, the results do not strongly
inform the understanding of the effects of NC>2 on respiratory infections.
January 2015
5-170
DRAFT: Do Not Cite or Quote
-------
5.2.5.5 Summary of Respiratory Infection
1 Animal lexicological studies provide clear evidence for short-term NCh exposure
2 impairing host defense by demonstrating increased mortality from bacterial or viral
3 infection following exposures of experimental animals to 1,500 to 4,500 ppb NC>2 for 1 to
4 8 hours. Several studies also demonstrated decreased bactericidal activity following
5 exposures of 1,000 to 5,000 ppb for 1 to 17 hours. Compared with animal toxicological
6 studies, controlled human exposure studies provide less consistent evidence for
7 NC>2-induced infectivity assessed as viral titers or inactivation of influenza virus. In
8 humans, NC>2 exposures spanned 600-3,000 ppb for 3 hours for a single or 3-day
9 exposure (Table 5-27). The evidence from animal toxicological studies provides
10 biological plausibility for the associations observed in epidemiologic studies between
11 increases in ambient NC>2 concentrations (5- to 7-day avg) and increases in respiratory
12 infections as ascertained by hospital admissions, ED visits, and parental reports. Studies
13 varied in the specific respiratory infection examined (e.g., bronchiolitis, ear infection, any
14 respiratory infection), and many studies observed null or imprecise associations with
15 wide 95% CIs (Figure 5-10). Epidemiologic associations were observed in children and
16 in study populations with respiratory disease (i.e., children with asthma, adults with
17 COPD). Whereas the association between NO2 and LTRI in adults with COPD was
18 robust to adjustment for PMio (Faustini et al., 2013). most epidemiologic studies did not
19 examine copollutant models, and respiratory infections also were associated with highly
20 correlated copollutants such as BS, PMio, and SO2 (r = 0.74 to 0.92).
21 Also providing some biological plausibility for NCh-induced impaired host defense, some
22 studies characterized potential mechanisms underlying susceptibility to infection.
23 Although results vary across studies, some animal toxicological studies found NC>2
24 exposure to decrease the ability of AMs to produce superoxide anion and decrease
25 phagocytic activity. Such observations were made with NC>2 exposures of 300 to
26 5,000 ppb (Table 5-31). There was heterogeneity across studies in animal species, strain,
27 and sex that may or may not have contributed to inconsistencies observed in response to
28 NC>2. Results for the effects of NC>2 exposure on pulmonary clearance were more
29 variable, with a majority of studies reporting increased pulmonary clearance after
30 ambient-relevant NO2 exposure.
5.2.6 Aggregated Respiratory Conditions
31 In addition to individual respiratory conditions, epidemiologic studies examined
32 respiratory effects as an aggregate of multiple respiratory conditions (e.g., asthma,
33 COPD, respiratory infections). The studies from the 2008 ISA for Oxides of Nitrogen
January 2015 5-171 DRAFT: Do Not Cite or Quote
-------
1 (U.S. EPA. 2008a) and recent studies consistently show associations between short-term
2 increases in ambient NO2 concentration and increases in aggregated respiratory
3 conditions. This evidence is based primarily on hospital admissions and ED visits for all
4 respiratory diseases, which are the focus of the discussion in this section. Other outcomes
5 include lung function in adults with asthma or COPD and sales of medication for asthma
6 and COPD combined or cough and mucus combined. As described in preceding sections,
7 evidence for the effects of short-term NO2 exposure varies among specific respiratory
8 outcome groups. Thus, it is not clear whether the evidence for aggregated respiratory
9 conditions reflects associations with each respiratory condition equally or a particular
10 condition(s).
5.2.6.1 Respiratory Symptoms, Lung Function, and Medication
Use
11 Outcomes such as lung function decrements in adults with asthma and/or COPD (Rice et
12 al.. 2013; Higgins et al.. 2000) and increases in the sale of medication for asthma and
13 COPD combined or for cough and mucus combined (Pitard et al.. 2004; Zeghnoun et al..
14 1999) were associated with ambient NO2 (24-h avg, lagged 0 to 7 days or 0-1 day avg) in
15 a small group of epidemiologic studies, with exception of the Higgins et al. (1995) study.
16 However, uncertainties in these studies result in weak inference of the independent
17 effects of NO2. Associations with medication sales were modeled with GAM in S-plus
18 (Pitard et al.. 2004). which can produce biased results (U.S. EPA. 2006). In the
19 Framingham cohort study, lung function was associated with PIVb 5 (r = 0.63) (Rice et al..
20 2013). and a copollutant model was not analyzed. The other lung function studies did not
21 report what potential confounding factors were examined (Higgins et al.. 2000; 1995).
22 Another uncertainty is potential exposure measurement error produced by the use of
23 central site ambient concentrations to represent ambient exposure. In the Framingham
24 study, sites in the Boston, MA area were averaged (Rice et al.. 2013). With one to two
25 observations per subject collected over 3-9 years, the analysis relied on both temporal
26 and spatial contrasts in exposure. With individuals distributed across a 40 km area, and
27 variability in ambient NO2 observed across a range of 3 to 10 km in Boston
28 (Section 2.5.2). is not clear how well the average area concentration represents ambient
29 exposure among study subjects.
January 2015 5-172 DRAFT: Do Not Cite or Quote
-------
5.2.6.2 Hospital Admissions and Emergency Department Visits
for All Respiratory Diseases
1 Epidemiologic studies examining the association between short-term NC>2 exposures and
2 respiratory-related hospital admissions or ED visits were not available until after the
3 completion of the 1993 AQCD for Oxides of Nitrogen. As a result, the 2008 ISA for
4 Oxides of Nitrogen (U.S. EPA. 2008a) contained the first thorough evaluation of
5 respiratory morbidity in the form of respiratory-related hospital admissions and ED visits.
6 The majority of the studies evaluated consisted of single-city, time-series studies that
7 examined all respiratory hospital admissions or ED visits with additional cause-specific
8 studies, as discussed in previous sections. Studies of all respiratory hospital admissions
9 and ED visits consistently reported positive associations with short-term NO2 exposures
10 (Figure 5-13 and Table 5-34). These associations were generally found to be robust and
11 independent of the effects of ambient particles or gaseous copollutants (U.S. EPA.
12 2008a). The evidence supporting NO2-associated increases in all respiratory disease
13 hospital admission and ED visits contributed heavily to the 2008 ISA for Oxides of
14 Nitrogen conclusion that "there is a likely causal relationship between short-term
15 exposure to NO2 and effects on the respiratory system" (U.S. EPA. 2008a). The air
16 quality characteristics of the cities and the exposure assignment approach used in each
17 study evaluated in this section are presented in Table 5-33. As detailed in Section 5.2.2.4.
18 other recent studies of all respiratory disease hospital admissions and ED visits are not
19 the focus of this evaluation, and the full list of these studies, as well as study details, can
20 be found in Supplemental Table S5-3 (U.S. EPA. 2014h).
January 2015 5-173 DRAFT: Do Not Cite or Quote
-------
Table 5-33 Mean and upper percentile concentrations of nitrogen dioxide in
studies of hospital admissions and emergency department visits for
aggregrated respiratory conditions.
Study
Yanq et al.
(2003)
Fund et al.
(2006)
Burnett et al.
(2001)
fCakmak et al.
(2006)
tWonq et al.
(2009)
fDales etal.
(2006)
Location
(Years)
Vancouver,
Canada
(1986-1998)
Vancouver,
Canada
(1995-1999)
Toronto,
Canada
(1980-1994)
10 Canadian
cities
(1993-2000)
Hong Kong
(1996-2002)
11 Canadian
cities
(1986-2000)
Exposure
Assignment Metric
Average of NO2 24-h avg
concentrations
from
30 monitors.
Average of NO2 24-h avg
concentrations
over all
monitors.
Average of NO2 1-h max
concentrations
from 4 monitors.
Average of NO2 24-h avg
concentrations
over all monitors
within each city.
Average of NO2 24-h avg
concentrations
across
8 monitors.
Average of NO2 24-h avg
concentrations
over all monitors
within each city.
Upper
Mean Percentile
Concentration Concentrations Copollutant
ppb ppb Examination
18.7 NR Correlations (r):
O3: -0.32
Copollutant
models: O3
16.8 Max: 33.9 Correlations (r):
CO: 0.74
CoH; 0.72
O3: -0.32
SO2: 0.57
PMio: 0.54
PlVh.s: 0.36
PMio-2.s: 0.52
Copollutant
model: none
44.1 146 Correlations (r):
O3: 0.52
Copollutant
model: Os
21.4 Max: 44-134 Correlations (r):
NR
Copollutant
models: none
31.2 75th: 37.0 Correlations (r):
Max: 89.4 NR
Copollutant
models: none
21.8 95th: 21 -43 Correlations (r),
across cities:
PMio: -0.26 to
0.69
O3: -0.55 to
0.05SO2:
0.20-0.67
CO: 0.13-0.76
Copollutant
models: none
January 2015
5-174
DRAFT: Do Not Cite or Quote
-------
Table 5-33 (Continued): Mean and upper percentile concentrations of nitrogen
dioxide in studies of hospital admissions and emergency
department visits for aggregated respiratory conditions.
Study
fSon et al.
(2013)
fAtkinson et al.
(2012)
tFaustini et al.
(2013)
Peel et al.
(2005)
Location
(Years)
8 South
Korean cities
(2003-2008)
Meta-
analysis
(Asia)
(Years NR)
6 Italian
cities
(2001-2005)
Atlanta, GA
(1993-2000)
Exposure
Assignment Metric
Hourly ambient 24-h avg
NO2
concentrations
from monitors in
each city.
NR 24-h avg
Average of NO2 24-h avg
concentrations
over all monitors
within each city.
Number of NO2
monitors in each
city ranged from
1-5.a
Average of NO2 1-h max
concentrations
from monitors
for several
monitoring
networks.
Upper
Mean Percentile
Concentration Concentrations Copollutant
ppb ppb Examination
11.5-36.9 NR Correlations (r):
PMio: 0.5
O3: -0.1SO2:0.6
CO: 0.7
Copollutant
models: none
NR NR Correlations (r):
NR
Copollutant
models: none
24.1-34.6 NR Correlations (r),
across cities:
PMio: 0.22-0.79
Copollutant
models: PMio
45.9 NR Correlations (r):
PlVh.s: 0.46
PMio: 0.49
PMio-2.s: 0.46
UFP: 0.26
PIvh.s Water
Soluble Metals:
0.32
PIvh.s Sulfate:
0.17
PlVh.s Acidity:
0.10
PlVh.s OC: 0.63
PlVh.s EC: 0.61
Oxygenated HCs:
0.30
O3: 0.42
CO: 0.68
SO2: 0.34
Copollutant
models: none
January 2015
5-175
DRAFT: Do Not Cite or Quote
-------
Table 5-33 (Continued): Mean and upper percentile concentrations of nitrogen
dioxide in studies of hospital admissions and emergency
department visits for aggregated respiratory conditions.
Study
Tolbert et al.
(2007)
fDarrow et al.
(2011)
Location Exposure
(Years) Assignment
Atlanta, GA Average of NO2
(1993-2004) concentrations
from monitors
for several
monitoring
networks.
Atlanta, GA Epidemiologic
(1993-2004) analysis used
N02
concentrations
from 1 centrally
located monitor.
Assessment of
spatial
heterogeneity
relied upon all
EPAAQSand
ARIES
monitors.
Mean
Concentration
Metric ppb
1-h max 81.7
1-h max 1-h max: 43
24-h avg 24-h avg: 22
Commute Commute: 21
(7a.m.- Daytime: 17
10a.m., Nighttime: 25
4 p.m.-
7p.m.)
Daytime
(8 a.m.-
7p.m.)
Nighttime
(12a.m.-
6 a.m.)
Upper
Percentile
Concentrations
ppb
306
75th:
1-h max: 53
24-h avg: 28
Commute: 27
Daytime: 22
Nighttime: 35
Max:
1-h max: 181
24-h avg: 74
Commute: 97
Daytime: 82
Night-time: 97
Copollutant
Examination
Correlations (r):
PlVh.s: 0.47
PMio: 0.53
PMio-2.s: 0.4
PIvh.s Sulfate:
0.14
PIvh.s OC: 0.62
PIvh.s EC: 0.64
PIvh.s TC: 0.65
PIvh.s Water
Soluble Metals:
0.32
Oxygenated HCs:
0.24
Os: 0.44
CO: 0.70
SO2: 0.36
Copollutant
models: CO,
PMio, Os
Correlations (r),
for averaging
times specified in
current NAAQS:
CO, 1-h: 0.61
Os, 8-h: 0.34
PIvh.s, 24-h: 0.42
Copollutant
models: none
CO = carbon monoxide, CoH = coefficient of haze, EC = elemental carbon, ED = emergency department, HC = hydrocarbon,
NAAQS = National Ambient Air Quality Standards, NO2 = nitrogen dioxide, NR = not reported, O3 = ozone, OC = organic carbon,
PM = particulate matter, SO2 = sulfur dioxide, TC = total carbon, UFP = ultrafine particles.
aMonitoring information obtained from Colais et al. (2012).
fStudies published since the 2008 ISA for Oxides of Nitrogen
January 2015
5-176
DRAFT: Do Not Cite or Quote
-------
Hospital Admissions
1 Multicity studies conducted in Canada (Cakmak et al.. 2006; Dales et al.. 2006). Italy
2 (Faustini et al., 2013) and Korea (Son etal.. 2013). as well as a single-city study
3 conducted in Hong Kong (Wong et al.. 2009) examined the association between
4 short-term NC>2 concentrations and hospital admissions for all respiratory diseases, each
5 focusing on a different age range (Figure 5-13 and Table 5-34). Additional supporting
6 evidence for an association between short-term NCh exposures and respiratory hospital
7 admissions comes from a meta-analysis of studies conducted in Asian cities (Atkinson et
8 al.. 2012).
9 Cakmak et al. (2006) focused on all ages in 10 Canadian cities with the primary objective
10 of the study being to examine the potential modification of the effect of ambient air
11 pollution on daily respiratory hospital admissions by education and income using a
12 time-series analysis conducted at the city level (the effect modification analysis is
13 discussed in Chapter 7). The authors calculated a pooled estimate across cities for each
14 pollutant using a random effects model by first selecting the lag day with the strongest
15 association from the city-specific models. For NC>2, the mean lag day across cities that
16 provided the strongest association and for which the pooled effect estimate was
17 calculated was 1.4 days. At this lag, Cakmak et al. (2006) reported a 2.3% increase
18 (95% CI: 0.2, 4.5%) in respiratory hospital admissions for a 20-ppb increase in 24-h avg
19 NO2 concentrations. This result is consistent with a study conducted in Hong Kong that
20 examined whether influenza modifies the relationship between air pollution exposure and
21 hospital admissions (Wong et al.. 2009). Wong et al. (2009) observed a 3.2% (95% CI:
22 1.9, 4.5) increase in all respiratory disease hospital admissions for all ages at lag 0-1 days
23 for a 20-ppb increase in 24-h avg NCh concentrations, with an association slightly smaller
24 in magnitude for acute respiratory disease (2.1% [95% CI: -0.1, 4.3]), which comprises
25 approximately 39% of all respiratory disease hospital admissions in Hong Kong. Cakmak
26 et al. (2006) also examined the potential confounding by other pollutants but only
27 through the use of a multipollutant model (i.e., two or more additional pollutants included
28 in the model). These models are difficult to interpret due to the multicollinearity between
29 pollutants and are not evaluated in this ISA.
30 In an additional multicity study conducted in 11 Canadian cities, Dales et al. (2006)
31 focused on NC>2-associated respiratory hospital admissions in neonatal infants (ages
32 0-27 days). The investigators used a statistical analysis approach similar to Cakmak et al.
33 (2006) (i.e., time-series analysis to examine city-specific associations, and then a random
34 effects model to pool estimates across cities). Dales et al. (2006) observed that the mean
35 lag day across cities that provided the strongest association for NO2 was 1 day, which
January 2015 5-177 DRAFT: Do Not Cite or Quote
-------
1 corresponded to 6.5% (95% CI: 3.5, 9.6%) increase in neonatal respiratory hospital
2 admissions for a 20-ppb increase in 24-h avg NO2 concentrations. Similar to Cakmak et
3 al. (2006). Dales et al. (2006) only examined the potential confounding effects of other
4 pollutants on the NO2-respiratory hospital admission association through the use of
5 multipollutant models, which are not informative due to multicollinearity between
6 pollutants.
7 The results of Cakmak et al. (2006) and Wong et al. (2009). which focus on all ages, are
8 further supported by Son etal. (2013). a study that examined the association between
9 short-term exposures to air pollution and respiratory-related hospital admissions in eight
10 South Korean cities. It is important to note that South Korea has unique demographic
11 characteristics with some indicators more in line with other more developed countries
12 (e.g., life expectancy, percentage of population living in urban areas), but because it
13 represents a rapidly developing Asian country, it is likely to have different air pollution,
14 social, and health patterns than less industrialized Asian nations or Western nations that
15 developed earlier (Son etal.. 2013). In a time-series analysis using a two-stage Bayesian
16 hierarchical model, Sonet al. (2013) examined both single-day lags and cumulative lags
17 up to 3 days (i.e., lag 0-3). The authors only presented NO2 results for the strongest lag
18 and observed a 3.6% increase (95% CI: 1.0, 6.1) in respiratory disease hospital
19 admissions at lag 0 for a 20-ppb increase in 24-h avg NO2 concentrations. These results
20 are consistent with those of a meta-analysis of studies conducted in Asian cities by
21 Atkinson et al. (2012). which in a random effects model based on five estimates, reported
22 a 3.5% increase (95% CI: 0.6, 6.5) in respiratory hospital admissions for a 20-ppb
23 increase in 24-h avg NO2 concentrations.
24 Son etal. (2013)did not conduct copollutant analyses; however, similar patterns of
25 associations were observed across pollutants that were moderately [PMio (r = 0.5); SO2
26 (r = 0.6)] to highly correlated [CO (r = 0.7)] with NO2. Son etal. (2013) also examined
27 potential seasonal differences in all respiratory disease hospital-admission associations.
28 The authors reported that the association with NO2 was largest in magnitude during the
29 summer (8.3% [95% CI: 2.8, 14.3], lag 0). However, across the eight cities, NO2
30 concentrations were lowest during the summer season (<20 ppb compared to >24 ppb in
31 the other seasons), which complicates the interpretation of these results.
32 Faustini etal. (2013) focused on examining the relationship between short-term air
33 pollution exposures and respiratory hospital admissions, specifically on the adult
34 population (i.e., individuals 35 years of age and older) in six Italian cities. In a time-series
35 analysis the authors examined the lag structure of associations through single-day lags as
36 well as cumulative lags, using cubic polynomial distributed lags, in an attempt to identify
37 whether the NO2 effect on respiratory-related hospital admissions was immediate (lag 0,
January 2015 5-178 DRAFT: Do Not Cite or Quote
-------
1 lag 0-1 days), delayed (lag 2-5 days), or prolonged (lag 0-3, 0-5 days). The authors
2 reported that NC>2 was most strongly associated with all respiratory hospital admissions at
3 lag 0-5 days (4.6% [95% CI: 0.87, 8.3] for a 20-ppb increase in 24-h avg NO2
4 concentrations), which differs from Cakmak et al. (2006) and Dales et al. (2006) where
5 the strongest effects were observed at lags less than 2 days. However, Faustini et al.
6 (2013) did observe positive associations, although smaller in magnitude (ranging from
7 2.5-2.9%) at the shorter lags (i.e., lag 0 and 0-1 days). Faustini etal. (2013) only
8 examined potential copollutant confounding of NC>2 associations in models with PMio,
9 and reported that the NO2 association with respiratory hospital admissions at lag
10 0-5 days was attenuated slightly, but remained positive (3.3% [95% CI: -1.1, 7.8]).
Emergency Department Visits
11 Studies of ED visits for aggregated respiratory conditions that were evaluated in the 2008
12 ISA for Oxides of Nitrogen were few in number and focused almost exclusively on study
13 populations consisting of all ages, and U.S. studies were limited to Atlanta, GA. Building
14 on the previous studies conducted in Atlanta, GA (Tolbert et al.. 2007; Peel et al.. 2005).
15 Darrow et al. (2011) also examined associations between short-term air pollution
16 exposures and all respiratory ED visits. To examine the association between the various
17 NO2 exposure metrics and respiratory ED visits, the authors conceptually used a
18 time-stratified case-crossover framework in which control days were selected as those
19 days within the same calendar month and maximum temperature as the case day.
20 However, instead of conducting a traditional case-crossover analysis, the authors used a
21 Poisson model with indicator variables for each of the strata (i.e., parameters of the
22 control days). Darrow etal. (2011) only reported results for a 1 day lag in NC>2
23 concentrations. For a 30-ppb increase in 1-h max NC>2 concentrations the authors reported
24 a 1.4% increase (95% CI: 0.8, 2.1) in all respiratory ED visits. These results are slightly
25 smaller than those reported by Peel et al. (2005) and Tolbert et al. (2007). but this could
26 be attributed to the fact that the latter two studies used a multiday average of NO2
27 concentrations (i.e., lag 0-2 days) instead of the single-day lag used in Darrow et al.
28 (2011).
Model Specification—Sensitivity Analyses
29 A question that often arises in the examination of associations between air pollution and a
30 health effect is whether the statistical model employed adequately controls for the
31 potential confounding effects of temporal trends and meteorological conditions. Son etal.
32 (2013). in the study of eight South Korean cities, conducted a sensitivity analysis to
33 identify whether risk estimates changed depending on the df used to control for temporal
34 trends and meteorology covariates (i.e., temperature, humidity, and barometric pressure).
January 2015 5-179 DRAFT: Do Not Cite or Quote
-------
1 Similar to the other respiratory-related hospital admission outcomes examined, the
2 authors reported that the association between short-term NO2 exposures and all
3 respiratory disease hospital admissions was sensitive to using less than 6 df per year to
4 control for temporal trends, but was stable when using 6-10 df per year. Additionally,
5 when varying the number of df used for the meteorology covariates from 3 to 6 df as well
6 as the lag structure (i.e., lag 0 and lag 0-3 days), the NCh association remained robust
7 (i.e., relatively unchanged).
Exposure Assignment
8 In addition to model specification, the method used to assign exposure in epidemiologic
9 studies has been suggested to influence the magnitude and direction of air
10 pollution-health effects associations. As discussed in Section 5.2.2.4, Strickland et al.
11 (2011) examined exposure assignment in the case of asthma ED visits in Atlanta, GA and
12 found that different exposure assignment approaches could influence the magnitude, but
13 not direction of associations. Darrow et al. (2011) also used data from Atlanta, GA to
14 examine the influence of alternative exposure metrics on the association between
15 short-term NC>2 concentrations and all respiratory ED visits along with the spatial
16 variability of each exposure metric.
17 To examine whether all respiratory ED visits associations differed depending on the
18 exposure metric used, Darrow et al. (2011) used five different exposure metrics:
19 (1) 1-h max; (2) 24-h avg; (3) commuting period (7:00 a.m. to 10:00 a.m. and 4:00 p.m.
20 to 7:00 p.m.); (4) daytime avg (8:00 a.m. to 7:00 p.m.); and (5) nighttime avg (12:00 a.m.
21 to 6:00 a.m.). The authors reported relatively consistent results (using an a priori lag of
22 1 day) across exposure metrics with the largest estimate found for the night-time avg and
23 the smallest for the daytime metrics (Figure 5-11). The larger risk estimate for the
24 nighttime metric could be a reflection of NC>2 during this exposure duration being a better
25 surrogate for NCh concentrations on the previous day (Darrow et al.. 2011). The
26 correlation between NC>2 metrics was not as high compared to that for other pollutants
27 examined in the study (i.e., r < 0.80 between 1-h max and all other metrics), but was
28 relatively high for the 24-h avg metric (r = 0.79), which is the other metric for NC>2 often
29 used in epidemiologic studies.
January 2015 5-180 DRAFT: Do Not Cite or Quote
-------
1.09
1.02 -
1.01 -
100
» 0.99 -
Partial
Spearman r.
i
1 0.79 0,59 0,55 0,44
CM
NO.
Note: Partial Spearman correlation coefficient between a priori metrics (shaded in gray) and other pollutant metrics shown above the
x-axis.
Source: Reprinted with permission of Nature Publishing Group (Darrow et al.. 2011).
Figure 5-11 Risk ratio and 95% confidence intervals for associations between
various lag 1 day nitrogen dioxide (NOa) metrics and respiratory
emergency department visits.
i
2
3
4
5
6
7
In the analysis of the spatial correlation of exposure metrics for NO2, Darrow et al.
(2011) found that unlike Os and PM2 5, which were spatially homogenous, there was
evidence that correlations for NO2 metrics decreased dramatically as distance from the
central monitor increased (Figure 5-12). This was especially true for the 1-h max and
nighttime metrics (r < 0.20) at 60 km. The 24-h avg metric was also reduced (r = -0.40),
but not as dramatically as the 1-h max. Although reduced at greater distances, moderate
correlations (r = -0.50) were reported with the central monitor for the daytime and
January 2015
5-181
DRAFT: Do Not Cite or Quote
-------
commute time metrics. Overall, these results suggest evidence of potential exposure
misclassification for NC>2 with increasing distance from the central monitor across
exposure metrics.
4
5
6
7
8
1.0
0.0
1-hr max
24-hr
commute
day
0 10 20 30 40 50 60 70
Distance from Central Monitor (km)
Source: Reprinted with permission of Nature Publishing Group Darrow et al. (2011).
Figure 5-12 Spatial correlations for nitrogen dioxide (NOa) metrics in the
Atlanta, GA area.
As detailed within this section, hospital admission and ED visit studies of all respiratory
diseases consistently report positive associations with short-term increases in ambient
NO2 concentrations. As presented in Figure 5-13 and Table 5-34. associations are
consistently observed in studies evaluated in the 2008 ISA for Oxides of Nitrogen as well
as recent studies.
January 2015
5-182
DRAFT: Do Not Cite or Quote
-------
Study
Fiino f^t ^1 f~)f\f\ft^\
Ung QL 3,1. l^ZUUOJ
Cakmak et al. (2006)
Atkinson et al. (2012)
Sonetal. (2013)
Dales et al. (2006)
Burnett et al. (2001)
Yang et al. (2005)
Faustini et al. (2013)
Wong et al. (2009)
Yang et al. (2005)
Wong et al. (2009)a
Peel et al. (2005)
Tolbert et al. (2007)
Darrow et al. (2009)
Location
Vancouver, CAN
10 Canadian cities
Meta-analysis (Asia)
8 South Korean cities
1 1 Canadian cities
Toronto, CAN
Vancouver, CAN
6 Italian cities
Hong Kong
Vancouver, CAN
Hong Kong
Atlanta, GA
Atlanta, GA
Atlanta, GA
Age
All
/\11
All
All
All
0-27 days
f~ •")
< 2
<" "5
<- j
35+
All
65+
/-C I
DJ +
All
65+
All
All
All
Lag
n 9
u-z
1.4
—
0
1
01
-1
0-5 DL
0-1
0-1
1
0-1
0-1
0-2
0-2
1
-5.0 0
— • —
— •—
• ^
— • —
-•-
Ł
-•—
ED Visi
-•-
-•-
«•
0 5.0 10.0 15.0 20.0 25.0 30.0 35.0
% Increase
Note: Black circles = U.S. and Canadian studies evaluated in the 2008 Integrated Science Assessment for Oxides of Nitrogen;
Red = recent studies, a = This estimate is for acute respiratory diseases, which comprise approximately 39% of all respiratory
disease hospital admissions in Hong Kong. DL = distributed lag. Effect estimates are standardized to a 20-ppb increase in 24-h NO2
or 30-ppb increase in 1-h max NO2.
Figure 5-13 Percentage increase in all respiratory disease hospital
admissions and emergency department (ED) visits in relation to
nitrogen dioxide concentrations from U.S. and Canadian studies
evaluated in the 2008 Integrated Science Assessment for Oxides
of Nitrogen and recent studies.
January 2015
5-183
DRAFT: Do Not Cite or Quote
-------
1
2
o
3
4
5
6
7
Table 5-34 Corresponding risk estimate for studies presented in
Study
Location
Age
Avg Time Lag
Figure 5-13.
% Increase
(95% Cl)
Hospital Admissions
Funq et al. (2006)
tCakmak et al. (2006)
t Atkinson et al. (2012)
tSon etal. (2013)
tDalesetal. (2006)
Burnett etal. (2001)
Yang et al. (2003)
tFaustini et al. (2013)
tWonq et al. (2009)a
Yanq et al. (2003)
Wong et al. (2009)
Emergency Department
Peel et al. (2005)
Tolbert et al. (2007)
tDarrow et al. (2011)
Vancouver, Canada
10 Canadian cities
Meta-analysis (Asia)
8 South Korean
cities
11 Canadian cities
Toronto, Canada
Vancouver, Canada
6 Italian cities
Hong Kong
Vancouver, Canada
Hong Kong
Visits
Atlanta, GA
Atlanta, GA
Atlanta, GA
All
All
All
All
0-27 days
<2
<3
35+
All
65+
65+
All
65+
All
All
All
24-h avg
24-h avg
24-h avg
24-h avg
24-h avg
1-h max
24-h avg
24-h avg
24-h avg
24-h avg
24-h avg
24-h avg
24-h avg
1-h max
1-h max
1-h max
Cl = confidence interval, DL = distributed lag.
aThis estimate is for acute respiratory diseases, which comprise approximately 39% of all
in Hong Kong.
fStudies published since the 2008 ISA for Oxides of Nitrogen.
0-2
1.4
0
1
0-1
1
0-5 DL
0-1
0-1
1
0-1
0-1
0-2
0-2
1
9.1 (1.5, 17.2)
2.3(0.2,4.5)
3.5(0.6,6.5)
3.6(1.0,6.1)
6.5(3.5, 9.6)
13.3(5.3,22.0)
19.1 (7.4,36.3)
4.6(0.9, 8.3)
3.2(1.9,4.5)
4.0(2.4, 5.7)
19.1 (11.2,27.5)
2.1 (-0.1,4.3)
1.7 (-0.6, 4.0)
2.4(0.9,4.1)
2.0(0.5, 3.3)
1.4(0.8,2.1)
respiratory disease hospital admissions
5.2.6.1 Summary of Aggregated Respiratory Conditions
Previous and recent epidemiologic studies consistently indicate associations between
short-term increases in ambient NC>2 concentrations and increases in respiratory effects
aggregated across specific conditions such as asthma, COPD, and respiratory infections.
A majority of the available evidence is for hospital admissions for all respiratory diseases
combined (Figure 5-13 and Table 5-34). with a few additional studies of ED visits for all
respiratory diseases, lung function in adults with asthma or COPD, or medication sales
for unspecified respiratory effects. Associations of NC>2 with respiratory disease hospital
January 2015
5-184
DRAFT: Do Not Cite or Quote
-------
1 admissions and ED visits are observed to be larger for children and older adults; limited
2 evidence points to differences in risk by sex and SES.
3 With respect to the lag structure of associations across studies, evidence indicates that the
4 strongest associations for all respiratory hospital admissions and ED visits are within the
5 first few days after NO2 exposure, specifically lags of 0 to 2 days. An examination of
6 model specification indicated the NO2-respiratory hospital admission relationship was
7 robust to alternative lags and df for weather covariates (Son etal.. 2013). Thus, varying
8 approaches to modeling weather does not appear to be a source of confounding. NO2
9 effect estimates were sensitive to using less than 6 df per year to account for temporal
10 trends, but most studies did not model temporal trends with fewer df. The limited analysis
11 of potential seasonal differences in associations, suggests that NO2 associations with all
12 respiratory disease hospital admissions are stronger during the summer (Son etal.. 2013).
13 In a study of all respiratory disease ED visits, similar associations were observed for
14 1-h max and 24-h avg NO2 (Darrow et al.. 2011).
15 The epidemiologic evidence for associations of NO2 with aggregated respiratory effects is
16 based on exposure assessment from central site monitors. In two study locations, Boston,
17 MA (Section 2.5.2) and Atlanta, GA (Darrow et al.. 2011). between-monitor correlation
18 in ambient NO2 concentration decreased with increasing distance. Thus, it is unclear the
19 extent to which temporal variation in central site NO2 concentrations represent variation
20 in exposure among subjects. Also, studies of aggregated respiratory effects did not
21 thoroughly examine potential confounding by traffic-related copollutants, which in many
22 studies, showed moderate to high (r = 0.61-0.76, Table 5-33) correlations with NO2.
23 Limited evidence indicates that NO2 associations with all respiratory hospital admissions
24 and ED visits persisted in copollutant models with CO or PM2 5 (Tolbert et al.. 2007)
25 (Figures 5-16 and 5-17). However, potential differential exposure measurement error
26 resulting from central site exposure assessment limits inference from the copollutant
27 model results. Further, given the variable nature of evidence for the effects of short-term
28 NO2 exposure among specific respiratory conditions (Sections 5.2.2. 5.2.4. 5.2.5). it is
29 not clear whether the evidence for aggregated respiratory conditions reflects associations
30 with each respiratory condition equally or a particular condition(s).
5.2.7 Respiratory Effects in Healthy Populations
31 Similar to populations with asthma and COPD, an array of respiratory outcomes has been
32 examined in relation to short-term exposure to NCh in healthy populations. The 2008 ISA
33 for Oxides of Nitrogen did not draw inferences specifically about respiratory effects of
34 NO2 exposure in healthy populations (U.S. EPA. 2008a) but described epidemiologic
January 2015 5-185 DRAFT: Do Not Cite or Quote
-------
1 associations of short-term increases in ambient NC>2 concentration with increases in
2 respiratory symptoms and decreases in lung function in children. Evidence from
3 experimental studies varied across outcomes, indicating no effects on respiratory
4 symptoms or lung function in healthy adults. However, NC>2 exposure did affect
5 underlying key events, inducing increases in airway responsiveness and PMNs in healthy
6 adults generally at 1,000 ppb NO2 exposure or higher. Recent evidence, which is from
7 epidemiologic studies, continues to indicate NO2-related respiratory effects in healthy
8 populations, most consistently seen as increases in pulmonary inflammation.
5.2.7.1 Airway Responsiveness in Healthy Individuals
9 The 2008 ISA for Oxides of Nitrogen reported that increases in nonspecific airway
10 responsiveness were observed in the range of 1,500 to 2,000 ppb NCh for
11 3-hour exposures in healthy adults (U.S. EPA. 2008a). Studies of airway responsiveness
12 in healthy individuals were generally conducted using volunteers ages 18 to 35+ years.
13 (Mohsenin. 1988) found that a 1-hour resting exposure to 2,000 ppb NC>2 increased
14 responsiveness to methacholine. A mild increase in responsiveness to carbachol was
15 observed following a 3-hour exposure to 1,500 ppb NC>2 with moderate intermittent
16 exercise (VE = 40 L/min; 10 of 30 minutes) (Frampton et al., 1991). Kulle and Clements
17 (1988) also showed a tendency for greater FEVi decrements from methacholine challenge
18 following 2-hour resting exposures to 2,000 and 3,000 ppb NC>2. Resting exposures to
19 100 ppb NC>2 for 1 hour have not affected carbachol or methacholine responsiveness in
20 healthy subjects (Ahmed etal.. 1983a: Hazucha et al.. 1983). Two meta-analyses of the
21 available literature confirm statistically significant effects of NO2 exposures above
22 1,000 ppb, but not below, on airway responsiveness in healthy individuals (Kjaergaard
23 and Rasmussen. 1996; Folinsbee. 1992). More recent studies of airway responsiveness in
24 healthy individuals following NC>2 exposure are not available.
5.2.7.2 Lung Function Changes in Healthy Populations
25 Compared with evidence for airway responsiveness, the 2008 ISA for Oxides of Nitrogen
26 reported weak evidence for the effects of NO2 exposure on changes in lung function in
27 the absence of a challenge agent in controlled human exposure and epidemiologic studies
28 of healthy adults (U.S. EPA. 2008a). A small body of epidemiologic studies of children
29 in the general population indicated associations between increases in ambient NO2
30 concentration and decrements in lung function measured by supervised spirometry.
31 Several recent studies, which are epidemiologic, contribute inconsistent evidence for
32 ambient NO2-associated lung function decrements in children in the general population.
January 2015 5-186 DRAFT: Do Not Cite or Quote
-------
Epidemiologic Studies of Children in the General Population
1
2
3
4
5
6
7
As in other populations, ambient NC>2 concentrations are more consistently associated
with lung function decrements in children in the general population as measured by
supervised spirometry than by home PEF. However, many studies of supervised
spirometry did not find associations with ambient NC>2 concentrations. Locations, time
periods, and ambient concentrations of oxides of nitrogen for these studies are presented
in Table 5-35. The studies recruited children from schools, supporting the likelihood that
study populations were representative of the population of children in the study areas.
Table 5-35 Mean and upper percentile oxides of nitrogen concentrations in
epidemiologic studies of lung function in the general population.
Study3
Location
Exposure
Metric
Study Period Analyzed
Upper
Mean/Median Percentile
Concentration Concentrations
ppb ppb
Altuqetal. (2014)
Castro et al. (2009)
Moshammer et al.
(2006)
Scarlett et al. (1996)
Linnetal. (1996)
Oftedal et al. (2008)
Padhi and Padhv
(2008)
Eenhuizen et al.
(2013)
Chanq et al. (2012)
Baqheri Lankarani
etal. (2010)
Eskisehir, Turkey
Rio de Janeiro, Brazil
Linz, Austria
Surrey, U.K.
Upland, Rubidoux,
Torrance, CA
Oslo, Norway
West Bengal, India
3 study areas, the
Netherlands
Taipei, Taiwan
Tehran, Iran
Feb-Mar, 2007
May, June,
Sept, Oct 2004
School yr,
2000-2001
June-July 1994
School yr,
1992-1994
Nov 2001 -Dec
2002
June2006-July
2007
Oct 2000-Nov
2001
Dec 1996-May
1997
NR
24-h avg NO2
24-h avg NO2
8-h avg NO2
(12-8 a.m.)
24-h avg NO2
1-h max NO2
24-h avg NO2
24-h avg NO2
24-h avg NO2
indoor
24-h avg NO
indoor
24-h avg NO2
6 day avg NO2
24-h avg NO2
24-h NO
24-avg NOx
Suburban: 9.4,
Urban: 13.0
Traffic: 21. 2
49.2b
NR
9.3b
34.9
33
14.4b
Biomass fuel:
71.7, LPG:27.5
Biomass fuel:
46.7, LPG: 37.7
16.0b
31.8
75.5, 17.6b
51.6, 40.4b
72.9, 38.8b
Max: 13.1
Max: 17.7
Max: 28.2
Max: 115b
NR
75th: 11. 4b
Max: 82
Max: 96
Max: 59.2b
75th:
Biomass fuel:
90, LPG: 44
75th:
Biomass fuel:
55, LPG: 30
75th: 23.2b
Max: 47.9b
75th: 41. 7
Max: 119, 25.5b
Max: 85.1, 11 Ob
Max: 122, 94.7b
January 2015
5-187
DRAFT: Do Not Cite or Quote
-------
Table 5-35 (Continued): Mean and upper percentile oxides of nitrogen
concentrations in epidemiologic studies of lung function
in the general population.
Study3
Steerenberq et al.
(2001)
Peacock et al.
(2003)
Correia-Deur et al.
(2012)
Van DerZee et al.
(2000)
van derZee et al.
(1999)
Ranzi et al. (2004)
Ward et al. (2000)
Roemer et al.
(1998)
Timonen and
Pekkanen (1997)
Schindler et al.
(2001)
Steinvil et al. (2009)
Cakmak et al.
(2011 a)
Lepeule et al.
(2014)
Sonetal. (2010)
Aqarwal et al.
(2012)
Weichenthal et al.
(2011)
Location
Utrecht, the
Netherlands
Bilthoven, the
Netherlands
Rochester upon
Medway, U.K.
Sao Paolo, Brazil
Rotterdam, Nunspeet,
Bodegraven/Reeuwij,
Amsterdam, Meppel,
the Netherlands
Emiglia-Romagna,
Italy
West Midlands, U.K.
Sweden, Germany,
Finland, Hungary,
Norway, Italy, Greece,
Czech Republic, the
Netherlands
Kuopio, Finland
Aarau, Basel, Davos,
Geneva, Lugano,
Montana, Payerne,
Wald, Switzerland
Tel Aviv, Israel
14 Canadian cities
Boston, MA area
Ulsan, Korea
Patiala, Punjab area,
India
Ottawa, ON, Canada
Study Period
Feb-Mar1998
Nov1996-Feb
1997
Apr-Jul 2004
Three winters
1992-1993
"1 QQQ -1 QQ/1
1994-1995
Feb-May 1999
Jan-Mar 1997
May-July 1997
Winter
1993-1994
Feb-Apr1994
NR
Sept 2002-Nov
2007
Mar2006-Mar
2007
1999-2009
2003-2007
Aug-Jan,
2007-2009
NR
Exposure
Metric
Analyzed
24-h
24-h
24-h
24-h
avg
avg
avg
avg
NO2
NO
NO2
NO
24-h avg NO2
1-h max NO2
24-h
24-h
24-h
24-h
24-h
24-h
24-h
24-h
24-h
24-h
24-h
1-mo
avg
avg
avg
avg
avg
avg
avg
avg
avg
avg
avg
avg
NO2
NO2
NO2
NO2
NO2
NO2
NO2
NO2
NO2
NO2
NO2
1-h avg NO2
Mean/Median
Concentration
ppb
28.2b
30.2b
25.5b
7.4b
17.4, 17.1, 19.2
28.5,28.1, 31.8
Mean: 69. 9b
27.1, 17.6b
25.5, 13.3b
25.0, 11. 7b
Urban: 37.0b
Rural: 18.51b
NR
Across
locations:
6.7-39.8b
Urban: 14.9b
Suburban: 7.4b
19.5b
19.3
12.6
20.2b
21.4
For 2008
Aug-Sep: 8.4b
Oct-Nov: 21. 9b
Dec-Jan: 17.4b
High traffic: 4.8
Low traffic: 4.6
Upper
Percentile
Concentrations
ppb
Max:
Max:
Max:
Max:
Max:
Max:
75th:
90th:
Max:
Max:
Max:
NR
NR
NR
NR
Max:
Max:
Max:
75th:
Max:
95th:
95th:
75th:
Max:
NR
Max:
Max:
44.7b
168b
49.
85.
39,
67,
5b
6b
39,
71,
43
98
84.5b
102b
50,
40.
43.
41.
27.
69.
25.
59.
29.
23.
26.
44.
11
10
44.
,2b
4, 28.7b
6, 30.3b
5b
1b
3b
3
9
4
9b
1
8
January 2015
5-188
DRAFT: Do Not Cite or Quote
-------
Table 5-35 (Continued): Mean and upper percentile oxides of nitrogen
concentrations in epidemiologic studies of lung function
in the general population.
Study3
Thaller etal. (2008)
Straketal. (2012)
Dales etal. (2013)
Location
Galveston, TX
Bilthoven, the
Netherlands
Sault Ste. Marie, ON,
Study Period
Summers 2002,
2003, 2004
Mar-Oct 2009
May-Aug 2010
Exposure
Metric
Analyzed
24-h avg NO2
1-h max NO2
5-h avg NOx
5-h avg NO2
10-h avg NO2
Mean/Median
Concentration
ppb
1.2
3.2
36
20
Near steel plant:
Upper
Percentile
Concentrations
ppb
Max: 7.1
Max: 27.7
Max: 96
Max: 34
NR
Canada
(8 a.m.-6 p.m.) 7.1
Distant site: 4.5
NR = not reported, NO = nitric oxide, NO2 = nitrogen dioxide, NOX = sum of NO and NO2, LPG = liquefied petroleum gas.
aStudies presented in order of first appearance in the text of this section.
""Concentrations converted from |jg/m3 to ppb multiplying by 0.532 assuming standard temperature (25°C) and pressure (1 atm).
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
The most informative studies are those examining NC>2 concentrations outdoor schools or
at a central site adjacent to schools, which may represent a component of the subjects'
ambient exposures. These metrics were inconsistently associated with lung function in
children (Altug etal.. 2014; Castro et al., 2009; Moshammer etal., 2006; Scarlettet al.,
1996). The inconsistent evidence does not appear to be related to the health status of the
study population. NC>2 was not associated with lung function in children without
respiratory symptoms (Altug etal.. 2014). but results were inconsistent in groups of
children with prevalence of asthma or wheeze of 5 or 9% (Castro et al.. 2009; Scarlett et
al.. 1996). Associations were found with same-day NCh and NC>2 averaged over 3 to 8
days but were inconsistent for lag day 1. Linn etal. (1996) found that a 20-ppb increase
in lag 0 of central site NO2 was associated with a -5.2 mL (95% CI: -13, 2.3) change in
evening FEVi among children in three southern California communities. These results
have relatively strong inference about an association with ambient NO2 exposure because
daily average personal and ambient NO2 were reported to be well correlated (r = 0.63).
In the studies examining ambient NC>2 metrics representing subjects' school or total
personal exposure, there is uncertainty regarding copollutant confounding. Associations
were found with CO, PMi, and PIVb 5 (Castro et al.. 2009; Moshammer et al.. 2006; Linn
etal.. 1996); other traffic-related pollutants were not examined. The lack of examination
of other traffic-related pollutants particularly weakens the implications of Oftedal etal.
(2008). who observed high correlations among NCh, PM2.5 and PMio estimated by a
dispersion model (r = 0.83-0.95). Potential confounding also is an uncertainty in a recent
study in India that found decreases in lung function in association with indoor NC>2, NO,
and CO from cooking fuel but did not specify model covariates (Padhi and Padhy. 2008).
Linn etal. (1996) did not provide quantitative results and indicated only that NO2 effect
January 2015
5-189
DRAFT: Do Not Cite or Quote
-------
1 estimates lost statistical significance with adjustment for PlVfc 5, which was weakly
2 correlated with NO2 (r = 0.25). Among children in Austria, with pollutants measured at a
3 site adjacent to the school, NO2 effect estimates were unchanged with adjustment for
4 moderately correlated PIVb 5 (r = 0.54). (Moshammer et al.. 2006). A 25-ppb increase in
5 lag 1 of 8-h avg NO2 (12-8 a.m.) was associated with a -4.1% change (95% CI: -6.4,
6 -1.7) in FEVi in the single-pollutant model and a -4.7% change (95% CI: -7.3, -2.0)
7 with adjustment for PM2 5. PM2 5 effect estimates were attenuated or became positive with
8 adjustment for NO2. While these results provide evidence for an independent association
9 with NO2, other model covariates were not specified, and potential confounding by other
10 factors such as weather cannot be assessed.
11 Among studies of supervised spirometry, evidence was inconsistent for associations with
12 NO2 and NO ascertained from central sites (Eenhuizen et al.. 2013; Chang etal.. 2012;
13 Bagheri Lankarani et al.. 2010; Oftedal et al.. 2008; Steerenberg etal.. 200IV Results
14 were inconsistent for PEF as well as FEVi, and no association was found with a measure
15 of airway resistance. Controlled human exposure studies, conducted in healthy adults, do
16 not consistently indicate effects on ambient-relevant NO2 exposures on FEVi (see below)
17 or airway resistance (Section 4.3.2.2). In addition to the inconsistent findings, there is
18 uncertainty as to whether the NO2 concentrations from an average of area central sites or
19 one central site represent the variability in NO2 concentrations across the study area or
20 subjects' ambient exposure, particularly in the many cross-sectional studies that make up
21 the evidence base. Inconsistencies also were found between studies that measured NO2 at
22 sites located 2 km from children's schools (Chang etal.. 2012; Steerenberg et al.. 2001).
23 Repeated measures and cross-sectional studies found associations with adjustment for
24 time-varying factors such as weather as well as between-subject factors such as height,
25 weight, smoking exposure, and SES. However, copollutant confounding was not
26 examined, and lung function also was associated with the traffic-related pollutants CO
27 and BS (Chang et al.. 2012; Steerenberg etal.. 2001) as well as PMio, SO2, and O3.
28 A fairly large body of studies, conducted in various European countries, does not strongly
29 support NO2-associated decrements in PEF in children. These studies were similar to
30 studies of supervised lung function in that they examined populations that included
31 children with respiratory symptoms, asthma, or atopy and measured NO2 concentrations
32 at central sites and schools. Outdoor school NO2 concentrations were associated with an
33 increase in PEF in children with 25% wheeze prevalence (Peacock et al.. 2003).
34 Associations with central site NO2 tended to be positive (Roemer et al.. 1998; Timonen
35 and Pekkanen. 1997) or null (Ranzi et al.. 2004; Ward et al.. 2000; van der Zee etal..
36 1999). A recent study found an NO2-associated decrease in PEF among children that was
37 independent of CO (Correia-Deur et al.. 2012) (Table 5-36). Both NO2 and CO were
38 averaged across multiple city sites.
January 2015 5-190 DRAFT: Do Not Cite or Quote
-------
Table 5-36 Epidemiologic studies of lung function in children and adults in the general population.
Study
Population Examined and Methodological Details
Oxide of Nitrogen
Metrics Analyzed
Effect Estimate (95% Cl) Copollutant
Lag day Single-Pollutant Model3 Examination
Children in the General Population
Linnetal. (1996)
Upland, Rubidoux, Torrance, CA
n = 269, 4th-5th grades
Repeated measures. Supervised spirometry. Examined 1
week/season for 2 yr. Recruitment from schools. 75-90% follow-
up participation across communities. Repeated measures
ANOVA adjusted for year, day, temperature, rain. Time spent
outdoors = 101-136 min across seasons and communities.
NO2-central site
24-h avg
# sites NR, no site in
Torrance
r = 0.63 correlation with
personal NO2
p.m. FEVi: -5.2 (-13, 2.3)
ml_
p.m. FVC:-3.6(-12, 4.6)
ml_
Diurnal change FEVi
-7.8 (-14, -1.5)
Diurnal change FVC:
-2.2 (-9.6, 4.9)
No quantitative results.
NO2 association reported
to lose statistical
significance with PlVh.s
adjustment.
Associations found with
PM2.5, weak for Os.
Weak correlation with
PM2.5. r=0.25.
tCastro et al. (2009)
Rio de Janeiro, Brazil
n = 118, ages 6-15 yr, 18.4% with asthma
Repeated measures. Supervised PEF. Recruitment from school.
Examined daily for 6 weeks. 9-122 observations/subject. No
information on participation rate. Mixed effects model with
random effect for subject and adjusted for weight, height, sex,
age, asthma, smoking exposure, time trend, temperature, relative
humidity.
NO2-school outdoor
24-h avg
School was within 2 km
of homes.
PEF, L/min:
1 0.04 (-0.58, 0.65)
1-2 avg -0.60 (-1.3, 0.14)
1-3 avg -0.83 (-1.7, 0.02)
No copollutant model.
Associations also found
with PM-io. Associations
with CO, SO2 had wide
95% CIs.
Scarlett et al. (1996)
Surrey, U.K.
n = 154, ages 7-11 yr, 9% with wheeze
Repeated measures. Supervised spirometry. Examined daily for
6 weeks. Recruitment from school. No information on
participation rate. Lung function adjusted for machine, operator,
day of week. Individual subject regressions adjusted for
temperature, humidity, pollen. Pooled estimates obtained using
weighting method.
NO2-school outdoor
1-h max
FEVo./s: 0.30% (-0.29, 0.89) No copollutant model.
FVC: 5.5% (-5.1, 17%)
Association found with
PMio.
No to moderate
correlations with NO2.
r= 0.07 for PMio, 0.50
for 8-h max Os.
January 2015
5-191
DRAFT: Do Not Cite or Quote
-------
Table 5-36 (Continued): Epidemiologic studies of lung function in children and adults in the general population.
Study
Population Examined and Methodological Details
Oxide of Nitrogen
Metrics Analyzed
Effect Estimate (95% Cl) Copollutant
Lag day Single-Pollutant Model3 Examination
Moshammer et al. (2006)
Linz, Austria
n = 163, ages 7-10 yr
Repeated measures. Supervised spirometry. Examined every
2 weeks for school yr. Recruitment from schools. No information
on participation rate. GEE model, covariates not specified.
NO2-central site
8-h avg
(12 a.m.-8 a.m.)
Site adjacent to school
0 FEVi: -4.1% (-6.4, -1.7)
FVC: -2.7% (-5.1, -0.33)
With PlVhs:
-4.7% (-7.3, -2.0)
PM2.5 results attenuated
or become positive.
Associations also found
for PM-i, PMm
Moderate correlations
with NO2. r= 0.53 for
PMi, 0.54 for PM2.5, 0.62
for PMm
tAltuqetal. (2014)
Eskisehir, Turkey, Feb-Mar, 2007
n = NR, ages 9-13 yr, no upper respiratory symptoms
Cross-sectional. Supervised spirometry. Recruitment from
schools of participants of a larger study. No information on
participation rate. Logistic regression adjusted for sex, age,
asthma, parental smoking, coal or wood stove use, parental
education, height, weight, daily average temperature.
NO2-outdoor school
24-h avg
1 site at each of
16 schools
0-6 avg FEVi: 0% (-14, 17)
FVC: 3.8% (-7.3, 16)
No copollutant model.
Os associated with PEF
only. Strong inverse
correlation with NO2.
Pearson r= -0.80.
NO2 and PlVh.s reported
to be highly correlated.
tOftedal et al. (2008)
Oslo, Norway
N =2,170, ages 9-10 yr, 5.5% with asthma
Cross-sectional. Supervised spirometry. Recruitment from a birth
cohort. 67% participation, 60% follow-up. Examined subjects had
more "Westernized" parents. Linear regression adjusted for age,
sex, height, BMI, current asthma, early life maternal smoking,
parental ethnicity, education, smoking, and atopy, lag 1-3
temperature, neighborhood variables (% married, % with
income < median, etc.), long-term NO2.
NO2-dispersion model 1-3 avg
NO2-central site 1-7 avg
24-h avg 1-30
1 city site av9
Quantitative results not
reported.
Association observed with
lag 1-3 day avg and
1-7 day avg. Larger effect
estimated for lag 1-30 day
avg.
Central site no association.
No copollutant model.
No association reported
for PM2.5.
Correlations among
pollutants = 0.83-0.95.
Short-term association
attenuated with
adjustment for early or
lifetime NO2.
r= 0.46-0.77 among
NO2 metrics.
January 2015
5-192
DRAFT: Do Not Cite or Quote
-------
Table 5-36 (Continued): Epidemiologic studies of lung function in children and adults in the general population.
Study
Population Examined and Methodological Details
Steerenberq et al. (2001)
Utrecht, Bilthoven, the Netherlands
n = 126, ages 8-13 yr, 28% respiratory disease, 20% allergy
Repeated measures. Supervised PEF. Examined 1/weekfor
7-8 weeks. Recruitment from urban and suburban schools. 65%
participation. Mixed effects model adjusted for sex, age, #
cigarettes smoked in home, presence of a cold, history of
respiratory symptoms and allergy. No consideration for potential
confounding by meteorological factors.
tChanqetal. (2012)
Taipei, Taiwan
n =2,919, ages 12-16 yr
Cross-sectional. Supervised spirometry. Recruitment from
schools. No information on participation rate. Regression model
adjusted for residence in district, age, sex, height, weight,
temperature, rainfall.
tEenhuizen etal. (2013)
3 study areas, the Netherlands
n = 880, age 8 yr
Cross-sectional. Recruitment from intervention study of mattress
allergy covers. Valid data on 49% subjects, who had higher
parental education, less likely to have pets. Linear regression
adjusted for sex, age, height, weight, prenatal smoke exposure,
smoking in home, gas stove, parental allergy, dampness in home,
parental education, season, temperature, humidity.
tBaqheri Lankarani et al. (2010)
Tehran, Iran
n = 562, elementary school age
Oxide of Nitrogen
Metrics Analyzed
NO2-central site
1 5-h avg
(8 a.m.-11 p.m.)
24-h avg
Site within 2 km of
schools
NO-central site
15-h avg
(8 a.m.-11 p.m.)
24-h avg
NO2-central site
4-h avg
(8 a.m.-12 p.m.)
1 0-h avg
(8 a.m.-6 p.m.)
Average of 5 city sites
within 2 km of schools
NO2-central site
1 site
NO-central site
24-h avg
2 city sites
Effect Estimate (95% Cl)
Lag day Single-Pollutant Model3
PEF mL/min:
0 Urban: -17 (-35,0)
Suburban: 7, p >0.05
0-2 avg Urban: 0, p > .05
Suburban: 6, p > 0.05
15-h avg Urban: 1, p> 0.05
0-2 avg Suburban: 0, p > 0.05
Urban: -6 (-12, 0)
Suburban: 6, p > 0.05
FEVi in ml_:
0 -25 (-57, 7.5)
1 -41 (-70, -11)
2 -2.5 (-50, 45)
Interrupter resistance in
kPA*s/L:
0 0 (-0.04, 0.04)
1 -0.02 (-0.06, 0.03)
Positive effect estimate
indicates increase in
resistance
PEF <50% predicted:
0-6 avg OR: 18 (1,326)
Copollutant
Examination
No copollutant model.
BS associated with PEF.
Correlation with NO2and
NONR.
Association also found
with PMm
No copollutant model.
Associations also found
with SO2, CO, 03, PMm
No associations with
PM-ioor BS.
Moderate correlations
with NO2. Pearson
r= 0.47 for PMio, 0.60
for BS.
No copollutant model.
PMio associated with
decreased odds of large
PEF decrement.
Repeated measures. Examined daily for 6 weeks. No information
on participation rate. 158 case-days. Case crossover with control
dates as 2 weeks before and after case date. Conditional logistic
regression adjusted for daily temperature, lag 0-6 day avg PMio.
January 2015
5-193
DRAFT: Do Not Cite or Quote
-------
Table 5-36 (Continued): Epidemiologic studies of lung function in children and adults in the general population.
Study
Population Examined and Methodological Details
Oxide of Nitrogen
Metrics Analyzed
Effect Estimate (95% Cl) Copollutant
Lag day Single-Pollutant Model3 Examination
tPadhiand Padhy(2008)
West Bengal, India
n = 755 from biomass fuel homes, 372 from liquified petroleum
gas homes, ages 5-10 yr
Cross-sectional. Supervised spirometry. Recruitment method and
participation not reported. Multiple regression adjusted for
unspecified covariates.
NO2-indoor home
24-h avg
NR Biomass fuel homes
Lung function units NR
FEVi: -1.05 (-1.75, -0.35)
FVC: -1.09 (-1.58, -0.61)
Liquified gas petroleum
homes
FEVi: -5.41 (-8.33, -2.50)
FVC: -5.17 (-9.17, -1.17)
No copollutant model.
CO also associated with
lung function. Correlation
with NO2NR.
SPM, SO2, Os also
associated with lung
function.
tCorreia-Deur et al. (2012)
Sao Paolo, Brazil
n = 31, ages 9-11 yr, no allergic sensitization
Repeated measures. Daily supervised spirometry for 15 school
days. Number of observations not reported. Recruitment from
school. 86% participation. Allergic sensitization ascertained by
skin prick test, blood eosinophils, and serum IgE. GEE with
autoregressive correlation matrix adjusted for date, school
absence, temperature, humidity.
NO2-outdoor school
24-h avg
PEF: -1.0% (-1.7, -0.35)
Lag 0, all subjects
with CO: -1.5% (-3.0, 0)
Moderate correlation
with NO2. r=0.51. CO
association persists with
NO2 adjustment.
withSO2: -1.9% (-3.3,
-0.4)
with PMio: -0.8% (-4.4,
3.1)
withOs: -1.5% (-3.3,
0.38)
Moderate correlations
with NO2. r= 0.59, 0.60,
0.40.
03 association persists
with NO2 adjustment.
SO2 & PMio attenuated.
Peacock et al. (2003)
Rochester upon Medway, U.K.
N = 177, ages 7-13 yr, 25% with wheeze
NO2-outdoor school
24-h avg
0-4 avg PEF: -0.20 (-3.0, 2.6)
OR for PEF > 20%:
2.3(1.0, 5.4)
No copollutant model.
PM2salso associated
with PEF decrement
>20%.
Correlation NR.
Repeated measures. Home PEF. Examined daily for 13 weeks.
14-63 observations/subject. Recruitment from rural and urban
schools. No information on participation rate. Individual subject
regressions adjusted for day of week, date, temperature. Pooled
estimates obtained using weighting method.
1-h max
PEF: 1.2 (-1.5, 3.9)
OR for PEF > 20%:
1.3(0.5, 3.4)
January 2015
5-194
DRAFT: Do Not Cite or Quote
-------
Table 5-36 (Continued): Epidemiologic studies of lung function in children and adults in the general population.
Study
Population Examined and Methodological Details
Oxide of Nitrogen
Metrics Analyzed
Effect Estimate (95% Cl) Copollutant
Lag day Single-Pollutant Model3 Examination
van derZee et al. (1999)
Rotterdam, Bodegraven/Reeuwijk, Amsterdam, Meppel,
Nunspeet, the Netherlands
n = 633, ages 7-11 yr, 63% with symptoms, 26 and 38% with
asthma
Repeated measures. Home PEF. Examined daily for 3 mo.
Recruitment from school and mail. 47% responded to initial
survey, 80% follow-up participation. Logistic regression adjusted
for minimum temperature, day of week, time trend, influenza.
NO2-central site
24-h avg
1 site per community
0-4 avg
ORs
Urban: 0.96(0.79, 1.2)
Suburban: 0.77(0.54, 1.1)
Urban: 1.1 (0.93, 1.3)
Suburban: 0.99(0.72, 1.4)
Associations found for
PM-io, BS, SO4, and SO2
in urban area.
Correlations NR.
Roemeret al. (1998)
Germany, Finland, the Netherlands, Czech Republic, Norway,
Italy, Greece, Hungary, Sweden—26 locations
n = 2,010, ages 6-12 yr, atopy prevalence: 7-81%
Repeated measures. Home PEF. Examined daily for 2 mo. 85%
of enrolled included in analysis. Regression model adjusted for
minimum temperature, school day, time trend. Individual panel
results combined in a meta-analysis.
NO2-central site
24-h avg
PEF, L/min:
0 0.15 (-0.19, 0.49)
0-6 avg 0.23 (-1.2, 1.6)
Association found with
PM-iqand BS, but not
consistently across lags.
Ranzi et al. (2004)
Emiglia-Romagna, Italy
n = 118, ages 6-11 yr, 77% with asthma, 67% with atopy
Repeated measures. Home PEF. Examined daily for 12 weeks.
98.4% follow-up participation. Recruited from schools. GLM
adjusted for sex, medication use, symptoms, temperature,
humidity
NO2-central site
24-h avg
# sites NR
No quantitative data. Figure
shows no association in
group with and without
atopy.
PlVh s associated with
PEF in urban group.
Ward et al. (2000)
West Midlands, U.K.
n = 147, age 9 yr, 24% with symptoms, 31% with atopy
Repeated measures. Home PEF. Examined daily for two 8-week
periods. Recruitment from schools. Individual subject regressions
adjusted for time trend, day of week, meteorological variables,
pollen count. Individual regressions pooled with weighting
method.
NO2-central site
24-h avg
2 sites
0, 1, 2, No quantitative data. Figure
3, 0-4 shows no association
avg across lags, except at lag
day 0 in symptomatic group.
No copollutant model.
Associations with PlVh.s
equally inconsistent.
January 2015
5-195
DRAFT: Do Not Cite or Quote
-------
Table 5-36 (Continued): Epidemiologic studies of lung function in children and adults in the general population.
Study
Population Examined and Methodological Details
Timonen and Pekkanen (1997)
Kuopio, Finland
n = 169, ages 7-12 yr, children with cough
Repeated measures. Home PEF. Examined daily for 3 mo.
Recruitment from schools. 86% participation. Linear mixed model
adjusted for time trend, weekend, minimum temperature, relative
humidity.
Oxide of Nitrogen
Metrics Analyzed
NO2-central site
24-h avg
# sites NR
26% missing data were
modeled, r= 0.58
Effect Estimate (95% Cl)
Lag day Single-Pollutant Model3
0 FEVi:
Urban: 11 (-14, 35)
Suburban: -6.5 (-40, 27)
1-4 avg PEF:
Urban: 13 (-24, 50)
Suburban: -22 (-87, 43)
Copollutant
Examination
Associations found for
SO2 in urban group.
Weak correlations with
NO2. r= 0.22.
Adults in the General Population
tStraketal. (2012)
Utrecht area, the Netherlands
n = 31, adults ages 19-26 yr, all healthy, non-smoking
Repeated measures. Supervised spirometry. Examined 3-7
times. 107 observations. Recruitment from university. No
information on participation rate. Well-defined outdoor exposures
at various traffic/non-traffic sites. Heart rate maintained during
intermittent exercise. Higher probability of associations found by
chance alone. Mixed effects model adjusted for temperature,
relative humidity, season, high/low pollen, respiratory infection.
NO2-personal outdoor
5-h avg
Measured next to
subjects during outdoor
exposures.
NOx-personal outdoor
5-h avg
FVC post-exposure:
0-h
2-h
18-h
0-h
2-h
18-h
-4,
-3,
-4,
-1
-2
-2
.3%
.5%
.5%
.6%
.0%
.5%
(-7.
(-6.
(-7.
(-2.
(-4.
(-5.
4,
5,
4,
,6,
9,
,4,
-1.
-0.
-1.
-0,
-0,
-0,
0)
.43)
4)
.51)
.16)
.69)
FVC with PNC:
NO2: -3.0% (-7.2, 1.4)
NOx: -0.11% (-2.6, 2
Moderate to high
correlation with NO2.
PNC association
attenuated with
adjustment for NO2 or
NOx.
.5)
-7C
ID.
tDalesetal. (2013)
Sault Ste. Marie, Ontario, Canada
n = 59, adults mean (SD) age 24.2 (5.8) yr, all healthy
Repeated measures. Supervised spirometry. Examined 10 times.
Total observations NR. Recruitment from university. No
information on participation rate. Well-defined outdoor exposures
near steel plant and university campus 4.5 km away. Exposures
occurred at rest except for 30-min exercise to increase heart rate
to 60% predicted maximum. Mixed effect model with
autoregressive correlation matrix and adjusted for site, day of
week, mean temperature, humidity.
NO2-on site of outdoor 0-h
exposure Post.
10-h avg exposure
(8 a.m.-6 p.m.)
% predicted FEVi:
-10.9 (-13.3, -8.6)
% predicted FVC:
-9.2 (-14.5,-3.9)
No copollutant model.
Associations found with
UFP and PM2.5.
Correlations NR. All
pollutants higher at steel
plant than at university
campus.
Associations also found
with SO2, 03.
January 2015
5-196
DRAFT: Do Not Cite or Quote
-------
Table 5-36 (Continued): Epidemiologic studies of lung function in children and adults in the general population.
Study
Population Examined and Methodological Details
tWeichenthal et al.
Ottawa, Canada
n = 42, adults ages
white, 62% with alle
(2011)
19-58 yr, from
irgies, 33% with
non-smoking homes, 95%
asthma
Oxide of Nitrogen
Metrics Analyzed
NO2-central site
1-h avg
1 site
Lag day
1-h
4-h
Pnct-
Effect Estimate (95% Cl) Copollutant
Single-Pollutant Model3 Examination
FEVi
0.54 (
0.40 (
in
-o
-o
L:
.15,
.12,
1.2) L
0.92) L
No copollutant model.
Lung function not
associated with Os or
VOCs, UFP, BC, PM2.5.
Repeated measures. Supervised spirometry. Most examined
3 times. 118 observations. 1-h outdoor exposures during cycling
in low and high traffic areas. Recruitment from public
advertisements. No information on participation rate. Differential
exposure measurement error for personal PM and VOCs and
central site NO2. Mixed effects models with random subject effect
adjusted for temperature during cycling, average heart rate.
Adjustment for relative humidity, day of week did not affect
results.
exposure
tThaller et al. (2008)
Galveston, TX
n = 142, lifeguards at work, ages 16-27 yr, 13% with asthma,
22% with allergies
Repeated measures. Supervised spirometry. Recruitment from
worksite. 1,140 observations. Self-report of physician-diagnosed
asthma. 81% follow up participation. GLM, covariates not
specified.
NO2 & N Ox-centra I site
24-h avg, 1-h max
1 site 4-12 km from
beaches
No quantitative data. NO2
and NOx reported not to be
significantly associated with
lung function.
No copollutant model.
Schindleretal. (2001)
Aarau, Basel, Davos, Geneva, Lugano, Montana, Payerne, Wald,
Switzerland
n = 3,912, ages 18-60 yr, non-smokers
Cross-sectional. Supervised spirometry. Recruitment from
registry and SALPADIA cohort. Sample representative of full
cohort. Regression model adjusted for sex, age, height, weight,
day of week, temperature, relative humidity. Adjustment for
asthma medication or wheeze did not alter results.
Van Per Zee etal. (2000)
Rotterdam, Bodegraven/Reeuwijk, Amsterdam, Meppel,
Nunspeet, the Netherlands
NO2-central site
24-h avg
1 site per city
NO2-central site
24-h avg
FEVi:
0 -2.5% (-4.5, -0.48)
0-3 avg -2.9% (-5.9, 0.21)
OR for PEF decrease >10%
0 Urban: 0.85(0.59, 1.2)
Suburban: 0.72(0.50, 1.05)
\A/ith TQD- "1 OOA / Q Q
1.6)
No copollutant model.
PEF associated with
PM-io and SO4 in urban
group.
n = 274, ages 50-70 yr, no symptoms in previous 12 mo
Repeated measures. Home PEF. Examined daily for 3 mo.
Recruitment from mailings. 81% enrolled included in final
analysis. Logistic regression adjusted for minimum temperature,
day of week, time trend, influenza.
1 site per community 0-4 avg
Urban: 0.46(0.20, 1.08)
Suburban: 0.56(0.27, 1.16)
Wide range of
correlations with NO2.
Spearman r= 0.16-0.72
for PMio, 0.25-0.65 for
BS.
January 2015
5-197
DRAFT: Do Not Cite or Quote
-------
Table 5-36 (Continued): Epidemiologic studies of lung function in children and adults in the general population.
Study
Population Examined and Methodological Details
Oxide of Nitrogen
Metrics Analyzed
Effect Estimate (95% Cl) Copollutant
Lag day Single-Pollutant Model3 Examination
tCakmaketal. (2011 a)
15 cities, Canada
n = 5,011, ages 6-79 yr, mean age 39 yr
Cross-sectional. Supervised spirometry. Recruitment by random
sampling of households. No information on participation rate.
GLMM adjusted forage, sex, income, education, smoking,
random effect for site. Adjustment for temperature and relative
humidity did not alter results.
NO2-central site
24-h avg
# sites NR
% predicted FEV-i:
-1.6 (-2.9,-0.35)
No copollutant model.
03 and PlVh.salso
associated with lung
function. Correlations
NR.
tLepeuleetal. (2014)
Boston, MA area
n = 776, all male, mean (SD) age at baseline 72.3 (6.8) yr,
Normative Aging Study
Longitudinal. Supervised spirometry. Examined 1-4 times over
10 yr. No information on recruitment over follow-up participation.
Linear mixed effects model adjusted forage, log-height, race,
education, standardized weight, smoking status, pack-years
smoking, chronic lung condition, methacholine responsiveness,
medication use season, day of week, visit number, temperature,
humidity. Adjusting for cardiovascular disease did not alter
results.
NO2-central site
24-h avg
Average of 5 sites in
Boston area. Median
21.4 km from subjects'
homes.
FEVi:
0 -0.18% (-1.89, 1.57)
0-2 avg -1.62% (-3.89, 0.70)
0-27 -13.0% (-17.9, 7.75)
av9 Low IL-6 gene methylation
-11.6% (-17.5, -5.32)
High IL-6 gene methylation
-13.0% (-18.7, -6.95)
No copollutant model.
Associations found with
BC, CO, PM2.5. Moderate
correlation with NO2.
Spearman r=0.59, 0.52,
0.62, respectively.
BC and PM2.5 measured
at one Boston site.
Association also found
with 03. r= -0.31.
tSteinvil et al. (2009)
Tel Aviv, Israel
n = 2,380, mean age 43 (SD:11) yr, healthy non-smokers
Cross-sectional. Supervised spirometry. Recruitment from
ongoing survey of individuals attending health center. No
information on participation rate. Linear regression adjusted for
sex, age, height, BMI, exercise intensity, education, temperature,
relative humidity, season, year.
NO2-central site
24-h avg
3 sites within 11 km of
homes
FEVi:
0 -16 (-64, 33) mL
5 -55 (-103,-6.3) mL
0-6 avg -97 (-181,-13) mL
w/CO(lag5): -19 (-88,
50)
w/SO2 (lag 5): -7.8 (-72,
56)
SO2 and CO results
persist with adjustment
for NO2.
High correlations with
NO2. Pearson r= 0.75
for CO, 0.70 for SO2.
January 2015
5-198
DRAFT: Do Not Cite or Quote
-------
Table 5-36 (Continued): Epidemiologic studies of lung function in children and adults in the general population.
Study
Population Examined and Methodological Details
Oxide of Nitrogen
Metrics Analyzed
Effect Estimate (95% Cl) Copollutant
Lag day Single-Pollutant Model3 Examination
tSon etal. (2010)
Ulsan, Korea
n = 2,102, mean age 45 (SD: 17) yr, mean % predicted FEV-i:
83%
Cross-sectional. Supervised spirometry. Recruitment during
meeting of residents. No information on participation. Regression
model adjusted for age, sex, BMI. Did not consider potential
confounding by weather, season, or time trend. High correlation
among exposure assessment methods. r= 0.84-0.96.
NO2-central site
13 site average
Nearest site
Inverse distance
weighting
Kriging
All 24-h avg
% predicted FVC:
0-2 avg -7.9 (-10,-5.6)
-6.9 (-8.8, -5.0)
-6.9 (-9.1,-4.7)
Associations found with
PMio, O3, SO2, CO. NO2
effect estimate slightly
reduced with adjustment
for Os. No copollutant
• model with PMio or SO2.
-7.4 (-9.8,-5.1)
tAqarwal et al. (2012)
5 locations with agricultural burning around Patiala City, Punjab,
India.
n = 50, ages 13-53 yr, 80% adults, no respiratory conditions
Repeated measures. Supervised spirometry. Examined
2 times/mo for 6 mo in each of 3 years. Total observations NR.
No information on recruitment method. 40% follow-up
participation. Linear regression. Did not report whether covariates
were included.
NO2-central site
24-h avg
1 site per location
1 mn FFV-r ft Q% n — 0 0^4 Nn rnnnlhitant mnrlpl
av9 FVC: -7.5%, p = 0.064 Association found with
PlVhs. Correlation with
NO2NR.
Association also found
with PMio and SO2.
Note: Studies are organized by population examined, and more informative studies in terms of exposure assessment method and potential confounding considered are presented
first.
GLM = Generalized linear model, BMI = body mass index, ICS = inhaled corticosteroid, SES = socioeconomic status, GEE = generalized estimating equations, Cl = confidence
interval, CO = carbon monoxide, FEVi = forced expiratory volume in 1 second, FVC = forced vital capacity, GLMM = generalized linear mixed model, kPa = kilopascal,
IL = interleukin, NO = nitric oxide, NO2 = nitrogen dioxide, NOX = sum of NO and NO2, NR = not reported, O3 = ozone, OR = odds ratio, PEF = peak expiratory flow,
PM = particulate matter, PNC = particle number concentration, SD = standard deviation, SO2 = sulfur dioxide, TSP = total suspended particles, UFP = ultrafine particles,
VOC = volatile organic compound..
aEffect estimates were standardized to a 20-ppb increase in 24-h avg NO2 and a 30-ppb increase 1-h max NO2. Effect estimates for other averaging times (1-h avg to 15-h avg) are
not standardized but presented as they are reported in their respective studies (Section 5.1.2.3).
fStudies published since the 2008 ISA for Oxides of Nitrogen.
January 2015
5-199
DRAFT: Do Not Cite or Quote
-------
Adults in the General Population
I In studies reviewed in the 2008 ISA for Oxides of Nitrogen (U.S. EPA. 2008a). increases
2 in ambient NC>2 concentration were associated with decrements in lung function in adults
3 in the general population as measured by supervised spirometry (Schindler et al.. 2001)
4 but not home peak flow (Van Per Zee et al.. 2000). Recent studies conducted supervised
5 spirometry, and while the results were inconsistent overall, the studies with stronger
6 exposure assessment and/or examination of copollutant confounding indicated ambient
7 NO2-associated decreases in lung function in healthy adults. Overall, studies examined a
8 wide range of ages (i.e., 18-79 years) and a mix of healthy populations and those
9 including adults with asthma or allergies, but these factors did not appear to influence the
10 results. Van Per Zee et al. (2000) found no association in adults with or without
11 respiratory symptoms. Locations, time periods, and ambient concentrations of oxides of
12 nitrogen for these studies are presented in Table 5-36.
13 While many studies found lung function decrements in adults in the general population in
14 association with higher ambient NO2 concentrations, they do not strongly inform the
15 independent effects of NO2 exposure (Lepeule etal.. 2014; Cakmak et al.. 201 la: Son et
16 al.. 2010; Steinvil etal.. 2009). A major uncertainty is potential confounding. Son et al.
17 (2010) did not examine confounding by meteorological or other time-varying factors.
18 Studies found associations with the traffic-related pollutants CO, BC, and PIVb 5, which
19 were moderately to highly correlated with NO2 (r = 0.56 to 0.75) (Lepeule etal.. 2014;
20 Agarwaletal.,2012: Cakmak etal.. 201 la: Son etal.. 2010; Steinvil etal.. 2009).
21 Copollutant models were not analyzed, except in Steinvil et al. (2009). where the NO2
22 effect estimate was attenuated with CO adjustment, and NO2 results were mixed among
23 the various lags examined. Lung function also was associated with PMio, total suspended
24 particles (TSP), SO2, and Os, and in copollutant models, NO2 associations remained with
25 adjustment for TSP or O3 (Son etal.. 2010: Schindler etal.. 2001) but not for highly
26 correlated SO2 (r = 0.70) (Steinvil et al.. 2009). Another uncertainty pertains to central
27 site ambient NO2, particularly whether NO2 concentrations measured at the nearest site,
28 one city site, averaged across multiple sites, or spatially interpolated by inverse distance
29 weighting or kriging were equally representative of ambient exposure among subjects
30 distributed within a city or across multiple communities. Differences in exposure
31 measurement error between subjects may influence results, particularly in cross-sectional
32 studies (Cakmak etal.. 201 la: Son etal.. 2010: Steinvil et al.. 2009: Schindler et al..
33 2001) and a longitudinal study collecting one to four measures of lung function over 10
34 years (Lepeule et al.. 2014).
35 Ambient concentrations may better represent exposures in situations when people are
36 outdoors. In adults cycling in various traffic and nontraffic locations or lifeguards
January 2015 5-200 DRAFT: Do Not Cite or Quote
-------
1 working outdoors, lung function before and after repeated outdoor exposures were not
2 associated with NO2 assessed from a central site (Weichenthal et al.. 2011; Thaller et al..
3 2008). However, in healthy adults, lung function decrements were associated with NO2
4 measured on site of outdoor activity in locations that varied in traffic (Strak etal.. 2012)
5 or distance from a steel plant (Dales etal.. 2013). Lung function decreased immediately
6 after and 2 to 18 hours after the outdoor exposure period (Dales etal.. 2013; Strak et al..
7 2012). These studies have stronger inference than the aforementioned central site studies
8 because NO2 measurements are aligned with subjects in both time and space. Both
9 outdoor exposure studies found associations with the moderately to highly correlated
10 traffic-related copollutants PM2 5 absorbance, EC, metal components of PM2 5,
11 UFP/particle number concentration (PNC), and/or PM2 5 (r = 0.67-0.87). Only Strak et
12 al. (2012) examined copollutant models and found that NO2 associations persisted with
13 adjustment for PNC, EC, PM2 5, and other PM2 5 components. A 25-ppb increase in
14 5-h avg NO2 was associated with a -4.3% (95% CI: -7.4, -1.0%) change in FVC and a
15 -3.0% (95% CI: -7.2, 1.4%) change with adjustment for PNC. The NOX association was
16 attenuated with adjustment for PNC. Effect estimates for EC, absorbance, and PNC were
17 attenuated with adjustment for NO2, indicating that NO2 may have confounded
18 associations for copollutants.
Controlled Human Exposure Studies
19 Similar to the epidemiologic studies, controlled human exposure studies generally did not
20 report effects of NO2 on lung function in healthy adults. Overall, exposures ranged from
21 200 to 4,000 ppb NO2 for 75 minutes to 6 hours, and most studies incorporated exercise
22 in the exposure period to assess lung function during various physiological conditions
23 (Table 5-37).
24 Huang et al. (2012b) examined the health effects of NO2 exposure alone and in
25 combination with exposure to concentrated ambient particles (CAPs) in healthy adults
26 and did not report any effects on pulmonary function during, immediately after, or
27 18 hours after exposure to 500 ppb NO2 for 2 hours with intermittent exercise. These
28 results are consistent with previously published studies in healthy adults. For example,
29 Hackney et al. (1978) demonstrated that exposure to 1,000 ppb for 2 h/day for
30 2 consecutive days did not induce pulmonary function changes with the exception of a
31 1.5% drop in FVC after exposure on the second day. Similarly, Frampton et al. (1989)
32 reported no differences in lung function before, during, or after exercise or after exposure
33 to 600 or 1,500 ppb NO2 for 3 hours or a 3-hour base of 50 ppb NO2 with intermittent
34 peaks of 2,000 ppb. These results were replicated in studies in healthy adults at similar
35 concentrations (Frampton et al.. 2002; Devlin etal.. 1999). Rasmussen et al. (1992)
36 reported that healthy subjects exposed to 2,300 ppb NO2 for 5 hours had slight
January 2015 5-201 DRAFT: Do Not Cite or Quote
-------
1
2
o
6
4
5
6
7
8
9
10
11
12
improvements, though not statistically significant, in FVC and FEVi during and after
NO2 exposure compared to air. Blomberg et al. (1999) reported a decrease in FEVi and
FVC after the first exposure to 2,000 ppb NO2 for 4 hours. This response was attenuated
after a second, third, and fourth exposure.
Other studies examined co-exposures to NO2 and Os, which often are weakly correlated
in the ambient air (Figure 3-6). Lung function was not affected by a 2-hour exposure to
600 ppb NO2, but statistically significant reductions in FEVi and forced expiratory flow
were observed after a subsequent Os exposure (HazuchaetaL 1994). Exposure of
aerobically trained young men and women to 600 ppb NC>2 or 600 ppb NC>2 + 300 ppb O
for 1 hour resulted in an increase in airway resistance with co-exposure, although the
increase in resistance with co-exposure was far less than with Os exposure alone, and
NO2 exposure alone did affect lung function (Adams etal.. 1987).
Table 5-37 Controlled human exposure studies of lung function and respiratory
symptoms in healthy adults.
Disease Status3; n;
Study Sex; Age (mean ± SD)
Exposure Details
(Concentration; Duration)
Endpoints Examined
Adams et al.
(1987)
(1-3)n=20M, 20 F;
F = 21.4± 1.5yr
M = 22.7 ± 3.3
(1)600ppbNO2for1 h,
(2) 300 ppb O3 for 1 h,
(3) 600 ppb NO2 and 300 ppb Os for 1 h;
(1-3) Exercise during entire exposure at
VE = 75 L/min (M) and VE = 50 L/min (F)
Pulmonary function before
and after exposure.
Symptoms following
exposure.
Blomberq et al.
(1999)
n = 8 M, 4 F; 26 yr
(range: 21-32 yr)
2,000 ppb, 4 h/day for 4 days; Exercise
15 min on/15 min off at workload of
75 watts
Pulmonary function before
and after exposure.
Devlin et al.
(1999)
n = 11 M (range: 18-35
yr)
2,000 ppb for 4 h;
Exercise for 15 min on/15 min off at
VE = 50 L/min
Aerosol bolus dispersion
(deposition, FEVi and
Sraw).
Frampton et al.
(1989)
(1)n = 7M, 2F; 30 yr
(range: 24-37 yr)
(2)n = 11 M, 4F;25yr
(range: 19-37 yr)
(1) 600 ppb for 3 h,
(2) 1,500 ppb for 3 h;
(1,2) Exercise 10 min on/20 min off at
VE = ~4 times resting
Pulmonary function tests
before, during, and after
exposure.
Frampton et al.
(1991)
(1)n = 7M, 2F;
29.9 ± 4.2 yr
(2)n = 12M, 3F;
25.3 ± 4.6 yr
(3) n = 11 M, 4 F;
23.5 ± 2.7 yr
(1) 600 ppb for 3 h,
(2) 1,500 ppb for 3 h,
(3) 50 ppb for 3 h + 2,000 ppb peak for
15 min/h;
(1-3) Exercise 10 min on/20 min off at
VE = ~4 times resting
Pulmonary function tests
before, during, and after
exposure, airway reactivity
30 min
post-exposure.
Symptoms following
exposure.
January 2015
5-202
DRAFT: Do Not Cite or Quote
-------
Table 5-37 (Continued): Controlled human exposure studies of lung function and
respiratory symptoms in healthy adults.
Disease Status3; n;
Study Sex; Age (mean ± SD)
Exposure Details
(Concentration; Duration)
Endpoints Examined
Frampton et al.
(2002)
(1,2) n = 12 M, 9F;
F = 27.1 ±4.1 yr
M = 26.9 ± 4.5 yr
(1)600ppbfor3h,
(2)1,500ppbfor3h;
(1,2) Exercise 10 min on/20 min off at
VE = 40 L/min
Pulmonary function tests
before and after exposure.
Gong et al.
(2005)
Healthy; n = 2 M, 4 F;
68 ± 11 yr
COPD; n = 9 M, 9 F;
72 ± 7 yr
(1)400ppbNO2for2 h
(2) 200 ug/m3 CAPs for 2 h
(3) 400 ppb NO2 + 200 ug/m3 CAPs for 2 h
(1-3) Exercise 15 min on/15 min off at
VE = ~2 times resting
Pulmonary function tests
before and immediately
Symptoms before, during,
and after exposure
Hackney et al.
(1978)
n = 16M;26.9±5.0yr
1,000 ppb, 2 h/day for 2 days; Exercise
15 min on/15 min off at VE = 2 times
resting
Pulmonary function tests
before and after each
exposure.
Symptoms after each
exposure.
Hazucha et al.
(1994)
n=21 F; 22.9 ± 3.6 yr
(1) 600 ppb NO2 for 2 h, air for 3 h,
300 ppb 03 for 2 h,
(2) air for 5 h, 300 ppb Os for 2 h;
(1,2) Exercise for 15 min on/15 min off at
VE = 35 L/min
Pulmonary function tests
before, during, and after
exposure; airway reactivity
after exposure.
Times for symptoms
assessment not reported.
Huang et al.
(2012b)
(1)n = 11 M, 3F
(2) n = 6 M, 7 F
(3) n = 7 M, 6 F;
24.6 ± 4.3 yr
(1) 500 ppb NO2 for 2 h, Pulmonary function tests
(2) 500 ppb NO2 + 73.4 ± 9.9 ug/m3 CAPs before, immediately after,
for 2 h and 18 n after exposure.
(3) 89.5 ± 10.7ug/m3for2h;
(1-3) Exercise 15 min on/15 min off at
VE = 25 L/min
Koeniq et al.
(1987)
Healthy;
(1)n = 3M, 7F
(2) n = 4 M, 6 F
Asthma;
(1)n = 4M,6F
(2) n = 7 M, 3 F
14.4 yr
(range: 12-19 yr)
(1)120ppbNO2,
(2)180ppbNO2;
(1-2) Exposures were 30 min at rest with
10 min of moderate exercise
Pulmonary function tests
before, during, and after
exposure.
Jorres et al.
Healthy; n = 5 M, 3 F;
27 yr (range: 21-33)
Asthma; n = 8 M, 4 F;
27 ± 5 yr
1,000 ppb for 3 h;
Exercise 10 min on/10 min off at
individual's maximum workload
Symptoms immediately
and 6 and 24 h after
exposure.
January 2015
5-203
DRAFT: Do Not Cite or Quote
-------
Table 5-37 (Continued): Controlled human exposure studies of lung function and
respiratory symptoms in healthy adults.
Disease Status3; n;
Study Sex; Age (mean ± SD)
Exposure Details
(Concentration; Duration)
Endpoints Examined
Linnetal. Healthy; n = 16 M, 9 F
(1985b) (range: 20-36 yr)
Asthma; n = 12 M, 11 F
(range: 18-34 yr)
4,000 ppbfor75 min;
Two 15-min periods of exercise at
VE = 25 L/min and 50 L/min
Airway resistance before,
during, and after
exposure.
Symptoms before, during,
and after exposure and
24-h
post-exposure.
Morrow et al.
(1992)
Healthy; n = 10M, 10 F
(13 never smokers,
4 former smokers,
3 current smokers)
COPD; n = 13M, 7F
(14 current smokers,
6 former smokers);
59.9 ± 7.0 yr
300 ppb for 4 h;
Three 7-min periods of exercise at
VE = ~4 times resting
Pulmonary function tests
before, during, and after
exposure and 24-h
post-exposure.
Rasmussen et
al. (1992).
n = 10M, 4F; 34.4 yr
(range: 22-66 yr)
2,300 ppb for 5 h
Pulmonary function tests
before, 2 times during,
and 3 times after
exposure.
Symptoms before, during,
and after exposure.
Vaqaqqini et al.
(1996)
Healthy: n = 7 M;
34 ± 5 yr
Asthma: n = 4 M, 4 F;
29 ± 14 yr
COPD: n = 7 M;
58 ± 12 yr
300 ppb for 1 h;
Exercise at VE = 25 L/min
Pulmonary function tests
before and 2 h after
exposure.
Witten et al.
n = 15; 32±8.6yr
400 ppb for 3 h;
Exercise 30 min on/15 min off at
VE = 25 L/min;
Inhalation challenge with house dust mite
antigen after NO2 exposure.
Symptoms before and
after exposure and 6 h
after allergen challenge.
CAPS = concentrated ambient particles, COPD = chronic obstructive pulmonary disease, F = female, FENA, = forced expiratory
volume in 1 second, M = male, NO2 = nitrogen dioxide, O3 = ozone, SD = standard deviation.
aSubjects were healthy individuals unless described otherwise.
5.2.7.3 Respiratory Symptoms in Healthy Populations
Epidemiologic Studies of Children in the General Population
Respiratory symptoms in relation to short-term NC>2 exposure have not been examined in
epidemiologic studies of healthy adults; however, associations are indicated in
January 2015
5-204
DRAFT: Do Not Cite or Quote
-------
1
2
o
6
4
5
6
7
8
9
10
11
school-aged children in the general population (Tables 5-38 and 5-39). NCh-associated
increases in respiratory symptoms also were found in infants (Stern etal.. 2013;
Andersen et al.. 2008a; Peel et al.. 2007) (Table 5-39). These results have weaker
implications because symptoms such as wheeze are common in infancy and may not
clearly distinguish children who do and do not develop respiratory conditions like asthma
later in life (Cano Garcinuno et al.. 2013). Further, Peel et al. (2007) examined apnea in
infants on home cardiorespiratory monitors, a group unrepresentative of the general
population. Another uncertainty is whether the temporal variation in ambient NC>2
concentrations from one central site in the area adequately represents variation in ambient
NO2 exposure in infants, particularly those on cardiorespiratory monitors, who may not
spend much time outdoors away from home.
Table 5-38 Mean and upper percentile concentrations of nitrogen dioxide (NO2)
in epidemiologic studies of respiratory symptoms in children in the
general population.
Study3
Stern etal. (2013)
Andersen et al.
(2008a)
Peel etal. (2011)
Rodriquez et al.
(2007)
Ward et al. (2002)
Pateletal. (2010)
Wendt et al.
(2014)
Schwartz et al.
(1994)
Moon et al. (2009)
Location
Bern, Basel,
Switzerland
Copenhagen, Denmark
Atlanta, GA
Perth, Australia
Birmingham, Sandwell,
U.K.
New York City and
nearby suburb, NY
Harris County, TX
(Houston area)
Watertown, MA;
Steubenville, OH;
Topeka, KS; St. Louis,
MO; Portage, Wl;
Kingston-Harriman, TN
Seoul, Incheon, Busan,
Jeju, Korea
Study Period
Apr 1999-
Feb2011
Dec 1998-Dec
2004
Aug 1998-Dec
2002
June 1996-July
1998
Jan-Mar 1997
May-July 1997
2003-2005, mo
NR
2005-2007
Apr-Aug, 1984-
1988
Apr-May 2003
NO2 Metric
Analyzed
24-h avg NO2
24-h avg NO2
1-h max NO2
1-h max NO2
24-h avg NO2
24-h avg NO2
24-h avg NO2
1-h max NO2
24-h avg NO2
24-h avg NO2
Mean/Median
Concentration
ppb
Rural: 8.1b
Urban: 25.6b
11.8
41.7
18
7
18
13.3
NR
39.26
13.3
NR
Upper Percentile
Concentrations
ppb
NR
NR
75th: 14.6
90th: 65.6
Max: 109.2
Max: 48
Max: 24
Max: 35
Max: 29
NR
75th: 48.00
Max: 108
75th: 24.1
Max: 44.2
NR
NO2 = nitrogen dioxide, NR = not reported.
aStudies presented in order of first appearance in the text of this section.
"•Concentrations converted from |jg/m3 to ppb using the conversion factor of 0.532 assuming standard temperature (25°C) and
pressure (1 atm).
January 2015
5-205
DRAFT: Do Not Cite or Quote
-------
1 In school-aged children, not all results were statistically significant, but a pattern of
2 elevated odds ratios indicates consistency in association between short-term NO2
3 exposure and respiratory symptoms (Table 5-39). Evidence is stronger for cough than
4 wheeze, which is identified more with asthma. Children were recruited primarily from
5 schools but also from a birth cohort, suggesting study populations were representative of
6 the general populations. A wide range of participation rates was reported (Table 5-39).
7 but no study reported issues with differential participation by a particular group. The
8 health status of study populations was not always specified, and it is not clear whether the
9 NO2-associated increases in respiratory symptoms reflect associations among all children
10 or those with a respiratory disease. For example, associations were reported in
11 populations with parental history of asthma (Rodriguez et al.. 2007) or with 27% asthma
12 prevalence (Ward et al.. 2002). Findings for symptoms are uncertain in healthy children.
13 NC>2 was not associated with respiratory symptoms in children without asthma (Patel et
14 al.. 2010) but was associated with new diagnosis of asthma in children (Wendt et al..
15 2014) (Table 5-39). Findings that an increase in a 5-day avg of ambient NC>2
16 concentrations may induce respiratory symptoms that precipitate an asthma diagnosis
17 have uncertain implications. Asthma diagnosis was ascertained from Medicaid claims as
18 a record of an outpatient or inpatient visit for asthma or dispensing events of asthma
19 medication during a three-year period (Wendt et al.. 2014). Among children older than
20 age 3 years, what is defined as a diagnosis instead could represent an exacerbation of
21 previously diagnosed asthma. Among children younger than age 3 years, the reliability of
22 an asthma diagnosis is uncertain. The uncertainty of basing a new asthma diagnosis on a
23 three-year review of medical records is underscored by observations that NC>2
24 associations were stronger in children older than age 4 years than in children ages
25 1-4 years (Table 5-39).
26 Despite the associations found with respiratory symptoms in children, there is uncertainty
27 regarding the extent to which the results reflect an independent relationship with NC>2.
28 Ambient NC>2 exposures were assigned from one central site per city or the average
29 across multiple sites per city. In the study of asthma diagnosis, effect estimates were
30 similar for 1-h max NC>2 assigned to subjects as the average NO2 across 17 sites covering
31 the 4,400-km2 area of Harris County, TX and the nearest site within 9.7 km of the zip
32 code centroid (Wendt et al.. 2014). The two NC>2 exposure metrics differed in mean
33 concentration, 39.3 ppb versus 27.7 ppb, but the temporal variability of each metric was
34 not reported to assess whether large differences in temporal variability in NC>2 occurred
35 across the study area.
36 Studies of respiratory symptoms in children also did not clearly identify NC>2 associations
37 that are independent of other traffic-related pollutants. Symptoms also were related to
38 CO, BS, UFP, and PM2 5, which tended to be highly correlated with NO2 (r = 0.61-0.75)
January 2015 5-206 DRAFT: Do Not Cite or Quote
-------
1 (Wendt et al.. 2014; Moon et al.. 2009; Andersen et al.. 2008a; Rodriguez et al.. 2007;
2 Wardetal.. 2002). Analysis of potential confounding was limited, and potentially
3 differential exposure measurement error for central site NO2 and copollutants limits
4 inference from the copollutant model results. In Harris County, TX, 1-h max NO2
5 remained associated with diagnosis of asthma in a copollutant model with 24-h avg PM2s
6 (Table 5-39) (Wendt etal.. 2014). NO2 was weakly correlated with PM2 5 (r = 0.21), and
7 while the variability in ambient PM2 5 concentrations was not reported, NO2
8 concentrations were reported to vary across the county. In infants, ORs for both NO2 and
9 UFP decreased with mutual adjustment (Table 5-39); thus, an independent or
10 confounding effect was not discerned for either pollutant (Andersen et al.. 2008a).
11 Copollutant models were examined in the U.S. Six Cities study for PMio and SO2. ORs
12 for cough decreased with PMio or SO2 adjustment to 1.37 (95% CI: 0.98,2.12) and 1.42
13 (95% CI: 0.90, 2.22 for a 20-ppb increase in NO2, respectively (Schwartz etal.. 1994).
14 The width of 95% CIs is inflated when presented for a 20-ppb increase in NO2, which is
15 double the 10-ppb interquartile range for the study areas. The OR for PMio was robust to
16 NO2 adjustment. Thus, PMio may partly confound NO2 associations. However, the
17 positive ORs for NO2 suggest an independent association for NO2 as well.
January 2015 5-207 DRAFT: Do Not Cite or Quote
-------
Table 5-39 Epidemiologic studies of respiratory symptoms in children in the general population.
Study
Population Examined and Methodological Details
NO2 Metrics Lag Odds Ratio (95% Cl)
Analyzed Day Single-Pollutant Model3
Copollutant Examination
tWendtetal. (2014)
Harris County, TX (Houston area)
n = 18,264 cases in incident asthma, ages 1-17 yr
Case-crossover. Incident asthma cases ascertained for 2004-2007
from Medicaid database. Medicaid enrollment required only for 13 mo.
Date of diagnosis defined as earliest date of asthma diagnosis on
inpatient or outpatient record or earliest of four asthma medication
dispensing events in a year. Uncertainty as to whether outcome
represents incident asthma. Conditional logistic regression adjusted
for temperature, humidity, mold spores, tree pollen, grass pollen, weed
pollen.
NO2-central site
1-h max
Average of
17 sites in
4,400 km2 area
Mean: 39.3 ppb
Site within
9.7 km of zip
code centroid
Mean: 27.6 ppb
0-5 avg Asthma diagnosis
May-Oct
Alleges: 1.22(1.09, 1.36)
1-4 yr: 1.05(1.00, 1.09)
15-17 yr: 1.57(1.06,2.32)
Nov-Apr
Alleges: 1.03(0.93, 1.14)
No quantitative results.
Slightly higher OR than that
for 17 site avg
All ages, NO2 average over
17 county sites, May-Oct.
With24-havgPM25:
1.20(1.06, 1.36)
With 8-h max Os.
1.11 (0.90, 1.37)
Low or moderate
correlations with NO2.
.r=0.21 forPM2.5, 0.49 for
Os. Os means similar for
county average and site
within 9 km.
PM2.sand Os attenuated with
adjustment for NO2.
Schwartz et al. (1994)
Watertown, MA; Kingston-Harriman, TN; St. Louis, MO; Steubenville,
OH; Portage, Wl; Topeka, KS
n = 1,844, grades 2-5
Repeated measures. Daily symptom diaries for 5 mo, collected every
2 weeks. Recruitment from schools. No information on participation
rate. Logistic regression adjusted for lag day 1 temperature, day of
week, city.
NO2-central site Cough:
24-havg 0 1.21(0.92,1.59)
1 site per 0-3 avg 1.61 (1.08, 2.40)
community Lower respiratory symptoms:
1 1.44(0.96,2.16)
For cough:
With PMio: 1.37(0.98,2.12)
WithOs: 1.61 (1.08,2.41)
WithSO2: 1.42(0.90,2.22)
PMio and Os robust to
adjustment for NO2. SO2
reduced.
Moderate correlations with
NO2. R = 0.36 for PMio, 0.35
forPM25, 0.51 forSO2,
—0.28 for O3.
tMoon et al. (2009) NO2-central site
Seoul, Incheon, Busan, Jeju, Korea 24-h avg
n =696, ages NR # sites NR
Repeated measures. Daily symptom diaries for 2 mo. Recruitment
from schools. 69% participation rate. GEE adjusted for temperature,
relative humidity.
Lower respiratory symptoms:
All subjects 1.02(1.00, 1.05)
Seoul 1.08(0.99, 1.18)
Incheon 1.08(0.99, 1.18)
Busan 1.04(0.96,1.12)
Jeju 0.97 (0.89, 1.06)
No copollutant model.
Association also found with
CO. Correlation NR.
January 2015
5-208
DRAFT: Do Not Cite or Quote
-------
Table 5-39 (Continued): Epidemiologic studies of respiratory symptoms in children in the general population.
Study
Population Examined and Methodological Details
NO2 Metrics
Analyzed
Lag
Day
Odds Ratio (95% Cl)
Single-Pollutant Model3
Copollutant Examination
tPateletal. (2010)
New York City and nearby suburb, NY
n = 192 children without asthma, ages 14-20 yr
Repeated measures. Daily symptom diaries for 4-6 weeks, collected
weekly. Recruitment from schools. Self-report of physician-diagnosed
asthma. 75-90% participation across schools. GLMM with random
effect for subject and school and adjusted for weekend, 8-h max Os,
urban location. Adjustment for season, pollen counts did not alter
results.
NO2-central site
24-h avg
1 site 2.2-9.0
km from
schools, 1 site
40 km from
schools
Wheeze:
0.88(0.75, 1.03)
Chest tightness
0.96(0.75, 1.23)
No copollutant model with
BC.
BC also associated with
symptoms. Across locations,
moderately to highly
correlated with NO2.
Spearman r= 0.56-0.90 for
BC.
Ward et al. (2002) NO2-central site
Birmingham, Sandwell, U.K. 24-h avg
n = 162, age 9 yr, 27% with asthma, 31% with atopy Multiple sites
Repeated measures. Daily symptom diaries for two 8-week periods,
collected weekly. Recruitment from schools. 61% participation rate.
Logistic regression adjusted for time trend, temperature, school day.
Cough:
Winter: 0.78(0.57, 1.09)
Summer: 1.14(1.01, 1.27)
No copollutant model.
PM2 s associated with cough.
Rodriguez et al. (2007) NCb-central site
Perth, Australia 24h max
n = 263, ages 0-5 yr, 1 parent with asthma or other atopic disease 10 site average
Repeated measures. Daily symptom diary from birth to age 5 yr.
Recruitment from birth cohort. >80% follow-up participation until yr 4 24-h avg
and 5. GEE adjusted for temperature, humidity.
Wheeze (unit NR):
1.00(0.99, 1.01)
Cough: 1.01 (1.00, 1.02)
No copollutant model.
Associations also found for
PlVh.s, BS at lag 0.
Wheeze: 1.01 (0.98, 1.04)
Cough: 1.03(1.00, 1.06)
tAndersen et al. (2008a)
Copenhagen, Denmark
n = 205, ages 0-3 yr, all with maternal asthma
Repeated measures. Daily symptom diaries from birth to 3 yrs,
collected every 6 mo. Recruitment from birth cohort. 95% follow-up
participation. Mean 805 observations/subject. GEE adjusted for age,
sex, smoking exposure, paternal asthma, temperature, calendar
season.
NO2-central site
24-h avg
1 site within
15 km of homes
N Ox-centra I site
24-h avg
Wheeze:
Age 0-1 yr
3.13(1.27,7.77)
Age 2-3 yr
1.71 (0.94,3.10)
Age 0-1 yr
3.26(1.14, 9.26)
Age 2-3 yr
1.80(0.87, 3.72)
For age 0-1 yr:
With UFP: 1.19(0.14, 75)
with PMio: 2.46 (0.72, 8.4)
UFP & PMio associations
attenuated with adjustment
. for NO2.
UFP & CO highly correlated
with NO2. Spearman
r= 0.67, 0.75. Moderate
correlation for PMio.
r=0.43.
January 2015
5-209
DRAFT: Do Not Cite or Quote
-------
Table 5-39 (Continued): Epidemiologic studies of respiratory symptoms in children in the general population.
Study
Population Examined and Methodological Details
NO2 Metrics
Analyzed
Lag
Day
Odds Ratio (95% Cl)
Single-Pollutant Model3
Copollutant Examination
tStern et al. (2013) NO2-central site
Bern, Base, Switzerland 1-week avg
n = 366, ages 0-1 yr 2 site, urban
Repeated measures. Symptoms reported weekly by telephone for 1 yr. and rural
Recruitment from birth cohort. 95% follow-up participation. GAM
adjusted for study week, sex, siblings, nursery care, maternal atopy,
birth weight, prenatal & post-natal maternal smoking, parental
education.
Daytime respiratory symptom
composite:
1.20(1.04, 1.39)
No copollutant model.
PM-io lag 7 associated with
respiratory symptoms.
Correlation NR.
tPeeletal. (2011)
Atlanta, GA area
n = 4,277, mean age 46 days, 84% premature births
Repeated measures. Followed for mean of 42 days. 111,000 person-
days. Recruitment from referral center for home cardiorespiratory
monitoring of infants. Limited generalizability. Apnea events collected
electronically. No information on participation rate. GEE adjusted for
long-term trends, age.
NO2-central site
1-h max
1 site
0-1 avg Apnea:
1.02(0.96, 1.08)
WithOs: 1.00(0.96, 1.05)
03 association robust to NO2
adjustment. Moderate
correlation with NO2.
Spearman r= 0.45. No
association with PM-io,
coarse PM.
Note: Studies are organized by population examined, and more informative studies in terms of the exposure assessment method and potential confounding considered are presented
first.
GEE = generalized estimating equations, GLMM = Generalized linear mixed model, GLM = generalized linear model, NR = not reported, GAM = generalized additive model,
Cl = confidence interval, CO = carbon monoxide, NO2 = nitrogen dioxide, NOX = sum of NO and NO2,03 = ozone, OR = odds ratio, PM = particulate matter, SO2 = sulfur dioxide,
UFP = ultrafine particles.
aEffect estimates are standardized to a 20 ppb for 24-h avg NO2, 25 ppb for 8-h max, a 30-ppb increase for 1-h max NO2, and 40-ppb increase in 24-h avg NOX.
fStudies published since the 2008 ISA for Oxides of Nitrogen.
January 2015
5-210
DRAFT: Do Not Cite or Quote
-------
Controlled Human Exposure Studies
1 Controlled human exposure studies do not provide strong evidence for NO2-induced
2 increases in respiratory symptoms in healthy adults. Most of these studies were reviewed
3 in the 2008 ISA for Oxides of Nitrogen (U.S. EPA. 2008a). Study details are presented in
4 Table 5-37; overall, studies involved NO2 exposures of 200-2,300 ppb for 2-5 hours and
5 assessment of symptoms 24 hours later. The majority of studies in healthy subjects
6 reported no change in symptoms, as measured by symptom score (Gong et al.. 2005;
7 Witten et al.. 2005; Frampton etal.. 2002; Torres etal.. 1995; Morrow et al.. 1992;
8 Rasmussen et al.. 1992; Linn etal.. 1985b). although a few studies reported statistically
9 nonsignificant increases in symptom score following NC>2 exposure (Frampton et al..
10 2002: Hackney et al.. 1978). NO2 exposure (600 ppb for 1-2 hour) also did not affect
11 respiratory symptoms with simultaneous or sequential Os (200-300 ppb for 1-2 hours)
12 co-exposures (Hazuchaetal.. 1994; Adams etal.. 1987).
5.2.7A Subclinical Respiratory Effects in Healthy Individuals:
Pulmonary Inflammation, Injury, and Oxidative Stress
13 Pulmonary inflammation, injury, and oxidative stress are mediators of respiratory
14 symptoms and decreases in lung function (Section 4.3.5). Consistent with the evidence
15 described in the preceding sections, epidemiologic studies show ambient NCh-related
16 increases in pulmonary inflammation and oxidative stress in children and adults in the
17 general population. A few analyses of copollutant models indicate associations for NO2
18 persist with adjustment for another traffic-related pollutant. Experimental studies report
19 evidence for increased pulmonary inflammation as PMN increases. Effects on other
20 indicators depended upon exposure concentration, duration, and frequency and were
21 observed more consistently at higher than ambient-relevant concentrations. Limited
22 evidence from experimental studies indicates development of an allergic phenotype with
23 repeated NO2 exposures.
Epidemiologic Studies
24 Together, the few epidemiologic studies from the 2008 ISA for Oxides of Nitrogen
25 (U.S. EPA. 2008a) and most recent studies found associations between increases in
26 ambient oxides of nitrogen and increases in pulmonary inflammation or oxidative stress
27 among children and adults in the general population and healthy populations. Locations,
28 time periods, and ambient concentrations of oxides of nitrogen for these studies are
January 2015 5-211 DRAFT: Do Not Cite or Quote
-------
presented in Table 5-40. In this group of studies are several with exposure assessment
methods that aim to account for the high variability in ambient oxides of nitrogen.
Table 5-40
Study3
Flamant-Hulin
etal. (2010)
Lin et al.
(2011)
Liu et al.
(2014a)
Berhane et al.
(2011)
Patel et al.
(2013)
Altuq et al.
(2014)
Holquin et al.
(2007)
Steerenberq et
al. (2001)
Chen etal.
(2012a)
Salam et al.
(2012)
Steerenberq et
al. (2003)
Steenhof et al.
(2013)
Strak et al.
(2012)
Mean and upper percentile concentrations of oxides of nitrogen in
epidemiologic studies of pulmonary inflammation and oxidative
stress in the general population.
Location
Clermont-
Ferrand, France
Beijing, China
Munich and
Wesel,
Germany
13 southern
California
communities
New York City,
NY
Eskisehir,
Turkey
Ciudad Juarez,
Mexico
Utrecht,
Bilthoven, the
Netherlands
New Taipei City,
Taiwan
13 southern
California
communities
the Netherlands,
city NR
the Netherlands,
city NR
Study Period
NR
Jun2007
Sept 2007
Dec 2007
Jun 2008
Sept 2008
NR
Sept-June
2004-2005
May-June 2005
Feb-Mar2007
2001-2002
Feb-Mar1998
Oct-June 2007;
June-Nov2009
2004-2007,
school year
May-June; year
not reported
Mar-Oct 2009
Exposure
Metric
Analyzed
5-day avg NO2
24-h avg NO2
24-h avg NO2
24-h avg NO2
24-h avg NO2
24-h avg NO2
1-week avg NO2
24-h avg NO2
24-h avg NO
24-h avg NO2
24-h avg NO
24-h avg NO2
24-h avg NO2
24-h avg NO2
24-h avg NO
5-h avg NO2
5-h avg NOx
Mean Concentration
ppb
Schools <14.0: 10.1
Schools >14.0: 17.4
24.3
30.4
45.3
26.6
25.9
15.9C
NR
Median: 23.3
Suburban: 9.4b
Urban: 13.0b
Urban-traffic: 21. 2b
18.2
28.2b
30.2b
25.5b
7.4b
21.7
19.0
17.3b
6.3b
36
20
Upper Percentile
Concentrations (ppb)
Across schools:
75th: 14.0b
Max: 19.7b
NR
NR
NR
NR
NR
95th: 29.7b
NR
NR
Max: 13.1b
Max: 17.7b
Max: 28.2b
NR
Max: 44.7b
Max: 168b
Max: 49. 5b
Max: 85.6b
NR
Max: 39.4
Max: 28.3b
Max: 34.5b
Max: 96
Max: 34
January 2015
5-212
DRAFT: Do Not Cite or Quote
-------
Table 5-40 (Continued): Mean and upper percentile concentrations of oxides of
nitrogen in epidemiologic studies of pulmonary
inflammation and oxidative stress in the general
population.
Study3
Adamkiewicz
et al. (2004)
Weichenthal et
al. (2011)
Chimenti et al.
(2009)
Madsen et al.
(2008)
Location
Steubenville,
OH
Ottawa, ON,
Canada
Palermo, Sicily,
Italy
Oslo, Norway
Study Period
Sept-Dec 2000
NR
Nov
Feb
July; year NR
Jan-June 2000
Exposure
Metric
Analyzed
1-h avg NO2
24-h avg NO2
1-h avg NO
24-h avg NO
1-h avg NO2
7-day avg NO2
24-h avg NO2
7-day avg NO2
Mean Concentration Upper Percentile
ppb Concentrations (ppb)
9.2
10.9
15
11.2
High traffic: 4.8
Low traffic: 4.6
31. 7b
27. 1b
33.9b
NR
NR
75th: 12.8, Max:
75th: 14.6, Max:
75th: 16.1, Max:
75th: 14.2, Max:
Max: 1 1
Max: 10
NR
NR
NR
NR
NR
32.9
23.8
215
70.7
NO = nitric oxide. NO2 = nitrogen dioxide, NOX = sum of NO and NO2, NR = not reported.
aStudies presented in order of first appearance in the text of this section.
""Concentrations converted from |jg/m3 to ppb using the conversion factor of 0.532 for NO2 and 0.815 for NO assuming standard
temperature (25°C) and pressure (1 atm).
1
2
3
4
5
6
7
10
11
12
13
14
15
16
17
18
19
Children in the General Population
Ambient NC>2 was associated with pulmonary inflammation and oxidative stress in
populations of children, in which the prevalence of asthma ranged from 7.5 to 59% and
prevalence of allergy ranged from 20 to 56% (Patel et al.. 2013; Berhane et al.. 2011; Lin
etal.. 2011; Steerenberg et al.. 2001). Except for Altugetal. (2014). studies
demonstrated associations in groups without asthma or allergy, with no consistent
difference in magnitude of association between children with and without respiratory
disease [(Liu etal.. 2014a; Berhane et al.. 2011; Lin etal.. 2011); Figure 5-14 and
Table 5-41]. These findings suggest associations between NO2 exposure and pulmonary
inflammation in healthy children.
Among children, associations for NC>2 varied among the various indicators of oxidative
stress and inflammation. As examined in one study each, associations were not observed
with PMNs, eosinophils, exhaled breath condensate pH, or methylation of inducible nitric
oxide synthase (iNOS) (Patel etal.. 2013: Chenet al.. 2012a: Salam etal.. 2012:
Steerenberg et al.. 2001). But several study results pointed to associations with eNO
(Berhane etal.. 2011: Linet al.. 2011: Steerenberg etal.. 2001) (Figure 5-14 and
Table 5-41). Most of these studies assigned exposure from one central site per
community located between 1 and 14 km of subjects' homes. Further, associations were
found with CO, BC, BS, and PM2 5, and copollutant models were not analyzed. Moderate
to strong correlations were reported for NC>2 with PM2 5 and BC (r = 0.47-0.80) (Patel et
January 2015
5-213
DRAFT: Do Not Cite or Quote
-------
1 al., 2013; Berhane et al., 2011). Thus, the extent to which the results for central site NO2
2 reflect an independent association with NO2 is uncertain.
3 Other studies examined copollutant confounding and/or ambient NO2 measurements
4 spatially aligned with a location of subjects, which may better represent ambient
5 microenvironmental exposure. Outdoor school NO2, averaged over 5 or 7 days, was not
6 associated with pulmonary inflammation in children without respiratory disease (Altug et
7 al.. 2014; Flamant-Hulin et al.. 2010; Holguin et al.. 2007). However. Holguin et al.
8 (2007) did not report quantitative results to assess whether there was suggestion of
9 association. And, the cross-sectional comparison of high versus low NO2 in Flamant-
10 Hulin et al. (2010) lacks the sensitivity to discern incremental changes in eNO that may
11 occur with incremental changes in NO2 exposure. Also, for some subjects, eNO was
12 measured days before NO2 was measured.
13 The study with the strongest inference about NO2-related increases in pulmonary
14 inflammation in healthy children was conducted in Beijing, China before and after the
15 2008 Olympics (Lin et al.. 2011). Although results were based on 28 children without
16 asthma, a large number of measurements was collected per child. NO2 and copollutants
17 were measured at a site 0.65 km from schools, improving the spatial alignment of
18 pollutants with subjects over the aforementioned central site studies. A 20-ppb increase in
19 lag 0 day of 24-h avg NO2 was associated with a 22% (95% CI: 18, 26) increase in eNO.
20 This effect estimate was attenuated two to fourfold with adjustment for BC or PM2 5 but
21 remained positive (e.g., 5.6% [95% CI: 0.38, 11] with adjustment for BC). Adjustment
22 for NO2 attenuated the association of eNO with PM2 5 but not BC. These results indicated
23 that the NO2 association was partly confounded, by BC in particular. But, they also
24 provide evidence for an independent association for NO2 in this population of children
25 without asthma.
26 Although (Lin et al.. 2011) found that eNO increased in relation to ambient NO2
27 measured at subjects' school and independently of the traffic-related pollutant BC or
28 PM2 5, other studies had weaker inference. Outdoor school NO2 was not associated with
29 eNO in children without respiratory disease, but these studies either did not report
30 quantitative results or had other methodological limitations. eNO was consistently
31 associated with NO2 measured at central sites and also with traffic-related copollutants
32 such as CO, BC, BS, and PM2 5. Thus, overall, there is uncertainty in the epidemiologic
33 evidence regarding an independent association of NO2 exposure with pulmonary
34 inflammation and/or oxidative stress in healthy children.
January 2015 5-214 DRAFT: Do Not Cite or Quote
-------
Study
Children
Linetal. (2011)
NO2 Metrics
Analyzed
24-h avg
lag 0 day
Exposure Assessment Subgroup
Central site
0.65 km from school
Berhane et al. (2011) 24-h avg Central sites
lag 1-6 day avg
Steerenberg et al.
(2001)
Adults
Straketal. (2012)
Weichenthal et al.
(2011)
Adamkiewicz et al.
(2004)
24-h avg Central site
lag 0-6 day avg 1.9 km from schools
All subjects
No asthma
No asthma
No respiratory allergy
Urban
Suburban
5-h avg
lag Oh
1 -h avg
lag 1 h
24-h avg
lag 0 day
On location of outdoor
exposure
Central site
Central site
-10 -50 5 10 15 20 25 30 35 40 45
Percent change in eNO per increase in NO2 (95% Cl)a
Note: Results are presented first for children then adults. Within each of these groups, results from more informative studies in terms
of the exposure assessment method and potential confounding considered are presented first. Red = recent studies,
Black = previous studies. Study details and quantitative results reported in Table 5-41.
aEffect estimates are standardized to a 20-ppb increase for 24-h avg NO2. Effect estimates for 5-h or 1-h avg NO2 are not
standardized but are presented as reported in their respective studies (Section 5.1.2.3).
Figure 5-14 Associations between ambient nitrogen dioxide (NO2)
concentrations and exhaled nitric oxide (eNO) among children
and adults in the general population.
January 2015
5-215
DRAFT: Do Not Cite or Quote
-------
Table 5-41 Epidemiologic studies of pulmonary inflammation, injury, and oxidative stress in children and adults
in the general population.
Study
Population Examined and Methodological Details
NO2 Metrics Effect Estimate (95% Cl)
Analyzed Lag Day Single-Pollutant Model3
Copollutant Examination
Children in the general population: studies with small spatial scale exposure assessment and/or examination of copollutant confounding
tZhu(2013): Lin etal. (2011) NO2-central site
Beijing, China 24-h avg
n = 36, ages 9-12 yr, 8 with asthma, 28 without asthma. Site g50 meters
Repeated measures before and after Olympics. Examined daily for from schools
five 2-week periods. 1,581 observations. Recruitment from school.
60% responded to initial survey, 95% follow-up participation. GEE
adjusted for temperature, relative humidity, body mass index.
eNO:
0 All subjects: 22% (18, 26)
No asthma: 22% (18, 26)
1 No asthma: 9.5% (5.8, 13)
w/BC: 5.6% (0.38, 11)
w/PIVh.s: 14% (9.5, 19)
BC robust to adjustment for
NO2, PM2.5 reduced but
positive.
Moderate correlations with
NO2. Spearman r= 0.30 for
PM2.5, 0.68 for BC.
tFlamant-Hulin et al. (2010)
Clermont-Ferrand, France
n = 70 without asthma, mean age: 10.7 (SD: 0.7) yr, 75% no atopy
Cross-sectional. Recruitment from schools. 69% participation. Self
or parental report of no asthma. For some subjects, eNO measured
up to 1 week before pollutants. GEE adjusted for atopy, mother's
birth region, parental education, family history of allergy, smoking
exposure. Did not consider confounding by meteorology.
NO2-school
outdoor
24-h avg
NO2-school
indoor
24-h avg
0-4 avg log eNO comparing >14.3
vs. <14.3 ppb NO2
-0.09 (-0.22, -0.04)
0-4 avg log eNO comparing >16.3
vs. <16.3 ppb NO2
-0.16 (-0.11, -0.20)
No copollutant model.
Acetylaldehyde and PlVhs
associated with eNO.
• Correlations with NO2 not
reported.
Children in the general population: studies with central site exposure assessment and no examination of copollutant confounding
tChen etal. (2012a)
New Taipei City, Taiwan
n = 100, mean age 10.6 (SD: 2.5) yr, 33% asthma, 33% atopy
Repeated measures. Examined 3-4 times/mo for 10 mo.
824 observations. Recruited from schools. A priori recruitment of
children with and without asthma or atopy. Participants similar to
. . I I r j. I r I I
nonparticipants. Mixed effects model adjusted for school, age, sex,
body mass index, upper respiratory infection, asthma/allergic rhinitis
attack, asthma medication use, temperature, humidity, day of week,
sampling time, parental education, smoking exposure at home.
NO2-central site
24-h avg
1 sjte 2 5 km
from scnc.ols
most homes '
-i k
' "^'
No quantitative data. NO2
reported not to affect
eosinophils, PMNs,
monocytes, IL-8.
No copollutant model.
Associations found for PM2.5,
Os but not CO.
Moderate to no correlation with
NO2. Pearson r= 0.61 for
PM2.5, -0.01 forOs.
January 2015
5-216
DRAFT: Do Not Cite or Quote
-------
Table 5-41 (Continued): Epidemiologic studies of pulmonary inflammation, injury, and oxidative stress in children
and adults in the general population.
Study
Population Examined and Methodological Details
NO2 Metrics Effect Estimate (95% Cl)
Analyzed Lag Day Single-Pollutant Model3
Copollutant Examination
Steerenberq et al. (2001)
Utrecht (Urban, near busy roadway) and Bilthoven (Suburban), the
Netherlands
n = 126, ages 8-13 yr, 28% respiratory disease, 20% allergy
Repeated measures. Examined 1/weekfor 7-8 weeks. Recruitment
from urban and suburban schools. 65% participation. Non-
standardized eNO collection. Mixed effects model adjusted for sex,
age, # cigarettes smoked in home, presence of a cold, history of
respiratory symptoms, allergy. No consideration for potential
confounding by meteorological factors.
NO2-central site 0-6 avg
eNO:
Urban: 35% (0, 70)b
Suburban: 3.0%, p > 0.05
IL-8 (units NR)
Urban OR: 1.08, p> 0.05
Suburban OR: 1.03, p>0.05
No copollutant model.
PM-io and BS also associated
with eNO, IL-8, uric acid, urea.
NO-central site 0-6 avg
All 24-h avg
Site within 1.9 km
of schools
eNO:
Urban: 6.6% (0, 13)b
Suburban: 7.3% (0, 15)b
IL-8 (units NR)
Urban OR: 1.05, p> 0.05
Suburban OR: 0.95, p>0.05
tPateletal. (2013)
New York City, NY
n = 36, ages 14-19 yr, 94% nonwhite, 50% with asthma
Repeated measures. EEC collected 2/weekfor4 weeks.
217 observations. Recruitment from schools. 89-90% participation
rate. A priori recruitment of children with and without asthma or
atopy. Self-report of physician-diagnosed asthma and symptoms in
previous 12 mo. Mixed effects model with random effects for subject
and adjusted for school, daily average temperature, 8-h max Os.
Adjustment for day of week and humidity did not alter results.
tBerhaneetal. (2011)
Anaheim, Glendora, Long Beach, Mira Loma, Riverside, San Dimas,
Santa Barbara, Upland, CA, Children's Health Study
n = 169, ages 6-9 yr
Cross-sectional. Recruitment from schools. Parental report of
physician-diagnosed asthma and history of respiratory allergy. Two
different methods used for eNO measurement. No information on
participation rate. Linear regression adjusted for community, age,
sex, race/ethnicity, asthma, asthma medication use, history of
respiratory allergy, eNO collection time, body mass index percentile,
smoking exposure, parental education, questionnaire language,
season, multiple temperature metrics, eNO collected outdoors.
NO2-central site
24-h avg
Site 14 km from
schools
NO2-central site
24-h avg
Sites in each
community. #
sites in each
community NR
EEC 8-isoprostane:
0 1.7(0.63, 2. 7) log units
0-3 avg 3.1 (1.3, 4.9) log units
EEC pH:
0 -0.05 (-0.79, 0.68)
0-3 avg -0.11 (-1.2, 1.0)
1-6 avg eNO:
No asthma
11% (-3.2, 28)
No respiratory allergy:
9.1% (-3.6, 23)
No copollutant model.
BC also associated with EEC
pH and 8-isoprostane.
School BC moderately to
highly correlated with NO2.
Pearson r= 0.62, 0.80.
No copollutant model.
PM2.5, PM-io, Os associated
with eNO.
Moderate or weak correlations
with NO2.
Pearson r for warm and cold
season = 0.47, 0.65 for PM25,
0.49, 0.55 for PMio, 0.15, -0.4
for Os.
January 2015
5-217
DRAFT: Do Not Cite or Quote
-------
Table 5-41 (Continued): Epidemiologic studies of pulmonary inflammation, injury, and oxidative stress in children
and adults in the general population.
Study
Population Examined and Methodological Details
NO2 Metrics Effect Estimate (95% Cl)
Analyzed Lag Day Single-Pollutant Model3
Copollutant Examination
tSalametal. (2012)
Same cohort as above
n = 940, ages 6-11 yr, 14% asthma, 56% respiratory allergy
Cross-sectional. Recruitment from schools. Subjects representative
of full cohort. Linear regression model adjusted for age, sex,
ethnicity, asthma, respiratory allergy, parental education, smoking
exposure, community, month of eNO collection. No consideration for
confounding by meteorology.
NO2-central site 1-7avg iNOS promoter methylation: No copollutant model.
24-h avg
Sites in each
community. #
sites in each
community NR
0.40% (-1.0, 1.8)
iNOS methylation not
strong predictor of eNO.
PM2.5 associated with higher
iNOS promoter methylation.
Moderate correlation with NO2.
Spearman r= 0.36.
Adults in the general population: studies with small spatial scale exposure assessment and/or examination of copollutant confounding
tStrak (2013): Strak et al. (2012)
tSteenhofetal. (2013)
Utrecht area, the Netherlands
n = 31, adults ages 19-26 yr, all healthy, nonsmoking
Repeated measures. Examined 3-7 times. 107 observations.
Recruitment from university. Well-defined outdoor exposures at
various sites: underground train station, two traffic sites, farm, and
urban background site. Heart rate maintained during intermittent
exercise. Multiple comparisons could results in higher probability of
associations found by chance alone. No information on participation
rate. Mixed effects model adjusted for temperature, relative
humidity, season, high/low pollen, respiratory infection.
NO2 and NOx-on
site of outdoor
activity
5-h avg
0-h eNO:
post- NO2:6.9%(-1.9, 16)
exposure NOX: 4.7% (-1.8, 11)
Per 10.54 ppb increase in
NO2and 28.05 ppb
increase in NOx
2-h NO2
IL-6:66%(-10, 142)
NAL protein: 60% (0, 121)b
w/PNC: -7.4% (-19, 3.9) for
NO2
-5.8% (-14, 2.4) for NOx
w/EC:4.1%(-6.0, 14)forNO2
2.0% (-7.3, 11) for NOx
PNC association persists
NO2/NOx adjustment. EC &
' Abs attenuated.
ForlL-6:
w/PNC: 95% (0, 190)
w/OC:67%(-10, 144)
Copollutant results robust.
Moderate to high correlations
with NO2 & NOx. Spearman
r=0.56, 0.75 for PNC, 0.74,
0.87 for Abs, 0.67, 0.87 for EC.
Adults in the general population: studies with central site exposure assessment and no examination of copollutant confounding
tChimenti et al. (2009)
Palermo, Sicily, Italy
n = 9, male adults mean age 40 (SD: 3.8) yr, all healthy,
nonsmoking
Repeated measures. Examined during 3 outdoor races. No
information on participation rate. Statistical analyses limited to
correlation analyses. No consideration for potential confounding
factors or repeated measures.
NO2-central site
Averaging time
NR
10 sites
No correlations with plasma No copollutant model.
PMN or eosinophils.
No results reported for
CC16.
Associations found with Os and
PM2.5.
January 2015
5-218
DRAFT: Do Not Cite or Quote
-------
Table 5-41 (Continued): Epidemiologic studies of pulmonary inflammation, injury, and oxidative stress in children
and adults in the general population.
Study
Population Examined and Methodological Details
NO2 Metrics Effect Estimate (95% Cl)
Analyzed Lag Day Single-Pollutant Model3
Copollutant Examination
tWeichenthal et al. (2011)
Ottawa, Canada
n = 42, adults ages 19-58 yr, from nonsmoking homes, 95% white,
62% with allergies, 33% with asthma
Repeated measures. Most examined 3 times. 118 observations. 1-h
outdoor exposures during cycling in low and high traffic areas.
Recruitment from public advertisements. No information on
participation rate. Mixed effects models with random subject effect
adjusted for temperature during cycling, average heart rate.
Adjustment for relative humidity, day of week did not affect results.
NO2-central site
1-h avg
1 site
1-h
4-h
post-
exposure
eNO:
-0.01% (-0.08, 0.06)
-0.04% (-0.09, 0.01)
Per4-ppb increase in NO2
No copollutant model.
PM2 s associated with eNO.
Moderate correlation with NO2.
Spearman r = 0.31 for low
traffic site, 0.45 for high traffic
site.
Potential differential exposure
error for personal PM species
and VOCs vs. central site NO2.
tMadsen et al. (2008)
Oslo, Norway
n = 1,004, male adults ages 67-77 yr, 10% with respiratory disease
Cross-sectional. Recruitment from a larger cohort to represent a
range of home outdoor NO2. No information on participation rate.
GLM adjusted for age, respiratory disease, alcohol consumption,
smoking status, # cigarettes/day, smoking exposure, education,
hour of exam, body mass index, temperature.
NO2-central site
NO2-dispersion
model
No information
on model
validation.
CC16:
0-7 avg 30% (7.8, 57)
3.8% (-7.3, 16)
No copollutant model.
PM2.5 (central site and home)
associated with CC16.
Moderate correlation with NO2.
Spearman rfor home = 0.59.
Adamkiewicz et al. (2004)
Steubenville, OH
n = 29, adults ages 53-90 yr, nonsmoking, 28% with asthma, 24%
with COPD
Repeated measures. Examined weekly for 12 weeks. 138-244 total
observations. No information on participation rate. GLM with subject-
specific intercept and adjusted for time of day, day of week, study
week, temperature, pressure, relative humidity. Several NO2
measurements missing.
NO2-central site
24-h avg
eNO:
0 3.8% (-7.3, 16%)
NO-central site
24-h avg
1 site
0 30% (7.8, 57%)
No copollutant model for NO2.
NOw/PM2.5: 9.2% (-1.7, 20)
• PM2.5 result robust.
Correlations NR.
Ambient NO robust to
adjustment for indoor NO.
Note: Studies are organized by population examined, and more informative studies in terms of the exposure assessment method and potential confounding considered are presented
first.
GLM = generalized linear mixed effects model, GEE = generalized estimating equation, EEC = exhaled breath condensate, eNO = exhaled nitric oxide, IL = interleukin, NR = not
reported, PMNs = polymorphonuclear leukocytes, iNOS = inducible nitric oxide synthase, NAL = nasal lavage, CC16 = club cell protein, Abs = absorbance coefficient,
Cl = confidence interval, CO = carbon monoxide, COPD = chronic obstructive pulmonary disease, EC = elemental carbon, eNO = exhaled nitric oxide, NO = nitric oxide.
NO2 = nitrogen dioxide, NOX = sum of NO and NO2, O3 = ozone, OC = organic carbon, OR = odds ratio, PM = particulate matter, PNC = particle number concentration, SD = standard
deviation, VOC = volatile organic compound.
aEffect estimates are standardized to a 20 ppb for 24-h avg NO2 or NO and 25 ppb for 8-h max NO2. Effect estimates for other averaging times (1-h avg to 15-h avg) are not
standardized but presented as they are reported in their respective studies (Section 5.1.2.3).
"95% Cl estimated for p = 0.05 based on reported p-value < 0.05.
fRecent studies published since the 2008 ISA for Oxides of Nitrogen.
January 2015
5-219
DRAFT: Do Not Cite or Quote
-------
Adults in the General Population
I Among a few studies reviewed in the 2008 ISA for Oxides of Nitrogen that were
2 conducted in older adults (U.S. EPA. 2008a) and recent studies conducted in older adults
3 and adults performing outdoor exercise, several results point to increases in pulmonary
4 inflammation in association with increases in ambient NC>2 concentrations. Pulmonary
5 inflammation was indicated as increases in eNO, nasal lavage IL-6, and indicators of
6 pulmonary injury and lung permeability such as Club cell protein (CC16) and nasal
7 lavage protein levels (Table 5-41). Copollutant-adjusted associations were found with
8 24-h avg NO (Adamkiewicz et al.. 2004) and with 5-h avg NO2 for some outcomes
9 (Steenhof et al., 2013; Strak et al., 2012). The epidemiologic findings have some support
10 from controlled human exposure and toxicological studies (described in sections that
11 follow), although the evidence for pulmonary injury is inconsistent.
12 In populations of mostly healthy adults performing outdoor exercise for <1 to 5 hours,
13 increases in pulmonary inflammation were associated with NO2 measured at the locations
14 of outdoor locations but not at central sites. Compared with studies that do not account
15 for time-activity patterns, examination of subjects during time spent outdoors may better
16 reflect effects related to ambient exposures, particularly for pollutants measured in
17 subjects' outdoor locations. In these studies, subjects had 3-5 separate outdoor exposure
18 periods. In some studies, exposures occurred in locations that represented a gradient of
19 traffic volume. Among adults running or cycling outdoors for 35-90 minutes, eNO and
20 inflammatory cell counts (as measured by PMNs and eosinophils) were not associated
21 with NO2 measured at central sites (Weichenthal et al.. 2011; Chimenti et al.. 2009)
22 (Figure 5-14 and Table 5-41). However, increases in eNO and nasal lavage IL-6 and
23 protein were found in healthy adults in association with 5-h avg NOx and NO2 measured
24 on the site of outdoor exposures (Steenhof et al.. 2013; Strak etal.. 2012). which account
25 for variability in exposure better than central site measurements. Increases in eNO and
26 nasal lavage IL-6 and protein were found immediately after and 2 hours after exposures
27 ended but not the morning after, indicating a transient increase in pulmonary
28 inflammation. The multiple analyses conducted across pollutants, including several PM2 5
29 components, increases the potential for associations to be found by chance alone (Strak et
30 al.. 2012). but there was good consistency in results.
31 Among healthy adults, eNO also was associated with EC, Absorbance coefficient (Abs),
32 and PNC (Strak etal.. 2012); IL-6 also was associated with PM2 5 and OC (Steenhof et
33 al.. 2013). In copollutant models, associations of eNO with NOx and NO2 were
34 attenuated with adjustment for EC or Abs and became negative with adjustment for PNC
35 (Strak etal.. 2012). The PNC effect estimate was robust to adjustment for NOx or NO2.
36 NOx and NO2 were highly correlated with PNC and EC (e.g., r = 0.75 for NOx and PNC
January 2015 5-220 DRAFT: Do Not Cite or Quote
-------
1 and 0.71 forNCh and EC). However, NC>2 remained associated with nasal lavage IL-6
2 and protein after adjustment for PNC or other copollutants (e.g., 67% [95% CI: -10, 144]
3 increase in IL-6 per 30-ppb increase in 5-h avg NO2 and 95% [95% CI: 0, 190] with
4 adjustment for PNC). Thus, in this study of well-defined outdoor exposures, there was
5 evidence of confounding of NC^-eNO associations by PNC but independent associations
6 of NO2 with IL-6 and nasal lavage protein as well as lung function (Figures 5-16 and
7 5-17).
8 Increases in pulmonary inflammation were associated with 24-h avg NO or NCh
9 measured at central monitoring sites among older adults (ages: 53-90 years) (Madsen et
10 al.. 2008; Adamkiewicz et al., 2004). Multiday averages of NO2 (e.g., lag 0-4 day avg,
11 0-7 day avg) were associated with CC16 (Madsen et al.. 2008). However, there is
12 uncertainty regarding independent associations of NC>2 as Madsen et al. (2008) found an
13 association with central site not home NC>2, and each study found associations with other
14 pollutants. There was evidence of an independent association between lag 0 of 24-h avg
15 NO and eNO among older adults in Steubenville, OH (Adamkiewicz et al.. 2004). In a
16 copollutant model, the NO effect estimate decreased, and the 24-h avg PM2 5 effect
17 estimate increased. However, the NO effect estimate remained positive.
Controlled Human Exposure Studies
18 The 2008 ISA for Oxides of Nitrogen (U.S. EPA. 2008a') cited several studies addressing
19 the effects of NO2 exposure on markers of pulmonary inflammation (i.e., differential cell
20 counts, cytokines, eicosanoids), injury (i.e., LDH and protein concentrations), and
21 oxidative stress (i.e., antioxidant molecules and enzymes). The study protocol typically
22 used in these studies includes a single- or multi-day exposure to NO2 (50-5,000 ppb)
23 followed 1 to 24 hours later by collection of bronchial wash or BAL fluid (Table 5-42).
24 The consistency and biological significance of effects across studies is difficult to
25 evaluate given the variety of exposures and timing of when effects were measured, but
26 there is evidence for pulmonary inflammation that is most consistently demonstrated by
27 increases in PMNs.
January 2015 5-221 DRAFT: Do Not Cite or Quote
-------
Table 5-42 Controlled human exposure studies of pulmonary inflammation,
injury, and oxidative stress in healthy adults.
Disease Status3; n, Sex;
Study Age (Mean ± SD)
Exposure Details
(Concentration; Duration)
Endpoints Examined
Azadnivet al.
(1998)
n = 11 M, 4F;
Early phase: 28.1 ± 3.Syr
Late phase: 27.4 ± 4.2 yr
2,000 ppb for 6 h;
Exercise for approximately 10 of
every 30 min at VE = 40 L/min
BAL fluid analysis 1 h and 18 h
after exposure. Protein
concentration, differential cell
counts.
Blomberq et al. n = 8 M, 4 F; 26 yr
(1999) (range: 21-32 yr)
2,000 ppb, 4 h/day for 4 days;
Exercise 15 min on/15 min off at
workload of 75 watts
Cell counts from bronchial biopsies,
BW, and BAL fluid 1.5-h post-
exposure; protein concentration,
IL-8, MPO, hyaluronic acid,
glutathione, ascorbic acid, and uric
acid in BAL fluid and BW 1.5-h
post-exposure, blood parameters.
Devlin et al.
n = 10 M;
range: 18-35 yr
2,000 ppb for 4 h;
Exercise for 15 min on/15 min off
at VE = 50 L/min
Bronchial and alveolar lavage fluid
contents 16-h post-exposure. LDH
activity, tissue plasminogen factor
activity, IL-6 activity, IL-8 activity,
PGE2 levels, total protein,
ascorbate, urate, and glutathione.
Frampton et al. (1) n = 7 M, 2 F;
(1989) 30 yr (range: 24-37 yr)
(2)n = 11 M, 4F;
25 yr (range: 19-37yr)
(1) 600 ppb for 3 h,
(2) 50 ppb for 3 h + 2,000 ppb
peak for 15 min/h;
(1,2) Exercise 10 min on/20 min
off at VE = ~4 times resting
BAL fluid cell viability and
differential counts 3.5-h
post-exposure, IL-1 activity in BAL
fluid cells.
Frampton et al. (1,2) n = 12 M, 9 F;
(2002) F = 27.1 ± 4.1 yr
M = 26.9 ± 4.5 yr
(1) 600 ppb for 3 h,
(2) 1,500 ppb for 3 h;
(1,2) Exercise 10 min on/20 min
off at VE = 40 L/min
Bronchial and alveolar lavage fluid
cell viability and differential counts
3.5-h post-exposure, peripheral
blood characterization.
Helledavetal.
(1994)
n = 8 nonsmokers;
median: 26 yr
(range: 24-35 yr),
n = 8 smokers,
median: 29 yr
(range: 28-32 yr)
3,500 ppb for 20 min;
Exercise last 15 min at 75 watts
Bronchial wash and BAL fluid
analysis. Protein concentration,
differential cell counts.
Huang et al.
(2012b)
(1)n = 11 M, 3F
(2)n=6M, 7F
(3) n = 7 M, 6 F;
24.6 ± 4.3 yr
(1)500ppbNO2for2h,
(2) 500 ppb NO2
+ 73.4 ± 9.9 ug/m3 CAPs for 2 h,
(3) 89.5 ± 10.7 ug/m3 CAPS for
2h;
(1-3) Exercise 15 min on/15 min
off at VE = 25 L/min
Cell counts and concentrations of
IL-6, IL-8, a1-antitrypsin, and LDH
in BAL fluid 18-h post-exposure.
January 2015
5-222
DRAFT: Do Not Cite or Quote
-------
Table 5-42 (Continued): Controlled human exposure studies of pulmonary
inflammation, injury, and oxidative stress in healthy
adults.
Disease Status3; n, Sex;
Study Age (Mean ± SD)
Exposure Details
(Concentration; Duration)
Endpoints Examined
Jorres et al.
(1995)
Healthy; n = 5 M, 3 F;
27 yr (range: 21-33 yr)
Asthma; n = 8 M, 4 F;
27 ± 5 yr
1,000 ppbforS h;
Exercise 10 min on/10 min off at
individual's maximum workload
BAL fluid analysis 1 h after
exposure (cell counts, histamine,
eicosanoids).
Kelly et al.
(1996)
n = 44;
median: 24 yr
(range: 19-45 yr)
2,000 ppbfor4 h;
Exercise 15 min on/15 min off at
75 watts
Antioxidant concentrations and
malondialdehyde in BAL fluid and
bronchial wash at 1.5, 6, or 24-h
post-exposure.
Mohsenin
(1991)
n = 10 M, 9F;
(range: 21-33 yr)
4,000 ppbforS h;
Prior to exposure, 4 week course
of daily placebo or vitamin C and
vitamin E.
BAL fluid immediately after
exposure (a1-protease inhibitor,
elastase inhibitory capacity,
TEARS, conjugated dienes, and
phospholipid phosphorus in lipid
extraction, albumin).
Pathmanathan
et al. (2003)
n = 8M, 4F
26 yr (range: 21-32 yr)
2,000 ppb for 4 h/day for 4 days;
Exercise 15 min on/15 min off at
75 watts
Biomarkers in bronchial epithelium-
exotoxin, GM-CSF, Gro-a, I-CAM
1, IL-5, IL-6, IL-8, IL-10, IL-13, total
and active NF-K(3, and TNF-a
(fiberoptic bronchoscopy after end
of last exposure).
Solomon et al.
(2000)
n = 11 M, 4F;
29.3 ± 4.8 yr
2,000 ppb for 4 h/day for 3 days;
Exercise 30 min on/30 min off at
VE = 25 L/min
Bronchial wash and BAL fluid
analysis immediately after
exposure. Differential cell counts,
LDH, peripheral blood parameters.
Vaqaqqini et
al. (1996)
Healthy; n = 7 M; 34 ± 5 yr 300 ppb for 1 h;
Asthma; n = 4 M, 4 F; Exercise at VE = 25 L/min
29 ± 14 yr
COPD; n = 7M; 58 ± 12 yr
Cell counts in sputum 2-h
post-exposure.
BAL = bronchoalveolar lavage, BW = bronchial wash, CAPS = concentrated ambient particles, COPD = chronic obstructive
pulmonary disease, F = female, GM-CSF = granulocyte macrophage-colony stimulating factor, IL = interleukin, LDH = lactate
dehydrogenase, M = male, NO = nitric oxide. NO2 = nitrogen dioxide, SD = standard deviation, TEARS - thiobarbituric acid
reactive substances, TNF-a = tumor necrosis factor alpha.
aSubjects were healthy individuals unless described otherwise.
January 2015
5-223
DRAFT: Do Not Cite or Quote
-------
1 Several studies reported increases in PMNs and inflammatory mediators following NO2
2 exposure. In a study by Frampton et al. (2002). adults exposed to 1,500 ppb NO2 had
3 increased PMNs in BAL fluid though PMNs were not statistically significantly increased
4 after exposure to 600 ppb, consistent with results from an earlier study (Frampton et al..
5 1989). No change in BAL fluid protein concentration was reported, but lymphocytes
6 were decreased in peripheral blood and increased in BAL fluid after 600 ppb NO2.
7 Consistent with Frampton et al. (2002). several studies reported an increase in PMNs in
8 BAL fluid or bronchial wash from adults exposed to 2,000 ppb NO2 under varying
9 exposure durations and patterns (Solomon et al.. 2000; Blomberg et al.. 1999; Devlin et
10 al.. 1999; Azadniv et al.. 1998). Other cell populations, LDH, and protein concentration
11 generally were not altered following NO2 exposure in these studies. Albumin levels in the
12 bronchial wash fluid were increased following 4 consecutive days of 4-hour exposure to
13 2,000 ppb NO2 (Blomberg et al.. 1999). In an additional study, Helledav et al. (1994)
14 found that bronchial PMNs were increased in nonsmoking adults while alveolar PMNs
15 were increased in smoking adults 24 hours after a brief exposure to 3,500 ppb NO2. With
16 respect to cytokine profiles, Devlin et al. (1999) reported increased IL-6 and IL-8 in BAL
17 fluid from adults 16 hours following exposure to 2,000 ppb. However, 2,000 ppb NO2 for
18 4 hours repeated over 4 consecutive days did not increase expression of IL-6 and IL-8 in
19 biopsies of the bronchial epithelium obtained from healthy human subjects who were
20 exercising at a light rate but did increase the ICAM-1 in the bronchial epithelium
21 (Pathmanathan et al.. 2003). Additionally, Torres et al. (1995) reported increases in
22 thromboxane B2 after a 3 hour exposure to 1,000 ppb NO2, but there were no changes in
23 other eicosanoids. Based on this group of studies, NO2 exposure can induce pulmonary
24 inflammation in healthy human adults, although evidence does not demonstrate
25 NO2-induced pulmonary injury.
26 Studies have also measured effects of NO2 exposure on antioxidant capacity, but results
27 across studies are mixed. Blomberg et al. (1999) found no changes in glutathione,
28 ascorbic acid, or uric acid levels 1.5 hours after 4 consecutive days of exposure to
29 2,000 ppb NO2 for 4 hours. Kelly et al. (1996) examined the kinetics of antioxidant
30 response in the respiratory tract after a single exposure to 2,000 ppb NO2 and found
31 reduced levels of uric acid and ascorbic acid in bronchial wash and BAL fluid 1.5 hours
32 post-exposure. The levels of these antioxidants were increased or at baseline levels at 6
33 and 24 hours after exposure. Glutathione was increased at 1.5 and 6 hours in the
34 bronchial wash, but no changes in glutathione were found in the BAL fluid or for reduced
35 glutathione and malondialdehyde at any time after exposure. Additionally, Mohsenin
36 (1991) found increased lipid peroxidation in BAL fluid following a 3-hour exposure to
37 4,000 ppb NO2. Supplementation with ascorbate and a-tocopherol decreased the
38 NO2-induced lipid peroxidation (Mohsenin. 1991). as well as NO2-induced increases in
39 airway responsiveness (Mohsenin. 1987b) (Section 5.2.2.1). Results from this study
January 2015 5-224 DRAFT: Do Not Cite or Quote
-------
suggest that NCh exposure may induce oxidative stress, and antioxidant status may
modulate the effects of inhaled NCh (Section 4.3.2.1).
3
4
5
6
7
lexicological Studies
Animal toxicological studies reported limited evidence of pulmonary inflammation,
injury, and oxidative stress with ambient-relevant, short-term exposures to NC>2 but more
consistently indicate effects with long-term exposure (Section 6.2.7.2). Few studies have
been published since the 2008 ISA for Oxides of Nitrogen (U.S. EPA. 2008a). and they
similarly report inconsistent results. Study details are presented in Table 5-43.
Table 5-43 Animal toxicological studies of pulmonary inflammation, injury, and
oxidative stress.
Species (Strain);
Study Age; Sex; n
Exposure Details
(Concentration; Duration)
Endpoints Examined
Barth etal.
(1995)
Rat (Sprague
Dawley);
M, n = 7/group
5,000, 10,000, and 20,000 ppb NO2 for
3 or 25 days
Histological evaluation,
morphometry, parenchymal and
vascular damage, pulmonary
arterial thickness, average medial
thickness.
Barth and
Muller(1999)
Rat (Sprague
Dawley);
M, n = 5/group
5,000, 10,000, and 20,000 ppb NO2 for
3 or 25 days
Club cell morphology, cellular
proliferation, epithelial
proliferation.
de Burbure et
al- <2007)
Rat (Wistar);
8 weeks; M;
n = 8/group
(1)1,OOOppbNO2for6h/day,
5 days/week for 4 weeks;
(2) 10,000 ppb NO2for6 h/day,
5 days/week for 4 weeks;
(3) 5,000 ppb NO2 for 6 h/day for
5 days;
(1-3) Animals had selenium-deficient or
selenium-supplemented diets.
BAL fluid lipid peroxidation,
antioxidant enzyme levels, protein
concentration, cell counts, oxidant
production, selenium levels, and
peripheral blood parameters.
Gregory et al.
(1983)
Rat (Fischer 344);
14-16 weeks;
n = 4-6/group
(1) 1,000 and 5,000 ppb NO2 for 7 h/day
for 5 days/week for up to 15 weeks;
(2) 1,000 ppb NO2 for 0.5 h, 5,000 ppb
NO2for1.5h;
(3) 1,000 ppb NO2 for 3 h, 5,000 ppb
NO2for1.5h;
(4) 1,000 ppb NO2 for 0.5 h for
5 days/week for up to 15 weeks
Histopathological evaluation, BAL
fluid and lung homogenate
biochemical analysis (protein
concentration, LDH, glucose-6-
phosphate dehydrogenase,
alkaline phosphatase, glutathione
reductase, and glutathione
peroxidase).
January 2015
5-225
DRAFT: Do Not Cite or Quote
-------
Table 5-43 (Continued): Animal toxicological studies of pulmonary inflammation,
injury, and oxidative stress.
Species (Strain);
Study Age; Sex; n
Exposure Details
(Concentration; Duration)
Endpoints Examined
Hatch et al.
(1986)
Guinea pig (Hartley);
young adult;
n = >3/group
4,800 ppb NO2 for 3 h in deficient and
normal animals; 4,500 ppb NO2 for 16 h;
Animals had Vitamin C deficient or
normal diets
BAL fluid protein and antioxidant
concentrations, 16 h after the 3 h
exposure and within 2 h after the
16 h exposure.
Ichinose et al. Mice (ICR), hamster 400 ppb NO2, 400 ppb Os, and 400 ppb
(1988) (Golden), rat (Wistar), NO2 + 400 ppb O3 for 24 h/day for
guinea pig (Hartley); 2 weeks
10 weeks; M; n = NR
Lipid peroxidation, antioxidant
protective enzymes, total
proteins, TEA reactants, non-
protein sulfhydryls in lung
homogenates, immediately after
exposure.
Last and
Warren (1987)
Rat (Sprague
Dawley);
M; n = > 4/group
5,000 ppb NO2, 1.0 mg/m3 NaCI or
H2SO4, 5,000 ppb NO2 + 1.0 mg/m3
NaCI, 5,000 ppb NO2 + 1.0 mg/m3
H2SO4for23.5 h/day for 1, 3, or 7 days
Collagen synthesis, BAL fluid
protein content and lavagable
enzyme activities, immediately
after exposure.
Mulleretal.
(1994)
Rat (Sprague
Dawley);
M; n = 4
800, 5,000, and 10,000 ppb NO2 for 1
and 3 days
BAL fluid cell counts and protein
concentration, phospholipid
component, SP-A, morphological
changes.
Mustafa et al.
(1984)
Mice (Swiss
Webster);
8 weeks; M;
n = 6/group
(1) 4,800 ppb NO2;
(2) 4,500 ppb O3;
(2) 4,800 ppb NO2 + 4,500 ppb O3;
(1-3) for 8 h/day for 7 days
Physical and biochemical lung
parameters (lung weight, DMA,
protein contents, oxygen
consumption, sulfhydryl
metabolism, NADPH generating
enzyme activities), immediately
after exposure.
Ohashi et al.
Guinea pig (Hartley);
F; n = 10/group
3,000 and 9,000 ppb NO2 for 6 h/day,
6 times/week for 2 weeks
Pathology of mucosal samples:
accumulation of eosinophils,
epithelial injury, mucociliary
dysfunction (taken 24 h after end
of exposure period).
Pagan! et al.
(1994)
Rat (CD Cobs);
M; n = 5/group
5,000 and 10,000 ppb NO2for24 h and
7 days
Analysis of BAL fluid and
superoxide anion production by
AMs.
Povnter et al.
(2006)
Mice (C57BL/6);
n = 5/group
5,000 and 25,000 ppb NO2 for 6 h/day
for 1, 3, or 5 days
Analysis of BAL fluid and
histopathological evaluation
immediately or 20 days
post-exposure.
Robison and
Forman
(1993)
Rat (Sprague
Dawley);
M; n = 3/group
100, 1,000, 5,000, and 20,000 ppb NO2
for 1, 2, and 4 h ex vivo
Enzymatic production of
arachidonate metabolites in AMs,
cyclooxygenase products.
Robison et al.
(1993)
Rat (Sprague
Dawley);
n > 4/group
500 ppb NO2 for 8 h/day for 0.5, 1, 5, or
10 days
BAL fluid cell counts and
arachidonate metabolite levels,
AM arachidonate metabolism,
respiratory burst activity, and
glutathione content.
January 2015
5-226
DRAFT: Do Not Cite or Quote
-------
Table 5-43 (Continued): Animal toxicological studies of pulmonary inflammation,
injury, and oxidative stress.
Species (Strain);
Study Age; Sex; n
Exposure Details
(Concentration; Duration)
Endpoints Examined
Rose et al. Mice (CD-1);
(1989) 4-6 weeks;
n > 4/group
(1) 1,000, 2,500, and 5,000 ppb NCtefor Infection 5 and 10 days
6 h/day for 2 days; intra-tracheal
inoculation with murine
Cytomegalovirus; 4 additional days
(6 h/day) of exposure
(2) re-inoculation 30 days (air)
post-primary inoculation
post-inoculation, histopathological
evaluation, and analysis of BAL
fluid (LDH, albumin,
macrophages).
Sherwin et al. Guinea pig; M;
(1972) n = 4/group
2,000 ppb NO2 continuously for 7, 14, or Histopathological evaluation,
21 days cellular damage by LDH staining.
Sherwin and
Carlson
Guinea pig; M;
n = 9/group
400 ppb NO2 continuously for 1 week Protein concentration in BAL fluid.
Schlesinqer Rabbit (New Zealand 0.5 mg/m3 hhSCk + 300 ppb NO2, Cell counts in BAL fluid, AM
(1987a) white); M; n = 5/group 0.5 mg/m3 H2SCM + 1,000 ppb NO2 for function 24 h after exposure.
2 h/day for 2, 6, or 13 days
Schlesinqer et Rabbit (New Zealand (1) 1,000, 3,000, or 10,000 ppb NCfe for Eicosanoids in BAL fluid,
al. (1990) white); M; n = 3/group 2 h; immediately and 24 h after
(2) 3,000 ppb NO2 + 300 ppb O3 for 2 h; exposure.
(3) 100, 300, or 1,000 ppbO3for2h
H2SO4 = sulfuric acid, F = female, LDH = lactate dehydrogenase, M = male, NADPH = reduced nicotinamide adenine dinucleotide
phosphate, NO = nitric oxide. NO2 = nitrogen dioxide, NR = not reported, O3 = ozone.
1
2
3
4
5
6
7
10
11
12
13
14
15
Pulmonary Inflammation
Animal studies examined similar endpoints to those in controlled human exposure studies
(discussed above) to assess pulmonary inflammation after NO2 exposure, but effects of
NC>2 are inconsistent between humans and laboratory animals. While studies in humans
demonstrated increases in BAL fluid PMNs after NO2 exposure, several studies found no
statistically significant changes in BAL fluid inflammatory cells and mediators in rodents
exposed to 5,000 ppb NO2 for up to 7 days (Poynter et al.. 2006; Muller et al.. 1994;
Pagani et al., 1994; Mustafa et al.. 1984). Schlesinger (1987a). however, did report an
increase in PMNs in BAL fluid from rabbits exposed to 1,000 ppb NC>2 for 2, 6, and
13 days, though all exposures included H2SO4.
A series of studies also investigated changes in arachidonic acid metabolism in response
to NCh exposure. Robison and Forman (1993) exposed rats or rat AMs ex vivo to NO2 at
concentrations as low as 100 ppb and found that in vivo exposure led to statistically
significant decreases in eicosanoid levels in as little as 4 hours, while ex vivo exposure of
AMs led to statistically significant increases in cyclooxygenase and lipoxygenase activity
and slight, but increases in eicosanoids that were statistically nonsignificant. Schlesinger
January 2015
5-227
DRAFT: Do Not Cite or Quote
-------
1 et al. (1990) studied similar endpoints in rabbits exposed to NO2 for 2 hours and found an
2 increase in thromboxane B2 in BAL fluid at 1,000 ppb, but not at 3,000 ppb. Results from
NO2 and Os co-exposure suggested that eicosanoid response is more sensitive to Os
4 exposure than NC>2.
5 A study in guinea pigs provides evidence for the development of pro-allergic responses
6 since increases in eosinophils were found in the nasal epithelium and submucosa
7 following a two-week exposure to 3,000 ppb NO2 (Ohashi et al.. 1994) (Table 5-43). The
8 observed increase in numbers of airway eosinophils in this study and expression of Th2
9 cytokines in Pathmanathan et al. (2003) suggest that inhaled NCh may promote Th2
10 skewing and allergic sensitization (Section 4.3.2.6.3).
Pulmonary Injury
11 In addition to NC>2-induced changes in inflammatory cells and mediators, toxicological
12 studies have also assessed pulmonary injury at the morphologic and molecular level. For
13 example, Muller et al. (1994) did not find evidence of changes in surfactant or lipid
14 content in BAL fluid at concentrations below 10,000 ppb; however histopathological
15 assessments in lung tissues from this study suggested morphologic changes in the
16 respiratory airways including thickened interstitium and inflammatory cell accumulation.
17 Earth et al. (1995) expanded upon these structural observations and reported pulmonary
18 injury at 10,000 ppb that includes diffuse alveolar damage, epithelial degeneration and
19 necrosis, proteinaceous oedema, inflammatory cell influx, and compensatory
20 proliferation and differentiation. Few morphologic studies have incorporated
21 ambient-relevant NC>2 exposures; however, Earth et al. (1995) reported that slight
22 interstitial edema was present following a 5,000 ppb exposure for 3 days, although this
23 edema was not present after a 25 -day exposure. In another study, Earth and Muller
24 (1999) also found slight modifications to the bronchiolar epithelium after 3 days of
25 exposure, though the bronchi appeared normal. The proliferative index of club cells
26 increased in the bronchioles and bronchi relative to air controls following a 3 -day
27 exposure to 5,000 ppb, but the number of club cells was only increased in the
28 bronchioles; no changes were observed following a 25-day exposure. Additionally, Last
29 and Warren (1987) found increased collagen synthesis, a feature of fibrosis, in lung
30 homogenates obtained from rats exposed to 5,000 ppb NO2, which was enhanced with
3 1 concurrent exposure to H2SO4 or NaCl. Overall, short term exposure to NO2 appears to
32 induce minor morphologic changes in the respiratory tract, although long-term exposure
33 studies (Section 6.2.6) report more profound impacts of exposure.
34 In addition to pulmonary injury observed at the morphologic level, molecular markers of
35 injury have been described in some studies. Continuous exposure to 400 ppb NCh for
36 1 week resulted in increased BAL fluid protein in guinea pigs on a Vitamin C-deficient
January 2015 5-228 DRAFT: Do Not Cite or Quote
-------
1 diet (Sherwin and Carlson. 1973). while a 2,000 ppb exposure for 1-3 weeks increased
2 LDH levels in alveolar lung sections (Sherwin et al.. 1972). Hatch etal. (1986) also
3 reported increased BAL fluid protein levels in NO2 exposed Vitamin C-deficient guinea
4 pigs. Gregory et al. (1983) exposed rats to 1,000 and 5,000 ppb NO2 for up to 15 weeks
5 and found early increases (1.7-2.7 weeks) in LDH in BAL fluid. Rose etal. (1989) did
6 not find any changes in LDH in BAL fluid following a 6-day exposure to 5,000 ppb,
7 though slight increases in albumin were reported, suggesting mild pulmonary injury. In
8 contrast to these studies, a number of studies have shown that NO2 exposure below
9 5,000 ppb does not result in an increase in BAL fluid protein and LDH levels in many
10 animal species (Robison et al.. 1993; Robison and Forman. 1993; Schlesinger et al..
11 1990; Last and Warren. 1987).
Oxidative Stress and Antioxidant Status
12 The 2008 ISA for Oxides of Nitrogen (U.S. EPA. 2008a) did not discuss the toxicological
13 evidence relating to effects of NO2 on antioxidants or oxidative stress, but in the few
14 previous studies, oxidative stress was not consistently induced by ambient-relevant NO2
15 concentrations. For example, Ichinose etal. (1988) exposed rats and guinea pigs to
16 400 ppb NO2 for 2 weeks and found that levels of lipid peroxides and antioxidants
17 (nonprotein sulfhydryls, Vitamin C, and Vitamin E) were not affected in lung
18 homogenates. Furthermore, there was no change in activity levels of in antioxidant
19 enzymes including glucose-6-phosphate dehydrogenase, 6-phosphogluconate
20 dehydrogenase, glutathione S-transferase (GST), glutathione peroxidase (GPx),
21 glutathione reductase, and superoxide dismutase (SOD) after NO2 exposure; however,
22 combined exposure with Os did demonstrate synergistic effects on antioxidant systems.
23 Studies also report variable effects of NO2 on glutathione and oxidized glutathione levels
24 in the BAL fluid and peripheral blood. Pagani etal. (1994) found that rats exposed to
25 5,000 ppb NO2 for 24 hours had increased total and oxidized glutathione in peripheral
26 blood, though the increase in oxidized glutathione alone was not statistically significant.
27 Conversely, statistically significant increases in oxidized glutathione were reported in the
28 BAL fluid, whereas total glutathione was slightly diminished, de Burbure et al. (2007)
29 reported decreased GPx in plasma immediately and 48 hours after exposure to 1,000 ppb
30 NO2 for 28 days, whereas GPx increased in the BAL fluid. GST and SOD also increased
31 in BAL fluid after exposure, although SOD returned to control levels by 48 hours
32 post-exposure. Rats exposed to 5,000 ppb NO2 for 5 days also had reduced levels of GPx
33 in plasma and increased levels of GPx and GST in BAL fluid. SOD also increased in
34 BAL fluid, but only 48 hours post-exposure. Oxidized lipids were transiently increased
35 immediately after exposure in BAL fluid and were not affected in the subacute exposure.
January 2015 5-229 DRAFT: Do Not Cite or Quote
-------
1 Other studies have reported effects of NO2 on antioxidant levels or enzyme activity, but
2 those exposures were above ambient-relevant concentrations of NO2.
3 Other studies have reported that Vitamin C or E deficiency enhances the effects of NO2 in
4 the lung, which is plausible given that both vitamins have antioxidant activity in the
5 airways and neutralize reactive oxygen species. Guinea pigs with a Vitamin C-deficient
6 diet had increased BAL fluid protein and lipids following exposure to 1,000 ppb NO2 for
7 72 hours or 4,800 ppb for 3 hours relative to air controls or guinea pigs with a normal diet
8 (Hatch etal., 1986; Selgrade etal.. 1981). Additionally, exposure to 5,000 ppb for
9 72 hours resulted in 50% mortality in Vitamin C-deficient guinea pigs. Similarly, rats
10 with diets deficient in Vitamin E had increases in lipid peroxidation and protein content
11 in lung homogenates following a 7-day exposure to 3,000 ppb NO2 (Elsayed and
12 Mustafa. 1982; Sevanian et al., 1982). Additional support for an influence of Vitamin E is
13 provided by observations that NO2-induced increases in BAL fluid protein or decreases in
14 glutathione peroxidase activity were attenuated in animals fed Vitamin E-supplemented
15 diets, relative to animals not supplemented with Vitamin E (Guth and Mavis. 1986; Ayaz
16 and Csallany. 1978). These studies demonstrate that antioxidants, particularly Vitamin C
17 and E, can modify the effects of NO2 on pulmonary injury and oxidative stress in
18 animals.
Development of a pro-allergic phenotype
19 A few experimental studies indicate that repeated exposures to NO2 may promote Th2
20 skewing, which may have implications for allergic sensitization and development of
21 Th2-related conditions such as asthma (Section 4.3.2.6). In guinea pigs, two-week
22 exposure to 3,000 ppb NO2 led to an increase in eosinophils in the nasal epithelium and
23 submucosa (Ohashi et al., 1994) (Table 5-43). In healthy adults, 2,000 ppb NO2 exposure
24 for 6 hours on 4 consecutive days increased expression of the Th2 cytokines IL-5, IL-10,
25 and IL-13 in the bronchial epithelium (Pathmanathan et al., 2003). IL-5 promotes
26 eosinophila, and IL-13 promotes airway hyperresponsiveness and mucus production.
5.2.7.5 Summary of Respiratory Effects in Healthy Individuals
27 The 2008 ISA for Oxides of Nitrogen did not make a specific assessment about the
28 respiratory effects of short-term exposure to oxides of nitrogen in healthy populations
29 (U.S. EPA. 2008a). However, previous and recent epidemiologic evidence indicates
30 ambient NO2-associated increases in respiratory symptoms and pulmonary inflammation
31 in children in the general population. Neither epidemiologic nor experimental studies
32 clearly support NO2-related effects on lung function in healthy populations. Experimental
January 2015 5-230 DRAFT: Do Not Cite or Quote
-------
1 studies do not clearly demonstrate effects on respiratory symptoms in healthy adults
2 either. Although experimental findings are variable across the exposure concentrations
3 and specific outcomes examined, they provide some support for the effects of NO2 on
4 events that may mediate the occurrence of respiratory symptoms.
5 Epidemiologic studies show a consistent pattern of elevated odds ratios for associations
6 between short-term increases in ambient NO2 and cough in school-aged children
7 (Table 5-39). Epidemiologic studies did not examine respiratory symptoms in healthy
8 adults. A majority of evidence was for 24-h avg NO2, and no clear difference was found
9 between lag day 0 and multiday (2- to 5-day) average concentrations. Most studies
10 recruited children from schools, increasing the likelihood that study populations were
11 representative of the general population. No study reported issues with differential
12 participation by a particular group. Despite the consistency of findings, there is
13 uncertainty as to whether these results support an independent effect of NO2 on
14 symptoms in healthy children. Associations were reported in study populations with 27%
15 asthma prevalence (Ward et al.. 2002) or parental asthma (Rodriguez et al.. 2007) but not
16 children without asthma (Patel etal.. 2010). Other studies, including a large U.S.
17 multicity study (Schwartz et al.. 1994). did not report the health status of study
18 populations. All studies assigned NO2 exposure from central sites. Symptoms also were
19 associated with CO, BS, and PM2 5, PMio, and SO2, and while there is limited evidence
20 indicating associations for NO2 with adjustment for PMio or SO2 (Schwartz et al.. 1994).
21 confounding by traffic-related copollutants was not examined. An association found
22 between indoor NO2 at ice arenas and respiratory symptoms in hockey players (Salonen
23 et al.. 2008) provides limited coherence for the findings for outdoor NO2.
24 Although evidence for NO2-associated decreases in lung function is inconsistent overall,
25 associations were observed in healthy adults in studies characterized as having strong
26 exposure assessment with NO2 and copollutants measured on site of outdoor exposures
27 near busy roads or a steel plant (Dales etal.. 2013; Straket al.. 2012). These studies
28 found lung function decrements associated with 5-hour or 10-hour outdoor NO2
29 exposures that persisted 0 to 18 hours after exposure. Strak etal. (2012) further indicated
30 NO2 associations in an array of copollutant models each with another traffic-related
31 pollutant: EC, PM25inetal components, PNC, or PM2s. That some of the copollutants
32 effect estimates were attenuated with adjustment for NO2 indicates that NO2 may have
33 confounded copollutant associations. These findings provide evidence for the
34 independent effects of NO2 exposure on lung function in healthy individuals, but such
35 informative epidemiologic studies are few in number.
36 Controlled human exposure studies do not indicate an effect of NO2 exposures of
37 200-4,000 ppb (30 minutes to 6 hours) on respiratory symptoms or lung function in
January 2015 5-231 DRAFT: Do Not Cite or Quote
-------
1 healthy adults (Sections 5.2.7.2 and 5.2.7.3) but provide some support for the
2 epidemiologic findings by demonstrating effects on some key events that can lead to
3 respiratory symptoms (Section 4.3.5). Controlled human exposure studies show increases
4 in airway responsiveness in healthy adults with 3-hour NO2 exposures above 1,000 ppb
5 but no change at lower concentrations (Section 5.2.7.1). Epidemiologic studies indicate
6 increases in pulmonary inflammation in association with NO2 measured near children's
7 school or on location of outdoor exposures with adjustment for PM2 5, BC, OC, or PNC
8 (Steenhof etal.. 2013: Lin etal.. 2011). In Steenhof etal. (2013). NO2 appeared to
9 confound copollutant associations. Experimental studies provide some support for the
10 independent effects of NO2 exposure. Although pulmonary inflammation and oxidative
11 stress were not always affected in controlled human exposure or toxicological studies
12 with a wide range of NO2 exposures and durations (800-5,000 ppb for 6 hours to
13 2 weeks) (Section 5.2.7.4). controlled human exposure studies of healthy adults showed
14 NO2-induced (1,500-3,500 ppb) increases in PMNs.
15 Overall, epidemiologic studies find associations of short-term increases in ambient NO2
16 concentrations with increases in respiratory symptoms and pulmonary inflammation in
17 children, and to a limited extent, in healthy adults. Evidence specifically attributing the
18 associations to healthy children or to NO2 independently of other traffic-related pollutants
19 is limited. Experimental studies do not indicate NO2 effects on respiratory symptoms or
20 lung function in healthy adults. There is experimental evidence for an effect of NO2 on
21 key events underlying respiratory symptoms, including increases in airway
22 responsiveness and pulmonary inflammation, but variable findings among exposure
23 concentrations and/or specific endpoints limit the extent of support to the epidemiologic
24 evidence for respiratory symptoms and respiratory effects overall in healthy populations.
5.2.8 Respiratory Mortality
25 Studies evaluated in the 2008 ISA for Oxides of Nitrogen that examined the association
26 between short-term NO2 exposure and cause-specific mortality consistently found
27 positive associations with respiratory mortality, with some evidence indicating that the
28 magnitude of the effect was larger compared to total and cardiovascular mortality. Recent
29 multicity studies conducted in Asia (Wong et al.. 2008). China (Meng et al.. 2013: Chen
30 etal.. 2012b). and Italy (Faustini etal.. 2013: Chiusolo et al.. 2011). as well as a
31 meta-analysis of studies conducted in Asian cities Atkinson et al. (2012) add to the initial
32 body of evidence indicating larger respiratory mortality effects (Section 5.4.3.
33 Figure 5-23). However, an additional multicity study conducted in Italy (Bellini et al..
34 2007). which is an extension of Biggeri et al. (2005). observed relatively consistent risk
35 estimates across mortality outcomes, which differs from the results of the original
January 2015 5-232 DRAFT: Do Not Cite or Quote
-------
1 analysis and complicates interpretation of whether there is differential risk among
2 mortality outcomes.
3 The initial observation of consistent positive NO2 associations with respiratory mortality
4 was examined in a few studies that conducted copollutant analyses. It is important to note
5 that, similar to interpreting NO2 associations with total mortality (Section 5.4.4). it is
6 difficult to examine whether NO2 is independently associated with respiratory mortality
7 because NO2 is often highly correlated with other traffic-related pollutants. Chen et al.
8 (2012b) in the 17 Chinese cities study [China Air Pollution and Health Effects Study
9 (CAPES)] found that NO2 risk estimates for respiratory mortality were slightly
10 attenuated, but remained positive in copollutant models with PMio and SO2 (9.8% [95%
11 CI: 5.5, 14.2]; with PMio: 6.7% [95% CI: 2.9, 10.7]; with SO2: 7.0% [95% CI: 3.2, 11.0];
12 for a 20-ppb increase in 24-h avg NO2 concentrations at lag 0-1 days). These results are
13 consistent with those of Meng et al. (2013) for COPD mortality in a study of four
14 Chinese cities (i.e., 7.1% [95% CI: 5.4, 8.9]; lag 0-1 for a 20-ppb increase in 24-h avg
15 NO2 concentrations; with PMi0: 6.0% [95% CI: 3.2, 8.8]; and with SO2: 6.9% [95% CI:
16 4.2, 9.5]). Chiusolo etal. (2011) also found evidence that associations between short-term
17 NO2 exposure and respiratory mortality remained robust in copollutant models in a study
18 of 10 Italian cities. In both an all-year analysis of NO2 with PMio (NO2: 13.7% [95% CI:
19 2.9, 25.8]; NO2 with PMio: 13.4% [95% CI: 2.9, 24.9]; for a 20-ppb increase in NO2
20 concentrations at lag 1-5 days), and a warm season (April-September) analysis of NO2
21 with O3 (NO2: 41.3% [95% CI: 16.2, 71.7]; NO2 with O3: 43.4% [95% CI: 14.6, 79.5]; for
22 a 20-ppb increase in NO2 concentrations at lag 1-5 days) NO2 associations with
23 respiratory mortality were relatively unchanged. However, when focusing on a subset of
24 respiratory mortality, specifically those deaths occurring out-of-hospital, in six Italian
25 cities, Faustini et al. (2013) reported evidence of an attenuation of the NO2-repiratory
26 mortality association in copollutant models with PMio (NO2: 24.5% [95% CI: 7.4, 44.2];
27 lag 0-5 for a 20-ppb increase in 24-h avg NO2 concentrations; NO2 with PMio: 11.8%
28 [95% CI: -7.5, 35.0]). Overall, the limited number of studies that have examined the
29 potential confounding effects of copollutants on the NO2-respiratory mortality
30 relationship generally indicate that associations remain relatively unchanged, but it is
31 remains difficult to disentangle the independent effects of NO2.
32 Of the studies evaluated, only the studies conducted in Italy examined potential seasonal
33 differences in the NO2-respiratory mortality relationship (Chiusolo et al.. 2011; Bellini et
34 al.. 2007). In a study of 15 Italian cities, Bellini et al. (2007) found that risk estimates for
35 respiratory mortality were dramatically increased in the summer from 1.4 to 9.1% for a
36 20-ppb increase in 24-h avg NO2 concentrations at lag 0-1 days, respectively, with no
37 evidence of an association in the winter. These results were further confirmed in a study
38 of 10 Italian cities (Chiusolo et al.. 2011). which also observed an increase in risk
January 2015 5-233 DRAFT: Do Not Cite or Quote
-------
1 estimates for respiratory mortality in the warm season (i.e., April-September) compared
2 to all-year analyses. Chiusolo etal. (2011) did not conduct analyses only for the winter
3 season . Although the respiratory mortality results are consistent with those observed in
4 the total mortality analyses conducted by Bellini et al. (2007) and Chiusolo et al. (2011).
5 as discussed in Section 5.4. studies conducted in Asian cities observed much different
6 seasonal patterns, and it remains unclear whether the seasonal patterns observed for total
7 mortality would be similar to those observed for respiratory mortality in these cities.
8 An uncertainty that often arises when examining the relationship between short-term air
9 pollution exposures and cause-specific mortality is whether the lag structure of
10 associations and the C-R relationship provide results that are consistent with what is
11 observed for total mortality. Chiusolo etal. (2011) in a study of 10 Italian cities found the
12 strongest evidence for an effect of NO2 on respiratory mortality at longer lags, with the
13 largest association at lag 2-5 days, which is indicative of a delayed effect (Figure 5-24).
14 These results are supported by the study of (Faustini et al.. 2013) in six Italian cities,
15 which found the strongest evidence of an NCh-association with out-of-hospital
16 respiratory mortality at lags 2-5 and 0-5 days. Evidence of an immediate effect at lag
17 0-1 day avg was also observed, but the magnitude of the association was smaller
18 compared to lags 2-5 and 0-5 days. However, Chen etal. (2012b) in the CAPES study
19 reported the largest effect at single-day lags of 0 and 1 and the average of lag 0-1 days
20 providing support for an immediate effect of NC>2 on respiratory mortality (Figure 5-25).
21 When examining longer lags Chen etal. (2012b) reported that the magnitude of the
22 association was similar, albeit slightly smaller, for a 0-4 day lag, suggesting a potential
23 prolonged effect. Meng etal. (2013) in a study of COPD mortality in four Chinese cities,
24 of which all are found within the CAPES study cities, reported slightly different results
25 than the CAPES study respiratory mortality results. When examining single-day lags
26 from 0 to 7 days, the authors reported the largest association for lag day 0. However,
27 larger associations were observed in multiday lag analyses with a similar magnitude of an
28 association observed for lags 0-1 and 0-7 days, and the largest magnitude of an
29 association overall for lag 0-4 days.
30 To date, analyses detailing the C-R relationship between air pollution and cause-specific
31 mortality have been limited. In the analysis of four Chinese cities, Meng etal. (2013) also
32 examined the air pollution and COPD mortality C-R relationship in each individual city.
33 To examine the assumption of linearity the authors fit both a linear and spline model to
34 the city-specific NO2-COPD mortality relationship. Meng etal. (2013) then computed the
35 deviance between the two models to determine if there was any evidence of non-linearity.
36 An examination of the deviance did not indicate that the spline model improved the
37 overall fit of the NO2-COPD mortality relationship across the cities examined
38 (Figure 5-15).
January 2015 5-234 DRAFT: Do Not Cite or Quote
-------
g
ro
o
E
Q
CL
O
o
(0
JO
0) CD
OJ
O
Beijing
Shanghai
Guangzhou
Hongkong
50 100 150
NO2 concentration at lag 01 day
Source: Reprinted with permission of Elsevier Meng et al. (2013).
Figure 5-15 City-specific concentration-response curves of nitrogen dioxide
and daily chronic obstructive pulmonary disease (COPD)
mortality in four Chinese cities.
5.2.9 Summary and Causal Determination
1 Evidence indicates that there is a causal relationship between short-term NO2 exposure
2 and respiratory effects based on the coherence among multiple lines of evidence and
3 biological plausibility for effects on asthma exacerbation. There is some support for
4 NO2-related exacerbation of respiratory allergy and COPD, respiratory infection,
5 respiratory mortality, and respiratory effects in healthy populations. However, because of
6 inconsistency among lines of evidence and consequent uncertainty about the effects of
7 NC>2 exposure, evidence for these other non-asthma respiratory effects does not strongly
8 inform the determination of a causal relationship.
January 2015
5-235
DRAFT: Do Not Cite or Quote
-------
1 The determination of a causal relationship represents a change from the "sufficient to
2 infer a likely causal relationship" concluded in the 2008 ISA for Oxides of Nitrogen
3 (U.S. EPA. 2008a). Consistent with previous findings, recent epidemiologic results
4 indicate associations between ambient NO2 concentrations and asthma-related respiratory
5 effects. Biological plausibility continues to be provided by the NCh-induced increases in
6 airway responsiveness and allergic inflammation demonstrated in experimental studies.
7 The 2008 ISA cited uncertainty as to whether NCh has an effect independent from other
8 traffic-related pollutants, and additional copollutant model results show ambient
9 NO2-associated increases in asthma-related effects with adjustment for PlVfc 5, BC/EC,
10 UFP, OC, metals, VOCs, or CO. Thus, much of the evidence for NO2-related respiratory
11 effects was available in the 2008 ISA. However, the 2008 ISA emphasized epidemiologic
12 findings and did not assess the coherence and biological plausibility for various
13 respiratory conditions separately, which is important given that the weight of evidence
14 varies among respiratory conditions. More than new findings, the evidence integrated
15 across outcomes related to asthma exacerbation, with due consideration of experimental
16 evidence, is sufficient to rule out chance, confounding, and other biases with reasonable
17 confidence and support a change in conclusion from likely to be causal to causal
18 relationship. The evidence for a causal relationship is detailed below using the framework
19 described in the Preamble (Table ID. The key evidence as it relates to the causal
20 framework is presented in Table 5-45.
5.2.9.1 Evidence on Asthma Exacerbation
21 A causal relationship between short-term NO2 exposure and respiratory effects is strongly
22 supported by evidence for effects across clinical asthma events and pulmonary responses
23 that mediate asthma exacerbation. Epidemiologic studies (Iskandar et al.. 2012;
24 Strickland etal.. 2010; Jalaludin et al.. 2008; Villeneuve et al.. 2007) demonstrate
25 associations between increases in ambient NO2 concentration and increases in asthma
26 hospital admissions and ED visits among subjects of all ages and children
27 (Section 5.2.2.4). Risk estimates ranged from a 4.5 to 34% increase per 20-ppb increase
28 in 24-h avg NO2 or 30-ppb increase in 1-h max NO2. These observations are coherent
29 with evidence in children with asthma for increases in respiratory symptoms (Zora et al..
30 2013: Gent et al.. 2009: Delfino et al.. 2003: Delfino etal.. 2002) (Section 5.2.2.3). the
31 major reason for seeking medical treatment. The recruitment of children from schools
32 supports the likelihood that study populations were representative of the general
33 population of children with asthma. Issues with selective participation by certain groups
34 was not reported. Individual epidemiologic studies examined multiple outcomes and lags
35 of exposure, and not all studies had statistically significant results. However, the pattern
January 2015 5-236 DRAFT: Do Not Cite or Quote
-------
1 of association observed for NCh supports the consistency of evidence and does not
2 indicate a high probability of associations found by chance alone. Consistency also is
3 demonstrated as NCh-related asthma exacerbation found across diverse locations in the
4 U.S., Canada, and Asia, including several recent multicity studies. Asthma hospital
5 admissions and ED visits were associated with 24-h avg and 1-h max NC>2, whereas
6 respiratory symptoms were associated primarily with 24-h avg NC>2. Most evidence was
7 for multiday lags of NC>2 exposure of 2 to 5 days, but associations also were found with
8 lags of 0 or 1 day. A larger magnitude of association is not clearly indicated for a
9 particular lag of NC>2 exposure. The concentration-response relationship was analyzed for
10 pediatric asthma ED visits in Atlanta, GA and Detroit, MI, and neither a threshold nor
11 deviation from linearity was found in the range of 24-h avg or 1-h max ambient NC>2
12 concentrations examined (Li et al.. 20lib: Strickland et al.. 2010).
13 Epidemiologic evidence for NCh-related decreases in lung function in populations with
14 asthma is inconsistent as a whole, but associations were found with lung function
15 measured under supervised conditions (Greenwald et al.. 2013; Martins etal.. 2012;
16 Delfino et al.. 2008a: Holguin et al.. 2007; McCreanoret al.. 2007; Delfino et al.. 2003)
17 (Section 5.2.2.2). Many populations had high prevalence of atopy (e.g., 53-84%), which
18 supports the evidence for asthma-related respiratory symptoms, ED visits, and hospital
19 admissions because airway obstruction in response to allergens can lead to lung function
20 decrements and respiratory symptoms. These epidemiologic findings are not clearly
21 supported by findings for respiratory symptoms and lung function in controlled human
22 exposure studies, as most studies found no effect of NO2 (120-4,000 ppb) in adults with
23 asthma and, in one study, adolescents with asthma. Many studies examined subjects with
24 allergic asthma but did not challenge subjects with an asthma trigger. Only as examined
25 in Riedl et al. (2012). symptoms were not increased after a 3-hour exposure to 350 ppb
26 NC>2 and methacholine challenge.
27 Although coherence among disciplines is weak for NC^-related increases in asthma
28 symptoms and decreases in lung function, NC>2-induced increases in airway
29 responsiveness and allergic inflammation in experimental studies provide sufficient
30 biological plausibility for the effects of NC>2 exposure on asthma exacerbation. Increased
31 airway responsiveness can contribute to increases in asthma symptoms such as wheeze.
32 Controlled human exposure studies demonstrated increases in airway responsiveness in
33 adults with asthma at rest and doubling reduction in provocative dose in response to 200
34 to 300 ppb NO2 for 30 minutes and 100 ppb for 1 hour [Section 5.2.2.1. (Brown. 2015;
35 Folinsbee. 1992)]. The findings for clinically-relevant airway responsiveness with NC>2
36 exposures not much higher than peak concentrations near roadways (Section 2.5.3). in
37 particular, support an effect of ambient NC>2 exposures on asthma exacerbation. Further
38 characterizing mechanisms for NC>2-induced airway responsiveness and asthma
January 2015 5-237 DRAFT: Do Not Cite or Quote
-------
1 symptoms is evidence for NC>2 exposures of 260-400 ppb enhancing allergic
2 inflammation (e.g., eosinophil activation, Th2 cytokines, PMNs) in humans with allergic
3 asthma and a rat model of allergic disease (Ezratty et al.. 2014; Barck et al.. 2005a; Barck
4 et al.. 2002; Wangetal.. 1999) (Sections 4.3.2.6 and 5.2.2.5). NO2-associated increases
5 in pulmonary inflammation also were found in epidemiologic studies of populations with
6 asthma (Section 5.2.2.5). As allergic inflammation promotes bronchoconstriction and
7 airway obstruction, the evidence describes key events in the mode of action for
8 NO2-associated increases in respiratory symptoms found in populations of children with
9 asthma with prevalence of atopy ranging from 47 to 77%. In experimental studies,
10 ambient-relevant NC>2 exposures increased eicosanoids (involved in PMN recruitment)
11 but did not consistently affect lung injury or pulmonary oxidative stress (Section 5.2.7.4).
12 These inconsistent findings for other key events in the mode of action for asthma
13 exacerbation are not considered to weaken the evidence for a relationship with NO2
14 because the effects were studied mostly in healthy humans and animal models.
5.2.9.2 Evidence on Nonasthma Respiratory Effects
15 Epidemiologic studies demonstrate associations of ambient NO2 concentrations with
16 hospital admissions and ED visits for all respiratory causes combined (Table 5-45).
17 suggesting that the respiratory effects of short-term NC>2 exposure may extend beyond
18 exacerbation of asthma. However, when other respiratory conditions are evaluated
19 individually, there is uncertainty about relationships with NC>2 because of inconsistency
20 among disciplines and/or inconsistency among outcomes ranging from clinical events to
21 key events in modes of action. Where epidemiologic associations were found, limited
22 examination of potential confounding by traffic-related copollutants results in weak
23 inference about NC>2 effects. Experimental evidence for NCh-induced increases in airway
24 responsiveness and allergic inflammation supports effects on allergy exacerbation, but
25 epidemiologic evidence is inconsistent (Section 5.2.3). For COPD exacerbation and
26 respiratory infection (Sections 5.2.4 and 5.2.5). evidence from epidemiologic, controlled
27 human exposure, and toxicological studies is inconsistent across outcomes such as
28 hospital admissions, ED visits, symptoms, lung function, and immune cell function; thus,
29 a direct effect of NC>2 exposure is not clearly demonstrated (Table 5-45). Epidemiologic
30 studies consistently found NCh-associated increases in respiratory mortality
31 (Section 5.2.8). but the spectrum of respiratory effects that can lead to mortality is not
32 entirely clear. Among the leading causes of mortality, COPD, and respiratory infections
33 are the ones related to respiratory causes (Hoyert and Xu. 2012). but these conditions are
34 not clearly related to NC>2 exposure. Epidemiologic evidence also indicates ambient
35 NO2-associated respiratory effects in healthy populations (Section 5.2.7). as cough and
January 2015 5-238 DRAFT: Do Not Cite or Quote
-------
1 pulmonary inflammation in children in the general population and healthy adults.
2 However, an independent effect of NO2 is uncertain because of limited support from
3 experimental studies.
5.2.9.3 Evaluation of Nitrogen Dioxide Exposure Assessment
4 Most epidemiologic evidence indicating ambient NC>2-related asthma exacerbation is
5 based on exposure assessment from central site concentrations. Substantiating the
6 evidence are several findings for associations with NO2 concentrations spatially aligned
7 with subjects' location(s), including total and outdoor personal NCh as well as NC>2
8 measured outside schools (Greenwald et al.. 2013; Zoraetal.. 2013; Martins etal.. 2012;
9 Sarnat et al.. 2012; Spira-Cohen et al.. 2011; Delfino et al.. 2008a; Holguin et al.. 2007;
10 McCreanor et al., 2007). Ambient NC>2 concentrations are highly variable across
11 locations (Sections 2.5.2 and 2.5.3). Thus, compared to area-wide central site
12 concentrations, NC>2 concentrations in subjects' locations may better represent temporal
13 variation in subjects' ambient exposures in those locations. NC>2 concentrations summed
14 across individuals' microenvironments have shown good agreement with total personal
15 NC>2 (Section 3.4.3.1). demonstrating that microenvironmental ambient concentrations
16 are important determinants of exposure. Further supporting asthma exacerbation in
17 relation to ambient NC>2 exposure, for some study areas, central site concentrations were
18 reported to be correlated with total personal NO2 (Delfino et al., 2008a). outdoor school
19 NO2 (Sarnat et al.. 2012). or NC>2 measured at other central sites in the area
20 (Section 2.5.2). In support of exposure assessment from central sites, larger ambient
21 NO2-associated increases in respiratory hospital admissions and ED visits were found in
22 the warm season. Personal-ambient NC>2 correlations are higher in the warm than cold
23 season (Section 3.4.4.3). pointing to lower potential NC>2 exposure error.
24 The studies with microenvironmental exposure assessment provide some, albeit far from
25 conclusive, indication that short-term NC>2 exposures near sources may be related to
26 respiratory effects. In some cases, respiratory effects were observed in association with
27 ambient NC>2 measured across locations with varying traffic intensities or distance to
28 highways (Steenhof etal.. 2013; Strak et al.. 2012; Spira-Cohen etal.. 2011). but NO2
29 associations were not compared among locations. Other studies provided stronger
30 support, observing respiratory effects in association with NC>2 at a school in a high but
31 not low traffic area (Greenwald et al.. 2013; Sarnat etal.. 2012) or larger respiratory
32 effects near a steel plant than in a residential area (Dales etal.. 2013). However, none of
33 these studies examined whether the findings were attributable to NO2 independently of
34 correlated copollutants or differences between locations in population characteristics such
35 as race/ethnicity, body mass index, or asthma medication use. Informing these
January 2015 5-239 DRAFT: Do Not Cite or Quote
-------
1 uncertainties, McCreanor et al. (2007) observed that adults with asthma had larger
2 decreases in lung function and increases in eosinophil activation after walking along a
3 high traffic road in London, U.K. than after walking in a park. Observations that
4 associations with personal ambient NO2 persisted with adjustment for personal ambient
5 EC, UFP, or PM2 5 provide evidence for an independent relationship between respiratory
6 effects and NO2 exposures near high traffic roads.
7 Whether estimated from central sites or subjects' locations, NO2 exposure metrics largely
8 were integrated over 24 hours or 2-15 hours. The diurnal temporal pattern of exposure
9 (e-g-, acute peaks) underlying the associations of respiratory effects with daily average or
10 multiday averages of NO2 is not well characterized. However, studies conducted in
11 outdoor locations with varying traffic intensities indicate increases in pulmonary
12 inflammation and decreases in lung function in association with 2- or 5-h avg NO2
13 exposures that ranged between 5.7 and 153.7 ppb (Strak et al.. 2012; McCreanor et al..
14 2007).
5.2.9.4 Evaluation of Confounding
15 Also supporting an independent effect of NC>2 exposure on asthma exacerbation are
16 epidemiologic associations found for NC>2 with statistical adjustment for potential
17 confounding factors such as temperature, humidity, season, medication use, and, in
18 particular, copollutants. Based on a common source and moderate to high correlations
19 with NCh, confounding by other traffic-related pollutants is a major concern
20 (Sections 1.4.3 and 5.1.2.1). Copollutant models were the predominant method used for
21 evaluating copollutant confounding, and most of these studies found that NC>2
22 associations persisted with adjustment for PIVb 5, BC/EC, OC, UFP, PNC (Figure 5-16
23 and Table 5-44). PM2 5 metal components, VOCs, or CO (Figure 5-17 and Table 5-44).
24 Copollutant models also indicated that NO2 associations with asthma and other
25 respiratory effects were independent of PMio-2.5, PMio, SO2, or Os [Supplemental
26 Figure S5-1 (U.S. EPA. 2014a)1. O3 (r = -0.61 to 0.45) and PMio (r = -0.71 to 0.59)
27 showed a wide range of correlations with NO2, from strongly negative to moderately
28 positive; SO2 was moderately correlated with NO2 (r = 0.31-0.56). Inference regarding
29 confounding by traffic-related copollutants is strongest for exposure assessment in
30 subjects' locations. Exposure measurement error due to spatial variability may be similar
31 for NO2 and copollutants, thereby improving the reliability of copollutant models. These
32 studies reported a wide range of correlations between NO2 and traffic-related copollutants
33 (r = -0.42 to 0.68). Also strengthening inference from copollutant models, in some
34 studies, personal NO2 was not strongly positively correlated with personal copollutants
35 (r = 0.20-0.33 for EC, OC, PM2 5; -0.42 to 0.08 for ethylbenzene and benzene) (Martins
January 2015 5-240 DRAFT: Do Not Cite or Quote
-------
1 et al.. 2012: Delfino et al.. 2006). As examined in multiple populations with asthma,
2 associations of lung function and pulmonary inflammation with total or outdoor personal
3 NO2 and NO2 measured within 650 m of a children's school persisted with adjustment for
4 PM2 5 or BC/EC (Martins etal.. 2012: Lin etal.. 2011: Delfino et al.. 2008a: McCreanor
5 et al.. 2007: Delfino et al.. 2006). Results were similar in a study of healthy adults
6 (Steenhof etal.. 2013: Strak etal.. 2012). In some cases, the 95% CIs for NO2
7 associations were exaggerated because the increment used to standardize effect estimates
8 is far larger than the variability in NO2 concentrations during the study period (Martins et
9 al.. 2012: Strak et al.. 2012). Most studies examining exposures at subjects' locations did
10 not examine CO, either in single- or co-pollutant models. However, among children in
11 the general population, the association between outdoor school NO2 and lung function
12 persisted with adjustment for CO (Correia-Deur et al.. 2012).
13 As examined in only one or two studies with exposures assessed in subjects' locations,
14 NO2 associations persisted with adjustment for OC or metal PM2 5 components such as
15 iron and copper (Steenhof et al.. 2013: Strak etal.. 2012: Delfino et al.. 2006).
16 Information on potential confounding by UFP/PNC or VOCs also is limited, and results
17 from copollutant models are more variable. However, rather than clearly demonstrating
18 confounding of NO2 associations, results show that adjustment for UFP/PNC or benzene
19 attenuated one outcome in a study but not another (Steenhof et al.. 2013: Martins et al..
20 2012: Strak etal.. 2012: McCreanor et al.. 2007). There also was some evidence that NO2
21 exposure confounded associations for traffic-related copollutants. Some associations of
22 personal ambient PM2 5, EC/BC, OC, copper, UFP/PNC, or benzene with pulmonary
23 inflammation and lung function were attenuated with adjustment for personal ambient
24 NO2 (Martins etal.. 2012: Strak etal.. 2012: McCreanor et al.. 2007). Also supporting an
25 independent association for NO2, some studies found associations with NO2 but not EC,
26 OC, or PM2.5 for school or personal measurements (Sarnat et al.. 2012: Delfino et al..
27 2008a: Holguin et al.. 2007).
28 Copollutant models based on central site concentrations indicate that ambient NO2
29 remains associated with asthma- and non-asthma-related respiratory effects with
30 adjustment for PM2 5 (Iskandar etal.. 2012: Dales et al.. 2009: Jalaludin et al.. 2008:
31 Villeneuve et al.. 2007: von Klot et al.. 2002) or as examined in fewer studies, UFP, a
32 source apportionment factor comprising EC and various metals, CO, or VOCs (Delfino et
33 al.. 2013: Gent et al.. 2009: Tolbert et al.. 2007: Delfino et al.. 2003). Several
34 traffic-related PM constituents are shown to induce oxidative stress (Table 5-1). and
35 Delfino etal. (2013) found an NO2 association with adjustment for the oxidative potential
36 of PM2.5 extracts. Observations of larger ambient NO2-associated increases in
37 respiratory-related hospital admissions and ED visits in the warm rather than cold season
38 also support an independent NO2 association. NO2-PM2 5 correlations are lower in the
January 2015 5-241 DRAFT: Do Not Cite or Quote
-------
1 warm season (Section 3.4.5.1). pointing to lower potential confounding by PIVb 5. NO2
2 and Os are not strongly positively correlated in the warm season. As with NO2 measured
3 in subjects' locations, some central site NO2 associations were attenuated with adjustment
4 for a traffic-related copollutant. However, in the same studies, NO2 associations with
5 other outcomes persisted with PIVb 5 or UFP adjustment (Dales et al.. 2009; Liu et al..
6 2009; von Klot et al.. 2002). and a clear confounding effect was not demonstrated.
7 Similar to NO2 measured in subjects' locations, central site NO2 showed a range of
8 correlations with traffic-related pollutants (r = 0.43-0.74). However, because of different
9 spatial distributions (Section 3.3.1.1). exposure measurement error may differ between
10 central site concentrations of NO2 and traffic-related copollutants, resulting in weaker
11 inference from copollutant models.
12 Confounding by any particular copollutant was examined to a limited extent, and not all
13 potentially correlated pollutants were examined. Further, inference from copollutant
14 models can be limited (Section 5.1.2.2). and methods to adjust for multiple copollutants
15 simultaneously are not reliable. Thus, residual confounding is likely. However,
16 copollutant model results, particularly for pollutants measured in subjects' locations,
17 integrated with experimental findings, support an effect of ambient NO2 exposure on
18 asthma exacerbation independent of other traffic-related pollutants.
5.2.9.5 Evaluation of Nitrogen Dioxide-Copollutant Mixture
Effects
19 As a component of an air pollution mixture, NC>2 potentially can induce health effects in
20 combination with other pollutants in the mixture. Controlled human exposure studies,
21 with well-defined NCh-copollutant co-exposures, do not provide strong evidence that the
22 effects of NC>2 exposure on lung function and airway responsiveness differ when
23 occurring alone or as part of a mixture with PIVb 5 (Gong et al.. 2005). SC>2 (Devalia et al..
24 1994). or Os (simultaneous or sequential exposure) (Jenkins etal.. 1999; Hazuchaet al..
25 1994; Adams etal.. 1987). Interactions with CO, EC/BC, or UFP have not been
26 examined in controlled human exposure studies. But, limited epidemiologic findings
27 point to increases in asthma-related symptoms and ED visits in association with
28 short-term increases in ambient NC>2 concentration examined alone and jointly with
29 traffic-related pollutants such as PM2 5, CO, and EC (Gass et al.. 2014; Winquist et al..
30 2014; Schildcrout et al.. 2006) and with O^ and SO? (Winquist et al.. 2014). However.
31 there is no clear indication of combined (i.e., synergistic) effects of NO2 with PIVbs, CO,
32 EC, or VOCs larger than effects of individual pollutants alone (Gass etal.. 2014;
33 Winquist etal.. 2014; Schildcrout et al.. 2006; Delfino et al.. 2003). Inference from these
January 2015 5-242 DRAFT: Do Not Cite or Quote
-------
1 epidemiologic findings of joint effects is limited because exposure measurement error in
2 the central site pollutant metrics may obscure interactions between personal exposures.
5.2.9.6 Indoor Nitrogen Dioxide-Related Asthma Exacerbation
3 A causal relationship between NO2 and respiratory effects also is supported by the
4 coherence of asthma-related effects associated with ambient and indoor NO2
5 (Table 5-45). In schools in Ciudad Juarez, Mexico, correlations of NC>2 with BC, PIVbs,
6 PMio, PMio-2.5, and SC>2 differed indoors and outdoors, suggesting that NC>2 was part of a
7 different pollutant mixture indoors and outdoors. NC>2 also may be part of different
8 pollutant mixtures inside homes and classrooms because gas heaters and not stoves are a
9 major source of classroom NC>2. Cooking has been shown to be a more important
10 determinant of indoor UFP than heating systems (Weichenthal et aL 2007). Thus,
11 associations with indoor classroom NCh may be less likely to be confounded by UFP than
12 are associations with indoor home NC>2 or ambient NC>2. Mean concentrations of indoor
13 NC>2, averaged over 3 to 7 days, were in the range of ambient concentrations
14 (Table 5-45). except for a mean of 121 ppb at one school. Indoor NO2 concentrations,
15 particularly at home, can exhibit acute peaks that deviate from the mean (Table 3-4). As
16 with ambient NC>2, the temporal pattern of NC>2 concentrations underlying associations of
17 respiratory effects with multiday averages of indoor NC>2 is not understood.
5.2.9.7 Conclusion
18 Multiple lines of evidence support a relationship between short-term NC>2 exposure and
19 asthma exacerbation. Some findings point to effects on allergy, COPD, respiratory
20 infection, respiratory effects in healthy populations, and respiratory mortality, but there is
21 inconsistency among disciplines and outcomes. Increases in asthma hospital admissions,
22 ED visits, as well as respiratory symptoms and pulmonary inflammation in populations
23 with asthma are found in association with 24-h avg and 1-h max NC>2 concentrations, at
24 lags of 0 or 1 day and multiday averages of 2 to 5 days. Across studies finding
25 NO2-associated effects on asthma, the range of mean ambient concentrations was
26 11.3-30.9 ppb for 24-h avg NO2, 75.5 ppb for 2-h avg NO2, and 23.0-44.4 ppb for
27 1-h max NCh. The epidemiologic evidence is substantiated by the consistency of findings
28 among central site ambient NC>2 and NCh measured in subjects' location(s), including
29 personal, ambient school, ambient near-road, or indoor concentrations. Further,
30 associations of ambient and personal NCh with asthma-related effects persist with
31 adjustment for meteorological factors and/or another traffic-related pollutants such as
32 PM2 5, BC/EC, UFP, OC, metals, VOCs, or CO. Inference from copollutant models is
January 2015 5-243 DRAFT: Do Not Cite or Quote
-------
1 limited as is the breadth of analysis of traffic-related copollutants and copollutant
2 interactions. Thus, the coherence of epidemiologic findings for ambient and indoor NC>2
3 with effects demonstrated on key events in the mode of action for asthma exacerbation
4 provide strong evidence for an independent effect of NO2 exposure. Epidemiologic
5 evidence for NCh-associated asthma exacerbation persisting with adjustment for another
6 traffic-related pollutant and biological plausibility from NCh-induced increases in airway
7 responsiveness and allergic inflammation in adults with asthma and animal models of
8 asthma are sufficient to conclude that there is a causal relationship between short-term
9 NC>2 exposure and respiratory effects.
January 2015 5-244 DRAFT: Do Not Cite or Quote
-------
Study
PM2.5
McCreanor et al. (2007)a FEF25-75%
Delfino et al. (2008)
FEV1 %pred
Lin etal. (2011) eNO
Moshammer et al. (2006)a FEV1
ATSDR (2006) Asthma ED visits
Delfino etal. (2013)b eNO
Jalaludin et al. (2008) Asthma ED visits
Liu etal. (2009) FEV1
Correlation with NO2
Exposure Assessment
0.60
Personal ambient
0.38
Total personal
0.36
Central site within 8,16 km of homes
0.30
Central site 0.65 km from school
0.54
Central site next to school
0.61
Central site -1 site
0.43
Central site -1 or 2 sites per city
0.45 warm, 0.68 cold
Central sites - city average
Central site within 10 km of homes
Iskandar etal. (2012) Asthma Hospital Ad. 0.33
Central site within 15 km of hospitals
von Klot et al. (2002) Wheeze
von Klot et al. (2002) Beta-agonist Use
EC/BC
McCreanor et al. (2007)a FEF25-75%
Strak et al. (2012)
Strak et al. (2012)
Lin etal. (2011)
Gentetal.(2009)c
FVC
eNO
eNO
Wheeze
0.68
Central site 0.5 km from traffic
0.68
Central site 0.5 km from traffic
0.58
Personal ambient
0.67
Location of outdoor exposures
0.67
Location of outdoor exposures
0.68
Central site 0.65 km from school
0.49
Central site
acock et al. (2011) Symptomatic Fall PEF NR
Central site -1 site
PNC/UFP
McCreanor et al. (2007)a FEF25-75%
Strak et al. (2012)
Strak et al. (2012)
FVC
eNO
0.58
Personal Ambient
0.56
Location of outdoor exposures
0.56
Location of outdoor exposures
Iskandar etal. (2012) Asthma Hospital Ad. 0.51
Central site within 15 km ofhospitals
von Klot et al. (2002) Wheeze
von Klot et al. (2002) Beta-agonist Use
0.66
Central site 50 m from traffic
0.66
Central site 50 m from traffic
Bruskeetal. (2009)
Lymphocytes in BALF 0.66
Central site within 3.5 km of homes
Percent change in respiratory outcome per increase
in NO2 (95% Cl)d in single- and co-pollutant models
Note: EC/BC = elemental/black carbon, NO2 = nitrogen dioxide, PM2.5 = particles with an aerodynamic diameter less than equal to a
nominal 2.5 Lim, PNC = particle number concentration, UFP = ultrafine particles. Magnitude and precision of effect estimates should
not be compared among different outcomes. Results are organized by copollutants analyzed then by exposure assessment method.
Percentage change in FEF2s-75%, FEVi, or FVC refers to percentage decrease. Studies in red = recent studies, studies in
black = studies reviewed in the 2008 ISA for Oxides of Nitrogen. Effect estimates in closed circles = NO2 in a single-pollutant model,
effect estimates in open circles = NO2 adjusted for a copollutant. Quantitative results and abbreviations are described in Table 5-44.
aTo fit results in the figure, effect estimates are multiplied by 10. bCopollutant is ROS generated from PM25 extract. °Copollutant is a
source apportionment factor comprising EC and various metals. dEffect estimates standardized to a 20-ppb increase for 24-h avg
NO2 and a 30-ppb increase for 1-h max NO2. Effect estimates for 2-h, 5-h, or 15-h avg NO2 are not standardized but presented as
reported in their respective studies (see Section 5.1.2.3).
Figure 5-16 Associations of ambient or personal NO2 with respiratory effects
adjusted for PM2.5, EC/BC, or PNC/UFP.
January 2015
5-245
DRAFT: Do Not Cite or Quote
-------
Study
Benzene
Martins et al. (2012)
Martins et al. (2012)
Outcome
EBCpH
FEV1
Ethvlbenzene
Martins etal. (2012) EEC pH
Martins etal. (2012) FEV1
CO
Correia-Deur et al. (2012) PEF
Tolbert et al. (2007)
Respiratory ED visits
Jalaludin et al. (2008) Asthma ED visits
Correlation with NO2
Exposure Assessment
-0.43 to 0.14
Individual TWA
-0.43-0.14
Individual TWA •*
-0.43 to 0.14
Individual TWA
-0.43 to 0.14
Individual TWA
0.51
Outdoor school
0.70
Central sites - city average
0.71 warm, 0.55 cold
Central sites - city average
-10
1°-
k>
O
O
10
20
30
40
Percent change in respiratory outcome per increase
in NO2 (95% Cl)ain single-and co-pollutant models
Note: VOC = volatile organic compound, CO = carbon monoxide. Magnitude and precision of effect estimates should not be
compared among different outcomes. Results are organized by copollutants analyzed then by exposure assessment method.
Percentage change in EEC pH, FENA,, and PEF refers to percentage decrease. Studies in red = recent studies, studies in
black = studies reviewed in the 2008 ISA for Oxides of Nitrogen. Effect estimates in closed circles = NO2 in a single-pollutant model,
effect estimates in open circles = NO2 adjusted for a copollutants. Quantitative results and abbreviations are described in
Table 5-44.
aEffect estimates standardized to a 20-ppb increase for 24-avg NO2 and a 30-ppb increase for 1-h max NO2.
Figure 5-17 Associations of ambient nitrogen dioxide (NO2) with respiratory
effects adjusted for VOCs or CO.
January 2015
5-246
DRAFT: Do Not Cite or Quote
-------
Table 5-44 Corresponding effect estimates for nitrogen dioxide (NO2)-associated
respiratory effects in single- and co-pollutant models presented in
Figures 5-16 and 5-17.
% change in outcome(95% Cl)
per increase in NO2a
Study
McCreanor et al.
(2007)
Delfino et al.
(2008a)
Zhu(2013)l; Lin
etal. (2011)
Moshammer et
al. (2006)
ATSDR (2006)
Delfino et al.
(2013)
Jalaludin et al.
(2008)
Liu (2013); Liuet
al. (2009)
Iskandar et al.
(2012)
Respiratory
Outcome
FEF25-75%
FEVi %
predicted
eNO
FEVi
Asthma ED
visits
eNO
Asthma ED
visits
FEVi
Asthma
hospital
admissions
NO2 Averaging
Time and Lag
2-h avg
Lag 0 h
24-h avg
Lag 0-1 -day
avg
24-h avg
Lag 0 day
8-h avg
(12-8 a.m.)
Lag 0 day
24-h avg
Lag 0-4 day avg
24-h avg
Lag 0-1 day avg
1-h max
Lag 0-1 day avg
24-h avg
Lag 0-2 day avg
24-h avg
Lag 0-4 day avg
Exposure
Assessment
Method
Personal
ambient
Total personal
Central site
within 8 or
16-km of
homes
Central site
0.65-km of
school
Central site
next to school
Central site:
1 site
Central site: 1
or 2 sites per
city
Central site:
city average
Central site
within 10 km
of homes
Central site
within 15 km
of hospital
Correlation
with NO2
0.60
0.58
0.58
0.38
0.36
0.30
0.68
0.54
0.61
0.43
0.45 warm
0.68 cold
0.71
0.33
0.51
Single-
Pollutant Model
7.8(1.3, 14)b
per 5.3 ppb NO2
1.7(0.19,3.2)
1.3(0.15,2.4)
22(18, 26)
8.9(3.7, 14)b
per 5.32 ppb
NO2
5.8(0.59, 11)
9.0(2.9, 15)
7.4(4.5, 10)
1.2 (-0.84, 3.2)
30(10,60)
Copollutant
model
With PlVh.s:
4.8 (-2.5, 13)b
With EC:
4. 3 (-2.6, 11)b
With UFP:
4.7 (-3.9, 13)b
With PIvh.s:
1.3 (-0.22, 2.8)
With PIvh.s:
0.86 (-0.89, 2.6)
With PIvh.s:
14(9.5, 19)
With BC:
5.6(0.38, 11)
With PlVh.s:
10(4.2, 16)b
With PlVh.s:
3.5 (-1.8, 9.0)
With PlVh.s:
5.8 (-1.9, 14)c
With PlVh.s:
4.5(1.3, 7.8)
With PlVh.s:
-1.2 (-6.4, 3.8)
With PlVh.s:
40 (20, 70)
With UFP:
50 (20, 80)
January 2015
5-247
DRAFT: Do Not Cite or Quote
-------
Table 5-44 (Continued): Corresponding effect estimates for nitrogen dioxide
(NO2)-associated respiratory effects in single- and co-
pollutant models presented in Figures 5-16 and 5-17.
% change in outcome(95% Cl)
per increase in NO2a
Study
von Klot et al.
(2002)
Strak(2013);
Strak et al.
(2012)
Gent et al.
(2009)
Peacock et al.
(2011)
Bruske(2014):
Bruske et al.
(2010)
Res p i rato ry N 02 Ave raging
Outcome Time and Lag
Wheeze 24-h avg
Lag 0-4 day avg
Beta agonist
use
FVC 5-h avg
Lag 0 h
eNO
Wheeze NR
Lag 0 day
Symptomatic 1-h max
fall in PEF Lag 1 day
Lymphocytes 24-h avg
in BAL fluid Lag 0-23 h
Exposure
Assessment
Method
Central site 50
m from traffic
Location of
outdoor
exposures
Central site
Central site:
1 site
Central site
within 3.5 km
of homes
Correlation Single- Copollutant
with NO2 Pollutant Model model
0.68 15(2.0,28)
0.66
0.68 20 (5.0, 37)
0.66
0.67 1.8(0.44,3.2)
per 10.54 ppb
NO2
0.56
0.67 6.9 (-1.9, 16)
per 10.54 ppb
NO2
0.56
0.49 NR
NR 13 (-3.0, 31)
0.66 8.4 (-5.0, 24)
With PlVh.s:
12 (-7.0, 35)
With UFP:
2.0 (-14, 21)
With PIvh.s:
25 (5.0, 49)
With UFP:
22 (5.0, 43)
With EC:
2.3(0,4.6)
With PNC:
1.3 (-0.58, 3.1)
With EC:
4.1 (-6.0, 14)
With PNC:
-7.4 (-19, 3.9)
With source
apportionment
factor of EC, zinc,
copper, lead:
8.0 (-1.0, 18)
With BS:
6.2 (-17, 34)
With UFP:
8.4 (-7.2, 27)
January 2015
5-248
DRAFT: Do Not Cite or Quote
-------
Table 5-44 (Continued): Corresponding effect estimates for nitrogen dioxide
(NO2)-associated respiratory effects in single- and co-
pollutant models presented in Figures 5-16 and 5-17.
% change in outcome(95% Cl)
per increase in NO2a
Study
Martins (201 3):
Martins et al.
(2012)
Correia-Deur et
al. (2012)
Tolbert (2009):
Tolbert et al.
(2007)
Jalaludin et al.
(2008)
Respiratory
Outcome
EBCpH
FEVi
PEF
Respiratory
ED visits
Asthma ED
visits
NO2 Averaging
Time and Lag
24-h avg
Lag 0-4 day avg
24-h avg
Lag 0 day
1-h max
Lag 0-2 day avg
1-h max
Lag 0-1 day avg
Exposure
Assessment
Method
Individual
TWA based
on outdoor
monitoring,
modeling,
time-activity
data
Outdoor
school
Central sites:
city average
Central sites:
city average
Correlation Single- Copollutant
with NO2 Pollutant Model model
With benzene:
-0.43 to 2.6(1.3,3.9) 1.7 (-0.26, 3.6)
0.14 across
time periods
With
ethylbenzene:
1.6 (-0.49, 3.7)
With benzene:
22(1.5,38) 3.6 (-31, 29)
With
ethylbenzene:
17 (-17, 41)
With CO:
0.51 1.9 (-0.38, 4.1) 1.5(0,3.0)
With CO:
0.70 2.6(1.3,3.9) 2.2(0.78,3.7)
With CO:
0.71 warm 7.4(4.5,10) 4.2(0.78,3.7)
0.55 cold
FEF25-75% = Forced expiratory flow between 25 and 75% of forced vital capacity, FENA = forced expiratory flow in 1 second,
eNO = exhaled breath condensate, ED = emergency department, FVC = forced vital capacity, PEF = peak expiratory flow, NR = not
reported, EEC = exhaled breath condensate, TWA = time-weighted average.
asingle- and copollutant model results are standardized to a 20-ppb increase in 24-h avg NO2 and 30-ppb increase in 1-h max NO2.
Results based on other averaging times are not standardized but presented as reported in their respective studies (Section 5.1.2.3).
Percentage change in FEF2s-75%, FE\A, FVC, PEF, and EEC pH refers to percentage decrease.
To fit results in Figure 5-16, results are multiplied by 10.
°Copollutant specifically is reactive oxygen species generated from ambient PM25 extracts.
January 2015
5-249
DRAFT: Do Not Cite or Quote
-------
Table 5-45 Summary of evidence for a causal relationship between short-term
nitrogen dioxide (NO2) exposure and respiratory effects.
Rationale for
Causal
Determination3
Key Evidence13
Key References'3
NO2
Concentrations
Associated with
Effects0
Asthma Exacerbation
Consistent
epidemiologic
evidence from
multiple,
high-quality studies
at relevant NO2
concentrations
Increases in asthma hospital
admissions, ED visits in diverse
populations in association with 24-h avg
and 1-h max NO2, lags 0 and 3 to 5-
day avg among all ages and children.
Strickland et al. (2010).
Villeneuve et al. (2007), Jalaludin
et al. (2008). Ito et al. (2007a).
Iskandar et al. (2012). ATSDR
(2006)
Section 5.2.2.4. Figure 5-7
Overall study
mean 24-h avg:
11.3-31.3 ppb
Overall study
mean 1-h max:
23-44 ppb
No association in recent Canadian
multicity study.
Stieb et al. (2009)
Mean 24-h avg:
21.4-41.2 ppb
Coherence with increases in respiratory
symptoms and decrements in lung
function in populations with asthma in
association with 24-h avg, 2-4 h avg
NO2, 1-h max, lags 0, 3 to 6-day avg.
Panel studies of children examined
representative populations recruited
from schools.
No reports of selective participation by
particular groups.
Schildcrout et al. (2006). Gent et
al. (2009), Zora et al. (2013),
Greenwald et al. (2013). Holquin
et al. (2007). Delfino et al.
(2008a), McCreanor et al. (2007)
Sections 5.2.2.2 and 5.2.2.3,
Figure 5-3 and Figure 5-4
Outdoor school
mean 1-week
avg:
3.4-18.2 ppb
Personal outdoor
2-h avg: 75.5 ppb
Total personal
mean 24-h avg:
28.6 ppb
City mean
24-h avg:
17.8-26 ppb
Consistent
evidence for NO2
metrics with lower
potential for
exposure
measurement error
Asthma-related effects associated with Greenwald et al. (2013). Hole
NO2 measured in subjects' locations:
total and outdoor personal, school
outdoor.
Better spatial alignment with subjects
compared to central site NO2.
et al. (2007). Delfino et al.
(2008a), McCreanor et al. (2007),
Sarnat et al. (2012). Zora et al.
(2013), Delfino et al. (2006)
Consistent
evidence from
multiple,
high-quality
controlled human
exposure studies
Rules out chance,
confounding, and
other biases with
reasonable
confidence
NO2 increases airway responsiveness in
adults with asthma exposed at rest
following nonspecific or allergen
challenge in several individual studies
and meta-analyses.
Clinical relevance supported by findings
of a doubling reduction in provocative
dose in response to NO2.
Folinsbee (1992), Brown (2015)
Section 5.2.2.1, Table 5-4, Table
5-5. and Table 5-6
Any change:
100 ppb for 1 h
200-300 ppb for
30 min
Doubling
reduction in PD:
100 ppb for 1 h,
140 ppb for
30 min
January 2015
5-250
DRAFT: Do Not Cite or Quote
-------
Table 5-45 (Continued): Summary of evidence for a causal relationship between
short-term nitrogen dioxide (NO2) exposure and
respiratory effects
Rationale for
Causal
Determination3
Key Evidence13
Key References'3
NO2
Concentrations
Associated with
Effects0
Epidemiologic
evidence helps rule
out chance,
confounding, and
other biases with
reasonable
confidence
NO2 associations with lung function and
pulmonary inflammation persist in
copollutant models with a traffic-related
copollutant: PlVhs, EC/BC, OC, UFP, or
VOCs in studies with exposure
assessment in subjects' locations.
Ambient and total personal NO2
weakly-moderately correlated with other
traffic-related pollutants in some studies
(r= -0.43 to 0.49).
Delfino et al. (2006). Delfino et al. Same as above
(2008a), Martins et al. (2012),
McCreanoret al. (2007)
Figure 5-16 and Figure 5-17,
Table 5-44
Most central site NO2 associations
persist with adjustment for PlVh.s,
EC/metals factor, UFP, or CO.
Differential exposure measurement
error limits inference from copollutant
models based on central site NO2 and
copollutants.
Villeneuve et al. (2007). Jalaludin
et al. (2008), Gent et al. (2009),
von Klot et al. (2002)
Some associations were attenuated
with adjustment for PlVh.s or UFP.
Liu et al. (2009), von Klot et al.
(2002)
Most associations for
microenvironmental and central site
NO2 persist in copollutant models with
PM-io, SO2, orOs.
Supplemental Figure S5-1
(U.S. EPA. 2014a)
Indoor NO2 associated with increases in Sarnat et al. (2012). Lu et al.
respiratory effects in children with (2013), Hansel et al. (2008)
asthma. |\|0 association in Greenwald et
al. (2013)
Means of 3-to
7-day avg:
18.7-121 ppb
75th: 31 ppb
Max: 394 ppb
NO2 associations persist with
adjustment for meteorology, time
trends, season, medication use.
Evidence for Key Events in Mode of Action
Allergic responses
Increases in eosinophil activation, IgE,
Th2 cytokines in adults with asthma.
Barck et al. (2005a), Barck et al.
(2002), Wang etal. (1995a),
Ezrattyetal. (2014)
Sections 4.3.2.6, and 5.2.2.5
Figure 4-1
Humans:
260 ppb
15-30 min,
400 ppb 6 h,
581 ppb 30 min
on 2 days
January 2015
5-251
DRAFT: Do Not Cite or Quote
-------
Table 5-45 (Continued): Summary of evidence for a causal relationship between
short-term nitrogen dioxide (NO2) exposure and
respiratory effects
Rationale for
Causal
Determination3
Key Evidence13
Key References'3
NO2
Concentrations
Associated with
Effects0
Inflammation Increases in PMNs and prostaglandins
in healthy adults.
Section 5.2.7.4
1,500-3,500 ppb
20 min or 3-4 h
Increases in eNO in children with
asthma in association with 24-h avg
NO2.
Delfino et al. (2006), Sarnat et al.
(2012). Martins et al. (2012)
Section 5.2.7.4
Total personal
mean 24-h avg:
24.3, 30.9 ppb
Ambient mean
1-week avg:
4.5-20 ppb
Inconsistent effects
on oxidative stress,
pulmonary injury
See Respiratory Effects in Healthy
Individuals below.
COPD Exacerbation
Inconsistent
epidemiologic
evidence and
uncertainty
regarding NO2
independent effects
Increases in COPD hospital admissions
and ED visits.
Faustini et al. (2013), Ko et al.
(2007b), Arbex et al. (2009)
Section 5.2.4.2
Mean 24-h avg:
24.1-63.0 ppb
Mean 1-h max:
63.0 ppb
Inconsistent associations with lung
function decrements and symptoms in
adults with COPD.
Section 5.2.4.1
Inconsistent
evidence from
controlled human
exposure studies
Decreased lung function not
consistently found in adults with COPD.
Morrow et al. (1992). Vaqaqqini
etal. (1996)
Section 5.2.4.1
300 ppb for 1 h,
4h
Weak evidence for
key events in mode
of action
Increased inflammation in healthy adults Bruske et al. (2010)
but not in adults with COPD. Sections 5.2.4.3 and 5.2.7.4
1,500-3,500 ppb
20 min or 3-4 h
Respiratory Infection
Consistent animal
toxicological
evidence with
relevant NO2
exposures
Mortality from bacterial or viral infection
in animals with relevant NO2 exposures.
Ehrlich et al. (1977). Ehrlich et al.
(1979), Ehrlich (1980), Graham
etal. (1987)
Section 5.2.5.1
1,500-5,000 ppb
forSh
1,500 ppb with
4,500 ppb spike
of 1-7.5 h
Inconsistent
epidemiologic
evidence and
uncertainty
regarding NO2
independent effects
Associations with hospital
admissions/ED visits for respiratory
infections. All results based on central
site NO2, and some have wide 95% CIs.
Inconsistent evidence for parental
reports of infection or laboratory-
confirmed infections.
Zemek etal. (2010). Mehta et al. Overall study
(2013), Stieb etal. (2009),
Faustini etal. (2013). Just et al.
(2002). Stern etal. (2013)
Sections 5.2.5.3, and 5.2.5.2.
mean 24-h avg:
11.7-28.6 ppb
City mean 24-h
avg:
9.3-34.6 ppb
January 2015
5-252
DRAFT: Do Not Cite or Quote
-------
Table 5-45 (Continued): Summary of evidence for a causal relationship between
short-term nitrogen dioxide (NO2) exposure and
respiratory effects
Rationale for
Causal
Determination3
Key Evidence13
Key References'3
NO2
Concentrations
Associated with
Effects0
Limited Evidence for Key Events in Mode of Action
Decreased alveolar Diminished superoxide production in Section 5.2.5.4
macrophage AM function.
function NO consjstent effect on pulmonary
clearance.
Respiratory Effects in Healthy Individuals
Limited
epidemiologic
evidence and
uncertainty
regarding NO2
independent effects
Consistent evidence for respiratory
symptoms in children. All based on
central site NO2 and no examination of
confounding by traffic-related
copollutants.
Schwartz et al. (1994)
Section 5.2.7.3, Table 5-39
Mean 24-h avg:
13 ppb
Lung function not consistently
associated with NO2 measured at
subjects' locations or central site NO2
correlated (r= 0.61) with total personal.
But, personal ambient NO2 associations
found with adjustment for PM2.5, EC,
OC, copper, iron, orUFP.
Straket al. (2012). Moshammer
etal. (2006), Linn et al. (1996)
Section 5.2.7.2. Table 5-36
Max for 5-h avg:
96 ppb
Max for 24-h avg:
96 ppb
75th for 24-h avg:
11.4 ppb
Limited and
inconsistent
evidence from
controlled human
exposure studies
Increases in airway responsiveness Folinsbee (1992), Kjaerqaard
found in healthy adults above 1,000 ppb and Rasmussen (1996)
NO2, not lower concentrations. Section 5.2.7.1
1,000-2,000 ppb
forSh
Respiratory symptoms or lung function
examined in adults; changes generally
not found.
Sections 5.2.7.2 and 5.2.7.3
200-4,000 ppb
for 2-5 h
Limited Evidence for Key Events in Mode of Action
Inflammation Increases in PMNs and prostaglandins
in healthy adults.
Frampton et al. (2002), Frampton
etal. (1989)
Section 5.2.7.4
1,500-3,500 ppb
forSh
Limited epidemiologic evidence for
associations of NO2 measured in
subjects' locations with increases in
pulmonary inflammation in children and
adults. Associations persist with
adjustment for BC/EC, OC, UFP, or
PM2.5.
Straket al. (2012). Steenhof et Mean 24-h avg:
al. (2013), Lin etal. (2011)
Section 5.2.7.4. Table 5-41
9.3, 33 ppb
Mean 5-h avg
across sites with
varying traffic:
20 ppb
January 2015
5-253
DRAFT: Do Not Cite or Quote
-------
Table 5-45 (Continued): Summary of evidence for a causal relationship between
short-term nitrogen dioxide (NO2) exposure and
respiratory effects
Rationale for
Causal
Determination3
Inconsistent effects
on oxidative stress,
pulmonary injury
Key Evidence13 Key References'3
Inconsistent chanaes in antioxidants in Sections 4.3.2.3 and 5.2.7.4
experimental studies but found in
humans and rodents with lower dietary
antioxidant vitamins.
Increases in LDH, CC16, BAL fluid Sections 4.3.2.4 and 5.2.7.4
protein inconsistently found in humans,
rodents. Limited evidence for impaired
epithelial barrier function.
NO2
Concentrations
Associated with
Effects0
Humans:
2,000 ppbfor4 h,
1 day or 4 days
Rodents:
1, 000-5,000 ppb
for 3-7 days
Humans:
600-2,000 ppb
for 3-4 h,
1-4 days
Animal models:
400-2,000 ppb
for 1-3 weeks
Respiratory Mortality
Consistent
epidemiologic
evidence but
uncertainty
regarding NO2
independent effect
Multicity studies consistently observe
associations of respiratory mortality with
24-h avg NO2 at lag 0-1 days.
Results based on NO2 averaged across
central sites.
Potential confounding by traffic-related
copollutants not assessed. NO2 results
robust to adjustment for PM-io, SO2, Os.
Wong et al. (2008). Chen et al.
(2012b), Chiusolo et al. (2011),
Bellini etal. (2007). Bigger! etal.
(2005)
Section 5.2.8
Means across
cities for
24-h avg:
13.5-55.5 ppb
Uncertainty due to
limited coherence
with respiratory
morbidity evidence
Evidence for asthma exacerbation in
adults but limited coherence among
lines of evidence for effects on COPD
and respiratory infection. Uncertainty
regarding spectrum of effects that can
lead to respiratory mortality.
CC16 = club cell protein, Cl = confidence interval, CO = carbon monoxide, COPD = chronic obstructive pulmonary disease,
EC = elemental carbon, ED = emergency department, eNO = exhaled nitric oxide, LDH = lactate dehydrogenase, NO2 = nitrogen
dioxide, O3 = ozone, OC = organic carbon, PM = particulate matter, PMN = polymorphonuclear cells, SO2 = sulfur dioxide,
Th2 = T-derived lymphocyte helper 2, UFP = ultrafine particles, VOC = volatile organic compound.
aBased on aspects considered in judgments of causality and weight of evidence in causal framework in Tables I and N. of the
Preamble.
"•Describes the key evidence and references, supporting or contradicting, contributing most heavily to causal determination and,
where applicable, to uncertainties or inconsistencies. References to earlier sections indicate where full body of evidence is
described.
°Describes the NO2 concentrations with which the evidence is substantiated (for experimental studies, < 5,000 ppb).
January 2015
5-254
DRAFT: Do Not Cite or Quote
-------
5.3 Cardiovascular and Related Metabolic Effects
5.3.1 Introduction
1 The 2008 ISA for Oxides of Nitrogen concluded that the "available evidence on the
2 effects of short-term exposure to NC>2 on cardiovascular health effects was inadequate to
3 infer the presence or absence of a causal relationship" (U.S. EPA. 2008a). Multiple
4 studies found associations between short-term exposure to NO2 and rates of hospital
5 admission or ED visits for cardiovascular diseases (CVDs), yet it was unclear at that time
6 whether these results supported a direct effect of short-term NO2 exposure on
7 cardiovascular morbidity or were confounded by other correlated pollutants.
8 Additionally, epidemiologic studies available at the time of the last review provided
9 inconsistent evidence for associations between short-term NC>2 exposure and other
10 cardiovascular events such as arrhythmia among patients with implanted cardioverter
11 defibrillators and subclinical measures associated with cardiovascular events, such as
12 heart rate variability (HRV) and electrocardiographic (ECG) markers of cardiac
13 repolarization. Experimental studies available at the time of the 2008 ISA for Oxides of
14 Nitrogen did not provide biological plausibility for the cardiovascular effects observed in
15 epidemiologic studies. There was limited evidence from controlled human exposure
16 studies demonstrating a reduction in hemoglobin and some evidence from toxicological
17 studies for effects of NO2 on various hematological parameters in animals, but these
18 studies were limited and inconsistent. Overall, the experimental studies could not address
19 the uncertainty related to copollutant confounding in epidemiologic studies of hospital
20 admission or ED visits for CVDs in the 2008 ISA for Oxides of Nitrogen.
21 The following sections review the published studies pertaining to the cardiovascular and
22 related metabolic effects of short-term exposure to oxides of nitrogen in humans, animals,
23 and cells. When compared to the 2008 ISA for Oxides of Nitrogen, the recent
24 epidemiologic and toxicological studies provide evidence for effects of NO2 exposure on
25 a broader array of cardiovascular effects and mortality. Still, substantial uncertainties
26 remain concerning potential confounding by other traffic-related pollutants, exposure
27 measurement error, and the limited mechanistic evidence to describe a role for NO2 in the
28 manifestation of cardiovascular diseases, including key events within the mode of action.
29 The majority of the recent evidence is from epidemiologic studies, which suggest that
30 exposure to NO2 may result in the triggering of MI. To clearly characterize the evidence
31 underlying causality, the discussion of the evidence is organized into groups of related
32 outcomes [e.g., MI including ischemic heart disease (IHD), arrhythmia and cardiac
33 arrest]. Evidence for subclinical effects (e.g., HRV, blood biomarkers of cardiovascular
January 2015 5-255 DRAFT: Do Not Cite or Quote
-------
1 effects) that potentially underlie the development, progression, or indication of various
2 clinical events is discussed in Section 5.3.11. and may provide biological plausibility for
3 multiple outcomes.
5.3.2 Myocardial Infarction
4 Several lines of evidence are discussed in support of a relationship between short-term
5 NC>2 exposure and MI. An MI or heart attack occurs as a consequence of IHD, resulting
6 in insufficient blood flow to the heart that overwhelms myocardial repair mechanisms
7 and leads to muscle tissue death. ICD codes for MI are classified within the group of
8 IHDs; thus, studies where IHD is evaluated will include any patients diagnosed with an
9 MI. In addition, IHD includes the diagnosis of angina. Symptoms of MI are similar to
10 those of angina; however, where MI results in damage to the heart muscle, angina does
11 not result in myocardial necrosis. As angina may indicate an increased risk for future MI,
12 studies analyzing outcomes of angina are discussed in support of a relationship to MI.
13 Finally, acute MI may be characterized by ST segment depression, a nonspecific marker
14 of myocardial ischemia. The evaluation of evidence supporting a relationship between
15 short-term NC>2 exposure and triggering an MI includes hospitalization and ED visits for
16 MI, IHD, or angina, and ST-segment amplitude changes.
5.3.2.1 Hospital Admissions and Emergency Department Visits
for Myocardial Infarction and Ischemic Heart Disease
17 The 2008 ISA for Oxides of Nitrogen concluded that the epidemiologic evidence
18 consistently supported the associations between short-term increases in ambient NC>2
19 concentrations and hospital admissions or ED visits for cardiac diseases (U.S. EPA.
20 2008a). This conclusion continues to be supported by studies published since the 2008
21 ISA, as reviewed below (Figure 5-18 and Table 5-46). However, potential copollutant
22 confounding, especially from other traffic-related pollutants (e.g., EC, CO), and limited
23 mechanistic evidence are still key uncertainties, and make it difficult to interpret the
24 results of these studies. Additionally, all of the studies in this section use central site
25 monitors to estimate ambient NO2 exposure, which may result in misclassification of the
26 exposure due to the high variability in NO2 (Section 3.4.5.1).
27 A number of studies rely on clinical registries, which are generally less susceptible to
28 misclassification of the outcome and exposure. The strongest evidence of an association
29 between ambient NO2 and the risk of MI comes from a study using clinical registry data
30 from the U.K.'s Myocardial Ischaemia National Audit Project (Bhaskaran et al.. 2011).
January 2015 5-256 DRAFT: Do Not Cite or Quote
-------
1 which found a 5.8% (95% CI: 1.7, 10.6) increase in risk of MI per 30-ppb increase in
2 1-h max NC>2 concentrations in the 6 hours preceding the event. This study is unique
3 because it included detailed data on the timing of MI onset in more than 79,000 patients
4 from 15 conurbations in England and Wales, which allowed examination of association
5 with ambient NC>2 in the hours preceding MI. NC>2 results were robust to a number of
6 sensitivity analyses that evaluated key aspects of study design and model specification
7 (e.g., stricter diagnosis criteria, different time strata). Additionally, Bhaskaran et al.
8 (2011) restricted analyses to urban areas to reduce heterogeneity that may have resulted
9 in measurement bias from the use of fixed site monitors to assess NC>2 exposure. The
10 findings for NC>2 were more pronounced in those aged between 60 and 80 years, among
11 those with prior coronary heart disease, and for events occurring in the autumn and
12 spring. Conversely, in a smaller study of only 429 MI events, Turin et al. (2012) did not
13 observe a consistent positive association using data from the Takashima County Stroke
14 and AMI Registry in Central Japan. Cases were cross-checked by research physicians,
15 epidemiologists, and cardiologists, thereby minimizing potential misclassification of the
16 outcome.
January 2015 5-257 DRAFT: Do Not Cite or Quote
-------
Study Outcon
Thach et a\. (2O1O) IHD
Qiu et al. {in Press) IHD
Hsieh et al. (2O1O) Ml
Tsaiet al (2O12) Ml
Cheng et al. (2OO9) Ml
Bell et al. (2OO8) IHD
Gogginset al. (2O14) Ml
Szyszkowicz (2OO9) Angina
Szyszkowicz (2OO7) IHD
Sliebet al (2OO9) Ml
Vencloviene et al, (2O11)MI
Turin et al. (2012) Ml
Nuvolone et al. (2O11) Ml
Wichmann et al. (2012) Ml
larrieu et al. (2OO7J IHD
Rich et al. (2O1O) Ml
Manner al. (2QO2) IHD
Poloniecki et al. (1997) IHD
Ml
Angina
Linn et al. (2OOO) Ml
Wonget al. (1999) IHD
Ponka(1996) IHO
Barnett et al. (2OO6) IHD
Mi
von Klot et al. (2OOS) Ml
Angina
Bhaskaranetal. (2O11) Ml
Atkinsonet al. (1999) IHD
Peel et aL (2OO7) IHD
Jalaludin et al. (2OO6) IHD
Simpson et al. (2OO5) IHD
Concentration
31.2
3O.8S
29.88
27.59
26.5
26.4
23.5-29. 0
20.1
19.4
18.4
18.4
15.2-21.1
11.9 24.9
37.2
35
28-41
27-3
20. 7
72.8
15.7-23.2
16.5
5O.3
45.9
23.2
16.3-23.7
Lag
0 1
a ••
O-3
0 |
0 '
'.
i
0 '
o
i
0 '
a
1
O-3
O
0
o
t
i
0
I
' • J
1
a
1
0
i
O-l
o
1
0-1
o
1
0-1
o
1
O-l
1
1
1
o
O-l
o
o
1
0-1
0-1
0 1
1 6 hrs
7-12 hr
13-18 h
19-24 h
O-2
O
O
O-2
O-2
O
1
O-l
Notes
whole year
warm season
I'.':, il '-.•• UOI
23+ C
<23C
23+ C; non-hypertensive
<23 C; non-hypertensive
25+ C
all monitors (13)
city monitors (5)
correlated monitors (8)
Hong Kong
Taipei< Taiwain
Khaosiung, Taiwan
<65 yrs
yrs
Cold
sARR
SCHF
sNO
NO2
NO
> 65 yrs
15-64 vrs
15-64 yrs
rs
O-64yrs
65+
case-crossover
time series
> 65 yrs
! 24-h
1 "*" m
1
I*
•t
1 .
1
|
|,
^
i
IS
1.
i
.1 •
1 •
-"% 1-h
1^
1 •
1 1.25 l.S 1.75
Relative Risk and 95°/o CI
Note: Studies in red are recent studies. Studies in black were included in the 2008 ISA for Oxides of Nitrogen. Relative risks are
standardized to a 20-ppb or 30-ppb increase in NO2 or NO concentration for 24-h and 1-h averaging times, respectively. Studies are
presented in descending order of mean NO2 concentration, with the largest mean concentration (ppb) at the top and the smallest at
the bottom of the figure (by averaging time and inclusion in previous ISA). Circles = NO2; Triangles = NO.
Figure 5-18 Results of studies of short-term exposure to oxides of nitrogen
and hospital admissions for ischemic heart disease.
January 2015
5-258
DRAFT: Do Not Cite or Quote
-------
Table 5-46 Corresponding risk estimates for hospital admissions for ischemic
heart disease for studies presented in Figure 5-18.
Study
Location
Health
Effect
Relative Risk3 (95% Cl) Copollutant Examination13
tThachetal. (2010) Hong Kong, China IHD
Lag 0-1: 1.04 (1.02, 1.05) No copollutant models.
tQiuetal.,2013
Hong Kong, China IHD
All year, lag 0-3:
1.09(1.08, 1.11)
Warm season, lag 0-3:
1.05(1.03, 1.08)
Cool season, lag 0-3:
1.15(1.12, 1.18)
No copollutant models.
NO2 and PM-io correlation:
Pearson r= 0.76.
tHsieh etal. (2010) Taipei, Taiwan
>23°C: 1.24(1.16, 1.35)
<23°C: 1.26(1.18, 1.35)
NO2: robust to PM-io, SO2,
CO, or Os inclusion in
copollutant models.
Copollutants: all but Os
attenuated by NO2
adjustment.
NO2 correlations (Pearson r):
PMio: 0.55; SO2: 0.51; CO:
0.71; O3:0.02.
tTsai etal. (2012) Kaohsiung, IV
Taiwan
tChenq et al. (2009) Kaohsiung, IV
Taiwan
II Hypertension
>23°C, lag 0-2:
1.24(1.12, 1.38)
<23°C, lag 0-2:
1.29(1.16, 1.44)
No Hypertension
>23°C, lag 0-2:
1.29(1.18, 1.40)
<23°C, lag 0-2:
1.24(1.14, 1.35)
II >25°C, lag 0-2:
1.23(1.06, 1.44)
<25°C, lag 0-2:
1.76(1.55,2.02)
No copollutant models
examined.
NO2 correlations (Pearson r):
PMio: 0.48; SO2: 0.45; CO:
0.77; O3:-0.01.
NO2: Attenuated by CO or Os
adjustment on warm days
and PMio on cool days.
Robust to SO2 adjustment.
Copollutants: All but CO and
Os on warm days attenuated
by NO2 adjustment.
NO2 correlations (Pearson r):
PMio: 0.73; SO2: 0.53; CO:
0.66; Os: 0.09.
tBell et al. (2008) Taipei, Taiwan
IHD All monitors
LagO: 1.10(1.02, 1.18)
Lag 1: 1.05(0.98, 1.13)
Lag 0-3: 1.03(0.98, 1.21)
City monitors
Lag 0: 1.09(1.02, 1.16)
Lag 1: 1.05(0.98, 1.12)
Lag 0-3:1.08 (0.99, 1.20)
Correlated monitors
LagO: 1.09(1.02, 1.17)
Lag 1: 1.05(0.98, 1.12)
No copollutant models.
January 2015
5-259
DRAFT: Do Not Cite or Quote
-------
Table 5-46 (Continued): Corresponding risk estimates for hospital admissions for
ischemic heart disease for studies presented in
Figure 5-18.
Study
Health
Location Effect
Relative Risk3 (95% Cl)
Copollutant Examination13
Lag 0-3: 1.08(0.97, 1.20)
tGoqqins et al.
(2013)
tSzyszkowicz (2009)
fSzyszkowicz (2007)
tStieb et al. (2009)
tVencloviene et al.
(2011)
Hong Kong, Ml
China; Taipei,
Taiwan;
Kaohsiung,
Taiwan
6 Canadian cities Angina
Montreal, Canada IHD
7 Canadian cities Ml
Kaunas, Lithuania Ml
Hong Kong
Lag 0: 1.04(1.02,
Taipei
LagO: 1.09(1.05,
Kaohsiung
LagO: 1.05(0.98,
Lag 0: 1.04(1.03,
Lag 1: 1.13(1.04,
LagO: 1.03(1.00,
Lag 1: 1.03(1.00,
<65 yrs, lag 0-1:
1.19(0.96, 1.48)
1.07)
1.13)
1.13)
1.05)
1.22)
1.05)
1.06)
No copollutant models.
No copollutant models.
No copollutant models.
NO2 association attenuated
by CO.
No results provided for other
pollutants.
>65yrs, lag 0-1:
0.97(0.81, 1.17)
tTurin etal. (2012)
Takashima
County, Japan
Lag 0: 1.14(0.92, 1.40)
Lag 1:0.87(0.70, 1.08)
NO2: robust to TSP or SO2
adjustment. Attenuated by
Os adjustment.
Copollutants: TSP and SO2
attenuated by NO2
adjustment. No associations
between Os and Ml
regardless of NO2
adjustment.
tNuvolone et al.
(2011)
Tuscany, Italy
LagO: 1.04(0.97, 1.12)
Lag 1: 1.08(1.00, 1.15)
NO2: robust to PM-io
adjustment; attenuated by
CO.
Copollutants: PM-io no longer
associated with Ml after NO2
adjustment. No association
between CO and Ml
regardless of NO2
adjustment.
No correlations provided.
tWichmann et al.
(2012)
Copenhagen,
Denmark
Warm season
Lag 0: 0.99 (0.86, 1.14)
Lag 1: 1.10(0.96, 1.26)
Lag 0-1: 1.07(0.93, 1.22)
Cool season
Lag 0: 1.01 (0.91, 1.11)
Lag 1: 1.08(0.98, 1.19)
Lag 0-1: 1.06(0.96, 1.17)
No copollutant models.
tLarrieu etal. (2007) 8 French cities
IHD
1.07(1.03, 1.10)
No copollutant models.
January 2015
5-260
DRAFT: Do Not Cite or Quote
-------
Table 5-46 (Continued): Corresponding risk estimates for hospital admissions for
ischemic heart disease for studies presented in
Figure 5-18.
Study
Location
Health
Effect
Relative Risk3 (95% Cl) Copollutant Examination13
tRich etal. (2010) New Jersey, U.S.
Lag 0, transmural:
1.14(0.96, 1.32)
NO2: slightly attenuated by
PM2.5 adjustment.
Copollutants: PIVh.s
association attenuated by
adjustment for NO2.
NO2 and PIVh.5 correlation:
r=0.44.
Mann et al. (2002) Los Anqeles, CA IHD
With secondary arrhythmia
LagO: 1.04(1.02, 1.06)
Lag 1: 1.02(1.00, 1.04)
Lag 0-1: 1.03(1.01, 1.05)
With secondary congestive
heart failure
LagO: 1.05(1.01, 1.08)
Lag 1: 1.04(1.01, 1.08)
Lag 0-1: 1.05(1.02, 1.09)
With no secondary disease
Lag 0: 1.03(1.01, 1.04)
Lag 1: 1.03(1.01, 1.04)
Lag 0-1: 1.03(1.02, 1.05)
No copollutant models.
Ml 1.04(1.02,1.06)
Poloniecki et al. London, U.K. IHD,
(1997) Ml,
angina
Linn et al. (2000) Los Anqeles, CA Ml
Wonq etal. (1999) Honq Konq, China IHD
Ponka and Virtanen Helsinki, Finland IHD
(1996)
Barnett et al. (2006) 7 Australian and IHD,
New Zealand
cities Ml
Von Klot et al. (2005) 5 European cities Ml,
angina
Lag1: 1.00(0.98, 1.02)
Lag 1: 1.02(1.01, 1.03)
Lag 1: 1.01 (1.00, 1.03)
Lag 0: 1.02(1.00, 1.04)
Lag 0-1: 1.04(1.00, 1.08)
NO2
Lag 0: 1.17(0.96, 1.42)
Lag 1:0.95(0.77, 1.16)
NO
Lag 0: 1.01 (0.96, 1.05)
Lag 1: 1.10(1.05, 1.15)
Lag 0-1 , >65 yrs:
1.10(1.04, 1.17)
Lag 0-1, 15-64 yrs:
1.03(0.96, 1.10)
Lag 0-1, >65yrs:
1.18(1.04, 1.35)
Lag 0-1, 15-64 yrs:
1.07(0.96, 1.20)
LagO: 1.14(0.99, 1.32)
LagO: 1.16(1.03, 1.31)
Ages > 35 yr, Ml survivors
NO2: Robust to O;
Attenuated by CO, SO2, and
BS.
No copollutant models.
No copollutant models.
No copollutant models.
No copollutant models.
NO2: Robust to PM-io or Os
adjustment.
January 2015
5-261
DRAFT: Do Not Cite or Quote
-------
Table 5-46 (Continued): Corresponding risk estimates for hospital admissions for
ischemic heart disease for studies presented in
Figure 5-18.
Study
Location
Health
Effect
Relative Risk3 (95% Cl) Copollutant Examination13
tBhaskaran et al. England and
(2011) Wales
Lag 1-6 h: 1.06(1.02, 1.11) No copollutant models.
Lag 7-12 h: 0.95(0.90, 0.99)
Lag 13-18 h: NO2 correlations: PMm 0.48;
0.99(0.94, 1.05)
Lag 19-24h:
1.00(0.96(1.05)
Lag 0-2 days:
0.98(0.93, 1.02)
O3: -0.58; CO: 0.61; SO2:
0.31.
Atkinson et al. (1999) London, U.K. IHD
Lag 0; 0-64 yrs:
24.h(0.99, 1.04)
Lag 0; 65+ yrs:
1.03(1.01, 1.04)
No copollutant models
analyzed for IHD.
Peel et al. (2007) Atlanta, GA
IHD Lag 0-2; case-crossover:
1.04(1.00, 1.07)
Lag 0-2; time-series:
1.04(1.01, 1.08)
No copollutant models.
Jalaludin et al.
(2006)
Sydney, Australia IHD
Lag 0: 1.07(1.01, 1.13)
Lag 1: 1.02(0.97, 1.08)
Lag 0-1: 1.01 (0.99, 1.03)
Ages > 65 yr
No copollutant models
analyzed for IHD.
Simpson et al.
(2005a)
4 Australian cities IHD
Lag 0-1: 1.06(1.03, 1.09)
No copollutant models
analyzed for IHD.
Cl = confidence interval, CO = carbon monoxide, IHD = ischemic heart disease, Ml = myocardial infarction, NO = nitric oxide.
NO2 = nitrogen dioxide, O3 = ozone, PM = particulate matter, SO2 = sulfur dioxide, TSP = total suspended particles.
aEffect estimates are standardized to a 20-ppb or 30-ppb increase in NO2 or NO for 24-h avg and 1-h max metrics, respectively.
"Relevant relative risks for copollutant models can be found in Figure S5-2. S5-3. S5-4. and S5-5 (U.S. EPA. 2014b. c, d, e).
fStudies published since the 2008 ISA for Oxides of Nitrogen.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
A number of studies based on administrative data have also been published since the
2008 ISA for Oxides of Nitrogen. In six areas in central Italy, Nuvolone et al. (2011)
found an 8% (95% Cl: 0, 15) increase in risk of hospital admission for MI per 20-ppb
increase in 24-h avg NCh on the previous day. Similar associations were seen in relation
to lags 2 to 4 days prior to hospital admission. The finding at lag 2 was robust to
adjustment for PMio in a copollutant model, and remained positive, though somewhat
attenuated, by adjustment for CO (Figure S5-2. (U.S. EPA. 2014b) and Figure S5-3.
(U.S. EPA. 2014c)). The association with NO2 was somewhat more pronounced among
females and in the cold season. Using data from 14 hospitals in seven Canadian cities,
Stiebetal. (2009) found a 2.8% (95% Cl: 0.2, 5.4) increase in risk of ED visits for the
composite endpoint of acute MI or angina per 20-ppb increase in 24-h avg NO2 on the
same day. However, the overall association was heavily influenced by the association
observed in Edmonton, and exclusion of the data from Edmonton from the analysis
attenuated the results. Furthermore, the association observed from the data including
Edmonton was weakened in magnitude and precision (wider 95% Cl) in a copollutant
January 2015
5-262
DRAFT: Do Not Cite or Quote
-------
1 model adjusting for CO (1.3% [95% CI: -2.9, 5.6] increase per 20 ppb increase in
2 24 h avg NO2 on the same day). Larrieu et al. (2007) observed a positive association
3 between hospital admissions for IHD and NO2 concentrations in eight French cities. The
4 magnitude of the association was higher for older adults (i.e., >65 years) than for the
5 general population. In large single-city studies.Szyszkowicz (2007). Thach et al. (2010).
6 and Franck etal. (2014) found that NO2 was associated with increased risk of hospital
7 admission for IHD in Montreal, Canada; Hong Kong, China; and Santiago, Chile,
8 respectively. Qiu et al. (2013b) also reported an overall association between NO2
9 concentrations risk of ED visits for IHD in Hong Kong that was stronger in the cool
10 season and on low humidity days.
11 In New Jersey, Rich etal. (2010) found a relative risk of 1.14 (95% CI: 0.96, 1.32) per
12 20-ppb increase in 24-h avg NO2 for hospitalization for transmural Mis, but that
13 association was attenuated by adjustment for PM2 5 in a copollutant model (1.05 [95% CI:
14 0.85, 1.28]). No results were reported for all Mis or for nontransmural infarcts. NO2 was
15 positively associated with hospital admissions for MI in Taipei, Taiwan (Goggins et al..
16 2013; Tsai etal.. 2012; Hsieh etal.. 2010). Kaohsiung, Taiwan (Tsai et al.. 2012; Cheng
17 et al.. 2009). and Hong Kong (Goggins et al.. 2013). The associations reported by Hsieh
18 etal. (2010) remained relatively unchanged after adjustment for PMio, SO2, CO, or Os in
19 copollutant models, as did the results from Cheng et al. (2009). with the exception of CO
20 and Os on warm days. NO2 was also positively associated with hospital admissions for
21 IHD in Taipei, Taiwan (Bell etal.. 2008). In an effort to reduce uncertainty related to the
22 use of central site monitors, Bell et al. (2008) estimated NO2 exposure over the entire
23 Taipei area (average of 13 monitors), within Taipei City only (average of 5 monitors),
24 and using a subset of monitors where all pairs of monitors had NO2 correlations greater
25 than 0.75 (8 monitors). The authors reported consistent results across the multiple
26 exposure metrics, with the exception of stronger associations observed using the city or
27 correlated monitors at lag 0-3 (Table 5-46). Wichmann et al. (2012) found that NO2 was
28 positively associated with risk of acute MI hospital admissions in Copenhagen, Denmark,
29 but only in the warm months of the year. NO2 was not associated with risk of hospital
30 admission for acute coronary syndrome in Lithuania (Vencloviene et al.. 2011).
5.3.2.2 Hospital Admissions and Emergency Department Visits
for Angina Pectoris
31 The preceding epidemiologic evidence describing associations between short-term
32 increases in ambient NO2 concentrations and increased hospital admissions and ED visits
33 for MI and IHD is supported by evidence for increases in hospital admissions and ED
34 visits for angina. Angina pectoris results from an imbalance between the demand for
January 2015 5-263 DRAFT: Do Not Cite or Quote
-------
1 oxygen in the heart and the delivery by the coronary artery. Reduction in coronary blood
2 flow due to atherosclerosis is a common cause of this imbalance. Unstable angina, where
3 the coronary artery is not completely occluded, can lead to MI.
4 The 2008 ISA for Oxides of Nitrogen did not include specific discussion of angina but
5 did report results from two studies that examined associations between ambient NO2
6 concentrations and angina hospital admissions (U.S. EPA, 2008a). In a study of five
7 European cities, Von Klot et al. (2005) examined the relationship between short-term air
8 pollution and hospital readmissions of myocardial infarction survivors. The authors
9 reported a 16% (95% CI: 3, 31) increase in risk of hospital readmissions for angina
10 pectoris per 20-ppb increase in 24-h avg NCh on the same day. Poloniecki et al. (1997)
11 observed a smaller, but statistically significant association between NO2 concentrations
12 on the previous day and angina hospital admissions in London, UK (Table 5-46). Neither
13 study evaluated copollutant models.
14 More recent studies add to the limited, but consistent evidence of an association between
15 ambient NC>2 exposure and angina hospital admissions and ED visits. Szyszkowicz
16 (2009) found that NC>2 concentrations were associated with risk of ED visits for chest
17 pain in six Canadian cities. The magnitude of association was stronger in the warm
18 season, with a 5.9% increase in risk (95% CI: 3.3, 8.6) than in the cold season, with a
19 3.2% increase in risk (95% CI: 1.5, 5.0) at lag 1 per 20-ppb increase in 24-h avg NC>2. As
20 discussed in Section 5.3.2.1. Stieb et al. (2009) examined the composite endpoint of acute
21 MI or angina ED visits in a study of seven Canadian cities that included overlapping data
22 with Szyszkowicz (2009). Stieb et al. (2009) observed a positive association between
23 ambient NC>2 and Ml/angina that was still positive, but attenuated, imprecise, and no
24 longer statistically significant after adjustment for CO in a copollutant model. In addition
25 to limited interpretability from using a composite endpoint, the results were also largely
26 influenced by data from one city, as detailed in Section 5.3.2.1.
5.3.2.3 ST-Segment Amplitude
27 ST-segment changes (either ST-segment elevation or depression) on the
28 electrocardiogram are considered a nonspecific marker of myocardial ischemia. While
29 the 2008 ISA for Oxides of Nitrogen did not review any epidemiologic studies of ambient
30 oxides of nitrogen concentrations and markers of myocardial ischemia (U.S. EPA.
31 2008a). a few recent studies report associations (Table 5-47). Chuang et al. (2008)
32 conducted a repeated-measures study of Boston-area adults with a history of coronary
33 heart disease and examined the association between ambient pollutants and ST-segment
34 changes. The authors reported an OR of 3.29 (95% CI: 1.82, 5.92) for ST-segment
January 2015 5-264 DRAFT: Do Not Cite or Quote
-------
1
2
o
J
4
5
6
7
8
9
10
11
12
13
depression of >0.1 mm per 20-ppb increase in 24-h avg NO2 concentrations over the
previous 24 hours. This finding was robust to additional adjustment for PM2 5 in a
copollutant model (OR: 3.29 [95% CI: 1.65, 6.59]).
Delfino etal. (2011) used a similar design to study 38 older, nonsmoking adult residents
of four retirement homes in the Los Angeles area with a documented history of coronary
artery disease. A particular strength of this study is that the authors measured pollutant
concentrations outside of the residence, which improved spatial matching of NC>2
concentrations to subjects' locations. The authors observed an OR of 10.13 (95% CI:
1.37, 74.23) for ST-segment depression > 1.0 mm per 30-ppb increase in mean
1-hour NO2 concentrations preceding measurement over the previous 3 days. Other
averaging periods from 8 hours to 4 days gave similar or slightly weaker results. NO2 was
more strongly associated with ST depression than was NOx. No copollutant models were
evaluated.
Table 5-47 Epidemiologic studies of ST-segment amplitude.
Study
tChuanq et al.
(2008)
fDelfino etal.
(2011)
Location
Sample Size
Boston, MA
(n = 48)
Los Angeles,
CA
(n = 38)
Mean NO2
ppb
24-h avg NO2
21.4
75th: 24.9
Max: 44.5
1-h NO2: 27.5
1-h NOx: 46.6
Exposure
assessment
City-wide avg
Outdoor monitor
at retirement
community
Selected Effect Estimates3
ST segment change (mm):
12-h: -0.02 (-0.05, 0.00)
24-h: -0.08 (-0.12, -0.05)
RR for ST-segment depression
12-h: 1.15(0.72, 1.82)
24-h: 3.29 (95% CI: 1.82, 5.92
OR for ST-segment depression
NO2:
1-h: 1.33(0.83,2.11)
(95% CI)
>0.1 mm:
>1.0 mm
8-h: 2.37 (1.14, 4.92)
24-h: 4.75 (1.51, 14.84)
2-day: 7.51 (1.49, 37.87)
3-day: 10.13(1.37, 74.23)
4-day: 5.47 (0.65, 45.89)
NOx:
1-h: 1.25(0.96, 1.64)
8-h: 1.49(0.94,2.40)
24-h: 1.88(0.83,4.20)
2-day: 2.10 (0.67, 6.45)
3-day: 2.31 (0.45, 11.83)
4-day: 1.82(0.21, 15.84)
CI = confidence interval, NO2 = nitrogen dioxide, NOx = sum of NO and NO2, OR = odds ratio, RR = relative risk.
aEffect estimates are standardized to a 20-ppb or 30-ppb increase in NO2 or 40 ppb or 60-ppb increase in NOX concentration for
24-h avg and 1-h max metric, respectively.
fStudies published since the 2008 ISA for Oxides of Nitrogen.
January 2015
5-265
DRAFT: Do Not Cite or Quote
-------
5.3.2.4 Summary of Myocardial Infarction
1 In summary, the epidemiologic data available continue to support potential associations
2 between ambient NO2 concentrations and risk of triggering an MI. However, potential
3 copollutant confounding by traffic-related pollutants was not examined extensively in
4 these studies. In the studies that did analyze copollutant models to adjust for another
5 traffic pollutant, the findings were generally inconsistent. Associations between ambient
6 NO2 and risk of hospital admissions or ED visits for MI and IHD were attenuated by
7 adjustment for CO in two studies (Nuvolone et al.. 2011; Stieb et al.. 2009) but remained
8 robust in three others (Hsieh et al.. 2010; Cheng et al.. 2009; Yang. 2008). Additionally,
9 Rich et al. (2010) reported that an association between short-term NO2 exposure and
10 hospitalization for MI was attenuated by the inclusion of PM2 5 in a copollutant model.
11 There is limited, but consistent evidence of an association between NO2 and angina
12 pectoris (Stieb et al.. 2009; Szyszkowicz. 2009; Von Klotetal.. 2005; Poloniecki et al..
13 1997). However, only one study included a copollutant model, in which the association
14 was attenuated in magnitude and precision after adjustment for CO (Stieb etal.. 2009).
15 None of the reviewed studies of MI, IHD, or angina utilized copollutant models to adjust
16 for potential confounding by EC or VOCs. Additionally, all of the studies in this section
17 used central site monitors to measure ambient NO2, which have noted limitations in
18 capturing the variation in NO2 (Section 3.4.4.2).
19 In addition to hospital admission and ED visit studies, a few available epidemiologic
20 studies report an association between short-term exposure to NO2 and ST-segment
21 changes on the electrocardiogram of older adults with a history of coronary artery
22 disease, potentially indicating an association between NO2 and increased risk of
23 myocardial ischemia in this patient population. No studies from the previous ISA are
24 available for comparison. Once again, there was limited assessment of potential
25 confounding by traffic pollutants in copollutant models, though Chuang et al. (2008)
26 reported that the association between NO2 and ST-segment changes was robust to PM2 5
27 adjustment.
5.3.3 Arrhythmia and Cardiac Arrest
5.3.3.1 Panel Epidemiologic Studies
28 The 2008 ISA for Oxides of Nitrogen found little epidemiologic evidence of an
29 association between short-term changes in ambient NO2 concentrations and cardiac
January 2015 5-266 DRAFT: Do Not Cite or Quote
-------
1 arrhythmias (U.S. EPA. 2008a). There continues to be limited epidemiologic evidence for
2 such an association, either from panel studies of patients with ICDs or panel studies of
3 arrhythmias detected on ambulatory ECG recordings (Table 5-48).
4 In a study of patients with ICDs, Ljungman et al. (2008) found that NC>2 was positively
5 associated with increased risk of confirmed ventricular tachyarrhythmias (VT). The
6 association with PMio and PlVfc 5 was stronger than the association for NC>2. The authors
7 observed no evidence of effect modification by city, distance from the nearest ambient
8 monitor at the time of the event, number of events, type of event (ventricular fibrillation
9 vs. ventricular tachycardia), age, history of IHD, left ventricular ejection fraction,
10 diabetes, body mass index, or use of beta blockers. They did, however, report effect
11 modification depending on whether the patient was indoors or outdoors at the time of the
12 event with a strong association between NO2 and risk of VT among the 22 subjects that
13 were outdoors at the time of ICD activation. Because the authors accounted for personal
14 activity/behavior, exposure measurement error may have been reduced in the effect
15 modification analysis, as more time spent outdoors is likely to correspond to a greater
16 personal-ambient correlation (Section 3.4.4.1). In a similar study, Anderson et al. (2010)
17 observed generally null associations between ICD activation and ambient NO, NO2, or
18 NOx concentrations. Anderson et al. (2010) only had the study cardiologist review the
19 electrocardiograms from about 60% of ICD activations (confirming 87% of those cases
20 as VT), potentially leading to greater misclassification of the outcome than in the study
21 by Ljungman et al. (2008). Recently, Link et al. (2013) examined a panel of patients with
22 dual chamber ICDs. They observed positive associations between ICD-detected
23 arrhythmias and atrial fibrillations >30 seconds and NC>2 concentrations that were
24 generally stronger when the authors used a 2-hour lag compared to a 2-day lag. Finally,
25 Metzger et al. (2007) observed generally null associations between NO2 concentrations
26 and VT events over a 10-year period in Atlanta, GA.
27 Using a different approach, Bartell etal. (2013) used ECG monitors to evaluate VT
28 events in 50 older adult, non-smoking residents of four retirement communities in the
29 greater Los Angeles area. The study reported a 35% (95% CI: -1, 82) increase in the
30 daily rate of VT events per 40-ppb increase in 24-h avg NOx. The estimated effect of
31 3 - and 5 -day avg NOx on the daily rate of VT was somewhat stronger, though markedly
32 less precise (i.e., wider confidence limits around the effect estimates). Bartell etal. (2013)
33 measured pollutant concentrations outside of each of the retirement communities, which
34 improved spatial matching of NO2 concentrations to subjects' locations. Conversely,
35 Barclay et al. (2009) generally observed weak and inconsistent associations between NO2
36 or NO and incident arrhythmias detected on ambulatory ECG recordings in a repeated
37 measures study of nonsmoking adults with stable heart failure.
January 2015 5-267 DRAFT: Do Not Cite or Quote
-------
Table 5-48 Epidemiologic studies of arrhythmia and cardiac arrest.
Study
tUunqman et
al. (2008)
Location
Sample Size
Gothenburg and
Stockholm,
Sweden
n=211
(266 events)
Mean NO2 (ppb)
24-h avg NO2
Gothenburg: 11.8
Stockholm: 8.3
Exposure
Assessment
Single monitor
in Gothenburg,
average of
2 monitors in
Stockholm
Selected Effect Estimates
(95% Cl)a
Ventricular tachyarrhythmia (OR)
2-havg: 1.37(0.53, 3.64)
24-h avg: 1.26(0.49, 3.32)
tAnderson et London, U.K.
al. (2010) n = 705
(5,462 device
activations)
24-h avg NO2: 12.1
24-h avg NOx: 24.1
24-h avg NO: 19.4
City-wide avg ICD activations (OR)
NO2; Lag 0-1: 0.93(0.70, 1.24)
NOx; Lag 0-1: 0.92 (0.86, 1.08)
NO: Lag 0-1: 0.96(0.93, 1.04)
tLink et al.
(2013)
Boston, MA
n = 176(328atrial
fibrillation
episodes >30 sec)
24-h avg NO2: City-wide avg ICD-detected arrhythmias (OR)
16.1 24-h lag: 1.23(0.75, 2.10)
2-hlag: 1.57(0.97,2.47)
Metzqer et al.
(2007)
Atlanta, GA
n = 518
1-h maxNO2: 44.9
90th: 68
Max: 181
Central monitor All arrhythmia events (OR)
Allyr: 1.00(0.95, 1.05)
Warm season: 1.00(0.93, 1.08)
Cold season: 1.01 (0.94, 1.08)
Events resulting in cardiac pacing or
defibrillation:
Allyr: 1.01 (0.94, 1.10)
Events resulting in defibrillation:
Allyr: 1.07(0.93, 1.23)
Ventricular tachyarrhythmia (RR)
24-h avg: 1.35(0.99, 1.82)
3-day avg: 1.74(0.47,6.40)
5-day avg: 1.65(0.56,4.93)
tBartell et al.
(2013)
Los Angeles, CA
n = 50 (302
subject h of VT
observed)
24-h avg NOx: 42.3
Max: 183.7
Monitors on
trailers at each
of 4 retirement
communities
tBarclay et al. Aberdeen,
(2009) Scotland
n = 132
24-h avg NO2: Central monitor All arrhythmias (regression coefficients)
30.1 NO2: 3.193 (-3.600, 9.985)
NO: 14.7 NO: 3.524 (-3.059, 10.107)
Ventricular ectopic beats
NO2: 3.642 (-4.837, 12.121)
NO: 4.588 (-3.628, 12.803)
Ventricular couplets
NO2: 0.356 (-7.395, 8.106)
NO: -0.085 (-7.601, 7.431)
Ventricular runs
NO2: 2.443 (-2.537, 7.422)
NO: 2.219 (-2.618, 7.055)
Supraventricular ectopic beats
NO2: 2.888 (-4.833, 10.608)
NO: -2.688 (-10.170, 4.794)
Supraventricular couplets
NO2: 5.209 (-1.896, 12.313)
NO: 1.366 (-5.542, 8.274)
Supraventricular runs
NO2: 3.441 (-1.760,8.641)
NO: 2.298 (-2.753, 7.348)
Cl = confidence interval, ICD = implantable cardioverter defibrillators, NO = nitric oxide, NO2 = nitrogen dioxide, NOX = sum of NO
and NO2, OR = odds ratio, RR = relative risk, VT = ventricular tachyarrhythmias.
aEffect estimates are standardized to a 20-ppb or 30-ppb increase in NO2 or NO or 40 ppb or 60-ppb increase in NOX
concentration for 24-h avg and 1-h max metrics, respectively.
fStudies published since the 2008 ISA for Oxides of Nitrogen.
January 2015
5-268
DRAFT: Do Not Cite or Quote
-------
5.3.3.2 Out-of-Hospital Cardiac Arrest
1 The majority of out-of-hospital cardiac arrests are due to cardiac arrhythmias.
2 Dennekamp et al. (2010) observed generally positive, though weak, associations between
3 NO2 concentrations and risk of out-of-hospital cardiac arrest (Table 5-49). A similar
4 approach was used by Silverman et al. (2010) using data from out-of-hospital cardiac
5 arrests in New York City and observed generally null associations with NO2
6 concentrations in all year and cold season analyses, and a positive association in the
7 warm season analysis. More recently, Straneyetal. (2014) also reported null associations
8 between out-of-hospital cardiac arrest and ambient NO2 concentrations from a
9 case-crossover study in Perth, Australia. In two other studies of out-of-hospital cardiac
10 arrest, (Ensoretal.. 2013) found inconsistent and weak associations with ambient NO2
11 concentrations in Houston, while (Wichmann et al.. 2013) reported similarly inconsistent
12 associations with NOx in Copenhagen. However, (Wichmann et al.. 2013) observed a
13 positive association between ambient NOx concentration and out-of-hospital cardiac
14 arrest in females (46% [95% CI: 8%, 99] increase per 40-ppb increase in 24-h avg NOx at
15 lag 3), although there were slightly under two thirds the amount of cases observed in
16 females compared to males. None of the out-of-hospital cardiac arrest studies examined
17 potential copollutant confounding of NO2 or NOx associations. No studies from the
18 previous ISA are available for comparison.
January 2015 5-269 DRAFT: Do Not Cite or Quote
-------
Table 5-49 Epidemiologic studies of out-of-hospital cardiac arrest.
Study
tDennekamp et al.
(2010)
fSilverman et al.
(2010)
tStraney et al.
(2014)
Location
Sample Size
Melbourne,
Australia
n = 8,434
New York City,
NY
n = 8,216
Perth, Australia
(n = 8,551)
Mean NO2 (ppb)
24-h avg NCfe:
12.0
75th: 15.16
24-h avg NCfe:
50th: 27
75th: 32
95th: 43
1-h max NO2:
50th: 3.0
75th: 8.1
95th: 19.8
Exposure
Assessment
Central monitor
City-wide avg
Nearest monitor
(avg and/or max
distances not
specified)
Selected Effect Estimates3
(95% Cl)
% Change in out-of-hospital
cardiac arrest
LagO: 3.23 (-10.19, 18.51)
Lag 1: 7.69 (-7.29, 25.11)
Lag 2: -4.51 (-16.48, 10.56)
Lag 3: 7.37 (-7.11, 24.13)
Lag 0-1: 9.28 (-7.54, 29.14)
No quantitative results presented for
NO2.
OR
LagO-h: 1.008(0.992, 1.030)
Lag 1-h: 1.000(0.979, 1.021)
Lag2-h: 0.987(0.967, 1.004)
Lag3-h: 0.992(0.971, 1.013)
LagO-1-h: 1.004(0.987, 1.026)
LagO-3-h: 0.996(0.975, 1.017)
Lag 0-12-h: 0.996 (0.971, 1.026)
tEnsoretal. (2013) Houston, TX
(n = 11,677)
24-h avg NCb:
9.11
75th: 11.66
95th: 16.87
City-wide avg
% Change in out-of-hospital
cardiac arrests
LagO: 3.2 (-20.3, 18.9)
Lag 1: -2.5 (-14.7, 11.0)
Lag 2: -1.4 (-13.8, 12.6)
Lag 3: 3.2 (-9.6, 17.7)
Lag 4: 1.1 (-11.5, 15.3)
Lag 0-1: -0.4 (-14.4, 16.1)
Lag 1-2: -2.8 (-16.3, 12.9)
tWichmann et al.
(2013)
Copenhagen,
Denmark
(n = 4,657)
24-h avg NOx:
14.75
75th: 18.35
Central monitor
% Change in out-of-hospital
cardiac arrests
LagO: -13.5 (-28.6, 5.0)
Lag 1:5.4 (-12.7, 27.4)
Lag 2:-9.2 (-25.0, 11.6)
Lag 3: 16.0 (-4.0, 40.2)
Lag 4: 8.7 (-10.2, 31.2)
Lag 5: 5.4 (-13.1, 27.9)
Males, lag 3: 2.2 (-19.7, 30.1)
Females, lag 3: 46.1 (7.8, 98.7)
Cl = confidene interval, NO2 = nitrogen dioxide, NOX = sum of NO and NO2, OR = odds ratio.
aEffect estimates are standardized to a 20-ppb or 30-ppb increase in NO2 or NO or 40 ppb or 60-ppb increase in NOX
concentration for 24- h avg and 1-h max metrics, respectively.
fStudies published since the 2008 ISA for Oxides of Nitrogen.
January 2015
5-270
DRAFT: Do Not Cite or Quote
-------
5.3.3.3 Hospital Admissions and Emergency Department Visits
1 There are a limited number of studies examining associations between short-term NO2
2 exposure and hospital admissions with a primary discharge diagnosis related to
3 arrhythmias. Using data from 14 hospitals in seven Canadian cities, Stieb et al. (2009)
4 found no association between NO2 and risk of hospital admission for arrhythmias.
5 However, Tsai et al. (2009) reported a positive association in Taipei, Taiwan that was
6 stronger on cool days (OR: 1.34 [95% CI: 1.25, 1.44] per 20 ppb increase in 24-h avg
7 NO2) than warm days (OR: 1.19 [95%CI: 1.10, 1.28] per 20 ppb increase in 24-h avg
8 NO2). Both cool and warm day associations remained robust in copollutant models
9 controlling for PMio, SO2, CO, or Os; however, they did not evaluate potential
10 confounding by most of the traffic-related pollutants of concern.
5.3.3.4 Summary of Arrhythmia and Cardiac Arrest
11 In summary, there is currently inconsistent epidemiologic evidence for an association
12 between 24-h avg NO2, NO, or NOx and risk of cardiac arrhythmias as examined in
13 patients with ICDs, continuous ECG recordings, out-of-hospital cardiac arrest, and
14 hospital admissions. The reviewed studies rarely adjusted for copollutant confounding by
15 traffic pollutants, focused almost exclusively on ventricular arrhythmias, and are
16 potentially limited by misclassification of the outcome. Additionally, the majority of
17 studies used central site monitors to estimate ambient NO2 exposure, which have noted
18 limitations in capturing the variation in NO2.
5.3.4 Cerebrovascular Disease and Stroke
5.3.4.1 Hospital Admissions and Emergency Department Visits
19 The 2008 ISA for Oxides of Nitrogen found that the epidemiologic evidence for
20 associations between short-term changes in NO2 levels and hospital admissions or ED
21 visits for cerebrovascular diseases was generally inconsistent and provided little support
22 for an independent NO2 effect (U.S. EPA. 2008a). Recent studies published since the
23 2008 ISA also provide inconsistent evidence (Figure 5-19 and Table 5-50).
24 Generally, studies based on clinical registries are less susceptible to misclassification of
25 the outcome and exposure, which may explain the stronger evidence provided by these
January 2015 5-271 DRAFT: Do Not Cite or Quote
-------
1 studies than that based on administrative data. Wellenius et al. (2012) reviewed the
2 medical records of 1,705 Boston-area patients hospitalized with neurologist-confirmed
3 acute ischemic stroke and found an OR for ischemic stroke onset of 1.32 (95% CI: 1.08,
4 1.63) per 20-ppb increase in NC>2 concentration averaged over the 24 hours preceding
5 hospitalization for stroke. A unique strength of this study was the availability of
6 information on the date and time of stroke symptom onset in most patients, thereby
7 potentially reducing misclassification of the exposure. Copollutant models were not
8 evaluated.
9 Andersen et al. (2010) obtained data on strokes in Copenhagen, Denmark from the
10 Danish National Indicator Project and found a positive association between ambient NOx
11 concentrations and risk of ischemic stroke but not hemorrhagic stroke. The strongest
12 association was observed in relation to NOx levels 4 days earlier and for those suffering a
13 mild stroke, but the association was attenuated after adjustment for UFP. Using data from
14 a stroke registry in Como, Italy, Vidale et al. (2010) found that NO2 was associated with
15 risk of ischemic stroke hospital admission. On the other hand, Turin etal. (2012) did not
16 observe any association using data from the Takashima County Stroke and AMI Registry
17 in Central Japan. Similarly, Oudin et al. (2010) found no association between modeled
18 residential NOx concentration and risk of ischemic or hemorrhagic stroke within the
19 context of a Swedish quality register for stroke.
January 2015 5-272 DRAFT: Do Not Cite or Quote
-------
Study
Thach et al. (2010)
Zheng et al. (2013)
Bell et al. (2008)
Xiang etal. (2013)
ViUen^uve ct al. (2012)
Turin etal. (2012)
Wai Oftl ">\
cLJctlltls Cl 91. (.ul_)
Amturscn el si. (2010)
Larrieu et al. (2007)
Pokraiecki et al. (1997)
Linn et al. (2000)
Tcai *>t al OAfV^I
1 adl cl Al. i. ^ ' ''•.'. • J
Wong etal. (1999)
Vffleneuve et al (2006)
WeHenhis et al. (2005)
Prtnl-a { 1 OOfil
r onxa ^ 1 yyv )
BaJester et al. (2001)
Peel et al. (2007)
Chen etal. (2014)
Jalahidin et al. (2006)
Mean
Outcome Concentration
Stroke 31.2
CBV 28.19
CBV 26.4
Stroke 24.36
Sffvil-p, 1 A T
UQpiC ID. /
l&cnenuc Stroke
Heiiiorrhamc Stroke
Transient Tschernic Sfroke
Stroke .
Ischctinc Stroke
HeniQirhagic Stroke
Inuisient Tsctieniic Stroke
Stroke 16
Cerebral Infarction
IiUracerebra] Heni
S ubaj"3cLuioid riL'in
Isclicinic Stroke
Isclienitc Stroke l?-3
\ • f i
rleniorrnagic Stroke
. »-, , , ,
Ivlllu ISCucTiilC
Stroke 11.9-24.9
CBV 35
CBV 28-41
Occ Stroke
r>rp1"»ral ^rrnVp ^S I 7
LcicUldl oirOKc _o.l t
Iscliennc Stroke
Cerebral Stroke
Isch^niic Stroke
CBV 27.3
Ischemic Stroke 24
Hemorrhagic Stroke
Cerebral Stroke
Ischemic Stroke 23.54
Hemorrhagic Stroke
PRV ">fi 7
^ D V .iU. /
CBV 61.8
CBV 45.9
Ischemic Stroke 23.68
Stroke 23.2
Lag
0-1
k
0
0-3
I
0-3
0
0-2
8-2
07
-Z
n "•
n "
1 '-„
0-)
-t.
r\ ~
0-2
!
1
1 A lir
•.•4 iir
0" ,
,". i
• .'- f
-
- 1
i
0
0
A.7
o3
0. -j
^A
0-1
0
0
\
0
0
1
1
2
§:i
11
I -r>
1-8
0
0-1
Notes
all monitors (13)
city monitors (5)
correlated monitors (8)
warm season
cool season
warm season
«
cool season
all aaes
2 65>s
^A-f C"
iU^^ V_-
<2v L- •
65+yrs
••
• .
•
case-crossover
tune series
all year
\vanti season
cold season
>65yrs
1
b
|t
>
f
<-
j^
l^1
1 »
3
52
™ •
i ^
-^
— »|^-
• ^
•
• •
— »t — *
_ i *
1 A
• ™
; -
}
i
I — *
• ^
!•
^^
+• —
~f9
_^j
|?
,
I
A
-1
^t
24-h
• ».
.
*
l-h
0.5 1 1.5
Relative Risk (95% CI)
2.5
Note: Studies in red are recent studies. Studies in black were included in the 2008 ISA for Oxides of Nitrogen. Relative risks are
standardized to a 20-ppb or 30-ppb increase in NO2 concentration and 40 ppb or 60 ppb for NOX concentrations for 24-h avg and
1-h max metrics, respectively. Studies are presented in descending order, with the largest mean concentration (ppb) at the top and
the smallest at the bottom of the figure (by averaging time and inclusion in previous ISA). Circles = NO2; Diamonds = NOX.
Figure 5-19 Results of studies of short-term exposure to oxides of nitrogen
and hospital admissions for cerebrovascular disease and stroke.
January 2015
5-273
DRAFT: Do Not Cite or Quote
-------
Table 5-50 Corresponding risk estimates for hospital admissions for
cerebrovascular disease and stroke for studies presented in
Figure 5-19.
Study
fThach et al.
(2010)
tZhenq et al.
(2013)
Location
Hong Kong,
China
Lanzhou,
China
Health Effect
Stroke
Cerebrovascular
disease
Selected Relative Risks3
95% Cl
Lag 0-1: 1.01 (1.00, 1.03)
LagO: 1.05(1.02, 1.08)
Lag 0-3: 1.06(1.02, 1.10)
Copollutant Examination13
No copollutant models.
NO2: associations were
robust to adjustment for
SO2, associations increased
with adjustment for PMm
Copollutants: SO2 (positive)
associations and PM-io
(negative) associations
robust to adjustment for
NO2.
NO2 correlations
(Spearman r): PMm 0.64;
SO2: 0.64.
tBell et al. (2008) Taipei, Cerebrovascular All monitors:
Taiwan disease Lag 0: 1.01 (0.95, 1.07)
Lag 1: 0.97(0.92, 1.03)
Lag 0-3: 1.04(0.96, 1.12)
City monitors:
LagO: 1.01 (0.96, 1.06)
Lag 1: 0.97(0.92, 1.02)
Lag 0-3: 1.04(0.96, 1.12)
Correlated monitors:
LagO: 1.01 (0.96, 1.06)
Lag 1: 0.97(0.92, 1.02)
Lag 0-3: 1.04(0.96, 1.12)
tXianq et al. (2013) Wuhan, Stroke Warm season:
China LagO: 0.99(0.93, 1.06)
Lag 0-2: 0.96(0.88, 1.05)
Cool season:
LagO: 1.11 (1.05, 1.18)
Lag 0-2: 1.12(1.04, 1.21)
No copollutant models.
NO2: cold season
association robust to PM-io
adjustment.
Copollutants: PM-io no
longer associated with
stroke hospital admissions
in the cold season after NO2
adjustment.
No correlations provided.
January 2015
5-274
DRAFT: Do Not Cite or Quote
-------
Table 5-50 (Continued): Corresponding risk estimates for hospital admissions for
cerebrovascular disease and stroke for studies
presented in Figure 5-19.
Study
Location
Health Effect
Selected Relative Risks3
95% Cl
Copollutant Examination13
tVilleneuve et al.
(2012)
Edmonton,
Canada
Stroke,
ischemic stroke,
hemorrhagic stroke,
transient ischemic
stroke
Stroke (Lag 0-2):
Warm: 1.42(0.94,2.17)
Cool: 0.98(0.84, 1.15)
Ischemic stroke (Lag 0-2):
Warm: 2.37 (1.27, 4.42)
Cool: 0.90(0.69, 1.13)
Hemorrhagic stroke
(Lag 0-2):
Warm: 1.50(0.59,4.32)
Cool: 0.98(0.64, 1.50)
Transient ischemic stroke
(Lag 0-2):
Warm: 0.67(0.33, 1.42)
Cool: 1.11 (0.80, 1.45)
Ischemic stroke during
warm season
NO2: associations robust to
adjustment for 862; slightly
attenuated but positive after
adjustment for CO, Os, or
PM2.5.
Copollutants: CO, Os, and
PM2.5 associations
attenuated by adjustment
for NO2. No association
between SO2 and ischemic
stroke.
Hemorrhagic stroke during
warm season
NO2: associations
attenuated after adjustment
forSO2and Os, but
increased after adjustment
for CO or PlVh.5.
Copollutants: SO2 and Os
associations robust to NO2
adjustment; CO no longer
associated with
hemorrhagic stroke after
NO2 adjustment.
tTurin et al. (2012) Takashima
County,
Japan
fWellenius et al. Boston, MA
(2012)
t Andersen et al. Copenhagen,
(2010) Denmark
tLarrieu et al. 8 French
(2007) cities
Stroke,
cerebral infarction,
intracerebral
hemorrhage,
subarachnoid
hemorrhage
Ischemic stroke
Ischemic stroke,
hemorrhagic stroke,
mild ischemic
stroke
Stroke
Stroke:
LagO: 0.98(0.89, 1.08)
Lag 1: 0.98(0.89, 1.08)
Cerebral infarction:
LagO: 0.98(0.87, 1.10)
Lag 1: 1.00(0.89, 1.12)
Intracerebral hemorrhage:
LagO: 1.06(0.85, 1.33)
Lag 1: 0.94(0.75, 1.16)
Subarachnoid hemorrhage:
LagO: 1.12(0.80, 1.56)
Lag 1: 1.02(0.73, 1.42)
Lag 24 h preceding event:
1.32(1.08, 1.63)
NOx:
Ischemic stroke:
LagO: 1.20(0.96, 1.48)
Lag 1: 0.96(0.79, 1.20)
Lag 0-4: 1.36(0.96, 1.89)
Hemorrhagic stroke:
LagO: 0.96(0.48, 1.81)
Lag 1: 1.14(0.59,2.20)
Lag 0-4: 0.43(0.13, 1.36)
Mild ischemic stroke :
Lag 0-4: 1.61 (0.79,3.30)
All ages: 0.99 (0.96, 1.03)
>65yrs: 1.01 (0.97, 1.05)
No evidence of an
association between NO2
and stroke. Copollutant
models did not change the
results.
No copollutant models.
NOx: no longer associated
with ischemic stroke after
adjustment for UFPs.
Copollutants: UFP
association robust after
adjustment for NOx.
No copollutant models.
January 2015
5-275
DRAFT: Do Not Cite or Quote
-------
Table 5-50 (Continued): Corresponding risk estimates for hospital admissions for
cerebrovascular disease and stroke for studies
presented in Figure 5-19.
Study
Poloniecki et al.
(1997)
Linnetal. (2000)
Tsai et al. (2003)
Wonqetal. (1999)
Villeneuve et al.
(2006a)
Location
London, U.K.
Los Angeles,
CA
Kaohsiung,
Taiwan
Hong Kong,
China
Edmonton,
Canada
Health Effect
Cerebrovascular
disease
Cerebrovascular
disease,
occlusive stroke
Cerebral stroke,
ischemic stroke
Cerebrovascular
disease
Ischemic stroke,
hemorrhagic stroke,
Selected Relative Risks3
95% Cl
Lag 1: 0.99(0.98, 1.00)
Cerebrovascular disease:
LagO: 1.01 (0.99, 1.02)
Occlusive stroke:
LagO: 1.04(1.02, 1.06)
Cerebral stroke:
Lag 0-2; 20+°C:
1.68(1.38,2.04)
Lag 0-2; <20°C:
0.78(0.44, 1.37)
Ischemic stroke:
Lag 0-2; 20+°C:
1.67(1.48, 1.87)
Lag 0-2; <20°C:
1.19(0.78, 1.84)
Lag 0-1: 1.03(0.99, 1.07)
Ages 65 yr and older
Ischemic stroke:
Copollutant Examination13
No copollutant models
examined.
No copollutant models.
NO2 correlations: PM-io:
0.67 to 0.88; O3: -0.23 to
0.35; CO: 0.84 to 0.94
NO2: Ischemic stroke and
hemorrhagic stroke
associations robust to SO2,
CO, or Os adjustment.
Attenuated, but positive
after PM-io adjustment.
Copollutants: PM-io, SO2,
CO, and Os ischemic stroke
and hemorrhagic stroke
associations attenuated by
adjustment for NO2.
No copollutant models.
Ischemic stroke during
warm season
cerebral stroke
LagO: 1.04(0.96, 1.15)
Lag 1: 1.07(0.97, 1.17)
Hemorrhagic stroke:
LagO: 1.07(0.96, 1.21)
Lag 1: 1.06(0.94, 1.20)
Cerebral stroke:
LagO: 0.99(0.90, 1.07)
Lag 1: 0.91 (0.84, 1.00)
NO2: warm season
associations robust to
adjustment for SO2 or CO;
increase with adjustment for
PM-io or PlVh.s; attenuated
with adjustment for Os.
Hemorrhagic stroke during
warm season
NO2: warm season
associations robust to SO2,
Os, PM2.5, orPM-io
adjustment (large increases
in CIs in models with PM);
but attenuated with
adjustment for CO.
NO2 warm season
correlations (Pearson r):
SO2: 0.22; O3: -0.09; CO:
0.59; PIvh.s: 0.52; PMm
0.57.
Wellenius et al.
(2005)
Ponka and
Virtanen(1996)
Ballester et al.
(2001)
Peel et al. (2007)
9 U.S. cities
Helsinki,
Finland
Valencia,
Spain
Atlanta, GA
Ischemic stroke,
hemorrhagic stroke
Cerebrovascular
disease
Cerebrovascular
disease
Cerebrovascular
disease
LagO: 1.05(1.03, 1.07)
LagO: 1.01 (0.96, 1.06)
LagO: 0.96(0.87, 1.07)
Lag 1: 0.98(0.87, 1.09)
Lag 2: 1.22(1.04, 1.44)
Lag 0-2; case-crossover:
1.05(1.01, 1.09)
Lag 0-2; time-series: 1.06
(1.02, 1.11)
No copollutant models.
No copollutant models.
NO2: associations were
robust to adjustment for
SO2 or BS.
No copollutant models.
January 2015
5-276
DRAFT: Do Not Cite or Quote
-------
Table 5-50 (Continued): Corresponding risk estimates for hospital admissions for
cerebrovascular disease and stroke for studies
presented in Figure 5-19.
Study
Location
Health Effect
Selected Relative Risks3
95% Cl
Copollutant Examination13
tChen et al. (2014) Edmonton,
Canada
Ischemic stroke All Year: No copollutant models.
Lag 1-8 h: 1.06(0.98, 1.14)
Warm Season:
Lag 1-8 h: 1.39(1.13, 1.71)
Cold Season:
Lag 1-8 h: 0.98(0.89, 1.08)
Jalaludin et al.
(2006)
Sydney,
Australia
Stroke
Ages 65 yr and older
LagO: 0.95(0.88, 1.02)
Lag 1: 0.96(0.90, 1.03)
Lag 0-1: 0.95(0.88, 1.02)
No copollutant models
analyzed for
cerebrovascular disease.
Cl = confidence interval, CO = carbon monoxide, NO2 = nitrogen dioxide, NOX = sum of NO and NO2,03 = ozone, PM = particulate
matter, SO2 = sulfur dioxide, UFP = ultrafine particles.
aEffect estimates are standardized to a 20-ppb or 30-ppb increase in NO2 or 40 ppb or 60-ppb increase in NOX concentration for
24-h avg and 1-h max metrics, respectively.
"Relevant relative risks for copollutant models can be found in Figure S5-2, S5-3, S5-4, and S5-5 (U.S. EPA, 2014b, c, d, e).
fStudies published since the 2008 ISA for Oxides of Nitrogen.
1
2
o
J
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
Additional studies based on administrative data are also available. A number of the
administrative data studies were conducted in Edmonton, Canada and used similar or
identical data sources. The most thorough Edmonton study observed an association
between NO2 and ED visits for ischemic stroke in the warm season (OR: 2.37 [95% Cl:
1.27, 4.41] per 20-ppb increase in 3 day average NO2) and a weak association between
hemorrhagic stroke in the warm season (OR: 1.50 [95% Cl: 0.59, 3.82] per 20-ppb
increase in 3 day average NO2) (Villeneuve et al.. 2012). Villeneuve et al. (2012) also
examined copollutant models in which ischemic stroke associations were robust to
adjustment for SO2 (OR: 2.34 [95% Cl: 1.25, 4.37]), and remained positive, but slightly
attenuated after adjustment for CO (OR: 2.05 [95% Cl: 0.92, 4.57]), O3 (OR: 1.92 [95%
Cl: 0.98, 3.78]), and PM25 (OR: 1.98 [95% Cl: 0.94, 4.20]). The hemorrhagic stroke
associations were attenuated after adjustment for SO2 or Os but were robust to adjustment
for the traffic-related pollutant CO or PM2s. There is the potential for misclassification of
the exposure due to differences in the timing of stroke symptoms and the corresponding
ED visit; however, after surveying a subset of the study population, Villeneuve et al.
(2012) observed that roughly 75% of patients visited the emergency room on the same
day that their symptoms presented. Further, when the authors adjusted the assigned
pollution levels from the day of the ED visit to the day of symptom presentation, they
observed no systematic differences in assigned pollution levels. Szvszkowicz (2008)
observed a positive association between 24-h avg NO2 and ED visits for ischemic stroke
in Edmonton, Canada but only within specific subgroups according to sex, season, and
age. In a recent related study, Chenet al. (2014) used the same data, but applied hourly
NO2 values to their models. The authors reported an association between NO2 and ED
January 2015
5-277
DRAFT: Do Not Cite or Quote
-------
1 visits for acute ischemic stroke that remained relatively consistent across lag days.
2 However, that association was almost entirely influenced by the association observed in
3 the warm season, as the association in the cold season was null.
4 Zheng etal. (2013) conducted a time-series study in Lanzhou, China and found a positive
5 association between NO2 and all cerebrovascular hospital admissions. The strongest
6 relationships were observed on same-day and 3-day cumulative lags. Zheng etal. (2013)
7 also reported stronger associations in women and the elderly. Xiang etal. (2013)
8 observed a positive association between NO2 and hospital admissions for all strokes in
9 the cold season in Wuhan, China that was robust in a copollutant model including PMio.
10 Conversely, in Taipei, Taiwan, Bell et al. (2008) did not observe an association between
11 NO2 and cerebrovascular disease. As mentioned in Section 5.3.2.1. Bell et al. (2008)
12 attempted to reduce uncertainty related to the use of central site monitors by estimating
13 NC>2 exposure over the entire Taipei area (average of 13 monitors), within Taipei City
14 only (average of 5 monitors), and using a subset of monitors where all pairs of monitors
15 had NC>2 correlations greater than 0.75 (8 monitors). The null findings were consistent
16 across the three exposure assignment techniques. In a 7-year study of Hong Kong
17 residents, Thach etal. (2010) also reported no association between NO2 and all
18 cerebrovascular hospital admissions.
5.3.4.2 Summary of Cerebrovascular Disease and Stroke
19 In summary, the epidemiologic data provide generally inconsistent evidence for a
20 potential association between ambient NO2 concentrations and risk of hospital admission
21 for cerebrovascular disease and stroke. Clinical registry studies reported both positive
22 (Wellenius etal.. 2012: Andersen et al.. 2010: Vidale etal.. 2010) and null (Turin et al..
23 2012: Oudin etal.. 2010) associations, while the majority of administrative database
24 evidence of an association came from studies using similar or identical data sets (Chen et
25 al.. 2014: Villeneuve et al.. 2012: Szyszkowicz. 2008). There were a limited number of
26 studies that evaluated potential confounding by traffic pollutants (UFP, CO, PIVb 5, BS,
27 and PMio), and the results were, again, inconsistent. Additionally, the majority of the
28 studies in this section used central site monitors to estimate ambient NC>2 exposure, which
29 have noted limitations in capturing the variation in NC>2 (Section 3.4.4.2).
5.3.5 Decompensation of Heart Failure
30 Two recent studies found associations between short-term increases in ambient NC>2
31 concentration and hospital admissions or ED visits for heart failure. In the study of seven
January 2015 5-278 DRAFT: Do Not Cite or Quote
-------
1 Canadian cities described above, Stieb et al. (2009) observed a 5.1% (95% CI: 1.3, 9.2)
2 increase in risk of ED visits for heart failure per 20-ppb increase in 24-h avg NO2. Unlike
3 the results for the composite endpoint of MI or acute angina, the increased risk of ED
4 visits for heart failure was not dominated by results from a single city. In Taipei, Taiwan,
5 Yang (2008) found that risk of hospital admission for heart failure were associated with
6 NO2 concentrations but only on days where the mean ambient temperature was >20°C.
7 The association on warm days remained relatively unchanged after copollutant
8 adjustment for PMio, SO2, CO, or O3.
5.3.6 Increased Blood Pressure and Hypertension
5.3.6.1 Epidemiologic Studies
9 The 2008 ISA for Oxides of Nitrogen did not review any epidemiologic studies of
10 ambient oxides of nitrogen and blood pressure (BP) (U.S. EPA. 2008a). Several studies
11 are now available for review (Table 5-51). There is little evidence from longitudinal
12 studies of the association between NO2 and BP. A number of longitudinal studies
13 measured BP in subjects in Beijing before, during, and after the 2008 Beijing Olympics
14 when city-wide air pollution control measures substantially reduced ambient levels of
15 most criteria pollutants. One study reported that NO2 concentrations during the Olympics
16 were reduced by close to 22% versus the previous month and 13% versus the same period
17 the previous summer (Huang et al.. 2012a). Other ambient pollutants (except Os) were
18 reduced by similar or larger amounts. Huang etal. (2012a) measured BP repeatedly in
19 participants with pre-existing cardiovascular disease in Beijing and found no association
20 between NO2 and either systolic or diastolic BP. Focusing on healthy young adults,
21 Zhang etal. (2013) and Rich etal. (2012) each observed no clear association between
22 NO2 and either systolic or diastolic BP among participants assessed before, during, and
23 after the 2008 Beijing Olympics.
January 2015 5-279 DRAFT: Do Not Cite or Quote
-------
Table 5-51 Epidemiologic studies of blood pressure.
Study
fWilliams et al.
(2012a)
tHuanq et al.
(2012a)
tRich etal. (2012)
and tZhanq et al.
(2013)
tLiuetal. (2014b)
fCakmak et al.
(2011 a)
tChuanq et al.
(2010)
tChen etal. (2012c)
Location
Sample
Size
Detroit, Ml
n=65
Beijing,
China
n=40
Beijing,
China
n = 125
Sault Ste.
Marie.
Ontario,
Canada
n=61
Canada
n = 5,604
Taiwan
n = 7,578
Taiwan
Mean NO2 (ppb)
24-h avg NO2: 24.0
75th: 28.0
Max: 100.0
2007, Visit 1: 33.8
2007, Visit 2: 26.3
2008, Visit 3: 29.2
2008, Visit 4: 22.9
24-h avg NCb:
Entire study: 27.0
Before: 26.0
During: 13.9
After: 41. 4
1-h max NO2:
Site 1: 3.9
95th: 9.5
Site 2: 5.8
95th: 13.8
24-h avg NCb:
12.6
24-h avg NCb:
22.4
Max: 65.5
24-h avg NCb:
Exposure
Assessment
Personal
monitoring
and central
monitor
Central
monitor
Central
monitor
Central
monitor
City-wide avg
Nearest
monitor
(within 10 km)
City-wide avg
Selected Effect Estimates3 (95% Cl)
No quantitative results presented.
Change in SBP (mmHg)
30-min: 3.73 (-1.04, 8.28)
2-h: 0.00 (-5.89, 5.89)
12-h: 1.69 (-7.73, 11.11)
24-h: -2.32 (-19.10, 14.47)
Change in DBP (mmHg)
30-min: 2.28 (-1.86, 6.221)
2-h: -0.19 (-4.45, 6.00)
12-h: 2.90 (-5.07, 10.87)
24-h: 4.34 (-9.84, 18.52)
No quantitative results presented; results
presented graphically. Generally
inconsistent results with SBP: positive and
negative associations across lags.
Generally null and inconsistent
associations with DBP across lags 0-6.
Change in SBP (mmHg)
Lag 0: -1.04 (-4.20, 2.12)
Lag 1:2.08 (-1.36, 5.52)
Change in DBP (mmHg)
Lag 0: -1.32 (-3. 88, 1.28)
Lag 1: 1.64 (-1.36, 4.60)
Change in resting SBP (mmHg)
LagO: 1.76(0.35, 3.17)
Change in resting DBP (mmHg)
Lag 0:2.11 (1.12, 3.10)
No quantitative results presented for NO2.
Change in SBP (mmHg)
(n = 9,238)
13.9 to 26.1 across
locations
Max: 34.3 to 49.1
Lag 0: -0.81 (-2.16, 0.55)
Lag 0-1: -1.17 (-2.34,-0.01)
Lag 0-2: -4.20 (-5.22,-3.17)
Change in DBP (mmHg)
LagO: 1.03(0.11, 1.95)
Lag 0-1: 1.54(0.75,2.32)
Lag 0-2: -0.01 (-0.71, 0.68)
Pulse Pressure Change
Lag 0:-2.55 (-3.62, -1.48)
Lag 0-1: -2.09 (-3.02, -1.18)
Lag 0-2: -3.22 (-4.04, -2.40)
January 2015
5-280
DRAFT: Do Not Cite or Quote
-------
Table 5-51 (Continued): Epidemiologic studies of blood pressure.
Study
Location
Sample
Size
Mean NO2 (ppb)
Exposure
Assessment
Selected Effect Estimates3 (95% Cl)
tChoi et al. (2007) Incheon, 24-h avg NCb: City wide avg Warm season
South Korea Warm season: Change in SBP (mmHg)
(n = 10,459) 22.5 Lag 0: 2.24 (p = 0.002)
75th: 26.9 Lag 1: 2.40 (p < 0.001)
Max: 49.3 Lag 2: -0.04 (p = 0.534)
Cool season: 29.2 Change in DBP (mmHg)
75th: 34.7 Lag 0: 2.02 (p = 0.645)
Max: 74.0 Lag 1: 2.12 (p = 0.016)
Lag 2: -0.04 (p = 0.331)
Cool season
Change in SBP (mmHg)
Lag 0: 2.06 (p = 0.181)
Lag 1: 2.06 (p = 0.195)
Lag 2: -0.06 (p = 0.223)
Change in DBP (mmHg)
Lag 0: -0.02 (p = 0.573)
Lag 1: 2.00 (p = 0.445)
Lag 2: 2.02 (p = 0.445)
Cl = confidence interval, DBP = diastolic blood pressure, NO2 = nitrogen dioxide, SBP = systolic blood pressure.
aEffect estimates are standardized to a 20-ppb or 30-ppb increase in NO2 for 24-h avg and 1-h max metrics, respectively.
fStudies published since the 2008 ISA for Oxides of Nitrogen.
1 In the Detroit area, Williams et al. (2012a) measured BP up to 10 times in each of
2 65 adult participants and found no association between BP and either total personal or
3 ambient NC>2 concentrations. A strength of this study was the authors' use of personal
4 exposure measurements, which are less susceptible to exposure misclassification due to
5 the variability in NC>2 concentrations and variation in time-activity patterns than central
6 site monitoring (Section 3.4.4). Similarly, in a randomized cross-over study designed to
7 examine the cardiovascular effects of exposure to steel plant emissions in Ontario, Liu et
8 al. (2014b) measured NO2 exposure near subjects' randomized exposure location and
9 reported no association between NO2 and either systolic or diastolic BP.
10 Results of cross-sectional studies of the association between NO2 and BP measured on
11 the same day or with the NO2 measurement lagged 1-3 days before the BP measurement
12 have also been mixed. Cakmak et al. (201 la) used cross-sectional data from a national
13 population-based survey of children and adults in Canada and found a 1.76 mmHg (95%
14 Cl: 0.35, 3.17 mmHg) increase in systolic BP and a 2.11 mmHg (95% Cl: 1.12, 3.10
15 mmHg) increase in diastolic BP per 20-ppb increase in 24-h avg NC>2 on the same day.
16 Chuang et al. (2010) used cross-sectional data from a national population-based health
17 screening of adults in Taiwan and reported finding no association between BP and NC>2
18 levels, although quantitative results were not presented. On the other hand, Chen et al.
19 (2012c) used cross-sectional data from a different population-based health screening in
20 adults across six townships in Taiwan and found a 4.20 mmHg decrease (95% Cl: -5.22,
21 -3.17 mmHg) in systolic BP per 20-ppb increase in 24-h avg NO2 at lag 3 and a
January 2015 5-281 DRAFT: Do Not Cite or Quote
-------
1 1.54 mmHg increase (95% CI: 0.75, 2.32 mmHg) in diastolic BP per 20-ppb increase in
2 24-h avg NO2 at lag 2. Choi et al. (2007) observed positive associations between NO2
3 concentrations and systolic BP during the warm and cold seasons at lags 0 and 1, though
4 the associations for diastolic BP were generally null.
5.3.6.2 Controlled Human Exposure Studies
5 The 2008 ISA for Oxides of Nitrogen (U.S. EPA. 2008a) reviewed controlled human
6 studies of cardiac output and BP (Table 5-56); several of these studies also examined
7 heart rate (HR) as described in Section 5.3.11.1. NCh exposure generally did not increase
8 cardiac output or BP in healthy adults or those with COPD. These endpoints have not
9 been evaluated in recent controlled human exposure studies of NC>2.
10 Cardiac output is the volume of blood pumped out by each of the two ventricles per
11 minute. It is directly related to HR, as the output of each ventricle is the product of the
12 HR (beats/minute) and the stroke volume (mL of blood/beat). BP is the product of
13 cardiac output and vascular resistance. Cardiac output, vascular resistance, and BP
14 interact moment-to-moment to ensure systemic circulatory demands are met.
15 Folinsbee et al. (1978) exposed three groups of five young healthy adult males to 600_ppb
16 NC>2 for 2_hours with intermittent exercise. The authors reported no changes in cardiac
17 output or BP. Drechsler-Parks (1995) exposed eight older healthy adults to filtered air,
18 600_ppb NC>2, 450 ppb Os, and NC>2 + Os for 2_hours with intermittent exercise. There
19 was no change in stroke volume or cardiac output following exposure to NC>2 or Os alone
20 compared to filtered air; however, a decrease in cardiac output was observed following
21 NO2 + Os exposure compared to Os and filtered air exposures (p < 0.05). Gong etal.
22 (2005) reported no change in BP after exposure to 400_ppb NC>2 for 2_hours with
23 intermittent exercise in volunteers with COPD and healthy volunteers. One controlled
24 human exposure study examined exposure to higher concentrations of NO2. Linn et al.
25 (1985b) reported a small, but statistically significant decrease in BP after exposure to
26 approximately 4,000_ppb NO2 for 75_minutes with exercise. In both healthy volunteers
27 and those with asthma, the mean BP decrease was about 5_mmHg relative to controls.
5.3.6.3 Hospital Admissions and Emergency Department Visits
28 In contrast with findings for changes in BP, the limited number of available studies report
29 associations between NO2 and ED visits for hypertension. In Beijing, China, Guo et al.
30 (2010) found that NO2 was associated with ED visits for hypertension, and the
31 association remained relatively unchanged in copollutant models adjusting for PMio or
January 2015 5-282 DRAFT: Do Not Cite or Quote
-------
1 SCh. Similarly, in Edmonton, Canada, Szyszkowicz et al. (2012) found that ED visits for
2 hypertension were positively associated with NC>2 in single-pollutant models. The
3 association was attenuated in a multipollutant model adjusting for 862 and PMio, but
4 results from these models are difficult to interpret given the potential for multicollinearity
5 among pollutants.
5.3.6.4 Summary of Blood Pressure and Hypertension
6 In summary, there is little evidence from available epidemiologic studies to suggest that
7 short-term exposure to ambient NC>2 is associated with increased BP in the population
8 overall. There is no evidence of an association from longitudinal studies and mixed
9 evidence from cross-sectional studies. However, cross-sectional studies do not assess
10 temporal relationships and are more prone to confounding by factors that differ between
11 individual participants. Controlled human exposure studies show no evidence to suggest
12 that short-term exposure to ambient relevant concentrations of NC>2 alone alter BP or
13 cardiac output.
5.3.7 Venous Thromboembolism
14 Venous thromboembolism is a term that includes both deep vein thrombosis (DVT) and
15 pulmonary embolism (PE). DVT occurs when a blood clot develops in the deep veins,
16 most commonly in the lower extremities. A part of the clot can break off and travel to the
17 lungs, causing a PE, which can be life threatening.
18 Two recent studies found associations between NO2 and venous thrombosis and/or PE;
19 however, both studies were small, and neither evaluated potential copollutant
20 confounding. A study covering the metropolitan region of Santiago, Chile, found a 9.7%
21 (95% CI: 4.1, 15.4) and 8.4% (95% CI: 5.0, 11.8) increase in hospital admissions for
22 venous thrombosis and PE, respectively, per 20-ppb increase in 24-h avg NC>2
23 concentrations (Dales et al.. 2010). Spiezia et al. (2014) also examined the association
24 between ambient air pollution and PE hospital admissions in a small case-control study of
25 105_adults in Padua, Italy. The authors observed an increase in the risk of unprovoked PE
26 for subjects who were in the upper tertile of NOx exposure (average exposure for the
27 month leading up to hospitalization >124jig/m3) compared to those in the bottom two
28 exposure tertiles (OR: 2.35 [95% CI: 0.76, 7.25]).
January 2015 5-283 DRAFT: Do Not Cite or Quote
-------
5.3.8 Cardiometabolic Effects
1 There were no epidemiologic studies of diabetes or insulin deficiency available for the
2 2008 ISA for Oxides of Nitrogen. Two recent studies reported contrasting findings
3 regarding short-term associations between air pollutants and measures of insulin
4 resistance, which plays a key role in the development of type two diabetes mellitus (DM).
5 In a panel study of older adults in Korea, Kim and Hong (2012) observed a 1.33_(iU/mL
6 (95% CI: 0.54, 2.11) increase in insulin resistance and a 0.52_mean (95% CI: 0.24, 0.77)
7 increase in the homeostatic model assessment index of insulin resistance [fasting insulin
8 * (fasting glucose -^ 22.5)] per 20-ppb increase in 24-h avg NC>2 at lag 7. The association
9 was stronger in participants with a history of DM but still present for those without. On
10 the other hand, Kelishadi et al. (2009) reported a lack of an association between 24-h avg
11 NO2 and insulin resistance in a study of 374_Iranian children aged 10-18_years. Both of
12 the recent studies relied on central site monitoring for exposure estimation, and neither
13 evaluated potential confounding by other traffic-related pollutants. There are a number of
14 recent prospective studies reporting associations between diabetes and long-term
15 exposure to oxides of nitrogen (Section 6.3.3).
5.3.9 Aggregated Cardiovascular Effects
16 Many epidemiologic studies consider the composite endpoint of all cardiovascular
17 diseases, which typically includes all diseases of the circulatory system (e.g., heart
18 diseases and cerebrovascular diseases). Most studies reviewed in the 2008 ISA for
19 Oxides of Nitrogen found positive associations between ambient NO2 concentrations and
20 risk of hospital admissions or ED visits for all cardiovascular diseases (U.S. EPA. 2008a)
21 (Figure 5-20 and Table 5-52). However, it was unclear at that time whether these results
22 truly indicated effects of NO2 or were confounded by other correlated pollutants. Several
23 additional studies are now available with broadly consistent results, though some
24 uncertainty still remains with regard to potential copollutant confounding.
25 Ito et al. (2011) observed that risk of CVD hospital admission was associated with NO2
26 concentrations at lag 0 in New York City. Results from copollutant models were not
27 reported. Zheng et al. (2013) and Son etal. (2013) observed seasonal variation in the
28 strength of association between NO2 and CVD hospital admission in Lanzhou, China and
29 Korea, respectively. In contrast, Ito et al. (2011) did not find any seasonal differences in
30 New York City. A study in Santiago, Chile reported an increase in risk of CVD hospital
31 admissions per increase in 24-h avg NO2, which remained relatively unchanged after
32 adjustment for highly correlated traffic-related copollutants, CO (r = 0.94) or PM2 5
33 (r = 0.87) (no quantitative results; results presented graphically) (Franck et al., 2014). In
January 2015 5-284 DRAFT: Do Not Cite or Quote
-------
1 Beijing, China, Guo et al. (2009) reported an association between ambient NO2
2 concentrations and risk of CVD hospital admissions at lag 0 (OR: 1.05 [95% CI: 1.00,
3 1.11] per 20-ppb increase in 24-h avg NO2), but this association was attenuated and had a
4 wide 95% CI in copollutant models adjusting for either PM2 5 (OR: 1.02 [95% CI: 0.96,
5 1.09])orSO2(OR: 1.01 [95% CI: 0.94, 1.081). Sarnat et al. (2013b) reported a positive
6 association between NOx concentrations and CVD ED visits in Atlanta. This study
7 compared the strength of the association across exposure assessment techniques and
8 estimated larger effects using spatially refined ambient concentration metrics (AERMOD,
9 Air Pollution Exposure model, and a hybrid model of background concentrations and
10 AERMOD) in contrast to central site monitoring data. However, there is uncertainty
11 regarding the extent to which an association with NOx reflects an association with NO2
12 (Sections 1.1 and 2.5).
13 In Shanghai, China, Chen et al. (2010) found a 1.02% (95% CI: -2.0, 4.0) increased risk
14 of hospital admission for CVD per 20-ppb increase in 24-h avg NO2 concentrations (lag
15 0-l_days). This association was robust to additional adjustment for PMio but was
16 attenuated after adjustment for SO2 (Table S5-5). A study in Sao Paulo, Brazil, also
17 found a positive association with some evidence that the association was stronger among
18 patients with a secondary diagnosis of diabetes mellitus (Filho et al.. 2008). (Jevtic et al..
19 2014) reported a positive association that was robust to the inclusion of SO2 in a
20 copollutant model in Novi Sad, Serbia. Studies from Copenhagen, Denmark (Andersen et
21 al.. 2008b); Madrid, Spain (Linares and Diaz. 2010); Reykjavik, Iceland (Carlsen et al..
22 2013); and Taipei, Taiwan (Chan et al.. 2008) reported null or negative associations
23 between NO2 concentrations and risk of hospital admission for CVD. A study in
24 Guangzhou, China also found no clear association between NO2 and CVD hospital
25 admissions, with observed associations alternating between positive and negative
26 depending on the lags examined (Zhang et al.. 2014).
27 In summary, consistent evidence reported in the 2008 ISA for Oxides of Nitrogen
28 combined with recent epidemiologic data continue to support the presence of an
29 association between ambient NO2 concentrations and risk of hospital admission for
30 cardiovascular diseases (Figure 5-20 and Table 5-52). However, despite generally
31 consistent associations, there were a limited number of studies that evaluated potential
32 confounding by correlated traffic-related copollutants, particularly EC, PM2 5, UFPs, and
33 VOCs, resulting in some uncertainty of the independent effect of NO2 on cardiovascular
34 disease hospital admissions and ED visits (Table 5-52).
January 2015 5-285 DRAFT: Do Not Cite or Quote
-------
Mean
Study Concentration Lag Notes
Guoetal. (2009) 36.3 0
Chen et al. (2010)
Zhang et al. (In Press)
Itoetal. (2011)
Sonetal. (2013)
Larrieu et al. (2007)
Samat et al. (2013)
Poloniecki et al. (1997)
Chang et al. (2005)
Yang et al. (2004)
Linn et al. (2000)
Wong et al. (1999)
von Klot et al. (2005)
Andersen et al. (2008)
Hinwood et al. (2006)
Llorca et al. (2005)
Filho et al. (2008)
Atkinson et al. (1999)
Peel et aL (2007)
Tolbert et al. (2007)
Fung et al. (2005)
Jalaludin et al. (2006)
Simpson et al. (2005)
Ballester et al. (2006)
Morgan et al. (1998)
30.3 0 -<
0-1
29.8 0
12 ^
I
17.9-26.8 0
11.9-24.9
6.3-30.0 0-2 CS
BG
AERMOD
Hybrid
35 1
31.5 >20C
28.17 0-2 25- C
0-2 <25C
28-41 0
27.3 0-1
12.1-37.2 0 > 35 yrs; MI survivors
11 0-3 — <
10,3 1 All Ages
11.3 N02
9.76 NO
61.1 0 diabetic '
0 non-diabetic <
j-i :
24-h
r
-
0
•
*
L
'
•
IM
•
ft
2. 54 (2. 27, 2.84) >
•
0
•
k
— ^~
i 1-h
i
i
i
p
i
50.3 0 h
•
45.9 0-2 case-crossover
0-2 time series
43.2 0-2
38.9 0 < 65 vrs ^1
fi 1 J
> 65 vrs
23.2 0 > 65 vrs
0-1
16.3-23.7 0
0-1
12.4-40.5 0-1 24-h
61.8 0 1-h
15 0 24-h
29 0 1-h
_ft
*_
_
•^
^
1.
• 24-h and 1-h
"•
_._
0.5 0.75 1 1.25
Relative Risk (95%CI)
1.5
1.75
Note: Studies in red are recent studies. Studies in black were included in the 2008 ISA for Oxides of Nitrogen. Relative risks are
standardized to a 20-ppb or 30-ppb increase in NO2 or NO concentrations and 40 ppb or 60 ppb for NOX concentrations for 24-h avg
and 1-h max metrics, respectively. Studies are presented in descending order of mean concentration, with the largest mean
concentration (ppb) at the top and the smallest at the bottom of the figure (by averaging time and inclusion in previous ISA).
Circles = NO2; Triangles = NO; Diamonds = NOX. Franck et al. (2014) not presented in table due to lack of quantitative results.
Figure 5-20 Results of studies of short-term exposure to oxides of nitrogen
and hospital admissions for all cardiovascular disease.
January 2015
5-286
DRAFT: Do Not Cite or Quote
-------
Table 5-52 Corresponding effect estimates for hospital admissions for all
cardiovascular disease studies presented in Figure 5-20.
Study
Location
Relative Risk3 (95% Cl)
Copollutant Examination13
tGuo et al. (2009) Beijing, China
LagO: 1.05(1.00, 1.11)
Lag 1: 1.03(0.985, 1.09)
NO2: associations attenuated by
PM2.5 or SO2 adjustment.
Copollutants: PlVb.s and SO2
associations robust to adjustment
for NO2.
NO2 correlations (Pearson r). PIVhs:
0.67; SO2: 0.53.
tChen et al. (2010) Shanghai, China
Lag 0: 0.997 (0.970, 1.025)
Lag 1: 1.02(0.99, 1.05)
Lag 0-1: 1.01 (0.98, 1.04)
NO2: associations robust to
adjustment for PM-io; attenuated by
SO2 adjustment.
Copollutants: PM-ioand SO2
associations attenuated by
adjustment for NO2.
NO2 correlations (Pearson r): PMm
0.70; SO2: 0.76.
tZhanqetal. (2014)
tltoetal. (2011)
tSonetal. (2013)
fLarrieu et al.
(2007)
fSarnat et al.
(2013b)
Guangzhou,
China
New York City, NY
8 Korean cities
8 French cities
Atlanta, GA
LagO: 1.03(0.98, 1.
Lag 1: 1.02(0.95, 1.
Lag 2: 0.97 (0.91, 1.
LagO: 1.04(1.03, 1.
Lag 1: 1.01 (1.00, 1.
LagO: 1.04(1.02, 1.
Lag 1: 1.00(0.98, 1.
1.02(1.00, 1.04)
Central site
Lag 0-2: 1.01 (1.00,
08)
08)
03)
05)
,02)
06)
01)
, 1-01)
No copollutant models.
NO2 correlations (Spearman
PMio: 0.82; SO2: 0.60.
No copollutant models.
r):
No copollutant models.
NO2 correlations (Pearson r): PMio:
0.5; SO2: 0.6; CO: 0.7; O3: -0.1.
No copollutant models.
No copollutant models.
Background (distance2 weighting)
Lag 0-2: 1.05(1.01, 1.09)
AERMOD
Lag 0-2: 1.01 (1.00, 1.02)
Hybrid—background and AERMOD
Lag 0-2: 1.01 (1.0, 1.02)
Poloniecki et al.
(1997)
London, U.K.
Lag 1: 1.02(1.00, 1.04)
No Copollutants models analyzed
for CVD.
Chang et al. (2005) Taipei, Taiwan
>20°C: 1.39(1.32, 1.45)
<20°C: 1.23(1.12, 1.37)
NO2: associations robust to
adjustment for PMio, SO2, CO, or
Os, with the exception of PMio on
cold days.
Yang et al. (2004) Kaohsiung,
Taiwan
>25°C, lag 0-2: 1.46(1.31, 1.62)
<25°C, lag 0-2: 2.54 (2.27, 2.84)
NO2: associations robust to
adjustment for PMio, SO2, CO, or
Os on cold days. Somewhat
attenuated after adjustment on
warm days, but still positive (except
SO2, still robust).
Linnetal. (2000) Los Angeles, CA Lag 0: 1.03 (1.02, 1.04)
No copollutant models.
NO2 correlations: PMio: 0.67 to
0.88; Os: -0.23 to 0.35; CO: 0.84 to
0.94.
January 2015
5-287
DRAFT: Do Not Cite or Quote
-------
Table 5-52 (Continued): Corresponding effect estimates for hospital admissions
for all cardiovascular disease studies presented in
Figure 5-20.
Study
Location
Relative Risk3 (95% Cl)
Copollutant Examination13
Wonqetal. (1999) Hong Kong, China Lag 0-1: 1.05 (1.03, 1.08)
No copollutant models.
Von Klot et al.
5 European cities Lag 0: 1.16 (1.07, 1.27)
Llorca et al. (2005) Torrelavega,
Spain
NO2: 1.11 (1.05, 1.17)
NO: 1.13(1.07, 1.19)
: associations robust to
(2005)
Andersen et al.
(2008b)
Hinwood et al.
(2006)
Copenhagen,
Denmark
Perth, Australia
Lag
Lag
0-3: 1.00(0.93, 1.10)
1: 1.08(1.04, 1.13)
adjustment for PM-io or Os.
No evidence of an association
between NO2 and CVD. Copollutant
models did not change the results.
No copollutant models.
No copollutant models.
tFilhoetal. (2008)
Sao Paulo, Brazil Diabetics
LagO: 1.00(1.00, 1.00)
Lag 1: 1.00(0.99, 1.00)
Lag 0-1: 1.00(1.00, 1.00)
Non-diabetics
LagO: 1.00(1.00, 1.00)
Lag 1: 1.00(0.99, 1.00)
Lag 0-1: 1.00(1.00, 1.00)
No copollutant models.
NO2 correlations (Pearson r): PM-io:
0.68; SO2: 0.62; CO: 0.58; O3: 0.41.
Atkinson et al.
(1999)
London, U.K.
LagO: 1.01 (1.00, 1.02)
2: association attenuated by
adjustment for BS.
Copollutants: BS robust to
adjustment for NO2.
Peel et al. (2007) Atlanta, GA
Case-crossover; lag 0-2:
1.04(1.02, 1.06)
Time-series; lag 0-2:
1.04(1.02, 1.06)
No copollutant models.
Tolbert et al. (2007) Atlanta, GA
Lag 0-2: 1.02(1.01, 1.03)
NO2: association attenuated after
adjustment for CO or PIVh.sTC.
Copollutants: CO and PIVh.sTC
associations robust to adjustment
for NO2.
NO2 correlations (Spearman r): CO:
0.70; PIVh.sTC: 0.65.
Funqetal. (2005)
Windsor, Ontario,
Canada
<65 yr
Lag 0: 0.99 (0.90, 1.13)
Lag 0-1: 1.04(0.93, 1.16)
>65yr
LagO: 1.02(0.96, 1.07)
Lag 0-1: 1.02(0.98, 1.05)
Copollutant results not reported for
NO2.
January 2015
5-288
DRAFT: Do Not Cite or Quote
-------
Table 5-52 (Continued): Corresponding effect estimates for hospital admissions
for all cardiovascular disease studies presented in
Figure 5-20.
Study
Location
Relative Risk3 (95% Cl)
Copollutant Examination13
Jalaludin et al.
Sydney, Australia
LagO: 1.06(1.02, 1.09)
Lag 0-1: 1.01 (0.98, 1.05)
Lag 1: 1.04(1.01, 1.08)
NO2: associations robust to
adjustment for PM-io, PIVh.s, SO2,
C>3, or BS in adults aged 65 yr and
older. Attenuated after CO
adjustment.
Copollutants: CO, PIVh.s, SO2, Os,
and BS associations robust to NO2
adjustment; PM-io association
attenuated.
NO2 correlations: BS: 0.35 to 0.59;
PMio: 0.44 to 0.67; PIVh.s: 0.45 to
0.68; Os: 0.21 to 0.45; CO: 0.55 to
0.71; SO2: 0.52 to 0.56.
Simpson et al.
(2005a)
4 Australian cities
LagO: 1.07(1.05, 1.09)
Lag 0-1: 1.07(1.05, 1.09)
NO2: associations robust to
adjustment for BS; attenuated, but
positive after Os adjustment.
Copollutants: Os negative
association robust to adjustment for
NO2; BS association attenuated, but
positive after adjustment for NO2.
Ballester et al.
(2006)
Spain
24-hNO2, lagO: 1.01 (1.00, 1.03)
1-h NO2, lagO: 1.04(0.99, 1.09)
NO2: associations robust to
adjustment for Os; attenuated but
positive with CO or SO2 adjustment.
Copollutants: CO, BS, PMio, SO2,
and Os associations robust to NO2
adjustment.
Morgan et al. (1998) Sydney, Australia
24-hNO2, lagO: 1.09(1.06, 1.12)
1-h NO2, lagO: 1.06(1.04, 1.09)
No copollutant models.
NO2 correlations: Os: -0.09; PM:
0.53.
Cl = confidence interval, CO = carbon monoxide, CVD = cardiovascular disese, ED = emergency department, NO = nitric oxide,
NO2 = nitrogen dioxide, O3 = ozone, PM = particulate matter, SO2 = sulfur dioxide.
aRelative risks are standardized to a 20-ppb or 30-ppb increase in NO2 or NO or 40 ppb or 60-ppb increase in NOX concentration
for 24-h avg and 1-h max metrics, respectively.
"Relevant relative risks for copollutant models can be found in Figure S5-2. S5-3. S5-4. and S5-5 (U.S. EPA. 2014b. c, d, e).
fStudies published since the 2008 ISA for Oxides of Nitrogen.
5.3.10 Cardiovascular Mortality
i
2
o
3
4
5
6
7
Studies evaluated in the 2008 ISA for Oxides of Nitrogen that examined the association
between short-term NCh exposure and cause-specific mortality consistently reported
positive associations with cardiovascular mortality. Across studies, there was evidence
that the magnitude of the NCh-cardiovascular mortality relationship was similar or
slightly larger than that for total mortality. Recent multicity studies as well as a
meta-analysis of studies conducted in Asian cities Atkinson et al. (2012) provide
evidence that is consistent with those studies evaluated in the 2008 ISA for Oxides of
Nitrogen (Section 5.4. Figure 5-23).
January 2015
5-289
DRAFT: Do Not Cite or Quote
-------
1 The NCh-cardiovascular mortality relationship was further examined in a few studies that
2 analyzed copollutant models. It is important to note that it is difficult to examine whether
3 NO2 is independently associated with cardiovascular mortality because NO2 often is
4 highly correlated with other traffic-related pollutants. In the 17_Chinese cities study
5 (CAPES), Chen et al. (2012b) found that NO2 risk estimates for cardiovascular mortality
6 were slightly attenuated but remained positive in copollutant models with PMio or SO2
7 (6.9% [95% CI: 3.8, 10.1] for a 20-ppb increase in 24-h avg NO2 concentrations at lag
8 0-1; 4.6% [95% CI: 1.1, 8.1] with PMi0; 5.7% [95% CI: 2.5, 9.0] with SO2). Chen et al.
9 (2013) reported similar results when examining stroke mortality in a subset of 8 CAPES
10 cities, i.e., 5.6% increase in stroke mortality (95% CI: 3.4, 8.0) at lag 0-1 for a 20-ppb
11 increase in 24-h avg NO2 concentrations with a slight attenuation of the association in
12 copolluant models with PMio (4.5% [95% CI: 1.8, 7.3]) or SO2 (5.2% [95% CI: 2.1, 8.3]).
13 Also, Chiusolo et al. (2011) found evidence that associations between short-term NO2
14 exposure and cardiovascular mortality remained robust in copollutant models in a study
15 of 10_Italian cities. In an all-year analysis, a 20-ppb increase in NO2 at lag 0-5 was
16 associated with a 10.5% (95% CI: 5.9, 14.8) increase in cardiovascular mortality and a
17 10.1% (95% CI: 4.0, 16.4) increase adjusted for PMio. In a warm season analysis
18 (April-September), the NO2 effect estimate was 19.2% (95% CI: 11.4, 27.4) and 18.8%
19 (95% CI: 10.7, 27.5) with adjustment for O3. Overall, the limited number of studies that
20 have examined the potential confounding effects on the NO2-cardiovascular mortality
21 relationship indicate that associations remain relatively unchanged with adjustment for
22 PMio or SO2, but it remains difficult to disentangle the independent effects of NO2 as
23 confounding by more highly traffic-related copollutants has not been examined.
24 Of the multicity studies evaluated, only the studies conducted in Italy examined potential
25 seasonal differences in the NO2-cause-specific mortality relationship (Chiusolo et al..
26 2011; Bellini et al.. 2007). Additional information with regard to whether there is
27 evidence of seasonal differences in NO2-cardiovascular mortality associations is provided
28 by single-city studies conducted in the U.S. (Sacks et al.. 2012; Ito et al.. 2011). In a
29 study of 15_Italian cities, Bellini et al. (2007) found that risk estimates for cardiovascular
30 mortality were dramatically increased in the summer from 1.5 to 7.3% for a 20-ppb
31 increase in 24-h avg NO2 concentrations at lag 0-1, respectively, with no evidence of an
32 association in the winter. These results were corroborated in a study of 10_Italian cities
33 (Chiusolo et al., 2011). which observed an increase in risk estimates for cardiovascular
34 mortality in the warm season (i.e., April-September) compared to all-year analyses.
35 Chiusolo etal. (2011) did not conduct analyses with only the winter season. U.S. studies
36 conducted in New York, NY (Ito etal.. 2011) and Philadelphia, PA (Sacks etal.. 2012)
37 do not provide consistent evidence indicating seasonal differences (Section 5.4.6).
38 Overall, the cardiovascular mortality results from the multicity studies conducted in Italy
39 are consistent with those observed in the total mortality analyses conducted by Bellini et
January 2015 5-290 DRAFT: Do Not Cite or Quote
-------
1 al. (2007) and Chiusolo et al. (2011). However, as discussed in Section 5.4.3. studies
2 conducted in Asian cities observed very different seasonal patterns, and it remains
3 unclear if the seasonal patterns observed for total mortality would be similar to those
4 observed for cardiovascular mortality in these cities.
5 An uncertainty that often arises when examining the relationship between short-term air
6 pollution exposures and cause-specific mortality is whether analyses that examine
7 statistical modeling parameters, the lag structure of associations, and the C-R relationship
8 provide results that are consistent with those observed for total mortality. In a study
9 conducted in Philadelphia, PA, Sacks et al. (2012) examined whether the various
10 modeling approaches to control for both temporal trends/seasonality and weather used in
11 a number of multicity studies [e.g., National Morbidity, Mortality, and Air Pollution
12 Study (NMMAPS), Air Pollution and Health: A European Approach (APHEA)]
13 influence air pollution-cardiovascular mortality associations when using the same data
14 set. Across models, the authors reported that associations of NO2 with cardiovascular
15 mortality were relatively consistent, with the percentage increase in cardiovascular
16 mortality for a 20-ppb increase in 24-h avg NC>2 concentrations ranging from 1.4 to 2.0%.
17 The results of Sacks et al. (2012) support those of Chen et al. (2013). which found that
18 NO2-stroke mortality associations were robust to using 4 to 10_df per year to control for
19 temporal trends.
20 Studies that examined the lag structure of associations for cardiovascular mortality
21 reported results consistent with those observed for total mortality (see Section 5.4.7). In a
22 study of 10_Italian cities, Chiusolo et al. (2011) reported evidence of an immediate effect
23 of NO2 at lag 0-1 on cardiovascular mortality but also provided evidence for a prolonged
24 effect due to the magnitude of the association being larger at lag 0-5 (Figure 5-24). These
25 results are consistent with those of Chen et al. (2012b) in the CAPES study. The authors
26 found the largest effect at single day lags of 0 and 1 and the average of lag 0-l_days
27 providing support for an immediate effect of NC>2 on cardiovascular mortality
28 (Figure 5-25). However, when examining longer lags Chen etal. (2012b) reported that
29 the magnitude of the association was slightly larger for a 0-4_day lag suggesting a
30 potential prolonged effect. In an analysis of stroke mortality Chen et al. (2013) reported
31 similar results in a subset of eight Chinese cities from the CAPES study.
32 To date, analyses detailing the C-R relationship between air pollution and cause-specific
33 mortality have been limited. In the analysis of eight Chinese cities, Chen et al. (2013)
34 also examined the air pollution and stroke mortality C-R relationship. To examine the
35 assumption of linearity, the authors fit both a linear and spline model to the NC^-stroke
36 mortality relationship. Chen etal. (2013) then computed the deviance between the two
37 models to determine if there was any evidence of non-linearity. An examination of the
January 2015 5-291 DRAFT: Do Not Cite or Quote
-------
deviance did not indicate that the spline model improved the overall fit of the NCh-stroke
mortality relationship (Figure 5-21).
•B
* s
"S o
i §
o
o
50
100 150
NO2 concentration
200
250
Source: Reprinted with permission of Wolters Kluwer Health Chen et al. (2013).
Note: The black line represents the mean estimate, and the dotted lines are 95% confidence intervals.
Figure 5-21 Pooled concentration-response curve for nitrogen dioxide (NO2)
and daily stroke mortality in eight Chinese cities for lag 0-1 day.
5.3.11 Subclinical Effects Underlying Cardiovascular Effects
3 The following subsections review studies of subclinical effects that serve as useful
4 measures of physiological and biochemical responses and could provide mechanistic
5 evidence to describe a role for NO2 in the manifestation of cardiovascular diseases. These
6 subclinical effects are not widely validated markers of specific clinical cardiovascular
7 outcomes but could potentially underlie the development, progression, or indication of
8 various clinical events and provide biological plausibility for multiple outcomes.
January 2015
5-292
DRAFT: Do Not Cite or Quote
-------
5.3.11.1 Heart Rate and Heart Rate Variability
1 HRV provides a noninvasive marker of cardiac autonomic nervous system function. The
2 rhythmic variation in the intervals between heart beats can be quantified in either the time
3 domain or the frequency domain (Task Force of the European Society of Cardiology and
4 the North American Society of Pacing and Electrophysiology. 1996). Common
5 time-domain measures of HRV include the standard deviation of all normal-to-normal
6 intervals (SDNN, an index of total HRV) and the root-mean-square of successive
7 differences (rMSSD, an index influenced mainly by the parasympathetic nervous
8 system). In the frequency domain, HRV is usually divided into the high frequency (HF)
9 and low frequency (LF) components, as well as the ratio of the LF to HF components
10 (LF/HF) (Task Force of the European Society of Cardiology and the North American
11 Society of Pacing and Electrophysiology. 1996). Decreases in indices of HRV have been
12 associated with increased risk of cardiovascular events in prospective cohort studies (La
13 Rovereetal.. 2003: Kikuva et al.. 2000: Tsujietal.. 1996: Tsujietal.. 1994).
Epidemiologic Studies
14 The 2008 ISA for Oxides of Nitrogen reported that there was insufficient evidence to
15 determine whether exposure to oxides of nitrogen was associated with changes in cardiac
16 autonomic control as assessed by indices of HRV (U.S. EPA. 2008a). Additional studies
17 are now available for review (Table 5-53) that provide evidence for an association
18 between exposure to NCh and HRV among those with pre-existing disease but not in
19 healthy individuals.
20 The multicountry ULTRA study assessed the longitudinal association between ambient
21 pollution and HRV among older adults with stable coronary artery disease in Amsterdam,
22 the Netherlands; Erfurt, Germany; and Helsinki, Finland (Timonen et al.. 2006). In each
23 participant, HRV was assessed multiple times over a 6-month period, resulting in a total
24 of l,266_repeated measures. Pooling results across the three cities, the authors found a
25 S.Ol.msec (95% CI: -5.94, -0.11) decrease in SDNN and a 17.67% (95% CI: -31.95,
26 -3.01) decrease in LF/HF associated with a 20-ppb increase in 24-hour average NO2
27 concentrations at lag day_2. The magnitudes of these associations were somewhat larger
28 in relation to the 5-day moving average of NO2. The authors report that these effects were
29 robust to adjustment for other pollutants in copollutant models, including UFPs, PM2 5, or
30 CO, but detailed results were not provided. These results were reportedly similar in men
31 and women and after exclusion of those exposed to environmental tobacco smoke at
32 home. Most associations with HF were positive.
January 2015 5-293 DRAFT: Do Not Cite or Quote
-------
Table 5-53
Study
Timonen et al.
(2006)
fZanobetti et al.
(2010)
fBartell et al.
(2013)
Epidemiologic
Location
Sample Size
Amsterdam,
Netherlands;
Erfurt, Germany;
Helsinki, Finland
n = 131
Boston, MA
n=46
(aged 43-75 yr)
Los Angeles, CA
n = 50
studies of
Pre-Existing
Condition
Coronary
artery disease
Coronary
artery disease
Coronary
artery disease
heart rate/heart rate
Mean NO2
ppb
24-h avg NO2
Amsterdam:
22.7
Erfurt: 15.4
Helsinki: 16.5
2-h avg NO2
50th: 21
75th: 27
95th: 36
72-h avg NO2
50th: 21
75th: 25
95th: 31
24-h avg
NOx: 42.3
Max: 183.7
Exposure
variability.
Assessment Selected Effect Estimates3 (95% Cl)
Central
monitor
City-wide
avg
SDNN (msec) LF/HF (%)
Lag 0: -1.05 (-3.50, 1.39) Lag 0: -3.01 (-15.41,
Lag 1: -1.28 (-3.98, 1.43) Lag 1: -16.54 (-30.08
Lag 2: -3.01 (-5.94, -0.11) Lag 2: -17.67 (-31.95
Lag 3: -0.68 (-3.42, 2.07) Lag 3: -1.88 (-15.41,
Lag 0-4: -4.59 (-9.32, 0.15) Lag 0-4: -25.9 (-50.0
HF (%)
Lag 0:4.51 (-33.46, 18.42)
Lag 1:4.51 (-10.15, 19.55)
Lag 2: 1.88 (-13.91, 18.05)
Lag 3: -1.88 (-16.92, 13.16)
Lag 0-4: 4.51 (-22.18, 30.83)
HF (% Change)
2-h: -18.27 (-29.45, -6.82)
Lag 0-4: -47.00 (-70.50, -22.00)
9.77)
, -3.01)
, -3.01)
11.65)
, -1.88)
All other results presented graphically, no quantitative results.
Monitors on
trailers
parked at
each of
4 retirement
communities
No quantitative results presented; results presented
graphically.
Generally null associations between NOx and SDNN
medication use in participants with and without acetylcholine
esterase inhibitors.
January 2015
5-294
DRAFT: Do Not Cite or Quote
-------
Table 5-53 (Continued): Epidemiologic studies of heart rate/heart rate variability.
Study
Location
Sample Size
Pre-Existing
Condition
Mean NO2
ppb
Exposure
Assessment
Selected Effect Estimates3 (95% Cl)
tBarclay et al. Aberdeen, Stable heart 24-h avg Central HR
(2009) Scotland, U.K. failure NO2: 30.1 monitor NO2: 0.398 (-0.003, 0.799)
n = 132 24-h avg NO: NO: 0.353 (-0.036, 0.742)
14.7 SDNN(msec)
NO2: 0.619 (-0,588, 1.826)
NO: 0.608 (-0.562, 1.778)
SDANN (msec)
NO2: 0.512 (-0.865, 1.890)
NO:0.570 (-0.766, 1.906)
rMSSD (msec)
NO2: 0.398 (-0.003, 0.799)
NO: 0.353 (-0.036, 0.742)
pNNSO(%)
NO2: 1.568 (-3.851, 6.986)
NO: 0.909 (-4.347, 6.165)
LF power
NO2: 2.353 (-1.052, 5.757)
NO: 1.940 (-1.328, 5.207)
LF normalized
NO2: -0.857 (-2.817, 1.102)
NO: -0.183 (-2.066, 1.699)
HF power
NO2: 3.365 (-1.169, 7.900)
NO: 2.895 (-1.457, 7.247)
HF normalized
NO2: 0.722 (-1.554, 2.998)
NO: 1.407 (-0.775, 3.558)
LF/HF ratio
NO2: -1.089 (-3.930, 1.753)
NO: -1.054 (-3.779, 1.672)
tGoldberq et al.
(2008)
Montreal,
Quebec, Canada
n = 31
Stable heart 24-h avg NO2 City-wide Pulse Rate (mean difference)
failure 17.9 avg Lag 0:-0.07 (-0.09, 0.80)
Max: 54.1 Lag 1: 0.78 (-0.14, 1.71)
Lag 0-2: 0.99 (-0.34, 2.32)
tSuh and
Zanobetti
(2010b)
Atlanta, GA
(n = 30)
MIorCOPD
24-h avg NO2 City-wide
Ambient: 17.1 avg
Personal: Personal
11.6
SDNN (% change)
Ambient:
-0.64-11.06, 10.43)
Personal: -3.48 (-10.69, 3.89)
rMSSD (% change)
Ambient:
-6.60 (-30.64, 20.85)
Personal:
-14.52 (-29.87, 1.70)
pNNSO (% change)
Ambient: 0.30 (-38.28, 47.38)
Personal:
-32.30 (-56.49, -5.65)
HF (% change)
Ambient:
-1.49 (-37.09, 41.32)
Personal:
-21.35 (-44.92, 4.48)
LF/HF (% change)
Ambient: 13.74 (-4.11,
33.13)
Personal: 9.69 (-2.34, 22.20)
January 2015
5-295
DRAFT: Do Not Cite or Quote
-------
Table 5-53 (Continued): Epidemiologic studies of heart rate/heart rate variability.
Study
tHuanq et al.
(2012a)
tWilliams et al.
(2012b)
tLaumbach et al.
(2010)
fPeel et al.
(2011)
tChuanq et al.
(2007a)b
tJiaetal. (2011)
fParket al.
(2010)
Location
Sample Size
Beijing, China
n=40
Detroit, Ml
n=65
New Brunswick,
NJ
n=21
Atlanta, GA
n= 4,277
Taipei, Taiwan
n = 76
Beijing, China
n=41
6 U.S.
communities:
Baltimore, MD;
Chicago, IL;
Forsyth County,
NC; Los Angeles,
CA; New York,
NY; St. Paul, MN
n = 5,465
Pre-Existing
Condition
CVD
CV risk
factors (i.e.,
hypertension,
hyperlipidemi
a, diabetes)
Diabetes
Healthy
infants
Healthy
Healthy
Healthy
Mean NO2
ppb
1-h max NO2
2007, visit 1:
33.8
2007, visit 2:
26.3
2008, visit 3:
29.2
2008, visit 4:
22.9
24-h avg NO2
24.0
75th: 28.0
Max: 100.0
NO2:
50th: 25.9
75th: 32.8
Max: 61.1
1-h max NO2
41.7
90th: 65.6
Max: 109.2
24-h avg NO2
17.3
Max: 53.1
24-h avg NOx
35.0
24-h avg NO2
Lag 0-1: 23.5
Exposure
Assessment
Central
monitor
Personal
monitor
In-vehicle
mean
Central
monitor
Central
monitor
Central
monitor
City-wide
avg
Selected Effect Estimates3 (95% Cl)
SDNN (% change) LF (% change)
1-h: -1.9 (-3.4, -0.3) 1-h: -5.4 (-9.3, -1.4)
4-h: -3.9 (-5.7, -2.2) 4-h: -8.9 (-13.2, -4.3)
12-h: -3.6 (-5.5, -1.6) 12-h: -7.9 (-12.8, -2.8)
r-MSSD (% change) HF (% change)
1-h: 1.4 (-1.1, 3.9) 1-h: -3.5 (-8.2, 1.4)
4-h: -2.2 (-5.7, 1.5) 4-h: -5.1 (-11.0, 1.3)
12-h: -2.2 (-6.1, 2.0) 12-h: -3.7 (-10.4, 3.5)
HR(bpm)
bpm: -2.95 (-4.82, -0.80)
HF (% change)
-11. 92 (-104.64, 80.79)
LF/HF ratio (% change)
-107.28 (-298.01, 83.44)
Bradycardia (OR)
1.04(1.00, 1.08)
"We found no associations between HRV indices and NO2."
No quantitative results presented.
"No significant effects are found between daily average...
on HRV indices."
No quantitative results presented.
NOx
"There were no significant associations of HRV with gaseous
pollutants (data not shown)."
No quantitative results presented.
January 2015
5-296
DRAFT: Do Not Cite or Quote
-------
Table 5-53 (Continued): Epidemiologic studies of heart rate/heart rate variability.
Location Pre-Existing Mean NO2 Exposure
Study Sample Size Condition ppb Assessment Selected Effect Estimates3 (95% Cl)
tChuanq et al. Taipei, Taiwan Healthy
(2007b) n=46
tMinetal. (2008) Taein Island, Healthy
South Korea
n = 1,349
fWeichenthal et Ottawa, Canada Healthy
al. (2011 )b (n = 42)
tShields et al. Mexico City, Healthy
(2013) Mexico
(n = 16)
1-h max NO2 Avg of "...NO2... exposures were not associated with any HRV indices
38.4 monitors in our study participants (data not shown)."
within 1 km No quantitative results presented.
of residence
24-h avg NO2 Central SDNN (% change)
24 monitor 6-h: -2.45 (-6.28, 1.53)
75th: 30 9-h: -3.89 (-8.31, 0.71)
Max: 119 12-h: -3.81 (-8.75, 1.34)
24-h: -1.72 (-6.71, 3.51)
48-h: 2.93 (-2.33, 8.42)
72-h: 1.20 (-3.81, 6.42)
LF (% change)
6-h: -8.61 (-16.85, 0.31)
9-h:- -12.24 (-21. 48, -2.11)
12-h: -12.28 (-22.58, -0.88)
24-h: -5.71 (-16.58,6.33)
48-h: 3.69 (-8.22, 16.92)
72-h:5.84(-6.19, 18.45)
1-hmaxNO2 Central ALF (ms2)
4.8 Monitor 1-h: -532.5 (-2872.5, 1807.5)
2-h: 12.0 (-2467.5, 2490.0)
3-h: 577.5 (-2055.0, 3217.5)
4-h: -397.5 (-3532.5, 2025.0)
AHF (ms2)
1-h: -420.0 (-1785.0, 952.5)
2-h: -487.5 (-1612.5, 637.5)
3-h: -24.0 (-1020.0, 975)
4-h: -247.5 (-1417.5, 922.5)
ALF:HF
1-h: 5.70 (-2.10, 13.50)
2-h: 10.50(2.63, 18.75)
3-h: 12.75(4.20,21.75)
4-h: 7.50 (-1.80, 17.25)
1-hmaxNO2 In-Vehicle LF (% Change)
130 Monitor -0.69 (-1.91, 0.57)
HF (% Change)
-0.24 (-1.80, 1.47)
HF (% change)
6-h: -1.08 (-10.75, 9.47)
9-h: -3.31 (-14.32, 8.88)
12-h: -2.38 (-14.73, 11.45)
24-h: -4.53 (-16.58, 8.94)
48-h: 4.42 (-8.72, 19.14)
72-h: 4. 18 (-8.52, 18.35)
ASDNN (msec)
1-h: -18.75 (-112.50, 72.00)
2-h: -75.0 (-150.0, -2.55)
3-h: -39.75 (-120.0, 40.50)
4-h: -12.00 (-82.50, 61.50)
ArMSSD (msec)
1-h: -12.00 (-48.75, 24.75)
2-h: -12.00 (-41.25, 17.25)
3-h: 2.33 (-30.0, 34.5)
4-h: -2. 10 (-33.0, 29.25)
ApNNSO (%)
1-h: -3.30 (-31. 50, 24.75)
2-h: -8.25 (-33.00, 15.75)
3-h: -3.23 (-29.25, 22.50)
4-h: 1.28 (-26.25, 29.25)
LF/HF (% Change)
-0.45 (-1.53, 0.64)
SDNN (% Change)
-1.03 (-1.55, -0.49)
January 2015
5-297
DRAFT: Do Not Cite or Quote
-------
Table 5-53 (Continued): Epidemiologic studies of heart rate/heart rate variability.
Study
tRich et al.
(2012)b
tZhanq et al.
(201 3)b
Location
Sample Size
Beijing, China
(n = 125)
Beijing, China
(n = 125)
Pre-Existing
Condition
Healthy
Healthy
Mean NO2
ppb
24-h avg NO2
Entire study:
27.0
Before: 26.0
During: 13.9
After: 41. 4
24-h avg NO2
Before: 25.6
During: 14.6
After: 41. 4
Exposure
Assessment
Central
monitor
Central
monitor
Selected Effect Estimates3 (95% Cl)
No quantitative results presented; results presented
graphically. Positive, but statistically nonsignificant increase in
heart rate, generally consistent across lags from 0 to 6.
No quantitative results presented; results presented
graphically. Statistically significant decreases in SDNN and
rMSSD in the early lags (0 to 1); measurable but non-
statistically significant decreases across lags 2 and 3; and
generally null associations in lags 4, 5, and 6.
Cl = confidence interval, COPD = chronic obstructive pulmonary disease, CVD = cardiovascular disease, HF = high frequency, HR = heart rate, HRV = heart rate
variability, LF = low frequency, LF/HF = LF to HF components, Ml = myocardial infarction, NO = nitric oxide, NO2 = nitrogen dioxide, NOX = sum of NO and NO2,
OR = odds ratio, pNNSO = proportion of pairs of successive normal simus intervals that exceeds 50 milliseconds divided by the total number of successive pairs of
normal simus intervals, rMSSD = root-mean-square of successive differences, SDNN = standard deviation of all normal-to-normal intervals.
"•Relative risks are standardized to a 20-ppb or 30-ppb increase in NO2 or NO or 40 ppb or 60-ppb increase in NOX concentration for 24-h avg and 1 -h max metrics,
respectively.
fStudies published since the 2008 ISA for Oxides of Nitrogen.
January 2015
5-298
DRAFT: Do Not Cite or Quote
-------
1 Huang et al. (2012a) measured HRV repeatedly in participants with pre-existing
2 cardiovascular disease in Beijing in the summer of 2007 and again in the summer of
3 2008, including one measurement period during the 2008 Beijing Olympics when
4 city-wide air pollution control measures substantially reduced ambient concentrations of
5 most criteria pollutants as described in more detail in Section 5.3.6.1. In single-pollutant
6 models, an unspecified IQR increase in 1-h max NO2 was associated with a 3.6%
7 decrease (95% CI: -5.5, -1.6) in SDNN, and a 7.9% decrease (95% CI: -12.8, -2.8) in
8 LF. The association with SDNN was stronger among those with a higher C-reactive
9 protein (CRP), women, and those without a history of diabetes, but BMI did not appear to
10 modify the association. Rich et al. (2012) also examined the association between heart
11 rate and NO2 concentrations before, during and after the 2008 Beijing Olympics. The
12 authors observed increases in heart rate that were generally consistent in magnitude
13 across lags from 0 to 6_days. In expanded results from the same protocol, Zhang etal.
14 (2013) reported that NO2 was inversely associated with SDNN and rMSSD, with stronger
15 associations in the earlier lags (0 to 3). The HR association with NO2 was somewhat
16 attenuated, but still positive after adjustment for PM2 5, CO, SO2, or OC, and no longer
17 positive after adjustment for EC. The decrements in rMSSD and SDNN associated with
18 increased ambient NO2 remained relatively unchanged after adjustment for PM2 5, CO,
19 SO2, OC, or EC.
20 Several studies (Weichenthal et al.. 2011; Laumbach et al.. 2010; Suh and Zanobetti.
21 2010a) used personal exposure assessment techniques that tend to reduce uncertainty in
22 the NO2 exposure estimate when compared to the use of city-wide averages, decreasing
23 the distance between the monitor and subject (Section 3.4.4.2). Suh and Zanobetti
24 (201 Ob) examined the association between HRV and short-term exposure to NO2 among
25 people that had either recently experienced an MI or had COPD. Same-day total personal
26 exposures to NO2 were associated with decreased HRV. Decreases in pNN50 (proportion
27 of pairs of successive normal simus intervals exceeds 50 milliseconds divided by the total
28 number of successive pairs of normal simus intervals) were the largest among the
29 individuals with COPD, while NO2-associated decrements in HF were the largest among
30 individuals with a recent MI, but were less precise when all individuals or individuals
31 with COPD were included in the analysis. Associations were most pronounced when
32 examining personal as opposed to ambient measures of NO2. Copollutant confounding
33 was not assessed. Laumbach et al. (2010) studied the effects of in-vehicle exposure to
34 traffic-related pollutants among a group of individuals with diabetes. The authors did not
35 observe any strong evidence of an association between HF HRV and in-vehicle exposure
36 to NO2. Weichenthal et al. (2011) carried out a cross-over trial with 42_healthy adults
37 who cycled for IJiour on high- and low-traffic routes as well as indoors. Mean
January 2015 5-299 DRAFT: Do Not Cite or Quote
-------
1 concentrations of NO2 measured at nearby stationary monitors were associated with
2 decreases in SDNN and increases in LF/HF.
3 In a repeated-measures study of Boston-area patients with clinically significant coronary
4 artery disease, Zanobetti et al. (2010) found that HF was inversely associated with
5 ambient NO2 concentrations. This association remained robust after adjustment for PM2 5
6 in a copollutant model. Among a population reporting a substantial prevalence of
7 cardiovascular risk factors (i.e., hypertension, diabetes, hyperlipidemia), Williams et al.
8 (2012a) observed a strong association between NO2 concentrations and reduced heart
9 rate. On the other hand, Barclay et al. (2009) reported no association between NO2 or NO
10 and indices of HRV in a repeated-measures study of non-smoking patients with stable
11 heart failure. (Bartell etal.. 2013) observed generally null associations between NO2 and
12 SDNN medication use in retirement residents with coronary artery disease in the greater
13 Los Angeles area. Also, Goldberg et al. (2008) followed 31 Montreal-area participants
14 with stable heart failure for 2 months and found no association between pulse rate and
15 NO2 concentrations.
16 Infants are potentially at greater risk of pollution-related health effects (American
17 Academy of Pediatrics. 2004). Peel etal. (2011) examined data from 4,277_Atlanta-area
18 infants prescribed home cardiorespiratory monitors and observed a slightly elevated risk
19 of bradycardia (OR: 1.04 [95% CI: 1.00, 1.08]) per 30-ppb increase in 1-h max NO2
20 concentrations averaged over the previous 2 days and measured at a central site monitor.
21 The clinical or public health significance of this finding is unclear.
22 The majority of the above studies focused on infants or participants with a documented
23 history of heart disease, with the exception of the Beijing Olympics studies (Zhang etal..
24 2013; Rich etal.. 2012). In contrast to the pre-existing disease studies, there is little
25 evidence that HRV is associated with NO2 concentrations in healthy participants. For
26 example, a repeated-measures study of young healthy participants in Taipei, Taiwan
27 found no association between NO2 and HRV indices (Chuang et al.. 2007a). Another
28 repeated-measures study in Mexico City observed small decrements in SDNN associated
29 with increases in NO2, but no association between NO2 and LF/HF (Shields etal.. 2013).
30 In Beijing, Jiaet al. (2011) assessed HRV two times in each of 20 healthy participants
31 and reported no association between oxides of nitrogen and HRV. However, this study
32 was quite small, and detailed results were not shown.
33 Cross-sectional analyses of populations with or without a history of heart disease have
34 also tended to yield null results. In a cross-sectional analysis of 5,465_participants, ages
35 45-84 years, from the multicity Multiethnic Study of Atherosclerosis, Park et al. (2010)
36 found no association between NO2 concentrations and indices of HRV. A cross-sectional
37 study from Taipei also observed no association between NO2 and HRV among 46 older
January 2015 5-300 DRAFT: Do Not Cite or Quote
-------
1 adults with cardiovascular disease risk factors (Chuang et al.. 2007b). A cross-sectional
2 study of 1,349 healthy participants in Taein Island, South Korea by Min et al. (2008).
3 found that NO2 was associated with decreases in the LF component of HRV, but not with
4 changes in SDNN or the HF component.
5 In summary, current evidence suggests that among participants with pre-existing or
6 elevated risk for cardiovascular disease, ambient NO2 concentrations are associated with
7 alterations in cardiac autonomic control as assessed by indices of HRV; however,
8 evidence for differential effects between populations with and without pre-existing
9 diseases and conditions is limited. In this specific subgroup of the population, NO2 seems
10 to be associated with changes in HRV, which is consistent with relative increases in
11 sympathetic nervous system activity and/or decreases in parasympathetic nervous system
12 activity. In contrast, this association has not been commonly apparent among healthier
13 participants. Additionally, the observed associations in the pre-existing disease
14 populations were generally persistent after adjustment for UFPs, PIVb 5, CO, SO2, or OC
15 in limited traffic pollutant models.
Controlled Human Exposure Studies
16 Controlled human exposure studies evaluating HRV were not available for review in the
17 2008 ISA for Oxides of Nitrogen; since then, two new studies have become available
18 (Table 5-56). Huang et al. (2012b) evaluated changes in various HRV parameters
19 following NO2 exposure in healthy adult volunteers performing intermittent exercise.
20 Exposure to 500_ppb NO2 did not alter HRV time domain intervals, but did slightly
21 increase LF/HF, although this change was not statistically significant. The authors
22 reported an 11.6 and 13% decrease in the HF domain normalized for heart rate (HFn) 1
23 and 18 hours after exposure, respectively. Combined exposure to NO2 and PM2s CAPs
24 increased LF/HF (IJiour; p = 0.062), as well as the low frequency domain normalized for
25 heart rate (1 hour;/? = 0.021) and cardiac t-wave amplitude (18_hour;/> = 0.057). CAPs
26 exposure alone did not induce such changes. Epidemiologic studies found
27 NO2-associated decreases in HRV primarily in adults with or at risk for cardiovascular
28 disease. However, a recent study in resting adults with stable coronary heart disease and
29 impaired left ventricular systolic function showed no statistically significant changes in
30 HRV after exposure to 400 ppb NO2 for 1 hour while seated (Scaife et al.. 2012);
31 however, it should be noted that the study had only 75% power to detect significant
32 differences in the HF domain of 50% or less.
33 The few studies reviewed in the previous assessments of oxides of nitrogen (U.S. EPA.
34 2008a. 1993) reported mixed effects of NO2 exposure on HR; a recent study shows no
35 effect. Folinsbee et al. (1978) and Drechsler-Parks (1995) exposed healthy adult males
January 2015 5-301 DRAFT: Do Not Cite or Quote
-------
1 and healthy older adults, respectively, to approximately 600_ppb NO2 for 2_hours and
2 reported no changes in HR. Changes in HR were also examined in potentially at-risk
3 populations exposed to NO2. Exposure to 400 ppb NO2 did not alter HR in adults with
4 coronary heart disease (Scaife et al.. 2012) and resulted in a statistically nonsignificant
5 increase in adults with COPD and healthy volunteers, (Gong et al.. 2005). Among healthy
6 volunteers and those with asthma, NO2 exposure resulted in no change in HR (Linn et al..
7 1985a).
8 In summary, there is limited, inconsistent evidence from controlled human exposure
9 studies to suggest NO2 alone or in combination with CAPs exposure during exercise
10 alters HRV. Additionally, there appears to be no evidence from controlled human
11 exposure studies to suggest NO2 alters HR.
lexicological Studies
12 Toxicology studies examining HRV changes were not available for review in the 2008
13 ISA for Oxides of Nitrogen. Similar to controlled human exposure studies, a recent study
14 in rats found mixed evidence for changes in HR and HRV (Table 5-57). Ramos-Bonilla
15 et al. (2010) examined body weight, HR, and HRV, following exposure of aged inbred
16 mice to an ambient mixture consisting of PM, CO, and NO2. Animals were exposed to
17 either filtered or unfiltered outdoor Baltimore air for 6 hours daily for 40_weekdays.
18 Health effects associated with daily exposure to each pollutant were ascertained with
19 multipollutant models and lagged covariates. Statistically significant declines in HR were
20 associated with NO2 at lag 3 and the 7-day cumulative concentration with adjustment for
21 PM and CO. However, HRV changes were not associated with NO2 exposure. The
22 independent effects of each pollutant are difficult to distinguish in this multipollutant
23 model because of multicollinearity among pollutants.
5.3.11.2QT-lnterval Duration
24 The QT interval provides an ECG marker of ventricular repolarization. Prolongation of
25 the QT interval is associated with increased risk of life-threatening ventricular
26 arrhythmias. Limited evidence from the epidemiologic study reviewed in the 2008 ISA
27 for Oxides of Nitrogen and the single recent study does not clearly indicate an association
28 between short-term NO2 exposure and markers of ventricular repolarization (U.S. EPA.
29 2008a). Epidemiologic evidence is similarly inconsistent for associations of NO2
30 exposure and arrhythmias (Section 5.3.3).
January 2015 5-302 DRAFT: Do Not Cite or Quote
-------
1
2
o
6
4
5
6
7
Within the context of the Veterans Administration Normative Aging Study, Baja et al.
(2010) found imprecise associations between heart-rate corrected QT interval (QTc) and
10-hour moving average of NCh concentrations among older, generally white men, but
was associated with NO2 concentrations at lags 3 and 4_hours (longer lags or moving
averages were not considered) (Table 5-54). The only prior study available for
comparison found that 24-h avg NC>2 concentrations were positively associated with
increased QTc duration, but this association was imprecise (i.e., had wide confidence
intervals), and the 6-hour moving average of NC>2 was not associated with an increase in
QTc duration (Henneberger et al., 2005).
Table 5-54 Epidemiologic studies of QT-interval duration.
Study
Location
Sample Size
Mean
Concentration
ppb
Exposure
Assessment
Selected Effect Estimates3
(95% Cl)
tBaja et al. (2010) Boston, MA
n = 580
1-h max NO2
19 during ECG
monitoring
21 ppb 10 h before
monitoring
Central site Change in QTc (msec)
10-hlag: 5.91 (-2.03, 13.85)
4-h lag: 6.28 (-0.02, 12.55)
Henneberqer et al. Erfurt, Germany
(2005) n = 56
24-h avg NCb: 18.2 City-wide avg
75th: 22.6
Max: 36.4
24-h avg NO: 19.4
75th: 24.2
Max: 110.1
QTc (msec)
NO2, lag 6-11 h: 9.77 (2.23, 17.33)
T-wave complexity (%)
NO, lag 0-23: 0.15 (0.02, 0.28)
T-wave amplitude (uV)
NO, lag 0-5 h: -2.10 (-4.16, -0.03)
Cl = confidence interval, ECG = electrocardiographic, NO = nitric oxide, NO2 = nitrogen dioxide, QTc = corrected QT interval.
aEffect estimates are standardized to a 20-ppb or 30-ppb increase in NO2 or NO for 24-h avg and 1-h max metrics, respectively.
fStudy published since the 2008 ISA for Oxides of Nitrogen.
10
11
12
13
14
15
16
Controlled Human Exposure Studies
There were no available controlled human exposure studies evaluating changes in the QT
interval for the 2008 ISA for Oxides of Nitrogen; since then, one new study has become
available. Huang etal. (2012b) found a near statistically significant (quantitative results
not reported) decrease in QTc at 1 and 18 hours after a 2-hour exposure to 500_ppb NO2
in healthy exercising adults (Table 5-56). NC>2 exposure also induced a 29.9% decrease
(p = 0.001) in the QT variability index (QTVI). However, when volunteers were exposed
to both PM2.5 and NO2, the QTVI synergistically increased.
January 2015
5-303
DRAFT: Do Not Cite or Quote
-------
5.3.11.3Vascular Reactivity
1 The vascular endothelium plays a fundamental role in the maintenance of vascular tone
2 that is involved in the regulation of blood pressure and blood flow. In a controlled human
3 exposure study, Langrish et al. (2010) examined the effects of NO2 on vascular
4 endothelial tone and fibrinolytic function. In a random crossover double-blind study,
5 healthy male volunteers were exposed to 4,000 ppb of NCh for 1 hour with intermittent
6 exercise. This study employed infusion of endothelial-dependent vasodilators, bradykinin
7 and acetylcholine, and endothelial-independent vasodilators, sodium nitroprusside and
8 verapamil, to examine vascular endothelial tone. The results demonstrated that NO2 did
9 not attenuate the vasodilator response to these vasoactive agents.
10 Epidemiologic studies provide inconsistent evidence regarding a potential association
11 between NCh and vascular function. In an analysis of data from the EPA's Detroit
12 Exposure and Aerosol Research Study, Williams et al. (2012a) found that total personal
13 NC>2 concentrations were associated with changes in brachial artery diameter (positive
14 association at lag 1 and negative association at lag 2) and positive (i.e., presumably
15 beneficial) changes in flow-mediated dilation. No associations were observed in
16 relationship to ambient measures of NC>2. Ljungman et al. (2014) reported no consistent
17 associations between 1-, 2-, 3-, 5-, and 7-day moving averages of NOx and peripheral
18 arterial tonometry ratio in the Offspring and Third Generation Cohorts of the
19 Framingham Heart Study.
5.3.11.4 Blood Biomarkers of Cardiovascular Effects
20 Several epidemiologic and toxicological studies have explored the potential relationship
21 between NCh and biomarkers of cardiovascular risk. In particular, markers of
22 inflammation, cell adhesion, coagulation, and thrombosis have been evaluated in a
23 number of epidemiologic studies published since the 2008 ISA for Oxides of Nitrogen
24 (U.S. EPA. 2008a) (Table 5-55). These biomarkers also have been examined in
25 controlled human exposure and animal toxicological studies.
January 2015 5-304 DRAFT: Do Not Cite or Quote
-------
Table 5-55 Epidemiologic studies of biomarkers of cardiovascular effects.
Study
Location Pre-existing Mean NO2 Exposure
Sample Size Condition ppb Assessment
Selected Effect Estimates3 (95% Cl)
tPelfinoetal. (2008b)
Los Angeles, Coronary 24-h avg NO2 Indoor and
CA artery Outdoor: outdoor home
n = 29 disease 33.1 measurements
Max: 59.8
Indoor:
32.3
Max: 53.5
Outdoor:
CRP (ng/mL)
LagO: 1,125 (-314, 2,565)
Lag 0-2: 1,027 (-465, 2,520)
Fibrinogen (ug/mL)
LagO:-110.31 (-504,283)
Lag 0-2: -110 (-502, 281)
IL-6 (pg/mL)
LagO: 1.32(0.48,2.18)
Lag 0-2: 1.17(0.28,2.08)
IL-6R (pg/mL)
Lag 0: -493 (-9,387, -249)
Lag 0-2: -3,212 (-7,789, 1,365)
TNF-a (pg/mL)
LagO: 0.13 (-0.26, 0.52)
Lag 0-2: 0.15 (-0.22, 0.54)
TNF-RII (pg/mL)
Lag 0:290 (-41, 623)
Lag 0-2: 240 (-82, 562)
P-selectin (ng/mL)
LagO: 5.13 (-1.02, 11.27)
Lag 0-2: 1.49 (-5.04, 8.02)
VCAM-1 (pg/mL)
LagO: 53,734 (-11,381,
118,849)
Lag 0-2: 18,266 (-45,532,
82,062)
ICAM-1 (pg/mL)
LagO: 5,381 (-8,987, 19,748)
Lag 0-2: 575 (-13,495, 14,643)
SOD (U/g Hb)
LagO: -541 (-1,021, -63)
Lag 0-2: -571 (-1,036, -106)
GPx (U/g Hb)
LagO: -1.99 (-3.68,-0.26)
Lag 0-2: 1.15 (-2.81, 0.58)
MPO (ng/mL)
LagO: -5.34 (-14.92, 4.33)
Lag 0-2: -1.15 (-10.81, 8.44)
tDelfino et al. (2009)
Los Angeles, Coronary 24-h avg NO2 Hourly outdoor
CA artery Phase 1: home air
n = 60 disease 26.4 measurements
Phase 2:
28.3
24-h avg NOx
Phase 1:
37.2
Phase 2:
53.9
NOx:
IL-6 (pg/mL)
Lag 0:0.23 (0.12, 0.35)
Lag 0-2: 0.23(0.10, 0.35)
P-selectin (ng/mL)
LagO: 1.52(0.09,2.94)
Lag 0-2: 2.29 (0.68, 3.90)
TNF-RII (pg/mL)
Lag 0.66 (7.93, 124.76)
Lag 0-2: 86.54(18.75, 155.05)
TNF-a (pg/mL)
Lag 0: 0.01 (-0.06, 0.07)
Lag 0-2: 0.04 (-0.03, 0.12)
NOx:
CRP (ng/mL)
Lag 0: 469.47 (212.74, 726.92)
Lag 0-2: 408.17 (111.06,
705.29)
SOD (U/g Hb)
LagO: -100.24 (-201.92, 2.16)
Lag 0-2: -95.91 (-214.90,
22.36)
GPx (U/g Hb)
LagO: -0.17 (-0.61, 0.26)
Lag 0-2: -0.14 (-0.63, 0.36)
January 2015
5-305
DRAFT: Do Not Cite or Quote
-------
Table 5-55 (Continued): Epidemiologic studies of biomarkers of cardiovascular effects.
Study
tDelfinoetal. (2010)
tWittkopp et al. (2013)
tUunqman etal. (2009)
tBruskeetal. (2011)
Location Pre-existing
Sample Size Condition
Los Angeles, Coronary
CA artery
n = 60 disease
Los Angeles, Coronary
CA artery
n = 36 disease
6 European Ml
cities
n = 955
(total n = 5,539
measurements)
Augsburg, Ml
Germany
n=200
Mean NO2 Exposure
ppb Assessment
Warm Hourly outdoor
season home air
24-h avg measurements
NO2: 26.4
24-h avg
NOx: 37.2
Cool season
24-h avg
NO2: 28.3
24-h avg
NOx: 53.9
24-h avg Hourly outdoor
NOx: 45.35 home air
Max: 188.00 measurements
24-h avg NO2 City-wide avg
22.6
24-h avg NO2 Central site
20.8 monitor
75th: 24.7
Max: 38.2
Selected Effect Estimates3 (95% Cl)
IL-6 (pg/mL)
NO2: 0.48 (-0.06, 1.05)
NOx: 0.60 (0.26, 0.96)
No quantitative results presented; results presented graphically.
Statistically significant positive associations between 1-, 2-, 3-, and
5-day avg NOx and IL-6 (pg/mL) and TNF-a (pg/mL) in haplogroup
H participants. Statistically nonsignificant, but negative associations
between 1-, 2-, 3-, and 5-day avg NOx and IL-6 (pg/mL) and TNF-a
(pg/mL) in haplogroup U participants.
IL-6 (% change)
Overall: 4.02 (0.47, 8.04)
IL-6 genetic variants
IL6rs2069832(1,1):
7.33(2.13, 12.77)
IL6rs2069832(1,2):
2.84 (-1.18, 7.09)
IL6 rs2069832 (2,2):
-1.18 (-8.27, 5.91)
IL6rs2069840(1,1):
4.26 (-0.95, 9.46)
IL6rs2069840(1,2):
4.02(0.00,8.04)
IL6 rs2069840 (2,2):
4.02 (-3.55, 11.58)
Lp-PLA2 (% Change)
NO2Lag4: 7.28(3.00, 11.56)
NO Lag 4: 2.74 (-0.21, 5.70)
IL6rs2069845(1,1):
6.62(1.18, 12.06)
IL6rs2069845(1,2):
3.07 (-0.95, 7.33)
IL6 rs2069845 (2,2):
0.47 (-6.15, 7.57)
IL6rs2070011 (1,1):
4.96 (-0.24, 10.17)
IL6rs2070011 (1,2):
3.78 (-0.24, 7.80)
IL6rs2070011 (2,2):
2.60 (-4.26, 9.69)
IL6rs1800790(1,1):
2.36 (-2.13, 6.86)
IL6rs1 800790 (1,2):
6.62(1.42, 11.82)
IL6rs1 800790 (2,2):
10.40(0.24,21.04)
January 2015
5-306
DRAFT: Do Not Cite or Quote
-------
Table 5-55 (Continued): Epidemiologic studies of biomarkers of cardiovascular effects.
Study
Location Pre-existing Mean NO2 Exposure
Sample Size Condition ppb Assessment
Selected Effect Estimates3 (95% Cl)
24-h NO
24.0
75th: 25.8
Max: 141.1
"Inverse associations were observed for... NO2with Lp-PLA2 at lag
days 1-2 and positive associations were estimated ...with Lp-PLA2
lagged 4 and 5 days."
tBarclay et al. (2009)
Aberdeen,
Scotland
n = 132
Stable
chronic heart
failure
24-h avg
NO2:
30.1
24-h avg NO:
14.7
Central monitor
Hemoglobin
NO2: 0.035 (-0.291, 0.361)
NO: -0.011 (-0.331,0.310)
Mean corpuscular hemoglobin
NO2: 0.050 (-0.158, 0.257)
NO: -0.039 (-0.243, 0.165)
Platelets
NO2: -0.049 (-0.867, 0.768)
NO: 0.247 (-0.556, 1.050)
Hematocrit
NO2: -0.017 (-0.350, 0.316)
NO: 0.101 (-0.226, 0.428)
WBC
NO2: -0.722 (-1.670, 0.226)
NO: -0.708 (-1.640, 0.224)
CRP
NO2: 0.423 (-5.263, 6.108)
NO: 0.890 (-4.694, 6.473)
IL-6
NO2: 6.276 (0.594, 11.940)
NO: 2.767 (-2.810, 8.344)
vWF
NO2: 2.164 (-0.328, 4.655)
NO: 3.522(1.091, 5.954)
E-selectin
NO2: 1.162 (-0.372, 2.696)
NO: 0.483 (-1.022, 1.989)
Fibrinogen
NO2: -0.219 (-1.759, 1.322)
NO: 0.195 (-1.320, 1.709)
Factor VII
NO2: 0.273 (-1.441, 1.987)
NO: 0.335 (-1.348, 2.018)
d-dimer
NO2: -0.243 (-2.781, 2.294)
NO: -0.316 (-2.807, 2.175)
Wellenius etal. (2007) Boston, MA
n=28
Stable 24-h avg NO2 City-wide avg
chronic heart 20.7
failure
"No significant associations were observed between (NO2)
BNP levels at any of the lags examined."
No quantitative results presented.
and
tHildebrandt et al. (2009) Erfurt,
Germany
n = 38
COPD 24-h avg NO2 Central monitor
13.5
24-h NO
10.7
Increases in fibrinogen and prothrombin fragment 1 + 2 associated
with NO concentrations. A decrease in vWF was associated with
NO2 concentrations. No quantitative results presented for NO or
NO2.
tDadvand etal. (2014) Barcelona,
Spain
n=242
COPD
24-h avg
NO2: 30.7
Residential land-
use regression
CRP (% change)
Lag 1:2.99 (-21.6, 34.16)
Lag 2: 26.05 (-3.47, 64.08)
Lag 5: 54.92 (23.21, 94.47)
TNF-a (% change)
Lag 1: 3.90 (-24.6, 41.5)
Lag 2: 10.61 (-19.43, 51.37)
Lag 5: (26.54 (-4.51, 66.85)
IL-6 (% change)
Lag 1: 10.47 (-13.46, 40.14)
IL-8 (% change)
Lag 1: 7.94 (-2.07, 18.98)
Lag 2: 8.09 (-2.00, 19.12)
Lag 5: 3.44 (-5.37, 13.11)
Fibrinogen (% change)
Lag 1: 3.57 (-1.84, 9.09)
Lag 2: 3.26 (-2.00, 8.72)
Lag 5: 10.42(5.59, 15.58)
HGF (% change)
Lag 1: 3.11 (-3.91, 10.58)
January 2015
5-307
DRAFT: Do Not Cite or Quote
-------
Table 5-55 (Continued): Epidemiologic studies of biomarkers of cardiovascular effects.
Study
Location Pre-existing Mean NO2 Exposure
Sample Size Condition ppb Assessment
Selected Effect Estimates3 (95% Cl)
Lag 2: 4.31 (-18.28, 32.88) Lag 2: 5.57 (-1.58, 13.24)
Lag 5: 21.28 (-2.47, 50.40) Lag 5: 9.99 (3.44, 16.98)
tKhafaie et al. (2013)
Pune City,
India
n = 1,392
Type 2
diabetes
24-h avg City-wide avg No quantitative results presented; results presented graphically.
NOx: 21.1 NOx was statistically significantly associated with increases in CRP
at lags 0, 1, 2, 4, and 0-7. There were no measurable differences
between winter and summer associations.
tBindetal. (2012) Boston, MA Healthy
n = 704
24-h avg NO2 City-wide avg
18
95th: 35
Fibrinogen (% change)
Lag 0-2: 8.18(4.73, 11.64)
tRenetal. (2011) Boston, MA Healthy
n = 320
24-h avg NO2 Central site
17.8 monitor
8-OhdG (% change)
Lag 0:28.48 (-19.39, 76.36)
Lag-6: 90.00 (-12.22, 191.67)
Lag 0-13: 166.88(28.75,
305.63)
Lag 0-20: 195.15(44.85,
344.85)
tThompson et al. (2010) Toronto,
Canada
n=45
Healthy 24-h avg NO2 Central site
23.8 monitor
Quantitative results not presented.
January 2015
5-308
DRAFT: Do Not Cite or Quote
-------
Table 5-55 (Continued): Epidemiologic studies of biomarkers of cardiovascular effects.
Study
Location Pre-existing Mean NO2 Exposure
Sample Size Condition ppb Assessment
Selected Effect Estimates3 (95% Cl)
tRudez et al. (2009)
Rotterdam, the Healthy 24-h avg NO2 Central site Maximal platelet aggregation
Netherlands 50th: 19.7 monitor (% change) NO; NO2
n=40 75th: 25.5 Lag 0-6 h: 5.42 (-18.33, 29.58);
Max: 43.1 -4.11 (-13.04, 4.82)
24-h NO: Lag 0-12 h: 2.92 (-22.50,
50th: 5.6 28.33);
75th: 12 -4.64 (-15.00, 5.89)
Max: 130.4 Lag 0-24 h: 7.92 (-12.50,
28.75);
-5.36 (-18.39, 7.68)
Lag 24-48 h:
5.00 (-17.08, 27.08);
-1.07 (-11.79, 9.46)
Lag 48-72 h: 25.42 (10.00,
40.42);
10.00(2.68, 17.32)
Late aggregation (% change)
NO; NO2
Lag 0-6 h: 33.75 (-5.00, 72.08);
5.89 (-9.46, 21.07)
Lag 0-12 h: 35.42 (-2.92,
73.33);
13.39 (-4.11, 30.71)
Lag 0-24 h: 37.08 (4.67, 69.17);
17.68 (-4.46, 39.82)
Lag 24-48 h:
22.92 (-6.25, 51.67);
3.39 (-16.07, 22.68)
Lag 48-72 h: 32.92 (9.58,
55.83);
15.89(4.64,27.14)
Lag 72-96 h:
14.17 (-23.75, 52.50);
8.57 (-7.68, 24.82)
Lag 0-96 h: 54.17(20.42,
87.92);
28.75(8.93,48.57)
Thrombin generation—Peak (%
change) NO; NO2
Lag 0-6 h:-1.67 (-15.00,
11.67);
-2.68 (-9.82, 4.46)
Lag 0-12 h: -1.67 (-12.92,
9.58);
-1.25 (-9.11, 6.61)
Lag 0-24 h: -2.50 (-16.25,
10.83);
-1.07 (-9.46, 7.32)
Lag 24-48 h: 17.08(4.58,
30.00);
14.29(4.29,24.29)
Lag 48-72 h: 5.00 (-6.67,
16.67);
6.61 (-2.68, 16.07)
Lag 72-96 h: 14.58(1.67,
27.92);
-0.36 (-8.57, 7.86)
Lag 0-96 h: 12.92 (-7.08,
32.50);
1.79 (-7.32, 10.71)
January 2015
5-309
DRAFT: Do Not Cite or Quote
-------
Table 5-55 (Continued): Epidemiologic studies of biomarkers of cardiovascular effects.
Study
Location Pre-existing Mean NO2 Exposure
Sample Size Condition ppb Assessment
Selected Effect Estimates3 (95% Cl)
tRudez et al. (2009)
(continued)
Rotterdam, the Healthy 24-h avg NO2 Central site Thrombin generation—ETP
Netherlands 50th: 19.7 monitor (% change) NO; NO2
n = 40 75th: 25.5 Lag 0-6 h: -1.67 (-9.58, 6.25);
Max: 43.1 -2.14 (-6.43, 2.14)
24-h NO: Lag 0-12 h: -1.67 (-8.33, 4.58);
50th: 5.6 -0.36 (-5.00, 4.29)
75th: 12 Lag 0-24 h: -1.25 (-9.17, 6.67);
Max: 130.4 0.54 (-4.46, 5.54)
Lag 24-48 h: 7.92 (0.42, 15.42);
6.25(0.36, 12.14)
Lag 48-72 h: 1.43 (-3.39, 6.25);
7.08 (-4.58, 18.75)
Lag 72-96 h: 8.75 (0.83, 16.67);
1.43 (-3.39, 6.25)
Lag 0-96 h: 7.08 (-4.58, 18.75);
1.96 (-3.04, 7.14)
Thrombin generation—Lag time
(% change) NO; NO2
Lag 0-6 h: -0.42 (-5.83, 4.58);
0.00 (-2.86, 2.86)
Lag 0-12 h: 0.00 (-4.58, 4.17);
0.00 (-3.21, 3.04)
Lag 0-24 h: 2.50 (-2.50, 7.50);
0.36 (-2.86, 3.57)
Lag 24-48 h: -7.50 (-12.08,
-2.92);-5.54 (-9.11, -1.79)
Lag 48-72 h: -3.33 (-7.50,
1.25);
-4.46 (-7.68, 1.07)
Lag 72-96 h: -5.83 (-10.83,
-0.83); 0.00 (-3.21, 3.04)
Lag 0-96 h: -4.58 (-11.67,
2.08);
-1.25 (-4.46, 1.96)
Fibrinogen—lag time (% change)
NO; NO2
Lag 24-48 h: 0.42 (-4.17, 5.42);
0.71 (-3.04, 4.46)
Lag 48-72 h: 1.25 (-3.33, 5.83);
2.50 (-1.07, 6.07)
Lag 72-96 h: 0.42 (-4.58, 5.83);
-0.71 (-4.11,2.50)
CRP—lag time (% change) NO;
NO2
Lag 24-48 h: 15.00 (-12.08,
41.67); 11.61 (-8.75, 31.79)
Lag 48-72 h: 0.42 (-27.08,
27.92);
-0.18 (-19.64, 19.29)
Lag 72-96 h: -19.17 (-50.00,
12.08);-12.32 (-30.71, 6.25)
tSteenhofetal. (2014)
Utrecht, the
Netherlands
n = 31
Healthy
5-h avg NO2:
20
Central monitors
at each of 5 sites
No quantitative results presented; results presented graphically.
NO2 was statistically significantly associated with decreases in
eosinophils and lymphocytes 2-h after exposure. Null associations
were observed between NO2 and WBC count, neutrophils, or
monocytes.
January 2015
5-310
DRAFT: Do Not Cite or Quote
-------
Table 5-55 (Continued): Epidemiologic studies of biomarkers of cardiovascular effects.
Study
tStraketal. (2013)
Steinvil et al. (2008)
tHildebrandtetal. (2009)
tKhafaie et al. (2013)
tKelishadi et al. (2009)
tLeeetal. (2011)
Location
Sample Size
Utrecht, the
Netherlands
n = 31
Tel Aviv, Israel
n = 3,659
Erfurt,
Germany
n = 38
Pune City,
India
n = 1,392
Isfahan, Iran
2004-2005
(n = 374)
Allegheny
County, PA
1997-2001
(n = 2,211)
Pre-existing
Condition
Healthy
Healthy
Healthy
Healthy
Healthy
Healthy
Mean NO2 Exposure
ppb Assessment
5-h avg NCb: Central monitors
20 at each of 5 sites
24-h avg NO2 City-wide avg
19.5
75th: 25.3
24-h avg NO2 Central monitor
13.5
24-h NO
10.7
24-h avg City-wide avg
NOx:21.1
24-h avg: City-wide avg
35.8
75th %: 47.2
Max: 271
7-d avg: 8.4 City-wide avg
75th %: 10.1
Max: 25.4
Selected Effect Estimates3 (95% Cl)
Endogenous thrombin [in FXII-mediated thrombin generation
pathway (% Change)].
All Sites: 65.47 (7.63, 144.68)
Outdoor: 76.14 (-2.23, 154.51)
CRP (% change) men; women WBC (cells/uL) men; women
Lag 0: 0.31 (-7.87, 12.60); Lag 0: 22.05 (-155.91, 200.00);
-4.72 (-17.32, 9.45) -83.46 (-305.51, 138.58)
Lag 1: -7.87 (-17.32, 9.45); Lag 1: 39.37 (-146.46, 223.62);
-3.15 (-15.75, 11.02) -20.47 (-244.09, 203.15)
Lag 2: -1.57 (-11. 02, 11.02); Lag 2: -36.22 (-226.77, 154.33);
0.00 (-12.60, 15.75) 18.90 (-218.90, 255.12)
Fibrinogen (mg/dL) men; women
Lag 0: -9.92 (-15.59, -4.25);
-12.44 (-19.84, -5.20)
Lag 1: -7.87 (-13.86, -2.05);
-5.51 (-12.91, 1.89)
Lag 2: -7.09 (-13.07, -1.10);
-1.42 (-9.45, 6.46)
Increases in fibrinogen and prothrombin fragment 1 + 2 associated
with NO concentrations. A decrease in vWF was associated with
NO2 concentrations. No quantitative results presented for NO or
NO2.
No quantitative results presented; results presented graphically.
NOx was statistically significantly associated with increases in CRP
at lags 0, 1,2, 4, and 0-7. There were no measurable differences
between winter and summer associations.
NO2 positively associated with CRP and markers of oxidative stress
(oxidized-LDL, malondialdehyde, and conjugated diene).
No quantitative results presented. "... NO2 ... associations [with
CRP] were negligible for both the entire population and nonsmokers
only."
Chuanq et al. (2007a)
Taipei, Taiwan Healthy
n = 76
24-h avg NO2 Central monitor
17.3
Max: 53.1
"There was no association between NO2 and any of the blood
markers."
No quantitative results presented.
January 2015
5-311
DRAFT: Do Not Cite or Quote
-------
Table 5-55 (Continued): Epidemiologic studies of biomarkers of cardiovascular effects.
Location Pre-existing Mean NO2 Exposure
Study
Baccarelli etal. (2007)
tRich etal. (2012)
tZhanqetal. (2013)
Sample Size Condition
Lombardia, Healthy
Italy
n = 1,213
Beijing, China Healthy
n = 125
Beijing, China Healthy
n = 125
ppb Assessment
24-h avg NO2 City-wide avg
Median: 22.7
75th: 33.7
Max: 194.2
24-h avg NO2 Central monitor
Entire study:
27.0
Before: 26.0
During: 13.9
After: 41. 4
24-h avg NO2 Central monitor
Before: 25.6
During: 14.6
After: 41. 4
Selected Effect Estimates3 (95% Cl)
Homocysteine difference Homocysteine, post-methionine-
(% change) load (% change)
Lag 24 h: 0.24 (-2.86, 3.57) Lag 24 h: 0.00 (-2.86, 2.86)
Lag 0-6 days: -2.21 (-6.01, Lag 0-6 days: 0.49 (-2.94, 4.17)
1.72)
No quantitative results presented; results presented graphically.
Positive and statistically significant increase in sCD62P, generally
consistent across lags from 0 to 6. Generally null associations with
sCD40L across lags from 0-6. Positive and statistically significant
increases in vWF and fibrinogen at early lags (lag 0, lag 1 ) but null,
or negative at later lags. Generally null or negative associations
with WBC across lags 0-6.
No quantitative results presented; results presented graphically.
Statistically significant increase in fibrinogen at lag 0. Positive, but
statistically nonsignificant at lags 1, 2, 3, and 6.
8-OhdG = urinary 8-hydroxy-29-deoxyguanosine, Cl = confidence interval, COPD = chronic obstructive pulmonary disease, CRP = C-reactive protein, GPx = glutathione peroxidase,
HGF = hepatocyte growth factor, ICAM-1 = inter-cellular adhesion molecule 1, IL = interleukin, Lp-PLA2 = lipoprotein-associated phospholipase A2, Ml = myocardial infarction,
NO = nitric oxide, NO2 = nitrogen dioxide, NOX = sum of NO and NO2, SOD = superoxide dismutase, TNF-a = tumor necrosis factor alpha, VCAM-1 = vascular adhesion molecule-1,
vWF = von Willebrand factor, WBC = white blood cells.
aEffect estimates are standardized to a 20-ppb or 30-ppb increase in NO2 or NO or 40-ppb or 60-ppb increase in NOX concentration for 24-h avg and 1-h max metrics, respectively.
fStudies published since the 2008 ISA for Oxides of Nitrogen.
January 2015
5-312
DRAFT: Do Not Cite or Quote
-------
Epidemiologic Studies
1 Levels of some circulating systemic inflammatory markers appear to be related to NC>2
2 concentrations among participants with a history of heart disease. Delfino et al. (2008b)
3 followed nonsmoking elderly subjects with a history of coronary artery disease living in
4 retirement communities in Los Angeles, CA and measured plasma biomarkers weekly
5 over a 12-week period. The authors observed that indoor and/or outdoor NO2
6 concentrations measured at the retirement homes were associated with increases IL-6 and
7 the soluble tumor necrosis factor a receptor II (sTNFa-RII), markers of systemic
8 inflammation, but not associated with a number of other biomarkers of inflammation and
9 vascular injury including CRP, P-selectin, D-dimer, TNF-a, soluble intercellular adhesion
10 molecule-1 (sICAM-1), or soluble vascular adhesion molecule-1 (sVCAM-1). In
11 subsequent analysis of overlapping populations, Delfino et al. (2009) and Delfino et al.
12 (2010) found that NC>2 and NOx were both associated with circulating levels of IL-6.
13 Delfino et al. (2009) also observed positive associations with P-selectin, TNF-RII, and
14 CRP. Working with the same study population, Wittkopp et al. (2013) also found an
15 association between NOx concentrations and increases in IL-6 and TNF-a, but only for
16 participants with mitochondrial haplogroup H, which has been linked to oxidative
17 damage and increased risk of age-related diseases. Similarly, Ljungman et al. (2009)
18 repeatedly measured plasma IL-6 in 955 MI survivors from six European cities and found
19 that NC>2 was associated with increased levels of IL-6, and that the strength of the
20 association varied in individuals with specific variants of inflammatory genes. However,
21 in studies conducted among patients with stable chronic heart failure, no associations
22 were observed between any biomarkers (including hematological markers and markers of
23 inflammation) and NC>2 concentrations (Barclay et al.. 2009; Wellenius et al., 2007).
24 None of these studies examined potential confounding by traffic-related copollutants.
25 In Augsburg, Germany, Briiske etal. (2011) measured lipoprotein-associated
26 phospholipase A2 (Lp-PLA2), a marker of vascular inflammation and an independent
27 predictor of coronary heart disease events and stroke, up to six_times in 200_participants
28 with a history of myocardial infarction. They found that Lp-PLA2 was associated with
29 both NO and NO2. However, the association was negative at short lags and positive at
30 longer lags, making interpretation of these results difficult.
31 The results have been more heterogeneous in participants without a history of heart
32 disease. One semi-experimental design assessed changes in blood biomarker levels in
33 healthy participants exposed to ambient air pollution at five locations in the Netherlands
34 with contrasting air pollution characteristics (Steenhof et al.. 2014; Straket al.. 2013). A
35 particular strength of these studies is the measurement of pollutants at the location of
36 participants' outdoor exposure, which minimizes measurement error from time-activity
January 2015 5-313 DRAFT: Do Not Cite or Quote
-------
1 patterns and variability in NO2 concentration (Sections 3.4.4.1 and 3.4.4.2). Steenhof et
2 al. (2014) reported that NO2 was negatively associated with both eosinophil and
3 lymphocyte counts. Importantly, this could either be due to eosinophils and lymphocytes
4 leaving the blood and infiltrating stressed tissue, or a decrease in formation of eosinophils
5 and lymphocytes. The associations were relatively unchanged after adjustment for PIVb 5,
6 EC, OC, or PMio in copollutant models. Straketal. (2013) observed an increase in
7 thrombin generation in the endogenous pathway (FXII-mediated) associated with
8 ambient outdoor NO2, that was robust to adjustment for EC or OC, and slightly
9 attenuated after adjustment for PM2 5.
10 Among older men participating in the Normative Aging Study, Bind et al. (2012) found
11 that NO2 was associated with fibrinogen, sVCAM-1, and sICAM-1, but not CRP. In this
12 same cohort, Ren etal. (2011) found that NO2 was positively linked with urinary
13 8-hydroxy-29-deoxyguanosine (8-OhdG) concentrations, a marker of oxidative stress
14 resulting in DNA damage. Thompson et al. (2010) analyzed the baseline data on IL-6 and
15 fibrinogen from 45 non-smoking subjects who participated in a controlled human
16 exposure study in Toronto, Canada. Using baseline blood samples allowed the authors to
17 measure the association between systemic inflammation and ambient NO2, prior to
18 controlled exposure. The authors found that NO2 concentrations were not associated with
19 either IL-6 or fibrinogen overall, but IL-6 was associated with NO2 in the winter months.
20 In Rotterdam, the Netherlands, Rudez et al. (2009) measured CRP, fibrinogen, and
21 markers of platelet aggregation and thrombin generation up to 13 times in 40 healthy
22 participants. Both NO2 and NO were associated with markers of platelet aggregation and
23 thrombin generation, but neither NO2 nor NO was associated with CRP or fibrinogen.
24 During the Beijing Olympics, NO2 was positively associated with increases in biomarkers
25 indicative of the thrombosis-endothelial dysfunction mechanism (i.e., sCD62P) and
26 increases in fibrinogen among healthy young adults (Zhang et al., 2013; Rich et al..
27 2012). The association with sCD62P was attenuated, but remained positive after
28 adjustment for PM2 5, CO, Os, SO2, EC, or OC; whereas the association between NO2 and
29 fibrinogen was generally robust to the above pollutants, with the exception of EC and
30 OC. Among 3,659 individuals in Tel-Aviv, Steinvil et al. (2008) found a null association
31 between NO2 concentrations and CRP and a negative association with fibrinogen and
32 white blood cell counts. Baccarelli et al. (2007) observed generally null associations
33 between NO2 concentrations and total homocysteine among subjects in Lombardia, Italy.
34 Similarly, Chuang et al. (2007a) observed no association between NO2 and any blood
35 markers, including markers of systemic inflammation and oxidative stress, as well as
36 fibrinolytic and coagulation factors.
January 2015 5-314 DRAFT: Do Not Cite or Quote
-------
1 Other subgroups that might be at increased risk of pollution-related health effects have
2 also been studied. In a cross-sectional study of COPD patients in Barcelona, Spain, there
3 was evidence of a positive association between NO2 and multiple biomarkers of
4 inflammation and tissue repair, including CRP, TNFa, IL-6, IL-8, fibrinogen, and
5 hepatocyte growth factor (HGF) (Dadvand et al., 2014). These associations were
6 generally strongest at lags of 4 or 5 days. A particular strength of this study is that the
7 authors used validated land use regression (LUR) models to estimate ambient NO2
8 exposure at residential locations. In a repeated-measures study of male patients with
9 chronic pulmonary disease in Germany, Hildebrandt et al. (2009) reported that NO was
10 positively associated with fibrinogen and prothrombin levels but not other markers of
11 coagulation; however, detailed results were not presented in the paper. Khafaie et al.
12 (2013) observed a positive association between NO2 and CRP in a cross-sectional study
13 of type 2 diabetes patients in Pune City, India. In another cross-sectional analysis of
14 pregnant women in Allegheny County, PA, there was no association between NO2 and
15 CRP (Lee et al., 2011). Among 374 Iranian children aged 10-18 years, Kelishadi et al.
16 (2009) found that NO2 was associated with CRP and markers of oxidative stress.
Controlled Human Exposure Studies
17 Markers of inflammation, oxidative stress, cell adhesion, coagulation, and thrombosis
18 have been evaluated in a few controlled human exposure studies published since the 2008
19 ISA for Oxides of Nitrogen (U.S. EPA. 2008a) (Table 5-56). Similar to epidemiologic
20 studies, controlled human exposure studies also report evidence for increases in some
21 inflammatory markers, but not consistently across all studies. There is also evidence for
22 hematological changes following NO2 exposure, and a recent study reported endothelial
23 cell activation.
January 2015 5-315 DRAFT: Do Not Cite or Quote
-------
Table 5-56 Controlled human exposure studies of short-term nitrogen dioxide
(NO2) exposure and cardiovascular effects.
Study
Age; Sex;n
Exposure Details Concentration;
Duration
Endpoints Examined
Channell et al.
(2012)
Drechsler-Parks
(1995)
Folinsbee et al.
(1978)
Frampton et al.
(2002)
Gona et al.
(2005)
Adult (25.3 ± 5.5 yr);
M/F; n = 7
Primary hCAECs
Adult (65.9 ± 9 yr);
M/F; n = 8
Adult (20-25 yr); M;
n = 5/group
Adult;
M (26.9 ± 4.5 yr);
n = 12
F(27.1 ±4.1 yr);
n = 9
Older adult:
Healthy
nonsmokers: 68 ± 11
Adults were exposed to 500 ppb
NO2; 2 h; intermediate intermittent
exercise (15 min on/off; VE = 25
L/min per m2 of BSA). Plasma
samples were collected before
exposures, immediately after, and
24-h post-exposure. hCAECswere
treated with a dilution of these
plasma samples (10 or 30% in
media) for 24 h.
600 ppb; 2 h; intermittent exercise
(20 min on/off)
VE = 26-29 L/min
600 ppb; 2 h; exercise (15, 30, or
60 min; VE = 33 L/min)
600 and 1,500 ppb; 3 h; intermittent
exercise (10 min on/20 min off);
VE = 40 L/min
400 ppb NO2; 2 h; intermittent
exercise (15 min on/off);
VE = 22-26 L/min
LOX-1 protein measured from
plasma pre, immediately post, and
24-h post-exposure ICAM-1 and
VCAM-1 mRNA from hCAECs and
IL-8 and MCP-1 protein from cell
supernatant measured immediately
post-exposure to plasma.
HR was calculated throughout
exposure, cardiac output was
measured during the last 2 min of
each exercise period.
HR, BP, and cardiac output were
measured during exposure.
Venous blood collected for
hematocrit, hemoglobin, and red
blood cell count 3.5 h after
exposure.
HR and BP were measured
immediately post, 4-h post, and
Day 2.
yr; n = 6
Ex-smokers with
COPD: 72 ± 7 yr;
n = 18
Huang et al.
(2012b)
Adult; M/F
(24.56 ± 4.28 yr);
n = 23
500 ppb NO2 and 500 ppb
NO2 + 73.4 ± 9.9 ug/m3 CAPs; 2 h;
intermittent exercise (15 min
on/off); VE = 25 L/min per m2 of
BSA
IL-6, coagulation factors, and lipid
panel in peripheral blood; HRV; and
HR measured 1 and 18 h after
exposure.
January 2015
5-316
DRAFT: Do Not Cite or Quote
-------
Table 5-56 (Continued): Controlled human exposure studies of short-term
nitrogen dioxide (NO2) exposure and cardiovascular
effects.
Study
Age; Sex;n
Exposure Details Concentration;
Duration
Endpoints Examined
Lanqrish et al.
(2010)
Adult; M; (median
age 24 yr); n = 10
4,000 ppb NO2; 1 h; intermittent
exercise; (VE = 25 L/min)
4 h after exposure 5,10, and
20 ug/min acetylcholine; 100, 300,
and 1,000 pmol/min bradykinin; 2,4,
and 8 ug/min sodium nitroprusside;
10, 30, and 100 ug/min verapamil
were infused in the brachial artery
for 6 min/dose during forearm
venous occlusion plethysmography.
Each vasodilator administration
was separated by a 20-min
washout period.
Hemoglobin concentration was
measured 4 and 6 h after exposure,
Forearm blood flow and tissue-
plasminogen activator and
plasminogen-activator inhibitor
Type 1 were measured 4 h after
exposure.
Linn et al.
Adult; M/F
With asthma
(18-34yr); n = 23,
Without asthma
(20-36 yr); n = 25
3,850-4,210 ppb NO2; 75 min;
intermittent exercise; light and
heavy exercise VE = 25 and
50 L/min (15 min of each; light
minute ventilation 25 L/min and
heavy minute ventilation 50 L/min)
HR and BP were measured
throughout exposure.
Posinetal. Adult; NR; NR; 1,000 or 2,000 ppb NO2; 2.5 h; light
(1978) n = 8-10 intermittent exercise (15 min on/off)
Acetylcholinesterase, glutathione,
glucose-6-phosphate
dehydrogenase, lactate
dehydrogenase, erythrocyte
glutathione reductase, erythrocyte
glutathione peroxidase, alpha-
tocopherol, TEARS, serum
glutathione reductase, 2,3
diphosphoglycerate, hemoglobin,
hematocrit.
Riedletal.
(2012)
Adult; M/F
(1) 37.33 ± 10.91 yr;
n = 10 M, 5F
(2) 36.13 ± 2.52 yr;
n = 6M, 9F
(1-2) 350 ppb NO2; 2 h; intermittent Serum levels of IL-6, ICAM-1,
exercise (15 min on/off);
VE = 15-20 L/min x m2 BSA
(1) Methacholine challenge after
exposure.
(2) Cat allergen challenge after
exposure.
fibrinogen, Factor VII, and vWF.
Serum collected 22.5 h after
exposure.
Scaife et al.
Adult (median age
68 yr); with stable
coronary heart
disease or impaired
left ventricular
systolic function;
M/F; n = 18
400 ppb NO2; 1 h
HR and HRV monitored continuously
for 24 h after exposure.
BSA = body surface area, COPD = chronic obstructive pulmonary disease, F = female, hCAEC = human coronary artery
endothelial cells, HR = heart rate, HRV = heart rate variability, ICAM-1 = inter-cellular adhesion molecule 1, IL = interleukin
M = male, MCP-1 = monocyte chemoattractant protein-1, mRNA= messenger RNA, NO2 = nitrogen dioxide, NR = not reported,
TEARS - thiobarbituric acid reactive substances, VCAM-1 = vascular adhesion molecule-1, vWF = von Willebrand factor.
January 2015
5-317
DRAFT: Do Not Cite or Quote
-------
1 In healthy adults, exposure to 500 ppb NC>2 for 2 hours with intermittent exercise did not
2 alter circulating IL-8, a pro-inflammatory cytokine, or coagulation factors but induced a
3 statistically nonsignificant increase in the pro-inflammatory cytokine, IL-6 (Huang et al..
4 2012b). Lipid profile changes were also reported. There was a 4.1% increase in blood
5 total cholesterol (p = 0.059) and a 5.9% increase in high density lipoprotein cholesterol
6 (p = 0.036) 18 hours after exposure, but no changes in low density lipoprotein or very
7 low density lipoprotein cholesterol or triglycerides.
8 The controlled human exposure study by Langrish et al. (2010) examined the effects of
9 NC>2 on fibrinolytic function. The endogenous fibrinolytic pathway was assessed by
10 sampling venous concentrations of tissue-plasminogen activator and
11 plasminogen-activator inhibitor Type I at baseline and 4 hours after exposure.
12 Concentrations of these proteins were not affected by exposure to NC>2.
13 Atherosclerosis is a chronic inflammatory disease; early stages of the disease include
14 inflammatory activation of endothelial cells and adhesion of leukocytes to the vascular
15 endothelium. Channell et al. (2012) reported endothelial cell activation in an in vitro
16 model following NC>2 exposure. Plasma samples were collected from healthy volunteers
17 exposed to filtered air or 500 ppb NO2 for 2 hours with intermittent exercise. Primary
18 human coronary artery endothelial cells (hCAECs) were then treated with a dilution of
19 these plasma samples (10 or 30% in media) for 24 hours. Increases in messenger RNA
20 (mRNA) expression levels of endothelial cell adhesion molecules, vascular adhesion
21 molecule-1 (VCAM-1) and ICAM-1, from hCAECs were observed for both
22 post-exposure time points compared to control. Cells treated with plasma (30%) collected
23 immediately post NC>2 exposure had a statistically significant greater release of IL-8 but
24 not monocyte chemoattractant protein-1 (MCP-1). In addition, plasma collected 24 hours
25 post NC>2 exposure had a significant increase (30%) in soluble lectin-like oxLDL receptor
26 (LOX-1) levels, a protein recently found to play a role in the pathogenesis of
27 atherosclerosis (Sections 4.3.2.9 and 4.3.5).
28 Riedl etal. (2012) reported on the cardiovascular effects of healthy volunteers and
29 individuals with asthma exposed to filtered air, diesel exhaust, or 350 ppb NO2 for
30 2 hours with intermittent exercise. No statistically significant differences were found in
31 IL-6, ICAM-1, and blood coagulation factors [i.e., factor VII, fibrinogen, and von
32 Willebrand factor (vWF)] the morning after NC>2 exposure.
33 Studies from the 2008 ISA for Oxides of Nitrogen (U.S. EPA. 2008a) reported
34 NO2-induced hematological changes. Frampton et al. (2002) reported decreases in
35 hematocrit, hemoglobin, and red blood cell count in healthy volunteers 3.5 hours after
36 exposure to 600 and 1,500 ppb NO2 for 3 hours with intermittent exercise. Results from
37 this study support those of Posinetal. (1978). in which hematocrit and hemoglobin levels
January 2015 5-318 DRAFT: Do Not Cite or Quote
-------
1 were decreased in young males exposed to 1,000 and 2,000 ppb NO2 for approximately
2 2.5 hours with intermittent exercise. However, a recent study reported no change in
3 hemoglobin levels 4 and 6 hours post-exposure to 4,000 ppb NO2 for 1 hour (Langrish et
4 al.. 2010).
5 lexicological Studies
6 The 2008 ISA for Oxides of Nitrogen (U.S. EPA. 2008a) reported on various
7 hematological parameters in animals including oxidative stress, red blood cell turnover,
8 and methemoglobin levels. Similar to epidemiologic and controlled human exposure
9 studies, several recently published toxicological studies have examined the potential
10 association between short-term NO2 exposure and biomarkers of cardiovascular effects,
11 including markers of oxidative stress, inflammation, and cell adhesion (Table 5-57).
12 Recently, the effects of NO2 on markers of oxidative stress were examined by Li et al.
13 (201 la). Rats exposed to 2,660 or 5,320 ppb NO2 for 7 days had a small, but statistically
14 significant decrease in the activity of the antioxidant enzyme Cu/Zn-SOD and, at the
15 higher dose, an increase in malondialdehyde (MDA, an indicator of lipid peroxidation) in
16 heart tissue. These changes were accompanied by mild pathological changes in the heart.
17 However, there were no changes in Mn-SOD or GPx activity or protein carbonyl (PCO)
18 levels at either exposure concentration. Campen et al. (2010) reported Apolipoprotein E
19 knockout mice (ApoE~'~) exposed to 200 and 2,000 ppb NO2 had a
20 concentration-dependent decrease (statistically significant linear trend) in the expression
21 of the antioxidant enzyme HO-1 in the aorta. Together, these results demonstrate the
22 ability of NO2 inhalation to perturb the oxidative balance in the heart and aorta.
23 The effects of NO2 on antioxidant capacity were also examined in the context of diet (de
24 Burbure et al.. 2007). Rats were placed on low (Se-L) or supplemented (Se-S) selenium
25 (Se) diets and were exposed to 5,000 ppb NO2 for 5 days. Se is an integral component of
26 the antioxidant enzyme GPx. GPx levels in red blood cells (RBC) increased in both
27 groups immediately and 48 hours after exposure; however, plasma levels were decreased
28 in Se-L diet rats at both time points. SOD activity in RBCs also decreased in Se-L diet
29 rats at both time points but increased in Se-S diet rats 48 hours after exposure. Overall,
30 NO2 exposure stimulated oxidative stress protective mechanisms with high Se but were
31 mixed with low Se.
32
January 2015 5-319 DRAFT: Do Not Cite or Quote
-------
Table 5-57 Animal toxicological studies of short-term nitrogen dioxide (NO2)
exposure and cardiovascular effects.
Species (Strain);
Study Age; Sex;n
Exposure Details (Concentration;
Duration)
Endpoints Examined
Campen et al. Mice (ApoE~'~);
(2010) 8 weeks; M;
n = 5-10/group
High fat diet; 200 ppb, 2,000 ppb
NO2; 6 h/day for 7 days
ET-1, MMP-9, HO-1, and TIMP-1 mRNA
expression in aorta; TEARS in aorta;
MMP-2/9 activity in aorta. Endpoints
measured 18 h after exposure.
de Burbure et
al. (2007)
Kunimoto et
al. (1984)
Li et al.
(2011 a)
Rats (Wistar);
8 weeks; M;
n = 8/group
Rats (Wistar);
16-20 weeks; M;
n = 6/group
Rats (Wistar);
Adults; M;
n = 6/group
Selenium: 6 ug/day or 1.3 ug/day:
1,000 ppb NO2; 28 day, 6 h/day,
5 days/week (Se+/Se-)
10,000 ppb NO2; 28 day, 6 h/day,
5 days/week;
5,000 ppb NO2; 5 days, 6 h/day
50,000 ppb; 30 min
4,000 ppb NO2; continuously for
1-10 days
2,660, 5,320, and 10,640 ppb NO2;
6 h/day for 7 days
GPx in plasma and red blood cell lysate;
SOD activity in red blood cell lysate; GST
activity in red blood cell lysate; TEARS in
plasma. Endpoints examined
immediately and 48 h after exposure.
ATPase activity, sialic acid, and hexose
in red blood cell membranes measured
after 1, 4, 7, and 10 days of exposure.
H&E staining of heart tissue; Cu/Zn-
SOD, Mn-SOD activity, GPx activity,
MDA level, and PCO level in heart tissue;
ET-1, eNOS, TNF-a, IL-1, and ICAM-1
mRNA and protein levels in heart tissue;
cardiac myocyte apoptosis. Endpoints
examined 18 h after exposure.
Mersch et al. Guinea pigs; NR; 360 ppb NCb; continuously for
(1973) n = 8 7 days
D-2,3-diphosphoglycerate content in red
blood cells; collection time NR.
Mochitate and Rats (Wistar); 4,000 ppb NCb; continuously for
Miura (1984) 16-20 weeks; M; 1-10days
n = 6
PK and PFK activity and hemoglobin
content in red blood cells was measured
after 1, 3, 5, 7, and 10 days of exposure.
Nakajima and Mice (ICR);
Kusumoto 4 weeks; M;
(1968) n = NR
800 ppb NO2; continuously for
5 days
Metahemoglobin in blood from the heart
taken immediately after exposure.
Ramos- Mice (AKR/J); Low-pollution chamber:
Bonillaetal. 180 days; M; (21.2 ppb NO2, 465 ppb CO,
(2010) n = 3/group 11.5 ug/m3 PM);
High-pollution chamber:
(36.1 ppb NO2, 744 ppb CO,
36.7 ug/m3 PM);
6 h/day, 5 days/week, 40 weekdays
ECG (HR, SDNN, r-MSSD, TP, LF, HF,
LH:HF), BW;
Endpoints measured throughout the
exposure.
CO = carbon monoxide, ECG = electrocardiograph^, eNOS = endothelial nitric oxide synthase, ET-1 = endothelin-1,
GPx = glutathione peroxidase. GST = Glutathione-S-transferase, HF = high frequency, HR = heart rate, ICAM-1 = inter-cellular
adhesion molecule 1, IL = interleukin, LF = low frequency, M = male, MDA = malondialdehyde, mRNA = messenger RNA,
NO2 = nitrogen dioxide, NR = not reported, PCO = protein carbonyl, PFK = phosphofructokinase, PK = pyruvate kinase,
PM = particulate matter, SDNN = standard deviation of all normal-to-normal intervals, Se = selenium, SOD = superoxide
dismutase, TEARS - thiobarbituric acid reactive substances.
January 2015
5-320
DRAFT: Do Not Cite or Quote
-------
1 The effects of NO2 on vascular tone modifiers, endothelin-1 (ET-1), and endothelial nitric
2 oxide synthase (eNOS) were recently examined in two studies (Li etal.. 201 la: Campen
3 et al., 2010). ET-1 is a potent vasoconstrictor while the enzyme eNOS catalyzes the
4 production of NO, which induces vasodilation. Campen et al. (2010) did not see a
5 statistically significant increase in ET-1 expression level in the aorta after exposure of
6 mice to 200 and 2,000 ppb NO2. However, exposure to higher NO2 concentrations
7 induced a statistically significant increase in ET-1 in the heart at the mRNA (10,640 ppb)
8 and protein level (5,320 and 10,640 ppb) (Li et al.. 201 la). eNOS mRNA and protein
9 levels were increased at both 2,660 and 5,320 ppb NO2 and decreased to control levels at
10 10,640 ppb NO2. At ambient-relevant concentrations of NO2 exposure, there was an
11 increase in eNOS, while higher concentrations elicited an increase in the vasoconstrictor,
12 ET-1.
13 Studies have also reported changes in some inflammatory markers and adhesion
14 molecules after NO2 exposure in animals. Li et al. (201 la) observed a statistically
15 significant increase in TNF mRNA levels in the heart at 5,320 ppb NO2. In addition, IL-1
16 expression and protein levels were increased; however, this effect was in response to a
17 higher NO2 concentration. ICAM-1 transcription and protein levels were increased in the
18 heart after both the 2,660 and 5,320 ppb NO2 exposures. These results are consistent with
19 the increase in ICAM-1 mRNA Channell et al. (2012) found in an in vitro model
20 described above.
21 The 2008 ISA for Oxides of Nitrogen (U.S. EPA. 2008a) reported on several animal
22 studies examining hematological parameters. Three studies indicate elevated levels of a
23 younger population of red blood cells following NO2 exposure. Red blood cell
24 D-2,3-diphosphoglycerate levels, important in hemoglobin-oxygen dissociation, were
25 increased in guinea pigs following a 7-day continuous exposure to 360 ppb NO2 (Mersch
26 et al.. 1973). Kunimoto et al. (1984) reported an increase in red blood cell sialic acid after
27 24 hours of exposure to 4,000 ppb NO2. Similarly, Mochitate and Miura (1984) reported
28 an elevation of the glycolytic enzymes pyruvate kinase and phosphofructokinase after a
29 7-day continuous exposure to 4,000 ppb NO2. These results suggest an increase in red
30 blood cell turnover after NO2 exposure. Nakajima and Kusumoto (1968) reported that
31 mice exposed to 800 ppb NO2 continuously for 5 days had no change in the
32 oxygen-carrying metalloprotein hemoglobin, methemoglobin.
Summary of Blood Biomarkers of Cardiovascular Effects
33 In summary, the evidence across disciplines for changes in blood biomarkers of
34 cardiovascular effects is inconsistent; however, there is limited but supportive evidence
35 for measures of NO2-induced systemic inflammation. Some epidemiologic evidence
January 2015 5-321 DRAFT: Do Not Cite or Quote
-------
1 suggests the presence of an association between NO2 concentrations and some markers of
2 systemic inflammation among participants with a history of heart disease. However,
3 potential copollutant confounding was not evaluated in these studies, so the possibility
4 remains that the associations observed were the artifact of correlated pollutants. This
5 association is not consistently observed in healthy individuals. Other potentially at-risk
6 populations have not been clearly identified due to contrasting or limited evidence.
7 Controlled human exposure studies evaluating systemic inflammation demonstrated
8 inconsistent results; however this may be due to a dose-dependent effect. Toxicological
9 studies reported an increase in some inflammatory mediators, as well as oxidative stress
10 effects in RBC, the heart, and aorta of rodents. Other biological markers of
11 cardiovascular effects, also discussed in Section 4.3.2.9. demonstrated that short-term
12 NO2 exposure causes a slight reduction in hematocrit and hemoglobin levels associated
13 with a decrease in RBC levels in controlled human exposure studies. Toxicological
14 studies reported an increase in RBC turnover. The clinical significance of these findings
15 is unknown (Section 4.3.2.9). Evidence has not shown NO2 to alter circulating blood
16 coagulation factors or modify the body's response to vasodilators in controlled human
17 exposure studies. However, in toxicological studies at higher concentrations, NO2 was
18 found to induce the expression and production of the vasoconstrictor ET-1.
19 Overall, there is preliminary evidence, albeit not entirely consistent, from controlled
20 human exposure and toxicological studies that suggests systemic inflammation and
21 oxidative stress can occur after exposure to NO2. However, changes in other blood
22 biomarkers, such as coagulation or vasomotor response, are not observed in relation to
23 NO2 exposure.
5.3.12 Summary and Causal Determination
24 Available evidence is suggestive of, but not sufficient to infer, a causal relationship
25 between short-term exposure to oxides of nitrogen and cardiovascular health effects. The
26 strongest evidence comes from epidemiologic studies of adults and consistently
27 demonstrates a relationship between short-term exposure to NO2 and triggering of an MI.
28 This is supported by epidemiologic studies reporting NO2-associated hospitalizations and
29 ED visits for MI, IHD, and angina, ST-segment alterations, and mortality from
30 cardiovascular disease. There is a lack of experimental studies that evaluate similar
31 clinical outcomes in order to assess the coherence across disciplines. However, some
32 controlled human exposure and toxicological studies provide limited evidence for
33 potential biologically plausible mechanisms, including inflammation and oxidative stress.
34 Evidence for other cardiovascular and related metabolic effects is inconsistent.
January 2015 5-322 DRAFT: Do Not Cite or Quote
-------
1 This conclusion represents a change from the 2008 ISA for Oxides of Nitrogen, which
2 concluded the "available evidence on the effects of short-term exposure to NO2 on
3 cardiovascular health effects was inadequate to infer the presence or absence of a causal
4 relationship at this time" (U.S. EPA. 2008a). Specifically, the epidemiologic panel
5 studies and toxicological studies available at the time of the last review were inconsistent.
6 Most epidemiologic studies reviewed in the 2008 ISA for Oxides of Nitrogen found
7 positive associations between ambient NO2 concentrations and risk of hospital
8 admissions or ED visits for all cardiovascular diseases (U.S. EPA. 2008a). However, it
9 was unclear at that time whether these results supported a direct effect of short-term NO2
10 exposure on cardiovascular morbidity or were confounded by other correlated pollutants.
11 Recent epidemiologic studies have further evaluated this uncertainty using copollutant
12 models and comparing associations of NO2 with those of other criteria pollutants. While
13 the recently reviewed studies provide suggestive evidence for independent associations of
14 NO2 with cardiovascular effects after adjusting for some pollutants, uncertainties still
15 remain regarding the potential for NO2 to serve as an indicator for other
16 combustion-related pollutants or mixtures. Specifically, there is a lack of epidemiologic
17 studies evaluating traffic-related pollutants (i.e., PM2 5, BC/EC, UFPs, or VOCs) in
18 copollutant models with NO2.
19 There continues to be a lack of experimental evidence in coherence with the
20 epidemiologic studies to strengthen the inference of causality for NO2-related
21 cardiovascular effects, including MI. Further, the limited mechanistic evidence to
22 describe a role for NO2 in the triggering of cardiovascular diseases, including key events
23 within the mode of action, remains from the 2008 ISA for Oxides of Nitrogen. The
24 evidence for cardiovascular effects, with respect to the causal determination for
25 short-term exposure to NO2, is detailed below using the framework described in the
26 Preamble (Tables I and II). The key evidence, supporting or contradicting, as it relates to
27 the causal framework, is summarized in Table 5-58.
5.3.12.1 Evidence on Triggering a Myocardial Infarction
28 The causal determination for the relationship between short-term NO2 exposure and
29 cardiovascular effects is based on the evidence for effects related to triggering an MI,
30 including findings for hospital admissions and ED visits for IHD, MI, or angina and
31 ST-segment amplitude changes. Time-series studies of adults in the general population
32 consistently report positive associations between 24-h avg and 1-h max NO2
33 concentrations and hospital admissions and ED visits for IHD and MI among adults
34 (Figure 5-18. Section 5.3.2.1). Risk estimates ranged from 0.87 to 1.76 per a 20 or 30 ppb
35 increase in NO2, with the magnitude of most of the risk estimates greater than 1.00.
January 2015 5-323 DRAFT: Do Not Cite or Quote
-------
1 Symptoms of MI are similar to those of angina; however, where MI results in damage to
2 the heart muscle, angina does not result in myocardial necrosis. However, angina may
3 indicate an increased risk for future MI. IHD is an over-arching category of related
4 ischemic events that includes both acute MI and angina, as well as events related to older
5 MI and other IHD-related events. Increased observations of hospital admissions and ED
6 visits for MI and IHD are coherent with epidemiologic studies reporting increased
7 hospital admissions and ED visits for angina (Section 5.3.2.2). Among those hospitalized,
8 ST-segment decreases are considered a nonspecific marker of myocardial ischemia. A
9 small number of epidemiologic panel studies have reported associations between
10 short-term exposure to NO2 and ST-segment changes on the electrocardiogram of older
11 adults with a history of coronary artery disease (Section 5.3.2.3).
12 Coherent with the increase in hospital admissions and ED visits for IHD, MI, and angina,
13 single-city studies from the U.S. (Ito et al.. 2011: Peel et al.. 2007: Tolbert et al.. 2007)
14 and multicity studies conducted in Europe and Australia and New Zealand (Larrieu et al..
15 2007: Ballester et al.. 2006: Barnett et al.. 2006: Von Klot et al.. 2005) report positive
16 associations with all CVD hospital admissions in adults with adjustment for numerous
17 potential confounding factors, including weather and time trends (Section 5.3.9).
18 Additionally, the evidence for associations observed in time-series studies is coherent
19 with positive associations reported in epidemiologic studies of short-term NO2 exposure
20 and cardiovascular mortality in adults (Section 5.3.10). These include studies reviewed in
21 the 2008 ISA for Oxides of Nitrogen and recent multicity studies that generally report a
22 similar or slightly larger magnitude for the NO2 cardiovascular mortality relationship
23 compared to total mortality.
24 Recent controlled human exposure and animal toxicological studies provide preliminary
25 evidence for a potentially biologically plausible mechanism for short-term exposure to
26 NO2 leading to cardiovascular disease, including IHD. Reactive intermediates or
27 inflammatory mediators that have "spilled over" from the respiratory tract into the
28 circulation may result in systemic inflammation and/or oxidative stress, which may
29 mediate effects in the heart and vasculature (Sections 4.3.2.9 and 4.3.5). These
30 nonspecific effects may promote the triggering of an MI. There is limited and supportive
31 evidence in humans and animals for increased systemic and tissue specific oxidative
32 stress (Channell et al.. 2012; Li et al.. 201 la). In addition, evidence in animal and cell
33 models and in some controlled human exposure studies report NO2-mediated increases in
34 inflammatory markers (Channell et al.. 2012; Huang et al.. 2012b; Li et al.. 201 la).
35 A key uncertainty that remains since the 2008 ISA for Oxides of Nitrogen is the potential
36 for confounding by other correlated traffic-related pollutants given a common source and
37 moderate to high correlations with NO2. Recent studies have evaluated this uncertainty
January 2015 5-324 DRAFT: Do Not Cite or Quote
-------
1 using copollutant models and comparing associations of NO2 with those of other
2 pollutants. A number of studies examined associations between short-term exposure to
3 NO2 and cardiovascular disease adjusted for PMio or TSP (Figure S5-2. (U.S. EPA.
4 2014b)), CO (Figure S5-3. (U.S. EPA. 2014c)). O3 (Figure S5-4: (U.S. EPA. 2014(1)).
5 and SC>2 (Figure S5-5; (U.S. EPA. 2014eVK and reported that the effect estimate was
6 generally robust to the inclusion of the copollutant in the models. However, not all
7 analyses reported NO2 as the strongest predictor of cardiovascular effects. One study
8 reported that associations with cardiovascular hospital admissions were not robust in
9 models adjusting for CO exposure (Barnett et al., 2006) and another reported associations
10 with CO, total carbon, and EC and OC components of PM2 5 that were stronger or similar
11 in magnitude to those for NO2 (Tolbert et al., 2007). However, other traffic-related
12 pollutants that may be potentially correlated with NO2 (i.e., PIVb 5, EC, BC, VOCs) were
13 generally not examined in copollutant models, resulting in the potential for unmeasured
14 confounding. A limited number of studies that examined copollutant confounding on the
15 NO2 cardiovascular mortality relationship indicate that associations remain robust to
16 adjustment for PMio, SO2, or O3 (Chenetal.. 2012b: Chiusolo etal.. 2011). Finally, while
17 copollutant models are a common statistical tool used to evaluate the potential for
18 copollutant confounding, inferences from their results can be limited (Section 5.1.2.2).
19 Until more reliable methods to adjust for multiple copollutants simultaneously become
20 available, there is potential for residual confounding due to unmeasured copollutants
21 (Section 5.1.2.2). Without consistent and reproducible experimental evidence that is
22 coherent with the effects observed in epidemiologic studies, uncertainty still exists
23 concerning the role of correlated pollutants in the associations observed with NO2.
24 Additionally, the lack of studies with copollutant models evaluating PIVbs, BC/EC, UFPs,
25 or VOCs in relation to NO2 raises the concern that these associations could be a result of
26 NO2 serving as a marker for effects of other traffic-related pollutants or mixtures of
27 pollutants.
5.3.12.2 Evidence on Other Cardiovascular and Related
Metabolic Effects
28 There is inconclusive evidence from epidemiologic, controlled human exposure, and
29 animal toxicological studies for other cardiovascular and related metabolic effects from
30 short-term exposure to NO2. There are a number of epidemiologic studies that provide
31 inconsistent evidence for an association between 24-h avg NO2, NO, or NOx and risk of
32 cardiac arrhythmias as examined in patients with ICDs, continuous ECG recordings,
33 out-of-hospital cardiac arrest, and hospital admissions (Section 5.3.3). Similarly,
34 epidemiologic studies provide inconsistent evidence for a potential association between
35 ambient NO2 concentrations and risk of hospital admission for cerebrovascular disease
January 2015 5-325 DRAFT: Do Not Cite or Quote
-------
1 and stroke (Section 5.3.4). Both epidemiologic and controlled human exposure studies
2 provide little to no evidence to indicate that short-term exposure to ambient NO2 is
3 associated with increased BP or hypertension (Section 5.3.6). Other outcomes have an
4 insufficient quantity of studies to evaluate the effects. A small number of epidemiologic
5 studies have found associations between NO2 concentrations and hospital admissions or
6 ED visits for heart failure (Section 5.3.5) and hospital admission for venous thrombosis
7 and pulmonary embolism (Section 5.3.7). One recent epidemiologic study reported a lack
8 of an association between 24-h avg NO2 and insulin resistance (Section 5.3.8).
9 Various subclinical effects have been investigated that are not clearly associated with a
10 particular clinical event observed in the population but may be key events within a mode
11 of action for cardiovascular effects other than MI. There is limited evidence from
12 epidemiologic and controlled human exposure studies to suggest that NO2 exposure
13 results in alterations of cardiac autonomic control. Recent epidemiologic studies
14 generally reported associations between ambient NO2 levels and decreases in indices of
15 HRV (Section 5.3.11.1) and changes in ventricular repolarization (Section 5.3.11.2)
16 among populations with pre-existing or at elevated risk for cardiovascular disease.
17 Experimental studies also evaluated changes in HRV and ventricular repolarization
18 parameters. Although changes were not observed across all endpoints, a recent controlled
19 human exposure study reported decreased HFn and QTVI in healthy exercising adults
20 exposed to NO2, indicating a potential disruption in the normal cardiac autonomic control
21 (Huang et al., 2012b). However, similar measures of autonomic control in another
22 controlled human exposure study showed statistically nonsignificant increases after
23 exposure to NO2 (Scaife etal.. 2012).
5.3.12.3Conclusion
24 In conclusion, consistent epidemiologic evidence from multiple studies at relevant NO2
25 concentrations is suggestive of, but not sufficient to infer, a causal relationship between
26 short-term NO2 exposure and cardiovascular health effects. The strongest evidence
27 supporting this determination comes from studies of triggering an MI. However,
28 uncertainty remains regarding exposure measurement error and potential confounding by
29 traffic-related copollutants. Experimental studies provide some evidence describing key
30 events within the mode of action but do not provide evidence that is coherent with the
31 epidemiologic studies to help rule out chance, confounding, and other biases. Evidence
32 for other cardiovascular and related metabolic effects is inconclusive, including effects on
33 cardiac arrest and arrhythmia, cerebrovascular disease and stroke, increased blood
34 pressure and hypertension, decompensation of heart failure, and diabetes. Studies of
35 adults consistently demonstrate NO2-associated hospital admissions and ED visits for
January 2015 5-326 DRAFT: Do Not Cite or Quote
-------
1 IHD, MI, and angina, as well as all cardiovascular diseases. This is coherent with
2 evidence for NC^-related ST segment decrements and mortality from cardiovascular
3 disease. These studies have been replicated by different researchers in different locations
4 and have adjusted for numerous potential confounding factors including meteorological
5 factors and time trends. However, due to limited analysis of potentially correlated
6 pollutants and recognized limitations of copollutant models, some uncertainty remains
7 regarding the extent to which NC>2 is independently associated with cardiovascular effects
8 or if NC>2 serves as a marker for the effects of another traffic-related pollutant or mix of
9 pollutants. Thus, the combined evidence from epidemiologic and experimental studies is
10 suggestive of, but not sufficient to infer, a causal relationship between short-term NCh
11 exposure and cardiovascular effects.
January 2015 5-327 DRAFT: Do Not Cite or Quote
-------
Table 5-58 Summary of evidence, which is suggestive, but not sufficient, to infer
a causal relationship between short-term nitrogen dioxide (NO2)
exposure and cardiovascular and related metabolic effects.
Rationale for
Causal
Determination3 Key Evidence13
NO2
Concentrations
Associated with
Key References'3 Effects0
Triggering a myocardial infarction
Consistent
epidemiologic
evidence from
multiple, high-
quality studies at
relevant NO2
concentrations
Increases in hospital admissions and
ED visits for IHD and Ml in adults in
multiple studies, including multicity
studies, in diverse locations.
Larrieu et al. (2007):
Stieb et al. (2009):
Peel et al. (2007):
Von Klot et al. (2005): Mann et al.
(2002)
Figure 5-13: Section 5.3.2.1
Mean 24-h avg:
11.9-37.2 ppb
Mean 1-h max:
45.9 ppb
Coherence with limited evidence for
increased hospital admissions and ED
visits for angina in adults in multiple
studies, including multicity studies.
Szvszkowicz (2009):
Poloniecki et al. (1997):
Von Klot et al. (2005)
Section 5.3.2.2
Increases in hospital admissions and Larrieu et al. (2007):
ED visits for all CVD in adults in
multiple studies, including multicity
studies, in diverse locations.
Itoetal. (2011):
Peel et al. (2007):
Tolbertetal. (2007):
Von Klot et al. (2005): Ballester et
al. (2006): Barnett et al. (2006)
Section 5.3.9
Mean 24-h avg:
11.9-40.5 ppb
Mean 1-h max:
43.2-45.9 ppb
Coherence with ST-segment
depression in adults with pre-existing
coronary heart disease in association
with 24-h avg and 1-h avg NO2.
Chuanq et al. (2008): Delfino et al.
(2011)
Section 5.3.2.3
24 h avg:
21.4 ppb
Mean 1-h max:
27.5 ppb
Consistent evidence for increased risk
of cardiovascular mortality in adults
applying differing model specifications
in diverse locations.
Bellini etal. (2007):
Wong et al. (2008): Chen et al.
(2012b): Chiusoloetal. (2011)
Section 5.3.10
Mean 24-h avg:
13.5-35.4 ppb
Uncertainty
regarding
exposure
measurement
error
Majority of evidence from time-series
studies that rely on central site
exposure estimates.
January 2015
5-328
DRAFT: Do Not Cite or Quote
-------
Table 5-58 (Continued): Summary of evidence, which is suggestive, but not
sufficient, to infer a causal relationship between short-
term nitrogen dioxide (NO2) exposure and cardiovascular
and related metabolic effects.
Rationale for
Causal
Determination3
Key Evidence13
Key References'3
NO2
Concentrations
Associated with
Effects0
Uncertainty
regarding
potential
confounding by
traffic-related
copollutants
NO2 associations with ED visits and Supplemental Figures S5-2, S5-3,
hospital admissions are generally S5-4, and S5-5 (U.S. EPA, 2014b,
robust in copollutant models containing c, d_, e)
PM-io, CO, 03, orSO2.
Inability to disentangle the effects of
traffic-related pollutants because of
lack of examination (e.g., PlVhs,
BC/EC, UFPs, orVOCs).
NO2 associations with ED visits,
hospital admissions, and mortality
found with adjustment for numerous
potential confounding factors including
meteorological factors and time trends.
Some evidence for key events within the mode of action
Oxidative stress Limited and supportive evidence of
increased oxidative stress in heart
tissue in rats with relevant NO2
exposures (i.e., MDA) and plasma
from NO2-exposed humans
(i.e., LOX-1).
Lietal. (2011 a)
Section 4.3.2.9, Figure 4-3
Rats: 5,320 ppb
but not 2,660 ppb
NO2
Inflammation Limited and supportive toxicological
evidence of increased transcription of
some inflammatory mediators in vitro
(i.e., IL-8, ICAM-1, VCAM-1) and in
rats (i.e., ICAM-1, TNF-a).
Channelletal. (2012)
Human cells
exposed to
plasma from
healthy adults:
500 ppb NO2
Lietal. (2011 a)
Rats: 2,660 and
5,320 ppb NO2
Limited and inconsistent evidence in
controlled human exposure studies
(i.e., IL-6, IL-8, ICAM-1).
Huang et al. (2012b): Riedl et al. Adults: 350,
(2012) 500 ppb NO2
Inconsistent epidemiologic evidence
for changes in CRP, IL-6, and TNF-RII.
Section 5.3.11.4
January 2015
5-329
DRAFT: Do Not Cite or Quote
-------
Table 5-58 (Continued): Summary of evidence, which is suggestive, but not
sufficient, to infer a causal relationship between short-
term nitrogen dioxide (NO2) exposure and cardiovascular
and related metabolic effects.
Rationale for
Causal
Determination3
Key Evidence13
Key References'3
NO2
Concentrations
Associated with
Effects0
Other cardiovascular and related metabolic effects
Inconclusive
evidence from
epidemiologic,
controlled human
exposure and
toxicological
studies
Inconsistent epidemiologic evidence
for an association between NO2, NO,
or NOx and risk of cardiac arrest and
arrhythmias, cerebrovascular disease
and stroke, and increased blood
pressure and hypertension.
Sections 5.3.3. 5.3.4. and 5.3.6
Insufficient quantity of studies
evaluating decompensation of heart
failure and venous thrombosis and
pulmonary embolism.
Stieb et al. (2009):
Yang (2008)
Section 5.3.5
Dales etal. (2010)
Section 5.3.7
Lack of an association between
24-h avg NO2 and diabetes (i.e.,
insulin resistance).
Kelishadi et al. (2009)
Section 5.3.8
Inconsistent changes in HRV in
controlled human exposure studies.
Huang etal. (2012b)
Healthy adults:
500 ppb NO2
Scaifeetal. (2012)
Section 5.3.11.1
Adults with
pre-existing CVD:
400 ppb NO2
Limited epidemiologic evidence for
changes in HRV and ventricular
repolarization.
Stronger associations observed in
groups of individuals with pre-existing
cardiovascular disease.
HRV:
Timonen et al. (2006):
Suh and Zanobetti (201 Ob):
Zanobetti et al. (2010)
Section 5.3.11.1
QT interval:
Henneberqer et al. (2005)
Section 5.3.11.2
CO = carbon monoxide, CRP = C-reactive protein, CVD = cardiovascular disease, EC = elemental carbon, HRV = heart rate
variability, ICAM-1 = inter-cellular adhesion molecule 1, IHD = ischemic heart disease, IL = interleukin, MDA = malondialdehyde,
Ml = myocardial infarction, NO = nitric oxide, NO2 = nitrogen dioxide, O3 = ozone, PM = particulate matter, SO2 = sulfur dioxide,
UFP = ultrafine particles, VCAM-1 = vascular adhesion molecule-1.
aBased on aspects considered in judgments of causality and weight of evidence in causal framework in Tables I and N. of the
Preamble.
""Describes the key evidence and references, supporting or contradicting, that contribute most heavily to causal determination.
References to earlier sections indicate where full body of evidence is described.
°Describes the NO2 concentrations with which the evidence is substantiated.
January 2015
5-330
DRAFT: Do Not Cite or Quote
-------
5.4 Total Mortality
5.4.1 Introduction and Summary of 2008 Integrated Science Assessment for
Oxides of Nitrogen
1 Prior to the 2008 ISA for Oxides of Nitrogen, epidemiologic studies had not been
2 identified that examined whether an association exists between short-term NO2 exposure
3 and mortality. The 2008 ISA for Oxides of Nitrogen evaluated a collection of studies,
4 including multicity studies, conducted in the U.S., Canada, and Europe, and a
5 meta-analysis (U.S. EPA. 2008a). All of these studies reported evidence of an association
6 between short-term NO2 exposure and mortality with estimates ranging from 0.5 to 3.6%
7 for a 20-ppb increase in 24-h avg or 30-ppb increase in 1-h max NO2 concentrations. A
8 limitation of this collection of studies was that the majority focused specifically on PM
9 and did not conduct extensive analyses to examine the relationship between short-term
10 NO2 exposure and mortality.
11 Multicity studies conducted in the U.S. (HEI. 2003). Canada (Brook et al.. 2007; Burnett
12 et al.. 2004) and Europe (Samoli et al.. 2006). as well as a large study conducted in the
13 Netherlands (Hoek. 2003). consistently reported positive associations between short-term
14 NO2 exposure and mortality, specifically at lag 1, with evidence that these associations
15 remain robust in copollutant models. These results were confirmed in a meta-analysis that
16 did not include any of the aforementioned multicity studies (Stieb et al.. 2002).
17 Of the studies evaluated in the 2008 ISA for Oxides of Nitrogen, a limited number
18 provided additional information (i.e., seasonal analyses, examination of cause-specific
19 mortality, examination of effect modifiers) on the NO2-mortality relationship. Initial
20 evidence indicated a larger NO2-mortality association during the warmer months (Brook
21 et al.. 2007; Burnett et al.. 2004; HEI. 2003). Additionally, an examination of total and
22 cause-specific mortality found associations similar in magnitude across mortality
23 outcomes (total, respiratory, and cardiovascular); however, some studies reported
24 stronger NO2 associations for respiratory mortality (Biggeri et al.. 2005; Simpson et al..
25 2005b). Potential effect modifiers of the NO2-mortality relationship were examined only
26 within the APHEA study, which found that within the European cities, geographic area
27 and smoking prevalence modified the NO2-mortality relationship. It is worth noting that
28 additional multicity European studies that focused on PM (Agaetal.. 2003; Katsouyanni
29 et al.. 2003) reported that cities with higher NO2 concentrations also had higher PM risk
30 estimates indicating that NO2 and PM may be potential effect modifiers of each other.
January 2015 5-331 DRAFT: Do Not Cite or Quote
-------
1 In summary, the multicity studies evaluated in the 2008 ISA for Oxides of Nitrogen
2 consistently observed positive associations between short-term NO2 exposure and
3 mortality. These studies indicated that associations were found to occur within the first
4 few days after exposure and are potentially influenced by season. However, uncertainties
5 remained in the NO2-mortality relationship, which led to the 2008 ISA for Oxides of
6 Nitrogen (U.S. EPA. 2008a) concluding that the evidence "was suggestive but not
7 sufficient to infer a causal relationship." These uncertainties and data gaps included
8 whether: NO2 is acting as an indicator for another pollutant or a mix of pollutants; there is
9 evidence for potential copollutant confounding; specific factors modify the
10 NO2-mortality relationship; there is seasonal heterogeneity in mortality associations; NO2
11 associations are stronger with specific mortality outcomes; and the shape of the
12 NO2-mortality concentration-response relationship is linear.
5.4.2 Associations between Short-Term Nitrogen Dioxide Exposure and Mortality
13 Since the completion of the 2008 ISA for Oxides of Nitrogen, the body of epidemiologic
14 literature that has examined the association between short-term NO2 exposure and
15 mortality has grown. However, similar to the collection of studies evaluated in the 2008
16 ISA for Oxides of Nitrogen, most of the recent studies did not focus specifically on the
17 NO2-mortality relationship but on other pollutants. Of the studies identified, a limited
18 number have been conducted in the U.S., Canada, and Europe, with the majority being
19 conducted in Asia due to the increased focus on examining the effect of air pollution on
20 health in developing countries. Although these studies are informative in evaluation of
21 the relationship between oxides of nitrogen and mortality, the broad implications of these
22 studies in the context of results from studies conducted in the U.S., Canada, and Western
23 Europe are limited. This is because studies conducted in Asia encompass cities with
24 meteorological (Tsai et al.. 2010; Wong et al.. 2008). outdoor air pollution (e.g.,
25 concentrations, mixtures, and transport of pollutants), and sociodemographic (e.g.,
26 disease patterns, age structure, and socioeconomic variables) (Kanet al.. 2010)
27 characteristics that differ from cities in North America and Western Europe, potentially
28 limiting the generalizability of results from these studies to other cities.
29 Overall, this section evaluates studies that examined the association between short-term
30 NO2 exposure and mortality and addresses the key uncertainties and data gaps in the
31 NO2-mortality relationship identified in the 2008 ISA for Oxides of Nitrogen: potential
32 confounding of NO2 associations, effect modification (i.e., sources of heterogeneity in
33 risk estimates across cities), seasonal heterogeneity in NO2 associations, and the
34 NO2-mortality C-R relationship. Other recent studies of mortality are not the focus of this
35 evaluation because they were conducted in small single-cities, encompass a short study
January 2015 5-332 DRAFT: Do Not Cite or Quote
-------
1 duration, had insufficient sample size, and/or did not examine potential copolluant
2 confounding. The full list of the studies can be found in Supplemental Table S5-4
3 (U.S. EPA. 20141).
5.4.3 Associations between Short-term Nitrogen Dioxide Exposure and Mortality
in All-Year Analyses
4 Multicity studies evaluated in the 2008 ISA for Oxides of Nitrogen reported consistent,
5 positive associations between short-term NC>2 exposure and mortality in all-year analyses
6 (U.S. EPA. 2008a). However, when focusing on specific causes of mortality, some
7 studies reported similar risk estimates across total (nonaccidental), cardiovascular, and
8 respiratory mortality (Samoli et al.. 2006; Burnett et al.. 2004). while others indicated
9 larger respiratory mortality risk estimates compared to both total and cardiovascular
10 mortality (Atkinson etal.. 2012; Biggeri et al.. 2005; Simpson et al.. 2005b). Additional
11 multicity studies focusing on COPD (Meng etal.. 2013) and stroke (Chen etal.. 2013)
12 mortality further support potential differences in the NC^-mortality association by
13 mortality outcome. Although only a small number of multicity studies have been
14 conducted since the completion of the 2008 ISA for Oxides of Nitrogen, these studies
15 build upon and provide additional evidence for an association between short-term NO2
16 exposure and total mortality along with potential differences by mortality outcome. Air
17 quality characteristics and study specific details for the studies evaluated in this section
18 are provided in Table 5-59.
January 2015 5-333 DRAFT: Do Not Cite or Quote
-------
Table 5-59 Air quality characteristics of studies evaluated in the 2008 Integrated Science Assessment for
Oxides of Nitrogen and recently published multicity and select single-city studies.
Study
Bigger! et al. (2005)
Brook etal. (2007)
Burnett et al. (2004)
HEI (2003)
Hoek (2003)
Samoli et al. (2006)
Simpson et al. (2005b)
Stieb et al. (2003)
tAtkinson etal. (2012)
Location (Years)
8 Italian cities
(1990-1999)
10 Canadian cities
(1984-2000)
12 Canadian cities
(1981-1999)
58 U.S. cities3
(1987-1994)
the Netherlands
(1986-1994)
30 European cities
(1990-1997)
4 Australian cities
(1996-1999)
Meta-analysis,
worldwide
(Years NR)
Meta-analysis,
Asia
(Years NR)
Mortality
Outcome(s)
Total,
cardiovascular,
respiratory
Total
Total,
cardiovascular,
respiratory
Total
Total
Total,
cardiovascular,
respiratory
Total,
cardiovascular,
respiratory
Total
Total,
cardiovascular,
respiratory
Averaging
Exposure Assignment Time
Average of NO2 concentrations across all 24-h avg
Monitors influenced by local traffic
excluded.
Average of NO2 concentrations across all 24-h avg
monitors in each city.
Average of NO2 concentrations across all 24-h avg
monitors in each city.
Average of NO2 concentrations across all 24-h avg
monitors in each city.
15 NO2 monitors across the study area, 24-h avg
mean concentration calculated in each
region then weighted by population
density in each region.
Average of NO2 concentrations across all 1-h maxb
monitors in each city.
Average of NO2 concentrations across all 1-h max
monitors in each city.
NA NR
NA NR
Mean Upper Percentile
Concentration Concentrations
ppb ppb
30.1-55.0 95th: 45.8-94.0
Max: 62.6-160.7
NR NR
10.0-26.4 NR
9.2-39.4 NR
NR NR
24.0-80.5 90th: 33.1-132.5
-100007 iv/i-av QP n -i -I -I c;
NR NR
NR NR
January 2015
5-334
DRAFT: Do Not Cite or Quote
-------
Table 5-59 (Continued): Air quality characteristics of studies evaluated in the 2008 Integrated Science
Assessment for Oxides of Nitrogen and recently published multicity and select single-
city studies.
Study
tBellini et al. (2007)
tBerqlindetal.(2009)
tChenetal.(2012b)
tChenetal.(2013)
tChiusolo et al. (2011)
tKanetal. (2010): Kan et al.
(2008)
tFaustini et al. (2013)
tltoetal. (2011)
tSacksetal. (2012)
tMenqetal. (2013)
tMoolgavkar et al. (2013)
tShinetal. (2012)
Location (Years)
15 Italian cities
(1996-2002)
5 European cities
(1992-2002)
17 Chinese cities
(1 996-201 Oc)
8 Chinese cities
(1996-2008d)
10 Italian cities8
(2001-2005)
Shanghai, China
(2001-2004)
6 Italian cities
(2001-2005)
New York, NY
(2000-2006)
Philadelphia, PA
(1992-1995)
4 Chinese cities
(1996-2008C)
72 U.S. cities^
(1987-2000)
24 Canadian cities
(1984-2004)
Mortality
Outcome(s)
Total,
cardiovascular,
respiratory
Total
Total,
cardiovascular,
respiratory
Stroke
Total,
cardiovascular,
cerebrovascular,
respiratory
Total,
cardiovascular,
respiratory
Respiratory
(out-of-hospital)
Cardiovascular
Cardiovascular
COPD
Total
Cardiopulmonary
Exposure Assignment
NR
Average of NO2 concentrations across all
monitors in each city.
Average of NO2 concentrations across all
Average of NO2 concentrations across all
If more than 1 monitor, average of NO2
concentrations across all monitors in
each city (1-5 monitors).
Average of NO2 concentrations across
6 monitors.
Average of NO2 concentrations over all
Average of NO2 concentrations across all
monitors.
Central site monitor.
Average of NO2 concentrations across all
Average of NO2 concentrations across all
monitors in each city.
If more than 1 monitor, average of NO2
concentrations across all monitors in
each city (1-8 monitors).
Averaging
Time
24-h avg
24-h avg
24-h avg
24-h avg
24-h avg
24-h avg
24-h avg
24-h avg
1-h max
24-h avg
24-h avg
24-h avg
Mean
Concentration
ppb
NR
-I -I n oc A
13.5-34.8
19.7-35.6
-i o Q oc n
35.4
•~)A C, QC, -1
28.7
47.4
30.6-35.4
NR
8.7-25.0
Upper Percentile
Concentrations
ppb
NR
NR
Max: 55.1-132.1
NR
90th: 21. 7-48.8
75th: 42.1
Max: 134.9
NR
NR
Max: 146.7
NR
NR
NR
January 2015
5-335
DRAFT: Do Not Cite or Quote
-------
Table 5-59 (Continued): Air quality characteristics of studies evaluated in the 2008 Integrated Science
Assessment for Oxides of Nitrogen and recently published multicity and select single-
city studies.
Study
tStieb et al. (2008)
tWonqetal. (2010); Wonqet
al. (2008)
Location (Years)
12 Canadian cities
(1981-2000)
4 Asian cities
(1996-2004h)
Mortality
Outcome(s)
Total
Total
cardiovascular
respiratory
Mean Upper Percentile
Averaging Concentration Concentrations
Exposure Assignment Time ppb ppb
If more than 1 monitor, average of NO2 3-h max
concentrations across all monitors in
each city.
Average of NO2 concentrations across all 24-h avg
1981-1990:
24.7-42.6
1991-2000:
16.3-39.2
23.2-34.6
NR
75th: 28.5-41. 2
Max: 72.6-131.9
COPD - chronic obstructive pulmonary disease, NA = not available, NO2 = nitrogen dioxide, NR = not reported.
aOf the 90 cities included in the NMMAPS analysis only 58 had NO2 data.
bSamoli et al. (2006) estimated 1-h max concentrations for each city by multiplying 24-h avg concentrations by 1.64.
°Study period varied for each city and encompassed 2 to 7 yr. Hong Kong was the only city that had air quality data prior to 2000.
dThese monitors were "mandated to not be in the direct vicinity of traffic or of industrial sources, and not be influenced by local pollution sources, and to avoid buildings, or those
housing large emitters, such as coal-, waste-, or oil-burning boilers, furnaces, and incinerators" (Chen etal.. 2013: Meng et al.. 2013: Chenetal.. 2012b).
eOnly 9 cities (Cagliari was excluded) were included in the formal analysis of examining potential factors that could increase the risk of mortality due to short-term NO2 exposure.
'Information on the monitors used in this study were obtained from Colaisetal. (2012).
9Of the 108 cities included in the analyses using NMMAPS data only 72 had NO2 data.
The study period varied for each city, Bangkok: 1999-2003, Hong Kong: 1996-2002, and Shanghai and Wuhan: 2001-2004.
fStudies published since the 2008 ISA for Oxides of Nitrogen.
January 2015
5-336
DRAFT: Do Not Cite or Quote
-------
1
2
o
6
4
5
6
7
8
9
10
11
12
As demonstrated in Figure 5-22 and Table 5-60 multicity studies evaluated in the 2008
ISA for Oxides of Nitrogen and those recently published, consistently provide evidence
of positive associations between short-term NO2 exposure and total (non-accidental)
mortality. In these multicity studies, the associations observed were in analyses that
primarily examined all ages, the exceptions being Chiusolo et al. (2011) and Berglind et
al. (2009). who both focused on the risk of mortality attributed to air pollution in the
population >35 years of age. Across these studies, associations between short-term NCh
exposure and mortality were examined primarily in the total population; however,
Berglind et al. (2009) focused on a subset of the population, i.e., MI survivors. The large
effect estimate for Berglind et al. (2009) could be attributed to the larger mortality rate
for MI survivors: 30-day mortality rate of 14-15% and 1-year mortality rate of 22-24%,
compared to populations examined in the other multicity studies (Berglind et al.. 2009).
Study
Dominici et al. (2003)
Stieb et al. (2003)
Samoli et al. (2006)
Burnett et al. (2004)
Hoek (2003)
Simpson et al. (2005)
Brook etal. (2007)
Biggen et al. (2005)
Stieb et al. (2008)
Moolgavkar et al. (201 3)
Bellini et al. (2007)
Atkinson et al. (201 2)
Wong et al. (2008)
Chiusolo et al. (201 1 )
Berglind et al. (2009)
Location
58 U.S. cities
Meta-analysis (Worldwide)
30 European cities
1 2 Canadian cities
Netherlands
4 Australian cities
1 0 Canadian cities
8 Italian cities
1 2 Canadian cities
7 2 U.S. cities
15 Italian cities
Meta-Analysis (Asia)
4 Asian cities
17 Chinese cities
10 Italian cities
5 European cities
Lag
1
0-1
0-2
0-6
n i
U-l
0-1
1
1
0-1
n i
U-l
n i
U-l
0 5
0-1 "^
-•-
-•-
-*-
•
— • —
— o —
•
— • —
• fc
w°
4.0 6.0 8.0 10.0
% Increase
14.0 16.0
Note: Results are presented for per a 20-ppb increase in 24-h avg nitrogen dioxide concentrations or a 30-ppb increase in 1-h max
nitrogen dioxide concentrations. Black = studies reviewed in the 2008 Integrated Science Assessment for Oxides of Nitrogen,
Red = recent studies.
Figure 5-22 Summary of multicity studies evaluated in the 2008 Integrated
Science Assessment for Oxides of Nitrogen and recently
published studies that examined the association between
short-term nitrogen dioxide exposure and total mortality.
January 2015
5-337
DRAFT: Do Not Cite or Quote
-------
Table 5-60 Corresponding percentage increase in total mortality (95% Cl) for
Figure 5-22
Study
Dominicietal. (2003)
Stieb et al. (2003)
Samoli et al.
(2006)
Burnett et al. (2004)
Hoek (2003)
Simpson et al. (2005b)
Brook etal. (2007)
Biqqeri et al.
tStiebetal.
(2005)
(2008)
tMoolgavkar etal. (2013)
tBellini et al.
tAtkinson et
tWonq et al.
tChenetal.
tChiusolo et
. (2007)
al. (2012)
(2008)
(201 2b)
al. (2011)
tBerglind etal. (2009)
1
2
3
4
5
6
7
8
Location
58 U.S. cities
Meta-analysis
(worldwide)
30 European
cities
12 Canadian
cities
the Netherlands
4 Australian cities
10 Canadian
cities
8 Italian cities
12 Canadian
cities
72 U.S. cities
15 Italian cities
Meta-analysis
(Asia)
4 Asian cities
17 Chinese cities
10 Italian cities
5 European cities
Age Lag Averaging Time
All
All
All
All
All
All
All
All
All
All
All
All
All
All
>35
>35
1
—
0-1
0-2
0-6
0-1
1
0-1
1
1
0-1
—
0-1
0-1
0-5
0-1
Cl = confidence interval.
fStudies published since the 2008 ISA for Oxides of Nitrogen.
When focusing on cause-specific mortality, recem
patterns of associations to those evaluated in the 2
some evidence of larger respiratory mortality risk
However, in a study of 15 Italian cities, Bellini et
cardiovascular and respiratory mortality risk estin
contradicts the results of Biaaeri et al. (2005) of v
extension. Additionally, the total mortality results
24-h avg
24-h avg
1-h max
24-h avg
24-h avg
1-h max
24-h avg
1-h max
3-h max
24-h avg
24-h avg
24-h avg
24-h avg
24-h avg
24-h avg
24-h avg
% Increase (95% Cl)
0.5(0.09,
0.
1
2
2
3
3
3
1.
2
2
3
4
6
8
.8
.0
.6
.4
.5
.6
9
.1
.2
.7
.7
.3
8.1
11.
6
(0.20
(1.3,
(1.1,
(1.2,
(1.1,
(1.4,
(2.3,
(0.80
(1.8,
(1.0,
(2.1,
(3.2,
(4.2,
(3.7,
(-5.9
0.
, 1
2.
2.
4.
5.
5.
5.
,2
2.
3.
5.
6.
8.
90)
.5)
2)
9)
0)
7)
5)
0)
.9)
3)
6)
4)
2)
4)
12.7
, 32.4)
t multeity studies have reported similar
!008 ISA for Oxides of Nitrogen with
estimates (Figure 5-23 and Table 5-61).
al. (2007) observed smaller
lates compared to total mortality, which
Mch Bellini et al. (2007) is an
of Bellini et al. (2007) are smaller in
magnitude than those observed in Biggeri etal. (2005).
January 2015
5-338
DRAFT: Do Not Cite or Quote
-------
Study
Samoli et al. (2006)
Burnett et al. (2004)
Simpson et al. (2005)
Biggeri et al. (2005)
Bellini et al. (2007)
Atkinson et al. (2012)
Wong et al. (2008)
Chen et al. (2012)
Chen etal. (2012)
Chen etal. (2013)
Chen et al. (2012)
Meng et al. (2013)a
Chiusolo et al. (2011 )b
Location
30 European cities (APHEA2)
12 Canadian cities
4 Australian cities
8 Italian cities (MISA-1)
15 Italian cities (MISA-2)
Meta-Analysis (Asia)
4 Asian cities (PAPA)
17 Chinese cities (CAPES)
17 Chinese cities (CAPES)
8 Chinese cities (CAPES)
17 Chinese cities (CAPES)
4 Chinese cities
Mortality
All
Cardiovascular
Respiratory
All
Cardiopulmonary
Respiratory
All
Cardiovascular
Respiratory
All
Cardiovascular
Respiratory
All
Cardiovascular
Respiratory
All
Cardiovascular
Respiratory
All
Cardiovascular
Respiratory
All
Cardiovascular
Stroke
Respiratory
COPD
All
Cardiac
Respiratory
Lag
0-1
0-1
0-1
0-2
0-2
0-1
0-1
0-1
0-1
0-1
NR
NU
NU
0-1
0-1
0-1
0-1
0-1
0-1
0-5
1-5
-•-
— ®
A
-2.0 0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0
% Increase
Note: Black symbols = multicity studies evaluated in the 2008 Integrated Science Assessment for Oxides of Nitrogen; Red
symbols = recent studies. Filled circle = total mortality; Crosshatch = cardiovascular mortality; Vertical lines = respiratory mortality.
a = Although the study was not part of the CAPES study, it included four of the cities also included in CAPES; b = Study focused on
individuals >35 years of age while the other studies focused on all ages.
Figure 5-23 Percentage increase in total, cardiovascular, and respiratory
mortality from multicity studies for a 20-ppb increase in
24-hour average or 30-ppb increase in 1-hour maximum nitrogen
dioxide concentrations.
January 2015
5-339
DRAFT: Do Not Cite or Quote
-------
Table 5-61 Corresponding percentage
Study
Samoli et al. (2006)
Burnett et al. (2004)
Simpson et al. (2005b)
Biqqeri et al. (2005)
tBellini et al. (2007)
tAtkinson et al. (2012)
tWonq et al. (2008)
tChenetal.(2012b)
tChenetal.(2013)
tChenetal.(2012b)
tMenqetal. (2013)
tChiusolo et al. (2011)
Location Age Lag
30 European All 0-1
cities
12 Canadian All 0-2
cities
4 Australian All 0-1
cities
8 Italian All 0-1
cities
15 Italian All 0-1
cities
Meta- All
analysis
(Asia)
4 Asian All 0-1
cities
17 Chinese All 0-1
cities
8 Chinese All 0-1
cities
17 Chinese All 0-1
cities
4 Chinese All 0-1
cities
10 Italian >35 0-5
cities
1-5
increase (95% CD for Figure 5-23.
Averaging
Time Mortality
1-h max Total
Cardiovascular
Respiratory
24-h avg Total
Cardiovascular
Respiratory
1-h max Total
Cardiovascular
Respiratory
1-h max Total
Cardiovascular
Respiratory
24-h avg Total
Cardiovascular
Respiratory
24-h avg Total
Cardiovascular
Respiratory
24-h avg Total
Cardiovascular
Respiratory
24-h avg Total
Cardiovascular
24-h avg Stroke
24-h avg Respiratory
24-h avg COPD
24-h avg Total
Cardiovascular
Respiratory
% Increase
(95% Cl)
1.8(1.3,2.2)
2.3(1.7, 3.0)
2.2(1.0, 3.4)
2.0(1.1,2.9)
2.0(0.5, 3.9)
2.1 (-0.3, 3.9)
3.4(1.1, 5.7)
4.3(0.9, 7.8)
11.4(3.4, 19.7)
3.5 (2.2, 4.9)
5.0(2.9, 7.1)
5.4(0.2, 11.0)
2.2(1.0, 3.6)
-1 c / -1 7 A r\\
1.4 (-2.4, 6.7)
3.7(2.1, 5.4)
4.1 (2.2, 6.0)
6.7(3.2, 10.3)
4.7 (3.2, 6.2)
5.2 (3.4, 7.0)
5.7 (2.6, 8.8)
6.3 (4.2, 8.4)
6.9(3.8, 10.1)
5.6 (3.4, 8.0)
9.8(5.5, 14.2)
7.1 (5.4, 8.9)
8.1 (3.7, 12.7)
10.3(5.9, 14.8)
13.7(2.9,25.8)
Cl = confidence interval, COPD = chronic obstructive pulmonary disease.
fStudies published since the 2008 ISA for Oxides of Nitrogen.
January 2015
5-340
DRAFT: Do Not Cite or Quote
-------
5.4.4 Potential Confounding of the Nitrogen Dioxide-Mortality Relationship
1 A key uncertainty of the NO2-mortality relationship identified in the 2008 ISA for Oxides
2 of Nitrogen (U.S. EPA. 2008a) was whether NO2 acts as a surrogate of another
3 unmeasured pollutant. As such, although the multicity studies evaluated in the 2008 ISA
4 for Oxides of Nitrogen reported consistent evidence of an association between short-term
5 NO2 exposure and mortality that persisted in copollutant models, these studies often
6 concluded that the observed mortality effects could not be attributed solely to NO2.
7 Copollutant analyses conducted in recent studies further attempted to identify whether
8 NO2 has an independent effect on mortality. Additionally, recent studies have examined
9 whether the extent of temporal adjustment employed adequately controls for the potential
10 confounding effects of season on the NO2-mortality relationship.
Copollutant Confounding
11 In the examination of the potential confounding effects of copollutants on the
12 NO2-mortality relationship, it is informative to evaluate whether NO2 risk estimates
13 remain robust in copollutant models, specifically with traffic-related pollutants (e.g.,
14 PM2 5, EC, CO), and whether NO2 modifies the effect of other pollutants. Recent
15 multicity studies examine the NO2-mortality relationship by taking into consideration
16 both of these aspects in different study designs and in different study locations (i.e., U.S.,
17 Canada, Europe, and Asia). However, copollutant analyses in these studies have not
18 explicitly focused on traffic-related pollutants, complicating the overall interpretation of
19 results regarding whether there is an independent effect of short-term NO2 exposures on
20 mortality.
21 In a study of 108 U.S. cities using data from the NMMAPS for 1987-2000 (of which 72
22 had NO2 data), Moolgavkar et al. (2013) used a sub-sampling approach where a random
23 sample of 4 cities was removed from the 108 cities over 5,000 bootstrap cycles to
24 examine associations between short-term air pollution concentrations and mortality. This
25 approach was used instead of the two-stage Bayesian hierarchical approach employed in
26 the original NMMAPS analysis, which assumes that city-specific risk estimates are
27 normally distributed around a national mean (Dominici et al.. 2003). In a single-pollutant
28 model using 100 degrees of freedom (~7 df/yr, which is consistent with NMMAPS) to
29 control for temporal trends, Moolgavkar et al. (2013) reported a 2.1% (95% CI: 1.8, 2.3)
30 increase in total (nonaccidental) mortality at lag 1 day for a 20-ppb increase in 24-h avg
31 NO2 concentrations. The single-pollutant result is larger in magnitude than that observed
32 in (Dominici et al.. 2003). which only included 58 cities in the NO2 analysis
January 2015 5-341 DRAFT: Do Not Cite or Quote
-------
1 (Figure 5-22). In a copollutant analysis with PMio, the NCh-mortality risk estimate was
2 relatively unchanged (1.9% [95% CI: 1.3, 2.4]), and similar to the copollutant results in
3 (Dominici etal.. 2003).
4 Stieb etal. (2008) reported results consistent with Moolgavkar et al. (2013) in a study
5 that focused on the development of a new air quality health index in Canada. Focusing on
6 lag day 1 and models using 10 df per year, Stieb et al. (2008) examined whether
7 copollutants confounded the single-pollutant results in both copollutant and
8 multipollutant models with CO, Os, PMio, PM2 5, and 862. However, the study did not
9 clearly identify which results pertained to which model. As stated previously in this ISA,
10 it is important to note that multipollutant models are difficult to interpret due to the
11 multicollinearity often observed between pollutants and as a result are not used to inform
12 upon whether there is evidence of copollutant confounding. In models using all available
13 data and limited to days with PM data the results of the copollutant and multipollutant
14 analyses conducted by Stieb et al. (2008) indicate that the NCh-mortality relationship
15 remain relatively unchanged when adjusted for other pollutants, including some
16 traffic-related pollutants (quantitative results not presented).
17 Additional studies conducted in Europe and Asia also provide evidence indicating that
18 NO2-mortality associations remain robust in copollutant models; however, these studies
19 have also not focused on traffic-related pollutants. Chiusolo etal. (2011) conducted a
20 multicity study of 10 Italian cities using a time-stratified, case-crossover approach as part
21 of the Italian Epi Air multicenter study "Air Pollution and Health: Epidemiological
22 Surveillance and Primary Prevention." The authors reported consistent, positive
23 associations for total and cause-specific mortality (i.e., cardiac, cerebrovascular, and
24 respiratory), ranging from an 8.1 to 13.7% increase for a 20-ppb increase in 24-hour NC>2
25 concentrations using an unconstrained distributed lag of 0-5 days (lag 1-5 days was used
26 for respiratory mortality). In copollutant analyses, NC>2 risk estimates remained robust in
27 models with PM in all-year analyses and with Os in analyses restricted to the summer
28 season (i.e., April-September) (Table 5-62).
January 2015 5-342 DRAFT: Do Not Cite or Quote
-------
Table 5-62 Percentage increase in total and cause-specific mortality for a
20-ppb increase in 24-hour average NO2 concentrations in single-
and co-pollutant models with PMio in all-year analyses or Os in
summer season analyses.
Mortality Season
All natural All-year
April-September
Cardiac All-year
April-September
Cerebrovascular All-year
April-September
Respiratory All-year
April-September
Model
NO2 (lag 0-5)
With PMio (lag 0-5)
NO2 (lag 0-5)
With Os (lag 0-5)
NO2 (lag 0-5)
With PMio (lag 0-5)
NO2 (lag 0-5)
With O3 (lag 0-5)
NO2 (lag 0-5)
With PMio (lag 0-5)
NO2 (lag 0-5)
With Os (lag 0-5)
NO2(lag 1-5)
With PMio (lag 0-5)
NO2(lag 1-5)
With Os (lag 0-5)
% Increase (95% Cl)
8.1 (3.7, 12.7)
7.5(1.9, 13.5)
17.8(12.3,23.6)
18.2(13.1,23.6)
10.3(5.9, 14.8)
10.1 (4.0, 16.4)
19.2(11.4,27.4)
18.8(10.7,27.5)
9.1 (-0.5, 19.7)
9.9 (-2.6, 24.1)
33.0(19.2,48.3)
30.2(13.9,48.8)
13.7(2.9,25.8)
13.4(2.9,24.8)
41.3(16.2, 71.7)
43.4(14.6, 79.5)
Note: Concentrations converted from |jg/m3 to ppb using the conversion factor of 0.532, assuming standard temperature (25°C)
and pressure (1 atm).
Cl = confidence interval, NO2 = nitrogen dioxide, O3 = ozone, PM = particulate matter.
Source: Reproduced with permission from Environmental Health Perspectives (Chiusolo et al.. 2011).
1
2
o
J
4
5
6
7
The Public Health and Air Pollution in Asia (PAPA) study as well as the CAPES
collectively found that the NCh-mortality association remains robust in copollutant
models with other criteria air pollutants in analyses conducted in Asian cities. The PAPA
study examined the effect of air pollution on mortality in four cities, one in Thailand (i.e.,
Bangkok) and three in China (i.e., Hong Kong, Shanghai, and Wuhan) (Wong et al..
2010; Wong et al.. 2008). In these study locations, PMio and SC>2 concentrations are
much higher than those reported in the U.S.; however, NC>2 and Os concentrations are
January 2015
5-343
DRAFT: Do Not Cite or Quote
-------
1 fairly similar (Wong et al.. 2010; Wong et al., 2008). Copollutant analyses were only
2 conducted in the individual cities; a combined four-city analysis was not conducted. In
3 models using lag 0-1 days NCh concentrations in the Chinese cities, NC>2 mortality risk
4 estimates were relatively unchanged in copollutant models (quantitative results not
5 presented). However, in Bangkok, the NCh-mortality risk estimate was attenuated in
6 models with PMio.
7 The results from the Chinese cities in the PAPA study are consistent with those found in
8 CAPES (Chenet al., 2012b). In a two-stage Bayesian hierarchical model, where the first
9 stage followed the PAPA protocol, Chen et al. (2012b) reported a 6.3% increase
10 (95% CI: 4.2, 8.4) in total mortality, 6.9% increase (95% CI: 3.8, 10.1) for cardiovascular
11 mortality, and 9.8% increase (95% CI: 5.5, 14.2) for respiratory mortality for a 20-ppb
12 increase in 24-h avg NCh concentrations at lag 0-1 days. Although NC>2 was moderately
13 correlated with both PMio and SC>2, 0.66 and 0.65, respectively, NCh-mortality
14 associations, although attenuated, remained positive across total, cardiovascular, and
15 respiratory mortality with the percentage increase in mortality ranging from 4.6-6.7% in
16 copollutant models with PMio and 5.2-7.0% in models with 862 for a 20-ppb increase in
17 24-h avg NC>2 concentrations.
18 In addition to examining whether copollutants confound the NCh-mortality relationship,
19 studies also conducted analyses to examine if there was any indication that NO2 modifies
20 the PM-mortality relationship. The Air Pollution and Health: A European and North
21 American Approach study, although it focused specifically on examining the
22 PMio-mortality relationship, also conducted an analysis to identify whether NC>2 modifies
23 the PMio-mortality relationship. In both the European and U.S. data sets, as mean NCh
24 concentrations and the NCVPMio ratio increased, there was evidence that the risk of PMio
25 mortality increased. These results are consistent with Katsouyanni et al. (2003) and
26 Katsouyanni et al. (2001). who reported higher PM risk estimates in cities with higher
27 NO2 concentrations, suggesting that NC>2 and PM may be effect modifiers of each other.
Temporal Confounding
28 Recent studies have also examined whether the NO2-mortality relationship is subject to
29 temporal confounding. These studies have focused on examining the effect of increasing
30 the number of df employed per year to control for temporal trends on NO2-mortality risk
31 estimates. Using the entire data set, which encompassed the years 1981-2000, Stieb et al.
32 (2008) examined the effect of using an alternative number of df to adjust for seasonal
33 cycles on NCh-mortality risk estimates. In analyses of single-day lags from 0 to 2 days in
34 single-pollutant models, the authors reported comparable risk estimates for each
35 individual lag day when using 6, 8, 10, 12, and 14 df per year. Similar to Stieb et al.
January 2015 5-344 DRAFT: Do Not Cite or Quote
-------
1 (2008). the PAPA study also examined the impact of alternative approaches to
2 controlling for temporal trends on mortality risk estimates. In models using 4, 6, 8, 10, or
3 12 df per year, Wong etal. (2010) also reported relatively similar results across the df per
4 year specified, with some evidence for a slight attenuation of the NCh-mortality
5 association in Wuhan, China as the df per year increased.
6 Unlike Stieb et al. (2008) and Wong etal. (2010). who conducted a systematic analysis of
7 the influence of increasing the df per year to control for temporal trends on the
8 NO2-mortality relationship, Moolgavkar et al. (2013) only compared models that used
9 50 df (~3.5 df per year) or 100 df (~7 df per year) in the statistical model. However,
10 similar to both Stieb et al. (2008) and Wong et al. (2010). Moolgavkar etal. (2013)
11 reported similar results regardless of the number of df used, 2.0% (95% CI: 1.8, 2.3) for a
12 20-ppb increase in 24-h avg NC>2 concentrations at lag 1 day in the 50 df model and 2.1%
13 (95% CI: 1.8, 2.3) in the 100 df model.
5.4.5 Modification of the Nitrogen Dioxide-Mortality Relationship
14 To date, a limited number of studies have examined potential effect measure modifiers of
15 the NO2-mortality relationship. In the 2008 ISA for Oxides of Nitrogen (U.S. EPA.
16 2008a). Samoli et al. (2006) provided evidence of regional heterogeneity in
17 NO2-mortality associations and higher NC^-mortality risk estimates in cities with a lower
18 prevalence of smoking as part of the APHEA-2 study. Recent multicity studies conducted
19 in Italy (Chiusolo etal.. 2011). Chile (Cakmak et al.. 20 lib), and Asia (Chen et al..
20 2012b) conducted extensive analyses of potential effect measure modifiers of the
21 NO2-mortality relationship and identified specific factors that may characterize
22 populations potentially at increased risk of NCh-related mortality (see Chapter 7). These
23 studies presented evidence indicating that older adults (>65 years of age), females,
24 individuals with pre-existing cardiovascular or respiratory diseases, and individuals of
25 lower SES, specifically lower income and educational attainment, are at greater risk. It
26 should be noted that demographic as well as socioeconomic differences between
27 countries may complicate the interpretation of results across these studies, and
28 subsequently the ability to make generalizations across locations regarding the factors
29 that may modify the NC^-mortality association.
January 2015 5-345 DRAFT: Do Not Cite or Quote
-------
5.4.6 Potential Seasonal Differences in the Nitrogen Dioxide-Mortality
Relationship
1 Studies evaluated in the 2008 ISA for Oxides of Nitrogen indicated seasonal differences
2 in the NO2-mortality relationship with evidence of larger associations in the warm or
3 summer season. Recent multicity studies conducted in Canada (Shin etal.. 2012: Stieb et
4 al.. 2008) and Italy (Chiusolo et al.. 2011: Bellini et al.. 2007) further support these
5 previous findings but also raise additional questions in light of the seasonal patterns in
6 NO2 concentrations observed in the U.S. and Canada (i.e., higher concentrations in the
7 winter months compared with the summer months) and the higher personal-ambient
8 relationship in the summer compared with the winter (Section 2.5.4).
9 In the 12 Canadian city study, Stieb et al. (2008) reported that NO2-mortality risk
10 estimates were larger in the warm season (April-September) compared with the cool
11 season (October-March) (quantitative results not presented). These results are consistent
12 with those reported by Shin et al. (2012) in a study that examined year-to-year changes in
13 the association between short-term NO2 exposure and mortality (i.e., cardiopulmonary
14 and non-cardiopulmonary) across 24 Canadian cities during 1984-2004. In seasonal
15 analyses, NO2 associations with cardiopulmonary mortality at lag 0-2 days were
16 observed to be stronger in the warm season (April-September) compared with the cold
17 season (October-March). Shin et al. (2012) suggest that the larger NO2 mortality effects
18 in the warm season could be due to the role of NO2 in the atmospheric reactions that form
19 Os, and subsequently suggests that the relationship between NO2 and Os does not allow
20 for a clear assessment of the independent effects of NO2. However, in Canada, as well as
21 the U.S., NO2 concentrations are higher in the cold season compared to the warm season.
22 Additionally, NO2 and Os are not well correlated during the summer (r ranging from 0.0
23 to 0.40), which makes it less likely Os is a confounder of the NO2-mortality relationship
24 (Section 3.4.4.1).
25 To date, U.S.-based multicity studies have not examined whether the seasonal patterns of
26 NO2-mortality associations observed in Canadian multicity studies are similar in the U.S.
27 However, a few single-city U.S.-based studies that focused on cardiovascular mortality
28 inform upon whether there is evidence of seasonal differences in NO2-total mortality
29 associations (Sacks et al.. 2012: Ito et al.. 2011). In a study conducted in New York City
30 that examined the association between short-term exposure to air pollution and
31 cardiovascular mortality, Ito etal. (2011) reported similar effect estimates in all-year
32 (1.8% [95% CI: 0.17, 3.3] for a 20-ppb increase in 24-h avg avg NO2 concentrations at
33 lag 1 d day) and seasonal (warm: 1.8% [95% CI: -0.4, 3.9]; cold: 2.3% [95% CI: 0.0,
34 4.7]) analyses. It should be noted that the study did not conduct copollutant analyses and
35 the NO2-mortality pattern of associations was similar to that observed for PM2 5 and EC.
January 2015 5-346 DRAFT: Do Not Cite or Quote
-------
1 Sacks etal. (2012) also examined potential seasonal differences in the
2 NO2-cardiovascular mortality association in a study conducted in Philadelphia, PA that
3 examined the influence of various approaches to control for seasonality and the potential
4 confounding effects of weather on the air pollution-cardiovascular mortality relationship.
5 Across models, the authors found that either: NCh-mortality associations were similar
6 between warm and cold seasons; or that associations were slightly larger in magnitude
7 during the warm season. These results suggest that the modeling approach employed may
8 influence the NCh-mortality associations observed, specifically with regard to whether
9 there is evidence of seasonal differences in associations, but the various approaches did
10 not influence the direction of the observed association.
11 Multicity studies conducted in Italy provide evidence consistent with that observed in the
12 Canadian multicity studies. In the MISA-2 study, Bellini et al. (2007) reported larger
13 NO2-mortality risk estimates in the summer (April-September) compared with the winter
14 (October-March) for total (6.4 vs. 0.9% for a 20-ppb increase in 24-h avg NO2
15 concentrations at lag 0-1 days), respiratory (9.1 vs. -0.04%), and cardiovascular (7.3 vs.
16 -0.2%) mortality. In an analysis of 10 Italian cities, Chiusolo etal. (2011) supports the
17 results of Bellini et al. (2007) by indicating larger NO2-mortality risk estimates in the
18 warm season compared with all-year (Table 5-62) for total (nonaccidental) mortality and
19 cause-specific mortality (i.e., cardiac, cerebrovascular, and respiratory).
20 The evidence for increased NO2-mortality associations in the warm season, as presented
21 in the Canadian and Italian multicity studies (Shin etal.. 2012; Stieb et al.. 2008; Brook
22 et al., 2007; Burnett et al.. 2004). differs from the seasonal patterns observed in a study
23 conducted in Shanghai as part of the PAPA study (Kan etal.. 2010; Kan et al.. 2008). The
24 authors reported evidence of increased NO2-mortality risk estimates in the cold season
25 compared with the warm for total (nonaccidental) mortality (cold: 4.7 vs. warm: 1.7% for
26 a 20-ppb increase in 24-h avg NO2 at lag 0-1 days), cardiovascular (cold: 4.8 vs. warm:
27 1.1%), and respiratory mortality (cold: 10.4 vs. warm: -5.1%). Across all of the gaseous
28 pollutants examined, mortality risk estimates were double the size or larger in the cool
29 season, whereas PMio mortality risk estimates were similar across seasons except for
30 respiratory mortality (larger in the cool season). The authors speculate these seasonal
31 differences could be due to seasonal exposure differences specific to Shanghai (i.e.,
32 limited time spent outdoors and increased air conditioning use in the warm season
33 because of high temperature and humidity and heavy rain, versus more time spent
34 outdoors and open windows in the cool season) (Kan etal.. 2010; Kan et al.. 2008). The
35 results of (Kan etal.. 2010; Kan et al.. 2008) highlight the complexity of clearly
36 identifying seasonal patterns in NO2-mortality associations across locations with
37 drastically different seasonal weather patterns.
January 2015 5-347 DRAFT: Do Not Cite or Quote
-------
5.4.7 Nitrogen Dioxide-Mortality Concentration-Response Relationship and
Related Issues
Lag Structure of Associations
1 The 2008 ISA for Oxides of Nitrogen found consistent evidence across studies indicating
2 that NO2-mortality effects occur within the first few days after exposure, with multiple
3 studies demonstrating the largest effect occurring the day after exposure (i.e., lag 1 day)
4 (U.S. EPA. 2008a). Recent multicity studies have conducted additional analyses
5 examining multiday lags, which further inform the lag structure of associations between
6 short-term NO2 exposure and mortality.
7 In the analysis of 10 Italian cities, Chiusolo et al. (2011) examined the lag structure of
8 associations between mortality and short-term NO2 exposure through both single-day and
9 multiday lag analyses. Multiday analyses consisted of a priori defined lags (i.e., 0-1, 2-5,
10 and 0-5 days) examined using an unconstrained distributed lag model. In addition to
11 examining single-day lags of 0 to 5 days, the authors also explored the pattern of
12 associations observed over each individual day using a constrained polynomial
13 distributed lag model. It is important to note that the individual lag days of a constrained
14 distributed lag model are not directly interpretable; however, this analysis allowed
15 Chiusolo etal. (2011) to visually display the potential latency of the NO2 effect on
16 mortality. Collectively, the single- and multi-day lag analyses support an immediate
17 effect of NO2 on mortality but also provide evidence for a prolonged effect extending out
18 to 5 days for all mortality outcomes (Figure 5-24).
January 2015 5-348 DRAFT: Do Not Cite or Quote
-------
3.00
2.50
8 2,00
CO
Ł 1.50
u
~ 1 .00
S 0.50
iu n
-0.50
1 nn
All natural mortality
,,tl
QSingle-lag models
• Distributed-lag models
0123
t
U '
1
[
1
1
Cardiac mortality
T T T IT
T I 1 -
1 5 1 \ 1 3 *
,i,
n
]
4 5 0-1 2-50-5 012345 0-1 2-5 0-5
Lag (days) Lag (days)
5.00
4.00
03
re 3.00
0)
« 2.00
u
0>
0_
1.00
0
-1.00
-2,00
Cerebrovascular mortality
n
Respiratory mortality I
34 5
Lag (days)
0-1 2-5 0-5
345
Lag (days)
0-1 2-5 0-5
Source: Reproduced with permission from Environmental Health Perspectives (Chiusolo et al.. 2011).
Figure 5-24 Percentage increase in total and cause-specific mortality due to
short-term nitrogen dioxide exposure at single day lags,
individual lag days of a constrained polynomial distributed lag
model, and multiday lags of an unconstrained distributed lag
model.
i
2
3
4
5
6
1
8
9
10
11
Chen etal. (2012b) also conducted an extensive analysis of the lag structure of
associations for the NCh-mortality relationship as part of CAPES. Multiday lags were
examined by averaging multiple single lag days and using a constrained polynomial
distributed lag model of 0-4 days. Chen etal. (2012b) reported the largest effect at single
day lags of 0 and 1 and the average of lags 0-1 days indicating an immediate effect of
NC>2 on mortality (Figure 5-25). However, the similar or larger magnitude 0-4 day avg
and distributed lag model results provide some evidence for a delayed NO2 effect on
total, cardiovascular, and respiratory mortality, which is consistent with the results of
Chiusolo etal. (2011) (Figure 5-24). These results were further supported by studies of
cause-specific mortality. Chen etal. (2013) as part of CAPES, in a subset of eight
Chinese cities, reported the largest magnitude of an NO2 effect on stroke mortality at lag
January 2015
5-349
DRAFT: Do Not Cite or Quote
-------
1
2
o
6
4
5
6
7
0-1 days, but the association remained positive and statistically significant in an analysis
of lag 0-4 days (Section 5.3.8). In an analysis of COPD mortality in four Chinese cities,
Meng etal. (2013) also provided evidence of associations larger in magnitude for
multiday averages, suggestive of a prolonged effect, with the largest association at lag
0-4 and slightly smaller associations for a lag of 0-1 days (Section 5.2.8). These results
are consistent with Faustini etal. (2013) in a study of out-of-hospital respiratory mortality
in six Italian cities that found upon examining both single- and multi-day lags the
strongest associations with NC>2 were for lags of 2-5 and 0-5 days.
4.0 -|
^ 3.5-
"cc 3.0-
•c
o
E 2.5-
c
i 2-°"
CC
2. 1.5-
o
Ł 1.0-
Q)
§ 0.5-
CD
Q_
0.0-
-0.5-
4
I
0
{
1
t
2
-•
3
Tola
4
i
I
01
<
i
i
i
04 pLivj
mortality
i
0
t
»
i
i
2
(
1
3
{
L
i
\
«
i
01
4
4
i
i
04 PLIVJ
Cardiovascular mortality
1
0
i
1
i
2
|
i
i.
1
1
i
i
4
i
i
01
i
i
i
04pLIV
Respiratory mortality
Percentage increase (mean and 95% Cl) of daily mortality associated with a 10 |jg/m3 (5.3 ppb) increase of NO2 concentrations,
using different lag structures in the CAPES cities. Multiday average lag 01 corresponds to 2-day moving average, and lag 04
corresponds to 5-day moving average of NO2 concentration of the current and previous 4 days. DIM: polynomial distributed lag
model, representing the cumulative effects of NO2.
Source: Reprinted with permission of Elsevier Ltd. (Chen et al.. 2012b).
Figure 5-25 Percentage increase in total and cause-specific mortality due to
short-term nitrogen dioxide exposure in single- and multi-day lag
models.
9
10
11
12
13
14
Additional studies that examined associations between NO2 and mortality at single-day
lags or multiday averages provide evidence that is consistent with those studies evaluated
in the 2008 ISA for Oxides of Nitrogen (U.S. EPA. 2008a). which demonstrated strong
associations between NO2 and mortality at lag 1. In the analysis of 12 Canadian cities,
Stieb etal. (2008) found the strongest association between short-term NCh exposure and
mortality at lag 1 when examining single-day lags of 0-2 days. Wong et al. (2008) and
January 2015
5-350
DRAFT: Do Not Cite or Quote
-------
1 Wong et al. (2010) examined single and multiday lags in each individual city in the
2 PAPA study. In the three Chinese cities, similar to Stieb et al. (2008). the authors
3 reported evidence of immediate effects of NO2 on mortality; with the strongest
4 association occurring for a 0-1 day lag. However, in Bangkok, the lag structure of
5 associations was different and more in line with those observed in Chiusolo et al. (2011)
6 and Chenetal. (2012b). with the strongest association occurring at a lag of 0-4 days.
Concentration-Response Relationship
7 The studies evaluated in the 2008 ISA for Oxides of Nitrogen (U.S. EPA. 2008a) that
8 examined the association between short-term NO2 exposure and mortality did not
9 conduct formal analyses of the C-R relationship. Recent studies published since the
10 completion of the 2008 ISA for Oxides of Nitrogen have examined the NO2-mortality
11 C-R relationship in both multi- and single-city analyses, focusing on the shape of the C-R
12 curve and whether a threshold exists.
13 Using a subsampling approach, Moolgavkar et al. (2013) examined the shape of the C-R
14 relationship between short-term air pollution exposures and mortality in the NMMAPS
15 data set by applying a nonlinear function (i.e., natural splines with 6 df) to each pollutant.
16 This analysis provides support for a linear relationship between short-term NO2
17 exposures and mortality (Figure 5-26). Although Moolgavkar et al. (2013) state that the
18 C-R relationship for NO2 "suggest(s) non-linearity and threshold like behavior" the
19 widening of the confidence intervals at the tails of the distribution prevents a clear
20 interpretation of the shape of the curve where the data density is low. It should be noted
21 that the confidence intervals approach zero at the low end of the NO2 distribution due to
22 the way the model is structured.
January 2015 5-351 DRAFT: Do Not Cite or Quote
-------
0.06
0.04
O)
o
0.02
0
-0.02
0
20
40
60
80
Lag-1 NO.
Source: Reproduced with permission from Environmental Health Perspectives (Moolgavkar et al.. 2013).
Note: RR = relative risk.
Figure 5-26 Flexible ambient concentration-response relationship between
short-term nitrogen dioxide (NOa, in ppb) exposure and mortality
at lag day 1. Pointwise means and 95% CIs adjusted for size of the
bootstrap sample.
i
2
3
4
The evidence for a linear C-R relationship between short-term NO2 exposure and
mortality was further supported by Stieb et al. (2008) in a pooled analysis of 12 Canadian
cities. The authors examined three functional forms (i.e., linear, quadratic, and cubic
polynomial) and assessed the model fit using the sum of the Akaike Information
January 2015
5-352
DRAFT: Do Not Cite or Quote
-------
1 Criterion. Stieb et al. (2008) indicated that the linear function was the best fit of the
2 NO2-mortality relationship (quantitative results not presented).
3 Multicity studies conducted in Asia examined the NC^-mortality C-R relationship
4 through either a combined analysis using data from all cities or by examining the C-R
5 relationship in individual cities. Chen et al. (2012b) examined the shape of the
6 NO2-mortality C-R curve across all cities as part of CAPES for total, cardiovascular, and
7 respiratory mortality using 24-h avg NCh concentrations at lag 0-1 days. To limit the
8 influence of extreme NC>2 concentrations on the shape of the C-R curve, concentrations
9 greater than 120 (ig/m3 (62.4 ppb), which represented only 3% of the data, were
10 excluded. The authors used a cubic spline with two knots at different concentrations for
11 each of the mortality outcomes [40 (ig/m3 (20.8 ppb) and 70 (ig/m3 (36.4 ppb) for total
12 mortality, 50 (ig/m3 (26.0 ppb) and 70 (ig/m3 (36.4 ppb) for cardiovascular mortality, and
13 40 (ig/m3 (20.8 ppb) and 60 (ig/m3 (31.2 ppb) for respiratory mortality]. Chen et al.
14 (2012b) found evidence of a linear relationship between short-term NC>2 exposure and
15 total and cause-specific mortality (Figure 5-27). which was confirmed by the lack of a
16 statistically significant difference in the deviance between the spline and linear fit
17 models. These results are further supported by examinations of the C-R relationship for
18 the cause-specific mortality outcomes of stroke [(ChenetaL. 2013): Section 5.3.10] and
19 COPD [Meng etal. (2013); Section 5.2.8]. which also provided evidence of a linear
20 relationship.
January 2015 5-353 DRAFT: Do Not Cite or Quote
-------
CM
>\
-^
To
o
E
CD
•o
V)
03
u
c
•«-•
c
Q)
O
o _
^^— Total mortality
~~ • Cardiovascular mortality
1 - '' Respiratory mortality
20
40 60 80 100
Two-day average N02 concentrations
120
Source: Reprinted with permission of Elsevier Ltd. (Chen et al.. 2012b).
Note: NO2 concentrations on the x-axis are in the unit of |jg/m3.
Figure 5-27 CAPES concentration-response curve for the association
between total and cause-specific mortality and 24-hour average
nitrogen dioxide (NOa) concentrations at lag 0-1 days.
3
4
5
6
7
8
9
10
The four-city PAPA study (Wong et al.. 2010; Wong et al., 2008) also examined the
NO2-mortality C-R relationship but only focused on the shape of the C-R curve in each
individual city. The C-R curve for the NCh-mortality relationship was assessed by
applying a natural spline smoother with 3 df to NC>2 concentrations. To examine whether
the NO2-mortality relationship deviates from linearity, the deviance between the
smoothed (nonlinear) pollutant model and the unsmoothed (linear) pollutant model was
examined. The C-R curves in the three Chinese cities further support the results from
Stieb etal. (2008) and Chen et al. (2012b) by indicating a linear relationship between
short-term NC>2 concentrations and mortality (Figure 5-28). Specifically, the evidence for
linearity was strongest between the 25th and 75th percentiles of the NO2 concentrations
January 2015
5-354
DRAFT: Do Not Cite or Quote
-------
1 in each city with some uncertainty in the shape of the C-R curve at lower concentrations
2 where the data density is low, generally below the 25th percentile. The results of the
3 analysis for Bangkok, which provides evidence for nonlinearity, are consistent with what
4 has been observed in examinations of city-specific C-R curves for other air pollutants
5 (e-g-, PM and Os). That is, the heterogeneity in city-specific risk estimates can translate
6 into heterogeneity in the shape of the C-R curve, which has often been hypothesized to be
7 due to city-specific exposure characteristics and demographics. The results from the
8 Bangkok analysis highlight the difficulty in interpreting a combined C-R curve across
9 cities, when there is evidence for city-to-city differences in the association between
10 short-term NC>2 exposure and mortality.
January 2015 5-355 DRAFT: Do Not Cite or Quote
-------
-0.1
0.3
0.2
0.1
0.0
-0.1
Hong Kong
_^-
20 40 60 80 100 120
N02 concentration (ug/m3)
20 40 60 80 100 120 140
N02 concentration (fig/m3)
0.3
0.2
.M
00
•= 0.1
3
0.0
-0.1
Shanghai
0.3
0.2
0.1
0.0
-0.1
Wuhan
50 100 150 200
N02 concentration (ug/m3)
20 40 60 80 100 120
N02 concentration (ug/m3)
Note: Thin vertical lines represent interquartile range of NO2 concentrations in each city. The thick line was included by Wong et al.
(2008) to depict where the World Health Organization 1 -year averaging time standard for NO2 of 40 |jg/m3 (20.8 ppb) could be found
along the distribution of NO2 concentrations in each city.
Source: Reproduced with permission from Environmental Health Perspectives (Wong et al.. 2008).
Figure 5-28 Concentration-response curve for association between total
mortality and 24-hour average nitrogen dioxide (NOa)
concentrations at lag 0-1 days in the four cities of the Public
Health and Air Pollution in Asia study.
January 2015
5-356
DRAFT: Do Not Cite or Quote
-------
5.4.8 Summary and Causal Determination
1 Recent multicity studies evaluated since the completion of the 2008 ISA for Oxides of
2 Nitrogen continue to provide consistent evidence of positive associations between
3 short-term NC>2 exposures and total mortality. Although the body of evidence is larger,
4 key uncertainties and data gaps still remain, which contribute to the conclusion that the
5 evidence for short-term NC>2 exposures and total mortality is suggestive, but not
6 sufficient to infer a causal relationship. This conclusion is consistent with that reached in
7 the 2008 ISA for Oxides of Nitrogen (U.S. EPA. 2008a). Recent multicity studies
8 evaluated have further informed key uncertainties and data gaps in the NO2-mortality
9 relationship identified in the 2008 ISA for Oxides of Nitrogen including confounding,
10 modification of the NO2-mortality relationship, potential seasonal differences in
11 NO2-mortality associations, and the shape of the NO2-mortality C-R relationship.
12 However, questions remain regarding whether NO2 is independently associated with
13 mortality, specifically due to the lack of copollutant model analyses that focus on
14 traffic-related pollutants. This section describes the evaluation of evidence for total
15 mortality with respect to the causal determination for short-term NO2 exposure, using the
16 framework described in Table II of the Preamble. The key evidence, as it relates to the
17 causal framework, is summarized in Table 5-63.
18 Collectively, the evidence from recent multicity studies of short-term NO2 exposures and
19 mortality consistently demonstrate the NO2-mortality association is robust in copollutant
20 models with PMio, Os, and SO2. However, NO2 is often highly correlated with other
21 traffic-related pollutants complicating the ability to disentangle the independent effects of
22 NO2 from those of other measured or unmeasured pollutants associated with traffic
23 (Section 1.4.3) (Figure 3-6). adding uncertainty to the interpretation of the association
24 between NO2 and total mortality. In addition, studies that focused on PM and examined
25 whether NO2 modified the PM-mortality relationship reported that PM risk estimates
26 increased as NO2 concentrations increased or the ratio of NO2/PM increased. These
27 results suggest that NO2 and PM may be effect modifiers of each other. This is consistent
28 with the conclusions of the 2008 ISA for Oxides of Nitrogen (U.S. EPA. 2008a). In
29 addition to copollutant analyses, recent studies examined the influence of the extent of
30 temporal adjustment on NO2-mortality risk estimates and reported similar results across a
31 range of degrees of freedom per year.
32 An examination of factors that may contribute to increased risk of NO2-related mortality,
33 as discussed in Chapter 7. found evidence indicating that older adults (>65 years of age),
34 females, individuals with pre-existing cardiovascular or respiratory diseases, and
35 individuals of lower SES, specifically lower income and educational attainment, are at
36 greater risk. Studies that examined whether there are seasonal differences in the
January 2015 5-357 DRAFT: Do Not Cite or Quote
-------
1 NCh-mortality relationship found greater effects in the warm or summer months in
2 multicity studies conducted in Canada and Europe. However, these results are
3 contradicted by a study conducted in Asia where larger effects were observed in the cold
4 season. These between-study differences in seasonal associations are more than likely a
5 reflection of the different seasonal weather patterns observed between countries (Kan et
6 al.. 2010: Kan etal.. 2008).
7 Those studies that examined the lag structure of associations for the NCh-mortality
8 relationship observed that there continues to be evidence of an immediate effect (i.e., lag
9 0 to 1 day), which is consistent with studies evaluated in the 2008 ISA for Oxides of
10 Nitrogen. Recent studies also provided evidence for a prolonged effect on mortality in
11 distributed lag models with lags ranging from 0-4 to 0-5 days (Chen etal.. 2012b:
12 Chiusolo etal.. 2011). Multicity studies have examined the shape of the C-R relationship
13 and whether a threshold exists in both a multi- and single-city setting. These studies have
14 used different statistical approaches and consistently demonstrated a linear relationship
15 with no evidence of a threshold within the range of NC>2 concentrations currently found in
16 the U.S. However, consistent with observations from C-R analyses conducted for other
17 criteria pollutants [e.g., PM (U.S. EPA. 2009) and O3 (U.S. EPA. 2013a)1. an
18 examination of the C-R relationship in individual cities, specifically in China, has
19 demonstrated heterogeneity in the shape of the curve across cities (Wong etal.. 2010;
20 Wong et al.. 2008).
21 Overall, recent epidemiologic studies build upon and support the conclusions of the 2008
22 ISA for Oxides of Nitrogen for total mortality. However, the biological mechanism that
23 could lead to mortality as a result of short-term NO2 exposures has not been clearly
24 characterized. This is evident when evaluating the underlying health effects (i.e.,
25 cardiovascular effects in Section 5.3 and respiratory effects in Section 5.2) that could lead
26 to cardiovascular (-35% of total mortality) and respiratory (~9% of total mortality)
27 mortality, the components of total mortality most thoroughly evaluated (Hoyert and Xu.
28 2012). An evaluation of epidemiologic studies that examined the relationship between
29 short-term NO2 exposure and cardiovascular effects found consistent evidence for
30 myocardial infarction and inconclusive epidemiologic and experimental evidence for
31 other cardiovascular endpoints. However, important uncertainties remain especially
32 regarding disentangling whether there is an independent effect of NO2 on cardiovascular
33 effects, which is the same uncertainty in total mortality studies. Overall this evidence
34 provides limited coherence and biological plausibility for NO2-related cardiovascular
35 mortality. For respiratory effects, there is causal evidence for NO2-related asthma
36 exacerbation supported by toxicological and controlled human exposure studies
37 demonstrating increased airway responsiveness (Section 5.2.2.1) in response to
38 short-term NO2 exposures as well as epidemiologic studies reporting respiratory-related
January 2015 5-358 DRAFT: Do Not Cite or Quote
-------
1
2
o
6
4
5
6
7
8
9
10
11
12
morbidity including hospital admissions and ED visits, specifically for asthma
(Section 5.2.2.4). However, the biological mechanism that explains the continuum of
effects that could lead to respiratory-related mortality also remains unclear. Additionally,
it is important to note studies that examine the association between short-term NCh
exposures and mortality rely on central site monitors, which may contribute to exposure
measurement error and underestimate associations observed (Section 3.4.5.1). In
conclusion, the consistent positive associations observed across various multicity studies
is limited by the uncertainty due to whether NC>2 is independently associated with total
mortality as well as the uncertainty in the biological mechanism that could lead to
NO2-induced mortality. Collectively, this body of evidence is suggestive, but not
sufficient to infer a causal relationship between short-term NO2 exposure and total
mortality.
Table 5-63 Summary of evidence, which is suggestive, but not sufficient to infer,
a causal relationship between short-term nitrogen dioxide (NO2)
exposure and total mortality.
Rationale for Causal
Determination3
Consistent
epidemiologic evidence
from multiple, high-
quality studies at
relevant NO2
concentrations
Key Evidence13
Increases in mortality in multicity
studies conducted in the U.S.,
Canada, Europe, and Asia.
NO2
Concentrations
Associated with
Key References'3 Effects0
Section 5.4.3 Mean 24-h avq:
Table 5-60 9.2-55.0 ppb
Mean 1-h max:
16.3-80.5 ppb
Mean 3-h max:
16. 3-42. 6 ppb.
Table 5-59
Uncertainty regarding
potential confounding by
traffic-related
copollutants
Although NO2 associations were
relatively unchanged in copollutant
models with PM-io, SO2, and Os; NO2
is often highly correlated with other
traffic-related pollutants (e.g., PM2.5,
EC, and CO) complicating the
interpretation of whether NO2 is
independently associated with total
mortality.
(Moolqavkar et al.
(2013): Chen et al.
(2012b): Chiusolo et al.
(2011): Wong etal.
(2010): Stiebetal.
(2008): Wong etal.
(2008))
Section 3.4.5, Figure 3-6:
Section 5.4.4
NO2 and PM may be effect modifiers
of each other.
(Katsouvanni et al.
(2009): Katsouvanni et
al. (2003): Katsouvanni
etal. (2001))
Uncertainty regarding
exposure measurement
error
Studies that examine the association
between short-term NO2 exposures
and mortality rely on central site
monitors.
Sections 3.4.5.1 and 3.5
January 2015
5-359
DRAFT: Do Not Cite or Quote
-------
Table 5-63 (Continued): Summary of evidence, which is suggestive, but not
sufficient to infer, a causal relationship between short-
term nitrogen dioxide (NO2) exposure and total mortality.
Rationale for Causal
Determination3
Key Evidence13
Key References'3
NO2
Concentrations
Associated with
Effects0
Uncertainty due to
limited coherence and
biological plausibility
with cardiovascular and
respiratory morbidity
evidence
Consistent epidemiologic evidence
for myocardial infarction.
Inconclusive epidemiologic and
experimental evidence for other
cardiovascular endpoints.
Uncertainties with respect to the
independent effect of NO2 on
cardiovascular effects contributing to
limited coherence and biological
plausibility for NO2-related
cardiovascular mortality, which
comprises -35% of total mortalityd.
Section 5.3.12
Table 5-58
Consistent evidence for asthma
exacerbation from experimental
studies demonstrating increased
airway responsiveness and
epidemiologic studies demonstrating
asthma-related morbidity.
Uncertainty as to the biological
mechanism that explains the
continuum of effects leading to NO2-
related respiratory mortality, which
comprises -8% of total mortality11.
Section 5.2.9
Table 5-45
CO = carbon monoxide, EC = elemental carbon, NO2 = nitrogen dioxide, O3 = ozone, PM = particulate matter, SO2 = sulfur
dioxide.
aBased on aspects considered in judgments of causality and weight of evidence in causal framework in Table I and Table II of the
Preamble.
Describes the key evidence and references, supporting or contradicting, contributing most heavily to causal determination and,
where applicable, to uncertainties or inconsistencies. References to earlier sections indicate where full body of evidence is
described.
°Describes the NO2 concentrations with which the evidence is substantiated (for experimental studies, below 5,000 ppb).
Statistics taken from American Heart Association (2011)
January 2015
5-360
DRAFT: Do Not Cite or Quote
-------
References for Chapter 5
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/10.1136/thorax.2003.006445
Adams. WC: Brookes. KA: Schelegle. ES. (1987). Effects of NO2 alone and in combination with O3 on
young men and women. J Appl Physiol 62: 1698-1704.
Aga. E: Samolj E: Touloumj G: Anderson. HR: Cadum. E: Forsberg. B: Goodman. P: Goren. A:
Kotesovec. F: Kriz. B: Macarol-Hiti M: Medina. S: Paldv. A: Schindler. C: Sunyer. J: Tittanen. P:
Wojtyniak. B: Zmirou. D: Schwartz. J: Katsouvannj K. (2003). Short-term effects of ambient particles
on mortality in the elderly: Results from 28 cities in the APHEA2 project. Eur Respir J 21: 28s-33s.
http://dx.doi.org/10.1183/09031936.03.00402803
Agarwal. R: Awasthi. A: Singh. N: Gupta. PK: Mittal. SK. (2012). Effects of exposure to rice-crop
residue burning smoke on pulmonary functions and Oxygen Saturation level of human beings in
Patiala (India). Sci Total Environ 429: 161-166. http://dx.doi.Org/10.1016/j.scitotenv.2012.03.074
Ahmed. T: Dougherty. R: Sackner. MA. (1983a). Effect of 0.1 ppm NO2 on pulmonary functions and
non-specific bronchial reactivity of normals and asthmatics [final report]. (CR-83/11/BI). Warren, MI:
General Motors Research Laboratories.
Ahmed. T: Dougherty. R: Sackner. MA. (1983b). Effect of NO2 exposure on specific bronchial reactivity
in subjects with allergic bronchial asthma [final report]. (CR-83/07/BI). Warren, MI: General Motors
Research Laboratories.
Altug. H: Gaga. EO: Dogeroglu. T: Brunekreef. B: Hoek. G: Van Doom. W. (2014). Effects of ambient air
pollution on respiratory tract complaints and airway inflammation in primary school children. Sci
Total Environ 479-480: 201-209. http://dx.doi.Org/10.1016/j.scitotenv.2014.01.127
American Academy of Pediatrics. (2004). Ambient air pollution: health hazards to children. Pediatrics
114: 1699-1707. http://dx.doi.org/10.1542/peds.2004-2166
American Heart Association. (2011). Cardiovascular disease statistics. Available online at
http://web.archive.Org/web/20110410093532/http://www.americanheart.org/presenter.jhtml7identifier
=4478 (accessed April 12, 2011).
Amoruso. MA: Witz. G: Goldstein. BD. (1981). Decreased superoxide anion radical production by rat
alveolar macrophages following inhalation of ozone or nitrogen dioxide. Life Sci 28: 2215-2221.
Andersen. ZJ: Loft S: Ketzel M: Stage. M: Scheike. T: Hermansen. MN: Bisgaard. H. (2008a). Ambient
air pollution triggers wheezing symptoms in infants. Thorax 63: 710-716.
http://dx.doi.org/10.1136/thx.2007.085480
Andersen. ZJ: Olsen. TS: Andersen. KK: Loft. S: Ketzel. M: Raaschou-Nielsen. O. (2010). Association
between short-term exposure to ultrafine particles and hospital admissions for stroke in Copenhagen,
Denmark. Eur Heart J 31: 2034-2040. http://dx.doi.org/10.1093/eurheartj/ehql88
Andersen. ZJ: Wahlin. P: Raaschou-Nielsen. O: Ketzel M: Scheike. T: Loft. S. (2008b). Size distribution
and total number concentration of ultrafine and accumulation mode particles and hospital admissions
in children and the elderly in Copenhagen, Denmark. Occup Environ Med 65: 458-466.
http://dx.doi.org/10.1136/oem.2007.033290
Anderson. HR: Armstrong. B: Hajat S: Harrison. R: Monk. V: Polonieckj 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
January 2015 5-361 DRAFT: Do Not Cite or Quote
-------
Annesi-Maesano. I: Rouve. S: Desqueyroux. H: Jankovskj R: Klossek. JM: Thibaudon. M: Demoly. P:
Didier. A. (2012). Grass pollen counts, air pollution levels and allergic rhinitis severity. Int Arch
Allergy Immunol 158: 397-404. http://dx.doi.org/10.1159/000332964
Arbex. MA: de Souza Conceicao. GM: Cendon. SP: Arbex. FF: Lopes. AC: Moyses. EP: Santiago. SL:
Saldiva. PHN: Pereira. LAA: Braga. ALF. (2009). Urban air pollution and chronic obstructive
pulmonary disease-related emergency department visits. J Epidemiol Community Health 63: 777-783.
http://dx.doi.org/10.1136/jech.2008.078360
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 Health 54: 398-411.
http://dx.doi.org/10.1080/00039899909603371
Atkinson. RW: Cohen. A: Mehta. S: Anderson. HR. (2012). Systematic review and meta-analysis of
epidemiological time-series studies on outdoor air pollution and health in Asia. Air Qual Atmos Health
5: 383-391. http://dx.doi.org/10.1007/sll869-010-0123-2
ATS (American Thoracic Society). (2000). Guidelines for methacholine and exercise challenge testing-
1999. Am J Respir Crit Care Med 161: 309-329. http://dx.doi.Org/10.1164/ajrccm.161.l.atsll-99
AT SDR (Agency for Toxic Substances and Disease Registry). (2006). A study of ambient air
contaminants and asthma in New York City: Part A and B. Atlanta, GA: U.S. Department of Health
and Human Services.
http://permanent.access.gpo.gov/lps88357/ASTHMA BRONX FINAL REPORT.pdf
Avol. EL: Linn. WS: Peng. RC: Valencia. G: Little. D: Hackney. JD. (1988). Laboratory study of
asthmatic volunteers exposed to nitrogen dioxide and to ambient air pollution. Am Ind Hyg Assoc J
49: 143-149. http://dx.doi.org/10.1080/15298668891379530
Avol. EL: Linn. WS: Peng. RC: Whynot JD: Shamoo. DA: Little. DE: Smith. MN: Hackney. JD. (1989).
Experimental exposures of young asthmatic volunteers to 0.3 ppm nitrogen dioxide and to ambient air
pollution. Toxicol Ind Health 5: 1025-1034.
Ayaz. KL: Csallany. AS. (1978). Long-term NO2 exposure of mice in the presence and absence of
vitamin E. II. Effect of glutathione peroxidase. Arch Environ Occup Health 33: 292-296.
Azadniv. M: Utell. MJ: Morrow. PE: Gibb. FR: Nichols. J: Roberts. NJ. Jr: Speers. DM: Torres. A: Tsai.
Y: Abraham. MK: Voter. KZ: Frampton. MW. (1998). Effects of nitrogen dioxide exposure on human
host defense. Inhal Toxicol 10: 585-601. http://dx.doi.org/10.1080/089583798197556
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
Bagheri Lankarani. N: Kreis. I: Griffiths. DA. (2010). Air pollution effects on peak expiratory flow rate in
children. Iran J Allergy Asthma Immunol 9: 117-126. http://dx.doi.org/09.02/ijaai.117126
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
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.
http://dx.doi.org/10.1136/jech.2005.037978
Ballester. F: Tenias. JM: Perez-Hoyos. S. (2001). Air pollution and emergency hospital admissions for
cardiovascular diseases in Valencia, Spain. J Epidemiol Community Health 55: 57-65.
http://dx.doi.0rg/10.1136/iech.55.l.57
January 2015 5-362 DRAFT: Do Not Cite or Quote
-------
Barck. C: Lundahl. J: Hallden. G: Bylin. G. (2005a). Brief exposures to NO2 augment the allergic
inflammation in asthmatics. Environ Res 97: 58-66. http://dx.doi.Org/10.1016/j.envres.2004.02.009
Barck. C: Lundahl. J: Holmstrom. M: Bylin. G. (2005b). Does nitrogen dioxide affect inflammatory
markers after nasal allergen challenge? Am J Rhinol 19: 560-566.
Barck. C: Sandstrom. T: Lundahl J: Hallden. G: Svartengren. M: Strand. V: Rak. S: Bylin. G. (2002).
Ambient level of NO2 augments the inflammatory response to inhaled allergen in asthmatics. Respir
Med 96: 907-917. http://dx.doi.org/10.1053/rmed.2002.1374
Barclay. JL: Miller. BG: Dick. S: Dennekamp. M: Ford. I: Hillis. GS: Ayres. JG: Seaton. A. (2009). A
panel study of air pollution in subjects with heart failure: negative results in treated patients. Occup
Environ Med 66: 325-334. http://dx.doi.org/10.1136/oem.2008.039032
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
Barr. RG: Herbstman. J: Speizer. FE: Camargo. CA. Jr. (2002). Validation of self-reported chronic
obstructive pulmonary disease in a cohort study of nurses. Am J Epidemiol 155: 965-971.
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
Bartell SM: Longhurst. J: Tjoa. T: Sioutas. C: Delfino. RJ. (2013). Particulate air pollution, ambulatory
heart rate variability, and cardiac arrhythmia in retirement community residents with coronary artery
disease. Environ Health Perspect 121: 1135-1141. http://dx.doi.org/10.1289/ehp.1205914
Barth. PJ: Miiller. B. (1999). Effects of nitrogen dioxide exposure on Clara cell proliferation and
morphology. Pathol Res Pract 195: 487-493. http://dx.doi.org/10.1016/S0344-0338(99)80052-l
Barth. PJ: Muller. B: Wagner. U: Bittinger. A. (1995). Quantitative analysis of parenchymal and vascular
alterations in NO2-induced lung injury in rats. Eur Respir J 8: 1115-1121.
http://dx.doi.org/10.1183/09031936.95.08071115
Bauer. MA: Utell MJ: Morrow. PE: Speers. DM: Gibb. FR. (1986). Inhalation of 0.30 ppm nitrogen
dioxide potentiates exercise-induced bronchospasm in asthmatics. Am Rev Respir Dis 134: 1203-
1208.
Beaupre. A: Orehek. J. (1982). Factors influencing the bronchodilator effect of a deep inspiration in
asthmatic patients with provoked bronchoconstriction. Thorax 37: 124-128.
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
Bellini P: Baccini. M: Biggeri. A: Terracini. B. (2007). The meta-analysis of the Italian studies on short-
term effects of air pollution (MISA): Old and new issues on the interpretation of the statistical
evidences. Environmetrics 18: 219-229. http://dx.doi.org/10.1002/env.836
Berglind. N: Bellander. T: Forastiere. F: von Klot S: Aalto. P: Elosua. R: Kulmala. M: Lanki. T: Lowel.
H: Peters. A: Picciotto. S: Salomaa. V: Stafoggia. M: Sunyer. J: Nyberg. F. (2009). Ambient air
pollution and daily mortality among survivors of myocardial infarction. Epidemiology 20: 110-118.
http://dx.doi.org/10.1097/EDE.Ob013e3181878b50
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
January 2015 5-363 DRAFT: Do Not Cite or Quote
-------
Bhaskaran. K: Hajat S: Armstrong. B: Haines. A: Herrett E: Wilkinson. P: Smeeth. L. (2011). The effects
of hourly differences in air pollution on the risk of myocardial infarction: case crossover analysis of
the MINAP database. B M J 343: d5531. http://dx.doi.org/10.1136/bmj.d5531
Biggeri. A: Baccini. M: Bellini. P: Terracinj 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.
http://dx.doi.0rg/10.1179/oeh.2005.ll.l.107
Bind. MA: Baccarelli A: Zanobetti A: Tarantinj L: Suh. H: Vokonas. P: Schwartz. J. (2012). Air
pollution and markers of coagulation, inflammation, and endothelial function: Associations and
epigene-environment interactions in an elderly cohort. Epidemiology 23: 332-340.
http://dx.doi.org/10.1097/EDE.Ob013e31824523fO
Blomberg. A: Krishna. MT: Helledav. R: Soderberg. M: Ledin. MC: Kelly. FJ: Frew. AJ: Holgate. ST:
Sandstrom. T. (1999). Persistent airway inflammation but accomodated antioxidant and lung function
responses after repeated daily exposure to nitrogen dioxide. Am J Respir Crit Care Med 159: 536-543.
http://dx.doi.0rg/10.1164/airccm.159.2.9711068
Blount RJ: Djawe. K: Daly. KR: Jarlsberg. LG: Fong. S: Balmes. J: Miller. RF: Walzer. PD: Huang. L.
(2013). Ambient air pollution associated with suppressed serologic responses to Pneumocystis
jirovecii in a prospective cohort of HIV-infected patients with Pneumocystis pneumonia. PLoS ONE
8: e80795. http://dx.doi.org/10.1371/journal.pone.0080795
Boezen. M: Schouten. J: Rijcken. B: Vonk. J: Gerritsen. J: Van Per Zee. S: Hoek. G: Brunekreef. B:
Postma. D. (1998). Peak expiratory flow variability, bronchial responsiveness, and susceptibility to
ambient air pollution in adults. Am J Respir Crit Care Med 158: 1848-1854.
http://dx.doi.0rg/10.1164/ajrccm.158.6.9804072
Brodv. DJ: Zhang. X: Kit. BK: Dillon. CF. (2013). Reference values and factors associated with exhaled
nitric oxide: U.S. youth and adults. Respir Med 107: 1682-1691.
http://dx.doi.0rg/10.1016/i.rmed.2013.07.006
Brook. JR: Burnett. RT: Dann. TF: Cakmak. S: Goldberg. MS: Fan. X: Wheeler. AJ. (2007). Further
interpretation of the acute effect of nitrogen dioxide observed in Canadian time-series studies. J Expo
Sci Environ Epidemiol 17: S36-S44. http://dx.doi.org/10.1038/sj.jes.7500626
Brown. JS. (2015). Nitrogen dioxide exposure and airway responsiveness in individuals with asthma.
Inhal Toxicol 27: 1-14. http://dx.doi.org/10.3109/08958378.2014.979960
Brusasco. V: Crimi. E: Gianiorio. P: Lantero. S: Rossi. GA. (1990). Allergen-induced increase in airway
responsiveness and inflammation in mild asthma. J Appl Physiol 69: 2209-2214.
Briiske. I. (2014). Email from Dr. Irene Brtiske to Molini Patel, cc Dr. Regina Hampel, Dr. Kathrin Wolf.
WG: question re: 2010 Inhalational toxicology paper (With attachments). Available online
Briiske. I: Hampel. R: Baumgartner. Z: Riickerl. R: Greven. S: Koenig. W: Peters. A: Schneider. A.
(2011). Ambient air pollution and lipoprotein-associated phospholipase A2 in survivors of myocardial
infarction. Environ Health Perspect 119: 921-926. http://dx.doi.org/10.1289/ehp.1002681
Bruske. I: Hampel. R: Socher. MM: Ruckerl. R: Schneider. A: Heinrich. J: Oberdorster. G: Wichmann.
HE: Peters. A. (2010). Impact of ambient air pollution on the differential white blood cell count in
patients with chronic pulmonary disease. Inhal Toxicol 22: 245-252.
http://dx.doi.org/10.3109/08958370903207274
Burnett. RT: Smith-Doiron. M: Stieb. D: Cakmak. S: Brook. JR. (1999). Effects of particulate and
gaseous air pollution on cardiorespiratory hospitalizations. Arch Environ Health 54: 130-139.
http://dx.doi.org/10.1080/00039899909602248
Burnett. RT: Smith-Doiron. M: Stieb. D: Raizenne. ME: Brook. JR: Dales. RE: Leech. JA: Cakmak. S:
Krewski. D. (2001). Association between ozone and hospitalization for acute respiratory diseases in
children less than 2 years of age. Am J Epidemiol 153: 444-452.
http://dx.doi.0rg/10.1093/aie/153.5.444
January 2015 5-364 DRAFT: Do Not Cite or Quote
-------
Burnett. RT: Stieb. D: Brook. JR: Cakmak. S: Dales. R: Raizenne. M: Vincent. R: Dann. T. (2004).
Associations between short-term changes in nitrogen dioxide and mortality in Canadian cities. Arch
Environ Occup Health 59: 228-236. http://dx.doi.Org/10.3200/AEOH.59.5.228-236
Burnev. PG: Laitinen. LA: Perdrizet S: Huckauf. H: Tattersfield. AE: Chinn. S: Poisson. N: Heeren. A:
Britton. JR: Jones. T. (1989). Validity and repeatability of the IUATLD (1984) Bronchial Symptoms
Questionnaire: an international comparison. Eur Respir J 2: 940-945.
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/j.envres.2009.03.004
Bylin. G. (1993). Health risk evaluation of nitrogen oxide: Controlled studies on humans. Scand J Work
Environ Health 19: 37-43.
Bylin. G: Hedenstierna. G: Lindvall. T: Sundin. B. (1988). Ambient nitrogen dioxide concentrations
increase bronchial responsiveness in subjects with mild asthma. Eur Respir J 1: 606-612.
Bylin. G: Lindvall. T: Rehn. T: Sundin. B. (1985). Effects of short-term exposure to ambient nitrogen
dioxide concentrations on human bronchial reactivity and lung function. Eur Respir J 66: 205-217.
Cakmak. S: Dales. R: Leech. J: Liu. L. (201 la). The influence of air pollution on cardiovascular and
pulmonary function and exercise capacity: Canadian Health Measures Survey (CHMS). Environ Res
111: 1309-1312. http://dx.doi.0rg/10.1016/j.envres.2011.09.016
Cakmak. S: Dales. RE: Judek. S. (2006). 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.61.1.5-
10
Cakmak. S: Dales. RE: Rubio. MA: Vidal CB. (20 lib). 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/j.envres.2011.01.003
Campen. MJ: Lund. AK: Doyle-Eisele. ML: Mcdonald. JD: Knuckles. TL: Rohr. AC: Knipping. EM:
Mauderly. JL. (2010). A comparison of vascular effects from complex and individual air pollutants
indicates a role for monoxide gases and volatile hydrocarbons. Environ Health Perspect 118: 921-927.
http://dx.doi.org/10.1289/ehp.0901207
Cano Garcinuno. A: Mora Gandarillas. I: Grp. SS. (2013). Early patterns of wheezing in asthmatic and
nonasthmatic children. Eur Respir J 42: 1020-1028. http://dx.doi.org/10.1183/09031936.00148712
Canova. C: Torresan. S: Simonato. L: Scapellato. ML: Tessari. R: Visentin. A: Lotti M: Maestrelli. P.
(2010). Carbon monoxide pollution is associated with decreased lung function in asthmatic adults. Eur
Respir J 35: 266-272. http://dx.doi.org/10.1183/09031936.00043709
Carlsen. HK: Forsberg. B: Meister. K: Gislason. T: Oudin. A. (2013). Ozone is associated with
cardiopulmonary and stroke emergency hospital visits in Reykjavik, Iceland 2003-2009. Environ
Health 12: 28. http://dx.doi.org/10.1186/1476-069X-12-28
Carlsen. HK: Zoega. H: Valdimarsdottir. U: Gislason. T: Hrafnkelsson. B. (2012). Hydrogen sulfide and
particle matter levels associated with increased dispensing of anti-asthma drugs in Iceland's capital.
Environ Res 113: 33-39. http://dx.doi.Org/10.1016/i.envres.2011.10.010
Carraro. S: Gottardi G: Bonetto. G: Baraldi E. (2007). Exhaled nitric oxide in children with asthma and
sinusitis [Review]. PAI 18: 28-30. http://dx.doi.Org/10.llll/i.1399-3038.2007.00629.x
Casset A: Purohit A: Birba. E: Chenard. MP: Uring Lambert. B: Bahrain. S: Meyer. P: Pauli. G: De
Slav. F. (2007). Bronchial challenge test in asthmatics sensitized to mites: Role of particle size in
bronchial response. J Aerosol Med 20: 509-518. http://dx.doi.org/10.1089/jam.2007.0591
Castro. HA: Cunha. MF: Mendonca. GA: Junger. WL: Cunha-Cruz. J: Leon. AP. (2009). Effect of air
pollution on lung function in schoolchildren in Rio de Janeiro, Brazil. Rev Saude Publica 43: 26-34.
January 2015 5-365 DRAFT: Do Not Cite or Quote
-------
Chan. CC: Chuang. KJ: Chen. WJ: Chang. WT: Lee. CT: Peng. CM. (2008). Increasing cardiopulmonary
emergency visits by long-range transported Asian dust storms in Taiwan. Environ Res 106: 393-400.
http://dx.doi.0rg/10.1016/j.envres.2007.09.006
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. http://dx.doi.Org/10.1016/j.envres.2004.07.005
Chang. YK: Wu. CC: Lee. IT: Lin. RS: Yu. YH: Chen. YC. (2012). The short-term effects of air pollution
on adolescent lung function in Taiwan. Chemosphere 87: 26-30.
http://dx.doi.0rg/10.1016/j.chemosphere.2011.ll.048
Channell. MM: Paffett ML: Devlin. RB: Madden. MC: Campen. MJ. (2012). Circulating factors induce
coronary endothelial cell activation following exposure to inhaled diesel exhaust and nitrogen dioxide
in humans: Evidence from a novel translational in vitro model. Toxicol Sci 127: 179-186.
http://dx.doi.org/10.1093/toxsci/kfs084
Chen. BY Chan. CC: Lee. CT: Cheng. TJ: Huang. WC: Jhou. JC: Han. YY Chen. CC: Quo. YL. (2012a).
The association of ambient air pollution with airway inflammation in schoolchildren. Am J Epidemiol
175: 764-774. http://dx.doi.org/10.1093/aie/kwr380
Chen. L: Villeneuve. PJ: Rowe. BH: Liu. L: Stieb. DM. (2014). The Air Quality Health Index as a
predictor of emergency department visits for ischemic stroke in Edmonton, Canada. J Expo Sci
Environ Epidemiol 24: 358-364. http://dx.doi.org/10.1038/jes.2013.82
Chen. R: Chu. C: Tan. J: Cao. J: Song. W: Xu. X: Jiang. C: Ma. W: Yang. C: Chen. B: Gui. Y Kan. H.
(2010). Ambient air pollution and hospital admission in Shanghai, China. J Hazard Mater 181: 234-
240. http://dx.doi.0rg/10.1016/j.jhazmat.2010.05.002
Chen. R: Samolj E: Wong. CM: Huang. W: Wang. Z: Chen. B: Kan. H. (2012b). Associations between
short-term exposure to nitrogen dioxide and mortality in 17 Chinese cities: The China Air Pollution
and Health Effects Study (CAPES). Environ Int 45: 32-38.
http://dx.doi.0rg/10.1016/j.envint.2012.04.008
Chen. R: Zhang. Y: Yang. C: Zhao. Z: Xu. X: Kan. H. (2013). Acute effect of ambient air pollution on
stroke mortality in the china air pollution and health effects study. Stroke 44: 954-960.
http://dx.doi.org/10.1161/STROKEAHA.lll.673442
Chen. SY Su. TC: Lin. YL: Chan. CC. (2012c). Short-term effects of air pollution on pulse pressure
among nonsmoking adults. Epidemiology 23: 341-348.
http://dx.doi.org/10.1097/EDE.Ob013e3182452fld
Cheng. MF: Tsai SS: Yang. CY. (2009). Air pollution and hospital admissions for myocardial infarction
in a tropical city: Kaohsiung, Taiwan. J Toxicol Environ Health A 72: 1135-1140.
http://dx.doi.org/10.1080/15287390903091756
Chetta. A: Foresi A: Del Donno. M: Consigli GF: Bertorelli G: Pesci A: Barbee. RA: Olivierj D.
(1996). Bronchial responsiveness to distilled water and methacholine and its relationship to
inflammation and remodeling of the airways in asthma. Am J Respir Crit Care Med 153: 910-917.
http://dx.doi.0rg/10.1164/airccm.153.3.8630572
Cheung. D: Dick. EC: Timmers. MC: de Klerk. EP: Spaan. WJ: Sterk. PJ. (1995). Rhinovirus inhalation
causes long-lasting excessive airway narrowing in response to methacholine in asthmatic subjects in
vivo. Am J Respir Crit Care Med 152: 1490-1496. http://dx.doi.Org/10.1164/ajrccm.152.5.7582282
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
Chiusolo. M: Cadum. E: Stafoggia. M: Galassi C: Berti G: Faustini A: Bisanti L: Vigotti MA: Dessi.
MP: Cernigliaro. A: Mallone. S: Pacelli B: Minerba. S: Simonato. L: Forastiere. F. (2011). Short term
effects of nitrogen dioxide on mortality and susceptibility factors in 10 Italian cities: The EpiAir
Study. Environ Health Perspect 119: 1233-1238. http://dx.doi.org/10.1289/ehp. 1002904
January 2015 5-366 DRAFT: Do Not Cite or Quote
-------
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.
http://dx.doi.org/10.1136/iech.2006.049205
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 particulate 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: Coull BA: Zanobetti A: Suh. H: Schwartz. J: Stone. PH: Litonjua. A: Speizer. FE: Gold.
DR. (2008). Particulate air pollution as a risk factor for ST-segment depression in patients with
coronary artery disease. Circulation 118: 1314-1320.
http://dx.doi.org/10.1161/CIRCULATIONAHA.108.765669
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
Cockcroft DW. (2010). Direct challenge tests: Airway hyperresponsiveness in asthma: Its measurement
and clinical significance [Review]. Chest 138: 18S-24S. http://dx.doi.org/10.1378/chest.10-0088
Cockcroft. DW: Berscheid. BA: Murdock. KY. (1983). Unimodal distribution of bronchial responsiveness
to inhaled histamine in a random human population. Chest 83: 751-754.
http://dx.doi.0rg/10.1378/chest.83.5.751
Cockcroft DW: Davis. BE. (2006a). Airway hyperresponsiveness as a determinant of the early asthmatic
response to inhaled allergen [Review]. J Asthma 43: 175-178.
http://dx.doi.org/10.1080/02770900600566348
Cockcroft. DW: Davis. BE. (2006b). The bronchoprotective effect of inhaling methacholine by using total
lung capacity inspirations has a marked influence on the interpretation of the test result [Review]. J
Allergy Clin Immunol 117: 1244-1248. http://dx.doi.Org/10.1016/i.iaci.2006.02.038
Cockcroft DW: Davis. BE. (2006c). Mechanisms of airway hyperresponsiveness [Review]. J Allergy
Clin Immunol 118: 551-559. http://dx.doi.Org/10.1016/j.jaci.2006.07.012
Cockcroft DW: Davis. BE: Todd. DC: Smycniuk. AJ. (2005). Methacholine challenge: Comparison of
two methods. Chest 127: 839-844. http://dx.doi.Org/10.1378/chest.127.3.839
Colais. P: Faustini. A: Stafoggia. M: Berti G: Bisanti L: Cadum. E: Cernigliaro. A: Mallone. S: Pacelli.
B: Serinelli M: Simonato. L: Vigotti MA: Forastiere. F. (2012). Particulate air pollution and hospital
admissions for cardiac diseases in potentially sensitive subgroups. Epidemiology 23: 473-481.
http://dx.doi.org/10.1097/EDE.Ob013e31824d5a85
Correia-Deur. J: Claudio. L: Imazawa. AT: Eluf-Neto. J. (2012). Variations in peak expiratory flow
measurements associated to air pollution and allergic sensitization in children in Sao Paulo, Brazil.
Am J Ind Med 55: 1087-1098. http://dx.doi.org/10.1002/aiim.22060
Cuinica. LG: Abreu. I: Esteves da Silva. J. (2014). Effect of air pollutant NO2 on Betula pendula, Ostrya
carpinifolia and Carpinus betulus pollen fertility and human allergenicity. Environ Pollut 186: 50-55.
http://dx.doi.0rg/10.1016/i.envpol.2013.12.001
Currell G: Dowman. AA. (2014). Sign test (Web). Available online at
http://calcscience.uwe.ac.uk/w2/am/Chl2/12 6 SignTest.pdf (accessed June 30, 2014).
Dadvand. P: Nieuwenhuijsen. MJ: Agusti. A: de Batlle. J: Benet M: Beelen. R: Cirach. M: Martinez. D:
Hoek. G: Basagana. X: Ferrer. A: Ferrer. J: Rodriguez-Roisin. R: Sauleda. J: Guerra. S: Anto. JM:
Garcia-Aymerich. J. (2014). Air pollution and biomarkers of systemic inflammation and tissue repair
in COPD patients. Eur Respir J 44: 603-613. http://dx.doi.org/10.1183/09031936.00168813
January 2015 5-367 DRAFT: Do Not Cite or Quote
-------
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
Dales. R: Kauri. LM: Cakmak. S: Mahmud. M: Weichenthal. SA: Van Ryswyk. K: Kumarathasan. P:
Thomson. E: Vincent. R: Broad. G: Liu. L. (2013). Acute changes in lung function associated with
proximity to a steel plant: Arandomized study. Environ Int 55: 15-19.
http://dx.doi.0rg/10.1016/j.envint.2013.01.014
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
Dales. RE: Cakmak. S: Vidal. CB. (2010). Air Pollution and hospitalization for venous thromboembolic
disease in Chile. J Thromb Haemost 8: 669-674. http://dx.doi.Org/10.llll/i.1538-7836.2010.03760.x
Darrow. LA: Klein. M: Sarnat JA: Mulholland. JA: Strickland. MJ: Sarnat SE: Russell AG: Tolbert. PE.
(2011). 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
Davis. JK: Davidson. M: Schoeb. TR. (1991). Murine respiratory mycoplasmosis: a model to study
effects of oxidants (pp. 1-29; discussion 31-43). (ISSN 1041-5505). Health Effects Institute.
http://pubs.healtheffects.org/view. php?id=65
Dawson. SV: Schenker. MB. (1979). Health effects of inhalation of ambient concentrations of nitrogen
dioxide [Editorial]. Am Rev Respir Dis 120: 281-292.
de Burbure. CY: Heilier. JF: Neve. J: Becker. A: Albrecht. C: Borm. PJA: Gromadzinska. J: Wasowicz.
W: Rydzynski. K: Bernard. AM. (2007). Lung permeability, antioxidant status, and NO2 inhalation: A
selenium supplementation study in rats. J Toxicol Environ Health A 70: 284-294.
http://dx.doi.org/10.1080/15287390600884875
Delfino. RJ. (2006). Think globally, breathe locally [Editorial]. Thorax 61: 184-185.
http://dx.doi.org/10.1136/thx.2005.047894
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: Staimer. N: Gillen. D: Tjoa. T: Sioutas. C: Fung. K: George. SC: Kleinman. MT. (2006).
Personal and ambient air pollution is associated with increased exhaled nitric oxide in children with
asthma. Environ Health Perspect 114: 1736-1743. http://dx.doi.org/10.1289/ehp.9141
Delfino. RJ: Staimer. N: Tjoa. T: Arhami. M: Polidori. A: Gillen. PL: George. SC: Shafer. MM: Schauer.
JJ: Sioutas. C. (2010). 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: Staimer. N: Tjoa. T: Gillen. D: Kleinman. MT: Sioutas. C: Cooper. D. (2008a). Personal and
ambient air pollution exposures and lung function decrements in children with asthma. Environ Health
Perspect 116: 550-558. http://dx.doi.org/10.1289/ehp.10911
Delfino. RJ: Staimer. N: Tjoa. T: Gillen. PL: Polidori. A: Arhami. M: Kleinman. MT: Vaziri. ND:
Longhurst. J: Sioutas. C. (2009). Air pollution exposures and circulating biomarkers of effect in a
susceptible population: Clues to potential causal component mixtures and mechanisms. Environ
Health Perspect 117: 1232-1238. http://dx.doi.org/10.1289/ehp.0800194
January 2015 5-368 DRAFT: Do Not Cite or Quote
-------
Delfino. RJ: Staimer. N: Tjoa. T: Gillen. PL: Schauer. JJ: Shafer. MM. (2013). Airway inflammation and
oxidative potential of air pollutant particles in a pediatric asthma panel. J Expo Sci Environ Epidemiol
23: 466-473. http://dx.doi.org/10.1038/jes.2013.25
Delfino. RJ: Staimer. N: Tjoa. T: Polidori. A: Arhami. M: Gillen. PL: Kleinman. MT: Vaziri. ND:
Longhurst. J: Zaldivar. F: Sioutas. C. (2008b). Circulating biomarkers of inflammation, antioxidant
activity, and platelet activation are associated with primary combustion aerosols in subjects with
coronary artery disease. Environ Health Perspect 116: 898-906. http://dx.doi.org/10.1289/ehp.11189
Delfino. RJ: Zeiger. RS: Seltzer. JM: Street. DH: McLaren. CE. (2002). Association of asthma symptoms
with peak particulate air pollution and effect modification by anti-inflammatory medication use.
Environ Health Perspect 110: A607-A617.
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
Desqueyroux. H: Pujet. JC: Prosper. M: Le Moullec. Y: Momas. I. (2002). Effects of air pollution on
adults with chronic obstructive pulmonary disease. Arch Environ Occup Health 57: 554-560.
http://dx.doi.org/10.1080/00039890209602088
Devalia. JL: Rusznak. C: Herdman. MJ: Trigg. CJ: Tarraf. H: Davies. RJ. (1994). Effect of nitrogen-
dioxide and sulfur-dioxide on airway response of mild asthmatic-patients to allergen inhalation.
Lancet 344: 1668-1671.
Devlin. RB: Horstman. DP: Gerritv. TR: Becker. S: Madden. MC. (1999). Inflammatory response in
humans exposed to 2.0 PPM nitrogen dioxide. Inhal Toxicol 11: 89-109.
http://dx.doi.org/10.1080/089583799197195
Dominici F: McDermott A: Daniels. M: Zeger. SL: Samet. JM. (2003). Mortality among residents of 90
cities [HEI]. In Revised analyses of time-series studies of air pollution and health (pp. 9-24). Boston,
MA: Health Effects Institute, http://pubs.healtheffects.org/view.php?id=4
Powell AR: Kilburn. KH: Pratt. PC. (1971). Short-term exposure to nitrogen dioxide: Effects on
pulmonary ultrastructure, compliance, and the surfactant system. Arch Intern Med 128: 74-80.
http://dx.doi.org/10.1001/archinte.1971.00310190078009
Downie. SR: Salome. CM: Verbanck. S: Thompson. B: Berend. N: King. GG. (2007). Ventilation
heterogeneity is a major determinant of airway hyperresponsiveness in asthma, independent of airway
inflammation. Thorax 62: 684-689. http://dx.doi.org/10.1136/thx.2006.069682
Drechsler-Parks. DM. (1995). Cardiac output effects of O3 and NO2 exposure in healthy older adults.
Toxicol Ind Health 11: 99-109. http://dx.doi.org/10.1177/074823379501100109
Dryden. 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.
http://www.ahrq.gov/research/findings/evidence-based-reports/eibeia-evidence-report.pdf
Eenhuizen. E: Gehring. U: Wijga. AH: Smit. HA: Fischer. PH: Brauer. M: Koppelman. GH: Kerkhof. M:
de Jongste. JC: Brunekreef. B: Hoek. G. (2013). Traffic related air pollution is related to interrupter
resistance in four-year old children. Eur Respir J 41: 1257-1263.
http://dx.doi.org/10.1183/09031936.00020812
Ehrlich. R. (1980). Interaction between environmental pollutants and respiratory infections. Environ
Health Perspect 35: 89-99.
Ehrlich. R: Findlav. JC: Fenters. JD: Gardner. DE. (1977). Health effects of short-term inhalation of
nitrogen dioxide and ozone mixtures. Environ Res 14: 223-231. http://dx.doi.org/10.1016/0013-
9351(77)90034-2
January 2015 5-369 DRAFT: Do Not Cite or Quote
-------
Ehrlich. R: Findlav. JC: Gardner. DE. (1979). Effects of repeated exposures to peak concentrations of
nitrogen dioxide and ozone on resistance to streptococcal pneumonia. J Toxicol Environ Health 5:
631-642. http://dx.doi.org/10.1080/15287397909529775
Elsaved. NM: Mustafa. MG. (1982). Dietary antioxidants and the biochemical response to oxidant
inhalation: I: Influence of dietary vitamin E on the biochemical effects of nitrogen dioxide exposure in
rat lung. Toxicol Appl Pharmacol 66: 319-328. http://dx.doi.org/10.1016/0041-008X(82)90298-8
Ensor. KB: Raun. LH: Persse. D. (2013). A case-crossover analysis of out-of-hospital cardiac arrest and
air pollution. Circulation 127: 1192-1199. http://dx.doi.org/10.1161/CIRCULATIONAHA.113.000027
Erjefalt JS: Greiff. L: Andersson. M: Matsson. E: Petersen. H: Linden. M: Ansari. T: Jeffery. PK:
Persson. CG. (1999). Allergen-induced eosinophil cytolysis is a primary mechanism for granule
protein release in human upper airways. Am J Respir Crit Care Med 160: 304-312.
http://dx.doi.0rg/10.1164/ajrccm.160.l.9809048
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
Esposito. S: Galeone. C: Lelii M: Longhj B: Ascolese. B: Senatore. L: Prada. E: Montinaro. V: Malerba.
S: Patria. MF: Principi N. (2014). Impact of air pollution on respiratory diseases in children with
recurrent wheezing or asthma. BMC Pulm Med 14: 130. http://dx.doi.org/10.1186/1471-2466-14-130
Ezrattv. V: Guillossou. G: Neukirch. C: Dehoux. M: Koscielny. S: Bonav. M: Cabanes. PA: Samet JM:
Mure. P: Ropert. L: Tokarek. S: Lambrozo. J: Aubier. M. (2014). Repeated nitrogen dioxide exposures
and eosinophilic airway inflammation in asthmatics: a randomized crossover study. Environ Health
Perspect 122: 850-855. http://dx.doi.org/10.1289/ehp.1307240
Faustinj A: Stafoggia. M: Colais. P: Berti G: Bisanti. L: Cadum. E: Cernigliaro. A: Mallone. S: Scarnato.
C: Forastiere. F. (2013). Air pollution and multiple acute respiratory outcomes. Eur Respir J 42: 304-
313. http://dx.doi.org/10.1183/09031936.00128712
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
Ferin. J: Leach. LJ. (1975). The effects of selected air pollutants on clearance of titanic oxide particles
from the lungs of rats. Inhaled Particles 4 Pt 1: 333-341.
Filho. MAP: Pereira. LAA: Arbex. FF: Arbex. M: Conceicao. GM: Santos. UP: Lopes. AC: Saldiva.
PHN: Braga. ALF: Cendon. S. (2008). Effect of air pollution on diabetes and cardiovascular diseases
in Sao Paulo, Brazil. Braz J Med Biol Res 41: 526-532. http://dx.doi.org/10.1590/S0100-
879X2008005000020
Flamant-Hulin. M: Caillaud. D: Sacco. P: Penard-Morand. C: Annesi-Maesano. I. (2010). Air pollution
and increased levels of fractional exhaled nitric oxide in children with no history of airway damage. J
Toxicol Environ Health A 73: 272-283. http://dx.doi.org/10.1080/15287390903249206
Folinsbee. LJ. (1992). Does nitrogen dioxide exposure increase airways responsiveness? Toxicol Ind
Health 8: 273-283.
Folinsbee. LJ: Horvath. SM: Bedi JF: Delehunt JC. (1978). Effect of 0.62 ppm NO2 on cardiopulmonary
function in young male nonsmokers. Environ Res 15: 199-205. http://dx.doi.org/10.1016/0013-
9351(78)90096-8
Forsberg. B: Stjernberg. N: Linne. R: Segerstedt B: Wall S. (1998). Daily air pollution levels and acute
asthma in southern Sweden. Eur Respir J 12: 900-905.
Fraenkel DJ: Bardin. PG: Sanderson. G: Lampe. F: Johnston. SL: Holgate. ST. (1995). Lower airways
inflammation during rhinovirus colds in normal and in asthmatic subjects. Am J Respir Crit Care Med
151: 879-886. http://dx.doi.Org/10.1164/airccm/151.3 Pt 1.879
January 2015 5-370 DRAFT: Do Not Cite or Quote
-------
Frampton. MW: Boscia. J: Roberts. NJ. Jr: Azadniv. M: Torres. A: Cox. C: Morrow. PE: Nichols. J:
Chalupa. D: Frasier. LM: Gibb. FR: Speers. DM: Tsai Y: Utell. MJ. (2002). Nitrogen dioxide
exposure: Effects on airway and blood cells. Am J Physiol 282: L155-L165.
Frampton. MW: Morrow. PE: Cox. C: Gibb. FR: Speers. DM: Utell MJ. (1991). Effects of nitrogen
dioxide exposure on pulmonary function and airway reactivity in normal humans. Am Rev Respir Dis
143: 522-527. http://dx.doi.Org/10.1164/ajrccm/143.3.522
Frampton. MW: Smeglin. AM: Roberts. NJ. Jr: Finkelstein. JN: Morrow. PE: Utell MJ. (1989). Nitrogen
dioxide exposure in vivo and human alveolar macrophage inactivation of influenza virus in vitro.
Environ Res 48: 179-192. http://dx.doi.org/10.1016/80013-9351(89)80033-7
Franck. U: Leitte. AM: Suppan. P. (2014). Multiple exposures to airborne pollutants and hospital
admissions due to diseases of the circulatory system in Santiago de Chile. Sci Total Environ 468-469:
746-756. http://dx.doi.0rg/10.1016/j.scitotenv.2013.08.088
Freedman. S: Lane. R: Gillett. MK: Guz. A. (1988). Abolition of methacholine induced
bronchoconstriction by the hyperventilation of exercise or volition. Thorax 43: 631-636.
Fung. KY: Khan. S: Krewski. D: Chen. Y. (2006). Association between air pollution and multiple
respiratory hospitalizations among the elderly in Vancouver, Canada. Inhal Toxicol 18: 1005-1011.
http://dx.doi.org/10.1080/08958370600904538
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.
Gardner. DE: Miller. FJ: Blommer. EJ: Coffin. PL. (1979). Influence of exposure mode on the toxicity of
NO2. Environ Health Perspect 30: 23-29.
Gass. K: Klein. M: Chang. HH: Flanders. WD: Strickland. MJ. (2014). Classification and regression trees
for epidemiologic research: an air pollution example. Environ Health 13: 17.
http://dx.doi.org/10.1186/1476-069X-13-17
Gent. JF: Koutrakis. P: Belanger. K: Triche. E: Holford. TR: Bracken. MB: Leaderer. BP. (2009).
Symptoms and medication use in children with asthma and traffic-related sources of fine particle
pollution. Environ Health Perspect 117: 1168-1174. http://dx.doi.org/10.1289/ehp.0800335
Ghosh. R: Joad. J: Benes. I: Postal. M: Sram. RJ: Hertz-Picciotto. I. (2012). Ambient nitrogen oxides
exposure and early childhood respiratory illnesses. Environ Int 39: 96-102.
http://dx.doi.0rg/10.1016/j.envint.2011.10.001
Gillespie-Bennett J: Pierse. N: Wickens. K: Crane. J: Howden-Chapman. P. (2011). The respiratory
health effects of nitrogen dioxide in children with asthma. Eur Respir J 38: 303-309.
http://dx.doi.org/10.1183/09031936.00115409
Gilmour. MI: Park. P: Selgrade. MK. (1996). Increased immune and inflammatory responses to dust mite
antigen in rats exposed to 5 ppm NO2. Fundam Appl Toxicol 31: 65-70.
http://dx.doi.org/10.1006/faat.1996.0076
Goggins. WB: Chan. EY: Yang. CY. (2013). Weather, pollution, and acute myocardial infarction in Hong
Kong and Taiwan. Int J Cardiol 168: 243-249. http://dx.doi.Org/10.1016/j.ijcard.2012.09.087
Goings. SAJ: Kulle. TJ: Bascom. R: Sauder. LR: Green. DJ: Hebel. JR: Clements. ML. (1989). Effect of
nitrogen dioxide exposure on susceptibility to influenza A virus infection in healthy adults. Am Rev
Respir Dis 139: 1075-1081. http://dx.doi.Org/10.1164/ajrccm/139.5.1075
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
Goldstein. BD: Hamburger. SJ: Falk. GW: Amoruso. MA. (1977). Effect of ozone and nitrogen dioxide
on the agglutination of rat alveolar macrophages by concanavalin A. Life Sci 21: 1637-1644.
January 2015 5-371 DRAFT: Do Not Cite or Quote
-------
Goldstein. E: Eagle. MC: Hoeprich. PP. (1973). Effect of nitrogen dioxide on pulmonary bacterial
defense mechanisms. Arch Environ Health 26: 202-204.
http://dx.doi.org/10.1080/00039896.1973.10666256
Goldstein. E: Warshauer. D: Lippert. W: Tarkington. B. (1974). Ozone and nitrogen dioxide exposure:
Murine pulmonary defense mechanisms. Arch Environ Health 28: 85-90.
http://dx.doi.org/10.1080/00039896.1974.10666444
Gong. H: Linn. WS: Clark. KW: Anderson. KR: Geller. MD: Sioutas. C. (2005). Respiratory responses to
exposures with fine particulates and nitrogen dioxide in the elderly with and without COPD. Inhal
Toxicol 17: 123-132. http://dx.doi.org/10.1080/08958370590904481
Goodman. JE: Chandalia. JK: Thakali S: Seelev. M. (2009). Meta-analysis of nitrogen dioxide exposure
and airway hyper-responsiveness in asthmatics. Crit Rev Toxicol 39: 719-742.
http://dx.doi.org/10.3109/10408440903283641
Graham. JA: Gardner. DE: Blommer. EJ: House. DE: Menache. MG: Miller. FJ. (1987). Influence of
exposure patterns of nitrogen dioxide and modifications by ozone on susceptibility to bacterial
infectious disease in mice. J Toxicol Environ Health 21: 113-125.
http://dx.doi.org/10.1080/15287398709531006
Greenwald. R: Sarnat SE: Ravsoni. AU: Li. WW: Johnson. BA: Stock. TH: Holguin. F: Sosa. T: Sarnat
JA. (2013). Associations between source-indicative pollution metrics and increases in pulmonary
inflammation and reduced lung function in a panel of asthmatic children. Air Qual Atmos Health 6:
487-499. http://dx.doi.org/10.1007/sll869-012-0186-3
Gregory. RE: Pickrell. JA: Hahn. FF: Hobbs. CH. (1983). Pulmonary effects of intermittent subacute
exposure to low-level nitrogen dioxide. J Toxicol Environ Health 11: 405-414.
http://dx.doi.org/10.1080/15287398309530354
Gryparis. A: Coull. BA: Schwartz. J. (2007). Controlling for confounding in the presence of measurement
error in hierarchical models: a Bayesian approach. J Expo Sci Environ Epidemiol 17: S20-S28.
http://dx.doi.org/10.1038/sj.jes.7500624
Guo. Y: Jia. Y: Pan. X: Liu. L: Wichmann. HE. (2009). The association between fine particulate air
pollution and hospital emergency room visits for cardiovascular diseases in Beijing, China. Sci Total
Environ 407: 4826-4830. http://dx.doi.Org/10.1016/j.scitotenv.2009.05.022
Guo. Y: Tong. S: Li. S: Barnett AG: Yu. W: Zhang. Y: Pan. X. (2010). Gaseous air pollution and
emergency hospital visits for hypertension in Beijing, China: a time-stratified case-crossover study.
Environ Health 9: 57. http://dx.doi.org/10.1186/1476-069X-9-57
Guth. DJ: Mavis. RD. (1986). The effect of lung alpha-tocopherol content on the acute toxicity of
nitrogen dioxide. Toxicol Appl Pharmacol 84: 304-314. http://dx.doi.org/10.1016/0041-
008X(86)90138-9
Hackney. JD: Thiede. FC: Linn. WS: Pedersen. EE: Spier. CE: Law. DC: Fischer. DA. (1978).
Experimental studies on human health effects of air pollutants. IV. Short-term physiological and
clinical effects of nitrogen dioxide exposure. Arch Environ Health 33: 176-181.
http://dx.doi.org/10.1080/00039896.1978.10667330
Hansel. NN: Brevsse. PN: Mccormack. MC: Matsui. EC: Curtin-Brosnan. J: Williams. DAL: Moore. JL:
Cuhran. JL: Diette. GB. (2008). A longitudinal study of indoor nitrogen dioxide levels and respiratory
symptoms in inner-city children with asthma. Environ Health Perspect 116: 1428-1432.
http://dx.doi.org/10.1289/ehp. 11349
Harre. ESM: Price. PD: Ayrev. RB: Toop. LJ: Martin. IR: Town. GI. (1997). Respiratory effects of air
pollution in chronic obstructive pulmonary disease: a three month prospective study. Thorax 52: 1040-
1044.
Hatch. GE: Slade. R: Selgrade. MK: Stead. AG. (1986). Nitrogen dioxide exposure and lung antioxidants
in ascorbic acid-deficient guinea pigs. Toxicol Appl Pharmacol 82: 351-359.
http://dx.doi.org/10.1016/0041-008X(86)90212-7
January 2015 5-372 DRAFT: Do Not Cite or Quote
-------
Hazucha. MJ: Folinsbee. LJ: Seal. E: Bromberg. PA. (1994). Lung function response of healthy women
after sequential exposures to NO2 and O3. Am J Respir Crit Care Med 150: 642-647.
http://dx.doi.0rg/10.1164/airccm.150.3.8087332
Hazucha. MJ: Ginsberg. JF: McDonnell WF: Haak. ED. Jr: Pimmel RL: Salaam. SA: House. DE:
Bromberg. PA. (1983). Effects of 0.1 ppm nitrogen dioxide on airways of normal and asthmatic
subjects. J Appl Physiol 54: 730-739.
Heckbert SR: Kooperberg. C: Safford. MM: Psatv. BM: Hsia. J: Mctiernan. A: Gaziano. JM: Frishman.
WH: Curb. JD. (2004). Comparison of self-report, hospital discharge codes, and adjudication of
cardiovascular events in the Women's Health Initiative. Am J Epidemiol 160: 1152-1158.
http://dx.doi.org/10.1093/aie/kwh314
HEI (Health Effects Institute). (2003). Revised analyses of time-series studies of air pollution and health
[HEI]. (HEI Special Report). Boston, MA: Health Effects Institute (HEI).
http://pubs.healtheffects.org/view. php?id=4
HEI (Health Effects Institute). (2010). Traffic-related air pollution: A critical review of the literature on
emissions, exposure, and health effects. (Special Report 17). Boston, MA: Health Effects Institute
(HEI). http://pubs.healtheffects.org/view.php?id=334
HEI (Health Effects Institute). (2012). Effects of short-term exposure to air pollution on hospital
admissions of young children for acute lower respiratory infections in Ho Chi Minh City, Vietnam (pp.
5-72; discussion 73-83). (ISSN 1041-5505 Research Report 169). Boston, MA: Health Effects
Institute, HEI Collaborative Working Group on Air Pollutino, Poverty, and Health in Ho Chi Minh
City. http://pubs.healtheffects.org/view.php?id=380
HEI Review Panel on Ultrafine Particles. (2013). Understanding the health effects of ambient ultrafine
particles. (HEI Perspectives 3). Boston, MA: Health Effects Institute.
http://pubs.healtheffects.org/view. php?id=394
Helledav. R: Huberman. D: Blomberg. A: Stiernberg. N: Sandstrom. T. (1995). Nitrogen dioxide exposure
impairs the frequency of the mucociliary activity in healthy subjects. Eur Respir J 8: 1664-1668.
http://dx.doi.org/10.1183/09031936.95.08101664
Helledav. R: Sandstrom. T: Stiernberg. N. (1994). Differences in bronchoalveolar cell response to
nitrogen dioxide exposure between smokers and nonsmokers. Eur Respir J 7: 1213-1220.
Henneberger. A: Zareba. W: Ibald-Mulli. A: Ruckerl. R: Cyrys. J: Couderc. JP: Mykins. B: Woelke. G:
Wichmann. HE: Peters. A. (2005). Repolarization changes induced by air pollution in ischemic heart
disease patients. Environ Health Perspect 113: 440-446. http://dx.doi.org/10.1289/ehp.7579
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
Higgins. BG: Francis. HC: Yates. C: Warburton. CJ: Fletcher. AM: Pickering. CAC: Woodcock. AA.
(2000). Environmental exposure to air pollution and allergens and peak flow changes. Eur Respir J 16:
66.
Higgins. BG: Francis. HC: Yates. CJ: Warburton. CJ: Fletcher. AM: Reid. JA: Pickering. CAC:
Woodcock. AA. (1995). Effects of air pollution on symptoms and peak expiratory flow measurements
in subjects with obstructive airways disease. Thorax 50: 149-155.
http://dx.doi.org/10.1136/thx.50.2.149
Hildebrandt K: Ruckerl. R: Koenig. W: Schneider. A: Pitz. M: Heinrich. J: Marder. V: Frampton. M:
Oberdorster. G: Wichmann. HE: Peters. A. (2009). Short-term effects of air pollution: A panel study of
blood markers in patients with chronic pulmonary disease. Part Fibre Toxicol 6: 25.
http://dx.doi.org/10.1186/1743-8977-6-25
January 2015 5-373 DRAFT: Do Not Cite or Quote
-------
Hiltermann. TJN: Stolk. J: van der Zee. SC: Brunekreef. B: de Bruijne. CR: Fischer. PH: Ameling. CB:
Sterk. PJ: Hiemstra. PS: van Bree. L. (1998). Asthma severity and susceptibility to air pollution. Eur
RespirJll: 686-693.
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. (2003). Daily mortality and air pollution in The Netherlands [HEI]. In Revised analyses of time-
series studies of air pollution and health (pp. 133-141). Boston, MA: Health Effects Institute.
http://pubs.healtheffects.org/getfile.php?u=21
Holguin. F: Flores. S: Ross. Z: Cortez. M: Molina. M: Molina. L: Rincon. C: Jerrett M: Berhane. K:
Granados. A: Romieu. I. (2007). Traffic-related exposures, airway function, inflammation, and
respiratory symptoms in children. Am J Respir Crit Care Med 176: 1236-1242.
http://dx.doi.org/10.1164/rccm.200611-1616OC
Hooftman. RN: Kuper. CF: Appelman. LM. (1988). Comparative sensitivity of histo-pathology and
specific lung parameters in the detection of lung injury. J Appl Toxicol 8: 59-65.
http://dx.doi.org/10.1002/jat.2550080110
Hovert. PL: Xu. J. (2012). Deaths: Preliminary data for 2011. National Vital Statistics Reports 61: 1-51.
Hsieh. YL: Yang. YH: Wu. TN: Yang. CY. (2010). Air pollution and hospital admissions for myocardial
infarction in a subtropical city: Taipei, Taiwan. J Toxicol Environ Health A 73: 757-765.
http://dx.doi.org/10.1080/15287391003684789
Huang. W: Zhu. T: Pan. X: Hu. M: Lu. SE: Lin. Y: Wang. T: Zhang. Y: Tang. X. (2012a). Air pollution
and autonomic and vascular dysfunction in patients with cardiovascular disease: Interactions of
systemic inflammation, overweight, and gender. Am J Epidemiol 176: 117-126.
http://dx.doi.org/10.1093/aje/kwr511
Huang. YC: Rappold. AG: Graff. DW: Ohio. AJ: Devlin. RB. (2012b). Synergistic effects of exposure to
concentrated ambient fine pollution particles and nitrogen dioxide in humans. Inhal Toxicol 24: 790-
797. http://dx.doi.org/10.3109/08958378.2012.718809
Hubbard. AK: Symanowicz. PT: Thibodeau. M: Thrall. RS: Schramm. CM: Cloutier. MM: Morris. JB.
(2002). Effect of nitrogen dioxide on ovalbumin-induced allergic airway disease in a murine model. J
Toxicol Environ Health A 65: 1999-2005. http://dx.doi.org/10.1080/00984100290071397
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.
Illing. JW: Miller. FJ: Gardner. DE. (1980). Decreased resistance to infection in exercised mice exposed
to NO2 and O3. J Toxicol Environ Health 6: 843-851. http://dx.doi.org/10.1080/15287398009529902
Inman. MD: Watson. RM: Killian. KJ: O'Byrne. PM. (1990). Methacholine airway responsiveness
decreases during exercise in asthmatic subjects. Am Rev Respir Dis 141: 1414-1417.
http://dx.doi.0rg/10.1164/airccm/141.6.1414
Iskandar. A: Andersen. ZJ: B0nnelykke. K: Ellermann. T: Andersen. KK: Bisgaard. H. (2012). Coarse and
fine particles but not ultrafine particles in urban air trigger hospital admission for asthma in children.
Thorax 67: 252-257. http://dx.doi.org/10.1136/thoraxinl-2011-200324
Ito. K: Mathes. R: Ross. Z: Nadas. A: Thurston. G: Matte. T. (2011). Fine paniculate matter constituents
associated with cardiovascular hospitalizations and mortality in New York City. Environ Health
Perspect 119: 467-473. http://dx.doi.org/10.1289/ehp.1002667
Ito. K: Thurston. GD: Silverman. RA. (2007a). Association between coarse particles and asthma
emergency department (ED) visits in New York City. American Thoracic Society International
Conference 2007, San Francisco, CA.
January 2015 5-374 DRAFT: Do Not Cite or Quote
-------
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
Jackson. AC: Murphy. MM: Rassulo. J: Celli. BR: Ingram. RH. (2004). Deep breath reversal and
exponential return of methacholine-induced obstruction in asthmatic and nonasthmatic subjects. J
Appl Physiol 96: 137-142. http://dx.doi.org/10.1152/iapplphysiol.00504.2003
Jaffe. DH: Singer. ME: Rimm. AA. (2003). Air pollution and emergency department visits for asthma
among Ohio Medicaid recipients, 1991-1996. Environ Res 91: 21-28.
http://dx.doi.org/10.1016/S0013-9351(02)00004-X
Jakab. GJ. (1988). Modulation of pulmonary defense mechanisms against viral and bacterial infections by
acute exposures to nitrogen dioxide (pp. 1-38). (ISSN 1041-5505 HEI Research Report no. 20).
Cambridge, MA: Health Effects Institute. http://pubs.healtheffects.org/view.php?id=238
Jalaludin. B: Khalaj. B: Sheppeard. V: Morgan. G. (2008). Air pollution and ED visits for asthma in
Australian children: A case-crossover analysis. Int Arch Occup Environ Health 81: 967-974.
http://dx.doi.org/10.1007/s00420-007-0290-0
Jalaludin. B: Morgan. G: Lincoln. D: Sheppeard. V: Simpson. R: Corbett S. (2006). Associations between
ambient air pollution and daily emergency department attendances for cardiovascular disease in the
elderly (65+ years), Sydney, Australia. J Expo Sci Environ Epidemiol 16: 225-237.
http://dx.doi.org/10.1038/sj.jea.7500451
Jalaludin. BB: OToole. 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/S0013-9351(03)00038-0
Jenkins. HS: Devalia. JL: Mister. RL: Sevan. AM: Rusznak. C: Davies. RJ. (1999). The effect of
exposure to ozone and nitrogen dioxide on the airway response of atopic asthmatics to inhaled
allergen: dose- and time-dependent effects. Am J Respir Crit Care Med 160: 33-39.
http://dx.doi.0rg/10.1164/ajrccm.160.l.9808119
Jevtic. M: Dragic. N: Bijelovic. S: Popovic. M. (2014). Cardiovascular diseases and air pollution in Novi
Sad, Serbia. Int J Occup Med Environ Health 27: 153-164. http://dx.doi.org/10.2478/sl3382-014-
0239-v
Jia. X: Song. X: Shima. M: Tamura. K: Deng. F: Guo. X. (2011). Acute effect of ambient ozone on heart
rate variability in healthy elderly subjects. J Expo Sci Environ Epidemiol 21: 541-547.
http://dx.doi.org/10.1038/ies.2011.18
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. http://dx.doi.Org/10.1164/airccm.164.5.2012125
Torres. R: Magnussen. H. (1990). Airways response of asthmatics after a 30 min exposure, at resting
ventilation, to 0.25 ppm NO2 or 0.5 ppm SO2. Eur Respir J 3: 132-137.
Torres. R: Magnussen. H. (1991). Effect of 0.25 ppm nitrogen dioxide on the airway response to
methacholine in asymptomatic asthmatic patients. Lung 169: 77-85.
Torres. R: Nowak. D: Grimminger. F: Seeger. W: Oldigs. M: Magnussen. H. (1995). The effect of 1 ppm
nitrogen dioxide on bronchoalveolar lavage cells and inflammatory mediators in normal and asthmatic
subjects. Eur Respir J 8: 416-424. http://dx.doi.org/10.1183/09031936.95.08030416
Just. J: Segala. C: Sahraoui F: Priol. G: Grimfeld. A: Neukirch. F. (2002). Short-term health effects of
particulate and photochemical air pollution in asthmatic children. Eur Respir J 20: 899-906.
http://dx.doi.org/10.1183/09031936.02.00236902
January 2015 5-375 DRAFT: Do Not Cite or Quote
-------
Kan. H: Chen. B: Zhao. N: London. SJ: Song. G: Chen. G: Zhang. Y: Jiang. L. (2010). Part 1. Atime-
series study of ambient air pollution and daily mortality in Shanghai, China. 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. 17-78). (ISSN 1041-5505). Boston, MA: Health Effects Institute.
http://pubs.healtheffects.org/view. php?id=348
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
Karakatsanj A: Analitis. A: Perifanou. D: Ayres. JG: Harrison. RM: Kotronarou. A: Kavouras. IG:
Pekkanen. J: Hameri. K: Kos. GP: de Hartog. JJ: Hoek. G: Katsouvanni. K. (2012). Paniculate matter
air pollution and respiratory symptoms in individuals having either asthma or chronic obstructive
pulmonary disease: A European multicentre panel study. Environ Health 11: 75.
http://dx.doi.org/10.1186/1476-069X-ll-75
Karr. C: Lumlev. T: Shepherd. K: Davis. R: Larson. T: Ritz. B: Kaufman. J. (2006). A case-crossover
study of wintertime ambient air pollution and infant bronchiolitis. Environ Health Perspect 114: 277-
281. http://dx.doi.org/10.1289/ehp.8313
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) (pp. 5-90). (ISSN
1041-5505 Research Report 142). Boston, MA: Health Effects Institute.
http://pubs.healtheffects.org/view. php?id=327
Katsouvanni. K: Touloumi. G: Samoli. E: Gryparis. A: Le Tertre. A: Monopolis. Y: Rossi. G: Zmirou. D:
Ballester. F: Boumghar. A: Anderson. HR: Wojtyniak. 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.
Katsouvanni. K: Touloumi. G: Samoli. E: Petasakis. Y: Analitis. A: Le Tertre. A: Rossi. G: Zmirou. D:
Ballester. F: Boumghar. A: Anderson. HR: Wojtyniak. B: Paldv. A: Braunstein. R: Pekkanen. J:
Schindler. C: Schwartz. J. (2003). Sensitivity analysis of various models of short-term effects of
ambient particles on total mortality in 29 cities in APHEA2. In Revised Analyses of Time-series
Studies of Air Pollution and Health (pp. 157-164). Boston, MA: Health Effects Institute.
Kehrl. HR: Peden. DB: Ball. B: Folinsbee. LJ: Horstman. D. (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. http://dx.doi.org/10.1016/S0091-6749(99)70013-8
Kelishadi R: Mirghaffari. N: Poursafa. P: Gidding. SS. (2009). Lifestyle and environmental factors
associated with inflammation, oxidative stress and insulin resistance in children. Atherosclerosis 203:
311-319. http://dx.doi.0rg/10.1016/j.atherosclerosis.2008.06.022
Kelly. FJ: Blomberg. A: Frew. A: Holgate. ST: Sandstrom. T. (1996). Antioxidant kinetics in lung lavage
fluid following exposure of humans to nitrogen dioxide. Am J Respir Crit Care Med 154: 1700-1705.
http://dx.doi.0rg/10.1164/airccm.154.6.8970358
Khafaie. MA: Salvi SS: Ojha. A: Khafaie. B: Gore. SS: Yajnik. CS. (2013). Systemic inflammation (C-
reactive protein) in type 2 diabetic patients is associated with ambient air pollution in Pune City, India.
Diabetes Care 36: 625-630. http://dx.doi.org/10.2337/dcl2-0388
Kharitonov. SA: Barnes. PJ. (2000). Clinical aspects of exhaled nitric oxide [Review]. Eur Respir J 16:
781-792.
Kikuya. M: Hozawa. A: Ohokubo. T: Tsuji I: Michimata. M: Matsubara. M: Ota. M: Nagai K: Arakj T:
Satoh. H: Ito. S: Hisamichi. S: Imai Y. (2000). Prognostic significance of blood pressure and heart
rate variabilities: the Ohasama study. Hypertension 36: 901-906.
http://dx.doi.Org/10.1161/01.HYP.36.5.901
January 2015 5-376 DRAFT: Do Not Cite or Quote
-------
Kilkenny. C: Browne. WJ: Cuthill 1C: Emerson. M: Altaian. DG. (2010). Improving bioscience research
reporting: the ARRIVE guidelines for reporting animal research [Review]. PLoS Biol 8: e!000412.
http://dx.doi.org/10.1371/journal.pbio.1000412
Kim. JH: Hong. YC. (2012). GSTM1, GSTT1, and GSTP1 polymorphisms and associations between air
pollutants and markers of insulin resistance in elderly Koreans. Environ Health Perspect 120: 1378-
1384. http://dx.doi.org/10.1289/ehp.1104406
Kim. S: Kim. Y: Lee. MR: Kim. J: Jung. A: Park. JS: Jang. AS: Park. SW: Uh. ST: Choi. JS: Kim. YH:
Buckley. T: Park. CS. (2012). Winter season temperature drops and sulfur dioxide levels affect on
exacerbation of refractory asthma in South Korea: a time-trend controlled case-crossover study using
soonchunhyang asthma cohort data. J Asthma 49: 679-687.
http://dx.doi.org/10.3109/02770903.2012.702839
Kinsella. J: Carter. R: Reid. WH: Campbell D: Clark. CJ. (1991). Increased airways reactivity after
smoke inhalation. Lancet 337: 595-597.
Kjaergaard. SK: Rasmussen. TR. (1996). Clinical studies of effects of nitrogen oxides in healthy and
asthmatic subjects. Cent Eur J Public Health 4: 23-26.
Kleinman. MT: Bailey. RM: Linn. WS: Anderson. KR: Whynot JD: Shamoo. DA: Hackney. JD. (1983).
Effects of 0.2 ppm nitrogen dioxide on pulmonary function and response to bronchoprovocation in
asthmatics. J Toxicol Environ Health 12: 815-826. http://dx.doi.org/10.1080/15287398309530472
Klimisch. HJ: Andreae. M: Tillmann. U. (1997). A systematic approach for evaluating the quality of
experimental toxicological and ecotoxicological data. Regul Toxicol Pharmacol 25: 1-5.
http://dx.doi.org/10.1006/rtph.1996.1076
Ko. FW: Tarn. W: Wong. TW: Chan. DP: Tung. AH: Lai CKW: Hui DSC. (2007a). Temporal
relationship between air pollutants and hospital admissions for chronic obstructive pulmonary disease
in Hong Kong. Thorax 62: 780-785. http://dx.doi.org/10.1136/thx.2006.076166
Ko. FWS: Tarn. W: Wong. TW: Lai. CKW: Wong. GWK: Leung. TF: Ng. SSS: Hui. DSC. (2007b).
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/i.1365-2222.2007.02791.x
Koenig. JQ: Covert PS: 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. http://dx.doi.Org/10.1164/ajrccm/136.5.1152
Kulle. TJ: Clements. ML. (1988). Susceptibility to virus infection with exposure to nitrogen dioxide (pp.
5-21). (ISSN 1041-5505). Cambridge, MA: Health Effects Institute.
http://pubs.healtheffects.org/view. php?id=243
Kunimoto. M: Mochitate. K: Kava. K: Miura. T: Kubota. K. (1984). Effects of nitrogen dioxide on red
blood cells of rats: Alterations of cell membrane components and populational changes of red blood
cells during in vivo exposure to NO2. Environ Res 33: 361-369.
La Rovere. MT: Pinna. GD: Maestri. R: Mortara. A: Capomolla. S: Febo. O: Ferrari. R: Franchinj M:
Gnemmj M: Opasich. C: Riccardi PG: Traversi E: Cobelli F. (2003). Short-term heart rate
variability strongly predicts sudden cardiac death in chronic heart failure patients. Circulation 107:
565-570.
Lagorio. S: Forastiere. F: Pistelli R: lavarone. I: Michelozzj P: Fano. V: Marconi. A: Ziemackj 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
Langrish. JP: Lundba'ck. M: Barath. S: Soderberg. S: Mills. NL: Newbv. DE: Sandstrom. T: Blomberg. A.
(2010). Exposure to nitrogen dioxide is not associated with vascular dysfunction in man. Inhal Toxicol
22: 192-198. http://dx.doi.org/10.3109/08958370903144105
January 2015 5-377 DRAFT: Do Not Cite or Quote
-------
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. http://dx.doi.Org/10.1016/j.scitotenv.2007.07.025
Last JA: Warren. PL. (1987). Synergistic interaction between nitrogen dioxide and respirable aerosols of
sulfuric acid or sodium chloride on rat lungs. Toxicol Appl Pharmacol 90: 34-42.
http://dx.doi.org/10.1016/0041-008X(87)90303-6
Laube. BL: Norman. PS: Adams. GK. (1992). The effect of aerosol distribution on airway responsiveness
to inhaled methacholine in patients with asthma. J Allergy Clin Immunol 89: 510-518.
http://dx.doi.org/10.1016/0091-6749(92)90317-U
Laumbach. RJ: Rich. DQ: Gandhi. S: Amorosa. L: Schneider. S: Zhang. J: Ohman-Strickland. P: Gong. J:
Lelyanov. O: Kipen. HM. (2010). Acute changes in heart rate variability in subjects with diabetes
following a highway traffic exposure. J Occup Environ Med 52: 324-331.
http://dx.doi.org/10.1097/JOM.Ob013e3181d241fa
Laurent. O: Pedrono. G: Filleul L: Segala. C: Lefranc. A: Schillinger. C: Riviere. E: Bard. D. (2009).
Influence of socioeconomic deprivation on the relation between air pollution and beta-agonist sales for
asthma. Chest 135: 717-723. http://dx.doi.org/10.1378/chest.08-1604
Lee. PC: Talbott EO: Roberts. JM: Catov. JM: Sharma. RK: Ritz. B. (2011). Paniculate air pollution
exposure and C-reactive protein during early pregnancy. Epidemiology 22: 524-531.
http://dx.doi.org/10.1097/EDE.Ob013e31821c6c58
Lepeule. J: Bind. MA: Baccarelli. AA: Koutrakis. P: Tarantini. L: Litonjua. A: Sparrow. D: Vokonas. P:
Schwartz. JD. (2014). Epigenetic influences on associations between air pollutants and lung function
in elderly men: the normative aging study. Environ Health Perspect 122: 566-572.
http://dx.doi.org/10.1289/ehp.1206458
Lew. D: Sheppard. L: Checkowav. H: Kaufman. J: Lumlev. T: Koenig. J: Siscovick. D. (2001). A case-
crossover analysis of paniculate matter air pollution and out-of-hospital primary cardiac arrest.
Epidemiology 12: 193-199.
Li. H: Han. M: Guo. L: Li. G: Sang. N. (201 la). Oxidative stress, endothelial dysfunction and
inflammatory response in rat heart to NO inhalation exposure. Chemosphere 82: 1589-1596.
http://dx.doi.0rg/10.1016/j.chemosphere.2010.ll.055
Li S: Batterman. S: Wasilevich. E: Wahl. R: Wirth. J: Su. FC: Mukherjee. B. (2011b). Association of
daily asthma emergency department visits and hospital admissions with ambient air pollutants among
the pediatric Medicaid population in Detroit: Time-series and time-stratified case-crossover analyses
with threshold effects. Environ Res 111: 1137-1147. http://dx.doi.Org/10.1016/j.envres.2011.06.002
Lin. M: Stieb. DM: Chen. Y. (2005). Coarse particulate matter and hospitalization for respiratory
infections in children younger than 15 years in Toronto: A case-crossover analysis. Pediatrics 116:
235-240. http://dx.doi.org/10.1542/peds.2004-2012
Lin. W: Huang. W: Zhu. T: Hu. M: Brunekreef. B: Zhang. Y: Liu. X: Cheng. H: Gehring. U: Li. C: Tang.
2L (2011). Acute respiratory inflammation in children and black carbon in ambient air before and
during the 2008 Beijing Olympics. Environ Health Perspect 119: 1507-1512.
http://dx.doi.org/10.1289/ehp.1103461
Linares. C: Diaz. J. (2010). Short-term effect of concentrations of fine particulate 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
Link. MS: Luttmann-Gibson. H: Schwartz. J: Mittleman. MA: Wessler. B: Gold. PR: Dockery. DW:
Laden. F. (2013). Acute exposure to air pollution triggers atrial fibrillation. J Am Coll Cardiol 62: 816-
825. http://dx.doi.0rg/10.1016/i.iacc.2013.05.043
January 2015 5-378 DRAFT: Do Not Cite or Quote
-------
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: Spier. CE: Valencia. LM: Anzar. UT: Venet TG: Avol EL: Hackney. JD.
(1985a). Controlled exposure of volunteers with chronic obstructive pulmonary disease to nitrogen
dioxide. Arch Environ Health 40: 313-317. http://dx.doi.org/10.1080/00039896.1985.10545939
Linn. WS: Solomon. JC: Trim. SC: Spier. CE: Shamoo. DA: Venet. TG: Avol EL: Hackney. JD. (1985b).
Effects of exposure to 4 ppm nitrogen dioxide in healthy and asthmatic volunteers. Arch Environ
Occup Health 40: 234-239.
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.
Liu. C: Flexeder. C: Fuertes. E: Cyrys. J: Bauer. CP: Koletzko. S: Hoffmann. B: von Berg. A: Heinrich. J:
Groups. ftGaLS. (2014a). Effects of air pollution on exhaled nitric oxide in children: Results from the
GINIplus and LISAplus studies. Int J Hyg Environ Health 217: 483-491.
http://dx.doi.0rg/10.1016/j.ijheh.2013.09.006
Liu. L. (2013). Email from Dr. Liu to Dr. Patel; Response to data request. Available online at
http://www.regulations.gov/#!documentDetail:D=EPA-HQ-ORD-2013-0232-0013
Liu. L: Kauri. LM: Mahmud. M: Weichenthal S: Cakmak. S: Shutt R: You. H: Thomson. E: Vincent. R:
Kumarathasan. P: Broad. G: Dales. R. (2014b). Exposure to air pollution near a steel plant and effects
on cardiovascular physiology: A randomized crossover study. Int J Hyg Environ Health 217: 279-286.
http://dx.doi.0rg/10.1016/j.ijheh.2013.06.007
Liu. L: Poon. R: Chen. L: Frescura. AM: Montuschj P: Ciabattoni. G: Wheeler. A: Dales. R. (2009).
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
Ljungman. P: Bellander. T: Schneider. A: Breitner. S: Forastiere. F: Hampel. R: Illig. T: Jacquemin. B:
Katsouvanni. K: von Klot. S: Koenig. W: Lanki. T: Nyberg. F: Pekkanen. J: Pistelli. R: Pitsavos. C:
Rosenqvist. M: Sunyer. J: Peters. A. (2009). Modification of the interleukin-6 response to air pollution
by interleukin-6 and fibrinogen polymorphisms. Environ Health Perspect 117: 1373-1379.
http://dx.doi.org/10.1289/ehp.0800370
Ljungman. PL: Wilker. EH: Rice. MB: Schwartz. J: Gold. PR: Koutrakis. P: Vita. JA: Mitchell GF:
Vasan. RS: Benjamin. EJ: Mittleman. MA: Hamburg. NM. (2014). Short-term exposure to air
pollution and digital vascular function. Am J Epidemiol 180: 482-489.
http://dx.doi.org/10.1093/aje/kwul61
Ljungman. PLS: Berglind. N: Holmgren. C: Gadler. F: Edvardsson. N: Pershagen. G: Rosenqvist M:
Sjogren. B: Bellander. T. (2008). Rapid effects of air pollution on ventricular arrhythmias. Eur Heart J
29: 2894-2901. http://dx.doi.org/10.1093/eurheartj/ehn463
Llorca. J: Salas. A: Prieto-Salceda. D: Chinchon-Bengoechea. V: Delgado-Rodriguez. M. (2005).
Nitrogen dioxide increases cardiorespiratory admissions in Torrelavega (Spain). J Environ Health 68:
30-35.
Lu. C: Deng. Q: Yu. CWF: Sundell J: Ou. C. (2014). Effects of ambient air pollution on the prevalence of
pneumonia in children: Implication for National Ambient Air Quality Standards in China. Indoor Built
Environ 23: 259-269. http://dx.doi.org/10.1177/1420326X13504423
Lu. KD: Brevsse. PN: Diette. GB: Curtin-Brosnan. J: Aloe. C: Williams. PL: Peng. RD: Mccormack.
MC: Matsui. EC. (2013). Being overweight increases susceptibility to indoor pollutants among urban
children with asthma. J Allergy Clinlmmunol 131: 1017-1023.e!013.
http://dx.doi.0rg/10.1016/i.iaci.2012.12.1570
Madsen. C: Durand. KL: Nafstad. P: Schwarze. PE: R0nningen. KS: Haheim. LL. (2008). Associations
between environmental exposures and serum concentrations of Clara cell protein among elderly men
in Oslo, Norway. Environ Res 108: 354-360. http://dx.doi.Org/10.1016/j.envres.2008.07.017
January 2015 5-379 DRAFT: Do Not Cite or Quote
-------
Maestrelli. P: Canova. C: Scapellato. ML: Visentin. A: Tessarj R: Bartolucci GB: Simonato. L: Lotti M.
(2011). Personal exposure to paniculate matter is associated with worse health perception in adult
asthma. J Investig Allergol Clin Immunol 21: 120-128.
Magnussen. H: Reuss. G: Torres. R. (1986). Airway response to methacholine during exercise induced
refractoriness in asthma. Thorax 41: 667-670.
Malm-ErjeFalt M: Persson. CG: ErjeFalt JS. (2001). Degranulation status of airway tissue eosinophils in
mouse models of allergic airway inflammation. Am J Respir Cell Mol Biol 24: 352-359.
http://dx.doi.0rg/10.1165/ajrcmb.24.3.4357
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
Mann. JK: Tager. IB: Lurmann. F: Segal M: Quesenberry. CP. Jr: Lugg. MM: Shan. J: Van den Eeden.
SK. (2002). Air pollution and hospital admissions for ischemic heart disease in persons with
congestive heart failure or arrhythmia. Environ Health Perspect 110: 1247-1252.
Martins. P. (2013). Email from Dr. Martins to Dr Patel; Response to data request. Martins, P.
Martins. PC: Valente. J: Papoila. AL: Caires. I: Araujo-Martins. J: Mata. P: Lopes. M: Torres. S: Rosado-
Pinto. J: Borrego. C: Annesi-Maesano. I: Neuparth. N. (2012). Airways changes related to air
pollution exposure in wheezing children. Eur Respir J 39: 246-253.
http://dx.doi.org/10.1183/09031936.00025111
McCreanor. J: Cullinan. P: Nieuwenhuijsen. MJ: Stewart-Evans. J: Malliarou. E: Jarup. L: Harrington. R:
Svartengren. M: Han. IK: Ohman-Strickland. P: Chung. KF: Zhang. J. (2007). Respiratory effects of
exposure to diesel traffic in persons with asthma. N Engl J Med 357: 2348-2358.
http://dx.doi.org/10.1056/NEJMoa071535
Mehta. S: Ngo. LH: Dzung. DV: Cohen. A: Thach. TQ: Dan. VX: Tuan. ND: Giang. LT. (2013). Air
pollution and admissions for acute lower respiratory infections in young children of Ho Chi Minh
City. Air Qual Atmos Health 6: 167-179. http://dx.doi.org/10.1007/sll869-011-0158-z
Meng. X: Wang. C: Cao. D: Wong. CM: Kan. H. (2013). Short-term effect of ambient air pollution on
COPD mortality in four Chinese cities. Atmos Environ 77: 149-154.
http://dx.doi.0rg/10.1016/j.atmosenv.2013.05.001
Mersch. J: Dyce. BJ: Haverback. BJ: Sherwin. RP (1973). Diphosphoglycerate content of red blood cells:
Measurements in guinea pigs exposed to 04 ppm nitrogen dioxide. Arch Environ Health 27: 94-95.
http://dx.doi.org/10.1080/00039896.1973.10666326
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
Min. KB: Min. JY: Cho. SI: Paek. D. (2008). The relationship between air pollutants and heart-rate
variability among community residents in Korea. Inhal Toxicol 20: 435-444.
http://dx.doi.org/10.1080/08958370801903834
Mochitate. K: Miura. T. (1984). In vivo effect of nitrogen dioxide on the activities of glycolytic enzymes
in red blood cells of rats. Toxicol Lett 22: 315-321.
Mochitate. K: Takahashi. Y: Ohsumi. T: Miura. T. (1986). Activation and increment of alveolar
macrophages induced by nitrogen dioxide. J Toxicol Environ Health 17: 229-239.
http://dx.doi.org/10.1080/15287398609530819
Mohsenin. V. (1987a). Airway responses to nitrogen dioxide in asthmatic subjects. J Toxicol Environ
Health 22: 371-380. http://dx.doi.org/10.1080/15287398709531080
Mohsenin. V. (1987b). Effect of vitamin C on NO2-induced airway hyperresponsiveness in normal
subjects: Arandomized double-blind experiment. Am Rev Respir Dis 136: 1408-1411.
http://dx.doi.0rg/10.1164/airccm/136.6.1408
January 2015 5-380 DRAFT: Do Not Cite or Quote
-------
Mohsenin. V. (1988). Airway responses to 2.0 ppm nitrogen dioxide in normal subjects. Arch Environ
Health 43: 242-246. http://dx.doi.org/10.1080/00039896.1988.9934941
Mohsenin. V. (1991). Lipid peroxidation and antielastase activity in the lung under oxidant stress: Role of
antioxidant defenses. J Appl Physiol 70: 1456-1462.
Moolgavkar. SH. (2003). Air pollution and daily deaths and hospital admissions in Los Angeles and Cook
counties. In Revised Analyses of the National Morbidity, Mortality, and Air Pollution Study, Part II
(pp. 183-198). Boston, MA: Health Effects Institute.
Moolgavkar. SH: Mcclellan. RO: Dewanji. A: Turim. J: Luebeck. EG: Edwards. M. (2013). Time-series
analyses of air pollution and mortality in the United States: A subsampling approach. Environ Health
Perspect 121: 73-78. http://dx.doi.org/10.1289/ehp.1104507
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. http://dx.doi.org/10.2105/AJPH.88.12.1761
Morrow. PE: Utell. MJ. (1989a). Responses of susceptible subpopulations to nitrogen dioxide [HEI] (pp.
1-45). (ISSN 1041-5505 Research Report 23). Cambridge, MA: Health Effects Institute.
http://pubs.healtheffects.org/view. php?id=235
Morrow. PE: Utell MJ. (1989b). Responses of susceptible subpopulations to nitrogen dioxide: Appendix
H. Selected raw data [HEI]. (Research Report 23). Cambridge, MA: Health Effects Institute.
Morrow. PE: Utell MJ: Bauer. MA: Smeglin. AM: Frampton. MW: Cox. C: Speers. DM: Gibb. FR.
(1992). Pulmonary performance of elderly normal subjects and subjects with chronic obstructive
pulmonary disease exposed to 0.3 ppm nitrogen dioxide. Am J Respir Crit Care Med 145: 291-300.
http://dx.doi.0rg/10.1164/airccm/145.2 Pt 1.291
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
Moshammer. H: Hutter. HP: Hauck. H: Neuberger. M. (2006). Low levels of air pollution induce changes
of lung function in a panel of schoolchildren. Eur Respir J 27: 1138-1143.
http://dx.doi.org/10.1183/09031936.06.00089605
Moss. OR: Oldham. MJ. (2006). Dosimetry counts: Molecular hypersensitivity may not drive pulmonary
hyperresponsiveness. J Aerosol Med 19: 555-564. http://dx.doi.org/10.1089/jam.2006.19.555
Muhajarine. N: Mustard. C: Roos. LL: Young. TK: Gelskev. DE. (1997). Comparison of survey and
physician claims data for detecting hypertension. J Clin Epidemiol 50: 711-718.
Miiller. B: Schafer. H: Barth. P: von Wichert. P. (1994). Lung surfactant components in bronchoalveolar
lavage after inhalation of NO2 as markers of altered surfactant metabolism. Lung 172: 61-72.
http://dx.doi.org/10.1007/BF00185077
Murgia. N: Brisman. J: Claesson. A: Muzi. G: Olin. AC: Toren. K. (2014). Validity of a questionnaire-
based diagnosis of chronic obstructive pulmonary disease in a general population-based study. BMC
Pulm Med 14: 49. http://dx.doi.org/10.1186/1471-2466-14-49
Murray. AB: Ferguson. AC: Morrison. B. (1981). Airway responsiveness to histamine as a test for overall
severity of asthma in children. J Allergy Clin Immunol 68: 119-124. http://dx.doi.org/10.1016/0091-
6749(81)90169-X
Mustafa. MG: Elsaved. NM: von Dohlen. FM: Hassett CM: Postlethwait EM: Quinn. CL: Graham. JA:
Gardner. DE. (1984). Acomparison of biochemical effects of nitrogen dioxide, ozone, and their
combination in mouse lung. Toxicol Appl Pharmacol 72: 82-90.
January 2015 5-381 DRAFT: Do Not Cite or Quote
-------
NAEPP (National Asthma Education and Prevention Program). (2007). Expert panel report 3: guidelines
for the diagnosis and management of asthma. J Allergy Clin Immunol 120: S93.
http://dx.doi.0rg/10.1016/i.iaci.2007.09.043
Nakajima. T: Kusumoto. S. (1968). Effect of nitrogen dioxide exposure on the contents of reduced
glutathione in mouse lung. Food Sanitation Research (English) 6: 17-21.
Nuvolone. D: Balzi. D: Chinj M: Scala. D: Giovannini. F: Barchielli A. (2011). Short-term association
between ambient air pollution and risk of hospitalization for acute myocardial infarction: Results of
the cardiovascular risk and air pollution in Tuscany (RISCAT) study. Am J Epidemiol 174: 63-71.
http://dx.doi.org/10.1093/aje/kwr046
O'Byrne. PM: Gauvreau. GM: Brannan. JD. (2009). Provoked models of asthma: What have we learnt?
[Review]. Clin Exp Allergy 39: 181-192. http://dx.doi.Org/10.llll/i.1365-2222.2008.03172.x
O'Connor. GT: Neas. L: Vaughn. B: Kattan. M: Mitchell H: Grain. EF: Evans. R. Ill: 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/j.jaci.2008.02.020
Odajima. H: Yamazaki. S: Nitta. H. (2008). Decline in peak expiratory flow according to hourly short-
term concentration of particulate matter in asthmatic children. Inhal Toxicol 20: 1263-1272.
http://dx.doi.org/10.1080/08958370802311151
Oftedal B: Brunekreef. B: Nystad. W: Madsen. C: Walker. SE: Nafstad. P. (2008). Residential outdoor air
pollution and lung function in schoolchildren. Epidemiology 19: 129-137.
http://dx.doi.org/10.1097/EDE.Ob013e31815c0827
Ohashi. Y: Nakai. Y: Sugiura. Y: Ohno. Y: Okamoto. H: Tanaka. A: Kakinoki. Y: Havashi. M. (1994).
Nitrogen dioxide-induced eosinophilia and mucosal injury in the nose of the guinea pig. Acta
Otolaryngol 114: 547-551.
Orazzo. F: Nespoli L: Ito. K: Tassinarj 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
Orehek. J: Massarj JP: Gayrard. P: Grimaud. C: Charpin. J. (1976). Effect of short-term, low-level
nitrogen dioxide exposure on bronchial sensitivity of asthmatic patients. J Clin Invest 57: 301-307.
http://dx.doi.org/10.1172/JCI108281
Orehek. J: Nicoli MM: Delpierre. S: Beaupre. A. (1981). Influence of the previous deep inspiration on
the spirometric measurement of provoked bronchoconstriction in asthma. Am Rev Respir Dis 123:
269-272.
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
Padhi. BK: Padhy. PK. (2008). Domestic fuels, indoor air pollution, and children's health. Ann N Y Acad
Sci 1140: 209-217. http://dx.doi.org/10.1196/annals.1454.015
Pagani. P: Romano. M: Erroi. A: Ferro. M: Salmona. M. (1994). Biochemical effects of acute and
subacute nitrogen dioxide exposure in rat lung and bronchoalveolar lavage fluid. Arch Environ
Contam Toxicol 27: 426-430. http://dx.doi.org/10.1007/BF00213182
Park. JW: Lim. YH: Kyung. SY: An. CH: Lee. SP: Jeong. SH: Ju. YS. (2005). 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/j.1440-1843.2005.00728.x
January 2015 5-382 DRAFT: Do Not Cite or Quote
-------
Park. SK: Auchincloss. AH: O'Neill MS: Prineas. R: Correa. JC: Keeler. J: Barr. RG: Kaufman. JD: Diez
Roux. AV. (2010). Participate air pollution, metabolic syndrome, and heart rate variability: The Multi-
Ethnic Study of Atherosclerosis (MESA). Environ Health Perspect 118: 1406-1411.
http://dx.doi.org/10.1289/ehp.0901778
Parker. RF: Davis. JK: Cassell. GH: White. H: Dziedzic. D: Blalock. DK: Thorp. RB: Simecka. JW.
(1989). Short-term exposure to nitrogen dioxide enhances susceptibility to murine respiratory
mycoplasmosis and decreases intrapulmonary killing of Mycoplasma pulmonis. Am J Respir Crit Care
Med 140: 502-512. http://dx.doi.Org/10.1164/ajrccm/140.2.502
Patel. MM: Chillrud. SN: Correa. JC: Hazi. Y: Feinberg. M: Deepti. KG: Prakash. S: Ross. JM: Lew. D:
Kinnev. PL. (2010). Traffic-related particulate matter and acute respiratory symptoms among New
York City area adolescents. Environ Health Perspect 118: 1338-1343.
http://dx.doi.org/10.1289/ehp.0901499
Patel MM: Chillrud. SN: Deepti. KG: Ross. JM: Kinnev. PL. (2013). Traffic-related air pollutants and
exhaled markers of airway inflammation and oxidative stress in New York City adolescents. Environ
Res 121: 71-78. http://dx.doi.Org/10.1016/j.envres.2012.10.012
Pathmanathan. S: Krishna. MT: Blomberg. A: Helledav. R: Kelly. FJ: Sandstrom. T: Holgate. ST: Wilson.
SJ: Frew. AJ. (2003). Repeated daily exposure to 2 ppm nitrogen dioxide upregulates the expression of
IL-5, IL-10, IL-13, and ICAM-1 in the bronchial epithelium of healthy human airways. Occup
Environ Med 60: 892-896. http://dx.doi.org/10.1136/oem.60.ll.892
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/thx.2010.155358
Peacock. JL: Symonds. P: Jackson. P: Bremner. SA: Scarlett. JF: Strachan. DP: Anderson. HR. (2003).
Acute effects of winter air pollution on respiratory function in schoolchildren in southern England.
Occup Environ Med 60: 82-89. http://dx.doi.Org/10.1136/oem.60.2.82
Peel JL: Klein. M: Flanders. WD: Mulholland. JA: Freed. G: Tolbert PE. (2011). Ambient air pollution
and apnea and bradycardia in high-risk infants on home monitors. Environ Health Perspect 119: 1321-
1327. http://dx.doi.org/10.1289/ehp.1002739
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/aje/kwk051
Peel JL: Tolbert. PE: Klein. M: Metzger. KB: Flanders. WD: Todd. K: Mulholland. JA: Ryan. PB:
Frumkin. H. (2005). Ambient air pollution and respiratory emergency department visits. Epidemiology
16: 164-174. http://dx.doi.org/10.1097/01.ede.0000152905.42113.db
Pilotto. LS: Nitschke. M: Smith. BJ: Pisaniello. D: Ruffin. RE: McElroy. HJ: Martin. J: Hiller. JE. (2004).
Randomized controlled trial of unflued gas heater replacement on respiratory health of asthmatic
schoolchildren. Int J Epidemiol 33: 208-214. http://dx.doi.org/10.1093/ije/dyh018
Pinkston. P: Smeglin. A: Roberts. NJ. Jr: Gibb. FR: Morrow. PE: Utell. MJ. (1988). Effects of in vitro
exposure to nitrogen dioxide on human alveolar macrophage release of neutrophil chemotactic factor
and interleukin-1. Environ Res 47: 48-58. http://dx.doi.org/10.1016/S0013-9351(88)80020-3
Pitard. A: Zeghnoun. A: Courseaux. A: Lambertv. J: Delmas. V: Fossard. JL: Villet H. (2004). Short-term
associations between air pollution and respiratory drug sales. Environ Res 95: 43-52.
http://dx.doi.0rg/10.1016/j.envres.2003.08.006
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.
Ponka. A: Virtanen. M. (1996). Low-level air pollution and hospital admissions for cardiac and
cerebrovascular diseases in Helsinki. Am J Public Health 86: 1273-1280.
http://dx.doi.Org/10.2105/AJPH.86.9.1273
January 2015 5-383 DRAFT: Do Not Cite or Quote
-------
Posin. C: Clark. K: Jones. MP: Patterson. JV: Buckley. RD: Hackney. JD. (1978). Nitrogen dioxide
inhalation and human blood biochemistry. Arch Environ Occup Health 33: 318-324.
Postma. PS: Boezen. HM. (2004). Rationale for the Dutch hypothesis. Allergy and airway
hyperresponsiveness as genetic factors and their interaction with environment in the development of
asthma and COPD [Review]. Chest 126: 96S-104S; discussion 159S-161S.
http://dx.doi.0rg/10.1378/chest.126.2 suppl 1.96S
Povnter. ME: Persinger. RL: Irvin. CG: Butnor. KJ: van Hirtum. H: Slav. W: Heintz. NH: Robbins. J:
Hemenwav. D: Taatjes. DJ: Janssen-Heininger. Y. (2006). Nitrogen dioxide enhances allergic airway
inflammation and hyperresponsiveness in the mouse. Am J Physiol Lung Cell Mol Physiol 290: L144-
L152. http://dx.doi.org/10.1152/aiplung.00131.2005
Proust B: Lacroix. G: Robidel. F: Marliere. M: Lecomte. A: Vargaftig. BB. (2002). Interference of a
short-term exposure to nitrogen dioxide with allergic airways responses to allergenic challenges in
BALB/c mice. Mediators Inflamm 11: 251-260. http://dx.doi.org/10.1080/096293502900000113
Purvis. MR: Ehrlich. R. (1963). Effect of atmospheric pollutants on susceptibility to respiratory infection:
II Effect of nitrogen dioxide. J Infect Dis 113: 72-76.
Qian. Z: Lin. HM: Chinchilli. VM: Lehman. EB: Duan. Y: Craig. TJ: Wilson. WE: Liao. D: Lazarus. SC:
Bascom. R. (2009a). 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
Qian. Z: Lin. HM: Chinchilli. VM: Lehman. EB: Stewart. WF: Shah. N: Duan. Y: Craig. TJ: Wilson. WE:
Liao. D: Lazarus. SC: Bascom. R. (2009b). Associations between air pollution and peak expiratory
flow among patients with persistent asthma. J Toxicol Environ Health A 72: 39-46.
http://dx.doi.org/10.1080/15287390802445517
Qiu. H: Yu. I: Wang. X: Tian. L: Tse. L: Wong. T. (2013a). Season and humidity dependence of the effects
of air pollution on COPD hospitalizations in Hong Kong. Atmos Environ 76: 74-80.
http://dx.doi.0rg/10.1016/j.atmosenv.2012.07.026
Qiu. H: Yu. IT: Wang. X: Tian. L: Tse. LA: Wong. TW. (2013b). Cool and dry weather enhances the
effects of air pollution on emergency IHD hospital admissions. Int J Cardiol 168: 500-505.
http://dx.doi.0rg/10.1016/i.iicard.2012.09.199
Ramos-Bonilla. JP: Brevsse. PN: Dominici F: Geyh. A: Tankerslev. CG. (2010). Ambient air pollution
alters heart rate regulation in aged mice. Inhal Toxicol 22: 330-339.
http://dx.doi.org/10.3109/08958370903349365
Ranzi. A: Gambini M: Spattini. A: Galassi C: Sesti. D: Bedeschi. M: Messori. A: Baroni. A: Cavagni. G:
Lauriola. P. (2004). Air pollution and respiratory status in asthmatic children: Hints for a locally based
preventive strategy AIRE study. Eur J Epidemiol 19: 567-576.
Rasmussen. TR: Kjaergaard. SK: Tarp. U: Pedersen. OF. (1992). Delayed effects of NO2 exposure on
alveolar permeability and glutathione peroxidase in healthy humans. Am Rev Respir Dis 146: 654-
659. http://dx.doi.0rg/10.1164/ajrccm/146.3.654
Reddel. HK: Taylor. PR: Bateman. ED: Boulet. LP: Boushev. HA: Busse. WW: Casale. TB: Chanez. P:
Enright. PL: Gibson. PG: de Jongste. JC: Kerstjens. HA: Lazarus. SC: Lew. ML: O'Byrne. PM:
Partridge. MR: Pavord. ID: Sears. MR: Sterk. PJ: Stoloff. SW: Sullivan. SD: Szefler. SJ: Thomas.
MD: Wenzel. SE. (2009). An official American Thoracic Society/European Respiratory Society
statement: Asthma control and exacerbations: Standardizing endpoints for clinical asthma trials and
clinical practice. Am J Respir Crit Care Med 180: 59-99. http://dx.doi.org/10.1164/rccm.200801-
060ST
Ren. C: Fang. S: Wright. RO: Suh. H: Schwartz. J. (2011). Urinary 8-hydroxy-2'-deoxyguanosine as a
biomarker of oxidative DNA damage induced by ambient pollution in the Normative Aging Study.
Occup Environ Med 68: 562-569. http://dx.doi.org/10.1136/oem.2010.056358
January 2015 5-384 DRAFT: Do Not Cite or Quote
-------
Rice. MB: Ljungman. PL: Wilker. EH: Gold. PR: Schwartz. JD: Koutrakis. P: Washko. GR: O'Connor.
GT: Mittleman. MA. (2013). Short-term exposure to air pollution and lung function in the
Framingham heart study. Am J Respir Crit Care Med 188: 1351-1357.
http://dx.doi.org/10.1164/rccm.201308-1414OC
Rich. DO: Kipen. HM: Huang. W: Wang. G: Wang. Y: Zhu. P: Ohman-Strickland. P: Hu. M: Philipp. C:
Diehl SR: Lu. SE: Tong. J: Gong. J: Thomas. D: Zhu. T: Zhang. JJ. (2012). Association between
changes in air pollution levels during the Beijing Olympics and biomarkers of inflammation and
thrombosis in healthy young adults. JAMA 307: 2068-2078. http://dx.doi.org/10.1001/jama.2012.3488
Rich. DO: 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
Riedl. MA: Diaz-Sanchez. D: Linn. WS: Gong. H. Jn Clark. KW: Effros. RM: Miller. JW: Cocker. PR:
Berhane. KT. (2012). Allergic inflammation in the human lower respiratory tract affected by exposure
to diesel exhaust [HEI] (pp. 5-43; discussion 45-64). (ISSN 1041-5505 Research Report 165). Boston,
MA: Health Effects Institute, http://pubs.healtheffects.org/view.php?id=373
Robison. TW: Duncan. DP: Forman. HJ. (1990). Chemoattractant and leukotriene B4 production from rat
alveolar macrophages exposed to nitrogen dioxide. Am J Respir Cell Mol Biol 3: 21-26.
http://dx.doi.0rg/10.1165/aircmb/3.l.21
Robison. TW: Forman. HJ. (1993). Dual effect of nitrogen dioxide on rat alveolar macrophage
arachidonate metabolism. Exp Lung Res 19: 21-36.
Robison. TW: Murphy. JK: Beyer. LL: Richters. A: Forman. HJ. (1993). Depression of stimulated
arachidonate metabolism and superoxide production in rat alveolar macrophages following in vivo
exposure to 0.5 ppm NO2. J Toxicol Environ Health 38: 273-292.
http://dx.doi.org/10.1080/15287399309531718
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
Roemer. W: Hoek. G: Brunekreef. B: Haluszka. J: Kalandidi A: Pekkanen. J. (1998). Daily variations in
air pollution and respiratory health in a multicentre study: The PEACE project. Eur Respir J 12: 1354-
1361.
Roger. LJ: Horstman. DH: McDonnell. W: Kehrl. H: Ives. PJ: Seal E: Chapman. R: Massaro. E. (1990).
Pulmonary function, airway responsiveness, and respiratory symptoms in asthmatics following
exercise in NO2. Toxicol Ind Health 6: 155-171.
Rombout. PJ: Dormans. JA: Marra. M: Van Esch. GJ. (1986). Influence of exposure regimen on nitrogen
dioxide-induced morphological changes in the rat lung. Environ Res 41: 466-480.
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.0rg/10.1016/i.iaci.2007.12.004
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
Roonev. AA: Boyles. AL: Wolfe. MS: Bucher. JR: Thaver. KA. (2014). Systematic review and evidence
integration for literature-based environmental health science assessments. Environ Health Perspect
122: 711-718. http://dx.doi.org/10.1289/ehp.1307972
January 2015 5-385 DRAFT: Do Not Cite or Quote
-------
Rose. RM: Fuglestad. JM: Skornik. WA: Hammer. SM: Wolfthal SF: Beck. BD: Brain. JD. (1988). The
pathophysiology of enhanced susceptibility to murine cytomegalovirus respiratory infection during
short-term exposure to 5 ppm nitrogen dioxide. Am Rev Respir Dis 137: 912-917.
http://dx.doi.0rg/10.1164/airccm/137.4.912
Rose. RM: Pinkston. P: Skornik. WA. (1989). Altered susceptibility to viral respiratory infection during
short-term exposure to nitrogen dioxide. (HEI Research Report no. 24). Cambridge, MA: Health
Effects Institute.
Rubinstein. I: Bigbv. BG: Reiss. TF: Boushev. HA. Jr. (1990). Short-term exposure to 0.3 ppm nitrogen
dioxide does not potentiate airway responsiveness to sulfur dioxide in asthmatic subjects. Am Rev
Respir Dis 141: 381-385. http://dx.doi.Org/10.1164/airccm/141.2.381
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
Sacks. JD: Ito. K: Wilson. WE: Neas. LM. (2012). Impact of covariate models on the assessment of the
air pollution-mortality association in a single- and multipollutant context. Am J Epidemiol 176: 622-
634. http://dx.doi.org/10.1093/aie/kwsl35
Salam. MT: Byun. HM: Lurmann. F: Breton. CV: Wang. X: Eckel SP: Gilliland. FD. (2012). Genetic and
epigenetic variations in inducible nitric oxide synthase promoter, particulate pollution, and exhaled
nitric oxide levels in children. J Allergy Clin Immunol 129: 232-239.e237.
http://dx.doi.0rg/10.1016/i.iaci.2011.09.037
Salonen. RO: Pennanen. AS: Vahteristo. M: Korkeila. P: Aim. S: Randell JT. (2008). Health risk
assessment of indoor air pollution in Finnish ice arenas. Environ Int 34: 51-57.
http://dx.doi.0rg/10.1016/j.envint.2007.06.012
Samoli. E: Aga. E: Touloumi. G: Nisiotis. K: Forsberg. B: Lefranc. A: Pekkanen. J: Wojtyniak. B:
Schindler. C: Niciu. E: Brunstein. R: Dodic Fikfak. M: Schwartz. J: Katsouvanni. K. (2006). Short-
term effects of nitrogen dioxide on mortality: An analysis within the APHEA project. Eur Respir J 27:
1129-1137. http://dx.doi.org/10.1183/09031936.06.00143905
Samoli. E: Nastos. PT: Paliatsos. AG: Katsouvanni. K: Priftis. KN. (2011). Acute effects of air pollution
on pediatric asthma exacerbation: Evidence of association and effect modification. Environ Res 111:
418-424. http://dx.doi.0rg/10.1016/i.envres.2011.01.014
Sarnat JA: Sarnat SE: Flanders. WD: Chang. HH: Mulholland. J: Baxter. L: Isakov. V: Ozkaynak. H.
(2013a). Spatiotemporally resolved air exchange rate as a modifier of acute air pollution-related
morbidity in Atlanta. J Expo Sci Environ Epidemiol 23: 606-615.
http://dx.doi.org/10.1038/ies.2013.32
Sarnat. SE: Ravsoni. AU: Li. WW: Holguin. F: Johnson. BA: Flores Luevano. S: Garcia. JH: Sarnat. JA.
(2012). Air pollution and acute respiratory response in a panel of asthmatic children along the U.S.-
Mexico border. Environ Health Perspect 120: 437444. http://dx.doi.org/10.1289/ehp.1003169
Sarnat. SE: Sarnat. JA: Mulholland. J: Isakov. V: Ozkaynak. H: Chang. HH: Klein. M: Tolbert. PE.
(2013b). Application of alternative spatiotemporal metrics of ambient air pollution exposure in a time-
series epidemiological study in Atlanta. J Expo Sci Environ Epidemiol 23: 593-605.
http://dx.doi.org/10.1038/ies.2013.41
Scaife. A: Barclay. J: Hillis. GS: Srinivasan. J: Macdonald. DW: Ross. JA: Ayres. JG. (2012). Lack of
effect of nitrogen dioxide exposure on heart rate variability in patients with stable coronary heart
disease and impaired left ventricular systolic function. Occup Environ Med 69: 587-591.
http://dx.doi.org/10.1136/oemed-2011-100126
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.
January 2015 5-386 DRAFT: Do Not Cite or Quote
-------
Schildcrout JS: Sheppard. L: Lumlev. T: Slaughter. JC: Koenig. JQ: 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/aje/kwj225
Schindler. C: Kiinzlj N: Bongard. JP: Leuenberger. P: Karrer. W: Rapp. R: Monn. C: Ackermann-
Liebrich. U. (2001). Short-term variation in air pollution and in average lung function among never-
smokers. The Swiss Study on Air Pollution and Lung Diseases in Adults (SAPALDIA). Am J Respir
Crit Care Med 163: 356-361. http://dx.doi.Org/10.1164/ajrccm.163.2.9911116
Schlesinger. RB. (1987a). Effects of intermittent inhalation exposures to mixed atmospheres of NO2 and
H2SO4 on rabbit alveolar macrophages. J Toxicol Environ Health 22: 301-312.
http://dx.doi.org/10.1080/15287398709531073
Schlesinger. RB. (1987b). Intermittent inhalation of nitrogen dioxide: Effects on rabbit alveolar
macrophages. J Toxicol Environ Health 21: 127-139. http://dx.doi.org/10.1080/15287398709531007
Schlesinger. RB: Driscoll. KE: Gunnison. AF: Zelikoff. JT. (1990). Pulmonary arachidonic acid
metabolism following acute exposures to ozone and nitrogen dioxide. J Toxicol Environ Health 31:
275-290. http://dx.doi.org/10.1080/15287399009531456
Schwartz. J: Coull. BA. (2003). Control for confounding in the presence of measurement error in
hierarchical models. Biostatistics 4: 539-553.
Schwartz. J: Dockery. DW: Neas. LM: Wvpij. D: Ware. JH: Spengler. JD: Koutrakis. P: Speizer. FE:
Ferris. BG. Jr. (1994). Acute effects of summer air pollution on respiratory symptom reporting in
children. Am J Respir Crit Care Med 150: 1234-1242. http://dx.doi.Org/10.1164/ajrccm.150.5.7952546
Segala. C: Fauroux. B: Just. J: Pascual. L: Grimfeld. A: Neukirch. F. (1998). Short-term effect of winter
air pollution on respiratory health of asthmatic children in Paris. Eur Respir J 11: 677-685.
Segala. C: Poizeau. D: Mesbah. M: Willems. S: Maidenberg. M. (2008). Winter air pollution and infant
bronchiolitis in Paris. Environ Res 106: 96-100. http://dx.doi.Org/10.1016/j.envres.2007.05.003
Selgrade. MK: Mole. ML: Miller. FJ: Hatch. GE: Gardner. DE: Hu. PC. (1981). Effect of NO2 inhalation
and vitamin C deficiency on protein and lipid accumulation in the lung. Environ Res 26: 422-437.
Sevanian. A: Hacker. AD: Elsaved. N. (1982). Influence of vitamin E and nitrogen dioxide on lipid
peroxidation in rat lung and liver microsomes. Lipids 17: 269-277.
Sherwin. RP: Carlson. DA. (1973). Protein content of lung lavage fluid of guinea pigs exposed to 0.4 ppm
nitrogen dioxide. Arch Environ Health 27: 90-93.
Sherwin. RP: Dibble. J: Weiner. J. (1972). Alveolar wall cells of the guinea pig: Increase in response to 2
ppm nitrogen dioxide. Arch Environ Health 24: 43-47.
Sherwood. RL: Lippert WE: Goldstein. E: Tarkington. B. (1981). Effect of ferrous sulfate aerosols and
nitrogen dioxide on murine pulmonary defense. Arch Environ Occup Health 36: 130-135.
Shields. KN: Cavallari. JM: Hunt. MJ: Lazo. M: Molina. M: Molina. L: Holguin. F. (2013). Traffic-
related air pollution exposures and changes in heart rate variability in Mexico City: A panel study.
Environ Health 12: 7. http://dx.doi.org/10.1186/1476-069X-12-7
Shin. HH: Stieb. D: Burnett. R: Takahara. G: Jessiman. B. (2012). Tracking national and regional spatial-
temporal mortality risk associated with NO(2) concentrations in Canada: A Bayesian hierarchical two-
Level model. Risk Anal 32: 513-530. http://dx.doi.Org/10.llll/i.1539-6924.2011.01684.x
Silbaugh. SA: Mauderly. JL: Macken. CA. (1981). Effects of sulfuric acid and nitrogen dioxide on airway
responsiveness of the guinea pig. J Toxicol Environ Health 8: 31-45.
http://dx.doi.org/10.1080/15287398109530048
Silkoff. PE: Zhang. L: Dutton. S: Langmack. EL: Vedal S: Murphy. J: Make. B. (2005). Winter air
pollution and disease parameters in advanced chronic obstructive pulmonary disease panels residing in
Denver, Colorado. J Allergy Clinlmmunol 115: 337-344. http://dx.doi.Org/10.1016/j.jaci.2004.ll.035
January 2015 5-387 DRAFT: Do Not Cite or Quote
-------
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/aje/kwq217
Simpson. R: Williams. G: Petroeschevsky. A: Best T: Morgan. G: Denison. L: Hinwood. A: Neville. G.
(2005a). The short-term effects of air pollution on hospital admissions in four Australian cities. Aust N
Z J Public Health 29: 213-221.
Simpson. R: Williams. G: Petroeschevsky. A: Best T: Morgan. G: Denison. L: Hinwood. A: Neville. G:
Neller. A. (2005b). The short-term effects of air pollution on daily mortality in four Australian cities.
Aust N Z J Public Health 29: 205-212. http://dx.doi.Org/10.llll/i.1467-842X.2005.tb00758.x
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.
http://dx.doi.org/10.1080/10473289.2004.10470979
Smargiassi. A: Goldberg. MS: Wheeler. AJ: Plante. C: Valois. MF: Mallach. G: Kauri. LM: Shutt R:
Harriett S: Raphoz. M: Liu. L. (2014). Associations between personal exposure to air pollutants and
lung function tests and cardiovascular indices among children with asthma living near an industrial
complex and petroleum refineries. Environ Res 132: 38-45.
http://dx.doi.0rg/10.1016/j.envres.2014.03.030
Solomon. C: Christian. PL: Chen. LL: Welch. BS: Kleinman. MT: Dunham. E: Erie. DJ: Balmes. JR.
(2000). Effect of serial-day exposure to nitrogen dioxide on airway and blood leukocytes and
lymphocyte subsets. Eur Respir J 15: 922-928.
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/j.envres.2010.08.003
Son. JY: Lee. JT: Park. YH: Bell ML. (2013). Short-term effects of air pollution on hospital admissions
in Korea. Epidemiology 24: 545-554. http://dx.doi.org/10.1097/EDE.Ob013e3182953244
Soto-Ramos. M: Castro-Rodriguez. JA: Hinojos-Gallardo. LC: Hernandez-Saldana. R: Cisneros-Castolo.
M: Carrillo-Rodriguez. V. (2013). Fractional exhaled nitric oxide has a good correlation with asthma
control and lung function in Latino children with asthma. J Asthma 50: 590-594.
http://dx.doi.org/10.3109/02770903.2013.792349
Sousa. R: Duque. L: Duarte. AJ: Gomes. CR: Ribeiro. H: Cruz. A: Esteves da Silva. JC: Abreu. I. (2012).
In vitro exposure of Acer negundo pollen to atmospheric levels of SO and NO: effects on allergenicity
and germination. Environ Sci Technol 46: 2406-2412. http://dx.doi.org/10.1021/es2034685
Spiezia. L: Campello. E: Bon. M: Maggiolo. S: Pelizzaro. E: Simioni. P. (2014). Short-term exposure to
high levels of air pollution as a risk factor for acute isolated pulmonary embolism. Thromb Res Suppl
134: 259-263. http://dx.doi.Org/10.1016/i.thromres.2014.05.011
Spira-Cohen. A. (2013). Email from Dr. Spira-Cohen to Dr. Patel; Response to data request. Available
online
Spira-Cohen. A: Chen. LC: Kendall M: Lall R: Thurston. GD. (2011). Personal exposures to traffic-
related air pollution and acute respiratory health among Bronx schoolchildren with asthma. Environ
Health Perspect 119: 559-565. http://dx.doi.org/10.1289/ehp. 1002653
Steenhof. M: Janssen. NA: Strak. M: Hoek. G: Gosens. I: Mudwav. IS: Kelly. FJ: Harrison. RM: Pieters.
RH: Cassee. FR: Brunekreef. B. (2014). Air pollution exposure affects circulating white blood cell
counts in healthy subjects: the role of particle composition, oxidative potential and gaseous pollutants
- the RAPTES project. Inhal Toxicol 26: 141-165. http://dx.doi.org/10.3109/08958378.2013.861884
January 2015 5-388 DRAFT: Do Not Cite or Quote
-------
Steenhof. M: Mudwav. IS: Gosens. I: Hoek. G: Godri. KJ: Kelly. FJ: Harrison. RM: Pieters. RH: Cassee.
FR: Lebret. E: Brunekreef. BA: Strak. M: Janssen. NA. (2013). Acute nasal pro-inflammatory
response to air pollution depends on characteristics other than particle mass concentration or oxidative
potential: the RAPTES project. Occup Environ Med 70: 341-348. http://dx.doi.org/10.1136/oemed-
2012-100993
Steerenberg. PA: Bischoff. EWM. A: de Klerk. A: Verlaan. APJ: Jongbloets. LMN: Van Loveren. H:
Opperhuizen. A: Zomer. G: Heisterkamp. SH: Hadv. M: Spieksma. FTM: Fischer. PH: Dormans.
JAM. A: van Amsterdam. JGC. (2003). Acute effect of air pollution on respiratory complaints, exhaled
NO and biomarkers in nasal lavages of allergic children during the pollen season. Int Arch Allergy
Immunol 131: 127-137. http://dx.doi.org/10.1159/000070928
Steerenberg. PA: Nierkens. S: Fischer. PH: Van Loveren. H: Opperhuizen. A: Vos. JG: van Amsterdam.
JG. (2001). Traffic-related air pollution affects peak expiratory flow, exhaled nitric oxide, and
inflammatory nasal markers. Arch Environ Health 56: 167-174.
http://dx.doi.org/10.1080/00039890109604069
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/j.envres.2007.08.006
Stern. G: Latzin. P: Roosli. M: Fuchs. O: Proietti. E: Kuehni. C: Frev. U. (2013). A prospective study of
the impact of air pollution on respiratory symptoms and infections in infants. Am J Respir Crit Care
Med 187: 1341-1348. http://dx.doi.org/10.1164/rccm.201211-2008OC
Stieb. DM: Burnett. RT: Smith-Doiron. M: Brion. O: Shin. HH: Economou. V. (2008). Anew
multipollutant, no-threshold air quality health index based on short-term associations observed in daily
time-series analyses. J Air Waste Manag Assoc 58: 435-450. http://dx.doi.org/10.3155/1047-
3289.58.3.435
Stieb. DM: Judek. S: Burnett. RT. (2002). Meta-analysis of time-series studies of air pollution and
mortality: Effects of gases and particles and the influence of cause of death, age, and season. J Air
Waste Manag Assoc 52: 470-484. http://dx.doi.org/10.1080/10473289.2002.10470794
Stieb. DM: Judek. S: Burnett. RT. (2003). Meta-analysis of time-series studies of air pollution and
mortality: Update in relation to the use of generalized additive models. J Air Waste Manag Assoc 53:
258-261.
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 8: 25.
http://dx.doi.org/10.1186/1476-069X-8-25
Strak. M. (2013). [Email from Dr. Strak to Dr. Patel; Response to data request]. Available online
Strak. M: Hoek. G: Steenhof. M: Kilinc. E: Godri. KJ: Gosens. I: Mudwav. IS: van Oerle. R: Spronk.
HM: Cassee. FR: Kelly. FJ: Harrison. RM: Brunekreef. B: Lebret E: Janssen. NA. (2013).
Components of ambient air pollution affect thrombin generation in healthy humans: the RAPTES
project. Occup Environ Med 70: 332-340. http://dx.doi.org/10.1136/oemed-2012-100992
Strak. M: Janssen. NA: Godri. KJ: Gosens. I: Mudwav. IS: Cassee. FR: Lebret E: Kelly. FJ: Harrison.
RM: Brunekreef. B: Steenhof. M: Hoek. G. (2012). Respiratory health effects of airborne particulate
matter: The role of particle size, composition and oxidative potential - The RAPTES project. Environ
Health Perspect 120: 1183-1189. http://dx.doi.org/10.1289/ehp.1104389
Strand. V: Rak. S: Svartengren. M: Bylin. G. (1997). Nitrogen dioxide exposure enhances asthmatic
reaction to inhaled allergen in subjects with asthma. Am J Respir Crit Care Med 155: 881-887.
http://dx.doi.0rg/10.1164/airccm.155.3.9117021
January 2015 5-389 DRAFT: Do Not Cite or Quote
-------
Strand. V: Salomonsson. P: Lundahl J: Bylin. G. (1996). Immediate and delayed effects of nitrogen
dioxide exposure at an ambient level on bronchial responsiveness to histamine in subjects with
asthma. Eur Respir J 9: 733-740. http://dx.doi.org/10.1183/09031936.96.09040733
Strand. V: Svartengren. M: Rak. S: Barck. C: Bylin. G. (1998). Repeated exposure to an ambient level of
NO2 enhances asthmatic response to a nonsymptomatic allergen dose. Eur Respir J 12: 6-12.
http://dx.doi.org/10.1183/09031936.98.12010006
Stranev. L: Finn. J: Dennekamp. M: Bremner. A: Tonkin. A: Jacobs. I. (2014). Evaluating the impact of
air pollution on the incidence of out-of-hospital cardiac arrest in the Perth Metropolitan Region: 2000-
2010. J Epidemiol Community Health 68: 6-12. http://dx.doi.org/10.1136/jech-2013-202955
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 10: 36.
http://dx.doi.org/10.1186/1476-069X-10-36
Suh. HH: Zanobetti A. (2010a). Exposure error masks the relationship between traffic-related air
pollution and heart rate variability. J Occup Environ Med 52: 685-692.
http://dx.doi.org/10.1097/JOM.Ob013e3181e8071f
Suh. HH: Zanobetti. A. (201 Ob). 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
Suzuki. T: Ikeda. S: Kanoh. T: Mizoguchi. I. (1986). Decreased phagocytosis and superoxide anion
production in alveolar macrophages of rats exposed to nitrogen dioxide. Arch Environ Contam Toxicol
15: 733-739. http://dx.doi.org/10.1007/BF01054920
Szyszkowicz. M. (2007). Air pollution and emergency department visits for ischemic heart disease in
Montreal, Canada. Int J Occup Med Environ Health 20: 167-173. http://dx.doi.org/10.2478/vl0001-
007-0015-3
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
Szyszkowicz. M. (2009). Air pollution and ED visits for chest pain. Am J Emerg Med 27: 165-168.
http://dx.doi.0rg/10.1016/j.ajem.2008.01.010
Szyszkowicz. M: Rowe. BH: Brook. RD. (2012). Even low levels of ambient air pollutants are associated
with increased emergency department visits for hypertension. Can J Cardiol 28: 360-366.
http://dx.doi.0rg/10.1016/i.cica.2011.06.011
Tashkin. DP: Simmons. MS: Chang. P: Liu. H: Coulson. AH. (1993). Effects of smoked substance abuse
on nonspecific airway hyperresponsiveness. Am Rev Respir Dis 147: 97-103.
http://dx.doi.0rg/10.1164/ajrccm/147.l.97
Task Force of the European Society of Cardiology and the North American Society of Pacing and
Electrophy siolo gv. (1996). Heart rate variability: Standards of measurement, physiological
interpretation and clinical use. Circulation 93: 1043-1065. http://dx.doi.Org/10.1161/01.CIR.93.5.1043
Thach. TO: Wong. CM: Chan. KP: Chau. YK: Thomas. GN: Ou. CO: Yang. L: Peiris. JSM: Lam. TH:
Hedlev. AJ. (2010). Air pollutants and health outcomes: Assessment of confounding by influenza.
Atmos Environ 44: 1437-1442. http://dx.doi.Org/10.1016/i.atmosenv.2010.01.036
January 2015 5-390 DRAFT: Do Not Cite or Quote
-------
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
Timonen. KL: Pekkanen. J. (1997). Air pollution and respiratory health among children with asthmatic or
cough symptoms. Am J Respir Crit Care Med 156: 546-552.
http://dx.doi.0rg/10.1164/ajrccm.156.2.9608044
Timonen. KL: Vanninen. E: De Hartog. J: Ibald-Mulli. A: Brunekreef. B: Gold. PR: Henrich. J: Hoek. G:
Lankj T: Peters. A: Tarkiainen. T: Tiittanen. P: Kreyling. W: Pekkanen. J. (2006). Effects of ultrafine
and fine particulate and gaseous air pollution on cardiac autonomic control in subjects with coronary
artery disease: The ULTRA study. J Expo Sci Environ Epidemiol 16: 332-341.
http://dx.doi.org/10.1038/sj.jea.7500460
Tolbert. P. (2009). Personal communication from Dr. Paige Tolbert, Emory University to Dr. Ellen
Kirrane, NCEA with attachment. Available online
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.7500625
Tolbert. PE: Mulholland. JA: Macintosh. PL: Xu. F: Daniels. D: Devine. OJ: Carlin. BP: Klein. M:
Dorlev. J: Butler. AJ: Nordenberg. DF: Frumkin. H: Ryan. PB: White. MC. (2000). Air quality and
pediatric emergency room visits for asthma in Atlanta, Georgia. Am J Epidemiol 151: 798-810.
Toren. K: Brisman. J: Jarvholm. B. (1993). Asthma and asthma-like symptoms in adults assessed by
questionnaires: a literature review [Review]. Chest 104: 600-608.
Tsai DH: Wang. JL: Chuang. KJ: Chan. CC. (2010). Traffic-related air pollution and cardiovascular
mortality in central Taiwan. Sci Total Environ 408: 1818-1823.
http://dx.doi.0rg/10.1016/j.scitotenv.2010.01.044
Tsai SS: Chen. PS: Yang. YH: Liou. SH: Wu. TN: Sung. FC: Yang. CY. (2012). Air pollution and hospital
admissions for myocardial infarction: are there potentially sensitive groups? J Toxicol Environ Health
A 75: 242-251. http://dx.doi.org/doi: 10.1080/15287394.2012.641202
Tsai SS: Chiu. HF: Wu. TN: Yang. CY. (2009). Air pollution and emergency room visits for cardiac
arrhythmia in a subtropical city: Taipei, Taiwan. Inhal Toxicol 21: 1113-1118.
http://dx.doi.org/10.3109/08958370902758939
Tsai SS: Goggins. WB: Chiu. HF: Yang. CY. (2003). Evidence for an association between air pollution
and daily stroke admissions in Kaohsiung, Taiwan. Stroke 34: 2612-2616.
http://dx.doi.org/10.1161/01.STR.0000095564.33543.64
Tsuji. H: Larson. MG: Venditti. FJ. Jr: Manders. ES: Evans. JC: Feldman. CL: Lew. D. (1996). Impact of
reduced heart rate variability on risk for cardiac events The Framingham Heart Study. Circulation 94:
2850-2855.
Tsuji H: Venditti FJ: Manders. ES: Evans. JC: Larson. MG: Feldman. CL: Lew. D. (1994). Reduced
heart rate variability and mortality risk in an elderly cohort: The Framingham Heart Study. Circulation
90: 878-883.
Tunnicliffe. WS: Surge. PS: Ayres. JG. (1994). Effect of domestic concentrations of nitrogen dioxide on
airway responses to inhaled allergen in asthmatic patients. Lancet 344: 1733-1736.
http://dx.doi.org/10.1016/s0140-6736(94)92886-x
January 2015 5-391 DRAFT: Do Not Cite or Quote
-------
Turin. TC: Kita. Y: Rumana. N: Nakamura. Y: Ueda. K: Takashima. N: Sugihara. H: Morita. Y: Ichikawa.
M: Hirose. K: Nitta. H: Okavama. A: Miura. K: Ueshima. H. (2012). Short-term exposure to air
pollution and incidence of stroke and acute myocardial infarction in a Japanese population.
Neuroepidemiology 38: 84-92. http://dx.doi.org/10.1159/000335654
U.S. EPA (U.S. Environmental Protection Agency). (1991). Guidelines for developmental toxicity risk
assessment. (EPA/600/FR-91/001). Washington, DC: U.S. Environmental Protection Agency, Risk
Assessment Forum, http://www.epa.gov/raf/publications/guidelines-dev-toxicitv-risk-assessment.htm
U.S. EPA (U.S. Environmental Protection Agency). (1993). Air quality criteria for oxides of nitrogen, vol.
1-3 [EPAReport]. (EPA/600/8-91/049aF-cF). Research Triangle Park, NC: U.S. Environmental
Protection Agency, Environmental Criteria and Assessment Office.
http://cfpub.epa. gov/ncea/cfm/recordisplav.cfm?deid=40179
U.S. EPA (U.S. Environmental Protection Agency). (1996). Guidelines for reproductive toxicity risk
assessment [EPA Report]. (EPA/630/R-96/009). Washington, DC: U.S. Environmental Protection
Agency, Risk Assessment Forum, http://www.epa.gov/raf/publications/pdfs/REPRO51.PDF
U.S. EPA (U.S. Environmental Protection Agency). (1998). Guidelines for neurotoxicity risk assessment.
(EPA/630/R-95/001F). Washington, DC: U.S. Environmental Protection Agency, Risk Assessment
Forum. http://www.epa.gov/raf/publications/pdfs/NEUROTOX.PDF
U.S. EPA (U.S. Environmental Protection Agency). (2005). Guidelines for carcinogen risk assessment.
(EPA/63O/P-03/00IF). Washington, DC: U.S. Environmental Protection Agency, Risk Assessment
Forum, http://www.epa. gov/cancerguidelines/
U.S. EPA (U.S. Environmental Protection Agency). (2006). Air quality criteria for ozone and related
photochemical oxidants [EPA Report]. (EPA/600/R-05/004AF). Research Triangle Park, NC: U.S.
Environmental Protection Agency, National Center for Environmental Assessment.
http://cfpub.epa.gov/ncea/cfm/recordisplav.cfm7de id=149923
U.S. EPA (U.S. Environmental Protection Agency). (2008a). Integrated science assessment for oxides of
nitrogen health criteria (final report, 2008) [EPAReport]. (EPA/600/R-08/071). Research Triangle
Park, NC: U.S. Environmental Protection Agency, National Center for Environmental Assessment.
http://cfpub.epa.gov/ncea/cfm/recordisplav.cfm7de id=194645
U.S. EPA (U.S. Environmental Protection Agency). (2008b). Integrated science assessment for sulfur
oxides: Health criteria [EPA Report]. (EPA/600/R-08/047F). Research Triangle Park, NC: U.S.
Environmental Protection Agency, National Center for Environmental Assessment.
http://cfpub.epa.gov/ncea/cfm/recordisplav.cfm7de id=198843
U.S. EPA (U.S. Environmental Protection Agency). (2009). Integrated science assessment for particulate
matter [EPA Report]. (EPA/600/R-08/139F). Research Triangle Park, NC: U.S. Environmental
Protection Agency, National Center for Environmental Assessment.
http://cfpub.epa.gov/ncea/cfm/recordisplav.cfm7de id=216546
U.S. EPA (U.S. Environmental Protection Agency). (2010). Integrated science assessment for carbon
monoxide [EPA Report]. (EPA/600/R-09/019F). Research Triangle Park, NC: U.S. Environmental
Protection Agency, National Center for Environmental Assessment.
http://cfpub.epa.gov/ncea/cfm/recordisplav.cfm7de id=218686
U.S. EPA (U.S. Environmental Protection Agency). (2013a). Integrated science assessment for ozone and
related photochemical oxidants (final report) [EPA Report]. (EPA/600/R-10/076F). Washington, DC.
http://cfpub.epa.gov/ncea/isa/recordisplav.cfm?deid=247492
U.S. EPA (U.S. Environmental Protection Agency). (2013b). Toxicological review of trimethylbenzenes
(CASRN 25551-13-7, 95-63-6, 526-73-8, and 108-67-8) in support of summary information on the
Integrated Risk Information System (IRIS): revised external review draft [EPA Report].
(EPA/635/R13/17la). Washington, D.C.: U.S. Environmental Protection Agency, National Center for
Environmental Assessment.
http://YOsemite.epa.gov/sab/SABPRODUCT.NSF/b5d8alce9b07293485257375007012b7/eele280e77
586de985257b65005d37e7!OpenDocument
January 2015 5-392 DRAFT: Do Not Cite or Quote
-------
U.S. EPA (U.S. Environmental Protection Agency). (2014a). Supplemental figure S5-1 associations of
NO2 with respiratory effects in copollutant models with PM10 PM10_25 SO2 O3.
U.S. EPA (U.S. Environmental Protection Agency). (2014b). Supplemental Figure S5-2. Results of single-
pollutant and copollutants models of short-term exposure to NO2 with and without PM and CVD HA.
U.S. EPA (U.S. Environmental Protection Agency). (2014c). Supplemental Figure S5-3. Results of single-
pollutant and copollutants models of short-term exposure to NO2 with CO and without CO and CVD
HA.
U.S. EPA (U.S. Environmental Protection Agency). (2014d). Supplemental Figure S5-4 Results of single-
pollutant and copollutants models of short-term exposure to NO2 or NOX with and without O3 and
hospital admissions CVD.
U.S. EPA (U.S. Environmental Protection Agency). (2014e). Supplemental Figure S5-5 Results of single-
pollutant and copollutants models of short-term exposure to NO2 or NOX with and without SO2 and
hospital admissions CVD.
U.S. EPA (U.S. Environmental Protection Agency). (2014f). Supplemental Table S5-1. Calculation of
Increments of Oxides of Nitrogen for Standardizing Epidemiologic Effect Estimates - Short-term
Averages.
U.S. EPA (U.S. Environmental Protection Agency). (2014g). Supplemental Table S5-2. Identification of
airway responsiveness data in papers.
U.S. EPA (U.S. Environmental Protection Agency). (2014h). Supplemental Table S5-3. Studies of short-
term exposure to oxides of nitrogen and respiratory hospital admissions and ED visits.
U.S. EPA (U.S. Environmental Protection Agency). (20141). Supplemental Table S5-4. Studies of short-
term exposure to oxides of nitrogen and mortality.
Vagaggini B: Paggiaro. PL: Gianninj D: Franco. AD: Cianchetti S: Carnevali S: Taccola. M: Bacci. E:
Bancalarj L: Dente. FL: Giuntini C. (1996). Effect of short-term NO2 exposure on induced sputum in
normal, asthmatic and COPD subjects. Eur Respir J 9: 1852-1857.
van der Zee. SC: Hoek. G: Boezen. HM: Schouten. JP: van Wijnen. JH: Brunekreef. B. (1999). Acute
effects of urban air pollution on respiratory health of children with and without chronic respiratory
symptoms. Occup Environ Med 56: 802-813. http://dx.doi.org/10.1136/oem.56.12.802
Van Der Zee. SC: Hoek. G: Boezen. MH: Schouten. JP: Van Wijnen. JH: Brunekreef. B. (2000). Acute
effects of air pollution on respiratory health of 50-70 yr old adults. Eur Respir J 15: 700-709.
Van Schavck. CP: Dompeling. E: Van Herwaarden. CL: Wever. AM: Van Weel. C. (1991). Interacting
effects of atopy and bronchial hyperresponsiveness on the annual decline in lung function and the
exacerbation rate in asthma. Am Rev Respir Dis 144: 1297-1301.
http://dx.doi.0rg/10.1164/ajrccm/144.6.1297
Vencloviene. J: Grazuleviciene. R: Babarskiene. R: Dedele. A: Grazulevicius. T. (2011). Short-term
nitrogen dioxide exposure and geomagnetic activity interaction: contribution to emergency
hospitalization for acute coronary syndrome. Int J Environ Health Res 21: 149-160.
http://dx.doi.org/10.1080/09603123.2010.515671
Verbanck. S: Schuermans. D: Noppen. M: Vincken. W: Paiva. M. (2001). Methacholine versus histamine:
paradoxical response of spirometry and ventilation distribution. J Appl Physiol 91: 2587-2594.
Vidale. S: Bonanomi. A: Guidotti. M: Arnaboldi. M: Sterzi. R. (2010). Air pollution positively correlates
with daily stroke admission and in hospital mortality: a study in the urban area of Como, Italy. Neurol
Sci31: 179-182. http://dx.doi.org/10.1007/sl0072-009-0206-8
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 6: 40. http://dx.doi.org/10.1186/1476-069X-6-40
January 2015 5-393 DRAFT: Do Not Cite or Quote
-------
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.Org/10.llll/j.1398-9995.2006.01070.x
Villeneuve. PJ: Johnson. JY: Pasichnyk. D: Lowes. J: Kirkland. S: Rowe. BH. (2012). Short-term effects
of ambient air pollution on stroke: who is most vulnerable? Sci Total Environ 430: 193-201.
http://dx.doi.0rg/10.1016/j.scitotenv.2012.05.002
Vollmuth. TA: Driscoll KE: Schlesinger. RB. (1986). Changes in early alveolar particle clearance due to
single and repeated nitrogen dioxide exposures in the rabbit. J Toxicol Environ Health 19: 255-266.
http://dx.doi.org/10.1080/15287398609530925
von Elm. E: Altman. DG: Egger. M: Pocock. SJ: G0tzsche. PC: Vandenbroucke. JP. (2007). The
Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement:
guidelines for reporting observational studies [Review]. PLoS Med 4: e296.
http://dx.doi.org/10.1371/journal.pmed.0040296
Von Klot S: Peters. A: Aalto. P: Bellander. T: Berglind. N: D'Ippoliti. D: Elosua. R: Hormann. A:
Kulmala. M: Lanki. T: Lowel H: Pekkanen. J: Picciotto. S: Sunyer. J: Forastiere. F. (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
von Klot. S: Wolke. G: Tuch. T: Heinrich. J: Dockery. DW: Schwartz. J: Kreyling. WG: Wichmann. HE:
Peters. A. (2002). Increased asthma medication use in association with ambient fine and ultrafine
particles. Eur Respir J 20: 691-702. http://dx.doi.org/10.1183/09031936.02.01402001
Wang. JH: Devalia. JL: Puddle. JM: Hamilton. SA: Davies. RJ. (1995a). Effect of six-hour exposure to
nitrogen dioxide on early-phase nasal response to allergen challenge in patients with a history of
seasonal allergic rhinitis. J Allergy Clin Immunol 96: 669-676.
Wang. JH: Devalia. JL: Rusznak. C: Bagnall. A: Sapsford. RJ: Davies. RJ. (1999). Effect of fluticasone
propionate aqueous nasal spray on allergen-induced inflammatory changes in the nasal airways of
allergic rhinitics following exposure to nitrogen dioxide. Clin Exp Allergy 29: 234-240.
http://dx.doi.0rg/10.1046/i.1365-2222.1999.00440.x
Wang. JH: Puddle. J: Devalia. JL: Davies. RJ. (1995b). Nitrogen dioxide increases eosinophil activation
in the early-phase response to nasal allergen provocation. Int Arch Allergy Immunol 107: 103-105.
http://dx.doi.org/10.1159/000236944
Wanner. A: Brodnan. JM: Perez. J: Henke. KG: Kim. CS. (1985). Variability of airway responsiveness to
histamine aerosol in normal subjects. Role of deposition. Am Rev Respir Dis 131: 3-7.
http://dx.doi.org/10.1002/ppul.1950010433
Ward. DJ: Miller. MR: Walters. S: Harrison. RM: Ayres. JG. (2000). Impact of correcting peak flow for
nonlinear errors on air pollutant effect estimates from a panel study. Eur Respir J 15: 137-140.
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.Org/10.1136/thorax.57.6.489
Weaklev. J: Webber. MP: Ye. F: Zeig-Owens. R: Cohen. HW: Hall. CB: Kelly. K: Prezant. DJ. (2013).
Agreement between obstructive airways disease diagnoses from self-report questionnaires and
medical records. Prev Med 57: 38-42. http://dx.doi.Org/10.1016/j.vpmed.2013.04.001
Weichenthal. S: Dufresne. A: Infante-Rivard. C: Joseph. L. (2007). Indoor ultrafine particle exposures and
home heating systems: a cross-sectional survey of Canadian homes during the winter months. J Expo
Sci Environ Epidemiol 17: 288-297. http://dx.doi.org/10.1038/sj.jes.7500534
January 2015 5-394 DRAFT: Do Not Cite or Quote
-------
Weichenthal S: Kulka. R: Dubeau. A: Martin. C: Wang. D: Dales. R. (2011). Traffic-related air pollution
and acute changes in heart rate variability and respiratory function in urban cyclists. Environ Health
Perspect 119: 1373-1378. http://dx.doi.org/10.1289/ehp.1003321
Weinmavr. G: Romeo. E: De Sario. M: Weiland. SK: Forastiere. F. (2010). Short-term effects of PM10
and NO2 on respiratory health among children with asthma or asthma-like symptoms: a systematic
review and meta-analysis [Review]. Environ Health Perspect 118: 449-457.
http://dx.doi.org/10.1289/ehp.0900844
Wellenius. GA: Burger. MR: Coull BA: Schwartz. J: Suh. HH: Koutrakis. P: Schlaug. G: Gold. PR:
Mittleman. MA. (2012). Ambient air pollution and the risk of acute ischemic stroke. Arch Intern Med
172: 229-234. http://dx.doi.org/10.1001/archinternmed.2011.732
Wellenius. GA: Schwartz. J: Mittleman. MA. (2005). Air pollution and hospital admissions for ischemic
and hemorrhagic stroke among medicare beneficiaries. Stroke 36: 2549-2553.
http://dx.doi.org/10.1161/01.STR.0000189687.78760.47
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: 26. http://dx.doi.org/10.1186/1476-069X-6-26
Wendt JK: Symanski. E: Stock. TH: Chan. W: Du. XL. (2014). Association of short-term increases in
ambient air pollution and timing of initial asthma diagnosis among Medicaid-enrolled children in a
metropolitan area. Environ Res 131: 50-58. http://dx.doi.Org/10.1016/j.envres.2014.02.013
Wichmann. J: Folke. F: Torp-Pedersen. C: Lippert F: Ketzel M: Ellermann. T: Loft. S. (2013). Out-of-
hospital cardiac arrests and outdoor air pollution exposure in Copenhagen, Denmark. PLoS ONE 8:
e53684. http://dx.doi.org/10.1371/journal.pone.0053684
Wichmann. J: Ketzel. M: Ellermann. T: Loft. S. (2012). Apparent temperature and acute myocardial
infarction hospital admissions in Copenhagen, Denmark: A case-crossover study. Environ Health 11:
19. http://dx.doi.org/10.1186/1476-069X-ll-19
Williams. R: Brook. R: Bard. R: Conner. T: Shin. H: Burnett. R. (2012a). Impact of personal and ambient-
level exposures to nitrogen dioxide and particulate matter on cardiovascular function. Int J Environ
Health Res 22: 71-91. http://dx.doi.org/10.1080/09603123.2011.588437
Williams. R: Jones. P: Croghan. C: Thornburg. J: Rodes. C. (2012b). The influence of human and
environmental exposure factors on personal NO(2) exposures. J Expo Sci Environ Epidemiol 22: 109-
115. http://dx.doi.org/10.1038/jes.2011.20
Winquist A: Kirrane. E: Klein. M: Strickland. M: Darrow. LA: Sarnat SE: Gass. K: Mulholland. J:
Russell A: Tolbert. P. (2014). Joint effects of ambient air pollutants on pediatric asthma emergency
department visits in Atlanta, 1998-2004. Epidemiology 25: 666-673.
http://dx.doi.org/10.1097/EDE.0000000000000146
Witten. A: Solomon. C: Abbritti. E: Arjomandi. M: Zhai. W: Kleinman. M: Balmes. J. (2005). Effects of
nitrogen dioxide on allergic airway responses in subjects with asthma. J Occup Environ Med 47:
1250-1259. http://dx.doi.org/10.1097/01.iom.0000177081.62204.8d
Wittkopp. S: Staimer. N: Tjoa. T: Gillen. D: Daher. N: Shafer. M: Schauer. JJ: Sioutas. C: Delfino. RJ.
(2013). Mitochondrial Genetic Background Modifies the Relationship between Traffic-Related Air
Pollution Exposure and Systemic Biomarkers of Inflammation. PLoS ONE 8: e64444.
http://dx.doi.org/10.1371/journal.pone.0064444
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/j.ijheh.2011.03.003
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
January 2015 5-395 DRAFT: Do Not Cite or Quote
-------
Wong. CM: Vichit-Vadakan. N: Kan. H: Qian. Z. (2008). Public Health and Air Pollution in Asia (PAPA):
A multicity study of short-term effects of air pollution on mortality. Environ Health Perspect 116:
1195-1202. http://dx.doi.org/10.1289/ehp.11257
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: Qin. 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). (ISSN 1041-5505). Boston, MA: Health Effects Institute.
http://pubs.healtheffects.org/getfile.php?u=595
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. (1999). Air pollution and hospital
admissions for respiratory and cardiovascular diseases in Hong Kong. Occup Environ Med 56: 679-
683.
Xiang. H: Mertz. KJ: Arena. VC: Brink. LL: Xu. X: Bi. Y: Talbott EO. (2013). Estimation of short-term
effects of air pollution on stroke hospital admissions in Wuhan, China. PLoS ONE 8: e61168.
http://dx.doi.org/10.1371/journal.pone.0061168
Xu. Z: Hu. W: Williams. G: Clements. AC: Kan. H: Tong. S. (2013). Air pollution, temperature and
pediatric influenza in Brisbane, Australia. Environ Int 59: 384-388.
http://dx.doi.0rg/10.1016/j.envint.2013.06.022
Xuan. W: Peat. JK: Toelle. BG: Marks. GB: Berry. G: Woolcock. AJ. (2000). Lung function growth and
its relation to airway hyperresponsiveness and recent wheeze: Results from a longitudinal population
study. Am J Respir Crit Care Med 161: 1820-1824. http://dx.doi.Org/10.1164/airccm.161.6.9809118
Yamazakj S: Shima. M: Ando. M: Nitta. H: Watanabe. H: Nishimuta. T. (2011). Effect of hourly
concentration of particulate matter on peak expiratory flow in hospitalized children: A panel study.
Environ Health 10: 15. http://dx.doi.org/10.1186/1476-069X-10-15
Yang. CL: To. T: Foty. RG: Stieb. DM: Dell SD. (2011). Verifying a questionnaire diagnosis of asthma in
children using health claims data. BMC Pulm Med 11: 52. http://dx.doi.org/10.1186/1471-2466-ll-52
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. http://dx.doi.org/10.1080/15287390490276502
Yang. Q: Chen. Y Krewski. D: Burnett. RT: Shi. Y Me Grail. KM. (2005). 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/j.envres.2004.09.014
Yang. Q: Chen. Y: Shi. Y: Burnett. RT: Mcgrail. KM: Krewski. D. (2003). Association between ozone and
respiratory admissions among children and the elderly in Vancouver, Canada. Inhal Toxicol 15: 1297-
1308. http://dx.doi.org/10.1080/08958370390241768
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
January 2015 5-396 DRAFT: Do Not Cite or Quote
-------
Zeghnoun. A: Beaudeau. P: Carrat F: Delmas. V: Boudhabhav. O: Gavon. F: Guincetre. D: Czernichow.
P (1999). Air pollution and respiratory drug sales in the city of Le Havre, France, 1993-1996. Environ
Res 81: 224-230. http://dx.doi.org/10.1006/enrs.1999.3974
Zemek. R: Szyszkowicz. M: Rowe. BH. (2010). Air pollution and emergency department visits for otitis
media: A case-crossover study in Edmonton, Canada. Environ Health Perspect 118: 1631-1636.
http://dx.doi.org/10.1289/ehp.0901675
Zhang. J: Zhu. T: Kipen. H: Wang. G: Huang. W: Rich. D: Zhu. P: Wang. Y: Lu. SE: Ohman-Strickland.
P: Diehl S: Hu. M: Tong. J: Gong. J: Thomas. D: Committee. HHR. (2013). Cardiorespiratory
biomarker responses in healthy young adults to drastic air quality changes surrounding the 2008
Beijing Olympics. Res Rep Health Eff Inst Feb: 5-174.
Zhang. Z: Wang. J: Chen. L: Chen. X: Sun. G: Zhong. N: Kan. H: Lu. W. (2014). Impact of haze and air
pollution-related hazards on hospital admissions in Guangzhou, China. Environ Sci Pollut Res Int 21:
4236-4244. http://dx.doi.org/10.1007/sll356-013-2374-6
Zheng. S: Wang. M: Wang. S: Tao. Y: Shang. K. (2013). Short-term effects of gaseous pollutants and
particulate matter on daily hospital admissions for cardio-cerebrovascular disease in Lanzhou:
evidence from a heavily polluted city in China. Int J Environ Res Public Health 10: 462-477.
http://dx.doi.org/10.3390/ijerphl0020462
Zhu. T. (2013). Email from Dr. Zhu to Dr. Patel; Response to data request. Available online
Zora. JE: Sarnat SE: Ravsoni. AU: Johnson. BA: Li. WW: Greenwald. R: Holguin. F: Stock. TH: Sarnat.
JA. (2013). Associations between urban air pollution and pediatric asthma control in El Paso, Texas.
Sci Total Environ 448: 56-65. http://dx.doi.Org/10.1016/i.scitotenv.2012.ll.067
January 2015 5-397 DRAFT: Do Not Cite or Quote
-------
CHAPTER 6 INTEGRATED HEALTH EFFECTS
OF LONG-TERM EXPOSURE TO OXIDES OF
NITROGEN
6.1 Scope and Issues Considered in Health Effects Assessment
6.1.1 Scope of Chapter
1 As in the preceding chapter for short-term exposure, with consideration of exposure
2 measurement error, effects of other correlated pollutants, and mode of action information
3 to support biological plausibility, this chapter summarizes, integrates, and evaluates the
4 evidence for a broad spectrum of health effects associated with long-term exposure
5 (i.e., more than 1 month to years) to oxides of nitrogen. This chapter comprises
6 evaluations of the epidemiologic and toxicological evidence for the effects of long-term
7 exposure to oxides of nitrogen on health outcomes related to respiratory effects
8 (Section 6.2). cardiovascular and related metabolic effects (Section 6.3). reproductive and
9 developmental effects (Section 6.4). and mortality (Section 6.5). Chapter 6 concludes
10 with a discussion of the evidence for the cancer effects of oxides of nitrogen
11 (Section 6.6). To characterize the weight of evidence in a cohesive manner, results from
12 both short-term (i.e., up to 1 month) and long-term exposure studies specific to
13 reproductive and developmental effects are included in this chapter. These results are
14 identified according to exposure duration in the text and tables throughout Section 6.4.
15 Individual sections for broad health categories (e.g., respiratory effects) begin with a
16 summary of conclusions from the 2008 ISA for Oxides of Nitrogen (U.S. EPA. 2008)
17 followed by an evaluation of recent (i.e., published since the completion of the 2008 ISA
18 for Oxides of Nitrogen) studies that builds upon evidence from previous reviews. Within
19 each of these sections, results are organized into smaller outcome groups (e.g., asthma
20 development) that are made up of a continuum of subclinical to clinical effects. The
21 discussion of individual effects is then organized by specific scientific discipline
22 (i.e., epidemiology, toxicology). This organization permits a clear description of the
23 extent of coherence and biological plausibility for the effects of oxides of nitrogen on a
24 group of related outcomes, and in turn, a transparent characterization of the weight of
25 evidence in drawing conclusions.
26 Sections for each of the broad health categories (e.g., respiratory effects, cardiovascular,
27 and related metabolic effects) conclude with an integrated assessment of evidence and
28 conclusions regarding causality. A determination of causality has been made for a broad
January 2015 6-1 DRAFT: Do Not Cite or Quote
-------
1 health category (e.g., respiratory effects) or smaller group of related outcomes (e.g., birth
2 outcomes) by evaluating the evidence for each category or group independently with the
3 causal framework (described in the Preamble to this ISA). A unique situation arises in the
4 evaluation of mortality. Findings for cause-specific mortality (i.e., respiratory,
5 cardiovascular) were used to assess the continuum of effects and inform the causal
6 determinations for respiratory and cardiovascular effects. A separate causal determination
7 was made for total mortality (Section 6.5). based primarily on the evidence for
8 nonaccidental causes of mortality combined but also informed by the extent to which
9 evidence for the spectrum of cardiovascular and respiratory effects provides biological
10 plausibility for NO2-related total mortality. Judgments regarding causality were made by
11 evaluating the evidence for the full range of exposures to oxides of nitrogen or ambient
12 concentrations in animal toxicological and epidemiologic studies defined in this ISA to
13 be relevant to ambient exposure (i.e., concentrations up to 5,000 ppb as described in
14 Section 1.2. Experimental studies that examined higher concentrations were evaluated
15 particularly to inform mode of action.
6.1.2 Evidence Evaluation and Integration to Form Causal Determinations
16 As for relationships of health effects with short-term exposure, judgments regarding
17 causality were made by evaluating evidence for the consistency of findings across
18 multiple studies, the coherence of findings across related endpoints and across
19 disciplines, and the extent to which chance, confounding (i.e., bias due to a correlation
20 with NO2 exposures or ambient concentrations and relationship with the outcome), and
21 other biases could be ruled out with reasonable confidence. This evaluation involved the
22 integration among various lines of evidence and consideration of the quality of individual
23 studies (detailed in Section 5.1.2 and Table 5-1).
24 Epidemiologic studies of long-term NOx or NCh exposure generally rely on
25 between-subject differences in exposure between subjects. These differences may be by
26 location of residence (spatial differences) or time periods that vary in long-term ambient
27 NOx or NO2 concentrations, for example. For the assessment of potential confounding,
28 long-term exposure epidemiologic studies were evaluated for the extent to which they
29 considered other factors associated with health outcomes and were spatially correlated
30 with NOx or NO2 exposure that varied between subjects. These potential confounding
31 factors can include socioeconomic status (SES), diet, smoking or exposure to
32 environmental tobacco smoke, medication use, and copollutant exposures (Table 5-1).
33 Epidemiologic studies varied in the extent to which they considered potential
34 confounding. Because no single study considered all potential confounding factors and
35 not all potential confounding factors were examined in the collective body of evidence,
January 2015 6-2 DRAFT: Do Not Cite or Quote
-------
1 residual confounding by unmeasured factors is possible. Residual confounding is also
2 possible by poorly measured factors. The evidence was examined based on factors well
3 documented in the literature to be associated with NO2 exposure and health outcomes.
4 The limitations of multivariable models, including copollutant models, to examine
5 potential confounding were considered in drawing inferences about the independent
6 effects of NO2 (Section 5.1.2.2). Specific to copollutant confounding, the magnitude of
7 correlations between NO2 and copollutants is considered. The potential for differential
8 measurement error for NO2 and copollutants is also considered.
9 This ISA presents epidemiologic effect estimates for associations with health outcomes
10 scaled to the same increment of oxides of nitrogen to increase comparability among
11 studies that report effect estimates scaled to various changes in concentrations of oxides
12 of nitrogen (e.g., interquartile range of concentrations for the study period or an arbitrary
13 unit such as 10 ppb). For long-term exposure metrics, effect estimates are scaled to a
14 10-ppb increase in NO2 or NO and a 20-ppb increase in NOx. These increments were
15 derived by calculating the U.S. nationwide percentile distributions for annual average
16 concentrations (Table 2-2) and then calculating the approximate difference between the
17 median (atypical pollution year) and the 95th percentile (a more polluted year) of annual
18 average concentrations among monitors in the State and Local Air Monitoring Stations
19 network. Long-term averages of ambient oxides of nitrogen are lower in concentration
20 than short-term averages, less variable across time, and do not differ widely among
21 averages of multiple months, annual averages, or multiyear averages [see Table S6-1;
22 (U.S. EPA. 2014)]. Thus, all long-term exposure metrics were scaled to the same
23 increment. Effect estimates that were reported in terms of ug/m3 are converted to ppb and
24 standardized for NO2 and NO but not NOx. Because the proportions of NO2 and NO are
25 unknown for the various NOx metrics, concentrations cannot be converted from ug/m3 to
26 ppb. And, data are not available to calculate the percentiles of NOx concentrations in
27 ug/m3 at a national scale for the U.S. or other countries. Therefore, the ISA presents
28 effect estimates based on ug/m3 of NOx as they are reported in individual studies.
29 To form causal determinations, evidence was integrated across a spectrum of related
30 endpoints, including cause-specific mortality, and across disciplines to assess the extent
31 to which chance, confounding, and other biases could be ruled out with reasonable
32 confidence. Animal toxicological studies can provide direct evidence for health effects
33 related to NO2 exposures. Coherence between toxicological and epidemiologic findings
34 can address uncertainties such as whether epidemiologic associations with health
35 outcomes reflect an independent effect of ambient NO2 exposure or are potentially
36 confounded by other factors. Experimental studies also can provide biological plausibility
37 by describing key events within the modes of action for health effects. Thus, integration
38 of evidence was used to inform uncertainties for any particular outcome or discipline due
January 2015 6-3 DRAFT: Do Not Cite or Quote
-------
1 to factors such as publication bias, selection bias, exposure measurement error, or
2 confounding by copollutant exposures. The subsequent sections assess study quality and
3 strength of inference and integrate evidence across multiple lines of evidence to evaluate
4 relationships between oxides of nitrogen and various health effects.
6.2 Respiratory Effects
6.2.1 Introduction
5 The 2008 ISA for Oxides of Nitrogen (U.S. EPA. 2008) examined the epidemiologic and
6 toxicological evidence for the relationship between long-term exposure to NO2 and
7 respiratory effects and concluded that the evidence was suggestive, but not sufficient, to
8 infer a causal relationship. The key supporting evidence comprised associations in
9 epidemiologic studies of higher NC>2 exposure with decrements in lung function and
10 partially irreversible decrements in lung development in children. However, several
11 sources of uncertainty were acknowledged. For example, results from the Southern
12 California Children's Health Study (CHS) indicated that decrements in lung development
13 in children were associated with higher ambient NC>2 concentrations (Gauderman et al.,
14 2004). but similar associations were also found for traffic-related pollutants such as PM2 5
15 and EC, proximity to traffic (<500 m), as well as acid vapor, Os, and PMio. Generally, the
16 high correlation among long-term averages of traffic-related pollutants made it difficult
17 to discern the independent effects of NCh. Further, although animal toxicological studies
18 demonstrated that long-term exposure to NC>2 resulted in permanent morphologic changes
19 to the lung, particularly in the centriacinar region and bronchiolar epithelium, they
20 provided little evidence for effects on key events within the mode of action for
21 NC>2-related decreases in lung function or development. Additional uncertainty was
22 related to the inconsistent cross-sectional evidence for associations between long-term
23 exposure to NCh and increases in asthma prevalence and incidence. For example, two
24 cohort studies, the CHS in Southern California (Gauderman et al.. 2005) and a birth
25 cohort study in the Netherlands (Brauer et al., 2007) observed positive associations, while
26 other studies did not find consistent associations between long-term NCh exposure and
27 asthma. Epidemiologic studies conducted in both the U.S. and Europe also reported
28 inconsistent results regarding an association between long-term exposure to NCh and
29 respiratory symptoms.
30 This section presents the current body of evidence examining the relationship between
31 long-term exposure to NC>2 and respiratory effects. A large body of recent studies has
32 evaluated the development of asthma and bronchitis and reduced lung function and
January 2015 6-4 DRAFT: Do Not Cite or Quote
-------
1 development in children and the incidence of asthma and bronchitis in adults.
2 Longitudinal studies of the incidence of asthma in children, the largest and strongest
3 evidence base, is presented first. Development of allergic disease, chronic disease
4 severity, lung function and development, respiratory infections, chronic obstructive
5 pulmonary disease (COPD), and respiratory morphology are discussed thereafter. No
6 recent animal toxicological studies evaluating respiratory effects of long-term NO2
7 exposure have been published since the release of the 2008 ISA for Oxides of Nitrogen
8 (U.S. EPA. 2008). but previous studies are evaluated to inform the biological plausibility
9 for the array of respiratory effects examined.
10 Emphasis is placed on the longitudinal cohort studies, which compared to cross-sectional
11 studies can better characterize the temporality between exposure and incidence of a
12 health effect. For the onset of asthma, the prospective designs take into account the
13 difference between the first occurrence of asthma and the exacerbation of asthma by
14 defining asthma incidence as diagnosis of asthma by a physician in the time since the
15 previous follow-up period. For other health effects, such as respiratory symptom
16 occurrence or pulmonary function changes, aspects of the study design or statistical
17 methods must consider potential effects of short-term exposure.
18 NO2 exposure assessment methods are discussed to describe the utility of various
19 methods in representing the variability in NO2 concentrations in the study areas and, in
20 turn, the strength of inference about relationships with respiratory effects. The potential
21 for confounding by traffic-related copollutants is discussed also. Correlations between
22 NO2 and traffic-related pollutants [e.g., CO, EC/BC, and PIVb 5 among others] are
23 presented in the summary tables and text. Copollutant regression modeling results are
24 discussed in the text when they are available. Long-term exposure to indoor
25 concentrations of NO2 has been examined in relation to some respiratory effects and is
26 discussed to describe the extent of coherence with the effects observed with ambient
27 exposure to NO2. NO2 exposure estimates examined in the epidemiologic studies in this
28 section are generally annual averages unless stated otherwise.
6.2.2 Development of Asthma or Chronic Bronchitis
29 Asthma is a chronic disease characterized by chronic inflammation, development of
30 airway hyperresponsiveness (AHR), and in some cases, airway remodeling. In
31 characterizing the evidence for a relationship between long-term NO2 exposure and
32 asthma development, this section evaluates asthma incidence in children in longitudinal
33 cohort studies. Cross-sectional and prevalence studies were reviewed and are discussed to
34 inform discussion of potential copollutant confounding and other policy-relevant issues
January 2015 6-5 DRAFT: Do Not Cite or Quote
-------
1 [see Annex tables, Tables AX6 3-15, AX6 3-16, and AX6 3-17 of the 2008 ISA for
2 Oxides of Nitrogen (U.S. EPA. 2008) for descriptions of previous studies]. This section
3 also characterizes evidence for airway responsiveness, allergic sensitization, and
4 pulmonary inflammation, which are key events in the mode of action for a relationship
5 between NO2 exposure and asthma development (Figure 4-2). A few studies examined a
6 composite index of asthma and chronic bronchitis or chronic bronchitis alone.
6.2.2.1 Asthma or Chronic Bronchitis in Children
7 Since the 2008 ISA for Oxides of Nitrogen (U.S. EPA. 2008). recent prospective and
8 retrospective longitudinal cohort studies provide a strong evidence base that generally
9 demonstrates a positive relationship between asthma incidence in children and long-term
10 NO2 exposure. Details from these key longitudinal studies are presented in Table 6-1 and
11 Figure 6-1. The evidence base includes studies from North America, Europe, and Asia
12 that use varied designs and analyses. Several exposure methods were used that were
13 designed to provide individual exposure estimates. A uniform health-effect indicator,
14 physician-diagnosed asthma, was used. The studies followed children from birth to ages
15 7 to 12 years of age or from ages 8 to 18 years of age.
January 2015 6-6 DRAFT: Do Not Cite or Quote
-------
Table 6-1 Prospective cohort studies of long-term exposure to nitrogen dioxide (NO2) or sum of NO2 and nitric
oxide (NOx) and asthma incidence in children.
Study3
Exposure Assessment Pollutant Correlation Statistical Methods
Comments
Results (95% Cl)b
Vancouver high asthma risk birth cohort
Carlsten et al. (2011 c)
n = 184 children at 7 yr of
age
Related publications:
Carlsten et al. (2011 a)
Carlsten et al. (2011b)
Henderson et al. (2007)
Cohort with high risk of
asthma: parental report
of at least 1 first-degree
relative with asthma or 2
first-degree relatives with
other allergic disease
(atopic dermatitis,
seasonal or perennial
allergic rhinitis, or food
allergy).
LUR used to estimate
annual concentrations at
the birth residential
address of each subject
for NO2, PM2.5, BC, NO
as previously developed,
described and validated
(Henderson et al., 2007).
Resolution 10m.
Annual means from
16 monitors show a
strong 1:1 relationship
with the mean
concentrations for sites
used in LUR models.
Slope = 0.89 (R2 = 0.98).
NO2 normally distributed
with a mean of 16.2 (SD
of5.6)ppb. TheR2
values for NO2 models
(114 sites) range from
0.56 to 0.60 are
consistent across models
built with different traffic
variables. Mean error
estimates based on leave
one out cross validation
had SD of about 15%
[0.0 (2.75)].
Pearson R:
NO-NO2 = 0.8
NO2-PM2.5 = 0.7
NO-PM2.5 = 0.5
BC-NO2 = 0.5
BC-NO = 0.3
Multiple logistic
regression analysis
adjusted for maternal
education, history of
asthma (in mother, father
or siblings), atopic status
at age 1 yr.
Follow-up at 7 yr of age,
represented 63% of the
cohort entering the study
at its onset. The key
characteristics of children
who returned for
assessment did not differ
from those in the original
cohort.
Air pollution estimates for
1995 generated from
2003 annual averages
adjusted for temporal
trends.
OR among all children:
2.9(0.8, 10.9)
Association observed
with PM2.5 with wide Cl;
no association observed
for BC.
OR in 13 children with
both allergist diagnosis of
asthma and bronchial
hyperactivity:
1.3(0.9, 2.2) per 10-ppb
NO
January 2015
6-7
DRAFT: Do Not Cite or Quote
-------
Table 6-1 (Continued): Prospective cohort studies of long-term exposure to nitrogen dioxide (NO2) or sum of NO2
and nitric oxide (NOx) and asthma incidence in children.
Study3
Exposure Assessment Pollutant Correlation Statistical Methods
Comments
Results (95% Cl)b
Children, Allergy, Milieu, Stockholm, Epidemiology Survey (BAMSE)
Gruzieva et al. (2013)
N = 3,633
Birth cohort followed up
to 12 yr of age. Enrolled
between 1994 and 1996.
Related publications:
Gruzieva et al. (2012)
Nordlinq et al. (2008)
Wickman et al. (2002)
Dispersion models used
to estimate NOx for all
addresses in the yr 1994
to 2008 representing
when the first child was
born until the end of the
12-yr follow-up.
Model is regularly
validated against
measurements at air
quality-monitoring
stations. The correlation
between measured NO2
concentrations and
calculated traffic-related
NO2 from dispersion
modeled NOx for
487 addresses in the
study had an rvalue of
0.74.
Used time and activity
patterns and different
places of exposure to
estimate exposure.
r= 0.96 between NOx
and PM-io concentrations
during the first yr of life.
Multinomial
regression/GEE adjusted
for municipality, SES, yr
the house was built, and
mother or father with
doctor diagnosis of
asthma and asthma
medication.
Associations were
stronger for the oldest
children and for
nonallergic asthma.
Participation at the 1 st yr
was 96%, at the 2nd yr
94%, at4yr91%, at 8 yr
84%, andat12yr82%.
The distribution of risk
factors was similar in the
study population and the
original cohort.
OR for NOx during the
first yr of life and
development of incident
asthma at 12 yr of age.
1.87(1.0, 3.44) per
46.8 ug/m3 NOx
January 2015
6-8
DRAFT: Do Not Cite or Quote
-------
Table 6-1 (Continued): Prospective cohort studies of long-term exposure to nitrogen dioxide (NO2) or sum of NO2
and nitric oxide (NOx) and asthma incidence in children.
Study3
Exposure Assessment Pollutant Correlation Statistical Methods
Comments
Results (95% Cl)b
Chiba prefecture, Japan cohort
Shima et al. (2002)
N = 1,910 children in
8 communities at age
6 yr.
1st graders between
1989 and 1992 were
followed to 6th grade.
The average annual
concentrations of air
pollutants for the 10-yr
period from 1988 to 1997
at ambient air monitoring
stations approximately 1
to 2 km from study
schools. Almost all the
childrens' homes and
schools were about 1 km
from the sites.
NR
Logistic regression model
adjusted for sex, history
of allergic diseases,
respiratory diseases prior
to age 2 yr, parental
history of allergic
diseases, maternal
smoking habits, type of
heater used in winter in
the home, and
construction elements of
the house.
The follow up included
1,910 children (66.9% of
the original cohort) at
6 yr. The percentage of
children who were not
followed was slightly
higher in urban than in
rural communities
because urban subjects
changed residence more
frequently. Questionnaire
responses were
unavailable for
944 children in the
follow-up period,
primarily because of
changed residence 3 yr
prior to entering 1st
grade, which provides an
exposure estimate
related to the child's
location.
OR: 1.71
(1.04,2.79)
January 2015
6-9
DRAFT: Do Not Cite or Quote
-------
Table 6-1 (Continued): Prospective cohort studies of long-term exposure to nitrogen dioxide (NO2) or sum of NO2
and nitric oxide (NOx) and asthma incidence in children.
Study3
Exposure Assessment Pollutant Correlation Statistical Methods
Comments
Results (95% Cl)b
Taiwan Children Health Study (TCHS)
Leeetal. (2012b)
N = 3,160 from
14 communities
Ages 12-14yrat
baseline in 2007.
Mean follow-up 2 yr.
Annual avg NO2
calculated from
14 monitoring stations
between 2007 and 2009.
Almost all childrens'
homes and schools were
1 km from the sites.
Two NO2 strata:
Low: < median 17.5 ppb
High: > median 17.5 ppb
Median annual avg NO2
22.1 ppb in high
communities
14.0 ppb in low
communities
NR
Poisson regression
models adjusted for
prenatal maternal
smoking, family history of
asthma, family history of
atopy, and community.
At the 2-yr follow-up
period, 96.9% of the
children completed the
questionnaire and
pulmonary function tests.
NO2 observed to modify
the protective effect of
high lung function on risk
of asthma incidence.
RR for asthma per
interquartile range
increase in percentage
predicted FVC:
Low NO2 communities:
0.82(0.72,0.93)
High NO2 communities:
1.00(0.88, 1.14)
January 2015
6-10
DRAFT: Do Not Cite or Quote
-------
Table 6-1 (Continued): Prospective cohort studies of long-term exposure to nitrogen dioxide (NO2) or sum of NO2
and nitric oxide (NOx) and asthma incidence in children.
Study3
Exposure Assessment Pollutant Correlation Statistical Methods
Comments
Results (95% Cl)b
British Columbia birth cohort
Clark etal. (2010)
N= 2,801
Mean age at follow-up:
48 mo (SD: 7).
All 1999 and 2000 births
in southwest British
Columbia eligible.
Related publication:
(Henderson etal.. 2007).
Central site monitors,
LUR, and point source
derived IDW summation
of emissions.
All estimated for the
postal code level. (92%
was at resolution of a city
block or block face).
For LUR, model R2 is
0.53 for NO2. In the
sampling yr(2003),
measurements at LUR
sites exhibit a strong 1:1
relationship with annual
means at central
monitoring sites. Slopes
are 1.03 (R2 = 0.96) and
0.89 (R2 = 0.98) for NO
and NO2, respectively
(Henderson etal.. 2007).
For IDW, exposures
assigned using the
3 closest monitors within
50 km weighted by their
distance to the postal
code of interest.
Correlations among
pollutants were generally
high. Quantitative results
reported only for Os.
r= -0.7 to -0.9.
Conditional logistic
regression adjusted for
native status,
breast-feeding, maternal
smoking, income quartile,
maternal age, birth
weight, and gestational
length.
The potential limitation of
the young age of the
children when wheezing
is more common was
addressed by restricting
asthma cases to children
with a hospital admission
or at least two outpatient
diagnoses of asthma
which indicate severe
ongoing symptoms.
LUR and IDW results for
NO2 were similar.
Results for PlVh.s were
smaller for both LUR and
IDW than for NO2.
OR for LUR:
1.26(1.08, 1.48)
OR for IDW:
1.24(1.14, 1.34)
Asthma diagnosis
associated with early life
exposure to CO, PM-io,
SO2, BC and proximity to
point sources.
Traffic-related pollutants
were associated with the
highest risks.
January 2015
6-11
DRAFT: Do Not Cite or Quote
-------
Table 6-1 (Continued): Prospective cohort studies of long-term exposure to nitrogen dioxide (NO2) or sum of NO2
and nitric oxide (NOx) and asthma incidence in children.
Study3
Exposure Assessment Pollutant Correlation Statistical Methods
Comments
Results (95% Cl)b
Childrens Health Study (CHS), Southern California communities
Jerrett et al. (2008)
N= 217 children
Ages 10-1 Syr
Enrolled in 1993 or 1996
from 11 communities.
Asthma assessed over
8 yr of follow-up.
Palmes tubes outside
home, 2 weeks summer
and winter to provide
annual and seasonal
levels.
Over the
11 communities, mean
(SD) annual avg NO2
ranged from 9.6 (2.5) to
51.3(4.4)ppb.
No quantitative data.
Correlations between
residential NO2 and
various measures of
traffic proximity or
modeled pollutant
concentrations reported
to be moderate to high.
Random-effects Cox
proportional hazards
models adjusted for
median household
income, proportion of
respondents with low
education, percentage of
males unemployed,
percentage living in
poverty, temperature,
and humidity.
Within-community effects HR: 1.51 (1.12, 2.05)
indicative of long-term
local traffic sources were
similar to effects of
community average NO2
across communities,
suggesting that both
regional and local
pollution contributed to
associations with
asthma. The range of
NO2 within communities
was smaller than that
between communities.
McConnell et al. (2010)
N = 120
Ages 4.8 to 9.0 yr
New cohort established
2002-2003 in
13 communities.
3 yr of follow-up
Related publications:
Wu et al. (2005)
Peters et al. (1999)
Benson (1984)
Community central site
pollutant measurements
and line source
dispersion model for
residential and school
NOx.
The overall
within-community
variability of personal
exposures using
time-activity categories
across communities was
highest for NO2, followed
by EC, PM-io, PM2.5, and
CO (Wuetal.,2005).
NR
In 2000, correlations for
measured NO2 or
residential NO2 with
freeway- and
nonfreeway-related NO2
from the dispersion
model were r = 0.56 and
r= 0.34, respectively,
indicating that the
measured and modeled
metrics may have some
level of independence
(Gauderman et al..
2005).
Multilevel Cox
proportional hazards
model adjusted for
sociodemographic
characteristics, exposure
to cigarette and wildfire
smoke, health insurance,
housing characteristics,
history of allergy, and
parental asthma.
After a 3-yr follow-up
period, 74% of the
baseline cohort
remained. Follow-up was
lower among Hispanic
children and children with
lower SES than
non-Hispanic white
children and children with
higher SES. However,
the NO2 association was
similar after adjusting for
these factors. Risk was
higher in children with
high parental stress
compared to low parental
stress (Shankardass
etal.. 2009).
HR for central site NO2:
1.39(1.07, 1.80)
HR for modeled NOx
from freeways:
1.67(1.32, 2.12) near
homes
1.88(1.10, 3.19) near
schools.
OR for PlVh.5 central site
(range 13.9 to
17.4ug/m3):
1.66(0.91,3.05)
January 2015
6-12
DRAFT: Do Not Cite or Quote
-------
Table 6-1 (Continued): Prospective cohort studies of long-term exposure to nitrogen dioxide (NO2) or sum of NO2
and nitric oxide (NOx) and asthma incidence in children.
Study3
Exposure Assessment Pollutant Correlation Statistical Methods
Comments
Results (95% Cl)b
Prevention and Incidence of Asthma and Mite Allergy (PIAMA) study, the Netherlands
Gehrinq et al. (2010)
N = 3,863
Follow-up from birth to
age 8 yr.
Related publications:
Wijqaetal. (2014)
Eeftensetal. (2011)
Hoek et al. (2008)
Braueretal. (2007)
Braueret al. (2003)
Braueretal. (2002)
LUR to estimate annual
concentrations for birth
address of each child.
NO2 measurements
conducted in 2007
agreed well with NO2
measurements taken in
1999-2000 at the same
locations (R2 = 0.86).
LUR models from
1999-2000 and 2007
explained 85 and 86% of
observed spatial
variance, respectively.
Developed from 40 sites.
16 urban/suburban,
12 regional, 12 traffic.
5-12% of population
lived near major roads.
Buffer >50 m.
NO2 annual mean:
(10-90%) 13.5 ppb
(7.8-18.5 ppb).
Copollutants were highly
correlated.
NO2-PM2.5:r=0.93
NO2-soot: r= 0.96
PlVh.s-soot: r= 0.97
Generalized estimating
equations adjusted for
sex, study arm, use of
mite-impermeable
mattress covers, allergies
of mother and father,
maternal and paternal
education, maternal
smoking during
pregnancy,
breastfeeding, presence
of a gas stove in the
child's home, presence of
older siblings, and any
smoking at home.
Participation was high at
all ages, starting with
94.4% in the first yr and
82% in the eighth yr.
Characteristics between
the original cohort and
studied groups are
similar.
Adjusted OR of 1.36
(1.09, 1.67) without
adjustment for study
region.
Similar results were
observed for PlVh.s and
soot.
Adjusted OR of 1.32
(0.93, 1.85) with
adjustment for study
region.
January 2015
6-13
DRAFT: Do Not Cite or Quote
-------
Table 6-1 (Continued): Prospective cohort studies of long-term exposure to nitrogen dioxide (NO2) or sum of NO2
and nitric oxide (NOx) and asthma incidence in children.
Study3
Exposure Assessment Pollutant Correlation Statistical Methods
Comments
Results (95% Cl)b
Gene and Environment Prospective Study in Italy (GASPII)
Ranzi etal. (2014)
N=672
Birth cohort enrolled from
2 large obstetric hospitals
in Rome from June 2003
to October 2004.
Related publications:
Cesaroni etal. (2012)
NO2 assessed for each
residential address
during the follow-up
period using LUR. NO2
measured simultaneously
at 78 locations in winter,
spring, and fall. Annual
average concentrations
calculated.
Mean (SD) NO2 across
78 sites: 23.7 (5.85) ppb
Os inversely correlated
on the spatial scale with
NO2.
Spearman r. -0.34.
Logistic regression
adjusted for sex, age,
breastfeeding at 3 mo,
day care attendance,
presence of any pets in
the home, siblings,
maternal and paternal
smoking, maternal
smoking during
pregnancy, maternal and
paternal education,
presence of molds or
dampness at home,
familial asthma or
allergies.
Information on
subsequent health
outcomes and additional
variables was obtained
by questionnaires at
6 mo, 15 mo, 4 yr, and
7 yr for 694, 664, 581
and 497 children,
respectively. Participation
was at 70% at the 7-yr
follow-up.
OR for time-weighted
average NO2:
1.17(0.63,2.20)
Adjusted for Os:
1.11 (0.54,2.25)
Genes-environment & admixture in Latino Americans and the study of African Americans, asthma, genes, & environments
Nishimura etal. (2013)
N= 4,320
Ages 8-21 yr
Multicity study: Chicago,
IL; Bronx, NY; Houston,
TX; San Francisco Bay
Area, CA and Puerto
Rico.
Average NO2 over the
first 3 yr of life were
calculated by averaging
NO2 using IDW from the
4 closest monitors within
50 km of the residence.
Mean (SD) NO2 across
cities:
9.9 (2.9) to 32.1 (5.7).
NR
Logistic regression
models adjusted for age,
sex, ethnicity, and
composite SES.
Sensitivity analysis
conducted with additional
covariates: maternal
gestational smoking, ETS
in the household
between 0 and 2 yr old,
and maternal language of
preference.
Region-specific results
suggest that risk of
asthma due to air
pollution may not be
uniform throughout the
nation and could depend
on local characteristics,
such as varying
proportions of different
racial/ethnic groups and
differing pollution sources
and/or weather patterns.
OR for NO2 during the
first yr of life, all cities
combined:
1.37(1.08, 1.72)
OR per 1 ug/m3 PM2.s:
1.03(0.90, 1.18)
January 2015
6-14
DRAFT: Do Not Cite or Quote
-------
Table 6-1 (Continued): Prospective cohort studies of long-term exposure to nitrogen dioxide (NO2) or sum of NO2
and nitric oxide (NOx) and asthma incidence in children.
Study3
Exposure Assessment Pollutant Correlation Statistical Methods
Comments
Results (95% Cl)b
Maternal-infant smoking study of East Boston
Clouqherty et al. (2007)
N= 413 full
cohort/255 lifetime
residents. Enrolled at
birth between 1987 and
1993.
Asthma ascertained in
1997 at mean age of
6.Syr.
LUR model (R2 = 0.83).
More variability in NO2
was explained by spatial
(R2 = 0.53) than temporal
variables (R2 = 0.29).
NR
Regression model
adjusted for maternal
asthma, education, and
smoking before and after
pregnancy, child's sex
and age.
Observed an association OR for NO2 in the yr of
between NO2 and
asthma solely among
urban children exposed
to violence.
diagnosis:
3.12(1.36, 7.15) in group
with high exposure to
violence.
Oslo Norway birth cohort
Oftedal et al. (2009a)
N= 2,329
Follow-up from birth in
1992-1993 to age
9-1 Oyr.
Related publications:
Oftedal et al. (2009b)
Laupsa and Slordal
(2003)
Walker etal. (1999)
Gronskei et al. (1993)
NO2 estimated by
dispersion model and
assigned at updated
individual addresses
during lifetime.
Dispersion model
estimates of long-term
NO2 averages were well
correlated with
measurements from
10 monitoring stations in
Oslo. r= 0.76.
PM-io and PlVh.s highly
correlated with NO2
(r= 0.79-0.91).
Cox proportional hazard
regression and logistic
regression adjusted for
sex, parental atopy,
maternal smoking in
pregnancy, paternal
education, and maternal
marital status at the
child's birth.
Several long-term
exposure periods
examined: exposure in
first yr of life, average
exposure from birth to
asthma onset, and
previous yr's exposure
before completing the
questionnaire.
RR for NO2 in first yr of
life and asthma onset at
any age:
0.87(0.76, 1.00)
Average NO2 onset not
associated with asthma
onset before or after age
4 yr. Increment of NO2
for RR not reported.
BAMSE = Children, Allergy, Milieu, Stockholm, Epidemiology Survey; BC = black carbon; CHS = Children's Health Study; Cl = confidence interval; CO = carbon monoxide;
EC = elemental carbon; ETS = environmental tobacco smoke; FVC = forced vital capacity; GASPII = Gene and Environmental Prospective Study in Italy; GEE = generalized
estimating equations; HR = hazard ratio; IDW = inverse distance weighting; LUR = land-use regression; NO = nitric oxide; NO2 = nitrogen dioxide; NOX = sum of NO and NO2;
OR = odds ratio; PIAMA = Prevention and Incidence of Asthma and Mite Allergy; PM25 = particulate matter with a nominal aerodynamic diameter less than or equal to 2.5 |jm;
PMio = particulate matter with a nominal aerodynamic diameter less than or equal to 10 (jm; RR = risk ratio(s), relative risk; SD = standard deviation; SES = socioeconomic status;
TCHS = Taiwan Children Health Study.
aStudies are presented in the order of appearance in the text.
""Results are presented for a 10 ppb change in NO2 unless otherwise specified.
January 2015
6-15
DRAFT: Do Not Cite or Quote
-------
Age in
Years of
Study
Clark et al. (2010)
Raiizi et ale (2014)
Gehring et al. (2010)
sten eta. (2011)
Gruzieva et al. (2013)
Shinia et al. (2002)
Nishimura et al. (2013)
f~*\ rtl 1 rtll a*+T 1 a+ o1 f'~)f\f\™\
ciougnerty et al. (200 /)
McConnell et al. (2010)
McConneU et al. (2010)
Oftedal et al. (2009)
Jon'ct c3! ;il. (JuoS)
Period of Exposure
birtli year
birtli up to 7 years
birth year
biitli year
lifetime
6 to 1 1 years old
birth year
year of diagnosis
5 to 9 years
5 to 9 years
birth up to 3 years
birth up to 10 years
14 to 17 years old
Years
4
Up to 7
Ito8
up to 7
up to 12
6f _ 1 ->
to 1^
8 to 21
up to 18
4.8 to 9
4.8 to 9
up to 10
1 1 K"k 1 fi
14 tO lo
Follow-up
3 to 4
-
/
8
-
/
1 t
12
2 to 7
3
3
9 to 10
Notes
row
LUR
LUR region adjusted
LUR not region adjusted
T T TTJ
LUK.
Dispersion
Central site
Central site
T T FD
LL K
Dispersion
Central site
LUR - Early Onset -•-
Residential
-• —
—
0.5 1 1.5 2 2.5 3 3.5
Risk or Odds Ratio and 95 % Cl
Note: Studies in red are recent studies. Studies in black were included in the 2008 ISA for Oxides of Nitrogen. All effect estimates in
this plot are standardized to 10 ppb, with the exception of Gruzieva et al. (2013) and Oftedal et al. (2009a). See Table 6-1 for further
study details and quantitative results. Circles = NO2; Triangles = NO; Diamond = NOX.
Figure 6-1 Associations of long-term exposure to nitrogen dioxide (NOa),
nitric oxide (NO), and the sum of NO and NOa (NOx) with asthma
incidence from prospective studies of children.
i
2
3
4
5
6
Transient wheezing is common in infants and often resolves as the child ages (Martinez
etal.. 1995). and thus the reliability of asthma diagnosis in infants is a factor to consider.
As a child progresses in age, the reliability of diagnosis of asthma would be expected to
strengthen. Consistent with this hypothesis, associations of NC>2 and asthma incidence in
children are greater in magnitude at later age evaluation and longer follow-up time. In
these studies, the methods used to determine the incident of asthma are similar across
January 2015
6-16
DRAFT: Do Not Cite or Quote
-------
1 studies. Children were required to be disease free at the start of the study. In the majority
2 of studies, asthma incidence was assessed using an annual respiratory questionnaire that
3 asked parents whether a doctor has ever diagnosed the child as having asthma, without
4 having fulfilled the definition of asthma at any previous time of follow-up. Several
5 studies assessed asthma incidence in a different manner but also demonstrated association
6 with NO2. For example, Carlsten et al. (201 la) used a pediatric allergist to assess asthma
7 in the children when they were 7 years old. Gruzieva et al. (2013) defined asthma
8 incidence as children at 12 years of age having at least 4 episodes of wheeze in the last
9 12 months, or at least one episode in combination with prescription of inhaled
10 corticosteroids, which would have been provided by a physician making a diagnosis of
11 asthma. Kramer et al. (2009) examined the incidence of doctor-diagnosed
12 asthmatic/spastic/obstructive bronchitis/asthma, which is not uniform with the other
13 studies evaluating asthma incidence, and thus not included here. The use of
14 questionnaires to determine asthma incidence is a best practice (Burr. 1992; Ferris. 1978)
15 and adds to the strength of inference from the available studies.
16 Associations between long-term NC>2 and asthma incidence are observed consistently
17 across studies that are diverse in the age of assessment of asthma, period of NO2 exposure
18 examined, and duration of follow-up (Table 6-1 and Figure 6-1). Within the birth cohorts,
19 results are reported at age 7 through age 12 years reflecting the length of follow-up time.
20 Additionally, various periods of exposure were examined, including the first year of life,
21 the year previous to asthma diagnosis, and cumulative exposure. The cohorts that start at
22 age 6 years and go up to 17 years examined a similar array of exposure periods. In the
23 birth cohorts examined by Gehringetal. (2010) and Gruzieva et al. (2013). results are
24 presented for multiple ages in the follow-up period. In Gehring et al. (2010). age-specific
25 associations, estimated from models with air pollution-age interaction terms, indicate
26 small differences in the associations of NC>2 with asthma and related symptoms with age
27 (Figure 6-2). However, larger odds ratios (ORs) were observed at ages 6-8 years, which
28 is consistent with the view that as a child progresses in age, the reliability of the diagnosis
29 of asthma would be expected to strengthen. Gruzieva et al. (2013) observed an
30 association for NOx estimated for the first year of life by dispersion models with asthma
31 incidence at age 12 years but not at earlier ages, consistent with lower reliability of
32 asthma diagnosis in young children and the extended follow-up to obtain newer cases. No
33 association was observed with two other exposure period evaluated: (1) NOx average
34 since the date of the previous follow-up or (2) during the preceding 12 months. Both were
35 periods of lower exposure.
January 2015 6-17 DRAFT: Do Not Cite or Quote
-------
J.U
2,8 •
2.0
1.2
•
A A
Incident asthma past 12 mo.
,
T
IT T- K
m ' 9]
>
i
T "
1 I
yc
^'L jo* H-i- i-]
i _^
T i <
)t o?4
;i * t I J
T
.
i
f
c
0
^
"iC
T
)1
] T
1 §1^
.-4— L | - ~—- ~ = •
1
4
7
8 overall
Note: Results are presented as adjusted except study region odds ratios (ORs) with 95% confidence intervals. Because the study
region is an important determinant of air pollution concentrations in the LUR models that were used to estimate exposures, the
adjustment for region may be an over adjustment. ORs were calculated for an interquartile range increase in air pollution levels of
5.5 ppb for NO2; blank circle = NO2; Gray circle = PM2 5; and Black circle = soot.
Source: Reprinted with permission of American Thoracic Society, Gehring et al. (2010).
Figure 6-2 Adjusted overall and age-specific association between annual
average levels of air pollution at the birth address and asthma
during the first 8 years of life.
2
3
4
5
6
1
8
9
10
11
12
13
14
15
16
A pooled analysis of six birth cohort studies (5,115 children) examined incident
physician-diagnosed asthma from birth to 7-8 years of age [confirmed by pediatric
allergist in two cohorts; (Macintyre et al., 2014a)1. Individual estimates of annual average
NO2 assigned to each child's birth address using land-use regression (LUR) was
associated with asthma with an OR of 1.48% (95% CI: 1.06, 2.06) per 10-ppb increase in
NC>2. Recent meta-analyses (Anderson et al.. 2013; Gasana et al.. 2012; Powers et al..
2012; Takenoue et al.. 2012; Braback and Forsberg. 2009) also report positive
associations between long-term NO2 exposure and asthma. Some of these meta-analyses
mixed children and adults, and some included both cross-sectional and prospective
studies. In Japan, Hasunuma et al. (2014) evaluated the health-improving effect of
anti-air pollution measures from 1997 to 2009. Analysis showed that a reduction in the
ambient NO2 concentrations was associated with a reduction in the prevalence of asthma.
In limited analysis, Shima et al. (2002) observed a linear concentration-response
relationship for NCh-related asthma across communities. Whereas, Carlsten et al. (201 Ic)
observed higher risk estimates in the second and third tertiles of NCh; the very wide and
January 2015
6-18
DRAFT: Do Not Cite or Quote
-------
1 overlapping CIs do not strongly demonstrate a linear relationship. These studies did not
2 conduct analysis to evaluate whether there is a threshold effect.
3 The set of studies examining asthma among children used a variety of exposure
4 assessment techniques, including LUR models, inverse distance weighting (IDW),
5 Palmes tubes outside subjects' homes, dispersion models with or without central site
6 monitoring data, and central site monitoring data (Table 6-1). Information in
7 Section 3.4.5.2 aids interpreting these methods with regard to potential exposure
8 measurement error. Many asthma incidence studies used exposure assessment techniques
9 intended to characterize the exposure for subjects in a representative manner. The
10 limitations and strengths of these methods that help inform the strength of inference
11 about the relationship between NCh exposure and asthma incidence are discussed next.
12 Recent prospective birth cohort studies that estimated residential ambient NCh
13 concentrations using LUR provide strong support for a relationship with asthma
14 development (Figure 6-1). Misrepresenting the differences between subjects in NCh
15 exposure due to the high variability often observed in ambient NCh concentrations can
16 lead to bias in health effects association (Section 3.4.5.2). The studies of LUR are
17 noteworthy in that the NCh exposure estimates were spatially aligned to subjects'
18 residences, and many provided validation that LUR estimates well represented the spatial
19 variation in ambient NCh concentrations in the study area (Table 6-1). Similar to NCh
20 estimated by LUR, asthma incidence was associated with NCh exposure estimated from
21 passive sampling outside the subjects' homes and from monitoring sites located 1 to 2 km
22 from the home and/or the school.
23 Unlike other LUR studies of asthma, Clark etal. (2010) assigned NCh exposure for the
24 first year of life at the postal code level. A similar effect was estimated for IDW for the
25 first year of life. Because the spatial resolution was at the postal code level, and not the
26 residential level, there may be more uncertainty associated with the exposure assessment
27 (Section 3.4.5.2). Similarly, the use of IDW may also introduce uncertainty due to poor
28 accounting for localized sources between modeling sites (Section 3.2.1.1). In Toronto,
29 individual level residential and school NCh estimated by IDW and LUR were similarly
30 associated with asthma prevalence in a retrospective cross-sectional analysis (Dell et al.,
31 2014). Thus, the representativeness of IDW estimates to the spatial pattern of NCh
32 concentrations may vary across locations.
33 Jerrett et al. (2008) modeled the effects of the within- and between-community variation
34 in NCh in 11 of the 13 CHS communities. This approach allowed examination of the
35 independent contributions of local NCh and regional NCh to the associations with asthma.
36 Both within-community variation and between-community variation in NCh were
37 associated with the development of asthma, providing evidence that both regional and
January 2015 6-19 DRAFT: Do Not Cite or Quote
-------
1 local pollution contributed to the observed associations. One strength of this analysis'
2 exposure assessment is that monitors were placed outside study participants' residences
3 to capture a good representation of the spatial distribution of the true NO2 exposures
4 (Section 3.4.5.2). Moreover, the 2-week measurement periods across cities were
5 synchronized within 4 hours, and the measurement methodology was identical for each
6 city. However, the Palmes tubes used to measure NO2 outside of the residences are
7 subject to positive biases that could negatively bias the effect estimate (Section 3.2.2).
8 Dispersion models have known limitations for accurately estimating within-community
9 conditions, including oversimplification of the NOx reaction model and inaccurate
10 representation of the meteorological conditions, which can add uncertainty to the effect
11 estimate (Section 3.2.1.2). As with IDW, exposure measurement error associated with
12 dispersion model estimates of NO2 or NOx may vary by location. Oftedal et al. (2009a)
13 indicated that dispersion model estimates of NO2 correlated well with central site
14 concentrations (r = 0.76). Specific to NOx, estimates may not represent NO2 exposure
15 equally among subjects. The very high correlations that have been observed between
16 NOx and other traffic-related pollutants such as CO, EC, and PM2 5 (r > 0.94) estimated
17 from dispersion models add uncertainty in attributing associations to NOx specifically.
18 Higher long-term exposure to ambient NO2 is consistently associated with the
19 development of asthma in children as examined in several longitudinal studies. Such
20 associations were found with NO2 assessed from LUR models that were demonstrated to
21 well represent the variability in NO2 in study locations and measurements made outside
22 subjects' homes (Table 6-1). These exposure estimates for subjects' residences provide a
23 strong basis for inferring associations of NO2 with asthma incidence. Associations also
24 were observed with NO2 assessed from central site monitors. Such measurements have
25 well-known limitations in capturing the spatial heterogeneity in ambient NO2
26 concentrations within an area; however, some studies limited monitors to those within
27 1-2 km from children's schools.
28 Another factor influencing inference about the effects of NO2 on asthma development is
29 the ability to disentangle the independent effects of NO2 from the effects of other
30 pollutants in the ambient mixture, particularly those related to traffic. The potential for
31 copollutant confounding in the studies of asthma was informed by examining the
32 correlations with NO2 and results of copollutant models. The available correlations
33 between NO2 and other pollutants for these long-term prospective studies are found in
34 Table 6-1. The level of detail varies from study to study; in some cases, the correlations
35 are not reported or a statement of moderate to high correlation is reported without
36 quantitative results. Specifically, in this evidence base, no data were reported for the
37 correlations between NO2 and ultrafine particles (UFP) or CO. For BC, the data show a
January 2015 6-20 DRAFT: Do Not Cite or Quote
-------
1 correlation of 0.5 in one study. For PIVbs, correlations range from about 0.7 to 0.93. The
2 strong correlations often observed between NCh and PM2 5 make it difficult to interpret
3 the results for NC>2 in the context of the traffic-related pollutant mixture, and introduces
4 uncertainty to the NO2 results.
5 Further, no prospective study evaluated copollutant models for asthma incidence among
6 children. Hwang et al. (2005), in a cross-sectional study of 32,672 school children in
7 Taiwan, observed that the association for NOx measured at monitoring stations within
8 1 km of the schools with physician-diagnosed asthma remained relatively unchanged
9 after including either sulfur dioxide (802), PMio, or Os in a copollutant model.
10 Importantly, this study did not analyze NC>2 or potential confounding by traffic-related
11 copollutants. The strong correlations between NO2 and PM2 5, the limited or no data on
12 correlations with other traffic-related pollutants, and the lack of examination of potential
13 confounding by traffic-related copollutants introduces uncertainty in distinguishing an
14 independent effect of NO2 based on just epidemiologic results.
15 In the groups of studies relating NC>2 exposure to asthma incidence in children, PIVbs was
16 evaluated in several of the studies presented in Table 6-1. McConnell et al. (2010)
17 assigning exposure from a central site observed a smaller effect for PIVbsthan for NC>2
18 with broader CIs; they did not report a correlation between NO2 and PIVb.5. Carlsten et al.
19 (20lie) reported a correlation between NO2 and PM25 of r = 0.7; and NC>2 and BC of
20 r = 0.5. Using LUR exposure estimates, they observed a stronger odds ratio but wider CI
21 for PM2.5 than NC>2, and no risk for BC. Nishimura et al. (2013) did not report
22 correlations between NC>2 and copollutants and observed for IDW exposure estimates a
23 smaller odds ratio for PM2 5 than NC>2, for which the odds ratio was larger and the CI was
24 broader. Gehring etal. (2010) reported a correlation between NC>2 and PM2 5 of 0.93 and,
25 using LUR estimates, observed similar ORs for NCh and PIVb .5. Clark etal. (2010)
26 reported high correlations between NC>2 and the other pollutants but did not provide
27 quantitative data. Based on exposure assessment by LUR and IDW at the postal code
28 level, Clark etal. (2010) observed PM2 5 effect estimates that were smaller than those for
29 NC>2 with broader CIs. However, for BC estimated by LUR, the odds ratio was larger
30 than that for NC>2. Thus, stronger effect estimates with smaller CIs were generally
31 observed for NCh than for PIVb 5.
32 Early-life influences have been implicated in the potential development of asthma as
33 discussed in recent reviews (Kim etal.. 2013; Kudo etal.. 2013). While the first year of
34 life is considered to be an important period for factors affecting asthma development,
35 other periods also play a role especially when considering the multifaceted aspect and the
36 natural history and pathophysiology of asthma. Studies examining the association
37 between long-term exposure to NC>2 and asthma in children vary in exposure period
January 2015 6-21 DRAFT: Do Not Cite or Quote
-------
1 evaluated, age of asthma diagnosis, and length of follow-up time. Several involve birth
2 cohorts followed to an age of 8 to 12 years. Across studies, no single critical time
3 window of NO2 exposure was identified. Associations were observed for NO2 in the birth
4 year (Carlsten et al.. 201 Ic) and lifetime average NO2 (Ranzi et al.. 2014). Other studies
5 found larger magnitudes of association with NO2 exposure in the first 3 years of life
6 (Nishimura et al.. 2013) or year of diagnosis (Clougherty et al.. 2007). Often, the various
7 early life exposure periods that are evaluated are highly correlated with one another,
8 making it difficult to interpret the results or identify a single exposure window of
9 concern. Exposure measurement error also may vary across time periods.
6.2.2.2 Asthma or Chronic Bronchitis in Adults
10 Since the 2008 ISA for Oxides of Nitrogen (U.S. EPA. 2008). recent longitudinal cohort
11 studies have examined asthma/chronic bronchitis in adults in relation to long-term
12 exposure to NO2 and observed positive associations. In the European Community
13 Respiratory Health Survey (ECRHS) cohort, this relationship in adults was examined
14 using various health effects definitions and exposure assessment approaches.
15 The relationship between home outdoor NO2 and chronic bronchitis defined as
16 productive (phlegm) chronic cough more than 3 months each year was analyzed in the
17 ECRHS cohort (Sunyer et al.. 2006). The follow-up time period was 8.9 years. Individual
18 level, indoor kitchen and outdoor (at the kitchen window) residential NO2 was measured
19 using Palmes tubes during a 14-day period in 1,634 households of subjects who did not
20 change residences during the follow-up. This was repeated in 659 households (45%)
21 6 months later. A linear concentration-response relationship with NO2 was observed only
22 in females. NO2 was associated with chronic bronchitis after adjustment for traffic
23 intensity at the residence.
24 In a meta-analysis, Cai etal. (2014) cross-sectionally assessed the associations of outdoor
25 air pollution on the prevalence of chronic bronchitis symptoms in adults in five cohort
26 studies participating in the European Study of Cohorts for Air Pollution Effects
27 (ESCAPE) project. Annual average NO2, NOx, as well as PMio, PM2 5, PMabsorbance,
28 PMcoarse between 2008-2011 were assigned to home addresses by LUR. Symptoms
29 examined were chronic bronchitis (cough and phlegm for >3 months of the year for
30 >2 years), chronic cough (with/without phlegm) and chronic phlegm (with/without
31 cough). Overall, there were no associations with any air pollutant or traffic exposure.
32 New onset asthma was related to NO2 exposure in the ECRHS adult cohort as ascertained
33 by a positive response to the question "Have you ever had asthma?" (Jacquemin et al..
34 2009b) and a continuous asthma score (Jacquemin et al.. 2009a). Asthma incidence was
January 2015 6-22 DRAFT: Do Not Cite or Quote
-------
1 defined as reporting asthma in the follow-up (1999 to 2001) but not at baseline (1991 to
2 1993). Subjects' home addresses were geocoded and linked to outdoor NO2 estimates
3 developed from NOx emissions. For "ever having asthma," the adjusted OR was 1.96
4 (95% CI: 1.04, 3.70) per 10 ppb increase in NO2. The OR for asthma incidence based on
5 the ratio of the mean asthma score was 1.48 (95% CI: 1.18, 1.85) for each increase of
6 10 ppb of NO2.
7 In a preliminary examination of a smaller group of the ECRHS cohort, the prospective
8 Respiratory Health in Northern Europe cohort study, Modig et al. (2009) used dispersion
9 models and two definitions of asthma among 3,824 adults aged 20-44 years at inclusion:
10 (1) the cumulative number of onset cases of asthma and (2) incident cases of asthma.
11 Asthma was defined as no asthma attacks during the last 12 months and no current use of
12 asthma medication in the first survey plus having asthma or ever being diagnosed with
13 asthma at follow-up. NO2 concentrations estimated at the home were associated with risk
14 of developing asthma. The OR was 2.04 (95% CI: 1.14, 3.65) for the cumulative number
15 of onset cases of asthma and 2.25 (95% CI: 1.00, 5.07) for the incident definition of cases
16 per 10-ppb increase in NO2 concentration. The OR for asthma increased across NO2
17 tertiles, indicating a concentration-dependent relationship. With the first tertile as the
18 reference, the OR was higher for the third tertile (ORonset 1.58 [95% CI: 0.96, 2.6];
19 ©Rodent 2.06 [95% CI: 0.98, 4.32]) than for the second tertile (ORonset 1.17 [95% CI:
20 0.70, 1.94]; ORmcident 1.77 [95% CI: 0.86, 3.64]).
21 In 13 European cities, Castro-Giner et al. (2009) prospectively examined asthma
22 incidence and prevalence in the large (2,577 subjects at follow-up) adult
23 population-based ECRHS cohort. In the longitudinal analysis, for the 120 subjects who
24 developed asthma during the follow-up period, NO2 was associated with new-onset
25 asthma with an OR of 2.20 (95% CI: 1.17, 4.10) per 10-ppb increase. For asthma
26 prevalence, an association was indicated among subjects who changed homes rather than
27 subjects who lived in the same home during follow-up (movers OR: 2.53 [95% CI: 1.16,
28 5.56]; nonmovers OR: 1.04 [95% CI: 0.66, 1.64]). However, for new-onset asthma,
29 evidence for association was stronger among nonmovers than movers (nonmovers OR:
30 2.39 [95% CI: 1.10, 5.22]; movers OR: 2.09 [95% CI: 0.70, 6.12]).
31 In summary, among adults, long-term NO2 exposure generally is associated with asthma
32 incidence and chronic bronchitis. The longitudinal design of the studies adds strength to
33 the interpretation of results as do the various approaches to defining asthma. However,
34 the strength of inference is limited because findings are based on one cohort. While the
35 studies aimed to produce more individual measures of exposure at the residence, most
36 exposures in these studies were assessed by dispersion models.
January 2015 6-23 DRAFT: Do Not Cite or Quote
-------
6.2.2.3 Subclinical Effects Underlying Development of Asthma or
Chronic Bronchitis
1 Animal toxicological studies demonstrate that long-term NO2 exposure enhances both
2 responsiveness of airways and the development of allergic responses. Animal
3 toxicological studies and epidemiological studies of long-term exposure show increases
4 in pulmonary inflammation and oxidative stress, providing evidence of NO2-induced
5 airway injury and suggesting a mechanistic basis for the development of asthma in
6 relation to NO2 exposure.
Airway Responsiveness
7 Animal toxicological studies have demonstrated that NO2 exposure enhances
8 responsiveness of airways to nonspecific and specific challenges. A subchronic-duration
9 study demonstrated concentration-dependent increases in airway responsiveness to
10 histamine in NO2-exposed guinea pigs (Kobavashi and Miura. 1995). In this study, one
11 experiment demonstrated AHR after 6 weeks of exposure to 4,000 ppb, but not 60 or
12 500 ppb NO2. In another experiment, AHR was observed in guinea pigs exposed to
13 4,000 ppb NO2 for 6 weeks and to 2,000 ppb for 6 and 12 weeks and to 1,000 ppb for
14 12 weeks. Specific airways resistance in the absence of a challenge agent was increased
15 in guinea pigs exposed to 2,000 and 4,000 ppb NO2 for 12 weeks, which indicates the
16 development of airways obstruction. Another subchronic-duration exposure study found
17 delayed bronchial responses, measured as increased respiration rate, in guinea pigs
18 sensitized and challenged with C. albicans and exposed to NO2 [4,760 ppb, 4 h/day,
19 5 days/week, 6 weeks (Kitabatake et al.. 1995)]. However, NO2 exposure (4,000 ppb,
20 2 h/day, 3 months) failed to alter airway responsiveness to a nonspecific challenge in
21 rabbits sensitized at birth with house dust mite antigen (Douglas et al.. 1995). Overall,
22 results are consistent with those demonstrated in rodent models with short-term
23 exposures to NO2 (Section 4.3.2.5) and are supported by effects demonstrated on key
24 events underlying these responses, including inflammation, allergic sensitization, and
25 airway remodeling (Section 4.3.2).
Development of Allergic Responses
26 Toxicological studies provide some experimental evidence that is coherent with the
27 development of allergic responses seen in some of the epidemiologic studies
28 (Section 6.2.4). One subchronic-duration toxicological study showed that exposure to
29 4,000 ppb NO2 for 12 weeks led to enhanced immunoglobulin E (IgE)-mediated release
30 of histamine from mast cells isolated from guinea pigs (Fujimaki and Nohara. 1994). This
January 2015 6-24 DRAFT: Do Not Cite or Quote
-------
1 response was not found in mast cells from rats similarly exposed in the same study.
2 Furthermore, two short-term studies provide evidence that exposure to NC>2 leads to
3 T-derived lymphocyte helper 2 (Th2) skewing and/or allergic sensitization in healthy
4 adults and naive animal models, as discussed in Sections 4.3.2.6 and 5.2.7.4
5 (Pathmanathan et al., 2003; Ohashi et al., 1994). Findings of increased histamine release
6 from mast cells, increased nasal eosinophils, and increased Th2 cytokines seen in humans
7 and animal models exposed to NO2 provide support for the epidemiologic evidence
8 relating NC>2 exposure to asthma development and the findings in some of the
9 epidemiologic studies for the association of NO2 exposure with the development of
10 allergic responses.
Pulmonary Inflammation and Oxidative Stress
11 Inflammation and oxidative stress are key events in the mode of action for development
12 of asthma (Figure 4-2). Long-term NC>2 exposure has been shown to induce pulmonary
13 inflammation or oxidative stress in toxicological and epidemiologic studies, but results
14 are not entirely consistent. Similarly, there is some evidence for a relationship of
15 short-term exposure to NC>2 with pulmonary inflammation and oxidative stress
16 (Section 5.2.7.4) to describe a potential pathophysiologic basis for development of
17 asthma in response to repeated NO2 exposures.
Epidemiologic Evidence in Children
18 In the CHS cohort of 1,211 schoolchildren from eight Southern California communities,
19 annual average NO2 was associated with a longitudinal increase in exhaled nitric oxide
20 (eNO; using a flow rate of 50 mL/sec) in 2006-2007 and 2007-2008 (Berhane et al..
21 2014). This association was observed with adjustment for short-term NC>2 assessed from
22 central monitoring sites and was independent of asthma status. Based on prior findings in
23 CHS (Bastain et al.. 2011) that elevated eNO is associated with increased risk of new
24 onset asthma, an effect of long-term exposure to NC>2 on increases in eNO over time is
25 consistent with a role for NO2 in asthma pathogenesis.
26 Using LUR models to estimate annual average NO2 exposure, Liu etal. (2014a) observed
27 null associations with eNO among all children (N = 1,985, ages 10 years, ESCAPE) and
28 those without asthma (n = 1,793) in both the single and copollutant (PMio) models. NO2
29 was positively associated with eNO in the 192 children with asthma.
30 Using a cross-sectional prevalence design, Dales et al. (2008) examined the relationship
31 of eNO and NO2 in a cohort of 2,402 healthy school children. NO2 was estimated for
32 each child's residential postal code. Quantitative results were not reported; the study
33 authors only indicate that NO2 showed positive but statistically nonsignificant
January 2015 6-25 DRAFT: Do Not Cite or Quote
-------
1 associations with eNO. An eNO-roadway density association persisted after adjustment
2 for air pollutant concentrations (NO2, SO2, and PM2 5) within the previous 24 and
3 48 hours of the eNO measure, indicating that the association was unlikely to be
4 confounded by an unmeasured short-term exposure effect.
5 The short-term evidence base provides support for the development of asthma in relation
6 to NO2 exposure. Evidence for short-term NO2-associated increases in oxidative stress
7 and pulmonary inflammation, particularly allergic inflammation (Section 5.2.7.4).
8 informs key events within the modes of action for AHR and asthma development
9 (Section 4.3.5 and Figure 4-2).
Toxicological Evidence
10 Similar to studies of short-term NO2 exposure (Section 5.2.4.2). some animal
11 toxicological studies of long-term exposure show increases in pulmonary inflammation
12 and oxidative stress. Compared with short-term exposure studies, long-term studies
13 provide more evidence of NO2-induced pulmonary injury. Details from these studies, all
14 of which were reviewed in the 2008 ISA for Oxides of Nitrogen (U.S. EPA. 2008). are
15 presented in Table 6-2.
16 Many studies investigating NO2-induced injury and oxidative stress in the airway
17 measured changes in lipids, which are necessary for both lung function and defense.
18 Sagai et al. (1982) and Ichinose et al. (1983) reported that rats exposed to 40 or 120 ppb
19 NO2 for 9 or 18 months had increased ethane exhalation and that exposure to 40 ppb for
20 9 months resulted in increased lipid peroxidation. Arner and Rhoades (1973) showed that
21 rats exposed to 2,900 ppb NO2 for 9 months had decreased lipid content leading to
22 increased surface tension of the lung surfactant and altered lung mechanics.
23 Histopathological assessment of lung tissue showed that long-term exposure to NO2
24 resulted in alveolar macrophage (AM) accumulation and areas of hyperinflation (Gregory
25 etal.. 1983). Kumae and Arakawa (2006) exposed rats to 200, 500, or 2,000 ppb NO2
26 from birth or the weanling period (5 weeks old) and assayed bronchoalveolar lavage fluid
27 (BALF) at 8 and 12 weeks of age. Lymphocytes increased at 8 weeks with exposure to
28 500 ppb NO2 in the embryonic group, and macrophages and neutrophils were increased at
29 12 weeks with exposure to 500 ppb NO2. No changes in differential cell counts were
30 observed in the weanling group at 8 weeks of age, but at 12 weeks of age, lymphocytes
31 were increased with exposures above 500 ppb and neutrophils were increased at
32 2,000 ppb. The embryonic group also had increased tumor necrosis factor alpha (TNF-a)
33 and interferon gamma (IFN-y) at 8 weeks but not at 12 weeks, while in the weanling
34 group, IFN^y was increased only at 12 weeks.
January 2015 6-26 DRAFT: Do Not Cite or Quote
-------
Table 6-2 Animal toxicological studies of the respiratory effects of long-term
nitrogen dioxide (NO2) exposure.
Study
Species (Strain);
Age; Sex; n
Exposure Details (Concentration;
Duration)
Endpoints Examined
Arnerand Rhoades Rats (Long Evans); M
(1973)
2,900 ppb 5 days/weeks for 9 mo
Histopathologic
evaluation and
morphometry
Aranvi et al. (1976) Mice
500 ppb continuously, 2,000 ppb
continuously, 100 ppb continuously with
daily 3-h peaks of 1,000 ppb, or 500 ppb
with daily 1-h peaks of 2,000 ppb for 4,
12, 21,24, 28, or 33 weeks
Morphometry
Ayaz and Csallany
(1978)
Mice (C57BL/6J); F;
n = 120
500 ppb or 1,000 ppb continuously for Morphometry
17 mo
Blair et al. (1969) Mice; n = 4/group
500 ppb for 6, 18, or 24 h/day,
7 days/week for 3-12 mo
Histopathologic
evaluation
Chang et al. (1986)
Rat(F344); 1-day or
6 weeks;
M; n = 8/group
(1) 500 ppb continuously with two, daily
1-h spikes of 1,500 ppb, 5 days/week for
6 weeks,
(2) 2,000 ppb continuously for
7 days/week for 6 weeks; Two 1-h
spikes daily to 6,000 ppb (6-week rats
only)
Histopathologic
evaluation and lung
morphometry
Crapo et al. (1984)
Rat (CD, Fischer 344);
6 week; M
2,000 ppb for 23 h/day; two daily 30-min
spikes of 6,000 ppb
Morphometric analysis of
proximal alveolar and
distal alveolar regions
Ehrlich and Henry
(1968)
Mice (Swiss albino); F;
n = >30/group,
n = 4-8/group
(1) 500 ppb continuously,
(2) 500 ppb for 6 h/day,
(3) 500 ppb for 18 h/day;
(1-3) for 1 to 12 mo; challenged with
Klebsiella pneumoniae after exposure
Mortality, hematology,
serum LDH, body weight,
bacterial clearance
Fuiimaki and
Nohara(1994)
Rats (Wistar); 8 weeks;
M; n = 10/group
Guinea pigs (Hartley);
8 weeks; n = 10/group
1,000, 2,000, or 4,000 ppb continuously
for 12 weeks
Mast cell counts and
histamine release
Furiosietal. (1973)
Monkey (Macaca
speciosa), rat
(Sprague-Dawley);
maturing (monkey),
weanling (rat);
M/F (monkey), M (rat);
n = 4-5/group (monkey),
n = 15-25/group (rat)
(1) 2,000 ppb NO2 continuously,
(2) 330 ug/m3 NaCI continuously,
(3) 2,000 ppb NO2 + 330 ug/m3 NaCI
continuously;
(1-3) for 14 mo
Histopathologic
evaluation, hematology
January 2015
6-27
DRAFT: Do Not Cite or Quote
-------
Table 6-2 (Continued): Animal toxicological studies of the respiratory effects of
long-term nitrogen dioxide (NO2) exposure.
Study
Species (Strain);
Age; Sex; n
Exposure Details (Concentration;
Duration)
Endpoints Examined
Greene and
Schneider (1978)
Baboons; 3 to 4 yr; M/F; 2,000 ppb 8 h/day, 5 days/week for 6 mo Immunologic and
n = 6 histopathologic
evaluation
Gregory et al.
(1983)
Rat (Fischer 344);
14-16 weeks;
n = 4-6/group
(1)1,000 ppb,
(2) 5,000 ppb,
(3) 1,000 ppb with two daily, 1.5-h spikes
of 5,000 ppb;
(1-3) 7 h/day for 5 days/week for up to
15 weeks
Histopathological
evaluation, BAL fluid
analysis (LDH, ALKP,
glutathione peroxidase),
antioxidant enzymes in
lung homogenates
Havashi et al.
(1987)
Rat (Wistar); M;
n = 18-160/group
500 ppb or 5,000 ppb continuously for
up to 19 mo
Morphological changes,
histology
Henry et al. (1970) Squirrel monkeys; M;
n = 37
5,000 ppb continuously for 2 mo;
challenge with Klebsiella pneumoniae or
influenza after exposure
Infection resistance,
mortality, peripheral
blood markers, and
respiratory function
Ichinose et al.
Rats (JCL, Wistar);
and 13 weeks; M
(1)10,000 ppb continuously for 2 weeks;
(2)400, 1,200, or 4,000 ppb
continuously for 1, 2, 4, 8, 12, or
16 weeks;
(3) 40, 400, or 4,000 ppb continuously
for 9, 18, or 27 mo
Histopathologic
evaluation and
morphometry
Kumae and
Arakawa (2006)
Rats (Brown -Norway);
prenatal exposure; F;
n=201
200, 500, or 2,000 ppb pre- and
post-natal for up to 12 post-natal weeks
Immunologic evaluation
(alveolar macrophage
activity)
Kubotaetal. (1987) Rat (JCL Wistar);
40, 400, or 4,000 ppb continuously for 9, Serological examination
2 mo; M; n = 3-4/group 18, and 27 mo
Lafuma et al.
(1987)
Mercer et al. (1995)
Hamster (Golden
Syrian); M;
n = 7-9/group
Rats (Fischer 344);
7 weeks; M;
n = 5/group
2,000 ppb NO2 for 8 h/day for
5 days/week for 2 mo
500 ppb continuously with 2 daily, 1-h
peaks of 1 ,500 ppb for 9 weeks
and lung morphometry
Lung histopathology and
morphometry, lung
mechanics, serum
elastase activity, and
protease inhibitor
capacity
Histopathologic
evaluation and
morphometry
Miller etal. (1987)
Mice(CD-l);
4-6 weeks; F;
n = 18-21/treatment
group
(1)200 ppb,
(2) 200 ppb daily continuously for
7 days/week with 2 daily, 1-h peaks of
780 ppb 5 days/week;
(1-2)16, 32, or 52 weeks
Histopathologic
evaluation, pulmonary
function, and
antibacterial host
defenses
January 2015
6-28
DRAFT: Do Not Cite or Quote
-------
Table 6-2 (Continued): Animal toxicological studies of the respiratory effects of
long-term nitrogen dioxide (NO2) exposure.
Study
Species (Strain);
Age; Sex; n
Exposure Details (Concentration;
Duration)
Endpoints Examined
Saqaietal. (1982)
Rats (JCL, Wistar);
8 weeks; M; n =
6-12/group
10,000 ppb continuously for 2 weeks
Antioxidant levels,
enzyme activity, lipid
peroxidation
Saqaietal. (1984)
Rats (JCL Wistar);
8 weeks; M;
n = 4-6/group
40, 400, or 4,000 ppb continuously for 9, Histopathologic
18, or 27 mo evaluation and
morphometry
Sherwin and
Richters(1982)
Stevens et al.
(1988)
Mice (Swiss Webster);
young adults;
M; n = 30/group
Rat (Fischer 344);
young adult, neonate;
M; n = 1 or 6/group
340 ppb for 6 h/day for 5 days/week for
6 weeks
500, 1,000, or 2, 000 ppb continuously
with two daily, 1-h spikes at 1,500,
3,000, or 6,000 ppb for 5 days/week for
6 weeks
Type II pneumocytes in
the lungs and alveolar
wall area
Pulmonary function
Tepperetal. (1993) Rats (Fischer 344);
60 days; M;
n = 11-16/group
500 ppb continuously 7 days/week with
two daily, 2-h spikes of 1,500 ppb,
5 days/week for up to 78 weeks
Pulmonary function and
lung disease
Wagner et al.
(1965)
Dog (Mongrels); M; n =
6-10/group
Rabbit; M; n = 4-8/group
Guinea pig (English); M;
15-31/group
Rat (Sherman); M; n =
20-40/group
Mice (HLA, C57BI/6J,
CAF/Jax); M; n =
60-110/group
1,000 or 5,000 ppb continuously for
10-18 months
Pulmonary function and
histopathology
ALKP = alkaline phosphatase; BAL = bronchoalveolar lavage; LDH = lactate dehydrogenase; NaCI = sodium chloride;
NO2 = nitrogen dioxide.
1
2
3
4
5
6
7
8
9
10
Oxidative stress resulting from NCh exposure has been further characterized in a number
of studies, and the varying effects of NO2 on antioxidant levels and enzyme activity were
presented in the 2008 ISA for Oxides of Nitrogen (U.S. EPA. 2008). After NO2 exposure,
studies have reported both increased and decreased activity of enzymes involved in the
glutathione cycle (Sagaietal. 1984; Gregory et al.. 1983; Ayaz and Csallany. 1978).
Sagai etal. (1984) reported increased nonprotein sulfhydryl levels and glutathione
S-transferase (GST) activity in adult male rats after 9 and 18 months of exposure to
400 ppb NC>2 and decreased glutathione peroxidase (GPx) activity, while
glucose-6-phosphate dehydrogenase activity increased after exposure to 4,000 ppb NCh.
There were no changes in the activity of 6-phosphogluconate dehydrogenase, superoxide
January 2015
6-29
DRAFT: Do Not Cite or Quote
-------
1 dismutase (SOD), or disulfide reductase after exposure to 400 ppb NC>2. Gregory et al.
2 (1983) reported increased glutathione peroxidase activity in BALF after 6 weeks of
3 exposure to 5,000 ppb NCh; however, at 15 weeks, enzyme activity returned to control
4 levels although slight changes in pathology were reported. Ayaz and Csallany (1978)
5 showed that continuous exposure to 1,000 ppb NCh for 17 months decreased GPx activity
6 in Vitamin E-deficient mice, while Vitamin E-supplemented mice had increased
7 glutathione peroxidase activity.
8 These studies demonstrate that long-term NO2 exposure modifies oxidant balance in the
9 airway and can initiate inflammation; however, the observations from these studies at
10 concentrations relevant to ambient exposures do not consistently show this to be the case
11 across species. Antioxidant enzymes are involved in response to NC>2 exposure, but this
12 response is variable and transient. Overall, these findings are similar to the effects
13 reported from short-term exposures (Section 5.2.2.5).
6.2.2.4 Summary of Development of Asthma or Chronic
Bronchitis
14 The recent evidence base adds several longitudinal studies, including prospective studies,
15 that consistently find a positive association of various NO2 exposure measures with
16 asthma incidence in children at several ages. The exposure estimates include residential
17 measurements and individual, residential ambient NO2 concentrations estimated by LUR.
18 Other studies use NO2 measurements at sites 1-2 km from the subjects' school or home.
19 Asthma incidence also was associated with neighborhood-level ambient NO2
20 concentrations estimated by IDW or NO2 or NOx estimated for residential locations by
21 dispersion models, which are associated with greater uncertainty in representing the
22 spatial variability in ambient NO2 concentrations. In adults, positive associations are also
23 observed; however, this evidence base is limited primarily to one adult cohort in Europe,
24 and exposure measures are from dispersion models. Overall, there is a consistency of
25 association that is observed across exposure assessment methods and ages of children
26 examined from 10 to 18 years. None of the studies examined whether there was evidence
27 for an association of NO2 with health effects independent from other traffic-related
28 pollutants, such as UFP, CO, BC/EC, and PM25.
29 Toxicological and controlled human exposure studies reduce some of the uncertainty in
30 the epidemiologic evidence by providing biological plausibility for a relationship
31 between long-term NO2 exposure and asthma development. In the pathophysiology of
32 asthma, recurrent pulmonary inflammation, allergic sensitization, and subsequent
33 development of AHR play important roles (Section 4.3.5. Figure 4-2).
January 2015 6-30 DRAFT: Do Not Cite or Quote
-------
1 Long-term-exposure toxicological studies demonstrate NCh-induced AHR, and
2 experimental studies of repeated short-term exposures provide evidence for NO2-induced
3 development of allergic responses in healthy adults and animal models as well as
4 increases in neutrophils in healthy adults. In one study of guinea pigs, NC>2-induced
5 (1,000-4,000 ppb) increases in AHR was accompanied by an increase in specific airways
6 resistance, suggesting that airway remodeling may contribute to the development of AHR
7 [(Kobayashi and Miura. 1995); Section 4.3.2.51. Mechanistic studies indicate that
8 inflammatory mediators and structural changes occurring due to airway remodeling can
9 alter the contractility of airway smooth muscle. Epidemiologic evidence points to
10 associations between short-term increases in ambient NO2 concentrations and increases in
11 pulmonary inflammation in healthy children and adults (Section 5.2.2.5). An
12 epidemiologic study also indicates an association of long-term NC>2 exposure with
13 longitudinal changes in pulmonary inflammation in healthy children that were
14 independent of short-term changes in NC>2 concentrations. There also is some evidence
15 for pulmonary oxidative stress induced by short-term NCh exposure in healthy adults
16 (Section 5.2.7.4) and long-term exposure in rodents (Section 6.2.2.3). although results
17 overall are not consistent. The positive relationship between NO2 exposures and asthma
18 in longitudinal epidemiological studies and the small body of evidence indicating NC>2
19 effects on inflammation, allergic sensitization, and AHR, which are key events in the
20 mode of action for the development of asthma, support an independent role for NCh
21 exposure in development of asthma.
6.2.3 Severity of Asthma, Chronic Bronchitis, and Chronic Obstructive
Pulmonary Disease: Respiratory Symptoms and Hospital Admissions
22 In the 2008 ISA for Oxides of Nitrogen (U.S. EPA. 2008). there was limited evidence,
23 consisting of a single prospective cohort study and several cross-sectional studies, to
24 support an association between long-term exposure to NC>2 and respiratory symptoms.
25 Evidence was inconsistent, with uncertainties in the cross-sectional studies related to the
26 temporality of exposure and occurrence of symptoms. Key recent prospective cohort
27 studies evaluating the relationship between respiratory symptoms in children and
28 long-term exposure to NC>2 are summarized in Table 6-3. Studies that evaluated indoor
29 NC>2 concentrations are discussed first, followed by studies related to outdoor NO2
30 concentrations. Cross-sectional studies generally are consistent with longitudinal studies
31 (Annesi-Maesano et al.. 2012; Ghosh etal.. 2012b: Dong etal.. 2011; Mi et al.. 2006;
32 Pattenden et al.. 2006; Nicolai etal.. 2003; Brauer etal.. 2002; Gehring et al.. 2002;
33 Zemp et al.. 1999) and were reviewed and are discussed in other sections as appropriate
34 and summarized in the Annex Table AX6.3-17 of the 2008 ISA for Oxides of Nitrogen
35 (U.S. EPA. 2008).
January 2015 6-31 DRAFT: Do Not Cite or Quote
-------
6.2.3.1 Indoor Nitrogen Dioxide and Respiratory Symptoms in
Children and Adults
1 Effects of indoor NC>2 may not be confounded by all of the same copollutants as outdoor
2 NO2, although potential confounding by other indoor pollutants, such as emissions from
3 heating sources could occur. For long-term NC>2 exposures, the recent indoor prospective
4 study of school-aged children (Belanger et al.. 2013) and the adult indoor prospective
5 study (Hansel etal., 2013) provide evidence that supports a relationship between
6 long-term NC>2 exposure and respiratory symptoms in children with asthma and former
7 smokers with COPD.
8 Belanger et al. (2013) observed positive associations of asthma severity score, wheeze,
9 nighttime symptoms, and rescue medication use with indoor residential NCh where the
10 mean monitoring length was 33 [standard deviation (SD): 7] days. Figure 6-3 illustrates
11 the concentration-response relationships between indoor NC>2 and asthma-related effects
12 using a constrained, natural spline function of ln(NO2) and 95% confidence limits as well
13 as threshold functions for each outcome. In adjusted models with quintiles of NC>2
14 exposure, concentrations >14.3 ppb compared with the reference level (<6 ppb,
15 designated as the threshold value) was associated with increased risk of a one-level
16 increase in asthma severity score (OR: 1.43 [95% CI: 1.08, 1.88]). These same exposures
17 were also associated with increased risks of wheeze (OR: 1.53 [95% CI: 1.16, 2.02]),
18 night symptoms (OR: 1.59 [95% CI: 1.24, 2.01]), and rescue medication use (OR: 1.74
19 [95% CI: 1.34, 2.26]). Every fivefold increase in NO2 exposure >6 ppb was associated
20 with an increase in asthma severity score (OR: 1.37 [95% CI: 1.01, 1.89]) and asthma
21 morbidity measured by wheeze (OR: 1.49 [95% CI: 1.09, 2.03]), night symptoms (OR:
22 1.52 [95% CI: 1.16, 2.00]), and rescue medication use (OR: 1.78 [95% CI: 1.33, 2.38]).
23 Recent infant studies are consistent with earlier results (Samet etal., 1993) that showed
24 no association between 2-week avg exposure to NO2 and the incidence and duration of
25 respiratory illness. Raaschou-Nielsen et al. (201 Ob) and Sonnenschein-Van der Voort
26 et al. (2012) found no associations between indoor NO2 exposure and wheezing in
27 infants.
January 2015 6-32 DRAFT: Do Not Cite or Quote
-------
Table 6-3 Prospective studies of long-term nitrogen dioxide exposure and respiratory symptoms in children.
Study3
Exposure
Assessment
Pollutant
Correlation
Statistical Methods
Comments
Results
95% Clb
Longitudinal New England Indoor Children's Asthma Study
Belanqeret al. (2013)
N = 1,642
Ages 5-10 yr
Followed for 1 yr.
Asthma severity score from
2006 through 2009. Symptoms:
wheeze, nighttime symptoms,
rescue medication use and an
asthma severity score [which
consist of symptoms and
medication use based on the
Global Initiative for Asthma
(NHLBI. 2002)1.
Palmes tubes in
bedrooms and
dayroom for 4 weeks
for 4 seasons.
NR
Hierarchical ordered logistic
regression adjusted for age, sex,
atopy, season of monitoring,
race/ethnicity, mother's
education, smoking in the home,
and all five variables for
combined specific sensitization
and exposure to indoor allergens.
Analyses were based on
repeated measures of both NO2
and asthma outcomes.
Included maintenance
medication use as a
covariate in models
because the use of
maintenance medication is
also associated with SES.
OR per fivefold
increase in NO2
exposure above
6 ppb
Asthma severity
score: 1.37
(1.01, 1.89)
Wheeze: 1.49
(1.09,2.03)
Night symptoms:
1.52(1.16,2.00)
Rescue
medication use:
1.78(1.33,2.38)
January 2015
6-33
DRAFT: Do Not Cite or Quote
-------
Table 6-3 (Continued): Prospective studies of long-term nitrogen dioxide exposure and respiratory symptoms in
children.
Study3
Exposure
Assessment
Pollutant
Correlation
Statistical Methods
Comments
Results
95% Clb
Childrens Health Study (CHS) Southern California Communities
McConnell et al. (2003)
N=475
Fourth graders (aged 9-10) and
seventh graders (aged 12-13)
with a history of asthma or
bronchitic symptoms at study
entry who completed two or
more follow-up questionnaires
any time during the yr 1996 to
1999. Yearly follow-up for 4 yr.
The overall participation rate in
the surveyed classrooms was
high (82%).
Annual average NO2
calculated from
monitoring sites
established in each of
the 12 communities.
4-yr mean (SD;
1996-1999) NO2
concentrations across
communities: 19.4
(11.3) ppb.
Within-community
correlations differed
in that NO2 could be
distinguished from
most other major
pollutants except
OC and inorganic
acid.
O3: 0.59
PMio: 0.20
PlVh.s: 0.54
PMio-2.5: -0.22
I ACID: 0.65
OACID: 0.48
EC: 0.54
OC: 0.67
Three-stage regression adjusted
for age, maternal smoking
history, child's sex maternal and
child's race; within-community
estimates were adjusted for
between-community effects of
the pollutant, and vice versa.
Copollutant model results:
Within communities
NO2With PIvh.s: 1.05 NS
NO2with I ACID: 1.09C
NO2 with OACID: 1.07d
NO2with EC: 1.05d
NO2withOC: 1.04NS
NO2withO3: 1.06NS
NO2with PMio: 1.07d
NO2with PMio-2.s: 1.08C
Between communities
NO2with PlVh.s: 1.01 NS
NO2with I ACID: 1.02NS
NO2 with OACID: 1.02NS
NO2with EC: 1.01 NS
NO2with OC: 1.01 NS
NO2withO3: 1.02d
NO2with PMio: 1.01 NS
NO2with PMio-2s.: 1.02d
OR within
communities
1.97(1.22,3.18)
OR between
communities
1.22(1.00, 1.49)
January 2015
6-34
DRAFT: Do Not Cite or Quote
-------
Table 6-3 (Continued): Prospective studies of long-term nitrogen dioxide exposure and respiratory symptoms in
children.
Study3
Exposure
Assessment
Pollutant
Correlation
Statistical Methods
Comments
Results
95% Clb
Prevention and Incidence of Asthma and Mite Allergy (PIAMA) study, the Netherlands
Gehrinq et al. (2010)
N = 3,863 children in birth
cohort aged 1 to 8 yr
LUR models used to
provide annual NO2
concentrations for
birth address of each
participant.
NO2-PM2.5: 0.93
NO2-soot: 0.96
Generalized estimating equations
adjusted for sex, study arm
(intervention or natural history),
use of mite-impermeable
mattress covers, allergies of
mother and father, maternal and
paternal education, maternal
prenatal smoking, breastfeeding,
presence of a gas stove in the
child's home, presence of older
siblings, smoking, signs of
dampness and pets in the child's
home, day-care attendance, and
Dutch nationality.
No associations were
found with atopic eczema,
allergic sensitization, and
bronchial
hyperresponsiveness.
No copollutant analyses
were conducted.
OR for asthma
symptoms:
1.17(0.98, 1.39)
without
adjustment for
study region.
OR for wheeze:
1.27(1.07, 1.50)
without
adjustment for
study region.
January 2015
6-35
DRAFT: Do Not Cite or Quote
-------
Table 6-3 (Continued): Prospective studies of long-term nitrogen dioxide exposure and respiratory symptoms in
children.
Study3
Exposure
Assessment
Pollutant
Correlation
Statistical Methods
Comments
Results
95% Clb
Children, Allergy, Milieu, Stockholm, Epidemiology Survey (BAMSE)
Gruzieva et al. (2013)
N = 3,633
Followed from birth
(1994-1996) to age 12yr.
Related publications:
Melen et al. (2008)
Nordlinq et al. (2008)
Dispersion models NOx-PM-io for first
used to calculate NOx yr of life: 0.96
for all addresses in
the yr 1994 to 2008,
representing when the
first child was born
until the end of the
12-yr follow-up.
Multinomial regression/GEE
adjusted for municipality, SES, yr
the house was built, and heredity.
OR for wheeze
at 12 yr of age, 3
or more
episodes: 1.35
(0.79, 2.29) per
20 ppb NOx.
No association of
NOx after the
first yr of life with
asthma
symptoms.
BAMSE = Children, Allergy, Milieu, Stockholm, Epidemiology Survey; CHS = Children's Health Study; Cl = confidence interval; EC = elemental carbon; GEE = generalized estimating
equations; I ACID = inorganic acid; LUR = land-use regression; NO2 = nitrogen dioxide; NOX = sum of NO and NO2; NS = not statistically significant; O ACID = organic acid;
OC = organic carbon; OR = odds ratio; PIAMA = Prevention and Incidence of Asthma and Mite Allergy; PM2.s = particulate matter with a nominal aerodynamic diameter less than or
equal to 2.5 |jm; PMio = particulate matter with a nominal aerodynamic diameter less than or equal to 10 |jm; PMi0-2.5 = particulate matter with a nominal aerodynamic diameter less
than or equal to 10 |jm and greater than a nominal 2.5 |jm; SD = standard deviation; SES = socioeconomic status.
aStudies are presented in the order of appearance in the text.
""Results are presented per 10-ppb increase in NO2 unless otherwise specified.
°p < 0.05.
dp<0.01.
January 2015
6-36
DRAFT: Do Not Cite or Quote
-------
A Asthma Severity Score
B wheeze
128
128
0.3
0.3
NO2 (ppb, log scale)
C Night Symptoms
NO2 (ppb, log scale)
D Rescue Medication Use
0.3 J
NO2 (ppb, log scale)
4 8 IS 32
NO2 (ppb, log scale)
64 128
Note: Also shown is a histogram of NO2 concentrations measured in subjects' homes (lower portion of panel D) for all observations
(thin border) and observations taken in homes of gas stove users (bold border). Indoor NO2 was modeled as a continuous variable
of log-transformed concentrations.
Source: Reprinted with permission of Wolters Kluwer Health, Belanger et al. (2013).
Figure 6-3 Concentration-response relationships between respiratory effects
and indoor nitrogen dioxide (NO2) illustrated with constrained,
natural spline functions (solid lines) with 95% confidence limits
(small dashed lines) and threshold function (bold dashed line)
from hierarchical ordered logistic regression models.
i
2
o
3
4
Hansel et al. (2013) investigated indoor NCh and PIVb 5 concentrations in relation to
respiratory effects among former smokers with COPD in Baltimore, MD. Air sampling
was performed for 1 week at baseline, 3 months, and 6 months in the participant's
bedroom and the main living area, which was identified as an additional room where the
January 2015
6-37
DRAFT: Do Not Cite or Quote
-------
1 participant reported spending the most time. Clinic visits occurred at baseline, 3 months,
2 and 6 months. No interaction was indicated between PM2 5 and NO2, and increasing NO2
3 concentrations in the main living area were independently associated with increased
4 dyspnea and increased rescue medication use with adjustment for PIVbs. Higher bedroom
5 NO2 concentrations were associated with increased risk of nocturnal awakenings (OR:
6 1.59 [95% CI: 1.05, 2.42] per 10-ppb increase) and severe exacerbations (OR: 1.65 [95%
7 CI: 1.02, 2.64]). NO2 concentrations were not associated with lung function. There was
8 indication of outdoor NO2 concentrations contributing to indoor NO2. Among the
9 26 subjects who lived within 4.8 km of a central monitoring site, outdoor NO2
10 concentrations explained 25% of the variance in indoor NO2 concentrations.
6.2.3.2 Outdoor Nitrogen Dioxide and Respiratory Symptoms in
Children
11 A number of studies (Table 6-3) have observed an association between various
12 respiratory symptoms in children and long-term exposure to outdoor NO2. McConnell
13 et al. (2003) examined children with asthma for bronchitic symptoms, including daily
14 cough for 3 months in a row, congestion or phlegm 3 months in a row, or bronchitis.
15 Thus, while these symptoms may have started with acute exacerbation of asthma, they
16 were likely to represent chronic indolent symptoms. In copollutant models, the effects of
17 yearly variation in NO2, ascertained from a central monitoring site in each community,
18 were only modestly reduced by adjusting for another traffic-related copollutant such as
19 PM2 5, EC, or organic carbon (OC; Figure 6-4 and Table 6-3).
January 2015 6-38 DRAFT: Do Not Cite or Quote
-------
Q.
°-
re
(Ł
C/5
T3
73
O
1.5 -i
1.3
1.2
1.1 -
1.0
0.9
Risk of Bronchitic Symptoms as a Function
of Yearly Deviation in NO2
I ? J
0Ł '" <^ \^ O^
Adjustment Air Pollutants
Source: McConnell et al. (2003). Reprinted with permission of the American Thoracic Society. Copyright © 2014 American Thoracic
Society. American Journal of Respiratory and Critical Care Medicine, 168(7): 790-797. Official Journal of the American Thoracic
Society.
Figure 6-4 Odds ratios for within-community bronchitis symptoms
associated with nitrogen dioxide (NO2), adjusted for other
pollutants in copollutant models for the 12 communities of the
Children's Health Study.
i
2
3
4
5
6
7
8
9
10
11
12
13
Gehring etal. (2010) examined a composite of asthma symptoms (one or more attacks of
wheeze, shortness of breath, prescription of inhalation steroids), wheeze (transient, late
onset, persistent), sneezing, hay fever, atopic eczema, and prevalence of asthma and
observed positive associations with LUR modeled NO2 exposures. Gruzieva et al. (2013)
examined wheeze, categorized as either one or more episodes or three or more episodes
in the past year and observed an association with NOx concentrations from a dispersion
model. Aguileraet al. (2013) observed an association between NO2 exposure and
increased risk of upper and lower respiratory tract infections in infants.
Cross-sectional studies provide information that informs the potential for copollutant
confounding of associations found with various NC>2 measures. Hwang and Lee (2010)
observed positive associations between NO2 and bronchitic symptoms in children with
asthma in copollutant models with PIVbs generally similar to what was observed for
single pollutant models (ORs for NO2 adjusted for PM2 5: 2.25 [95% CI: 1.17, 4.33] for
January 2015
6-39
DRAFT: Do Not Cite or Quote
-------
1 bronchitis; 1.60 [95% CI: 0.76, 3.34] for chronic phlegm; 1.30 [95% CI: 0.53, 3.12] for
2 chronic cough; 2.21 [95% CI: 1.23, 3.97] for bronchitic symptoms per 10-ppb increase in
3 NO2).
Outdoor Nitrogen Dioxide and Respiratory Symptoms in Adults with
Asthma
4 Studies examining the relationship between long-term NO2 exposure and respiratory
5 symptoms in adults include prospective studies of asthma incidence in adults discussed in
6 Section 6.2.2.2. Most studies assessed NO2 exposure from dispersion models. Jacquemin
7 et al. (2009b) report that all the associations between NO2 and asthma symptoms at
8 ECHRS II were positive. The strongest was for waking "with a feeling of tightness in the
9 last 12 months." Symptoms in the last 12 months at ECRHS II among people without
10 asthma at baseline were also associated with NC>2.
11 Zemp et al. (1999) report, in a cross-sectional study, the Swiss Study on Air Pollution and
12 Lung Disease in Adults (SAPALDIA), an association between NCh and prevalence of
13 respiratory symptoms in adults. Bentayeb et al. (2010) report cross-sectional associations
14 to be weakly positive for cough and phlegm in adults (>65 years old, in Bordeaux,
15 France) in relation to NCh exposure.
Outdoor Nitrogen Dioxide and Asthma Hospital Admissions in Adults
16 Recent studies represent the first evaluation of the association between long-term NCh
17 exposure and hospital admissions for asthma. The relationship between long-term
18 pollutant exposures on the risk for asthma hospital admissions [International
19 Classification of Diseases (ICD)-IO: J45-46] in people aged 50-65 years at baseline was
20 evaluated in the Danish Diet, Cancer and Health cohort study (Andersen et al.. 2012a).
21 Associations between NO2 concentration estimated by the Danish Air geographic
22 information system (GIS) dispersion modelling system and hospital admission were
23 found in the full cohort [hazard ratio (HR) per-10 ppb NO2: 1.44 [95% CI: 1.14, 1.84]).
24 NO2 was estimated to have a similar effect on the first asthma hospital admission (HR:
25 1.36 [95% CI: 1.03, 1.80]), but people with a previous asthma hospital admission were at
26 greater risk for re-admission (HR: 3.05 [95% CI: 1.57, 5.90]). NO2 was associated with a
27 much larger risk of asthma hospital admission among people with previous admission for
28 COPD (HR: 2.34 [95% CI: 1.25, 4.40]). Some of the observed effects could possibly be
29 ascribed to the short-term effects of increases in air pollution on the days prior to asthma
30 admission. The 35-yr avg, 15-yr avg, and 1-yr avg NCh at follow-up were highly
31 correlated (r = 0.88, 0.92) and were more strongly associated with asthma hospital
32 admission than was 1-yr avg NCh at baseline. The authors indicated that they could not
January 2015 6-40 DRAFT: Do Not Cite or Quote
-------
1 discern whether the results reflected the importance of more recent exposures or the
2 better performance of the dispersion model in more recent years. The NC>2 exposure
3 estimates may be less certain for earlier time periods because of uncertainty in emission
4 factors and traffic counts that are used as inputs to the dispersion model.
5 An ecological time-series study, (Delamater et al.. 2012) and a cross-sectional study,
6 (Meng et al.. 2010) provide inconsistent results in regards to asthma-related emergency
7 department (ED) visits or hospital admissions. Meng etal. (2010) examined subjects ages
8 1 to 65+ years who reported physician-diagnosed asthma (N = 1,502) in the San Joaquin
9 Valley, CA from among participants of the 2001 California Health Interview Survey.
10 Subjects were assigned annual average concentrations for NO2 based on residential ZIP
11 code and the closest air monitoring station within 8 km, but data on duration of residence
12 were not available. No associations were found between average annual concentrations of
13 NC>2 and the odds of asthma-related ED visits or hospital admissions. No quantitative
14 results were shown for NC>2.
6.2.3.3 Summary of Severity of Asthma, Chronic Bronchitis, and
Chronic Obstructive Pulmonary Disease
15 Longitudinal studies conducted in children observed associations between long-term
16 ambient NC>2 exposure metrics and an array of respiratory symptoms in school-age
17 children. Results in infants were inconsistent, but transient symptoms are common in
18 infants, but these symptoms may not have strong implications for developing respiratory
19 disease. For children, ambient NC>2 exposure was assessed with outdoor residential
20 measurements, LURthat estimated exposure at subjects' homes, and central site
21 measurements. The McConnell et al. (2003) study is unique in that it is the only
22 prospective study examining bronchitic symptoms in children with asthma. The study
23 authors report stronger associations for NC>2 variation within communities
24 (within-community associations cannot be confounded by any time-fixed personal
25 covariates) than for NCh variation between communities.
26 Further supporting a relationship with NC>2, indoor NC>2 was associated with asthma
27 symptoms and medication use in children with asthma (Belanger et al.. 2013) and
28 respiratory symptoms in former smokers with COPD (Hansel et al.. 2013). The effect
29 estimates for indoor NC>2 were generally larger than those reported in the studies of
30 outdoor NC>2, and Belanger et al. (2013) provided evidence for a concentration-dependent
31 increase in NC>2-related symptoms. These indoor NC>2 exposures may be part of a
32 different mix of air pollutants than in the ambient air and support an independent effect of
33 NO2.
January 2015 6-41 DRAFT: Do Not Cite or Quote
-------
1 An uncertainty in the evidence base is the potential influence of short-term NC>2
2 exposure. While many studies aimed to characterize chronic symptoms, they did not
3 examine whether associations were independent of short-term NC>2 exposure. Another
4 uncertainty is the potential confounding by other traffic-related pollutants. The available
5 correlations between NC>2 and other pollutants for these respiratory-symptom, long-term
6 prospective studies are found in Table 6-3. Specifically, no data for CO are available. The
7 data show a correlation of 0.54 for ambient EC and 0.96 for soot. For PM2 5, the
8 correlations range from about 0.54 to 0.93 (Gehring et al.. 2010; McConnell et al.. 2003).
9 In limited analysis, NO2 associations with symptoms persisted with adjustment for EC or
10 PM2 5 as measured at central sites and was somewhat attenuated with adjustment for OC.
11 The collective evidence from this group of prospective studies is supportive of a
12 relationship of long-term exposure to NO2 and increased respiratory symptoms using
13 various indicators in children with asthma, but evidence identifying an independent
14 association of long-term NO2 exposure is limited.
6.2.4 Development of Allergic Disease
6.2.4.1 Epidemiologic Studies of Children or Adults
15 Recent cross-sectional studies report results for various aspects of allergic responses and
16 long-term exposure to NC>2. Allergic sensitization indicators included measures of IgE,
17 allergic rhinitis, skin prick test, and reporting respiratory allergy/hay fever. Various age
18 groups were examined, including children less than 6 years old, children aged about
19 10 years, and adults. As described in Section 6.2.2.3. the few available experimental
20 studies provide support for an effect of long-term or repeated short-term NC>2 exposure
21 on development of allergic responses.
22 In a nationally representative sample of the U.S. population, Weir etal. (2013) linked
23 annual average concentrations of NC>2 to allergen-specific IgE data for participants
24 6 years old and older in the 2005-2006 National Health and Nutrition Examination
25 Survey using both monitor-based (within 32.2 km) air pollution estimates and the
26 Community Multiscale Air Quality model (36 km) and observed that increased
27 concentrations of NC>2 were associated with positive IgE to any allergen, inhalant, and
28 indoor allergens.
29 In the German Infant Nutritional Intervention (GINI) and Lifestyle-Related Factors on
30 the Immune System and the Development of Allergies in Childhood (LISA) cohorts,
31 analysis of individual-based exposure to NCh derived from LUR and allergic disease
32 outcomes during the first 6 years of life (Morgenstern et al.. 2008) indicated associations
January 2015 6-42 DRAFT: Do Not Cite or Quote
-------
1 with eczema. Some associations with allergen-specific IgE and hay fever were positive
2 but imprecise with wide 95% CIs. Previous analyses of these cohorts did not indicate
3 associations with runny nose and sneezing at age 2 years (Morgenstern et al., 2007;
4 Gehring et al.. 2002). A longitudinal study of the LISA and GINI cohorts (Tuertes et al..
5 2013) found no evidence that NC>2 exposure in the birth year increases the prevalence of
6 allergic rhinitis or increases risk of aeroallergen sensitization as determined by
7 allergen-specific IgE in children examined at age 10 years. Air pollution concentrations
8 decreased in the study areas during this time.
9 Annesi-Maesano et al. (2007) related individual data on allergy from
10 5,338 schoolchildren (ages 10.4 ± 0.7 years) attending 108 randomly chosen schools in
11 six French cities to the concentration of NC>2 measured in school yards with passive
12 diffusion samplers and at fixed-site monitoring stations. In examining associations of a
13 5-day avg NCh concentration with lifetime prevalence of allergic conditions, the authors
14 used the short-term exposure metric to represent long-term exposure NC>2. NCh was
15 positively associated with flexural dermatitis and skin prick test to indoor allergens but
16 not with allergic rhinitis or atopic dermatitis. In a large cross-sectional study of school
17 children in Taiwan, Hwang et al. (2006) observed that a 10-ppb increase in NC>2 was
18 associated with a higher prevalence of allergic rhinitis, with an OR of 1.11 (95% CI:
19 1.08, 1.15). Parker et al. (2009) evaluated the association between ambient pollution
20 monitoring data and childhood respiratory allergies in the U.S. using the 1999-2005
21 National Health Interview Survey of approximately 70,000 children and observed no
22 associations between NC>2 and the reporting of respiratory allergy/hay fever.
23 As part of the same study of schools in six French cites, Annesi-Maesano et al. (2012)
24 evaluated the relationship between indoor air quality in schools and the previous-year
25 allergic and respiratory health of schoolchildren (mean age 10.4 years). For each
26 pollutant, a 5-day mean concentration in the classroom was computed and categorized
27 into tertiles, independent of the city (low <9.7 ppb, medium >9.7 to <12.9 ppb, high
28 >12.9 ppb NO2). Between-school and within-school variability of the measured indoor
29 pollutants were estimated using linear mixed models for longitudinal data. Among
30 children with atopy (N = 1,719), high NCh was related to previous-year allergic asthma
31 but not allergic rhinitis.
32 Nordling et al. (2008) reported that exposure to dispersion-modeled NOx from traffic
33 during the first year of life was associated with sensitization (measured as specific IgE) to
34 inhalant allergens, especially pollen (OR: 1.24 [95% CI: 1.04, 1.49] per 10-ppb increase
35 in NO2). The relationship between the development of allergic sensitization in children
36 during the first 8 years of life and long-term exposure to NOx was evaluated in a
37 prospective analysis of the Children, Allergy, Milieu, Stockholm, Epidemiology Survey
January 2015 6-43 DRAFT: Do Not Cite or Quote
-------
1 (BAMSE) cohort (Gruzieva et al., 2012). There was no overall risk of sensitization at
2 4 years of age associated with NOx assessed from a dispersion model.
3 In a cross-sectional analysis of 30,139 Chinese children aged 3 to 12 years, Dong et al.
4 (2011) evaluated the relationship between 3-yr avg of NCh and allergy rhinitis. Among
5 children without allergic predisposition (N = 26,004), several positive associations with
6 NC>2 were observed, mainly in males. Among children with an allergic predisposition,
7 associations were detected in males and females.
8 Pujades-Rodriguez et al. (2009) examined a cohort of 2,644 adults aged 18-70 years
9 living in Nottingham, U.K. to evaluate the relationship between NC>2 exposure and
10 allergy-related effects. In cross-sectional analyses, they found generally null associations
11 between NCh concentration and skin test positivity, total IgE, and questionnaire-reported
12 eczema or hay fever. Total IgE levels were not related to NO2 concentrations in
13 369 adults with asthma in five French centers as part of the Epidemiological Study on the
14 Genetics and Environment of Asthma (Rage et al.. 2009) but were related to Os
15 concentrations.
6.2.4.2 Summary of Development of Allergic Responses
16 As described in Section 6.2.2.3. a few available experimental studies demonstrate the
17 effects of repeated short- or long-term NO2 exposure on development of an allergic
18 phenotype in healthy adults and animal models. These findings not only suggest the
19 possibility that chronic or recurrent exposure to NCh may lead to the development of
20 asthma but also support a role for NC>2 exposure in the development of allergic
21 conditions.
22 Long-term NCh exposure has been linked to various indicators of allergic sensitization in
23 a few cross-sectional studies of children, but not consistently for outcomes such as
24 allergic rhinitis or hay fever. In children 6 years and younger and for adults, the various
25 allergic indicators are not related to NC>2 exposure. For the age group of children about
26 10 years old, NCh was related to allergic sensitization as assessed by allergen-specific
27 IgE or skin prick test in cross-sectional studies but not the longitudinal study. NC>2
28 metrics aimed at characterizing individual exposures, such as 5-day measurements in
29 school yards and residential estimates from LUR, produced inconsistent results.
January 2015 6-44 DRAFT: Do Not Cite or Quote
-------
6.2.5 Lung Function and Lung Development
1 The 2008 ISA for Oxides of Nitrogen (U.S. EPA. 2008) characterized long-term
2 prospective studies as showing a relationship between NO2 concentrations and
3 decrements in lung function and reduction in lung development in children. A key
4 uncertainty associated with these studies was the high correlation of NO2 concentrations
5 with other ambient pollutants. Recent prospective cohort studies add to the evidence base
6 that evaluates the relationship between supervised lung function tests and long-term NO2
7 exposure.
6.2.5.1 Lung Function and Development in Children
8 The key longitudinal prospective studies summarized in Table 6-4 and Figure 6-7
9 continue to show a relationship between long-term NC>2 exposure and decrements in lung
10 function, especially as children reach later ages. Lung function continues to increase
11 through early adulthood with growth and development, then declines with aging
12 (Stanojevicetal.. 2008: Zeman and Bennett. 2006: Thurlbeck. 1982). Thus, the
13 relationship between long-term NC>2 exposure and decreased lung function over time in
14 school-age children into early adulthood is an indicator of decreased lung development.
15 The CHS has examined three separate cohorts for pollutant effects on lung function and
16 development [1993 cohort in Gauderman et al. (2004). 1993 and 1996 cohorts in Breton
17 etal. (2011) and Gauderman et al. (2007). and 2002 cohort in Urman etal. (2014)1. The
18 results of Breton et al. (2011) are consistent with earlier results from Gauderman et al.
19 (2004). Both Gauderman et al. (2007) and Urman etal. (2014) assessed copollutant
20 models that included another regional or near-roadway pollutant, and Urman etal. (2014)
21 examined the joint effect of NOx and PlVfc 5.
January 2015 6-45 DRAFT: Do Not Cite or Quote
-------
Table 6-4 Prospective studies of long-term nitogen dioxide exposure and lung function and lung development
in children.
Study3
Exposure Assessment Pollutant Correlation Statistical Methods
Comments
Results (95% Cl)b
FEVi and FVC
(mL)
MMEF, PEF, and
FEF (mL/sec)
Children's Health Study (CHS) California Communities
Gauderman et al. (2004)
N = 1,759
Ages 10-18 yr from 1993 cohort
8-yr follow-up.
Central monitoring station
in each of 12 study
communities, beginning in
1994.
Average hourly
concentrations of NO2
used to compute annual
averages. Then, calculated
long-term mean pollutant
concentrations (from 1994
through 2000).
NO2 range across
communities: 34.6 ppb
NO2-acid vapor: 0.87
NO2-PMio: 0.67
NCb-PIVh.s: 0.79
NO2-EC: 0.94
NO2-OC: 0.64
2 stage linear
regression adjusted
for log values for
height, BMI, BMI
squared, race,
Hispanic ethnic
background,
doctor-diagnosed
asthma, any tobacco
smoking by the child in
the preceding yr,
exposure to
environmental tobacco
smoke, and exercise;
or respiratory tract
illness on the day of
the test and indicator
variables for the field
technician and the
spirometer.
Children who moved away
from their recruitment
community were classified
as lost to follow-up and not
tested further. The number
of children available for
follow-up was 1,414 in
1995, 1,252 in 1997, 1,031
in 1999, and 747 in 2001,
reflecting the attrition of
approximately 10% of
subjects peryr. Model fit
was not better in
copollutantthan in
single-pollutant models.
No quantitative data
shown.
Adjustment for 3-day avg
NO2 before each child's
lung function
test did not alter
association for long-term
NO2.
FVC: -27.5
(-54.7, -0.2)
FEVi: -29.3
(-47.5,-11.1)
MMEF: -61.0
(-109.1, -12.8)
January 2015
6-46
DRAFT: Do Not Cite or Quote
-------
Table 6-4 (Continued): Prospective studies of long-term nitogen dioxide exposure and lung function and lung
development in children.
Results (95% Cl)b
FEVi and FVC
(mL)
MMEF, PEF, and
Comments FEF (mL/sec)
Study3
Exposure Assessment Pollutant Correlation Statistical Methods
Gauderman et al. (2007)
N = 3,677 children
Mean age 10 yr(SD 0.44)
12 CHS communities
Central community NR
monitoring sites for NO2.
Range of mean NO2
across communities:
34.6 ppb
Identified several
indicators of residential
exposure to traffic from
large roads.
Regression analysis.
Examined models
containing both local
traffic exposure and
regional air pollutants.
Adjusted for height,
height squared, BMI,
BMI squared, present
asthma status,
exercise or respiratory
illness on the day of
the test, any tobacco
smoking by the child in
the previous yr, and
indicator variables for
field technician.
NO2 and distance to
freeways were
independently associated
with decrements in lung
development. Compared
with living >1,500 away
from a freeway, living
within 500 m of a freeway
was associated with a
mean
percentage-predicted
FEVi of 97.0% (94.6, 99.4)
and MMEF of 93.4% (89.1,
97 7)
^ ' • ' /•
Change in FEVi
over 8-yr period:
-32 mL/sec
95% Cl not
reported.
Breton et al. (2011)
2 cohorts of fourth-grade
children
1993 cohort 1: n = 1,759
1996 cohort 2: n = 2,004
Mean age at baseline: 10.0 yr.
Monitored for 8 yr, through 12th
grade.
Central monitoring stations
in each of the original
12 study communities from
1994 to the present.
Average hourly
concentrations of NO2
used to compute annual
averages.
NO2 range across
communities: 33.9 ppb
NO2-PM2.5 = 0.79
NO2-O3 = -0.11
Hierarchical mixed
effects with adjustment
for height, height
squared, BMI, BMI
squared, current
asthma status,
exercise or respiratory
illness on the day of
the test, any tobacco
smoking by the child in
the last yr, glutathione
S-transferase mu 1
genotype, and
indicator variables for
the field technician.
Main purpose was to
determine whether
sequence variation in
genes in the glutathione
synthesis pathway alters
susceptibility to air
pollution effects on lung
function.
Haplotype "0100000" was
associated with a 39.6 mL,
29.1 mL, and 51.0 mL/sec
reduction in 8-yr growth of
FEVi, FVC, and MMEF,
respectively.
FEVi: -29.83
(-49.96, -9.70)
FVC: -29.84
(-54.73, -4.95)
MMEF: -54.38
(-90.80, -17.96)
January 2015
6-47
DRAFT: Do Not Cite or Quote
-------
Table 6-4 (Continued): Prospective studies of long-term nitogen dioxide exposure and lung function and lung
development in children.
Study3
Exposure Assessment Pollutant Correlation Statistical Methods
Comments
Results (95% Cl)b
FEVi and FVC
(mL)
MMEF, PEF, and
FEF (mL/sec)
Urmanetal. (2014)
N = 1,811 children (82% of the
active cohort) from 8 CHS
communities
Cohort established in 2002.
LUR developed from
900 monitoring sites in
CHS communities
(Franklin etal.. 2012).
Near roadway NO2, NO,
and NOx assessed based
on (1) residential distance
to the nearest freeway or
major road and (2)
estimated near-roadway
contributions to residential
NO2, NO, and NOx.
Cross-validation R2 for
NO2 = 0.69, 0.72
Regional Os, NO2, PM-io,
PlVh.s assessed from
central sites.
Community mean centered
distribution for NO2:
6.4 ppb
For regional pollutant
concentrations:
NO2-PMio: 0.06
NO2-PM2.5: 0.60
For near-roadway NO,
NO2, and NOx (within
communities): >0.90
Linear regression
models (with fixed
effects for each study
community) for
near-roadway
estimates.
Mixed model (with
random intercept for
community) for
regional pollutants and
joint effects with
near-roadway NO2,
NO, and NOx.
Adjusted for log of
height and its squared
value, BMI, BMI
squared, sex, age,
sex x age interaction,
race, Hispanic
ethnicity, respiratory
illness at time of test,
and field technician
and study community.
Lung function deficits of
2-3% were associated
with regional PM-io, PlVh.s,
and Os across the range of
exposure between
communities.
NO2 was associated with
FEVi but not FVC.
Associations with regional
pollution and
near-roadway NOx were
independent in models
adjusted for each.
The effects of
near-roadway NOx were
not modified by regional
pollutant concentrations.
Per 20 ppb increase
in near-roadway
NOx:
FVC: -1.74%
(-2.93, -0.55%)
FEVi: -1.23%
(-2.44,-0.01%)
Associations were
observed in all
communities and
were similar for NO2
and NO.
Residential
proximity to a
freeway was
associated with a
reduction in FVC.
January 2015
6-48
DRAFT: Do Not Cite or Quote
-------
Table 6-4 (Continued): Prospective studies of long-term nitogen dioxide exposure and lung function and lung
development in children.
Results (95% Cl)b
FEVi and FVC
(mL)
MMEF, PEF, and
Comments FEF (mL/sec)
Study3
Exposure Assessment Pollutant Correlation Statistical Methods
Children, Allergy, Milieu, Stockholm, Epidemiology Survey (BAMSE)
Schultzetal. (2012)
N = 1,924 children followed from
birth until 8 yr of age.
Related Publications:
Nordlinq et al. (2008)
Long-term NOx exposure NR
estimated using dispersion
model and emission
inventories. Time-weighted
average exposures for
various time windows
estimated based on
lifetime residential, day
care, and school
addresses.
Short-term exposure
assessed from central
sites.
Linear regression
adjusted for
municipality, sex, age,
height and heredity for
asthma and/or allergy.
Additional adjustments
for temperature,
relative humidity,
ozone, and PM-io
levels during 3-7 days
before each child's
pulmonary function
test showed little effect
on the estimates of the
long-term effects of air
pollution.
The odds ratios associated
with 80 and 85% of
predicted FEVi were 2.1
(95% Cl: 0.6, 8.1) and 3.4
(95% Cl: 1.6, 7.4),
respectively, for first yr
exposure to NOx.
Specific adjustment for
short-term NOx was not
discussed.
Per 47 ug/m3
increase in NOx in
first yr of life:
-34.9 mL
(-80.1, 10.4) in
FEVi
Group sensitized
against any
common inhalant or
food allergens, and
those with asthma
of Q \/r-
dl O yi .
-98.9 mL (-169.4,
-28.4)
No clear association
seen with NOx after
infancy.
January 2015
6-49
DRAFT: Do Not Cite or Quote
-------
Table 6-4 (Continued): Prospective studies of long-term nitogen dioxide exposure and lung function and lung
development in children.
Results (95% Cl)b
FEVi and FVC
(mL)
MMEF, PEF, and
Comments FEF (mL/sec)
Study3
Exposure Assessment Pollutant Correlation Statistical Methods
PIAMA
Eenhuizen et al. (2013)
N = 880 children age 4 yr
Long-term average air
pollution concentrations of
NO2, PlVh.s, and soot at the
residential address at birth
were assessed using LUR
models.
Mean NO210.4 ppb
Daily average air pollution
concentrations on the day
of interrupter resistance
(Rint) measurement
obtained from central sites.
For birth address:
NCb-PIVh.s: 0.93
NO2-soot: 0.96
NO2 on test day and
long-term NO2: 0.55
NO2 on the day before
the test and long-term
NO2: 0.57
Multiple linear
regression adjusted
for sex, age at
examination (days),
height, weight,
maternal prenatal
smoking, any smoking
in the child's home,
use of gas for cooking,
parental allergy,
dampness in the
home, education of
the parents, season,
temperature, and
humidity on the day of
the Rint measurement.
First report of an
association in 4-yr old
children. Rint at age 4 yr
predicted asthma and
wheeze at age 8 yr.
Long-term average PM2.5
and soot associated with
Rint. Adjustment for
individual level
confounders, season and
weather on the test day
reduced air pollution effect
estimates only slightly. A
monotonic increase of Rint
with increasing NO2
concentration was seen,
with no threshold
identified.
Change in Rint:
0.05(0.001, 0.11)
kPa*s/L
January 2015
6-50
DRAFT: Do Not Cite or Quote
-------
Table 6-4 (Continued): Prospective studies of long-term nitogen dioxide exposure and lung function and lung
development in children.
Results (95% Cl)b
FEVi and FVC
(mL)
MMEF, PEF, and
Comments FEF (mL/sec)
Study3
Exposure Assessment Pollutant Correlation Statistical Methods
Manchester Asthma and Allergy Study (MAAS)
Molteretal. (2013)
Molteretal. (201 Oa)
Molteretal. (201 Ob)
Molteretal. (2012)
N = 1,185
Birth cohort recruited between
1995 and 1997 and evaluated at
ages 3. 5. 8 and 11 yr.
Microenvironmental PM-io and NO2
Exposure Model that moderately to strongly
estimates personal correlated in all
exposure by incorporating exposure time
children's time-activity windows. Pearson
patterns and LUR modeled r= 0.59 to 0.89.
concentrations.
The modeled estimates
agreed well with measured
NO2 concentrations.
Mean (SD) NO2:
14.76 (6.6) ppb
Generalized
estimating equations.
Potential confounding
variables evaluated
included sex, age,
ethnicity, older
siblings, sensitization,
asthma or current
wheeze, family history
of asthma, parental
smoking, parental
atopy, daycare
attendance during the
first 2 yr of life,
hospitalization during
the first 2 yr of life,
presence of a gas
cooker in the home,
presence of a dog or
cat in the home, visible
signs of dampness or
mold in the home,
body height, body
weight, BMI, maternal
age at birth,
gestational age,
duration of breast
feeding, Tanner stage
(age 11 only), and
socioeconomic status
(paternal income).
Change in
percentage
predicted FEVi from
onci Ł. -1 "1 \/r- "1 Ł. KC\
aye D 1 1 yi . 1 D.DU
(-26.13, -5.26).
Based on the
average predicted
FEVi in cohort of
1.65 L, change
equals total
decrease in FEVi of
263 mL.
Change in
post-bronchodilator
FEVi from age
5-11 yr: -22
/ QT Q (~\\
(~of, — o.U).
Equivalent to total
413 mL decrease.
January 2015
6-51
DRAFT: Do Not Cite or Quote
-------
Table 6-4 (Continued): Prospective studies of long-term nitogen dioxide exposure and lung function and lung
development in children.
Results (95% Cl)b
FEVi and FVC
(mL)
MMEF, PEF, and
Comments FEF (mL/sec)
Study3
Exposure Assessment Pollutant Correlation Statistical Methods
Mexico City School Children Cohort
Roias-Martinez et al. (2007a)
Roias-Martinez et al. (2007b)
N = 3, 170 children
Age 8 yr at baseline, from
31 schools
Examined April 1996 through
May 1999.
NO2 assigned from closest 24-h avg NO2 and
air-quality monitoring 8-h avg Os. 0.166
station within 2 km of 24-h avg NO2 and
school. 24-h avg PMm 0.250
NO2 mean (SD) across
communities:
27.2 (10.9) to 42.6 (13.2)
General linear mixed
models adjusted for
age, BMI, height,
height by age,
weekday time spent in
outdoor activities, and
environmental tobacco
smoke.
NO2, 03, and PM-io were
associated with
decrements in lung
development after
adjusting for short-term
averages (day before lung
function measurement) for
the pollutants.
Girls
FVC: -40
(-46, -34)
FEVi: -27
(-33, -22)
FEF25-75%: 7
(-8, 18)
Boys
FVC: -38
(-44, -31)
FEVi: -22
(-28, -16)
FEF25-75%: 3
(-10, 16)
January 2015
6-52
DRAFT: Do Not Cite or Quote
-------
Table 6-4 (Continued): Prospective studies of long-term nitogen dioxide exposure and lung function and lung
development in children.
Results (95% Cl)b
FEVi and FVC
(mL)
MMEF, PEF, and
Comments FEF (mL/sec)
Study3
Exposure Assessment Pollutant Correlation Statistical Methods
Oslo Birth Cohort
Oftedal et al. (2008)
N = 2,307 children ages 9 and
10 yr
Children lived in Oslo, Norway,
since birth.
Spirometry conducted
2001-2002.
Related Publications: SI0rdal
et al. (2003)
NO2 estimated for 1 km2 of
residence with dispersion
model based on
emissions, meteorology,
topography, and central
site measurements.
Modeled values
represented ambient
concentrations reasonably
well.
Mean NC>2:
16 ppb in the first yr
11.86 ppb for lifetime
exposure
NO2 and PM
r= 0.83-0.95
Multiple linear
regression adjusted
for sex, height, age,
BMI, birth weight,
temperature Lags
1-3 days before the
lung function test,
current asthma,
indicator for
participation in the
Oslo Birth Cohort
study, maternal
smoking in early
lifetime, parental
ethnicity, education,
and smoking.
Associations with NO2
were stronger in girls.
In models that included
both short- and long-term
NO2 exposures, only the
association with long-term
NO2 remained.
NO2 in first yr of life
among all children:
FEVi: -6.0
(-18.0,6.2)
FVC: -1.4
(-14.6, 11.8)
PEF: -57.9
(-92.5, -22.3)
FEFso%: -37.3
(-71.2,-3.5)
BAMSE = Children, Allergy, Milieu, Stockholm, Epidemiology Survey; BMI = body mass index; CHS = Children's Health Study; Cl = confidence interval; EC = elemental carbon;
FEF = forced expiratory flow; FEV, = forced expiratory volume in 1 second; FVC = forced vital capacity; kPa*s/L = kilopascals per second per liter; LUR = land-use regression;
MAAS = Manchester Asthma and Allergy Study; MMEF = maximum (or maximal) midexpiratory flow; NO = nitric oxide; NO2 = nitrogen dioxide; NOX = sum of NO and NO2;
O3 = ozone; OC = organic carbon; PEF = peak expiratory flow; PIAMA = Prevention and Incidence of Asthma and Mite Allergy; PM = particulate matter; PM2.5 = particulate matter
with a nominal aerodynamic diameter less than or equal to 2.5 |jm; PMio = particulate matter with a nominal aerodynamic diameter less than or equal to 10 |jm;SD = standard
deviation.
Note: FEF50% = forced expiratory flow at 50% of forced vital capacity. FEF2s-75% = forced expiratory flow between 25 and 75% of forced vital capacity.
aStudies are presented in the order of appearance in the text.
"•Results are presented for a 10 ppb change in NO2 and 20 ppb change in NOX unless otherwise specified.
January 2015
6-53
DRAFT: Do Not Cite or Quote
-------
c
O
E
LU
I
1450-
1420-
1390-
1360-
1330-
1300-
1270-
1240-
1210-
0
O Girls
• Boys
O
-2485
-2455 =*
-2425 &
-2395
Oo
i
10
i
20
15 20 25
N02 (ppb)
30
i
35
00
c
O
-2365 |
-2335 Ł
E
-2305 ^
4->
-2275 Ł
-2245
40
Source: Reprinted with permission of the Massachusetts Medical Society, Gauderman et al. (2004).
Figure 6-5 Community-specific average growth in forced expiratory volume
in 1 second (FEVi; ml_) among girls and boys during the 8-year
period from 1993 to 2001, plotted against average nitrogen
dioxide (NOa) concentrations from 1994 through 2000.
Urman et al. (2014) examined lung function in 1,811 cohort children (82% of the active
cohort) from eight communities in the CHS cohort established in 2002-2003 for
near-roadway and regional air pollution exposure effects. Since the beginning of this
study, regional pollutant levels have been measured continuously at central monitoring
locations in the study communities. LUR models were developed from 900 monitoring
sites in the CHS communities (Franklin et al.. 2012). Distance to roadway and central site
air pollution were also analyzed. LUR models also included NOx estimated from a
dispersion model to estimate near-roadway NOx. For forced expiratory volume in
1 second (FEVi), there was little change in the association of near-roadway NOx after
adjusting for a copollutant (i.e., Os, PM2s or PMio). Central site NO2 remained associated
with FEVi after adjustment for near-roadway NOx. Near-roadway NO2 also was
January 2015
6-54
DRAFT: Do Not Cite or Quote
-------
1 associated with decrements in lung function. Near-roadway NOx was associated with
2 lung function decrements in children with and without asthma, suggesting that
3 traffic-related pollution may potentially affect all children. The association for
4 within-community near-roadway NOx was somewhat less than that for
5 between-community regional NOx, although the strength of inferences from quantitative
6 comparisons of effect estimates for regional and near-roadway pollution are limited.
7 Gauderman et al. (2007) reported results of an 8-year follow-up on 3,677 children who
8 participated in the CHS. The FEVi reduction was -31.5 mL (95% CI not reported) for an
9 increase in NO2 of 10 ppb. Children living <500 m from a freeway (N = 440) had deficits
10 in lung development over the 8-year follow-up compared to children who lived at least
11 1,500 m from a freeway. When examined in the same model, both distance to freeway
12 and NO2 measured at community central sites were associated with decrements in lung
13 development. Acid vapor, EC, PMio, and PIVb 5, but not Os, were associated with reduced
14 lung development. There was no evidence that the association for NO2 differed according
15 to distance to freeway or vice versa. Throughout the 8-year follow-up, around an 11%
16 loss of study participants per year was observed.
17 Gauderman et al. (2004) examined the 1993 CHS cohort of 1,759 children aged 10 to
18 18 years and states that although the average increase in FEVi over time was larger in
19 boys than in girls, the associations of lung development with NO2 measured at
20 community air monitoring sites did not differ between the sexes, as shown in Figure 6-5.
21 As depicted by the regression line in Figure 6-5. for both sexes combined, the average
22 difference in FEVi growth over the 8-year period between the communities with the
23 lowest and highest 8-yr avg NO2 concentration (34.6 ppb difference) was -101.4 mL
24 (95% CI:-164.5,-38.4).
25 Gauderman et al. (2004) further indicated that NO2 exposure over the 8-year follow-up
26 was associated with clinically relevant decrements in attained lung function at the age of
27 18 years (Figure 6-6). Clinical importance was defined as FEVi less than 80% of the
28 predicted value for height, body mass index (BMI), sex, race/ethnicity, and asthma status.
29 Across the 12 communities, higher NO2 was associated with an increase in the percentage
30 of children with FEVi less than 80% predicted.
31 A recent study examined the relationship between long-term exposure to air pollution and
32 lung function in 1,924 school-age children in the Swedish birth cohort BAMSE (Schultz
33 etal.. 2012). NOx exposure during the first year of life was associated with a deficit in
34 FEVi. The odds ratios of having a deficit of 80% and 85% of predicted FEVi were 1.69
35 (95% CI: 0.67, 4.2), and 2.6 (95% CI: 1.4, 4.5), respectively, for a 20 ppb increase in
36 NOx. No impact of short-term air pollution exposure on the estimates of the long-term
January 2015 6-55 DRAFT: Do Not Cite or Quote
-------
effects of air pollution was observed in analyses of PMio; similar analyses were not
conducted for NOx.
u
10n
6-
4-
2-
0
0
R=0.75
P=0.005
UP
ML
rSM
4LE
i
10
» LN,
20
N02(ppb)
i
30
40
Note: The correlation coefficient (R) and P value are shown for each comparison. AL = Alpine, AT = Atascadero, LE = Lake
Elsinore, LA = Lake Arrowhead, LN = Lancaster, LM = Lompoc, LB = Long Beach, ML = Mira Loma, RV = Riverside, SD = San
Dimas, SM = Santa Maria, and UP = Upland. NO2 = nitrogen dioxide.
Source: Reprinted with permission of the Massachusetts Medical Society, Gauderman et al. (2004).
Figure 6-6 Community-specific proportion of 18-year-olds with a forced
expiratory volume in 1 second (FEVi) below 80% of the predicted
value, plotted against the average concentrations of nitrogen
dioxide (NO2) from 1994 through 2000.
3
4
5
6
7
8
9
10
11
Because of the difficulties of lung function examinations in young children, such as those
4 years old or younger, limited data are available. Eenhuizen et al. (2013) assessed the
relationship between long- and short-term exposure to traffic-related air pollution and
interrupter resistance (Rint), an indicator of airway resistance, in 4-year-old children
participating in the Prevention and Incidence of Asthma and Mite Allergy (PIAMA)
Dutch birth cohort study. Of the original invited 1,808 children, a total of 880 children
were in the final analysis. The children with valid Rint data did not have different
characteristics than the population recruited for the study. Long-term average
concentrations of NC>2, PIVbs, and soot at the residential address at birth were assessed
January 2015
6-56
DRAFT: Do Not Cite or Quote
-------
1 using LUR models as discussed in Section 6.2.2.1. and daily average air pollution
2 concentrations on the day of clinical examination were obtained. Positive associations
3 were observed between long-term average NO2 concentrations and Rint. Such findings
4 are supported by the study showing NO2-induced increased airway resistance in guinea
5 pigs (Section 6.2.2.3). They also support a relationship between NO2 and asthma, given
6 that Rint at age 4 years was a predictor of asthma and wheeze at age 8 years. A
7 monotonic increase in Rint with increasing NO2 concentration, with no suggestion of a
8 threshold was observed. Short-term exposure was not associated with interrupter
9 resistance. NO2 concentrations on the test day and the day before the test were
10 moderately correlated with long-term concentrations. This is the first report of an
11 association in 4-year-old children. Because of the high correlation between modelled
12 PM2 5, NO2 and soot (quantitative data not reported), the study could not disentangle an
13 independent association for any of the examined pollutants.
14 The long-term effects of PMio and NO2 exposure on specific airway resistance (sRaw)
15 and FEVi before and after bronchodilator treatment was examined within the Manchester
16 Asthma and Allergy Study (MAAS) birth cohort [N = 1,185; (Molteretal.. 2013)1. At
17 age 11 years, the cohort size was 813. The authors utilized an LUR model that
18 incorporated children's time-activity patterns to produce total exposure estimates with
19 spatial resolution at the individual level rather than community level (the
20 Microenvironmental Exposure Model). The model was validated and there was good
21 agreement between modeled and measured total personal NO2 concentrations for
22 short-term averaging times (Molteretal.. 2012). Lifetime exposure to NO2 was
23 associated with less growth in FEVi (percentage predicted) over time, both before (16.0%
24 [95% CI: -26.0, -0.5] for a 10-ppb increase in NO2) and after bronchodilator treatment
25 (23% [95% CI: -37.0, -9.0]).
26 As part of ESCAPE, Gehring et al. (2013) analyzed data from birth cohort studies
27 conducted in Germany, Sweden, the Netherlands, and the U.K. that measured lung
28 function at 6 to 8 years of age (N = 5,921). The five birth cohorts [BAMSE, MAAS,
29 German Infant Nutritional Intervention covers urban Munich, Germany, and its
30 surrounding areas (GINI SOUTH, GINI/LISA), and PIAMA] were discussed in
31 Section 6.2.2.1. Annual average exposures to NO2, NOx, PM2 5, PMio, PM coarse, and
32 PM absorbance at the birth address and current address were estimated by LUR models,
33 except for the BAMSE cohort, for which a dispersion model was used. Associations of
34 lung function with estimated air pollution concentrations and traffic indicators were
35 examined for each cohort using linear regression analysis, and then combined by random
36 effects meta-analysis. Across the five cohorts, annual mean (SD) for NO2 ranged from
37 7.44 (2.87) to 12.6 (1.91) ppb. Long-term associations were adjusted for short-term
38 changes in pollutants measured at central sites. NO2 and NOx estimated for the current
January 2015 6-57 DRAFT: Do Not Cite or Quote
-------
1 address were associated with decrements in both FEVi and forced vital capacity (FVC) as
2 were PM2 5 and PM2 5 absorbance. NO2 and PM2 5 at the current address also were
3 associated with peak expiratory flow (PEF). For the five cohorts combined, NO2 at the
4 birth address was associated with a smaller decrease in lung function than was NO2 for
5 the current address. Short-term (7-day avg) exposure to NO2 and PMio also were
6 associated with lung function decrements. Two traffic measures (i.e., traffic intensity
7 nearest street and traffic load on major roads within a 100-m buffer) were associated with
8 deficits in lung function, although the effect estimates had wide confidence intervals,
9 indicating imprecise associations. Annual average concentrations of NO2, NOx, PM2 5,
10 and PMio at the current address were associated with clinically relevant lung function
11 decrements (FEVi < 85% predicted). In copollutant models with NO2 and PM2 5, effect
12 estimates for both pollutants were reduced, but the relative impact on NO2 and PM2 5
13 differed among lung function indices. The association for NO2 was reduced more than
14 that for PM25, for FEVi, and PEF. In contrast, the association for PM25 was reduced
15 more for FVC than for NO2. These findings add to the notion that exposure to NO2 may
16 result in reduction in lung function in school children, but uncertainties related to the
17 potential for confounding by traffic-related copollutants do not conclusively support
18 independent effects.
19 In Mexico City. Mexico. Rojas-Martinez et al. (2007b) and Rojas-Martinez et al. (2007a)
20 evaluated lung development in a prospective cohort of children aged 8 years at baseline.
21 Long-term pollutant exposures were assigned from the closest central monitoring site
22 located within 2 km of schools. An unspecified number of children were lost to follow-up
23 during the study, mainly because they moved to another area of the city or to another city
24 altogether. Information was obtained from atotal of 3,170 children. A 10-ppb increase in
25 NO2 was associated with an annual deficit in FEVi of 27 (95% CI: 22, 33) mL in girls
26 and 22 (95% CI: 16, 28) mL in boys. The negative association for NO2 persisted in
27 copollutant models with Os or PMio. A deficit in lung development was observed for Os,
28 PMio, and NO2 after adjusting for the short-term associations of these pollutants
29 (previous-day concentrations).
30 A cohort study in Oslo, Norway, examined associations of short- and long-term NO2 and
31 other pollutant exposures on PEF and forced expiratory flow at 25% of forced vital
32 capacity and 50% of forced vital capacity in 2,307 children ages 9-10 years (Oftedal
33 et al.. 2008). In models that included both short- and long-term NO2 exposures estimated
34 from dispersion models, only the association with long-term NO2 remained. Adjusting for
35 a contextual socioeconomic factor diminished the association with NO2.
January 2015 6-58 DRAFT: Do Not Cite or Quote
-------
Years of
Study Period of Exposure Age in Years
Breton et al. (2011) continuously during follow-up 18 (SD: 0.6)
Gauderman et al. (2004) continuously during follow-up 10 to IS
Rojas-Martinez et al. (2007) continuously during follow-up 11
Schultzetal. (2012)
Oftedal el al. (200S)
Molter et al. (2013)
Urman elal. (2014)
birth yean 1 to 4: 4 to 8 8
birth year: second year lifetime 8
continuously during follow-up 3 to 11
continuously during follow-up 5 to 7
i
Lung function Mem
FEVl (mL)
FEVl (ml.)
FEVl (mL)
FEVl (mL)
FVC (mL)
FVC (mL)
FEF25-7J (mL'sec)
FEF25-7J (mL/sec)
FEVl (mL)
FVC (mL)
FEVl (%)
FEVl (%)
FVC {%)
Central site — • —
Central site — • —
Central site - girls -*
Central site -boys •*-
Central site - girls *-
Central site -boys -•-
Central site - girls
Central site - boys
Dispersion — •
"I
t
Dispersion »
LUR 4
4
•
to
t
>
-120 -80 -40 0 40 80 120
Change in Lung Function Estimate and 95% CI
Note: Studies in red are recent studies. Studies in black were included in the 2008 ISA for Oxides of Nitrogen. All mean changes in this plot are standardized to a 10-ppb increase in
NO2 and a 20-ppb increase in NOX concentration. Effect estimates from studies measuring NOX in |jg/m3 (Schultz et al., 2012) have not been standardized. Circles = NO2;
Diamonds = NOX.
Figure 6-7 Associations of nitrogen dioxide (NO2) or the sum of nitric oxide and NO2 (NOx) with lung
function indices from prospective studies.
January 2015
6-59
DRAFT: Do Not Cite or Quote
-------
6.2.5.2 Lung Function in Adults
1 Prospective studies evaluating long-term NCh exposure and pulmonary function in adults
2 are discussed. The limited number of cross-sectional studies (Forbes et al., 2009b; Sekine
3 et al.. 2004) have results that are consistent with those from prospective studies.
4 Gotschi et al. (2008) examined the relationship between air pollution and lung function in
5 adults in the ECRHS cohort. FEVi and FVC were assessed at baseline and after 9 years
6 of follow-up from 21 European centers (followed-up sample N = 5,610). Quantitative
7 results were not reported; NC>2 was reported only to show no statistically significant
8 association with average lung function. This is in contrast to the results from Ackermann-
9 Liebrich et al. (1997) (SAPALDIA) and Schikowski et al. (2005) [Study on the Influence
10 of Air Pollution on Lung, Inflammation, and Aging (SALIA)], which examined far more
11 homogenous populations than the population assessed in the ECRHS.
12 A recent study (Boogaard et al.. 2013) evaluated the impact on pulmonary function of a
13 reduction in outdoor pollution concentrations resulting from the policy implementation of
14 forbidding old heavy duty vehicles in all inner cities and other related policies. At
15 12 locations in the Netherlands, air pollution concentrations were measured on the street
16 where participants lived within 500 m of subjects' homes. Respiratory health was
17 measured in 2008 and 2010, during which air pollution concentrations decreased. The
18 study population included both children and adults. Eighty-four percent were above
19 30 years of age at baseline. The participation rate in the study was around 10%. Over the
20 two time periods, 585 subjects were re-evaluated for spirometry. Reductions in
21 concentrations of NC>2 and NOx as well as soot, copper (Cu), and Fe were associated with
22 increases in FVC. Airway resistance decreased with a decline in PMio and PIVb 5,
23 although these associations were somewhat less consistent. No associations were found
24 with eNO. Results were driven largely by the small group of residents living at the one
25 urban street where traffic flow as well as air pollution were drastically reduced.
26 In a Nottingham, U.K. cohort of adults aged 18-70 years, lung function changes were
27 evaluated in a cross-sectional analysis of 2,599 subjects at baseline and a longitudinal
28 analysis of 1,329 subjects followed up 9 years later (Pujades-Rodriguez et al.. 2009).
29 There were no substantial cross-sectional associations between home proximity to the
30 roadside and NC>2 concentration with lung function or any other outcome. Also, neither
31 exposure was associated with a decline in FEVi over time. Insufficient contrast in NC>2
32 exposure (interquartile range: 18.1-19.1 ppb) may be a factor in the inability to detect
33 any associations for NC>2 in this study population.
January 2015 6-60 DRAFT: Do Not Cite or Quote
-------
6.2.5.3 Toxicological Studies of Lung Function
1 Studies included in the 2008 ISA for Oxides of Nitrogen (U.S. EPA. 2008) showed
2 inconsistent evidence of changes in lung function in animals after long-term exposure to
3 concentrations of NC>2 relevant to ambient exposure. No recent studies are available.
4 Arner and Rhoades (1973) published early studies that exposed rats to 2,900 ppb NO2
5 continuously for 5 days/week for 9 months and reported changes in lipid composition in
6 the airway that could be related to observed functional consequences, including decreased
7 lung volume and compliance and increased surface tension, although these changes have
8 not been consistently observed in animal studies.
9 Tepper et al. (1993) exposed rats to a background concentration of 500 ppb NO2 for
10 16 h/day followed by a 6 hour peak of 1,500 ppb and 2 hours of downtime for up to
11 78 weeks. Frequency of breath was significantly slower in these animals and was
12 paralleled by a trend toward increased tidal volume, expiratory resistance, and inspiratory
13 and expiratory time, although changes were not statistically significant. Mercer et al.
14 (1995) and Miller et al. (1987) published studies with similar exposures in rats and mice,
15 respectively, and also reported that NO2 exposure did not alter lung function, although
16 mice tended to have slightly decreased vital capacity from 16 to 52 weeks of exposure.
17 Inconsistent effects were also described in studies of long-term NO2 exposure in the
18 range of 6-7 weeks. Stevens etal. (1988) exposed 1-day and 7-week old rats to 500,
19 1,000, and 2,000 ppb NC>2 continuously with two daily peaks at three times the baseline
20 concentration (1,500, 3,000, and 6,000 ppb) for 1-7 weeks and observed different results
21 among age groups. Rats exposed from 1 day of age had increased lung compliance after
22 3 weeks of exposure that returned to control levels by 6 weeks (1,000 ppb with 3,000 ppb
23 peaks). In rats exposed from 7 weeks of age, compliance was decreased after 6 weeks of
24 exposure at 1,000 and 2,000 ppb NC>2. In an 8-week study, Lafuma et al. (1987). reported
25 increased lung volumes in animals exposed to 2,000 ppb (8 h/day, 5 days/week), but vital
26 capacity and compliance were not affected.
6.2.5.4 Summary of Lung Function
27 In children, recent findings from longitudinal studies provide further support that
28 early-life NC>2 exposure is associated with long-lasting impact on the lung development.
29 Additionally, cross-sectional studies (Gao etal.. 2013; Svendsen et al.. 2012; Lee et al..
30 201 Ib: Rosenlund et al.. 2009b: Tager etal.. 2005; Sekine et al.. 2004; Moseler etal..
31 1994) report associations between NC>2 exposure and decrements in lung function in
32 children. A meta-analysis across five birth cohorts in Europe using LUR exposure
January 2015 6-61 DRAFT: Do Not Cite or Quote
-------
1 estimates reported results consistent with the rest of the evidence base. Investigation in
2 adults is limited, but some longitudinal studies indicate associations. Associations are
3 observed in young children aged 4 years. Effects were mainly on FEVi, which reflects
4 the mechanical properties of the airways and not as much on FVC, representing lung size.
5 Two longitudinal studies examined associations with short-term NC>2 and report that the
6 short-term exposures do not impact the association between long-term NCh exposure and
7 lung function. Recent studies use exposure methods such as LUR that provide more
8 individual exposure estimates than central sites do. Results have been observed in various
9 locations in studies using varied exposure assessment methods, lung function
10 measurements, and time of follow-up with children. A linear concentration-response
11 relationship was observed in one study. In limited analysis of copollutant models, NO2
12 associations persisted with adjustment for Os, and the NOx association persisted with
13 adjustment for PIVbs. In copollutant models with PIVbs, NC>2 remained associated with
14 FVC but not FEVi or PEF.
15 The limited analysis of potential confounding by traffic-related copollutants, inconsistent
16 results with PM2 5 adjustment, and high copollutant correlations often observed produces
17 uncertainty as to whether NC>2 has an independent effect on lung function. Animal studies
18 do not address this uncertainty in the epidemiologic evidence as they demonstrate
19 inconsistent effects of long-term NC>2 exposure on lung function. However, age may be
20 an important factor that influences the effect of NC>2 on lung function that has not been
21 adequately addressed by the existing body of toxicological evidence.
6.2.6 Changes in Lung Morphology
22 While no recent studies are available, the 2008 ISA for Oxides of Nitrogen (U.S. EPA.
23 2008) reported that animal toxicological studies demonstrate morphological changes to
24 the respiratory tract resulting from exposure to NC>2. Details from the available studies
25 are presented in Table 6-2. Studies examined long-term exposures to NC>2 to determine
26 effects on lung structure and morphology and report variations in response to
27 concentrations below 5,000 ppb. Wagner etal. (1965) exposed dogs, rabbits, guinea pigs,
28 rats, hamsters, and mice to 1,000, 5,000, or 25,000 ppb NC>2 for up to 18 months and
29 found enlarged air space and edema and areas of mild to moderately thickened septae
30 with chronic inflammatory cells. However, some of these observations were also made in
31 control animals and were not considered to be significant in any species. Importantly, this
32 study demonstrated differences in sensitivity to NC>2 across species. Furiosi etal. (1973)
33 exposed monkeys and rats to 2,000 ppb NC>2 continuously for 14 months and also found
34 species-specific responses; monkeys experienced hypertrophy of the bronchiolar
35 epithelium that was most notable in the respiratory bronchioles in addition to
January 2015 6-62 DRAFT: Do Not Cite or Quote
-------
1 development of a cuboidal phenotype in the squamous proximal bronchiolar epithelium.
2 In rats, these effects were more occasional under identical exposure conditions.
3 The majority of other morphologic studies employed rodent models to evaluate effects of
4 NO2 exposure. Chang etal. (1986) compared responses in mature and juvenile rats to an
5 urban exposure pattern of NC>2 for 6 weeks (500 ppb continuously with two daily peaks at
6 1,500 ppb). Mature rats were more sensitive to NC>2 exposure and exhibited increased
7 surface density of the alveolar basement membrane and decreased air space in the
8 proximal alveolar regions, accompanied by an increase in lung volume attributable to
9 Type II cell hyperplasia and increases in fibroblasts, alveolar macrophages, and
10 extracellular matrix. In the juvenile rats, effects of exposure were limited to thinning of
11 Type II cells that were spread over more surface area compared to controls. Mercer et al.
12 (1995) found more subtle effects in rats with this exposure; lungs did not appear to have
13 differences in alveolar septal thickness, parenchymal cell populations, or cellular size and
14 surface area after 9 weeks of exposure. Although the frequency of fenestrae was
15 increased in the alveolar epithelium, there were no changes found in the extracellular
16 matrix or interstitial cells. Crapo etal. (1984) conducted a 6-week study in rats with a
17 similar exposure pattern at higher concentrations (2,000 ppb NC>2 for 23 h/day with two
18 30-minute peaks of 6,000 ppb) and reported hypertrophy and hyperproliferation of the
19 alveolar epithelium. In another study, rats were exposed to a similar urban exposure
20 pattern in addition to a single high concentration for up to 15 weeks; these animals had
21 subpleural alveolar macrophage accumulation and areas of focal hyperinflation, though
22 the mean linear intercept (MLI), a measure of free distance in the air space, was not
23 changed (Gregory et al.. 1983). Conversely, Lafumaetal. (1987) reported that hamsters
24 exposed to 2,000 ppb NC>2 for 8 h/day, 5 days/week for 8 weeks had increased MLI and
25 decreased internal surface area, but no lesions were found in the bronchiole or
26 bronchiolar epithelium, alveolar ducts, or alveolar epithelium.
27 Kubotaetal. (1987) conducted a 27-month study in rats that included pathological
28 assessments of the airways after continuous exposure to 40, 400, or 4,000 ppb NC>2. At
29 the highest exposure, rats had increased bronchial epithelial proliferation after 9 and
30 18 months, and by 27 months, proliferation and edema resulted in fibrosis. Exposure to
31 400 ppb produced similar morphological changes in the bronchial epithelium that was not
32 apparent until 27 months. Exposure to 40 ppb NC>2 did not result in morphological
33 changes that could be identified by microscopic techniques. Studies conducted at similar
34 concentrations and durations reported analogous effects. Blair etal. (1969) and Hayashi
35 etal. (1987) exposed mice and rats, respectively to 500 ppb for up to 19 months. Blair
36 etal. (1969) described an increase in alveolar size after 3 months of exposure with loss of
37 cilia in respiratory bronchioles, which persisted at 12 months. After 4 months of
38 exposure, Hayashi et al. (1987) reported Type II cell hypertrophy and interstitial edema
January 2015 6-63 DRAFT: Do Not Cite or Quote
-------
1 leading to thickened alveolar septa at 6 months and fibrous pleural thickening at
2 9 months. Similarly, exposure to 500 ppb for 7 months resulted in interstitial edema and
3 Type II cell hyperplasia in rats, and additional injury at 1,000 ppb included loss of cilia in
4 the terminal bronchioles (Yamamoto and Takahashi. 1984). Type II cell hyperplasia was
5 also documented by Sherwin and Richters (1982) as well as an increase in the MLI.
6 These studies demonstrate that long-term exposure to relatively high ambient
7 concentrations of NO2 can result in subtle changes in lung morphology including Type II
8 cell hyperplasia, loss of cilia in the bronchiolar region, and enlarged airspace.
6.2.7 Respiratory Infection
9 Toxicological studies, as reviewed in the 2008 ISA for Oxides of Nitrogen (U.S. EPA.
10 2008), demonstrated NC>2-induced mortality from infection in experiment animals as well
11 as changes in host defense mechanisms. Epidemiologic investigation of the relationship
12 between long-term NC>2 exposure and respiratory infection is limited, particularly in
13 prospective studies.
6.2.7.1 Epidemiologic Studies of Respiratory Infection
14 Respiratory infection and symptoms of infection in infants and young children up to
15 3 years of age were examined in recent studies (Aguilera et al.. 2013; Sunyer etal..
16 2004). In a multicenter prospective cohort study, Sunyer et al. (2004) observed no
17 associations between 2-week indoor NO2 exposure and lower respiratory tract infections
18 during the first year of life. Aguilera etal. (2013) observed an association between
19 increased NC>2 exposure estimated by LUR and increased risk of upper and lower
20 respiratory tract infections in infants.
21 In a population-based case-control study in Hamilton, Ontario, Canada, Neupane et al.
22 (2010) examined the relationship between ambient NO2, PM2 5, and SO2 and hospital
23 admission for community-acquired pneumonia in 345 hospitalized patients aged 65 years
24 or more. Control participants (n = 494) aged 65 years or more were randomly selected by
25 telephone calls from the same community as cases from July 2003 to April 2005. Air
26 pollutants in this study were assessed as exposures over the previous 12 months to test
27 the chronic effect rather than an acute effect. Three methods were used to estimate the
28 annual average NO2 concentrations: daily ambient data, LUR models, and IDW.
29 Participants had to present to the emergency room with at least two signs and symptoms
30 for pneumonia and have a new opacity on a chest radiograph interpreted by a radiologist
31 as being compatible with pneumonia. NO2 and PlVfc 5 were associated with hospital
January 2015 6-64 DRAFT: Do Not Cite or Quote
-------
1 admission for community-acquired pneumonia, but SC>2 was not. NC>2 exposures
2 estimated by all three methods were associated with pneumonia (ORs per 10 ppb
3 increase: 3.20 [95% CI: 1.37, 7.45] for IDW; 1.97 [95% CI: 1.21, 3.19] for bicubic
4 spline; 1.93 [95% CI: 1.00, 3.74] for LUR). There was no mention of adjustment for
5 short-term exposure effects, and it is not clear what the relative impacts on respiratory
6 infections are for short-term versus long-term exposure. While there are associations
7 observed between short-term NO2 exposure and hospital admissions for pneumonia
8 (Table 5-30). quantitative comparisons with long-term NC>2 exposure effect estimates
9 may not be informative given the differences in exposure assessment methods and
10 distribution of NC>2 concentrations.
11 The association between parent report of physician-diagnosed pneumonia, croup, and
12 otitis media during early childhood and annual average concentrations of NC>2, NOx,
13 PM2 5, PM2 5 absorbance, PMio, and particulate matter with a nominal aerodynamic
14 diameter less than or equal to 10 um and greater than a nominal 2.5 urn (PIVb.5-10; coarse
15 PM) was examined in 10 European birth cohorts: BAMSE (Sweden), Gene and
16 Environmental Prospective Study in Italy [GASPII (Italy)], Gene and Environmental
17 Prospective Study in Italy plus environmental and genetic influences (GINIplus),
18 Lifestyle-Related factors on the Immune System and the Development of Allergies in
19 Childhood plus the influence of traffic emissions and genetics [LISAplus (Germany)],
20 MAAS (U.K.), PIAMA (the Netherlands), and four Infancia y Medio Ambiente cohorts
21 [Spain; (Maclntyre et al., 2014b)1. Exposures were estimated using LUR models and
22 assigned to children based on their residential address at birth. Identical protocols were
23 used to develop LUR models for each study area as part of the ESCAPE project. There
24 was a complete outcome (at least one), exposure (a minimum of NC>2 and NOx), and
25 potential confounder information for 16,059 children across all 10 cohorts. For
26 pneumonia, the meta-analysis produced a combined adjusted OR of 1.64 (95% CI: 1.02,
27 1.65) per 10-ppb increase in NO2. NO2 was associated with otitis media but not croup.
28 The air pollution data used to build the LUR models were measured in 2008-2011, but
29 children in the study cohorts were born as early as 1994. To address this temporal
30 mismatch, a sensitivity analyses was conducted using routine monitoring data to
31 back-extrapolate LUR estimates and produced results generally consistent with the main
32 findings. Correlations between PIVbs and NO2 ranged between 0.42 and 0.80, and
33 correlations between PIVbs absorbance and NO2 ranged between 0.40 and 0.93. For
34 pneumonia, the ORs (95% CI) for NO2 in copollutant models with PM2 5 and PM2 5
35 absorbance were respectively: 1.32 (0.72, 2.42) and 1.36 (0.57, 3.28), showing
36 attenuation and large increase in width of 95% CIs from the single pollutant model for
37 NO2.
January 2015 6-65 DRAFT: Do Not Cite or Quote
-------
6.2.7.2 Toxicological Studies of Respiratory Infection
1 Long-term NCh exposure has been shown to increase susceptibility of experimental
2 animals to infection. Henry etal. (1970) published a study showing that squirrel monkeys
3 exposed to 5,000 ppb NO2 for a period of 2 months and then exposed to Klebsiella
4 pneumoniae or Influenza had increased markers of infection, white blood cell counts and
5 erythrocyte sedimentation rate (ESR), 3 days post-infection. Furthermore, 2 of the
6 1 monkeys exposed to NC>2 died at 3 and 10 days post-infection. When influenza virus
7 was given 24 hours prior to NC>2 exposure and after NC>2 exposure, tidal volume and
8 respiratory rate increased and the ESR increased. One of the three exposed monkeys died
9 5 days post-infection. Ehrlich and Henry (1968) and Ehrlich (1980) also studied the
10 effects of NO2 on Klebsiella pneumoniae infection in mice. Exposures were either
11 continuous or intermittent (6 or 18 h/day) at a concentration of 500 ppb NCh and bacterial
12 challenge was administered at 1, 3, 6, 9, and 12 months. Continuous exposure to NC>2 for
13 3 months or longer resulted in increased mortality rates after infection, whereas
14 intermittent exposures led to increased mortality at 6, 9, and 12 months. Likewise, Miller
15 etal. (1987) showed increased mortality in mice exposed to a base of 200 ppb NO2 with
16 two daily 1-hour peaks of 800 ppb and subsequent challenge with Streptococcus
17 zooepidemicus at 16, 32, and 52 weeks.
6.2.7.3 Subclinical Effects Underlying Respiratory Infection
18 Impaired host defense mechanisms can increase susceptibility to bacterial and viral
19 infection, and toxicological studies have demonstrated that experimental animals exposed
20 to concentrations of NC>2 relevant to ambient exposure for periods greater than 6 weeks
21 have modulated lung host defense including altered characteristics of AMs. Details from
22 these studies, which were also reviewed in the 2008 ISA for Oxides of Nitrogen
23 (U.S. EPA. 2008). are presented in Table 6-2.
24 Alveolar macrophages play a critical role in removing pathogens from the airways and
25 impaired function can increase susceptibility to infection and injury. Aranyi et al. (1976)
26 found that AM morphology was abnormal after 21 weeks of continuous exposure to
27 2,000 ppb NO2 and/or a base of 500 ppb NO2 with 3-hour peaks of 2,000 ppb, although
28 exposures at lower concentrations had no effects on AM morphology. Chang etal. (1986)
29 showed that exposure to 500 ppb NO2 continuously with 1,500 ppb 1-hour peaks twice
30 daily for 6 weeks increased the number of macrophages in the alveoli and their cellular
31 volume. Gregory etal. (1983) reported similar findings and observed AM accumulation
32 in lung sections by light microscopy after exposure to 5,000 ppb NO2 or a base of
33 1,000 ppb NO2 with 5,000 ppb spikes twice each day for 15 weeks.
January 2015 6-66 DRAFT: Do Not Cite or Quote
-------
1 Greene and Schneider (1978) investigated the effects of NO2 exposure on the function of
2 AMs isolated from antigen-sensitized baboons exposed to 2,000 ppb NO2 for 8 h/day,
3 5 days/week for 6 months and found that AMs had diminished response to migration
4 inhibitory factor obtained from antigen-stimulated lymphocytes. However, sample size in
5 this study was small: 3 exposed to NO2 and antigen, 1 exposed to NO2 alone, 1 exposed
6 to antigen alone, and 1 air control. Other studies have not reported on this endpoint.
7 In addition to AMs, mast cells play an important role in host defense and inflammatory
8 processes. Fujimaki and Nohara (1994) investigated the effects of a 12-week continuous
9 exposure to 1,000, 2,000, and 4,000 ppb NO2 in both rats and guinea pigs. Although the
10 number of mast cells in the airway increased after exposure to 2,000 and 4,000 ppb, these
11 changes were not statistically significant. Histamine, released by mast cells, was reduced
12 in rats at 2,000 ppb NO2 and increased in guinea pigs at 4,000 ppb. This observation
13 suggests species differences in response to NO2 exposure.
6.2.7.4 Summary of Respiratory Infection
14 In the small body of epidemiologic studies, long-term NC>2 exposure estimated for
15 subjects' homes by LUR was associated with respiratory infections in school children and
16 pneumonia hospital admissions in adults ages 65 years or older. Results were inconsistent
17 in infants. Particularly for hospital admissions, it is not clear whether the association
18 observed for long-term NC>2 exposure is independent of an association with short-term
19 exposure. As examined in school children, associations for long-term NC>2 exposure were
20 positive with adjustment for PM2 5 or PM2 5 absorbance, but the 95% CIs were very wide.
21 Thus, an independent association for NCh is not clearly indicated. A small body of
22 toxicological studies provide support for an independent effect of NC>2 exposure on
23 respiratory infections, showing that mice and monkeys exposed to NC>2 concentrations in
24 the range of 200 to 5,000 ppb for periods greater than 6 weeks have increased
25 infection-induced mortality and altered AM morphology and function.
6.2.8 Chronic Obstructive Pulmonary Disease
26 Recent epidemiologic studies have examined associations between long-term NCh
27 exposure and effects related to COPD, including the study of indoor NC>2 and respiratory
28 symptoms in adults with COPD (Hansel et al., 2013) described in Section 6.2.3.1. There
29 are few studies examining COPD development, and results are inconsistent. In a
30 prospective cohort study, Andersen et al. (2011) estimated outdoor annual average NO2
31 and NOx since 1971 by a validated LUR model for residential locations and calculated
January 2015 6-67 DRAFT: Do Not Cite or Quote
-------
1 time-weighted averages for 15-, 25- and 35-year periods (Raaschou-Nielsen et al., 2000).
2 No other pollutants were considered. COPD hospital admissions were ascertained from
3 1976, and incidence of COPD was defined as first hospital admission between
4 1993-1997 and June 2006. COPD incidence was associated with the 35- and 25-year
5 mean concentration of NO2 (HR: 1.28; [95% CI: 1.07, 1.54] and 1.22 [95% CI: 1.03,
6 1.45] per 10-ppb increase) and 35-year mean concentration of NOx (1.16 [95% CI: 1.04,
7 1.31] per 20-ppb increase). Weaker positive associations were observed with 25-year
8 mean NOx, 15-year mean NO2 and NOx, and baseline residence traffic proxies (major
9 road within 50m, traffic load within 200 m). The associations with NO2 were stronger
10 than those with NOx. The association was stronger in people with diabetes and asthma
11 compared to the rest of the cohort, but no difference in association was observed by
12 smoking or occupational exposure. COPD incidence was most strongly associated with
13 35-yr avg NO2, suggesting that long-duration, possibly lifetime exposure may be
14 associated with development of COPD.
15 Gan etal. (2013) evaluated a population-based cohort in Canada that included a 5-year
16 exposure period and a 4-year follow-up period. All residents aged 45-85 years who
17 resided in Metropolitan Vancouver, Canada, during the exposure period and did not have
18 known COPD at baseline were included in this study (N = 467,994). Five-year average
19 residential exposures to NO2 and NO as well as BC, PM2 5, and wood-smoke were
20 estimated using LUR models, incorporating changes in exposure over time due to
21 changes in residences. COPD incidence was ascertained from a hospital admissions
22 database and defined as admission during the follow-up period. Mortality data were also
23 studied and are discussed in Section 6.2. The Spearman correlations for NO2 with BC and
24 PM2 5 were respectively 0.39 and 0.47. The association of 5-year NO2 with COPD
25 hospital admission was null. The exposure period examined in this study was shorter than
26 that in Andersen et al. (2011) (i.e., 25-35 years).
27 In the ECRHS cohort, the association of NOx with prevalence of COPD and related
28 symptoms was investigated by two methods for assessing exposure to power
29 plant-specific emissions of NOx and SO2 (Amster et al.. 2014). NOx exposures (8-yr avg)
30 related to power plant emissions were estimated for subjects' residences (N = 2,244)
31 based on kriging ambient concentrations from 20 central site monitors downwind of the
32 power plant (source approach), and peak emission events (event approach) were defined
33 as 30-minute concentrations that exceeded 125 ppb NO2. Neither source-based nor
34 event-based power plant NOx emissions was associated with COPD prevalence.
35 Respiratory symptoms were associated with source-based NOx but not event-based NOx.
36 In a cross-sectional study, Wood et al. (2009) examined the association of outdoor air
37 pollution with respiratory phenotype (PiZZ type) in alpha 1-antitrypsin deficiency
January 2015 6-68 DRAFT: Do Not Cite or Quote
-------
1 (a-ATD) from the U.K. a-ATD registry. This deficiency leads to exacerbated responses
2 to inflammatory stimuli. In total, 304 PiZZ subjects underwent full lung-function testing
3 and quantitative high-resolution computed tomography to identify the presence and
4 severity of COPD emphysema. Annual average NO2 was estimated for subjects' homes
5 with dispersion models. NO2 was associated with improved gas transfer and less severe
6 emphysema. Similar associations were observed with SO2 and particles. In contrast, Os
7 was associated with worse gas transfer and more severe emphysema, albeit accounting
8 for only a small proportion of the lung function variability. NO2 was negatively
9 correlated with Os, which might explain NO2 associations with gas transfer and
10 emphysema severity. The dispersion model also may not well represent long-term NO2
11 exposures.
6.2.9 Summary and Causal Determination
12 There is likely to be a causal relationship between long-term NO2 exposure and
13 respiratory effects, based strongly on evidence integrated across disciplines for a
14 relationship with asthma development. There is continued epidemiologic evidence for
15 effects on decrements in lung function and partially irreversible decrements in lung
16 development in children. Other, more limited lines of evidence include NO2-related
17 increases in respiratory symptoms in children with asthma, chronic bronchitis/asthma
18 incidence in adults, COPD hospital admissions, and respiratory infection.
19 The conclusion of a likely to be causal relationship represents a change from the
20 "suggestive, but not sufficient, to infer a causal relationship" determined in the 2008 ISA
21 for Oxides of Nitrogen (U.S. EPA. 2008). The main difference in the evidence base in
22 this review compared to the 2008 ISA is recent epidemiologic evidence from several
23 longitudinal studies that indicate associations between asthma incidence in children and
24 long-term NO2 exposures estimated for outside children's homes. In contrast, the 2008
25 ISA for Oxides of Nitrogen reported inconsistent findings from a limited number of
26 cross-sectional studies that examined asthma prevalence. An additional uncertainty
27 identified in the 2008 ISA for Oxides of Nitrogen (U.S. EPA. 2008) was the potential for
28 NO2 to serve as an indicator for another combustion-related pollutant or mixture. Because
29 of the high correlations among traffic-related pollutants and limited examination of
30 copollutant confounding, the independent effects of long-term NO2 exposure could not be
31 clearly discerned at the time of the last review. While this uncertainty continues to apply
32 to the epidemiologic evidence across the respiratory effects examined, coherence of
33 epidemiologic evidence for asthma with the limited previous toxicological evidence for
34 both AHR and development of allergic responses, which are key events in the mode of
35 action for asthma development, provides support for an independent effect of long-term
January 2015 6-69 DRAFT: Do Not Cite or Quote
-------
1 exposure to NCh on development of asthma. The key evidence supporting the likely to a
2 causal relationship is detailed in Table 6-5 using the framework described in Table II of
3 the Preamble to this ISA.
6.2.9.1 Evidence on Development of Asthma
4 Multiple longitudinal, prospective and retrospective studies demonstrate associations
5 between higher ambient NC>2 concentrations measured in the first year of life, in the year
6 of diagnosis, or over a lifetime and asthma incidence in children. Results are consistent
7 across locations based on various study designs and cohorts (Table 6-1). Consistency
8 across studies in the use of questionnaires to ascertain parent report of
9 physician-diagnosed asthma, a best practice (Burr. 1992; Ferris. 1978). adds to the
10 strength of inference about associations with NC>2.
11 A pooled analysis across six birth cohorts relating NC>2 with ever asthma (OR: 1.48 [95%
12 CI: 1.06, 2.04] per 10 ppb increase) (Macintyre et al.. 2014a) is consistent with results
13 from individual studies. Individual studies relied upon various methods to estimate
14 exposure, and several characterize the exposure at the subject's residence using LUR
15 models that were demonstrated to well represent the spatial variability in the study areas.
16 Further, associations were observed across a range of ambient NO2 concentrations
17 [(Carlsten etal.. 20lie: Gehring etal.. 2010; Jerrett et al.. 2008; Cloughertv et al.. 2007):
18 Table 6-6] and resolutions or buffers used in the LUR models [e.g., 10 m (Carlsten et al..
19 201 Ic) to 500 m (Cloughertv et al.. 2007). In limited analysis of the
20 concentration-response relationship, results did not consistently indicate a linear
21 relationship in the range of ambient NC>2 concentrations examined (Carlsten et al.. 201 Ic;
22 Shima et al.. 2002). These studies did not conduct analyses to evaluate whether there is a
23 threshold for effects. Limited supporting evidence for chronic bronchitis incidence in
24 adults is provided in the ECRHS cohort, which prospectively evaluated individual NC>2
25 concentrations outdoor at the home using Palmes tubes (Sunyer et al.. 2007). For asthma
26 incidence in adults, recent studies in the ECRHS cohort report relationships with NCh
27 estimated from dispersion models.
28 Epidemiologic studies of asthma development in children have not clearly characterized
29 potential confounding by other traffic-related pollutants or mixtures pollutants [e.g., CO,
30 BC/EC, volatile organic compounds (VOCs)]. In the longitudinal studies of asthma
31 incidence, correlations with PIVbs and BC were often high (e.g., r = 0.7-0.96), and no
32 studies of asthma incidence evaluated copollutant models to address copollutant
33 confounding, making it difficult to evaluate the independent effect of NO2. Across studies
34 that examined both NO2 and PIVb 5, positive associations were observed between PM2 5
January 2015 6-70 DRAFT: Do Not Cite or Quote
-------
1 concentrations and asthma development, although the effects are smaller in magnitude
2 compared to those for NC>2 (Nishimura et al.. 2013; Clark et al.. 2010; McConnell et al..
3 2010). Correlations between NC>2 and PM2 5 were not reported in these three studies. One
4 cross-sectional study (Hwang and Lee. 2010). examined copollutant models but not with
5 traffic-related copollutants.
6 The uncertainty in the epidemiologic evidence base is partly reduced by the biological
7 plausibility provided by findings from experimental studies that demonstrate
8 NO2-induced effects on key events that are specific to the mode of action for
9 development of asthma (Figure 4-2). Though not observed in all studies, AHR and
10 airway remodeling was reported following 6-12 weeks of exposure of guinea pigs to
11 NO2 [1,000-4,000 ppb; (Kobayashi and Miura. 1995)]. Experimental studies also
12 indicate that short-term exposure repeated over several days and long-term NO2 exposure
13 can induce Th2 skewing/allergic sensitization by showing increased Th2 cytokines,
14 airway eosinophils, and IgE-mediated responses (Sections 4.3.5. 6.2.2.3). Findings for
15 short-term NC>2 exposure support asthma development by describing a potential role for
16 repeated exposures to lead to recurrent episodes of inflammation and allergic responses.
17 Epidemiologic evidence for NCh-related allergic responses is inconsistent but reported in
18 a few studies (Section 6.2.4).
19 Recurrent pulmonary inflammation and oxidative stress are key early events in the
20 pathophysiology of asthma (Figure 4-2). While the effects of long-term NC>2 exposure on
21 oxidative stress in toxicological studies are variable and transient (Section 6.2.2.3). there
22 is evidence supporting a relationship between short-term NC>2 exposure and increased
23 pulmonary inflammation. Evidence from controlled human exposure studies indicate
24 NO2-induced increases in neutrophils in healthy adults, and epidemiologic evidence also
25 points to associations between ambient NC>2 concentrations and increases in pulmonary
26 inflammation in healthy children and adults (Section 5.2.2.5). While epidemiologic
27 evidence overall is inconsistent in showing associations of long-term NCh exposure with
28 pulmonary inflammation (Section 6.2.2.3). a recent longitudinal study observed that
29 annual average NC>2 was associated with increases in eNO over time in children without
30 asthma (Berhane et al.. 2014). Such findings support a relationship between long-term
31 NO2 exposure and asthma development because the association was independent of
32 short-term change in NC>2 concentrations, and elevated eNO was associated with
33 increased risk of new onset asthma in the cohort. This limited evidence base for
34 NO2-related development of AHR and allergic responses and increases in pulmonary
35 inflammation combined with the consistent evidence for NO2-related development of
36 asthma in children describe a coherent and biologically plausible sequence of events by
37 which long-term NO2 exposure could lead to asthma development.
January 2015 6-71 DRAFT: Do Not Cite or Quote
-------
6.2.9.2 Evidence on Lung Function
1 Another line of evidence indicating a relationship between long-term NC>2 exposure and
2 respiratory effects includes multiple, longitudinal epidemiologic studies observing
3 associations between long-term NC>2 exposure and decrements in lung function and
4 partially irreversible decrements in lung development in children. Expanding on evidence
5 reviewed in the 2008 ISA (Figure 6-5). recent studies consistently demonstrate
6 associations with individual-level NC>2 exposure estimates based on time-activity patterns
7 and/or LUR (Urtnan etal.. 2014: Eenhuizenetal.. 2013: Molteretal.. 2013).
8 Associations also are observed with NO2 assessed from central sites. Some studies found
9 an NO2 concentration-dependent decrement in lung development (Rojas-Martinez et al..
10 2007a: Gauderman et al.. 2004) based on comparisons among communities or a
11 multipollutant model (with Os and PMio) which has uncertain reliability.
12 Potential confounding of long-term NC^-related decrements in lung function and lung
13 development by traffic-related copollutants has not been evaluated, although an
14 association was observed with adjustment for Os or PMio. Toxicological studies do not
15 clearly support epidemiologic findings. NC>2-induced changes in lung function were
16 inconsistently demonstrated in animal models [(Tepper et al.. 1993: Stevens etal.. 1988:
17 Lafumaetal.. 1987): Section 6.2.5.31. Long-term NC>2 exposure was observed to alter
18 lung morphology in adult experimental animals but not juvenile animals (Section 6.2.6).
19 but the changes observed do not appear to contribute to altered lung function or the
20 effects observed in epidemiologic studies.
6.2.9.3 Evidence on Respiratory Symptoms
21 Several longitudinal studies consistently demonstrate increases in respiratory symptoms
22 in children with asthma in relation to increased ambient NC>2 concentrations
23 (Section 6.2.3. Table 6-3). Associations were observed with NC>2 estimated from central
24 sites and NC>2 estimated for children's homes using LUR. Studies did not examine
25 whether associations of long-term NC>2 were independent of short-term exposure;
26 however, McConnell et al. (2003) assessed chronic symptoms as a daily cough for
27 3 months or congestion/phlegm for 3 months. Limited information from longitudinal
28 studies of indoor NC>2 support an association with respiratory symptoms in children with
29 asthma and adults with COPD (Belanger etal.. 2013: Hansel etal.. 2013). Findings for
30 indoor NCh exposure provide support for an independent relationship between NC>2 and
31 respiratory effects because NO2 may exist as part of a different air pollutant mixture
32 indoors than in the ambient air (Section 5.2.9.6). In limited analysis of copollutant
33 models, associations of NC>2 with respiratory symptoms in children persisted with
January 2015 6-72 DRAFT: Do Not Cite or Quote
-------
1 adjustment for the traffic-related copollutants PlVfc 5, EC, or OC. Potentially limiting
2 inference from these results, pollutants were measured from central sites, and correlations
3 with NCh were high in some cases (0.75 for PIVb 5, 0.92 for EC).
6.2.9.4 Evidence on Respiratory Infection
4 In the limited body of epidemiologic studies, findings do not consistently indicate
5 associations between long-term NC>2 exposure and respiratory infection. Findings in
6 infants are inconsistent, and associations with pneumonia hospital admissions in adults
7 could be due to short-term exposure. An evaluation of 10 European births demonstrated
8 associations of residential estimates of NCh exposure with parent report of
9 physician-diagnosed pneumonia and otitis media (Maclntyre et al.. 2014b). Adjustment
10 for PM2 5 or PIVb 5 absorbance produced associations for NO2 with wide 95% CIs,
11 limiting inferences about independent NC>2 associations (Section 6.2.7.1). The strongest
12 evidence for effects of long-term NC>2 exposure (500-2,000 ppb for 1 month up to 1 year)
13 in toxicological studies indicates increased respiratory infection (Section 6.2.7.2). NCh
14 exposure has been observed to increase mortality in rodents and squirrel monkeys
15 following bacterial challenge (Miller et al.. 1987; Henry et al.. 1970). and findings for
16 alterations to alveolar macrophage function and morphology describe key events in the
17 underlying mode of action (Gregory et al.. 1983; Aranyi et al.. 1976).
6.2.9.5 Analysis of Potential Confounding by Traffic-Related
Copollutants
18 Potential confounding of long-term NCh associations with respiratory effects by
19 traffic-related copollutants has been examined to a limited extent, particularly in
20 longitudinal analyses of asthma incidence. In longitudinal cohorts of children, copollutant
21 models were analyzed for chronic bronchitic symptoms in the CHS (McConnell et al..
22 2003) and lung function and respiratory infection in ESCAPE (Maclntyre et al.. 2014b;
23 Gehring et al.. 2013). Copollutants evaluated include EC, OC, PIVb 5 absorbance, and
24 PM2.5. The results ranged from persistence of the NCh effect or only modest reduction of
25 NCh effect to attenuation of NCh effect. In some cases, correlations with NCh were
26 moderate (r = 0.37-0.46 for PM2 5, 0.52 for PM2 5 absorbance, and 0.58 for OC).
27 However, high correlations often were reported (r = 0.72-0.80 for PlVfc 5 or 0.75-0.92 for
28 PM2 5 absorbance or EC). Although some studies support an independent association for
29 NO2, inconsistencies in the evidence base and limited analysis of the array of potential
30 confounding traffic-related copollutants, leave uncertainty in the epidemiologic studies in
January 2015 6-73 DRAFT: Do Not Cite or Quote
-------
1 disentangling the independent effect of NO2 from other traffic-related pollutants or
2 mixtures.
6.2.9.6 Conclusion
3 Taken together, recent epidemiologic studies and previous experimental studies provide
4 evidence that there is likely to be a causal relationship between long-term NO2 exposure
5 and respiratory effects (Table 6-5). The strongest evidence is provided by the consistent
6 findings for associations between NC>2 exposure and increases in asthma incidence and
7 decrements in pulmonary function in children, particularly for NCh exposures estimated
8 for children's homes. Experimental evidence indicating AHR induced by long-term NCh
9 exposure, and development of an allergic phenotype with repeated short-term and
10 long-term NC>2 exposure provides a pathophysiological basis for the effects of long-term
11 NO2 exposure on the development of asthma and reduces previous uncertainty related to
12 the independent effect of NC>2. However, because the experimental evidence is limited,
13 particularly for long-term exposure, there remains some uncertainty regarding an
14 independent effect of long-term NC>2 exposure on asthma. Indoor studies are limited in
15 number but support associations of respiratory symptoms with indoor NC>2, which may
16 exist as part of a different pollutant mixture than in the ambient air. In addition, other
17 studies that characterized the concentration-response for the relationship between NO2
18 and an array of respiratory effects ranging from respiratory symptoms and airway
19 resistance in children to chronic bronchitis and asthma in adults, generally observed a
20 linear relationship. Copollutant models with traffic-related copollutants such as PM2 5 and
21 PM2 5 absorbance provide mixed evidence for respiratory symptoms and lung function
22 decrements to inform an independent relationship for NC>2. Potential confounding by
23 other traffic-related copollutants is unexamined, and largely unavailable for studies of
24 asthma in children. Overall, the consistent epidemiologic and consistent but limited
25 experimental evidence for development of asthma is sufficient to conclude that there is
26 likely to be a causal relationship between long-term NO2 exposure and respiratory effects.
January 2015 6-74 DRAFT: Do Not Cite or Quote
-------
Table 6-5 Summary of key evidence for a likely to be a causal relationship
between long-term nitrogen dioxide (NO2) exposure and respiratory
effects.
Rationale for Causal
Determination3
Key Evidence13
Key References'3
NO2 Concentrations
Associated with Effects0
Asthma Development
Consistent epidemiologic
evidence from multiple,
high quality studies with
relevant NO2
concentrations
Consistent evidence for
increases in asthma
incidence in diverse
cohorts of children in U.S.,
Europe, Canada, and Asia.
Asthma ascertainment by
parental report of doctor
diagnosis.
Carlsten etal. (2011 c),
Clouqhertv et al. (2007).
Gehrinq etal. (2010).
Jerrett et al. (2008),
Shimaetal. (2002)
Weak evidence: Ranzi
etal. (2014)
Section 6.2.2.1. Table 6-1.
Figure 6-1
Means across studies of
LUR: 13.5, 17.3, 27.5 ppb
Upper percentiles: 75th:
15.1, 15.4 ppb
Max: 69.4 ppb
Range in mean residential
NO2 measurements across
communities:
9.6 to 51.3 ppb
Range in mean central site
NO2 across communities:
7.3-31.4 ppb
Supporting evidence for
asthma incidence or
chronic bronchitis in the
ECHRS cohort of adults.
Jacquemin et al. (2009b),
Modiq et al. (2009),
Sunveret al. (2006)
Section 6.2.2.2
Consistent evidence for
NO2 metrics with lower
potential for exposure
measurement error
In children, asthma
associated with residential
NO2 estimated using well
validated LUR models or
by monitoring.
Carlsten et al. (2011 c).
Clouqhertv et al. (2007).
Gehrinq etal. (2010),
Jerrett et al. (2008)
Section 6.2.2.1
Uncertainty regarding
potential confounding by
traffic-related copollutants
When reported,
correlations with PM2.5 and
EC often were high
(r= 0.7-0.96). No
copollutant models
analyzed.
Associations found with
adjustment for SES, family
history of asthma, smoking
exposure, housing
characteristics, and
presence of gas stove.
Table 6-1
January 2015
6-75
DRAFT: Do Not Cite or Quote
-------
Table 6-5 (Continued): Summary of key evidence for a likely to be a causal
relationship between long-term nitrogen dioxide (NO2)
exposure and respiratory effects.
Rationale for Causal
Determination3
Key Evidence13
Key References'3
NO2 Concentrations
Associated with Effects0
Coherence with respiratory
effects of short-term NO2
exposure
Limited epidemiologic
evidence for increases in
pulmonary inflammation in
healthy children and adults
with exposures assessed in
subjects' locations and
associations adjusted for
BC/EC, OC, PNC, or PM2.s.
Straketal. (2012).
Steenhofetal. (2013),
Linetal. (2011)
Section 5.2.7.4
Max for 5-h avg: 96 ppb
Means for 24-h avg across
seasons: 24.3-45.3 ppb
Evidence from controlled
human exposure studies
for increased airway
responsiveness in healthy
adults.
Section 5.2.7.1
1,000-2,000 ppb for 3 h
but not below
Limited and supporting
toxicological evidence at
relevant NO2 exposures
Increased AHR in guinea
pigs with long-term or
short-term NO2 exposure.
Kobavashi and Miura
(1995),
Kobavashi and Shinozaki
(1990)
1,000-4,000 ppb for
6-12 weeks
4,000 ppb for 7 days
Some Evidence for Key Events in Mode of Action
Allergic responses
Increased IgE-mediated
histamine release in mast
cells from rodents.
Fujimaki and Nohara
(1994)
Experimental findings for
development of Th2
phenotype with short-term
NO2.
Pathmanathan et al.
(2003),
Ohashietal. (1994)
Inconsistent epidemiologic
evidence for allergic
diseases or responses in
children with long-term
exposure.
Section 6.2.4.1
4,000 ppb for 12 weeks
2,000 ppb over
4 consecutive days;
3,000 ppb for 2 weeks
Airway remodeling
Increased airway
resistance with airway
hyperresponsiveness in
guinea pigs.
Kobavashi and Miura
(1995)
1,000-4,000 ppb for
6-12 weeks
January 2015
6-76
DRAFT: Do Not Cite or Quote
-------
Table 6-5 (Continued): Summary of key evidence for a likely to be a causal
relationship between long-term nitrogen dioxide (NO2)
exposure and respiratory effects.
Rationale for Causal
Determination3
Key Evidence13
Key References'3
NO2 Concentrations
Associated with Effects0
Inflammation
Increases in lymphocytes,
PMNs, in rats with
long-term exposure.
Increases in PMNs in
healthy adults with
repeated short-term
exposure.
Kumae and Arakawa
(2006),
Blomberq et al. (1999)
Longitudinal changes in
eNO in children
independent of asthma
status.
Berhaneetal. (2014)
500 or 2,000 ppb for
5 weeks in rats
2,000 ppb for 4 h/day for
4 days
Inconsistent associations in
other studies with exposure
assessment by LUR and
central site.
Liuetal. (2014a)
Residential by LUR:
Mean 11.3
95th: 15.3 ppb
Oxidative stress
Varying and transient
effects on antioxidant
levels and enzyme activity.
Avaz and Csallanv (1978),
Gregory et al. (1983),
Saqai et al. (1984)
400, 1,000, 5,000 ppb for
6 weeks to 18 mo
January 2015
6-77
DRAFT: Do Not Cite or Quote
-------
Table 6-5 (Continued): Summary of key evidence for a likely to be a causal
relationship between long-term nitrogen dioxide (NO2)
exposure and respiratory effects.
Rationale for Causal
Determination3
Key Evidence13
Key References'3
NO2 Concentrations
Associated with Effects0
Severity of Asthma
Consistent epidemiologic
evidence but uncertainty
regarding NO2 independent
effects
Consistent evidence for
increases in respiratory
symptoms in children with
asthma.
Exposure assessment by
central site measurements
and LUR.
McConnell et al. (2003),
Gehrinq et al. (2010)
Figure 6-5, Table 6-3.
Residential NO2 by LUR:
Mean: 13.5 ppb
10th-90th percentile:
7.8-18.5 ppb
Central site:
Mean, Max 4-yr avg for
12 communities:
19.4, 38.0 ppb
Associations with
respiratory symptoms
remain robust with
adjustment for a
traffic-related copollutant:
PM2.5, EC, orOC. But,
analysis is limited and
based on central site
exposure assessment.
In limited analysis,
associations with
respiratory symptoms
remain robust with
adjustment for Os, SO2,
PMio-2.5, or PM-io.
McConnell etal. (2003),
Hwang and Lee (2010)
Table 6-3
Evidence for associations
between indoor NO2 and
respiratory symptoms in
children with asthma ages
5-10 yr; inconsistent
evidence in younger
children and infants.
Belanqeret al. (2013)
Mean daily indoor NCb:
10.6 ppb
75th: 12.5 ppb
January 2015
6-78
DRAFT: Do Not Cite or Quote
-------
Table 6-5 (Continued): Summary of key evidence for a likely to be a causal
relationship between long-term nitrogen dioxide (NO2)
exposure and respiratory effects.
Rationale for Causal
Determination3
Key Evidence13
Key References'3
NO2 Concentrations
Associated with Effects0
Lung Function and Development
Consistent epidemiologic
evidence from multiple,
high-quality studies but
uncertainty regarding NO2
independent effects
Epidemiologic evidence for
decrements in lung function
and partially irreversible
decrements in lung
development in children.
Gauderman et al. (2004),
Rojas-Martinez et al.
(2007a).
Molteretal. (2013).
Gehrinq et al. (2013),
Urmanetal. (2014).
Eenhuizen et al. (2013)
In limited analysis,
associations are
inconsistent with
adjustment for PlVh.s but
robust with adjustment for
PM-io or 03.
Residential NCb-PIVh.s
correlations vary across
cohorts. Pearson
r= 0.31-0.76.
Gehrinq et al. (2013),
Rojas-Martinez et al.
(2007b)
NO2 by LUR:
Means across
communities: 7.4-12.6 ppb
Overall study mean:
13.5 ppb, 75th: 15.4 ppb
Central site NO2 mean
across communities:
27.2-42.6 ppb
Uncertain relevance of
toxicological evidence
Changes in lung
morphology including
increases in edema,
hypertrophy of lung
epithelium, fibrotic changes
in adult not juvenile
animals. Uncertain
relevance to epidemiologic
findings.
Kubotaetal. (1987),
Havashietal. (1987)
500 ppb for 19 mo,
4,000 ppb for 9-27 mo
Respiratory Infection
Consistent toxicological
evidence
Increased mortality of mice
and monkeys with NO2
exposure and challenge
with bacterial or viral
infection.
Henry et al. (1970),
Ehrlich and Henry (1968),
Ehrlich(1980),
Miller etal. (1987)
500 ppb for 3 mo,
5,000 ppb for 2 mo,
200 ppb base plus daily
spike of 800 ppb for
16-52 weeks
Limited and inconsistent
epidemiologic evidence
Associations found with
physician-diagnosed
pneumonia, otitis media,
and croup in multicounty
European cohort study but
not consistently in other
studies.
Macintvre et al. (2014a)
Range in mean across
10 birth cohorts:
7.5-23.7 ppb
Limited evidence for key
events in mode of action
Increased macrophage
infiltration to lung tissue or
increased lymphocytes in
BAL fluid of experimental
animals.
Gregory etal. (1983)
5,000 ppb for 15 weeks
January 2015
6-79
DRAFT: Do Not Cite or Quote
-------
Table 6-5 (Continued): Summary of key evidence for a likely to be a causal
relationship between long-term nitrogen dioxide (NO2)
exposure and respiratory effects.
Rationale for Causal
Determination3
Key Evidence13
Key References'3
NO2 Concentrations
Associated with Effects0
COPD
Limited and inconsistent
epidemiologic evidence
Inconsistent evidence for
hospital admissions for
COPD in adults. Unclear
whether independent of
short-term exposure
effects.
Andersen et al. (2011),
Ganetal. (2013)
BALF = bronchoalveolar lavage fluid; BC = black carbon; COPD = chronic obstructive pulmonary disease; EC = elemental
carbon; eNO = exhaled nitric oxide; IgE = immunoglobulin E; LUR = land-use regression; NO2 = nitrogen dioxide; O3 = ozone;
OC = organic carbon; PM25 = particulate matter with a nominal aerodynamic diameter less than or equal to 2.5 |jm;
PM10 = particulate matter with a nominal aerodynamic diameter less than or equal to 10 |jm; PM10-2.s = particulate matter with a
nominal aerodynamic diameter less than or equal to 10 |jm and greater than a nominal 2.5 |jm; PMN = polymorphonuclear cell(s),
polymorphonuclear leukocyte; PNC = particle number concentration; SES = socioeconomic status; SO2 = sulfur dioxide;
Th2 = T-derived lymphocyte helper 2.
aBased on aspects considered in judgments of causality and weight of evidence in causal framework in Tables I and N. of the
Preamble.
""Describes the key evidence and references contributing most heavily, but not necessarily exclusively, to causal determination.
Where applicable, uncertainties and inconsistencies are described. References to earlier sections indicate where full body of
evidence is described.
°Describes the NO2 concentrations with which the evidence is substantiated (for experimental studies, below 5,000 ppb).
6.3 Cardiovascular and Related Metabolic Effects
6.3.1 Introduction
i
2
o
6
4
5
6
1
8
9
10
11
12
13
The 2008 ISA for Oxides of Nitrogen (U.S. EPA. 2008) concluded that "the available
epidemiologic and toxicological evidence was inadequate to infer the presence or absence
of a causal relationship" between cardiovascular effects and long-term NCh exposure.
This section updates the previous review with the inclusion of recent studies on the
cardiovascular and related cardiometabolic effects of NO2 and NOx exposure in humans,
animals, and cells. Data from individual studies can be found in summary tables at the
end of each section and an integrated summary of the evidence is presented in
Section 6.3.9.
At the completion of the 2008 ISA for Oxides of Nitrogen (U.S. EPA. 2008) one
epidemiologic study of the association of cardiovascular disease (CVD) with long-term
exposure to NO2 was available for inclusion in the document. Miller et al. (2007) studied
65,893 post-menopausal women (50-79 years old) without previous CVD from 36 U.S.
metropolitan areas. Exposures to air pollution were estimated by assigning the annual
January 2015
6-80
DRAFT: Do Not Cite or Quote
-------
1 (2000) mean air pollutant concentration measured at the monitor nearest to the subject's
2 five-digit residential ZIP Code centroid. In single-pollutant models, PM2 5 showed the
3 strongest associations with the CVD events [myocardial infarction (MI),
4 revascularization, angina, congestive heart failure, coronary heart disease (CHD) death],
5 followed by SCh. The association of NC>2 with overall CVD events was 1.04 (95% CI:
6 0.96, 1.12) per 10-ppb increase and NC>2 not associated with CVD events when the data
7 set was restricted to those with nonmissing exposure data. The available animal
8 toxicological evidence was limited to studies of changes in HR, vagal response, and
9 alterations in specific hematological parameters [e.g., hematocrit, hemoglobin,
10 ethrythrocytes; (U.S. EPA. 2008. 1993)1.
11 Large, prospective studies with consideration of potential confounding and other sources
12 of bias are emphasized in this section (see Table 5-1 for study evaluation guidelines). The
13 exposure assessment method was also an important consideration in the evaluation of
14 long-term exposure and cardiovascular and related cardiometabolic health effects, given
15 the spatial variability typically observed in ambient NC>2 concentrations (Section 2.5.3).
16 Exposure assessment was evaluated drawing upon discussions in Section 3.2 and
17 Section 3.4.5. In general, LUR model predictions have been found to correlate well with
18 outdoor NO2 concentration measurements (Section 3.2.1.1). A select number of recent
19 studies have employed exposure assessment methods such as LUR to represent the spatial
20 variability of NC>2. Statistics indicating the correlation between predicted and measured
21 NC>2 concentrations are presented where available.
22 Several recent epidemiologic studies report positive associations of NO2 and NOx
23 exposure with heart disease, diabetes, stroke, and hypertension. The body of evidence is
24 strongest for heart disease and diabetes and includes several large, longitudinal studies
25 with consideration of multiple potential confounding factors including age, sex, BMI,
26 smoking, and pre-existing conditions (Tables 6-6 and 6-7). Some of these studies
27 employed validated exposure assessment methods such as LUR, which were
28 demonstrated to capture the spatial variability of NC>2 concentration. The extent to which
29 the studies inform the independent effect of NC>2 exposure through their consideration of
30 correlated copollutants (i.e., CO, BC, PM2 5) and noise is discussed in the section. A
31 small number of experimental animal studies examining the effect of NO2 on oxidative
32 stress and the progression of vascular disease provided limited support for the biological
33 plausibility of the effects observed in the epidemiologic studies.
January 2015 6-81 DRAFT: Do Not Cite or Quote
-------
6.3.2 Heart Disease
i
2
3
4
5
6
7
Several studies published since the 2008 ISA for Oxides of Nitrogen examine the
association of long-term NOx exposure and heart disease. Although the evidence from
the epidemiologic studies is not entirely consistent, several prospective studies and/or
studies with exposure assessment strategies designed to capture the spatial variability of
NO2 report positive associations. Most studies adjust for a wide array of potential
confounders such as age, sex, BMI, smoking, and pre-existing conditions (Table 6-6) but
uncertainty remains regarding the extent to which findings can be explained by correlated
copollutant exposures and noise.
Table 6-6 Epidemiologic studies of the association of long-term exposure to
nitrogen dioxide (NO2) or the sum of nitric oxide and NO2 (NOx) with
heart disease.
Study
Cohort Location
Study Period
Mean (ppb)
Exposure
Assessment
Effect Estimates (95% Cl)
Cesaroni et al. (2014)
ESCAPE Project, NC>2:
11 Cohorts,
5 countries in
Europe
2008-2012
Range of
means
across
cohorts: 4.2
(3.2-5.8) to
31.9
(22.3-40.9)
Annual avg NO2,
NOx, LUR,
40 monitoring sites,
linked to geocoded
addresses.
Coronary events
NO2HR: 1.06(0.96, 1.16)
per 10-ppb increase NO2
NOxHR: 1.01 (0.98, 1.05)
per 20 ug/m3 NOx*
Covariate adjustment:
marital status, education,
occupation, smoking status
duration and intensity, and
Copollutant adjustment:
none
Gan et al. (2011) Population based
cohort in
Vancouver,
Canada)
1999-2002
N = 452,735
NO2:
Mean:
16.3
IQR: 4.5
NO:
Mean:
26.1
IQR: 10.8
LUR, 5-yr avg
concentration (NO2
and NO,
1995-1 998) and
4 yr avg
concentration
(1999-2002; 10-m
spatial resolution).
Concentrations
assigned to postal
code centroids
(typically ~1 city
block in urban areas
and larger in less
populated areas).
CHD hospitalization (ICD-9
410-414)
RR (NO2): 0.93 (0.89, 0.98)
RR(NO): 0.96(0.92, 1.00)
per 10 ppb NO2 and NO
Covariates: age, sex,
pre-existing diabetes,
COPD, hypertension, and
QPQ
O^O.
Copollutant adjustment:
none
January 2015
6-82
DRAFT: Do Not Cite or Quote
-------
Table 6-6 (Continued): Epidemiologic studies of the association of long-term
exposure to nitrogen dioxide (NO2) or the sum of nitric
oxide and NO2 (NOx) with heart disease.
Cohort Location
Study Study Period
Beckerman et al. (2012) Cohort of
pulmonary
patients in
Toronto, Canada
1992-1999
N= 2,414
Rosenlund et al. (2009a) SHEEP Study in
Stockholm,
Sweden
1985-1996
N = 24,347
cases, 276,926
controls
Hartetal. (2013) NHS
11 States in the
U.S.
1990-2008
N = 121,700
Exposure
Mean (ppb) Assessment Effect Estimates (95% Cl)
NO2: LUR, avg of fall 2002 IHD prevalence (ICD-9
Median- and spring 2004 NO2 412-414)— old Ml, angina
229 concentration, or other IHD
_, assigned atthe RR: 1.24 (1.01, 1.53) per
IQR: 4'° postal code centroid 10 ppb
(typically 1 block, or Covariate adjustment:
single building and gender, age, pack-yr
larger in less smoking, BMI, deprivation
populated areas). indeX] and diabetes.
RR 1.17(1.01, 1.36) per
10 ppb after adjustment for
covariates above plus Os
and PM2.5 (in same model).
5th-95th: 5-yr avg NO2 First nonfatal Ml
15.9 cases concentration OR: 0.96 (0.93, 1.00) per
Median assessed by 10 ppb
(cases): 6.9 dispersion modeling covariates: age, sex,
Median of traffic-related calendar yr, and SES.
Median emissions 25-m
(controls): reso|ution inner ^
6-3 1 00-m urban, 500-m
regional/countryside.
Concentrations
assigned to
residential address.
NR Dispersion model to Incident Ml
predict annual avg HR: 1 22 (0.99, 1 .50) per
(2000) NO2 1-ppb increase in NO2
concentration, between addresses
assigned to „ . . ,. .
residential address Covariate adjustment: BMI,
Main results were for Phfical *«"**• h,ealthV
traffic proximity. diet score, a cohol,
hypercholesterolemia, high
blood pressure, diabetes,
family history of Ml, smoking
status, mental health status,
father's occupation, marital
status, husband's
education, education level,
employment, and median
income/home value.
Copollutants adjustment:
none
January 2015
6-83
DRAFT: Do Not Cite or Quote
-------
Table 6-6 (Continued): Epidemiologic studies of the association of long-term
exposure to nitrogen dioxide (NO2) or the sum of nitric
oxide and NO2 (NOx) with heart disease.
Study
Lipsettetal. (2011)
Atkinson et al. (2013)
de Kluizenaaret al. (2013)
Dong etal. (2013a)
Cohort Location
Study Period
CIS Cohort in
CA.
June
1996-Dec2005
N = 124,614
National GP
Patient Cohort in
the U.K.
2003
N = 836,557
Eindhoven,
Netherlands
1991-2003
N = 18,213
33 communities
in 11 districts of
3 cities in
Liaoning
Province, China
2006-2008
N — OA RA^L
— Z.*i,O*iO
Mean (ppb)
NO2:
Mean
(5th-95th):
33.59
IQR: 10.29
NOx:
Mean: 95.6
IOP- ^ft Tl
|Vo4r\. oo.o I
Mean (SD):
12.0
IQR:
5.7 ppb
NO2
5th-95th:
7.5
NO2:
Mean: 18.7
Median:
17.5
IQR: 4.8
Exposure
Assessment
Gridded pollutant
surface (250-m
spatial resolution)
developed using
IDW fixed site
monitor
concentrations
(1995-2005) linked
to geocoded
residential address.
Defined
representative range
of 3-5 km
(neighborhood and
regional monitors,
respectively) for NOx
and NO2 to account
for spatial variability
of pollutant.
Annual average NO2
concentration (2002)
derived from
dispersion models
identifying all known
emissions sources
(1 by 1 km
resolution), linked to
residential
post-codes.
Concentrations
linked to post-code
centroids that
typically include
13 residential
addresses.
Dispersion model
1 x 1 km resolution,
linked to residential
address.
District-specific
3-yr avg NO2
concentrations for
communities within
1 km of an air
monitoring station
(selected to
maximize intra- and
inter-city gradients).
Effect Estimates (95% Cl)
Ml incidence
NOx: HR1.01 (0.91, 1.11)
per 20 ppb NOx
NO2: HR 1.06(0.88, 1.27)
per 10 ppb NO2
Covariate adjustment: age,
race, smoking second-hand
smoke, BMI, lifetime
physical activity, nutritional
factors, alcohol, marital
status, menopausal status,
hormone therapy,
hypertension medication
and aspirin, and family
history of Ml/stroke.
Copollutant adjustment:
none
Ml incidence
HR: 0.97(0.90, 1.04)
Arrhythmia incidence
HR: 0.98(0.91, 1.04)
Heart failure incidence
HR: 1.11 (1.02, 1.21)
per 10 ppb
Covariates: age, sex,
smoking, BMI, diabetes,
hypertension, and index of
multiple deprivation.
Copollutant adjustment:
none
Self-reported CVD
OR: 1.04(0.60, 1.82)
per 10-ppb increase
Copollutant adjustment:
none
January 2015
6-84
DRAFT: Do Not Cite or Quote
-------
Table 6-6 (Continued): Epidemiologic studies of the association of long-term
exposure to nitrogen dioxide (NO2) or the sum of nitric
oxide and NO2 (NOx) with heart disease.
CohortLocation Exposure
Study Study Period Mean (ppb) Assessment Effect Estimates (95% Cl)
Miller et al. (2007) WHI Cohort in 36 NR Annual avg (2000): Incident CVD events
U.S. cities nearest monitor to HR: 1.04 (0.96, 1.12)
1994-1998 residence ZIP code per 10 ppb
centroid (overall Covariates: age, ethnicity,
effect based on education, household
intra- and inter-city jncome, smoking, diabetes,
gradients). hypertension, systolic blood
pressure, BMI, and
hypercholesterolemia.
Copollutant adjustment:
none
BMI = body mass index; CBVD = cerebrovascular disease; CHD = coronary heart disease; Cl = confidence interval;
COPD = chronic obstructive pulmonary diesease; CIS = California Teachers Study; CVD = cardiovascular disease;
ESCAPE = European Study of Cohorts for Air Pollution Effects; GP = general practice; HR = hazard ratio; ICD = international
Classification of Diseases; IDW = inverse distance weighting; IHD = ischemic heart disease; IQR = interquartile range;
LUR = land-use regression; Ml = myocardial infarction; NHS = Nurses Health Study; NO = nitric oxide; NO2 = nitrogen dioxide;
NOx = sum of NO and NO2; NR = no quantitative results reported; O3 = ozone; OR = odds ratio; PM25 = particulate matter with a
nominal aerodynamic diameter less than or equal to 2.5 |jm; RR = risk ratio(s), relative risk; SES = socioeconomic status;
SHEEP = Stockholm Heart Epidemiology Program; WHI = Women's Health Initiative.
*NOX results that are originally reported in jxg/m3 are not standardized if the molecular weight needed to convert to ppb is not
reported.
1 Cesaroni et al. (2014) reported an increased risk of incident coronary events of 1.06
2 (95% Cl: 0.96, 1.16) per 10-ppb increase in NC>2. This large study of 11 cohorts from
3 5 countries used LUR to assign exposure at each participant's residence. Authors
4 reported good performance of exposure models based on their comparison of predicted
5 estimates and concentrations measured at 40 sites (R2 > 0.61). A study of pulmonary
6 patients in Toronto, Canada (Beckerman et al.. 2012) reported an increased risk of 1.17
7 (95% Cl 1.01, 1.36) per 10-ppb increase in NC>2 with ischemic heart disease (HD)
8 prevalence after adjustment for individual covariates as well as simultaneous adjustment
9 for Os and PM2 5. In this study, LUR was also used to estimate NO2 concentrations, which
10 were assigned at the post-code centroid level (typically one block area or specific
11 building in this study area). In another prospectively designed study, Gan etal. (2011)
12 examined the association of long-term exposure to BC, PIVb 5, NCh, and NO with CHD
13 hospitalization and mortality among participants (45-85 year-olds) residing in
14 Vancouver, Canada enrolled in the universal health insurance system. In this study, LUR
15 was used to predict NO2 concentrations at a resolution of 10 m. These predicted
16 concentrations were adjusted using factors derived from regulatory monitoring data and
17 then linked to each participant's postal code of residence (typically one block or specific
18 building in this study area). After adjustment for potential confounders, NC>2 and NO
19 were inversely associated with CHD hospitalization (HR: 0.93 [95%CI: 0.89, 0.98] and
January 2015 6-85 DRAFT: Do Not Cite or Quote
-------
1 HR: 0.96 [95% CI: 0.92, 1.00] per 10 ppb); however, positive associations of NO2 and
2 NO with CHD mortality were observed (Section 6.5.2).
3 Several other studies characterized NO2 exposure using IDW estimates of concentration
4 from central site monitors or dispersion models that captured a range of spatial
5 resolutions. Uncertainties associated with these models are described in detail in
6 Sections 3.2.2 and 3.2.3. Briefly, estimates derived from IDW monitor concentrations
7 may not capture the true variability in NO2 concentration from local sources if the
8 monitor coverage is not adequately dense, thus underestimating concentrations. Biases in
9 dispersion model output can occur in either direction and depend on the complexity of the
10 topography, meteorology, and sources that are modelled.
11 Lipsett etal. (2011) determined the association of incident MI with long-term exposure to
12 NO2, NOx, other gases (CO, Os, 802) and PM in a prospective study. These authors
13 followed a cohort of California public school teachers aged 20-80 years old
14 (N = 124,614). Each participant's geocoded residential address was linked to a pollutant
15 surface with a spatial resolution of 250 m, which was determined by IDW interpolation
16 of pollutant concentrations measured at fixed site monitors within a representative range
17 of 3-5 km. Those living outside the radial range for which the monitor was intended to
18 provide representative data were excluded from the analysis. The authors observed a
19 positive association between NO2 and incident MI (HR: 1.06 [95% CI: 0.88, 1.27] per
20 10 ppb). In a study of women enrolled in the Nurses Health Study (NHS), Hart etal.
21 (2013) reported an increased risk of incident MI associated with living consistently near
22 sources of traffic. Although the exposure assessment for the main analyses in this study
23 was based on distance to roadway, the authors used a dispersion model to predict the
24 change in NO2 concentration among those who moved from one address to another. They
25 observed an increased risk of incident MI in association with current NO2 compared with
26 NO2 concentration at the previous address (Table 6-6).
27 Rosenlund et al. (2009a) conducted a case-control study of first MI using the Swedish
28 registry of hospital discharges and deaths for Stockholm County and randomly selected
29 population-based controls. Predicted 5-yr avg NO2 concentrations were determined and
30 linked to each participant's geocoded address using dispersion models. The resolution of
31 the predicted concentrations corresponded to 500 m in the countryside, 100 m in urban
32 areas, and 25 m in the inner city. Five-year average NO2 concentration was associated
33 with fatal MI (OR: 1.14 [95% CI: 1.09, 1.19] per 10 ppb) but not with nonfatal MI (OR:
34 0.96 [95% CI: 0.93, 1.00] per 10 ppb). CO and PMio were also associated with fatal cases
35 of MI in this population. Atkinson et al. (2013) examined the association of incident
36 cardiovascular disease with NO2. These authors studied patients (aged 40-89 years)
37 registered with 205 general practices across the U.K. Predicted annual average NO2
January 2015 6-86 DRAFT: Do Not Cite or Quote
-------
1 concentrations within 1 km by 1 km grids, estimated using dispersion models, were
2 assigned to participants based on their residential postal code. Cardiovascular disease
3 outcomes included in the analysis were MI, arrhythmias, and heart failures. Authors
4 reported a positive association between NO2 and heart failure in fully adjusted models
5 (HR: 1.11 [95% CI: 1.02, 1.21] per 10 ppb). Incident MI and arrhythmia were not
6 associated with NO2 concentration in this analysis. A similar pattern of findings were
7 observed for the associations between PM and these outcomes (associations with CHD
8 and MI were null while the association of PMio with heart failure was increased).
9 de Kluizenaar et al. (2013) assigned NO2 exposure to participants' residential addresses,
10 based on dispersion modelling with a 1 * 1 km resolution, reported an association of NO2
11 with CHD and cerebrovascular disease (CBVD) hospitalizations that was robust to
12 adjustment for individual level covariates and noise (HR 1.18 [95% CI: 0.9, 1.48] per
13 10 ppb increase in NO2). As discussed above, the only study available for inclusion in the
14 previous assessment reported a null association between NO2 and incident CVD events
15 (Miller etal.. 2007) comparing annual average concentration assigned at the ZIP code
16 centroid level to study participants across 36 U.S. cities. Dong etal. (2013 a) reported a
17 small, imprecise increase in the prevalence of self-reported CVD comparing 3-yr avg
18 concentrations for communities within 1 km of an air monitoring station across 3 cities in
19 Liaoning, China (OR: 1.04 [95%CI: 0.60, 1.82] per 10 ppb increase NO2).
20 Overall, several epidemiologic studies, including some large studies with prospective
21 designs, adjustment for known risk factors for cardiovascular disease such as age, sex,
22 BMI, and smoking, and use of exposure assessment methods designed to achieve high
23 spatial resolution (Cesaroni et al.. 2014; Beckerman et al.. 2012). provide evidence that
24 long-term exposure to NO2 is associated with the risk of heart disease. Although positive
25 associations between MI and CHD were not observed consistently across studies, some
26 studies reporting null or inverse associations with CHD or MI morbidity reported
27 increased risk of mortality from these causes (Ganet al.. 2011; Rosenlund et al..
28 2009a)and a positive association with heart failure was reported by Atkinson et al.
29 (2013). The few studies that accounted for confounding by PM2 5 (Beckerman et al..
30 2012) or noise (de Kluizenaar et al.. 2013) provide limited evidence that estimates are
31 robust to adjustment for these factors. In general these studies were not designed to
32 distinguish the independent effect of NO2 from the effects of other traffic-related
33 pollutants (e.g., BC, EC, CO), noise, or stress.
January 2015 6-87 DRAFT: Do Not Cite or Quote
-------
6.3.3 Diabetes
1 There were no epidemiologic studies examining the association of NO2 exposure with
2 diabetes or insulin deficiency in the 2008 ISA for Oxides of Nitrogen. Recent large
3 prospective studies using exposure assessment methods designed to achieve high spatial
4 resolution, provide some evidence of an association (Table 6-7). However, studies overall
5 have not distinguished an independent effect of NO2 on diabetes.
6 Coogan et al. (2012) examined the association of long-term NOx exposure with incident
7 diabetes among black women residing in Los Angeles, CA. An LUR model was applied
8 to estimate exposure at each participant's residential address and cross validation of the
9 model performance produced an R2 value of approximately 92%. An increased risk of
10 1.25 (95% CI: 1.07, 1.46) per interquartile range (IQR) increase (12.4 j^g/m3) in NOx
11 after adjustment for a wide array of potential confounders including traffic-related noise
12 exposure. Negligible attenuation in the effect estimate for NOx was reported after
13 adjustment for PM25. The correlation between NOx and PM2 5 concentrations (PM2 5
14 estimated using kriging), was low (r = 0.27) but correlations between NOx and NO2 or
15 other traffic-related copollutants were not reported. An increased risk of type II diabetes
16 in association with LUR estimates of NO2 was reported among older adult women living
17 in the Ruhr district of West Germany (HR: 1.55 [95% CI: 1.20, 2.00] per 10 ppb increase
18 in NO2; (Kramer et al.. 2010)). In this study, nondiabetic women (age 54-55) were
19 followed over 16 years (1990-2006), and alternate NO2 exposure assessment methods
20 (mean monitor concentration and emission inventory-based methods) were compared.
21 Relative risks determined by these alternative methods were smaller and less precise
22 compared to those obtained using validated LUR models, which were reported to explain
23 92% of the variance in NO2. Although diabetes status was self-reported in this study, a
24 validation study comparing self-reported diabetes from the questionnaire to answers
25 obtained during a clinical exam interview indicated 99% concordance. In an analysis of a
26 subgroup (n = 363) of these women, Teichert et al. (2013) observed positive associations
27 of NO2 and NOx exposure (estimated for the period 10-20 years prior to the baseline
28 exam) with impaired glucose metabolism (IGM). Risk estimates were robust to
29 adjustment for an array of biomarkers of subclinical inflammation.
30 In a large study of prevalent diabetes ascertained by self-report among randomly selected
31 adults (ages 18-65 years), which also used LUR, Eze et al. (2014) reported a positive
32 association (OR: 1.43 [95% CI: 1.08, 1.90] per 10 ppb increase in NO2). Noise was
33 included among the array of potential confounders for which the final model was
34 adjusted. The association of NO2 with diabetes was attenuated after adjustment for PMio
35 in a copollutant model (OR: 1.07 [95% CI: 0.72, 1.61] per 10 ppb increase in NO2). In
36 another study of diabetes prevalence using LUR, Brook et al. (2008) reported an
January 2015 6-88 DRAFT: Do Not Cite or Quote
-------
1 increased risk in the prevalence of diabetes mellitus (Types I and II) of 1.48 (95% CI:
2 1.00, 2.19) per 10 ppb increase in NO2 among female respiratory patients in two
3 Canadian cities. NO2 exposure was not associated with diabetes in male patients,
4 however. Prevalent diabetes was not associated with NO2 exposure estimated by LUR in
5 a semirural population in the Netherlands (Dijkema et al.. 2011).
6 A large prospective study examined the association of NO2 exposure with diabetes
7 incidence among participants of the Danish Diet, Cancer, and Health Cohort (Andersen
8 etal.. 2012c). A validated dispersion model (r > 0.75 for correlation between measured
9 and predicted 1/2-yr avg NO2 concentration; (Hertel et al.. 2014)) was used to assign
10 mean NO2 concentration since 1971 based on residential address history. No association
11 between NO2 exposure and diabetes was observed in fully adjusted models (HR: 1.00
12 [95% CI: 0.89, 1.12] per 10 ppb increase in NO2); however, after restricting the analyses
13 to confirmed cases of diabetes, a weak positive association was observed (HR: 1.04 [95%
14 CI: 1.00, 1.08] per 10 ppb increase in NO2). Long-term exposure to traffic noise was not
15 associated with a higher risk of diabetes in this population (S0rensen etal.. 2013).
16 Another study designed to evaluate the association of long-term exposure to aircraft noise
17 with diabetes found that, although associations with metabolic outcomes such as waist
18 circumference were observed, no association of Type II diabetes or BMI with noise was
19 present. (Eriksson et al.. 2014).
20 The association of NO2 concentration with insulin resistance in children was examined in
21 one study. In a study of Homeostatic Model Assessment (HOMA) of insulin resistance, a
22 metric derived from blood glucose and serum insulin measurements among 10 year old
23 children (n = 397), Thiering et al. (2013) reported that NO2 was associated with a 28%
24 increase in insulin resistance (95% CI: 6.7, 51.7%) per 10 ppb NO2.
25 Generally consistent associations of NO2 (Teichert et al.. 2013; Andersen et al.. 2012c;
26 Kramer et al.. 2010) as well as NOx (Coogan et al.. 2012) with diabetes or impaired
27 insulin metabolism are reported in prospective studies using LUR to assign exposure.
28 Associations of NO2 with prevalent diabetes among females and respiratory patients are
29 reported in some (Eze etal.. 2014; Brook et al.. 2008) but not all studies (Dijkema et al..
30 2011). Findings regarding the potential for noise exposure to confound observed
31 associations of NO2 and NOx with diabetes are limited. Overall, studies have not
32 distinguished an independent effect of NO2 from other traffic related exposures on
33 diabetes.
January 2015 6-89 DRAFT: Do Not Cite or Quote
-------
Table 6-7 Epidemiologic studies of the association of long-term exposure to
nitrogen dioxide (NO2) or the sum of nitric oxide and NO2 (NOx) with
cardiometabolic disorders.
Cohort Location
Study Study Period
Mean (ppb) Exposure Assessment
Effect Estimates (95% Cl)
Cooqan et al.
(2012)
BWHS Cohort in NOx(|jg/m3):
Los Angeles, CA Mean: 43.3
1995-2005 Median: 41.6
N = 3,236 IQR: 12.4
LUR to estimate
exposure at
participant's residences
(summer and winter
measurements were
taken at 183 sites in Los
Angeles to estimate an
annual mean).
Diabetes [per IQR (12.4 ug/m3)
increase in NOx]:
IRR: 1.25(1.07,1.46)
Covariate adjustment: age, BMI, yrof
education, income, number of people
in the household, smoking, drinks per
week, h/week of physical activity,
neighborhood socioeconomic status
score, and family history of diabetes.
IRR: 1.24 (1.05, 1.45), adjusted for
above covariates plus PlVhs
Kramer et al.
(2010)
SALIA Cohort in
Ruhr district, West
Germany
1990-2006
N = 1,755
Older adult
women 54-55 yr
at enrollment
NO2 (monitors):
Median: 22.15
IQR: 13.23
NO2 (estimated
using—emissions
inventories):
Median: 6.37
IQR: 10.09
LUR:
Median: 18.33
IQR: 7.97
4 methods: 5 yr
(1986-1998)mean
monitor concentrations
(8 x 8 km grid monitor
coverage) to capture
broad scale variability in
NO2. Emission
inventories used to
determine exposure to
traffic pollution (1-km
grid). LUR to estimate
the NO2 exposure at
participant's residence.
Distance from residence
to roadway <100 km.
Diabetes HR (per 10 ppb NO2)
Monitor concentration:
HR: 1.25(1.01, 1.53)
Emission Inventory:
HR: 1.15(1.04,1.27)
LUR:
HR: 1.55(1.20,2.00)
Covariate adjustment: age, BMI,
heating with fossil fuels, workplace
exposure to dust/fumes, extreme
temperature, smoking, and education.
Copollutant adjustment: none
Teichert et al.
(2013)
SALIA Cohort in
Ruhr district, West
Germany
2003-2009
Subgroup,
N = 363
Older adult
women 54-55 yr
at enrollment
NOx (ug/m3):
Population with
IGM
Mean: 74.1
SD: 31.2
Population without
IGM
Mean: 69.3
SD: 30.0
N02(ppb):
Population with
IGM
Mean: 21.09
SD: 5.79
Population without
IGM
Mean: 20.08
SD: 4.09
3 methods:
5-yr (2003-2007) mean
monitor concentration
nearest to residence
(8 x 8 km grid monitor
coverage) to capture
broad scale variability in
NO2. LUR to estimate
the NO2 exposure at
participant's residences
(40 sites). Back
extrapolation using the
ratio method to estimate
concentrations 10-20yr
prior to disease.
IGM:
OR: 1.41 (1.01,1.97)
per IQR NOx 43.16 ug/m3
IGM:
OR: 1.63(1.06, 2.51)
per 10 ppb NO2
Covariate adjustment: age, BMI,
smoking status, passive smoking,
education, exposure to indoor mold,
and season of blood sampling.
Copollutant adjustment: none
January 2015
6-90
DRAFT: Do Not Cite or Quote
-------
Table 6-7 (Continued): Epidemiologic studies of the association of long-term
exposure to nitrogen dioxide (NO2) or the sum of nitric
oxide and NO2 (NOX) with cardiometabolic disorders.
Study
Brocket al.
(2008)
Thierinq et al.
(2013)
Cohort Location
Study Period
Patients who
attended 2
respiratory clinics
in Hamilton and
Toronto, Canada
2002-2004
N = 7,634
GINIplus and
LISAplus Cohorts
in Germany
Oct 2008-Nov
2009
n = 397
Mean (ppb)
NO2:
Hamilton
C^KYisil^
Fcl 1 Idle
Median: 15.3
IQR: 3
Male
Median: 15.2
IQR: 3.2
Toronto
Female
Median: 22.9
IQR: 3.9
Male
Median: 23
IQR: 20.8
NO2:
Mean: 11.53
on- i KO
OLI. ^L.O^L
Exposure Assessment
LUR to estimate
participants' exposures.
Measurements collected
at -250 sites selected
using a
location-allocation
model.
Exposure estimates at
each participant's
residence were
calculated using an
LUR. Annual average
concentration
extrapolated from NO2
measurements taken
during three 2 week
periods (warm, cold,
and intermediate
seasons) at 40 sites.
Effect Estimates (95% Cl)
Odds ratio for diabetes mellitus (per
10ppbNO2):
Female
OR: 1.48(1.00,2.19)
Male
OR: 0.90(0.60,1.37)
Both sexes combined
OR: 1.16(0.82, 1.65)
Covariate adjustment: age, BMI, and
neighborhood income.
Copollutant adjustment: none
Percentage difference in insulin
resistance per 10 ppb NO2:
28.06(6.7, 51.7)
Covariate adjustment: sex, birth
weight, study center, parental
education, study, study design,
puberty status, age, BMI, and
exposure to smoke.
Copollutant adjustment: none
Ezeetal. (2014) SAPALDIA Cohort
in Switzerland
1991-2002
N = 6,372
NO2:
Mean: 15.03
IQR: 6.06
LUR to estimate
10-yr avg NO2
concentrations using
dispersion model
(200 x 200 m
resolution). Model
incorporated LUR
components to help
prevent underestimation
at background sites.
Annual NO2 trends were
combined with
residential histories to
estimate long-term
exposure.
Percentage increase in diabetes
prevalence (per 10 ppb NO2):
OR: 1.43(1.08,1.90)
Covariate adjustments: age, sex,
educational level, neighborhood SEI,
lifestyle, BMI, noise, hypertension,
hs-CRP, and dyslipidemia.
OR: 1.07 (0.72, 1.61), adjusted for
above covariates plus PM-io
January 2015
6-91
DRAFT: Do Not Cite or Quote
-------
Table 6-7 (Continued): Epidemiologic studies of the association of long-term
exposure to nitrogen dioxide (NO2) or the sum of nitric
oxide and NO2 (NOX) with cardiometabolic disorders.
Study
Diikema et al.
(2011)
Andersen et al.
(2012c)
Cohort Location
Study Period
Wesfriesland,
Netherlands
(semirural)
2007
N = 8,018
Diet, Cancer, and
Health Cohort in
Copenhagen or
Aarhus, Denmark
Jan.1995-June
2006
~
Mean (ppb)
NO2 range:
Q1: 4.7-7.5
Q2: 7.5-8.1
Q3: 8.1-8.8
Q4: 8.8-19.1
NO2:
1971 -end of
follow-up
iv /I a H 1 3 n ' 7 7
ivi cu id 1 1 . / . /
IQR: 2.6
1991 -end of
follow-up
Median: 8.13
IQR: 2.97
Exposure Assessment
LUR to estimate NO2
concentration at
residential address at
the time of recruitment.
Source dispersion
model that sums local
air pollution from street
traffic based on traffic
location and density.
AirGIS used to predict
the NOx, NO2, and NO
concentrations at each
participant's residence.
Effect Estimates (95% Cl)
OR (prevalence of Type II diabetes)
Q1: referent
Q2: 1.03(0.82, 1.31)
Q3: 1.25(0.99, 1.56)
Q4: 0.8, (0.63, 1.02)
Covariate adjustment: average
monthly income, age, and sex.
Diabetes HR (per 10 ppb NO2)
1971 -end of follow-up
HR all diabetes: 1.00(0.89,1.12)
HR confirmed diabetes*:
1.16(1.00, 1.35)
1991 -end of follow-up
HR all diabetes:
1.04(0.93, 1.17)
HR confirmed diabetes*:
1.16(1.04, 1.3)
Covariate adjustment: sex, BMI,
waist-to-hip ratio, smoking status,
smoking duration, smoking intensity,
environmental tobacco smoke,
educational level, physical/sports
activity in leisure time, alcohol
consumption, fruit consumption, fat
consumption, and calendar yr.
Copollutant adjustment: none
Note: confirmed diabetes defined as
exclusion of cases included solely
because of glucose blood test.
BMI = body mass index; BWHS = Black Women's Health Study; Cl = confidence interval; GINIplus = German Infant Nutritional
Intervention plus environmental and genetic influences.; HOMA = homeostatic model assessment; HR = hazard ratio; hs-CRP = high
sensitivity C-reactive protein; IGM = impaired glucose metabolism; IQR = interquartile range; IRR = incidence rate ratios;
LISAplus = Lifestyle-Related Factors on the Immune System and the Development of Allergies in Childhood plus the influence of traffic
emissions and genetics: LUR = land-use regression model; NO = nitric oxide; NO2 = nitrogen dioxide; NOX = sum of NO and NO2;
OR = odds ratio; PM2.s = particulate matter with a nominal aerodynamic diameter less than or equal to 2.5 |jm; PM10 = particulate matter
with a nominal aerodynamic diameter less than or equal to 10 |jm; Q1 = first quartile; Q2 = second quartile; Q3 = third quartile; Q4 = fourth
quartile; SALIA = Study on the Influence of Air Pollution on Lung, Inflammation, and Aging; SAPALDIA = Swiss study on Air Pollution and
Lung Disease in adults; SD = standard deviation; SEI = socio-economic index.
6.3.4 Cerebrovascular Disease and Stroke
Several studies published since the 2008 ISA for Oxides of Nitrogen examine the
association of long-term NC>2 exposure and stroke (Table 6-8). Evidence from
January 2015
6-92
DRAFT: Do Not Cite or Quote
-------
1 epidemiologic studies is not consistent and there is uncertainty regarding the extent to
2 which findings can be explained by noise or copollutant exposures.
3 A hospital-based case-control study in Edmonton, Canada reported a positive association
4 of NO2 exposure with ischemic stroke (OR: 1.06 [95% CI: 0.88, 1.27] per 10 ppb
5 increase) and a stronger positive association with hemorrhagic stroke (OR: 1.14 [95% CI
6 0.85, 1.55]) but not with transient ischemic attack (TIA) (OR: 0.90 [95% CI: 0.74, 1.10]
7 (Johnson et al.. 2013). This was the only study of stroke to use LUR to estimate NO2
8 concentration at the participants' residences. Findings were similar in an ecological
9 analysis of annual incidence of stroke also conducted in Edmonton, Canada. Positive,
10 imprecise associations with hemorrhagic and nonhemorrhagic stroke incidence were
11 observed with IDW weighted average NO2 concentration assigned based on residential
12 ZIP code (Johnson et al.. 2010). Associations of stroke with CO and traffic density were
13 also observed in this study.
14 Andersen et al. (2012b) conducted a study of long-term traffic-related NO2 exposure and
15 incident stroke using data from a large cohort study of residents of Copenhagen,
16 Denmark enrolled in the Danish Diet, Cancer, and Health Study. The Danish GIS-based
17 air pollution and human exposure dispersion modelling system was used to predict NO2
18 concentrations for geocoded residential address histories up to approximately 35 years in
19 duration. Authors report an increase in ischemic stroke incidence (HR 1.19 [95% CI:
20 0.88, 1.61]) but not hemorrhagic stroke incidence (HR: 0.80 [95% CI: 0.52, 1.22). In an
21 analysis of the same data set that adjusted for traffic-related road noise as a potential
22 confounder, S0rensen et al. (2014) reported a substantially attenuated risk estimate
23 between NO2 concentration at the time of diagnosis and ischemic stroke [incidence rate
24 ratios (IRR: 1.04 [95% CI: 0.85, 1.26]). The association for the combined effect of the
25 highest tertile of noise and the highest tertile of NO2 was increased (IRR: 1.28 [95% CI:
26 1.09, 1.52]) and the association with fatal strokes persisted after adjustment for noise,
27 however. A study of IHD and cerebrovascular diseases combined reported that
28 associations with NO2 were robust to adjustment for noise and other individual-level
29 covariates (HR: 1.18 [95% CI 0.9, 1.48] per 10 ppb increase) (de Kluizenaar et al.. 2013V
30 Lipsett etal. (2011) analyzed the association of incident stroke with long-term exposure
31 to NO2, NOx, other gases (CO, Os, SO2), and PM. These authors analyzed data from a
32 cohort of California public school teachers and assigned exposure by linking IDW
33 pollution concentrations from monitors within a representative range of 3-5 km to
34 participants' geocoded addresses. An association with incident stroke that was close to
35 the null value (HR: 1.02 [95%CI: 0.90, 1.15] per 10-ppb increase in NO2) was observed.
36 Estimates for the association of other pollutants (PMio, PM2 5, SO2, and Os) with incident
37 stroke were increased.
January 2015 6-93 DRAFT: Do Not Cite or Quote
-------
1 Atkinson et al. (2013) examined the association of incident cardiovascular disease with
2 NO2. These authors studied patients (aged 40-89 years) registered with 205 general
3 practices across the U.K. Predicted annual average NO2 concentrations within 1 km by
4 1 km grids, estimated using dispersion models, were assigned to participants based on
5 their residential postal code. Incident stroke was not associated with NCh concentration in
6 this analysis. An increase in NCh concentration for communities within 1 km of an air
7 monitoring station was associated with self-reported stroke prevalence in a multicity
8 study in China (OR 1.27 [95% CI: 0.92, 1.74] per 10 ppb increase) (Dong et al.. 2013a).
9 Oudin et al. (2011) reported no association between long-term NOx exposure and
10 ischemic stroke in a population-based registry-based case-control study conducted in
11 Scania, Sweden. Exposure was characterized using dispersion models to estimate outdoor
12 NOx concentrations within 500 m by 500 m grids and linking those predicted
13 concentrations to geocoded residential addresses. Although no association of NOx
14 exposure with stroke was observed, modification of the association of diabetes and stroke
15 by long-term NOx exposure was reported in this study.
16 Although several studies report an increased risk between NO2 exposure and stroke
17 and/or cerebrovascular disease, estimates are generally imprecise (de Kluizenaar et al..
18 2013; Dong et al.. 2013a: Johnson et al.. 2013; Andersen et al.. 2012b: Johnson et al..
19 2010). Some studies reported weak or null associations (Atkinson et al.. 2013; Lipsett
20 et al.. 2011). The positive associations observed for stroke were not consistent across
21 stroke subtype. Johnson et al. (2013) observed a larger increased risk for hemorrhagic
22 compared to ischemic stroke in a LUR study while Andersen et al. (2012b) observed an
23 increase for ischemic not hemorrhagic stroke in the Danish Diet, Cancer, and Health
24 Study. The association with ischemic stroke observed by Andersen et al. (2012b) was
25 diminished after further adjustment for noise although an interaction between highest
26 tertile of NO2 and highest tertile of noise was observed (S0rensen et al.. 2014). Evidence
27 from epidemiologic studies is not consistent, only one study using LUR to capture the
28 fine scale variability of NO2 concentrations was available, and there is uncertainty
29 regarding the extent to which findings can be explained by noise or copollutant
30 exposures.
January 2015 6-94 DRAFT: Do Not Cite or Quote
-------
Table 6-8 Epidemiologic studies of the association of long-term exposure to
nitrogen dioxide (NO2) or the sum of nitric oxide and NO2 (NOx) with
cerebrovascular disease or stroke.
Study
Cohort
Location
Study Period
Mean (ppb) Exposure Assessment
Effect Estimates (95% Cl)
Johnson et al.
(2013)
Edmonton,
Canada
Jan 2007-Dec
2009
N = 4,696 cases,
37,723 controls
NO2:
Mean
Cases: 15.4
Controls:
15.2
LUR model of NO2
concentrations matched to
residential postal code (spatial
resolution <50 m).
Stroke hospitalizations:
All stroke: 1.02(0.88, 1.18)
IS: 1.06(0.88, 1.27)
TIA: 0.90(0.74, 1.10)
HS: 1.14(0.85, 1.55)
per 10 ppb NO2
Covariate adjustment: age,
sex, and SES.
Copollutant adjustment: none
Johnson et al.
(2010)
Edmonton,
Canada
Jan 2003-Dec
2007
NO2:
Mean: 15.7
IQR: 2.2
IDW average monitor NO2
concentration assigned at
postal code centroid level.
Ecological analysis of stroke
incidence rates:
HS ED visits
Q1 RR 1.0 (reference)
Q2RR: 0.88(0.68, 1.14)
Q3RR: 1.03(0.79, 1.20)
Q4RR: 1.13(0.90, 1.43)
Q5RR: 1.14(0.92, 1.52)
non-HS ED visits
Q1 RR 1.0 (reference)
Q2RR: 1.0(0.85, 1.18)
Q3RR: 1.05(0.92, 1.20)
Q4RR: 1.02(0.87, 1.18)
Q5RR: 1.08(0.91, 1.27)
* Results for all stroke and TIA
also presented.
Covariate adjustment: age,
sex, and household income.
Copollutant adjustment: none
S0rensen et al.
(2014)
Danish Diet,
Cancer, and
Health Cohort
1993/1997-June
2006
N = 57,053
NO2:
Median: 8.1
10th: 6.3
90th: 12.4
IQR: 3.0
Predicted traffic-related NO2
concentration using dispersion
models, linked to geocoded
residential address.
Noise-adjusted IS incidence
NO2: IRR 1.04 (0.85, 1.26)
per 10 ppb NO2 (at the time of
diagnosis)
NOx*: IRR: 0.97(0.92, 1.03)
per 20 |ig/m3 NOx
Covariate adjustment: age,
sex, education, municipality,
SES, smoking status and
intensity, intake of fruits,
vegetables, alcohol, and
coffee, physical activity, BMI,
and calendar y.
Copollutant adjustment: none
January 2015
6-95
DRAFT: Do Not Cite or Quote
-------
Table 6-8 (Continued): Epidemiologic studies of the association of long-term
exposure to nitrogen dioxide (NO2) or the sum of nitric
oxide and NO2 (NOx) with cerebrovascular disease or
stroke.
Study
Andersen et al.
(2012b)
Cohort
Location
Study Period
Danish Diet,
Cancer, and
Health Cohort
1993/1997- June
2006
Mean (ppb)
NO2:
Median: 8.1
10th: 6.3
90th: 12.4
IQR: 3.0
Exposure Assessment
Predicted traffic-related NO2
concentration using AirGIS
(validated dispersion model),
linked to geocoded residential
address.
Effect Estimates (95% Cl)
IS incidence
NO2: HR 1.19(0.88, 1.61)
HS Incidence
NO2: HR: 0.80(0.52, 1.22)
Covariate adjustment: smoking
status, duration, intensity; ETS,
gender, BMI, education, sports
activity, alcohol consumption,
fruit consumption, fat
consumption, hypertension,
and hypercholesterolemia.
Copollutant adjustment: none
de Kluizenaar
etal. (2013)
Eindhoven,
Netherlands
1991-2003
N = 18,213
NO2
5th-95th:
7.5 ppb
Dispersion model, 1
resolution, linked to
residential address.
1 km Hospital admissions for IHD or
CBVD
HR: 1.16(0.9, 1.43)
Noise-adjusted HR: 1.18 (0.9,
1.48)
Covariate adjustment: age, sex
BMI, smoking, education,
exercise, marital status,
alcohol use, work situation,
and financial difficulty.
Copollutant adjustment: none
Lipsett et al.
(2011)
CIS Cohort in
California, U.S.
June
1996-Dec2005
N = 133,479
NO2:
Mean:
33.59
IQR: 10.29
NOx:
Mean: 95.6
IQR: 58.31
Geocoded residential address
linked to gridded pollutant
surface (250-m spatial
resolution) developed using
IDW fixed site monitor
concentrations (1995-2005).
Defined representative range
of 3-5 km (neighborhood and
regional monitors respectively)
for NOx and NO2 to account
for spatial variability of
pollutant.
Stroke incidence
NOx: HR 1.02 (0.96, 1.09)
NO2: HR 1.02(0.90, 1.15)
per 10 ppb NO2 and 20 ppb
NOx
Covariate adjustment: age,
race, smoking, second-hand
smoke, BMI, lifetime physical
activity, nutritional factors,
alcohol, marital status,
menopausal status, hormone
therapy, hypertension,
medication and aspirin, and
family history of Ml/stroke.
Copollutant adjustment: none
Atkinson et al.
National GP
Patient Cohort in
the U.K.
2003
N = 836,557
Mean (SD):
12.0
IQR: 5.7
ppb
Annual average NO2
concentration for 2002 at a 1
by 1 km resolution derived
from dispersion models
identifying all known emissions
sources and linked to
residential post-codes.
Stroke incidence
HR: 0.98(0.91, 1.06)
Covariates: age, sex, smoking
BMI, diabetes, hypertension,
and index of multiple
deprivation.
Copollutant adjustment: none
January 2015
6-96
DRAFT: Do Not Cite or Quote
-------
Table 6-8 (Continued): Epidemiologic studies of the association of long-term
exposure to nitrogen dioxide (NO2) or the sum of nitric
oxide and NO2 (NOx) with cerebrovascular disease or
stroke.
Study
Cohort
Location
Study Period
Mean (ppb) Exposure Assessment
Effect Estimates (95% Cl)
Dong et al.
(2013a)
33 communities in
11 districts of
3 cities in Liaoning
Province, China
2006-2008
N=24,845
NO2:
Mean: 18.7
Median 17.5
IQR4.8
District-specific 3-yr avg NO2
concentrations for
communities within 1 km of an
air monitoring station (selected
to maximize intra- and
inter-city gradients).
Self-reported stroke:
OR: 1.27(0.92, 1.74)
per 10-ppb increase
*Sex-specific results also
presented.
Covariate adjustment: age,
gender, education, occupation,
family income, BMI,
hypertension, family history of
stroke, family history of CVD,
smoking status, drinking, diet,
and exercise.
Copollutant adjustment: none
BMI = body mass index; Cl = confidence interval; CIS = California Teachers Study; CBVD = cerebrovascular disease;
CVD = cardiovascular disease; ED = emergency department; ETS = environmental tobacco smoke; GP = general practice;
HR = hazard ratio; HS = hemorrhagic stroke; IDW = inverse distance weighting; IHD = ischemic heart disease; IQR = interquartile
range; IS = ischemic stroke; LUR = land-use regression; Ml = myocardial infarction; NO2 = nitrogen dioxide; NOX = sum of NO and
NO2; non-HS = non-hemhorragic stroke; OR = odds ratio; Q1 = first quantile; Q2 = second quantile: Q3 = third quantile;
Q4 = fourth quantile; Q5 = fifth quantile; RR = risk ratio(s), relative risk; SD = standard deviation; SES = socioeconomic status;
TIA = transient ischemic attack.
*NOX results that are originally reported in jxg/m3 are not standardized if the molecular weight needed to convert to ppb is not
reported.
6.3.5 Hypertension
i
2
3
4
5
6
There were no studies of the effect of long-term NO2 or NOx exposure on hypertension in
the 2008 ISA for Oxides of Nitrogen. Several recent studies of both children and adults
add to the evidence base (Table 6-9). Overall, findings from both studies of adults and
children were inconsistent. Further, the independent effect of NO2 was not distinguished
from the effect of noise and other traffic pollutants in the epidemiologic studies reporting
positive associations.
January 2015
6-97
DRAFT: Do Not Cite or Quote
-------
Table 6-9 Epidemiologic studies of the association of long-term exposure to
nitrogen dioxide (NO2) or the sum of nitric oxide and NO2 (NOx) with
hypertension and blood pressure.
Cohort Mean/
Location Median
Study Study Period (ppb)
Exposure Assessment
Effect Estimates (95% Cl)
Studies of Adults
Cooqan et al. (2012)
BWHS Cohort
in Los Angeles,
CA
2006
N = 3,236
NOx:
Mean:
43.3 |jg/m3
Median:
41.6 |jg/m3
IQR:
12.4 |jg/m3
LUR incorporating
traffic, land use,
population, and physical
geography was used to
estimate exposure at the
participants' residences.
Summer and winter field
measurements taken at
183 sites in Los Angeles
and averaged to
estimate an annual
mean.
Hypertension (per IQR
increase in NOx)*
IRR = 1.14(1.03, 1.25)
Copollutant adjustment: none
IRR = 1.11 (1.00, 1.23)
Copollutant adjustment: PM2.5
Covariate adjustment: age,
BMI, yr of education, income,
number of people in the
household, smoking, drinks
per week, h/week of physical
activity, neighborhood
socioeconomic status score,
and noise.
Foraster et al. (2014)
REGICOR
Cohort in
Girona, Spain
2007-2009
N = 3,836
NO2:
Median:
14.1
IQR: 6.22
LUR to estimate NO2
concentrations at
participant's geocoded
address. Primary model
inputs were air sampler
height and traffic-related
variables.
Blood pressure change
(mmHg per 10 ppb NO2)
Participants taking
BP-lowering medications
Systolic: 2.24 (-2.58, 7.06)
Participants not taking
BP-lowering medications
Systolic: 2.52 (0.26, 4.80)
Covariates: age, age squared,
sex, living alone, education,
diabetes, BMI, nighttime
railway noise, smoking,
alcohol consumption,
deprivation, daily NO2,
temperature, and nighttime
railway, and traffic noise.
Copollutant adjustment: none
January 2015
6-98
DRAFT: Do Not Cite or Quote
-------
Table 6-9 (Continued): Epidemiologic studies of the association of long-term
exposure to nitrogen dioxide (NO2) or the sum of nitric
oxide and NO2 (NOX) with hypertension and blood
pressure.
Study
Cohort
Location
Study Period
Mean/
Median
(PPb)
Exposure Assessment Effect Estimates (95% Cl)
S0rensen etal. (2012)
Diet, Cancer,
and Health
Cohort in
Copenhagen or
Aarhus,
Denmark
2000-2002
N = 45,271
NOx:
(ug/m3)
1 -yr avg
baseline
Median:
20.2
IQR: 72.5
Follow-up
Median:
20.0
IQR: 71.1
5-yr avg
baseline
Median:
19.6
IQR: 73.2
Follow-up
Median:
19.3
IQR: 71.4
Air pollution modeled
using source dispersion
model that sums local
air pollution from street
traffic based on traffic
location and density.
The AirGIS modeling
system was used to
predict the NOx, NO2,
and NO concentrations
at each participant's
residence.
Blood pressure change
(mmHg in response to
doubling NOx)
1-yr period
Systolic:-0.53 (-0.88, -0.19)
5-yr period
Systolic: -0.84 (-0.84, -0.16)
5-yr period hypertension
OR: 0.96(0.91,1.00)
Covariates: traffic noise, short
term NOx, temperature,
relative humidity, age, sex,
calendar yr, center of
enrollment, length of school
attendance, body mass index,
smoking status, alcohol
intake, intake of fruit and
vegetables, sport during
leisure time, and season.
Copollutant adjustment: none
Dong etal. (2013b)
33 communities
in 11 districts of
3 cities in
Liaoning
Province, China
2006-2008
N = 24,845
NO2:
Mean: 18.7
Median:
17.5
IQR: 4.8
District-specific 3-y avg
NO2 concentrations for
communities within 1 km
of an air monitoring
station (selected to
maximize intra- and
inter-city gradients).
Blood pressure change
(mmHg per 10 ppb NO2)
Diastolic: 0.46 (-0.10, 1.02)
Systolic: 0.48 (-0.44,0.088)
Estimated change in the
prevalence of hypertension
OR: 1.20(1, 1.43)
Covariate adjustment:
smoking status, duration,
intensity; ETS, gender, BMI,
education, sports activity,
alcohol consumption, fruit
consumption, fat consumption,
hypertension, and
hypercholesterolemia.
Copollutant adjustment: none
January 2015
6-99
DRAFT: Do Not Cite or Quote
-------
Table 6-9 (Continued): Epidemiologic studies of the association of long-term
exposure to nitrogen dioxide (NO2) or the sum of nitric
oxide and NO2 (NOX) with hypertension and blood
pressure.
Cohort Mean/
Location Median
Study Study Period (ppb)
Exposure Assessment
Effect Estimates (95% Cl)
Studies of Children
Liuetal. (2014b) GINIplus and
LISAplus
Cohorts in
Germany
October 2008-
November
2009
N = 2,368
NO2: Exposure estimates at
Mean: 12.4 each participant's
residence were
"p'an: calculated using an
1 8.85 LUR An annua| average
IQR: 3.42 was extrapolated from
NO2 measurements
taken during three
2-week periods (warm,
cold, and intermediate
seasons) at 40 sites.
Blood pressure change
(mmHg per 10 ppb NO2)
Systolic: 0.32 (-1.32, 1.96)
Diastolic: -0.18 (-1.40, 1.05)
Covariate adjustment: Cohort
study, area, gender, age of
child, BMI, physical activity,
maternal smoking during
pregnancy, parental education
level, parental history of
hypertension, 7-day level of
air pollutants, and 7-day
temperature.
Copollutant adjustment: None
Systolic: -0.56 (-3.30, 2.19)
Diastolic: -2.25 (-4.59, 0.088)
N=605
Covariate Adjustment:
Additionally adjusted for
road-traffic noise.
Copollutant adjustment: none
Bilenkoetal. (2013)
PIAMA Birth
Cohort in the
Netherlands
February
2009-February
2010
N = 1,432
NO2:
Long-term
Median:
11.6
IQR: 4.1
Short-term
Median: 8.8
IQR: 7.1
Long term: Exposure
estimates at each
participant's residence
were calculated using an
LUR. An annual average
was extrapolated from
NO2 measurements
taken during three
2-week periods (warm,
cold, and intermediate
seasons) at 80 sites.
Note: short-term
exposure also
estimated.
Blood pressure change
(mmHg per 10 ppb NCfe)
Long-term
Diastolic: 0.80 (-0.44, 2.05)
Systolic: -0.073(-0.17,0.16)
Short-term
Diastolic: 0.24 (-0.77, 1)
Systolic: 0.11 (-0.44, 0.92)
Covariate adjustment: age,
sex height, BMI, cuff size,
gestational age at birth,
birthweight, weight gain during
the first yr of life, breast
feeding, maternal smoking
during pregnancy, parental
smoking in the child's home,
physical activity, puberty
development scale, maternal
education, maternal
hypertension during
pregnancy, pneumonia, and/or
otitis media during the first
2 yrs of life, ambient
temperature, and room
temperature.
Copollutant adjustment: none
January 2015
6-100
DRAFT: Do Not Cite or Quote
-------
Table 6-9 (Continued): Epidemiologic studies of the association of long-term
exposure to nitrogen dioxide (NO2) or the sum of nitric
oxide and NO2 (NOX) with hypertension and blood
pressure.
Study
Cohort
Location
Study Period
Mean/
Median
(PPb)
Exposure Assessment Effect Estimates (95% Cl)
Clark etal. (2012)
UK RANCH
Study
2001-2003
N = 719
children
(9-1 Oyr),
N = 11 schools
Combined
emission-dispersion and
regression model to
assign annual avg NO2
concentration at
20 x 20 m resolution at
each school.
No associations reported
between systolic or diastolic
blood pressure.
Covariate adjustment: age,
gender, employment status,
crowding, home ownership,
mother's educational level,
long-standing illness,
language spoken at home,
parental support for school
work, classroom window
glazing type, and noise.
Copollutant adjustment: none
BMI = body mass index; BWHS = Black Women's Health Study; Cl = confidence interval; ETS = environmental tobacco smoke;
GINIplus = German Infant Nutritional Intervention plus environmental and genetic influences; IQR = interquartile range;
IRR = incidence rate ratios; LISAplus = Lifestyle-Related factors on the Immune System and the Development of Allergies in
Childhood plus the influence of traffic emissions and genetics; NO = nitric oxide; NO2 = nitrogen dioxide; NOX = sum of NO and NO2;
RANCH = Road Traffic and Aircraft Noise Exposure and Children's Cognition and Health; LUR = land-use regression model;
OR = odds ratio; PIAMA = Prevention and Incidence of Asthma and Mite Allergy; PM25 = particulate matter with a nominal
aerodynamic diameter less than or equal to 2.5 |jm; REGICOR = Registre Gironi del Cor.
*NOX results that are originally reported in ng/m3 are not standardized if the molecular weight needed to convert to ppb is not
reported.
1
2
o
5
4
5
6
1
8
9
10
11
12
13
14
15
16
17
18
Coogan et al. (2012) examined the association of long-term NOx exposure with incident
hypertension among black women residing in Los Angeles, CA. An LUR model was
used to estimate exposure at each participant's residential address, and cross validation of
the model performance produced an R2 value of approximately 92%. These authors
reported an increased risk of 1.14 (95% Cl: 1.03, 1.25%) per IQR (12.4 ug/m3) increase
in NOx after adjustment for a wide array of potential confounders including
traffic-related noise exposure. Slight attenuation in the effect estimate for NOx was
reported after adjustment for PIVb 5. Although the correlation between NOx and PIVb 5
(PM2 5 concentration was estimated using kriging) was low (r = 0.27), correlations
between NOx and other traffic-related copollutants were not reported.
In a cross-sectional study, Foraster et al. (2014) reported that NO2 was associated with an
increase in systolic blood pressure that was attenuated to varying degrees depending on
the method adjustment for medication use but not with hypertension. In this study, LUR
(R2 = 0.63) was used to estimate NO2 exposure among participants (35-83 years) of a
large population-based cohort study. Both short-term exposure to NO2 and noise were
adjusted for in the analysis, in addition to an array of other potential confounders.
Associations with systolic blood pressure were stronger among those with cardiovascular
disease, those living alone, and those living in areas with high traffic load and traffic
January 2015
6-101
DRAFT: Do Not Cite or Quote
-------
1 noise. In a large study involving Danish adults (50-64 year old), S0rensen et al. (2012)
2 reported an inverse association between NO2 and both systolic and diastolic blood
3 pressure at baseline while largely null findings were report for self-reported incident
4 hypertension. A validated dispersion model was used in this study to predict annual and
5 5-yr avg NO2 concentration and to assign exposure based on residential address history
6 \r> 0.9 for correlation between measured and predicted 1/2-yr avg NO2 concentration;
7 (S0rensen et al., 2012)1 Dong etal. (2013b) reported imprecise associations of average
8 NO2 concentration from monitoring stations located within 1 km of the community where
9 study participants resided, with prevalent hypertension and increased blood pressure.
10 Stronger associations of hypertension with PMio, SO2, and Os were reported in this study.
11 Some additional studies examined the association of NO2 with blood pressure in children.
12 Liu etal. (2014b) reported an increase in diastolic blood pressure that was diminished
13 after adjustment for traffic-related noise exposure. Bilenko et al. (2013). however,
14 reported an association between NO2 and diastolic blood pressure that was robust to
15 adjustment for noise. Both studies used LUR models to assign NO2 exposure at each
16 participant's residential address. A study designed to evaluate the effect of aircraft noise
17 on cognition among school children (9-10 years old) reported no association between
18 NO2 and blood pressure after adjustment for noise (Clark etal.. 2012).
19 Overall, findings from studies of adults and studies of children report weak, inconsistent
20 results for the association between NO2 and hypertension and increased blood pressure,
21 although one prospective study using LUR to estimate exposure and adjusting for a
22 cardiovascular disease risk factors reported an association of NOx with hypertension
23 (Coogan etal.. 2012). Uncertainties remain regarding the independent effect of NO2 on
24 hypertension and blood pressure and specifically whether confounding by correlated
25 traffic-related pollutants or noise can explain the positive associations observed.
6.3.6 Markers of Cardiovascular Disease
26 Some epidemiologic and toxicological studies published since previous assessments have
27 investigated the effects of long-term NCh exposure on risk factors and markers of
28 cardiovascular disease risk, such as arterial stiffness, subclinical atherosclerosis,
29 circulating lipids, and heart rate variability (HRV). Previous information was limited. The
30 1993 Air Quality Criteria Document (AQCD) for Oxides of Nitrogen (U.S. EPA. 1993)
31 reported a significant reduction in HR in rats exposed to 1,200 and 4,000 ppb NO2 for
32 1 month but not after lower concentrations or longer durations of exposure (Suzuki et al..
33 1981). There were no changes in vagal responses in rats exposed to 400 ppb NO2 for
34 4 weeks (Tsubone and Suzuki. 1984).
January 2015 6-102 DRAFT: Do Not Cite or Quote
-------
1 Several recent cross-sectional analyses of long-term exposure to NCh evaluated vascular
2 markers of cardiovascular disease. Rivera et al. (2013) used LUR to estimate NO2
3 concentrations. Increases in carotid intima-media thickness (cIMT) observed in crude
4 models were attenuated in fully adjusted models while a positive association between
5 NC>2 and high ankle brachial index (ABI >1.3) remained. LUR was also used to assess
6 vascular damage in health young adults (Lenters et al.. 2010). Increases in pulse wave
7 velocity, augmentation index, but not cIMT in association with NC>2 exposure were
8 observed in this study. In a study that used IDW methods for spatial interpolation and
9 linked to participants' residential addresses, exposure to NO2 during childhood lifestages
10 was not associated with an increase in cIMT among young adults (Breton et al.. 2012).
11 The effects of NO2 in relation to autonomic function in a random selection of Swiss
12 cohort study participants have also been examined. In this study, Felber Dietrich et al.
13 (2008) linked measures of HRV to annual NC>2 concentration at each participant's
14 residential address using dispersion model predictions supplemented with land use and
15 meteorological data. Annual average NO2 concentration was associated with decreased
16 standard deviation of beat-to-beat (NN) intervals, an index of total HRV, nighttime
17 low-frequency component of HRV (LF), and LF/high frequency component of HRV
18 (HF) ratio in women. No associations with other parameters of HRV were observed in
19 these data.
20 A recent experimental animal study by Seilkop etal. (2012) reported changes in markers
21 that are characteristic of vascular disease development and progression (see Table 6-10
22 for toxicological study details). Mice were exposed for 50 days to various multipollutant
23 atmospheres (diesel or gasoline exhaust, wood smoke, or simulated "downwind" coal
24 emissions) comprising varying concentrations of NC>2 (0-3,670 ppb). A data mining
25 technique known as Multiple Additive Regression Trees analysis was employed to
26 identify associations between the 45 different exposure component categories, including
27 NO2, and various effects [markers of oxidative stress (discussed in Section 5.3.4) and
28 cardiovascular disease stability and progression (endothelin-1 (ET-1), matrix
29 metalloproteinase (MMP)-3, MMP-7, MMP-9, tissue inhibitor of metalloproteinase-2
30 (TIMP-2))]. The results demonstrated NCh was among one of the strongest predictors of
31 responses. More specifically, NCh ranked among the top three predictors for ET-1 and
32 TIMP-2; however, the study design did not allow for the independent effects of NCh to be
33 evaluated.
January 2015 6-103 DRAFT: Do Not Cite or Quote
-------
Table 6-10 Study details for toxicological studies examining cardiovascular
effects from long-term nitrogen dioxide (NO2) exposure.
Study
Species (Strain);
Age;
Sex; n
Exposure Details (Concentration;
Duration)
Endpoints Examined
de Burbure
et al. (2007)
Rats (Wistar);
8 weeks; M;
n = 8/group
High (6 ug/day) Or low (1.3 ug/day) GPx in plasma and RBC lysate; SOD
selenium;
(1)1,000ppb, 28 days, 6 h/day,
5 days/week (Se+/Se~);
(2) 10,000 ppb, 28 days, 6 h/day,
5 days/week;
(3) 5,000 ppb, 5 days, 6 h/day;
(4) 50,000 ppb, 30 min
activity in RBC lysate; GST activity in RBC
lysate; TEARS in plasma. Endpoints
examined immediately and 48 h after
exposure.
Renters et al.
Squirrel monkeys;
adult; M; n = 4
1,000 ppb NO2, continuously for
16 mo; challenged with influenza
virus
Hemoglobin and hematocrit levels were
measured throughout the study.
Furiosi et al. Monkeys (Macaca (1) 2,000 ppb NO2, continuously for Erythrocyte, hematocrit, and hemoglobin
spec/osa); adult;
M/F; n = 4
Rats
(Sprague-Dawley);
4 weeks; M; n = 8
14 mo
levels were measured throughout the
study.
Seilkop et al. Mice (ApoE-/-); NO2 (along with 700 other
(2012) 10 weeks; M; components) Fed a high-fat diet;
n = 8-10 260, 745, and 3,670 ppb (along
with dilutions of 1/3 and 1/10);
6 h/day, 7 days/week for 50 days
ET-1, VEGF, MMP-3, MMP-7, MMP-9,
TIMP-2, HO-1, TEARS in proximal aorta
18-h after exposure.
Suzuki etal. Rats; NR; NR; 400, 1,200, and 4,000 ppb NO2;
(1981) n = 6 1,2, and 3 mo
HR and hemoglobin levels measured after
1, 2, and 3 mo exposures.
Takano et al.
(2004)
Rats (OLETF and
LETO); 4 weeks;
M; n = 10-14
160, 800, or 4,000 ppb NO2;
continuously for 32 weeks
BW, Triglyceride, HDL, total cholesterol,
HDL/total cholesterol, sugar measured
8 weeks following exposure.
Tsubone and Rats (Wistar);
Suzuki (1984) 9-13 weeks; M;
n = 6
400 and 4,000 ppb NCb;
continuously for 1 and 4 weeks,
respectively; Immediately after
exposure, animals were injected
with 5 ug/kg BW phenyl diguanide
HR was measured 10 sec after injection.
Wagner et al.
(1965)
Dogs; adult; M;
n = 6-10/group
1,000 or 5,000 ppb NO2;
continuously for 18 mo
Hemoglobin and hematocrit levels were
measured quarterly throughout exposure.
BW = body weight; ET-1 = endothelin-1; GPx = glutathione peroxidase; GST = glutathione s-transferase; HDL = high density
lipoprotein; HO-1 = heme oxygenase-1; HR = hazard ratio; LETO = Long-Evans Tokushima; NO2 = nitrogen dioxide; NR = no
quantitative results reported; OLETF = Otsuka Long-Evans Tokushima Fatty; RBC = red blood cell; SOD = superoxide dismutase;
TEARS = thiobarbituric acid reactive substances; TIMP-2 = tissue inhibitor of metalloproteinase-2; VEGF = vascular endothelial
growth factor
January 2015
6-104
DRAFT: Do Not Cite or Quote
-------
1 In another study, Takano et al. (2004) reported that obese rats (Otsuka Long-Evans
2 Tokushima Fatty) had elevated levels of triglycerides and decreased high-density
3 lipoprotein (HDL) and HDL/total cholesterol levels after long-term exposure to 160 ppb
4 NO2 compared to clean air. HDL levels were also decreased after 800 ppb NO2 exposure
5 in the obese strain and in the nonobese rats (Long-Evans Tokushima). The authors
6 suggested that obese animals were at greater risk of dyslipidemia following NO2
7 exposure.
8 Overall, a limited number of epidemiologic and toxicological studies have evaluated
9 long-term NO2 exposure on markers of cardiovascular disease. There is some evidence
10 for increased arterial stiffness, increased markers for cardiovascular disease stability and
11 progression, dyslipidemia, decreased HRV, and reduced HR; however, these effects have
12 only been reported in one study each. Findings from several studies of cIMT are
13 inconsistent. Further, the independent effect of NO2 is not consistently distinguished in
14 the available body of epidemiologic and toxicological evidence.
6.3.7 Inflammation and Oxidative Stress
15 Inflammation and oxidative stress have been shown to play a role in the progression of
16 chronic cardiometabolic disorders including heart disease and diabetes. Although studies
17 of inflammation and oxidative stress were not generally available for inclusion in the
18 2008 ISA for Oxides of Nitrogen (U.S. EPA. 2008). a number of null findings related to
19 changes in hematological parameters were reported. Hematocrit and hemoglobin levels
20 were unchanged in squirrel monkeys (Fenters etal. 1973). rats (Suzuki etal.. 1981). or
21 dogs exposed to <5,000 ppb NO2 (Wagner etal.. 1965). However, Furiosi etal. (1973)
22 reported polycythemia due to reduced mean corpuscular volume and an increased trend in
23 the ratio of neutrophil to lymphocytes in the blood of NO2-exposed monkeys and similar
24 increases in erythrocyte counts in NO2-exposed rats.
25 A limited number of studies published since the 2008 ISA for Oxides of Nitrogen have
26 evaluated markers of inflammation and oxidative stress. Forbes et al. (2009a) examined
27 the association of predicted annual average NO2 concentrations with C-reactive protein
28 (CRP) and fibrinogen among an English population. Multilevel linear regression models
29 were used to determine pooled estimates across three cross-sectional surveys conducted
30 during different years. Each participant's postal code of residence was linked to a
31 predicted annual average NO2 concentration derived from dispersion models. NO2 was
32 not associated with increased CRP or fibrinogen in these data nor were PMio, SO2, or Os.
33 A study conducted among men and women (45-70 year-olds) in Stockholm reported an
34 association of 30-yr avg traffic-related NO2 concentrations estimated using dispersion
December 2014 6-105 DRAFT: Do Not Cite or Quote
-------
1 models with increases in interleukin-6 (IL-6) and CRP but not with TNF-a, fibrinogen, or
2 PAI-1 (Panasevich et al.. 2009). Associations between several metrics of SO2 exposure
3 and increased IL-6 and CRP were observed in this study. In another analysis from this
4 study, long-term exposure to NC>2 interacted with IL-6 and TNF polymorphisms on an
5 additive scale with regard to increased MI risk (Panasevich et al., 2013). In a study of
6 COPD patients, annual average NO2 concentration was associated with increases in
7 interleukin-8 (IL-8) but not with the other markers studied including CRP, TNF-a IL-6,
8 fibrinogen, and hepatocyte growth factor (Dadvand et al.. 2014b).
9 de Burbure et al. (2007) examined oxidative stress markers in rats on a low selenium
10 (Se-L) or supplemented selenium (Se-S) diet exposed to 1,000 ppb NO2 for 28 days.
11 Blood Se levels decreased significantly in both groups immediately after the 28-day
12 exposure and continued to decrease in the Se-S diet rats following a 48-hour recovery
13 period. GPx, of which Se is an integral component, also decreased immediately and
14 48 hours after exposure only in the plasma of Se-S diet rats. However, GPx levels
15 increased in red blood cells (RBC) of Se-L diet rats immediately after the 28-day
16 exposure and increased in both groups 48 hours later. RBC SOD activity increased in
17 both groups immediately after the exposure and decreased in Se-L diet rats 48 hours later.
18 GST was increased for both groups immediately after the 28-day exposure and continued
19 to increase after the 48-hour recovery period, potentially compensating for the increase in
20 thiobarbituric acid reactive substances (TEARS) immediately after exposure.
21 As discussed in Section 6.3.6. Seilkop et al. (2012). examined the effects of NO2
22 exposure, in a multipollutant context, on markers of oxidative stress [heme oxygenase-1
23 (HO-1) expression and TBARs, indicator of lipid peroxidation] in ApoE-/- mice fed a
24 high-fat diet. Mice were exposed to various atmospheres (diesel or gasoline exhaust,
25 wood smoke, or simulated "downwind" coal emissions) with varying concentrations of
26 NC>2 (0-745 ppb) for 50 days. Associations between the oxidative stress indicators and
27 the 45 different exposure component categories were determined using a data mining
28 technique known as Multiple Additive Regression Trees analysis. The results
29 demonstrated NC>2 was among one of the strongest predictors of response for TBARS but
30 not HO-1.
31 Overall, a limited number of epidemiologic and toxicological studies have evaluated
32 long-term NO2 exposure on inflammation and oxidative stress with some, but not all,
33 studies reporting positive associations. In general, findings of the epidemiologic studies
34 are mixed, and the animal toxicological studies do not consistently separate the effect of
35 NO2 from other copollutants.
January 2015 6-106 DRAFT: Do Not Cite or Quote
-------
6.3.8 Cardiovascular Mortality
1 Results of studies of long-term exposure to NCh and cardiovascular diseases are coherent
2 with findings reporting associations of long-term NC>2 exposure with total and
3 cardiovascular mortality. Consistent, positive associations with total mortality, as well as
4 deaths due to cardiovascular disease have been observed in cohort studies conducted in
5 the U.S. and Europe (Section 6.5.2. Figure 6-9. and Table 6-16). Specifically, the
6 strongest evidence comes from a number of recent studies that have observed positive
7 associations between exposure to NO2 and NOx and IHD mortality (Cesaroni et al.. 2013;
8 Chen et al.. 2013; Lipsett et al.. 2011; Yorifuji et al.. 2010). mortality due to coronary
9 heart disease (Ganet al.. 2011; Rosenlund et al.. 2008). and circulatory mortality
10 (Yorifuji et al.. 2010; Jerrett et al.. 2009). Coherence is also provided for the effect of
11 long-term exposure and cardiovascular effects by the evidence from studies of short-term
12 cardiovascular mortality and morbidity (Section 5.3).
6.3.9 Summary and Causal Determination
13 Overall, the evidence is suggestive, but not sufficient, to infer a causal relationship
14 between long-term exposure to NC>2 and cardiovascular and related cardiometabolic
15 effects. This conclusion is based on the consideration of recent epidemiologic studies
16 reporting associations of NC>2 with heart disease, diabetes, stroke, and hypertension.
17 While well-conducted studies of NOx are also available, these studies are less
18 informative regarding the independent effect of NO2 exposure on the cardiovascular
19 system. Although associations with these cardiovascular and related cardiometabolic
20 outcomes were not entirely consistent across studies, several studies reporting positive
21 associations were prospective in design and did not report evidence that findings were
22 likely to be biased by selective participation or missing data. Additionally, these studies
23 adjusted for a wide array of cardiovascular risk factors and used exposure assessment
24 methods that captured the fine scale variability in NO2 concentration. Uncertainty
25 remains, however, regarding the independent effect of NO2 relative to other traffic-related
26 exposures and noise. Some support is provided by a limited body of evidence
27 demonstrating biological plausibility, as well as consistent associations between
28 long-term NO2 exposure and cardiovascular mortality. This current conclusion represents
29 a change from the conclusion drawn in the 2008 ISA for Oxides of Nitrogen, which
30 stated that the evidence was inadequate to infer the presence or absence of a causal
31 relationship. The evidence for cardiovascular effects with respect to the causal
32 determination for long-term NO2 exposure is detailed below using the framework
33 described in Table II of the Preamble to this ISA. The key evidence as it relates to the
34 causal determination is summarized in Table 6-11.
January 2015 6-107 DRAFT: Do Not Cite or Quote
-------
1 Briefly, the 2008 ISA for Oxides of Nitrogen concluded that the available evidence was
2 inadequate to infer the presence or absence of a causal relationship between long-term
3 NO2 exposure and cardiovascular disease. Although Miller etal. (2007) reported a
4 positive association between long-term NO2 exposure and cardiovascular events among
5 post-menopausal women, an independent effect of NO2 was not distinguished in this
6 study. Several studies evaluating hematological parameters reported mixed results that
7 included no changes in hematocrit or hemoglobin and increased erythrocyte count.
8 Evidence from recent, large and well-conducted prospective epidemiologic studies
9 generally supports the association of long-term exposure to NO2 with heart disease and
10 diabetes. The strongest evidence for heart disease comes from a large multicohort
11 prospective study using LUR to predict NO2 concentrations on a finely resolved spatial
12 scale (Cesaroni et al.. 2014) with supporting evidence from cross-sectional study of CHD
13 hospitalizations also using LUR (Beckerman et al.. 2012). Consistent findings from
14 multiple epidemiologic studies of cardiovascular mortality support these morbidity
15 findings (Section 6.5.2). Studies using dispersion models or IDW for exposure
16 assessments were less consistent, with some reporting positive associations (Lipsett et al..
17 2011) and others reporting null associations with morbidity outcomes (Atkinson et al..
18 2013; Rosenlund et al.. 2009a). As noted in Section 3.2.3. IDW and dispersion modeling
19 may not adequately capture the spatial variability in NO2 concentrations, resulting in
20 biased exposure estimates. Several large prospective studies also using LUR report
21 increased risk of diabetes, impaired glucose metabolism, and increased insulin resistance
22 with NOx and NO2 exposure (Coogan et al.. 2012; Kramer etal.. 2010) (Teichertet al..
23 2013; Thiering et al.. 2013). The evidence for the association of long-term NO2 exposure
24 with stroke and hypertension is less consistent.
25 Studies in human populations offer limited support that long-term NO2 exposures may be
26 associated with increased high ABI (Rivera et al.. 2013). arterial stiffness (Lenters et al..
27 2010) and markers of inflammation [CRP and IL-6; (Panasevich et al.. 2009)].
28 Epidemiologic studies of the effect of NO2 on cIMT are inconsistent. Toxicological
29 studies provide limited evidence that NO2 may be independently associated with the
30 effects observed in epidemiologic studies by demonstrating dyslipidemia and oxidative
31 stress in animals after long-term exposure (de Burbure et al.. 2007; Takano etal.. 2004).
32 Evidence for cardiovascular effects is provided by both controlled human exposure and
33 animal toxicological studies that report increased markers of oxidative stress (Channell
34 etal.. 2012; Li etal.. 2011) and inflammation (Huang etal.. 2012; Riedl etal.. 2012) after
35 short-term NO2 exposure, however.
36 Although several epidemiologic studies report positive associations of NO2 or NOx
37 exposure with heart disease and diabetes, confounding by correlated traffic-related
January 2015 6-108 DRAFT: Do Not Cite or Quote
-------
1 pollutants and noise remains an uncertainty. Animal toxicological studies provide only
2 limited support for the biological plausibility of associations observed in the
3 epidemiologic studies. In addition, the annual average or 5-yr avg residential exposures
4 were typically considered surrogates for long-term exposure, and residential stability was
5 assumed (or sometimes required for eligibility). Further, most studies did not disentangle
6 the effects of long-term from short-term exposure. These general limitations introduce
7 some uncertainty with regard to the specific patterns of exposure associated with the
8 observed effects. Overall, the evidence from some epidemiologic studies of
9 cardiovascular and cardiometabolic effects is suggestive, but not sufficient, to infer a
10 causal relationship between long-term NO2 exposure and cardiovascular and related
11 cardiovascular effects.
January 2015 6-109 DRAFT: Do Not Cite or Quote
-------
Table 6-11 Summary of evidence, which is suggestive, but not sufficient, to infer
a causal relationship between long-term nitrogen dioxide (NO2)
exposure and cardiovascular and related metabolic effects.
Rationale for Causal
Determination3
Key Evidence13
Key References13
NO2 Concentrations
(ppb) Associated with
Effects0
CHD and Ml
Evidence from
epidemiologic studies
generally supportive but
not entirely consistent
Findings from a large
multicohort prospective
study using LUR provides
evidence that NO2 is
associated with coronary
events.
Cesaroni et al. (2014)
Range of means across
cohorts: 4.2-31.9
Supporting evidence from
cross-sectional
epidemiologic study of CHD
hospitalizations and study
of Ml incidence using IDW.
Beckerman et al. (2012)
Median: 22.9
Lipsett et al. (2011)
Mean (5th-95th): 33.59
Inverse or null associations Atkinson et al. (2013)
of NO2 with CHD or Ml
reported in some
epidemiologic studies.
Mean: 12.0
Rosenlund et al. (2009a)
Median; 6.9 (cases), 6.3
(controls)
Ganetal. (2011)
Mean: 16.3
Consistent associations
from multiple,
high-quality
epidemiologic studies of
cardiovascular mortality
support morbidity
findings
Strongest evidence of
mortality from IHD, CHD,
and circulatory diseases,
including supporting
evidence of positive
mortality associations from
studies noted above that
report weak or null
associations with
cardiovascular morbidity
outcomes.
Rosenlund et al. (2009a):
Ganetal. (2011)
Section 6.5.2
See above
Uncertainty remains
regarding potential
confounding by
traffic-related pollutants
and noise
Overall, studies did not
consistently adjust for
PM2.5, BC, EC, CO, or
noise.
Section 6.3.32
January 2015
6-110
DRAFT: Do Not Cite or Quote
-------
Table 6-11 (Continued): Summary of evidence, which is suggestive, but not
sufficient, to infer a causal relationship between long-
term nitrogen dioxide (NO2) exposure and cardiovascular
and related metabolic effects.
Rationale for Causal
Determination3
Key Evidence13
Key References13
NO2 Concentrations
(ppb) Associated with
Effects0
Diabetes
Evidence from
epidemiologic studies
generally consistent and
supportive
Large prospective studies
using LUR report increased
risk of diabetes incidence,
impaired glucose
metabolism, and increased
insulin resistance with NOx
and NO2 exposure.
Cooqan et al. (2012)
Mean (NOx): 43.3
Kramer et al. (2010)
Median: 18.33
Teichert et al. (2013)
Mean: -21 (among those w
IGM); Mean: -20 (among
those w/o IGM)
Thierinqetal. (2013)
Mean: 11.53
Supporting evidence that Eze et al. (2014)
NO2 exposure is associated
with prevalent diabetes.
Mean: 15.03
Brook et al. (2008)—females Mean:-15 (Hamilton);
only Mean: -23 (Toronto)
Association observed Andersen et al. (2012c)
among confirmed cases of
diabetes but not overall.
Median: -8
Uncertainty remains
regarding potential
confounding by traffic
pollutants
Consistent but limited Eze et al. (2014): S0rensen
evidence that associations et al. (2013): Eriksson et al.
are robust to adjustment for (2014)
noise.
See above
Overall, studies did not
consistently adjust for
PM2.5, BC, EC, CO, or
noise.
Section 6.3.4
Associations observed with Cooqan et al. (2012)
NOx may not inform the
independent effect of NO2.
January 2015
6-111
DRAFT: Do Not Cite or Quote
-------
Table 6-11 (Continued): Summary of evidence, which is suggestive, but not
sufficient, to infer a causal relationship between long-
term nitrogen dioxide (NO2) exposure and cardiovascular
and related metabolic effects.
Rationale for Causal
Determination3
Key Evidence13
Key References13
NO2 Concentrations
(ppb) Associated with
Effects0
Cerebrovascular Disease and Stroke
Inconsistent evidence
from epidemiologic
studies
Some studies of variable
quality report increased,
typically imprecise, risks of
stroke and/or
cerebrovascular disease
with NO2. Inconsistency
across stroke subtype.
Johnson et al. (2013),
Johnson et al. (2010).
Andersen et al. (2012b),
Mean: 15.4 (cases), 15.2
(controls)
Mean: 15.7
Median: 8.1
de Kluizenaaret al. (2013)
NR
Dong et al. (2013a)
Mean: 18.7
Other studies reported Atkinson et al. (2013).
weak or null associations.
Mean: 12.0
Lipsettetal. (2011)
Mean: 33.59
Confounding bias cannot Studies not consistently
be ruled out robust to adjustment for
noise.
S0rensen et al. (2014)
See above
de Kluizenaaret al. (2013)
Uncertainty regarding
potential confounding by
traffic pollutants.
Section 6.3.54
January 2015
6-112
DRAFT: Do Not Cite or Quote
-------
Table 6-11 (Continued): Summary of evidence, which is suggestive, but not
sufficient, to infer a causal relationship between long-
term nitrogen dioxide (NO2) exposure and cardiovascular
and related metabolic effects.
Rationale for Causal
Determination3
Key Evidence13
Key References13
NO2 Concentrations
(ppb) Associated with
Effects0
Hypertension
Overall, studies of
hypertension and
increased blood pressure
are inconsistent
Large prospective study
using LUR to estimate NOx
exposure reports
association with
hypertension.
Cooqan et al. (2012)
Mean (NOx): 43.3
Cross-sectional association
between NO2 and
increased blood pressure
but not hypertension.
Foraster et al. (2014)
Median: 14.1
Large prospective study
using LUR does not report
positive associations with
hypertension and/or
increased blood pressure.
S0rensen et al. (2012)
Median: 20.2
Uncertainty regarding
potential confounding by
traffic related pollutants
and noise
Overall, studies did not
consistently adjust for
PM2.5, BC, EC, CO, or
noise.
Section 5.3.6
Association observed with
NOx may not inform the
independent effect of NO2.
Cooqan et al. (2012)
Biological Plausibility and Coherence for Cardiovascular and Related Cardiometabolic Effects
Limited and supportive
evidence from
epidemiologic and
toxicological studies for
effects on cardiovascular
disease risk provides
biological plausibility
Associations of NO2
exposure with some
markers of vascular
damage were observed in
epidemiologic studies.
Lenters etal. (2010)
Rivera et al. (2013)
Mean: 18.3
Dyslidemia—increased
triglycerides and decreased
HDL—in rats.
Takano et al. (2004)
Rats: 160
January 2015
6-113
DRAFT: Do Not Cite or Quote
-------
Table 6-11 (Continued): Summary of evidence, which is suggestive, but not
sufficient, to infer a causal relationship between long-
term nitrogen dioxide (NO2) exposure and cardiovascular
and related metabolic effects.
Rationale for Causal
Determination3
Key Evidence13
Key References13
NO2 Concentrations
(ppb) Associated with
Effects0
Some evidence for key
events within the mode
of action
Lietal. (2011),
Section 4.3.2.9, Figure 4-3
Rats: 5,320 but not 2,660
Limited and supportive
evidence of increased
oxidative stress in rats with
relevant short-term and
long-term NO2 exposures
(i.e., MDA, TEARS, GPx,
GST) and in plasma from
NO2-exposed humans
(i.e., LOX-1).
de Burbure et al. (2007)
Rats: 1,000
Channelletal. (2012)
Healthy adults: 500
Limited and supportive
toxicological evidence of
increased transcription of
some inflammatory
mediators in vitro after
short-term exposure to NO2
(i.e., IL-8, ICAM-1,
VCAM-1) and in rats
(i.e., ICAM-1, TNF-a).
Channelletal. (2012)
Human cells exposed to
plasma from healthy
adults: 500
Lietal. (2011)
Rats: 2,660 and 5,320
Limited and inconclusive
evidence in controlled
human exposure studies
(i.e., IL-6, IL-8, ICAM-1).
Huang et al. (2012)
Adults: 350
Riedletal. (2012)
Adults: 500
January 2015
6-114
DRAFT: Do Not Cite or Quote
-------
Table 6-11 (Continued): Summary of evidence, which is suggestive, but not
sufficient, to infer a causal relationship between long-
term nitrogen dioxide (NO2) exposure and cardiovascular
and related metabolic effects.
NO2 Concentrations
Rationale for Causal (ppb) Associated with
Determination3 Key Evidence13 Key References'3 Effects0
Some evidence for key Inconsistent epidemiologic Section 6.3.7
events within the mode evidence for increases in
of action (continued) CRP and IL-6 in adults.
BC = black carbon; CHD = coronary heart disease; CO = carbon monoxide; CRP = C-reactive protein; GPx = glutathione
peroxidase; GST = glutathione s-transferase; EC = elemental carbon; HDL = high-density lipoprotein; ICAM = intercellular
adhesion molecule 1; IDW = inverse distance weighting; IGM = impaired glucose metabolismn; IHD = ischemic heart disease;
IL-6 = interleuken-6; IL-8 = interleukin-8; LOX-1 = lectin-like oxidized low density lipoprotein receptor 1; LUR = land-use
regression; MDA = malondialdehyde; Ml = myocardial infarction; NO2 = nitrogen dioxide; NOX = sum of NO and NO2; NR = no
quantitative results reported; PM2.5 = particulate matter with a nominal aerodynamic diameter less than or equal to 2.5 |jm;
TEARS = thiobarbituric acid reactive substances; TNF-a = tumor necrosis factor alpha; VCAM-1 = vascular cell adhesion molecule
1.
aBased on aspects considered in judgments of causality and weight of evidence in causal framework in Table I and Table II of the
Preamble.
""Describes the key evidence and references that contribute most heavily to causal determination, and where applicable, to
uncertainties and inconsistencies. References to earlier sections indicate where the full body of evidence is described.
°Describes the NO2 concentrations with which the evidence is substantiated (for experimental studies, below 5,000 ppb).
6.4 Reproductive and Developmental Effects
6.4.1 Introduction
1 The body of literature characterizing the health effects associated with exposure to NO2 is
2 large and continues to grow; much of the research focuses on birth outcomes, for which
3 the body of evidence has grown considerably since the 2008 ISA for Oxides of Nitrogen
4 (U.S. EPA. 2008). Due to the growth in the quantity of literature, as well as in the breadth
5 of the health endpoints evaluated, the reproductive and developmental effects will be
6 divided into three separate categories: (1) Fertility, Reproduction, and Pregnancy (i.e., the
7 ability to achieve and maintain a healthy pregnancy, with emphasis on the health of
8 potential parents); (2) Birth Outcomes [i.e., measures of birth weight and fetal growth,
9 preterm birth (PTB), birth defects, and infant mortality, with emphasis on the perinatal
10 health of the child] and (3) Developmental Effects (i.e., effect on development through
11 puberty/adolescence). Separate causal determinations are made for each of these
12 categories at the end of this section. Among the epidemiologic studies of birth outcomes,
13 various measures of birth weight and fetal growth, such as low birth weight (LEW), small
14 for gestational age (SGA), intrauterine growth restriction (IUGR), and preterm birth
15 (<37-week gestation) have received more attention in air pollution research, while
January 2015 6-115 DRAFT: Do Not Cite or Quote
-------
1 congenital malformations are less studied. There is some examination of effects on
2 fertility and pregnancy conditions; however, studies on any particular endpoint remain
3 limited. The toxicological studies of outcomes analogous to fetal growth and birth weight
4 in humans measured litter size and birth weight in animals. Nervous system and
5 respiratory outcomes after early life exposures to NC>2 are examined in the developmental
6 toxicological and epidemiologic literature.
7 A major issue in studying environmental exposures and reproductive and developmental
8 effects (including infant mortality) is selecting the relevant exposure period because the
9 biological mechanisms leading to these outcomes and the critical periods of exposure are
10 poorly understood. To account for this, many epidemiologic studies evaluate multiple
11 exposure periods (including long-term exposure periods, such as the entirety of
12 pregnancy, individual trimesters or months of pregnancy; or short-term (days to weeks)
13 exposure periods, such as the days and weeks immediately preceding birth). Due to the
14 shorter length of gestation in rodents (18-24 days, on average), animal toxicological
15 studies investigating the effects of NO2 on pregnancy generally utilize short-term
16 exposure periods, which cover an entire lifestage. Thus, a study in humans that uses the
17 entire pregnancy as the exposure period is considered to have a long-term exposure
18 period (about 40 weeks, on average), while a toxicological study conducted with rats that
19 also uses the entire pregnancy as the exposure period (about 18-24 days, on average) is
20 defined as a short-term exposure. In order to characterize the weight of evidence for the
21 effects of NO2 on reproductive and developmental effects in a consistent, cohesive, and
22 integrated manner, results from both short-term and long-term exposure periods are
23 included in this section and are identified accordingly in the text and tables throughout
24 this section.
25 Due to the paucity of data for biological mechanisms and uncertainty regarding relevant
26 exposure periods, all of the studies of reproductive and developmental outcomes,
27 including infant mortality, are evaluated in this section. Exposures proximate to death
28 may be most relevant if exposure causes an acute effect. However, exposure occurring in
29 early life might affect critical growth and development, with results observable later in
30 the first year of life, or cumulative exposure during the first year of life may be the most
31 important determinant. In dealing with the uncertainties surrounding these issues, studies
32 have considered several exposure metrics based on different periods of exposure,
33 including both short- and long-term exposure periods. These studies are characterized
34 here as they contribute to the weight of evidence for an effect of NO2 on reproductive and
35 developmental effects.
36 Although the biological mechanisms are not fully understood, several hypotheses have
37 been proposed for the effects of NCh on reproductive and developmental outcomes; these
January 2015 6-116 DRAFT: Do Not Cite or Quote
-------
1 include oxidative stress, systemic inflammation, vascular dysfunction, and impaired
2 immune function. The study of these outcomes can be difficult given the need for
3 detailed exposure data and potential residential movement of mothers during pregnancy.
4 Air pollution epidemiologic studies reviewed in the 2008 ISA for Oxides of Nitrogen
5 (U.S. EPA. 2008) examined impacts on birth-related endpoints, including intrauterine,
6 perinatal, post-neonatal, and infant deaths; premature births; intrauterine growth
7 restriction; very low birth weight (weight <1,500 g) and low birth weight (weight
8 <2,500 g); and birth defects. However, in the limited number of studies included in the
9 2008 ISA for Oxides of Nitrogen, no associations were found between NO2 and birth
10 outcomes, with the possible exception of birth defects. Overall, the evidence evaluated in
11 the 2008 ISA for Oxides of Nitrogen was inconsistent and lacked coherence and
12 plausibility, and was determined to be inadequate to infer the presence or absence of a
13 causal relationship.
14 Several recent articles reviewed methodological issues relating to the study of outdoor air
15 pollution and adverse birth outcomes (Chen etal.. 2010; Woodruff etal.. 2009; Ritz and
16 Wilhelm. 2008; Slama et al.. 2008). Some of the key challenges to interpretation of these
17 study results include the difficulty in assessing exposure as most studies use existing
18 monitoring networks to estimate individual exposure to ambient air pollution; the
19 inability to control for potential confounders such as other risk factors that affect birth
20 outcomes (e.g., smoking); evaluating the exposure window (e.g., trimester) of
21 importance; and limited evidence on the physiological mechanism of these effects (Ritz
22 and Wilhelm. 2008: Slama etal.. 2008).
23 Overall, the number of studies examining the association between exposure to ambient
24 NO2 and reproductive and developmental outcomes has increased tremendously, yet
25 evidence for an association with these outcomes remains relatively uncertain. Recently,
26 an international collaboration was formed to better understand the relationships between
27 air pollution and adverse birth outcomes and to examine some of these methodological
28 issues through standardized parallel analyses in data sets from different countries
29 (Woodruff et al.. 2010). Initial results from this collaboration have examined PM and
30 birth weight (Parker et al.. 2011); work on NO2 has not yet been performed. Although
31 early animal studies (Shalamberidze and Tsereteli. 197la. b) found that exposure to NO2
32 during pregnancy in rats led to some abnormal birth outcomes, human studies to date
33 have reported inconsistent results for the association of ambient NO2 concentrations and a
34 range of reproductive and developmental outcomes, though the evidence has been
35 generally supportive for some particular outcomes (e.g., fetal growth).
January 2015 6-117 DRAFT: Do Not Cite or Quote
-------
6.4.2 Fertility, Reproduction, and Pregnancy
6.4.2.1 Effects on Sperm
1 A limited amount of research has been conducted to examine the association between air
2 pollution and male reproductive outcomes, specifically semen quality. To date, the
3 epidemiologic studies have considered various exposure durations before semen
4 collection that encompass either the entire period of spermatogenesis (i.e., 90 days) or
5 key periods of sperm development that correspond to epididymal storage, development of
6 sperm motility, and spermatogenesis.
7 An occupational study of male motorway company employees reported that men with the
8 highest NC>2 exposures in the workplace (near-road environment, -160 ppb) had lower
9 sperm motility, but no difference in sperm count, compared to men with lower exposures
10 [~80 ppb; (Boggia et al.. 2009)]. Two epidemiologic studies evaluated the relationship
11 between ambient concentrations of NO2 and sperm quality and observed no associations
12 (Rubes etal.. 2010; Sokol et al.. 2006): while a cross-sectional study observed
13 associations between NCh and some semen quality parameters (Zhou etal.. 2014). No
14 recent toxicological studies have examined the effect of NO2 exposure on male
15 reproductive outcomes, specifically semen quality. Kripke and Sherwin (1984) found no
16 significant effects on spermatogenesis, or on germinal and interstitial cells of the testes of
17 a small group of LEW/f mai rats (n = 6) after 21 days of exposure to a single
18 concentration of NC>2, 1,000 ppb 7 h/day, 5 days/week (Table 6-13). Overall, there is
19 little epidemiologic evidence for an association and no toxicological evidence of effects
20 of NO2 exposure on sperm or semen quality.
6.4.2.2 Effects on Reproduction
21 Several recent studies have examined the association between exposure to air pollutants
22 during pregnancy and the ability to reproduce. Gametes (i.e., ova and sperm) may be
23 even more at risk, especially outside of the human body, as occurs with assisted
24 reproduction. Smokers require twice the number of in vitro fertilization (FVF) attempts to
25 conceive as nonsmokers (Feichtinger et al.. 1997). suggesting that a preconception
26 exposure can be harmful to pregnancy. A recent study estimated daily concentrations of
27 criteria pollutants at addresses of women undergoing their first FVF cycle and at their FVF
28 labs from 2000 to 2007 in the northeastern U.S. (Legro et al., 2010). Increasing NO2
29 concentration at the patient's address during ovulation induction (short-term exposure,
30 -12 days) was associated with a decreased chance of live birth (OR: 0.80 [95% CI: 0.71,
January 2015 6-118 DRAFT: Do Not Cite or Quote
-------
1 0.91] per 10-ppb increase). Similar risks were observed when the exposure period was
2 the daily concentration averaged over the days from oocyte retrieval through embryo
3 transfer, and the days from embryo transfer through the pregnancy test (14 days). The
4 authors also observed a decreased odds of live birth when exposed from embryo transfer
5 to live birth [long-term exposure, -200 days; OR: 0.76 [95% CI: 0.56, 1.02] per 10-ppb
6 increase). After adjusting for Os in a copollutant model, NO2 continued to be associated
7 with IVF failure. The results of this study suggest that both short- and long-term exposure
8 to NO2 during ovulation and gestation was detrimental and reduced the likelihood of a
9 live birth. In a more general population, increased NO2 exposure in the 30 days before
10 initiation of unprotected intercourse also was associated with reduced fecundability
11 [fecundability ratio per 10 ppb: 0.50 [95% CI: 0.32, 0.76]) (SlamaetaL. 2013)1.
12 Similarly, in a cross-sectional study of fertility rates, Nieuwenhuijsen et al. (2014)
13 observed decreased fertility in areas with higher NO2 and NOx concentrations.
14 In contrast, NO2 exposure has not been shown to induce such effects in animals. Breeding
15 studies by Shalamberidze and Tsereteli (1971b) and Shalamberidze and Tsereteli (197la)
16 with exposures of animals to 67 or 1,300 ppb NO2 12 h/day for 3 months found that
17 long-term NO2 exposure had no effect on fertility; NO2 exposure produced no change in
18 the number of dams that became pregnant after mating with an unexposed male. At the
19 higher dose, Shalamberidze and Tsereteli (1971b) and Shalamberidze and Tsereteli
20 (1971a) did see impaired estrous cyclicity (cycle prolongation, increased duration of
21 diestrus, decreased number of normal and total estrus cycles), and the exposed females
22 had a decreased number of ovarian primordial follicles.
6.4.2.3 Effects on Pregnancy
Epidemiologic Evidence
23 Evidence suggests that exposure to air pollutants may affect maternal and fetal health
24 during pregnancy. One such health effect, systemic inflammation, has been proposed as a
25 potential biological mechanism through which air pollution could result in other adverse
26 pregnancy outcomes (Slamaetal.. 2008; Kannan et al.. 2006). Recent studies have
27 investigated the relationship between CRP, a marker for systemic inflammation,
28 measured in maternal blood during early pregnancy and in umbilical cord blood (as a
29 measure of fetal health) and the association with NO2 concentrations, van den Hooven
30 et al. (2012a) observed generally null associations between exposure to NO2 and elevated
31 maternal CRP levels but did observe a positive, linear relationship between quartiles of
32 NO2 exposure and elevated fetal CRP levels. This association was evident when exposure
33 was measured 1, 2, and 4 weeks prior to delivery but was strongest when exposure to
January 2015 6-119 DRAFT: Do Not Cite or Quote
-------
1 NC>2 was measured over the entire pregnancy. Similarly, Lee et al. (2011 a) observed
2 generally null associations between short-term exposure (i.e., 1 to 29 days) to NC>2 and
3 elevated maternal CRP levels.
4 Pregnancy-associated hypertension is a leading cause of perinatal and maternal mortality
5 and morbidity. A large body of research has linked changes in blood pressure to ambient
6 air pollution; however, evidence is inconsistent for NC>2 (see Sections 5.3.5 and 6.3.5). A
7 few recent studies have examined whether increases in NO2 concentrations are associated
8 with blood pressure changes in women who are pregnant. The results of these studies
9 were not consistent. Hampel et al. (2011) observed that increases in NO2 were associated
10 with decreases in systolic blood pressure but found no clear associations between NO2
11 concentrations and diastolic blood pressure. Lee et al. (2012a) observed associations
12 between exposure to NCh and changes in blood pressure that were null for the entire
13 population and when the population was restricted to nonsmokers. van den Hooven et al.
14 (2011) observed small increases in systolic blood pressure associated with increases in
15 NO2 concentrations across all three trimesters of pregnancy but did not observe a similar
16 association with diastolic blood pressure. Mobasher et al. (2013) observed a positive
17 association between exposure to NO2 during the first trimester and hypertensive disorders
18 of pregnancy, though the association was imprecise and was reduced when exposure was
19 averaged over the second and third trimesters. The same pattern was observed when
20 analyses were restricted to nonobese women, but among obese women, the effect
21 estimate was below 1.00 for each trimester. Xu et al. (2014) observed positive
22 associations between NC>2 exposure during the entire pregnancy and first trimester and
23 hypertensive disorders of pregnancy; associations remained positive after adjustment for
24 O3, CO, SO2, or PM2 5.
25 New-onset gestational hypertension can contribute to pre-eclampsia, a common
26 pregnancy complication diagnosed after 20 weeks of pregnancy. Wu et al. (2009)
27 observed a 45% increase (95% CI: 23%, 64%) in the risk of pre-eclampsia associated
28 with a 20-ppb increase in NOx measured over the entire pregnancy; when the exposure
29 was examined categorically, the association between pre-eclampsia risk and NOx
30 concentration was consistent with a linear concentration-response relationship. Similarly,
31 NO2 concentrations during pregnancy were associated with an increased risk of
32 pre-eclampsia among a cohort of Australian women (Pereiraetal.. 2013). with the
33 strongest association observed when exposure was limited to the third trimester.
34 Malmqvist et al. (2013) also observed a positive association between NOx concentrations
35 in the third trimester of pregnancy and pre-eclampsia consistent with a linear
36 concentration-response relationship in a Swedish cohort. Dadvand et al. (2013) observed
37 increases in odds of pre-eclampsia, particularly late-onset pre-eclampsia, with increased
38 NO2 exposure during the third and first trimesters, and with entire pregnancy exposures.
January 2015 6-120 DRAFT: Do Not Cite or Quote
-------
1 A number of other studies, of similar study quality and using similar study designs, did
2 not observe positive associations for NO2 exposure and risks of pregnancy-induced
3 hypertension or pre-eclampsia across different exposure periods including exposure over
4 entire pregnancy (Nahidi et al.. 2014; van den Hooven et al.. 2011) and the first trimester
5 exposure (Olsson et al.. 2013)1. Exposure to NOx (Malmqvist et al.. 2013; Wu etal..
6 2009) or NO2 (Dadvand et al.. 2013; Pereira et al.. 2013) was estimated at each subject's
7 residential address using LUR or dispersion models. Details on these exposure
8 assessment techniques can be found in Sections 3.2.1.1 and 3.2.3. respectively.
9 Information in Section 3.4.5.2 aids interpreting these methods with regard to potential
10 exposure measurement error. A meta-analysis of pre-eclampsia studies reported a
11 combined OR for NO2 of 1.23 (95% CI: 1.04, 1.42), though there was a large amount of
12 heterogeneity between studies particularly in outcome definition; removal of an
13 influential study produced an OR of 1.11 (95% CI: 1.04, 1.17) with no observed
14 heterogeneity (Pedersen et al.. 2014).
15 Other pregnancy complications that have recently been evaluated and found to be
16 associated with NO2 include gestational diabetes (Malmqvist et al.. 2013) and markers of
17 placental growth and function (van den Hooven et al.. 2012c). Overall, the evidence for
18 the effects of NO2 on pregnancy effects is inconsistent. Key studies examining the
19 association between exposure to NO2 and pregnancy-related effects can be found in
20 Table 6-12. A supplemental Table S6-2 (U.S. EPA. 2013b) provides an overview of all of
21 the epidemiologic studies of pregnancy-related health effects.
January 2015 6-121 DRAFT: Do Not Cite or Quote
-------
Table 6-12 Key reproductive and developmental epidemiologic studies for
nitrogen dioxide (NO2).
Study
Location
Sample Size
Mean NO2a
or NOx
(PPb)
Exposure
Assessment
Selected Effect Estimates'3 (95% Cl)
Fertility, Reproduction, and Pregnancy
Pereira et al.
(2013)
Wu et al. (2009)
Malmqvist et al.
(2013)
Perth, Australia
N = 23,452
Southern
California
N = 81,186
Sweden
N = 81,110
NR
NOx:
Entire
pregnancy:
7.23
T1: 7.45
T2: 7.29
T3: 7.14
NOx:
7.5
LUR model
NOx from
dispersion
model.
NOx from
dispersion model
with a spatial
resolution of
Pre-eclampsia
T1: 1.04(0.94, 1.16)
T2: 1.02(0.91, 1.15)
T3: 1.17(1.04, 1.32)
Entire pregnancy: 1.22 (1
Pre-eclampsia
Entire pregnancy: 1.44 (1
Pre-eclampsia
Third trimester
Q1: (reference)
.02, 1.49)
.23, 1.68)
500 x 500 m.
Q2: 1.28(1.13, 1.46)
Q3: 1.33(1.17, 1.52)
Q4: 1.51 (1.32, 1.73)
Gestational diabetes
Third trimester
Q1: (reference)
Q2: 1.19(0.99, 1.44)
Q3: 1.52(1.28, 1.82)
Q4: 1.69(1.41,2.03)
Dadvand et al.
Barcelona, Spain T1: 30
N = 8,398 T2: 31
T3: 31
EP: 30
LUR model
Pre-eclampsia
T1: 1.07(0.94, 1.22)
T2: 1.03(0.90, 1.19)
T3: 1.11 (0.99, 1.23)
EP: 1.09(0.94, 1.27)
January 2015
6-122
DRAFT: Do Not Cite or Quote
-------
Table 6-12 (Continued): Key reproductive and developmental epidemiologic
studies for nitrogen dioxide (NO2).
Study
Leqro et al.
(2010)
Slama et al.
(2013)
Location
Sample Size
Northeastern U.S.
N = 7,403
Teplice, Czech
Republic
N = 1,916
Mean NO2a
or NOx Exposure
(ppb) Assessment
19 Spatially
interpolated
concentrations
from kriging
based on
monitoring data.
19 Central site
monitor (within
12 km) of
residence.
Selected Effect Estimates13 (95% Cl)
Odds of live birth following IVF
Medication start to oocyte retrieval: 0.80
(0.71,0.91)
Oocyte retrieval to embryo transfer: 0.87
(0.79, 0.96)
Embryo transfer to pregnancy test (14 days):
0.76 (0.66, 0.86)
Embryo transfer to live birth: 0.76 (0.56,
1.02)
Fecundity ratio
Last 30 days before unprotected intercourse
(Lag 1):
0.50 (0.32, 0.76)
Last 30 days before Lag 1 (Lag 2):
1.10(0.69, 1.80)
Lag 1 + Lag 2:
0.52 (0.28, 0.94)
Mo post-outcome:
1.17(0.76, 1.85)
Birth Outcomes
Aquilera et al.
(2010)
Catalonia, Spain
N = 562
16.9-17.2 LUR
Fetal length (% change)
T1: -2.04 (-7.01, 2.95)
T2: -1.69 (-7.05, 3.69)
T3: 0.33 (-4.06, 4.72)
Head circumference (% change)
T1: 0.25 (-5.42, 5.91)
T2: 1.70 (-3.69, 7.07)
T3: 0.23 (-4.32, 4.77)
Abdominal circumference (% change)
T1: -2.82 (-8.24, 2.59)
T2: -0.13 (-5.64, 5.38)
T3: 0.74 (-3.926, 5.40)
Biparietal diameter (% change)
T1: 3.87 (-2.04, 9.75)
T2: 4.90 (-0.34, 10.11)
T3: 1.48 (-3.41, 6.35)
Estimated fetal weight (% change)
T1: -2.22 (-7.39, 2.98)
T2: 0.46 (-5.82, 6.72)
T3: 0.91 (-3.65, 5.45)
January 2015
6-123
DRAFT: Do Not Cite or Quote
-------
Table 6-12 (Continued): Key reproductive and developmental epidemiologic
studies for nitrogen dioxide (NO2).
Study
Location
Sample Size
Mean NO2a
or NOx
(PPb)
Exposure
Assessment
Selected Effect Estimates'3 (95% Cl)
Ballesteret al.
Valencia, Spain
N = 785
19.1-20.2 LUR
Head circumference (cm)
Entire pregnancy: -0.11 (-0.25, 0.03)
Birth length (cm)
Entire pregnancy: -0.09 (-0.27, 0.10)
SGA—weight
Entire pregnancy: 1.59 (0.89, 2.84)
SGA—length
Entire pregnancy: 1.48 (0.628, 3.49)
Estarlich et al.
(2011)
Hansen et al.
(2007)
Hansen et al.
(2008)
Asturias,
Gipuzkoa,
Sabadell,
Valencia, Spain
N= 2,337
Brisbane,
Australia
N= 26,617
Brisbane,
Australia
N = 15,623
Overall:
15.5
Urban: 15.9
Rural: 8.7
Median: 7.8
75th: 11.4
Max: 24.2
9.8
LUR
Central sites,
city-wide avg.
Closest central
site monitor
(within 2-14 km
of 1 of
17 monitors).
Birth length (cm)
Entire pregnancy: -1.69 (-0.34,
Head circumference (cm)
Entire pregnancy: -0.01 (0.13, 0
Head circumference (cm)
T1: 0.05 (-0.05, 0.17)
T2: 0.08 (-0.02, 0.19)
T3: 0.00 (-0.10, 0.10)
Crown-heel length (cm)
T1: 0.24(0.05,0.42)
T2: 0.07 (-0.1 0,0.24)
T3: -0.1 5 (-0.25, -0.05)
Head circumference (mm)
M1: 0.54 (-1.88, 2.94)
M2: -0.16 (-2.54, 2.20)
M3: -0.60 (-3.18, 2.00)
-0.02)
.11)
M4: -0.30 (-2.30, 1.68)
Biparietal diameter (mm)
M1: 0.14 (-0.62, 0.88)
M2: -0.20 (-0.88, 0.50)
M3: -0.12 (-0.82, 0.58)
M4: -0.16 (-0.74, 0.42)
Abdominal circumference (mm)
M1: 0.48 (-1.98, 2.94)
M2: 0.98 (-1.40, 3.34)
M3: 0.20 (-2.12, 2.52)
M4: 0.30 (-1.80, 2.40)
Femur length (mm)
M1: 0.06 (-0.50, 0.62)
M2: -0.18 (-0.78, 0.44)
M3: 0.02 (-0.52, 0.56)
M4: -0.26 (-0.80, 0.26)
January 2015
6-124
DRAFT: Do Not Cite or Quote
-------
Table 6-12 (Continued): Key reproductive and developmental epidemiologic
studies for nitrogen dioxide (NO2).
Study
Iniquez et al.
(2012)
Location
Sample Size
Valencia, Spain
N = 818
Mean NO2a
or NOx
(PPb)
Median:
20.2
Exposure
Assessment
LUR model
Selected Effect Estimates'3
Fetal length (% change)
T1: 0.97(0.92, 1.02)
(95% Cl)
T2: 0.96(0.92, 1.00)
T3: 0.97(0.92, 1.02)
Abdominal circumference (% change)
T1: 0.96(0.92, 0.99)
T2: 0.98(0.94, 1.02)
T3: 0.98(0.94, 1.03)
Biparietal diameter (% change)
T1: 0.96(0.92, 1.00)
T2: 0.97(0.92, 1.01)
T3: 0.98(94, 1.02)
Estimated fetal weight (% change)
T1: 0.96(0.92, 1.00)
T2: 0.98(0.94, 1.02)
T3: 0.97(0.93, 1.02)
van den Hooven Rotterdam, Mean: 21. 2 Combination of
etal. (2012b) Netherlands Median' continuous
l\l = 7 772 21 1 monitoring data
v^th. oo / and GIS-based
75th. 22.4 dispersion
Max: 30.3 modeling
techniques.
Head circumference (mm)
T3:
Q1: (reference)
Q2: -0.40 (-1.00, 0.20)
Q3: -0.81 (-1.42, -0.20)
Q3: -1.28 (-1.96, -0.61)
Length (mm)
T3:
Q1: Ref
Q2: -0.02 (-0.17, 0.13)
Q3: -0.09 (-0.24, 0.06)
Q4: -0.33 (-0.50, -0.16)
SGA
Entire pregnancy:
Q1 : ref
Q2: 0.93(0.66, 1.31)
Q3: 1.25(0.90, 1.73)
Q4: 1.35(0.94, 1.94)
Ritz etal. (2014) Los Angeles, CA
LUR: n = 501;
Monitors: n = 98
22.7-39.3
Biparietal diameter (mm)
LUR model GW 0-19:-0.41 (-1.07,0.23)
GW 19-29: 0.39 (-0.25, 1.02)
GW 29-37: -0.50 (-1.23, 0.23)
Central site
monitors
combined by
IDW.
GW 0-19:-4.45 (-10.55, 1.55)
GW 19-29: 4.92 (0.03, 9.83)
GW 29-37: -8.33 (-13.83, -2.83)
January 2015
6-125
DRAFT: Do Not Cite or Quote
-------
Table 6-12 (Continued): Key reproductive and developmental epidemiologic
studies for nitrogen dioxide (NO2).
Study
Bell et al. (2007)
Darrow et al.
(2011)
Mean NO2a
Location or NOx Exposure
Sample Size (ppb) Assessment
Connecticut and 17.4 County-level
Massachusetts average of
N = 358,504 central site
monitors.
Atlanta, GA 23.6 Population-
N = 406,627 weighted spatial
average.
Selected Effect Estimates13 (95% Cl)
Birth weight (g)
Entire Pregnancy: -18.54 (-22.50, -14.58)
Black mothers: -26.46 (-37.50, -15.63)
White mothers: -17.29 (-21.67 -13.13)
LEW: 1.06(1.00, 1.11)
Birth weight (g)
Entire pregnancy: -18.40 (-28.00, -9.00)
First 28 days: 0.8 (-3.60, 5.20)
T3: -9.00 (-17.00, -1.20)
Non-Hispanic white: -9.20 (-18.60, 0.20)
Non-Hispanic black: -7.8 (-17.40, 1.60)
Hispanic: -11.60 (-24.80, 1.40)
Postnatal Development
Guxens et al.
(2014)
Ruhr, Germany Medians LUR — home
NR Prenatal
exposure
Heraklion, Greece 6.1
Asturias, Spain NR
Gipuzkoa, Spain NR
Valencia, Spain NR
Sabadell, Spain 23.4
Granada, Spain NR
Rotterdam, NR
Netherlands
Poitiers, France NR
Nancy, France NR
Rome, Italy NR
Change in cognitive or motor function score
Mental development: -3.61 (-8.53, 1.32)
Motor function: -5.04 (-11, 0.49)
Mental development: 1.90 (-2.33, 6.13)
Motor function: -0.83 (-5.39, 3.74)
Mental development: -1.39 (-3.12, 0.34)
Motor function: -2.03 (-3.82, -0.24)
Mental development: -1.11 (-6.65, 4.44)
Motor function: 0.17 (-1.73, 5.34)
General cognition: -1.35 (-3.74, 1.03)
Motor function: -3.72 (-6.37, -1.07)
General cognition: -0.15 (-2.42, 2.12)
Motor function: 0.71 (-1.71, -3.14)
General cognition: 3.18 (-0.26, 6.62)
Motor function: 1.80 (-1.73, 5.34)
Motor function: -0.1 7 (-1.60, 1.26)
Motor function: -0.64 (-6.75, 5.47)
Motor function: -2.84 (-5.64, -0.04)
Motor function: -1.97 (-4.44, 0.49)
January 2015
6-126
DRAFT: Do Not Cite or Quote
-------
Table 6-12 (Continued): Key reproductive and developmental epidemiologic
studies for nitrogen dioxide (NO2).
Study
Location
Sample Size
Mean NO2a
or NOx
(PPb)
Exposure
Assessment
Selected Effect Estimates'3 (95% Cl)
Guxens et al.
(2014)
(continued)
Cohorts combined NR
Motor function
All subjects: -1.28 (-2.35, -0.21)
Birth dates temporally matched with NO2
monitoring: -0.79 (-1.84, 0.26)
Freire et al.
(2010)
Spain
N=210
11.1
LUR—home
Concurrent
exposure
Change in score in group with
NO2 >13.2 ppb compared with group with
NO2<8.2 ppb
General cognitive index: -4.19 (-14.02, 5.64)
Verbal:-3.09 (-13.31, 7.13)
Quantitative: -6.71 (-17.91, 4.49)
Memory: -5.52 (-16.18, 5.13)
Executive function: -4.93 (-14.90, 5.05)
Gross motor function: -8.61 (-18.96, 1.74)
Fine motor skills: 0.91 (-10.22, 12.05)
van Kempen the Netherlands School: LUR—school Change in score, adjusted for traffic noise
etal. (2012) N = 485 16-5 Concurrent Memory:-0.30 (-0.55, 0.04)
Home: 16.4 exposure Measures of attention
SRTT, reaction time (ms):
-2.23 (-22.1, 17.7)
SAT block, # errors: -0.02 (-0.42, 0.38)
SAT block, reaction time (ms):
13.9 (-16.7, 43.9)
SAT switch, # errors: -1.19 (-3.62, 1.26)
SAT switch, reaction time (ms):
21.5 (-45.2, 88.2)
Locomotion: 0.08 (-0.08, 0.25)
LUR—home Change in score, adjusted for traffic noise
Concurrent Memory: 0.17 (-0.08, 0.42)
exposure Measures of attention
SRTT, reaction time (ms):
-2.11 (-21.0, 16.7)
SAT block, # errors: -0.04 (-0.40, 0.32)
SAT block, reaction time (ms):
15.9 (-11.3, 43.0)
SAT switch, # errors: -1.23 (-3.32, 0.87)
SAT switch, reaction time (ms):
-20.2 (-74.9, 34.5)
Locomotion: 0.06 (-0.08, 0.21)
January 2015
6-127
DRAFT: Do Not Cite or Quote
-------
Table 6-12 (Continued): Key reproductive and developmental epidemiologic
studies for nitrogen dioxide (NO2).
Mean NO2a
Location or NQx
Study Sample Size (ppb)
Clark etal. U.K. 22.7
(2012) N = 719
Guxens et al. Spain Overall:
(2012) N = 1,889 15-4
Valencia:
19.6
Sabadell:
17.1
Asturias:
12.3
Gipuzkoa:
10.7
Volketal. (2013) California NOX:
n = 524 Q1:<9.7
Q2:
9.7-16.9
Q3:
16.9-31.8
Q4: Ł31.8
Exposure
Assessment
LUR— school
Concurrent
exposure
LUR
Prenatal
exposure
NOx dispersion
model (within
5 km of child's
home).
NO2 from central
sites within
50 km of homes,
combined by
IDW.
Selected Effect Estimates13 (95% Cl)
Change in score, adjusted for traffic noise
Reading comprehension:
0.08 (-0.17, 0.34)
Information recall:
0.28 (-0.62, 1.17)
Working memory:
0.06 (-5.55, 5.66)
Physiological distress:
0.47 (-0.62, 1.56)
Change in mental development indexc
Location
All regions: -0.95 (-3.90, 1.89)
Gipuzkoa: -5.15 (-8.04, -2.27)
Asturias: 0.17 (-2. 71, 3.04)
Sabadell: 1.98 (-1.69, 5.66)
Valencia: -0.43 (-2.86, 2.01)
Maternal fruit and vegetable intake
<405 g/day: -4.13 (-7.06, -1.21)
>405g/day: 0.25 (-3.63, 4.12)
Maternal Vitamin D circulation
Low: -2.49 (-6.87, 1.89)
Medium: -0.55 (-3.48, 2.39)
High: -0.11 (-2.72,2.49)
Odds ratio for autism relative to Q1
First yr of life
Q2: 0.91 (0.56, 1.47)
Q3: 1.00(0.62, 1.62)
Q4: 3.10(1.76, 5.57)
Average prenatal
Q2: 1.26(0.77,2.06)
Q3: 1.09(0.67, 1.79)
Q4: 1.98(1.20, 3.31)
Odds ratio for autism
First yr of life: 1.67(1.25,2.23)
Average prenatal: 1.52 (1.16, 2.00)
January 2015
6-128
DRAFT: Do Not Cite or Quote
-------
Table 6-12 (Continued): Key reproductive and developmental epidemiologic
studies for nitrogen dioxide (NO2).
Study
Location
Sample Size
Mean NO2a
or NOx
(PPb)
Exposure
Assessment
Selected Effect Estimates'3 (95% Cl)
Becerra et al.
California
n = 83,385
30.8
LUR—home Odds ratio for autism
Average prenatal: 1.05 (0.98, 1.12)
T1: 1.03(0.98, 1.08)
T2: 1.03(0.98, 1.08)
T3: 1.04(0.98, 1.09)
Nearest central
site monitor to
birth residence.
Average prenatal: 1.04 (0.98, 1.10)
T1: 1.04(0.99, 1.08)
T2: 1.01 (0.97, 1.06)
T3: 1.02(0.97, 1.07)
Cl = confidence interval; EP = entire pregnancy; GIS = geographic information system; GW = gestational week; IDW = inverse
distance weighting; IVF = in vitro fertilization; LEW = low birth weight; LUR = land-use regression; M1 = Month 1; M2 = Month 2;
M3 = Month 3; M4 = Month 4; NO2 = nitrogen dioxide; NOX = sum of NO and NO2; NR = No quantitative results reported; Q1 = first
quartile; Q2 = second quartile; Q3 = third quartile; Q4 = fourth quartile; SAT = Switching Attention Test; SGA = small for gestational
age; SRTT = Simple Reaction Time Test; T1 = first trimester; T2 = second trimester; T3 = third trimester.
aNO2 unless otherwise specified.
"•Relative risk per 10-ppb change in NO2, or 20-ppb change in NOX, unless otherwise noted.
°Per doubling in NO2 concentration estimated using IDW.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
Toxicological Evidence
Evidence from animal toxicological studies suggests that exposure to NO2 may affect
pregnancy. Maternal toxicity or deficits in maternal weight gain during gestation was
reported in pregnant rats with inhalation exposure to 5,300 ppb NC>2 for 6 h/day
7 days/week throughout gestation [21 days; (Tabacova et al.. 1984)]. Another study
reported dam weight gain over pregnancy as a percentage of body weight at conception
or GDO, which is an unusual metric, and found no differences in maternal weight gain
with 1,500 or 3,000 ppb NC>2 exposure over the duration of pregnancy (Di Giovanni
etal.. 1994).
Fetal lethality in toxicological studies is measured by counting pup loss or resorption
sites. This directly affects litter size, or number of live pups born. Shalamberidze and
Tsereteli (1971b) and Shalamberidze and Tsereteli (197la) reported decreased litter sizes
(fewer pups born) to dams that received 1,300 ppb NCh 12 h/day for 3 months during
pregnancy. Litter size was not affected in dams exposed to 1,500 or 3,000 ppb NO2
exposure over the duration of pregnancy (Di Giovanni et al.. 1994).
January 2015
6-129
DRAFT: Do Not Cite or Quote
-------
6.4.3 Birth Outcomes
6.4.3.1 Fetal Growth
1 Fetal growth is influenced by maternal, placental, and fetal factors. The biological
2 mechanisms by which air pollutants may influence the developing fetus remain largely
3 unknown. LEW has often been used as an outcome measure because it is easily available
4 and accurately recorded on birth certificates. However, LEW may result from either short
5 gestation or inadequate growth in utero. Most of the studies investigating air pollution
6 exposure and LEW limited their analyses to term infants to focus on inadequate growth.
7 A number of studies were identified that specifically addressed growth restriction in utero
8 by identifying infants who failed to meet specific growth standards. Usually, these infants
9 had birth weight less than the 10th percentile for gestational age, using an external
10 standard.
11 A limitation of environmental studies that use birth weight as a proxy measure of fetal
12 growth is that patterns of fetal growth during pregnancy cannot be assessed. This is
13 particularly important when investigating pollutant exposures during early pregnancy as
14 birth weight is recorded many months after the exposure period. The insult of air
15 pollution may have a transient effect on fetal growth, where growth is hindered at one
16 point in time but catches up at a later point. For example, maternal smoking during
17 pregnancy can alter the growth rate of individual body segments of the fetus at variable
18 developmental stages, as the fetus experiences selective growth restriction and
19 augmentation (Lampl and Jeanty. 2003).
20 The terms SGA, which is defined as a birth weight <10th percentile for gestational age
21 (and often sex and/or race), and IUGR are often used interchangeably. However, this
22 definition of SGA does have limitations. For example, using it for IUGR may
23 overestimate the percentage of "growth-restricted" neonates as it is unlikely that 10% of
24 neonates have growth restriction (Wollmann. 1998). On the other hand, when the 10th
25 percentile is based on the distribution of live births at a population level, the percentage
26 of SGA among PTB is most likely underestimated (Hutcheon and Platt 2008).
27 Nevertheless, SGA represents a statistical description of a small neonate, whereas the
28 term IUGR is reserved for those with clinical evidence of abnormal growth. Thus, all
29 IUGR neonates will be SGA, but not all SGA neonates with be IUGR rWollmann. 1998).
30 In the following section, the terms SGA and IUGR are referred to as each cited study
31 used the terms.
32 The 2008 ISA for Oxides of Nitrogen reviewed three studies that evaluated the
33 relationship between exposure to NO2 and fetal growth (Mannes et al., 2005; Salam et al.,
January 2015 6-130 DRAFT: Do Not Cite or Quote
-------
1 2005; Liu et al., 2003) and concluded that they "did not consistently report associations
2 between NCh exposure and intrauterine growth retardation" [(U.S. EPA. 2008); p. 3-73].
3 In recent years, a number of studies have examined various metrics of fetal growth
4 restriction. Several of these more recent studies have used anthropometric measurements
5 (e-g-, head circumference, abdominal circumference) measured via ultrasound at different
6 periods of pregnancy in order to evaluate patterns of fetal growth during pregnancy and
7 to detect growth restriction that may occur early in pregnancy, but which may no longer
8 be detectable at birth. In a mother and child cohort study conducted in Spain, ultrasound
9 measurements were recorded at 12, 20, and 32 weeks of gestation, and these
10 anthropometric measurements were recorded again at birth (Iniguez et al., 2012; Aguilera
11 etal.. 2010). Aguilera et al. (2010) observed that exposure to NO2 early in pregnancy was
12 associated with impaired growth in head circumference from Weeks 12 to 20 of gestation
13 and abdominal circumference and estimated fetal weight from Weeks 20 to 32. Similarly,
14 Iniguez et al. (2012) reported decreased fetal length and decreased biparietal diameter
15 measured by ultrasound in association with exposure to NC>2 during Weeks 12-20 of
16 gestation. Decreased birth length and head circumference measured at birth were also
17 associated with exposure to NO2 during this same period. Examining fetal growth
18 characteristics assessed by ultrasound during each trimester of pregnancy, van den
19 Hooven et al. (2012b) observed decreases in head circumference and fetal length in the
20 second and third trimesters associated with exposure to NO2. Hansen et al. (2008) used
21 ultrasound measurements during Weeks 13-26 of pregnancy and did not observe
22 associations between exposure to relatively low concentrations of NC>2 (mean = 9.8 ppb)
23 and head circumference, biparietal diameter, abdominal circumference, or fetal length.
24 Ritz etal. (2014) used multiple ultrasound measures to examine fetal growth parameters
25 across gestation and observed that higher exposure to NO2 during Gestational Weeks
26 29-37 was associated with decrements in biparietal diameter at 37 weeks; no consistent
27 associations were found for head circumference, femur length, or abdominal
28 circumference.
29 Several studies made use of anthropometric measurements made immediately after birth
30 to evaluate fetal growth. Estarlich et al. (2011), Ballester et al. (2010). and Hansen et al.
31 (2007) observed decreases in body length associated with exposure to NO2. This
32 association persisted when NC>2 exposure was estimated for each trimester of pregnancy
33 in the study by Estarlich et al. (2011). Ballester et al. (2010) observed the strongest
34 association with NC>2 exposure during the first trimester, while Hansen et al. (2007)
35 reported that the association was strongest for NC>2 exposure measured at the end of the
36 pregnancy.
January 2015 6-131 DRAFT: Do Not Cite or Quote
-------
1 When using SGA as an indicator of fetal growth restriction, several studies observed
2 associations with exposure to NCh, NOx, or NO (Sathyanarayana et al.. 2013; Le et al..
3 2012; Pereira et al.. 2012; Malmqvist et al.. 2011; Ballester etal.. 2010; Rich et al.. 2009;
4 Brauer et al.. 2008; Mannes et al.. 2005). These associations were most often observed
5 for exposure to NC>2 during the second trimester (Pereira et al.. 2012; Ballester et al..
6 2010: Rich et al.. 2009: Mannes etal.. 2005). Gehring et al. (2011 a). Hansen et al.
7 (2007). Olssonetal. (2013). Kashimaet al. (2011). and Hannametal. (2014) did not
8 observe an increased risk of SGA associated with exposure to NC>2. All of the studies that
9 used IUGR as an indicator of fetal growth restriction observed an association with
10 exposure to NCh, and this association was strongest for exposures at the beginning of
11 pregnancy [i.e., first month or first trimester; (Liu et al.. 2007: Salam et al.. 2005: Liu
12 et al.. 2003)]. Generally, studies of fetal growth restriction did not examine the potential
13 for confounding by traffic-related copollutants.
14 When evaluating the association between fetal growth and exposure to NO2, many
15 studies relied on modeled concentrations of NO2 at maternal residence coming from LUR
16 models (Iniguez et al.. 2012: Pereira et al.. 2012: Estarlich et al.. 2011: Gehring et al..
17 201 la: Aguilera et al.. 2010: Ballester etal.. 2010: Brauer et al.. 2008) and emissions or
18 dispersion models (van den Hooven et al.. 2012b: Malmqvist et al.. 2011). Generally, the
19 results of studies that relied on estimates of NO2 from LUR models were not substantially
20 different from those that estimated exposure to NC>2 using concentrations measured at
21 central site monitors. However, in a study that assigned exposure to NCh using both a
22 LUR model and IDW of measured NC>2 concentration from monitors, Brauer et al. (2008)
23 found higher risks for SGA using the monitoring data (OR: 1.28 [95% CI: 1.18, 1.36])
24 compared to the risks observed with the NO2 estimates from the LUR model (OR 0.94
25 [95% CI: 0.80, 1.10]). Given the differences among the study designs, it cannot be
26 concluded that the inconsistencies are related to exposure assessment method or length of
27 follow-up periods. In general, studies using central site monitors for exposure estimates
28 carry uncertainty in long-term NO2 exposure studies because the exposure error resulting
29 from spatial misalignment between subjects' and monitor locations can overestimate or
30 underestimate associations with health effects (Section 3.4.5.2).
31 Several studies were able to incorporate data on activity patterns in order to help reduce
32 uncertainty related to exposure assessment. Some analyses attempted to decrease the
33 potential exposure measurement error associated with exposure to ambient NO2 by
34 limiting inclusion to subjects that spent 15 or more hours per day at home or subjects that
35 spent less than 2 hours a day in an outdoor environment other than at their primary
36 residence. In such analyses, Aguilera et al. (2010) and Estarlich et al. (2011) found
37 stronger associations between measures of decreased fetal growth and exposure to NO2.
38 In contrast, when Gehring et al. (201 la) limited their analyses to participants that did not
January 2015 6-132 DRAFT: Do Not Cite or Quote
-------
1 move during pregnancy or did not have paid employment outside of the home, there were
2 no consistent associations between SGA and exposure to NO2.
3 In summary, there is generally consistent evidence for an association between exposure
4 to NO2 and fetal growth restriction, including recent evidence from studies that have used
5 fetal anthropometric measurements made via ultrasound and anthropometric
6 measurements made immediately after birth. These are consistent with the studies of the
7 clinical measurement of IUGR and the statistical definition of SGA. The evidence is less
8 certain when it comes to assessing the time period of pregnancy when exposure to NO2 is
9 associated with the highest risks. Some studies find the highest risks associated with NO2
10 when NO2 is measured in early pregnancy, while in other studies, the time period
11 associated with the greatest risk is toward the end of pregnancy. Others find the greatest
12 risk when exposure is assigned for the entire pregnancy period. Key studies examining
13 the association between exposure to NO2 and fetal growth effects can be found in
14 Table 6-12. A supplemental Table S6-3 (U.S. EPA. 2013c) provides an overview of all of
15 the epidemiologic studies of fetal growth effects.
6.4.3.2 Preterm Birth
16 PTB is a syndrome (Romero et al.. 2006) that is characterized by multiple etiologies. It is,
17 therefore, unusual to be able to identify an exact cause for each PTB. In addition, PTB is
18 not an adverse outcome in itself but an important determinant of health status
19 (i.e., neonatal morbidity and mortality). Although some overlap exists for common risk
20 factors, different etiologic entities related to distinct risk factor profiles and leading to
21 different neonatal and post-neonatal complications are attributed to PTB and measures of
22 fetal growth. Although both restricted fetal growth and PTB can result in LEW,
23 prematurity does not have to result in LEW or growth restricted babies.
24 A major issue in studying environmental exposures and PTB is selecting the relevant
25 exposure period because the biological mechanisms leading to PTB and the critical
26 periods of vulnerability are poorly understood (Bobak. 2000). Short-term exposures
27 proximate to birth may be most relevant if exposure causes an acute effect. However,
28 exposure occurring in early gestation might affect placentation, with results observable
29 later in pregnancy, or cumulative exposure during pregnancy may be the most important
30 determinant. The studies reviewed have dealt with this issue in different ways. Many
31 have considered several exposure metrics based on different periods of exposure. Often
32 the time periods used are the first month (or first trimester) of pregnancy and the
33 last month (or 6 weeks) prior to delivery. Using a time interval prior to delivery
34 introduces an additional problem because cases and controls are not in the same stage of
January 2015 6-133 DRAFT: Do Not Cite or Quote
-------
1 development when they are compared. For example, a preterm infant delivered at
2 36 weeks is a 32-week fetus 4 weeks prior to birth, while an infant born at term
3 (40 weeks) is a 36-week fetus 4 weeks prior to birth.
4 Recently, investigators have examined the association of PTB with both short-term
5 (i.e., hours, days, or weeks) and long-term (i.e., months or years) exposure periods.
6 Time-series studies have been used to examine the association between air pollution
7 concentrations during the days immediately preceding birth. An advantage of these
8 time-series studies is that this approach can remove the influence of covariates that vary
9 across individuals over a short period of time. Retrospective cohort and case-control
10 studies have been used to examine long-term exposure periods, often averaging air
11 pollution concentrations over months or trimesters of pregnancy.
12 Studies of PTB fail to show consistency in the periods during pregnancy when pollutants
13 are associated with an effect. For example, while some studies find the strongest effects
14 associated with exposures early in pregnancy, others report effects when the exposure is
15 limited to the second or third trimester. However, the effect of air pollutant exposure
16 during pregnancy on PTB has a biological basis. There is an expanding list of possible
17 mechanisms that may explain the association between NC>2 exposure and PTB.
18 Many studies of PTB compare exposure in quartiles, using the lowest quartile as the
19 reference (or control) group. No studies use a truly unexposed control group. If exposure
20 in the lowest quartile confers risk, then it may be difficult to demonstrate additional risk
21 associated with a higher quartile. Thus, negative studies must be interpreted with caution.
22 Preterm birth occurs both naturally (idiopathic PTB), and as a result of medical
23 intervention (iatrogenic PTB). Ritz et al. (2000) excluded all births by Cesarean section
24 to limit their studies to idiopathic PTB. No other studies attempted to distinguish the type
25 of PTB, although air pollution exposure maybe associated with only one type. This is a
26 source of potential effect misclassification. One study examined preterm premature
27 rupture of membranes, observing positive ORs with NC>2 exposure (Dadvand etal..
28 2014a).
29 A number of recent studies have evaluated the association between exposure to NO2 and
30 PTB, and the results have generally been inconsistent. The body of literature that has
31 observed an association between NO2 and PTB (Gehring et al.. 2014; Trasande et al..
32 2013: Le etal.. 2012: Olssonetal.. 2012: Wuet al.. 2011: Llop etal.. 2010: Darrow
33 et al.. 2009: Wu et al.. 2009: Jiang et al.. 2007: Leem etal.. 2006: Maroziene and
34 Grazuleviciene. 2002: Bobak. 2000) is generally the same (in both the quantity and
35 quality of studies) to those that find no consistent pattern in the association between NC>2
36 and PTB (Hannam etal.. 2014: Olsson etal.. 2013: Gehring etal.. 201 la: Gehring et al..
January 2015 6-134 DRAFT: Do Not Cite or Quote
-------
1 201 Ib; Kashima etal.. 2011; Basu etal.. 2010; Brauer etal.. 2008; Jalaludin et al.. 2007;
2 Ritzetal.. 2007; Hansen et al.. 2006; Liu etal.. 2003; Ritz et al.. 2000). Among the
3 studies that observe an association between exposure to NO2 and PTB, the association
4 seems to be strongest for exposure to NO2 late in pregnancy, including the third trimester
5 (Llop etal.. 2010; Leem etal.. 2006; Bobak. 2000). the last 8 weeks of pregnancy (Jiang
6 et al.. 2007). the last 6 weeks of pregnancy (Darrow et al.. 2009). month of birth
7 (Trasande et al.. 2013). or the last week of pregnancy (Olsson et al.. 2012).
8 Several studies examined very preterm birth (VPTB, <30 weeks gestation), and observed
9 positive associations with NO2 for VPTB when none were observed for PTB (Brauer
10 et al.. 2008). or observed stronger associations for VPTB compared to those for PTB (Wu
11 et al.. 201 l;Wu etal.. 2009).
12 When evaluating the association between PTB and exposure to NO2 at maternal
13 residence, several studies relied on estimates of NO2 concentrations coming from LUR
14 models (Gehring et al.. 2014; Gehring et al.. 201 la; Gehring et al.. 201 Ib; Kashima et al..
15 2011; Wu etal.. 2011; Llop et al.. 2010; Brauer et al.. 2008) and dispersion models (Wu
16 etal.. 2011; Wu etal.. 2009). Generally, the results of studies that relied on modeled
17 estimates of NO2 were similarly inconsistent, and not substantially different from those
18 that used measured NO2 concentrations. In a study that assigned exposure to NO2 using
19 both a LUR model and IDW of measured NO2 concentration from monitors, Brauer et al.
20 (2008) found generally comparable risk estimates per 10 ppb for VPTB using the
21 monitoring data (OR: 1.24, [95% CI: 0.80, 1.88]) and NO2 estimates from the LUR
22 model (OR: 1.16 [95% CI: 0.93, 1.61]). Given the differences among the study designs, it
23 cannot be concluded that the inconsistencies are related to exposure assessment method
24 or length of follow-up periods. In general, studies using central site monitors for exposure
25 estimates carry uncertainty in long-term NO2 exposure studies because the exposure error
26 resulting from spatial misalignment between subjects' and monitor locations can
27 overestimate or underestimate associations with health effects (Section 3.4.5.2).
28 In summary, the evidence is generally inconsistent, with some studies observing
29 associations between NO2, using both estimates from LUR models and concentrations
30 measured at central site monitors to estimate exposure, and PTB. These studies are
31 characterized in supplemental Table S6-4 (U.S. EPA. 2013d).
6.4.3.3 Birth Weight
32 With birth weight routinely collected in vital statistics and being a powerful predictor of
33 infant mortality, it is the most studied outcome within air pollution-birth outcome
January 2015 6-135 DRAFT: Do Not Cite or Quote
-------
1 research. Air pollution researchers have analyzed birth weight as a continuous variable
2 and/or as a dichotomized variable in the form of LEW [<2,500 g (5 Ibs, 8 oz)].
3 Birth weight is primarily determined by gestational age and intrauterine growth but also
4 depends on maternal, placental, and fetal factors as well as on environmental influences.
5 In both developed and developing countries, LEW is the most important predictor for
6 neonatal mortality and is an important determinant of post-neonatal mortality and
7 morbidity. Studies report that infants who are smallest at birth have a higher incidence of
8 diseases and disabilities, which continue into adulthood (Hack and Fanaroff. 1999).
9 A number of recent studies have evaluated the association between exposure to NO2 and
10 birth weight, and the results have generally been inconsistent. When examining birth
11 weight as a continuous variable, several studies have observed decreases in birth weight
12 associated with increases in NCh exposure (Gehring et al.. 2014; Laurent et al.. 2014;
13 Savitz etal.. 2014; Darrow etal.. 2011; Estarlich etal.. 2011; Ballester etal.. 2010;
14 Morello-Frosch et al.. 2010; Bell etal.. 2007). Generally, these studies observed the
15 largest decreases in birth weight when exposure to NO2 was averaged over the entire
16 pregnancy. There were also a number of studies that examined birth weight as a
17 continuous variable that found no consistent decreases in birth weight associated with
18 increases in NC>2 exposure averaged over the entire pregnancy or specific trimesters of
19 pregnancy (Hannam et al.. 2014; Sellier et al.. 2014; Pedersen et al.. 2013; Geeretal.
20 2012; Rahmalia etal.. 2012; Gehring etal.. 2011 a; Gehring etal.. 20 lib; Kashima et al..
21 2011; Lepeule etal..2010; Aguilera etal.. 2009; Hansen et al.. 2007; Salam etal.. 2005;
22 Gouveia et al.. 2004). With low birth weight examined as the risk of having a baby
23 weighing less than 2,500 g, the study results remained inconsistent, with some study
24 authors observing an association between LEW and exposure to NC>2 (Dadvand etal..
25 2014c; Ebisu and Bell. 2012; Ghosh etal.. 2012a; Wilhelm etal.. 2012; Morello-Frosch
26 etal.. 2010; Brauer et al.. 2008; Bell et al.. 2007; Lee et al.. 2003). while others reported
27 no consistent association (Pedersen et al.. 2013; Kashima et al.. 2011; Slama et al.. 2007;
28 Salametal.. 2005; Wilhelm and Ritz. 2005; Gouveia et al.. 2004; Liu et al.. 2003;
29 Maroziene and Grazuleviciene. 2002; Bobak. 2000). One study observed decreases in
30 effect estimates for both LEW and change in birth weight with increases in NO2 exposure
31 (Laurent et al.. 2013). Generally, the studies that observed the largest risks for LEW
32 averaged exposure to NO2 over the entire pregnancy.
33 Several studies were able to incorporate data on activity patterns in order to help reduce
34 uncertainty related to exposure assessment based on maternal residence. In analyses
35 limited to subjects that spent 15 or more hours per day at home or subjects that spent less
36 than 2 hours a day in an outdoor environment other than at their primary residence,
37 Estarlich et al. (2011) found stronger associations between birth weight and exposure to
January 2015 6-136 DRAFT: Do Not Cite or Quote
-------
1 NC>2 at the maternal residence. These sensitivity analyses did not consistently change the
2 associations observed by (Aguilera et al.. 2009). When Gehring et al. (201 la) limited
3 their analyses to participants that did not move during pregnancy, or did not have paid
4 employment outside of the home, they continued to observe no consistent associations
5 between birth weight and exposure to NC>2.
6 When evaluating the association between birth weight and exposure to NO2, several
7 studies relied on estimated residential concentrations of NO2 coming from LUR models
8 (Ghosh etal.. 2012a: Wilhelm etal.. 2012: Estarlich etal.. 2011: Gehring etal.. 201 la:
9 Gehring etal.. 20 lib: Kashima etal.. 2011: Ballester etal.. 2010: Lepeule etal.. 2010:
10 Aguilera et al.. 2009: Brauer et al.. 2008: Slama et al.. 2007) and dispersion models
11 (Rahmalia et al.. 2012: van den Hooven et al.. 2012c: Madsen et al.. 2010). Generally, the
12 results of studies that relied on estimates of NO2 from LUR models were similarly
13 inconsistent, and not substantially different from those that used NC>2 concentrations
14 measured at central site monitors. Several studies compared the use of statistical models
15 and the use of routinely collected monitoring data to assign exposure to NO2, and
16 concluded that while the monitoring data may include larger errors in estimated exposure,
17 these errors had little impact on the association between exposure to NO2 and birth
18 weight calculated using the two different methods for exposure assessment (Lepeule
19 etal.. 2010: Madsen etal.. 2010). Details on deriving exposure estimates using LUR and
20 dispersion models can be found in Sections 3.2.1.1 and 3.2.3. respectively. Given the
21 differences among the study designs, it cannot be concluded that the inconsistencies are
22 related to exposure assessment method or length of follow-up periods. In general, studies
23 using central site monitors for exposure estimates carry uncertainty in long-term NCh
24 exposure studies because the exposure error resulting from spatial misalignment between
25 subjects' and monitor locations can overestimate or underestimate associations with
26 health effects (Section 3.4.5.2). However, an association was observed with residential
27 NO2 exposure estimated for subjects' homes, and the improved spatial resolution of the
28 exposure estimate lends more confidence in the association.
29 In animal toxicological studies by Shalamberidze and Tsereteli (1971b) and
30 Shalamberidze and Tsereteli (197la), albino rats with exposures to 67 or 1,300 ppb NC>2
31 12 h/day for 3 months prior to breeding produced pups with significantly decreased birth
32 weights. These body-weight decrements continued to be significantly decreased at
3 3 post-natal day (PND)4 and PND12.
34 In summary, the evidence is generally inconsistent, with some studies observing
35 associations between NC>2 exposure and birth weight, while other studies observe no
36 consistent pattern of association. Key studies examining the association between
37 exposure to NCh and birth weight can be found in Table 6-12. A supplemental Table S6-5
January 2015 6-137 DRAFT: Do Not Cite or Quote
-------
1 (U.S. EPA. 2013e) provides an overview of all of the epidemiologic studies of birth
2 weight.
6.4.3.4 Birth Defects
3 Despite the growing body of literature evaluating the association between ambient air
4 pollution and various adverse birth outcomes, relatively few studies have investigated the
5 effect of temporal variations in ambient air pollution on birth defects. Heart defects and
6 oral clefts have been the focus of the majority of these recent studies, given their higher
7 prevalence than other birth defects and associated mortality. Mechanistically, air
8 pollutants could be involved in the etiology of birth defects via a number of key events.
9 A recent study investigated the association between NO or NO2 and cardiac birth defects
10 (Padula et al.. 2013a) and other noncardiac birth defects (Padula et al.. 2013b) in the San
11 Joaquin Valley in California. The authors observed no associations between heart defects
12 and NO or NO2 but did observe an association between neural tube defects and both NO
13 and NO2. In a further analysis of noncardiac/nonneural tube defects, Padula etal. (2013c)
14 observed no associations between NO or NO2 and any of the defects studied. A nine-state
15 cardiac birth defect case-control study observed associations between NO2 and
16 coarctation of the aorta, pulmonary valve stenosis, and left ventricular outflow tract
17 obstructions (Stingone et al.. 2014). A Barcelona, Spain-based case-control study of
18 18 congenital anomaly groups found coarctation of the aorta and digestive system defects
19 associated with increases in NO2 (Schembari et al.. 2014). Two studies examining
20 trisomy risk observed no correlations/associations with NO2, and a correlation between
21 NO and trisomy 21 (Chung etal.. 2014; Jurewicz et al.. 2014). In general, however,
22 studies of birth defects have focused on cardiac defects, and the results from these studies
23 are not entirely consistent. This inconsistency could be due to the absence of true
24 associations between NO2 and risks of cardiovascular malformations; it could also be due
25 to differences in populations, pollution concentrations, outcome definitions, or analytical
26 approaches. The lack of consistency of associations between NO2 and cardiovascular
27 malformations might be due to issues relating to statistical power or measurement error.
28 A recent meta-analysis of air pollution and congenital anomalies observed elevated
29 summary effects for NO2 and coarctation of the aorta (OR: 1.17 (1.00, 1.36)), Tetralogy
30 of Fallot (OR: 1.20 (1.02, 1.42)), and atrial septal defects (1.10 (0.91, 1.33)); ventral
31 septal defects exhibited an elevated summary estimate but also high heterogeneity
32 between studies (Vrijheid et al.. 2011). Another meta-analysis found association only
33 between coarctation of the aorta and NO2 (Chen etal.. 2014). These authors note that
34 heterogeneity in the results of these studies may be due to inherent differences in study
35 location, study design, and/or analytic methods, and comment that these studies have not
January 2015 6-138 DRAFT: Do Not Cite or Quote
-------
1 employed some recent advances in exposure assessment used in other areas of air
2 pollution research that may help refine or reduce this heterogeneity. These studies are
3 characterized in supplemental Table S6-6 (U.S. EPA. 2013F).
6.4.3.5 Early Life Mortality
4 An important question regarding the association between NC>2 and infant mortality is the
5 critical window of exposure during development for which infants are at risk. Several age
6 intervals have been explored: neonatal (<1 month); post-neonatal (1 month to 1 year);
7 and an overall interval for infants that includes both the neonatal and post-neonatal
8 periods (<1 year). During the neonatal and post-neonatal periods, the developing lung is
9 highly sensitive to environmental toxicants. The lung is not well developed at birth, with
10 80% of alveoli being formed post-natally. The studies below reflect a variety of study
11 designs, exposure periods, regions, and included relevant potential confounders. As
12 discussed below, a handful of studies have examined the effect of ambient air pollution
13 on neonatal and post-neonatal mortality, with the former the least studied. These studies
14 varied somewhat with regard to the outcomes and exposure periods examined and study
15 designs employed.
16 Overall, the evidence for an association between exposure to NO2 and infant mortality is
17 inconsistent. In an animal toxicological study, Tabacova et al. (1985) examined post-natal
18 development of rodent pups from dams that were exposed to 50, 500, or 5,300 ppb NCh
19 [6 h/day, 7 days/week, gestation day (GD)0-GD21J. Significantly decreased pup viability
20 was seen at PND21 with 5,300 ppb NC>2. Another study in which male pups received
21 prenatal exposure to NC>2 via a different daily exposure duration (dam continuous
22 exposure to 1,500 or 3,000 ppb, GDO-GD20) showed no significant effects on pup
23 post-natal mortality to PND21 (Di Giovanni et al., 1994). Recent epidemiologic studies
24 have examined the association between long-term exposure to NO2 measured at central
25 site monitors and stillbirths, with one study (Faiz et al.. 2012) observing an association
26 and another (Hwang et al.. 2011) observing associations near the null value. Faiz et al.
27 (2013) observed positive ORs for stillbirth with NCh exposures 2 days before birth. A
28 case-control study of spontaneous abortion before 14 weeks of gestation found a positive
29 OR for NO2 exposure (Moridi et al.. 2014). Houetal. (2014) also found positive ORs for
30 fetal loss before 14 weeks of gestation with NO2 exposure; however, these estimates have
31 large confidence intervals and may not be reliable. Enkhmaa et al. (2014) found high
32 correlations between air pollutants and fetal loss before 20 weeks in Mongolia, China,
33 including NO2, but these results were unadjusted for other factors or copollutants. One
34 study investigated the association between short-term exposure to NO2 and mortality
35 during the neonatal period (Lin et al.. 2004) and did not observe a positive association.
January 2015 6-139 DRAFT: Do Not Cite or Quote
-------
1 More studies have examined the association between exposure to NO2 and mortality
2 during the post-neonatal period. Son et al. (2008). Tsai et al. (2006). and Yang et al.
3 (2006) examined the association between short-term exposure to NC>2 and post-neonatal
4 mortality, while Ritz et al. (2006) investigated the association between long-term
5 exposure to NCh and post-neonatal mortality; none observed a consistent, positive
6 association. Finally, two studies examined the association between NO2 and sudden
7 infant death syndrome. Dales et al. (2004) and Ritz et al. (2006) observed positive
8 associations with short-term and long-term exposure to NC>2, respectively. Supplemental
9 Table S6-8 (U.S. EPA. 2013h) provides a brief overview of the epidemiologic studies of
10 infant mortality
6.4.4 Postnatal Development
11 The role of prenatal air pollution exposure has assumed increasing importance over time
12 for effects on post-natal development. Ambient air pollution exposures of pregnant
13 women have been associated with negative birth outcomes. Additionally, the prenatal and
14 early post-natal periods are critical periods for extensive growth and development, and air
15 pollution exposures during this period have been linked to health effects in the first years
16 of life. Thus, air pollution-related effects in both the developing fetus and infant have
17 implications for effects on post-natal development. This evaluation of the relationship of
18 post-natal developmental with NO2 exposure consists primarily of neurodevelopmental
19 outcomes and limited toxicological information on physical development. Studies
20 examining the effects of NC>2 exposure on development of the respiratory system
21 (Sections 6.2.6 and 6.2.7) inform the evaluation of the relationship between long-term
22 NC>2 exposure and respiratory effects (Section 6.2.9).
6.4.4.1 Neurodevelopmental Effects
23 Epidemiologic studies of neurodevelopment in children were not available for the 2008
24 ISA for Oxides of Nitrogen (U.S. EPA. 2008). but several have been published since
25 then. As described in the sections that follow, associations with NC>2 are inconsistent for
26 cognitive function, which was most extensively examined, and for attention-related
27 behaviors, motor function, psychological distress, and autism, which were examined in a
28 few studies each. Table 6-12 details the key studies, and Supplemental Table S6-7
29 (U.S. EPA. 2013g) provides an overview of all of the epidemiologic studies of
30 neurodevelopmental effects. Strengths of the studies overall are assessment of
31 neurodevelopment with widely used, structured neuropsychological tests, spatial
32 alignment of ambient NC>2 concentrations to subjects' school or home locations, and
January 2015 6-140 DRAFT: Do Not Cite or Quote
-------
1 examination of potential confounding by multiple SES indicators. While some studies
2 considered birth outcomes and traffic noise exposure as potential confounding factors,
3 smoking and stress were inconsistently or not considered. Further, in most studies, NC>2
4 was the only air pollutant examined, and uncertainty remains regarding potential
5 confounding by copollutants that are well-characterized risk factors for decrements in
6 neurodevelopmental function such as lead, PM2 5, or PM components such as polycyclic
7 aromatic hydrocarbons.
Cognitive Function
8 NC>2 is not consistently associated with cognition in children. Among children age
9 4 years, indoor home NCh at age 3 months was associated with multiple measures of
10 cognitive function, from a general cognition index to memory, verbal, and quantitative
11 skills (Morales et al.. 2009). These associations were limited to children with a
12 glutathione s-transferase P (GSTP1) valine (Val)-105 allele [isoleucine (Ile)/Val or
13 Val/Val vs. lie/lie genotype], which is associated with lower oxidative metabolism. In
14 contrast with indoor NC>2, ambient NC>2 assessed concurrently with cognitive function or
15 for the prenatal period was not clearly associated with cognitive function in
16 schoolchildren or infants (Guxens et al., 2014; Clark et al., 2012; Guxens et al., 2012; van
17 Kempen etal.. 2012; Freire etal.. 2010; Wang et al.. 2009a). Within studies, results were
18 inconsistent among the multiple indices of cognitive function examined. Results also
19 were inconsistent across studies, including those for indices of memory, which was
20 examined in most studies. As was done in the 2013 ISA for Lead (U.S. EPA. 2013a),
21 evidence is evaluated separately for cognition in schoolchildren and infants.
22 A common strength of studies conducted in schoolchildren is the assessment of NC>2
23 exposures for home or school locations using well-validated LUR models (see
24 Section 3.2.1.1 for details on using LUR models to assign exposure), van Kempen et al.
25 (2012) was particularly noteworthy for assessing exposures outside both home and school
26 and examining potential confounding by traffic noise. The LUR model well predicted
27 ambient NC>2 concentrations in the study area (R2 for cross-validation = 0.85). School
28 NC>2, not home NC>2, was associated with memory with adjustment for noise
29 (Table 6-12). Neither school nor home NC>2 was associated with ability to process
30 information. A similar study found school-based aircraft noise but not NC>2 to be
31 associated with cognitive function (Clark et al.. 2012). NCh estimated for home locations
32 also was inconsistently associated with cognitive function. Another study of concurrent
33 NO2 exposure observed that higher home outdoor NCh was associated with poorer
34 cognitive function, but the wide 95% CIs call into question the reliability of findings
35 (Freire etal.. 2010). Associations for prenatal residential NC>2 exposure were similarly
36 inconsistent among cohorts in three cities in Spain, with negative, null, and positive
January 2015 6-141 DRAFT: Do Not Cite or Quote
-------
1 associations observed with cognition [(Guxens et al.. 2014); Table 6-121. In these studies,
2 NO2 exposures were estimated from LUR models that varied in performance across
3 locations; however, the inconsistency in findings does not appear to be related to model
4 performance. In Freire etal. (2010). the LUR model performed better in the nonurban
5 than urban locations (R2 for building = 0.75 and 0.45, respectively; R2 for
6 validation = 0.64 for locations combined), and 84% of subjects lived in nonurban areas.
7 Additionally, NC>2 was not associated with decrements in cognitive function in cities in
8 Spain where LUR models predicted ambient NO2 concentrations well [cross-validation
9 R2 = 0.75, 0.77; (Guxens etal.. 2012; Estarlich etal.. 2011)1. Wang et al. (2009a) also
10 produced inconsistent findings but has weaker implications because of its ecological
11 comparison of school locations that differed in ambient NO2 concentrations not direct
12 analysis of NC>2.
13 The Bayley Scales of Infant Development is a widely used and reliable test for infant
14 development. The mental development index is a measure of sensory acuity, memory,
15 and early language skills. However, the Bayley Scales of Infant Development scores are
16 not necessarily correlated with development of children at older ages, and the tests at age
17 1 year or younger do not assess many outcomes that are analogous to those assessed at 2
18 and 3 years old. Across studies, mental development in infants ages 6 to 24 months was
19 not associated with trimester-specific prenatal NO2 exposure or NO2 exposure from birth
20 to age 6 months (Lin et al.. 2014) and was inconsistently associated with NC>2 averaged
21 over pregnancy (Guxens et al.. 2014; Kim etal.. 2014; Guxens et al.. 2012). In an
22 analysis of cohorts across multiple European counties, associations were found in
23 locations where the LUR model better predicted ambient NO2 [cross-validation
24 R2 = 0.69-0.84 vs. 0.45, 0.51; (Beelen etal.. 2013; Estarlich etal.. 2011)1. Associations
25 for four cohorts in Spain differed between publications [(Guxens et al.. 2014; Guxens
26 etal.. 2012); Table 6-12]. Other than sample sizes and possibly different pregnancy
27 addresses used to estimate NO2 exposure, an explanation for divergent results is not clear
28 (Guxens et al.. 2014; Estarlich et al.. 2011). Mental development of infants also was
29 inconsistently associated with NO2 exposure assessed by averaging concentrations across
30 city central site monitors (Lin etal.. 2014) or combining concentrations by IDW (Kim
31 etal.. 2014). The uncertainty of these methods in capturing the spatial heterogeneity in
32 ambient NO2 concentrations is illustrated by the weak or moderate correlation (r = 0.42,
33 0.21) observed between ambient NO2 estimated by IDW and measured outside homes in
34 a subset of subjects in Kim etal. (2014). Another uncertainty in these studies is the lack
35 of examination of potential confounding by benzene (Guxens et al.. 2012) or PMio (Kim
36 etal.. 2014). which showed a similar pattern of association as did NO2 and were highly or
37 moderately correlated with NO2 (r = 0.70 and 0.40, respectively).
January 2015 6-142 DRAFT: Do Not Cite or Quote
-------
Attention-Related Behaviors
1 The few studies of attention-related behaviors produced contrasting results for
2 associations with NCh. Morales et al. (2009) observed that exposure to gas appliances
3 and higher indoor NC>2 at age 3 months were associated with elevated odds ratios for
4 symptoms of Attention Deficit Hyperactivity Disorder at age 4 years. The association
5 was attributable mainly to inattention, as hyperactivity was not associated with NC>2. As
6 was observed for cognitive function, associations were limited to children with a GSTP1
7 Val-105 allele. In contrast, outdoor school and home NO2 concentrations (with or without
8 adjustment for road traffic and aircraft noise) were not associated with poorer
9 performance on multiple tests of sustained and switching attention (van Kempen et al..
10 2012). There was some evidence of an MVroad traffic noise interaction, as home
11 outdoor NC>2 was associated with poorer attention switching among children in the
12 highest noise category. Home NC>2 and road traffic noise were moderately correlated
13 (r = 0.30). The ecological study did not find attention performance test results to differ
14 consistently with respect to school locations (Wang et al., 2009a).
Motor Function
15 Evidence does not strongly indicate that NC>2 exposure affects motor function of children.
16 Whereas higher indoor home NCh exposure at age 3 months was associated with poorer
17 motor function in 4-year olds (Morales et al., 2009). findings were inconsistent for
18 ambient NC>2 exposure. A combined analysis of multiple European cohorts found that
19 NC>2 exposure ascertained for the birth address by LUR was associated with poorer motor
20 function overall, but associations were limited to half of the individual cohorts (Guxens
21 etal.. 2014). PM2 5, PM2 5 absorbance (an indicator of EC), PMio, and coarse PM were
22 not associated with motor function in all of the same cohorts as NC>2. Thus, confounding
23 by these copollutants does not seem to fully explain the NC>2 associations. There was no
24 clear pattern of association by gross or fine motor function or by age at which motor
25 function was assessed. NCh was associated with poorer motor function among children
26 between ages 1-6 years in some locations but not others. The inconsistency in findings
27 does not appear to depend on the adequacy of the LUR models to represent ambient NCh
28 concentrations in the study area. LUR models have shown a similar range in performance
29 [cross-validation R2 = 0.49 to 0.87; (Beelen etal.. 2013; Estarlich etal.. 2011)1 in
30 locations where associations were observed and not observed. A limitation of the LUR
31 models is that they were constructed based on ambient concentrations measured after the
32 birth of some subjects. When the analysis was restricted to subjects whose birth dates
33 coincided with the period of ambient monitoring, the effect estimate decreased; however,
34 there was evidence of association between NO2 and motor function for the combined
35 cohorts (Table 6-12).
January 2015 6-143 DRAFT: Do Not Cite or Quote
-------
1 Results also are inconsistent for concurrent exposure. Among children ages 9-11 years,
2 neither concurrent NC>2 nor noise exposure, alone or combined, at school or home, was
3 associated with fine motor function (van Kempen et al.. 2012). Among children age
4 4 years, higher concurrent outdoor home NC>2 exposure was associated with poorer gross
5 motor function but not fine motor skills (Freire et al., 2010). Children attending schools
6 with higher ambient NC>2 had poorer motor function compared to children attending
7 schools with lower NC>2 (Wang et al., 2009a); however, attributing the findings to NC>2
8 versus another factor that differed between schools is not possible. Like home- or
9 school-based exposure measures, NO2 exposure assessed from central sites was not
10 clearly associated with motor function in infants. Prenatal and lifetime exposure was
11 inconsistently associated with motor function at 6 months of age, and no associations
12 were observed in infants ages 12 to 24 months (Kim etal.. 2014; Lin et al.. 2014).
13 Limited evidence from toxicological studies also shows mixed effects of NO2 exposure
14 on motor function. Tabacova et al. (1985) found deficits in motor function and postural
15 gait in rat pups exposed gestationally to 50, 500, or 5,300 ppb NC>2 (dam exposure,
16 6 h/day, 7 days/week, Gestation Day 0-20). In the open field test, female and male
17 animals exhibited retarded locomotor development, with stronger effects earlier in life
18 (testing done until 3 months of age). On PND9, reductions were noted in horizontal
19 motility and head raising with prolonged periods of immobility, hypotonia, tremor, and
20 equilibrium deficits. Gait deficits including hindlimb dragging, crawling en lieu of
21 walking, pivoting, and impaired body raising ability were observed out to PND14 even in
22 animals in the lowest dose group. Tabacova et al. (1985) also found deficits in righting
23 reflex and the auditory startle reflex. In a separate study, prenatal exposure to 1,500 or
24 3,000 ppb NO2 [(Di Giovanni et al.. 1994) dam exposure over GD 0-20 of pregnancy]
25 did not significantly affect motor function in 10- to 15-day-old male pups as measured by
26 infrared sensors.
Psychological Distress
27 The two available studies produced equivocal evidence for the effects of NO2 exposure
28 on psychological distress. Among children ages 9-10 years, an index of emotional,
29 social, and conduct problems was not associated with concurrent NO2 or with aircraft or
30 road traffic noise at school, either alone or after mutual adjustment (Clark etal.. 2012). In
31 rats, Di Giovanni et al. (1994) reported that 3,000 ppb continuous NO2 exposure of dams
32 during GDO-GD21 resulted in decreased pup vocalization, an indicator of emotionality,
33 in males removed from the nest at PND5, PND10, or PND15.
January 2015 6-144 DRAFT: Do Not Cite or Quote
-------
Autism
1 Autism is a neurodevelopmental disorder characterized by impaired social interaction,
2 verbal and nonverbal communication deficits, and repetitive or stereotypic behavior.
3 Although the causes of autism are not fully understood, genetic conditions, family
4 history, and older parental age have been implicated as risk factors. Case-control studies
5 in California observed that higher NC>2 concentrations during the prenatal period and
6 during the first year of life were associated with higher odds ratios for autism in children
7 ages 24-71 months (Becerra et al.. 2013; Volk et al.. 2013). In both studies, cases were
8 identified from regional referral centers contracted by the Department of Developmental
9 Services. Controls were selected as birth certificate records not having a matching record
10 of autism with the referral centers. Controls were matched to cases by age, sex, and wide
11 geographic area. However, matching by area of residence was based only on birth
12 addresses. These studies also observed stronger associations for autism among children
13 with mothers with less than a high school education compared with higher education
14 (Becerra et al.. 2013) and children with the CC MET genotype compared with CC/GG
15 genotype (Volk et al.. 2014). The CC MET genotype is associated with decreased MET
16 protein in the brain and has been associated with autism risk.
17 Between studies, inference about NO2 is stronger in (Becerra et al., 2013). Residential
18 NO2 exposure was assessed using a well-validated LUR model [cross-validation
19 R2 = 0.87; (Su et al., 2009)1. In contrast, Volk etal. (2013) examined central site ambient
20 NC>2 concentrations and NOx estimated with a dispersion model. There are large
21 uncertainties with these exposure measures. Central site NO2 concentrations were
22 assigned from a site 5 km from homes, if available, or by inverse distance weighting over
23 a 50-km area. The authors did not report what proportion of subjects were assigned NC>2
24 exposures at the 5-km or 50-km scale, but neither scale may adequately capture the
25 spatial heterogeneity in NCh concentrations (Section 3.4.5.2). Inference is poorer for NOx
26 as it was nearly perfectly correlated (r -0.99) with EC and CO. In each study, PM2 5 also
27 was associated with autism, and (Becerra et al.. 2013) found that NO2 associations were
28 robust to adjustment for traffic-related copollutants PM2 5 or CO as well as the
29 copollutants Os or PMio. However, the reliability of the copollutant model results is
30 uncertain as copollutant concentrations were assessed from central sites, and exposure
31 measurement error likely varies between central site copollutant concentrations and
32 residential NO2.
Neuronal Degeneration and Nervous System Oxidative Stress
33 A recent study found that short-term NO2 exposure induced neuronal degeneration and
34 oxidative stress in the brains of adult male Wistar rats. Seven-day (6 h/day) exposure to
January 2015 6-145 DRAFT: Do Not Cite or Quote
-------
1 2,500-10,000 ppb NO2 (Li et al., 2012) had no effect on body weight; however,
2 concentration-dependent reductions in brain-to-body-weight ratios were observed, with
3 statistically significant differences at 5,320 and 10,000 ppb NC>2. Histopathological
4 analysis of cerebral cortex demonstrated a concentration-dependent increase in swollen or
5 shrunken nuclei and a concentration-dependent, statistically significant increase in
6 apoptotic cell number in all NC>2-exposed rats. Statistically significant changes in
7 antioxidant enzyme activities [Cu/zinc (Zn) SOD, MnSOD, and glutathione peroxidase],
8 protein carbonyls, and malondialdehyde were observed in response to 5,320 and
9 10,000 ppb NO2. While rats exposed to 2,500 ppb NO2 demonstrated a statistically
10 significant increase in the protein level of p53, rats exposed to the higher concentrations
11 exhibited statistically significant increases in mRNA and protein levels of c-fos, c-jun,
12 p-53, and bax. These results are consistent with oxidative stress especially at higher
13 concentrations of NO2.
6.4.4.2 Physical Development
14 Limited information from toxicological studies does not clearly indicate that NC>2
15 exposure affects physical post-natal development. Distinct exposure periods and test
16 endpoints produced disparate results in two studies on post-natal body-weight gain in
17 pups whose dams were exposed to NC>2. Shalamberidze and Tsereteli (1971b) and
18 Shalamberidze and Tsereteli (197la) showed decrements in post-natal body weight at
19 PND4 and 12 in albino rats with prenatal exposures to 67 or 1,300 ppb NC>2 during the
20 3 months prior to breeding for 12 hours a day. Continuous NC>2 exposure during gestation
21 [continuous exposure of dam to 1,500 or 3,000 ppb NC>2, GDO-GD20; (Di Giovanni
22 et al., 1994)1 produced no significant differences in weight gain at PND1, 11, or 21 in
23 male Wistar rat pups. Tabacova et al. (1985) saw concentration-dependent delays in eye
24 opening and incisor eruption in rodents after maternal exposure to NCh during pregnancy
25 (dam exposure: GDO-GD21, 5 h/day, 25, 50, 500, or 5,300 ppb NO2) (Table 6-13).
6.4.4.3 Summary of Postnatal Development
26 The collective evidence does not consistently indicate a relationship between NO2
27 exposure and effects on post-natal development. Very few outcomes were similar
28 between epidemiologic and toxicological studies. Physical development, examined as
29 post-natal weight gain, eye opening, and incisor eruption in only a few studies of rats,
30 was not clearly affected by prenatal NC>2 exposures in the range of 53 to 5,320 ppb. As
31 examined primarily in epidemiologic studies, prenatal, early life, or concurrent
32 school-age NC>2 exposure was not consistently associated with cognitive function,
January 2015 6-146 DRAFT: Do Not Cite or Quote
-------
1
2
o
6
4
5
6
7
attention-related behaviors, motor function, or psychological distress in infants or
schoolchildren. While epidemiologic associations were observed for indoor home NCh
(Morales et al., 2009). evidence is equivocal for ambient NO2, including exposure metrics
spatially aligned with subjects' home and school locations using LUR models that well
represented the spatial heterogeneity in the study areas (Guxens et al., 2014; van Kempen
et al.. 2012; Freire et al.. 2010). In limited examination of children in California, autism
was associated with residential prenatal NO2 exposure (Becerra et al., 2013).
Table 6-13 Reproductive and developmental lexicological studies for nitrogen
dioxide (NO2).
Reference
Tabacova et al.
(1985)
Shalamberidze and
Tsereteli(1971a).
Shalamberidze and
Tsereteli(1971b)
Di Giovanni et al.
(1994)
Kripke and Sherwin
(1984)
Species
(Strain);
Age;
Concentration Sex;
NO2 (ppb) n
25, 50, 500, or Rats
5,300 ppb (Wistar);
adult;
F; n = 20
1,300 ppb Rats
(albino);
adult;
F; n = 7
1,500 or Rats
3,000 ppb (Wistar);
pups; M;
n = 7
1,000 ppb Rats
(LEW/f
mai);
young
adults; M;
n = 6
Exposure Conditions
Developmental exposure with
post-natal neurotoxicity testing.
0,25, 50,500, or 5,300 ppb,
5 h/day during Gestational Days
0 through 21 ; progeny followed
up to PND60.
1,300 ppb for 12 h/day for 3 mo
(further specifics unavailable).
Direct exposure of dams at 0,
1 ,500, or 3,000 ppb continuously
throughout Gestational Days
0-20, male offspring tested for
vocalizations on PND5, PND10,
and PND15.
0 or 1,000 ppb for 7 h/day,
5 days/week for 21 days.
Endpoints Examined
Pup viability, developmental
endpoints (eye opening,
incisor eruption);
neuromotor (righting reflex,
postural gait, geotaxis);
hepatic lipid peroxidation;
hepatic drug-metabolizing
enzyme activity.
Litter size, birth weight,
post-natal weight gain (body
weight).
Neurobehavior (ultrasonic
vocalization), maternal body
weight during pregnancy,
litter size, post-natal body
weight, early life mortality,
motor function.
Spermatogenesis, germinal
cells histology, and
testicular interstitial cell
histology.
NR = no quantitative results reported; PND = post-natal day.
8 In addition to the inconsistent or limited evidence for NCh-related neurodevelopmental
9 effects, there is uncertainty regarding confounding by factors spatially correlated with
10 NO2 at the level of individuals or communities. Studies observed associations with NO2
11 with adjustment for SES indicators and birth outcomes. However, examination of
January 2015
6-147
DRAFT: Do Not Cite or Quote
-------
1 confounding was absent for stress and was very limited for smoking, noise, and
2 traffic-related copollutants. In van Kempen et al. (2012). school NO2 was associated with
3 memory after adjustment for traffic noise, and there was evidence of a home NO2-noise
4 interaction for attention switching. The few studies that examined pollutants other than
5 NO2 found associations of neurodevelopmental effects with traffic-related copollutants
6 such as PM2 5, benzene, and CO. An association between NO2 and autism was observed
7 with adjustment for CO or PM2 5 as well as for Os or PMio (Becerra et al., 2013).
8 However, the reliability of results is uncertain because of potential differential exposure
9 measurement error between residential NO2 and central site copollutants. Other pollutants
10 characterized to be associated with neurodevelopment such as lead or polycyclic aromatic
11 hydrocarbons (U.S. EPA. 2013a. 2009) were not examined as potential confounding
12 copollutants. Toxicological evidence is far more limited than epidemiologic evidence and
13 is similarly uncertain. Prenatal NO2 exposure induced psychological distress (Di
14 Giovanni et al.. 1994) in rat offspring but showed mixed effects on motor function (Di
15 Giovanni et al.. 1994; Tabacova et al., 1985). Further, although some studies showed
16 effects on post-natal development, there was very limited information to characterize
17 possible modes of action for ambient-relevant NO2 exposures. In a recent study of adult
18 rats, short-term exposure to 5,320 ppb NO2 induced increases in the neuronal apoptotic
19 and oxidative stress markers (Li et al.. 2012). which have been linked to cognitive
20 function.
6.4.5 Summary and Causal Determination
21 Overall, the evidence is suggestive, but not sufficient, to infer a causal relationship
22 between exposure to NO2 and birth outcomes and is inadequate to infer the presence or
23 absence of a causal relationship between exposure to NO2 and fertility, reproduction and
24 pregnancy, and post-natal development. Separate conclusions are made for these three
25 smaller groups of outcomes because they are likely to have different etiologies and
26 critical exposure patterns over different lifestages. In past reviews, a limited number of
27 epidemiologic and toxicological studies had assessed the relationship between exposure
28 to NO2 and reproductive and developmental effects. The 2008 ISA for Oxides of
29 Nitrogen concluded that there was not consistent evidence for an association between
30 NO2 and birth outcomes and concluded that evidence was inadequate to infer the
31 presence or absence of a causal relationship with reproductive and developmental effects
32 overall. The change in the causal determination for birth outcomes reflects the larger
33 number of studies that have evaluated these outcomes, as well as the improved exposure
34 assessment (e.g., LUR models) and outcome assessment (e.g., fetal growth measured
35 throughout pregnancy via ultrasound) employed by these studies. Exposure assessment
January 2015 6-148 DRAFT: Do Not Cite or Quote
-------
1 was evaluated drawing upon discussions in Sections 3.2 and 3.4.5. In general, LUR
2 model predictions have been found to correlate well with outdoor NO2 concentration
3 measurements (Section 3.2.1.1). A select number of recent studies have employed
4 exposure assessment methods such as LUR to represent the spatial variability of NC>2. All
5 available evidence, including more than 100 recent epidemiologic studies, examining the
6 relationship between exposure to NO2 and reproductive and developmental effects were
7 evaluated using the framework described in Table II of the Preamble. The key evidence
8 as it relates to the causal determination is summarized in Table 6-14.
Fertility, Reproduction, and Pregnancy
9 A number of studies examined the association between exposure to NO2 concentrations
10 measured at central site monitors or estimates of NO2 or NOx concentration from LUR
11 models and effects on fertility, reproduction, and pregnancy. These types of health
12 endpoints and their relationship with air pollution have only recently begun to be
13 evaluated, and, thus, the number of studies for any one endpoint is limited. There is
14 generally no evidence for an association between NO2 concentrations and sperm quality
15 in either epidemiologic or toxicology studies. One study (Les.ro et al.. 2010) observed a
16 decreased odds of live birth associated with higher NO2 concentrations during ovulation
17 induction and the period after embryo transfer; while another (SlamaetaL 2013)
18 observed decreased fecundability with higher NO2 exposure near conception. There is
19 inconsistent evidence for an association between NO2 concentrations estimated from
20 LUR models and pre-eclampsia, pregnancy-induced hypertension, gestational diabetes,
21 and reduced placental growth and function. Maternal weight gain showed mixed effects
22 during pregnancy with one of two studies reporting decrements in weight gain with NCh
23 exposure; the same trend was seen with NCh-dependent smaller litter size. Impaired
24 estrus cyclicity in NCh-exposed animals was reported as was a decrease in number of
25 primordial follicles in NCh-exposed rodents. Collectively, the limited evidence is of
26 insufficient consistency and is inadequate to infer a causal relationship between exposure
27 to NC>2 and effects on fertility, reproduction, and pregnancy.
Birth Outcomes
28 While the collective evidence for many of the birth outcomes examined is not entirely
29 consistent, there are several well-designed, well-conducted studies that indicate an
30 association between NO2 and adverse birth outcomes. For example, the Spanish cohort
31 that utilized anthropometric fetal measurements throughout pregnancy (Iniguez et al.,
32 2012: Estarlichetal..2011: Aguileraetal.. 2010: Ballester et al.. 2010) observed small,
33 yet consistent associations with impaired fetal growth and NCh concentrations.
January 2015 6-149 DRAFT: Do Not Cite or Quote
-------
1 NC>2-dependent decrements in pup birth weight were reported in an animal toxicology
2 study (Shalamberidze and Tsereteli. 1971a) and provide supporting evidence for the
3 associations with fetal growth restriction observed in epidemiologic studies. Studies that
4 examined PTB, birth weight, birth defects, and infant mortality generally found
5 inconsistent results, with some studies observing positive associations, while others
6 observed negative associations, regardless of whether NO2 or NOx were used to estimate
7 exposure. Many of the studies examining PTB observed associations very close to the
8 null value. Generally, studies of birth outcomes did not evaluate other traffic-related
9 pollutants along with NO2 or NOx in copollutants models, and this is a source of
10 uncertainty that can not be addressed. Several different methods for exposure assessment
11 were used in these studies of birth outcomes, but there were no trends observed across
12 studies based on the method of exposure assessment. Collectively, the limited and
13 inconsistent evidence is suggestive, but not sufficient, to infer a causal relationship
14 between exposure to NO2 and effects on birth outcomes, with the strongest evidence
15 coming from studies of fetal growth restriction.
Postnatal Development Effects
16 There is inconsistent evidence from both epidemiologic and animal toxicological studies
17 for a relationship between prenatal and childhood NO2 exposure and post-natal
18 development effects. Findings across the several recent epidemiologic studies of
19 neurodevelopment do not consistently support associations of NO2 with cognitive
20 function, attention-related behaviors, motor function, and psychological distress in
21 children. Many of these studies estimate ambient NO2 exposures for children's homes or
22 schools using LUR models that well represent the variability in ambient concentrations in
23 study areas. Toxicological evidence for effects on neurodevelopment is limited, and
24 evidence for impaired physical development shows mixed results including null findings.
25 NO2 exposures were related to autism in children in recent epidemiologic studies, but
26 such findings are limited to a few studies. In the small group of epidemiologic studies
27 observing associations with neurodevelopmental effects, potential confounding by
28 traffic-related copollutants was not adequately examined. Motor function testing showed
29 multiple endpoints affected by NO2 exposure in animal models including reflexes
30 (auditory startle and righting reflex), postural gait, impaired walking, and head raising.
31 But one study found no motor function impairment with NO2 exposure. Communication
32 of psychological distress was impaired in NO2-exposed pups. Collectively, the evidence
33 is of insufficient consistency or quantity and is inadequate to infer a causal relationship
34 between exposure to NO2 and effects on post-natal development.
January 2015 6-150 DRAFT: Do Not Cite or Quote
-------
Table 6-14 Summary of evidence supporting the causal determinations for
relationships between long-term nitrogen dioxide (NO2) exposure
and reproductive and developmental effects.
Rationale for Causal
Determination3
Key Evidence13
Key References'3
NO2 Concentrations
Associated with Effects0
Fertility, Reproduction, and Pregnancy—Inadequate to Infer a Causal Relationship
Available Inconsistent associations
epidemiologic studies when exposure is assessed
of pre-eclampsia are of across entire pregnancy and
insufficient consistency individual trimesters,
between NO2 concentration
and pre-eclampsia, after
adjustment for common
potential confounders.
Uncertainty regarding
potential confounding by
traffic-related copollutants.
Wu et al. (2009),
Pereiraetal. (2013).
Malmqvist et al. (2013),
Dadvandetal. (2013)
Mean NCb: 7.2 ppb
Mean NOx: 23 ppb
Mean NOx: 7.5 ppb
Mean NCb: 30 ppb
Available
epidemiologic and
toxicological studies
for other
pregnancy-related
health effects are of
insufficient consistency
Limited and inconsistent
epidemiologic evidence for
associations with
pregnancy-induced
hypertension, gestational
diabetes, and placental
growth and function.
Hampeletal. (2011),
Leeetal. (2012a),
Mobasheret al. (2013),
Malmqvist et al. (2013),
Xuetal. (2014).
van den Hooven et al.
Means: 8.7-28.6 ppb
Limited and inconsistent
evidence in rats for deficits in
maternal weight gain during
pregnancy.
Tabacova etal. (1985),
Pi Giovanni et al. (1994)
1,300, 1500, and 3,000 ppb
Limited evidence for
key events in mode of
action
Impaired estrus cyclicity and
decreased number of
primordial follicles in rodents
exposed to NO2.
Shalamberidze and
Tsereteli(1971a),
(Shalamberidze and
Tsereteli, 1971b)
67 or 1300 ppb
Available
epidemiologic studies
for in vitro fertilization
failure are of
insufficient consistency
Decreased odds of live birth
associated with higher NO2
concentrations during
ovulation induction and the
period after embryo transfer.
Leqroetal. (2010).
Slamaetal. (2013)
Mean: 19 ppb
Median: 19 ppb
Available toxicological
and epidemiologic
studies on sperm
quality are of
insufficient consistency
Overall, a limited number of
toxicological and
epidemiologic studies
provide no evidence for an
association between
exposure to ambient NO2
concentrations and effects
on sperm.
Rats: Kripke and Sherwin
(1984)
Humans: Rubes et al.
(2010),
Sokol et al. (2006)
1,000 ppb 7 h/day,
5 days/week
Mean ambient exposure
averaged over 90-days:
16.8 ppb
Mean ambient daily
concentration: 30.1 ppb
January 2015
6-151
DRAFT: Do Not Cite or Quote
-------
Table 6-14 (Continued): Summary of evidence supporting the causal
determinations for relationships between long-term
nitrogen dioxide (NO2) exposure and reproductive and
developmental effects.
Rationale for Causal
Determination3
Key Evidence13
Key References'3
NO2 Concentrations
Associated with Effects0
Birth Outcomes—Suggestive, But Not Sufficient, to Infer a Causal Relationship
Evidence from multiple Strongest evidence from
epidemiologic studies well-conducted Spanish
of fetal growth cohort studies that observe
restriction is generally associations with NO2
supportive but not concentrations and fetal
entirely consistent growth restriction.
Supported by consistent
evidence for SGA and IUGR.
Outcomes assessed with
anthropometric fetal
measurements.
Section 6.4.3.31
Mean exposure averaged over
trimesters: 7.8-36.1 ppb
Limited and
inconsistent
epidemiologic
evidence for other birth
outcomes
Some studies observe an
association between NO2
exposure and PTB, birth
weight, birth defects, and
infant mortality while other
studies observe no
consistent pattern of
association.
Section 6.4.3.2
Section 6.4.3.3
Section 6.4.3.4
Section 6.4.3.5
Mean exposure averaged over
trimesters: 8.8-37.6 ppb
Mean exposure averaged over
trimesters: 6.2-62.7 ppb
Mean exposure averaged over
early pregnancy
(e.g., Weeks 3-8):
8.2-28.0 ppb
Mean daily concentrations:
20.3-50.3 ppb
Limited and
inconsistent
toxicological evidence
with relevant NO2
exposures
Mixed evidence of effects on
litter size and mixed
evidence of late embryonic
lethality in rats.
Shalamberidze and
Tsereteli(1971a),
Shalamberidze and
Tsereteli(1971b)
Pi Giovanni et al. (1994)
Tabacova etal. (1985)
1,300-5,300 ppb
Limited evidence for key events in mode of action
Inflammation
Increase in CRP
concentration in human
umbilical cord blood
associated with NO2
concentration.
van den Hooven et al.
Mean exposure averaged over
week before delivery: 21.4 ppb
January 2015
6-152
DRAFT: Do Not Cite or Quote
-------
Table 6-14 (Continued): Summary of evidence supporting the causal
determinations for relationships between long-term
nitrogen dioxide (NO2) exposure and reproductive and
developmental effects.
Rationale for Causal
Determination3
Key Evidence13
Key References'3
NO2 Concentrations
Associated with Effects0
Postnatal Development—Inadequate to Infer a Causal Relationship
Limited and
inconsistent
epidemiologic and
toxicological evidence
for effects on
neurodevelopment
Some epidemiologic studies
showed cognitive function
decrements in infants and
schoolchildren in association
with NO2 exposure.
Uncertainty regarding
potential confounding by
traffic-related copollutants.
van Kempen et al. (2012).
Morales et al. (2009),
Guxens et al. (2012)
Mean concurrent:
16.5, 16.9 ppb
Mean prenatal:
15.7 ppb
Some studies did not
indicate associations with
cognitive function.
Evidence is inconsistent for
NO2 exposures estimated for
childrens' homes or schools.
Clark etal. (2012),
Freireetal. (2010),
Guxens etal. (2014)
More limited and inconsistent
epidemiologic evidence for
attention-related behaviors,
motor function, psychological
distress.
Section 6.4.4.1
In utero NO2 exposure
increased emotionality in rat
pups, but effects on motor
function inconsistent.
Tabacova etal. (1985),
Pi Giovanni etal. (1994)
50, 500 or 5,300 ppb
1,500 and 3,000 ppb
Prenatal NO2 exposure
associated with autism in first
yr of life or at ages 3-6 yr in
California.
Becerra etal. (2013)
Mean: 30.8 ppb
Limited evidence for
key events within
mode of action
Neurodegeneration
Oxidative stress
Short-term exposure
increased apoptotic factors
and oxidative stress in brain
of adult rats.
Li etal. (2012)
5,320 ppb for 7 days
January 2015
6-153
DRAFT: Do Not Cite or Quote
-------
Table 6-14 (Continued): Summary of evidence supporting the causal
determinations for relationships between long-term
nitrogen dioxide (NO2) exposure and reproductive and
developmental effects.
Rationale for Causal NO2 Concentrations
Determination3 Key Evidence13 Key References'3 Associated with Effects0
Limited and NO2 exposure delayed Tabacova et al. (1985), 530 and 5,300 ppb
inconsistent post-natal eye opening and p| Giovanni et al. (1994) 1,500 and 3,000 ppb
toxicological evidence incisor eruption but had
for physical mixed effects on post-natal
development growth.
CRP = C-reactive protein; IUGR = intrauterine growth restriction; NO2 = nitrogen dioxide; NOX = sum of NO and NO2;
PTB = preterm birth; SGA = small for gestational age.
aBased on aspects considered in judgments of causality and weight of evidence in causal framework in Tables I and N. of the
Preamble.
""Describes the key evidence and references contributing most heavily to causal determination and where applicable to
uncertainties and inconsistencies. References to earlier sections indicate where the full body of evidence is described.
°Describes the NO2 concentrations with which the evidence is substantiated (for experimental studies, below -5,000 ppb).
6.5 Total Mortality
1 In past reviews, a limited number of epidemiologic studies had assessed the relationship
2 between long-term exposure to NC>2 and mortality in adults, including cause-specific and
3 total mortality. The 2008 ISA for Oxides of Nitrogen concluded that the amount of
4 evidence was "inadequate to infer the presence or absence of a causal relationship"
5 (U.S. EPA. 2008). In the current ISA, findings for cause-specific mortality
6 (i.e., respiratory, cardiovascular) are used to assess the continuum of effects and inform
7 the causal determinations for respiratory and cardiovascular effects. The causal
8 determination for total mortality contained herein (Section 6.5) is based primarily on the
9 evidence for nonaccidental mortality but also is informed by the extent to which evidence
10 for the spectrum of cardiovascular and respiratory effects provides biological plausibility
11 for NO2-related total mortality. A supplemental Table S6-9 (U.S. EPA. 2013i) provides
12 an overview of the epidemiologic studies of long-term exposure to NCh or NOx and
13 mortality, including details on exposure assessment and mean concentrations from the
14 study locations.
6.5.1 Review of Mortality Evidence from 2008 Integrated Science Assessment for
Oxides of Nitrogen
15 Two seminal studies of long-term exposure to air pollution and mortality among adults
16 have been conducted in the United States; the American Cancer Society (ACS) and the
January 2015 6-154 DRAFT: Do Not Cite or Quote
-------
1 Harvard Six Cities (HSC) cohorts have undergone extensive independent reanalyses and
2 have reported extended results including additional years of follow-up. The initial reports
3 from the ACS (Popeetal.. 1995) and the HSC (Dockery et al.. 1993) studies did not
4 include results for gaseous pollutants. However, as reported in the 2008 ISA for Oxides
5 of Nitrogen (U.S. EPA. 2008). in reanalyses of these studies, Krewski et al. (2000)
6 examined the association between gaseous pollutants, including NO2, and mortality.
7 Krewski et al. (2000) observed a positive association between long-term exposure to NO2
8 and mortality in the HSC cohort, with effect estimates1 similar in magnitude to those
9 observed with PIVb 5. The effect estimates were positive for different causes of mortality
10 but were the strongest for cardiopulmonary and total mortality. In a re-analyses of the
11 ACS cohort data (Krewski et al., 2000). long-term exposure to NO2 estimated from
12 central site monitors was not associated with mortality. An extended study of the ACS
13 cohort (Pope et al.. 2002) doubled the follow-up time and tripled the number of deaths
14 compared to the original study but still observed no association between long-term
15 exposure to NCh and mortality.
16 A series of studies (Lipfert et al.. 2006a; Lipfert et al.. 2006b; Lipfert et al.. 2003. 2000)
17 characterized a national cohort of over 70,000 male U.S. military veterans who were
18 diagnosed as having hypertension in the mid-1970s and were followed through 2001. In
19 the earlier studies, the authors reported increased risk of mortality associated with
20 exposure to NCh; these excess risks were in the range of 5-9% (Lipfert et al.. 2003.
21 2000). In the later studies, the authors focused on traffic density in this cohort. Lipfert
22 et al. (2006b) and Lipfert et al. (2006a) reported that traffic density was a better predictor
23 of mortality than ambient air pollution variables, though they still observed a positive
24 association between mortality and NC>2 exposure. The results from the series of studies
25 characterizing the Veterans cohort are indicative of a traffic-related air pollution effect on
26 mortality, but the study population (lower SES, males with hypertension and a very high
27 smoking rate) was not representative of the general U.S. population.
28 In another cohort conducted in the U.S. [the California Seventh-day Adventist cohort
29 (AHSMOG)], Abbey etal. (1999) enrolled young adult, nonsmoking Seventh-day
30 Adventists throughout California. Generally, NC>2 was not associated with total,
31 cardiopulmonary, or respiratory mortality in either men or women. The authors observed
32 large risk estimates for lung cancer mortality for most of the air pollutants examined,
33 including NC>2, but the number of lung cancer deaths in this cohort was very small (12 for
1 Quantitative effect estimates from studies reviewed in the 2008 ISA for Oxides of Nitrogen (U.S. EPA. 2008)
can be found alongside effect estimates from more recent studies in Figures 6-8. 6-9. and 6-10 and the
corresponding Tables 6-15. 6-16. and 6-17. respectively).
January 2015 6-155 DRAFT: Do Not Cite or Quote
-------
1 females and 18 for males out of atotal of 5,652 subjects); therefore, it is difficult to
2 interpret these results.
3 Several studies conducted in European countries have examined the relationship between
4 long-term exposure to traffic-related pollutants (including NO2 and NOx) and mortality
5 among adults. Hoek et al. (2002) observed an association between NO2 and mortality in
6 the Netherlands Cohort Study on Diet and Cancer (NLCS), though the association with
7 living near a major road was stronger in magnitude. On the other hand, Gehring et al.
8 (2006) observed that NO2 was generally more strongly associated with mortality than an
9 indicator for living near a major road in a cohort of women from Germany. Results from
10 the Air Pollution and Chronic Respiratory Diseases survey conducted in France,
11 demonstrated increased risk between long-term exposure to NO2 and total,
12 cardiopulmonary, and lung cancer mortality (Filleul etal. 2005). Similarly, Nafstad et al.
13 (2004) observed an association between NOx and total mortality, as well as deaths due to
14 respiratory causes, lung cancer, and ischemic heart disease in a cohort of Norwegian men.
15 Nyberg et al. (2000) observed similar results for lung cancer mortality in a case-control
16 study of men in Stockholm, Sweden. Naess et al. (2007) investigated the
17 concentration-response relationships between NO2 and cause-specific mortality among a
18 cohort from Oslo, Norway, aged 51-90 years. Total mortality, as well as death due to
19 cardiovascular causes, lung cancer, and COPD were associated with NO2 for both men
20 and women in two different age groups, 51-70 and 71-90 years. Naess et al. (2007)
21 reported that the effects appeared to increase at NO2 levels higher than 21 ppb in the
22 younger age group (with little evidence of an association below 21 ppb), while a linear
23 effect was observed between 10 and 31 ppb in the older age group.
24 The results from these studies led to the conclusion that the evidence was inadequate to
25 infer the presence or absence of a causal relationship in the 2008 ISA for Oxides of
26 Nitrogen (U.S. EPA. 2008). The 2008 ISA for Oxides of Nitrogen (U.S. EPA. 2008)
27 noted that potential confounding by copollutants was an important uncertainty when
28 interpreting the evidence for the association between long-term exposure to NO2 and
29 mortality. Collinearity among criteria pollutants is another uncertainty that needs to be
30 considered; several studies reported moderate-to-high correlations between NO2 and PM
31 indices (i.e., >0.5). The 2008 ISA for Oxides of Nitrogen (U.S. EPA. 2008) also
32 acknowledged that NO2 could be a surrogate or marker for traffic-related pollution. These
33 uncertainties do not preclude the possibility of an independent effect of NO2, or of NO2
34 playing a role in interactions among traffic-related pollutants.
January 2015 6-156 DRAFT: Do Not Cite or Quote
-------
6.5.2 Recent Evidence for Mortality from Long-term Exposure to Oxides of
Nitrogen
1 Several recent studies provide extended analyses of existing cohort studies of adult
2 populations. In a reanalysis that extended the follow-up period for the ACS cohort to
3 18 years (1982-2000), Krewski et al. (2009) reported generally null associations between
4 long-term exposure to NC>2 estimated from central site monitors and total and
5 cause-specific mortality, similar to what was reported in the initial reanalysis of this
6 cohort (Krewski et al.. 2000). In an update to the ACS study including cohort members
7 residing in California, Jerrett etal. (2013) used LUR models to predict long-term
8 (i.e., 15 years) exposures to NC>2 at the home addresses of each of the cohort members.
9 The authors observed positive associations between predicted NO2 exposures and total,
10 CVD, IHD, stroke, and lung cancer mortality, but not for respiratory mortality. The
11 strongest associations were observed for deaths due to lung cancer and stroke. The
12 associations with CVD and IHD mortality were attenuated in copollutant models that
13 included PIVb 5 (also estimated from an LUR model), while the association with lung
14 cancer was generally unchanged in two-pollutant models. In an update to the Veterans
15 cohort study, Lipfert et al. (2009) looked at markers for specific emission sources,
16 including NOx as a marker of traffic, and their relationship with mortality, utilizing a
17 26-year follow-up period now available for this cohort. The authors observed an
18 association between long-term exposures to NOx estimated from a plume-in-grid model
19 and mortality, and noted that this association was stronger among men living in areas
20 with high traffic density compared to men living in areas with lower traffic density. The
21 authors also demonstrate that traffic-related air pollutants (including NOx) are belter
22 predictors of mortality than a measure of traffic density in this cohort. Updated results
23 have also been reported for the NLCS cohort [the same effect estimates are reported by
24 both Beelen et al. (2008b) and Brunekreef et al. (2009)1. Consistent with previous results
25 from this cohort, the authors observe an association with total mortality. In the updated
26 results, the authors observe the strongest effect between long-term exposure to NO2
27 estimated from central site monitors and respiratory mortality; this association is stronger
28 than any observed with the traffic variables and total or cause-specific mortality.
29 In updates to a cohort of women in Germany (Gehring et al., 2006). Schikowski et al.
30 (2007) observed a positive association between ambient NO2 concentrations measured at
31 central site monitors and cardiovascular mortality among older women, though this
32 association was not modified by lung function status (i.e., normal vs. impaired lung
33 function). Heinrich et al. (2013) includes five additional years of follow-up and twice as
34 many fatalities compared to the original analysis. In the updated analyses, the authors
35 observed positive associations between NO2 concentrations measured at central site
36 monitors and total and cardiopulmonary mortality. The effect estimates were highest for
January 2015 6-157 DRAFT: Do Not Cite or Quote
-------
1 women living within 50 m of a road with median daily traffic volume of 5,000 cars or
2 greater. The effect estimates for the associations between total and cardiopulmonary
3 mortality and NO2 were generally lower for the follow-up period compared to the
4 original analysis.
5 Several recent U.S. cohort studies examined the association between long-term exposures
6 to NO2 and mortality in occupational cohorts. Hart etal. (2010) examined the association
7 between residential exposure to NO2 estimated from a spatial smoothing model and
8 mortality among men in the U.S. trucking industry in the Trucking Industry Particle
9 Study (TrIPS). The authors observed an increase in cardiovascular disease mortality and
10 a decrease in COPD mortality associated with NO2 exposure. The association between
11 NO2 exposure and total mortality was robust to the inclusion of PMio or SO2 in
12 copollutant models. This association was stronger when the cohort was restricted to truck
13 drivers that maintained local routes, and long haul drivers were excluded. COPD
14 mortality was positively associated with NO2 exposure in the sensitivity analysis
15 excluding long haul drivers. The associations for other causes of death (i.e., lung cancer,
16 IHD, respiratory disease) were generally positive. Another recent U.S. cohort study, The
17 California Teachers Study (Lipsett et al.. 2011) examined the association between
18 long-term exposure to NOx and NO2 measured at central site monitors and mortality
19 among current and former female public school teachers. The authors observed the
20 strongest associations between IHD mortality and exposure to NOx and NO2; the
21 associations for other causes of death (i.e., CVD, cerebrovascular, respiratory, lung
22 cancer and total) were less consistent and generally close to the null value. Hart etal.
23 (2013) examined the association between long-term exposure to NO2 and total mortality
24 among a cohort of female nurses in the Nurses' Health Study. The authors used spatial
25 modeling to estimate exposure to NO2 and observed a small increase in the risk of total
26 mortality. In a sensitivity analysis examining women that moved during study follow-up,
27 the authors observed even higher risks among women that moved to areas with higher
28 concentrations of NO2.
29 A number of recent studies have examined the association between long-term exposure to
30 NO2 and mortality in Canadian cities. Chen etal. (2013) conducted a cohort study of air
31 pollution and cardiovascular mortality in three cities in Ontario. They used LUR models
32 to assign exposure to NO2 and observed that long-term exposure to NO2 was associated
33 with an increased risk of cardiovascular mortality. The association was stronger when
34 mortality from IHD was evaluated separately. In a single-city study conducted in
35 Toronto, Ontario, Jerrett et al. (2009) examined the association between long-term
36 exposure to NO2 (estimated from a LUR model) and total mortality among subjects from
37 a respiratory clinic. The authors observed positive associations with total and circulatory
38 mortality; the associations with respiratory and lung cancer mortality were also positive,
January 2015 6-158 DRAFT: Do Not Cite or Quote
-------
1 though less precise. In a model that included both NO2 and proximity to traffic, the effect
2 estimate for NO2 remained robust, and the effect attributable to traffic was attenuated. In
3 a single-city study conducted in Vancouver, British Columbia, Ganet al. (2013) and Gan
4 et al. (2011) conducted a population-based cohort study to evaluate the association
5 between traffic-related pollutants and risk of mortality due to CHD and COPD,
6 respectively. LUR models were used to estimate exposure over a 5-year period,
7 (1994-1998) and the cohort was followed for 4 years (1999-2002). The authors observed
8 the strongest associations (i.e., highest magnitude) for exposures to NO2 and CHD
9 mortality; however, these associations were greatly attenuated when PM2 5 and BC were
10 included in the model. The correlations between NO2 and PM2 5 and BC were low to
11 moderate (i.e., <0.5). The authors observed positive associations between both NO and
12 NO2 concentrations and COPD mortality, which were slightly attenuated when PM2 5 and
13 BC were included in the model.
14 A recent multicenter European study pooled data from 22 existing cohort studies and
15 used a strictly standardized protocol to investigate the associations between long-term
16 concentrations of NO2 and NOx and total (Beelen et al.. 2014a). respiratory,
17 (Dimakopoulou et al.. 2014) and cardiovascular (Beelen et al.. 2014b) mortality. The
18 authors used LUR models to assign exposure and observed generally null associations
19 with total, respiratory, and cardiovascular mortality. The total mortality associations were
20 evaluated in copollutants models and remained unchanged after adjustment for PM2 5 and
21 PMlo-2.5.
22 Several studies have examined the association between long-term exposure to NO2 and
23 mortality in England. Carey etal. (2013) conducted a cohort study using an
24 emissions-based model to assign exposure. Model validation was good, and model
25 estimates for NO2 were highly correlated with PMio and PM2 5 (r = 0.9). The authors
26 observed positive associations with total mortality; these associations were stronger for
27 respiratory and lung cancer deaths, and somewhat attenuated when restricted to
28 cardiovascular deaths. Tonne and Wilkinson (2013) evaluated the association between
29 long-term exposure to NO2 and NOx estimated from a Gaussian dispersion model among
30 survivors of hospital admissions for acute coronary system in England and Wales and
31 observed evidence of a null association after adjustment for PM2 5. In a single-city study,
32 Maheswaran et al. (2010) compiled a cohort of stroke survivors and modeled NO2
33 concentrations estimated from an emissions model across London. The authors observed
34 a nearly 30% increase in total mortality associated with exposure to NO2.
35 Rome, Italy was the setting for a number of single-city cohort studies. Cesaroni et al.
36 (2013) observed positive associations between long-term exposure to NO2 estimated from
37 an LUR model and total, cardiovascular, IHD, respiratory, and lung cancer mortality
January 2015 6-159 DRAFT: Do Not Cite or Quote
-------
1 among the adult population in the Rome Longitudinal Study (RoLS). These associations
2 were robust to the inclusion of PM2 5 in the model. Later, these authors used several
3 different LUR models to predict NO2 in Rome (Cesaroni et al.. 2012) and observed that
4 the modest, positive association between total mortality and NO2 concentrations was
5 consistent across all models evaluated. Rosenlund et al. (2008) conducted a cohort study
6 in Rome to investigate the effects of long-term exposure to NO2 and cardiovascular
7 deaths, including mortality among previous MI survivors. The authors observed a
8 positive association between long-term exposures to NO2 estimated from an LUR model
9 and fatal coronary events, though they did not observe an association with mortality
10 among survivors of a first coronary event.
11 A Danish study evaluated the association between long-term exposure to NO2 (estimated
12 from a dispersion model) and diabetes-related mortality Raaschou-Nielsen et al. (2013b).
13 The authors reported a 30% increase in risk of diabetes-related mortality associated with
14 NO2 concentrations. In Brisbane, Australia, Wang et al. (2009b) examined the association
15 between long-term exposure to NO2 estimated from central site monitors and
16 cardio-respiratory mortality. The relative risk for NO2 and cardio-respiratory mortality
17 was near the null value.
18 A number of studies were conducted in Asian countries to evaluate the association
19 between long-term-exposure to NO2 and mortality. In a national study covering
20 16 provinces in eastern China, Cao et al. (2011) observed positive associations between
21 ambient NOx concentrations from central site monitors and total, cardiovascular,
22 respiratory, and lung cancer mortality. The association between total mortality and NOx
23 was relatively unchanged in a copollutant model with total suspended particles (TSP) but
24 was reduced by half in a copollutants model with SO2. The associations between NOx
25 and cardiovascular, respiratory, and lung cancer mortality were all attenuated in
26 copollutants models including either TSP or SO2. In a single-city study in Shenyang,
27 China, the authors observed a strong, positive association between long-term exposure to
28 NO2 estimated from central site monitors and respiratory mortality (Dong etal. 2012)
29 and total, cardiovascular, and cerebrovascular mortality (Zhang et al.. 2011). In Shizuoka,
30 Japan, Yorifuji et al. (2010) observed positive associations between NO2 estimated from
31 an LUR model and total, cardiopulmonary, IHD, and respiratory disease mortality, with
32 the strongest effects observed for IHD mortality. When the analysis was restricted to
33 nonsmokers, a positive association was observed with lung cancer mortality. Similar
34 observations were reported for lung cancer by Katanoda et al. (2011) among a cohort in
35 Tokyo, Japan, and Liu et al. (2008) for a study of women living in Taiwan. In a related
36 study, Liu et al. (2009) also observed a positive association between long-term exposure
37 to NO2 and bladder cancer mortality.
January 2015 6-160 DRAFT: Do Not Cite or Quote
-------
These quantitative results of these studies are characterized in Figures 6-8. 6-9. and 6-10;
and Tables 6-15. 6-16. and 6-17.
MEAN
STUDY
Abbey et al. 1999
Abbey et al. 1999
Lipsettetal. 2011
Krewski et al. 2000
Cao etal. 2011*
Filleul et al. 2005
Zhang etal. 2011
Cesaroni et al. 2012
Cesaroni et al. 2013
Tonne etal. 2013*
Lipfert et al. 2000
COHORT CONCENTRATION
AHSMOG
AHSMOG
CA Teachers
ACS
China National Hypertension
PAARC
Shenyang Cohort
RoLS
Rome
Veterans
ivianeswaran et al. zuiut5ouui LOIIUOII ^LIOKC
Pope et al. 2002 ACS -Extended
Lipfert et al. 2006b
Brunekreefetal. 2009
Gehring et al. 2006
Heinrich et al. 2013
HoeK et al ^OOz
Lipfert et al. 2006a
Hart etal. 2010
Hart etal. 2013
Yorifuji et al. 2010
Jerrett et al. 2013
Carey etal. 2013
Nafstad et al. 2004
Veterans
NLCS-AIR
German Women's Health
German Women's Health
X,
1 oronto
Veterans
TrIPS
Nurses Health Study
Shizuoka Elderly
ACS-CA
National English
Norwegian Men
Tonne and Wilkinson 2013
Krewski et al. 2000
Lipfert et al. 2009
Beelenetal. 2014a
Beelenetal. 2014a*
Six Cities Study-Reanalysis
Veterans
ESCAPE
ESCAPE
36.7
36.7
33.6
27.9
50
24.5
24.4
23.3-24.2
23.4
22.6
21.5-27.8
21.4-27.9
19.8-27.2
20.7
20.7
20.7
1Q S
16.3
14.2
13.9
13.3
12.3
11.9
10.6
10; 6.5
6.1-21.9
5.58
2.8-31.7
8.7-107.3
i
NOTES |
Men -•-
Women -^-
Women -•-
.
jk
24 areas •
18 areas
otroKe survivors
1
Full Cohort
Case Cohort — •-
1-yr avg
5-yr avg
-
Full Cohort
Excluding Long Haul Drivers
Female Nurses
— •—
-•—
O
O
A
•
r1
o-
1 —
— • —
•
<•-
•*•
_9_
»
• —
-o-
-o
*-
-•-
A
I*
Pooled Analysis *
Pooled Analysis Jk
1 ! I 1 1 I I I 1 1
0.5 5
Hazard Ratios and 95% Cl
ACS = American Cancer Society; AHSMOG = California Seventh-day Adventists Cohort; ESCAPE = European Study of Cohorts for
Air Pollution Effects; NLCS = Netherlands Cohort Study on Diet and Cancer; NLCS-AIR = Netherlands Cohort Study on Air Pollution
and Mortality; PAARC = Air Pollution and Chronic Respiratory Diseases; RoLS = Rome Longitudinal Study; TrIPS = Trucking
Industry Particle Study.
Note: Red = recent studies; black = studies reviewed in the 2008 ISA for Oxides of Nitrogen. Hazard ratios are standardized to a
10-ppb increase in NO2 and a 20-ppb increase in NOX concentration.
aEffect estimates from studies measuring NOX in |jg/m3 have not been standardized. Studies are presented in descending order,
with the largest mean concentration (ppb) at the top and the smallest at the bottom of the figure. Circles = NO2; triangles = NOX.
Figure 6-8 Results of studies of long-term exposure to nitrogen dioxide
(NO2) or the sum of nitric oxide and NO2 (NOx) and total mortality.
January 2015
6-161
DRAFT: Do Not Cite or Quote
-------
Table 6-15 Correspon
Study
Abbey et al. (1999)
Abbey et al. (1999)
Lipsett et al. (2011)
Lipsettetal. (2011)
Krewski et al. (2000)
Caoetal. (201 1)b
Filleuletal. (2005)
Filleuletal. (2005)
Zhanqetal. (2011)
Cesaroni et al. (2012)
Cesaroni et al. (2013)
Tonne and Wilkinson (2013)b
Lipfert et al. (2000)
Maheswaran et al. (2010)
Popeetal. (2002)
Lipfert et al. (2006b)
Brunekreef et al. (2009)
Brunekreef et al. (2009)
Gehrinq et al. (2006)
Gehrinq et al. (2006)
Heinrich etal. (2013)
Hoek et al. (2002)
Jerrett et al. (2009)
Lipfert et al. (2006a)
Hart etal. (2010)
Hart etal. (2010)
ding risk estimates for Figure 6-8.
Location Notes
U.S. Men
U.S. Women
California Women, NO2
California Women, NOx
U.S.
China NOx
France 24 areas
France 18 areas
China
Italy
Italy
England and Wales NOx
U.S.
England Stroke survivors
U.S.
U.S.
the Netherlands Full cohort
the Netherlands Case cohort
Germany 1-yravg
Germany 5-yr avg
Germany
the Netherlands
Canada
U.S.
U.S. Full cohort
U.S. Excluding long haul drivers
Hazard Ratio3 (95% Cl)
1.02(0.95, 1.08)
0.99(0.94, 1.05)
0.97(0.94, 1.05)
1.02(0.98, 1.06)
0.99(0.99, 1.00)
1.02(1.00, 1.03)
0.98(0.96, 1.00)
1.22(1.10, 1.34)
5.39 (4.94, 5.94)
1.07(1.05, 1.11)
1.06(1.04, 1.06)
1.03(1.01, 1.05)
1.07(1.04, 1.10)
1.59(1.22,2.09)
1.00(0.98, 1.02)
1.03(0.98, 1.02)
1.05(1.00, 1.10)
0.92(0.79, 1.06)
1.20(1.02, 1.41)
1.23(1.02, 1.47)
1.21 (1.08, 1.36)
1.15(0.89, 1.49)
1.48(1.00,2.16)
1.04(0.97, 1.13)
1.10(1.06, 1.15)
1.19(1.13, 1.26)
January 2015
6-162
DRAFT: Do Not Cite or Quote
-------
Table 6-15 (Continued): Corresponding risk estimates for Figure 6-8.
Study
Hartetal. (2013)
Yorifuji et al. (2010)
Jerrettetal. (2013)
Carey et al. (2013)
Nafstad et al. (2004)
Tonne and Wilkinson (2013)
Krewski et al. (2000)
Lipfert et al. (2009)
Beelen etal. (2014a)
Beelen etal. (2014a)c
Location Notes
U.S. Female nurses
Japan
U.S.
England
Norway
England and Wales
U.S. NOx
U.S.
Europe NCb-pooled analysis
Europe NOx-pooled analysis
Hazard Ratio3 (95% Cl)
1.03(1.00, 1.06)
1.04(0.93, 1.32)
1.08(1.02, 1.14)
1.11 (1.05, 1.15)
1.16(1.12, 1.22)
1.12(1.06, 1.20)
1.15(1.04, 1.27)
1.04(1.03, 1.05)
1.02(0.98, 1.06)
1.02(1.00, 1.04)
Cl = confidence interval; NO2 = nitrogen dioxide; NOx = sum of NO and NO2
Note: Studies correspond to those presented in Figure 6-8.
aEffect estimates are standardized to a 10-ppb increase in NO2 and a 20-ppb increase in NOX concentration.
bNOx measured in |jg/m3. Effect estimate is per 10 |jg/m3 increase.
CNOX measured in |jg/m3. Effect estimate is per 20 |jg/m3 increase.
January 2015
6-163
DRAFT: Do Not Cite or Quote
-------
MEAN
STUDY
Naess et al. 2007
Lipsettetal. 2011
Cao etal. 2011*
Schikowski et al. 2007
Zhang etal. 2011
Cesaroni et al. 2013
Brunekreefetal. 2009
Hart etal. 2010
Jerrettetal. 2013
Chen etal. 2013
Carey etal. 2013
Beelenetal. 2014b
Beelenetal. 2014b+
Abbey etal. 1999
Krewski et al. 2000
Filleul et al. 2005
Gehring et al. 2006
Heinrich et al. 2013
\Vang et al. 2009b
HnclXct al 9nn')
oeK et ai. zuuz
Yorifuji et al. 2010
Krewsfci et al. 2000
Rosenlund et al. 2008
Ganetal. 2011
Carey etal. 2013
Lipsettetal. 2011
Krewski et al. 2000
Cesaroni et al. 2013
Hart etal. 2010
Yorifuji et al. 2010
Jerrettetal. 2013
Nafstad et al. 2004*
Chen etal. 2013
Carey etal. 2013
Beelenetal. 2014b
Beelenetal. 2014b+
Beelenetal. 2014b
Beelenetal. 2014b+
Jerrett et al. 2009
Yorifuji etal. 2010
Lipsettetal. 2011
Zhang etal. 2011
Cesaroni et al. 2013
Yonfuii et al. 2010
Nafstad et al. 2004*
Chen etal. 2013
Beelenetal. 2014b
Beelenetal. 2014b+
Jerrettetal. 2013
Carey etal. 2013
Carey etal. 2013
Raaschou-Nielsen 201
COHORT CONCENTRATION
Oslo
CA Teachers
National Hypertension
SALT A cohort
Shenyang Cohort
Rome
NLCS-AIR
TrIPS
ACS-CA
Ontario Cohort
National English
ESCAPE
ESCAPE
AHSMOG Cohort
ACS Cohort
PAARC Survey Cohort
German Women's Health
German Women's Health
Brisbane Cohort
Shizuoka Elderly Cohort
Six Cities Reanalysis
Rome Cohort
Vancouver Cohort
National English
CA Teachers
ACS Cohort
Rome
TrIPS
Shizuoka Elderly Cohort
ACS-CA
Norwegian Men Cohort
Ontario Cohort
National English
ESCAPE
ESCAPE
ESCAPE
ESCAPE
Toronto Cohort
Shizuoka Elderly Cohort
CA Teachers
Shenyang Cohort
Rome
Shizuoka Elderly Cohort
Norwegian Men Cohort
Ontario Cohort
ESCAPE
ESCAPE
ACS-CA
National English
National English
BDiet, Cancer and Health
42.1
33.0
26.5
24^4
23.4
20.7
14.2
12.3
12.1-21.7
11.9
2.8-31.7
8.7-107.3
36.7
27.9
24.5
20.7
20.7
1 Q Q
iy.o
133
6.1-21.9
-23.9
17.01
11.9
33.6
27.9
23.4
14.2
ITT
iJ.J
12.3
11.5-21.7
12.1-21.7
11.9
2.8-31.7
8.7-107.3
2.8-31.7
8.7-107.3
1 Q S
iy. j
13.3
33.6
24.4
23.4
ITT
iJ.J
11.5-21.7
12.1-21.7
2.8-31.7
8.7-107.3
12.3
11.9
11.9
9
NOTES i
Women
\Vomen
Full Cohort
Case Cohort
Full Cohort
Excluding Long
Men
Women
24 areas
1 8 areas
1-yr avg
*-9- Cardiovascular
q.
-10 —
f— |
' 0
Haul Drivers | — • —
— 0—
| -0-
-• —
jj*~
t
1 Cardiopulmonary
— 0 —
0
•
f .
A
Out of hospital
In hospital
Following non-f
Women
Full Cohort
— 0
— •—
— = — 0
tal event — •— *-
— 0^-
!• -
p
_ 0-
— 0 —
Excluding Long Haul Drivers • * —
i — • —
MI
A11IHD
A11IHD
MI
MI
[•*,
I
—A—
0C
CHD
IHD
-i-
Circulatory
•
1 Terbrovascular
Women • r-
m
V
(A
^ •
— •-!—
»
-A-
1 0
-f-
•
— (-•
-•
Stroke
Heart Failure
Diabetes
0.5
Hazard Ratio and 95% Cl
ACS = American Cancer Society; AHSMOG = California Seventh-day Adventists Cohort; ESCAPE = European Study of Cohorts for
Air Pollution Effects; NLCS = Netherlands Cohort Study on Diet and Cancer; NLCS-AIR = Netherlands Cohort Study on Air Pollution
and Mortality; PAARC = Air Pollution and Chronic Respiratory Diseases; RoLS = Rome Longitudinal Study; TrIPS = Trucking
Industry Particle Study.
Note: Red = recent studies/studies not reviewed in 2008 ISA; black = studies reviewed in the 2008 ISA for Oxides of Nitrogen.
Hazard ratios are standardized to a 10-ppb increase in NO2 and NO, and a 20-ppb increase in NOX concentration.
* NOX measured in |jg/m3. Effect estimate is per 10 |jg/m3 increase. + NOX measured in |jg/m3. Effect estimate is per 20 |jg/m3
increase.Studies are presented in descending order, with the largest mean concentration (ppb) at the top and the smallest at the
bottom of the figure. Circles = NO2; triangles = NOX; Squares = NO.
Figure 6-9 Results of studies of long-term exposure to nitrogen dioxide
(NO2), nitric oxide (NO), or the sum of NO and NO2 (NOx) and
cardiovascular mortality.
January 2015
6-164
DRAFT: Do Not Cite or Quote
-------
Table 6-16 Correspom
Study
ding risk estimates for Figure 6-9.
Location Notes
Hazard Ratio3 (95% Cl)
Cardiovascular Disease
Naess et al. (2007)
Lipsettetal. (2011)
Lipsettetal. (2011)
Caoetal. (201 1)b
Schikowski et al. (2007)
Zhang et al. (2011)
Cesaroni et al. (2013)
Brunekreef et al. (2009)
Brunekreef et al. (2009)
Hartetal. (2010)
Hartetal. (2010)
Jerrettetal. (2013)
Chen etal. (2013)
Carey et al. (2013)
Beelenetal. (2014b)
Beelenetal. (2014b)c
Norway Women
California Women, NO2
California Women, NOx
China NOx
Germany
China
Italy
the Netherlands Full cohort
the Netherlands Case cohort
U.S. Full cohort
U.S. Excluding long haul drivers
U.S.
Canada
England
Europe Pooled analysis, NO2
Europe Pooled analysis, NOx
1.06(1.00,
0.98 (0.88,
1.05(0.99,
1.02(1.01,
1.86(1.26,
5.43(4.82,
1.06(1.04,
1.04(0.96,
0.92(0.77,
1.09(1.01,
1.14(1.03,
1.12(1.02,
1.17(1.10,
1.05(1.00,
1.02(0.94,
1.02(0.99,
1.12)
1.09)
1.12)
1.04)
2.74)
6.16)
1.08)
1.13)
1.10)
1.17)
1.25)
1.22)
1.23)
1.13)
1.12)
1.06)
Cardiopulmonary Disease
Abbey etal. (1999)
Abbey etal. (1999)
Krewski et al. (2000)
Filleuletal. (2005)
Filleuletal. (2005)
Gehrinq et al. (2006)
Gehrinq et al. (2006)
Heinrich etal. (2013)
U.S. Men
U.S. Women
U.S.
France 24 areas
France 18 areas
Germany 1-yravg
Germany 5-yr avg
Germany
1.01 (0.93,
1.02(0.95,
1.01 (1.00,
1.00(0.96,
1.16(0.93,
1.70(1.28,
1.92(1.35,
1.67(1.36,
1.09)
1.09)
1.02)
1.04)
1.45)
2.26)
2.71)
2.05)
January 2015
6-165
DRAFT: Do Not Cite or Quote
-------
Table 6-16 (Continued): Corresponding risk estimates for Figure 6-9.
Study
Wanq et al. (2009b)
Hoek et al. (2002)
Yorifuji et al. (2010)
Krewski et al. (2000)
Location
Australia
the Netherlands
Japan
U.S.
Notes Hazard Ratio3 (95% Cl)
0.95(0.74,
1.45(0.99,
1.32(1.12,
1.17(1.03,
1.22)
2.13)
1.54)
1.34)
CHD
Rosenlund et al. (2008)
Rosenlund et al. (2008)
Rosenlund et al. (2008)
Ganetal. (2011)
Ganetal. (2011)
Carey et al. (2013)
Italy
Italy
Italy
Canada
Canada
England
Out of hospital 1.16(1.04,
In hospital 1.10(0.94,
Following nonfatal coronary 0.91 (0.80,
event
NO2 1.09(1.02,
NO 1.09(1.03,
0.98(0.90,
1.26)
1.30)
1.04)
1.19)
1.15)
1.07)
IHD
Lipsettetal. (2011)
Lipsettetal. (2011)
Krewski et al. (2000)
Cesaroni et al. (2013)
Hartetal. (2010)
Hartetal. (2010)
Yorifuii et al. (2010)
Jerrettetal. (2013)
Nafstad et al. (2004)b
Chenetal. (2013)
Carey et al. (2013)
Beelen etal. (2014b)
Beelen etal. (2014b)c
Beelen etal. (2014b)
Beelen etal. (2014b)c
California
California
U.S.
Italy
U.S.
U.S.
Japan
U.S.
Norway
Canada
England
Europe
Europe
Europe
Europe
Women 1.07(0.92,
Women, NOx 1.09(1.00,
1.02(1.00,
1.10(1.06,
Full cohort 1.01(0.92,
Excluding long haul drivers 1.07 (0.95,
1.57(1.04,
1.17(1.04,
NOx 1.08(1.03,
1.19(1.08,
Ml only 1.00(0.88,
Pooled analysis, all IHD; NO2 1.00 (0.84,
Pooled analysis all IHD; NOx 1 .02 (0.95,
Pooled analysis, Ml only; NO2 0.96 (0.79,
Pooled analysis, Ml only; NOx 0.99 (0.90,
1.24)
1.19)
1.03)
1.14)
1.11)
1.21)
2.36)
1.31)
1.12)
1.30)
1.13)
1.18)
1.09)
1.18)
1.07)
January 2015
6-166
DRAFT: Do Not Cite or Quote
-------
Table 6-16 (Continued): Corresponding risk estimates for Figure 6-9.
Study
Location
Notes Hazard Ratio3 (95% Cl)
Circulatory Disease
Jerrett et al. (2009)
Yorifuji et al. (2010)
Canada
Japan
2.53(1.27,
1.30(1.06,
5.11)
1.59)
Cerebrovascular Disease
Lipsett et al. (2011)
Lipsettetal. (2011)
Zhang et al. (2011)
Cesaroni et al. (2013)
Yorifuji et al. (2010)
Nafstad et al. (2004)b
Chen etal. (2013)
Beelen etal. (2014b)
Beelenetal. (2014b)c
California
California
China
Italy
Japan
Norway
Canada
Europe
Europe
Women, NO2 0.86(0.71,
Women, NOx 1.01(0.90,
5.35(4.67,
1.02(0.98,
1.18(0.89,
NOx 1.04(0.94,
0.92(0.81,
Pooled analysis, NO2 1.02 (0.87,
Pooled analysis, NOx 1.00 (0.93,
1.06)
1.14)
6.11)
1.06)
1.57)
1.15)
1.10)
1.20)
1.08)
Stroke
Jerrett etal. (2013)
Carey etal. (2013)
U.S.
England
1.20(1.04,
1.00(0.91,
1.39)
1.09)
Heart Failure
Carey etal. (2013)
England
1.09(0.91,
1.32)
Diabetes
Raaschou-Nielsen et al. (201 3b)
Denmark
1.66(0.96,
2.89)
CHD = coronary heart disease; Cl = confidence interval; IHD = ischemic heart disease; Ml = myocardial infarction; NO = nitric
oxide; NO2 = nitrogen dioxide; NOx = sum of NO and NO2
Note: Studies correspond to those presented in Figure 6-9.
aEffect estimates are standardized to a 10-ppb increase in NO2 and NO, or a 20-ppb increase in NOX concentration.
bNOx measured in |jg/m3. Effect estimate is per 10 |jg/m3 increase.
CNOX measured in |jg/m3. Effect estimate is per 20 |jg/m3 increase.
January 2015
6-167
DRAFT: Do Not Cite or Quote
-------
MEAN
STUDY
Abbey etal. 1999
Lipsettetal. 2011
Caoetal. 2011*
Dong etal. 2012
Cesarom et al. 2013
Brunekreef et al. 2009
Heinrich et al. 2013
Jerrett et al. 2009
Hart etal. 2010
Yorifuji et al. 2010
Jerrett etal. 2013
Carey etal. 2013
Nafstad et al. 2004*
Dimakopoulou et al. 2014
Dimakopoulou et al. 2014
Katanoda et al. 2011
Naess et al. 2007
Ganetal. 2013
Hart etal. 2010
Yorifuii et al. 2010
Carey etal. 2013
Abbey etal. 1999
Lipsettetal. 2011
Naess et al. 2007
Krewski et al. 2000
Caoetal. 2011*
Filleul et al. 2005
Cesarom et al. 2013
Brunekreef et al. 2009
Heinrich etal. 2013
Hart etal. 2010
Nyberg et al. 2000
Yorifuji et al. 2010
Jerrett etal. 2013
Carey etal. 2013
Nafstad et al. 2004*
Krewski et al. 2000
Katanoda et al. 2011
COHORT CONCENTRATIOP
AHSMOG Cohort 36.7
CA Teachers 33.6
National Hypertension 26.5
Shenyang Cohort 24.5
Rome 23.4
German Women 20.7
Toronto Cohort 19.5
TrIPS 14.2
Shizuoka Elderly 13.3
ACS-CA 12.3
National English 11.9
Norwegian Men 10.6
^ESCAPE 2.8-28.1
Three -prefecture 1.2-33.7
Oslo Cohort 27.5-44.9
Vancouver Cohort 17.1
26
TrIPS 14.2
Shizuoka Elderly 13.3
National English 11.9
AHSMOG Cohort 36.7
CA Teachers 33.6
Oslo Cohort 27.5-44.9
ACS Cohort 27.9
National Hypertension 26.5
PAARC Survey 24.5
Rome 23.4
NLCS-AIR 20.7
German Women 20.7
TrIPS 14.2
Stockhom 13.3
Shizuoka Elderly 13.3
ACS-CA 12.3
National English 11.9
Norwegian Men 10.6
Six Cities -Reanalysis 6.1-21.9
Three-prefecture 1.2-33.7
*NOTES 1
Men •J—
Women ^^
Women «H —
Full Cohort — ' • ~ —
. *
All • -•-
Male " -O-
Female | -•—
Women — ;-•
Women — Ł —
Men T-«—
Women . — • —
24 areas -+^
Full Cohort^ • | —
Excluding Long Haul Drivers — 1 — *
* |
i_^*~
All . *-•-
Respiratory
COPD
Lung Cancer
Men
Women
H 1 1 1 1 1 1-
0.5
Hazard Ratio and 95% Cl
ACS = American Cancer Society; AHSMOG = California Seventh-day Adventists Cohort; ESCAPE = European Study of Cohorts for
Air Pollution Effects; NLCS-AIR = Netherlands Cohort Study of Air Pollution and Mortality; PAARC = Air Pollution and Chronic
Respiratory Diseases; TrIPS = Traffic Industry Particle Study.
Note: Red = recent studies; black = studies reviewed in the 2008 ISA for Oxides of Nitrogen. Hazard ratios are standardized to a
10-ppb increase in NO2 and NO, and a 20-ppb increase in NOX concentration. 'Effect estimates from studies measuring NOX in
|jg/m3 have not been standardized. Studies are presented in descending order, with the largest mean concentration (ppb) at the top
and the smallest at the bottom of the figure. Circles = NO2; triangles = NOX; squares = NO.
Figure 6-10 Results of studies of long-term exposure to nitrogen dioxide
(NO2), nitric oxide (NO), or the sum of NO and NO2 (NOx), and
respiratory mortality.
January 2015
6-168
DRAFT: Do Not Cite or Quote
-------
Table 6-17 Corresponding risk estimates for Figure 6-10.
Study
Location Notes
Hazard
Ratio3 (95% Cl)
Respiratory
Abbey et al. (1999)
Abbey et al. (1999)
Lipsettetal. (2011)
Lipsettetal. (2011)
Caoetal. (201 1)b
Donqetal. (2012)
Cesaroni et al. (2013)
Brunekreef et al. (2009)
Brunekreef et al. (2009)
Heinrichetal. (2013)
Jerrett et al. (2009)
Hartetal. (2010)
Hartetal. (2010)
Yorifuii et al. (2010)
Jerrett et al. (2013)
Carey et al. (2013)
Nafstad et al. (2004)
Dimakopoulou etal. (2014)
Dimakopoulou et al.
(2014)c
Katanoda et al. (2011)
Katanoda et al. (2011)
Katanoda et al. (2011)
U.S. Men
U.S. Women
California Women, NO2
California Women, NOx
China NOx
China
Italy
the Netherlands Full cohort
the Netherlands Case cohort
Germany
Canada
U.S. Full cohort
U.S. Excluding long haul drivers
Japan
U.S.
England
Norway NOx
Europe Pooled analysis, NO2
Europe Pooled analysis, NOx
Japan All
Japan Men
Japan Women
0.93 (0.82,
0.98 (0.87,
0.93 (0.76,
0.94 (0.83,
1.03(1.00,
6.1 (5.2, 7.
1.06(1.00,
1.22(1.00,
1.16(0.83,
1.15(0.67,
1.16(0.37,
1.07(0.91,
1.26(1.01,
1.39(1.04,
1.00(0.83,
1.28(1.18,
1.16(1.06,
0.94 (0.80,
0.99 (0.90,
1.16(1.12,
1.11 (1.05,
1.25(1.18,
1.07)
1.11)
1.15)
1.07)
1.06)
2)
1.12)
1.48)
1.62)
2.00)
2.71)
1.27)
1.56)
1.83)
1.20)
1.38)
1.26)
1.10)
1.09)
1.21)
1.18)
1.33)
COPD
Naess et al. (2007)
Naess et al. (2007)
Norway Men
Norway Women
1.18(1.04,
1.05(0.93,
1.33)
1.18)
January 2015
6-169
DRAFT: Do Not Cite or Quote
-------
Table 6-17 (Continued): Corresponding risk estimates for Figure 6-10.
Study
Ganetal. (2013)
Ganetal. (2013)
Hartetal. (2010)
Hartetal. (2010)
Yorifuii et al. (2010)
Carey et al. (2013)
Location
Canada
Canada
U.S.
U.S.
Japan
England
Notes
NO2
NO
Full cohort
Excluding long haul drivers
Hazard
1.09(0.91,
1.06(0.97,
0.97 (0.79,
1.01 (0.82,
1.22(0.63,
1.13(0.98,
Ratio3 (95% Cl)
1.32)
1.15)
1.19)
1.44)
2.31)
1.28)
Lung Cancer
Abbey et al. (1999)
Abbey et al. (1999)
Lipsett et al. (2011)
Lipsettetal. (2011)
Naess et al. (2007)
Naess et al. (2007)
Krewski et al. (2000)
Caoetal. (201 1)b
Filleuletal. (2005)
Filleuletal. (2005)
Cesaroni et al. (2013)
Brunekreef et al. (2009)
Brunekreef et al. (2009)
Heinrich etal. (2013)
Hartetal. (2010)
Hartetal. (2010)
Nyberq et al. (2000)
Nyberq et al. (2000)
Yorifuii et al. (2010)
Jerrettetal. (2013)
U.S.
U.S.
California
California
Norway
Norway
U.S.
China
France
France
Italy
the Netherlands
the Netherlands
Germany
U.S.
U.S.
Sweden
Sweden
Japan
U.S.
Men
Women
Women, NO2
Women, NOx
Men
Women
NOx
24 areas
18 areas
Full cohort
Case cohort
Full cohort
Excluding long haul drivers
30-yr exposure
10-yr exposure
1.35(0.96,
1.69(1.07,
1.00(0.76,
0.98 (0.90,
1.06(0.97,
1.20(1.09,
0.99 (0.97,
1.03(0.99,
0.94 (0.89,
1.12(0.77,
1.08(1.04,
0.94(0.81,
0.87 (0.66,
1.56(0.91,
1.07(0.96,
1.09(0.95,
1.10(0.87,
1.20(0.94,
0.91 (0.63,
1.29(1.05,
1.90)
2.65)
1.32)
1.07)
1.15)
1.32)
1.01)
1.07)
1.02)
1.61)
1.14)
1.09)
1.14)
2.69)
1.19)
1.25)
1.36)
1.48)
1.34)
1.59)
January 2015
6-170
DRAFT: Do Not Cite or Quote
-------
Table 6-17 (Continued): Corresponding risk estimates for Figure 6-10.
Study
Carey et al. (2013)
Nafstad et al. (2004)b
Krewski et al. (2000)
Katanoda et al. (2011)
Katanoda et al. (2011)
Katanoda et al. (2011)
Location Notes
England
Norway
U.S.
Japan
Japan
Japan
NOx
All
Men
Women
Hazard
1.20(1.09,
1.11 (1.03,
1.09(0.76,
1.17(1.10,
1.18(1.11,
1.13(1.01,
Ratio3 (95% Cl)
1.32)
1.19)
1.57)
1.26)
1.26)
1.27)
Cl = confidence interval; COPD = chronic obstructive pulmonary disease; NO = nitric oxide; NO2 = nitrogen dioxide; NOx = sum of
NO and NO2.
Note: Studies correspond to those presented in Figure 6-10.
aEffect estimates are standardized to a 10-ppb increase in NO2 and NO or a 20-ppb increase in NOX concentration.
bNOx measured in |jg/m3. Effect estimate is per 10 |jg/m3 increase.
CNOX measured in |jg/m3. Effect estimate is per 20 |jg/m3 increase.
6.5.3 Summary and Causal Determination
1 Collectively, the evidence is suggestive, but not sufficient, to infer a causal relationship
2 between long-term exposure to NC>2 and mortality among adults. The strongest evidence
3 comes from cohort studies conducted in the U.S., Canada, and Europe, which show
4 consistent, positive associations with total mortality, as well as deaths due to respiratory
5 and cardiovascular disease (Chen et al.. 2013; Gan etal.. 2013; Hart et al.. 2013; Heinrich
6 etal.. 2013; Jerrett etal.. 2013; Gan etal.. 2011; Lipsett etal.. 2011; Hartet al.. 2010;
7 Brunekreef etal.. 2009; Jerrett et al.. 2009; Beelen et al.. 2008b; Schikowski et al.. 2007;
8 Krewski et al.. 2000). The results from these studies are coherent with studies that have
9 observed associations between long-term exposure to NCh and respiratory hospital
10 admissions (Andersen et al.. 2012a; Andersen et al.. 2011) and cardiovascular effects
11 (Lipsett et al.. 2011; Hart etal.. 2010). Additionally, the evidence for short- and
12 long-term respiratory and cardiovascular morbidity provides some biological plausibility
13 for mortality.
14 Many of the studies evaluating the associations between long-term exposure to NO2 and
15 mortality have used concentrations measured at central site monitors to assign exposure.
16 A select number of recent studies have employed exposure assessment methods such as
17 LUR to represent the spatial variability of NC>2. There was no distinguishable pattern or
18 trend in the results of this body of evidence that could be attributed to the use of central
19 site monitors or LUR in order to assign exposure. Exposure assessment was evaluated
January 2015 6-171 DRAFT: Do Not Cite or Quote
-------
1 drawing upon discussions in Sections 3.2 and 3.4.5. In general, LURmodel predictions
2 have been found to correlate well with outdoor NO2 concentration measurements
3 (Section 3.2.1.1). which may explain why the results for this evidence base were
4 consistent across these exposure assessment types.
5 In past reviews, a limited number of epidemiologic studies had assessed the relationship
6 between long-term exposure to NO2 and mortality in adults. The 2008 ISA for Oxides of
7 Nitrogen concluded that the scarce amount of evidence was "inadequate to infer the
8 presence or absence of a causal relationship" (U.S. EPA. 2008). Recent studies provide
9 evidence for an association between long-term exposure to NO2 or NOx and mortality
10 from extended analyses of existing cohorts as well as original results from new cohorts in
11 the U.S., Europe, and Asia. Recent studies have examined the potential for copollutant
12 confounding by evaluating copollutant models that include PM2 5, PMio, PMio-2.5, and
13 SO2 (Beelenetal. 2014a: Jerrettet al.. 2013; Hartet al. 2010). These recent studies
14 address a previously identified data gap. The NO2 results from these models were
15 generally attenuated with the inclusion of copollutants, though a key traffic-related
16 copollutant (i.e., EC) was not evaluated. It remains difficult to disentangle the
17 independent effect of NO2 from the potential effect of the traffic-related pollution mixture
18 or other components of that mixture.
19 While the results were generally consistent across studies, there were several
20 well-designed, well-conducted studies that did not observe an association between
21 long-term exposure to NO2 and mortality (Beelen et al.. 2014a: Beelen et al.. 2014b:
22 Dimakopoulou et al.. 2014; Krewski et al.. 2009; Pope et al.. 2002; Abbey et al.. 1999).
23 All available evidence for mortality due to long-term exposure to NO2 or NOx was
24 evaluated using the framework described in Table II of the Preamble. The key evidence
25 as it relates to the causal determination is summarized in Table 6-18. The overall
26 evidence is suggestive, but not sufficient, to infer a causal relationship between long-term
27 exposure to NO2 and total mortality among adults.
January 2015 6-172 DRAFT: Do Not Cite or Quote
-------
Table 6-18 Summary of evidence, which is suggestive, but not sufficient, to infer
a causal relationship between long-term nitrogen dioxide (NO2)
exposure and total mortality.
Rationale for
Causal
Determination3 Key Evidence13
NOx or NO2
Concentrations
Associated with
Key References13 Effects0
High-quality
epidemiologic
studies are
generally
supportive but not
entirely consistent
Positive association between long-term Krewski et al. (2000)
exposure to NO2 and mortality in the HSC
cohort and a subset of the ACS cohort,
with effect estimates similar in magnitude
to those observed with PIVh.s, even after
adjustment for common potential
confounders.
Mean concentrations
across cities (1980):
6.1-21.9 ppb
Jerrettetal. (2013)
Mean (1988-2002):
12.3 ppb
Updated results from the NLCS report a Beelen etal. (2008b)
positive association with total mortality,
effects for respiratory mortality stronger
than any observed with traffic variables
and total or other cause-specific mortality.
Mean (1987-1996):
20.7 ppb
Brunekreef et al. (2009) Max: 35.5 ppb
Updated results from the German
women's cohort report positive
associations with total and
cardiopulmonary mortality.
Heinrichetal. (2013) Mean: 20.7 ppb
Schikowski et al. (2007) Median: 24.4 ppb
Recent cohort studies in the U.S. observe Hart et al. (2010)
increases in total mortality and mortality
due to cardiovascular disease in separate
cohorts of men and women.
Mean (1985-2000):
14.2 ppb
Lipsettetal. (2011)
Mean (1996-2005):
33.6 ppb; Max: 67.2 ppb
Hart etal. (2013)
Median (2000):
13.9 ppb
Positive associations with total,
cardiovascular, respiratory, and lung
cancer mortality in Canadian cities.
Chen etal. (2013)
Mean (across cities):
12.1-21.7 ppb
Jerrett et al. (2009)
Median: 22.9 ppb
Ganetal. (2011)
Mean: 17.0 ppb
Ganetal. (2013)
Mean: 17.1 ppb
January 2015
6-173
DRAFT: Do Not Cite or Quote
-------
Table 6-18 (Continued): Summary of evidence, which is suggestive, but not
sufficient, to infer a causal relationship between long-
term nitrogen dioxide (NO2) exposure and total mortality.
Rationale for
Causal
Determination3
Key Evidence13
Key References13
NOx or
Concentrations
Associated with
Effects0
Uncertainty
remains regarding
independent
effects of NO2
Associations with mortality generally
attenuated with adjustment for PM-io,
PM2.5, PM-io-2.5, or SO2, but analysis is
limited and does not include other
traffic-related copollutants (i.e., EC).
When reported, correlations with
copollutants were highly variable (low to
high).
Beelen et al. (2014a).
Jerrettetal. (2013),
Hartetal. (2010)
Mean 2.8-31.7 ppb
Mean (1988-2002):
12.3 ppb
Mean (1985-2000):
14.2 ppb
Some studies
show no
association
No association in several reanalyses of
the ACS cohort.
Krewski et al. (2000),
Pope etal. (2002),
Krewski et al. (2009)
Mean (1982-1998):
21.4-27.9 ppb
Mean (1982-1998) 27.9
ppb; Max 51.1 ppb
No association observed in a multicenter
European study of pooled data from 22
existing cohort studies for total,
respiratory, or cardiovascular mortality.
Beelen etal. (2014a).
Dimakopoulou etal.
(2014).
Beelen etal. (2014b)
Range of means across
cohorts: 2.8-31.7 ppb
No association with total,
cardiopulmonary, or respiratory mortality
intheAHSMOG.
Abbey etal. (1999)
Mean (1973-1992):
36.8 ppb
Limited coherence
with evidence for
respiratory and
cardiovascular
morbidity
Limited evidence for respiratory
hospitalizations in adults coherent with
evidence for respiratory mortality.
Andersen etal. (2011), 35-yr mean: 9.0 ppb
Andersen et al. (2012a) 25-yr mean: 9.5 ppb
Some inconsistencies reported for
cardiovascular morbidity. Evidence for Ml
and heart failure coherent with evidence
for cardiovascular mortality.
Lipsettetal. (2011)
Mean: 33.6 ppb;
Max: 67.2 ppb
Atkinson et al. (2013)
Mean: 12.0 ppb;
Max: 32.3 ppb
ACS = American Cancer Society; AHSMOG = Adventist Health Study of Smog; EC = elemental carbon; HSC = Harvard Six
Cities; Ml = myocardial infarction; NLCS = Netherlands Cohort Study on Diet and Cancer; NO2 = nitrogen dioxide;
PM2.5 = particulate matter with a nominal aerodynamic diameter less than or equal to 2.5 |jm; PM10 = particulate matter with a
nominal aerodynamic diameter less than or equal to 10 |jm; PMio-2.5 = particulate matter with a nominal aerodynamic diameter
less than or equal to 10 |jm and greater than a nominal diameter of 2.5 |jm; SO2 = sulfur dioxide.
aBased on aspects considered in judgments of causality and weight of evidence in causal framework in Tables I and N. of the
Preamble.
""Describes the key evidence and references contributing most heavily to causal determination and, where applicable, to
uncertainties or inconsistencies. References to earlier sections indicate where the full body of evidence is described.
°Describes the NOX or NO2 concentrations with which the evidence is substantiated (for experimental studies, below 5,000 ppb).
January 2015
6-174
DRAFT: Do Not Cite or Quote
-------
6.6 Cancer
1 The 1993 AQCD for Oxides of Nitrogen and the 2008 ISA for Oxides of Nitrogen
2 reported that there was no clear evidence that NO2 or other oxides of nitrogen act as a
3 complete carcinogen. The U.S. Department of Health and Human Services, the
4 International Agency for Research on Cancer, and the U.S. Environmental Protection
5 Agency (EPA) have not classified nitrogen oxides for potential carcinogenicity. The
6 American Conference of Industrial Hygienists has classified NO2 as A4 (Not classifiable
7 for humans or animals). The 2008 Oxides of Nitrogen ISA (U.S. EPA. 2008) included a
8 few epidemiologic studies of oxides of nitrogen and cancer, both examining lung cancer
9 incidence and reporting positive associations. Since the 2008 ISA for Oxides of Nitrogen
10 (U.S. EPA. 2008). additional studies have been published exploring this relationship. In
11 addition, epidemiologic studies have been performed examining the relationship between
12 NO2 and leukemia, bladder cancer, breast cancer, and prostate cancer. These are all
13 described in more detail in supplementary Table S6-10 (U.S. EPA. 2013J). which
14 includes information on the exposure assessment and duration, as well as effect estimates.
15 Many studies assigned exposure using local air monitors, but others derived exposure
16 estimates using LUR and dispersion models. Details on these methods of exposure
17 assessment can be found in Sections 3.2.1.1 and 3.2.3. respectively. Information
18 in Section 3.4.5.2 aids interpreting these methods with regard to potential exposure
19 measurement error.
6.6.1 Lung Cancer
6.6.1.1 Epidemiologic Studies
Lung Cancer Incidence
20 Two previous studies included in the 2008 ISA for Oxides of Nitrogen (U.S. EPA. 2008)
21 reported positive associations between NO2 or NOx and lung cancer incidence (Nafstad
22 etal.. 2003; Nyberg et al.. 2000). Nyberg et al. (2000) reported an association between
23 NO2 and lung cancer at the highest 10-yr avg concentrations of NO2 with a 20-year lag.
24 This association was robust to inclusion of SO2, which was not observed to be associated
25 with lung cancer (Pearson correlation coefficient between SO2 and NO2 ranged from 0.5
26 to 0.7). Nafstad et al. (2003) performed a study with 24 years of follow-up and reported a
27 positive association between NOx concentrations and lung cancer incidence during the
January 2015 6-175 DRAFT: Do Not Cite or Quote
-------
1 early years of the study, but the authors report more recent years had weaker associations
2 (results were not provided). The Pearson r between NOx and 862 was 0.63, and no
3 association was observed between 862 concentration and cancer.
4 A recent study examined the association between NO2 concentration and lung cancer
5 incidence within the NLCS using over 11 years of follow-up (Brunekreef et al.. 2009;
6 Beelen et al., 2008a). The researchers observed no association in analyses using
7 case-cohort and full cohort approaches. The associations between lung cancer and 862
8 (correlation coefficient with NC>2 >0.6) and PIVb 5 (correlation coefficient with NC>2 >0.8)
9 were also examined and found to be null.
10 A meta-analytical study in Europe combined individual estimates from cohort studies
11 across nine countries in Europe (Raaschou-Nielsen et al., 2013a). Although positive
12 associations between concentrations of NOx and NO2 and lung cancer were detected in
13 models adjusting for age, sex, and calendar time, these associations became null when
14 other confounders, such as smoking-related covariates, fruit intake, and area-level SES,
15 were included. No associations were observed for PIVbs, PNfe.s absorbance, and PMio-25,
16 but PMio was positively associated with lung cancer.
17 A Danish study combined three cohorts and reported an association between increased
18 NOx concentrations and lung cancer incidence (Raaschou-Nielsen et al.. 2010a). This
19 increased incidence with NOx exposure persisted in some models of specific cancer
20 types, such as squamous cell carcinomas. When examining the associations by sex,
21 length of education, and smoking status, the precision was decreased (i.e., wider 95%
22 CI), and no differences were observed between the groups. One of these cohorts was used
23 in another study where the follow-up period was extended 5 years to include more cases
24 (Raaschou-Nielsen et al.. 201 lb). This study detected an increased incidence rate of lung
25 cancer in the highest quartile of NOx concentrations. Further analyses evaluated
26 interactions with sex, smoking status, length of school attendance, and daily fruit intake.
27 An increased association between NOx concentration and lung cancer incidence was
28 observed among individuals with at least 8 years of schooling, but no association was
29 apparent among those with less schooling.
30 A case-control study of molecular changes and genetic susceptibility in relation to air
31 pollution reported on nonsmokers and lung cancer incidence (Papathomas et al.. 2011).
32 This study used multiple statistical analysis techniques to evaluate the associations
33 between air pollutants and lung cancer incidence. Although profile regression analyses
34 reported higher NO2 exposures for the higher risk grouping, logistic regression analyses
35 did not find an association between NO2 and lung cancer incidence. The same was true of
36 PMio. In another statistical model by the authors, NO2 was not chosen as a predictor,
37 whereas PMio concentration was chosen.
January 2015 6-176 DRAFT: Do Not Cite or Quote
-------
1 A study in Canada reported a positive association between NO2 concentrations from
2 national spatiotemporal models and lung cancer incidence (Hystad et al. 2013). When
3 examining types of lung cancer, the association was present for adenocarcinomas but not
4 squamous cell carcinoma. Confidence intervals for estimates of small cell and large cell
5 carcinomas were wide and included the null. Associations were not present between Os
6 and lung cancer. PM2 5 demonstrated some associations with lung cancer incidence,
7 especially in the third and fourth, but not fifth, quintiles of exposures. When NC>2 and Os
8 were considered in copollutant models, the ORs increased for both pollutants.
9 Copollutant models were not examined for NC>2 and PIVb 5 because of the high correlation
10 between the pollutants. Odds of lung cancer increased in association with NO2
11 concentrations when the analysis was limited to the closest monitor within 50 km, but the
12 authors state that NO2 estimates, "are also capturing a component of PlVfc 5, due to the
13 correlation between the two pollutants." The authors also believe it could be the result of
14 more accurate exposure assessment or restriction of the study area. In stratified analyses,
15 associations appear to be greater among men, with null results among women. No
16 differences were clear in stratified analyses of education or smoking status. Another
17 Canadian study also reported a positive association between NO2 concentrations and lung
18 cancer when using a population-based control group (Villeneuve et al., 2014). The
19 association was present when using NC>2 concentration at the time of the interview, from
20 10 years prior, and from a time-weighted average. However, in analyses adjusted for
21 personal (e.g., age, smoking, BMI) and ecological (e.g., neighborhood unemployment
22 rate) covariates using hospital-based and population-based controls, no association was
23 detected. Associations were observed between lung cancer and benzene and total
24 hydrocarbons when using population-based controls but only for total hydrocarbons and
25 time-weighted average of benzene when using all controls.
26 An ecologic study examined NC>2 and NO concentrations in relation to lung cancer rates
27 (Tseng etal.. 2012). No associations were observed for NO2 concentrations and all lung
28 cancer cases combined, adenocarcinomas, or squamous cell carcinomas. A positive
29 association was observed in the highest quartile of NO for adenocarcinomas but not for
30 squamous cell carcinomas. Associations were also reported for SO2 concentrations but
31 not Os concentrations, CO concentrations, or PMio concentrations.
32 In summary, multiple studies have examined the associations between concentrations of
33 oxides of nitrogen and lung cancer incidence. Positive associations were reported in
34 multiple studies, but other studies reported no associations. The inconsistency observed
35 between studies does not appear to be related to the length of the exposure or follow-up
36 period. Inconsistency is observed with NO2 exposure assessed from central site
37 monitors/spatiotemporal pollutant models (Brunekreef et-al.. 2009; Beelen et-al.. 2008a.
38 Papathomas et-al., 2011. Hystad et-al., 2013. Tseng et-al.. 2012), NO2 estimated for
January 2015 6-177 DRAFT: Do Not Cite or Quote
-------
1 subject's residences using LUR (Raaschou-Nielsen et-al., 2013. Villeneuve et-al.. 2014).
2 and NC>2 estimated for subject's residences using dispersion models (Raaschou-Nielsen
3 et-al.. 2010a. Raaschou-Nielsen et-al.. 201 Ib. Nafstad et-al.. 2003; Nvberg et-al.. 2000).
4 Given the differences among the study designs, it cannot be concluded that the
5 inconsistencies are related to exposure assessment method or length of follow-up periods.
6 In general, studies using central site monitors for exposure estimates carry uncertainty in
7 long-term NC>2 exposure studies because the exposure error resulting from spatial
8 misalignment between subjects' and monitor locations can overestimate or underestimate
9 associations with health effects (Section 3.4.5.2). However, an association was observed
10 with residential NC>2 exposure estimated for subjects' homes, and the improved spatial
11 resolution of the exposure estimate lends more confidence in the association.
Lung Cancer Mortality
12 Two HEI Research Reports have investigated the association between NO2 concentration
13 and lung cancer mortality using large cohorts with follow-ups of at least 10 years.
14 Brunekreef et al. (2009) (see also. Beelen et al.. 2008b) reported no association between
15 NC>2 and lung cancer mortality using the NLCS, and results were not changed with the
16 inclusion of a traffic-intensity variable. No association was observed between lung cancer
17 mortality and other pollutants (862, correlation coefficient with NC>2 >0.6, or PM2 5,
18 correlation coefficient with NCh >0.8). Krewski et al. (2009) utilized an extended
19 follow-up of the American Cancer Society Study and reported no associations between
20 NC>2 and lung cancer mortality. An association with lung cancer mortality for PIVb 5 was
21 noted in this report but not for CO, Os, or SCh. However, a study utilizing the American
22 Cancer Society's Cancer Prevention Study II cohort reported a positive association
23 between NCh concentrations and lung cancer mortality (Jerrett et al.. 2013). Pearson
24 correlation coefficients were about 0.55 for the association between NO2 and PIVb 5 as
25 well as for the association between NO2 and Os. The positive association between NO2
26 and lung cancer mortality was robust to adjustment with PM2 5 or Os, although when both
27 PM2 5 and Os were included in the model with NC>2, the 95% CIs widened.
28 Inconsistent findings between NC>2 and lung cancer mortality have been reported in
29 studies conducted across Europe. A positive association was observed between NO2 and
30 lung cancer mortality in a large study conducted in Rome, Italy (Cesaroni et al., 2013).
31 The association demonstrated a linear relationship. No effect measure modification was
32 apparent by age, sex, educational level, area-based socioeconomic position, or moving
33 history. NCh was highly correlated with PM2 5, which was also associated with lung
34 cancer mortality. In England, a study using a large nationally representative database
35 reported a positive association between NO2 concentrations and lung cancer mortality
36 (Carey et al.. 2013). Among other pollutants, no associations were observed in fully
January 2015 6-178 DRAFT: Do Not Cite or Quote
-------
1 adjusted models for PMio, PNfe.s, and Os. SO2 concentrations were positively associated
2 with lung cancer mortality in some adjusted analyses. A study in France reported a
3 positive association between NO2 and lung cancer mortality only after exclusion of areas
4 with air monitoring sites reporting a high ratio of NO to NC>2 [which implied a strong
5 influence of heavy traffic near the monitor that may not represent the air pollution
6 concentrations in the entire area (Filleul etal.. 2005)]. Correlations between NC>2 and
7 other air pollutants ranged from 0.22 to 0.86. No other air pollutants examined in the
8 study (SC>2, total suspended particles, black smoke, and NO) were associated with lung
9 cancer mortality. A study in Norway examined 4 years of air pollution and mortality data
10 (Naess etal.. 2007). Positive associations between NO2 and lung cancer mortality were
11 observed among women aged 51-70 years and 71-90 years but not among men in these
12 age groups (although a positive association was reported in the crude HR for
13 71-90 year-old men). Correlations between the pollutants examined (NO2, PMio, and
14 PM2.5) were not reported individually but ranged from 0.88 to 0.95. Associations between
15 lung cancer and the other pollutants were similar to those observed for NO2. In a
16 nonparametric smooth analysis that combined the sexes, the increase in log odds for lung
17 cancer appears to begin around 21.3 ppb for 51-70 year-olds while the increase appears
18 to be at lower concentrations among those aged 71-90 years. A large study of women
19 from Germany followed up women who were originally enrolled in cross-sectional
20 studies in the 1980s and 1990s (Heinrich etal., 2013). Using NO2 concentration from
21 their address at the baseline examination, the authors reported no association between
22 NO2 concentration and lung cancer mortality. The Spearman's correlation coefficient for
23 PMio, which was observed to be associated with lung cancer, and NO2 was 0.5. A large
24 cohort of men employed by the U.S. trucking industry in 1985 were matched to records
25 in the National Death Index through 2000 (Hart etal.. 2011). Using NO2 concentrations
26 at their residential address, the association with lung cancer mortality was examined. No
27 association was detected, and this persisted when long-haul drivers who are away from
28 the home at least one night per week were excluded from the analyses. Similar results
29 were observed for PMio and SO2.
30 Multiple studies of NO2 and lung cancer mortality have been conducted in Asia. A study
31 in Japan followed individuals aged 65-84 years at enrollment for about 6 years (Yorifuji
32 etal.. 2010). No overall association was reported between NO2 concentration and lung
33 cancer mortality. In stratified analyses, the association between NO2 concentration and
34 lung cancer mortality was higher among nonsmokers compared to former/current
35 smokers, but the findings were imprecise and the 95% confidence intervals overlapped.
36 No difference in the association was observed among other stratification variables (age,
37 sex, BMI, hypertension, diabetes, and financial ability). A continuation of this study was
38 conducted using additional years of follow-up and NO2 concentrations assigned at the
39 year of the outcome (Yorifuji et al., 2013). This follow-up reported a positive association
January 2015 6-179 DRAFT: Do Not Cite or Quote
-------
1 between NCh concentration and lung cancer mortality. Results were similar for
2 nonsmokers compared to former and current smokers. When examining exposure over
3 various years (ranging from 1 year before death to an average from baseline to death), the
4 association remained. Another study in Japan followed individuals for 10 years and
5 observed a positive association between NO2 concentration and lung cancer mortality
6 (Katanoda et al.. 2011). An association was also observed for suspended PM (Pearson
7 correlation coefficient with NCh = 0.26) but not for SC>2. When the association between
8 NC>2 concentration and lung cancer mortality was examined by region, the association
9 appears to persist only in the areas of study with the highest NC>2 concentration (data on
10 association by region only presented in figures; numerical estimates not provided). A
11 national study of urban areas in China had a follow-up of less than 10 years and reported
12 no association between NOx and lung cancer mortality (Cao et al.. 2011). This lack of
13 association was robust to inclusion of TSP or SC>2, of which SCh concentrations were
14 found to be associated with lung cancer mortality. A study performed in Taiwan used a
15 case-control approach, comparing women who died of lung cancer or other
16 nonrespiratory related causes (Liu et al.. 2008). The highest tertile of NC>2 concentration
17 was positively associated with lung cancer mortality. Associations between pollution
18 concentrations and lung cancer mortality were also observed for CO, but not 862, PMio,
19 or Os. A combined exposure category was created, examining those women with
20 estimated exposure concentrations of CO and NO2 in the highest tertiles compared to
21 those in the lowest tertiles. The results were similar to those of the single-pollutant
22 estimates.
23 Overall, there are inconsistent findings among studies of NO2 and lung cancer mortality
24 (see Table S6-10 for quantitative results). Most of these studies controlled for
25 confounders, such as smoking. Inconsistency is observed with NO2 exposure assessed
26 from central site monitors/spatiotemporal pollutant models (Brunekreef et al.. 2009;
27 Beelen et al.. 2008a.. Krewski et al. (2009). Filleul et al.. 2005. Heinrich et al.. 2013.
28 Katanoda et al.. 2011. Cao et al.. 2011. Liu et al.. 2008X NO? estimated for subject's
29 residences using LUR (Cesaroni et al.. 2013. Jerrett et al., 2013. Hart et al., 2011.
30 Yorifuji et al.. 2010. Yorifuji et al.. 2013). and NO2 estimated for subject's residences
31 using dispersion models (Carey et al.. 2013. Naess et al.. 2007). Given the differences
32 among the study designs, it cannot be concluded that the inconsistencies are related to
33 exposure assessment method or length of follow-up periods. In general, studies using
34 central site monitors for exposure estimates carry uncertainty in long-term NO2 exposure
35 studies because the exposure error resulting from spatial misalignment between subjects'
36 and monitor locations can overestimate or underestimate associations with health effects
37 (Section 3.4.5.2). However, an association was observed with residential NO2 exposure
38 estimated for subjects' homes, and the improved spatial resolution of the exposure
39 estimate lends more confidence in the association.
January 2015 6-180 DRAFT: Do Not Cite or Quote
-------
6.6.1.2 Animal Toxicological Studies
Lung Tumors with Co-exposure with Known Carcinogens
1 The 1993 AQCD for Oxides of Nitrogen and the 2008 ISA for Oxides of Nitrogen
2 detailed NO2 co-exposure with known carcinogens. NO2 has been reported to act as a
3 tumor promoter at the site of contact, i.e., in the respiratory tract after inhalation
4 exposure. Toxicological studies of NO2 and carcinogenicity and genotoxicity are
5 described in Table 6-18. This is consistent with mechanistic evidence of observed
6 hyperplasia of respiratory epithelium with NO2 exposure (see Section 6.2.6). Rats
7 injected with the carcinogen N-bis (2-hydroxy-propyl) nitrosamine (BHPN) and
8 continuously exposed to 40, 400, or 4,000 ppb NO2 for 17 months developed a
9 nonstatistically significant fivefold increase in incidence of adenomas or
10 adenocarcinomas of the lungs versus control animals [4,000 ppb NO2; (Ichinose et al..
11 1991)1. Another study by the same lab (Ichinose and Sagai. 1992) showed statistically
12 significant increases in BHPN-induced lung tumors with combined NO2 + Os exposure, a
13 multipollutant effect absent with exposure to either single pollutant (BHPN injection
14 followed the next day by either clean air 0% NO2, 500 ppb NO2, 50 ppb NO2 + 400 ppb
15 Os, or 400 ppb Os + 1 mg/m3 H2SO4 for 13 months, and then recovery with clean air for
16 another 11 months); continuous NO2 exposure, 11 h/day H2SO.4, or Os exposure).
17 Another study with coexposure of F344 male rats to diesel exhaust particle extract-coated
18 carbon black particles (DEPcCBP) and NO2 and/or SO2 found significantly increased
19 incidences of lung tumors (alveolar adenomas) for the animals coexposed to DEPcCBP
20 and NO2 and/or SO2 but not in those with DEPcCBP exposure alone (Ohyama et al..
21 1999). The National Toxicology Program's Report on Carcinogens has stated DEP is
22 reasonably anticipated to be a human carcinogen (NTP. 2011). Exposed rats received
23 intratrachael (IT) installation of DEPcCBP once per week for 4 weeks, and 6,000 ppb
24 NO2, 4,000 ppb SO2, or 6,000 ppb NO2 + 4,000 ppb SO2 was administered 16 h/day for
25 8 months, and followed by 8 months of clean air exposure.
Lung Tumors in Animals with Spontaneously High Tumor Rates
26 The previous ISA and AQCDs described studies in animals with spontaneously high
27 tumor rates including strain A/J mice, AKR/cum mice, and CAFl/Jax mice. Strain A/J
28 mice exposed to 10,000 ppb NO2 for 6 h/day, 5 days/week for 6 months (Adkins et al..
29 1986) had a small, but statistically significant increase in pulmonary adenomas (increased
30 tumor multiplicity) with NO2 exposure (1,000 and 5,000 ppb NO2 had no effect). In
31 another study, increased survival rates of NO2-exposed animals were reported in a model
32 of spontaneous T cell lymphoma, i.e., AKR/cum mice that were exposed intermittently
January 2015 6-181 DRAFT: Do Not Cite or Quote
-------
1 (7 h/day, 5 days/week) to 250 ppb NO2 for up to 26 weeks (Richters and Damii. 1990).
2 Another study using CAP 1/Jax mice (Wagner etal.. 1965) showed that continuous
3 exposure to 5,000 ppb NO2 produced significant increases in the number of year-old
4 animals with pulmonary tumors when compared with control; this finding was no longer
5 significant at 14 or 16 months exposure.
Facilitation of Lung Cancer Metastases
6 The previous ISA and AQCDs summarized a group of experiments by one lab that
7 focused on the role of NO2 in metastases facilitation. Richters and Kuraitis (1981).
8 Richters and Kuraitis (1983). and Richters et al. (1985) exposed mice to multiple
9 concentrations and durations of NO2, and after exposure, the mice were injected
10 intravenously (I.V.) with the B16 melanoma cell line. Lung tumors were then counted,
11 with results of some of the experiments showing significantly increased numbers of
12 tumors.
Genotoxicity in Airway Cells
13 Ex vivo exposure of human nasal epithelial mucosa cells cultured at the air-liquid
14 interface to 10 ppb NO2 (Koehler etal.. 2013: Koehler etal.. 2010) or 100 ppb NO2
15 (Koehler et al.. 2011) produced increased deoxyribonucleic acid (DNA) fragmentation
16 measured with the single cell gel electrophoresis (COMET) assay as early as 30 minutes
17 after exposure and micronuclei formation after 3-hour exposure to 100 ppb NO2 (Koehler
18 etal.. 2011). Percentage of DNA content in the tail as detected with the COMET assay
19 decreased with increasing exposure duration [0.5, 1, 2, and 3-hour exposure; (Koehler
20 etal.. 2013)1. Of the in vivo assays reported in the previous ISA [see U.S. EPA (2008):
21 Annex Tables 4-11, 4-12, and 4-13, on pages 4-36 and 4-37 of the 2008 Annex], results
22 were mixed with positive findings of genotoxicity seen in two studies that employed rat
23 lung cells (mutations and chromosome abnormalities, 50,000-560,000 ppb
24 NO2 >12 days; 27,000 ppb NO2, 3 hours) and negative findings of genotoxicity seen in
25 tests employing Drosophila recessive lethals (500,000-7,000,000 ppb NO2, 1 hour),
26 Drosophila wing spot test (50,000-280,000 ppb NO2, 2 days), mouse bone marrow
27 micronuclei (20,000 ppb, 23 hours), and mouse spermatocyte and lymphocyte
28 chromosomal aberrations (100-10,000 ppb NO2, 6 hours). In vitro exposures to NO2
29 yielded positive findings in a majority of the tests in rodent (2,000-3,000 ppb NO2,
30 10 minutes) and human cell lines, bacteria (5,000-90,000 ppb NO2, 30 minutes), and
31 plants (5,000 ppb NO2, 24 hours).
January 2015 6-182 DRAFT: Do Not Cite or Quote
-------
Table 6-19 Animal toxicological studies of carcinogenicity and genotoxicity
exposure to nitrogen dioxide (NO2).
with
Reference
Koehler et al.
(2013)
Koehler et al.
(2010)
Koehler et al.
(2011)
Ohvama et al.
(1999)
Species
(Strain);
Concentration Age;
NO2 Sex; n
10 ppb Human
cells
100, 1,000, or Human
10, 000 ppb cells
100 ppb Human
cells
1,000, 5,000, or Rats
6,000 ppb (F344);
Adult M,
n=26
Exposure Conditions
Ex vivo cell culture at the air
liquid interface, primary
human nasal epithelia cells
from n = 10 donors, NO2
exposure for 0, 0.5, 1 , 2 and
3h.
Ex vivo cell culture at the air
liquid interface, primary
human nasal epithelia cells
from n = 10 donors, NO2
exposure for 0 or 0.5 h.
Ex vivo cell culture at the air
liquid interface, primary
human nasal epithelia cells
from n = 10 donors, NO2
exposure for 0, 0.5, 1, 2, and
3h.
Exposure to DEPcCBP and
NO2. IT installation of
DEPcCBP 1 x/week for
4 weeks. 6,000 ppb NO2 was
administered 16 h/day for
8 mo, and followed by 8 mo of
clean air exposure.
Endpoints Examined
Comet assay, micronucleus
formation, proliferation assay,
apoptosis, necrosis,
cytotoxicity.
Comet assay, micronucleus
formation, proliferation assay,
cytotoxicity.
Comet assay, micronucleus
formation, proliferation assay,
apoptosis, necrosis,
cytotoxicity.
Lung tumor incidence
(alveolar adenomas).
Adkins et al. 10,000 ppb
(1986)
Mice (A/J); Exposure of mice with
adult F, spontaneous high tumor rates
n=30. to NO2 for 6 h/day,
5 days/week for 6 mo.
Lung tumor multiplicity
(pulmonary adenomas).
Richters and Damii 250 ppb
(1990)
Mice
(AKR/cum),
adult
female,
n=50
Exposure of mice
intermittently (7 h/day,
5 days/week) to NO2 for up to
26 weeks.
Rodent survival rate.
Wagner et al. 1000, or
(1965) 5,000 ppb
Mice
(CAF1/Jax),
adult male,
n=20.
Continuous exposure to 1,000
or 5,000 ppb NO2.
Lung tumor multiplicity at 12,
14, and 16 mo.
January 2015
6-183
DRAFT: Do Not Cite or Quote
-------
Table 6-19 (Continued): Animal toxicological studies of carcinogenicity and
genotoxicity with exposure to nitrogen dioxide (NO2).
Reference
Richters and
Kuraitis(1981)
Species
(Strain);
Concentration Age;
NO2 Sex; n
400 or 800 ppb Mice (Swiss
Webster);
adult M; n =
24
Mice
(C57BL/6J);
Adult M; n
= 90
Exposure Conditions
NO2 exposure 8 h/day,
5 days/week for 1 0 weeks
(Swiss mice) or 12 weeks
(C57BL/6J mice); then all
animals were infused i.v. with
B16 melanoma cells that are
known to metastasize to the
lung. 3 weeks post-infusion,
animals were sacrificed and
lungs scored for tumor
incidence.
Endpoints Examined
Facilitation of lung tumor
metastasis (incidence of lung
tumors).
Richters and 300, 400, or Mice NO2 exposure 7 h/day,
Kuraitis(1983) 500 ppb (C57BL/6J); 5 days/week for 10 weeks.
adult M; n = Then all animals were infused
25, 51, 23 i.v. with B16 melanoma cells
that are known to metastasize
to the lung. 3 weeks
post-infusion, animals were
sacrificed and lungs scored
for tumor incidence.
Facilitation of lung tumor
metastasis (incidence of lung
tumors).
Richters et al.
(1985)
400 ppb
Mice 12 weeks of continuous
(C57BL/6J); exposure to NO2. Then all
adult M, animals were infused i.v. with
n=48. B16 melanoma cells. 3 weeks
post-infusion, animals were
sacrificed and lungs scored
for tumor incidence.
Facilitation of lung tumor
metastasis (incidence of lung
tumors).
Ichinose et al.
(1991)
40, 400, or
4,000 ppb
Rats
(Wistar);
adult M;
n=30.
Coexposure with carcinogen
BHPN and NO2. NO2
exposure for 17 mo.
Incidence of BHPN-induced
lung tumors (adenoma or
adenocarcinomas).
Ichinose and Saqai
(1992)
500 ppb NO2; Rats
50 ppb (Wistar);
NO2 + 400ppbO3 adultM:
n=36.
Carcinogen exposure plus air
pollutant mixture exposure
(O3 + NO2). 500 ppb NO2,
50 ppb NO2 + 400 ppb Os, for
13 mo, and then recovery with
clean air for another 11 mo;
continuous NO2 exposure,
11 h/day Os exposure.
Incidence of BHPN-induced
lung tumors (adenoma or
adenocarcinomas).
DEPcCBP = diesel exhaust particle extract-coated carbon black particles; COMET = single cell gel electrophoresis;
I.V. = intravenously; IT = intratrachael instillation; BHPN = N-bis(2-hydroxypropvl) nitrosamine; NO2 = nitrogen dioxide; O3 = ozone.
January 2015
6-184
DRAFT: Do Not Cite or Quote
-------
6.6.2 Leukemia Incidence and Mortality
1 A study of acute leukemia incidence identified cases from the French National Registry
2 of Childhood Blood Malignancies. Controls were randomly selected from the population
3 with a distribution of age and sex that matched that of the cases (Amigou et al.. 2011).
4 NC>2 concentrations over 6.5 ppb were positively associated with the odds of leukemia
5 incidence. This was also observed for specific types of leukemia. There was no difference
6 in results based on urban or rural residence. The authors stated that results were
7 strengthened when including only children who had been in the residences utilized in the
8 study for at least 2 years (data not included in the paper). In a case-control study in Italy,
9 no association was observed between NO2 concentrations and leukemia incidence
10 (Badaloni et al.. 2013). This was true in analyses limited to children aged 0-4 years and
11 children who never moved. NC>2 concentrations were highly correlated with PIVb 5 (0.78)
12 and inversely correlated with Os (-0.74). No association was reported between leukemia
13 incidence and these other pollutants. A study in Denmark reported no association
14 between NOx and leukemia incidence (Raaschou-Nielsen et al.. 201 la). One key
15 difference between this and the other studies described in this section is the age of the
16 study participants, with this study including adults, whereas the prior studies included
17 children.
18 A study performed in the U. S. examined the associations between NO, NO2, and NOx
19 concentrations during pregnancy with multiple cancers among children ages 0-5 years
20 old (Ghosh et al.. 2013). The exposure period examined the overall pregnancy, as well as
21 each trimester. Positive associations were observed between NO, NO2, and NOx and
22 acute lymphoblastic leukemia for the entire pregnancy and the first and second trimesters
23 but not the third trimester. Similar associations were not reported for acute myeloid
24 leukemia; results were null for all oxides of nitrogen during all trimesters.
25 A study in Taiwan matched children with a cause of death related to leukemia to children
26 with a cause of death unrelated to neoplasms or respiratory problems based on sex, year
27 of birth, and year of death (Weng et al.. 2008). NO2 concentrations were positively
28 associated with the odds of leukemia mortality.
29 These studies used multiple methods of exposure assessment, such as LUR (Badaloni
30 et-al.. 2013; Ghosh et-al.. 2013), dispersion models (Raaschou-Nielsen et al., 201 la), and
31 air monitors (Weng et al.. 2008). There were not enough studies within each type of
32 exposure metric to examine consistency across studies. Given the differences among the
33 study designs, it cannot be concluded that the inconsistencies are related to exposure
34 assessment method or length of follow-up periods. In general, studies using central site
35 monitors for exposure estimates carry uncertainty in long-term NO2 exposure studies
36 because the exposure error resulting from spatial misalignment between subjects' and
January 2015 6-185 DRAFT: Do Not Cite or Quote
-------
1 monitor locations can overestimate or underestimate associations with health effects
2 (Section 3.4.5.2). However, an association was observed with residential NO2 exposure
3 estimated for subjects' homes, and the improved spatial resolution of the exposure
4 estimate lends more confidence in the association.
6.6.3 Bladder Cancer Incidence and Mortality
5 A study in Denmark examined the association between NOx concentration and bladder
6 cancer incidence (Raaschou-Nielsen et al., 201 la). This longitudinal study calculated
7 incidence rate ratios for various types of cancer, and no association was demonstrated
8 between NOx concentration and bladder cancer incidence.
9 A study performed in Taiwan examined mortality records, comparing individuals
10 (matched on sex, year of birth, and year of death) with and without mortality due to
11 bladder cancer (Liuet al., 2009). Increased odds of bladder cancer mortality was
12 associated with increased NO2 concentrations. This trend was also observed for 862. The
13 highest tertile of PMio concentration was also associated with bladder cancer mortality,
14 but no association was observed for CO or Os concentrations. Liu et al. (2009) further
15 examined a three-level variable, with the lowest level being individuals in the lowest
16 tertile of 862 and NC>2 concentrations (<4.32 ppb and <20.99 ppb, respectively), the
17 highest level being individuals in the highest tertile of SC>2 and NC>2 concentrations
18 (>6.49 ppb and >27.33 ppb, respectively), and all others being categorized in the middle.
19 The resulting ORs, adjusted for urbanization of residential area and marital status, were
20 1.37 (95% CI: 1.03, 1.82) for the middle level and 1.98 (95% CI: 1.36, 2.88) for the
21 highest level. Although the point estimates for NC>2 and 862 combined are higher than
22 those observed for NO2 or SO2 alone [see Supplemental Table S6-10: (U.S. EPA. 2013jVI.
23 the 95% confidence intervals overlap. Therefore, the conclusion that NC>2 and SC>2
24 combined contribute to higher odds of mortality than either alone cannot be drawn.
6.6.4 Breast Cancer Incidence
25 A Canadian study of post-menopausal breast cancer incidence using a hospital-based
26 case-control study design estimated NC>2 concentrations at residential addresses using two
27 methods: (1) extrapolating data from fixed-site monitoring stations or (2) extrapolating
28 data from predicted concentrations determined with LUR using a dense network of air
29 samplers (Grouse et al.. 2010). Although point estimates were elevated for some of the
30 associations between NO2 concentrations and post-menopausal breast cancer incidence,
31 most of the associations were not statistically significant (see Table S6-10). In sensitivity
January 2015 6-186 DRAFT: Do Not Cite or Quote
-------
1 analyses limited to subjects who were residents of the same address for at least 10 years
2 prior to the study, the point estimates were slightly higher, but precision was reduced.
3 The results of this study may be biased due to the selection of controls, which were
4 hospital-based and limit the generalizability of the study. This study suggests a possible
5 association between post-menopausal breast cancer incidence and NO2 concentration. A
6 study in Denmark reported no association between NOx and breast cancer incidence
7 (Raaschou-Nielsen et al. 201 la). This study, unlike the one performed in Canada,
8 included all breast cancer cases instead of limiting to post-menopausal cases.
9 An ecologic study was performed by Wei et al. (2012) using data from the Surveillance,
10 Epidemiology, and End Results program to determine the breast cancer incidence rate of
11 various states and metropolitan areas and data from the EPA's Geographic Area AirData
12 to determine NOx emissions. Results of Pearson's correlations demonstrated a
13 relationship between NOx emissions and breast cancer incidence. The state with the
14 highest NOx emissions also had the highest breast cancer incidence rate, and the state
15 with the lowest emissions had the lowest breast cancer incidence rate. However, this
16 study is limited by its ecologic nature and the lack of individual-level data. There is no
17 control for potential confounders or examination of factors other than air pollutants (of
18 which CO, SO2, and VOCs, but not PMio, also had positive correlations) that could be
19 associated with breast cancer incidence rates.
6.6.5 Prostate Cancer Incidence
20 Men enrolled in the Prostate Cancer and Environment Study were included in an
21 investigation of NO2 concentration and prostate cancer incidence (Parent et al.. 2013).
22 Cases were men diagnosed with prostate cancer and recruited through pathology
23 departments. Population-based controls were identified through electoral lists and
24 frequency matched by 5-year age groups. A positive association was observed between
25 recent NO2 concentration and odds of prostate cancer. The association was also observed
26 using back-extrapolated estimates of NO2 10 years prior. Multiple sensitivity analyses
27 were performed, including back-extrapolation of NO2 estimates for 20 years, addition of
28 smoking and alcohol consumption as confounders, exclusion of proxy subjects, exclusion
29 of subjects without a prostate cancer screening in the past 5 years, exclusion of subjects at
30 their residence for less than 10 years, and comparisons of subjects with geo-coding to
31 their exact address or to a centroid of their postal code. The results, while not always
32 statistically significant (in some parts due to decreases in sample size and precision),
33 were similar to the overall results reported. However, a study in Denmark examined the
34 association between NOx concentration and prostate cancer incidence using the Diet,
January 2015 6-187 DRAFT: Do Not Cite or Quote
-------
1 Cancer and Health cohort study (Raaschou-Nielsen et al., 201 la). No association was
2 reported between prostate cancer and NOx concentrations in the study.
6.6.6 Other Cancers Incidence and Mortality
3 A study in Denmark utilized participants in the Diet, Cancer and Health cohort and
4 examined the relationship between NOx concentrations and multiple cancers, including
5 many not examined in other studies (Raaschou-Nielsen et al.. 201 la). A positive
6 association was demonstrated between NOx concentrations and cervical and brain cancer
7 incidences. No association was observed between NOx concentrations and buccal
8 cavity/pharynx cancer, esophageal cancer, stomach cancer, colon cancer, rectal cancer,
9 liver cancer, pancreatic cancer, laryngeal cancer, uterine cancer, ovarian cancer, kidney
10 cancer, melanoma, non-Hodgkin lymphoma, or myeloma.
11 A study performed in the United States examined maternal exposure to air pollution
12 (i.e., NO, NO2, and NOx concentrations during pregnancy) and multiple cancers in
13 children at age 0-5 years (Ghosh etal. 2013). Elevated point estimates were observed
14 for NO and NOx, but not NO2, in the third trimester and incidence of bilateral
15 retinoblastoma. Null associations were found for all other cancers examined with NOx
16 (non-Hodgkin lymphoma, central nervous system tumors, ependymoma, astrocytoma,
17 intracranial and intraspinal embryonal tumors, primitive neuroectodermal tumor, other
18 gliomas, neuroblastoma, retinoblastoma, unilateral retinoblastoma, Wilms tumor,
19 hepatoblastoma, rhabdomyosarcomas, germ cell tumors, extracranial and extragonadal
20 germ cell tumors, and teratoma).
21 Finally, a study in Canada reported the association between daily NO2 concentrations and
22 cancer mortality [any cancer type; (Goldberg et al., 2013)1. A positive association was
23 observed between cancer mortality and NO2 as well as for some other pollutants (warm
24 season Os and 802). No association was reported for CO or PM2 5. This study is limited
25 by the use of daily concentrations and the grouping of all types of cancer together.
6.6.7 Production of N-Nitroso Compounds and other Nitro Derivatives
26 Daily chemical transformations involving ultraviolet, NO2 and hydrocarbons, products of
27 automobile exhaust, and oxygen/ozone can generate peroxyacetyl nitrate (PAN) in the
28 gas fraction as part of photochemical smog. Mutagenicity assays demonstrated that PAN
29 is weakly mutagenic in the lungs of the highly susceptible big Blue (R) mice and in
30 Salmonella and that PAN produces a unique signature mutation (DeMarini et al., 2000).
January 2015 6-188 DRAFT: Do Not Cite or Quote
-------
1 N-nitroso compounds can be generated endogenously in the human body from NO2 via
2 processes that generate nitrite (NO2 ) or nitrate. Further, NO2 is known to react with
3 amines to produce nitrosamines, known animal carcinogens. The possibility that NO2
4 could produce cancer via nitrosamine formation has been investigated and was reported
5 in the 2008 ISA for Oxides of Nitrogen U.S. EPA (2008).
6.6.8 Genotoxicity
6.6.8.1 lexicological Studies
6 A number of animal toxicology studies have examined the genotoxicity of NO2. Ex vivo
7 exposure of human nasal epithelial mucosa cells cultured at the air-liquid interface to
8 10 ppb NO2 (Koehleretal.. 2013; Koehler etal.. 2010) or 100 ppb NO2 (Koehler et al.
9 2011) produced increased DNA fragmentation measured with the COMET assay as early
10 as 30 minutes after exposure and micronuclei formation after 3-hours of exposure to
11 100 ppb NO2 (Koehleretal.. 2011). Percentage of DNA content in the tail as detected
12 with the COMET assay decreased with increasing exposure duration [0.5, 1, 2, and
13 3-hour exposure; (Koehler et al.. 2013)]. Of the in vivo assays reported in the previous
14 ISA [see U.S. EPA (2008). Annex Tables 4-11, 4-12, and 4-13, on pages 4-36 and 4-37
15 of the 2008 Annex], results were mixed with positive findings of genotoxicity seen in two
16 studies that employed rat lung cells (mutations and chromosome abnormalities,
17 50,000-560,000 ppb NO2 >12 days; 27,000 ppb NO2, 3 hours) and negative findings of
18 genotoxicity seen in tests employing Drosophila recessive lethals
19 (500,000-7,000,000 ppb NO2, 1 hour), Drosophila wing spot test (50,000-280,000 ppb
20 NO2, 2 days), mouse bone marrow micronuclei (20,000 ppb, 23 hours), and mouse
21 spermatocyte and lymphocyte chromosomal aberrations (100-10,000 ppb NO2, 6 hours).
22 In vitro exposures to NO2 yielded positive findings in a majority of the tests in rodent
23 (2,000-3,000 ppb NO2, 10 minutes) and human cell lines, bacteria (5,000-90,000 ppb
24 NO2, 30 minutes), and plants (5,000 ppb NO2, 24 hours) (Table 6-19).
25 NO2-induced genotoxicity in various organs of male rats was demonstrated after
26 inhalation exposure to 2,660, 5,320, or 10,640 ppb NO2 for 6 h/day for 7 days (Han et al..
27 2013). In the COMET assay, NO2 inhalation exposure generated significant increases in
28 DNA damage in all dose groups in the liver, lung, and kidney; brain and spleen had
29 significantly increased DNA damage at the two highest doses. Bone marrow PCE
30 micronuclei testing, a marker of chromosomal damage, revealed significant increases in
31 Mn formation with NO2 inhalation across all dose groups. DNA-protein cross-links,
32 another form of DNA damage, was significantly increased in all dose groups in the brain
January 2015 6-189 DRAFT: Do Not Cite or Quote
-------
1 and liver; in the two highest dose groups in the heart and spleen; and in the highest dose
2 group in the lung and kidney (Han et al.. 2013).
6.6.8.2 Epidemiologic Studies
3 A study in Italy of children living near chipboard industries examined the relationship
4 between NO2 concentrations and genotoxicity (Marcon et al., 2014). NO2 concentrations
5 were not associated with results of Comet Assays (i.e., tail length, moment, or intensity)
6 but were associated with some results of micronucleus assays. A 10-ppb increase in NO2
7 concentration was associated with a 1.16% (95% CI: 0.6%, 1.7%) change in binucleated
8 cells. A 10-ppb increase in NO2 concentration was also associated with an increased risk
9 of nuclear buds [risk ratio (RR): 3.72 [95% CI: 1.67, 7.73). The other pollutant examined,
10 formaldehyde, was also associated with nuclear buds but not with binucleated cells.
11 Additionally, formaldehyde was associated with Comet tail intensity and moment.
12 Although this study is limited by potential for bias by confounding, this study supports
13 the findings that NO2 concentrations may be associated with cancer, as some measures of
14 genotoxicity were associated with NO2 concentrations and genotoxicity may lead to
15 cancer.
6.6.9 Summary and Causal Determination
16 The overall evidence for long-term NC>2 exposure and cancer is suggestive, but not
17 sufficient, to infer a causal relationship. This conclusion is based on evidence from some
18 prospective epidemiologic studies reporting associations between NO2 exposure and
19 cancer incidence and mortality, with the strongest evidence coming from studies of lung
20 cancer incidence and mortality. Animal toxicology studies employing NC>2 exposure with
21 other known carcinogens provide further supporting evidence, showing that inhaled NC>2
22 can increase tumor load in laboratory rodents. Nonetheless, toxicological data provide no
23 clear evidence of NC>2 acting as a complete carcinogen, and not all epidemiologic studies
24 report positive associations.
25 In past reviews, a limited number of epidemiologic studies had assessed the relationship
26 between long-term NC>2 or NOx exposure and cancer incidence and mortality. The 2008
27 ISA for Oxides of Nitrogen concluded that the evidence was "inadequate to infer the
28 presence or absence of a causal relationship" (U.S. EPA. 2008). Recent studies include
29 evidence on lung cancer as well as new types of cancer, evaluating both incidence and
30 mortality. All available evidence for cancer due to long-term NO2 or NOx exposure was
31 evaluated using the framework described in Table II of the Preamble. The key evidence
January 2015 6-190 DRAFT: Do Not Cite or Quote
-------
1 as it relates to the causal determination is summarized in Table 6-20 and demonstrates the
2 addition of recent studies, some of which support an association but are not consistent
3 throughout the body of literature, as well as limited toxicological evidence providing
4 coherence.
5 Epidemiologic studies of NC>2 or NOx and lung cancer incidence have had mixed results,
6 with some studies reporting no associations while other studies report positive
7 associations. Most of these studies included large sample sizes, similar NOx or NCh
8 concentrations, and control for many potential confounders, including smoking
9 exposures, although many studies lacked investigation into potential copollutant
10 confounding. Most studies of NC>2 or NOx and lung cancer mortality reported no
11 association, but some studies reported positive associations. Evidence for cancer in other
12 organ systems is accumulating, but is more difficult to interpret due to questions
13 regarding biological plausibility. Recent studies of leukemia have reported associations
14 with NO2 concentration. Similarly, a study of bladder cancer mortality reported an
15 association with NO2. Breast cancer incidence was positively correlated with NOx
16 concentration in an ecologic analysis, but a study of post-menopausal women observed
17 no increase in odds with higher NO2 concentrations. A positive association was observed
18 between NO2 concentration and prostate cancer incidence. Overall, the epidemiologic
19 studies use multiple methods, such as nearest air monitor and dispersion models, to
20 estimate NOx concentrations. No patterns or trends are apparent in the results based on
21 the type of exposure assessment method. Toxicological data provide no clear evidence of
22 NO2 acting as a complete carcinogen, and agencies that classify carcinogens including the
23 Department of Health and Human Services, the International Agency for Research on
24 Cancer, and the EPA have not classified oxides of nitrogen for potential carcinogenicity.
25 The American Conference of Industrial Hygienists has classified NO2 as A4 (not
26 classifiable for humans or animals). However, in some animal toxicological models, NO2
27 may act as a tumor promoter at the site of contact, possibly due to its ability to produce
28 cellular damage, induce respiratory epithelial hyperplasia (Section 6.2.6). or promote
29 regenerative cell proliferation. Genotoxic and mutagenic studies with NO2 have mixed
30 results. Some studies with coexposure to other known carcinogens demonstrated that
31 inhaled NO2 can increase tumor burden in rodents. Collectively, while some studies
32 observed no associations, the evidence from several toxicological and epidemiologic
33 studies is suggestive, but not sufficient, to infer a causal relationship between long-term
34 exposure to NO2 and cancer incidence and mortality.
January 2015 6-191 DRAFT: Do Not Cite or Quote
-------
Table 6-20 Summary of evidence, which is suggestive, but not sufficient, to infer
a causal relationship between long-term nitrogen dioxide (NO2)
exposure and cancer.
Rationale for Causal
Determination3
Key Evidence13
Key References13
NO2 or
Concentrations
Associated with
Effects0
Evidence from
epidemiologic studies
generally supportive
but not consistent
Positive associations were
observed between overall lung
cancer incidence and mortality in
multiple studies conducted in the
U.S., Canada, Europe and Asia.
Nafstad et al. (2003).
Nvberq et al. (2000).
Raaschou-Nielsen et al.
(201 Oa),
Raaschou-Nielsen et al.
(2011 b),
Cesaroni et al. (2013).
Filleuletal. (2005),
Carey et al. (2013).
Jerrettetal. (2013),
Hvstadetal. (2013)
Means varied with some
studies including areas
estimating
concentrations of NO2 or
NOx as low as 1.2 ppb to
studies with areas
estimated at 33.7 ppb.
No associations were observed
between overall lung cancer
incidence and mortality in multiple
other studies conducted in the
United States, Europe, and Asia.
Brunekreef et al. (2009),
Beelen et al. (2008a),
Papathomas et al. (2011).
Beelen et al. (2008b),
Caoetal. (2011).
Hartetal. (2011),
Heinrichetal. (2013),
Krewski et al. (2009),
Raaschou-Nielsen et al.
(2013a)
Means varied with
estimated concentrations
of NO2 NOx ranging from
2.8 to 34.1 ppb.
Positive associations were also
observed in some studies of NO2
concentrations and leukemia,
bladder cancer, and prostate
cancer.
Amiqou etal. (2011).
Wenq et al. (2008),
Liu et al. (2009),
Parent etal. (2013),
Ghosh etal. (2013),
Raaschou-Nielsen et al.
(2011 a)
Associations observed at
levels as low as
6.5-8.6 ppb for
leukemia.
January 2015
6-192
DRAFT: Do Not Cite or Quote
-------
Table 6-20 (Continued): Summary of evidence, which is suggestive, but not
sufficient, to infer a causal relationship between long-
term nitrogen dioxide (NO2) exposure and cancer.
Rationale for Causal
Determination3
Limited toxicological
evidence provides
coherence
Limited supporting
evidence of
carcinogenesis,
mutagenesis, or
genotoxicity provides
biological plausibility
Key Evidence13
Studies of facilitation of
metastasis and coexposures with
known carcinogens show NO2
related effects. Studies of NO2 as
a direct carcinogen are lacking.
Finding of mutagenicity and
micronucleus formation in ex vivo
culture of primary human nasal
epithelial cells exposed to NO2.
Key References'3
Adkinsetal. (1986),
Richters and Damii (1990).
Wagner et al. (1965).
Richters and Kuraitis (1981),
Richters and Kuraitis (1983).
Richters et al. (1985),
Ichinose et al. (1991),
Ichinose and Saqai (1992)
Sections 6. 6.1 and 6.6.8
Koehleretal. (2013),
Koehleretal. (2011),
Koehleretal. (2010)
Section 6.6.7
NO2 or NOx
Concentrations
Associated with
Effects0
10,000 ppb
250 ppb
5,000 ppb
400, 800 ppb
300, 400, 500 ppb
400 ppb
4,000 ppb
500 ppb
100, 1,000, 10,000 ppb
Mixed findings of mutagenicity
and carcinogenicity in various
models of NO2 exposure in older
studies, mainly in nonhuman
species.
U.S. EPA (2008). Annex
Table AX4-11, Table AX 4-12,
and Table AX 4-13
NO2 = nitrogen dioxide; NOX = sum of NO and NO2.
aBased on aspects considered in judgments of causality and weight of evidence in causal framework in Tables I and N. of the
Preamble.
Describes the key evidence and references contributing most heavily to causal determination and, where applicable, to
uncertainties or inconsistencies. References to earlier sections indicate where the full body of evidence is described.
°Describes the NO2 concentrations with which the evidence is substantiated.
January 2015
6-193
DRAFT: Do Not Cite or Quote
-------
References for Chapter 6
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. http://dx.doi.Org/10.1164/ajrccm.159.2.9806020
Ackermann-Liebrich. U: Leuenberger. P: Schwartz. J: Schindler. C: Monn. C: Bologninj G: Bongard. JP:
Brandli O: Domenighetti G: Elsasser. S: Grize. L: Karrer. W: Keller. R: Keller-Wossidlo. H: Kiinzli.
N: Martin. BW: Medici. TC: Perruchoud. AP: Schoni. MH: Tschopp. JM: Villiger. B: Wuthrich. B:
Zellweger. JP: Zemp. E. (1997). Lung function and long term exposure to air pollutants in
Switzerland. Am J Respir Crit Care Med 155: 122-129.
http://dx.doi.0rg/10.1164/ajrccm.155.l.9001300
Adkins. B. Jr: Van Stee. EW: Simmons. JE: Eustis. SL. (1986). Oncogenic response of strain A/J mice to
inhaled chemicals. J Toxicol Environ Health 17: 311-322.
http://dx.doi.org/10.1080/15287398609530825
Aguilera. I: Garcia-Esteban. R: Iniguez. C: Nieuwenhuijsen. MJ: Rodriguez. A: Paez. M: Ballester. F:
Sunyer. J. (2010). Prenatal exposure to traffic-related air pollution and ultrasound measures of fetal
growth in the INMA Sabadell cohort. Environ Health Perspect 118: 705-711.
http://dx.doi.org/10.1289/ehp.0901228
Aguilera. I: Guxens. M: Garcia-Esteban. R: Corbella. T: Nieuwenhuijsen. MJ: Foradada. CM: Sunyer. J.
(2009). Association between GIS-based exposure to urban air pollution during pregnancy and birth
weight in the INMA Sabadell cohort. Environ Health Perspect 117: 1322-1327.
http://dx.doi.org/10.1289/ehp.0800256
Aguilera. I: Pedersen. M: Garcia-Esteban. R: Ballester. F: Basterrechea. M: Esplugues. A: Fernandez-
Somoano. A: Lertxundi A: Tardon. A: Sunyer. J. (2013). Early-life exposure to outdoor air pollution
and respiratory health, ear infections, and eczema in infants from the INMA study. Environ Health
Perspect 121: 387-392. http://dx.doi.org/10.1289/ehp.1205281
Amigou. A: Sermage-Faure. C: Orsi. L: Leverger. G: Baruchel A: Bertrand. Y: Nelken. B: Robert. A:
Michel G: Margueritte. G: Perel Y: Mechinaud. F: Bordigoni. P: Hemon. D: Clavel J. (2011). Road
traffic and childhood leukemia: The ESCALE Study (SFCE). Environ Health Perspect 119: 566-572.
http://dx.doi.org/10.1289/ehp.1002429
Amster. ED: Haim. M: Dubnov. J: Brodav. DM. (2014). Contribution of nitrogen oxide and sulfur dioxide
exposure from power plant emissions on respiratory symptom and disease prevalence. Environ Pollut
186: 20-28. http://dx.doi.Org/10.1016/j.envpol.2013.10.032
Andersen. ZJ: B0nnelykke. K: Hvidberg. M: Jensen. SS: Ketzel. M: Loft. S: S0rensen. M: Tj0nneland. A:
Overvad. K: Raaschou-Nielsen. O. (2012a). Long-term exposure to air pollution and asthma
hospitalisations in older adults: A cohort study. Thorax 67: 6-11. http://dx.doi.org/10.1136/thoraxjnl-
2011-200711
Andersen. ZJ: Hvidberg. M: Jensen. SS: Ketzel M: Loft. S: S0rensen. M: Tj0nneland. A: Overvad. K:
Raaschou-Nielsen. O. (2011). Chronic obstructive pulmonary disease and long-term exposure to
traffic-related air pollution: Acohort study. Am J Respir Crit Care Med 183: 455-461.
http://dx.doi.org/10.1164/rccm.201006-0937OC
Andersen. ZJ: Kristiansen. LC: Andersen. KK: Olsen. TS: Hvidberg. M: Jensen. SS: Ketzel. M: Loft. S:
S0rensen. M: Tj0nneland. A: Overvad. K: Raaschou-Nielsen. O. (2012b). Stroke and long-term
exposure to outdoor air pollution from nitrogen dioxide: a cohort study. Stroke 43: 320-325.
http://dx.doi.org/10.1161/STROKEAHA.lll.629246
January 2015 6-194 DRAFT: Do Not Cite or Quote
-------
Andersen. ZJ: Raaschou-Nielsen. O: Ketzel. M: Jensen. SS: Hvidberg. M: Loft. S: Tj0nneland. A:
Overvad. K: S0rensen. M. (2012c). 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
Anderson. HR: Favarato. G: Atkinson. RW. (2013). Long-term exposure to air pollution and the incidence
of asthma: Meta-analysis of cohort studies. Air Qual Atmos Health 6: 47-56.
http://dx.doi.org/10.1007/sll869-011-0144-5
Annesi-Maesano. I: Hulin. M: Lavaud. F: Raherison. C: Kopferschmitt C: de Slav. F: Charpin. DA:
Denis. C. (2012). Poor air quality in classrooms related to asthma and rhinitis in primary
schoolchildren of the French 6 Cities Study. Thorax 67: 682-688. http://dx.doi.org/10.1136/thoraxjnl-
2011-200391
Annesi-Maesano. I: Moreau. D: Caillaud. D: Lavaud. F: Le Moullec. Y: Tavtard. A: Pauli G: Charpin. D.
(2007). Residential proximity fine particles related to allergic sensitisation and asthma in primary
school children. Respir Med 101: 1721-1729. http://dx.doi.Org/10.1016/j.rmed.2007.02.022
Aranyi. C: Fenters. J: Erhlich. R: Gardner. D. (1976). Scanning electron microscopy of alveolar
macrophages after exposure to oxygen, nitrogen dioxide, and ozone [Abstract]. Environ Health
Perspect 16: 180.
Arner. EC: Rhoades. RA. (1973). Long-term nitrogen dioxide exposure: Effects on lung lipids and
mechanical properties. Arch Environ Health 26: 156-160.
http://dx.doi.org/10.1080/00039896.1973.10666244
Atkinson. RW: Carey. IM: Kent. AJ: van Staa. TP: Anderson. HR: Cook. DG. (2013). Long-term
exposure to outdoor air pollution and incidence of cardiovascular diseases. Epidemiology 24: 44-53.
http://dx.doi.org/10.1097/EDE.Ob013e318276ccb8
Ayaz. KL: Csallany. AS. (1978). Long-term NO2 exposure of mice in the presence and absence of
vitamin E. II. Effect of glutathione peroxidase. Arch Environ Occup Health 33: 292-296.
Badaloni. C: Ranucci. A: Cesaroni. G: Zanini. G: Vienneau. D: Al-Aidrous. F: De Hoogh. K: Magnani. C:
Forastiere. F: Group. SS. (2013). Air pollution and childhood leukaemia: a nationwide case-control
study in Italy. Occup Environ Med 70: 876-883. http://dx.doi.org/10.1136/oemed-2013-101604
Ballester. F: Estarlich. M: Iniguez. C: Llop. S: Ramon. R: Esplugues. A: Lacasana. M: Rebagliato. M.
(2010). Air pollution exposure during pregnancy and reduced birth size: Aprospective birth cohort
study in Valencia, Spain. Environ Health 9: 6. http://dx.doi.org/10.1186/1476-069X-9-6
Bastain. T: Islam. T: Berhane. K: McConnell R: Rappaport. E: Salam. M: Linn. W: Avol E: Zhang. Y:
Gilliland. F. (2011). Exhaled nitric oxide, susceptibility and new-onset asthma in the Children's Health
Study. Eur Respir J 37: 523-531. http://dx.doi.org/10.1183/09031936.00021210
Basu. R: Malig. B: Ostro. B. (2010). High ambient temperature and the risk of preterm delivery. Am J
Epidemiol 172: 1108-1117. http://dx.doi.org/10.1093/aie/kwal70
Becerra. TA: Wilhelm. M: Olsen. J: Cockburn. M: Ritz. B. (2013). Ambient air pollution and autism in
Los Angeles county, California. Environ Health Perspect 121: 380-386.
http://dx.doi.org/10.1289/ehp.1205827
Beckerman. BS: Jerrett M: Finkelstein. M: Kanaroglou. P: Brook. JR: Arain. MA: Sears. MR: Stieb. D:
Balmes. J: Chapman. K. (2012). The association between chronic exposure to traffic-related air
pollution and ischemic heart disease. J Toxicol Environ Health A 75: 402-411.
http://dx.doi.org/10.1080/15287394.2012.670899
Beelen. R: Hoek. G: van den Brandt. PA: Goldbohm. RA: Fischer. P: Schouten. LJ: Armstrong. B:
Brunekreef. B. (2008a). Long-term exposure to traffic-related air pollution and lung cancer risk.
Epidemiology 19: 702-710. http://dx.doi.org/10.1097/EDE.Ob013e318181b3ca
January 2015 6-195 DRAFT: Do Not Cite or Quote
-------
Beelen. R: Hoek. G: van den Brandt. PA: Goldbohm. RA: Fischer. P: Schouten. LJ: Jerrett. M: Hughes. E:
Armstrong. B: Brunekreef. B. (2008b). Long-term effects of traffic-related air pollution on mortality
in a Dutch cohort (NLCS-AIR study). Environ Health Perspect 116: 196-202.
http://dx.doi.org/10.1289/ehp.10767
Beelen. R: Hoek. G: Vienneau. D: Eeftens. M: Dimakopoulou. K: Pedeli. X: Tsai. MY. i: Kuenzli. N:
Schikowski. T: Marcon. A: Eriksen. KT: Raaschou-Nielsen. O. le: Stephanou. E: Patelarou. E: Lanki.
T: Yli-Toumi. T: Declercq. C: Falq. G: Stempfelet. M: Birk. M: Cyrys. J: von Klot. S: Nador. G:
Varro. MJ: Dedele. A: Grazuleviciene. R: Moelter. A: Lindlev. S: Madsen. C: Cesaroni. G: Ranzi. A:
Badaloni. C: Hoffmann. B: Nonnemacher. M: Kraemer. U: Kuhlbusch. T: Cirach. M: de Nazelle. A:
Nieuwenhuijsen. M: Bellander. T. om: Korek. M: Olsson. D: Stromgren. M: Dons. E. vi: Jerrett. M:
Fischer. P: Wang. M: Brunekreef. B: de Hoogh. K. (2013). Development of NO2 and NOx land use
regression models for estimating air pollution exposure in 36 study areas in Europe - The ESCAPE
project. Atmos Environ 72: 10-23. http://dx.doi.Org/10.1016/j.atmosenv.2013.02.037
Beelen. R: Raaschou-Nielsen. O: Stafoggia. M: Andersen. ZJ: Weinmayr. G: Hoffmann. B: Wolf. K:
Samoli. E: Fischer. P: Nieuwenhuijsen. M: Vineis. P: Xun. WW: Katsouvanni. K: Dimakopoulou. K:
Oudin. A: Forsberg. B: Modig. L: Havulinna. AS: Lanki. T: Turunen. A: Oftedal. B: Nystad. W:
Nafstad. P: De Faire. U: Pedersen. NL: Ostenson. CG: Fratiglioni. L: Penell. J: Korek. M: Pershagen.
G: Eriksen. KT: Overvad. K: Ellermann. T: Eeftens. M: Peeters. PH: Meliefste. K: Wang. M: Bueno-
De-Mesquita. B: Sugiri. D: Kramer. U: Heinrich. J: de Hoogh. K: Key. T: Peters. A: Hampel R:
Concin. H: Nagel. G: Ineichen. A: Schaffner. E: Probst-Hensch. N: Kiinzli. N: Schindler. C:
Schikowski. T: Adam. M: Phuleria. H: Vilier. A: Clavel-Chapelon. F: Declercq. C: Grioni. S: Krogh.
V: Tsai. MY: Riccerj F: Sacerdote. C: Galassi C: Migliore. E: Ranzi. A: Cesaroni. G: Badaloni. C:
Forastiere. F: Tamavo. I: Amiano. P: Dorronsoro. M: Katsoulis. M: Trichopoulou. A: Brunekreef. B:
Hoek. G. (2014a). Effects of long-term exposure to air pollution on natural-cause mortality: an
analysis of 22 European cohorts within the multicentre ESCAPE project. Lancet 383: 785-795.
http://dx.doi.org/10.1016/S0140-6736(13)62158-3
Beelen. R: Stafoggia. M: Raaschou-Nielsen. O: Andersen. ZJ: Xun. WW: Katsouvannj K:
Dimakopoulou. K: Brunekreef. B: Weinmayr. G: Hoffmann. B: Wolf. K: Samoli. E: Houthuijs. D:
Nieuwenhuijsen. M: Oudin. A: Forsberg. B: Olsson. D: Salomaa. V: Lanki. T: Yli-Tuomi. T: Oftedal.
B: Aamodt G: Nafstad. P: De Faire. U: Pedersen. NL: Ostenson. CG: Fratiglioni. L: Penell. J: Korek.
M: Pyko. A: Eriksen. KT: Tj0nneland. A: Becker. T: Eeftens. M: Bots. M: Meliefste. K: Wang. M:
Bueno-De-Mesquita. B: Sugiri. D: Kramer. U: Heinrich. J: de Hoogh. K: Key. T: Peters. A: Cyrys. J:
Concin. H: Nagel. G: Ineichen. A: Schaffner. E: Probst-Hensch. N: Dratva. J: Ducret-Stich. R: Vilier.
A: Clavel-Chapelon. F: Stempfelet. M: Grioni. S: Krogh. V: Tsai. MY: Marcon. A: Ricceri. F:
Sacerdote. C: Galassi. C: Migliore. E: Ranzi. A: Cesaroni. G: Badaloni. C: Forastiere. F: Tamavo. I:
Amiano. P: Dorronsoro. M: Katsoulis. M: Trichopoulou. A: Vineis. P: Hoek. G. (2014b). Long-term
exposure to air pollution and cardiovascular mortality: An analysis of 22 European cohorts.
Epidemiology 25: 368-378. http://dx.doi.org/10.1097/EDE.0000000000000076
Belanger. K: Holford. TR: Gent. JF: Hill. ME: Kezik. JM: Leaderer. BP (2013). Household levels of
nitrogen dioxide and pediatric asthma severity. Epidemiology 24: 320-330.
http://dx.doi.org/10.1097/EDE.Ob013e318280e2ac
Bell. ML: Ebisu. K: Belanger. K. (2007). Ambient air pollution and low birth weight in Connecticut and
Massachusetts. Environ Health Perspect 115: 1118-1124. http://dx.doi.org/10.1289/ehp.9759
Benson. PE. (1984). CALINE4 - a dispersion model for predicting air pollutant concentrations near
roadways. (FHWA/CA/TL-84/15). Sacramento, California: California Department of Transportation,
Office of Transportation Laboratory.
https://ntrl.ntis.gov/NTRL/dashboard/searchResults.xhtml?searchQuery=PB85211498
Bentaveb. M: Helmer. C: Raherison. C: Dartigues. JF: Tessier. JF: Annesi-Maesano. I. (2010). Bronchitis-
like symptoms and proximity air pollution in French elderly. Respir Med 104: 880-888.
http://dx.doi.0rg/10.1016/i.rmed.2010.01.004
January 2015 6-196 DRAFT: Do Not Cite or Quote
-------
Berhane. K: Zhang. Y: Salam. MT: Eckel SP: Linn. WS: Rappaport EB: Bastain. TM: Lurmann. F:
Gilliland. FD. (2014). Longitudinal effects of air pollution on exhaled nitric oxide: the Children's
Health Study. Occup Environ Med 71: 507-513. http://dx.doi.org/10.1136/oemed-2013-101874
Bilenko. N: Rossem. LV: Brunekreef. B: Beelen. R: Eeftens. M: Hoek. G: Houthuijs. D: de Jongste. JC:
Kempen. EV: Koppelman. GH: Meliefste. K: Oldenwening. M: Smit. HA: Wijga. AH: Gehring. U.
(2013). Traffic-related air pollution and noise and children's blood pressure: results from the PIAMA
birth cohort study. European Journal of Preventive Cardiology.
http://dx.doi.org/10.1177/2047487313505821
Blair. WH: Henry. MC: Ehrlich. R. (1969). Chronic toxicity of nitrogen dioxide: II. Effect on
histopathology of lung tissue. Arch Environ Health 18: 186-192.
http://dx.doi.org/10.1080/00039896.1969.10665391
Blomberg. A: Krishna. MT: Helledav. R: Soderberg. M: Ledin. MC: Kelly. FJ: Frew. AJ: Holgate. ST:
Sandstrom. T. (1999). Persistent airway inflammation but accomodated antioxidant and lung function
responses after repeated daily exposure to nitrogen dioxide. Am J Respir Crit Care Med 159: 536-543.
http://dx.doi.0rg/10.1164/airccm.159.2.9711068
Bobak. M. (2000). Outdoor air pollution, low birth weight, and prematurity. Environ Health Perspect 108:
173-176.
Boggia. B: Carbone. U: Farinaro. E: Zarrilli. S: Lombardi. G: Colao. A: De Rosa. N: De Rosa. M. (2009).
Effects of working posture and exposure to traffic pollutants on sperm quality. J Endocrinol Invest 32:
430-434.
Boogaard. H: Fischer. PH: Janssen. NA: Kos. GP: Weijers. EP: Cassee. FR: van der Zee. SC: de Hartog.
JJ: Meliefste. K: Wang. M: Brunekreef. B: Hoek. G. (2013). Respiratory effects of a reduction in
outdoor air pollution concentrations. Epidemiology 24: 753-761.
http://dx.doi.org/10.1097/EDE.Ob013e31829el639
Braback. L: Forsberg. B. (2009). Does traffic exhaust contribute to the development of asthma and
allergic sensitization in children: findings from recent cohort studies [Review]. Environ Health 8: 17.
http://dx.doi.org/10.1186/1476-069X-8-17
Brauer. M: Hoek. G: Smit. HA: de Jongste. JC: Gerritsen. J: Postma. PS: Kerkhof. M: Brunekreef. B.
(2007). Air pollution and development of asthma, allergy and infections in a birth cohort. Eur Respir J
29: 879-888. http://dx.doi.org/10.1183/09031936.00083406
Brauer. M: Hoek. G: van Vliet P: Meliefste. K: Fischer. P: Gehring. U: Heinrich. J: Cyrys. J: Bellander.
T: Lewne. M: Brunekreef. B. (2003). Estimating long-term average particulate air pollution
concentrations: application of traffic indicators and geographic information systems. Epidemiology
14: 228-239.
Brauer. M: Hoek. G: Van Vliet. P: Meliefste. K: Fischer. PH: Wijga. A: Koopman. LP: Neijens. HJ:
Gerritsen. J: Kerkhof. M: Heinrich. J: Bellander. T: Brunekreef. B. (2002). Air pollution from traffic
and the development of respiratory infections and asthmatic and allergic symptoms in children. Am J
Respir Crit Care Med 166: 1092-1098. http://dx.doi.org/10.1164/rccm.200108-007OC
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
Breton. CV: Wang. X: Mack. WJ: Berhane. K: Lopez. M: Islam. TS: Feng. M: Lurmann. F: Mcconnell. R:
Hodis. HN: Kiinzli. N: Avol E. (2012). Childhood air pollutant exposure and carotid artery intima-
media thickness in young adults. Circulation 126: 1614-1620.
http://dx.doi.org/10.1161/CIRCULATIONAHA.112.096164
January 2015 6-197 DRAFT: Do Not Cite or Quote
-------
Brook. RD: Jerreft M: Brook. JR: Bard. RL: Finkelstein. MM. (2008). The relationship between diabetes
mellitus and traffic-related air pollution. J Occup Environ Med 50: 32-38.
http://dx.doi.org/10.1097/JOM.Ob013e31815dba70
Brunekreef. B: Beelen. R: Hoek. G: Schouten. L: Bausch-Goldbohm. S: Fischer. P: Armstrong. B:
Hughes. E: Jerrett. M: van den Brandt. P. (2009). Effects of long-term exposure to traffic-related air
pollution on respiratory and cardiovascular mortality in the Netherlands: The NLCS-AIR Study. (139).
Boston, MA: Health Effects Institute.
Burr. ML. (1992). Diagnosing asthma by questionnaire in epidemiological surveys [Editorial]. Clin Exp
Allergy 22: 509-510. http://dx.doi.Org/10.llll/i.1365-2222.1992.tb00158.x
Cai Y: Schikowski. T: Adam. M: Buschka. A: Carsin. AE: Jacquemin. B: Marcon. A: Sanchez. M:
Vierkotter. A: Al-Kanaani. Z: Beelen. R: Birk. M: Brunekreef. B: Cirach. M: Clavel-Chapelon. F:
Declercq. C: de Hoogh. K: de Nazelle. A: Ducret-Stich. RE: Valeria Ferretti. V: Forsberg. B: Gerbase.
MW: Hardy. R: Heinrich. J: Hoek. G: Jarvis. D: Keidel D: Kuh. D: Nieuwenhuijsen. MJ: Ragettli.
MS: Ranzi. A: Rochat T: Schindler. C: Sugiri. D: Temam. S: Tsai. MY: Varraso. R: Kauffmann. F:
Kramer. U: Sunyer. J: Kiinzli. N: Probst-Hensch. N: Hansell. AL. (2014). Cross-sectional associations
between air pollution and chronic bronchitis: an ESCAPE meta-analysis across five cohorts. Thorax.
http://dx.doi.org/10.1136/thoraxinl-2013-204352
Cao. J: Yang. C: Li. J: Chen. R: Chen. B: Gu. B: Kan. H. (2011). Association between long-term exposure
to outdoor air pollution and mortality in China: A cohort study. J Hazard Mater 186: 1594-1600.
http://dx.doi.0rg/10.1016/j.jhazmat.2010.12.036
Carey. IM: Atkinson. RW: Kent. AJ: van Staa. T: Cook. DG: Anderson. HR. (2013). Mortality
associations with long-term exposure to outdoor air pollution in a national english cohort. Am J Respir
Crit Care Med 187: 1226-1233. http://dx.doi.org/10.1164/rccm.201210-1758OC
Carlsten. C: Brauer. M: Dimich-Ward. H: Dybuncio. A: Becker. AB: Chan-Yeung. M. (201 la). Combined
exposure to dog and indoor pollution: incident asthma in a high-risk birth cohort. Eur Respir J 37:
324-330. http://dx.doi.org/10.1183/09031936.00187609
Carlsten. C: Dybuncio. A: Becker. A: Chan-Yeung. M: Brauer. M. (2011V). GSTP1 polymorphism
modifies risk for incident asthma associated with nitrogen dioxide in a high-risk birth cohort [Letter].
Occup Environ Med In Press: In Press, http://dx.doi.org/10.1136/oem.2010.063560
Carlsten. C: Dybuncio. A: Becker. A: Chan-Yeung. M: Brauer. M. (2011c). Traffic-related air pollution
and incident asthma in a high-risk birth cohort. Occup Environ Med 68: 291-295.
http://dx.doi.org/10.1136/oem.2010.055152
Castro-Giner. F: Kiinzli. N: Jacquemin. B: Forsberg. B: de Cid. R: Sunyer. J: Jarvis. D: Briggs. D:
Vienneau. D: Norback. D: Gonzalez. JR: Guerra. S: Janson. C: Anto. JM: Wist M: Heinrich. J:
Estivill. X: Kogevinas. M. (2009). Traffic-related air pollution, oxidative stress genes, and asthma
(ECHRS). Environ Health Perspect 117: 1919-1924. http://dx.doi.org/10.1289/ehp.0900589
Cesaroni. G: Badaloni. C: Gariazzo. C: Stafoggia. M: Sozzi. R: Davoli M: Forastiere. F. (2013). Long-
term exposure to urban air pollution and mortality in a cohort of more than a million adults in rome.
Environ Health Perspect 121: 324-331. http://dx.doi.org/10.1289/ehp.1205862
Cesaroni. G: Forastiere. F: Stafoggia. M: Andersen. ZJ: Badaloni C: Beelen. R: Caracciolo. B: de Faire.
U: Erbel R: Eriksen. KT: Fratiglioni. L: Galassi. C: Hampel. R: Heier. M: Hennig. F: Hilding. A:
Hoffmann. B: Houthuijs. D: Jockel KH: Korek. M: Lanki. T: Leander. K: Magnusson. PK: Migliore.
E: Ostenson. CG: Overvad. K: Pedersen. NL: J. JP: Penell J: Pershagen. G: Pyko. A: Raaschou-
Nielsen. O: Ranzi. A: Ricceri. F: Sacerdote. C: Salomaa. V: Swart. W: Turunen. AW: Vineis. P:
Weinmayr. G: Wolf. K: de Hoogh. K: Hoek. G: Brunekreef. B: Peters. A. (2014). Long term exposure
to ambient air pollution and incidence of acute coronary events: prospective cohort study and meta-
analysis in 11 European cohorts from the ESCAPE Project. B M J (Online) 348: f7412.
http://dx.doi.org/10.1136/bmi.f7412
January 2015 6-198 DRAFT: Do Not Cite or Quote
-------
Cesaronj G: Porta. D: Badalonj C: Stafoggia. M: Eeftens. M: Meliefste. K: Forastiere. F. (2012).
Nitrogen dioxide levels estimated from land use regression models several years apart and association
with mortality in a large cohort study. Environ Health 11: 48. http://dx.doi.org/10.1186/1476-069X-ll-
48
Chang. LY: Graham. JA: Miller. FJ: Ospital. JJ: Crapo. JD. (1986). Effects of subchronic inhalation of
low concentrations of nitrogen dioxide: I. The proximal alveolar region of juvenile and adult rats.
Toxicol Appl Pharmacol 83: 46-61. http://dx.doi.org/10.1016/0041-008X(86)90321-2
Channel!. MM: Paffett ML: Devlin. RB: Madden. MC: Campen. MJ. (2012). Circulating factors induce
coronary endothelial cell activation following exposure to inhaled diesel exhaust and nitrogen dioxide
in humans: Evidence from a novel translational in vitro model. Toxicol Sci 127: 179-186.
http://dx.doi.org/10.1093/toxsci/kfs084
Chen. EK: Zmirou-Navier. D: Padilla. C: Deguen. S. (2014). Effects of air pollution on the risk of
congenital anomalies: a systematic review and meta-analysis. Int J Environ Res Public Health 11:
7642-7668. http://dx.doi.org/10.3390/ijerphll0807642
Chen. H: Goldberg. MS: Burnett. RT: Jerrett. M: Wheeler. AJ: Villeneuve. PJ. (2013). Long-term
exposure to traffic-related air pollution and cardiovascular mortality. Epidemiology 24: 35-43.
http://dx.doi.org/10.1097/EDE.Ob013e318276c005
Chen. L: Bell EM: Caton. AR: Druschel. CM: Lin. S. (2010). Residential mobility during pregnancy and
the potential for ambient air pollution exposure misclassification. Environ Res 110: 162-168.
http://dx.doi.0rg/10.1016/j.envres.2009.ll.001
Chung. MK: Lao. TT: Ting. YH: Wong. TW: Leung. TY. (2014). Seasonality of fetal trisomy 21 - have
ambient air pollutants played a role? J Matern Fetal Neonatal Medl-6.
http://dx.doi.org/10.3109/14767058.2014.924104
Clark. C: Crombie. R: Head. J: van Kamp. I: van Kempen. E: Stansfeld. SA. (2012). Does traffic-related
air pollution explain associations of aircraft and road traffic noise exposure on children's health and
cognition? A secondary analysis of the United Kingdom sample from the RANCH project. Am J
Epidemiol 176: 327-337. http://dx.doi.org/10.1093/aje/kws012
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
Cloughertv. JE: Levy. JI: Kubzansky. LD: Ryan. PB: Suglia. SF: Canner. MJ: Wright. RJ. (2007).
Synergistic effects of traffic-related air pollution and exposure to violence on urban asthma etiology.
Environ Health Perspect 115: 1140-1146. http://dx.doi.org/10.1289/ehp.9863
Coogan. PF: White. LF: Jerrett. M: Brook. RD: Su. JG: Seto. E: Burnett. R: Palmer. JR: Rosenberg. L.
(2012). Air pollution and incidence of hypertension and diabetes mellitus in black women living in
Los Angeles. Circulation 125: 767-772. http://dx.doi.org/10.1161/CIRCULATIONAHA.lll.052753
Crapo. JD: Barry. BE: Chang. LY: Mercer. RR. (1984). Alterations in lung structure caused by inhalation
of oxidants. J Toxicol Environ Health 13: 301-321. http://dx.doi.org/10.1080/15287398409530500
Grouse. PL: Goldberg. MS: Ross. NA: Chen. H: Labreche. F. (2010). Postmenopausal breast cancer is
associated with exposure to traffic-related air pollution in Montreal, Canada: Acasecontrol study.
Environ Health Perspect 118: 1578-1583. http://dx.doi.org/10.1289/ehp.1002221
Dadvand. P: Basagana. X: Figueras. F: Martinez. D: Beelen. R: Cirach. M: de Nazelle. A: Hoek. G:
Ostro. B: Nieuwenhuijsen. MJ. (2014a). Air pollution and preterm premature rupture of membranes: a
spatiotemporal analysis. Am J Epidemiol 179: 200-207. http://dx.doi.org/10.1093/aje/kwt240
Dadvand. P: Figueras. F: Basagana. X: Beelen. R: Martinez. D: Cirach. M: Schembari. A: Hoek. G:
Brunekreef. B: Nieuwenhuijsen. MJ. (2013). Ambient air pollution and preeclampsia: a spatiotemporal
analysis. Environ Health Perspect 121: 1365-1371. http://dx.doi.org/10.1289/ehp.1206430
January 2015 6-199 DRAFT: Do Not Cite or Quote
-------
Dadvand. P: Nieuwenhuijsen. MJ: Agusti. A: de Batlle. J: Benet M: Beelen. R: Cirach. M: Martinez. D:
Hoek. G: Basagana. X: Ferrer. A: Ferrer. J: Rodriguez-Roisin. R: Sauleda. J: Guerra. S: Anto. JM:
Garcia-Aymerich. J. (2014b). Air pollution and biomarkers of systemic inflammation and tissue repair
in COPD patients. Eur Respir J 44: 603-613. http://dx.doi.org/10.1183/09031936.00168813
Dadvand. P: Ostro. B: Figueras. F: Foraster. M: Basagana. X: Valentin. A: Martinez. D: Beelen. R:
Cirach. M: Hoek. G: Jerrett M: Brunekreef. B: Nieuwenhuijsen. MJ. (2014c). Residential proximity
to major roads and term low birth weight: the roles of air pollution, heat, noise, and road-adjacent
trees. Epidemiology 25: 518-525. http://dx.doi.org/10.1097/EDE.0000000000000107
Dales. R: Burnett. RT: Smith-Doiron. M: Stieb. DM: Brook. JR. (2004). Air pollution and sudden infant
death syndrome. Pediatrics 113: 628-631.
Dales. R: Wheeler. A: Mahmud. M: Frescura. AM: Smith-Doiron. M: Nethery. E: Liu. L. (2008). The
influence of living near roadways on spirometry and exhaled nitric oxide in elementary
schoolchildren. Environ Health Perspect 116: 1423-1427. http://dx.doi.org/10.1289/ehp.10943
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. (2011). 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
de Burbure. CY: Heilier. JF: Neve. J: Becker. A: Albrecht. C: Borm. PJA: Gromadzinska. J: Wasowicz.
W: Rydzynski. K: Bernard. AM. (2007). Lung permeability, antioxidant status, and NO2 inhalation: A
selenium supplementation study in rats. J Toxicol Environ Health A 70: 284-294.
http://dx.doi.org/10.1080/15287390600884875
de Kluizenaar. Y: van Lenthe. FJ: Visschedijk. AJ: Zandveld. PY: Miedema. HM: Mackenbach. JP
(2013). Road traffic noise, air pollution components and cardiovascular events. Noise Health 15: 388-
397. http://dx.doi.org/10.4103/1463-1741.121230
Delamater. PL: Finlev. AO: Banerjee. S. (2012). An analysis of asthma hospitalizations, air pollution, and
weather conditions in Los Angeles County, California. Sci Total Environ 425: 110-118.
http://dx.doi.0rg/10.1016/i.scitotenv.2012.02.015
Dell SD: Jerrett. M: Beckerman. B: Brook. JR: Fotv. RG: Gilbert. NL: Marshall L: Miller. JD: To. T:
Walter. SD: Stieb. DM. (2014). Presence of other allergic disease modifies the effect of early
childhood traffic-related air pollution exposure on asthma prevalence. Environ Int 65: 83-92.
http://dx.doi.0rg/10.1016/j.envint.2014.01.002
DeMarini. DM: Shelton. ML: Kohan. MJ: Hudgens. EE: Kleindienst TE: Ball LM: Walsh. D: de Boer.
JG: Lewis-Sevan. L: Rabinowitz. J. R.: Claxton. LD: Lewtas. J. (2000). Mutagenicity in lung of Big
Blue(R) mice and induction of tandem-base substitutions in Salmonella by the air pollutant
peroxyacetyl nitrate (PAN): Predicted formation of intrastrand cross-links. Mutat Res-Fundam Mol
Mech Mutagen 457: 41-55. http://dx.doi.org/10.1016/80027-5107(00)00121-4
Pi Giovanni. V: Cagiano. R: Carratu. MR: De Salvia. MA: Giustino. A: Cuomo. V. (1994). Alterations in
the ontogeny of rat pup ultrasonic vocalization produced by prenatal exposure to nitrogen dioxide.
Psychopharmacology 116: 423-427.
Dijkema. MBA: Mallant. SF: Gehring. U: van den Hurk. K: Alssema. M: van Strien. RT: Fischer. PH:
Nijpels. G: Stehouwer. CDA: Hoek. G: Dekker. JM: Brunekreef. B. (2011). Long-term exposure to
traffic-related air pollution and type 2 diabetes prevalence in a cross-sectional screening-study in the
Netherlands. Environ Health 10: 76. http://dx.doi.org/10.1186/1476-069X-10-76
January 2015 6-200 DRAFT: Do Not Cite or Quote
-------
Dimakopoulou. K: Samolj E: Beelen. R: Stafoggia. M: Andersen. ZJ: Hoffmann. B: Fischer. P:
Nieuwenhuijsen. M: Vineis. P: Xun. W: Hoek. G: Raaschou-Nielsen. O: Oudin. A: Forsberg. B:
Modig. L: Jousilahti. P: Lankj T: Turunen. A: Oftedal B: Nafstad. P: Schwarze. PE: Penell J:
Fratiglioni. L: Andersson. N: Pedersen. N: Korek. M: De Faire. U: Eriksen. KT: Tj0nneland. A:
Becker. T: Wang. M: Bueno-De-Mesquita. B: Tsai. MY: Eeftens. M: Peelers. PH: Meliefste. K:
Marcon. A: Kramer. U: Kuhlbusch. TA: Vossoughi. M: Key. T: de Hoogh. K: Hampel R: Peters. A:
Heinrich. J: Weinmayr. G: Concin. H: Nagel. G: Ineichen. A: Jacquemin. B: Stempfelet. M: Vilier. A:
Riccerj F: Sacerdote. C: Pedeli. X: Katsoulis. M: Trichopoulou. A: Brunekreef. B: Katsouvanni. K.
(2014). Air pollution and nonmalignant respiratory mortality in 16 cohorts within the ESCAPE
project. Am J Respir Crit Care Med 189: 684-696. http://dx.doi.org/10.1164/rccm.201310-1777OC
Dockerv. DW: Pope. CA. Ill: Xu. X: Spengler. JD: Ware. JH: Fay. ME: Ferris. BG. Jn 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
Dong. G: Qian. Z: Wang. J: Chen. W: Ma. W: Trevathan. E: Xaverius. PK: Declue. R: Wiese. A:
Langston. M: Liu. MM: Wang. D: Ren. W. (2013a). Associations between ambient air pollution and
prevalence of stroke and cardiovascular diseases in 33 Chinese communities. Atmos Environ 77: 968-
973. http://dx.doi.0rg/10.1016/j.atmosenv.2013.06.034
Dong. GH: Chen. T: Liu. MM: Wang. D: Ma. YN: Ren. WH: Lee. YL: Zhao. YD: He. PC. (2011).
Gender differences and effect of air pollution on asthma in children with and without allergic
predisposition: Northeast Chinese children health study. PLoS ONE 6: e22470.
http://dx.doi.org/10.1371/journal.pone.0022470
Dong. GH: Qian. ZM: Xaverius. PK: Trevathan. E: Maalouf. S: Parker. J: Yang. L: Liu. MM: Wang. D:
Ren. WH: Ma. W: Wang. J: Zelicoff. A: Fu. Q: Simckes. M. (2013b). Association between long-term
air pollution and increased blood pressure and hypertension in China. Hypertension 61: 578-584.
http://dx.doi.org/10.1161/HYPERTENSIONAHA.lll.00003
Dong. GH: Zhang. P: Sun. B: Zhang. L: Chen. X: Ma. N: Yu. F: Quo. H: Huang. H: Lee. YL: Tang. N:
Chen. J. (2012). Long-term exposure to ambient air pollution and respiratory disease mortality in
Shenyang, China: A 12-year population-based retrospective cohort study. Respiration 84: 360-368.
http://dx.doi.org/10.1159/000332930
Douglas. GJ: Price. JF: Page. CP (1995). The effect of prolonged exposure to NO2 from birth on airways
responsiveness in rabbits sensitized at birth. Eur Respir J 8: 246-252.
http://dx.doi.org/10.1183/09031936.95.08020246
Ebisu. K: Bell ML. (2012). Airborne PM2.5 chemical components and low birth weight in the
Northeastern and Mid-Atlantic regions of the United States. Environ Health Perspect 120: 1746-1752.
http://dx.doi.org/10.1289/ehp.1104763
Eeftens. M: Beelen. R: Fischer. P: Brunekreef. B: Meliefste. K: Hoek. G. (2011). Stability of measured
and modelled spatial contrasts in NO2 over time. Occup Environ Med 68: 765-770.
http://dx.doi.org/10.1136/oem.2010.061135
Eenhuizen. E: Gehring. U: Wijga. AH: Smit HA: Fischer. PH: Brauer. M: Koppelman. GH: Kerkhof. M:
de Jongste. JC: Brunekreef. B: Hoek. G. (2013). Traffic related air pollution is related to interrupter
resistance in four-year old children. Eur Respir J 41: 1257-1263.
http://dx.doi.org/10.1183/09031936.00020812
Ehrlich. R. (1980). Interaction between environmental pollutants and respiratory infections. Environ
Health Perspect 35: 89-99.
Ehrlich. R: Henry. MC. (1968). Chronic toxicity of nitrogen dioxide: I: Effect on resistance to bacterial
pneumonia. Arch Environ Health 17: 860-865. http://dx.doi.org/10.1080/00039896.1968.10665342
Enkhmaa. D: Warburton. N: Javzandulam. B: Uyanga. J: Khishigsuren. Y: Lodovsamba. S: Enkhtur. S:
Warburton. D. (2014). Seasonal ambient air pollution correlates strongly with spontaneous abortion in
Mongolia. BMC Pregnancy Childbirth 14: 146. http://dx.doi.org/10.1186/1471-2393-14-146
January 2015 6-201 DRAFT: Do Not Cite or Quote
-------
Eriksson. C: Hilding. A: Pyko. A: Bluhm. G: Pershagen. G: Ostenson. CG. (2014). Long-term aircraft
noise exposure and body mass index, waist circumference, and type 2 diabetes: A prospective study.
Environ Health Perspect 122: 687-694. http://dx.doi.org/10.1289/ehp.1307115
Estarlich. M: Ballester. F: Aguilera. I: Fernandez-Somoano. A: Lertxundi. A: Llop. S: Freire. C: Tardon.
A: Basterrechea. M: Sunyer. J: Iniguez. C. (2011). Residential exposure to outdoor air pollution during
pregnancy and anthropometric measures at birth in a multicenter cohort in Spain. Environ Health
Perspect 119: 1333-1338. http://dx.doi.org/10.1289/ehp.1002918
Eze. 1C: Schaffner. E: Fischer. E: Schikowski. T: Adam. M: Imboden. M: Tsai. M: Carballo. D: von
Eckardstein. A: Kiinzli. N: Schindler. C: Probst-Hensch. N. (2014). Long-term air pollution exposure
and diabetes in a population-based Swiss cohort. Environ Int 70: 95-105.
http://dx.doi.0rg/10.1016/j.envint.2014.05.014
Faiz. AS: Rhoads. GG: Demissie. K: Kruse. L: Lin. Y: Rich. DQ. (2012). Ambient air pollution and the
risk of stillbirth. Am J Epidemiol 176: 308-316. http://dx.doi.org/10.1093/aje/kws029
Faiz. AS: Rhoads. GG: Demissie. K: Lin. Y: Kruse. L: Rich. DQ. (2013). Does ambient air pollution
trigger stillbirth? Epidemiology 24: 538-544. http://dx.doi.org/10.1097/EDE.Ob013e3182949ce5
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
Felber Dietrich. D: Gemperli. A: Gaspoz. JM: Schindler. C: Liu. LJ: Gold. PR: Schwartz. J: Rochat T:
Barthelemy. JC: Pons. M: Roche. F: Probst Hensch. NM: Bridevaux. PO: Gerbase. MW: Neu. U:
Ackermann-Liebrich. U. (2008). Differences in heart rate variability associated with long-term
exposure to NO2. Environ Health Perspect 116: 1357-1361. http://dx.doi.org/10.1289/ehp.11377
Fenters. JD: Findlav. JC: Port. CD: Ehrlich. R: Coffin. PL. (1973). Chronic exposure to nitrogen dioxide:
Immunologic, physiologic, and pathologic effects in virus-challenged squirrel monkeys. Arch Environ
Health 27: 85-89. http://dx.doi.org/10.1080/00039896.1973.10666324
Ferris. BG. (1978). Epidemiology standardization project: II recommended respiratory disease
questionnaires for use with adults and children in epidemiological research. Am Rev Respir Dis 118:
7-53.
Filleul. L: Rondeau. V: Vandentorren. S: Le Moual. N: Cantagrel. A: Annesi-Maesano. I: Charpin. D:
Declercq. C: Neukirch. F: Paris. C: Vervloet D: Brochard. P: Tessier. JF: Kauffmann. F: Baldi. I.
(2005). Twenty five year mortality and air pollution: results from the French PAARC survey. Occup
Environ Med 62: 453-460. http://dx.doi.org/10.1136/oem.2004.014746
Foraster. M: Basagana. X: Aguilera. I: Rivera. M: Agis. D: Bouso. L: Deltell A: Marrugat J: Ramos. R:
Sunyer. J: Vila. J: Elosua. R: Kiinzli. N. (2014). Association of long-term exposure to traffic-related
air pollution with blood pressure and hypertension in an adult population-based cohort in Spain (the
REGICOR Study). Environ Health Perspect 122: 404-411. http://dx.doi.org/10.1289/ehp. 1306497
Forbes. LJ: Patel MD: Rudnicka. AR: Cook. DG: Bush. T: Stedman. JR: Whincup. PH: Strachan. DP:
Anderson. HR. (2009a). Chronic exposure to outdoor air pollution and diagnosed cardiovascular
disease: meta-analysis of three large cross-sectional surveys. Environ Health 8: 30.
http://dx.doi.org/10.1186/1476-069X-8-30
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
Franklin. M: Vora. H: Avol. E: Mcconnell. R: Lurmann. F: Liu. F: Penfold. B: Berhane. K: Gilliland. F:
Gauderman. WJ. (2012). Predictors of intra-community variation in air quality. J Expo Sci Environ
Epidemiol 22: 135-147. http://dx.doi.org/10.1038/jes.2011.45
Freire. C: Ramos. R: Puertas. R: Lopez-Espinosa. MJ: Julvez. J: Aguilera. I: Cruz. F: Fernandez. MF:
Sunyer. J: Olea. N. (2010). Association of traffic-related air pollution with cognitive development in
children. J Epidemiol Community Health 64: 223-228. http://dx.doi.org/10.1136/jech.2008.084574
January 2015 6-202 DRAFT: Do Not Cite or Quote
-------
Fuertes. E: Standl. M: Cyrys. J: Berdel. D: von Berg. A: Bauer. CP: Kramer. U: Sugiri. D: Lehmann. I:
Koletzko. S: Carlsten. C: Brauer. M: Heinrich. J. (2013). A longitudinal analysis of associations
between traffic-related air pollution with asthma, allergies and sensitization in the GINIplus and
LISAphis birth cohorts. 1: e!93. http://dx.doi.org/10.7717/peerj.193
Fujimaki. H: Nohara. O. (1994). Changes in the response of lung mast cells isolated from rats and guinea
pigs exposed to nitrogen dioxide. Inhal Toxicol 6: 515-520.
http://dx.doi.org/10.3109/08958379409040508
Furiosi NJ: Crane. SC: Freeman. G. (1973). Mixed sodium chloride aerosol and nitrogen dioxide in air:
Biological effects on monkeys and rats. Arch Environ Occup Health 27: 405-408.
Gan. W: Koehoorn. M: Davies. H: Demers. P: Tamburic. L: Brauer. M. (2011). Long-term exposure to
traffic-related air pollution and the risk of coronary heart disease hospitalization and mortality.
Environ Health Perspect 119: 501-507. http://dx.doi.org/10.1289/ehp.1002511
Gan. WQ: Fitzgerald. JM: Carlsten. C: Sadatsafavi. M: Brauer. M. (2013). Associations of ambient air
pollution with chronic obstructive pulmonary disease hospitalization and mortality. Am J Respir Crit
Care Med 187: Ill-Ill. http://dx.doi.org/10.1164/rccm.201211-2004OC
Gao. Y: Chan. EY: Li. LP: He. QQ: Wong. TW. (2013). Chronic effects of ambient air pollution on lung
function among Chinese children. Arch Dis Child 98: 128-135. http://dx.doi.org/10.1136/archdischild-
2011-301541
Gasana. J: Dillikar. D: Mendv. A: Forno. E: Ramos Vieira. E. (2012). Motor vehicle air pollution and
asthma in children: Ameta-analysis. Environ Res 117: 36-45.
http://dx.doi.0rg/10.1016/j.envres.2012.05.001
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.
http://dx.doi.org/10.1056/NEJMoa040610
Gauderman. WJ: Avol E: Lurmann. F: Kuenzli. N: Gilliland. F: Peters. J: McConnell R. (2005).
Childhood asthma and exposure to traffic and nitrogen dioxide. Epidemiology 16: 737-743.
http://dx.doi.org/10.1097/01.ede.0000181308.51440.75
Gauderman. WJ: Vora. H: McConnell. R: Berhane. K: Gilliland. F: Thomas. D: Lurmann. F: Avol. E:
Kunzli. N: Jerrett M: Peters. J. (2007). Effect of exposure to traffic on lung development from 10 to
18 years of age: Acohort study. Lancet 369: 571-577. http://dx.doi.org/10.1016/S0140-
6736(07)60037-3
Geer. LA: Weedon. J: Bell. ML. (2012). Ambient air pollution and term birth weight in Texas from 1998
to 2004. J Air Waste Manag Assoc 62: 1285-1295. http://dx.doi.org/10.1080/10962247.2012.707632
Gehring. U: Cyrys. J: Sedlmeir. G: Brunekreef. B: Bellander. T: Fischer. P: Bauer. CP: Reinhardt D:
Wichmann. HE: Heinrich. J. (2002). Traffic-related air pollution and respiratory health during the first
2 yrs of life. Eur Respir J 19: 690-698. http://dx.doi.org/10.1183/09031936.02.01182001
Gehring. U: Gruzieva. O: Agius. RM: Beelen. R: Custovic. A: Cyrys. J: Eeftens. M: Flexeder. C: Fuertes.
E: Heinrich. J: Hoffmann. B: de Jongste. JC: Kerkhof. M: Klumper. C: Korek. M: Molter. A: Schultz.
ES: Simpson. A: Sugiri. D: Svartengren. M: von Berg. A: Wijga. AH: Pershagen. G: Brunekreef. B.
(2013). Air pollution exposure and lung function in children: the ESCAPE project. Environ Health
Perspect 121: 1357-1364. http://dx.doi.org/10.1289/ehp.1306770
Gehring. U: Heinrich. J: Kramer. U: Grote. V: Hochadel M: Sugiri. D: Kraft. M: Rauchfuss. K:
Eberwein. HG: Wichmann. HE. (2006). Long-term exposure to ambient air pollution and
cardiopulmonary mortality in women. Epidemiology 17: 545-551.
http://dx.doi.org/10.1097/01.ede.0000224541.38258.87
Gehring. U: Tamburic. L: Sbihj H: Davies. HW: Brauer. M. (2014). Impact of noise and air pollution on
pregnancy outcomes. Epidemiology 25: 351-358. http://dx.doi.org/10.1097/EDE.0000000000000073
January 2015 6-203 DRAFT: Do Not Cite or Quote
-------
Gehring. U: van Eijsden. M: Dijkema. MB: van der Wai MF: Fischer. P: Brunekreef. B. (201 la). Traffic-
related air pollution and pregnancy outcomes in the Dutch ABCD birth cohort study. Occup Environ
Med 68: 36-43. http://dx.doi.org/10.1136/oem.2009.053132
Gehring. U: Wijga. AH: Brauer. M: Fischer. P: de Jongste. JC: Kerkhof. M: Oldenwening. M: Smit. HA:
Brunekreef. B. (2010). Traffic-related air pollution and the development of asthma and allergies
during the first 8 years of life. Am J Respir Crit Care Med 181: 596-603.
http://dx.doi.org/10.1164/rccm.200906-0858OC
Gehring. U: Wijga. AH: Fischer. P: de Jongste. JC: Kerkhof. M: Koppelman. GH: Smit. HA: Brunekreef.
R (20 lib). Traffic-related air pollution, preterm birth and term birth weight in the PIAMA birth
cohort study. Environ Res 111: 125-135. http://dx.doi.Org/10.1016/j.envres.2010.10.004
Ghosh. JK: Heck. JE: Cockburn. M: Su. J: Jerrett M: Ritz. B. (2013). Prenatal exposure to traffic-related
air pollution and risk of early childhood cancers. Am J Epidemiol 178: 1233-1239.
http://dx.doi.org/10.1093/aje/kwtl29
Ghosh. JKC: Wilhelm. M: Su. J: Goldberg. D: Cockburn. M: Jerrett. M: Ritz. B. (2012a). Assessing the
influence of traffic-related air pollution on risk of term low birth weight on the basis of land-use-based
regression models and measures of air toxics. Am J Epidemiol 175: 1262-1274.
http://dx.doi.org/10.1093/aje/kwr469
Ghosh. R: Joad. J: Benes. I: Postal. M: Sram. RJ: Hertz-Picciotto. I. (2012b). Ambient nitrogen oxides
exposure and early childhood respiratory illnesses. Environ Int 39: 96-102.
http://dx.doi.0rg/10.1016/j.envint.2011.10.001
Goldberg. MS: Burnett. RT: Stieb. DM: Brophy. JM: Daskalopoulou. SS: Valois. MF: Brook. JR. (2013).
Associations between ambient air pollution and daily mortality among elderly persons in Montreal,
Quebec. Sci Total Environ 463-464: 931942. http://dx.doi.Org/10.1016/j.scitotenv.2013.06.095
Gotschi. T: Sunyer. J: Chinn. S: de Marco. R: Forsberg. B: Gauderman. JW: Garcia-Esteban. R: Heinrich.
J: Jacquemin. B: Jarvis. D: Ponzio. M: Villani. S: Kiinzli. N. (2008). Air pollution and lung function in
the European Community Respiratory Health Survey. Int J Epidemiol 37: 1349-1358.
http://dx.doi.org/10.1093/ije/dvnl36
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.
http://dx.doi.0rg/10.1136/iech.58.l.ll
Powers. AM: Cullinan. P: Ayres. JG: Anderson. HR: Strachan. DP: Holgate. ST: Mills. 1C: Maynard. RL.
(2012). Does outdoor air pollution induce new cases of asthma? Biological plausibility and evidence;
a review [Review]. Respirology 17: 887-898. http://dx.doi.Org/10.llll/j.1440-1843.2012.02195.x
Greene. ND: Schneider. SL. (1978). Effects of NO2 on the response of baboon alveolar macrophages to
migration inhibitory factor. J Toxicol Environ Health 4: 869-880.
http://dx.doi.org/10.1080/15287397809529708
Gregory. RE: Pickrell. JA: Halm. FF: Hobbs. CH. (1983). Pulmonary effects of intermittent subacute
exposure to low-level nitrogen dioxide. J Toxicol Environ Health 11: 405-414.
http://dx.doi.org/10.1080/15287398309530354
Gronskei. KE: Walker. SE: Gram. F. (1993). EVALUATION OF A MODEL FOR HOURLY SPATIAL
CONCENTRATION DISTRIBUTIONS. Atmos Environ 27: 105-120.
Gruzieva. O: Bellander. T: Eneroth. K: Kull. I: Melen. E: Nordling. E: van Hage. M: Wickman. M:
Moskalenko. V: Hulchiy. O: Pershagen. G. (2012). Traffic-related air pollution and development of
allergic sensitization in children during the first 8 years of life. J Allergy Clin Immunol 129: 240-246.
http://dx.doi.0rg/10.1016/i.iaci.2011.ll.001
Gruzieva. O: Bergstrom. A: Hulchiy. O: Kull I: Lind. T: Melen. E: Moskalenko. V: Pershagen. G:
Bellander. T. (2013). Exposure to air pollution from traffic and childhood asthma until 12 years of age.
Epidemiology 24: 54-61. http://dx.doi.org/10.1097/EDE.Ob013e318276clea
January 2015 6-204 DRAFT: Do Not Cite or Quote
-------
Guxens. M: Aguilera. I: Ballester. F: Estarlich. M: Fernandez-Somoano. A: Lertxundi. A: Lertxundi. N:
Mendez. MA: Tardon. A: Vrijheid. M: Sunyer. J. (2012). Prenatal exposure to residential air pollution
and infant mental development: Modulation by antioxidants and detoxification factors. Environ Health
Perspect 120: 144-149. http://dx.doi.org/10.1289/ehp.1103469
Guxens. M: Garcia-Esteban. R: Giorgis-Allemand. L: Forns. J: Badaloni. C: Ballester. F: Beelen. R:
Cesaroni. G: Chatzi. L: de Agostini. M: de Nazelle. A: Eeftens. M: Fernandez. MF: Fernandez-
Somoano. A: Forastiere. F: Gehring. U: Ghassabian. A: Heude. B: Jaddoe. VW: Kliimper. C:
Kogevinas. M: Kramer. U: Larroque. B: Lertxundi. A: Lertxuni. N: Murcia. M: Navel. V:
Nieuwenhuijsen. M: Porta. D: Ramos. R: Roumeliotaki. T: Slama. R: S0rensen. M: Stephanou. EG:
Sugiri. D: Tardon. A: Tiemeier. H: Tiesler. CM: Verhulst. FC: Vrijkotte. T: Wilhelm. M: Brunekreef.
B: Pershagen. G: Sunyer. J. (2014). Air pollution during pregnancy and childhood cognitive and
psychomotor development: Six European birth cohorts. Epidemiology 25: 636-647.
http://dx.doi.org/10.1097/EDE.0000000000000133
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
Hampel. R: Lepeule. J: Schneider. A: Bottagisi. S: Charles. MA: Ducimetiere. P: Peters. A: Slama. R.
(2011). Short-term impact of ambient air pollution and air temperature on blood pressure among
pregnant women. Epidemiology 22: 671-679. http://dx.doi.org/10.1097/EDE.Ob013e318226e8d6
Han. M: Guo. Z: Li. G: Sang. N. (2013). Nitrogen dioxide inhalation induces genotoxicity in rats.
Chemosphere 90: 2737-2742. http://dx.doi.Org/10.1016/j.chemosphere.2012.ll.057
Hannam. K: Mcnamee. R: Baker. P: Siblev. C: Agius. R. (2014). Air pollution exposure and adverse
pregnancy outcomes in a large UK birth cohort: use of a novel spatio-temporal modelling technique.
Scand J Work Environ Health 40: 518-530. http://dx.doi.org/10.5271/sjweh.3423
Hansel NN: Mccormack. MC: Belli AJ: Matsui EC: Peng. RD: Aloe. C: Paulin. L: Williams. PL:
Diette. GB: Brevsse. PN. (2013). In-home air pollution is linked to respiratory morbidity in former
smokers with chronic obstructive pulmonary disease. Am J Respir Crit Care Med 187: 1085-1090.
http://dx.doi.org/10.1164/rccm.201211-1987OC
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.0rg/10.llll/j.1471-0528.2006.01010.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/j.envres.2006.06.010
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
Hart. JE: Garshick. E: Dockery. DW: Smith. TJ: Ryan. L: Laden. F. (2010). Long-term ambient multi-
pollutant exposures and mortality. Am J Respir Crit Care Med 183: 73-78.
http://dx.doi.org/10.1164/rccm.200912-1903OC
Hart. JE: Garshick. E: Dockery. DW: Smith. TJ: Ryan. L: Laden. F. (2011). Long-term ambient
multipollutant exposures and mortality. Am J Respir Crit Care Med 183: 73-78.
http://dx.doi.org/10.1164/rccm.200912-1903OC
Hart. JE: Rimm. EB: Rexrode. KM: Laden. F. (2013). Changes in traffic exposure and the risk of incident
myocardial infarction and all-cause mortality. Epidemiology 24: 734-742.
http://dx.doi.org/10.1097/EDE.Ob013e31829d5dae
January 2015 6-205 DRAFT: Do Not Cite or Quote
-------
Hasunuma. H: Ishimaru. Y: Yoda. Y: Shima. M. (2014). Decline of ambient air pollution levels due to
measures to control automobile emissions and effects on the prevalence of respiratory and allergic
disorders among children in Japan. Environ Res 131: 111-118.
http://dx.doi.0rg/10.1016/j.envres.2014.03.007
Havashj Y: Kohno. T: Ohwada. H. (1987). Morphological effects of nitrogen dioxide on the rat lung.
Environ Health Perspect 73: 135-145.
Heinrich. J: Thiering. E: Rzehak. P: Kramer. U: Hochadel. M: Rauchfuss. KM: Gehring. U: Wichmann.
HE. (2013). Long-term exposure to NO2 and PM10 and all-cause and cause-specific mortality in a
prospective cohort of women. Occup Environ Med 70: 179-186. http://dx.doi.org/10.1136/oemed-
2012-100876
Henderson. SB: Beckerman. B: Jerrett M: Brauer. M. (2007). Application of land use regression to
estimate long-term concentrations of traffic-related nitrogen oxides and fine paniculate matter.
Environ Sci Technol 41: 2422-2428. http://dx.doi.org/10.1021/es0606780
Henry. MC: Findlav. J: Spangler. J: Ehrlich. R. (1970). Chronic toxicity of NO2 in squirrel monkeys: III.
Effect on resistance to bacterial and viral infection. Arch Environ Health 20: 566-570.
http://dx.doi.org/10.1080/00039896.1970.10665663
Hertel O: Jensen. SS: Ketzel M: Becker. T: Peel RG: Orbv. PV: Skjoth. CA: Ellermann. T: Raaschou-
Nielsen. O: Sorensen. M: Brauner. EV: Andersen. ZJ: Loft. S: Schlunssen. V: Bonlokke. JH:
Sigsgaard. T. (2014). Utilizing monitoring data and spatial analysis tools for exposure assessment of
atmospheric pollutants in Denmark [Chapter 6]. In LL McConnell; J Dachs; CJ Hapeman (Eds.),
Occurrence, fate and impact fo atmospheric pollutants on environmental and human health (pp. 95-
122). Washington, D.C.: American Chemical Society, http://dx.doi.org/10.1021/bk-2013-1149
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. http://dx.doi.Org/10.1016/j.atmosenv.2008.05.057
Hoek. G: Brunekreef. B: Goldbohm. S: Fischer. P: Van den Brandt. PA. (2002). Association between
mortality and indicators of traffic-related air pollution in the Netherlands: A cohort study. Lancet 360:
1203-1209. http://dx.doi.org/10.1016/s0140-6736(02)11280-3
Hou. HY: Wang. D: Zou. XP: Yang. ZH: Li TC: Chen. YQ. (2014). Does ambient air pollutants increase
the risk of fetal loss? A case-control study. Arch Gynecol Obstet 289: 285-291.
http://dx.doi.org/10.1007/s00404-013-2962-l
Huang. YC: Rappold. AG: Graff. DW: Ohio. AJ: Devlin. RB. (2012). Synergistic effects of exposure to
concentrated ambient fine pollution particles and nitrogen dioxide in humans. Inhal Toxicol 24: 790-
797. http://dx.doi.org/10.3109/08958378.2012.718809
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. JJK: Lee. YL: Lin. YC: Y-LL. G. (2006). Relation between air pollution and
allergic rhinitis in Taiwanese schoolchildren. Respir Res 7: 23. http://dx.doi.org/10.1186/1465-9921-7-
23
Hwang. BF: Lee. YL. (2010). Air pollution and prevalence of bronchitic symptoms among children in
Taiwan. Chest 138: 956-964. http://dx.doi.org/10.1378/chest.09-2600
Hwang. BF: Lee. YL: Jaakkola. JJ. (2011). Air pollution and stillbirth: A population-based case-control
study in Taiwan. Environ Health Perspect 119: 1345-1349. http://dx.doi.org/10.1289/ehp. 1003056
Hwang. BF: Lee. YL: Lin. YC: Jaakkola. JJK: Quo. YL. (2005). Traffic related air pollution as a
determinant of asthma among Taiwanese school children. Thorax 60: 467-473.
http://dx.doi.org/10.1136/thx.2004.033977
January 2015 6-206 DRAFT: Do Not Cite or Quote
-------
Hystad. P: Demers. PA: Johnson. KG: Carpiano. RM: Brauer. M. (2013). Long-term Residential Exposure
to Air Pollution and Lung Cancer Risk. Epidemiology 24: 762-772.
http://dx.doi.org/10.1097/EDE.Ob013e3182949ae7
Ichinose. T: Fujij K: Sagai M. (1991). Experimental studies on tumor promotion by nitrogen dioxide.
Toxicology 67: 211-225. http://dx.doi.org/10.1016/0300-483X(91)90144-P
Ichinose. T: Sagai M. (1992). Combined exposure to NO2, O3 and H2SO4-aerosol and lung tumor
formation in rats. Toxicology 74: 173-184. http://dx.doi.org/10.1016/0300-483X(92)90137-4
Ichinose. T: Sagai. M: Kubota. K. (1983). Changes of lipid peroxidation and antioxidative protective
systems in lungs or rats exposed acutely, subacutely and chronically to nitrogen dioxide. J Jpn Soc Air
Pollut 18: 132-146.
Iniguez. C: Ballester. F: Estarlich. M: Esplugues. A: Murcia. M: Llop. S: Plana. A: Amoros. R:
Rebagliato. M. (2012). Prenatal exposure to traffic-related air pollution and fetal growth in a cohort of
pregnant women. Occup Environ Med 69: 736-744. http://dx.doi.org/10.1136/oemed-2011-100550
Jacquemin. B: Sunyer. J: Forsberg. B: Aguilera. I: Bouso. L: Briggs. D: de Marco. R: Garcia-Esteban. R:
Heinrich. J: Jarvis. D: Maldonado. JA: Pavo. F: Rage. E: Vienneau. D: Kiinzli. N. (2009a).
Association between modelled traffic-related air pollution and asthma score in the ECRHS. Eur Respir
J 34: 834-842. http://dx.doi.org/10.1183/09031936.00138208
Jacquemin. B: Sunyer. J: Forsberg. B: Aguilera. I: Briggs. D: Garcia-Esteban. R: Gotschi. T: Heinrich. J:
Jarvholm. B: Jarvis. D: Vienneau. D: Kiinzli. N. (2009b). Home outdoor NO2 and new onset of self-
reported asthma in adults. Epidemiology 20: 119-126.
http://dx.doi.org/10.1097/EDE.Ob013e3181886e76
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: Beckerman. BS: Turner. MC: Krewski. D: Thurston. G: Martin. RV: van
Donkelaar. A: Hughes. E: Shi. Y: Gapstur. SM: Thun. MJ: Pope. CA. III. (2013). Spatial Analysis of
Air Pollution and Mortality in California. Am J Respir Crit Care Med 188: 593-599.
http://dx.doi.org/10.1164/rccm.201303-0609OC
Jerrett. M: Finkelstein. MM: Brook. JR: Arain. MA: Kanaroglou. P: Stieb. DM: Gilbert. NL: Verma. D:
Finkelstein. N: Chapman. KR: Sears. MR. (2009). A cohort study of traffic-related air pollution and
mortality in Toronto, Ontario, Canada. Environ Health Perspect 117: 772-777.
http://dx.doi.org/10.1289/ehp.11533
Jerrett. M: Shankardass. K: Berhane. K: Gauderman. WJ: Kiinzli. N: Avol. E: Gilliland. F: Lurmann. F:
Molitor. JN: Molitor. JT: Thomas. DC: Peters. J: Mcconnell. R. (2008). Traffic-related air pollution
and asthma onset in children: A prospective cohort study with individual exposure measurement.
Environ Health Perspect 116: 1433-1438. http://dx.doi.org/10.1289/ehp.10968
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.
Johnson. JY: Rowe. BH: Villeneuve. PJ. (2010). Ecological analysis of long-term exposure to ambient air
pollution and the incidence of stroke in Edmonton, Alberta, Canada. Stroke 41: 1319-1325.
http://dx.doi.org/10.1161/STROKEAHA.110.580571
Johnson. JYM: Rowe. BH: Allen. RW: Peters. PA: Villeneuve. PJ. (2013). A case-control study of
medium-term exposure to ambient nitrogen dioxide pollution and hospitalization for stroke. BMC
Public Health 13: 368. http://dx.doi.org/10.1186/1471-2458-13-368
Jurewicz. J: Radwan. M: Sobala. W: Polanska. K: Radwan. P: Jakubowski. L: Ulanska. A: Hanke. W.
(2014). The relationship between exposure to air pollution and sperm disomy. Environ Mol Mutagen.
http://dx.doi.org/10.1002/em.21883
January 2015 6-207 DRAFT: Do Not Cite or Quote
-------
Kannan. S: Misra. DP: Dvonch. JT: 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.
http://dx.doi.org/10.1289/ehp.9081
Kashima. S: Naruse. H: Yorifuji. T: Ohki. S: Murakoshj T: Takao. S: Tsuda. T: Pol H. (2011).
Residential proximity to heavy traffic and birth weight in Shizuoka, Japan. Environ Res 111: 377-387.
http://dx.doi.0rg/10.1016/j.envres.2011.02.005
Katanoda. K: Sobue. T: Satoh. H: Tajima. K: Suzuki. T: Nakatsuka. H: Takezaki. T: Nakavama. T: Nitta.
H: Tanabe. K: Tominaga. S. (2011). An association between long-term exposure to ambient air
pollution and mortality from lung cancer and respiratory diseases in Japan. J Epidemiol 21: 132-143.
http://dx.doi.org/10.2188/iea.JE20100098
Kim. E: Park. H: Hong. YC: Ha. M: Kim. Y: Kim. BN: Kim. Y: Roh. YM: Lee. BE: Ryu. JM: Kim. BM:
Ha. EH. (2014). Prenatal exposure to PM10 and NO2 and children's neurodevelopment from birth to
24months of age: Mothers and Children's Environmental Health (MOCEH) study. Sci Total Environ
481: 439-445. http://dx.doi.Org/10.1016/j.scitotenv.2014.01.107
Kim. YM: Kim. YS: Jeon. SG: Kim. YK. (2013). Immunopathogenesis of allergic asthma: more than the
Th2 hypothesis. Allergy, Asthma & Immunology Research 5: 189-196.
http://dx.doi.0rg/10.4168/aair.2013.5.4.189
Kitabatake. M: Yamamoto. H: Yuan. PF: Manjurul. H: Murase. S: Yamauchj T. (1995). Effects of
exposure to NO2 or SO2 on bronchopulmonary reaction induced by Candida albicans in guinea pigs. J
Toxicol Environ Health 45: 75-82. http://dx.doi.org/10.1080/15287399509531981
Kobavashj T: Miura. T. (1995). Concentration- and time-dependent increase in specific airway resistance
after induction of airway hyperresponsiveness by subchronic exposure of guinea pigs to nitrogen
dioxide. Fundam Appl Toxicol 25: 154-158. http://dx.doi.org/10.1006/faat.1995.1049
Kobavashj T: Shinozakj Y. (1990). Effect of subacute exposure to nitrogen dioxide on the airway
responsiveness of guinea pig. Agents Actions Suppl 31: 71-74.
Koehler. C: Ginzkev. C: Friehs. G: Hackenberg. S: Froelich. K: Scherzed. A: Burghartz. M: Kessler. M:
Kleinsasser. N. (2010). Aspects of nitrogen dioxide toxicity in environmental urban concentrations in
human nasal epithelium. Toxicol Appl Pharmacol 245: 219-225.
http://dx.doi.0rg/10.1016/i.taap.2010.03.003
Koehler. C: Ginzkev. C: Friehs. G: Hackenberg. S: Froelich. K: Scherzed. A: Burghartz. M: Kessler. M:
Kleinsasser. N. (2011). Ex vivo toxicity of nitrogen dioxide in human nasal epithelium at the WHO
defined 1-h limit value. Toxicol Lett 207: 89-95. http://dx.doi.Org/10.1016/j.toxlet.2011.08.004
Koehler. C: Thielen. S: Ginzkev. C: Hackenberg. S: Scherzed. A: Burghartz. M: Paulus. M: Hagen. R:
Kleinsasser. NH. (2013). Nitrogen dioxide is genotoxic in urban concentrations. Inhal Toxicol 25:
341-347. http://dx.doi.org/10.3109/08958378.2013.788104
Kramer. U: Herder. C: Sugiri. D: Strassburger. K: Schikowski. T: Ranft. U: Rathmann. W. (2010). Traffic-
related air pollution and incident type 2 diabetes: Results from the SALIA cohort study. Environ
Health Perspect 118: 1273-1279. http://dx.doi.org/10.1289/ehp.0901689
Kramer. U: Sugiri. D: Ranft. U: Krutmann. J: von Berg. A: Berdel D: Behrendt H: Kuhlbusch. T:
Hochadel. M: Wichmann. HE: Heinrich. J. (2009). Eczema, respiratory allergies, and traffic-related air
pollution in birth cohorts from small-town areas. J Dermatol Sci 56: 99-105.
http://dx.doi.0rg/10.1016/j.jdermsci.2009.07.014
Krewski. D: Burnett. RT: Goldberg. MS: Hoover. K: Siemiatvcki. J: Jerrett M: Abrahamowicz. M:
White. WH. (2000). Reanalysis of the Harvard Six Cities study and the American Cancer Society
study of particulate air pollution and mortality. Cambridge, MA: Health Effects Institute.
http://pubs.healtheffects.org/view. php?id=6
January 2015 6-208 DRAFT: Do Not Cite or Quote
-------
Krewski. D: Jerrett M: Burnett. RT: Ma. R: Hughes. E: Shi. Y: Turner. MC: Pope AC. I: 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 particulate air pollution and mortality [HEI] (pp. 5-114;
discussion 115-136). (ISSN 1041-5505
HEI Research Report 140). Boston, MA: Health Effects Institute.
http://pubs.healtheffects.org/view.php?id=315
Kripke. BJ: Sherwin. RR (1984). Nitrogen dioxide exposure-influence on rat testes. Anesth Analg 63:
526-528.
Kubota. K: Murakami. M: Takenaka. S: Kawai K: Kyono. H. (1987). Effects of long-term nitrogen
dioxide exposure on rat lung: Morphological observations. Environ Health Perspect 73: 157-169.
Kudo. M: Ishigatsubo. Y: Aokj I. (2013). Pathology of asthma [Review]. FMICB 4: 1-16.
http://dx.doi.org/10.3389/fmicb.2013.00263
Kumae. T: Arakawa. H. (2006). Comparison of effects of in vivo nitrogen dioxide exposure starting from
different periods on alveolar macrophage activity, assessed by a chemiluminescence technique in
Brown-Norway rats. Luminescence 21: 226-232. http://dx.doi.org/10.1002/bio.911
Lafuma. C: Harf. A: Lange. F: Bozzi. L: Poncv. JL: Bignon. J. (1987). Effect of low-level NO2 chronic
exposure on elastase-induced emphysema. Environ Res 43: 75-84. http://dx.doi.org/10.1016/S0013-
9351(87)80059-2
Lampl M: Jeantv. P. (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
Laupsa. H: Slordal LH. (2003). Applying model calculations to estimate urban air quality with respect to
the requirements of the EU directives on NO2, PM10 and C6H6. Int J Environ Pollut 20: 309-320.
Laurent O: Hu. J: Li. L: Cockburn. M: Escobedo. L: Kleeman. MJ: Wu. J. (2014). Sources and contents
of air pollution affecting term low birth weight in Los Angeles County, California, 2001-2008.
Environ Res. http://dx.doi.Org/10.1016/j.envres.2014.05.003
Laurent O: Wu. J: Li. L: Chung. J: Bartell S. (2013). Investigating the association between birth weight
and complementary air pollution metrics: a cohort study. Environ Health 12: 18.
http://dx.doi.org/10.1186/1476-069X-12-18
Le. HQ: Batterman. SA: Wirtfa. JJ: Wahl. RL: Hoggatt. KJ: Sadeghnejad. A: Hultin. ML: Depa. M.
(2012). Air pollutant exposure and preterm and term small-for-gestational-age births in Detroit,
Michigan: Long-term trends and associations. Environ Int 44: 7-17.
http://dx.doi.0rg/10.1016/j.envint.2012.01.003
Lee. BE: Ha. EH: Park. HS: Kim. YJ: Hong. YC: Kim. H: Lee. JT. (2003). Exposure to air pollution
during different gestational phases contributes to risks of low birth weight. Hum Reprod 18: 638-643.
http://dx.doi.org/10.1093/humrep/degl02
Lee. PC: Talbott EO: Roberts. JM: Catov. JM: Bilonick. RA: Stone. RA: Sharma. RK: Ritz. B. (2012a).
Ambient air pollution exposure and blood pressure changes during pregnancy. Environ Res 117: 46-
53. http://dx.doi.0rg/10.1016/j.envres.2012.05.011
Lee. PC: Talbott. EO: Roberts. JM: Catov. JM: Sharma. RK: Ritz. B. (2011 a). Particulate air pollution
exposure and C-reactive protein during early pregnancy. Epidemiology 22: 524-531.
http://dx.doi.org/10.1097/EDE.Ob013e31821c6c58
Lee. YL: Hwang. BF: Chen. YA: Chen. JM: Wu. YF. (2012b). Pulmonary function and incident bronchitis
and asthma in children: a community-based prospective cohort study. PLoS ONE 7: e32477.
http://dx.doi.org/10.1371/iournal.pone.0032477
January 2015 6-209 DRAFT: Do Not Cite or Quote
-------
Lee. YL: Wang. WH: Lu. CW: Lin. YH: Hwang. BF. (20lib). Effects of ambient air pollution on
pulmonary function among schoolchildren. Int J Hyg Environ Health 214: 369-375.
http://dx.doi.0rg/10.1016/i.iiheh.2011.05.004
Leem. JH: Kaplan. BM: Shim. YK: Pohl HR: Gotwav. CA: Bullard. SM: Rogers. JF: Smith. MM:
Tylenda. CA. (2006). Exposures to air pollutants during pregnancy and preterm delivery. Environ
Health Perspect 114: 905-910. http://dx.doi.org/10.1289/ehp.8733
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
Lenters. V: Uiterwaal CS: Beelen. R: Bots. ML: Fischer. P: Brunekreef. B: Hoek. G. (2010). Long-term
exposure to air pollution and vascular damage in young adults. Epidemiology 21: 512-520.
http://dx.doi.org/10.1097/EDE.Ob013e3181dec3a7
Lepeule. J: Caini. F: Bottagisi. S: Galineau. J: Hulin. A: Marquis. N: Bohet. A: Siroux. V: Kaminski. M:
Charles. MA: Slama. R. (2010). Maternal exposure to nitrogen dioxide during pregnancy and
offspring birth weight: Comparison of two exposure models. Environ Health Perspect 118: 1483-1489.
http://dx.doi.org/10.1289/ehp.0901509
Li. H: Chen. L: Guo. Z: Sang. N: Li G. (2012). In vivo screening to determine neurological hazards of
nitrogen dioxide (NO2) using Wistar rats. J Hazard Mater 225-226: 46-53.
http://dx.doi.0rg/10.1016/j.jhazmat.2012.04.063
Li. H: Han. M: Guo. L: Li. G: Sang. N. (2011). Oxidative stress, endothelial dysfunction and
inflammatory response in rat heart to NO inhalation exposure. Chemosphere 82: 1589-1596.
http://dx.doi.0rg/10.1016/j.chemosphere.2010.ll.055
Lin. CA: Pereira. LAA: Nishioka. DC: Conceicao. GMS: Graga. ALF: Saldiva. PHN. (2004). Air
pollution and neonatal deaths in Sao Paulo, Brazil. Braz J Med Biol Res 37: 765-770.
http://dx.doi.org/10.1590/S0100-879X2004000500019
Lin. CC: Yang. SK: Lin. KG: Ho. WC: Hsieh. WS: Shu. BC: Chen. PC. (2014). Multilevel analysis of air
pollution and early childhood neurobehavioral development. Int J Environ Res Public Health 11:
6827-6841. http://dx.doi.org/10.3390/iierphll0706827
Lin. W: Huang. W: Zhu. T: Hu. M: Brunekreef. B: Zhang. Y: Liu. X: Cheng. H: Gehring. U: Li. C: Tang.
2L (2011). Acute respiratory inflammation in children and black carbon in ambient air before and
during the 2008 Beijing Olympics. Environ Health Perspect 119: 1507-1512.
http://dx.doi.org/10.1289/ehp.1103461
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 12: 41-73.
http://dx.doi.org/10.1080/713856640
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.
http://dx.doi.org/10.1080/08958370304463
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. http://dx.doi.Org/10.1016/j.atmosenv.2005.09.027
Lipfert. FW: Wyzga. RE: Batv. JD: Miller. JP. (2009). Air pollution and survival within the Washington
University-EPRI veterans cohort: Risks based on modeled estimates of ambient levels of hazardous
and criteria air pollutants. J Air Waste Manag Assoc 59: 473-489. http://dx.doi.org/10.3155/1047-
3289.59.4.473
January 2015 6-210 DRAFT: Do Not Cite or Quote
-------
Lipsett MJ: Ostro. BD: Reynolds. P: Goldberg. D: Hertz. A: Jerrett M: Smith. DF: Garcia. C: Chang.
ET: Bernstein. L. (2011). Long-term exposure to air pollution and cardiorespiratory disease in the
California teachers study cohort. Am J Respir Crit Care Med 184: 828-835.
http://dx.doi.org/10.1164/rccm.201012-2082OC
Liu. C: Flexeder. C: Fuertes. E: Cyrys. J: Bauer. CP: Koletzko. S: Hoffmann. B: von Berg. A: Heinrich. J:
Groups. ftGaLS. (2014a). Effects of air pollution on exhaled nitric oxide in children: Results from the
GINIplus and LISAplus studies. Int J Hyg Environ Health 217: 483-491.
http://dx.doi.0rg/10.1016/j.ijheh.2013.09.006
Liu. C: Fuertes. E: Tiesler. CMT: Birk. M: Babisch. W: Bauer. CP: Koletzko. S: von Berg. A: Hoffmann.
B: Heinrich. J. (2014b). The associations between traffic-related air pollution and noise with blood
pressure in children: Results from the GINIplus and LISAplus studies. Int J Hyg Environ Health 217:
499-505. http://dx.doi.0rg/10.1016/j.ijheh.2013.09.008
Liu. CC: Tsai SS: Chiu. HF: Wu. TN: Chen. CC: Yang. CY. (2009). Ambient exposure to criteria air
pollutants and risk of death from bladder cancer in Taiwan. Inhal Toxicol 21: 48-54.
http://dx.doi.org/10.1080/08958370802207326
Liu. CC: Tsai SS: Chiu. HF: Wu. TN: Yang. CY. (2008). Ambient exposure to criteria air pollutants and
female lung cancer in Taiwan. Inhal Toxicol 20: 311-317.
http://dx.doi.org/10.1080/08958370701866107
Liu. S: Krewski. D: Shi. Y Chen. Y Burnett. R. (2007). 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/si.ies.7500503
Liu. S: Krewski D: Shi. Y Chen. Y Burnett. RT. (2003). Association between gaseous ambient air
pollutants and adverse pregnancy outcomes in Vancouver, Canada. Environ Health Perspect 111:
1773-1778.
Llop. S: Ballester. F: Estarlich. M: Esplugues. A: Rebagliato. M: Iniguez. C. (2010). Preterm birth and
exposure to air pollutants during pregnancy. Environ Res 110: 778-785.
http://dx.doi.0rg/10.1016/j.envres.2010.09.009
Macintyre. EA: Brauer. M: Melen. E: Bauer. CP: Bauer. M: Berdel D: Bergstrom. A: Brunekreef. B:
Chan-Yeung. M: Klumper. C: Fuertes. E: Gehring. U: Gref. A: Heinrich. J: Herbarth. O: Kerkhof. M:
Koppelman. GH: Kozyrskyj. AL: Pershagen. G: Postma. PS: Thiering. E: Tiesler. CM: Carlsten. C:
Group. TS. (2014a). GSTP1 and TNF Gene Variants and Associations between Air Pollution and
Incident Childhood Asthma: The Traffic, Asthma and Genetics (TAG) Study. Environ Health Perspect
122: 418-424. http://dx.doi.org/10.1289/ehp.1307459
Maclntyre. EA: Gehring. U: Molter. A: Fuertes. E: Klumper. C: Kramer. U: Quass. U: Hoffmann. B:
Gascon. M: Brunekreef. B: Koppelman. GH: Beelen. R: Hoek. G: Birk. M: de Jongste. JC: Smit HA:
Cyrys. J: Gruzieva. O: Korek. M: Bergstrom. A: Agius. RM: de Vocht F: Simpson. A: Porta. D:
Forastiere. F: Badaloni. C: Cesaroni. G: Esplugues. A: Fernandez-Somoano. A: Lerxundi. A: Sunyer.
J: Cirach. M: Nieuwenhuijsen. MJ: Pershagen. G: Heinrich. J. (2014b). Air pollution and respiratory
infections during early childhood: an analysis of 10 European birth cohorts within the ESCAPE
Project. Environ Health Perspect 122: 107-113. http://dx.doi.org/10.1289/ehp.1306755
Madsen. C: Gehring. U: Walker. SE: Brunekreef. B: Stigum. H: Naess. O: Nafstad. P. (2010). Ambient air
pollution exposure, residential mobility and term birth weight in Oslo, Norway. Environ Res 110: 363-
371. http://dx.doi.0rg/10.1016/j.envres.2010.02.005
Maheswaran. R: Pearson. T: Smeeton. NC: Beevers. SD: Campbell MJ: Wolfe. CD. (2010). Impact of
outdoor air pollution on survival after stroke: population-based cohort study. Stroke 41: 869-877.
http://dx.doi.org/10.1161/STROKEAHA.109.567743
Malmqvist E: Jakobsson. K: Tinnerberg. H: Rignell-Hydbom. A: Rylander. L. (2013). Gestational
diabetes and preeclampsia in association with air pollution at levels below current air quality
guidelines. Environ Health Perspect 121: 488-493. http://dx.doi.org/10.1289/ehp.1205736
January 2015 6-211 DRAFT: Do Not Cite or Quote
-------
Malmqvist E: Rignell-Hydbom. A: Tinnerberg. H: Bjork. J: Stroh. E: Jakobsson. K: Rittner. R: Rylander.
L_. (2011). Maternal exposure to air pollution and birth outcomes. Environ Health Perspect 119:
553558. http://dx.doi.org/10.1289/ehp.1002564
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
Marcon. A: Fracasso. ME: Marchetti P: Doria. D: Girardi P: Guarda. L: Pesce. G: Pironj V: Ricci P: de
Marco. R. (2014). Outdoor Formaldehyde and NO2 Exposures and Markers of Genotoxicity in
Children Living Near Chipboard Industries. Environ Health Perspect 122: 639-645.
http://dx.doi.org/10.1289/ehp.1307259
Maroziene. L: Grazuleviciene. R. (2002). Maternal exposure to low-level air pollution and pregnancy
outcomes: A population-based study. Environ Health 1: 6. http://dx.doi.org/10.1186/1476-069X-l-6
Martinez. FD: Wright. AL: Taussig. LM: Holberg. CJ: Halonen. M: Morgan. WJ. (1995). Asthma and
wheezing in the first six years of life. N Engl J Med 332: 133-138.
http://dx.doi.org/10.1056/NEJM199501193320301
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. http://dx.doi.org/10.1164/rccm.200304-466OC
McConnell. R: Islam. T: Shankardass. K: Jerrett M: Lurmann. F: Gilliland. F: Gauderman. J: Avol. E:
Kiinzli. 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.
http://dx.doi.org/10.1289/ehp.0901232
Melen. E: Nyberg. F: Lindgren. CM: Berglind. N: Zucchelli M: Nordling. E: Hallberg. J: Svartengren.
M: Morgenstern. R: Kere. J: Bellander. T: Wickman. M: Pershagen. G. (2008). Interactions between
glutathione S-transferase PI, tumor necrosis factor, and traffic-related air pollution for development of
childhood allergic disease. Environ Health Perspect 116: 1077-1084.
http://dx.doi.org/10.1289/ehp.11117
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
Mercer. RR: Costa. PL: Crapo. JD. (1995). Effects of prolonged exposure to low doses of nitric oxide or
nitrogen dioxide on the alveolar septa of the adult rat lung. Lab Invest 73: 20-28.
Mi. YH: Norback. D: Tao. J: Mi. YL: Perm. M. (2006). Current asthma and respiratory symptoms among
pupils in Shanghai, China: Influence of building ventilation, nitrogen dioxide, ozone, and
formaldehyde in classrooms. Indoor Air 16: 454-464. http://dx.doi.Org/10.llll/j. 1600-
0668.2006.00439.x
Miller. FJ: Graham. JA: Raub. JA: Illing. JW: Menache. MG: House. DE: Gardner. DE. (1987).
Evaluating the toxicity of urban patterns of oxidant gases. II. Effects in mice from chronic exposure to
nitrogen dioxide. J Toxicol Environ Health 21: 99-112. http://dx.doi.org/10.1080/15287398709531005
Miller. KA: Siscovick. PS: Sheppard. L: Shepherd. K: Sullivan. JH: Anderson. GL: Kaufman. JD. (2007).
Long-term exposure to air pollution and incidence of cardiovascular events in women. N Engl J Med
356: 447-458.
Mobasher. Z: Salam. MT: Goodwin. TM: Lurmann. F: Ingles. SA: Wilson. ML. (2013). Associations
between ambient air pollution and Hypertensive Disorders of Pregnancy. Environ Res 123: 9-16.
http://dx.doi.0rg/10.1016/j.envres.2013.01.006
Modig. L: Toren. K: Janson. C: Jarvholm. B: Forsberg. B. (2009). Vehicle exhaust outside the home and
onset of asthma among adults. Eur Respir J 33: 1261-1267.
http://dx.doi.org/10.1183/09031936.00101108
January 2015 6-212 DRAFT: Do Not Cite or Quote
-------
Molter. A: Agius. RM: de Vocht F: Lindlev. S: Gerrard. W: Lowe. L: Belgrave. D: Custovic. A: Simpson.
A.. (2013). Long-term exposure to PM10 and NO2 in association with lung volume and airway
resistance in the MAAS birth cohort. Environ Health Perspect 121: 1232-1238.
http://dx.doi.org/10.1289/ehp.1205961
Molter. A: Lindlev. S: de Vocht. F: Agius. R: Kerry. G: Johnson. K: Ashmore. M: Terry. A:
Dimitroulopoulou. S: Simpson. A. (2012). Performance of a microenviromental model for estimating
personal NO2 exposure in children. Atmos Environ 51: 225-233.
http://dx.doi.0rg/10.1016/j.atmosenv.2012.01.030
Molter. A: Lindlev. S: de Vocht. F: Simpson. A: Agius. R. (2010a). Modelling air pollution for
epidemiologic research—Part I: A novel approach combining land use regression and air dispersion.
Sci Total Environ 408: 5862-5869. http://dx.doi.Org/10.1016/j.scitotenv.2010.08.027
Molter. A: Lindlev. S: de Vocht. F: Simpson. A: Agius. R. (201 Ob). Modelling air pollution for
epidemiologic research—Part II: Predicting temporal variation through land use regression. Sci Total
Environ 409: 211-217. http://dx.doi.Org/10.1016/j.scitotenv.2010.10.005
Morales. E: Julvez. J: Torrent. M: de Cid. R: Guxens. M: Bustamante. M: Kiinzli. N: Sunyer. J. (2009).
Association of early-life exposure to household gas appliances and indoor nitrogen dioxide with
cognition and attention behavior in preschoolers. Am J Epidemiol 169: 1327-1336.
http://dx.doi.org/10.1093/aje/kwp067
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
Morgenstern. V: Zutavern. A: Cyrys. J: Brockow. I: Gehring. U: Koletzko. S: Bauer. CP: Reinhardt D:
Wichmann. HE: Heinrich. J. (2007). Respiratory health and individual estimated exposure to traffic-
related air pollutants in a cohort of young children. Occup Environ Med 64: 8-16.
http://dx.doi.org/10.1136/oem.2006.028241
Morgenstern. V: Zutavern. A: Cyrys. J: Brockow. I: Koletzko. S: Kramer. U: Behrendt H: Herbarth. O:
von Berg. A: Bauer. CP: Wichmann. HE: Heinrich. J. (2008). Atopic diseases, allergic sensitization,
and exposure to traffic-related air pollution in children. Am J Respir Crit Care Med 177: 1331-1337.
http://dx.doi.org/10.1164/rccm.200701-036OC
Moridi. M: Ziaei. S: Kazemnejad. A. (2014). Exposure to ambient air pollutants and spontaneous
abortion. J Obstet Gynaecol Res 40: 743-748. http://dx.doi.org/10.llll/jog.12231
Moseler. M: Hendel-Kramer. A: Karmaus. W: Forster. J: Weiss. K: Urbanek. R: Kuehr. J. (1994). Effect
of moderate NO2 air pollution on the lung function of children with asthmatic symptoms. Environ Res
67: 109-124. http://dx.doi.org/10.1006/enrs.1994.1068
Naess. O: Nafstad. P: Aamodt G: Claussen. B: Rosland. P. (2007). Relation between concentration of air
pollution and cause-specific mortality: four-year exposures to nitrogen dioxide and particulate matter
pollutants in 470 neighborhoods in Oslo, Norway. Am J Epidemiol 165: 435-443.
http://dx.doi.org/10.1093/aje/kwk016
Nafstad. P: Haheim. LL: Oftedal. B: Gram. F: Holme. I: Hjermann. I: Leren. P. (2003). Lung cancer and
air pollution: A 27 year follow up of 16 209 Norwegian men. Thorax 58: 1071-1076.
http://dx.doi.org/10.1136/thorax.58.12.1071
Nafstad. P: Haheim. LL: Wisloff. T: Gram. F: Oftedal B: Holme. I: Hjermann. I: Leren. P. (2004). Urban
air pollution and mortality in a cohort of Norwegian men. Environ Health Perspect 112: 610-605.
Nahidi F: Gholami. R: Rashidi Y: Maid. HA. (2014). Relationship between air pollution and pre-
eclampsia in pregnant women: a case-control study. East Mediterr Health J 19: S60-S66.
Neupane. B: Jerrett M: Burnett. RT: Marrie. T: Arain. A: Loeb. M. (2010). Long-term exposure to
ambient air pollution and risk of hospitalization with community-acquired pneumonia in older adults.
Am J Respir Crit Care Med 181: 47-53. http://dx.doi.org/10.1164/rccm.200901-0160OC
January 2015 6-213 DRAFT: Do Not Cite or Quote
-------
NHLBI (National Institutes of Health, National Heart Lung and Blood Institute). (2002). Global initiative
for asthma: Global strategy for asthma management and prevention. U.S. Department of Health and
Human Services. http://www.ginasthma.org/Guidelines/guidelines-2002-original%3a-workshop-
report%2c-global-strate gv-for-asthmamanagement-and-prevention.html
Nicolai. T: Carr. D: Weiland. SK: Duhme. H: Von Ehrenstein. O: Wagner. C: Von Mutius. E. (2003).
Urban traffic and pollutant exposure related to respiratory outcomes and atopy in a large sample of
children. Eur Respir J 21: 956-963.
Nieuwenhuijsen. MJ: Basagana. X: Dadvand. P: Martinez. D: Cirach. M: Beelen. R: Jacquemin. B.
(2014). Air pollution and human fertility rates. Environ Int 70: 9-14.
http://dx.doi.0rg/10.1016/j.envint.2014.05.005
Nishimura. KK: Galanter. JM: Roth. LA: Oh. SS: Thakur. N: Nguyen. EA: Thyne. S: Farber. HJ:
Serebrisky. D: Kumar. R: Brigino-Buenaventura. E: Davis. A: Lenoir. MA: Meade. K: Rodriguez-
Cintron. W: Avila. PC: Borrell LN: Bibbins-Domingo. K: Rodriguez-Santana. J. R.: Sen. S: Lurmann.
F: Balmes. J. R.: Burchard. EG. (2013). Early-life air pollution and asthma risk in minority children:
The GALA II and SAGE II studies. Am J Respir Crit Care Med 188: 309-318.
http://dx.doi.org/10.1164/rccm.201302-0264OC
Nordling. E: Berglind. N: Melen. E: Emenius. G: Hallberg. J: Nyberg. F: Pershagen. G: Svartengren. M:
Wickman. M: Bellander. T. (2008). Traffic-related air pollution and childhood respiratory symptoms,
function and allergies. Epidemiology 19: 401-408. http://dx.doi.org/10.1097/EDE.Ob013e31816alce3
NTP (National Toxicology Program). (2011). Diesel exhaust particulates. In Report on Carcinogens,
Twelfth Edition (pp. 153-155). Research Triangle Park, NC: U.S. Department of Health and Human
Services, Public Health Service, National Toxicology Program.
Nyberg. F: Gustavsson. P: Jarup. L: Bellander. T: Berglind. N: Jakobsson. R: Pershagen. G. (2000). Urban
air pollution and lung cancer in Stockholm. Epidemiology 11: 487-495.
Oftedal B: Brunekreef. B: Nystad. W: Madsen. C: Walker. SE: Nafstad. P. (2008). Residential outdoor air
pollution and lung function in schoolchildren. Epidemiology 19: 129-137.
http://dx.doi.org/10.1097/EDE.Ob013e31815c0827
Oftedal. B: Nystad. W: Brunekreef. B: Nafstad. P. (2009a). Long-term traffic-related exposures and
asthma onset in schoolchildren in Oslo, Norway. Environ Health Perspect 117: 839-844.
http://dx.doi.org/10.1289/ehp.11491
Oftedal B: Walker. S: Gram. F: Mcinnes. H: Nafstad. P. er. (2009b). Modelling long-term averages of
local ambient air pollution in Oslo, Norway: evaluation of nitrogen dioxide, PM10 and PM2.5. Int J
Environ Pollut 36: 110-126.
Ohashi. Y: Nakai. Y: Sugiura. Y: Ohno. Y: Okamoto. H: Tanaka. A: Kakinoki. Y: Havashi. M. (1994).
Nitrogen dioxide-induced eosinophilia and mucosal injury in the nose of the guinea pig. Acta
Otolaryngol 114: 547-551.
Ohyama. K: Ito. T: Kanisawa. M. (1999). The roles of diesel exhaust particle extracts and the promotive
effects of NO2 and/or SO2 exposure on rat lung tumorigenesis. Cancer Lett 139: 189-197.
http://dx.doi.org/10.1016/S0304-3835(99)00040-3
Olsson. D: Ekstrom. M: Forsberg. B. (2012). Temporal variation in air pollution concentrations and
preterm birth-a population based epidemiological study. Int J Environ Res Public Health 9: 272-285.
http://dx.doi.org/10.3390/ijerph9010272
Olsson. D: Mogren. I: Forsberg. B. (2013). Air pollution exposure in early pregnancy and adverse
pregnancy outcomes: a register-based cohort study. BMJ Open 3. http://dx.doi.org/10.1136/bmjopen-
2012-001955
Oudin. A: Stromberg. U: Jakobsson. K: Stroh. E: Lindgren. AG: Norrving. B: Pessah-Rasmussen. H:
Engstrom. G: Bjork. J. (2011). Hospital admissions for ischemic stroke: does long-term exposure to
air pollution interact with major risk factors. Cerebrovasc Dis 31: 284-293.
http://dx.doi.org/10.1159/000322600
January 2015 6-214 DRAFT: Do Not Cite or Quote
-------
Padula. AM: Tager. IB: Carmichael SL: Hammond. SK: Lurmann. F: Shaw. GM. (2013a). The
association of ambient air pollution and traffic exposures with selected congenital anomalies in the
San Joaquin Valley of California. Am J Epidemiol 177: 1074-1085.
http://dx.doi.org/10.1093/aje/kws367
Padula. AM: Tager. IB: Carmichael SL: Hammond. SK: Yang. W: Lurmann. F: Shaw. GM. (2013b).
Ambient air pollution and traffic exposures and congenital heart defects in the San Joaquin Valley of
California. Paediatric and Perinatal Epidemiology (Online) 27: 329-339.
http://dx.doi.org/10.llll/ppe.12055
Padula. AM: Tager. IB: Carmichael SL: Hammond. SK: Yang. W: Lurmann. FW: Shaw. GM. (2013c).
Traffic-related air pollution and selected birth defects in the San Joaquin Valley of California. Birth
Defects Res AClin Mol Teratol 97: 730-735. http://dx.doi.org/10.1002/bdra.23175
Panasevich. S: Leander. K: Ljungman. P: Bellander. T: de Faire. U: Pershagen. G: Nyberg. F. (2013).
Interaction between air pollution exposure and genes in relation to levels of inflammatory markers and
risk of myocardial infarction. BMJ Open 3: e003058. http://dx.doi.org/10.1136/bmjopen-2013-003058
Panasevich. S: Leander. K: Rosenlund. M: Ljungman. P: Bellander. T: de Faire. U: Pershagen. G: Nyberg.
E. (2009). Associations of long- and short-term air pollution exposure with markers of inflammation
and coagulation in a population sample. Occup Environ Med 66: 747-753.
http://dx.doi.org/10.1136/oem.2008.043471
Papathomas. M: Molitor. J: Richardson. S: Riboli E: Vineis. P. (2011). Examining the joint effect of
multiple risk factors using exposure risk profiles: Lung cancer in non-smokers. Environ Health
Perspect 119: 84-91. http://dx.doi.org/10.1289/ehp.1002118
Parent ME: Goldberg. MS: Grouse. PL: Ross. NA: Chen. H: Valois. MF: Liautaud. A. (2013). Traffic-
related air pollution and prostate cancer risk: A case-control study in Montreal, Canada. Occup
Environ Med 70: 511-518. http://dx.doi.org/10.1136/oemed-2012-101211
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: Sathyanaravana. 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
Pathmanathan. S: Krishna. MT: Blomberg. A: Helledav. R: Kelly. FJ: Sandstrom. T: Holgate. ST: Wilson.
SJ: Frew. AJ. (2003). Repeated daily exposure to 2 ppm nitrogen dioxide upregulates the expression of
IL-5, IL-10, IL-13, and ICAM-1 in the bronchial epithelium of healthy human airways. Occup
Environ Med 60: 892-896. http://dx.doi.org/10.1136/oem.60.ll.892
Pattenden. S: Hoek. G: Braun-Fahrlander. C: Forastiere. F: Kosheleva. A: Neuberger. M: Fletcher. T.
(2006). NO2 and children's respiratory symptoms in the PATY study. Occup Environ Med 63: 828-
835. http://dx.doi.org/10.1136/oem.2006.025213
Pedersen. M: Giorgis-Allemand. L: Bernard. C: Aguilera. I: Andersen. AM: Ballester. F: Beelen. RM:
Chatzi. L: Cirach. M: Danileviciute. A: Dedele. A: Eijsden. M. v: Estarlich. M: Fernandez-Somoano.
A: Fernandez. MF: Forastiere. F: Gehring. U: Grazuleviciene. R: Gruzieva. O: Heude. B: Hoek. G: de
Hoogh. K: van Den Hooven. EH: Haberg. SE: Jaddoe. VW: Kliimper. C: Korek. M: Kramer. U:
Lerchundi A: Lepeule. J: Nafstad. P: Nystad. W: Patelarou. E: Porta. D: Postma. D: Raaschou-
Nielsen. O: Rudnai. P: Sunyer. J: Stephanou. E: S0rensen. M: Thiering. E: Tuffnell. D: Varro. MJ:
Vrijkotte. TG: Wijga. A: Wilhelm. M: Wright. J: Nieuwenhuijsen. MJ: Pershagen. G: Brunekreef. B:
Kogevinas. M: Slama. R. (2013). Ambient air pollution and low birthweight: a European cohort study
(ESCAPE). 1: 695-704. http://dx.doi.org/10.1016/S2213-2600(13)70192-9
January 2015 6-215 DRAFT: Do Not Cite or Quote
-------
Pedersen. M: Stayner. L: Slama. R: S0rensen. M: Figueras. F: Nieuwenhuijsen. MJ: Raaschou-Nielsen.
O: Dadvand. P. (2014). Ambient air pollution and pregnancy-induced hypertensive disorders: a
systematic review and meta-analysis [Review]. Hypertension 64: 494-500.
http://dx.doi.org/10.1161/HYPERTENSIONAHA.114.03545
Pereira. G: Cook. AG: Haggar. F: Bower. C: Nassar. N. (2012). Locally derived traffic-related air
pollution and fetal growth restriction: A retrospective cohort study. Occup Environ Med 69: 815-822.
http://dx.doi.org/10.1136/oemed-2011-100509
Pereira. G: Haggar. F: Shand. AW: Bower. C: Cook. A: Nassar. N. (2013). Association between pre-
eclampsia and locally derived traffic-related air pollution: A retrospective cohort study. J Epidemiol
Community Health 67: 147-152. http://dx.doi.org/10.1136/iech-2011-200805
Peters. JM: Avol. E: Gauderman. WJ: Linn. WS: Navidi. W: London. SJ: Margolis. H: Rappaport E:
Vora. H: Gong. H. Jr: Thomas. DC. (1999). 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. http://dx.doi.Org/10.1164/ajrccm.159.3.9804144
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 particulate air pollution. JAMA 287: 1132-
1141.
Pope. CA. Ill: Thun. MJ: Namboodiri. MM: Dockery. DW: Evans. JS: Speizer. FE: Heath. CW. Jr.
(1995). Particulate air pollution as a predictor of mortality in a prospective study of U.S. adults. Am J
Respir Crit Care Med 151: 669-674. http://dx.doi.Org/10.1164/ajrccm/151.3 Pt 1.669
Pujades-Rodriguez. M: Mckeever. T: Lewis. S: Whyatt. D: Britton. J: Venn. A. (2009). Effect of traffic
pollution on respiratory and allergic disease in adults: Cross-sectional and longitudinal analyses. BMC
Pulm Med 9: 42. http://dx.doi.org/10.1186/1471-2466-9-42
Raaschou-Nielsen. O: Andersen. ZJ: Beelen. R: Samoli. E: Stafoggia. M: Weinmayr. G: Hoffmann. B:
Fischer. P: Nieuwenhuijsen. MJ: Brunekreef. B: Xun. WW: Katsouvanni. K: Dimakopoulou. K:
Sommar. J: Forsberg. B: Modig. L: Oudin. A: Oftedal. B: Schwarze. PE: Nafstad. P: De Faire. U:
Pedersen. NL: Qstenson. CG: Fratiglioni. L: Penell. J: Korek. M: Pershagen. G: Eriksen. KT:
S0rensen. M: Tj0nneland. A: Ellermann. T: Eeftens. M: Peeters. PH: Meliefste. K: Wang. M: Bueno-
De-mesquita. B: Key. TJ: De Hoogh. K: Concin. H: Nagel. G: Vilier. A: Grioni. S: Krogh. V: Tsai.
MY: Ricceri. F: Sacerdote. C: Galassi. C: Migliore. E: Ranzi. A: Cesaroni. G: Badaloni. C: Forastiere.
F: Tamavo. I: Amiano. P: Dorronsoro. M: Trichopoulou. A: Bamia. C: Vineis. P: Hoek. G. (2013a).
Air pollution and lung cancer incidence in 17 European cohorts: Prospective analyses from the
European Study of Cohorts for Air Pollution Effects (ESCAPE). Lancet Oncol 14: 813-822.
http://dx.doi.org/10.1016/S1470-2045(13)70279-l
Raaschou-Nielsen. O: Andersen. ZJ: Hvidberg. M: Jensen. SS: Ketzel. M: Sorensen. M: Hansen. J: Loft.
S: Overvad. K: Tjonneland. A. (2011a). Air pollution from traffic and cancer incidence: a Danish
cohort study. Environ Health 10: 67. http://dx.doi.org/10.1186/1476-069X-10-67
Raaschou-Nielsen. O: Andersen. ZJ: Hvidberg. M: Jensen. SS: Ketzel. M: S0rensen. M: Loft S: Overvad.
K: Tj0nneland. A. (20lib). Lung cancer incidence and long-term exposure to air pollution from
traffic. Environ Health Perspect 119: 860-865. http://dx.doi.org/10.1289/ehp.1002353
Raaschou-Nielsen. O: Bak. H: S0rensen. M: Jensen. SS: Ketzel. M: Hvidberg. M: Schnohr. P:
Tj0nneland. A: Overvad. K: Loft. S. (2010a). Air pollution from traffic and risk for lung cancer in
three Danish cohorts. Cancer Epidemiol Biomarkers Prev 19: 1284-1291.
http://dx.doi.org/10.1158/1055-9965.EPI-10-0036
Raaschou-Nielsen. O: Hermansen. MN: Loland. L: Buchvald. F: Pipper. CB: S0rensen. M: Loft. S:
Bisgaard. H. (2010b). Long-term exposure to indoor air pollution and wheezing symptoms in infants.
Indoor Air 20: 159-167. http://dx.doi.Org/10.llll/i.1600-0668.2009.00635.x
January 2015 6-216 DRAFT: Do Not Cite or Quote
-------
Raaschou-Nielsen. O: Hertel. O: Vignati E: Berkowicz. R: Jensen. SS: Larsen. VB: Lohse. C: Olsen. JH.
(2000). An air pollution model for use in epidemiological studies: evaluation with measured levels of
nitrogen dioxide and benzene. J Expo Sci Environ Epidemiol 10: 4-14.
http://dx.doi.org/10.1038/sj.jea.7500070
Raaschou-Nielsen. O: S0rensen. M: Ketzel. M: Hertel. O: Loft. S: Tj0nneland. A: Overvad. K: Andersen.
ZJ. (2013b). Long-term exposure to traffic-related air pollution and diabetes-associated mortality: a
cohort study. Diabetologia 56: 36-46. http://dx.doi.org/10.1007/s00125-012-2698-7
Rage. E: Jacquemin. B: Nadif. R: Oryszczyn. MP: Siroux. V: Aguilera. I: Kauffmann. F: Kiinzli. N.
(2009). Total serum IgE levels are associated with ambient ozone concentration in asthmatic adults.
Allergy 64: 40-46. http://dx.doi.Org/10.llll/j.1398-9995.2008.01800.x
Rahmalia. A: Giorgis-Allemand. L: Lepeule. J: Philippat C: Galineau. J: Hulin. A: Charles. MA: Slama.
R. (2012). Pregnancy exposure to atmospheric pollutants and placental weight: An approach relying
on a dispersion model. Environ Int 48: 47-55. http://dx.doi.Org/10.1016/j.envint.2012.06.013
Ranzi. A: Porta. D: Badaloni. C: Cesaroni. G: Lauriola. P: Davoli. M: Forastiere. F. (2014). Exposure to
air pollution and respiratory symptoms during the first 7 years of life in an Italian birth cohort. Occup
Environ Med 71: 430-436. http://dx.doi.org/10.1136/oemed-2013-101867
Rich. DQ: Demissie. K: Lu. SE: Kamat L: Wartenberg. D: Rhoads. GG. (2009). Ambient air pollutant
concentrations during pregnancy and the risk of fetal growth restriction. J Epidemiol Community
Health 63: 488-496. http://dx.doi.org/10.1136/iech.2008.082792
Richters. A: Damji KS. (1990). The relationship between inhalation of nitrogen dioxide, the immune
system, and progression of a spontaneously occurring lymphoma in AKR mice. J Environ Pathol
Toxicol Oncol 10: 225-230.
Richters. A: Kuraitis. K. (1981). Inhalation of NO2 and blood borne cancer cell spread to the lungs. Arch
Environ Health 36: 36-39.
Richters. A: Kuraitis. K. (1983). Air pollutants and the facilitation of cancer metastasis. Environ Health
Perspect 52: 165-168.
Richters. A: Richters. V: Alley. WP. (1985). The mortality rate from lung metastases in animals inhaling
nitrogen dioxide (NO2). J Surg Oncol 28: 63-66. http://dx.doi.org/10.1002/jso.2930280115
Riedl. MA: Diaz-Sanchez. D: Linn. WS: Gong. H. Jn Clark. KW: Effros. RM: Miller. JW: Cocker. PR:
Berhane. KT. (2012). Allergic inflammation in the human lower respiratory tract affected by exposure
to diesel exhaust [HEI] (pp. 5-43; discussion 45-64). (ISSN 1041-5505
Research Report 165). Boston, MA: Health Effects Institute.
http://pubs.healtheffects.org/view. php?id=373
Ritz. B: Qiu. J: Lee. PC: Lurmann. F: Penfold. B: Erin Weiss. R: Mcconnell R: Arora. C: Hobel C:
Wilhelm. M. (2014). Prenatal air pollution exposure and ultrasound measures of fetal growth in Los
Angeles, California. Environ Res 130: 7-13. http://dx.doi.Org/10.1016/j.envres.2014.01.006
Ritz. B: Wilhelm. M. (2008). Ambient air pollution and adverse birth outcomes: Methodologic issues in
an emerging field [Review]. Basic Clin Pharmacol Toxicol 102: 182-190.
http://dx.doi.0rg/10.llll/i.1742-7843.2007.00161.x
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. http://dx.doi.org/10.1093/aje/kwml81
Ritz. B: Wilhelm. M: Zhao. Y. (2006). Air pollution and infant death in southern California, 1989-2000.
Pediatrics 118: 493-502. http://dx.doi.org/10.1542/peds.2006-0027
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.
January 2015 6-217 DRAFT: Do Not Cite or Quote
-------
Rivera. M: Basagana. X: Aguilera. I: Foraster. M: Agis. D: de Groot E: Perez. L: Mendez. MA: Bouso.
L: Targa. J: Ramos. R: Sala. J: Marrugat J: Elosua. R: Kiinzli. N. (2013). Association between long-
term exposure to traffic-related air pollution and subclinical atherosclerosis: The REGICOR study.
Environ Health Perspect 121: 223-230. http://dx.doi.org/10.1289/ehp. 1205146
Rojas-Martinez. R: Perez-Padilla. R: Olaiz-Fernandez. G: Mendoza-Alvarado. L: Moreno-Macias. H:
Fortoul. T: McDonnell. W: Loomis. D: Romieu. I. (2007a). 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
Rojas-Martinez. R: Perez-Padilla. R: Olaiz-Fernandez. G: Mendoza-Alvarado. L: Moreno-Macias. H:
Fortoul. T: McDonnell. W: Loomis. D: Romieu. I. (2007b). Lung function growth in children with
long-term exposure to air pollutants in Mexico City. Online data supplement [Supplemental Data]. Am
J Respir Crit Care Med [176].
Romero. R: Espinoza. J: Kusanovic. JP: Gotsch. F: Hassan. S: Erez. O: Chaiworapongsa. T: Mazor. M.
(2006). The preterm parturition syndrome [Review]. BJOG 113: 17-42.
http://dx.doi.0rg/10.llll/i.1471-0528.2006.01120.x
Rosenlund. M: Bellander. T: Nordquist T: Alfredsson. L. (2009a). Traffic-generated air pollution and
myocardial infarction. Epidemiology 20: 265-271. http://dx.doi.org/10.1097/EDE.Ob013e318190ea68
Rosenlund. M: Forastiere. F: Porta. D: De Sario. M: Badaloni. C: Perucci CA. (2009b). Traffic-related air
pollution in relation to respiratory symptoms, allergic sensitisation and lung function in
schoolchildren. Thorax 64: 573-580. http://dx.doi.org/10.1136/thx.2007.094953
Rosenlund. M: Picciotto. S: Forastiere. F: Stafoggia. M: Perucci CA. (2008). Traffic-related air pollution
in relation to incidence and prognosis of coronary heart disease. Epidemiology 19: 121-128.
http://dx.doi.org/10.1097/EDE.Ob013e31815cl921
Rubes. J: Rybar. R: Prinosilova. P: Veznik. Z: Chvatalova. I: Solansky. I: Sram. RJ. (2010). Genetic
polymorphisms influence the susceptibility of men to sperm DNA damage associated with exposure to
air pollution. Mutat Res 683: 9-15. http://dx.doi.Org/10.1016/j.mrfmmm.2009.09.010
Sagai. M: Ichinose. T: Kubota. K. (1984). Studies on the biochemical effects of nitrogen dioxide. IV.
Relation between the change of lipid peroxidation and the antioxidative protective system in rat lungs
upon life span exposure to low levels of NO2. Toxicol Appl Pharmacol 73: 444-456.
Sagai. M: Ichinose. T: Oda. H: Kubota. K. (1982). Studies on biochemical effects of nitrogen dioxide. II.
Changes of the protective systems in rat lungs and of lipid peroxidation by acute exposure. J Toxicol
Environ Health 9: 153-164. http://dx.doi.org/10.1080/15287398209530150
Salam. MT: Millstein. J: Li. YF: Lurmann. FW: Margolis. HG: Gilliland. FD. (2005). Birth outcomes and
prenatal exposure to ozone, carbon monoxide, and particulate matter: Results from the Children's
Health Study. Environ Health Perspect 113: 1638-1644. http://dx.doi.org/10.1289/ehp.8111
Samet JM: Lambert. WE: Skipper. BJ: Gushing. AH: Hunt. WC: Young. SA: McLaren. LC: Schwab. M:
Spengler. JD. (1993). Nitrogen dioxide and respiratory illness in children. Part I: Health outcomes.
(HEI Research Report no. 58). Cambridge, MA: Health Effects Institute.
Sathyanaravana. S: Zhou. C: Rudra. CB: Gould. T. im: Larson. T. im: Koenig. J: Karr. CJ. (2013).
Prenatal ambient air pollution exposure and small for gestational age birth in the Puget Sound Air
Basin. Air Qual Atmos Health 6: 455-463. http://dx.doi.org/10.1007/sll869-012-0182-7
Savitz. DA: Bobb. JF: Carr. JL: Cloughertv. JE: Dominici F: Elston. B: Ito. K: Ross. Z: Yee. M: Matte.
TD. (2014). Ambient fine particulate matter, nitrogen dioxide, and term birth weight in new york, new
york. Am J Epidemiol 179: 457-466. http://dx.doi.org/10.1093/aje/kwt268
Schembarj A: Nieuwenhuijsen. MJ: Salvador. J: de Nazelle. A: Cirach. M: Dadvand. P: Beelen. R: Hoek.
G: Basagana. X: Vrijheid. M. (2014). Traffic-Related Air Pollution and Congenital Anomalies in
Barcelona. Environ Health Perspect 122: 317-323. http://dx.doi.org/10.1289/ehp.1306802
January 2015 6-218 DRAFT: Do Not Cite or Quote
-------
Schikowskj T: Sugiri D: Ranft U: Gehring. U: Heinrich. J: Wichmann. HE: Kramer. U. (2007). Does
respiratory health contribute to the effects of long-term air pollution exposure on cardiovascular
mortality? Respir Res 8: 20. http://dx.doi.org/10.1186/1465-9921-8-20
Schikowski. T: Sugiri. D: Ranft. U: Gehring. U: Heinrich. J: Wichmann. HE: Kramer. U. (2005). Long-
term air pollution exposure and living close to busy roads are associated with COPD in women. Respir
Res 22: 152-161. http://dx.doi.org/10.1186/1465-9921-6-152
Schultz. ES: Gruzieva. O: Bellander. T: Bottai. M: Hallberg. J: Kull I: Svartengren. M: Melen. E:
Pershagen. G. (2012). Traffic-related air pollution and lung function in children at 8 years of age: A
birth cohort study. Am J Respir Crit Care Med 186: 1286-1291.
http://dx.doi.org/10.1164/rccm.201206-1045OC
Seilkop. SK: Campen. MJ: Lund. AK: Mcdonald. JD: Mauderly. JL. (2012). Identification of chemical
components of combustion emissions that affect pro-atherosclerotic vascular responses in mice. Inhal
Toxicol 24: 270-287. http://dx.doi.org/10.3109/08958378.2012.667455
Sekine. K: Shima. M: Nitta. Y: Adachi. M. (2004). Long term effects of exposure to automobile exhaust
on the pulmonary function of female adults in Tokyo, Japan. Occup Environ Med 61: 350-357.
http://dx.doi.org/10.1136/oem.2002.005934
Sellier. Y: Galineau. J: Hulin. A: Caini F: Marquis. N: Navel V: Bottagisi. S: Giorgis-Allemand. L:
Jacquier. C: Slama. R: Lepeule. J: group. tEmcs. (2014). Health effects of ambient air pollution: Do
different methods for estimating exposure lead to different results? Environ Int 66: 165-173.
http://dx.doi.0rg/10.1016/j.envint.2014.02.001
Shalamberidze. OP: Tsereteli. NT. (1971a). Effect of low concentrations of sulfur and nitrogen dioxides
on the estrual cycle and reproductive functions of experimental animals. Gig Sanit 36: 178-182.
Shalamberidze. OP: Tsereteli. NT. (1971b). [Effect of small concentrations of sulfur dioxide and nitrogen
peroxide on the estrus cycle and reproductive function in experiments on animals]. Gig Sanit 36: 13-
17.
Shankardass. K: Mcconnell. R: Jerrett M: Milam. J: Richardson. J: Berhane. K. (2009). Parental stress
increases the effect of traffic-related air pollution on childhood asthma incidence. PNAS 106: 12406-
12411. http://dx.doi.org/10.1073/pnas.0812910106
Sherwin. RP: Richters. V. (1982). Hyperplasia of type 2 pneumocytes following 0.34 ppm nitrogen
dioxide exposure: Quantitation by image analysis. Arch Environ Health 37: 306-315.
http://dx.doi.org/10.1080/00039896.1982.10667584
Shima. M: Nitta. Y: Ando. M: Adachi. M. (2002). Effects of air pollution on the prevalence and incidence
of asthma in children. Arch Environ Health 57: 529-535.
http://dx.doi.org/10.1080/00039890209602084
Slama. R: Bottagisi S: Solansky. I: Lepeule. J: Giorgis-Allemand. L: Sram. R. (2013). Short-term impact
of atmospheric pollution on fecundability. Epidemiology 24: 871-879.
http://dx.doi.org/10.1097/EDE.Ob013e3182a702c5
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. http://dx.doi.org/10.1289/ehp.11074
Slama. R: Morgenstern. V: Cyrys. J: Zutavern. A: Herbarth. O: Wichmann. HE: Heinrich. J: Lisa Group.
S.. (2007). Traffic-related atmospheric pollutants levels during pregnancy and offspring's term birth
weight: A study relying on a land-use regression exposure model. Environ Health Perspect 115: 1283-
1292. http://dx.doi.org/10.1289/ehp.10047
S10rdal LH: Walker. SE: Solberg. S. (2003). The urban air dispersion model EPISODE applied in
AirQUIS2003: Technical description. (TR 12/2003). Kjeller, Norway: Norwegian Institute for Air
Research.
January 2015 6-219 DRAFT: Do Not Cite or Quote
-------
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
Sonnenschein-Van der Voort AM: de Kluizenaar. Y: Jaddoe. VW: Gabriele. C: Raat H: Moll HA:
Hofman. A: Pierik. FH: Miedema. HM: de Jongste. JC: Duijts. L. (2012). Air pollution, fetal and
infant tobacco smoke exposure, and wheezing in preschool children: a population-based prospective
birth cohort. Environ Health 11:91. http://dx.doi.org/10.1186/1476-069X-ll-91
S0rensen. M: Andersen. ZJ: Nordsborg. RB: Becker. T: Tj0nneland. A: Overvad. K: Raaschou-Nielsen.
Q (2013). Long-term exposure to road traffic noise and incident diabetes: a cohort study. Environ
Health Perspect 121: 217-222. http://dx.doi.org/10.1289/ehp. 1205503
S0rensen. M: Hoffmann. B: Hvidberg. M: Ketzel. M: Jensen. SS: Andersen. ZJ: Tj0nneland. A: Overvad.
K: Raaschou-Nielsen. O. (2012). Long-term exposure to traffic-related air pollution associated with
blood pressure and self-reported hypertension in a Danish cohort. Environ Health Perspect 120: 418-
424. http://dx.doi.org/10.1289/ehp.1103631
S0rensen. M: Liihdorf. P: Ketzel M: Andersen. ZJ: Tj0nneland. A: Overvad. K: Raaschou-Nielsen. O.
(2014). Combined effects of road traffic noise and ambient air pollution in relation to risk for stroke?
Environ Res 133: 49-55. http://dx.doi.Org/10.1016/j.envres.2014.05.011
Stanojevic. S: Wade. A: Stocks. J: Hankinson. J: Coates. AL: Pan. H: Rosenthal. M: Corey. M: Lebecque.
P: Cole. TJ. (2008). Reference ranges for spirometry across all ages: a new approach. Am J Respir Crit
Care Med 177: 253-260.
Steenhof. M: Mudwav. IS: Gosens. I: Hoek. G: Godri. KJ: Kelly. FJ: Harrison. RM: Pieters. RH: Cassee.
FR: Lebret E: Brunekreef. BA: Strak. M: Janssen. NA. (2013). Acute nasal pro-inflammatory
response to air pollution depends on characteristics other than particle mass concentration or oxidative
potential: the RAPTES project. Occup Environ Med 70: 341-348. http://dx.doi.org/10.1136/oemed-
2012-100993
Stevens. MA: Menache. MG: Crapo. JD: Miller. FJ: Graham. JA. (1988). Pulmonary function in juvenile
and young adult rats exposed to low-level NO2 with diurnal spikes. J Toxicol Environ Health 23: 229-
240. http://dx.doi.org/10.1080/15287398809531109
Stingone. JA: Luben. TJ: Daniels. JL: Fuentes. M: Richardson. DB: Aylsworth. AS: Herring. AH:
Anderka. M: Botto. L: Correa. A: Gilboa. SM: Langlois. PH: Moslev. B: Shaw. GM: Siffel C: Olshan.
AF: Study. NBDP (2014). Maternal exposure to criteria air pollutants and congenital heart defects in
offspring: results from the national birth defects prevention study. Environ Health Perspect 122: 863-
872. http://dx.doi.org/10.1289/ehp.1307289
Strak. M: Janssen. NA: Godri KJ: Gosens. I: Mudwav. IS: Cassee. FR: Lebret E: Kelly. FJ: Harrison.
RM: Brunekreef. B: Steenhof. M: Hoek. G. (2012). Respiratory health effects of airborne particulate
matter: The role of particle size, composition and oxidative potential - The RAPTES project. Environ
Health Perspect 120: 1183-1189. http://dx.doi.org/10.1289/ehp.1104389
Su. JG: Jerrett. M: Beckerman. B. (2009). A distance-decay variable selection strategy for land use
regression modeling of ambient air pollution exposures. Sci Total Environ 407: 3890-3898.
http://dx.doi.0rg/10.1016/i.scitotenv.2009.01.061
Sunyer. J: Jarvis. D: Gotschi. T: Garcia-Esteban. R: Jacquemin. B: Aguilera. I: Ackerman. U: de Marco.
R: Forsberg. B: Gislason. T: Heinrich. J: Norback. D: Villani. S: Kiinzli. N. (2006). Chronic bronchitis
and urban air pollution in an international study. Occup Environ Med 63: 836-843.
http://dx.doi.org/10.1136/oem.2006.027995
Sunyer. J: Pekkanen. J: Garcia-Esteban. R: Svanes. C: Kiinzli. N: Janson. C: de Marco. R: Anto. JM:
Burnev. P. (2007). Asthma score: Predictive ability and risk factors. Allergy 62: 142-148.
http://dx.doi.0rg/10.llll/i.1398-9995.2006.01184.x
January 2015 6-220 DRAFT: Do Not Cite or Quote
-------
Sunyer. J: Puig. C: Torrent. M: Garcia-Algar. O: Calico. I: Munoz-Ortiz. L: Barnes. M: Cullinan. P.
(2004). Nitrogen dioxide is not associated with respiratory infection during the first year of life. Int J
Epidemiol 33: 116-120. http://dx.doi.org/10.1093/ije/dyh037
Suzuki. AK: Tsubone. H: Ichinose. T: Oda. H: Kubota. K. (1981). [Effects of subchronic nitrogen dioxide
exposure on arterial blood pHa, PaCO2 and PaO2 in rats]. Nippon Eiseigaku Zasshi 36: 816-823.
Svendsen. ER: Gonzales. M: Mukerjee. S: Smith. L: Ross. M: Walsh. D: Rhonev. S: Andrews. G:
Ozkaynak. H: Neas. LM. (2012). GIS-modeled indicators of traffic-related air pollutants and adverse
pulmonary health among children in El Paso, Texas. Am J Epidemiol 176 Suppl 7: S131-S141.
http://dx.doi.org/10.1093/aje/kws274
Tabacova. S: Balabaeva. L: Vardev. F. (1984). Nitrogen dioxide: maternal and fetal effects. 25th Congress
of the European Society of Toxicology, June 11-14, 1984, Budapest, Hungary.
Tabacova. S: Nikiforov. B: Balabaeva. L. (1985). Postnatal effects of maternal exposure to nitrogen
dioxide. Neurotoxicol Teratol 7: 785-789.
Tager. IB: Balmes. J: Lurmann. F: Ngo. L: Alcorn. S: Kiinzli. 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
Takano. H: Yanagisawa. R: Inoue. KI: Shimada. A: Ichinose. T: Sadakane. K: Yoshino. S: Yamaki. K:
Morita. M: Yoshikawa. T. (2004). Nitrogen dioxide air pollution near ambient levels is an atherogenic
risk primarily in obese subjects: A brief communication. Exp Biol Med 229: 361-364.
Takenoue. Y: Kaneko. T: Miyamae. T: Mori. M: Yokota. S. (2012). The influence of outdoor NO(2)
exposure on asthma in childhood: Meta-analysis [Review]. Pediatrics International 54: 762-769.
http://dx.doi.Org/10.llll/j.1442-200X.2012.03674.x
Teichert. T: Vossoughi. M: Vierkotter. A: Sugiri. D: Schikowski. T: Schulte. T: Roden. M: Luckhaus. C:
Herder. C: Kramer. U. (2013). Association between traffic-related air pollution, subclinical
inflammation and impaired glucose metabolism: results from the SALIA study. PLoS ONE 8: e83042.
http://dx.doi.org/10.1371/journal.pone.0083042
Tepper. JS: Costa. PL: Winsett DW: Stevens. MA: Doerfler. PL: Watkinson. WP. (1993). Near-lifetime
exposure of the rat to a simulated urban profile of nitrogen dioxide: Pulmonary function evaluation.
Fundam Appl Toxicol 20: 88-96. http://dx.doi.Org/10.1093/toxsci/20.l.88
Thiering. E: Cyrys. J: Kratzsch. J: Meisinger. C: Hoffmann. B: Berdel D: von Berg. A: Koletzko. S:
Bauer. CP: Heinrich. J. (2013). Long-term exposure to traffic-related air pollution and insulin
resistance in children: results from the GINIplus and LISAplus birth cohorts. Diabetologia 56: 1696-
1704. http://dx.doi.org/10.1007/s00125-013-2925-x
Thurlbeck. WM. (1982). Postnatal human lung growth. Thorax 37: 564-571.
Tonne. C: Wilkinson. P. (2013). Long-term exposure to air pollution is associated with survival following
acute coronary syndrome. European Heart Journal (Online) 34: 1306-1311.
http://dx.doi.org/10.1093/eurheartj/ehs480
Trasande. L: Wong. K: Roy. A: Savitz. DA: Thurston. G. (2013). Exploring prenatal outdoor air pollution,
birth outcomes and neonatal health care utilization in a nationally representative sample. J Expo Sci
Environ Epidemiol 23: 315-321. http://dx.doi.org/10.1038/jes.2012.124
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.
http://dx.doi.org/10.1080/08958370500434214
Tseng. CY: Huang. YC: Su. SY Huang. JY: Lai. CH: Lung. CC: Ho. CC: Liaw. YP (2012). Cell type
specificity of female lung cancer associated with sulfur dioxide from air pollutants in Taiwan: an
ecological study. BMC Public Health 12: 4. http://dx.doi.org/10.1186/1471-2458-12-4
January 2015 6-221 DRAFT: Do Not Cite or Quote
-------
Tsubone. H: Suzuki. AK. (1984). Reflex cardiopulmonary responses by stimulation to type J receptors in
rats exposed to NO2. J Toxicol Environ Health 13: 905-917.
http://dx.doi.org/10.1080/15287398409530550
U.S. EPA (U.S. Environmental Protection Agency). (1993). Air quality criteria for oxides of nitrogen, vol.
1-3 [EPAReport]. (EPA/600/8-91/049aF-cF). Research Triangle Park, NC: U.S. Environmental
Protection Agency, Environmental Criteria and Assessment Office.
http://cfpub.epa. gov/ncea/cfm/recordisplav.cfm?deid=40179
U.S. EPA (U.S. Environmental Protection Agency). (2008). Integrated science assessment for oxides of
nitrogen Health criteria [EPA Report]. (EPA/600/R-08/071). Research Triangle Park, NC: U.S.
Environmental Protection Agency, National Center for Environmental Assessment.
http://cfpub.epa.gov/ncea/cfm/recordisplav.cfm7de id=194645
U.S. EPA (U.S. Environmental Protection Agency). (2009). Integrated science assessment for particulate
matter - Annexes with errata [EPAReport]. (EPA/600/R-08/139). Washington, DC: U.S.
Environmental Protection Agency, National Center for Environmental Assessment.
http://cfpub.epa.gov/ncea/cfm/recordisplav.cfm7de id=216546
U.S. EPA (U.S. Environmental Protection Agency). (2013a). Integrated science assessment for lead [EPA
Report]. (EPA/600/R-10/075F). Research Triangle Park, NC: U.S. Environmental Protection Agency,
National Center for Environmental Assessment.
http://cfpub.epa.gov/ncea/cfm/recordisplav.cfm7de id=255721
U.S. EPA (U.S. Environmental Protection Agency). (2013b). Supplemental table S5-2. Epidemiologic
studies of exposure to oxides of nitrogen and pregnancy effects [EPAReport].
U.S. EPA (U.S. Environmental Protection Agency). (2013c). Supplemental table S5-3. Epidemiologic
studies of exposure to oxides of nitrogen and fetal growth [EPA Report].
U.S. EPA (U.S. Environmental Protection Agency). (2013d). Supplemental table S5-4. Epidemiologic
studies exposure to oxides of nitrogen and preterm birth [EPA Report].
U.S. EPA (U.S. Environmental Protection Agency). (2013e). Supplemental table S5-5. Epidemiologic
studies of exposure to oxides of nitrogen and birth weight [EPA Report].
U.S. EPA (U.S. Environmental Protection Agency). (20131). Supplemental table S5-6. Epidemiologic
studies of exposure to oxides of nitrogen and birth defects [EPA Report].
U.S. EPA (U.S. Environmental Protection Agency). (2013g). Supplemental table S5-7. Epidemiologic
studies of exposure to nitrogen dioxide and neurodevelopmental effects [EPAReport].
U.S. EPA (U.S. Environmental Protection Agency). (2013h). Supplemental table S5-8. Epidemiologic
studies of exposure to nitrogen dioxide and infant mortality [EPA Report].
U.S. EPA (U.S. Environmental Protection Agency). (2013i). Supplemental table S5-9. Epidemiologic
studies of long-term exposure to oxides of nitrogen and mortality [EPAReport].
U.S. EPA (U.S. Environmental Protection Agency). (2013J). Supplemental table S5-10. Epidemiologic
studies of exposure to oxides of nitrogen and cancer incidence and mortality [EPA Report].
U.S. EPA (U.S. Environmental Protection Agency). (2014). Supplemental Table S6-1 Calculation of
increments of oxides of nitrogen for standardizing effect estimates.
Urman. R: Mcconnell R: Islam. T: Avol. EL: Lurmann. FW: Vora. H: Linn. WS: Rappaport EB:
Gilliland. FD: Gauderman. WJ. (2014). Associations of children's lung function with ambient air
pollution: joint effects of regional and near-roadway pollutants. Thorax 69: 540-547.
http://dx.doi.org/10.1136/thoraxinl-2012-203159
van den Hooven. EH: de Kluizenaar. Y: Pierik. FH: Hofman. A: van Ratingen. SW: Zandveld. PY:
Lindemans. J: Russcher. H: Steegers. EA: Miedema. HM: Jaddoe. VW. (2012a). Chronic air pollution
exposure during pregnancy and maternal and fetal C-reactive protein levels: The generation R study.
Environ Health Perspect 120: 746-751. http://dx.doi.org/10.1289/ehp.1104345
January 2015 6-222 DRAFT: Do Not Cite or Quote
-------
van den Hooven. EH: de Kluizenaar. Y: Pierik. FH: Hofman. A: van Ratingen. SW: Zandveld. PY:
Mackenbach. JP: Steegers. EA: Miedema. HM: Jaddoe. VW. (2011). Air pollution, blood pressure,
and the risk of hypertensive complications during pregnancy: The generation R study. Hypertension
57: 406-412. http://dx.doi.org/10.1161/HYPERTENSIONAHA.110.164087
van den Hooven. EH: Pierik. FH: de Kluizenaa. Y: Willemsen. SP: Hofman. A: van Ratingen. SW:
Zandvel PYJ: Mackenbach. JP: Steegers. EAP: Miedema. HME: Jaddoe. VWV. (2012b). Air
pollution exposure during pregnancy, ultrasound measures of fetal growth, and adverse birth
outcomes: A prospective cohort study. Environ Health Perspect 120: 150-156.
http://dx.doi.org/10.1289/ehp.1003316
van den Hooven. EH: Pierik. FH: de Kluizenaar. Y: Hofman. A: van Ratingen. SW: Zandveld. PY:
Russcher. H: Lindemans. J: Miedema. HM: Steegers. EA: Jaddoe. VW. (2012c). Air pollution
exposure and markers of placental growth and function: The Generation R Study. Environ Health
Perspect 120: 1753-1759. http://dx.doi.org/10.1289/ehp.1204918
van Kempen. E: Fischer. P: Janssen. N: Houthuijs. D: van Kamp. I: Stansfeld. S: Cassee. F. (2012).
Neurobehavioral effects of exposure to traffic-related air pollution and transportation noise in primary
schoolchildren. Environ Res 115: 18-25. http://dx.doi.Org/10.1016/j.envres.2012.03.002
Villeneuve. PJ: Jerrett M: Brenner. D: Su. J: Chen. H: Mclaughlin. J. R. (2014). A case-control study of
long-term exposure to ambient volatile organic compounds and lung cancer in toronto, Ontario,
Canada. Am J Epidemiol 179: 443-451. http://dx.doi.org/10.1093/aje/kwt289
Volk. HE: Kerin. T: Lurmann. F: Hertz-Picciotto. I: Mcconnell R: Campbell DB. (2014). Autism
spectrum disorder: interaction of air pollution with the MET receptor tyrosine kinase gene.
Epidemiology 25: 44-47. http://dx.doi.org/10.1097/EDE.0000000000000030
Volk. HE: Lurmann. F: Penfold. B: Hertz-Picciotto. I: Mcconnell R. (2013). Traffic-related air pollution,
particulate matter, and autism. Arch Gen Psychiatry 70: 71-77.
http://dx.doi.org/10.1001/jamapsvchiatry.2013.266
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
Wagner. WD: Duncan. BR: Wright. PG: Stokinger. HE. (1965). Experimental study of threshold limit of
NO2. Arch Environ Health 10: 455-466. http://dx.doi.org/10.1080/00039896.1965.10664029
Walker. SE: Slordal. LH: Guerreiro. C: Gram. F: Gronskei. KE. (1999). Air pollution exposure
monitoring and estimation Part II Model evaluation and population exposure. J Environ Monit 1: 321-
326.
Wang. SO: Zhang. JL: Zeng. XD: Zeng. YM: Wang. SC: Chen. SY (2009a). Association of traffic-related
air pollution with children's neurobehavioral functions in Quanzhou, China. Environ Health Perspect
117: 1612-1618. http://dx.doi.org/10.1289/ehp.0800023
Wang. XY: Hu. W: Tong. S. (2009b). Long-term exposure to gaseous air pollutants and cardio-respiratory
mortality in Brisbane, Australia. Geospat Health 3: 257-263.
Wei. Y Davis. J: Bina. WF. (2012). Ambient air pollution is associated with the increased incidence of
breast cancer in US. Int J Environ Health Res 22: 12-21.
http://dx.doi.org/10.1080/09603123.2011.588321
Weir. CH: Yeatts. KB: Sarnat. JA: Vizuete. W: Salo. PM: Jaramillo. R: Cohn. RD: Chu. H: Zeldin. DC:
London. SJ. (2013). Nitrogen dioxide and allergic sensitization in the 2005-2006 National Health and
Nutrition Examination Survey. Respir Med 107: 1763-1772.
http://dx.doi.0rg/10.1016/j.rmed.2013.08.010
Weng. HH: Tsai. SS: Chen. CC: Chiu. HF: Wu. TN: Yang. CY (2008). Childhood leukemia development
and correlation with traffic air pollution in Taiwan using nitrogen dioxide as an air pollutant marker. J
Toxicol Environ Health A 71: 434-438. http://dx.doi.org/10.1080/15287390701839042
January 2015 6-223 DRAFT: Do Not Cite or Quote
-------
Wickman. M: Kull I: Pershagen. G: Nordvall SL. (2002). The BAMSE project: presentation of a
prospective longitudinal birth cohort study. PAI 13: 11-13. http://dx.doi.Org/10.1034/j. 1399-
3038.13.s.15.10.x
Wijga. AH: Kerkhof. M: Gehring. U: de Jongste. JC: Postma. PS: Aalberse. R: Wolse. A: Koppelman.
GH: van Rossem. L: Oldenwening. M: Brunekreef. B: Smit. HA. (2014). Cohort profile: the
prevention and incidence of asthma and mite allergy (PIAMA) birth cohort. Int J Epidemiol 43: 527-
535. http://dx.doi.org/10.1093/iie/dvs231
Wilhelm. M: Ghosh. JK: Su. J: Cockburn. M: Jerrett M: Ritz. B. (2012). Traffic-related air toxics and
term low birth weight in Los Angeles County, California. Environ Health Perspect 120: 132-138.
http://dx.doi.org/10.1289/ehp.1103408
Wilhelm. M: Ritz. B. (2005). Local variations in CO and paniculate 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
Wollmann. HA. (1998). Intrauterine growth restriction: Definition and etiology. Horm Res 49: 1-6.
http://dx.doi.org/10.1159/000053079
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
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/ijerph7062638
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/j.envres.2008.12.012
Wu. J: Lurmann. F: Winer. A: Lu. R: Turco. R: Funk. T. (2005). Development of an individual exposure
model for application to the Southern California children's health study. Atmos Environ 39: 259-273.
http://dx.doi.0rg/10.1016/j.atmosenv.2004.09.061
Wu. J: Ren. C: Delfino. RJ: Chung. J: Wilhelm. M: Ritz. B. (2009). Association between local traffic-
generated air pollution and preeclampsia and preterm delivery in the south coast air basin of
California. Environ Health Perspect 117: 17731779. http://dx.doi.org/10.1289/ehp.0800334
Wu. J: Wilhelm. M: Chung. J: Ritz. B. (2011). Comparing exposure assessment methods for traffic-
related air pollution in an adverse pregnancy outcome study. Environ Res 111: 685-692.
http://dx.doi.0rg/10.1016/j.envres.2011.03.008
Xu. X: Hu. H: Ha. S: Roth. J. (2014). Ambient air pollution and hypertensive disorder of pregnancy. J
Epidemiol Community Health 68: 13-20. http://dx.doi.org/10.1136/jech-2013-202902
Yamamoto. I: Takahashi. M. (1984). Ultrastructural observations of rat lung exposed to nitrogen dioxide
for 7 months. Kitasato Archives of Experimental Medicine 57: 57-65.
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
Yorifuji. T: Kashima. S: Tsuda. T: Ishikawa-Takata. K: Ohta. T: Tsuruta. K: Doi. H. (2013). Long-term
exposure to traffic-related air pollution and the risk of death from hemorrhagic stroke and lung cancer
in Shizuoka, Japan. Sci Total Environ 443: 397-402. http://dx.doi.Org/10.1016/j.scitotenv.2012.10.088
Yorifuji. T: Kashima. S: Tsuda. T: Takao. S: Suzuki. E: Doi. H: Sugiyama. M: Ishikawa-Takata. K: Ohta.
T\ (2010). Long-term exposure to traffic-related air pollution and mortality in Shizuoka, Japan. Occup
Environ Med 67: 111-117. http://dx.doi.org/10.1136/oem.2008.045542
January 2015 6-224 DRAFT: Do Not Cite or Quote
-------
Zeman. KL: Bennett. WD. (2006). Growth of the small airways and alveoli from childhood to the adult
lung measured by aerosol-derived airway morphometry. J Appl Physiol 100: 965-971.
http://dx.doi.org/10.1152/japplphvsiol.00409.2005
Zemp. E: Elsasser. S: Schindler. C: Kiinzlj N: Perruchoud. AP: Domenighetti G: Medici. T: Ackermann-
Liebrich. U: Leuenberger. P: Monn. C: Bologninj G: Bongard. JP: Brandli O: Karrer. W: Keller. R:
Schonj MH: Tschopp. JM: Villiger. B: Zellweger. JP. (1999). Long-term ambient air pollution and
respiratory symptoms in adults (SAPALDIA study). Am J Respir Crit Care Med 159: 1257-1266.
http://dx.doi.0rg/10.1164/ajrccm.159.4.9807052
Zhang. PF: Dong. GH: Sun. BJ: Zhang. LW: Chen. X: Ma. NN: Yu. F: Quo. HM: Huang. H: Lee. YL:
Tang. NJ: Chen. J. (2011). Long-term exposure to ambient air pollution and mortality due to
cardiovascular disease and cerebrovascular disease in Shenyang, China. PLoS ONE 6: e20827.
http://dx.doi.org/10.1371/journal.pone.0020827
Zhou. N: Cui. Z: Yang. S: Han. X: Chen. G: Zhou. Z: Zhai. C: Ma. M: Li. L: Cai. M: Li. Y: Ao. L: Shu.
W: Liu. J: Cao. J. (2014). Air pollution and decreased semen quality: A comparative study of
Chongqing urban and rural areas. Environ Pollut 187: 145-152.
http://dx.doi.0rg/10.1016/i.envpol.2013.12.030
January 2015 6-225 DRAFT: Do Not Cite or Quote
-------
CHAPTER 7 POPULATIONS AND LIFESTAGES
POTENTIALLY AT RISK FOR HEALTH EFFECTS
RELATED TO NITROGEN DIOXIDE EXPOSURE
7.1 Introduction
1 Interindividual variation in human responses to air pollution exposure can result in some
2 groups being at increased risk for detrimental effects in response to ambient exposure to
3 an air pollutant. The NAAQS are intended to protect public health with an adequate
4 margin of safety. Protection is provided for both the population as a whole and those
5 potentially at increased risk for health effects in response to exposure to a criteria air
6 pollutant (e.g., NCh) (see Preamble to the ISA). The scientific literature has used a variety
7 of terms to identify factors and subsequently populations that may be at increased risk of
8 an air pollutant-related health effect including susceptible, vulnerable, sensitive, and
9 at-risk, with recent literature introducing the term response-modifying factor
10 (Vinikoor-Imler et al.. 2014) (see Preamble to the ISA). Due to the inconsistency in
11 definitions for these terms across the scientific literature and the lack of a consensus on
12 terminology in the scientific community, as detailed in the Preamble to this ISA, this
13 chapter focuses on identifying those populations potentially "at-risk" of an NCh-related
14 health effect. This leads to a focus on the identification, evaluation, and characterization
15 of factors to address the main question of what populations and lifestages are at increased
16 risk of an NC>2-related health effect. It is recognized that some factors may lead to a
17 reduction in risk, and these are recognized during the evaluation process, but for the
18 purposes of identifying those populations or lifestages at greatest risk to inform decisions
19 on the NAAQS, the focus of this chapter is on characterizing those factors that may
20 increase risk.
21 Individuals, and ultimately populations, could be at increased risk of an air
22 pollutant-related health effect via multiple avenues. As discussed in the Preamble, risk
23 may be modified by intrinsic or extrinsic factors, differences in internal dose, or
24 differences in exposure to air pollutant concentrations. It is important to note that the
25 emphasis of this chapter is on identifying, evaluating, and characterizing the evidence for
26 factors that potentially increase the risk of health effects related to exposure to NO2,
27 regardless of whether the change in risk is due to intrinsic factors, extrinsic factors,
28 increased internal dose, increased exposure, or a combination. Some studies examined
29 potential at-risk populations or lifestages based on NOx exposures, but these studies are
30 not discussed in this chapter because they do not directly inform whether a population is
31 at increased risk of an NC>2-related health effect (Section 1.1). It is important to note that
January 2015 7-1 DRAFT: Do Not Cite or Quote
-------
1 although individual factors that may increase the risk of an NO2-related health effect are
2 discussed in this chapter, it is likely in many cases that portions of the population are at
3 increased risk of an NO2-related health effect due to a combination of multiple factors
4 (e-g-, residential location and SES), but information on the interaction among factors
5 remains limited. Thus, the following sections indentify, evaluate, and characterize the
6 overall weight-of-evidence for individual factors to determine if they potentially lead to
7 increased risk for NO2-related health effects (see Preamble to the ISA).
7.2 Approach to Evaluating and Characterizing the Evidence for
At-Risk Factors
8 The systematic approach used to evaluate factors that may increase the risk of a
9 population or specific lifestage to an air pollutant-related health effect is described in
10 more detail in the Preamble. The evidence evaluated includes relevant studies discussed
11 in Chapter 5 and Chapter 6 of this ISA building on the evidence presented in the 2008
12 ISA for Oxides of Nitrogen (U.S. EPA. 2008) and the 1993 Air Quality Criteria for
13 Oxides of Nitrogen (U.S. EPA. 1993). Additionally, within this chapter each factor is
14 evaluated using the current weight-of-evidence framework to charcaterize whether the
15 factor may lead to increased risk of an air pollutant-related health effect in a population
16 or specific lifestage as detailed in previous ISAs (U.S. EPA. 2013a. b). In general, the
17 current approach builds on the causal framework used throughout the ISA. The
18 characterization of each factor consists of evaluating the evidence across scientific
19 disciplines for individual health effects and assessing the overall weight-of-evidence to
20 detail the level of confidence that a specific factor may result in a population or lifestage
21 being at increased risk of an NO2-related health effect
22 As discussed in the Preamble, this evaluation focuses on epidemiologic studies that
23 conduct stratified analyses to compare populations or lifestages exposed to similar air
24 pollutant concentrations within the same study design. During the evaluation of these
25 studies, important considerations include whether the stratified analyses were planned a
26 priori or were post-hoc analyses, whether the study conducted multiple comparisons, and
27 whether there were small sample sizes in individual strata. These study design issues can
28 increase the probability of finding associations by chance or reduce power to detect
29 associations in subgroup analyses. Experimental studies in human subjects or animal
30 models that focus on factors, such as genetic background or health status, are also
31 important lines of evidence to evaluate because they inform coherence and biological
32 plausibility of effects observed in epidemiologic studies, as well as the independent effect
33 of NO2. Additionally, studies examining whether factors may result in differential
January 2015 7-2 DRAFT: Do Not Cite or Quote
-------
1
2
o
J
4
5
6
7
8
9
10
11
exposure to NCh and subsequently increased risk of an NCh-related health effects are also
included.
The objective of this chapter is to identify, evaluate, and characterize the evidence
regarding factors that may increase the risk an NC>2-related health effect in a population
or lifestage, building on the conclusions drawn in the ISA with respect to NC>2 exposure
and health effects. The factors that are evaluated in this chapter include pre-existing
disease (Section 7.3). genetic background (Section 7.4). sociodemographics (Section 7.5).
and behavioral and other factors (Section 7.6). These categories are described in more
detail in Table 7-1, and a summary of the characterization of the evidence for each factor
considered that may increase the risk of NCh-related health effects is presented in
Section 7.7.
Table 7-1 Characterization of evidence for factors potentially increasing the
risk for nitrogen dioxide-related health effects.
Classification
Health Effects
Adequate There is substantial, consistent evidence within a discipline to conclude that a factor results in a
evidence population or lifestage being at increased risk of air pollutant-related health effect(s) relative to
some reference population or lifestage. Where applicable, this evidence includes coherence
across disciplines. Evidence includes multiple high-quality studies.
Suggestive The collective evidence suggests that a factor results in a population or lifestage being at
evidence increased risk of air pollutant-related health effect(s) 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 The collective evidence is inadequate to determine whether a factor results in a population or
evidence lifestage being at increased risk of air pollutant-related health effect(s) 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.
Evidence of no There is substantial, consistent evidence within a discipline to conclude that a factor does not
effect result in a population or lifestage being at increased risk of air pollutant-related health effect(s)
relative to some reference population or lifestage. Where applicable, the evidene includes
coherence across disciplines. Evidence includes multiple high-quality studies.
7.3 Pre-Existing Disease/Conditions
12 Individuals with pre-existing disease may be considered at greater risk for some air
13 pollution-related health effects because they may be in a compromised biological state
14 depending on the disease and severity. The 2008 ISA for Oxides of Nitrogen (U.S. EPA.
15 2008) concluded that those with pre-existing pulmonary conditions were likely to be at
January 2015
7-3
DRAFT: Do Not Cite or Quote
-------
1 greater risk for NCh-related health effects, especially individuals with asthma. Among
2 recent studies evaluating effect modification by pre-existing disease, the largest group
3 examined asthma (Section 7.3.1). Several studies are available on other diseases,
4 including chronic pulmonary respiratory disease (COPD, Section 7.3.2). cardiovascular
5 disease (CVD, Section 7.3.3). diabetes (Section 7.3.4) and obesity (Section 7.3.5). Table
6 7-2 presents the prevalence of these diseases according to the Center for Disease
7 Control's National Center for Health Statistics (Schiller etal.. 2012). including the
8 proportion of adults with a current diagnosis categorized by age and geographic region.
9 The large proportions of the U.S. population affected by many chronic diseases,
10 including various cardiovascular diseases, indicates the potential public health impact of
11 characterizing the risk of NC>2-related health effects for affected populations.
Table 7-2 Prevalence of respiratory diseases,
diabetes, and obesity among adults
2010.
Chronic
Disease/Condition
All (N, in thousands)
Adults (18+)
N
(in thousands)
229,505
cardiovascular diseases,
by age and region in the
Age (%)a
18-44
110,615
45-64
80,198
65-74
21,291
75+
17,401
North-
east
40,577
u
.S. in
Region(%)b
Midwest
53,316
South
81,721
West
53,891
Selected respiratory diseases
Asthmac
18,734
8.1
8.4
8.7
7.4
8.7
8.
2
7,
.7
8.4
COPD - ______ __
Chronic bronchitis
Emphysema
9,883
4,314
3.0
0.3
5.3
2.1
6.0
5.4
6.3
6.3
3.8
1.7
4.
2.
,7
,3
4,
1
.7
.9
3.1
1.2
Selected cardiovascular diseases/conditions
All heart disease
Coronary heart
disease
Hypertension
Stroke
Diabetes
Obesity
27,066
15,262
59,259
6,226
20,974
62,026
4.4
1.4
9.3
0.6
2.8
25.4
13.2
7.3
34.4
3.0
12.3
32.8
24.3
16.5
54.2
6.1
22.0
31.5
37.1
25.8
57.3
10.7
21.7
18.2
10.7
6.1
24.0
2.0
7.1
25.1
12.
6.
24.
2.
8.
29.
2
,6
,7
,9
,9
,6
12,
7,
27,
2
10,
29
.3
.2
.1
.9
.1
.4
10.1
5.4
21.7
2.5
8.3
24.4
Percentage of individual adults within each age group with disease, based on N (at the top of each age column).
""Percentage of individual adults (18+) within each geographic region with disease, based on N (at the top of each region column).
°Asthma prevalence is reported for "still has asthma."
Source: Schiller et al. (2012): National Center for Health Statistics: Data from Tables 1 and 2; Tables 3 and 4; and Tables 7 and 8
of the Centers for Disease Control and Prevention report.
January 2015 7-4 DRAFT: Do Not Cite or Quote
-------
7.3.1 Asthma
1 Approximately 8.2% of adults and 9.5% of children (age <18 yr) in the U.S. currently
2 have asthma (Schiller et al.. 2012: Bloom etal.. 2011). and it is the leading chronic
3 illness affecting children. This ISA concludes that a causal relationship exists short-term
4 NC>2 exposure and respiratory effects, based primarily on evidence for effects on asthma
5 exacerbation (Section 5.2.9). The evidence demonstrating NC>2-induced asthma
6 exacerbation formed the basis for the 2008 ISA for Oxides of Nitrogen (U.S. EPA. 2008)
7 concluding that individuals with pre-existing pulmonary conditions, particularly those
8 with asthma, are likely at greater risk of NO2-related health effects. This section evaluates
9 controlled human exposure and epidemiologic studies that compare NO2-related health
10 effects between groups with or without asthma (Tables 7-3 and 7-4). As a whole, results
11 of these comparisons are variable; however, controlled human exposure studies provide
12 compelling evidence that people with asthma are at greater risk for NO2-related
13 respiratory effects than people without asthma.
14 Controlled human exposure studies demonstrating NO2-induced increases in airway
15 responsiveness in adults with asthma provide key evidence for an independent, causal
16 relationship between NO2 exposure and respiratory effects (Section 5.2.9). This evidence
17 also demonstrates greater sensitivity of adults with asthma to short-term NO2 exposure
18 compared to adults without asthma. A meta-analysis conducted by Folinsbee (1992)
19 demonstrates that NO2 exposures in the range of 100-300 ppb increased airway
20 responsiveness in adults with asthma [(Folinsbee. 1992); Table 7-31. In this study
21 Folinsbee (1992) combined groups that varied with respect to respiratory symptoms and
22 medication use at the time of assessment but had high prevalence (50-100%) of atopic
23 asthma. Additionally, in many studies these participants were characterized as having
24 mild asthma. Although fewer studies of healthy adults examined airway responsiveness
25 for NO2 exposures below 300 ppb, results were statistically significant only for NO2
26 exposures >1,000 ppb. In comparison to airway responsiveness, there is inconsistent
27 evidence for the effects of short-term NO2 exposure on lung function in adults with
28 asthma in the absence of a challenge agent, and the limited evidence is inconclusive in
29 demonstrating differences in response between healthy adults and those with asthma
30 KVagaggini et al.. 1996: Torres etal.. 1995: Linn etal.. 1985b): Table 7-31.
31 Epidemiologic evidence does not clearly show differences in NO2-related health effects
32 between children with and without asthma (Table 7-4). Studies characterized as having
33 strong exposure assessment, such as monitoring NO2 at or near children's school, also did
34 not show differences in NO2-related respiratory effects between children with and
35 without asthma [(Lin et al.. 2011: Flamant-Hulin et al.. 2010: Holguin et al.. 2007):
January 2015 7-5 DRAFT: Do Not Cite or Quote
-------
1 Table 7-41. However, asthma is a heterogeneous disease as demonstrated in Section 5.2.2.
2 and a limitation of epidemiologic studies that may obscure potential differences among
3 people with and without asthma is the practice often used of grouping all people with
4 asthma together even though there are varying phenotypes and triggers of asthma as well
5 as varying degrees of response among people with asthma. Additionally, compared to
6 controlled human exposure studies, when examining potential differences among people
7 with and without asthma, epidemiologic studies examined a more diverse set of
8 respiratory outcomes and asthma phenotypes.
9 Several lines of evidence indicate that people with asthma are at increased risk for
10 NC>2-related health effects. The causal relationship determined for short-term NCh
11 exposure and respiratory effects is based on the evidence for asthma exacerbation
12 (Section 5.2.9). Controlled human exposure studies demonstrate that NC>2 has an
13 independent effect on increasing airway responsiveness in adults with asthma and show
14 increased sensitivity of adults with asthma compared to healthy adults. People with
15 asthma also tend to have oronasal breathing, although the implications on differential
16 uptake of NC>2 in the respiratory tract are not known (Section 4.2.2). Epidemiologic
17 studies do not clearly demonstrate differences between populations with and without
18 asthma. The study populations represent an array of asthma phenotypes, and limited
19 epidemiologic evidence indicates larger NCh-related respiratory effects in children with
20 asthma not using asthma medication. Thus, the epidemiologic results are not necessarily
21 incoherent with experimental findings from populations of mostly mild, atopic asthma.
22 Because of clear evidence for an effect of NO2 exposure on asthma exacerbation and for
23 increased sensitivity of adults with asthma to NC>2-induced increases in airway
24 responsiveness in controlled human exposure studies, there is adequate evidence to
25 conclude that people with asthma are at increased risk for NCh-related health effects.
January 2015 7-6 DRAFT: Do Not Cite or Quote
-------
Table 7-3 Controlled human exposure studies evaluating pre-existing asthma.
Factor
Evaluated
Asthma
n = 33
<300 ppb
NO2
Asthma
n = 12
Asthma
n=4
Asthma
n=23
Atopic
asthma
Asthma
Asthma
Asthma
Reference
Category
Healthy
n = 36
<1,000 ppb
NO2
Healthy
n = 8
Healthy
n = 7
Healthy
n=25
None
None
None
None
Direction of
Effect
Modification
or Effect3 Outcome
* Airway
' respons-
iveness
_ Airway
inflamma-
tion
* Lung
' function
decrement
_ Lung
function
decrement
_ Airway
resistance
_ Lung
function
decrement
Allergen
response
Lung
function
decrement
| Lung
' function
decrement
Airway
resistance
Study
Population
N = 355
N=20
Ages 1-33 yr
N = 11
Mean age:
31. Syr
N=48
Ages
18-36yr
N = 11
Mean age:
31.2yr
N=41
Mean age:
31 yr
N = 15
Mean age:
33 yr
N = 11
Ages 7-55 yr
Study Details
Range: 100 ppb NO2for
1 h to 7, 500 ppb NO2 for
2 h of exposure;
Exposures at rest
1,000 ppb for 3 h;
Exercise 10 min on/10 min
off at individual's
maximum workload
300 ppb for 1 h;
Exercise at VE = 25 L/min
4,000 ppb for 75 min;
Two 15-min periods of
exercise at VE = 25 L/min
and 50 L/min
(1)200ppbNO2for6h
(2) 200 ppb
NO2+ 100 ppb OsforB h
(3) 400 ppb NO2 for 3 h
(4) 400 ppb
NO2 + 200 ppb Os for 3 h
(1-4) Exercise 10 min
on/40 min off at
VE = 32 L/min
200 ppb for 2 h;
Exercise 15 min on/15 min
off at VE = ~2 times resting
300 ppb for 30 min
Exercise 10 min on/20 min
off at VE >3 times resting
250 ppb for 30 min;
Exercise 10 min on/20 min
off/ at VE 3 times resting
Study
Folinsbee
(1992)
Jorres et al.
(1995)
Vaqaqqini
etal. (1996)
Linn et al.
(1985b)
Jenkins et al.
(1999)
Kleinman et al.
(1983)
Bauer etal.
(1986)
Jorres and
Maqnussen
(1991)
aUp facing arrow indicates that the effect of NO2 is greater (e.g., larger lung function decrement, larger increase in airway
inflammation) in the group with the factor evaluated than in the reference group. Down facing arrow indicates that the effect of NO2 is
smaller in the group with the factor evaluated than in the reference group. A dash indicates no difference in NO2-related health effect
between groups. In some studies, only a population with pre-existing disease was examined; therefore, the arrow or dash represents
the direction of the effect in that population after exposure to NO2 relative to exposure to filtered air.
January 2015
7-7
DRAFT: Do Not Cite or Quote
-------
Table 7-4 Epidemiologic studies evaluating
Factor Reference
Evaluated Category
Direction of
Effect
Modification3
Outcome
pre-existing
Study
Population
asthma.
Study Details
Study
Short-term exposure
Asthma No asthma
n = 57 n = 192
Asthma No asthma
n - 50 8Q% n - 1 58
with atopy 72% with
atopy
Asthma No asthma
n - 8 n - 30
Asthma No asthma
n = 100 n = 100
Asthma No asthma
n = 34 n = 70
Asthma No asthma
n = 169 n= 2,071
t
1
-
1
Respiratory
symptoms
Pulmonary
inflammation,
lung function
decrement
Pulmonary
inflammation
Pulmonary
inflammation
Pulmonary
inflammation
Pulmonary
inflammation
N = 249
Ages 14-20 yr
N = 208
Ages 7.9-11 yr
N = 38
Ages 9-1 2 yr
N = 200
Ages 6-12 yr
N = 104
Mean age:
10.3 yr
N = 2,240
Ages 5-7 yr
New York, NY,
2003-2005
Mexico City,
2003-2005
Beijing, China,
2008
Ciudad Juarez,
Mexico,
2001-2002
Clermont-
Ferrand, France
Southern
California,
2004-2005
Patel et al.
(2010)
Barraza-
Villarreal et al.
(2008)
Lin etal. (2011)
Holquin et al.
(2007)
Flamant-Hulin
etal. (2010)
Berhane et al.
(2011)
Long-term exposure
Asthma No asthma
n - 50 n - 1,934
Asthma No asthma
n - 1,273 n - 50,545
t
t
Incident stroke
Fatal stroke
Diabetes
N = 1,984
Ages 50-65 yr at
baseline
N = 51,818
Ages 50-65 yr at
baseline
Copenhagen,
Aarhus
counties,
Denmark,
1993-2006
Long-term NO2
Copenhagen,
Aarhus
counties,
Denmark,
1993-2006
Long-term NO2
Andersen et al.
(2012a)
Andersen et al.
(2012b)
aUp facing arrow indicates that the effect of NO2 is greater (e.g., larger lung function decrement, larger increase in pulmonary
inflammation) in the group with the factor evaluated than in the reference group. Down facing arrow indicates that the effect of NO2 is
smaller in the group with the factor evaluated than in the reference group. A dash indicates no difference in NO2-related health effect
between groups.
7.3.2 Chronic Obstructive Pulmonary Disease
1 Chronic lower respiratory disease, including COPD, was ranked as the third leading
2 cause of death in the United States in 2011 (Hoyert and Xu. 2012). COPD comprises
3 chronic bronchitis and emphysema which affect approximately 9.9 million and 4.3
January 2015
7-8
DRAFT: Do Not Cite or Quote
-------
1 million adults in the U.S., respectively [Table 7-2; (Schiller et al., 2012)1. Given that
2 people with COPD have compromised respiratory function and systemic inflammation,
3 they may be at increased risk for an array of NO2-related health effects. Evidence for
4 differential NO2-related respiratory or cardiovascular effects between adults with COPD
5 and those without COPD is weak (Table 7-5 and 7-6). Based on the ability to inform a
6 direct effect of NO2, controlled human exposure studies of lung function provide a
7 stronger basis for drawing conclusions about whether pre-existing COPD leads to
8 increased risk of NO2-related health effects.
9 Compared with asthma exacerbation, there is greater uncertainty regarding a relationship
10 between short-term NO2 exposure and COPD exacerbation (Section 5.2.2.4). This is
11 illustrated by the lack of consistent evidence from controlled human exposure studies for
12 changes in lung function or pulmonary inflammation in adults with COPD following NO2
13 exposure (Gong et al.. 2005; Linn etal.. 1985a). Among the limited number of studies
14 that compared adults with COPD and healthy adults, only some indicated larger
15 NO2-induced decrements in lung function in adults with COPD (Table 7-5). Among
16 adults with COPD, NO2 exposures of 300 ppb for 1 or 4 hours induced decreases in lung
17 function of 4.8, 8.2 (Morrow et al.. 1992) or 10% (Vagaggini et al.. 1996) relative to air
18 control exposures. In contrast, in healthy adults, NO2 did not have any effect on lung
19 function or resulted in increased lung function. However, adults with COPD were older
20 than healthy adults and had a higher prevalence of smoking, which could have influenced
21 results (Table 7-5). For example, in one study, smokers had larger NO2-induced
22 decrements in lung function independently of COPD (Morrow et al.. 1992).
23 COPD was not observed to modify associations between long-term NO2 exposure and
24 diabetes (Eze etal.. 2014; Andersen et al., 2012b). but some epidemiologic studies show
25 larger association between short-term exposures and cardiovascular-related emergency
26 department (ED) visits and decreases in heart rate variability (HRV) among adults with
27 COPD, as well as long-term exposures and stroke (Table 7-6). Inference is limited from
28 studies of cardiovascular effects because of lack of comparison to healthy groups in the
29 short-term exposure studies (Suh and Zanobetti. 2010; Peel et al.. 2007) and uncertainty
30 regarding an independent relationship for both short-term and long-term NO2 exposure
31 with cardiovascular effects (Sections 5.3.12 and 6.3.9).
32 In conclusion, some but not all epidemiologic evidence points to larger NO2-related
33 cardiovascular effects in adults with COPD, and there is uncertainty as to whether the
34 findings reflect an independent effect of NO2. Controlled human exposure studies do not
35 clearly demonstrate that NO2 exposure induces respiratory effects in adults with COPD,
36 and the limited findings for larger lung function decrements in adults with COPD relative
37 to healthy adults may be influenced by differences between groups in age or smoking.
January 2015 7-9 DRAFT: Do Not Cite or Quote
-------
The limited and inconsistent evidence for NCh-related changes in lung function in adults
with COPD is inadequate to determine whether people with COPD are at increased risk
for NO2-related health effects.
Table 7-5 Controlled human exposure studies evaluating pre-existing COPD.
Factor
Evaluated
COPD
n=20
COPD
n = 7
COPD
n = 18
COPD
Reference
Category
Healthy
n=20
Healthy
n = 7
Healthy
n=6
None
Direction of
Effect
Modification or
Effect3 Outcome
* Lung function
' decrements
_ Symptoms,
Respiratory
conductance
* Lung function
' decrements
_ Sputum cell
counts,
symptoms
_ Lung function
decrements,
heart rate,
blood
pressure,
symptoms
Lung function
decreases,
heart rate,
symptoms
Study
Population
N = 40
Mean age
' 59.9 yr
N = 14
Mean age
' COPD: 58 yr
Healthy:
34 yr
N = 24
Mean age
COPD: 72 yr
Healthy:
68 yr
N = 22
Mean age:
60.8 yr
Study Details Study
300ppbfor4h; Morrow et al.
Three 7 min periods of (1992)
exercise at VE - ~4 times
resting
300 ppb for 1 h; Vaqaqqini
Exercise at VE = 25 L/min et al. (1996)
(1)400 ppb NO2 for 2 h Gonqetal.
(2) 200 ug/m3 CAPs for (2005)
2h
(3) 400 ppb
NO2 + 200 ug/m3 CAPs
for2h
(1-3) Exercise 15 min
on/15 min off at
VE = ~2 times resting
500, 1,000, and Linn et al.
2,000 ppb for 1 h; (1985a)
Exercise 15 min
on/15 min off
VE = 16 L/min
aUp facing arrow indicates that the effect of NO2 is greater (e.g., larger FENA, decrement, larger increase in airway inflammation) in the
group with the factor evaluated than in the reference group. Down facing arrow indicates that the effect of NO2 is smaller in the group
with the factor evaluated than in the reference group. A dash indicates no difference in NO2-related health effect between groups. In
some studies, only a population with pre-existing disease was examined; therefore, the arrow or dash represents the direction of the
effect in that population after exposure to NO2 relative to exposure to filtered air.
January 2015
7-10
DRAFT: Do Not Cite or Quote
-------
Table 7-6
Factor
Evaluated
Epidemiologic studies evaluating pre-existing COPD.
Reference
Category
Direction of
Effect
Modification3 Outcome
Study
Population
Study Details
Study
Short-term exposure
COPD
n = 18
COPD
n = 8% ED
visits
Recent
myocardial
infarction
n = 12
No COPD
n = 92% ED
visits
f HRV pNNSO
' decrements
HRV r-MSSD
decrements
•j> Cardiovascular-
' related ED visits
N = 30
Ages not
reported
31 participating
hospitals,
103,551 ED visits
for
cardiovascular
disease
Atlanta, GA,
1999-2000
Atlanta, GA,
1993-2000
Suh and
Zanobetti
(2010)
Peel et al.
(2007)
Long-term exposure
COPD
n = 121
COPD
n=6
COPD
n= 2,058
COPD
n = 1,268
No COPD
n = 1,863
No COPD
n = 136
No COPD
n= 49,760
No COPD
n = 5,124
_ Incident stroke
-j- Fatal stroke
_ Diabetes
_ Diabetes
N = 1,984
Ages 50-65 yr at
baseline
N = 142
Ages 50-65 yr at
baseline
N = 51,818
Ages 50-65 yr at
baseline
N = 6,392
Ages 29-73 yr
Copenhagen,
Aarhus counties,
Denmark,
1993-2006
Copenhagen,
Aarhus counties,
Denmark,
1993-2006
Switzerland
2002
Andersen
etal. (2012a)
Andersen
etal. (2012b)
Eze et al.
(2014)
aUp facing arrow indicates that the effect of NO2 is greater (e.g., larger decrement in HRV, larger increase in ED visits) in the group
with the factor evaluated than in the reference group. Down facing arrow indicates that the effect of NO2 is smaller in the group with
the factor evaluated than in the reference group. A dash indicates no difference in NO2-related health effect between groups.
7.3.3 Cardiovascular Disease
1 Cardiovascular disease is the primary cause of death in the U.S., and approximately 12%
2 of adults report a diagnosis of heart disease. In addition, hypertension has been diagnosed
3 in roughly 25% of the adult U.S. population [Table 7-2: (Schiller etal.. 2012)1. For both
4 short-term (Section 5.3.12) and long-term (Section 6.3.9) NC>2 exposure, the collective
5 body of evidence is suggestive, but not sufficient to infer a causal relationship with
6 cardiovascular effect. These conclusions were primarily based on the uncertainty in
7 distinguishing an independent effect of NC>2 on cardiovasucular effects. In addition to this
8 uncertainty, the evidence base does not consistently show that pre-existing CVD
January 2015
7-11
DRAFT: Do Not Cite or Quote
-------
1 increases the risk for NCh-related health effects (Table 7-7). Evidence is equally
2 inconsistent for short-term and long-term NC>2 exposure.
3 For short-term exposure, NC>2-related mortality was higher among individuals with CVD
4 (Chiusolo et al.. 2011); however, the majority of evidence, which is for cardiovascular
5 hospital admissions and ED visits, is inconsistent (Table 7-7). The strongest evidence for
6 a relationship between short-term NC>2 exposure and cardiovascular effects is for
7 myocardial infarction (MI, Section 5.4.8). and most studies show no difference in the
8 association in groups with hypertension (Tsai etal.. 2012; Peel et al.. 2007; D'Ippoliti
9 et al.. 2003). arrhythmia (Tsai et al.. 2012; Mann et al.. 2002). or congestive heart failure
10 (Tsai et al.. 2012; D'Ippoliti et al.. 2003). Many studies of long-term NC>2 exposure
11 compared groups with and without hypertension and found no difference in the
12 association with diabetes (Eze etal.. 2014; Andersen et al.. 2012b) and no consistent
13 difference in the association with stroke (Andersen et al.. 2012a). For studies that
14 examined hypertension or blood pressure examined as an outcome, associations with
15 long-term NC>2 exposure were larger in groups with pre-existing CVD (Foraster etal..
16 2014; S0rensen etal.. 2012).
17 In studies examining subclinical cardiovascular effects such as changes in HRV,
18 interleukin (IL)-6, or arrhythmic events recorded on electrocardiograms, most did not
19 observe that associations with short- or long-term NC>2 exposure differed between groups
20 with or without pre-existing CVD, whether defined as any CVD, ischemic heart disease
21 (HD), or hypertension (Panasevich et al.. 2009; Felber Dietrich et al.. 2008; Ljungman
22 etal.. 2008). Risk factors for CVD, including higher systemic inflammation and
23 hypercholesterolemia, do not consistently modify NCh-related cardiovascular effects
24 (Andersen et al.. 2012a; Huang etal.. 2012). Experimental studies (Table 7-8) also do not
25 clearly support greater NC>2-induced subclinical cardiovascular effects as examined in
26 adults with CVD (Scaife etal.. 2012) and a mouse model of CVD (Campen etal.. 2010).
27 For associations with short-term and long-term NC>2 exposure, people with and without
28 pre-existing CVD have been compared with respect to an array of cardiovascular
29 diseases, events, and subclinical effects. Studies are also diverse in the conditions by
30 which they define pre-existing CVD. No consistent difference in NC>2-related
31 cardiovascular effects is demonstrated between groups with and without pre-existing
32 CVD. Additionally, there was limited biological plausibility from the experimental
33 evidence demonstrating NC>2-related health effects in response to pre-existing
34 cardiovascular conditions. In conclusion, the large evidence base lacks sufficient
35 consistency in demonstrating that pre-existing CVD modifies NC>2-related cardiovascular
36 effects, and an independent effect of NO2 is uncertain overall. Therefore, the current
January 2015 7-12 DRAFT: Do Not Cite or Quote
-------
evidence is inadequate to determine whether people with CVD are at increased risk for
NO2-related health effects.
Table 7-7
Factor
Evaluated
Epidemiologic studies
disease.
Direction of
Reference Effect
Category Modification3
evaluating
Outcome
pre-existing cardiovascular
Study
Population Study Details Study
Short-term exposure
Hypertension
n - 30% visits
Hypertension
n = 40% visits
CHF
n = 15% visits
Cardiac
arrhythmia
n = 11% visits
Hypertension
n = 1,648
Heart
conduction
disorder
n =414
Cardiac
dysrhythmia
n = 1,296
Heart failure
n = 703
Secondary
diagnosis of
arrhythmia
n = 34.5%
admissions
Secondary
diagnosis of
CHF
n = 14.1%
admissions
No |
hypertension '
n - 70% visits
No
hypertension
n = 60% visits
No CHF
n = 85% visits
No cardiac _
arrhythmia
n = 89% visits
No
hypertension
n = 4,883
No heart y
conduction '
disorder
n = 6,117
No cardiac _
dysrhythmia
n = 5,235
No heart _
failure
n = 5,828
No secondary _
diagnosis of
arrhythmia
n = 65.5%
admissions
No secondary *
diagnosis of '
CHF
n = 85.9%
admissions
Arrhythmia ED
visits
ED visits for
IHDor
congestive
heart failure
(CHF)
Myocardial
infarction
hospital
admission
First acute
myocardial
infarction
hospital
admission
Hospital
admission for
IHD
31 participatinq Atlanta, GA, Peel et al.
hospitals 1993-2000 (2007)
N - 103,551 ED
visits for
cardiovascular
disease
N = 27,563 Taipei, Taiwan Tsai et al.
hospital 1999-2009 (2012)
admissions
N=6,531 Rome, Italy, D'lppoliti et al.
hospital records 1995-1997 (2003)
N = 54,863 Southern Mann et al.
hospital California, (2002)
admissions 1988-1995
January 2015
7-13
DRAFT: Do Not Cite or Quote
-------
Table 7-7 (Continued): Epidemiologic studies evaluating pre-existing
cardiovascular disease.
Factor
Evaluated
Pre-existing
heart disease
n = 525 with
stroke
IHD
n = 56
Pre-existing
CVD
n = 1.2-1 4%
Reference
Category
No pre-
existing heart
disease
n = 2, 214 with
stroke
No IHD
n = 32
No pre-
existing CVD
n = 86-98.8%
Direction of
Effect
Modification3 Outcome
y Hospital
' admission for
ischemic
stroke
_ Hospital
admission for
hemorrhagic
stroke
_ Ventricular
tachy-
arrhythmia
•j> Total mortality
Study
Population
N = 5,927
hospital
admissions
N = 88 with
implantable
cardioverter
defibrillators
Age 28-85 yr
N 276,205
natural deaths
Study Details
Edmonton,
Canada,
2003-2009
Gothenburg,
Stockholm,
Sweden,
2001-2006
10 Italian cities
12% of
population
2001-2005
Study
Villeneuve
etal. (2012)
Ljunqman
et al. (2008)
Chiusolo et al.
(2011)
Long-term exposure
Hypertension
n = 575
Hypertension
n = 38
Hypercholes-
terolemia
n=230
Hypercholes-
terolemia
n = 19
Pre-existing
CVD
n=269
Hypertension
n = 867
Hypertension
n = 19.4%
No
hypertension
n = 1,409
No
hypertension
n = 104
No
hypercholes-
terolemia
n = 1,754
No
hypercholes-
terolemia
n = 123
No
pre-existing
CVD
n = 3,431
No
hypertension
n = 669
No hyper-
tension
n = 80.6%
_ Incident stroke
_ Fatal stroke
Incident stroke
(confirmed by
hospital
admission)
_ Fatal stroke
(confirmed by
hospital
admission)
y Systolic/
' diastolic blood
pressure
Blood IL-6
levels
_ Diabetes
N = 1,984
Ages 50-65 yr at
baseline
N = 142
Ages 50-65 yr at
baseline
N = 1,984
Ages 50-65 yr at
baseline
N = 142
Ages 50-65 yr at
baseline
N = 3,700
Ages 35-83 yr
N = 1,536
Ages 45-70 yr
N= 6,392
Ages 29-73 yr
Copenhagen,
Aarhus
counties,
Denmark,
iyyo ZUUD
Copenhagen,
Aarhus
counties,
Denmark,
-1993-2006
Girona, Spain
Stockholm
county,
Sweden,
1992-1994
Switzerland,
2002
Andersen
etal. (2012a)
Andersen
etal. (2012a)
Foraster et al.
(2014)
Panasevich
et al. (2009)
Eze et al.
(2014)
aUp facing arrow indicates that the effect of NO2 is greater (e.g., larger FEVi decrement, larger increase in airway inflammation) in the
group with the factor evaluated than in the reference group. Down facing arrow indicates that the effect of NO2 is smaller in the group
with the factor evaluated than in the reference group. A dash indicates no difference in NO2-related health effect between groups.
January 2015
7-14
DRAFT: Do Not Cite or Quote
-------
Table 7-8 Controlled human exposure and toxicological studies informing
pre-existing cardiovascular disease.
Factor Reference
Evaluated Category
Stable None
coronary
heart disease
or impaired
left ventricular
systolic
function
Athero- None
sclerosis
Direction of
Effect3 Outcome
_ Heart rate,
HRV
decrement
* Oxi dative
' stress in heart
tissue
Study
Population/
Animal Model
N = 18 humans
Mean age 68 yr
N = 5-10
mice/group,
ApoE-'-
Study Details
400 ppb NO2 for 1 h
200 or 2, 000 ppb
NO2, 6 h/day, 7 days
High fat diet
Study
Scaife et al.
(2012)
Campen et al.
(2010)
aThese studies only examined subjects with cardiovascular disease and have no reference group. A dash indicates that NO2 was not
observed to induce an effect in the group with cardiovascular disease evaluated relative to clean air exposure. An up-facing arrow
indicates that NO2 induced an effect on the outcome (e.g., cause a decrement in HRV) in the group with cardiovascular disease.
7.3.4 Diabetes
1 Diabetes mellitus is a group of diseases characterized by high blood glucose levels that
2 result from defects in the body's ability to produce and/or use insulin. An estimated
3 20 million Americans had diagnosed diabetes mellitus in 2010, representing 9.1% of the
4 adult population (Schiller et al.. 2012). High blood glucose levels can damage blood
5 vessels, increasing the risk of people with diabetes for heart disease or stroke. Diabetes
6 and cardiovascular disease are also linked by common risk factors such as hypertension
7 and obesity. These relationships provide support for diabetes influencing the risk of
8 cardiovascular disease; however, diabetes has not consistently been observed to modify
9 epidemiologic associations in studies of short-term or long-term NCh exposure and
10 cardiovascular effects (Table 7-9). No difference by diabetes was observed in studies of
11 short-term NC>2 exposure with hospital admissions or ED visits for IHD or MI (Tsai et al..
12 2012; Filho et al.. 2008) or of long-term NCh exposure with stroke (Andersen etal..
13 2012a). Diabetes also did not clearly modify associations of short-term or long-term NC>2
14 exposure with the subclinical effects heart rate variability, ventricular tachyarrhythmia,
15 blood pressure, and blood IL-6 levels (Foraster et al., 2014; Huang etal., 2012;
16 Panasevich et al.. 2009; Ljungman et al.. 2008). Associations of short-term and long-term
17 NO2 exposure with total mortality also did not consistently differ between people with
18 and without diabetes (Faustini et al.. 2013; Chiusolo et al.. 2011; Maheswaran et al..
19 2010).
January 2015
7-15
DRAFT: Do Not Cite or Quote
-------
1
2
o
6
4
5
6
7
Across studies there was not evidence of a consistent difference in associations of NO2
exposure with cardiovascular effects between people with and without diabetes.
Additionally, there was no consistent pattern observed for short-term or long-term
exposure or for a particular cardiovascular endpoint (Table 7-9). In conclusion, the large
evidence base lacks sufficient consistency in demonstrating that diabetes modifies
NO2-related cardiovascular effects along with evidence demonstration an independent
effect of NO2 on cardiovascular outcomes (Sections 5.3.12 and 6.3.9). As a result, the
evidence is inadequate to determine whether people with diabetes are at increased risk for
NO2-related health effects.
Table 7-9 Epidemiologic studies evaluating
Factor
Evaluated
Short-term
Diabetes
29.6%
admission
Diabetes
n = 700 ED
visits
Diabetes
n = 9
Diabetes
n = 12
Diabetes
n = 30,260
Diabetes
n = 11.3%
admission
Long-term
Diabetes
n = 97
Diabetes
n = 9
Direction of
Reference Effect
Category Modification3
exposure
No diabetes _
70.4%
admission
No diabetes *
n= 44,300 '
ED visits
No diabetes i
n = 31 ^
No diabetes _
n = 76
No diabetes y
n = 245,945 '
No diabetes _
n = 88.7%
admission
Exposure
No diabetes _
n = 1,887
No diabetes _
n = 133
Outcome
Hospital
admission for Ml
ED visits for
hypertension and
cardiac ischemic
disease
HRV decrements
Ventricular
tachy-arrhythmia
Total mortality
Respiratory
hospital
admissions
Incident stroke
Fatal stoke
pre-existing
Study
Population
N= 27,563
hospital
admissions
N= 45,000
ED visits
N = 40 with CVD
Mean age 66 yr
N = 88 with
implantable
cardioverter
defibrillators
Ages 28-85 yr
N = 276,205
natural deaths
N = 100,690
hospital
admissions
N = 1,984
Ages 50-65 yr at
baseline
N = 142
Age 50-65 yr at
baseline
diabetes.
Study Details
Taipei, Taiwan
1999-2009
Sao Paulo
Hospital,
January 2001-
July2003
Beijing, China,
2008
Gothenburg,
Stockholm,
Sweden,
2001-2006
10 Italian cities
2001-2005
6 Italian cities,
2001-2005
Copenhagen,
Aarhus counties,
Denmark,
1993-2006
Study
Tsai et al.
(2012)
Filho et al.
(2008)
Huanq et al.
(2012)
Ljunqman
et al. (2008)
Chiusolo et al.
(2011)
Faustini et al.
(2013)
Andersen
etal. (2012a)
January 2015
7-16
DRAFT: Do Not Cite or Quote
-------
Table 7-9 (Continued): Epidemiologic studies evaluating pre-existing diabetes.
Factor
Evaluated
Diabetes
n = 580
Diabetes
n = 121
Diabetes
n = 315
Diabetes
n = 1,045
Reference
Category
No diabetes
n = 3,120
No diabetes
n = 1,415
No diabetes
n = 1,541
No diabetes
n = 12,399
Direction of
Effect
Modification3
-
1
1
-
Outcome
Systolic blood
pressure
Diastolic blood
pressure
Relative IL-6
levels
Total mortality
Lung cancer
mortality
Study
Population
N = 3,700
Ages 35-83 yr
N = 1,536
Ages 45-70 yr
N = 3,320
Mean age 70 yr
N = 13,444
Mean age 74 yr
Study Details
Girona, Spain
Stockholm
county, Sweden
1992-1994
London, England
Follow-up:
1995-2005
NO2 assessed
for 2002
Shizuok, Japan
1999-2006
Study
Foraster et al.
(2014)
Panasevich
et al. (2009)
Maheswaran
etal. (2010)
Yorifuii et al.
(2010)
aUp facing arrow indicates that the effect of NO2 is greater (e.g., larger risk of ED visit, larger decrement in HRV) in the group with the
factor evaluated than in the reference group. Down facing arrow indicates that the effect of NO2 is smaller in the group with the factor
evaluated than in the reference group. A dash indicates no difference in NO2-related health effect between groups.
7.3.5 Obesity
1 Obesity can be defined as a body mass index (BMI) of 30 kg/m2 or greater. It is a public
2 health issue of increasing importance as obesity rates in adults have continually increased
3 over several decades in the U.S., reaching an estimated 27% in 2010 (Schiller et al..
4 2012). Obesity or high BMI could increase the risk of NO2-related health effects through
5 multiple mechanisms including persistent, low-grade inflammation. Obesity often occurs
6 with poor diet and chronic diseases. As a result, the combination of these factors could be
7 part of the pathway by which obesity increases the risk of NO2-related health effects or
8 they could act in combination with obesity to increase risk.
9 The 2008 ISA for Oxides of Nitrogen (U.S. EPA. 2008) did not evaluate obesity as a
10 potential factor that could increase the risk of NO2-related health effects. More recently
11 studies have included obesity or BMI as a potential effect measure modifier, but overall
12 the evidence is inconsistent as to whether obesity leads to increased risk of NO2-related
13 health effects. Most studies examined whether obesity increases the risk of NO2-related
14 cardiovascular and related metabolic effects. Obesity has been shown to be a risk factor
15 for both cardiovascular disease and diabetes, providing biological plausibility for the
16 potential increased risk of both diseasse in response to NO2 exposure. A study in rats
17 provides evidence that long-term NO2 exposure has larger effects on dyslipidemia, a
18 known risk factor for cardiovascular disease, in obese rats compared to nonobese rats
January 2015
7-17
DRAFT: Do Not Cite or Quote
-------
1 [(Takano et al.. 2004); Table 7-101. However, differences between obese and nonobese
2 strains were limited to 160 ppb NC>2 and not observed at higher NC>2 exposures.
3 The epidemiologic evidence does not provide coherence with the results from
4 toxicological studies. There is some indication of larger NCh-associated cardiovascular or
5 diabetes-related mortality in obese groups (Raaschou-Nielsen et al.. 2012). However,
6 most studies did not provide evidence that associations between long-term NCh exposure
7 and cardiovascular disease or diabetes morbidity differed between obese and nonobese
8 groups [(Ezeetal.. 2014; Atkinson etal.. 2013; Hart etal.. 2013; Mobasher etal.. 2013;
9 Andersen et al.. 2012b; Andersen et al.. 2012a); Table 7-23]. Most studies used a similar
10 definition of obese, i.e., BMI > 30 kg/m2. The limited number of studies that examined
11 NO2 assocaitions with mortality from respiratory causes or lung cancer also did not
12 provide any evidence of differences by obesity (Dimakopoulou et al.. 2014; Yorifuji
13 etal.. 2010). Studies that examined short-term and long-term NC>2 exposure with
14 subclinical cardiovascular effects found that most associations did not differ between
15 obese and nonobese groups of people (Dadvand et al.. 2014; Huang etal.. 2012; Baja
16 etal.. 2010; Ljungman etal.. 2008).
17 Toxicological evidence from a single study in rats demonstrated that obesity increases
18 dyslipidemia in relation to long-term NC>2 exposure. However, the majority of
19 epidemiologic studies reported that associations of long-term NO2 exposure with
20 cardiovascular diseases, events, and mortality as well as diabetes mostly did not differ
21 between obese and nonobese adults. In addition to the limited evidence indicating that
22 obese people may be at increased risk of NCh-related health effects uncertainties remain
23 in the overall body of evidence regarding the independent effects of NO2 on
24 cardiovascular and related metabolic effects (Section 6.3.9) and mortality (Section 6.5.3).
25 Therefore, the evidence is inadequate to determine whether obese individuals are at
26 increased risk for NC>2-related health effects.
Table 7-10 Toxicological study evaluating
Factor
Evaluated
Obesity
n = 9-13
Reference
Category
No obesity
n = 10-14
Direction of
Effect
Modification3
t
Outcome
Triglycerides,
HDL, total
cholesterol,
blood sugar
pre-existing obesity.
Animal Model Study Details Study
Rats (OLETF and Takano et al.
LETO) (2004)
N = 10-14
males/group
aUp facing arrow indicates that the effect of NO2 is greater (e.g., larger increase in triglycerides) in the group with the factor evaluated
than in the reference group.
January 2015 7-18 DRAFT: Do Not Cite or Quote
-------
Table 7-11
Factor
Evaluated
Epidemiologic studies evaluating
Direction of
Reference Effect
Category Modification3
Outcome
pre-existing
Study
Population
obesity.
Study Details
Study
Short-term exposure
HighBMI
(>27)
n=29
HighBMI
(>25)
n = 16
HighBMI
(>30)
n = 27.6%
Low BMI _
(<27)
n=69
Low BMI _
(<25)
n=24
Low BMI *
(<30) '
n = 72.4%
Ventricular
arrhythmia
HRV decrements
Change in
ventricular
repolarization
N = 98 with
implantable
cardioverter
defibrillators
Age 28-85 yr
N = 40
nonsmoking
adults with CVD
Mean age 66 yr
N = 580 males
Mean age 75 yr
Gothenburg,
Stockholm,
Sweden,
2001-2006
Beijing, China,
2007-2008
Boston, MA
area, Follow-up:
2000-2008
Ljunqman
et al. (2008)
Huana et al.
(2012)
Baia et al.
(2010)
Long-term exposure
HighBMI
(>30)b
HighBMI
(>30)
n = 109,104
HighBMI
(>30)
n = 84
HighBMI
(>30)
n = 366
HighBMI
(>30)
n = 19
High
(>25-<30)
and very
highBMI
(>30)
n= 28,937
Low BMI _
(<30)b
Low BMI _
(25-30)
n = 243,556
Low BMI
«so)
n = 158 _
-
1
-
-
Low BMI _
(<30)
n = 1,618
Low BMI _
(<30)
n = 123
Low BMI _
(<25)
n= 22,881
Incidence Ml
Heart failure
C-reactive protein
Tumor necrosis
factor (TNF)-a
IL-6
IL-8
Fibrinogen
Hepatocyte
Growth Factor
Incident stroke
Fatal stroke
Diabetes
N = 84,562
Ages 30-55 yr at
enrollment
N = 836,557
Ages 40-89 yr in
2003
N = 242 adults
with clinically
stable COPD
Mean age 68 yr
N = 1,984
Ages 50-65 yr at
baseline
N = 142
Ages 50-65 yr at
baseline
N = 51,818
Ages 50-65 yr at
baseline
U.S.,
1990-2008
England,
2003-2007
Barcelona,
Spain,
2004-2006
Copenhagen,
Aarhus counties,
Denmark,
1993-2006
Copenhagen,
Aarhus counties,
Denmark,
1993-2006
Hartetal.
(2013)
Atkinson et al.
(2013)
Dadvand et al.
(2014)
Andersen
etal. (2012a)
Andersen
etal. (2012b)
January 2015
7-19
DRAFT: Do Not Cite or Quote
-------
Table 7-11 (Continued): Epidemiologic studies evaluating pre-existing obesity.
Factor
Evaluated
High waist-
to-hip ratio
(>0.90
male, >0.85
female)
n= 26,183
HighBMIb
(>30)
HighBMI
(>30)
n=68
HighBMIb
(>30)
HighBMI
(>25)
n = 1,950
HighBMI
(>30)
n = 96,076
person-
years
HighBMI
(>21.8)b
Reference
Category
Low waist-
to-hip ratio
(<0.90 male,
<0.85
female)
n= 25,635
LowBMIb
(<30)
Low BMI
(<30)
n=213
LowBMIb
(<30)
Low BMI
(<18.5)
n = 1,010
Low BMI
(<25)
n= 298,503
person-years
Low BMI
(<21.8)b
Direction of
Effect
Modification3 Outcome
t
Diabetes
_ Hypertensive
disorders of
pregnancy
i Respiratory
^ mortality
_ CVD mortality
* Diabetes-related
' mortality
Lung cancer
mortality
Study
Population
N = 6,392
Ages 29-73 yr
N = 298
predominantly
Hispanic women
N = 307,553
Mean age at
baseline 41.9 to
73.0 yr across
16 cohorts
N = 9,941, 256
deaths
Ages 35-103 yr
N = 52,061
Ages 50-64 yr
N = 14,001
Ages >65 yr
Study Details
Switzerland,
2002
Los Angeles,
CA,
1996-2008
Europe
Follow-up:
1985-2007
NO2 exposure
assessed for
2008-2011
Shenyang,
China
Follow-up:
1998-2009
NO2 exposure
assessed for
1998-2009
Denmark
Follow-up:
1971-2009
NO2 exposure
assessed for
1971-2009
Shizuoka,
Japan,
1999-2006
Study
Eze et al.
(2014)
Mobasher
etal. (2013)
Dimakopoulou
etal. (2014)
Zhanq et al.
(2011)
Raaschou-
Nielsen et al.
(2012)
Yorifuii et al.
(2010)
aUp facing arrow indicates that the effect of NO2 is greater (e.g., larger change in ventricular repolarization) in the group with the
factor evaluated than in the reference group. Down facing arrow indicates that the effect of NO2 is smaller in the group with the
evaluated factor than in the reference group. A dash indicates no difference in NO2-related health effect between groups.
bSample size not reported.
7.4
Genetic Factors
i
2
3
4
5
Genetic variation in the human population is known to contribute to numerous diseases
and differential physiologic responses. The 2008 ISA for Oxides of Nitrogen (U.S. EPA.
2008) concluded that "it remains plausible that there are genetic factors that can influence
health response to NO2, through the few available studies did not provide specific
support." Since then a growing body of studies have examined whether specific genetic
January 2015
7-20
DRAFT: Do Not Cite or Quote
-------
1 polymorphisms increase the risk of NCh-related health effects. A strength of these studies
2 is that they evaluate genetic factors using a targeted approach to focus on specific genes
3 that potentially are involved in signaling pathways involved in biological responses to
4 NC>2. For example, NCh exposure can lead to the formation of oxidation products
5 (Section 4.3.2.1) and also modulate immune function (Section 4.3.2.6), and studies
6 examined variants for genes encoding antioxidant enzymes [e.g., glutathione
7 S-transferases (GST)Ml and GSTP1]) and mediators of immune response [e.g., tumor
8 necrosis factor-alpha (TNF-a)]. A potential limitation for drawing conclusions about
9 genetic variants is the large number of variants examined within studies which increases
10 the probability of finding associations by chance alone. Thus, consistency in findings
11 across genetic variants is considered. Further, the functional or biological consequence of
12 some of the gene variants is unknown, and some variants may be surrogates for another
13 linked gene or a group of related genes. Thus, where available, the variant effect is
14 described (Table 7-12) and considered in conclusions. With in this section gene variants
15 have been examined primarily for NCh-associated respiratory effects, with a few studies
16 examining cardiovascular effects and cognitive function. Because evidence supporting an
17 independent effect of NO2 exposure is strongest for respiratory effects (causal for
18 short-term NC>2 exposure, Section 5.2.9) and there is uncertainty for relationships with
19 other health effects, conclusions for genetic variants emphasize respiratory effects
20 evidence.
21 Oxidative stress has been described as a key process underlying the respiratory effects
22 attributed to NC>2 exposure (Section 4.3.2.1); however, studies that examined functional
23 variants of GSTM1 or GSTP1 estimated similar (Castro-Giner et al.. 2009) or lower
24 (Romieu et al.. 2006) effects of short-term or long-term NCh exposure on asthma
25 symptoms or asthma prevalence in groups with variants encoding enzymes with null or
26 reduced oxidative metabolizing activity (Table 7-12). These variants are common in the
27 population, and NC>2-related health effects were compared between groups with fairly
28 similar numbers of people (Table 7-12). Genetic variants with the potential for elevated
29 oxidative stress have been observed to increase NCh-related subclinical cardiovascular
30 and metabolic effects [(Kim and Hong. 2012; Bajaetal.. 2010): Table 7-12]. A strength
31 of the studies of cardiovascular effects is that rather than performing multiple
32 comparisons of individual variants, genetic variants were analyzed as a cumulative index
33 of oxidative stress potential, either the number of variants with increased oxidative stress
34 potential (Bajaetal.. 2010). Further, independent relationships between short-term NC>2
35 exposure and cardiovascular and metabolic effects are uncertain. Thus, it is not clear the
36 extent to which the findings for modification by gene variants can be attributed to NC>2
37 specifically.
January 2015 7-21 DRAFT: Do Not Cite or Quote
-------
1 Similar to variants with known functional differences, no clear evidence exists that
2 genetic variants in oxidative metabolism enzymes with unknown functional differences
3 increase the risk of NC>2-related respiratory effects. Associations of long-term NC>2
4 exposure with decrements in lung development in children were larger for some variants
5 in glutathione metabolism pathway genes, such as glutathione synthetase [GSS; (Breton
6 etal.. 2011)]. However, results were not consistent across the multiple gene variants of
7 glutathione examined [glutathione reductase (GSR), glutamate-cysteine ligase, modifier
8 subunit (GCLM), glutamate-cysteine ligase, catalytic subunit (GCLC)] or
9 NADPH-quinone oxidoreductase (NQO1) (rs 10517). Variants for NQO1 (Castro-Giner
10 et al.. 2009) were observed to increase associations with asthma prevalence in adults.
11 However, NQO1 was one among many variants examined.
12 Mediators of immune response including TNF-a and toll-like receptor (TLR)4 are known
13 to have a role in oxidant-induced inflammation and asthma pathogenesis, but variants in
14 these genes were not observed to modify associations of long-term NO2 exposure with
15 asthma prevalence [(Castro-Giner et al.. 2009): Table 7-12]. Variants in immune response
16 genes also did not modify associations of long-term NC>2 exposure with myocardial
17 infarction (Panasevich et al.. 2013).
18 The beta-2-adrenergic receptor (ADRB2) is an encoded G protein-coupled receptor that
19 plays an important role in regulation of airway smooth muscle tone and is the
20 pharmacological target of beta-agonist asthma medications (Hizawa. 2011). NCh
21 exposure has been shown to increase airway responsiveness in adults with asthma
22 (Section 5.2.2.1), providing a plausible role for variants in ADRB2 in modifying the risk
23 of NO2-associated respiratory effects. The association between ambient NO2 estimated
24 for residential locations and asthma prevalence in adults was not modified by ADRB2
25 variants with unknown functional differences (Castro-Giner et al.. 2009). However,
26 higher methylation of the ADRB2 promoter, which is associated with reduced expression
27 of the receptor, was observed to increase the risk of asthma severity in children
28 associated with indoor residential NCh exposure (Fu etal.. 2012). There is mixed
29 evidence for beta-agonist medication use in modifying NC^-associated respiratory effects
30 (Section 5.2.2.2), and it is not known whether the response to beta-agonists is influenced
31 by genetic variants in ADRB2.
32 There is evidence for independent relationships of short-term and long-term NC>2
33 exposure, respectively, with exacerbation (Section 5.2.9) and development
34 (Section 6.2.9) of asthma, and antioxidant modulation, immune-mediated inflammation,
35 and airway responsiveness are described as key events in the underlying modes of action
36 (Section 4.3.5). Evidence in rodents that dietary antioxidants modify NC>2-induced
37 pulmonary oxidative stress (Section 7.6.1) would suggest a role for variants in oxidative
January 2015 7-22 DRAFT: Do Not Cite or Quote
-------
1
2
o
6
4
5
6
7
8
9
10
11
12
13
14
15
Table 7-12
metabolism genes in modifying NCh-related respiratory effects. However, gene variants
with greater potential for oxidative stress were not observed to modify associations of
short-term NC>2 with asthma-related effects. Variants in antioxidant and immune-related
genes did modify some associations of long-term NC>2 exposure with respiratory effects,
but results are inconsistent for any particular gene variant or respiratory effect. Further,
several results are based on multiple comparisons and post-hoc analyses. There is more
limited but consistent evidence for gene variants for antioxidant enzymes and
inflammatory cytokines modifying subclinical cardiovascular and related metabolic
effects but not myocardial infarction. Overall, the findings for effect measure
modification by genetic variants are inconsistent for respiratory effects and the
interpretation of results from studies of cardiovascular and related metabolic effects is
complicated by the uncertainty as to whether the effects can be attributed specifically to
NC>2 exposure. Therefore, the evidence is inadequate to determine whether genetic
variants, particularly for antioxidant enzymes and immune responses, increase the risk for
NO2-related health effects.
Epidemiologic studies evaluating genetic factors.
Factor Direction of
Evaluated/Gene Reference Effect Study Location
Function Category Modification3 Outcome and Population Study
GSTM1 null,
n = 58
Null oxidant
metabolizing
capacity
GSTM1 null
n = 49%
Null oxidant
metabolizing
capacity
GSTM1 null,
n=299
Null oxidant
metabolizing
capacity
GSTT1 null,
n=270
Null oxidant
metabolizing
capacity
GSTM1 positive, i Respiratory Mexico City, Romieu et al.
n = 93 "*" symptoms and Mexico (2006)
medication use N = 151 children Short-term NO2
with asthma
Ages not reported
GSTM1 positive, _ Asthma Umea and Uppsala, Castro-Giner
n = 51% prevalence Sweden; Ipswich et al. (2009)
and Norwich, U.K.;
Albacete, Long-term NO2
Barcelona, Huelva,
Galdakao, Oviedo,
Spain; Erfurt,
Germany; Paris,
Grenoble, France;
Antwerp, Belgium
N= 2,920
Mean age 43 yr
GSTM1 positive, f Fastinq qlucose Seoul, Korea Kim and Honq
n - 225 ' N - 560 (2012)
Ageb 60 87 yi
| Insulin levels Short-term NO2
GSTT1 positive, * Fasting glucose
n = 254 !
-j- Insulin levels
January 2015
7-23
DRAFT: Do Not Cite or Quote
-------
Table 7-12 (Continued) Epidemiologic studies evaluating genetic variants.
Factor
Evaluated/Gene
Function
GSTP1 VaMOSVal
orlle105Val
(rs1695), n = 179
Reduced oxidant
metabolizing
capacity (Val/Val)
GSTP1 VaMOSVal,
n = 54
(rsID 947894)
Reduced oxidant
metabolizing
capacity
GSTP1 VaMOSVal
(rsID 1695)
n = 32%
Reduced oxidant
metabolizing
capacity
>4 variants with
increased oxidative
stress potential13
(GSTT1, GSTP1,
GSTM1, HMOX,
NQO1, HFE)
GSTP1 VaMOSVal
orlle105Val
n = 198
Reduced oxidant
metabolizing
capacity
GSS haplotype
0100000,
n = 1,010
(rs1801310)
Unknown function
GSR, various SNPs
n = 3-21%
Unknown function
GCLM, various
SNPs,
n = 6-35%
Unknown function
GCLC, various
SNPs,
n = 4-54%
Unknown function
NQO1 CC
(rs2917666)
n = 32%
Unknown function
Direction of
Reference Effect
Category Modification3 Outcome
GSTP1 Ile105lle,
n = 359
Ile105lle,
n - 97
GSTP1 Ile105lle
orlle105Val
n = 68%
<4 variants with
increased
oxidative stress
potential13
GSTP1 lie/lie
n = 152
Other
haplotypes,
n = 1,096
Other haplotypes
n = 3-21%
Other
haplotypes,
n = 6-35%
Other
haplotypes,
n = 4-54%
NQO1 GC or
GG,
n = 68%
* Fasting glucose
_ Insulin levels
i Respiratory
^ symptoms and
medication use
_ Asthma
prevalence
* Heart rate-
' corrected QT
interval
(ventricular
repolarization)
* Cognitive
' function
decrements
* Lung
' development
decrements
—
"
^™
* Asthma
' prevalence
Study Location
and Population
Mexico City,
Mexico
N = 151 children
with asthma
Ages not reported
Multiple European
countries (see
above)
N= 2,920
Mean age 43 yr
Boston, MA area
N = 580 males
Mean age 75 yr
Menorca, Spain
N = 350 children
followed from birth
to age 4 yr
Alpine, Atascadero,
Lake Elsinore, Lake
Arrowhead,
Lancaster, Lompoc,
Long Beach, Mira
Loma, Riverside,
San Dimas, Santa
Maria, Upland, CA
N= 2, 106 children
followed ages
10-18 yr
Multiple European
countries (see
above)
N= 2,920
Study
Romieu et al.
(2006)
Short-term NO2
Castro-Giner
et al. (2009)
Long-term NO2
Baiaetal. (2010)
Morales et al.
(2009)
Long-term indoor
NO2
Breton et al.
(2011)
Long-term NO2
Castro-Giner
et al. (2009)
Long-term NO2
January 2015
7-24
DRAFT: Do Not Cite or Quote
-------
Table 7-12 (Continued) Epidemiologic studies evaluating genetic variants.
Factor
Evaluated/Gene
Function
TLR4 GG
(rs1 1536889)
n = 14%
Unknown function
TNF-a 308 GA/AA
(rs1 800629)
n = 16%
Increased
expression
TNF-a 308 GA/AA
Increased
expression
n = 17%
IL-6 174CC
n = 48%
Increased blood
IL-6 levels
IL-6 598 AA
n = 47%
Increased blood
IL-6 levels
ADRB2b
Intermediate or
high methylation
levels
Reduced
expression
ADRB2
rs 104271 3 G/G
rs 104271 4 C/C
rs 104271 8 C/C
rs 10427 19 G/G
n = 18-40%
Unknown function
MET Tyrosine
receptor kinase CC
(rs1 858830)
n = 102
Reference
Category
TLR4 GC or CC
n = 86%
TNF-a 308 GG
n = 84%
TNF-a 308 GG
n = 83%
IL-6174GG
n = 52%
IL-6 598 GG
n = 53%
ADRB2b
Low methylation
levels
G/AorA/A
C/G or G/G
C/A or A/A
G/C or C/C
n = 60-82%
MET Tyrosine
receptor kinase
CG/GG
n = 305
Direction of
Effect
Modification3 Outcome
t
_ Asthma
prevalence
_ Myocardial
infarction
^™
"
* Asthma severity
_ Asthma
prevalence
y Autism
Study Location
and Population
Mean age 43 yr
Multiple European
countries (see
above)
N= 2,920
Mean age 43 yr
Stockholm County,
Sweden
N= 2,698
Ages 45-70 yr
CT and Springfield,
Worcester, MA
N = 182
Ages 5-12 yr,
followed for 1 yr
Multiple European
countries (see
above)
N= 2,920
Mean age 43 yr
Multiple unspecified
locations, California
N = 252 with
autism, 156 without
Ages 2-5 yr
Study
Castro-Giner
et al. (2009)
Long-term NO2
Panasevich et al.
(2013)
Long-term NO2
Fuetal. (2012)
Long-term indoor
NO2
Castro-Giner
et al. (2009)
Long-term NO2
Volketal. (2014)
aUp facing arrow indicates that the effect of NO2 is greater (e.g., larger increase in symptoms) in the group with the factor
evaluated than in the reference group. Down facing arrow indicates that the effect of NO2 is smaller (e.g., smaller increase in
symptoms) in the group with the factor evaluated than in the reference group. A dash indicates no difference in NO2-related health
effect between groups.
bsample size not reported
January 2015
7-25
DRAFT: Do Not Cite or Quote
-------
7.5 Sociodemographic Factors
7.5.1 Lifestage
1 Lifestage refers to a distinguishable time frame in an individual's life characterized by
2 unique and relatively stable behavioral and/or physiological characteristics that are
3 associated with development and growth (U.S. EPA. 2014). The 2008 ISA for Oxides of
4 Nitrogen (U.S. EPA. 2008) indicated there was supporting evidence for increased risk of
5 health effects related to NO2 exposure among different lifestages, i.e., children and older
6 adults. Differential health effects of NO2 across lifestages theoretically could be due to
7 several factors:
6) The human respiratory system is not fully developed until 18-20 years of age,
and therefore, it is plausible to consider children to have intrinsic risk for
respiratory effects due to potential perturbations in normal lung development.
7) Older adults (typically considered those 65 years of age or greater) have
weakened immune function, impaired healing, decrements in pulmonary and
cardiovascular function, and greater prevalence of chronic disease (Table 7-2).
8) Exposure/internal dose of NO2 may vary across lifestages due to varying
ventilation and time-activity patterns.
8 Studies in this ISA add to the evidence presented in the 2008 ISA indicating increased
9 risk of NO2-related health effects for children and older adults. Further, this evaluation of
10 lifestage as a factor that may lead to increased risk for NO2-related health effects draws
11 upon information about time activity patterns and ventilation patterns among different
12 lifestages to assess the potential for differences in NO2 exposure or internal dose among
13 lifestages.
7.5.1.1 Children
14 According to the 2010 census, 24% of the U.S. population is less than 18 years of age,
15 with 6.5% less than age 6 (Howden and Meyer. 2011). The large proportion of children
16 within the U.S. supports the public health significance of characterizing the risk of
17 NO2-related health effects among children.
18 In evaluating risk for NO2-related health effects in children, it is important to consider
19 exposure or dose differences; however, these are not well characterized. Children and
20 adults differ with respect to time-activity patterns, which are determinants of
21 inter-individual variability in NO2 exposure [(Molter et al., 2012; Kousaetal.. 2001) and
January 2015 7-26 DRAFT: Do Not Cite or Quote
-------
1 Section 3.4.41. In a comparison of children (mostly less than age 8 years), adults mostly
2 under age 55 years, and adults older than age 55 years, a larger proportion of children
3 reported spending over 30 minutes performing vigorous outdoor physical activity (Wu
4 etal.. 2011). However, there was little variation among groups in time spent in various
5 microenvironments (Wu etal.. 2011). Although a recent meta-analysis suggested a
6 weaker association between ambient NO2 concentrations and personal NO2 exposure of
7 children (Meng etal.. 2012). some studies found ambient NC>2-related respiratory effects
8 in children for whom there were moderate personal-ambient correlations [r values of 0.43
9 and 0.63; (Delfino et al.. 2006; Linn etal.. 1996)1. Such personal-ambient NO2
10 relationships are consistent with greater time spent outdoor among children and could be
11 an explanation for larger risk of NC^-related health effects for children. A recent analysis
12 found children more likely than adults to take part in vigorous activity or aerobic exercise
13 [indoors and outdoors; (Wu etal.. 2011)1. Higher activity along with higher ventilation
14 rates relative to lung volume could potentially result in greater NC>2 penetration to the
15 lower respiratory tract of children; however, this has not been examined for NC>2
16 (Section 4.2.2.3).
17 Epidemiologic evidence across diverse locations including the U.S., Europe, Asia, and
18 Australia consistently demonstrates that short-term increases in ambient NC>2
19 concentration are associated with larger increases in asthma-related hospital admissions,
20 ED visits, or outpatient visits among children than adults [(Sonet al.. 2013; Sinclair et al..
21 2010; Ko et al.. 2007; Hinwood et al.. 2006; Peel etal.. 2005; Atkinson et al.. 1999;
22 Anderson et al.. 1998): Table 7-13]. Most results are based on comparisons between
23 children ages 0-14 years and people ages 15-64 years, and these found that
24 NO2-associated increases in asthma hospital admissions were 1.8 to 3.4 fold greater in
25 children (Son etal.. 2013; Ko et al.. 2007; Atkinson et al.. 1999; Anderson et al.. 1998).
26 Not all results demonstrated increased risk for children, with some studies of asthma
27 hospital admissions, outpatient visits, and medication sales showing no difference in
28 association with NC>2 between children and adults or no association in either group (Burra
29 et al.. 2009; Laurent etal.. 2009; Migliaretti etal.. 2005; Petroeschevsky et al.. 2001). A
30 few results point to larger NC>2-related increases in asthma hospital admissions or ED
31 visits among younger children (e.g., age 0-4 years, 2-4 years) than older children ages
32 5-14 years (Samoli etal.. 2011; Villeneuve et al.. 2007); however, inference from these
33 findings is limited because of the questionable reliability of asthma diagnosis in children
34 below the age of 5 years (Section 5.2.2.4). Limited comparisons of lifestage in
35 toxicological studies do not indicate larger NC>2-related effects on lung injury,
36 inflammation, or lung host defense among juvenile than mature rodents (Azoulay-Dupuis
37 etal.. 1983) or between rodents with prenatal/weaning exposure and exposure only
38 during weaning [(Kumae and Arakawa. 2006); Table 7-14]. The endpoints examined in
January 2015 7-27 DRAFT: Do Not Cite or Quote
-------
1 experimental animals do not have direct coherence with asthma-related effects and are
2 not considered to be in conflict with epidemiologic evidence.
3 Risk may vary among children according to the time window of exposure because there
4 are differences in lung development over the course of childhood. In this ISA, critical
5 time windows of exposure for NC^-related health effects in children were assessed from
6 longitudinal studies that permitted within-subject comparisons as children were followed
7 over time. Across studies, respiratory effects were associated with long-term NCh
8 exposures assessed for various time windows, including birth, the first year of life, year
9 of asthma diagnosis, and lifetime exposure (Section 6.2.2.1). In limited comparisons of
10 time periods, no single critical time window of exposure was identified for the
11 association of short-term or long-term NC>2 exposure and asthma exacerbation or
12 diagnosis in children. In cohorts of children diagnosed with asthma at a median age of 2
13 or 5 years, NC>2 in the first year of life was associated with similar or lower risk of asthma
14 compared with NC>2 assessed for later in childhood [average of ages 1-3 years or average
15 in year of diagnosis; (Nishimura et al.. 2013; Clougherty et al.. 2007)]. The young age of
16 diagnosis in most of these children limits inference about critical time windows of NC>2
17 exposure. In the Children's Health Study (CHS) cohorts, both exposures and respiratory
18 outcomes were examined at various ages during follow-up from ages 5 or 10 years to
19 18 years. The heterogeneity among studies in exposure assessment methods, statistical
20 methods, and examination of incidence or prevalence of outcomes is not amenable to
21 quantitative comparisons. However, NCh exposure was associated with asthma and
22 respiratory symptoms in childhood (ages 9-13 or 10 years) and into adolescence [ages
23 13-16 years or 10-18 years; (Jerrett et al.. 2008; McConnell et al.. 2006; Gauderman
24 etal.. 2005; McConnell et al.. 2003; McConnell et al.. 1999)], also pointing to risk of
25 NO2-associated respiratory effects throughout childhood.
26 In conclusion, epidemiologic evidence generally demonstrates that NC>2-related asthma
27 exacerbation is greater in children compared to adults. In a few cases, no difference was
28 observed by age, i.e., for NCh-associated asthma outpatient visits and medication use.
29 However, there is sufficient consistency for asthma hospital admissions and ED visits and
30 for similar age comparisons (ages 0-14 years vs. 15-64 years). Limited toxicological
31 results suggest greater NO2-induced pulmonary injury and impaired host defense in
32 mature compared to juvenile animals, but these endpoints are not directly related to
33 asthma and are not considered to contradict epidemiologic evidence. As examined for
34 asthma incidence or prevalence in children, no single critical time window of exposure
35 (e.g., infancy, later childhood) has been identified. Children have different time-activity
36 and ventilation patterns than adults, but it is not clear whether these contribute to higher
37 NC>2 exposure or internal dose or increased risk for NC^-related asthma exacerbation in
38 children. Overall, the consistent epidemiologic evidence for larger NO2-related asthma
January 2015 7-28 DRAFT: Do Not Cite or Quote
-------
exacerbation is adequate to conclude that children are at increased risk for NCh-related
health effects.
Table 7-13
Factor
Evaluated
Age of health
Childhood
Ages 0-14 yr
n= 23,596
Childhood
Ages 0-14 yr
Childhood
Ages 0-4 yr
Childhood
Ages 2-4 yr
n = 7,247
Childhood
Ages not
reported
n= 28,487
Childhood
Ages 1-17 yr
Childhood
Ages 0-19 yr
n = 7,774
Time window
Exposure in
year of
diagnosis
Exposure in
first year of
life
Epidemiologic studies evaluating
Direction of
Reference Effect
Category Modification3
Outcome
childhood lifestage.
Study
Population
Study Details
Study
effect: short-term NO2 exposure
Adulthood
Ages 15-65 yr
n = 21,204
Adulthood
Ages 15-64 yr
Childhood
Ages 5-14 yr
Childhood
Ages 5-14 yr
n = 13,145
Adulthood
Ages not
reported
n = 19,085
Adulthood
Ages 18-64 yr
Adulthood
Ages 20-39 yr
n = 7,347
of exposure: long-term
Exposure in
first year of life
Exposure in
first 3 yr of life
t
t
t
t
t
—
—
exposure
t
Asthma
hospital
admissions
Asthma
hospital
admissions
Asthma
hospital
admissions
Asthma ED
visits
Asthma
outpatient
visits
Asthma
outpatient
visits
Asthma
medication
sales
Asthma
incidence
Median age of
diagnosis:
Syr
Asthma
prevalence
Median age of
diagnosis:
2yr
15 hospitals
N = 69,176
admissions
Database
accounting for
48% of Korean
population
3 children's
hospitals
5 hospitals
N = 57, 192 visits
Records from
managed care
organization for
270,000 people
N = 417 children
N = 4,320
children enrolled
between ages of
8 and 21 yr
Hong Kong,
2000-2005
8 South Korean
cities,
2003-2008
Athens, Greece,
2001-2004
Edmonton,
Canada,
1992-2002
Atlanta, GA,
1998-2002
Toronto,
Canada,
1992-2001
Strasbourg,
France, 2004
Boston, MA,
Follow up:
prenatally
(1987-1 993) to
1997
5 U.S. cities,
1996-2001
Central site NO2
Ko et al.
(2007)
Son et al.
(2013)
Samoli et al.
(2011)
Villeneuve
et al. (2007)
Sinclair et al.
(2010)
Burra et al.
(2009)
Laurent et al.
(2009)
Clouqhertv
et al. (2007)
Nishimura
etal. (2013)
aUp facing arrow indicates that the effect of NO2 is greater (e.g., larger increase in hospital admission) in the group with the factor
evaluated than in the reference group. Down facing arrow indicates that the effect of NO2 is smaller in the group with the factor
evaluated than in the reference group. A dash indicates no difference in NO2-related health effect between groups.
January 2015
7-29
DRAFT: Do Not Cite or Quote
-------
Table 7-14 Toxicological studies informing childhood lifestage.
Factor
Evaluated
Prenatal/
weanling
exposure
Juvenile
age
Juvenile
age
Juvenile
age
Juvenile
age
Direction of
Reference Effect
Category Modification3 Outcome
Weanling t Alveolar
exposure macrophage
activity
Adult age — Mortality
Adult age I
Adult age — Lung injury,
inflammation
Adult age I
Animal Model
Rats (Brown
Norway)
N = 5-7/group
Females
Rats (Wistar)
N = 5-8/group
Guinea pigs
(Hartley)
N = 5-8/group
Rats (Wistar)
N = 5-8/group
Guinea pigs
(Hartley)
N = 5-8/group
Study Details
Prenatal/weaning
exposure: Breeding pairs
mated in 200, 500, or
2,000 ppb NO2. Litters
continuously exposed until
8 and 12 weeks
Weanling exposure:
5 week old rats exposed
to 200, 500, or 2, 000 ppb
NO2 continuously until 8
and 12 weeks
2,000 ppb for 3 days at 5,
10,21,45, 55, and
60 days of age
2,000 ppb for 3 days at 5,
10,21,45, 55, and
60 days of age
Study
Kumae and
Arakawa
(2006)
Azoulav-
Dupuis et al.
(1983)
aUp facing arrow indicates that the effect of NO2 is greater (e.g., greater increase in alveolar macrophage activity) in the group with the
factor evaluated than in the reference group. Down facing arrow indicates that the effect of NO2 is smaller in the group with the factor
evaluated than in the reference group. A dash indicates no difference in NO2-related health effect between groups.
1
2
o
5
4
5
6
7
8
9
10
11
12
7.5.1.2 Older Adults
According to the 2008 National Population Projections issued by the U.S. Census
Bureau, approximately 12.9% of the U.S. population is age 65 years or older, and by
2030, this fraction is estimated to grow to 20% (Vincent and Velkoff. 2010). Thus, this
lifestage represents a substantial proportion of the U.S. population that is potentially at
increased risk for health effects related to NO2 exposure.
The 2008 ISA for Oxides of Nitrogen (U.S. EPA. 2008) indicated that compared with
younger adults, older adults (ages 65 years and older) may be at increased risk for
NO2-related respiratory effects and morality but not cardiovascular effects. Recent
epidemiologic findings add to this body of evidence (Table 7-15). As described in the
preceding section, time-activity patterns were found to differ across age groups; however,
there were no differences in time spent in particular microenvironments or time in
vigorous or outdoor activity (Wu etal. 2011) that might inform differences in NC>2
January 2015
7-30
DRAFT: Do Not Cite or Quote
-------
1 exposure among older adults than younger adults. Comparisons of older and younger
2 adults with respect to NC>2-related asthma exacerbation are limited and generally show
3 larger (one- to three-fold) effects in adults ages 65 years or older than among individuals
4 ages 15-64 years or 15-65 years (Ko et al.. 2007: Villeneuve et al.. 2007: Migliaretti
5 etal.. 2005: Anderson et al.. 1998). A few studies showed no difference between older
6 and younger adults [(Son etal.. 2013: Hinwood et al.. 2006): Table 7-15]. Results for all
7 respiratory hospital admissions combined generally also showed larger associations with
8 NC>2 among older adults ages 65 years or older (Arbex et al.. 2009: Wong et al.. 2009:
9 Hinwood et al.. 2006: Atkinson et al.. 1999). Evidence for increased risk in older adults
10 for NO2-related respiratory outcomes is more limited in controlled human exposure
11 studies, where only statistically nonsignificant decrements in lung function were found
12 with NO2 exposure in healthy, older adults [(Gong et al.. 2005: Morrow et al.. 1992):
13 Table 7-161.
14 For nonrespiratory effects, associations of short-term NC>2 with total mortality in most
15 studies were larger in adults ages 65 or older than in younger adults Table 7-15). with
16 evidence pointing to elevated risk among the oldest adults ages greater than 75 or
17 85 years (Chen etal.. 2012: Cakmak etal.. 2011: Chiusolo etal.. 2011). Studies of
18 long-term NC>2 exposure do not provide strong evidence of elevated risk of health effects
19 among older adults, with inconsistent effect modification observed for total or
20 cause-specific mortality (Dimakopoulou et al.. 2014: Carey etal.. 2013: Cesaroni et al..
21 2013: Zhang etal..2011: Maheswaran et al.. 2010: Yorifuii etal.. 2010) and generally no
22 difference by age group observed for associations with cardiovascular effects or diabetes
23 (Eze etal..2014: Atkinson et al.. 2013: Rivera et al.. 2013: Wichmann et al.. 2013:
24 Andersen etal.. 2012a: Rosenlund et al.. 2009a: Ljungman et al.. 2008: Min et al.. 2008).
25 The age to define older adults varied among mortality and cardiovascular effect studies
26 from 50 to 75 years. Further, it is uncertain the extent to which the inconsistent findings
27 can be attributable to NO2 because of uncertainty in whether NO2 has independent
28 relationships with cardiovascular effects (Sections 5.3.12 and 6.3.9) and mortality
29 (Sections 5.4.8 and 6.5.3).
30 There is substantial and consistent evidence for larger NCh-related respiratory effects in
31 older adults compared to younger adults. Such evidence is based on hospital admissions
32 and ED for asthma in a few cases, all respiratory conditions combined in several cases,
33 and comparisons between adults ages 65 years or older and individuals 15-64 years of
34 age. Controlled human exposure studies do not indicate NC>2-induced respiratory effects
35 in older, healthy adults, but evidence is far more limited compared with epidemiologic
36 evidence. Older adults had larger NO2-related increases in total mortality but not
37 cardiovascular effects. However, inferences about the risk for older adults from the total
38 mortality and cardiovascular effects evidence is limited because of uncertainties
January 2015 7-31 DRAFT: Do Not Cite or Quote
-------
1
2
o
6
4
regarding the independent effect of NO2 on those outcomes. Overall, the consistent
epidemiologic evidence for larger NCh-related asthma and all respiratory hospital
admissions and ED visits is adequate to conclude that older adults are at increased risk
for NO2-related health effects.
Table 7-15 Epidemiologic studies evaluating older adult lifestage.
Factor
Evaluated
Direction of
Reference Effect
Category Modification3 Outcome
Study
Population
Study Details
Study
Short-term exposure
Older
adulthood
Ages >65 yr
n= 24,916
Older
adulthood
Ages >65 yr
n= 4,705
Older
adulthood
Ages >65 yr
Older
adulthood
Ages >65 yr
Older
adulthood
Ages >64 yr,
n = 789
Older
adulthood
Ages >65 yrb
Younger
adulthood
Ages 15-65 yr
n=21,204
Younger
adulthood
Ages 15-64 yr
n = 32,815
Younger
adulthood
Ages 15-64 yr
All ages
Younger
adulthood
Ages 40-64 yr,
n = 980
Younger
adulthood,
childhood Ages
5-64 yrb
t
t
-
t
"
t
t
t
Asthma hospital
admissions
Asthma ED
visits
Asthma and
allergic disease
hospital
admissions
COPD hospital
admissions
COPD hospital
admissions with
influenza
Acute
respiratory
disease hospital
admissions
Cardiovascular
hospital
admissions
COPD ED visits
Total mortality
15 hospitals
N = 69,176
admissions
5 hospitals
N = 57,912 visits
Hospital
admission
database
accounting for
48% of Korean
population
14 hospitals
40 hospitals
Data from
Municipal Center
for Disease
Control and
Prevention
Hong Kong,
2000-2005
Edmonton,
Canada,
1992-2002
8 South Korean
cities,
2003-2008
Hong Kong,
1996-2002
Sao Paulo,
Brazil,
2001-2003
17 Chinese cities
Ko et al.
(2007)
Villeneuve
et al. (2007)
Son et al.
(2013)
Wong et al.
(2009)
Arbex et al.
(2009)
Chen et al.
(2012)
January 2015
7-32
DRAFT: Do Not Cite or Quote
-------
Table 7-15 (Continued) Epidemiologic studies evaluating older adult lifestage.
Factor
Evaluated
Older
adulthood
Ages >65 yr,
n = 187,608
Older
adulthood
Ages >85 yr
n = 90,070
Older
adulthood
Ages
65-74 yr
n = 52,689
Reference
Category
Younger
adulthood,
childhood Ages
<65 yr,
n = 91,253
Younger
adulthood
Ages 35-64 yr
n = 181,031
Younger
adulthood
Ages 35-64 yr
n = 35,803
Direction of
Effect
Modification3 Outcome
y Total mortality
y Total mortality
1
Study
Population
Mean daily
mortality across
locations 7.29 to
15.8
N = 276,205
natural deaths
Study Details
Santiago
Province, Chile
(7 urban
centers),
1997-2007
10 Italian cities,
2001-2005
Study
Cakmak et al.
(2011)
Chiusolo et al.
(2011)
Long-term exposure
Older
adulthood
Ages >75 yrb
Older
adulthood
Ages >75 yra
Older
adulthood
Ages >60 yr
n = 365,368
Older
adulthood
Ages >60 yr
n= 4,061
Older
adulthood
Ages >60 yrb
Older
adulthood
Ages >70 yr
n = 1,329
Older
adulthood
Ages >50 yr
n=635
Younger
adulthood
Ages <60 yrb
Younger
adulthood
Ages 65-75 yra
Younger
adulthood
Ages 40-60 yr
n = 470,239
Younger
adulthood
Ages <60 yr
n = 5,880
Younger
adulthood
Ages <60 yrb
Younger
adulthood
Ages <70 yr
n = 527
Younger
adulthood
Ages 20-50 yr
n=242
1 Total mortality,
"*" Cardiovascular
mortality
_ Lung cancer
mortality
_ Lung cancer or
cardiopulmonary
mortality
y Total Mortality
_ Cardiovascular
mortality
y Respiratory
' Mortality
1 Total mortality
f HRV
' decrements in
low frequency
domain
N = 1,265,058
Ages >30 yr
N = 14,001
N = 835,607
deaths
Ages 40-89 yr
N = 9,941, 256
deaths
Ages 35-103 yr
16 cohorts
N = 307,553
Mean age at
baseline 41.9 to
73.0 yr across
cohorts
N = 3,320
Mean age 70 yr
N = 1,349
healthy subjects
Mean age 44 yr
Rome, Italy,
2001-2010
Shizuoka, Japan,
1999-2006
England
Follow-up:
2003-2007
NO2 exposure
assessed for
2002
Shenyang, China
Follow-up:
1998-2009
NO2 exposure
assessed for
1998-2009
Europe
Follow-up:
1985-2007
NO2 exposure
assessed for
2008-2011
London, England
Follow-up:
1995-2005
NO2 exposure
assessed for
2002
Taein Island,
South Korea,
2003-2004
Cesaroni et al.
(2013)
Yorifuii et al.
(2010)
Carey et al.
(2013)
Zhanq et al.
(2011)
Dimakopoulou
etal. (2014)
Maheswaran
etal. (2010)
Min et al.
(2008)
January 2015
7-33
DRAFT: Do Not Cite or Quote
-------
Table 7-15 (Continued) Epidemiologic studies evaluating older adult lifestage.
Factor
Evaluated
Older
adulthood
Ages >75 yr
n = 1,995
Older
adulthood
Ages >65 yr
n = 50
Older
adulthood
Ages >65 yr
n = 137,184
Older
adulthood
Ages
65-89 yrb
Older
adulthood
Ages >56 yr
n = 1,297
Older
adulthood
Ages >56 yr,
n = 106
Older
adulthood
Ages >60 yr
Femaleb
Maleb
Older
adulthood
Ages >65 yr
n = 2,234
Older
adulthood
Ages >50 yrb
Reference
Category
Younger
adulthood
Ages <60 yr,
n = 1,252
Younger
adulthood
Ages 60-75 yr,
n = 1,410
Younger
adulthood
Ages <65 yr
n=60
Younger
adulthood
Ages <65 yr
n= 417,156
Younger
adulthood
Ages 40-64 yrb
Younger
adulthood
Ages <56 yr
n=687
Younger
adulthood
Ages <56 yr,
n = 36
Younger
adulthood
Ages <60 yr
Femaleb
Maleb
Younger
adulthood
Ages 55-65 yr
n = 3,913
Younger
adulthood
Ages <50 yrb
Direction of
Effect
Modification3 Outcome
_ Out-of-hospital
cardiac arrest
t
_ Ventricular
arrhythmia
_ Myocardial
infarction
_ Heart failure
_ Incidence stroke
_ Fatal stroke
_ Intimedia
thickness cca
i Intimedia
^ thickness 6seg
* Intimedia
' thickness cca
_ Intimedia
thickness 6seg
_ Prevalent
hypertension
_ Diabetes
Study
Population
N = 4,657 events
N = 211 with
implantable
cardioverter
defibrillators
Age 28-85 yr
N = 43,275
cases,
51 1,065 controls
Ages 15-79 yr
N = 836,557
Ages 40-89 yr at
baseline
N = 1,984
Ages 50-65 yr at
baseline
N = 142
Ages 50-65 yr at
baseline
N = 2,780
Median age:
'58yr
N = 24,845
Mean age
45.59 yr
N = 6,392
Ages 29-73 yr
Study Details
Copenhagen,
Denmark,
2000-2010
Gothenburg,
Stockholm,
Sweden,
2001-2006
Stockholm
County, Sweden,
1985-1996
England,
2003-2007
Copenhagen,
Aarhus counties,
Denmark,
1993-2006
Girona Province,
Spain,
2007-2010
Shenyan,
Anshan and
Jinzhou, China,
2006-2008
Switzerland,
2002
Study
Wichmann
etal. (2013)
Ljunqman
et al. (2008)
Rosenlund
et al. (2009a)
Atkinson et al.
(2013)
Andersen
etal. (2012a)
Rivera et al.
(2013)
Dong etal.
(2013b)
Eze et al.
(2014)
aUp facing arrow indicates that the effect of NO2 is greater (e.g., larger risk of hospital admission, larger decrement in HRV) in the
group with the factor evaluated than in the reference group. Down facing arrow indicates that the effect of NO2 is smaller in the group
with the factor evaluated than in the reference group. A dash indicates no difference in NO2-related health effect between groups.
bSample size not reported.
January 2015
7-34
DRAFT: Do Not Cite or Quote
-------
Table 7-16 Controlled human exposure studies informing older adult lifestage.
Factor Reference
Evaluated Category
Older None
adulthood
Older None
adulthood
Direction of
Effect3 Outcome
Lung function
Lung function,
airway
inflammation
Study Population
N=20
(10 males,
10 females)
Mean age 61 yr
N=6
(2 males,
4 females)
Mean age 68 yr
Study Details
300 ppb NO2 for
4 h with exercise
400 ppb NO2 for
2 h with exercise
Study
Morrow et al.
(1992)
Gonq et al.
(2005)
aThese studies only examined older adults and have no reference group. A dash indicates that NO2 was not observed to induce an
effect in the older adults relative to clean air exposure.
7.5.2 Socioeconomic Status
i
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
SES is a composite measure that usually consists of economic status, measured by
income; social status measured by education, and work status measured by occupation.
Persons with lower SES have been generally found to have a higher prevalence of
pre-existing diseases; potential inequities in access to resources such as healthcare; and
possibly increased nutritional deficiences, which may increase this population's risk to
NC>2-related health effects. According to U.S. Census data, 15.9% (approximately
48.5 million) of Americans were of poverty status in 2011 as defined by household
income, which is one metric used to define SES (Bishaw. 2012). Across the indicators of
SES examined (e.g., education level, employment status, insurance status, social
deprivation, and access to health care) there is some evidence indicating higher NO2
exposure and larger risk of NCh-related health effects among low SES groups in the
population, but these relationships are not uniformly observed (Table 7-15). A challenge
in synthesizing findings across studies is the array of SES indicators examined. Further,
many studies were conducted outside of the U.S., and definitions of SES can vary across
countries based on population demographics, bureaucracy, and the local economy which
can contribute to varying degrees of deprivation or inequities.
Several studies relate higher NO2 exposure with indicators of low SES, but the
relationship varies across communities, levels of SES, and indicators of SES. Results
from studies conducted in the U.S., Canada, and Europe point to a relationship between
higher ambient NC>2 exposure among populations of low SES as determined by
household income, job class (e.g., unskilled, professional, skilled manual labor), or
education. Higher ambient NC>2 concentrations have been measured in communities in
Montreal, Canada and Los Angeles, CA with high proportions of nonwhite residents and
low SES residents (Suetal.. 2012: Molitoretal.. 2011: Grouse et al.. 2009: Su et al..
January 2015
7-35
DRAFT: Do Not Cite or Quote
-------
1 2009). While many of these studies examined community-level correlations based on
2 census block or tract SES and NC>2 modeled at the neighborhood scale (Clark etal.. 2014;
3 Deguen and Zmirou-Navier. 2010; Namdeo and Stringer. 2008; Kruize et al.. 2007;
4 Chaix et al.. 2006; Mitchell. 2005). studies with data from individuals also found
5 relationships between higher residential or personal NC>2 and lower SES (Llop et al..
6 2011; Deguen and Zmirou-Navier. 2010). A U.S.-wide analysis of census blocks suggests
7 inequities in NC>2 exposure in the low SES communities by age, with higher exposures
8 indicated for children and older adults (Clark et al.. 2014).
9 While most results indicate higher NC>2 exposure in low SES groups, some indicate that
10 the relationship between NCh exposure and SES varies in strength and direction. In some
11 cases, a nonlinear relationship is observed with either no difference in NC>2
12 concentrations among communities in the higher end of the income distribution [e.g., top
13 50%; (Kruize et al.. 2007)] or higher NC>2 concentrations in some affluent communities in
14 the downtown core of a city (Grouse et al.. 2009). Other studies find that the relationship
15 varies across communities (Stroh et al.. 2005) and among particular SES indicators [e.g.,
16 education but not occupation, country of birth but not education; (Stroh et al.. 2005;
17 Rotko et al.. 2001)]. The relationship between NCh exposure and SES also may weaken
18 over time as was forecasted for Leeds, U.K. Over the period 1993-2005, fleet renewal
19 and congestion pricing were predicted to reduce the discrepancy in NO2 exposures
20 between groups with high and low deprivation index [combining unemployment, noncar
21 ownership, nonhome ownership, and household overcrowding; (Mitchell. 2005)1. O'Neill
22 et al. (2003) noted that several factors might alter the relationship between NO2 exposure
23 and SES, including changing development, migration, and transportation patterns all of
24 which could result in individuals of high socioeconomic status having high NO2
25 exposures.
26 There is also the possibility that a multitude of factors may interact to influence the risk
27 of NO2-related health effects in populations of low SES. The hypothesis of "double
28 jeopardy" describes interactions between higher air pollution exposure and social
29 inequities in health, whereby risk of health effects for low SES and/or nonwhite
30 populations may be increased because of increased exposure and psychosocial stress or
31 access to health services (O'Neill et al.. 2003). An index combining risk factors such as
32 air pollution concentrations, including NO2, with nonwhite population and low SES
33 population has been constructed for communities in a few California cities (Su et al..
34 2012; Su etal.. 2009). However, relationships between such indices and health effects
35 have not been examined, and for the studies evaluated in this ISA, the risk for certain
36 SES or nonwhite populations resulting from multiple stressors have not been
37 characterized.
January 2015 7-36 DRAFT: Do Not Cite or Quote
-------
1 While there is strong evidence for relationships between short-term and long-term NC>2
2 exposure and asthma exacerbation (Section 5.2.9) and development (Section 6.2.9).
3 evidence does not clearly indicate differences among groups of varying SES
4 (Table 7-15). Children with higher psychosocial stress due to exposure to community
5 violence were observed to have increased risk for asthma incidence related to long-term
6 NO2 exposure (Clougherty et al.. 2007). Associations between short-term NC>2 exposure
7 and asthma-related effects mostly do not differ by SES (Burra et al.. 2009; Laurent et al..
8 2009; Kim et al.. 2007; Lin et al.. 2004). although a stronger association was found
9 among children in Phoenix, AZ with no insurance (Grineski et al.. 2010). The latter study
10 observed interactions between race/ethnicity and SES. NCh-related asthma hospital
11 admissions did not differ between Hispanic and white children, except in the group
12 without health insurance (Grineski et al.. 2010). Such results indicate the potential for
13 multiple co-occurring factors in certain populations to influence risk of NCh-related
14 health effects.
15 Several multeity studies in various countries found larger associations between
16 short-term NC>2 exposures and total mortality in low SES compared to high SES groups
17 as indicated by education, income, or employment [(Chen etal.. 2012; Cakmak et al..
18 2011; Chiusolo et al.. 2011); Table 7-17]. Despite the consistency, there is uncertainty in
19 the extent to which the findings can be attributed specifically to NO2 because uncertainty
20 regarding confounding by traffic-related copollutants is noted for a relationship between
21 short-term NC>2 exposure and total mortality (Section 5.4.8).
22 Evidence that SES modifies associations of long-term NC>2 exposure with cardiovascular
23 and related metabolic effects, mortality, reproductive effects, developmental effects, or
24 cancer is unclear (Table 7-17). However, independent effects of NO2 exposure on these
25 health effects is uncertain (Sections 6.3.9. 6.4.5. 6.5.3. 6.6.9). Some studies found larger
26 associations among lower SES groups (Becerraet al.. 2013; Carey etal.. 2013; Cesaroni
27 etal.. 2013; Morello-Frosch et al.. 2010). but studies equally found no difference among
28 SES groups (Eze etal.. 2014; Andersen etal.. 2012b; Guxens etal.. 2012; Pereira et al..
29 2012; Zhang etal.. 2011; Lenters etal.. 2010; Yorifuii etal.. 2010; Rosenlund et al..
30 2009a) or inconsistent effect modification among the outcomes examined (Foraster etal..
31 2014; Andersen et al.. 2012a; S0rensen et al.. 2012). A few studies observed weaker
32 NO2-related effects among lower SES groups (Atkinson et al.. 2013; Rivera etal.. 2013).
33 A diverse set of SES indicators was examined, and results were inconsistent even among
34 studies examining education or income.
35 Evidence indicates higher NC>2 exposure among low SES communities, although elevated
36 concentrations are also reported for some high SES communities. For short-term or
37 long-term NC>2 exposure, associations with respiratory effects, cardiovascular effects,
January 2015 7-37 DRAFT: Do Not Cite or Quote
-------
1
2
o
6
4
5
6
7
mortality, reproductive effects, developmental effects, and cancer do not consistently
vary by SES. However, evidence consistently demonstrates larger associations between
short-term NC>2 exposure and mortality among low SES groups. Interpreting this body of
evidence is challenging given the diversity of SES indicators used across studies, breadth
of countries where studies have been conducted, and varying certainty in the independent
effect of NO2 on the array of health effects examined. Overall, there is consistent
evidence for larger NCh-related increases in mortality in low SES groups, but uncertainty
remains in attributing the findings specifically to NC>2. As a result, the evidence is
suggestive that low SES populations are at increased risk for NCh-related health effects.
Table 7-17 Epidemiologic studies evaluating
Factor
Evaluated
Reference
Category
Direction of
Effect
Modification3 Outcome
socioeconomic status.
Study
Population
Study Details
Study
Short-term exposure
No
insurance
n=205
Low income
census
tracts'3
Low income-
based
insurance
premiums
n = 24%
quintile 1
Lowest
quintile for
census tract
income
n =610,121
Low SES
composite of
income,
education,
job, housing
factors
n = 43,674
lowest
stratum
Insurance
n= 2,508
private
n= 2,015
Medicaid
High income
census
tracts'3
High
income-
based
insurance
premiums
n = 17%
quintile 5
Highest
quintile for
census tract
income
n = 527,385
High SES
n=49,111
highest
stratum
y Asthma hospital
' admissions
_ Asthma hospital
admissions
_ Asthma
emergency
outpatient visits
_ Asthma
physician visits
_ Asthma
medication sales
Children
Ages <14 yr
N = 3,822
admissions
Children 6-12 yr
Mean
254 visits/day
Prior asthma
diagnosis
required
Data from
Ontario Health
Insurance Plan
N = 261, 063
Ages 0-39 yr
Phoenix, AZ,
2001-2003
Vancouver,
Canada (13
subdivisions),
1987-1998
Seoul, Korea,
2002
Toronto,
Canada,
1992-2001
Strasbourg,
France, 2004
Grineski et al.
(2010)
Lin et al. (2004)
Kim et al.
(2007)
Burra et al.
(2009)
Laurent et al.
(2009)
January 2015
7-38
DRAFT: Do Not Cite or Quote
-------
Table 7-17 (Continued): Epidemiologic studies evaluating socioeconomic status.
Factor
Evaluated
Low income
census
tracts <20th
percentile
n = 33,565
Low or
middle
income
census
tracts
n = 126,605
Low
education
(80th
percentile
n = 38,681
High
education
(university
diploma)13
High income
census area13
White collar
worker13
High
education
(middle
school and
above)13
Direction of
Effect
Modification3 Outcome
y Total mortality
* Total mortality
t
t
y Total mortality
Study
Population
N = 276,205
natural deaths
Ages >35 yr
SES available for
44% of study
population
Mean daily
mortality across
locations 7.29 to
15.8
Data from
Municipal Center
for Disease
Control and
Prevention
Study Details
10 Italian cities,
2001-2005
Santiago
Province, Chile
7 urban centers
1997-2007
17 Chinese
cities
Study
Chiusolo et al.
(2011)
Cakmak et al.
(2011)
Chen et al.
(2012)
Long-term exposure
High
exposure to
violence13
Blue collar
work, Low-
level white
collarwork
n = 57.9%
Low income
(mean of
controls)
n = 55.6%
High
education
(>high
school)
n=41.5%
High
education
n = 35%
y Asthma
' incidence
_ Myocardial
infarction
"
Atherosclerosis
(carotid intima-
media
thickness)
N = 417 children
followed from
prenatal period
N = 43,275
cases,
51 1,065 controls
N = 745
Ages 26-30 yr
Boston, MA,
1987-1993
Followed to
1997
Stockholm
county,
Sweden,
1985-1996
Utrecht,
Netherlands,
1999-2000
Clouqherty
et al. (2007)
Rosenlund
et al. (2009a)
Lenters et al.
(2010)
January 2015
7-39
DRAFT: Do Not Cite or Quote
-------
Table 7-17 (Continued): Epidemiologic studies evaluating socioeconomic status.
Factor
Evaluated
Primary or
secondary
school
n = 2,234
Lowest
deprivation
index
n = 20%
Low or
medium
education
(<10yr)
n = 1,628
Low or
medium
education
(<10yr)
n = 110
Low or
medium
education
(<10yr)
n= 40,956
Low or
medium
education
(primary or
secondary)
n= 4,586
Low or
medium
education
(<11 yr)
n = 79%
Low/
medium SES
municipality
n = 78%
Illiterate/
primary
education
n = 1,540
Low income
(<200
month)
n = 1,817
Low
education
n = 5,970
Reference
Category
Higher
education/
technician
n = 526
Highest
deprivation
index
n = 20%
High
education
(>10yr)
n = 356
High
education
(>10yr)
n = 32
High
education
(>10yr)
n = 10,862
High
education
(college or
university)
n = 1,806
High
education
(>1 1 yr)
n=21%
High SES
municipality
n = 22%
Secondary/
university
education
n= 2,160
High income
(>800
month)
n= 2,618
High
education
n = 3,971
Direction of
Effect
Modification3
1
1
t
t
-
t
"
—
Outcome
Atherosclerosis
(carotid intima
meda thickness)
Heart failure
Incident stroke
Fatal stroke
Diabetes
Diabetes
Hypertension
(change in
systolic BP)
Systolic blood
pressure
Diastolic blood
pressure
Cardiovascular
mortality
Study
Population
N = 2,780
Median age 58 yr
N = 836,557,
Ages 40-89 yr in
2003
N = 1984
Ages age
50-65 yr at
baseline
N = 142
Ages 50-65 yr at
baseline
N = 51,818
Ages 50-65 yr at
baseline
N = 6,392
Age 29-73 yr
N = 44,436
Ages 50-65 yr at
baseline
N = 3,700, age
35-83 yr
N = 9,941, 256
deaths
Ages 35-103 yr
Study Details
Girona
Province,
Spain,
2007-2010
England,
2003-2007
Copenhagen,
Aarhus
counties,
Denmark,
1993-2006
Copenhagen,
Aarhus
counties,
Denmark,
1993-2006
Switzerland,
2002
Copenhagen,
Aarhus
counties,
Denmark,
1993-2006
Girona, Spain
Shenyang,
China
Follow-up:
1998-2009
NO2 assessed
for 1998-2009
Study
Rivera et al.
(2013)
Atkinson et al.
(2013)
Andersen et al.
(2012a)
Andersen et al.
(2012b)
Eze et al.
(2014)
S0rensen et al.
(2012)
Foraster et al.
(2014)
Zhanq et al.
(2011)
January 2015
7-40
DRAFT: Do Not Cite or Quote
-------
Table 7-17 (Continued): Epidemiologic studies evaluating socioeconomic status.
Factor
Evaluated
Low
education
(<8 yr)
n = 33%
Low socio-
economic
position
census
block13
Lowest
quintile for
area-level
income
n = 12.5%
Financially
incapable
(self-
reported)
n= 4,054
Lowest SES
fertile
n = 7,556
High
neighbor-
hood level
poverty13
Low parental
social class
(semi-
skilled/un-
skilled
occupation
high
school)
n = 3,926
Direction of
Effect
Modification3
t
t
t
t
Outcome
Diabetes-
related Mortality
Mortality - total,
cardiovascular,
IHD, lung cancer
Total Mortality
Lung cancer
mortality
Small for
gestational age
or intrauterine
growth
restriction
Low birth weight
Mental
development in
infants at age
14 months
Autistic disorder
in children
Study
Population
N = 52,061
Ages 50-64 yr
N = 1,265,058
Ages >30 yr
N = 835,607
Ages 40-89 yr
N = 14,001
Ages >65 yr
N = 23,452
women/infants
N = 3,545,177
births
N = 1,889
children followed
from prenatal
period
N = 7,603
children with
autism,
10 controls per
case
Study Details
Denmark
Follow-up:
1971-2009
NO2 exposure
assessed for
1971-2009
Rome, Italy,
2001-2010
England
Follow-up:
2003-2007
NO2 exposure
assessed for
2002
Shizuoka,
Japan,
1999-2006
Perth, Western
Australia,
2000-2006
California,
1996-2006
4 Spanish
cities,
2003-2008
Los Angeles,
CA,
1998-2009
Study
Raaschou-
Nielsen et al.
(2012)
Cesaroni et al.
(2013)
Carey et al.
(2013)
Yorifuji et al.
(2010)
Pereira et al.
(2012)
Morello-Frosch
etal. (2010)
Guxens et al.
(2012)
Becerra et al.
(2013)
aUp facing arrow indicates that the effect of NO2 is greater (e.g., larger risk of hospital admission) in the group with the factor
evaluated than in the reference group. Down facing arrow indicates that the effect of NO2 is smaller in the group with the factor
evaluated than in the reference group. A dash indicates no difference in NO2-related health effect between groups.
bSample size not reported.
January 2015
7-41
DRAFT: Do Not Cite or Quote
-------
7.5.3 Race/Ethnicity
1 Based on the 2010 U.S. Census, 63.7% of the U.S. population identified themselves as
2 non-Hispanic whites; 12.6% reported their race as non-Hispanic black; and 16.3%
3 reported being Hispanic (Humes etal.. 2011). Race and ethnicity are complex factors that
4 are often closely correlated with other factors including particular genetics, diet, and
5 socioeconomic status. Therefore, race and ethnicity may influence any potential
6 differences in NC>2-related health effects through both intrinsic and extrinsic mechanisms.
7 Information characterizing racial/ethnic differences in NC>2 exposure is sparse but
8 suggests higher exposure among nonwhite people independent of SES. For U.S. urban
9 areas, the population-weighted mean annual average NO2 for nonwhites was estimated to
10 be 4.6 ppb (38%) higher than for whites (Clark etal.. 2014). This difference was
11 observed across the distribution of census block household income. However, NO2 was
12 estimated from a national scale LUR model and may reflect census block differences
13 other than race or in combination with race.
14 In contrast with exposure, NCh-related health effects have not been shown clearly to
15 differ between groups of nonwhite and white populations (Table 7-18). This is the case
16 for NO2-related asthma ED visits among children, but interestingly there was a difference
17 between Hispanic and white children when examining insurance status (Section 7.5.2)
18 (Grineski et al.. 2010). Additionally, Grineski et al. (2010) reported evidence of larger
19 NO2-related asthma ED visit associations among black children compared to Hispanic
20 children. Racial and ethnic differences in NC>2-related health effects also are not
21 consistently found for birth outcomes, although the implications of these findings are
22 weak because an independent relationship between NO2 exposure and birth outcomes is
23 not certain. Some studies estimated larger effects on birth weight or gestational age
24 among babies of black or Hispanic mothers (Rich et al.. 2009; Bell et al.. 2007); whereas
25 others estimated larger effects for babies of white mothers (Morello-Frosch et al.. 2010).
26 or no difference among races (Darrow et al.. 2011; Madsen et al.. 2010).
27 There is some indication that NC>2 exposure may be higher among nonwhite compared to
28 white populations, but information on NCh exposure at the individual-level is lacking.
29 NO2-related health effects do not consistently differ among racial and ethnic groups,
30 particularly, for asthma exacerbation, which is concluded to have an independent
31 relationship with short-term NC>2 exposure (Section 5.2.9). Additionally, it is unclear
32 whether higher NC>2 exposure and higher prevalence of potential at-risk factors in
33 combination impact the health of nonwhite populations (Section 7.5.2). Overall, the
34 evidence for potential differences in the risk of NO2-related health effects by race and
January 2015 7-42 DRAFT: Do Not Cite or Quote
-------
ethnicity is inconsistent and largely based on birth outcomes, for which an independent
relationship with NC>2 exposure is uncertain. Therefore, the evidence is inadequate to
determine whether race or ethnicity increases the risk for NC^-related health effects.
Table 7-18 Epidemiologic studies evaluating race/ethnicity.
Factor Reference
Evaluated Category
Direction of
Effect
Modification3 Outcome
Study
Population
Study Details Study
Short-term exposure
Black race White race
n = 635 n = 2,227
Black race Hispanic
race
n=635 n = 1,454
Hispanic White race
race
n = 1,454 n= 2,227
_ Asthma hospital
admissions
t
—
N = 4,316
Children ages
<14yr
Phoenix, AZ, Grineski et al.
2001-2003 (2010)
Long-term exposure
Black White
maternal maternal
race race
n = 10.7% n = 83.4%
Hispanic White
maternal maternal
race race
n = 14.3% n=45.2%
Non-
Hispanic
black
maternal
race
n = 40.5%
Non- Western
Western ethnicity
ethnicity
n = 24.3% n = 75.7%
Hispanic Non-
maternal Hispanic
race white
n = 51.5% maternal
race
Non- n = 32 20/o
Hispanic
black
maternal
race
n = 5.8%
-j- Birth weight
' decrements
_ Birth weight
decrements
Birth weight
decrements
i Birth weight
"*" decrements
1
N = 358,504
births
N = 406,627
full-term,
singleton births
N = 25,229
Full-term,
singleton births
N = 3,545,177
singleton births,
37-44 week
gestation
Massachusetts, Bell et al.
Connecticut, (2007)
1999-2002
Atlanta, GA, Darrow et al.
1994-2004 (2011)
Oslo, Norway Madsen et al.
(2010)
California, Morello-
1996-2006 Froschetal.
(2010)
January 2015
7-43
DRAFT: Do Not Cite or Quote
-------
Table 7-18 (Continued): Epidemiologic studies evaluating race/ethnicity.
Factor
Evaluated
Hispanic
maternal
race
n = 31%
Reference
Category
White or
African-
American
maternal
race
n = 69%
Direction of
Effect
Modification3
t
Outcome
Very small for
gestational age
(VSGA)
Study
Population
N = 178,198
singleton births,
37-42 week
gestation,
birth weight
>500 g
Study Details
New Jersey,
1999-2003
Study
Rich et al.
(2009)
aUp facing arrow indicates that the effect of NO2 is greater (e.g., larger risk of hospital admission, larger decrement in birth weight) in
the group with the factor evaluated than in the reference group. Down facing arrow indicates that the effect of NO2 is smaller in the
group with the factor evaluated than in the reference group. A dash indicates no difference in NO2-related health effect between
groups.
7.5.4 Sex
i
2
o
5
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
A vast number of health conditions and diseases have been shown to differ by sex with
some indication that there may be differences by sex in the relationship between air
pollution and health effects. The 2010 U.S. Census indicates an approximately equal
distribution of males and females in the U.S.: 49.2% male and 50.8% female (Howden
and Meyer. 2011). However, the distribution varies by age with a greater prevalence of
females above 65 years of age compared to males. Thus, the public health implications of
potential sex-based differences in air pollution-related health effects may vary among age
groups within the population.
With respect to NO2 exposure, limited evidence from a large (N = 1,634) multi-European
country study indicates no difference between males and females as demonstrated by
similar 2-week avg residential outdoor NO2 concentrations (Sunyer et al.. 2006). With
respect to NCh-related health effects, the strongest basis for inferring differences between
males and females comes from studies of asthma exacerbation, for which an independent
effect of short-term NO2 exposure is determined (Section 5.2.9). Associations between
short-term increases in ambient NC>2 concentration and asthma-related effects, as
examined in children, did not clearly differ between males and females: with
observations of larger effects in females (Lin et al.. 2004). males (Mann et al.. 2010). or
no difference between sexes (Sarnat et al.. 2012; Liu et al.. 2009). Inconsistent evidence
for potential differences by sex was also observed in studies examining associations
between short-term NCh exposures and respiratory infections in children (Zemek et al..
2010; Lin et al.. 2005). In contrast with findings for short-term NC>2 exposure, studies of
long-term NC>2 exposure were consistent in showing that associations with respiratory
effects such as asthma prevalence (Kim et al.. 2004) and lung function decrements
(Rosenlundetal.. 2009b: Oftedal et al.. 2008; Roias-Martinez et al.. 2007; Peters et al..
January 2015
7-44
DRAFT: Do Not Cite or Quote
-------
1 1999) were larger in female than male children. The majority of evidence is for lung
2 function decrements, and effect modification by sex is not explained by lower baseline
3 lung function in females. While many studies estimated NC>2 exposures at or near (2 km)
4 subjects" homes or schools (Rosenlund et al.. 2009b; Oftedal et al.. 2008; Rojas-Martinez
5 et al.. 2007). there is uncertainty regarding potential confounding by other traffic-related
6 pollutants (Section 6.2.9). Thus, it is not clear to what extent the larger NO2-related
7 decreases in lung function among females reflect an independent effect of NC>2.
8 Beyond respiratory effects, the majority of studies observed no difference between males
9 and females in associations of long-term NC>2 with an array of cardiovascular effects
10 (e-g-, myocardial infarction, heart failure, hypertension, stroke), diabetes, total mortality,
11 cause-specific mortality, or lung cancer incidence as described in Table 7-19 (Beelen
12 etal.. 2014; Eze etal.. 2014; Atkinson etal.. 2013; Cesaroni etal.. 2013; Dong et al..
13 2013a; Johnson etal.. 2013; Andersen etal.. 2012b; Raaschou-Nielsen et al.. 2012;
14 Raaschou-Nielsen et al.. 2011; Zhang etal.. 2011; Raaschou-Nielsen et al.. 2010; Yorifuji
15 etal.. 2010; Rosenlund et al.. 2009a; Abbey etal.. 1999). In most cases, no difference
16 between males and females was observed for NO2 associations with subclinical effects
17 such as changes in blood pressure, atherosclerosis, HRV, systemic inflammation, or
18 insulin resistance (Bilenko et al.. 2015; Foraster et al.. 2014; Atkinson et al.. 2013; Dong
19 etal.. 2013b; Rivera etal.. 2013; Thiering etal.. 2013; Lenters etal.. 2010; Panasevich
20 et al.. 2009; Felber Dietrich et al.. 2008). In the relatively small group of studies that
21 found differences between males and females, most observed greater risk among females
22 for associations of short-term NC>2 exposure with cardiac arrest ("Wichmann et al.. 2013)
23 or total mortality (Cakmak et al.. 2011; Kan et al.. 2008) and for associations of
24 long-term NC>2 exposure with total mortality or mortality from cardiovascular or
25 respiratory causes or lung cancer (Carey etal.. 2013; Katanoda et al.. 2011; Naess et al..
26 2007; Abbey etal.. 1999). Although most epidemiologic evidence indicates no difference
27 between males and females for cardiovascular morbidity and mortality and total mortality
28 related to long-term NCh exposure, similar to the evidence for long-term NCh exposure
29 and lung function, there is uncertainty as to whether NC>2 has an effect independent of
30 other traffic-related pollutants (Sections 6.3.9 and 6.5.3). Thus, the extent to which the
31 lack of effect modification by sex can be attributable to NO2 versus correlated
32 copollutants is not clear.
33 The collective body of evidence does not clearly indicate that NO2 exposure or
34 NO2-related health effects differ between males and females. Evidence demonstrates that
35 short-term NC>2 exposure has an independent effect on asthma exacerbations, but the few
36 studies do not indicate differences in associations between males and females.
37 Additionally, there is limited but inconsistent evidence for short-term NCh exposure and
38 cardiovascular disease and mortality, but uncertainty remains regarding an independent
January 2015 7-45 DRAFT: Do Not Cite or Quote
-------
1
2
o
6
4
5
effect of NC>2 on these outcomes. There is some support for larger associations in females
for long-term NC>2 exposure and pulmonary function, cardiovascular disease, and
mortality, though results are consistent only for pulmonary function. Therefore, the
current evidence is suggestive that females may be at increased risk for NCh-related
health effects.
Table 7-19 Epidemiologic studies evaluating
Factor
Evaluated
Reference
Category
Direction of
Effect
Modification3
Outcome
sex.
Study
Population
Study Details
Study
Short-term exposure
Female
n = 1,454
Female
n= 2,137
Female
n=20
Female
n=68
Female
n = 43.5%
Female
n= 2,784
Female
n = 8,055
Female
n=24
Female
n = 1,846
Female
n = 51.9%
Male
n= 2,368
Male
n - 2,077
Male
n = 38
Male
n = 114
Male
n = 56.5%
Male
n = 3,998
Male
n = 6,472
Male
n = 16
Male
n=2,811
Male
n - 48.1%
t
t
—
^™
1
"
t
t
t
1
Asthma hospital
admissions
Respiratory
hospital
admissions
Pulmonary
inflammation
Lung function
decrements,
Pulmonary
inflammation
Wheeze
Respiratory
infection hospital
admissions
Otitis media ED
visits
HRV
Out-of-hospital
cardiac arrest
Total mortality
N = 3,822
admissions Ages
6-12 yr
4 hospitals
N = 58 children
with asthma
Ages 6-1 2 yr
N = 182 children
with asthma
Ages 9-14 yr
N = 315 children
with asthma
Ages 6-11 yr
N = 6,782
admissions in
children
Ages 0-14 yr
N = 14,527 ED
visits in children
Ages 1-3 yr
N = 40
nonsmoking
adults with CVD
Mean age:
65.6 yr
N = 4,657 events
N = 276,205
natural deaths,
>35yr
Vancouver,
Canada,
1987-1998
Windsor,
Canada,
1995-2000
Ciudad Juarez,
Mexico and El
Paso, TX
Windsor,
Canada, 2005
Fresno, CA,
2000-2005
Toronto,
Canada,
1998-2001
Edmonton,
Canada,
1992-2002
Beijing, China,
2007-2008
Copenhagen,
Denmark,
2000-2010
10 Italian cities,
2001-2005
Lin et al.
(2004)
Luqinaah et al.
(2005)
Sarnat et al.
(2012)
Liu et al.
(2009)
Mann et al.
(2010)
Lin et al.
(2005)
Zemek et al.
(2010)
Huanq et al.
(2012)
Wichmann
etal. (2013)
Chiusolo et al.
(2011)
January 2015
7-46
DRAFT: Do Not Cite or Quote
-------
Table 7-19 (Continued): Epidemiologic studies evaluating sex.
Factor
Evaluated
Femaleb
Female
Mean daily
deaths =
56.5
Reference
Category
Maleb
Male
Mean daily
deaths =
62.5%
Direction of
Effect
Modification3 Outcome
y Total mortality
y Total mortality
Study
Population
Mean daily
mortality across
locations 7.29 to
15.8
N = 173,911
deaths
Study Details
Santiago
Province, Chile
(7 urban
centers),
1997-2007
Shanghai,
China,
2001-2004
Study
Cakmak et al.
(2011)
Kan et al.
(2008)
Long-term exposure
Female
n = 49%
Female
n = 52.6%
Female
n =
942-1,161
Female
n=648
Female
n = 832
Femaleb
Female
n = 49.4%
Female
n = 508
Female
n = 395
Female
n = 53.6%
Female
n = 1,980
Female
n= 431, 388
Male
n = 51%
Male
n = 47.4%
Male
n =
890-1,160
Male
n = 711
Male
n = 924
Maleb
Male
n = 50.6%
Male
n = 1,028
Male
n = 350
Male
n = 46.4%
Male
n = 1,720
Male
n= 405,169
y Lung function
' decrements
y Asthma diagnosis
_ Bronchitis
| Lung
' development
decrements
y Lung function
' decrements
_ Respiratory
symptoms
y Lung function
' decrements
_ Myocardial
infarction
_ Blood IL-6 levels
Mean carotid
artery intima-
media thickness
_ Atherosclerosis
(carotid intima
media thickness)
_ Systolic/diastolic
blood pressure
_ Heart failure
N = 2,307
children
Ages 9-10 yr
N = 1,109
children
Grades 3-5
N = 3,170
healthy children
Age 8 yr
N = 1,760
children
Ages 9-14 yr
N = 1,756 full-
term infants
Assessed at
ages 1 and 2 yr
N = 3,293
children
Grades 4, 7, 10
N = 43,275
cases,
511,065 controls
N = 1,536
Ages 45-70 yr
N = 745
Ages 26-30 yr
N = 2,780
Median age 58 yr
N = 3,700
Age 35-83 yr
N = 836,557,
Age 40-89 yr in
2003
Oslo, Norway,
2001-2002
San Francisco,
CA, 2001
Mexico City,
Mexico,
1996-1999
Rome, Italy,
2000-2001
3 German cities,
1995-1999
Southern
California,
1986-1990
Stockholm
county, Sweden,
1985-1996
Stockholm
county, Sweden,
1992-1994
Utrecht,
Netherlands,
1999-2000
Girona Province,
Spain,
2007-2010
Girona, Spain
England,
2003-2007
Oftedal et al.
(2008)
Kim et al.
(2004)
Roias-
Martinez et al.
(2007)
Rosenlund
et al. (2009b)
Gehrinq et al.
(2002)
Peters et al.
(1999)
Rosenlund
et al. (2009a)
Panasevich
et al. (2009)
Lenters et al.
(2010)
Rivera et al.
(2013)
Foraster et al.
(2014)
Atkinson et al.
(2013)
January 2015
7-47
DRAFT: Do Not Cite or Quote
-------
Table 7-19 (Continued): Epidemiologic studies evaluating sex.
Factor
Evaluated
Female
n = 573
Female
n= 23,092
Female
n = 12,184
Female
n = 12,184
Female
n = 829
Female
n=63
Female
n = 18,085
Female
n = 725
Female with
CVD
n = 115
Female
without
CVD
n =610
Female
n= 27,273
Female
n = 51.3%
Female
n=222
Reference
Category
Male
n = 574
Male
n=21,344
Male
n = 12,661
Male
n = 12,661
Male
n = 1,155
Male
n = 79
Male
n= 29,030
Male
n - 683
Male with
CVD
n = 121
Male
without CVD
n = 562
Male
n = 24,545
Male
n = 48.7%
Male
n = 175
Direction of
Effect
Modification3 Outcome
_ Systolic/diastolic
blood pressure
i Hypertension
"*" (change in
systolic BP)
_ Incident CVD
_ Incident stroke
_ Incidence
hypertension
_ Absolute increase
in arterial blood
pressure
_ Incident stroke
_ Fatal stroke
_ Stroke
_ HRV decrements
(SDNN)
—
t
Diabetes
_ Diabetes
_ Insulin resistance
Study
Population
N = 1,147
children
Age 12 yr
N = 44,436
Age 50-65 yr at
baseline
N = 24,845
Mean age 41.7 yr
N = 24,845
Mean age
45.59 yr
N = 1,984
Ages 50-65 yr at
baseline
N = 142
Ages 50-65 yr at
baseline
N = 4,696 cases
and
37,723 controls
Age >20 yr
N = 1,408
Ages >50 yr
N = 51,818
Ages 50-65 yr at
baseline
N = 6,392
Ages 29-73 yr
N = 397 children
Age 10 yr
Study Details
The
Netherlands,
Long-term and
short-term NO2
exposure
Copenhagen,
Aarhus,
Denmark,
1993-2006
Shenyan,
Anshan and
Jinzhou, China,
2009
Shenyan,
Anshan and
Jinzhou, China,
2009-2010
Copenhagen,
Aarhus counties,
Denmark,
'1993-2006
(Long-term)
Edmonton,
Alberta, Canada,
2007-2009
Switzerland
Followed 1991 to
2001-2003
Copenhagen,
Aarhus counties,
Denmark,
1993-2006
Switzerland,
2002
Munich and
Wesel, Germany
Study
Bilenko et al.
(2013)
S0rensen
etal. (2012)
Dong et al.
(2013a)
Donq et al.
(2013b)
Andersen
etal. (2012a)
Johnson et al.
(2013)
Felber Dietrich
et al. (2008)
Andersen
etal. (2012b)
Eze et al.
(2014)
Thierinq et al.
(2013)
January 2015
7-48
DRAFT: Do Not Cite or Quote
-------
Table 7-19 (Continued): Epidemiologic studies evaluating sex.
Factor
Evaluated
Female
Female
n = 51.47%
Female
n = 63%
Female
n = 52.5%
Female
n= 430,891
Femaleb
Female
n = 111
deaths
Female
n = 203-
7,840
deaths
Reference
Category
Male
Male
n = 48.53%
Male
n = 37%
Male
n = 47.5%
Male
n= 404,716
Maleb
Male
n=407
deaths
Male
n = 233-
4,531 deaths "
Direction of
Effect
Modification3 Outcome
_ Respiratory
mortality
CVD mortality
CVD mortality
Diabetes-related
mortality
t Total mortality
Mortality
(total, CV, IHD,
lung cancer)
— Lung cancer
mortality
t Respiratory
mortality
t Lung cancer
mortality
CVD mortality
I COPD mortality
Study
Population
16 cohorts
N = 307,553
Mean age at
baseline 41.9 to
73.0 yr across
cohorts
N = 9,941, 256
deaths
Ages 35-103 yr
N = 52,061
Ages 50-64 yr
N = 835,607
deaths
Ages 40-89 yr
N = 1,265,058
Ages >30 yr
N = 63,520
Ages >40 yr
N = 138,977,518
deaths
Ages 51-90 yr
Study Details
Europe
Follow-up:
1985-2007
NO2 exposure
assessed for
2008-2011
Shenyang,
China
Follow-up:
1998-2009
NO2 exposure
assessed for
1998-2009
Europe
Follow-up:
1985-2007
NO2 exposure
assessed for
2008-2011
Denmark
Follow-up:
1971-2009
NO2 exposure
assessed for
1971-2009
England
Follow-up:
2003-2007
NO2 exposure
assessed for
2002
Rome, Italy,
2001-2010
3 Japanese
prefectures,
1983-1985
Oslo, Norway,
1992-1998
Study
Dimakopoulou
etal. (2014)
Zhana et al.
(2011)
Beelen et al.
(2014)
Raaschou-
Nielsen et al.
(2012)
Carey et al.
(2013)
Cesaroni et al.
(2013)
Katanoda
etal. (2011)
Naess et al.
(2007)
January 2015
7-49
DRAFT: Do Not Cite or Quote
-------
Table 7-19 (Continued): Epidemiologic studies evaluating sex.
Factor
Evaluated
Female
n= 4,060
Female
n = 49%
Female
n = 1,206
Female
n= 27,788
Reference
Category
Male
n= 2,278
Male
n = 51%
Male
n= 2,275
Male
n= 25,182
Direction of
Effect
Modification3 Outcome
t Lung cancer
mortality
- Total mortality
- Cardiopulmonary
mortality
- Respiratory
mortality
— Lung cancer
mortality
— Lung cancer
incidence
— Lung cancer
incidence
Study
Population
N = 6,338
nonsmoking,
" non-Hispanic
adults
Ages 27-95 yr
N = 14,001 Ages
>65yr
N = 3,481
Ages 20-93 yr at
enrollment
N = 52,970
Ages 50-64 yr
Study Details
California,
Follow-up:
1977-1992
NO2 exposure
assessed for
1973-1992
Shizuoka,
Japan,
1999-2006
Copenhagen,
Aarhus counties,
Denmark,
1970-1997
Copenhagen,
Aarhus counties,
Denmark,
1993-1997
Study
Abbey et al.
(1999)
Yorifuii et al.
(2010)
Raaschou-
Nielsen et al.
(2010)
Raaschou-
Nielsen et al.
(2011)
aUp facing arrow indicates that the effect of NO2 is greater (e.g., larger risk of hospital admission, larger decrement in HRV) in the
group with the factor evaluated than in the reference group. Down facing arrow indicates that the effect of NO2 is smaller in the group
with the factor evaluated than in the reference group. A dash indicates no difference in NO2-related health effect between groups.
bSample size not reported.
7.5.5 Residence in Urban Areas
i
2
3
4
5
6
7
8
9
10
11
12
13
14
A majority (81%) of the U.S. populations lives in urban areas, and U.S. Census data
indicate that the urban population grew 12% from 2000 to 2010. Higher ambient NC>2
concentrations in urban than suburban areas and the large numbers of people potentially
having higher exposures highlights the public health significance of potential differences
in NCh-related health effects in urban residents. Higher ambient NC>2 concentrations have
been described for downtown versus suburban areas [14.9 ppb vs. 11.7 ppb; (Rotko et al..
2001)1. Higher ambient NC>2 concentrations also were related to building characteristics
such as high-rise building versus single family home and older versus newer construction
(before or after 1970). Proximity to roads has been shown to be a determinant of personal
NC>2 exposure (Section 7.5.6). and the higher road density in urban areas and proximity to
major roads also may result in higher exposure of urban residents. The topography of
urban communities also may contribute to higher NC>2 exposure among residents as the
presence of street canyons enhances mixing at elevations closer to the street
canyon-urban boundary layer interface, resulting in higher NC>2 concentrations at lower
January 2015
7-50
DRAFT: Do Not Cite or Quote
-------
1 elevations (Section 2.5.3). This may have implications for higher NC>2 exposures for
2 pedestrians, outdoor workers, and those living on lower floors of buildings. The evidence
3 indicates that residing in urban areas may lead to increased exposure to NC>2.
4 Although the potential for higher exposure to NCh of urban residents is well
5 characterized, epidemiologic comparisons of NC>2-related health effects between urban
6 and nonurban residents are limited and use variable definitions of urban and nonurban
7 residence (Table 7-20). A larger association between short-term increases in NO2 and
8 lung function decrements among urban than suburban children was observed for children
9 in the general population (Steerenberg et al.. 2001) but not for children with asthma
10 (Ranzi et al., 2004). Associations of long-term NC>2 and cardiovascular effects did not
11 differ by urban residence (Atkinson et al.. 2013; S0rensen et al.. 2012). but there is
12 uncertainty regarding the extent to which there is an independent effect of NC>2 exposure
13 on cardiovascular effects (Section 6.3.9).
14 As detailed above, urban residents potentially have higher ambient NO2 exposure.
15 However, the limited epidemiologic evidence based on variable definitions of urban and
16 nonurban residence does not provide a strong basis for inferring whether urban residence
17 leads to increased risk for NCh-related health effects. Overall, the limited number of
18 epidemiologic studies that examined urban residence is inconsistent and based primarily
19 on health effects for which independent relationships with NC>2 exposure are uncertain.
20 As a result, the evidence is inadequate to determine whether residence in urban areas
21 increases the risk for NC^-related health effects.
January 2015 7-51 DRAFT: Do Not Cite or Quote
-------
Table 7-20 Epidemiologic studies evaluating urban residence.
Factor
Evaluated
Direction of
Reference Effect
Category Modification3 Outcome
Study
Population
Study Details
Study
Short-term exposure
Urban
residence
n = 38
Urban-
industrial
residence
n=67
Suburban * Lung function
residence ' decrements
n=44
Rural i Lung function
residence "*" decrements
n = 51
N = 82
Ages 8-1 Syr
N = 118
Ages 6-11 yr
Children with
asthma or
respiratory
symptoms
Utrecht,
Bilthoven, the
Netherlands
Urban & rural
areas Emilia-
Romagna, Italy,
1999
Steerenberq
etal. (2001)
Ranzi et al.
(2004)
Long-term exposure
Residence
in London
n = 91,992
Residence
near city
center
n= 24,514
Residence in _ Heart failure
North/ South
UK
(excluding
London)
n = 744,565
Residence _ Systolic blood
outside of pressure
city center
n = 19,922
N = 836,557
Ages 40-89 yr in
2003
N = 44,436
Ages 50-65 yr at
baseline
England,
2003-2007
Copenhagen,
Aarhus counties,
Denmark,
1993-2006
Atkinson et al.
(2013)
S0rensen
etal. (2012)
aUp facing arrow indicates that the effect of NO2 is greater (e.g., larger decrement in lung function) in the group with the factor
evaluated than in the reference group. Down facing arrow indicates that the effect of NO2 is smaller (e.g., smaller decrement in lung
function) in the group with the factor evaluated than in the reference group. A dash indicates no difference in NO2-related health effect
between groups.
7.5.6 Proximity to Roadways
i
2
o
J
4
5
6
7
8
9
10
11
NC>2 concentrations can be 30 to 100% higher within 10-20 m of roads compared to
locations 80-500 m away (Sections 2.5.3 and 3.3.1.1). Thus, individuals spending a
substantial amount of time on or near high-traffic roadways, including those living or
working near highways and commuters, are likely to be exposed to elevated NO2
concentrations.
Large proportions of the U.S. population potentially have elevated NO2 exposures as a
result of proximity to roadways. Seventeen percent of U.S. homes are located within
91m of a highway with four or more lanes, a railroad, or an airport (U.S. Census Bureau.
2009). Specific to road traffic, Rowangould (2013) found that over 19% of the U.S.
population lives within 100 m of roads with an annual average daily traffic (AADT) of
25,000 vehicles, and 1.3% lives near roads with AADT greater than 200,000. The
January 2015
7-52
DRAFT: Do Not Cite or Quote
-------
1 proportion is much larger in certain parts of the country, mostly coinciding with urban
2 areas. For example in Los Angeles, CA, primary and secondary roads run through
3 neighborhoods with population density as high as 15,000-18,000 people per km2 [(U.S.
4 Census Bureau. 2014. 2013); Figure 7-1]. Among California residents, 40% lives within
5 100 m of roads with AADT of 25.000 (Rowangould. 2013).
6 Though far from generalizable across populations, there are examples indicating that
7 residence near a busy roadway may be associated with higher NO2 exposure. In the
8 Southern California CHS cohort, closer proximity to a freeway showed a range of
9 correlations with residential NC>2 measurements, with values of-0.73 to -0.90 in some
10 communities (Gauderman et al.. 2005). In a cohort of pregnant women who spent on
11 average 60% of time home indoors, traffic intensity within 250 and 500 m of homes was
12 moderately correlated with personal NC>2 exposures [r = 0.3, 0.4, respectively;
13 (Schembari et al.. 2013)]. However, the strongest correlation was not observed for traffic
14 intensity in closest proximity to homes as the correlation between personal NC>2 and
15 traffic intensity within a 100-m buffer was 0.2. Such results may be explained by the
16 atmospheric chemistry of NCh. Depending on atmospheric stability, NC>2 concentrations
17 can dilute with distance from the road or extend beyond 1 km of the roadway
18 (Section 2.5.3) and may be higher after some of the NO reacts photochemically to
19 become NO2.
January 2015 7-53 DRAFT: Do Not Cite or Quote
-------
0 4,250 8,500
17.000
Population Density per Sq Kilometer
j^H 110-4.600
I 4,700-9,100
| | 9.200 - 14,000
| | 15,000-18,000
| | 19,000-23.000
HH 24,000 - 27,000
Primary and Secondary Roads
J Los Angeles city boundry
A
N
Source: National Center for Environmental Assessment analysis of U.S. Census data (U.S. Census Bureau, 2014, 2013).
Figure 7-1 Map of population density in Los Angeles, CA in relation to
primary and secondary roads.
1 Exposure to NC>2 in transport is found to be an important determinant of total personal
2 exposure (Son et al., 2004; Lee et al. 2000). although time in transport makes up a
3 relatively small proportion of people's activities. Such findings have implications for
4 commuters as well as for occupational drivers. Among the populace working outside the
5 home, 15.6% spend 45 minutes or more commuting to work each day (U.S. Census
6 Bureau. 2007). Average one-way commuting times for the U.S. labor force working
January 2015
7-54
DRAFT: Do Not Cite or Quote
-------
1 outside the home are 19.3 minutes for bicyclists, 11.5 minutes for walkers, and
2 25.9 minutes for all other modes of transportation. The potential of higher NC>2 exposure
3 of commuters and professional drivers is supported by many observations of in-vehicle
4 NC>2 concentrations approaching roadside concentrations (Figure 3-2) and some evidence
5 of higher personal NC>2 exposure during transport than in outdoor or indoor environments
6 (Delgado-Saborit. 2012). The relationship between NO2 exposure during commute or
7 while driving for work and health effects is not well characterized. In the CHS cohort,
8 increasing commuting time to school was associated with wheeze but not asthma onset;
9 commute-time NC>2 exposures were not measured (McConnell et al.. 2010).
10 Children are characterized to be at increased risk for NCh-related health effects
11 (Section 7.5.1.1). and time spent near major roads could potentially be a source of higher
12 NO2 exposure contributing to health effects. Attendance at schools or daycare near major
13 roads may be an important determinant of NC>2 exposure, and ambient NC>2
14 concentrations at schools have been associated with respiratory effects in children with
15 asthma (Sections 5.2.2.2 and 5.2.2.5). Seven percent of U.S. schools serving
16 3,152,000 school children are located within 100 m of a major roadway, and 15% of U.S.
17 schools serving 6,357,000 school children are located within 250 m of a major roadway
18 [not specifically defined in terms of AADT, number of lanes, or other criteria; (Kingsley
19 et al.. 2014)1. In California, 2.3% of public schools serving 150,323 children were
20 estimated to be located within 150 m of high-traffic roads [>50,000 vehicles per day;
21 (Green et al.. 2004)1. Also in California, 1,534 daycare facilities serving 57,173 (7% of
22 those in daycare) children were within 200 m of roadways with AADT of >5 0,000, and
23 4,479 facilities serving 171,818 (21%) children were within 200 m of roadways with
24 AADT of 25,000-49,999 (Houston et al.. 2006). Though neither of these analyses
25 assessed NCh exposures, they identify the large numbers of children potentially exposed
26 to higher NC>2 concentrations in locations where they spend several hours per day.
27 There is some indication that traffic exposures differ among sociodemographic groups. In
28 California, schools or daycare in close proximity to high-traffic roadways had a higher
29 percentage of nonwhite students (Green et al.. 2004) or tended to be located in areas with
30 higher percentages of nonwhite residents (Houston et al.. 2006). Analyses of U.S. Census
31 blocks or tracts indicate associations of higher traffic or road density or proximity to
32 roadways with higher proportion of nonwhite residents (Rowangould. 2013; Tian et al..
33 2013). In some (Rowangould. 2013: Green et al.. 2004) but not all (Tian etal.. 2013)
34 cases, closer proximity to roadways or higher traffic density was associated with lower
35 SES at the school or census block level. In analyses not considering proximity or density
36 of traffic, higher NC>2 exposures are suggested among nonwhite (Section 7.5.3) or low
37 SES (Section 7.5.2). However, it is not understood whether observations of higher NC>2
38 exposures in certain sociodemographic groups is related to disparities in traffic exposure.
January 2015 7-55 DRAFT: Do Not Cite or Quote
-------
1 Large proportions of the U.S. population live or attend school near roads or travel on
2 roads, and some evidence indicates higher NO2 exposure with proximity to roads. Traffic
3 proximity may be more prevalent among nonwhite and low SES groups, but the influence
4 of traffic proximity on differential NC>2 exposure in these groups is unclear. While traffic
5 proximity (HEI. 2010) and NC>2 exposure near traffic (Section 5.2.9.3) are linked to
6 asthma exacerbation or prevalence, evidence does not clearly indicate larger NCh-related
7 health effects in populations living near traffic (Table 7-21). Closer proximity to freeway
8 was associated with larger NCh-related decrements in lung development among children
9 (Gauderman et al.. 2007), but NC>2 concentrations as measured at central sites were
10 weakly correlated with traffic counts near homes, and an independent effect of NCh
11 exposure on lung development is uncertain. Additionally, results are inconclusive for
12 cardiovascular effects and leukemia (Toraster et al.. 2014; Hart et al.. 2013; Amigou
13 et al.. 2011). The insufficient quantity and consistency of evidence and uncertainty
14 regarding the independent effects of NO2 exposure is inadequate to determine whether
15 populations in close proximity to roadways are at increased risk for NCh-related health
16 effects.
January 2015 7-56 DRAFT: Do Not Cite or Quote
-------
Table 7-21 Epidemiologic studies evaluating proximity to roadways (all
long-term exposure).
Factor
Evaluated
Residence
<500 m
from
freeway
n=440
Near road
n = 539MI
cases
Moved near
road
n=48MI
cases
Moved
away from
road
n=603MI
cases
Traffic
intensity of
nearest
road
>median
Traffic load
at 500 m
>median
Residence
near main
roads
(<500 m)
n=48
Reference
Category
Residence
1,000-1,500
m from
freewayb
Far from
road
(>50 m from
secondary
road [> 2
lanes] or
>150 m from
primary road
[highway])
n= 2,841 Ml
cases
Traffic
intensity of
nearest road
500 m)
n = 954
Direction of
Effect
Modification3 Outcome
y Decrements in
' lung development
(FEV-i change
overtime)
* Incident Ml
t
Systolic blood
pressure
_ Diastolic blood
pressure
* Systolic blood
' pressure
_ Diastolic blood
pressure
_ Leukemia
Study
Population
N = 3,677
children followed
ages 10-18 yr
N = 84,562
Age 30-55 yr at
enrollment
N = 3,700
Age 35-83 yr
N = 763 cases,
1,681 controls
Ages <15 yr
Study Details
Alpine, Lake
Elsinore, Lake
Arrowhead,
Atascadero,
Lancaster, San
Dimas, Long
Beach, Mira
Loma, Lompoc,
Riverside, Santa
Maria, Upland, CA
Follow-up:
1993/1 996 to
2001/2004
U.S.,
1990-2008
Girona, Spain
France,
2003-2004
Study
Gauderman
et al. (2007)
Hartetal.
(2013)
Foraster
etal. (2014)
Amiqou et al.
(2011)
aUp facing arrow indicates that the effect of NO2 is greater (e.g., larger decrement in lung function, larger risk of Ml) in the group with
the factor evaluated than in the reference group. Down facing arrow indicates that the effect of NO2 is smaller in the group with the
factor evaluated than in the reference group. A dash indicates no difference in NO2-related health effect between groups.
bSample size not reported.
January 2015
7-57
DRAFT: Do Not Cite or Quote
-------
7.6 Behavioral and Other Factors
7.6.1 Diet
1 Diet is an important influence on health and thus, plausibly could influence air
2 pollutant-related health effects. The 2008 ISA for Oxides of Nitrogen (U.S. EPA. 2008)
3 did not discuss whether diet influences the risk of NCh-related health effects; however,
4 evidence from previous experimental studies indicates reduced or greater respiratory
5 effects in humans and rodents with supplementation of or deficiencies in antioxidant
6 vitamins, respectively (Table 7-22). A controlled human exposure study demonstrated
7 that healthy adults with diets supplemented with Vitamin C had less airway
8 responsiveness following 2,000 ppb NCh for 1 hour compared to adults with a normal
9 diet (Mohsenin. 1987). Airway responsiveness is a hallmark of asthma exacerbation
10 (Figure 4-1). The evidence that higher antioxidant vitamin intake reduces NC>2-induced
11 airway responsiveness is supported by experimental evidence in humans and rodents that
12 higher dietary vitamin E and/or C reduces NC>2-induced pulmonary inflammation and
13 modulates the oxidant/antioxidant balance [(Mohsenin. 1991; Hatch et al.. 1986; Elsayed
14 and Mustafa. 1982: Sevanian et al.. 1982: Selgrade et al.. 1981: Avaz and Csallanv.
15 1978); Table 7-221. which are early events in the mode of action for NC>2 effects on
16 asthma exacerbation (Section 4.3.5. Figure 4-1). Despite the consistency and coherence
17 of evidence, findings are limited, particularly for changes that are indicative of health
18 effects. The changes in NC>2-induced lipid peroxidation, antioxidant levels, and
19 antioxidant enzyme activity observed in relation to vitamin deficiencies or
20 supplementation may or may not lead to health effects.
21 Epidemiologic studies have not examined whether diet modifies NCh-related respiratory
22 effects. Limited information indicates that associations of long-term NCh exposure with
23 mental development in infants are larger in groups with low fruit intake (maternal
24 prenatal or concurrent, respectively) than groups with high fruit intake [(Guxens et al..
25 2012): Table 7-23]. Fruits are a source of antioxidants; thus, the results for modification
26 by fruit intake are consistent with those for dietary antioxidant vitamins. However,
27 because evidence for NCh-related neurodevelopmental effects is overall inconclusive
28 (Section 6.4.5). the available epidemiologic evidence cannot adequately inform whether
29 diet deficiencies increase the risk for NCh-related health effects.
30 Experimental studies in humans and animals provide evidence that dietary intake of
31 Vitamin C or E modifies airway responsiveness, pulmonary inflammation, and oxidant
January 2015 7-58 DRAFT: Do Not Cite or Quote
-------
1 balance following NO2 exposure, with high vitamin intake reducing these effects and low
2 intake increasing effects. Epidemiologic evidence is available only for health effects for
3 which relationships with NC>2 are uncertain. Oxidative stress, pulmonary inflammation,
4 and airway responsiveness are key events in the mode of action for asthma exacerbation
5 (Figure 4-1); thus, a biologically plausible mechanism exists for dietary antioxidants to
6 reduce the risk of NCh-related health effects. Although biological plausibility exists for
7 dietary deficiencies influencing the risk for NC^-related health effects, most findings are
8 for changes in oxidant/antioxidant balance rather than changes clearly indicative of health
9 effects such as airway responsiveness. Therefore, there is suggestive evidence that
10 insufficient dietary antioxidant intake increases the risk for NC^-related health effects.
January 2015 7-59 DRAFT: Do Not Cite or Quote
-------
Table 7-22 Controlled human exposure and toxicological studies evaluating diet
Factor Evaluated
Vitamin C
supplemented diet
3 days
n = 11
Vitamin C and E
supplemented diet
4 weeks
n = 10
Vitamin C
supplemented diet
n=4-5
Vitamin C deficient
diet
n = 5-8
Vitamin E deficient
diet,
birth-adolescence
n=6-7
Vitamin E deficient
diet,
birth -adolescence
n=6
Vitamin E deficient
diet
n=6-10
Direction of
Reference Effect
Category Modification3 Outcome
Normal diet
n = 11
Normal diet
n = 9
Vitamin C normal
diet
n = 9
Vitamin C
supplemented
diet
n = 15
Vitamin E
supplemented
diet
n = 6-8
Vitamin E
supplemented
diet
n = 6
Vitamin E
supplemented
diet
n = 6-10
1
1
t
t
t
t
t
Airway
responsive-
ness
Lipid
peroxidation
in lavage fluid
Pulmonary
inflammation
Pulmonary
inflammation,
Antioxidant
reduction
Lipid
peroxidation,
Pulmonary
inflammation
Lipid
peroxidation,
Induction of
antioxidant
enzymes
Glutathione
peroxidase
activity
reduction
Study
Population/
Animal Model
Humans
n = 8 male,
3 female
Ages 18-37 yr
Humans
n = 10 male,
9 female
Ages 21 -33 yr
Guinea pigs
(Hartley)
n = 2-6
males/group
Guinea pigs
(Hartley)
n = 3-15
males/group
Rats (Sprague-
Dawley)
n = 6-8/group
Rats (Sprague-
Dawley)
n = 6/group
Mice
(C57BL/6J)
n = 120 females
Study Details
2,000 ppbfor
1 hour,
randomized,
double-blind
4,000 ppbfor
3 hour
400, 1,000,
3,000, or
5,000 ppbfor
3 days
4,800 ppbfor
3 hours
3,000 ppbfor
1 week
3,000 ppbfor
1 week
500 or
1,000 ppbfor
17 months
Study
Mohsenin
(1987)
Mohsenin
(1991)
Belgrade
etal. (1981)
Hatch et al.
(1986)
Sevanian
etal. (1982)
Elsaved and
Mustafa
(1982)
Avaz and
Csallany
(1978)
aUp facing arrow indicates that the effect of NO2 is greater (e.g., larger increase in airway responsiveness, larger increase in lipid
peroxidation) in the group with the factor evaluated than in the reference group.
January 2015
7-60
DRAFT: Do Not Cite or Quote
-------
Table 7-23 Epidemiologic studies evaluating
Factor Reference
Evaluated Category
Low healthy High healthy
eating index eating index
(<109)b (>109)b
Low Medium/
maternal high
fruit intake maternal fruit
in 1st intake in 1st
trimester trimester
(<405 (405 g/day)
g/day) n = 66.5%
n = 33.5%
Direction of
Effect
Modification3 Outcome
_ Ml Incidence
-j- Decrement in
' mental
development
score in infants at
age 14 mo
diet (all long-term exposure).
Study
Population Study Details Study
N = 84,562 U.S., Hart et al.
Age 30-55 yr at 1 990-2008 (2013)
enrollment
N = 1,889 4 Spanish cities Guxens et al.
children followed 2003-2008 (2012)
from prenatal
period
aUp facing arrow indicates that the effect of NO2 is greater (e.g., larger decrement in mental development score) in the group with the
factor evaluated than in the reference group. A dash indicates no difference in NO2-related health effect between groups.
bSample size not reported.
7.6.2 Smoking
i
2
o
6
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
Smoking is a common behavior, the 2010 National Health Interview Survey estimated
that within the U.S. adult population approximately 19.2% of individuals report being
current smokers and 21.5% report being a former smoker (Schiller et al., 2012). Smoking
is a well-documented risk factor for many diseases, but it is unclear whether smoking
exacerbates health effects associated with air pollutant exposures, including NC>2.
Although many controlled human exposure studies report smoking status, comparisons
between smokers and nonsmokers are infrequent due to small sample size. However, in
limited examination, a 4-hour exposure to 300 ppb NC>2 induced a larger decrement in
mean forced expiratory volume in 1 second (FEVi) among 7 smoking, healthy adults than
among 13 nonsmoking subjects (Morrow et al.. 1992). There is a lack of epidemiologic
studies to draw direct coherence with this experimental evidence for respiratory effects.
As examined primarily for cardiovascular or diabetes morbidity and mortality, most
associations with long-term NC>2 do not differ between smokers and nonsmokers
(Dadvand et al.. 2014; Atkinson etal.. 2013; Hart etal.. 2013; Rivera et al.. 2013;
Andersen et al.. 2012a: Zhang etal.. 2011; Lenters et al.. 2010; Panasevich et al.. 2009)
or are larger among nonsmokers [(Carey etal.. 2013; Andersen et al.. 2012b:
Raaschou-Nielsen et al., 2012; Maheswaran et al.. 2010); Table 7-241. A similar lack of
difference between smokers and nonsmokers was observed for NO2 associations with
lung cancer incidence (Raaschou-Nielsen et al.. 2011; Raaschou-Nielsen et al.. 2010).
January 2015
7-61
DRAFT: Do Not Cite or Quote
-------
1
2
o
J
4
5
6
7
8
9
10
11
12
13
although the association with lung cancer mortality was larger in smokers when limited
to males (Katanoda et al.. 2011).
A controlled human exposure study demonstrated larger NCh-induced decrements in lung
function among smokers compared to nonsmokers, but epidemiologic evidence is not
available for respiratory effects. Most epidemiologic studies examine NCh-related
cardiovascular effects, diabetes, or mortality and do not report differences between
smokers and nonsmokers. Many studies similarly defined smoking as current or former
smoking, providing a basis for comparisons across studies. Although there is lack of
evidence for differences in risk for NC>2-related health effects by smoking status, there is
also uncertainty as to whether NO2 has an independent relationship with cardiovascular
effects, diabetes (Section 6.3.9). and mortality (Section 6.5.3). the health effects for
which smoking status was examined. Therefore, the evidence is inadequate to determine
whether smoking increases the risk of NC>2-related health effects.
Table 7-24 Epidemiologic studies evaluating smoking status.
Factor Reference
Evaluated Category
Direction of
Effect
Modification3 Outcome
Study
Population
Study Details
Study
Short-term exposure
Current or Never
former smoking
smoking n = 28.4%
n = 74%
i Change in
"*" ventricular
repolarization
N = 580 males
Mean age 75 yr
Boston, MA,
Follow-up:
2000-2008
Baia et al.
(2010)
Long-term exposure
Current or Never
former smokingb
smokingb
Current or Never
former smoking
smoking n = 396,647
n = 313,487
Current or Never
former smoking
smoking
n= 28,612 n = 15,824
Current or Never
former smoking
_ Incident Ml
_ Heart failure
-j- Hypertension
' (change in
systolic BP)
_ Incident stroke
N = 84,562
Ages 30-55 yr at
enrollment
N = 836,557
Ages 40-89 yr in
2003
N = 44,436
Ages 50-65 yr at
baseline
N = 1,984
Ages 50-65 yr at
U.S.,
1990-2008
England,
2003-2007
Copenhagen,
Aarhus counties,
Denmark,
1993-2006
Copenhagen,
Aarhus counties,
Hartetal.
(2013)
Atkinson et al.
(2013)
S0rensen
etal. (2012)
Andersen
etal. (2012a)
smoking n = 481
n = 1,503
Current or Never
former smoking
smoking n = 24
n = 118
baseline Denmark,
1993-2006
Fatal stroke
N = 142
Ages 50-65 yr at
baseline
January 2015
7-62
DRAFT: Do Not Cite or Quote
-------
Table 7-24 (Continued) Epidemiologic studies evaluating smoking status.
Factor
Evaluated
Current or
former
smoking
n = 33,380
Current or
former
smoking
n = 45%
Current or
former
smoking
n = 45.5%
Current
smoking
n = 90
Current or
former
smoking
n = 917
Current or
former
smoking
n=25.5-
65% across
cohorts
Current or
former
smoking
n= 2,850
Current or
former
smoking
n = 64%
Current
smoking
n=608
Direction of
Reference Effect
Category Modification3 Outcome
Never
smoking
n = 18,438
Never
smoking
n = 55%
Never
smoking
n = 54.5%
Former
smoking
n = 152
Never
smoking
n=619
Never
smoking
n = 35-
74.5%
across
cohorts
Never
smoking
n= 4,359
Never
smoking
n = 36%
Never or
former
smoking
n = 1,248
1
^™
1
-
-
-
1
-
^™
t
1
1
Diabetes
Atherosclerosis
(carotid artery
intima-media
thickness
Atherosclerosis
(carotid intima
media thickness)
C-reactive protein
TNF-a
IL-6
IL-8
Fibrinogen
Hepatocyte
growth factor
Blood
IL-6 levels
Respiratory
mortality
CVD Mortality
Diabetes-related
mortality
Total mortality
Study
Population
N = 51,818
Ages 50-65 yr at
baseline
N = 745
Ages 26-30 yr
N = 2,780
Median age 58 yr
N = 242 adults
with clinically
stable COPD
Mean age 68 yr
N = 1,536
Ages 45-70 yr
N = 307,553
Mean age at
baseline 41.9 to
73.0 yr across
16 cohorts
N = 9,941, 256
deaths
Ages 35-103 yr
N = 52,061
Ages 50-64 yr
N = 3,320
Mean age 70 yr
Study Details
Copenhagen,
Aarhus counties,
Denmark,
1993-2006
Utrecht, the
Netherlands,
1999-2000
Girona Province,
Spain,
2007-2010
Barcelona,
Spain,
2004-2006
Stockholm
county, Sweden,
1992-1994
Europe
Follow-up:
1985-2007
NO2 exposure
assessed for
2008-2011
Shenyang,
China
Follow-up:
1998-2009
NO2 exposure
assessed for
1998-2009
Denmark
Follow-up:
1971-2009. NO2
exposure
assessed for
1971-2009
London, England
Follow-up:
1995-2005
NO2 assessed
for 2002
Study
Andersen
etal. (2012b)
Lenters et al.
(2010)
Rivera et al.
(2013)
Dadvand et al.
(2014)
Panasevich
et al. (2009)
Dimakopoulou
etal. (2014)
Zhana et al.
(2011)
Raaschou-
Nielsen et al.
(2012)
Maheswaran
etal. (2010)
January 2015
7-63
DRAFT: Do Not Cite or Quote
-------
Table 7-24 (Continued) Epidemiologic studies evaluating smoking status.
Factor
Evaluated
Current or
former
smoking
n = 322,766
Current
smoking,
male
n=292
deaths
Current or
former
smoking
n = 3,713
Current or
former
smoking
n = 3,372
Current or
former
smoking
n = 19,253
Reference
Category
Never
smoking
n = 386,591
Former
smoking,
male
n = 90
deaths
No smoking
n = 9,135
No smoking
n = 109
No smoking
n = 33,717
Direction of
Effect
Modification3 Outcome
i Total mortality
* Lung cancer
' mortality
i Lung cancer or
^ cardiopulmonary
mortality
_ Lung cancer
incidence
_ Lung cancer
incidence
Study
Population
N = 835,607
deaths
Ages 40-89 yr
N = 63,520
Ages >40 yr
N = 14,001
Ages >65 yr
N = 3,481
Ages 20-93 yr at
enrollment
N = 52,970
Ages 50-64 yr
Study Details
England
Follow-up:
2003-2007
NO2 exposure
assessed for
2002
3 Japanese
prefectures,
1983-1985
Shizuoka,
Japan,
1999-2006
Copenhagen,
Aarhus counties,
Denmark,
1970-1997
Copenhagen,
Aarhus counties,
Denmark,
1993-1997
Study
Carev et al.
(2013)
Katanoda
etal. (2011)
Yorifuii et al.
(2010)
Raaschou-
Nielsen et al.
(2010)
Raaschou-
Nielsen et al.
(2011)
aUp facing arrow indicates that the effect of NO2 is greater (e.g., larger risk hypertension) in the group with the factor evaluated than in
the reference group. Down facing arrow indicates that the effect of NO2 is smaller in the group with the factor evaluated than in the
reference group. A dash indicates no difference in NO2-related health effect between groups.
bSample size not reported.
7.6.3 Physical Activity
i
2
o
5
4
5
6
7
8
9
10
11
12
There is some evidence indicating that during physical activity, increased respiratory rate
and oronasal breathing can increase the deposition of NO2 in the lower respiratory tract
(Section 4.2.2). which could have implications for increasing the risk of NCh-related
health effects. However, the effect of concurrent physical activity on NC^-related health
effects has not been characterized. Rather, physical activity has been examined as a
modifier of health effects related to long-term NO2 exposure, as an indicator of active
versus sedentary lifestyle or fitness. Further, outdoor activity has not been assessed. The
influence of general activity or fitness on NCh exposure and internal dose are not known.
Epidemiologic studies examined physical activity or exercise as a modifier of
cardiovascular effects and mortality, for which independent relationships with NC>2 are
uncertain (Sections 6.3.9 and 6.5.3). These studies have found inconsistent results with
respect to whether physical activity increases the risk for NC>2-related health effects
January 2015
7-64
DRAFT: Do Not Cite or Quote
-------
1 (Table 7-25). Associations between long-term NO2 exposure and mortality from diabetes
2 was higher in the group not engaging in exercise (Raaschou-Nielsen et al.. 2012). but risk
3 of diabetes was similar or lower among those with low levels of physical activity (Eze
4 et al.. 2014; Andersen et al.. 2012b). Similarly, NCh-related cardiovascular mortality was
5 greater in the group with no exercise (Zhang et al.. 2011). but associations with other
6 cardiovascular effects were similar between groups with low or high physical activity
7 (Hart etal. 2013; Panasevich et al.. 2009). Contributing to the uncertainty in the
8 evidence base is the heterogeneity across studies in how physical activity was defined, for
9 example, the frequency or intensity of activity. Overall, there is inconsistent evidence
10 indicating that physical activity increases the risk for NO2-related health effects.
11 Therefore, the evidence is inadequate to determine whether low physical activity
12 increases the risk for NO2-related health effects.
January 2015 7-65 DRAFT: Do Not Cite or Quote
-------
Table 7-25
Factor
Evaluated
Low physical
activity
(<18METS/
week)b
Physical
inactivity
n= 23,536
Low physical
activity (<0.5 h/
week)
n = 38%
Physical
inactivity
(inactive
leisure time)
n = 543
No Exercise
n = 45.7%
No Exercise
n = 5,795
Epidemiologic studies evaluating physical activity (all long-term
exposure).
Direction of
Reference Effect
Category Modification3 Outcome
High physical _ Incident Ml
activity
(>18METS/
week)b
Physical i Diabetes
activity or ^
playing sports
in leisure time
n = 28,282
Physical _ Diabetes
activity
(>2 h/week)
n = 28%
Physical _ Blood IL-6
activity levels
n = 993
Exercise * Diabetes-
n = 54.3% ' related
mortality
Exercise * CV mortality
n = 4,146 !
Study
Population
N = 84,562
Age 30-55 yr
at enrollment
N = 51,818
Ages 50-65 yr
at baseline
N = 6,392
Ages 29-73 yr
N = 1,536
Age 45-70 yr
N = 52,061
Ages 50-64 yr
N = 9,941, 256
deaths
Ages
35-1 03 yr
Study Details
U.S.,
1990-2008
Copenhagen,
Aarhus counties,
Denmark,
1993-2006
Switzerland,
2002
Stockholm county,
Sweden,
1992-1994
Denmark
Follow-up and NO2
exposure assessed
for 1971 -2009
Shenyang, China
Follow-up and NO2
exposure assessed
for 1998-2009
Study
Hartetal.
(2013)
Andersen
etal. (2012b)
Eze et al.
(2014)
Panasevich
et al. (2009)
Raaschou-
Nielsen et al.
(2012)
Zhanq et al.
(2011)
aUp facing arrow indicates that the effect of NO2 is greater (e.g., larger risk of mortality) in the group with the factor evaluated than in
the reference group. Down facing arrow indicates that the effect of NO2 is smaller in the group with the factor evaluated than in the
reference group. A dash indicates no difference in NO2-related health effect between groups.
bSample size not reported.
7.7 Conclusions
i
2
3
4
5
6
This chapter evaluates factors that may characterize populations and lifestages at
increased risk for health effects related to NO2 exposure (Table 7-26). The evidence for
each factor was classified based on judgments of the consistency, coherence, and
biological plausibility of evidence integrated across epidemiologic, controlled human
exposure, and toxicological studies using the weight-of-evidence approach detailed in
Table 7-1. The evaluation also drew upon information presented in preceding chapters on
January 2015
7-66
DRAFT: Do Not Cite or Quote
-------
1 exposure, dosimetry, modes of action, and independent relationships of NCh exposure
2 with health effects.
3 Consistent with observations made in the 2008 ISA for Oxides of Nitrogen (U.S. EPA.
4 2008). there is adequate evidence to conclude that people with asthma, children, and
5 older adults are at increased risk for NO2-related health effects. Not only does evidence
6 consistently indicate increased risk for these groups, but the evidence is based on findings
7 for short-term NO2 exposure and respiratory effects (particularly asthma exacerbation),
8 for which a causal relationship exists (Section 5.2.9). In addition to the strong evidence
9 for a relationship between short-term NO2 exposure and asthma exacerbation, the
10 conclusion that people with asthma are at increased risk of NO2-related health effects is
11 supported by results from a meta-analysis of controlled human exposure studies
12 demonstrating that NO2 exposure increases airway responsiveness, a key feature of
13 asthma exacerbation, at lower concentrations in people with asthma compared to healthy
14 individuals. Epidemiologic evidence does not consistently demonstrate differences in
15 NO2-related respiratory effects in people with asthma. It is important to note that there is
16 evidence of heterogeneity in asthma severity and triggers within study populations; thus,
17 the epidemiologic evidence is not considered to be in conflict with experimental
18 evidence. Children and older adults consistently have larger magnitude associations
19 between NO2 exposure and asthma hospital admissions and ED visits, compared to adults
20 or the general population. There is not clear evidence from controlled human exposure
21 studies that NO2 induces respiratory effects in healthy, older adults, but examination is
22 much more limited compared with epidemiologic evidence. Time-activity patterns and
23 ventilation rates differ among age groups, but it is not understood whether these factors
24 contribute to the increased risk of NO2-related health effects for children and older adults.
January 2015 7-67 DRAFT: Do Not Cite or Quote
-------
Table 7-26 Summary of evidence for potential increased NO2 exposure and
increased risk of NO2-related health effects.
Evidence
Classification
Factor Evaluated
Rationale for Classification
Adequate Asthma (Section 7.3.1)
evidence Lifestage: (Section 7.5.1.1): Children
(Section 7.5.1.2): Older adults
Each factor: consistent evidence for
increased risk for NCb-related asthma
exacerbation.
Asthma: evidence from controlled human
exposure studies.
Lifestage: different time-activity patterns
and ventilation patterns but unclear
implications for differences in NO2
exposure or internal dose.
Suggestive SES (Section 7.5.2): Low SES
evidence Sex (Section 7.5.4): Females
Diet (Section 7.6.1): Reduced antioxidant intake
Each factor: limited and generally
supporting evidence for differences in
NO2-related health effects.
SES and females: findings based primarily
on short-term NO2 exposure and mortality
for SES and long-term NO2 exposure and
lung function for females. Uncertainty in
independent relationships with NO2 for
some health effects provides limited basis
for inferences about differential risk.
Reduced dietary antioxidant vitamin intake:
consistent evidence from experimental
studies for modification of NCb-related
respiratory effects, but changes in oxidant
balance may not necessarily indicate health
effects.
Inadequate COPD (Section 7.3.2)
evidence Cardiovascular disease (Section 7.3.3)
Diabetes (Section 7.3.4)
Genetic background (Section 7.4)
Obesity (Section 7.3.5)
Smoking (Section 7.6.2)
Physical activity (Section 7.6.3)
Race/ethnicity (Section 7.5.3)
Residence in urban areas (Section 7.5.5)
Proximity to roadways (Section 7.5.6)
Epidemiologic findings inconsistently show
differences in NO2-related health effects,
show no difference, or are limited in
quantity.
Findings based primarily on cardiovascular
effects, diabetes, birth outcomes, and
mortality. Uncertainty in independent
relationships with NO2 provides limited
basis for inferences about differential risk.
Indication of higher NO2 exposure among
nonwhite populations, urban residents, and
people spending time or living near
roadways, but insufficient information to
assess increased risk of NO2-related health
effects.
Evidence of no
effect
None
1
2
o
J
4
There is suggestive evidence that people with low antioxidant diets, people of low SES,
and females are at increased risk for NCh-related health effects because of some
uncertainties in the evidence bases. While experimental studies indicate that dietary
intake of Vitamin C or E modifies NC^-related effects on airway responsiveness, much of
January 2015
7-68
DRAFT: Do Not Cite or Quote
-------
1 the evidence is for effects on oxidant balance, which is not necessarily indicative of
2 health effects. Evidence indicates that low SES populations have higher NO2 exposure
3 and larger NCh-related risk of mortality. For females, limited epidemiologic evidence
4 points to greater risk for NCh-related decrements in lung function. However, for low SES
5 populations and females, the evidence is based on studies of health effects for which
6 independent relationships with NCh exposure are uncertain (Sections 5.4.8 and 6.2.9).
7 There is inadequate evidence to determine whether pre-existing cardiovascular disease,
8 diabetes, COPD, genetic variants, obesity, smoking, or physically active lifestyle
9 increases the risk for NC^-related health effects. Studies show either inconsistent or no
10 modification of NC>2-related health effects by these factors, and information is based
11 primarily on cardiovascular effects (Sections 5.3.12 and 6.3.9) and mortality
12 (Sections 5.4.8 and 6.5.3) for which independent relationships with NC>2 are uncertain.
13 Evidence also is inadequate to determine whether race/ethnicity, urban residence, or
14 proximity to roadways increase the risk for NC>2-related health effects. While nonwhite
15 populations, urban residents, and people with close proximity to roadways (i.e., living,
16 attending school, working, or commuting on or near roadways) may have increased
17 exposure to NCh, there is limited or inconsistent evidence for larger NC^-related health
18 effects in these populations. Further, inferences about the potential differential risk for
19 these populations are limited by evidence that is based on cardiovascular effects
20 (Section 6.3.9) and birth outcomes (Section 6.4.5). for which independent effects of NC>2
21 exposure are uncertain. Additionally, it is important to note that many factors may be
22 acting in combination, which may lead to a different public health impact than is
23 reflected when evaluating any one factor in isolation. However, at this time information
24 remains limited as to the impact of multiple factors and how they affect the risk for
25 NO2-related health effects.
26 In conclusion, evidence is adequate to conclude that people with asthma, children, and
27 older adults are at increased risk for NC^-related health effects. The large proportions of
28 the U.S. potentially that encompass each of these groups and lifestages underscores the
29 potential public health significance of NC^-related health effects.
January 2015 7-69 DRAFT: Do Not Cite or Quote
-------
References for Chapter 7
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. http://dx.doi.Org/10.1164/ajrccm.159.2.9806020
Amigou. A: Sermage-Faure. C: Orsi. L: Leverger. G: Baruchel A: Bertrand. Y: Nelken. B: Robert. A:
Michel. G: Margueritte. G: Perel. Y: Mechinaud. F: Bordigoni. P: Hemon. D: Clavel. J. (2011). Road
traffic and childhood leukemia: The ESC ALE Study (SFCE). Environ Health Perspect 119: 566-572.
http://dx.doi.org/10.1289/ehp.1002429
Andersen. ZJ: Kristiansen. LC: Andersen. KK: Olsen. TS: Hvidberg. M: Jensen. SS: Ketzel. M: Loft S:
S0rensen. M: Tj0nneland. A: Overvad. K: Raaschou-Nielsen. O. (2012a). Stroke and long-term
exposure to outdoor air pollution from nitrogen dioxide: a cohort study. Stroke 43: 320-325.
http://dx.doi.org/10.1161/STROKEAHA.lll.629246
Andersen. ZJ: Raaschou-Nielsen. O: Ketzel M: Jensen. SS: Hvidberg. M: Loft. S: Tj0nneland. A:
Overvad. K: S0rensen. M. (2012b). 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
Anderson. HR: Ponce de Leon. A: Bland. JM: Bower. JS: Emberlin. J: Strachen. DP. (1998). Air
pollution, pollens, and daily admissions for asthma in London 1987-92. Thorax 53: 842-848.
Arbex. MA: de Souza Conceicao. GM: Cendon. SP: Arbex. FF: Lopes. AC: Moyses. EP: Santiago. SL:
Saldiva. PHN: Pereira. LAA: Braga. ALF. (2009). Urban air pollution and chronic obstructive
pulmonary disease-related emergency department visits. J Epidemiol Community Health 63: 777-783.
http://dx.doi.org/10.1136/iech.2008.078360
Atkinson. RW: Anderson. HR: Strachan. DP: Bland. JM: Bremner. SA: Ponce de Leon. A. (1999). Short-
term associations between outdoor air pollution and visits to accident and emergency departments in
London for respiratory complaints. Eur Respir J 13: 257-265.
Atkinson. RW: Carey. IM: Kent. AJ: van Staa. TP: Anderson. HR: Cook. DG. (2013). Long-term
exposure to outdoor air pollution and incidence of cardiovascular diseases. Epidemiology 24: 44-53.
http://dx.doi.org/10.1097/EDE.Ob013e318276ccb8
Ayaz. KL: Csallany. AS. (1978). Long-term NO2 exposure of mice in the presence and absence of
vitamin E. II. Effect of glutathione peroxidase. Arch Environ Occup Health 33: 292-296.
Azoulav-Dupuis. E: Torres. M: Soler. P: Moreau. J. (1983). Pulmonary NO2 toxicity in neonate and adult
guinea pigs and rats. Environ Res 30: 322-339.
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
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
Bauer. MA: Utell. MJ: Morrow. PE: Speers. DM: Gibb. FR. (1986). Inhalation of 0.30 ppm nitrogen
dioxide potentiates exercise-induced bronchospasm in asthmatics. Am Rev Respir Dis 134: 1203-
1208.
Becerra. TA: Wilhelm. M: Olsen. J: Cockburn. M: Ritz. B. (2013). Ambient air pollution and autism in
Los Angeles county, California. Environ Health Perspect 121: 380-386.
http://dx.doi.org/10.1289/ehp.1205827
January 2015 7-70 DRAFT: Do Not Cite or Quote
-------
Beelen. R: Stafoggia. M: Raaschou-Nielsen. O: Andersen. ZJ: Xun. WW: Katsouvanni. K:
Dimakopoulou. K: Brunekreef. B: Weinmayr. G: Hoffmann. B: Wolf. K: Samoli. E: Houthuijs. D:
Nieuwenhuijsen. M: Oudin. A: Forsberg. B: Olsson. D: Salomaa. V: Lanki. T: Yli-Tuomi. T: Oftedal
B: Aamodt. G: Nafstad. P: De Faire. U: Pedersen. NL: Ostenson. CG: Fratiglioni. L: Penell J: Korek.
M: Pyko. A: Eriksen. KT: Tj0nneland. A: Becker. T: Eeftens. M: Bots. M: Meliefste. K: Wang. M:
Bueno-De-Mesquita. B: Sugiri. D: Kramer. U: Heinrich. J: de Hoogh. K: Key. T: Peters. A: Cyrys. J:
Concin. H: Nagel. G: Ineichen. A: Schaffner. E: Probst-Hensch. N: Dratva. J: Ducret-Stich. R: Vilier.
A: Clavel-Chapelon. F: Stempfelet M: Grioni. S: Krogh. V: Tsai. MY: Marcon. A: Ricceri. F:
Sacerdote. C: Galassi C: Migliore. E: Ranzi. A: Cesaroni. G: Badaloni. C: Forastiere. F: Tamavo. I:
Amiano. P: Dorronsoro. M: Katsoulis. M: Trichopoulou. A: Vineis. P: Hoek. G. (2014). Long-term
exposure to air pollution and cardiovascular mortality: An analysis of 22 European cohorts.
Epidemiology 25: 368-378. http://dx.doi.org/10.1097/EDE.0000000000000076
Bell. ML: Ebisu. K: Belanger. K. (2007). Ambient air pollution and low birth weight in Connecticut and
Massachusetts. Environ Health Perspect 115: 1118-1124. http://dx.doi.org/10.1289/ehp.9759
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
Bilenko. N: Rossem. LV: Brunekreef. B: Beelen. R: Eeftens. M: Hoek. G: Houthuijs. D: de Jongste. JC:
Kempen. EV: Koppelman. GH: Meliefste. K: Oldenwening. M: Smit HA: Wijga. AH: Gehring. U.
(2013). Traffic-related air pollution and noise and children's blood pressure: results from the PIAMA
birth cohort study. European Journal of Preventive Cardiology.
http://dx.doi.org/10.1177/2047487313505821
Bilenko. N: Rossem. LV: Brunekreef. B: Beelen. R: Eeftens. M: Hoek. G: Houthuijs. D: de Jongste. JC:
Kempen. EV: Koppelman. GH: Meliefste. K: Oldenwening. M: Smit. HA: Wij ga. AH: Gehring. U.
(2015). Traffic-related air pollution and noise and children's blood pressure: results from the PIAMA
birth cohort study. European Journal of Preventive Cardiology 22: 4-12.
http://dx.doi.org/10.1177/2047487313505821
Bishaw. A. (2012). Poverty: 2010 and 2011. Washington, DC: U.S. Department of Commerce, U.S.
Census Bureau, http://www.census.gov/prod/2012pubs/acsbrll-01.pdf
Bloom. B: Cohen. RA: Freeman. G. (2011). Summary health statistics for U.S. children: National Health
Interview Survey, 2010. Vital Health Stat 10 0: 1-80.
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
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/j.envres.2009.03.004
Cakmak. S: Dales. RE: Rubio. MA: Vidal CB. (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/j.envres.2011.01.003
Campen. MJ: Lund. AK: Doyle-Eisele. ML: Mcdonald. JD: Knuckles. TL: Rohr. AC: Knipping. EM:
Mauderly. JL. (2010). A comparison of vascular effects from complex and individual air pollutants
indicates a role for monoxide gases and volatile hydrocarbons. Environ Health Perspect 118: 921-927.
http://dx.doi.org/10.1289/ehp.0901207
Carey. IM: Atkinson. RW: Kent. AJ: van Staa. T: Cook. DG: Anderson. HR. (2013). Mortality
associations with long-term exposure to outdoor air pollution in a national english cohort. Am J Respir
Crit Care Med 187: 1226-1233. http://dx.doi.org/10.1164/rccm.201210-1758OC
January 2015 7-71 DRAFT: Do Not Cite or Quote
-------
Castro-Giner. F: Kiinzli. N: Jacquemin. B: Forsberg. B: de Cid. R: Sunyer. J: Jarvis. D: Briggs. D:
Vienneau. D: Norback. D: Gonzalez. JR: Guerra. S: Janson. C: Anto. JM: Wist M: Heinrich. J:
Estivill. X: Kogevinas. M. (2009). Traffic-related air pollution, oxidative stress genes, and asthma
(ECHRS). Environ Health Perspect 117: 1919-1924. http://dx.doi.org/10.1289/ehp.0900589
Cesaroni. G: Badaloni. C: Gariazzo. C: Stafoggia. M: Sozzi. R: Davoli M: Forastiere. F. (2013). Long-
term exposure to urban air pollution and mortality in a cohort of more than a million adults in rome.
Environ Health Perspect 121: 324-331. http://dx.doi.org/10.1289/ehp.1205862
Chaix. B: Gustafsson. S: Jerrett M: Kristersson. H: Lithman. T: Boalt. A: Merlo. J. (2006). Children's
exposure to nitrogen dioxide in Sweden: Investigating environmental injustice in an egalitarian
country. J Epidemiol Community Health 60: 234-241. http://dx.doi.org/10.1136/jech.2005.038190
Chen. R: Samoli. E: Wong. CM: Huang. W: Wang. Z: Chen. B: Kan. H. (2012). Associations between
short-term exposure to nitrogen dioxide and mortality in 17 Chinese cities: The China Air Pollution
and Health Effects Study (CAPES). Environ Int 45: 32-38.
http://dx.doi.0rg/10.1016/j.envint.2012.04.008
Chiusolo. M: Cadum. E: Stafoggia. M: Galassi. C: Berti. G: Faustini A: Bisanti L: Vigotti. MA: Dessi.
MP: Cernigliaro. A: Mallone. S: Pacelli. B: Minerba. S: Simonato. L: Forastiere. F. (2011). Short term
effects of nitrogen dioxide on mortality and susceptibility factors in 10 Italian cities: The EpiAir
Study. Environ Health Perspect 119: 1233-1238. http://dx.doi.org/10.1289/ehp. 1002904
Clark. LP: Millet DB: Marshall JD. (2014). National patterns in environmental injustice and inequality:
outdoor NO2 air pollution in the United States. PLoS ONE 9: e94431.
http://dx.doi.org/10.1371/journal.pone.0094431
Cloughertv. JE: Levy. JI: Kubzansky. LD: Ryan. PB: Suglia. SF: Canner. MJ: Wright. RJ. (2007).
Synergistic effects of traffic-related air pollution and exposure to violence on urban asthma etiology.
Environ Health Perspect 115: 1140-1146. http://dx.doi.org/10.1289/ehp.9863
Grouse. PL: Ross. NA: Goldberg. MS. (2009). Double burden of deprivation and high concentrations of
ambient air pollution at the neighbourhood scale in Montreal, Canada. Soc Sci Med 69: 971-981.
http://dx.doi.0rg/10.1016/j.socscimed.2009.07.010
D'Ippoliti D: Forastiere. F: Ancona. C: Agabiti. N: Fusco. D: Michelozzi. P: Perucci. CA. (2003). Air
pollution and myocardial infarction in Rome: a case-crossover analysis. Epidemiology 14: 528-535.
http://dx.doi.org/10.1097/01.ede.0000082046.22919.72
Dadvand. P: Nieuwenhuijsen. MJ: Agusti. A: de Batlle. J: Benet M: Beelen. R: Cirach. M: Martinez. D:
Hoek. G: Basagana. X: Ferrer. A: Ferrer. J: Rodriguez-Roisin. R: Sauleda. J: Guerra. S: Anto. JM:
Garcia-Aymerich. J. (2014). Air pollution and biomarkers of systemic inflammation and tissue repair
in COPD patients. Eur Respir J 44: 603-613. http://dx.doi.org/10.1183/09031936.00168813
Darrow. LA: Klein. M: Strickland. MJ: Mulholland. JA: Tolbert. PE. (2011). 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
Deguen. S: Zmirou-Navier. D. (2010). Social inequalities resulting from health risks related to ambient air
quality—A European review [Review]. Eur J Public Health 20: 27-35.
http://dx.doi.org/10.1093/eurpub/ckp220
Delfino. RJ: Staimer. N: Gillen. D: Tjoa. T: Sioutas. C: Fung. K: George. SC: Kleinman. MT. (2006).
Personal and ambient air pollution is associated with increased exhaled nitric oxide in children with
asthma. Environ Health Perspect 114: 1736-1743. http://dx.doi.org/10.1289/ehp.9141
Delgado-Saborit. JM. (2012). Use of real-time sensors to characterise human exposures to combustion
related pollutants. J Environ Monit 14: 1824-1837. http://dx.doi.org/10.1039/c2eml0996d
January 2015 7-72 DRAFT: Do Not Cite or Quote
-------
Dimakopoulou. K: Samolj E: Beelen. R: Stafoggia. M: Andersen. ZJ: Hoffmann. B: Fischer. P:
Nieuwenhuijsen. M: Vineis. P: Xun. W: Hoek. G: Raaschou-Nielsen. O: Oudin. A: Forsberg. B:
Modig. L: Jousilahti P: Lankj T: Turunen. A: Oftedal B: Nafstad. P: Schwarze. PE: Penell J:
Fratiglionj L: Andersson. N: Pedersen. N: Korek. M: De Faire. U: Eriksen. KT: Tj0nneland. A:
Becker. T: Wang. M: Bueno-De-Mesquita. B: Tsai MY: Eeftens. M: Peelers. PH: Meliefste. K:
Marcon. A: Kramer. U: Kuhlbusch. TA: Vossoughj M: Key. T: de Hoogh. K: Hampel R: Peters. A:
Heinrich. J: Weinmayr. G: Concin. H: Nagel. G: Ineichen. A: Jacquemin. B: Stempfelet. M: Vilier. A:
Riccerj F: Sacerdote. C: Pedeli X: Katsoulis. M: Trichopoulou. A: Brunekreef. B: Katsouvannj K.
(2014). Air pollution and nonmalignant respiratory mortality in 16 cohorts within the ESCAPE
project. Am J Respir Crit Care Med 189: 684-696. http://dx.doi.org/10.1164/rccm.201310-1777OC
Dong. G: Qian. Z: Wang. J: Chen. W: Ma. W: Trevathan. E: Xaverius. PK: Declue. R: Wiese. A:
Langston. M: Liu. MM: Wang. D: Ren. W. (2013a). Associations between ambient air pollution and
prevalence of stroke and cardiovascular diseases in 33 Chinese communities. Atmos Environ 77: 968-
973. http://dx.doi.0rg/10.1016/j.atmosenv.2013.06.034
Dong. GH: Qian. ZM: Xaverius. PK: Trevathan. E: Maalouf. S: Parker. J: Yang. L: Liu. MM: Wang. D:
Ren. WH: Ma. W: Wang. J: Zelicoff. A: Fu. Q: Simckes. M. (2013b). Association between long-term
air pollution and increased blood pressure and hypertension in China. Hypertension 61: 578-584.
http://dx.doi.org/10.1161/HYPERTENSIONAHA.lll.00003
Elsaved. NM: Mustafa. MG. (1982). Dietary antioxidants and the biochemical response to oxidant
inhalation: I: Influence of dietary vitamin E on the biochemical effects of nitrogen dioxide exposure in
rat lung. Toxicol Appl Pharmacol 66: 319-328. http://dx.doi.org/10.1016/0041-008X(82)90298-8
Eze. 1C: Schaffner. E: Fischer. E: Schikowski. T: Adam. M: Imboden. M: Tsai. M: Carballo. D: von
Eckardstein. A: Kiinzlj N: Schindler. C: Probst-Hensch. N. (2014). Long-term air pollution exposure
and diabetes in a population-based Swiss cohort. Environ Int 70: 95-105.
http://dx.doi.0rg/10.1016/j.envint.2014.05.014
Faustinj A: Stafoggia. M: Colais. P: Berti G: Bisanti L: Cadum. E: Cernigliaro. A: Mallone. S: Scarnato.
C: Forastiere. F. (2013). Air pollution and multiple acute respiratory outcomes. Eur Respir J 42: 304-
313. http://dx.doi.org/10.1183/09031936.00128712
Felber Dietrich. D: Gemperlj A: Gaspoz. JM: Schindler. C: Liu. LJ: Gold. PR: Schwartz. J: Rochat T:
Barthelemy. JC: Pons. M: Roche. F: Probst Hensch. NM: Bridevaux. PO: Gerbase. MW: Neu. U:
Ackermann-Liebrich. U. (2008). Differences in heart rate variability associated with long-term
exposure to NO2. Environ Health Perspect 116: 1357-1361. http://dx.doi.org/10.1289/ehp.11377
Filho. MAP: Pereira. LAA: Arbex. FF: Arbex. M: Conceicao. GM: Santos. UP: Lopes. AC: Saldiva.
PHN: Braga. ALF: Cendon. S. (2008). Effect of air pollution on diabetes and cardiovascular diseases
in Sao Paulo, Brazil. Braz J Med Biol Res 41: 526-532. http://dx.doi.org/10.1590/S0100-
879X2008005000020
Flamant-Hulin. M: Caillaud. D: Sacco. P: Penard-Morand. C: Annesi-Maesano. I. (2010). Air pollution
and increased levels of fractional exhaled nitric oxide in children with no history of airway damage. J
Toxicol Environ Health A 73: 272-283. http://dx.doi.org/10.1080/15287390903249206
Folinsbee. LJ. (1992). Does nitrogen dioxide exposure increase airways responsiveness? Toxicol Ind
Health 8: 273-283.
Foraster. M: Basagana. X: Aguilera. I: Rivera. M: Agis. D: Bouso. L: Deltell A: Marrugat J: Ramos. R:
Sunyer. J: Vila. J: Elosua. R: Kiinzli. N. (2014). Association of long-term exposure to traffic-related
air pollution with blood pressure and hypertension in an adult population-based cohort in Spain (the
REGICOR Study). Environ Health Perspect 122: 404-411. http://dx.doi.org/10.1289/ehp. 1306497
Fu. A: Leaderer. BP: Gent. JF: Leaderer. D: Zhu. Y. (2012). An environmental epigenetic study of
ADRB2 5'-UTR methylation and childhood asthma severity. Clin Exp Allergy 42: 1575-1581.
http://dx.doi.0rg/10.llll/i.1365-2222.2012.04055.x
January 2015 7-73 DRAFT: Do Not Cite or Quote
-------
Gauderman. WJ: Avol E: Lurmann. F: Kuenzli. N: Gilliland. F: Peters. J: McConnell. R. (2005).
Childhood asthma and exposure to traffic and nitrogen dioxide. Epidemiology 16: 737-743.
http://dx.doi.org/10.1097/01.ede.0000181308.51440.75
Gauderman. WJ: Vora. H: McConnell. R: Berhane. K: Gilliland. F: Thomas. D: Lurmann. F: Avol. E:
Kunzli. N: Jerrett M: Peters. J. (2007). Effect of exposure to traffic on lung development from 10 to
18 years of age: Acohort study. Lancet 369: 571-577. http://dx.doi.org/10.1016/S0140-
6736(07)60037-3
Gehring. U: Cyrys. J: Sedlmeir. G: Brunekreef. B: Bellander. T: Fischer. P: Bauer. CP: Reinhardt. D:
Wichmann. HE: Heinrich. J. (2002). Traffic-related air pollution and respiratory health during the first
2 yrs of life. Eur Respir J 19: 690-698. http://dx.doi.org/10.1183/09031936.02.01182001
Gong. H: Linn. WS: Clark. KW: Anderson. KR: Geller. MD: Sioutas. C. (2005). Respiratory responses to
exposures with fine particulates and nitrogen dioxide in the elderly with and without COPD. Inhal
Toxicol 17: 123-132. http://dx.doi.org/10.1080/08958370590904481
Green. RS: Smorodinsky. S: Kim. JJ: McLaughlin. R: Ostro. B. (2004). Proximity of California public
schools to busy roads. Environ Health Perspect 112: 61-66. http://dx.doi.org/10.1289/ehp.6566
Grineski. SE: Staniswalis. JG: Peng. Y: Atkinson-Palombo. C. (2010). Children's asthma hospitalizations
and relative risk due to nitrogen dioxide (NO2): Effect modification by race, ethnicity, and insurance
status. Environ Res 110: 178-188. http://dx.doi.Org/10.1016/j.envres.2009.10.012
Guxens. M: Aguilera. I: Ballester. F: Estarlich. M: Fernandez-Somoano. A: Lertxundi. A: Lertxundi. N:
Mendez. MA: Tardon. A: Vrijheid. M: Sunyer. J. (2012). Prenatal exposure to residential air pollution
and infant mental development: Modulation by antioxidants and detoxification factors. Environ Health
Perspect 120: 144-149. http://dx.doi.org/10.1289/ehp.1103469
Hart. JE: Rimm. EB: Rexrode. KM: Laden. F. (2013). Changes in traffic exposure and the risk of incident
myocardial infarction and all-cause mortality. Epidemiology 24: 734-742.
http://dx.doi.org/10.1097/EDE.Ob013e31829d5dae
Hatch. GE: Slade. R: Selgrade. MK: Stead. AG. (1986). Nitrogen dioxide exposure and lung antioxidants
in ascorbic acid-deficient guinea pigs. Toxicol Appl Pharmacol 82: 351-359.
http://dx.doi.org/10.1016/0041-008X(86)90212-7
HEI (Health Effects Institute). (2010). Traffic-related air pollution: A critical review of the literature on
emissions, exposure, and health effects. (Special Report 17). Boston, MA: Health Effects Institute
(HEI). http://pubs.healtheffects.org/view.php?id=334
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
Hizawa. N. (2011). Pharmacogenetics of p2-agonists [Review]. Allergol Int 60: 239-246.
http://dx.doi.org/10.2332/allergolint.ll-RAI-0317
Holguin. F: Flores. S: Ross. Z: Cortez. M: Molina. M: Molina. L: Rincon. C: Jerrett. M: Berhane. K:
Granados. A: Romieu. I. (2007). Traffic-related exposures, airway function, inflammation, and
respiratory symptoms in children. Am J Respir Crit Care Med 176: 1236-1242.
http://dx.doi.org/10.1164/rccm.200611-1616OC
Houston. D: Ong. P: Wu. J: Winer. A. (2006). Proximity of licensed child care facilities to near-roadway
vehicle pollution. Am J Public Health 96: 1611-1617. http://dx.doi.org/10.2105/AJPH.2005.077727
Howden. LM: Meyer. JA. (2011). Age and sex composition: 2010. (2010 Census Briefs, C2010BR-03).
Washington, DC: U.S. Department of Commerce, Economics and Statistics Administration, U.S.
Census Bureau. http://www.census.gov/prod/cen2010/briefs/c201 Obr-03.pdf
Hovert. PL: Xu. J. (2012). Deaths: Preliminary data for 2011. National Vital Statistics Reports 61: 1-51.
January 2015 7-74 DRAFT: Do Not Cite or Quote
-------
Huang. W: Zhu. T: Pan. X: Hu. M: Lu. SE: Lin. Y: Wang. T: Zhang. Y: Tang. X. (2012). Air pollution and
autonomic and vascular dysfunction in patients with cardiovascular disease: Interactions of systemic
inflammation, overweight, and gender. Am J Epidemiol 176: 117-126.
http://dx.doi.org/10.1093/aje/kwr511
Humes. KR: Jones. NA: Ramirez. RR. (2011). Overview of race and Hispanic origin: 2010. (2010 Census
Briefs, C2010BR-02). Washington, DC: U.S. Department of Commerce, Economics and Statistics
Administration, U.S. Census Bureau, http://www.census.gov/prod/cen2010/briefs/c2010br-02.pdf
Jenkins. HS: Devalia. JL: Mister. RL: Sevan. AM: Rusznak. C: Davies. RJ. (1999). The effect of
exposure to ozone and nitrogen dioxide on the airway response of atopic asthmatics to inhaled
allergen: dose- and time-dependent effects. Am J Respir Crit Care Med 160: 33-39.
http://dx.doi.0rg/10.1164/ajrccm.160.l.9808119
Jerrett M: Shankardass. K: Berhane. K: Gauderman. WJ: Kiinzli. N: Avol. E: Gilliland. F: Lurmann. F:
Molitor. JN: Molitor. JT: Thomas. DC: Peters. J: Mcconnell. R. (2008). Traffic-related air pollution
and asthma onset in children: A prospective cohort study with individual exposure measurement.
Environ Health Perspect 116: 1433-1438. http://dx.doi.org/10.1289/ehp.10968
Johnson. JYM: Rowe. BH: Allen. RW: Peters. PA: Villeneuve. PJ. (2013). A case-control study of
medium-term exposure to ambient nitrogen dioxide pollution and hospitalization for stroke. BMC
Public Health 13: 368. http://dx.doi.org/10.1186/1471-2458-13-368
Torres. R: Magnussen. H. (1991). Effect of 0.25 ppm nitrogen dioxide on the airway response to
methacholine in asymptomatic asthmatic patients. Lung 169: 77-85.
Torres. R: Nowak. D: Grimminger. F: Seeger. W: Oldigs. M: Magnussen. H. (1995). The effect of 1 ppm
nitrogen dioxide on bronchoalveolar lavage cells and inflammatory mediators in normal and asthmatic
subjects. Eur Respir J 8: 416-424. http://dx.doi.org/10.1183/09031936.95.08030416
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
Katanoda. K: Sobue. T: Satoh. H: Tajima. K: Suzuki. T: Nakatsuka. H: Takezaki. T: Nakavama. T: Nitta.
H: Tanabe. K: Tominaga. S. (2011). An association between long-term exposure to ambient air
pollution and mortality from lung cancer and respiratory diseases in Japan. J Epidemiol 21: 132-143.
http://dx.doi.org/10.2188/iea.JE20100098
Kim. JH: Hong. YC. (2012). GSTM1, GSTT1, and GSTP1 polymorphisms and associations between air
pollutants and markers of insulin resistance in elderly Koreans. Environ Health Perspect 120: 1378-
1384. http://dx.doi.org/10.1289/ehp.1104406
Kim. JJ: Smorodinsky. S: Lipsett M: Singer. BC: Hodgson. AT: Ostro. B. (2004). Traffic-related air
pollution near busy roads: the East Bay children's Respiratory Health Study. Am J Respir Crit Care
Med 170: 520-526. http://dx.doi.org/10.1164/rccm.200403-281OC
Kim. SY: O'Neill MS: Lee. JT: Cho. Y: Kim. J: Kim. H. (2007). Air pollution, socioeconomic position,
and emergency hospital visits for asthma in Seoul, Korea. Int Arch Occup Environ Health 80: 701-
710. http://dx.doi.org/10.1007/s00420-007-0182-3
Kingslev. SL: Eliot. MN: Carlson. L: Finn. J: Macintosh. PL: Suh. HH: Wellenius. GA. (2014). Proximity
of US schools to major roadways: a nationwide assessment. J Expo Sci Environ Epidemiol 24: 253-
259. http://dx.doi.0rg/10.1038/jes.2014.5
Kleinman. MT: Bailey. RM: Linn. WS: Anderson. KR: Whynot JD: Shamoo. DA: Hackney. JD. (1983).
Effects of 0.2 ppm nitrogen dioxide on pulmonary function and response to bronchoprovocation in
asthmatics. J Toxicol Environ Health 12: 815-826. http://dx.doi.org/10.1080/15287398309530472
Ko. FWS: Tarn. W: Wong. TW: Lai. CKW: Wong. GWK: Leung. TF: Ng. SSS: Hui. DSC. (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/i.1365-2222.2007.02791.x
January 2015 7-75 DRAFT: Do Not Cite or Quote
-------
Kousa. A: Monn. C: Rotko. T: Aim. S: Oblesbv. L: Jantunen. MJ. (2001). Personal exposures to NO2 in
the EXPOLIS-study: Relation to residential indoor, outdoor and workplace concentrations in Basel,
Helsinki and Prague. Atmos Environ 35: 3405-3412.
Kruize. H: Driessen. PP: Glasbergen. P: van Egmond. KN. (2007). Environmental equity and the role of
public policy: Experiences in the Rijnmond region. Environ Manage 40: 578-595.
http://dx.doi.org/10.1007/s00267-005-0378-9
Kumae. T: Arakawa. H. (2006). Comparison of effects of in vivo nitrogen dioxide exposure starting from
different periods on alveolar macrophage activity, assessed by a chemiluminescence technique in
Brown-Norway rats. Luminescence 21: 226-232. http://dx.doi.org/10.1002/bio.911
Laurent O: Pedrono. G: Filleul L: Segala. C: Lefranc. A: Schillinger. C: Riviere. E: Bard. D. (2009).
Influence of socioeconomic deprivation on the relation between air pollution and beta-agonist sales for
asthma. Chest 135: 717-723. http://dx.doi.org/10.1378/chest.08-1604
Lee. K: Yang. W: Bofinger. ND. (2000). Impact of microenvironmental nitrogen dioxide concentrations
on personal exposures in Australia. J Air Waste Manag Assoc 50: 1739-1744.
Lenters. V: Uiterwaal CS: Beelen. R: Bots. ML: Fischer. P: Brunekreef. B: Hoek. G. (2010). Long-term
exposure to air pollution and vascular damage in young adults. Epidemiology 21: 512-520.
http://dx.doi.org/10.1097/EDE.Ob013e3181dec3a7
Lin. M: Chen. Y: Villeneuve. PJ: Burnett. RT: Lemyre. L: Hertzman. C: McGrail. KM: Krewski. D.
(2004). Gaseous air pollutants and asthma hospitalization of children with low household income in
Vancouver, British Columbia, Canada. Am J Epidemiol 159: 294-303.
http://dx.doi.org/10.1093/aje/kwh043
Lin. M: Stieb. DM: Chen. Y. (2005). Coarse particulate matter and hospitalization for respiratory
infections in children younger than 15 years in Toronto: A case-crossover analysis. Pediatrics 116:
235-240. http://dx.doi.org/10.1542/peds.2004-2012
Lin. W: Huang. W: Zhu. T: Hu. M: Brunekreef. B: Zhang. Y: Liu. X: Cheng. H: Gehring. U: Li. C: Tang.
2L (2011). Acute respiratory inflammation in children and black carbon in ambient air before and
during the 2008 Beijing Olympics. Environ Health Perspect 119: 1507-1512.
http://dx.doi.org/10.1289/ehp.1103461
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: Spier. CE: Valencia. LM: Anzar. UT: Venet TG: Avol EL: Hackney. JD.
(1985a). Controlled exposure of volunteers with chronic obstructive pulmonary disease to nitrogen
dioxide. Arch Environ Health 40: 313-317. http://dx.doi.org/10.1080/00039896.1985.10545939
Linn. WS: Solomon. JC: Trim. SC: Spier. CE: Shamoo. DA: Venet. TG: Avol EL: Hackney. JD. (1985b).
Effects of exposure to 4 ppm nitrogen dioxide in healthy and asthmatic volunteers. Arch Environ
Occup Health 40: 234-239.
Liu. L: Poon. R: Chen. L: Frescura. AM: Montuschi. P: Ciabattoni. G: Wheeler. A: Dales. R. (2009).
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
Ljungman. PLS: Berglind. N: Holmgren. C: Gadler. F: Edvardsson. N: Pershagen. G: Rosenqvist M:
Sjogren. B: Bellander. T. (2008). Rapid effects of air pollution on ventricular arrhythmias. Eur Heart J
29: 2894-2901. http://dx.doi.org/10.1093/eurheartj/ehn463
Llop. S: Ballester. F: Estarlich. M: Iniguez. C: Ramon. R: Gonzalez. MC: Murcia. M: Esplugues. A:
Rebagliato. M. (2011). Social factors associated with nitrogen dioxide (NO2) exposure during
pregnancy: The INMA-Valencia project in Spain. Soc Sci Med 72: 890-898.
http://dx.doi.0rg/10.1016/i.socscimed.2010.12.018
January 2015 7-76 DRAFT: Do Not Cite or Quote
-------
Luginaah. IN: Fung. KY: Gorev. KM: Webster. G: Wills. C. (2005). Association of ambient air pollution
with respiratory hospitalization in a government-designated "area of concern": The case of Windsor,
Ontario. Environ Health Perspect 113: 290-296. http://dx.doi.org/10.1289/ehp.7300
Madsen. C: Gehring. U: Walker. SE: Brunekreef. B: Stigum. H: Naess. O: Nafstad. P. (2010). Ambient air
pollution exposure, residential mobility and term birth weight in Oslo, Norway. Environ Res 110: 363-
371. http://dx.doi.0rg/10.1016/j.envres.2010.02.005
Maheswaran. R: Pearson. T: Smeeton. NC: Beevers. SD: Campbell MJ: Wolfe. CD. (2010). Impact of
outdoor air pollution on survival after stroke: population-based cohort study. Stroke 41: 869-877.
http://dx.doi.org/10.1161/STROKEAHA.109.567743
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
Mann. JK: Tager. IB: Lurmann. F: Segal M: Quesenberry. CP. Jr: Lugg. MM: Shan. J: Van den Eeden.
SK. (2002). Air pollution and hospital admissions for ischemic heart disease in persons with
congestive heart failure or arrhythmia. Environ Health Perspect 110: 1247-1252.
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. http://dx.doi.org/10.1164/rccm.200304-466OC
McConnell. R: Berhane. K: Molitor. J: Gilliland. F: Kunzli. N: Thorne. PS: Thomas. D: Gauderman. WJ:
Avol. E: Lurmann. F: Rappaport. E: Jerrett M: Peters. J. (2006). Dog ownership enhances
symptomatic responses to air pollution in children with asthma. Environ Health Perspect 114: 1910-
1915. http://dx.doi.org/10.1289/ehp.8548
McConnell R: Liu. FF: Wu. J: Lurmann. F: Peters. J: Berhane. K. (2010). Asthma and School
Commuting Time. J Occup Environ Med 52: 827-829.
http://dx.doi.org/10.1097/JOM.Ob013e3181ebfla9
Meng. Q: Williams. R: Pinto. JP (2012). Determinants of the associations between ambient
concentrations and personal exposures to ambient PM2.5, NO2, and O3 during DEARS. Atmos
Environ 63: 109-116. http://dx.doi.Org/10.1016/j.atmosenv.2012.09.019
Migliaretti G: Cadum. E: Migliore. E: Cavallo. F. (2005). Traffic air pollution and hospital admission for
asthma: a case-control approach in a Turin (Italy) population. Int Arch Occup Environ Health 78: 164-
169. http://dx.doi.org/10.1007/s00420-004-0569-3
Min. KB: Min. JY: Cho. SI: Paek. D. (2008). The relationship between air pollutants and heart-rate
variability among community residents in Korea. Inhal Toxicol 20: 435-444.
http://dx.doi.org/10.1080/08958370801903834
Mitchell. G. (2005). Forecasting environmental equity: Air quality responses to road user charging in
Leeds, UK. J Environ Manage 77: 212-226. http://dx.doi.Org/10.1016/j.jenvman.2005.04.013
Mobasher. Z: Salam. MT: Goodwin. TM: Lurmann. F: Ingles. SA: Wilson. ML. (2013). Associations
between ambient air pollution and Hypertensive Disorders of Pregnancy. Environ Res 123: 9-16.
http://dx.doi.0rg/10.1016/j.envres.2013.01.006
Mohsenin. V. (1987). Effect of vitamin C on NO2-induced airway hyperresponsiveness in normal
subjects: Arandomized double-blind experiment. Am Rev Respir Dis 136: 1408-1411.
http://dx.doi.0rg/10.1164/ajrccm/136.6.1408
Mohsenin. V. (1991). Lipid peroxidation and antielastase activity in the lung under oxidant stress: Role of
antioxidant defenses. J Appl Physiol 70: 1456-1462.
January 2015 7-77 DRAFT: Do Not Cite or Quote
-------
Molitor. J: Su. JG: Molitor. NT: Rubio. VG: Richardson. S: Hastie. D: Morello-Frosch. R: Jerrett. M.
(2011). Identifying vulnerable populations through an examination of the association between
multipollutant profiles and poverty. Environ Sci Technol 45: 7754-7760.
http://dx.doi.org/10.1021/esl04017x
Molter. A: Lindlev. S: de Vocht F: Agius. R: Kerry. G: Johnson. K: Ashmore. M: Terry. A:
Dimitroulopoulou. S: Simpson. A. (2012). Performance of a microenviromental model for estimating
personal NO2 exposure in children. Atmos Environ 51: 225-233.
http://dx.doi.0rg/10.1016/j.atmosenv.2012.01.030
Morales. E: Julvez. J: Torrent. M: de Cid. R: Guxens. M: Bustamante. M: Kiinzli. N: Sunyer. J. (2009).
Association of early-life exposure to household gas appliances and indoor nitrogen dioxide with
cognition and attention behavior in preschoolers. Am J Epidemiol 169: 1327-1336.
http://dx.doi.org/10.1093/aje/kwp067
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
Morrow. PE: Utell MJ: Bauer. MA: Smeglin. AM: Frampton. MW: Cox. C: Speers. DM: Gibb. FR.
(1992). Pulmonary performance of elderly normal subjects and subjects with chronic obstructive
pulmonary disease exposed to 0.3 ppm nitrogen dioxide. Am J Respir Crit Care Med 145: 291-300.
http://dx.doi.0rg/10.1164/airccm/145.2 Pt 1.291
Naess. O: Nafstad. P: Aamodt G: Claussen. B: Rosland. P. (2007). Relation between concentration of air
pollution and cause-specific mortality: four-year exposures to nitrogen dioxide and particulate matter
pollutants in 470 neighborhoods in Oslo, Norway. Am J Epidemiol 165: 435-443.
http://dx.doi.org/10.1093/aje/kwk016
Namdeo. A: Stringer. C. (2008). Investigating the relationship between air pollution, health and social
deprivation in Leeds, UK. Environ Int 34: 585-591. http://dx.doi.Org/10.1016/j.envint.2007.12.015
Nishimura. KK: Galanter. JM: Roth. LA: Oh. SS: Thakur. N: Nguyen. EA: Thyne. S: Farber. HJ:
Serebrisky. D: Kumar. R: Brigino-Buenaventura. E: Davis. A: Lenoir. MA: Meade. K: Rodriguez-
Cintron. W: Avila. PC: Borrell. LN: Bibbins-Domingo. K: Rodriguez-Santana. J. R.: Sen. S: Lurmann.
F: Balmes. J. R.: Burchard. EG. (2013). Early-life air pollution and asthma risk in minority children:
The GALA II and SAGE II studies. Am J Respir Crit Care Med 188: 309-318.
http://dx.doi.org/10.1164/rccm.201302-0264OC
O'Neill MS: Jerrett. M: Kawachi. I: Lew. JL Cohen. AJ: Gouveia. N: Wilkinson. P: Fletcher. T:
Cifuentes. L: Schwartz. J. (2003). Health, wealth, and air pollution: Advancing theory and methods
[Review]. Environ Health Perspect 111: 1861-1870.
Oftedal B: Brunekreef. B: Nystad. W: Madsen. C: Walker. SE: Nafstad. P. (2008). Residential outdoor air
pollution and lung function in schoolchildren. Epidemiology 19: 129-137.
http://dx.doi.org/10.1097/EDE.Ob013e31815c0827
Panasevich. S: Leander. K: Ljungman. P: Bellander. T: de Faire. U: Pershagen. G: Nyberg. F. (2013).
Interaction between air pollution exposure and genes in relation to levels of inflammatory markers and
risk of myocardial infarction. BMJ Open 3: e003058. http://dx.doi.org/10.1136/bmjopen-2013-003058
Panasevich. S: Leander. K: Rosenlund. M: Ljungman. P: Bellander. T: de Faire. U: Pershagen. G: Nyberg.
F (2009). Associations of long- and short-term air pollution exposure with markers of inflammation
and coagulation in a population sample. Occup Environ Med 66: 747-753.
http://dx.doi.org/10.1136/oem.2008.043471
Patel. MM: Chillrud. SN: Correa. JC: Hazi. Y: Feinberg. M: Deepti. KG: Prakash. S: Ross. JM: Lew. D:
Kinnev. PL. (2010). Traffic-related particulate matter and acute respiratory symptoms among New
York City area adolescents. Environ Health Perspect 118: 1338-1343.
http://dx.doi.org/10.1289/ehp.0901499
January 2015 7-78 DRAFT: Do Not Cite or Quote
-------
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/aje/kwk051
Peel JL: Tolbert. PE: Klein. M: Metzger. KB: Flanders. WD: Todd. K: Mulholland. JA: Ryan. PB:
Frumkin. H. (2005). Ambient air pollution and respiratory emergency department visits. Epidemiology
16: 164-174. http://dx.doi.org/10.1097/01.ede.0000152905.42113.db
Pereira. G: Cook. AG: Haggar. F: Bower. C: Nassar. N. (2012). Locally derived traffic-related air
pollution and fetal growth restriction: A retrospective cohort study. Occup Environ Med 69: 815-822.
http://dx.doi.org/10.1136/oemed-2011-100509
Peters. JM: Avol. E: Gauderman. WJ: Linn. WS: Navidi. W: London. SJ: Margolis. H: Rappaport E:
Vora. H: Gong. H. Jr: Thomas. DC. (1999). 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. http://dx.doi.Org/10.1164/ajrccm.159.3.9804144
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.
http://dx.doi.org/10.1080/00039890109604053
Raaschou-Nielsen. O: Andersen. ZJ: Hvidberg. M: Jensen. SS: Ketzel. M: S0rensen. M: Loft S: Overvad.
K: Tj0nneland. A. (2011). Lung cancer incidence and long-term exposure to air pollution from traffic.
Environ Health Perspect 119: 860-865. http://dx.doi.org/10.1289/ehp.1002353
Raaschou-Nielsen. O: Andersen. ZJ: Jensen. SS: Ketzel. M: S0rensen. M: Hansen. J: Loft S: Tj0nneland.
A: Overvad. K. (2012). Traffic air pollution and mortality from cardiovascular disease and all causes:
a Danish cohort study. Environ Health 11: 60. http://dx.doi.org/10.1186/1476-069X-ll-60
Raaschou-Nielsen. O: Bak. H: S0rensen. M: Jensen. SS: Ketzel. M: Hvidberg. M: Schnohr. P:
Tj0nneland. A: Overvad. K: Loft. S. (2010). Air pollution from traffic and risk for lung cancer in three
Danish cohorts. Cancer Epidemiol Biomarkers Prev 19: 1284-1291. http://dx.doi.org/10.1158/1055-
9965.EPI-10-0036
Ranzi. A: Gambini M: Spattini. A: Galassi C: Sesti. D: Bedeschi. M: Messori. A: Baroni. A: Cavagni. G:
Lauriola. P. (2004). Air pollution and respiratory status in asthmatic children: Hints for a locally based
preventive strategy AIRE study. Eur J Epidemiol 19: 567-576.
Rich. DQ: Demissie. K: Lu. SE: Kamat L: Wartenberg. D: Rhoads. GG. (2009). Ambient air pollutant
concentrations during pregnancy and the risk of fetal growth restriction. J Epidemiol Community
Health 63: 488-496. http://dx.doi.org/10.1136/jech.2008.082792
Rivera. M: Basagana. X: Aguilera. I: Foraster. M: Agis. D: de Groot E: Perez. L: Mendez. MA: Bouso.
L: Targa. J: Ramos. R: Sala. J: Marrugat J: Elosua. R: Kiinzli. N. (2013). Association between long-
term exposure to traffic-related air pollution and subclinical atherosclerosis: The REGICOR study.
Environ Health Perspect 121: 223-230. http://dx.doi.org/10.1289/ehp. 1205146
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
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
Rosenlund. M: Bellander. T: Nordquist T: Alfredsson. L. (2009a). Traffic-generated air pollution and
myocardial infarction. Epidemiology 20: 265-271. http://dx.doi.org/10.1097/EDE.Ob013e318190ea68
Rosenlund. M: Forastiere. F: Porta. D: De Sarkx M: Badalonj C: Perucci CA. (2009b). Traffic-related air
pollution in relation to respiratory symptoms, allergic sensitisation and lung function in
schoolchildren. Thorax 64: 573-580. http://dx.doi.org/10.1136/thx.2007.094953
January 2015 7-79 DRAFT: Do Not Cite or Quote
-------
Rotko. T: Kousa. A: Aim. S: Jantunen. M. (2001). Exposures to nitrogen dioxide in EXPOLIS-Helsinki:
Microenvironment, behavioral and sociodemographic factors. J Expo Sci Environ Epidemiol 11: 216-
223. http://dx.doi.org/10.1038/sj.jea.7500162
Rowangould. GM. (2013). A census of the US near-roadway population: Public health and environmental
justice considerations. Transport Res Transport Environ 25: 59-67.
http://dx.doi.0rg/10.1016/j.trd.2013.08.003
Samoli E: Nastos. PT: Paliatsos. AG: Katsouvannj K: Priftis. KN. (2011). Acute effects of air pollution
on pediatric asthma exacerbation: Evidence of association and effect modification. Environ Res 111:
418-424. http://dx.doi.0rg/10.1016/i.envres.2011.01.014
Sarnat SE: Ravsonj AU: Li. WW: Holguin. F: Johnson. BA: Flores Luevano. S: Garcia. JH: Sarnat JA.
(2012). Air pollution and acute respiratory response in a panel of asthmatic children along the U.S.-
Mexico border. Environ Health Perspect 120: 437444. http://dx.doi.org/10.1289/ehp.1003169
Scaife. A: Barclay. J: Hillis. GS: Srinivasan. J: Macdonald. DW: Ross. JA: Ayres. JG. (2012). Lack of
effect of nitrogen dioxide exposure on heart rate variability in patients with stable coronary heart
disease and impaired left ventricular systolic function. Occup Environ Med 69: 587-591.
http://dx.doi.org/10.1136/oemed-2011-100126
Schembarj A: Triguero-Mas. M: De Nazelle. A: Dadvand. P: Vrijheid. M: Cirach. M: Martinez. D:
Figueras. F: Querol. X: Basagana. X: Eeftens. M: Meliefste. K: Nieuwenhuijsen. MJ. (2013).
Personal, indoor and outdoor air pollution levels among pregnant women. Atmos Environ 64: 287-
295. http://dx.doi.0rg/10.1016/j.atmosenv.2012.09.053
Schiller. JS: Lucas. JW: Ward. BW: Peregov. JA. (2012). Summary health statistics for U.S. adults:
National Health Interview Survey, 2010. (DHHS Publication No. 20121580). Hyattsville, MD:
National Center for Health Statistics, http://www.cdc.gov/nchs/data/series/sr 10/srlO 252.pdf
Selgrade. MK: Mole. ML: Miller. FJ: Hatch. GE: Gardner. DE: Hu. PC. (1981). Effect of NO2 inhalation
and vitamin C deficiency on protein and lipid accumulation in the lung. Environ Res 26: 422-437.
Sevanian. A: Hacker. AD: Elsaved. N. (1982). Influence of vitamin E and nitrogen dioxide on lipid
peroxidation in rat lung and liver microsomes. Lipids 17: 269-277.
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
Son. B: Yang. W: Brevsse. P: Chung. T: Lee. Y. (2004). Estimation of occupational and nonoccupational
nitrogen dioxide exposure for Korean taxi drivers using a microenvironmental model. Environ Res 94:
291-296. http://dx.doi.0rg/10.1016/j.envres.2003.08.004
Son. JY: Lee. JT: Park. YH: Bell ML. (2013). Short-term effects of air pollution on hospital admissions
in Korea. Epidemiology 24: 545-554. http://dx.doi.org/10.1097/EDE.Ob013e3182953244
S0rensen. M: Hoffmann. B: Hvidberg. M: Ketzel. M: Jensen. SS: Andersen. ZJ: Tj0nneland. A: Overvad.
K: Raaschou-Nielsen. O. (2012). Long-term exposure to traffic-related air pollution associated with
blood pressure and self-reported hypertension in a Danish cohort. Environ Health Perspect 120: 418-
424. http://dx.doi.org/10.1289/ehp.1103631
Steerenberg. PA: Nierkens. S: Fischer. PH: Van Loveren. H: Opperhuizen. A: Vos. JG: van Amsterdam.
JG. (2001). Traffic-related air pollution affects peak expiratory flow, exhaled nitric oxide, and
inflammatory nasal markers. Arch Environ Health 56: 167-174.
http://dx.doi.org/10.1080/00039890109604069
Stroh. E: Oudin. A: Gustafsson. S: Pilesjo. P: Harrie. L: Stromberg. U: Jakobsson. K. (2005). Are
associations between socio-economic characteristics and exposure to air pollution a question of study
area size? An example from Scania, Sweden. Int J Health Geogr 4: 30-42.
January 2015 7-80 DRAFT: Do Not Cite or Quote
-------
Su. JG: Jerrett M: Morello-Frosch. R: Jesdale. BM: Kyle. AD. (2012). Inequalities in cumulative
environmental burdens among three urbanized counties in California. Environ Int 40: 79-87.
http://dx.doi.0rg/10.1016/i.envint.2011.ll.003
Su. JG: Morello-Frosch. R: Jesdale. BM: Kyle. AD: Shamasunder. B: Jerrett M. (2009). An index for
assessing demographic inequalities in cumulative environmental hazards with application to Los
Angeles, California. Environ Sci Technol 43: 7626-7634. http://dx.doi.org/10.1021/es901041p
Suh. HH: Zanobetti A. (2010). Exposure error masks the relationship between traffic-related air pollution
and heart rate variability. J Occup Environ Med 52: 685-692.
http://dx.doi.org/10.1097/JOM.Ob013e3181e8071f
Sunyer. J: Jarvis. D: Gotschi. T: Garcia-Esteban. R: Jacquemin. B: Aguilera. I: Ackerman. U: de Marco.
R: Forsberg. B: Gislason. T: Heinrich. J: Norback. D: Villani. S: Kiinzli. N. (2006). Chronic bronchitis
and urban air pollution in an international study. Occup Environ Med 63: 836-843.
http://dx.doi.org/10.1136/oem.2006.027995
Takano. H: Yanagisawa. R: Inoue. KI: Shimada. A: Ichinose. T: Sadakane. K: Yoshino. S: Yamaki. K:
Morita. M: Yoshikawa. T. (2004). Nitrogen dioxide air pollution near ambient levels is an atherogenic
risk primarily in obese subjects: A brief communication. Exp Biol Med 229: 361-364.
Thiering. E: Cyrys. J: Kratzsch. J: Meisinger. C: Hoffmann. B: Berdel D: von Berg. A: Koletzko. S:
Bauer. CP: Heinrich. J. (2013). Long-term exposure to traffic-related air pollution and insulin
resistance in children: results from the GINIplus and LISAplus birth cohorts. Diabetologia 56: 1696-
1704. http://dx.doi.org/10.1007/s00125-013-2925-x
Tian. N: Xue. J: Barzyk. TM. (2013). Evaluating socioeconomic and racial differences in traffic-related
metrics in the United States using a GIS approach. J Expo Sci Environ Epidemiol 23: 215-222.
http://dx.doi.org/10.1038/ies.2012.83
Tsai. SS: Chen. PS: Yang. YH: Liou. SH: Wu. TN: Sung. FC: Yang. CY. (2012). Air pollution and hospital
admissions for myocardial infarction: are there potentially sensitive groups? J Toxicol Environ Health
A 75: 242-251. http://dx.doi.org/doi: 10.1080/15287394.2012.641202
U.S. Census Bureau. (2007). 2007 American community survey [Database]. Washington, DC: U.S.
Department of Commerce, U.S. Census Bureau. Retrieved from
http://www.census.gov/acs/www/index.html
U.S. Census Bureau. (2009). American housing survey for the United States: 2009. Washington, DC: U.S.
Government Printing Office, http://www.census.gov/prod/2011pubs/hl50-09.pdf
U.S. Census Bureau. (2013). 2013 TIGER/Line shapefiles. Suitland, MD: U.S. Census Bureau,
Geography Division. https://www.census.gov/cgi-bin/geo/shapefiles2013/main
U.S. Census Bureau. (2014). TIGER/Line with selected demographic and economic data. Suitland, MD:
U.S. Census Bureau, Geography Division. https://www.census.gov/geo/maps-data/data/tiger-data.html
U.S. EPA (U.S. Environmental Protection Agency). (1993). Air quality criteria for oxides of nitrogen, vol.
1-3 [EPAReport]. (EPA/600/8-91/049aF-cF). Research Triangle Park, NC: U.S. Environmental
Protection Agency, Environmental Criteria and Assessment Office.
http://cfpub.epa. gov/ncea/cfm/recordisplav.cfm?deid=40179
U.S. EPA (U.S. Environmental Protection Agency). (2008). Integrated science assessment for oxides of
nitrogen Health criteria [EPA Report]. (EPA/600/R-08/071). Research Triangle Park, NC: U.S.
Environmental Protection Agency, National Center for Environmental Assessment.
http://cfpub.epa.gov/ncea/cfm/recordisplav.cfm7de id=194645
U.S. EPA (U.S. Environmental Protection Agency). (2013a). Integrated science assessment for lead [EPA
Report]. (EPA/600/R-10/075F). Research Triangle Park, NC: U.S. Environmental Protection Agency,
National Center for Environmental Assessment.
http://cfpub.epa.gov/ncea/cfm/recordisplav.cfm?deid=255721
January 2015 7-81 DRAFT: Do Not Cite or Quote
-------
U.S. EPA (U.S. Environmental Protection Agency). (2013b). Integrated science assessment for ozone and
related photochemical oxidants (final report) [EPAReport]. (EPA/600/R-10/076F). Washington, DC.
http://cfpub.epa.gov/ncea/isa/recordisplav.cfm?deid=247492
U.S. EPA (U.S. Environmental Protection Agency). (2014). Integrated review plan for the primary
national ambient air quality standards for nitrogen dioxide [EPA Report]. (EPA-452/R-14/003).
Research Triangle Park, NC: U.S. Environmental Protection Agency, National Center for
Environmental Assessment.
http://www.epa.gov/ttn/naaqs/standards/nox/data/201406finalirpprimarvno2.pdf
Vagaggini B: Paggiaro. PL: Giannini. D: Franco. AD: Cianchetti S: Carnevali S: Taccola. M: Bacci E:
Bancalari. L: Dente. FL: Giuntini C. (1996). Effect of short-term NO2 exposure on induced sputum in
normal, asthmatic and COPD subjects. Eur Respir J 9: 1852-1857.
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 6: 40. http://dx.doi.org/10.1186/1476-069X-6-40
Villeneuve. PJ: Johnson. JY: Pasichnyk. D: Lowes. J: Kirkland. S: Rowe. BH. (2012). Short-term effects
of ambient air pollution on stroke: who is most vulnerable? Sci Total Environ 430: 193-201.
http://dx.doi.0rg/10.1016/j.scitotenv.2012.05.002
Vincent. GK: Velkoff. VA. (2010). The next four decades: The older population in the United States: 2010
to 2050. (P25-1138). Washington, DC: U.S. Department of Commerce, Economics and Statistics
Administration, U.S. Census Bureau, http://www.census.gov/prod/2010pubs/p25-1138.pdf
Vinikoor-Imler. LC: Owens. EO: Nichols. JL: Ross. M: Brown. JS: Sacks. JD. (2014). Evaluating
potential response-modifying factors for associations between ozone and health outcomes: a weight-
of-evidence approach [Review]. Environ Health Perspect 122: 1166-1176.
http://dx.doi.org/10.1289/ehp.1307541
Volk. HE: Kerin. T: Lurmann. F: Hertz-Picciotto. I: Mcconnell R: Campbell DB. (2014). Autism
spectrum disorder: interaction of air pollution with the MET receptor tyrosine kinase gene.
Epidemiology 25: 44-47. http://dx.doi.org/10.1097/EDE.0000000000000030
Wichmann. J: Folke. F: Torp-Pedersen. C: Lippert F: Ketzel M: Ellermann. T: Loft. S. (2013). Out-of-
hospital cardiac arrests and outdoor air pollution exposure in Copenhagen, Denmark. PLoS ONE 8:
e53684. http://dx.doi.org/10.1371/journal.pone.0053684
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. http://dx.doi.org/10.1289/ehp.11605
Wu. X: Bennett. DH: Lee. K: Cassadv. PL: Ritz. B: Hertz-Picciotto. I. (2011). Longitudinal variability of
time-location/activity patterns of population at different ages: a longitudinal study in California.
Environ Health 10: 80. http://dx.doi.org/10.1186/1476-069X-10-80
Yorifuji. T: Kashima. S: Tsuda. T: Takao. S: Suzuki. E: Doi H: Sugiyama. M: Ishikawa-Takata. K: Ohta.
X, (2010). Long-term exposure to traffic-related air pollution and mortality in Shizuoka, Japan. Occup
Environ Med 67: 111-117. http://dx.doi.org/10.1136/oem.2008.045542
Zemek. R: Szyszkowicz. M: Rowe. BH. (2010). Air pollution and emergency department visits for otitis
media: A case-crossover study in Edmonton, Canada. Environ Health Perspect 118: 1631-1636.
http://dx.doi.org/10.1289/ehp.0901675
Zhang. PF: Dong. GH: Sun. BJ: Zhang. LW: Chen. X: Ma. NN: Yu. F: Quo. HM: Huang. H: Lee. YL:
Tang. NJ: Chen. J. (2011). Long-term exposure to ambient air pollution and mortality due to
cardiovascular disease and cerebrovascular disease in Shenyang, China. PLoS ONE 6: e20827.
http://dx.doi.org/10.1371/iournal.pone.0020827
January 2015 7-82 DRAFT: Do Not Cite or Quote
------- |