vvEPA
United States March 2008
|ng™mental Protection EPA/600/R-07/093bB
Annexes for the
Integrated Science Assessment
for Oxides of Nitrogen -
Health Criteria
(Second External Review Draft)
-------
EPA/600/R-07/093bB
March 2008
Annexes for the
Integrated Science Assessment
for Oxides of Nitrogen - Health Criteria
National Center for Environmental Assessment-RTF Division
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, NC
-------
DISCLAIMER
This document is a first external review draft being released for review purposes only and
does not constitute U.S. Environmental Protection Agency (EPA) policy. Mention of trade
names or commercial products does not constitute endorsement or recommendation for use.
March 2008 ii DRAFT-DO NOT QUOTE OR CITE
-------
Annexes for the Integrated Science Assessment
for Oxides of Nitrogen - Health Criteria
ANNEX CHAPTERS
AXl. CHAPTER 1 ANNEX - FRAMEWORK FOR REVIEW AX1-1
AX2. CHAPTER 2 ANNEX - ATMOSPHERIC CHEMISTRY OF NITROGEN
AND SULFUR OXIDES AX2-1
AX3. CHAPTER 3 ANNEX - AMBIENT CONCENTRATIONS AND
EXPOSURES AX3-1
AX4. CHAPTER 4 ANNEX - TOXICOLOGICAL EFFECTS OF NITROGEN
DIOXIDE AND RELATED OXIDES OF NITROGEN AX4-1
AX5. CHAPTER 5 ANNEX - CONTROLLED HUMAN EXPOSURE STUDIES
OF NITROGEN OXIDES AX5-1
AX6. CHAPTER 6 ANNEX - EPIDEMIOLOGICAL STUDIES OF HUMAN
HEALTH EFFECTS ASSOCIATED WITH AMBIENT OXIDES OF
NITROGEN EXPOSURE AX6-1
March 2008 iii DRAFT-DO NOT QUOTE OR CITE
-------
Annex Table of Contents
AX1. CHAPTER 1 ANNEX - FRAMEWORK FOR REVIEW AX1-1
AX1.1 LEGISLATIVE REQUIREMENTS AX1-2
AX1.2 HISTORY OF REVIEWS OF THE PRIMARY NAAQS FORNO2.... AX1-4
AX1.3 LITERATURE SELECTION AX1-5
AX1.4 EVALUATION GUIDELINES AX1-9
AX1.4.1 Background on Causality Decision Framework AX1-9
AX1.4.2 Approaches to the Determination of Causality AX1-11
AX1.5 REFERENCES AX1-47
AX2. CHAPTER 2 ANNEX - ATMOSPHERIC CHEMISTRY OF NITROGEN
AND SULFUR OXIDES AX2-1
AX2.1 INTRODUCTION AX2-1
AX2.2 CHEMISTRY OF NITROGEN OXIDES IN THE TROPOSPHERE.. AX2-2
AX2.2.1 Basic Chemistry AX2-2
AX2.2.2 Nonlinear Relations between Nitrogen Oxide
Concentrations and Ozone Formation AX2-9
AX2.2.3 Multiphase Chemistry Involving NOX AX2-13
AX2.3 CHEMISTRY OF SULFUR OXIDES IN THE TROPOSPHERE AX2-24
AX2.4 MECHANISMS FOR THE AQUEOUS PHASE FORMATION
OF SULFATE AND NITRATE AX2-28
AX2.5 TRANSPORT OF NITROGEN AND SULFUR OXIDES IN
THE ATMOSPHERE AX2-31
AX2.6 SOURCES AND EMISSIONS OF NITROGEN OXIDES,
AMMONIA, AND SULFUR DIOXIDE AX2-35
AX2.6.1 Interactions of Nitrogen Oxides with the Biosphere AX2-35
AX2.6.2 Emissions of NOX, NH3, and SO2 AX2-49
AX2.6.3 Field Studies Evaluating Emissions Inventories AX2-56
AX2.7 METHODS USED TO CALCULATE CONCENTRATIONS OF
NITROGEN OXIDES AND THEIR CHEMICAL INTERACTIONS
IN THE ATMOSPHERE AX2-58
AX2.7.1 Chemistry-Transport Models AX2-59
AX2.7.2 CTM Evaluation AX2-74
AX2.8 SAMPLING AND ANALYSIS OF NITROGEN AND
SULFUR OXIDES AX2-87
AX2.8.1 Availability and Accuracy of Ambient Measurements
forNOY AX2-87
AX2.8.2 Measurements of HNO3 AX2-94
AX2.8.3 Techniques for Measuring Other NOy Species AX2-96
March 2008 iv DRAFT-DO NOT QUOTE OR CITE
-------
Annex Table of Contents
(cont'd)
Page
AX2.8.4 Remote Sensing of Tropospheric NO2 Columns for
Surface NOx Emissions and Surface NO2
Concentrations AX2-96
AX2.8.5 SAMPLING AND ANALYSIS FOR SO2 AX2-98
AX2.8.6 Sampling and Analysis for Sulfate, Nitrate, and
Ammonium AX2-102
AX2.9 POLICY RELEVANT BACKGROUND CONCENTRATIONS
OF NITROGEN AND SULFUR OXIDES AX2-110
AX2.10 REFERENCES AX2-127
AX3. CHAPTER 3 ANNEX - AMBIENT CONCENTRATIONS
AND EXPOSURES AX3-1
AX3.1 INTRODUCTION AX3-1
AX3.2 AMBIENT CONCENTRATIONS OF NITROGEN OXIDES
AND RELATED SPECIES AX3-2
AX3.2.1 Spatial and Temporal Variability in Ambient
Concentrations of NO2 and Related Species in
Urban Areas AX3-4
AX3.2.2 Temporal Variability in Nitrogen Oxides AX3-7
AX3.2.4 Relationships between NO2 and Other Pollutants AX3-20
AX3.2.5 Abundance of NOY Species AX3-23
AX3.3 METHODS FOR MEASURING PERSONAL AND INDOOR
NO2 CONCENTRATIONS AX3-30
AX3.3.1 Issues in Measuring Personal/Indoor NO2 AX3-30
AX3.4 NITROGEN OXIDES IN INDOOR AIR AX3-40
AX3.4.1 Indoor Sources and Concentrations of Nitrogen
Oxides AX3-40
AX3.4.2 Reactions of NO2 in Indoor Air AX3-47
AX3.4.3 Contributions from Outdoor NO2 AX3-53
AX3.5 PERSONAL EXPOSURE AX3-55
AX3.5.1 Personal Exposures and Ambient (Outdoor)
Concentrations AX3-57
AX3.5.2 Personal Exposure in Microenvironments AX3-67
AX3.5.3 Exposure Indicators AX3-83
AX3.6 CONFOUNDING AND SURROGATE IS SUES AX3-85
AX3.7 A FRAMEWORK FOR MODELING HUMAN EXPOSURES
TO NO2 AND RELATED PHOTOCHEMICAL AIR
POLLUTANTS AX3-94
AX3.7.1 Introduction: Concepts, Terminology, and Overall
Summary AX3-94
March 2008 v DRAFT-DO NOT QUOTE OR CITE
-------
Annex Table of Contents
(cont'd)
Page
AX3.7.2 Population Exposure Models: Their Evolution and
Current Status AX3-100
AX3.7.3 Characterization of Ambient Concentrations of NO2
and Related Air Pollutants AX3-103
AX3.7.4 Characterization of Microenvironmental
Concentrations AX3-106
AX3.7.5 Concluding Comments AX3-114
AX3.8 EXPOSURE ERROR AX3-115
AX3.9 REFERENCES AX3-170
AX4. CHAPTER 4 ANNEX - TOXICOLOGICAL EFFECTS OF NITROGEN
DIOXIDE AND RELATED OXIDES OF NITROGEN AX4-1
AX4.1 PULMONARY EFFECTS OF NITROGEN DIOXIDE AND
RELATED OXIDES OF NITROGEN AX4-1
AX4.1.1 Effects of Nitrogen Dioxide on Antioxidant and
Antioxidant Metabolism AX4-1
AX4.1.2 Lipid Metabolism and Content of the Lung AX4-3
AX4.1.3 Emphysema Following Nitrogen Dioxide Exposure AX4-5
AX4.1.4 Nitrates (NO3 ) AX4-6
AX4.2 DOSIMETRY OF INHALED NITROGEN OXIDES AX4-7
AX4.2.1 Mechanisms of NO2 Absorption AX4-8
AX4.2.2 Regional and Total Respiratory Absorption of NO2 AX4-11
AX4.3 EXPERIMENTAL STUDIES OF NO2 UPTAKE AX4-13
AX4.3.1 Upper Respiratory Tract Absorption AX4-14
AX4.3.2 Lower Respiratory Tract Absorption AX4-14
AX4.3.3 Total Respiratory Tract Absorption AX4-15
AX4.4 METABOLISM, DISTRIBUTION AND ELIMINATION OF
NO2 PRODUCTS AX4-15
AX4.5 EXTRA-PULMONARY EFFECTS OF NO2 AND NO AX4-17
AX4.5.1 Body Weight, Hepatic, and Renal Effects AX4-17
AX4.5.2 Brain Effects AX4-18
AX4.5.3 NO AX4-18
AX4.6 EFFECTS OF MIXTURES CONTAINING NO2 AX4-19
AX4.6.1 Simple Mixtures Containing NO2 AX4-19
AX4.6.2 Complex Mixtures Containing NO2 AX4-21
AX4.7 REFERENCES AX4-68
AX5. CHAPTER 5 ANNEX - CONTROLLED HUMAN EXPOSURE STUDIES
OF NITROGEN OXIDES AX5-1
AX5.1 INTRODUCTION AX5-1
March 2008 vi DRAFT-DO NOT QUOTE OR CITE
-------
Annex Table of Contents
(cont'd)
Page
AX5.1.1 Considerations in Controlled Human Exposure Studies AX5-2
AX5.2 EFFECTS OF NITROGEN DIOXIDE IN HEALTHY SUBJECTS .... AX5-4
AX5.3 THE EFFECTS OF NITROGEN OXIDE EXPOSURE IN
SENSITIVE SUBJECTS AX5-4
AX5.4 EFFECTS OF MIXTURES CONTAINING NITROGEN OXIDES .... AX5-5
AX5.5 REFERENCES AX5-16
AX6. CHAPTER 6 ANNEX - EPIDEMIOLOGICAL STUDIES OF HUMAN
HEALTH EFFECTS ASSOCIATED WITH AMBIENT OXIDES OF
NITROGEN EXPOSURE AX6-1
AX6.1 CONSIDERATIONS IN THE INTERPRETATION OF
EPIDEMIOLOGIC STUDIES OF OXIDES OF NITROGEN
HEALTH EFFECTS AX6-1
AX6.1.1 Exposure Assessment and Measurement Error in
Epidemiologic Studies and Related Surrogate
Discussion AX6-2
AX6.1.2 NO2 Exposure Indices Used AX6-7
AX6.1.3 Lag Time: Period between NO2 Exposure and
Observed Health Effect AX6-8
AX6.1.4 Model Specification to Adjust for Temporal Trends
and Meteorological Effects AX6-9
AX6.1.5 Confounding Effects of Copollutants AX6-10
AX6.1.6 Generalized Estimating Equations (GEE) AX6-11
AX6.1.7 Hypothesis Testing and Model Selection in NO2
Epidemiologic Studies AX6-11
AX6.1.8 Impact of Generalized Additive Models Convergence
Issue onNO2 Risk Estimates AX6-12
AX6.2 CARDIOVASCULAR EFFECTS ASSOCIATED WITH
SHORT-TERM NO2 EXPOSURE AX6-13
AX6.2.1 Studies Hospital Admissions and ED Visits for
Cardiovascular Disease (CVD) AX6-13
AX6.2.2 Heart Rate Variability, Repolarization, Arrhythmia,
and Other Measures Cardiovascular Function
Associated with Short-Term NO2 Exposure AX6-22
AX6.4 REFERENCES AX6-194
March 2008 vii DRAFT-DO NOT QUOTE OR CITE
-------
Annex List of Figures
Number
AX1.3-1. Selection process for studies included in ISA AX1-6
AX1.4-1. Focusing on unmeasured confounders/covariates, or other
sources of spurious association from bias AX1-21
AX1.4-2. Example posterior distribution for the determination of Sufficient AX1 -21
AX1.4-3. Example posterior distribution for the determination of Equipoise
and Above AX1-22
AX1.4-4. Example posterior distribution for the determination of Against AX1 -23
AX2.2-1. Schematic diagram of the cycle of reactive nitrogen species
in the atmosphere AX2-3
AX2.2-2. Measured values of O3 and NOZ (NOY - NOX) during the afternoon
at rural sites in the eastern United States (gray circles) and in urban
areas and urban plumes associated with Nashville, TN (gray dashes),
Paris, FR (black diamonds) and Los Angeles, CA (X's) AX2-12
AX2.2-3. Structures of nitro-polycyclic aromatic hydrocarbons AX2-16
AX2.2-4. Formation of 2-nitropyrene (2NP) from the reaction of OH with
gaseous pyrene (PY) AX2-17
AX2.3-1. Transformations of sulfur compounds in the atmosphere AX2-26
AX2.4-1. Comparison of aqueous-phase oxidation paths AX2-30
AX2.6-1. Diel cycles of median concentrations (upper panels) and fluxes
(lower panels) for the Northwest clean sector, left panels) and
Southwest (polluted sector, right panels) wind sectors at Harvard
Forest, April-November, 2000, for NO, NO2, and O3/10. NO and
O3 were sampled at a height of 29 m, andNO2 at 22m AX2-43
AX2.6-2. Simple NOX photochemical canopy model outputs AX2-44
AX2.6-3. Hourly (dots) and median nightly (pluses) NO2 flux vs. concentration,
with results of least-squares fit on the hourly data (curve) AX2-45
AX2.6-4. Averaged profiles at Harvard Forest give some evidence of some
NO2 input near the canopy top from light-mediated ambient reactions,
or emission from open stomates AX2-46
AX2.6-5. Summer (June-August) 2000 median concentrations (upper panels),
fractions of NOy (middle panels), and fluxes (lower panels) of NOy
and component species separated by wind direction (Northwest on the
left and Southwest on the right) AX2-48
March 2008
vin
DRAFT-DO NOT QUOTE OR CITE
-------
Annex List of Figures
(cont'd)
Number
age
AX2.7-1. Scatter plot of total nitrate (HNO3 plus aerosol nitrate) wet deposition
(mg(N)m2yr~1) of the mean model versus measurements for the North
American Deposition Program (NADP) network AX2-71
AX2.7-2. Same as Figure AX2.7-1 but for sulfate wet deposition
(mg^m'V1) AX2-72
AX2.7-3a,b. Impact of model uncertainty on control strategy predictions for O3 for
two days (August lOa and lib, 1992) in Atlanta, GA AX2-77
AX2.7-4. Ozone isopleths (ppb) as a function of the average emission rate for
NOx and VOC (1012 molec. cm"2 s"1) in zero dimensional box model
calculation AX2-78
AX2.7-5a. Time series for measured gas-phase species in comparison with
results from a photochemical model AX2-79
AX2.7-5b. Time series for measured gas-phase species in comparison with
results from a photochemical model AX2-80
AX2.7-6. Correlations for O3 versus NOz (NOy-NOx) in ppb from chemical
transport models for the northeast corridor, Lake Michigan, Nashville,
the San Joaquin Valley, and Los Angeles AX2-82
AX.7-7a,b. Evaluation of model versus measured O3 versus NOy for two model
scenarios for Atlanta AX2-83
AX2.7-8a,b. Evaluation of model versus: (a) measured O3 versus NOz and
(b) O3 versus the sum 2H2O2 + NOZ for Nashville, TN AX2-85
AX2.7-9. Time series of concentrations of RO2, HO2, and OH radicals, local
O3 photochemical production rate and concentrations of NOx from
measurements made during BERLIOZ AX2-86
AX2.8-1. Tropospheric NO2 columns (molecules NO2/ cm2) retrieved from the
SCIAMACHY satellite instrument for 2004-2005 AX2-97
AX2.9-1. Annual mean concentrations of NO2 (ppbv) in surface air over the
United States in the present-day (upper panel) and policy relevant
background (middle panel) MOZART-2 simulations AX2-112
AX2.9-2. Same as Figure AX2.9-1 but for SO2 concentrations AX2-113
AX2.9-3. Same as for Figure AX2.9-1 but for wet and dry deposition of
HNO3, NH4NO3, NOX, HO2NO2, and organic nitrates
(mgNm'V1) AX2-114
AX2.9-4. Same as Figure AX2.9-1 but for SOX deposition (SO2 + SO4)
mSnr1 AX2-115
March 2008
IX
DRAFT-DO NOT QUOTE OR CITE
-------
Annex List of Figures
(cont'd)
Number Page
AX2.9-5. July mean soil NO emissions (upper panels; 1 x 10 9 molecules
cm"2 s1) and surface PRB NOx concentrations (lower panels; pptv)
over the United States from MOZART-2 (left) and GEOS-Chem
(right) model simulations in which anthropogenic Os precursor
emissions were set to zero in North America AX2-116
AX3.2-1. Location of ambient NO2 monitors in the United States AX3-3
AX3.2-2. NC>2 concentrations measured at 4 m (Van) and at 15 m at NY
Department of Environmental Conservation sites (DEC709406
and DEC709407) AX3-7
AX3.2-4a-e. Time series of 24-h average NO2 concentrations at
individual sites in New York City from 2003 through 2006 AX3-10
AX3.2-5a-e. Time series of 24-h average NO2 concentrations at individual sites
in Atlanta, GA from 2003 through 2005 AX3-11
AX3.2-6a-g. Time series of 24-h average NC>2 concentrations at individual sites
in Chicago, IL from 2003 through 2005 AX3-12
AX3.2-7a-b. Time series of 24-h average NC>2 concentrations at individual sites
in Baton Rouge, LA from 2003 through 2005 AX3-13
AX3.2-8a-g. Time series of 24-h average NC>2 concentrations at individual sites
in Houston, TX from 2003 through 2005 AX3-14
AX3.2-9a-h. Time series of 24-h average NC>2 concentrations at individual sites in
Los Angeles, CAfrom 2003 through 2005 AX3-15
AX3.2-9i-n. Time series of 24-h average NC>2 concentrations at individual sites in
Los Angeles, CA from 2003 through 2006 AX3-16
AX3.2-10a-d. Time series of 24-h average NC>2 concentrations at individual sites in
Riverside, CA from 2003 through 2006 AX3-17
AX3.2-10e-i. Time series of 24-h average NC>2 concentrations at individual sites in
Riverside, CA from 2003 through 2006 AX3-18
AX3.2-11. Nationwide trends in annual meanNO2 concentrations AX3-19
AX3.2-12. Trends in annual meanNO2 concentrations by site type AX3-19
AX3.2-13a-d. Correlations of NC>2 to 63 vs. correlations of NC>2 to CO for
Los Angeles, CA (2001-2005) AX3-22
AX3.2-14. Relationship between 63, NO, and NO2 as a function of NOx
concentration AX3-23
March 2008 x DRAFT-DO NOT QUOTE OR CITE
-------
Annex List of Figures
(cont'd)
Number Page
AX3.2-15. Variation of odd oxygen (= O3 + NO2) with NOX AX3-24
AX3.2-16a-d. Measured Os (ppbv) versus PAN (pptv) in Tennessee, including
(a) aircraft measurements, and (b, c, and d) suburban sites
near Nashville AX3-26
AX3.2-17. Relationship between benzene and NOy at a measurement site in
Boulder, CO AX3-27
AX3.2-18. Ratios of PAN to NO2 observed at Silwood Park, Ascot, Berkshire,
U.K. from July 24 to August 12 1999 AX3-28
AX3.2-20. Concentrations of particulate nitrate measures as part of the
Environmental Protection Agency PA's speciation network AX3-30
AX3.4-1. Percentage of time people spend in different environments AX3-42
AX3.5-1. Average residential outdoor concentration versus concentration
during commuting forNO2 AX3-77
AX3.7-1. Schematic description of a general framework identifying the processes
(steps or components) involved in assessing inhalation exposures
and doses for individuals and populations AX3-97
AX3.8-1. Errors associated with components of the continuum from ambient air
pollution to adverse health outcome AX3-116
AX3.8-2. A systematic approach to evaluate whether NO2 itself is causing
the observed adverse health outcome orNO2 is acting as a surrogate
for other pollutants AX3-120
AX6.2-1. Relative risks (95% CI) for associations of 24-h NO2 (per 20 ppb) and
daily 1 hour maximum NO2* with hospitalizations or emergency
department visits for all cardiovascular diseases (CVD) AX6-14
AX6.2-2. Relative risks (95% CI) for associations of 24-h NO2 (per 20 ppb) and
daily 1 hour maximum NO2* (per 30 ppb) with hospitalizations for
Ischemic Heart Disease (MD) AX6-15
AX6.2-3. Relative risks (95% CI) for associations between 24-h NO2 (per 20 ppb)
and hospitalizations for myocardial infarction (MI) AX6-17
AX6.2-4. Relative risks (95% CI) for associations of 24-h NO2 (per 20 ppb) and
1-hour maximum NO2* with hospitalizations for congestive heart
failure (CHF). Primary author and year of publication, city, stratification
variable(s), and lag are listed. Results for lags 0 or 1 are presented
as available AX6-18
March 2008 xi DRAFT-DO NOT QUOTE OR CITE
-------
Annex List of Figures
(cont'd)
Number Page
AX6.2-5. Relative risks (95% CI) for associations of 24-h NO2 exposure (per
20 ppb) and daily 1-hour maximum NO2* (per 30 ppb) with
hospitalizations or emergency department visits for CVD: Studies with
2 pollutant model results AX6-20
March 2008 xii DRAFT-DO NOT QUOTE OR CITE
-------
Annex List of Tables
Number
AX1.3-1. Literature Search Strategy for Epidemiologic Studies:
Examples of Keywords AX1-46
AX1.3-2. Literature Search Strategy for The Atmospheric Sciences AX1-46
AX2.3-1. Atmospheric Lifetimes of Sulfur Dioxide and Reduced Sulfur
Species with Respect to Reaction With OH, NO3, and Cl Radicals AX2-120
AX2.4-la. Relative Contributions of Various Reactions to the Total S(IV)
Oxidation Rate within a Sunlit Cloud, 10 Minutes after
Cloud Formation AX2-120
AX2.4-lb. Relative Contributions of Various Gas and Aqueous Phase Reactions
to Aqueous Nitrate Formation within a Sunlit Cloud, 10 Minutes after
Cloud Formation AX2-121
AX2.6-1. Emissions of Nitrogen Oxides, Ammonia, and Sulfur Dioxide in the
United States in 2002 AX2-122
AX2.8-1. Satellite Instruments Used to Retrieve Tropospheric NO2 Columns AX2-126
AX3.2-1. Summary of Percentiles of NO2 Data Pooled Across Monitoring
Sites (2003-2005) Concentrations are in ppm AX3-121
AX3.2-2. Spatial Variability of NO2 in Selected United states Urban Areas AX3-122
AX3.2-3. NOX and NOY Concentrations at Regional Background Sites in the
Eastern United States. Concentrations are GIVEN in ppb AX3-122
AX3.2-4. Range of Pearson Correlation Coefficients Between NO2 and O3,
COandPM2.5 AX3-123
AX3.3-1. Passive Samplers Used inNO2 Measurements AX3-124
AX3.3-2. The Performance of Sampler/Sampling Method for NO2
Measurements in the Air AX3-125
AX3.4-1. NO2 Concentrations (ppb) in Homes in Latrobe Valley, Victoria,
Australia AX3-126
AX3.4-2. NO2 Concentrations (ppb) in Homes in Connecticut AX3-126
AX3.4-3. NO2 Concentrations Near Indoor Sources - Short-Term Averages AX3-127
AX3.4-4. NO2 Concentrations Near Indoor Sources - Long-Term Averages AX3-128
AX3.5-1. Summary of Regression Models of Personal Exposure to
Ambient/Outdoor NO2 AX3-129
AX3.5-2. Average Ambient and Nonambient Contributions to
Population Exposure AX3-130
March 2008
xin
DRAFT-DO NOT QUOTE OR CITE
-------
Annex List of Tables
(cont'd)
Number Page
AX3.5-3. The Association Between Personal Exposures and Ambient
Concentrations AX3-131
AX3.5-4. Indoor/Outdoor Ratio and the Indoor vs. Outdoor Regression Slope AX3-134
AX3.5-5. NO2 Concentrations (ppb) in Different Rooms AX3-141
AX3.5-6. Indoor and Outdoor Contributions to Indoor Concentrations AX3-143
AX3.5-7. The Association Between Indoor, Outdoor, and Personal NO2 AX3-145
AX3.5-9. Personal NO2 Levels Stratified by Demographic and Socioeconomic
Factors (Concentrations are in ppb and Slopes are Dimensionless) AX3-164
AX3.6-1. Correlations (Pearson Correlation Coefficient) Between Ambient NO2
and Ambient Copollutants AX3-165
AX3.6-2. Correlations (Pearson Correlation Coefficient) Between Personal NO2
and Personal Copollutants AX3-166
AX3.6-3. Correlations (Pearson Correlation Coefficient) Between Personal NO2
and Ambient Copollutants AX3-167
AX3.6-4. Correlations (Pearson Correlation Coefficient) Between Ambient NO2
and Personal Copollutants AX3-168
AX3.7-1. The Essential Attributes of the pNEM, HAPEM, APEX, SHEDS, and
MENTOR-1A AX3-169
AX4.1. Effects of Nitrogen Dioxide on Oxidant and Antioxidant Homeostasis AX4-22
AX4.2. Effects of Nitrogen Dioxide on Lung Amino Acids, Proteins,
Lipids, and Enzymes AX4-25
AX4.3. Effects of Nitrogen Dioxide on Alveolar Macrophages and
Lung Host Defense AX4-31
AX4.4. Effects of Nitrogen Dioxide on Lung Permeability and Inflammation AX4-38
AX4.5. Effects of Nitrogen Dioxide on Immune Responses AX4-43
AX4.6. Effect of Nitrogen Dioxide on Susceptibility to Infectious Agents AX4-49
AX4.7. Effects of Nitrogen Dioxide on Lung Structure AX4-56
AX4.8. Effects of Nitrogen Dioxide on Pulmonary Function AX4-62
AX4.9. Effect of Nitrogen Dioxide on Hematological Parameters AX4-63
AX4.10. Effects of Nitric Oxide on Iron, Enzymes, and Nucleic Acids AX4-65
AX4.11 A. Genotoxicity of Nitrogen Dioxide In Vitro and In Plants AX4-66
March 2008
xiv
DRAFT-DO NOT QUOTE OR CITE
-------
Annex List of Tables
(cont'd)
Number
age
AX4.11B. Genoticity of Nitrogen Dioxide In Vivo AX4-67
AX4.11C. Genotoxicity of Nitric Oxide AX4-67
AX5.2-1. Clinical Studies of NO2 Exposure in Healthy Subjects AX5-6
AX5.3-1. Effects of NO2 Exposure in Subjects with Respiratory Disease
(see Table AX5.3-2 for Studies with Allergen Challenge) AX5-11
AX5.3-2. Effects of NO2 Exposure on Response to Inhaled Allergen AX5-12
AX5.4-1. Effects of Exposure to NO2 with Other Pollutants AX5-14
AX6.3-1. Studies Examining Exposure to Indoor NO2 and Respiratory Symptoms... AX6-29
AX6.3-2. Studies Examining Exposure to Ambient NO2 and Acute Respiratory
Symptoms Using Generalized Estimating Equations (GEE) in the
Analysis Method AX6-32
AX6.3-3. Respiratory Health Effects of Oxides of Nitrogen: Hospital Admissions.. AX6-35
AX6.3-4. Respiratory Health Effects of Oxides of Nitrogen: Emergency
Department Visits AX6-81
AX6.3-5. Respiratory Health Effects of Oxides of Nitrogen: General
Practitioner/Clinic Visits AX6-104
AX6.3-6. Human Health Effects of Oxides of Nitrogen: CVD Hospital
Admissions and Visits: United States and Canada AX6-110
AX6.3-7. Human Health Effects of Oxides of Nitrogen: CVD Hospital
Admissions and Visits: Australia and New Zealand AX6-125
AX6.3-8. Human Health Effects of Oxides of Nitrogen: CVD Hospital
Admissions and Visits: Europe AX6-130
AX6.3-9. Human Health Effects of Oxides of Nitrogen: CVD Hospital
Admissions and Visits: Asia AX6-141
AX6.3-10. Studies Examining Exposure to Ambient NO2 and Heart Rate
Variability as Measured by Standard Deviation of Normal-to-Normal
Intervals (SDNN) AX6-147
AX6.3-11. Studies Examining Exposure to Ambient NO2 and Heart Rate Variability
as Measured by Variables Recorded on Implantable Cardioverter
Defibrillators (ICDs) AX6-148
AX6.3-12. Birth Weight and Long-Term NO2 Exposure Studies AX6-149
AX6.3-13. Preterm Delivery and Long-Term NO2 Exposure Studies AX6-153
AX6.3-14. Fetal Growth and Long-Term NO2 Exposure Studies AX6-155
March 2008
xv
DRAFT-DO NOT QUOTE OR CITE
-------
Number
AX6.3-15.
AX6.3-16.
AX6.3-17.
AX6.3-18.
AX6.3-19.
AX6.3-20.
Annex List of Tables
(cont'd)
age
Lung Function and Long-Term NO2 Exposure AX6-156
Asthma and Long-Term NO2 Exposure AX6-158
Respiratory Symptoms and Long-Term NO2 Exposure AX6-162
Lung Cancer AX6-169
Effects of Acute NOX Exposure on Mortality. Risk Estimates are
Standardized for per 20 ppb 24-h AVG NO2 Increment AX6-170
NO2 Exposure Affects Asthmatics AX6-192
March 2008
xvi
DRAFT-DO NOT QUOTE OR CITE
-------
Authors, Contributors, and Reviewers
Authors
Dr. Dennis J. Kotchmar (NOx Team Leader)—National Center for Environmental Assessment
(B243-01), U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
Dr. Mary Ross (Branch Chief)—National Center for Environmental Assessment (B243-01),
U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
Dr. Kathleen Belanger, Yale University, Epidemiology and Public Health, 60 College Street,
New Haven, CT 06510-3210
Dr. James S. Brown—National Center for Environmental Assessment (B243-01),
U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
Dr. Douglas Bryant—Cantox Environmental Inc., 1900 Minnesota Court, Mississauga, Ontario
L8S IPS
Dr. Ila Cote—National Center for Environmental Assessment (B243-01), U.S. Environmental
Protection Agency, Research Triangle Park, NC 27711
Dr. Mark Frampton—Strong Memorial Hospital, 601 Elmwood Ave., Box 692, Rochester, NY
14642-8692
Dr. Janneane Gent—Yale University, CPPEE, One Church Street, 6th Floor, New Haven, CT
06510
Dr. Vic Hasselblad—Duke University, 29 Autumn Woods Drive, Durham, NC 27713
Dr. Kazuhiko Ito—New York University School of Medicine, 57 Old Forge Road, Tuxedo, NY
10987
Dr. Jee Young Kim—National Center for Environmental Assessment (B243-01),
U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
Dr. Ellen Kirrane—National Center for Environmental Assessment (B243-01),
U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
Dr. Thomas Long—National Center for Environmental Assessment (B243-01),
U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
Dr. Thomas Luben—National Center for Environmental Assessment (B243-01),
U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
March 2008 xvii DRAFT-DO NOT QUOTE OR CITE
-------
Authors, Contributors, and Reviewers
(cont'd)
Authors
(cont'd)
Dr. Andrew Maier—Toxicology Excellence for Risk Assessment, 2300 Montana Avenue,
Suite 409, Cincinnati, OH 45211
Dr. Qingyu Meng—National Center for Environmental Assessment (B243-01),
U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
Dr. Joseph Pinto—National Center for Environmental Assessment (B243-01),
U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
Dr. Paul Reinhart—National Center for Environmental Assessment (B243-01),
U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
Dr. David Svendsgaard—National Center for Environmental Assessment (B243-01),
U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
Dr. Lori White—National Center for Environmental Assessment (B243-01),
U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
Contributors
Dr. Dale Allen, University of Maryland, College Park, MD
Dr. Jeffrey Arnold—National Center for Environmental Assessment (B243-01),
U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
Dr. Barbara Buckley—National Center for Environmental Assessment (B243-01),
U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
Ms. Louise Camalier, U.S. EPA, OAQPS, Research Triangle Park, NC
Ms. Rebecca Daniels, MSPH—National Center for Environmental Assessment (B243-01),
U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
Dr. Russell Dickerson, University of Maryland, College Park, MD
Dr. Tina Fan, EOHSI/UMDNJ, Piscataway, NJ
Dr. Arlene Fiore, NOAA/GFDL, Princeton, NJ
March 2008 xviii DRAFT-DO NOT QUOTE OR CITE
-------
Authors, Contributors, and Reviewers
(cont'd)
Contributors
(cont'd)
Dr. Panos Georgopoulos, EOHSI/UMDNJ, Piscataway, NJ
Dr. Larry Horowitz, NOAA/GFDL, Princeton, NJ
Dr. William Keene, University of Virginia, Charlottesville, VA
Dr. Randall Martin, Dalhousie University, Halifax, Nova Scotia
Dr. Maria Morandi, University of Texas, Houston, TX
Dr. William Munger, Harvard University, Cambridge, MA
Mr. Charles Piety, University of Maryland, College Park, MD
Dr. Jason Sacks—National Center for Environmental Assessment (B243-01),
U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
Dr. Sandy Sillman, University of Michigan, Ann Arbor, MI
Dr. Jeffrey Stehr, University of Maryland, College Park, MD
Dr. Helen Suh, Harvard University, Boston, MA
Ms. Debra Walsh—National Center for Environmental Assessment (B243-01),
U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
Dr. Charles Wechsler, EOHSI/UMDNJ, Piscataway, NJ
Dr. Clifford Weisel, EOHSI/UMDNJ, Piscataway, NJ
Dr. Jim Zhang, EOHSI/UMDNJ, Piscataway, NJ
Reviewers
Dr. Tina Bahadori—American Chemistry Council, 1300 Wilson Boulevard, Arlington, VA
22209
March 2008 xix DRAFT-DO NOT QUOTE OR CITE
-------
Authors, Contributors, and Reviewers
(cont'd)
Reviewers
(cont'd)
Dr. Tim Benner—Office of Science Policy, Office of Research and Development, Washington,
DC 20004
Dr. Daniel Costa—National Program Director for Air, Office of Research and Development,
U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
Dr. Robert Devlin—National Health and Environmental Effects Research Laboratory, Office of
Research and Development, U.S. Environmental Protection Agency, Chapel Hill, NC
Dr. Judy Graham—American Chemistry Council, LRI, 1300 Wilson Boulevard, Arlington, VA
22209
Dr. Stephen Graham—Office of Air and Radiation, U.S. Environmental Protection Agency,
Research Triangle Park, NC 27711
Ms. Beth Hassett-Sipple—U.S. Environmental Protection Agency (C504-06), Research Triangle
Park, NC 27711
Dr. Gary Hatch—National Health and Environmental Effects Research Laboratory, Office of
Research and Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
27711
Dr. Scott Jenkins—Office of Air Quality Planning and Standards, U.S. Environmental Protection
Agency (C504-02), Research Triangle Park, NC 27711
Dr. David Kryak—National Exposure Research Laboratory, Office of Research and
Development, U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
Mr. John Langstaff—U.S. Environmental Protection Agency (C504-06), Research Triangle Park,
NC27711
Dr. Morton Lippmann—NYU School of Medicine, 57 Old Forge Road, Tuxedo, NY 10987
Dr. Thomas Long—National Center for Environmental Assessment (B243-01),
U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
Dr. Karen Martin—Office of Air and Radiation, U.S. Environmental Protection Agency
(C504-06), Research Triangle Park, NC 27711
March 2008 xx DRAFT-DO NOT QUOTE OR CITE
-------
Authors, Contributors, and Reviewers
(cont'd)
Reviewers
(cont'd)
Dr. William McDonnell—William F. McDonnell Consulting, 1207 Hillview Road, Chapel Hill,
NC27514
Dr. Dave McKee—Office of Air and Radiation/Office of Air Quality Planning and Standards,
U.S. Environmental Protection Agency (C504-06), Research Triangle Park, NC 27711
Dr. Lucas Neas—National Health and Environmental Effects Research Laboratory, Office of
Research and Development, U.S. Environmental Protection Agency, Chapel Hill, NC 27711
Dr. Russell Owen—National Health and Environmental Effects Research Laboratory, Office of
Research and Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
27711
Dr. Haluk Ozkaynak—National Exposure Research Laboratory, Office of Research and
Development, U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
Dr. Jennifer Peel—Colorado State University, 1681 Campus Delivery, Fort Collins, CO 80523-
1681
Mr. Harvey Richmond—Office of Air Quality Planning and Standards/Health and
Environmental Impacts Division, U.S. Environmental Protection Agency (C504-06), Research
Triangle Park, NC 27711
Mr. Joseph Somers—Office of Transportation and Air Quality, U.S. Environmental Protection
Agency, 2000 Traverwood Boulevard, Ann Arbor, MI 48105
Ms. Susan Stone—U.S. Environmental Protection Agency (C504-06), Research Triangle Park,
NC27711
Dr. John Vandenberg—National Center for Environmental Assessment (B243-01), Office of
Research and Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
27711
Dr. Alan Vette—National Exposure Research Laboratory, Office of Research and Development,
U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
Mr. Ron Williams—National Exposure Research Laboratory, Office of Research and
Development, U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
March 2008 xxi DRAFT-DO NOT QUOTE OR CITE
-------
Authors, Contributors, and Reviewers
(cont'd)
Reviewers
(cont'd)
Dr. William Wilson—Office of Research and Development, National Center for Environmental
Assessment (B243-01), U.S. Environmental Protection Agency, Research Triangle Park, NC
27711
March 2008 xxii DRAFT-DO NOT QUOTE OR CITE
-------
U.S. Environmental Protection Agency Project Team
for Development of Integrated Scientific Assessment
for Oxides of Nitrogen
Executive Direction
Dr. Ila Cote (Acting Director)—National Center for Environmental Assessment-RTF Division,
(B243-01), U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
Ms. Debra Walsh (Deputy Director)—National Center for Environmental Assessment-RTF
Division, (B243-01), U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
Scientific Staff
Dr. Dennis Kotchmar (NOX Team Leader)—National Center for Environmental Assessment
(B243-01), U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
Dr. Jeff Arnold—National Center for Environmental Assessment (B243-01),
U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
Dr. James S. Brown—National Center for Environmental Assessment (B243-01),
U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
Dr. Barbara Buckley—National Center for Environmental Assessment (B243-01),
U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
Ms. Rebecca Daniels—National Center for Environmental Assessment (B243-01),
U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
Dr. Jee Young Kim—National Center for Environmental Assessment (B243-01),
U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
Dr. Ellen Kirrane—National Center for Environmental Assessment (B243-01),
U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
Dr. Tom Long—National Center for Environmental Assessment (B243-01),
U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
Dr. Thomas Luben—National Center for Environmental Assessment (B243-01),
U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
Dr. Qingyu Meng—National Center for Environmental Assessment (B243-01),
U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
March 2008 xxiii DRAFT-DO NOT QUOTE OR CITE
-------
U.S. Environmental Protection Agency Project Team
for Development of Integrated Scientific Assessment
for Oxides of Nitrogen
(cont'd)
Scientific Staff
(cont'd)
Dr. Joseph Pinto—National Center for Environmental Assessment (B243-01),
U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
Dr. Paul Reinhart—National Center for Environmental Assessment (B243-01),
U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
Dr. Mary Ross—National Center for Environmental Assessment (B243-01),
U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
Dr. Jason Sacks—National Center for Environmental Assessment (B243-01),
U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
Dr. David Svendsgaard—National Center for Environmental Assessment (B243-01),
U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
Dr. Lori White—National Center for Environmental Assessment (B243-01),
U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
Dr. William Wilson—National Center for Environmental Assessment (B243-01),
U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
Technical Support Staff
Ms. Ella King—Executive Secretary, National Center for Environmental Assessment (B243-01),
U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
Ms. Emily R. Lee—Management Analyst, National Center for Environmental Assessment
(B243-01), U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
Ms. Ellen F. Lorang—Information Manager, National Center for Environmental Assessment
(B243-01), U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
Ms. Christine Searles—Management Analyst, National Center for Environmental Assessment
(B243-01), U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
March 2008 xxiv DRAFT-DO NOT QUOTE OR CITE
-------
U.S. Environmental Protection Agency Project Team
for Development of Integrated Scientific Assessment
for Oxides of Nitrogen
(cont'd)
Technical Support Staff
(cont'd)
Mr. Richard Wilson—Clerk, National Center for Environmental Assessment (B243-01),
U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
Document Production Staff
Ms. Barbra H. Schwartz—Task Order Manager, Computer Sciences Corporation,
2803 Slater Road, Suite 220, Morrisville, NC 27560
Mr. John A. Bennett—Technical Information Specialist, Library Associates of Maryland,
11820 Parklawn Drive, Suite 400, Rockville, MD 20852
Mr. David Casson—Publication/Graphics Specialist, TekSystems, 1201 Edwards Mill Road,
Suite 201, Raleigh, NC 27607
Mrs. Melissa Cesar—Publication/Graphics Specialist, Computer Sciences Corporation,
2803 Slater Road, Suite 220, Morrisville, NC 27560
Mr. Eric Ellis—Records Management Technician, InfoPro, Inc., 8200 Greensboro Drive, Suite
1450, McLean, VA 22102
Ms. Kristin Hamilton—Publication/Graphics Specialist, TekSystems, 1201 Edwards Mill Road,
Suite 201, Raleigh, NC 27607
Ms. Stephanie Harper—Publication/Graphics Specialist, TekSystems, 1201 Edwards Mill Road,
Suite 201, Raleigh, NC 27607
Ms. Sandra L. Hughey—Technical Information Specialist, Library Associates of Maryland,
11820 Parklawn Drive, Suite 400, Rockville, MD 20852
Dr. Barbara Liljequist—Technical Editor, Computer Sciences Corporation, 2803 Slater Road,
Suite 220, Morrisville, NC 27560
Ms. Molly Windsor—Graphic Artist, Computer Sciences Corporation, 2803 Slater Road,
Suite 220, Morrisville, NC 27560
March 2008 xxv DRAFT-DO NOT QUOTE OR CITE
-------
U.S. Environmental Protection Agency
Science Advisory Board (SAB)
Staff Office Clean Air Scientific Advisory Committee (CASAC)
CASAC NOX and SOX Primary NAAQS Review Panel
Chair
Dr. Rogene Henderson*, Scientist Emeritus, Lovelace Respiratory Research Institute,
Albuquerque, NM
Members
Mr. Ed Avol, Professor, Preventive Medicine, Keck School of Medicine, University of Southern
California, Los Angeles, CA
Dr. John R. Balmes, Professor, Department of Medicine, Division of Occupational and
Environmental Medicine, University of California, San Francisco, CA
Dr. Ellis Cowling*, University Distinguished Professor At-Large, North Carolina State
University, Colleges of Natural Resources and Agriculture and Life Sciences, North Carolina
State University, Raleigh, NC
Dr. James D. Crapo [M.D.]*, Professor, Department of Medicine, National Jewish Medical and
Research Center, Denver, CO
Dr. Douglas Crawford-Brown*, Director, Carolina Environmental Program; Professor,
Environmental Sciences and Engineering; and Professor, Public Policy, Department of
Environmental Sciences and Engineering, University of North Carolina at Chapel Hill, Chapel
Hill, NC
Dr. Terry Gordon, Professor, Environmental Medicine, NYU School of Medicine, Tuxedo, NY
Dr. Dale Hattis, Research Professor, Center for Technology, Environment, and Development,
George Perkins Marsh Institute, Clark University, Worcester, MA
Dr. Patrick Kinney, Associate Professor, Department of Environmental Health Sciences,
Mailman School of Public Health, Columbia University, New York, NY
Dr. Steven Kleeberger, Professor, Laboratory Chief, Laboratory of Respiratory Biology,
NIH/NIEHS, Research Triangle Park, NC
Dr Timothy Larson, Professor, Department of Civil and Environmental Engineering, University
of Washington, Seattle, WA
March 2008 xxvi DRAFT-DO NOT QUOTE OR CITE
-------
U.S. Environmental Protection Agency
Science Advisory Board (SAB)
Staff Office Clean Air Scientific Advisory Committee (CASAC)
CASAC NOX and SOX Primary NAAQS Review Panel
(cont'd)
Members
(cont'd)
Dr. Kent Pinkerton, Professor, Regents of the University of California, Center for Health and
the Environment, University of California, Davis, CA
Mr. Richard L. Poirot*, Environmental Analyst, Air Pollution Control Division, Department of
Environmental Conservation, Vermont Agency of Natural Resources, Waterbury, VT
Dr. Edward Postlethwait, Professor and Chair, Department of Environmental Health Sciences,
School of Public Health, University of Alabama at Birmingham, Birmingham, AL
Dr. Armistead (Ted) Russell*, Georgia Power Distinguished Professor of Environmental
Engineering, Environmental Engineering Group, School of Civil and Environmental
Engineering, Georgia Institute of Technology, Atlanta, GA
Dr. Richard Schlesinger, Associate Dean, Department of Biology, Dyson College, Pace
University, New York, NY
Dr. Christian Seigneur, Vice President, Atmospheric and Environmental Research, Inc., San
Ramon, CA
Dr. Elizabeth A. (Lianne) Sheppard, Research Professor, Biostatistics and Environmental &
Occupational Health Sciences, Public Health and Community Medicine, University of
Washington, Seattle, WA
Dr. Frank Speizer [M.D.]*, Edward Kass Professor of Medicine, Channing Laboratory,
Harvard Medical School, Boston, MA
Dr. George Thurston, Associate Professor, Environmental Medicine, NYU School of Medicine,
New York University, Tuxedo, NY
Dr. James Ultman, Professor, Chemical Engineering, Bioengineering Program, Pennsylvania
State University, University Park, PA
Dr. Ronald Wyzga, Technical Executive, Air Quality Health and Risk, Electric Power Research
Institute, P.O. Box 10412, Palo Alto, CA
March 2008 xxvii DRAFT-DO NOT QUOTE OR CITE
-------
U.S. Environmental Protection Agency
Science Advisory Board (SAB)
Staff Office Clean Air Scientific Advisory Committee (CASAC)
CASAC NOX and SOX Primary NAAQS Review Panel
(cont'd)
SCIENCE ADVISORY BOARD STAFF
Dr. Angela Nugent, CASAC Designated Federal Officer, 1200 Pennsylvania Avenue, N.W.,
Washington, DC, 20460, Phone: 202-343-9981, Fax: 202-233-0643 (nugent.angela@epa.gov)
(Physical/Courier/FedEx Address: Angela Nugent, Ph.D, EPA Science Advisory Board Staff
Office (Mail Code 1400F), Woodies Building, 1025 F Street, N.W., Room 3614, Washington,
DC 20004, Telephone: 202-343-9981)
* Members of the statutory Clean Air Scientific Advisory Committee (CASAC) appointed by the EPA
Administrator
March 2008 xxviii DRAFT-DO NOT QUOTE OR CITE
-------
Annex Abbreviations and Acronyms
a
ACS
ADP
a;
AIRE
AM
APEX
APHEA
AQCD
AQS
ATS
BAL
BALF
BHPN
BHR
Br
Cx T
Ca++
CAA
CALINE4
CAMP
CAPS
CAPs
CARS
CASAC
CC16
CDC
CHAD
CHF
CHS
CI
CMAQ
CO
C02
COD
brackets signifying concentration(s)
alpha; the ratio of a person's exposure to a pollutant of ambient
origin to the pollutant's ambient concentration
American Cancer Society
adenosine dinucleotide phosphate
air exchange rate for microenvironment /'
Asma Infantile Ricerca (Italian study)
alveolar macrophage
Air Pollution Exposure (model)
Air Pollution on Health: a European Approach (study)
Air Quality Criteria Document
Air Quality System (database)
American Thoracic Society
bronchoalveolar lavage
bronchoalveolar lavage fluid
7V-bis(2-hydroxyl-propyl)nitrosamine
bronchial hyperresponsiveness
bromine
concentration x time; concentration times duration of exposure
calcium ion
Clean Air Act
California line source dispersion (model)
Childhood Asthma Management Program
cavity attenuated phase shift (monitor)
concentrated ambient particles
California Air Resources Board
Clean Air Scientific Advisory Committee
Clara cell 16-kDa protein
Centers for Disease Control and Prevention
Consolidated Human Activity Database
congestive heart failure
Children's Health Study
confidence interval
Community Multiscale Air Quality (model)
carbon monoxide
carbon dioxide
coefficient of divergence
March 2008
XXIX
DRAFT-DO NOT QUOTE OR CITE
-------
CoH
COPD
CRP
CTM
CVD
DEPcCBP
DHHS
DMA
DMN
DNA
DOAS
Ea
EC
ECP
ED
ELF
Ena
EPA
EPO
ER
ETS
FEF25
FEF75
F£NO
FEVo.s
FEVi
Finfi
FRM
FVC
GAM
GEE
GEOS-CHEM
GIS
GM-CSF
GSH
GST
H+
HCHO
HDL
coefficient of haze
chronic obstructive pulmonary disease
C-reactive protein
Chemistry-transport model
cardiovascular disease
diesel exhaust particulates extract-coated carbon black particles
U.S. Department of Health and Human Services
dimethylamine
dimethylnitrosamine
deoxyribonucleic acid
differential optical absorption spectroscopy
a person's exposure to pollutants of ambient origin
elemental carbon
eosinophil cationic protein
emergency department
epithelial lining fluid
a person's exposure to pollutants that are not of ambient origin
U.S. Environmental Protection Agency
eosinophil peroxidase
emergency room
environmental tobacco smoke
forced expiratory flow at 25% of vital capacity
forced expiratory flow at 25 to 75% of vital capacity
forced expiratory flow at 75% of vital capacity
fractional exhaled nitric oxide
forced expiratory volume in 0.5 second
forced expiratory volume in 1 second
the infiltration factor for microenvironment /'
Federal Reference Method
forced vital capacity
Generalized Additive Model(s)
generalized estimating equation(s)
three-dimensional, global model of atmospheric chemistry driven by
assimilated Goddard Earth Orbiting System observations
Geographic Information System
granulocyte-macrophage colony stimulating factor
glutathione
glutathione S-transferase (e.g., GSTM1, GSTP1, GSTT1)
hydrogen ion
formaldehyde
high-density lipoprotein cholesterol
March 2008
XXX
DRAFT-DO NOT QUOTE OR CITE
-------
HNO3
HNO4
HONO
HR
HRV
HS
H2SO4
hv
ICAM-1
ICD, ICD9
id
Ig
fflD
IIASA
IL
He
IN
IOM
IQR
IS
ISA
ISAAC
ki
LDH
LIF
LOESS
LRD
LT
MEF25
MEF50
MEF75
MENTOR
MI
MMEF
MoOx
MOZART
MPO
MPP
MSA
N
n
nitric acid
pernitric acid
nitrous acid
heart rate
heart rate variability
hemorrhagic stroke
sulfuric acid
solar ultraviolet proton
intercellular adhesion molecule-1
International Classification of Diseases, Ninth Revision
identification
immunoglobulin (e.g., IgA, IgE, IgG)
ischemic heart disease
International Institute for Applied Systems Analysis
interleukin (e.g., IL-6, IL-8)
isoleucine
inorganic particulate species
Institute of Medicine
interquartile range
ischemic stroke
Integrated Science Assessment
International Study of Asthma and Allergies in Children
pollutant specific decay rate in microenvironment /'
lactate dehydrogenase
laser-induced fluorescence
locally estimated smoothing splines
lower respiratory disease
leukotriene (e.g., LTB4, LTC4, LTD4, LTE4)
maximal expiratory flow at 25%
maximal expiratory flow at 50%
maximal expiratory flow at 75%
Modeling Environment for Total Risk
myocardial infarction
maximal midexpiratory flow
molybdenum oxide
Model for Ozone and Related Chemical Tracers
myeloperoxidase
multiphase processes
metropolitan statistical area
nitrogen
number of observations
March 2008
XXXI
DRAFT-DO NOT QUOTE OR CITE
-------
Na+
NAAQS
NaAsO2
NAL
NAMS
NAS
NCo.oi-o.io
NCHS
NCICAS
NDMA
NEI
NERL
2NF
NHAPS
NHIS
NHX
NK
NLCS
NMMAPS
NMOR
NN
NO
NO2
NO2
NO3
NO3
NOX
NOY
NOZ
NOAANCEP
1NP
2NP
NR,N/R
NRC
NSA
03
oc
OH
sodium ion
National Ambient Air Quality Standards
sodium arsenite
nasal lavage
National Air Monitoring Stations
National Academy of Sciences
particle number concentration for particle aerodynamic diameter
between 10 and 100 nm
National Center for Health Statistics
National Cooperative Inner-City Asthma Study
7V-nitrosodimethylamine
National Emissions Inventory
National Exposure Research Laboratory
2-nitrofluoranthene
National Human Activity Pattern Survey
National Health Interview Survey
reduced nitrogen compounds (NH3, NH4+)
natural killer (lymphocytes)
the Netherlands Cohort Study on Diet and Cancer
National Morbidity, Mortality, and Air Pollution Study
7V-nitrosomorpholine
nitronapthalene
nitric oxide
nitrogen dioxide
nitrite ion
nitrate radical
nitrate ion
sum of NO and NO2
sum of NOX and NOZ, total oxidized nitrogen
sum of all inorganic and organic reaction products of NOX (HONO,
HNO3, HNO4, organic nitrates, particulate nitrate, nitro-PAHS, etc.)
U.S. National Oceanic and Atmospheric Administration's National
Center for Environmental Prediction
1-nitropyrene
2-nitropyrene
not reported
National Research Council
nitrosating agent
ozone
organic carbon
hydroxyl radical
March 2008
XXXll
DRAFT-DO NOT QUOTE OR CITE
-------
OR
OVA
P,P
P90
PAARC
PAF
PAHs
PAMS
PAN
PANs
PaO2
Pb
PD20-FEVi
PD100
PEACE
PEF
PEFR
Pi
PIH
PM
PMiQ-2.5
PM2.5
PMN
pN03
POM
ppb
ppm
ppt
PRB
PT
PUFA
R
r
R2
RAPS
RCS
RONO2
odds ratio
ovalbumin
probability value
90th percentile
French air pollution and chronic respiratory diseases study
paroxysmal atrial fibrillation
polycyclic aromatic hydrocarbons
Photochemical Aerometric Monitoring System
peroxyacetyl nitrate
peroxyacyl nitrates
pressure of arterial oxygen
lead
provocative dose that produces a 20% decrease in FEVi
provocative dose that produces a 100% increase in SRaw
Pollution Effects on Asthmatic Children in Europe (study)
peak expiratory flow
peak expiratory flow rate
pollutant specific penetration coefficient for microenvironment /
primary intracerebral hemorrhage
particulate matter
particulate matter with an aerodynamic diameter of < lOjim
coarse particulate matter
fine particulate matter
polymorphonuclear leukocytes
particulate nitrate
particulate organic matter
parts per billion (by volume)
parts per million (by volume)
parts per trillion (by volume)
Policy Relevant Background
prothombin time
polyunsaturated fatty acids
intraclass correlation coefficient; organic radical
correlation coefficient
coefficient of determination
Pearson's correlation coefficient
Spearman's rank correlation coefficient
Regional Air Pollution Study
random component superposition
organic nitrates
March 2008
XXXlll
DRAFT-DO NOT QUOTE OR CITE
-------
ROS
RR
RSV
S
SAPALDIA
SAR
SCE
SD
SE
SEP
SES
SGA
SHEDS
SIDS
SLAMS
SO2
SO42"
SRaw
STN
T
TEA
Th2
TNF
TSP
TVOCs
TX
UFP
URI
V
Val
VOCs
VWF
WBC
Yi
reactive oxygen species
relative risk
respiratory syncytial virus
microenvironmental source strength
Study of Air Pollution and Lung Diseases in Adults
Site Audit Report
sister chromatid exchange
standard deviation
standard error
social-economic position
social-economic status
small for gestational age
Simulation of Human Exposure and Dose System
sudden infant death syndrome
State and Local Air Monitoring Stations
sulfur dioxide
sulfate ion
specific airways resistance
Speciation Trends Network
tau; atmospheric lifetime
triethanolamine
T-derived helper 2 lymphocyte
tumor necrosis factor (e.g., TNF-a)
total suspended particulates
total volatile organic compounds
thromboxane (e.g., TXA2, TXB2)
ultrafine particles; <0.1 jim diameter
upper respiratory infections
volume of the microenvironment
valine
volatile organic compounds
von Willibrand Factor
white blood cell
the fraction of time people spend in microenvironment /
the fraction of time people spend outdoors
Fisher's transform of the correlation coefficient
March 2008
XXXIV
DRAFT-DO NOT QUOTE OR CITE
-------
i AX1. CHAPTER 1 ANNEX - FRAMEWORK FOR
2 REVIEW
o
4 The Integrated Science Assessment (ISA) presents a concise synthesis of the most policy-
5 relevant science to form the scientific foundation for the review of the primary (health-based)
6 National Ambient Air Quality Standards (NAAQS) for nitrogen dioxide (NO2) (U.S.
7 Environmental Protection Agency, 2007). The Annexes: (1) provide more details of the most
8 pertinent scientific literature relative to the review of the NO2 NAAQS in the areas of
9 atmospheric sciences, air quality analyses, exposure assessment, dosimetry, controlled human
10 exposure studies, toxicology, and epidemiology; and (2) focus on the key policy relevant
11 questions and studies published since the last EPA review.
12 Annex 1 provides the legislative background and history of previous reviews of the
13 NAAQS for oxides of nitrogen. Annex 1 also includes more detailed information on the
14 methods used to identify and select studies, and on frameworks for evaluating scientific evidence
15 relative to causality determination. The overarching framework for evaluation of evidence for
16 causality used in the draft ISA is outlined in the introduction to that document, and this Annex
17 provides supporting information for that framework, including excerpts from decision
18 frameworks or criteria developed by other organizations.
19 Annex 2 presents evidence related to the physical and chemical processes controlling the
20 production, destruction, and levels of reactive nitrogen compounds in the atmosphere, including
21 both oxidized and reduced species. Annex 3 presents information on environmental
22 concentrations, patterns, and human exposure to ambient oxides of nitrogen; however, most
23 information relates to NO2. Annex 4 presents results from toxicological studies as well as
24 information on dosimetry of oxides of nitrogen. Annex 5 discusses results from controlled
25 human exposure studies, and Annex 6 presents evidence from epidemiologic studies. Finally,
26 Annex 7 is comprised of tables of the findings of epidemiologic studies of respiratory health
27 outcomes that also include descriptive statistics (e.g., mean, maximum) on the NO2 air quality
28 data used in the studies. These Annexes include more detailed information on health or exposure
29 studies that is summarized in tabular form, as well as more extensive discussion of atmospheric
30 chemistry, source, exposure, and dosimetry information. Annex tables for health studies are
31 generally organized to include information about (1) concentrations of oxides of nitrogen levels
March 2008 AX1-1 DRAFT-DO NOT QUOTE OR CITE
-------
1 or doses and exposure times, (2) description of study methods employed, (3) results and
2 comments, and (4) quantitative outcomes for oxides of nitrogen measures.
3
4
5 AX1.1 LEGISLATIVE REQUIREMENTS
6 Two sections of the Clean Air Act (CAA) govern the establishment and revision of the
7 national ambient air quality standards (NAAQS). Section 108 (U.S. Code, 2003a) directs the
8 Administrator to identify and list "air pollutants" that "in his judgment, may reasonably be
9 anticipated to endanger public health and welfare" and whose "presence in the ambient air results
10 from numerous or diverse mobile or stationary sources" and to issue air quality criteria for those
11 that are listed. Air quality criteria are intended to "accurately reflect the latest scientific
12 knowledge useful in indicating the kind and extent of identifiable effects on public health or
13 welfare which may be expected from the presence of [a] pollutant in ambient air."
14 Section 109 (U.S. Code, 2003b) directs the Administrator to propose and promulgate
15 "primary" and "secondary" NAAQS for pollutants listed under Section 108. Section 109(b)(l)
16 defines a primary standard as one "the attainment and maintenance of which in the judgment of
17 the Administrator, based on such criteria and allowing an adequate margin of safety, are requisite
18 to protect the public health."1 A secondary standard, as defined in Section 109(b)(2), must
19 "specify a level of air quality the attainment and maintenance of which, in the judgment of the
20 Administrator, based on such criteria, is required to protect the public welfare from any known
21 or anticipated adverse effects associated with the presence of [the] pollutant in the ambient air."2
22 The requirement that primary standards include 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 reasonable
25 degree of protection against hazards that research has not yet identified. See Lead Industries
26 Association v. EPA, 647 F.2d 1130, 1154 (D.C. Cir 1980), cert, denied, 449 U.S. 1042 (1980);
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" [U.S. Senate (1970)].
2 Welfare effects as defined in Section 302(h) [U.S. Code, (2005)] 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."
March 2008 AX1-2 DRAFT-DO NOT QUOTE OR CITE
-------
1 American Petroleum Institute v. Costle, 665 F.2d 1176, 1186 (D.C. Cir. 1981), cert, denied, 455
2 U.S. 1034(1982). Both kinds of uncertainties are components of the risk associated with
3 pollution at levels below those at which human health effects can be said to occur with
4 reasonable scientific certainty. Thus, in selecting primary standards that include an adequate
5 margin of safety, the Administrator is seeking not only to prevent pollution levels that have been
6 demonstrated to be harmful but also to prevent lower pollutant levels that may pose an
7 unacceptable risk of harm, even if the risk is not precisely identified as to nature or degree.
8 In selecting a margin of safety, the U.S. Environmental Protection Agency (EPA)
9 considers such factors as the nature and severity of the health effects involved, the size of
10 sensitive population(s) at risk, and the kind and degree of the uncertainties that must be
11 addressed. The selection of any particular approach to providing an adequate margin of safety is
12 a policy choice left specifically to the Administrator's judgment. See Lead Industries
13 Association v. EPA, supra, 647 F.2d at 1161-62.
14 In setting standards that are "requisite" to protect public health and welfare, as provided
15 in Section 109(b), EPA's task is to establish standards that are neither more nor less stringent
16 than necessary for these purposes. In so doing, EPA may not consider the costs of implementing
17 the standards. See generally Whitman v. American Trucking Associations, 531 U.S. 457, 465-
18 472, and 475-76 (U.S. Supreme Court, 2001).
19 Section 109(d)(l) requires that "not later than December 31, 1980, and at 5-year intervals
20 thereafter, the Administrator shall complete a thorough review of the criteria published under
21 Section 108 and the national ambient air quality standards and shall make such revisions in such
22 criteria and standards and promulgate such new standards as may be appropriate ...." Section
23 109(d)(2) requires that an independent scientific review committee "shall complete a review of
24 the criteria ... and the national primary and secondary ambient air quality standards ... and shall
25 recommend to the Administrator any new standards and revisions of existing criteria and
26 standards as may be appropriate ...." Since the early 1980s, this independent review function
27 has been performed by the Clean Air Scientific Advisory Committee (CAS AC) of EPA's
28 Science Advisory Board.
29
30
March 2008 AX1-3 DRAFT-DO NOT QUOTE OR CITE
-------
1 AX1.2 HISTORY OF REVIEWS OF THE PRIMARY NAAQS FOR NO2
2 On April 30, 1971, EPA promulgated identical primary and secondary NAAQS for
3 nitrogen dioxide (NO2), under Section 109 of the Act, set at 0.053 parts per million (ppm),
4 annual average (Federal Register, 1971). In 1982, EPA published Air Quality Criteria for
5 Oxides of Nitrogen (1982 NOX AQCD) (U.S. Environmental Protection Agency, 1982), which
6 updated the scientific criteria upon which the initial NO2 standards were based. On February 23,
7 1984, EPA proposed to retain these standards (Federal Register, 1984). After taking into account
8 public comments, EPA published the final decision to retain these standards on June 19, 1985
9 (Federal Register, 1985).
10 On July 22, 1987, EPA announced that it was undertaking plans to revise the 1982 NOx
11 air quality criteria (Federal Register, 1987). In November 1991, EPA released an updated draft
12 air quality criteria document (AQCD) for CAS AC and public review and comment (Federal
13 Register, 1991). The draft document provided a comprehensive assessment of the available
14 scientific and technical information on heath and welfare effects associated with NO2 and other
15 oxides of nitrogen. The CAS AC reviewed the document at a meeting held on July 1, 1993, and
16 concluded in a closure letter to the Administrator that the document "provides a scientifically
17 balanced and defensible summary of current knowledge of the effects of this pollutant and
18 provides an adequate basis for EPA to make a decision as to the appropriate NAAQS for MV
19 (Wolff, 1993).
20 The EPA also prepared a draft Staff Paper that summarized and integrated the key studies
21 and scientific evidence contained in the revised AQCD and identified the critical elements to be
22 considered in the review of the NO2 NAAQS. The Staff Paper received external review at a
23 December 12, 1994 CAS AC meeting. CAS AC comments and recommendations were reviewed
24 by EPA staff and incorporated into the final draft of the Staff Paper as appropriate. CAS AC
25 reviewed the final draft of the Staff Paper in June 1995 and responded by written closure letter
26 (Wolff, 1996). In September of 1995, EPA finalized the document entitled, "Review of the
27 National Ambient Air Quality Standards for Nitrogen Dioxide Assessment of Scientific and
28 Technical Information" (U.S. Environmental Protection Agency, 1995).
29 Based on that review, the Administrator announced her proposed decision not to revise
30 either the primary or the secondary NAAQS for NO2 (Federal Register, 1995). The decision not
31 to revise the NO2 NAAQS was finalized after careful evaluation of the comments received on the
March 2008 AX1-4 DRAFT-DO NOT QUOTE OR CITE
-------
1 proposal. The level for both the existing primary and secondary NAAQS for NC>2 is 0.053 ppm
2 annual arithmetic average, calculated as the arithmetic mean of the 1-h NC>2 concentrations.
3
4
5 AX1.3 LITERATURE SELECTION
6 Literature searches are conducted routinely, on a quarterly or monthly basis, to identify
7 studies published since the last review; the review includes publications from 1-2 years prior to
8 the publication of 1993 AQCD for Oxides of Nitrogen (U.S. Environmental Protection Agency,
9 1993). Examples of strategies used for literature searches are presented below. The search
10 strategies are periodically reexamined and modified in an effort to optimize the identification of
11 pertinent published papers. Additional papers are identified for inclusion in several ways. These
12 include the review of pre-publication tables of contents for journals in which relevant papers may
13 be published independent identification of relevant literature by expert authors. In addition,
14 publications that may be pertinent are identified by both the public and CAS AC during the
15 external review process. Generally, only information that has undergone scientific peer review
16 and that has been published (or accepted for publication) in the open literature is considered.
17 The following sections briefly summarize criteria for selection of studies for this draft ISA.
18 Figure AX1.3-1 depicts the selection process for studies included in the ISA, and two
19 tables are included below that offer examples of the keywords and strategies used to search the
20 literature. Table AX1.3-1 lists examples of the keywords used for identifying epidemiologic
21 studies on oxides of nitrogen for this review. The search strategy for atmospheric science and
22 exposure studies is outlined in Table AX1.3-2.
23 The studies identified through literature searches are further evaluated by EPA staff and
24 outside experts to determine if the studies merit inclusion in the ISA and/or its Annexes. The
25 criteria used for study selection are summarized below.
26
27 General Criteria for Study Selection
28 In assessing the scientific quality and relevance of epidemiological and human or animal
29 toxicological studies, the following considerations have been taken into account.
30 • To what extent are the aerometric data, exposure, or dose metrics of adequate quality
31 and sufficiently representative to serve as indicators of exposure to ambient NOx?
March 2008 AX1-5 DRAFT-DO NOT QUOTE OR CITE
-------
Identification of Studies for Inclusion in the ISA
Continuous,
comprehensive
literature review
of peer-reviewed
journal articles
Studies added
to the docket
during public
comment period.
Informative studies
are identified.
Studies identified
during EPA
sponsored kickoff
meeting (including
studies in
preparation).
Studies that do
not address
exposure and/or
effects of air
pollutant(s) under
review are
excluded.
Selection of
studies
discussed and
additional studies
identified during
CASACpeer
review of draft
document.
INFORMATIVE studies are well-designed,
properly implemented, thoroughly
described.
HIGHLY INFORMATIVE studies reduce
uncertainty on critical issues, may include
analyses of confounding or effect
modification by copollutants or other
variables, analyses of concentration-
response or dose-response relationships,
analyses related to time between
exposure and response, and offer
innovation in method or design.
POLICY-RELEVANT studies may include
those conducted at or near ambient
concentrations and studies conducted in
.U.S. and Canadian airsheds.
>w_ -~*
tudies are
evaluated for
inclusion in the ISA
and included
in the
Annex
Studies
summarized
in figures are
included
because they are
sufficiently
comparable
to be displayed
together. A study
highlighted in the
text may not
appear in a
summary figure.
A figure from a
highlighted study
maybe
reproduced
in its entirely.
Studies
REFERENCED
in the text include
those that provide
a basis for or
describe
the association
between the
criteria pollutant
and effects. In
addition, policy
relevant
and highly
informative
studies
are discussed
in the text.
Figure AX1.3-1. Selection process for studies included in ISA.
March 2008
AX1-6
DRAFT-DO NOT QUOTE OR CITE
-------
1 • Were the study populations adequately selected and are they sufficiently well defined
2 to allow for meaningful comparisons between study groups?
3 • Are the statistical analyses appropriate, properly performed, and properly interpreted?
4 • Are likely covariates (i.e., potential confounders or effect modifiers) adequately
5 controlled or taken into account in the study design and statistical analysis?
6 • Are the reported findings internally consistent, biologically plausible, and coherent in
7 terms of consistency with other known facts?
8 Consideration of these issues informs our judgments on the relative quality of individual studies
9 and allows us to focus the assessment on the most pertinent studies.
10
11 Criteria for Selecting Epidemiological Studies
12 In selecting epidemiological studies for this assessment, EPA considered whether a given
13 study contains information on (1) associations with measured oxides of nitrogen concentrations
14 using short- or long-term exposures at or near ambient levels of oxides of nitrogen, (2) health
15 effects of specific oxides of nitrogen species or indicators related to oxides of nitrogen sources
16 (e.g., motor vehicle emissions, combustion-related particles), (3) health endpoints and
17 populations not previously extensively researched, (4) multiple pollutant analyses and other
18 approaches to address issues related to potential confounding and modification of effects, and/or
19 (5) important methodological issues (e.g., lag of effects, model specifications, thresholds,
20 mortality displacement) related to interpretation of the health evidence. Among the
21 epidemiological studies, particular emphasis has been placed on those most relevant to reviews
22 of the NAAQS. Specifically, studies conducted in the United States or Canada may be discussed
23 in more detail than those from other geographic regions. Particular emphasis has been placed on:
24 (A) new multicity studies that employ standardized methodological analyses for evaluating
25 effects of oxides of nitrogen and that provide overall estimates for effects based on combined
26 analyses of information pooled across multiple cities, (B) new studies that provide quantitative
27 effect estimates for populations of interest, and (C) studies that consider oxides of nitrogen as a
28 component of a complex mixture of air pollutants.
29 Not all studies are accorded equal weight in the overall interpretive assessment of
30 evidence regarding NO2-associated health effects. Among well-conducted studies with adequate
31 control for confounding, increasing scientific weight is accorded in proportion to the precision of
March 2008 AX1-7 DRAFT-DO NOT QUOTE OR CITE
-------
1 their effect estimates. Small-scale studies without a wide range of exposures generally produce
2 less precise estimates compared to larger studies with a broad exposure gradient. For time-series
3 studies, the size of the study, as indicated by the length of the study period and total number of
4 events, and the variability of NO2 exposures are important components that help to determine the
5 precision of the health effect estimates. In evaluating the epidemiologic evidence in this chapter,
6 more weight is accorded to estimates from studies with narrow confidence bands.
7 The goal is to perform a balanced and objective evaluation that summarizes, interprets,
8 and synthesizes the most important studies and issues in the epidemiologic database pertaining to
9 oxides of nitrogen exposure, illustrated using newly created or previously published summary
10 tables and figures. For each study presented, the quality of the exposure and outcome data as
11 well as the quality of the statistical analysis methodology are discussed. The discussion
12 incorporates the magnitude and statistical strengths of observed associations between NO2
13 exposure and health outcomes.
14
15 Criteria for Selecting Animal and Human Toxicological Studies
16 Criteria for the selection of research evaluating animal toxicological or controlled human
17 exposure studies include a focus on those studies conducted at levels within about an order of a
18 magnitude of ambient NC>2 concentrations and those studies that approximate expected human
19 exposure conditions in terms of concentration and duration. Studies that elucidate mechanisms
20 of action and/or susceptibility, particularly if the studies were conducted under atmospherically
21 relevant conditions, are emphasized whenever possible.
22 The selection of research evaluating controlled human exposures to oxides of nitrogen is
23 mainly limited to studies in which subjects were exposed to <5 ppm NC>2. For these controlled
24 human exposures, emphasis is placed on studies that (1) investigate potentially susceptible
25 populations such as asthmatics, particularly studies that compare responses in susceptible
26 individuals with those in age-matched healthy controls; (2) address issues such as concentration-
27 response or time-course of responses; (3) investigate exposure to NC>2 separately and in
28 combination with other pollutants such as 63 and 862; (4) include control exposures to filtered
29 air; and (5) have sufficient statistical power to assess findings.
30
31
March 2008 AX1-8 DRAFT-DO NOT QUOTE OR CITE
-------
1 AX1.4 EVALUATION GUIDELINES
2
3 AX1.4.1 Background on Causality Decision Framework
4 The critical assessment of health evidence presented in the ISA is conceptually based
5 upon consideration of salient aspects of the evidence so as to reach fundamental judgments as to
6 the likely causal significance of the observed associations. In so doing, it is appropriate to draw
7 from those aspects initially presented in Hill's classic monograph (Hill, 1965) and widely used
8 by the scientific community in conducting such evidence-based reviews. A number of these
9 aspects are judged to be particularly salient in evaluating the body of evidence available in this
10 review, including the aspects described by Hill as strength, experiment, consistency, plausibility,
11 and coherence. Other aspects identified by Hill, including temporality and biological gradient,
12 are also relevant and considered here (e.g., in characterizing lag structures and concentration-
13 response relationships), but are more directly addressed in the design and analyses of the
14 individual epidemiologic studies included in this assessment. (As noted below, Hill's remaining
15 aspects of specificity and analogy are not considered to be particularly salient in this
16 assessment.) As discussed below, these salient aspects are interrelated and considered
17 throughout the evaluation of the evidence presented in this chapter, and are more generally
18 reflected in the ISA.
19 In the following sections, the general evaluation of the strength of the epidemiological
20 evidence reflects consideration not only of the magnitude of reported oxides of nitrogen effects
21 estimates and their statistical significance, but also of the precision of the effects estimates and
22 the robustness of the effects associations. Consideration of the robustness of the associations
23 takes into account a number of factors, including in particular the impact of alternative models
24 and model specifications and potential confounding by copollutants, as well issues related to the
25 consequences of measurement error. Another aspect that is related to the strength of the
26 evidence in this assessment is the availability of evidence from "found experiments", or so-called
27 intervention studies, which have the potential to provide particularly strong support for making
28 causal inferences.
29 Consideration of the consistency of the effects associations, as discussed in the following
30 sections, involves looking across the results of multi- and single-city studies conducted by
31 different investigators in different places and times. In this assessment of ambient oxides of
March 2008 AX1-9 DRAFT-DO NOT QUOTE OR CITE
-------
1 nitrogen—health effects associations, it is important to consider the aspect of consistency. Other
2 relevant factors are also known to exhibit much variation across studies. These include, for
3 example, the presence and levels of copollutants, the relationships between central measures of
4 oxides of nitrogen and exposure-related factors, relevant demographic factors related to sensitive
5 subpopulations, as well as climatic and meteorological conditions. Thus, in this case,
6 consideration of consistency, and the related heterogeneity of effects issue, is appropriately
7 understood as an evaluation of the similarity or general concordance of results, rather than an
8 expectation of finding quantitative results within a very narrow range. Particular weight is given
9 in this assessment, consistent with Hill's views, to the presence of "similar results reached in
10 quite different ways, e.g., prospectively and retrospectively" (Hill, 1965). On the other hand, in
11 light of complexities of exposure and surrogate issues and its spatial and temporal variations,
12 Hill's specificity of effects and analogy aspects are not viewed as being particularly salient here.
13 Looking beyond the epidemiological evidence, evaluation of the biological plausibility of
14 the oxides of nitrogen—health effect associations observed in epidemiologic studies reflects
15 consideration of both exposure-related factors and dosimetric/toxicologic evidence relevant to
16 identification of potential biological mechanisms. Similarly, consideration of the coherence of
17 health effects associations reported in the epidemiologic literature reflects broad consideration of
18 information pertaining to the nature of the various respiratory- and cardiac-related mortality and
19 morbidity effects and biological markers evaluated in toxicologic and epidemiologic studies.
20 In identifying these aspects as being particularly salient in this assessment, it is also
21 important to recognize that no one aspect is either necessary or sufficient for drawing inferences
22 of causality. As Hill (1965) emphasized:
23 "None of my nine viewpoints can bring indisputable evidence for or against the
24 cause-and-effect hypothesis and none can be required as a sine qua non. What
25 they can do, with greater or less strength, is to help us to make up our minds on
26 the fundamental question — is there any other way of explaining the set of facts
27 before us, is there any other answer equally, or more, likely than cause and
28 effect?"
29 Thus, while these aspects frame considerations weighed in assessing the epidemiologic
30 evidence, they do not lend themselves to being considered in terms of simple formulas or hard-
31 and-fast rules of evidence leading to answers about causality (Hill, 1965). One, for example,
32 cannot simply count up the numbers of studies reporting statistically significant results for oxides
33 of nitrogen and health endpoints evaluated in this assessment and reach credible conclusions
March 2008 AX1-10 DRAFT-DO NOT QUOTE OR CITE
-------
1 about the relative strength of the evidence and the likelihood of causality. Rather, these
2 important considerations are taken into account throughout this assessment with a goal of
3 producing an objective appraisal of the evidence (informed by peer and public comment and
4 advice), which includes the weighing of alternative views on controversial issues.
5
6 AX1.4.2 Approaches to the Determination of Causality
7 The following sections include excerpts from several reports that have documented
8 approaches for the determination of causality, or related decision-making processes. These
9 sections provide supplementary documentation of approaches that are similar in nature to EPA's
10 framework for evaluation of health evidence.
11
12 AX1.4.2.1 Surgeon General's Report: The Health Consequences of Smoking (CDC, 2004)
13 The Surgeon General's Report (CDC, 2004) evaluates the health effects of smoking; it
14 builds upon the first Surgeon General's report published in 1964 (USDEHW, 1964). It also
15 updates the methodology for evaluating evidence that was first presented in the 1964 report. The
16 2004 report acknowledges the effectiveness of the previous methodology, but attempts to
17 standardize the language surrounding causality of associations.
18 The Surgeon General's Reports on Smoking play a central role in the translation of
19 scientific evidence into policy. As such, it is important that scientific evidence is presented in a
20 manner that conveys most succinctly the link between smoking and a health effect. Specifically,
21 the report states:
22 The statement that an exposure "causes" a disease in humans represents a
23 serious claim, but one that carries with it the possibility of prevention. Causal
24 determinations may also carry substantial economic implications for society and
25 for those who might be held responsible for the exposure or for achieving its
26 prevention.
27 To address the issue of identifying causality, the 2004 report provides the following
28 summary of the earlier 1964 report:
29 When a relationship or an association between smoking... and some condition in
3 0 the host was noted, the significance of the association was assessed.
31 The characterization of the assessment called for a specific term. ... The word
32 cause is the one in general usage in connection with matters considered in this
3 3 study, and it is capable of conveying the notion of a significant, effectual
34 relationship between an agent and an associated disorder or disease in the host.
March 2008 AX1-11 DRAFT-DO NOT QUOTE OR CITE
-------
1 No member was so naive as to insist upon mono-etiology in pathological
2 processes or in vital phenomena. All were thoroughly aware... that the end
3 results are the net effect of many actions and counteractions.
4 Granted that these complexities were recognized, it is to be noted clearly that the
5 Committee's considered decision to use the words "a cause," or "a major cause,"
6 or "a significant cause," or "a causal association" in certain conclusions about
7 smoking and health affirms their conviction (USDHEW, 1964, p. 21).
8 This 2004 report creates uniformly labeled conclusions that are used throughout the document.
9 The following excerpts from the report include a description of the methodology and the
10 judgments used to reach a conclusion:
11 Terminology of Conclusions and Causal Claims
12 The first step in introducing this revised approach is to outline the language that
13 will be used for summary conclusions regarding causality, which follows
14 hierarchical language used by Institute of Medicine committees (Institute of
15 Medicine, 1999) to couch causal conclusions, and by IARC to classify
16 carcinogenic substances (IARC, 1986). These entities use a four-level hierarchy
17 for classifying the strength of causal inferences based on available evidence as
18 follows:
19 A. Evidence is sufficient to infer a causal relationship.
20 B. Evidence is suggestive but not sufficient to infer a causal relationship.
21 C. Evidence is inadequate to infer the presence or absence of a causal
22 relationship (which encompasses evidence that is sparse, of poor
23 quality, or conflicting).
24 D. Evidence is suggestive of no causal relationship.
25 For this report, the summary conclusions regarding causality are expressed in
26 this four-level classification. Use of these classifications should not constrain
27 the process of causal inference, but rather bring consistency across chapters and
28 reports, and greater clarity as to what the final conclusions are actually saying.
29 As shown in Table 1.1 [see original document], without a uniform classification
30 the precise nature of the final judgment may not always be obvious, particularly
31 when the judgment is that the evidence falls below the "sufficient" category.
32 Experience has shown that the "suggestive" category is often an uncomfortable
33 one for scientists, since scientific culture is such that any evidence that falls
34 short of causal proof is typically deemed inadequate to make a causal
3 5 determination. However, it is very useful to distinguish between evidence that is
3 6 truly inadequate versus that which just falls short of sufficiency.
37 There is no category beyond "suggestive of no causal relationship" as it is
3 8 extraordinarily difficult to prove the complete absence of a causal association.
39 At best, "negative" evidence is suggestive, either strongly or weakly. In
40 instances where this category is used, the strength of evidence for no
41 relationship will be indicated in the body of the text. In this new framework,
42 conclusions regarding causality will be followed by a section on implications.
43 This section will separate the issue of causal inference from recommendations
44 for research, policies, or other actions that might arise from the causal
March 2008 AX1-12 DRAFT-DO NOT QUOTE OR CITE
-------
1 conclusions. This section will assume a public health perspective, focusing on
2 the population consequences of using or not using tobacco and also a scientific
3 perspective, proposing further research directions. The proportion of cases in
4 the population as a result of exposure (the population attributable risk), along
5 with the total prevalence and seriousness of a disease, are more relevant for
6 deciding on actions than the relative risk estimates typically used for etiologic
7 determinations. In past reports, the failure to sharply separate issues of
8 inference from policy issues resulted in inferential statements that were
9 sometimes qualified with terms for action. For example, based on the evidence
10 available in 1964, the first Surgeon General's report on smoking and health
11 contained the following statement about the relationship between cardiovascular
12 diseases and smoking:
13 It is established that male cigarette smokers have a higher death rate
14 from coronary artery disease than non-smoking males. Although the
15 causative role of cigarette smoking in deaths from coronary disease is
16 not proven, the Committee considers it more prudent from the public
17 health viewpoint to assume that the established association has
18 causative meaning, than to suspend judgment until no uncertainty
19 remains (USDHEW, 1964, p. 32).
20 Using this framework, this conclusion would now be expressed differently,
21 probably placing it in the "suggestive" category and making it clear that
22 although it falls short of proving causation, this evidence still makes causation
23 more likely than not. The original statement makes it clear that the 1964
24 committee judged that the evidence fell short of proving causality but was
25 sufficient to justify public health action. In this report, the rationale and
26 recommendations for action will be placed in the implications section, separate
27 from the causal conclusions. This separation of inferential from action-related
28 statements clarifies the degree to which policy recommendations are driven by
29 the strength of the evidence and by the public health consequences acting to
30 reduce exposure. In addition, this separation appropriately reflects the
31 differences between the processes and goals of causal inference and decision
32 making.
33 Judgment in Causal Inference
34 Making causal inferences from observational data can be a challenging task,
3 5 requiring expert judgment as to the likely sources and magnitude of
3 6 confounding, together with judgments about how well the existing constellation
37 of study designs, results, and analyses addresses this potential threat to
3 8 inferential validity. To aid this judgment, criteria for the determination of a
3 9 cause have been proposed by many philosophers and scientists over the
40 centuries. The most widely cited criteria in epidemiology and public health
41 more generally were set forth by Sir Austin Bradford Hill in 1965 (Weed, 2000).
42 Five of the nine criteria he listed were also put forward in the 1964 Surgeon
43 General's report as the criteria for causal judgment: consistency, strength,
44 specificity, temporality, and coherence of an observed association. Hill also
45 listed biologic gradient (dose-response), plausibility, experiment (or natural
46 experiment), and analogy. Many of these criteria have been cited in earlier
47 epidemiologic writings (Lilienfeld, 1959; Yerushalmy and Palmer, 1959;
48 Sartwell, 1960), and Susser has extensively refined them by exploring their
49 justification, merits, and interpretations (Susser, 1973, 1977; Kaufman and
50 Poole, 2000).
51
March 2008 AX1-13 DRAFT-DO NOT QUOTE OR CITE
-------
1 Hill (1965) clearly stated that these criteria were not intended to serve as a checklist:
2 Here are then nine different viewpoints from all of which we should study
3 association before we cry causation. What I do not believe... is that we can
4 usefully lay down some hard-and-fast rules of evidence that must be obeyed
5 before we accept cause and effect. None of my nine viewpoints can bring
6 indisputable evidence for or against the cause-and-effect hypothesis and none
7 can be required as a sine qua non. What they can do, with greater or less
8 strength, is to help us to make up our minds on the fundamental question—is
9 there any other way of explaining the facts before us, is there any other answer
10 equally, or more, likely than cause and effect? (Hill, 1965, p. 299)
11 All of these criteria were meant to be applied to an already established statistical
12 association; if no association has been observed, then these criteria are not
13 relevant. Hill explained how, if a given criterion were satisfied, it strengthened a
14 causal claim. Each of these nine criteria served one of two purposes: either as
15 evidence against competing noncausal explanations or as evidence supporting
16 causal ones. Noncausal explanations for associations include chance; residual or
17 unmeasured confounding; model misspecification; selection bias; errors in
18 measurement of exposure, confounders, or outcome; and issues regarding
19 missing data (which can also include missing studies, e.g., publication bias).
20 The criteria are briefly discussed below.
21 Consistency
22 This criterion refers to the persistent finding of an association between exposure
23 and outcome in multiple studies of adequate power, and in different persons,
24 places, circumstances, and times. Consistency can serve two purposes. The first
25 purpose, which was discussed previously, is to make unmeasured confounding
26 an unlikely alternative explanation for an observed association. Such
27 confounding would have to persist across diverse populations, exposure
28 opportunities, and measurement methods. The confounding is still possible if
29 the exposure (in this case smoking) were very strongly tied to an alternative
3 0 cause, as was claimed in the form of the "constitutional hypothesis" put forward
31 in the early days of the smoking-disease debate (USDHEW, 1964). This
32 hypothesis held that there was a constitutional (i.e., genetic) factor that made
33 people more likely to both smoke and develop cancer. So consistency serves
34 mainly to rule out the hypothesis that the association is produced by an ancillary
3 5 factor that differs across studies, but not one factor that is common to all or most
36 of them (Rothman and Greenland, 1998).
3 7 The second purpose of the consistency criterion is to make the hypothesis of a
3 8 chance effect unlikely by increasing the statistical strength of a finding through
3 9 the accumulation of a larger body of data. It does not include the qualitative
40 strength of such studies, which Susser subsumes under his subsidiary concept of
41 "survivability," relating to the rigor and severity of tests of association (Susser,
42 1991).
43 Strength of Association
44 This criterion includes two dimensions of strength: the magnitude of the
45 association and its statistical strength. An association strong in both aspects
46 makes the alternative explanations of chance and confounding unlikely. The
47 larger the measured effect, the less likely that an unmeasured or poorly
48 controlled confounder could account for it completely. Associations that have a
49 small magnitude or a weak statistical strength are more likely to reflect chance,
50 modest bias, or unmeasured weak confounding. However, the magnitude of
March 2008 AX1-14 DRAFT-DO NOT QUOTE OR CITE
-------
1 association is reflective of underlying biologic processes and should be
2 consistent with understanding the role of smoking in these processes.
3 Specificity
4 Specificity has been interpreted to mean both a single (or few) effect(s) of one
5 cause, or no more than one possible cause for one effect. In addition to specific
6 infectious diseases that are caused by specific infectious agents, some other
7 examples include asbestos exposure and mesothelioma and thalidomide
8 exposure during gestation and the resulting unusual constellation of birth
9 defects. This criterion is rarely used as it was originally proposed, having been
10 derived primarily from the Koch Postulates for infectious causes of disease
11 (Evans, 1993). When specificity exists, it can strengthen a causal claim, but its
12 absence does not weaken it (Sartwell, 1960). For example, most cancers are
13 known to have multifactorial etiologies, many cancer-causing agents can cause
14 several types of cancer, and these agents can also have noncancerous effects.
15 Similarly, there are multiple causes of cardiovascular disease.
16 In considering specificity in relation to the smoking-lung cancer association, the
17 1964 Surgeon General's report (USDHEW, 1964) provides a rich discussion of
18 this criterion. The committee recognized the linkage between this criterion and
19 strength of association and offered a symmetric formulation of specificity in the
20 relationship between exposure and disease; that is, a particular exposure always
21 results in a particular disease and the disease always results from the exposure.
22 The committee acknowledged that smoking does not always result in lung
23 cancer and that lung cancer has other causes. The report notes the extremely
24 high relative risk for lung cancer in smokers and the high attributable risk, and
25 concludes that the association between smoking and lung cancer has "a high
26 degree of specificity."
27 Temporality
28 Temporality refers to the occurrence of a cause before its purported effect.
29 Temporality is the sine qua non of causality, as a cause clearly cannot occur
30 after its purported effect. Failure to establish temporal sequence seriously
31 weakens a causal claim, but establishing temporal precedence is by itself not
3 2 very strong evidence in favor of causality.
33 Coherence, Plausibility, and Analogy
34 Although the original definitions of these criteria were subtly different, in
3 5 practice they have been treated essentially as one idea: that a proposed causal
3 6 relationship not violate known scientific principles, and that it be consistent with
37 experimentally demonstrated biologic mechanisms and other relevant data, such
38 as ecologic patterns of disease (Rothman and Greenland, 1998). In addition, if
3 9 biologic understanding can be used to set aside explanations other than a causal
40 association, it offers further support for causality. Together, these criteria can
41 serve both to support a causal claim (by supporting the proposed mechanism) or
42 refute it (by showing that the proposed mechanism is unlikely).
43 Biologic understanding, of course, is always evolving as scientific advances
44 make possible an ever deeper exploration of disease pathogenesis. For example,
45 in 1964 the Surgeon General's committee found a causal association of smoking
46 with lung cancer to be biologically plausible. Nearly 40 years later, this
47 association remains biologically plausible, but that determination rests not only
48 on the earlier evidence but on more recent findings that address the genetic and
49 molecular basis of carcinogenesis.
March 2008 AX1-15 DRAFT-DO NOT QUOTE OR CITE
-------
1 Biologic Gradient (Dose-Response)
2 The finding of an increment in effect with an increase in the strength of the
3 possible cause provides strong support in favor of a causal hypothesis. This is
4 not just because such an observation is predicted by many cause-effect models
5 and biologic processes, but more importantly, because it makes most noncausal
6 explanations very unlikely. One would have to posit that some unmeasured
7 factor was changing in the same manner as the exposure of interest if that factor,
8 rather than the factor of interest, is to explain the gradient. Except for
9 confounders that are very closely related to a causal factor, it is very difficult for
10 such a pattern to be created by virtually any of the noncausal explanations for an
11 association listed earlier. The finding of a dose-response relationship has long
12 been a mainstay of causal arguments in smoking investigations; virtually all
13 health outcomes causally linked to smoking have shown an increase in risk
14 and/or severity with an increase in the lifetime smoking history, generally
15 number of cigarettes smoked per day, duration of smoking, or a cumulative
16 measure of consumption. This criterion is not based on any specific shape of the
17 dose-response relationship.
18 Experiment
19 This criterion refers to situations where natural conditions might plausibly be
20 thought to imitate conditions of a randomized experiment, producing a "natural
21 experiment" whose results might have the force of a true experiment. An
22 experiment is typically a situation in which a scientist controls who is exposed
23 in a way that does not depend on any of the subject's characteristics. Sometimes
24 nature produces similar exposure patterns. The reduction in risk after smoking
25 cessation serves as one such situation that approximates an experiment; an
26 alternative noncausal explanation would have to posit that an unmeasured causal
27 factor of that health outcome was more frequent among those who did not stop
28 smoking than among those who did. The causal interpretation is further
29 strengthened if risk continues to decline in former smokers with increasing
30 length of time since quitting. Similar to the dose-response criteria, observations
31 of risk reduction after quitting smoking have the dual effects of making most
32 noncausal explanations unlikely, and supporting the biologic model that
33 underlies the causal claim.
34
35 AX1.4.2.2 The EPA Guidelines for Carcinogen Risk Assessment
36 The EPA Guidelines for Carcinogen Risk Assessment, published in 2005 (U.S.
37 Environmental Protection Agency, 2005), is an update to the previous risk assessment document
38 published in 1986. This document serves to guide EPA staff and public about the Agency's risk
39 assessment development and methodology. In the 1986 Guidelines, a step-wise approach was
40 used to evaluate the scientific findings. However, this newer document is similar to the Surgeon
41 General's Report on Smoking in that it uses single integrative step after assessing all of the
42 individual lines of evidence. Five standard descriptors are used to evaluate the weight of
43 evidence:
44 • Carcinogenic to Humans
45 • Likely to Be Carcinogenic to Humans
March 2008 AX1-16 DRAFT-DO NOT QUOTE OR CITE
-------
1 • Suggestive Evidence of Carcinogenic Potential
2 • Inadequate Information to Assess Carcinogenic Potential
3 • Not Likely to Be Carcinogenic to Humans.
4 The 2005 Guidelines recommend that a separate narrative be prepared on the weight of evidence
5 and the descriptor. The Guidelines further recommend that the descriptors should only be used
6 in the context of a weight-of-evidence discussion.
7 The following excerpt describes how a weight of evidence narrative should be developed
8 and a how a descriptor should be selected (U.S. Environmental Protection Agency, 2005):
9 The weight of the evidence should be presented as a narrative laying out the
10 complexity of information that is essential to understanding the hazard and its
11 dependence on the quality, quantity, and type(s) of data available, as well as the
12 circumstances of exposure or the traits of an exposed population that may be
13 required for expression of cancer. For example, the narrative can clearly state to
14 what extent the determination was based on data from human exposure, from
15 animal experiments, from some combination of the two, or from other data.
16 Similarly, information on mode of action can specify to what extent the data are
17 from in vivo or in vitro exposures or based on similarities to other chemicals.
18 The extent to which an agent's mode of action occurs only on reaching a
19 minimum dose or a minimum duration should also be presented. A hazard
20 might also be expressed disproportionately in individuals possessing a specific
21 gene; such characterizations may follow from a better understanding of the
22 human genome. Furthermore, route of exposure should be used to qualify a
23 hazard if, for example, an agent is not absorbed by some routes. Similarly, a
24 hazard can be attributable to exposures during a susceptible lifestage on the
25 basis of our understanding of human development.
26 The weight of evidence-of-evidence narrative should highlight:
27 • the quality and quantity of the data;
28 • all key decisions and the basis for these major decisions; and
29 • any data, analyses, or assumptions that are unusual for or new to EPA.
30 To capture this complexity, a weight of evidence narrative generally includes
31 • conclusions about human carcinogenic potential (choice of
32 descriptor(s), described below)
33 • a summary of the key evidence supporting these conclusions (for each
34 descriptor used), including information on the type(s) of data (human
3 5 and/or animal, in vivo and/or in vitro) used to support the conclusion(s)
36 • available information on the epidemiologic or experimental conditions
37 that characterize expression of carcinogenicity (e.g., if carcinogenicity
38 is possible only by one exposure route or only above a certain human
39 exposure level),
40 • a summary of potential modes of action and how they reinforce the
41 conclusions,
42 • indications of any susceptible populations or lifestages, when available,
43 and
44 • a summary of the key default options invoked when the available
45 information is inconclusive.
March 2008 AX1-17 DRAFT-DO NOT QUOTE OR CITE
-------
1 To provide some measure of clarity and consistency in an otherwise free-form
2 narrative, the weight of evidence descriptors are included in the first sentence of
3 the narrative. Choosing a descriptor is a matter of judgment and cannot be
4 reduced to a formula. Each descriptor may be applicable to a wide variety of
5 potential data sets and weights of evidence. These descriptors and narratives are
6 intended to permit sufficient flexibility to accommodate new scientific
7 understanding and new testing methods as they are developed and accepted by
8 the scientific community and the public. Descriptors represent points along a
9 continuum of evidence; consequently, there are gradations and borderline cases
10 that are clarified by the full narrative. Descriptors, as well as an introductory
11 paragraph, are a short summary of the complete narrative that preserves the
12 complexity that is an essential part of the hazard characterization. Users of
13 these cancer guidelines and of the risk assessments that result from the use
14 of these cancer guidelines should consider the entire range of information
15 included in the narrative rather than focusing simply on the descriptor.
16 In borderline cases, the narrative explains the case for choosing one descriptor
17 and discusses the arguments for considering but not choosing another. For
18 example, between "suggestive" and "likely" or between "suggestive" and
19 "inadequate," the explanation clearly communicates the information needed to
20 consider appropriately the agent's carcinogenic potential in subsequent
21 decisions.
22 Multiple descriptors can be used for a single agent, for example, when
23 carcinogenesis is dose- or route-dependent. For example, if an agent causes
24 point-of-contact tumors by one exposure route but adequate testing is negative
25 by another route, then the agent could be described as likely to be carcinogenic
26 by the first route but not likely to be carcinogenic by the second. Another
27 example is when the mode of action is sufficiently understood to conclude that a
28 key event in tumor development would not occur below a certain dose range. In
29 this case, the agent could be described as likely to be carcinogenic above a
30 certain dose range but not likely to be carcinogenic below that range.
31 Descriptors can be selected for an agent that has not been tested in a cancer
32 bioassay if sufficient other information, e.g., toxicokinetic and mode of action
33 information, is available to make a strong, convincing, and logical case through
34 scientific inference. For example, if an agent is one of a well-defined class of
3 5 agents that are understood to operate through a common mode of action and if
3 6 that agent has the same mode of action, then in the narrative the untested agent
37 would have the same descriptor as the class. Another example is when an
3 8 untested agent's effects are understood to be caused by a human metabolite, in
3 9 which case in the narrative the untested agent could have the same descriptor as
40 the metabolite. As new testing methods are developed and used, assessments
41 may increasingly be based on inferences from toxicokinetic and mode of action
42 information in the absence of tumor studies in animals or humans.
43 When a well-studied agent produces tumors only at a point of initial contact, the
44 descriptor generally applies only to the exposure route producing tumors unless
45 the mode of action is relevant to other routes. The rationale for this conclusion
46 would be explained in the narrative.
47 When tumors occur at a site other than the point of initial contact, the descriptor
48 generally applies to all exposure routes that have not been adequately tested at
49 sufficient doses. An exception occurs when there is convincing information,
50 e.g., toxicokinetic data that absorption does not occur by another route.
March 2008 AX1-18 DRAFT-DO NOT QUOTE OR CITE
-------
1 When the response differs qualitatively as well as quantitatively with dose, this
2 information should be part of the characterization of the hazard. In some cases
3 reaching a certain dose range can be a precondition for effects to occur, as when
4 cancer is secondary to another toxic effect that appears only above a certain
5 dose. In other cases exposure duration can be a precondition for hazard if
6 effects occur only after exposure is sustained for a certain duration. These
7 considerations differ from the issues of relative absorption or potency at
8 different dose levels because they may represent a discontinuity in a dose-
9 response function.
10 When multiple bioassays are inconclusive, mode of action data are likely to hold
11 the key to resolution of the more appropriate descriptor. When bioassays are
12 few, further bioassays to replicate a study' s results or to investigate the potential
13 for effects in another sex, strain, or species may be useful.
14 When there are few pertinent data, the descriptor makes a statement about the
15 database, for example, "Inadequate Information to Assess Carcinogenic
16 Potential," or a database that provides "Suggestive Evidence of Carcinogenic
17 Potential." With more information, the descriptor expresses a conclusion about
18 the agent's carcinogenic potential to humans. If the conclusion is positive, the
19 agent could be described as "Likely to Be Carcinogenic to Humans" or, with
20 strong evidence, "Carcinogenic to Humans." If the conclusion is negative, the
21 agent could be described as "Not Likely to Be Carcinogenic to Humans."
22 Although the term "likely" can have a probabilistic connotation in other
23 contexts, its use as a weight of evidence descriptor does not correspond to a
24 quantifiable probability of whether the chemical is carcinogenic. This is
25 because the data that support cancer assessments generally are not suitable for
26 numerical calculations of the probability that an agent is a carcinogen. Other
27 health agencies have expressed a comparable weight of evidence using terms
28 such as "Reasonably Anticipated to Be a Human Carcinogen" (NTP) or
29 "Probably Carcinogenic to Humans" (International Agency for Research on
30 Cancer).
31 AX1.4.2.3 Improving the Presumptive Disability Decision-Making Process for Veterans
32 A recent publication by the Institute of Medicine (IOM) also provides foundation for the
33 causality framework adapted in this ISA (IOM, 2007). The Committee on Evaluation of the
34 Presumptive Disability Decision-Making Process for Veterans was charged by the Veterans
35 Association to describe how presumptive decisions are made for veterans with health conditions
36 arising from military service currently, as well as recommendations for how such decisions could
37 made in the future. The committee proposes a multiple-element approach that includes a
38 quantification of the extent of disease attributable to an exposure. This process involves a review
39 of all relevant data to decide the strength of evidence for causation, using one of four categories:
40 • Sufficient: the evidence is sufficient to conclude that a causal relationship exists.
41 • Equipoise and Above: the evidence is sufficient to conclude that a causal relationship
42 is at least as likely as not, but not sufficient to conclude that a causal relationship
43 exists.
March 2008 AX1-19 DRAFT-DO NOT QUOTE OR CITE
-------
1 • Below Equipoise: the evidence is not sufficient to conclude that a causal relationship
2 is at least as likely as not, or is not sufficient to make a scientifically informed
3 judgment.
4 • Against: the evidence suggests the lack of a causal relationship.
5
6 The following is an excerpt from this report and describes these four categories in detail:
7 In light of the categorizations used by other health organizations and agencies as
8 well as considering the particular challenges of the presumptive disability
9 decision-making process, we propose a four-level categorization of the strength
10 of the overall evidence for or against a causal relationship from exposure to
11 disease.
12 We use the term "equipoise" to refer to the point at which the evidence is in
13 balance between favoring and not favoring causation. The term "equipoise" is
14 widely used in the biomedical literature, is a concept familiar to those concerned
15 with evidence-based decision-making and is used in VA processes for rating
16 purposes as well as being a familiar term in the veterans' community.
17 Below we elaborate on the four-level categorization which the Committee
18 recommends.
19 Sufficient
20 If the overall evidence for a causal relationship is categorized as Sufficient, then
21 it should be scientifically compelling. It might include:
22 • replicated and consistent evidence of a causal association: that is,
23 evidence of an association from several high-quality epidemiologic
24 studies that cannot be explained by plausible noncausal alternatives
25 (e.g., chance, bias, or confounding)
26 • evidence of causation from animal studies and mechanistic knowledge
27 • compelling evidence from animal studies and strong mechanistic
28 evidence from studies in exposed humans, consistent with (i.e., not
29 contradicted by) the epidemiologic evidence.
3 0 Using the Bayesian framework to illustrate the evidential support and the
31 resulting state of communal scientific opinion needed for reaching the Sufficient
32 category (and the lower categories that follow), consider again the causal
33 diagram in [Figure AX1.4-1]. In this model, used to help clarify matters
34 conceptually, the observed association between exposure and health is the result
35 of: (1) measured confounding, parameterized by a; (2) the causal relation,
36 parameterized by (3; and (3) other, unmeasured sources such as bias or
3 7 unmeasured confounding, parameterized by y. The belief of interest, after all
3 8 the evidence has been weighed, is in the size of the causal parameter (3. Thus,
39 for decision making, what matters is how strongly the evidence supports the
40 proposition that (3 is above 0. As it is extremely unlikely that the types of
41 exposures considered for presumptions reduce the risk of developing disease, we
42 exclude values of (3 below 0. If we consider the evidence as supporting degrees
43 of belief about the size of (3, and we have a posterior distribution over the
44 possible size of (3, then a posterior like [Figure AX1.4-2] illustrates a belief state
45 that might result when the evidence for causation is considered Sufficient.
March 2008 AX1-20 DRAFT-DO NOT QUOTE OR CITE
-------
Measured
Confounders/Covariates
I
Exposure
to Substance
I
Health
Outcome
1
Unmeasured Confounders/Covariates, or
Other Sources of Spurious Association from Bias
Figure AX1.4-1.
Focusing on unmeasured confounders/covariates, or other sources of
spurious association from bias.
Posterior Mass
Over an Effect
Size of the Causal Effect 6
Figure AX1.4-2. Example posterior distribution for the determination of Sufficient.
Source: IOM(2007).
1
2
3
4
5
6
7
8
9
10
11
As the "mass" over a positive effect (the area under the curve to the right of the
zero) vastly "outweighs" the small mass over no effect (zero), the evidence is
considered sufficient to conclude that the association is causal. Put another way,
even though the scientific community might be uncertain as to the size of (3,
after weighing all the evidence, it is highly confident that the probability that (3
is greater than zero is substantial; that is, that exposure causes disease.
Equipoise and Above
To be categorized as Equipoise and Above, the scientific community should
categorize the overall evidence as making it more confident in the existence of a
causal relationship than in the non-existence of a causal relationship, but not
sufficient to conclude causation.
March 2008
AX1-21
DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
For example, if there are several high-quality epidemiologic studies, the
preponderance of which show evidence of an association that cannot readily be
explained by plausible noncausal alternatives (e.g., chance, bias, or
confounding), and the causal relationship is consistent with the animal evidence
and biological knowledge, then the overall evidence might be categorized as
Equipoise and Above. Alternatively, if there is strong evidence from animal
studies or mechanistic evidence, not contradicted by human or other evidence,
then the overall evidence might be categorized as Equipoise and Above.
Equipoise is a common term employed by VA and the courts in deciding
disability claims (see Appendix D).
Again, using the Bayesian model to illustrate the idea of Equipoise and Above,
[Figure AX 1.4-3] shows a posterior probability distribution that is an example of
belief compatible with the category Equipoise and Above.
Posterior Mass
Over an Effect
Size of the Causa! Effect 6
Figure AX1.4-3. Example posterior distribution for the determination of Equipoise and
Above.
Source: IOM(2007).
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
In this figure, unlike the one for evidence classified as Sufficient, there is
considerable mass over zero, which means that the scientific community has
considerable uncertainty as to whether exposure causes disease at all; that is,
whether (3 is greater than zero. At least half of the mass is to the right of the
zero, however, so the community judges causation to be at least as likely as not,
after they have seen and combined all the evidence available.
Below Equipoise
To be categorized as Below Equipoise, the overall evidence for a causal
relationship should either be judged not to make causation at least as likely as
not, or not sufficient to make a scientifically informed judgment.
This might occur:
• when the human evidence is consistent in showing an association, but
the evidence is limited by the inability to rule out chance, bias, or
confounding with confidence, and animal or mechanistic evidence is
weak
March 2008
AX1-22
DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
• when animal evidence suggests a causal relationship, but human and
mechanistic
• evidence is weak or inconsistent
• when mechanistic evidence is suggestive but animal and human
evidence is weak or inconsistent
• when the evidence base is very thin.
Against
To be categorized as Against, the overall evidence should favor belief that there
is no causal relationship from exposure to disease. For example, if there is
human evidence from multiple studies covering the full range of exposures
encountered by humans that are consistent in showing no causal association, or
there are is animal or mechanistic evidence supporting the lack of a causal
relationship, and combining all of the evidence results in a posterior resembling
[Figure AX 1.4-4] then the scientific community should categorize the evidence
as, Against causation.
22
23
24
25
26
27
28
Posterior Over p
Size of the Causal Effect
Figure AX1.4-4. Example posterior distribution for the determination of Against.
Source: IOM(2007).
16 AX1.4.2.4 Guidelines for Formulation of Scientific Findings to be Used for Policy
17 Purposes
18 The following guidelines in the form of checklist questions were developed and
19 published in 1991 by the NAPAP Oversight Review Board for the National Acid Precipitation
20 Assessment Program to assist scientists in formulating presentations of research results to be
21 used in policy decision processes.
1. IS THE STATEMENT SOUND ? Have the central issues been clearly
identified? Does each statement contain the distilled essence of present
scientific and technical understanding of the phenomenon or process to
which it applies? Is the statement consistent with all relevant evidence -
evidence developed either through NAPAP research or through analysis
of research conducted outside of NAPAP? Is the statement contradicted
by any important evidence developed through research inside or outside
March 2008
AX1-23
DRAFT-DO NOT QUOTE OR CITE
-------
1 of NAPAP? Have apparent contradictions or interpretations of available
2 evidence been considered in formulating the statement of principal
3 findings?
4 2 IS THE STATEMENT DIRECTIONAL AND, WHERE
5 APPROPRIATE, QUANTITATIVE? Does the statement correctly
6 quantify both the direction and magnitude of trends and relationships in
7 the phenomenon or process to which the statement is relevant? When
8 possible, is a range of uncertainty given for each quantitative result?
9 Have various sources of uncertainty been identified and quantified, for
10 example, does the statement include or acknowledge errors in actual
11 measurements, standard errors of estimate, possible biases in the
12 availability of data, extrapolation of results beyond the mathematical,
13 geographical, or temporal relevancy of available information, etc. In
14 short, are there numbers in the statement? Are the numbers correct? Are
15 the numbers relevant to the general meaning of the statement?
16 3 IS THE DEGREE OF CERTAINTY OR UNCERTAINTY OF THE
17 STATEMENT INDICATED CLEARLY? Have appropriate statistical
18 tests been applied to the data used in drawing the conclusion set forth in
19 the statement? If the statement is based on a mathematical or novel
20 conceptual model, has the model or concept been validated? Does the
21 statement describe the model or concept on which it is based and the
22 degree of validity of that model or concept?
23 4 IS THE STATEMENT CORRECT WITHOUT QUALIFICATION?
24 Are there limitations of time, space, or other special circumstances in
25 which the statement is true? If the statement is true only in some
26 circumstances, are these limitations described adequately and briefly?
27 5 IS THE STATEMENT CLEAR AND UNAMBIGUOUS? Are the
28 words and phrases used in the statement understandable by the decision
29 makers of our society? Is the statement free of specialized jargon? Will
30 too many people misunderstand its meaning?
31 6 IS THE STATEMENT AS CONCISE AS IT CAN BE MADE
32 WITHOUT RISK OF MISUNDERSTANDING? Are there any excess
3 3 words, phrases, or ideas in the statement which are not necessary to
34 communicate the meaning of the statement? Are there so many caveats
35 in the statement that the statement itself is trivial, confusing, or
36 ambiguous?
37 7 IS THE STATEMENT FREE OF SCIENTIFIC OR OTHER
3 8 BIASES OR IMPLICATIONS OF SOCIETAL VALUE
3 9 JUDGMENTS? Is the statement free of influence by specific schools of
40 scientific thought? Is the statement also free of words, phrases, or
41 concepts that have political, economic, ideological, religious, moral, or
42 other personal-, agency-, or organization-specific values, overtones, or
43 implications? Does the choice of how the statement is expressed rather
44 than its specific words suggest underlying biases or value judgments? Is
45 the tone impartial and free of special pleading? If societal value
46 judgments have been discussed, have these judgments been identified as
47 such and described both clearly and objectively?
48 8 HAVE SOCIETAL IMPLICATIONS BEEN DESCRIBED
49 OBJECTIVELY? Consideration of alternative courses of action and
March 2008 AX1-24 DRAFT-DO NOT QUOTE OR CITE
-------
1 their consequences inherently involves judgments of their feasibility and
2 the importance of effects. For this reason, it is important to ask if a
3 reasonable range of alternative policies or courses of action have been
4 evaluated? Have societal implications of alternative courses of action
5 been stated in the following general form?:
6 9. "If this [particular option] were adopted then that [particular outcome]
7 would be expected."
8 10 HAVE THE PROFESSIONAL BIASES OF AUTHORS AND
9 REVIEWERS BEEN DESCRIBED OPENLY? Acknowledgment of
10 potential sources of bias is important so that readers can judge for
11 themselves the credibility of reports and assessments.
12
13 AX1.4.2.5 International Agency for Research on Cancer Guidelines for Scientific Review
14 and Evaluation
15 The following is excerpted from the International Agency for Research on Cancer
16 (IARC) Monographs on the evaluation of carcinogenic risks to humans (LARC, 2006).
17 The available studies are summarized by the Working Group, with particular
18 regard to the qualitative aspects discussed below. In general, numerical findings
19 are indicated as they appear in the original report; units are converted when
20 necessary for easier comparison. The Working Group may conduct additional
21 analyses of the published data and use them in their assessment of the evidence;
22 the results of such supplementary analyses are given in square brackets. When
23 an important aspect of a study that directly impinges on its interpretation should
24 be brought to the attention of the reader, a Working Group comment is given in
25 square brackets.
26 The scope of the IARC Monographs programme has expanded beyond
27 chemicals to include complex mixtures, occupational exposures, physical and
28 biological agents, lifestyle factors and other potentially carcinogenic exposures.
29 Over time, the structure of a Monograph has evolved to include the following
30 sections:
31 1. Exposure data
32 2. Studies of cancer in humans
33 3. Studies of cancer in experimental animals
34 4. Mechanistic and other relevant data
35 5. Summary
36 6. Evaluation and rationale
37 In addition, a section of General Remarks at the front of the volume discusses
3 8 the reasons the agents were scheduled for evaluation and some key issues the
39 Working Group encountered during the meeting.
March 2008 AX1-25 DRAFT-DO NOT QUOTE OR CITE
-------
1 This part of the Preamble discusses the types of evidence considered and
2 summarized in each section of a Monograph, followed by the scientific criteria
3 that guide the evaluations.
4 1. Exposure data
5 Each Monograph includes general information on the agent: this information
6 may vary substantially between agents and must be adapted accordingly. Also
7 included is information on production and use (when appropriate), methods of
8 analysis and detection, occurrence, and sources and routes of human
9 occupational and environmental exposures. Depending on the agent, regulations
10 and guidelines for use may be presented.
11 (a) General information on the agent
12 For chemical agents, sections on chemical and physical data are included: the
13 Chemical Abstracts Service Registry Number, the latest primary name and the
14 IUPAC systematic name are recorded; other synonyms are given, but the list is
15 not necessarily comprehensive. Information on chemical and physical
16 properties that are relevant to identification, occurrence and biological activity is
17 included. A description of technical products of chemicals includes trade
18 names, relevant specifications and available information on composition and
19 impurities. Some of the trade names given may be those of mixtures in which
20 the agent being evaluated is only one of the ingredients.
21 For biological agents, taxonomy, structure and biology are described, and the
22 degree of variability is indicated. Mode of replication, life cycle, target cells,
23 persistence, latency, host response and clinical disease other than cancer are also
24 presented.
25 For physical agents that are forms of radiation, energy and range of the radiation
26 are included. For foreign bodies, fibres and respirable particles, size range and
27 relative dimensions are indicated.
28 For agents such as mixtures, drugs or lifestyle factors, a description of the agent,
29 including its composition, is given.
30 Whenever appropriate, other information, such as historical perspectives or the
31 description of an industry or habit, may be included.
32 (b) Analysis and detection
33 An overview of methods of analysis and detection of the agent is presented,
34 including their sensitivity, specificity and reproducibility. Methods widely used
3 5 for regulatory purposes are emphasized. Methods for monitoring human
3 6 exposure are also given. No critical evaluation or recommendation of any
3 7 method is meant or implied.
3 8 (c) Production and use
3 9 The dates of first synthesis and of first commercial production of a chemical,
40 mixture or other agent are provided when available; for agents that do not occur
41 naturally, this information may allow a reasonable estimate to be made of the
42 date before which no human exposure to the agent could have occurred. The
43 dates of first reported occurrence of an exposure are also provided when
March 2008 AX1-26 DRAFT-DO NOT QUOTE OR CITE
-------
1 available. In addition, methods of synthesis used in past and present commercial
2 production and different methods of production, which may give rise to different
3 impurities, are described.
4 The countries where companies report production of the agent, and the number
5 of companies in each country, are identified. Available data on production,
6 international trade and uses are obtained for representative regions. It should
7 not, however, be inferred that thoseareas or nations are necessarily the sole or
8 major sources or users of the agent. Some identified uses may not be current or
9 major applications, and the coverage is not necessarily comprehensive. In the
10 case of drugs, mention of their therapeutic uses does not necessarily represent
11 current practice nor does it imply judgement as to their therapeutic efficacy.
12 (d) Occurrence and exposure
13 Information on the occurrence of an agent in the environment is obtained from
14 data derived from the monitoring and surveillance of levels in occupational
15 environments, air, water, soil, plants, foods and animal and human tissues.
16 When available, data on the generation, persistence and bioaccumulation of the
17 agent are also included. Such data may be available from national databases.
18 Data that indicate the extent of past and present human exposure, the sources of
19 exposure, the people most likely to be exposed and the factors that contribute to
20 the exposure are reported. Information is presented on the range of human
21 exposure, including occupational and environmental exposures. This includes
22 relevant findings from both developed and developing countries. Some of these
23 data are not distributed widely and may be available from government reports
24 and other sources. In the case of mixtures, industries, occupations or processes,
25 information is given about all agents known to be present. For processes,
26 industries and occupations, a historical description is also given, noting
27 variations in chemical composition, physical properties and levels of
28 occupational exposure with date and place. For biological agents, the
29 epidemiology of infection is described.
30 (e) Regulations and guidelines
31 Statements concerning regulations and guidelines (e.g. occupational exposure
32 limits, maximal levels permitted in foods and water, pesticide registrations) are
33 included, but they may not reflect the most recent situation, since such limits are
34 continuously reviewed and modified. The absence of information on regulatory
3 5 status for a country should not be taken to imply that that country does not have
3 6 regulations with regard to the exposure. For biological agents, legislation and
3 7 control, including vaccination and therapy, are described.
38 2. Studies of cancer in humans
39 This section includes all pertinent epidemiological studies (see Part A, Section
40 4). Studies of biomarkers are included when they are relevant to an evaluation
41 of carcinogenicity to humans.
42 (a) Types of study considered
43 Several types of epidemiological study contribute to the assessment of
44 carcinogenicity in humans — cohort studies, case-control studies, correlation
45 (or ecological) studies and intervention studies. Rarely, results from randomized
March 2008 AX1-27 DRAFT-DO NOT QUOTE OR CITE
-------
1 trials may be available. Case reports and case series of cancer in humans may
2 also be reviewed.
3 Cohort and case-control studies relate individual exposures under study to the
4 occurrence of cancer in individuals and provide an estimate of effect (such as
5 relative risk) as the main measure of association. Intervention studies may
6 provide strong evidence for making causal inferences, as exemplified by
7 cessation of smoking and the subsequent decrease in risk for lung cancer.
8 In correlation studies, the units of investigation are usually whole populations
9 (e.g. in particular geographical areas or at particular times), and cancer
10 frequency is related to asummary measure of the exposure of the population to
11 the agent under study. In correlation studies, individual exposure is not
12 documented, which renders this kind of study more prone to confounding. In
13 some circumstances, however, correlation studies may be more informative than
14 analytical study designs (see, for example, the Monograph on arsenic in
15 drinking-water; IARC, 2004).
16 In some instances, case reports and case series have provided important
17 information about the carcinogenicity of an agent. These types of study
18 generally arise from a suspicion, based on clinical experience, that the
19 concurrence of two events — that is, a particular exposure and occurrence of a
20 cancer — has happened rather more frequently than would be expected by
21 chance. Case reports and case series usually lack complete ascertainment of
22 cases in any population, definition or enumeration of the population at risk and
23 estimation of the expected number of cases in the absence of exposure.
24 The uncertainties that surround the interpretation of case reports, case series and
25 correlation studies make them inadequate, except in rare instances, to form the
26 sole basis for inferring a causal relationship. When taken together with case-
27 control and cohort studies, however, these types of study may add materially to
28 the judgement that a causal relationship exists.
29 Epidemiological studies of benign neoplasms, presumed preneoplastic lesions
30 and other end-points thought to be relevant to cancer are also reviewed. They
31 may, in some instances, strengthen inferences drawn from studies of cancer
32 itself.
3 3 (b) Quality of studies considered
34 It is necessary to take into account the possible roles of bias, confounding and
3 5 chance in the interpretation of epidemiological studies. Bias is the effect of
3 6 factors in study design or execution that lead erroneously to a stronger or weaker
3 7 association than in fact exists between an agent and disease. Confounding is a
3 8 form of bias that occurs when the relationship with disease is made to appear
3 9 stronger or weaker than it truly is as a result of an association between the
40 apparent causal factor and another factor that is associated with either an
41 increase or decrease in the incidence of the disease. The role of chance is
42 related to biological variability and the influence of sample size on the precision
43 of estimates of effect.
44 In evaluating the extent to which these factors have been minimized in an
45 individual study, consideration is given to a number of aspects of design and
46 analysis as described in the report of the study. For example, when suspicion of
47 carcinogenicity arises largely from a single small study, careful consideration is
48 given when interpreting subsequent studies that included these data in an
March 2008 AX1-28 DRAFT-DO NOT QUOTE OR CITE
-------
1 enlarged population. Most of these considerations apply equally to case-
2 control, cohort and correlation studies. Lack of clarity of any of these aspects in
3 the reporting of a study can decrease its credibility and the weight given to it in
4 the final evaluation of the exposure.
5 Firstly, the study population, disease (or diseases) and exposure should have
6 been well defined by the authors. Cases of disease in the study population
7 should have been identified in a way that was independent of the exposure of
8 interest, and exposure should have been assessed in a way that was not related to
9 disease status.
10 Secondly, the authors should have taken into account — in the study design and
11 analysis — other variables that can influence the risk of disease and may have
12 been related to the exposure of interest. Potential confounding by such variables
13 should have been dealt with either in the design of the study, such as by
14 matching, or in the analysis, by statistical adjustment. In cohort studies,
15 comparisons with local rates of disease may or may not bemore appropriate than
16 those with national rates. Internal comparisons of frequency of disease among
17 individuals at different levels of exposure are also desirable in cohort studies,
18 since they minimize the potential for confounding related to the difference in
19 risk factors between an external reference group and the study population.
20 Thirdly, the authors should have reported the basic data on which the
21 conclusions are founded, even if sophisticated statistical analyses were
22 employed. At the very least, they should have given the numbers of exposed
23 and unexposed cases and controls in a case-control study and the numbers of
24 cases observed and expected in a cohort study. Further tabulations by time since
25 exposure began and other temporal factors are also important. In a cohort study,
26 data on all cancer sites and all causes of death should have been given, to reveal
27 the possibility of reporting bias. In a case-control study, the effects of
28 investigated factors other than the exposure of interest should have been
29 reported.
30 Finally, the statistical methods used to obtain estimates of relative risk, absolute
31 rates of cancer, confidence intervals and significance tests, and to adjust for
32 confounding should have been clearly stated by the authors. These methods
33 have been reviewed for case-control studies (Breslow & Day, 1980) and for
34 cohort studies (Breslow & Day, 1987).
3 5 (c) Meta-analyses and pooled analyses
36 Independent epidemiological studies of the same agent may lead to results that
37 are difficult to interpret. Combined analyses of data from multiple studies are a
3 8 means of resolving this ambiguity, and well-conducted analyses can be
39 considered. There are two types of combined analysis. The first involves
40 combining summary statistics such as relative risks from individual studies
41 (meta-analy sis) and the second involves a pooled analysis of the raw data from
42 the individual studies (pooled analysis) (Greenland, 1998).
43 The advantages of combined analyses are increased precision due to increased
44 sample size and the opportunity to explore potential confounders, interactions
45 and modifying effects that may explain heterogeneity among studies in more
46 detail. A disadvantage of combined analyses is the possible lack of compatibility
47 of data from various studies due to differences in subject recruitment,
48 procedures of data collection, methods of measurement and effects of
49 unmeasured co-variates that may differ among studies. Despite these
March 2008 AX1-29 DRAFT-DO NOT QUOTE OR CITE
-------
1 limitations, well-conducted combined analyses may provide a firmer basis than
2 individual studies for drawing conclusions about the potential carcinogenicity of
3 agents.
4 IARC may commission a meta-analysis or pooled analysis that is pertinent to a
5 particular Monograph (see Part A, Section 4). Additionally, as a means of
6 gaining insight from the results of multiple individual studies, ad-hoc
7 calculations that combine data from different studies may be conducted by the
8 Working Group during the course of a Monograph meeting. The results of such
9 original calculations, which would be specified in the text by presentation in
10 square brackets, might involve updates of previously conducted analyses that
11 incorporate the results of more recent studies or de-novo analyses. Irrespective
12 of the source of data for the meta-analyses and pooled analyses, it is important
13 that the same criteria for data quality be applied as those that would be applied
14 to individual studies and to ensure also that sources of heterogeneity between
15 studies be taken into account.
16 (d) Temporal effects
17 Detailed analyses of both relative and absolute risks in relation to temporal
18 variables, such as age at first exposure, time since first exposure, duration of
19 exposure, cumulative exposure, peak exposure (when appropriate) and time
20 since cessation of exposure, are reviewed and summarized when available.
21 Analyses of temporal relationships may be usefulin making causal inferences.
22 In addition, such analyses may suggest whether a carcinogen acts early or late in
23 the process of carcinogenesis, although, at best, they allow only indirect
24 inferences about mechanisms of carcinogenesis.
25 (e) Use of biomarkers in epidemiological studies
26 Biomarkers indicate molecular, cellular or other biological changes and are
27 increasingly used in epidemiological studies for various purposes (IARC, 1991;
28 Vainio et al., 1992; Toniolo et al., 1997; Vineis et al., 1999; Buffler et al.,
29 2004). These may include evidence of exposure, of early effects, of cellular,
3 0 tissue or organism responses, of individual susceptibility or host responses, and
31 inference of a mechanism (see Part B, Section 4b). This is a rapidly evolving
32 field that encompasses developments in genomics, epigenomics and other
33 emerging technologies.
34 Molecular epidemiological data that identify associations between genetic
3 5 polymorphisms and interindividual differences in susceptibility to the agent(s)
3 6 being evaluated may contribute to the identification of carcinogenic hazards to
3 7 humans. If the polymorphism has been demonstrated experimentally to modify
38 the functional activity of the gene product in a manner that is consistent with
39 increased susceptibility, these data may be useful in making causal inferences.
40 Similarly, molecular epidemiological studies that measure cell functions,
41 enzymes or metabolites that are thought to be the basis of susceptibility may
42 provide evidence that reinforces biological plausibility. It should be noted,
43 however, that when data on genetic susceptibility originate from multiple
44 comparisons that arise from subgroup analyses, this can generate false-positive
45 results and inconsistencies across studies, and such data therefore require careful
46 evaluation. If the known phenotype of a genetic polymorphism can explain the
47 carcinogenic mechanism of the agent being evaluated, data on this phenotype
48 may be useful in making causal inferences.
March 2008 AX1-30 DRAFT-DO NOT QUOTE OR CITE
-------
1 (f) Criteria for causality
2 After the quality of individual epidemiological studies of cancer has been
3 summarized and assessed, a judgement is made concerning the strength of
4 evidence that the agent in question is carcinogenic to humans. In making its
5 judgement, the Working Group considers several criteria for causality (Hill,
6 1965). A strong association (e.g. a large relative risk) is more likely to indicate
7 causality than a weak association, although it is recognized that estimates of
8 effect of small magnitude do not imply lack of causality and may be important if
9 the disease or exposure is common. Associations that are replicated in several
10 studies of the same design or that use different epidemiological approaches or
11 under different circumstances of exposure are more likely to represent a causal
12 relationship than isolated observations from single studies. If there are
13 inconsistent results among investigations, possible reasons are sought (such as
14 differences in exposure), and results of studies that are judged to be of high
15 quality are given more weight than those of studies that are judged to be
16 methodologically less sound.
17 If the risk increases with the exposure, this is considered to be a strong
18 indication of causality, although the absence of a graded response is not
19 necessarily evidence against a causal relationship. The demonstration of a
20 decline in risk after cessation of or reduction in exposure in individuals or in
21 whole populations also supports a causal interpretation of the findings.
22 A number of scenarios may increase confidence in a causal relationship. On the
23 one hand, an agent may be specific in causing tumours at one site or of one
24 morphological type. On the other, carcinogenicity may be evident through the
25 causation of multiple tumour typesTemporality, precision of estimates of effect,
26 biological plausibility and coherence of the overall database are considered.
27 Data on biomarkers may be employed in an assessment of the biological
28 plausibility of epidemiological observations.
29 Although rarely available, results from randomized trials that show different
3 0 rates of cancer among exposed and unexposed individuals provide particularly
31 strong evidence for causality.
32 When several epidemiological studies show little or no indication of an
33 association between an exposure and cancer, a judgement may be made that, in
34 the aggregate, they show evidence of lack of carcinogenicity. Such a judgement
3 5 requires firstly that the studies meet, to a sufficient degree, the standards of
36 design and analysis described above. Specifically, the possibility that bias,
3 7 confounding or misclassification of exposure or outcome could explain the
3 8 observed results should be considered and excluded with reasonable certainty.
39 In addition, all studies that are judged to be methodologically sound should (a)
40 be consistent with an estimate of effect of unity for any observed level of
41 exposure, (b) when considered together, provide a pooled estimate of relative
42 risk that is at or near to unity, and (c) have a narrow confidence interval, due to
43 sufficient population size. Moreover, no individual study nor the pooled results
44 of all the studies should show any consistent tendency that the relative risk of
45 cancer increases with increasing level of exposure. It is important to note that
46 evidence of lack of carcinogenicity obtained from several epidemiological
47 studies can apply only to the type(s) of cancer studied, to the dose levels
48 reported, and to the intervals between first exposure and disease onset observed
49 in these studies. Experience with human cancer indicates that the period from
50 first exposure to the development of clinical cancer is sometimes longer than 20
March 2008 AX1-31 DRAFT-DO NOT QUOTE OR CITE
-------
1 years; latent periods substantially shorter than 30 years cannot provide evidence
2 for lack of carcinogenicity.
3 3. Studies of cancer in experimental animals
4 All known human carcinogens that have been studied adequately for
5 carcinogenicity in experimental animals have produced positive results in one or
6 more animal species (Wilbourn et al., 1986; Tomatis et al., 1989). For several
7 agents (e.g. aflatoxins, diethylstilbestrol, solar radiation, vinyl chloride),
8 carcinogenicity in experimental animals was established or highly suspected
9 before epidemiological studies confirmed their carcinogenicity in humans
10 (Vainio et al., 1995). Although this association cannot establish that all agents
11 that cause cancer in experimental animals also cause cancer in humans, it is
12 biologically plausible that agents for which there is sufficient evidence of
13 carcinogenicity in experimental animals (see Part B, Section 6b) also present a
14 carcinogenic hazard to humans. Accordingly, in the absence of additional
15 scientific information, these agents are considered to pose a carcinogenic hazard
16 to humans. Examples of additional scientific information are data that
17 demonstrate that a given agent causes cancer in animals through a species-
18 specific mechanism that does not operate in humans or data that demonstrate
19 that the mechanism in experimental animals also operates in humans (see Part B,
20 Section 6).
21 Consideration is given to all available long-term studies of cancer in
22 experimental animals with the agent under review (see Part A, Section 4). In all
23 experimental settings, the nature and extent of impurities or contaminants
24 present in the agent being evaluated are given when available. Animal species,
25 strain (including genetic background where applicable), sex, numbers per group,
26 age at start of treatment, route of exposure, dose levels, duration of exposure,
27 survival and information on tumours (incidence, latency, severity or multiplicity
28 of neoplasms or preneoplastic lesions) are reported. Those studies in
29 experimental animals that are judged to be irrelevant to the evaluation or judged
30 to be inadequate (e.g. too short aduration, too few animals, poor survival; see
31 below) may be omitted. Guidelines for conducting long-term carcinogenicity
32 experiments have been published (e.g. OECD, 2002).
3 3 Other studies considered may include: experiments in which the agent was
34 administered in the presence of factors that modify carcinogenic effects (e.g.
3 5 initiation-promotion studies, co-carcinogenicity studies and studies in
3 6 genetically modified animals); studies in which the end-point was not cancer but
37 a defined precancerous lesion; experiments on the carcinogenicity of known
3 8 metabolites and derivatives; and studies of cancer in non-laboratory animals
39 (e.g. livestock and companion animals) exposed to the agent.
40 For studies of mixtures, consideration is given to the possibility that changes in
41 the phy sicochemical properties of the individual substances may occur during
42 collection, storage, extraction, concentration and delivery. Another
43 consideration is that chemical and toxicological interactions of components in a
44 mixture may alter dose-response relationships. The relevance to human
45 exposure of the test mixture administered in the animal experiment is also
46 assessed. This may involve consideration of the following aspects of the
47 mixture tested: (i) physical and chemical characteristics, (ii) identified
48 constituents that may indicate the presence of a class of substances and (iii) the
49 results of genetic toxicity and related tests.
March 2008 AX1-32 DRAFT-DO NOT QUOTE OR CITE
-------
1 The relevance of results obtained with an agent that is analogous (e.g. similar in
2 structure or of a similar virus genus) to that being evaluated is also considered.
3 Such results may provide biological and mechanistic information that is relevant
4 to the understanding of the process of carcinogenesis in humans and may
5 strengthen the biological plausibility that the agent being evaluated is
6 carcinogenic to humans (see Part B, Section 2f).
7 (a) Qualitative aspects
8 An assessment of carcinogenicity involves several considerations of qualitative
9 importance, including (i) the experimental conditions under which the test was
10 performed, including route, schedule and duration of exposure, species, strain
11 (including genetic background where applicable), sex, age and duration of
12 follow-up; (ii) the consistency of the results, for example, across species and
13 target organ(s); (iii) the spectrum of neoplastic response, from preneoplastic
14 lesions and benign tumours to malignant neoplasms; and (iv) the possible role of
15 modifying factors.
16 Considerations of importance in the interpretation and evaluation of a particular
17 study include: (i) how clearly the agent was defined and, in the case of mixtures,
18 how adequately the sample characterization was reported; (ii) whether the dose
19 was monitored adequately, particularly in inhalation experiments; (iii) whether
20 the doses, duration of treatment and route of exposure were appropriate; (iv)
21 whether the survival of treated animals was similar to that of controls; (v)
22 whether there were adequate numbers of animals per group; (vi) whether both
23 male and female animals were used; (vii) whether animals were allocated
24 randomly to groups; (viii) whether the duration of observation was adequate;
25 and (ix) whether the data were reported and analysed adequately.
26 When benign tumours (a) occur together with and originate from the same cell
27 type as malignant tumours in an organ or tissue in a particular study and (b)
28 appear to represent a stage in the progression to malignancy, they are usually
29 combined in the assessment of tumour incidence (Huff., 1989). The occurrence
30 of lesions presumed to be preneoplastic may in certain instances aid in assessing
31 the biological plausibility of any neoplastic response observed. If an agent
32 induces only benign neoplasms that appear to be end-points that do not readily
33 undergo transition to malignancy, the agent should nevertheless be suspected of
34 being carcinogenic and requires further investigation.
3 5 (b) Quantitative aspects
36 The probability that tumours will occur may depend on the species, sex, strain,
3 7 genetic background and age of the animal, and on the dose, route, timing and
3 8 duration of the exposure. Evidence of an increased incidence of neoplasms with
3 9 increasing levels of exposure strengthens the inference of a causal association
40 between the exposure and the development of neoplasms.
41 The form of the dose-response relationship can vary widely, depending on the
42 particular agent under study and the target organ. Mechanisms such as
43 induction of DNA damage or inhibition of repair, altered cell division and cell
44 death rates and changes in intercellular communication are important
45 determinants of dose-response relationships for some carcinogens. Since many
46 chemicals require metabolic activation before being converted to their reactive
47 intermediates, both metabolic and toxicokinetic aspects are important in
48 determining the dose-response pattern. Saturation of steps such as absorption,
49 activation, inactivation and elimination may produce non-linearity in the dose-
March 2008 AX1-33 DRAFT-DO NOT QUOTE OR CITE
-------
1 response relationship (Hoel et al., 1983; Gart et al., 1986), as could saturation of
2 processes such as DNA repair. The dose-response relationship can also be
3 affected by differences in survival among the treatment groups.
4 (c) Statistical analyses
5 Factors considered include the adequacy of the information given for each
6 treatment group: (i) number of animals studied and number examined
7 histologically, (ii) number of animals with a given tumour type and (iii) length
8 of survival. The statistical methods used should be clearly stated and should be
9 the generally accepted techniques refined for this purpose (Peto et al., 1980;
10 Gart et al, 1986; Portier & Bailer, 1989; Bieler & Williams, 1993). The choice
11 of the most appropriate statistical method requires consideration of whether or
12 not there are differences in survival among the treatment groups; for example,
13 reduced survival because of non-tumour-related mortality can preclude the
14 occurrence of tumours later in life. When detailed information on survival is not
15 available, comparisons of the proportions of tumour-bearing animals among the
16 effective number of animals (alive at the time the first tumour was discovered)
17 can be useful when significant differences in survival occur before tumours
18 appear. The lethality of the tumour also requires consideration: for rapidly fatal
19 tumours, the time of death provides an indication of the time of tumour onset
20 and can be assessed using life-table methods; non-fatal or incidental tumours
21 that do not affect survival can be assessed using methods such as the Mantel-
22 Haenzel test for changes in tumour prevalence. Because tumour lethality is
23 often difficult to determine, methods such as the Poly-K test that do not require
24 such information can also be used. When results are available on the number
25 and size of tumours seen in experimental animals (e.g. papillomas on mouse
26 skin, liver tumours observed through nuclear magnetic resonance tomography),
27 other more complicated statistical procedures may be needed (Sherman et al.,
28 1994; Dunson et al, 2003).
29 Formal statistical methods have been developed to incorporate historical control
3 0 data into the analysis of data from a given experiment. These methods assign an
31 appropriate weight to historical and concurrent controls on the basis of the
32 extent of between-study and within-study variability: less weight is given to
33 historical controls when they show a high degree of variability, and greater
34 weight when they show little variability. It is generally not appropriate to
3 5 discount a tumour response that is significantly increased compared with
3 6 concurrent controls by arguing that it falls within the range of historical controls,
3 7 particularly when historical controls show high between-study variability and
3 8 are, thus, of little relevance to the current experiment. In analysing results for
39 uncommon tumours, however, the analysis may be improved by considering
40 historical control data, particularly when between-study variability is low.
41 Historical controls should be selected to resemble the concurrent controls as
42 closely as possible with respect to species, gender and strain, as well as other
43 factors such as basal diet and general laboratory environment, which may affect
44 tumour-response rates in control animals (Haseman et al, 1984; Fung et al,
45 1996;Greime/a/., 2003).
46 Although meta-analyses and combined analyses are conducted less frequently
47 for animal experiments than for epidemiological studies due to differences in
48 animal strains, they can be useful aids in interpreting animal data when the
49 experimental protocols are sufficiently similar.
March 2008 AX1-34 DRAFT-DO NOT QUOTE OR CITE
-------
1 4. Mechanistic and other relevant data
2 Mechanistic and other relevant data may provide evidence of carcinogenicity
3 and also help in assessing the relevance and importance of findings of cancer in
4 animals and in humans. The nature of the mechanistic and other relevant data
5 depends on the biological activity of the agent being considered. The Working
6 Group considers representative studies to give a concise description of the
7 relevant data and issues that they consider to be important; thus, not every
8 available study is cited. Relevant topics may include toxicokinetics,
9 mechanisms of carcinogenesis, susceptible individuals, populations and life-
10 stages, other relevant data and other adverse effects. When data on biomarkers
11 are informative about the mechanisms of carcinogenesis, they are included in
12 this section.
13 These topics are not mutually exclusive; thus, the same studies may be discussed
14 in more than one subsection. For example, a mutation in a gene that codes for
15 an enzyme that metabolizes the agent under study could be discussed in the
16 subsections on toxicokinetics, mechanisms and individual susceptibility if it also
17 exists as an inherited polymorphism.
18 (a) Toxicokinetic data
19 Toxicokinetics refers to the absorption, distribution, metabolism and elimination
20 of agents in humans, experimental animals and, where relevant, cellular systems.
21 Examples of kinetic factors that may affect dose-response relationships include
22 uptake, deposition, biopersistence and half-life in tissues, protein binding,
23 metabolic activation and detoxification. Studies that indicate the metabolic fate
24 of the agent in humans and in experimental animals are summarized briefly, and
25 comparisons of data from humans and animals are made when possible.
26 Comparative information on the relationship between exposure and the dose that
27 reaches the target site may be important for the extrapolation of hazards between
28 species and in clarifying the role of in-vitro findings.
29 (b) Data on mechanisms of carcinogenesis
30 To provide focus, the Working Group attempts to identify the possible
31 mechanisms by which the agent may increase the risk of cancer. For each
32 possible mechanism, a representative selection of key data from humans and
33 experimental systems is summarized. Attention is given to gaps in the data and
34 to data that suggests that more than one mechanism may be operating. The
3 5 relevance of the mechanism to humans is discussed, in particular, when
3 6 mechanistic data are derived from experimental model systems. Changes in the
37 affected organs, tissues or cells can be divided into three non-exclusive levels as
3 8 described below, (i) Changes in physiology
3 9 Physiological changes refer to exposure-related modifications to the physiology
40 and/or response of cells, tissues and organs. Examples of potentially adverse
41 physiological changes include mitogenesis, compensatory cell division, escape
42 from apoptosis and/or senescence, presence of inflammation, hyperplasia,
43 metaplasia and/or preneoplasia, angiogenesis, alterations in cellular adhesion,
44 changes in steroidal hormones and changes in immune surveillance.
March 2008 AX1-35 DRAFT-DO NOT QUOTE OR CITE
-------
1 (ii) Functional changes at the cellular level
2 Functional changes refer to exposure-related alterations in the signalling
3 pathways used by cells to manage critical processes that are related to increased
4 risk for cancer. Examples of functional changes include modified activities of
5 enzymes involved in the metabolism of xenobiotics, alterations in the expression
6 of key genes that regulate DNA repair, alterations in cyclin-dependent kinases
7 that govern cell cycle progression, changes in the patterns of post-translational
8 modifications of proteins, changes in regulatory factors that alter apoptotic rates,
9 changes in the secretion of factors related to the stimulation of DNA replication
10 and transcription and changes in gap-junction-mediated intercellular
11 communication.
12 (iii) Changes at the molecular level
13 Molecular changes refer to exposure-related changes in key cellular structures at
14 the molecular level, including, in particular, genotoxicity. Examples of
15 molecular changes include formation of DNA adducts and DNA strand breaks,
16 mutations in genes, chromosomal aberrations, aneuploidy and changes in DNA
17 methylation patterns. Greater emphasis is given to irreversible effects.
18 The use of mechanistic data in the identification of a carcinogenic hazard is
19 specific to the mechanism being addressed and is not readily described for every
20 possible level and mechanism discussed above.
21 Genotoxicity data are discussed here to illustrate the key issues involved in the
22 evaluation of mechanistic data.
23 Tests for genetic and related effects are described in view of the relevance of
24 gene mutation and chromosomal aberration/aneuploidy to carcinogenesis
25 (Vainio et al., 1992; McGregor et al., 1999). The adequacy of the reporting of
26 sample characterization is considered and, when necessary, commented upon;
27 with regard to complex mixtures, such comments are similar to those described
28 for animal carcinogenicity tests. The available data are interpreted critically
29 according to the end-points detected, which may include DNA damage, gene
30 mutation, sister chromatid exchange, micronucleus formation, chromosomal
31 aberrations and aneuploidy. The concentrations employed are given, and
3 2 mention is made of whether the use of an exogenous metabolic system in vitro
3 3 affected the test result. These data are listed in tabular form by phylogenetic
34 classification.
3 5 Positive results in tests using prokaryotes, lower eukaryotes, insects, plants and
3 6 cultured mammalian cells suggest that genetic and related effects could occur in
37 mammals. Results from such tests may also give information on the types of
3 8 genetic effect produced and on the involvement of metabolic activation. Some
39 end-points described are clearly genetic in nature (e.g. gene mutations), while
40 others are associated with genetic effects (e.g. unscheduled DNA synthesis). In-
41 vitro tests formay be sensitive to changes that are not necessarily the result of
42 genetic alterations but that may have specific relevance to the process of
43 carcinogenesis. Critical appraisals of these tests have been published
44 (Montesano et al., 1986; McGregor et al., 1999).
45 Genetic or other activity manifest in humans and experimental mammals is
46 regarded to be of greater relevance than that in other organisms. The
47 demonstration that an agent can induce gene and chromosomal mutations in
48 mammals in vivo indicates that it may have carcinogenic activity. Negative
March 2008 AX1-36 DRAFT-DO NOT QUOTE OR CITE
-------
1 results in tests for mutagenicity in selected tissues from animals treated in vivo
2 provide less weight, partly because they do not exclude the possibility of an
3 effect in tissues other than those examined. Moreover, negative results in short-
4 term tests with genetic end-points cannot be considered to provide evidence that
5 rules out the carcinogenicity of agents that act through other mechanisms (e.g.
6 receptor-mediated effects, cellular toxicity with regenerative cell division,
7 peroxisome proliferation) (Vainio et al., 1992). Factors that may give
8 misleading results in short-term tests have been discussed in detail elsewhere
9 (Montesano et al, 1986; McGregor et al, 1999).
10 When there is evidence that an agent acts by a specific mechanism that does not
11 involve genotoxicity (e.g. hormonal dysregulation, immune suppression, and
12 formation of calculi and other deposits that cause chronic irritation), that
13 evidence is presented and reviewed critically in the context of rigorous criteria
14 forthe operation of that mechanism in carcinogenesis (e.g. Capenet al., 1999).
15 For biological agents such as viruses, bacteria and parasites, other data relevant
16 to carcinogenicity may include descriptions of the pathology of infection,
17 integration and expression of viruses, and genetic alterations seen in human
18 tumours. Other observations that might comprise cellular and tissue responses
19 to infection, immune response and the presence of tumour markers are also
20 considered.
21 For physical agents that are forms of radiation, other data relevant to
22 carcinogenicity may include descriptions of damaging effects at the
23 physiological, cellular and molecular level, as for chemical agents, and
24 descriptions of how these effects occur. 'Physical agents' may also be
25 considered to comprise foreign bodies, such as surgical implants of various
26 kinds, and poorly soluble fibres, dusts and particles of various sizes, the
27 pathogenic effects of which are a result of their physical presence in tissues or
28 body cavities. Other relevant data for such materials may include
29 characterization of cellular, tissue and physiological reactions to these materials
3 0 and descriptions of pathological conditions other than neoplasia with which they
31 may be associated.
32 (c) Other data relevant to mechanisms
33 A description is provided of any structure-activity relationships that may be
34 relevant to an evaluation of the carcinogenicity of an agent, the lexicological
3 5 implications of the physical and chemical properties, and any other data relevant
36 to the evaluation that are not included elsewhere.
3 7 High-output data, such as those derived from gene expression microarrays, and
3 8 high-throughput data, such as those that result from testing hundreds of agents
39 for a single end-point, pose a unique problem for the use of mechanistic data in
40 the evaluation of a carcinogenic hazard. In the case of high-output data, there is
41 the possibility to overinterpret changes in individual end-points (e.g. changes in
42 expression in one gene) without considering the consistency of that finding in
43 the broader context of the other end-points (e.g. other geneswith linked
44 transcriptional control). High-output data can be used in assessing mechanisms,
45 but all end-points measured in a single experiment need to be considered in the
46 proper context. For high-throughput data, where the number of observations far
47 exceeds the number of end-points measured, their utility for identifying common
48 mechanisms across multiple agents is enhanced. These data can be used to
49 identify mechanisms that not only seem plausible, but also have a consistent
50 pattern of carcinogenic response across entire classes of related compounds.
March 2008 AX1-37 DRAFT-DO NOT QUOTE OR CITE
-------
1 (d) Susceptibility data
2 Individuals, populations and life-stages may have greater or lesser susceptibility
3 to an agent, based on toxicokinetics, mechanisms of carcinogenesis and other
4 factors. Examples of host and genetic factors that affect individual susceptibility
5 include sex, genetic polymorphisms of genes involved in the metabolism of the
6 agent under evaluation, differences in metabolic capacity due to life-stage or the
7 presence of disease, differences in DNA repair capacity, competition for or
8 alteration of metabolic capacity by medications or other chemical exposures,
9 pre-existing hormonal imbalance that is exacerbated by a chemical exposure, a
10 suppressed immune system, periods of higher-than-usual tissue growth or
11 regeneration and genetic polymorphisms that lead to differences in behaviour
12 (e.g. addiction). Such data can substantially increase the strength of the evidence
13 from epidemiological data and enhance the linkage of in-vivo and in-vitro
14 laboratory studies to humans.
15 (e) Data on other adverse effects
16 Data on acute, subchronic and chronic adverse effects relevant to the cancer
17 evaluation are summarized. Adverse effects that confirm distribution and
18 biological effects at the sites of tumour development, or alterations in
19 physiology that could lead to tumour development, are emphasized. Effects on
20 reproduction, embryonic and fetal survival and development are summarized
21 briefly. The adequacy of epidemiological studies of reproductive outcome and
22 genetic and related effects in humans is judged by the same criteria as those
23 applied to epidemiological studies of cancer, but fewer details are given.
24 5. Summary
25 This section is a summary of data presented in the preceding sections.
26 Summaries can be found on the Monographs programme website
27 (http://monographs.iarc.fr).
28 (a) Exposure data
29 Data are summarized, as appropriate, on the basis of elements such as
30 production, use, occurrence and exposure levels in the workplace and
31 environment and measurements in human tissues and body fluids. Quantitative
32 data and time trends are given to compare exposures in different occupations
33 and environmental settings. Exposure to biological agents is described in terms
34 of transmission, prevalence and persistence of infection.
3 5 (b) Cancer in humans
3 6 Results of epidemiological studies pertinent to an assessment of human
37 carcinogenicity are summarized. When relevant, case reports and correlation
3 8 studies are also summarized. The target organ(s) or tissue(s) in which an
39 increase in cancer was observed is identified. Dose-response and other
40 quantitative data may be summarized when available.
41 (c) Cancer in experimental animals
42 Data relevant to an evaluation of carcinogenicity in animals are summarized.
43 For each animal species, study design and route of administration, it is stated
44 whether an increased incidence, reduced latency, or increased severity or
March 2008 AX1-38 DRAFT-DO NOT QUOTE OR CITE
-------
1 multiplicity of neoplasms or preneoplastic lesions were observed, and the
2 tumour sites are indicated. If the agent produced tumours after prenatal
3 exposure or in single-dose experiments, this is also mentioned. Negative
4 findings, inverse relationships, dose-response and other quantitative data are
5 also summarized.
6 (d) Mechanistic and other relevant data
7 Data relevant to the toxicokinetics (absorption, distribution, metabolism,
8 elimination) and the possible mechanism(s) of carcinogenesis (e.g. genetic
9 toxicity, epigenetic effects) are summarized. In addition, information on
10 susceptible individuals, populations and life-stages is summarized. This section
11 also reports on other toxic effects, including reproductive and developmental
12 effects, as well as additional relevant data that are considered to be important.
13 6. Evaluation and rationale
14 Evaluations of the strength of the evidence for carcinogenicity arising from
15 human and experimental animal data are made, using standard terms. The
16 strength of the mechanistic evidence is also characterized.
17 It is recognized that the criteria for these evaluations, described below, cannot
18 encompass all of the factors that may be relevant to an evaluation of
19 carcinogenicity. In considering all of the relevant scientific data, the Working
20 Group may assign the agent to a higher or lower category than a strict
21 interpretation of these criteria would indicate.
22 These categories refer only to the strength of the evidence that an exposure is
23 carcinogenic and not to the extent of its carcinogenic activity (potency). A
24 classification may change as new information becomes available.
25 An evaluation of the degree of evidence is limited to the materials tested, as
26 defined physically, chemically or biologically. When the agents evaluated are
27 considered by the Working Group to be sufficiently closely related, they may be
28 grouped together for the purpose of a single evaluation of the degree of
29 evidence.
30 (a) Carcinogenicity in humans
31 The evidence relevant to carcinogenicity from studies in humans is classified
32 into one of the following categories:
33 Sufficient evidence of carcinogenicity. The Working Group considers that a
34 causal relationship has been established between exposure to the agent and
3 5 human cancer. That is, a positive relationship has been observed between the
3 6 exposure and cancer in studies in which chance, bias and confounding could be
37 ruled out with reasonable confidence. A statement that there is sufficient
3 8 evidence is followed by a separate sentence that identifies the target organ(s) or
3 9 tissue(s) where an increased risk of cancer was observed in humans.
40 Identification of a specific target organ or tissue does not preclude the possibility
41 that the agent may cause cancer at other sites.
42 Limited evidence of carcinogenicity. A positive association has been observed
43 between exposure to the agent and cancer for which a causal interpretation is
March 2008 AX1-39 DRAFT-DO NOT QUOTE OR CITE
-------
1 considered by the Working Group to be credible, but chance, bias or
2 confounding could not be ruled out with reasonable confidence.
3 Inadequate evidence of carcinogenicity. The available studies are of
4 insufficient quality, consistency or statistical power to permit a conclusion
5 regarding the presence or absence of a causal association between exposure and
6 cancer, or no data on cancer in humans are available.
7 Evidence suggesting lack of carcinogenicity. There are several adequate
8 studies covering the full range of levels of exposure that humans are known to
9 encounter, which are mutually consistent in not showing a positive association
10 between exposure to the agent and any studied cancer at any observed level of
11 exposure. The results from these studies alone or combined should have narrow
12 confidence intervals with an upper limit close to the null value (e.g. a relative
13 risk of 1.0). Bias and confounding should be ruled out with reasonable
14 confidence, and the studies should have an adequate length of follow-up. A
15 conclusion of evidence suggesting lack of carcinogenicity is inevitably limited to
16 the cancer sites, conditions and levels of exposure, and length of observation
17 covered by the available studies. In addition, the possibility of a very small risk
18 at the levels of exposure studied can never be excluded.
19 In some instances, the above categories may be used to classify the degree of
20 evidence related to carcinogenicity in specific organs or tissues.
21 When the available epidemiological studies pertain to a mixture, process,
22 occupation or industry, the Working Group seeks to identify the specific agent
23 considered most likely to be responsible for any excess risk. The evaluation is
24 focused as narrowly as the available data on exposure and other aspects permit.
25 (b) Carcinogenicity in experimental animals
26 Carcinogenicity in experimental animals can be evaluated using conventional
27 bioassays, bioassays that employ genetically modified animals, and other in-
28 vivo bioassays that focus on one or more of the critical stages of carcinogenesis.
29 In the absence of data from conventional long-term bioassays or from assays
30 with neoplasia as the end-point, consistently positive results in several models
31 that address several stages in the multistage process of carcinogenesis should be
32 considered in evaluating the degree of evidence of carcinogenicity in
33 experimental animals.
34 The evidence relevant to carcinogenicity in experimental animals is classified
3 5 into one of the following categories:
36 Sufficient evidence of carcinogenicity. The Working Group considers that a
3 7 causal relationship has been established between the agent and an increased
3 8 incidence of malignant neoplasms or of an appropriate combination of benign
39 and malignant neoplasms in (a) two or more species of animals or (b) two or
40 more independent studies in one species carried out at different times or in
41 different laboratories or under different protocols. An increased incidence of
42 tumours in both sexes of a single species in a well-conducted study, ideally
43 conducted under Good Laboratory Practices, can also provide sufficient
44 evidence.
45 A single study in one species and sex might be considered to provide sufficient
46 evidence of carcinogenicity when malignant neoplasms occur to an unusual
March 2008 AX1-40 DRAFT-DO NOT QUOTE OR CITE
-------
1 degree with regard to incidence, site, type of tumour or age at onset, or when
2 there are strong findings of tumours at multiple sites.
3 Limited evidence of carcinogenicity. The data suggest a carcinogenic effect but
4 are limited for making a definitive evaluation because, e.g. (a) the evidence of
5 carcinogenicity is restricted to a single experiment; (b) there are unresolved
6 questions regarding the adequacy of the design, conduct or interpretation of the
7 studies; (c) the agent increases the incidence only of benign neoplasms or
8 lesions of uncertain neoplastic potential; or (d) the evidence of carcinogenicity is
9 restricted to studies that demonstrate only promoting activity in a narrow range
10 of tissues or organs.
11 Inadequate evidence of carcinogenicity. The studies cannot be interpreted as
12 showing either the presence or absence of a carcinogenic effect because of major
13 qualitative or quantitative limitations, or no data on cancer in experimental
14 animals are available.
15 Evidence suggesting lack of carcinogenicity: Adequate studies involving at
16 least two species are available which show that, within the limits of the tests
17 used, the agent is not carcinogenic. A conclusion of evidence suggesting lack of
18 carcinogenicity is inevitably limited to the species, tumour sites, age at
19 exposure, and conditions and levels of exposure studied.
20 (c) Mechanistic and other relevant data
21 Mechanistic and other evidence judged to be relevant to an evaluation of
22 carcinogenicity and of sufficient importance to affect the overall evaluation is
23 highlighted. This may include data on preneoplastic lesions, tumour pathology,
24 genetic and related effects, structure-activity relationships, metabolism and
25 toxicokinetics, physicochemical parameters and analogous biological agents.
26 The strength of the evidence that any carcinogenic effect observed is due to a
27 particular mechanism is evaluated, using terms such as 'weak', 'moderate' or
28 'strong'. The Working Group then assesses whether that particular mechanism
29 is likely to be operative in humans. The strongest indications that a particular
3 0 mechanism operates in humans derive from data on humans or biological
31 specimens obtained from exposed humans. The data may be considered to be
32 especially relevant if they show that the agent in question has caused changes in
33 exposed humans that are on the causal pathway to carcinogenesis. Such data
34 may, however, never become available, because it is at least conceivable that
3 5 certain compounds may be kept from human use solely on the basis of evidence
36 of their toxicity and/or carcinogenicity in experimental systems.
3 7 The conclusion that a mechanism operates in experimental animals is
3 8 strengthened by findings of consistent results in different experimental systems,
39 by the demonstration of biological plausibility and by coherence of the overall
40 database. Strong support can be obtained from studies that challenge the
41 hypothesized mechanism experimentally, by demonstrating that the suppression
42 of key mechanistic processes leads to the suppression of tumour development.
43 The Working Group considers whether multiple mechanisms might contribute to
44 tumour development, whether different mechanisms might operate in different
45 dose ranges, whether separate mechanisms might operate in humans and
46 experimental animals and whether a unique mechanism might operate in a
47 susceptible group. The possible contribution of alternative mechanisms must be
48 considered before concluding that tumours observed in experimental animals are
49 not relevant to humans. An uneven level of experimental support for different
March 2008 AX1-41 DRAFT-DO NOT QUOTE OR CITE
-------
1 mechanisms may reflect that disproportionate resources have been focused on
2 investigating a favoured mechanism.
3 For complex exposures, including occupational and industrial exposures, the
4 chemical composition and the potential contribution of carcinogens known to be
5 present are considered by the Working Group in its overall evaluation of human
6 carcinogenicity. The Working Group also determines the extent to which the
7 materials tested in experimental systems are related to those to which humans
8 are exposed.
9 (d) Overall evaluation
10 Finally, the body of evidence is considered as a whole, in order to reach an
11 overall evaluation of the carcinogenicity of the agent to humans.
12 An evaluation may be made for a group of agents that have been evaluated by
13 the Working Group. In addition, when supporting data indicate that other related
14 agents, for which there is no direct evidence of their capacity to induce cancer in
15 humans or in animals, may also be carcinogenic, a statement describing the
16 rationale for this conclusion is added to the evaluation narrative; an additional
17 evaluation may be made for this broader group of agents if the strength of the
18 evidence warrants it.
19 The agent is described according to the wording of one of the following
20 categories, and the designated group is given. The categorization of an agent is
21 a matter of scientific judgement that reflects the strength of the evidence derived
22 from studies in humans and in experimental animals and from mechanistic and
23 other relevant data.
24 Group 1: The agent is carcinogenic to humans.
25 This category is used when there is sufficient evidence of carcinogenicity in
26 humans. Exceptionally, an agent may be placed in this category when evidence
27 of carcinogenicity in humans is less than sufficient but there is sufficient
28 evidence of carcinogenicity in experimental animals and strong evidence in
29 exposed humans that the agent acts through a relevant mechanism of
3 0 carcinogenicity.
31 Group 2.
32 This category includes agents for which, at one extreme, the degree of evidence
33 of carcinogenicity in humans is almost sufficient, as well as those for which, at
34 the other extreme, there are no human data but for which there is evidence of
3 5 carcinogenicity in experimental animals. Agents are assigned to either Group
36 2A (probably carcinogenic to humans) or Group 2B (possibly carcinogenic to
3 7 humans) on the basis of epidemiological and experimental evidence of
3 8 carcinogenicity and mechanistic and other relevant data. The terms probably
39 carcinogenic and possibly carcinogenic have no quantitative significance and
40 are used simply as descriptors of different levels of evidence of human
41 carcinogenicity, with probably carcinogenic signifying a higher level of
42 evidence than possibly carcinogenic.
March 2008 AX1-42 DRAFT-DO NOT QUOTE OR CITE
-------
1 Group 2A: The agent improbably carcinogenic to humans.
2 This category is used when there is limited evidence of carcinogenicity in
3 humans and sufficient evidence of carcinogenicity in experimental animals. In
4 some cases, an agent may be classified in this category when there is inadequate
5 evidence of carcinogenicity in humans and sufficient evidence of carcinogenicity
6 in experimental animals and strong evidence that the carcinogenesis is mediated
7 by a mechanism that also operates in humans. Exceptionally, an agent may be
8 classified in this category solely on the basis of limited evidence of
9 carcinogenicity in humans. An agent may be assigned to this category if it
10 clearly belongs, based on mechanistic considerations, to a class of agents for
11 which one or more members have been classified in Group 1 or Group 2A.
12 Group 2B: The agent is possibly carcinogenic to humans.
13 This category is used for agents for which there is limited evidence of
14 carcinogenicity in humans and less than sufficient evidence of carcinogenicity in
15 experimental animals. It may also be used when there is inadequate evidence of
16 carcinogenicity in humans but there is sufficient evidence of carcinogenicity in
17 experimental animals. In some instances, an agent for which there is inadequate
18 evidence of carcinogenicity in humans and less than sufficient evidence of
19 carcinogenicity in experimental animals together with supporting evidence from
20 mechanistic and other relevant data may be placed in this group. An agent may
21 be classified in this category solely on the basis of strong evidence from
22 mechanistic and other relevant data.
23 Group 3: The agent is not classifiable as to its carcinogenicity to humans.
24 This category is used most commonly for agents for which the evidence of
25 carcinogenicity is inadequate in humans and inadequate or limited in
26 experimental animals.
27 Exceptionally, agents for which the evidence of carcinogenicity is inadequate in
28 humans but sufficient in experimental animals may be placed in this category
29 when there is strong evidence that the mechanism of carcinogenicity in
3 0 experimental animals does not operate in humans.
31 Agents that do not fall into any other group are also placed in this category.
32 An evaluation in Group 3 is not a determination of non-carcinogenicity or
33 overall safety. It often means that further research is needed, especially when
34 exposures are widespread or the cancer data are consistent with differing
3 5 interpretations.
36 Group 4: The agent improbably not carcinogenic to humans.
3 7 This category is used for agents for which there is evidence suggesting lack of
3 8 carcinogenicity in humans and in experimental animals. In some instances,
3 9 agents for which there is inadequate evidence of carcinogenicity in humans but
40 evidence suggesting lack of carcinogenicity in experimental animals,
41 consistently and strongly supported by a broad range of mechanistic and other
42 relevant data, may be classified in this group.
March 2008 AX1-43 DRAFT-DO NOT QUOTE OR CITE
-------
1 (e) Rationale
2 The reasoning that the Working Group used to reach its evaluation is presented
3 and discussed. This section integrates the major findings from studies of cancer
4 in humans, studies of cancer in experimental animals, and mechanistic and other
5 relevant data. It includes concise statements of the principal line(s) of argument
6 that emerged, the conclusions of the Working Group on the strength of the
7 evidence for each group of studies, citations to indicate which studies were
8 pivotal to these conclusions, and an explanation of the reasoning of the Working
9 Group in weighing data and making evaluations. When there are significant
10 differences of scientific interpretation among Working Group Members, a brief
11 summary of the alternative interpretations is provided, together with their
12 scientific rationale and an indication of the relative degree of support for each
13 alternative.
14
15 AX1.4.2.6 National Toxicology Program Criteria
16
17 The criteria for listing an agent, substance, mixture, or exposure circumstance in the
18 National Toxicology Program's Report on Carcinogens (NTP, 2005) are as follows:
19 Known To Be Human Carcinogen:
20 There is sufficient evidence of carcinogenicity from studies in humans*, which
21 indicates a causal relationship between exposure to the agent, substance, or
22 mixture, and human cancer.
23 Reasonably Anticipated To Be Human Carcinogen:
24 There is limited evidence of carcinogenicity from studies in humans*, which
25 indicates that causal interpretation is credible, but that alternative explanations,
26 such as chance, bias, or confounding factors, could not adequately be excluded,
27
33
or
28 there is sufficient evidence of carcinogenicity from studies in experimental
29 animals, which indicates there is an increased incidence of malignant and/or a
3 0 combination of malignant and benign tumors (1) in multiple species or at
31 multiple tissue sites, or (2) by multiple routes of exposure, or (3) to an unusual
32 degree with regard to incidence, site, or type of tumor, or age at onset,
or
34 there is less than sufficient evidence of carcinogenicity in humans or laboratory
3 5 animals; however, the agent, substance, or mixture belongs to a well-defined,
3 6 structurally related class of substances whose members are listed in a previous
3 7 Report on Carcinogens as either known to be a human carcinogen or reasonably
3 8 anticipated to be a human carcinogen, or there is convincing relevant
3 9 information that the agent acts through mechanisms indicating it would likely
40 cause cancer in humans.
41 Conclusions regarding carcinogenicity in humans or experimental animals are
42 based on scientific judgment, with consideration given to all relevant
March 2008 AX1-44 DRAFT-DO NOT QUOTE OR CITE
-------
1 information. Relevant information includes, but is not limited to, dose response,
2 route of exposure, chemical structure, metabolism, pharmacokinetics, sensitive
3 sub-populations, genetic effects, or other data relating to mechanism of action or
4 factors that may be unique to a given substance. For example, there may be
5 substances for which there is evidence of carcinogenicity in laboratory animals,
6 but there are compelling data indicating that the agent acts through mechanisms
7 which do not operate in humans and would therefore not reasonably be
8 anticipated to cause cancer in humans.
9 This evidence can include traditional cancer epidemiology studies, data from
10 clinical studies, and/or data derived from the study of tissues or cells from
11 humans exposed to the substance in question that can be useful for evaluating
12 whether a relevant cancer mechanism is operating in people.
March 2008 AX1-45 DRAFT-DO NOT QUOTE OR CITE
-------
TABLE AX1.3-1. LITERATURE SEARCH STRATEGY FOR EPIDEMIOLOGIC
STUDIES: EXAMPLES OF KEYWORDS
Search engines used include: MEDLINE, BIOSYS,...
Search Key Word Examples:
Nitrogen Oxides or Nitric Acid or Nitrous Oxide or Nitrogen Dioxide or Nitrogen Tetroxide or Nitrogen Trioxide
or NOX or NO2 or HNO3 or Peroxyacetyl Nitrate or HNO4 or NO3 or N2O5 or Pan or Ch3COOONO2 or HNO2 or
HONO or Organic Nitrate or Peroxynitric Acid or Nitrogen Pentoxide or Nitrous Acid
Mortality or Epidemiologic Studies or Hospitals
Asthma or Bronchial Hyperactivity or Lung Diseases, Obstructive or Respiratory Hypersensitivity and
Immunology, Respiratory Tract Diseases or Respiratory Tract Infections, Lung Infection or Respiratory Disease
or Respiratory System
Neoplasm or Neoplastic or Cancer or Carcinogen, Mutation or Chromosome Aberrations or Mutagenicity Tests
Pregnancy Complication or Prenatal Exposure or Delayed Effects or Teratogens
TABLE AX1.3-2. LITERATURE SEARCH STRATEGY FOR THE
ATMOSPHERIC SCIENCES
Search Engine: Web of Knowledge
Search Key Words:
Exposure and (NO2 or NO or Nitrogen Dioxide or Nitrogen Oxide(s) or Nitrous Oxide or Oxide(s) of Nitrogen or
HNO3 or HONO or Nitric Acid or Nitrous Acid or PAN(s) or Nitro-PAH(s) or NO3 Radical)
Indoor and (NO2 or NO or Nitrogen Dioxide or Nitrogen Oxide(s) or Nitrous Oxide or Oxide(s) of Nitrogen or
HNO3 or HONO or Nitric Acid or Nitrous Acid or PAN(s) or Nitro-PAH(s) or NO3 Radical)
(Source Apportionment or Source(s) or PMF or CMB or Receptor Model) and (NO2 or NO or Nitrogen Dioxide
or Nitrogen Oxide(s) or Nitrous Oxide or Oxide(s) of Nitrogen or HNO3 or HONO or Nitric Acid or Nitrous Acid
or PAN(s) or Nitro-PAH(s) or NO3 Radical)
(Traffic or Street Canyon) and (NO2 or NO or Nitrogen Dioxide or Nitrogen Oxide(s) or Nitrous Oxide or
Oxide(s) of Nitrogen or HNO3 or HONO or Nitric Acid or Nitrous Acid or PAN(s) or Nitro-PAH(s) or NO3
Radical)
Sampler and (NO2 or NO or Nitrogen Dioxide or Nitrogen Oxide(s) or Nitrous Oxide or Oxide(s) of Nitrogen or
HNO3 or HONO or Nitric Acid or Nitrous Acid or PAN(s) or Nitro-PAH(s) or NO3 Radical)
March 2008 AX1-46 DRAFT-DO NOT QUOTE OR CITE
-------
1 AX1.5 REFERENCES
2 Bieler, G. S.; Williams, R. L. (1993) Ratio estimates, the delta method, and quantal response
3 tests for increased carcinogenicity. Biometrics 49: 793-801.
4 Breslow, N. E.; Day, N. E. (1980) Statistical methods in cancer research, volume 1 - the analysis
5 of case-control studies. Lyon, France: International Agency for Research on Cancer.
6 (IARC scientific publications series: no. 32).
7 Breslow, N. E.; Day, N. E. (1987) Statistical methods in cancer research, volume 2 - the design
8 and analysis of cohort studies. Lyon, France: International Agency for Research on
9 Cancer. (IARC scientific publications series: no. 82).
10 Buffler, P.; Rice, J.; Baan, R.; Bird, M.; Boffetta, P., eds. (2004) Mechanisms of carcinogenesis:
11 contributions of molecular epidemiology. Lyon, France: International Agency for
12 Research on Cancer. (IARC scientific publications series: no. 157).
13 Capen, C. C.; Dybing, E.; Rice, J. M.; Wilbourn, J. D. (1999) Species differences in throid,
14 kidney and urinary bladder carcinogenesis: proceedings of a workshop; November; Lyon,
15 France. Lyon, France: International Agency of Cancer Research (IARC). [IARC
16 scientific publications series: no. 147].
17 Centers for Disease Control and Prevention. (2004) Introduction and approach to causal
18 inference. In: The health consequences of smoking: a report of the Surgeon General.
19 Atlanta, GA: U.S. Department of Health and Human Services, Centers for Disease
20 Control and Prevention, National Center for Chronic Disease Prevention and Health
21 Promotion, Office on Smoking and Health. Available:
22 http://www.cdc.gov/tobacco/sgr/sgr_2004/chapters.htm (18 August, 2004).
23 Cogliano, V. J.; Baan, R. A.; Straif, K.; Grosse, Y.; Secretan, M. B.; El Ghissassi, F.; Kleihues,
24 P. (2004) The science and practice of carcinogen identification and evaluation. Environ.
25 Health Perspect. 112: 1269-1274.
26 Dunson, D. B.; Chen, Z.; Harry, J. (2003) A bayesian approach for joint modeling of cluster size
27 and subunit-specific outcomes. Biometrics 59: 521-530.
28 Evans, A. S. (1993) Causation and disease: a chronological journey. New York, NY: Plenum
29 Publishing Co.
30 Federal Register. (1971) National primary and secondary ambient air quality standards. F. R.
31 (April 30) 36: 8186-8201.
32 Federal Register. (1984) Proposed reaffirmation of the National Ambient Air Quality Standards
33 for Nitrogen Dioxide. F. R. 49 (February 23): 6866-6879.
34 Federal Register. (1985) Retention of the national ambient air quality standards for nitrogen
35 dioxide: final rule. F. R. (June 19) 50: 25,532-25,545.
36 Federal Register. (1987) Air quality criteria for carbon monoxide; air quality criteria for oxides
37 of nitrogen. F. R. 52 (July 22): 27580.
38 Federal Register. (1991) Draft criteria document for oxides of nitrogen. F. R. 56 (November 25):
39 59285.
March 2008 AX1-47 DRAFT-DO NOT QUOTE OR CITE
-------
1 Federal Register. (1995) National ambient air quality standards for nitrogen dioxide: proposed
2 decision. F. R. 60 (October 11): 52874-52889.
3 Fung, K. Y.; Krewski, D.; Smythe, R. T. (1996) A comparison of tests for trend with historical
4 controls in carcinogen biosaasy. Can. J. Stat. 24: 431-454.
5 Gait, J. J.; Krewski, D.; Lee, P. N.; Tarone, R. E.; Wahrendorf, J. (1986) Statistical methods in
6 cancer research, volume 3 - the design and analysis of long-term animal experiments.
7 Lyon, France: International Agency for Research on Cancer. (IARC scientific
8 publications series: no. 79).
9 Greenland, S. (1998) Meta-analysis. In: Rothman, K. J.; Greenland, S., eds. Modern
10 epidemiology. Philadelphia, PA: Lippincott Williams & Wilkins; pp. 643-673.
11 Greim, H.; Gelbke, H.-P.; Reuter, R.; Thielmann, H. W.; Edler, L. (2003) Evaluation of
12 historical control data in carcinogenicity studies. Hum. Exp. Toxicol. 22: 541-549.
13 Haseman, J. K.; Huff, J.; Boorman, G. A. (1984) Use of historical control data in carcinogenicity
14 studies in rodents. Toxicol. Pathol. 12: 126-135.
15 Hill, A. B. (1965) The environment and disease: association or causation? Proc. R. Soc. Med. 58:
16 295-300.
17 Hoel, D. G.; Kaplan, N. L.; Anderson, M. W. (1983) Implication of nonlinear kinetics on risk
18 estimation in carcinogenesis. Science (Washington, DC) 219: 1032-1037.
19 Huff, J. E.; Eustis, S. L.; Haseman, J. K. (1989) Occurrence and relevance of chemically induced
20 benign neoplasms in long-term carcinogenicity studies. Cancer Metastasis Rev. 8: 1-21.
21 Institute of Medicine (IOM). (1999) Veterans and agent orange: update 1998. Washington, DC:
22 National Academy Press. Available: http://www.nap.edu/catalog.php?record_id=6415
23 [15 February, 2008].
24 Institute of Medicine (IOM). (2007) Improving the presumptive disability decision-making
25 process for veterans. Washington, DC: The National Academies Press. Available:
26 http://www.nap.edu/catalog.php?record_id=l 1908 (11 February, 2008).
27 International Agency for Research on Cancer (IARC). (1986) Tobacco smoking. Lyon, France:
28 International Agency for Research on Cancer. (IARC monographs on the evaluation of
29 the carcinogenic risk of chemicals to humans, v. 38).
30 International Agency for Research on Cancer. (1991) A consensus report of an IARC
31 monographs working group on the use of mechanisms of carcinogenesis in risk
32 identification. Lyon, France: International Agency for Research on Cancer. (IARC
33 internal technical report no. 91/002.
34 International Agency for Research on Cancer. (2004) Some drinking-water disinfectants and
35 contaminants, including arsenic. Lyon, France: International Agency for Research on
36 Cancer. (IARC monographs on the evaluation of carcinogenic risks to humans: v. 84).
37 International Agency for Research on Cancer. (2006) Report of the advisory group to review the
38 amended preamble to the IARC monographs. Lyon, France: International Agency for
39 Research on Cancer. (IARC internal report no. 06/001.
March 2008 AX1-48 DRAFT-DO NOT QUOTE OR CITE
-------
1 Kaufman, J. S.; Poole, C. (2000) Looking back on "causal thinking in the health sciences." Annu.
2 Rev. Public Health 21: 101-119.
3 Lilienfeld, A. M. (1959) "On the methodology of investigations of etiologic factors in chronic
4 diseases"~some comments. J. Chronic Dis. 10: 41-46.
5 McGregor, D. B.; Rice, J. M.; Venitt, S., eds. (1999) The use of short- and medium-term tests for
6 carcinogens and data on genetic effects in carcinogenic hazard evaluation. Lyon, France:
7 International Agency for Research on Cancer. (IARC scientific publications series: no.
8 146).
9 Montesano, R.; Bartsch, H.; Vainio, H.; Wilbourn, J.; Yamasaki, H., eds. (1986) Long-term and
10 short-term assays for carcinogenesis—a critical appraisal. Lyon, France: International
11 Agency for Research on Cancer. (IARC scientific publications series: no. 83).
12 National Acid Precipitation Program (NAPAP). (1991) The experience and legacy of NAPAP.
13 Report of the Oversight Review Board. Washington, DC: National Acid Precipitation
14 Program.
15 National Toxicology Program. (2005) Report on carcinogens. 11th ed. Washington, DC:
16 National Toxicology Program.
17 OECD. (2002) Guidance notes for analysis and evaluation of chronic toxicity and
18 carcinogenicity studies. Paris, France: Organisation for Economic Co-operation and
19 Development. Series on testing and assessment no. 35 and series on pesticides no. 14;
20 publication no. ENV/JM/MONO(2002)19. Available:
21 http://www.olis.oecd.org/olis/2002doc.nsf/LinkTo/env-jm-mono(2002)19 [28 February,
22 2008].
23 Peto, R.; Pike, M. C.; Day, N. E.; Gray, R. G.; Lee, P. N.; Parish, S.; Peto, J.; Richards, S.;
24 Wahrendorf, J. (1980) Guidelines for simple, sensitive significance tests for carcinogenic
25 effects in long-term animal experiments. In: Long-term and short-term screening assays
26 for carcinogens: a critical appraisal. Lyon, France: International Agency for Research on
27 Cancer; pp. 311-426. (IARC monographs on the evaluation of the carcinogenic risk of
28 chemicals to humans: suppl. 2).
29 Portier, C. J.; Bailer, A. J. (1989) Testing for increased carcinogenicity using a survival-adjusted
30 quantal response test. Fundam. Appl. Toxicol. 12: 731-737.
31 Rothman, K. J.; Greenland, S., eds. (1998) Modern epidemiology. 2nd ed. Philadelphia, PA:
32 Lippincott-Raven Publishers.
33 Sartwell, P. E. (1960) "On the methodology of investigations of etiologic factors in chronic
34 diseases"—further comments. J. Chronic Dis. 11: 61-63.
35 Sherman, C.D.; Portier, C.J.; Kopp-Schneider, A. (1994) Multistage models of carcinogenesis:
36 an approximation for the size and number distribution of late-stage clones. Risk Anal. 14:
37 1039-1048.
38 Susser, M. (1973) Causal thinking in the health sciences: concepts and strategies of
39 epidemiology. New York, NY: Oxford University Press.
40 Susser, M. (1977) Judgment and causal inference: criteria in epidemiologic studies. Am. J.
41 Epidemiol. 105: 1-15.
March 2008 AX1-49 DRAFT-DO NOT QUOTE OR CITE
-------
1 Susser, M. (1991) What is a cause and how do we know one? A grammar for pragmatic
2 epidemiology. Am. J. Epidemiol. 133: 635-648.
3 Tomatis, L.; Aitio, A.; Wilbourn, J.; Shuker, L. (1989) Human carcinogens so far identified. Jpn.
4 J. Cancer Res. 80: 795-807.
5 Toniolo, P.; Boffetta, P.; Shuker, D. E. G.; Rothman, N.; Hulka, B.; Pearce, N., eds. (1997)
6 Application of biomarkers in cancer epidemiology. Lyon, France: International Agency
7 for Research on Cancer. (IARC scientific publications series: no. 142).
8 U.S. Code. (2003a) Clean Air Act, §108, air quality criteria and control techniques. U. S. C. 42:
9 §7408.
10 U.S. Code. (2003b) Clean Air Act, §109, national ambient air quality standards. U. S. C. 42:
11 §7409.
12 U.S. Code. (2005) Clean Air Act, §302, definitions. U. S. C. 42: §7602(h).
13 U.S. Court of Appeals for the District of Columbia. (1980) Lead Industries v. U.S.
14 Environmental Protection Agency. 647 F2d 1130, 1154 (DC Cir. 1980).
15 U.S. Court of Appeals for the District of Columbia. (1981) American Petroleum Institute v.
16 Costle. 665 F2d 1176, 1186 (DC Cir. 1981).
17 U.S. Department of Health, Education, and Welfare (DHEW). (1964) Smoking and health: report
18 of the advisory committee to the Surgeon General of the Public Health Service.
19 Washington, DC: Public Health Service; PHS publication no. 1103.
20 U.S. Environmental Protection Agency. (1982) Air quality criteria for oxides of nitrogen.
21 Research Triangle Park, NC: Office of Health and Environmental Assessment,
22 Environmental Criteria and Assessment Office; EPA report no. EPA-600/8-82-026.
23 Available from: NTIS, Springfield, VA; PB83-131011.
24 U.S. Environmental Protection Agency. (1993) Air quality criteria for oxides of nitrogen.
25 Research Triangle Park, NC: Office of Health and Environmental Assessment,
26 Environmental Criteria and Assessment Office; report nos. EPA/600/8-91/049aF-cF. 3v.
27 Available from: NTIS, Springfield, VA; PB95-124533, PB95-124525, and PB95-124517.
28 U.S. Environmental Protection Agency. (1995) Review of the national ambient air quality
29 standards for nitrogen dioxide: assessment of scientific and technical information.
30 Research Triangle Park, NC: Office of Air Quality Planning and Standards; report no.
31 EPA/452/R-95-005.
32 U.S. Environmental Protection Agency. (2005) Guidelines for carcinogen risk assessment.
33 Washington, DC: Risk Assessment Forum; report no. EPA/630/P-03/001F. Available:
34 http://cfpub.epa.gov/ncea/index.cfm [30 November, 2005].
35 U.S. Environmental Protection Agency. (2007) Integrated science assessment for oxides of
36 nitrogen: health criteria (first external review draft). Washington, DC: National Center
37 for Environmental Assessment. EPA/600/R-07/093. Available:
38 http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=181712 [11 October, 2007].
March 2008 AX1-50 DRAFT-DO NOT QUOTE OR CITE
-------
1 U.S. Senate. (1970) National Air Quality Standards Act of 1970: report of the Committee on
2 Public Works, United States Senate together with individual views to accompany S.
3 4358. Washington, DC: Committee on Public Works; report no. CONG/91-1196.
4 U.S. Supreme Court. (2001) Whitman v. American Trucking Association. 531 U.S. 457 (nos. 99-
5 1257 and 99-1426).
6 Vainio, H.; Magee, P. N.; McGregor, D. B.; McMichael, A. J., eds. (1992) Mechanisms of
7 carcinogenesis in risk identification. Lyon, France: World Health Organization,
8 International Agency for Research on Cancer. (IARC scientific publications series: no.
9 116).
10 Vainio, H.; Wilbourn, J. D.; Sasco, A. J.; Partensky, C.; Gaudin, N.; Heseltine, E.; Eragne, I.
11 (1995) Identification des facteurs cancerogenes pour rhomme dans les monographies du
12 CIRC [Identification of human carcinogenic risk in IARC monographs]. Bull. Cancer 82:
13 339-348.
14 Vineis, P.; Malats, N.; Lang, M.; d'Errico, A.; Caporaso, N.; Cuzick, J.; Boffetta, P., eds. (1999)
15 Metabolic polymorphisms and susceptibility to cancer. Lyon, France: International
16 Agency for Research on Cancer. (IARC scientific publications series: no. 148).
17 Weed, D. L. (2000) Epidemiologic evidence and causal inference. Hematol./Oncol. Clin. N. Am.
18 14: 797-807.
19 Wilbourn, J.; Haroun, L.; Heseltine, E.; Kaldor, J.; Partensky, C.; Vainio, H. (1986) Response of
20 experimental animals to human carcinogens: an analysis based upon the IARC
21 monographs programme. Carcinogenesis 7: 1853-1863.
22 Wolff, G. T. (1993) On a NOx-focused control strategy to reduce O3. J. Air Waste Manage.
23 Assoc. 43: 1593-1596.
24 Wolff, G. T. (1996) Closure by the Clean Air Scientific Advisory Committee (CASAC) on the
25 staff paper for parti culate matter [letter to Carol M. Browner, Administrator, U.S. EPA].
26 Washington, DC: U.S. Environmental Protection Agency, Clean Air Scientific Advisory
27 Committee; EPA-SAB-CASAC-LTR-96-008; June 13.
28 Yerushalmy, J.; Palmer, C. E. (1959) On the methodology of investigations of etiologic factors
29 in chronic diseases. J. Chronic Dis. 10: 27-40.
March 2008 AX1-51 DRAFT-DO NOT QUOTE OR CITE
-------
i AX2. CHAPTER 2 ANNEX-ATMOSPHERIC
2 CHEMISTRY OF NITROGEN AND SULFUR OXIDES
o
4
5 AX2.1 INTRODUCTION
6 Nitrogen oxides (NOX) along with volatile organic compounds (VOCs) including
7 anthropogenic and biogenic hydrocarbons, aldehydes, etc. and carbon monoxide (CO) serve as
8 precursors in the formation of ozone (O3) and other elements of photochemical smog. Nitrogen
9 oxides are defined here as nitric oxide (NO) and nitrogen dioxide (NO2), the latter of which is a
10 U.S. EPA Air Pollutant; similarly, oxides of sulfur (SOx) are defined here to be sulfur monoxide
11 (SO), sulfur dioxide (802), the largest component of SOx and also a U.S. EPA Criteria Air
12 Pollutant, and sulfur trioxide (SO3). SO3 rapidly reacts with water vapor to form H2SO4, and
13 only SO2 is present in the atmosphere at detectable levels.
14 Nitrogen dioxide is an oxidant and can further react to form other photochemical
15 oxidants, in particular the organic nitrates, including peroxy acetyl nitrates (PAN) and higher
16 PAN analogues. It can also react with toxic compounds such as poly cyclic aromatic
17 hydrocarbons (PAHs) to form nitro-PAHs, which may be even more toxic than the precursors.
18 Nitrogen dioxide together with sulfur dioxide (802), another U.S. EPA criteria air pollutant, can
19 be oxidized to the strong mineral acids, nitric acid (HNO3) and sulfuric acid (H2SO4), which
20 contribute to the acidity of cloud, fog, and rainwater, and can form ambient particles.
21 The role of NOx in O3 formation was reviewed in Chapter 2 (Section 2.2) of the latest
22 AQCD for Ozone and Other Photochemical Oxidants (U.S. Environmental Protection Agency,
23 2006 CD06), and in numerous texts (e.g., Seinfeld and Pandis, 1998; Jacob, 2000; Jacobson,
24 2002). Mechanisms for transporting O3 precursors, the factors controlling the efficiency of O3
25 production from NOx, methods for calculating O3 from its precursors, and methods for
26 measuring NOx were all reviewed in Section 2.6 of CD06. The main points from those
27 discussions in CD06 and updates, based on new materials will be presented here. Ammonia
28 (NH3) is included here because its oxidation can be a source of NOx, and it is a precursor for
29 ammonium ions (NH4+), which play a key role in neutralizing acidity in ambient particles and in
30 cloud, fog, and rain water. Ammonia is also involved in the ternary nucleation of new particles,
31 and it reacts with gaseous HNO3 to form ammonium nitrate (NH4NO3), which is a major
March 2008 AX2-1 DRAFT-DO NOT QUOTE OR CITE
-------
1 constituent of ambient Parti culate Matter (PM) in many areas. Ammonia is also involved in over
2 nitrification of aqueous and terrestrial ecosystems and participates in the N cascade (Galloway
3 et al., 2003)
4 The atmospheric chemistry of NOx is discussed in Section AX2.2, and of SC>2 in Section
5 AX2.3. Mechanisms for the formation of aqueous-phase sulfate (SC>42 ) and nitrate (NOs ) are
6 reviewed in Section AX2.4. Sources and emissions of NOx, NHa, and SC>2 are discussed in
7 Section AX2.5. Modeling methods used to calculate the atmospheric chemistry, transport, and
8 fate of NOx and SO2 and their oxidation products are presented in Section AX2.6. Measurement
9 techniques for the nitrogen-containing compounds and for SO2, nitrates, sulfates, and ammonium
10 ion are discussed in Section AX2.8. Estimates of policy-relevant background concentrations of
1 1 NOx and SOx are given in Section AX2.9. An overall review of key points in this chapter is
12 given in Section AX2. 1 1 .
13 The overall chemistry of reactive nitrogen compounds in the atmosphere is summarized
14 in Figure AX2.2-1 and is described in greater detail in the following sections. Nitrogen oxides
15 are emitted primarily as NO with smaller quantities of NO2. Emissions of NOx are spatially
16 distributed vertically with some occurring at or near ground level and others aloft as indicated in
17 Figure AX2.2-1. Because of atmospheric chemical reactions, the relative abundance of different
18 compounds contributed by different sources varies with location. Both anthropogenic and
19 natural (biogenic) processes emit NOX. In addition to gas phase reactions, multiphase processes
20 are important for forming aerosol-phase pollutants, including aerosol
21
22
23 AX2.2 CHEMISTRY OF NITROGEN OXIDES IN THE TROPOSPHERE
24
25 AX2.2.1 Basic Chemistry
26 There is a rapid photochemical cycle in the troposphere that involves photolysis of NO2
27 by solar UV-A radiation to yield NO and a ground-state oxygen atom, O(3P)
28 (AX2M)
29 This ground-state oxygen atom can then combine with molecular oxygen (©2) to form Os; and,
30 colliding with any molecule from the surrounding air (M = N2, 62, etc.), the newly formed Os
3 1 molecule, transfers excess energy and is stabilized
March 2008 AX2-2 DRAFT-DO NOT QUOTE OR CITE
-------
Long range transport to remote
regions at low temperatures
; PAH__^ nitro.pAHs
R-C=C R
>• RONO.
nitrosamines,
nitro-phenols, etc.
>-RONO,
deposition
emissions
Figure AX2.2-1.
Schematic diagram of the cycle of reactive nitrogen species in the
atmosphere. MPP refers to multi-phase process; hv to a photon of
solar energy.
(AX2.2-2)
5
6
7
8
9
where M = N2, 62. Reaction AX2.2-2 is the only significant reaction forming Oj in the
troposphere.
NO and 63 react to reform
(AX2.2-3)
NO + O3 -> M?2 + O2
Reaction AX2.2-3 is responsible for O3 decreases and NO2 increases found near sources of NO
(e.g., highways), especially at night when the actinic flux is 0. Oxidation of reactive VOCs leads
to the formation of reactive radical species that allow the conversion of NO to NO2 without the
participation of Os (as in Reaction AX2.2-3)
March 2008
AX2-3
DRAFT-DO NOT QUOTE OR CITE
-------
„„ HO,,RO,
N° ~ •
1 (AX2.2-4)
2 Ozone, therefore, can accumulate as NO2 photolyzes as in Reaction AX2.2-1, followed
3 by Reaction AX2.2-2. Specific mechanisms for the oxidation of a number of VOCs were
4 discussed in the Os AQCD (U.S. Environmental Protection Agency, 2006).
5 It is often convenient to speak about families of chemical species defined in terms of
6 members that interconvert rapidly among themselves on time scales that are shorter than those
7 for formation or destruction of the family as a whole. For example, an "odd oxygen" (Ox)
8 family can be defined as
9 Ox = I(0(-lP) + 0('D) + 03 + N02) (AX2.2-5)
10 In much the same way, NOX is sometimes referred to as "odd nitrogen". Hence, we see that
1 1 production of Ox occurs by the schematic Reaction AX2.2-4, and that the sequence of reactions
12 given by reactions AX2.2-1 through AX2.2-3 represents no net production of Ox. Definitions of
13 species families and methods for constructing families are discussed in Jacobson (1999) and
14 references therein. Other families that include nitrogen-containing species (and which will be
1 5 referred to later in this chapter) include:
(AX22_6)
17 One can then see that production of Ox occurs by the schematic Reaction AX2.2-4, and that the
18 sequence of reactions given by reactions AX2.2-1 through AX2.2-3 represents no net production
19 of Ox. Definitions of species families and methods for constructing families are discussed in
20 Jacobson (1999) and references therein. Other families that include nitrogen-containing species,
21 and which will be referred to later in this chapter, are: (which is the sum of the products of the
22 oxidation of NOX)
NOZ = X HNO3 + HNO4 + NO3 + 2NO2O5 + PAN(CH3CHO - OO - NO2) + other
23 organic nitrates + halogen nitrates + paniculate nitrate)
NOY = NOX + NO2 + HONO;
24 and NHX = NH3 + NH4 + (AX2.2-7)
March 2008 AX2-4 DRAFT-DO NOT QUOTE OR CITE
-------
The reaction of NO2 with O3 leads to the formation of N(V radical
3 However, because the NOs radical photolyzes rapidly (lifetime of ~5 s during the
4 photochemically most active period of the day around local solar noon (Atkinson et al., 1992),
(AX2.2-9a)
6 "^ ' ^v ' '(yuyf°) (AX2.2-9b)
7 its concentration remains low during daylight hours, but can increase after sunset to nighttime
8 concentrations of <5 x 107 to 1 x 1010 molecules cm"3 (<2 to 430 parts per trillion (ppt)) over
9 continental areas influenced by anthropogenic emissions of NOx (Atkinson et al., 1986). At
10 night, NOs, rather than the hydroxyl radical (OH), is the primary oxidant in the system.
11 Nitrate radicals can combine with NO2 to form dinitrogen pentoxide (N2Os)
M . « r /-i
12 ~* ^2 '";~:> (AX2.2-10)
13 and N2Os both photolyzes and thermally decomposes back to NO2 and NOs during the day;
14 however, N2Os concentrations ([N2Os]) can accumulate during the night to parts per billion (ppb)
15 levels in polluted urban atmospheres.
16 The tropospheric chemical removal processes for NOx include reaction of NO2 with the
17 OH radical and hydrolysis of N2Os in aqueous aerosol solutions if there is no organic coating.
18 Both of these reactions produce
19 23 (AX22_n)
20 25
21 The gas-phase reaction of the OH radical with NO2 (Reaction AX2.2-1 1) initiates one of
22 the major and ultimate removal processes for NOx in the troposphere. This reaction removes
23 OH and NO2 radicals and competes with hydrocarbons for OH radicals in areas characterized by
24 high NOx concentrations, such as urban centers (see Section AX2.2.2). The timescale (T) for
25 conversion of NOX to HNO3 in the planetary boundary layer at 40 N latitude ranges from about
March 2008 AX2-5 DRAFT-DO NOT QUOTE OR CITE
-------
1 4 hours in July to about 16 hours in January. The corresponding range in T at 25 N latitude is
2 between 4 and 5 hours, while at 50 N latitude, HNOs T ranges from about 4 to 20 hours (Martin
3 et al., 2003). In addition to gas-phase HNO3, Golden and Smith (2000) have shown on the basis
4 of theoretical studies that pernitrous acid (HOONO) is also produced by the reaction of NC>2 and
5 OH radicals. However, this channel of production most likely represents a minor yield
6 (approximately 15% at the surface) (Jet Propulsion Laboratory, 2003). Pernitrous acid will also
7 thermally decompose and can photolyze. Gas-phase HNOs formed from Reaction AX2.2-1 1
8 undergoes wet and dry deposition to the surface, and uptake by ambient aerosol particles.
9 Reaction AX2.2-1 1 limits NOx T to a range of hours to days.
10 In addition to the uptake of HNOs on particles and in cloud drops, it photolyzes and
1 1 reacts with OH radicals via
12
13 and
14 HN03 + OH^ N03 + H20 (AX2.2-14)
15 The lifetime of HNOs with respect to these two reactions is long enough for HNOs to act as a
16 reservoir species for NOx during long-range transport, contributing in this way to NO2 levels and
17 to Os formation in areas remote from the source region of the NOx that formed this HNOs.
1 8 Geyer and Platt (2002) concluded that Reaction AX2 .2- 1 2 constituted about 1 0% of the
19 removal of NOx at a site near Berlin, Germany during spring and summer. However, other
20 studies found a larger contribution to HNOs production from Reaction AX2.2-12. Dentener and
21 Crutzen (1993) estimated 20% in summer and 80% of HNOs production in winter is from
22 Reaction AX2.2-12. Tonnesen and Dennis (2000) found between 16 to 3 1% of summer HNO3
23 production was from Reaction AX2.2-12. The contribution of Reaction AX2.2-12 to HNO3
24 formation is highly uncertain during both winter and summer. The importance of Reaction
25 AX2.2-12 could be much higher during winter than during summer because of the much lower
26 concentration of OH radicals and the enhanced stability of ^Os due to lower temperatures and
27 less sunlight. Note that Reaction AX2. 2-12 proceeds as a heterogeneous reaction. Recent work
28 in the northeastern United States indicates that this reaction is proceeds at a faster rate in power
29 plant plumes than in urban plumes (Brown et al., 2006a,b; Frost et al., 2006).
March 2008 AX2-6 DRAFT-DO NOT QUOTE OR CITE
-------
1 OH radicals also can react with NO to produce nitrous acid (HONO or HNO2)
2 (AX22.15)
3 In the daytime, HNO2 is rapidly photolyzed back to the original reactants
5 Reaction AX2.2-15 is, however, a negligible source of HONO, which is formed mainly
6 by multiphase processes (see Section AX2.2.3). At night, heterogeneous reactions of NO2 in
7 aerosols or at the earth's surface result in accumulation of HONO (Lammel and Cape, 1996;
8 Jacob, 2000; Sakamaki et al., 1983; Pitts et al., 1984; Svensson et al., 1987; Jenkin et al., 1988;
9 Lammel and Perner, 1988; Notholt et al., 1992a,b). Harris et al. (1982) (e.g.) suggested that
10 photolysis of this HNO2 at sunrise could provide an important early-morning source of OH
1 1 radicals to drive Os formation.
12 Hydroperoxy (HO2) radicals can react with NO2 to produce pernitric acid (HNO4)
13 HO 2 + NO2 + M^> HNO4 + M 2 2
14 which then can thermally decompose and photolyze back to its original reactants. The acids
15 formed in these gas-phase reactions are all water soluble. Hence, they can be incorporated into
16 cloud drops and in the aqueous phase of particles.
17 Although the lifetimes of HNO4 and N2O5 are short (minutes to hours) during typical
18 summer conditions, they can be much longer at the lower temperatures and darkness found
19 during the polar night. Under these conditions, species such as PAN, HNOs, HNO/t, and ^Os
20 serve as NOx reservoirs that can liberate NO2 upon the return of sunlight during the polar spring.
21 A broad range of organic nitrogen compounds can be directly emitted by combustion sources or
22 formed in the atmosphere from NOx emissions. Organic nitrogen compounds include the PANs,
23 nitrosamines, nitro-PAHs, and the more recently identified nitrated organics in the quinone
24 family. Oxidation of VOCs produces organic peroxy radicals (RO2), as discussed in the latest
25 AQCD for Ozone and Other Photochemical Oxidants (U.S. Environmental Protection Agency,
26 2006). Reaction of RO2 radicals with NO and NO2 produces organic nitrates (RONO2) and
27 peroxynitrates (RO2NO2)
March 2008 AX2-7 DRAFT-DO NOT QUOTE OR CITE
-------
(AX2.2-18)
2 R02 + N02 M »R02N02 (AX2
3 Reaction (AX2.2-18) is a minor branch for the reaction of RO2 with NO. The major
4 branch produces RO and NO2, as discussed in the next section; however, the organic nitrate yield
5 increases with carbon number (Atkinson, 2000).
6 The most important of these organic nitrates is PAN, the dominant member of the
7 broader family of peroxyacylnitrates which includes peroxypropionyl nitrate (PPN) of
8 anthropogenic origin and peroxymethacrylic nitrate (MPAN) produced from isoprene oxidation.
9 The PANs are formed by the combination reaction of acetyl peroxy radicals with NO2
1Q CH3C(0)-00 + N02 -> CH3C(0)OON02 (AX2.2-20)
11 where the acetyl peroxy radicals are formed mainly during the oxidation of ethane (^He).
12 Acetaldehyde (CHsCHO) is formed as an intermediate species during the oxidation of ethane.
13 Acetaldehyde can be photolyzed or react with OH radicals to yield acetyl radicals
14 CH3-C(0)H + hv -> CH3-C(0) + H (AX2.2-21)
15 CH3-C(0)H +OH^> CH3-C(0) + H2O (AX2.2-22)
16 Acetyl radicals then react with O2 to yield acetyl peroxy radicals
1? CH3-C(0) + 02 + M^ CH3C(0}-00 + M
18 However, acetyl peroxy radicals will react with NO in areas of high NO concentrations
CH3(CO)-00 + NO -> C//5(C0)-0 + N02
(AX2.2-24)
20 and the acetyl-oxy radicals will then decompose
21 CH3(CO)-0 -> CH3 + C02 (AX2.2-25)
22 Thus, the formation of PAN is favored at conditions of high ratios of NO2 to NO, which are most
23 typically found under low NOx conditions. The PANs both thermally decompose and photolyze
March 2008 AX2-8 DRAFT-DO NOT QUOTE OR CITE
-------
1 back to their reactants on timescales of a few hours during warm sunlit conditions, with lifetimes
2 with respect to thermal decomposition ranging from ~1 hour at 298 K to -2.5 days at 273 K, up
3 to several weeks at 250 K. Thus, they can provide an effective sink of NOX at cold temperatures
4 and high solar zenith angles, allowing release of NC>2 as air masses warm, in particular by
5 subsidence. The PANs are also removed by uptake to vegetation (Sparks et al., 2003;
6 Teklemariam and Sparks, 2004).
7 The organic nitrates may react further, depending on the functionality of the R group, but
8 they will typically not return NOx and can therefore be viewed mainly as a permanent sink for
9 NOx, as alkyl nitrates are sparingly soluble and will photolyze. This sink is usually small
10 compared to HNOs formation, but the formation of isoprene nitrates may be a significant sink for
11 NOX in the United States in summer (Liang et al., 1998).
12 The peroxynitrates produced by AX2.2-19 are thermally unstable and most have very
13 short lifetimes (less than a few minutes) owing to thermal decomposition back to the original
14 reactants. They are thus not effective sinks of NOx.
15
16 AX2.2.2 Nonlinear Relations between Nitrogen Oxide Concentrations and
17 Ozone Formation
18 Ozone is unlike some other species whose rates of formation vary directly with the
19 emissions of their precursors in that Os production (P(Os)) changes nonlinearly with the
20 concentrations of its precursors. At the low NOx concentrations found in most environments,
21 ranging from remote continental areas to rural and suburban areas downwind of urban centers,
22 the net production of Os increases with increasing NOx. At the high NOx concentrations found
23 in downtown metropolitan areas, especially near busy streets and roadways, and in power plant
24 plumes, there is net destruction of O3 by (titration) reaction with NO. Between these two
25 regimes is a transition stage in which O3 shows only a weak dependence on NOx concentrations.
26 In the high NOx regime, NO2 scavenges OH radicals which would otherwise oxidize VOCs to
27 produce peroxy radicals, which in turn would oxidize NO to NO2. In the low NOx regime, VOV
28 (VOC) oxidation generates, or at least does not consume, free radicals, and O3 production varies
29 directly with NOx. Sometimes the terms ' VOC-limited' and 'NOx-limited' are used to describe
30 these two regimes. However, there are difficulties with this usage because: (l)VOC
31 measurements are not as abundant as they are for NOx, (2) rate coefficients for reaction of
32 individual VOCs with free radicals vary over an extremely wide range, and (3) consideration is
March 2008 AX2-9 DRAFT-DO NOT QUOTE OR CITE
-------
1 not given to CO nor to reactions that can produce free radicals without invoking VOCs. The
2 terms NOx-limited and NOx-saturated (used by, e.g., Jaegle et al., 2001) will be used wherever
3 possible to describe these two regimes more adequately. However, the terminology used in
4 original articles will also be kept. The chemistry of OH radicals, which are responsible for
5 initiating the oxidation of hydrocarbons, shows behavior similar to that for Os with respect to
6 NOx concentrations (Hameed et al., 1979; Pinto et al., 1993; Poppe et al., 1993; Zimmerman and
7 Poppe, 1993). These considerations introduce a high degree of uncertainty into attempts to relate
8 changes in Os concentrations to emissions of precursors. It should also be noted at the outset that
9 in a NOx-limited (or NOx-sensitive) regime, Os formation is not insensitive to radical production
10 or the flux of solar UV photons, just that Os formation is more sensitive to NOx. For example,
11 global tropospheric O3 is sensitive to the concentration of CH4 even though the troposphere is
12 predominantly NOx-limited.
13 Various analytical techniques have been proposed that use ambient NOx and VOC
14 measurements to derive information about Os production and Os-NOx-VOC sensitivity.
15 Previously (e.g., National Research Council, 1991), it was suggested that Os formation in
16 individual urban areas could be understood in terms of measurements of ambient NOx and VOC
17 concentrations during the early morning. In this approach, the ratio of summed (unweighted by
18 chemical reactivity) VOC to NOx concentrations is used to determine whether conditions are
19 NOx-sensitive or VOC sensitive. This technique is inadequate to characterize Os formation
20 because it omits many factors recognized as important for P(O3), including: the effect of
21 biogenic VOCs (which are not present in urban centers during early morning); important
22 individual differences in the ability of VOCs to generate free radicals, rather than just from total
23 VOC concentration and other differences in Os-forming potential for individual VOCs (Carter,
24 1995); the effect of multiday transport; and general changes in photochemistry as air moves
25 downwind from urban areas (Milford et al., 1994).
26 Jacob et al. (1995) used a combination of field measurements and a chemical transport
27 model (CTM) to show that the formation of Os changed from NOx-limited to NOx-saturated as
28 the season changed from summer to fall at a monitoring site in Shenandoah National Park, VA.
29 Photochemical production of Os generally occurs together with production of various other
30 species including HNOs, organic nitrates, and hydrogen peroxide (^02). The relative rates of
March 2008 AX2-10 DRAFT-DO NOT QUOTE OR CITE
-------
and the production of other species varies depending on photochemical conditions, and can
2 be used to provide information about (Vprecursor sensitivity.
3 There are no hard and fast rules governing the levels of NOx at which the transition from
4 NOx-limited to NOx-saturated conditions occurs. The transition between these two regimes is
5 highly spatially and temporally dependent. In the upper troposphere, responses to NOx additions
6 from commercial aircraft have been found which are very similar to these in the lower
7 troposphere (Briihl et al., 2000). Briihl et al. (2000) found that the NOX levels for O3 production
8 versus loss are highly sensitive to the radical sources included in model calculations. They found
9 that inclusion of only CH4 and CO oxidation leads to a decrease in net O3 production in the
10 North Atlantic flight corridor due to NO emissions from aircraft. However, the additional
11 inclusion of acetone photolysis was found to shift the maximum in O3 production to higher NOX
12 mixing ratios, thereby reducing or eliminating areas in which O3 production rates decreased due
13 to aircraft emissions.
14 Trainer et al. (1993) suggested that the slope of the regression line between O3 and
15 summed NOx oxidation products (NOz, equal to the difference between measured total reactive
16 nitrogen, NOy, and NOx) can be used to estimate the rate of P(O3) per NOx (also known as the
17 O3 production efficiency, or OPE). Ryerson et al. (1998, 2001) used measured correlations
18 between O3 and NOz to identify different rates of O3 production in plumes from large point
19 sources.
20 Sillman (1995) and Sillman and He (2002) identified several secondary reaction products
21 that show different correlation patterns for NOx-limited conditions and NOx-saturated
22 conditions. The most important correlations are for O3 versus NOy, O3 versus NOz, O3 versus
23 HNO3, and H2O2 versus HNO3. The correlations between O3 and NOy, and O3 and NOz are
24 especially important because measurements of NOy and NOx are widely available. Measured O3
25 versus NOz (Figure AX2.2-2) shows distinctly different patterns in different locations. In rural
26 areas and in urban areas such as Nashville, TN, O3 shows a strong correlation with NOz and a
27 relatively steep slope to the regression line. By contrast, in Los Angeles O3 also increases with
28 NOZ, but the rate of increase of O3 with NOZ is lower and the O3 concentrations for a given NOZ
29 value are generally lower.
March 2008 AX2-11 DRAFT-DO NOT QUOTE OR CITE
-------
Figure AX2.2-2.
x
X
X
X X
X
10
20
NOZ (ppb)
30
x
40
Measured values of Os and NOz (NOy - NOx) during the afternoon at
rural sites in the eastern United States (gray circles) and in urban
areas and urban plumes associated with Nashville, TN (gray dashes),
Paris, FR (black diamonds) and Los Angeles, CA (X's)
1 The difference between NOx-limited and NOx-saturated regimes is also reflected in
2 measurements of H2O2. Formation of H2O2 takes place by self-reaction of photochemically
3 generated HC>2 radicals, so that there is large seasonal variation of H2O2 concentrations, and
4 values in excess of 1 ppb are mainly limited to the summer months when photochemistry is more
5 active (Kleinman, 1991). Hydrogen peroxide is produced in abundance only when 63 is
6 produced under NOx-limited conditions. When the photochemistry is NOx-saturated, much less
7 H2O2 is produced. In addition, increasing NOx tends to slow the formation of H2O2 under NOx-
8 limited conditions. Differences between these two regimes are also related to the preferential
9 formation of sulfate during summer and to the inhibition of sulfate and hydrogen peroxide during
10 winter (Stein and Lamb, 2003). Measurements in the rural eastern United States (Jacob et al.,
11 1995), at Nashville (Sillman et al., 1998), and at Los Angeles (Sakugawa and Kaplan, 1989)
12 show large differences in H2O2 concentrations likely due to differences in NOx availability at
13 these locations.
March 2008
AX2-12
DRAFT-DO NOT QUOTE OR CITE
-------
1 AX2.2.3 Multiphase Chemistry Involving NOX
2 Recent laboratory studies on sulfate and organic aerosols indicate that the reaction
3 probability yN2O5 is in the range of 0.01 to 0.05 (Kane et al., 2001; Hallquist et al., 2003;
4 Thornton et al., 2003). Tie et al. (2003) found that a value of 0.04 in their global model gave the
5 best simulation of observed NOx concentrations over the Arctic in winter.
6 Using aircraft measurements over the northeastern United States, Brown et al. (2006b)
7 found that the uptake coefficient for N2Os, yN2Os, on the surfaces of particles depends strongly
8 on their sulfate content. They found that yN2Os was highest (0.017) in regions where the aerosol
9 sulfate concentration was highest and lower elsewhere (<0.0016). This result contrasts with that
10 of Dentener and Crutzen (1993) who concluded that yN2Os would be independent of aerosol
11 composition, based on a value for yN2Os of 0.1, implying that the heterogeneous hydrolysis of
12 N2O5 would be saturated for typical ambient aerosol surface areas. The importance of this
13 reaction to tropospheric chemistry depends on the value of yN2Os. If it is 0.01 or lower, there
14 may be difficulty in explaining the loss of NOy and the formation of aerosol nitrate, especially
15 during winter. A decrease in N2Os slows down the removal of NOx by leaving more NO2
16 available for reaction and thus increases Os production. Based on the consistency between
17 measurements of NOy partitioning and gas-phase models, Jacob (2000) considers it unlikely that
18 HNOs is recycled to NOx in the lower troposphere in significant concentrations. However, only
19 one of the reviewed studies (Schultz et al., 2000) was conducted in the marine troposphere and
20 none was conducted in the MBL. An investigation over the equatorial Pacific reported
21 discrepancies between observations and theory (Singh et al., 1996) which might be explained by
22 HNOs recycling. It is important to recognize that both Schultz et al. (2000) and Singh et al.
23 (1996) involved aircraft sampling at altitude which, in the MBL, can significantly under-
24 represent sea salt aerosols and thus most total NOs (defined to be HNOs + NOs ) and large
25 fractions of NOy in marine air (e.g., Huebert et al., 1996). Consequently, some caution is
26 warranted when interpreting constituent ratios and NOy budgets based on such data.
27 Recent work in the Arctic has quantified significant photochemical recycling of NOs to
28 NOX and attendant perturbations of OH chemistry in snow (Honrath et al., 2000; Dibb et al.,
29 2002; Domine and Shepson, 2002) which suggest the possibility that similar multiphase
30 pathways could occur in aerosols. As mentioned above, NOs is photolytically reduced to NO2
31 (Zafiriou and True, 1979) in acidic sea salt solutions (Anastasio et al., 1999). Further photolytic
March 2008 AX2-13 DRAFT-DO NOT QUOTE OR CITE
-------
1 reduction of NO2 to NO (Zafiriou and True, 1979) could provide a possible mechanism for
2 HNOs recycling. Early experiments reported production of NOx during the irradiation of
3 artificial seawater concentrates containing NOs (Petriconi and Papee, 1972). Based on the
4 above, HNOs recycling in sea salt aerosols is potentially important and warrants further
5 investigation. Other possible recycling pathways involving highly acidic aerosol solutions and
6 soot are reviewed by Jacob (2000).
7 Stemmler et al. (2006) studied the photosensitized reduction of NO2 to HONO on humic
8 acid films using radiation in the UV-A through the visible spectral regions. They also found
9 evidence for reduction occurring in the dark, reactions which may occur involving surfaces
10 containing partly oxidized aromatic structures. For example, Simpson et al. (2006) found that
11 aromatic compounds constituted -20% of organic films coating windows in downtown Toronto.
12 They calculated production rates of HONO that are compatible with observations of high HONO
13 levels in a variety of environments. The photolysis of HONO formed this way could account for
14 up to 60% of the integrated source of OH radicals in the inner planetary boundary layer. A
15 combination of high NO2 levels and surfaces of soil and buildings and other man-made structures
16 exposed to diesel exhaust would then be conducive to HONO formation and, hence, to high
17 [OH].
18 Ammann et al. (1998) reported the efficient conversion of NO2 to HONO on fresh soot
19 particles in the presence of water. They suggest that interaction between NO2 and soot particles
20 may account for high mixing ratios of HONO observed in urban environments. Conversion of
21 NO2 to HONO and subsequent photolysis and HONO to NO + OH would constitute a NOX"
22 catalyzed Os sink involving snow. High concentrations of HONO can lead to the rapid growth in
23 OH concentrations shortly after sunrise, giving a "jump start" to photochemical smog formation.
24 Prolonged exposure to ambient oxidizing agents appears to deactivate this process. Broske et al.
25 (2003) studied the interaction of NO2 on secondary organic aerosols and concluded that the
26 uptake coefficients were too low for this reaction to be an important source of HONO in the
27 troposphere.
28 Choi and Leu (1998) evaluated the interactions of HNOs on model black carbon soot
29 (FW2), graphite, hexane, and kerosene soot. They found that HNOs decomposed to NO2 and
30 H2O at higher HNOs surface coverages, i.e., P(HNO3) > 10~4 Torr. None of the soot models used
31 were reactive at low HNOs coverages, at P(HNO3) = 5 x 10~7 Torr or at temperatures below 220
March 2008 AX2-14 DRAFT-DO NOT QUOTE OR CITE
-------
1 K. They conclude that it is unlikely that aircraft soot in the upper troposphere/lower stratosphere
2 reduces HNO3.
3 Heterogeneous production on soot at night is believed to be the mechanism by which
4 HONO accumulates to provide an early morning source of HOx in high NOx environments
5 (Harrison et al., 1996; Jacob, 2000). HONO has been frequently observed to accumulate to
6 levels of several ppb overnight, and this has been attributed to soot chemistry (Harris et al., 1982;
7 Calvert et al., 1994; Jacob, 2000).
8 Longfellow et al. (1999) observed the formation of HONO when methane, propane,
9 hexane, and kerosene soots were exposed to NO2. They suggested that this reaction may account
10 for some part of the unexplained high levels of HONO observed in urban areas. They comment
11 that without details about the surface area, porosity, and amount of soot available for this
12 reaction, reactive uptake values cannot be estimated reliably. They comment that soot and NO2
13 are produced in close proximity during combustion, and that large quantities of HONO have
14 been observed in aircraft plumes.
15 Saathoff et al. (2001) studied the heterogeneous loss of NO2, HNO3, NO3/N2O5,
16 HO2/HO2NO2 on soot aerosol using a large aerosol chamber. Reaction periods of up to several
17 days were monitored and results used to fit a detailed model. Saathoff et al. derived reaction
18 probabilities at 294 K and 50% RH for NO2, NO3, HO2, and HO2NO2 deposition to soot; HNO3
19 reduction to NO2; and ^Os hydrolysis. When these probabilities were included in
20 photochemical box model calculations of a 4-day smog event, the only noteworthy influence of
21 soot was a 10% reduction in the second day O3 maximum, for a soot loading of 20 jig nT3, i.e.,
22 roughly a factor of 10 times observed black carbon loadings seen in U.S. urban areas, even
23 during air pollution episodes.
24 Mufioz and Rossi (2002) conducted Knudsen cell studies of HNO3 uptake on black and
25 grey decane soot produced in lean and rich flames, respectively. They observed HONO as the
26 main species released following HNO3 uptake on grey soot, and NO and traces of NO2 from
27 black soot. They conclude that these reactions would only have relevance in special situations in
28 urban settings where soot and HNO3 are present in high concentrations simultaneously.
March 2008 AX2-15 DRAFT-DO NOT QUOTE OR CITE
-------
1 Formation ofNitro PAHs
2 Nitro-polycyclic aromatic hydrocarbons (nitro-PAHs) (see Figure AX2.2-3 for some
3 example nitro-PAHs) are generated from incomplete combustion processes through electrophilic
4 reactions of polycyclic aromatic hydrocarbons (PAHs) in the presence of NC>2 (International
5 Agency for Research on Cancer [IARC], 1989; World Health Organization [WHO], 2003).
6 Among combustion sources, diesel emissions have been identified as the major source of nitro-
7 PAHs in ambient air (Bezabeh et al., 2003; Gibson, 1983; Schuetzle, 1983; Tokiwa and Ohnishi,
8 1986). Direct emissions of NPAHs in PM vary with type of fuel, vehicle maintenance, and
9 ambient conditions (Zielinska et al., 2004).
2-nitronaphthalene 9-nitroanthracene 2-nitrofluoranthene 6-nitrobenzo(a)pyrene
Figure AX2.2-3. Structures of nitro-polycyclic aromatic hydrocarbons.
10 In addition to being directly emitted, nitro-PAHs can also be formed from both gaseous
11 and heterogeneous reactions of PAHs with gaseous nitrogenous pollutants in the atmosphere
12 (Arey et al., 1986, 1989, Arey, 1998; Perrini, 2005; Pitts, 1987; Sasaki et al., 1997; Zielinska
13 et al., 1989). Different isomers of nitro-PAHs are formed through different formation processes.
14 For example, the most abundant nitro-PAH in diesel particles is 1-nitropyene (1NP), followed by
15 3-nitrofluoranthene (3NF) and 8-nitrofluoranthene (8NF) (Bezabeh et al., 2003; Gibson, 1983;
16 Schuetzle, 1983; Tokiwa and Ohnishi, 1986). However, in ambient particulate organic matter
17 (POM), 2-nitrofluoranthene (2NF) is the dominant compound, followed by 1NP and 2-
18 nitropyrene (2NP) (Arey et al., 1989; Bamford et al., 2003; Reisen and Arey, 2005; Zielinska
19 et al., 1989), although 2NF and 2NP are not directly emitted from primary combustion
20 emissions. The reaction mechanisms for the different nitro-PAH formation processes have been
21 well documented and are presented in Figure AX2.2-3.
March 2008
AX2-16
DRAFT-DO NOT QUOTE OR CITE
-------
1 The dominant process for the formation of nitro-PAHs in the atmosphere is gas-phase
2 reaction of PAHs with OH radicals in the presence of NOx (Arey et al., 1986, Arey, 1998;
3 Atkinson and Arey, 1994; Ramdahl et al., 1986; Sasaki et al., 1997). Hydroxyl radicals can be
4 generated photochemically or at night through ozone-alkene reactions, (Finlayson-Pitts and Pitts,
5 2000). The postulated reaction mechanism of OH with PAHs involves the addition of OH at the
6 site of highest electron density of the aromatic ring, for example, the 1-position for pyrene (PY)
7 and the 3-position for fluoranthene (FL). This reaction is followed by the addition of NO2 to the
8 OH-PAH adduct and elimination of water to form the nitroarenes (Figure AX2.2-4) (Arey et al.,
9 1986; Atkinson et al., 1990; Pitts, 1987). After formation, nitro-PAHs with low vapor pressures
10 (such as 2NF and 2NP) immediately migrate to particles under ambient conditions (Fan et al.,
11 1995; Feilberg et al., 1999). The second order rate-constants for the reactions of OH with most
12 PAHs range from 10~10 to 10~12 cm3molecule~y * at 298 K with the yields ranging from -0.06 to
13 -5% (Atkinson and Arey, 1994). 2NF and 2NP have been found as the most abundant nitro-
14 PAHs formed via reactions of OH with gaseous PY and FL, respectively in ambient air.
Figure AX2.2-4.
OH
Hv OH
4NO;
H, OH
2NP
Formation of 2-nitropyrene (2NP) from the reaction of OH with
gaseous pyrene (PY).
March 2008
AX2-17
DRAFT-DO NOT QUOTE OR CITE
-------
1 The second important process for the formation of nitro-PAHs in the atmosphere is the
2 nitration of PAHs by NOs in the presence of NOx at night (Atkinson et al., 1990; Atkinson and
3 Arey, 1994; Sasaki et al., 1997). Nitrate radicals can be generated by reaction of ozone (O3) with
4 NC>2 in the atmosphere by Reaction AX2.2-26
5 3
6 Similar to the mechanism of OH reactions with PAHs, NOs initially adds to the PAH ring
7 to form an NOs-PAH adduct, followed by loss of HNOs to form nitro-PAHs (Atkinson et al.,
8 1990; Atkinson and Arey, 1994; Sasaki et al., 1997). For example, in the mixture of naphthalene
9 and ^Os-NOs-NCh, the major products formed through the NOs reaction are 1- and 2-nitro-
10 naphthalene (INN and 2NN) (Atkinson et al., 1990; Feilberg et al., 1999; Sasaki et al., 1997).
1 1 2NF and 4NP were reported as the primary products of the gas-phase reactions of FL and PY
12 with NOs radical, respectively (Atkinson et al., 1990; Atkinson and Arey, 1994).
13 The reaction with NOs is of minor importance in the daytime because NOs radical is not
14 stable in sunlight. In addition, given the rapid reactions of NO with NOs and with Os in the
15 atmosphere (Finlayson-Pitts and Pitts 2000), concentrations of NOs at ground level are low
16 during daytime. However, at night, concentrations of NOs radicals formed in polluted ambient
17 air are expected to increase. According to Atkinson (1991), the average NOs concentration is
18 about 20 ppt in the lower troposphere at night and can be as high as 430 ppt. It is also worth
19 noting that significant NOs radical concentrations are found at elevated altitudes where Os is
20 high but NO is low (Reissell and Arey, 2001; Stutz et al., 2004a). When NO3 reaches high
21 concentrations, the formation of nitro-PAHs by the reaction of gaseous PAHs with NOs may be
22 of environmental significance. At 10~17 - 10~18 cm3 molecule V1, the rate constants of NOs
23 with most PAHs are several orders of magnitude lower than those of OH with the same PAHs;
24 however, the yields of nitro-PAHs from NOs reactions are generally much higher than those of
25 OH reactions. For example, the yields of 1-NN and 2NF are 0.3% and 3%, respectively from
26 OH reactions, but the yields are 17% and 24% for these two compounds generated from the NO3
27 radical reactions (Atkinson and Arey, 1994). Therefore, formation of nitro-PAHs via reactions
28 of NOs at nighttime under certain circumstances can be significant.
29 The third process of nitro-PAH formation in the atmosphere is nitration of PAHs by
30 NO2/N2Os in the presence of trace amounts of HNOs (HNOs) in both gas and particle phases.
March 2008 AX2- 1 8 DRAFT-DO NOT QUOTE OR CITE
-------
1 This mechanism could be operative throughout the day and night (Pitts, 1983, 1985a,b; Grosjean
2 et al., 1983; Ramdahl et al., 1984; Kamens et al., 1990). The formation of nitro-fluoranthenes
3 was observed when adsorbed FL was exposed to gaseous N2O5, and the distribution of product
4 NF isomers was 3- > 8- > 7- > 1- NF (Pitts et al., 1985a,b). The proposed mechanism for this
5 reaction was an ionic electrophilic nitration by nitronium ion (NC>2+). It was speculated that
6 N2Os became ionized prior to the reaction with FL (Zielinska et al., 1986). Only 1NP was
7 observed for the reaction of PY with ^Os on filters (Pitts et al., 1985b). Compared to the
8 reactions of OH and NOs, nitration of PAHs by NO2/N2Os is less important.
9 Measurements of nitro-PAHs in ambient air provide evidence for the proposed reaction
10 mechanism, i.e. the reactions of OH and NOs radicals with PAHs are the major sources of
11 nitro-PAHs (Bamford and Baker, 2003; Reisen and Arey, 2005; and references therein). 2NF is
12 a ubiquitous component of ambient POM, much higher than 1NP, itself a marker of combustion
13 sources. Nitro-PAH isomer ratios show strong seasonality. For instance, the mean ratios of
14 2NF/1NP were higher in summer than in winter (Bamford et al., 2003; Reisen and Arey, 2005),
15 indicating that reactions of OH and NOs with FL are the major sources of nitro-PAHs in ambient
16 air in summer. The ratio of 2NF/1NP was lower in winter than in summer because of lower OH
17 concentrations and, therefore, less production of 2NF via atmospheric reactions. A ratio of
18 1NP/2NF greater than 1 was observed in locations with major contributions from vehicle
19 emissions (Dimashki et al., 2000; Feilberg et al., 2001). In addition, the ratio of 2NF/2NP was
20 also used to evaluate the contribution of OH and NO3 initiated reactions to the ambient nitro-
21 PAHs (Bamford et al., 2003; Reisen and Arey, 2005).
22 The concentrations for most nitro-PAHs found in ambient air are much lower than
23 1 pg/m3, except NNs, 1NP, and 2NF, which can be present at several pg/m3. These levels are
24 much lower (~2 to -1000 times lower) than their parent PAHs. However, nitro-PAHs are much
25 more toxic than PAHs (Durant et al., 1996; Grosovsky et al., 1999; Salmeen et al., 1982; Tokiwa
26 et al., 1998; Tokiwa and Ohnishi, 1986). Moreover, most nitro-PAHs are present in particles
27 with a mass median diameter <0.1 |im.
28 Esteve et al. (2006) examined the reaction of gas-phase NO2 and OH radicals with
29 various PAHs adsorbed onto model diesel particulate matter (SRM 1650a, NIST). Using pseudo
30 second order rate coefficients, they derived lifetimes for conversion of the particle-bound PAHs
31 to nitro-PAHs of a few days (for typical urban NO2 levels of 20 ppb). They also found that the
March 2008 AX2-19 DRAFT-DO NOT QUOTE OR CITE
-------
1 rates of reaction of OH with the PAHs were about four orders of magnitude larger than for the
2 reactions involving NC>2. However, since the concentrations of NC>2 used above are more than
3 four orders of magnitude larger than those for OH (106-107/cm3), they concluded that the
4 pathway involving NO2 is expected to be favored over that involving OH radicals. Consistent
5 with the importance of the gas-phase formation of NPAHS, both the mutagenic potency of PM
6 and the content of NPAHs in PM vary by particle size, and are higher in the submicron size
7 range (Xu and Lee, 2000; Kawanaka et al., 2004).
8 The major loss process of nitro-PAHs is photodecomposition (Fan et al., 1996; Feilberg
9 et al., 1999; Feilberg and Nielsen, 2001), with lifetimes on the order of hours. However, lacking
10 direct UV light sources indoors, nitro-PAHs are expected have a longer lifetimes (days) indoors
11 than outdoors; and may therefore pose increased health risks. Many nitro-PAHs are semi- or
12 nonvolatile organic compounds. As stated above, indoor environments have much greater
13 surface areas than outdoors. Thus, it is expected that gas/particle distribution of nitro-PAHs
14 indoors will be different from those in ambient air. A significant portion of nitro-PAHs will
15 probably be adsorbed by indoor surfaces, such as carpets, leading to different potential exposure
16 pathways to nitro-PAHs in indoor environments. The special characteristics of indoor
17 environments, which can affect the indoor chemistry and potential exposure pathways
18 significantly, should be taken into consideration when conducting exposure studies of nitro-
19 PAHs.
20 Reaction with OH and NO3 radicals is a major mechanism for removing gas-phase PAHs,
21 with OH radical initiated reactions predominating depending on season (Vione et al., 2004;
22 Bamford et al., 2003). Particle-bound PAH reactions occur but tend to be slower.
23 Nitronaphthalenes tend to remain in the vapor phase, but because phase partitioning depends on
24 ambient temperature, in very cold regions these species can condense (Castells et al., 2003)
25 while the higher molecular weight PAHs such as the nitroanthracenes, nitrophenantrenes and
26 nitrofluoranthenes condense in and on PM (Ciganek et al., 2004; Cecinato, 2003).
27
28 Multiphase Chemical Processes Involving Nitrogen Oxides and Halogens
29 Four decades of observational data on Os in the troposphere have revealed numerous
30 anomalies not easily explained by gas-phase HOX-NOX photochemistry. The best-known
31 example is the dramatic depletion of ground-level Os during polar sunrise due to multiphase
32 catalytic cycles involving inorganic Br and Cl radicals (Barrie et al., 1988; Martinez et al., 1999;
March 2008 AX2-20 DRAFT-DO NOT QUOTE OR CITE
-------
1 Foster et al., 2001). Other examples of anomalies in tropospheric 63 at lower latitudes include
2 low levels of 63 (<10 ppbv) in the marine boundary layer (MBL) and overlying free troposphere
3 (FT) at times over large portions of the tropical Pacific (Kley et al., 1996), as well as post-sunrise
4 Os depletions over the western subtropical Pacific Ocean (Nagao et al., 1999), the temperate
5 Southern Ocean (Galbally et al., 2000), and the tropical Indian Ocean (Dickerson et al., 1999).
6 The observed Os depletions in near-surface marine air are generally consistent with the model-
7 predicted volatilization of Br2, BrCl, and C^ from sea salt aerosols through autocatalytic halogen
8 "activation" mechanisms (e.g., Vogt et al., 1996; Von Glasow et al., 2002a) involving these
9 aqueous phase reactions.
1Q HOBr + Br- + H- -> Br2 + H2O
1 1 HOCL + Bf + H+ -» BrCl + H2O (AX2.2-28)
12 HOCl + Cr + H+^C!2+H20 (AX22-29)
13 In polluted marine regions at night, the heterogeneous reaction
14 N20S + CI-->CIN02 + NOS-
15 may also be important (Finlayson-Pitts et al., 1989; Behnke et al., 1997; Erickson et al., 1999).
16 Diatomic bromine, BrCl, Cb, and C1NO2 volatilize and photolyze in sunlight to produce atomic
17 Br and Cl. The acidification of sea salt aerosol via incorporation of HNOs (and other acids)
18 leads to the volatilization of HC1 (Erickson et al., 1999), e.g.
19 — ~., , ~, , ,,v^ ~v.f (AX2.2-31)
20 and the corresponding shift in phase partitioning can accelerate the deposition flux to the surface
21 of total NO3 (Russell et al., 2003; Fischer et al., 2006). However, Pryor and Sorensen (2000)
22 have shown that the dominant form of nitrate deposition is a complex function of wind speed. In
23 polluted coastal regions where HC1 from Reaction 35 often exceeds 1 ppbv, significant
24 additional atomic Cl~ is produced via
March 2008 AX2-21 DRAFT-DO NOT QUOTE OR CITE
-------
, we/ + o» - (AX22.32)
2 (Singh and Kasting, 1988; Keene et al., 2007). Following production, Br and Cl atoms
3 catalytically destroy Os via
(AX2.2-33)
7 where (X = Br and Cl).
8 Formation of Br and Cl nitrates via
9 (AX2M6)
10 and the subsequent reaction of XNO3 with sea salt and sulfate aerosols via
1 1 XN03 + H20 -» HOX + H+ + NOf
12 and
13 ""-,, ,,Vj (AX2.2-38)
14 (where Y = Cl, Br, or I) accelerates the conversion of NOX to particulate NO3 and thereby
15 contributes indirectly to net O3 destruction (Sander et al., 1999; Vogt et al., 1999, Pszenny et al.,
16 2004). Most XNO3 reacts via Reaction AX2.2-38 on sea salt whereas reaction 33 is more
17 important on sulfate aerosols. Partitioning of HC1 on sulfate aerosols following Henry's Law
18 provides Cl" for Reaction AX2.2-38 to form BrCl. Product NO3 from both Reactions AX2.2-37
19 and AX2.2-38 partitions with the gas-phase HNO3 following Henry's Law. Because most
20 aerosol size fractions in the MBL are near equilibrium with respect to HNO3 (Erickson et al.,
21 1999; Keene et al., 2004), both sulfate and sea salt aerosol can sustain the catalytic removal of
22 NOx and re-activation of Cl and Br with no detectable change in composition. The photolytic
23 reduction of NO3 in sea salt aerosol solutions recycles NOx to the gas phase (Pszenny et al.,
March 2008 AX2-22 DRAFT-DO NOT QUOTE OR CITE
-------
1 2004). Halogen chemistry also impacts 63 indirectly by altering OH/HO2 ratios (XO + HC>2 —»•
2 HOX + O2 -> OH + X) (e.g., Stutz et al., 1999; Bloss et al., 2005).
3 In addition to O3 destruction via Reaction AX2.3-3, atomic Cl oxidizes hydrocarbons
4 (HCs) primarily via hydrogen abstraction to form HC1 vapor and organz products (Jobson et al.,
5 1994; Pszenny et al., 2006). The enhanced supply of odd-H radicals from HC oxidation leads to
6 net Os production in the presence of sufficient NOx (Pszenny et al., 1993). Available evidence
7 suggests that Cl~ radical chemistry may be a significant net source for 63 in polluted
8 coastal/urban air (e.g., Tanaka et al., 2003; Finley and Saltzman, 2006).
9 An analogous autocatalyic 63 destruction cycle involving multiphase iodine chemistry
10 also operates in the marine atmosphere (Alicke et al., 1999, Vogt et al., 1999; McFiggans et al.,
11 2000; Ashworth et al., 2002). In this case, the primary source of I is believed to be either
12 photolysis of CH^, other I-containing gases (Carpenter et al., 1999; Carpenter, 2003), and/or
13 perhaps I2 (McFiggans et al., 2004; Saiz-Lopez and Plane, 2004; McFiggans, 2005) emitted by
14 micro-and macro flora. Sea salt and sulfate aerosols provide substrates for multiphase reactions
15 that sustain the catalytic I-IO cycle. The IO radical has been measured by long-path (LP) and/or
16 multi axis (MAX) differential optical absorption spectroscopy (DOAS) at Mace Head, Ireland;
17 Tenerife, Canary Islands; Cape Grim, Tasmania; and coastal New England, USA; having
18 average daytime levels of about 1 ppt with maxima up to 7 ppt (e.g., Allan et al., 2000; Pikelnaya
19 et al., 2006). Modeling suggests that up to 13% per day of O3 in marine air may be destroyed via
20 multiphase iodine chemistry (McFiggans et al., 2000). The reaction of IO with NO2 followed by
21 uptake of INOs into aerosols (analogous to Reactions AX2.2-12 through AX2.2-14) accelerates
22 the conversion of NOx to particulate MV and thereby contributes to net Os destruction. The
23 reaction IO + NO —»• I + NO2 also influences NOX cycling.
24 Most of the above studies have focused on halogen-radical chemistry and related
25 influences on NOx cycling in coastal and urban air. However, available evidence suggests that
26 similar chemical transformations proceed in other halogen-rich tropospheric regimes. For
27 example, Cl, Br, and/or I oxides have been measured at significant concentrations in near-surface
28 air over the Dead Sea, Israel, the Great Salt Lake, Utah (e.g., Hebestreit et al., 1999; Stutz et al.,
29 1999, 2002; Zingler and Platt, 2005), and the Salar de Uyuni salt pan in the Andes mountains
30 (U. Platt, unpublished data, 2006); high column densities of halogenated compounds have also
31 been observed from satellites over the northern Caspian Sea (Wagner et al., 2001; Hollwedel
March 2008 AX2-23 DRAFT-DO NOT QUOTE OR CITE
-------
1 et al., 2004). The primary source of reactive halogens in these regions is thought to be from
2 activation along the lives of that in Reactions AX2.2-27 through AX2.2-29 involving
3 concentrated salt deposits on surface evaporite pans. High concentrations of BrO have also been
4 measured in volcanic plumes (Bobrowski et al., 2003, Gerlach, 2004). Although virtually
5 unexplored, the substantial emissions of inorganic halogens during biomass burning (Lobert
6 et al., 1999; Keene et al., 2006) and in association with crustal dust (Keene et al., 1999; Sander
7 et al., 2003) may also support active halogen-radical chemistry and related transformations
8 involving NOx downwind of sources. Finally, observations from satellites, balloons, and aircraft
9 indicate that BrO is present in the free troposphere at levels sufficient to significantly influence
10 photochemistry (e.g., Von Glasow et al., 2004).
11
12
13 AX2.3 CHEMISTRY OF SULFUR OXIDES IN THE TROPOSPHERE
14 The four known monomeric sulfur oxides are sulfur monoxide (SO), sulfur dioxide
15 (802), sulfur tri oxide (SOs), and disulfur monoxide (82©). SO can be formed by photolysis of
16 SO2 at wavelengths less than 220 nm, and so could only be found in the middle and upper
17 stratosphere (Pinto et al., 1989). SOs can be emitted from the stacks of power plants and
18 factories however, it reacts extremely rapidly with H2O in the stacks or immediately after release
19 into the atmosphere to form H^SO/t. Of the four species, only SO2 is present at concentrations
20 significant for atmospheric chemistry and human exposures.
21 Sulfur dioxide can be oxidized either in the gas phase, or, because it is soluble, in the aqueous
22 phase in cloud drops. The gas-phase oxidation of SO2 proceeds through the reaction
23 SO 2 + OH+M^ HSO3 + M 2 3
24 followed by
25
26 ^ + H20 -> H2S04 (AX2.3-3)
27 Since H2SO4 is extremely soluble, it will be removed rapidly by transfer to the aqueous phase of
28 aerosol particles and cloud drops. Rate coefficients for reaction of SO2 with HO2 or NO3 are too
29 low to be significant (JPL, 2003).
March 2008 AX2-24 DRAFT-DO NOT QUOTE OR CITE
-------
1 SO2 is chiefly but not exclusively primary in origin; it is also produced by the
2 photochemical oxidation of reduced sulfur compounds such as dimethyl sulfide (CH3-S-CH3),
3 hydrogen sulfide (H2S), carbon disulfide (CS2), carbonyl sulfide (OCS), methyl mercaptan
4 (CH3-S-H), and dimethyl disulfide (CH3-S-S-CH3) which are all mainly biogenic in origin.
5 Their sources are discussed in Section AX2.5. Table AX2.3-1 lists the atmospheric lifetimes of
6 reduced sulfur species with respect to reaction with various oxidants. Except for OCS, which is
7 lost mainly by photolysis (x~6 months), all of these species are lost mainly by reaction with OH
8 and NO3 radicals. Because OCS is relatively long-lived in the troposphere, it can be transported
9 upwards into the stratosphere. Crutzen (1976) proposed that its oxidation serves as the major
10 source of sulfate in the stratospheric aerosol layer sometimes referred to the "Junge layer,"
11 (Junge et al., 1961) during periods when volcanic plumes do not reach the stratosphere.
12 However, the flux of OCS into the stratosphere is probably not sufficient to maintain this
13 stratospheric aerosol layer. Myhre et al. (2004) propose instead that SO2 transported upwards
14 from the troposphere is the most likely source, have become the upward flux of OCS is too small
15 to sustain observed sulfate loadings in the Junge layer. In addition, insitu measurements of the
16 isotopic composition of sulfur do not match those of OCS (Leung et al., 2002). Reaction with
17 NO3 radicals at night most likely represents the major loss process for dimethyl sulfide and
18 methyl mercaptan. The mechanisms for the oxidation of DMS are still not completely
19 understood. Initial attack by NO3 and OH radicals involves H atom abstraction, with a smaller
20 branch leading to OH addition to the S atom. The OH addition branch increases in importance as
21 temperatures decrease and becoming the major pathway below temperatures of 285 K
22 (Ravishankara, 1997). The adduct may either decompose to form methane sulfonic acid (MSA),
23 or undergo further reactions in the main pathway, to yield dimethyl sulfoxide (Barnes et al.,
24 1991). Following H atom abstraction from DMS, the main reaction products include MSA and
25 SO2. The ratio of MSA to SO2 is strongly temperature dependent, varying from about 0.1 in
26 tropical waters to about 0.4 in Antarctic waters (Seinfeld and Pandis, 1998). Excess sulfate (over
27 that expected from the sulfate in seawater) in marine aerosol is related mainly to the production
28 of SO2 from the oxidation of DMS. Transformations among atmospheric sulfur compounds are
29 summarized in Figure AX2.3-1.
March 2008 AX2-25 DRAFT-DO NOT QUOTE OR CITE
-------
DCS
Av, O
so
OH
Tropopause
Figure AX2.3-1. Transformations of sulfur compounds in the atmosphere.
Source: Adapted from Berresheim et al. (1995).
1 Multiphase Chemical Processes Involving Sulfur Oxides and Halogens
2 Chemical transformations involving inorganic halogenated compounds effect changes in
3 the multiphase cycling of sulfur oxides in ways analogous to their effects on NOx. Oxidation of
4 dimethylsulfide (CH3)2S by BrO produces dimethyl sulfoxide (CH3)2SO (Barnes et al., 1991;
5 Toumi, 1994), and oxidation by atomic chloride leads to formation of SO2 (Keene et al., 1996).
6 (CH3)2SO and SO2 are precursors for methanesulfonic acid (CH3SO3H) and H2SO4. In the MBL,
7 virtually all H2SO4 and CHsSOsH vapor condenses onto existing aerosols or cloud droplet, which
8 subsequently evaporate, thereby contributing to aerosol growth and acidification. Unlike
9 CHsSOsH, H2SO4 also has the potential to produce new particles (Korhonen et al., 1999;
March 2008
AX2-26
DRAFT-DO NOT QUOTE OR CITE
-------
1 Kulmala et al., 2000), which in marine regions is thought to occur primarily in the free
2 troposphere. Saiz-Lopez et al. (2004) estimated that observed levels of BrO at Mace Head
3 would oxidize (CH3)2S about six times faster than OH and thereby substantially increase
4 production rates of H2SO4 and other condensible S species in the MBL. Sulfur dioxide is also
5 scavenged by deliquesced aerosols and oxidized to H2SO4 in the aqueous phase by several
6 strongly pH-dependent pathways (Chameides and Stelson, 1992; Vogt et al., 1996; Keene et al.,
7 1998). Model calculations indicate that oxidation of S(IV) by O3 dominates in fresh, alkaline sea
8 salt aerosols, whereas oxidation by hypohalous acids (primarily HOC1) dominates in moderately
9 acidic solutions. Additional paniculate non-sea salt (nss) SO42 is generated by SO2 oxidation in
10 cloud droplets (Clegg and Toumi, 1998). Ion-balance calculations indicate that most nss SC>42
11 in short-lived (two to 48 hours) sea salt size fractions accumulates in acidic aerosol solutions
12 and/or in acidic aerosols processed through clouds (e.g., Keene et al., 2004). The production,
13 cycling, and associated radiative effects of S-containing aerosols in marine and coastal air are
14 regulated in part by chemical transformations involving inorganic halogens (Von Glasow et al.,
15 2002b). These transformations include: dry-deposition fluxes of nss SC>42 in marine air
16 dominated, naturally, by the sea salt size fractions (Huebert et al., 1996; Turekian et al., 2001);
17 HC1 phase partitioning that regulates sea salt pH and associated pH-dependent pathways for
18 S(IV) oxidation (Keene et al., 2002; Pszenny et al., 2004); and potentially important oxidative
19 reactions with reactive halogens for (CH3)2S and S(IV). However, both the absolute magnitudes
20 and relative importance of these processes in MBL S cycling are poorly understood.
21 Iodine chemistry has been linked to ultrafme particle bursts at Mace Head (O'Dowd
22 et al., 1999, 2002). Observed bursts coincide with the elevated concentrations of IO and are
23 characterized by particle concentrations increasing from background levels to up to
24 300,000 cm"3 on a time scale of seconds to minutes. This newly identified source of marine
25 aerosol would provide additional aerosol surface area for condensation of sulfur oxides and
26 thereby presumably diminish the potential for nucleation pathways involving H2SO4. However,
27 a subsequent investigation in polluted air along the New England, USA coast found no
28 correlation between periods of nanoparticle growth and corresponding concentrations of I oxides
29 (Russell et al., 2006). The potential importance of I chemistry in aerosol nucleation and its
30 associated influence on sulfur cycling remain highly uncertain.
March 2008 AX2-27 DRAFT-DO NOT QUOTE OR CITE
-------
1 AX2.4 MECHANISMS FOR THE AQUEOUS PHASE FORMATION OF
2 SULFATE AND NITRATE
3 The major species containing sulfur in clouds are HSCV and SOs2 , which are derived
4 from the dissolution of 862 in water and are referred to as S(IV); and HSCV and SC>42 , which
5 are referred to as S(VI). The major species capable of oxidizing S(IV) to S(VI) in cloud water
6 are 63, peroxides (either H2O2 or organic peroxides), OH radicals, and ions of transition metals
7 such as Fe and Cu that can catalyze the oxidation of S(IV) to S(VI) by 62.
8 The basic mechanism of the aqueous phase oxidation of SO2 has long been studied and
9 can be found in numerous texts on atmospheric chemistry, e.g., Seinfeld and Pandis (1998),
10 Jacob (2000), and Jacobson (2002). The steps involved in the aqueous phase oxidation of SC>2
1 1 can be summarized as follows (Jacobson, 2002):
12 Dissolution of SO2
13 S°2(g)^S02(aq} (AX2.4-1)
14 The formation and dissociation of H2SO3
15 aq}(aq (AX2.4-2)
16 In the pH range commonly found in rainwater (2 to 6), the most important reaction converting
17 S(IV) to S(VI) is
18
19 as SOs2 is much less abundant than
+H20+2H+ (AX2.4-3)
20 Major pathways for the aqueous phase oxidation of S(IV) to S(VI) as a function of pH are
21 shown in Figure AX2.4-1 . For pH up to about 5.3, H2O2 is seen to be the dominant oxidant;
22 above 5.3, Os, followed by Fe(III) becomes dominant. Higher pHs are expected to be found
23 mainly in marine aerosols. However, in marine aerosols, the chloride-catalyzed oxidation of
24 S(IV) may be more important (Zhang and Millero, 1991; Hoppel and Caffrey, 2005). Because
25 NH4+ is so effective in controlling acidity, it affects the rate of oxidation of S(IV) to S(VI) and
26 the rate of dissolution of 862 in particles and cloud drops.
March 2008 AX2-28 DRAFT-DO NOT QUOTE OR CITE
-------
1 Nitrogen dioxide is also taken up in cloud drops and can be oxidized to NO3 , although it
2 is much less soluble than 862 and this pathway is of minor importance. Instead, the uptake of
3 more highly soluble nitrogen-containing acids initiates aqueous-phase chemistry of NO3
4 formation.
5 Warneck (1999) constructed a box model describing the chemistry of the oxidation of
6 SC>2 and NO2 including the interactions of N and S species and minor processes in sunlit cumulus
7 clouds. The relative contributions of different reactions to the oxidation of S(IV) species to
8 S(VI) and NC>2 to NO3 10 minutes after cloud formation are given in Tables AX2.4-la and
9 AX2.4-lb. The two columns show the relative contributions with and without transition metal
10 ions. As can be seen from Table AX2.4-la, 862 within a cloud (gas + cloud drops) is oxidized
11 mainly by H2O2 in the aqueous phase, while and the gas-phase oxidation by OH radicals is small
12 by comparison. A much smaller contribution in the aqueous phase is made by methyl
13 hydroperoxide (CH3OOH) because it is formed mainly in the gas phase and its Henry's Law
14 constant is several orders of magnitude smaller that of H2O2. After H2O2, HNO4 is the major
15 contributor to S(IV) oxidation. The contribution from the gas phase oxidation of SC>2 to be small
16 by comparison to the aqueous -phase reactions given above.
17 In contrast to the oxidation of 862, Table AX2.4-lb shows that the oxidation of NC>2
18 occurs mainly in the gas phase within clouds, implying that the gas phase oxidation of NC>2 by
19 OH radicals predominates. Clouds occupy about 15%, on average, of the volume of the
20 troposphere.
21 The values shown in Tables AX2.4-la and AX2.4-lb indicate that only about 20% of
22 SO2 is oxidized in the gas phase, but about 90% of NO2 is oxidized in the gas phase. Thus, SO2
23 is oxidized mainly by aqueous-phase reactions, but NO2 is oxidized mainly by gas phase
24 reactions.
25
26 Multiphase Chemical Processes Involving Sulfur Oxides and Ammonia
27 The phase partitioning of NH3 with deliquesced aerosol solutions is controlled primarily
28 by the thermodynamic properties of the system expressed as follows:
9Q jg I _KK/J L""4 J w'if* j TAX2 4-4^
March 2008 AX2-29 DRAFT-DO NOT QUOTE OR CITE
-------
Figure AX2.4-1.
10'6
T) 1Q-10
s
s"
zz io-12
io-14
10
,-16
10
,-18
I I
J I
I
0123456
PH
Comparison of aqueous-phase oxidation paths. The rate of
conversion of S(IV) to S(VI) is shown as a function of pH. Conditions
assumed are: [SO2(g)] = 5 ppb; [NO2(g)] = 1 ppb; [H2O2(g)] = 1 ppb;
[03(g)] = 50 ppb; [Fe(III)(aq)] = 0.3 uM; [Mn(II)(aq)] = 0.3 uM.
Source: Seinfeld and Pandis (1998).
1 where KH and Kb are the temperature-dependent Henry's Law and dissociation constants
2 (62 M atnT1) (1.8 x 1CT5 M), respectively, for NH3, and Kw is the ion product of water (1.0 x
3 10~14 M) (Chameides, 1984). It is evident that for a given amount of NHX (NHs + particulate
4 NH4+) in the system, increasing aqueous concentrations of particulate H+ will shift the
5 partitioning of NH3 towards the condensed phase. Consequently, under the more polluted
6 conditions characterized by higher concentrations of acidic sulfate aerosol, ratios of gaseous
7 to particulate NH4+ decrease (Smith et al., 2007). It also follows that in marine air, where
March 2008
AX2-30
DRAFT-DO NOT QUOTE OR CITE
-------
1 aerosol acidity varies substantially as a function of particle size, NH3 partitions preferentially to
2 the more acidic sub-jam size fractions (e.g., Keene et al., 2004; Smith et al., 2007).
3 Because the dry-deposition velocity of gaseous NH3 to the surface is substantially greater
4 than that for the sub-|im, sulfate aerosol size factions with which most particulate NH4+ is
5 associated, dry-deposition fluxes of total NH3 are dominated by the gas phase fraction (Russell
6 et al., 2003; Smith et al., 2007). Consequently, partitioning with highly acidic sulfate aerosols
7 effectively increases the atmospheric lifetime of total NH3 against dry deposition. This shift has
8 important consequences for NH3 cycling and potential ecological effects. In coastal New
9 England during summer, air transported from rural eastern Canada contains relatively low
10 concentrations of particulate non-sea salt (nss) SC>42 and total NH3 (Smith et al., 2007). Under
11 these conditions, the roughly equal partitioning of total NH3 between the gas and particulate
12 phases sustains substantial dry-deposition fluxes of total NH3 to the coastal ocean (median of
13 10.7 (imol nT2 day l). In contrast, heavily polluted air transported from the industrialized
14 midwestern United States contains concentrations of nss SC>42 and total NH3 that are, about a
15 factory of 3 greater, based on median values. Under these conditions, most total NH3 (>85%)
16 partitions to the highly acidic sulfate aerosol size fractions and, consequently, the median dry-
17 deposition flux of total NH3 is 30% lower than that under the cleaner northerly flow regime. The
18 relatively longer atmospheric lifetime of total NH3 against dry deposition under more polluted
19 conditions implies that, on average, total NH3 would accumulate to higher atmospheric
20 concentrations under these conditions and also be subject to atmospheric transport over longer
21 distances. Consequently, the importance NHx of removal via wet deposition would also
22 increase. Because of the inherently sporadic character of precipitation, we might expect by
23 greater heterogeneity in NH3 deposition fields and any potential responses by sensitive
24 ecosystems downwind of major S-emission regions.
25
26
27 AX2.5 TRANSPORT OF NITROGEN AND SULFUR OXIDES IN
28 THE ATMOSPHERE
29 Major episodes of high O3 concentrations in the eastern United Sates and in Europe are
30 associated with slow moving high-pressure systems. High-pressure systems during the warmer
31 seasons are associated with subsidence, resulting in warm, generally cloudless conditions with
32 light winds. The subsidence results in stable conditions near the surface, which inhibit or reduce
March 2008 AX2-31 DRAFT-DO NOT QUOTE OR CITE
-------
1 the vertical mixing of 63 precursors (NOx, VOCs, and CO). Photochemical activity is enhanced
2 because of higher temperatures and the availability of sunlight. However, it is becoming
3 increasingly apparent that transport of O3 and NOX and VOC from distant sources can provide
4 significant contributions to local [Os] even in areas where there is substantial photochemical
5 production. There are a number of transport phenomena occurring either in the upper boundary
6 layer or in the free troposphere which can contribute to high Os values at the surface. These
7 phenomena include stratospheric-tropospheric exchange (STE), deep and shallow convection,
8 low-level jets, and the so-called "conveyor belts" that serve to characterize flows around frontal
9 systems.
10
11 Convective Transport
12 Crutzen and Gidel (1983), Gidel (1983), and Chatfield and Crutzen (1984) hypothesized
13 that convective clouds played an important role in rapid atmospheric vertical transport of trace
14 species and first tested simple parameterizations of convective transport in atmospheric chemical
15 models. At nearly the same time, evidence was shown of venting the boundary layer by shallow,
16 fair weather cumulus clouds (e.g., Greenhut et al., 1984; Greenhut, 1986). Field experiments
17 were conducted in 1985 which resulted in verification of the hypothesis that deep convective
18 clouds are instrumental in atmospheric transport of trace constituents (Dickerson et al., 1987).
19 Once pollutants are lofted to the middle and upper troposphere, they typically have a much
20 longer chemical lifetime and with the generally stronger winds at these altitudes, they can be
21 transported large distances from their source regions. Transport of NOx from the boundary layer
22 to the upper troposphere by convection tends to dilute the higher in the boundary layer
23 concentrations and extend the NOx lifetime from less than 24 hours to several days.
24 Photochemical reactions occur during this long-range transport. Pickering et al. (1990)
25 demonstrated that venting of boundary layer NOx by convective clouds (both shallow and deep)
26 causes enhanced Os production in the free troposphere. The dilution of NOx at the surface can
27 often increase Os production efficiency. Therefore, convection aids in the transformation of
28 local pollution into a contribution to global atmospheric pollution. Downdrafts within
29 thunderstorms tend to bring air with less NOx from the middle troposphere into the boundary
30 layer. Lightning produces NO which is directly injected chiefly into the middle and upper
31 troposphere. The total global production of NO by lightning remains uncertain, but is on the
3 2 order of 10% of the total.
March 2008 AX2-32 DRAFT-DO NOT QUOTE OR CITE
-------
1 Observations of the Effects ofConvective Transport
2 The first unequivocal observations of deep convective transport of boundary layer
3 pollutants to the upper troposphere were documented by Dickerson et al. (1987).
4 Instrumentation aboard three research aircraft measured CO, Os, NO, NOx, NOy, and
5 hydrocarbons in the vicinity of an active mesoscale convective system near the
6 Oklahoma/Arkansas border during the 1985 PRE-STORM experiment. Anvil penetrations about
7 two hours after maturity found greatly enhanced mixing ratios inside the cloud of all of the
8 aforementioned species compared with outside it. Nitric oxide mixing ratios in the anvil
9 averaged 3 to 4 ppbv, with individual 3-min observations reaching 6 ppbv; boundary layer NOx
10 was typically 1.5 ppbv or less outside the cloud. Therefore, the anvil observations represent a
11 mixture of boundary layer NOx and NOx contributed by lightning. Luke et al. (1992)
12 summarized the air chemistry data from all 18 flights during PRE-STORM by categorizing each
13 case according to synoptic flow patterns. Storms in the maritime tropical flow regime
14 transported large amounts of CO, Os, and NOy into the upper troposphere with the
15 midtroposphere remaining relatively clean. During frontal passages a combination of stratiform
16 and convective clouds mixed pollutants more uniformly into the middle and upper levels.
17 Prather and Jacob (1997) and Jaegle et al. (1997) noted that precursors of HOx are also
18 transported to the upper troposphere by deep convection, in addition to primary pollutants (e.g.,
19 NOX, CO, VOCs). The HOX precursors of most importance are water vapor, HCHO, H2O2,
20 CHaOOH, and acetone. The hydroperoxyl radical is critical for oxidizing NO to NO2 in the Os
21 production process as described above.
22 Over remote marine areas, the effects of deep convection on trace gas distributions differ
23 from those over moderately polluted continental regions. Chemical measurements taken by the
24 NASA ER-2 aircraft during the Stratosphere-Troposphere Exchange Project (STEP) off the
25 northern coast of Australia show the influence of very deep convective events. Between 14.5
26 and 16.5 km on the February 2-3, 1987 flight, chemical profiles that included pronounced
27 maxima in CO, water vapor, and CCN, and minima of NOY, and O3 (Pickering et al., 1993).
28 Trajectory analysis showed that these air parcels likely were transported from convective cells
29 800-900 km upstream. Very low marine boundary layer mixing ratios of NOy and Os in this
30 remote region were apparently transported upward in the convection. A similar result was noted
31 in Central Equatorial Pacific Experiment (CEPEX) (Kley et al., 1996) and in Indian Ocean
March 2008 AX2-33 DRAFT-DO NOT QUOTE OR CITE
-------
1 Experiment (INDOEX) (DeLaat et al., 1999) where a series of ozonesonde ascents showed very
2 low upper tropospheric 63 following deep convection. It is likely that similar transport oflow-
3 ozone tropical marine boundary layer air to the upper troposphere occurs in thunderstorms along
4 the east coast of Florida. Deep convection occurs frequently over the tropical Pacific. Low-
5 ozone and low-NOx convective outflow likely will descend in the subsidence region of the
6 subtropical eastern Pacific, leading to some of the cleanest air that arrives at the west coast of the
7 United States.
8 The discussion above relates to the effects of specific convective events. Observations
9 have also been conducted by NASA aircraft in survey mode, in which the regional effects of
10 many convective events can be measured. The Subsonic Assessment Ozone and Nitrogen
11 Oxides Experiment (SONEX) field program in 1997 conducted primarily upper tropospheric
12 measurements over the North Atlantic. The regional effects of convection over North America
13 and the Western Atlantic on upper tropospheric NOx were pronounced (Crawford et al., 2000;
14 Allen et al., 2000). A discussion of the results of model calculations of convection and its effects
15 can be found in Section AX2.7.
16
17 Effects on Photolysis Rates and Wet Scavenging
18 Thunderstorm clouds are optically very thick, and, therefore, have major effects on
19 radiative fluxes and photolysis rates. Madronich (1987) provided modeling estimates of the
20 effects of clouds of various optical depths on photolysis rates. In the upper portion of a
21 thunderstorm anvil, photolysis is likely to be enhanced by a factor of 2 or more due to multiple
22 reflections off the ice crystals. In the lower portion and beneath the cloud, photolysis is
23 substantially decreased. With enhanced photolysis rates, the NO/NO2 ratio in the upper
24 troposphere is driven to larger values than under clear-sky conditions.
25 Thunderstorm updraft regions, which contain copious amounts of water, are regions
26 where efficient scavenging of soluble species can occur (Balkanski et al., 1993). Nitrogen
27 dioxide itself is not very soluble and therefore wet scavenging is not a major removal process for
28 it. However, a major NOx reservoir species, HNOs is extremely soluble. Very few direct field
29 measurements of the amounts of specific trace gases that are scavenged in storms are available.
30 Pickering et al. (2001) used a combination of model estimates of soluble species that did not
31 include wet scavenging and observations of these species from the upper tropospheric outflow
32 region of a major line of convection observed near Fiji. Over 90% of the and in the outflow air
March 2008 AX2-34 DRAFT-DO NOT QUOTE OR CITE
-------
1 appeared to have been removed by the storm. About 50% of CH3OOH and about 80% of HCHO
2 had been lost.
3 Convective processes and small-scale turbulence transport pollutants both upward and
4 downward throughout the planetary boundary layer and the free troposphere. Ozone and its
5 precursors (NOx, CO, and VOCs) can be transported vertically by convection into upper part of
6 the mixed layer on one day, then transported overnight as a layer of elevated mixing ratios,
7 perhaps by a nocturnal low-level jet, and then entrained into a growing convective boundary
8 layer downwind and brought back to the surface.
9 Because NO and NO2 are only slightly soluble, they can be transported over longer
10 distances in the gas phase than can more soluble species which can be depleted by deposition to
11 moist surfaces, or taken up more readily on aqueous surfaces of particles. During transport, they
12 can be transformed into reservoir species such as HNOs, PANs, and ^Os. These species can
13 then contribute to local NOx concentrations in remote areas. For example, it is now well
14 established that PAN decomposition provides a major source of NOx in the remote troposphere
15 (Staudt et al., 2003). PAN decomposition in subsiding air masses from Asia over the eastern
16 Pacific could make an important contribution to Os and NOx enhancement in the United States
17 (Kotchenruther et al., 2001; Hudman et al., 2004). Further details about mechanisms for
18 transporting ozone and its precursors were described at length in CD06.
19
20
21 AX2.6 SOURCES AND EMISSIONS OF NITROGEN OXIDES,
22 AMMONIA, AND SULFUR DIOXIDE
23 All three of the species listed in the title to this section have both natural and
24 anthropogenic sources. In Section AX2.6.1, interactions of NOx with the terrestrial biosphere
25 are discussed. Because of the tight coupling between processes linking emissions and
26 deposition, they are discussed together. In Section AX2.6.2, emissions of NOx, NHa, and SO2
27 are discussed. Field studies evaluating emissions inventories are discussed in Section AX2.6.3.
28
29 AX2.6.1 Interactions of Nitrogen Oxides with the Biosphere
30 Nitrogen oxides affect vegetated ecosystems, and in turn the atmospheric chemistry of
31 NOx is influenced by vegetation. Extensive research on nitrogen inputs from the atmosphere to
32 forests was conducted in the 1980s as part of the Integrated Forest Study, and is summarized by
March 2008 AX2-3 5 DRAFT-DO NOT QUOTE OR CITE
-------
1 Johnson and Lindberg (1992). The following sections discuss sources of NOx from soil,
2 deposition of NOx to foliage, reactions with biogenic hydrocarbons, and ecological effects of
3 nitrogen deposition.
4
5 NOx Sources
6
1 Soil NO
8 Nitric oxide NO from soil metabolism is the dominant, but not exclusive, source of
9 nitrogen oxides from the biosphere to the atmosphere. As noted below, our understanding of
10 NO2 exchange with vegetation suggests that there should be emission of NO2 from foliage when
11 ambient concentrations are less than about 1 ppb. However, Lerdau et al. (2000) have pointed
12 out that present understanding of the global distribution of NOx is not consistent with a large
13 source that would be expected in remote forests if NO2 emission was important when
14 atmospheric concentrations were below the compensation point.
15 The pathways for nitrification and denitrification include two gas-phase intermediates,
16 NO and N2O, some of which can escape. While N2O is of interest for its greenhouse gas
17 potential and role in stratospheric chemistry it is not considered among the reactive nitrogen
18 oxides important for urban and regional air quality and will not be discussed further.
19 Temperature and soil moisture are critical factors that control the rates of reaction and
20 importantly the partitioning between NO and N2O which depend on oxygen levels: in flooded
21 soils where oxygen levels are low, N2O is the dominant soil nitrogen gas; as soil dries, allowing
22 more O2 to diffuse, NO emissions increase. In very dry soils microbial activity is inhibited and
23 emissions of both N2O and NO decrease. Nitrogen metabolism in soil is strongly dependent on
24 the substrate concentrations. Where nitrogen is limiting, nitrogen is efficiently retained and little
25 gaseous nitrogen is released. Where nitrogen is in excess of demand, gaseous nitrogen emissions
26 increase; consequently, soil NO emissions are highest in fertilized agriculture and tropical soils
27 (Davidson and Kingerlee, 1997; Williams et al., 1992).
28
29 Sinks
30 Several reactive nitrogen are species are deposited to vegetation, among them,
31 NO2, PAN, and organic nitrates.
March 2008 AX2-36 DRAFT-DO NOT QUOTE OR CITE
-------
1 HNO3
2 Deposition of HNO3 appears to be relatively simple. Field observations based on
3 concentration gradients and recently using eddy covariance demonstrate rapid deposition that
4 approaches the aerodynamic limit (as constrained by atmospheric turbulence) in the Wesely
5 (1989) formulation based on analogy to resistance. Surface resistance for HNOs uptake by
6 vegetation is negligible. Deposition rates are independent of leaf area or stomatal conductance,
7 implying that deposition occurs to branches, soil, and leaf cuticle as well as internal leaf surfaces.
8 Deposition velocities (Vd) typically exceed 1 cm s^and exhibit a daily pattern controlled
9 by turbulence characteristics: midday maximum and lower values at night when there is stable
10 boundary layer.
11
12 Deposition of NO 2
13 Nitrogen dioxide interaction with vegetation is more complex. Application of 15N-
14 labeled Nitrogen Dioxide demonstrates that Nitrogen Dioxide is absorbed and metabolized by
15 foliage (Siegwolf et al., 2001; Mocker et al., 1998; Segschneider et al., 1995; Weber, et al.,
16 1995). Exposure to NC>2 induces nitrate reductase (Weber et al., 1995, 1998), a necessary
17 enzyme for assimilating oxidized nitrogen. Understanding of NC>2 interactions with foliage is
18 largely based on leaf cuvette and growth chamber studies, which expose foliage or whole plants
19 to controlled levels of NO2 and measure the fraction of NO2 removed from the chamber air. A
20 key finding is that the fit of NO2 flux to NO2 concentration, has a non-zero intercept, implying a
21 compensation point or internal concentration. In studies at very low NC>2 concentrations
22 emission from foliage is observed (Teklemariam and Sparks, 2006). Evidence for a
23 compensation point is not solely based on the fitted intercept. Nitrogen dioxide uptake rate to
24 foliage is clearly related to stomatal conductance. Internal resistance is variable, and may be
25 associated with concentrations of reactive species such as ascorbate in the plant tissue that react
26 with NC>2 (Teklemariam and Sparks, 2006). Foliar NC>2 emissions show some dependence on
27 nitrogen content (Teklemariam and Sparks, 2006). Internal NC>2 appears to derive from plant
28 nitrogen metabolism.
29 Two approaches to modeling NO2 uptake by vegetation are the resistance-in-series
30 analogy which considers flux (F) as the product of concentration (C) and Vd, where is related to
31 the sum of aerodynamic, boundary layer, and internal resistances (Ra, Rb, and RC; positive fluxes
32 are from atmosphere to foliage)
March 2008 AX2-37 DRAFT-DO NOT QUOTE OR CITE
-------
(AX2.6-1)
2 'a v"a-»6 »t-y (AX2.6-2)
3 Ra and Rb and controlled by turbulence in the mixed layer; RC is dependent on
4 characteristics of the foliage and other elements of the soil, and may be viewed as 2 combination
5 of resistance internal to the foliage and external on the cuticle, soils, and bark. This approach is
6 amenable to predicting deposition in regional air quality models (Wesely, 1989). Typically, the
7 NO2, Vd is less than that for Os, due to the surface's generally higher resistance to NO2 uptake,
8 consistent with N(Vs lower reactivity.
9 Alternatively, NO2 exchange with foliage can be modeled from a physiological viewpoint
10 where the flux from the leaf is related to the stomatal conductance and a concentration gradient
11 between the ambient air and interstitial air in the leaf. This approach best describes results for
12 exchange with individual foliage elements, and is expressed per unit leaf (needle) area. While
13 this approach provides linkage to leaf physiology, it is not straightforward to scale up from the
14 leaf to ecosystem scale
15 J=Ss(Ca-Ci) (AX2.6-3)
16 This model implicitly associates the compensation point with a finite internal
17 concentration. Typically observed compensation points are around 1 ppb. Finite values of
18 internal NO2 concentration are consistent with metabolic pathways that include oxides of
19 nitrogen. In this formulation, the uptake will be linear with NO2 concentration, which is
20 typically observed with foliar chamber studies.
21 Several studies have shown the UV dependence of NO2 emission, which implies some
22 photo-induced surface reactions that release NO2. Rather than model this as a UV-dependent
23 internal concentration, it would be more realistic to add an additional term to account for
24 emission that is dependent on light levels and other surface characteristics
25 a-
26 The mechanisms for surface emission are discussed below. Measurement of NC>2 flux is
27 confounded by the rapid interconversion of NO, NC>2, and Os (Gao et al., 1991).
March 2008 AX2-3 8 DRAFT-DO NOT QUOTE OR CITE
-------
1 PAN Deposition
2 Peroxyacetyl nitrate is phytotoxic, so clearly it is absorbed at the leaf. Observations
3 based on inference from concentration gradients and rates of decline at night (Shepson et al.,
4 1992; Schrimpf et al., 1996) and leaf chamber studies (Teklemariam and Sparks, 2004) have
5 indicated that PAN uptake is slower than that of Os; however, recent work in coniferous canopy
6 with direct eddy covariance PAN flux measurements indicated a Vd more similar to that of O^.
1 Uptake of PAN is under stomatal control, has a non-zero deposition at night, and is influenced by
8 leaf wetness (Turnipseed et al., 2006). On the other hand, flux measurements determined by
9 gradient methods over a grass surface showed a Vd closer to 0.1 cm s'1, with large uncertainty
10 (Doskey et al., 2004). A factor of 10 uncertainty remains in Vd 0.1-1 cm s'1 giving a range.
11 Whether the discrepancies are methodological or indicate intrinsic differences between different
12 vegetation is unknown. Uptake of PAN is smaller than its thermal decomposition in all cases.
13
14 Organic Nitrates
15 The biosphere also interacts with NOx through hydrocarbon emissions and their
16 subsequent reactions to form multi-functional organic nitrates. Isoprene nitrates are an important
17 class of these. Isoprene reacts with OH to form a radical that adds NO2 to form a hydroxyalkyl
18 nitrate. The combination of hydroxyl and nitrate functional group makes these compounds
19 especially soluble with low vapor pressures; they likely deposit rapidly (Shepson et al., 1996;
20 Treves et al., 2000). Many other unsaturated hydrocarbons react by analogous routes.
21 Observations at Harvard Forest show a substantial fraction of total NOy not accounted for by
22 NO, NO2 and PAN, which is attributed to the organic nitrates (Horii et al., 2006, Munger et al.,
23 1998). Furthermore, the total NOY flux exceeds the sum of HNO3, NOX, and PAN, which
24 implies that the organic nitrates are a substantial fraction of nitrogen deposition. Other
25 observations that show evidence of hydoxyalkyl nitrates include those of Grossenbacher et al.
26 (2001) and Day et al. (2003).
27 Formation of the hydroxyalkyl nitrates occurs after VOC + OH reaction. In some sense,
28 this mechanism is just an alternate pathway for OH to react with NOX to form a rapidly
29 depositing species. If VOC were not present, OH would be available to react with NO2 when it
30 is present instead to form
March 2008 AX2-39 DRAFT-DO NOT QUOTE OR CITE
-------
1 HONO
2 Nitrous acid formation on vegetative surfaces at night has long been observed based on
3 measurements of positive gradients (Harrison and Kitto, 1994). Surface reactions of NO2
4 enhanced by moisture were proposed to explain these results. Production was evident at sites
5 with high ambient NO2; at low concentration, uptake of HONO exceeded the source.
6 Daytime observations of HONO when rapid photolysis is expected to deplete ambient
7 concentrations to very low levels implies a substantial source of photo-induced HONO formation
8 at a variety of forested sites where measurements have been made. Estimated source strengths
9 are 200-1800 pptv hr"1 in the surface layer (Zhou et al., 2002a, 2003), which is about 20 times
10 faster than all nighttime sources. Nitrous acid sources could be important to OH/HO2 budgets as
11 HONO is rapidly photolyzed by sunlight to OH and NO. Additional evidence of light-dependent
12 reactions to produce HONO comes from discovery of a HONO artifact in pyrex sample inlet
13 lines exposed to ambient light. Either covering the inlet or washing it eliminated the HONO
14 formation (Zhou et al., 2002b). Similar reactions might serve to explain observations of UV-
15 dependent production of NOx in empty foliar cuvettes that had been exposed to ambient air (Hari
16 et al., 2003; Raivonen et al., 2003).
17 Production of HONO in the dark is currently believed to occur via a heterogeneous
18 reaction involving NO2 on wet surfaces (Jenkin et al., 1988; Pitts et al., 1984; He et al., 2006;
19 Sakamaki et al., 1983), and it is proposed that the mechanism has first-order dependence in both
20 NO2 and H2O (Kleffmann et al., 1998; Svensson et al., 1987) despite the stoichiometry.
21 However, the molecular pathway of the mechanism is still under debate. Jenkin et al. (1988)
22 postulated a H2O-NO2 water complex reacting with gas phase NO2 to produce HONO, which is
23 inconsistent with the formation of an N2O4 intermediate leading to HONO as proposed by
24 Finlayson-Pitts et al. (2003). Another uncertainty is whether the reaction forming HONO is
25 dependent on water vapor (Svensson et al., 1987; Stutz et al., 2004b) or water adsorbed on
26 surfaces (Kleffmann et al., 1998). Furthermore, the composition of the surface and the available
27 amount of surface or surface-to-volume ratio can significantly influence the HONO production
28 rates (Kaiser and Wu, 1977; Kleffmann et al., 1998; Svensson et al., 1987), which may explain
29 the difference in the rates observed between laboratory and atmospheric measurements.
30 There is no consensus on a chemical mechanism for photo-induced HONO production.
31 Photolysis of HNOs or MV absorbed on ice or in surface water films has been proposed
March 2008 AX2-40 DRAFT-DO NOT QUOTE OR CITE
-------
1 (Honrath et al., 2002; Ramazan et al., 2004; Zhou et al., 2001, 2003). Alternative pathways
2 include NO2 interaction with organic surfaces such as humic substances (George et al., 2005;
3 Stemmler et al., 2006). Note that either NOs photolysis or heterogeneous reaction of NO2 are
4 routes for recycling deposited nitrogen oxides back to the atmosphere in an active form. Nitrate
5 photolysis would return nitrogen that heretofore was considered irreversibly deposited, surface
6 reactions between NO2 and water films or organic molecules would decrease the effectiveness of
7 observed NO2 deposition if the HONO were re-emitted.
8
9 Fast Homogeneous Reactions
10 Inferences from observations at Blodgett Forest (Cohen et al. in prep) suggest that
11 radicals from Os + VOC react with NOx in the canopy to produce HNOs and organic nitrates
12 among other species. This mechanism would contribute to canopy retention of soil NO emission
13 in forests with high VOC possibly more effectively than the NO to NO2 conversion and foliar
14 uptake of NO2 that has been proposed to reduce the amount of soil NO that escapes to the supra-
15 canopy atmosphere (Jacob and Bakwin, 1991).
16
17 Some NO 2 and HNOs Flux Data from Harvard Forest
18
19 Observations from TDL Measurements of NO 2
20 Harvard Forest is a rural site in central Massachusetts, where ambient NOx, NOy, and
21 other pollutant concentrations and fluxes of total NOy have been measured since 1990 (Munger
22 et al., 1996). An intensive study in 2000 utilized a Tunable Diode Laser Absorption
23 Spectrometer (TDLAS) to measure NO2 and HNOs. TDLAS has an inherently fast response, and
24 for species such as NO2 and HNOs with well-characterized spectra it provides an absolute and
25 specific measurement. Absolute concentrations of HNO3 were measured, and the flux inferred
26 based on the dry deposition inferential method that uses momentum flux measurements to
27 compute a deposition velocity and derives an inferred flux (Wesely and Hicks, 1977; Hicks et al.,
28 1987). Direct eddy covariance calculations for HNOs were not possible because the atmospheric
29 variations were attenuated by interaction with the inlet walls despite very short residence time
30 and use of fluorinated silane coatings to make the inlet walls more hydrophobic. Nitrogen Oxide
31 response was adequate to allow both concentration and eddy covariance flux determination.
32 Simultaneously, NO and NOy eddy covariance fluxes were determined with two separate Os
March 2008 AX2-41 DRAFT-DO NOT QUOTE OR CITE
-------
1 chemiluminescence detectors, one equipped with a H2-gold catalyst at the inlet to convert all
2 reactive nitrogen compounds to NO. Additionally, the measurements include concentration
3 gradients for NO, NO2, and O3 over several annual cycles to examine their vertical profiles in the
4 forest canopy.
5 Overall, the results show typical NO2 concentrations of 1 ppb under clean-air conditions
6 and mean concentrations up to 3 ppb at night and 1 ppb during daytime for polluted conditions.
7 Net positive fluxes (emission) of NO2 were evident in the daytime and negative fluxes
8 (deposition) were observed at night (Figure AX2.6-1). Nitric oxide fluxes were negative during
9 the daytime and near zero at night.
10 In part the opposite NO and NO2 fluxes are simply consequences of variable NO/NO2
11 distributions responding to vertical gradients in light intensity and O3 concentration, which
12 resulted in no net flux of NOx (Gao et al., 1993). In the Harvard Forest situation, the NO and
13 NO2 measurements were not at the same height above the canopy, and the resulting
14 differences derive at least in part from the gradient in flux magnitude between the two inlets
15 (Figure AX2.6-2).
16 At night, when NO concentrations are near 0 due to titration by ambient O3 there is not a
17 flux of NO to offset NO2 fluxes. Nighttime data consistently show NO2 deposition (Figure
18 AX2.6-3), which increases with increasing NO2 concentrations. Concentrations above 10 ppb
19 were rare at this site, but the few high NO2 observations suggest a nonlinear dependence on
20 concentration. The data fit a model with Vd of-0.08 plus an enhancement term that was second
21 order in NO2 concentration. The second order term implies that NO2 deposition rates to
22 vegetation in polluted urban sites would be considerably larger than what was observed at this
23 rural site.
24 After accounting for the NO-NO2 null cycle the net NOx flux could be derived. Overall,
25 there was a net deposition of NOx during the night and essentially zero flux in the day, with large
26 variability in the magnitude and sign of individual flux observations. For the periods with [NO2]
27 > 2 ppb, deposition was always observed. These canopy-scale field observations are consistent
28 with a finite compensation point for NO2 in the canopy that offsets foliar uptake or even reverses
29 it when concentrations are especially low. At concentrations above the compensation point, NOx
30 is absorbed by the canopy. Examination of concentration profiles corroborates the flux
31 measurements (Figure AX2.6-4). During daytime for low-NOx conditions, there is a local
March 2008 AX2-42 DRAFT-DO NOT QUOTE OR CITE
-------
NW
SW
, . 8-
^
"o
6-
"o
1 -
o
§ 2
o
0-
Z~~ 4
1 2
O n
1
•*-" -2
X
3
E -4
[NO] •
[Ncy -— ^ —
. [U^J/IO
*^*fi- ^i^^^^it^M
1 1-i
1! |U-^,
, , . .* 1 a
r-.-.^.l.T -. - - ---i'f -. i*,:.»*
\ ' /" < '
F NO • '*••-. ..
- FNO, ^*^-
FO /10 ------ -
i i i i I t i i i «i i i * i i ii
0 6 12 18
Hours
*rv*s..- "
[*fn7rT , *-*' -v
[-.* *%• : 1-
i*4* *
"i i ' • i 7 t i i i i i i t~\~r~»~i~tm
/'•"-*-*-*
,4
i-i-^*-K , "r * >^*-4^f
V_--''
1 1 1 1 i i • I I 1 1 1 1 ( 1 1 1 • 1
0 6 12 18
Hours
Figure AX2.6-1.
Diel cycles of median concentrations (upper panels) and fluxes (lower
panels) for the Northwest clean sector, left panels) and Southwest
(polluted sector, right panels) wind sectors at Harvard Forest, April-
November, 2000, for NO, NO2, and O3/10. NO and O3 were sampled
at a height of 29 m, and NOi at 22 m. Vertical bars indicate 25th and
27th quartiles for NO and NOi measurements. NOi concentration
and nighttime deposition are enhanced under southwesterly
conditions, as are O3 and the morning NO maximum.
Source: Horn et al. (2004).
4
5
maximum in the concentration profile near the top of the canopy where Oj has a local minimum,
which is consistent with foliar emission or light-dependent production of NOx in the upper
canopy. Depletion is evident for both NOx and Os near the forest floor. Air reaching the ground
has passed through the canopy where uptake is efficient and the vertical exchange rates near the
ground are slow. At night, the profiles generally decrease with decreasing height above the
March 2008
AX2-43
DRAFT-DO NOT QUOTE OR CITE
-------
Simple Model
100
80
,, 60
O)
"0
40
20
0
NO,
NO,
NO
0.0 0.2 0.4 0.6 0.8 1.0-2-1 0 1 2
Concentration (nmol mol"1) Flux (nmol mol"1 cm s"1)
Figure AX2.6-2.
Simple NOx photochemical canopy model outputs. Left panel,
concentrations of NO (dashed) and NO2 (solid); right, fluxes of NO
(dashed) and NO2 (solid). Symbols indicate measurement heights for
NO (29m) and NO2 (22m) at Harvard Forest. The model solves the
continuity equation for NO concentration at 200 levels, d/dz(-
Kc(dNO/dz)) = PNO - LNO, where PNO = [NO]/tl, LNO = [NO]/t2,
and zero net deposition or emission of NOx is allowed. NOx (NO +
NO2) is normalized to Ippb. tl = 70s in this example. Due to the
measurement height difference, observed upward NO2 flux due to
photochemical cycling alone should be substantially larger than
observed downward NO flux attributable to the same process.
Source: Horn (2002).
March 2008
AX2-44
DRAFT-DO NOT QUOTE OR CITE
-------
FNO2 (night) = F0 + V0 [NO2] + a [NO2]2
5-
-5 —
U)
u
"o
E
"o
J -10
u_
CM
O
-15-
-20-
Hourly Data (fitted) *
Nightly Medians +
F0=0
V0= -0.08 ±0.03 (ems'1)
a = -0.013 ± 0.001 (nmo!"1 mol cm s"1)
R2=0.63
0
10 15 20
[N02] (nmol mol'1)
25
30
Figure AX2.6-3.
Hourly (dots) and median nightly (pluses) NOi flux vs. concentration,
with results of least-squares fit on the hourly data (curve). The flux is
expressed in units of concentration times velocity (nmol mol'1 cm s'1)
in order to simplify the interpretation of the coefficients in the least-
squares fit. Pressure and temperature corrections have been taken
into account in the conversion from density to mixing ratio.
Source: Horn et al. (2004).
March 2008
AX2-45
DRAFT-DO NOT QUOTE OR CITE
-------
N(X PROFILES
Canopy
Top *""l
30- '
25- -
20- -
.S>
"3
-C
10
xo
xo
xo
xo
NO2 //NO,
xo
xo
Night Low NOx o
30"
25
20"
15"
10"
Night High NOx <>
. >-
O.TO 0.75 0.80 0.85 15 20 25 30 34 3.8 4.2 4,814 16 18 20 22 24
Canopy
Top ~~
30
25- -
20"
10"
o -•
N02 NO
x o
X O
Day Low NOx o
20 •'
15"
10--
Day High NOx o
0,60 0.65 0.70 0,75 0,80 28 30 32 34 36 4,2 4,4 4.6 4,8 5,0 22 23 24 25 26 27 28
Concentration (nmol mol'1) Concentration (nrnol mo!*1)
Figure AX2.6-4.
Averaged profiles at Harvard Forest give some evidence of some
input near the canopy top from light-mediated ambient reactions, or
emission from open stomates.
Source: Horn et al. (2004).
March 2008
AX2-46 DRAFT-DO NOT QUOTE OR CITE
-------
1 ground, showing only uptake. At higher concentrations, the daytime NOx concentrations are
2 nearly constant through the canopy; no emission is evident from the sunlit leaves.
3 Figure AX2.6-5 compares observed fluxes of all the observed species. The measured
4 NOx and estimated PAN fluxes are small relative to the observed total NOy flux. In clean air,
5 HNOs accounts for nearly all the NOy flux and the sum of all measured species is about equal to
6 the NOy concentration. However, in polluted conditions, unmeasured species are up to 25% of
7 the NOy, and HNOs fluxes cannot account for all the total NOy flux observed. Likely these
8 unmeasured NOy species are hydroxyalkyl nitrates and similar compounds and are rapidly
9 deposited. Although NC>2 uptake may be important to the plant, because it is an input directly to
10 the interior of foliage that can be used immediately in plant metabolism, it is evidently not a
11 significant part of overall nitrogen deposition to rural sites. The deposition of HNO3 and
12 multifunctional organic nitrates are the largest elements of the nitrogen dry deposition budget.
13 Two key areas of remaining uncertainty are the production of HONO over vegetation and the
14 role of very reactive biogenic VOCs. HONO is important because its photolysis is a source of
15 OH radicals, and its formation may represent an unrecognized mechanism to regenerate
16 photochemically active NOx from nitrate that had been considered terminally removed from the
17 atmosphere.
18
19 Ecosystem Effects
20 In addition to the contribution to precipitation acidity, atmospheric nitrogen oxides have
21 ecological effects. Total loading by both and wet and dry deposition is the relevant metric for
22 considering ecosystem impacts. At low inputs, nitrogen deposition adds essential nutrients to
23 terrestrial ecosystems. Most temperate forests are nitrogen limited; thus the inputs stimulate
24 growth. Anthropogenic nitrogen may influence some plant species different and alter the
25 distribution of plant species (cf. Wedin and Tilman, 1996). At high nitrogen loading, where
26 nitrogen inputs exceed nutrient requirements, deleterious effects including forest decline
27 associated with 'nitrogen saturation' are seen (Aber at al., 1998; Driscoll et al., 2003). In aquatic
28 ecosystems, however, nitrogen is may or may not be limiting, but in brackish waters atmospheric
29 deposition of anthropogenic nitrogen is suspected of contributing to eutrophication of some
30 coastal waters and lakes (see Bergstrom and Jansson, 2006; Castro and Driscoll, 2002).
March 2008 AX2-47 DRAFT-DO NOT QUOTE OR CITE
-------
Summer 2000
NW
sw
-» 12
o 10
E
_ 8
E 6
c
4 •
0
0 2
•
""^ 1 2
0
^•^
"5 0.8
c
0
o 0.4
2
IL
0.0
^= -5
CM
E
"5
1 -15
X
3
ul
-25
NOxV+HNO3+PAN
"NO +HNO,
NOx=NO+NO2
NO
- *^<^^?*^^- ...^-- :-- - -
- ' S ' ' •
L \ -N / y
\ '/ --, _ . ,- "* •--.
VA / '-/ \ ~ . /
V ' — ' v-,v-'/
, ^ \/
- ' \ •- ^ '
V \ '
^ ^- ^ — "*
, - ' '
" [E m O i ' / " C 1
FPAN (est.) X
- FNO (param.) T
FNO2 (param.) A
_FHNO3(DDIM) o
FNOy (e.c.) n
- I I I I I I ill LJJLLJL
0 5 10 15 20
He
*
•
1 ~~~~^^^^\ y\ -
'^- \/** — ~^\_— ^—/
">'x\\1— ^ _"^«— "" N .^ - ^
S____ ^ ^^^
1 1 1 1 1)11 1 1 1 1 1 1 1 1 1 1 *
.
/•'"_ Vx'7/ ••X\'x/%v^'~^ '^,^- ", ~- "*'•
• /' Vv_x ^^^ ~~" --"-"; — •
/ - -.^ _ . *
.. — -
. S | ' ff) _ r^-^' J^ /
\ /f\ a. o ,^- /
'[1^ J, n 1 c> | ° °' '\ / ° I
i( F n'
^ M\ -[I-[]'[I
[] []'
1
0 5 10 15 20
•ur
Figure AX2.6-5.
Summer (June-August) 2000 median concentrations (upper panels),
fractions of NOY (middle panels), and fluxes (lower panels) of NOY
and component species separated by wind direction (Northwest on the
left and Southwest on the right). Vertical lines in the flux panels show
25th and 75th quartiles of F(NOY) and F(HNO3); negative fluxes
represent deposition; F(NOX) is derived from eddy covariance F(NO)
and F(NOi) measurements (corrected for photochemical cycling),
F(HNOs) is inferred, and F(NOy) was measured by eddy covariance.
The sum of NOx, HNOs, and PAN accounts for all of the NOy
concentration and flux for Northwesterly (unpolluted background)
flows, whereas up to 50% of NOy and F(NOy) under Southwesterly
flows are in the form of reactive nitrogen species whose fluxes are not
measured or estimated here.
Source: Horn et al. (2006).
March 2008
AX2-48
DRAFT-DO NOT QUOTE OR CITE
-------
1 AX2.6.2 Emissions of NOX, NH3, and SO2
2
3 Emissions ofNOx
4 Estimated annual emissions of NOx, NHa, and SC>2 for 2002 (U.S. Environmental
5 Protection Agency, 2006) are shown in Table AX2.6-1 . Methods for estimating emissions of
6 criteria pollutants, quality assurance procedures, and examples of emissions calculated by using
7 data are given in U.S. Environmental Protection Agency (1999). Discussions of uncertainties in
8 current emissions inventories and strategies for improving them can be found in NARSTO
9 (2005).
10 As can be seen from the table, combustion by stationary sources, such as electrical
1 1 utilities and various industries, accounts for roughly half of total anthropogenic emissions of
12 NOx. Mobile sources account for the other half, with highway vehicles representing the major
13 mobile source component. Approximately half the mobile source emissions are contributed by
14 diesel engines, the remainder are emitted by gasoline-fueled vehicles and other sources.
15 Emissions of NOx associated with combustion arise from contributions from both fuel
16 nitrogen and atmospheric nitrogen. Combustion zone temperatures greater than about 1300 K
17 are required to fix atmospheric N2
18 (AX26.5)
19 Otherwise, NO can be formed from fuel N according to this reaction
20 CaHhOcNd + 02 -> xC02 + yH20 + zNO (AX2.6-6)
21 In addition to NO formation by the schematic reactions given above, some NO2 and CO
22 are also formed depending on temperatures, concentrations of OH and HO2 radicals and O2
23 levels. Fuel nitrogen is highly variable in fossil fuels, ranging from 0.5 to 2.0 percent by weight
24 (wt %) in coal to 0.05% in light distillates (e.g., diesel fuel), to 1 .5 wt % in heavy fuel oils (UK
25 AQEG, 2004). The ratio of NO2 to NOx in primary emissions ranges from 3 to 5 % from
26 gasoline engines, 5 to 12% from heavy-duty diesel trucks, 5 to 10% from vehicles fueled by
27 compressed natural gas and from 5 to 10% from stationary sources. In addition to NOx, motor
28 vehicles also emit HONO, with ratios of HONO to NOX ranging from 0.3% in the Caldecott
29 Tunnel, San Francisco Bay (Kirchstetter and Harley, 1996) to 0.5 to 1.0% in studies in the
March 2008 AX2-49 DRAFT-DO NOT QUOTE OR CITE
-------
1 United Kingdom (UK AQEG, 2004). The NC>2 to NOx ratios in emissions from turbine jet
2 engines are as high as 32 to 35 % during taxi and takeoff (CD93). Sawyer et al. (2000) have
3 reviewed the factors associated with NOX emissions by mobile sources. Marine transport
4 represents a minor source of NOx, but it constitutes a larger source in the EU where it is
5 expected to represent about two-thirds of land-based sources (UK AQEG, 2004).
6
7 NOx Emissions from Natural Sources (Soil, Wild Fires, and Lightning)
8
9 Soil
10 Emission rates of NO from cultivated soil depend mainly on fertilization levels and soil
11 temperature. About 60% of the total NOx emitted by soils occurs in the central corn belt of the
12 United States. The oxidation of NH3, emitted mainly by livestock and soils, leads to the
13 formation of NO, also NH4+ and NO3 fertilizers lead to NO emissions from soils. Estimates of
14 emissions from natural sources are less certain than those from anthropogenic sources. On a
15 global scale, the contribution of soil emissions to the oxidized nitrogen budget is on the order of
16 10% (Van Aardenne et al., 2001; Finlayson-Pitts and Pitts, 2000; Seinfeld and Pandis, 1998), but
17 NOx emissions from fertilized fields are highly variable. Soil NO emissions can be estimated
18 from the fraction of the applied fertilizer nitrogen emitted as NOx, but the flux varies strongly
19 with land use and temperature. Estimated globally averaged fractional applied nitrogen loss as
20 NO varies from 0.3% (Skiba et al., 1997) to 2.5% (Yienger and Levy, 1995). Variability within
21 biomes to which fertilizer is applied, such as shortgrass versus tallgrass prairie, accounts for a
22 factor of three in uncertainty (Williams et al., 1992; Yienger and Levy, 1995; Davidson and
23 Kingerlee, 1997).
24 The local contribution can be much greater than the global average, particularly in
25 summer and especially where corn is grown extensively. Williams et al. (1992) estimated that
26 contributions to NO budgets from soils in Illinois are about 26% of the emissions from industrial
27 and commercial processes in that State. In Iowa, Kansas, Minnesota, Nebraska, and South
28 Dakota, all states with smaller human populations, soil emissions may dominate the NO budget.
29 Conversion of NH3 to NO3 (nitrification) in aerobic soils appears to be the dominant pathway to
30 NO. The mass and chemical form of nitrogen (reduced or oxidized) applied to soils, the
31 vegetative cover, temperature, soil moisture, and agricultural practices such as tillage all
32 influence the amount of fertilizer nitrogen released as NO.
March 2008 AX2-50 DRAFT-DO NOT QUOTE OR CITE
-------
1 Emissions of NO from soils peak in summer when Os formation is also at a maximum.
2 An NRC panel report (NRC, 2002) outlined the role of agriculture in emissions of air pollutants
3 including NO and NH3. That report recommends immediate implementation of best
4 management practices to control these emissions, and further research to quantify the magnitude
5 of emissions and the impact of agriculture on air quality. Civerolo and Dickerson (1998) report
6 that use of the no-till cultivation technique on a fertilized cornfield in Maryland reduced NO
7 emissions by a factor of seven.
8
9 NOxfrom Biomass Burning
10 During biomass burning, nitrogen is derived mainly from fuel nitrogen and not from
11 atmospheric N2, since temperatures required to fix atmospheric N2 are likely to be found only in
12 the flaming crowns of the most intense boreal forest fires. Nitrogen is present mainly in plants as
13 amino (NH2) groups in amino acids. During combustion, nitrogen is released mainly in
14 unidentified forms, presumably as N2, with very little remaining in fuel ash. Apart from N2, the
15 most abundant species in biomass burning plumes is NO. Emissions of NO account for only
16 about 10 to 20% relative to fuel N (Lobert et al., 1991). Other species such as NO2, nitriles,
17 ammonia, and other nitrogen compounds account for a similar amount. Emissions of NOx are
18 about 0.2 to 0.3% relative to total biomass burned (e.g., Andreae, 1991; Radke et al., 1991).
19 Westerling et al. (2006) have noted that the frequency and intensity of wildfires in the western
20 United States have increased substantially since 1970.
21
22 Lightning Production of NO
23 Annual global production of NO by lightning is the most uncertain source of reactive
24 nitrogen. In the last decade, literature values of the global average production rate range from
25 2 to 20 Tg N per year. However, the most likely range is from 3 to 8 Tg N per year, because the
26 majority of the recent estimates fall in this range. The large uncertainty stems from several
27 factors: (1) a large range of NO production rates per meter of flash length (as much as two
28 orders of magnitude); (2) the open question of whether cloud-to-ground (CG) flashes and
29 intracloud flashes (1C) produce substantially different amounts of NO; (3) the global flash rate;
30 and (4) the ratio of the number of 1C flashes to the number of CG flashes. Estimates of the
31 amount of NO produced per flash have been made based on theoretical considerations (e.g.,
32 Price et al., 1997), laboratory experiments (e.g., Wang et al., 1998); field experiments (e.g., Stith
March 2008 AX2-51 DRAFT-DO NOT QUOTE OR CITE
-------
1 et al., 1999; Huntrieser et al., 2002, 2007) and through a combination of cloud-resolving model
2 simulations, observed lightning flash rates, and anvil measurements of NO (e.g., DeCaria et al.,
3 2000, 2005; Ott et al., 2007). The latter method was also used by Pickering et al. (1998), who
4 showed that only ~5 to 20% of the total NO produced by lightning in a given storms exists in the
5 boundary layer at the end of a thunderstorm. Therefore, the direct contribution to boundary layer
6 Os production by lightning NO is thought to be small. However, lightning NO production can
7 contribute substantially to Os production in the middle and upper troposphere. DeCaria et al.
8 (2005) estimated that up to 10 ppbv of ozone was produced in the upper troposphere in the first
9 24 hours following a Colorado thunderstorm due to the injection of lightning NO. A series of
10 midlatitude and subtropical thunderstorm events have been simulated with the model of DeCaria
11 et al. (2005), and the derived NO production per CG flash averaged 500 moles/flash while
12 average production per 1C flash was 425 moles/flash (Ott et al., 2006).
13 A major uncertainty in mesoscale and global chemical transport models is the
14 parameterization of lightning flash rates. Model variables such as cloud top height, convective
15 precipitation rate, and upward cloud mass flux have been used to estimate flash rates. Allen and
16 Pickering (2002) have evaluated these methods against observed flash rates from satellite, and
17 examined the effects on ozone production using each method.
18
19 Uses of Satellite Data to Derive Emissions
20 Satellite data have been shown to be useful for optimizing estimates of emissions of NO2.
21 (Leue et al., 2001; Martin et al., 2003; Jaegle et al., 2005). Satellite-borne instruments such as
22 Global Ozone Monitoring Experiment (GOME) (Martin et al., 2003; and references therein) and
23 Scanning Imaging Absorption Spectrometer for Atmospheric Chartography (SCIAMACHY)
24 (Bovensmann et al., 1999) retrieve tropospheric columns of NO2, which can then be combined
25 with model-derived chemical lifetimes of NOx to yield emissions of NOx.
26 Top-down inference of NOx emission inventory from the satellite observations of NO2
27 columns by mass balance requires at minimum three pieces of information: the retrieved
28 tropospheric NO2 column, the ratio of tropospheric NOx to NO2 columns, and the NOx lifetime
29 against loss to stable reservoirs. A photochemical model has been used to provide information
30 on the latter two pieces of information. The method is generally applied exclusively to land
31 surface emissions, excluding lightning. Tropospheric NO2 columns are insensitive to lightning
32 NOx emissions since most of the lightning NOx in the upper troposphere is present as NO at the
March 2008 AX2-52 DRAFT-DO NOT QUOTE OR CITE
-------
1 local time of the satellite measurements (Ridley et al., 1996), owing to the slower reactions of
2 NO with O3 there.
3 Jaegle et al. (2005) applied additional information on the spatial distribution of emissions
4 and on fire activity to partition NOx emissions into sources from fossil fuel combustion, soils,
5 and biomass burning. Global a posteriori estimates of soil NOx emissions are 68% larger than
6 the a priori estimates. Large increases are found for the agricultural region of the western United
7 States during summer, increasing total U.S. soil NOx emissions by a factor of 2 to 0.9 Tg N yr'1.
8 Bertram et al. (2005) found clear signals in the SCIAMACHY observations of short intense NOx
9 pulses following springtime fertilizer application and subsequent precipitation over agricultural
10 regions of the western United States. For the agricultural region in North-Central Montana, they
11 calculate a yearly SCIAMACHY top-down estimate that is 60% higher than a commonly used
12 model of soil NOx emissions by Yienger and Levy (1995).
13 Martin et al. (2006) retrieved tropospheric nitrogen dioxide (NO2) columns for
14 May 2004 to April 2005 from the SCIAMACHY satellite instrument to derive top-down NOX
15 emissions estimates via inverse modeling with a global chemical transport model (GEOS-Chem).
16 The top-down emissions were combined with a priori information from a bottom-up emission
17 inventory with error weighting to achieve an improved a posteriori estimate of the global
18 distribution of surface NOx emissions. Their a posteriori inventory improves the GEOS-Chem
19 simulation of NOX, PAN, and HNO3 with respect to airborne in situ measurements over and
20 downwind of New York City. Their a posteriori inventory shows lower NOx emissions from the
21 Ohio River valley during summer than during winter, reflecting recent controls on NOx
22 emissions from electric utilities. Their a posteriori inventory is highly consistent (R2 = 0.82,
23 bias = 3%) with the NEI99 inventory for the United States. In contrast, their a posteriori
24 inventory is 68% larger than a recent inventory by Streets et al. (2003) for East Asia for the
25 year 2000.
26
27 Emissions ofNHs
28 Emissions of NH3 show a strikingly different pattern from those of NOx. Three-way
29 catalysts used in motor vehicles emit small amounts of NH3 as a byproduct during the reduction
30 of NOX. Stationary combustion sources make only a small contribution to emissions of NH3
31 because efficient combustion favors formation of NOx and, NH3 from combustion is produced
March 2008 AX2-53 DRAFT-DO NOT QUOTE OR CITE
-------
1 mainly by inefficient, low temperature fuel combustion. For these reasons, most emissions of
2 NHs arise from fertilized soils and from livestock.
3 The initial step in the oxidation of atmospheric NH3 to NO is by reaction with OH
4 radicals. However, the lifetime of NH? from this pathway is sufficiently long (-1-2 months
5 using typical OH values 1-2 x 106/cm3) that it is a small sink compared to uptake of NH? by
6 cloud drops, dry deposition, and aerosol particles. Thus, the gas-phase oxidation of NH? makes a
7 very small contribution as a source of NO. Holland et al. (2005) estimated wet and dry
8 deposition of NHX, based on measurements over the continental United States, and found that
9 emissions of NH3 in the National Emissions Inventory are perhaps underestimated by about a
10 factor of two to three. Reasons for this imbalance include under-representation of deposition
11 monitoring sites in populated areas and the neglect of off-shore transport in their estimate. The
12 use of fixed deposition velocities that do not reflect local conditions at the time of measurement
13 introduces additional uncertainty into their estimates of dry deposition.
14
15 Emissions of SO2
16 As can be seen from Table AX2.6-1, emissions of SO2 are due mainly to the combustion
17 of fossil fuels by electrical utilities and industry. Transportation related sources make only a
18 minor contribution. As a result, most SO2 emissions originate from point sources. Since sulfur
19 is a volatile component of fuels, it is almost quantitatively released during combustion and
20 emissions can be calculated on the basis of the sulfur content of fuels to greater accuracy than for
21 other pollutants such as NOx or primary PM.
22 The major natural sources of SO2 are volcanoes and biomass burning and DMS oxidation
23 over the oceans. SO2 constitutes a relatively minor fraction (0.005% by volume) of volcanic
24 emissions (Holland, 1978). The ratio of H2S to SO2 is highly variable in volcanic gases. It is
25 typically much less than one, as in the Mt. Saint Helen's eruption (Turco et al., 1983). However,
26 in addition to being degassed from magma, H2S can be produced if ground waters, especially
27 those containing organic matter, come into contact with volcanic gases. In this case, the ratio of
28 H2S to SO2 can be greater than one. H2S produced this way would more likely be emitted
29 through side vents than through eruption columns (Pinto et al., 1989). Primary particulate sulfate
30 is a component of marine aerosol and is also produced by wind erosion of surface soils.
31 Volcanic sources of SO2 are limited to the Pacific Northwest, Alaska, and Hawaii. Since
32 1980, the Mount St. Helens volcano in the Washington Cascade Range (46.20 N, 122.18 W,
March 2008 AX2-54 DRAFT-DO NOT QUOTE OR CITE
-------
1 summit 2549 m asl) has been a variable source of 862. Its major effects came in the explosive
2 eruptions of 1980, which primarily affected the northern part of the mountainous western half of
3 the United States. The Augustine volcano near the mouth of the Cook Inlet in southwestern
4 Alaska (59.363 N, 153.43 W, summit 1252 m asl) has had variable SC>2 emission since its last
5 major eruptions in 1986. Volcanoes in the Kamchatka peninsula of eastern region of Siberian
6 Russia do not significantly effect surface SO2 concentrations in northwestern North America.
7 The most serious effects in the United States from volcanic SC>2 occurs on the island of Hawaii.
8 Nearly continuous venting of 862 from Mauna Loa and Kilauea produces 862 in such large
9 amounts that >100 km downwind of the island SC>2 concentrations can exceed 30 ppbv
10 (Thornton and Bandy, 1993). Depending on wind direction, the west coast of Hawaii (Kona
11 region) has had significant deleterious effects from SO2 and acidic sulfate aerosols for the past
12 decade.
13 Emissions of SC>2 from burning vegetation are generally in the range of 1 to 2% of the
14 biomass burned (see e.g., Levine et al., 1999). Sulfur is bound in amino acids in vegetation.
15 This organically bound sulfur is released during combustion. However, unlike nitrogen, about
16 half of the sulfur initially present in vegetation is found in the ash (Delmas, 1982). Gaseous
17 emissions are mainly in the form of SC>2 with much smaller amounts of H2S and OCS. The ratio
18 of gaseous nitrogen to sulfur emissions is about 14, very close to their ratio in plant tissue
19 (Andreae, 1991). The ratio of reduced nitrogen and sulfur species such as NH3 and H2S to their
20 more oxidized forms, such as NO and SO2, increases from flaming to smoldering phases of
21 combustion, as emissions of reduced species are favored by lower temperatures and C>2 reduced
22 availability.
23 Emissions of reduced sulfur species are associated typically with marine organisms living
24 either in pelagic or coastal zones and with anaerobic bacteria in marshes and estuaries.
25 Mechanisms for their oxidation were discussed in Section AX2.2. Emissions of dimethyl sulfide
26 (DMS) from marine plankton represent the largest single source of reduced sulfur species to the
27 atmosphere (e.g., Berresheim et al., 1995). Other sources such as wetlands and terrestrial plants
28 and soils probably account for less than 5% of the DMS global flux, with most of this coming
29 from wetlands.
30 The coastal and wetland sources of DMS have a dormant period in the fall/winter from
31 senescence of plant growth. Marshes die back in fall and winter, so dimethyl sulfide emissions
March 2008 AX2-55 DRAFT-DO NOT QUOTE OR CITE
-------
1 from them are lower, reduced light levels in winter at mid to high latitudes reduce cut
2 phytoplankton growth which also tends to reduce DMS emissions. Western coasts at mid to high
3 latitudes have reduced levels of the light that drive photochemical production and oxidation of
4 DMS. Freezing at mid and high latitudes affects the release of biogenic sulfur gases, particularly
5 in the nutrient-rich regions around Alaska. Transport of SC>2 from regions of biomass burning
6 seems to be limited by heterogeneous losses that accompany convective processes that ventilate
7 the surface layer and the lower boundary layer (Thornton et al., 1996, TRACE-P data archive).
8 However, it should be noted that reduced sulfur species are also produced by industry.
9 For example, DMS is used in petroleum refining and in petrochemical production processes to
10 control the formation of coke and carbon monoxide. In addition, it is used to control dusting in
11 steel mills. It is also used in a range of organic syntheses. It also has a use as a food flavoring
12 component. It can also be oxidized by natural or artificial means to dimethyl sulfoxide (DMSO),
13 which has several important solvent properties.
14
15 AX2.6.3 Field Studies Evaluating Emissions Inventories
16 Comparisons of emissions model predictions with observations have been performed in a
17 number of environments. A number of studies of ratios of concentrations of CO to NOx and
18 NMOC to NOx during the early 1990s in tunnels and ambient air (summarized in Air Quality
19 Criteria for Carbon Monoxide (U.S. Environmental Protection Agency, 2000)) indicated that
20 emissions of CO and NMOC were systematically underestimated in emissions inventories.
21 However, the results of more recent studies have been mixed in this regard, with many studies
22 showing agreement to within ±50% (U.S. Environmental Protection Agency, 2000).
23 Improvements in many areas have resulted from the process of emissions model development,
24 evaluation, and further refinement. It should be remembered that the conclusions from these
25 reconciliation studies depend on the assumption that NOx emissions are predicted correctly by
26 emissions factor models. Roadside remote sensing data indicate that over 50% of NMHC and
27 CO emissions are produced by less than about 10% of the vehicles (Stedman et al., 1991). These
28 "super-emitters" are typically poorly maintained vehicles. Vehicles of any age engaged in off-
29 cycle operations (e.g., rapid accelerations) emit much more than if operated in normal driving
30 modes. Bishop and Stedman (1996) found that the most important variables governing CO
31 emissions are fleet age and owner maintenance.
March 2008 AX2-56 DRAFT-DO NOT QUOTE OR CITE
-------
1 Emissions inventories for North America can be evaluated by comparison to measured
2 long-term trends and or ratios of pollutants in ambient air. A decadal field study of ambient CO
3 at a rural site in the eastern United States (Hallock-Waters et al., 1999) indicates a downward
4 trend consistent with the downward trend in estimated emissions over the period 1988 to 1999
5 (U.S. Environmental Protection Agency, 1997), even when a global downward trend is
6 accounted for. Measurements at two urban areas in the United States confirmed the decrease in
7 CO emissions (Parrish et al., 2002). That study also indicated that the ratio of CO to NOx
8 emissions decreased by almost a factor of three over 12 years (such a downward trend was noted
9 in AQCD 96). Emissions estimates (U.S. Environmental Protection Agency, 1997) indicate a
10 much smaller decrease in this ratio, suggesting that NOx emissions from mobile sources may be
11 underestimated and/or increasing. Parrish et al. (2002) conclude that O3 photochemistry in U.S.
12 urban areas may have become more NOx-limited over the past decade.
13 Pokharel et al. (2002) employed remotely sensed emissions from on-road vehicles and
14 fuel use data to estimate emissions in Denver. Their calculations indicate a continual decrease in
15 CO, HC, and NO emissions from mobile sources over the 6-year study period. Inventories based
16 on the ambient data were 30 to 70% lower for CO, 40% higher for HC, and 40 to 80% lower for
17 NO than those predicted by the MOBILE6 model.
18 Stehr et al. (2000) reported simultaneous measurements of CO, SO2, and NOy at an East
19 Coast site. By taking advantage of the nature of mobile sources (they emit NOx and CO but
20 little SO2) and power plants (they emit NOX and SO2 but little CO), the authors evaluated
21 emissions estimates for the eastern United States. Results indicated that coal combustion
22 contributes 25 to 35% of the total NOx emissions in rough agreement with emissions inventories
23 (U.S. Environmental Protection Agency, 1997).
24 Parrish et al. (1998) and Parrish and Fehsenfeld (2000) proposed methods to derive
25 emission rates by examining measured ambient ratios among individual VOC, NOx and NOy.
26 There is typically a strong correlation among measured values for these species because emission
27 sources are geographically collocated, even when individual sources are different. Correlations
28 can be used to derive emissions ratios between species, including adjustments for the impact of
29 photochemical aging. Investigations of this type include correlations between CO and NOy (e.g.,
30 Parrish et al., 1991), between individual VOC species and NOY (Goldan et al., 1995, 1997, 2000)
31 and between various individual VOC (Goldan et al., 1995, 1997; McKeen and Liu, 1993;
March 2008 AX2-57 DRAFT-DO NOT QUOTE OR CITE
-------
1 McKeen et al., 1996). Buhr et al. (1992) derived emission estimates from principal component
2 analysis (PC A) and other statistical methods. Many of these studies are summarized in Trainer
3 et al. (2000), Parrish et al. (1998), and Parrish and Fehsenfeld (2000). Goldstein and Schade
4 (2000) also used species correlations to identify the relative impacts of anthropogenic and
5 biogenic emissions. Chang et al. (1996, 1997) and Mendoza-Dominguez and Russell (2000,
6 2001) used the more quantative technique of inverse modeling to derive emission rates, in
7 conjunction with results from chemistry-transport models.
8
9
10 AX2.7 METHODS USED TO CALCULATE CONCENTRATIONS OF
11 NITROGEN OXIDES AND THEIR CHEMICAL
12 INTERACTIONS IN THE ATMOSPHERE
13 Atmospheric chemistry and transport models are the major tools used to calculate the
14 relations among Os, other oxidants, and their precursors, the transport and transformation of air
15 toxics, the production of secondary organic aerosol, the evolution of the particle size distribution,
16 and the production and deposition of pollutants affecting ecosystems. Chemical transport
17 models are driven by emissions inventories for primary species such as the precursors for 63 and
18 PM and by meterological fields produced by other numerical models. Emissions of precursor
19 compounds can be divided into anthropogenic and natural source categories. Natural sources can
20 be further divided into biotic (vegetation, microbes, animals) and abiotic (biomass burning,
21 lightning) categories. However, the distinction between natural sources and anthropogenic
22 sources is often difficult to make as human activities affect directly, or indirectly, emissions from
23 what would have been considered natural sources during the preindustrial era. Emissions from
24 plants and animals used in agriculture have been referred to as anthropogenic or natural in
25 different applications. Wildfire emissions may be considered to be natural, except that forest
26 management practices may have led to the buildup of fuels on the forest floor, thereby altering
27 the frequency and severity of forest fires. Needed meteorological quantities such as winds and
28 temperatures are taken from operational analyses, reanalyses, or circulation models. In most
29 cases, these are off-line analyses, i.e., they are not modified by radiatively active species such as
30 Os and particles generated by the model.
31 A brief overview of atmospheric chemistry-transport models is given in Section AX2.7.1.
32 A discussion of emissions inventories of precursors used by these models is given in Section
March 2008 AX2-58 DRAFT-DO NOT QUOTE OR CITE
-------
1 AX2.7.2. Uncertainties in emissions estimates have also been discussed in Air Quality Criteria
2 for Particulate Matter (U.S. Environmental Protection Agency, 2004). Chemistry-transport
3 model evaluation and an evaluation of the reliability of emissions inventories are presented in
4 Section AX2.7.4.
5
6 AX2.7.1 Chemistry-Transport Models
7 Atmospheric CTMs have been developed for application over a wide range of spatial
8 scales ranging from neighborhood to global. Regional scale CTMs are used: (1) to obtain better
9 understanding of the processes controlling the formation, transport, and destruction of gas-and
10 particle-phase criteria and hazardous air pollutants; (2) to understand the relations between Os
11 concentrations and concentrations of its precursors such as NOx and VOCs, the factors leading to
12 acid deposition, and hence to possible damage to ecosystems; and (3) to understand relations
13 among the concentration patterns of various pollutants that may exert adverse health effects.
14 Chemistry Transport Models are also used for determining control strategies for 63 precursors.
15 However, this application has met with varying degrees of success because of the highly
16 nonlinear relations between Os and emissions of its precursors, and uncertainties in emissions,
17 parameterizations of transport, and chemical production and loss terms. Uncertainties in
18 meteorological variables and emissions can be large enough to lead to significant errors in
19 developing control strategies (e.g., Russell and Dennis, 2000; Sillman et al., 1995).
20 Global scale CTMs are used to address issues associated with climate change,
21 stratospheric ozone depletion, and to provide boundary conditions for regional scale models.
22 CTMs include mathematical (and often simplified) descriptions of atmospheric transport, the
23 transfer of solar radiation through the atmosphere, chemical reactions, and removal to the surface
24 by turbulent motions and precipitation for pollutants emitted into the model domain. Their upper
25 boundaries extend anywhere from the top of the mixing layer to the mesopause (about 80 km in
26 height), to obtain more realistic boundary conditions for problems involving stratospheric
27 dynamics. There is a trade-off between the size of the modeling domain and the grid resolution
28 used in the CTM that is imposed by computational resources.
29 There are two major formulations of CTMs in current use. In the first approach, grid-
30 based, or Eulerian, air quality models, the region to be modeled (the modeling domain) is
31 subdivided into a three-dimensional array of grid cells. Spatial derivatives in the species
March 2008 AX2-59 DRAFT-DO NOT QUOTE OR CITE
-------
1 continuity equations are cast in finite-difference there are also some finite-element models, but
2 not many applications form over this grid, and a system of equations for the concentrations of all
3 the chemical species in the model are solved numerically at each grid point. Time dependent
4 continuity (mass conservation) equations are solved for each species including terms for
5 transport, chemical production and destruction, and emissions and deposition (if relevant), in
6 each cell. Chemical processes are simulated with ordinary differential equations, and transport
7 processes are simulated with partial differential equations. Because of a number of factors such
8 as the different time scales inherent in different processes, the coupled, nonlinear nature of the
9 chemical process terms, and computer storage limitations, all of the terms in the equations are
10 not solved simultaneously in three dimensions. Instead, operator splitting, in which terms in the
11 continuity equation involving individual processes are solved sequentially, is used. In the second
12 CTM formulation, trajectory or Lagrangian models, a large number of hypothetical air parcels
13 are specified as following wind trajectories. In these models, the original system of partial
14 differential equations is transformed into a system of ordinary differential equations.
15 A less common approach is to use a hybrid Lagrangian/Eulerian model, in which certain
16 aspects of atmospheric chemistry and transport are treated with a Lagrangian approach and
17 others are treaded in an Eulerian manner (e.g., Stein et al., 2000). Each approach has its
18 advantages and disadvantages. The Eulerian approach is more general in that it includes
19 processes that mix air parcels and allows integrations to be carried out for long periods during
20 which individual air parcels lose their identity. There are, however, techniques for including the
21 effects of mixing in Lagrangian models such as FLEXPART (e.g., Zanis et al., 2003), ATTILA
22 (Reithmeier and Sausen, 2002), and CLaMS (McKenna et al., 2002).
23
24 Regional Scale Chemistry Transport Models
25 Major modeling efforts within the U.S. Environmental Protection Agency center on the
26 Community Multiscale Air Quality modeling system (CMAQ) (Byun and Ching, 1999; Byun
27 and Schere, 2006). A number of other modeling platforms using Lagrangian and Eulerian
28 frameworks have been reviewed in the 96 AQCD for Os (U.S. Environmental Protection
29 Agency, 1997), and in Russell and Dennis (2000). The capabilities of a number of CTMs
30 designed to study local- and regional-scale air pollution problems are summarized by Russell and
31 Dennis (2000). Evaluations of the performance of CMAQ are given in Arnold et al. (2003), Eder
32 and Yu (2005), Appel et al. (2005), and Fuentes and Raftery (2005). The domain of CMAQ can
March 2008 AX2-60 DRAFT-DO NOT QUOTE OR CITE
-------
1 extend from several hundred km to the hemispherical scale. In addition, both of these classes of
2 models allow the resolution of the calculations over specified areas to vary. CMAQ is most
3 often driven by the MM5 mesoscale meteorological model (Seaman, 2000), though it may be
4 driven by other meteorological models (e.g., RAMS). Simulations of Os episodes over regional
5 domains have been performed with a horizontal resolution as low as 1 km, and smaller
6 calculations over limited domains have been accomplished at even finer scales. However,
7 simulations at such high resolutions require better parameterizations of meteorological processes
8 such as boundary layer fluxes, deep convection and clouds (Seaman, 2000), and finer-scale
9 emissions. Finer spatial resolution is necessary to resolve features such as urban heat island
10 circulations; sea, bay, and land breezes; mountain and valley breezes, and the nocturnal low-level
11 jet.
12 The most common approach to setting up the horizontal domain is to nest a finer grid
13 within a larger domain of coarser resolution. However, there are other strategies such as the
14 stretched grid (e.g., Fox-Rabinovitz et al., 2002) and the adaptive grid. In a stretched grid, the
15 grid's resolution continuously varies throughout the domain, thereby eliminating any potential
16 problems with the sudden change from one resolution to another at the boundary. Caution
17 should be exercised in using such a formulation, because certain parameterizations that are valid
18 on a relatively coarse grid scale (such as convection) may not be valid on finer scales. Adaptive
19 grids are not fixed at the start of the simulation, but instead adapt to the needs of the simulation
20 as it evolves (e.g., Hansen et al., 1994). They have the advantage that they can resolve processes
21 at relevant spatial scales. However, they can be very slow if the situation to be modeled is
22 complex. Additionally, if adaptive grids are used for separate meteorological, emissions, and
23 photochemical models, there is no reason a priori why the resolution of each grid should match,
24 and the gains realized from increased resolution in one model will be wasted in the transition to
25 another model. The use of finer horizontal resolution in CTMs will necessitate finer-scale
26 inventories of land use and better knowledge of the exact paths of roads, locations of factories,
27 and, in general, better methods for locating sources and estimating their emissions.
28 The vertical resolution of these CTMs is variable, and usually configured to have higher
29 resolution near the surface and decreasing aloft. Because the height of the boundary layer is of
30 critical importance in simulations of air quality, improved resolution of the boundary layer height
31 would likely improve air quality simulations. Additionally, current CTMs do not adequately
March 2008 AX2-61 DRAFT-DO NOT QUOTE OR CITE
-------
1 resolve fine scale features such as the nocturnal low-level jet in part because little is known about
2 the nighttime boundary layer.
3 CTMs require time-dependent, three-dimensional wind fields for the period of
4 simulation. The winds may be either generated by a model using initial fields alone or with four-
5 dimensional data assimilation to improve the model's performance, fields (i.e., model equations
6 can be updated periodically or "nudged", to bring results into agreement with observations.
7 Modeling efforts typically focus on simulations of several days' duration, the typical time scale
8 for individual 63 episodes, but there have been several attempts at modeling longer periods. For
9 example, Kasibhatla and Chameides (2000) simulated a four-month period from May to
10 September of 1995 using MAQSIP. The current trend in modeling applications is towards
11 annual simulations. This trend is driven in part by the need to better understand observations of
12 periods of high wintertime PM (e.g., Blanchard et al., 2002) and the need to simulate Os episodes
13 occurring outside of summer.
14 Chemical kinetics mechanisms (a set of chemical reactions) representing the important
15 reactions occurring in the atmosphere are used in CTMs to estimate the rates of chemical
16 formation and destruction of each pollutant simulated as a function of time. Unfortunately,
17 chemical mechanisms that explicitly treat the reactions of each individual reactive species are too
18 computationally demanding to be incorporated into CTMs. For example, a master chemical
19 mechanism includes approximately 10,500 reactions involving 3603 chemical species (Derwent
20 et al., 2001). Instead, "lumped" mechanisms, that group compounds of similar chemistry
21 together, are used. The chemical mechanisms used in existing photochemical Os models contain
22 significant uncertainties that may limit the accuracy of their predictions; the accuracy of each of
23 these mechanisms is also limited by missing chemistry. Because of different approaches to the
24 lumping of organic compounds into surrogate groups, chemical mechanisms can produce
25 somewhat different results under similar conditions. The CB-IV chemical mechanism (Gery
26 et al., 1989), the RADMII mechanism (Stockwell et al., 1990), the SAPRC (e.g., Wang et al.,
27 2000a,b; Carter, 1990) and the RACM mechanisms can be used in CMAQ. Jimenez et al. (2003)
28 provide brief descriptions of the features of the main mechanisms in use and they compared
29 concentrations of several key species predicted by seven chemical mechanisms in a box model
30 simulation over 24 h. The average deviation from the average of all mechanism predictions for
31 Os and NO over the daylight period was less than 20%, and was 10% for NC>2 for all
March 2008 AX2-62 DRAFT-DO NOT QUOTE OR CITE
-------
1 mechanisms. However, much larger deviations were found for HNOs, PAN, HC>2, H2O2, C2H4,
2 and CsH8 (isoprene). An analysis for OH radicals was not presented. The large deviations
3 shown for most species imply differences between the calculated lifetimes of atmospheric
4 species and the assignment of model simulations to either NOx-limited or radical quantity
5 limited regimes between mechanisms. Gross and Stockwell (2003) found small differences
6 between mechanisms for clean conditions, with differences becoming more significant for
7 polluted conditions, especially for NC>2 and organic peroxy radicals. They caution modelers to
8 consider carefully the mechanisms they are using. Faraji et al. (2005) found differences of 40%
9 in peak 1 h 63 in the Houston-Galveston-Brazoria area between simulations using SAPRAC and
10 CB4. They attributed differences in predicted 63 concentrations to differences in the
11 mechanisms of oxidation of aromatic hydrocarbons.
12 CMAQ and other CTMs (e.g., PM-CAMx) incorporate processes and interactions of
13 aerosol-phase chemistry (Mebust et al., 2003). There have also been several attempts to study
14 the feedbacks of chemistry on atmospheric dynamics using meteorological models, like MM5
15 (e.g., Grell et al., 2000; Liu et al., 2001a; Lu et al., 1997; Park et al., 2001). This coupling is
16 necessary to simulate accurately feedbacks such as may be caused by the heavy aerosol loading
17 found in forest fire plumes (Lu et al., 1997; Park et al., 2001), or in heavily polluted areas.
18 Photolysis rates in CMAQ can now be calculated interactively with model produced 63, NO2,
19 and aerosol fields (Binkowski et al., 2007).
20 Spatial and temporal characterizations of anthropogenic and biogenic precursor emissions
21 must be specified as inputs to a CTM. Emissions inventories have been compiled on grids of
22 varying resolution for many hydrocarbons, aldehydes, ketones, CO, NH?, and NOx. Emissions
23 inventories for many species require the application of some algorithm for calculating the
24 dependence of emissions on physical variables such as temperature and to convert the
25 inventories into formatted emission files required by a CTM. For example, preprocessing of
26 emissions data for CMAQ is done by the Spare-Matrix Operator Kernel Emissions (SMOKE)
27 system. For many species, information concerning the temporal variability of emissions is
28 lacking, so long-term (e.g., annual or O3-season) averages are used in short-term, episodic
29 simulations. Annual emissions estimates are often modified by the emissions model to produce
30 emissions more characteristic of the time of day and season. Significant errors in emissions can
31 occur if an inappropriate time dependence or a default profile is used. Additional complexity
March 2008 AX2-63 DRAFT-DO NOT QUOTE OR CITE
-------
1 arises in model calculations because different chemical mechanisms are based on different
2 species, and inventories constructed for use with another mechanism must be adjusted to reflect
3 these differences. This problem also complicates comparisons of the outputs of these models
4 because one chemical mechanism may produce some species not present in another mechanism
5 yet neither may agree with the measurements.
6 In addition to wet deposition, dry deposition (the removal of chemical species from the
7 atmosphere by interaction with ground-level surfaces) is an important removal process for
8 pollutants on both urban and regional scales and must be included in CTMs. The general
9 approach used in most models is the resistance in series method, in which where dry deposition
10 is parameterized with a Vd, which is represented as Vd = (ra + rb + r^1 where ra, rt,, and rc
11 represent the resistance due to atmospheric turbulence, transport in the fluid sublayer very near
12 the elements of surface such as leaves or soil, and the resistance to uptake of the surface itself.
13 This approach works for a range of substances, although it is inappropriate for species with
14 substantial emissions from the surface or for species whose deposition to the surface depends on
15 its concentration at the surface itself. The approach is also modified somewhat for aerosols: the
16 terms it, and rc are replaced with a surface Vd to account for gravitational settling. In their
17 review, Wesely and Hicks (2000) point out several shortcomings of current knowledge of dry
18 deposition. Among those shortcomings are difficulties in representing dry deposition over
19 varying terrain where horizontal advection plays a significant role in determining the magnitude
20 of ra and difficulties in adequately determining a Vd for extremely stable conditions such as
21 those occurring at night (e.g., Mahrt, 1998). Under the best of conditions, when a model is
22 exercised over a relatively small area where dry deposition measurements have been made,
23 models still commonly show uncertainties at least as large as ±30% (e.g., Massman et al., 1994;
24 Brook et al., 1996; Padro, 1996). Wesely and Hicks (2000) state that an important result of these
25 comparisons is that the current level of sophistication of most dry deposition models is relatively
26 low, and that deposition estimates therefore must rely heavily on empirical data. Still larger
27 uncertainties exist when the surface features in the built environment are not well known or
28 when the surface comprises a patchwork of different surface types, as is common in the eastern
29 United States.
30 The initial conditions, i.e., the concentration fields of all species computed by a model,
31 and the boundary conditions, i.e., the concentrations of species along the horizontal and upper
March 2008 AX2-64 DRAFT-DO NOT QUOTE OR CITE
-------
1 boundaries of the model domain throughout the simulation must be specified at the beginning of
2 the simulation. It would be best to specify initial and boundary conditions according to
3 observations. However, data for vertical profiles of most species of interest are sparse. The
4 results of model simulations over larger, preferably global, domains can also be used. As may be
5 expected, the influence of boundary conditions depends on the lifetime of the species under
6 consideration and the time scales for transport from the boundaries to the interior of the model
7 domain (Liu et al., 2001 b).
8 Each of the model components described above has an associated uncertainty, and the
9 relative importance of these uncertainties varies with the modeling application. The largest
10 errors in photochemical modeling are still thought to arise from the meteorological and
11 emissions inputs to the model (Russell and Dennis, 2000). Within the model itself, horizontal
12 advection algorithms are still thought to be significant source of uncertainty (e.g., Chock and
13 Winkler, 1994), though more recently, those errors are thought to have been reduced (e.g.,
14 Odman and Ingram, 1996). There are also indications that problems with mass conservation
15 continue to be present in photochemical and meteorological models (e.g., Odman and Russell,
16 1999); these can result in significant simulation errors. The effects of errors in initial conditions
17 can be minimized by including several days "spin-up" time in a simulation to allow the model to
18 be driven by emitted species before the simulation of the period of interest begins.
19 While the effects of poorly specified boundary conditions propagate through the model's
20 domain, the effects of these errors remain undetermined. Because many meteorological
21 processes occur on spatial scales which are smaller than the model grid spacing (either
22 horizontally or vertically) and thus are not calculated explicitly, parameterizations of these
23 processes must be used and these introduce additional uncertainty.
24 Uncertainty also arises in modeling the chemistry of Os formation because it is highly
25 nonlinear with respect to NOx concentrations. Thus, the volume of the grid cell into which
26 emissions are injected is important because the nature of 63 chemistry (i.e., 63 production or
27 titration) depends in a complicated way on the concentrations of the precursors and the OH
28 radical as noted earlier. The use of ever-finer grid spacing allows regions of O3 titration to be
29 more clearly separated from regions of Os production. The use of grid spacing fine enough to
30 resolve the chemistry in individual power-plant plumes is too demanding of computer resources
31 for this to be attempted in most simulations. Instead, parameterizations of the effects of sub-
March 2008 AX2-65 DRAFT-DO NOT QUOTE OR CITE
-------
1 grid-scale processes such as these must be developed; otherwise serious errors can result if
2 emissions are allowed to mix through an excessively large grid volume before the chemistry step
3 in a model calculation is performed. In light of the significant differences between atmospheric
4 chemistry taking place inside and outside of a power plant plume (e.g., Ryerson et al., 1998 and
5 Sillman, 2000), inclusion of a separate, meteorological module for treating large, tight plumes is
6 necessary. Because the photochemistry of Os and many other atmospheric species is nonlinear,
7 emissions correctly modeled in a tight plume may be incorrectly modeled in a more dilute plume.
8 Fortunately, it appears that the chemical mechanism used to follow a plume's development need
9 not be as detailed as that used to simulate the rest of the domain, as the inorganic reactions are
10 the most important in the plume see (e.g., Kumar and Russell, 1996). The need to include
11 explicitly plume-in-grid chemistry only down to the level of the smallest grid disappears if one
12 uses the adaptive grid approach mentioned previously, though such grids are more
13 computationally intensive. The differences in simulations are significant because they can lead
14 to significant differences in the calculated sensitivity of Os to its precursors (e.g., Sillman et al.,
15 1995).
16 Because the chemical production and loss terms in the continuity equations for individual
17 species are coupled, the chemical calculations must be performed iteratively until calculated
18 concentrations converge to within some preset criterion. The number of iterations and the
19 convergence criteria chosen also can introduce error.
20
21 Intra-urban Scale Dispersion Modeling
22 The grid spacing in regional chemistry transport models is typically too coarse to resolve
23 spatial variations on the neighborhood scale. CTM grid spacing is typically 4 km at best,
24 although there are efforts to increase the horizontal resolution to 1 km. The interface between
25 regional scale models and models of personal exposure described in Annex 3, Section 3.7 is
26 provided by smaller scale dispersion models. AERMOD is a steady-state plume model that was
27 formulated as a replacement to the ISC3 dispersion model. In the stable boundary layer (SBL), it
28 assumes the concentration distribution to be Gaussian in both the vertical and horizontal. In the
29 convective boundary layer (CBL), the horizontal distribution is also assumed to be Gaussian, but
30 the vertical distribution is described with a bi-Gaussian probability density function (pdf).
31 AERMOD has provisions to be applied to flat and complex terrain, and multiple source types
32 (including, point, area and volume sources) in both urban and rural areas. It incorporates air
March 2008 AX2-66 DRAFT-DO NOT QUOTE OR CITE
-------
1 dispersion based on planetary boundary layer turbulence structure and scaling concepts, and it is
2 meant to treat both surface and elevated sources, and both simple and complex terrain in both
3 rural and urban areas. The dispersion of emissions from line sources, such as highways is treated
4 as the sum of emissions from a number of point sources placed side by side. However,
5 emissions are usually not in steady state and there are different functional relationships between
6 buoyant plume rise in point and line sources. In contrast, there are models that are non-steady
7 state and can incorporate plume rise explicitly from different types of sources. For example,
8 CALPUFF (http://www.src.com/calpuff/calpuffl .htm) is a non-steady-state puff dispersion
9 model that simulates the effects of time- and space-varying meteorological conditions on
10 pollution transport, transformation, and removal and has provisions for calculating dispersion
11 from surface sources.
12
13 Global Scale CTMs
14 The importance of global transport of Os and Os precursors and their contribution to
15 regional Os levels in the United States is slowly becoming apparent. There are presently on the
16 order of 20 three-dimensional global models that have been developed by various groups to
17 address problems in tropospheric chemistry. These models resolve synoptic meteorology,
18 (VNOx-CO-hydrocarbon photochemistry, have parameterizations for wet and dry deposition,
19 and parameterize sub-grid scale vertical mixing processes such as convection. Global models
20 have proven useful for testing and advancing scientific understanding beyond what is possible
21 with observations alone. For example, they can calculate quantities of interest that cannot be
22 measured directly, such as the export of pollution from one continent to the global atmosphere or
23 the response of the atmosphere to future perturbations to anthropogenic emissions.
24 Global simulations are typically conducted at a horizontal resolution of about 200 km2.
25 Simulations of the effects of transport from long-range transport link multiple horizontal
26 resolutions from the global to the local scale. Finer resolution will only improve scientific
27 understanding to the extent that the governing processes are more accurately described at that
28 scale. Consequently, there is a critical need for observations at the appropriate scales to evaluate
29 the scientific understanding represented by the models.
30 During the recent IPCC-AR4 tropospheric chemistry study coordinated by the European
31 Union project Atmospheric Composition Change: the European Network of excellence
32 (ACCENT), 26 atmospheric CTMs were used to estimate the impacts of three emissions
March 2008 AX2-67 DRAFT-DO NOT QUOTE OR CITE
-------
1 scenarios on global atmospheric composition, climate, and air quality in 2030 (Dentener et al.,
2 2006a). All models were required to use anthropogenic emissions developed at IIASA (Dentener
3 et al., 2005) and GFED version 1 biomass burning emissions (Van der Werf et al., 2003) as
4 described in Stevenson et al. (2006). The base simulations from these models were evaluated
5 against a suite of present-day observations. Most relevant to this assessment report are the
6 evaluations with ozone and NC>2, and for nitrogen and sulfur deposition (Stevenson et al., 2006;
7 Van Noije et al., 2006; Dentener et al., 2006a), which are summarized briefly below.
8 An analysis of the standard deviation of zonal mean and tropospheric column 63 reveals
9 large inter-model variability in the tropopause region and throughout the polar troposphere,
10 likely reflecting differences in model tropopause levels and the associated stratospheric injection
11 of O3 to the troposphere (Stevenson et al., 2006). Ozone distributions in the tropics also exhibit
12 large standard deviations (-30%), particularly as compared to the mid-latitudes (-20%),
13 indicating larger uncertainties in the processes that influence ozone in the tropics: deep tropical
14 convection, lightning NOx, isoprene emissions and chemistry, and biomass burning emissions
15 (Stevenson et al., 2006).
16 Stevenson et al., (2006) found that the model ensemble mean (MEM) typically captures
17 the observed seasonal cycles to within one standard deviation. The largest discrepancies
18 between the MEM and observations include: (1) an underestimate of the amplitude of the
19 seasonal cycle at 30°-90°N with a 10 ppbv overestimate of winter ozone, possibly due to the lack
20 of a seasonal cycle in anthropogenic emissions or to shortcomings in the stratospheric influx of
21 Os, and (2) an overestimate of Os throughout the northern tropics. However, the MEM was
22 found to capture the observed seasonal cycles in the southern hemisphere, suggesting that the
23 models adequately represent biomass burning and natural emissions.
24 The mean present-day global ozone budget across the current generation of CTMs differs
25 substantially from that reported in the IPCC TAR, with a 50% increase in the mean chemical
26 production (to 5100 Tg O3 yr"1), a 30% increase in the chemical and deposition loss terms (to
27 4650 and 1000 Tg Oj yr'1, respectively) and a 30% decrease in the mean stratospheric input flux
28 (to 550 Tg Oj yr"1) (Stevenson et al., 2006). The larger chemical terms as compared to the IPCC
29 TAR are attributed mainly to higher NOx (as well as an equatorward shift in distribution) and
30 isoprene emissions, although more detailed NMHC schemes and/or improved representations of
March 2008 AX2-68 DRAFT-DO NOT QUOTE OR CITE
-------
1 photolysis, convection, and stratospheric-tropospheric exchange may also contribute (Stevenson
2 et al., 2006).
3 A subset of 17 of the 26 models used in the Stevenson et al. (2006) study was used to
4 compare with three retrievals of NO2 columns from the GOME instrument (van Noije et al.,
5 2006) for the year 2000. The higher resolution models reproduce the observed patterns better,
6 and the correlation among simulated and retrieved columns improved for all models when
7 simulated values are smoothed to a 5° x 5° grid, implying that the models do not accurately
8 reproduce the small-scale features of NO2 (Van Noije et al., 2006). Van Noije et al. (2006)
9 suggest that variability in simulated NO2 columns may reflect a model differences in OH
10 distributions and the resulting NOx lifetimes, as well as differences in vertical mixing which
11 strongly affect partitioning between NO and NO2. Overall, the models tend to underestimate
12 concentrations in the retrievals in industrial regions (including the eastern United States) and
13 overestimate them in biomass burning regions (Van Noije et al., 2006).
14 Over the eastern United States, and industrial regions more generally, the spread in
15 absolute column abundances is generally larger among the retrievals than among the models,
16 with the discrepancy among the retrievals particularly pronounced in winter (Van Noije et al.,
17 2006), suggesting that the models are biased low, or that the European retrievals may be biased
18 high as the Dalhousie/SAO retrieval is closer to the model estimates. The lack of seasonal
19 variability in fossil fuel combustion emissions may contribute to a wintertime model
20 underestimate (Van Noije et al., 2006) that is manifested most strongly over Asia. In biomass
21 burning regions, the models generally reproduce the timing of the seasonal cycle of the
22 retrievals, but tend to overestimate the seasonal cycle amplitude, partly due to lower values in the
23 wet season, which may reflect an underestimate in wet season soil NO emissions (Van Noije
24 et al., 2006, Jaegle et al., 2004, 2005).
25
26 Deposition in Global CTMs
27 Both wet and dry deposition are highly parameterized in global CTMs. While all current
28 models implement resistance schemes for dry deposition, the generated Vd generated from
29 different models can vary highly across terrains (Stevenson et al., 2006). The accuracy of wet
30 deposition in global CTMs is tied to spatial and temporal distribution of model precipitation and
31 the treatment of chemical scavenging. Dentener et al. (2006b) compared wet deposition across
March 2008 AX2-69 DRAFT-DO NOT QUOTE OR CITE
-------
1 23 models with available measurements around the globe. Figures AX2.7-1 and AX2.7-2 below
2 extract the results of a comparison of the 23-model mean versus observations from Dentener
3 et al. (2006b) over the eastern United States for nitrate and sulfate deposition, respectively. The
4 mean model results are strongly correlated with the observations (r > 0.8), and usually capture
5 the magnitude of wet deposition to within a factor of 2 over the eastern United States (Dentener
6 et al., 2006b). Dentener et al. (2006b) conclude that 60-70% of the participating models capture
7 the measurements to within 50% in regions with quality controlled observations. This study then
8 identified world regions receiving >1000 mg (N) nT2 yr"1 (the "critical load") and found that
9 20% of the natural vegetation (non-agricultural) in the United States is exposed to nitrogen
10 deposition in excess of the critical load threshold (Dentener et al., 2006b).
11
12 Modeling the Effects of Convection
13 The effects of deep convection can be simulated using cloud-resolving models, or in
14 regional or global models in which the convection is parameterized. The Goddard Cumulus
15 Ensemble (GCE) model (Tao and Simpson, 1993) has been used by Pickering et al. (1991,
16 1992a,b, 1993, 1996), Scala et al. (1990), and Stenchikov et al. (1996) in the analysis of
17 convective transport of trace gases. The cloud model is nonhydrostatic and contains a detailed
18 representation of cloud microphysical processes. Two- and three-dimensional versions of the
19 model have been applied in transport analyses. The initial conditions for the model are usually
20 from a sounding of temperature, water vapor and winds representative of the region of storm
21 development. Model-generated wind fields can be used to perform air parcel trajectory analyses
22 and tracer advection calculations.
23 Such methods were used by Pickering et al. (1992b) to examine transport of urban
24 plumes by deep convection. Transport of an Oklahoma City plume by the 10-11 June 1985
25 PRE-STORM squall line was simulated with the 2-D GCE model. This major squall line passed
26 over the Oklahoma City metropolitan area, as well as more rural areas to the north. Chemical
27 observations ahead of the squall line were conducted by the PRE-STORM aircraft. In this event,
28 forward trajectories from the boundary layer at the leading edge of the storm showed that almost
29 75% of the low level inflow was transported to altitudes exceeding 8 km. Over 35% of the air
30 parcels reached altitudes over 12 km. Tracer transport calculations were performed for CO,
31 NOx, Os, and hydrocarbons. Rural boundary layer NOx was only 0.9 ppbv, whereas the urban
March 2008 AX2-70 DRAFT-DO NOT QUOTE OR CITE
-------
600
400
o
T3
O
200
0
'/ -. '1 . x? '•
/ '*. .*/
Ave- model: 227 Ave. Mea. 195 r: 0.82 n = 226
2 param. fit: y = 51.1 + 0.90.x
1 param. fit: y = 1.08x
Percentage within ± 50%: 74.8
0
200 400
Measurement
600
Figure AX2.7-1.
Scatter plot of total nitrate (HNOs plus aerosol nitrate) wet deposition
(mg(N)m2yr~1) of the mean model versus measurements for the North
American Deposition Program (NADP) network. Dashed lines
indicate factor of 2. The gray line is the result of a linear regression
fitting through 0.
Source: Dentener et al. (2006b).
1 plume contained about 3 ppbv. In the rural case, mixing ratios of 0.6 ppbv were transported up
2 to 11 km. Cleaner air descended at the rear of the storm lowering NOx at the surface from 0.9 to
3 0.5 ppbv. In the urban plume, mixing ratios in the updraft core reached 1 ppbv between 14 and
4 15 km. At the surface, the main downdraft lowered the NOx mixing ratios from 3 to 0.7 ppbv.
5 Regional chemical transport models have been used for applications such as simulations
6 of photochemical Os production, acid deposition, and fine PM. Walcek et al. (1990) included a
March 2008
AX2-71
DRAFT-DO NOT QUOTE OR CITE
-------
1000
800 -
600 -
TJ
O
t
5 f
400 -
200
Ave, model: 383 Ave. Meas: 322 r; 0.87 n = 226
2 param fit: y = 114.0 + 0.77x
1 param fit: y ™ 1.00x
percentage within t 50%; 66,0
200
400 600
Measurement
800
1000
2 J
Figure AX2.7-2. Same as Figure AX2.7-1 but for sulfate wet deposition (mg(S)m yr
Source: Dentener et al. (2006b).
1 parameterization of cloud-scale aqueous chemistry, scavenging, and vertical mixing in the
2 chemistry model of Chang et al. (1987). The vertical distribution of cloud microphysical
3 properties and the amount of sub-cloud-layer air lifted to each cloud layer are determined using a
4 simple entrainment hypothesis (Walcek and Taylor, 1986). Vertically integrated O3 formation
5 rates over the northeast U. S. were enhanced by -50% when the in-cloud vertical motions were
6 included in the model.
7 Wang et al. (1996) simulated the 10-11 June 1985 PRE-STORM squall line with the
8 NCAR/Penn State Mesoscale Model (MM5) (Grell et al., 1994; Dudhia, 1993). Convection was
9 parameterized as a sub-grid-scale process in MM5 using the Kain and Fritsch (1993) scheme.
March 2008
AX2-72
DRAFT-DO NOT QUOTE OR CITE
-------
1 Mass fluxes and detrainment profiles from the convective parameterization were used along with
2 the 3-D wind fields in CO tracer transport calculations for this convective event.
3 Convective transport in global chemistry and transport models is treated as a sub-grid-
4 scale process that is parameterized typically using cloud mass flux information from a general
5 circulation model or global data assimilation system. While GCMs can provide data only for a
6 "typical" year, data assimilation systems can provide "real" day-by-day meteorological
7 conditions, such that CTM output can be compared directly with observations of trace gases.
8 The NASA Goddard Earth Observing System Data Assimilation System (GEOS-1 DAS and
9 successor systems; Schubert et al., 1993; Bloom et al., 1996; Bloom et al., 2005) provides
10 archived global data sets for the period 1980 to present, at 2° x 2.5° or better resolution with
11 20 layers or more in the vertical. Deep convection is parameterized with the Relaxed
12 Arakawa-Schubert scheme (Moorthi and Suarez, 1992) in GEOS-1 and GEOS-3 and with the
13 Zhang and McFarlane (1995) scheme in GEOS-4. Pickering et al. (1995) showed that the cloud
14 mass fluxes from GEOS-1 DAS are reasonable for the 10-11 June 1985 PRE-STORM squall line
15 based on comparisons with the GCE model (cloud-resolving model) simulations of the same
16 storm. In addition, the GEOS-1 DAS cloud mass fluxes compared favorably with the regional
17 estimates of convective transport for the central United States presented by Thompson et al.
18 (1994). However, Allen et al. (1997) have shown that the GEOS-1 DAS overestimates the
19 amount and frequency of convection in the tropics and underestimates the convective activity
20 over midlatitude marine storm tracks.
21 Global models with parameterized convection and lightning have been run to examine
22 the roles of these processes over North America. Lightning contributed 23% of upper
23 tropospheric NOy over the SONEX region according to the UMD-CTM modeling analysis of
24 Allen et al. (2000). During the summer of 2004 the NASA Intercontinental Chemical Transport
25 Experiment - North America (INTEX-NA) was conducted primarily over the eastern two-thirds
26 of the United States, as a part of the International Consortium for Atmospheric Research on
27 Transport and Transformation (ICARTT). Deep convection was prevalent over this region
28 during the experimental period. Cooper et al. (2006) used a particle dispersion model simulation
29 for NOx to show that 69-84% of the upper tropospheric Os enhancement over the region in
30 Summer 2004 was due to lightning NOx. The remainder of the enhancement was due to
31 convective transport of Os from the boundary layer or other sources of NOx. Hudman et al.
March 2008 AX2-73 DRAFT-DO NOT QUOTE OR CITE
-------
1 (2007) used a GEOS-Chem model simulation to show that lightning was the dominant source of
2 upper tropospheric NOx over this region during this period. Approximately 15% of North
3 American boundary layer NOx emissions were shown to have been vented to the free
4 troposphere over this region based on both the observations and the model.
5
6 AX2.7.2 CTM Evaluation
7 The comparison of model predictions with ambient measurements represents a critical
8 task for establishing the accuracy of photochemical models and evaluating their ability to serve
9 as the basis for making effective control strategy decisions. The evaluation of a model's
10 performance, or its adequacy to perform the tasks for which it was designed can only be
11 conducted within the context of measurement errors and artifacts. Not only are there analytical
12 problems, but there are also problems in assessing the representativeness of monitors at ground
13 level for comparison with model values which represent typically an average over the volume of
14 a grid box.
15 Evaluations of CMAQ are given in Arnold et al. (2003) and Fuentes and Raftery (2005).
16 Discrepancies between model predictions and observations can be used to point out gaps in
17 current understanding of atmospheric chemistry and to spur improvements in parameterizations
18 of atmospheric chemical and physical processes. Model evaluation does not merely involve a
19 straightforward comparison between model predictions and the concentration field of the
20 pollutant of interest. Such comparisons may not be meaningful because it is difficult to
21 determine if agreement between model predictions and observations truly represents an accurate
22 treatment of physical and chemical processes in the CTM or the effects of compensating errors in
23 complex model routines. Ideally, each of the model components (emissions inventories,
24 chemical mechanism, meteorological driver) should be evaluated individually. However, this is
25 rarely done in practice.
26 Chemical transport models for Os formation at the urban/regional scale have traditionally
27 been evaluated based on their ability to simulate correctly 03. A series of performance statistics
28 that measure the success of individual model simulations to represent the observed distribution
29 of ambient 63, as represented by a network of surface measurements at the urban scale were
30 recommended by the EPA (U.S. Environmental Protection Agency, 1991; see also Russell and
31 Dennis, 2000). These statistics consist of the following:
March 2008 AX2-74 DRAFT-DO NOT QUOTE OR CITE
-------
1 • Unpaired peak 63 concentration within a metropolitan region (typically for a
2 single day).
3 • Normalized bias equal to the summed difference between model and measured hourly
4 concentrations divided by the sum of measured hourly concentrations.
5 • Normalized gross error, equal to the summed unsigned (absolute value) difference
6 between model and measured hourly concentrations divided by the sum of measured
7 hourly concentrations.
8 Unpaired peak prediction accuracy, Au
=
C (x* t')
9 ^o<-*"lSmax (AX2.7-1)
10 Normalized bias, D;
/ N {Cn(Xi,t)-Cn(Xj,t)}
D = — y f ' ° ' tt = 1,24.
11 N i=1 Co(*i.n (AX2.7-2)
12 Gross error, Ed (for hourly observed values of Os >60 ppb)
Ed =-Lj[ I Cp(Xi,t)-C0(Xi,t) \ f=l 24
13 ^;W Co(*/.0 (AX2.7-3)
14 The following performance criteria for regulatory models were recommended in U.S.
15 Environmental Protection Agency (1991): unpaired peak Oj to within ±15% or ±20%;
16 normalized bias within ± 5% to ± 15%; and normalized gross error less than 30% to 35%, but
17 only when Os the concentration >60 ppb. This can lead to difficulties in evaluating model
18 performance since nighttime and diurnal cycles are ignored. A major problem with this method
19 of model evaluation is that it does not provide any information about the accuracy of Os-
20 precursor relations predicted by the model. The process of Os formation is sufficiently complex
21 that models can predict 63 correctly without necessarily representing the 63 formation process
22 properly. If the 63 formation process is incorrect, then the modeled source-receptor relations
23 will also be incorrect.
24 Studies by Sillman et al. (1995, 2003), Reynolds et al. (1996), and Pierce et al. (1998)
25 have identified instances in which different model scenarios can be created with very different
March 2008 AX2-75 DRAFT-DO NOT QUOTE OR CITE
-------
1 (Vprecursor sensitivity, but without significant differences in the predicted 63 fields.
2 Figures AX2.7-3a,b provide an example. Referring to the (VNOx-VOC isopleth plot (Figure
3 AX2.7-4), it can be seen that similar O3 concentrations can be found for photochemical
4 conditions that have very different sensitivity to NOx and VOCs.
5 Global-scale CTMs have generally been evaluated by comparison with measurements for
6 a wide array of species, rather than just for Os (e.g., Wang et al., 1998; Emmons et al., 2000; Bey
7 et al., 2001; Hess, 2001; Fiore et al., 2002). These have included evaluation of major primary
8 species (NOx, CO, and selected VOCs) and an array of secondary species (HNOs, PAN, H2O2)
9 that are often formed concurrently with Os. Models for urban and regional Os have also been
10 evaluated against a broader ensemble of measurements in a few cases, often associated with
11 measurement intensives (e.g., Jacobson et al., 1996; Lu et al., 1997; Sillman et al., 1998). The
12 results of a comparison between observed and computed concentrations from Jacobson et al.
13 (1996) for the Los Angeles Basin are shown in Figures AX2.7-5a,b.
14 The highest concentrations of primary species usually occur in close proximity to
15 emission sources (typically in urban centers) and at times when dispersion rates are low. The
16 diurnal cycle includes high concentrations at night, with maxima during the morning rush hour,
17 and low concentrations during the afternoon (Figure AX2.7-5a). The afternoon minima are
18 driven by the much greater rate of vertical mixing at that time. Primary species also show a
19 seasonal maximum during winter, and are often high during fog episodes in winter when vertical
20 mixing, is suppressed. By contrast, secondary species such as O3 are typically highest during the
21 afternoon (the time of greatest photochemical activity), on sunny days and during summer.
22 During these conditions, concentrations of primary species may be relatively low. Strong
23 correlations between primary and secondary species are generally observed only in downwind
24 rural areas where all anthropogenic species are simultaneously elevated. The difference in the
25 diurnal cycles of primary species (CO, NOx and ethane) and secondary species (Os, PAN, and
26 HCHO) is evident in Figure AX2.7-5b.
27 Models for urban and regional chemistry have been evaluated less extensively than
28 global-scale models in part because the urban/regional context presents a number of difficult
29 challenges. Global-scale models typically represent continental-scale events and can be
30 evaluated effectively against a sparse network of measurements. By contrast, urban/regional
March 2008 AX2-76 DRAFT-DO NOT QUOTE OR CITE
-------
.Q
Q.
Q.
25
20
~ 10
o
3
"O 5
tt
O
0
-5
100 120 140 160
Ozone (ppb)
180
200
O
35
30
25
20
15
0} 10
CO r-
o 5
0
x x
X
XX
X
X
X
X
100 120 140 160
Ozone (ppb)
180
200
Figure AX2.7-3a,b. Impact of model uncertainty on control strategy predictions for Os for
two days (August lOa and lib, 1992) in Atlanta, GA. The figures
show the predicted reduction in peak Os resulting from 35%
reductions in anthropogenic VOC emissions (crosses) and from 35%
reductions in NOx (solid circles) in a series of model scenarios with
varying base case emissions, wind fields, and mixed layer heights.
Source: Results are plotted from tabulated values published in Sillman et al. (1995, 1997).
March 2008
AX2-77
DRAFT-DO NOT QUOTE OR CITE
-------
~ 10.00
E
o
c
o
LU
x
O
Figure AX2.7-4.
3.16
o
es
O
C, 1.0
0)
•+•»
(0
0.316
0.1
a.
a.
CO
o
1.0 3.16 10.0 31.6 100.0
VOC Emission Rate (1012 molec. cm-2 s-1)
Ozone isopleths (ppb) as a function of the average emission rate for
NOx and VOC (1012 molec. cm"2 s"1) in zero dimensional box model
calculations. The isopleths (solid lines) represent conditions during
the afternoon following 3-day calculations with a constant emission
rate, at the hour corresponding to maximum Os. The ridge line
(shown by solid circles) lies in the transition from NOx-saturated to
NOx-limited conditions.
1 models are critically dependent on the accuracy of local emission inventories and event-specific
2 meteorology, and must be evaluated separately for each urban area that is represented.
3 The evaluation of urban/regional models is also limited by the availability of data.
4 Measured NOx and speciated VOC concentrations are widely available through the EPA PAMs
5 network, but questions have been raised about the accuracy of those measurements and the data
6 have not yet been analyzed thoroughly. Evaluation of urban/regional models versus
7 measurements has generally relied on results from a limited number of field studies in the United
8 States. Short-term, research-grade measurements for species relevant to Os formation, including
March 2008
AX2-78
DRAFT-DO NOT QUOTE OR CITE
-------
E
Q.
a
_o
+!
m
en
c
'x
i
E
a
a.
03
a:
en
.E
'x
E
Q.
a.
a
IT
en
c
0.30
0.25
0.20
0.15
0.10
0.05
0.00
0.30
0.25
0.20
0.15
0.10
0.05
0.00
6
5
4
3
2
1
0
Reseda
O3(g)
Predicted
Observed
8 16 24 32 40 48 56 64 72
Hour After First Midnight
Reseda
NOX (g)
Predicted
Observed
16 24 32 40 48 56 64 72
Hour After First Midnight
Riverside
C0(g)
Predicted
Observed
16 24 32 40 48 56 64 72
Hour After First Midnight
Figure AX2.7-5a.
Time series for measured gas-phase species in comparison with results
from a photochemical model. The dashed lines represent
measurements, and solid lines represent model predictions (in parts
per million, ppmv) for August 26-28,1988 at sites in southern
California. The horizontal axis represents hours past midnight,
August 25. Results represent Os and NOx at Reseda, and CO at
Riverside.
Source: Jacobsonetal. (1996).
March 2008
AX2-79
DRAFT-DO NOT QUOTE OR CITE
-------
0.060
| 0.050
Q.
S 0.040
O
'•JS 0.030
IT
0> 0.020
g 0.010
0.000
0.030
| 0.025
Q.
3: 0.020
O
"• 0.015
TO 0.010
c.
~ 0.005
0.000
0.020
£ 0.016
Q.
^O 0.012
'^
re
** 0.008
O)
1 0.004
s
0.000
Claremont
Ethane (g)
Predicted
O Observed
I
I
I
I
I
16 24 32 40 48 56
Hour After First Midnight
64 72
Claremont
Formaldehyde (g)
Predicted
O Observed
I
I
I
I
I
16 24 32 40 48 56
Hour After First Midnight
64 72
Los Angeles
PAN (g)
Predicted
Observed
JL
16 24 32 40 48 56
Hour After First Midnight
64 72
Figure AX2.7-5b.
Time series for measured gas-phase species in comparison with results
from a photochemical model. The circles represent measurements,
and solid lines represent model predictions (in parts per million,
ppmv) for August 26-28,1988 at sites in southern California. The
horizontal axis represents hours past midnight, August 25. Results
represent ethane and formaldehyde at Claremont, and PAN at Los
Angeles.
Source: Jacobsonetal. (1996).
March 2008
AX2-80
DRAFT-DO NOT QUOTE OR CITE
-------
1 VOCs, NOx, PAN, HNOs, and H2O2 are also available at selected rural and remote sites (e.g.,
2 Daum et al., 1990, 1996; Martin et al., 1997; Young et al., 1997; Thompson et al., 2000; Hoell
3 et al., 1997, 1999; Fehsenfeld et al., 1996; Emmons et al., 2000; Hess, 2001; Carroll et al., 2001).
4 The equivalent measurements are available for some polluted rural sites in the eastern United
5 States, but only at a few urban locations (Meagher et al., 1998; Hiibler et al., 1998; Kleinman
6 et al., 2000, 2001; Fast et al., 2002; new SCAQS-need reference). Extensive measurements have
7 also been made in Vancouver (Steyn et al., 1997) and in several European cities (Staffelbach
8 et al., 1997; Prevot et al., 1997, Dommen et al., 1999; Geyer et al., 2001; Thielman et al., 2001;
9 Martilli et al., 2002; Vautard et al., 2002).
10 The results of straightforward comparisons between observed and predicted
11 concentrations of O3 can be misleading because of compensating errors, although this possibility
12 is diminished when a number of species are compared. Ideally, each of the main modules of a
13 CTM system (for example, the meteorological model and the chemistry and radiative transfer
14 routines) should be evaluated separately. However, this is rarely done in practice. To better
15 indicate how well physical and chemical processes are being represented in the model,
16 comparisons of relations between concentrations measured in the field and concentrations
17 predicted by the model can be made. These comparisons could involve ratios and correlations
18 between species. For example, correlation coefficients could be calculated between primary
19 species as a means of evaluating the accuracy of emission inventories or between secondary
20 species as a means of evaluating the treatment of photochemistry in the model. In addition,
21 spatial relations involving individual species (correlations, gradients) can also be used as a means
22 of evaluating the accuracy of transport parameterizations. Sillman and He (2002) examined
23 differences in correlation patterns between O3 and NOz in Los Angeles, CA, Nashville, TN, and
24 various sites in the rural United States. Model calculations (Figure AX2.7-6) show differences in
25 correlation patterns associated with differences in the sensitivity of O3 to NOx and VOCs.
26 Primarily NOx-sensitive (NOx-limited) areas in models show a strong correlation between O3
27 and NOz with a relatively steep slope, while primarily VOC-sensitive (NOx-saturated) areas in
28 models show lower O3 for a given NOZ and a lower O3-NOZ slope. They found that differences
29 found in measured data ensembles were matched by predictions from chemical transport models.
30 Measurements in rural areas in the eastern United States show differences in the pattern of
31 correlations for O3 versus NOz between summer and autumn (Jacob et al., 1995; Hirsch et al.,
March 2008 AX2-81 DRAFT-DO NOT QUOTE OR CITE
-------
250
0
Figure AX2.7-6.
20
NOZ (ppb)
Correlations for Os versus NOz (NOy-NOx) in ppb from chemical
transport models for the northeast corridor, Lake Michigan,
Nashville, the San Joaquin Valley, and Los Angeles. Each location is
classified as NOx-limited or NOx-sensitive (circles), NOx-saturated or
VOC-sensitive (crosses), mixed or with near-zero sensitivity (squares),
and dominated by NOX titration (asterisks) based on the model
response to reduced NOx and VOC.
Source: Sillman and He (2002).
1 1996), corresponding to the transition from NOx-limited to NOx-saturated patterns, a feature
2 which is also matched by CTMs.
3 The difference in correlations between secondary species in NOx-limited to NOx-
4 saturated environments can also be used to evaluate the accuracy of model predictions in
5 individual applications. Figures AX2.7-7a and AX2.7-7b show results for two different model
6 scenarios for Atlanta. As shown in the figures, the first model scenario predicts an urban plume
7 with high NOy and Os formation apparently suppressed by high NOy. Measurements show
8 much lower NOy in the Atlanta plume. This error was especially significant because the model
March 2008
AX2-82
DRAFT-DO NOT QUOTE OR CITE
-------
0
200
10 20
NOy (Ppb)
30
Figure AX.7-7a,b.
Evaluation of model versus measured Os versus NOy for two model
scenarios for Atlanta. The model values are classified as NOx-limited
(circles), NOx-saturated (crosses), or mixed or with low sensitivity to
NOx (squares). Diamonds represent aircraft measurements.
Source: Sillmanetal. (1997).
March 2008
AX2-83
DRAFT-DO NOT QUOTE OR CITE
-------
1 locations sensitive to NOx. The second model scenario (with primarily NOx-sensitive
2 conditions) shows much better agreement with measured values. Figure AX2.7-8a,b shows
3 model-measurement comparisons for secondary species in Nashville, showing better agreement
4 with measured with conditions. Greater confidence in the predictions made by CTMs will be
5 gained by the application of techniques such as these on a more routine basis.
6 The ability of chemical mechanisms to calculate the concentrations of free radicals under
7 atmospheric conditions was tested in the Berlin Ozone Experiment, BERLIOZ (Volz-Thomas
8 et al., 2003) during July and early August at a site located about 50 km NW of Berlin. (This
9 location was chosen because Os episodes in central Europe are often associated with SE winds.)
10 Concentrations of major compounds such as Os, hydrocarbons, etc., were fixed at
11 observed values. In this regard, the protocol used in this evaluation is an example of an
12 observationally high NOy were not sensitive to NOx, while locations with lower NOy were
13 primarily based method. Figure AX2.7-9 compares the concentrations of RO2, HO2, and OH
14 radicals predicted by RACM and MCM with observations made by the laser-induced
15 fluorescence (LIF) technique and by matrix isolation ESR spectroscopy (MTESR). Also shown
16 are the production rates of Os calculated using radical concentrations predicted by the
17 mechanisms and those obtained by measurements, and measurements of NOx concentrations.
18 As can be seen, there is good agreement between measurements of RO2, HO2, OH, radicals with
19 values predicted by both mechanisms at high concentrations of NOx (>10 ppb). However, at
20 lower NOX concentrations, both mechanisms substantially overestimate OH concentrations and
21 moderately overestimate HO2 concentrations. Agreement between models and measurements is
22 generally better for organic peroxy radicals, although the MCM appears to overestimate their
23 concentrations somewhat. In general, the mechanisms reproduced the HO2 to OH and RO2 to
24 OH ratios better than the individual measurements. The production of O?, was found to increase
25 linearly with NO (for NO < 0.3 ppb) and to decrease with NO (for NO > 0.5 ppb).
26 OH and HO2 concentrations measured during the PM2.s Technology Assessment and
27 Characterization Study conducted at Queens College in New York City in the summer of 2001
28 were also compared with those predicted by RACM (Ren et al., 2003). The ratio of observed to
29 predicted HO2 concentrations over a diurnal cycle was 1.24 and the ratio of observed to predicted
30 OH concentrations was about 1.10 during the day, but the mechanism significantly
31 underestimated OH concentrations during the night.
March 2008 AX2-84 DRAFT-DO NOT QUOTE OR CITE
-------
160
140
0
160
140
10
20
NOZ (ppb)
30
40
10 20 30
2H2O2 + NOZ (ppb)
Figure AX2.7-8a,b. Evaluation of model versus: (a) measured Os versus NOz and (b) Os
versus the sum 2H2O2 + NOz for Nashville, TN. The model values are
classified as NOx-limited (gray circles), NOx-saturated (X's), mixed or
near-zero sensitivity (squares), or dominated by NOX titration (filled
circles). Diamonds represent aircraft measurements.
Source: Sillmanetal. (1998).
March 2008
AX2-85
DRAFT-DO NOT QUOTE OR CITE
-------
0
«?" 6
I 4
O
£. 2
CM
O
x 0
| 10
CD
I 5
O 0
£• 10
B
a.
= 5
CO
O
r o
t 20
Q.
5 10
0
O LIF
• MiESR
VVV
J(d1D)*106(s-1)
8
10 12 14
UT 20.7.98
16
Figure AX2.7-9.
Time series of concentrations of ROi, HOi, and OH radicals, local Os
photochemical production rate and concentrations of NOX from
measurements made during BERLIOZ. Also shown are comparisons
with results of photochemical box model calculations using the RACM
and MCM chemical mechanisms.
Source: Volz-Thomas et al. (2003).
March 2008
AX2-86
DRAFT-DO NOT QUOTE OR CITE
-------
1 AX2.8 SAMPLING AND ANALYSIS OF NITROGEN AND
2 SULFUR OXIDES
o
4 AX2.8.1 Availability and Accuracy of Ambient Measurements for NOy
5 Section AX2.8.1-AX2.8.4 focus on current methods and on promising new technologies,
6 but no attempt is made here to cover the extensive development of these methods or of methods
7 such as wet chemical techniques, no longer in widespread use. More detailed discussions of
8 these methods may be found elsewhere (U.S. Environmental Protection Agency, 1993, 1996).
9 McClenny (2000), Parrish and Fehsenfeld (2000), and Clemitshaw (2004) reviewed methods for
10 measuring NOx and NOy compounds. Discussions in Sections 2.8.1-2.8.4 center on
11 chemiluminescence and optical Federal Reference and Equivalent Methods (FRM and FEM,
12 respectively).
13 The use of methods such as observationally based methods or source apportionment
14 models, either as stand-alone methods or as a basis for evaluating chemical transport models, is
15 often limited by the availability and accuracy of measurements. Measured NOx and speciated
16 VOC concentrations are widely available in the United States through the PAMS network.
17 However, challenges have been raised about both the accuracy of the measurements and their
18 applicability.
19 The PAMs network currently includes measured NO and NOx. However, Cardelino and
20 Chameides (2000) reported that measured NO during the afternoon was frequently at or below
21 the detection limit of the instruments (1 ppb), even in large metropolitan regions (Washington,
22 DC; Houston, TX; New York, NY). Nitric dioxide measurements are made with commercial
23 chemilluminescent detectors with hot molybdenum converters. However, these measurements
24 typically include a wide variety of other reactive N species, such as organic nitrates in addition to
25 NOx, and cannot be interpreted as a "pure" NOx measurement (see summary in Parrish and
26 Fehsenfeld, 2000). Detection of these species can be considered an interference or a cross
27 sensitivity useful for understanding the chemistry of the air.
28 Total reactive nitrogen (NOy) is included in the PAMS network only at a few sites. The
29 possible expansion of PAMS to include more widespread NOY measurements has been
30 suggested (McClenny, 2000). NOy measurements are also planned for inclusion in the NCore
31 network (U.S. Environmental Protection Agency, 2005). A major issue to be considered when
32 measuring NOx and NOy is the possibility that HNOs, a major component of NOy, is sometimes
March 2008 AX2-87 DRAFT-DO NOT QUOTE OR CITE
-------
1 lost in inlet tubes and not measured (Luke et al., 1998; Parrish and Fehsenfeld, 2000). This
2 problem is especially critical if measured NOy is used to identify NOx-limited versus NOx-
3 saturated conditions. The problem is substantially alleviated although not necessarily completely
4 solved by using much shorter inlets on NOy monitors than on NOx monitors and by the use of
5 surfaces less likely to take up HNOs. The correlation between Os and NOy differs for NOx-
6 limited versus NOx-saturated locations, but this difference is driven primarily by differences in
7 the ratio of Os to HNOs. If HNOs were omitted from the NOy measurements, then the
8 measurements would represent a biased estimate and their use would be problematic.
9
10 AX2.8.1.1 Calibration Standards
11 Calibration gas standards of NO, in N2 (certified at concentrations of approximately 5 to
12 40 ppm) are obtainable from the Standard Reference Material (SRM) Program of the National
13 Institute of Standards and Technology (NIST), formerly the National Bureau of Standards
14 (NBS), in Gaithersburg, MD. These SRMs are supplied as compressed gas mixtures at about
15 135 bar (1900 psi) in high-pressure aluminum cylinders containing 800 L of gas at standard
16 temperature and pressure, dry (STPD) National Bureau of Standards, 1975; Guenther et al.,
17 1996). Each cylinder is supplied with a certificate stating concentration and uncertainty. The
18 concentrations are certified to be accurate to ±1 percent relative to the stated values. Because of
19 the resources required for their certification, SRMs are not intended for use as daily working
20 standards, but rather as primary standards against which transfer standards can be calibrated.
21 Transfer stand-alone calibration gas standards of NO in N2 (at the concentrations
22 indicated above) are obtainable from specialty gas companies. Information as to whether a
23 company supplies such mixtures is obtainable from the company, or from the SRM Program of
24 NIST. These NIST Traceable Reference Materials (NTRMs) are purchased directly from
25 industry and are supplied as compressed gas mixtures at approximately 135 bar (1900 psi) in
26 high-pressure aluminum cylinders containing 4,000 L of gas at STPD. Each cylinder is supplied
27 with a certificate stating concentration and uncertainty. The concentrations are certified to be
28 accurate to within ±1 percent of the stated values (Guenther et al., 1996). Additional details can
29 be found in the previous AQCD for Os (U.S. Environmental Protection Agency, 1996).
March 2008 AX2-88 DRAFT-DO NOT QUOTE OR CITE
-------
1 AX2.8.1.2 Measurement of Nitric Oxide
2
3 Gas-phase Chemiluminescence (CL) Methods
4 Nitric oxide can be measured reliably using the principle of gas-phase
5 chemiluminescence induced by the reaction of NO with Os at low pressure. Modern commercial
6 NOx analyzers have sufficient sensitivity and specificity for adequate measurement in urban and
7 many rural locations (U.S. Environmental Protection Agency, 1993, 1996, 2006). Research
8 grade CL instruments have been compared under realistic field conditions to spectroscopic
9 instruments, and the results indicate that both methods are reliable (at concentrations relevant to
10 smog studies) to better than 15 percent with 95 percent confidence. Response times are on the
11 order of 1 minute. For measurements meaningful for understanding Os formation, emissions
12 modeling, and N deposition, special care must be taken to zero and calibrate the instrument
13 frequently. A chemical zero, obtained by reacting the NO up-stream and out of view of the
14 photomultiplier tube, is preferred because it accounts for interferences such as light emitting
15 reactions with unsaturated hydrocarbons. Calibration should be performed with NTRM-of
16 compressed NO in N2. Standard additions of NO at the inlet will account for NO loss or
17 conversion to NO2 in the lines. In summary, CL methods, when operated carefully in an
18 appropriate manner, can be suitable for measuring or monitoring NO (e.g., Crosley, 1996).
19
20 Spectroscopic Methods for Nitric Oxide
21 Nitric oxide has also been successfully measured in ambient air with direct spectroscopic
22 methods; these include two-photon laser-induced fluorescence (TPLIF), tunable diode laser
23 absorption spectroscopy (TDLAS), and two-tone frequency-modulated spectroscopy (TTFMS).
24 These were reviewed thoroughly in the previous AQCD and will be only briefly summarized
25 here. The spectroscopic methods demonstrate excellent sensitivity and selectivity for NO with
26 detection limits on the order of 10 ppt for integration times of 1 min. Spectroscopic methods
27 compare well with the CL method for NO in controlled laboratory air, ambient air, and heavily
28 polluted air (e.g., Walega et al., 1984; Gregory et al., 1990; Kireev et al., 1999). These
29 spectroscopic methods remain in the research arena due to their complexity, size, and cost, but
30 are essential for demonstrating that CL methods are reliable for monitoring NO concentrations
31 involved in O3 formation—from around 20 ppt to several hundred of ppb.
March 2008 AX2-89 DRAFT-DO NOT QUOTE OR CITE
-------
1 Atmospheric pressure laser ionization followed by mass spectroscopy has also been
2 deployed for detection of NO and NC>2. Garnica et al. (2000) describe a technique involving
3 selective excitation at one wavelength followed by ionization at a second wavelength. They
4 report good selectivity and detection limits well below 1 ppb. The practicality of the instrument
5 for ambient monitoring, however, has yet to be demonstrated.
6
7 AX2.8.1.3 Measurements of Nitrogen Dioxide
8
9 Gas-Phase Chemiluminescence Methods
10 Reduction of NC>2 to NO, on the surface of a heated (to 300 to 400 °C) molybdenum
11 oxide substrate followed by detection of the chemiluminescence produced during the reaction of
12 NO with Os at low pressure as described earlier for measurement of NO serves as the basis of the
13 FRM for measurement of ambient NO2. However, the substrate used in the reduction of NO2 to
14 NO is not specific to NO2; hence the chemiluminescence analyzers are subject to interference
15 nitrogen oxides other than NO2 produced by oxidized NOY compounds, or NOZ. Thus, this
16 technique will overestimate NO2 concentrations particularly in areas downwind of sources of NO
17 and NO2 as NOx is oxidized to NOz in the form of PANs and other organic nitrates, and HNOs
18 and HNO4. Many of these compounds are reduced at the catalyst with nearly the same efficiency
19 as NO2. Interferences have also been found from a wide range of other compounds as described
20 in the latest AQCD for NO2.
21
22 Other Methods
23 Nitrogen dioxide can be selectively converted to NO by photolysis. For example,
24 (Ryerson et al., 2000) developed a gas-phase chemiluminescence method using a photolytic
25 converter based on a Hg lamp with increased radiant intensity in the region of peak NO2
26 photolysis (350 to 400 nm) and producing conversion efficiencies of 70% or more in less than
27 1 s. Metal halide lamps with conversion efficiency of about 50% and accuracy on the order of
28 20% (Nakamura, et al., 2003) have been used. Because the converter produces little radiation at
29 wavelengths less than 350 nm, interferences from HNOs and PAN are minimal. Alternative
30 methods to photolytic reduction followed by CL are desirable to test the reliability of this widely
31 used technique. Any method based on a conversion to measured species presents potential for
March 2008 AX2-90 DRAFT-DO NOT QUOTE OR CITE
-------
1 interference a problem. Several atmospheric species, PAN and HO2NO2 for example, dissociate
2 to NO2 at higher temperatures.
3 Laser induced fluorescence for NO2 detection involves excitation of atmospheric NO2
4 with laser light emitted at wavelengths too long to induce photolysis. The resulting excited
5 molecules relax in a photoemissive mode and the fluorescing photons are counted. Because
6 collisions would rapidly quench the electronically excited NO2, the reactions are conducted at
7 low pressure. Matsumi et al. (2001) describe a comparison of LIF with a photofragmentation
8 chemiluminescence instrument. The LIF system involves excitation at 440 nm with a multiple
9 laser system. They report sensitivity of 30 ppt in 10 s and good agreement between the two
10 methods under laboratory conditions at mixing ratios up to 1.0 ppb. This high-sensitivity LIF
11 system has yet to undergo long-term field tests. Cleary et al. (2002) describe field tests of a
12 system that uses continuous, supersonic expansion followed by excitation at 640 nm with a
13 commercial cw external-cavity tunable diode laser. More recently, LIF has been successfully
14 used to detect NO2 with accuracy of about 15% and detections limits well below 1 ppb. When
15 coupled with thermal dissociation, the technique also measures peroxy nitrates such as PAN,
16 alkyl nitrates, HNO4 and HNO3 (Cohen, 1999; Day et al., 2002; Farmer et al., 2006; Perez et al.,
17 2007; Thornton et al., 2003). This instrument can have very fast sampling rates be fast (>1 Hz)
18 and shows good correlation with chemiluminescent techniques, but remains a research-grade
19 device.
20 Nitrogen dioxide can be detected by differential optical absorption spectroscopy (DOAS)
21 in an open, long-path system by measuring narrow band absorption features over a background
22 of broad band extinction (e.g., Stutz et al., 2000; Kim and Kim, 2001). A DOAS system
23 manufactured by OPSIS is designated as a Federal Equivalent Method for measuring NO2.
24 DOAS systems can also be configured to measure NO, HONO, and NO3 radicals. Typical
25 detection limits are 0.2 to 0.3 ppbv for NO, 0.05 to 0.1 ppbv for NO2, 0.05 to 0.1 ppbv for
26 HONO, and 0.001 to 0.002 ppbv for NO3, at path lengths of 0.2, 5, 5, and 10 km, respectively.
27 The obvious advantage compared to fixed point measurements is that concentrations relevant to
28 a much larger area are obtained, especially if multiple targets are used. At the same time, any
29 microenvironmental artifacts are minimized over the long path integration. A major limitation in
30 this technique had involved inadequate knowledge of absorption cross sections. Harder et al.
31 (1997) conducted an experiment in rural Colorado involving simultaneous measurements of NO2
March 2008 AX2-91 DRAFT-DO NOT QUOTE OR CITE
-------
1 by DOAS and by photolysis followed by chemiluminescence. They found differences of as
2 much as 110% in clean air from the west, but for NO2 mixing ratios in excess of 300 ppt, the two
3 methods agreed to better than 10%. Stutz et al. (2000) cites two intercomparisons of note. Nitric
4 oxide was measured by DOAS, by photolysis of NO2 followed by chemiluminescence, and by
5 LIF during July 1999 as part of the SOS in Nashville, TN. On average, the three methods agreed
6 to within 2%, with some larger differences likely caused by spatial variability over the DOAS
7 path. In another study in Europe, and a multi-reflection set-up over a 15 km path, negated the
8 problem of spatial averaging here agreement with the chemiluminescence detector following
9 photolytic conversion was excellent (slope = 1.006 ± 0.005; intercept = 0.036 ± 0.019; r = 0.99)
10 over a concentration range from about 0.2 to 20 ppbv.
11 Nitric oxide can also be detected from space with DOAS-like UV spectroscopy
12 techniques (Kim et al., 2006; Ma et al., 2006). These measurements appear to track well with
13 emissions estimates and can be a useful indicator of column content as well as for identifying hot
14 spots in sources. See also Richter et al., 2005. Leigh (2006) report on a DOAS method that uses
15 the sun as a light source and compares well with an in situ chemiluminescence detector in an
16 urban environment.
17 Chemiluminescence on the surface of liquid Luminol has also been used for measurement
18 of NO2 (Gaffney et al., 1998; Kelly et al., 1990; Marley et al., 2004; Nikitas et al., 1997; Wendel
19 etal., 1983). This technique is sensitive and linear, and more specific than hot MoOx. Luminol
20 does not emit light when exposed to NHOs or alkyl nitrates, but does react with PAN. This
21 interference can be removed by chromatographic separation prior to detection and the resulting
22 measurement compares well with more specific techniques for moderate to high (> 1 ppb) mixing
23 ratios of NO2.
24 Several tunable diode laser spectroscopy techniques have been used successfully for NO2
25 detection (Eisele et al., 2003; Osthoff et al., 2006). These devices remain research grade
26 instruments, not yet practical for urban monitoring.
27
28 Measurements of Total Oxidized Nitrogen Species, NOy
29 Gold catalyzed CO, or H2 reduction or as conversion on hot molybdenum oxide catalyst
30 have been used to reduce NOY to NO before then detection by chemiluminescence (Fehsenfeld
31 et al., 1987; Crosley, 1996). Both techniques offer generally reliable measurements, with
March 2008 AX2-92 DRAFT-DO NOT QUOTE OR CITE
-------
1 response times on the order of 60 s and a linear dynamic range demonstrated in field
2 intercomparisons from about 10 ppt to 10s of ppb. Under certain conditions, HCN, NH3, RNO2,
3 and CH3CN can be converted to NO, but at normal concentrations and humidity these are minor
4 interferences. Thermal decomposition followed by LIF has also been used for NOy detection, as
5 described above. In field comparisons, instruments based on these two principles generally
6 showed good agreement (Day et al., 2002). The experimental uncertainty is estimated to be of
7 15-30%.
8
9 AX2.8.1.4 Monitoring for NOi Compliance Versus Monitoring for Ozone Formation
10 Regulatory measurements of NC>2 have been focused on demonstrating compliance with
11 the NAAQS for NC>2. Today, few locations violate that standard, but NO2 and related NOy
12 compounds remain among the most important atmospheric trace gases to measure and
13 understand. Commercial instruments for NO/NOX detection are generally constructed with an
14 internal converter for reduction of NC>2 to NO, and generate a signal referred to as NOx. These
15 converters, generally constructed of molybdenum oxides (MoOx), reduce not only NO2 but also
16 most other NOy species. Unfortunately, with an internal converter, the instruments may not give
17 a faithful indication of NOy either—reactive species such as HNOs will adhere to the walls of
18 the inlet system. Most recently, commercial vendors such as Thermo Environmental (Franklin,
19 MA) have offered NO/NOy detectors with external Mo converters. If such instruments are
20 calibrated through the inlet with a reactive nitrogen species such as propyl nitrate, they give
21 accurate measurements of total NOy, suitable for evaluation of photochemical models. (Crosley,
22 1996; Fehsenfeld et al., 1987; Nunnermacker et al., 1998; Rodgers and Davis, 1989). Under
23 conditions of fresh emissions, such as in urban areas during the rush hour, NOy ~ NOx and these
24 monitors can be used for testing emissions inventories (Dickerson, et al., 1995; Parrish, 2006).
25 The state of Maryland for example is making these true NOy measurements at the Piney Run site
26 in the western part of the state. These data produced at this site can be more reliably compared
27 to the output of CMAQ and other chemical transport models.
28
29 Summary of Methods for Measuring NO 2
30 A variety of techniques exist for reliable monitoring of atmospheric NO2 and related
31 reactive nitrogen species. For demonstration of compliance with the NAAQS for NO2,
32 commercial chemiluminescence instruments are adequate. For certain conditions, luminol
March 2008 AX2-93 DRAFT-DO NOT QUOTE OR CITE
-------
1 chemiluminescence is adequate. Precise measurements of NO2 can be made with research grade
2 instruments such as LIF and TDLS. For path-integrated concentration determinations UV
3 spectroscopic methods provide useful information. Commercial NOX instruments are sensitive
4 to other NOy species, but do not measure NOy quantitatatively. NOy instruments with external
5 converters offer measurements more useful for comparison to chemical transport model
6 calculations.
7
8 AX2.8.2 Measurements of HNO3
9 Accurate measurement of HNOs, has presented a long-standing analytical challenge to
10 the atmospheric chemistry community. In this context, it is useful to consider the major factors
11 that control HNOs partitioning between the gas and deliquesced-particulate phases in ambient
12 air. In equation form,
Kll Ka
13 •*£ L -'tw3rj L J L -' (AX2.8-1)
14 where KH is the Henry's Law constant in M ataf * and Ka is the acid dissociation constant in M.
15 Thus, the primary controls on HNOs phase partitioning are its thermodynamic properties
16 (KH, Ka, and associated temperature corrections), aerosol liquid water content (LWC), solution
17 pH, and kinetics. Aerosol LWC and pH are controlled by the relative mix of different acids and
18 bases in the system, hygroscopic properties of condensed compounds, and meteorological
19 conditions (RH, temperature, and pressure). It is evident from relationship AX2.8-1 that, in the
20 presence of chemically distinct aerosols of varying acidities (e.g., super-jam predominantly sea
21 salt and sub-jam predominantly S aerosol), HNOs will partition preferentially with the less-acidic
22 particles; and this is consistent with observations (e.g., Huebert et al., 1996; Keene and Savoie,
23 1998; Keene et al., 2002). Kinetics are controlled by atmospheric concentrations of HNOs vapor
24 and particulate MV and the size distribution and corresponding atmospheric lifetimes of
25 particles against deposition. Sub-jam diameter aerosols typically equilibrate with the gas phase
26 in seconds to minutes while super-um aerosols require hours to a day or more (e.g., Meng and
27 Seinfeld, 1996; Erickson et al., 1999). Consequently, smaller aerosol size fractions are typically
28 close to thermodynamic equilibrium with respect to HNOs whereas larger size fractions (for
29 which atmospheric lifetimes against deposition range from hours to a few days) are often
30 undersaturated (e.g., Erickson et al., 1999; Keene and Savioe, 1998).
March 2008 AX2-94 DRAFT-DO NOT QUOTE OR CITE
-------
1 Many sampling techniques for HNOs (e.g., annular denuder, standard filterpack and mist-
2 chamber samplers) employ upstream prefilters to remove particulate species from sample air.
3 However, when chemically distinct aerosols with different pHs (e.g., sea salt and S aerosols) mix
4 together on a bulk filter, the acidity of the bulk mixture will be greater than that of the less acidic
5 aerosols with which most NOs is associated. This change in pH may cause the bulk mix to be
6 supersaturated with respect to HNOs leading to volatilization and, thus, positive measurement
7 bias in HNOs sampled downstream. Alternatively, when undersaturated super-Jim size fractions
8 (e.g., sea salt) accumulate on a bulk filter and chemically interact over time with HNOs in the
9 sample air stream, scavenging may lead to negative bias in HNOs sampled downstream.
10 Because the magnitude of both effects will vary as functions of the overall composition and
11 thermodynamic state of the multiphase system, the combined influence can cause net positive or
12 net negative measurement bias in resulting data. Pressure drops across particle filters can also
13 lead to artifact volatilization and associated positive bias in HNOs measured downstream.
14 Widely used methods for measuring HNOs include standard filterpacks configured with
15 nylon or alkaline-impregnated filters (e.g., Goldan et al., 1983; Bardwell et al., 1990), annular
16 denuders (EPA Method IP-9), and standard mist chambers (Talbot et al., 1990). Samples from
17 these instruments are typically analyzed by ion chromatography. Intercomparisons of these
18 measurement techniques (e.g., Hering et al., 1988; Tanner et al., 1989; Talbot et al., 1990) report
19 differences on the order of a factor of two or more.
20 More recently, sensitive HNOs measurements based on the principle of Chemical
21 lonization Mass Spectroscopy (CIMS) have been reported (e.g., Huey et al., 1998; Mauldin
22 et al., 1998; Furutani and Akimoto, 2002; Neuman et al., 2002). CIMS relies on selective
23 formation of ions such as SiFs -HNOs or HSC>4 -HNOs followed by detection via mass
24 spectroscopy. Two CIMS techniques and a filter pack technique were intercompared in Boulder,
25 CO (Fehsenfeld et al., 1998). Results indicated agreement to within 15% between the two CIMS
26 instruments and between the CIMS and filterpack methods under relatively clean conditions with
27 HNO3 mixing ratios between 50 and 400 pptv. In more polluted air, the filterpack technique
28 generally yielded higher values than the CIMS suggesting that interactions between chemically
29 distinct particles on bulk filters is a more important source of bias in polluted continental air.
30 Differences were also greater at lower temperature when particulate NOs corresponded to
31 relatively greater fractions of total N(V.
March 2008 AX2-95 DRAFT-DO NOT QUOTE OR CITE
-------
1 AX2.8.3 Techniques for Measuring Other NOY Species
2 Methods for sampling and analysis of alkyl nitrates in the atmosphere have been
3 reviewed by Parrish and Fehsenfeld (2000). Peroxyacetylnitrate, PPN, and MPAN are typically
4 measured using a chromatograph followed by electron capture detectors or GC/ECD (e.g.,
5 Gaffney et al., 1998), although other techniques such as FTIR could also be used. Field
6 measurements are made using GC/ECD with a total uncertainty of ± 5 pptv + 15% (Roberts
7 etal., 1998).
8 In the IMPROVE network and in the EPA's speciation network, particulate nitrate in the
9 PM2.5 size range is typically collected on nylon filters downstream of annular denuders coated
10 with a basic solution capable of removing acidic gases such as HNOs, HNO2, and 862. Filter
11 extracts are then analyzed by ion chromatography (1C) for nitrate, sulfate, and chloride. Nitrite
12 ions are also measured by this technique but their concentrations are almost always beneath
13 detection limits. However, both of these networks measure nitrate only in the PM2.5 fraction.
14 Because of interactions with more highly acidic components on filter surfaces, there could be
15 volatilization of nitrate in PMio samples. These effects are minimized if separate aerosol size
16 fractions are collected, i.e., the more acidic PM2.s and the more alkaline PMi0-2.5 as in a
17 dichotomous sampler or multistage impactor.
18
19 AX2.8.4 Remote Sensing of Tropospheric NOi Columns for Surface NOx
20 Emissions and Surface NOi Concentrations
21 Table AX2.8-1 contains an overview of the three satellite instruments that are used
22 retrieve tropospheric NC>2 columns from measurements of solar backscatter. All three
23 instruments are in polar sun-synchronous orbits with global measurements in the late morning
24 and early afternoon. The spatial resolution of the measurement from SCIAMACHY is 7 times
25 better than that from Ozone Monitoring Instrument (GOME), and that from Ozone Monitoring
26 Instrument (OMI) is 40 times better than that from GOME.
27 Figure AX2.8-1 shows tropospheric NO2 columns retrieved from SCIAMACHY.
28 Pronounced enhancements are evident over major urban and industrial emissions. The high
29 degree of spatial heterogeneity over the southwestern United States provides empirical evidence
30 that most of the tropospheric NO2 column is concentrated in the lower troposphere.
31 Tropospheric NO2 columns are more sensitive to NOx in the lower troposphere than in the upper
32 troposphere (Martin et al., 2002). This sensitivity to NOx in the lower troposphere is due to the
March 2008 AX2-96 DRAFT-DO NOT QUOTE OR CITE
-------
Figure AX2.8-1.
u
SCIAMACHYTropospheric N0 (10 15 moleccm 2)
Tropospheric NOi columns (molecules NOi/ cm ) retrieved from the
SCIAMACHY satellite instrument for 2004-2005.
Source: Martin et al. (2006).
1 factor of 25 decrease in the NO2/NO ratio from the surface to the upper troposphere (Bradshaw
2 et al., 1999) that is driven by the temperature dependence of the NO + Os reaction. Martin et al.
3 (2004a) integrated in situ airborne measurements of NO2 and found that during summer the
4 lower mixed layer contains 75% of the tropospheric NO2 column over Houston and Nashville.
5 However, it should be noted that these measurements are also sensitive to surface albedo and
6 aerosol loading.
7 The retrieval involves three steps: (1) determining total NO2 line-of-sight (slant) columns
8 by spectral fitting of solar backscatter measurements, (2) removing the stratospheric columns by
9 using data from remote regions where the tropospheric contribution to the column is small, and
10 (3) applying an air mass factor (AMF) for the scattering atmosphere to convert tropospheric slant
11 columns into vertical columns. The retrieval uncertainty is determined by (1) and (2) over
March 2008
AX2-97
DRAFT-DO NOT QUOTE OR CITE
-------
1 remote regions where there is little tropospheric NO2, and by (3) over regions in regions of
2 elevated tropospheric NO2 (Martin et al., 2002; Boersma et al., 2004).
3 The paucity of in situ NO2 measurements motivates the inference of surface NO2
4 concentrations from satellite measurements of tropospheric NO2 columns. This prospect would
5 take advantage of the greater sensitivity of tropospheric NO2 columns to NOx in the lower
6 troposphere than in the upper troposphere as discussed earlier. Tropospheric NO2 columns show
7 a strong correlation with in situ NO2 measurements in northern Italy (Ordonez et al., 2006).
8 Quantitative calculation of surface NO2 concentrations from a tropospheric NO2 column
9 would require information on the relative vertical profile. Comparison of vertical profiles of
10 NO2 in a chemical transport model (GEOS-Chem) versus in situ measurements over and
11 downwind of North America shows a high degree of consistency (Martin et al., 2004b, 2006),
12 suggesting that chemical transport models could be used to infer the relationship between surface
13 NO2 concentrations and satellite observations of the tropospheric NO2 column.
14 However, the satellites carrying the spectrometer (GOME/SCIAMACHY/OMI) are in
15 near polar, sun-synchronous orbits. As a result, these measurements are made only once per day,
16 typically between about 10:00 to 11:00 a.m. or 1 p.m. local time, during a brief overflight. Thus
17 the utility of these measurements is limited as they would likely miss short-term features.
18
19 AX2.8.5 SAMPLING AND ANALYSIS FOR SO2
20 Currently, ambient SO2 is measured using instruments based on pulsed fluorescence. The
21 UV fluorescence monitoring method for atmospheric SO2 was developed to improve upon the
22 flame photometric detection (FPD) method for SO2, which in turn had displaced the
23 pararosaniline wet chemical method for SO2 measurement. The pararosaniline method is still the
24 FRM for atmospheric SO2, but is rarely used because of its complexity and slow response, even
25 in its automated forms. Both the UV fluorescence and FPD methods are designated as FEMs by
26 the EPA, but UV fluorescence has largely supplanted the FPD approach because of the UV
27 method's inherent linearity, sensitivity, and the absence of consumables, such as the hydrogen
28 gas needed for the FPD method.
29 Basically, SO2 molecules absorb ultraviolet (UV) light at one wavelength and emit UV
30 light at longer wavelengths. This process is known as fluorescence, and involves the excitation
31 of the SO2 molecule to a higher energy (singlet) electronic state. Once excited, the molecule
March 2008 AX2-98 DRAFT-DO NOT QUOTE OR CITE
-------
1 decays non-radiatively to a lower energy electronic state from which it then decays to the
2 original, or ground, electronic state by emitting a photon of light at a longer wavelength (i.e.,
3 lower energy) than the original, incident photon. The process can be summarized by the
4 following equations
(2.8-2)
, S02*^>S02+hv2 ,._.,
6 (2.Q-3)
7 where SO 2* represents the excited state of SO2, h v/, and h V2 represent the energy of the
8 excitation and fluorescence photons, respectively, and hv2
-------
1 positive interference. To remove this source of interference, the high sensitivity 862 analyzers,
2 such as those to be used in the NCore network (U.S. Environmental Protection Agency, 2005),
3 have hydrocarbon scrubbers to remove these compounds from the sample stream before the
4 sample air enters the optical chamber.
5 Another potential source of positive interference is nitric oxide (NO). NO fluoresces in a
6 spectral region that is close to the SO2 fluorescence. However, in high sensitivity SO2 analyzers,
7 the bandpass filter in front of the PMT is designed to prevent NO fluorescence from reaching the
8 PMT and being detected. Care must be exercised when using multicomponent calibration gases
9 containing both NO and SO2 that the NO rejection ratio of the SO2 analyzer is sufficient to
10 prevent NO interference. The most common source of positive bias (as constrasted with positive
11 spectral interference) in high-sensitivity SO2 monitoring is stray light reaching the optical
12 chamber. Since SO2 can be electronically excited by a broad range of UV wavelengths, any
13 stray light with an appropriate wavelength that enters the optical chamber can excite SO2 in the
14 sample and increase the fluorescence signal.
15 Furthermore, stray light at the wavelength of the SO2 fluorescence that enters the optical
16 chamber may impinge on the PMT and increase the fluorescence signal. Several design features
17 are incorporated to minimize the stray light that enters the chamber. These features include the
18 use of light filters, dark surfaces, and opaque tubing to prevent light from entering the chamber.
19 Luke (1997) reported the positive artifacts of a modified pulsed fluorescence detector
20 generated by the co-existence of NO, CS2, and a number of highly fluorescent aromatic
21 hydrocarbons such as benzene, toluene, o-xylene, m-xylene, p-xylene, m-ethyltoluene,
22 ethylbenzene, and 1,2,4-trimethylbenzene. The positive artifacts could be reduced by using a
23 hydrocarbon "kicker" membrane. At a flow rate of 300 standard cc min"1 and a pressure drop of
24 645 torr across the kicker, the interference from ppm levels of many aromatic hydrocarbons was
25 eliminated entirely.
26 Nicks and Benner (2001) described a sensitive SO2 chemiluminescence detector, which
27 was based on a differential measurement where response from ambient SO2 is determined by the
28 difference between air containing SO2 and air scrubbed of SO2 where both air samples contain
29 other detectable sulfur species, and the positive artifact could also be reduced through this way.
March 2008 AX2-100 DRAFT-DO NOT QUOTE OR CITE
-------
1 Sources of Negative Interference
2 Nonradiative deactivation (quenching) of excited SO2 molecules can occur from
3 collisions with common molecules in air, including nitrogen, oxygen, and water. During
4 collisional quenching, the excited SC>2 molecule transfers energy, kinetically allowing the SC>2
5 molecule to return to the original lower energy state without emitting a photon. Collisional
6 quenching results in a decrease in the SC>2 fluorescence and results in the underestimation of SC>2
7 concentration in the air sample. The concentrations of nitrogen and oxygen are constant in the
8 ambient air, so quenching from those species at a surface site is also constant, but the water
9 vapor content of air can vary. Luke (1997) reported that the response of the detector could be
10 reduced by about 7% and 15% at water vapor mixing ratios of 1 and 1.5 mole percent (RH = 35
11 to 50% at 20-25 °C and 1 atm for a modified pulsed fluorescence detector (Thermo
12 Environmental Instruments, Model 43s). Condensation of water vapor in sampling lines must be
13 avoided, as it can absorb SC>2 from the sample air. The simplest approach to avoid condensation
14 is to heat sampling lines to a temperature above the expected dew point, and within a few
15 degrees of the controlled optical bench temperature. At very high 862 concentrations, reactions
16 between electronically excited 862 and ground state 862 to form 863 and SO might occur
17 (Calvert et al., 1978). However, this possibility has not been examined.
18
19 Other Techniques for Measuring SO 2
20 A more sensitive SO2 measurement method than the UV-fluorescence method was
21 reported by Thornton et al. (2002). Thornton et al (2002) reported an atmospheric pressure
22 ionization mass spectrometer. The high measurement precision and instrument sensitivity were
23 achieved by adding isotopically labeled SC>2 (34S16C>2) continuously to the manifold as an internal
24 standard. Field studies showed that the method precision was better than 10% and the limit of
25 detection was less than 1 pptv for a sampling interval of Is.
26 Sulfur Dioxide can be measured by LIF at around 220 nm (Matsumi et al. (2005).
27 Because the laser wavelength is alternately tuned to an 862 absorption peak at 220.6 and bottom
28 at 220.2 nm, and the difference signal at the two wavelengths is used to extract the 862
29 concentration, the technique eliminates interference from either absorption or fluorescence by
30 other species and has high sensitivity (5 pptv in 60 sec). Sulfur Dioxide can also be measured by
31 the same DOAS instrument that can measure NC>2.
March 2008 AX2-101 DRAFT-DO NOT QUOTE OR CITE
-------
1 Photoacoutsic techniques have been employed for SO2 detection, but they generally have
2 detection limits suitable only for source monitoring (Gondal, 1997; Gondal and Mastromarino,
3 2001).
4 Chemical lonization Mass Spectroscopy (CIMS) utilizes ionization via chemical
5 reactions in the gas phase to determine an unknown sample's mass spectrum and identity. High
6 sensitivity (10 ppt or better) has been achieved with uncertainty of-15% when a charcoal
7 scrubber is used for zeroing and the sensitivity is measured with isotopically labeled 34SO2
8 (Hanke et al., 2003; Huey et al., 2004; Hennigan et al., 2006).
9
10 AX2.8.6 Sampling and Analysis for Sulfate, Nitrate, and Ammonium
11
12 Sampling Artifacts
13 Sulfate, nitrate, and ammonium are commonly present in PM2 5. Most PM2 5 samplers
14 have a size-separation device to separate particles so that only those particles approximately
15 2.5 |im or less are collected on the sample filter. Air is drawn through the sample filter at a
16 controlled flow rate by a pump located downstream of the sample filter. The systems have two
17 critical flow rate components for the capture of fine paniculate: (1) the flow of air through the
18 sampler must be at a flow rate that ensures that the size cut at 2.5 jim occurs; and (2) the flow
19 rate must be optimized to capture the desired amount of particulate loading with respect to the
20 analytical method detection limits.
21 When using the system described above to collect sulfate, nitrate and parti culate
22 ammonium, sampling artifacts can occur because of: (1) positive sampling artifact for sulfate,
23 nitrate, and particulate ammonium due to chemical reaction; and (2) negative sampling artifact
24 for nitrate and ammonium due to the decomposition and evaporation.
25
26 Sampling and Analysis Techniques
27
28 Denuder-Filter Based Sampling and Analysis Techniques for Sulfate, Nitrate, and Ammonium
29 There are two major PM speciation ambient air-monitoring networks in the United States:
30 the Speciation Trend Network (STN), and the Interagency Monitoring of Protected Visual
31 Environments (IMPROVE) network. The current STN samplers include three filters: (1) Teflon
32 for equilibrated mass and elemental analysis including elemental sulfur; (2) a HNOs denuded
33 nylon filter for ion analysis including NOs and SO/t, (3) a quartz-fiber filter for elemental and
March 2008 AX2-102 DRAFT-DO NOT QUOTE OR CITE
-------
1 organic carbon. The IMPROVE sampler, which collects two 24-h samples per week,
2 simultaneously collects one sample of PMio on a Teflon filter, and three samples of PM2.5 on
3 Teflon, nylon, and quartz filters. PM2 5 mass concentrations are determined gravimetrically from
4 the PM2.5 Teflon filter sample. The PM2.5 Teflon filter sample is also used to determine
5 concentrations of selected elements. The PM2.5 nylon filter sample, which is preceded by a
6 denuder to remove acidic gases, is analyzed to determine nitrate and sulfate aerosol
7 concentrations. Finally, the PM2 5 quartz filter sample is analyzed for OC and EC using the
8 thermal-optical reflectance (TOR) method. The STN and the IMPROVE networks represent a
9 major advance in the measurement of nitrate, because the combination of a denuder (coated with
10 either Na2COs or MgO) to remove HNOs vapor and a Nylon filter to adsorb HNOs vapor
11 volatilizing from the collected ammonium nitrate particles overcomes the loss of nitrate from
12 Teflon filters.
13 The extent to which sampling artifacts for particulate NH3+ have been adequately
14 addressed in the current networks is not clear. Recently, new denuder-filter sampling systems
15 have been developed to measure sulfate, nitrate, and ammonium with an adequate correction of
16 ammonium sampling artifacts. The denuder-filter system, Chembcomb Model 3500 speciation
17 sampling cartridge developed by Rupprecht & Patashnick Co, Inc. could be used to collect
18 nitrate, sulfate, and ammonium simultaneously. The sampling system contains a single-nozzle
19 size-selective inlet, two honeycomb denuders, the aerosol filter and two backup filters (Keck and
20 Wittmaack, 2005). The first denuder in the system is coated with 0.5% sodium carbonate and
21 1% glycerol and collects acid gases such as HCL, SO2, HONO, and HNOs. The second denuder
22 is coated with 0.5% phosphoric acid in methanol for collecting NHa. Backup filters collect the
23 gases behind denuded filters. The backup filters are coated with the same solutions as the
24 denuders. A similar system based on the same principle was applied by Possanzini et al. (1999).
25 The system contains two NaCl-coated annular denuders followed by other two denuders coated
26 with NaCOs/glycerol and citric acid, respectively. This configuration was adopted to remove
27 HNOs quantitatively on the first NaCl denuder. The third and forth denuder remove SO2 and
28 NH3, respectively. A polyethylene cyclone and a two-stage filter holder containing three filters
29 is placed downstream of the denuders. Aerosol fine particles are collected on a Teflon
30 membrane. A backup nylon filter and a subsequent citric acid impregnated filter paper collect
March 2008 AX2-103 DRAFT-DO NOT QUOTE OR CITE
-------
1 dissociation products (HNOs and NH3) of ammonium nitrate evaporated from the filtered
2 particulate matter.
3 Several traditional and new methods could be used to quantify elemental S collected on
4 filters: energy dispersive X-ray fluorescence, synchrotron induced X-ray fluorescence, proton
5 induced X-ray emission (PIXE), total reflection X-ray fluorescence, and scanning electron
6 microscopy. Energy dispersive X-ray fluorescence (EDXRF) (Method IO-3.3, U.S.
7 Environmental Protection Agency, 1997; see 2004 PM CD for details) and PIXE are the most
8 commonly used methods. Since sample filters often contain very small amounts of particle
9 deposits, preference is given to methods that can accommodate small sample sizes and require
10 little or no sample preparation or operator time after the samples are placed into the analyzer. X-
11 ray fluorescence (XRF) meets these needs and leaves the sample intact after analysis so it can be
12 submitted for additional examinations by other methods as needed. To obtain the greatest
13 efficiency and sensitivity, XRF typically places the filters in a vacuum which may cause volatile
14 compounds (nitrates and organics) to evaporate. As a result, species that can volatilize such as
15 ammonium nitrate and certain organic compounds can be lost during the analysis. The effects of
16 this volatilization are important if the PTFE filter is to be subjected to subsequent analyses of
17 volatile species.
18 Polyatomic ions such as sulfate, nitrate, and ammonium are quantified by methods such
19 as ion chromatography (1C) (an alternative method commonly used for ammonium analysis is
20 automated colorimetry). All ion analysis methods require a fraction of the filter to be extracted
21 in deionized distilled water for sulfate and NaCOs/NaHCOs solution for nitrate and then filtered
22 to remove insoluble residues prior to analysis. The extraction volume should be as small as
23 possible to avoid over-diluting the solution and inhibiting the detection of the desired
24 constituents at levels typical of those found in ambient PM2 5 samples. During analysis, the
25 sample extract passes through an ion-exchange column which separates the ions in time for
26 individual quantification, usually by an electroconductivity detector. The ions are identified by
27 their elution/retention times and are quantified by the conductivity peak area or peak height.
28 In a side-by-side comparison of two of the major aerosol monitoring techniques (Hains
29 et al., 2007), PM2.5 mass and major contributing species were well correlated among the different
30 methods with r-values in excess of 0.8. Agreement for mass, sulfate, OC, TC, and ammonium
31 was good while that for nitrate and BC was weaker. Based on reported uncertainties, however,
March 2008 AX2-104 DRAFT-DO NOT QUOTE OR CITE
-------
1 even daily concentrations of PM2.5 mass and major contributing species were often significantly
2 different at the 95% confidence level. Greater values of PM2.5 mass and individual species were
3 generally reported from Speciation Trends Network methods than from the Desert Research
4 Institute Sequential Filter Samplers. These differences can only be partially accounted for by
5 known random errors. The authors concluded that the current uncertainty estimates used in the
6 STN network may underestimate the actual uncertainty.
7
8 Positive Sampling Artifacts
9 The reaction of SO2 (and other acid gases) with basic sites on glass fiber filters or with
10 basic coarse particles on the filter leads to the formation of sulfate (or other nonvolatile salts,
11 e.g., nitrate, chloride). These positive artifacts lead to the overestimation of total mass, and
12 sulfate, and probably also nitrate concentrations. These problems were largely overcome by
13 changing to quartz fiber or Teflon filters and by the separate collection of PM2.5. However, the
14 possible reaction of acidic gases with basic coarse particles remains a possibility, especially with
15 PMio and PMio-2.s measurements. These positive artifacts could be effectively eliminated by
16 removing acidic gases in the sampling line with denuders coated with NaCl or Na2CO3.
17 Positive sampling artifacts also occur during measurement of particulate NH4. The
18 reaction of NH3 with acidic particles (e.g. 2NH3 + H2SO4 ^ (NH4)2SO4), either during sampling
19 or during transportation, storage, and equilibration could lead to an overestimation of parti culate
20 NH4 concentrations. Techniques have been developed to overcome this problem: using a
21 denuder to remove NH3 during sampling and to protect the collected PM from NH3 (Suh et al.,
22 1992, 1994; Brauer et al., 1991; Koutrakis et al., 1988a,b; Keck and Wittmaack, 2006;
23 Possanzini et al., 1999; Winberry et al., 1999). Hydrogen fluoride, citric acid, and phosphorous
24 acids have been used as coating materials for the NH3 denuder. Positive artifacts for particulate
25 NH4 can also be observed during sample handling due to contamination. No chemical analysis
26 method, no matter how accurate or precise, can adequately represent atmospheric concentrations
27 if the filters to which these methods are applied are improperly handled. Ammonia is emitted
28 directly from human sweat, breath and smoking. It can then react with acidic aerosols on the
29 filter to form ammonium sulfate, ammonium bisulfate and ammonium nitrate if the filter was not
30 properly handled (Sutton el al., 2000). Therefore, it is important to keep filters away from
31 ammonia sources, such as human breath, to minimize neutralization of the acidic compounds.
March 2008 AX2-105 DRAFT-DO NOT QUOTE OR CITE
-------
1 Also, when filters are handled, preferably in a glove box, the analyst should wear gloves that are
2 antistatic and powder-free to act as an effective contamination barrier.
3
4 Negative Sampling Artifact
5 Although sulfate is relatively stable on a Teflon filter, it is now well known that
6 volatilization losses of particulate nitrates occur during sampling.
7 For nitrate, the effect on the accuracy of atmospheric parti culate measurements from
8 these volatilization losses is more significant for PM2.5 than for PMio. The FRM for PM2.5 will
9 likely suffer a loss of nitrates similar to that experienced with other simple filter collection
10 systems. Sampling artifacts resulting from the loss of parti culate nitrates represents a significant
11 problem in areas such as southern California that experience high loadings of nitrates. Hering
12 and Cass (1999) discussed errors in PM2.5 mass measurements due to the volatilization of
13 parti culate nitrate. They examined data from two field measurement campaigns that were
14 conducted in southern California: (1) the Southern California Air Quality Study (SCAQS)
15 (Lawson, 1990) and (2) the 1986 CalTech study (Solomon et al., 1992). In both these studies,
16 side-by-side sampling of PM2.5 was conducted. One sampler collected particles directly onto a
17 Teflon filter. The second sampler consisted of a denuder to remove gaseous HMOs followed by
18 a nylon filter that absorbed the HNOs as it evaporated from NITXNOs. In both studies, the
19 denuder consisted of MgO-coated glass tubes (Appel et al., 1981). Fine particulate nitrate
20 collected on the Teflon filter was compared to fine particulate nitrate collected on the denuded
21 nylon filter. In both studies, the PM2.5 mass lost because of ammonium nitrate volatilization
22 represented a significant fraction of the total PM2.5 mass. The fraction of mass lost was higher
23 during summer than during fall (17% versus 9% during the SCAQS study, and 21% versus 13%
24 during the CalTech study). In regard to percentage loss of nitrate, as opposed to percentage loss
25 of mass discussed above, Hering and Cass (1999) found that the amount of nitrate remaining on
26 the Teflon filter samples was on average 28% lower than that on the denuded nylon filters.
27 Hering and Cass (1999) also analyzed these data by extending the evaporative model
28 developed by Zhang and McMurry (1987). The extended model used by Hering and Cass (1999)
29 takes into account the dissociation of collected particulate ammonium nitrate on Teflon filters
30 into HNOs and NH3 via three mechanisms: (1) the scrubbing of HNOs and NH3 in the sampler
31 inlet (John et al. (1988) showed that clean PMio inlet surfaces serve as an effective denuder for
32 HNOs); (2) the heating of the filter substrate above ambient temperature by sampling; and (3) the
March 2008 AX2-106 DRAFT-DO NOT QUOTE OR CITE
-------
1 pressure drop across the Teflon filter. For the sampling systems modeled, the flow-induced
2 pressure drop was measured to be less than 0.02 atm, and the corresponding change in vapor
3 pressure was 2%, so losses driven by pressure drop were not considered to be significant in this
4 work. Losses from Teflon filters were found to be higher during the summer than during the
5 winter, higher during the day compared to night, and reasonably consistent with modeled
6 predictions.
7 Finally, during the SCAQS (Lawson, 1990) study, particulate samples also were collected
8 using a Berner impactor and greased Tedlar substrates in size ranges from 0.05 to 10 jim in
9 aerodynamic diameter. The Berner impactor PM2.5 nitrate values were much closer to those
10 from the denuded nylon filter than those from the Teflon filter, the impactor nitrate values being
11 -2% lower than the nylon filter nitrate for the fall measurements and -7% lower for the summer
12 measurements. When the impactor collection was compared to the Teflon filter collection for a
13 nonvolatile species (sulfate), the results were in agreement. Chang et al. (2000) discuss reasons
14 for reduced loss of nitrate from impactors.
15 Brook and Dann (1999) observed much higher nitrate losses during a study in which they
16 measured parti culate nitrate in Windsor and Hamilton, Ontario, Canada, by three techniques:
17 (1) a single Teflon filter in a dichotomous sampler, (2) the Teflon filter in an annular denuder
18 system (ADS), and (3) total nitrate including both the Teflon filter and the nylon back-up filter
19 from the ADS. The Teflon filter from the dichotomous sampler averaged only 13% of the total
20 nitrate, whereas the Teflon filter from the ADS averaged 46% of the total nitrate. The authors
21 concluded that considerable nitrate was lost from the dichotomous sampler filters during
22 handling, which included weighing and X-ray fluorescence (XRF) measurement in a vacuum.
23 Kim et al. (1999) also examined nitrate sampling artifacts by comparing denuded and
24 non-denuded quartz and nylon filters during the PMi0 Technical Enhancement Program (PTEP)
25 in the South Coast Air Basin of California. They observed negative nitrate artifacts (losses) for
26 most measurements; however, for a significant number of measurements, they observed positive
27 nitrate artifacts. Kim et al. (1999) pointed out that random measurement errors make it difficult
28 to measure true amounts of nitrate loss.
29 Diffusion denuder samplers, developed primarily to measure particle strong acidity
30 (Koutrakis et al., 1988b, 1992), also can be used to study nitrate volatilization. Such techniques
31 were used to measure loss of parti culate nitrate from Teflon filters in seven U.S. cities (Babich
March 2008 AX2-107 DRAFT-DO NOT QUOTE OR CITE
-------
1 et al., 2000). Measurements were made with two versions of the Harvard-EPA Annular Denuder
2 System (HEADS). HNOs vapor was removed by a Na2CC>3-coated denuder. Paniculate nitrate
3 was the sum of nonvolatile nitrate collected on a Teflon filter and volatized nitrate collected on a
4 Na2CO3-coated filter downstream of the Teflon filter (full HEADS) or on a Nylon filter
5 downstream of the Teflon filter (Nylon HEADS). It was found that the full HEADS (using a
6 Na2COs filter) consistently underestimated the total particulate nitrate by approximately 20%
7 compared to the nylon HEADS. Babich et al. (2000) found significant nitrate losses in
8 Riverside, CA; Philadelphia, PA; and Boston, MA, but not in Bakersfield, CA; Chicago, IL;
9 Dallas, TX; or Phoenix, AZ, where measurements were made only during the winter. Tsai and
10 Huang (1995) used a diffusion denuder to study the positive and negative artifacts on glass and
11 quartz filters. They found positive artifacts attributed to SO2 and HNO3 reaction with basic sites
12 on glass fibers and basic particles and negative artifacts attributed to loss of HNOs and HC1 due
13 to volatilization of NH4NO3 and NH4C1 and reaction of these species with acid sulfates.
14 Volatile compounds can also leave the filter after sampling and prior to filter weighing or
15 chemical analysis. Losses of NOs, NH4, and Cl from glass and quartz-fiber filters that were
16 stored in unsealed containers at ambient air temperatures for 2 to 4 weeks prior to analysis
17 exceeded 50 percent (Witz et al., 1990). Storing filters in sealed containers and under
18 refrigeration will minimize these losses.
19 Negative sampling artifacts due to decomposition and volatilization are also significant
20 for particulate ammonium. Ammonium particulates, especially NH4 N3 nitrate NH4 Cl are very
21 sensitive to some environmental factors, such as temperature, relative humidity, acidity of
22 aerosols, as well as to filter type (Spurny, 1999; Keck and Wittmaack, 2005). Any change in
23 these parameters during the sampling period influences the position of the equilibrium between
24 the particle phase and the gas phase. Keck and Wittmaack (2005) observed that at temperatures
25 below 0 °C, acetate-nitrate, quartz fiber, and Teflon filters could properly collect particulate NH4
26 NET? and Cl. At temperature above 0 °C, the salts were lost from quartz fiber and Teflon filters,
27 more so the higher the temperature and with no significant difference between quartz fiber and
28 Teflon filters. The salts were lost completely from denuded quartz fiber filters above about
29 20 °C, and from non-undenuded quartz fiber and Teflon filters above about 25 °C. It is
30 anticipated that current sampling techniques underestimate NH4 concentrations due to the
31 volatilization of NH4, but fine particle mass contains many acidic compounds and consequently,
March 2008 AX2-108 DRAFT-DO NOT QUOTE OR CITE
-------
1 a fraction of volatilized NH4 (in the form of NHa) can be retained on a PTFE filter by reaction
2 with the acid compounds. Therefore, it is reasonable to assume that NH4 loss will be less than
3 the nitrate loss. Techniques have been applied to paniculate ammonium sampling to correct
4 particulate ammonium concentrations due to evaporation: a backup filter coated with
5 hydrofluoric acids, citric acid, or phosphorous acids, is usually introduced to absorb the
6 evaporated ammonium (as ammonia); the total ammonium concentration is the sum of the
7 particle phase ammonium collected on the Teflon filter and the ammonia concentration collected
8 on the backup filter.
9
10 Other Measurement Techniques
11
12 Nitrate
13 An integrated collection and vaporization cell was developed by Stolzenburg and Hering
14 (2000) that provides automated, 10-min resolution monitoring of fine-particulate nitrate. In this
15 system, particles are collected by a humidified impaction process and analyzed in place by flash
16 vaporization and chemiluminescent detection of the evolved NOx. In field tests in which the
17 system was collocated with two FRM samplers, the automated nitrate sampler results followed
18 the results from the FRM, but were offset lower. The system also was collocated with a HEADS
19 and a SASS speciation sampler (MetOne Instruments). In all these tests, the automated sampler
20 was well correlated to other samplers with slopes near 1 (ranging from 0.95 for the FRM to 1.06
21 for the HEADS) and correlation coefficients ranging from 0.94 to 0.996. During the Northern
22 Front Range Air Quality Study in Colorado (Watson et al., 1998), the automated nitrate monitor
23 captured the 12-min variability in fine-particle nitrate concentrations with a precision of
24 approximately ± 0.5 |ig/m3 (Chow et al., 1998). A comparison with denuded filter
25 measurements followed by ion chromatographic (1C) analysis (Chow and Watson, 1999) showed
26 agreement within ± 0.6 |ig/m3 for most of the measurements, but exhibited a discrepancy of a
27 factor of two for the elevated nitrate periods. More recent intercomparisons took place during
28 the 1997 Southern California Ozone Study (SCOS97) in Riverside, CA. Comparisons with
29 14 days of 24-h denuder-filter sampling gave a correlation coefficient of R2 = 0.87 and showed
30 no significant bias (i.e., the regression slope is not significantly different from 1). As currently
31 configured, the system has a detection limit of 0.7 |ig/m3 and a precision of 0.2 |ig/m3.
March 2008 AX2-109 DRAFT-DO NOT QUOTE OR CITE
-------
1 Sulfate
2 Continuous methods for the quantification of aerosol sulfur compounds first remove
3 gaseous sulfur (e.g., SO2, H^S) from the sample stream by a diffusion tube denuder followed by
4 the analysis of particulate sulfur (Cobourn et al., 1978; Durham et al., 1978; Huntzicker et al.,
5 1978; Mueller and Collins, 1980; Tanner et al., 1980). Another approach is to measure total
6 sulfur and gaseous sulfur separately by alternately removing particles from the sample stream.
7 Particulate sulfur is obtained as the difference between the total and gaseous sulfur (Kittelson
8 et al., 1978). The total sulfur content is measured by a flame photometric detector (FPD) by
9 introducing the sampling stream into a fuel-rich, hydrogen-air flame (e.g., Stevens et al., 1969;
10 Farwell and Rasmussen, 1976) that reduces sulfur compounds and measures the intensity of the
11 chemiluminescence from electronically excited sulfur molecules (S2*). Because the formation
12 of S2* requires two sulfur atoms, the intensity of the chemiluminescence is theoretically
13 proportional to the square of the concentration of molecules that contain a single sulfur atom.
14 In practice, the exponent is between 1 and 2 and depends on the sulfur compound being analyzed
15 (Dagnall et al., 1967; Stevens et al., 1971). Calibrations are performed using both particles and
16 gases as standards. The FPD can also be replaced by a chemiluminescent reaction with ozone
17 that minimizes the potential for interference and provides a faster response time (Benner and
18 Stedman, 1989, 1990). Capabilities added to the basic system include in situ thermal analysis
19 and sulfuric acid speciation (Cobourn et al., 1978; Huntzicker et al., 1978; Tanner et al., 1980;
20 Cobourn and Husar, 1982). Sensitivities for particulate sulfur as low as 0.1 |ig/m3, with time
21 resolution ranging from 1 to 30 min, have been reported. Continuous measurements of
22 particulate sulfur content have also been obtained by on-line XRF analysis with resolution of
23 30 min or less (Jaklevic et al., 1981). During a field-intercomparison study of five different
24 sulfur instruments, Camp et al. (1982) reported four out of five FPD systems agreed to within
25 ± 5% during a 1-week sampling period.
26
27
28 AX2.9 POLICY RELEVANT BACKGROUND CONCENTRATIONS OF
29 NITROGEN AND SULFUR OXIDES
30 Background concentrations of nitrogen and sulfur oxides used for purposes of informing
31 decisions about NAAQS are referred to as Policy Relevant Background (PRB) concentrations.
32 Policy Relevant Background concentrations are those concentrations that would occur in the
March 2008 AX2-110 DRAFT-DO NOT QUOTE OR CITE
-------
1 United States in the absence of anthropogenic emissions in continental North America (defined
2 here as the United States, Canada, and Mexico). Policy Relevant Background concentrations
3 include contributions from natural sources everywhere in the world and from anthropogenic
4 sources outside these three countries. Background levels so defined facilitate separation of
5 pollution levels that can be controlled by U.S. regulations (or through international agreements
6 with neighboring countries) from levels that are generally uncontrollable by the United States.
7 EPA assesses risks to human health and environmental effects from NC>2 and SC>2 levels in
8 excess of PRB concentrations.
9 Contributions to PRB concentrations include natural emissions of NC>2, 862, and
10 photochemical reactions involving natural emissions of reduced nitrogen and sulfur compounds,
11 as well as their long-range transport from outside North America. Natural sources of NO2 and its
12 precursors include biogenic emissions, wildfires, lightning, and the stratosphere. Natural sources
13 of reduced nitrogen compounds, mainly NHa, include biogenic emissions and wildfires. Natural
14 sources of reduced sulfur species include anaerobic microbial activity in wetlands and volcanic
15 activity. Volcanos and biomass burning are the major natural source of SC>2. Biogenic
16 emissions from agricultural activities are not considered in the formation of PRB concentrations.
17 Discussions of the sources and estimates of emissions are given in Section AX2.6.2.
18
19 Analysis of PRB Contribution to Nitrogen and Sulfur oxide Concentrations and Deposition
20 over the United States
21 The MOZART-2 global model of tropospheric chemistry (Horowitz et al., 2003) is used
22 to diagnose the PRB contribution to nitrogen and sulfur oxide concentrations, as well as to total
23 (wet plus dry) deposition. The model setup for the present-day simulation has been published in
24 a series of papers from a recent model intercomparison (Dentener et al., 2006a,b; Shindell et al.,
25 2006; Stevenson et al., 2006; Van Noije et al., 2006). MOZART-2 is driven by National Center
26 for Environmental Prediction meteorological fields and IIASA 2000 emissions at a resolution of
27 1.9° x 1.9° with 28 sigma levels in the vertical, and it includes gas- and aerosol phase chemistry.
28 Results shown in Figures AX2.9-1 to AX2.9-5 are for the meteorological year 2001. Note that
29 color images are available on the web. An additional "policy relevant background" simulation
30 was conducted in which continental North American anthropogenic emissions were set to zero.
31 We first examine the role of PRB in contributing to NC>2 and SC>2 concentrations in
32 surface air. Figure AX2.9-1 shows the annual mean NC>2 concentrations in surface air in the base
March 2008 AX2-111 DRAFT-DO NOT QUOTE OR CITE
-------
Total
12D°W
100°W
290
530
770
1010
1250
Background
120°W
10Q°W
25
45
65
85
105
125
Percent Background Contribution
iaa°w
100°W
14
23
32
41
50
Figure AX2.9-1.
Annual mean concentrations of NOi (ppbv) in surface air over the
United States in the present-day (upper panel) and policy relevant
background (middle panel) MOZART-2 simulations. The bottom
panel shows the percentage contribution of the background to the
present-day concentrations. Please see text for details.
March 2008
AX2-112
DRAFT-DO NOT QUOTE OR CITE
-------
Total
so°w
< D.D1
1.21
2.41
3.BO 4.80
B.OO
ppb
Background
100°W
oo°w
< 0.001 0.006 0.011 0.015 0.020 0.025
ppb
Percent Background Contribution
ioo°w
so°w
10
15
20
25
Figure AX2.9-2. Same as Figure AX2.9-1 but for SO2 concentrations.
March 2008
AX2-113
DRAFT-DO NOT QUOTE OR CITE
-------
Total
12Q°W
100°W
ao°w
50
290
530
770 1010
1250
Background
25
45
B5
105
125
Percent Background Contribution
so°w
14
23
32
41
50
Figure AX2.9-3. Same as for Figure AX2.9-1 but for wet and dry deposition of HNO3,
-2 -K
NH4NO3, NOX, HO2NO2, and organic nitrates (mg N m" y ).
March 2008
AX2-114
DRAFT-DO NOT QUOTE OR CITE
-------
Total
10Q°W
ao°w
< 100 800 1500 2200 2900 3BOO
10
Background
ioo°w
so°w
16
22
34
40
Percent Background Contribution
10
20
30
40
50
Figure AX2.9-4. Same as Figure AX2.9-1 but for SOX deposition (SO2 + SO4)
(mg S m-y1).
March 2008
AX2-115 DRAFT-DO NOT QUOTE OR CITE
-------
W.OZART-2 SOIL WO
GEOS-Cam SOIL UD.
-. n a ii *fl 22 2B
MOZAHT-2 SL*rac* MO JUL
GEQS-ChWfi Surface NO
-: SD 150 2» S'JB 450
» EH) 190 23O 3SD 490 33O
Figure AX2.9-5.
July mean soil NO emissions (upper panels; 1 x 10 9 molecules cm 2 s1)
and surface PRB NOx concentrations (lower panels; pptv) over the
United States from MOZART-2 (left) and GEOS-Chem (right) model
simulations in which anthropogenic Os precursor emissions were set
to zero in North America.
1 case simulation (top panel) and the PRB simulation (middle panel), along with the percentage
2 contribution of the background to the total base case NO2 (bottom panel). Maximum
3 concentrations in the base case simulation occur along the Ohio River Valley and in the
4 Los Angeles basin. While present-day concentrations are often above 5 ppbv, PRB is less than
5 300 pptv over most of the continental United States, and less than 100 pptv in the eastern United
6 States. The distribution of PRB (middle panel of Figure AX2.9-1) largely reflects the
7 distribution of soil NO emissions, with some local enhancements due to biomass burning such as
8 is seen in western Montana. In the northeastern United States, where present-day
9 concentrations are highest, PRB contributes <1% to the total.
March 2008
AX2-116
DRAFT-DO NOT QUOTE OR CITE
-------
1 The spatial pattern of present-day 862 concentrations over the United States is similar to
2 that of NO2, with highest concentrations (>5 ppbv) along the Ohio River valley (upper panel
3 Figure AX2.9-2). Background SO2 concentrations are orders of magnitude smaller, below 10
4 pptv over much of the United States (middle panel of Figure AX2.9-2). Maximum PRB
5 concentrations of SC>2 are 30 ppt. In the Northwest where there are geothermal sources of SO2,
6 the contribution of PRB to total SC>2 is 70 to 80%. However, with the exception of the West
7 Coast where volcanic SC>2 emissions enhance PRB concentrations, the PRB contributes <1% to
8 present-day 862 concentrations in surface air (bottom panel Figure AX2.9-2).
9 The spatial pattern of NOY (defined here as HNO3, NH4NO3, NOX, HO2NO2, and organic
10 nitrates) wet and dry deposition is shown in Figure AX2.9-3. Figure AX2.9-3 (upper panel)
11 shows that highest values are found in the eastern United States in and downwind of the Ohio
12 River Valley. The pattern of nitrogen deposition in the PRB simulation (Figure AX2.9-3, middle
13 panel), however, shows maximum deposition centered over Texas and in the Gulf Coast region,
14 reflecting a combination of nitrogen emissions from lightning in the Gulf region, biomass
15 burning in the Southeast, and from microbial activity in soils (maximum in central Texas and
16 Oklahoma). The bottom panel of Figure AX2.9-3 shows that the PRB contribution to nitrogen
17 deposition is less than 20% over the eastern United States, and typically less than 50% in the
18 western United States where NOy deposition is low (25-50 mg N nT2 yr"1).
19 Present-day SOx (SO2 + SO4 ~) deposition is largest in the Ohio River Valley, likely due
20 to coal-burning power plants in that region, while background deposition is typically at least an
21 order of magnitude smaller (Figure AX2.9-4). Over the eastern United States, the background
22 contribution to SOx deposition is <10%, and it is even smaller (<1%) where present-day SOx
23 deposition is highest. The contribution of PRB to sulfate deposition is highest in the western
24 United States (>20%) because of geothermal sources of SO2 and oxidation of dimethyl sulfide in
25 the surface of the eastern Pacific.
26 Thus far, the discussion has focused on results from the MOZART-2 tropospheric
27 chemistry model. In Figure AX2.9-5, results from MOZART-2 are compared with those from
28 another tropospheric chemistry model, GEOS-Chem (Bey et al., 2001), which was previously
29 used to diagnose PRB O3 (Fiore et al., 2003; U.S. Environmental Protection Agency, 2006). In
30 both models, the surface PRB NOx concentrations tend to mirror the distribution of soil NO
31 emissions, which are highest in the Midwest. The higher soil NO emissions in GEOS-Chem (by
March 2008 AX2-117 DRAFT-DO NOT QUOTE OR CITE
-------
1 nearly a factor of 2) as compared to MOZART-2 reflect different assumptions regarding the
2 contribution to soil NO emissions largely through fertilizer, since GEOS-Chem total soil NO
3 emissions are actually higher than MOZART-2 (0.07 versus 0.11 Tg N) over the United States in
4 July. Even with the larger PRB soil NO emissions, surface NOx concentrations in GEOS-Chem
5 are typically below 500 pptv.
6 It is instructive to also consider measurements of SO2 at relatively remote monitoring
7 sites, i.e., site located in sparsely populate areas not subject to obvious local sources of pollution.
8 Berresheim et al. (1993) used a type of atmospheric pressure ionization mass spectrometer
9 (APIMS) at Cheeka Peak, WA (48.30N 124.62W, 480 m asl), in April 1991 during a field study
10 for DMS oxidation products. Sulfur Dioxide concentrations ranged between 20 and 40 pptv.
11 Thornton et al. (2002) have also used an APIMS with an isotopically labeled internal standard to
12 determine background SO2 levels. SO2 concentrations of 25 to 40 pptv were observed in
13 northwestern Nebraska in October 1999 at 150m above ground using the NCAR C-130
14 (Thornton, unpublished data). These data are comparable to remote central south Pacific
15 convective boundary layer SO2 (Thornton et al., 1999).
16 Volcanic sources of SO2 in the United States are limited to the Pacific Northwest, Alaska,
17 and Hawaii. Since 1980 the Mt. St. Helens volcano in Washington Cascade Range (46.20 N,
18 122.18 W, summit 2549 m asl) has been a variable source of SO2. Its major impact came in the
19 explosive eruptions of 1980, which primarily affected the northern part of the mountain west of
20 the United States. The Augustine volcano near the mouth of the Cook Inlet in southwestern
21 Alaska (59.363 N, 153.43 W, summit 1252 m asl) has had SO2 emissions of varying extents
22 since its last major eruptions in 1986. Volcanoes in the Kamchatka peninsula of eastern region
23 of Siberian Russia do not particularly impact the surface concentrations in the northwestern NA.
24 The most serious impact in the United States from volcanic SO2 occurs on the island of Hawaii.
25 Nearly continuous venting of SO2 from Mauna Loa and Kilauea produce SO2 in such large
26 amounts so that >100 km downwind of the island SO2 concentrations can exceed 30 ppbv
27 (Thornton and Bandy, 1993). Depending on the wind direction the west coast of Hawaii (Kona
28 region) has had significant impacts from SO2 and acidic sulfate aerosols for the past decade.
29 Indeed, SO2 levels in Volcanoes National Park, HI exceeded the 3-h and the 24-h NAAQS in
30 2004 -2005. The area's design value is 0.6 ppm for the 3-h, and 0.19 ppm for the 24-h NAAQS
31 (U.S. Environmental Protection Agency, 2006).
March 2008 AX2-118 DRAFT-DO NOT QUOTE OR CITE
-------
1 Overall, the background contribution to nitrogen and sulfur oxides over the United States
2 is relatively small, except for 862 in areas where there is volcanic activity.
March 2008 AX2-119 DRAFT-DO NOT QUOTE OR CITE
-------
TABLE AX2.3-1. ATMOSPHERIC LIFETIMES OF SULFUR DIOXIDE AND
REDUCED SULFUR SPECIES WITH RESPECT TO REACTION WITH OH, NO3,
AND CL RADICALS
Compound
S02
CH3-S-CH3
H2S
CS2
DCS
CH3-S-H
CH3-S-S-CH3
OH
k x 1012
1.6
5.0
4.7
1.2
0.0019
33
230
T
7.2d
2.3d
2.2 d
9.6 d
17 y
8.4 h
1.2 h
N03
k x 1012
NA
1.0
NA
0.0004
0.0001
0.89
0.53
T
l.lh
>116d
>1.3y
1.2 h
2.1 h
Cl
k x 1012
NA
400
74
0.004
0.0001
200
NA
T
29 d
157 d
NR
NR
58 d
Notes:
NA = Reaction rate coefficient not available. NR = Rate coefficient too low to be relevant as an atmospheric loss mechanism. Rate
coefficients were calculated at 298 K and 1 atmosphere.
y = year, d = day. h = hour. OH = 1 x 106/cm3; NO3 = 2.5 x 108/cm3; Cl = 1 x 103/cm3.
'Rate coefficients were taken from JPL Chemical Kinetics Evaluation No. 14 (JPL, 2003).
Source: Seinfeld and Pandis (1998).
TABLE AX2.4-la. RELATIVE CONTRIBUTIONS OF VARIOUS REACTIONS TO
THE TOTAL S(IV) OXIDATION RATE WITHIN A SUNLIT CLOUD, 10 MINUTES
AFTER CLOUD FORMATION
Reaction
Gas Phase
OH + SO2
Aqueous Phase
O3 + HSCV
O3 + SO32~
H2O2 + SCV
CH3OOH + HSCV
HNO4 + HSCV
HOONO + HSCV
HSCV + HSCV
SCV + SO32~
HSCV + Fe2+
% of Total3
3.5
0.6
7.0
78.4
0.1
9.0
O.I
1.2
0.1
% of Total b
3.1
0.7
8.2
82.1
0.1
4.4
O.I
O.I
0.1
0.6
aln the absence of transition metals.
bln the presence of iron and copper ions.
Source: Adapted from Warneck (1999).
March 2008
AX2-120
DRAFT-DO NOT QUOTE OR CITE
-------
TABLE AX2.4-lb. RELATIVE CONTRIBUTIONS OF VARIOUS GAS AND
AQUEOUS PHASE REACTIONS TO AQUEOUS NITRATE FORMATION WITHIN
A SUNLIT CLOUD, 10 MINUTES AFTER CLOUD FORMATION
Reaction
Gas Phase
OH + NO2 + M
Aqueous Phase
N2O5g + H2O
NO3 + cr
NO3 + HSO3~
NO3 + HCOCT
HNO4 + HSO3~
HOONO+NCV
O3 + NO2
% of Total3
57.7
8.1
0.1
0.7
0.6
31.9
0.8
O.I
% of Total b
67.4
11.2
0.1
1.0
0.8
20.5
0.1
O.I
a In the absence of transition metals.
b In the presence of iron and copper ions.
Source: Adapted from Warneck (1999).
March 2008 AX2-121 DRAFT-DO NOT QUOTE OR CITE
-------
TABLE AX2.6-1. EMISSIONS OF NITROGEN OXIDES, AMMONIA, AND SULFUR
DIOXIDE IN THE UNITED STATES IN 2002
2002 Emissions (Tg/yr)
Source Category
TOTAL ALL SOURCES
FUEL COMBUSTION TOTAL
FUEL COMB. ELEC. UTIL
Coal
Bituminous
Subbituminous
anthracite & lignite
Other
Oil
Residual
Distillate
Gas
Natural
Process
Other
Internal Combustion
FUEL COMBUSTION INDUSTRIAL
Coal
Bituminous
Subbituminous
Anthracite & Lignite
Other
Oil
Residual
Distillate
Other
Gas
Natural
Process
Other
Other
wood/bark waste
liquid waste
Other
Internal Combustion
FUEL COMB. OTHER
Commercial/Institutional Coal
Commercial/Institutional Oil
Commercial/Institutional Gas
Misc. Fuel Comb. (Except Residential)
Residential Wood
NOx1
23.19
9.11
5.16
4.50
2.90
1.42
0.18
<0.01
0.14
0.13
0.01
0.30
0.29
0.01
0.05
0.17
3.15
0.49
0.25
0.07
0.04
0.13
0.19
0.09
0.09
0.01
1.16
0.92
0.24
0.01
0.16
0.11
0.01
0.04
1.15
0.80
0.04
0.08
0.25
0.03
0.03
NH3
4.08
0.02
0.01
0.01
O.01
0.01
O.01
O.01
O.01
O.01
0.01
O.01
0.01
0.01
0.01
0.01
0.01
O.01
O.01
S02
16.87
14.47
11.31
10.70
8.04
2.14
0.51
0.38
0.36
0.01
0.01
0.21
0.01
2.53
1.26
0.70
0.10
0.13
0.33
0.59
0.40
0.16
0.02
0.52
0.15
0.01
0.63
0.16
0.28
0.02
0.01
O.01
March 2008 AX2-122 DRAFT-DO NOT QUOTE OR CITE
-------
TABLE AX2.6-1 (cont'd). EMISSIONS OF NITROGEN OXIDES, AMMONIA, AND
SULFUR DIOXIDE IN THE UNITED STATES IN 2002
2002 Emissions (Tg/yr)
Residential Other
distillate oil
bituminous/subbituminous
Other
INDUSTRIAL PROCESS TOTAL
CHEMICAL & ALLIED PRODUCT
MFC
Organic Chemical Mfg
Inorganic Chemical Mfg
Sulfur compounds
Other
Polymer & Resin Mfg
Agricultural Chemical Mfg
ammonium nitrate/urea mfg.
Other
Paint, Varnish, Lacquer, Enamel Mfg
Pharmaceutical Mfg
Other Chemical Mfg
METALS PROCESSING
Non-Ferrous Metals Processing
Copper
Lead
Zinc
Other
Ferrous Metals Processing
Metals Processing
PETROLEUM & RELATED
INDUSTRIES
Oil & Gas Production
natural gas
Other
Petroleum Refineries & Related
Industries
fluid catalytic cracking units
Other
Asphalt Manufacturing
OTHER INDUSTRIAL PROCESSES
Agriculture, Food, & Kindred Products
Textiles, Leather, & Apparel Products
Wood, Pulp & Paper, & Publishing
Products
NOx1
0.36
0.06
0.26
0.04
1.10
0.12
0.02
0.01
<0.01
0.05
0.00
0.00
0.03
0.09
0.01
0.07
0.01
0.16
0.07
0.05
0.04
0.54
0.01
<0.01
0.09
NH3
0.21
0.02
0.01
O.01
O.01
0.02
0.01
0.02
O.01
O.01
O.01
0.01
O.01
O.01
O.01
0.01
0.01
0.01
0.05
O.01
O.01
0.01
S02
0.16
0.15
0.01
0.01
1.54
0.36
0.01
0.18
0.17
0.02
O.01
0.05
0.00
0.00
0.12
0.30
0.17
0.04
0.07
0.01
0.01
0.11
0.02
0.38
0.11
0.11
0.01
0.26
0.16
0.07
0.01
0.46
0.01
O.01
0.10
March 2008 AX2-123 DRAFT-DO NOT QUOTE OR CITE
-------
TABLE AX2.6-1 (cont'd). EMISSIONS OF NITROGEN OXIDES, AMMONIA, AND
SULFUR DIOXIDE IN THE UNITED STATES IN 2002
2002 Emissions (Tg/yr)
Rubber & Miscellaneous Plastic
Products
Mineral Products
cement mfg
glass mfg
Other
Machinery Products
Electronic Equipment
Transportation Equipment
Miscellaneous Industrial Processes
SOLVENT UTILIZATION
Degreasing
Graphic Arts
Dry Cleaning
Surface Coating
Other Industrial
Nonindustrial
Solvent Utilization NEC
STORAGE & TRANSPORT
Bulk Terminals & Plants
Petroleum & Petroleum Product Storage
Petroleum & Petroleum Product
Transport
Service Stations: Stage II
Organic Chemical Storage
Organic Chemical Transport
Inorganic Chemical Storage
Inorganic Chemical Transport
Bulk Materials Storage
WASTE DISPOSAL & RECYCLING
Incineration
Industrial
Other
Open Burning
Industrial
Land clearing debris
Other
Public Operating Treatment Works
Industrial Waste Water
Treatment, Storage, and Disposal
Facility
Landfills
Industrial
NOx1
O.01
0.42
0.24
0.01
0.10
0.01
0.01
O.01
0.01
0.01
O.01
0.01
0.01
0.01
0.01
O.01
O.01
O.01
O.01
0.01
0.01
0.01
0.01
0.01
O.01
O.01
0.01
0.17
0.06
0.10
O.01
0.01
0.01
0.01
NH3
O.01
O.01
0.01
0.01
0.05
O.01
O.01
0.01
0.01
0.01
0.01
O.01
O.01
0.01
0.01
0.01
O.01
O.01
0.14
O.01
0.01
0.14
0.01
0.01
0.01
S02
O.01
0.33
0.19
0.09
0.01
0.01
O.01
0.02
O.01
O.01
0.01
0.01
0.01
0.01
0.01
O.01
0.01
0.01
0.01
0.01
0.01
O.01
O.01
O.01
0.03
0.02
0.01
0.01
0.01
O.01
O.01
0.01
0.01
0.01
0.01
March 2008 AX2-124 DRAFT-DO NOT QUOTE OR CITE
-------
TABLE AX2.6-1 (cont'd). EMISSIONS OF NITROGEN OXIDES, AMMONIA, AND
SULFUR DIOXIDE IN THE UNITED STATES IN 2002
2002 Emissions (Tg/yr)
Other
Other
TRANSPORTATION TOTAL
HIGHWAY VEHICLES
Light-Duty Gas Vehicles & Motorcycles
light-duty gas vehicles
Motorcycles
Light-Duty Gas Trucks
light-duty gas trucks 1
light-duty gas trucks 2
Heavy-Duty Gas Vehicles
Diesels
heavy-duty diesel vehicles
light-duty diesel trucks
light-duty diesel vehicles
OFF-HIGHWAY
Non-Road Gasoline
Recreational
Construction
Industrial
lawn & garden
Farm
light commercial
Logging
airport service
railway maintenance
recreational marine vessels
Non-Road Diesel
Recreational
Construction
Industrial
lawn & garden
Farm
light commercial
Logging
airport service
railway maintenance
recreational marine vessels
Aircraft
Marine Vessels
Diesel
residual oil
Other
NOX 1 NH3
<0.01 <0.01
12.58 0.32
8.09 0.32
2.38 0.20
2.36
0.02
1.54 0.10
1.07
0.47
0.44 <0.01
3.73 0.01
3.71
0.01
0.01
4.49 O.01
0.23 O.01
0.01
0.01
0.01
0.10
0.01
0.04
0.01
O.01
O.01
0.05
1.76 O.01
0.00
0.84
0.15
0.05
0.57
0.08
0.02
0.01
0.01
0.03
0.09
1.11
1.11
S02
O.01
O.01
0.76
0.30
0.10
0.10
0.00
0.07
0.05
0.02
0.01
0.12
0.46
0.01
0.22
0.01
0.18
March 2008
AX2-125
DRAFT-DO NOT QUOTE OR CITE
-------
TABLE AX2.6-1 (cont'd). EMISSIONS OF NITROGEN OXIDES, AMMONIA, AND
SULFUR DIOXIDE IN THE UNITED STATES IN 2002
2002 Emissions (Tg/yr)
Railroads
Other
liquefied petroleum gas
compressed natural gas
MISCELLANEOUS
Agriculture & Forestry
agricultural crops
agricultural livestock
Other Combustion
Health Services
Cooling Towers
Fugitive Dust
Other
Natural Sources
NOx1
0.98
0.32
0.29
0.04
0.39
0.01
3.10
NH3
O.01
3.53
3.45
O.01
2.66
0.08
0.03
S02
0.05
0.00
0.10
0.01
0.10
1 Emissions are expressed in terms of NO2.
Emissions based on Guenther et al. (2000).
Source: U.S. Environmental Protection Agency (2006).
TABLE AX2.8-1. SATELLITE INSTRUMENTS USED TO RETRIEVE
TROPOSPHERIC NO2 COLUMNS.
Instrument
GOME
SCIAMACHY
OMI
Coverage
1995-2002
2002-
2004-
Typical U.S.
Measurement Time
10:30-11:30 AM
10:00-11:00 AM
12:45-1:45 PM
Typical
Resolution
(km)
320 x 40
30x60
13 x24
Return
Time
(days)1
3
6
1
Instrument
Overview
Burrows et al.
(1999)
Bovensmann
etal. (1999)
Levelt et al.
(2006)
1 Return time is reported here for cloud free conditions. Note that due to precession of the satellite's orbit, return
measurements are close to but not made over the same location. In practice, clouds decrease observation frequency by a
factor of 2.
March 2008
AX2-126
DRAFT-DO NOT QUOTE OR CITE
-------
1 AX2.10 REFERENCES
2 Aber, I; McDowell, W.; Nadelhoffer, K.; Magill, A.; Berntson, G.; Kamakea, M.; McNulty, S.;
3 Currie, W.; Rustad, L.; Fernandez, I. (1998) Nitrogen saturation in temperate forest
4 ecosystems. BioScience 48: 921-934.
5 Alicke, B.; Hebestreit, K.; Stutz, J.; Platt, U. (1999) Iodine oxide in the marine boundary layer.
6 Nature (London, U.K.) 397: 572-573.
7 Allan, B. J.; McFiggans, G.; Plane, J. M. C.; Coe, H. (2000) Observations of iodine monoxide in
8 the remote marine boundary layer. J. Geophys. Res. [Atmos.] 105: 14,363-14,369.
9 Allen, D. J.; Pickering, K. E. (2002) Evaluation of lightning flash rate parameterizations for use
10 in a global chemical transport model. J. Geophys. Res. [Atmos.] 107(D23):
11 10.1029/2002JD002066.
12 Allen, D. J.; Pickering, K. E.; Molod, A. (1997) An evaluation of deep convective mixing in the
13 Goddard Chemical Transport Model using ISCCP cloud parameters. J. Geophys. Res.
14 [Atmos.] 102: 25,467-25,476.
15 Allen, D.; Pickering, K.; Stenchikov, G.; Thompson, A.; Kondo, Y. (2000) A three-dimensional
16 total odd nitrogen (NOy) simulation during SONEX using a stretched-grid chemical
17 transport model. J. Geophys. Res. [Atmos.] 105: 3851-3876.
18 Ammann, M.; Kalberer, M.; lost, D. T.; Tobler, L.; Rossler, E.; Piguet, D.; Gaggeler, H. W.;
19 Baltensperger, U. (1998) Heterogeneous production of nitrous acid on soot in polluted air
20 masses. Nature (London) 395: 157-160.
21 Anastasio, C.; Newberg, J. T.; Williams, D. K.; Chu, G. B.; Matthew, B. M. (1999)
22 Photoformation of hydroxyl radical in sea salt particles. EOS Trans. Am. Geophys. Union
23 80:F147.
24 Andreae, M. O. (1991) Biomass burning: its history, use, and distribution and its impact on
25 environmental quality and global climate. In: Levine, J. S., ed. Global biomass burning:
26 atmospheric, climatic, and biospheric implications. Cambridge, MA: MIT Press; pp. 1-
27 21.
28 Appel, B. R.; Tokiwa, Y.; Haik, M. (1981) Sampling of nitrates in ambient air. Atmos. Environ.
29 15:283-289.
30 Appel, K. W.; Gilliland, A.; Eder, B. (2005) An operational evaluation of the 2005 release of
31 models-3 CMAQ version 4.5. Washington, DC: National Oceanic and Atmospheric
32 Administration (NOAA) Air Resources Laboratory, Atmospheric Sciences Modeling
33 Division.
34 Arey, J. (1998) Atmospheric reactions of PAHs including formation of nitroarenes. In: Neilson,
35 A. N., vol. ed. Anthropogenic Compounds. V. 3, Part I. PAHs and Related Compounds -
36 Chemistry. New York, NY: Springer-Verlag; pp. 347-385. (Hutzinger, O., ed. The
37 Handbook of Environmental Chemistry series).
38 Arey, J.; Zielinska, B.; Atkinson, R.; Winer, A. M.; Ramdahl, T.; Pitts, J. N., Jr. (1986) The
39 formation of nitro-PAH from the gas-phase reactions of fluoranthene and pyrene with the
40 OH radical in the presence of NOX. Atmos. Environ. 20: 2339-2345.
March 2008 AX2-127 DRAFT-DO NOT QUOTE OR CITE
-------
1 Arey, J.; Atkinson, R.; Zielinska, B.; McElroy, P. A. (1989) Diurnal concentrations of volatile
2 polycyclic aromatic hydrocarbons and nitroarenes during a photochemical air pollution
3 episode in Glendora, California. Environ. Sci. Technol. 23: 321-327.
4 Arnold, J. R.; Dennis, R. L.; Tonnesen, G. S. (2003) Diagnostic evaluation of numerical air
5 quality models with specialized ambient observations: testing the Community Multiscale
6 Air Quality modeling system (CMAQ) at selected SOS 95 ground sites. Atmos. Environ.
7 37:1185-1198.
8 Ashworth, S. H.; Allan, B. J.; Plane, J. M. C. (2002) High resolution spectroscopy of the OIO
9 radical: implications for the ozone-depleting potential of iodine. Geophys. Res. Lett. 29:
10 10.1029/GL013851.
11 Atkinson, R. (1991) Kinetics and mechanisms of the gas-phase reactions of the NOs radical with
12 organic compounds. J. Phys. Chem. Ref Data 20: 459-507.
13 Atkinson, R. (2000) Atmospheric chemistry of VOCs and NOX. Atmos. Environ. 34: 2063-2101.
14 Atkinson, R.; Arey, J. (1994) Atmospheric chemistry of gas-phase polycyclic aromatic
15 hydrocarbons: formation of atmospheric mutagens. Environ. Health Perspect. 102(suppl.
16 4): 117-126.
17 Atkinson, R.; Winer, A. M.; Pitts, J. N., Jr. (1986) Estimation of night-time ^Os concentrations
18 from ambient NC>2 and NOs radical concentrations and the role of ^Os in night-time
19 chemistry. Atmos. Environ. 20: 331-339.
20 Atkinson, R.; Arey, J.; Zielinska, B.; Aschmann, S. M. (1990) Kinetics and nitro-products of the
21 gas-phase OH and NOs radical-initiated reactions of naphthalene-fife, fluoranthene-t/io and
22 pyrene. Int. J. Chem. Kinet. 22: 999-1014.
23 Atkinson, R.; Baulch, D. L.; Cox, R. A.; Hampson, R. F., Jr.; Kerr, J. A.; Troe, J. (1992)
24 Evaluated kinetic and photochemical data for atmospheric chemistry: supplement IV,
25 IUPAC Subcommittee on Gas Kinetic Data Evaluation for Atmospheric Chemistry. J.
26 Phys. Chem. Ref. Data 21: 1125-1568.
27 Babich, P.; Davey, M.; Allen, G.; Koutrakis, P. (2000) Method comparisons for particulate
28 nitrate, elemental carbon, and PM2.5 mass in seven U.S. cities. J. Air Waste Manage.
29 Assoc. 50: 1095-1105.
30 Balkanski, Y. J.; Jacob, D. J.; Gardener, G. M.; Graustein, W. C.; Turekian, K. K. (1993)
31 Transport and residence times of tropospheric aerosols inferred from a global three-
32 dimensional simulation of 210Pb. J. Geophys. Res. [Atmos.] 98: 20,573-20,586.
33 Bamford, H. A.; Baker, J. E. (2003) Nitro-polycyclic aromatic hydrocarbon concentrations and
34 sources in urban and suburban atmospheres of the mid-Atlantic region. Atmos. Environ.
35 37:2077-2091.
36 Bamford, H. A.; Bezabeh, D. Z.; Schantz, M. M.; Wise, S. A.; Baker, J. E. (2003) Determination
37 and comparison of nitrated-polycyclic aromatic hydrocarbons measured in air and diesel
38 particulate reference materials. Chemosphere 50: 575-587.
39
40
March 2008 AX2-128 DRAFT-DO NOT QUOTE OR CITE
-------
1 Bardwell, C. A.; Maben, J. R.; Hurt, J. A.; Keene, W. C.; Galloway, J. N.; Boatman, J. F.;
2 Wellman, D. L. (1990) A technique using high-flow, dichotomous filter packs for
3 measuring major atmospheric chemical constituents. Global Biogeochem. Cycles 4: 151-
4 163.
5 Barnes, I.; Bastian, V.; Becker, K. H.; Overath, R. D. (1991) Kinetic studies of the reactions of
6 IO, BrO and CIO with Dimethylsulfide. Int. J. Chem. Kinet. 23: 579-591.
7 Barrie, L. A.; Bottenheim, J. W.; Schnell, R. C.; Crutzen, P. J.; Rasmussen, R. A. (1988) Ozone
8 destruction and photochemical reactions at polar sunrise in the lower Arctic atmosphere.
9 Nature (London, U.K.) 334: 138-141.
10 Behnke, W.; George, C.; Sheer, V.; Zetzsch, C. (1997) Production and decay of C1NO2 from the
11 reaction of gaseous ^Os with NaCl solution: bulk and aerosol experiments. J. Geophys.
12 Res. [Atmos.] 102: 3795-3804.
13 Benner, R. L.; Stedman, D. H. (1989) Universal sulfur detection by chemiluminescence. Anal.
14 Chem. 61: 1268-1271.
15 Benner, R. L.; Stedman, D. H. (1990) Field evaluation of the sulfur chemiluminescence detector.
16 Environ. Sci. Technol. 24: 1592-1596.
17 Bergstrom, A.-K.; Jansson, M. (2006) Atmospheric nitrogen deposition has caused nitrogen
18 enrichment and eutrophication of lakes in the northern hemisphere. Glob. Change Biol.
19 12: 635-643.
20 Berresheim, H.; Eisele, F. L.; Tanner, D. J.; Mclnnes, L. M.; Ramsey-Bell, D. C.; Covert, D. S.
21 (1993) Atmospheric sulfur chemistry and cloud condensation nuclei (CCN)
22 concentrations over the northeastern Pacific coast. J. Geophys. Res. [Atmos.] 98: 12,701-
23 12,711.
24 Berresheim, H.; Wine, P. H.; Davis, D. D. (1995) Sulfur in the atmosphere. In: Singh, H. B., ed.
25 Composition, chemistry, and climate of the atmosphere. New York, NY: Van Nostrand
26 Reinhold; pp. 251-307.
27 Bertram, T. H.; Heckel, A.; Richter, A.; Burrows, J. R.; Cohen, R. C. (2005) Satellite
28 measurements of daily variations in soil NO* emissions. Geophys. Res. Lett. 32(L24812):
29 10.1029/2005GL024640.
30 Bey, L; Jacob, D. J.; Yantosca, R. M.; Logan, J. A.; Field, B.; Fiore, A. M.; Li, Q.; Liu, H.;
31 Mickley, L. J.; Schultz, M. G. (2001) Global modeling of tropospheric chemistry with
32 assimilated meteorology: model description and evaluation. J. Geophys. Res. [Atmos.]
33 106:23,073-23,095.
34 Bezabeh, D. Z.; Bamford, H. A.; Schantz, M. M.; Wise, S. A. (2003) Determination of nitrated
35 poly cyclic aromatic hydrocarbons in diesel parti culate-r elated standard reference
36 materials by using gas chromatography/mass spectrometry with negative ion chemical
37 ionization. Anal. Bioanal. Chem. 375: 381-388.
38 Binkowski, F. S.; Arunachalam, S.; Adelman, Z.; Pinto, J. P. (2007) Examining photolysis rates
39 with a prototype online photolysis module in CMAQ. J. Appl. Meteorol. Climatol. 46:
40 1252-1256.
March 2008 AX2-129 DRAFT-DO NOT QUOTE OR CITE
-------
1 Bishop, G. A.; Stedman, D. H. (1996) Measuring the emissions of passing cars. Ace. Chem. Res.
2 29:489-495.
3 Blanchard, P.; Brook, J. R.; Brazil, P. (2002) Chemical characterization of the organic fraction of
4 atmospheric aerosol at two sites in Ontario, Canada. J. Geophys. Res. [Atmos.]
5 107(D21): 10.1029/2001JD000627.
6 Bloom, S. C.; Takacs, L. L.; Da Silva, A. M.; Ledvina, D. (1996) Data assimilation using
7 incremental analysis updates. Mon. Weather Rev. 124: 1256-1271.
8 Bloom, S. C.; da Silva, A.; Dee, D.; Bosilovich, M.; Chern, J.-D.; Pawson, S.; Schubert, S.;
9 Sienkiewicz, M.; Stajner, L; Tan, W.-W.; Wu, M.-L. (2005) Documentation and
10 validation of the Goddard Earth Observing System (GEOS) Data Assimilation System—
11 version 4. Greenbelt, MD: National Aeronautics and Space Administration (NASA),
12 Goddard Space Flight Center. (Technical Report Series on Global Modeling and Data
13 Assimilation, 104606, v. 26).
14 Bloss, W. J.; Lee, J. D.; Johnson, G. P.; Sommariva, R.; Heard, D. E.; Saiz-Lopez, A.; Plane, J.
15 M. C.; McFiggans, G.; Coe, H.; Flynn, M.; Williams, P.; Rickard, A. R.; Fleming, Z. L.
16 (2005) Impact of halogen monoxide chemistry upon boundary layer OH and HO2
17 concentrations at a coastal site. Geophys. Res. Lett. 32: 10.1029/2004GL022084.
18 Bobrowski, N.; Honninger, G.; Galle, B.; Platt, U. (2003) Detection of bromine monoxide in a
19 volcanic plume. Nature (London, U.K.) 423: 273-276.
20 Boersma, K. F.; Eskes, H. J.; Brinksma, E. J. (2004) Error analysis for tropospheric NO2 retrieval
21 from space. J. Geophys. Res. [Atmos.] 109(D04311): 10.1029/2003JD003962.
22 Bovensmann, H.; Burrows, J. P.; Buchwitz, M.; Frerick, J.; Noel, S.; Rozanov, V. V. (1999)
23 SCIAMACHY: mission objectives and mesurement modes. J. Atmos. Sci. 56: 127-150.
24 Bradshaw, J.; Davis, D.; Crawford, J.; Chen, G.; Shelter, R.; Muller, M.; Gregory, G.; Sachse,
25 G.; Blake, D.; Heikes, B.; Singh, H.; Mastromarino, J.; Sandholm, S. (1999)
26 Photofragmentation two-photon laser-induced fluorescence detection of NO2 and NO:
27 comparison of measurements with model results based on airborne observations during
28 PEM-Tropics A. Geophys. Res. Lett. 26: 471-474.
29 Brauer, M.; Koutrakis, P.; Keeler, G. J.; Spengler, J. D. (1991) Indoor and outdoor
30 concentrations of inorganic acidic aerosols and gases. J. Air Waste Manage. Assoc. 41:
31 171-181.
32 Brook, J. R.; Dann, T. F. (1999) Contribution of nitrate and carbonaceous species to PM2.5
33 observed in Canadian cities. J. Air Waste Manage. Assoc. 49: 193-199.
34 Brook, J. R.; Sirois, A.; Clarke, J. F. (1996) Comparison of dry deposition velocities for SO2,
35 HNO3, and SO42" estimated with two inferential models. Water Air Soil Pollut. 87: 205-
36 218.
37 Broske, R.; Kleffmann, J.; Wiesen, P. (2003) Heterogeneous conversion of NO2 on secondary
38 organic aerosol surfaces: a possible source of nitrous acid (HONO) in the atmosphere?
39 Atmos. Chem. Phys. 3:469-474.
40
March 2008 AX2-130 DRAFT-DO NOT QUOTE OR CITE
-------
3 Brown, S. S.; Neuman, J. A.; Ryerson, T. B.; Trainer, M.; Dube, W. P.; Holloway, J. S.;
4 Warneke, C.; De Gouw, J. A.; Donnelly, S. G.; Atlas, E.; Matthew, B.; Middlebrook, A.
5 M.; Peltier, R.; Weber, R. J.; Stohl, A.; Meagher, J. F.; Fehsenfeld, F. C.; Ravishankara,
6 A. R. (2006a) Nocturnal odd-oxygen budget and its implications for ozone loss in the
7 lower troposphere. Geophys. Res. Lett. 33(L08801): 10.1029/2006GL025900.
8 Brown, S. S.; Ryerson, T. B.; Wollny, A. G.; Brock, C. A.; Peltier, R.; Sullivan, A. P.; Weber, R.
9 J.; Dube, W. P.; Trainer, M.; Meagher, J. F.; Fehsenfeld, F. C.; Ravishankara, A. R.
10 (2006b) Variability in nocturnal nitrogen oxide processing and its role in regional air
11 quality. Science (Washington, DC, U.S.) 311: 67-70.
12 Briihl, C.; Poschl, U.; Crutzen, P. J.; Steil, B. (2000) Acetone and PAN in the upper troposphere:
13 impact on ozone production from aircraft emissions. Atmos. Environ. 34: 3931-3938.
14 Buhr, M. P.; Trainer, M.; Parrish, D. D.; Sievers, R. E.; Fehsenfeld, F. C. (1992) Assessment of
15 pollutant emission inventories by principal component analysis of ambient air
16 measurements. Geophys. Res. Lett. 19: 1009-1012.
17 Burrows, J. P.; Weber, M.; Buchwitz, M.; Rozanov, V.; Ladstatter-WeiBenmayer, A.; Richter,
18 A.; DeBeek, R.; Hoogen, R.; Bramstedt, K.; Eichmann, K.-U.; Eisinger, M. (1999) The
19 Global Ozone Monitoring Experiment (GOME): mission concept and first scientific
20 results. J. Atmos. Sci. 56: 151-175.
21 Byun, D. W.; Ching, J. K. S., eds. (1999) Science algorithms of the EPA models-3 community
22 multiscale air quality model (CMAQ) modeling system. Washington, DC: U.S.
23 Environmental Protection Agency, Office of Research and Development; EPA/600/R-
24 99/030. Available: http://www.epa.gov/asmdnerl/CMAQ/CMAQscienceDoc.html [4
25 October, 2006].
26 Byun, D. W.; Schere, K. L. (2006) Review of the governing equations, computational
27 algorithms, and other components of the models-3 community multiscale air quality
28 (CMAQ) modeling system. Appl. Mech. Rev. 59: 51-77.
29 Calvert, J. G.; Su, F.; Bottenheim, J. W.; Strausz, O. P. (1978) Mechanism of the homogeneous
30 oxidation of sulfur dioxide in the troposphere. In: Sulfur in the atmosphere: proceedings
31 of the international symposium; September 1977; Dubrovnik, Yugoslavia. Atmos.
32 Environ. 12: 197-226.
33 Calvert, J. G.; Yarwood, G.; Dunker, A. M. (1994) An evaluation of the mechanism of nitrous
34 acid formation in the urban atmosphere. Res. Chem. Intermed. 20: 463-502.
35 Camp, D. C.; Stevens, R. K.; Cobourn, W. G.; Husar, R. B.; Collins, J. F.; Huntzicker, J. J.;
36 Husar, J. D.; Jaklevic, J. M.; McKenzie, R. L.; Tanner, R. L.; Tesch, J. W. (1982)
37 Intercomparison of concentration results from fine particle sulfur monitors. Atmos.
38 Environ. 16: 911-916.
39 Cardelino, C. A.; Chameides, W. L. (2000) The application of data from photochemical
40 assessment monitoring stations to the observation-based model. Atmos. Environ. 34:
41 2325-2332.
March 2008 AX2-131 DRAFT-DO NOT QUOTE OR CITE
-------
1 Carpenter, L. J. (2003) Iodine in the marine boundary layer. Chem. Rev. (Washington, DC, U.S.)
2 103:4953-4962.
3 Carpenter, L. J.; Sturges, W. T.; Penkett, S. A.; Liss, P. S.; Alicke, B.; Hebestreit, K.; Platt, U.
4 (1999) Short-lived alkyl iodides and bromides at Mace Head, Ireland: links to biogenic
5 sources and halogen oxide production. J. Geophys. Res. [Atmos.] 104: 1679-1689.
6 Carroll, M. A.; Bertman, S. B.; Shepson, P. B. (2001) Overview of the program for research on
7 oxidants, photochemistry, emissions and transport (PROPHET) summer 1998
8 measurements intensive. J. Geophys. Res. [Atmos.] 106: 24,275-24,288.
9 Carter, W. P. L. (1990) A detailed mechanism for the gas-phase atmospheric reactions of organic
10 compounds. Atmos. Environ. Part A 24: 481-518.
11 Carter, W. P. L. (1995) Computer modeling of environmental chamber studies of maximum
12 incremental reactivities of volatile organic compounds. Atmos. Environ. 29: 2513.
13 Castells, P.; Santos, F. J.; Gaiceran, M. T. (2003) Development of a sequential supercritical fluid
14 extraction method for the analysis of nitrated an oxygenated derivatives of polycyclic
15 aromatic hydrocarbons in urban aerosols. J. Chromatogr. A 1010: 141-151.
16 Castro, M. S.; Driscoll, C. T. (2002) Atmospheric nitrogen deposition has caused nitrogen
17 enrichment and eutrophication of lakes in the northern hemisphere. Environ. Sci.
18 Technol. 36: 3242-3249.
19 Cecinato, A. (2003) Nitrated polynuclear aromatic hydrocarbons in ambient air in Italy. A brief
20 overview. J. Separation Sci. 26: 402-408.
21 Chameides, W. L. (1984) The photochemistry of a remote marine stratiform cloud. J. Geophys.
22 Res. [Atmos.] 89: 4739-4755.
23 Chameides, W. L.; Stelson, A. W. (1992) Aqueous-phase chemical processes in deliquescent
24 sea-salt aerosols: a mechanism that couples the atmospheric cycles of S and sea salt. J.
25 Geophys. Res. [Atmos.] 97: 20,565-20,580.
26 Chang, J. S.; Brost, R. A.; Isaken, I. S. A.; Madronich, S.; Middleton, P.; Stockwell, W. R.;
27 Walcek, C. J. (1987) A three-dimensional Eulerian acid deposition model: physical
28 concepts and formulation. J. Geophys. Res. [Atmos.] 92: 14,681-14,700.
29 Chang, M. E.; Hartley, D. E.; Cardelino, C.; Chang, W.-L. (1996) Inverse modeling of biogenic
30 emissions. Geophys. Res. Lett. 23: 3007-3010.
31 Chang, M. E.; Hartley, D. E.; Cardelino, C.; Hass-Laursen, D.; Chang, W. L. (1997) On using
32 inverse methods for resolving emissions with large spatial inhomogeneities. J. Geophys.
33 Res. [Atmos.] 102: 16,023-16,036.
34 Chang, M. C.; Sioutas, C.; Kim, S.; Gong, H., Jr.; Linn, W. S. (2000) Reduction of nitrate losses
35 from filter and impactor samplers by means of concentration enrichment. Atmos.
36 Environ. 34: 85-98.
37 Chatfield, R. B.; Crutzen, P. J. (1984) Sulfur dioxide in remote oceanic air: cloud transport of
38 reactive precursors. J. Geophys. Res. [Atmos.] 89: 7111-7132.
39 Chock, D. P.; Winkler, S. L. (1994) A comparison of advection algorithms coupled with
40 chemistry. Atmos. Environ. 28: 2659-2675.
March 2008 AX2-132 DRAFT-DO NOT QUOTE OR CITE
-------
1 Choi, W.; Leu, M.-T. (1998) Nitric acid uptake and decomposition on black carbon (soot)
2 surfaces: its implications for the upper troposphere and lower stratosphere. J. Phys.
3 Chem. A 102: 7618-7630.
6 Chow, J. C.; Watson, J. G. (1999) Ion chromatography in elemental analysis of airborne
7 particles. In: Landsberger, S.; Creatchman, M., eds. Elemental analysis of airborne
8 particles. Amsterdam, The Netherlands: Gordon and Breach Science Publishers; pp. 97-
9 137. (Vo-Dinh, T., ed. Advances in environmental industrial and process control
10 technologies: v. 1).
11 Chow, J. C.; Zielinska, B.; Watson, J. G; Fujita, E. M.; Richards, H. W.; Neff, W. D.; Dietrich,
12 D.; Hering, S. V. (1998) Northern Front Range Air Quality Study. Volume A: Ambient
13 measurements. Fort Collins, CO: Colorado State University, Cooperative Institute for
14 Research in the Atmosphere. Available: http://www.nfraqs.colostate.edu/dri.html [4
15 February 2002].
16 Ciganek, M.; Neca, J.; Adamec, V.; Janosek, J.; Machala, M. (2004) A combined chemical and
17 bioassay analysis of traffic-emitted polycyclic aromatic hydrocarbons. Sci. Total Environ.
18 334-334(spec. issue): 141-148.
19 Civerolo, K. L.; Dickerson, R. R. (1998) Nitric oxide soil emissions from tilled and untilled
20 cornfields. Agric. For. Meteorol. 90: 307-311.
21 deary, P. A.; Wooldridge, P. J.; Cohen, R. C. (2002) Laser-induced fluorescence detection of
22 atmospheric NO2 with a commercial diode laser and a supersonic expansion. Appl.
23 Optics 41: 6950-6956.
24 Clegg, N. A.; Toumi, R. (1998) Non-sea-salt-sulphate formation in sea-salt aerosol. J. Geophys.
25 Res. [Atmos.] 103: 31,095-31,31,102.
26 Clemitshaw, K. C. (2004) A review of instrumentation and measurement techniques for ground-
27 based and airborne field studies of gas-phase tropospheric chemistry. Crit. Rev. Environ.
28 Sci. Technol. 34: 1-108.
29 Cobourn, W. G.; Husar, R. B. (1982) Diurnal and seasonal patterns of particulate sulfur and
30 sulfuric acid in St. Louis, July 1977-June 1978. Atmos. Environ. 16: 1441-1450.
31 Cobourn, W. G.; Husar, R. B.; Husar, J. D. (1978) Continuous in situ monitoring of ambient
32 particulate sulfur using flame photometry and thermal analysis. In: Husar, R. B.; Lodge,
33 J. P., Jr.; Moore, D. J., eds. Sulfur in the atmosphere: proceedings of the international
34 symposium; September 1977; Dubrovnik, Yugoslavia. Atmos. Environ. 12: 89-98.
35 Code of Federal Regulations. (2006) Lower detectable limit—(1) technical definition. C. F. R.
36 40: §53.23c.
37 Cohen, R. C. (1999) Laser-induced fluorescence detection of atmospheric NO2 at parts per
38 trillion mixing ratios: implications for nitrogen oxide photochemistry in the stratosphere
39 and troposphere [abstract]. Abstr. Pap. Am. Chem. Soc. (Phys.) 218(Pt. 2, Aug. 22): 262.
March 2008 AX2-133 DRAFT-DO NOT QUOTE OR CITE
-------
1 Cooper, O. R.; Stohl, A.; Trainer, M.; Thompson, A. M.; Witte, J. C.; Oltmans, S. J.; Morris, G.;
2 Pickering, K. E.; Crawford, J. H.; Chen, G.; Cohen, R. C.; Bertram, T. H.; Wooldridge,
3 P.; Perring, A.; Brune, W.; Merrill, J.; Moody, J. L.; Tarasick, D.; Nedelec, P.; Forbes,
4 G.; Newchurch, M.; Schmidlin, F.; Johnson, B. J.; Turquety, S.; Baughcum, S. L.; Ren,
5 X.; Fehsenfeld, F. C.; Meagher, J. F.; Spichtinger, N.; Brown, C. C.; McKeen, S. A.;
6 McDermid, IS.; Leblanc, T. (2006) Large upper tropospheric ozone enhancements
7 above mid-latitude North America during summer: in situ evidence from the IONS and
8 MOZAIC ozone measurement network. J. Geophys. Res. [Atmos.]
9 111(D24S05): 10.1029/2006JD007306.
10 Crawford, J.; Davis, D.; Olson, J.; Chen, G.; Liu, S.; Fuelberg, H.; Hannan, J.; Kondo, Y.;
11 Anderson, B.; Gregory, G.; Sachse, G.; Talbot, R.; Viggiano, A.; Heikes, B.; Snow, J.;
12 Singh, H.; Blake, D. (2000) Evolution and chemical consequences of lightning-produced
13 NOX observed in the North Atlantic upper troposphere. J. Geophys. Res. [Atmos.] 105:
14 19,795-19,809.
15 Crosley, D. R. (1996) NOy blue ribbon panel. J. Geophys. Res. [Atmos.] 101: 2049-2052.
16 Crutzen, P. J. (1976) Upper limits on atmospheric ozone reductions following increased
17 application of fixed nitrogen to the soil. Geophys. Res. Lett. 3: 169-172.
18 Crutzen, P. J.; Gidel, L. T. (1983) A two-dimensional photochemical model of the atmosphere. 2.
19 The tropospheric budgets of the anthropogenic chlorocarbons, CO, CFL;, CHaCl and the
20 effect of various NOX sources on tropospheric ozone. J. Geophys. Res. [Atmos.] 88:
21 6641-6661.
22 Dagnall, R. M.; Thompson, K. C.; West, T. S. (1967) Molecular-emission in cool flames. Part I.
23 the behaviour of sulphur species in a hydrogen-nitrogen diffusion flame and in a shielded
24 air-hydrogen flame. Analyst 92: 506-512.
25 Daum, P. H.; Kleinman, L. L; Hills, A. J.; Lazrus, A. L.; Leslie, A. C. D.; Busness, K.; Boatman,
26 J. (1990) Measurement and interpretation of concentrations of H2O2 and related species
27 in the upper midwest during summer. J. Geophys. Res. [Atmos.] 95: 9857-9871.
28 Daum, P. H.; Kleinman, L. L; Newman, L.; Luke, W. T.; Weinstein-Lloyd, J.; Berkowitz, C. M.;
29 Busness, K. M. (1996) Chemical and physical properties of plumes of anthropogenic
30 pollutants transported over the North Atlantic during the North Atlantic Regional
31 Experiment. J. Geophys. Res. [Atmos.] 101: 29,029-29,042.
32 Davidson, E. A.; Kingerlee, W. (1997) A global inventory of nitric oxide emissions from soils.
33 Nutr. Cycling Agroecosyst. 48: 37-50.
34 Day, D. A.; Wooldridge, P. J.; Dillon, M. B.; Thornton, J. A.; Cohen, R. C. (2002) A thermal
35 dissociation laser-induced fluorescence instrument for in situ detection of NO2, peroxy
36 nitrates, alkyl nitrates, andHNOs. J. Geophys. Res. [Atmos.] 107(D6):
37 10.1029/2001JD000779.
38 Day, D. A.; Dillon, M. B.; Woolridge, P. J.; Thornton, J. A.; Rosen, R. S.; Wood, E. C.; Cohen,
39 R. C. (2003) On alkyl nitrates, O3, and the "missing NOy". J. Geophys. Res. [Atmos.]
40 108: 10.1029/2003JD003685.
41 DeCaria, A. J.; Pickering, K. E.; Stenchikov, G. L.; Scala, J. R.; Stith, J. L.; Dye, J. E.; Ridley, B.
42 A.; Laroche, P. (2000) A cloud-scale model study of lightning-generated NOX in an
March 2008 AX2-134 DRAFT-DO NOT QUOTE OR CITE
-------
1 individual thunderstorm during STERAO-A. J. Geophys. Res. [Atmos.] 105: 11,601-
2 11,616.
3 DeCaria, A. J.; Pickering, K. E.; Stenchikov, G. L.; Ott, L. E. (2005) Lightning-generated NOX
4 and its impact on tropospheric ozone production: a three-dimensional modeling study of a
5 stratosphere-troposphere experiment: radiation, aerosols and ozone (STERAO-A)
6 thunderstorm. J. Geophys. Res. [Atmos.] 110(D14303): 10.1029/2004JD005556.
9 De Laat, A. T. J.; Zachariasse, M.; Roelofs, G. J.; Van Velthoven, P.; Dickerson, R. R.; Rhoads,
10 K. P.; Oltmans, S. J.; Lelieveld, J. (1999) Tropospheric O3 distribution over the Indian
11 Ocean during spring 1995 evaluated with a chemistry-climate model. J. Geophys. Res.
12 [Atmos.] 104: 13,881-13,893.
13 Delmas, R. (1982) On the emissions of carbon, nitrogen and sulfur to the atmosphere during
14 bushfires in intertropical savannah zones. Geophys. Res. Lett. 9: 761-764.
15 Dentener, F. J.; Crutzen, P. J. (1993) Reaction of ^Os on tropospheric aerosols: impact on the
16 global distributions of NOX, O3, and OH. J. Geophys. Res. [Atmos.] 98: 7149-7163.
17 Dentener, F.; Stevenson, D.; Cofala, J.; Mechler, R.; Amann, M.; Bergamaschi, P.; Raes, F.;
18 Derwent, R. (2005) The impact of air pollutant and methane emission controls on
19 tropospheric ozone and radiative forcing: CTM calculations for the period 1990-2030.
20 Atmos. Chem. Phys. 5: 1731-1755.
21 Dentener, F.; Stevenson, D.; Ellingsen, K.; VanNoije, T.; Schultz, M.; Amann, M.; Atherton, C.;
22 Bell, N.; Bergmann, D.; Bey, L; Bouwman, L.; Butler, T.; Cofala, J.; Collins, B.; Drevet,
23 J.; Doherty, R.; Eickhout, B.; Eskes, H.; Fiore, A.; Gauss, M.; Hauglustaine, D.;
24 Horowitz, L.; Isaksen, I. S. A.; Josse, B.; Lawrence, M.; Krol, M.; Lamarque, J. F.;
25 Montanaro, V.; Muller, J. F.; Peuch, V. H.; Pitari, G.; Pyle, J.; Rast, S.; Rodriguez, J.;
26 Sanderson, M.; Savage, N. H.; Shindell, D.; Strahan, S.; Szopa, S.; Sudo, K.; Van
27 Dingenen, R.; Wild, O.; Zeng, G. (2006a) The global atmospheric environment for the
28 next generation. Environ. Sci. Technol. 40: 3586-3594.
29 Dentener, F.; Drevet, J.; Lamarque, J. F.; Bey, L; Eickhout, B.; Fiore, A. M.; Hauglustaine, D.;
30 Horowitz, L. W.; Krol, M.; Kulshrestha, U. C.; Lawrence, M.; Galy-Lacaux, C.; Rast, S.;
31 Shindell, D.; Stevenson, D.; Van Noije, T.; Atherton, C.; Bell, N.; Bergman, D.; Butler,
32 T.; Cofala, J.; Collins, B.; Doherty, R.; Ellingsen, K.; Galloway, J.; Gauss, M.;
33 Montanaro, V.; Muller, J. F.; Pitari, G.; Rodriguez, J.; Sanderson, M.; Strahan, S.;
34 Schultz, M.; Sudo, K.; Szopa, S.; Wild, O. (2006b) Nitrogen and sulfur deposition on
35 regional and global scales: a multi-model evaluation. Global Biogeochem. Cycles:
36 20(GB4003): 10.1029/2005GB002672.
37 Derwent, R. G.; Collins, W. J.; Johnson, C. E.; Stevenson, D. S. (2001) Transient behaviour of
38 tropospheric ozone precursors in a global 3-D CTM and their indirect greenhouse effects.
39 Climatic Change 49: 463-487.
40 Dibb, J. E.; Arsenault, M.; Peterson, M. C. (2002) Fast nitrogen oxide photochemistry in
41 Summit, Greenland snow. Atmos. Environ. 36: 2501-2511.
March 2008 AX2-135 DRAFT-DO NOT QUOTE OR CITE
-------
1 Dickerson, R. R.; Huffman, G. J.; Luke, W. T.; Nunnermacker, L. J.; Pickering, K. E.; Leslie, A.
2 C. D.; Lindsey, C. G.; Slinn, W. G. N.; Kelly, T. J.; Daum, P. H.; Delany, A. C.;
3 Greenberg, J. P.; Zimmerman, P. R.; Boatman, J. F.; Ray, J. D.; Stedman, D. H. (1987)
4 Thunderstorms: an important mechanism in the transport of air pollutants. Science
5 (Washington, DC) 23 5: 460-465.
6 Dickerson, R. R.; Doddridge, B. G.; Kelley, P.; Rhoads, K. P. (1995) Large-scale pollution of the
7 atmosphere over the remote Atlantic Ocean: evidence from Bermuda. J. Geophys. Res.
8 [Atmos.] 100: 8945-8952.
9 Dickerson, R. R.; Rhoads, K. P.; Carsey, T. P.; Oltmans, S. J.; Burrows, J. P.; Crutzen, P. J.
10 (1999) Ozone in the remote marine boundary layer: a possible role for halogens. J.
11 Geophys. Res. [Atmos.] 104: 21,385-21,395.
12 Dimashki, M.; Harrad, S.; Harrison, R. M. (2000) Concentrations and phase distribution of nitro-
13 PAH in the Queensway road tunnel in Birmingham, United Kingdom. Polycyclic
14 Aromat. Cmpd. 20: 205-223.
15 Domine, F.; Shepson, P. B. (2002) Air-snow interactions and atmospheric chemistry. Science
16 (Washington, DC, U.S.) 297: 1506-1510.
17 Dommen, J.; Prevot, A. S. H.; Hering, A. M.; Staffelbach, T.; Kok, G. L.; Schillawski, R. D.
18 (1999) Photochemical production and aging of an urban air mass. J. Geophys. Res.
19 [Atmos.] 104: 5493-5506.
20 Doskey, P. V.; Kotamarthi, V. R.; Fukui, Y.; Cook, D. R.; Breitbeil, F. W., Ill; Wesely, M. L.
21 (2004) Air-surface exchange of peroxyacetyl nitrate at a grassland site. J. Geophys. Res.
22 [Atmos.] 109(D10310): 10.1029/2004JD004533.
23 Driscoll, C. T.; Whitall, D.; Aber, J.; Boyer, E.; Castro, M.; Cronan, C.; Goodale, C. L.;
24 Groffman, P.; Hopkinson, C.; lambert, K.; Lawrence, G.; Ollinger, S. (2003) Nitrogen
25 pollution in the northeastern United States: sources, effects, and management options.
26 BioScience 53: 357-374.
27 Dudhia, J. (1993) A nonhydrostatic version of the Penn State-NCAR mesoscale model:
28 validation tests and simulation of an Atlantic cyclone and cold front. Mon. Weather Rev.
29 121: 1493-1513.
30 Durant, J. L.; Busby, W. F., Jr.; Lafleur, A. L.; Penman, B. W.; Crespi, C. L. (1996) Human cell
31 mutagenicity of oxygenated, nitrated and unsubstituted polycyclic aromatic hydrocarbons
32 associated with urban aerosols. Mutat. Res. 371: 123-157.
33 Durham, J. L.; Wilson, W. E.; Bailey, E. B. (1978) Application of an SO2-denuder for
34 continuous measurement of sulfur in submicrometric aerosols. Atmos. Environ. 12: 883-
35 886.
36 Eder, B.; Yu, S. (2005) A performance evaluation of the 2004 release of Models-3 CMAQ.
37 Atmos. Environ. 40: 4811-4824.
38 Eisele, F. L.; Mauldin, L.; Cantrell, C.; Zondlo, M.; Apel, E.; Fried, A.; Walega, J.; Shelter, R.;
39 Lefer, B.; Flocke, F.; Weinheimer, A.; Avery, M.; Vay, S.; Sachse, G.; Podolske, J.;
40 Diskin, G.; Barrick, J. D.; Singh, H. B.; Brune, W.; Harder, H.; Martinez, M.; Bandy, A.;
41 Thornton, D.; Heikes, B.; Kondo, Y.; Riemer, D.; Sandholm, S.; Tan, D.; Talbot, R.;
March 2008 AX2-136 DRAFT-DO NOT QUOTE OR CITE
-------
1 Dibb, J. (2003) Summary of measurement intercomparisons during TRACE-P. J.
2 Geophys. Res. [Atmos.] 108(D20): 10.1029/2002JD003167.
3 Emmons, L. K.; Hauglustaine, D. A.; Miiller, J.-F.; Carroll, M. A.; Brasseur, G. P.; Brunner, D.;
4 Staehelin, J.; Thouret, V.; Marenco, A. (2000) Data composites of airborne observations
5 of tropospheric ozone and its precursors. J. Geophys. Res. [Atmos.] 105: 20,497-20,538.
6 Erickson, D. J., Ill; Seuzaret, C.; Keene, W. C.; Gong, S. L. (1999) A general circulation model
7 based calculation of HC1 and CINO2 production from sea-salt dechlorination: reactive
8 chlorine emissions inventory. J. Geophys. Res. [Atmos.] 104: 8347-8372.
10
11 Esteve, W.; Budzinski, H.; Villenave, E. (2006) Relative rate constants for the heterogeneous
12 reactions of NC>2 and OH radicals with polycyclic aromatic hydrocarbons adsorbed on
13 carbonaceous particles. Part 2: PAHs adsorbed on diesel particulate exhaust SRM 1650a.
14 Atmos. Environ. 40: 201-211.
15 Fan, Z.; Chen, D.; Birla, P.; Kamens, R. M. (1995) Modeling of nitro-polycyclic aromatic
16 hydrocarbon formation and decay in the atmosphere. Atmos. Environ. 29: 1171-1181.
17 Fan, Z.; Kamens, R. M.; Hu, J.; Zhang, J.; McDow, S. (1996) Photostability of nitro-poly cyclic
18 aromatic hydrocarbons on combustion soot particles in sunlight. Environ. Sci. Technol.
19 30: 1358-1364.
20 Faraji, M.; Kimura, Y.; Allen, D. (2005) Comparison of the carbon bond and SAPRC
21 photochemical mechanisms under conditions relevent to southeast Texas. Houston, TX:
22 Texas Environmental Research Consortium (TERC); project report H12.8HRB.
23 Available: http://files.harc.edu/Projects/AirQuality/Projects/H012.2004/8HRB/H12-
24 8HRBMechanismReport.pdf [4 October, 2007].
25 Farmer, D. K.; Wooldridge, P. J.; Cohen, R. C. (2006) Application of thermal-dissociation laser
26 induced fluorescence (TD-LIF) to measurement of HNO3, Ealkyl nitrates, Eperoxy
27 nitrates, and NO2 fluxes using eddy covariance. Atmos. Chem. Phys. 6: 3471-3486.
28 Farwell, S. O.; Rasmussen, R. A. (1976) Limitations of the FPD and ECD in atmospheric
29 analysis: a review. J. Chromatogr. Sci. 14: 224-234.
30 Fast, J. D.; Zaveri, R. A.; Bian, R. X.; Chapman, E. G.; Easter, R. C. (2002) Effect of regional-
31 scale transport on oxidants in the vicinity of Philadelphia during the 1999 NE-OPS Field
32 Campaign. J. Geophys. Res. [Atmos.] 107(D16): 10.1029/2001JD000980.
33 Fehsenfeld, F. C.; Dickerson, R. R.; Hiibler, G.; Luke, W. T.; Nunnermacker, L. J.; Williams, E.
34 J.; Roberts, J. M.; Calvert, J. G.; Curran, C. M.; Delany, A. C.; Eubank, C. S.; Fahey, D.
35 W.; Fried, A.; Gandrud, B. W.; Langford, A. O.; Murphy, P. C.; Norton, R. B.; Pickering,
36 K. E.; Ridley, B. A. (1987) A ground-based intercomparison of NO, NOX, and NOy
37 measurement techniques. J. Geophys. Res. [Atmos.] 92: 14,710-14,722.
38 Fehsenfeld, F. C.; Trainer, M.; Parrish, D. D.; Volz-Thomas, A.; Penkett, S. (1996) North
39 Atlantic Regional Experiment (NARE) 1993 summer intensive: foreword. J. Geophys.
40 Res. [Atmos.] 101: 28,869-28,875.
March 2008 AX2-137 DRAFT-DO NOT QUOTE OR CITE
-------
1 Fehsenfeld, F. C.; Huey, L. G.; Sueper, D. T.; Norton, R. B.; Williams, E. I; Eisele, F. L.;
2 Mauldin, R. L., Ill; Tanner, D. J. (1998) Ground-based intercomparison of nitric acid
3 measurement techniques. J. Geophys. Res. [Atmos.] 103: 3343-3353.
4 Feilberg, A.; Nielsen, T. (2001) Photodegradation of nitro-PAHs in viscous organic media used
5 as models of organic aerosols. Environ. Sci. Technol. 35: 108-113.
6 Feilberg, A.; Kamens, R. M.; Strommen, M. R.; Nielsen, T. (1999) Modeling the formation,
7 decay, and partitioning of semivolatile nitro-polycyclic aromatic hydrocarbons
8 (nitronaphthalenes) in the atmosphere. Atmos. Environ. 33: 1231-1243.
9 Feilberg, A.; Poulsen, M. W. B.; Nielsen, T.; Skov, B. (2001) Occurrence and sources of
10 particulate nitro-polycyclic aromatic hydrocarbons in ambient air in Denmark. Atmos.
11 Environ. 35: 353-366.
12 Finlayson-Pitts, B. J.; Pitts, J. N., Jr. (2000) Chemistry of the upper and lower atmosphere:
13 theory, experiments and applications. San Diego, CA: Academic Press.
14 Finlayson-Pitts, B. J.; Ezell, M. J.; Pitts, J. N., Jr. (1989) Formation of chemically active chlorine
15 compounds by reactions of atmospheric NaCl particles with gaseous ^Os and C1ONO2.
16 Nature (London) 337: 241-244.
17 Finlayson-Pitts, B. J.; Wingen, L. M.; Sumner, A. L.; Syomin, D.; Ramazan, K. A. (2003) The
18 heterogeneous hydrolysis of NC>2 in laboratory system and in outdoor and indoor
19 atmospheres: an integrated mechanism. Phys. Chem. Chem. Phys. 5: 223-242.
20 Finley, B. D.; Saltzman, E. S. (2006) Measurement of C12 in coastal urban air. Geophys. Res.
21 Lett. 33(L11809): 10.1029/2006GL025799.
22 Fiore, A. M.; Jacob, D. J.; Field, B. D.; Streets, D. G.; Fernandes, S. D.; Jang, C. (2002) Linking
23 ozone pollution and climate change: the case for controlling methane. Geophys. Res.
24 Lett. 29(19): 10.1029/2002GLO 15601.
25 Fiore, A.; Jacob, D. J.; Liu, H.; Yantosca, R. M.; Fairlie, T. D.; Li, Q. (2003) Variability in
26 surface ozone background over the United States: implications for air quality policy. J.
27 Geophys. Res. [Atmos.] 108(D24): 10.1029/2003JD003855.
28 Fischer, E.; Pszenny, A.; Keene, W.; Maben, J.; Smith, A.; Stohl, A.; Talbot, R. (2006) Nitric
29 acid phase partitioning and cycling in the New England coastal atmosphere. J. Geophys.
30 Res. [Atmos.] 111(D23S09): 10.1029/2006JD007328.
31 Foster, K. L.; Plastridge, R. A.; Bottenheim, J. W.; Shepson, P. B.; Finlayson-Pitts, B. J.; Spicer,
32 C. W. (2001) The role of Br2 and BrCl in surface ozone destruction at polar sunrise.
33 Science (Washington, DC, U.S.) 291: 471-474.
34 Fox-Rabinovitz, M. S.; Takacs, L. L.; Govindaraju, R. C. (2002) A variable-resolution stretched-
35 grid general circulation model and data assimilation system with multiple areas of
36 interest: studying the anomalous regional climate events of 1998. J. Geophys. Res.
37 [Atmos.] 107(D24): 10.1029/2002 JD002177.
38 Frost, G. J.; McKeen, S. A.; Trainer, M.; Ryerson, T. B.; Neuman, J. A.; Roberts, J. M.;
39 Swanson, A.; Holloway, J. S.; Sueper, D. T.; Fortin, T.; Parrish, D. D.; Fehsenfeld, F. C.;
40 Flocke, F.; Peckham, S. E.; Grell, G. A.; Kowal, D.; Cartwright, J.; Auerbach, N.;
41 Habermann, T. (2006) Effects of changing power plant NOX emissions on ozone in the
March 2008 AX2-138 DRAFT-DO NOT QUOTE OR CITE
-------
1 eastern United States: proof of concept. J. Geophys. Res. [Atmos.] 111(D12306):
2 10.1029/2005 JD0063 54.
3 Fuentes, M.; Raftery, A. E. (2005) Model evaluation and spatial interpolation by Bayesian
4 combination of observations with outputs from numerical models. Biometrics 61: 36-45.
5 Furutani, H.; Akimoto, H. (2002) Development and characterization of a fast measurement
6 system for gas-phase nitric acid with a chemical ionization mass spectrometer in the
7 marine boundary layer. J. Geophys. Res. [Atmos.] 107(D2): 10.1029/2000JD000269.
8 Gaffney, J. S.; Bornick, R. M.; Chen, Y.-H.; Marley, N. A. (1998) Capillary gas
9 chromatographic analysis of nitrogen dioxide and pans with luminol chemiluminescent
10 detection. Atmos. Environ. 32: 1445-1454.
11 Galbally, I. E.; Bentley, S. T.; Meyer, C. P. (2000) Mid-latitude marine boundary-layer ozone
12 destruction at visible sunrise observed at Cape Grim, Tasmania, 41°S. Geophys. Res.
13 Lett. 27: 3841-3844.
14 Galloway, J. N.; Aber, J. D.; Erisman, J. W.; Seitzinger, S. P.; Howarth, R. W.; Cowling, E. B.;
15 Cosby, B. J. (2003) The nitrogen cascade. BioScience 53: 341-356.
16 Gao, W.; Wesely, M. L.; Lee, I. Y. (1991) A numerical study of the effects of air chemistry on
17 fluxes of NO, NO2, and O3 near the surface. J. Geophys. Res. 96: 18,761-18,769.
18 Gao, W.; Wesely, M. L.; Doskey, P. V. (1993) Numerical modeling of the turbulent diffusion
19 and chemistry of NOX, 63, isoprene, and other reactive trace gases in and above a forest
20 canopy. J. Geophys. Res. [Atmos.] 98: 18339-18353.
21 Garnica, R. M.; Appel, M. F.; Eagan, L.; McKeachie, J. R.; Benter, T. (2000) A REMPI method
22 for the ultrasensitive detection of NO and NO2 using atmospheric pressure laser
23 ionization mass spectrometry. Anal. Chem. 72: 5639-5646.
24 George, C.; Strekowski, R. S.; Kleffmann, J.; Stemmler, K.; Ammann, M. (2005) Photoenhanced
25 uptake of gaseous NO2 on solid organic compounds: a photochemical source of HONO?
26 Faraday Discuss. 130: 195-210.
27 Gerlach, T. M. (2004) Volcanic sources of tropospheric ozone-depleting trace gases. Geochem.
28 Geophys. Geosys. 5: 10.1029/2004GC000747.
29 Gery, M. W.; Whitten, G. Z.; Killus, J. P.; Dodge, M. C. (1989) A photochemical kinetics
30 mechanism for urban and regional scale computer modeling. J. Geophys. Res. [Atmos.]
31 94: 12,925-12,956.
32 Geyer, A.; Platt, U. (2002) Temperature dependence of the NOs loss frequency: A new indicator
33 for the contribution of NO3 to the oxidation of monoterpenes and NOX removal in the
34 atmosphere. J. Geophys. Res. [Atmos.] 107(D20): 10.1029/2001JD001215.
35 Geyer, A.; Alicke, B.; Konrad, S.; Schmitz, T.; Stutz, J.; Platt, U. (2001) Chemistry and
36 oxidation capacity of the nitrate radical in the continental boundary layer near Berlin. J.
37 Geophys. Res. [Atmos.] 106: 8013-8025.
38 Gibson, T. L. (1983) Sources of direct-acting nitroarene mutagens in airborne particulate matter.
39 Mutat. Res. 122: 115-121.
March 2008 AX2-139 DRAFT-DO NOT QUOTE OR CITE
-------
1 Gidel, L. T. (1983) Cumulus cloud transport of transient tracers. J. Geophys. Res. [Atmos.] 88:
2 6587-6599.
3 Goldan, P. D.; Kuster, W. C.; Albritton, D. L.; Fehsenfeld, F. C.; Connell, P. S.; Norton, R. B.;
4 Huebert, B. J. (1983) Calibration and tests of the filter-collection method for measuring
5 clean-air, ambient levels of nitric acid. Atmos. Environ. 17: 1355-1364.
6 Goldan, P. D.; Trainer, M.; Kuster, W. C.; Parrish, D. D.; Carpenter, J.; Roberts, J. M.; Yee, J.
7 E.; Fehsenfeld, F. C. (1995) Measurements of hydrocarbons, oxygenated hydrocarbons,
8 carbon monoxide, and nitrogen oxides in an urban basin in Colorado: implications for
9 emission inventories. J. Geophys. Res. [Atmos.] 100: 22,771-22,783.
10 Goldan, P. D.; Kuster, W. C.; Fehsenfeld, F. C. (1997) Nonmethane hydrocarbon measurements
11 during the tropospheric OH photochemistry experiment. J. Geophys. Res. [Atmos.] 102:
12 6315-6324.
13
14
15 Goldan, P. D.; Parrish, D. D.; Kuster, W. C.; Trainer, M.; McKeen, S. A.; Holloway, J.; Jobson,
16 B. T.; Sueper, D. T.; Fehsenfeld, F. C. (2000) Airborne measurements of isoprene, CO,
17 and anthropogenic hydrocarbons and their implications. J. Geophys. Res. [Atmos.] 105:
18 9091-9105.
19 Golden, D. M.; Smith, G. P. (2000) Reaction of OH + NO2 + M: a new view. J. Phys. Chem. A
20 104: 3991-3997.
21 Goldstein, A. H.; Schade, G. W. (2000) Quantifying biogenic and anthropogenic contributions to
22 acetone mixing ratios in a rural environment. Atmos. Environ. 34: 4997-5006.
23 Gondal, M. A. (1997) Laser photoacoustic spectrometer for remote monitoring of atmospheric
24 pollutants. Appl. Opt. 36: 3195-3201.
25 Gondal, M. A.; Mastromarino, J. (2001) Pulsed laser photoacoustic detection of SO2 near 225.7
26 nm. Appl. Opt. 40: 2010-2016.
27 Greenhut, G. K. (1986) Transport of ozone between boundary layer and cloud layer by cumulus
28 clouds. J. Geophys. Res. [Atmos.] 91: 8613-8622.
29 Greenhut, G. K.; Ching, J. K. S.; Pearson, R., Jr.; Repoff, T. P. (1984) Transport of ozone by
30 turbulence and clouds in an urban boundary layer. J. Geophys. Res. [Atmos.] 89: 4757-
31 4766.
32 Gregory, G. L.; Hoell, J. M., Jr.; Torres, A. L.; Carroll, M. A.; Ridley, B. A.; Rodgers, M. O.;
33 Bradshaw, J.; Sandholm, S.; Davis, D. D. (1990) An intercomparison of airborne nitric
34 oxide measurements: a second opportunity. J. Geophys. Res. [Atmos.] 95: 10,129-10,138.
35 Grell, G. A.; Dudhia, J.; Stauffer, D. R. (1994) Description of the fifth-generation Penn
36 State/NCAR mesoscale model (MM5). Boulder, CO: National Center for Atmospheric
37 Research, Mesoscale and Microscale Meteorology Division; report no. NCAR/TN-
38 398+STR. Available from: NTIS, Springfield, VA; PB95-206348.
March 2008 AX2-140 DRAFT-DO NOT QUOTE OR CITE
-------
1 Grell, G. A.; Emeis, S.; Stockwell, W. R.; Schoenemeyer, T.; Forkel, R.; Michalakes, J.; Knoche,
2 R.; Seidl, W. (2000) Application of a multiscale, coupled MM5/chemistry model to the
3 complex terrain of the VOTALP valley campaign. Atmos. Environ. 34: 1435-1453.
4 Grosjean, D.; Fung, K.; Harrison, J. (1983) Interactions of polycyclic aromatic hydrocarbons
5 with atmospheric pollutants. Environ. Sci. Technol. 17: 673-679.
6 Grosovsky, A. J.; Sasaki, J. C.; Arey, J.; Eastmond, D. A.; Parks, K. K.; Atkinson, R. (1999)
7 Evaluation of the potential health effects of the atmospheric reaction products of
8 polycyclic aromatic hydrocarbons. Cambridge, MA: Health Effects Institute; research
9 report no. 84.
10 Gross, A.; Stockwell, W. R. (2003) Comparison of the EMEP, RADM2 and RACM
11 mechanisms. J. Atmos. Chem. 44: 151-170.
12 Grossenbacher, J. W.; Couch, T.; Shepson, P. B.; Thornberry, T.; Witmer-Rich, M.; Carroll, M.
13 A.; Faloona, I; Tan, D.; Brune, W.; Ostling, K.; Bertman, S. (2001) Measurements of
14 isoprene nitrates above a forest canopy. J. Geophys. Res. [Atmos.] 106: 24,429-24,438.
15 Guenther, F. R.; Dorko, W. D.; Miller, W. R.; Rhoderick, G. C. (1996) The NIST traceable
16 reference material program for gas standards. Washington, DC: U.S. Department of
17 Commerce, National Institute of Standards and Technology; NIST special publication
18 260-126.
19 Guenther, A.; Geron, C.; Pierce, T.; Lamb, B.; Harley, P.; Fall, R. (2000) Natural emissions of
20 non-methane volatile organic compounds, carbon monoxide, and oxides of nitrogen from
21 North America. Atmos. Environ. 34: 2205-2230.
22 Hains, J. C.; Chen, L.-W. A.; Taubman, B. F.; Doddridge, B. G.; Dickerson, R. R. (2007) A side-
23 by-side comparison of filter-based PM2.5 measurements at a suburban site: a closure
24 study. Atmos. Environ. 41: 6167-6184.
25 Hallock-Waters, K. A.; Doddridge, B. G.; Dickerson, R. R.; Spitzer, S.; Ray, J. D. (1999) Carbon
26 monoxide in the U.S. mid-Atlantic troposphere: evidence for a decreasing trend.
27 Geophys. Res. Lett. 26: 2861-2864.
28 Hallquist, M.; Stewart, D. J.; Stephenson, S. K.; Cox, R. A. (2003) Hydrolysis of N205 on sub-
29 micron sulfate aerosols. Phys. Chem. Chem. Phys. 5: 3453-3463.
30 Hameed, S.; Pinto, J. P.; Stewart, R. W. (1979) Sensitivity of the predicted CO-OH-CH4
31 perturbation to tropospheric NOX concentrations. J. Geophys. Res. C: Oceans Atmos. 84:
32 763-768.
33 Hanke, M.; Umann, B.; Uecker, J.; Arnold, F.; Bunz, H. (2003) Atmospheric measurements of
34 gas-phase HNOs and SO2 using chemical ionization mass spectrometry during the
35 MINATROC field campaign 2000 on Monte Cimone. Atmos. Chem. Phys. 3: 417-436.
36 Hansen, K.; Draaijers, G. P. J.; Ivens, W. P. M. F.; Gundersen, P.; van Leeuwen, N. F. M. (1994)
37 Concentration variations in rain and canopy throughfall collected sequentially during
38 individual rain events. Atmos. Environ. 28: 3195-3205.
39 Harder, J. W.; Williams, E. J.; Baumann, K.; Fehsenfeld, F. C. (1997) Ground-based comparison
40 of NO2, H2O, and Os measured by long-path and in situ techniques during the 1993
March 2008 AX2-141 DRAFT-DO NOT QUOTE OR CITE
-------
1 Tropospheric OH Photochemistry Experiment. J. Geophys. Res. [Atmos.] 102: 6227-
2 6243.
3 Hari, P.; Raivonen, M.; Vesala, T.; Munger, J. W.; Pilegaard, K.; Kulmala, M. (2003) Ultraviolet
4 light and leaf emission of NOX. Nature 422: 134.
5 Harris, G. W.; Carter, W. P. L.; Winer, A. M.; Pitts, J. N., Jr.; Platt, U.; Perner, D. (1982)
6 Observations of nitrous acid in the Los Angeles atmosphere and implications for
7 predictions of ozone-precursor relationships. Environ. Sci. Technol. 16: 414-419.
8 Harrison, R. M.; Kitto, A.-M. N. (1994) Evidence for a surface source of atmospheric nitrous
9 acid. Atmos. Environ. 28: 1089-1094.
10 Harrison, R. M.; Peak, J. D.; Collins, G. M. (1996) Tropospheric cycle of nitrous acid. J.
11 Geophys. Res. [Atmos.] 101: 14,429-14,439.
12 He, Y.; Zhou, X.; Hou, J.; Gao, H.; Bertman, S. B. (2006) Importance of dew in controlling the
13 air-surface exchange of HONO in rural forested environments. Geophys. Res. Lett. 33:
14 10.1029/2005GL024348.
15 Hebestreit, K.; Stutz, J.; Rosen, D.; Matveiv, V.; Peleg, M.; Luria, M.; Platt, U. (1999) DOAS
16 measurements of tropospheric bromine oxide in mid-latitudes. Science (Washington, DC)
17 283: 55-57.
18 Hennigan, C. J.; Sandholm, S.; Kim, S.; Stickel, R. E.; Huey, L. G; Weber, R. J. (2006)
19 Influence of Ohio River valley emissions on fine particle sulfate measured from aircraft
20 over large regions of the eastern United States and Canada during INTEX-NA. J.
21 Geophys. Res. [Atmos.] 111(D24S04): 10.1029/2006JD007282.
22 Hering, S.; Cass, G. (1999) The magnitude of bias in the measurement of PM25 arising from
23 volatilization of particulate nitrate from Teflon filters. J. Air Waste Manage. Assoc. 49:
24 725-733.
25 Hering, S. V.; Lawson, D. R.; Allegrini, L; Febo, A.; Perrino, C.; Possanzini, M.; Sickles, J. E.,
26 II; Anlauf, K. G; Wiebe, A.; Appel, B. R.; John, W.; Ondo, J.; Wall, S.; Braman, R. S.;
27 Sutton, R.; Cass, G. R.; Solomon, P. A.; Eatough, D. J.; Eatough, N. L.; Ellis, E. C.;
28 Grosjean, D.; Hicks, B. B.; Womack, J. D.; Horrocks, J.; Knapp, K. T.; Ellestad, T. G.;
29 Paur, R. J.; Mitchell, W. J.; Pleasant, M.; Peake, E.; MacLean, A.; Pierson, W. R.;
30 Brachaczek, W.; Schiff, H. I; Mackay, G. L; Spicer, C. W.; Stedman, D. H.; Winer, A.
31 M.; Biermann, H. W.; Tuazon, E. C. (1988) The nitric acid shootout: field comparison of
32 measurement methods. Atmos. Environ. 22: 1519-1539.
33 Hess, P. G. (2001) Model and measurement analysis of springtime transport and chemistry of the
34 Pacific Basin. J. Geophys. Res. [Atmos.] 106: 12,689-12,717.
35 Hicks, B. B.; Baldocchi, D. D.; Meyers, T. P.; Hosker, R. P., Jr.; Matt, D. R. (1987) A
36 preliminary multiple resistance routine for deriving dry deposition velocities from
37 measured quantities. Water Air Soil Pollut. 36: 311-330.
38 Hirsch, A. I; Munger, J. W.; Jacob, D. J.; Horowitz, L. W.; Goldstein, A. H. (1996) Seasonal
39 variation of the ozone production efficiency per unit NOX at Harvard Forest,
40 Massachusetts. J. Geophys. Res. [Atmos.] 101: 12,659-12,666.
March 2008 AX2-142 DRAFT-DO NOT QUOTE OR CITE
-------
1 Hoell, J. M.; Davis, D. D.; Liu, S. C.; Newell, R. E.; Akimoto, H.; McNeal, R. J.; Bendura, R. J.
2 (1997) The Pacific exploratory mission-west phase B: February-March, 1994. J.
3 Geophys. Res. [Atmos.] 102: 28,223-28,239.
4 Hoell, J. M.; Davis, D. D.; Jacob, D. J.; Rodgers, M. O.; Newell, R. E.; Fuelberg, H. E.; McNeal,
5 R. J.; Raper, J. L.; Bendura, R. J. (1999) Pacific exploratory mission in the tropical
6 Pacific: PEM-Tropics A, August-September 1996. J. Geophys. Res. [Atmos.] 104: 5567-
7 5583.
8 Holland, H. D. (1978) The chemistry of the atmosphere and oceans. New York, NY: Wiley.
9 Holland, E. A.; Braswell, B. H.; Sulzman, J.; Lamarque, J.-F. (2005) Nitrogen deposition onto
10 the United States and Western Europe: synthesis of observations and models. Ecol. Appl.
11 15:38-57.
12 Hollwedel, J.; Wenig, M.; Beirle, S.; Kraus, S.; Kiihl, S.; Wilms-Grabe, W.; Platt, U.; Wagner,
13 T. (2004) Year-to-year variations of spring time polar tropospheric BrO as seen by
14 GOME. Adv. Space Res. 34: 804-808.
15 Honrath, R. E.; Guo, S.; Peterson, M. C.; Dziobak, M. P.; Dibb, J. E.; Arsenault, M. A. (2000)
16 Photochemical production of gas phase NOX from ice crystal NCV. J. Geophys. Res.
17 [Atmos.] 105:24,183-24,190.
18 Honrath, R. E.; Lu, Y.; Peterson, M. C.; Dibb, J. E.; Arsenault, M. A.; Cullen, N. J.; Steffen, K.
19 (2002) Vertical fluxes of NOX, HONO, and HNO3 above the snowpack at Summit,
20 Greenland. Atmos. Environ. 36: 2629-2640.
21 Hoppel, W. A.; Caffrey, P. F. (2005) Oxidation of S(IV) in sea-salt aerosol at high pH: ozone
22 versus aerobic reaction. J. Geophys. Res. [Atmos.] 110(D23202):
23 10.1029/2005JD006239.
24 Horii, C. V. (2002) Tropospheric reactive nitrogen speciation, deposition, and chemistry at
25 Harvard Forest [dissertation]. Cambridge, MA: Harvard University.
26 Horii, C. V.; Munger, J. W.; Wofsy, S. C.; Zahniser, M.; Nelson, D.; McManus, J. B. (2004)
27 Fluxes of nitrogen oxides over a temperate deciduous forest. J. Geophys. Res. [Atmos.]
28 109(D08305): 10.1029/2003 JD004326.
29 Horii, C.; Munger, J. W.; Wofsy, S.; Zahniser, M.; Nelson, D.; McManus, J. B. (2006)
30 Atmospheric reactive nitrogen concentrations and flux budgets at a northeastern U.S
31 forest site. Agric. For. Meteorol. 136: 159-174.
32 Horowitz, L. W.; Walters, S.; Mauzerall, D. L.; Emmons, L. K.; Rasch, P. J.; Granier, C.; Tie,
33 X.; Lamarque, J.-F.; Schultz, M. G.; Tyndall, G. S.; Orlando, J. J.; Brasseur, G. P. (2003)
34 A global simulation of tropospheric ozone and related tracers: description and evaluation
35 of MOZART, version 2. J. Geophys. Res. [Atmos.] 108(D24): 10.1029/2002JD002853.
36 Hiibler, G.; Alvarez, R., II; Daum, P.; Dennis, R.; Gillani, N.; Kleinman, L.; Luke, W.; Meagher,
37 J.; Rider, D.; Trainer, M.; Valente, R. (1998) An overview of the airborne activities
38 during the Southern Oxidants Study (SOS) 1995 Nashville/Middle Tennessee ozone
39 study. J. Geophys. Res. [Atmos.] 103: 22,245-22,259.
40 Hudman, R. C.; Jacob, D. J.; Cooper, O. R.; Evans, M. J.; Heald, C. L.; Park, R. J.; Fehsenfeld,
41 F.; Flocke, F.; Holloway, J.; Hubler, G.; Kita, K.; Koike, M.; Kondo, Y.; Neuman, A.;
March 2008 AX2-143 DRAFT-DO NOT QUOTE OR CITE
-------
1 Nowak, J.; Oltmans, S.; Parrish, D.; Roberts, J. M.; Ryerson, T. (2004) Ozone production
2 in transpacific Asian pollution plumes and implications for ozone air quality in
3 California. J. Geophys. Res. [Atmos.] 109(D23): 10.1029/2004JD004974.
4 Hudman, R. C.; Jacob, D. J.; Turquety, S.; Leibensperger, E. M.; Murray, L. T.; Wu, S.;
5 Gilliland, A. B.; Avery, M.; Bertram, T. H.; Brune, W.; Cohen, R. C.; Dibb, J. E.; Flocke,
6 F. M.; Fried, A.; Holloway, J.; Neuman, J. A.; Orville, R.; Perring, A.; Ren, X.; Ryerson,
7 T. B.; Sachse, G. W.; Singh, H. B.; Swanson, A.; Wooldridge, P. J. (2007) Surface and
8 lightning sources of nitrogen oxides over the United States: magnitudes, chemical
9 evolution, and outflow. J. Geophys. Res. [Atmos.] 112(D12S05):
10 10.1029/2006JD007912.
11 Huebert, B. J.; Zhuang, L.; Howell, S.; Noone, K.; Noone, B. (1996) Sulfate, nitrate,
12 methanesulfonate, chloride, ammonium, and sodium measurements from ship, island, and
13 aircraft during the Atlantic Stratocumulus Transition Experiment/Marine Aerosol Gas
14 Exchange. J. Geophys. Res. [Atmos.] 101: 4413-4423.
15 Huey, L. G.; Dunlea, E. J.; Lovejoy, E. R.; Hanson, D. R.; Norton, R. B.; Fehsenfeld, F. C.;
16 Howard, C. J. (1998) Fast time response measurements of HNOs in air with a chemical
17 ionization mass spectrometer. J. Geophys. Res. [Atmos.] 103: 3355-3360.
18 Huey, L. G.; Tanner, D. J.; Slusher, D. L.; Dibb, J. E.; Arimoto, R.; Chen, G.; Davis, D.; Buhr,
19 M. P.; Nowak, J. B.; Mauldin, R. L., Ill; Eisele, F. L.; Kosciuch, E. (2004) CIMS
20 measurements of HNOs and SC>2 at the South Pole during ISCAT 2000. Atmos. Environ.
21 38:5411-5421.
22 Huntrieser, H.; Feigl, C.; Schlager, H.; Schroder, F.; Gerbig, C.; Van Velthoven, P.; Flatey, F.;
23 Thery, C.; Petzold, A.; Holler, H.; Schumann, U. (2002) Airborne measurements of NOX,
24 tracer species, and small particles during the European Lightning Nitrogen Oxides
25 Experiment. J. Geophys. Res. [Atmos.] 107(D11): 10.1029/2000JD000209.
26 Huntrieser, H.; Schlager, H.; Roiger, A.; Lichtenstern, M.; Schumann, U.; Kurz, C.; Brunner, D.;
27 Schwierz, C.; Richter, A.; Stohl, A. (2007) Lightning-produced NOX over Brazil during
28 TROCCINOX: airborne measurements in tropical and subtropical thunderstorms and the
29 importance of mesoscale convective systems. Atmos. Chem. Phys. 7: 2987-3013.
30 Huntzicker, J. J.; Hoffman, R. S.; Ling, C.-S. (1978) Continuous measurement and speciation of
31 sulfur-containing aerosols by flame photometry. Atmos. Environ. 12: 83-88.
32 International Agency for Research on Cancer (IARC). (1989) Diesel and gasoline engine
33 exhausts and some nitroarenes. Lyon, France: International Agency for Research on
34 Cancer. (IARC monographs on the evaluation of carcinogenic risks to humans: v. 46).
35 Jacob, D. J. (2000) Heterogeneous chemistry and tropospheric ozone. Atmos. Environ. 34: 2131-
36 2159.
37 Jacob, D. J.; Bakwin, P. S. (1991) Cycling of NOX in tropical forest canopies. In: Rogers, J. E.;
38 Whitman, W. B., eds. Microbial production and consumption of greenhouse gases.
39 Washington, DC: American Society of Microbiology; pp. 237-253.
40 Jacob, D. J.; Horowitz, L. W.; Munger, J. W.; Heikes, B. G.; Dickerson, R. R.; Artz, R. S.;
41 Keene, W. C. (1995) Seasonal transition from NOX- to hydrocarbon-limited conditions
March 2008 AX2-144 DRAFT-DO NOT QUOTE OR CITE
-------
1 for ozone production over the eastern United States in September. J. Geophys. Res.
2 [Atmos.] 100:9315-9324.
3 Jacobson, M. Z. (1999) Isolating nitrated and aromatic aerosols and nitrated aromatic gases as
4 sources of ultraviolet light absorption. J. Geophys. Res. [Atmos.] 104: 3527-3542.
5 Jacobson, M. Z. (2002) Atmospheric pollution: history, science, and regulation. New York, NY:
6 Cambridge University Press.
7 Jacobson, M. Z.; Lu, R.; Turco, R. P.; Toon, O. P. (1996) Development and application of a new
8 air pollution modeling system-Part I: gas-phase simulations. Atmos. Environ. 30: 1939-
9 1963.
10 Jaegle, L.; Jacob, D. J.; Wennberg, P. O.; Spivakovsky, C. M.; Hanisco, T. F.; Lanzendorf, E. J.;
11 Hintsa, E. J.; Fahey, D. W.; Keim, E. R.; Proffitt, M. H.; Atlas, E. L.; Flocke, F.;
12 Schauffler, S.; McElroy, C. T.; Midwinter, C.; Pfister, L.; Wilson, J. C. (1997) Observed
13 OH and HO2 in the upper troposphere suggests a major source from convective injection
14 of peroxides. Geophys. Res. Lett. 24: 3181-3184.
15 Jaegle, L.; Jacob, D. J.; Brune, W. H.; Wennberg, P. O. (2001) Chemistry of HOX radicals in the
16 upper troposphere. Atmos. Environ. 35: 469-489.
17 Jaegle, L.; Martin, R. V.; Chance, K.; Steinberger, L.; Kurosu, T. P.; Jacob, D. J.; Modi, A. L;
18 Yoboue, V.; Sigha-Nkamdjou, L.; Galy-Lacaux, C. (2004) Satellite mapping of rain-
19 induced nitric oxide emissions from soils. J. Geophys. Res. [Atmos.] 109(D21):
20 10.1029/2004JD004787.
21 Jaegle, L.; Steinberger, L.; Martin , R. V.; Chance, K. (2005) Global partitioning of NOX sources
22 using satellite observations: Relative roles of fossil fuel combustion, biomass burning and
23 soil emissions. Faraday Discuss. 130: 407-423.
24 Jaklevic, J. M.; Loo, B. W.; Fujita, T. Y. (1981) Automatic particulate sulfur measurements with
25 a dichotomous sampler and on-line x-ray fluorescence analysis. Environ. Sci. Technol.
26 15: 687-690.
27 Jenkin, M. E.; Cox, R. A.; Williams, D. J. (1988) Laboratory studies of the kinetics of formation
28 of nitrous acid from the thermal reaction of nitrogen dioxide and water vapour. Atmos.
29 Environ. 22: 487-498.
30 Jet Propulsion Laboratory. (2003) Chemical kinetics and photochemical data for use in
31 atmospheric studies. Pasadena, CA: California Institute of Technology; JPL publication
32 no. 02-25. Available: http://jpldataeval.jpl.nasa.gov/pdf/JPL_02-25_l_intro_rev0.pdf (18
33 December, 2003).
34 Jimenez, P.; Baldasano, J. M.; Dabdub, D. (2003) Comparison of photochemical mechanisms for
35 air quality modeling. Atmos. Environ. 37: 4179-4194.
36 Jobson, B. T.; Niki, H.; Yokouchi, Y.; Bottenheim, J.; Hopper, F.; Leaitch, R. (1994)
37 Measurements of C2-Ce hydrocarbons during the Polar Sunrise 1992 Experiment:
38 evidence for Cl atom and Br atom chemistry. J. Geophys. Res. [Atmos.] 99: 25,355-
39 25,368.
40 John, W.; Wall, S. M.; Ondo, J. L. (1988) A new method for nitric acid and nitrate aerosol
41 measurement using the dichotomous sampler. Atmos. Environ. 22: 1627-1635.
March 2008 AX2-145 DRAFT-DO NOT QUOTE OR CITE
-------
1 Johnson, D. W.; Lindberg, S. E., eds. (1992) Atmospheric deposition and forest nutrient cycling:
2 a synthesis of the integrated forest study. New York, NY: Springer-Verlag, Inc. (Billings,
3 W. D.; Golley, F.; Lange, O. L.; Olson, J. S.; Remmert, H., eds. Ecological studies:
4 analysis and synthesis: v. 91).
5 Junge, C. E.; Chagnon, C. W.;; Manson, J. E. (1961) Stratospheric aerosols. J. Meteorol. 18: 81-
6 108.
7 Kain, J. S.; Fritsch, J. M. (1993) Convective parameterization in mesoscale models: the Kain-
8 Fritsch scheme. In: Emanuel, K. A.; Raymond, D. J., eds. The Representation of Cumulus
9 Convection in Numerical Models. Boston, MA: American Meteorological Society; pp.
10 165-170. (Meteorological Monographs, v. 24, no. 46).
11 Kaiser, E. W.; Wu, C. H. (1977) A kinetic study of the gas phase formation and decomposition
12 reactions of nitrous acid. J. Phys. Chem. 81: 1701-1706.
13 Kamens, R. M.; Guo, J.; Guo, Z.; McDow, S. R. (1990) Polynuclear aromatic hydrocarbon
14 degradation by heterogenous reactions with N2Os on atmospheric particles. Atmos.
15 Environ. Part A 24: 1161-1173.
16 Kane, S. M.; Caloz, F.; Leu, M.-T. (2001) Heterogeneous uptake of gaseous N2O5 by (NH4)2SO4,
17 NH4HSO4, and H2SO4 aerosols. J. Phys. Chem. A 105: 6465-6470.
18 Kasibhatla, P.; Chameides, W. L. (2000) Seasonal modeling of regional ozone pollution in the
19 eastern United States. Geophys. Res. Lett. 27: 1415-1418.
20 Kawanaka, Y.; Matsumoto, E.; Sakamoto, K.; Wang, N.; Yun, S. J. (2004) Size distributions of
21 mutagenic compounds and mutagenicity in atmospheric particulate matter collected with
22 a low-pressure cascade impactor. Atmos. Environ. 38: 2125-2132.
23 Keck, L.; Wittmaack, K. (2005) Effect of filter type and temperature on volatilisation losses from
24 ammonium salts in aerosol matter. Atmos. Environ. 39: 4093-4100.
25 Keck, L.; Wittmaack, K. (2006) Simplified approach to measuring semivolatile inorganic
26 particulate matter using a denuded cellulose filter without backup filters. Atmos. Environ.
27 40:7106-7114.
28 Keene, W. C.; Savoie, D. L. (1998) The pH of deliquesced sea-salt aerosol in polluted marine
29 air. Geophys. Res. Lett. 25: 2181-2194.
30 Keene, W. C.; Jacob, D. J.; Fan, S.-M. (1996) Reactive chlorine: a potential sink for
31 dimethylsulfide and hydrocarbons in the marine boundary layer. Atmos. Environ. 30(6):
32 i-iii.
33 Keene, W. C.; Sander, R.; Pszenny, A. A. P.; Vogt, R.; Crutzen, P. J.; Galloway, J. N. (1998)
34 Aerosol pH in the marine boundary layer: a review and model evaluation. J. Aerosol Sci.
35 29: 339-356.
36 Keene, W. C.; Khalil, M. A. K.; Erickson, D. J., Ill; McCulloch, A.; Graedel, T. E.; Lobert, J.
37 M.; Aucott, M. L.; Gong, S. L.; Harper, D. B.; Kleiman, G.; Midgley, P.; Moore, R. M.;
38 Seuzaret, C.; Sturges, W. T.; Benkovitz, C. M.; Koropalov, V.; Barrie, L. A.; Li, Y. F.
39 (1999) Composite global emissions of reactive chlorine from anthropogenic and natural
40 sources: reactive chlorine emissions inventory. J. Geophys. Res. [Atmos.] 104(D7):
41 8429-8440.
March 2008 AX2-146 DRAFT-DO NOT QUOTE OR CITE
-------
1 Keene, W. C.; Pszenny, A. A. P.; Maben, J. R.; Sander, R. (2002) Variation of marine aerosol
2 acidity with particle size. Geophys. Res. Lett. 29(7): 10.1029/2001GLO13881.
3 Keene, W. C.; Pszenny, A. A. P.; Maben, J. R.; Stevenson, E.; Wall, A. (2004) Closure
4 evaluation of size-resolved aerosol pH in the New England coastal atmosphere during
5 summer. J. Geophys. Res. [Atmos.] 109(D23307): 10.1029/2004JD004801.
6 Keene, W. C.; Lobert, R. M.; Crutzen, P. J.; Maben, J. R.; Scharffe, D. H.; Landmann, T.; Hely,
7 C.; Brain, C. (2006) Emissions of major gaseous and particulate species during
8 experimental burns of southern African biomass. J. Geophys. Res. [Atmos.]
9 111(D04301): 10.1029/2005 JD006319.
10 Keene, W. C.; Stutz, J.; Pszenny, A. A. P.; Maben, J. R.; Fischer, E.; Smith, A. M.; Von Glasow,
11 R.; Pechtl, S.; Sive, B.C.; Varner, R. K. (2007) Inorganic chlorine and bromine in coastal
12 New England air during summer. J. Geophys. Res. [Atmos.]:
13 doi:10.1029/2006JD007689, in press.
14 Kelly, T. J.; Spicer, C. W.; Ward, G. F. (1990) An assessment of the luminol chemiluminescence
15 technique for measurement of NO2 in ambient air. Atmos. Environ. Part A 24: 2397-
16 2403.
17 Kim, K.-H.; Kim, M.-Y. (2001) Comparison of an open path differential optical absorption
18 spectroscopy system and a conventional in situ monitoring system on the basis of long-
19 term measurements of SO2, NO2, and O3. Atmos. Environ. 35: 4059-4072.
20 Kim, B. M.; Lester, J.; Tisopulos, L.; Zeldin, M. D. (1999) Nitrate artifacts during PM2.5
21 sampling in the South Coast Air Basin of California. J. Air Waste Manage. Assoc.
22 49(special issue): PM142-153.
23 Kim, S.-W.; Heckel, A.; McKeen, S. A.; Frost, G. J.; Hsie, E.-Y.; Trainer, M. K.; Richter, A.;
24 Burrows, J. P.; Peckham, S. E.; Grell, G. A. (2006) Satellite-observed U.S. power plant
25 NOX emission reductions and their impact on air quality. Geophys. Res. Lett.
26 33(L22812): 10.1029/2006GL027749.
27 Kirchstetter, T. W.; Harley, R. A. (1996) Measurement of nitrous acid in motor vehicle exhaust.
28 Environ. Sci. Technol. 30: 2843-2849.
29 Kireev, S. V.; Shnyrev, S. L.; Zhiganov, A. A. (1999) A laser fluorimeter for NO and NO2 in
30 atmosphere. Instrum. Exp. Tech. 42: 701-703.
31 Kittelson, D. B.; McKenzie, R.; Vermeersch, M.; Dorman, F.; Pui, D.; Linne, M.; Liu, B.;
32 Whitby, K. (1978) Total sulfur aerosol concentration with an electrostatically pulsed
33 flame photometric detector system. Atmos. Environ. 12: 105-111.
34 Kleffmann, J.; Becker, K. H.; Wiesen, P. (1998) Heterogeneous NO2 conversion processes on
35 acid surfaces: possible atmospheric implications. Atmos. Environ. 32: 2721-2729.
36 Kleinman, L. I. (1991) Seasonal dependence of boundary-layer peroxide concentration: the low
37 and high NOX regimes. J. Geophys. Res. [Atmos.] 96: 20,721-20,733.
38 Kleinman, L. L; Daum, P. H.; Imre, D. G.; Lee, J. H.; Lee, Y.-N.; Nunnermacker, L. J.;
39 Springston, S. R.; Weinstein-Lloyd, J.; Newman, L. (2000) Ozone production in the New
40 York City urban plume. J. Geophys. Res. [Atmos.] 105: 14,495-14,511.
March 2008 AX2-147 DRAFT-DO NOT QUOTE OR CITE
-------
1 Kleinman, L. I; Daum, P. H.; Lee, Y.-N.; Nunnermacker, L. J.; Springston, S. R.; Weinstein-
2 Lloyd, J.; Rudolph, J. (2001) Sensitivity of ozone production rate to ozone precursors.
3 Geophys. Res. Lett. 28: 2903-2906.
4 Kley, D.; Crutzen, P. J.; Smit, H. G. J.; Vomel, H.; Oltmans, S. J.; Grassl, H.; Ramanathan, V.
5 (1996) Observations of near-zero ozone concentrations over the convective Pacific:
6 effects on air chemistry. Science (Washington, DC) 274: 230-233.
7 Korhonen, P.; Kulmala, M.; Laaksonen, A.; Viisanen, Y.; McGraw, R.; Seinfeld, J. H. (1999)
8 Ternary nucleation of H2SO4, NH3, and H2O in the atmosphere. J. Geophys. Res.
9 [Atmos.] 104: 26,349-26,353.
10 Kotchenruther, R. A.; Jaffe, D. A.; Jaegle, L. (2001) Ozone photochemistry and the role of
11 peroxyacetyl nitrate in the springtime northeastern Pacific troposphere: results from the
12 Photochemical Ozone Budget of the Eastern North Pacific Atmosphere (PHOBEA)
13 campaign. J. Geophys. Res. [Atmos.] 106: 28,731-28,742.
14 Koutrakis, P.; Wolfson, J. M.; Slater, J. L.; Brauer, M.; Spengler, J. D.; Stevens, R. K.; Stone, C.
15 L. (1988a) Evaluation of an annular denuder/filter pack system to collect acidic aerosols
16 and gases. Environ. Sci. Technol. 22: 1463-1468.
17 Koutrakis, P.; Wolfson, J. M.; Spengler, J. D. (1988b) An improved method for measuring
18 aerosol strong acidity: results from a nine-month study in St. Louis, Missouri and
19 Kingston, Tennessee. Atmos. Environ. 22: 157-162.
20 Koutrakis, P.; Thompson, K. M.; Wolfson, J. M.; Spengler, J. D.; Keeler, G. J.; Slater, J. L.
21 (1992) Determination of aerosol strong acidity losses due to interactions of collected
22 particles: results from laboratory and field studies. Atmos. Environ. Part A 26: 987-995.
23 Kulmala, M.; Pirjola, L.; Makela, J. M. (2000) Stable sulphate clusters as a source of new
24 atmospheric particles. Nature (London, U.K.) 404: 66-69.
25 Kumar, N.; Russell, A. G. (1996) Development of a computationally efficient, reactive subgrid-
26 scale plume model and the impact in the northeastern United States using increasing
27 levels of chemical detail. J. Geophys. Res. [Atmos.] 101: 16,737-16,744.
28 Lammel, G.; Cape, J. N. (1996) Nitrous acid and nitrite in the atmosphere. Chem. Soc. Rev. 25:
29 361-369.
30 Lammel, G.; Perner, D. (1988) The atmospheric aerosol as a source of nitrous acid in the
31 polluted atmosphere. J. Aerosol Sci. 19: 1199-1202.
32 Lawson, D. R. (1990) The Southern California Air Quality Study. J. Air Waste Manage. Assoc.
33 40:156-165.
34 Leigh, R. J. (2006) Concurrent multiaxis differential optical absorption spectroscopy system for
35 the measurement of tropospheric nitrogen dioxide. Appl. Opt. 45: 7504-7518.
36 Lerdau, M. T.; Munger, J. W.; Jacob, D. J. (2000) The NO2 flux conundrum. Science 289: 2291,
37 2293.
38 Leue, C.; Wenig, M.; Wagner, T.; Klimm, O.; Platt, U.; Jahne, B. (2001) Quantitative analysis of
39 NOx emissions from Global Ozone Monitoring Experiment satellite image sequences. J.
40 Geophys. Res. [Atmos.] 106(D6): 5493-5505.
March 2008 AX2-148 DRAFT-DO NOT QUOTE OR CITE
-------
1 Leung, F.-Y. T.; Colussi, A. J.; Hoffmann, M. R. (2002) Isotopic fractionation of carbonyl
2 sulfide in the atmosphere: implications for the source of background stratospheric sulfate
3 aerosol. Geophys. Res. Lett. 29: 10.1029/2001GL013955.
4 Levine, J. S.; Bobbe, T.; Ray, N.; Witt, R. G.; Singh, A. (1999) Wildland fires and the
5 environment: a global synthesis. Nairobi, Kenya: United Nations Environment
6 Programme (UNEP), Division of Environmental Information, Assessment an Early
7 Warning (DEIA&EW); report no. UNEP/DEIAEW/TR.99.1. Available:
8 http://www.na.unep.net/publications/wildfire.pdffl7 August, 2007].
9 Levelt, P. F.; Yen Den Oord, G. H. J.; Dobber, M. R.; Malkki, A.; Visser, H.; De Vries, J.;
10 Stammes, P.; Lundell, J. O. V.; Saari, H. (2006) The ozone monitoring instrument. IEEE
11 Trans. Geosci. Remote Sens. 44: 1093-1101.
12
13
14 Liang, J.; Horowitz, L. W.; Jacob, D. J.; Wang, Y.; Fiore, A. M.; Logan, J. A.; Gardner, G. M.;
15 Munger, J. W. (1998) Seasonal budgets of reactive nitrogen species and ozone over the
16 United States, and export fluxes to the global atmosphere. J. Geophys. Res. [Atmos.] 103:
17 13,435-13,450.
18 Liu, X. H.; Hegg, D. A.; Stoelinga, M. T. (200la) Numerical simulation of new particle
19 formation over the northwest Atlantic using the MM5 mesoscale model coupled with
20 sulfur chemistry. J. Geophys. Res. [Atmos.] 106: 9697-9715.
21 Liu, H.; Jacob, D. J.; Bey, I; Yantosca, R. M. (2001b) Constraints from 210Pb and 7Be on wet
22 deposition and transport in a global three-dimensional chemical tracer model driven by
23 assimilated meteorological fields. J. Geophys. Res. [Atmos.] 106: 12,109-12,128.
24 Lobert, J. M.; Scharffe, D. H.; Hao, W.-M.; Kuhlbusch, T. A.; Seuwen, R.; Warneck, P.;
25 Crutzen, P. J. (1991) Experimental evaluation of biomass burning emissions: nitrogen
26 and carbon containing compounds. In: Levine, J. S., ed. Global biomass burning:
27 atmospheric, climatic, and biospheric implications. Cambridge, MA: MIT Press; pp. 289-
28 304.
29 Lobert, J. M.; Keene, W. C.; Logan, J. A.; Yevich, R. (1999) Global chlorine emissions from
30 biomass burning: Reactive Chlorine Emissions Inventory. J. Geophys. Res. [Atmos.]
31 104(D7): 8373-8389.
32 Longfellow, C. A.; Ravishankara, A. R.; Hanson, D. R. (1999) Reactive uptake on hydrocarbon
33 soot: focus on NO2. J. Geophys. Res. [Atmos.] 104: 13,833-13,840.
34 Lu, R.; Turco, R. P.; Jacobson, M. Z. (1997) An integrated air pollution modeling system for
35 urban and regional scales: 1. Structure and performance. J. Geophys. Res. [Atmos.] 102:
36 6063-6079.
37 Luke, W. T. (1997) Evaluation of a commercial pulsed fluorescence detector for the
38 measurement of low-level SC>2 concentrations during the Gas-Phase Sulfur
39 Intercomparison Experiment. J. Geophys. Res. [Atmos.] 102: 16,255-16,265.
March 2008 AX2-149 DRAFT-DO NOT QUOTE OR CITE
-------
1 Luke, W. T.; Dickerson, R. R.; Ryan, W. F.; Pickering, K. E.; Nunnermacker, L. J. (1992)
2 Tropospheric chemistry over the lower Great Plains of the United States. 2. Trace gas
3 profiles and distributions. J. Geophys. Res. [Atmos.] 97: 20,647-20,670.
4 Luke, W. T.; Watson, T. B.; Olszyna, K. J.; Gunter, R. L.; McMillen, R. T.; Wellman, D. L.;
5 Wilkison, S. W. (1998) A comparison of airborne and surface trace gas measurements
6 during the Southern Oxidants Study (SOS). J. Geophys. Res. [Atmos.] 103: 22,317-
7 22,337.
8 Ma, J.; Richter, A.; Burrows, J. P.; NuB, H.; Van Aardenne, J. A. (2006) Comparison of model-
9 simulated tropospheric NC>2 over China with GOME-satellite data. Atmos. Environ. 40:
10 593-604.
11 Madronich, S. (1987) Photodissociation in the atmosphere. 1. Actinic flux and the effects of
12 ground reflections and clouds. J. Geophys. Res. [Atmos.] 92: 9740-9752.
13 Mahrt, L. (1998) Stratified atmospheric boundary layers and breakdown of models. Theor.
14 Comput. Fluid Dyn. 11: 263-279.
15
16
17 Marley, N. A.; Gaffney, J. S.; White, R. V.; Rodriguez-Cuadra, L.; Herndon, S. E.; Dunlea, E.;
18 Volkamer, R. M.; Molina, L. T.; Molina, M. J. (2004) Fast gas chromatography with
19 luminol chemiluminescence detection for the simultaneous determination of nitrogen
20 dioxide and peroxyacetyl nitrate in the atmosphere. Rev. Sci. Instrum. 75: 4595-4605.
21 Martilli, A.; Neftel, A.; Favaro, G.; Kirchner, F.; Sillman, S.; Clappier, A. (2002) Simulation of
22 the ozone formation in the northern part of the Po Valley. J. Geophys. Res. [Atmos.]
23 107(D22): 10.1029/2001JD000534.
24 Martin, D.; Tsivou, M.; Bonsang, B.; Abonnel, C.; Carsey, T.; Springer-Young, M.; Pszenny, A.;
25 Suhre, K. (1997) Hydrogen peroxide in the marine atmospheric boundary layer during the
26 Atlantic Stratocumulus Transition Experiment/Marine Aerosol and Gas Exchange
27 experiment in the eastern subtropical North Atlantic. J. Geophys. Res. [Atmos.] 102:
28 6003-6015.
29 Martin, R. V.; Chance, K.; Jacob, D. J.; Kurosu, T. P.; Spurr, R. J. D.; Bucsela, E.; Gleason, J. F.;
30 Palmer, P. L; Bey, L; Fiore, A. M.; Li, Q.; Yantosca, R. M.; Koelemeijer, R. B. A. (2002)
31 An improved retrieval of tropospheric nitrogen dioxide from GOME. J. Geophys. Res.
32 [Atmos.] 107(D20): 10.1029/2001DOO1027.
33 Martin, R. V.; Jacob, D. J.; Chance, K. V.; Kurosu, T. P.; Palmer, P. L; Evans, M. J. (2003)
34 Global inventory of nitrogen oxide emissions constrained by space-based observations of
35 NO2 columns. J. Geophys. Res. [Atmos.] 108(D17): 10.1029/2003JD003453.
36 Martin, R. V.; Fiore, A. M.; Van Donkelaar, A. (2004a) Space-based diagnosis of surface ozone
37 sensitivity to anthropogenic emissions. Geophys. Res. Lett. 31(L06120):
38 10.1029/2004GL019416.
39 Martin, R. V.; Parrish, D. D.; Ryerson, T. B.; Nicks, D. K., Jr.; Chance, K.; Kurosu, T. P.; Jacob,
40 D. J.; Sturges, E. D.; Fried, A.; Wert, B. P. (2004b) Evaluation of GOME satellite
41 measurements of trophospheric NO2 and HCHO using regional data from aircraft
March 2008 AX2-150 DRAFT-DO NOT QUOTE OR CITE
-------
1 campaigns in the southeastern United States. J. Geophys. Res. [Atmos.] 109(D24307):
2 10.1029/2004JD004869.
3 Martin, R. V.; Sioris, C. E.; Chance, K.; Ryerson, T. B.; Bertram, T. H.; Wooldridge, P. J.;
4 Cohen, R. C.; Neuman, J. A.; Swanson, A.; Flocke, F. M. (2006) Evaluation of space-
5 based constraints on nitrogen oxide emissions with regional aircraft measurements over
6 and downwind of eastern North America. J. Geophys. Res. [Atmos.] 111(D15308):
7 10.1029/2005 JD006680.
8 Martinez, M.; Arnold, T.; Perner, D. (1999) The role of bromine and chlorine chemistry for
9 arctic ozone depletion events in Ny-Alesund and comparison with model calculations.
10 Ann. Geophys. 17: 941-956.
11 Massman, W. J.; Pederson, J.; Delany, A.; Grantz, D.; Denhartog, G.; Neumann, H. H.; Oncley,
12 S. P.; Pearson, R.; Shaw, R. H. (1994) An evaluation of the regional acid deposition
13 model surface module for ozone uptake at 3 sites in the San-Joaquin valley of California.
14 J. Geophys. Res. [Atmos.] 99: 8281-8294.
15 Matsumi, Y.; Murakami, S.-L; Kono, M.; Takahashi, K.; Koike, M.; Kondo, Y. (2001) High-
16 sensitivity instrument for measuring atmospheric NC>2. Anal. Chem. 73: 5485-5493.
17 Matsumi, Y.; Shigemori, H.; Takahashi, K. (2005) Laser-induced fluorescence instrument for
18 measuring atmospheric SC>2. Atmos. Environ. 39: 3177-3185.
19 Mauldin, R. L., Ill; Tanner, D. J.; Eisele, F. L. (1998) A new chemical ionization mass
20 spectrometer technique for the fast measurement of gas phase nitric acid in the
21 atmosphere. J. Geophys. Res. [Atmos.] 103: 3361-3367.
22 McClenny, W. A., ed. (2000) Recommended methods for ambient air monitoring of NO, NC>2,
23 NOy, and individual NOZ species. Research Triangle Park, NC: U.S. Environmental
24 Protection Agency, National Exposure Research Laboratory; report no. EPA/600/R-
25 01/005.
26 McFiggans, G. (2005) Marine aerosols and iodine emissions. Nature (London, U.K.) 433(7026):
27 E13.
28 McFiggans, G.; Plane, J. M. C.; Allan, B.; Carpenter, L. J.; Coe, H.; O'Dowd, C. (2000) A
29 modeling study of iodine chemistry in the marine boundary layer. J. Geophys. Res.
30 [Atmos.] 105: 14,371-14,385.
31 McFiggans, G.; Coe, H.; Burgess, R.; Allan, J.; Cubison, M.; Alfarra, M. R.; Saunders, R.; Saiz-
32 Lopez, A.; Plane, J. M. C.; Wevill, D.; Carpenter, L. J.; Rickard, A. R.; Monks, P. S.
33 (2004) Direct evidence for coastal iodine particles from Laminaria macroalgae - linkage
34 to emissions of molecular iodine. Atmos. Chem. Phys. 4: 701-713.
35 McKeen, S. A.; Liu, S. C. (1993) Hydrocarbon ratios and photochemical history of air masses.
36 Geophys. Res. Lett. 20: 2363-2366.
37 McKeen, S. A.; Liu, S. C.; Hsie, E.-Y.; Lin, X.; Bradshaw, J. D.; Smyth, S.; Gregory, G. L.;
38 Blake, D. R. (1996) Hydrocarbon ratios during PEM-WEST A: a model perspective. J.
39 Geophys. Res. [Atmos.] 101: 2087-2109.
40 McKenna, D. S.; Konopka, P.; Groofi, J.-U.; Gunther, G.; Muller, R.; Spang, R.; Offermann, D.;
41 Orsolini, Y. (2002) A new chemical Lagrangian model of the stratosphere (CLaMS) 1.
March 2008 AX2-151 DRAFT-DO NOT QUOTE OR CITE
-------
1 formulation of advection and mixing. J. Geophys. Res. [Atmos.] 107(D16):
2 10.1029/2000JD000114.
3 Meagher, J. F.; Cowling, E. B.; Fehsenfeld, F. C.; Parkhurst, W. J. (1998) Ozone formation and
4 transport in southeastern United States: overview of the SOS Nashville/Middle Tennessee
5 Ozone Study. J. Geophys. Res. [Atmos.] 103: 22,213-22,223.
6 Mebust, M. R.; Eder, B. K.; Binkowski, F. S.; Roselle, S. J. (2003) Models-3 community
7 multiscale air quality (CMAQ) model aerosol component. 2. Model evaluation. J.
8 Geophys. Res. [Atmos.] 108(D6): 10.1029/2001JD001410.
9 Mendoza-Dominguez, A.; Russell, A. G. (2000) Iterative inverse modeling and direct sensitivity
10 analysis of a photochemical air quality model. Environ. Sci. Technol. 3: 4974-4981.
11 Mendoza-Dominguez, A.; Russell, A. G. (2001) Estimation of emission adjustments from the
12 application of four-dimensional data assimilation to photochemical air quality modeling.
13 Atmos. Environ. 35: 2879-2894.
14 Meng, Z.; Seinfeld, J. H. (1996) Time scales to achieve atmospheric gas-aerosol equilibrium for
15 volatile species. Atmos. Environ. 30: 2889-2900.
16 Milford, J. B.; Gao, D.; Sillman, S.; Blossey, P.; Russell, A. G. (1994) Total reactive nitrogen
17 (NOy) as an indicator of the sensitivity of ozone to reductions in hydrocarbon and NOX
18 emissions. J. Geophys. Res. 99: 3533-3542.
19 Mocker, D.; Jung, K.; Forstel, H.; Schuiirmann, G. (1998) Isotopic and enzymatic investigations
20 into the assimilation and effect of NO2 on C3 and C4 plants. J. Appl. Bot. 72: 186-190.
21 Moorthi, S.; Suarez, M. J. (1992) Relaxed Arakawa-Schubert: a parameterization of moist
22 convection for general circulation models. Mon. Weather Rev. 120: 978-1002.
23 Mueller, P. K.; Collins, J. F. (1980) Development of a particulate sulfate analyzer. Palo Alto,
24 CA: Electric Power Research Institute; report no. P-1382F. Available from: NTIS,
25 Springfield, VA; EPRI-EA-1492.
26 Munger, J. W.; Wofsy, S. C.; Bakwin, P. S.; Fan, S.-M.; Goulden, M. L.; Daube, B. C.;
27 Goldstein, A. H. (1996) Atmospheric deposition of reactive nitrogen oxides and ozone in
28 a temperate deciduous forest and a subarctic woodland. 1. Measurements and
29 mechanisms. J. Geophys. Res. [Atmos.] 101: 12639-12657.
30 Munger, J. W.; Fan, S.-M.; Bakwin, P. S.; Goulden, M. L.; Goldstein, A. H.; Colman, A. S.;
31 Wofsy, S. C. (1998) Regional budgets of nitrogen oxides from continental sources:
32 variations of rates for oxidation and deposition with season and distance from source
33 regions. J. Geophys. Res. [Atmos.] 103: 8355-8368.
34 Mufioz, M. S. S.; Rossi, M. J. (2002) Heterogeneous reactions of HNOs with flame soot
35 generated under different combustion conditions. Reaction mechanism and kinetics.
36 Phys. Chem. Chem. Phys. 4: 5110-5118.
37 Myhre, G.; Berglen, T. F.; Myhre, C. E.; Isaksen, I. S. A. (2004) The radiative effect of the
38 anthropogenic influence on the stratospheric sulfate aerosol layer. Tellus Ser. B 56: 294-
39 299.
March 2008 AX2-152 DRAFT-DO NOT QUOTE OR CITE
-------
1 Nagao, I; Matsumoto, K.; Tanaka, H. (1999) Sunrise ozone destruction found in the sub-tropical
2 marine boundary layer. Geophys. Res. Lett. 26: 3377-3380.
3 Nakamura, K.; Kondo, Y.; Chen, G.; Crawford, J. H.; Takegawa, N.; Koike, M.; Kita, K.;
4 Miyazaki, Y.; Shelter, R. E.; Lefer, B. L.; Avery, M.; Matsumoto, J. (2003) Measurement
5 of NO2 by the photolysis conversion technique during the Transport and Chemical
6 Evolution Over the Pacific (TRACE-P) campaign. J. Geophys. Res. [Atmos.] 108(D24):
7 10.1029/2003 JD003712.
8 NARSTO. (2005) Improving emission inventories for effective air quality management across
9 North America. A NARSTO assessment. Pasco, WA: The NARSTO Emission Inventory
10 Assessment Team; report no. NARSTO-05-001. Available:
11 http://www.narsto.org/section.src?SID=8 [18 April, 2006].
12 National Bureau of Standards. (1975) Catalog of NBS standard reference materials, 1975-76
13 edition. Washington, DC: U.S. Department of Commerce, National Bureau of Standards;
14 NBS special publication no. 260.
15 National Research Council. (1991) Rethinking the ozone problem in urban and regional air
16 pollution. Washington, DC: National Academy Press. Available:
17 http://www.nap.edu/books/0309046319/html/ [26 March, 2004].
18 National Research Council. (2002) Estimating the public health benefits of proposed air
19 pollution regulations. Washington, DC: National Academy of Sciences. Available at:
20 http://books.nap.edu/books/0309086094/html/index.html [7 November, 2002].
21 Neuman, J. A.; Huey, L. G; Dissly, R. W.; Fehsenfeld, F. C.; Flocke, F.; Holecek, J. C.;
22 Holloway, J. S.; Hubler, G.; Jakoubek, R.; Nicks, D. K., Jr.; Parrish, D. D.; Ryerson, T.
23 B.; Sueper, D. T.; Weinheimer, A. J. (2002) Fast-response airborne in-situ measurements
24 of HNO3 during the Texas 2000 Air Quality Study. J. Geophys. Res. [Atmos.] 107:
25 10.1029/2001JD001437.
26 Nicks, D. K.; Benner, R. L. (2001) Subminute measurements of SO2 and low parts per trillion by
27 volume mixing ratios in the atmosphere. J. Geophys. Res. [Atmos.] 106: 27,769-27,776.
28 Nikitas, C.; Clemitshaw, K. C.; Oram, D. E.; Penkett, S. A. (1997) Measurements of PAN in the
29 polluted boundary layer and free troposphere using a luminol-NO2 detector combined
30 with a thermal converter. J. Atmos. Chem. 28: 339-359.
31 Notholt, J.; Hjorth, J.; Raes, F. (1992a) Formation of HNO2 on aerosol surfaces during foggy
32 periods in the presence of NO and NO2. Atmos. Environ. Part A 26: 211-217.
33 Notholt, J.; Hjorth, J.; Raes, F.; Schrems, O. (1992b) Simultaneous long path field measurements
34 of HNO2, CH2O and aerosol. Ber. Bunsen Ges. Phys. Chem. 96: 290-293.
35 Nunnermacker, L. J.; Imre, D.; Daum, P. H.; Kleinman, L.; Lee, Y.-N.; Lee, J. H.; Springston,
36 S. R.; Newman, L.; Weinstein-Lloyd, J.; Luke, W. T.; Banta, R.; Alvarez, R.; Senff, C.;
37 Sillman, S.; Holdren, M.; Keigley, G. W.; Zhou, X. (1998) Characterization of the
38 Nashville urban plume on July 3 and July 18, 1995. J. Geophys. Res. [Atmos.] 103:
39 28,129-28,148.
40 Odman, M. T.; Ingram, C. I. (1996) Multiscale air quality simulation platform (MAQSIP):
41 source code documentation and validation. Research Triangle Park, NC: MCNC;
March 2008 AX2-153 DRAFT-DO NOT QUOTE OR CITE
-------
1 technical report ENV-96TR002-vl.0. Available:
2 http://www.ce.gatech.edu/~todman/maqsip.pdf (9 September 2003).
3 Odman, M. T.; Russell, A. G. (1999) Mass conservative coupling of non-hydrostatic
4 meteorological models with air quality models. In: Grying, S.-E.; Batchvarova, E., eds.
5 Air Pollution Modeling and its Application XIII. New York, NY: Plenum Press.
6 O'Dowd, C.; McFiggans, G.; Creasey, D. I; Pirjola, L.; Hoell, C.; Smith, M. H.; Allan, B. J.;
7 Plane, J. M. C.; Heard, D. E.; Lee, J. D.; Pilling, M. J.; Kulmala, M. (1999) On the
8 photochemical production of new particles in the coastal boundary layer. Geophys. Res.
9 Lett. 26: 1707-1710.
10 O'Dowd, C. D.; Jimenez, J. L.; Bahrein!, R.; Flagan, R. C.; Seinfeld, J. H.; Hameri, K.; Pirjola,
11 L.; Kulmala, M.; Jennings, S. G.; Hoffman, T. (2002) Marine aerosol formation from
12 biogenic iodine emissions. Nature (London, U.K.) 417: 632-636.
13 Ordonez, C.; Richter, A.; Steinbacher, M.; Zellweger, C.; NuB, H.; Burrows, J. P.; Prevot, A. S.
14 H. (2006) Comparison of 7 years of satellite-borne and ground-based tropospheric NO2
15 measurements around Milan, Italy. J. Geophys. Res. [Atmos.] 111(D05310):
16 10.1029/2005JD006305.
17 Osthoff, H. D.; Brown, S. S.; Ryerson, T. B.; Fortin, T. J.; Lerner, B. M.; Williams, E. J.;
18 Pettersson, A.; Baynard, T.; Dube, W. P.; Ciciora, S. J.; Ravishankara, A. R. (2006)
19 Measurement of atmospheric NO2 by pulsed cavity ring-down spectroscopy. J. Geophys.
20 Res. [Atmos.] 111(D12): 10.1029/2005JD006942.
21 Ott, L. E.; Pickering, K. E.; Stenchikov, G. L.; Huntrieser, H.; Schumann, U. (2007) Effects of
22 lightning NOX production during the July 21 European Lightning Nitrogen Oxides Project
23 storm studied with a three-dimensional cloud-scale chemical transport model. J.
24 Geophys. Res. [Atmos.] 112(D05307): 10.1029/2006JD007365.
25 Padro, J. (1996) Summary of ozone dry deposition velocity measurements and model estimates
26 over vineyard, cotton, grass and deciduous forest in summer. Atmos. Environ. 30: 2363-
27 2369.
28 Park, R. J.; Stenchikov, G. L.; Pickering; Dickerson, R. R.; Allen, D. J.; Kondragunta, S. (2001)
29 Regional air pollution and its radiative forcing: studies with a single column chemical
30 and radiation transport model. J. Geophys. Res. [Atmos.] 106: 28,751-28,770.
31 Parrish, D. D. (2006) Critical evaluation of US on-road vehicle emission inventories. Atmos.
32 Environ. 40: 2288-2300.
33 Parrish, D. D.; Fehsenfeld, F. C. (2000) Methods for gas-phase measurements of ozone, ozone
34 precursors and aerosol precursors. Atmos. Environ. 34: 1921-1957.
35 Parrish, D. D.; Trainer, M.; Buhr, M. P.; Watkins, B. A.; Fehsenfeld, F. C. (1991) Carbon
36 monoxide concentrations and their relation to concentrations of total reactive oxidized
37 nitrogen at two rural U.S. sites. J. Geophys. Res. [Atmos.] 96: 9309-9320.
38 Parrish, D. D.; Trainer, M.; Young, V.; Goldan, P. D.; Kuster, W. C.; Jobson, B. T.; Fehsenfeld,
39 F. C.; Lonneman, W. A.; Zika, R. D.; Farmer, C. T.; Riemer, D. D.; Rodgers, M. O.
40 (1998) Internal consistency tests for evaluation of measurements of anthropogenic
41 hydrocarbons in the troposphere. J. Geophys. Res. [Atmos.] 103: 22,339-22,359.
March 2008 AX2-154 DRAFT-DO NOT QUOTE OR CITE
-------
1 Parrish, D. D.; Trainer, M.; Hereid, D.; Williams, E. J.; Olszyna, K. J.; Harley, R. A.; Meagher,
2 J. F.; Fehsenfeld, F. C. (2002) Decadal change in carbon monoxide to nitrogen oxide
3 ratio in U.S. vehicular emissions J. Geophys. Res. [Atmos.] 107(D13):
4 10.1029/2001JD000720.
5 Perez, I.; Wooldridge, P. J.; Cohen, R. C. (2007) Laboratory evaluation of a novel thermal
6 dissociation chemiluminescence method for in situ detection of nitrous acid. Atmos.
7 Environ. 41: 3993-4001.
8 Perrini, G.; Tomasello, M.; Librando, V.; Minniti, Z. (2005) Nitrated polycyclic aromatic
9 hydrocarbons in the environment: formation, occurrences and analysis. Ann. Chim. 95:
10 567-577.
11 Petriconi, G. L.; Papee, H. M. (1972) On the photolytic separation of halogens from sea water
12 concentrates. Water Air Soil Pollut. 1: 117-131.
13 Pickering, K. E.; Thompson, A. M.; Dickerson, R. R.; Luke, W. T.; McNamara, D. P.;
14 Greenberg, J. P.; Zimmerman, P. R. (1990) Model calculations of tropospheric ozone
15 production potential following observed convective events. J. Geophys. Res. [Atmos.]
16 95: 14,049-14,062.
17 Pickering, K. E.; Thompson, A. M.; Scala, J. R.; Tao, W.-K.; Simpson, J.; Garstang, M. (1991)
18 Photochemical ozone production in tropical squall line convection during NASA Global
19 Tropospheric Experiment/Amazon Boundary Layer Experiment 2A. J. Geophys. Res.
20 [Atmos.] 96: 3099-3114.
21 Pickering, K. E.; Thompson, A. M.; Scala, J. R.; Tao, W.-K.; Simpson, J. (1992a) Ozone
22 production potential following convective redistribution of biomass burning emissions. J.
23 Atmos. Chem. 14: 297-313.
24 Pickering, K. E.; Thompson, A. M.; Scala, J. R.; Tao, W.-K.; Dickerson, R. R.; Simpson, J.
25 (1992b) Free tropospheric ozone production following entrainment of urban plumes into
26 deep convection. J. Geophys. Res. [Atmos.] 97: 17,985-18,000.
27 Pickering, K. E.; Thompson, A. M.; Tao, W.-K.; Kucsera, T. L. (1993) Upper tropospheric ozone
28 production following mesoscale convection during STEP/EMEX. J. Geophys. Res.
29 [Atmos.] 98: 8737-8749.
30 Pickering, K. E.; Thompson, A. M.; Tao, W.-K.; Rood, R. B.; McNamara, D. P.; Molod, A. M.
31 (1995) Vertical transport by convective clouds: comparisons of three modeling
32 approaches. Geophys. Res. Lett. 22: 1089-1092.
33 Pickering, K. E.; Thompson, A. M.; Wang, Y.; Tao, W.-K.; McNamara, D. P.; Kirchhoff, W. J.
34 H.; Heikes, B. G.; Sachse, G. W.; Bradshaw, J. D.; Gregory, G. L.; Blake, D. R. (1996)
35 Convective transport of biomass burning emissions over Brazil during TRACE A. J.
36 Geophys. Res. [Atmos.] 101: 23,993-24,012.
37 Pickering, K. E.; Wang, Y.; Tao, W.-K.; Price, C.; Muller, J.-F. (1998) Vertical distributions of
38 lightning NOX for use in regional and global chemical transport models. J. Geophys. Res.
39 [Atmos.] 103:31,203-31,216.
40 Pickering, K. E.; Thompson, A. M.; Kim, H.; DeCaria, A. J.; Pfister, L.; Kucsera, T. L.; Witte, J.
41 C.; Avery, M. A.; Blake, D. R.; Crawford, J. H.; Heikers, B. G; Sachse, G. W.;
March 2008 AX2-155 DRAFT-DO NOT QUOTE OR CITE
-------
1 Sandholm, S. T.; Talbot, R. W. (2001) Trace gas transport and scavenging in PEM-
2 Tropics B South Pacific Convergence Zone convection. J. Geophys. Res. [Atmos.] 106:
3 32,591-32,602.
4 Pierce, T; Geron, C.; Bender, L.; Dennis, R.; Tonnesen, G.; Guenther, A. (1998) Influence of
5 increased isoprene emissions on regional ozone modeling. J. Geophys. Res. [Atmos.]
6 103:25,611-25,629.
7 Pikelnaya, O.; Hurlock, S. C.; Trick, S.; Stutz, J. (2006) Measurements of reactive iodine species
8 on the Isles of Shoals, Gulf of Maine. J. Geophys. Res. [Atmos.]: submitted.
9 Pinto, J. P.; Turco, R. P.; Toon, O. B. (1989) Self-limiting physical and chemical effects in
10 volcanic eruption clouds. J. Geophys. Res. [Atmos.] 94: 11,165-11,174.
11 Pinto, J. P.; Bruhl, C.; Thompson, A. M. (1993) The current and future envirionmental role of
12 atmospheric methane. In: Khalil, M. A. K., ed. Atmospheric methane sources, sinks, and
13 role in global change, p. 514-531. (NATO ASI Series, v. 113).
14 Pitts, J. N., Jr. (1983) Formation and fate of gaseous and particulate mutagens and carcinogens in
15 real and simulated atmospheres. Environ. Health Perspect. 47: 115-140.
16 Pitts, J. N., Jr. (1987) Nitration of gaseous polycyclic aromatic hydrocarbons in simulated and
17 ambient urban atmospheres: a source of mutagenic nitroarenes. Atmos. Environ. 21:
18 2531-2547.
19 Pitts, J. N., Jr.; Biermann, H. W.; Atkinson, R.; Winer, A. M. (1984) Atmospheric implications
20 of simultaneous nighttime measurements of NOs radicals and HONO. Geophys. Res.
21 Lett. 11: 557-560.
22 Pitts, J. N., Jr.; Atkinson, R.; Sweetman, J. A.; Zielinska, B. (1985a) The gas-phase reaction of
23 naphthalene with ^Os to form nitronaphthalenes. Atmos. Environ. 19: 701-705.
24 Pitts, J. N., Jr.; Sweetman, J. A.; Zielinska, B.; Atkinson, R.; Winer, A. M.; Harger, W. P.
25 (1985b) Formation of nitroarenes from the reaction of polycyclic aromatic hydrocarbons
26 with dinitrogen pentoxide. Environ. Sci. Technol. 19: 1115-1121.
27 Pokharel, S. S.; Bishop, G. A.; Stedman, D. H. (2002) An on-road motor vehicle emissions
28 inventory for Denver: an efficient alternative to modeling. Atmos. Environ. 36: 5177-
29 5184.
30 Poppe, D.; Wallasch, M.; Zimmermann, J. (1993) The dependence of the concentration of OH on
31 its precursors under moderately polluted conditions: a model study. J. Atmos. Chem. 16:
32 61-78.
33 Possanzini, M.; De Santis, F.; Di Palo, V. (1999) Measurements of nitric acid and ammonium
34 salts in lower Bavaria. Atmos. Environ. 33: 3597-3602.
35 Prather, M. J.; Jacob, D. J. (1997) A persistent imbalance in HOX and NOX photochemistry in the
36 upper troposphere driven by deep tropical convection. Geophys. Res. Lett. 24: 3189-
37 3192.
38 Prevot, A. S. H.; Staehelin, J.; Kok, G. L.; Schillawski, R. D.; Neininger, B.; Staffelbach, T.;
39 Neftel, A.; Wernli, H.; Dommen, J. (1997) The Milan photooxidant plume. J. Geophys.
40 Res. [Atmos.] 102: 23,375-23,388.
March 2008 AX2-156 DRAFT-DO NOT QUOTE OR CITE
-------
1 Price, C.; Penner, J.; Prather, M. (1997) NOX from lightning: 1. Global distribution based on
2 lightning physics. J. Geophys. Res. [Atmos.] 102: 5929-5941.
3 Pryor, S. C.; S0rensen, L. L. (2000) Nitric acid—sea salt reactions: implications for nitrogen
4 deposition to water surfaces. J. Appl. Meteorol. 39: 725-731.
5 Pszenny, A. A. P.; Keene, W. C.; Jacob, D. J.; Fan, S.; Maben, J. R.; Zetwo, M. P.; Springer-
6 Young, M.; Galloway, J. N. (1993) Evidence of inorganic chlorine gases other than
7 hydrogen chloride in marine surface air. Geophys. Res. Lett. 20: 699-702.
8 Pszenny, A. A. P.; Moldanova, J.; Keene, W. C.; Sander, R.; Maben, J. R.; Martinez, M.;
9 Crutzen, P. J.; Perner, D.; Prinn, R. G. (2004) Halogen cycling and aerosol pH in the
10 Hawaiian marine boundary layer. Atmos. Chem. Phys. 4: 147-168.
11 Pszenny, A. A. P.; Fischer, E. V.; Russo, R. S.; Sive, B. C.; Varner, R. K. (2006) Estimates of Cl
12 atom concentrations and hydrocarbon kinetic reactivity in surface air at Appledore Island,
13 Maine (USA), during International Consortium for Atmospheric Research on Transport
14 and Transformation/Chemistry of Halogens at the Isles of Shoals. J. Geophys. Res.
15 [Atmos.] 112(D10S13): 10.1029/2006JD007725.
16 Radke, L. F.; Hegg, D. A.; Hobbs, P. V.; Nance, J. D.; Lyons, J. H.; Laursen, K. K.; Weiss, R.
17 E.; Riggan, P. J.; Ward, D. E. (1991) Particulate and trace gas emissions from large
18 biomass fires in North America. In: Levine, J. S., ed. Global biomass burning:
19 atmospheric, climatic, and biospheric implications. Cambridge, MA: MIT Press; pp. 209-
20 224.
21 Raivonen, M.; Keronen, P.; Vesala, T.; Kulmala, M.; Hari, P. (2003) Measuring shoot-level NOX
22 flux in field conditions: the role of blank chambers. Boreal Environ. Res. 8: 445-455.
23 Ramazan, K. A.; Syomin, D.; Finlayson-Pitts, B. J. (2004) The photochemical production of
24 HONO during the heterogeneous hydrolysis of NO2. Phys. Chem. Chem. Phys. 6: 3836-
25 3843.
26 Ramdahl, T.; Schjoldager, J.; Currie, L. A.; Hanssen, J. E.; M011er, M.; Klouda, G. A.; Alfheim,
27 I. (1984) Ambient impact of residential wood combustion in Elverum, Norway. Sci. Total
28 Environ. 36: 81-90.
29 Ramdahl, T.; Zielinska, B.; Arey, J.; Atkinson, R.; Winer, A. M.; Pitts, J. N., Jr. (1986)
30 Ubiquitous occurrence of 2-nitrofluoranthene and 2-nitropyrene in air. Nature (London)
31 321:425-427.
32 Ravishankara, A. R. (1997) Heterogeneous and multiphase chemistry in the troposphere. Science
33 (Washington, DC) 276: 1058-1065.
34 Reisen, F.; Arey, J. (2005) Atmospheric reactions influence seasonal PAH and nitro-PAH
35 concentrations in the Los Angeles basin. Environ. Sci. Technol. 39: 64-73.
36 Reissell, A.; Arey, J. (2001) Biogenic volatile organic compounds at Azusa and elevated sites
37 during the 1997 Southern California Ozone Study. J. Geophys. Res. [Atmos.] 106: 1607-
38 1621.
39 Reithmeier, C.; Sausen, R. (2002) ATTILA: atmospheric tracer transport in a Lagrangian model.
40 Tellus 54B: 278-299.
March 2008 AX2-157 DRAFT-DO NOT QUOTE OR CITE
-------
1 Ren, X. R.; Harder, H.; Martinez, M.; Lesher, R. L.; Oliger, A.; Shirley, T.; Adams, J.; Simpas,
2 J. B.; Brune, W. H. (2003) HOX concentrations and OH reactivity observations in New
3 York City during PMTACS-NY2001. Atmos. Environ. 37: 3627-3637.
4 Reynolds, S.; Michaels, H.; Roth, P.; Tesche, T. W.; McNally, D.; Gardner, L.; Yarwood, G.
5 (1996) Alternative base cases in photochemical modeling: their construction, role, and
6 value. Atmos. Environ. 30: 1977-1988.
7 Richter, A.; Burrows, J. P.; NuB, H.; Granier, C.; Niemeier, U. (2005) Increase in tropospheric
8 nitrogen dioxide over China observed from space. Nature (London, U.K.) 437: 129-132.
9 Ridley, B. A.; Dye, J. E.; Walega, J. G.; Zheng, J.; Grahek, F. E.; Rison, W. (1996) On the
10 production of active nitrogen by thunderstorms over New Mexico. J. Geophys. Res.
11 [Atmos.] 101:20,985-21,005.
12 Roberts, J. M.; Williams, J.; Baumann, K.; Buhr, M. P.; Goldan, P. D.; Holloway, J.; Hubler, G.;
13 Kuster, W. C.; McKeen, S. A.; Ryerson, T. B.; Trainer, M.; Williams, E. J.; Fehsenfeld,
14 F. C.; Bertman, S. B.; Nouaime, G.; Seaver, C.; Grodzinsky, G.; Rodgers, M.; Young, V.
15 L. (1998) Measurements of PAN, PPN, and MPAN made during the 1994 and 1995
16 Nashville Intensives of the Southern Oxidant Study: implications for regional ozone
17 production from biogenic hydrocarbons. J. Geophys. Res. [Atmos.] 103: 22,473-22,490.
18 Rodgers, M. O.; Davis, D. D. (1989) A UV-photofragmentation/laser-induced fluorescence
19 sensor for the atmospheric detection of HONO. Environ. Sci. Technol. 23: 1106-1112.
20 Russell, A.; Dennis, R. (2000) NARSTO critical review of photochemical models and modeling.
21 Atmos. Environ. 34: 2283-2324.
22 Russell, K. M.; Keene, W. C.; Maben, J. R.; Galloway, J. N.; Moody, J. L. (2003) Phase-
23 partitioning and dry deposition of atmospheric nitrogen at the mid-Atlantic U.S. coast. J.
24 Geophys. Res. [Atmos.] 108(D21): 10.1029/2003JD003736.
25 Russell, L. M.; Mensah, A. A.; Fischer, E. V.; Sive, B. C.; Varner, R. K.; Keene, W. C.; Stutz, J.;
26 Pszenny, A. A. P. (2006) Nanoparticle growth following photochemical a- and p-pinene
27 oxidation at Appledore Island during International Consortium for Research on Transport
28 and Transformation/Chemistry of Halogens at the Isles of Shoals 2004. J. Geophys. Res.
29 [Atmos.] 112: 10.1029/2006JD007736.
30 Ryerson, T. B.; Buhr, M. P.; Frost, G. J.; Goldan, P. D.; Holloway, J. S.; Hubler, G.; Jobson, B.
31 T.; Kuster, W. C.; McKeen, S. A.; Parrish, D. D.; Roberts, J. M.; Sueper, D. T.; Trainer,
32 M.; Williams, J.; Fehsenfeld, F. C. (1998) Emissions lifetimes and ozone formation in
33 power plant plumes. J. Geophys. Res. [Atmos.] 103: 22,569-22,583.
34 Ryerson, T. B.; Williams, E J.; Fehsenfeld, F. C. (2000) An efficient photolysis system for fast-
35 response NO2 measurements. J. Geophys. Res. [Atmos.] 105: 26,447-26,461.
36 Ryerson, T. B.; Trainer, M.; Holloway, J. S.; Parrish, D. D.; Huey, L. G.; Sueper, D. T.; Frost, G.
37 J.; Donnelly, S. G.; Schauffler, S.; Atlas, E. L.; Kuster, W. C.; Goldan, P. D.; Hubler, G.;
38 Meagher, J. F.; Fehsenfeld, F. C. (2001) Observations of ozone formation in power plant
39 plumes and implications for ozone control strategies. Science (Washington, DC) 292:
40 719-723.
March 2008 AX2-158 DRAFT-DO NOT QUOTE OR CITE
-------
1 Saathoff, H.; Naumann, K. H.; Riemer, N.; Kamm, S.; Mohler, O.; Schurath, U.; Vogel, H.;
2 Vogel, B. (2001) The loss of NO2, HNO3, NO3/N2O5, and HO2/HOONO2 on soot aerosol:
3 a chamber and modeling study. Geophys. Res. Lett. 28: 1957-1960.
4 Saiz-Lopez, A.; Plane, J. M. C. (2004) Novel iodine chemistry in the marine boundary layer.
5 Geophys. Res. Lett. 31(L04112): 10.1029/2003GL019215.
6 Saiz-Lopez, A.; Plane, J. M. C.; Shillito, J. A. (2004) Bromine oxide in the mid-latitude marine
7 boundary layer. Geophys. Res. Lett. 31(L03111): 10.1029/2003GL018956.
8 Sakamaki, F.; Hatakeyama, S.; Akimoto, H. (1983) Formation of nitrous acid and nitric oxide in
9 the heterogeneous dark reaction of nitrogen dioxide and water vapor in a smog chamber.
10 Int. J. Chem. Kinet. 15: 1013-1029.
11 Sakugawa, H.; Kaplan, I. R. (1989) H2O2 and O3 in the atmosphere of Los Angeles and its
12 vicinity: factors controlling their formation and their role as oxidants of SO2. J. Geophys.
13 Res. [Atmos.] 94: 12,957-12,973.
14 Salmeen, I; Durisin, A. M.; Prater, T. J.; Riley, T.; Schuetzle, D. (1982) Contribution of 1-
15 nitropyrene to direct-acting Ames assay mutagenicities of diesel particulate extracts.
16 Mutat. Res. 104: 17-23.
17 Sander, R.; Rudich, Y.; Von Glasow, R.; Crutzen, P. J. (1999) The role of BrNO3 in marine
18 tropospheric chemistry: a model study. Geophys. Res. Lett. 26: 2858-2860.
19 Sander, R.; Keene, W. C.; Pszenny, A. A. P.; Arimoto, R. Ayers, G. P.; Chainey, J. M.; Crutzen,
20 P. J.; Duce, R. A.; Huebert, B. J.; Maenhaut, W.; Turekian, V. C.; Van Dingenan, R.
21 (2003) Inorganic bromine in the marine boundary layer: a critical review. Atmos. Chem.
22 Phys. 3: 1301-1336.
23 Sasaki, J. C.; Arey, J.; Eastmond, D. A.; Parks, K. K.; Grosovsky, A. J. (1997) Genotoxicity
24 induced in human lymphoblasts by atmospheric reaction products of naphthalene and
25 phenanthrene. Mutat. Res. 393: 23-35.
26 Sawyer, R. F.; Harley, R. A.; Cadle, S. H.; Norbeck, J. M.; Slott, R.; Bravo, H. A. (2000) Mobile
27 sources critical review: 1998 NARSTO assessment. Atmos. Environ. 34: 2161-2181.
28 Scala, J. M.; Garstang, M.; Tao, W.-K.; Pickering, K. E.; Thompson, A. M.; Simpson, J.;
29 Kirchhoff, V. W. J. H.; Browell, E. V.; Sachse, G. W.; Torres. A. L.; Gregory, G. L.;
30 Rasmussen, R. W.; Khalil, M. A. K. (1990) Cloud draft structure and trace gas transport.
31 J. Geophys. Res. [Atmos.] 95: 17,015-17,030.
32 Schrimpf, W.; Lienaerts, K.; Muller, K. P.; Rudolph, J.; Neubert, R.; SchuBler, W.; Levin, I.
33 (1996) Dry deposition of peroxyacetyl nitrate (PAN): determination of its deposition
34 velocity at night from measurements of the atmospheric PAN and 222radon concentration
35 gradient. Geophys. Res. Lett. 23: 3599-3602.
36 Schubert, S. D.; Rood, R. B.; Pfaendtner, J. (1993) An assimilated dataset for earth science
37 applications. Bull. Am. Meteorol. Soc. 74: 2331-2342.
38 Schuetzle, D. (1983) Sampling of vehicle emissions for chemical analysis and biological testing.
39 Environ. Health Perspect. 47: 65-80.
March 2008 AX2-159 DRAFT-DO NOT QUOTE OR CITE
-------
1 Schultz, M. G.; Jacob, D. J.; Bradshaw, J. D.; Sandholm, S. T.; Dibb, J. E.; Talbot, R. W.; Singh,
2 H. B. (2000) Chemical NOX budget in the upper troposphere over the tropical South
3 Pacific. J. Geophys. Res. [Atmos.] 105: 6669-6679.
4 Seaman, N. L. (2000) Meteorological modeling for air quality assessments. Atmos. Environ. 34:
5 2231-2259.
6 Segschneider, H.-J.; Wildt, J.; Forstel, H. (1995) Uptake of 15NO2 by sunflower (Helianthus-
1 annuus) during exposures in light and darkness: quantities, relationship to stomatal
8 aperture and incorporation into different nitrogen pools within the plant. New Phytol.
9 131:109-119.
10 Seinfeld, J. H.; Pandis, S. N. (1998) Atmospheric chemistry and physics: from air pollution to
11 climate change. New York, NY: John Wiley & Sons, Inc.
12 Shepson, P. B.; Bottenheim, J. W.; Hastie, D. R.; Venkatram, A. (1992) Determination of the
13 relative ozone and PAN deposition velocities at night. Geophys. Res. Lett. 19: 1121-
14 1124.
15 Shepson, P. B.; Mackay, E.; Muthuramu, K. (1996) Henry's Law constants and removal
16 processes for several atmospheric p-hydroxy alkyl nitrates. Environ. Sci. Technol. 30:
17 3618-3623.
18 Shindell, D. T.; Faluvegi, G.; Stevenson, D. S.; Krol, M. C.; Emmons, L. K.; Lamarque, J.-F.;
19 Petron, G.; Dentener, F. J.; Ellingsen, K.; Schultz, M. G.; Wild, O.; Amann, M.;
20 Atherton, C. S.; Bergmann, D. J.; Bey, L; Butler, T.; Cofala, J.; Collins, W. J.; Derwent,
21 R. G.; Doherty, R. M.; Drevet, J.; Eskes, H. J.; Fiore, A. M.; Gauss, M.; Hauglustaine, D.
22 A.; Horowitz, L. W.; Isaksen, I. S. A.; Lawrence, M. G.; Montanaro, V.; Muller, J.-F.;
23 Pitari, G.; Prather, M. J.; Pyle, J. A.; Rast, S.; Rodriguez, J. M.; Sanderson, M. G.;
24 Savage, N. H.; Strahan, S. E.; Sudo, K.; Szopa, S.; Unger, N.; Van Noije, T. P. C.; Zeng,
25 G. (2006) Multimodel simulations of carbon monoxide: comparison with observations
26 and projected near-future changes. J. Geophys. Res. [Atmos.] 111(D19306):
27 10.1029/2006JD007100.
28 Siegwolf, R. T. W.; Matyssek, R.; Sauer, M.; Maurer, S.; Gunthardt-Goerg, M. S.; Schmutz, P.;
29 Bucher, J. B. (2001) Stable isotope analysis reveals differential effects of soil nitrogen
30 and nitrogen dioxide on the water use efficiency in hybrid poplar leaves. New Phytol.
31 149:233-246.
32 Sillman, S. (1995) The use of NOy, H2O2 and HNOs as indicators for ozone-NOx-hydrocarbon
33 sensitivity in urban locations. J. Geophys. Res. [Atmos.] 100: 14,175-14,188.
34 Sillman, S. (2000) Ozone production efficiency and loss of NOx in power plant plumes:
35 photochemical model and interpretation of measurements in Tennessee. J. Geophys. Res.
36 [Atmos.] 105: 9189-9202.
37 Sillman, S.; He, D.-Y. (2002) Some theoretical results concerning O3-NOX-VOC chemistry and
38 NOX-VOC indicators. J. Geophys. Res. [Atmos.] 107: 10.1029/2001JDOO1123.
39 Sillman, S.; Al-Wali, K. L; Marsik, F. J.; Nowacki, P.; Samson, P. J.; Rodgers, M. O.; Garland,
40 L. J.; Martinez, J. E.; Stoneking, C.; Imhoff, R.; Lee, J.-H.; Newman, L.; Weinstein-
41 Lloyd, J.; Aneja, V. P. (1995) Photochemistry of ozone formation in Atlanta, GA—
42 models and measurements. Atmos. Environ. 29: 3055-3066.
March 2008 AX2-160 DRAFT-DO NOT QUOTE OR CITE
-------
1 Sillman, S.; He, D.; Cardelino, C.; Imhoff, R. E. (1997) The use of photochemical indicators to
2 evaluate ozone-NOx-hydrocarbon sensitivity: case studies from Atlanta, New York, and
3 Los Angeles. J. Air Waste Manage. Assoc. 47: 1030-1040.
4 Sillman, S.; He, D.; Pippin, M. R.; Daum, P. H.; Imre, D. G.; Kleinman, L. I; Lee, J. H.;
5 Weinstein-Lloyd, J. (1998) Model correlations for ozone, reactive nitrogen, and
6 peroxides for Nashville in comparison with measurements: implications for O3-NOX-
7 hydrocarbon chemistry. J. Geophys. Res. [Atmos.] 103: 22,629-22,644.
8 Sillman, S.; Vautard, R.; Menut, L.; Kley, D. (2003) O3-NOX-VOC sensitivity and NOX-VOC
9 indicators in Paris: results from models and atmospheric pollution over the Paris area
10 (ESQUIF) measurements. J. Geophys. Res. [Atmos.] 108: 10.1029/2002JDOO1561.
11 Simpson, A. J.; Lam, B.; Diamond, M. L.; Donaldson, D. J.; Lefebvre, B. A.; Moser, A. Q.;
12 Williams, A. J.; Larin, N.; Kvasha, M. P. (2006) Assessing the organic composition of
13 urban surface films using nuclear magnetic resonance spectroscopy. Chemosphere 63:
14 142-152.
15 Singh, H. B.; Kasting, J. F. (1988) Chlorine-hydrocarbon photochemistry in the marine
16 troposphere and lower stratosphere. J. Atmos. Chem. 7: 261-285.
17 Singh, H. B.; Herlth, D.; Kolyer, R.; Salas, L.; Bradshaw, J. D.; Sandholm, S. T.; Davis, D. D.;
18 Crawford, J.; Kondo, Y.; Koike, M.; Talbot, R.; Gregory, G. L.; Sachse, G. W.; Browell,
19 E.; Blake, D. R.; Rowland, F. S.; Newell, R.; Merrill, J.; Heikes, B.; Liu, S. C.; Crutzen,
20 P. J.; Kanakidou, M. (1996) Reactive nitrogen and ozone over the western Pacific:
21 distribution, partitioning, and sources. J. Geophys. Res. [Atmos.] 101: 1793-1808.
22 Skiba, U.; Fowler, D.; Smith, K. A. (1997) Nitric oxide emissions from agricultural soils in
23 temperate and tropical climates: sources, controls and mitigation options. Nutr. Cycling
24 Agroecosyst. 48: 139-153.
25 Smith, A. M.; Keene, W. C.; Maben, J. R.; Pszenny, A. A. P.; Fischer, E.; Stohl, A. (2007)
26 Ammonia sources, transport, transformation, and deposition in coastal New England
27 during summer. J. Geophys. Res. [Atmos.] 112(D10S08): 10.1029/2006JD007574.
28 Solomon, P. A.; Salmon, L. G.; Fall, T.; Cass, G. R. (1992) Spatial and temporal distribution of
29 atmospheric nitric acid and particulate nitrate concentrations in the Los Angeles area.
30 Environ. Sci. Technol. 26: 1596-1601.
31 Sparks, J. P.; Roberts, J. M.; Monson, R. K. (2003) The uptake of gaseous organic nitrogen by
32 leaves: a significant global nitrogen transfer process. Geophys. Res. Lett. 30(23):
33 10.1029/2003GL018578.
34 Spurny, K. R. (1999) Trends in the chemical analysis of aerosols. In: Spurny, K. R., ed.
35 Analytical chemistry of aerosols. Boca Raton, FL: Lewis Publishers; pp. 23-48.
36 Staffelbach, T.; Neftel, A.; Blatter, A.; Gut, A.; Fahrni, M.; Stahelin, J.; Prevot, A.; Hering, A.;
37 Lehning, M.; Neininger, B.; Baumle, M.; Kok, G. L.; Dommen, J.; Hutterli, M.; Anklin,
38 M. (1997) Photochemical oxidant formation over southern Switzerland 1. results from
39 summer 1994. J. Geophys. Res. [Atmos.] 102: 23,345-23,362.
40 Staudt, A. C.; Jacob, D. J.; Ravetta, F.; Logan, J. A.; Bachiochi, D.; Krishnamurti, T. N.;
41 Sandholm, S. T.; Ridley, B. A.; Singh, H. B.; Talbot, B. (2003) Sources and chemistry of
March 2008 AX2-161 DRAFT-DO NOT QUOTE OR CITE
-------
1 nitrogen oxides over the tropical Pacific. J. Geophys. Res. [Atmos.] 108(8239):
2 10.1029/2002JD002139.
3 Stedman, D. H.; Bishop, G.; Peterson, J. E.; Guenther, P. L. (1991) On-road CO remote sensing
4 in the Los Angeles Basin: final report. Sacramento, CA: California Air Resources Board,
5 ARE Contract No. A932-189.
6 Stehr, J. W.; Dickerson, R. R.; Hallock-Waters, K. A.; Doddridge, B. G.; Kirk, D. (2000)
7 Observations of NOy , CO, and SO2 and the origin of reactive nitrogen in the eastern
8 United States. J. Geophys. Res. [Atmos.] 105: 3553-3563.
9 Stein, A. F.; Lamb, D. (2003) Empirical evidence for the low- and high-NOx photochemical
10 regimes of sulfate and nitrate formation. Atmos. Environ. 37: 3615-3625.
11 Stein, A. F.; Lamb, D.; Draxler, R. R. (2000) Incorporation of detailed chemistry into a three-
12 dimensional Lagrangian-Eulerian hybrid model: application to regional tropospheric
13 ozone. Atmos. Environ. 34: 4361-4372.
14 Stemmler, K.; Ammann, M.; Bonders, C.; Kleffmann, J.; George, C. (2006) Photosensitized
15 reduction of nitrogen dioxide on humic acid as a source of nitrous acid. Nature 440: 195-
16 198.
17 Stenchikov, G.; Dickerson, R.; Pickering, K.; Ellis, W. Jr.; Doddridge, B.; Kondragunta, S.;
18 Poulida, O.; Scala, J.; Tao, W.-K. (1996) Stratosphere-troposphere exchange in a
19 midlatitude mesoscale convective complex. 2. Numerical simulation. J. Geophys. Res.
20 [Atmos.] 101:6837-6851.
21 Stevens, R. K.; O'Keeffe, A. E.; Ortman, G. C. (1969) Absolute calibration of a flame
22 photometric detector to volatile sulfur compounds at sub-part-per-million levels. Environ.
23 Sci. Technol. 3:652-655.
24 Stevens, R. K.; Mulik, J. D.; O'Keefe, A. E.; Krost, K. J. (1971) Gas chromatography of reactive
25 sulfur gases in air at the parts-per-billion level. Anal. Chem. 43: 827-831.
26 Stevenson, D.; Dentener, F. J.; Schultz, M. G.; Ellingsen, K.; Van Noije, T. P. C.; Wild, O.;
27 Zeng, G.; Amann, M.; Atherton, C. S.; Bell, N.; Bergmann, D. J.; Bey, I; Butler, T.;
28 Cofala, J.; Collins, W. J.; Derwent, R. G.; Doherty, R. M.; Drevet, J.; Eskes, H. J.; Fiore,
29 A. M.; Gauss, M.; Hauglustaine, D. A.; Horowitz, L. W.; Isaksen, I. S. A.; Krol, M. C.;
30 Lamarque, J.-F.; Lawrence, M. G.; Montanaro, V.; Miiller, J.-F.; Pitari, G.; Prather, M. J.;
31 Pyle, J. A.; Rast, S.; Rodriguez, J. M.; Sanderson, M. G. (2006) Multimodel ensemble
32 simulations of present-day and near-future tropospheric ozone. J. Geophys. Res. [Atmos.]
33 111(D08301): 10.1029/2005JD006338.
34 Steyn, D. J.; Bottenheim, J. W.; Thomson, R. B. (1997) Overview of tropospheric ozone in the
35 Lower Fraser Valley, and the Pacific '93 field study. Atmos. Environ. 31: 2025-2036.
36 Stith, J.; Dye, J.; Ridley, B.; Laroche, P.; Defer, E.; Baumann, K.; Hubler, G.; Zerr, R.;
37 Venticinque, M. (1999) NO signatures from lightning flashes. J. Geophys. Res. [Atmos.]
38 101: 16081-16089.
39 Stockwell, W. R.; Middleton, P.; Chang, J. S.; Tang, X. (1990) The second generation Regional
40 Acid Deposition Model chemical mechanism for regional air quality modeling. J.
41 Geophys. Res. [Atmos.] 95: 16,343-16,367.
March 2008 AX2-162 DRAFT-DO NOT QUOTE OR CITE
-------
1 Stolzenburg, M. R.; Hering, S. V. (2000) Method for the automated measurement of fine particle
2 nitrate in the atmosphere. Environ. Sci. Technol. 34: 907-914.
3 Streets, D.; Bond, T. C.; Carmichael, G. R.; Fernandes, S. D.; Fu, Q.; He, D.; Klimont, Z.;
4 Nelson, S. M.; Tsai, N. Y.; Wang, M. Q.; Woo, J.-H.; Yarber, K. F. (2003) An inventory
5 of gaseous and primary aerosol emissions in Asia in the year 2000. J. Geophys. Res.
6 [Atmos.] 108(D21): 10.1029/2002JD003093.
7 Stutz, J.; Hebestreit, K.; Alicke, B.; Platt, U. (1999) Chemistry of halogen ozides in the
8 troposphere: comparison of model calculations with recent field data. J. Atmos. Chem.
9 34: 65-85.
10 Stutz, J.; Kim, E. S.; Platt, U.; Bruno, P.; Perrino, C.; Febo, A. (2000) UV-visible absorption
11 cross sections of nitrous acid. J. Geophys. Res. [Atmos.] 105: 14,585-14,592.
12 Stutz, J.; Ackermann, R.; Fast, J. D.; Barrie, L. (2002) Atmospheric reactive chlorine and
13 bromine at the Great Salt Lake, Utah. Geophys. Res. Lett. 29: 10.1029/2002GL014812.
14 Stutz, J.; Alicke, B.; Ackermann, R.; Geyer, A.; White, A.; Williams, E. (2004a) Vertical profiles
15 of NOs, N2Os, 63, and NOX in the nocturnal boundary layer: 1. Observations during the
16 Texas Air Quality Study 2000. J. Geophys. Res. [Atmos.] 109(D12306):
17 10.1029/2003 JD004209.
18 Stutz, J.; Alicke, B.; Ackermann, R.; Geyer, A.; Wang, S.; White, A. B.; Williams, E. J.; Spicer,
19 C. W.; Fast, J. D. (2004b) Relative humidity dependence of HONO chemistry in urban
20 areas. J. Geophys. Res. [Atmos.] 109: 10.1029/2003JD004135.
21 Suh, H. H.; Spengler, J. D.; Koutrakis, P. (1992) Personal exposures to acid aerosols and
22 ammonia. Environ. Sci. Technol. 26: 2507-2517.
23 Suh, H. H.; Koutrakis, P.; Spengler, J. D. (1994) The relationship between airborne acidity and
24 ammonia in indoor environments. J. Exposure Anal. Environ. Epidemiol. 4: 1-23.
25 Sutton, M. A.; Dragosits, U.; Tang, Y. S.; Fowler, D. (2000) Ammonia emissions from non-
26 agricultural sources in the UK. Atmos. Environ. 34: 855-869.
27 Svensson, R.; Ljungstroem, E.; Lindqvist, O. (1987) Kinetics of the reaction between nitrogen
28 dioxide and water vapour. Atmos. Environ. 21: 1529-1539.
29 Talbot, R. W.; Vijgen, A. S.; Harriss, R. C. (1990) Measuring tropospheric HNO3: problems and
30 prospects for nylon filter and mist chamber techniques. J. Geophys. Res. [Atmos.] 95:
31 7553-7561.
32 Tanaka, P. L.; Riemer, D. D.; Chang, S.; Yarwood, G.; McDonald-Buller, E. C.; Apel, E. C.;
33 Orlando, J. J.; Silva, P. J.; Jimenez, J. L.; Canagaratna, M. R.; Neece, J. D.; Mullins, C.
34 B.; Allen, D. T. (2003) Direct evidence for chlorine-enhanced urban ozone formation in
35 Houston, Texas. Atmos. Environ. 37: 1393-1400.
36 Tanner, R. L.; D'Ottavio, T.; Garber, R.; Newman, L. (1980) Determination of ambient aerosol
37 sulfur using a continuous flame photometric detection system. I. Sampling system for
38 aerosol sulfate and sulfuric acid. Atmos. Environ. 14: 121-127.
March 2008 AX2-163 DRAFT-DO NOT QUOTE OR CITE
-------
1 Tanner, R. L.; Kelly, T. J.; Dezaro, D. A.; Forrest, J. (1989) A comparison of filter, denuder, and
2 real-time chemiluminescence techniques for nitric acid determination in ambient air.
3 Atmos. Environ. 23: 2213-2222.
4 Tao, W.-K.; Simpson, J. (1993) The Goddard Cumulus Ensemble Model. Part I: model
5 description. Terr. Atmos. Oceanic Sci. (TAO) 4: 35-71.
6 Teklemariam, T. A.; Sparks, J. P. (2004) Gaseous fluxes of peroxyacetyl nitrate (PAN) into plant
7 leaves. Plant Cell Environ. 27: 1149-1158.
8 Teklemariam, T. A.; Sparks, J. P. (2006) Leaf fluxes of NO and NO2 in four herbaceous plant
9 species: the role of ascorbic acid. Atmos. Environ. 40: 2235-2244.
10 Thielman, A.; Prevot, A. S. H.; Griiebler, F. C.; Staelhelin, J. (2001) Empirical ozone isopleths
11 as a tool to identify ozone production regimes. Geophys. Res. Lett. 28: 2369-2372.
12 Thompson, A. M.; Pickering, K. E.; Dickerson, R. R.; Ellis, W. G., Jr.; Jacob, D. J.; Scala, J. R.;
13 Tao, W.-K.; McNamara, D. P.; Simpson, J. (1994) Convective-transport over the central
14 United States and its role in regional CO and O3 budgets. J. Geophys. Res. [Atmos.] 99:
15 18,703-18,711.
16 Thompson, A. M.; Singh, H. B.; Schlager, H. (2000) Subsonic assessment ozone and nitrogen
17 oxide experiment (SONEX) and pollution from aircraft emissions in the North Atlantic
18 Flight Corridor (POLINAT 2). J. Geophys. Res. [Atmos.] 105: 3595-3603.
19 Thornton, D. C.; Bandy, A. R. (1993) Sulfur dioxide and dimethyl sulfide in the central Pacific
20 troposphere. J. Atmos. Chem. 17: 1-13.
21 Thornton, D. C.; Bandy, A. R.; Blomquist, B. W.; Anderson, B. E. (1996) Impact of
22 anthropogenic and biogenic sources and sinks on carbonyl sulfide in the North Pacific
23 troposphere. J. Geophys. Res. [Atmos.] 101: 1873-1881.
24 Thornton, D. C.; Bandy, A. R.; Blomquist, B. W.; Driedger, A. R.; Wade, T. P. (1999) Sulfur
25 dioxide distribution over the Pacific Ocean 1991-1996. J. Geophys. Res. [Atmos.] 104:
26 5845-5854.
27 Thornton, D. C.; Bandy, A. R.; Tu, F. H.; Blomquist, B. W.; Mitchell, G. M.; Nadler, W.;
28 Lenschow, D. H. (2002) Fast airborne sulfur dioxide measurements by atmospheric
29 pressure ionization mass spectrometry (APIMS). J. Geophys. Res. [Atmos.] 107(D22):
30 10.1029/2002JD02289.
31 Thornton, J. A.; Braban, C. F.; Abbatt, J. P. D. (2003) ^Os hydrolysis on sub-micron organic
32 aerosol: the effect of relative humidity, particle phase, and particle size. Phys. Chem.
33 Chem. Phys. 5: 4593-4603.
34 Tie, X. X.; Emmons, L.; Horowitz, L.; Brasseur, G.; Ridley, ,B.; Atlas, E.; Stround, C.; Hess, P.;
35 Klonecki, A.; Madronich, S.; Talbot, R.; Dibb, J. (2003) Effect of sulfate aerosol on
36 tropospheric NOX and ozone budgets: model simulations and TOPSE evidence. J.
37 Geophys. Res. [Atmos.] 108(D4): 10.1029/2001JD001508.
38 Tokiwa, H.; Ohnishi, Y. (1986) Mutagenicity and carcinogenicity of nitroarenes and their
39 sources in the environment. Crit. Rev. Toxicol. 17: 23-60.
March 2008 AX2-164 DRAFT-DO NOT QUOTE OR CITE
-------
1 Tokiwa, H.; Nakanishi, Y.; Sera, N.; Kara, N.; Inuzuka, S. (1998) Analysis of environmental
2 carcinogens associated with the incidence of lung cancer. Toxicol. Lett. 99: 33-41.
3 Tonnesen, G. S.; Dennis, R. L. (2000) Analysis of radical propagation efficiency to assess ozone
4 sensitivity to hydrocarbons and NOX: 2. Long-lived species as indicators of ozone
5 concentration sensitivity. J. Geophys. Res. [Atmos.] 105: 9227-9241.
6 Toumi, R. (1994) BrO as a sink for dimethylsulfide in the marine atmosphere. Geophys. Res.
7 Lett. 21: 117-120.
8 Trainer, M.; Parrish, D. D.; Buhr, M. P.; Norton, R. B.; Fehsenfeld, F. C.; Anlauf, K. G.;
9 Bottenheim, J. W.; Tang, Y. Z.; Wiebe, H. A.; Roberts, J. M.; Tanner, R. L.; Newman,
10 L.; Bowersox, V. C.; Meagher, J. F.; Olszyna, K. J.; Rodgers, M. O.; Wang, T.;
11 Berresheim, H.; Demerjian, K. L.; Roychowdhury, U. K. (1993) Correlation of ozone
12 with NOy in photochemically aged air. J. Geophys. Res. [Atmos.] 98: 2917-2925.
13 Trainer, M.; Parrish, D. D.; Golday, P. D.; Roberts, J.; Fehsenfeld, F. C. (2000) Review of
14 observation-based analysis of the regional factors influencing ozone concentrations.
15 Atmos. Environ. 34: 2045-2061.
16 Treves, K.; Shragina, L.; Rudich, Y. (2000) Henry's Law constants of some P-, y-, and 5-hydroxy
17 alkyl nitrates of atmospheric interest. Environ. Sci. Technol. 34: 1197-1203.
18 Tsai, C.-J.; Huang, H.-Y. (1995) Atmospheric aerosol sampling by an annular denuder system
19 and a high-volume PMio sampler. Environ. Int. 21: 283-291.
20 Turco, R. P.; Toon, O. B.; Whitten, R. C.; Hamill, P.; Keesee, R. G. (1983) The 1980 eruptions
21 of Mount St. Helens: physical and chemical processes in the stratospheric clouds. J.
22 Geophys. Res. C: Oceans Atmos. 88: 5299-5319.
23 Turekian, V. C.; Macko, S. A.; Keene, W. C. (2001) Application of stable sulfur isotopes to
24 differentiate sources of size-resolved paniculate sulfate in polluted marine air at Bermuda
25 during spring. Geophys. Res. Lett. 28: 1491-1494.
26 Turnipseed, A. A.; Huey, L. G.; Nemitz, E.; Stickel, R.; Higgs, J.; Tanner, D. J.; Slusher, D. L.;
27 Sparks, J. P.; Flocke, F.; Guenther, A. (2006) Eddy covariance fluxes of peroxyacetyl
28 nitrates (PANs) and NOy to a coniferous forest. J. Geophys. Res. [Atmos.] 111(D09304):
29 10.1029/2005 JD006631.
30 United Kingdom Air Quality Expert Group (U.K. AQEG). (2004) Nitrogen dioxide in the United
31 Kingdom. London, United Kingdom: Department for Environment, Food and Rural
32 Affairs. Available:
33 http://www.defira.gov.uk/environment/airquality/panels/aqeg/index.htm [12 April, 2007].
34 U.S. Environmental Protection Agency. (1991) Guideline for regulatory application of the urban
35 airshed model. Research Triangle Park, NC: Office of Air Quality Planning and
36 Standards; report no. EPA-450/4-91-013. Available from: NTIS, Springfield, VA; PB92-
37 108760.
38 U.S. Environmental Protection Agency. (1993) Ambient air quality surveillance. Final rule. F. R.
39 (February 12): 8452-8475.
40 U.S. Environmental Protection Agency. (1996) Air quality criteria for ozone and related
41 photochemical oxidants. Research Triangle Park, NC: Office of Research and
March 2008 AX2-165 DRAFT-DO NOT QUOTE OR CITE
-------
1 Development; report nos. EPA/600/AP-93/004aF-cF. 3v. Available from: NTIS,
2 Springfield, VA; PB96-185582, PB96-185590, and PB96-185608. Available:
3 http ://cfpub2. epa. gov/ncea/.
4 U.S. Environmental Protection Agency. (1997) National air pollutant emission trends 1990-1996.
5 Research Triangle Park, NC: Office of Air Quality Planning and Standards; report no.
6 EPA/454/R-97/011.
7 U.S. Environmental Protection Agency. (1999) Getting started: emissions inventory methods for
8 PM-2.5. Research Triangle Park, NC: Office of Air Quality Planning and Standards,
9 Emission Factor and Inventory Group; EIIP document series - volume IX. Available:
10 http://www.epa.gov/ttn/chief/eiip/techreport/volume09/index.html [7 May 2003].
11 U.S. Environmental Protection Agency. (2000) Air quality criteria for carbon monoxide.
12 Research Triangle Park, NC: National Center for Environmental Assessment; report no.
13 EPA/600/P-99/001F. Available: http://www.epa.gov/ncea/pdfs/coaqcd.pdf [19 April
14 2007].
15 U.S. Environmental Protection Agency. (2004) Air quality criteria for particulate matter.
16 Research Triangle Park, NC: National Center for Environmental Assessment; report no.
17 EPA/600/P-99/002aF-bF. 2v. Available: http://cfpub.epa.gov/ncea/ [9 November, 2004].
18 U.S. Environmental Protection Agency. (2005) Technical assistance document (TAD) for
19 precursor gas measurements in the NCore multi-pollutant monitoring network. Version 4.
20 Research Triangle Park, NC: Office of Air Quality Planning and Standards; report no.
21 EPA-454/R-05-003.
22 U.S. Environmental Protection Agency. (2006) Air quality criteria for ozone and related
23 photochemical oxidants. Research Triangle Park, NC: National Center for Environmental
24 Assessment; report no. EPA/600/R-05/004aF-cF. 3v. Available:
25 http://cfpub.epa.gov/ncea/ [24 March, 2006].
26 Van Aardenne, J. A.; Dentener, F. J.; Olivier, J. G. J.; Klein Goldewijk, C. G. M.; Lelieveld, J.
27 (2001) A 1° x 1° resolution data set of historical anthropogenic trace gas emissions for
28 the period 1980-1990. Global Biogeochem. Cycles 15: 909-928.
29 Van der Werf, G. R.; Randerson, J. T.; Collatz, J.; Giglio, L. (2003) Carbon emissions from fires
30 in tropical and subtropical ecosystems. Glob. Change Biol. 9: 547-562.
31 Van Noije, T. P. C.; Eskes, H. J.; Dentener, F. J.; Stevenson, D. S.; Ellingsen, K.; Schultz, M. G.;
32 Wild, O.; Amann, M.; Atherton, C. S.; Bergmann, D. J.; Bey, I; Boersma, K. F.; Butler,
33 T.; Cofala, J.; Drevet, J.; Fiore, A. M.; Gauss, M.; Hauglustaine, D. A.; Horowitz, L. W.;
34 Isaksen, I. S. A.; Krol, M. C.; Lamarque, J.-F.; Lawrence, M. G.; Martin, R. V.;
35 Montanaro, V.; Muller, J.-F.; Pitari, G.; Prather, M. J.; Pyle, J. A.; Richter, A.;
36 Rodriguez, J. M.; Savage, N. H.; Strahan, S. E.; Sudo, K.; Szopa, S.; Van Roozendael, M.
37 (2006) Multi-model ensemble simulations of tropospheric NO2 compared with GOME
38 retrievals for the year 2000. Atmos. Chem. Phys. Discuss. 6: 2965-3047.
39 Vautard, R.; Martin, D.; Beekman, M.; Drobinski, P.; Friedrich, R.; Jaubertie, A.; Kley, D.;
40 Lattuati, M.; Moral, P.; Neininger, B.; Theloke, J. (2002) Paris emission inventory
41 diagnostics from the ESQUIF airborne measurements and a chemistry transport model. J.
42 Geophys. Res. [Atmos.] 108(D17): 10.1029/2002JD002797.
March 2008 AX2-166 DRAFT-DO NOT QUOTE OR CITE
-------
1 Vione, D.; Barra, S.; De Gennaro, G.; De Rienzo, M.; Gilardoni, S.; Perrone, M. G.; Pozzoli, L.
2 (2004) Polycyclic aromatic hydrocarbons in the atmosphere: monitoring, sources, sinks
3 and fate. II: sinks and fate. Ann. Chim. 94: 257-268.
4 Vogt, R.; Crutzen, P. J.; Sander, R. (1996) A mechanism for halogen release from sea-salt
5 aerosol in the remote marine boundary layer. Nature (London, U.K.) 383: 327-330.
6 Vogt, R.; Sander, R.; Von Glasow, R.; Crutzen, P. J. (1999) Iodine chemistry and its role in
7 halogen activation and ozone loss in the marine boundary layer: a model study. J. Atmos.
8 Chem. 32: 375-395.
9 Volz-Thomas, A.; Geiss, H.; Hofzumahaus, A.; Becker, K.-H. (2003) Introduction to special
10 section: photochemistry experiment in BERLIOZ. J. Geophys. Res. [Atmos.] 108(D4):
11 10.1029/JD002029.
12 Von Glasow, R.; Sander, R.; Bott, A.; Crutzen, P. J. (2002a) Modeling halogen chemistry in the
13 marine boundary layer. 1. Cloud-free MBL. J. Geophys. Res. [Atmos.] 107(D17):
14 10.1029/2001JD000942.
15 Von Glasow, Sander, R.; Bott, A.; Crutzen, P. J. (2002b) Modeling halogen chemistry in the
16 marine boundary layer. 2. Interactions with sulfur and cloud-covered MBL. J. Geophys.
17 Res. [Atmos.] 107(D17): 10.1029/2001JD000943.
18 Von Glasow, R.; Von Kuhlmann, R.; Lawrence, M. G.; Platt, U.; Crutzen, P. J. (2004) Impact of
19 reactive bromine chemistry in the troposphere. Atmos. Chem. Phys. 4: 2481-2497.
20 Wagner, T.; Leue, C.; Wenig, M.; Pfeilsticker, K.; Platt, U. (2001) Spatial and temporal
21 distribution of enhanced boundary layer BrO concentrations measured by the GOME
22 instrument aboard ERS-2. J. Geophys. Res. [Atmos.] 106: 24,225-24,235.
23 Walcek, C. J.; Taylor, G. R. (1986) A theoretical method for computing vertical distributions of
24 acidity and sulfate production within cumulus clouds. J. Atmos. Sci. 43: 339-355.
25 Walcek, C. J.; Stockwell, W. R.; Chang, J. S. (1990) Theoretical estimates of the dynamic,
26 radiative and chemical effects of clouds on tropospheric trace gases. Atmos. Res. 25: 53-
27 69.
28 Walega, J. G; Stedman, D. H.; Shelter, R. E.; Mackay, G. I; Iguchi, T.; Schiff, H. I. (1984)
29 Comparison of a chemiluminescent and a tunable diode laser absorption technique for the
30 measurement of nitrogen oxide, nitrogen dioxide, and nitric acid. Environ. Sci. Technol.
31 18:823-826.
32 Wang, Y.; Tao, W.-K.; Pickering, K. E.; Thompson, A. M.; Kain, J. S.; Adler, R. F.; Simpson, J.;
33 Keehn, P. R.; Lai, G. S. (1996) Mesoscale model simulations of TRACE A and
34 preliminary regional experiment for storm-scale operational and research meteorology
35 convective systems and associated tracer transport. J. Geophys. Res. [Atmos.] 101:
36 24,013-24,027.
37 Wang, Y.; DeSilva, A. W.; Goldenbaum, G. C.; Dickerson, R. R. (1998) Nitric oxide production
38 by simulated lightning: dependence on current, energy, and pressure. J. Geophys. Res.
39 [Atmos.] 103: 19,149-19,159.
March 2008 AX2-167 DRAFT-DO NOT QUOTE OR CITE
-------
1 Wang, L. H.; Milford, J. B.; Carter, W. P. L. (2000a) Reactivity estimates for aromatic
2 compounds. Part I: uncertainty in chamber-derived parameters. Atmos. Environ. 34:
3 4337-4348.
4 Wang, L. H.; Milford, J. B.; Carter, W. P. L. (2000b) Reactivity estimates for aromatic
5 compounds. Part 2. uncertainty in incremental reactivities. Atmos. Environ. 4349-4360.
6 Warneck, P. (1999) The relative importance of various pathways for the oxidation of sulfur
7 dioxide and nitrogen dioxide in sunlit continental fair weather clouds. Phys. Chem.
8 Chem. Phys. 1: 5471-5483.
9 Watson, J. G.; Fujita, E. M.; Chow, J. C.; Zielinska, B.; Richards, L. W.; Neff, W.; Dietrich, D.
10 (1998) Northern front range air quality study. Final report. Fort Collins, CO: Colorado
11 State University, Cooperative Institute for Research in the Atmosphere. Available:
12 http://www.nfraqs.colostate.edu/index2.html (16 Jan 2002).
13 Weber, P.; NuBbaum, S.; Fuhrer, J.; Gfeller, H.; Schlunegger, U. P.; Brunold, C.; Rennenberg,
14 H. (1995) Uptake of atmospheric 15NC>2 and its incorporation into free amino-acids in
15 wheat (Triticum-aestivuni). Physiol. Plant. 94: 11-11.
16 Weber, P.; Thoene, B.; Rennenberg, H. (1998) Absorption of atmospheric NO2 by spruce (Picea
17 abies) trees. III. Interaction with nitrate reductase activity in the needles and phloem
18 transport. Bot. Acta 111: 377-382.
19 Wedin, D. A.; Tilman, D. (1996) Influence of nitrogen loading and species composition on the
20 carbon balance of grasslands. Science (Washington, DC) 274: 1720-1723.
21 Wendel, G. J.; Stedman, D. H.; Cantrell, C. A.; Damrauer, L. (1983) Luminol-based nitrogen
22 dioxide detector. Anal. Chem. 55: 937-940.
23 Wesely, M. L. (1989) Parameterization of surface resistances to gaseous dry deposition in
24 regional-scale numerical models. Atmos. Environ. 23: 1293-1304.
25 Wesely, M. L.; Hicks, B. B. (1977) Some factors that affect the deposition rates of sulfur dioxide
26 and similar gases on vegetation. J. Air Pollut. Control Assoc. 27: 1110-1116.
27 Wesely, M. L.; Hicks, B. B. (2000) A review of the current status of knowledge on dry
28 deposition. Atmos. Environ. 34: 2261-2282.
29 Westerling, A. L.; Hildalgo, H. G.; Cayan, D. R.; Swetnam, T. W. (2006) Warming and earlier
30 spring increase western U.S. forest wildfire activity. Science (Washington, DC, U.S.)
31 313:940-943.
32 Williams, E. J.; Guenther, A.; Fehsenfeld, F. C. (1992) An inventory of nitric oxide emissions
33 from soils in the United States. J. Geophys. Res. [Atmos.] 97: 7511-7519.
34 Winberry, W. T., Jr.; Ellestad, T.; Stevens, B. (1999) Compendium method IO-4.2:
35 determination of reactive acidic and basic gases and strong acidity of atmospheric fine
36 particles (< 2.5 um). Compendium of methods for the determination of inorganic
37 compounds in ambient air. Cincinnati, OH: U.S. Environmental Protection Agency,
38 Center for Environmental Research Information; report no. EPA/625/R-96/010a.
39 Available: http://www.epa.gov/ttn/amtic/files/ambient/inorganic/mthd-4-2.pdff2 May,
40 2007].
March 2008 AX2-168 DRAFT-DO NOT QUOTE OR CITE
-------
1 Witz, S.; Eden, R. W.; Wadley, M. W.; Dunwoody, C.; Papa, R.; Torre, K. J. (1990) Rapid loss
2 of particulate nitrate, chloride and ammonium on quartz fiber filters during storage. J. Air
3 Waste Manage. Assoc. 40: 53-61.
4 World Health Organization (WHO). (2003) Nitrogenated polycyclic aromatic hydrocarbons.
5 Geneva, Switzerland: World Health Organization. (Environmental Health Criteria 229).
6 Xu, J. H.; Lee, F. S. C. (2000) Quantification of nitrated polynuclear aromatic hydrocarbons in
7 atmospheric parti culate matter. Anal. Chim. Acta 416: 111-115.
8 Yienger, J. J.; Levy, H., II. (1995) Empirical model of global soil-biogenic NOX emissions. J.
9 Geophys. Res. [Atmos.] 100: 11,447-11,464.
10 Young, V. L.; Kieser, B. N.; Chen, S. P.; Niki, H. (1997) Seasonal trends and local influences on
11 nonmethane hydrocarbon concentrations in the Canadian boreal forest. J. Geophys. Res.
12 [Atmos.] 102: 5913-5918.
13 Zafiriou, O. C.; True, M. B. (1979) Nitrate photolysis in seawater by sunlight. Mar. Chem. 8: 33-
14 42.
15 Zanis, P.; Trickl, T.; Stohl, A.; Wernli, H.; Cooper, O.; Zerefos, C.; Gaeggeler, H.; Schnabel, C.;
16 Tobler, L.; Kubik, P. W.; Priller, A.; Scheel, H. E.; Kanter, H. J.; Cristofanelli, P.;
17 Forster, C.; James, P.; Gerasopoulos, E.; Delcloo, A.; Papayannis, A.; Claude, H. (2003)
18 Forecast, observation and modelling of a deep stratospheric intrusion event over Europe.
19 Atmos. Chem. Phys. 3: 763-777.
20 Zhang, G. J.; McFarlane, N. A. (1995) Sensitivity of climate simulations to the parameterization
21 of cumulus convection in the Canadian Climate Centre general circulation model. Atmos.
22 Ocean 33: 407-446.
23 Zhang, X. Q.; McMurry, P. H. (1987) Theoretical analysis of evaporative losses from impactor
24 and filter deposits. Atmos. Environ. 21: 1779-1789.
25 Zhang, J.-Z.; Millero, F. J. (1991) The rate of sulfite oxidation in seawater. Geochim.
26 Cosmochim. Acta 55: 677-685.
27 Zhou, X.; Beine, H. J.; Honrath, R. E.; Fuentes, J. D.; Simpson, W.; Shepson, P. B.; Bottenheim,
28 J. W. (2001) Snowpack photochemical production of HONO: a major source of OH in
29 the Arctic boundary layer in springtime. Geophys. Res. Lett. 28: 4087-4090.
30 Zhou, X.; Civerolo, K.; Dai, H. (2002a) Summertime nitrous acid chemistry in the atmospheric
31 boundary layer at a rural site in New York state. J. Geophys. Res. [Atmos.] 107(D21):
32 10.1029/2001JD001539.
33 Zhou, X. L.; He, Y.; Huang, G; Thornberry, T. D.; Carroll, M. A.; Bertman, S. B. (2002b)
34 Photochemical production of nitrous acid on glass sample manifold surface. Geophys.
35 Res. Lett. 29(1681): 10.1029/2002GLO 15080.
36 Zhou, X.; Gao, H.; He, Y.; Huang, G. (2003) Nitric acid photolysis on surfaces in low-NOx
37 environments: significant atmospheric implications. Geophys. Res. Lett. 30:
38 10.1029/2003GL018620.
39 Zielinska, B.; Arey, J.; Atkinson, R.; Ramdahl, T.; Winer, A. M.; Pitts, J. N., Jr. (1986) Reaction
40 of dinitrogen pentoxide with fluoranthene. J. Am. Chem. Soc. 108: 4126-4132.
March 2008 AX2-169 DRAFT-DO NOT QUOTE OR CITE
-------
1 Zielinska, B.; Arey, J.; Atkinson, R.; Winer, A. M. (1989) The nitroarenes of molecular weight
2 247 in ambient particulate samples collected in southern California. Atmos. Environ. 23:
3 223-229.
4 Zielinska, B.; Sagebiel, J.; McDonald, J. D.; Whitney, K.; Lawson, D. R. (2004) Emission rates
5 and comparative chemical composition from selected in-use diesel and gasoline-fueled
6 vehicles. J. Air Waste Manage. Assoc. 54: 1138-1150.
7 Zimmermann, J.; Poppe, D. (1993) Nonlinear chemical couplings in the tropospheric NOX—HOX
8 gas phase chemistry. J. Atmos. Chem. 17: 141-155.
9 Zingler, J.; Platt, U. (2005) Iodine oxide in the Dead Sea Valley: evidence for inorganic sources
10 of boundary layer IO. J. Geophys. Res. [Atmos.] 110(D07307): 10.1029/2004JD004993.
March 2008 AX2-170 DRAFT-DO NOT QUOTE OR CITE
-------
i AX3. CHAPTER 3 ANNEX-AMBIENT
2 CONCENTRATIONS AND EXPOSURES
o
4
5 AX3.1 INTRODUCTION
6 Topics discussed in this chapter include the characterization of ambient air quality for
7 nitrogen dioxide (NO2), the uses of these data in assessing human exposures to NO2;
8 concentrations and sources of NC>2 in different microenvironments, and personal exposures to
9 NC>2. The NC>2 data contained in this chapter are taken mainly from the U.S. Environmental
10 Protection Agency's Air Quality System (AQS) database (formerly the AIRS database) (U.S.
11 Environmental Protection Agency, 2007).
12
13 Characterizing Ambient NO2 Concentrations
14 The "concentration" of a specific air pollutant is typically defined as the amount (mass)
15 of that material per unit volume of air. However, most of the data presented in this chapter are
16 expressed as "mixing ratios" in terms of a volume-to-volume ratio (e.g., parts per million [ppm]
17 or parts per billion [ppb]. Data expressed this way are often referred to as concentrations, both in
18 the literature and in the text, following common usage. Human exposures are expressed in units
19 of mixing ratio times time.
20
21 Relationship to the 1993 Air Quality Criteria Document for Nitrogen Oxides
22 The 1993 AQCD for Oxides of Nitrogen emphasized NC>2 indoor sources (gas stoves)
23 and the relationship between personal total exposure and indoor or outdoor NC>2 concentrations
24 (U.S. Environmental Protection Agency, 1993). At that time, only few personal exposure studies
25 had been conducted with an emphasis on residential indoor NC>2 sources and concentrations.
26 Although the concept of microenvironment had been introduced in the document, NC>2
27 concentrations were seldom reported for microenvironments other than residences. Exposure
28 measurements at that time relied on Palmes tubes and Yanagisawa badges; and exposure-
29 modeling techniques were limited mainly to simple linear regression. In the 1993 AQCD, NC>2
30 was treated as an independent risk factor, and confounding issues were not mentioned in the
31 human environmental exposure chapters.
March 2008 AX3-1 DRAFT-DO NOT CITE OR QUOTE
-------
1 The current chapter summarizes and discusses the state-of-the-science and technology
2 regarding NC>2 human exposures since 1993. Since then, numerous human exposure studies
3 have been conducted with new measurement and modeling techniques. Microenvironmental
4 measurements were not limited to residential indoor environments; NC>2 concentrations were also
5 measured in vehicles, schools and offices, and microenvironments close to traffic. More indoor
6 sources have been identified and more NO2 formation and transformation mechanisms in the
7 indoor environment have been reported. Both indoor and outdoor NC>2 have been treated as
8 components of a pollutant mixture, and therefore the concepts of confounding and surrogacy
9 have been discussed in the current chapter.
10
11
12 AX3.2 AMBIENT CONCENTRATIONS OF NITROGEN OXIDES AND
13 RELATED SPECIES
14 As discussed in Chapter 2, most measurements of NOx are made by instruments that
15 convert NC>2 to NO, which is then measured by chemiluminescence. However, the surface
16 converters that reduce NC>2 to NO also reduce other reactive NOy species. As indicated in
17 Chapter 2, NOy compounds consist of NOx, gas phase inorganic nitrates, such as CINOs; organic
18 nitrates, such as PANs; inorganic acids, given by the formulas HNOy (Y = 2 to 4); and
19 particulate nitrate. In urban areas or in rural areas where there are large local sources, NO and
20 NO2 are expected to be the major forms of NOy. Thus, interference from PANs and other NOy
21 species near sources are expected to minor; in most rural and remote areas, interference may be
22 substantial as concentrations of other NOy species may be much larger than those for NO and
23 NO2 (National Research Council, 1991). Examples will be presented in Section AX3.3.5.
24 Data for NOx in addition to NO2 is reported into the U.S. Environmental Protection
25 Agency's Air Quality System (AQS), but data for NO is not reported, even though measurements
26 of NO are not affected by artifacts caused by products of NO2 oxidation and therefore should be
27 the most reliable. By definition, NOx is equal to the sum of NO and NO2, so the concentration
28 of NO can be found by subtraction. However, measurements are obtained for NO and NOx
29 every 2 to 3 min, but hourly averages for NO2 and NOx are reported into AQS. The locations of
30 NO2 monitoring sites are shown in Figure AX3.2-1. As can be seen from Figure AX3.2-1, there
31 are large areas of the United States for which data for ambient NO2 are not collected. The
March 2008 AX3-2 DRAFT-DO NOT CITE OR QUOTE
-------
Monitor Locator Map - Criteria Air Pollutants
United States
Monitor Location: A NO2 (n=375)
Shaded states have monitors
Figure AX3.2-1. Location of ambient NOi monitors in the United States.
1 percentile distribution of NC>2 concentrations in urban and nonurban areas in the U.S. for
2 different averaging periods is shown in Table AX3.2-1.
3 Because of their short lifetime with respect to oxidation to PANs and HNOs, NOx
4 concentrations are highly spatially and temporally variable. Average concentrations range from
5 tens of ppt in remote areas of the globe to tens of ppb in urban cores, i.e., by three orders of
6 magnitude. Median NO, NOx, and NOy concentrations at the surface are typically below 0.01,
7 0.05, and 0.3 ppb, respectively, in remote areas such as Alaska, northern Canada, and the eastern
8 Pacific; median NOy concentrations range from about 0.7 to about 4.3 ppb at regional
9 background sites in the eastern United States (Emmons et al., 1997). Note that the last two
10 values, especially, contain a substantial contribution from pollution. Maximum short-term
11 average (1-h) NOx concentrations near heavy traffic (e.g., in Los Angeles, CA) approach 1 ppm,
12 but these levels decrease rapidly away from sources. Even at sites where such high hourly
13 values are found, 24-h average concentrations are much lower. For example, the maximum 24-h
14 average NOx concentration at any site in Los Angeles in 2004 was 82 ppb.
March 2008
AX3-3
DRAFT-DO NOT CITE OR QUOTE
-------
1 NO2 concentrations are likewise highly spatially and temporally variable. The overall
2 annual mean concentration of NO2 at U.S. monitoring sites is about 15 ppb. Most sites
3 monitoring NO2 are located in populated areas and values outside of urban and suburban areas
4 can be much lower. Perhaps the most comprehensive characterization of ambient NO2 levels has
5 been performed by the California Air Resources Board (CARB) as part of the review of the air
6 quality standards for California (CARB, 2007). On a statewide basis, the average NO2
7 concentration was about 15 ppb from 2002 to 2004. Highest average values of about 27 ppb
8 were found in the South Coast Air Basin. The maximum 1-h average NO2 concentration during
9 the same period was 262 ppb, again in the South Coast Air Basin. However, maximum 1-h
10 concentrations of NO2 were about 150 ppb in Los Angeles, CA in 2004, implying that the high
11 NOx level (~lppm) cited above for Los Angeles consisted mainly of NO. It is highly unlikely
12 that NOx oxidation products constituted a significant fraction of the NOx reported.
13
14 AX3.2.1 Spatial and Temporal Variability in Ambient Concentrations of
15 NO2 and Related Species in Urban Areas
16 As noted earlier, the number of monitoring sites reporting data for NO2 is considerably
17 smaller than for other criteria pollutants. As a result, there are few urban areas where there exist
18 sufficient data to evaluate the spatial variability in NO2 even though most of the NO2 monitors
19 are found in urban or suburban areas. Analyses of spatial variability in NO and NO2 are thus
20 limited to Los Angeles, CA and Chicago, IL. Also, as noted in Chapter 2, current methods for
21 measuring NO2 are subject to interference from its oxidation products. Hence the reported
22 values represent upper limits for the true NO2 concentration. Near highways or other NOx
23 sources, the measurements should give more accurate values, but because of variability in the
24 time needed for conversion of NOx to NOz, no firm rules can be applied to account for the
25 presence of NOz species such as HNOs and PANs. These considerations introduce additional
26 uncertainty into the interpretation of any metrics (e.g., correlation coefficients, concentration
27 differences) that are used to characterize spatial variability in NO2 concentrations.
28 The spatial variability in 1 h average NO2 concentrations in New York, NY; Atlanta, GA;
29 Chicago, IL; Houston, TX; Los Angeles, CA; and Riverside, CA is characterized in this section.
30 These areas were chosen to provide analyses to help guide risk assessment and to provide a
31 general overview of the spatial variability of NO2 in cities where health outcome studies have
32 been conducted. Statistical analyses of the human health effects of airborne pollutants based on
March 2008 AX3-4 DRAFT-DO NOT CITE OR QUOTE
-------
1 aggregate population time-series data have often relied on ambient concentrations of pollutants
2 measured at one or more central sites in a given metropolitan area. In particular, cities with low
3 traffic densities that are located downwind of major sources of precursors are heavily influenced
4 by long range transport and tend to show smaller spatial variability (e.g., New Haven, CT) than
5 those source areas with high traffic densities located upwind (e.g., New York, NY).
6 Metrics for characterizing spatial variability include the use of Pearson correlation
7 coefficients, values of the 90th percentile (P90) of the absolute difference in concentrations, and
8 coefficients of divergence (COD) The COD is defined as follows:
ViU
•">ii ' "-it
(AX3.2-1)
Y--+ y
Aij ^ A-A
10 where xy- and x;k represent observed concentrations averaged over some measurement averaging
11 period (hourly, daily, etc.), for measurement period /' at site7 and site k andp is the number of
12 observations. These methods of analysis follow those used for characterizing PM2 5 and PMi0-2.5
13 concentrations in Pinto et al. (2004) and in the latest edition of the PM AQCD (U.S.
14 Environmental Protection Agency, 2004).
15 Summary statistics for the spatial variability in several urban areas across the United
16 States are shown in Table AX3.2-2. These areas were chosen because they are the major urban
17 areas with at least five monitors operating from 2003 to 2005. Values in parentheses below the
18 city name refer to the number of sites colleting data. The second column shows the mean 1 h
19 average concentration across all sites and the range in means at individual sites. The third
20 column gives the range of Pearson correlation coefficients between individual site pairs in the
21 urban area. The fourth column shows the 90th percentile absolute difference in concentrations
22 between site pairs. The fifth column gives the coefficient of divergence (COD).
23 As can be seen from the table, mean concentrations at individual sites vary by factors of
24 1.5 to 6 in the MS As examined. Correlations between individual site pairs range from slightly
25 negative to highly positive in a given urban area. The sites in New York City tend to be the most
26 highly correlated and show the highest mean levels, reflecting their proximity to traffic, as
27 evidenced by the highest mean concentration of all the entries. However, correlation coefficients
28 are not sufficient for describing spatial variability as concentrations at two sites may be highly
29 correlated but show differences in levels. Thus, the range in mean concentrations is given. Even
March 2008 AX3-5 DRAFT-DO NOT CITE OR QUOTE
-------
1 in New York City, the spread in mean concentrations is about 40% of the city-wide mean
2 (12/29). The relative spread in mean concentrations is larger in the other urban areas shown in
3 Table AX3.2-2. As might be expected, the 90th percentile concentration spreads are even larger
4 than the spreads in the means.
5 The same statistics shown in Table AX3.2-2 have been used to describe the spatial
6 variability of PM2.5 (U.S. Environmental Protection Agency, 2004; Pinto et al., 2004) and 63
7 (U.S. Environmental Protection Agency, 2006a). However, because of relative sparseness in
8 data coverage for NO2, spatial variability in all cities that were considered for PM2.5 and 63
9 could not be considered here. Thus, the number of cities included below is much smaller than
10 for either 63 (24 urban areas) or PM2.5 (27 urban areas). Even in those cities where there are
11 monitors for all three pollutants, data may not have been collected at the same locations and even
12 if they were, there would be variable influence from local sources. For example, concentrations
13 of NO2 collected near traffic will be highest in an urban area, but concentrations of Os will tend
14 to be lowest because of titration by NO forming NO2. However, some general observations can
15 still be made. Mean concentrations of NO2 at individual monitoring sites are not as highly
16 variable as for 63 but are more highly variable than PM2.5. Lower bounds on inter-site
17 correlation coefficients for PM2.5 and for 63 tend to be much higher than NO2 in the same areas
18 shown in Table AX3.2-2. CODs for PM2 5 are much lower than for Os, whereas CODs for NO2
19 tend to be the largest among the three pollutants. Therefore, it is apparent that there is the
20 potential for errors from the use of ambient monitors to characterize exposures either at the
21 community or personal level, and that this potential may be higher than for either Os or PM2 5.
22
23 Small Scale Vertical Variability
24 Inlets to instruments for monitoring gas phase criteria pollutants can be located from 3 to
25 15m above ground level (Code of Federal Regulations, 2002). Depending on the pollutant,
26 either there can be positive, negative or no vertical gradient from the ground to the monitor inlet.
27 Pollutants that are formed over large areas by atmospheric photochemical reactions and are
28 destroyed by deposition to the surface or by reaction with pollutants emitted near the surface
29 show positive vertical gradients. Pollutants that are emitted by sources at or just above ground
30 level show negative vertical gradients. Pollutants with area sources and have minimal deposition
31 velocities show little or no vertical gradient. Restrepo et al. (2004) compared data for criteria
32 pollutants collected at fixed monitoring sites at 15 m above the surface on a school rooftop to
March 2008 AX3-6 DRAFT-DO NOT CITE OR QUOTE
-------
1 those measured by a van whose inlet was 4 m above the surface at monitoring sites in the South
2 Bronx during two sampling periods in November and December 2001. They found that CO,
3 SO2, and NO2 showed positive vertical gradients, whereas O3 showed a negative vertical gradient
4 and PM2.5 showed no significant vertical gradient. As shown in Figure AX3.2-2, NO2 mixing
5 ratios obtained at 4 m (mean -74 ppb) were about a factor of 2.5 higher than at 15 m (mean -30
6 ppb). Because tail pipe emissions occur at lower heights, NO2 values could have been much
7 higher nearer to the surface, and the underestimation of NO2 values by monitoring at 15 m even
8 larger. Restrepo et al. (2004) note that the use of the NO2 data obtained by the stationary
9 monitors would result in an underestimate of human exposures to NO2 in the South Bronx.
10 However, this issue is most likely not unique to the South Bronx and could arise in other large
11 urban areas in the United States with populations of similar demographic and socioeconomic
12 characteristics.
£
a.
Q.
0 08 •
0 06 -
0 G4 -
n -
*.
•* * * .*
• * * * »
* * », +
• u** .•» .
* » * * »
A * ** _ **
* • ^f Ik ' « ^»^ ij*
-------
1 AX3.2.2.1 Diurnal Variability in NOi Concentrations
2 As might be expected from a pollutant having a major traffic source, the diurnal cycle of
3 NC>2 in typical urban areas is characterized by traffic emissions, with peaks in emissions
4 occurring during morning and evening rush hour traffic. Motor vehicle emissions consist mainly
5 of NO, with only about 10% of primary emissions in the form of NC>2. The diurnal pattern of
6 NO and NO2 concentrations is also strongly influenced by the diurnal variation in the mixing
7 layer height. Thus, during the morning rush hour when mixing layer heights are still low, traffic
8 produces a peak in NO and NO2 concentrations. As the mixing layer height increases during the
9 day, dilution of emissions occurs. During the afternoon rush hour, mixing layer heights are at or
10 are near their daily maximum values resulting in dilution of traffic emissions through a larger
11 volume than in the morning. Starting near sunset, the mixing layer height drops and conversion
12 of NO to NO2 occurs without photolysis of NO2 recycling NO.
13 The composite diurnal variability of NO2 in selected urban areas with multiple sites
14 (New York, NY; Atlanta, GA; Baton Rouge, LA; Chicago. IL; Houston, TX; Riverside, CA;
15 and Los Angeles, CA) is shown in Figure AX3.2-3. Figure AX3.2-3 shows that lowest hourly
16 median concentrations are typically found at around midday and that highest hourly median
17 concentrations are found either in the early morning or in mid-evening. Median values range by
18 about a factor of two from about 13 ppb to about 25 ppb. However, individual hourly
19 concentrations can be considerably higher than these typical median values, and hourly NO2
20 concentrations > 0.10 ppm can be found at any time of day.
21
22 AX3.2.2.2 Seasonal Variability in NO2 Concentrations
23
24 Urban Sites
25 As might be expected from an atmospheric species that behaves essentially like a primary
26 pollutant emitted from surface sources, there is strong seasonal variability in NOx and NO2
27 concentrations. Highest concentrations are found during winter, consistent with lowest mixing
28 layer heights found during the year. Mean and peak concentrations in winter can be up to a
29 factor of two larger than in the summer at several sites in Los Angeles County.
March 2008 AX3-8 DRAFT-DO NOT CITE OR QUOTE
-------
o
'=.
I
0>
o
o
O
0.20-
0.19-
0.18-
0.17-
0.16-
0.15-
0.14-
0.13-
0.12-
0.11-
0.10-
0.09-
0.08-
0.07-
0.06-
0.05-
0.04-
0.03-
0.02-
0.01-
0.00-
x
x
X
012345
i
6
I ' I
89
I ' I ' I • I ' t ' I • I ' I ' I ' I ' I ' I ' 1 ' I ' I
101112131415161718192021222324
Hour
Figure AX3.2-3. Composite, diurnal variability in 1-h average NO2 in urban areas. Values
shown are averages from 2003 through 2005. Boxes define the
interquartile range, and the whiskers the 5th and 95th percentile values.
Asterisks denote individual values above the 95th percentile.
1 The month-to-month variability in NO2 at individual sites in selected urban areas is
2 illustrated in Figures AX3.2-4 to AX3.2-10. Seasonal patterns can be found at some sites but not
3 in others. There appears to be a somewhat regular pattern for the southern cities with winter
4 maxima and summer minima. Monthly maxima tend to be found from late winter to early spring
5 in Chicago and New York with minima occurring from summer through the fall. However, in
6 Los Angeles and Riverside, monthly maxima tend to occur from autumn through early winter
7 with minima occurring from spring through early summer.
8
9 Regional Background Sites
10 Surface NOx and NOy data obtained in Shenandoah National Park, VA from 1988 to
11 1989 show wintertime maxima and summertime minima (Doddridge et al., 1991, 1992; Poulida
12 et al., 1991). NOX and NOY data collected in Harvard Forest, MA from 1990 to 1993 show a
13 similar seasonal pattern (Munger et al., 1996). In addition the within-season variability was
14 found to be smaller in the summer than in the winter as shown in Table AX3.2-3.
March 2008
AX3-9
DRAFT-DO NOT CITE OR QUOTE
-------
a. New York. NY.
SUBURBAN
b. New York, NY.
URBAN and CENTER CITY
Q.
& 006
C
.2 005
£ 004
g 0.03
§ 0.02
001
033
Alttid « 3Mfl10124 p
| -M. • Kalura1 Spire Fit »,' a dl j
c
.2 005.
2 004
g 003
O 002
001
100
01(01(2003 07(01(2003 01(01(2004 07(01(2001 01(01(2005 07(01(2005 01(01(2006
Sample Date (mm/dd/yyyy)
01(01(2003 07mi200} 01/01(2004 DT«100M 01(01)2005 07W1/20M 01(01(2006
Sample Date {mm/dd/yyyy)
c. New York, NY,
URBAN and CENTER CITY
oca
oca.
0-07-
OCG
005
CM-
CM-
OCJ-
001
000-
titori =M00601!0[>oc =
d New York, NY URBAN and CENTER CITY
Q.
c
O
0
01(01(2003 07(010003
01(01(2005 07(010005
01(01(2003 07(01(2003
07.01(2005 01(01(2006
Sample Date (mm/dd/yyyy)
Sample Date (mm/dd/yyyy)
e. New York, NY. URBAN and CENTER CITY
C
O
O
01(01(2003 07(01/2033 01(01(2004 07(01(2004 01(01(2005 07(01j2005 01j01(20C6
Sample Date (mm/dd/yyyy)
Figure AX3.2-4a-e. Time series of 24-h average NOi concentrations at individual sites in
New York City from 2003 through 2006. A natural spline function
(with 9 degrees of freedom) was fit and overlaid to the data (dark
solid line).
March 2008
AX3-10
DRAFT-DO NOT CITE OR QUOTE
-------
a. Atlanta, GA.
SUBURBAN
site id=130890002 poc=1
E
a
a.
o
'•M
ro
*•>
CD
O
c
o
o
= Natural Spline Fit w/ 9 df
0.09-I
0.08
0.07-
0.06-
0.05
0.04-
0.03:
0.02-
001.
0.00^
01/01/2003 07/01/2003 01/01/2004 07/01/2004 01/01/2005 07/01/2005 01/01/2006
b. Atlanta, GA.
Sample Date (mm/dd/yyyy)
URBAN and CENTER CITY
a.
Q.
C
O
0)
o
o
o
0.09:
0.08:
0.07-
0.06 •:
0.05-
0.04:
0.03-;
0.02-i
0.01 •:
0.00-
site id=131210048 poc=1
01/01/2003 07/01/2003 01/01/2004 07/01/2004 01/01/2005 07/01/2005 01/01/2006
Sample Date (mm/dd/yyyy)
Figure AX3.2-5a-e. Time series of 24-h average NOi concentrations at individual sites in
Atlanta, GA from 2003 through 2005. A natural spline function (with
9 degrees of freedom) was fit and overlaid to the data (dark solid line).
March 2008
AX3-11
DRAFT-DO NOT CITE OR QUOTE
-------
a. Chicago, IL.
RURAL
b. Chicago, IL.
SUBURBAN
007-
006-
001'
DCO-
«UA j^-. ij=—to
wT^^f^t ^rjpfjj "fTOfvm
Sample Date {mm/dd/yyyy}
Sample Date (mm/dd/yyyy)
C. Chicago, IL
SUBURBAN
5.
S 006-
C
O (
2 DW:
d- Chicago, IL
SUBURBAN
Sampl« Date {mm/dd,'yyyyj
DI/QICO&* effiflusodw oifli^OB
Sample Date (mm/dd/yyyy)
e. Chicago, IL,
.« 005-
5
t> O'-W"
g Q.DJ.
u
SUBURBAN
00:-
COC-
01^1/2003 07/010003 01rfH3004 07fl)1/20M 01^31/2005
Sample Date {mm/dd/yyyy}
f. Chicago, IL.
•—• aos-
.2 0.05 4
?
g
2 003.
O W.
aoi.
URBAN and CENTER CITY
14* J • 170310W3 po4 •
01*1/2006 oirtjriaooa
OliOlQOO* 07*l^!flOfl Ol/Ol/SOOS
Sample Date {mm/dd/yyyy)
g. Chicago, IL. URBAN and CENTER CITY
-------
a. Baton Rouge, LA.
SUBURBAN
a
Q.
c
o
I
O
o
0.09:
0.08-i
0.07-i
0.06^
0.05-i
0.04 •;
0.03-j
0.02-i
0.01-j
0.00^
siteid=221210001 poc=1
= Natural Spline Fit w/ 9 df
01/01/2003 07/01/2003 01/01/2004 07/01/2004 01/01/2005 07/01/2005 01/01/2006
Sample Date (mm/dd/yyyy)
b. Baton Rouge, LA.
URBAN and CENTER CITY
Q.
O
'+">
re
o>
o
c
o
o
0.09:
0.08-
0.07-
0.06-
0.05:
0.04:
0.03:
0.02:
0.01-
0.00-
site id=220330009 poc=1
I I
01/01/2003 07/01/2003 01/01/2004 07/01/2004 01/01/2005 07/01/2005 01/01/2006
Sample Date (mm/dd/yyyy)
Figure AX3.2-7a-b. Time series of 24-h average NOi concentrations at individual sites in
Baton Rouge, LA from 2003 through 2005. A natural spline function
(with 9 degrees of freedom) was fit and overlaid to the data (dark
solid line).
March 2008
AX3-13
DRAFT-DO NOT CITE OR QUOTE
-------
a. Houston, TX.
SUBURBAN
b. Houston, TX.
SUBURBAN
^-, oa
£ „
i'.L&".-:ji'.a:'t:- '
Sample Date (mm/dd/yyyy)
9 £4.
003-
3C2-
001-
0:0-
Sample Date (mm/dd/yyyy)
c. Houston, TX.
SUBURBAN
=.
ace-
s
o
o
CO' •
cce-
URBAN and CENTER CITY
Cn^OUSBJ QTflJfEQM 01^)1W»4
Sample Date (mm/dd/yyyy)
f, Houston, TX,
URBAN and CENTER CITY
Sample Date (mm/dd/yyyy)
g. Houston. TX. U RBAN and C ENTER CITY
E 0.07
Q.
S 0»<
c
o Q.K.
O CK
o
001
Sample Date (mm/dd/yyyy)
Figure AX3.2-8a-g. Time series of 24-h average NOi concentrations at individual sites in
Houston, TX from 2003 through 2005. A natural spline function (with
9 degrees of freedom) was fit and overlaid to the data (dark solid line).
March 2008
AX3-14
DRAFT-DO NOT CITE OR QUOTE
-------
a. Los Angeles, CA.
SUBURBAN
b. Los Angeles, CA.
SUBURBAN
G1«1J2fKB 07J01/2003
Sample Date (mm/dd/yyyy)
Sample Date (mm/dd/yyyy)
c. Los Angeles, CA.
SUBURBAN
d. Los Angeles, CA.
SUBURBAN
01*1^003 OT/01/2003 OMQiraOW OWHaHXl
Sample Date (mm/dd/yyyy)
Sample Date (mm/dd/yyyy)
e. Los Angeles, CA,
SUBURBAN
f. Los Angeles, CA.
SUBURBAN
OWIEOM Q7rtJ1fflB4
Sample Date (mm/dd/yyyy)
Sample Date (mm/dd/yyyy)
g. Los Angeles, CA.
SUBURBAN
h. LOS Angeles, CA. URBAN and CENTER CITY
D1/Q1/20W {J7>D1/3XM
Sample Date (mm/dd/yyyy)
OTfllQOtB 07/0tia»3 01*1J20O1 07/01COM QI/QltfQQS 07flli2006
Sample Date (mm/dd/yyyy)
Figure AX3.2-9a-h. Time series of 24-h average NOi concentrations at individual sites in
Los Angeles, CA from 2003 through 2005. A natural spline function
(with 9 degrees of freedom) was fit and overlaid to the data (dark
solid line).
March 2008
AX3-15
DRAFT-DO NOT CITE OR QUOTE
-------
LOS Angeles, CA. URBAN and CENTER CITY
j. LOS Angeles, CA. URBAN and CENTER CITY
I
0.
c
01
o
c
o
o
07/01/2003 OM31/2004 07(01/2004 01j01/2CG5
Sample Date (mm/dd/yyyy)
07)01/2003 01*1/2004 07)0112004 01(01(2005 07/010005
Sample Date (mm/dd/yyyy}
k. Los Angeles, CA. URBAN and CENTER CITY
i. LOS Angeles, CA. URBAN and CENTER CITY
a.
a.
O
o
c
o
o
01)01(2006 07/0112005
0110112003 07/01(2003 01101/2004 07|01I2004 QlfflteGOB 07(01(2005
Sample Date (mm/dd/yyyy)
Sample Date (mm/dd/yyyy)
n. LOS Angeles, CA. URBAN and CENTER CITY
•rfe d = GSQ3750Q1 poe = 1
n. Los Angeles, CA, URBAN and CENTER CITY
01(01(2003 074)10003 01/01/2004 07,01(2004 OKOIftX* 07(01/2006 01(01(200$ 01(01(2003 07/01(2003 01/01(2004 07/OIG004 01(01(2005 07(01(2005 01/01(2006
Sample Date (mm/dd/yyyy)
Sample Date (mm/dd/yyyy)
Figure AX3.2-9i-n.
Time series of 24-h average NOi concentrations at individual sites in
Los Angeles, CA from 2003 through 2006. A natural spline function
(with 9 degrees of freedom) was fit and overlaid to the data (dark
solid line).
March 2008
AX3-16
DRAFT-DO NOT CITE OR QUOTE
-------
a. Riverside, CA.
RURAL
b. Riverside, CA.
SUBURBAN
MS-
006.
COS-
004-
003^
002-
0.01 • 1
000- '
01(01(2003
-~, - Nnluial Spline fe-ntl &dl
a
_a
B
o
2
c. Riverside, CA.
01101(2004 07(01)2004 01)01(2005 07)01)2005 01(01(2008
Sample Date (mm/dd/yyyy)
SUBURBAN
01(01(2003 07(01(2003 01101)2004 07)01)2004 01(01(2005 07(01(2005 01BK2006
d. Riverside, CA.
Sample Date (mm/dd/yyyy)
SUBURBAN
oca
oos
mar
0.06'
005
VM
003
002
001
ooc
01*1)2003 07)01(2003 01(01(20M 07(010X14 OWK2005 07)01(2005 01OT200S
Sample Date (mm/dd/yyyy}
o
o
01(01(2003 07(01(2003 01(01)2004 07)0112004 01(01(2005 07)01(2005 0101)2005
Sample Date (mm/dd/yyyy)
Figure AX3.2-10a-d. Time series of 24-h average NOi concentrations at individual sites in
Riverside, CA from 2003 through 2006. A natural spline function
(with 9 degrees of freedom) was fit and overlaid to the data (dark
solid line).
1 AX3.2.2.3 Trends in NO2 Concentrations
2 Figure AX3.2-11 shows the nationwide trend in annual mean NC>2 concentrations from
3 1983 to 2002. As can be seen from the figure, NC>2 concentrations have decreased by about 10%
4 per decade. As can be seen from Figure AX3.2-12, most monitoring sites are located in either
5 urban (49) or suburban (58) areas and comparatively few monitoring sites are located in rural
6 areas (14). Figure AX3.2-12 also shows that decreases have been at least twice as large in urban
7 and suburban areas than in rural areas and that NC>2 concentrations in urban and suburban areas
8 are roughly twice those in rural areas. Note that a land use characterization of rural does not
March 2008
AX3-17
DRAFT-DO NOT CITE OR QUOTE
-------
e. Riverside, CA.
SUBURBAN
f. Riverside, CA.
SUBURBAN
a. uu'
Q.
— 005
a ri = 060712002 pDC = 1
01(01/2003 07/0112003 01(01/2004 07/01(2004 01(01/2005 07/01(2005 01(01(2006
005^
004-
003-
002-
101. |
0.00-
01(01(2003 07)01/2003 01/01/2004 07/01/2KM 01/01(2005 07/OU20CS 0«)1(2006
Sample Date (mm/dd/yyyy)
Sample Date (mm/dd/yyyy)
g. Riverside, CA.
0.09-
SUBURBAN
h Riverside. CA URBAN and CENTER CITY
01(01/2003 07/01(2003 01(01(2004 07/01(2004 01101(2005 07/010005 Olffll/2006
000-
01(01/2003 07/0112003 0101/2004 07/01(2004 01/01(2005 07/01(2005 01(01(2006
Sample Date (mm/dd/yyyy)
Sample Date (mm/dd/yyyy)
i. Riverside, CA. URBAN and CENTER CITY
01/01M03 07*1/2003 01/OW2OM 07)01/2004 01/01/2OS 07(01/2005 01/01.0003
Sample Date (mm/dd/yyyy)
Figure AX3.2-10e-i. Time series of 24-h average NOi concentrations at individual sites in
Riverside, CA from 2003 through 2006. A natural spline function
(with 9 degrees of freedom) was fit and overlaid to the data (dark
solid line).
March 2008
AX3-18 DRAFT-DO NOT CITE OR QUOTE
-------
0.06
0.05 -
sites have* concentraliens wlcw this line
0.0
83 64 85 86 87 88 83 90 91 92 93 94 95 &6 97 98 99 00 01 02
1983-02: 21% decrease
1993-02: 11% decrease
Figure AX3.2-11. Nationwide trends in annual mean NOi concentrations.
Source: U.S. Environmental Protection Agency (2003).
.C'30
.025
,03,
1 .015
I
§ -010
.COS
.000
19B2-2O01
Rural Sil«e 14
Suburban Sites
Urban SHes
sa
49
62 83 84 85 86 87 88 6S> go 01 02 93 S4 SiS 96 87 98 99 00 01
Year
Figure AX3.2-12. Trends in annual mean NOi concentrations by site type.
4
5
6
Source: U.S. Environmental Protection Agency (2003)
imply that a site is free of local pollution influences, as evidenced by the still relatively high
values at rural sites compared to those found in remote areas of the globe. Rural sites can be
affected by nearby highways, power plants, and other sources.
In addition to the downward trend in annual mean concentrations of NC>2 shown in
Figures AX3.2-11 and AX3.2-12, hourly maximum concentrations have also declined, as
evidenced by a number of peak values above 250 ppb across the United States in 1988. In
March 2008
AX3-19
DRAFT-DO NOT CITE OR QUOTE
-------
1 contrast only one hourly maximum concentration above 250 ppb was found in 2004 (however,
2 this may have been a measurement artifact as it represented a one h spike that was many times
3 the next highest concentration at this site), and all other values were less than about 150 ppb.
4
5 AX3.2.4 Relationships between NOi and Other Pollutants
6 Determining the relationships between NO2 and other pollutants is important for better
7 understanding the findings of time-series epidemological studies relating NO2 to mortality
8 (e.g., Burnett et al., 2004). Correlations between NO2 and CO, Os, and PM2.5 were calculated for
9 monitoring sites in Los Angeles and Riverside, CA; Chicago, IL; Washington, D.C.; and New
10 York City. Correlations were calculated using both hourly and 24-h average data with similar
11 results. The ranges of Pearson correlation coefficients between 24-h average NO2 and Os, CO
12 and PM2.5 for 2000 through 2004 at monitoring sites in a few urban areas are shown in Table
13 AX3.2-4. As can be seen from the table, correlations of NO2 with Os range from negative to
14 slightly positive; with CO they range from slightly negative to highly positive, and with PM2.5
15 they range from slightly to moderately positive. However, it should be noted that these
16 correlations are based on annual data from sites influenced by local sources. In general, there is
17 a strong seasonal variation in the correlations, r, with lowest values of r between NO2 and Os
18 found in winter.
19 In order to understand the relations between atmospheric species as shown in Table
20 AX3.2-4, an important distinction must be made between primary (directly emitted) species and
21 secondary (photochemically produced) species. In general, it is more likely that primary species
22 will be more highly correlated with each other, and that secondary species will be more highly
23 correlated with each other. By contrast, primary and secondary species are less likely to be
24 correlated with each other. Secondary reaction products tend to correlate with each other, but
25 there is considerable variation. Some species (e.g., Os and organic nitrates) are closely related
26 photochemically and correlate with each other strongly.
27 Although NO2 is produced mainly by the reaction of directly emitted NO with Os with
28 a small contribution from direct emissions, in practice, it behaves like a primary species. The
29 timescale for conversion of NO to NO2 is relatively rapid (~1 or 2 min for Os = 40 ppb and
30 ambient temperatures from 273 to 298 K), so NO and NO2 ambient concentrations rapidly
31 approach values determined by the photochemical steady state. The sum of NO and NO2 (NOx)
March 2008 AX3-20 DRAFT-DO NOT CITE OR QUOTE
-------
1 behaves like a typical primary species, while NO and NO2 reflect some additional complexity
2 based on photochemical interconversion. Chemical interactions among O3, NO and NO2 have
3 the effect of converting O3 to NO2 and vice versa, which can result in a significant negative
4 correlation between O3 and NO2.
5 Most CO in urban air is emitted from motor vehicles and so is primary in origin. O3 is a
6 secondary pollutant. Figures AX3.2-13a-d show seasonal plots of correlations between NO2 and
7 Os versus correlations between NO2 and CO. As can be seen from the figures, NO2 is positively
8 correlated with CO during all seasons at all sites. However, the sign of the correlation of NO2
9 with O3 varies with season, ranging from negative during winter to slightly positive during
10 summer. There are at least two main factors contributing to the observed seasonal behavior.
11 O3 and radicals correlated with it tend to be higher during the summer, thereby tending to
12 increase the NO2 to NO ratio according to the expression below (Equation AX3.2-2).
_ k,(O3) + k2(HO2) + k3(RO3)
13 N° (AX3.2-2)
14 NOz compounds formed from the oxidation of NOx are also expected to be correlated
15 with O3 and increased photochemical activity. Because of interference of NOZ compounds with
16 the measurement of NO2 by conventional chemiluminescent monitors, they may also tend to
17 increase the correlation of NO2 with O3 during the warmer months. However, there is not
18 enough information on the seasonal behavior in their concentrations to quantify the contribution
19 of NOz compounds.
20 Relationships between O3, NO, and NO2 are shown in Figures AX3.2-14 and AX3.2-15.
21 Figure AX3.2-14 shows daylight average concentrations based on data collected from November
22 1998 and 1999 at several sites in the United Kingdom representing a wide range of pollution
23 conditions (open symbols). The solid lines represent calculations of photostat! onary state values
24 subj ect to the constraint that Ox = 3 1 . 1 + 0. 1 04(NOX), where Ox = O3 + NO2. Note that Ox is
25 defined in the UK AQG report as oxidant, as used in this document, and in the latest AQCD for
26 Ozone and other Photochemical Oxidants (U.S. Environmental Protection Agency, 2006a) it is
27 taken to refer to "odd oxygen" as defined in Section 2.2. The reason is that oxidants also include
28 PANs, peroxides, and reactive oxygen species in particles etc., in addition to O3 and NO2. The
29
March 2008 AX3-21 DRAFT-DO NOT CITE OR QUOTE
-------
o
*CS|
O
^
CO
o
CM
o
Winter
0.8-
0.6-
0.4-
0.2-
-0.8 -0.6 -0.4 -0.2 (
-0.2-
-0.4 •
-0.6-
-0.8-
0.2 0.4 0.6 0.8
• #*++}
* **
NO2: CO
Spring
0.8-
0.6-
0.4-
0.2-
1 -0.8 -0.6 -0.4 -0.2
% * ++1 +
I Q^ o5 0.6 Js
O
N
o
z
CO
o
o-
Summer
0.8-
0.6-
0.4-
0.2-
1 -0.8 -0.6 -0.4 -0.2 [
-0.2-
-0.4-
-0.6-
-0.8-
/ * t
) 0.2 0.4 0.6 0.8
NO2: CO
Fall
0.8-
0.6-
0.4-
0.2-
1 -0.8 -0.6 -0.4 -0.2 (
* ** \ *
) 0.2 0.4 ^0.6 *.8
NO2: CO
NO,: CO
Figure AX3.2-13a-d. Correlations of NOi to Os vs. correlations of NOi to CO for Los
Angeles, CA (2001-2005).
4
5
6
emissions of NC>2 (an oxidant and a component of odd oxygen) varying linearly with emissions
of NOx, especially after NO has reacted with Os to form NO2 as shown in Figure AX3.2-14.
Thus the concentration of Ox (and not Os, as is often stated) can be taken to be the sum of
regional and local contributions.
Figure AX3.2-15 shows that primary emissions from motor vehicles are major sources of
oxidant in the form of NO2, as evidenced by the high values of Ox at elevated NOx.
March 2008
AX3-22
DRAFT-DO NOT CITE OR QUOTE
-------
a) 100
80
Q.
o
o
Z
40
20
100 200 300 400
[NOxHppb)
500
600
Figure AX3.2-14.
Relationship between Os, NO, and NOi as a function of NOx
concentration. Open circles represent data collected at a number of
sites in the United Kingdom. Lines represent calculated relationships
based on photostationary state.
Source: Clapp and Jenkin (2001).
1 AX3.2.5 Abundance of NOY Species
2 Data for individual NOy species are much less abundant than for either oxides of nitrogen
3 or for total NOy. Data for NOy species are collected typically as part of research field studies,
4 e.g., the Southern Oxidant Study (SOS), Texas Air Quality Study (TexAQS I and TexAQS II) in
5 the United States. So this information is simply not available for a large number of areas in the
6 United States.
7
8 PANs
9 Organic nitrates consist of PAN, a number of higher-order species with photochemistry
10 similar to PAN (e.g., PPN), and species such as alkyl nitrates with somewhat different
11 photochemistry. These species are produced by a photochemical process very similar to that of
12 Os. Photochemical production is initiated by the reaction of primary and secondary VOCs with
13 OH radicals, the resulting organic radicals subsequently react with NO2 (producing
March 2008
AX3-23
DRAFT-DO NOT CITE OR QUOTE
-------
100
local contribution to ox id ant
(NOx-dependent)
regional contribution to oxidant
(NO x-independent)
0 100 200 300 400 500 600
[N0x](ppb)
Figure AX3.2-15. Variation of odd oxygen (= Os + NOi) with NOx. The figure shows
the "regional" and the "local" contributions. Note that Ox refers to
odd oxygen in the document and the latest Os AQCD.
Source: Clapp and Jenkin (2001).
1 PAN and analogous species) or with NO (producing alkyl nitrates). The same sequence (with
2 organic radicals reacting with NO) leads to the formation of Os.
3 In addition, at warm temperatures, the concentration of PAN forms a photochemical
4 steady state with its radical precursors on a timescale of roughly 30 min. This steady state value
5 increases with the ambient concentration of Os (Sillman et al., 1990). Ozone and PAN may
6 show different seasonal cycles, because they are affected differently by temperature. Ambient
7 Os increases with temperature, driven in part by the photochemistry of PAN (see description in
8 Chapter 2). The atmospheric lifetime of PAN decreases rapidly with increasing temperature due
9 to thermal decomposition. Based on the above, the ratio of O3 to PAN is expected to show
10 seasonal changes, with highest ratios in summer, although there is no evidence from
11 measurements. Measured ambient concentrations (Figures AX3.2-16a-d) show a strong
12 nonlinear association between Os and PAN, and between Os and other organic nitrates (Pippin
13 et al., 2001; Roberts et al., 1998). Moreover, the uncertainty in the relationship between Os and
March 2008
AX3-24
DRAFT-DO NOT CITE OR QUOTE
-------
1 PAN grows as the level of PAN increases. Individual primary VOCs are generally highly
2 correlated with each other and with NOx (Figure AX3.2-17).
3 Measurements and models show that PAN in the United States includes major
4 contributions from both anthropogenic and biogenic VOC precursors (Horowitz et al., 1998;
5 Roberts et al., 1998). Measurements in Nashville during the 1999 summertime Southern
6 Oxidants Study (SOS) showed PPN and MPAN amounting to 14% and 25% of PANs,
7 respectively (Roberts et al., 2002). Measurements during the TexAQS 2000 study in Houston
8 indicated PAN concentrations of up to 6.5 ppbv (Roberts et al., 2003). PAN measurements in
9 southern California during the SCOS97-NARSTO study indicated peak concentrations of
10 5-10 ppbv, which can be contrasted to values of 60-70 ppbv measured back in 1960 (Grosjean,
11 2003). Vertical profiles measured from aircraft over the United States and off the Pacific coasts
12 typically show PAN concentrations above the boundary layer of only a few hundred pptv,
13 although there are significant enhancements associated with long-range transport of pollution
14 plumes from Asia (Kotchenruther et al., 2001; Roberts et al., 2004).
15 Observed ratios of PAN to NO2 as a function of NOx at a site at Silwood Park, Ascot,
16 Berkshire, UK are shown in Figure AX3.2-18 United Kingdom Air Quality Expert Group (U.K.
17 AQEG, 2004). As can be seen there is a very strong inverse relation between the ratio and the
18 NOx concentration, indicating photochemical oxidation of NOx has occurred in aged air masses
19 and that PAN can make a significant contribution to measurements of NO2 especially at low
20 levels of NO2 (cf. Section 2-8). It should be noted that these ratios will likely differ from those
21 found in the United States because of differences in the composition of precursor emissions, the
22 higher solar zenith angles found in the UK compared to the United States., and different
23 climactic conditions.
24 Nevertheless, these results indicate the potential importance of interference from NOy
25 compounds in measurements of NO2.
26
27 HONO
28 The ratio of HONO to NO2 as a function of NOx measured at a curbside site in a street
29 canyon in London, UK is shown in Figure AX3.2-19 (U.K. AQEG, 2004). The ratio is highly
30 variable, ranging from about 0.01 to 0.1, with a mean -0.05. As NO2 constitutes several percent
March 2008 AX3-25 DRAFT-DO NOT CITE OR QUOTE
-------
0 1000 2000 3000 4000 0 1000 2000 3000 4000 5000
PAN (pptv)
Figure AX3.2-16a-d. Measured Os (ppbv) versus PAN (pptv) in Tennessee, including (a)
aircraft measurements, and (b, c, and d) suburban sites near
Nashville.
Source: Roberts etal. (1998).
1
2
3
4
5
6
7
8
of motor vehicle emissions of NOx, the above implies that emissions of HONO represent a few
tenths of a percent of mobile NOx emissions. A similar range of ratios have been observed at
other urban sites in the United Kingdom (Lammel and Cape, 1996). The ratios of HONO to NO2
shown in Figure AX3.2-19 indicate that HONO can make a measurable contribution to
measurements of NO2 (cf. Section 2-8). However, similar arguments about extrapolating the
use of UK data to the United States can be made for HONO as for PAN.
HNO3 andNO3
9 Elevated Os is generally accompanied by elevated HNOs, although the correlation is not
10 as strong as between Os and organic nitrates. Ozone is often associated with HNOs, because
11 they have the same precursor (NOx). However, HNOs can be produced in significant quantities
12 in winter, even when Os is low. The ratio between Os and HNOs also shows great variation in
13 air pollution events, with NOx-saturated environments having much lower ratios of Os to HNOs
14 (Ryerson et al., 2001). Aerosol nitrate is formed primarily by the combination of nitrate
15 (supplied by HNOs) with ammonia, and may be limited by the availability of either nitrate or
March 2008
AX3-26
DRAFT-DO NOT CITE OR QUOTE
-------
0.0
20
40 60 80
NOy (ppbv)
100 120
Figure AX3.2-17.
Relationship between benzene and NOy at a measurement site in
Boulder, CO. Instances with SO2 >10 ppb are identified separately
(open circles), because these may reflect different emission sources.
Source: Goldanetal. (1995).
1 ammonia. Nitrate is expected to correlate loosely with Os (see above), whereas ammonia is not
2 expected to correlate with 03. Concentrations of particulate nitrate measured as part of the
3 Environmental Protection Agency's speciation network at several locations are shown in Figure
4 AX3.2-20. Concentrations shown are annual averages for 2003. Also shown are the estimated
5 contributions from regional and local sources. A concentration of 1 |ig/m3 corresponds to -0.40
6 ppb equivalent gas phase concentration for NOs~.
March 2008
AX3-27
DRAFT-DO NOT CITE OR QUOTE
-------
o
EL
z*
Q_
0.20
0.15
0.10 >
0.05
0.00
AQ
20
40 60
[NOX] (ppb)
80
100
Figure AX3.2-18. Ratios of PAN to NOi observed at Silwood Park, Ascot, Berkshire,
U.K. from July 24 to August 12 1999. Each data point represents a
measurement averaged over 30 minutes.
Source: UK AQEG (2004).
0.3 -1—
CD0'2
z
*:£
O
z
O
I 0.1 ^
200
400
600
800
1000
(ppb)
Figure AX3.2-19.
Ratios of HONO to NOi observed in a street canyon (Marylebone
Road) in London, U.K. from 11 a.m. to midnight during October
1999. Data points reflect 15-min average concentrations of HONO
and NO2.
Source: UK AQEG (2004).
March 2008
AX3-28 DRAFT-DO NOT CITE OR QUOTE
-------
1 Thus, annual average particulate nitrate can account for several ppb of NOy, with the higher
2 values in the West. There is a strong seasonal variation, which is especially pronounced in
3 western areas where there is extensive wood burning in the winter resulting in a larger fractional
4 contribution of local sources. Areas in the East where there are topographic barriers might be
5 expected to show higher fractional contributions from local sources than other eastern areas that
6 are influenced by regionally dispersed sources.
7 However, depending on the acidity of the particles, which in turn depends strongly on
8 their sulfate and ammonium contents, higher nitrate concentrations could be found in coarse
9 mode particles PMi0-2.5 than in PM2.5 samples. The average nitrate content of PM2.5 and PMio is
10 typically about a percent in the eastern United States; and 15.7% and 4.5% in the western United
11 States (U.S. Environmental Protect!on Agency, 1996). These values suggest that most of the
12 nitrate was in the PM2.5 size fraction in the studies conducted in the western United States, but
13 nitrate in the studies in the eastern United States was mainly in the PMio-2.5 size fraction.
14
15 Nitro-PAHs
16 Nitro-PAHs (NPAHs) are widespread and found even in high altitude, relatively
17 unpolluted environments (Schauer et al., 2004) but there are differences in composition and
18 concentration profiles both within and between sites (rural vs. urban) as well as between and
19 within urban areas (Albinet et al., 2006; Soderstrom et al., 2005; Naumova et al., 2002, 2003),
20 with some differences in relative abundances of nitro- and oxo-PAHs also reported. Source
21 attribution has remained largely qualitative with respect to concentrations or mutagenicity (Eide
22 et al., 2002). The spatial and temporal concentration pattern for the NPAHs may differ from that
23 of the parent compounds (PAHs) because concentrations of the latter are dominated by direct
24 emission from local combustion sources. These emissions results in higher concentrations
25 during atmospheric conditions more typical of wintertime when mixing heights tend to be low.
26 The concentrations of secondary NPAHs are elevated under conditions that favor hydroxyl and
27 nitrate radical formation, i.e., during conditions more typical of summertime, and are enhanced
28 downwind of areas of high emission density of parent PAHs and show diurnal variation (Fraser
29 et al., 1998; Reisen and Arey, 2005; Kameda et al., 2004). Nitro-napthalene concentrations in
30 Los Angeles, CA varied between about 0.15 to almost 0.30 ng/m3 compared to 760 to
31 1500 ng/m3 for napthalene. Corresponding values for Riverside, CA were 0.012 to more than
32 0.30 ng/m3 for nitro-napthalene and 100 to 500 ng/m3 for napthalene. Nitro-pyrene
March 2008 AX3-29 DRAFT-DO NOT CITE OR QUOTE
-------
Nitrates
Figure AX3.2-20.
Fresno
Missoula
Salt Lake City
Tulsa
St. Louis
Birmingham
Indianapolis
Atlanta
Cleveland
Charlotte
Richmond
Baltimore
New York City
WEST
EAST
Q Regional
Contribution
• Local
Contribution
8
10
12
Annual Average Concentration
of Nitrates, MB'1113
Concentrations of particulate nitrate measures as part of the
Environmental Protection Agency PA's speciation network. 1 ug/m3
-0.45 ppb equivalent gas phase concentration for NOs". (Note:
Regional concentrations are derived from the rural IMPROVE
monitoring network, http://vista.cira.colostate.edu/improve.
Source: U.S. Environmental Protection Agency (2004).
1
2
3
4
5
6
1
8
9
10
11
concentrations in LA varied between approximately 0.020 to 0.060 ng/m3 compared to 3.3 to
6.9 ng/m3 pyrene, whereas corresponding values for Riverside were 0.012 to 0.025 ng/m3 and 0.9
to 2.7 ng/m3.
AX3.3 METHODS FOR MEASURING PERSONAL AND INDOOR NO2
CONCENTRATIONS
AX3.3.1 Issues in Measuring Personal/Indoor
Background
12 Nitrogen dioxide, a criteria air pollutant, has been sampled in ambient and indoor air
13 using active pumped systems both for continuous monitoring and collection onto adsorbents, and
14 by diffusive samplers of various designs, including badges and tubes. Nitrogen dioxide
March 2008
AX3-30
DRAFT-DO NOT CITE OR QUOTE
-------
1 concentrations in personal air have been typically measured using diffusive samplers because
2 they are: (1) small in size and light-weight, (2) unobtrusive and thus more readily used by study
3 participants, (3) comparatively easier to use and handle in field studies because they do not
4 require power (e.g., battery or extra electrical sources), (4) cost-effective, and (5) usable not only
5 for residential indoor and outdoor air sampling but also personal monitoring. However, diffusive
6 samplers usually have lower equivalent sampling rates than active methods and so require
7 relatively long sampling times (24 h or longer). Consequently, diffusive samplers including
8 those used for NO2 monitoring provide integrated but not short-term concentration
9 measurements.
10 Both active and passive sampling methods can collect other gas-phase nitrogen oxide
11 species. However, semivolatile nitrogen oxide compounds require separation of the gas- and
12 particle-bound phases. This selective separation of gases from gas-particle matrices is
13 commonly done by means of diffusion denuders (Vogel, 2005), an approach also useful for
14 measuring other gas phase airborne contaminants such as SC>2 (Rosman et al., 2001).
15 Application of denuder sampling to personal exposure or indoor air monitoring has been
16 relatively limited.
17 Active air sampling with a pump can collect larger volumes of air and thus detect the
18 lower concentrations found in community environments within relatively short time periods.
19 Automated active sampling methods have been the preferred method used to monitor NO2
20 continuously at ambient sites for environmental regulation compliance purposes. However,
21 practical considerations impede the use of these continuous monitors in residential air and
22 exposure monitoring studies. Small, low flow active samplers using battery-operated pumps
23 have been used instead, however, there are only a few such studies.
24 The first passive sampling devices for NC>2 were intended for occupational exposure
25 monitoring, but were later adapted for environmental monitoring purposes. Since this sampler,
26 the Palmes tubes (Palmes et al., 1976), was first developed, other tube, badge-type (Yanagisawa
27 and Nishimura, 1982) and radial (Cocheo et al., 1996) diffusive samplers have been employed as
28 monitors in exposure studies worldwide. The theories behind and applications of Palmes Tubes
29 and Yanagisawa badges have been described in the last AQCD for Oxides of Nitrogen (U.S.
30 Environmental Protection Agency, 1993). There are currently several commercially available
31 samplers (e.g., Ogawa, 1998; Radiello®, 2006) which are modifications of the original Palmes
March 2008 AX3-31 DRAFT-DO NOT CITE OR QUOTE
-------
1 tube design. Most modifications are directed at reducing effects related to meteorological
2 conditions (e.g., insufficient or too high a wind speed, humidity, temperature), increasing the
3 sampling uptake rate, and improving analytical sensitivity.
4
5 Active (Pumped) Sampling
6 Nitrogen dioxide measurement by active pumping systems as part of continuous monitors
7 has been widely employed for ambient air monitoring as these instruments require relatively
8 little maintenance; however they have been used less frequently for indoor sampling. Devices
9 needing a pump to draw air can measure average concentrations of pollutants over short time
10 periods, but are not generally suitable for measuring personal exposures because they are heavy
1 1 and large. Some exposure studies employed this approach for active sampling with stationary
12 chemiluminescent analyzers or portable monitors to measure nitrogen dioxide levels in
13 residential indoor air (Mourgeon et al., 1997; Levesque et al., 2000; Chau et al., 2002).
14 Recently, Staimer and his colleagues (2005) evaluated a miniaturized active sampler, suitable for
15 personal exposure monitoring, to estimate the daily exposure of pediatric asthmatics to nitrogen
16 dioxide, and reported that this small active sampling system is useful for this purpose in
17 environmental exposure epidemiology studies where daily measurements are desired.
18
19 Passive (Diffusive) Sampling
20 Passive samplers are based on the well known diffusion principle described by Pick's law
21 (Krupa and Legge, 2000). A convenient formulation of this law that can be easily related to
22 sampler design considerations is:
23 arsot (AX3.3-1)
24 where:
25 J = flux (mg/s)
26 D = diffusion coefficient in air (cm2/s)
27 A = diffusion cross-sectional area of the sampler (cm2)
28 L = diffusion path length from the inlet to sorbent (cm),
29 Cair = concentration of analyte in air (mg/cm3)
30 Csor = concentration of analyte at the sorbent (mg/cm3)
March 2008 AX3-32 DRAFT-DO NOT CITE OR QUOTE
-------
1 The term D(A/L) can be related to the uptake or sampling rate (cmVs) which is
2 conceptually analogous to the sampling rate in an active monitor. Once the amount of analyte in
3 the passive sampler sorbent is determined, the concentration in air (Ca;r) can be calculated as:
4 Concentration(mg/cm3) - M(mg)/D(A/L)(cm3/s)/t($ec) (AX3 3-2)
5 where:
6 M = mass of analyte collected in the sorbent
7 t = sampling time
8
9 Pick's law strictly applies only under ideal, steady state conditions assuming that the
10 sorbent is a perfect sink. However, there can be deviations between the theoretical sampling rate
1 1 for a given analyte and the actual rate depending on sampling conditions. It is also clear that
12 sampling rate can be optimized by modifying the geometry of the diffusive sampler, either by
13 reducing L, increasing A or a suitable combination. However, the impact of deviations from
14 ideality on actual sampling rate due to geometry also poses a limit to the extent of possible
15 modifications. Thus, passive samplers, either diffusive or permeation, are prepared as tubes or
16 badges. These two main designs are the basis for all further modifications which, as indicated
17 above, have been made in order to improve efficiency, reduce sensitivity to wind turbulence of
18 the samplers, and to simplify analyte desorption. Tube-type samplers are characterized by a
19 long, axial diffusion length, and a low cross-sectional area; this results in relatively low sampling
20 rates (Namiesnik et al., 2005). Badge-type samplers have a shorter diffusion path length and a
21 greater cross-sectional area which results in uptake rates that are typically higher than diffusion
22 tubes (Namiesnik et al., 2005) but the sampling rate may be more variable because it is more
23 affected by turbulence. Physical characteristics of these two fundamental passive sampler types,
24 tube-type and badge-type, are summarized and provided in Table AX3.3-1. Performance
25 characteristics are presented in Table AX3.3-2.
26 The sorbent can be either physically sorptive or chemisorptive; passive samplers for NC>2
27 are chemisorptive, that is, a reagent coated on a support (e.g., metal mesh, filter) reacts with the
28 NC>2. The sorbent is extracted and analyzed for one or more reactive derivatives; the mass of
29 NC>2 collected is derived from the concentration of the derivative(s) based on the stoichiometry
30 of the reaction. Thus, an additional approach to reducing detection limits associated with passive
March 2008 AX3-33 DRAFT-DO NOT CITE OR QUOTE
-------
1 samplers is to modify the chemisorptive reaction and the extraction and analysis methods to
2 increase analytical sensitivity. However, although chemisorption is less prone to the back
3 diffusion phenomenon of sorptive-only methods, analyte losses could occur due to interferences
4 from other pollutants that also react with the sorbent or the derivatives. The most commonly
5 used NC>2 passive samplers rely on the classical reaction with triethanolamine (TEA). TEA
6 requires hydration for quantitative NC>2 sampling (i.e., 1:1 conversion to nitrite) and the reaction
7 products have been subject to a number of investigations and several have been reported,
8 including TEA-nitrate and nitrite, triethanolammonium nitrate, nitrosodiethanolamine, and
9 triethanolamine N-oxide (Glasius et al., 1999). Known interferences include HONO, PAN, and
10 nitric acid (Gair et al., 1991.).
11 The tube-type passive samplers (Palmes tubes) require week-long sampling periods and
12 have been extensively used for residential indoor/outdoor measurements, mostly for exploring
13 the relationship between indoor and outdoor levels (Cyrys et al., 2000; Raw et al., 2004; Simoni
14 et al., 2004; Janssen et al., 2001). Passive diffusion tubes have also been widely used for
15 measurements of NC>2 in ambient air (Gonzales et al., 2005; Gauderman et al., 2005; Da Silva
16 et al., 2006; Lewne et al., 2004; Stevenson et al., 2001; Glasius et al., 1999). Personal exposure
17 studies have also been conducted using the Palmes tubes (Mukala et al., 1996; Kousa et al.,
18 2001). Some of these studies evaluated passive sampler performance by collocating them with
19 chemiluminescence analyzers during at least some portion of the field studies (Gair et al., 1991;
20 Gair and Penkett, 1995; Plaisance et al., 2004; Kirby et al., 2001). The majority of these studies
21 indicate that these samplers have very good precision (generally within 5%) but tend to
22 overestimate NO2 by 10 to 30%. However, there has not been a methodical evaluation of
23 variables contributing to variance for the range of samplers available when used in field
24 conditions. Thus, it is not clear if the bias is due to deviations from ideal sampling conditions
25 that can affect actual sampling rates, contributions from co-reacting contaminants or, most
26 probably, a combination of these variables.
27 A badge-type sampler was introduced by Yanagisawa and Nishimura (1982) to overcome
28 the long sampling time required by Palmes tubes. Since then, these sensitive NO2 short path
29 length samplers (Toyo Roshi Ltd) have been optimized and evaluated for indoor air and for
30 personal monitoring (Lee et al., 1993a,b). They have been used extensively for personal
31 exposure studies (Ramirez-Aguilar et al., 2002; Yanagisawa et al., 1986; Berglund et al., 1994,
March 2008 AX3-34 DRAFT-DO NOT CITE OR QUOTE
-------
1 Lee et al., 2004) and indoor air measurements (Kodama et al., 2002; Bae et al., 2004; Algar
2 et al., 2004; Shima and Adachi, 2000; Smedje, et al., 1997) and to a more limited amount for
3 ambient monitoring (Tashiro and Taniyama, 2002; Levy et al., 2006; Norris and Larson, 1999).
4 Due to the greater uptake rate resulting from the larger cross sectional area of the badges and
5 shorter diffusion length compared to the tube-type samplers, sampling times can be decreased
6 from one-week to one-day for typical environmental air concentrations. This makes diffusive
7 filter-badges more suitable for shorter-term sampling while long-term ambient monitoring can
8 still be conducted using the Palmes-tubes.
9
10 Tube Type Samplers
11 Gradko Sampler (http://www.gradko.co.uk)
12 The Gradko sampler is based on the Palmes tube design (Gerboles et al., 2006a).
13 It collects O3 or NO2 by molecular diffusion along an inert tube by chemisorption. A stable
14 complex is formed with triethanolamine coated on a stainless steel screen in the tube. The
15 complex is spectroscopically analyzed by adding an azo die (Chao and Law, 2000). The sampler
16 has a detection limit of 0.5 ppb for NO/NO2 and the precision of + 6% above 5 ppb levels when
17 used for two weeks (Table AX3.3-2). This sampler has been used to measure personal
18 exposures, concentrations of residential air indoors such as in the kitchen and bedroom, and
19 concentrations of outdoor air (Chao and Law, 2000; Gallelli et al., 2002; Lai et al., 2004). It has
20 been used to measure ambient NO2 levels in Southern California as a marker of traffic-related
21 pollution in San Diego County (Ross et al., 2006).
22
23 Passam Sampler (http://www.passam.ch)
24 This sampler is also based on the design of the Palmes tube (Palmes et al., 1976).
25 It collects NO2 by molecular diffusion along an inert polypropylene tube to an absorbent,
26 triethanolamine. The collected NO2 is determined spectrophotometrically by the well-
27 established Saltzmann method. When used outdoors the samplers are placed in a special shelter
28 to protect them from rain and minimize wind turbulence effects. The Passam sampler is sold in
29 two different models, one for long-term and one for short-term sampling.
30 Analyst™Sampler (http://www.monitoreurope.com)
March 2008 AX3-35 DRAFT-DO NOT CITE OR QUOTE
-------
1 The Analyst™ sampler is also a modification of the open-Palmes-tube design and was
2 developed by the Italian National Research Council (CNR - Institute Inquinamento
3 Atmosferico) in 2000 (Bertoni et al., 2001). The Analyst™ consists of a glass vessel, which
4 contains a reactant supported on a stainless steel grid. It is suitable for long-term monitoring
5 (typically one month) of oxides of nitrogen, sulfur dioxide, and volatile organic compounds in
6 ambient air. The target compound is analyzed by gas chromatography with minimum detection
7 limit of 0.1 mg/m3 (-52 ppb) for a twelve-week sample duration, and has relatively high
8 precision. The Analyst™ method development (De Santis et al., 1997, 2002) and actual field
9 application (De Santis et al., 2004) have been described. The primary use for Analyst™ is as a
10 reliable tool for long-term determination of concentration in indoor as well as outdoor
11 environments (Bertoni et al., 2001) and as a screening tool for ambient monitoring to identify
12 pollution "hot spots" (De Santis et al., 2004).
13
14 Badge-Types Samplers
15 Ogawa Passive Sampler (http://www.ogawausa.com)
16 This sampler is a double face badge that can monitor NO, NOx, and NC>2. The design
17 can be used also for the determination of 862, 63, and NH3 levels in air. The manufacturer-
18 reported detection limits for nitrogen oxides are 2.3 ppb and 0.32 ppb for 24-h and 168-h
19 sampling, respectively. Reported actual sampling rates for NO2 are two to three times higher
20 than the manufacturer's values. The normal operation ranges are 0 to 25 ppm for 24-h exposure
21 and 0 to 3.6 ppm for 168-h exposure. The manufacturer recommends a sampling height of
22 2.5 meters and storage time of up to 1 year when kept frozen. Ogawa passive samplers have
23 been extensively used for human exposure studies to measure personal air concentrations and
24 (or) indoor/outdoor levels for residents in a number of locations, including adults of Richmond,
25 Virginia (Zipprich et al., 2002), children of Santiago, Chile (Rojas-Bracho et al., 2002), office
26 workers of Paris, France (Mosqueron et al., 2002), and cardiac compromised individuals of
27 Toronto, Canada (Kim et al., 2006). The samplers have been used also in air monitoring
28 networks to assess traffic-related pollutant exposure (Singer et al., 2004), as well as to evaluate
29 spatial variability of nitrogen dioxide ambient concentrations in Montreal, Canada (Gilbert et al.,
30 2005).
31 IVL Sampler (http//www.ivl.se/en/business/monitoring/diffusive_samplers.asp)
March 2008 AX3-36 DRAFT-DO NOT CITE OR QUOTE
-------
1 The IVL method development has been described in detail by Perm and Svanberg (1998).
2 It was developed by Swedish Environmental Research Institute in the mid of 1980s (Sjodin et al.,
3 1996), is designed to minimize turbulent wind effects outdoors as well as "starvation effects"
4 indoors (i.e., very low face velocities), interferences from within sampling tube chemistry,
5 temperature and humidity effects, and artifacts and losses during post-sampling storage.
6 Manufacturer-reported detection limits for this sampler with sampling times of ~1 month are
7 0.1 |ig/m3 (0.05 ppb) for NO2, and 0.5 |ig/m3 (0.42 ppb) for NO, respectively. Due to its long
8 sampling time, this sampler has been extensively used for NO2 background monitoring in
9 ambient air of rural or urban (Fagundez et al., 2001; Sjodin et al., 1996; Pleijel et al., 2004).
10
11 Willems Badge Sampler
12 The Willems badge, a short-term diffusion sampler, was developed at the University of
13 Wageningen, Netherlands, originally for airborne ammonia measurements and later for
14 measuring NO2 (Hagenbjork-Gustafsson et al., 1996). It consists of a cylinder of polystyrene
15 with a Whatman GF-A glass fiber filter impregnated with triethanolamine at its based held in
16 place by a 6 mm distance ring. A Teflon filter is placed on the 6 mm polystyrene ring, which is
17 secured with a polystyrene ring of 3 mm (Hagenbjork-Gustafsson et al., 1996). The badge is
18 closed by a polyethylene cap to limit influences by air turbulence. The diffusion length in the
19 badge is 6 mm. This sampler was evaluated for ambient air measurements in laboratory and
20 field tests (Hagenbjork-Gustafsson et al., 1999). It has a manufacturer's reported detection limit
21 of 2 |ig/m3 (~1 ppb) for 48 h sampling duration. When used for personal sampling in an
22 occupational setting with a minimum wind velocity of 0.3 m/s, detection limits of 18 (-9.4 ppb)
23 and 2 |ig/m3 (~1 ppb) for 1-h and 8-h sampling, respectively, have been reported (Hagenbjork-
24 Gustafsson et al., 2002, Glas et al., 2004).
25
26 Radial Sampler Types
27 Radiello® -the radial diffusive sampler (http://www.radiello.com)
28 Radiello® samplers use radial diffusion over a microporous cylinder into an absorbing
29 inner cylinder, instead of axial diffusion, which increases the uptake rate by a factor of about
30 100 (Hertel et al., 2001). Nitrogen dioxide is chemiadsorbed onto triethanolamine as nitrite,
31 which is quantified by visible spectrometry. Sample collection of up to 15 days is feasible but
32 relative humidity higher than 70% can cause interferences when used for extended periods of
March 2008 AX3-37 DRAFT-DO NOT CITE OR QUOTE
-------
1 more than 7 days. The manufacturer-reported typical sampling rate for nitrogen dioxide
2 sampling is 75 + 3.72 ml/min at temperatures between -10 and 40 °C. The rate can vary with
3 humidity in the range of 15 to 90% and wind speed between 0.1 and 10 m/s (Radiello® Manual,
4 2006). A Danish study (S0rensen et al., 2005) recruited 30 subjects during each of four seasons
5 in Copenhagen, and measured the subjects' personal exposures, home indoor/front door air
6 concentrations during 2-day periods with this sampler.
7
8 EMD (Ecole des Mines de Douai) Sampler
9 A new high-uptake rate diffusive sampler has been recently developed by the Ecole
10 des Mines de Douai (EMD) laboratory (Piechocki-Minguy et al., 2003) and evaluated in the
11 laboratory and field for measurement of NC>2 levels in ambient air. It is composed of a porous
12 cartridge impregnated with triethanolamine and fitted in a cylindrical protective box equipped
13 with caps at its extremities (Piechocki-Minguy et al., 2006). The large sampling area (cartridge
14 surface) and the two circular openings provide a high uptake rate (exceeding 50 cmVmin). The
15 sampling rate was reported to be on average 0.89 cm3/s for indoor sampling and 1.00 cm3/s for
16 outdoor sampling. Detection limits were determined to be 11 |ig/m3 (-5.8 ppb) for 1-h
17 measurement. The sampling rate was not significantly influenced by wind at speeds higher than
18 0.3 m/s (Piechocki-Minguy et al., 2003). This sampler has been used in France to assess
19 personal exposures in a series of microenvironments (home, other indoor places, transport and
20 outdoor) for two 24-h time periods (weekday and weekend) (Piechocki-Minguy et al., 2006).
21
22 NO2 Measurements in Epidemological Studies
23 Since passive samplers are the most frequently used monitoring method in epidemiology
24 studies of NC>2 effects, their performance compared to the long established chemiluminescence
25 monitoring method is critical for determining the contribution of measurement error to exposure
26 estimates. First, most passive samplers developed and used for personal and indoor exposure
27 studies need to be employed for at least 24 h to collect sufficient NO2 to be detected. Therefore,
28 the majority of measurements of personal exposure concentrations done to date represents daily
29 or longer integrated or average exposure and cannot be used to assess acute, peak exposure
30 concentrations. Some newer passive samplers for nitrogen dioxide have higher uptake rates and
31 active pump samplers with traditional battery operated sampling pumps and appropriate
32 adsorbents can collect sufficient NC>2 in approximately one h and have been used in a few studies
March 2008 AX3-38 DRAFT-DO NOT CITE OR QUOTE
-------
1 providing information on exposure in microenvironments and shorter term exposure
2 concentration. Hourly fluctuations in nitrogen dioxide concentrations may be important to the
3 evaluation of exposure-health effects relationship, so continuous monitors, such as those used at
4 central site monitoring stations are still the only approach for estimating short-term exposures.
5 Second, interferences for other nitrogen oxide species can contribute to NC>2 exposure
6 monitoring errors. Both the chemiluminescence analyzer and passive samplers experience these
7 interferences but the kinetics and stoichiometry of interferent compound reactions have not been
8 well established, especially for the passive samplers. As indicated earlier, TEA-based diffusive
9 sampling methods tend to overestimate NO2 concentrations in field comparisons with
10 chemiluminescence analyzers. This could be in part the result of chemical reactions between
11 ozone and nitric oxide (NO) within the diffusion tube, leading to as much as an overestimate up
12 to 30%, or differential sensitivity to other nitrogen oxides between the passive and active
13 samplers. Due to spatially and temporally variability of NO and NO2 concentrations, especially
14 at roadsides where nitric oxide concentrations are relatively high and when sufficient ozone is
15 present for interconversion between the species, lack of agreement between the passive sampler
16 and central continuous monitor can represent differences in sampler response (Heal et al., 1999;
17 Cox, 2003). In the U.K., an alternative nitrogen dioxide monitoring plan using cost-effective and
18 simpler tube-type passive sampler has been proposed and implemented countrywide. However,
19 careful investigation of nitrogen dioxide levels revealed an overestimation, around 30% by the
20 passive sampler (Campbell et al., 1994). Another evaluation study (Bush et al., 2001) showed
21 that the overall average NO2 concentrations calculated from diffusion tube measurements were
22 likely to be within 10% of chemiluminescent measurement data.
23 Third, the effect of environmental conditions (e.g., temperature, wind speed, and
24 humidity) on the performance of passive samplers is still a concern when using it for residential
25 indoor, outdoor, and personal exposure studies, because of sampling rates that deviate from ideal
26 and can vary through the sampling period. Overall, field test results of passive sampler
27 performance are not consistent and they have not been extensively studied over a wide range of
28 concentrations, wind velocities, temperatures and relative humidities (Varshney and Singh,
29 2003). Therefore, studies directed at investigating the contributions from environmental
30 conditions to the performance of diffusive samplers in multiple locations need to be undertaken.
31
March 2008 AX3-39 DRAFT-DO NOT CITE OR QUOTE
-------
1 AX3.4 NITROGEN OXIDES IN INDOOR AIR
2
3 AX3.4.1 Indoor Sources and Concentrations of Nitrogen Oxides
4 Penetration of outdoor NO2 and combustion in various forms are the major sources of
5 NO2 to indoor environments. These environments include homes, schools, restaurants, theaters
6 etc. As might be expected, indoor concentrations of NO2 in the absence of combustion sources
7 are determined by the infiltration of outdoor NO2 (Spengler et al., 1994; Weschler et al., 1994;
8 Levy et al., 1998a), with a much smaller contribution from chemical reactions in indoor air.
9 Indoor sources of nitrogen oxides have been characterized in several reviews, namely the last
10 AQCD for Oxides of Nitrogen (U.S. Environmental Protection Agency, 1993); the Review of the
11 Health Risks Associated with Nitrogen Dioxide and Sulfur Dioxide in Indoor Air for Health
12 Canada (Brauer et al., 2002); and the Staff Recommendations for revision of the NO2 Standard in
13 California (CARB, 2007). Mechanisms by which nitrogen oxides are produced in the
14 combustion zones of indoor sources were reviewed in the last AQCD for Oxides of Nitrogen
15 (U.S. Environmental Protection Agency, 1993) and will not be repeated here. Sources of
16 ambient NO2 are reviewed in Chapter 2 of this document. It should also be noted that indoor
17 sources can affect ambient NO2 levels, particularly in areas in which atmospheric mixing is
18 limited.
19 Because most people spend most of their time indoors, personal exposure is primarily
20 determined by indoor air quality as shown in Figure AX3.4-1. Ideally, exposure to NO2 should
21 be cumulated over all indoor environments in which an individual spends time. These indoor
22 environments may include homes, schools, offices, restaurants, theaters, ice skating rinks, stores,
23 etc. However, in a study by Leaderer et al. that used two-week integrated measures,
24 concentrations of NO2 inside the home accounted for 80% of the variance in total personal
25 exposure, indicating that home concentrations are a reasonable proxy for personal exposure
26 (Leaderer etal., 1986).
27
28 Homes
29 Combustion of fossil and biomass fuels produce nitrogen oxides and the importance of such
30 sources for determining human exposures depends on how emissions are allowed to mix into
31 living areas and whether emissions are vented to the outdoors or not. Combustion of fossil fuels
32 occurs in gas-fired appliances used for cooking, heating, and drying clothes; oil furnaces;
March 2008 AX3-40 DRAFT-DO NOT CITE OR QUOTE
-------
1 kerosene space heaters; and coal stoves. Motor vehicles and various types of generators also
2 contribute in structures attached to living areas. Biomass fuels include mainly wood used in
3 fireplaces and wood stoves and tobacco.
4
5 Gas Cooking Appliances
6 A large number of studies, as described in the reviews cited above, have all noted the
7 importance of gas cooking appliances as sources of NO2 emissions. Depending on geographical
8 location, season, other sources, length of monitoring period, and household characteristics,
9 homes with gas cooking appliances have approximately 50% to over 400% higher NC>2
10 concentrations than homes with electric cooking appliances (Gilbert et al., 2006; Lee et al., 2000,
11 2002; Garcia-Algar et al., 2004; Raw et al., 2004; Leaderer et al., 1986; Garcia-Algar, 2003).
12 Gas cooking appliances remain significantly associated with indoor NC>2 concentrations after
13 adjusting for several potential confounders including season, type of community, socioeconomic
14 status, use of extractor fans, household smoking, and type of heating (Garcia-Algar et al., 2004;
15 Garrett et al., 1999).
16 Gas appliances with pilot lights emit more NC>2 than gas appliances with electronic
17 ignition. Spengler et al. (1994) found that NC>2 concentrations in bedrooms of homes with a gas
18 range without a pilot light averaged 4 ppb higher than in homes with an electric range, but were
19 15 ppb higher in homes with gas ranges with pilot lights. Lee et al. (1998) found somewhat
20 larger differences in NC>2 concentrations in homes in the Boston area, with minor seasonal
21 variation. Homes with gas stoves without pilot lights averaged between 11 ppb (summer) and
22 18 ppb (fall) higher than homes with electric stoves, while those with pilot lights averaged
23 between 19 ppb (summer) and 27 ppb (fall) higher than electric stove homes.
24 Use of extractor fans reduces NC>2 concentrations in homes with gas cooking appliances
25 (Gallelli et al., 2002; Garcia-Algar et al., 2003), although absolute NO2 levels tend to remain
26 higher than in homes with electric stoves. In a multivariate analysis, Garcia-Algar et al. (2004)
27 found that having a gas cooker remained significantly increased NC>2 concentrations even after
28 adjusting for extractor fan use. Raw et al. (2004) found only a small effect of extraction fan use
29 on NC>2 levels in the bedroom in gas cooker homes. Among homes with gas cooking, geometric
30 mean bedroom NC>2 levels were 1.7 ppb lower in homes with an extractor fan than in homes
31 without one. As expected, among homes with no fossil fuel cooking, there were no differences
32 in mean bedroom levels of NC>2 in homes with and without extractor fans.
March 2008 AX3-41 DRAFT-DO NOT CITE OR QUOTE
-------
NHAPS - Nation, Percentage Time Spent
Total n = 9,
IN A RESIDENCE iMS 7",i
TOTAL TIME SPENT
IN DOORS 2. Kodama et al. (2002) examined the associations between secondary heating sources and
1 1 NC>2 concentrations measured over a 48-h exposure period in the living rooms of homes in
12 Tokyo, Japan. They found much higher NO2 concentrations during February 1998 and January
13 1999 in homes with kerosene heaters in both southern (152.6 ppb and 139.7 ppb for 1998 and
14 1999, respectively) and northern (102.4 and 93.1 ppb for 1998 and 1999, respectively) areas of
March 2008
AX3-42
DRAFT-DO NOT CITE OR QUOTE
-------
1 Tokyo compared to homes with electric heaters (30.8 and 31.1 for the southern and 37.2 and
2 31.6 for northern areas, 1998 and 1999, respectively).
3 In a study by Garrett et al. (1999) of 78 homes in Latrobe Valley, Australia, the two
4 highest indoor NO2 levels recorded in the study were 129 ppb for the only home with an
5 unvented gas heater and 69 ppb for a home with a vented gas heater. Levels of NO2 in the
6 kitchens and living rooms of homes with a vented gas heater (mean = 6.9 ppb in living room,
7 7.3 ppb in kitchen, n = 15) were comparable to homes with gas stoves (mean = 6.7 ppb in living
8 room, 8.0 ppb in kitchen, n = 15) (Table AX3.4-1). These concentrations include results from all
9 seasons combined, so the levels are somewhat lower than those found by Triche et al. (2005) for
10 winter monitoring periods only.
11 Triche et al. (2005) also found high levels of NO2 in homes with gas space heaters,
12 although information on whether the appliance was vented or unvented was not available. Data
13 from this study were analyzed in more detail and are shown in Table AX3.4-2. The median NO2
14 concentration in the 6 homes with gas space heater use during monitoring periods with no gas
15 stove use was 15.3 ppb; a similar incremental increase in total NO2 levels was noted for homes
16 with gas space heater use during periods when gas stoves were also used (Median = 36.6 ppb)
17 compared to homes where gas stoves were used but no secondary heating sources were present
18 (Median = 22.7 ppb) (Table AX3.4-2).
19 Shima and Adachi (1998) examined associations between household characteristics,
20 outdoor NO2, and indoor NO2 in 950 homes during the heating season (640 with unvented and
21 310 vented heaters) and 905 homes during the non-heating season in urban, suburban, and rural
22 areas of Japan. While no information is provided on gas stove use, the authors note that nearly
23 all homes in Japan have gas stoves, though relatively few have pilot lights. During the heating
24 season, geometric mean NO2 levels in homes with unvented heaters (66.4 ppb) are about three
25 times higher than in homes with vented heaters (20.6 ppb). In the non-heating season, the mean
26 levels were lower at only 13.8 ppb, suggesting a contribution from vented heaters as well.
27 In multivariate analyses, Gilbert et al. (2006) found that gas and mixed/other heating
28 systems were significantly associated with NO2 levels, adjusting for presence of gas stoves and
29 air exchange rates in 96 homes in Quebec City, Canada during the winter/early spring period.
30 Many homes with gas space heaters also have gas stoves, and the contribution from multiple
31 sources is much higher than from any single source alone (Garrett et al., 1999). In the Garrett
March 2008 AX3-43 DRAFT-DO NOT CITE OR QUOTE
-------
1 et al. (1999) study, homes were classified into five categories: no indoor source (n = 15), gas
2 stove only (n = 15), gas heater only (n = 14), smoker in the household only (n = 7), and multiple
3 sources (n = 29). Homes with multiple sources had much higher NO2 concentrations homes with
4 either a gas stove only or gas heater only (Table AX3.4-3).
5 Kerosene heaters are also important contributors to indoor NO2 levels. Leaderer et al.
6 (1986) enrolled a cohort of kerosene heater users identified from local kerosene dealers and a
7 cohort of controls systematically chosen from the same neighborhoods with each matched pair
8 treated as a sampling unit (i.e., sampled at the same randomly assigned time period). A total of
9 302 homes were monitored for at least one two-week period. While outdoor concentrations
10 never exceeded 100 |ig/m3 (53 ppb), approximately 5% of homes with either no gas but
11 1 kerosene heater or gas but no kerosene heater had levels exceeding 53 ppb. Between
12 17%-33% of homes with both gas and kerosene heater(s) exceeded this limit, while nearly one
13 quarter of homes with no gas, but two or more kerosene heaters had these levels.
14 Data from Triche et al. (2005) (Table AX3.4-2) also indicated increased levels of NO2 for
15 kerosene heater homes during monitoring periods with no gas stove use (Median = 18.9 ppb)
16 compared to homes with no sources (Median = 6.3 ppb), which is similar to levels found in
17 homes using gas space heaters (Median =15.3 ppb). However, these NO2 concentrations are of
18 the same magnitude as those in homes with gas stove use (Median = 17.2 ppb).
19 Data are available for unvented gas hot water heaters from a number of studies conducted
20 in the Netherlands. Results summarized by Brauer et al. (2002) indicate that concentrations of
21 NO2 in homes with unvented gas hot water heaters were 10 to 21 ppb higher than in homes with
22 vented heaters, which in turn, had NO2 concentrations 7.5 to 38 ppb higher than homes without
23 gas hot water heaters.
24 The contribution from combustion of biomass fuels has not been studied as extensively as
25 that from gas. A main conclusion from the previous AQCD was that properly vented wood
26 stoves and fireplaces would make only minor contributions to indoor NO2 levels. Several studies
27 conclude that use of wood burning appliances does not increase indoor NO2 concentrations.
28 Levesque et al. (2001) examined the effects of wood-burning appliances on indoor NO2
29 concentrations in 49 homes in Quebec City, Canada. The homes, which had no other
30 combustion source, were sampled for 24 h while the wood-burning appliance was being used.
31 No significant differences in mean NO2 levels were found in homes with (6.6 + 3.6 ppb) and
March 2008 AX3-44 DRAFT-DO NOT CITE OR QUOTE
-------
1 without (8.8 + 1.9 ppb) a wood-burning appliance. Data from Triche et al. (2005) confirm these
2 findings (Table AX3.4-2). Homes with wood burning sources had comparable NC>2
3 concentrations to homes without other secondary heating sources, with (Median = 5.9 ppb) and
4 without (Median = 16.7 ppb) gas stove use.
5 Table AX3.4-3 shows short-term average (minutes to a few hours) concentrations of NC>2
6 in homes with combustion sources. The concentrations represent those found in different rooms
7 in houses sampled. However, concentrations are much higher in those persons directly exposed
8 to emissions. For example, Dennekamp et al. (2001) found NCh concentrations of about 1 ppm
9 at face level in front of a 4-burner gas range. Table AX3.4-4 shows long-term average (24-h to
10 2 week) concentrations of NC>2 in homes with combustion sources (mainly gas fired).
11 Data are available for unvented gas hot water heaters from a number of studies conducted
12 in the Netherlands. Results summarized by Brauer et al. (2002) indicate that concentrations of
13 NC>2 in homes with unvented gas hot water heaters were 10 to 21 ppb higher than in homes with
14 vented heaters, which in turn, had NC>2 concentrations 7.5 to 38 ppb higher than homes without
15 gas hot water heaters.
16 As can be seen from the tables, shorter-term average concentrations tend to be much
17 higher than longer term averages. However, as Triche et al. (2005) point out, the 90th percentile
18 concentrations can be substantially greater than the medians, even for two week long samples.
19 This finding illustrates the high variability found among homes. This variability reflects
20 differences in ventilation of emissions from sources, air exchange rates, the size of rooms etc.
21 The concentrations for short averaging periods that are listed in Table AX3.4-3 correspond to
22 about 10 to 30 ppb on a 24-h average basis. As can be seen from inspection of Table AX3.4-4,
23 these sources would contribute significantly to the longer term averages reported there if
24 operated on a similar schedule on a daily basis. This implies that measurements made with long
25 averaging periods may not capture the nature of the diurnal pattern of indoor concentrations in
26 homes with strong indoor sources. This problem becomes more evident as ambient NC>2 levels
27 decrease due to more efficient controls on outdoor sources.
28 In 10% of homes with fireplaces studied by Triche et al. (2005), NO2 concentrations were
29 greater than or equal to 80 ppb, or about twice the level found in homes with no indoor
30 combustion source (see Figure AX3.30). In a study of students living in Copenhagen, S0rensen
31 et al. (2005) found that personal exposures to NC>2 were significantly associated with time
March 2008 AX3-45 DRAFT-DO NOT CITE OR QUOTE
-------
1 exposed to burning candles in addition to other sources. However, they did not provide data for
2 concentrations in spaces in which candles were burned. Results of studies relating NO2
3 concentrations and exposures to environmental tobacco smoke (ETS) have been mixed. Several
4 studies found positive associations between NO2 levels and ETS (e.g., Linaker et al., 1996);
5 Farrow et al., 1997; Aim et al., 1998; Levy et al., 1998a; Monn et al., 1998; Cyrys et al., 2000;
6 Lee et al., 2000; Garcia-Algar et al., 2004) whereas others have not (e.g., Hackney et al., 1992;
7 Kawamoto et al., 1993). In a study of 57 homes in Brisbane, Australia (Lee et al., 2000), levels
8 of NC>2 were higher in homes with smokers present (14.9 + 7.7 ppb) than without smokers (9.9 +
9 5.0 ppb). However, these concentrations did not account for presence of a gas range (n = 18 of
10 57 homes had a gas range). Garrett et al. (1999) found that smoking in the home increased levels
11 of NC>2 in the winter, but not in the summer when windows tended to be opened. In a study of
12 students living in Copenhagen, S0rensen et al. (2005) did not find a significant association
13 between ETS and personal exposures to NC>2. However, they found that burning candles was a
14 significant prediction of bedroom levels of NC>2.
15
16 Other Indoor Environments
17 Indoor ice skating rinks have been cited as environments containing high levels of NC>2
18 when fuel powered ice resurfacing machines are used especially without ventilation. As part of a
19 three year study, Levy et al. (1998b) measured NC>2 concentrations at 2 locations at the outside of
20 the ice surface in 19 skating rinks in the Boston area over 3 winters. Although different passive
21 samplers were used in the first year (Palmes tubes, 7 day sampling time) and in years 2 and
22 3 (Yanagisawa badges, 1 day working hours) of the study, consistently high mean NO2
23 concentrations were associated with the use of propane fueled resurfacers (248 ppb in the first
24 year and 206 ppb in the following years) and gasoline fueled resurfacers (54 ppb in the first year
25 and 132 ppb in the following years) than with electric resurfacers (30 ppb in the first year and
26 37 ppb in the following years). During all three years of the study peak NC>2 concentrations were
27 several times higher in the rinks with propane and gasoline fueled resurfacers than the values
28 given above. A number of earlier studies have also indicated NC>2 concentrations of this order
29 and even higher (Paulozzi et al., 1993; Berglund et al., 1994; Lee et al., 1994; Brauer et al.,
30 1997). In these studies peak averages were in the range of a few ppm.
31
March 2008 AX3-46 DRAFT-DO NOT CITE OR QUOTE
-------
1 AX3.4.2 Reactions of NO2 in Indoor Air
2 Chemistry in indoor settings can be both a source and a sink for NC>2 (Weschler and
3 Shields, 1997). NC>2 is produced by reactions of NO with ozone or peroxy radicals, while NC>2 is
4 removed by gas phase reactions with ozone and assorted free radicals and by surface promoted
5 hydrolysis and reduction reactions. The concentration of indoor NC>2 also affects the
6 decomposition of peroxyacyl nitrates. Each of these processes is discussed in the following
7 paragraphs. They are important not only because they influence the indoor NC>2 concentrations
8 to which humans are exposed, but also because certain products of indoor chemistry may
9 confound attempts to examine associations between NC>2 and health.
10 Indoor NO can be oxidized to NO2 by reaction with ozone or peroxy radicals; the latter
1 1 are generated by indoor air chemistry involving Os and unsaturated hydrocarbons such as
12 terpenes found in air fresheners and other household products (Sarwar et al., 2002a,b; Nazaroff
13 and Weschler, 2004; Carslaw, 2007). The rate coefficient for the reaction
14
15 at room temperature (298 K) is 1.9 x 1CT14 cmVmolec-sec or 4.67 x 10~4 ppb"1 s"1 (Jet
16 Propulsion Laboratory, 2006). At an indoor Os concentration of 10 ppb and an indoor NO
17 concentration that is significantly less than that of Os, the half-life of NO is 2.5 min. This
18 reaction is sufficiently fast to compete with even relatively fast air exchange rates. Hence, the
19 amount of NO2 produced from NO tends to be limited by the amount of Os available. The
20 indoor concentrations of NO and Os are negatively correlated; significant concentrations of NO
21 can only accumulate when small amounts of Os are present and vice versa (Weschler et al.,
22 1994).
23 The rapid reaction between NO and Os also means that humans, themselves, can be
24 indirect sources of NO2 in the rooms they occupy. Exhaled human breath contains NO that is
25 generated endogenously (Gustafsson et al., 1991). For a typical adult male, the average nasal
26 NO output is 325 nL min"1 or 23.9 jig IT1 (Imada et al., 1996). If ozone is present in the indoor
27 air, some or all of these exhaled NO molecules will be oxidized to NO2. To put this source in
28 perspective, consider the example of an adult male in a 30 m3 room ventilated at 1 air change per
29 hour (h-1) with outdoor air. The steady-state concentration of NO in the room as a consequence
30 of NO in exhaled breath is 0.80 jig m3 or 0.65 ppb if none of the NO were to be oxidized.
March 2008 AX3-47 DRAFT-DO NOT CITE OR QUOTE
-------
1 However, assuming a meaningful concentration of ozone in the ventilation air (>5 ppb), most of
2 this NO is oxidized to NO2 before it is exhausted from the room. In this scenario, the single
3 human occupant is indirectly a source for 0.65 ppb of NO2 in the surrounding air. At higher
4 occupant densities, lower air exchange rates and elevated concentrations of Os in the ventilation
5 air, human exhaled breath could contribute as much as 5 ppb to the total concentration of indoor
6 NO2.
7 The reaction of NO2 with ozone produces nitrate radicals
8 (AX34_2)
9 The second order rate-constant for this reaction at room temperature (298 K) is
10 3.2 x 1Q~17 cmVmolec-sec or 7.9 x 10~7 ppb"1 s"1 (Jet Propulsion Labatory, 2006). For indoor
1 1 concentrations of 20 ppb and 30 ppb for 63 and NO2, respectively, the production rate of
12 NOs radicals is 1.7 ppb h"1. This reaction is strongly temperature dependent, an important
1 3 consideration given the variability of indoor temperatures with time of day and season. The
14 nitrate radical is photolytically unstable (Finlayson-Pitts and Pitts, 2000). As a consequence,
15 it rapidly decomposes outdoors during daylight hours. Indoors, absent direct sunlight, nitrate
16 radical concentrations may approach those measured during nighttime hours outdoors. To date
17 there have been no indoor measurements of the concentration of nitrate radicals in indoor
18 settings. Modeling studies by Nazaroff and Cass (1986), Weschler et al. (1992), Sarwar et al.
19 (2002b), and Carslaw (2007) estimate indoor nitrate radical concentrations in the range of 0.01 to
20 5 ppt, depending on the indoor levels of 63 and NO2.
21 The nitrate radical and NO2 are in equilibrium with dinitrogen pentoxide (N2Os):
22 (AX34.3)
23 Dinitrogen pentoxide reacts with water to form nitric acid. The gas phase reaction with water is
24 too slow (Sverdrup et al., 1987) to compete with air exchange rates in most indoor environments.
25 Due to mass transport limits on the rate at which N2Os is transported to indoor surfaces, reactions
26 of N2Os with water sorbed to indoor surfaces are much slower than gas phase reactions between
27 nitrate radicals and commonly occurring indoor alkenes.
28 Once formed, NOs radicals can oxidize organic compounds by either adding to an
29 unsaturated carbon bond or abstracting a hydrogen atom (Wayne et al., 1991). In certain indoor
March 2008 AX3-48 DRAFT-DO NOT CITE OR QUOTE
-------
1 settings, the nitrate radical may be a more important indoor oxidant than either ozone or the
2 hydroxyl radical. Table 8 in Nazaroff and Weschler (2004) illustrates this point. Assuming
3 indoor concentrations of 20 ppb, 5 x 10~6 ppb, and 0.001 ppb for Os, OH, and NOs, respectively,
4 the pseudo first-order rate constants for reactions of most terpenoids are larger for reactions with
5 NOs than for reactions with either Os or OH. For example, for the stated conditions, the half-
6 lives of d-limonene and a-pinene are roughly three times shorter as a consequence of reaction
7 with NOs versus reaction with 03. The products of reactions between NOs and various organic
8 compounds include nitric acid, aldehydes, ketones, organic acids and organic nitrates; these have
9 been summarized by Wayne et al. (1991). Nitrate radicals and the products of nitrate radical
10 chemistry may be meaningful confounders in NO2 exposure studies.
11 Reactions between NO2 and various free radicals can be an indoor source of organo-
12 nitrates, analogous to the chain-terminating reactions observed in photochemical smog
13 (Weschler and Shields, 1997). Additionally, based on laboratory measurements and
14 measurements in outdoor air (Finlayson-Pitts and Pitts, 2000), one would anticipate that NO2,
15 in the presence of trace amounts of HNOs, can react with PAHs sorbed on indoor surfaces to
16 produce mono- and dinitro-PAHs.
17 As noted earlier in Chapter 2, HONO occurs in the atmosphere mainly via multiphase
18 processes involving NO2. HONO is observed to form on surfaces containing partially oxidized
19 aromatic structures (Stemmler et al., 2006) and on soot (Ammann et al., 1998). Indoors, surface-
20 to-volume ratios are much larger than outdoors, and the surface mediated hydrolysis of NO2 is a
21 major indoor source of HONO (Brauer et al., 1990; Febo and Perrino, 1991; Spicer et al., 1993;
22 Brauer et al., 1993; Spengler et al., 1993; Wainman et al., 2001; Lee et al., 2002). Spicer et al.
23 (1993) made measurements in a test house that demonstrated HONO formation as a consequence
24 of NO2 surface reactions and postulated the following mechanism to explain their observations
25 2NO2 + H2O/stirface -> HONO(aq) + H++NO3~
26 HONO(aq) o HONO(g) (AX3.4-5)
27 In a series of chamber studies, Brauer et al. (1993) reported HONO formation as a consequence
28 of NO2 surface reactions and further reported that HONO production increased with increasing
29 relative humidity. Wainman et al. (2001) confirmed Brauer's findings regarding the influence of
March 2008 AX3-49 DRAFT-DO NOT CITE OR QUOTE
-------
1 relative humidity. They also found that NC>2 removal and concomitant HONO production was
2 greater on synthetic carpet surfaces compared to Teflon surfaces, and that the affinity of a
3 surface for water influences HONO's desorption from that surface. Lee et al. (2002) measured
4 HONO and NO2 concentrations in 119 Southern California homes. Average indoor HONO
5 levels were about 6 times larger than outdoors (4.6 ppb versus 0.8 ppb). Indoor HONO
6 concentrations averaged 17% of indoor NO2 concentrations, and the two were strongly
7 correlated. Indoor HONO levels were higher in homes with humidifiers compared to homes
8 without humidifiers (5.9 ppb versus 2.6 ppb). This last observation is consistent with the studies
9 of Brauer et al. (1993) and Wainman et al. (2001) indicating that the production rate of HONO
10 from NO2/surface reactions is larger at higher relative humidities. Based on detailed laboratory
11 studies, the hydrolysis mechanism, Equations AX3.4-4 and AX3.4-5, have been refined.
12 Finlayson-Pitts et al. (2003) hypothesize that the symmetric form of the NO2 dimer is sorbed on
13 surfaces, isomerizes to the asymmetric dimer which auto ionizes to NO^Os ; the latter then
14 reacts with water to form HONO and surface adsorbed HNOs. FTIR-based analyses indicate that
15 the surface adsorbed HNOs exists as both undissociated nitric acid-water complexes,
16 (HNO3)x(H2O)Y, and nitrate ion-water complexes, (NO3 )x(H2O)Y (Dubowski et al., 2004,
17 Ramazan et al., 2006). Such adsorbed species may serve as oxidizing agents for organic
18 compounds sorbed to these same surfaces (Ramazan et al., 2006).
19 HONO and much smaller amounts of HNO3 are also emitted directly by combustion by
20 gas appliances and can infiltrate from outdoors. Spicer et al. (1993) compared the measured
21 increase in HONO in a test house resulting from direct emissions of HONO from a gas range and
22 from production by surface reactions of NO2. They found that emissions from the gas range
23 could account for about 84% of the measured increase in HONO and surface reactions for 11%
24 in an experiment that lasted several hours. An equilibrium between adsorption of HONO from
25 the gas range (or other indoor combustion sources) and HONO produced by surface reactions
26 (see Equation AX3.4-5) also determines the relative importance of these processes in producing
27 HONO in indoor air. In a study of Southern CA homes (Lee et al., 2002), indoor levels of NO2
28 and HONO were positively associated with the presence of gas ranges.
29 It is known that the photolysis of HONO (g) in the atmosphere (outdoors) is a major
30 source of the hydroxyl radical (OH). Given high indoor HONO concentrations and the presence
31 of lighting (sun light penetrating windows, incandescent lights, fluorescent lights), the photolysis
March 2008 AX3-50 DRAFT-DO NOT CITE OR QUOTE
-------
1 of indoor HONO may be a meaningful source of indoor hydroxyl radical, under favorable
2 reaction conditions. Given the large suite of man-made chemicals present indoors at elevated
3 concentrations, indoor free radicals (e.g., OH and NO3) can initiate and drive a complex series of
4 indoor chemical reactions.
5 NO2 can also be reduced on certain surfaces, forming NO. Spicer et al. (1989) found that
6 as much as 15% of the NO2 removed on the surfaces of masonite, ceiling tile, plywood,
7 plasterboard, bricks, polyester carpet, wool carpet, acrylic carpet and oak paneling was re-
8 emitted as NO. Weschler and Shields (1996) found that the amount of NO2 removed by charcoal
9 building filters were almost equally matched by the amount of NO subsequently emitted by these
10 same filters.
11 Spicer et al. (1993) determined the 1st order rate constants for removal of several NOY
12 components by reaction with indoor surfaces. They found lifetimes (e-folding times) of about
13 half an hour for HNOs, an hour for NO2, and hours for NO and HONO. Thus the latter two
14 components, if generated indoors are more likely to be lost to the indoor environment through
15 exchange with outside air than by removal on indoor surfaces. However, HONO is in
16 equilibrium with the nitrite ion (NO2 ) in aqueous surface films
17 HONO(ag) ^ H* + NOf (AX3 4_6)
18 Ozone oxidation of nitrite ions in such films is a potential sink for indoor HONO (Lee et al.,
19 2002).
20 Jakobi and Fabian (1997) measured indoor and outdoor concentrations of ozone and
21 peroxyacetyl nitrate (PAN) in several offices, private residences, a classroom, a gymnasium and
22 a car. They found that indoor levels of PAN were 70% to 90% outdoor levels, and that PAN's
23 indoor half-life ranged from 0.5 to 1 h. The primary indoor removal process is thermal
24 decomposition
25
CH3C(O)OONO2 <-> CH3C(O)OO + NO2
*" ~"
26 As is indicated by Equation AX3.4-7, PAN is in equilibrium with the peroxylacetyl radical and
27 NO2. Hence, the indoor concentration of NO2 affects the thermal decomposition of PAN and,
28 analogously, other peroxyacyl nitrates. Peroxylalkyl radicals rapidly oxidize NO to NO2, so the
March 2008 AX3-5 1 DRAFT-DO NOT CITE OR QUOTE
-------
1 indoor concentration of NO also influences the thermal decomposition of PAN type species
2 (Finlayson-Pitts and Pitts, 2000).
3 Reactions between hydroxyl radicals and aldehydes in the presence of NO2 can lead to
4 the formation of peroxyacyl nitrates. Weschler and Shields (1997) have speculated that such
5 chemistry may sometimes occur indoors. For example, the requisite conditions for the formation
6 of the highly irritating compound peroxybenzoyl nitrate may occur when ozone, certain terpenes,
7 styrene and NO2 are present simultaneously at low air exchange rates. This relatively common
8 indoor mixture of pollutants produces hydroxyl radicals and benzaldehyde, which can
9 subsequently react as noted above. In her detailed model of indoor chemistry, Carslaw (2007)
10 explores the indoor formation of PAN-type species (see Figure 2 in the cited reference).
11 Recent work indicates that indoor NO2 also can affect the formation of secondary organic
12 aerosols (SOA) resulting from the reaction of O3 with terpenes such as d-limonene and a-pinene
13 (N0jgaard et al., 2006). At concentrations of 50 ppb for O3 and the terpenes, NO2 decreased the
14 formation of SOA compared to the levels formed in the absence of NO2. The effect was more
15 pronounced for SOA derived from a-pinene than d-limonene, and at lower NO2 concentrations,
16 appears to be explained by the O3 loss resulting from its reaction with NO2. The resultant nitrate
17 radicals apparently are not as efficient at producing SOA as the lost O3.
18 Nitro-PAHs have been found in indoor environments (Mumford et al., 1991; Wilson
19 et al., 1991). The major indoor sources of nitro-PAHs include cooking, wood burning, and the
20 use of kerosene heater (World Health Organization (WHO), 2003). It is also likely that nitro-
21 PAHs outdoors can infiltrate indoors. One of the potential sources of nitro-PAHs indoors, which
22 has not been characterized, is reactions via indoor chemistry. The reactions of PAHs with OH
23 and NO3 may occur in indoor environments. Although no direct measurements of OH or NO3 in
24 indoor environments, OH and NO3 can be formed via indoor chemistry and may present at
25 significant levels indoors (Nazaroff and Cass 1986, Sarwar et al., 2002a; Carslaw, 2007).
26 Concentrations of ~10~6 ppb for OH and 0.01-5 ppt of NO3 have been predicted through indoor
27 chemical reactions (Nazaroff and Cass 1986, Sarwar et al., 2002a, Carslaw, 2007), depending on
28 the indoor levels of O3, alkenes, and NO2. Observation of secondary organic aerosols (SOA)
29 formation in a simulated indoor environment also suggested that ~10~5 ppb steady-state OH
30 radicals were generated from the reactions of O3 with terpenes (Fan et al., 2003). PAHs are
31 common indoor air pollutants (Chuang et al., 1991; Naumova et al., 2002), and the
March 2008 AX3-52 DRAFT-DO NOT CITE OR QUOTE
-------
1 concentrations of some PAHs indoors are often higher than outdoors (Naumova et al., 2002).
2 Therefore, the reactions of OH and NOs with PAHs may occur at rates comparable to air
3 exchange rates to form nitro-PAHs indoors. In addition, the reactions of NOs with PAHs may be
4 more significant indoors than outdoors because indoor NOs is more stable due to the low uv in
5 indoor environments. Given the high surface areas available indoors, the formation of nitro-
6 PAHs via surface reactions of PAHs with nitrating species may be more important compared to
7 heterogeneous reactions outdoors.
8 In summary, indoor chemistry can meaningfully alter the indoor concentration of NC>2.
9 Indoor exposure to NC>2 may be accompanied by indoor exposures to nitrate radicals, organic
10 nitrates, and nitro-PAHs.
11
12 AX3.4.3 Contributions from Outdoor NO2
13 As might be expected, indoor concentrations of NC>2 in the absence of combustion
14 sources are primarily determined by outdoor NC>2 concentrations (Spengler et al., 1994;
15 Weschler et al., 1994; Levy et al., 1998a), with a much smaller contribution from chemical
16 reactions in indoor air.
17 The exchange between NC>2 in ambient air and in the indoor environment is influenced by
18 infiltration (air leakage), natural ventilation (air flow through intentional openings such as
19 windows), and mechanical ventilation (rarely used in residences) (Yang et al., 2004).
20 In temperate climates, winter is associated with lower indoor/outdoor ratios of NC>2 since
21 windows and doors are usually tightly closed and the only source of exchange is infiltration.
22 Newer homes tend to be built more tightly than older homes, so have even lower rates of
23 infiltration. During warmer weather, air conditioner use and opening of windows increase air
24 exchange between outdoors and indoors.
25 Yang et al. (2004) used multiple integrated (7-day) NC>2 measurements indoors and
26 outdoors to calculate penetration and source strength factors in Seoul, Korea and Brisbane,
27 Australia using a mass balance model considering a residence as a single chamber (Yang et al.,
28 2004). They showed that, while penetration factors did not differ significantly between gas and
29 electric range homes, source strength factors were much higher in homes with gas ranges in both
30 Brisbane and Seoul (5.77 ± 3.55 and 9.12 ± 4.50, respectively) than in electric range homes in
31 Brisbane (1.49 ± 1.25). Similarly, calculated NC>2 source strengths (|ig/m3/h) were
March 2008 AX3-53 DRAFT-DO NOT CITE OR QUOTE
-------
1 21.9 ± 21.8 and 44.7 ± 38.1 in gas homes in Brisbane and Seoul, respectively, and 6.6 ± 6.3 in
2 electric homes in Brisbane.
3
4 Household Characteristics
5 Yang et al. (2004) found that levels of indoor NO2 (in |ig/m3) were associated with house
6 characteristics in 28 homes in Brisbane (where there were both electric and gas range homes).
7 Homes with a gas water heater had higher levels than those without (34.5 ± 16.4 versus 22.8 ±
8 12.1, p = 0.048), but these were unadjusted associations, and it is likely that many of the homes
9 with gas water heaters also had gas ranges. Homes with an attached garage had higher levels of
10 NO2 (33.1 ± 18.3) compared to homes without one (21.8 ± 8.8) (p = 0.039). Attached garages
11 were not, however, associated with NO2 levels in a study in Quebec City, Canada (Gilbert et al.,
12 2006). The authors suggested that the lack of association might be attributed to small numbers
13 (n = 18 homes with attached garages) or to the airtightness of homes in Canada compared to
14 those in Australia.
15 Location in a city center was associated with higher NO2 levels in homes in Menorca
16 (one of the Balearic Islands off the coast of Spain with rural and small town residences), after
17 adjusting for gas cooker, extractor fan use, smoking in the home, type of central heating, season,
18 and social class (Garcia-Algar et al., 2004). In the same study, levels of indoor NO2 in
19 Barcelona (a large coastal city in Spain) and Ashford (a medium-sized town in the southeast UK)
20 were significantly higher than those in Menorca
21 In a study of a random sample of 845 homes in England (Raw et al., 2004), levels of NO2
22 were significantly associated with dwelling type and age of home, but the authors attributed
23 these effects to the geographical location of the home (e.g., inner city). Garrett et al. (1999) also
24 found that age of house was significantly associated with NO2 levels in winter and summer. In
25 the study by Shima and Adachi, (1998), differences in concentrations of NO2 between homes
26 with and without unvented heaters in the heating season were slightly lower among homes with
27 wood compared to aluminum window frames. Type of window frames, but not structure type,
28 was associated with NO2 concentrations in the heating period for homes with unvented heaters
29 (76.2 ±1.4 ppb versus 55.9 ±3.9 ppb in homes with aluminum and wood windows,
30 respectively), but not in homes with vented heaters. In the non-heating season, mean NO2 levels
31 in the home varied by type of structure (steel/concrete or wood) and type of window frames
32 (aluminum or wood), with wood structures and frames indicating a less airtight dwelling.
March 2008 AX3-54 DRAFT-DO NOT CITE OR QUOTE
-------
1 AX3.5 PERSONAL EXPOSURE
2
3 Components of Personal Exposure
4 Human exposure to NC>2 consists of contact at the air boundary layer between the human
5 and the environment at a specific concentration for a specified period of time. People spend
6 various amount of time in different microenvironments with various NC>2 concentrations. The
7 integrated NO2 exposure is the sum of the individual NO2 exposures over all possible time
8 intervals for all environments. Therefore, the assessment of human exposures to NC>2 can be
9 represented by the following equation:
10 i=l (AX3.5-1)
11 where ET is the time-weighted personal exposure concentration over a certain period of time, n is
12 the total number of environments that a person encounters,/ is the fraction of time spent in the
13 /'th environment, and C\ is the average NC>2 concentration in the rth environment during the time
14 fraction/. Depending upon the time fraction and environmental concentration we consider
15 during exposure assessment, the exposure a person experiences can be classified into
16 instantaneous exposure, peak exposure, averaged exposure, or integrated exposure. These
17 distinctions are important because health effects caused by long-term low-level exposures may
18 be different from those resulting from short-term peak exposures.
19 The equation above represents the average personal exposure concentration is a linear
20 combination of the average concentration in the ambient environment and each
21 microenvironment, weighted by an individual's fraction of time spent in that environment.
22 Hence, personal exposure to NC>2 is influenced by the microenvironmental concentration and the
23 amount of time spent in each microenvironment. In theory, a microenvironment could be any
24 three-dimensional space having a volume in which people spend a certain amount of time.
25 In practice, microenvironments typically used to determine NC>2 exposures include residential
26 indoor environment, other indoor locations, near-traffic outdoor environment, other outdoor
27 locations, and in-vehicles. In other words, total personal exposure to NO2 can be decomposed
28 into exposure to NC>2 in different environments. An individual's total exposure (Ex) can also be
29 represented by the following equation
March 2008 AX3-55 DRAFT-DO NOT CITE OR QUOTE
-------
ET = Ea
2 subject to the constraint
10
15
,wna
(AX3.5-2)
3 " / (AX3.5-3)
4 where Ea is the person's exposure to pollutants of ambient origin; Enona is the person's exposure
5 to pollutants that are not of ambient origin; y0 is the fraction of time people spend outdoors and yi
6 is the fraction of time they spend in microenvironment /'; F-m^ P[, a;, and k; are the infiltration
7 factor, penetration coefficient, air exchange rate, and decay rate for microenvironment /'.
8 In the case where microenvironmental exposures are dominated by one
9 microenvironment, Equation AX3.5-2 may be approximated by
nona
{ v
-y)[Pa/(a + k)] } Ca + Emma = aCa
mma
1 1 where Et is the total personal exposure, Ea is the exposure to ambient generated pollutants, Enon!lg
12 is the nonambient generated pollutants, andy is the time fraction people spent outdoors. Other
13 symbols have the same definitions in Equation AX. 5-2. If microenvironmental concentrations
14 are considered, then Equation AX3.5-5 can be recast as
Cme = Ca+ Cna = [Pa /(a + k)}Ca + S/{V(a
(AX3.5-5)
16 where Cme is the concentration in a microenvironment; Ca and Cnona the contributions to Cme from
17 ambient and nonambient sources; S is the microenvironmental source strength; Fis the volume
18 of the microenvironment, and the symbols in brackets have the same meaning as in Equation
19 AX3.5-5. In this equation, it is assumed that microenvironments do not exchange air with each
20 other, but only with ambient air.
21 The NC>2 concentration in each microenvironment can show substantial spatial and
22 temporal variability, which is determined by many factors, such as season, day of the week,
23 personal age, occupation, house characteristics, personal activities, source emission rate, air
24 exchange rate, and transport and removal mechanisms of NC>2. Failure to disaggregate total
25 human exposure and assess human exposure in various microenvironments may result in
26 exposure misclassification, which may obscure the true relations between ambient air pollution
27 and health outcomes.
March 2008
AX3-56
DRAFT-DO NOT CITE OR QUOTE
-------
1 Studies reviewed in this section were generally conducted in North America (Canada, the
2 United States, and Mexico) and European countries. Studies conducted in other parts of the
3 world were not the primary focus of this science review because exposure patterns may not be
4 similar to those in the United States. However, studies which might support general conclusions
5 (not country or cultural specific conclusions) about NC>2 exposures will be included.
6 Either Palmes tubes or Yanagisawa badges or Ogawa samplers were used to measure
7 personal exposures in most of the reviewed studies, and sometimes residential indoor and
8 outdoor concentrations. Sampling time for each cartridge varied from 8 h to two weeks, and the
9 study design covered (1) longitudinal, in which each subject is measured for many days;
10 (2) pooled, in which each subject is measured for only one or two days, different days for
11 different subjects; and (3) daily-average, in which many subjects are measured on the same day.
12 Most studies focused primarily on children, and in some studies adults or people with respiratory
13 diseases were taken as study population.
14
15 AX3.5.1 Personal Exposures and Ambient (Outdoor) Concentrations
16 Numerous epidemiological studies have shown a positive association between ambient
17 (outdoor) NC>2 concentrations and adverse health effects. Since a causal association requires
18 exposure, it is very important to evaluate personal exposure to ambient (outdoor) generated NC>2.
19 In this section, topics related to the total personal exposure and ambient (outdoor) generated NC>2
20 will be evaluated, such as the levels of personal exposure and ambient (outdoor) NC>2, the
21 attenuation factor of personal exposure to NC>2, the correlation between personal and ambient
22 (outdoor) NC>2, and the factors determining the associations between personal exposure and
23 ambient (outdoor) level. Based on the science review, the following key questions will be
24 addressed: (1) When, where, how and how much are people exposed to ambient (outdoor)
25 generated NC>2? and (2) Is ambient (outdoor) NC>2 a good surrogate for personal total exposure
26 or personal exposure to ambient (outdoor) NC>2?
27 Personal exposures in most of the studies considered here were less than the
28 corresponding outdoor or ambient concentrations. In the presence of local sources (indoor or
29 local traffic sources), personal exposure levels could be higher than outdoor or ambient levels
30 (Spengler et al., 1994; Nakai et al., 1995; Linn et al., 1996; Spengler et al., 1996; Raaschou-
31 Nielsen et al., 1997; Aim et al., 1998; Levy et al., 1998a; Monn et al., 1998; Liard et al., 1999;
March 2008 AX3-57 DRAFT-DO NOT CITE OR QUOTE
-------
1 Kramer et al., 2000; Linaker et al., 2000; Mukala et al., 2000; Gauvin et al., 2001; Monn, 2001;
2 Rotko et al., 2001; Sarnat et al., 2001; Kodama et al., 2002; Mosqueron et al., 2002; Ramirez-
3 Aguilar et al., 2002; Rojas-Bracho et al., 2002; Lai et al., 2004; Nerriere et al., 2005; Sarnat
4 et al., 2005; S0rensen et al., 2005; Kim et al., 2006; Sarnat et al., 2006).
5 In a probability based population exposure study in Los Angeles Basin, 48 h indoor,
6 outdoor and personal exposures (pooled exposures) were reported for 682 participants (Spengler
7 et al., 1994). Spengler et al. (1994) found that the median personal exposure was 35 ppb and the
8 median outdoor level was 36 ppb. Linn et al. (1996) reported the results of a personal exposure
9 study for 269 school children from three Southern California communities. During this
10 longitudinal study, 24 h averaged personal exposures, as well as inside school, outside school
11 and ambient central site NO2 levels, were measured by Yanagisawa badges for one week for
12 each season from 1992 to 1994. Results showed that mean personal exposure was 22 ppb and
13 the mean central site concentration was 37 ppb. Kim et al. (2006) conducted a longitudinal,
14 multi-pollutant exposure study in Toronto, Canada. During the study, personal exposures (24-h
15 integrated by Ogawa sampler) to PM2 5, NO2 and CO were measured for 28 subjects with
16 coronary artery disease one day a week for a maximum of 10 weeks, and were compared with
17 ambient fixed site measurements. The mean NO2 personal exposure was 14.4 ppb, which was
18 lower than the ambient site concentrations (20-26 ppb). Sarnat et al. (2001) and Sarnat et al.
19 (2005) reported multi-pollutant exposure studies in Baltimore and Boston. In the Baltimore
20 study, 24 h averaged personal exposure and ambient PM2.5, O3, NO2, SO2, and CO were
21 measured for 56 subjects (20 older adults, 21 children and 15 individuals with COPD) in the
22 summer of 1998 and the winter of 1999. All subjects were monitored for 12 or 8 consecutive
23 days in each of the one or two seasons. Median ambient NO2 levels were higher than the median
24 personal levels in both seasons (about 10 ppb in difference). During the winter, both ambient
25 and personal exposure to NO2 were higher than the summer, the difference between ambient and
26 personal exposure in winter was 1 to 2 ppb smaller than the difference in the summer. In the
27 Boston study, 24-h averaged personal and ambient PM2.5, Os, NO2, and SO2 were measured for
28 20 healthy seniors and 23 schoolchildren. All subjects were measured for 12 consecutive 24-h
29 periods in each of the 1 or 2 seasons. Ambient NO2 levels were on average 6 to 20 ppb higher
30 than the personal exposure levels for seniors during all sampling sessions. For children's
31 exposure, ambient NO2 levels were 7 to 13 ppb higher than the personal exposures in 4 out of
March 2008 AX3-58 DRAFT-DO NOT CITE OR QUOTE
-------
1 6 sampling sessions, and in the other two sampling sessions (one in summer and one in winter)
2 ambient levels were 1.8 to 2.6 ppb lower than personal exposures. Sarnat et al. (2006) measured
3 24-h averaged ambient and personal PM2.5, sulfate, elemental carbon, O3, and SO2 for 10 non-
4 smoking seniors in Steubenville, Ohio during the summer and fall of 2000. For each subject,
5 two consecutive 24 h personal exposure measurements were collected during each week for
6 23 weeks. Data were stratified by the presence of gas stoves in homes. Personal exposure was
7 lower than the ambient level for homes without gas stoves (9.0 ppb for personal exposure versus
8 9.5 ppb for ambient level during the summer and 9.9 ppb versus 11.3 ppb during the fall), and
9 higher than ambient levels for homes with gas stoves (12.3 ppb for personal exposure versus
10 9.5 ppb for ambient level during the summer and 15.7 ppb versus 11.3 ppb during the fall).
11 Nerriere et al. (2005) investigated factors determining the discrepancies between personal
12 exposure and ambient levels in the Genotox ER study. During the study, forty-eight h averaged
13 PM2.5, PMio, and NO2 were collected in both summer and winter for each person in a cohort,
14 with 60 to 90 nonsmoking volunteers composed of two groups of equal size for adults and
15 children at four metropolitan areas in France (Grenoble, Paris, Rouen, and Strasbourg). In each
16 city, subjects were selected so as to live in three different urban sectors contrasted in terms of air
17 pollution: one highly exposed to traffic emissions, one influenced by local industrial sources,
18 and a background urban environment. In each urban sector, a fixed ambient air monitoring
19 station was used to simultaneously collect the same air pollutants as personal exposure samplers.
20 Factors affecting the concentration discrepancies between personal exposure and corresponding
21 ambient monitoring site were investigated by a multiple linear regression model. Results showed
22 that the discrepancies were season, city and land use dependent. During the winter, city and land
23 use can interpret 31% of the variation of the discrepancy, and during the summer 54% of the
24 variation in the discrepancy can be interpreted by those factors. In most cases, ambient
25 concentrations were higher than the corresponding personal exposures. When using the ambient
26 site to represent ambient levels, the largest difference between ambient and personal exposure
27 was found at the "proximity to traffic" site, while the smallest difference was found at the
28 "background" site. When using urban background site as ambient level, the largest difference
29 was observed at the "industry" site, and the smallest difference was observed at the background
30 site, which reflected the heterogeneous distribution of NO2 in an urban area. During winter,
31 differences between ambient site and personal exposure were larger than those in the summer.
March 2008 AX3-59 DRAFT-DO NOT CITE OR QUOTE
-------
1 Age was not found to be a significant factor interpreting the discrepancies between ambient level
2 and personal exposure.
3 S0rensen et al. (2005) reported that during the cold season, median personal exposure
4 was higher than residential indoor and urban background concentrations, but lower than the
5 residential outdoor and street station concentrations (designed to capture the close to traffic
6 exposure). During the warm season, personal exposure was again lower than the street station
7 concentration but higher than the residential indoor, outdoor, and urban background
8 concentrations. The implication of these findings is that ambient concentrations are the primary
9 factor in determining exposures when there is no or little contribution from indoor sources and
10 that traffic is the most significant NC>2 source in this study.
11 The relative levels of ambient and personal exposure can also be expressed as ratios of
12 personal/ambient (Levy et al., 1998a; Rojas-Bracho et al., 2002; Sarnat et al., 2006). As shown
13 in Equation AX3.5-4, personal exposure is related to ambient concentration through the
14 infiltration factor, the fraction of time people spend outdoors, indoor sources and outdoor
15 concentration. In the absence of indoor sources, the ratio of personal exposure to ambient
16 concentration is sometimes also called the attenuation factor (a), which is always less than or
17 equal to one, and it is a function of infiltration factor (F;nf) and the fraction of time people spend
18 outdoors (y). The attenuation factor can be derived directly from measured personal and outdoor
19 concentrations or calculated from measured or estimated values of the parameters a, k, and P
20 (see Equation AX3.5-2 and Equation AX3.5-4) and the time spent in various microenvironments
21 from activity pattern diaries (Wilson et al., 2000). Because a depends on building and lifestyle
22 factors, air exchange rate, and NC>2 decay rate, it will vary to a certain extent from region-to-
23 region, season-to-season, and by the type of indoor microenvironment. Consequently, predicted
24 exposures based on these physical modeling concepts provide exposure distributions derived
25 conceptually as resulting from building, lifestyles, and meteorological considerations. For any
26 given population, the distribution of the coefficient a may represent substantial intra- and inter-
27 personal variability based on personal activity patterns, building and other microenvironmental
28 characteristics, and proximity to ambient and indoor sources. Distributions of a should be
29 determined using population studies in order to evaluate the uncertainty and variability
30 associated with model exposures. Unfortunately, only a few studies have reported the value and
31 distribution of the ratio of personal to ambient, and even fewer studies reported the value and
March 2008 AX3-60 DRAFT-DO NOT CITE OR QUOTE
-------
1 distribution of attenuation factors based on sophisticated study designs. Rojas-Bracho et al.
2 (2002) reported the median personal/outdoor ratio was 0.64 (with an interquartile range (IQR) of
3 0.45). Although it was less than one, the authors also reported the indoor/outdoor ratio (0.95
4 with an IQR of 0.48) of NC>2 and based on the indoor/outdoor ratio, the authors pointed out that
5 the high median indoor/outdoor ratio was greater than the estimated effective penetration
6 efficiency, which supports the argument of the importance of indoor sources to indoor NC>2
7 levels. Therefore, the attenuation factor in this study should be smaller than the ratio of
8 personal/ambient, which was 0.64. Sarnat et al. (2006) reported that the ratio of
9 personal/ambient for NC>2 was 2.05 and 1.27 for subjects with and without gas stoves in their
10 homes. The large personal/ambient ratio for the latter might be attributed to the influence of
11 indoor or local sources that were not identified and/or partly to measurement error.
12 The attenuation factor is one of the keys to evaluate personal exposure to ambient
13 generated NO2, or ambient contribution to personal exposure. However, the ratio of personal
14 exposure/ambient concentration will not accurately reflect the attenuation factor in the presence
15 of indoor sources. As shown above, in many cases, the ratio of personal exposure and ambient
16 concentration was above one, which is physically impossible for the attenuation factor. The
17 random component superposition (RCS) model is an alternative way to calculate attenuation
18 factor using observed ambient and personal exposure concentrations (Ott et al., 2000). The
19 Random Component Superposition (RCS) statistical model (shown in Equation AX3.5-4) uses
20 the slope of the regression line of personal concentration on the ambient or outdoor NO2
21 concentration to estimate the population average attenuation factor and means and distributions
22 of ambient/outdoor and nonambient contributions to personal NC>2 concentrations (the intercept
23 of the regression is the averaged nonambient contribution to personal exposure). This model
24 assumes a linear superposition of the ambient and nonambient components of exposure and lack
25 of correlation between these two components.
26 The RCS model derives a mean a across all homes (assuming the infiltration behavior
27 and time budget for all people are the same) from the linear regression of measured values of Et
28 on Ca. The product of the constant a and Ca from each home provides an estimate of the mean
29 and distribution of Ea for the population of study homes. In practice, the mean and distribution
30 of nonambient contributions (En0na) are given by the difference, Et - E&, on a home-by-home
31 basis. The RCS-predicted distribution of E& across the population of study homes is given by the
March 2008 AX3-61 DRAFT-DO NOT CITE OR QUOTE
-------
1 product of the constant a and Ca from each home, and the mean of the ambient contribution is
2 the difference between the mean total personal exposure and the intercept of the regression line.
3 The RCS model has been widely applied to PM exposure studies PTEAM, THEES, Toronto, and
4 RIOPA studies (Ott et al., 2000; Meng et al., 2005), but researchers have not intentionally used
5 this model for NC>2 exposure assessments. Although many studies explored the relationship
6 between personal exposure and ambient NC>2 concentrations using regression models, most of
7 those studies are not useful for evaluating the attenuation factor or helping answer the question
8 of how much personal NCh exposure comes from ambient air, either because only R2 was
9 reported, or because log-transformed concentrations were used in the regression model, or
10 because physically meaningless multiple linear regression models (exploratory variables were
11 not independent of each other, e.g., both indoor, outdoor, indoor sources from questionnaire
12 responses and air exchange rate were used as exploratory variables) were used to interpret
13 personal exposure variations. Only those simple linear regression models (personal versus
14 ambient or personal versus outdoor) and physically meaningful multiple linear regression models
15 (personal versus ambient + indoor source measured or identified by questionnaire) are useful for
16 evaluating the attenuation factor, and those models are summarized in Table AX3.5-1. The
17 intercept of the regressions (i.e., the nonambient contribution to personal exposure) varies widely
18 from study to study (5 ppb to 18 ppb) and thus depends strongly on time and location. The slope
19 of these regression models (i.e., the population average attenuation factor) varies between 0.3 to
20 0.6 in most of the studies. The attenuation factor is determined by air exchange rate, penetration
21 and decay rate of NC>2 and also the fraction of time people spend outdoors. S0rensen et al.
22 (2005) found that the attenuation factor was larger in the summer than in the winter. However,
23 Sarnat et al. (2006) found opposing results and said the reason was unknown. Based on the
24 regression model and reported mean personal exposure values, the ambient and nonambient
25 contribution to personal exposure could be calculated using the method described above. Since
26 most researchers did not report the mean personal exposure and the regression model at the same
27 time, ambient and nonambient contributions can only be calculated in four studies as shown in
28 Table AX3.5-2. The ambient contribution to population exposures varies from 20% to 50% in
29 these four studies.
30 The RCS model calculates ambient contributions to indoor concentrations and personal
31 exposures based on the statistical inferences of regression analysis. However, personal-outdoor
March 2008 AX3-62 DRAFT-DO NOT CITE OR QUOTE
-------
1 regressions could be affected by extreme values (outliers either on the x or the y axis), such as a
2 high nonambient exposure on a day with low ambient concentration or vice versa. For this
3 reason outliers must be identified and their influence on the infiltration factor or attenuation
4 factor in the RCS model must be evaluated in order to obtain a robust result. Another limitation
5 of the RCS model is that this model is not designed to estimate ambient and nonambient
6 contributions for individuals, in part because the use of a single attenuation factor does not
7 account from the large home-to home variations in actual air exchange rates, and penetration and
8 decay rates of NC>2. As suggested by Meng et al. (2005) the use of a fixed attenuation factor
9 might underestimate ambient contributions to indoor concentrations and personal exposures and
10 could also overlook some of the exposure errors and cause large uncertainties in risk estimates.
11 The estimation of the ambient and nonambient contribution to personal exposure could be
12 improved by allowing for variations in air exchange rate, penetration and decay rate of NC>2, and
13 the variations in the fraction of time people spend outdoors. The mass balance model described
14 in Equation AX3-5-4 gives more flexibility than the RCS model if the distributions of P, k, a,
15 and y are known. A comprehensive assessment of the impact of ambient sources on personal
16 exposure would require detailed consideration of the mechanisms of NO2 formation,
17 transformation, transport and decay. In the research field of NC>2 exposure assessment, no
18 published reports were found that use the mass balance model to explore the relationship of
19 personal exposures to ambient NC>2 concentrations. As mentioned in Section 3.4.2, the only
20 reported k values were 0.99 h-1 by Yamanaka (1984), and people always assumes the
21 penetration coefficient (P) is one for NC>2, which might overestimate the ambient contribution
22 due to the chemical reactivity of NC>2 during penetration.
23 The association between personal exposure and ambient NC>2 was quantified by Pearson
24 correlation coefficient (rp), Spearman correlation coefficient (rs), or coefficient of determination
25 (R2) in regression models (Spengler et al., 1994; Linn et al., 1996; Spengler et al., 1996;
26 Raaschou-Nielsen et al., 1997; Aim et al., 1998; Levy et al., 1998a; Monn et al., 1998; Liard
27 et al., 1999; Kramer et al., 2000; Linaker et al., 2000; Mukala et al., 2000; Gauvin et al., 2001;
28 Monn, 2001; Rotko et al., 2001; Sarnat et al., 2001; Kodama et al., 2002; Rojas-Bracho et al.,
29 2002; Lai et al., 2004; Sarnat et al., 2005; Kim et al., 2006; Sarnat et al., 2006). In Table
30 AX3.5-3, the associations between personal exposure and ambient concentration found in these
31 studies are summarized.
March 2008 AX3-63 DRAFT-DO NOT CITE OR QUOTE
-------
1 The association between personal NO2 exposure and ambient/outdoor NO2 concentration
2 varied from poor to good as shown in Table AX3.5-3. The strength of the correlation between
3 personal exposure and ambient/outdoor concentration for a population is determined by the
4 variations in indoor or other local sources, air exchange rate, penetration and decay rate of NO2
5 in different microenvironment, and time people spend in different microenvironments with
6 different NC>2 concentrations. The relationship is also a function of season and location
7 (rural/urban). Aim et al. (1998) indicated that the association between personal exposure and
8 outdoor concentration was stronger than the correlation between personal exposure and central
9 site concentration. However, Kim et al. (2006) pointed out that the association was not improved
10 using the ambient sampler closest to a home. Home ventilation is another important factor
11 modifying the personal-ambient relationships; we expect to observe the strongest associations for
12 subjects spending time indoors with open windows. Aim et al. (1998) and Kodama et al. (2002)
13 observed the association between personal exposure and ambient concentration became stronger
14 during the summer than the winter. However, Sarnat et al. (2006) reported that R2 decreased
15 from 0.34 for low ventilation population to 0.16 for high ventilation population in the summer,
16 and from 0.47 to 0.34 in the fall. This might be a caution that the association between personal
17 exposure and ambient concentration is complicated and is determined by many factors.
18 Exposure misclassification might happen if a single factor, such as season or ventilation status, is
19 used as an exposure indicator. Another factor affecting the personal to ambient association is the
20 subject's location, with higher correlation for subjects living in the rural areas and lower
21 correlation with subjects living in the urban areas (Rojas-Bracho et al., 2002; Aim et al., 1998).
22 Spengler et al. (1994) also observed that the relationship between personal exposure and outdoor
23 concentration was highest in areas with lower ambient NC>2 levels (R2 = 0.47) and lowest in areas
24 with higher ambient NC>2 levels (R2 = 0.33). This might reflect the highly heterogeneous
25 distribution, or the effect of local sources of NC>2 in an urban area, and personal activities are
26 more diverse in an urban area. However, this factor (location: urban vs. rural) might also interact
27 with indoor sources because indoor sources could explain more personal exposure when ambient
28 concentrations become lower and more homogeneously distributed.
29 The association is also affected by indoor or local sources, and the association becomes
30 stronger after those sources are controlled in the model. Raaschou-Nielsen et al. (1997) observed
31 that R2 increased from 0.15 for general population to 0.49 for a population who spent less than
March 2008 AX3-64 DRAFT-DO NOT CITE OR QUOTE
-------
1 2% of their time close to gas appliances and passive smoking in Copenhagen urban area, and R2
2 increased from 0.35 to 0.45 in the rural area for the population with the same characteristics.
3 When those who reported exposure to either gas appliances or passive smoking were excluded,
4 R2 increased to 0.59 in urban and 0.46 in the rural districts. Spengler et al. (1994) observed that
5 less of the variation in personal exposure was explained by outdoor concentrations for those who
6 had gas ranges with pilot lights (R2 = 0.44) than it is for the other two groups (R2 = 0.52). When
7 there is little or no contribution from indoor sources, ambient concentrations are the primary
8 factor in determining exposure, but if there are continuous indoor sources, the influence of
9 outdoor levels decrease. In the VESTA study, Gauvin et al. (2001) reported low R2s in all three
10 cities. R2s increased for all three cities after controlling indoor air sources (e.g., gas cooking)
11 and ambient traffic densities: R2 increased from 0.01 to 0.43 for Grenoble, from 0.04 to 0.50 for
12 Toulouse, and from 0.02 to 0.37 for Paris. Other factors, such as cross-sectional vs. longitudinal
13 study design, and sampling duration might also affect the strength of the association. However,
14 the current science review cannot give a clear picture of the effects by those factors due the lack
15 of key studies and data.
16 The correlation coefficient between personal exposure and ambient/outdoor concentration
17 has different meanings for different study designs. There are three types of correlations
18 generated from different study designs: longitudinal, "pooled," and daily-average correlations.
19 Longitudinal correlations are calculated when data from a study includes measurements over
20 multiple days for each subject (longitudinal study design). Longitudinal correlations describe the
21 temporal relationship between daily personal NCh exposure or microenvironment concentration
22 and daily ambient NC>2 concentration for each individual subject. The longitudinal correlation
23 coefficient may differ for each subject. The distribution of correlations across a population could
24 be obtained with this type of data. Pooled correlations are calculated when a study involves one
25 or only a few measurements per subject and when different subjects are studied on subsequent
26 days. Pooled correlations combine individual subject/individual day data for the calculation of
27 correlations. Pooled correlations describe the relationship between daily personal NO2 exposure
28 and daily ambient NO2 concentration across all subjects in the study. Daily-average correlations
29 are calculated by averaging exposure across subjects for each day. Daily-average correlations
30 then describe the relationship between the daily average exposure and daily ambient NO2
31 concentration.
March 2008 AX3-65 DRAFT-DO NOT CITE OR QUOTE
-------
1 The type of correlation analysis can have a substantial effect on the value of the resultant
2 correlation coefficient. Mage et al. (1999) mathematically demonstrated that very low
3 correlations between personal exposure and ambient concentrations could be obtained when
4 people with very different nonambient exposures are pooled, even though their individual
5 longitudinal correlations are high. Data shown in Table AX3.5-3 demonstrate that the
6 longitudinal correlations between personal exposure and ambient NO2 concentrations were
7 higher than the correlations obtained from a pooled data set.
8 In conclusion, personal exposure to ambient/outdoor NC>2 is determined by many factors.
9 Physically, the determinant factors are ambient concentration, air exchange rate, NC>2 penetration
10 and decay rate, and also the fraction of time people spend outdoors. These factors are in turn
11 determined by factors, such as season, location of home, outdoor temperature and so on. These
12 factors all help determine the contribution of ambient/outdoor generated NO2 to personal
13 exposures. Personal activities determine when, where and how people are exposed to NC>2. The
14 variations of these physical factors and indoor sources determine the strength of the association
15 between personal exposure and ambient concentrations both longitudinally and cross-sectionally.
16 In the absence of indoor and local sources, the personal exposure level is in between the ambient
17 level and the indoor level, but in the presence of indoor and local sources, personal exposures
18 could be much higher than both indoor and outdoor concentrations. Again, the discrepancies
19 between personal exposures and ambient levels are determined by the considerations given
20 above. Most researchers found that personal NO2 was significantly associated with ambient NO2
21 but the strength of the association ranged from poor to good. Based on that finding, some
22 researchers concluded that ambient NC>2 is a good surrogate for personal exposure, while others
23 reminded us that caution must be exercised if ambient NC>2 is used as a surrogate for personal
24 exposure. The crude association between personal exposure and ambient monitors could be
25 improved when indoor or other local sources are well controlled during exposure assessment.
26 The ambient contribution to personal exposure could be evaluated by the attenuation factor,
27 which is the ratio of personal exposure to ambient level in the absence of indoor sources, or the
28 slope of the RCS regression model. The attenuation factor in the studies shown in Table
29 AX3.5-1 ranged from 0.3 to 0.6. The ambient and nonambients contributions could also be
30 calculated from the RCS model, although only a few studies provide enough information for us
31 to calculate them. The accuracy and precision of the estimation of ambient and nonambient
March 2008 AX3-66 DRAFT-DO NOT CITE OR QUOTE
-------
1 contributions to personal exposures could be improved if the variations for the physical factors
2 given above were known. The mass balance model could give a more accurate and precise
3 estimation if we knew the distributions of these key physical factors.
4 Because people are exposed to ambient NO2 in microenvironments, and the fact that NO2
5 is heterogeneously distributed in urban areas (as shown in Section AX3.3.2), the association of
6 personal exposure to ambient NO2 could be modified by microenvironmental characteristics.
7 Personal total exposure will be decomposed and further evaluated in each microenvironment in
8 the following section.
9
10 AX3.5.2 Personal Exposure in Microenvironments
11
1 2 Personal Exposure in the Residential Indoor Environment
13 People spend most of their daily time in a residential indoor environment (Klepeis et al.,
14 2001). NO2 found in an indoor environment originates both indoor and outdoors; and therefore,
1 5 people in an indoor environment are exposed to both indoor and outdoor generated NC>2. The
16 physical parameters, which determine personal exposure to ambient and nonambient generated
17 NC>2, have been shown in Equations AX3.5-2 to AX3.5-5. In a residential indoor environment,
18 personal exposure to NC>2 can be summarized below (notations are the same as those in
19 Equations AX3. 5-2 to AX3. 5-5)
+ Enona = {y + (/ - y)[Pa/(a + k)]}Ca + Eimw =
20 ^' ^ ~ ^! ^ "?/' a nona (AX3 5-6)
21 if people spend 100% of their time indoors, the equation above can be recast as
Et = Ea + Enona = «Q + Enona = Finf^a + Enona
22 a nona (AX3.5-7)
23 In other words, in a residential indoor environment, personal exposure concentration equals the
24 residential indoor concentration (if there is no personal cloud) which can be broken down into
25 two parts: indoor generation and ambient contribution.
26 In a residential indoor environment, the relationship between personal NC>2 exposure and
27 ambient NC>2 can be modified by the indoor environment in the following ways: (1) during the
28 infiltration processes, ambient NC>2 can be lost through penetration and decay (chemical and
March 2008 AX3-67 DRAFT-DO NOT CITE OR QUOTE
-------
1 physical processes) in the indoor environment, and therefore, the concentration of indoor NC>2 of
2 ambient origin is not the ambient NC>2 concentration but the product of the ambient NC>2
3 concentration and the infiltration factor (Finf, or a if people spend 100% of their time indoors);
4 (2) in an indoor environment, people are exposed to not only ambient generated NC>2 but also
5 indoor generated NC>2, and therefore, the relative contribution of ambient and nonambient NC>2 to
6 personal exposure depends not only on the ambient NC>2 concentration but also on the infiltration
7 factor (attenuation factor) and the indoor source contribution; (3) the strength of the association
8 between personal exposure to NC>2 of ambient origin and ambient NC>2 concentration is
9 determined by the temporal and spatial variation in the infiltration factor; and (4) the strength of
10 the association between personal total exposure and ambient NC>2 is determined by the variation
11 in the indoor source contribution and the variation in the infiltration factor. Below, factors
12 affecting infiltration factor and the indoor source contribution will be evaluated, and the key
13 issues, such as those mentioned above, related to ambient contribution to personal NC>2 exposure
14 will be addressed.
15 Infiltration factor (F^f) of NC>2, the physical meaning of which is the fraction of ambient
16 NC>2 found in the indoor environment, is determined by the NC>2 penetration coefficient (P), air
17 exchange rate (a), and the NC>2 decay rate (&), through the equation F-mf= Pa/(a + k). Information
18 on P and k for NC>2 is sparse. In most mass balance modeling work, researchers assume P
19 equals 1 because NC>2 is a gas, and assume k equals 0.99 IT1, which is cited from Yamanaka
20 (1984). Yamanaka (1984) systematically studied the decay rates of NO2 in a typical Japanese
21 living room. The author used a chemical luminescence method to monitor the decay process of
22 indoor-originated NC>2. The author observed that the decay process of NC>2 followed
23 approximately first-order kinetics. The author also pointed out that the NC>2 decay processes was
24 both surface type and relative humidity (RH) dependent: Under low RH (43.5-50%), the sink
25 rate of NC>2 was 0.99 ± 0.19 IT1, independent of interior surface properties; however, the NC>2
26 decay rate increased in proportion to RH above 50%, and in that RH range, the decay rate
27 depended on the interior surface properties. Yang et al. (2004) estimated a decay rate of 0.94 IT1
28 for Seoul and 1.05 IT1 for Brisbane. As it is well known, the decay rate is dependent on lots of
29 indoor parameters, such as indoor temperature, relative humidity, surface properties, surface-to-
30 volume ratio, the turbulence of air flow, and co-existing pollutants, et al. However, in the indoor
31 air modeling studies, a decay rate of 0.99 IT1 is a widely accepted parameter (Dimitroulopoulou
March 2008 AX3-68 DRAFT-DO NOT CITE OR QUOTE
-------
1 et al., 2001; Kulkarni et al., 2002). As a result, it will over- or underestimate the real NO2 decay
2 rate. A penetration coefficient (P) of 1 is also widely accepted for NO2 (Kulkarni et al., 2002;
3 Yang et al., 2004). No systematic investigations have been found on NO2 penetration behaviors.
4 As a general principle, the upper limit of the penetration coefficient is 1, and it would be less
5 than 1 if NO2 lost during penetration due to diffusion and chemical reactions. Therefore, using a
6 penetration coefficient of 1 gives an upper bound to the estimated infiltration coefficient.
7 Among P, k, and a, air exchange rate (a) is the most solidly based parameter and can be obtained
8 from a nationwide database (Pandian et al., 1998).
9 Although specific P, k, and a were not reported by most studies, a number of studies
10 investigated factors affecting P, k, and a (or indicators of P, £, and a), and their effects on indoor
11 and personal exposures (Lee et al., 1996; Cotterill and Kingham, 1997; Monn et al., 1998;
12 Garcia-Algar et al., 2003; S0rensen et al., 2005; Zota et al., 2005). Garcia-Algar et al. (2003)
13 observed that double-glazed windows had significant effect on indoor NO2 concentrations.
14 Homes with double-glazed windows had lower indoor concentrations (6 ppb lower) than homes
15 with single glazed windows. Cotterill and Kingham (1997) reported that single or double glazed
16 window was a significant factor affecting NO2 concentrations in kitchen in the gas-cooker homes
17 (31.4 ppb and 39.8 ppb for homes with single and double glazed windows, respectively). The
18 reduction of ventilation can block outdoor NO2 from coming into the indoor environment, and at
19 the same time it can also increase the accumulation of indoor generated NO2. The same effect
20 was found for homes using air conditioners. Lee et al. (2002) observed that NO2 was 9 ppb
21 higher in homes with an air conditioner than homes without. The authors also observed that the
22 use of humidifier would reduce indoor NO2 by 6 ppb. House type was another factor reported
23 affecting ventilation (Lee et al., 1996; Garcia-Algar et al., 2003). Lee et al. (1996) reported that
24 the building type was significantly associated with air exchange rate: the air exchange rate
25 ranged from 1.04 IT1 for single dwelling unit to 2.26 IT1 for large multiple dwelling unit. Zota
26 et al. (2005) reported that the air exchange rates were significantly lower in the heating season
27 than the non-heating season (0.49 IT1 for the heating season and 0.85 IT1 for the non-heating
28 season respectively). It should be pointed out that both P and k are functions of complicated
29 mass transfer mechanisms on the indoor surfaces, and therefore they are associated with air
30 exchange rate, which has an impact on the turbulence of air flows indoors. However, the
31 relationship between P, k, and a has not been thoroughly investigated. Factors mentioned above
March 2008 AX3-69 DRAFT-DO NOT CITE OR QUOTE
-------
1 can significantly affect P, &, and a, and thus affect the relationship between indoor and outdoor
2 NO2 concentration, and personal exposure and outdoor NC>2 concentration.
3 Due to the lack of specific P, &, and a for study homes or a study population, instead of
4 using P, k, and a, alternative approaches to obtain the infiltration factor are the ratio of
5 indoor/outdoor NC>2 and the regression based RCS model. The basic rationale of the RCS model
6 has been introduced in the previous section. Without indoor sources, the ratio between indoor
7 NC>2 and outdoor NC>2 should be always less than or equal to 1. If the indoor to outdoor ratio is
8 larger than 1 (after adjusting for measurement error), we can surely say that indoor sources exist.
9 However, if an indoor/outdoor ratio is less than one, we cannot exclude the effect of indoor
10 sources; otherwise, the infiltration factor would be overestimated. In order to use an
11 indoor/outdoor ratio as the infiltration factor, study designs and questionnaires must be carefully
12 read, and only the ratio for homes without identified indoor sources can be used as an indicator
13 of infiltration factor. The population averaged infiltration factor is the slope of the regression
14 line of indoor concentration vs. outdoor concentration. The reliability of the regression slope is
15 dependent upon the sample size and how to deal with the outlier effects. Indoor/outdoor ratios
16 and the regression slopes are summarized in Table AX3.5-4. Those numbers, which can be
17 considered as an infiltration factor, are underlined and marked with bold font. Most of the
18 infiltration factors ranges from 0.4 to 0.7. Theoretically, infiltration factor is a function of air
19 exchange rate, which has been indicated by season in some studies. However, most studies do
20 not report the infiltration factor by season, and therefore, a seasonal trend of infiltration factor
21 could not be observed in Table AX3.5-4.
22 As mentioned before, personal NC>2 exposure is not only affected by air infiltrating from
23 outdoors but also by indoor sources. The NC>2 residential indoor sources reported are gas
24 cooking, gas heating, kerosene heating, smoking and burning candles (Schwab et al., 1994;
25 Spengler et al., 1994; Nakai et al., 1995; Lee et al., 1996; Linaker et al., 1996; Cotterill and
26 Kingham, 1997; Farrow et al., 1997; Kawamoto et al., 1997; Lee, 1997; Raaschou-Nielsen et al.,
27 1997; Aim et al., 1998; Levy et al., 1998a; Monn et al., 1998; Garrett et al., 1999; Chao, 2001;
28 Dennekamp et al., 2001; Dutton et al., 2001; Emenius et al., 2003; Kodama et al., 2002; Lee
29 et al., 2002; Mosqueron et al., 2002; Garcia-Algar et al., 2003; Garcia-Algar et al., 2004; Lai
30 et al., 2004; Lee et al., 2004; Yang et al., 2004; Zota et al., 2005; S0rensen et al., 2005; Lai et al.,
31 2006). Spengler et al. (1994) reported that personal exposures in homes with gas range with
March 2008 AX3-70 DRAFT-DO NOT CITE OR QUOTE
-------
1 pilot light were 15 ppb higher than those in homes with electric range, and it was 5 ppb higher in
2 homes with gas range without pilot light than homes with electric ranges. Schwab et al. (1994)
3 reported that homes with gas stove with pilot light had higher indoor NO2 concentrations (peak
4 concentrations ranging from 30 to 35 ppb), followed by homes with gas stove without a pilot
5 light (peak concentrations ranging from 15 to 20 ppb) and then homes with electric stoves (peak
6 concentrations ranging from 5 to 10 ppb). In an international study, Levy et al., (1998a) reported
7 that the use of a gas stove in the home was the dominant activity influencing NO2 concentrations
8 with a 67% increase in mean personal NO2 exposure and an increase in indoor-outdoor ratios
9 from 0.7 to 1.2. Smoking was found to be another significant factor elevating personal and
10 indoor NO2 exposure. Monn et al. (1998) reported that during 1-week integrated measurement,
11 smoking contributed 1 ppb more NO2 exposure. Aim et al. (1998) reported that one-week
12 integrated personal NO2 exposure for smokers and nonsmokers were 12.9 ppb and 10.7 ppb,
13 respectively. Zota et al. (2005) observed that smoking was not a significant indoor source.
14 However, the authors pointed out that the effect of smoking might have been overwhelmed by
15 the presence of the gas stove. S0rensen et al. (2005) found that burning candles were
16 significantly associated with the elevation of indoor NO2 (p = 0.02). NO2 concentration in an
17 indoor environment affected by the indoor sources is not homogeneously distributed: NO2
18 concentration is usually the highest in the kitchen, lowest in the bedroom and the concentration
19 in a livingroom is in between as shown in Table AX3.5-5. The concentration differences
20 between a bedroom and a kitchen ranged from 1 ppb to 28 ppb, and largest difference occurred
21 in homes with gas stoves.
22 The concentration differences in indoor microenvironments reflect the differences in
23 personal exposure in those microenvironments, which is related to personal activities and
24 behaviors. People who spend more time in a kitchen are expected to have higher NO2 exposures.
25 Also, in most exposure studies, integrated indoor and personal exposures were measured from
26 2 days to 2 weeks with passive samplers. Therefore, the peak exposure concentration could be
27 even higher.
28 Indoor source contributions to indoor and personal exposure are determined by indoor
29 source strength (S), house volume (F), air exchange rate (a) and the NO2 decay rate (K) in an
30 indoor environment, through the equation Cnona = SI\V(a + k)]. Indoor source strength has been
31 summarized in a previous section (Indoor sources and concentrations of nitrogen oxides). With a
March 2008 AX3-71 DRAFT-DO NOT CITE OR QUOTE
-------
1 mass balance approach, Yang et al. (2004) reported that the source strength for electric range
2 was 3.5 ppb/h, 11.5 ppb/h for gas range in Brisbane, and 23.4 ppb/h for gas range in Seoul. The
3 age of house and the house type are associated with ventilation, indoor sources, and house
4 volume. As mentioned before, Lee et al. (1996) reported that the building type was significantly
5 associated with volume of dwelling unit, and air exchange rate. Garrett et al. (1999) reported
6 that older houses were associated with higher nitrogen dioxide levels, possibly as a result of
7 older and less efficient appliances in older homes or due to smaller rooms.
8 The relative contribution of indoor and outdoor NO2 to personal and indoor exposures
9 can be easily and precisely calculated if we know the physical determinants, such as P, &, a, and
10 indoor source strength. Probability based exposure models, such as SHEDS and APEX, could be
11 used to evaluate the personal exposure to indoor and outdoor generated NO2. Basically, those
12 exposure models incorporate the physical and chemical processes determining indoor pollutant
13 concentrations as a function of outdoor concentration, indoor emission rates and building
14 characteristics; the combination of a microenvironment model and personal activity model will
15 allow researchers to evaluate the personal exposure to indoor and outdoor generated NO2. Due
16 to the lack of those parameters in publications, we are going to use a regression based RCS
17 model to evaluate the contribution of indoor and outdoor generated NO2 to personal exposure.
18 The rationale to use the RCS model to estimate indoor and outdoor contribution to indoor and
19 personal NO2 have been introduced in the previous section. In summary, the regression intercept
20 of indoor NO2 concentration vs. outdoor NO2 concentration is the population mean indoor
21 contribution to indoor NO2; and the difference between the population mean NO2 and the
22 intercept in the population mean of outdoor contribution to indoor NO2. The RCS model results
23 are summarized in Table AX3.5-6. As shown in Table AX3.5-6, the overall ambient
24 contribution to indoor NO2 is around 70% with a wide range from 40 to 90%. Indoor generated
25 NO2 contribution is 10-20% less for homes with electric stoves if electric stove then indoor
26 contribution is usually zero. With the lack of indoor sources, the role of indoor environment is a
27 sink for outdoor generated NO2 due to physical and chemical losses of NO2 in the indoor
28 environment (Yamanaka, 1984; Ekberg 1996; Kraenzmer 1999; Chao, 2001). Chao (2001)
29 reported that the average sink strength of NO2 in an indoor environment in Hong Kong was 0.42
30 mg/h.
March 2008 AX3-72 DRAFT-DO NOT CITE OR QUOTE
-------
1 In theory, personal exposure of ambient origin should be at least as much as the indoor
2 NC>2 of ambient origin in that people spend time in either an indoor or an outdoor environment.
3 However, it was shown in the previous part (Table AX3.5-2) that the ambient contribution to
4 population exposure ranged from 20% to 50% based on four studies (Rojas-Bracho et al., 2002;
5 Monn et al., 1998; Levy et al., 1998a; Spengler et al., 1994); and results here show that the
6 ambient contribution to indoor NC>2 is around 70% with a wide range from 40 to 90% based on
7 another four studies (Mosqueron et al., 2002; Yang et al., 2004; Kulkarni et al., 2002; Monn
8 et al., 1998). It is not clear at present why the indoor NC>2 of ambient origin is larger than the
9 personal NO2 exposure of ambient origin.
10 The strength of the indoor, outdoor and personal NC>2 associations (rp: Pearson
11 correlation coefficient; rs: Spearman correlation coefficient; and R2: coefficient of
12 determination) are summarized in Table AX3.5-7. The strength of the associations are
13 determined by the variation in F;nf (P, k, and a) and indoor source contributions from home to
14 home and from day to day. In general, the correlation between indoor and outdoor NC>2 ranges
15 from poor to good (rp: 0.06 to 0.86). When we break down the correlation coefficient by season
16 and indoor sources, it is obvious that the association between indoor and outdoor NC>2 is stronger
17 during spring and summer but weaker during wintertime, and the association is stronger for
18 homes without indoor sources but weaker for homes with strong indoor sources. Mukala et al.
19 (2000) reported an rp of 0.86 for the indoor and outdoor NC>2 association during the spring and it
20 reduced to 0.54 during the winter. Spengler et al. (1994) reported that the associations were
21 0.66 and 0.75 (rp) for homes with and without air conditioning system, respectively. Emenius
22 et al. (2003) reported that the association between indoor and outdoor NC>2 was 0.69 (rp) for
23 homes without smoker and without gas stove using, but the association was not significant for
24 homes with gas stove or smokers. Yang et al. (2004) reported that the indoor and outdoor NC>2
25 association was 0.70 (R2) for homes with electric ranges, and was 0.57 (R2) for homes with gas
26 ranges. In other words, personal exposure to ambient NC>2 in a residential indoor environment
27 will be modified the least when the air exchange rate is high and the indoor source contribution
28 is not significant. Considering the large spatial variation in ambient NO2 concentrations and the
29 relative sparseness of ambient NC>2 monitors, the associations between indoor and outdoor
30 concentrations are usually stronger than the associations between indoor and ambient
31 concentrations. As shown in Table AX3.5-7, a stronger personal vs. residential indoor
March 2008 AX3-73 DRAFT-DO NOT CITE OR QUOTE
-------
1 relationship than the personal vs. outdoor relationship has been reported by most studies (Lai
2 et al., 2004; Monn et al., 1998, Levy et al., 1998a; Spengler et al., 1994; Kousa et al., 2001;
3 Linaker et al., 1996), which is a reminder that personal exposure to ambient NO2 mostly happens
4 in the residential indoor environment. It should be pointed out that the association between
5 indoor, outdoor and personal NO2 and the relative contributions of indoor and outdoor NO2 to
6 indoor and personal exposures were calculated based on time integrated indoor, outdoor and
7 personal NO2 measurement with passive samplers and an average measurement time of a couple
8 of days to two weeks. In most studies, an equilibrium condition was assumed and the effects of
9 dynamics on the indoor, outdoor, and personal association were not evaluated, which could result
10 in missing the peak exposure and obscuring the real short-term outdoor contribution to indoor
11 and personal exposure. For example, the NO2 concentrations at locations close to busy streets in
12 urban environments may vary drastically with time. If the measurement is carried out during a
13 non-steady-state period, the indoor/outdoor concentration ratio may indicate either a too low
14 relative importance of indoor sources (if the outdoor concentration is in an increasing phase) or a
15 too high relative importance of indoor resources (if the outdoor concentration is in a decreasing
16 phase). The lower the air exchange rate, the greater the error due to the effects of transients
17 (Ekberg, 1996).
18
19 School and Office
20 Workplaces (schools and offices) are the places where people spend most of their time
21 after homes in an urban area. The location, indoor sources as well as the ventilation pattern of
22 schools and offices could be different from people's homes. Therefore, personal exposure
23 patterns in schools and offices could be different from exposure patterns at home. However,
24 NO2 concentrations in schools and offices have only been measured in only a few exposure
25 studies.
26 Most studies reported the personal exposure levels were lower than or equal to office
27 NO2 levels. Lai et al. (2004) reported that a cohort in Oxford spent 17.5% of their daily time in
28 offices, and mean personal total NO2 exposure was 15 ppb and 16.8 ppb for mean office
29 concentrations. Mosqueron et al. (2002) reported Paris office worker exposure levels and no
30 significant difference was found between personal total exposure (22.8 ppb) and NO2
31 concentrations in office (23.5 ppb). Personal exposures in schools were studied in Helsinki,
32 Southampton and Southern California. Aim et al. (1998) and Mukala et al. (2000) reported the
March 2008 AX3-74 DRAFT-DO NOT CITE OR QUOTE
-------
1 personal exposure levels in Helsinki for pre-school children. They reported that median personal
2 exposures were lower than the median NO2 concentrations measured inside the day care center
3 (13.1 ppb for personal exposure versus 18.8 ppb for inside day-care center for downtown winter;
4 14.7 ppb versus 24.1 ppb for downtown spring; 8.9 ppb versus 15.2 ppb for suburban winter; and
5 8.9 ppb versus 13.1 ppb for suburban spring). Linaker et al. (1996) found that the geometric
6 mean of school children exposures (18.8 ppb) was higher than geometric means of the NC>2
7 concentrations in classrooms (8.4 to 14.1 ppb) in a study of children's exposures to NC>2 in
8 Southampton, UK. A similar exposure pattern was found by Linn et al. (1996) during the
9 Southern California school children exposure study. During the study, personal exposure
10 (22 ppb) was higher than the NC>2 concentration inside school (16 ppb). NC>2 concentration in
11 school/office is determined by ambient NO2 level, local traffic sources, floor height and building
12 ventilation pattern. Partti-Pellinen et al. (2000) studied the effect of ventilation and air filtration
13 systems on indoor air quality in a children's day-car center in Finland. Without filtration, NOx
14 and PM generated by nearby motor traffic penetrated readily indoors. With chemical filtration,
15 50 to 70% of nitrogen oxides could be removed. The authors suggested that the possible adverse
16 health effects of nitrogen oxides and particles indoors could be countered by efficient filtration.
17 Mosqueron et al. (2002) reported 24% of variations in in-office NC>2 concentrations could be
18 explained by outdoor NC>2 levels (18%), and floor height (6%) and an inverse relation was
19 observed between in-office concentration and floor height. Aim et al. (1998) attributed the high
20 NO2 concentration in the day-care center to its close to major roads. Obviously, the relative
21 scale of personal exposure and school concentration also depends on personal activities outside
22 schools and workplaces.
23 Significant associations between personal exposure and workplace concentrations were
24 reported by most studies. Mosqueron et al. (2002) reported office NC>2 was a significant
25 predictor of personal exposure and 15% of the personal exposure was explained by time
26 weighted office NC>2 concentrations. Aim et al. (1998) reported population NC>2 exposures were
27 highly correlated with the NC>2 levels inside the day-care centers (R2 = 0.88). However, Lai et al.
28 (2004) reported a nonsignificant Pearson correlation coefficient (0.15) between personal
29 exposure and workplace indoor concentration and the authors suggested that the strong
30 residential indoor sources and long time indoors obscured the personal versus office relationship.
31 Personal total exposure is a function of NC>2 concentrations in different indoor and outdoor
March 2008 AX3-75 DRAFT-DO NOT CITE OR QUOTE
-------
1 microenvironments and how long a person stays in that microenvironment. The large variation
2 of NC>2 exposure in some microenvironments could obscure the association between personal
3 exposure and NC>2 concentrations in other microenvironments.
4
5 In Traffic
6 On-road NC>2 concentrations could be substantially higher than ambient or residential
7 outdoor NC>2 concentrations, especially in a street canyon, which are narrow with enclosing
8 architecture and slow-moving traffic. As shown in Figure AX3.5-1, NC>2 in heavy traffic
9 (-60 ppb) can be over twice the concentration in a residential outdoor level (-26 ppb) in North
10 America (Lee et al., 2000). The UK and Scandinavian data in the plot may have been obtained
11 outside homes close to traffic. Westerdahl et al. (2005) reported on-road NC>2 concentrations in
12 Los Angeles ranging from 40 to 70 ppb on freeways, and 20 to 40 ppb on residential or arterial
13 roads. People in traffic can potentially experience such high concentrations and NO2 exposures
14 due to the high air exchange rates for vehicles. Park et al. (1998) measured the air exchange
15 rates in three stationary automobiles under four conditions: windows closed and no mechanical
16 ventilation, windows closed with fan set on recirculation, windows open with no mechanical
17 ventilation, and windows closed with the fan set on fresh air. The reported air exchange rates
18 varied from 1.0 to 3.0 IT1 with windows closed and no mechanical ventilation to 36.2 to 47.5 IT1
19 with windows closed and the fan set on fresh air. It implies that the NC>2 concentration inside a
20 vehicle is at least the same as the surrounding NC>2 concentration, or in other words, "on-road"
21 NC>2 can quickly and almost completely infiltrate into the "in-vehicle" environment contribute to
22 in-vehicle personal exposures. Although people only spend a small fraction of their time in
23 traffic (5% to 7%), exposure while commuting could be a significant contributor to personal
24 exposure to NC>2 due to the high concentration of NC>2 in traffic. Liard et al. (1999) reported that
25 both NO and NC>2 exposure levels increased with the number of hours spent in a car. During the
26 study, NO and NO2 concentrations were separated into three levels according to the distribution
27 tertiles. Personal exposure levels increased from low to high when accordingly people spent
28 from 2.5 h in a car to 6.7 h in a car. The same relationship only held for one of the two sampling
29 periods, in which personal NO2 exposures increased from low to high when the time people
30 spent in a car increased from 3.5 h to 5.7 h.
31 Bell and Ashenden (1997) and Kirby et al. (1998) reported the NO2 concentration along
32 major roads and street canyons in UK, and they found that monthly mean NO2 concentrations on
March 2008 AX3-76 DRAFT-DO NOT CITE OR QUOTE
-------
^^
Q.
Q.
yj
0
O
2
0
o
X
0
O)
2
0)
>
<
^u
45-
40-
35-
30-
25-
20-
15-
10-
5-
t
r
t
'
t
t
t
,oUK
*
i
Western Europe ° North America
' O
o Japan
,'O o Brisbane
Scandinavia
/
/
*
o
Korea
0 50 100 150
Estimated NO2 Level during Commuting (ppb)
Figure AX3.5-1. Average residential outdoor concentration versus concentration
during commuting for
Source: Lee et al. (2000).
1 major roads were consistently higher (up to 20 ppb) than those found 250 m away from the
2 major roads. It is important to distinguish between short-term peak exposure and chronic
3 exposures because health effects associated with short-term peak exposures might be different
4 from chronic exposures to ambient NC>2.
5 Other than infiltration of ambient air, the intrusion of the vehicle's own exhaust into the
6 passenger cabin is another NC>2 source contributing to personal exposure while commuting. The
7 intrusion of a school bus's own exhaust into the bus cabin was found by Sabin et al. (2005), but
8 the fraction of air inside the bus cabin from the bus's own exhaust was small, ranging from
9 0.02% to 0.28%. Marshall and Behrentz (2005) also reported the intrusion of exhaust into the
10 bus cabin and indicated that average per capita inhalation of emissions from any single bus is
11 105-106 times greater for a passenger on that school bus than for a typical resident in the same
March 2008
AX3-77
DRAFT-DO NOT CITE OR QUOTE
-------
1 area. CARB (2007) reported that self-pollution increased with increasing age of the bus. Fuel
2 type could be another factor affecting personal exposure while commuting. Sonetal. (2004)
3 found that the two-day averaged NO2 exposures for taxi drivers using LPG fueled vehicles
4 (26.3 ppb) were significantly lower than those using diesel-fueled vehicle (38.1 ppb). However,
5 in another taxi driver exposure study, Lewne et al. (2006) did not find an effect on taxi driver
6 exposures to NO2 due to fuel differences (diesel versus petrol). Sabin et al. (2005) reported that
7 NO2 concentrations were significantly higher inside diesel buses than inside the compressed
8 natural gas buses. CARB (2007) showed that the NO2 concentrations on a conventional diesel
9 bus was 2.8 times higher than the ambient concentration (76 ppb in cabin versus 27 ppb in
10 ambient) while windows were closed, and 3.85 times higher than the ambient concentration
11 (77 ppb in cabin versus 20 ppb in ambient) while windows were open. However, the ratio of
12 cabin NO2 to ambient NO2 was much lower for a compressed natural gas bus: 1.2 for windows
13 closed and 2.2 for windows opened.
14 While commuting, concentrations for personal exposure or in a vehicle cabin could be
15 substantially higher than corresponding residential indoor, outdoor, and ambient concentrations.
16 Sabin et al. (2005) measured concentrations of a number of pollutants (black carbon, particulate
17 PAHs and NO2 in school buses on routes in Los Angeles. Mean cabin concentrations for
18 individual runs ranged from 24 to 120 ppb. Concentrations of NO2 tended to be slightly higher
19 for open compared to closed windows on urban routes. These concentrations were typically
20 factors of 2.3 to 3.4 higher than at ambient monitors in the area. However, the highest ratios
21 found ranged from 3.9 to 5.3. They concluded that children commuting in areas such as Los
22 Angeles may be exposed to much higher levels of pollutants than are obtained at ambient, central
23 site monitors. Lewne et al. (2006) reported work hour exposures to NO2 for taxi drivers
24 (25.1 ppb), bus drivers (31.4 ppb) and lorry drivers (35.6 ppb). The ratios of in-vehicle
25 exposures to urban background were 1.8, 2.7, and 2.8 for taxi drivers, bus drivers and lorry
26 drivers, respectively. Due to the high peak exposures during commuting, total personal exposure
27 could be underestimated if exposure in traffic are not considered; and sometimes exposure in
28 traffic can dominate personal exposure to NO2. In a personal exposure study in Brisbane and
29 Queensland, Australia, two-day averaged indoor, outdoor, and personal NO2 were measured by
30 Yanagisawa badges (Lee et al., 2000). Lee et al. (2000) found that estimated personal exposures
31 (22.5 ppb) significantly underestimated the measured personal exposures (28.8 ppb) if personal
March 2008 AX3-78 DRAFT-DO NOT CITE OR QUOTE
-------
1 exposures in traffic were not considered. Son et al. (2004) reported two-day averaged indoor,
2 outdoor, in vehicle and personal NC>2 concentrations measured by passive filter badges for
3 31 taxi drivers in Korea. Measured personal concentrations (30.3 ppb) were higher than both
4 residential indoor (24.7 ppb) and residential outdoor concentrations (23.8 ppb). A stronger
5 correlation was observed between personal NO2 exposures and interior vehicle NO2 levels, than
6 for residential indoor and residential outdoor levels (rp = 0.89 for Personal versus Vehicle,
7 rp = 0.74 for Personal versus Indoor; and rp = 0.71 for Personal versus Outdoor).
8 Variations in traffic exposure could be attributed to time spent in traffic, type of vehicle,
9 traffic congestion levels, encounters with other diesel vehicles, type of fuel and driving location
10 (urban/rural) (Sabin et al., 2005; Son et al., 2004; Chan et al., 1999).
11
12 Microenvironments Close to NO2 Sources
13 As suggested previously in this chapter, both large and small-scale variations exist in
14 ambient NC>2 concentrations. In this section, those microenvironments and associated personal
15 exposures, which are close to traffic sources and might make significant contributions to total
16 personal NC>2 exposures are analyzed. These microenvironments could be residential outdoor
17 environments and some other outdoor environments, such as parking lots and playgrounds; they
18 could also be indoor environments as well, such as homes and classrooms. Concentrations in
19 these microenvironments and personal exposure characteristics in these microenvironments will
20 be summarized below.
21 Many studies show that outdoor NC>2 levels are strongly associated with distance from
22 major roads (the closer to a major road, the higher the NO2 concentration) (Gilbert et al., 2005;
23 Roorda-Knape et al., 1998; Lai and Patil, 2001; Kodama et al., 2002; Gonzales et al., 2005;
24 Cotterill and Kingham, 1997; Nakai et al., 1995). Meteorological factors (wind direction and
25 wind speed), and traffic density are also important for interpreting measured NC>2 concentrations
26 (Gilbert et al., 2005; Roorda-Knape et al., 1998; Rotko et al., 2001; Aim et al., 1998; Singer
27 et al., 2004; Nakai et al., 1995). Gonzales et al. (2005) found an inverse correlation between
28 NO2 concentration and distance from a highway (rp = -0.81, p < 0.001) in the El Paso region.
29 Nakai et al. (1995) reported the results of a study designed to explore the differences of indoor,
30 outdoor and personal exposure levels among residence zones located varying distances from
31 major roads with heavy traffic in Tokyo. The authors found that outdoor NC>2 concentrations in
March 2008 AX3-79 DRAFT-DO NOT CITE OR QUOTE
-------
1 Zone A (0-20 m from the road) was always the highest among the three zones (Zone B was 20-
2 150 m from the road, and Zone C was a reference zone in a suburban area). The differences of
3 the mean levels between Zone A and Zone C ranged from 11 ppb to 39 ppb. Kodama et al.
4 (2002) reported NC>2 levels for indoor, outdoor and personal exposure among 150 junior high
5 school student homes in two major traffic areas in Tokyo. Forty-eight h average NC>2
6 concentrations were measured by Yanagisawa badge. NO2 tended to decrease according to
7 distance from the roadside; the difference was about 10 ppb between the roadside (0-50 m) and
8 the site far away from the road (200 m). Singer et al. (2004) reported results of the East Bay
9 Children's Respiratory Health Study. The authors reported weekly integrated NC>2 and NOx
10 concentrations measured by Ogawa passive samplers placed outside ten elementary schools and
11 selected student residences during 14 weeks in spring and 8 weeks in fall 2001. The authors
12 found that NC>2 concentrations increased with decreasing downwind distance for school and
13 neighborhood sites within 350m downwind of a freeway, and schools located upwind or far
14 downwind of freeways were generally indistinguishable from one another and regional pollution
15 levels. An exponential equation was used to fit the measured concentrations to distance from the
16 freeway: C(x) = KixK2 where C is the measured concentration and x is the distance (m) from a
17 freeway. A high R2 was observed (R2 = 0.80, KI = 128, and K2 = -0.356 for NO2; R2 = 0.76, KI
18 = 376, and K2 = -0.468). According to this equation, NC>2 concentrations 100 m away from the
19 freeway are about 20% of those at roadside.
20 Elevated NC>2 concentrations were also observed and reported in parking lots and school
21 playgrounds. Lee et al. (1999) reported the concentration of NC>2 at a parking lot in Hong Kong
22 was 60 ppb, and the level was about the same for NO. Colbeck (1998) reported that
23 concentrations in two parking lots in Colchester, UK were similar to those measured at the curb
24 side. Exposure of car parking lot users to NC>2 is comparable to that arising in the vicinity of
25 roads with moderate traffic density (-9000 vehicles per day). NC>2 concentrations in one parking
26 lot ranged from 30.4 to 47.1 ppb, while those in the payment booth ranged from 22.5 to 31.4 ppb.
27 Rundell et al. (2006) reported PMi, NO2, SO2, CO, and O3 concentrations at four elementary
28 school playgrounds and one university soccer field in Pennsylvania. NO2 concentrations were
29 below 100 ppb. The number concentration in the PMi size fraction decreased with distance
30 away from the highway (from 140,000 number/cm3 within 10 m of the road to 40,000
31 number/cm3 at 80 m).
March 2008 AX3-80 DRAFT-DO NOT CITE OR QUOTE
-------
1 Indoor environments, which are close to traffic, include buildings and houses along
2 major, busy roads. Most studies show that indoor NC>2 is correlated with outdoor NC>2, and is
3 also a function of distance to traffic, traffic density and meteorological parameters. The level of
4 indoor NC>2 in those microenvironments is also affected by indoor sources. Bae et al. (2004)
5 reported indoor and outdoor concentrations of NC>2 in 32 shoe stalls in Seoul, which were located
6 on busy streets. Working-hour (10 ± 2.1 h) NC>2 was measured by Yanagisawa passive filter
7 badges. Mean indoor and outdoor NC>2 concentrations were 57.4 and 58.1 ppb with a mean
8 indoor vs. outdoor ratio of 0.93. Maximum indoor and outdoor NC>2 concentrations were 94.1
9 and 96.3 ppb. In this study, outdoor traffic generated NC>2 is likely the main source of indoor
10 exposures due to the lack of indoor NC>2 sources. Outdoor and in-classroom NC>2 were measured
11 using Palmes tubes during three 2-week periods in six city districts near motor ways in the West
12 of the Netherlands (Roorda-Knape et al., 1998). NC>2 concentrations in classrooms were
13 significantly correlated with car and total traffic density (rp = 0.68), percentage of time
14 downwind (rp = 0.88) and distance of the school from the motorway (rp = -0.83). Cotterill and
15 Kingham (1997) measured indoor and outdoor NC>2 in 40 homes in Huddersfield, UK, over three
16 consecutive two-week periods in late 1994 using Palmes tubes. The authors found that
17 proximity to a main road had little effect on indoor levels of nitrogen dioxide (a mean of 1 ppb
18 indoor concentration difference was found for homes close to main roads and homes close to
19 side roads). A t-test suggested that there was no difference in indoor levels of nitrogen dioxide
20 due to proximity to the main road after indoor sources were controlled by the type of cookers.
21 In this study, meteorological parameters were measured, but meteorological parameters were not
22 controlled during data analysis.
23 Personal exposure is determined by both indoor and outdoor levels of NC>2. Most studies
24 show significant associations between personal exposure and the traffic density. The influence
25 of indoor sources on personal exposure was also observed in those studies. Aim et al. (1998)
26 reported the weekly personal NC>2 exposures of 246 children aged 3-6 years in Helsinki. Weekly
27 personal exposures were measured for 13 weeks in winter and spring in 1991 using Palmes
28 tubes. The 13 week geometric mean of the NC>2 exposures was higher for the children living in
29 the downtown (13.9 ppb) than in the suburban area (9.2 ppb, p = 0.0001). Rotko et al. (2001)
30 reported the EXPOLIS-Helsinki study results and observed that the NC>2 exposure was
31 significantly associated with traffic volume near homes. The average exposure level of
March 2008 AX3-81 DRAFT-DO NOT CITE OR QUOTE
-------
1 138 subjects having low or moderate traffic near their homes was 12.3 ppb, while the level was
2 15.8 ppb for the 38 subjects having high traffic volume near home. Gauvin et al. (2001) reported
3 the VESTA study results. An index of traffic density and proximity was constructed as the ratio
4 of traffic density to distance from a roadway. The index was one of the significant interpreters of
5 personal exposure in all three cities (p < 0.05 for Grenoble and Toulouse, and 0.05 < p < 0.15 for
6 Paris). Kodama et al. (2002) showed that personal exposure was similar to residential home NO2
7 concentration for residences along busy roads. The authors also observed that personal exposure
8 levels were higher than outdoor levels during the winter, while during the summer, personal
9 exposure levels were lower than ambient levels, due to the influence of indoor sources and low
10 ventilations in the winter. Although the personal to outdoor relationship was dominated by
11 indoor sources, the effects of outdoor NO2 on personal exposure could still be observed after
12 controlling the indoor source effects. Nakai et al. (1995) observed that personal exposure levels
13 basically followed the ambient concentrations patterns given above; i.e., exposures in Zone A
14 (0-20 m from the road) were the highest and exposures in Zone C (the suburban background
15 area) were the lowest for residents not using an unvented heater (as defined before, Zone A was
16 0-20 m from the road; Zone B was 20-150 m from the road. The maximum difference of
17 personal exposure between Zone A and Zone C was approximately 20 ppb. The NO2 exposure
18 for a special population, athletes, was addressed by Carlisle and Sharp (2001). The authors
19 pointed out that athletes could be a potential population at risk, if the ambient NO2 concentration
20 is high because (1) inhalation rate increases during exercise, (2) a large fraction of air is inhaled
21 through the mouth during exercise, effectively bypassing the normal nasal mechanisms for the
22 filtration of large particles and soluble vapors, and (3) the increased air flow velocity carries
23 pollutants deeper into the respiratory tract and pulmonary diffusion capacity increases during
24 exercise. This might also be true for outdoor workers but few data are available to perform the
25 exposure assessment.
26 Although traffic is a maj or source of ambient NO2, industrial point sources are also
27 contributors to ambient NO2. However, no published reports were found to address the effect of
28 those sources on population exposure within the United States. Nerriere et al. (2005) measured
29 personal exposures to PM2 5, PMio, and NO2 in traffic dominated, urban background and
30 industrial settings in Paris, Grenoble, Rouen, and Strasbourg, France. They always found highest
31 ambient concentrations and personal exposures close to traffic. In some cases, urban and
March 2008 AX3-82 DRAFT-DO NOT CITE OR QUOTE
-------
1 background, concentrations of NC>2 were higher than in the industrial zone. However, PM levels
2 and personal exposures tended to be higher in the industrial area than in the traffic dominated
3 area. It should be remembered that there can be high traffic emissions in industrial zones, such
4 as in the Ship Channel in Houston, TX. In rural areas where traffic is sparse, other sources could
5 dominate. For example, Martin et al. (2003) found pulses of NC>2 release from agricultural areas
6 following rainfall and there are contributions from wildfires and residential wood burning.
7
8 Exposure Reconstruction
9 Personal exposure has been evaluated in each major microenvironment, where either the
10 NC>2 concentration is high or people spend most of their daily time. As shown in Equation
11 AX3.5-2, personal exposure could be reasonably reconstructed if we know the NO2
12 concentration in each microenvironment and the duration of personal exposure in each
13 microenvironment. Levy et al., (1998a) reconstructed personal exposures measured in an
14 international study with a time-weighted average exposure model (Equation AX3.5-1). The
15 personal exposure was reconstructed based on the measured NC>2 concentrations in residential
16 indoor, residential outdoor, and workplace microenvironments, and the time people spent in
17 those environments. The mean measured personal NC>2 exposure was 28.8 ppb and a mean of
18 estimated NC>2 exposure was 27.2 ppb. The Spearman correlation coefficient between personal
19 measured exposure and reconstructed exposure was 0.81. The same approach was applied by
20 Kousa et al. (2001) to reconstruct the personal exposures in the EXPOLIS study. A correlation
21 coefficient of 0.86 was observed for the association between measured NC>2 exposure and
22 reconstructed NO2 exposure (data were log-transformed), and the slope and the intercept were
23 0.90 and 0.22 respectively for the reconstructed exposure vs. measured exposure. In the two
24 studies mentioned above, NC>2 exposure during commuting was not measured. Probably that is
25 part of the reason why reconstructed NC>2 exposure was lower than the measured NC>2 exposure.
26
27 AX3.5.3 Exposure Indicators
28 Physically, personal exposure levels are determined by those physical parameters in
29 Equations AX3.5-1 to AX3.5-5, i.e., the time people spend in each microenvironment and the
30 NC>2 concentrations in each microenvironment, which is determined by source emission strength,
31 air exchange rate, penetration coefficient, the NC>2 decay rate and the volume of the
32 microenvironment. Any factors that can influence the above physical parameters can modify the
March 2008 AX3-83 DRAFT-DO NOT CITE OR QUOTE
-------
1 level of personal exposure. These factors are defined as exposure indicators in this section. The
2 indoor, outdoor and personal NO2 levels on each stratum of those factors will be summarized.
3 Those factors can be classified in to the following categories: (1) factors associated with
4 environmental conditions, such as weather and season; (2) factors associated with dwelling
5 conditions, such as the location of the house and ventilation system; (3) factors associated with
6 indoor sources, such as the type of range and the fuel type; (4) factors associated with personal
7 activities, such as the time spent on cooking or commuting; (5) socioeconomic status, such as the
8 level of education and the income level; and (6) demographic factors, such as age and gender.
9 Most studies addressed the influences of dwelling condition and indoor sources on indoor
10 and personal exposures. A few studies explored the impacts of environmental factors and
11 personal activities on personal exposures. Indoor and personal exposures have rarely been
12 stratified by socioeconomic and demographic factors. Indoor, outdoor, and personal exposure
13 levels are presented in Table AX3.5-8, stratified by environmental factors, dwelling conditions,
14 indoor sources, and personal activities factors. The effects of socioeconomic and demographic
15 factors on the indoor, outdoor, and personal levels are summarized in Table AX3.5-9.
16 Season is an environmental factor affecting both indoor and outdoor levels, and thus
17 personal NO2 levels. During the wintertime, the mixing height is usually lower than during the
18 summer, and therefore concentrations of many primary pollutants are higher than in the summer.
19 Wintertime is also a heating season, which usually leads to higher indoor source emissions and
20 lower air exchange rates. Therefore, a higher indoor NO2 concentration can be expected during
21 the winter. For most cases, the differences of indoor or personal NO2 exposure between the
22 heating and non-heating season are within several ppb, but sometimes the difference could be
23 close to 20 ppb (Zota et al., 2005). Other environmental factors include day of the week
24 (weekday versus weekend), and the wind direction, as shown in Table AX3.5-8.
25 The dwelling conditions are also associated with indoor, outdoor, and personal NO2
26 levels. Location of the dwelling unit is an indicator of ambient NO2 source strength. A house
27 located in an urban center or close to a major road is expected to have higher outdoor and indoor
28 NO2 levels, and the differences in NO2 exposures are often within 20 ppb based on passive
29 sampler monitoring. The age of the house, house type, and window type can affect the
30 ventilation of dwelling units, and sometimes the type of heating and cooking applicances in a
31 house. Range and fuel type are the indoor source factors discussed the most in the literature. It
March 2008 AX3-84 DRAFT-DO NOT CITE OR QUOTE
-------
1 is common to see differences larger than 10 ppb in indoor and personal NO2 exposures between a
2 gas range home (especially gas range with pilot light) and an electric range home. Sometimes
3 the differences could be as high as 40 ppb. For peak short-term exposures, the difference could
4 reach 100 ppb.
5 The level of personal exposure is dependent upon the time a person spends in each
6 microenvironment. Kawamoto et al. (1997), Levy et al. (1998a), and Chao and Law (2000)
7 clearly showed that personal NO2 exposure increases with time spent cooking or commuting.
8 The common findings are summarized above. However, there are inconsistencies in the
9 literature. For example, smoking is claimed to be a significant factor in some studies but not in
10 others, and the same can be said for proximity to a major road. For another example, a higher
11 indoor NO2 level could be found in a rural home rather than in an urban home (Table AX3.5-8),
12 although most studies found the opposite. Part of the reason is that exposure indicators function
13 together, as a multidimensional parameter space, on indoor and personal exposures. They are
14 not independent of each other. Unfortunately, studies have rarely been conducted to understand
15 the associations between these exposure indicators and to use the study findings to explain
16 indoor and personal NO2 exposures.
17 More effort put on exposure indicator studies should help in finding better surrogate
18 measurements for personal exposures. Although misclassifying exposures in epidemiological
19 studies is almost inevitable, and it is unlikely that the personal exposures of all subjects will be
20 measured, a better knowledge of the effects of exposure indicators on personal exposure will
21 help reduce exposure errors in exposure and epidemiological studies and help interpret those
22 study results.
23
24
25 AX3.6 CONFOUNDING AND SURROGATE ISSUES
26 Confounding is the technical term for finding an association for the wrong reason. It is
27 associated with both the exposure and the disease being studied, but is not a consequence of the
28 exposure. The confounder does not need to be an exposure for the disease under study. The
29 confounding variable can either inflate or deflate the true relative risk.
30 Since epidemological studies of NO2 often use ambient concentrations to reflect
31 exposures, whether confounding of NO2 findings is possible can be determined by examining
32 associations among ambient concentrations and personal exposures to NO2 and its relevant
March 2008 AX3-85 DRAFT-DO NOT CITE OR QUOTE
-------
1 copollutants. Importantly, by examining these associations, it is also possible to evaluate
2 whether a copollutant may act as a confounder or as a proxy of ambient NO2 concentrations.
3 The potential for confounding of ambient NO2 health effects is discussed in terms of four
4 relationships: (1) ambient NO2 and ambient copollutant concentrations, (2) personal NO2 and
5 personal copollutant exposures, (3) personal NO2 exposures and ambient copollutant
6 concentrations, and (4) ambient NO2 concentrations and personal copollutant exposures.
7
8 1) Associations between Ambient NO 2 and Ambient Copollutant Concentrations
9 Confounding of NO2 health effects is often examined at the ambient level, since ambient
10 concentrations are generally used to reflect exposures in epidemiological studies. The majority
11 of studies examining pollutant associations in the ambient environment have focused on ambient
12 NO2, PM2.5 (and its components), and CO, with fewer studies reporting the relationship between
13 ambient NO2 and ambient O3 or SO2.
14 Correlations between concentrations of ambient NO2 and other ambient pollutants, PM2 5
15 (and its components where available), CO, Os and SO2 are summarized in Table AX3.6-1. Data
16 were compiled from Environmental Protection Agency's Air Quality System and a number of
17 exposure studies. Mean values of site-wise correlations are shown. As can be seen from the
18 table, NO2 is moderately correlated with PM2.5 (range: 0.37 to 0.78) and with CO (0.41 to 0.76)
19 in suburban and urban areas. At rural locations, such as Riverside, CA, associations between
20 ambient NO2 and ambient CO concentrations (both largely traffic-related pollutants) are much
21 lower, likely as the result of other sources of both CO and NO2 increasing in importance in rural
22 areas. These sources include oxidation of CH4 and other biogenic compounds, wood burning
23 and wildfires (for CO); and soil emissions, lightning, and wood burning and wildfires for NO2.
24 In urban areas, the ambient NO2-CO correlations vary widely. The strongest correlations are
25 seen between NO2 and elemental carbon. Note that the results of Hochadel et al. (2006) for
26 PM2 5 optical absorbance have been interpreted in terms of elemental carbon (EC). Correlations
27 between ambient NO2 and ambient Os are mainly negative, with again considerable variability in
28 the observed correlations. Only one study (Sarnat et al., 2001) examined associations between
29 ambient NO2 and ambient SO2 concentrations, showing a negative correlation during winter.
30 The robustness of this result needs to be examined in other cities.
31 Kim et al. (2006) reported the associations between 24 h averaged NO2 and other
32 pollutions for personal exposures and ambient concentrations in a study in Toronto, Canada from
March 2008 AX3-86 DRAFT-DO NOT CITE OR QUOTE
-------
1 August 1999 to November 2001. The median, mean, and standard deviation of the correlations
2 between ambient NO2 and ambient PM2.5 were 0.52, 0.44, and 0.35 respectively; and 0.81, 0.72,
3 and 0.22 respectively for the correlation between NO2 and CO.
4 In an exposure study in Steubenville, Ohio, Sarnat et al. (2006) reported the associations
5 between ambient concentrations and personal exposures for different pollutants. Ambient NO2
6 was significantly associated with ambient PM2.5, sulfate and EC during the fall (slope = 0.38,
7 0.96, and 7.01; and R2 = 0.61, 0.49, and 0.68 respectively) but not during the summer
8 (slope = -0.01, -0.17, and 3.76; and R2 = 0.0, 0.01, and 0.06 respectively).
9 In a related study, Connell et al. (2005) reported the correlation between ambient NOx
10 and PM2.5 during a comprehensive air monitoring program in Steubenville, Ohio. Across the two
11 year study (August 2000-April 2002), the Spearman correlation coefficient (rs) between hourly
12 ambient PM2.5 and NO concentrations was 0.33, and between hourly ambient PM2.5 and NO2
13 concentrations was 0.50. The authors suggested the importance of a common factor influencing
14 ambient concentrations of these species.
15 Kim et al. (2005) analyzed particle composition and gas phase data collected during the
16 RAPS/RAMS study on St. Louis, MO from 1975 to 1977 in terms of source contributions to
17 PM2.s. This study examined the spatial variability of source contributions to PM2.5 at the ten
18 monitoring sites in that study.
19 Sarnat et al. (2001) and reported associations between personal exposures and ambient
20 concentration across pollutants in a study conducted in the Baltimore area. At the ambient level,
21 NO2 was significantly correlated with PM2.5 (rs = 0.37) and CO (rs = 0.75) during the summer
22 and with CO (rs = 0.76), SO2 (rs = -0.17), PM2.5 (rs = 0.75) and O3 (rs = -0.71) during the winter.
23 Linn et al. (1996) reported short-term air pollution exposures in Los Angeles area school
24 children. Correlations between different pollutants were weaker: rp = 0.11 for ambient NO2 and
25 O3; rp = 0.25 for ambient NO2 and outdoor PM2.5.
26 Lee et al. (2002) found that ambient NO2 was significantly correlated with O3
27 (rp = -0.34).
28 Foreign Studies
29 Hochadel et al. (2006) reported the results of research which is part of a cohort study on
30 the impact of traffic-related air pollution on respiratory health, conducted at the western end of
31 the Ruhr-area in North-Rhine Westphalia, Germany. Strong correlations across the measurement
March 2008 AX3-87 DRAFT-DO NOT CITE OR QUOTE
-------
1 sites were observed between annual average PM2.5 absorbance and NO2 concentrations
2 (rp = 0.93), whereas PM2.5 mass concentration was less strongly correlated with NO2 (rp = 0.41).
3 The only major absorbing agent in PM2.5 is elemental carbon (EC) as other components (sulfate,
4 nitrate, organic carbon) either do not absorb or at best are only weakly absorbing. Therefore,
5 correlations between PM2 5 absorbance and NO2 may be inferred as correlations between EC
6 and NO2.
7 Hazenkamp-von Ark et al. (2004) reported the PM2 5 and NO2 associations across 21
8 European study centers during ECRHS II. The correlation between annual NO2 and PM2.5
9 concentrations is fair (Spearman correlation coefficient rs = 0.75), but when considered as
10 monthly means, the correlation is far less consistent and varies substantially between centers.
11 The authors pointed out that NO2 is attributed to traffic emissions, a relatively constant source of
12 pollution throughout the year. PM2 5 on the other hand, can be driven by other sources such as
13 wind-blown dust, although usually it consists predominantly of primary and secondary particles
14 from combustion processes. Sources, such as Saharan dust in Spain, probably cause some of the
15 observed patterns. The wide range of correlations between PM2 5 and NO2 evokes the hypothesis
16 that monthly PM2 5 mass concentrations in some centers may be driven by traffic emissions,
17 whereas in other centers, particles from other sources may be of further relevance.
18 Cyrys et al. (2003) reported the results of a source apportionment study in Erfurt,
19 Germany. Hourly NO2 was correlated with NO, CO, PM0.oi-2.5 number concentration, and
20 PM0.oi-2.5 mass concentration (rp = 0.73, 0.74, 0.55, and 0.50 respectively). Stronger correlations
21 were found daily correlation between NO2 and NO, CO, PMo.oi-2.5 number concentration, and
22 PMo.oi-2.5 mass concentration (rp = 0.87, 0.76, 0.71, and 0.66 respectively). The observed high
23 correlations between CO, NO, and NO2 indicate that direct emissions from mobile sources might
24 be the major contributors to the concentrations of these gaseous pollutants.
25 Rojas-Bracho et al. (2002) conducted a study of children's exposures in Santiago, Chile.
26 During the study, indoor, outdoor, and personal PM2.5, PMio, PMio-2.5, and NO2 were measured
27 24 h averaged for five consecutive days). Outdoor NO2 was significantly associated with all PM
28 fractions (slope = 1.82 and R2 = 0.59 for PM2.5; slope = 3.12 and R2 = 0.57 for PMi0; and slope =
29 1.11 and R2 = 0.32 for PM2.5-io).
30 Modig et al. (2004) investigated whether NO2 can be used to indicate ambient and
31 personal levels of benzene and 1, 3-butadiene in air. The stationary measurements showed
March 2008 AX3-88 DRAFT-DO NOT CITE OR QUOTE
-------
1 strong relations between 1,3-butadiene, benzene and NO2 (rp = 0.70 for NO2 and benzene; and
2 r = 0.77 for NC>2 and 1,3-butadiene). This study supports NC>2 as a potential indicator for
3 1,3 butadiene and benzene levels in streets or urban background air.
4 In summary, ambient NC>2 was moderately correlated with corresponding ambient
5 concentrations of its co- pollutants. Based on associations in the ambient environment, results
6 suggest a possibility of confounding of ambient NC>2 health effects by ambient PM2.5 (and its
7 components) and by ambient CO.
8
9 2) Associations between Personal (NO 2) and Personal Copollutant Exposures
10 For this section, measured personal NC>2 exposures are regarded as the "true" personal
11 exposure. The correlation between personal NC>2 exposure and personal exposure to other
12 pollutants are summarized below in Table AX3.6-2.
13 In Kim et al. (2006), the median, mean and standard deviation of the correlation between
14 NC>2 and PM2.5 personal exposures for eleven subjects were 0.43, 0.41, and 0.28 respectively;
15 and 0.16, 0.12, and 0.42 respectively for the correlation between NO2 and CO (Kim et al., 2006).
16 Although Sarnat et al. (2001) found that personal exposures to PM2.5 were generally not
17 significantly associated with personal exposures to gases in Baltimore, personal NO2 was
18 significantly associated with personal PM2.5 (slope = 0.18, intercept = 18.65, p < 0.01, and
19 n = 213) and personal PM2.5 of ambient origin (slope = 0.17, intercept = 12.77, p < 0.05, and
20 n = 150) during the summer. There was some evidence to indicate that the strength of the
21 association was driven largely by the cohort of older adult subjects, and not by the children's or
22 COPD patients cohorts. They noted that gas stove usage did not significantly affect personal
23 NO2 to PM2.5 relations, but did affect relations between personal NO2 and personal PM2.5 of
24 ambient origin. They further pointed out that associations observed among pollutants in ambient
25 air may not be reflected in personal exposures and that they may not persist across seasons.
26 However, Lai et al. (2004) found that personal exposure to NO2 was slightly negatively
27 correlated with personal exposure to PM2.5 and total VOCs in a study conducted from 1998 to
28 2000 in Oxford, UK (- 0.1 for PM2.5, 0.3 for CO, and -0.11 for TVOCs).
29 Modig et al. (2004) investigated whether NO2 can be used to indicate ambient and
30 personal levels of benzene and 1, 3-butadiene in air. The results from the personal
31 measurements showed a negligible association of NO2 with 1,3-butadiene (rp = 0.06) as well as
March 2008 AX3-89 DRAFT-DO NOT CITE OR QUOTE
-------
1 with benzene (rp = 0.10), while the correlation coefficient between benzene and 1,3-butadiene
2 was high and significant (rp = 0.67). The weak relations found for the personal measurements do
3 not support the use of NO2 as an indicator for personal 1,3-butadiene and benzene exposure.
4 Although gas stove and kerosene heaters were almost absent in the study area, this study
5 included both smokers and non-smokers, but the data were not stratified. Smoking is a major
6 source of both benzene and 1,3-butadiene, in addition to motor vehicles. If smoking were the
7 major cause of the poor association between NO2 and the gases in the personal measurements,
8 then this would indicate that smoking was not a major source of personal NO2. Thus, this study
9 cannot determine whether personal NO2 is an indicator of traffic generated VOCs and so the
10 interpretation of results in this paper is problematic.
11 In the Paris office worker study, no relation was observed between personal NO2 and
12 PM2.5 exposures (rp = 0.12, n = 53, p = 0.38) (Mosqueron et al., 2002). In addition, NO2 and
13 PM2.s concentrations were correlated neither in-home (rp = 0.06, n = 54, p = 0.69) nor in-office
14 (rp = 0.05, n = 55, p = 0.74).
15
16 Associations with HONO
17 Spicer et al. (1993) and Wainman et al. (2000) suggested the presence of a strong indoor
18 source of HONO from heterogeneous reactions involving NO2 and water films on indoor
19 surfaces. Hence, combustion appliances are sources for exposures to both NO2 exposure and
20 HNO2. Epidemiological studies of NO2 health effects should consequently consider the potential
21 confounding effects of NO2 and vice versa.
22 Jarvis et al. (2005) reported the indoor nitrous acid and lung function in adults as part of
23 European Community Respiratory Health Survey (ECRHS). Indoor HONO and indoor and
24 outdoor NO2 were measured. Indoor NO2 were correlated with HONO (rp = 0.77) but no
25 significant association of indoor NO2 with symptoms or lung function was observed.
26 Lee et al. (2002) studied the nitrous acid, nitrogen dioxide, and ozone concentrations in
27 residential environments. The authors found that indoor NO2 was significantly correlated with
28 HONO (rp = 0.511).
29 As shown above, very few studies showed the relationship between personal NO2
30 exposure and other pollutant exposures. In general, personal NO2 was moderately correlated
31 with PM2.s and CO. Due to the lack of personal HONO exposure data, indoor HONO was used
32 as an indicator for personal exposure, and current studies showed that indoor HONO was
March 2008 AX3-90 DRAFT-DO NOT CITE OR QUOTE
-------
1 correlated with indoor NC>2 with high correlation coefficients, which suggested that the collect
2 ion of HONO exposure data would help interpret adverse health outcome in the NO2 health risk
3 assessment.
4
5 3) Personal (NO2) -Ambient Copollutants
6 The relationship between personal NO2 exposure and other ambient pollutants are
7 summarized in Table AX3.6-3.
8 In Steubenville, Ohio, Sarnat et al. (2006) found that personal NO2 was significantly
9 associated with ambient PM2.s and ambient sulfate during the fall (slope = 0.17 and R2 = 0.21 for
10 PM2.s; and slope = 0.34 and R2 = 0.12 for sulfate); and was significantly associated with ambient
11 EC in both summer and fall (slope =1.81 and R2 = 0.03 for the summer; and slope = 3.71 and
12 R2 = 0.32 for the fall).
13 Kim et al. (2006) also reported correlations between personal exposure and ambient
14 measurements across pollutants. The median, mean, and standard deviation of the correlation
15 between personal NC>2 and ambient PM2.5 were 0.36, 0.30, and 0.30 respectively; and 0.17, 0.20,
16 and 0.41 respectively for the correlation between personal NC>2 and ambient CO. The authors
17 suggested that the existing correlation between PM2.5 and NO2 for both ambient measurements
18 and personal exposures suggests that there is potential for NO2 to be a confounder of PM2.5, and
19 vice versa. Therefore, it may be appropriate for time-series epidemiological studies to control
20 for confounding by NO2 in PM2.5 risk models, and vice versa.
21 In a study conducted in Santiago, Chile (Rojas-Bracho et al., 2002) personal NO2 was
22 moderately associated with PM2.5 (slope = 1.99 and r2 = 0.42) and PMi0 (slope = 2.13 and
23 r2 = 0.15) but not coarse particles. At the indoor level, the same observation held (slope = 0.86
24 and r2 = 0.22 for PM2.5; slope =1.0 and R2 = 0.2 for PMio). "However, in comparing the indoor
25 and outdoor associations, we find that the latter is more highly significant and that the intercept
26 is smaller. It is likely that in outdoor environments, there are more high-temperature combustion
27 processes, which are associated with nitrogen oxide emissions. Since nitrogen oxides are
28 precursors of secondary particles, which partly form PM2.5, our results showed a stronger
29 association between these two pollutions outdoors."
March 2008 AX3-91 DRAFT-DO NOT CITE OR QUOTE
-------
1 Lee et al. (2002) studied nitrous acid, nitrogen dioxide, and ozone concentrations in
2 residential environments. The authors found that indoor NC>2 was significantly correlated with
3 outdoor O3 (rp = - 0.220).
4 These studies above show moderate correlations between personal NO2 exposure and
5 ambient PM2.5, PMio, EC, sulfate, and CO. Based on our knowledge that, moderate to strong
6 personal-ambient correlations exist for all the other pollutants mentioned above all of those
7 species might serve as confounders for NC>2 exposure (detailed evaluation of the personal vs.
8 ambient relationship for these pollutants are beyond the scope of this document).
9
10 4) Ambient NO2-Personal Copollutant
11 Correlation between ambient NC>2 and personal exposure to copollutants are summarized
12 in Table AX3.6-4.
13 Sarnat et al. (2006) found that ambient NO2 was significantly associated with personal
14 PM2.s and personal sulfate during the fall (slope = 0.93 and R2 = 0.25 for PM2.5; and
15 slope = 0.28 and R2 = 0.27 for sulfate); and was significantly associated with personal EC during
16 both summer and fall (slope = 0.02 and R2 = 0.07 during the summer; and slope = 0.08 and
17 R2 = 0.49 during the fall) in Steubenville, OH. Sarnat et al. (2006) suggested that for most cases,
18 ambient gas concentrations, although not suitable proxies of gas exposures are equally not
19 suitable for particle exposures in time-series health studies. Despite this, numerous
20 epidemological studies have linked 24-h ambient gas concentrations to adverse health impacts,
21 suggesting that the gases may indeed elicit biological responses alone or in combination with
22 other pollutants, or are acting as proxies for shorter-term exposures. The authors pointed out that
23 for Steubenville in the fall, a season with strong associations between ambient particle and NO2
24 concentrations, the separation of particle and NO2 health effects in daily time-series studies may
25 be difficult, and more precise exposure metrics may be needed. The authors suggested that
26 personal-ambient relationships are greatly dependent on ambient conditions (e.g., season and
27 meteorology) and behavior (e.g., use of windows). However, further factors such as building
28 design will also be extremely important, further exposure assessment work, particularly in
29 different geographic and climatic zones, is needed.
30 During both summer and winter in Baltimore (Sarnat et al., 2001), ambient NO2 was
31 significantly associated with personal PM2.5 (slope = 0.42, intercept = 12.38, and n = 225 during
March 2008 AX3-92 DRAFT-DO NOT CITE OR QUOTE
-------
1 the summer; and slope = 0.24, intercept = 13.16, and n = 487 during the winter). Also significant
2 relationships held for ambient NC>2 and personal exposures to PM2.5 of ambient origin. Ambient
3 NO2 was also significantly associated with personal EC (slope = 0.05 and p = 0.0001), as an
4 indicator of mobile source pollution. In conclusion, the authors suggested that ambient gases
5 were acting as surrogates for personal PM2.5 exposure instead of confounding effects of personal
6 PM2.5 exposure.
7 Vinzents et al. (2005) found that ambient temperature and NC>2 concentrations at one of
8 the street stations were the only significant predictors of ultra fine particle exposure during
9 bicycling in traffic (R2 = 0.74). Kim et al. (2006) also reported correlations between personal
10 exposure and ambient measurements across pollutants. The median, mean, and standard
11 deviation of the correlation between ambient NC>2 and personal PM2.s were 0.24, 0.29, and 0.33
12 respectively; and 0.26, 0.22, and 0.32 respectively for the correlation between ambient NC>2 and
13 personal CO.
14 Studies above shows that ambient NC>2 is moderately correlated with personal EC and
15 ultrafme particle exposures, but only weakly to moderately correlated with personal PM2.5 mass
16 and sulfate exposures. Since ambient NC>2 concentrations has been shown to be significant
17 proxy for corresponding personal NC>2 exposures, these findings suggest that ambient NC>2 may
18 be acting as a proxy not only for its own exposures but also to exposures to EC and ultrafme
19 particles. As a result, it may not be possible to separate the health effects of from those of other
20 pollutants, especially from the same source.
21 In the analysis of the confounding effect of exposure, we are limited by the lack of key
22 data: (1) multipollutant exposure studies were rarely conducted and even fewer studies reported
23 the cross-level (ambient and personal exposure) and cross-pollutant correlations; (2) most studies
24 focus on a several copollutants (PM and its components, CO, Os, and some VOCs) with little
25 data available for other possibly important copollutants; (3) the impact of indoor and personal
26 sources on the possibility of confounding has not yet been examined; and (4) the impact of
27 measurement uncertainties, which can be large as mentioned in Section AX3.4.1, on
28 confounding needs to be examined. Finally, the analysis shown above in the exposure
29 assessment should be integrated with other analysis in other parts of the risk assessment.
30
31
March 2008 AX3-93 DRAFT-DO NOT CITE OR QUOTE
-------
1 AX3.7 A FRAMEWORK FOR MODELING HUMAN EXPOSURES TO
2 NO2 AND RELATED PHOTOCHEMICAL AIR POLLUTANTS
o
4 AX3.7.1 Introduction: Concepts, Terminology, and Overall Summary
5 Predictive (or prognostic) exposure modeling studies1, specifically focusing on NO2,
6 could not be identified in the literature, though, often, statistical (diagnostic) analyses have been
7 reported using data obtained in various field exposure studies (see Section AX3.5.1). However,
8 existing prognostic modeling systems for the assessment of inhalation exposures can in principle
9 be directly applied to, or adapted for, NO2 studies; specifically, such systems include APEX,
10 SHEDS, and MENTOR-1 A, to be discussed in the following sections. Nevertheless, it should be
11 mentioned that such applications will be constrained by data limitations, such as the degree of
12 ambient concentration characterization (e.g., concentrations at the local level) and quantitative
13 information on indoor sources and sinks.
14 Predictive models of human exposure to ambient air pollutants such as NO2 can be
15 classified and differentiated based upon a variety of attributes. For example, exposure models
16 can be classified as:
17 • models of potential (typically maximum) outdoor exposure versus models of
18 actual exposures (the latter including locally modified microenvironmental
19 exposures, both outdoor and indoor),
20 • Population Based Exposure Models (PBEM) versus Individual Based Exposure
21 Models (IBEM),
22 • deterministic versus probabilistic (or statistical) exposure models,
23 • observation-driven versus mechanistic air quality models (see Section AX3.7.3
24 for discussions about the construction, uses and limitations of this class of
25 mathematical models.
26 Some points should be made regarding terminology and essential concepts in exposure
27 modeling, before proceeding to the overview of specific developments reported in the current
28 research literature:
1 i.e. assessments that start from emissions and demographic information and explicitly consider the physical and
chemical processes of environmental and microenvironmental transport and fate, in conjunction with human
activities, to estimate inhalation intake and uptake.
March 2008 AX3-94 DRAFT-DO NOT CITE OR QUOTE
-------
1 First, it must be understood that there is significant variation in the definitions of many of
2 the terms used in the exposure modeling literature; indeed, the science of exposure modeling is a
3 rapidly evolving field and the development of a standard and commonly accepted terminology is
4 an ongoing process (see, e.g., WHO, 2004).
5 Second, it should also be mentioned that, very often, procedures that are called exposure
6 modeling, exposure estimation, etc. in the scientific literature, may in fact refer to only a sub-set
7 of the complete set of steps or components required for a comprehensive exposure assessment.
8 For example, certain self-identified exposure modeling studies focus solely on refining the sub-
9 regional or local spatio-temporal dynamics of pollutant concentrations (starting from raw data
10 representing monitor observations or regional grid-based model estimates). Though not
11 exposure studies per se, such efforts have value and are included in the discussion of the next
12 sub-section, as they provide potentially useful tools that can be used in a complete exposure
13 assessment. On the other hand, formulations that are self-identified as exposure models but
14 actually focus only on ambient air quality predictions, such as chemistry-transport models, are
15 not included in the discussion that follows.
16 Third, the process of modeling human exposures to photochemical pollutants
17 (traditionally focused on ozone) is very often identified explicitly with population-based
18 modeling, while models describing the specific mechanisms affecting the exposure of an actual
19 individual (at specific locations) to an air contaminant (or to a group of co-occurring gas and/or
20 aerosol phase pollutants) are usually associated with studies focusing specifically on indoor air
21 chemistry modeling.
22 Finally, fourth, the concept of microenvironments, introduced in earlier sections of this
23 document, should be clarified further, as it is critical in developing procedures for exposure
24 modeling. In the past, microenvironments have typically been defined as individual or aggregate
25 locations (and sometimes even as activities taking place within a location) where a homogeneous
26 concentration of the pollutant is encountered. Thus a microenvironment has often been
27 identified with an ideal (i.e. perfectly mixed) compartment of classical compartmental modeling.
28 More recent and general definitions view the microenvironment as a control volume, either
29 indoors or outdoors, that can be fully characterized by a set of either mechanistic or
30 phenomenological governing equations, when appropriate parameters are available, given
31 necessary initial and boundary conditions. The boundary conditions typically would reflect
March 2008 AX3-95 DRAFT-DO NOT CITE OR QUOTE
-------
1 interactions with ambient air and with other microenvironments. The parameterizations of the
2 governing equations generally include the information on attributes of sources and sinks within
3 each microenvironment. This type of general definition allows for the concentration within a
4 microenvironment to be non-homogeneous (non-uniform), provided its spatial profile and
5 mixing properties can be fully predicted or characterized. By adopting this definition, the
6 number of microenvironments used in a study is kept manageable, but variability in
7 concentrations in each of the microenvironments can still be taken into account.
8 Microenvironments typically used to determine exposure include indoor residential
9 microenvironments, other indoor locations (typically occupational microenvironments), outdoors
10 near roadways, other outdoor locations, and in-vehicles. Outdoor locations near roadways are
11 segregated from other outdoor locations (and can be further classified into street canyons,
12 vicinities of intersections, etc.) because emissions from automobiles alter local concentrations
13 significantly compared to background outdoor levels. Indoor residential microenvironments
14 (kitchen, bedroom, living room, etc. or aggregate home microenvironment) are typically
15 separated from other indoor locations because of the time spent there and potential differences
16 between the residential environment and the work/public environment.
17 Once the actual individual and relevant activities and locations (for Individual Based
18 Modeling), or the sample population and associated spatial (geographical) domain (for
19 Population Based Modeling) have been defined along with the temporal framework of the
20 analysis (time period and resolution), the comprehensive modeling of individual/population
21 exposure to NO2 (and related pollutants) will in general require seven steps (or components, as
22 some of them do not have to be performed in sequence) that are listed below. This list represents
23 a composite based on approaches and frameworks described in the literature over the last twenty -
24 five years (Ott, 1982; Ott, 1985; Lioy, 1990; U.S. Environmental Protection Agency, 1992;
25 Georgopoulos and Lioy, 1994; U.S. Environmental Protection Agency, 1997; Buck et al., 2003;
26 Price et al., 2003; Georgopoulos et al., 2005; WHO, 2005; U.S. Environmental Protection
27 Agency, 2006a; Georgopoulos and Lioy, 2006) as well on the structure of various inhalation
28 exposure models (NEM/pNEM, HAPEM, SHEDS, REHEX, EDMAS, MENTOR, ORAMUS,
29 APEX, AIRPEX, AIRQUIS, etc., to be discussed in the following section) that have been used in
30 the past or in current studies to specifically assess inhalation exposures. Figure AX3.7-1,
March 2008 AX3-96 DRAFT-DO NOT CITE OR QUOTE
-------
r
Calculate
potential
outdoor
exposures
L
•'i,a. Emissions: NEt (NET, NTI), state; Processing with SMOKE,
EMS-HAP, MOBILE/MOVES, NONROAD, FAAED, BEIS, etc.
i.b. Meteorology: NWS, NCDC; Modeling MM5, RAMS, CALMET
i.e. Land Use/ land Cover, Topography: NLDC (USGS), etc.
*
gg Estimate background
levels of air pollutants
through
a. mutttvariate spatio-
temporal analysis of
monitor data
b. emissions-base^ air
quality modeling
{with regional,
grid -based models:
Models- 3/ CMAQ, CAMx
and REMSAD)
•. Develop database of
' individual subjects
attributes (residence &
wortc location, housing
characteristics, age,
gender, race, income, file.)
a. collect study-specific
information
b. supplement with
available relevant local,
regional, and national
demographic
information
X Study-specific survey
(also US Census,
US Housing Survey)
•*•
"+>
*
' estimate local outdoor
. pollutant levels that
characterize the ambient air of
an administrative unit (such as
census tract) or a conveniently
defined grid through
a. spatiotemporaS statistical
analysis of monitor data
t>. application of urban scale
model at high resolution
c subgrid (e.g. pfume-m-grid)
modeling
d. data/model assimilation
4 Develop activity event
(or exposure event)
sequences for each individual
of the study for the exposure
peHml
a. collect stu4y-#pedfic
information
b. supplement with ottier
available data
c, organize time-activity
database in format
compatible wtth CHAD
/ Study-specific survey
i (or defaw It from
I CMAD^ NMAF^)
; ii.a. Emissions: EMS-HAP
il.b. Local Meteorology - Local
Effects: RAMS, FLUENT
s +
•*
/*
»•#
;v; Estimate levels and
^'-~ temporal profiles of
pollutants m various
mlcroenvironments (streets,
residences, offices, restaurants,
vehicles, etc.) through
a. regression of observational
data
b. simple linear mass balance
€. lumped (nonlinear) ||E
9as/aerosol chemistry models
d. combined chemistry & CR>
(DNS, LES, RANS) models
*
i " Calculate appropriate
, initiation rates for the
members of ttie sample
population,, combining the
physiological attilbtites of the
study subjects and the
activities pursued during the
mdivMuaS exposure evem^s
f
IOM» and Other
WifsiologicalftHfTS
Oalabines
— -v
^>
1 1
Calculate
exposures/
intakes
_^l
[ "' Biologically
' based
target tissue
dose modeling
J
Figure AX3.7-1.
Schematic description of a general framework identifying the
processes (steps or components) involved in assessing inhalation
exposures and doses for individuals and populations. In general
terms, existing comprehensive exposure modeling systems such as
SHEDS, APEX, and MENTOR-1A follow this framework.
1
2
3
4
5
6
7
8
9
10
Source: Figure adapted with modifications from Georgopoulos et al. (2005).
adapted from Georgopoulos et al. (2005), schematically depicts the sequence of steps involved
that are summarized here (and further discussed in the following sub-sections).
1. Estimation of the background or ambient levels of both NO2 and related
photochemical pollutants. This is done through either (or a combination of):
a. multivariate spatio-temporal analysis of fixed monitor data, or
b. emissions-based, photochemical, air quality modeling (typically with a
regional, grid-based model such as Models-3/CMAQ or CAMx) applied in
a coarse resolution mode.
March 2008
AX3-97
DRAFT-DO NOT CITE OR QUOTE
-------
1 2. Estimation of local outdoor pollutant levels of both NC>2 and related photochemical
2 pollutants. These levels could typically characterize the ambient air of either an
3 administrative unit (such as a census tract, a municipality, a county, etc.) or a
4 conveniently defined grid cell of an urban scale air quality model. Again, this may
5 involve either (or a combination of):
6
7 a. spatio-temporal statistical analysis of monitor data, or
8 b. application of an urban multi-scale, grid based model (such as CMAQ or
9 CAMx) at its highest resolution (typically around 2-4 km), or
10 c. correction of the estimates of the regional model using some scheme that
11 adjusts for observations and/or for subgrid chemistry and mixing
12 processes.
13 3. Characterization of relevant attributes of the individuals or populations under study
14 (residence and work locations, occupation, housing data, income, education, age,
15 gender, race, weight, and other physiological characteristics). For Population Based
16 Exposure Modeling (PBEM) one can either:
17
18 a. select a fixed-size sample population of virtual individuals in a way that
19 statistically reproduces essential demographics (age, gender, race,
20 occupation, income, education) of the administrative population unit used
21 in the assessment (e.g., a sample of 500 people is typically used to
22 represent the demographics of a given census tract, whereas a sample of
23 about 10,000 may be needed to represent the demographics of a county),
24 or
25 b. divide the population-of interest into a set of cohorts representing selected
26 subpopulations where the cohort is defined by characteristics known to
27 influence exposure.
28 4. Development of activity event (or exposure event) sequences for each member of
29 the sample population (actual or virtual) or for each cohort for the exposure period.
30 This could utilize:
March 2008 AX3-98 DRAFT-DO NOT CITE OR QUOTE
-------
1 a. study-specific information, if available
2 b. existing databases based on composites of questionnaire information from
3 past studies
4 c. time-activity databases, typically in a format compatible with U.S.
5 Environmental Protection Agency's Consolidated Human Activity
6 Database (CHAD - McCurdy et al., 2000)
7
8 5. Estimation of levels and temporal profiles of both NC>2 and related photochemical
9 pollutants in various outdoor and indoor microenvironments such as street canyons,
10 roadway intersections, parks, residences, offices, restaurants, vehicles, etc. This is
11 done through either:
12
13 a. linear regression of available observational data sets,
14 b. simple mass balance models (with linear transformation and sinks) over
15 the volume (or a portion of the volume) of the microenvironment,
16 c. lumped (nonlinear) gas or gas/aerosol chemistry models, or
17 d. detailed combined chemistry and Computational Fluid Dynamics
18 modeling.
19 6. Calculation of appropriate inhalation rates for the members of the sample
20 population, combining the physiological attributes of the (actual or virtual) study
21 subjects and the activities pursued during the individual exposure events.
22 7. Calculation of target tissue dose through biologically based modeling estimation
23 (specifically, respiratory dosimetry modeling in the case of NC>2 and related reactive
24 photochemical pollutants) if sufficient information is available.
25 Implementation of the above framework for comprehensive exposure modeling has
26 benefited significantly from recent advances and expanded availability of computational
27 technologies such as Relational Database Management Systems (RDBMS) and Geographic
March 2008 AX3-99 DRAFT-DO NOT CITE OR QUOTE
-------
1 Information Systems (GIS) (Purushothaman and Georgopoulos, 1997, 1999a,b; Georgopoulos
2 et al., 2005).
3 In fact, only relatively recently comprehensive, predictive, inhalation exposure modeling
4 studies for ozone, PM, and various air toxics, have attempted to address/incorporate all the
5 components of the general framework described here. In practice, the majority of past exposure
6 modeling studies have either incorporated only subsets of these components or treated some of
7 them in a simplified manner, often focusing on the importance of specific factors affecting
8 exposure. Of course, depending on the objective of a particular modeling study, implementation
9 of only a limited number of steps may be necessary. For example, in a regulatory setting, when
10 comparing the relative effectiveness of emission control strategies, the focus can be on expected
11 changes in ambient levels (corresponding to those observed at NAAQS monitors) in relation to
12 the density of nearby populations. The outdoor levels of pollutants, in conjunction with basic
13 demographic information, can thus be used to calculate upper bounds of population exposures
14 associated with ambient air (as opposed to total exposures that would include contributions from
15 indoor sources) useful in comparing alternative control strategies. Though the metrics derived
16 would not be quantitative indicators of actual human exposures, they can serve as surrogates of
17 population exposures associated with outdoor air, and thus aid in regulatory decision making
18 concerning pollutant standards and in studying the efficacy of emission control strategies. This
19 approach has been used in studies performing comparative evaluations of regional and local
20 emissions reduction strategies in the eastern United States (e.g., Purushothaman and
21 Georgopoulos, 1997; Georgopoulos et al., 1997a; Foley et al., 2003).
22
23 AX3.7.2 Population Exposure Models: Their Evolution and Current Status
24 Existing comprehensive inhalation exposure models consider the trajectories of
25 individual human subjects (actual or virtual), or of appropriately defined cohorts, in space and
26 time as sequences of exposure events. In these sequences each event is defined by time, a
27 geographic location, a microenvironment, and the activity of the subject. U.S. Environmental
28 Protection Agency offices (OAQPS and NERL) have supported the most comprehensive efforts
29 in developing models implementing this general concept (see, e.g., Johnson, 2002), and these
30 efforts have resulted in the NEM/pNEM (National Exposure Model and Probabilistic National
31 Exposure Model - Whitfield et al., 1997), HAPEM (Hazardous Air Pollutant Exposure Model -
March 2008 AX3-100 DRAFT-DO NOT CITE OR QUOTE
-------
1 Rosenbaum, 2005), SHEDS (Simulation of Human Exposure and Dose System - Burke et al.,
2 2001), APEX (Air Pollutants Exposure model - U.S. Environmental Protection Agency,
3 2006b,c), and MENTOR (Modeling Environment for Total Risk studies - Georgopoulos et al.,
4 2005; Georgopoulos and Lioy, 2006) families of models. European efforts have produced some
5 formulations with similar general attributes as the above U.S. models but, generally, involving
6 simplifications in some of their components. Examples of European models addressing
7 exposures to photochemical oxidants (specifically ozone) include the AirPEx (Air Pollution
8 Exposure) model (Freijer et al., 1998), which basically replicates the pNEM approach and has
9 been applied to the Netherlands, and the AirQUIS (Air Quality Information System) model
10 (Clench-Aas et al., 1999).
11 The NEM/pNEM, SHEDS, APEX, and MENTOR-1A (MENTOR for One-Atmosphere
12 studies) families of models provide exposure estimates defined by concentration and breathing
13 rate for each individual exposure event, and then average these estimates over periods typically
14 ranging from one h to one year. These models allow simulation of certain aspects of the
15 variability and uncertainty in the principal factors affecting exposure. An alternative approach is
16 taken by the HAPEM family of models that typically provide annual average exposure estimates
17 based on the quantity of time spent per year in each combination of geographic locations and
18 microenvironments. The NEM, SHEDS, APEX, and MENTOR-type models are therefore
19 expected to be more appropriate for pollutants with complex chemistry such as NO2, and could
20 provide useful information for enhancing related health assessments.
21
22 More specifically, regarding the consideration of population demographics and activity patterns:
23 1. pNEM divides the population of interest into representative cohorts based on the
24 combinations of demographic characteristics (age, gender, and employment),
25 home/work district, residential cooking fuel and replicate number, and then assigns
26 activity diary record from CHAD (Consolidated Human Activities Database) to each
27 cohort according to demographic characteristic, season, day-type
28 (weekday/weekend) and temperature.
29 2. HAPEM6 divides the population of interest into demographic groups based on age,
30 gender and race, and then for each demographic group/day-type (weekday/weekend)
31 combination, select multiple activity patterns randomly (with replacement) from
March 2008 AX3-101 DRAFT-DO NOT CITE OR QUOTE
-------
1 CHAD and combine them to find the averaged annual time allocations for group
2 members in each census tract for different day types.
3 3. SHEDS, APEX, and MENTOR-1A generate population demographic files, which
4 contain a user-defined number of person records for each census tract of the
5 population based on proportions of characteristic variables (age, gender,
6 employment, and housing) obtained from the population of interest, and then assign
7 a matching activity diary record from CHAD to each individual record of the
8 population based on the characteristic variables. It should be mentioned that, in the
9 formulations of these models, workers may commute from one census tract to
10 another census tract for work. So, with the specification of commuting patterns, the
11 variation of exposure concentrations due to commuting between different census
12 tracts can be captured.
13
14 The essential attributes of the pNEM, HAPEM, APEX, SHEDS, and MENTOR-1A
15 models are summarized in Table AX3.7-1.
16 The conceptual approach originated by the SHEDS models was modified and expanded
17 for use in the development of MENTOR-1 A (Modeling Environment for Total Risk - One
18 Atmosphere). Flexibility was incorporated into this modeling system, such as the option of
19 including detailed indoor chemistry of the Os-NOx system and other relevant
20 microenvironmental processes, and providing interactive linking with CHAD for consistent
21 definition of population characteristics and activity events (Georgopoulos et al., 2005).
22 NEM/pNEM implementations have been extensively applied to ozone studies in the
23 1980s and 1990s. The historical evolution of the pNEM family of models of OAQPS started
24 with the introduction of the first NEM model in the 1980's (Biller et al., 1981). The first such
25 implementations of pNEM/Os in the 1980's used a regression-based relationship to estimate
26 indoor ozone concentrations from outdoor concentrations. The second generation of pNEM/Os
27 was developed in 1992 and included a simple mass balance model to estimate indoor ozone
28 concentrations. A report by Johnson et al. (2000) describes this version of pNEM/Os and
29 summarizes the results of an initial application of the model to 10 cities. Subsequent
30 enhancements to pNEM/Os and its input databases included revisions to the methods used to
31 estimate equivalent ventilation rates, to determine commuting patterns, and to adjust ambient
March 2008 AX3-102 DRAFT-DO NOT CITE OR QUOTE
-------
1 ozone levels to simulate attainment of proposed NAAQS. During the mid-1990's,
2 Environmental Protection Agency applied updated versions of pNEM/Os to three different
3 population groups in selected cities: (1) the general population of urban residents, (2) outdoor
4 workers, and (3) children who tend to spend more time outdoors than the average child. This
5 version of pNEM/Os used a revised probabilistic mass balance model to determine ozone
6 concentrations over one-h periods in indoor and in-vehicle microenvironments (Johnson, 2001).
7 In recent years, pNEM has been replaced by (or "evolved to") the Air Pollution Exposure
8 Model (APEX). APEX differs from earlier pNEM models in that the probabilistic features of the
9 model are incorporated into a Monte Carlo framework (Langstaff, 2007; U.S. Environmental
10 Protection Agency, 2006b,c). Like SHEDS and MENTOR-1A, instead of dividing the
11 population-of-interest into a set of cohorts, APEX generates individuals as if they were being
12 randomly sampled from the population. APEX provides each generated individual with a
13 demographic profile that specifies values for all parameters required by the model. The values
14 are selected from distributions and databases that are specific to the age, gender, and other
15 specifications stated in the demographic profile. Environmental Protection Agency has applied
16 APEX to the study of exposures to ozone and other criteria pollutants; APEX can be modified
17 and used for the estimation of NC>2 exposures, if required.
18 Reconfiguration of APEX for use with NC>2 or other pollutants would require significant
19 literature review, data analysis, and modeling efforts. Necessary steps include determining
20 spatial scope and resolution of the model; generating input files for activity data, air quality and
21 temperature data; and developing definitions for microenvironments and pollutant-
22 microenvironment modeling parameters (penetration and proximity factors, indoor source
23 emissions rates, decay rates, etc.) (ICF Consulting 2005). To take full advantage of the
24 probablistic capabilities of APEX, distributions of model input parameters should be used
25 wherever possible.
26
27 AX3.7.3 Characterization of Ambient Concentrations of NO2 and Related
28 Air Pollutants
29 As mentioned earlier, background and regional outdoor concentrations of pollutants, over
30 a study domain, may be estimated either through emissions-based mechanistic modeling, through
31 ambient-data-based modeling, or through a combination of both. Emissions-based models
32 calculate the spatio-temporal fields of the pollutant concentrations using precursor emissions and
March 2008 AX3-103 DRAFT-DO NOT CITE OR QUOTE
-------
1 meteorological conditions as inputs. The ambient-data-based models typically calculate spatial
2 or spatio-temporal distributions of the pollutant through the use of interpolation schemes, based
3 on either deterministic or stochastic models for allocating monitor station observations to the
4 nodes of a virtual regular grid covering the region of interest. The geostatistical technique of
5 kriging provides various standard procedures for generating an interpolated spatial distribution
6 for a given time, from data at a set of discrete points. Kriging approaches were evaluated by
7 Georgopoulos et al. (Georgopoulos et al., 1997b) in relation to the calculation of local ambient
8 ozone concentrations for exposure assessment purposes, using either monitor observations or
9 regional/urban photochemical model outputs. It was found that kriging is severely limited by the
10 nonstationary character of the concentration patterns of reactive pollutants; so the advantages this
11 method has in other fields of geophysics do not apply here. The above study showed that the
12 appropriate semivariograms had to be hour-specific, complicating the automated reapplication of
13 any purely spatial interpolation over an extended time period.
14 Spatio-temporal distributions of pollutant concentrations, such as ozone, PM, and various
15 air toxics have alternatively been obtained using methods of the Spatio-Temporal Random Field
16 (STRF) theory (Christakos and Vyas, 1998a,b). The STRF approach interpolates monitor data in
17 both space and time simultaneously. This method can thus analyze information on temporal
18 trends, which cannot be incorporated directly in purely spatial interpolation methods such as
19 standard kriging. Furthermore, the STRF method can optimize the use of data that are not
20 uniformly sampled in either space or time. STRF was further extended within the Bayesian
21 Maximum Entropy (BME) framework and applied to ozone interpolation studies (Christakos and
22 Hristopulos, 1998; Christakos and Kolovos, 1999; Christakos, 2000). It should be noted that
23 these studies formulate an over-arching scheme for linking air quality with population dose and
24 health effects; however they are limited by the fact that they do not include any
25 microenvironmental effects. MENTOR has incorporated STRF/BME methods as one of the
26 steps for performing a comprehensive analysis of exposure to ozone and PM (Georgopoulos
27 et al., 2005).
28 Subgrid spatial variability is a major issue with respect to characterizing local
29 concentrations of NC>2. Indeed, the fast rates of the reactions involving the Os-NOx system result
30 in significant concentration gradients in the vicinity of sources of NOx. These gradients are not
31 resolved directly by currently operational grid photochemical air quality simulation models
March 2008 AX3-104 DRAFT-DO NOT CITE OR QUOTE
-------
1 (PAQSMs) such as CMAQ and CAMx. However, both these models include a plume-in-grid.
2 (PinG) option (AER, 2004; Emery and Yarwood, 2005; Gillani and Godowitch, 1999; U.S.
3 Environmental Protection Agency, 2006d) that can be used for large point NOX sources (such as
4 smokestacks). Nevertheless, PinG formulations typically will resolve gradients in upper
5 atmospheric layers and thus are not necessarily relevant to human exposure calculations, which
6 are affected by gradients caused by a multiplicity of smaller ground level or near ground level
7 combustion sources such as motor vehicles.
8 Currently PAQSMs are typically applied with horizontal resolutions of 36 km, 12 km,
9 and 4 km and a surface layer thickness that is typically of the order of 30 m. Though
10 computationally it is possible to increase the resolution of these simulations, there are critical
11 limits that reflect assumptions inherent in the governing equations for both (a) the fluid
12 mechanical processes embodied in the meteorological models (e.g., typically MM5 and RAMS)
13 that provide the inputs for the PAQSMs, and (b) the dispersion processes which become more
14 complex at fine scales (see, e.g., Georgopoulos and Seinfeld, 1989) and thus cannot be described
15 by simple formulations (such as constant dispersion coefficients) when the horizontal resolutions
16 is 2 km or finer.
17 Application of PAQSMs to urban domains is further complicated by urban topography,
18 the urban heat island, etc. It is beyond the scope, however, of the present discussion, to overview
19 the various issues relevant to urban fluid dynamics and related transport/fate processes of
20 contaminants. However, the issue of modeling subgrid atmospheric dispersion phenomena
21 within complex urban areas in a consistent manner is a very active research area. Reviews of
22 relevant issues and of available approaches for modeling urban fluid mechanics and dispersion
23 can be found in, e.g., Fernando et al. (2001) and Britter and Hanna (2003).
24 The issue of subgrid variability (SGV) from the perspective of interpreting and evaluating
25 the outcomes of grid-based, multiscale, PAQSMs is discussed in Ching et al. (2006), who
26 suggest a framework that can provide for qualitative judgments on model performance based on
27 comparing observations to the grid predictions and its SGV distribution. From the perspective of
28 Population Exposure Modeling, the most feasible/practical approach for treating subgrid
29 variability of local concentrations is probably through (1) the identification and proper
30 characterization of an adequate number of outdoor microenvironments (potentially related to
31 different types of land use within the urban area as well as to proximity to different types of
March 2008 AX3-105 DRAFT-DO NOT CITE OR QUOTE
-------
1 roadways) and (2) then, concentrations in these microenvironments will have to be adjusted from
2 the corresponding local background ambient concentrations through either regression of
3 empirical data or various types of local atmospheric dispersion/transformation models. This is
4 discussed further in the next subsection.
5
6 AX3.7.4 Characterization of Microenvironmental Concentrations
7 Once the background and local ambient spatio-temporal concentration patterns have been
8 derived, microenvironments that can represent either outdoor or indoor settings when individuals
9 come in contact with the contaminant of concern (e.g., NO2) must be characterized. This process
10 can involve modeling of various local sources and sinks, and interrelationships between ambient
11 and microenvironmental concentration levels. Three general approaches have been used in the
12 past to model microenvironmental concentrations:
13 • Empirical (typically linear regression) fitting of data from studies relating ambient/local
14 and microenvironmental concentration levels to develop analytical relationships.
15 • Parameterized mass balance modeling over, or within, the volume of the
16 microenvironment. This type of modeling has ranged from very simple formulations, i.e.
17 from models assuming ideal (homogeneous) mixing within the microenvironment (or
18 specified portions of it) and only linear physicochemical transformations (including
19 sources and sinks), to models incorporating analytical solutions of idealized dispersion
20 formulations (such as Gaussian plumes), to models that take into account aspects of
21 complex multiphase chemical and physical interactions and nonidealities in mixing.
22 • Detailed Computational Fluid Dynamics (CFD) modeling of the outdoor or indoor
23 microenvironment, employing either a Direct Numerical Simulation (DNS) approach, a
24 Reynolds Averaged Numerical Simulation (RANS) approach, or a Large Eddy
25 Simulation (LES) approach, the latter typically for outdoor situations (see, e.g., Milner
26 et al., 2005; Chang and Meroney, 2003; Chang, 2006).
27
28 Parameterized mass balance modeling is the approach currently preferred for exposure
29 modeling for populations. As discussed earlier, the simplest microenvironmental setting
30 corresponds to a homogeneously mixed compartment, in contact with possibly both
31 outdoor/local environments as well as other microenvironments. The air quality of this idealized
32 microenvironment is affected mainly by the following processes:
March 2008 AX3-106 DRAFT-DO NOT CITE OR QUOTE
-------
1 a. Transport processes: These can include advection/convection and dispersion that
2 are affected by local processes and obstacles such as vehicle induced turbulence,
3 street canyons, building structures, etc.
4 b. Sources and sinks: These can include local outdoor emissions, indoor emissions,
5 surface deposition, etc.
6 c. Transformation processes: These can include local outdoor as well as indoor gas
7 and aerosol phase chemistry, such as formation of secondary organic and inorganic
8 aerosols.
9
10 Examples of the above are discussed next, specifically for outdoor and for indoor
11 microenvironments.
12
13 AX3.7.4.1 Characterization of Outdoor Microenvironments
14 Empirical regression analyses have been used in some studies to relate specific outdoor
15 locations - that can be interpreted as generalized types of exposure microenvironments - to
16 spatial variability of NC>2 concentrations. For example, Gilbert et al. (2005) in May 2003
17 measured NO2 for 14 consecutive days at 67 sites across Montreal, Canada. Concentrations
18 ranged from 4.9 to 21.2 ppb (median 11.8 ppb), and they used linear regression analysis to assess
19 the association between logarithmic values of NO2 concentrations and land-use variables via a
20 geographic information system. In univariate analyses, NC>2 was negatively associated with the
21 area of open space and positively associated with traffic count on nearest highway, the length of
22 highways within any radius from 100 to 750 m, the length of major roads within 750 m, and
23 population density within 2000 m. Industrial land-use and the length of minor roads showed no
24 association with NC>2. In multiple regression analyses, distance from the nearest highway, traffic
25 count on the nearest highway, length of highways and major roads within 100 m, and population
26 density showed significant associations with NC>2. The authors of that study point out the value
27 of using land-use regression modeling to assign exposures in large-scale epidemological studies.
28 Similar analyses have been performed in a predictive setting by Sahsuvaroglu et al. (2006) for
29 Hamilton, Ontario, Canada.
30 The category of parameterized mass balance models for outdoor microenvironments
31 includes various local roadway, intersection, and street canyon models. For example, Fraigneau
March 2008 AX3-107 DRAFT-DO NOT CITE OR QUOTE
-------
1 et al. (1995) developed a simple model to account for fast nitrogen oxide - ozone
2 reaction/dispersion in the vicinity of a motorway. Venegas and Mazzeo (2004) applied a
3 combination of simple point and area source analytical plume models to characterize NO2
4 concentration patterns in Buenos Aires, Argentina, which they used for a simplified (potential)
5 population exposure study. ROADWAY-2 (Rao, 2002), is another near-highway pollutant
6 dispersion model that incorporates vehicle wake parameterizations derived from canopy flow
7 theory and wind tunnel measurements. The atmospheric velocity and turbulence fields are
8 adjusted to account for velocity-deficit and turbulence production in vehicle wakes and a
9 turbulent kinetic energy closure model of the atmospheric boundary layer is used to derive the
10 mean velocity, temperature, and turbulence profiles from input meteorological data.
11 In parameterized street canyon models, typically, concentrations of exhaust gases are
12 calculated using a combination of a plume model for the direct contribution and a box model for
13 the recirculating part of the pollutants in the street. Parameterization of flow and dispersion
14 conditions in these models is usually deduced from analysis of experimental data and model tests
15 that considered different street configurations and various meteorological conditions.
16 An example of a current model that belongs in the parameterized mass balance category is the
17 Danish Operational Street Pollution Model (OSPM) (Berkowicz, 2002), which updates earlier
18 formulations of street canyon models such as STREET of Johnson et al. (1973) and CPBM
19 (Canyon Plume-Box Model) of Yamartino and Weigand (1986). A variation of this simple
20 approach is the model of Proyou et al. (1998), which uses a three-layer photochemical box model
21 to represent a street canyon.
22 A variety of CFD based street canyon models have been developed in recent years (see,
23 e.g., the series of International Conferences on Harmonization - http://www.harmo.org),
24 employing various alternatives for closure of the turbulent transport equations. A review and
25 intercomparison of five of these models (CHENSI, CHENSI-2, MIMO, MISKAM, TASCflow)
26 vis-a-vis field data from a street canyon in Hannover, Germany can be found in the articles by
27 Sahm et al. (2002) and by Ketzel et al. (2002).
28 These complex localized models could be useful for improving population exposure
29 model estimates by calculating pollutant concentrations at the microenvironmental level. Lack
30 of input parameter data and parameter variation across the modeling domain (spatial and
31 temporal) contributes to uncertainty in microenvironmental concentrations calculated by exposre
March 2008 AX3-108 DRAFT-DO NOT CITE OR QUOTE
-------
1 models. In such cases, parameterized mass balance models could provide outdoor concentration
2 values for estimating exposure. If infiltration factors are known, these concentrations could also
3 be used to estimate indoor exposures.
4
5 AX3.7.4.2 Characterization of Indoor Microenvironments
6 Numerous indoor air quality modeling studies have been reported in the literature;
7 however, depending on the modeling scenario, only few of them address (and typically only a
8 limited subset of) physical and chemical processes that affect photochemical oxidants indoors
9 (Nazaroff and Cass, 1986; Hayes, 1989,1991; Freij er and Bloemen, 2000).
10 It is beyond the scope of the present discussion to review in detail the current status of
11 indoor air modeling. Existing indoor air concentration models indeed are available as a wide
12 range of (a) empirical regression relationships, (b) parameterized mass balance models (that can
13 be either single-zone—that is, single well-mixed room—or multi-zone models), and (c) CFD
14 formulations. Recent overviews of this area can be found in Milner et al. (2005), who focus, in
15 particular, on the issue of entrainment from outdoor sources, and in Teshome and Haghighat,
16 (2004), who focus on different formulations of zonal models and on how they compare with
17 more complex CFD models.
18 Few indoor air models have considered detailed nonlinear chemistry, which, however,
19 can have a significant effect on the indoor air quality, especially in the presence of strong indoor
20 sources (e.g., gas stores and kerosene heaters, in the case of NCh). Indeed, the need for more
21 comprehensive models that can take into account the complex, multiphase processes that affect
22 indoor concentrations of interacting gas phase pollutants and particulate matter has been
23 recognized and a number of formulations have appeared in recent years. For example, the
24 Exposure and Dose Modeling and Analysis System (EDMAS) (Georgopoulos et al., 1997c)
25 included an indoor model with detailed gas-phase atmospheric chemistry to estimate indoor
26 concentrations resulting from penetration and reaction of ambient pollutants. This indoor model
27 was dynamically coupled with (a) the outdoor photochemical air quality models UAM-IV and
28 UAM-V, which provided the gas-phase composition of influent air; and (b) with a
29 physiologically based uptake and dosimetry model. Subsequent work (Isukapalli et al., 1999)
30 expanded the approach of the EDMAS model to incorporate alternative representations of gas-
31 phase chemistry as well as multiphase photochemistry and gas/aerosol interactions. The
32 microenvironmental model corresponding to this more general formulation is mathematically
March 2008 AX3-109 DRAFT-DO NOT CITE OR QUOTE
-------
1 represented by the following equation, when an assumption of uniform mixing is used for each
2 component (e.g., individual room) of the indoor environment. Sarwar et al. (2001) presented a
3 more comprehensive modeling study of the gas phase aspects of ozone indoor chemistry
4 focusing on the impact of different factors (such as outdoor ozone, indoor emissions, ventilation
5 rates, etc.) on the levels of indoor hydroxyl radicals (OH), which in turn are expected to control
6 the rate of formation of secondary toxicants indoors.
j
V, i /
Jj ,— 1 J J .— j j J . — j
1 dt i i j-i j-i (AX3.7-1)
8 where,
9 Vi = volume of compartment (m3)
10 C; = concentration of species in compartment (mol/m3)
11 K[j = mass transfer coefficient from compartment (m/h)
12 ay = interfacial air exchange area between compartments (m2)
13 QJ = concentration in compartment i in equilibrium with concentration in j (mol/m3)
14 Qij = volumetric flow rate from compartment i to j (m3/h)
15 R[ = rate of formation of species in compartment i (gmol/h)
16
17 and,
c _ c _ c
c / °/, emis °i, aepos °i,conaens ;for gases
'(C _C -t-C 4-C* 4-C 4-C
ig ^i, emis ^i.depos ^i.restisp ^i.condens ^i.nucl ^i.coag; Jor PM TAX3 7-2^)
19 More recent work (S0rensen and Weschler, 2002) has coupled CFD calculations with
20 gas-phase atmospheric chemistry mechanisms to account for the impact of nonideal flow mixing
21 (and associated concentration gradients) within a room on the indoor spatial distribution of ozone
22 and other secondary pollutants. This work has identified potential limitations associated with the
23 assumption of uniform mixing in indoor microenvironments when calculating personal
24 exposures.
25 A recent indoor air model that specifically focuses on NC>2 (along with CO, PMio, and
26 PM2.5 is INDAIR (Dimitroulopoulou et al., 2006). The INDAIR model considers three
27 interconnected residential microenvironments: kitchen, lounge, and bedroom. Removal
March 2008 AX3-1 10 DRAFT-DO NOT CITE OR QUOTE
-------
1 processes are lumped together and quantified via an apparent deposition velocity. Specifically, a
2 loss rate of 0.99 ± 0.19 IT1 (Yamanaka, 1984), is used in this model corresponding to a mean
3 deposition velocity of 1.2 x 10~4 m s~l. The sources of NC>2 considered in INDAIR are from gas
4 stove cooking and from cigarette smoking, but only the former contributes significantly to indoor
5 NC>2 levels, based on available model parameterizations.
6 Estimation of NC>2 emission rates from gas cooking utilized the following empirical
7 information: (a) NOx emission rate equal to 0.125 g kWlT1 (Wooders, 1994); (b) an assumption
8 that NO2 represents 25% of the total NOx emissions and (c) gas consumption per household in
9 cooking equal to 5-7 kWh day'1, assuming 1 h cooking per day. By multiplying the estimates in
10 (a), (b), and (c) together, NO2 gas cooking emission rates were calculated to be in the range 0.16
11 to 0.22 g h'1, with a uniform distribution.
12 In a range of simulations performed with INDAIR for houses in the UK, it was found that
13 the predicted maximum 1-h mean concentrations in the kitchen were increased, compared to no-
14 source simulations, by a factor of 10 for NC>2 (30 for PMi0 and 15 for PM2.5) and were higher in
15 winter than in summer. Cooking activity in the kitchen resulted in significantly elevated 24 h
16 mean concentrations of NC>2, PMio, and PM2.5 in the lounge, as well as the kitchen, while there
17 was a relatively small effect in the bedroom, which was not connected directly to the kitchen in
18 the model structure (i.e., the direct internal air exchange rate was zero).
19 A very wide range of predictions was derived from the INDAIR simulations. The 95th
20 percentile concentrations were typically 50% higher than mean concentrations during periods of
21 average concentration, and up to 100% higher than mean concentrations during concentration
22 peaks, which were associated with cooking emissions. There was approximately a factor of
23 2 variation in concentrations, and all modeled concentrations were below those outdoors. The
24 effect of cooking was to shift the distribution to the right, but the degree of variation was not
25 greatly increased. This may reflect the fact that for the fixed emission scenarios that were used,
26 the additional variation in emission rates was small compared to that of other factors such as
27 deposition rate and air exchange rate. In this scenario, modeled concentrations in the lounge all
28 remained below those outdoors, but a proportion of kitchens (16%) had modeled values above
29 the outdoor concentration. For the gas-cooking scenario, indoor/outdoor ratios for NC>2 ranged
30 from 0.5 to 0.8 for the bedroom, 0.7 to 1.6 for the lounge and 0.9 to 3.6 for the kitchen.
31 According to Dimitrolopoulou et al. (2006), these results were broadly consistent with
March 2008 AX3-111 DRAFT-DO NOT CITE OR QUOTE
-------
1 indoor/outdoor ratios reported for the UK. Modeled peak concentrations associated with gas
2 cooking, of about 300 ppb in the kitchen and 100 ppb in the lounge, were also consistent with
3 results from UK studies.
4
5 AX3.7.4.3 Characterization of Activity Events
6 An important development in inhalation exposure modeling has been the consolidation of
7 existing information on activity event sequences in the Consolidated Human Activity Database
8 (CHAD) (McCurdy, 2000; McCurdy et al., 2000). Indeed, most recent exposure models are
9 designed (or have been re-designed) to obtain such information from CHAD which incorporates
10 24-h time/activity data developed from numerous surveys. The surveys include probability -
11 based recall studies conducted by Environmental Protection Agency and the California Air
12 Resources Board, as well as real-time diary studies conducted in individual U.S. metropolitan
13 areas using both probability-based and volunteer subject panels. All ages of both genders are
14 represented in CHAD. The data for each subject consist of one or more days of sequential
15 activities, in which each activity is defined by start time, duration, activity type (140 categories),
16 and microenvironment classification (110 categories). Activities vary from one min to one h in
17 duration, with longer activities being subdivided into clock-hour durations to facilitate exposure
18 modeling. A distribution of values for the ratio of oxygen uptake rate to body mass (referred to
19 as metabolic equivalents or METs) is provided for each activity type listed in CHAD. The forms
20 and parameters of these distributions were determined through an extensive review of the
21 exercise and nutrition literature. The primary source of distributional data was Ainsworth et al.
22 (1993), a compendium developed specifically to facilitate the coding of physical activities and to
23 promote comparability across studies.
24
25 AX3.7.4.4 Characterization of Inhalation Intake and Uptake
26 Use of the information in CHAD provides a rational way for incorporating realistic
27 intakes into exposure models by linking inhalation rates to activity information. As mentioned
28 earlier, each cohort of the pNEM-type models, or each (virtual or actual) individual of the
29 SHEDS, MENTOR, APEX, and HAPEM4 models, is assigned an exposure event sequence
30 derived from activity diary data. Each exposure event is typically defined by a start time, a
31 duration, assignments to a geographic location and microenvironment, and an indication of
32 activity level. The most recent versions of the above models have defined activity levels using
March 2008 AX3-112 DRAFT-DO NOT CITE OR QUOTE
-------
1 the activity classification coding scheme incorporated into CHAD. A probabilistic module
2 within these models converts the activity classification code of each exposure event to an energy
3 expenditure rate, which in turn is converted into an estimate of oxygen uptake rate. The oxygen
4 uptake rate is then converted into an estimate of total ventilation rate (VE), expressed in liters
5 min"1. Johnson (2001) reviewed briefly the physiological principles incorporated into the
6 algorithms used in pNEM to convert each activity classification code to an oxygen uptake rate
7 and describes the additional steps required to convert oxygen uptake to VE.
8 McCurdy (1997a,b, 2000) has recommended that the ventilation rate should be estimated
9 as a function of energy expenditure rate. The energy expended by an individual during a
10 particular activity can be expressed as EE = (MET)(RMR) in which EE is the average energy
11 expenditure rate (kcal min"1) during the activity and RMR is the resting metabolic rate of the
12 individual expressed in terms of number of energy units expended per unit of time (kcal min"1).
13 MET (the metabolic equivalent of tasks) is a ratio specific to the activity and is dimensionless. If
14 RMR is specified for an individual, then the above equation requires only an activity-specific
15 estimate of MET to produce an estimate of the energy expenditure rate for a given activity.
16 McCurdy et al. (2000) developed distributions of MET for the activity classifications appearing
17 in the CHAD database.
18 Finally, in order to relate intake to dose delivered to the lungs, it is important to take into
19 account the processes affecting uptake following inhalation intake of NO2, in a biologically
20 based dosimetry modeling framework. As a reactive gas, NO2 participates in transformation
21 reactions in the lung epithelial lining fluid, and products of these reactions are thought to be
22 responsible for toxic effects (Postlethwait et., 1991), although kinetic modeling of these reactions
23 has not been performed. Dosimetry models indicate that deposition varies spatially within the
24 lung and that this spatial variation is dependent on ventilation rate (Miller et al., 1982; Overton
25 and Graham, 1995). Controlled exposure studies found that fractional uptake of NO2 increases
26 with exercises and ventilation rate (e.g., Bauer et al., 1986), making activities with high MET
27 values important for quantifying total NO2 exposure. Further discussion of NO2 dosimetry
28 modeling is provided in Section 4.2.
29
March 2008 AX3-113 DRAFT-DO NOT CITE OR QUOTE
-------
1 AX3.7.5 Concluding Comments
2 An issue that should be mentioned in closing is that of evaluating comprehensive
3 prognostic exposure modeling studies, for either individuals or populations, with field data.
4 Although databases that would be adequate for performing a comprehensive evaluation are not
5 expected to be available any time soon, there have been a number of studies, reviewed in earlier
6 sections of this Chapter, that can be used to start building the necessary information base. Some
7 of these studies report field observations of personal, indoor, and outdoor ozone levels and have
8 also developed simple semi-empirical personal exposure models that were parameterized using
9 the observational data and regression techniques.
10 In conclusion, though existing inhalation exposure modeling systems have evolved
11 considerably in recent years, limitations of available modeling methods and data, in relation to
12 potential NO2 studies that include the following, should be taken into account and be addressed
13 by future research efforts:
14 • Ambient photochemical modeling systems are not optimized for estimating NO2 at a
15 local scale.
16 • Subgrid scale modeling (LES, RANS, DNS) is needed to properly characterize effects of
17 nonhomogeneous mixing (i.e., of spatial subgrid variability) on fast nonlinear chemical
18 transformations; the outcomes of this characterization then should be incorporated in
19 simpler models, appropriate for use in conjunction with exposure modeling systems.
20 • Microenvironmental modeling efforts need to balance mechanistic detail and usability by
21 developing:
22 — A simplified but adequate indoor chemistry mechanism for NO2 and related
23 oxidants,
24 — Databases of realistic distributions of indoor NO2 source magnitudes and
25 activities,
26 — Flexible, multi-zonal models of indoor residential and occupational
27 microenvironments.
28 Existing prognostic modeling systems for inhalation exposure can in principle be directly
29 applied to, or adapted for, NO2 studies; APEX, SHEDS, and MENTOR-1A are candidates.
30 However, such applications would be constrained by data limitations such as ambient
March 2008 AX3-114 DRAFT-DO NOT CITE OR QUOTE
-------
1 characterization at the local scale and by lack of quantitative information for indoor sources and
2 sinks.
3
4
5 AX3.8 EXPOSURE ERROR
6 Discussions in this section focus on the errors associated with exposure assessments and,
7 in particular, with those that may be associated with using ambient NC>2 as a surrogate of
8 personal NO2 exposure in epidemological time series studies. As shown in Figure AX3.8-1,
9 exposure error is one of the errors associated with epidemological studies linking pollutant
10 concentrations in ambient air and human health responses. How exposure errors influence the
11 epidemological findings depend upon the design of the epidemological study. In this section, the
12 exposure errors will be discussed in the context of two common environmental epidemological
13 study designs, time-series studies and chronic studies, in which central site NC>2 concentrations
14 are used as surrogates of personal exposure.
15 In a broader sense, NO2 is an indicator of a chemical mixture, which might be the real
16 agent(s) leading to epidemological findings. Ambient, indoor or personal NC>2 might indicate
17 different chemical mixtures because of differences in the infiltration efficiency or chemical
18 reactivity of other NOY species or in the composition of nearby sources. When using ambient
19 NC>2 as a surrogate of personal exposure, issues of confounding and surrogate are raised.
20 Confounding issues have been discussed in Section AX3.6. A brief summary of the confounding
21 issues and a brief discussion of the surrogate issues will be provided in this section.
22 Usually when discussing errors in the context of exposure assessments, errors resulting
23 from limitations of analytical capabilities of monitoring instruments are lumped together with
24 those caused by environmental factors such as spatial heterogeneity in ambient concentrations,
25 the lack of identification of indoor and neighborhood sources etc. In certain instances these
26 different errors may be linked.
March 2008 AX3-115 DRAFT-DO NOT CITE OR QUOTE
-------
EPI
__Exposure
Assignment
I
Error >L
Uncertainty t?): Measurement
Variability <-): Spatial Distribution)
f Error "\
I Numerical: Convergence!
\ModeJ1 Selection ^/
EPI
Model
Error
? . Measurement, Misclass.
*• -" Levoi ditf. from person to
person
Health
Outcome
Different/able?
Error
7 ; Case Ascertainment
^: Susceptibility
'
Study
Design
Figure AX3.8-1.
Errors associated with components of the continuum from ambient
air pollution to adverse health outcome.
1 Measurements of NO2 are subject to artifacts both at the ambient level and at the personal
2 level. A discussion of the errors associated with ambient monitors is given in Section 2.8, and
3 one for errors associated with personal monitors is given in Section AX3.4. As noted earlier,
4 measurements of ambient NC>2 are subject to variable interference caused by other NOy
5 compounds, in particular PANs, organic nitrates, particulate nitrate and HNC>2 and HNOs. The
6 latter is taken up on inlet walls to varying degrees and likely causes variable (positive) artifacts
7 in NC>2 measurements.
8 Personal monitors are subject to interference by SC>2 and HONO and it is not clear to
9 what extent they are affected by interference by the NOy species interfering with the ambient
10 monitors. In addition, personal monitors generally require longer sampling times (typically from
March 2008
AX3-116
DRAFT-DO NOT CITE OR QUOTE
-------
1 about a day to two weeks) and so will not be able to identify peak exposures occurring on time
2 scales of a few hours or less. As noted by Pilotto et al. (1997) these exposures would have been
3 averaged out and associated health outcomes would not be properly attributed by monitors
4 requiring longer sampling times. Often personal concentrations may either be below or not very
5 much above detection limits for the most commonly used personal samplers (see Table
6 AX3.3-2). Thus, associations between ambient and personal concentrations could be weakened
7 between ambient and personal concentrations of a given pollutant. In studies of multiple
8 pollutants, personal concentrations of one pollutant may be more strongly associated with
9 ambient concentrations of another pollutant if the measurements of the latter at the personal level
10 are subject to larger analytical errors than are measurements of the former at the personal level.
11 Spatial heterogeneity in ambient concentrations helps determine how well concentrations
12 measured at ambient monitoring sites reflect exposures at the community and personal levels.
13 Correlations between different pairs of monitoring sites are not sufficient for characterizing
14 spatial variability, as there may be significant differences in concentrations among monitoring
15 sites. This point has been demonstrated in Chapter 3 the latest AQCD for PM (U.S.
16 Environmental Protection Agency, 2004) and Chapter 3 the latest AQCD for ozone and other
17 photochemical oxidants (U.S. Environmental Protection Agency, 2006a). As described earlier in
18 Section AX3.2, concentrations of NO2 are highly variable across the urban areas examined and
19 will result in exposure characterization errors at least as significant as, if not larger, than those
20 for O3 and PM2.5. The problem is exacerbated for NO2 because of the sparseness of NOX
21 monitors, compared to monitors for PM and Os. Thus, the use of central site monitors may be
22 more problematic for NO2 than for PM2 5 (e.g.). As a result, little relation might be found
23 between ambient central site monitors and personal exposures and/or indoor concentrations and
24 stronger associations might be found between cross pollutants at the ambient and personal levels.
25 In this case, it may be necessary to supplement existing ambient measurements to derive ambient
26 concentrations that are consistent with those of other pollutants, e.g., by the use of supplemental
27 'outdoor' monitors. Additional complexity arises if horizontal spatial gradients are large enough,
28 as might happen in going from urban to rural environments, as the lowest values measured might
29 be beneath quantification limits or even beneath detection. Small scale horizontal variability
30 especially as found near roads could be large.
March 2008 AX3-117 DRAFT-DO NOT CITE OR QUOTE
-------
1 As noted earlier in Section AX3.2, variability in the vertical must be considered in
2 addition to horizontal variability. NO2 emitted at or near ground level exhibits strong vertical
3 gradients. Restrepo et al. (2004) found that NO2 measured at 15 m above the surface was a
4 factor of higher than measurements of NO2 at 4 m. Monitors placed at heights such as these will
5 be found in many inner urban areas.
6 In the framework developed by Zeger (2000) for analyzing errors in time-series
7 epidemological studies associated with exposure measurement errors, exposure errors could be
8 classified into three components: (1) the difference between true ambient concentration and the
9 measured ambient concentration, (2) the difference between the measured ambient concentration
10 and the community ambient exposure, and (3) the difference between the community ambient
11 exposure and the personal ambient exposure. These differences mentioned above are determined
12 by (1) the reliability of measurement techniques, (2) the spatial and temporal variation of
13 ambient NO2 concentrations, and (3) personal activity and microenvironment characteristics.
14 In the context of chronic epidemological studies, the issue of misclassification also arises.
15 Personal exposure is composed of exposures to both ambient sources and nonambient sources. If
16 total personal NO2 exposure is assumed to be responsible for the observed health outcomes, the
17 use of ambient concentration as a surrogate for personal exposure could lead to misclassification
18 and bias the epidemological findings. The degree of the misclassification also depends on the
19 spatial and temporal variation of ambient NO2, personal activities and microenvironment
20 characteristics.
21 In the Danish children exposure study, front door NO2 as well as personal NO2
22 concentrations were measured (Raaschou-Nielsen et al., 1997). To evaluate the extent of
23 misclassification using outdoor NO2 as an indicator of personal exposure, Raaschou-Nielsen
24 et al. (1997) reported that both the sensitivity (the proportion of correctly classified highly
25 exposure) and the specificity (the proportion of correctly classified low exposure) were 81% in
26 Copenhagen and 74% in the rural areas. Similar results were reported by Lee, et al., (2004).
27 Exposure measurement errors could also be evaluated by comparing the within subject
28 and between subject variations of individual exposures. The higher the ratio of within variance
29 and between variance, the more the true exposure-effect relationship is biased (Armstrong et al.,
30 1992). During the Los Angeles NO2 exposure study, Spengler et al. (1994) reported that the
31 within personal variation was 61.2 |ig/m3 and the variation between personal exposure was
March 2008 AX3-118 DRAFT-DO NOT CITE OR QUOTE
-------
1 608.2 |ig/m3. Aim et al. (1998) reported that within personal variation explained 59% of the total
2 personal exposure variation and 41% of the total variation was accounted by between-subject
3 variation.
4 Simply speaking, two parameters could be used to evaluate the feasibility of using
5 ambient NO2 concentrations as a surrogate for personal exposure: the correlation coefficient
6 between personal exposure and ambient concentrations (especially in the context of longitudinal
7 design and daily-averaged design), and the difference between personal exposure and ambient
8 concentration. Extensive discussions of this issue have been provided in Section AX3.5, such
9 discussions are not repeated here and only general conclusions will be provided. The correlation
10 between personal exposure and ambient concentrations range from moderate to good. Personal
11 exposure concentrations are generally lower than ambient concentrations for homes with no
12 indoor or local sources but higher than ambient concentration for homes with indoor or local
13 sources.
14 In a broader context, NO2 serves as an indicator of a pollutant mixture whose components
15 have different physical and chemical properties that may be the real agent(s) causing the adverse
16 health effects. The components of the mixture are either primary or secondary, i.e., they either
17 come from direct emissions or form through atmospheric chemical reactions. When the ambient
18 mixture infiltrates into microenvironments, some components are lost due to absorption and
19 chemical reaction, while some new components are formed through chemical reactions in indoor
20 air. At the same time, indoor primary sources could add more NO2 along with other pollutants in
21 the indoor environments. When evaluating the question of whether ambient NO2 is the agent
22 causing the observed adverse health effects, the two issues of confounding and surrogacy are
23 raised.
24 The definition and discussion of the confounding issue from the perspective of exposure
25 analysis could be found in Section AX3.6. In Section AX3.6, the following five questions were
26 evaluated (the five arrows in Figure AX3.8-2): (1) Are ambient copollutant concentrations
27 significantly associated with ambient NO2? (2) Are personal exposures to copollutants
28 significantly associated with personal exposures to NO2? (3) Are ambient pollutant
29 concentrations associated with their respective personal exposures? (4) Are ambient copollutants
30 surrogates for personal exposure to NO2? (5) Is ambient NO2 a surrogate for personal exposure
31 to copollutants? Based on the fact that NO2 is correlated with other copollutants at both ambient
March 2008 AX3-119 DRAFT-DO NOT CITE OR QUOTE
-------
1 level and personal exposure levels and that cross-level correlations were also observed, we
2 concluded that caution should be exercised when dealing with the observed NO2 health effect
3 and a more comprehensive analysis should be performed in conjunction with other components
4 of the risk assessment.
5 Another issue raised is the surrogate issue. There are different meanings associated with,
6 to use the word "surrogate". In summary, there are three scenarios involving the concept of a
7 surrogate and each one is associated with a question: (1) At ambient level, is ambient NC>2 a
8 good surrogate (tracer) for some ambient chemical or chemical mixture? (2) At personal
9 exposure levels, is personal NO2 exposure a good surrogate (tracer) for some chemical or
10 chemical mixture of personal exposure? and (3) At health effect levels, is NO2 a good surrogate
11 for some chemical or chemical mixture causing an adverse health outcome? The first two
12 questions could be sufficiently answered by various source apportionment approaches to
13 evaluate the co-variation of NC>2 with other pollutants. The third question is evaluated in Figure
14 AX3.8-2 with a systematic approach considering biological plausibility and exposure
15 assessment.
Yes -^PersonalNO, Exposure
correlated witti ambient
c&pofatants?
Figure AX3.8-2.
A systematic approach to evaluate whether NOi itself is causing the
observed adverse health outcome or NOi is acting as a surrogate for
other pollutants.
March 2008
AX3-120
DRAFT-DO NOT CITE OR QUOTE
-------
TABLE AX3.2-1. SUMMARY OF PERCENTILES OF NO2 DATA POOLED ACROSS MONITORING SITES (2003-
o
to
o
o
oo
Pooled Group/ Avg
Time
1-h Max Concentrations
Monitors in CMSAs
Monitors not in
CMSAs
1-h Avg. Concentrations
Monitors in CMSAs
Monitors not in
CMSAs
Number of
Values Mean
288008 0.029
460913 0.008
6163408 0.015
460913 0.008
zuu3j ^*jr>^i^r> iK/\injr>a AKJL ir> rrwi
Percentiles
1 5 10 25 30 50 70 75 90 95 99 Max
0.003 0.007 0.010 0.017 0.019 0.027 0.036 0.038 0.048 0.055 0.072 0.201
0.001 0.001 0.001 0.002 0.003 0.005 0.009 0.010 0.019 0.026 0.040 0.189
0.001 0.003 0.003 0.006 0.007 0.012 0.019 0.022 0.033 0.040 0.053 0.201
0.001 0.001 0.001 0.002 0.003 0.005 0.009 0.010 0.019 0.026 0.040 0.189
Daily 24-h Avg. Concentrations
X
OJ
to
Monitors in CMSAs
riusAc
282810 0.015
20635 0.008
0.002 0.003 0.005 0.008 0.009 0.012 0.019 0.021 0.028 0.034 0.045 0.129
0.001 0.001 0.001 0.003 0.003 0.006 0.010 0.011 0.017 0.021 0.030 0.081
2-week Avg. Concentrations
Monitors in CMSAs 21779
Monitors not in
CMSAs 1588
0.015 0.003 0.005 0.006 0.009 0.010 0.014 0.019 0.020 0.026 0.031 0.038 0.076
0.008 0.001 0.001 0.001 0.003 0.003 0.007 0.009 0.012 0.016 0.020 0.030 0.039
H
6
o
2|
0
H
O
H
W
O
o
n
Yearly Avg. Concentrations
Monitors in CMSAs 758
CMSAs 51
3-yr Avg. Concentrations
Monitors in CMSAs 247
Monitors not in
CMSAs 15
0.015 0.004 0.006 0.007 0.011 0.012 0.015 0.018 0.019 0.025 0.028 0.033 0.037
0.008 0.001 0.001 0.002 0.003 0.005 0.009 0.012 0.012 0.015 0.016 0.017 0.017
0.015 0.004 0.006 0.007 0.011 0.012 0.015 0.018 0.019 0.025 0.028 0.032 0.033
0.008 0.001 0.001 0.002 0.003 0.006 0.008 0.012 0.012 0.014 0.016 0.016 0.016
-------
TABLE AX3.2-2. SPATIAL VARIABILITY OF NO2 IN SELECTED UNITED STATES
URBAN AREAS
Mean 1-h
Concentration(ppb) r P90 (ppb)
New York, NY
(5)
Atlanta, GA
(5)
Chicago, IL
(7)
Houston, TX
(7)
Los Angeles, CA
(14)
Riverside, CA
(9)
29
(25-37) 0.77-0.90 7-19
** 0.22-0.89 7-24
(5-16)
(6-230) -°05-°-83 10-39
(T-'lS) °31-°-8° 6-2°
25
(14-33) 0.01-0.90 8-32
(, 2* 0.03-0.84 10-40
COD
0.08-0.23
0.15-0.59
0.13-0.66
0.13-0.47
0.08-0.51
0.14-0.70
TABLE AX3.2-3. NOX AND NOY CONCENTRATIONS AT REGIONAL
BACKGROUND SITES IN THE EASTERN UNITED STATES. CONCENTRATIONS
ARE GIVEN IN PPB
NO
Winter
Summer
NOX
Winter
Summer
NOY
Winter
Summer
Shenandoah NP, VA
0.39-2.2 l
0.12-0.28
—
—
—
—
2.7-8.6
2.3-5.7
Harvard Forest, MA
—
—
—
1-15
0.4-1.2
—
4.4 2
2.7 2
1 Ranges represent lo limits.
2 Values represent medians.
March 2008 AX3-122 DRAFT-DO NOT CITE OR QUOTE
-------
TABLE AX3.2-4. RANGE OF PEARSON CORRELATION COEFFICIENTS
BETWEEN NO2 AND O3, CO AND PM2.5
Monitoring Sites in Copollutant
Selected Areas
Los Angeles, CA
Riverside, CA
03
-0.59 to 0.19
-0.26 to 0.28
CO
0.11 to 0.83
0.15 to 0.65
PM25
0.45 to 0.56
Chicago, IL -0.20 to-0.13 -0.10 to 0.53 0.21 to 0.49
Washington, DC — — —
New York City, NY — — —
March 2008 AX3-123 DRAFT-DO NOT CITE OR QUOTE
-------
TABLE AX3.3-1. PASSIVE SAMPLERS USED IN NO2 MEASUREMENTS
o
to
o
o
oo
X
OJ
to
o
H
6
o
0
H
O
HH
H
W
Passive Sampler
Palmes tube
Gradko sampler
Passam Short
sampler Long
Analyst™
Yanagisawa badge
Ogawa sampler
IVL sampler
Willems badge
Radiello®
EMD sampler
Dimension
(diffusion length x
cross-sectional area) Absorbent
7.1cm x 0.71cm2 Triethanolamine
7.1cm x 0.93cm2 Triethanolamine
0.74cm x 0.75cm2 Triethanolamine
2.54cm x 3.27cm2 Active charcoal
1.0cm x 20cm2 Triethanolamine
0.6cm x 0.79cm2 Triethanolamine
Potassium iodide
1.0cm x 3.14cm2 & sodium arsenite
Triethanolamine-
0 . 6cm x 5 . 3 1 cm2 acetone
1 . 8cm x 2 . Ocm2 Triethanolamine
N.A. Triethanolamine
Analytical
Method
Spectrophotometry
Spectrophotometry
Spectrophotometry
Gas
chromatography
Spectrophotometry
Spectrophotometry
Spectrophotometry
Spectrophotometry
Spectrophotometry
Ion
chromatography
Sampling Rate
Manufacturer Experiment Reference
N.A. 0.92 cnrYmin Palmes et al. (1976)
Plaisance et al.
(2004)
1.2 cnrYmin 1.212 cnrYmin Gradko (2007)
15.5 cnrYmin N.A.
0.854 cnrYmin 0.833 cnrYmin Passam (2007)
De Santis et al.
N.A. 12.3 cnrYmin (2002)
Yanagisawa and
N.A. N.R. Nishimura (1982)
Ogawa & Company
(1998a) Gerboles
N.A. 16.2 cnrYmin et al. (2006a)
Perm and Svanberg
N.A. 29 cnrYmin (1998)
Hagenbjork-
Gustafsson et al.
N.A. 46 cnrYmin (2002)
75 cnrYmin N.R. Radiello® (2006)
Piechocki-Minguy
N.A. 53.4 cnrYmin etal. (2006)
*N.A.: not available; N.R.: notreported.
O
-------
O
TABLE AX3.3-2. THE PERFORMANCE OF SAMPLER/SAMPLING METHOD FOR NO2 MEASUREMENTS
IN THE AIR
to
O
O
oo
X
OJ
1
to
O
H
6
O
0
H
O
H
W
Type Sampler
Active Impinger method
sampling
Chemiluminescence
Personal monitor
Passive Palmes tube
sampling
Gradko sampler
Passam Short
sampler
Long
Analyst™
Yanagisawa badge
Ogawa sampler
IVL sampler
Willems badge
Radiello®
EMD sampler
N.R.: not reported.
Optimal Duration of
Sampling
2-24 h
Continuous
Real-time
1-4 wks
2-4 wks
8-48 h
1-4 wks
1-3 mos
1-14 days
24-168 h
1 mo +
2-8 h& 1-7 days
1-24 h& 1-7 days
1-24 h
Concentration
Range
10 - 400 ppb
0.5 - 1000 ppb
0.1 -50ppm
10 - 100 ppb
1.0 -10,000 ppb
5 - 240 ug/m3
1 - 200 ug/m3
24 - 1,237 ug/m3
N.R.
0 - 3,600 ppb
0.1-400 ug/m3
2.0 - 150 ug/m3
1.0 -496 ppb
N.R.
Detection Limit
0.2 ppb
0.05 ppb
0.1 ppm
10 ppb
0.5 ppb
2-5 ug/m3
0.64 ug/m3
100 ug/m3
3.0 ppb
2.3 ppb
0.1 ug/m3
2 ug/m3
1.0 ppb
11 ug/m3
RSD < 5%
Accuracy ±
Precision ±
Uncertainty
Uncertainty
Accuracy ±
3%
RSD - 4%
Uncertainty
Uncertainty
Uncertainty
Comment
5%
5% above 5 ppb
-27% at 80 ug/m3
- 25% at 20-40 ug/m3
5%; Precision within
- 24%; RSD 22%
- 12%
-28%
O
-------
TABLE AX3.4-1. NO2 CONCENTRATIONS (PPB) IN HOMES IN LATROBE VALLEY,
VICTORIA, AUSTRALIA
Living Room
No source
Gas stove only
Gas heater only
Smoking only
Multiple sources
Mean ppb
3.77
6.70
6.86
6.02
14.50
Min ppb
<0.37
1.57
2.20
0.94
2.25
Max ppb
9.27
18.32
18.06
14.61
114.66
Mean ppb
3.82
8.01
7.33
6.60
10.73
Kitchen
Min ppb
<0.37
2.62
2.88
1.83
2.62
Max ppb
8.17
24.14
26.23
16.44
128.80
Source: Garrettetal. (1999).
TABLE AX3.4-2. NO2
CONCENTRATIONS (PPB) IN
No Gas Stove Used in Monitoring Period
Secondary
Heating
Source
None
Gas space
heater
Wood
burning
source
Kerosene
heater
GSH +
Wood
GSH + KH
Wood + KH
GSH +
Wood + KH
N
1018
6
200
159
3
0
73
0
10th
1.7
0.1
1.8
3.3
12.6
—
1.9
—
25th
3.5
9.2
3.6
7.1
12.6
—
8.2
—
Median
6.3
15.3
5.9
18.9
80.6
—
16.4
..
75th
12.3
68
12.2
42.7
81.9
—
35.2
—
90th
28.2
69.6
28.2
88.3
81.9
—
66.8
—
HOMES IN CONNECTICUT
Yes Gas Stove Used in Monitoring Period
N
564
6
78
14
5
1
5
1
10th
8.4
19.5
6
0
36.2
n/a
8.9
n/a
25th
14.5
34.6
9.5
9.6
44.8
n/a
12.7
n/a
Median
22.7
36.6
16.7
17.2
57.1
147.7
17.3
107.8
75th
33.8
54.8
31.4
33.6
114.
2
n/a
23.5
n/a
90th
48.1
147.2
58.6
46.1
156.6
n/a
72.9
n/a
Source: Triche et al. (2005).
March 2008
AX3-126
DRAFT-DO NOT CITE OR QUOTE
-------
TABLE AX3.4-3. NO2 CONCENTRATIONS NEAR INDOOR SOURCES -
SHORT-TERM AVERAGES
Average
Concentration
(PPb)
Peak Concentration
(PPb)
Comment
Reference
191 kitchen
195 living room
184 bedroom
400 kitchen,
living room,
bedroom
90 (low setting)
350 (med setting)
360 (high setting)
N/R
N/R
180 to 650
375 kitchen
401 living room
421 bedroom
673 bedroom
N/R1
1000
1500
N/R
Cooked full meal with
use of gas stove and
range for 2 h, 20 min;
avg cone, is time-
weighted over 7 h.
Automatic oven
cleaning of gas stove.
Avgs are over the entire
cycle.
Natural gas unvented
fireplace,2 2-h-time-
weighted avg in main
living area of house
(177m3).
Room concentration
with kerosene heater
operating for 46 min.
Room concentration
with gas heater
operating for 10 min.
Calculated steady-state
concentration from
specific unvented gas
space heaters operating
in a 1400 ft2 house,
1.0 ach.
Fortmann et al.
(2001)
Fortmann et al.
(2001)
Duttonetal. (2001)
Girman et al.
(1982)
Girman et al.
(1982)
Girman et al.
(1982)
1 N/R = Not Reported.
2 Unvented fireplaces are not permitted in many areas such as California.
Source: Adapted from CARB (2007).
March 2008
AX3-127
DRAFT-DO NOT CITE OR QUOTE
-------
TABLE AX3.4-4. NO2 CONCENTRATIONS NEAR INDOOR SOURCES -
LONG-TERM AVERAGES
Average Concentration
(ppb) Comment Reference
30 to 33
22
6 to 11
55 (Median)
41 (90th %-ile)
80 (90th %-ile)
84 (90th %-ile)
147 (90th %-ile)
52 (90th %-ile)
18 bedrooms
19 living rooms
15 outdoors
Gas stoves with pilot lights.
Gas stoves without pilot lights.
Electric ranges. Study conducted in 517
homes in Boston, values represent 2-wk
avgs.
Gas space heaters.
No indoor combustion source.
Fireplaces.
Kerosene heater.
Gas space heaters.
Wood stove.
All values represent 2-wk avgs in living
rooms.
Almost all homes had gas stoves. Values
represent 2-wk avgs.
Lee etal. (1998)
Triche et al.
(2005)
Zipprich et al.
(2002)
March 2008
AX3-128
DRAFT-DO NOT CITE OR QUOTE
-------
TABLE AX3.5-1. SUMMARY OF REGRESSION MODELS OF PERSONAL EXPOSURE TO AMBIENT/OUTDOOR NO2
O
to
O
O
oo
X
OJ
to
VO
H
6
o
O
H
O
HH
H
W
Study Location
Rojas-Bracho et al. Santiago, urban
(2002)
Aim et al. (1998) Helsinki, downtown +
Monn et al.
suburban
(1998) Four urban + two rural
+ two alpine
Levy et al. (1998a) 15 cities in 18
Spengler et
(1994)
S0rensen et
(2005)
Sarnat et al.
Sarnat et al.
Sarnat et al.
countries
al. Los Angeles Basin
al. Copenhagen, urban
(2001) Baltimore
(2005) Boston
(2006) Steubenville
Season
Winter
Winter +
Spring
All
Winter
All
All
(>8 °C)
(<8 °C)
All
Summer
Winter
Summer
Winter
Summer
Fall
Model type
Personal vs. outdoor
Personal vs. central
Personal vs. outdoor
Personal (all subjects) vs. outdoor
Personal (no smokers and gas
cooking) vs. outdoor
Personal vs. outdoor
Personal vs. outdoor
Personal vs. outdoor
Personal vs. outdoor
Personal vs. outdoor
Personal vs. central
Personal vs. central
Personal vs. central
Personal vs. central
Personal vs. central
Personal vs. central
Personal vs. central
Slope (SE) Intercept / ppb R2
0.33 (0.05)
0.3
0.4
0.45
0.38
0.49
0.56
0.60 (0.07)
0.68 (0.09)
0.32(0.13)
0.56 (0.09)
0.04*
-0.05*
0.19
-0.03*
0.25 (0.06)
0.49 (0.05)
7.2 0.27
5.0 0.37
4.7 0.86
7.2 0.33
7.2 0.27
14.5 —
15.8 0.51
— —
— —
— —
— —
9.5 —
18.2 —
— —
— —
— 0.14
— 0.43
*Not significant at the 5% level.
O
-------
TABLE AX3.5-2. AVERAGE AMBIENT AND NONAMBIENT CONTRIBUTIONS TO POPULATION EXPOSURE
o
to
o
o
oo
OJ
i
o
O
H
6
o
0
H
O
H
W
O
o
o
Study
Rojas-B radio
et al. (2002)
Aim et al.
(1998)
Monn et al.
(1998)
Levy et al.
(1998a)
Spengler et al.
(1994)
** Not reported.
Slope
Model Type (SE)
Personal vs. outdoor 0.33
(0.05)
Personal vs. central 0.3
Personal vs. outdoor 0.4
Personal (all subjects) 0.45
vs. outdoor
Personal (no smokers 0.38
and gas cooking) vs.
outdoor
Personal vs. outdoor 0.49
Personal vs. outdoor 0.56
Mean of Percent Percent
Personal Total Mean Ambient Ambient Nonambient
Intercept / Exposure / Contribution / Contribution Contribution
ppb ppb Ppb °/° 0//0
7.2 36.4 7.2 19.8 80.2
5.0 — 5.0 — —
4.7 — 4.7 — —
7.2 14.1 7.2 51.1 48.9
7.2 — 7.2 — —
14.5 28.8 14.5 50.3 49.7
15.8 37.6 15.8 42.0 58.0
-------
O
TABLE AX3.5-3. THE ASSOCIATION BETWEEN PERSONAL EXPOSURES AND
AMBIENT CONCENTRATIONS
1— '
to
o
o
oo
X
OJ
1
^
o
H
6
o
2;
o
H
O
HH
H
W
O
O
O
Study
Linn et al.
(1996)
Kramer et al.
(2000)
Rojas-B radio
et al. (2002)
Raaschou-
Nielsen et al.
(1997)
Aim et al.
(1998)
Monn et al.
(1998)
Study Design
Children, Southern California, 24 h averaged, one wk
consecutive measurement for each season (fall, winter, and
spring 1992-1994) for each child.
Children, West Germany, two one-wk averaged
measurements for each child each in March and Sept 1996
Children, Santiago, 24 h averaged sample for five
consecutive days for each child, winters of 1998 and 1999
Children, Copenhagen and rural areas, one-wk averaged, 2
measurements for each child in each month (Oct 1994,
April, May, and June 1995)
Children, Helsinki, one-week averaged, 13 wks for each
child in each season (winter and spring 1991)
Adults, Switzerland, eight regions in Swiss (four
urban/suburban, two rural and two alpine regions), one-wk
averaged, one measurement each mo (the first wk of the
mo) for each subject, between Dec 1993 to Dec 1994
Association Variable
Personal vs. central
Personal vs. outdoor
Personal vs. outdoor
Personal vs. outdoor
Personal vs. outdoor
Personal vs. outdoor
Personal vs. outdoor
Personal vs. outdoor
Personal vs. outdoor
Personal vs. outdoor
Personal vs. central
Personal vs. central
Personal vs. outdoor
Personal vs. central
Personal vs. outdoor
Location
pooled
pooled
urban
urban
urban
rural
downtown
suburban
downtown
suburban
downtown
suburban
pooled
pooled
pooled
Season
pooled
pooled
pooled
winter
pooled
pooled
winter
winter
spring
spring
spring
spring
pooled
pooled
pooled
rp, rs, or R2
0.63 (rp)
0.37 (rp)
0.06 (rp)
0.27 (R2)
0.15 (R2)
0.35 (R2)
0.46 (rp)
0.49 (rp)
0.80 (rp)
0.82 (rp)
0.64 (rp)
0.78 (rp)
0.86 (R2)
0.37 (R2)
0.33 (R2)
-------
O
TABLE AX3.5-3 (cont'd). THE ASSOCIATION BETWEEN PERSONAL EXPOSURES AND
AMBIENT CONCENTRATIONS
1— '
to
o
o
oo
X
OJ
1
to
o
fj*
i-rj
H
6
o
o
^^
H
o
i — 1
H
W
O
Study
Levy et al.
(1998a)
Kodama et al.
(2002)
Liard et al.
(1999)
Gauvin et al.
(2001)
Spengler et al.
(1994)
Kousa
etal. (2001)
Study Design
Adults, 18 cities across 15 countries, two-day averaged,
one measurement for each person, all people were
measured on the same winter day in February or March
1996
Junior high school students and their family members,
Tokyo, three-day averaged, samples were simultaneously
collected on Feb 24-26, Jun 2-4, July 13-15, and Oct 14-16
in 1998 and Jan 26-28 in 1999
Adults and Children, Paris, 4-day averaged, three
measurements for each person, during each measurement
session, all subjects were measured at the same time
during May/June 1996
Children, three French metropolitan areas, 48-h averaged,
one measurement for each child, all children in the same
city were measured on the same day. The study occurred
between April-June 1998 in Grenoble, May-June 1998 in
Toulouse, and June-Oct 1998 in Paris.
Probability based population, Los Angeles Basin, 48-h
averaged, one measurement per person in one of the eight
sampling cycles (microenvironmental component of the
study), from May 1987 to May 1988
Probability based population, Helsinki, Basel, and Prague,
48-h averaged, one measurement per person, during 1996
and 1997
Association Variable Location Season
Personal vs.
Personal vs.
Personal vs.
outdoor
outdoor
outdoor
Adults vs. central
Children vs.
Personal vs.
(Grenoble)
Personal vs.
(Toulouse)
Personal vs.
(Paris)
Personal vs.
Personal vs.
central
central
central
central
outdoor
outdoor
urban
urban
urban
urban
urban
urban
urban
urban
pooled
urban
winter
summer
winter
summer
summer
pooled
pooled
pooled
pooled
pooled
rp, rs, or R2
0.57 (rs)
0.24 (rp)
0.08 (rp)
0.41 (R2)
0.17(R2)
0.01 (R2)
0.04 (R2)
0.02 (R2)
0.48 (R2)
0.40 (R2)
O
-------
TABLE AX3.5-3 (cont'd). THE ASSOCIATION BETWEEN PERSONAL EXPOSURES AND
AMBIENT CONCENTRATIONS
to
o
o
oo
>
OJ
1
OJ
>
H
6
O
0
H
O
H
W
Study
Linaker et al.
(2000)
Lai et al.
(2004)
Kim et al.
(2006)
Sarnat et al.
(2005)
Sarnat et al.
(2006)
Study Design
Asthmatic children, Southampton, one-wk averaged,
13 mos for each child, until Dec 1995
Adults, Oxford, 48-h averaged, once per person, between
Dec 1998 and Feb 2000
Coronary artery adults, Toronto, 24-h averaged, one day
a wk for 10 wks for each person, from Aug 1999 to Nov
2001
Seniors and schoolchildren, Boston, 24-h averaged, 12
consecutive days in each of the 1 or 2 seasons, summer of
1999 and winter of 2000
Seniors, Steubenville, 24-h averaged, the same two
consecutive days each wk for 23 wks, summer and fall of
2000
Association Variable
Personal vs. outdoor
(Overall
measurements across
children and time)
Personal vs. outdoor
(subject-wise)
Personal vs. outdoor
Personal vs. central
(ambient)
Personal vs. central
(subject wise)
Personal vs. central
Location
pooled,
urban, no
major indoor
sources
By person
urban
urban
urban
urban
Season
pooled
pooled
pooled
pooled
summer
winter
summer
fall
rp, rs, or R2
Not
significant
-0.77 to 0.68
and median
-0.02(rp)
0.41 (rp)
0.57 (rs)
-0.25 to 0.5
(rs) with a
median of
0.3*
-0.5 to 0.9
(rs) with a
median of
0.4*
0.14(R2)
0.43 (R2)
* Values were estimated from figures in the original paper.
O
-------
TABLE AX3.5-4. INDOOR/OUTDOOR RATIO AND THE INDOOR VS. OUTDOOR REGRESSION SLOPE
O
to
O
O
oo
Study
Description
Season
Regression Format or
Ratio
Indoor
Characteristics
Slope/Ratio/
Fjnf
Comments
>
X
Mosqueron et al.
(2002)
Lee etal. (1999)
48 h residential indoor,
workplace, outdoor and
personal exposure were
measured for 62 Paris office
workers using Ogawa badges
from Dec 1999 to Sept 2000
Overall study Residential indoor vs. Cooking
seasons ambient and using gas
cooking
Office indoor vs. ambient None
and floor height
0.26
0.56
The indoor and outdoor air Overall study Indoor vs. outdoor
quality of 14 public places seasons
with mechanical ventilation
systems in Hong Kong; from
Oct 1996 to March 1997; Indoor/outdoor ratio
Teflon bags were used to
collect indoor and outdoor
NO and NO2 during peak h
0.59
0.53-1.03
(mean: 0.75)
The overall R is 0.14,
and ambient NO2 and
indoor cooking
account for 0.07 each.
The overall R2 is 0.24,
partial R2 for ambient
and floor height were
0.18 and 0.06,
respectively.
R2 was 0.59. The
slopes for NO and
NOX were 1.11 and
1.04 respectively.
0.83-2.68 for NO
(mean: 0.99)
0.78-1.68 for NOX
(mean: 0.94)
H
6
O
2
O
H
O
HH
H
W
O
-------
TABLE AX3.5-4 (cont'd). INDOOR/OUTDOOR RATIO AND THE INDOOR VS. OUTDOOR REGRESSION SLOPE
Regression Format or Indoor Slope/Ratio/
Study Description
O
to
O
O
oo
Season
Ratio
Characteristics
Finf
Comments
>
X
H
6
o
2
o
H
O
HH
H
W
Monn et al.
(1997)
Monn et al.
(1997)
Garcia-Algar
et al. (2003)
During the SAPALDIA
(Spain) study, 48-72 h
indoor, outdoor, and
personal NO2 were
measured by Palmes tubes
between the winter of
1994 to the summer of
1995, and between May
and July of 1996
During the SAPALDIA
(Spain) study, 48-72 h
indoor, outdoor, and
personal NO2 were
measured by Palmes tubes
between the winter of
1994 to the summer of
1995, and between May
and July of 1996
Yanagisawa passive filter
badges were used to
measure indoor NO2
concentrations for 7~15
days for 340 homes in
Barcelona, Spain during
1996-1999. Outdoor
NO2 concentrations were
obtained from the fixed
monitoring stations by the
method of CL.
Overall study
seasons
Indoor/outdoor ratio
Overall study
seasons
Indoor/outdoor ratio
With gas-cooking
Without gas
cooking
With gas-cooking
Without gas
cooking
> 1.2
0.4-0.7
> 1.2
0.4-0.7
Overall study
seasons
Indoor/outdoor ratio
0.8-1.0
Including
both homes
with and
without
indoor
sources.
O
-------
TABLE AX3.5-4 (cont'd). INDOOR/OUTDOOR RATIO AND THE INDOOR VS. OUTDOOR REGRESSION SLOPE
o
to
o
o
oo
X
UJ
1
s
o
£5
H
6
o
0
H
O
HH
H
W
Study Description Season
Lee et al. Two-wk averaged indoor Summer
(1995) (kitchen, living room, and
bedroom) and outdoor NO2
were measured by Palmes
tube for 517 homes from
November 1984 to Oct 1986
in Boston area.
Lee et al. Six-day integrated indoor and Overall
(2002) outdoor concentrations of study
NO2 in two communities in seasons
Southern California were
measured using Yanagisawa
badges for 119 homes in
April and May 1996.
Lee (1997) Indoor and outdoor air Overall
quality at two staff quarters study
in Hong Kong were measured seasons
from January to Feb of 1996
by Chemical Luminescent
method in two staff quarters
in Hong Kong (TSTE, in a
downtown area; and ST in a
suburban area).
Regression Format Indoor
or Ratio Characteristics
Indoor/outdoor Electric stove
ratio homes
Indoor/outdoor With gas range
ratio ± SD
Without gas range
With air
conditioner
Without air
conditioner
Indoor/outdoor Downtown area
ratio (Range)
Suburban area
Slope/Ratio/Finf
0.81 (kitchen)
0.81 (living room)
0.77 (bedroom)
2.27 ±1.88
1.22 ±0.52
1.07 ±0.26
3.03 ±2.01
0.78 (0.70-0.87)
forNO2
0.92(0.77-1.10)
for NO
0.86(0.78-0.95)
forNOx
0.97(0.89-1.03)
forNO2
0.92(0.77-3.14)
for NO
Comments
Homes with gas
stove and gas
stove with pilot
light have an I/O
ratio >1, but the
values were not
reported.
—
—
—
—
—
O
0.86(0.89-1.03)
forNOx
-------
TABLE AX3.5-4 (cont'd). INDOOR/OUTDOOR RATIO AND THE INDOOR VS. OUTDOOR REGRESSION SLOPE
o
to
o
o
oo
>
r*^
X
UJ
i
OJ
o
rr5
H
6
o
0
o
1 — I
H
W
O
O
^^
Study Description
Garrett et al. Four-day averaged indoor
(1999) (bedroom, living room, and
kitchen) and outdoor NO2
was monitored using
Yanagisawa passive samplers
for 80 homes in the Latrobe
Valley, Victoria, Australia, in
March-April 1994, and Jan-
Feb, 1995.
Zota et al. Two-wk integrated NO2 was
(2005) measured in 77 homes within
three Boston public housing
developments (low-income,
urban neighborhoods, where
asthma prevalence is high),
using Palmes tubes. Homes
were sampled between June
2002 and May 2003 for 2-wk
periods with up to three
sampling sessions in each
home.
Yang et al. Daily indoor and outdoor
(2004) NO2 concentrations were
measured for 30 consecutive
days in 28 house in Brisbane
(between April and May in
1999), and for 21 consecutive
days in 37 houses in Seoul
(between June and Aug in
2000) using Yanagisawa
badges.
Season
Overall
study
seasons
Overall
study
seasons
Overall
study
seasons
Regression Format or Indoor
Ratio Characteristics
Indoor/outdoor ratio No major indoor
sources (major
sources were gas
stove, vented gas
heater, and smoking)
Residential indoor vs. —
residential outdoor
Residential indoor vs. Brisbane with
residential outdoor electric range
Brisbane with gas
range
Seoul with gas range
Indoor/outdoor ratio Brisbane
Seoul
Slope/Ratio/
Finf Comments
0.8 The ratio
increased to
1.3, to 1.8 and
to 2.2 for
homes with
one, two, and
three major
indoor sources.
1.21 —
0.65 ±0.18 R2 was 0.70.
0.56 ±0.12 R2 was 0.57.
0.58 ±0.12 R2 was 0.52.
0.82 ±0.41 —
0.88 ±0.32 —
-------
TABLE AX3.5-4 (cont'd). INDOOR/OUTDOOR RATIO AND THE INDOOR VS. OUTDOOR REGRESSION SLOPE
O
to
O
O
oo
>
X
oo
H
6
O
2
O
H
O
HH
H
W
Regression Format or
Indoor
Study
Chao (2001)
Kulkarni and
Patil (2002)
Monn et al.
(1998)
Description
48-h averaged indoor
and outdoor NO, and
NO2 were measured in
ten non-smoking
residential buildings
using Ogawa passive
samplers in the summer
of 1997 in Hong Kong.
48-h averaged indoor
and outdoor NO2 were
measured using passive
filter badge sampler in
the winter (Feb 1996)
and summer of 1996
(April) for 43 residence
in Mumbai.
One-wk averaged
indoor, outdoor, and
personal NO2 were
measured for more than
500 subjects between
Dec 1993 to Dec 1994
for a SAPALDIA study
subpopulation, once per
home.
Season
Overall
study
seasons
Overall
study
seasons
Overall
study
seasons
Ratio
Indoor/outdoor ratio
Residential indoor vs.
residential outdoor
Residential indoor vs.
residential outdoor
Residential indoor vs.
residential outdoor + gas
cooking + smoking +
ventilation
Indoor/outdoor ratio
Characteristics
—
Homes using LPG
Homes using
Kerosene
All homes
Homes without
smokers and gas-
cooking
All homes
All homes
Slope/Ratio/Finf
0.79 ±0.30
(range: 0.75 -
1.36)forNO2
0.98 ±0.19
(range: 0.29 -
1.25) for NO
0.92
0.73
0.47
0.40
0.55
0.7-0.8
Comments
—
R2 was 0.80.
R2 was 0.40.
R2 was 0.37.
R2 was 0.33.
Overall R2 was
0.58, but partial
R2 cannot be
derived.
—
O
-------
TABLE AX3.5-4 (cont'd). INDOOR/OUTDOOR RATIO AND THE INDOOR VS. OUTDOOR REGRESSION SLOPE
Study Description Season
O
to
O
O
oo
Regression Format or
Ratio
Slope/Ratio
Indoor Characteristics / Finf Comments
>
X
VO
H
6
O
2
O
H
O
HH
H
W
O
Levy et al.
(1998a);
Spengler et al.
(1996)
48-h averaged indoor,
outdoor, and personal
exposures to NO2 were
measured in 18 cities in
15 countries around the
world during a 2-day
period in Feb or March
1996.
Overall
study
seasons
Indoor/outdoor ratio
Boston, U.S. 0.6 ± 0.4
Ottawa, Canada 0.5 ± 0.2
Mexico City, Mexico 1.9 ± 1.0
London, UK 0.6 ± 0.4
Watford, UK 0.8 ± 0.4
Geneva, Switzerland 0.8 ± 0.6
Kjeller, Norway 0.7 ± 0.4
Kuopio, Finland 0.5 ± 0.5
Berlin, Germany 0.3 ± 0.2
Erfurt, Germany 0.8 ± 0.7
Homes without gas stove 0.7
Homes with gas stove 1.2
Homes without kerosene 0.85
heater
Homes with kerosene 2.27
heater
Homes without gas space 0.96
heater
Homes with gas space 1.93
heater
Homes without gas water 0.94
heater
Homes with gas water 1.07
heater
Homes without smokers 0.92
present
Homes with smokers 1.16
present
-------
O
to
O
O
oo
TABLE AX3.5-4 (cont'd). INDOOR/OUTDOOR RATIO AND THE INDOOR VS. OUTDOOR REGRESSION SLOPE
Regression Format or Indoor Slope/Ratio/
Study Description Season Ratio Characteristics Finf
Comments
>
X
Spengleretal.
(1994)
Lai et al. (2004)
A Yanagisawa type of
passive sample was used
to measure the 48-h
integrated indoor, outdoor
and personal NO2 levels
from the May of 1987 to
the May of 1988.
48-h averaged personal,
indoor, outdoor and
workplace NO2 levels
were measured by passive
filter badges for 50 adults
in Oxford between 1998
and 2000, once per person.
Overall study Residential indoor vs.
seasons residential outdoor
Overall study Indoor/outdoor ratio
seasons
Gas range with pilot
light
Gas range without
pilot light
Electric stove
All homes
0.49 R2 was 0.44.
0.4 R2 was 0.39.
0.4 R2 was 0.41.
0.9 —
Smoking homes
Non-smoking homes
1.5
1
H
6
o
2
o
H
O
HH
H
W
Note: *Only data that are marked by underline and bold font can be considered as an infiltration factor.
O
-------
TABLE AX3.5-5. NO2 CONCENTRATIONS (PPB) IN DIFFERENT ROOMS
O
to
O
O
oo
>
X
H
6
o
2
o
H
O
HH
H
W
Study
Topp et al.
(2004)
Garrett et al.
(1999)
Cotterill and
Kingham(1997)
Zota et al.
(2005)
Conditions
First visit
Second visit
No identified
indoor sources
Gas stove homes
Gas heater
homes
Smoking homes
Homes with
multiple sources
Gas Stove homes
Electric cooker
homes
Gas cooker home
with single
glazing window
Gas cooker home
with double
glazing window
Overall
Heating season
Non-heating
season
Outdoor
12.4
12.5
4.7
4.7
4.7
4.7
4.7
20.9
20.9
20.9
20.9
19
21
17
Kitchen
—
—
3.8
8.0
7.3
6.6
10.7
35.6
9.9
31.4
39.8
43
50
33
Living Room
7.8
8.0
3.8
6.7
6.9
6.0
14.5
17.3
8.9
16.8
18.3
36
43
26
Bedroom
7.2
7.6
3.0
6.3
5.0
5.7
11.2
11.5
7.3
11.0
12.0
—
Comments
Indoor and outdoor NO2 concentrations for 777 residential
homes in five study areas were measured: Erfurt,
Hamburg, Zerbst, Bitterfeld and Hettstedt during two
visits (from June 1995 to May 1997, and from April 1996
to Sept 1998). In the study, one-week averaged NO2 were
measured by Palmes tube.
Garrett (1999) investigated the levels and sources of NO2
in Australian homes. During the study, four-day averaged
NO2 was monitored using Yanagisawa passive samplers in
80 homes in the Latrobe Valley, Victoria in March-April
1994, and Jan-Feb 1995.
Three consecutive two-week averaged outdoor, kitchen,
living room, and bedroom NO2 were measured using
Palmes tubes in 40 houses in Huddersfield, UK in late
1994. Half the houses were located close to a busy main
road and half on residential roads set back and parallel to
the main road. The sample was split so that half had gas
cookers and half had electric cookers. These subsets were
split again so that half had double glazing and half had
single glazed windows.
The indoor and outdoor NO2 concentrations for low-
income, urban neighborhoods were measured, where
asthma prevalence is high. NO2 was measured in 77
homes within three Boston public housing developments,
using Palmes tubes (two-wk integrated sample) placed in
the kitchen, living room, and outdoors. Air exchange rate
for each home was also measured.
O
-------
TABLE AX3.5-5 (cont'd). NO2 CONCENTRATIONS (PPB) IN DIFFERENT ROOMS
o
1— >
to
o
o
oo
>
!x!
rS
OJ
I
to
Study
Gallelli et al.
(2002)
Linaker
etal. (1996)
Kodama
et al. (2002)
Conditions
Overall study
With vent
Without vent
Overall study
Feb 1998
June 1998
July 1998
Oct 1998
Jan 1999
Outdoor Kitchen
— 24.6
— 18.1
— 30.9
— 27.2
40,31.3 81.8
38,28 33.2
29, 26.7 24.8
40,35 23.5
49, 50 70.9
Living Room
—
—
20.9
73.5
28.8
21.9
24.7
65.8
Bedroom
13.0
—
—
55.2
24
17.4
18.2
50.7
Comments
During the study, one-wk integrated indoor (kitchen and
bedroom) and personal NO2 were measured in Genoa,
Italy, for 89 subjects with Palmes samplers. Study
volunteers included students, workers, and housewives
living in three areas of Genoa differing by street traffic and
industrial plant location.
During the study, one-wk integrated personal, indoor
(kitchen, living room), classroom, and playground NO2
were measured using Palmes tubes for school children in
Southampton.
The first number in outdoor column was the ambient
concentration in the South Area; and the second number is
the ambient concentration in the North Area. During the
study, personal, indoor (kitchen, living room, bedroom and
study room), and outdoor NO2 were measured for 150
junior high school students with Yanagisawa badges in
Tokyo. The investigation was conducted five times
seasonally, 3 days each, from February 1998 to January
1999.
H
6
o
2
o
H
O
HH
H
W
Chao and
Law (2000)
Overall study
37.6
51.9
28.2
26.4 Personal and indoor exposures were monitored with
passive sampler in Hong Kong for 60 subjects. Twelve of
the subjects were selected to conduct more detailed study
to examine the behavioral and microenvironmental effects
on personal exposure to NO2.
O
-------
TABLE AX3.5-6. INDOOR AND OUTDOOR CONTRIBUTIONS TO INDOOR CONCENTRATIONS
o
to
o
o
oo
X
OJ
i
OJ
o
H
6
o
0
H
O
H
W
O
o
o
Study
Mosquero
net al.
(2002)
Yang
etal.
(2004)
Monn
etal.
(1998)
Percent
Mean Indoor Mean Outdoor Outdoor
Condition Slope Intercept Concentration Concentration Contribution
Overall 0.258 — 18.4 31.5 44.2
study
Brisbane, 0.65 0.8 10.3 — 92.4
electric
range
Brisbane, 0.56 3.0 18.3 — 83.5
gas range
Seoul, gas 0.58 4.8 33.4 40.4 85.7
range
Overall 0.47 3.2 11.0 16.2 70.5
study
Homes 0.40 3.2 6.8 16.2 53.1
without
smokers
and gas
cooking
Percent Indoor
Indoor Source
Contribution Strength Comments
55.8 — —
7.6 3.5ppb/h —
16.5 11.5ppb/ —
h
14.3 23.4 ppb/ —
h
29.5 — —
46.9 — Mean indoor
was
estimated
based on the
text
description.
-------
TABLE AX3.5-6 (cont'd). INDOOR AND OUTDOOR CONTRIBUTIONS TO INDOOR CONCENTRATIONS
o
to
o
o
oo
>
X
OJ
i
£
"^
O
H
6
o
0
H
O
H
W
Percent Percent
Mean Indoor Mean Outdoor Outdoor Indoor
Study Condition Slope Intercept Concentration Concentration Contribution Contribution
Spengler Gas range 0.49 — 30 37 60.4 39.6
etal. (1994) with pilot
light
Gas range 0.4 — 22 33 60.0 40.0
without
pilot light
Electric 0.4 — 17 33 77.6 22.4
stove
Overall 0.49 8.64 27.2 38.3 68.2 31.8
Indoor
Source
Strength Comments
— Mean indoor
and mean
outdoor are
estimated from
Figure 2 in
Spengler et al.
(1994).
— Mean indoor
and mean
outdoor are
estimated from
Figure 2 in
Spengler et al.
(1994).
— Mean indoor
and mean
outdoor are
estimated from
Figure 2 in
Spengler et al.
(1994).
— —
O
-------
O
to
O
O
oo
TABLE AX3.5-7. THE ASSOCIATION BETWEEN INDOOR, OUTDOOR, AND PERSONAL NO2
Study
Summary
Condition
Indoor vs.
Outdoor
Personal vs.
Indoor
Personal vs.
Outdoor
Comments
Mosqueron
et al. (2002)
Simultaneous personal,
indoor, and in-office
48-h averaged NO2
concentrations were
measured with Ogawa
badges for 62 people,
and ambient
concentrations were
provided by local air
monitoring network.
Overall study
0.07
(partial R2)
Gas cooking
interpreted another
7% of indoor NO2
variation
>
X
H
6
O
2
O
H
O
HH
H
W
Emenius
et al. (2003)
Palmes tubes were
used to measure indoor
(in the main living
room) and outdoor
(outside the window of
this room) NO2
concentrations during
a four-wk period
(mean 28 days, range
26-31) in the first
winter season
following recruitment
in the case-control
study.
Without smoker and 0.69 (rp)
gas stove was not
used
With gas stove and 0.13 (rp)
with smoker
With gas stove but 0.06 (rp)
without smoker
p< 0.001
p = 0.43
p = 0.75
O
-------
O
to
O
O
oo
TABLE AX3.5-7 (cont'd). THE ASSOCIATION BETWEEN INDOOR, OUTDOOR, AND PERSONAL NO2
Study
Summary
Condition
Indoor vs.
Outdoor
Personal vs.
Indoor
Personal vs.
Outdoor
Comments
Lee et al. Indoor and outdoor air
(1999) quality of 14 public places
with mechanical
ventilation systems in
Hong Kong were measured
fromOct 1996 to March
1997. Traffic peak h NO,
NO2 was sampled using
Teflon bags and then
shipped back to the
laboratory for further
analysis.
Overall study
0.59 (R2)
0.92 for NO and
0.92 for NOX.
>
X
H
6
o
2
o
H
O
HH
H
W
Garcia-Algar
et al. (2003)
Yanagisawa passive filter
badges were used to
measure indoor NO2
concentrations for 7~15
days for 340 homes in
Barcelona, Spain during
1996-1999. Outdoor NO2
concentrations were
obtained from the fixed
monitoring stations by the
method of CL.
Overall study
0.15(rp)
p = 0.007
O
-------
O
to
O
O
oo
>
X
H
6
o
2
o
H
O
HH
H
W
TABLE AX3.5-7 (cont'd). THE ASSOCIATION BETWEEN INDOOR, OUTDOOR, AND PERSONAL NO2
Indoor vs. Personal vs. Personal vs.
Study Summary Condition Outdoor Indoor Outdoor Comments
Lai et al. The study was conducted
(2006) between 1996 and 2000 in six
EU cities: Athens, Basel,
Helsinki, Milan, Oxford, and
Prague. 48 h averaged indoor
and outdoor NO2 were
collected each home using
diffusion tubes for 302 homes.
Lee et al. Six-day integrated indoor and
(2002) outdoor concentrations of NO2
were measured in two
communities in Southern
California using Yanagisawa
badges for 119 homes in April
and May 1996.
Mukala The one-week averaged
et al. indoor (day-care center),
(2000) outdoor (outside day care
center) and personal NO2 for
162 children aged 3-6 years
old nitrogen dioxide exposure
were measured by Palmes
tube in Helsinki, in 1991.
Overall study
0.13 (partial R")
The overall R for the
multiple linear
regression was 0.67
Overall study
0.60 (rp)
Spring
Winter
Spring (ambient vs.
indoor)
Winter (ambient vs.
indoor)
0.86 (rp)
0.54 (rp)
0.45 (rp)
0.36 (rp)
O
-------
O
to
O
O
oo
TABLE AX3.5-7 (cont'd). THE ASSOCIATION BETWEEN INDOOR, OUTDOOR, AND PERSONAL NO2
Indoor vs. Personal vs. Personal vs.
Study Summary Condition Outdoor Indoor Outdoor Comments
Garrett et al. Four-day averaged NO2 was
(1999) monitored using Yanagisawa passive
samplers in 80 homes in the Latrobe
Valley, Victoria, Australia in March-
April 1994, and Jan-Feb 1995.
Overall study
0.28 (R2)
Log 10
transformed data
>
X
oo
H
6
O
2
O
H
O
HH
H
W
Cotterill and Three consecutive two-week averaged
Kingham outdoor, kitchen, living room, and
(1997) bedroom NO2 were measured using
Palme's tubes in 40 houses in
Huddersfield, UK in late 1994. Half
the houses were located close to a
busy main road and half on residential
roads set back and parallel to the main
road. The sample was split so that
half had gas cookers and half had
electric cookers. These subsets were
split again so that half had double
glazing and half had single glazed
windows.
Overall study
0.59 (rp)
O
-------
O
to
O
O
oo
>
X
TABLE AX3.5-7 (cont'd). THE ASSOCIATION BETWEEN INDOOR, OUTDOOR, AND PERSONAL NO2
Indoor vs. Personal vs. Personal vs.
Study Summary Condition Outdoor Indoor Outdoor Comments
Yangetal. Daily indoor and outdoor
(2004) NO2 concentrations were
measured for 30 consecutive
days in 28 house in Brisbane
(between April and May in
1999), and for 21
consecutive days in 37
houses in Seoul (between
June and Aug in 2000) using
Yanagisawa badges.
Lai et al. During the study, 48-
(2004) averaged personal,
residential indoor,
residential outdoor, and
workplace indoor pollutants
were measured for 50 adults
between 1998 and 2000 in
Oxford, once per person.
NO2 were measured using
passive sampling badges.
Brisbane, electric
range house
Brisbane, gas range
house
Seoul, gas range
house
Overall study
0.70 (R2)
0.57 (R2)
0.52 (R2)
0.29 (rp) 0.47 (rp)
(not (p<0.01)
significant)
-0.41 (rp)
(p < 0.05)
Data were log
transformed
H
6
O
2
O
H
O
HH
H
W
O
Monn et al. During the study, one-wk
(1998) integrated indoor, outdoor
and personal samples were
collected for a
subpopulation (n = 140) of
SAPALDIA study using
Pamles tube between Dec
1993 and Dec 1994 at eight
study centers in Switzerland.
Overall study
Homes without
smoker and without
gas-cooking
0.37 (R2)
0.34 (R2)
0.51(R2)
0.47 (R2)
0.33 (R2)
0.27 (R2)
-------
TABLE AX3.5-7 (cont'd). THE ASSOCIATION BETWEEN INDOOR, OUTDOOR, AND PERSONAL NO2
O
to
O
O
oo
>
X
H
6
O
2
O
H
O
HH
H
W
Study
Levy
etal.
(1998a)
Spengler
etal.
(1994)
Summary
48-h averaged indoor,
outdoor and personal
NO2 were measured in
18 cities in 15
countries around the
world with passive
filter badges in Feb or
March, 1996.
Probability based
population, Los
Angeles Basin, 48-h
averaged indoor,
outdoor and personal
NO2 were measured
(microenvironmental
component of the
study), from May 1987
to May 1988
Condition
Overall study
Overall study
Electric range
Gas range without
pilot light
Gas range with pilot
light
With air conditioning
Without air
conditioning
High ambient
concentration
Low ambient
concentration
Indoor vs. Personal vs.
Outdoor Indoor
— 0.75 (rs)
0.4 (R2) 0.6 (R2)
0.41 (R2) —
0.39 (R2) —
0.44 (R2) —
0.66 (rp) —
0.75 (rp) -
— —
— —
Personal vs.
Outdoor Comments
0.57 (rs) —
0.51 (R2) —
0.52 (R2) —
—
0.44 (R2) —
— —
— —
0.47 (R2) —
0.33 (R2) —
O
-------
TABLE AX3.5-7 (cont'd). THE ASSOCIATION BETWEEN INDOOR, OUTDOOR, AND PERSONAL NO2
O
to
O
O
oo
X
Study
Kousa
etal.
(2001)
Linaker
etal.
(1996)
Summary Condition
The indoor, outdoor, and Overall study
personal NO2 relationship in
three EXPOLIS centers
(Basel, Helsinki, and Prague)
were reported. During the
study, 48-averaged indoor, Helsinki
outdoor, and personal NO2
were measured with Palmes
tubes during 1996-1997.
During the study, one-wk Overall study
integrated personal, indoor
(kitchen, living room),
classroom and playground
NO2 were measured using
Palmes tubes for 46 school
children aged 9-11 in
Southampton, UK.
Indoor vs. Personal vs. Personal vs.
Outdoor Indoor Outdoor Comments
0.44 (R2) 0.53 (R2) 0.37 (R2) Data were log-
transformed
— 0.45 (R2) 0.40 (R2) Data were log-
transformed
— 0.53-0.76 (rp) 0.61-0.65 (rp) Data were log-
transformed
H
6
o
2
o
H
O
HH
H
W
O
-------
TABLE AX3.5-7 (cont'd). THE ASSOCIATION BETWEEN INDOOR, OUTDOOR, AND PERSONAL NO2
O
to
O
O
oo
>
X
to
H
6
O
2
O
H
O
HH
H
W
O
Study
Aim et al.
(1998)
Summary
During the study,
weekly personal,
indoor (day care
center), outdoor (day
care center), and
ambient site NO2
exposures of 246
children aged 3-6 yrs
O J
were measured with
Palmes tubes during
13 wks in winter and
spring in 1991 in
Helsinki.
Indoor vs.
Condition Outdoor
Overall study —
Winter —
Spring —
Winter downtown 0.44 (rp)
Spring downtown 0.84 (rp)
Winter suburban 0.22 (rp)
Spring suburban 0.46 (rp)
Downtown electric —
stove
Downtown gas stove —
Downtown non- —
smoking
Downtown smoking —
Suburban electric —
stove
Suburban gas stove —
Suburban non- —
smoking
Suburban smoking —
Personal vs.
Indoor
0.88 (R2)
—
—
0.32 (rp)
0.75 (rp)
0.04 (rp)
0.75 (rp)
0.67 (rp)
0.50 (rp)
0.67 (rp)
0.47 (rp)
0.55 (rp)
0.50 (rp)
0.48 (rp)
Personal vs.
Outdoor
0.86 (R2)
0.04 (partial R2)
0.50 (partial R2)
0.46 (rp)
0.80 (rp)
0.49 (rp)
0.82 (rp)
0.55 (rp)
0.59 (rp)
0.73 (rp)
0.51(rp)
0.63 (rp)
0.59 (rp)
0.46 (rp)
Comments
0.37 (R2) for personal
vs. ambient
p = 0.01; log
transformed data
p = 0.0001; log
transformed data
Personal vs. indoor
was not significant
(day-care center, not
residential indoor).
Personal vs. indoor,
and indoor vs. outdoor
were not significant
—
—
—
—
—
—
—
—
—
-------
O
to
O
O
oo
>
X
TABLE AX3.5-7 (cont'd). THE ASSOCIATION BETWEEN INDOOR, OUTDOOR, AND PERSONAL NO2
Indoor vs. Personal vs. Personal vs.
Study Summary Condition Outdoor Indoor Outdoor Comments
Kodama During the study,
et al. personal, indoor
(2002) (kitchen, living room,
bedroom, and study
room), and outdoor
NO2 were measured
for 150 junior high
school students with
Yanagisawa badges in
Tokyo. The
investigation was
conducted five times
seasonally, 3 days
each, from Feb 1998 to
Jan 1999.
Summer
Winter
0.31(rp)
0.57 (rp)
0.24 (rp)
0.08 (rp)
H
6
O
2
O
H
O
HH
H
W
O
-------
O
TABLE AX3.5-8. INDOOR, OUTDOOR, AND PERSONAL NO2 LEVELS STRATIFIED BY
EXPOSURE INDICATORS
l->*
O
O
oo
>
X
OJ
1
L/l
r\
O
>
H
1
O
O
1— J
z;
0
H
O
H
W
O
O
O
Ambient
NO2
References Factor Name Factor levels Level
Environmental conditions
Singer etal. Wind Direction Upwind of freeway 20.5
(2004) Downwind and close 26.5
to freeway
Downward and far 21
from freeway
Zotaetal. Season Heating 21
(2005) Non-Heating 17
Serensen et al. Season < 8C 14.6
(2005) > 8C 7.8
Aim et al. Season Winter downtown —
(1998) smoker
Spring downtown —
smoker
Winter downtown —
nonsmoker
Spring downtown —
nonsmoker
Winter suburban —
smoker
Spring suburban —
smoker
Winter suburban —
nonsmoker
Spring suburban —
nonsmoker
Indoor Personal
Ambient NO2 Indoor NO2 Personal
Slope Level Slope Level Slope Comments
— — — — — —
— — — — — —
— — — — — —
— 43 — — — —
— 26 — — — —
— 8.9 — 11.4 — —
— 6.6 — 9.2 — —
— — — 13.5 — —
— — — 15.4 — —
— — — 13.0 — —
— — — 14.1 — —
— — — 11.2 — —
— — — 10.7 — —
— — — 9.2 — —
— — — 8.7 — —
-------
O
>
X
TABLE AX3.5-8 (cont'd). INDOOR, OUTDOOR, AND PERSONAL NO2 LEVELS STRATIFIED
EXPOSURE INDICATORS
BY
l->*
O
O
oo
References Factor Name Factor Levels
Zota et al. Heating season —
(2005)
Vukovich Day Weekday
(2000)
Lee (1997) Day Weekday
Ambient Indoor Personal
NO2 Ambient NO2 Indoor NO2 Personal
Level Slope Level Slope Level Slope Comments
— 3.87 — 17.3 — — —
— — — — — 39% more than
weekend
— — — — — — The effect of
weekday/week-
end is clear but
the paper didn't
give a value to
cite
Weekend
Dwelling conditions
01
01
o
^
H
6
o
0
H
O
H
W
O
7s
Levy et al. Window open
(1998a)
Cotterill and Window
Kingham (1997)
Partti-Pellinen Type of Filtration
et al. (2000)
With —
Without —
Single Glazing —
Double Glazing —
Single Glazing —
Double Glazing —
Mechanical filter 12.3
Mechanical intake 11.5
and mechanical filter
Mechanical intake 12.4
and mechanical and
chemical filter
— — — 30
— — — 26.7
— 9.4 — —
— 9.4 — —
— 11.0 — —
— 12.0 — —
— 9.6 — —
— 12.5 — —
— 6.5 — —
— —
— —
— —
— —
— Gas cooker
homes
— Gas cooker
homes
— —
— —
— —
-------
o
TABLE AX3.5-8 (cont'd). INDOOR, OUTDOOR, AND PERSONAL NO2 LEVELS STRATIFIED BY
EXPOSURE INDICATORS
to
o
o
oo
.
X
OJ
1
01
Oi
O
^
i-rj
H
6
o
0
H
O
H
W
O
o
o
(CONCENTRATIONS ARE IN PPB AND SLOPES ARE DIMENSIONLESS)
References
Yamanaka
(1984)
Zota et al.
(2005)
Levy et al.
(1998a)
Emenius et al.
(2003)
Cotterill and
Kingham (1997)
Zota et al.
(2005)
Lee et al. (2004)
Liard et al.
(1999)
Factor Name
Surface type
Occupancy
Occupancy
Location
Location
Location
Location
Location
Factor Levels
—
—
1
2
Urban
Semi-urban
Suburban
On Main Road
50-85m from Main
Road
—
Industrial
Residential
Main Road
Side Road
Ambient Indoor Personal
NO2 Ambient NO2 Indoor NO2 Personal
Level Slope Level Slope Level Slope Comments
— — — — — Affect decay
rate
— — — 3.2 — — —
— — — — 25.9 — —
— — — — 30.8 — —
16.5 — 9.6 — — — —
11.3 — 6.4 — — — —
7.2 — 4.2 — — — —
— — 7.9 — — — Electric cooker
homes
— — 6.8 — — — Electric cooker
homes
— -0.0093 — — — — —
— — — — 34.9 — —
— — — — 27.8 — —
— — — — 28.1 — —
— — — — 24.3 — —
-------
O
TABLE AX3.5-8 (cont'd). INDOOR, OUTDOOR, AND PERSONAL NO2 LEVELS STRATIFIED
EXPOSURE INDICATORS
BY
l->*
o
o
oo
X
UJ
1
^
References Factor Name Factor Levels
Nakai et al. Location < 20 m
(1995)
20-150 m
>150m
Aim et al. Location Downtown smoker
(1998) Suburban smoker
Downtown
nonsmoker
Suburban
nonsmoker
Ambient Indoor Personal
NO2 Ambient NO2 Indoor NO2
Level Slope Level Slope Level
42.4 — 43.8 — 43.1
34.9 — 38.4 — 35.9
20.3 — 36.4 — 30.1
— — — — 14.6
— — — — 10.9
— — — — 13.6
— — — — 9.0
Personal
Slope Comments
— Recalculated
based published
data
— Recalculated
based published
data
— Recalculated
based published
data
— —
— —
— —
fe
-LJ
H
1
O
o
0
H
O
HH
H
W
O
&
Lee etal. (1996) House structure Single DU
Small multi-DU
Large multi-DU
Single DU
Small multi-DU
Large multi-DU
Single DU
Small multi-DU
Large multi-DU
17
23
23.6
18.4
25.1
25.1
15.9
23.7
24.5
— 17 — —
— 28.9 — —
— 26.8 — —
— 17.8 — —
— 30.2 — —
— 25.4 — —
— 17.3 — —
— 27.8 — —
— 29.1 — —
— Winter
— Winter
— Winter
— Fall
— Fall
— Fall
— Summer
— Summer
— Summer
O
-------
O
TABLE AX3.5-8 (cont'd). INDOOR, OUTDOOR, AND PERSONAL NO2 LEVELS STRATIFIED
EXPOSURE INDICATORS
BY
l->*
O
O
oo
;>
X
OJ
1
01
oo
O
[>
H
1
O
O
z;
0
H
o
1 — I
H
W
O
O
O
References
Gallelli et al.
(2002)
Zota et al.
(2005)
Mosqueron
et al. (2002)
Liard et al.
(1999)
Gallelli et al.
(2002)
Yang et al.
(2004)
Garrett et al.
(1999)
Indoor sources
Zota et al.
(2005)
Lai et al. (2004)
Levy et al.
(1998a)
Belanger et al.
(2006)
Factor Name
Heating system
Frames
Floor level
Floor level
Extractor fan over
cooker
Chimney
Attached garage
Age of house
Supplemental
Heating with stove
Smoking
Smokers present
Ranges
Factor Levels
Individual
Central
Metal
Wood
—
—
Without
With
With vent
Without vent
With
Without
—
—
Smoking
Nonsmoking
With
Without
Electric
Gas
Ambient Indoor Personal
NO2 Ambient NO2 Indoor NO2
Level Slope Level Slope Level
— — 13.7 — —
— — 12.5 — —
— — 12.6 — —
— — 15.0 — —
— 2 — — —
— — — -1.78 —
— — — — 27.5
— — — — 24.8
— — 18.1 — —
— — 30.9 — —
— — 17.3 — —
— — 11.4 — —
— — — 0.5 —
— — — 7.84 —
— — 10.9 — 10.8
— — 11.5 — 14.1
— — — — 34.8
— — — — 26.8
— — 8.6 — —
— — 25.9 — —
Personal
Slope Comments
— Bedroom data
— Bedroom data
— Bedroom data
— Bedroom data
— —
— —
— —
— —
— Kitchen data
— Kitchen data
— —
— —
— —
— —
— —
— —
— —
— —
— —
-------
O
TABLE AX3.5-8 (cont'd). INDOOR, OUTDOOR, AND PERSONAL NO2 LEVELS STRATIFIED BY
EXPOSURE INDICATORS
l->*
o
o
oo
X
r N
OJ
L/l
VO
o
^
i-rj
H
6
o
o
H
O
H
W
O
References
Cotterill and
Kingham(1997)
Yang et al.
(2004)
Schwab et al.
(1994)
Monn et al.
(1998)
Spengler et al.
(1994)
Aim et al.
(1998)
Raaschou-
Nielsen et al.
(1997)
Factor Name
Ranges
Ranges
Ranges
Ranges
Ranges
Ranges
Near fire
Factor Levels
Gas
Electric
Gas
Electric
Gas
Not Gas
Gas with pilot light
Gas without pilot
light
Electric
Gas Geneva
Electric Geneva
Gas Basle
Electric Basle
Gas Lugano
Electric Lugano
Electric smoker
Ambient Indoor
NO2 Ambient NO2 Indoor
Level Slope Level Slope
— — 35.6 —
— — 9.9 —
— — 11.5 —
— — 7.3 —
— — 18.3 —
— — 10.3 —
— — 20.3 —
— — 11.7 —
— — 8 —
— — 20.9 —
— — 16.8 —
— — 15.2 —
— — 12.6 —
— — 18.8 —
— — 15.7 —
— — — —
— — — —
Personal
NO2
Level
—
—
—
—
—
—
—
—
—
23.6
19.9
18.3
16.2
20.9
18.3
—
13.0
Personal
Slope Comments
— Kitchen
— Kitchen
— Bedroom
— Bedroom
— —
— —
— Summer 1998 data
— Summer 1998 data
— Summer 1998 data
— —
— —
— —
— —
— —
— —
— Gas with pilot was
15 ppb higher than
electric; gas without
pilot was 4 ppb
higher than electric
— —
0.052 —
O
-------
O
X
ON
TABLE AX3.5-8 (cont'd). INDOOR, OUTDOOR, AND PERSONAL NO2 LEVELS STRATIFIED
EXPOSURE INDICATORS
BY
l->*
o
o
oo
References
Kawamoto et al.
(1997)
Lee et al. (2004)
Liardetal. (1999)
Kodama et al.
(2002)
Factor Name
Heating time
Heating fuel
Heating appliance
Heater
Factor Levels
Oil fan heater
Kerosene heater
Clean heater
Coal briquette
Petroleum
Gas
Other
Kerosene heater
Gas stove
Ambient Indoor
NO2 Ambient NO2 Indoor
Level Slope Level Slope
— — — —
— — — —
— — — —
— — — —
— — — —
— — — —
— — — —
— — 152.6 —
— — 77.5 —
Personal
NO2
Level
—
—
—
22.2
33.1
27.9
25.2
—
—
Personal
Slope Comments
2.59 —
1.17 —
— —
— —
— —
— —
— —
— Sourth area, Feb
1998
— Sourth area, Feb
Electric heater
30.8
1998
Sourth area, Feb
1998
o
O
[>
H
1
O
o
0
H
O
HH
H
W
O
O
Yang et al. (2004)
Levy et al.
(1998a)
Monn et al.
(1997)
Mosqueron et al.
(2002)
Raaschou-Nielsen
etal. (1997)
Gas water heater
Gas water heater
Gas range
Gas cooking
Gas cooking
Gas appliances at
home
With
Without
With
Without
With
Without
With
Without
With
Without
— — 18.1 — — — —
— — 11.9 — — — —
— — — — 30.5 — —
— — — — 28.2 — —
— — — — 36.4 — —
— — — — 28.5 — —
— — — — 34.8 — —
— — — — 20.5 — —
— — — — — — I/O > 1.2
— — — — — — I/O -0.4 -0.7
— — — 0.068 — — —
— — — — — 0.202 —
-------
O
to
O
O
oo
>
X
H
6
o
2
o
H
O
HH
H
W
TABLE AX3.5-8 (cont'd). INDOOR, OUTDOOR, AND PERSONAL NO2 LEVELS STRATIFIED BY
EXPOSURE INDICATORS
(CONCENTRATIONS ARE IN PPB AND SLOPES ARE DIMENSIONLESS)
References
Garrett et al.
(1999)
Button et al.
(2001)
Serensen et al.
(2005)
Liard et al.
(1999)
Raaschou-
Nielsen et al.
(1997)
Lee et al. (2004)
Liard et al.
(1999)
Factor Name
Gas and smoking
Fireplace setting
Exposure to burning
candle
Exposure to ETS
Exposure to ETS
Cooking fuel
Cooking appliance
Factor Levels
None
Gas stove
Gas heater
Smoking
Multiple
Low
Middle
High
—
With
Without
Petroleum
Gas
Coal briquette
Gas
Electric
Ambient Indoor
NO2 Ambient NO2 Indoor
Level Slope Level Slope
— — 3.0 —
— — 6.3 —
— — 5.0 —
— — 5.7 —
— — 11.2 —
— — 90 —
— — 350 —
— — 360 —
— — — —
— — — —
— — — —
— — — —
— — — —
— — — —
— — — —
— — — —
— — — —
Personal
NO2 Personal
Level Slope Comments
— — I/O ratio increase
from 0.8 to 1.3 to
1.8 to 2.2 in houses
with no, one, two,
or three major
indoors sources
— — —
— — —
— — —
— — —
— — —
— — —
— — —
— 0.031 —
25.1 — —
26.3 — —
— 0.056 —
26.1 — —
33.1 — —
20.6 — —
25.8 — —
25.5 — —
O
-------
O
TABLE AX3.5-8 (cont'd). INDOOR, OUTDOOR, AND PERSONAL NO2 LEVELS STRATIFIED BY
EXPOSURE INDICATORS
l->*
o
o
oo
j>
X
OJ
1
ON
to
o
s>
H
6
o
0
H
O
H
W
References Factor Name Factor Levels
Dennekamp Cooking 1 ring
etal. (2001)
2 rings
3 rings
4 rings
Boil water
Stir fry
Fry bacon
Bake cake
Roast meat
Bake potatoes
Ambient Indoor Personal
NO2 Ambient NO2 Indoor NO2
Level Slope Level Slope Level
— — 437 — —
— — 310 — —
— — 584 — —
— — 996 — —
— — 184 — —
— — 92 — —
— — 104 — —
— — 230 — —
— — 296 — —
— — 373 — —
Personal
Slope Comments
— The max 5 min
concentrations
— The max 5 min
concentrations
— The max 5 min
concentrations
— The max 5 min
concentrations
— The max 5 min
concentrations
— The max 5 min
concentrations
— The max 5 min
concentrations
— The max 5 min
concentrations
— The max 5 min
concentrations
— The max 5 min
concentrations
O
-------
O
>
X
TABLE AX3.5-8 (cont'd). INDOOR, OUTDOOR, AND PERSONAL NO2 LEVELS STRATIFIED BY
EXPOSURE INDICATORS
References
Personal activities
Levy et al.
(1998a)
Chao and Law
(2000)
Kawamoto et al.
(1997)
Factor
Name
Commute
Commute
Cooking to
stay home h
ratio
Cooking time
Ambient Indoor
NO2 Ambient NO2
Factor Levels Level Slope Level
Commuting less than — — —
Ih
Without commuting — — —
< 1 h — — —
1-2 h — — —
2-3 h — — —
3-4 h — — —
4-6 h — — —
— — — —
— — — —
Personal
Indoor NO2 Personal
Slope Level Slope Comments
— 29.9 — —
— 27.9 — —
— 21.7 — —
— 24.7 — —
— 24.6 — —
— 20.1 — —
— 27.9 — —
— — 55.4 —
— — 1.61 —
H
6
o
2
o
H
O
HH
H
W
O
-------
TABLE AX3.5-9. PERSONAL NO2 LEVELS STRATIFIED BY DEMOGRAPHIC AND SOCIOECONOMIC FACTORS
(CONCENTRATIONS ARE IN PPB AND SLOPES ARE DIMENSIONLESS)
to
o
o
oo
^
X
OJ
1
ON
DRAFT-DO I
^
0
H
O
H
W
References
Rotkoetal. (2001)
Rotkoetal. (2001)
Raaschou-Nielsen (1997)
Lee etal., (2004)
Lee etal., (2004)
Rotkoetal. (2001)
Rotkoetal. (2001)
Raaschou-Nielsen (1997)
Rotkoetal. (2001)
Rotkoetal. (2001)
Rotkoetal. (2001)
Rotkoetal. (2001)
Rotkoetal. (2001)
Rotkoetal. (2001)
Algar et al. (2004)
Algar et al. (2004)
Algar et al. (2004)
Factor Type
Demography
Demography
Demography
Demography
Demography
Demography
Demography
Demography
Socioeconomic
Socioeconomic
Socioeconomic
Socioeconomic
Socioeconomic
Socioeconomic
Socioeconomic
Socioeconomic
Socioeconomic
Factor Name
Age
Age
Age
Gender
Gender
Gender
Gender
Gender
Education years
Education years
Employment
Employment
Occupational status
Occupational status
Employment
Employment
Employment
Factor levels Personal NO2 Level Personal Slope
25-34
35-55
Female
Male
Female
Male
<14 years
> 14 years
Employed
Not employed
Non white collar
White collar
Managerial, technical and
professional (Barcelona)
Skilled (manual and non-
manual) (Barcelona)
Unskilled and partly skilled
(Barcelona)
13.1
13.1
0.056
33
29
12.9
13.4
0.267
13.8
12.8
13.3
11.5
13.4
13.0
12.2
12.3
12.1
O
-------
TABLE AX3.6-1. CORRELATIONS (PEARSON CORRELATION COEFFICIENT)
BETWEEN AMBIENT NO2 AND AMBIENT COPOLLUTANTS
Study
(ambient)
This CD
This CD
This CD
This CD
Kim et al.
(2006)
Sarnat et al.
(2006)
Sarnat et al.
(2006)
Connell et al.
(2005)
Kim et al.
(2005)
Sarnat et al.
(200 1)4
Sarnat et al.
(2001)
Hochadel et al.
(2006)
Location
Los Angeles
Riverside, CA
Chicago
New York City
Toronto
Steubenville, OH
(autumn)
Steubenville, OH
(summer)
Steubenville, OH
St. Louis (RAPS)
Baltimore, MD
(summer)
Baltimore, MD
(winter)
Ruhr area,
Germany
PM25
0.49 (u3), 0.56 (s)
0.49 (s)
0.58 (u)
0.44
0.78 (0.70 for sulfate,
0.82 for EC)
0.00
(0.1 for sulfate, 0.24
for EC)
0.50
0.37
0.75
0.41, (0.93 for EC2)
CO O3 SO2
0.59 (u), -0.29 (u),
0.64 (s) -O.ll(s)
0.43 (u), 0.045 (u),
0.41 (s), 0.10(s),
0.15 (r) -0.31 (r)
0.53 (u), -0.20 (u)
0.46 (s)
0.46(u) -0.06(u)
0.72
0.641
0.75 0.02
not significant
0.76 -0.71 -0.17
March 2008
AX3-165
DRAFT-DO NOT QUOTE OR CITE
-------
TABLE AX3.6-1 (cont'd). CORRELATIONS (PEARSON CORRELATION
COEFFICIENT) BETWEEN AMBIENT NO2 AND AMBIENT COPOLLUTANTS
Study
(ambient)
Location
PM2
CO
03
SO2
Hazenkamp-
von Arx et al.
(2004)
Cyrys et al.
(2003)
Mosqueron
et al. (2002)
Rojas-B radio
et al. (2002)
21 European cities 0.75
Erfurt, Germany 0.50
0.74
Paris
0.69
Santiago, Chile 0.77
'Value with respect to NOX.
Inferred based on EC as dominant contributor to PM2 5 absorbance.
3u: urban; s: suburban; and r: rural
4Spearman correlation coefficient was reported
TABLE AX3.6-2. CORRELATIONS (PEARSON CORRELATION COEFFICIENT)
BETWEEN PERSONAL NO2 AND PERSONAL COPOLLUTANTS
Study
Kim et al.
(2006)
Modig et al.
(2004)
Mosqueron et
al. (2002)
Jarvis et al.
(2005)
Lee et al.
(2002)
Lai et al. (2004)
Location PM2.5
Toronto 0.41
Umea
Paris 0.1 2 but not
significant
21
European
cities
Oxford -0.1
CO VOCs HONO
0.12
0.06 for 1,3-butadiene;
and 0.10 for benzene
0.77 for indoor
NO2 and indoor
HONO
0.51 for indoor
NO2 and indoor
HONO
0.3 - 0.11 for TVOCs
March 2008
AX3-166
DRAFT-DO NOT QUOTE OR CITE
-------
TABLE AX3.6-3. CORRELATIONS (PEARSON CORRELATION COEFFICIENT)
BETWEEN PERSONAL NO2 AND AMBIENT COPOLLUTANTS
Study Location PM2.5 Sulfate EC PMio CO
Sarnatetal. Steubenville / 0.46 0.35 0.57
(2006) Fall
Sarnatetal. Steubenville/ 0.00 0.1 0.17
(2006) Summer not significant
Kim et al. (2006) Toronto 0.30 0.20
Rojas-Bracho Santiago 0.65 0.39
et al. (2002)
March 2008 AX3-167 DRAFT-DO NOT QUOTE OR CITE
-------
TABLE AX3.6-4. CORRELATIONS (PEARSON CORRELATION COEFFICIENT)
BETWEEN AMBIENT NO2 AND PERSONAL COPOLLUTANTS
Study Location PM2.5 Sulfate EC Ultrafine-particle
Sarnatetal. Steubenville / 0.71 0.52 0.70
(2006) Fall
Sarnatetal. Steubenville/ 0.00 0.1 0.26
(2006) Summer not significant
Vinzents Copenhagen 0.49 (R2) explained by
et al. (2005) ambient NO2 and ambient
temperature
March 2008 AX3-168 DRAFT-DO NOT QUOTE OR CITE
-------
TABLE AX3.7-1. THE ESSENTIAL ATTRIBUTES OF THE PNEM, HAPEM, APEX,
SHEDS, AND MENTOR-1A
Exposure Estimate
Characterization of
the High-End
Exposures
Typical Spatial
Scale/Resolution
Temporal
Scale/Resolution
Population
Activity Patterns
Assembly
Microenvironment
Concentration
Estimation
Microenviron-
mental (ME)
Factors
Specification of
Indoor Source
Emissions
Commuting
Patterns
Exposure Routes
Potential Dose
Calculation
Physiologically
Based Dose
Variability/
Uncertainty
pNEM
Hourly averaged
Yes
Urban
areas/Census
tract level
A yr/one h
Top-down
approach
Non-steady-state
and steady-state
mass balance
equations (hard-
coded)
Random samples
from probability
distributions
Yes (gas-stove,
tobacco
smoking)
Yes
Inhalation
Yes
No
Yes
HAPEM
Annual averaged
No
Ranging from
urban to national/
Census tract level
A yr/one h
Top-down
approach
Linear
relationship
method (hard-
coded)
Random samples
from probability
distributions
Available; set to
zero in HAPEM6
Yes
Inhalation
No
No
No
APEX
Hourly averaged
Yes
Urban
area/Census tract
level
A yr/one h
Bottom-up
"person-
oriented"
approach
Non-steady-state
mass balance and
linear regression
(flexibility of
selecting
algorithms)
Random samples
from probability
distributions
Yes (multiple
sources defined
by the user)
Yes
Inhalation
Yes
No
Yes
SHEDS
Activity event
based
Yes
Urban
areas/Census
tract level
A yr/event based
Bottom-up
"person-oriented"
approach
Steady-state mass
balance equation
(residential) and
linear regression
(non-residential)
(hard-coded)
Random samples
from probability
distributions
Yes (gas-stove,
tobacco smoking,
other sources)
Yes
Inhalation
Yes
Yes
Yes
MENTOR-1A
Activity event
based
Yes
Multiscale/
Census tract level
A yr/activity
event based time
step
Bottom-up
"person-oriented"
approach
Non-steady-state
mass balance
equation with
indoor air
chemistry
module or
regression
methods
(flexibility of
selecting
algorithms)
Random samples
from probability
distributions
Yes (multiple
sources defined
by the user)
Yes
Multiple
(optional)
Yes
Yes
Yes (Various
"Tools")
March 2008
AX3-169
DRAFT-DO NOT QUOTE OR CITE
-------
1 AX3.9 REFERENCES
2 Ainsworth, B. E.; Haskell, W. L.; Leon, A. S.; Jacobs, D. R., Jr.; Montoye, H. J.; Sallis, J. F.;
3 Paffenbarger, R. S., Jr. (1993) Compendium of physical activities: classification of
4 energy costs of human physical activities. Med. Sci. Sports Exer. 25: 71-80.
5 Albinet, A.; Leoz-Garziandia, E.; Budzinski, H.; Villenave, E. (2006) Simultaneous analysis of
6 oxygenated and nitrated polycyclic aromatic hydrocarbons on standard reference material
7 1646a (urban dust) and on natural ambient air samples by gas chromatography-mass
8 spectrometry with negative ion chemical ionization. J. Chromatogr. A 1121: 106-113.
9 Algar, O. G.; Pichini, S.; Basagana, X.; Puig, C.; Vail, O.; Torrent, M.; Harris, J.; Sunyer, J.;
10 Cullinan, P., the AMICS group. (2004) Concentrations and determinations of NO2 in
11 homes of Ashford, UK and Barcelona and Menorca, Spain. Indoor Air 14: 298-304.
12 Aim, S.; Mukala, K.; Pasanen, P.; Tiittanen, P.; Ruuskanen, J.; Tuomisto, J.; Jantunen, M. J.
13 (1998) Personal NO2 exposures of preschool children in Helsinki. J. Exposure Anal.
14 Environ. Epidemiol. 8: 79-100.
15 Ammann, M.; Kalberer, M.; Jost, D. T.; Tobler, L.; Rossler, E.; Piguet, D.; Gaggeler, H. W.;
16 Baltensperger, U. (1998) Heterogeneous production of nitrous acid on soot in polluted air
17 masses. Nature (London) 395: 157-160.
18 Armstrong, B. K.; White, E.; Saracci, R. (1992) Principles of exposure measurement in
19 epidemiology. Oxford, United Kingdom: Oxford University Press. (Monographs in
20 epidemiology and biostatistics: v. 21).
21 Atmospheric and Environmental Research, Inc. (AER). (2004) Subgrid scale modeling of plumes
22 in three-dimensional air quality models for ozone and particulate matter (plume-in-grid).
23 Available: http://www.aer.com/scienceResearch/aq/web_update_3.htm (5 March, 2007).
24 Bae, H.; Yang, W.; Chung, M. (2004) Indoor and outdoor concentrations of RSP, NO2 and
25 selected volatile organic compounds at 32 shoe stalls located near busy roadways in
26 Seoul, Korea. Sci. Total Environ. 323: 99-105.
27 Bauer, M. A.; Utell, M. J.; Morrow, P. E.; Speers, D. M.; Gibb, F. R. (1986) Inhalation of 0.30
28 ppm nitrogen dioxide potentiates exercise-induced bronchospasm in asthmatics. Am.
29 Rev. Respir. Dis. 134: 1203-1208.
30 Belanger, K.; Gent, J. F.; Triche, E. W.; Bracken, M. B.; Leaderer, B. P. (2006) Association of
31 indoor nitrogen dioxide exposure with respiratory symptoms in children with asthma.
32 Am. J. Respir. Crit. Care Med. 173: 297-303.
33 Bell, S.; Ashenden, T. W. (1997) Spatial and temporal variation in nitrogen dioxide pollution
34 adjacent to rural roads. Biomed. Life Sci. Earth Environ. Sci. 95: 87-98.
35 Berglund, M.; Braback, L.; Bylin, G.; Jonson, J.-O.; Vahter, M. (1994) Personal NO2 exposure
36 monitoring shows high exposure among ice-skating schoolchildren. Arch. Environ.
37 Health 49: 17-24.
38
39
40
March 2008 AX3-170 DRAFT-DO NOT QUOTE OR CITE
-------
1 Berkowicz, R.; Ketzel, M.; Vachon, G.; Louka, P.; Rosant, J.-M.; Mestayer, P. G.; Sini, J.-F.
2 (2002) Examination of traffic pollution distribution in a street canyon using the Nantes'99
3 experimental data and comparison with model results. Water Air Soil Pollut. Focus 2:
4 311-324.
5 Bertoni, G.; Toppa, R.; Allegerni, I. (2001) The internal consistency of the "analyst" diffusive
6 sampler—a long-term field test. Chromatographia 54: 653-657.
7 Biller, W. F.; Feagans, T. B.; Johnson, T. R.; Duggan, G. M.; Paul, R. A.; McCurdy, T.; Thomas,
8 H. C. (1981) A general model for estimating exposure associated with alternative
9 NAAQS. Presented at: 74th annual meeting of the Air Pollution Control Association;
10 June; Philadelphia, PA. Pittsburgh, PA: Air Pollution Control Association; paper no. Si-
ll 18.4.
12 Brauer, M.; Ryan, P. B.; Suh, H. H.; Koutrakis, P.; Spengler, J. D.; Leslie, N. P.; Billick, I. H.
13 (1990) Measurements of nitrous acid inside two research houses. Environ. Sci. Technol.
14 24:1521-1527.
15 Brauer, M.; Rasmussen, T. R.; Kjaergaard, S. K.; Spengler, J. D. (1993) Nitrous acid formation in
16 an experimental exposure chamber. Indoor Air 3: 94-105.
17 Brauer, M.; Lee, K.; Spengler, J. D.; Salonen, R. O.; Pennanen, A.; Braathen, O. A.; Mihalikova,
18 E.; Miskovic, P.; Nozaki, A.; Tsuzuki, T.; Rui-Jin, S.; Xu, Y.; Qing-Xiang, Z.;
19 Drahonovska, H.; Kjaergaard, S. (1997) Nitrogen dioxide in indoor ice skating facilities:
20 an international survey. J. Air Waste Manage. Assoc. 47: 1095-1102.
21 Brauer, M.; Henderson, S.; Kirkham, T.; Lee, K. S.; Rich, K.; Teschke, K. (2002) Review of the
22 health risks associated with nitrogen dioxide and sulfur dioxide in indoor air. Vancouver,
23 British Columbia, Canada: University of British Columbia, School of Occupational and
24 Environmental Hygiene. Report to Health Canada.
25 Britter, R. E.; Hanna, S. R. (2003) Flow and dispersion in urban areas. Ann. Rev. Fluid Mech.
26 35:469-496.
27 Buck, J. W.; Tolle, D. A.; Whelan, G; Mast, T. J.; Peffers, M. S.; Evers, D. P.; Corley, R. A.;
28 Eslinger, M. A.; Kirk, J. L.; Pelton, M. A.; Townsend, C. C.; Nishioka, M. G.; Kogan, V.;
29 Mahasenan, S.; Dorow, K. E.; Stenner, R. D.; Strenge, D. L. (2002) Design of the
30 comprehensive chemical exposure framework and identification of research needs for
31 American Chemistry Council. Prepared for: American Chemistry Council Long Range
32 Research Initiative Team; Arlington, VA. Richland, WA: Battelle Pacific Northwest
33 Division; PNWD-3184.
34 Burke, J. M.; Zufall, M. J.; Ozkaynak, H. (2001) A population exposure model for particulate
35 matter: case study results for PM2.5 in Philadelphia, PA. J. Exposure Anal. Environ.
36 Epidemiol. 11:470-489.
37 Burnett, R. T.; Stieb, D.; Brook, J. R.; Cakmak, S.; Dales, R.; Raizenne, M.; Vincent, R.; Dann,
38 T. (2004) Associations between short-term changes in nitrogen dioxide and mortality in
39 Canadian cities. Arch. Environ. Health 59: 228-236.
40 Bush, T.; Smith, S.; Stevenson, K.; Moorcroft, S. (2001) Validation of nitrogen dioxide diffusion
41 tube methodology in the UK. Atmos. Environ. 35: 289-296.
March 2008 AX3-171 DRAFT-DO NOT QUOTE OR CITE
-------
1 California Air Resources Board. (2007) Review of the California ambient air quality standard for
2 nitrogen dioxide. Staff report: initial statement of reasons for proposed rulemaking.
3 Sacramento, CA: California Environmental Protection Agency, Air Resources Board.
4 Available: http://www.arb.ca.gov/research/aaqs/no2-rs/no2tech.pdffl4 August, 2007].
5 Campbell, G. W.; Stedman, J. R.; Stevenson, K. (1994) A survey of nitrogen dioxide
6 concentrations in the United Kingdom using diffusion tubes, July-December 1991.
7 Atmos. Environ. 28: 477-486.
8 Carlisle, A. J.; Sharp, N. C. C. (2001) Exercise and outdoor ambient air pollution. Br. J. Sports
9 Med. 35: 214-222.
10 Carslaw, N. (2007) A new detailed chemical model for indoor air pollution. Atmos. Environ. 41:
11 1164-1179.
12 Chan, L. Y.; Chan, C. Y.; Qin, Y. (1999) The effect of commuting microenvironment on
13 commuter exposures to vehicular emission in Hong Kong. Atmos. Environ. 33: 1777-
14 1787.
15 Chang, C.-H. (2006) Computational fluid dynamics simulation of concentration distributions
16 from a point source in the urban street canyons. J. Aerospace Eng. 19: 80-86.
17 Chang, C.-H.; Meroney, R. N. (2003) Concentration and flow distributions in urban street
18 canyons: wind tunnel and computational data. J. Wind Eng. Ind. Aerodyn. 91: 1141-
19 1154.
20 Chao, C. Y. (2001) Comparison between indoor and outdoor air contaminant levels in residential
21 buildings from passive sampler study. Build. Environ. 36: 999-1007.
22 Chao, C. Y. H.; Law, A. (2000) A study of personal exposure to nitrogen dioxide using passive
23 samplers. Build. Environ. 35: 545-553.
24 Chau, C. K.; Tu, E. Y.; Chan, D. W. T.; Burnett, J. (2002) Estimating the total exposure to air
25 pollutants for different population age groups in Hong Kong. Environ. Int. 27: 617-630.
26 Ching, J.; Herwehe, J.; Swall, J. (2006) On joint deterministic grid modeling and sub-grid
27 variability conceptual framework for model evaluation. Atmos. Environ. 40: 4935-4945.
28 Christakos, G. (2000) Modern spatiotemporal geostatistics. New York, NY: Oxford University
29 Press. (International Association for Mathematical Geology; studies in mathematical
30 geology: 6).
31 Christakos, G.; Hristopulos, D. T. (1998) Spatiotemporal environmental health modelling. A
32 tractatus stochasticus. Boston, MA: Kluwer Academic Publishers.
33 Christakos, G.; Kolovos, A. (1999) A study of the spatiotemporal health impacts of ozone
34 exposure. J. Exposure Anal. Environ. Epidemiol. 9: 322-335.
35 Christakos, G.; Vyas, V. M. (1998a) A composite space/time approach to studying ozone
36 distribution over eastern United States. Atmos. Environ. 32: 2845-2857.
37 Christakos, G.; Vyas, V. M. (1998b) A novel method for studying population health impacts of
38 spatiotemporal ozone distribution. Soc. Sci. Med. 47: 1051-1066.
March 2008 AX3-172 DRAFT-DO NOT QUOTE OR CITE
-------
1 Chuang, J. C.; Mack, G. A.; Kuhlman, M. R.; Wilson, N. K. (1991) Polycyclic aromatic
2 hydrocarbons and their derivatives in indoor and outdoor air in an eight-home study.
3 Atmos. Environ. Part B 25: 369-380.
4 Clapp, L.; Jenkin, M. (2001) Analysis of the relationship between ambient levels of Os, NC>2 and
5 NO as a function of NOX in the UK. Atmos. Environ. 35: 6391-6405.
6 Clench-Aas, J.; Bartonova, A.; B0hler, T.; Gr0nskei, K. E.; Sivertsen, B.; Larssen, S. (1999) Air
7 pollution exposure monitoring and estimating. Part I. Integrated air quality monitoring
8 system. J. Environ. Monit. 1: 313-319.
9 Cocheo, V.; Boaretto, C.; Sacco, P. (1996) High uptake rate radial diffusive sampler suitable for
10 both solvent and thermal desorption. Am. Ind. Hyg. Assoc. J. 57: 897-904.
11 Code of Federal Regulations. (2002) Ambient air quality surveillance; appendix E - probe and
12 monitoring path citing criteria for ambient air quality monitoring. C. F. R. 40: §58.
13 Colbeck, I. (1998) Nitrogen dioxide in the workplace environment. Environ. Monit. Assess. 52:
14 123-130.
15 Connell, D. P.; Withum, J. A.; Winter, S. E.; Statnick, R. M. (2005) The Steubenville
16 Comprehensive Air Monitoring Program (SCAMP): analysis of short-term and episodic
17 variations in PM2.5 concentrations using hourly air monitoring data. J. Air Waste
18 Manage. Assoc. 55: 559-573.
19 Cotterill, A.; Kingham, S. (1997) Nitrogen dioxide in the home: cooking, double glazing, or
20 outdoor air? Indoor Built Environ. 6: 344-349.
21 Cox, R. M. (2003) The use of passive sampling to monitor forest exposure to 63, NC>2, and SC^:
22 a review and some case studies. Environ. Pollut. 126: 301-311.
23 Cyrys, J.; Heinrich, J.; Richter, K.; Wolke, G.; Wichmann, H. E. (2000) Sources and
24 concentrations of indoor nitrogen dioxide in Hamburg (west Germany) and Erfurt (east
25 Germany). Sci. Total Environ. 250: 51-62.
26 Cyrys, J.; Stolzel, M.; Heinrich, J.; Kreyling, W. G.; Menzel, N.; Wittmaack, K.; Tuch, T.;
27 Wichmann, H.-E. (2003) Elemental composition and sources of fine and ultrafme
28 ambient particles in Erfurt, Germany. Sci. Total Environ. 305: 143-156.
29 Da Silva, A. S.; Cardoso, M. R.; Meliefste, K.; Brunekreef, B. (2006) Use of passive diffusion
30 sampling method for defining NC>2 concentrations gradient in Sao Paulo, Brazil. Environ.
31 Health: Global Access Sci. Source 5: 19.
32 De Santis, F.; Allegrini, L; Fazio, M. C.; Pasella, D.; Piredda, R. (1997) Development of a
33 passive sampling technique for the determination of nitrogen dioxide and sulphur dioxide
34 in ambient air. Anal. Chim. Acta 346: 127-134.
35 De Santis, F.; Dogeroglu, T.; Fino, A.; Menichelli, S.; Vazzana, C.; Allergrini, I. (2002)
36 Laboratory development and field evaluation of a new diffusive sampler to collect
37 nitrogen oxides in the ambient air. Anal. Bioanal. Chem. 373: 901-907.
38 De Santis, F.; Fino, A.; Menichelli, S.; Vazzana, C.; Allegrini, I. (2004) Monitoring the air
39 quality around an oil refinery through the use of diffusive sampling. Anal. Bioanal.
40 Chem. 378: 782-788.
March 2008 AX3-173 DRAFT-DO NOT QUOTE OR CITE
-------
1 Dennekamp, M.; Howarth, S.; Dick, C. A. I; Cherrie, J. W.; Donaldson, K.; Seaton, A. (2001)
2 Ultrafme particles and nitrogen oxides generated by gas and electric cooking. Occup.
3 Environ. Med. 58: 511-516.
4 Dimitroulopoulou, C.; Ashmore, M. R.; Byrne, M. A.; Kinnersley, R. P. (2001) Modelling of
5 indoor exposure to nitrogen dioxide in the UK. Atmos. Environ. 35: 269-279.
6 Dimitroulopoulou, C.; Ashmore, M. R.; Hill, M. T. R.; Byrne, M. A.; Kinnersley, R. (2006)
7 INDAIR: a probabilistic model of indoor air pollution in UK homes. Atmos. Environ. 40:
8 6362-6379.
9 Doddridge, B. G.; Dickerson, R. R.; Holland, J. Z.; Cooper, J. N.; Wardell, R. G.; Poulida, O.
10 (1991) Observations of tropospheric trace gases and meteorology in rural Virginia using
11 an unattended monitoring system: Hurricane Hugo (1989), a case study. J. Geophys. Res.
12 96: 9341-9360.
13 Doddridge, B. G.; Dickerson, R. R.; Wardell, R. G.; Civerolo, K. L.; Nunnermacker, L. J. (1992)
14 Trace gas concentrations and meteorology in rural Virginia. 2. Reactive nitrogen
15 compounds. J. Geophys. Res. 97: 20631-20646.
16 Dubowski, Y.; Sumner, A. L.; Menke, E. J.; Gaspar, D. J.; Newberg, J. T.; Hoffman, R. C.;
17 Penner, R. M.; Hemminger, J. C.; Finlayson-Pitts, B. J. (2004) Interactions of gaseous
18 nitric acid with surfaces of environmental interest. Phys. Chem. Chem. Phys. 6: 3879-
19 3888.
20 Dutton, S. J.; Hannigan, M. P.; Miller, S. L. (2001) Indoor pollutant levels from the use of
21 unvented natural gas fireplaces in Boulder, Colorado. J. Air Waste Manage. Assoc. 51:
22 1654-1661.
23 Eide, L; Neverdal, G.; Thorvaldsen, B.; Grung, B.; Kvalheim, O. M. (2002) Toxicological
24 evaluation of complex mixtures by pattern recognition: correlating chemical fingerprints
25 to mutagenicity. Environ. Health Perspect. Suppl. 110(6): 985-988.
26 Ekberg, L. E. (1996) Relationships between indoor and outdoor contaminants in mechanically
27 ventilated buildings. Indoor Air 6: 41-47.
28 Emenius, G.; Pershagen, G.; Berglind, N.; Kwon, H.-J.; Lewne, M.; Nordvall, S. L.; Wickman,
29 M. (2003) NC>2, as a marker of air pollution, and recurrent wheezing in children: a nested
30 case-control study within the BAMSE birth cohort. Occup. Environ. Med. 60: 876-881.
31 Emery, C.; Yarwood, G. (2005) Implementing PM chemistry in the CAMx "IRON PiG" plume-
32 in-grid module. Prepared for: Lake Michigan Air Directors Consortium; Des Plaines, IL.
33 Novato, CA: ENVIRON International Corporation. Available:
34 http://www.ladco.org/reports/rpo/MWRPOprojects/Modeling/CAMx_PiG.pdf [5 March,
35 2007].
36 Emmons, L. K.; Carroll, M. A.; Hauglustaine, D. A.; Brasseur, G. P.; Atherton, C.; Penner, J.;
37 Sillman, S.; Levy, H., II; Rohrer, F.; Wauben, W. M. F.; van Velthoven, P. F. J.; Wang,
38 Y.; Jacob, D.; Bakwin, P.; Dickerson, R.; Doddridge, B.; Gerbig, C.; Honrath, R.; Hubler,
39 G; Jaffe, D.; Kondo, Y.; Munger, J. W.; Torres, A.; Volz-Thomas, A. (1997)
40 Climatologies of NOX and NOy: a comparison of data and models. Atmos. Environ. 31:
41 1837-1850.
March 2008 AX3-174 DRAFT-DO NOT QUOTE OR CITE
-------
1 Fagundez, L. A.; Fernandez, V. L.; Marino, T. H.; Martin, I; Persano, D. A.; Benitez, R. M. Y.;
2 Sadafiiowski, I. V.; Codnia, J.; Zalts, A. (2001) Preliminary air pollution monitoring in
3 San Miguel, Buenos Aires. Environ. Monitor. Assess. 71: 61-70.
4 Fan, Z.; Lioy, P.; Weschler, C.; Fiedler, N.; Kipen, H.; Zhang, J. (2003) Ozone-initiated
5 reactions with mixtures of volatile organic compounds under simulated indoor
6 conditions. Environ. Sci. Technol. 37: 1811-1821.
7 Farrow, A.; Greenwood, R.; Preece, S.; Golding, J. (1997) Nitrogen dioxide, the oxides of
8 nitrogen, and infants' health symptoms. Arch. Environ. Health 52: 189-194.
9 Febo, A.; Perrino, C. (1991) Prediction and experimental evidence for high air concentration of
10 nitrous acid in indoor environments. Atmos. Environ. Part A 25: 1055-1061.
11 Ferm, M.; Svanberg, P.-A. (1998) Cost-efficient techniques for urban and background
12 measurements of SC>2 andNC>2. Atmos. Environ. 32: 1377-1381.
13 Fernando, H. J. S.; Lee, S. M.; Anderson, J.; Princevac, M.; Pardyjak, E.; Grossman-Clarke, S.
14 (2001) Urban fluid mechanics: air circulation and contaminant dispersion in cities.
15 Environ. Fluid Mech. 1: 107-164.
16 Finlayson-Pitts, B. J.; Pitts, J. N., Jr. (2000) Chemistry of the upper and lower atmosphere:
17 theory, experiments and applications. San Diego, CA: Academic Press.
18 Finlayson-Pitts, B. J.; Wingen, L. M.; Sumner, A. L.; Syomin, D.; Ramazan, K. A. (2003) The
19 heterogeneous hydrolysis of NC>2 in laboratory system and in outdoor and indoor
20 atmospheres: an integrated mechanism. Phys. Chem. Chem. Phys. 5: 223-242.
21 Foley, G. J.; Georgopoulos, P. G.; Lioy, P. J. (2003) Accountability within new ozone standards.
22 Environ. Sci. Technol. 37: 392A-399A.
23 Fortmann, R.; Kariher, P.; Clayton, R. (2001) Indoor air quality: residential cooking exposures.
24 Final report. Sacramento, CA: State of California Air Resources Board; ARB Contract
25 No. 97-330. Available: http://arb.ca.gov/research/abstracts/97-330.htm [22 May, 2007].
26 Fraigneau, Y. C.; Gonzalez, M.; Coppalle, A. (1995) Dispersion and chemical reaction of a
27 pollutant near a motorway. Sci. Total Environ. 169: 83-91.
28 Fraser, M. P.; Cass, G. R.; Simoneit, B. R. T.; Rasmussen, R. A. (1998) Air quality model
29 evaluation data for organics. 5. C6-C22 nonpolar and semipolar aromatic compounds.
30 Environ. Sci. Technol. 32: 1760-1770.
31 Freijer, J. L; Bloemen, H. J. T. (2000) Modeling relationships between indoor and outdoor air
32 quality. J. Air Waste Manage. Assoc. 50: 292-300.
33 Freijer, J. L; Bloemen, H. J. T.; de Loos, S.; Marra, M.; Rombout, P. J. A.; Steentjes, G. M.; Van
34 Veen, M. P. (1998) Modelling exposure of the Dutch population to air pollution. J.
35 Hazard. Mater. 61: 107-114.
36 Gair, A. J.; Penkett, S. A. (1995) The effects of wind speed and turbulence on the performance of
37 diffusion tube samplers. Atmos. Environ. 29: 2529-2533.
38 Gair, A. J.; Penkett, S. A.; Oyola, P. (1991) Development of a simple passive technique for the
39 determination of nitrogen-dioxide in remote continental locations. Atmos. Environ. Part
40 A 25: 1927-1939.
March 2008 AX3-175 DRAFT-DO NOT QUOTE OR CITE
-------
1 Gallelli, G.; Orlando, P.; Perdelli, F.; Panatoo, D. (2002) Factors affecting individual exposure to
2 NO2 in Genoa (northern Italy). Sci. Total Environ. 287: 31-36.
3 Garcia-Algar, 6.; Zapater, M.; Figueroa, C.; Vail, O.; Basagafia, X.; Sunyer, J.; Freixa, A.;
4 Guardino, X.; Pichini, S. (2003) Sources and concentrations of indoor nitrogen dioxide in
5 Barcelona, Spain. J. Air Waste Manage. Assoc. 53: 1312-1317.
6 Garcia Algar, O.; Pichini, S.; Basagafia, X.; Puig, C.; Vail, O.; Torrent, M.; Harris, J.; Sunyer, J.;
7 Cullinan, P. (2004) Concentrations and determinants of NC>2 in homes of Ashford, UK
8 and Barcelona and Menorca, Spain. Indoor Air 14: 298-304.
9 Garrett, M. H.; Hooper, M. A.; Hooper, B. M. (1999) Nitrogen dioxide in Australian homes:
10 levels and sources. J. Air Waste Manage. Assoc. 49: 76-81.
11 Gauderman, W. J.; Avol, E.; Lurmann, F.; Kuenzli, N.; Gilliland, F.; Peters, J.; McConnell, R.
12 (2005) Childhood asthma and exposure to traffic and nitrogen dioxide. Epidemiology 16:
13 737-743.
14 Gauvin, S.; Le Moullec, Y.; Bremont, F.; Momas, I; Balducci, F.; Ciognard, F.; Poilve, M.-P.;
15 Zmirou, D.; VESTA Investigators. (2001) Relationships between nitrogen dioxide
16 personal exposure and ambient air monitoring measurements among children in three
17 French metropolitan areas: VESTA study. Arch. Environ. Health 56: 336-341.
18 Georgopoulos, P. G.; Lioy, P. J. (1994) Conceptual and theoretical aspects of human exposure
19 and dose assessment. J. Exposure Anal. Environ. Epidemiol. 4: 253-285.
20 Georgopoulos, P. G.; Lioy, P. J. (2006) From theoretical aspects of human exposure and dose
21 assessment to computational model implementation: the Modeling Environment for
22 TOtal Risk studies (MENTOR). J. Toxicol. Environ. Health Part B 9: 457-483.
23 Georgopoulos, P. G.; Seinfeld, J. H. (1989) Nonlocal description of turbulent dispersion. Chem.
24 Eng. Sci. 44: 1995-2016.
25 Georgopoulos, P. G.; Arunachalam, S.; Wang, S. (1997a) Alternative metrics for assessing the
26 relative effectiveness of NOX and VOC emission reductions in controlling ground-level
27 ozone. J. Air Waste Manage. Assoc. 47: 838-850.
28 Georgopoulos, P. G.; Purushothaman, V.; Chiou, R. (1997b) Comparative evaluation of methods
29 for estimating potential human exposure to ozone: photochemical modeling and ambient
30 monitoring. J. Exposure Anal. Environ. Epidemiol. 7: 191-215.
31 Georgopoulos, P. G.; Walia, A.; Roy, A.; Lioy, P. J. (1997c) Integrated exposure and dose
32 modeling and analysis system. 1. Formulation and testing of microenvironmental and
33 pharmacokinetic components. Environ. Sci. Technol. 31: 11-21.
34 Georgopoulos, P. G.; Wang, S.-W.; Vyas, V. M.; Sun, Q.; Burke, J.; Vedantham, R.; McCurdy,
35 T.; Ozkaynak, H. (2005) A source-to-dose assessment of population exposures to fine
36 PM and ozone in Philadelphia, PA, during a summer 1999 episode. J. Exposure Anal.
37 Environ. Epidemiol. 15: 439-457.
38 Gerboles, M.; Buzica, D.; Amantini, L.; Lagler, F. (2006a) Laboratory and field comparison of
39 measurement obtained using the available diffusive samplers for ozone and nitrogen
40 dioxide in ambient air. J. Environ. Monit. 8: 112-119.
March 2008 AX3-176 DRAFT-DO NOT QUOTE OR CITE
-------
1 Gerboles, M.; Buzica, D.; Amantini, L.; Lagler, F.; Hafkenscheid, T. (2006b) Feasibility study of
2 preparation and certification of reference materials for nitrogen dioxide and sulfur
3 dioxide in diffusive samplers. J. Environ. Monit. 8: 174-182.
4 Gilbert, N. L.; Goldberg, M. S.; Beckerman, B.; Brook, J. R.; Jerrett, M. (2005) Assessing spatial
5 variability of ambient nitrogen dioxide in Montreal, Canada, with a land-use regression
6 model. J. Air Waste Manage. Assoc. 55: 1059-1063.
7 Gilbert, N. L.; Gauvin, D.; Guay, M.; Heroux, M.-E.; Dupuis, G.; Legris, M.; Chan, C. C.; Dietz,
8 R. N.; Levesque, B. (2006) Housing characteristics and indoor concentrations of nitrogen
9 dioxide and formaldehyde in Quebec City, Canada. Environ. Res. 102: 1-8.
10 Gillani, N. V.; Godowitch, J. M. (1999) Plume-in-grid treatment of major point source
11 emissions. In: Byun, D. W.; Ching, J. K. S., eds. Science algorithms of the EPA models-3
12 community multiscale air quality modeling system; EPA/600/R-99/030. Research
13 Triangle Park, NC: U.S. Environmental Protection Agency, National Exposure Research
14 Laboratory; chapter 9 . Available: http://www.epa.gov/asmdnerl/CMAQ/ch09.pdf [6
15 March, 2007].
16 Girman, J. R.; Apte, M. G.; Traynor, G. W.; Allen, J. R.; Hollowell, C. D. (1982) Pollutant
17 emission rates from indoor combustion appliances and sidestream cigarette smoke.
18 Environ. Int. 8:213-221.
19 Glas, B.; Levin, J.-O.; Stenberg, B.; Stenlund, H.; Sunesson, A.-L. (2004) Variability of personal
20 chemical exposure in eight office buildings in Sweden. J. Exposure Anal. Environ.
21 Epidemiol. 14(suppl.): S49-S57.
22 Glasius, M.; Carlsen, M. F.; Hansen, T. S.; Lohse, C. (1999) Measurements of nitrogen dioxide
23 on Funen using diffusion tubes. Atmos. Environ. 33: 1177-1185.
24 Goldan, P. D.; Trainer, M.; Kuster, W. C.; Parrish, D. D.; Carpenter, J.; Roberts, J. M.; Yee, J.
25 E.; Fehsenfeld, F. C. (1995) Measurements of hydrocarbons, oxygenated hydrocarbons,
26 carbon monoxide, and nitrogen oxides in an urban basin in Colorado: implications for
27 emission inventories. J. Geophys. Res. [Atmos.] 100: 22,771-22,783.
28 Gonzales, M.; Quails, C.; Hudgens, E.; Neas, L. (2005) Characterization of a spatial gradient of
29 nitrogen dioxide across a United States-Mexico border city during winter. Sci. Total
30 Environ. 337: 163-173.
31 Gradko International Limited. (2007) Technical data sheet: TDS 1. DIF 100 RTU - nitrogen
32 dioxide (NC^). Winchester, England: Gradko Environmental. Available:
33 http://www.gradko.co.uk/pdf/Nitrogen_Dioxide.pdf [15 August, 2007].
34 Grosjean, D. (2003) Ambient PAN and PPN in southern California from 1960 to the SCOS97-
35 NARSTO. Atmos. Environ. 37(suppl. 2): S221-S238.
36 Gustafsson, L. E.; Leone, A. M.; Persson, M. G.; Wiklund, N. P.; Moncada, S. (1991)
37 Endogenous nitric oxide is present in the exhaled air of rabbits, guinea pigs and humans.
38 Biochem. Biophys. Res. Commun. 181: 852-857.
39 Hackney, J. D.; Linn, W. S.; Avol, E. L.; Shamoo, D. A.; Anderson, K. R.; Solomon, J. C.;
40 Little, D. E.; Peng, R.-C. (1992) Exposures of older adults with chronic respiratory illness
March 2008 AX3-177 DRAFT-DO NOT QUOTE OR CITE
-------
1 to nitrogen dioxide: a combined laboratory and field study. Am. Rev. Respir. Dis. 146:
2 1480-1486.
3 Hagenbjork-Gustafsson, A.; Forsberg, B.; Hestvik, G.; Karlsson, D.; Wahlberg, S.; Sandstrom,
4 T. (1996) Measurements of indoor and outdoor nitrogen dioxide concentrations using a
5 diffusive sampler. Analyst (Cambridge, U.K.) 121: 1261-1264.
6 Hagenbjork-Gustafsson, A.; Lindahl, R.; Levin, J. O.; Karlsson, D. (1999) Validation of a
7 diffusive sampler for NO2. J. Environ. Monit. 1: 349-352.
8 Hagenbjork-Gustafsson, A.; Lindahl, R.; Levin, J.-O.; Karlsson, D. (2002) Validation of the
9 Willems badge diffusive sampler for nitrogen dioxide determinations in occupational
10 environments. Analyst (Cambridge, U.K.) 127: 163-168.
11 Hayes, S. R. (1989) Estimating the effect of being indoors on total personal exposure to outdoor
12 air pollution. JAPCA 39: 1453-1461.
13 Hayes, S. R. (1991) Use of an indoor air quality model (IAQM) to estimate indoor ozone levels.
14 J. Air Waste Manage. Assoc. 41: 161-170.
15 Hazenkamp-von Arx, M. E.; Gotschi, T.; Ackermann-Liebrich, U.; Bono, R.; Burney, P.; Cyrys,
16 J.; Jarvis, D.; Lillienberg, L.; Luczynska, C.; Maldonado, J. A.; Jaen, A.; de Marco, R.;
17 Mi, Y.; Modig, L.; Bayer-Oglesby, L.; Payo, F.; Soon, A.; Sunyer, J.; Villani, S.; Weyler,
18 J.; Kiinzli, N. (2004) PM2.s and NO2 assessment in 21 European study centres of ECRHS
19 II: annual means and seasonal differences. Atmos. Environ. 38: 1943-1953.
20 Heal, M. R.; O'Donoghue, M. A.; Cape, J. N. (1999) Overestimation of urban nitrogen dioxide
21 by passive diffusion tubes: a comparative exposure and model study. Atmos. Environ. 33:
22 513-524.
23 Hertel, O.; Jensen, S. S.; Andersen, H. V.; Palmgren, F.; Wahlin, P.; Skov, H.; Nielsen, I. V.;
24 S0rensen, M.; Loft, S.; Raaschou-Nielsen, O. (2001) Human exposure to traffic pollution
25 experience from Danish studies. Pure Appl. Chem. 73: 137-145.
26 Hochadel, M.; Heinrich, J.; Gehring, U.; Morgenstern, V.; Kuhlbusch, T.; Link, E.; Wichmann,
27 H.-E.; Kramer, U. (2006) Predicting long-term average concentrations of traffic-related
28 air pollutants using GIS-based information. Atmos. Environ. 40: 542-553.
29 Horowitz, L. W.; Liang, J.; Gardner, G. M.; Jacob, D. J . (1998) Export of reactive nitrogen from
30 North America during summertime: sensitivity to hydrocarbon chemistry. J. Geophys.
31 Res. [Atmos.] 103(D11): 13451-13476.
32 Imada, M.; Iwamoto, J.; Nonaka, S.; Kobayashi, Y.; Unno, T. (1996) Measurement of nitric
33 oxide in human nasal airway. Eur. Respir. J. 9: 556-559.
34 ICF Consulting. (2005) Decision points for consulting APEX for air toxics exposure
35 assessments. Research Triangle Park, NC: U.S. Environmental Protection Agency, Office
36 of Air Quality Planning and Standards. Available:
3 7 http ://www.epa.gov/ttnmain l/fera/data/apex/APEXDecisionPoints_l 21205 .pdf [ 19
38 September, 2007].
39 Isukapalli, S. S.; Purushothaman, V.; Georgopoulos, P. G. (1999) Mechanistic modeling of the
40 interrelationships between indoor/outdoor air quality and human exposure in a GIS
March 2008 AX3-178 DRAFT-DO NOT QUOTE OR CITE
-------
1 framework. Presented at: 92nd Air & Waste Management Association Annual Meeting;
2 June; St. Louis, MO. Pittsburgh, PA: Air & Waste Management Association.
3 Jakobi, G.; Fabian, P. (1997) Indoor/outdoor concentrations of ozone and peroxyacetyl nitrate
4 (PAN). Int. J. Biometeorol. 40: 162-165.
5 Janssen, N. A. H.; van Vliet, P. H. N.; Aarts, F.; Harssema, H.; Brunekreef, B. (2001)
6 Assessment of exposure to traffic related air pollution of children attending schools near
7 motorways. Atmos. Environ. 35: 3875-3884.
8 Jarvis, D. L.; Leaderer, B. P.; Chinn, S.; Burney, P. G. (2005) Indoor nitrous acid and respiratory
9 symptoms and lung function in adults. Thorax 60: 474-479.
10 Jet Propulsion Laboratory. (2006) Chemical kinetics and photochemical data for use in
11 atmospheric studies. Evaluation number 15. Pasadena, CA: California Institute of
12 Technology. JPL publication 06-2.
13 Johnson, T. (2001) A guide to selected algorithms, distributions, and databases used in exposure
14 models developed by the Office of Air Quality Planning and Standards (DRAFT).
15 Research Triangle Park, NC: U.S. Environmental Protection Agency, CERM report
16 TR01.
17 Johnson, T. (2002) A guide to selected algorithms, distributions, and databases used in exposure
18 models developed by the Office of Air Quality Planning and Standards [revised draft].
19 Research Triangle Park, NC: U.S. Environmental Protection Agency, Office of Research
20 and Development; EPA grant no. CR827033. Available:
21 http://www.epa.gov/ttn/fera/data/human/report052202.pdf [6 March, 2007].
22 Johnson, W. B.; Ludwig, F. L.; Dabbert, W. F.; Allen, R. J. (1973) An urban diffusion simulation
23 model for carbon monoxide. J. Air Pollut. Control Assoc. 23: 490-498.
24 Johnson, T.; Long, T.; Ollison, W. (2000) Prediction of hourly microenvironmental
25 concentrations of fine particles based on measurements obtained from the Baltimore
26 scripted activity study. J. Exposure Anal. Environ. Epidemiol. 10: 403-411.
27 Kameda, T.; Takenaka, N.; Bandow, H.; Inazu, K.; Hisamatsu, Y. (2004) Determination of
28 atmospheric nitro-polycyclic aromatic hydrocarbons and their precursors at a heavy
29 traffic roadside and at a residential area in Osaka, Japan. Poly cyclic Aromat. Compd. 24:
30 657-666.
31 Kawamoto, T.; Yoshikawa, M.; Matsuno, K.; Kayama, F.; Oyama, T.; Arashidani, K.; Kodama,
32 Y. (1993) Effect of side-stream cigarette smoke on the hepatic cytochrome P450. Arch.
33 Environ. Contam. Toxicol. 25: 255-259.
34 Kawamoto, T.; Matsuno, K.; Arashidani, K.; Kodama, Y. (1997) Personal exposure to indoor
35 nitrogen dioxide. In: Subramanian, K. S.; lyengar, G. V., eds. Environmental
36 biomonitoring: exposure assessment and specimen banking. Washington, DC: American
37 Chemical Society; pp. 178-182. (ACS symposium series: v. 654).
38 Ketzel, M.; Louka, P.; Sahm, P.; Guilloteau, E.; Sini, J.-F.; Moussiopoulos, N. (2002)
39 Intercomparison of numerical urban dispersion models - Part II: street canyon in
40 Hannover, Germany. Water Air Soil Pollut. Focus 2: 603-613.
March 2008 AX3-179 DRAFT-DO NOT QUOTE OR CITE
-------
1 Kim, E.; Hopke, P. K.; Pinto, J. P.; Wilson, W. E. (2005) Spatial variability of fine particle mass,
2 components, and source contributions during the regional air pollution study in St. Louis.
3 Environ. Sci. Technol. 39: 4172-4179.
4 Kim, D.; Sass-Kortsak, A.; Purdham, J. T.; Dales, R. E.; Brook, J. R. (2006) Associations
5 between personal exposures and fixed-site ambient measurements of fine particulate
6 matter, nitrogen dioxide, and carbon monoxide in Toronto, Canada. J. Exposure Sci.
7 Environ. Epidemiol. 16: 172-183.
8 Kirby, C.; Greig, A.; Drye, T. (1998) Temporal and spatial variations in nitrogen dioxide
9 concentrations across an urban landscape: Cambridge, UK. Environ. Monit. Assess. 52:
10 65-82.
11 Kirby, C.; Fox, M.; Waterhouse, J.; Drye, T. (2001) Influences of environmental parameters on
12 the accuracy of nitrogen dioxide passive diffusion tubes for ambient measurement. J.
13 Environ. Monit. 3: 150-158.
14 Klepeis, N. E.; Nelson, W. C.; Ott. W. R.; Robinson, J. P. Tsang, A. M.; Switzer, P.; Behar, J.
15 V.; Hern, S. C.; Engelmann, W. H. (2001) The National Human Activity Pattern Survey
16 (NHAPS): a resource for assessing exposure to environmental pollutants. J. Exposure
17 Anal. Environ. Epidemiol. 11: 231-252.
18 Kodama, Y.; Arashidani, K.; Tokui, N.; Kawamoto, T.; Matsuno, K.; Kunugita, N.; Minakawa,
19 N. (2002) Environmental NO2 concentration and exposure in daily life along main roads
20 in Tokyo. Environ. Res. A 89: 236-244.
21 Kotchenruther, R. A.; Jaffe, D. A.; Beine, H. J.; Anderson, T. L.; Bottenheim, J. W.; Harris, J.
22 M.; Blake, D. R.; Schmitt, R. (2001) Observations of ozone and related species in the
23 northeast Pacific during the PHOBEA campaigns: 2. Airborne observations. J. Geophys.
24 Res. (Atmos.) 106: 7463-7483.
25 Kousa, A.; Monn, C.; Rotko, T.; Aim, S.; Oblesby, L.; Jantunen, M. J. (2001) Personal exposures
26 to NO2 in the EXPOLIS-study: relation to residential indoor, outdoor and workplace
27 concentrations in Basel, Helsinki and Prague. Atmos. Environ. 35: 3405-3412.
28 Kraenzmer, M. (1999) Modeling and continuous monitoring of indoor air pollutants for
29 identification of sources and sinks. Environ. Int. 25: 541-551.
30 Kramer, U.; Koch, T.; Ranft, U.; Ring, J.; Behrendt, H. (2000) Traffic-related air pollution is
31 associated with atopy in children living in urban areas. Epidemiology 11: 64-70.
32 Krupa, S. V.; Legge, A. H. (2000) Passive sampling of ambient, gaseous air pollutants: an
33 assessment from an ecological perspective. Environ. Pollut. 107: 31-45.
34 Kulkarni, M. M.; Patil, R. S. (2002) An empirical model to predict indoor NO2 concentrations.
35 Atmos. Environ. 36: 4777-4785.
36 Lai, H. K.; Kendall, M.; Ferrier, H.; Lindup, L; Aim, S.; Hanninen, O.; Jantunen, M.; Mathys, P.;
37 Colvile, R.; Ashmore, M. R.; Cullinan, P.; Nieuwenhuijsen, M. J. (2004) Personal
38 exposures and microenvironment concentrations of PM2.5, VOC, NO2 and CO in Oxford,
39 UK. Atmos. Environ. 38: 6399-6410.
40 Lai, H. K.; Bayer-Oglesby, L.; Colvile, R.; Gotschi, T.; Jantunen, M. J.; Kiinzli, N.; Kulinskaya,
41 E.; Schweizer, C.; Nieuwenhuij sen, M. J. (2006) Determinants of indoor air
March 2008 AX3-180 DRAFT-DO NOT QUOTE OR CITE
-------
1 concentrations of PM2.5, black smoke and NO2 in six European cities (EXPOLIS study).
2 Atmos. Environ. 40: 1299-1313.
3 Lai, S.; Patil, R. S. (2001) Monitoring of atmospheric behaviour of NOx from vehicular traffic.
4 Environ. Monit. Assess. 68: 37-50.
5 Lammel, G.; Cape, J. N. (1996) Nitrous acid and nitrite in the atmosphere. Chem. Soc. Rev. 25:
6 361-369.
7 Langstaff, J. E. (2007) Analysis of uncertainty in ozone population exposure modeling [technical
8 memorandum]. Research Triangle Park, NC: U.S. Environmental Protection Agency,
9 Office of Air Quality Planning and Standards.
10 Leaderer, B. P.; Zagraniski, R. T.; Berwick, M.; Stolwijk, J. A. J. (1986) Assessment of exposure
11 to indoor air contaminants from combustion sources: methodology and application. Am.
12 J. Epidemiol. 124: 275-289.
13 Lee, S.-C. (1997) Comparison of indoor and outdoor air quality at two staff quarters in Hong
14 Kong. Environ. Int. 23: 791-797.
15 Lee, K.; Yanagisawa, Y.; Spengler, J. D.; Ozkaynak, H.; Billick, I. H. (1993a) Sampling rate
16 evaluation of NO2 badge: (I) in indoor environments. Indoor Air 3: 124-130.
17 Lee, K.; Yanagisawa, Y.; Spengler, J. D.; Ozkaynak, H.; Billick, I. H. (1993b) Sampling rate
18 evaluation for NO2 badge: (II) in personal monitoring. Environ. Int. 19: 3-7.
19 Lee, K.; Yanagisawa, Y.; Spengler, J. D.; Nakai, S. (1994) Carbon monoxide and nitrogen
20 dioxide exposures in indoor ice skating rinks. J. Sports Sci. 12: 279-283.
21 Lee, K.; Yanagisawa, Y.; Spengler, J. D.; Billick, I. H. (1995) Classification of house
22 characteristics based on indoor nitrogen dioxide concentrations. Environ. Int. 21: 277-
23 282.
24 Lee, K.; Yanagisawa, Y.; Spengler, J. D.; Fukumura, Y.; Billick, I. H. (1996) Classification of
25 house characteristics in a Boston residential nitrogen dioxide characterization study.
26 Indoor Air 6: 211-216.
27 Lee, K.; Levy, J. I; Yanagisawa, Y.; Spengler, J. D.; Billick, I. H. (1998) The Boston Residential
28 Nitrogen Dioxide Characterization Study: classification and prediction of indoor NO2
29 exposure. J. Air Waste Manage. Assoc. 48: 736-742.
30 Lee, S. C.; Chan, L. Y.; Chiu, M. Y. (1999) Indoor and outdoor air quality investigation at 14
31 public places in Hong Kong. Environ. Int. 25: 443-450.
32 Lee, K.; Yang, W.; Bofinger, N. D. (2000) Impact of microenvironmental nitrogen dioxide
33 concentrations on personal exposures in Australia. J. Air Waste Manage. Assoc. 50:
34 1739-1744.
35 Lee, K.; Xue, J.; Geyh, A. S.; Ozkaynak, H.; Leaderer, B. P.; Weschler, C. J.; Spengler, J. D.
36 (2002) Nitrous acid, nitrogen dioxide, and ozone concentrations in residential
37 environments. Environ. Health Perspect. 110: 145-150.
38 Lee, K.; Bartell, S. M.; Paek, D. (2004) Interpersonal and daily variability of personal exposures
39 to nitrogen dioxide and sulfur dioxide. J. Exposure Anal. Environ. Epidemiol. 14: 137-
40 143.
March 2008 AX3-181 DRAFT-DO NOT QUOTE OR CITE
-------
1 Levesque, B.; Allaire, S.; Prud'homme, H.; Dupuis, K.; Bellemare, D. (2000) Air quality
2 monitoring during indoor monster truck and car demolition shows. J. Exposure Anal.
3 Environ. Epidemiol. 10: 58-65.
4 Levesque, B.; Allaire, S.; Gauvin, D.; Koutrakis, P.; Gingras, S.; Rhainds, M.; Prud'Homme, H.;
5 Duchesne, J.-F. (2001) Wood-burning appliances and indoor air quality. Sci. Total
6 Environ. 281: 47-62.
7 Levy, J. I.; Lee, K.; Spengler, J. D.; Yanagisawa, Y. (1998a) Impact of residential nitrogen
8 dioxide exposure on personal exposure: an international study. J. Air Waste Manage.
9 Assoc. 48: 553-560.
10 Levy, J. I; Lee, K.; Yanagisawa, Y.; Hutchinson, P.; Spengler, J. D. (1998b) Determinants of
11 nitrogen dioxide concentrations in indoor ice skating rinks. Am. J. Public Health 88:
12 1781-1786.
13 Levy, J. L.; Baxter, L. K.; Clougherty, J. E. (2006) The air quality impacts of road closures
14 associated with the 2004 Democratic National Convention in Boston. Environ. Health:
15 Global Access Sci. Source 5: 16.
16 Lewne, M.; Cyrys, J.; Meliefste, K.; Hoek, G.; Brauer, M.; Fischer, P.; Gehring, U.; Heinrich, J.;
17 Brunekreef, B.; Bellander, T. (2004) Spatial variation in nitrogen dioxide in three
18 European areas. Sci. Total Environ. 332: 217-230.
19 Lewne, M.; Nise, G.; Lind, M. L.; Gustavsson, P. (2006) Exposure to particles and nitrogen
20 dioxide among taxi, bus and lorry drivers. Int. Arch. Occup. Environ. Health 79: 220-226.
21 Liard, R.; Zureik, M.; Le Moullec, Y.; Soussan, D.; Glorian, M.; Grimfeld, A.; Neukirch, F.
22 (1999) Use of personal passive samplers for measurement of NO2, NO, and O3 levels in
23 panel studies. Environ. Res. 81: 339-348.
24 Linaker, C. H.; Chauhan, A. J.; Inskip, H.; Frew, A. J.; Sillence, A.; Coggon, D.; Holgate, S. T.
25 (1996) Distribution and determinants of personal exposure to nitrogen dioxide in school
26 children. Occup. Environ. Med. 53: 200-203.
27 Linaker, C. H.; Chauhan, A. J.; Inskip, H. M.; Holgate, S. T.; Coggon, D. (2000) Personal
28 exposures of children to nitrogen dioxide relative to concentrations in outdoor air. Occup.
29 Environ. Med. 57: 472-476.
30 Linn, W. S.; Shamoo, D. A.; Anderson, K. R.; Peng, R.-C.; Avol, E. L.; Hackney, J. D.; Gong,
31 H., Jr. (1996) Short-term air pollution exposures and responses in Los Angeles area
32 schoolchildren. J. Exposure Anal. Environ. Epidemiol. 6: 449-472.
33 Lioy, P. J. (1990) Assessing total human exposure to contaminants. Environ. Sci. Technol. 24:
34 938-945.
35 Mage, D.; Wilson, W.; Hasselblad, V.; Grant, L. (1999) Assessment of human exposure to
36 ambient particulate matter. J. Air Waste Manage. Assoc. 49: 1280-1291.
37 Marshall, J. D.; Behrentz, E. (2005) Vehicle self-pollution intake fraction: children's exposure to
38 school bus emissions. Environ. Sci. Technol. 39: 2559-2563.
March 2008 AX3-182 DRAFT-DO NOT QUOTE OR CITE
-------
1 Martin, R. V.; Jacob, D. I; Chance, K. V.; Kurosu, T. P.; Palmer, P. I; Evans, M. J. (2003)
2 Global inventory of nitrogen oxide emissions constrained by space-based observations of
3 NO2 columns. J. Geophys. Res. [Atmos.] 108(D17): 10.1029/2003JD003453.
4 McCurdy, T. (1997a) Human activities that may lead to high inhaled intake doses in children
5 aged 6-13. Environ. Toxicol. Pharmacol. 4: 251-260.
6 McCurdy, T. (1997b) Modeling the dose profile in human exposure assessments: ozone as an
7 example. Rev. Toxicol. 1: 3-23.
8 McCurdy, T. (2000) Conceptual basis for multi-route intake dose modeling using an energy
9 expenditure approach. J. Exposure Anal. Environ. Epidemiol. 10: 86-97.
10 McCurdy, T.; Glen, G.; Smith, L.; Lakkadi, Y. (2000) The National Exposure Research
11 Laboratory's Consolidated Human Activity Database. J. Exposure Anal. Environ.
12 Epidemiol. 10: 566-578.
13 Meng, Q. Y.; Turpin, B. J.; Polidori, A.; Lee, J. H.; Weisel, C.; Morandi, M.; Colome, S.; Stock,
14 T.; Winer, A.; Zhang, J. (2005) PM25 of ambient origin: estimates and exposure errors
15 relevant to PM epidemiology. Environ. Sci. Technol. 39: 5105-5112.
16 Miller, F. J.; Overton, J. H.; Myers, E. T.; Graham, J. A. (1982) Pulmonary dosimetry of nitrogen
17 dioxide in animals and man. In: Schneider, T.; Grant, L., eds. Air pollution by nitrogen
18 oxides: proceedings of the US-Dutch international symposium; May; Maastricht, The
19 Netherlands. Amsterdam, The Netherlands: Elsevier Scientific Publishing Company; pp.
20 377-386. (Studies in environmental science 21).
21 Milner, J. T.; Dimitroulopoulou, C.; ApSimon, H. (2005) Indoor concentrations in buildings
22 from sources outdoors. In: Atmospheric Dispersion Modelling Liaison Committee
23 Annual Report 2004-2005, Annex B. Available: http:www.admlc.org.uk/ar04-05.htm [6
24 March, 2007].
25 Modig, L.; Sunesson, A.-L.; Levin, J.-O.; Sundgren, M.; Hagenbjork-Gustafsson, A.; Forsberg,
26 B. (2004) Can NO2 be used to indicate ambient and personal levels of benzene and 1,3-
27 butadiene in air? J. Environ. Monit. 6: 957-962.
28 Monn, C. (2001) Exposure assessment of air pollutants: a review on spatial heterogeneity and
29 indoor/outdoor/personal exposure to suspended particulate matter, nitrogen dioxide and
30 ozone. Atmos. Environ. 35: 1-32.
31 Monn, C.; Fuchs, A.; Hogger, D.; Junker, M.; Kogelschatz, D.; Roth, N.; Wanner, H.-U. (1997)
32 Particulate matter less than 10|i (PMio) and fine particles less than 2.5|i (PM2.5):
33 relationships between indoor, outdoor and personal concentrations. Sci. Total Environ.
34 208: 15-21.
35 Monn, C.; Brandli, O.; Schindler, C.; Ackermann-Liebrich, U.; Leuenberger, P.; SAPALDIA
36 team. (1998) Personal exposure to nitrogen dioxide in Switzerland. Sci. Total Environ.
37 215:243-251.
38 Mosqueron, L.; Momas, L; Le Moullec, Y. (2002) Personal exposure of Paris office workers to
39 nitrogen dioxide and fine particles. Occup. Environ. Med. 59: 550-555.
March 2008 AX3-183 DRAFT-DO NOT QUOTE OR CITE
-------
1 Mourgeon, E.; Levesque, E.; Duveau, C.; Law-Koune, J.-D.; Charbit, B.; Ternissien, E.; Coriat,
2 P.; Rouby, J-J. (1997) Factors influencing indoor concentrations of nitric oxide in a
3 Parisian intensive care unit. Am. J. Respir. Crit. Care Med. 156: 1692-1695.
4 Mukala, K.; Pekkanen, J.; Tiittanen, P.; Aim, S.; Salonen, R. O.; Jantunen, M.; Tuomisto, J.
5 (1996) Seasonal exposure to NC>2 and respiratory symptoms in preschool children. J.
6 Exposure Anal. Environ. Epidemiol. 6: 197-210.
7 Mukala, K.; Aim, S.; Tiittanen, P.; Salonen, R. O.; Jantunen, M.; Pekkanen, J. (2000) Nitrogen
8 dioxide exposure assessment and cough among preschool children. Arch. Environ.
9 Health. 55:431-438.
10 Mumford, J. L.; Williams, R. W.; Walsh, D. B.; Burton, R. M.; Svendsgaard, D. J.; Chuang, J.
11 C.; Houk, V. S.; Lewtas, J. (1991) Indoor air pollutants from unvented kerosene heater
12 emissions in mobile homes: studies on particles, semivolatile organics, carbon monoxide,
13 and mutagenicity. Environ. Sci. Technol. 25: 1732-1738.
14 Munger, J. W.; Wofsy, S. C.; Bakwin, P. S.; Fan, S.-M.; Goulden, M. L.; Daube, B. C.;
15 Goldstein, A. H. (1996) Atmospheric deposition of reactive nitrogen oxides and ozone in
16 a temperate deciduous forest and a subarctic woodland. 1. Measurements and
17 mechanisms. J. Geophys. Res. 101: 12639-12657.
18 Nakai, S.; Nitta, H.; Maeda, K. (1995) Respiratory health associated with exposure to automobile
19 exhaust II. Personal NC>2 exposure levels according to distance from the roadside. J.
20 Exposure Anal. Environ. Epidemiol. 5: 125-136.
21 Namiesnik, J.; Zabiegala, B.; Kot-Wasik, A.; Partyka, M.; Wasik, A. (2005) Passive sampling
22 and/or extraction techniques in environmental analysis: a review. Anal. Bioanal. Chem.
23 381:279-301.
24 National Research Council. (1991) Rethinking the ozone problem in urban and regional air
25 pollution. Washington, DC: National Academy Press. Available:
26 http://www.nap.edu/books/0309046319/html/ [26 March, 2004].
27 Naumova, Y. Y.; Eisenreich, S. J.; Turpin, B. J.; Weisel, C. P.; Morandi, M. T.; Colome, S. D.;
28 Totten, L. A.; Stock, T. H.; Winer, A. M.; Alimokhtari, S.; Kwon, J.; Shendell, D.; Jones,
29 J.; Maberti, S.; Wall, S. J. (2002) Poly cyclic aromatic hydrocarbons in the indoor and
30 outdoor air of three cities in the U.S. Environ. Sci. Technol. 36: 2552-2559.
31 Naumova, Y. Y.; Offenberg, J. H.; Eisenreich, S. J.; Meng, Q. Y.; Polidori, A.; Turpin, B. J.;
32 Weisel, C. P.; Morandi, M. T.; Colome, S. D.; Stock, T. H.; Winer, A. M.; Alimokhtari,
33 S.; Kwon, J.; Maberti, S.; Shendell, D.; Jones, J.; Farrar, C. (2003) Gas/particle
34 distribution of poly cyclic aromatic hydrocarbons in coupled outdoor/indoor atmospheres.
35 Atmos. Environ. 37: 703-719.
36 Nazaroff, W. W.; Cass, G. R. (1986) Mathematical modeling of chemically reactive pollutants in
37 indoor air. Environ. Sci. Technol. 20: 924-934.
38 Nazaroff, W. W.; Weschler, C. J. (2004) Cleaning products and air fresheners: exposure to
39 primary and secondary air pollutants. Atmos. Environ. 38: 2841-2865.
40
41
March 2008 AX3-184 DRAFT-DO NOT QUOTE OR CITE
-------
1 Nerriere, E.; Zmirou-Navier, D.; Blanchard, O.; Momas, I; Ladner, J.; Le Moullec, Y.;
2 Personnaz, M.-B.; Lameloise, P.; Delmas, V.; Target, A.; Desqueyroux, H. (2005) Can
3 we use fixed ambient air monitors to estimate population long-term exposure to air
4 pollutants? The case of spatial variability in the Genotox ER study. Environ. Res. 97: 32-
5 42.
6 N0jgaard, J. K.; Bilde, M.; Stenby, C.; Nielsen, O. J.; Wolkoff, P. (2006) The effect of nitrogen
7 dioxide on particle formation during ozonolysis of two abundant monoterpenes indoors.
8 Atmos. Environ. 40: 1030-1042.
9 Norris, G.; Larson, T. (1999) Spatial and temporal measurements of NC>2 in an urban area using
10 continuous mobile monitoring and passive samplers. J. Exposure Anal. Environ.
11 Epidemiol. 9: 586-593.
12 Ogawa & Company. (1998) NO, NC>2, NOX and SC>2 sampling protocol using the Ogawa
13 sampler. Version 3. Pompano Beach, FL: Ogawa & Co., USA, Inc. Available:
14 http://www.ogawausa.com/protocol.html [30 March, 2007].
15 Ott, W. R. (1982) Concepts of human exposure to air pollution. Environ. Int. 7: 179-196.
16 Ott, W. R. (1985) Total human exposure: an emerging science focuses on humans as receptors of
17 environmental pollution. Environ. Sci. Technol. 19: 880-886.
18 Ott, W.; Wallace, L.; Mage, D. (2000) Predicting particulate (PMi0) personal exposure
19 distributions using a random component superposition statistical model. J. Air Waste
20 Manage. Assoc. 50: 1390-1406.
21 Overton, J. H.; Graham, R. C. (1995) Simulation of the uptake of a reactive gas in a rat
22 respiratory tract model with an asymmetric tracheobronchial region patterned on
23 complete conducting airway cast data. Comput. Biomed. Res. 28: 171-190.
24 Palmes, E. D.; Gunnison, A. F.; DiMattio, J.; Tomczyk, C. (1976) Personal sampler for nitrogen
25 dioxide. Am. Ind. Hyg. Assoc. J. 37: 570-577.
26 Pandian, M. D.; Behar, J. V.; Ott, W. R.; Wallace, L. A.; Wilson, A. L.; Colome, S, D. (1998)
27 Correcting errors in the nationwide data base of residential air exchange rates. J.
28 Exposure Anal. Environ. Epidemiol. 8: 577-586.
29 Park, J.-H.; Spengler, J. D.; Yoon, D.-W.; Dumyahn, T.; Lee, K.; Ozkaynak, H. (1998)
30 Measurement of air exchange rate of stationary vehicles and estimation of in-vehicle
31 exposure. J. Exposure Anal. Environ. Epidemiol. 8: 65-78.
32 Partti-Pellinen, K.; Marttila, O.; Ahonen, A.; Suominen, O.; Haahtela, T. (2000) Penetration of
33 nitrogen oxides and particles from outdoor into indoor air and removal of the pollutants
34 through filtration of incoming air. Indoor Air. 10: 126-132.
35 Passam ag. (2007) Nitrogen oxides NOX [sampler]. Mannedorf, Switzerland: Passam ag,
36 Laboratory for Environmental Analysis. Available: http://www.passam.ch/nox.htm [15
37 August, 2007].
38 Paulozzi, L. J.; Spengler, R. F.; Vogt, R. L.; Carney, J. K. (1993) A survey of carbon monoxide
39 and nitrogen dioxide in indoor ice arenas in Vermont. J. Environ. Health 56(5): 23-25.
March 2008 AX3-185 DRAFT-DO NOT QUOTE OR CITE
-------
1 Piechocki-Minguy, A.; Plaisance, H.; Garcia-Fouque, S.; Galloo, J. C.; Guillermo, R. (2003)
2 Validation tests of a new high uptake rate passive sampler for nitrogen dioxide
3 measurements. Environ. Technol. 24: 1527-1535.
4 Piechocki-Minguy, A.; Plaisance, H.; Schadkowski, C.; Sagnier, I; Saison, J. Y.; Galloo, J. C.;
5 Guillermo, R. (2006) A case study of personal exposure to nitrogen dioxide using a new
6 high sensitive diffusive sampler. Sci. Total Environ. 366: 55-64.
7 Pilotto, L. S.; Douglas, R. M.; Attewell, R. G.; Wilson, S. R. (1997) Respiratory effects
8 associated with indoor nitrogen dioxide exposure in children. Int. J. Epidemiol. 26: 788-
9 796.
10 Pinto, J. P.; Lefohn, A. S.; Shadwick, D. S. (2004) Spatial variability of PM2.5 in urban areas in
11 the United States. J. Air Waste Manage. Assoc. 54: 440-449.
12 Pippin, M. R.; Bertman, S.; Thornberry, T.; Town, M.; Carroll, M. A.; Sillman, S. (2001)
13 Seasonal variation of PAN, PPN, and O3 at the upper Midwest PROPHET site. J.
14 Geophys. Res. (Atmos.) 106: 24,451-24,463.
15 Plaisance, H.; Piechocki-Minguy, A.; Garcia-Fouque, S.; Galloo, J. C. (2004) Influences of
16 meteorological factors on the NO2 measurements by passive diffusion tube. Atmos.
17 Environ. 38: 573-580.
18 Pleijel, H.; Karlsson, G. P.; Gerdin, E. B. (2004) On the logarithmic relationship between NO2
19 concentration and the distance from a highroad. Sci. Total Environ. 332: 261-264.
20 Postlethwait, E. M.; Langford, S. D.; Bidani, A. (1991) Transfer of NO2 through pulmonary
21 epithelial lining fluid. Toxicol. Appl. Pharmacol. 109: 464-471.
22 Poulida, O.; Dickerson, R. R.; Doddridge, B. G.; Holland, J. Z.; Wardell, R. G.; Watkins, J. G.
23 (1991) Trace gas concentrations and meteorology in rural Virginia. 1. Ozone and carbon
24 monoxide. J. Geophys. Res. [Atmos.] 96: 22,461-22,475.
25 Price, P. S.; Chaisson, C. F.; Koontz, M.; Wilkes, C.; Ryan, B.; Macintosh, D.; Georgopoulos, P.
26 G. (2003) Construction of a comprehensive chemical exposure framework using person
27 oriented modeling. Prepared for: The Exposure Technical Implementation Panel,
28 American Chemistry Council; contract #1338. Annandale, VA: The LifeLine Group.
29 Available: http://www.thelifelinegroup.org/lifeline/docs.htm [6 March, 2007].
30 Proyou, A. G.; Ziomas, I. C.; Stathopoulos, A. (1998) Application of a three-layer photochemical
31 box model in an Athens street canyon. J. Air Waste Manage. Assoc. 48: 427-433.
32 Purushothaman, V.; Georgopoulos, P. G. (1997) Computational tools to aid the estimation and
33 visualization of potential human exposure to ozone. In: Proceedings of the Air & Waste
34 Management Association Specialty Conference on Computing in Environmental
35 Resource Management; December, 1996; Research Triangle Park, NC. Pittsburgh, PA:
36 A&WMA; VIP-68; pp. 17-28.
37 Purushothaman, V.; Georgopoulos, P. G. (1999a) Evaluation of regional emissions control
38 strategies for ozone utilizing population exposure metrics: a new GIS-based approach
39 applied to the July 1995 episode over the OTAG domain. Piscataway, NJ: Environmental
40 and Occupational Health Sciences Institute, Ozone Research Center; ORC technical
41 report, ORC-TR99-02.
March 2008 AX3-186 DRAFT-DO NOT QUOTE OR CITE
-------
1 Purushothaman, V.; Georgopoulos, P. G. (1999b) Integrating photochemical modeling,
2 geostatistical techniques and geographical information systems for ozone exposure
3 assessment. Piscataway, NJ: Environmental and Occupational Health Sciences Institute,
4 Ozone Research Center; ORC technical report, ORC-TR99-01.
5 Raaschou-Nielsen, O.; Skov, H.; Lohse, C.; Thomsen, B. L.; Olsen, J. H. (1997) Front-door
6 concentrations and personal exposures of Danish children to nitrogen dioxide. Environ.
7 Health Perspect. 105: 964-970.
8 Radiello®. (2006) Radiello® manual. Padova, Italy: Fondazione Salvatore Maugeri Clinica del
9 Lavoro e Delia Riabilitazione I.R.C.C.S., Centre di Ricerche Ambientali. Available:
10 http://www.radiello.com/english/Radiello%27s%20manual%2001-06.pdf [15 August,
11 2007].
12 Ramazan, K. A.; Wingen, L. M.; Miller, Y.; Chaban, G. M.; Gerber, R. B.; Xantheas, S. S.;
13 Finlayson-Pitts, B. J. (2006) New experimental and theoretical approach to the
14 heterogeneous hydrolysis of NO2: key role of molecular nitric acid and its complex. J.
15 Phys. Chem. A 110: 6886-6897.
16 Ramirez-Aguilar, M.; Cicero-Fernandez, P.; Winer, A. M.; Romieu, L; Meneses-Gonzales, F.;
17 Hernandez-Avila, M. (2002) Measurements of personal exposure to nitrogen dioxide in
18 four Mexican cities in 1996. J. Air Waste Manage. Assoc. 52: 50-57.
19 Rao, K. S. (2002) ROADWAY-2: a model for pollutant dispersion near highways. Water Air
20 Soil Pollut. Focus 2: 261-277.
21 Raw, G. J.; Coward, S. K. D.; Brown, V. M.; Crump, D. R. (2004) Exposure to air pollutants in
22 English homes. J. Exposure Anal. Environ. Epidemiol. 14(suppl. 1): S85-S94.
23 Reisen, F.; Arey, J. (2005) Atmospheric reactions influence seasonal PAH and nitro-PAH
24 concentrations in the Los Angeles basin. Environ. Sci. Technol. 39: 64-73.
25 Restrepo, C.; Zimmerman, R.; Thurston, G.; Clemente, J.; Gorczynski, J.; Zhong, M.; Blaustein,
26 M.; Chen, L. C. (2004) A comparison of ground-level air quality data with New York
27 State Department of Environmental Conservation monitoring stations data in South
28 Bronx, New York. Atmos. Environ. 38: 5295-5304.
29 Roberts, J. M.; Williams, J.; Baumann, K.; Buhr, M. P.; Goldan, P. D.; Holloway, J.; Hubler, G.;
30 Kuster, W. C.; McKeen, S. A.; Ryerson, T. B.; Trainer, M.; Williams, E. J.; Fehsenfeld,
31 F. C.; Bertman, S. B.; Nouaime, G.; Seaver, C.; Grodzinsky, G.; Rodgers, M.; Young, V.
32 L. (1998) Measurements of PAN, PPN, and MPAN made during the 1994 and 1995
33 Nashville Intensives of the Southern Oxidant Study: implications for regional ozone
34 production from biogenic hydrocarbons. J. Geophys. Res. [Atmos.] 103: 22,473-22,490.
35 Roberts, J. M.; Flocke, F.; Stroud, C. A.; Hereid, D.; Williams, E.; Fehsenfeld, F.; Brune, W.;
36 Martinez, M.; Harder, H. (2002) Ground-based measurements of peroxycarboxylic nitric
37 anhydrides (PANs) during the 1999 Southern Oxidants Study Nashville Intensive. J.
38 Geophys. Res. [Atmos.] 107(D21): 10.1029/2001JD000947.
39 Roberts, J. M.; Jobson, B. T.; Kuster, W.; Goldan, P.; Murphy, P.; Williams, E.; Frost, G.;
40 Riemer, D.; Apel, E.; Stroud, C.; Wiedinmyer, C.; Fehsenfeld, F. (2003) An examination
41 of the chemistry of peroxycarboxylic nitric anhydrides and related volatile organic
March 2008 AX3-187 DRAFT-DO NOT QUOTE OR CITE
-------
1 compounds during Texas Air Quality Study 2000 using ground-based measurements. J.
2 Geophys. Res. [Atmos.] 108(D16): 10.1029/2003JD003383.
3 Roberts, J. M.; Flocke, F.; Chen, G.; De Gouw, J.; Holloway, J. S.; Hiibler, G.; Neuman, J. A.;
4 Nicks, D. K., Jr.; Nowak, J. B.; Parrish, D. D.; Ryerson, T. B.; Sueper, D. T.; Warneke,
5 C.; Fehsenfeld, F. C. (2004) Measurement of peroxycarboxylic nitric anhydrides (PANs)
6 during the ITCT 2K2 aircraft intensive experiment. J. Geophys. Res. [Atmos.]
7 109(D23S21): 10.1029/2004JD004960.
8 Rojas-Bracho, L.; Suh, H. H.; Oyola, P.; Koutrakis, P. (2002) Measurements of children's
9 exposures to particles and nitrogen dioxide in Santiago, Chile. Sci. Total Environ. 287:
10 249-264.
11 Roorda-Knape, M. C.; Janssen, N. A. H.; De Hartog, J. J.; Van Vliet, P. H. N.; Harssema, H.;
12 Brunekreef, B. (1998) Air pollution from traffic in city districts near major motorways.
13 Atmos. Environ. 32: 1921-1930.
14 Rosenbaum, A. (2005) The HAPEM5 user's guide: hazardous air pollutant exposure model,
15 version 5. Prepared for: U.S. Environmental Protection Agency, Office of Air Quality
16 Planning and Standards. Research Triangle Park, NC. San Francisco, CA: ICF
17 Consulting. Available: http://www.epa.gov/ttn/fera/hapem5/hapem5_guide.pdf [12
18 February, 2008].
19 Rosman, K.; Shimmo, M.; Karlsson, A.; Hansson, H.-C.; Keronen, P.; Allen, A.; Hoenninger, G.
20 (2001) Laboratory and field investigations of a new and simple design for the parallel
21 plate denuder. Atmos. Environ. 35: 5301-5310.
22 Ross, Z.; English, P. B.; Scalf, R.; Gunier, R.; Smorodinsky, S.; Wall, S.; Jerrett, M. (2006)
23 Nitrogen dioxide prediction in southern California using land use regression modeling:
24 potential for environmental health analyses. J. Exposure Sci. Environ. Epidemiol. 16:
25 106-114.
26 Rotko, T.; Kousa, A.; Aim, S.; Jantunen, M. (2001) Exposures to nitrogen dioxide mEXPOLIS-
27 Helsinki: microenvironment, behavioral and sociodemographic factors. J. Exposure Anal.
28 Environ. Epidemiol. 11: 216-223.
29 Rundell, K. W.; Caviston, R.; Hollenbach, A. M.; Murphy, K. (2006) Vehicular air pollution,
30 playgrounds, and youth athletic fields. Inhalation Toxicol. 18: 541-547.
31 Ryerson, T. B.; Trainer, M.; Holloway, J. S.; Parrish, D. D.; Huey, L. G.; Sueper, D. T.; Frost, G.
32 J.; Donnelly, S. G.; Schauffler, S.; Atlas, E. L.; Kuster, W. C.; Goldan, P. D.; Hubler, G.;
33 Meagher, J. F.; Fehsenfeld, F. C. (2001) Observations of ozone formation in power plant
34 plumes and implications for ozone control strategies. Science (Washington, DC) 292:
35 719-723.
36 Sabin, L. D.; Kozawa, K.; Behrentz, E.; Winer, A. M.; Fitz, D. R.; Pankratz, D. V.; Colome, S.
37 D.; Fruin, S. A. (2005) Analysis of real-time variables affecting children's exposure to
38 diesel-related pollutants during school bus commutes in Los Angeles. Atmos. Environ.
39 39: 5243-5254.
40 Sahm, P.; Louka, P.; Ketzel, M.; Guilloteau, E.; Sini, J.-F. (2002) Intercomparison of numerical
41 urban dispersion models - Part I: street canyon and single building configurations. Water
42 Air Soil Pollut. Focus 2: 587-601.
March 2008 AX3-188 DRAFT-DO NOT QUOTE OR CITE
-------
1 Sahsuvaroglu, T.; Arain, A.; Kanaroglou, P.; Finkelstein, N.; Newbold, B.; Jerrett, M.;
2 Beckerman, B.; Brook, J.; Finkelstein, M.; Gilbert, N. L. (2006) A land use regression
3 model for predicting ambient concentrations of nitrogen dioxide in Hamilton, Ontario,
4 Canada. J. Air Waste Manage. Assoc. 56: 1059-1069.
5 Sarnat, J. A.; Schwartz, J.; Catalano, P. J.; Suh, H. H. (2001) Gaseous pollutants in particulate
6 matter epidemiology: confounders or surrogates? Environ. Health Perspect. 109: 1053-
7 1061.
8 Sarnat, J. A.; Brown, K. W.; Schwartz, J.; Coull, B. A.; Koutrakis, P. (2005) Ambient gas
9 concentrations and personal particulate matter exposures: implications for studying the
10 health effects of particles. Epidemiology 16: 385-395.
11 Sarnat, S. E.; Coull, B. A.; Schwartz, J.; Gold, D. R; Suh, H. H. (2006) Factors affecting the
12 association between ambient concentrations and personal exposures to particles and
13 gases. Environ. Health Perspect. 114: 649-654.
14 Sarwar, M.; Corsi, R.; Kimura, Y.; Allen, D.; Weschler, C. (2001) Hydroxyl radicals in indoor
15 environments. In: Proceedings of the Air & Waste Management Association's 94th
16 Annual Conference & Exhibition; June; Orlando, FL. Pittsburgh, PA: Air & Waste
17 Management Association.
18 Sarwar, G.; Corsi, R.; Allen, D.; Weschler, C. J. (2002a) Production and levels of selected indoor
19 radicals: a modeling assessment. In: Proceedings of 9th International Conference on
20 Indoor Air Quality and Climate, Indoor Air 2002; June-July; Monterey, CA.
21 Sarwar, G.; Corsi, R.; Kumura, Y.; Allen, D.; Weschler, C. J. (2002b) Hydroxyl radicals in
22 indoor environments. Atmos. Environ. 36: 3973-3988.
23 Schauer, C.; Niessner, R.; Poschl, U. (2004) Analysis of nitrated polycyclic aromatic
24 hydrocarbons by liquid chromatography with fluorescence and mass spectrometry
25 detection: air particulate matter, soot, and reaction product studies. Anal. Bioanal. Chem.
26 378: 725-736.
27 Schwab, M.; McDermott, A.; Spengler, J. D.; Samet, J. M.; Lambert, W. E. (1994) Seasonal and
28 yearly patterns of indoor nitrogen dioxide levels: data from Albuquerque, New Mexico.
29 Indoor Air 4: 8-22.
30 Shima, M.; Adachi, M. (1998) Indoor nitrogen dioxide in homes along trunk roads with heavy
31 traffic. Occup. Environ. Med. 55: 428-433.
32 Shima, M.; Adachi, M. (2000) Effect of outdoor and indoor nitrogen dioxide on respiratory
33 symptoms in schoolchildren. Int. J. Epidemiol. 29: 862-870.
34 Sillman, S.; Logan, J. A.; Wofsy, S. C. (1990) The sensitivity of ozone to nitrogen oxides and
35 hydrocarbons in regional ozone episodes. J. Geophys. Res. [Atmos.] 95: 1837-1851.
36 Simoni, M.; Scognamiglio, A.; Carrozzi, L.; Baldacci, S.; Angino, A.; Pistelli, F.; Di Pede, F.;
37 Viegi, G. (2004) Indoor exposures and acute respiratory effects in two general population
38 samples from a rural and an urban area in Italy. J. Exposure Anal. Environ. Epidemiol.
39 14(suppl 1): S144-S152.
March 2008 AX3-189 DRAFT-DO NOT QUOTE OR CITE
-------
1 Singer, B. C.; Hodgson, A. T.; Hotchi, T.; Kim, J. J. (2004) Passive measurement of nitrogen
2 oxides to assess traffic-related pollutant exposure for the East Bay Children's Respiratory
3 Health Study. Atmos. Environ. 38: 393-403.
4 Sjodin, A.; Sjoberg, K.; Svanberg, P. A.; Backstrom, H. (1996) Verification of expected trends in
5 urban traffic NO* emissions from long-term measurements of ambient NO2
6 concentrations in urban air. Sci. Total Environ. 189/190: 213-220.
7 Slack, H. H.; Heumann, M. A. (1997) Use of unvented residential heating appliances—United
8 States, 1988-1994. Morb. Mortal. Wkly. Rep. 46: 1221-1224.
9 Smedje, G.; Norback, D.; Edling, C. (1997) Subjective indoor air quality in schools in relation to
10 exposure. Indoor Air 7: 143-150.
11 Soderstrom, H.; Hajslova, J.; Kocourek, V.; Siegmund, B.; Kocan, A.; Obiedzinski, M. W.;
12 Tysklind, M.; Bergqvist, P.-A. (2005) PAHs and nitrated PAHs in air of five European
13 countries determined using SPMDs as passive samplers. Atmos. Environ. 39: 1627-1640.
14 Son, B.; Yang, W.; Breysse, P.; Chung, T.; Lee, Y. (2004) Estimation of occupational and
15 nonoccupational nitrogen dioxide exposure for Korean taxi drivers using a
16 microenvironmental model. Environ. Res. 94: 291-296.
17 S0rensen, D. N.; Weschler, C. J. (2002) Modeling-gas phase reactions in indoor environments
18 using computational fluid dynamics. Atmos. Environ. 36: 9-18.
19 S0rensen, M.; Loft, S.; Andersen, H. V.; Raaschou-Nielsen, O.; Skovgaard, L. T.; Knudsen, L.
20 E.; Nielsen, I. V.; Hertel, O. (2005) Personal exposure to PM2 5, black smoke and NO2 in
21 Copenhagen: relationship to bedroom and outdoor concentrations covering seasonal
22 variation. J. Exposure Anal. Environ. Epidemiol. 15: 413-422.
23 Spengler, J. D.; Brauer, M.; Samet, J. M.; Lambert, W. E. (1993) Nitrous acid in Albuquerque,
24 New Mexico, homes. Environ. Sci. Technol. 27: 841-845.
25 Spengler, J.; Schwab, M.; Ryan, P. B.; Colome, S.; Wilson, A. L.; Billick, L; Becker, E. (1994)
26 Personal exposure to nitrogen dioxide in the Los Angeles Basin. J. Air Waste Manage.
27 Assoc. 44: 39-47.
28 Spengler, J. D.; Lee, K.; Yanagisawa, Y.; Bischof, W.; Braathan, O.; Chung, S.; Coward, K.;
29 Gutschmidt, V.; Isidorov, V.; Jahng, D.; Jin, K.; Korenaga, T.; Maroni, M.; Ohkoda, Y.;
30 Pastuszka, J.; Patil, R. S.; Qing, X.; Raizenne, M.; Romieu, L; Salonen, R.; Sega, K.;
31 Seifert, B.; Shah, S.; Torres, E.; Yoon, D.; Zhang, X. (1996) Impact of residential
32 nitrogen exposure on personal exposure: an international study. In: Indoor air '96: the 7th
33 international conference on indoor air quality, volume I; July; Nagoya, Japan. Tokyo,
34 Japan: Institute of Public Health; pp. 931-936.
35 Spicer, C. W.; Coutant, R. W.; Ward, G. F.; Joseph, D. W.; Gaynor, A. J.; Billick, I. H. (1989)
36 Rates and mechanisms of NO2 removal from indoor air by residential materials. Environ.
37 Int. 15: 643-654.
38 Spicer, C. W.; Kenny, D. V.; Ward, G. F.; Billick, I. H. (1993) Transformations, lifetimes, and
39 sources of NO2, HONO, and HNOs in indoor environments. Air Waste 43: 1479-1485.
March 2008 AX3-190 DRAFT-DO NOT QUOTE OR CITE
-------
1 Staimer, N.; Delfmo, R. 1; Bufalino, C.; Fine, P. M.; Sioutas, C.; Kleinman, M. T. (2005) A
2 miniaturized active sampler for the assessment of personal exposure to nitrogen dioxide.
3 Anal. Bioanal. Chem. 383: 955-962.
4 Stemmler, K.; Ammann, M.; Bonders, C.; Kleffmann, J.; George, C. (2006) Photosensitized
5 reduction of nitrogen dioxide on humic acid as a source of nitrous acid. Nature (London,
6 U.K.) 440: 195-198.
7 Stevenson, K.; Bush, T.; Mooney, D. (2001) Five years of nitrogen dioxide measurement with
8 diffusion tube samplers at over 1000 sites in the UK. Atmos. Environ. 35: 281-287.
9 Sverdrup, G. M.; Spicer, C. W.; Ward, G. F. (1987) Investigation of the gas-phase reaction of
10 dinitrogen pentoxide with water-vapor. Int. J. Chem. Kinet. 19: 191-205.
11 Tashiro, Y.; Taniyama, T. (2002) Atmospheric NO2 and CO concentration in Lima, Peru.
12 Environ. Int. 28: 227-233.
13 Teshome, E. J.; Haghighat, F. (2004) Zonal models for indoor air flow - a critical review. Int. J.
14 Ventil. 3: 119-129.
15 Topp, R.; Cyrys, J.; Gebefiigi, I; Schnelle-Kreis, J.; Richter, K.; Wichmann, H.-E.; Heinrich, J.
16 (2004) Indoor and outdoor air concentrations of BTEX and NO2: correlation of repeated
17 measurements. J. Environ. Monit. 6: 807-812.
18 Triche, E. W.; Belanger, K.; Bracken, M. B.; Beckett, W. S.; Holford, T. R.; Gent, J. F.;
19 McSharry, J.-E.; Leaderer, B. P. (2005) Indoor heating sources and respiratory symptoms
20 in nonsmoking women. Epidemiology 16: 377-384.
21 United Kingdom Air Quality Expert Group (U.K. AQEG). (2004) Nitrogen dioxide in the United
22 Kingdom. London, United Kingdom: Department for Environment, Food and Rural
23 Affairs. Available:
24 http://www.defira.gov.uk/environment/airquality/panels/aqeg/index.htm [12 April, 2007].
25 U.S. Environmental Protection Agency. (1992) Final guidelines for exposure assessment.
26 Research Triangle Park, NC: National Center for Environmental Assessment; EPA/600Z-
27 92/001.
28 U.S. Environmental Protection Agency. (1993) Air quality criteria for oxides of nitrogen.
29 Research Triangle Park, NC: Office of Health and Environmental Assessment,
30 Environmental Criteria and Assessment Office; report nos. EPA/600/8-91/049aF-cF. 3v.
31 Available from: NTIS, Springfield, VA; PB95-124533, PB95-124525, and PB95-124517.
32 U.S. Environmental Protection Agency. (1996) Air quality criteria for particulate matter.
33 Research Triangle Park, NC: National Center for Environmental Assessment-RTF Office;
34 report nos. EPA/600/P-95/001aF-cF. 3v.
35 U.S. Environmental Protection Agency. (1997) Exposure factors handbook. Washington, DC:
36 Office of Research and Development, National Center for Environmental Assessment;
37 report nos. EPA/600/P-95/002Fa-c.
38 U.S. Environmental Protection Agency. (2003) National air quality and emissions trends report.
39 2003 special studies edition. Research Triangle Park, NC: Office of Air Quality
40 Standards; Emissions Monitoring and Analysis Division; report no. EPA 454/R-03-005.
41 Available: http://www.epa.gov/air/airtrends/aqtrnd03/toc.html (27 August, 2004).
March 2008 AX3-191 DRAFT-DO NOT QUOTE OR CITE
-------
1 U.S. Environmental Protection Agency. (2004) Air quality criteria for particulate matter.
2 Research Triangle Park, NC: National Center for Environmental Assessment; report no.
3 EPA/600/P-99/002aF-bF. 2v. Available: http://cfpub.epa.gov/ncea/ [9 November, 2004].
4 U.S. Environmental Protection Agency. (2006a) Air quality criteria for ozone and related
5 photochemical oxidants. Research Triangle Park, NC: National Center for Environmental
6 Assessment; report no. EPA/600/R-05/004aF-cF. 3v. Available:
7 http://cfpub.epa.gov/ncea/ [24 March, 2006].
8 U.S. Environmental Protection Agency. (2006b) Total risk integrated methodology (TRIM) air
9 pollutants exposure model documentation (TRIM.Expo/APEX, version 4). Volume I:
10 User's guide. Research Triangle Park, NC: U.S. Environmental Protection Agency.
11 Available: http://www.epa.gov/ttn/fera/data/apex/APEX4UsersGuideJuly2006.pdf [7
12 March, 2007].
13 U.S. Environmental Protection Agency. (2006c) Total risk integrated methodology (TRIM) air
14 pollutants exposure model documentation (TREVI.Expo/APEX, version 4). Volume II:
15 technical support document. Research Triangle Park, NC: U.S. Environmental Protection
16 Agency. Available: http://www.epa.gov/ttn/fera/data/apex/APEX4TSDJuly2006.pdff7
17 March, 2007].
18 U.S. Environmental Protection Agency. (2006d) Plume-in-grid model. Available:
19 http://www.epa.gov/asmdnerl/ModelDevelopment/plumeInGrid.html [7 March, 2007].
20 U.S. Environmental Protection Agency. (2007) Air Quality System (AQS). Washington, DC:
21 Office of Air and Radiation. Available: http://www.epa.gov/ttn/airs/airsaqs/ [12 April,
22 2007].
23 Varshney, C. K.; Singh, A. P. (2003) Passive samplers for NOX monitoring: a critical review.
24 Environmentalist 23: 127-136.
25 Venegas, L. E.; Mazzeo, N. A. (2004) Aplication of atmospheric dispersion models to evaluate
26 population exposure to NO2 concentration in Buenos Aires. In: Proceedings of the 9th
27 International Conference on Harmonisation Within Atmospheric Dispersion Modelling
28 for Regulatory Purposes; p. 182-186.
29 Vinzents, P. S.; M011er, P.; S0rensen, M.; Knudsen, L. E.; Herte, L. Q.; Jensen, F. P.; Schibye,
30 B.; Loft, S. (2005) Personal exposure to ultrafme particles and oxidative DNA damage.
31 Environ. Health Perspect. 113: 1485-1490.
32 Vogel, M. (2005) Sampling of airborne pollutants: strategies and developments. Anal. Bioanal.
33 Chem. 381: 84-86.
34
35 Vukovich, F. (2000) The spatial variation of the weekday/weekend differences in the Baltimore
36 area. J. Waste Manage. Assoc. 50: 2067-2072.
37 Wainman, T.; Zhang, J.; Weschler, C. J.; Lioy, P. J. (2000) Ozone and limonene in indoor air: a
38 source of submicron particle exposure. Environ. Health Perspect. 108: 1139-1145.
39 Wainman, T.; Weschler, C.; Lioy, P.; Zhang, J. (2001) Effects of surface type and relative
40 humidity on the production and concentration of nitrous acid in a model indoor
41 environment. Environ. Sci. Technol. 35: 2200-2206.
March 2008 AX3-192 DRAFT-DO NOT QUOTE OR CITE
-------
1 Wayne, R. P.; Barnes, I; Biggs, P.; Burrows, J. P.; Canosa-Mas, C. E.; Hjorth, J.; Le Bras, G.;
2 Moortgat, G. K.; Perner, D.; Poulet, G.; Restelli, G.; Sidebottom, H. (1991) The nitrate
3 radical: physics, chemistry, and the atmosphere. Atmos. Environ. Part A 25: 1-203.
4 Weschler, C. J.; Shields, H. C. (1996) The conversion (reduction) of nitrogen dioxide to nitric
5 oxide as a consequence of charcoal filtration. In: Yoshizawa, S.; Kimura, K.; Ikeda, K.;
6 Tanabe, S.; Iwata, T., eds. Indoor Air '96: proceedings of the 7th international conference
7 on indoor air quality and climate, v. 3, July; Nagoya, Japan. Toykyo, Japan: Indoor Air
8 '96; pp. 453-458.
9 Weschler, C. J.; Shields, H. C. (1997) Potential reactions among indoor pollutants. Atmos.
10 Environ. 31: 3487-3495.
11 Weschler, C. J.; Brauer, M.; Koutrakis, P. (1992) Indoor ozone and nitrogen dioxide: a potential
12 pathway to the generation of nitrate radicals, dinitrogen pentaoxide, and nitric acid
13 indoors. Environ. Sci. Technol. 26: 179-184.
14 Weschler, C. J.; Shields, H. C.; Naik, D. V. (1994) Indoor chemistry involving O3, NO, and NO2
15 as evidenced by 14 months of measurements at a site in southern California. Environ. Sci.
16 Technol. 28: 2120-2132.
17 Westerdahl, D.; Fruin, S.; Sax, T.; Fine, P. M.; Sioutas, C. (2005) Mobile platform
18 measurements of ultrafine particles and associated pollutant concentrations on freeways
19 and residential streets in Los Angeles. Atmos. Environ. 39: 3597-3610.
20 Whitfield, R. G.; Richmond, H. M.; Johnson, T. R. (1997) Overview of ozone human exposure
21 and health risk analyses used in the U.S. EPA's review of the ozone air quality standard.
22 In: Proceedings of the U.S.-Dutch International Symposium on Air Pollution in the 21st
23 Century; April; Noordwijk, The Netherlands. Argonne, IL: Argonne National Laboratory;
24 ANL/DIS/CP-92660. Available: http://www.osti.gov/dublincore/gpo/servlets/purl/9748-
25 Nh3avx/webviewable/9748.pdf (22 July 2003).
26 Wilson, N. K.; Chuang, J. C.; Kuhlman, M. R. (1991) Sampling polycyclic aromatic
27 hydrocarbons and other semivolatile organic compounds in indoor air. Indoor Air 4: 513-
28 521.
29 Wilson, W. E.; Mage, D. T.; Grant, L. D. (2000) Estimating separately personal exposure to
30 ambient and nonambient particulate matter for epidemiology and risk assessment: why
31 and how. J. Air Waste Manage. Assoc. 50: 1167-1183.
32 Wooders, P. J. (1994) The external costs of UK domestic gas use (thesis). London, United
33 Kingdom: London University, Imperial College.
34 World Health Organization (WHO). (2003) Nitrogenated polycyclic aromatic hydrocarbons.
35 Geneva, Switzerland: World Health Organization. (Environmental Health Criteria 229).
36 World Health Organization (WHO). (2004) IPCS risk assessment terminology. Part 2: IPCS
37 glossary of key exposure assessment terminology. Geneva, Switzerland: IPCS
38 Harmonization Project document no. 1. Available:
39 http://www.who.int/ipcs/methods/harmonization/areas/ipcsterminologypartsland2.pdf
40 [19 April, 2007].
March 2008 AX3-193 DRAFT-DO NOT QUOTE OR CITE
-------
1 World Health Organization (WHO). (2005) Principles of characterizing and applying human
2 exposure models. Geneva, Switzerland: World Health Organization. (IPCS
3 harmonization project document; no. 3). Available:
4 http://whqlibdoc.who.int/publications/2005/9241563117_eng.pdf [7 March, 2007].
5 Yamanaka, S. (1984) Decay rates of nitrogen oxides in a typical Japanese living room. Environ.
6 Sci. Technol. 18: 566-570.
7 Yamartino, R. J.; Wiegand, G. (1986) Development and evaluation of simple models for flow,
8 turbulence and pollutant concentration fields within an urban street canyon. Atmos.
9 Environ. 20: 2137-2156.
10 Yanagisawa, Y.; Nishimura, H. (1982) A badge-type personal sampler for measurement of
11 personal exposure to NO2 and NO in ambient air. Environ. Int. 8: 235-242.
12 Yanagisawa, Y.; Nishimura, H.; Matsuki, H.; Osaka, F.; Kasuga, H. (1986) Personal exposure
13 and health effect relationship for NO2 with urinary hydroxyproline to creatinine ratio as
14 indicator. Arch. Environ. Health 41: 41-48.
15 Yang, W.; Lee, K.; Chung, M. (2004) Characterization of indoor air quality using multiple
16 measurements of nitrogen dioxide. Indoor Air 14: 105-111.
17 Zeger, S. L.; Thomas, D.; Dominici, F.; Samet, J. M.; Schwartz, J.; Dockery, D.; Cohen, A.
18 (2000) Exposure measurement error in time-series studies of air pollution: concepts and
19 consequences. Environ. Health Perspect. 108: 419-426.
20 Zipprich, J. L.; Harris, S. A.; Fox, J. C.; Borzelleca, J. F. (2002) An analysis of factors that
21 influence personal exposure to nitrogen oxides in residents of Richmond, Virginia. J.
22 Exposure Anal. Environ. Epidemiol. 12: 273-285.
23 Zota, A.; Adamkiewicz, G.; Levy, J. L; Spengler, J. D. (2005) Ventilation in public housing:
24 implications for indoor nitrogen dioxide concentrations. Indoor Air 15: 393-401.
March 2008 AX3-194 DRAFT-DO NOT QUOTE OR CITE
-------
i AX4. CHAPTER 4 ANNEX - TOXICOLOGICAL
2 EFFECTS OF NITROGEN DIOXIDE AND RELATED
3 OXIDES OF NITROGEN
4
5
6 AX4.1 PULMONARY EFFECTS OF NITROGEN DIOXIDE AND
7 RELATED OXIDES OF NITROGEN
8
9 AX4.1.1 Effects of Nitrogen Dioxide on Antioxidant and Antioxidant
10 Metabolism
11 Nitrogen dioxide is an oxidant and lipid peroxidation is believed to be a major molecular
12 event responsible for its toxicity. As a result, there has been considerable attention paid to the
13 effect of NO2 on the antioxidant defense system in the epithelial lining fluid and in pulmonary
14 cells. Repeated exposures to NO2 at concentrations ranging from 0.04 to 33 ppm have revealed
15 effects on low molecular weight antioxidants such as glutathione, vitamin E, and vitamin C, as
16 well as some enzymes involved in cell oxidant homeostasis.
17 A number of studies have investigated the hypothesis, originally proposed by Menzel
18 (1970), that antioxidants might protect the lung from NC>2 damage by inhibiting lipid
19 peroxidation (see Table AX4.1). Changes in the activity of enzymes in the lungs of NC>2-
20 exposed animals that regulate levels of glutathione (GSH) have been reported at relatively low
21 exposure concentrations. Sagai et al. (1984) studied the effects of prolonged (9 and 18 months)
22 exposure to 0.04, 0.4, and 4.0 ppm NC>2 on rats. After either exposure duration, non-protein
23 sulfhydryl levels were increased at 0.4 ppm or greater, and exposure to 4.0 ppm decreased the
24 activity of GSH peroxidase but increased glucose-6-phosphate dehydrogenase activity.
25 Glutathione peroxidase activity was also decreased in rats exposed to 0.4 ppm NC>2 for 18
26 months. Three GSH S-transferases were also studied, two of which (aryl S-transferase and
27 aralkyl S-transferase) exhibited decreased activities after 18 months of exposure to 0.4 ppm or
28 greater NC>2. No effects were observed on the activities of 6-phosphogluconate dehydrogenase,
29 superoxide dismutase, or disulfide reductase. Effects followed a concentration- and exposure-
30 duration response function. The decreases in glutathione-related enzyme activities were
31 inversely related to the apparent formation of lipid peroxides (see lipid peroxidation subsection).
32 Shorter exposures (4 months) to NC>2 between 0.4 and 4.0 ppm also caused concentration- and
March 2008 AX4-1 DRAFT-DO NOT QUOTE OR CITE
-------
1 duration-dependent effects on antioxidant enzyme activities (Ichinose and Sagai, 1982). For
2 example, glucose-6-phosphate dehydrogenase increased, reaching a peak at 1 month, and then
3 decreased towards the control value. Shorter (2-week) exposures to 0.4 ppm NO2 caused no
4 such effects in rats or guinea pigs (Ichinose and Sagai, 1989).
5 The activities of GSH reductase and glucose-6-phosphate dehydrogenase were
6 significantly increased during exposure to 6.2 ppm NO2 for 4 days; GSH peroxidase activity was
7 not affected (Chow et al., 1974). The possible role of edema and cellular inflammation in these
8 findings was not examined. Since NO2 had no significant effect on lung GSH peroxidase
9 activity in this study, but did significantly increase the activities of GSH reductase and glucose-
10 6-phosphate dehydrogenase, the authors concluded that NO2 attacks mainly GSH and NADPH.
11 Newer studies also identified effects on glutathione. Changes in glutathione status in the
12 blood and lung (bronchoalveolar lavage (BAL) fluid) occurred in rats exposed to 5 ppm and 10
13 ppm NO2 continuously for 24 h, but not for 7 days (Pagani et al., 1994). Total glutathione - total
14 of reduced (GSH) and oxidized (GSSG) form - was significantly increased in blood but not in
15 BAL fluid; however, GSSG was elevated in BAL fluid only. A decreased GSH/GSSG ratio was
16 observed in the blood and BAL fluid, but not in lung type II cells, in rats continuously exposed to
17 10 ppm NO2 for 3 or 20 days (Hochscheid et al., 2005). Interestingly, lipid peroxidation was
18 decreased in type II cells at 3 days, but was similar to controls at 20 days. Gene expression, as
19 measured by mRNA levels of the enzymes involved in the biosynthesis of glutathione - gamma-
20 glutamylcysteine synthetase (yGCS) and glutathione synthetase (GS), was decreased at both time
21 points, but gamma-glutamyltranspeptidase (yGT) mRNA expression was increased. No GSH
22 peroxidase activity (important for hydroperoxide reduction of complex lipids) was detected at 3
23 days, and was barely detected at 20 days.
24 Malnutrition of animals can drastically affect their response to toxicants, including NO2.
25 Experimental interest in this area has mainly focused on dietary lipids, vitamin E and other lipid-
26 soluble antioxidants, and vitamin C and other water-soluble antioxidants. Ayaz and Csallany
27 (1978) exposed vitamin E-deficient and vitamin E-supplemented (30 or 300 mg/kg of diet)
28 weanling mice continuously for 17 months to 0.5 or 1.0 ppm NO2 and assayed blood, lung, and
29 liver tissues for GSH peroxidase activity. Exposure to 1.0 ppm NO2 alone or combined with
30 vitamin E deficiency decreased the enzyme activity in the blood and lungs. Neither vitamin E
March 2008 AX4-2 DRAFT-DO NOT QUOTE OR CITE
-------
1 deficiency nor NC>2 exposure affected liver GSH peroxidase activity. However, in vitamin E-
2 supplemented mice, GSH peroxidase activity increased at 0.5 ppm and 1.0 ppm NC>2.
3
4 AX4.1.2 Lipid Metabolism and Content of the Lung
5 Lipid peroxidation is an important mechanism of cell damage arising from changes in
6 cell membrane structure and function. The ability of NC>2 exposure to induce lipid peroxidation
7 in the respiratory tract has been well demonstrated in available studies as measured by increased
8 ethane exhalation in the breath, as thiobarbituric acid (TEA) reactive substances in tissues, and
9 as the content of conjugated dienes in tissue homogenates.
10 A number of studies have investigated the effects of NO2 exposure on lipid metabolism
11 and content of the lung. Lipid peroxidation induced by NC>2 exposure has been detected at
12 exposure concentrations as low as 0.04 ppm. Increased ethane exhalation was observed in rats
13 exposed to 0.04 or 0.12 ppm after 9 and 18 months of exposure (Sagai et al., 1984). Exposure to
14 0.4 ppm NC>2 for 9 months or longer and to 4.0 ppm for 6 months resulted in increased TEA
15 reactants (Ichinose et al., 1983). NO2 exposures for shorter durations also increased lipid
16 peroxidation in rats. For example, NC>2 concentrations of 1.2 ppm or greater for 1 week
17 (Ichinose and Sagai, 1982; Ichinose et al., 1983) increased ethane exhalation in rats, while
18 exposure of pregnant rats to 0.53 or 5.3 ppm NC>2 for 5 h/day for 21 days rats resulted in
19 increases in lung lipid peroxidation products (Balabaeva and Tabakova, 1985). These results
20 indicate at least some degree of duration-dependence in the formation of lipid peroxidation, with
21 lower effect thresholds identified with longer durations of exposure.
22 Lipid peroxidation results in altered phospholipid composition, which in turn may affect
23 membrane fluidity and thus, membrane function. Significant depression of lipid content and
24 total content of saturated fatty acids such as phosphatidyl-ethanolamine, lecithin
25 (phosphatidylcholine), phosphatidylinositol, and phosphatidylserine were found in rats exposed
26 to 2.9 ppm NC>2 for 24 h/day, 5 days/week for 9 months (Arner and Rhoades, 1973). Exposure
27 of rabbits to 1.0 ppm NC>2 for 2 weeks also caused depression of lecithin synthesis after one
28 week of exposure (Seto et al., 1975), while exposure of rats to 5.5 ppm NC>2 for 3 h/day for 7 or
29 14 days elicited only few changes in lipid metabolism (Yokoyama et al., 1980). In beagle dogs,
30 the amount of unsaturated fatty acids in the phospholipids from the lungs was increased after
31 exposure to concentrations ranging from 5 to 16 ppm, but not to 3 ppm (Dowell et al., 1971).
March 2008 AX4-3 DRAFT-DO NOT QUOTE OR CITE
-------
1 Exposure of either mice or guinea pigs to an NC>2 level of 0.4 ppm for a week resulted in a
2 decreased concentration of phosphatidylethanolamine and a relative increase in the
3 phosphatidylcholine concentration (Sagai et al., 1987). Concentration-and exposure duration-
4 dependent increases were reported in phospholipid components in BAL fluid, when rats were
5 exposed to 10 ppm NO2 continuously for 1 day or 3 days (Miiller et. al., 1994).
6 Functional studies conducted on surfactant phospholipid extracts indicated that NO2
7 exposures of 5 ppm or greater, but not to 0.8 ppm, directly impaired surface tension, although the
8 structure of the surfactant protein A (SP-A) was not altered by NC>2 exposure. Changes in the
9 phospholipid composition of membranes may result in disruption of the cell membrane barrier.
10 Miiller et al. (2003) found that uptake of liposomes by type II lung cells occurred more easily
11 from animals exposed to 10 ppm NO2 for 3 to 28 days, possibly as a result of increased demand
12 of phosphatidylcholine during lung injury.
13 Lipid peroxidation can also activate phospholipases. Activation of phospholipase Al in
14 cultured endothelial cells occurred at NC>2 concentration of 5 ppm after 40 h of exposure and was
15 speculated to depend on a specific NO2-induced increase in phosphatidyl serine in the plasma
16 membranes (Sekharam et al., 1991).
17 One function of phospholipases is the release of arachidonic acid (AA), which serves as a
18 mediator of inflammatory response. NC>2 exposure affects the release and metabolism of
19 arachidonic acid both in vivo and in vitro. The products of arachidonic acid metabolism, such as
20 prostaglandins, prostacyclin, thromboxanes, and leukotrienes play an important role (such as
21 recruitment of neutrophils to sights of local irritation) in modulating inflammatory response.
22 Schlesinger et al. (1990) reported elevated concentrations of thromboxane B2 (1x82) following
23 NC>2 exposures of 1.0 ppm for 2 h, depressed concentrations at 3.0 ppm, and significant
24 depression 24 h postexposure at 10 ppm NC>2. The same investigators also reported depressed
25 level of 6-keto-prostaglandin F1 a at 1.0 ppm NC>2, but exposure to NC>2 did not affect
26 prostaglandins E2 and F2 and leukotriene B4 (LTB4) levels.
27 Changes in activation of arachidonate metabolism were also reported in rat alveolar
28 macrophages (AMs) when these animals were exposed to 0.5 ppm NO2 for 0.5, 1, 5, and 10 days
29 (Robison et al., 1993). Unstimulated AM synthesis of LTB4 was depressed after 0.5 days and
30 again after 5 days of exposure to NC>2. Alveolar macrophage production of TxB2, LTB/t, and 5-
31 hydroxyeicosatetraenoic acid (5-HETE) in response to stimulation with the calcium ionophore,
March 2008 AX4-4 DRAFT-DO NOT QUOTE OR CITE
-------
1 A23187, was depressed after 0.5 days of exposure and recovered to air-control values with
2 longer exposure periods. 5-HETE levels were increased after 10 days of exposure. However,
3 AM production of LTB4 in response to zymosan-activated rat serum was depressed only after 5
4 days of exposure.
5 The effects of NO2 on structural proteins of the lungs have been of concern because
6 elastic recoil is lost after exposure. Collagen synthesis rates are increased in rats exposed to NO2
7 concentrations as low as 5.0 ppm NC>2. It has been assumed that increased collagen synthesis
8 reflect increases in total lung collagen which, if sufficient, could result in pulmonary fibrosis
9 after longer periods of exposure. Such correlation has yet to be confirmed by in vivo studies
10 involving NO2 exposure.
11 Alterations in xenobiotic metabolism pathways following NO2 exposure are also
12 summarized in Table AX4.2, in addition to changes in phase I enzymes (such as cytochrome
13 P450s) and phase II enzymes (GST as described earlier). While these changes are not
14 necessarily toxic manifestations of NC>2 per se, such changes may impact the metabolism and
15 toxicity of other chemicals. Glycolytic pathways are also apparently affected. For example,
16 glycolytic metabolism was increased by NC>2 exposure, apparently due to a concurrent increase
17 in type II cells (Mochitate et al., 1985).
18
19 AX4.1.3 Emphysema Following Nitrogen Dioxide Exposure
20 Emphysema as a result of chronic exposure to NC>2 has been reported in animal studies.
21 The definition of emphysema has changed since the time that some of the studies have been
22 published; thus, it is important to compare the findings of the studies with the current definition
23 of emphysema. U.S. Environmental Protection Agency (1993) evaluated the animal studies
24 reporting emphysema from chronic exposure to NC>2 based upon the most recent definition of
25 emphysema from the report of the National Heart, Lung and Blood Institute (NHLBI), Division
26 of Lung Diseases Workshop (Snider et al., 1985); see U.S. Environmental Protection Agency
27 (1993) for the definitions of emphysema. Because the focus of this document is extrapolation of
28 NC>2 exposures to potential hazards for humans, only those studies showing emphysema of the
29 type seen in human lungs will be discussed.
30 Emphysema was reported by Hay don et al. (1967) in rabbits exposed continuously
31 (presumably 24 h/day) for 3 to 4 months to 8.0 or 12.0 ppm NC>2. The investigators reported
March 2008 AX4-5 DRAFT-DO NOT QUOTE OR CITE
-------
1 enlarged lungs that failed to collapse when the thorax was opened. When the lungs were fixed in
2 an expanded state via the trachea using formaldehyde, there was evidence of enlarged airspaces
3 with destructive changes in alveolar walls. Although no stereology was performed, the changes
4 observed appear to be emphysema of the type seen in human lungs.
5 WHO (1997) has also reported a study by Freeman et al. (1972) in which rats were
6 exposed to 20.0 ppm NO2, which was reduced during the exposure to 15.0 ppm or to 10.0 ppm,
7 for varying periods up to 33 months. The lungs were fixed via the trachea, and morphometric
8 analysis of the lung and alveolar size indicated an enlargement of alveolar, reduction in alveolar
9 surface, and alveolar destruction. Although the investigators concluded that their study
10 demonstrated emphysema in their NO2-exposed rats, WHO (1997) noted that it was not entirely
11 clear whether the experimental groups or only the group exposed to 15.0 ppm had emphysema.
12 Although many of the papers reviewed (U.S. Environmental Protection Agency, 1993)
13 reported finding emphysema, some of these studies were reported according to previous,
14 different criteria; some reports did not fully describe the methods used; and/or the results
15 obtained were not in sufficient detail to allow independent confirmation of the presence of
16 emphysema. For example, Hyde et al. (1978) reported no emphysema in beagle dogs exposed
17 16 h daily for 68 months to 0.64 ppm NO2 with 0.25 ppm NO or to 0.14 ppm NO2 with 1.67 ppm
18 NO. The dogs then breathed clean air during a 32-to 36-month post-exposure period. The right
19 lungs were fixed via the trachea at 30-cm fixative pressure in a distended state. Semiautomated
20 image analysis was used for morphometry of air spaces. The dogs exposed to 0.64 ppm NO2
21 with 0.25 ppm NO had significantly larger lungs with enlarged air spaces and evidence of
22 destruction of alveolar walls. These effects were not observed in dogs exposed to 0.14 ppm NO2
23 with 1.67 ppm NO, implying a significant role of the NO2 in the production of the lesions. The
24 lesions in the dogs exposed to the higher NO2 concentration meet the criteria of the 1985 NHLBI
25 workshop for emphysema of the type seen in human lungs.
26
27 AX4.1.4 Nitrates (NO3~)
28 Busch et al. (1986) exposed rats and guinea pigs with either normal lungs or elastase-
29 induced emphysema to ammonium nitrate aerosols at 1 mg/m3 for 6 h/day, 5 days/week for
30 4 weeks. Using light and electron microscopy, the investigators concluded that there were no
31 significant effects of exposure on lung structure.
March 2008 AX4-6 DRAFT-DO NOT QUOTE OR CITE
-------
1 AX4.2 DOSIMETRY OF INHALED NITROGEN OXIDES
2 This section provides an overview of NO2 dosimetry and updates information provided in
3 the 1993 AQCD for Oxides of Nitrogen. Dosimetry of NO2 refers to the measurement or
4 estimation of the amount of NO2 or its reaction products reaching and persisting at specific sites
5 in the respiratory tract following an exposure. Nitrogen dioxide, classified as a reactive gas,
6 interacts with surfactants, antioxidants, and other compounds in the epithelial lining fluid (ELF).
7 The compounds thought responsible for adverse pulmonary effects of inhaled NO2 are the
8 reaction products themselves or the metabolites of these products in the ELF. At the time of the
9 1993 AQCD for Oxides of Nitrogen, it was thought that inhaled NO2 probably reacted with the
10 water molecules in the ELF to form nitrous acid (FDSTO2) and nitric acid (HNOs). However,
11 some limited data suggested that the absorption of NO2 was linked to reactive substrates in the
12 ELF and subsequent nitrite production. Since then, the reactive absorption of NO2 has been
13 examined in a number of studies (see Section 4.2.2). These studies have characterized the
14 absorption kinetics and reactive substrates for NO2 delivered to various sites in the respiratory
15 tract. Researchers have attempted to obtain a greater understanding of how these complex
16 interactions affect NO2 absorption and NO2-induced injury.
17 With respect to quantifying absolute NO2 absorption, the following were reported in the
18 1993 AQCD for Oxides of Nitrogen. The principles of Os uptake were generally assumed
19 applicable for NO2 modeling studies. The results indicated that NO2 is absorbed throughout the
20 lower respiratory tract, but the major delivery site is the centriacinar region, i.e., the junction
21 between the conducting and respiratory airways in humans and animals. Experimental studies
22 have found that the total respiratory tract uptake in humans ranges from 72 to 92% depending on
23 the study and the breathing conditions. The percent total uptake increases with increasing
24 exercise level. In laboratory animals, upper respiratory tract uptakes ranged from as low as 25%
25 to as high as 94% depending on the study, species, air flow rate, and mode of breathing (nasal or
26 oral). Upper respiratory tract uptake of NO2 was found to decrease with increasing ventilation.
27 Uptake during nasal breathing was determined to be significantly greater than during oral
28 breathing.
29
March 2008 AX4-7 DRAFT-DO NOT QUOTE OR CITE
-------
1 AX4.2.1 Mechanisms of NO2 Absorption
2 The ELF is the initial barrier against NO2 delivery to the underlying epithelial cells.
3 Postlethwait and Bidani (1990) suggested that acute NO2 uptake in the lower respiratory tract
4 was rate limited by chemical reactions of NO2 with ELF constituents rather than by gas solubility
5 in the ELF. Subsequently, Postlethwait et al. (1991) reported that inhaled NO2 (10 ppm) does
6 not penetrate the ELF to reach underlying sites and suggested that cytotoxicity may be due to
7 NO2 reactants formed in the ELF. Since then, the reactive absorption of NO2 has been examined
8 in a number studies that have sought to identify reactive substrates for NO2 and quantify the
9 absorption kinetics of NO2 in the respiratory tract.
10 Postlethwait and Bidani (1994) concluded that the reaction between NO2 and water does
11 not significantly contribute to the absorption of inhaled NO2. Uptake is a first-order process for
12 NO2 concentrations less than 10 ppm, is aqueous substrate-dependent, and is saturable. The
13 absorption of inhaled NO2 is thought to be coupled with free radical-mediated hydrogen
14 abstraction to form FDSTO2 and an organic radical (Postlethwait and Bidani, 1989, 1994). At
15 physiologic pH, the HNO2 subsequently dissociates to H+ and nitrite (NO2 ). The concentration
16 of the resulting nitrite is thought insufficient to be toxic, so effects are thought to be due to the
17 organic radical and/or the proton load. Nitrite may enter the underlying epithelial cells and
18 blood. In the presence of red blood cells, nitrite is oxidized to nitrate (N(V) (Postlethwait and
19 Mustafa, 1981). Beyond cell susceptibility and the concentration of NO2 in the lumen, site-
20 specific injury was proposed to depend on rate of 'toxic' reaction product formation and the
21 quenching of these products within the ELF. Related to the balance between reaction product
22 formation and removal, it was further suggested that cellular responses may be nonlinear with
23 greater responses being possible at low levels of NO2 uptake versus higher levels of uptake.
24 Since the ELF may vary throughout the respiratory tract, the uptake of inhaled NO2 and reaction
25 with constituents of the pulmonary ELF may be related to the heterogeneous distribution of
26 epithelial injury observed from NO2 exposure.
27
28 Postlethwait et al. (1995) sought to determine the absorption substrates for NO2 in the
29 ELF lavaged from male Sprague-Dawley rats. Since the bronchoalveolar lavage fluid (BALF)
30 collected from the rats may be diluted up to 100-fold relative to the native ELF, the effect of
31 concentrating the BAL fluid on NO2 absorption was investigated. A linear association was
March 2008 AX4-8 DRAFT-DO NOT QUOTE OR CITE
-------
1 found between the first-order rate constant for NO2 absorption and the concentration of the
2 BALF. This suggests that concentration of the reactive substrates in the ELF determines the rate
3 of NO2 absorption. The absorption due to specific ELF constituents was also examined in
4 chemically pure solutions. Albumin, cysteine, reduced glutathione (GSH), ascorbic acid, and
5 uric acid were hydrophilic moieties found to be active substrates for NO2 absorption.
6 Unsaturated fatty acids (such as oleic, linoleic, and linolenic) were also identified as active
7 absorption substrates and thought to account for up to 20% of NO2 absorption. Vitamins A and
8 E exhibited the greatest reactivity of the substrates that were examined. However, the low
9 concentrations of uric acid and vitamins A and E were thought to preclude them from being
10 appreciable substrates in vivo. The authors concluded that ascorbate and GSH were the primary
11 NO2 absorption substrates in rat ELF. Postlethwait et al. (1995) also found that the pulmonary
12 surfactant, dipalmitoyl phosphatidylcholine, was not an effective substrate for NO2 absorption.
13 Later, Connor et al. (2001) suggested that dipalmitoyl phosphatidylcholine may actually inhibit
14 NO2 absorption.
15 In a subsequent study, Velsor and Postlethwait (1997) investigated the mechanisms of
16 acute epithelial injury from NO2 exposure. The impetus for this work was to evaluate the
17 supposition that NO2 reaction products rather than NO2 itself cause epithelial injury. Red blood
18 cell membranes were immobilized to the bottom of Petri dishes, covered with a variety of well
19 characterized aqueous layers, and exposed to gaseous NO2 (10 ppm for 20 min). The study
20 focused on the potential roles of GSH and ascorbic acid reaction products in mediating cellular
21 injury. Based on negligible membrane oxidation when covered with only an aqueous phosphate
22 buffer, the diffusive/reactive resistance of a thin aqueous layer clearly prevented direct
23 interaction between NO2 and the underlying membrane. The presence of unsaturated fatty acids
24 was not observed to affect NO2 absorption, but a sufficiently thin liquid layer was required for
25 membrane oxidation to occur. Interestingly, membrane oxidation was not a simple monotonic
26 function of GSH and ascorbic acid levels. The maximal levels of membrane oxidation were
27 observed at low antioxidant levels versus null or high antioxidant levels. Glutathione and
28 ascorbic acid related membrane oxidation were superoxide and hydrogen peroxide dependent,
29 respectively. The authors suggested that at the higher antioxidant concentrations, there was
30 increased absorption of NO2, but little secondary oxidation of the membrane because the reactive
31 species (e.g., superoxide and hydrogen peroxide) generated during absorption were quenched.
March 2008 AX4-9 DRAFT-DO NOT QUOTE OR CITE
-------
1 At the low antioxidant concentrations, there was a lower rate of NO2 absorption, but oxidants
2 were not quenched and so were available to interact with the cell membrane.
3 Kelly et al. (1996a) examined the effect of a 4-h NO2 (2 ppm) exposure on antioxidant
4 levels in bronchial lavage fluid (BLF) and BALF of 44 healthy nonsmoking adults (19-45 year,
5 median 24 years). Subjects were randomly assigned to three groups and lavaged at either 1.5 h
6 (n = 15), 6 h (n = 15), or 24 h (n = 14) after the NO2 exposure. The baseline concentrations of
7 uric acid and ascorbic acid were strongly correlated between the BLF and BALF within
8 individuals (r = 0.88, p < 0.001; r = 0.78, p = 0.001; respectively), whereas the concentrations of
9 GSH in the BLF and BALF were not correlated. Uric acid levels in both lavage fractions were
10 significantly reduced at 1.5 h (p < 0.04), significantly increased at 6 h (p < 0.05), and back to
11 baseline at 24 h postexposure. A statistically significant loss of ascorbic acid was also found in
12 both lavage fractions at 1.5 h (p < 0.05). At 6 and 24 h postexposure, the ascorbic acid levels
13 had returned to baseline. In contrast, GSH levels were significantly increased at both 1.5 h
14 (p < 0.01) and 6 h (p < 0.03) in BLF. At 24 h postexposure, the GSH levels in BLF returned to
15 baseline. Although GSH in BLF increased at 1.5 and 6 h postexposure, oxidized GSH levels
16 remained similar to baseline in both BLF and BALF. No changes in BALF levels of GSH were
17 observed at any time point.
18 The depletion of uric acid and ascorbic acid, but not GSH has also been observed with
19 ex vivo exposure of human BALF to NO2. Kelly et al. (1996b) collected BALF from male lung
20 cancer patients (n = 16) and exposed the BALF ex vivo at 37°C to NO2 (0.05 to 2.0 ppm; 4 h) or
21 O3 (0.05 to 1.0 ppm; 4 h). Kelly and Tetley (1997) also collected BALF from lung cancer
22 patients (n = 12, 54 + 16 years) and exposed the BALF ex vivo to NO2 (0.05 to 1.0 ppm; 4 h).
23 Both studies found that NO2 depletes uric acid and ascorbic acid, but not GSH from BALF.
24 Kelly et al. (1996b) noted a differential consumption of the antioxidants with uric acid loss being
25 greater than that of ascorbic acid which was lost at a much greater rate than GSH. Kelly and
26 Tetley (1997) found that the rates of uric acid and ascorbic acid consumption were correlated
27 with their initial concentrations in the BAL fluid, such that higher initial antioxidant
28 concentrations were associated with a greater rate of antioxidant depletion. Illustrating the
29 complex interaction of antioxidants, these studies also suggest that GSH oxidized by NO2 may
30 be again reduced by uric acid and/or ascorbic acid.
31
March 2008 AX4-10 DRAFT-DO NOT QUOTE OR CITE
-------
1 AX4.2.2 Regional and Total Respiratory Absorption of NO2
2 There has been very limited work related to the quantification of NC>2 uptake since the
3 1993 AQCD for Oxides of Nitrogen. As a result, there is an abbreviated discussion here of some
4 papers that were reviewed in the 1993 AQCD for Oxides of Nitrogen.
5
6 AX4.2.2.1 Dosimetry Models
7 There is a paucity of theoretical studies investigating NO2 dosimetry. Like Os, NO2 is
8 highly reactive in ELF and is not very soluble. An Os model has been utilized to predict the
9 uptake of NO2 in the lower respiratory tract of humans, rats, guinea pigs, and rabbits (Miller
10 et al., 1982; Overton, 1984). In this model, there was a strong distinction between uptake and
11 dose. Uptake referred to the amount of NO2 being removed from gas phase per lung surface area
12 (|ig/cm2), whereas, dose referred to the amount of NO2 per lung surface area (jig/cm2) that
13 diffused through the ELF and reached the underlying tissues. These investigators assessed NO2
14 uptake and dose on a breath by breath basis. Miller et al. (1988) provided uptake and dose rates
15 (|ig/cm2-min) for Oj, in the same species.
16 Miller et al. (1982) and subsequently Overton (1984) did not attempt to predict the
17 amount of reactants in the ELF or the transport of reactants to the tissues. Rather, they focused
18 mainly on the sensitivity of NO2 tissue dose on NO2 reaction rates in the ELF and the Henry's
19 law constant. Reaction rates of NO2 in the ELF were varied from zero, 50%, and 100% of the
20 reaction rate for Os in ELF. The Henry's law constant was varied from half to double the
21 Henry's law constant for NO2 in water at 37 °C. Effects of species, lung morphology, and tidal
22 volume (Vx) were also examined. In general, the model predicted that NO2 is taken up
23 throughout the lower respiratory tract. In humans, NO2 uptake was fairly constant from the
24 trachea to the terminal bronchioles, beyond which uptake decreased with distal progression.
25 This pattern of NO2 uptake predicted for humans is very similar to the pattern of Os uptake per
26 unit time predicted for humans, rats, rabbits, and guinea pigs by Miller et al. (1988). Thus, it is
27 reasonable to expect that the pattern NO2 uptake per unit time will also be similar between these
28 species. The NO2 tissue dose was highly dependent on the Henry's law constant and reaction
29 rate in the ELF. In the conducting airways, the NO2 tissue dose decreased as the Henry's law
30 constant increased (i.e., decreased gas solubility), whereas the NO2 tissue dose in the alveolar
31 region increased. The site of maximal NO2 tissue dose was fairly similar between species,
March 2008 AX4-11 DRAFT-DO NOT QUOTE OR CITE
-------
1 ranging from the first generation of respiratory bronchioles in humans to the alveolar ducts in
2 rats. In guinea pigs and rabbits, the maximal NC>2 tissue dose was predicted to occur in the last
3 generation of respiratory bronchioles. Based on Miller et al. (1988), the dose rate of NO2 is also
4 expected to be similar between species. The simulations also showed that exercise increases the
5 NC>2 tissue dose in the pulmonary region relative to rest. Miller et al. (1982) also reported that
6 increasing the NC>2 reaction rate decreased NO2 tissue dose in the conducting airways, but had no
7 effect on the dose delivered to the pulmonary region.
8 Simultaneously occurring diffusion and chemical reactions in the ELF have been
9 suggested as the limiting factors in 63 (Santiago et al., 2001) and NO2 uptake (Postlethwait and
10 Bidani, 1990). Hence, Miller et al. (1982) should have found an increase in the uptake of NC>2 in
11 the conducting airways with increasing the rate of chemical reactions in the ELF. This increase
12 in NC>2 uptake in the conducting airways would then lead to a reduction in the amount of NC>2
13 reaching and taken up in the pulmonary region. The Miller et al. (1982) model considered
14 reactions of NC>2 with constituents in the ELF as protective in that these reactions reduced the
15 flux of NC>2 to the tissues. Others have postulated that NCVreactants formed in the ELF, rather
16 than NC>2 itself, could actually cause adverse responses (Overton, 1984; Postlethwait and Bidani,
17 1994; Velsor and Postlethwait, 1997).
18 Overton and Graham (1995) examined NO2 uptake in an asymmetric anatomic model of
19 the rat lung. The multiple path model of Overton and Graham (1995) allowed for variable path
20 lengths from the trachea to the terminal bronchioles, whereas Miller et al. (1982) used a single or
21 typical path model of the conducting airways. The terms dose and uptake were used
22 synonymously to describe the amount of NO2 gas lost from the gas phase in a particular lung
23 region or generation by Overton and Graham (1995). Reactions of NO2 in the ELF were not
24 explicitly considered. Their simulations were conducted for rats breathing at 2 mL VT at a
25 frequency of 150 breaths per minute. The mass transfer coefficients of 0.173, 0.026, and 0.137
26 cm/sec were assumed for the upper respiratory tract, the tracheobronchial airways, and the
27 pulmonary region, respectively. Uptake was predicted to decrease with distal progression into
28 the lung. In general, the modeled NO2 dose varied among anatomically equivalent ventilatory
29 units as a function of path length from the trachea with shorter paths showing greater dose. A
30 sudden increase in NO2 uptake was predicted in the proximal alveolar region (PAR) which was
31 due to the increase in the assumed mass transfer coefficient relative to the adjacent terminal
March 2008 AX4-12 DRAFT-DO NOT QUOTE OR CITE
-------
1 bronchiole. Overton et al. (1996) showed that increasing the mass transfer coefficient of the
2 tracheobronchial airways would decrease the dose to the PAR and vice versa. Additionally, the
3 PAR dose would also be reduced by the more realistic modeling of tracheobronchial airways
4 expansion during inspiration versus the static condition employed by Overton and Graham
5 (1995).
6 More recently, two studies examined the influence of age on reactive gas dosimetry in
7 humans (Ginsberg et al., 2005; Sarangapani et al., 2003). Both studies specifically considered
8 the dosimetry 63 during light activity (on average) in their analysis. It is assumed here that their
9 general findings should also be applicable to NO2. Sarangapani et al. (2003) used a
10 physiologically based pharmacokinetic model and found that regional uptake of 63 is relatively
11 insensitive to age (range: infants to elderly). Ozone uptake per unit surface area was 2- to 8-fold
12 higher in infants compared to adults. However, this finding (i.e., uptake per unit surface area) is
13 a less informative expression of dose than the rate of uptake per unit surface area. The rate of
14 uptake, obtained by multiplying by the ventilation rate, adjusts for the greater rate of gas intake
15 by adults relative to children. Ginsberg et al. (2005) utilized the U.S. EPA (1994) reference
16 concentration methodology and found no effect of age (infants vs. adults) on the uptake rate of
17 63 per unit surface area.
18 In summary, these modeling studies predict that the net NO2 dose (NO2 flux to air-liquid
19 interface) is relatively constant from the trachea to the terminal bronchioles and then rapidly
20 decreases in the pulmonary region. The pattern of net NO2 dose rate or uptake rate is expected to
21 be similar between species and unaffected by age in humans. The predicted tissue dose and dose
22 rate of NO2 (NO2 flux to liquid-tissue interface) is low in the trachea, increases to a maximum in
23 the terminal bronchioles and the first generation of the pulmonary region, and then decreases
24 rapidly with distal progression. The site of maximal NO2 tissue dose is predicted to be fairly
25 similar between species, ranging from the first generation of respiratory bronchioles in humans
26 to the alveolar ducts in rats. The production of toxic NO2-reactants in the ELF and the
27 movement of the reactants to the tissues have not been modeled.
28
29
30 AX4.3 EXPERIMENTAL STUDIES OF NO2 UPTAKE
31
March 2008 AX4-13 DRAFT-DO NOT QUOTE OR CITE
-------
1 AX4.3.1 Upper Respiratory Tract Absorption
2 The nasal uptake of NO2 has been experimentally measured in dogs, rabbits, and rats
3 under conditions of unidirectional flow. Yokoyama (1968) reported 42.1 + 14.9%
4 (Mean + StDev) uptake of NO2 in the isolated nasal passages of two dogs (3.5 L/min) and three
5 rabbits (0.75 L/min) exposed to 4 and 41 ppm NC>2. Uptake did not appear to depend on the
6 exposure concentration and was relatively constant over a 10 to 15 min period. Cavanagh and
7 Morris (1987) measured uptakes of 28% and 25% uptake of NC>2 (40.4 ppm) in the noses of four
8 naive and four previously exposed rats (0.10 L/min), respectively. Uptake was not affected by a
9 4-h prior exposure (naive versus previously exposed rats) to 40.4 ppm NC>2 and was constant
10 over the 24-min period during which uptake was determined.
11 Kleinman and Mautz (1991) measured the penetration of NO2 through the upper airways
12 during inhalation in six tracheotomized dogs exposed to 1.0 or 5.0 ppm NC>2. Uptake in the nasal
13 passages was significantly greater at 1.0 ppm than at 5.0 ppm, although the magnitude of this
14 difference was not reported. The mean uptake of NC>2 (1.0 ppm) in the nasal passages decreased
15 from 55% to 40% as the ventilation rate increased from about 2 to 8 L/min. During oral
16 breathing, uptake was not dependent on concentration. The mean oral uptake of NC>2 (1.0 and
17 5.0 ppm) decreased from 65% to 30% as the ventilation rate increased from 2 to 8 L/min.
18
19 AX4.3.2 Lower Respiratory Tract Absorption
20 Postlethwait and Mustafa (1989) investigated the effect of exposure concentration and
21 breathing frequency on the uptake of NC>2 in isolated perfused rat lungs. To evaluate the effect
22 of exposure concentration, the lungs were exposed to NC>2 (4 to 20 ppm) while ventilated at 50
23 breaths/min with a VT of 2.0 mL. To examine the effect of breathing frequency, the lungs were
24 exposed to NC>2 (5 ppm) while ventilated at 30-90 breaths/min with a VT of 1.5 mL. All
25 exposures were for 90 min. The uptake of NC>2 ranged from 59 to 72% with an average of 65%
26 and was not affected by exposure concentration or breathing frequency. A combined regression
27 showed a linear relationship between NC>2 uptake and total inspired dose (25 to 330 jig NO2).
28 Illustrating variability in NC>2 uptake measurements, Postlethwait and Mustafa (1989) observed
29 59% NO2 uptake in lungs ventilated at 30 breaths/min with a VT of 1.5 mL, whereas,
30 Postlethwait and Mustafa (1981) measured 35% NC>2 uptake for the same breathing condition.
31 In another study, 73% uptake of NC>2 was reported for rat lungs ventilated 50 breaths/min with a
March 2008 AX4-14 DRAFT-DO NOT QUOTE OR CITE
-------
1 VT of 2.3 mL (Postlethwait et al., 1992). It should be noted that typical breathing frequencies are
2 around 80, 100, and 160 breaths/min for rats during sleep, rest, and light exercise, respectively
3 (Winter-Sorkina and Cassee, 2002). Hence, the breathing frequencies at which NO2 uptake has
4 been measured are lower than for rats breathing normally.
5 In addition to measuring upper respiratory tract uptakes, Kleinman and Mautz (1991) also
6 measured NO2 uptake in the dog lung. In general, there was about 90% NO2 uptake in the lungs
7 which was independent of ventilation rates from 3 to 16 L/min.
8
9 AX4.3.3 Total Respiratory Tract Absorption
10 Bauer et al. (1986) measured the uptake of NO2 (0.3 ppm) in 15 adult asthmatics exposed
11 for 30 min (20 min at rest, then 10 min exercising on a bicycle ergometer) via a mouthpiece
12 during rest and exercise. There was a statistically significant increase in uptake from 72% during
13 rest to 87% during exercise. The minute ventilation also increased from 8.1 L/min during rest to
14 30.4 L/min during exercise. Hence, exercise increased the dose rate of NO2 by 5-fold in these
15 subjects. In an earlier study of seven healthy adults in which subjects were exposed to a nitric
16 oxide (NO)/NO2 mixture containing 0.29 to 7.2 ppm NO2 for brief (but unspecified) periods,
17 Wagner (1970) reported that NO2 uptake increased from 80% during normal respiration (Vx, 0.4
18 L) to 90% during maximal respiration (Vx, 2 to 4 L).
19 Kleinman and Mautz (1991) also measured the total respiratory tract uptake of NO2 (5
20 ppm) in female beagle dogs while standing at rest or exercising on a treadmill. The dogs
21 breathed through a small face mask. Total respiratory tract uptake of NO2 was 78% during rest
22 and increased to 94% during exercise. In large part, this increase in uptake may be due to the
23 increase in Vx from 0.18 L during rest to 0.27 L during exercise. Coupled with an increase in
24 minute ventilation from 3.8 L/min during rest to 10.5 L/min during exercise, the dose rate of
25 NO2 was 3-fold greater for the dogs during exercise than rest.
26
27
28 AX4.4 METABOLISM, DISTRIBUTION AND ELIMINATION OF NO2
29 PRODUCTS
30 As stated earlier, NO2 absorption is coupled with nitrous acid (HNO2) formation, which
31 subsequently dissociates to H+ and nitrite (NO2 ). Nitrite enters the underlying epithelial cells
32 and subsequently the blood. In the presence of red blood cells and possibly involving
March 2008 AX4-15 DRAFT-DO NOT QUOTE OR CITE
-------
1 oxyhemoglobin, nitrite is oxidized to nitrate (NO3 ) (Postlethwait and Mustafa, 1981). Nitrate
2 may subsequently be excreted in the urine. There has been concern that inhaled NO2 may lead to
3 N-nitrosamine production, many of which are carcinogenic, since NO2 can produce nitrite and
4 nitrate (in blood). Nitrate can be converted to nitrite by bacterial reduction in saliva, the
5 gastrointestinal tract, and the urinary bladder. Nitrite has been found to react with secondary
6 amines to form N-nitrosamines. This remains speculative since nitrosamines are not detected in
7 tissues of animals exposed by inhalation to NO2 unless precursors to nitrosamines and/or
8 inhibitors of nitrosamine metabolism are co-administered. Rubenchik et al. (1995) could not
9 detect N-nitrosodimethylamine (NDMA) in tissues of mice exposed to 4 to 4.5 ppm NO2 for 1 h.
10 NDMA was found in tissues, however, if mice were simultaneously given oral doses of
11 amidopyrine and 4-methylpyrazole, an inhibitor of NDMA metabolism. Nevertheless, the main
12 source of NO2 in the body is formed endogenously, and food is also a contributing source of
13 nitrite from the conversion of nitrates. Thus, the relative importance of inhaled NO2 to N-
14 nitrosamine formation has yet to be demonstrated.
15 Metabolism of inhaled NO2 may also transform other chemicals that may be present in
16 the body, in some cases into mutagens and carcinogens. Van Stee et al. (1983) reported
17 N-nitrosomorpholine (NMOR), production in mice gavaged with 1 g of morpholine/kg body
18 weight per day and then exposed (5-6 h daily for 5 days) to 16.5 to 20.5 ppm NO2.
19 N-nitrosomorpholine is a nitrosamine that is a potent animal carcinogen. The single site
20 containing the greatest amount of NMOR was the gastrointestinal tract. Later, Van Stee et al.
21 (1995) exposed mice to approximately 20 ppm 15NO2 and to 1 g/kg morpholine simultaneously.
22 N-nitrosomorpholine was found in the body of the exposed mice. Ninety-eight point four
23 percent was labeled with 15N that was derived from the inhaled 15NO2 and 1.6% was derived
24 presumably from endogenous sources.
25 Miyanishi et al. (1996) co-exposed rats, mice, guinea pigs and hamsters to NO2 and
26 various polycyclic aromatic hydrocarbons (PAHs) such as pyrene, fluorene, or anthracene. Nitro
27 derivatives of these PAHs were excreted in the urine of co-exposed animals, which were found
28 to be highly mutagenic in the Ames/S. typhimurium assay. Specifically, the nitrated metabolite
29 of pyrene (l-nitro-6/8-hydroxypyrene and l-nitro-3hydroxypyrene) was detected in the urine.
30 Further studies indicated that these metabolites are nitrated by an ionic reaction in vivo after the
31 hydroxylation of pyrene in the liver.
March 2008 AX4-16 DRAFT-DO NOT QUOTE OR CITE
-------
1 Inhaled NO2 may also be involved in the production of mutagenic (and carcinogenic)
2 nitro derivatives of other co-exposed compounds, such as poly cyclic aromatic hydrocarbons
3 (PAHs), via nitration reactions. Miyanishi et al. (1996) co-exposed rats, mice, guinea pigs and
4 hamsters to 20 ppm NC>2 and various PAHs (pyrene, fluoranthene, fluorene, anthracene, or
5 chrysene). Nitro derivatives of these PAHs were excreted in the urine of these animals, which
6 were found to be highly mutagenic in the AmesA?. typhimurium assay. Specifically, the nitrated
7 metabolite of pyrene (l-nitro-6/8-hydroxypyrene and l-nitro-3hydroxypyrene) was detected in
8 the urine. Further studies indicated that these metabolites are nitrated by an ionic reaction in
9 vivo after the hydroxylation of pyrene in the liver.
10
11
12 AX4.5 EXTRA-PULMONARY EFFECTS OF NO2 AND NO
13 Exposure to NC>2 produces a wide array of health effects beyond the confines of the lung.
14 Thus, NC>2 and/or some of its reactive products penetrate the lung or nasal epithelial and
15 endothelial layers to enter the blood and produce alteration in blood and various other organs.
16 Effects on the systemic immune system were discussed above and the summary of other
17 systemic effects is quite brief because the database suggests that effects on the respiratory tract
18 and immune response are of greatest concern. A more detailed discussion of extrapulmonary
19 responses can be found in U.S. Environmental Protection Agency (1993).
20
21 AX4.5.1 Body Weight, Hepatic, and Renal Effects
22 Conflicting results have been reported on whether NO2 affects body weight gain in
23 experimental animals as a general indicator of toxicity (U.S. Environmental Protection Agency,
24 1993). Newer subchronic studies show no significant effects on body weight in rats, guinea
25 pigs, and rabbits exposed up to 4 ppm NC>2 (Tepper et al., 1993; Douglas et al., 1994; Fujimaki
26 and Nohara, 1994).
27 Effects on the liver, such as changes in serum chemistry and xenobiotic metabolism, have
28 been reported by various investigators to result from exposure to NC>2 (U.S. Environmental
29 Protection Agency 1993). Drozdz et al. (1976) found decreased total liver protein and sialic
30 acid, but increased protein-bound hexoses in guinea pigs exposed to 1.05 ppm NO2, 8 h/day for
31 180 days. Liver alanine and aspartate aminotransferase activity was increased in the
32 mitochondrial fraction but decreased in the cytoplasmic fraction of the liver. Electron
March 2008 AX4-17 DRAFT-DO NOT QUOTE OR CITE
-------
1 micrographs of the liver showed intracellular edema and inflammatory and parenchymal
2 degenerative changes.
3 No new studies on liver effects were located in the literature since the 1993 AQCD for
4 Oxides of Nitrogen. Several older studies have shown changes in kidney function and
5 xenobiotic metabolism in animals following NO2, although no histopathological changes were
6 reported.
7 Increases in urinary protein and specific gravity of the urine were reported by Sherwin
8 and Layfield (1974) in guinea pigs exposed continuously to 0.5 ppm NO2 for 14 days.
9 Proteinuria (albumin and alpha-, beta-, and gamma-globulins) was found in another group of
10 animals when the exposure was reduced to 0.4 ppm NO2 for 4 h/day. However, differences in
11 water consumption or in the histology of the kidney were not found. No new studies were
12 located in the literature since the 1993 AQCD for Oxides of Nitrogen.
13
14 AX4.5.2 Brain Effects
15 There are several studies suggesting that NO2 affects the brain. Decreased activity of
16 protein metabolizing enzymes, increased glycolytic enzymes, changes in neurotransmitter levels
17 (5-HT and noradrenaline), and increased lipid peroxidation, accompanied by lipid profile and
18 antioxidant changes, have been reported (Farahani and Hasan, 1990, 1991, 1992; Sherwin et al.,
19 1986; Drozdz et al., 1975). The U.S. Environmental Protection Agency (1993) concluded that
20 "none of these effects have been replicated and all reports lack sufficient methodological rigor;
21 thus, the implications of these findings, albeit important, are not clear and require further
22 investigation".
23 A developmental neurotoxicity study by Tabacova et al.(1985) suggest that in utero
24 exposure to NO2 may result in postnatal neurobehavioral development changes as described in
25 the section on reproductive and developmental toxicology.
26 AX4.5.3 NO
27 The genotoxicity of NO has been studied both in vitro and in vivo (Arroyo et al., 1992;
28 Nguyen et al., 1992) (see Table AX4.8). Overall, the synthesis of these older studies suggests
29 that NO has some genotoxic potential; however, the effect is slight and to a lesser extent when
30 compared to NO2.
March 2008 AX4-18 DRAFT-DO NOT QUOTE OR CITE
-------
1 AX4.6 EFFECTS OF MIXTURES CONTAINING NO2
2 Humans are generally exposed to NO2 in a mixture with other air pollutants. A limitation
3 of animal toxicity studies is the extrapolation of dose-response data from controlled exposures to
4 NO2 alone to air pollutant mixtures that are typically found in the environment. It is difficult to
5 predict the effects of NO2 in a mixture based on the effects of NO2 alone. In order to understand
6 how NO2 is affected by mixtures of other air pollutants, studies are typically conducted with
7 mixtures containing NO2 and one or two other air pollutants, such as Os and/or H2SC>4. The
8 result of exposure to two or more pollutants may be simply the sum of the responses to
9 individual pollutants (additivity), may be greater than the sum of the individual responses,
10 suggesting some type of interaction or augmentation of the response (synergism) or may be less
11 than additive (antagonism).
12 Animal toxicity studies have shown an array of interactions, including no interaction,
13 additivity or synergism. Because no clear understanding of NO2 interactions has yet emerged
14 from this database, only a brief overview is provided here. A more substantive review can be
15 found in U.S. Environmental Protection Agency (1993). There are animal studies, which have
16 studied the effects of ambient air mixtures containing NO2 or gasoline or diesel combustion
17 exhausts containing NOx. Generally these studies provide useful information on the mixtures,
18 but lack NO2-only groups, making it impossible to discern the influence of NO2. Therefore, this
19 class of research is not described here, but is reviewed elsewhere (U.S. Environmental Protection
20 Agency, 1993).
21
22 AX4.6.1 Simple Mixtures Containing NO2
23 Most of the interaction studies have involved NO2 and 63. After subchronic exposure,
24 lung morphology studies did not show any interaction of NO2 with O3 (Freeman et al., 1974) or
25 with SO2 (Azoulay et al., 1980). Some biochemical responses to NO2 plus Os display no
26 positive interaction or synergism. For example, Mustafa et al. (1984) found synergism for some
27 endpoints (e.g., increased activities of O2 consumption and antioxidant enzymes), but no
28 interaction for others (e.g., DNA or protein content) in rats exposed for 7 days. Ichinose and
29 Sagai (1989) observed a species dependence in regard to the interaction of 63 (0.4 ppm) and NO2
30 (0.4 ppm) after 2 weeks of exposure. Guinea pigs, but not rats, had a synergistic increase in lung
31 lipid peroxides. Rats, but not guinea pigs, had synergistic increases in antioxidant factors (e.g.,
March 2008 AX4-19 DRAFT-DO NOT QUOTE OR CITE
-------
1 non-protein thiols, vitamin C, glucose-6-phosphate dehydrogenase, GSH peroxidase). Duration
2 of exposure can have an impact. Schlesinger et al. (1990) observed a synergistic increase in
3 prostaglandin E2 and F2a in the lung lavage of rabbits exposed acutely for 2 h to 3.0 ppm NO2
4 plus 0.3 ppm Os; the response appeared to have been driven by Os. However, with 7 or 14 days
5 of repeated 2-h exposures, only prostaglandin E2 was decreased and appeared to have been
6 driven by NO2; there was no synergism (Schlesinger et al., 1991).
7 Using an infectivity model, Ehrlich et al. (1977) found additivity after acute exposure to
8 mixtures of NO2 and 63 and synergism after subchronic exposures. Exposure scenarios
9 involving NO2 and 63 have also been performed using a continuous baseline exposure to one
10 concentration or mixture, with superimposed short-term peaks to a higher level (Ehrlich et al.,
11 1979; Gardner, 1980, 1982; Graham et al., 1987). Differences in the pattern and concentrations
12 of the exposure are responsible for the increased susceptibility to pulmonary infection, without
13 indicating clearly the mechanism controlling the interaction.
14 Some aerosols may potentiate response to NO2 by producing local changes in the lungs
15 that enhance the toxic action of co-inhaled NO2. The impacts of NO2 and H2SC>4 on lung host
16 defenses have been examined by Schlesinger and Gearhart (1987) and Schlesinger (1987). In the
17 former study, rabbits were exposed for 2 h/day for 14 days to either 0.3 ppm or 1.0 ppm NO2, or
18 500 |ig/m3 H2SC>4 alone, or to mixtures of the low and high NO2 concentrations with H2SC>4.
19 Exposure to either concentration of NO2 accelerated alveolar clearance, whereas H2SC>4 alone
20 retarded clearance. Exposure to either concentration of NO2 with H2SO4 resulted in retardation
21 of clearance in a similar manner to that seen with H2SC>4 alone. Using a similar exposure design
22 but different endpoints, exposure of rabbits to 1.0 ppm NO2 increased the numbers of PMNs in
23 lavage fluid at all time points (not seen with either pollutant alone), and increased phagocytic
24 capacity of AMs after two or six exposures (Schlesinger et al., 1987). Exposure to 0.3 ppm NO2
25 with acid, however, resulted in depressed phagocytic capacity and mobility. The NO2/H2SO4
26 mixture was generally either additive or synergistic, depending on the specific cellular endpoint
27 being examined.
28 Exposure to high levels of NO2 (<5.0 ppm) with very high concentrations of H2SC>4
29 (1 mg/m3) caused a synergistic increase in collagen synthesis rate and protein content of the
30 lavage fluid of rats (Last and Warren, 1987; Last, 1989).
March 2008 AX4-20 DRAFT-DO NOT QUOTE OR CITE
-------
1 AX4.6.2 Complex Mixtures Containing NO2
2 Although many studies have examined the response to NC>2 with only one additional
3 pollutant, the atmosphere in most environments is a complex mixture of more than two materials.
4 A number of studies have attempted to examine the effects of multi-component atmospheres
5 containing NO2, but as mentioned before, in many cases the exact role of NC>2 in the observed
6 responses is not always clear. One study by Stara et al. (1980) deserves mention because
7 pulmonary function changes appeared to progress after exposure ceased.
8 In the study by Stara et al. (1980), dogs were exposed for 68 months (16 h/day) to raw or
9 photochemically reactive vehicle exhaust which included mixtures of NOx: one with a high NC>2
10 level and a low NO level (0.64 ppm, NO2; 0.25 ppm, NO), and one with a low NO2 level and a
11 high NO level (0.14 ppm, NO2; 1.67 ppm, NO). At the end of exposure, the animals were
12 maintained for about 3 years in normal indoor air. Numerous pulmonary functions,
13 hematological and histological endpoints were examined at various times during and after
14 exposure. The lack of an NO2-only or NO-only group precludes determination of the nature of
15 the interaction. Nevertheless, the main findings are of interest. Pulmonary function changes
16 appeared to progress after exposure ceased. Dogs in the high NO2 group had morphological
17 changes considered to be analogous to human centrilobular emphysema. Because these
18 morphological measurements were made after a 2.5- to 3-year holding period in clean air, it
19 cannot be determined with certainty whether these disease processes abated or progressed during
20 this time. This study suggests progression of damage after exposure ends.
March 2008 AX4-21 DRAFT-DO NOT QUOTE OR CITE
-------
TABLE AX4.1. EFFECTS OF NITROGEN DIOXIDE ON OXIDANT AND ANTIOXIDANT HOMEOSTASIS
O
to
O
O
oo
>
^
to
to
O
>
Tj
H
6
O
0
H
/O
r*^x
O
H
W
O
O
HH
H
W
ppm
0.04
0.4
4.0
0.4
0.4
1.2
4.0
0.4-0.5
0.5
1.0
1.0
1.0
2.3
6.2
1.2
1.8
2.0
10.0
Exposure Gender Age
Continuous, M 8 wks
9 and 18 mos
2 wks NS NS
Continuous, M 13 wks
4 mos
Continuous, F NS
1.5yrs
Continuous, F 4 wks
17 mos
4 h/day, NS NS
6 days
Continuous, M 8 wks
4 days
Continuous, M 12 wks
3 days
3 days M/F 5->60
days
Species (Strain)
Rat (Wistar)
Rat
Guinea Pig
Rat (Wistar)
Mouse
(NS)
Mouse (C57B1/6J)
Rat (Sprague-Dawley)
Rat
(Sprague-Dawley)
Rat (Sprague-Dawley)
Rat (Wistar) Guinea pig
(Dunkin Hartley)
Effects
NPSHs increased at >0.4 ppm after 9 or 18 mos;
GSH peroxidase activity increased after a 9-mo
exposure to 4.0 ppm; G-6-P dehydrogenase was
increased after a 9- and 18-mo exposure to 4.0 ppm;
no effects on 6-P-G dehydrogenase , SOD disulfide
reductase; some GSH S-transferase had decreased
activities after 18-mo exposure to >0.4 ppm.
No effect on TEA reactants, antioxidants, or
antioxidant enzyme activities.
Duration dependent pattern for increase in activities
of antioxidant enzymes; increase, peaking at wk 4
and then decreasing. Concentration-dependent
effects.
Growth reduced; Vitamin E (30 or 300 mg/kg diet)
improved growth.
At 1 ppm, GSH-peroxidase activity decreased in
vitamin E-deficient mice and increased in Vitamin
E- supplemented mice.
Vitamin E-supplement reduced lipid peroxidation.
Activities of GSH reductase and G-6-P
dehydrogenase increased at 6.2 ppm proportional to
duration of exposure; plasma lysozyme and GSH
peroxidase not affected at 6.2 ppm; no effects at 1.0
or 2.3 ppm.
Increases in G-6-P dehydrogenase, isocitrate
dehydrogenase, disulfide reductase, and NADPH
cytochrome c reductase activities at 1.8 ppm only.
Decreased SOD activity in 21 -day -old animals.
References
Sagai etal. (1984)
Ichinose etal. (1983)
Ichinose and Sagai
(1989)
Ichinose and Sagai
(1982)
Csallany (1975)
Ayaz and Csallany
(1978)
Thomas etal. (1967)
Chow etal. (1974)
Lee etal. (1989, 1990)
Azoulay-Dupuis et al.
(1983)
-------
TABLE AX4.1 (cont'd). EFFECTS OF NITROGEN DIOXIDE ON OXIDANT AND ANTIOXIDANT HOMEOSTASIS
o Ppm
5 2.0
8 4.0
00 10.0
3.0
9.5
3.0
Exposure
14 days
10 days
7 days
7 days
7 h/day,
5 days/wks,
6 mos
4 days
Gender Age
M
M/F
M
M
12-24 wks
1 day to
>8wks
In utero
and 6 mos
NS
Species (Strain)
Rat (Wistar)
Rat (Sprague-Dawley)
Rat (Fischer 344)
Rat (Sprague-Dawley)
Effects
G-6-P dehydrogenase increased at >2 ppm; at
2 ppm, 14 days of exposure needed
Increased lipid peroxidation (TBA-reactive
substances) with vitamin E deficiency.
Increase in GSH reductase activity in younger rats
and SDH peroxidase activity in older rats.
No effects on parameters tested.
References
Mochitate et al.
(1985)
Sevanian et al. (1982)
Mauderly et al. (1987)
Mustafa etal. (1979)
7.0
4 days
Increase in lung weight, G-6-P dehydrogenase,
GSH reductase, and GSH peroxidase activities.
Increased lung weight, G-6-P dehydrogenase; and
GSH reductase activities.
to
H
6
o
o
H
O
O
H
W
O
O
HH
H
W
10 4 days
15 1-7 days
4.0 3h
M/F 21-33yrs Human
5.0 Continuous, M NS
10.0 24 h 7 days
Rats (CD Cobs)
Increase in lung weight, DNA content, G-6-P
dehydrogenase, 6-P-G dehydrogenase, GSH
reductase, disulfide reductase, GSH peroxidase,
disulfide reductase, succinate oxidase, and
cytochrome oxidase activities; no effect on lung
protein
Decreased elastase inhibitory capacity and
increased lipid peroxidation products in B AL of
subjects not administered supplement of vitamin C
and E prior to NO2 exposure.
Changes in the GSH levels in blood and lung
occurred in rats exposed for 24 h, but returned to
normal after 7 days.
Mohsenin(1991)
Paganietal. (1994)
-------
O
to
O
O
oo
to
TABLE AX4.1 (cont'd). EFFECTS OF NITROGEN DIOXIDE ON OXIDANT AND ANTIOXIDANT HOMEOSTASIS
ppm Exposure Gender Age Species (Strain) Effects References
6.0
15
28
9.5
10.0
4 h/day,
30 days
7 days
7 h/day,
5 days/wk,
24 mos
Continuous
3 days,
20 days
NS
Mouse
(NS)
M
NS
18 wks Rat (Fischer 344)
NS Rat (Fischer 344)
14.0
NS
NS
NS
Human
Increase in GSH reductase and G-6-P
dehydrogenase activities.
Increase in GSH levels, G-6-P dehydrogenase, and
GSH peroxidase activities.
Increase in GSH reductase activity in BAL.
Decreased GSH/GSSG ratio in blood and BAL
fluid, but not in lung type II cells. Lipid
peroxidation was decreased in type II cells at
3 days, but was similar to controls at 20 days.
mRNA expression of the enzymes involved in the
biosynthesis (yGCS and GS) was decreased at
both time points. yGT (redox of GSH) mRNA
expression was increased.
Rapid depletion of vitamin C, glutathione and
vitamin E
Csallany (1975)
Mauderlyetal., (1990)
Hochscheid et al. (2005)
Halliwelletal. (1992)
H
6
o
o
H
O
O
H
W
O
O
HH
H
W
M = Male
NPSHs = Nonprotein sulfhydryls
G-6-P dehydrogenase = Glucose-6-phosphate dehydrogenase
G6-P-G dehydrogenase = 6-phosphosgluconate dehydrogenase
SOD = superoxide dismutase
F = Female
NS = Not started
NADP = Nicotinamide-adenine dinucleotide phosphate (reduced form)
TEA = Thiobarbituric acid
BAL = Bronchoalveolar lavage
GR = Glutathione reductase
GS = Glutathione synthetase
GSH = Reduced glutathione
GSSG = Oxidized glutathione
y-GCS - y-Glutamyl-cystein synthetase
y-GT - y-Glutamyltranspeptidase
-------
O
to
O
O
oo
ppm
TABLE AX4.2. EFFECTS OF NITROGEN DIOXIDE ON LUNG AMINO ACIDS, PROTEINS,
LIPIDS, AND ENZYMES
Exposure Gender Age Species (Strain)
Effects
References
to
H
6
o
o
H
O
O
H
W
O
O
HH
H
W
0.4
1.0
3.0
5.0
5.0
0.4
0.4
0.4
1.2
4.0
0.5
1.0
1.0
7.5
15
25
30
72 h M NS Guinea Pig No effect at 0.4 ppm; increase in BAL protein in vitamin
(Hartley) C-depleted, but not normal, animals at 1.0 ppm and
above.
Selgradeetal. (1981)
3h
Continuous,
Iwk
Continuous, M
Iwk
1-14 wks M
NS
Guinea Pig
22-24 wks Rat (Wistar)
6 h/day,
5 days/wk,
4 wks
M
NS
Rat (Fischer 344)
6 h/day,
2 days
Increased BAL protein in vitamin C-depleted guinea
pigs 15 h postexposure.
No effect on BAL protein.
Increased protein content of BAL from
vitamin-C-deficient guinea-pigs.
Complex concentration and duration dependence of
effects. Example: at 0.4 ppm, cytochrome P-450 levels
decreased at 2 wks, returned to control level by 5 wks.
At 1.2 ppm, cytochrome P-450 levels decreased initially,
increased at 5 wks, and decreased at 10 wks. Effects on
succinate-cytochrome c reductase also.
0.5 ppm; increase in urinary hydroxylysine output
starting during wk 1; BAL hydroxylysine level,
angiotensin-converting enzyme level, and BAL protein
content unchanged.
1.0 ppm: gradual increase in urinary hydoxylysine
output, becoming significant the week after exposure
ended; BAL hydroxylysine level lower following
exposure and 4 wks postexposure; andiotensin-
converting enzyme level increased.
Concentration dependent increase in urinary
hydroxylysine output and BAL hydroxyxlysine content,
but only significant at >7.5 ppm and 15 ppm,
respectively; angiotensin-converting enzyme levels and
BAL protein increased in highest-exposed groups.
Sherwin and Carlson
(1973)
Takahashietal. (1986)
Evans etal. (1989)
-------
TABLE AX4.2 (cont'd). EFFECTS OF NITROGEN DIOXIDE ON LUNG AMINO ACIDS, PROTEINS,
LIPIDS, AND ENZYMES
to
O
O
oo
X
•^
Oi
O
i
H
6
O
0
H
O
O
H
W
O
O
HH
H
W
ppm
1.0
5.0
1.2
1.2
4.0
2.0
0.8
5
10
2.0
4.0
10
3.0
3.6
7.2
10.8
14.4
Exposure
7 h/day,
5 days/wk,
up to 15 wks
7 days
1, 2, or 3 wks
1 or 3 days
14 days
10 days
7 days
7 days
24 h
12 h
8h
6h
Gender Age Species (Strain)
M/F 14-16 wks Rat (Fischer 344)
M 10 wks Rat (Wistar)
M NS Guinea pig
M NS Rat
(Sprague-Dawley)
M 12-24 wks Rats (Wistar)
M/F 8 wks Rat
(Sprague-Dawley)
M 10-12 wks Rat
(Sprague-Dawley)
Effects
Change in BAL and tissue levels of enzymes early in
exposure, resolved by 15 wks.
Decrease in levels of cytochrome P-450 at 1.2 ppm.
Increased lactate dehydrogenase (LDH) content of
the lower lobes of the lung.
BAL protein content significantly increased in a
concentration- and exposure duration-dependent
manner, with the change becoming significant at
5 ppm for 3 days and at 10 ppm for > 1 day of
exposure.
Increase activity of lung glycolytic enzymes.
Various changes in lung homogenate protein and
DNA content and enzyme activities, changes more
severe in vitamin E-deficient rats.
Increased BAL protein >7.2 ppm.
References
Gregory etal. (1983)
Mochitate et al. (1984)
Sherwin and Carlson
(1973)
Muller etal. (1994)
Mochitate et al. (1985)
Elsayed and Mustafa
(1982)
Gelzleichter etal. (1992)
-------
TABLE AX4.2 (cont'd). EFFECTS OF NITROGEN DIOXIDE ON LUNG AMINO ACIDS,
PROTEINS, LIPIDS, AND ENZYMES
to
O
O
oo
1
DRAFT-DO
^
0
H
O
O
H
W
O
O
HH
H
W
ppm
4.0
10
4.0
10
25
4.5
4.8
4.8
5.0
5.0
5.0
5.0-25.0
Exposure
10 days
7 days
6 h/day
5 days/wk,
7, 14, and
21 days
16hrs
3h
8 h/day,
7 days
14-72 h
2 wks
6 h/day,
6 days
Continuous,
7 days
Gender Age Species (Strain)
M 2 1-24 wks Rat(Wistar)
M NS Rat (Wistar)
M/F NS Guinea pig
(Hartley)
M
M 8 wks Mouse (Swiss
Webster)
F NS Mouse (NS)
M 5 wks Rat
(Fischer 344)
NS NS Mice
M 10- 11 wks Rat(Sprague-
Dawley)
Effects
Initial decrease in lung protein content followed by an
increase; changes on microsomal enzyme activities.
Increased gamma-glutamyl transferase on days 14 and
21; no consistent effect on alkaline phosphatase, LDH,
or total protein.
Increased lung wet weight, alterations in lung
antioxidant levels in Vitamin C- deficient animals.
Increased lung lavage fluid protein content in vitamin C-
deficient animals.
No significant changes in lung homogenate parameters.
Increase in lung protein (14 to 58 h) by radioactive label
incorporation.
Increased amounts of the tryptophan metabolites and
xanthurenic and kynurenic acids excreted in urine during
wk 2 of exposure, but had returned to normal levels by
wk4.
Modest increase in albumin in BAL; no effect on LDH
or lysosomal enzyme peroxidase.
Concentration-related increase in collagen synthesis rate;
125% increase in rats exposed to 5.0 ppm.
References
Mochitate et al. (1984)
Hooftman etal. (1988)
Hatch etal. (1986)
Mustafa etal. (1984)
Csallany (1975)
Suzuki etal. (1988)
Rose etal. (1989)
Last etal. (1983)
-------
TABLE AX4.2 (cont'd). EFFECTS OF NITROGEN DIOXIDE ON LUNG AMINO ACIDS,
PROTEINS, LIPIDS, AND ENZYMES
to
o
o
oo
X
r
oo
>
H
1
O
o
o
H
O
o
H
W
O
O
HH
H
W
ppm
5.0
20.0
50.0
5.0
8.0
9.5
9.5
10
10
10
20
30
40
Exposure
3h
Continuous,
1, 3, or 7 days
Continuous,
14 days
7 h/day,
5 days/wk,
6 mos
7 h/day, 5
days/wk,
24 mos
24 h or 7 days
Continuous,
14 days
4h
Gender Age Species (Strain)
NS NS Rabbit
(New Zealand)
M NS Rat
(Sprague-
Dawley)
F NS Mouse (NS)
M In utero Rat
and (Fischer 344)
6 mos
M 18 wks Rat
(Fischer 344)
M NS Rat (CD cobs)
M 8 wks Rat (Wistar)
M NS Rat
(Long Evans)
Effects
Benzo [a] pyrene hydroxylase activity of tracheal
mucosa not affected.
Increased BAL protein at 3 days (day 7 not measured);
increased (120% collagen synthesis at 7 days (not
measured other days).
Increase in lung protein.
Increase in BAL alkaline phosphatase, acid
phosphatase, and LDH in older rats only.
Increase in BAL levels of LDH and alkaline
phosphatase activities and in collagenous peptides.
Protein content of BALF increased significantly in rats
after only 24 h. BALF elastase activity was not
affected. Concentration-dependent increase in a-
1 proteinase inhibitor content after 24 h of exposure,
but not with longer exposures.
Changes in several enzymes in whole lung
homogenates.
Increased activities of various enzymes, sialic acid, and
BAL protein; attenuation by high dietary levels of
vitamin E.
References
Palmer etal. (1972)
Last & Warren (1987)
Csallany (1975)
Mauderly etal. (1987)
Mauderly etal. (1990)
Pagani etal. (1994)
Sagai etal. (1982)
Guth and Mavis
(1985, 1986)
-------
O
to
O
O
oo
TABLE AX4.2 (cont'd). EFFECTS OF NITROGEN DIOXIDE ON LUNG AMINO ACIDS,
PROTEINS, LIPIDS, AND ENZYMES
ppm
Exposure
Duration
Gender Age Species (Strain)
Effects
Reference
10
24 h/day,
0 (control),
3 days or
20 days
NS
NS
Rat (Fischer 344)
LPO decreased in type II pneumocytes after 3 days compared
to controls but remained comparable to controls after 20 days.
Authors stated that LPO is a very early reaction during
oxidative stress and that the decrease after 3-day exposure
suggests an adaptation mechanism.
Exposure duration-dependent, statistically significant increase
in GPX and GR enzyme activities after 3 and 20 days over
control values. Changes in mRNA expression of GSH
synthesizing enzymes in type II pneumocytes were also
observed.
Hochscheid et al.
(2005)
to
VO
H
6
o
o
H
O
O
H
W
O
O
HH
H
W
10 24 h/day, for 0,
3,20, or 28 days
M NS Rat (Sprague- Uptake of surfactant-like liposomes by type II pneumocytes in
Dawley) the presence or absence of SP-A was faster and significantly
higher in cells from all NO2 exposed groups than in control
cells. No difference in the uptake kinetics between cells from
exposed groups of different duration. Increase in liposome
uptake suggests NO2 exposure likely disrupted cell membranes
to allow liposomes to enter the cells easily. Lipid uptake
associated with duration-dependent increase in internalization
of label found in PC fraction. Suggests increased demand of
PC in lung injury.
Muller et al. (2003)
-------
O
to
O
O
oo
TABLE AX4.2 (cont'd). EFFECTS OF NITROGEN DIOXIDE ON LUNG AMINO ACIDS,
PROTEINS, LIPIDS, AND ENZYMES
ppm
Exposure
Duration
Gender Age
Species
(Strain)
Effects
Reference
0.8 Presumably M NS Rat Phospholipid component in BAL increased in a concentration- and
5.0 continuous, (Sprague- exposure duration-dependent manner, with significance only at
10 1 day or 3 days Dawley) 10 ppm, but not at 5 ppm or below, for > 1 day of exposure. PC
content in BAL of exposed animal did not change significantly
compared to controls. At 10 ppm, but not at 5 ppm or below,
percentage of saturated PC decreased and that of unsaturated PC
increased statistically significant, while significant decreases in
palmitic acid (16:0) and increases in arachidonic acid (20:4)
contained in PC were observed. Sphingomyelin - as a marker for an
influx of serum into the alveolar airspace - was also not changed by
NO2 exposure.
Mulleretal. (1994)
Functional studies on surfactant phospholipid extracts indicated
increased values for the surface tension at equilibrium, and for the
maximal and minimal surface tension of animals exposed to > 5 ppm,
but not to 0.8 ppm. Suggests NO2 directly impaired surface tension
at > 5 ppm. Structure of the SP-A not altered by NO2 exposure.
Authors suggested exposure to NO2 impaired surfactant components
may be used as markers of altered surfactant metabolism.
H
6
o
o
H
O
O
H
W
O
O
HH
H
W
5 24 h or 24 h/day M
10 for 7 days
NS Rat (CD Concentration-dependent increase in a-1 PI content. Exposure to 5
Cobs) and 10 ppm NO2 significantly increased
a-1 PI content only after 24 h and returned to control values after 7
days at each exposure concentration. Since blood GSH content was
also increased together with
a-1 PI content at 24 h, authors suggested the increase in these
parameters can be considered a prompt protective response resulting
in no further increase of a-1 PI.
Pagani et al. (1994)
LPO = Lipid peroxidation
PC = Phosphatidylcholine
SP-A = Surfactant protein-AGPx = Glutathione peroxidase
GPX = Glutathione peroxidase
GR = Glutathione reductase
GSH = Glutathione
a-1 PI = a-1 proteinase inhibitor
-------
O
to
O
O
oo
TABLE AX4.3. EFFECTS OF NITROGEN DIOXIDE ON ALVEOLAR MACROPHAGES AND
LUNG HOST DEFENSE
ppm
Exposure
Gender
Age
Species (Strain)
Effects
Reference
0.05 base +
2.0 peaks
0.6
3 h base + three
15-minpeaks
NS
NS
Human
No effects at 0.05 ppm NO2 with peaks; trend
(p < 0.07) towards AMs losing ability to
inactivate influenza virus at 0.6 ppm.
Framptonetal. (1989)
0.1
1.0
5.0
20
Ih
NS
NS Rat
(Sprague-Dawley)
(in vitro)
At 5.0 ppm: increase in LTB4; concentration-
related decrease in SOD production in AMs at
> 1.0 ppm; increase in LDH in AMS at 5.0 and
20 ppm
Robinson etal. (1990)
>
0.2
0.5
2.0
Gestation Rat (Brown- Reactive oxygen species generation from
12 wks Norway) alveolar macrophages was significantly
suppressed in NO2 exposed weanling animals;
no changes in reactive oxygen generating
capability in the embryonic exposed animals.
Kumae and Arakawa
(2006)
0.5
Continuous,
24wks
NS
NS
Mouse
No effects on AM morphology at 0.5 ppm
continuous or 0.1 ppm base + peak.
Aranyietal. (1976)
H
6
O
O
H
O
O
H
W
O
O
HH
H
W
0.1 base +
1.0 peak
2.0
0.5 base +
2.0 peak
Continuous
base + 3-h peak,
5 days/wk,
24wks
Continuous,
33wks
Continuous
base + 1-hpeak,
5 days/wk,
33wks
After 21 wks of exposure to 2.0 ppm
continuous or 0.5 ppm base + peak,
morphological changes were identified, such a
loss of surface processes, appearance of
fenestrae, bleb formation, and denuded surface
areas.
-------
TABLE AX4.3 (cont'd). EFFECTS OF NITROGEN DIOXIDE ON ALVEOLAR MACROPHAGES AND
LUNG HOST DEFENSE
to
o
o
oo
ppm
0.3
1.0
Exposure
2h/day
2, 6, 13 days
Gender Age Species (Strain)
M NS Rabbit
(New Zealand)
Effects
Decreased phagocytic ability of AMs at
0.3 ppm after 2 days of exposure;
increased at 1.0 ppm after 2 days of
exposure; no effect on cell number or
viability; random mobility reduced at
0.3 ppm only; no effects after 6 days of
exposure.
Reference
Schlesinger (1987)
0.3
1.0
2 h/day up to 14 days M
NS
Rabbit
(New Zealand)
Increase in alveolar clearance.
Schlesinger and
Gearhart (1987)
to
0.3
1.0
3.0
10
1.0
10
2h
M
NS
2 h/day, 14 days
Rabbit Concentration-related acceleration in
(New Zealand) clearance of particles from lung with the
greatest increase at two lowest
concentrations, effects from repeated
exposures similar to those seen after acute
exposures to same concentrations.
Vollmuthetal. (1986)
0.5 0.5, 1, 5 and 10 days
exposure
NS
NS
Rat (NS) Superoxide production in alveolar
macrophages from B ALF, stimulated by
phorbol myrisate acetate (PMA), was
Robinson etal. (1993)
H
6
o
0
— ]
o
o
H
W
O
o
HH
H
W
0.5 base +
1.5 peak
2.0 base
+6.0 peak
Base 22 h/day, M
7 days/wk + two 1-h
peaks, 5 days/wk,
6 wks
continued to be depressed after 1,5, and
10 days.
1 day and Rat (Fischer Trend towards increase in number of Crapo et al. (1984)
6 wks 344) AMs and cell volume in younger animals; Chang etal. (1986)
increase in number of AMs and cell
volume in older rats.
-------
TABLE AX4.3 (cont'd). EFFECTS OF NITROGEN DIOXIDE ON ALVEOLAR MACROPHAGES AND
LUNG HOST DEFENSE
to
O
O
oo
i
ppm
0.5
1.3
2.7
1.0
2.0
4.0
1.0
5.0
15
1.0
2.0
4.0
1.0 +
0.9 ppm No
Exposure Gender Age
Continuous, 28 days M 6 wks
24 h/day, 12 wks
6 h/day, 2 days NS 4-6 wks
24 h/day, 12 wks
7 h/day, 5 days/wks NS NS
for 1 1 or 22 exposures
Species (Strain)
Rat (Wistar)
Guinea pig (NS)
Rat (NS)
Mouse (CD1)
Guinea pig (NS)
Rat (NS)
Rat (Long Evans)
Effects
Increase in AMs in highest exposed group; no
effects noted in 2 lowest exposure groups.
IgE-mediated histamine release from lung
mast cells was enhanced in guinea pigs, but
not rats exposed to 4.0 ppm. No effect
observed at lower concentrations.
Exposure-related decrease in AM
phagocytosis from 1.0-5.0 ppm, decrease was
not further affected by 15 ppm.
IgE-mediated histamine release from lung
mast cells was enhanced in guinea pigs, but
not rats exposed to 4.0 ppm. No effect
observed at lower concentrations.
Stimulated clearance of particles from lung at
lowest concentration, but decreased clearance
Reference
Rombout etal. (1986)
Fujimaki and Nohara
(1994)
Rose etal. (1989)
Fujimaki and Nohara,
(1994)
Ferin and Leach
(1977)
15
24
rate at two highest concentrations.
H
6
o
o
^^
H
O
O
H
W
o
O
HH
H
W
1.0
5.0
base +
5.0 peaks
1.3-17
7 h/day, 5 days/wks M/F
Base 7 h/day,
5 days/wks;
two 1.5-h peaks/day;
15 wks
NS ("acute") F
14-16 wks Rat (Fischer 344) Accumulation of AMs. Superimposed peak
exposures produced changes that may persist
with continued exposures.
NS Rat (Sprague-Dawley) Decreased production of superoxide anion
radical.
Gregory etal. (1983)
Amorusoetal. (1981)
-------
TABLE AX4.3 (cont'd). EFFECTS OF NITROGEN DIOXIDE ON ALVEOLAR MACROPHAGES AND
LUNG HOST DEFENSE
to
O
O
oo
X
£
M
%•/
>
H
1
O
O
n
\~s
H
O
O
H
W
O
O
HH
H
W
ppm
2.0
10
2.0
2.0
2.7
3-6
3.6
12.1
4
10
25
4.0
Exposure Gender Age Species (Strain)
3 days M/F 5,10,21, Guinea pig (Dunkin
45, 55, 60, Hartley)
and Rat (Wistar)
>60 days
8 h/day, 5 days/wk, M/F 3-4 yrs Baboon
6 mo
4h NS NS Human
24 h M 6wks Rat (Wistar)
3 h NS NS Dog (Beagle)
Ih F NS Rat (Sprague-Dawley)
2h (in vitro)
6 h/day, 7, 14, or M NS Rat (Wistar)
21 days
10 days 19-23 wks
Effects
Newborns were less affected than adults
when AMs were tested for SOD levels.
Impaired AM responsiveness to migration
inhibitory factor.
Decreased phagocytosis and superoxide
anion release.
Increase in number of AMs.
Enhanced swelling of AMs.
Enhanced macrophage agglutination with
concanavalin A at both concentrations
tested.
Changes in morphology at all
concentrations; increase in number of AMs
at > 10 ppm; phagocytic capacity reduced
after 14 and 21 days of exposure to 25 ppm.
Increase in number of AMs; no increase in
PMNs; increased metabolic activity, protein,
and DNA synthesis; all responses peaked on
day 4 and returned to normal on day 10.
Reference
Azoulay-Dupuis et al.
(1983)
Green and Schneider
(1978)
Devlin etal. (1992)
Romboutetal. (1986)
Dowell etal. (1971)
Goldstein etal. (1977)
Hooftman etal. (1988)
Mochitate et al. (1986)
-------
O
to
O
O
oo
ppm
TABLE AX4.3 (cont'd). EFFECTS OF NITROGEN DIOXIDE ON ALVEOLAR MACROPHAGES AND
LUNG HOST DEFENSE
Exposure
Gender
Age Species (Strain)
Effects
Reference
4.0
8.0
Up to 10 days
NS
NS
Rat (Fischer 344)
Increase in number of AMs at both
concentrations, reaching a peak on day 3 and
5; no increase in number of PMNs; decrease
in AM viability throughout exposure period.
Suppression of phagocytic activity after 7
days of exposure to 4 ppm and after 5 days
of exposure to 8 ppm; returned to normal
value at 10 days. Decrease in superoxide
radical production, but at 4 ppm, the effect
became significant on days 3,5, and 10; at 8
ppm, the effect was significant at all time
periods tested.
Suzuki etal. (1986)
5.0
7 days
No effect on phagocytic activity.
Lefkowitzetal. (1986)
H
6
o
o
H
O
O
H
W
O
O
HH
H
W
5
15
5
10
15
5-60
7.0
9.5
3 h after infection
with parainfluenza
3 virus
3h
3h
24 h
7 h/day; 5 days/wk;
18-22 mo
NS
M
Fb
NS
NS
M
NS
NS
NS
NS
18wks
Rabbit (New
Zealand)
Humans (in vitro
exposure)
Rabbit (New
Zealand)
Rabbit
Rat (Fischer 344)
AMs lost resistance to challenge with rabbit
pox virus after exposure to 15 ppm.
No change in cell viability, release of
neutrophil chemotactic factor, or
interleukin-1.
Inhibition of phagocytic activity.
Increased rosette formation in AMs treated
with lipase.
No effect on long-term clearance of
radiolabeled tracer particles.
Acton and Myrvik (1972)
Pinkston et al. (1988)
Gardner etal. (1969)
Acton and Myrvik (1972)
Hadley etal. (1977)
Mauderly etal. (1990)
-------
TABLE AX4.3 (cont'd). EFFECTS OF NITROGEN DIOXIDE ON ALVEOLAR MACROPHAGES AND
LUNG HOST DEFENSE
to
o
o
oo
X
£
ON
O
§
H
6
o
0
H
O
O
H
W
O
O
HH
H
W
ppm Exposure Gender Age Species (Strain) Effects
10 Continuous 7 days NS NS Rat(NS) High influx of PMNsinthe lung(BALF)
after 24 h of exposure, reversed for
macrophages; no change in the lymphocyte
population.
10 35 days NS NS Guinea pig 63% increase in epithelial cells positive for
macrophage congregation.
10 4 h F NS Mouse (Swiss) Increase in total pulmonary cells in animals
infected with some species of bacteria.
10 24 h M 12-13 Rat(Sprague- Decreased phagocytosis at 25 ppm only.
25 wks Dawley)
ppm Exposure Gender Age Species (Strain) Effects
Clearance
3 6 h/day, 6 days/wk, F NS Guinea Pig Significant, dose -dependent decrease in
9 for 2 wks ciliary activity, significant at 3 ppm (12%)
and 9 ppm (30%), and increase in
eosinophil accumulation on epithelium and
submucosal layer.
10 14 h/day, for 15 M NS Mouse (C57BL/5) 20 ppm NO2 induced an increased mucus
20 days, 20 or 25 days production due to goblet cell hyperplasia in
the central airways.
Reference
Paganietal. (1994)
Sherwinetal. (1968)
Jakab (1988)
Katz and Laskin( 1976)
Reference
Ohashietal. (1994)
Wegman and Herz (2002)
-------
O
to
O
O
oo
TABLE AX4.3 (cont'd). EFFECTS OF NITROGEN DIOXIDE ON ALVEOLAR MACROPHAGES AND
LUNG HOST DEFENSE
ppm Exposure
Gender
Age
Species
(Strain)
Effects
Reference
Alveolar Macrophage Endpoints
H
6
o
o
H
O
O
H
W
O
O
HH
H
W
0.5
8 h/day,
5 days/wk,
for 0.5, 1,5,
or 10 days
M
NS
0.2
0.5
2.0
Continuous,
presumably
7 days/wk, up
to 12 wks
Neonates or
5 wks old
Rat (Sprague- Acute depression of pulmonary arachidonate metabolism observed. Robinson et al.
Dawley) Unstimulated AM synthesis of LTB4 depressed within 1 day and (1993)
also on day 5 of exposure. Acute depression of AM synthesis of
TxB2, LBT4, and 5-HETE on stimulation by the calcium
ionophore, A23187, within 1 day of exposure but not with longer
exposure, while 5-HETE increased significantly at 10 days of
exposure only. Suggests rapid depression of cyclooxygenase and
5-lipoxygenase activities. ZAS-stimulated LTB4 production
delayed until 5 days and remained lower at 10 days. AM
superoxide production stimulated by PMA was rapidly and
continuously depressed throughout the study. BAL fluid levels of
LTB4 and TxB2 paralleled ex vivo depression of AM production.
Rat (Brown- Animals were exposed during embryonic or weanling (5-wks old) Kumae and
Norway) period. ROS generation was significantly suppressed at 0.5 and Arakawa
2.0 ppm NO2 in animals exposed during the weanling period. (2006)
Cytokine level measurement in AM culture mediums indicated that
inflammatory reactions (significant increases in TFNa and IFNy)
were initiated at 8 wks and terminated at 12 wks in animals
exposed during the embryonic period while inflammatory reactions
(significantly increased TFNa level) were not initiated at 8 wks but
take place at 12 wks in animals exposed as weanlings. Results
suggest that NO2 exposure from the weanling period has stronger
effects on AM activity.
-------
O
to
O
O
oo
TABLE AX4.4. EFFECTS OF NITROGEN DIOXIDE ON LUNG PERMEABILITY AND INFLAMMATION
ppm
0.8
5
10
Exposure
Presumably
continuous,
1 day or 3 days
Gender
M
Age
NS
Species
(Strain)
Rat
(Sprague-
Dawley)
Effects
BAL protein content significantly increased in a concentration- and
exposure duration-dependent manner, with the change becoming significant
at 5 ppm for 3 days and 10 ppm for > 1 day of exposure.
Reference
Muller et al.
(1994)
oo
H
6
o
o
H
O
O
H
W
O
O
HH
H
W
5 24hor M NS Rat (CD Exposure induced inflammatory response in the lungs. At 10 ppm, influx Paganietal.
10 24 h/day for Cobs) of PMN, maximal at 24 h, but no influx observed after 7 days of exposure, (1994)
7 days a trend that was also observed with protein content in BAL fluid. In
contrast, no influx of macrophages was observed, but the influx was
maximal after 7 days of exposure. No significant changes in lymphocyte
counts at any exposure concentration and protein content in BAL fluid not
significantly affected at 5 ppm.
Protein content of BAL fluid increased significantly only after 24 h of
exposure to 10 ppm NO2.
5 6 h/day for 1, 3, NS NS Mouse Exposure to 5 ppm NO2 did not cause any lung inflammation or injury. Poynteretal.
25 or 5 days (C57BL/6) Exposure to 25 ppm NO2 induced acute lung injury (characterized by (2006)
increases in protein, LDH, macrophages, and neutrophils recovered by
BAL) that peaked after 3 days, lesions within terminal bronchioles, and
AHR. OVA-sensitized animals exposed to 25 ppm, but not 5 ppm, NO2
showed augmentation of eosinophilic inflammation and terminal
bronchiolar lesions, which extended significantly into the alveoli. No
increased expression of mucus cell-associated gene products in non-
sensitized or OVA-sensitized animals exposed at any concentration.
5 3 h M NS Mouse Exposure of OVA-challenged animals to 20 ppm produced BHR and Proust et al.
20 (BALB/c) caused significant increase in neutrophils and fibronectin concentration, (2002)
significant reduction in eosinophil count, and exudation and release of IL-5
in BAL fluid 24 h and/or 72 h after exposure. Exposure to 5 ppm did not
modify BHR, but significantly reduced pulmonary eosiniphilic
inflammation (reduced esinophilic count and eosinophil peroxidase
activity) and the production of IL-5 in the BAL fluid. Exposure to NO2 did
not cause any significant changes in IgE anti-OVA antibody in exposed
animals, while IgGl liters were significantly increased in animals exposed
only lo 5 ppm NO2, compared lo controls. There was no development of
mucosal melaplasia in any NO2 exposed group compared lo controls.
-------
O
to
O
O
oo
TABLE AX4.4 (cont'd). EFFECTS OF NITROGEN DIOXIDE ON LUNG PERMEABILITY AND INFLAMMATION
Species
ppm Exposure Gender Age (Strain)
Effects
Reference
5 24 h/day, for M NS Rat (Sprague- Exposure to NO2 exhibited concentration- and exposure duration-
10 3 or 25 days Dawley) dependent, and tissue localization-specific differences in Clara cell
20 proliferation. Increased proliferation (measured as BrdU-LI) in both
bronchial and bronchiolar epithelium of all exposed groups, with
significance starting at exposure to 5 ppm for 3 or 25 days. Exposure to 5
and 10 ppm NO2 for 3 days showed significantly higher proliferative
activity in bronchiolar epithelium than in the bronchial epithelium for
corresponding exposure groups, while the difference in proliferation
between proximal and distal airways diminished after 25 days of exposure.
Clara cell proliferation was not accompanied by a change in Clara cell
numbers in any of exposure groups, a finding the authors explained by the
assumption that the process of proliferation takes place at the same rate as
Clara cell differentiates into other cells, a transition that leads to the loss of
phenotypic Clara properties in differentiating cells.
Earth and
Muller (1999)
VO
H
6
o
o
H
O
O
H
W
O
O
HH
H
W
1.2
10
20
3.6
7.2
10.8
14.4
24 h/day, for M
3 days
14 h/day, for M
15 days, 20 or
25 days
24 h, 12h, 8h, M
and 6 h,
respectively,
for 3 days,
giving a C x T
of 86.4ppm-h
NS Rat (Sprague-
Dawley)
NS Mouse
(C57BL/5)
NS Rat (Sprague-
Dawley)
No significant differences in cell viability and percentages of pulmonary
AMs or PMNs between animals exposed to 1.2 ppm NO2 and controls.
Exposure to 10 ppm for 15 days - the only exposure duration used for this
concentration - did not significantly affect influx of inflammatory cells
(leukocyte subpopulations and macrophages). 20 ppm NO2 induced
airway and parenchymal inflammation - dominated by a significant influx
of macrophages and neutrophils -, peaking at 15 days of exposure and
declining at day 25.
Significant cell proliferation (increased labeling index) in peripheral
airways compared to controls, regardless of concentration or exposure
duration. Suggests Haber's law (c x t = k) was not followed over the
concentration ranges studied. No proliferative response was observed in
alveolar epithelium.
Bermudez
(2001)
Wegman and
Herz (2002)
Rajini et al.
(1993)
-------
TABLE AX4.4 (cont'd). EFFECTS OF NITROGEN DIOXIDE ON LUNG PERMEABILITY AND INFLAMMATION
Species
ppm Exposure Gender Age (Strain)
O
to
O
O
oo
Effects
Reference
0.5 8 h/day, M NS Rat(Sprague- No effect on weight gain. No effects on neutrophil, lymphocyte
5 days/wk, for Dawley) macrophage/monocyte levels or cell population percentages in BAL.
0.5, 1, 5, or Suggests no significant influx of inflammatory cells into lung airways and
10 days alveolar spaces.
Robinson et al.
(1993)
H
6
o
o
H
O
O
H
W
O
O
HH
H
W
5 24 h/day, for M NS Rat (Sprague- Compared to controls, proliferative activity significantly increased, but
10 3 or 25 days Dawley) with no concentration-dependence in respiratory bronchiolar epithelium at
20 5 ppm and above after 3-day exposure, but increase was concentration-
dependent, with significance at > 10 ppm following 25-day exposure.
Proliferative activity increased in a concentration-dependent manner in
bronchial epithelium, with significance only at 20 ppm after 3 days of
exposure and at > 10 ppm following 25 days of exposure.
Concentration-dependent thickening of alveolar septa and bronchiolar
walls, significant only at > 10 ppm after 3- and 25-day exposure, but 5 ppm
caused significant thickening only in bronchiolar walls after 25 days, but
not after 3 days, of exposure compared to controls.
Statistically significant reduction in alveoli surface density at > 10 ppm
after 25 days of exposure and at 20 ppm only after 3 days of exposure,
while alveolar circumference increased statistically significantly at
> 10 ppm after 25 days of exposure and at 20 ppm following 3 days of
exposure. 5 ppm did not significantly affect these endpoints after 3 or 25
days of exposure compared to controls. Data suggest concentration- and
exposure duration-dependent development of emphysema, significant only
at > 10 ppm, but not at 5 ppm, after 3 or 25 days of exposure. Alveolar
duct length increased significantly at > 10 ppm after 25 days and at 20 ppm
after 3 days of exposure; no effect at 5 ppm.
Radial alveolar count values decreased significantly at 20 ppm after 3 and
25 days and at 10 ppm after 25 days exposure; no effect at 5 ppm.
Increase in avg medial thickness at 10 ppm with significance only after 25
days of exposure and at 20 ppm with significance after 3 and 25 days of
exposure. In contrast, 5 ppm caused significant increase at both 3 and 25
days of exposure. Authors suggested effect at 5 ppm was due a reduction
of pulmonary arterial mass and not as a result of vasodilation induced by
nitrogen oxide.
Earth et al.
(1994b)
-------
O
to
O
O
oo
TABLE AX4.4 (cont'd). EFFECTS OF NITROGEN DIOXIDE ON LUNG PERMEABILITY AND INFLAMMATION
ppm
Exposure Gender Age
Species
(Strain)
Effects
Reference
>
-k
H
6
o
o
H
O
O
H
W
O
O
HH
H
W
0, 1.0,
2.0, 4.0
24 h/day,
for 12 wks
M
10 wks Rat (Wistar)
No change in body weight or absolute lung weight in any exposed
group compared to controls, but relative lung weight was significantly
increased in animals exposed to 4.0 ppm NO2. Number of lung cells
from animals exposed to >2.0 ppm significantly reduced (but extent of
reduction at 4.0 ppm < 2.0 ppm), whereas number of mast cells not
significantly different in exposed animals compared to controls. IgE-
mediated histamine release from lung mast cells significantly reduced
at 2.0 ppm NO2 but not at 4.0 ppm or 1.0 ppm, but no difference
observed in A23187-stimulated histamine release in lung mast cells in
any exposed group compared to controls. Results from histamine
release suggest that NO2 exposure to rat lung mast cells did not induce
histamine-releasing activity.
Guinea Pig NO change in body weight or absolute or relative lung weight in any
(Hartley) exposed group compared to controls. Number of lung cells or mast
cells was not significantly different in exposed animals compared to
controls. Increasing trend observed in IgE-mediated histamine release
from lung mast cells, the change becoming significant only at 4.0
ppm. No significant change in ionophore A2318-stimulated histamine
release in exposed groups compared to controls. Data suggest NO2
exposure in guinea pigs enhanced histamine-releasing activity in lung
mast cells.
Fujimaki and
Nohara (1994)
-------
TABLE AX4.4 (cont'd). EFFECTS OF NITROGEN DIOXIDE ON LUNG PERMEABILITY AND INFLAMMATION
Species
ppm Exposure Gender Age (Strain)
O
to
O
O
oo
Effects
Reference
to
0.2
0.5
2
Continuous,
presumably
7 days/wk, up
to 12 wks
Neonates Rat (Brown- Rats were exposed from embryonic or weanlings (5-wks old)
or 5 wks Norway) period up to 12 wks of age. Significantly decreased levels of AM +
old Mo in weanling animals exposed to 0.5 and 2.0 ppm, and
significantly increased levels of neutrophils in animals exposed to
2.0 ppm at 12 wks. Levels of AM + Mo significantly increased in
animals exposed to 0.5 ppm during embryonic period while levels
of neutrophil population significantly decreased, compared to
controls, at 12 wks. [Changes in cell populations in BAL fluid not
investigated in animals exposed to 2.0 ppm during the embryonic
period due to loss of sample.] Mean level of lymphocytes
significantly increased in the embryonic group exposed to 0.2 ppm
and in the weanling group exposed to 0.5 and 2.0 ppm, but increase
not concentration-dependent in the weanling group. Except for
exposure to 0.2 ppm, NO2 exposure appeared to improve allergic
conditions in the BAL fluid of the embryonic group, but caused
inflammatory changes in the BAL fluid of the weanling group.
Data suggest NO2 exposure from the weanling period has stronger
effects on AM activity.
Kumae and
Arakawa (2006)
H
6
o
o
H
O
O
H
W
O
O
HH
H
W
AHR = Airway hyperresponsiveness
AM = Alveolar macrophage
BAL = Bronchoalveolar lavage
BHR = Bronchopulmonary hyperreactivity
BrdU-LI = Bromodeoxyuridine-laebling index
IgE = Immunoglobulin E
IgG = Immunoglobulin G
IL-5 = Interleukin-5
LDH = Lactate dehydrogenase
Mo = Monocytes
OVA = Ovalbumin
PMN = Polynorphonuclear neutrophil
-------
O
to
O
O
oo
ppm
0.5
0.1 base +
0.25,0.5,
or 1.0
peak
0.25
TABLE
Exposure
Continuous
Continuous base +
3 h/day, 5 days/wk
peak for 1, 3, 6, 9,
12 mos
7 h/day,
5 days/wk, 7wks
AX4.5.
Gender
NS
F
EFFECTS OF NITROGEN DIOXIDE ON IMMUNE RESPONSES
Age
NS
6 wks
Species
(Strain)
Mouse
Mouse
(AKR/cum)
Effects
Suppression of splenic T and B cell responsiveness to
mitogens variable and not related to concentration or
duration, except for the 940 ug/m3 continuous group, which
had a linear decrease in PHA-induced mitogenesis with
NO2 duration.
Reduced percentage of total T-cell population and trend
towards reduced percentage of certain T-cell
References
Maigetter et al. (1978)
Richters and Damji
(1988)
subpopulations; no reduction of mature T cells or natural
killer cells.
H
6
o
o
H
O
O
H
W
O
O
HH
H
W
0.25 7 h/day, F 5 wks
5 days/ wk, 36 wks
0.35 7 h/day, M 6 wks
5 days/ wk, 12 wks
0.4 24 h/day M 7 wks
1.6 4 wks
0.5 base + 22 h/day, 7 days/wk M
1.5 peak base+ 6 h/day,
5 days/wk peak for
1,3, 13, 52, 78 wks
0.5 base + 24 h/day, 5 days/wk M 6 wks
2.0 peak base + 1 h/day,
5 days/wk peak for
3 mos
Mouse Reduced percentage of total T-cell population and
(AKR/cum) percentages of T helper/inducer cells on days 3 7 and 181.
Mouse
(C57BL/6J)
Mouse
(BALB/c)
10 wks Rat
(Fischer
344)
Mouse
(CD-I)
Richters and Damji
(1990)
Richters and Damji
(1988)
Trend towards suppression in total percentage of T-cells.
No effects on percentages of other T-cell subpopulations.
Decrease in primary PFC response at >752 ug/m3. Increase Fujimaki et al. (1982)
in secondary PFC response at 3010 ug/m3.
No effect on splenic or circulatory B or T cell response to
mitogens. After 3 weeks of exposure only, decrease in
splenic natural killer cell activity. No histological changes
in lymphoid tissues.
Vaccination with influenza A2/Taiwan virus after exposure.
Decrease in serum neutralizing antibody; hemagglutination
inhibition antibody liters unchanged. Before virus
challenge, NO2 exposure decreased serum IgA and
increased IgGl, IgM, and IgG2; after virus, serum IgA
unchanged and IgM increased.
Selgrade et al. (1991)
Ehrlichetal. (1975)
-------
TABLE AX4.5 (cont'd). EFFECTS OF NITROGEN DIOXIDE ON IMMUNE RESPONSES
O
to
O
O
oo
ppm
Exposure
Gender
Age
Species
(Strain)
Effects
Reference
H
6
o
o
H
O
O
H
W
O
O
HH
H
W
0.5 8 h/day, 5 days/wk,
for 0.5, 1,5, or
10 days
5 3h
M NS Rat(Sprague- Levels of TxB2, LTB4, and PGE2 in BAL fluid were depressed
Dawley) within 4 h of exposure. Suggests acute depression of pulmonary
arachidonate metabolism observed in BAL fluid. Unstimulated AM
synthesis of LTB4 was depressed within 1 day and also on day 5 of
exposure. The AM synthesis of TxB2, LTB4, and 5-HETE on
stimulation by the calcium ionophore, A23187, was acutely depressed
within 1 day of exposure but not with longer exposure, while 5-HETE
was significantly increased at 10 days. Suggests rapid depression of
cyclooxygenase and 5-lipoxygenase activities. Z AS-stimulated LTB4
production was delayed until 5 days and remained lower at 10 days.
BAL fluid levels of LTB4 and TxB2 paralleled ex vivo depression of
AM production. AM superoxide production stimulated by PMA was
rapidly and continuously depressed throughout the study.
F 7 wks Rat (Brown- Rats were immunized intraperitoneally and challenged intratracheally
Norway) with house mite dust. Animals were exposed to NO2 after
sensitization or challenge or after sensitization and challenge (double
exposure). A single exposure after sensitization or challenge caused a
significant decrease in antigen-specific IgG in BAL fluid, single
exposure after sensitization caused significant increase in serum IgE,
while exposure only after challenge caused significant decrease in
antigen-specific IgA levels in BAL fluid. Double NO2 exposure
caused significantly higher levels of antigen-specific serum IgE and
local IgA, IgG, and IgE antibody, and significant increase in
lymphocyte responsiveness to antigen in the spleen and mediastinal
lymph nodes. Double exposure also significantly increased the ratio
of inflammatory cells to alveolar macrophages without affecting the
total number of lavageable cells.
Data suggest a 3-h exposure to 5 ppm NO2 after intraperitoneal
sensitization and pulmonary challenge with house dust mite allergen
was necessary to enhance specific immune responses to the allergen
and increased the number of inflammatory cells in the lungs.
Additionally, data suggest an upregulation of specific immune
responses and subsequent immune-mediated pulmonary inflammation.
Robinson et al.
(1993)
Gilmour et al.
(1996)
-------
TABLE AX4.5 (cont'd). EFFECTS OF NITROGEN DIOXIDE ON IMMUNE RESPONSES
O
to
O
O
oo
ppm Exposure Gender
Age
Species
(Strain)
Effects
Reference
H
6
o
o
H
O
O
H
W
O
O
HH
H
W
5 3 h M NS Mouse Animals were sensitized and challenged with the antigen OVA
20 (BALB/c) to generate airway inflammation before being exposed to NO2.
Exposure of OVA-challenged animals to 20 ppm produced
BHR and caused significant increase in neutrophils and
fibronectin concentration, significant reduction in eosinophil
count, and exudation and release of IL-5 in broncholaveolar
fluid 24 h and/or 72 h after exposure. Exposure to 5 ppm did
not modify BHR, but significantly reduced pulmonary
eosiniphilic inflammation (reduced esinophilic count and
eosinophil peroxidase activity) and the production of IL-5 in the
B AL fluid. Authors suggested that potentiation of BHR by 20
ppm NO2 in allergic mice may be accounted for by an increased
vascular/epithelial permeability, facilitating the allergen
availability and accelerating the inflammatory process.
Exposure to NO2 did not cause any significant changes in IgE
anti-OVA antibody in exposed animals, while IgGl liters were
significantly increased in animals exposed only lo 5 ppm NO2,
compared lo controls. There was no developmenl of mucosal
melaplasia in any NO2 exposed group compared lo controls.
2 h/day M and F From birth Rabbil (NZW) No effecl on mortality, heallh, behavior, body weighl, or basal
until 3 mos pulmonary function (lung resistance, dynamic compliance,
of age respiration rales, tidal volume, and minule volumes) in animals
immunized againsl house mile dusl compared lo lillermales
exposed lo air. Immune parameters were nol evaluated.
Proust et al. (2002)
Douglas et al.
(1994)
-------
O
to
O
O
oo
TABLE AX4.5 (cont'd). EFFECTS OF NITROGEN DIOXIDE ON IMMUNE RESPONSES
ppm
Exposure
Gender
Age
Species
(Strain)
Effects
Reference
4.76 4 h/day, 5 days/wk,
for 6 wks
(30 exposures, total)
M
NS Guinea Pig Animals were intraperitoneally sensitized twice and then
(Hartley) challenged pulmonarily with C. albicans. Animals were
exposed, from the first day of sensitization, and
throughout the study period. Exposure to NO2 resulted in
significantly increased respiratory rate (tachypnea)
(2.3-2.7 times/s) 15 h after antigen challenge compared to
controls, but did not significantly affect the
expiration/inspiration ratio. Authors indicated that
delayed-type dyspneic symptoms in this study were
increased by exposure to NO2.
Kitabatakeetal. (1995)
H
6
o
o
H
O
O
H
W
O
O
HH
H
W
0.06
0.5
1
2
4
24 h/day, for 6 or
12 wks
M
NS Guinea Pig Concentration- and exposure duration-dependent
(Hartley) increases in airway responsiveness to inhaled histamine
aerosol, significant at 1.0 ppm and above in animals
exposed for 12 wks and at 2 ppm and above in those
exposed for 6 wks. No significant increase in specific
airway resistance (SRawo) values at any NO2
concentration at 6 wks of exposure, but there was a
concentration-dependent increase in this parameter at
12 wks of exposure, with significance at 2.0 ppm and
above. Authors concluded NO2 could be a potent risk
factor for alteration of pulmonary function and airway
responsiveness.
Kobayashi and Miura
(1995)
-------
O
to
O
O
oo
TABLE AX4.5 (cont'd). EFFECTS OF NITROGEN DIOXIDE ON IMMUNE RESPONSES
ppm Exposure
Gender
Age
Species
(Strain)
Effects
Reference
24 h/day, for 12
wks
M
10 wks Rat (Wistar) No change in body weight or absolute lung weight in any
exposed group compared to controls, but relative lung
weight significantly increased in animals exposed to 4.0
ppm NO2. Number of lung cells from animals exposed to
>2.0 ppm significantly reduced (but extent of reduction at
4.0 ppm < 2.0 ppm), but number of mast cells not
significantly different in exposed animals compared to
controls. IgE-mediated histamine release from lung mast
cells significantly reduced at 2.0 ppm NO2 but not at 4.0
ppm or 1.0 ppm, but no difference observed in A23 187-
stimulated histamine release in lung mast cells in any
exposed group compared to controls. Results from
histamine release suggest that NO2 exposure to rat lung
mast cells did not induce histamine-releasing activity in
rats.
Fujimaki and Nohara
(1994)
M
H
6
o
o
H
O
O
H
W
O
O
HH
H
W
10 wks Guinea Pig No change in body weight or absolute or relative lung
(Hartley) weight in any exposed group compared to controls.
Number of lung cells or mast cells not significantly
different in exposed animals compared to controls.
Increasing trend observed in IgE-mediated histamine
release from lung mast cells, the change becoming
significant only at 4.0 ppm. No significant change in
ionophore A2318-stimulated histamine release in exposed
groups compared to controls. Data suggest NO2 exposure
in guinea pigs enhanced histamine-releasing activity in
lung mast cells. Thus, species differences exist between
the rat and guinea pig in response to induction of
_ histamine-releasing activity following NO2 exposure.
-------
O
to
O
O
oo
TABLE AX4.5 (cont'd). EFFECTS OF NITROGEN DIOXIDE ON IMMUNE RESPONSES
ppm Exposure
Gender
Age
Species
(Strain)
Effects
Reference
5 6 h/day for 1, 3,
25 or 5 days
NS
NS Mouse Exposure to 25 ppm, but not 5 ppm, NO2 induced acute lung
(C57BL/6) injury (characterized by increases in protein, LDH,
macrophages, and neutrophils recovered by bronchoalveolar
lavage) that peaked after 3 days, lesions within terminal
bronchioles, and AHR. OVA-sensitized animals exposed to
25 ppm, but not 5 ppm, NO2 showed augmentation of
eosinophilic inflammation and terminal bronchiolar lesions,
which extended significantly into the alveoli. There was no
increased expression of mucus cell-associated gene products
in non-sensitized or OVA-sensitized animals exposed at any
concentration.
Poynter et al. (2006)
oo
5-HETE = 5-Hydroxyeicosatrtraenoate
AM = Alveolar macrophage
AHR = AHR = Airway hyperresponsiveness
BAL = Bronchoalveolar lavage
BHR = Bronchopulmonary hyperreactivity
IFNy = Interferon y
IgA = Immunoglobulin A
IgE = Immunoglobulin E
IgG = Immunoglobulin G
IL-5 = Interleukin-5
LDH = Lactate dehydrogenase
LTB4 = Leukotriene B4
OVA = Ovalbumin
PGE2 = Prostaglandin E2
PMA = Phorbol myristate acetate
ROS = Reactive oxygen species
SRawo = Specific airway resistance
TNFa = Tumor necrosis factor a
TxB2 = Thromboxane B2
ZAS = Zymosan-activated rat serum
H
6
o
o
H
O
O
H
W
O
O
HH
H
W
-------
TABLE AX4.6. EFFECT OF NITROGEN DIOXIDE ON SUSCEPTIBILITY TO
INFECTIOUS AGENTS
to
O
O
oo
ppm
0.05 base +
0.1 peak
Exposure
Continuous, base +
twice/day 1 h peaks,
5 days/wk for 15 days
Gender
F
Age
NS
Species
(Strain)
Mouse
(CD-I)
Infective Agent Effects
Streptococcus sp. No effect.
References
Gardner (1980, 1982)
Graham etal. (1987)
0.5
+
peak
Increased mortality.
H
6
o
o
H
O
O
H
W
O
O
HH
H
W
1.2 base +
2.5 peak
0.2 base + 23 h/day, 7 days/wk F
0.8 peak base+ twice daily 1 h
peaks, 5 days/wk for
lyr
0.3-0.5 Continuous, 3 mos F
Continuous, 6 mos
Increased mortality.
6-8 Mouse Streptococcus sp. Peak plus baseline caused
wks (CD-I) significantly greater mortality
than baseline.
4 wks Mouse A/PR/8 virus High incidence of
(ICR:JCL) adenomatous proliferation
peripheral and bronchial
epithelial cells; NO2 alone
and virus alone caused less
severe alterations.
No enhancement of effect of
NO2 and virus.
Miller etal. (1987)
Motomiya et al. (1973)
-------
o
to
o
o
oo
TABLE AX4.6 (cont'd). EFFECT OF NITROGEN DIOXIDE ON SUSCEPTIBILITY TO
INFECTIOUS AGENTS
H
6
O
O
H
O
O
H
W
O
O
HH
H
W
ppm
0.5
0.5-1.0
10
0.5-28
0.5
0.5
1.0
1.5
5.0
Exposure
Intermittent, 6 or
18 h/day, up to
12 mos
Continuous,
90 days
Continuous,
39 days
2 h/day, 1, 3, and
5 days
Varied
3 h/day, 3 mos
24 h/day,
7 days/wk,
3 mos
3 days
Species
Gender Age (Strain) Infective Agent Effects
F NS Mouse (Swiss) K. pneumoniae Increased mortality after 6 mos
intermittent exposure or after 3, 6,
9, or 12 mos continuous exposure,
increased mortality was
significant only in continuously
exposed mice.
F NS Mouse (ICR, A/PR/8 Increased susceptibility to
dd) virus infection.
F NS Mouse (CD-I) Streptococcus sp. Increase mortality with increased
time and concentration;
concentrations is more important
than time.
F 6-8 Mouse (CD2 Streptococcus sp. Increase in mortality with
wks F!, CD-I) reduction in mean survival time.
F NS Mouse (CF-1) K. pneumoniae Significant increase in mortality
after 3-day exposure to 5.0 ppm;
no effect at other concentrations,
but control mortality very high.
References
Ehrlich and Henry
(1968)
Ito (1971)
Gardner et al.
(1977 a,b)
Coffin etal. (1977)
Ehrlich etal. (1979)
McGrath and
Oyervides (1985)
-------
TABLE AX4.6 (cont'd). EFFECT OF NITROGEN DIOXIDE ON SUSCEPTIBILITY TO
INFECTIOUS AGENTS
to
o
o
oo
>
£
i
(^
o
>
H
6
o
0
H
O
O
H
W
O
O
HH
H
W
ppm
0.5
1.0
2.0
5.0
1.0
2.3
6.6
1.0
2.5
5.0
10.0
1.0
1.0
3.0
Species
Exposure Gender Age (Strain) Infective Agent Effects
4h M/F 8-10 wks Mouse Mycoplasma Decrease in intrapulmonary killing only
(C57BL/6N) pulmonis at 5.0 ppm.
17 h M NS Mouse (Swiss) S. aureus after No difference in number of bacteria
exposure deposited, but at the two highest
concentrations, there was a decrease in
pulmonary bactericidal activity of 6 and
35%, respectively; no effect at 1.0 ppm
4h F NS Mouse (Swiss) S. aureus Injection with corticosteroids increased
NO2-induced impairment of bactericidal
activity at >2.5 ppm.
48 h M NS Mouse (Swiss Streptococcus sp. Increased proliferation of Streptococcus
Webster) S. aureus in lung of exposed mice but no effect
withX aureus.
3h F 5-6 wks Mouse (CD-I) Streptococcus sp. Exercise on continuously moving
wheels during exposure increased
mortality at 3.0 ppm.
References
Davis et al.
(1991, 1992)
Goldstein et al.
(1974)
Jakab (1988)
Sherwood et al.
(1981)
Illingetal. (1980)
-------
to
H
O
O
H
TABLE AX4.6 (cont'd). EFFECT OF NITROGEN DIOXIDE ON SUSCEPTIBILITY TO
INFECTIOUS AGENTS
to
o
o
oo
ppm
1.0
2.5
5.0
1.5-
50
1.5
Species
Exposure Gender Age (Strain)
6 h/day, 6 days NS 4-6 wks Mouse
(CD-I)
2h NS NS Mouse (NS)
Hamster (NS)
Monkey
(Squirrel)
Continuous or F NS Mouse
intermittent, 7 h/day, (CD-I)
Infective Agent
Cytomegalovirus
K. pneumoniae
Streptococcus sp.
Effects
Increase in virus susceptibility at
5.0 ppm only.
Increased mortality in mice,
hamsters, and monkeys at >3.5,
>35, and 50 ppm NO2, respectively
After 1 wk, mortality with
continuous exposure was greater
References
Rose etal. (1988,
1989)
Ehrlich(1980)
Gardner etal. (1979)
Coffin etal. (1977)
7 days/wk, up to
15 days
5.5
than that for intermittent after 2
wks, no significant difference
between continuous and
intermittent exposure.
Increased mortality with increased
duration of exposure; no significant
difference between continuous and
intermittent exposure; with data
adjusted for total difference in
C x T, mortality essentially the
same.
o
H
W
O
^
O
HH
H
W
-------
o
to
o
o
oo
TABLE AX4.6 (cont'd). EFFECT OF NITROGEN DIOXIDE ON SUSCEPTIBILITY TO
INFECTIOUS AGENTS
ppm
Exposure
Gender
Age
Species
(Strain)
Infective Agent
Effects
References
1.5 base + Continuous 64 h,
4.5 peak then peak for 1,
3.5, or 7 h, then
continuous 18 h
base
NS
Mouse
(CD-I)
Streptococcus sp.
Mortality increased with 3.5- and
7 h single peak when bacterial
challenge was after an 18 h
baseline exposure.
Gardner (1980)
Gardner (1982)
Graham etal. (1987)
1,3.5, or 7 h
>
1.5 7 h/day, 4, 5, and 7 NS
days
NS
Mouse (NS) Streptococcus sp.
Mortality proportional to duration
when bacterial challenge was
immediate, but not 18 h
postexposure.
Elevated temperature (32°C)
increased mortality after 7 days.
Gardner (1982)
wv
O
E>
CTJ
H
1
O
o
0
H
O
o
H
W
O
O
HH
H
W
1.9 4h
3.8
7.0
9.2
14.8
1.5- 3h
5.0
M NS Mouse (NS) S. aureus Physical removal of bacteria
unchanged by exposure.
Bactericidal activity decreased by
7, 14, and 50%, respectively, in
three highest NO2-exposed
groups.
F 6-10 wks Mouse (CF- Streptococcus sp. Increased mortality in mice
1, CD2F!) exposed to >2.0 ppm
Goldstein etal. (1973)
Ehrlich etal. (1977)
Ehrlich(1980)
-------
TABLE AX4.6 (cont'd). EFFECT OF NITROGEN DIOXIDE ON SUSCEPTIBILITY TO
INFECTIOUS AGENTS
to
O
O
oo
ppm Exposure
1.5 2h
2.5
3.5
5.0
10
15
Species
Gender Age (Strain)
NS 6-8 wks Mouse
(Swiss
Webster)
Infective Agent Effects
K. pneumoniae No effect at 1.5 or 2.5 ppm; increased
mortality at 3.5 ppm and above. Increase
in mortality when K. pneumoniae
challenge 1 and 6 h after 5 or 10 pm NO2
exposure; when K. pneumoniae challenge
27 h following NO2 exposure, effect only
at 15 ppm.
References
Purvis and Ehrlich
(1963)
Ehrlich (1979)
2.0
1.5 h/day,
5 days/wk for
1, 2, and 3 wks
NS
2 wks
Hamster
(Golden
Syrian)
(in vitro)
A/PR/8/34
influenza virus
Peak virus production in tracheal explants Schiff (1977)
occurred earlier.
H
6
O
O
H
O
O
H
W
O
O
HH
H
W
2.5
4.0
5.0
10
15
5.0
4h
NS
10
Continuous,
2 mos
M
NS
Continuous,
1 mo
Mouse
(Swiss)
Monkey
(Squirrel)
S. aureus,
Proteus mirabilis,
Pasteurella
pneumotropica
K. pneumoniae or
A/PR/8 influenza
virus
Concentration-related decrease in
bactericidal activity at >4.0 ppm with S.
aureus when NO2 exposure after bacterial
challenge; when NO2 exposure was
before challenge, effect at 10 ppm; NO2
concentrations >5.0 ppm required to
affect bactericidal activity for other tested
microorganisms.
Increased viral-induced mortality (1/3).
Increase in Klebsiella-induced mortality
(2/7); no control deaths.
Increased virus-induced mortality (6/6)
within 2-3 days after infection; no control
deaths. Increase in Klebsiella-mduced
mortality (1/4), no control deaths.
Jakab (1987, 1988)
Henry etal. (1970)
-------
O
to
O
O
oo
TABLE AX4.6 (cont'd). EFFECT OF NITROGEN DIOXIDE ON SUSCEPTIBILITY TO
INFECTIOUS AGENTS
ppm
Exposure
Gender Age Species (Strain) Infective Agent
Effects
References
5.0
10
4h
M/F
6-10
wks
Mouse
(C57BL6N,
C3H/HeN)
Mycoplasma
pulmonis
NO2 increased incidence and severity Parker et al. (1989)
of pneumonia lesions and decreased
the number of organisms needed to
induce pneumonia; no effect on
physical clearance; decreased
mycoplasmal killing and increased
growth; no effect on specific IgM in
serum; C57B1/6N mice generally
more sensitive than C3H/HeN mice.
At 10 ppm, one strain (C57B1/6N) of
mice had increased mortality.
10 2h
15
35
50
5 NS
M/F NS Monkey
(Squirrel)
NS NS Mice (NS)
K. pneumoniae
Parainfluenza
(murine sendei
virus)
Clearance of bacteria from lungs of
10-, 15-, and 35-ppm groups delayed
or prevented. All three animals in
highest exposed group died.
Altered the severity but not the course
of the infection
Henry etal. (1969)
Jakab (1988)
Source: Modified from U.S. Environmental Protection Agency (1993).
H
6
o
o
H
O
O
H
W
O
O
HH
H
W
-------
TABLE AX4.7. EFFECTS OF NITROGEN DIOXIDE ON LUNG STRUCTURE
O
to
O
O
oo
ppm
Exposure Gender
Age
Species (Strain)
Effects
Reference
6 h/day,
6 days/wk, for
2 wks
NS
Guinea Pig Morphological changes observed at both
concentrations. Ciliated cells displayed pathological
changes such as cytoplasmic vacuolation and protrusion
at 3 ppm, but eosinophils displayed normal
morphology. Minor to major changes in epithelial cells
(decreased number of specific granules, cytoplasmic
vacuolization, and morpholigcal changes in specific
granules) and ciliated cells showed compound cilia,
cytoplasmic vacuolization, and sloughing) at 9 ppm.
Ohashietal. (1994)
5
25
6 h/day for 1,
3, or 5 days
NS
NS
Mouse No lung inflammation or injury observed at 5 ppm at
(C57BL/6) any time, compared to controls. 25 ppm NO2 induced
acute lung injury (characterized by increases in protein,
LDH, macrophages, and neutrophils recovered by
BAL) at 25 ppm that peaked after 3 days, lesions within
terminal bronchioles, and AHR. Another group of mice
exposed for 5 days to 25 ppm and allowed to recover in
room air for 20 days demonstrated resolution of the
pattern of acute lung injury in these animals.
Poynter et al. (2006)
H
6
o
o
H
O
O
H
W
O
O
HH
H
W
5
10
20
5
10
20
24 h/day, for
3 days
24 h/day, for
25 days
M
NS
M
NS
Rat (Sprague- Significant alteration in morphology of Clara cells (loss
Dawley) of apical intra-luminal projects and damaged epithelium
covered by a layer of CClO-reactive material) at
> 5 ppm.
Rat (Sprague- No significant alteration of morphology of Clara cells
Dawley) compared to controls.
Earth and Muller
(1999)
Earth and Muller
(1999)
-------
O
to
O
O
oo
TABLE AX4.7 (cont'd). EFFECTS OF NITROGEN DIOXIDE ON LUNG STRUCTURE
ppm
Exposure Gender Age Species (Strain)
Effects
Reference
5 24 h/day, M NS Rat (Sprague- Exposure to 5 ppm showed no significant qualitative changes of
10 for 25 days Dawley) the lung tissue, but animals exhibited slight fibrosis of the
20 centroacinar alveolar septa, respiratory bronchioli, and
interstitium, and irregularly shaped alveolar spaces at > 10 ppm.
Morphometric analysis showed significantly diminished alveolar
surface density at > 10 ppm. Suggests development of
emphysema at > 10 ppm. The avg medial thickness of the
pulmonary artery was significantly increased at > 10 ppm, but at
5 ppm, this parameter was significantly decreased, compared to
controls. Authors reported negative correlation between avg
medial thickness and alveolar surface density.
Earth etal. (1995)
H
6
o
o
H
O
O
H
W
O
O
HH
H
W
5 24 h/day, M NS Rat (Sprague- Histopathology revealed structural alterations extending from
10 for 3 days Dawley) slight interstitial edema after exposure to 5 ppm, to epithelial
20 necrosis and interstitial inflammatory infiltration after exposure
to 10 ppm, and an additional intra-alveolar edema after 20 ppm.
Light microscopic examination did not confirm the qualitative
histological changes, particularly muscularization of intra-acinar
vessels. Exposure to > 10 ppm for 25 days caused emphysema
and slight centrilobular interstitial fibrosis.
Morphometric analysis showed significantly diminished alveolar
surface density at 10 ppm after 25 days of exposure and at 20
ppm after 3 and 25 days of exposure. Avg medial thickness of
the pulmonary artery significantly increased at > 10 ppm, but at
5 ppm, this parameter was significantly decreased both during
after 3-day and 25-day exposures, compared to controls.
Authors regarded the decrease in the medial thickness at 5 ppm
as reflecting a reduction of pulmonary arterial mass and not as a
result of vasodilation. Avg medial thickness and alveolar
surface density were negatively correlated. Study indicates
exposures to as low as 5 ppm is not likely to induce structural
changes in the lung of rats. Effect on the morphometry of the
alveolar region appeared to be time-dependent, since significant
changes were seen at 10 ppm after 25 days, but only at 20 ppm
in the 3-day exposure study.
Earth etal. (1995)
-------
O
to
O
O
oo
TABLE AX4.7 (cont'd). EFFECTS OF NITROGEN DIOXIDE ON LUNG STRUCTURE
ppm
Exposure Gender
Age
Species
(Strain)
Effects
Reference
10
20
14h/day,for
15 days, 20 or
25 days
M
NS
Mouse Initial dose response experiment identified 20 ppm NO2 as
(C57/BL/5) concentration causing lung injury and air inflammation
(marked influx of inflammatory cells in the airways,
predominated by macrophages and neutrophils and to a
lesser extent by lymphocytes) for exposure that lasted 15
days, whereas 10 ppm did not induce significant increase in
leukocyte amount in BAL fluid. Actual study using 20 ppm
NO2 observed induction of air space enlargement
(evidenced by a significant increase in mass-specific lung
volume and volume-weighted alveolar volume). No
significant changes in total alveolar surface area.
Wegman and Herz
(2002)
oo
H
6
O
O
H
O
O
H
W
O
O
HH
H
W
0.8
5
10
24 h/day, for 1
or 3 days
M
NS
Rat (Sprague- Significant increase in Type II cell proliferation (evidenced
Dawley) by increases in AgNOR-number and BrdU-LI) after
exposure to 5 ppm NO2 for 3 days and 10 ppm for 1 and
3 days. Significant increase in bronchiolar epithelial
proliferation (increases in AgNOR-number and BrdU-LI) at
>0.8 ppm for 1 and 3 days. In the bronchial epithelium,
statistically significant increase in proliferation as increase
in AgNOR-number at 10 ppm only after 3 days of exposure
and as increase in BrdU-LI after exposure to 5 and 10 ppm
for > 1 day. Results showed highest rate of epithelial
proliferation in the bronchiolar epithelium compared to
bronchial epithelium and Type II cells. Study indicates cell
proliferation changes beginning at concentrations as low
0.8 ppm NO2 following a single day of exposure.
Earth etal. (1994a)
-------
TABLE AX4.7 (cont'd). EFFECTS OF NITROGEN DIOXIDE ON LUNG STRUCTURE
O
to
O
O
oo
ppm
Exposure Gender
Age
Species
(Strain)
Effects
Reference
H
6
o
o
H
O
O
H
W
O
O
HH
H
W
5 24 h/day, for 3 M
10 or 25 days
20
NS Rat Compared to controls, proliferative activity (evidenced by increase
(Sprague- in AgNOR-number) significantly increased, but with no
Dawley) concentration-dependence in respiratory bronchiolar epithelium at
5 ppm and above after 3-day exposure, but the increase was
concentration-dependent, with significance at > 10 ppm following
25-day exposure. Activity was increased in a concentration-
dependent manner in bronchial epithelium with significance only at
20 ppm after 3 days of exposure and at > 10 ppm following 25 days
of exposure.
Concentration-dependent thickening of alveolar septa and
bronchiolar walls, significant only at > 10 ppm after 3- and 25-day
exposure, but 5 ppm caused significant thickening only in
bronchiolar walls after 25 days, but not 3 days, of exposure
compared to controls.
Statistically significant reduction in alveoli surface density at > 10
ppm after 25 days of exposure and at 20 ppm only after 3 days of
exposure, while alveolar circumference increased statistically
significantly at > 10 ppm after 25 days of exposure and at 20 ppm
following 3 days of exposure. 5 ppm did not significantly affect
these endpoints after 3 or 25 days of exposure compared to controls.
Data suggest concentration- and exposure duration-dependent
development of emphysema, significant only at > 10 ppm, but not at
5 ppm, after 3 or 25 days of exposure.
Alveolar duct length increased significantly at > 10 ppm after 25 days
and at 20 ppm after 3 days of exposure; no effect at 5 ppm.
Radial alveolar count values decreased significantly at 20 ppm after
3 and 25 days and at 10 ppm after 25 days exposure; no effect at
5 ppm.
Increase in avg medial thickness at 10 ppm with significance only
after 25 days of exposure and at 20 ppm with significance after 3 and
25 days of exposure. However, 5 ppm caused significant decrease at
both 3 and 25 days of exposure. Authors suggested effect at 5 ppm
was due to vasodilation induced by nitrogen oxide.
Earth etal. (1994b)
-------
TABLE AX4.7 (cont'd). EFFECTS OF NITROGEN DIOXIDE ON LUNG STRUCTURE
O
to
O
O
oo
ppm
Exposure Gender Age
Species
(Strain)
Effects
Reference
25
50
75
100
150
200
250
5, 15, or 30 min M
2,5, 15, or
30 min
2, 5, or 15 min
NS Rat Animals exposed to >200 ppm NO2 for 30 min died within 24 h
(Fischer- after exposure.
344) NO2 exposure induced proportional increases in LWW, indicative
of pulmonary edematous responses, over the same exposure period
after the 15-min exposures. Animals exposed to 25 ppm NO2 did
not produce observable lung injury (i.e., occurrence of alveolar
fibrin and Type II cell hyperplasia) on exposure for 5 min.
Animals exposed to all exposure concentrations for 15 min showed
alveolar fibrin, while Type II hyperplasia occurred at 50 ppm and
its level of expression correlated with exposure concentration.
After 30 min of exposure, the occurrence of fibrin increased as a
function of exposure concentration, while over a concentration
range of 25-150 ppmNO2, the Type II cell hyperplastic response
increased with increasing exposure concentration. Data suggest
exposure concentration was evidently more important than
exposure time in terms of causing lung injury when high
concentrations of NO2 are inhaled.
Lehnert et al.
(1994)
o
>
H
1
O
o
0
H
O
o
H
W
O
O
HH
H
W
3.6 24 h, 12 h, 8 h, M NS Rat Short-term exposure was not sufficient to produce significant type Rajinietal.
7.2 and6h, (Sprague- I alveolar cell necrosis or a significant migration of inflammatory (1993)
10.8 respectively, Dawley) cells across the interstitium and alveolar epithelium.
14 4 for 3 days,
giving a C x T
of 86.4 ppm h
-------
O
to
O
O
oo
TABLE AX4.7 (cont'd).
ppm Exposure Gender Age
0.5 base + Base M 7 wks
1.5 peak presumably
continuous, two
1-h peaks/day,
for 9 wks
EFFECTS
Species
(Strain)
Rat
(Fischer-
344)
OF NITROGEN DIOXIDE ON LUNG STRUCTURE
Effects
No significant differences in thickness of the alveolar septal
components, between controls and exposed group.
Analysis of parenchymal cell populations showed no significant
differences in the avg volumes of different cell types or in their
surface areas. Total number of fenestrae in the lungs of
NO2~exposed animals occurred at a greater frequency than in
controls, but no significant alterations were found in the connective
tissue matrix or interstitial cell population, suggesting that
connective tissue matrix and interstitial cells of the lung
parenchyma did not undergo significant degeneration on exposure
to the low level of NO2 used in this study.
Reference
Mercer et al.
(1995)
AgNOR = Silver-stainable nucleolar organizer regions
AHR = Airway hyperresponsiveness
BAL = Bronchoalveolar lavage
BrdU-LI = Bromodeoxyuridine-laebling index
LDH = Lactate dehydrogenase
LWW = Lung wet weight
H
6
O
O
H
O
O
H
W
O
O
HH
H
W
-------
TABLE AX4.8. EFFECTS OF NITROGEN DIOXIDE ON PULMONARY FUNCTION
O
to
O
O
oo
Species
ppm Exposure Gender Age (Strain)
0.5 0.5 ppm background M 67 days Rat(Fischer-
1.5 level for 16 h, a 6-h 344)
exposure spike, and
a 2-h downtime;
profile was run each
day for 1, 3, 13, 52
or 78 wks
Effects
No NO2 related-effects on body weight or pulmonary function
(total lung capacity, vital capacity, residual volume (difference
between total lung capacity and vital capacity), respiratory system
compliance, single-breath diffusing capacity of carbon monoxide,
slope of nitrogen wash-out curve, and end-expiratory volume)
observed following 13-, 52-, or 78-wk exposure. NO2~related
effects on small airway function [FVC, peak flow, flow at 50%
(FEF 50%), 25% (FEF 25%) and 10% (FEF 10%) of FVC]
evaluated at 52- or 78-wk exposure or at 26- and 17-wks
postexposure were not significantly different from controls.
However, breathing patterns and mechanics (tidal volume,
expiratory resistance, inspiratory and expiratory time) were
generally greater and FOB was significantly slower in
NO2~exposed animals compared to controls at all of the time
points evaluated.
Reference
Tepper et al.
(1993)
to
H
6
o
o
H
O
O
H
W
O
O
HH
H
W
Base presumably M
continuous, two 1-h
peaks/day, for 9 wks
0.5 base + Base presumably M 7 wks Rat (Fischer- No significant differences in lung volume, total air volume of the Mercer etal.
1.5 peak continuous, two 1-h 344) lungs, total lung tissue volume, surface area, body weight, or (1995)
thickness of the alveolar septal components, between controls and
exposed group.
Analysis of parenchymal cell populations showed no significant
differences in the avg volumes of different cell types or in their
surface areas. Total number of fenestrae in the lungs of
NO2~exposed animals occurred at a greater frequency than in
baseline controls, but no significant alterations were found in the
connective tissue matrix or interstitial cell population, suggesting
that connective tissue matrix and interstitial cells of the lung
parenchyma did not undergo significant degeneration on exposure
to the low level of NO2 used in this study.
NS Mouse 20 ppm NO2 induced development of progressive airflow Wegmanand
(C57/BL/6) obstruction (evidenced by decreases in midexpiratory airflow, Herz (2002)
breathing frequency and tidal volume, with statistical significance
only at day 25 of exposure).
FVC = Forced vital capacity
FOB = Frequency of breathing
10
20
14 h/day, for 15
days, 20 or 25 days
M
-------
TABLE AX4.9. EFFECT OF NITROGEN DIOXIDE ON HEMATOLOGICAL PARAMETERS
o
to
o
o
oo
>
OJ
O
H
6
o
0
H
O
o
H
W
O
O
HH
H
W
ppm Exposure Gender Age
0.05 Continuous NS NS
90 days
0.36 Iwk NS NS
0.5-0.8 + Continuous 1 to M/F 4 wks
1.5 mos
0.8 Continuous, 5 days M 7 wks
1.0 Continuous, M NS
16 mos
1.0 Continuous, M NS
5.0 18 mos
1-30 18 h NS NS
1.3-3.0 2h/day, 15 and NS NS
17 wks
2.0 Continuous, M/F NS
14 mos
M
Species (Strain)
Rat
Guinea Pig
Mouse (ICR:JCL)
Mouse (ICR)
Monkey (Squirrel)
Dog (Mongrel)
Mouse (NS)
Rabbit
(NS)
Monkey (Macaca
speciosa)
Rat
(Sprague-Dawley)
Effects
No effect on blood hemoglobin or RBCs.
Increase of red blood cell D-2,3-diphosphoglycerate
Addition of 50 ppm CO to NO2 failed to affect
carboxy hemoglobin.
No effect on methemoglobin.
No effect on hematocrit or hemoglobin with NO2
and influenza exposure.
No changes in hemoglobin or hematocrit.
Concentration-related increase in methemoglobin
and nitrosylhemoglobin
Decreased RBCs.
With or without NaCl (330 ug/m3): polycythemia
with reduced mean corpuscular volume and normal
mean corpuscular hemoglobin.
References
Shalamberidze (1969)
Mersch etal. (1973)
Nakajima and
Kusumoto (1970)
Nakajima and
Kusumoto (1968)
Fenters etal. (1973)
Wagner etal. (1965)
Case et al. (1979)
Mitina (1962)
Furiosi etal. (1973)
-------
TABLE AX4.9 (cont'd). EFFECT OF NITROGEN DIOXIDE ON HEMATOLOGICAL PARAMETERS
O
to
O
O
oo
ppm
2.0
4.0
Exposure
Continuous, up to
6 wks
1-10 days
Gender
M
NS
Age
8 wks
NS
Species (Strain)
Rat
(Wistar)
Rat
(NS)
Effects
No effect on hemoglobin, hematocrit or RBC count;
no methemoglobin was observed.
Increase in RBC sialic acid.
References
Azoulayetal. (1978)
Kunimoto et al. (1984)
H
6
o
o
H
O
O
H
W
O
O
HH
H
W
4.0
NS
NS
NS NS
Decrease in RBCs.
Mochitate and Miura
(1984)
5-40
Ih
4 mos Mouse (JCL:ICR)
No increase in methemoglobin. Increased nitrite Oda et al. (1981)
and especially nitrate.
10 2 h/day, 5 days/wk, F
up to 30 wks
6-8 Mouse (BALB/c)
wks
Small decrease in hemoglobin and mean corpuscular Holt et al. (1979)
hemoglobin concentration.
Source: Modified from U.S. Environmental Protection Agency (1993).
-------
TABLE AX4.10. EFFECTS OF NITRIC OXIDE ON IRON, ENZYMES, AND
NUCLEIC ACIDS
Effect
Reference
Sodium nitroprusside (NO donor) mobilizes iron from ferritin Reif and Simmons (1990)
Modulation of arachidonic acid metabolism via interference Kanner et al. (1991, 1992)
with iron
Inhibition of aconitase (an enzyme in the Krebs cycle, and
also complex 1 and 2 of the respiratory chain)
Permanent modification of hemoglobin, possibly via
deamination
Deamination of DNA
DNA strand breaks
Inhibition of DNA polymerase and ribonucleotide reductase
Antimitogenic; inhibition of T cell proliferation in rat spleen
cells
Inhibition of DNA synthesis, cell proliferation, and
mitogenesis in vascular tissue
Inhibition of mitogenesis and cell proliferation (vascular
smooth muscle cells)
Adenosine diphosphate ribosylation is stimulated by NO-
generating agents
Hibbsetal. (1988)
Perssonetal. (1990)
Stadleretal. (1991)
Moriguchi et al. (1992)
Wink etal. (1991)
Nguyenetal. (1992)
Lepoivre et al. (1991)
Kwon etal. (1991)
Fu & Blankenhorn (1992)
Nakaki etal. (1990)
Garg and Hassid (1989)
Nakaki etal. (1990)
March 2008
AX4-65
DRAFT-DO NOT QUOTE OR CITE
-------
TABLE AX4.11 A. GENOTOXICITY OF NITROGEN DIOXIDE IN VITRO AND IN PLANTS
o
to
o
o
oo
Test Organism
End Point
Exposure
Comments
Results
Reference
Salmonella TA 100
Salmonella TA 100
Salmonella TA 100
and TA 102
Mutations
Mutations
Mutations
SalmonellaTAlOO SOS repair
E. coll, WP2
E. coll
Bacillus sub tills
spores
V79 hamster cells
Mutations
SOS repair
Mutations
6-10 ppm, 40 mins
10-15 ppm, 6h
Bubbling of 10-90 ppm
through bact. susp., 30 mins
as above
Bubbling of 10-90 ppm
through bact. susp., 30 mins
Bubbling of 10-90 ppm
through bact. susp., 30 mins
Bubbling of 10-90 ppm
through bact. susp., 30 mins
500 ppm, 2-3 h
Concentrations >10 ppm were
bacteriotoxic
Effect not considered solely
attributed to nitrite in suspension.
No effect seen with NO gas.
Chromatid-type 10-100 ppm, 10 mins
aberrations, SCE
Effect shown not to be solely due
to nitric acid or nitrite. No effect
if cells not washed with Hank's
salt solution prior to exposure
Isomuraetal. (1984)
Victorin and
Stahlberg (1988)
Kosakaetal. (1985)
+ Kosakaetal. (1985)
Kosakaetal. (1986,
1987)
Kosakaetal. (1986,
1987)
Sasaki etal. (1980)
Tsudaetal. (1981)
O
>
H
6
o
0
H
O
H
W
O
O
HH
H
W
V79 hamster cells
Don hamster cells
V79 hamster cells
Tradescantia
Tradescantia
Source: Victorin (1994).
SCE
Mutations (8-
G resistance)
DNA single-
strand breaks
Micronuclei in
pollen
Mutations in
stamen hair
2-3 ppm, 10 mins
2-3 ppm, 10 mins
10 ppm, 20 mins
5 ppm, 24 h
50 ppm, 6 h
+ Shiraishi and
Bandow (1985)
Slight response - Isomuraetal. (1984)
Effect not due to formation of + Gorsdorf et al.
nitrite (1990)
+ Ma etal. (1982)
+ Schairer etal. (1979)
-------
TABLE AX4.11B. GENOTICITY OF NITROGEN DIOXIDE IN VIVO
fa
O
to
O
O
oo
Test Organism
Drosophila
Drosophila
Rats
Rats
Mice
Mice
End Point
Recessive lethals
Somatic mutations (wing spot test)
Mutations in lung cells (oubain res.)
Chromosome aberrations in lung cells
Chromosome aberrations in lymphocytes and
spermatocytes
Micronuclei in bone marrow
Exposure
500-7000 ppm, 1 h
50-280 ppm, 2 days
50-560 ppm, >12 days
27 ppm, 3 h
0. 1-10 ppm, 6 h
20 ppm, 23 h
Result Reference
Inoueetal. (1981)
Victorinetal. (1990)
+ Isomuraetal. (1984)
+ Isomuraetal. (1984)
Goochetal. (1977)
Victorinetal. (1990)
Source: Victorin (1994).
TABLE AX4.11C. GENOTOXICITY OF NITRIC OXIDE
Test Organism
Salmonella TA100
Salmonella
Don hamster cells
V79 hamster cells
TK 6 human cells
End Point
Mutations
SOS repair
Mutations (8-AG resistance)
DNA single-strand breaks
Mutations, DNA single-strand breaks
Exposure
25-30 ppm, 40 min
Bubbling of 10-90 ppm
2-3 ppm, 10 min
500 ppm, 30 min
Injection of 0.12-0.38 ml NO
gas/ml of culture medium, 1 h
Result Reference
+ Isomuraetal. (1984)
Kosakaetal. (1985)
+ Isomuraetal. (1984)
Gorsdorfetal. (1990)
+ Nguyenetal. (1992)
H
6
o
2
0
H
O
c
o
H
W
O
V
O
HH
H
W
Salmonella Mutations
TA1535
Rats Mutations in lung cells (oubain res.)
Source: Victorin (1994); Arroyo et al. (1992) added.
30 min to 5-90 ppm + Arroyo et al. (1992)
27 ppm, 3 h - Isomura et al. (1984)
-------
1 AX4.7 REFERENCES
2 Acton, J. D.; Myrvik, Q. N. (1972) Nitrogen dioxide effects on alveolar macrophages. Arch.
3 Environ. Health 24: 48-52.
4 Amoruso, M. A.; Witz, G.; Goldstein, B. D. (1981) Decreased superoxide anion radical
5 production by rat alveolar macrophages following inhalation of ozone or nitrogen
6 dioxide. Life Sci. 28:2215-2221.
7 Aranyi, C.; Fenters, J.; Erhlich, R.; Gardner, D. (1976) Scanning electron microscopy of alveolar
8 macrophages after exposure to oxygen, nitrogen dioxide, and ozone. Environ. Health
9 Perspect. 16: 180.
10 Arner, E. C.; Rhoades, R. A. (1973) Long-term nitrogen dioxide exposure: effects on lung lipids
11 and mechanical properties. Arch. Environ. Health 26: 156-160.
12 Arroyo, P. L.; Hatch-Pigott, V.; Mower, H. F.; Cooney, R. V. (1992) Mutagenicity of nitric
13 oxide and its inhibition by antioxidants. Mutat. Res. Lett. 281: 193-202.
14 Ayaz, K. L.; Csallany, A. S. (1978) Long-term NO2 exposure of mice in the presence and
15 absence of vitamin E. II. Effect of glutathione peroxidase. Arch. Environ. Health 33: 292-
16 296.
17 Azoulay, E.; Soler, P.; Blayo, M. C. (1978) The absence of lung damage in rats after chronic
18 exposure to 2 ppm nitrogen dioxide. Bull. Eur. Physiopathol. Respir. 14: 311-325.
19 Azoulay, E.; Soler, P.; Moreau, J.; Blayo, M.-C. (1980) Effects of low-concentration NOXSO2
20 gas mixtures on lung structure and blood-oxygen affinity in rats. J. Environ. Pathol.
21 Toxicol. 4: 399-409.
22 Azoulay-Dupuis, E.; Torres, M.; Soler, P.; Moreau, J. (1983) Pulmonary NO2 toxicity in neonate
23 and adult guinea pigs and rats. Environ. Res. 30: 322-339.
24 Balabaeva, L.; Tabakova, S. (1985) Lipidnata peroksidatsiya v dvye pokoleniya zhenski beli
25 plukhove, inkhalirani s azoten dvuokis [Lipid peroxidation in two progenies of female
26 albino rats inhaling nitrogen dioxide]. Khig. Zdraveopaz. 28: 41-46.
27 Barth, P. J.; Miiller, B. (1999) Effects of nitrogen dioxide exposure on Clara cell proliferation
28 and morphology. Pathol. Res. Pract. 195: 487-493.
29 Barth, P. J.; Miiller, B.; Wagner, U.; Bittinger, A. (1994a) Assessment of proliferative activity in
30 type II pneumocytes after inhalation of NO2 by agNOR-analysis. Exp. Toxicol. Pathol.
31 46:335-342.
32 Barth, P. J.; Uhlarik, S.; Bittinger, A.; Wagner, U.; Riischoff, J. (1994b) Diffuse alveolar damage
33 in the rat lung after short and long term exposure to nitrogen dioxide. Pathol. Res. Pract.
34 190:33-41.
35 Barth, P. J.; Muller, B.; Wagner, U.; Bittinger, A. (1995) Quantitative analysis of parenchymal
36 and vascular alterations in NO2-induced lung injury in rats. Eur. Respir. J. 8: 1115-1121.
37 Bauer, M. A.; Utell, M. J.; Morrow, P. E.; Speers, D. M.; Gibb, F. R. (1986) Inhalation of 0.30
38 ppm nitrogen dioxide potentiates exercise-induced bronchospasm in asthmatics. Am.
39 Rev. Respir. Dis. 134: 1203-1208.
March 2008 AX4-68 DRAFT-DO NOT QUOTE OR CITE
-------
1 Bermiidez, E. (2001) Detection of poly(ADP-ribose) synthetase activity in alveolar macrophages
2 of rats exposed to nitrogen dioxide and ozone. Inhalation Toxicol. 13: 69-84.
3 Busch, R. H.; Buschbom, R. L.; Cannon, W. C.; Lauhala, K. E.; Miller, F. I; Graham, J. A.;
4 Smith, L. G. (1986) Effects of ammonium nitrate aerosol exposure on lung structure of
5 normal and elastase-impaired rats and guinea pigs. Environ. Res. 39: 237-252.
6 Case, G. D.; Dixon, J. S.; Schooley, J. C. (1979) Interactions of blood metalloproteins with
7 nitrogen oxides and oxidant air pollutants. Environ. Res. 20: 43-65.
8 Cavanagh, D. G.; Morris, J. B. (1987) Mucus protection and airway peroxidation following
9 nitrogen dioxide exposure in the rat. J. Toxicol. Environ. Health 22: 313-328.
10 Chang, L.-Y.; Graham, J. A.; Miller, F. J.; Ospital, J. J.; Crapo, J. D. (1986) Effects of
11 subchronic inhalation of low concentrations of nitrogen dioxide. I. The proximal alveolar
12 region of juvenile and adult rats. Toxicol. Appl. Pharmacol. 83: 46-61.
13 Chow, C. K.; Dillard, C. J.; Tappel, A. L. (1974) Glutathione peroxidase system and lysozyme in
14 rats exposed to ozone or nitrogen dioxide. Environ. Res. 7: 311-319.
15 Coffin, D. L.; Gardner, D. E.; Sidorenko, G. L; Pinigin, M. A. (1977) Role of time as a factor in
16 the toxicity of chemical compounds in intermittent and continuous exposures. Part II.
17 Effects of intermittent exposure. J. Toxicol. Environ. Health 3: 821-828.
18 Connor, L. M.; Bidani, A.; Goerke, J.; Clements, J. A.; Postlethwait, E. M. (2001) NO2
19 interfacial transfer is reduced by phospholipid monolayers. J. Appl. Physiol. 91: 2024-
20 2034.
21 Crapo, J. D.; Barry, B. E.; Chang, L.-Y.; Mercer, R. R. (1984) Alterations in lung structure
22 caused by inhalation of oxidants. J. Toxicol. Environ. Health 13: 301-321.
23 Csallany, A. S. (1975) The effect of nitrogen dioxide on the growth of vitamin E deficient,
24 vitamin E supplemented and DPPD supplemented mice. Fed. Proc. Fed. Am. Soc. Exp.
25 Biol. 34: 913.
26 Davis, J. K.; Davidson, M.; Schoeb, T. R. (1991) Murine respiratory mycoplasmosis: a model to
27 study effects of oxidants. Cambridge, MA: Health Effects Institute; research report no.
28 47.
29 Davis, J. K.; Davidson, M. K.; Schoeb, T. R.; Lindsey, J. R. (1992) Decreased intrapulmonary
30 killing of Mycoplasma pulmonis after short-term exposure to NO2 is associated with
31 damaged alveolar macrophages. Am. Rev. Respir. Dis. 145: 406-411.
32 Devlin, R.; Horstman, D.; Becker, S.; Gerrity, T.; Madden, M.; Koren, H. (1992) Inflammatory
33 response in humans exposed to 2.0 ppm NC>2. Am. Rev. Respir. Dis. 145: A456.
34 Douglas, G. J.; Price, J. F.; Page, C. P. (1994) A method for the long-term exposure of rabbits to
35 environmental pollutant gases. Eur. Respir. J. 7: 1516-1526.
36 Dowell, A. R.; Kilburn, K. H.; Pratt, P. C. (1971) Short-term exposure to nitrogen dioxide:
37 effects on pulmonary ultrastructure, compliance, and the surfactant system. Arch. Intern.
38 Med. 128: 74-80.
39 Drozdz, M.; Luciak, M.; Kosmider, S.; Molska-Drozdz, T.; Ludyga, K.; Pasiewicz, J. (1975)
40 Zmiany enzymatyczne i morfologiczne w osrodkowym ukladzie nerwowym u swinek
March 2008 AX4-69 DRAFT-DO NOT QUOTE OR CITE
-------
1 morskich w przewleklym zatruciu dwutlenkiem azotu [Brain enzyme activity and
2 morphological changes in the central nervous system of guinea pigs chronically
3 intoxicated with nitrogen dioxide]. Bromatol. Chem. Toksykol. 8: 241-249.
4 Drozdz, M.; Kucharz, E.; Ludyga, K.; Molska-Drozdz, T. (1976) Studies on the effect of long-
5 term exposure to nitrogen dioxide on serum and liver proteins level and enzyme activity
6 in guinea pigs. Eur. J. Toxicol. 9: 287-293.
7 Ehrlich, R. (1979) Interaction between environmental pollutants and respiratory infections. In:
8 Gardner, D. E.; Hu, E. P. C.; Graham, J. A., eds. Proceedings of the symposium on
9 experimental models for pulmonary research; February; Hilton Head Island, SC.
10 Research Triangle Park, NC: U.S. Environmental Protection Agency, Health Effects
11 Research Laboratory; EPA report no. EPA-600/9-79-022; pp. 145-163. Available from:
12 NTIS, Springfield, VA; PB80-121049.
13 Ehrlich, R. (1980) Interaction between environmental pollutants and respiratory infections.
14 Environ. Health Perspect. 35: 89-100.
15 Ehrlich, R.; Henry, M. C. (1968) Chronic toxicity of nitrogen dioxide: I. Effect on resistance to
16 bacterial pneumonia. Arch. Environ. Health 17: 860-865.
17 Ehrlich, R.; Silverstein, E.; Maigetter, R.; Fenters, J. D.; Gardner, D. (1975) Immunologic
18 response in vaccinated mice during long-term exposure to nitrogen dioxide. Environ. Res.
19 10:217-223.
20 Ehrlich, R.; Findlay, J. C.; Fenters, J. D.; Gardner, D. E. (1977) Health effects of short-term
21 inhalation of nitrogen dioxide and ozone mixtures. Environ. Res. 14: 223-231.
22 Ehrlich, R.; Findlay, J. C.; Gardner, D. E. (1979) Effects of repeated exposures to peak
23 concentrations of nitrogen dioxide and ozone on resistance to streptococcal pneumonia. J.
24 Toxicol. Environ. Health 5: 631-642.
25 Elsayed, N. M.; Mustafa, M. G. (1982) Dietary antioxidants and the biochemical response to
26 oxidant inhalation: I. Influence of dietary vitamin E on the biochemical effects of
27 nitrogen dioxide exposure in rat lung. Toxicol. Appl. Pharmacol. 66: 319-328.
28 Evans, J. N.; Hemenway, D. R.; Kelley, J. (1989) Early markers of lung injury. Cambridge, MA:
29 Health Effects Institute; research report no. 29. Available from: NTIS, Springfield, VA;
30 PB91-171983.
31 Farahani, H.; Hasan, M. (1990) Effect of NC>2 on lipids and lipid peroxidation in the CNS of the
32 guinea-pig. Pharmacol. Toxicol. (Copenhagen) 66: 146-149.
33 Farahani, H.; Hasan, M. (1991) Effect of nitrogen dioxide on sulfhydryl groups in the central
34 nervous system of guinea pigs. Pharmacol. Toxicol. 69: 56-59.
35 Farahani, H.; Hasan, M. (1992) Nitrogen dioxide induced changes in level of free fatty acids,
36 triglyceride, esterified fatty acid, ganglioside and lipase activity in the guinea pig brain. J.
37 Environ. Sci. Health Part B 27: 53-71.
38 Fenters, J. D.; Findlay, J. C.; Port, C. D.; Ehrlich, R.; Coffin, D. L. (1973) Chronic exposure to
39 nitrogen dioxide: immunologic, physiologic, and pathologic effects in virus-challenged
40 squirrel monkeys. Arch. Environ. Health 27: 85-89.
March 2008 AX4-70 DRAFT-DO NOT QUOTE OR CITE
-------
1 Ferin, J.; Leach, L. J. (1977) The effects of selected air pollutants on clearance of titanic oxide
2 particles from the lungs of rats. In: Walton, W. H.; McGovern, B., eds. Inhaled particles
3 IV: proceedings of an international symposium, part 1; September 1975; Edinburgh,
4 United Kingdom. Oxford, United Kingdom: Pergamon Press, Ltd.; pp. 333-341.
5 Frampton, M. W.; Smeglin, A. M.; Roberts, N. J., Jr.; Finkelstein, J. N.; Morrow, P. E.; Utell, M.
6 J. (1989) Nitrogen dioxide exposure in vivo and human alveolar macrophage inactivation
7 of influenza virus in vitro. Environ. Res. 48: 179-192.
8 Freeman, G.; Crane, S. C.; Furiosi, N. J.; Stephens, R. J.; Evans, M. J.; Moore, W. D. (1972)
9 Covert reduction in ventilatory surface in rats during prolonged exposure to subacute
10 nitrogen dioxide. Am. Rev. Respir. Dis. 106: 563-579.
11 Freeman, G.; Juhos, L. T.; Furiosi, N. J.; Mussenden, R.; Stephens, R. J.; Evans, M. J. (1974)
12 Pathology of pulmonary disease from exposure to interdependent ambient gases (nitrogen
13 dioxide and ozone). Arch. Environ. Health 29: 203-210.
14 Fu, Y.; Blankenhorn, E. P. (1992) Nitric oxide-induced anti-mitogenic effects in high and low
15 responder rat strains. J. Immunol. 148: 2217-2222.
16 Fujimaki, H.; Nohara, O. (1994) Changes in the response of lung mast cells isolated from rats
17 and guinea pigs exposed to nitrogen dioxide. Inhalation Toxicol. 6: 515-520.
18 Fujimaki, H.; Shimizu, F.; Kubota, K. (1982) Effect of subacute exposure to NO2 on
19 lymphocytes required for antibody responses. Environ. Res. 29: 280-286.
20 Furiosi, N. J.; Crane, S. C.; Freeman, G. (1973) Mixed sodium chloride aerosol and nitrogen
21 dioxide in air: biological effects on monkeys and rats. Arch. Environ. Health 27: 405-
22 408.
23 Gardner, D. E. (1980) Influence of exposure patterns of nitrogen dioxide on susceptibility to
24 infectious respiratory disease. In: Lee, S. D., ed. Nitrogen oxides and their effects on
25 health. Ann Arbor, MI: Ann Arbor Science Publishers, Inc.; pp. 267-288.
26 Gardner, D. E. (1982) Use of experimental airborne infections for monitoring altered host
27 defenses. Environ. Health Perspect. 43: 99-107.
28 Gardner, D. E.; Holzman, R. S.; Coffin, D. L. (1969) Effects of nitrogen dioxide on pulmonary
29 cell population. J. Bacteriol. 98: 1041-1043.
30 Gardner, D. E.; Coffin, D. L.; Pinigin, M. A.; Sidorenko, G. I. (1977a) Role of time as a factor in
31 the toxicity of chemical compounds in intermittent and continuous exposures. Part I.
32 Effects of continuous exposure. J. Toxicol. Environ. Health 3: 811-820.
33 Gardner, D. E.; Miller, F. J.; Blommer, E. J.; Coffin, D. L. (1977b) Relationships between
34 nitrogen dioxide concentration, time, and level of effect using an animal infectivity
35 model. In: Dimitriades, B., ed. International conference on photochemical oxidant
36 pollution and its control: proceedings, v. I; September 1976; Raleigh, NC. Research
37 Triangle Park, NC: U.S. Environmental Protection Agency, Environmental Sciences
38 Research Laboratory; pp. 513-525; EPA report no. EPA-600/3-77-00la. Available from:
39 NTIS, Springfield, VA; PB-264232. (Ecological research series).
40 Gardner, D. E.; Miller, F. J.; Blommer, E. J.; Coffin, D. L. (1979) Influence of exposure mode on
41 the toxicity of NO2. Environ. Health Perspect. 30: 23-29.
March 2008 AX4-71 DRAFT-DO NOT QUOTE OR CITE
-------
1 Garg, U. C.; Hassid, A. (1989) Inhibition of rat mesangial cell mitogenesis by nitric oxide-
2 generating vasodilators. Am. J. Physiol. 257: F60-F66.
3 Gelzleichter, T. R.; Witschi, H.; Last, J. A. (1992) Synergistic interaction of nitrogen dioxide and
4 ozone on rat lungs: acute responses. Toxicol. Appl. Pharmacol. 116: 1-9.
5 Gilmour, M. I; Park, P.; Belgrade, M. K. (1996) Increased immune and inflammatory responses
6 to dust mite antigen in rats exposed to 5 ppm NC>2. Fundam. Appl. Toxicol. 31: 65-70.
7 Ginsberg, G. L.; Foos, B. P.; Firestone, M. P. (2005) Review and analysis of inhalation
8 dosimetry methods for application to children's risk assessment. J. Toxicol. Environ.
9 Health Part A 68: 573-615.
10 Goldstein, E.; Eagle, M. C.; Hoeprich, P. D. (1973) Effect of nitrogen dioxide on pulmonary
11 bacterial defense mechanisms. Arch. Environ. Health 26: 202-204.
12 Goldstein, E.; Warshauer, D.; Lippert, W.; Tarkington, B. (1974) Ozone and nitrogen dioxide
13 exposure: murine pulmonary defense mechanisms. Arch. Environ. Health 28: 85-90.
14 Goldstein, B. D.; Hamburger, S. J.; Falk, G. W.; Amoruso, M. A. (1977) Effect of ozone and
15 nitrogen dioxide on the agglutination of rat alveolar macrophages by concanavalin A.
16 LifeSci. 21: 1637-1644.
17 Gooch, P. C.; Luippold, H. E.; Creasia, D. A.; Brewen, J. G. (1977) Observations on mouse
18 chromosomes following nitrogen dioxide inhalation. Mutat. Res. 48: 117-119.
19 Gorsdorf, S.; Appel, K. E.; Engeholm, C.; Obe, G. (1990) Nitrogen oxide induces DNA single-
20 strand breaks in cultured Chinese hamster cells. Carcinogenesis 11: 37-41.
21 Graham, J. A.; Gardner, D. E.; Blommer, E. J.; House, D. E.; Menache, M. G.; Miller, F. J.
22 (1987) Influence of exposure patterns of nitrogen dioxide and modifications by ozone on
23 susceptibility to bacterial infectious disease in mice. J. Toxicol. Environ. Health 21: 113-
24 125.
25 Green, N. D.; Schneider, S. L. (1978) Effects of NO2 on the response of baboon alveolar
26 macrophages to migration inhibitory factor. J. Toxicol. Environ. Health 22: 655-662.
27 Gregory, R. E.; Pickrell, J. A.; Hahn, F. F.; Hobbs, C. H. (1983) Pulmonary effects of
28 intermittent subacute exposure to low-level nitrogen dioxide. J. Toxicol. Environ. Health
29 11:405-414.
30 Guth, D. J.; Mavis, R. D. (1985) Biochemical assessment of acute nitrogen dioxide toxicity in rat
31 lung. Toxicol. Appl. Pharmacol. 81: 128-138.
32 Guth, D. J.; Mavis, R. D. (1986) The effect of lung a-tocopherol content on the acute toxicity of
33 nitrogen dioxide. Toxicol. Appl. Pharmacol. 84: 304-314.
34 Hadley, J. G.; Gardner, D. E.; Coffin, D. L.; Menzel, D. B. (1977) Inhibition of antibody-
35 mediated rosette formation by alveolar macrophages: a sensitive assay for metal toxicity.
36 J. Reticuloendothel. Soc. 22: 417-426.
37 Halliwell, B.; Hu, M. L.; Louie, S.; Duvall, T. R.; Tarkington, B. K.; Motchnik, P.; Cross, C. E.
38 (1992) Interaction of nitrogen dioxide with human plasma. Antioxidant depletion and
39 oxidative damage. FEES Lett. 313: 62-66.
March 2008 AX4-72 DRAFT-DO NOT QUOTE OR CITE
-------
1 Hatch, G. E.; Slade, R.; Belgrade, M. K.; Stead, A. G. (1986) Nitrogen dioxide exposure and
2 lung antioxidants in ascorbic acid-deficient guinea pigs. Toxicol. Appl. Pharmacol. 82:
3 351-359.
4 Haydon, G. B.; Davidson, J. T.; Lillington, G. A.; Wasserman, K. (1967) Nitrogen dioxide-
5 induced emphysema in rabbits. Am. Rev. Respir. Dis. 95: 797-805.
6 Henry, M. C.; Ehrlich, R.; Blair, W. H. (1969) Effect of nitrogen dioxide on resistance of squirrel
7 monkeys to Klebsiellapneumoniae infection. Arch. Environ. Health 18: 580-587.
8 Henry, M. C.; Findlay, J.; Spangler, J.; Ehrlich, R. (1970) Chronic toxicity of NO2 in squirrel
9 monkeys: III. effect on resistance to bacterial and viral infection. Arch. Environ. Health
10 20: 566-570.
11 Hibbs, J. B.; Taintor, R. R.; Vavrin, Z.; Rachlin, E. M. (1988) Nitric oxide: a cytotoxic activated
12 macrophage effector molecule. Biochem. Biophys. Res. Commun. 157: 87-94.
13 Hochscheid, R.; Schuchmann, U.; Kotte, E.; Kranz, S.; Heinrichs, S.; Muller, B. (2005) NO2-
14 induced acute and chronic lung injury cause imbalance of glutathione metabolism in type
15 II pneumocytes. Med. Sci. Monit. 11: BR273-BR279.
16 Holt, P. G.; Finlay-Jones, L. M.; Keast, D.; Papadimitrou, J. M. (1979) Immunological function
17 in mice chronically exposed to nitrogen oxides (NOX). Environ. Res. 19: 154-162.
18 Hooftman, R. N.; Kuper, C. F.; Appelman, L. M. (1988) Comparative sensitivity of histo-
19 pathology and specific lung parameters in the detection of lung injury. J. Appl. Toxicol.
20 8: 59-65.
21 Hyde, D.; Orthoefer, J.; Dungworth, D.; Tyler, W.; Carter, R.; Lum, H. (1978) Morphometric
22 and morphologic evaluation of pulmonary lesions in beagle dogs chronically exposed to
23 high ambient levels of air pollutants. Lab. Invest. 38: 455-469.
24 Ichinose, T.; Sagai, M. (1982) Studies on biochemical effects of nitrogen dioxide: III. changes of
25 the antioxidative protective systems in rat lungs and of lipid peroxidation by chronic
26 exposure. Toxicol. Appl. Pharmacol. 66: 1-8.
27 Ichinose, T.; Sagai, M. (1989) Biochemical effects of combined gases of nitrogen dioxide and
28 ozone. III. Synergistic effects on lipid peroxidation and antioxidative protective systems
29 in the lungs of rats and guinea pigs. Toxicology 59: 259-270.
30 Ichinose, T.; Sagai, M.; Kubota, K. (1983) [Changes of lipid peroxidation and antioxidative
31 protective systems in lungs of rats exposed acutely, subacutely and chronically to
32 nitrogen dioxide]. Taiki Osen Gakkaishi 18: 132-146.
33 Uling, J. W.; Miller, F. J.; Gardner, D. E. (1980) Decreased resistance to infection in exercised
34 mice exposed to NO2 and Os. J. Toxicol. Environ. Health 6: 843-851.
35 Inoue, H.; Fukunaga, A.; Okubo, S. (1981) Mutagenic effects of nitrogen dioxide combined with
36 methylurea and ethylurea in Drosophila melanogaster. Mutat. Res. 88: 281-290.
37 Isomura, K.; Chikahira, M.; Teranishi, K.; Hamada, K. (1984) Induction of mutations and
38 chromosome aberrations in lung cells following in vivo exposure of rats to nitrogen
39 oxides. Mutat. Res. 136: 119-125.
March 2008 AX4-73 DRAFT-DO NOT QUOTE OR CITE
-------
1 Ito, K. (1971) [Effect of nitrogen dioxide inhalation on influenza virus infection in mice].
2 Nippon Eiseigaku Zasshi 26: 304-314.
3 Jakab, G. J. (1987) Modulation of pulmonary defense mechanisms by acute exposures to
4 nitrogen dioxide. Environ. Res. 42: 215-228.
5 Jakab, G. J. (1988) Modulation of pulmonary defense mechanisms against viral and bacterial
6 infections by acute exposures to nitrogen dioxide. Cambridge, MA: Health Effects
7 Institute; research report no. 20.
8 Kanner, J.; Harel, S.; Granit, R. (1991) Nitric oxide as an antioxidant. Arch. Biochem. Biophys.
9 289: 130-136.
10 Kanner, J.; Harel, S.; Granit, R. (1992) Nitric oxide, an inhibitor of lipid oxidation by
11 lipoxygenase, cyclooxygenase and hemoglobin. Lipids 27: 46-49.
12 Katz, G. V.; Laskin, S. (1976) Pulmonary macrophage response to irritant gases. In: Aharonson,
13 E. F.; Ben-David, A.; Klingberg, M. A.; Kaye, M., eds. Air pollution and the lung:
14 proceedings of the twentieth annual "Oholo" biological conference; March 1975; Ma'alot,
15 Israel. New York, NY: John Wiley & Sons; pp. 83-100.
16 Kelly, F. J.; Tetley, T. D. (1997) Nitrogen dioxide depletes uric acid and ascorbic acid but not
17 glutathione from lung lining fluid. Biochem. J. 325: 95-99.
18 Kelly, F. J.; Blomberg, A.; Frew, A.; Holgate, S. T.; Sandstrom, T. (1996a) Antioxidant kinetics
19 in lung lavage fluid following exposure of humans to nitrogen dioxide. Am. J. Respir.
20 Crit. CareMed. 154: 1700-1705.
21 Kelly, F. J.; Cotgrove, M.; Mudway, I. S. (1996b) Respiratory tract lining fluid antioxidants: the
22 first line of defence against gaseous pollutants. Cent. Eur. J. Public Health 4(suppl.): 11-
23 14.
24 Kitabatake, M.; Yamamoto, H.; Yuan, P. F.; Manjurul, H.; Murase, S.; Yamauchi, T. (1995)
25 Effects of exposure to NC>2 or SC>2 on bronchopulmonary reaction induced by Candida
26 albicans in guinea pigs. J. Toxicol. Environ. Health 45: 75-82.
27 Kleinman, M. T.; Mautz, W. J. (1991) The effects of exercise on dose and dose distribution of
28 inhaled automotive pollutants. Cambridge, MA: Health Effects Institute; research report
29 no. 45.
30 Kobayashi, T.; Miura, T. (1995) Concentration- and time-dependent increase in specific airway
31 resistance after induction of airway hyperresponsiveness by subchronic exposure of
32 guinea pigs to nitrogen dioxide. Fundam. Appl. Toxicol. 25: 154-158.
33 Kosaka, H.; Oda, Y.; Uozumi, M. (1985) Induction ofumuC gene expression by nitrogen dioxide
34 in Salmonella typhimurium. Mutat. Res. 142: 99-102.
35 Kosaka, H.; Yamamoto, K.; Oda, Y.; Uozumi, M. (1986) Induction of SOS functions by nitrogen
36 dioxide in Escherichia coli with different DNA-repair capacities. Mutat. Res. 162: 1-5.
37 Kosaka, H.; Uozumi, M.; Nakajima, T. (1987) Induction of SOS functions in Escherichia coli
38 and biosynthesis of nitrosamine in rabbits by nitrogen dioxide. Environ. Health Perspect.
39 73:153-156.
March 2008 AX4-74 DRAFT-DO NOT QUOTE OR CITE
-------
1 Kumae, T.; Arakawa, H. (2006) Comparison of effects of in vivo nitrogen dioxide exposure
2 starting from different periods on alveolar macrophage activity, assessed by a
3 chemiluminescence technique in Brown-Norway rats. Luminescence 21: 226-232.
4 Kunimoto, M.; Mochitate, K.; Kaya, K.; Miura, T.; Kubota, K. (1984) Effects of nitrogen
5 dioxide on red blood cells of rats: alterations of cell membrane components and
6 populational changes of red blood cells during in vivo exposure to NO2. Environ. Res.
7 33:361-369.
8 Kwon, N. S.; Stuehr, D. J.; Nathan, C. F. (1991) Inhibition of tumor cell ribonucleotide reductase
9 by macrophage-derived nitric oxide. J. Exp. Med. 174: 761-767'.
10 Last, J. A. (1989) Effects of inhaled acids on lung biochemistry. Environ. Health Perspect. 79:
11 115-119.
12 Last, J. A.; Warren, D. L. (1987) Synergistic interaction between nitrogen dioxide and respirable
13 aerosols of sulfuric acid or sodium chloride on rat lungs. Toxicol. Appl. Pharmacol. 90:
14 34-42.
15 Last, J. A.; Gerriets, J. E.; Hyde, D. M. (1983) Synergistic effects on rat lungs of mixtures of
16 oxidant air pollutants (ozone or nitrogen dioxide) and respirable aerosols. Am. Rev.
17 Respir. Dis. 128: 539-544.
18 Lee, J.-S.; Afifi, A. A.; Mustafa, M. G. (1989) Effects of short-term, single and combined
19 exposure of rats to NO2 and Os on lung tissue enzyme activities. Inhalation Toxicol. 1:
20 21-35.
21 Lee, J.-S.; Mustafa, M. G.; Afifi, A. A. (1990) Effects of short-term, single and combined
22 exposure to low-level NO2 and O3 on lung tissue enzyme activities in rats. J. Toxicol.
23 Environ. Health 29: 293-305.
24 Lefkowitz, S. S.; McGrath, J. J.; Lefkowitz, D. L. (1986) Effects of NO2 on immune responses.
25 J. Toxicol. Environ. Health 17: 241-248.
26 Lehnert, B. E.; Archuleta, D. C.; Ellis, T.; Session, W. S.; Lehnert, N. M.; Gurley, L. R.; Stavert,
27 D. M. (1994) Lung injury following exposure of rats to relatively high mass
28 concentrations of nitrogen dioxide. Toxicology 89: 239-277.
29 Lepoivre, M.; Fieschi, F.; Coves, J.; Thelander, L.; Fontecave, M. (1991) Inactivation of
30 ribonucleotide reductase by nitric oxide. Biochem. Biophys. Res. Commun. 179: 442-
31 448.
32 Ma, T.-H.; Anderson, V. A.; Shmed, I. (1982) Environmental clastogens detected by meiotic
33 pollen mother cells of Tradescantia. Environ. Sci. Res. 25: 141-157.
34 Maigetter, R. Z.; Fenters, J. D.; Findlay, J. C.; Ehrlich, R.; Gardner, D. E. (1978) Effect of
35 exposure to nitrogen dioxide on T and B cells in mouse spleens. Toxicol. Lett. 2: 157-
36 161.
37 Mauderly, J. L.; Bice, D. E.; Carpenter, R. L.; Gillett, N. A.; Henderson, R. F.; Pickrell, J. A.;
38 Wolff, R. K. (1987) Effects of inhaled nitrogen dioxide and diesel exhaust on developing
39 lung. Cambridge, MA: Health Effects Institute; research report no. 8.
March 2008 AX4-75 DRAFT-DO NOT QUOTE OR CITE
-------
1 Mauderly, J. L.; Cheng, Y. S.; Gillett, N. A.; Griffith, W. C.; Henderson, R. F.; Pickrell, J. A.;
2 Wolff, R. K. (1990) Influence of preexisting pulmonary emphysema on susceptibility of
3 rats to chronic inhalation exposure to nitrogen dioxide. Inhalation Toxicol. 2: 129-150.
4 McGrath, J. J.; Oyervides, J. (1985) Effects of nitrogen dioxide on resistance to Klebsiella
5 pneumoniae in mice. J. Am. Coll. Toxicol. 4: 227-231.
6 Menzel, D. B. (1970) Toxicity of ozone, oxygen, and radiation. Annu. Rev. Pharmacol. 10: 379-
7 394.
8 Mercer, R. R.; Costa, D. L.; Crapo, J. D. (1995) Effects of prolonged exposure to low doses of
9 nitric oxide or nitrogen dioxide on the alveolar septa of the adult rat lung. Lab. Invest. 73:
10 20-28.
11 Mersch, J.; Dyce, B. J.; Haverback, B. J.; Sherwin, R. P. (1973) Diphosphoglycerate content of
12 red blood cells: measurements in guinea pigs exposed to 0.4 ppm nitrogen dioxide. Arch.
13 Environ. Health 27: 94-95.
14 Miller, F. J.; Overton, J. H.; Myers, E. T.; Graham, J. A. (1982) Pulmonary dosimetry of nitrogen
15 dioxide in animals and man. In: Schneider, T.; Grant, L., eds. Air pollution by nitrogen
16 oxides: proceedings of the US-Dutch international symposium; May; Maastricht, The
17 Netherlands. Amsterdam, The Netherlands: Elsevier Scientific Publishing Company; pp.
18 377-386. (Studies in environmental science 21).
19 Miller, F. J.; Graham, J. A.; Raub, J. A.; Illing, J. W.; Menache, M. G.; House, D. E.; Gardner,
20 D. E. (1987) Evaluating the toxicity of urban patterns of oxidant gases. II. Effects in mice
21 from chronic exposure to nitrogen dioxide. J. Toxicol. Environ. Health 21: 99-112.
22 Miller, P. D.; Ainsworth, D.; Lam, H. F.; Amdur, M. O. (1988) Indomethacin and cromolyn
23 sodium alter ozone-induced changes in lung function and plasma eicosanoid
24 concentrations in guinea pigs. Toxicol. Appl. Pharmacol. 93: 175-186.
25 Mitina, L. S. (1962) K voproso o sovmestnom vliyanii na organizm dvuokici azota i sernistogo
26 gaza v malykh kontsentratsiyakh [The combined effect of small concentrations of
27 nitrogen dioxide and sulfur dioxide gases]. Gig. Sanit. 27: 3-8.
28 Miyanishi, K.; Kinouchi, T.; Kataoka, K.; Kanoh, T.; Ohnishi, Y. (1996) In vivo formation of
29 mutagens by intraperitoneal administration of poly cyclic aromatic hydrocarbons in
30 animals during exposure to nitrogen dioxide. Carcinogenesis 17: 1483-1490.
31 Mochitate, K.; Miura, T. (1984) In vivo effect of nitrogen dioxide on the activities of glycolytic
32 enzymes in red blood cells of rats. Toxicol. Lett. 22: 315-321.
33 Mochitate, K.; Kaya, K.; Miura, T.; Kubota, K. (1984) In vivo effects of nitrogen dioxide on
34 membrane constituents in lung and liver of rats. Environ. Res. 33: 17-28.
35 Mochitate, K.; Miura, T.; Kubota, K. (1985) An increase in the activities of glycolytic enzymes
36 in rat lungs produced by nitrogen dioxide. J. Toxicol. Environ. Health 15: 323-331.
37 Mochitate, K.; Takahashi, Y.; Ohsumi, T.; Miura, T. (1986) Activation and increment of alveolar
38 macrophages induced by nitrogen dioxide. J. Toxicol. Environ. Health 17: 229-239.
39 Mohsenin, V. (1991) Lipid peroxidation and antielastase activity in the lung under oxidant stress:
40 role of antioxidant defenses. J. Appl. Physiol. 70: 1456-1462.
March 2008 AX4-76 DRAFT-DO NOT QUOTE OR CITE
-------
1 Moriguchi, M.; Manning, L. R.; Manning, J. M. (1992) Nitric oxide can modify amino acid
2 residues in proteins. Biochem. Biophys. Res. Commun. 183: 598-604.
3 Motomiya, K.; Ito, K.; Yoshida, A.; Idewara, S.; Otsu, Y.; Nakajima, Y. (1973) [The effects of
4 exposure to NC>2 gas on the infection of influenza virus of mouse - long term experiment
5 in low concentration]. Kankyo Kagaku Kenkyu Hokoku (Chiba Daigaku) 1: 27-33.
6 Miiller, B.; Schafer, H.; Barth, P.; Von Wichert, P. (1994) Lung surfactant components in
7 bronchoalveolar lavage after inhalation of NC>2 as markers of altered surfactant
8 metabolism. Lung 172: 61-72.
9 Miiller, B.; Garn, H.; Hochscheid, R. (2003) Impaired recycling of surfactant-like liposomes in
10 type II pneumocytes from injured lungs. Thorax 58: 127-134.
11 Mustafa, M. G.; Elsayed, N.; Lim, J. S. T.; Postlethwait, E.; Lee, S. D. (1979) Effects of nitrogen
12 dioxide on lung metabolism. In: Grosjean, D., ed. Nitrogenous air pollutants: chemical
13 and biological implications. Ann Arbor, MI: Ann Arbor Science Publishers, Inc.; pp.
14 165-178.
15 Mustafa, M. G.; Elsayed, N. M.; Von Dohlen, F. M.; Hassett, C. M.; Postlethwait, E. M.; Quinn,
16 C. L.; Graham, J. A.; Gardner, D. E. (1984) A comparison of biochemical effects of
17 nitrogen dioxide, ozone, and their combination in mouse lung. I. Intermittent exposures.
18 Toxicol. Appl. Pharmacol. 72: 82-90.
19 Nakajima, T.; Kusumoto, S. (1968) [Effect of nitrogen dioxide exposure on the contents of
20 reduced glutathione in mouse lung]. Osaka-Furitsu Koshu Eisei Kenkyusho Kenkyu
21 Hokoku Rodo Ei sei Hen 6: 17-21.
22 Nakajima, T.; Kusumoto, S. (1970) [On carboxyhemoglobin levels in mice continuously exposed
23 to a mixed carbon monoxide and nitrogen dioxide gas]. Osaka-furitsu Koshu Eisei
24 Kenkyusho Kenkyu Hokoku Rodo Eisei Hen 8: 25-28.
25 Nakaki, T.; Nakayama, M.; Kato, R. (1990) Inhibition by nitric oxide and nitric oxide-producing
26 vasodilators of DNA synthesis in vascular smooth muscle cells. Eur. J. Pharmacol. 189:
27 347-353.
28 Nguyen, T.; Brunson, D.; Crespi, C. L.; Penman, B. W.; Wishnok, J. S.; Tannenbaum, S. R.
29 (1992) DNA damage and mutation in human cells exposed to nitric oxide in vitro. Proc.
30 Natl. Acad. Sci. U. S. A. 89: 3030-3034.
31 Oda, H.; Tsubone, H.; Suzuki, A.; Ichinose, T.; Kubota, K. (1981) Alterations of nitrite and
32 nitrate concentrations in the blood of mice exposed to nitrogen dioxide. Environ. Res. 25:
33 294-301.
34 Overton, J. H., Jr. (1984) Physicochemical processes and the formulation of dosimetry models.
35 In: Miller, F. J.; Menzel, D. B., eds. Fundamentals of extrapolation modeling of inhaled
36 toxicants: ozone and nitrogen dioxide. Washington, DC: Hemisphere Publishing
37 Corporation; pp. 93-114.
38 Overton, J. H.; Graham, R. C. (1995) Simulation of the uptake of a reactive gas in a rat
39 respiratory tract model with an asymmetric tracheobronchial region patterned on
40 complete conducting airway cast data. Comput. Biomed. Res. 28: 171-190.
March 2008 AX4-77 DRAFT-DO NOT QUOTE OR CITE
-------
1 Overton, J. H.; Graham, R. C.; Menache, M. G.; Mercer, R. R.; Miller, F. J. (1996) Influence of
2 tracheobronchial region expansion and volume on reactive gas uptake and interspecies
3 dose extrapolations. Inhalation Toxicol. 8: 723-745.
4 Pagani, P.; Romano, M.; Erroi, A.; Ferro, M.; Salmona, M. (1994) Biochemical effects of acute
5 and subacute nitrogen dioxide exposure in rat lung and bronchoalveolar lavage fluid.
6 Arch. Environ. Contam. Toxicol. 27: 426-430.
7 Palmer, M. S.; Exley, R. W.; Coffin, D. L. (1972) Influence of pollutant gases on benzpyrene
8 hydroxylase activity. Arch. Environ. Health 25: 439-442.
9 Parker, R. F.; Davis, J. K.; Cassell, G. H.; White, H.; Dziedzic, D.; Blalock, D. K.; Thorp, R. B.;
10 Simecka, J. W. (1989) Short-term exposure to nitrogen dioxide enhances susceptibility to
11 murine respiratory mycoplasmosis and decreases intrapulmonary killing ofMycoplasma
12 pulmonis. Am. Rev. Respir. Dis. 140: 502-512.
13 Persson, M. G.; Gustafsson, L. E.; Wiklund, N. P.; Moncada, S.; Hedqvist, P. (1990)
14 Endogenous nitric oxide as a probable modulator of pulmonary circulation and hypoxic
15 pressor response in vivo. Acta Physiol. Scand. 140: 449-457.
16 Pinkston, P.; Smeglin, A.; Roberts, N. J., Jr.; Gibb, F. R.; Morrow, P. E.; Utell, M. J. (1988)
17 Effects of in vitro exposure to nitrogen dioxide on human alveolar macrophage release of
18 neutrophil chemotactic factor and interleukin-1. Environ. Res. 47: 48-58.
19 Postlethwait, E. M.; Bidani, A. (1989) Pulmonary disposition of inhaled NO2-nitrogen in isolated
20 rat lungs. Toxicol. Appl. Pharmacol. 98: 303-312.
21 Postlethwait, E. M.; Bidani, A. (1990) Reactive uptake governs the pulmonary air space removal
22 of inhaled nitrogen dioxide. J. Appl. Physiol. 68: 594-603.
23 Postlethwait, E. M.; Bidani, A. (1994) Mechanisms of pulmonary NO2 absorption. Toxicology
24 89:217-237.
25 Postlethwait, E. M.; Mustafa, M. G. (1981) Fate of inhaled nitrogen dioxide in isolated perfused
26 rat lung. J. Toxicol. Environ. Health 7: 861-872.
27 Postlethwait, E. M.; Mustafa, M. G. (1989) Effect of altered dose rate on NO2 uptake and
28 transformation in isolated lungs. J. Toxicol. Environ. Health 26: 497-507.
29 Postlethwait, E. M.; Langford, S. D.; Bidani, A. (1991) Transfer of NO2 through pulmonary
30 epithelial lining fluid. Toxicol. Appl. Pharmacol. 109: 464-471.
31 Postlethwait, E. M.; Langford, S. D.; Bidani, A. (1992) Kinetics of NO2 air space absorption in
32 isolated rat lungs. J. Appl. Physiol. 73: 1939-1945.
33 Postlethwait, E. M.; Langford, S. D.; Jacobson, L. M.; Bidani, A. (1995) NO2 reactive absorption
34 substrates in rat pulmonary surface lining fluids. Free Radical Biol. Med. 19: 553-563.
35 Poynter, M. E.; Persinger, R. L.; Irvin, C. G.; Butnor, K. J.; Van Hirtum, H.; Blay, W.; Heintz,
36 N. H.; Robbins, J.; Hemenway, D.; Taatjes, D. J.; Janssen-Heininger, Y. (2006) Nitrogen
37 dioxide enhances allergic airway inflammation and hyperresponsiveness in the mouse.
38 Am. J. Physiol. 290: L144-L152.
March 2008 AX4-78 DRAFT-DO NOT QUOTE OR CITE
-------
1 Proust, B.; Lacroix, G.; Robidel, F.; Marliere, M.; Lecomte, A.; Vargaftig, B. B. (2002)
2 Interference of a short-term exposure to nitrogen dioxide with allergic airways responses
3 to allergenic challenges in BALB/c mice. Mediators Inflammation 11: 251-260.
4 Purvis, M. R.; Ehrlich, R. (1963) Effect of atmospheric pollutants on susceptibility to respiratory
5 infection: II. Effect of nitrogen dioxide. J. Infect. Dis. 113: 72-76.
6 Rajini, P.; Gelzleichter, T. R.; Last, J. A.; Witschi, H. (1993) Alveolar and airway cell kinetics in
7 the lungs of rats exposed to nitrogen dioxide, ozone, and a combination of the two gases.
8 Toxicol. Appl. Pharmacol. 121: 186-192.
9 Reif, D. W.; Simmons, R. D. (1990) Nitric oxide mediates iron release from ferritin. Arch.
10 Biochem. Biophys. 283: 537-541.
11 Richters, A.; Damji, K. S. (1988) Changes in T-lymphocyte subpopulations and natural killer
12 cells following exposure to ambient levels of nitrogen dioxide. J. Toxicol. Environ.
13 Health 25: 247-256.
14 Richters, A.; Damji, K. S. (1990) The relationship between inhalation of nitrogen dioxide, the
15 immune system, and progression of a spontaneously occurring lymphoma in AKR mice.
16 J. Environ. Pathol. Toxicol. Oncol. 10: 225-230.
17 Robison, T. W.; Duncan, D. P.; Forman, H. J. (1990) Chemoattractant and leukotriene B4
18 production from rat alveolar macrophages exposed to nitrogen dioxide. Am. J. Respir.
19 Cell Mol.Biol. 3:21-26.
20 Robison, T. W.; Murphy, J. K.; Beyer, L. L.; Richters, A.; Forman, H. J. (1993) Depression of
21 stimulated arachidonate metabolism and superoxide production in rat alveolar
22 macrophages following in vivo exposure to 0.5 ppm NC>2. J. Toxicol. Environ. Health 38:
23 273-292.
24 Rombout, P. J. A.; Dormans, J. A. M. A.; Marra, M.; Van Esch, G. J. (1986) Influence of
25 exposure regimen on nitrogen dioxide-induced morphological changes in the rat lung.
26 Environ. Res. 41: 466-480.
27 Rose, R. M.; Fuglestad, J. M.; Skornik, W. A.; Hammer, S. M.; Wolfthal, S. F.; Beck, B. D.;
28 Brain, J. D. (1988) The pathophysiology of enhanced susceptibility to murine
29 cytomegalovirus respiratory infection during short-term exposure to 5 ppm nitrogen
30 dioxide. Am. Rev. Respir. Dis. 137: 912-917.
31 Rose, R. M.; Pinkston, P.; Skornik, W. A. (1989) Altered susceptibility to viral respiratory
32 infection during short-term exposure to nitrogen dioxide. Cambridge, MA: Health Effects
33 Institute; research report no. 24. Available from: NTIS, Springfield, VA; PB90-111139.
34 Rubenchik, B. L.; Glavin, A. A.; Galenko, P. M.; Kilkichko, A. A.; Olenick, I. O.; Artemov, K.
35 V. (1995) Gaseous nitrogen dioxide increases the endogenous synthesis of carcinogenic
36 7V-nitrosodimethylamine in animals. J. Environ. Pathol. Toxicol. Oncol. 14: 111-115.
37 Sagai, M.; Ichinose, T.; Oda, H.; Kubota, K. (1982) Studies on biochemical effects of nitrogen
38 dioxide. II. Changes of the protective systems in rat lungs and of lipid peroxidation by
39 acute exposure. J. Toxicol. Environ. Health 9: 153-164.
40 Sagai, M.; Ichinose, T.; Kubota, K. (1984) Studies on the biochemical effects of nitrogen
41 dioxide. IV. Relation between the change of lipid peroxidation and the antioxidative
March 2008 AX4-79 DRAFT-DO NOT QUOTE OR CITE
-------
1 protective system in rat lungs upon life span exposure to low levels of NC>2. Toxicol.
2 Appl. Pharmacol. 73:444-456.
3 Sagai, M.; Arakawa, K.; Ichinose, T.; Shimojo, N. (1987) Biochemical effects on combined
4 gases of nitrogen dioxide and ozone. I. Species differences of lipid peroxides and
5 phospholipids in lungs. Toxicology 46: 251-265.
6 Santiago, L. Y.; Hann, M. C.; Ben-Jebria, A.; Ultman, J. S. (2001) Ozone adsorption in the
7 human nose during unidirectional airflow. J. Appl. Physiol. 91: 725-732.
8 Sarangapani, R.; Gentry, P. R.; Covington, T. R.; Teeguarden, J. G.; Clewell, H. J., III. (2003)
9 Evaluation of the potential impact of age- and gender-specific lung morphology and
10 ventilation rate on the dosimetry of vapors. Inhalation Toxicol. 15: 987-1016.
11 Sasaki, Y.; Endo, R.; Koido, Y. (1980) Direct mutagens in the gaseous component of automobile
12 exhaust detected with Bacillus subtilis spores. Mutat. Res. 79: 181-184.
13 Schairer, L. A.; Van't Hof, J.; Hayes, C. G.; Burton, R. M.; de Serres, F. J. (1979) Measurement
14 of biological activity of ambient air mixtures using a mobile laboratory for in situ
15 exposures: preliminary results from the Tradescantia plant test system. In: Waters, M. D.;
16 Nesnow, S.; Huisingh, J. L.; Sandhu, S. S.; Claxton, L., eds. Application of short-term
17 bioassays in the fractionation of complex environmental mixtures. New York, NY:
18 Plenum Publishing Corporation; pp. 419-440. (Environmental science research: v. 15).
19 Schiff, L. J. (1977) Effect of nitrogen dioxide on influenza virus infection in hamster trachea
20 organ culture. Proc. Soc. Exp. Biol. Med. 156: 546-549.
21 Schlesinger, R. B. (1987) Intermittent inhalation of nitrogen dioxide: effects on rabbit alveolar
22 macrophages. J. Toxicol. Environ. Health 21: 127-139.
23 Schlesinger, R. B.; Gearhart, J. M. (1987) Intermittent exposures to mixed atmospheres of
24 nitrogen dioxide and sulfuric acid: effect on particle clearance from the respiratory region
25 of rabbit lungs. Toxicology 44: 309-319.
26 Schlesinger, R. B.; Driscoll, K. E.; Vollmuth, T. A. (1987) Effect of repeated exposures to
27 nitrogen dioxide and sulfuric acid mist alone or in combination on mucociliary clearance
28 from the lungs of rabbits. Environ. Res. 44: 294-301.
29 Schlesinger, R. B.; Driscoll, K. E.; Gunnison, A. F.; Zelikoff, J. T. (1990) Pulmonary
30 arachidonic acid metabolism following acute exposures to ozone and nitrogen dioxide. J.
31 Toxicol. Environ. Health 31: 275-290.
32 Schlesinger, R. B.; Weideman, P. A.; Zelikoff, J. T. (1991) Effects of repeated exposures to
33 ozone and nitrogen dioxide on respiratory tract prostanoids. Inhalation Toxicol. 3: 27-36.
34 Sekharam, K. M.; Patel, J. M.; Block, E. R. (1991) Plasma membrane-specific phospholipase-Al
35 activation by nitrogen dioxide in pulmonary artery endothelial cells. Toxicol. Appl.
36 Pharmacol. 107: 545-554.
37 Selgrade, M. K.; Mole, M. L.; Miller, F. J.; Hatch, G. E.; Gardner, D. E.; Hu, P. C. (1981) Effect
38 of NC>2 inhalation and vitamin C deficiency on protein and lipid accumulation in the lung.
39 Environ. Res. 26: 422-437.
March 2008 AX4-80 DRAFT-DO NOT QUOTE OR CITE
-------
1 Belgrade, M. K.; Daniels, M. J.; Grose, E. C. (1991) Evaluation of immunotoxicity of an urban
2 profile of nitrogen dioxide: acute, subchronic, and chronic studies. Inhalation Toxicol. 3:
3 389-403.
4 Seto, K.; Kon, M.; Kawakami, M.; Yagishita, S.; Sugita, K.; Shishido, M. (1975) [Influence of
5 nitrogen dioxide inhalation on the formation of protein in the lung]. Igaku to
6 Seibutsugaku 90: 103-106.
7 Sevanian, A.; Hacker, A. D.; Elsayed, N. (1982) Influence of vitamin E and nitrogen dioxide on
8 lipid peroxidation in rat lung and liver microsomes. Lipids 17: 269-277'.
9 Shalamberidze, O. P. (1969) Sovmestnoe deystvie malykh konistogo gaza i dviokici azota v
10 usloviyakh khroichskogo eksperimenta [The joint action of small concentrations of sulfur
11 dioxide and nitrogen dioxide gases under conditions of a chronic test]. Gig. Sanit. (4): 10-
12 14.
13 Sherwin, R. P.; Carlson, D. A. (1973) Protein content of lung lavage fluid of guinea pigs exposed
14 to 0.4 ppm nitrogen dioxide: disc-gel electrophoresis for amount and types. Arch.
15 Environ. Health 27: 90-93.
16 Sherwin, R. P.; Layfield, L. J. (1974) Proteinuria in guinea pigs exposed to 0.5 ppm nitrogen
17 dioxide. Arch. Environ. Health 28: 336-341.
18 Sherwin, R. P.; Richters, V.; Brooks, M.; Buckley, R. D. (1968) The phenomenon of
19 macrophage congregation in vitro and its relationship to in vivo NC>2 exposure of guinea
20 pigs. Lab. Invest. 18: 269-277.
21 Sherwin, R. P.; Shih, J. C.; Lee, J. D.; Ransom, R. (1986) Serotonin content of the lungs, brains,
22 and blood of mice exposed to 0.45 ppm nitrogen dioxide. J. Am. Coll. Toxicol. 5: 583-
23 588.
24 Sherwood, R. L.; Lippert, W. E.; Goldstein, E.; Tarkington, B. (1981) Effect of ferrous sulfate
25 aerosols and nitrogen dioxide on murine pulmonary defense. Arch. Environ. Health 36:
26 130-135.
27 Shiraishi, F.; Bandow, H. (1985) The genetic effects of the photochemical reaction products of
28 propylene plus NC>2 on cultured Chinese hamster cells exposed in vitro. J. Toxicol.
29 Environ. Health 15: 531-538.
30 Snider, G. L.; Kleinerman, J.; Thurlbeck, W. M.; Bengali, Z. H. (1985) The definition of
31 emphysema: report of a National Heart, Lung, and Blood Institute, Division of Lung
32 Diseases workshop. Am. Rev. Respir. Dis. 132: 182-185.
33 Stadler, J.; Curran, R. D.; Ochoa, J. B.; Harbrecht, B. G.; Hoffman, R. A.; Simmons, R. L.;
34 Billiar, T. R. (1991) Effect of endogenous nitric oxide on mitochondrial respiration of rat
35 hepatocytes in vitro and in vivo. Arch. Surg. 126: 186-191.
36 Stara, J. F.; Dungworth, D. L.; Orthoefer, J. G.; Tyler, W. S., eds. (1980) Long-term effects of air
37 pollutants: in canine species. Cincinnati, OH: U.S. Environmental Protection Agency,
38 Office of Health and Environmental Assessment, Environmental Criteria and Assessment
39 Office; EPA report no. EPA-600/8-80-014.
March 2008 AX4-81 DRAFT-DO NOT QUOTE OR CITE
-------
1 Suzuki, T.; Ikeda, S.; Kanoh, T.; Mizoguchi, I. (1986) Decreased phagocytosis and superoxide
2 anion production in alveolar macrophages of rats exposed to nitrogen dioxide. Arch.
3 Environ. Contam. Toxicol. 15: 733-739.
4 Suzuki, T.; Kanoh, T.; Mizoguchi, I. (1988) Increased urinary excretion of tryptophan
5 metabolites in rats exposed to nitrogen dioxide. Bull. Environ. Contam. Toxicol. 40: 139-
6 146.
7 Tabacova, S.; Nikiforov, B.; Balabaeva, L. (1985) Postnatal effects of maternal exposure to
8 nitrogen dioxide. Neurobehav. Toxicol. Teratol. 7: 785-789.
9 Takahashi, Y.; Mochitate, K.; Miura, T. (1986) Subacute effects of nitrogen dioxide on
10 membrane constituents of lung, liver, and kidney of rats. Environ. Res. 41: 184-194.
11 Tepper, J. S.; Costa, D. L.; Winsett, D. W.; Stevens, M. A.; Doerfler, D. L.; Watkinson, W. P.
12 (1993) Near-lifetime exposure of the rat to a simulated urban profile of nitrogen dioxide:
13 pulmonary function evaluation. Fund. Appl. Toxicol. 20: 88-96.
14 Thomas, H. V.; Mueller, P. K.; Wright, R. (1967) Response of rat lung mast cells to nitrogen
15 dioxide inhalation. J. Air Pollut. Control Assoc. 17: 33-35.
16 Tsuda, H.; Kushi, A.; Yoshida, D.; Goto, F. (1981) Chromosomal aberrations and sister-
17 chromatic exchanges induced by gaseous nitrogen dioxide in cultured Chinese hamster
18 cells. Mutat. Res. 89: 303-309.
19 U.S. Environmental Protection Agency. (1993) Air quality criteria for oxides of nitrogen.
20 Research Triangle Park, NC: Office of Health and Environmental Assessment,
21 Environmental Criteria and Assessment Office; report nos. EPA/600/8-91/049aF-cF. 3v.
22 Available from: NTIS, Springfield, VA; PB95-124533, PB95-124525, and PB95-124517.
23 U.S. Environmental Protection Agency. (1994) Methods for derivation of inhalation reference
24 concentrations and application of inhalation dosimetry. Research Triangle Park, NC:
25 Office of Health and Environmental Assessment, Environmental Criteria and Assessment
26 Office; report no. EPA/600/8-88/066F. Available:
27 http://cfpubl.epa.gov/ncea/cfm/recordisplay.cfm?deid=71993 (11 April, 2005).
28 Van Stee, E. W.; Sloane, R. A.; Simmons, J. E.; Brunnemann, K. D. (1983) In vivo formation of
29 7V-nitrosomorpholine in CD-I mice exposed by inhalation to nitrogen dioxide and by
30 gavage to morpholine. JNCI J. Natl. Cancer Inst. 70: 375-379.
31 Van Stee, E. W.; Sloane, R. A.; Simmons, J. E.; Moorman, M. P.; Brunnemann, K. D. (1995)
32 Endogenous formation of 7V-nitrosomorpholine in mice from 15NO2 by inhalation and
33 morpholine by gavage. Carcinogenesis 16: 89-92.
34 Velsor, L. W.; Postlethwait, E. M. (1997) MVinduced generation of extracellular reactive
35 oxygen is mediated by epithelial lining layer antioxidants. Am. J. Physiol. 17: L1265-
36 L1275.
37 Victorin, K. (1994) Review of the genotoxicity of nitrogen oxides. Mutat. Res. 317: 43-55.
38 Victorin, K.; Stahlberg, M. (1988) A method for studying the mutagenicity of some gaseous
39 compounds in Salmonella typhimurium. Environ. Mol. Mutagen. 11: 65-11.
March 2008 AX4-82 DRAFT-DO NOT QUOTE OR CITE
-------
1 Victorin, K.; Busk, L.; Cederberg, H.; Magnusson, J. (1990) Genotoxic activity of 1,3-butadiene
2 and nitrogen dioxide and their photochemical reaction products in Drosophila and in the
3 mouse bone marrow micronucleus assay. Mutat. Res. 228: 203-209.
4 Vollmuth, T. A.; Driscoll, K. E.; Schlesinger, R. B. (1986) Changes in early alveolar particle
5 clearance due to single and repeated nitrogen dioxide exposures in the rabbit. J. Toxicol.
6 Environ. Health 19: 255-266.
7 Wagner, H.-M. (1970) Absorption von NO und NO2 in MIK- und MAK-Konzentrationen bei der
8 Inhalation [Absorption of NO and NO2 in mik- and mak-concentrations during
9 inhalation]. Staub Reinhalt. Luft 30: 380-381.
10 Wagner, W. D.; Duncan, B. R.; Wright, P. G.; Stokinger, H. E. (1965) Experimental study of
11 threshold limit of NO2. Arch. Environ. Health 10: 455-466.
12 Wegmann, M.; Renz; Herz, U. (2002) Long-term NO2 exposure induces pulmonary
13 inflammation and progressive development of airflow obstruction in C57BL/6 mice: a
14 mouse model for chronic obstructive pulmonary disease. Pathobiology 70: 284-286.
15 Wink, D. A.; Kasprzak, K. S.; Maragos, C. M.; Elespuru, R. K.; Misra, M.; Dunams, T. M.;
16 Cebula, T. A.; Koch, W. H.; Andrews, A. W.; Allen, J. S. (1991) DNA deaminating
17 ability and genotoxicity of nitric oxide and its progenitors. Science 254: 1001-1003.
18 Winter-Sorkina, R. de; Cassee, F. R. (2002) From concentration to dose: factors influencing
19 airborne particulate matter deposition in humans and rats. Bilthoven, The Netherlands:
20 National Institute of Public Health and the Environment (RIVM); report no.
21 650010031/2002. Available: http://www.rivm.nl/bibliotheek/rapporten/650010031.html
22 (13 June 2003).
23 World Health Organization. (1997) Nitrogen oxides. 2nd ed. Geneva, Switzerland: World Health
24 Organization. (Environmental health criteria 188).
25 Yokoyama, E. (1968) Uptake of SO2 and NO2 by the isolated upper airways. Bull. Inst. Public
26 Health (Tokyo) 17: 302-306.
27 Yokoyama, E.; Ichikawa, I; Kawai, K. (1980) Does nitrogen dioxide modify the respiratory
28 effects of ozone? In: Lee, S. D., ed. Nitrogen oxides and their effects on health. Ann
29 Arbor, MI: Ann Arbor Science Publishers, Inc.; pp. 217-229.
March 2008 AX4-83 DRAFT-DO NOT QUOTE OR CITE
-------
i AX5. CHAPTER 5 ANNEX - CONTROLLED HUMAN
2 EXPOSURE STUDIES OF NITROGEN OXIDES
o
4
5 AX5.1 INTRODUCTION
6 This annex summarizes the effects of nitrogen oxides (NOx) on human volunteers
7 exposed under controlled conditions. The goal is to review the scientific literature on human
8 clinical studies of NOx exposure published since the 1993 Air Quality Criteria Document
9 (AQCD) for Oxides of Nitrogen (U.S. Environmental Protection Agency, 1993). Summary
10 findings from the 1993 AQCD are provided below. The primary focus will be on nitrogen
11 dioxide because it is the most abundant NOx species in the atmosphere and there are few human
12 studies of exposure to other NOx species.
13 The following are the conclusions drawn from the review of clinical studies of nitrogen
14 oxide exposure in the 1993 criteria document.
15 1. Nitrogen dioxide causes decrements in lung function, particularly increased
16 airway resistance in healthy subjects at concentrations exceeding 2.0 ppm for
17 2 hours.
18 2. Nitrogen dioxide exposure results in increased airway responsiveness in healthy,
19 nonsmoking subjects exposed to concentrations exceeding 1.0 ppm for exposure
20 durations of 1 hour or longer.
21 3. Nitrogen dioxide exposure at levels above 1.5 ppm may alter numbers and types
22 of inflammatory cells in the distal airways or alveoli, but these responses depend
23 upon exposure concentration, duration, and frequency. Nitrogen dioxide may
24 alter function of cells within the lung and production of mediators that may be
25 important in lung host defenses.
26 4. Nitrogen dioxide exposure of asthmatics causes, in some subjects, increased
27 airway responsiveness to a variety of provocative mediators, including cholinergic
28 and histaminergic chemicals, SO2 and cold air. However, the presence of these
29 responses appears to be influenced by the exposure protocol, particularly whether
30 or not the exposure includes exercise.
31 5. Modest decrements in spirometric measures of lung function (3 to 8%) may occur
32 in some asthmatics and COPD patients under certain NO2 exposure conditions.
33 6. Nitric acid levels in the range of 50 to 200 ppb may cause some pulmonary
34 function responses in adolescent asthmatics, but not in healthy adults. Other
35 commonly occurring NOx species do not appear to cause any pulmonary function
36 responses at concentrations expected in the ambient environment, even at higher
37 levels than in worst-case scenarios. However, not all nitrogen oxides acid species
38 have been studied sufficiently.
March 2008 AX5-1 DRAFT-DO NOT QUOTE OR CITE
-------
1 7. No association between lung function responses and respiratory symptom
2 responses were observed. Furthermore, there is little evidence of a concentration-
3 response relationship for changes in lung function, airway responsiveness, or
4 symptoms at the NO2 levels that are reviewed here.
5
6 In the summary and integration chapter of the 1993 NOX criteria document, one of the
7 key health effects of most concern at near ambient concentrations of NO2 was increases in
8 airway responsiveness of asthmatic individuals after short-term exposures. The 1993 AQCD
9 notes the absence of a concentration-response relationship for NO2 exposure and airways
10 responsiveness in asthmatics. For example, most responses to NO2 that had been observed in
11 asthmatics occurred at concentrations between 0.2 and 0.5 ppm. However, other studies showed
12 an absence of effects on airways responsiveness at much higher concentrations, up to 4 ppm.
13 Since 1993, additional studies suggest that exposure to low concentrations of NO2, either alone
14 or in combination with other pollutants such as SO2, may enhance allergen responsiveness in
15 asthmatic subjects.
16 In the years since the preparation of the 1993 AQCD, many studies from a variety of
17 disciplines have convincingly demonstrated that exposure to particulate air pollution increases
18 the risk for cardiovascular events. In addition, a number of epidemiological studies have shown
19 associations between ambient NO2 levels and adverse cardiovascular outcomes, at concentrations
20 well below those shown to cause respiratory effects. However, to date there remain very few
21 clinical studies of NO2 that include endpoints relevant to cardiovascular disease.
22
23 AX5.1.1 Considerations in Controlled Human Exposure Studies
24
25 Strengths and Limitations of Controlled Human Studies
26 The database for air pollution risk assessment arises from four investigative approaches:
27 epidemiology, animal toxicology, in vitro studies, and human inhalation studies. Each possesses
28 advantages but also carries significant limitations. For example, the epidemiological
29 investigation examines exposures in the "real world" but struggles with the realities of
30 conducting research in the community, where cigarette smoking, socioeconomic status,
31 occupational exposures, meteorological variability, and exposure characterization are important
32 confounders. Outcomes are often evaluated from available health or mortality records or from
33 administered questionnaires, all of which have inherent limitations. Sophisticated measures of
March 2008 AX5-2 DRAFT-DO NOT QUOTE OR CITE
-------
1 physiological responses are often not practical in studies involving large populations, although
2 they may be used in panel studies. In contrast, inhalation studies in animals allow precision in
3 quantifying exposure duration and concentration, measurement of a wide variety of physiologic,
4 biochemical, and histological endpoints, and examination of extremes of the exposure-response
5 relationship. Often, however, interpretation of these studies is constrained by difficulty in
6 extrapolating findings from animals to humans, especially when exposure concentrations are
7 unrealistically high.
8 Controlled, quantitative studies of exposed humans offer a third approach (Frampton
9 et al., 2006). Human clinical studies attempt to engineer laboratory atmospheric conditions
10 relevant to ambient pollutant atmospheres, with careful control of concentrations, duration,
11 timing, and other conditions which may impact responses. These studies provide the opportunity
12 to measure symptoms and physiological markers of health effects that result from breathing the
13 atmospheres. The carefully controlled environment allows investigators to identify responses to
14 individual pollutants, to characterize exposure-response relationships, to examine interactions
15 among pollutants, and to study the effects of other variables such as exercise, humidity, or
16 temperature. Susceptible populations can participate, including individuals with acute and
17 chronic respiratory and cardiovascular diseases, with appropriate limitations based on subject
18 comfort and protection from risk. Endpoint assessment traditionally has included symptoms and
19 pulmonary function, but more recently a variety of markers of pulmonary, systemic, and
20 cardiovascular function have been used to assess pollutant effects.
21 Human clinical studies have limitations. For practical and ethical reasons, studies must
22 be limited to relatively small groups, to short durations of exposure, and to pollutant
23 concentrations that are expected to produce only mild and transient responses. Findings from the
24 short-term exposures in clinical studies may provide limited insight into the health effects of
25 chronic or repeated exposures.
26 Specific issues of protocol design in human clinical studies have been reviewed
27 (Frampton et al., 2006), and will not be considered further here, except in the context of specific
28 studies of NC>2 exposure described in the following pages.
29
30 Assessing the Findings from Controlled Human Studies
31 In clinical studies, humans are the species of interest, so findings have particular
32 relevance in risk assessment. However, the utility of clinical studies in risk assessment is
March 2008 AX5-3 DRAFT-DO NOT QUOTE OR CITE
-------
1 tempered by the obvious need to avoid adverse health effects of the study itself. This usually
2 means selecting subjects that are not the most susceptible to the pollutant being studied.
3 Furthermore, clinical studies depend on outcome markers with variable relevance or validation
4 as markers of true health effects. The statement from the American Thoracic Society, "What
5 constitutes an adverse health effect?" (American Thoracic Society, 2000) addresses issues
6 relevant to selection and interpretation of outcome markers in clinical studies.
7 The 1993 NC>2 AQCD included a description of key outcome measures that had been in
8 use to that date. These included primarily respiratory outcomes, including pulmonary function
9 tests such as spirometry, lung volumes, and airways resistance, and tests of pulmonary clearance
10 of inhaled aerosols. A brief description of bronchoalveolar lavage was also included, which had
11 come into use prior to 1993 to assess airway inflammation and changes in the epithelial lining
12 fluid in response to NC>2 exposure.
13
14
15 AX5.2 EFFECTS OF NITROGEN DIOXIDE IN HEALTHY SUBJECTS
16 Table AX5.2-1 summarizes the key clinical studies of NC>2 exposure in healthy subjects
17 since 1993, with a few key studies included prior to that date.
18
19
20 AX5.3 THE EFFECTS OF NITROGEN OXIDE EXPOSURE IN
21 SENSITIVE SUBJECTS
22 Table AX5.3-1 summarizes studies of potentially sensitive subjects. The potential for
23 NC>2 exposure to enhance responsiveness to allergen challenge in asthmatics deserves special
24 mention. Several recent studies, summarized in Table AX5.3-2, have reported that low-level
25 exposures to NC>2, both at rest and with exercise, enhance the response to specific allergen
26 challenge in mild asthmatics.
27 These recent studies involving allergen challenge suggest that NC>2 may enhance the
28 sensitivity to allergen-induced decrements in lung function, and increase the allergen-induced
29 airway inflammatory response.
March 2008 AX5-4 DRAFT-DO NOT QUOTE OR CITE
-------
1 AX5.4 EFFECTS OF MIXTURES CONTAINING NITROGEN OXIDES
2 Table AX5.4-1 summarizes human clinical studies of NO2-containing mixtures or
3 sequential exposures that are most relevant to ambient exposure scenarios.
March 2008 AX5-5 DRAFT-DO NOT QUOTE OR CITE
-------
O
to
O
O
oo
TABLE AX5.2-1. CLINICAL STUDIES OF NO2 EXPOSURE IN HEALTHY SUBJECTS
Reference
Avissar et al.
(2000)
Location
Rochester,
NY, USA
Participants
21 healthy
nonsmokers
Approach & Methods
Measurements of extracellular
glutathione peroxidase (eGPx)
activity and protein levels in
epithelial lining fluid from NO2
exposure study described in
Frampton et al. (2002) (see
below).
Findings
No effects of NO2 exposure on eGPx
activity and protein concentrations.
(Ozone exposure decreased eGPx activity
and protein concentrations.)
Comments
NO2 up to 1.5 ppm for
3 h did not deplete this
mode of antioxidant
defense in the epithelial
lining fluid.
Azadniv et al.
(1998)
Rochester,
NY, USA
2 studies,
12 healthy
nonsmokers in
each
Air vs. 2 ppm NO2 for 6 h with
intermittent exercise.
Phase 1: BAL 18 h after
exposure;
Phase 2: BAL immediately after
exposure.
Increased BAL neutrophils, decreased
blood CD8+ and null T lymphocytes 18 h
after exposure. No effects on symptoms
or lung function.
2 ppm NO2 for 6 h
caused mild
inflammation.
X
Blomberg et al. Sweden 30 healthy Air vs. 2 ppm NO2 for 4 h, with
(1997) nonsmokers intermittent exercise.
Increased neutrophils and interleukin-8 in
bronchial wash. Increases in specific
lymphocyte subsets in BAL fluid.
Symptoms/lung function not reported.
2 ppm NO2 for 4 h
caused airway
inflammation.
H
6
o
o
H
O
O
H
W
O
O
HH
H
W
Blomberg et al.
(1999)
Sweden
Devlin et al.
(1999)
Chapel Hill,
NC, USA
12 healthy
nonsmokers
8 healthy
nonsmokers
Air vs. 2 ppm NO2 for 4 h on
4 days, with intermittent
exercise.
Air and 2.0 ppm NO2 for 4 h
with intermittent exercise.
After 4 days of NO2, increased
neutrophils in bronchial wash but
decreased neutrophils in bronchial biopsy.
2% decrease in FEVi after first exposure
to NO2, attenuated with repeated
exposure. Symptoms not reported.
Increased bronchial lavage neutrophils,
IL-6, IL-8, alphai-antitrypsin, and tissue
plasminogen activator. Decreased
alveolar macrophage phagocytosis and
superoxide production. No effects on
pulmonary function. Symptoms not
reported.
Decreased lung function,
not confirmed in other
studies at this
concentration.
Conflicting information
on airway inflammation.
2 ppm NO2 for 4 h
caused airway
inflammation.
-------
TABLE AX5.2-1 (cont'd). CLINICAL STUDIES OF NO2 EXPOSURE IN HEALTHY SUBJECTS
O
to
O
O
oo
Reference Location Participants
Approach & Methods
Findings
Comments
Drechsler-
Parks (1995)
Santa
Barbara,
CA, USA
Frampton et al. Rochester,
(1991) NY, USA
8 older healthy 4 2-h exposures with intermittent
nonsmokers exercise: air, 0.60 ppm NO2,
0.45 ppm O3, and 0.60 ppm NO2
+ 0.45 ppm O3.
39 healthy 3 protocols, all for 3 h with
nonsmokers control air exposure:
(1) continuous 0.06 ppm NO2,
(2) baseline 0.05 ppm NO2 with
peaks of 2.0 ppm, and
(3) continuous 1.5 ppmNO2.
Significant reduction in cardiac output
during exercise, estimated using
noninvasive impedance cardiography,
with NO2 + O3. Symptoms and
pulmonary function not reported.
No symptoms or direct effects on
pulmonary function. Increased airways
responsiveness to carbachol after
1.5ppmNO2.
Suggests cardiac effects of
NO2 + O3. Small number
of subjects limits statistical
power, has not been
replicated.
Evidence for increased
nonspecific airways
responsiveness with NO2 as
low as 1.5 ppm for 3 h.
X
Frampton et al. Rochester, 21 healthy
(2002) NY, USA nonsmokers
Exposure to air, 0.6, 1.5 ppm
NO2 for 3 h with intermittent
exercise.
Dose-related decrease in hematocrit,
hemoglobin, blood lymphocytes, and T
lymphocytes. Mild increase in
neutrophils recovered in bronchial
portion of B AL fluid. In vitro viral
challenge of bronchial epithelial cells
showed increased cytotoxicity after
1.5 ppm NO2. No effects on symptoms
or pulmonary function.
Indicates NO2 causes
airway inflammation below
1.5 ppm for 3 h. Suggest
subtle effects on red blood
cells, possibly RBC
destruction (hemolysis).
H
6
o
o
H
O
O
H
W
O
O
HH
H
W
Gong et al.
(2005)
Helleday et al.
(1994)
Downey,
CA, USA
Sweden
6 healthy
nonsmokers and
18 ex-smokers
with COPD
8 healthy
smokers, 8
healthy
nonsmokers
2 h exposures with intermittent
exercise to:
(1) air,
(2)0.4ppmNO2,
(3) 200 ug/m3 concentrated
ambient paniculate matter
(CAPs), (4) NO2 + CAPs.
3.5 ppm NO2 for 20 min with
15 min exercise. BAL 24 h after
exposure compared with non-
exposure control BAL.
Reduced maximum mid-expiratory flow
rate and oxygen saturation with CAPs
exposures; no effects of NO2 alone or
additive effect with CAPs.
Different inflammatory cell increases in
smokers and nonsmokers. No effects on
symptoms. Pulmonary function not
reported.
Exposures not fully
randomized. Small number
of healthy subjects limits
interpretation for healthy
group.
Lack of control air
exposure with exercise is
problematic.
-------
O
to
O
O
oo
TABLE AX5.2-1 (cont'd). CLINICAL STUDIES OF NO2 EXPOSURE IN HEALTHY SUBJECTS
Reference
Location
Participants
Approach & Methods
Findings
Comments
Helleday et al.
(1995)
Sweden
24 healthy
nonsmokers, 8 in
each of 3 groups
Bronchoscopic assessment of
mucociliary activity:
(1) 45 min after 1.5 ppm NO2
for 20 min,
(2) 45 min after 3.5 ppm NO2
for 20 min, and
(3) 24 h after 3.5 ppm NO2 for
4h.
Complete abolition of mucociliary
activity 20 min after NO2; increased
activity 24 h after NO2. Symptoms/
pulmonary function not reported.
No true air control
exposure, order of
procedures not randomized,
subjects not blinded.
X
-------
O
to
O
O
oo
TABLE AX5.2-1 (cont'd). CLINICAL STUDIES OF NO2 EXPOSURE IN HEALTHY SUBJECTS
Reference Location
Participants
Approach & Methods
Findings
Comments
Posin et al. Downey, 10 healthy 3 daily exposures for 2.5 h.
(1978) CA, USA nonsmokers 1st day: air;
2nd and 3rd days: 1 or 2 ppm NO2.
Intermittent exercise. Subsequent
control series of 3 daily air
exposures.
Reduced hemoglobin and hematocrit,
and red blood cell acetyl
cholinesterase.
Suggests red blood cell
effects of NO2(see
Frampton et al., 2002).
Exposures not randomized.
Rasmussen
etal. (1992)
Denmark
14 healthy
nonsmokers
Air vs. 2.3 ppm NO2 for 5 h.
Small increases in FVC
Reduced lung permeability and blood
glutathione peroxidase after exposure.
Only 1 wk between
exposures may have
confounded results.
X
Rigas et al.
(1997)
Sandstrom Sweden
etal. (1990)
12 healthy 2 h of 0.36 ppm NO2, 0.75 ppm
nonsmokers NO2, 0.36 ppm SO2, or 0.36 ppm
O3. Boluses of O3 every 30 min to
measure O3 absorption.
32 healthy 4 ppm NO2 for 20 min with 15 min
nonsmokers, exercise. BAL 4, 8, 24, 72 h after
4 groups of exposure, compared with non-
8 subjects exposure control BAL.
NO2 and SO2 increased O3 absorption
by increasing biochemical substrates.
Increase in BAL mast cells and
lymphocytes 4-24 h after exposure.
Suggests breathing
mixtures of NO2 and O3
would increase O3 dose to
airways.
Study weakened by lack of
control air exposure.
H
6
o
o
H
O
O
H
W
O
O
HH
H
W
Sandstrom Sweden
etal. (1991)
Sandstrom Sweden
etal.
(1992a)
Sandstrom Sweden
etal.
(1992b)
18 healthy
nonsmokers
10 healthy
nonsmoking men
8 healthy
nonsmokers
2.25, 4.0, 5.5 ppm NO2 for 20 min
with light exercise. BAL 24 h after
exposure, compared with non-
exposure control BAL.
4 daily exposures to 4 ppm NO2 for
20 min with 15 min exercise. BAL
24 h after exposure, compared with
non-exposure control BAL.
1.5 ppm NO2 for 20 min with
15 min exercise, every 2nd day x 6.
BAL 24 h after exposure compared
with non-exposure control BAL.
Increase in BAL mast cells (all
concentrations) and lymphocytes
(4.0 and 5.5 ppm).
Reduction in alveolar macrophages,
NK cells, and CDS lymphocytes in
BAL; reduction in total lymphocytes
in blood.
Reduced CD8+ T lymphocytes and NK
cells in BAL fluid.
Study weakened by lack of
control air exposure.
Study weakened by lack of
control air exposure.
Study weakened by lack of
control air exposure.
-------
I TABLE AX5.2-1 (cont'd). CLINICAL STUDIES OF NO2 EXPOSURE IN HEALTHY SUBJECTS
^j Reference Location Participants Approach & Methods Findings Comments
o Solomon San 15 healthy Air or 2.0 ppm NC>2 with intermittent Increased neutrophils in bronchial Airway inflammation with
etal. (2000) Francisco, nonsmokers exercise, for 4 h daily x 4. lavage decreased CD4+T lymphocytes 2 ppm NO2 for 4 daily 4 h
CA, USA BAL18 hours after exposure. in BAL. No changes in blood. exposures.
Vagaggini Italy 7 healthy Air vs. 0.3 ppm NO2 for 1 h with Mild increase in symptoms. No Small number of subjects
et al. (1996) nonsmokers intermittent exercise. effects on lung function, nasal lavage, limits statistical power.
or induced sputum.
X
H
6
o
o
H
O
O
H
W
O
O
HH
H
W
-------
O
to
O
O
oo
TABLE AX5.3-1. EFFECTS OF NO2 EXPOSURE IN SUBJECTS WITH RESPIRATORY DISEASE
(SEE TABLE AX5.3-2 FOR STUDIES WITH ALLERGEN CHALLENGE)
Reference
Location
Participants
Approach & Methods
Findings
Comments
Gong et al.
(2005)
Downey,
CA, USA
6 healthy
nonsmokers and
18 ex-smokers with
COPD
2 h exposures with intermittent
exercise to: (1) air,
(2) 0.4 ppm NO2,
(3) 200 ug/m3 concentrated
ambient participate matter (CAPs),
(4) NO2 + CAPs.
Reduced maximum mid-expiratory flow
rate and oxygen saturation with CAPs
exposures; no effects of NO2 alone or
additive effect with CAPs.
Exposures not fully randomized.
Small number of subjects limits
interpretation for healthy group.
Hackney et al.
(1992)
Downey,
CA, USA
26 smokers with
symptoms and
reduced FEV]
Personal monitoring and chamber
exposure to air and 0.3 ppm NO2
for 4 h with intermittent exercise.
No significant effects on lung function.
>
X
Jorres and
Magnussen
(1991)
Jorres et al.
(1995)
Germany
Germany
11 mild asthmatics
8 healthy
nonsmokers &
12 mild asthmatics
Air vs. 0.25 ppm NO2 for 30 min
with 10 min exercise.
Air or 1 ppm NO2 exposure for 3 h
with intermittent exercise.
No effects on lung function or airways
responsiveness to methacholine.
In asthmatics, 2.5% decrease FEVj after
NO2 vs. 1.3% decrease after air, p = 0.01.
FEV! decreased 20% in 1 subject after
NO2. No significant lung function effect
in healthy subjects. Changes in
eicosanoids (more pronounced in
asthmatics), but not inflammatory cells,
in BAL fluid.
Lung function effects consistent
with other studies, suggesting
some asthmatics susceptible.
Evidence for mild airway
inflammation. Small number of
healthy subjects limits statistical
power.
H
6
o
o
H
O
O
H
W
O
O
HH
H
W
Morrow et al.
(1992)
Strand et al.
(1996)
Vagaggini
etal. (1996)
Rochester,
NY, USA
Sweden
Italy
20 COPD,
20 healthy elderly
19 mild asthmatics
8 mild asthmatics,
7 COPD
Air vs. 0.3 ppm NO2 for 4 h with
intermittent exercise.
Air vs. 0.26 ppm NO2 for 30 min
with intermittent exercise.
Air vs. 0.3 ppm NO2 for 1 h with
intermittent exercise.
Equivocal reduction in FVC with COPD
patients, but not healthy subjects.
Increased airway responsiveness to
histamine 5 h after exposure. No effects
on lung function.
Mild decrease in FEV] in COPD subjects
in comparison with air exposure, but not
with baseline. No effects on nasal lavage
or induced sputum.
Suggests increased nonspecific
airways responsiveness at much
lower concentration than healthy
subjects. Differs from findings in
Jorres and Magnussen (1991).
No convincing effect of NO2 in
this study. Small number of
subjects limits statistical power.
-------
O
to
O
O
oo
TABLE AX5.3-2. EFFECTS OF NO2 EXPOSURE ON RESPONSE TO INHALED ALLERGEN
Reference
Location Participants
Approach & Methods
Findings
Comments
Barck et al. Sweden 13 mild 30 min exposures to air and 0.26 ppm
(2002) asthmatics, NO2 (at rest?), allergen challenge 4 h and
4 ex-smokers BAL 19 h after exposure. Randomized,
crossover, double blind.
Increased PMN in bronchial wash and
BAL fluid, increased eosinophil cationic
protein in bronchial wash, and reduced
cell viability and BAL volume with NO2
+ allergen. No effects on lung function
response to allergen.
Key study suggesting that NO2
enhances inflammatory response
to allergen in mild asthmatics.
Barck etal. Sweden 18 mild Day 1: 15 min exposures,
(2005a) asthmatics, Day 2:215-min exposures to air and
4 ex-smokers 0.26 ppm NO2 separated by 1 h, at rest.
Allergen challenge 4 h after exposure on
day 1 and 3 h after exposure on day 2.
Sputum induction before exposure on
days 1 & 2, and morning of day 3.
Randomized, crossover, single blind.
Increased eosinophilic cationic protein in
sputum and blood, and increased
myeloperoxidase in blood with NO2 +
allergen. No differences in lung function
or sputum cells.
Provides supporting evidence that
NO2 enhances the airway
inflammatory response to allergen.
X
(Si
I
to
Barck etal. Sweden 16 mild 30 min exposures to air and 0.26 ppm
(2005b) asthmatics NO2 at rest, nasal allergen challenge 4 h
with rhinitis after exposure. Nasal lavage before and
at intervals after exposure and challenge.
Devaliaetal. United 8 mild 6 h exposures to combination of 0.4 ppm
(1994) Kingdom asthmatics NO2 and 0.2 ppm SO2.
No significant differences between air
and NO2 exposure.
Increased allergen responsiveness 10 min
after exposure to combination of NO2 and
SO2, but not to individual gases.
0.26 ppm NO2 did not enhance
nasal inflammatory response to
allergen challenge.
Small number of subjects limits
statistical power.
H
6
o
o
H
O
O
H
W
O
O
HH
H
W
Jenkinsetal. United llmild (1) 6-h exposures to air, 0.1 ppm ozone,
(1999) Kingdom asthmatics 0.2 ppm NO2, and combination followed
by allergen challenge;
(2) 3-h exposures to air, 0.2 ppm ozone,
0.4 ppm NO2, and combination;
All exposures with intermittent exercise.
All of the second exposure scenarios
(ozone, NO2, and combination), but none
of the first exposure scenarios, resulted in
reduced concentration of allergen causing
a 20% decline in FEVj. Authors
conclude that concentration more
important than total inhaled pollutant.
Suggests 0.4 ppm for 3 h with
intermittent exercise increases
allergen responsiveness.
-------
O
to
O
O
oo
TABLE AX5.3-2 (cont'd). EFFECTS OF NO2 EXPOSURE ON RESPONSE TO INHALED ALLERGEN
Reference Location Participants Approach & Methods Findings Comments
Rusznak et al.
(1996)
United
Kingdom
13 mild asthmatics
6 h exposures to combination of
0.4 ppm NO2 and 0.2 ppm SO2
Increased allergen responsiveness to
combination of NO2 and SO2, 10 min, 24,
and 48 h after exposure.
Confirms findings of Devalia
et al. (1994), thatNO2 + SO2
for 6 h increases allergen
responsiveness.
Strand et al. Sweden 18 patients with mild
(1997) asthma, age 18-50 yrs
Exposure to 0.26 ppm NO2 for
30 min at rest, allergen challenge 4 h
after exposure.
Late phase, but not early phase, response
to allergen enhanced by NO2.
Suggests 0.26 ppm NO2 for
30 min at rest increases late
response.
Strand et al. Sweden 16 patients with mild to
(1998) moderate asthma, age
21-52 yrs
4 daily repeated exposures to
0.26 ppm NO2 for 30 min at rest.
Significant increases in both early and
late phase response to allergen after 4th
day of exposure.
Suggests repeated 0.26 ppm
NO2 at rest increases allergen
response.
X
Tunnicliffe United 10 nonsmoking mild
etal. (1994) Kingdom asthmatics age
16-60 yrs.
8 subjects completed.
Wang et al. United 2 groups of 8 subjects
(1995a,b) Kingdom with allergic rhinitis
Exposure to air, 0.1 ppm, and
0.4 ppm NO2 for 1 h at rest,
separated by at least 1 wk, followed
by allergen challenge.
Exposure to 0.4 ppm NO2 (at rest?)
for 6 h followed by nasal allergen
challenge and nasal lavage.
Post-challenge reduction in FEV! after
0.4 ppm NO2 was greater than after air,
for both the early (p < 0.009) and late
(p< 0.02) responses. No difference in
nonspecific airway responsiveness.
Increase in myeloperoxidase and
eosinophil cationic protein in nasal
lavage fluid following allergen challenge.
Suggests threshold for
allergen responsiveness effect
is between 0.1 and 0.4 ppm
for 1 h resting exposure.
Suggests enhanced nasal
inflammatory response to
allergen with 0.4 ppm.
H
6
o
o
H
O
O
H
W
O
O
HH
H
W
Wang et al. United 16 subjects with allergic
(1999) Kingdom rhinitis
Treatment with nasal fluticasone or
placebo for 4 wks followed by
exposure to 0.4 ppm NO2 for 6 h,
allergen challenge, and nasal lavage.
Fluticasone suppressed the NO2 and
allergen-induced increase in eosinophil
cationic protein in nasal lavage fluid.
Confirms earlier findings of
this group that 0.4 ppm NO2
enhances nasal allergen
response.
-------
TABLE AX5.4-1. EFFECTS OF EXPOSURE TO NO2 WITH OTHER POLLUTANTS
O
to
O
O
oo
Reference Location Participants
Approach & Methods
Findings
Comments
Devalia et al. United 8 mild asthmatics 6 h exposures to combination of
(1994) Kingdom 0.4 ppm NO2 and 0.2 ppm SO2.
Increased allergen responsiveness
10 min after exposure to combination
of NO2 and SO2, but not to individual
gases.
Small number of subjects
limits statistical power.
Drechsler- Santa 8 older healthy
Parks (1995) Barbara, nonsmokers
CA, USA
4 2-h exposures with intermittent
exercise: air, 0.60 ppm NO2,
0.45 ppm O3, and 0.60 ppm NO2 +
0.45 ppmO3.
Significant reduction in cardiac
output during exercise, estimated
using noninvasive impedance
cardiography, with NO2 + O3.
Symptoms and pulmonary function
not reported.
Suggests cardiac effects of
NO2 + O3. Small number
of subjects limits statistical
power, has not been
replicated.
X
Gong et al. Downey, 6 healthy
(2005) CA, USA nonsmokers and
18 ex-smokers
with COPD
2 h exposures with intermittent
exercise to:
(1) air,
(2)0.4ppmNO2,
(3) 200 ug/m3 concentrated ambient
paniculate matter (CAPs),
(4) NO2 + CAPs.
Reduced maximum mid-expiratory
flow rate and oxygen saturation with
CAPs exposures; no effects of NO2
alone or additive effect with CAPs.
Exposures not fully
randomized. Small
number of healthy subjects
limits interpretation for
healthy group.
H
6
O
O
H
O
O
H
W
O
O
HH
H
W
Hazucha
etal. (1994)
Torres and
Magnussen
(1990)
Chapel
Hill, NC,
USA
Germany
21 healthy female
nonsmokers
14 nonsmoking
mild asthmatics
2 h exposure to air or 0.6 ppm NO2
followed 3 h later by exposure to
0.3 ppm O3, with intermittent
exercise.
NO2 enhanced spirometric responses
and airways responsiveness
following subsequent O3 exposure.
30 min exposures to air, 0.25 ppm NO2 but not SO2 increased airways
NO2, or 0.5 ppm SO2 at rest followed responsiveness to SO2 challenge.
15 min later by 0.75 ppm SO2
hyperventilation challenge.
0.6 ppm NO2 enhanced
ozone responses.
Findings contrast with
Rubenstein et al. (1990).
-------
O
to
O
O
oo
TABLE AX5.4-1 (cont'd). EFFECTS OF EXPOSURE TO NO2 WITH OTHER POLLUTANTS
Reference Location Participants
Approach & Methods
Findings
Comments
Koenig et al. Seattle, 28 asthmatic
(1994) WA, USA adolescents; 6
subjects did not
complete.
Exposure for 90 min with
intermittent exercise to:
(1) 0.12 ppm ozone + 0.3 ppm NO2
(2) 0.12 ppm ozone + 0.3 ppm NO2
68 ug/m3 H2SO4, or
(3) 0.12 ppm ozone + 0.3 ppm NO2
0.05 ppm nitric acid.
No effects on pulmonary
function.
Absence of lung function
effects of 0.3 ppmNO2
consistent with other studies;
no effects of mixtures.
H
6
o
o
H
O
O
H
W
O
O
HH
H
W
Rubenstein San 9 stable 30 min exposures to air or 0.3 ppm
etal. (1990) Francisco, asthmatics NO2 with 20 min exercise, followed
CA, USA 1 h later by SO2 inhalation challenge.
No effects on pulmonary
function or SO2 responsiveness.
Findings contrast with Torres
and Magnussen et al. (1990).
X
Rudell et al. Sweden 10 healthy Air and diesel exhaust for 1 h, with
(1999) nonsmokers and without particle trap. NO2
concentration 1.2-1.3 ppm. BAL
24 h after exposures.
Rusznak United 13 mild 6 h exposures to combination of
et al. (1996) Kingdom asthmatics 0.4 ppm NO2 and 0.2 ppm SO2.
Increased neutrophils in BAL
fluid, no significant reduction in
effect with particle trap.
Increased allergen
responsiveness to combination
of NO2 and SO2, 10 min, 24, and
48 h after exposure.
Filter only partially trapped
particles. Unable to draw
conclusions about role of NO2
in causing effects.
Confirms findings of Devalia
et al. (1994), that NO2 + SO2
for 6 h increases allergen
responsiveness.
-------
1 AX5.5 REFERENCES
2 American Thoracic Society. (2000) What constitutes an adverse health effect of air pollution?
3 Am. J. Respir. Crit. Care Med. 161: 665-673.
4 Avissar, N. E.; Reed, C. K.; Cox, C.; Frampton, M. W.; Finkelstein, J. N. (2000) Ozone, but not
5 nitrogen dioxide, exposure decreases glutathione peroxidases in epithelial lining fluid of
6 human lung. Am. J. Respir. Crit. Care Med. 162: 1342-1347.
7 Azadniv, M.; Utell, M. J.; Morrow, P. E.; Gibb, F. R.; Nichols, J.; Roberts, N. J., Jr.; Speers, D.
8 M.; Torres, A.; Tsai, Y.; Abraham, M. K.; Voter, K. Z.; Frampton, M. W. (1998) Effects
9 of nitrogen dioxide exposure on human host defense. Inhalation Toxicol. 10: 585-602.
10 Barck, C.; Sandstrom, T.; Lundahl, J.; Hallden, G.; Svartengren, M.; Strand, V.; Rak, S.; Bylin,
11 G. (2002) Ambient level of NO2 augments the inflammatory response to inhaled allergen
12 in asthmatics. Respir. Med. 96: 907-917.
13 Barck, C.; Lundahl, J.; Hallden, G.; Bylin, G. (2005a) Brief exposures to NO2 augment the
14 allergic inflammation in asthmatics. Environ. Res. 97: 58-66.
15 Barck, C.; Lundahl, J.; Holmstrom, M.; Bylin, G. (2005b) Does nitrogen dioxide affect
16 inflammatory markers after nasal allergen challenge? Am. J. Rhinol. 19: 560-566.
17 Blomberg, A.; Krishna, M. T.; Bocchino, V.; Biscione, G. L.; Shute, J. K.; Kelly, F. J.; Frew, A.
18 J.; Holgate, S. T.; Sandstrom, T. (1997) The inflammatory effects of 2 ppm NO2 on the
19 airways of healthy subjects. Am. J. Respir. Crit. Care Med. 156: 418-424.
20 Blomberg, A.; Krishna, M. T.; Helleday, R.; Soderberg, M.; Ledin, M.-C.; Kelly, F. J.; Frew, A.
21 J.; Holgate, S. T.; Sandstrom, T. (1999) Persistent airway inflammation but accomodated
22 antioxidant and lung function responses after repeated daily exposure to nitrogen dioxide.
23 Am. J. Respir. Crit. Care Med. 159: 536-543.
24 Devalia, J. L.; Rusznak, C.; Herdman, M. J.; Trigg, C. J.; Tarraf, H.; Davies, R. J. (1994) Effect
25 of nitrogen dioxide and sulphur dioxide on airway response of mild asthmatic patients to
26 allergen inhalation. Lancet 344: 1668-1671.
27 Devlin, R. B.; Horstman, D. P.; Gerrity, T. R.; Becker, S.; Madden, M. C. (1999) Inflammatory
28 response in humans exposed to 2.0 PPM nitrogen dioxide. Inhalation Toxicol. 11: 89-
29 109.
30 Drechsler-Parks, D. M. (1995) Cardiac output effects of 63 and NO2 exposure in healthy older
31 adults. Toxicol. Ind. Health 11: 99-109.
32 Frampton, M. W.; Morrow, P. E.; Cox, C.; Gibb, F. R.; Speers, D. M.; Utell, M. J. (1991) Effects
33 of nitrogen dioxide exposure on pulmonary function and airway reactivity in normal
34 humans. Am. Rev. Respir. Dis. 143: 522-527.
35 Frampton, M. W.; Boscia, J.; Roberts, N. J., Jr.; Azadniv, M.; Torres, A.; Cox, C.; Morrow, P.
36 E.; Nichols, J.; Chalupa, D.; Frasier, L. M.; Gibb, F. R.; Speers, D. M.; Tsai, Y.; Utell, M.
37 J. (2002) Nitrogen dioxide exposure: effects on airway and blood cells. Am. J. Physiol.
38 282:L155-L165.
March 2008 AX5-16 DRAFT-DO NOT QUOTE OR CITE
-------
1 Frampton, M. W.; Pietropaoli, A. P.; Morrow, P. E.; Utell, M. J. (2006) Human clinical studies
2 of airborne pollutants. In: Gardner, D. E. Toxicology of the lung. Boca Raton, FL: CRC
3 Press; pp. 29-82. (Target organ toxicology series).
4 Gong, H., Jr., Linn, W. S.; Clark, K. W.; Anderson, K. R.; Geller, M. D.; Sioutas, C. (2005)
5 Respiratory responses to exposures with fine particulates and nitrogen dioxide in the
6 elderly with and without COPD. Inhalation Toxicol. 17: 123-132.
7 Hackney, J. D.; Linn, W. S.; Avol, E. L.; Shamoo, D. A.; Anderson, K. R.; Solomon, J. C.;
8 Little, D. E.; Peng, R.-C. (1992) Exposures of older adults with chronic respiratory illness
9 to nitrogen dioxide: a combined laboratory and field study. Am. Rev. Respir. Dis. 146:
10 1480-1486.
11 Hazucha, M. J.; Folinsbee, L. J.; Seal, E.; Bromberg, P. A. (1994) Lung function response of
12 healthy women after sequential exposures to NO2 and O^. Am. J. Respir. Crit. Care Med.
13 150: 642-647.
14 Helleday, R.; Sandstrom, T.; Stjernberg, N. (1994) Differences in bronchoalveolar cell response
15 to nitrogen dioxide exposure between smokers and nonsmokers. Eur. Respir. J. 7: 1213-
16 1220.
17 Helleday, R.; Huberman, D.; Blomberg, A.; Stjernberg, N.; Sandstrom, T. (1995) Nitrogen
18 dioxide exposure impairs the frequency of the mucociliary activity in healthy subjects.
19 Eur. Respir. J. 8: 1664-1668.
20 Jenkins, H. S.; Devalia, J. L.; Mister, R. L.; Bevan, A. M.; Rusznak, C.; Davies, R. J. (1999) The
21 effect of exposure to ozone and nitrogen dioxide on the airway response of atopic
22 asthmatics to inhaled allergen: dose- and time-dependent effects. Am. J. Respir. Crit.
23 Care Med. 160:33-39.
24 Jorres, R.; Magnussen, H. (1990) Airways response of asthmatics after a 30 min exposure, at
25 resting ventilation, to 0.25 ppm NO2 or 0.5 ppm SO2. Eur. Respir. J. 3: 132-137.
26 Jorres, R.; Magnussen, H. (1991) Effect of 0.25 ppm nitrogen dioxide on the airway response to
27 methacholine in asymptomatic asthmatic patients. Lung 169: 77-85.
28 Jorres, R.; Nowak, D.; Grimminger, F.; Seeger, W.; Oldigs, M.; Magnussen, H. (1995) The
29 effect of 1 ppm nitrogen dioxide on bronchoalveolar lavage cells and inflammatory
30 mediators in normal and asthmatic subjects. Eur. Respir. J. 8: 416-424.
31 Kim, S. U.; Koenig, J. Q.; Pierson, W. E.; Hanley, Q. S. (1991) Acute pulmonary effects of
32 nitrogen dioxide exposure during exercise in competitive athletes. Chest 99: 815-819.
33 Koenig, J. Q.; Covert, D. S.; Pierson, W. E.; Hanley, Q. S.; Rebolledo, V.; Dumler, K.;
34 McKinney, S. E. (1994) Oxidant and acid aerosol exposure in healthy subjects and
35 subjects with asthma. Part I: effects of oxidants, combined with sulfuric or nitric acid, on
36 the pulmonary function of adolescents with asthma. Cambridge, MA: Health Effects
37 Institute; pp. 1-36; research report no. 70.
38 Morrow, P. E.; Utell, M. J.; Bauer, M. A.; Smeglin, A. M.; Frampton, M. W.; Cox, C.; Speers,
39 D. M.; Gibb, F. R. (1992) Pulmonary performance of elderly normal subjects and
40 subjects with chronic obstructive pulmonary disease exposed to 0.3 ppm nitrogen
41 dioxide. Am. Rev. Respir. Dis. 145: 291-300.
March 2008 AX5-17 DRAFT-DO NOT QUOTE OR CITE
-------
1 Pathmanathan, S.; Krishna, M. T.; Blomberg, A.; Helleday, R.; Kelly, F. J.; Sandstrom, T.;
2 Holgate, S. T.; Wilson, S. J.; Frew, A. J. (2003) Repeated daily exposure to 2 ppm
3 nitrogen dioxide upregulates the expression of IL-5, IL-10, IL-13, and ICAM-1 in the
4 bronchial epithelium of healthy human airways. Occup. Environ. Med. 60: 892-896.
5 Posin, C.; Clark, K.; Jones, M. P.; Patterson, J. V.; Buckley, R. D.; Hackney, J. D. (1978)
6 Nitrogen dioxide inhalation and human blood biochemistry. Arch. Environ. Health 33:
7 318-324.
8 Rasmussen, T. R.; Kjaergaard, S. K.; Tarp, U.; Pedersen, O. F. (1992) Delayed effects of NO2
9 exposure on alveolar permeability and glutathione peroxidase in healthy humans. Am.
10 Rev. Respir. Dis. 146: 654-659.
11 Rigas, M. L.; Ben-Jebria, A.; Ultman, J. S. (1997) Longitudinal distribution of ozone absorption
12 in the lung: effects of nitrogen dioxide, sulfur dioxide, and ozone exposures. Arch.
13 Environ. Health 52: 173-178.
14 Rubenstein, I; Bigby, B. G.; Reiss, T. F.; Boushey, H. A., Jr. (1990) Short-term exposure to 0.3
15 ppm nitrogen dioxide does not potentiate airway responsiveness to sulfur dioxide in
16 asthmatic subjects. Am. Rev. Respir. Dis. 141: 381-385.
17 Rudell, B.; Blomberg, A.; Helleday, R.; Ledin, M.-C.; Lundback, B.; Stjernberg, N.; Horstedt,
18 P.; Sandstrom, T. (1999) Bronchoalveolar inflammation after exposure to diesel exhaust:
19 comparison between unfiltered and particle trap filtered exhaust. Occup. Environ. Med.
20 56: 527-534.
21 Rusznak, C.; Devalia, J. L.; Davies, R. J. (1996) Airway response of asthmatic subjects to
22 inhaled allergen after exposure to pollutants. Thorax 51: 1105-1108.
23 Sandstrom, T.; Andersson, M. C.; Kolmodin-Hedman, B.; Stjernberg, N.; Angstrom, T. (1990)
24 Bronchoalveolar mastocytosis and lymphocytosis after nitrogen dioxide exposure in man:
25 a time-kinetic study. Eur. Respir. J. 3: 138-143.
26 Sandstrom, T.; Stjernberg, N.; Eklund, A.; Ledin, M.-C.; Bjermer, L.; Kolmodin-Hedman, B.;
27 Lindstrom, K.; Rosenhall, L.; Angstrom, T. (1991) Inflammatory cell response in
28 bronchoalveolar lavage fluid after nitrogen dioxide exposure of healthy subjects: a dose-
29 response study. Eur. Respir. J. 4: 332-339.
30 Sandstrom, T.; Helleday, R.; Bjermer, L.; Stjernberg, N. (1992a) Effects of repeated exposure to
31 4 ppm nitrogen dioxide on bronchoalveolar lymphocyte subsets and macrophages in
32 healthy men. Eur. Respir. J. 5: 1092-1096.
33 Sandstrom, T.; Ledin, M.-C.; Thomasson, L.; Helleday, R.; Stjernberg, N. (1992b) Reductions in
34 lymphocyte subpopulations after repeated exposure to 1.5 ppm nitrogen dioxide. Br. J.
35 Ind. Med. 49: 850-854.
36 Solomon, C.; Christian, D. L.; Chen, L. L.; Welch, B. S.; Kleinman, M. T.; Dunham, E.; Erie, D.
37 J.; Balmes, J. R. (2000) Effect of serial-day exposure to nitrogen dioxide on airway and
38 blood leukocytes and lymphocyte subsets. Eur. Respir. J. 15: 922-928.
39 Strand, V.; Salomonsson, P.; Lundahl, J.; Bylin, G. (1996) Immediate and delayed effects of
40 nitrogen dioxide exposure at an ambient level on bronchial responsiveness to histamine in
41 subjects with asthma. Eur. Respir. J. 9: 733-740.
March 2008 AX5-18 DRAFT-DO NOT QUOTE OR CITE
-------
1 Strand, V.; Rak, S.; Svartengren, M.; Bylin, G. (1997) Nitrogen dioxide exposure enhances
2 asthmatic reaction to inhaled allergen in subjects with asthma. Am. J. Respir. Crit. Care
3 Med. 155: 881-887.
4 Strand, V.; Svartengren, M.; Rak, S.; Barck, C.; Bylin, G. (1998) Repeated exposure to an
5 ambient level of NO2 enhances asthmatic response to nonsymptomatic allergen dose. Eur.
6 Respir. J. 12: 6-12.
7 Tunnicliffe, W. S.; Burge, P. S.; Ayres, J. G. (1994) Effect of domestic concentrations of
8 nitrogen dioxide on airway responses to inhaled allergen in asthmatic patients. Lancet
9 344:1733-1736.
10 U.S. Environmental Protection Agency. (1993) Air quality criteria for oxides of nitrogen.
11 Research Triangle Park, NC: Office of Health and Environmental Assessment,
12 Environmental Criteria and Assessment Office; report nos. EPA/600/8-91/049aF-cF. 3v.
13 Available from: NTIS, Springfield, VA; PB95-124533, PB95-124525, and PB95-124517.
14 Vagaggini, B.; Paggiaro, P. L.; Giannini, D.; Franco, A. D.; Cianchetti, S.; Carnevali, S.;
15 Taccola, M.; Bacci, E.; Bancalari, L.; Dente, F. L.; Giuntini, C. (1996) Effect of short-
16 term NO2 exposure on induced sputum in normal, asthmatic and COPD subjects. Eur.
17 Respir. J. 9: 1852-1857.
18 Wang, J. H.; Devalia, J. L.; Duddle, J. M.; Hamilton, S. A.; Davies, R. J. (1995a) Effect of six-
19 hour exposure to nitrogen dioxide on early-phase nasal response to allergen challenge in
20 patients with a history of seasonal allergic rhinitis. J. Allergy Clin. Immunol. 96: 669-
21 676.
22 Wang, J. H.; Duddle, J.; Devalia, J. L.; Davies, R. J. (1995b) Nitrogen dioxide increases
23 eosinophil activation in the early-phase response to nasal allergen provocation. Int. Arch.
24 Allergy Immunol. 107: 103-105.
25 Wang, J. H.; Devalia, J. L.; Rusznak, C.; Bagnall, A.; Sapsford, R. J.; Davies, R. J. (1999) Effect
26 of fluticasone propionate aqueous nasal spray on allergen-induced inflammatory changes
27 in the nasal airways of allergic rhinitics following exposure to nitrogen dioxide. Clin.
28 Exp. Allergy 29: 234-240.
March 2008 AX5-19 DRAFT-DO NOT QUOTE OR CITE
-------
i AX6. CHAPTER 6 ANNEX - EPIDEMIOLOGICAL
2 STUDIES OF HUMAN HEALTH EFFECTS ASSOCIATED
3 WITH AMBIENT OXIDES OF NITROGEN EXPOSURE
4
5
6 This annex provides supplemental information on various epidemiologic methods and
7 studies that are referenced in the oxides of nitrogen integrated science assessment. The first
8 section describes considerations in the interpretation of epidemiologic studies. This is followed
9 by a section on cardiovascular effects that are associated with short-term exposure to nitrogen
10 dioxide (NO2). This topic is discussed in the Integrated Science Assessment (ISA), but more
11 detail is provided in this annex due to inconsistency with supporting studies. The second section
12 of this annex presents tables detailing the epidemiologic studies presented in the ISA. In general,
13 these tables are divided into sections based on the endpoint of concern. Tables AX6.1 through
14 AX6.4 cover respiratory endpoints, while tables in section AX6.5 address cardiovascular disease.
15 The two tables in AX6.6 cover heart rate variability, section AX6.7 addresses birth weight, AX
16 6.7 looks at lung function, AX6.9 focuses on lung cancer, and lastly, table AX6.10 covers
17 mortality.
18
19
20 AX6.1 CONSIDERATIONS IN THE INTERPRETATION OF
21 EPIDEMIOLOGIC STUDIES OF OXIDES OF NITROGEN
22 HEALTH EFFECTS
23 Issues and questions arising from the study designs and analysis methods used in the
24 assessment of NC>2 effect estimates will be presented briefly in this section. Study design can
25 restrict the health effect parameters that can be estimated. Separate considerations need to be
26 made for acute versus chronic effect studies, as well as individual- versus aggregate-level
27 analyses. Time-series studies and panel studies are most frequently conducted in air pollution
28 epidemiologic research. Aggregate-level exposure data are often used in these types of studies.
29 Time series studies also use aggregate level health outcome data while panel studies collect
30 individual level data on health outcomes. Analyses using administrative health outcome data
31 (e.g., numbers of deaths and emergency hospital admissions) have inherent limitations as well as
32 strengths (Virnig and McBean, 2001). The impact of study design or the loss of information due
33 to aggregation depends on the source of exposure (Sheppard et al., 2005).
March 2008 AX6-1 DRAFT-DO NOT QUOTE OR CITE
-------
1 This section mainly focuses on the topics of exposure assessment and model specification
2 in air pollution epidemiologic studies. Potential biases that may result from NC>2 exposure
3 measurement error and from the choice of exposure index and lag period are discussed first.
4 Model specification issues and potential confounding by temporal factors, meteorological
5 effects, seasonal trends, and copollutants are then discussed.
6
7 AX6.1.1 Exposure Assessment and Measurement Error in Epidemiologic
8 Studies and Related Surrogate Discussion
9 In many air pollution epidemiologic studies, especially time-series studies with
10 administrative data on mortality and hospitalization outcomes, data from central ambient
11 monitoring sites generally are used as the estimate of exposure. Personal exposures of individual
12 study subjects generally are not directly measured in epidemiologic studies. The relation
13 between NO2 concentrations from ambient monitors and personal NC>2 exposures was discussed
14 previously (Chapter 2). Routinely collected ambient monitor data, though readily available and
15 convenient, may not represent true personal exposure, which includes both ambient and
16 nonambient (i.e., indoor) source exposures. Also, personal exposure measurements may or may
17 not be subject to the same artifacts as the ambient measurements. Therefore, they may not be
18 measuring the same quantities.
19 Zeka and Schwartz (2004) state that each pollutant, as measured at a central site in each
20 city, is a surrogate for exposure to the same pollutant. However, Sarnat et al. (2001) have
21 proposed that ambient concentration of gaseous air pollutants may be serving as a surrogate not
22 for exposure to the gas itself, but for exposure to ambient PM from sources where NO2 is
23 primarily a surrogate for particles from traffic. These data are specific to the data in the Sarnat
24 et al. (2001) study. Studies in other cities provide different results. In Boston, Sarnat et al.
25 (2005) noted seasonal differences in the relationship and stronger associations between ambient
26 NC>2 and personal NC>2. Another aspect is noted by Gilbert et al. (2005) who report regression
27 models including traffic and land-use variables to provide exposure estimates for epidemiologic
28 studies that they state may be more representative than the fixed monitors.
29 Studies evaluating exposure to NC>2 that recorded indoor environment measurements
30 report air exchange rates (AERs) of 0.49/h versus 0.85/h for heating versus non-heating season
31 noting significant univariate predictors of indoor concentrations to include outdoor NC>2 levels
32 and AERs. Indoor sources of NC>2 include gas cooking and heating. Outdoor NC>2 levels are
March 2008 AX6-2 DRAFT-DO NOT QUOTE OR CITE
-------
1 impacted by combustion sources and vehicular traffic emissions and related to distance to
2 major highways.
3 In considering exposure error, it should be noted that total personal exposure can be
4 partitioned into two types of sources, ambient and nonambient. Sheppard (2005) notes that
5 nonambient source exposures typically vary across individuals, but the community averages do
6 not vary across communities. In addition, nonambient exposures are not likely to have strong
7 temporal correlations. In contrast, ambient concentrations across individuals should be highly
8 correlated, as they tend to vary over time similarly for everyone because of changes in source
9 generation, weather, and season. The independence of ambient and non-ambient exposure
10 sources has important implications. Sheppard et al. (2005) observe that when ambient and
11 nonambient sources are independent, exposure variation due to nonambient source exposures
12 behaves like Berkson measurement error and does not bias the effect estimates.
13 A simulation study by Sheppard et al. (2005) also considered attenuation of the risk based
14 on personal behavior, their microenvironment, and the qualities of the pollutant in time-series
15 studies. Of particular interest is their finding that significant variation in nonambient exposure or
16 in ambient source exposure that is independent of ambient concentration does not further bias the
17 effect estimate. In other words, risk estimates were not further attenuated in time-series studies
18 even when the correlations between personal exposures and ambient concentrations were weak.
19 In the case of NO2, there are nonambient indoor sources; thus, the nonambient source
20 exposures may be independent from ambient source exposures depending on the exchange rate.
21 However, unlike PM, NC>2 is a reactive pollutant, while less so than Os. In applying these
22 conclusions to NC>2, an additional assumption needs to be made, i.e., that its chemical reactivity
23 does not introduce strong temporal correlations.
24 Other complications for NO2 in the relationship between personal exposures and ambient
25 concentrations include expected strong seasonal variation of personal behaviors and building
26 ventilation practices that can modify exposure. Also, there may be potential differential errors
27 based on different measurement techniques for ambient and personal measurement. In addition,
28 the relationship may be affected by temperature (e.g., high temperature may increase air
29 conditioning use, which may reduce NO2 penetration indoors), further complicating the role of
30 temperature as a confounder of NO2 health effects. It should be noted that the pattern of
31 exposure misclassification error and influence of confounders may differ across the outcomes of
March 2008 AX6-3 DRAFT-DO NOT QUOTE OR CITE
-------
1 interest as well as in susceptible populations and by study design. For example, those who may
2 be suffering from chronic cardiovascular or respiratory conditions may be in a more protective
3 environment (i.e., with less exposure to both NO2 and its confounders, such as temperature and
4 PM) than those who are healthy.
5 As discussed thoroughly in the 2004 PM AQCD (Section 8.4.5), the resulting exposure
6 measurement error and its effect on the estimates of relative risk must be considered. In theory,
7 there are three components to exposure measurement error in time-series studies as described by
8 Zeger et al. (2000): (1) the use of average population rather than individual exposure data;
9 (2) the difference between average personal ambient exposure and ambient concentrations at
10 central monitoring sites; and (3) the difference between true and measured ambient
11 concentrations. The first error component, having aggregate rather than individual exposure
12 data, is a Berksonian measurement error, which in a simple linear model increases the standard
13 error, but does not bias the risk estimate. The second error component resulting from the
14 difference between average personal ambient exposure and outdoor ambient concentration level
15 has the greatest potential to introduce bias. If the error is of a fixed amount (i.e., absolute
16 differences do not change with increasing concentrations), there is no bias. However, if the error
17 is not a fixed difference, this error will likely attenuate the NO2 risk estimate as personal NO2
18 exposures are generally lower than ambient NO2 concentrations in homes without sources, while
19 they are higher in homes with sources. The third error component, the instrument measurement
20 error in the ambient levels, is referred to as nondifferential measurement error and while unlikely
21 to cause substantial bias, can lead to a bias toward the null.
22 The impact of exposure measurement error on NO2 effect estimates was demonstrated in
23 a study by Kim et al. (2006) that is a longitudinal study investigating personal exposures to NO2,
24 PM2 5, and CO for cardiac compromised individuals in Toronto, Canada. The mean (SD)
25 personal exposure for NO2 was 14 ppb (6). NO2 personal exposures were less than central-fixed-
26 site ambient measurements. Ambient NO2 was correlated with the personal NO2 (median
27 Spearman's correlation coefficient of 0.57). Personal exposures to PM2.5 were correlated with
28 the personal exposure to NO2 (median Spearman's correlation coefficient of 0.43). This study
29 suggests that central-fixed-site measurements of PM2 5 and NO2 may be treated as surrogates for
30 both exposure to PM2 5 and NO2 in time-series epidemiology studies and that NO2 is a potential
31 confounder of PM2 5 and vice versa. Nerriere et al. (2005) proved more data from European
March 2008 AX6-4 DRAFT-DO NOT QUOTE OR CITE
-------
1 cities, noting some differences by city and impact of indoor sources and activity patterns and
2 recommend a site-specific analysis for a specific study.
3 Zidek (1997) noted that a statistical analysis must balance bias and imprecision (error
4 variance). Ignoring measurement error in air pollution epidemiologic studies often results in
5 underestimated risk estimates and standard errors.
6 In addition to overestimation of exposure and the resulting underestimation of effects, the
7 use of ambient NO2 concentrations may obscure the presence of thresholds in epidemiologic
8 studies at the population level. Using PM2.5 as an example, Brauer et al. (2002) examined the
9 relationship between ambient concentrations and mortality risk in a simulated population with
10 specified common individual threshold levels. They found that no population threshold was
11 detectable when a low threshold level was specified. Even at high-specified individual threshold
12 levels, the apparent threshold at the population level was much lower than specified. Brauer
13 et al. (2002) concluded that surrogate measures of exposure (i.e., those from centrally-located
14 ambient monitors) that were not highly correlated with personal exposures obscured the presence
15 of thresholds in epidemiologic studies at the population level, even if a common threshold exists
16 for individuals within the population.
17 As discussed in Chapter 3, NO2 concentrations measured at central ambient monitors
18 may explain, at least partially, the variance of individual personal exposures; however, this
19 relationship is influenced by factors such as air exchange rates in housing and time spent
20 outdoors, which may vary by city. Other studies conducted in various cities observed that the
21 daily averaged personal NO2 exposures from the population were well correlated with monitored
22 ambient NO2 concentrations, although substantial variability existed among the personal
23 measurements. Thus, there is supportive evidence that ambient NO2 concentrations from central
24 monitors may serve as valid surrogate measures for mean personal NO2 exposures experienced
25 by the population, which is of most relevance to time-series studies (See Chapter 3). Respiratory
26 hospital visit and admission studies are influenced by the visits and admission of asthmatics. In
27 children, for whom asthma is more prevalent, ambient monitors may correlate to some extent
28 with personal exposure to NO2 of ambient origin because children spend more time outdoors in
29 the warm season and have an increased potential for exposure due to traffic. However, of some
30 concern for mortality and hospitalization time-series studies is the extent to which ambient NO2
31 concentrations are representative of personal NO2 exposures in another particularly susceptible
March 2008 AX6-5 DRAFT-DO NOT QUOTE OR CITE
-------
1 group of individuals, the debilitated elderly, as the correlation between the two measurements
2 has not been examined in this population. A better understanding of the relationship between
3 ambient concentrations and personal exposures, as well as of the factors that affect the
4 relationship will improve the interpretation of ambient concentration-population health response
5 associations observed.
6 Existing epidemiologic models may not fully take into consideration all the biologically
7 relevant exposure history or reflect the complexities of all the underlying biological processes.
8 Using ambient concentrations to determine exposure may overestimate true personal NO2
9 exposures (depending on indoor sources), resulting in biased descriptions of underlying
10 concentration-response relationships (i.e., in attenuated risk estimates). The implication is that
11 the effects being estimated occur at exposures that are uncertain and the potency of NO2 is
12 different than these effect estimates indicate. As very few studies evaluating NO2 health effects
13 with personal NO2 exposure measurements exist in the literature, effect estimates determined
14 from ambient NO2 concentrations must be evaluated and used with caution to assess the health
15 risks of NO2. Ambient NO2 levels are regulated and can consist of exposure to NO2 in the
16 ambient air and NO2 exposure to NO2 of ambient origin in vehicles or indoors as opposed to
17 personal NO2 levels.
18 The question of what the NO2 measurements made at ambient monitoring sites represent
19 impacts the interpretation of epidemiology studies where the exposure estimate is derived from
20 such data. Time-series studies for hospitalization and mortality that show a relationship with
21 such measurements must be interpreted with this question in mind. For example, if the NO2
22 measurement is a surrogate for some other pollutant such as particles or, more generally, a
23 traffic-related mix, what interpretations are possible? Further, if not related to NO2 levels but for
24 some unmeasured mixture, how is this quantified? Additionally, the discussion and data related
25 to surrogates for NO2 in the literature are not quantitative or extensively researched, but do
26 provide a hypothesis for the relationships observed. Obviously, other data from clinical studies
27 and animal toxicology became important in interpreting the meaning of the relationship. Also,
28 epidemiology studies that use personal exposure measurements related to health outcomes
29 providing direct evidence of an association between exposure to NO2 and respiratory health
30 unconfounded by surrogate issues or copollutants are very informative in this evaluation.
March 2008 AX6-6 DRAFT-DO NOT QUOTE OR CITE
-------
1 The ultimate goal of the NO2 NAAQS is to set a standard for the ambient level, not
2 personal exposure level, of NC>2. Confidence in the use of ambient concentrations in
3 epidemiologic studies is greatly strengthened if they are shown to be associated with personal
4 exposures. However, until more data on personal NO2 exposure become available, the use of
5 routinely monitored ambient NC>2 concentrations as a surrogate for personal exposures is not
6 generally expected to change the principal conclusions from NC>2 epidemiologic studies. An
7 issue to be addressed is if estimates of NC>2 exposure are a surrogate for another pollutant and/or
8 a surrogate for NC>2. Evidence related to this will be presented in this chapter. More discussion
9 evaluating this will be presented in the ISA. Therefore, population health risk estimates derived
10 using ambient NC>2 levels from currently available observational studies (with appropriate
11 caveats taking into account personal exposure considerations) remain useful. These conclusions
12 must be evaluated to better determine associated uncertainties.
13
14 AX6.1.2 NOi Exposure Indices Used
15 The NO2 related effect estimates for mortality and morbidity health outcomes are usually
16 presented in this document as a relative risk, i.e., the risk rate relative to a baseline mortality or
17 morbidity rate. Relative risks are based on an incremental change in exposure. To enhance
18 comparability between studies, presenting these relative risks by a uniform exposure increment is
19 needed. However, determining a standard increment is complicated by the use of different NC>2
20 exposure indices in the existing health studies. The daily NC>2 exposure indices that most often
21 appear in the literature are the maximum 1-h average within a 24-h period (1-h max) and 24-h
22 average (24-h avg) concentrations. As levels are lower and less variable for the longer averaging
23 times, relative risks of adverse health outcomes for a specific numeric concentration range are
24 not directly comparable across metrics. Using the nationwide distributional data for NC>2
25 monitors in U.S. Metropolitan Statistical Areas, increments representative of a low-to-high
26 change in NC>2 concentrations were approximated on the basis of annual mean to 95th percentile
27 differences (Langstaff, 2006), as follows:
Daily Exposure Index
l-havgNO2 30
24-h avg NO2 20
2-wk avg NO2 2p_
Exposure Increment (ppb)
March 2008
AX6-7
DRAFT-DO NOT QUOTE OR CITE
-------
1 In the following chapter sections, efforts were made to standardize the NO2 risk estimates
2 using these increments, except as noted. The specified incremental change for each daily NO2
3 exposure index ensures that risk estimates are comparable across the different metrics. The
4 different increments for each NC>2 exposure index do not represent inconsistencies; rather, they
5 are appropriately scaled to facilitate comparisons between the various studies that used different
6 indices. Note that in the Chapter 6 Annex Tables (see Annex Section AX6.1), effect estimates
7 are not standardized; there, the results are presented in the tables as reported in the published
8 papers.
9
10 AX6.1.3 Lag Time: Period between NO2 Exposure and Observed Health
11 Effect
12 Exposure lags may reflect the distribution of effects across time in a population and the
13 potential mechanisms of effects. The choice of lag days for the relationship between exposure
14 and health effects depends on the hypothesis being tested and the mechanism involved in the
15 expression of the outcome. Effects can occur acutely with exposure on the same or previous day,
16 cumulatively over several days, or after a delayed period of a few days. With knowledge of the
17 mechanism of effect, the choice of lag days can be determined prior to analysis. As one
18 example, one could expect cough to occur acutely after exposure with a lag of 0 or 1 day, given
19 that NO2 can act as a short-term irritant. However, an NCVrelated inflammatory response may
20 not lead to asthma exacerbation until several days later. An asthmatic may be impacted by NO2
21 on the first day of exposure, have further effects triggered on the second day, and then report to
22 the emergency room for an asthmatic attack three days after exposure. Further, within a
23 population of asthmatics, exacerbation of asthma symptoms may be observed over a period of
24 several days, since each asthmatic may have varying individual aspects of the disease and may
25 be affected by the exposure differently depending on his/her sensitivity and disease severity.
26 The results from controlled human studies may be useful in assessing the adequacy of lags for
27 some respiratory health outcomes.
28 The concepts of lags are well discussed in the O3 AQCD (2006) and are only briefly
29 reviewed here, as the concept for Os pertains to NC>2 as well. Selection of lag periods should
30 depend on the hypothesis of the study and the potential mechanism of the effect. When the
31 mechanism of the health effect is unknown, investigating the association between outcome and
32 exposure using cumulative distributed lag models may be informative. Analyzing a large
March 2008 AX6-8 DRAFT-DO NOT QUOTE OR CITE
-------
1 number of lags and simply choosing the largest and most significant results may bias the air
2 pollution risk estimates away from the null. Most studies have shown that NC>2 has a fairly
3 consistent, immediate effect on health outcomes, including respiratory hospitalizations and
4 mortality. Several studies also observed significant NO2 effects over longer cumulative lag
5 periods, suggesting that in addition to single-day lags, multiday lags should be investigated to
6 fully capture a delayed NO2 effect on health outcomes. In this document, discussion largely
7 focuses on effect estimates from 0- and 1-day lags, with some consideration of cumulative,
8 multiday lag effects. It is not straightforward to compare and contrast results from single-day
9 versus multiday lag models, because the parameters estimated from these models are not the
10 same. These complications need to be taken into consideration when interpreting results from
11 various lag models.
12
13 AX6.1.4 Model Specification to Adjust for Temporal Trends and
14 Meteorological Effects
15 Several challenges are encountered with respect to designing and interpreting time-series
16 studies. The principal challenge facing the analyst in the daily time-series context is avoiding
17 bias due to confounding by short-term temporal factors operating over time scales from days to
18 seasons, thus adjusting for long-term trends in the evaluation of acute or short-term associations.
19 In the current regression models used to estimate short-term effects of air pollution, two major
20 potential confounders generally need to be considered: (1) seasonal trend and other "long-wave"
21 temporal trends; and (2) weather effects. Both of these variables tend to predict a significant
22 fraction of fluctuations in time-series.
23 Current weather models used in time-series analyses can be classified by their use of:
24 (1) quantile (e.g., quartile, quintile) indicators; (2) parametric functional forms such as V- or
25 U-shape functions; and (3) parametric (e.g., natural splines) or nonparametric (e.g., locally
26 estimated smoothing splines [LOESS]) smoothing functions. More recent studies tend to use
27 smoothing functions. While these methods provide flexible ways to fit health outcomes as a
28 function of temperature and other weather variables, there are two major issues that need further
29 examination to enable more meaningful interpretation of NC>2 morbidity and mortality effects.
30 The first issue is the interpretation of weather or temperature effects. Most researchers
31 agree about the morbidity and mortality effects of extreme temperatures (i.e., heat waves or cold
32 spells). However, as extreme hot or cold temperatures, by definition, happen rarely, much of the
March 2008 AX6-9 DRAFT-DO NOT QUOTE OR CITE
-------
1 health effects occur in the mild or moderate temperature range. Given the significant correlation
2 between NO2 and temperature, ascribing the association between temperature and health
3 outcomes solely to temperature effects may underestimate the effect of NO2. The second issue is
4 that weather model specifications are fitted for year-round data in most studies. Such models
5 will ignore the correlation structure that can change across seasons, resulting in inefficiency and
6 model mis-specification. This is particularly important for NC>2, which appears to change its
7 relationship with temperature as well as with other pollutants across seasons.
8 This changing relationship between NC>2 and temperature, as well as between NO2 and
9 other pollutants across seasons, and its potential implications for health effects modeling have
10 not been examined thoroughly in the time-series literature. Even the flexible smoother-based
11 adjustments for seasonal and other time-varying variables cannot fully take into account these
12 complex relationships. One obvious way to alleviate or avoid this complication is to analyze
13 data by season. While this practice reduces sample size, its extent would not be as serious as for
14 PM (which is collected only every sixth day in most locations) because NC>2 is collected daily.
15 An alternative approach is to include separate NC>2 concentration variables for each season (by
16 multiplying NC>2 concentrations by a season indicator variable). However, this approach
17 assumes that all effects in the model that are not indicated to be season-specific do not vary
18 seasonally.
19 In locations where seasonal variability may be a factor, NC>2 effect estimates calculated
20 using year-round data can be misleading, as the changing relationship between NO2,
21 temperature, and other pollutants across seasons may have a significant influence on the
22 estimates. Analyses have indicated that confounding from seasonal variability may be controlled
23 effectively by stratifying the data by season.
24
25 AX6.1.5 Confounding Effects of Copollutants
26 Extensive discussion of issues related to confounding effects among air pollutants in
27 time-series studies are provided in Section 8.4.3 of the 2004 PM AQCD (U.S. Environmental
28 Protection Agency, 2004). Since the general issues discussed there are applicable to all
29 pollutants, such discussions are not repeated here.
March 2008 AX6-10 DRAFT-DO NOT QUOTE OR CITE
-------
1 AX6.1.6 Generalized Estimating Equations (GEE)
2 Since the publication of the 1993 NC>2 AQCD (U.S. Environmental Protection Agency,
3 1993), methods have been improved to analyze panel and longitudinal studies. The general
4 mixed model method of Stiratelli et al. (1984) was an improvement over the method of Korn and
5 Whittemore (1979) in that all the data could be used, including that from subjects with
6 insufficient data to permit fitting of a separate logistical regression model. Generalized
7 Estimating Equations, Liang and Zeger (1986) is an extension of generalized linear models. The
8 joint distribution of the subject's observations does not have to be specified to derive the
9 estimating equations. This is avoided by assuming a marginal distribution at each time.
10 However, a covariate that is constant for a subject cannot be included in this model. Besides
11 Gaussian outcome variables, the method can also be used for binomial or Poisson variables.
12
13 AX6.1.7 Hypothesis Testing and Model Selection in NO2 Epidemiologic
14 Studies
15 Epidemiologic studies investigated the association between various measures of NO2
16 (e.g., multiple lags, different metrics, etc.) and various health outcomes using different model
17 specifications. Statistically testing a null hypothesis (i.e., there is no effect of NCh) requires one
18 to calculate the value of a test statistic (i.e., the t-value). If the observed test statistic exceeds a
19 critical value (oftentimes the 95th percentile) or is outside a range of values, the null hypothesis
20 is rejected. However, when multiple testing is done using a critical value determined for a single
21 test, the chance that at least one of the hypotheses is significant may be greater than the specified
22 error rate. Procedures are available to ensure that the rejection error rate does not exceed the
23 expected error rate (usually 5%) when conducting multiple hypothesis testing. However, often
24 the multiple hypotheses being tested are not statistically independent, thus some corrections,
25 such as the Bonferroni adjustment, may be overly conservative.
26 Multiple hypothesis testing and model selection also contribute to publication bias.
27 Publication bias is the tendency of investigators to submit and/or editors to accept manuscripts
28 for publication based on the strength of the study findings. Although publication bias commonly
29 exists for many topics of research, it may be present to a lesser degree in the air pollution
30 literature. Several air pollutants often are examined in a single study, leading to the publication
31 of significant, as well as nonsignificant, individual pollutant results. For example, many air
32 pollution papers with a focus on PM health effects also published NO2 results. NO2 was largely
March 2008 AX6-11 DRAFT-DO NOT QUOTE OR CITE
-------
1 considered a potentially confounding copollutant of PM; thus, NC>2 effect estimates were often
2 presented regardless of the statistical significance of the results. Another aspect of publication
3 bias is only selecting the largest or statistically strongest effect estimate to report and not the
4 array of models evaluated. See a full discussion in the Os AQCD (U.S. Environmental
5 Protection Agency, 2006).
6 The summary of health effects in this chapter is vulnerable to the errors of publication
7 bias and multiple testing. Efforts have been made to reduce the impact of multiple testing errors
8 on the conclusions in this document. To address multiple hypothesis testing, emphasis will be
9 placed in this chapter on a priori hypotheses. As identifying a priori hypotheses is difficult in the
10 majority of the studies, the most common hypotheses will be considered. For example, although
11 many studies examined multiple single-day lag models, priority would be given to effects
12 observed at 0- or 1-day lags rather than at longer lags. Both single- and multiple-pollutant
13 models that include NC>2 will be considered and examined for robustness of results. Analyses of
14 multiple model specifications for adjustment of temporal or meteorological trends will be
15 considered sensitivity analyses. Sensitivity analyses shall not be granted the same inferential
16 weight as the original hypothesis-driven analysis; however, these analyses will be discussed in
17 this chapter as appropriate given their valuable insights that may lead scientific knowledge in
18 new directions.
19
20 AX6.1.8 Impact of Generalized Additive Models Convergence Issue on NO2
21 Risk Estimates
22 Generalized Additive Models (GAM) have been widely utilized for epidemiologic
23 analysis of the health effects attributable to air pollution. The impact of the GAM convergence
24 issue was thoroughly discussed in Section 8.4.2 of the 2004 PM AQCD. Reports have indicated
25 that using the default convergence criteria in the Splus software package for the GAM function
26 can lead to biased regression estimates for PM and an underestimation of the standard error of
27 the effect estimate (Dominici et al., 2002; Ramsay et al., 2003). The GAM default convergence
28 criterion in the Splus software package is 10 and a maximum number of 10 iterations. The user
29 can specify convergence criteria, that is orders of magnitude smaller than the default value and
30 can also allow for many more iterations before terminating the program. The use of the default
31 convergence criterion was found to be a problem when the estimated relative risks were small
32 and two or more nonparametric smoothing curves were included in the GAM (Dominici et al.,
March 2008 AX6-12 DRAFT-DO NOT QUOTE OR CITE
-------
1 2002). The magnitude and direction of the bias depend in part on the concurvity of the
2 independent variables in the GAM and the magnitude of the risk estimate. Recent focus has
3 been on the influence of the GAM function on effect estimates for PM.
4 The GAM convergence problem appears to vary depending on data sets, and likely
5 depends upon the intercorrelation among covariates and the magnitude of the risk estimate; thus,
6 its impact on the results of individual studies cannot be known without a reanalysis. Consistent
7 with the approach used in the 2004 PM AQCD, the results from studies that analyzed data using
8 GAM with the default convergence criterion and at least two nonparametric smoothing terms are
9 generally not considered in this chapter, with some exceptions as noted.
10
11
12 AX6.2 CARDIOVASCULAR EFFECTS ASSOCIATED WITH SHORT-
13 TERM NO2 EXPOSURE
14
15 AX6.2.1 Studies Hospital Admissions and ED Visits for Cardiovascular
16 Disease (CVD)
17
18 AX6.2.1.1 All CVD (ICD9 390-459)
19 Results of studies of short-term NC>2 exposure and hospitalization or ED visits for CVD
20 are summarized in Figure AX6.2-1. Studies of both 1-h maximum NC>2 level and 24-h average
21 NC>2 level are included. With the exception of lag 1 results reported by Jalaludin et al. 2006,
22 most point estimates are positive with confidence limits excluding the null value. Jalaludin et al.
23 report a lag 0 cool season relative risk (not pictured) of 1.09 (95%CI: 1.05, 1.13) per 30 ppb
24 increase in 1 hour maximum NC>2 level (Jalaludin et al., 2006). (Note that the IQR reported by
25 Jalaludin et al. is 9.3 ppb so the 30-ppp increase into which the results were standardized may be
26 unlikely in Sydney, where the study was conducted.) Although results for cardiovascular
27 diseases were not tabulated for a reanalysis of GAM impacted study of Los Angeles and Cook
28 County hospital admissions, authors note that they observed an association of NC>2 with CVD
29 hospital admissions in the reanalyses (Moolgavkar, 2003. The association was diminished with
30 the use of increasingly stringent convergence criteria, however (Moolgavkar, 2003).
March 2008 AX6-13 DRAFT-DO NOT QUOTE OR CITE
-------
Reference
Linn et al. (2000)
Metzger et al. (2004)*
Tolbertetal.(2007)*
Baliesteretal. (2006)
Andersen et al. (2007a)
Andersen et al. (2007b)
Atkinson etal. (1999a)*
Atkinson et a!. (1999a)*
Baliesteretal. (2001)*
Poloniecki etal. (1997)
Barnett etal. (2006)
Barnett et al. (2006)
Hinwood et al. (2006)
Hinwood et al. (2006)
Jalaludinetal. (2006)*
Jalaludin et al. (2006)*
Jalaludin et al (2006)*
Chang et al. (2005)
Chang et al. (2005)
Wong etal. (1999)
Wong etal. (1999)
Yang et al. (2004a)
Yang et al. (2004a)
Relative risk
Figure AX6.2-1. Relative risks (95% CI) for associations of 24-h NO2 (per 20 ppb) and
daily 1 hour maximum NOi* with hospitalizations or emergency
department visits for all cardiovascular diseases (CVD). Primary
author and year of publication, city, stratification variable(s), and lag
are listed. Results for lags 0 or 1 are presented, as available.
Location Season
Metro LA
Atlanta
Atlanta
Spain
Copenhagen
Copenhagen
London
London
Valencia
London
Australia Multicity
Australia Multicity
Perth
Perth
Sydney
Sydney
Sydney
Taipei Warm
Taipei Cold
Hong Kong
Hong Kong
Kaohsiung Warm
Kaohsiung Cool
Age Lag
All ages 0
All ages 3d av .
All ages 0-2
All ages 0-1
65 + 0-3
65 + 0-3 .
0-64 0
65+ 0
All ages 0
All ages 1
65+ 0-1
15-64 0-1
65+ 1
All ages 1
65+ 0
65+ 1
65+ 0-1
All ages 0-2
All ages 0-2
5-64 0-1
65+ 0-1 .
All ages 0-2
All ages 0-2
1
1
ff
1
4L
4-
|
_^_
1
ll
Hh
^_
_j_
._,_
_!_
,
_l_
_^
_,_
_|_
1 1 1 1 1 1 1
.9 1.1 1.3 1.5 1.7 1.9 2.1 2.32.52.72
1 AX6.2.1.2 Ischemic Heart Disease (IHD) ICD9 410-414
2 Studies that further narrow the cardiac disease grouping to evaluate Ischemic Heart
3 Diseases (IHD) are summarized in Figure AX6.2-2. Several US studies examined the
4 association of ambient NO2 level with IHD (Ito, 2004; Mann et al. 2002; Metzger et al. 2004;
March 2008
AX6-14
DRAFT-DO NOT QUOTE OR CITE
-------
Reference
Lippmannet al. (2000)
!to (2004)
Mann et al. (2002)
Mann et al. (2002)
Mann et al. (2002)
Mann et al. (2002)
Metzgeretai. (2004)*
Peel et al. (2007)*
Peel et al. (2007)*
Atkinson etal.(1999a)*
Atkinson etal. (1999a)*
Poloniecki et al. (1997)
Barnett et a!. (2006)
Barnett et al. (2006)
Simpson et al. (2005a,b)*
Simpson et al. (2005a,b)*
Jalaludin et al. (2006)*
Jalaludinetal. (2006)*
Jalaludin etal. (2006)*
Wong etal. (1999)
Location
Ontario
Ontario
South Coast Air Basin
South Coast Air Basin
South Coast Air Basin
South Coast Air Basin
Atlanta
Atlanta
Atlanta
London
London
London
Australia, NZ
Australia, NZ
Australia, Multicity
Australia, Multicity
Sydney
Sydney
Sydney
Hong Kong
Disease Age
All Ages
All Ages
All Ages
All Ages
Arrhythmia All Ages
CHF All Ages
No Secondary All Ages
AIIIHD All Ages
All Ages
Hypertension 0-64
Diabetes 65+
All Ages
65+
15-64
All Ages
65+
65+
65+
65+
All Ages
Lag
1.01
0.98 .
1.02
1.02 .
1.03 .
1.02
1.03 .
1.04
1.00 .
1.02 .
1.03 .
1.00
1.03 .
1.01 .
1.00 .
1.00 .
1.02
1.01 .
1.02 .
1.01 .
H
-1-
i
tt
+
1
I—
I I
9 1.1
Figure AX6.2-2.
Relative risk
Relative risks (95% CI) for associations of 24-h NOi (per 20 ppb) and
daily 1 hour maximum NO2* (per 30 ppb) with hospitalizations for
Ischemic Heart Disease (IHD). Primary author and year of
publication, city, stratification variable(s), and lag are listed. Results
for lags 0 or 1 are presented, as available.
4
5
6
Peel et al. 2007). Ito (2004) reports a null association of 24-h average NO2 level with IHD
admissions in Ontario where the mean ambient level was 21.3 ppb. Mann et al. (2002) examined
the association of 24-h average NO2 level with IHD and secondary diagnoses of arrhythmia, IHD
and secondary diagnosis of CHF, IHD and no secondary diagnosis, and all IHD regardless of
secondary diagnosis in single pollutant models. The authors note that the strongest effect
observed (IHD with secondary diagnosis of CHF) may have been driven by the MI primary
March 2008
AX6-15
DRAFT-DO NOT QUOTE OR CITE
-------
1 diagnoses. The 24-h average NO2 level in the South Coast Air Basin of California, where this
2 study was conducted was approximately 37 ppb. This study was novel in that exposure level
3 was assigned based on the zip code of the health insurance participant and proximity to the
4 monitoring station (Mann et al., 2002). A non-significant increased risk of ED visit for IHD was
5 observed in single pollutant models, among those with hypertension but not diabetes in a study
6 conducted in Atlanta where the daily 1-h maximum NC>2 level is approximately 46 ppb (Peel
7 et al., 2007).
8 Two studies of IHD and hospital admissions conducted in Europe have produced
9 conflicting results (Atkinson et al., 1999a; Poloniecki et al., 1997). Atkinson et al. (1999a)
10 reports a significant increase in IHD admission in a study in London. A study conducted in
11 Helsinki reports an association of NO with both hospitalization and ED visits for IHD while no
12 association with NC>2 was observed (Ponka and Virtanen, 1996). In addition, Several Australian
13 studies, including two multicity studies, support an association of hospital admissions and
14 emergency visits for IHD and ambient NC>2 level among older adults in single pollutant models
15 (Jalaludin et al., 2006; Barnett et al., 2006; Simpson et al., 2005a,b). One study conducted in
16 Hong Kong reports slightly elevated non-significant association of IHD with 24-h average NO2
17 level (Wong et al., 1999). In addition, Lee et al. (2003a) report an increase in IHD admissions
18 associated with 24-h NO2 level at lag 5.
19
20 AX6.2.1.3 Hospital Admissions for Myocardial Infarction (MI) (ICD9 410)
21 Studies of hospital admissions for MI are summarized in Figure AX6.2-3. In the United
22 States, positive single pollutant associations for emergency admissions for MI and increases in
23 ambient NO2 level were reported in Boston (Zanobetti and Schwartz, 2006) and California (Linn
24 et al., 2000; Mann et al., 2002).
25 Pooled results from two European multicity studies are inconsistent. Von Klot et al.
26 (2005) report an increase in MI readmissions at lag 0 while Lanki et al. (2006) report a null
27 effect at lag 1. The NO2 levels were similar in the cities studied with Lanki et al. (2006)
28 reporting the 24 h average level of 23 ppb and Von Klot et al. (2005) reporting a range in 24-h
29 average across the cities studied of 12-37 ppb. A single-city study in Italy (D'Ippoliti et al.,
30 2003) found positive significant associations between 24-h average NO2 level and admission for
31 the first episode of MI. The 24-h average NO2 level reported by D'Ippoliti was approximately
32 45 ppb. A study conducted in London reports a positive association of ED visit for MI with 24-h
March 2008 AX6-16 DRAFT-DO NOT QUOTE OR CITE
-------
Reference
Linn et al. (2000)
Mann et al. (2002)
Location
Metropolitan LA All ages
So Coast Air Basin All Ages
Zanobetti and Schwartz (2006) Boston
Zanobetti and Schwartz (2006) Boston
Lanki et al. (2006)
Von Klot et al. (2005)
D'lppoliti et al. (2003)
D'lppoliti et al. (2003)
Barnettetal. (2006)
Barnettetal.(2006)
Poloniecki et al. (1997)
Figure AX6.2-3.
Europe
Europe
Italy
Italy
Australia, NZ
Australia, NZ
London
65+
65+
65+
je
ages
Ages
+
+
ages
Survivors 35+
ages
ages
+
-64
ages
Lis
0 .
0
0
0-1
0
n
0
0-2
0-1
0-1
1 .
-H
_|_
fi
I I I
.9 1.1 1.3
Relative risk
Relative risks (95% CI) for associations between 24-h NOi (per 20
ppb) and hospitalizations for myocardial infarction (MI). Primary
author and year of publication, city, stratification variable(s), and lag
are listed. Results for lags 0 or 1 are presented as available.
4
5
6
1
8
9
10
average NC>2 where the 24-h average reported was approximately 35 ppb. Finally, positive
associations are reported for MI in a multicity study conducted in Australia and New Zealand
among older adults and adults from 15 to 64 years old (Barnett et al., 2006). The 24-h average
NC>2 level ranged from approximately 7-12 ppb in the Australian cities studied (Barnett et al.,
2006).
AX6.2.1.4 Arrhythmia (ICD9 427) and Congestive Heart Failure (CHF) (ICD9 428)
Hospital or ED admissions for arrhythmia were inconsistently associated with increases
in ambient NC>2 level. Some studies report positive associations (Rich et al., 2006a; Mann et al.,
2002; Barnett et al., 2006) while others report null associations (Metzger et al., 2004; Lippmann
March 2008
AX6-17
DRAFT-DO NOT QUOTE OR CITE
-------
1 et al., 2000; reanalysis Ito, 2003, 2004). Single pollutant models of hospital admissions and ED
2 visits for CHF have also produced mixed results (Figure AX6.2-4). A seven city study
Reference Location
Morris et al. (1995) US, multeity
Morris et al. (1995) US, multeity
Morris et al. (1995) US, multeity
Morris et al. (1995) US, multeity
Morris et al. (1995) US, multeity
Morris et al. (1995) US, multeity
Morris etal. (1995) US, multeity
Linn et al. (2000) Metro LA
Ito etal. (2004) Ontario
Ito et al. (2004) Ontario
Metzger etal. (2004) Atlanta
Peel et al. (2007) Atlanta
Peel et al. (2007) Atlanta
Wellenius e! al. (2005a) Pittsburg
Poloniecki et al. (1997) London
Barnett et a!. (2006) Australia, HZ
Barnett et al. (2006) Australia, NZ
Wong et al . ( 1 999) Hong Kong
Other Age Lag
LA 65+0
Chicago 65+ 0
Philadelphia 65+ 0
New York 65+ 0
Detroit 65+ 0
Houston 65+ 0
Milwaukee 65+ 0
All Ages 0
All Ages 0
All Ages 1
All Ages 3 d moving .
Hypertension All Ages 0-2
Diabetes All Ages 0-2
65+ 0
All Ages 1
65+ 0-1
15-64 0-1
All Ages 0-3
-
}
-1—
Ir-
-i —
-I —
+
L
1
-t-
—
I I I I I I I
.7 .9 1.1 1.3 1.5 1.7 1.9
Relative risk
Figure AX6.2-4.
Relative risks (95% CI) for associations of 24-h NO2 (per 20 ppb) and
1-hour maximum NOi* with hospitalizations for congestive heart
failure (CHF). Primary author and year of publication, city,
stratification variable(s), and lag are listed. Results for lags 0 or 1 are
presented as available.
March 2008
AX6-18
DRAFT-DO NOT QUOTE OR CITE
-------
1 conducted in the U.S. among the elderly found positive associations in Los Angeles (RR: =1.52
2 [1.35, 1.71]), Chicago (RR: = 1.60 [1.24, 2.07]) and New York (RR: = 1.23 [1.05, 1.43]) per
3 30-ppb increase in NO2 (Morris et al., 1995). Estimates were close to the null value in
4 Philadelphia, Detroit, Houston, and Milwaukee and only the estimate for New York remained
5 significant in multi-pollutant models (Morris et al., 1995). The 1 h maximum NO2 level in the
6 cities studied ranged from 40 ppb in Milwaukee to 77 ppb in Los Angeles (Morris et al. 1995).
7 Ito et al. 2004 report null associations for CHF and NO2 in Ontario where the 24 h average NO2
8 level is approximately 21 ppb (Ito et al., 2004). Elevated but non-significant associations were
9 reported in Atlanta (Metzger et al. 2004; Peel et al. 2007) and elevated significant associations
10 were reported in Pittsburgh (Wellenius et al., 2005a). Null associations were reported in London
11 (Poloniecki et al., 1997) while positive significant associations were reported in a multicity study
12 in Australia and New Zealand (Barnett et al. 2006) and in Hong Kong (Wong et al., 1999).
13
14 Cerebrovascular Disease (ICD9 430-448)
15
16 AX6.2.1.5 Vaso-occlusion in Sickle Cell
17 A recent study evaluated the association of pain in Sickle Cell patients, which is thought
18 to be caused by vaso-occlusion, with air pollution (Yallop et al., 2007). A time series analysis
19 was performed to link daily hospital admissions for acute pain among sickle cell patients with
20 daily air pollution levels in London using the cross correlation function. No association was
21 reported for NO2. However, Yallop et al. observed an association (CCF = -0.063, lag 0) for NO,
22 CO, and O3.
23
24 AX6.2.1.6 Multipollutant Modeling Results
25 As noted in Annex ISA 3B, multipollutant models may have limited utility to distinguish
26 the independent effects of specific pollutants if model assumptions are not met. However, these
27 models are widely used in air pollution research and results for CVD hospital admissions and ED
28 visits are summarized in Figure AX6.2-5. This figure includes only those studies that present
29 two pollutant results in tabular form. Studies with qualitative descriptions or figures
30 summarizing two pollutant results are discussed in the text that follows (Linn et al., 2000; Mann
31 et al., 2002; Metzger et al., 2004; Tolbert et al., 2007; Zanobetti and Schwartz, 2006; Jalaludin
32 et al., 2006; Von Klot et al., 2005; Ballester et al., 2006; Wong et al., 1999). In addition, we
March 2008 AX6-19 DRAFT-DO NOT QUOTE OR CITE
-------
Relative risk
Reference Location Season Age
Yang et al,
(2004a)
Kaohsiung, Warm All
Taiwan
Lag Polluants °-5
0-2 NO,
Cool
N02PM10
N02S010
N02CO
N0203
N02
N02 PM!0
N02 S02
N02CO
N0203
Andersen et al.
(2007a)
Andersen et al.
(2007a)
Simpson et ai.
(2005a,b)
Barnett et al.
(2006)*
Chang et al.
(2005)
Copenhagen
Copenhagen
4 Cities, All
Australia year
7 Cities,
Australia
andNZ
Taipei, Warm
Taiwan
65+
65+
All
65+
0-3 NO,
N02PM
0-3 NO,
N0
N02 BSP
N0203
0-1 NO,
NOZCO
0-2 NO,
Poloniecki et al. London
(1997)
Cool
N02PM10
N02 S02
N02CO
N02
N02 PM10
N02 S02
N02CO
N020j
Wellenius et al. Allegheny All
(2005a) Co PA year
65+ 0 NO
N02PM,0
N02CO
N0203
N02 S02
Cool
All 0-1 NO,
N02 S02
N02CO
N02BS
N0203
1.0 1.5 2.0 2.5 3.0 3.5
I i i i
Cardiovascular Disease
Congestive heart failure
Figure AX6.2-5. Relative risks (95% CI) for associations of 24-h NOi exposure (per 20
ppb) and daily 1-hour maximum NOi* (per 30 ppb) with
hospitalizations or emergency department visits for CVD: Studies
with 2 pollutant model results. Primary author and year of
publication, city, stratification variable(s), and lag are listed. Results
for lags 0 or 1 are presented as available.
March 2008
AX6-20
DRAFT-DO NOT QUOTE OR CITE
-------
1 include text discussion of studies that simultaneous adjust for several pollutants (Morris et al.,
2 1995; Llorca et al., 2005) and several cerebrovascular disease studies that report multipollutant
3 results (Ballester et al., 2001; Villeneuve et al., 2006; Tsai et al., 2003a; Chan et al., 2006).
4 A U.S. study shows a diminishment of the relative risk for CHF in two pollutant models
5 (Wellenius et al., 2005a). Morris et al. (1995) also observes a similar diminishment of the CHF
6 association in multipollutant models containing SO2, CO and Ozone (Morris et al., 1995). The
7 association of cardiac disease admissions with NO2 in several non-U.S. studies was not robust in
8 two pollutant models (Simpson et al., 2005a,b; Poloniecki et al., 1997; Barnett et al., 2006;
9 Ballester et al., 2006). Llorca et al. (2005) reports a similar lack of robustness in models
10 containing NO2, TSP, H2S, NO and SO2. Estimates from studies conducted in Taiwan reporting
11 relatively high associations of NO2 with CVD in single pollutant models remained robust in 2-
12 pollutant models during the cool (Yang et al., 2004a) or warm (Chang et al., 2005) seasons only.
13 In an Australian study of the older adults (65+ years), the effect estimate for NO2 was robust to
14 simultaneous adjustment for Os and particles (Morgan et al., 1998a).
15 In two additional U.S. studies, investigators provide text descriptions of multi-pollutant
16 model results and indicate that their analyses were unable to distinguish the effects of NO2 from
17 PM, CO and other traffic pollutants (Linn et al. 2000; Mann et al. 2002.) In both studies, CO
18 was more highly correlated with NO2 than PM. In a Canadian study in which default GAM
19 procedures were used, the significant association of NO2 with ED visits for cardiac disease was
20 reduced and non-significant in multipollutant models (Stieb et al., 2000). Further, in a study of
21 emergency department visits to Atlanta hospitals, Metzger et al. 2004, who present results via
22 figure, observed a diminishment of the effect of NO2 on visits for cardiovascular disease when
23 CO was modeled with NO2, while the effect of CO remained robust (Metzger et al., 2004). This
24 finding was repeated in an analysis that included several additional years of data (Tolbert et al.,
25 2007). In this paper, Tolbert et al. discuss the limitations of multipollutant models in detail and
26 conclude that these models may help researchers identify the strongest predictor of disease but
27 may not isolate the independent effect of each pollutant (Tolbert et al., 2007). In an Australia
28 study (Jalaludin et al., 2006) and a Spanish multicity study (Ballester et al., 2006) presenting
29 multipollutant results in figures, the association of NO2 with cardiac disease was not robust to
30 adjustment for other pollutants (CO, SO2, particles). However, in a European multicity study
March 2008 AX6-21 DRAFT-DO NOT QUOTE OR CITE
-------
1 investigators report that the effect of NO2 on cardiac readmissions among MI survivors was not
2 diminished in multipollutant models (Von Klot et al., 2005).
3 Burnett et al. (1997a) reported robust estimates for cardiac disease hospital admissions
4 and NO2, whereas the observed association for cardiac hospitalizations and PM were explained
5 by gaseous pollutants. In another multicity study conducted in the same area, associations of
6 NC>2 with cardiac disease were not attenuated when CO, SO2, and PM variables were included in
7 the models (Burnett et al., 1999).
8 The association of NC>2 with stroke was not robust to adjustment for CO in a Canadian
9 study (Villeneuve et al., 2006). The association of NO2 with all cerebrovascular disease was not
10 robust to adjustment for BS and SO2 in a Spanish single city study (Ballester et al., 2001).
11 Although results from a Taiwanese study indicate the effect of NO2 on stroke admissions is not
12 diminished in 2 pollutant models, the authors note that the association of NO2 with stroke may
13 not be causal if NO2 is a surrogate other components of the air pollution mixture
14 (Tsai et al., 2003a).
15 Studies using alternative methods to investigate the influence of co-pollutants on
16 observed associations of NO2 with cardiovascular disease are few in number. In an study of
17 emergency admissions for MI and ambient pollution in Boston investigators attempt to
18 distinguish traffic from non-traffic related pollutants through their definition of an exposure
19 metric for non-traffic PM (residuals in model of PM2.s regressed against BC) but found NO2,
20 PM2.s and non-traffic PM each may trigger MI during the warm season (Zanobetti and Schwartz,
21 2006). In a study conducted in Hong Kong, investigators looked at the association of NO2 with
22 CVD during high PMio and high ozone days (Wong et al., 1999). An interaction between NO2
23 and Ozone was observed (in the single pollutant model NO2 associated with heart failure,
24 RR: 1.1895%CI: 1.10, 1.26 per 20 ppb, lag 0-3).
25
26 AX6.2.2 Heart Rate Variability, Repolarization, Arrhythmia, and Other
27 Measures Cardiovascular Function Associated with Short-Term
28 NO2 Exposure
29
30 AX6.2.2.1 Heart Rate Variability
31 Liao et al. (2004) investigated short-term associations between ambient pollutants and
32 cardiac autonomic control from the fourth cohort examination (1996 to 1998) of the population-
March 2008 AX6-22 DRAFT-DO NOT QUOTE OR CITE
-------
1 based Atherosclerosis Risk in Communities (ARIC) Study. PMio, NO2, and other gaseous air
2 pollutants were examined in this study. PMio (24-h average) and NO2 (24-h average) 1 day prior
3 to the randomly allocated examination date were used. The mean (SD) NO2 level was 21 (8)
4 ppb. PMio concentrations measured 1 day prior to the HRV measurements were inversely
5 associated with both frequency- and time-domain HRV indices. Ambient NO2 concentrations
6 were inversely associated with high-frequency power and SDNN. In single-pollutant models, a
7 20-ppb increase in ambient NO2 was associated with a 5% reduction (95% CI: 0.7, 9.2), in mean
8 SDNN. Consistently more pronounced associations were suggested between PMio and HRV
9 among persons with a history of hypertension.
10 Various measures of HRV have been examined in relation to daily levels of ambient air
11 pollution in other studies (Chan et al., 2005; Wheeler et al., 2006; Holguin et al., 2003;
12 Luttmann-Gibson et al., 2006; Schwartz et al., 2005). Chan et al. (2005) recruited 83 patients
13 from the cardiology section of a hospital in Taiwan. Patients included 39 with coronary heart
14 disease (CHD) and 44 with more than one risk factor for CHD. The authors reported finding
15 significant associations between increases in NO2 and decline in SDNN (NO2 lagged 4 to 8 h)
16 and LF (NO2 lagged 5 or 7 h) (see Annex Table AX6.5.1 for quantitative results). There were no
17 significant associations for r-MSSD or HF and NO2. None of the other pollutants tested (PMio,
18 CO, SO2, 63) were significantly associated with any of the HRV measured. Wheeler et al.
19 (2006) examined HRV and ambient air pollution in Atlanta in 12 patients who had an MI from
20 3 to 12 months prior to enrollment and 18 COPD patients. The results in the two patient groups
21 were quite different: increasing concentration of NO2 in the previous 4-h significantly reduced
22 SDNN in MI patients and significantly increased SDNN in COPD patients (see Annex Table
23 AX6.10). Similar significant associations were seen with increases in 4-h ambient PM2.s. The
24 PM2 5 concentrations were moderately correlated with NO2 levels (r = 0.4).
25 In contrast, Holguin et al. (2003) found PM2.5 concentrations were moderately correlated
26 with NO2 levels (v = 0.04) in 34 elderly adults in Mexico City and found no significant
27 associations with increases in NO2, but did find significant effects of PM2.5 on HF, particularly
28 among hypertensive subjects. Similarly, Luttmann-Gibson et al. (2006) also found significant
29 effects of PM2.s and SO4 on HRV measured in a panel of 32 senior adults in Steubenville, OH,
30 but observed no effect of increasing NO2. Likewise, Schwartz et al. (2005) found significant
31 effects of increases in PM2 5 on measures of HRV, while no associations with NO2 were
March 2008 AX6-23 DRAFT-DO NOT QUOTE OR CITE
-------
1 observed. A population-based study of air pollutants and HRV was conducted in Boston, MA on
2 497 men from the VA Normative Aging Study (NAS) (Park et al., 2005). The mean (SD) 24-h
3 average NO2 concentration was 22.7 (6.2) ppb. Associations with HRV outcomes were observed
4 with a 4-h moving average of Os and PM2.5 concentrations, but not with NC>2.
5
6 AX6.2.2.2 Repolarization Changes
7 A prospective panel study, conducted in East Germany, analyzed 12 repeat ECG
8 recordings for 56 males with IHD (Henneberger et al., 2005). Ambient air pollutants measured
9 at fixed monitoring sites were used to assign individual exposures for 0 to 5, 5 to 11, 12 to 17, 18
10 to 23, 0 to 23 h and for 2 to 5 days prior to the EEG. Pollutants considered were ultrafme
11 particles (UFP), accumulation mode particle (ACP), PM2.5, elemental carbon (EC), organic
12 carbon (OC), 862, NC>2, NO, and CO. Associations were observed between (1) QT duration and
13 EC and OC; (2) T-wave amplitude and UFP, ACP and PM2.s; and (3) T-wave complexity and
14 PMio, EC, and OC. NO (r = 0.83) and NO2 (0.76) were highly correlated with UFP but were not
15 associated with repolarization abnormalities.
16
17 AX6.2.2.3 Arrhythmias Recorded on Implanted Defibrillators
18 In a pilot study, Peters et al. (2000a) abstracted device records for 3 years for each of 100
19 patients with ICDs. Defibrillator discharge events were positively associated with the previous
20 day and 5-day mean NO2 concentrations: each 20-ppb increase in the previous day's NO2 level
21 was associated with an increased risk of a discharge event (OR = 1.55 [95% CI: 1.05, 2.29]) (see
22 Annex Table AX6.5.2 for the increase associated with a 20-ppb increase in NO2).
23 Three papers by the same team of investigators examined the association between air
24 pollution and the incidence of ventricular arrhythmias (Dockery et al., 2005; Rich et al., 2005)
25 and PAF episode (Rich et al. 2006b) in Boston. A total of 203 patients with ICDs who lived
26 within 25 miles of the ambient monitoring site were monitored. Data included a total of
27 635 person-years of follow-up or an average of 3.1 years per subject. The median (IQR) 48-h
28 average NO2 concentration was 22.7 (7.7) ppb. In the study by Dockery et al. (2005), significant
29 positive associations were observed between ventricular arrhythmias within 3 days of a prior
30 event, and a 2-day mean exposure to several air pollutants including PM2.s, BC, NO2, CO, and
31 SO2. Rich et al. (2005) examined associations between ambient air pollution levels less than 24
32 hours before the occurrence of a ventricular arrhythmia to make use of the precise time definition
March 2008 AX6-24 DRAFT-DO NOT QUOTE OR CITE
-------
1 available from the implantable cardioverter defibrillator (Rich et al., 2005). In single-pollutant
2 models, each 20-ppb increase in the mean NO2 level over the previous 2 days was associated
3 with an increased likelihood of ventricular arrhythmia, OR = 1.54 (95% CI: 1.11,2.18). The
4 association with NC>2 was not significant in two pollutant models with PM2.5, but remained
5 marginally significant in models with Os (2.0-ppb increase in 24-h moving average NC>2 was
6 associated with an OR = 1.36 [95% CI: 1.00, 1.80]). There was a strong association between an
7 increase of NO2 (by 20 ppb) and ventricular arrhythmia in the presence of ventricular arrhythmia
8 within the previous 72 h (OR = 2.09 [95% CI: 1.26, 3.51]). Increased but non-significant
9 associations were observed in this population between NO2 levels and PAF, as well as fine
10 particles and black carbon (Rich et al., 2006b).
11 A study conducted in St. Louis, which also examined the association of air pollutant level
12 within 24 hours of a ventricular arrhythmia, reports non-significant increases for NO2 and
13 elemental carbon, while SO2 was significantly associated with increased occurrence of
14 arrhythmia (Rich et al. 2006a). Metzger et al. (2007) examined the association of ventricular
15 tachyarrhythmias with air pollutants in the largest study to date (N =518), which was conducted
16 in Atlanta. These investigators report "suggestive" findings for course particulate but generally
17 no evidence of an association of NO2 and other pollutants with tachyarrhythmias (Metzger et al.,
18 2007).
19
20 AX6.2.2.4 Markers of Cardiovascular Disease
21
22 Epidemiological Studies
23 In a large cross-sectional study of 7,205 office workers in London, Pekkanen et al. (2000)
24 collected blood samples and analyzed the association between plasma fibrinogen, a risk factor
25 for CVD, and ambient levels of air pollution. In models adjusting for weather, demographic, and
26 socioeconomic factors, there was an increased likelihood of blood levels of fibrinogen >3.19 g/1
27 (90th percentile) for each 20-ppb increase in NO2 lagged by 3 days (OR =1.14 [95% CI: 1.03,
28 1.25]). The correlation between daily NO2 and other traffic-related pollutants were high: daily
29 levels of black smoke (r = 0.75), PMio (r = 0.76), SO2 (r = 0.62), CO (r = 0.81). The authors
30 suggest that the increased concentrations of fibrinogen, a mediator of cardiovascular morbidity
31 and mortality, may be an indicator of inflammatory reactions caused by air pollution.
March 2008 AX6-25 DRAFT-DO NOT QUOTE OR CITE
-------
1 Schwartz (2001) examined the association between fibrinogen, platelet count and white
2 blood cell (WBC) count in the Third National Health and Nutrition Examination Survey
3 (NHANES III). In single pollutant models NO2 was associated with platelet counts and
4 fibrinogen. However, in a two-pollutant model with PMio these associations became negative.
5 Pekkanen et al. (2002) enrolled a panel of 45 adults with coronary heart disease in order
6 to examine associations between heart function as measured by risk of ST-segment depression
7 and particulate pollution. Level of particulate and gaseous pollutants, including NO2, lagged by
8 2 days was found to have the strongest effect on risk of ST-segment depression during mild
9 exercise tests (OR= 14.1 [95% CI: 3.0, 65.4] for ST-segment depression of >0.1mV with a
10 20-ppb increase in NO2 lagged by 2 days). A large (n = 863) cross-sectional study of resting
11 heart rate (HR) of adults in France found significant associations between elevated levels of NO2
12 within 8-h of measurement and resting HR of >75 beats per minute (bpm) (OR = 2.7 [95% CI:
13 1.2, 5.4] for resting HR >75 bpm for each 20-ppb increase in NO2) (Ruidavets et al., 2005).
14 In a population based study of participants in the Atherosclerosis Risk in Communities
15 (ARIC) study, Liao et al. 2005 did not observe differences in White Blood Cell (WBC) count,
16 Factor VIII C, fibrinogen, Von Willibrand Factor (VWF) or albumin depending on 24-h average
17 NO2 level lagged 1 to 3 days prior to the examination date. However, PMio was associated with
18 factor VIII-C in this cohort. An association between PMio and serum albumin was observed
19 only among persons with a history of CVD (Liao et al. 2005).
20 Ruckerl et al. (2006) examined several markers of inflammation, cell adhesion, and
21 coagulation among a panel of 57 male patients with CHD. These authors primary hypothesis
22 was that C-reactive protein (CRP) would be increased with increases in air pollution. They also
23 investigated the effect of air pollution on other markers including serum amyloid A (S AA),
24 E-selectin, von Wildebrand factor antigen, intercellular adhesion molecule 1 (ICAM-1),
25 fibrinogen, factor VII, prothrombin fragment 1+2, and D-dimer. A significant association was
26 observed for NO2 with CRP greater than the 90th percentile but the strongest effect on CRP was
27 observed for ultrafme particles.
28 Steinvil et al. investigated the association of air pollutants with several markers of
29 inflammation (fibrinogen, CRP and WBC). Significant decreases in fibrinogen associated with
30 increases of 13 ppb in ambient NO2 were reported among men (all lags 0-7 and 7 day average)
31 and women (lag 0, 7 day average). The absolute change in fibrinogen ranged from 7.9 to 16.7
March 2008 AX6-26 DRAFT-DO NOT QUOTE OR CITE
-------
1 mg/dL (Steinvil et al., 2007). The mean NO2 level was 19.5 ppb (Steinvil et al., 2007).
2 was significantly associated with increased fibrinogen only at day 7. No correlations with CRP
3 and WBC were observed (Steinvil et al. 2007).
4 Baccarelli et al. (2007) investigated the effect of ambient NO2 with prothrombin time
5 (PT) and activated partial thromboplastin time (APTT) in 1218 normal subjects in Italy. Both
6 NO2 (coefficient = -0.08 95%CI: -0.15, 0.00) and PMio (coefficient = -0.08 95% CI: -0.14,
7 -0.01) on the same day and the average for 30 days prior to the examination were negatively
8 correlated with PT (e.g. PT became shorter indicating hypercoagulability), while no effect on
9 APTT was reported (Baccarelli et al., 2007).
10
11 AX6.2.2.5 Controlled Human Exposure and Animal Studies
12 Folinsbee et al. (1978) studied three groups of 5 healthy males exposed to 0.62-ppm NO2
13 for 2 h. The groups differed by duration of exercise during exposure: 15, 30, or 60 min. In
14 addition to pulmonary function, outcome measures included indirect calorimetry, cardiac output
15 using the CO2 rebreathing technique, blood pressure, HR, and diffusing capacity of the lung for
16 carbon monoxide (DLCO). There were no significant effects for the individual groups, or for the
17 15 subjects analyzed together. However, the small number of subjects in each group limited
18 statistical power.
19 Drechsler-Parks (1995) assessed changes in cardiac output using noninvasive impedance
20 cardiography. Eight older adults (56 to 85 years of age) were exposed to 0.60-ppm NO2,
21 0.45-ppm Os, and the combination of 0.60-ppm NO2 + 0.45-ppm Os for 2-h with intermittent
22 exercise. The exercise-induced increase in cardiac output was smaller with the NO2 + Os
23 exposures than with the filtered air or Os exposures alone. There were no significant differences
24 in minute ventilation, HR, or cardiac stroke volume, although the mean stroke volume was lower
25 for NO2 + Os than for air. The author speculated that chemical interactions between Os and NO2
26 at the level of the epithelial lining fluid led to the production of nitrite, leading to vasodilatation,
27 with reduced cardiac preload and cardiac output. This study has not been repeated.
28 Linn et al. (1985) reported small but statistically significant reductions in blood pressure
29 after exposure to 4-ppm NO2 for 75 min, a finding consistent with systemic vasodilatation in
30 response to the exposure. However, many subsequent studies at concentrations generally less
31 than 4 ppm have not reported changes in blood pressure in response to NO2 exposure.
March 2008 AX6-27 DRAFT-DO NOT QUOTE OR CITE
-------
1 There is also evidence that NO2 exposure may affect circulating red blood cells. Posin
2 et al. (1978) exposed 10 healthy males to 1- or 2-ppm NO2 for 2.5- to 3.0-h daily for 2 days.
3 Blood obtained immediately after the second exposure showed a reduced hemoglobin and
4 hematocrit (NO2: 41.96 ± 2.75; sham exposure: 43.18 ± 2.83, p = 0.001) and reduced red blood
5 cell acetyl cholinesterase levels. However, the control air exposures were not identical to and
6 concurrent with the NO2 exposures, a potential flaw in the study design.
7 In the study by Frampton et al. (2002), healthy subjects were exposed to air or 0.6- or
8 1.5-ppm NO2 for 3-h with intermittent exercise, and blood was obtained 3.5-h after exposure.
9 There was a significant, concentration-related reduction in hematocrit and hemoglobin in both
10 males and females, confirming the findings of Posin et al. (1978). These studies suggest that
11 NO2 exposure in the range of 1- to 2-ppm for a few hours is sufficient to alter the red blood cell
12 membrane. The reductions in blood hemoglobin were not sufficiently large to result in health
13 effects for these healthy subjects. However, in the Frampton study, the reduction in hemoglobin
14 represented the equivalent of about 200 mL of blood loss for a 70-kg male. This could
15 conceivably have adverse cardiovascular consequences for someone with significant underlying
16 lung disease, heart disease, or anemia.
March 2008 AX6-28 DRAFT-DO NOT QUOTE OR CITE
-------
TABLE AX6.3-1. STUDIES EXAMINING EXPOSURE TO INDOOR NO2 AND RESPIRATORY SYMPTOMS
NO2 Measurement
O
to
O
O
oo
Author, Year,
Location
Study Design
Exposure
Time
Mid-Range (ppb)
Range
(ppb)
Outcome
OR or RR (95% CI)
Pilotto et al. (2004) Subjects: 118 asthmatic 6 h
Australia children
Analysis: negative
binomial
Monitoring Device:
passive diffusion badges
mean(sd) 7,38
intervention 16 (7)
mean(sd) 12,116
control 47 (27)
daytime symptoms
difficulty breathing RR 2.44 (1.02, 14.29)=*
chest tightness
asthma attacks
difficulty breathing,
night
RR 2.22 (1.23, 4.00)*
RR 2.56 (1.08, 5.88)*
RR 3.12 (1.45, 7.14)*
X
to
VO
Pilotto et al. (1997) Subjects: 388 children 6 h
Australia
Analysis: generalized
linear mixed models
Monitoring Device:
passive diffusion badges
4, 132
wheeze (>40 ppb)
OR 1.41 (0.63, 3.15)
C Nitschke et al.
^ (2006)
' Australia
o
O
2|
0
H
O
O
H
W
O
HH
H
W
Subjects: 174 asthmatic 6 h mean home 20 (22)
children
Analysis: negative mean school 34 (28)
binomial
Monitoring Device:
passive diffusion badges
night symptoms
difficulty breathing
school max
home max
chest tightness
school max
RR 1.23 (1.10, 1.39)
RR 1.06 (1.02, 1.10)
RR 1.25 (1.14, 1.37)
-------
TABLE AX6.3-1 (cont'd). STUDIES EXAMINING EXPOSURE TO INDOOR NO2 AND RESPIRATORY SYMPTOMS
NO2 Measurement
O
to
O
O
oo
Author, Year,
Location
Study Design
Exposure Time Mid-range (ppb) Range (ppb)
Outcome
OR or RR (95% CI)
Garrettetal. (1998)
Australia
Subjects: 148 children
Analysis: multiple
logistic regression
Monitoring Device:
passive monitors
4 days
med6
plO-p90, 3, 15 chest tightness
OR 1.53 (0.45, 5.32)
X
ON
OJ
O
H
6
o
o
H
O
O
H
W
O
HH
H
W
Smith et al. (2000)
Australia
Subjects: 125
asthmatic
adults/children
Analysis: GEE
Monitoring Device:
passive diffusion
badges
4.5 h
4, 147
children (n = 49, 0 14) OR 1.12 (1.07, 1.18)
chest tightness
Belanger et al.
(2006)
Northeast U.S.
Chauhan et al.
(2003)
Southampton, U.K.
Subjects: 728 2 wks
asthmatic children
Analysis: logistic,
Poisson regression
Monitoring Device:
Palmes tubes
Subjects: 114 7d
asthmatic children
Monitoring Device:
Palmes diffusion tubes
mean (sd) gas
home 26 (18)
mean (sd) elect
home 9 (9)
Exposure tertiles:
<4; 4-7; >7
multifamily housing
wheeze
chest tightness
Increased symptom
score, comparing first
and second tertiles of
NO2 exposure
Increased symptom
score, comparing first
and third tertiles of
NO2 exposure
RR 1.33 (1.05, 1.68)
RR 1.51 (1.18, 1.91)
0.6(0.01, 1.18)
2.1(0.52,3.81)
-------
O
to
O
O
oo
TABLE AX6.3-1 (cont'd). STUDIES EXAMINING EXPOSURE TO INDOOR NO2 AND RESPIRATORY SYMPTOMS
Author, Year,
Location
Van Strien et al.
(2004)
Northeast U.S.
Study Design
Subjects: 762 infants
Analysis: Poisson
regression
NO2 Measurement
Exposure Time Mid-range (ppb)
medlO
Range
(ppb)
Outcome
persistent cough
ORorRR(95%CI)
<5.1 ppb
5.1,9.9 ppb
9.9, 17.4 ppb
>17.4 ppb
shortness of breath
RR1.0
RR 0.96 (0.69, 1.36)
RR 1.33 (0.94, 1.88)
RR 1.52 (1.00, 2.31)
X
H
6
o
o
H
O
O
H
W
<5.1ppb
5.1,9.9 ppb
9.9, 17.4 ppb
>17.4 ppb
RR1.0
RR 1.59 (0.96, 2.62)
RR 1.95 (1.17, 3.27)
RR 2.38 (1.31, 4.34)
Notes:
Unless otherwise noted, results given for 20-ppb increase in NO2.
*For purpose of comparison, RRs from Pilotto et al. (2004) are shown here as risk of symptoms given greater exposure to NO2, i.e., control (unflued gas heater) vs intervention (flued or electric
replacement heater).
RRs reported by Pilotto et al. (2004) as protective effects for intervention vs. control.
O
HH
H
W
-------
TABLE AX6.3-2. STUDIES EXAMINING EXPOSURE TO AMBIENT NO2 AND ACUTE RESPIRATORY SYMPTOMS
USING GENERALIZED ESTIMATING EQUATIONS (GEE) IN THE ANALYSIS METHOD
to
o
o
oo
Author, Year,
Location
Subjects
NO2 Measurement
Avg Copollutants &
Time Mid-range (ppb) Range (ppb) Correlations
Outcome
OR (95% CI)
Children: Multi-City Studies
>
X
ON
i
OJ
to
o
^
H
O
2
0
H
O
O
^H
C
H
W
O
P
O
HH
H
W
Schwartz et al.
(1994)
6 U.S. Cities
Mortimer et al.
(2002)
U.S., NCICAS
Schildcrout et al.
(2006)
North America
CAMP
1844 children
864 asthmatic
children
990 asthmatic
children
24 h medlS plO-p90, 5, 24 PM25: r=0.35
PM10: r = 0.36
O3: r=-0.28
SO2: r=0.51
4h med25 7,90 O3: r=0.27
24 h med23 minplOtomax PM10:r = 0.26,
p90, 10, 37 0.64
O3: r= 0.04, 0.47
SO2: r= 0.23, 0.68
CO: r= 0.63, 0.92
cough, incidence:
lag 1-4 mean
asthma symptoms:
lag 1-6 mean
asthma symptoms:
lagO
lagl
lag 2
3 -day moving sum
1.61(1.08,2.43)
1.48(1.02,2.16)
1.06(1.00, 1.13)
1.04(0.97,1.10)
1.09(1.03,1.15)
1.04(1.01, 1.07)
Children: Single-City Studies
Pino et al. (2004)
Chile
Ostroetal. (2001)
Southern California
504 infants
138 asthmatic
children, African
American
24 h mean (sd) 4 1(19) p5-p95, 20, 81
1 h mean (sd) 80 (4) 20, 220 PM25: r = 0.34
PM10: r = 0.63
O3: r=0.48
wheezy bronchitis:
6 day lag
cough, incidence:
Iag3
wheeze, incidence:
Iag3
1.14(1.04, 1.30)
1.07(1.00,1.14)
1.05(1.01,1.09)
-------
TABLE AX6.3-2 (cont'd). STUDIES EXAMINING EXPOSURE TO AMBIENT NO2 AND ACUTE RESPIRATORY
SYMPTOMS USING GENERALIZED ESTIMATING EQUATIONS (GEE) IN THE ANALYSIS METHOD
^j NO2 Measurement
o
o Author, Year, Avg Mid-range Copollutants &
Location Subjects Time (ppb) Range (ppb) Correlations Outcome
OR (95% CI)
Children: Single-City Studies (cont'd)
Deffino et al. (2002) 22 asthmatic 8h mean (sd) 15 (7) 6,34
Southern California children
Segalaetal. (1998) 84 asthmatic 24 h mean (sd) 30 (8) 13,65
Paris children
X
Oi
OJ
Just et al. (2002) 82 asthmatic 24 h mean (sd) 29 (9) 12, 59
Paris children
O
£1
H
6
o
Z; Jalaludin et al. (2004) 148 children with 15 h mean (sd) 15 (6) 3, 79
O Austrailia wheeze history
O
O
H
W
O
O
HH
H
W
PM10: r = 0.55 asthma symptoms:
O3: r=0.26 lagO
PM25: r = (0.61)* asthma symptoms:
PM10: r = 0.55 incidence: lagO
SO2: r=0.54
lagl
lag 4
nocturnal cough:
incidence: lag 3
lag 4
PM25: r = 0.92* nocturnal cough:
PM10: r = 0.54 incidence: lagO
O3: r=0.09
SO2: r=0.69
lag 0-2
lag 0-4
PM10: r = 0.26 Wet cough: lagO
O3: r=-0.31
1.91(1.07,
1.89(1.13,
1.36 (0.70,
1.80(1.07,
1.44 (0.99,
1.74(1.20,
2.11(1.20,
1.80 (0.89,
1.58(0.73,
1.13(1.00,
3.39)
3.17)
2.64)
3.01)
2.08)
2.52)
3.74)
3.84)
3.54)
1.26)
-------
H
6
o
o
H
O
O
H
W
TABLE AX6.3-2 (cont'd). STUDIES EXAMINING EXPOSURE TO AMBIENT NO2 AND ACUTE RESPIRATORY
SYMPTOMS USING GENERALIZED ESTIMATING EQUATIONS (GEE) IN THE ANALYSIS METHOD
^j NO2 Measurement
qj
o Author, Year, Avg Copollutants &
Location Subjects Time Mid-range (ppb) Range (ppb) Correlations Outcome
OR (95% CI)
Adults:
Segala et al. (2004) 46 nonsmoking adults 24 h mean (sd) 30 (9) 12, 71
Paris
Von Klot et al. 53 asthmatic adults 24 h med 24 4, 63
(2002)
Germany
X
Oi
PM25: r = 0.82* sore throat, cough:
PM10: r = 0.83 lag 0-4
PM10: r = 0.74 wheeze, prev:
SO2: r = 0.36 5-day mean
CO: r=0.82
phlegm, prev:
5 -day mean
cough, prev:
5 -day mean
breathing prob in
a.m. : 5-day mean
4.05 (1.20,
1.15(1.02,
1.22(1.10,
1.15(1.00,
1.25(1.10,
13.60)
1.31)
1.39)
1.31)
1.39)
Odds ratios (OR) given for 20-ppb increase in NO2 with 24-h averaging time, or 30 ppb for 1-h averaging time. (20-ppb increases also used for times between 1 and 24 h.) *BS
O
HH
H
W
-------
O
to
O
O
oo
X
H
6
o
o
H
O
O
H
W
TABLE AX6.3-3. RESPIRATORY HEALTH EFFECTS OF OXIDES OF NITROGEN:
HOSPITAL ADMISSIONS
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels &
Monitoring Stations
Copollutants &
Correlations
Effects: Relative Risk or Percent Change &
Confidence Intervals (95%)
UNITED STATES
Moolgavkar (2000a,b,c)
Moolgavkar (2003)
Multi-city, United States:
Chicago, Los Angeles,
Maricopa County,
(Phoenix).
Period of Study:
1987-1995
Outcomes (ICD 9 codes): COPD
including asthma (490-496)
Age groups analyzed: 0-19,20-64,65+
(LA only)
Study Design: Time-series
Statistical Analyses: Poisson regression,
GAM
Covariates: Day of wk, temporal trends,
temperature, relative humidity
Lag: 0-5 days
Chicago
Median: 25 ppb
IQR: 10 ppb
Los Angeles
Median: 38 ppb
IQR: 18 ppb
Maricopa
Median: 19 ppb
IQR: 12 ppb
Chicago:
PM10;r=0.49
CO; r= 0.63
S02; r = 0.44
O3; r = 0.02
LA:
PM2.5; r = 0.73
PM10; r = 0.70
CO; r= 0.80
SO2; r = 0.74
O3;r=-0.10
Maricopa:
PM10; r = 0.22
CO; r= 0.66
S02; r = 0.02
03;r=-0.23
Increment: 10 ppb
COPD, >65 yrs
Chicago 1.7% [CI 0.36, 3.05] lag 0 - GAM default
Chicago 2.04% [t = 2.99] lag 0 - GAM-100
Los Angeles 2. 5% [CI 1.85, 3.15] lag 0 - GAM
default
Los Angeles 2.84% [t = 13.32] lag 0 - GAM - 30
Los Angeles 1 .80% [t = 9.60] lag 0 - GAM - 100
Los Angeles 1 .78% [t = 7.72] lag 0 - NS-100
Phoenix 4.4% [CI 1.07, 7.84] lag 5
Chronic Respiratory Disease
Los Angeles
0-19 yrs 4.9% [CI 4.1, 5.7] lag 2
20-64 yrs 1.7% [CI 0.9, 2.1] lag 2
Multi-pollutant model
Moolgavkar*et al. (1997) Outcomes (ICD 9 codes): COPD
United States:
Minneapolis-St. Paul
Period of Study:
1986-1991
including asthma (490-496), Pneumonia
(480-487)
Age groups analyzed: 65+
Study Design: Time-series
Statistical Analyses: Semi-parametric
Poisson regression, GAM
Covariates: Day of wk, season, temporal
trends, temperature
Statistical Package: S Plus
Lag: 0-3 days
NO2 24-h avg (ppb)
16.3 ppb
IQR: 9.5 ppb
PM10;r=0.31
SO2;r = 0.09
CO; r= 0.58
NO2 andPM10: 1.72% [t = 3.18] lag 0 - GAM-100
NO2 and PM2.5: 1.51% [t = 2.07] lag 0 - GAM-100
Increment: 10 ppb
Sum of Pneumonia and COPD
2.2% [0.2, 4.2] lag 1
Pneumonia Only
3.1% [0.6, 5.6] lag 1,20 df
1.7% [-0.8, 4.2] lag 1,130 df
O
HH
H
W
-------
O
to
O
O
oo
X
H
6
o
o
H
O
O
H
W
O
HH
H
W
TABLE AX6.3-3 (cont'd). RESPIRATORY HEALTH EFFECTS OF OXIDES OF NITROGEN:
HOSPITAL ADMISSIONS
Reference, Study
Location, & Period Outcomes, Design, & Methods
Mean Levels &
Monitoring Stations
Copollutants &
Correlations
Effects: Relative Risk or Percent Change &
Confidence Intervals (95%)
UNITED STATES (cont'd)
Neidell (2004) Outcomes (ICD 9 codes): Asthma
California Age groups analyzed: <18; 0-1; 1-3; 3-6;
N02 (ppb)
Mean: 45.947
03
CO
Increment: NR
Period of Study:
1992-1998
6-12; 12-18
Study Design: Time-series
Statistical Analyses: NR
Covariates: Temperature, precipitation,
influenza epidemic
Seasons: Nov-Maronly
Lag: 0-4 days
SD=17.171
Age 0-1
Fixed effects: 0.009(0.014)
Controlled for avoidance behavior: 0.009 (0.014)
Single pollutant: 0.001(0.011)
Adjusted for SES: 0.021(0.017)
Interaction with Low SES: -0.017 (0.029)
Age: 1-3
Fixed effects: 0.002(0.016)
Controlled for avoidance behavior: 0.002 (0.016)
Single pollutant: 0.009 (0.013)
Adjusted for SES: -0.001(0.020)
Interaction with Low SES: -0.004 (0.032)
Age 3-6
Fixed effects: 0.006(0.016)
Controlled for avoidance behavior: 0.006 (0.016)
Single pollutant: 0.028 (0.013)
Adjusted for SES: 0.020 (0.020)
Interaction with Low SES: -0.037 (0.033)
Age 6-12
Fixed effects: 0.041(0.015)
Controlled for avoidance behavior: 0.042 (0.015)
Single pollutant: 0.047(0.012)
Adjusted for SES: 0.040(0.018)
Interaction with Low SES: -0.016 (0.031)
Age: 12-18
Fixed effects: 0.005(0.013)
Controlled for avoidance behavior: 0.005 (0.013)
Single pollutant: 0.015(0.010)
Adjusted for SES: 0.013 (0.017)
Interaction with Low SES: -0.020 (0.026)
-------
O
to
O
O
oo
X
H
6
o
o
H
O
O
H
W
O
HH
H
W
TABLE AX6.3-3 (cont'd). RESPIRATORY HEALTH EFFECTS OF OXIDES OF NITROGEN:
HOSPITAL ADMISSIONS
Reference, Study
Location, & Period Outcomes, Design, & Methods
Mean Levels &
Monitoring Stations
Copollutants &
Correlations
Effects: Relative Risk or Percent Change &
Confidence Intervals (95%)
UNITED STATES (cont'd)
Karr et al. (2006) Outcomes (ICD 9 codes): Acute
Southern LA County, CA, bronchiolitis (466. 1 )
United States Age groups analyzed: 0-1 yr
1-h max NO2 (ppb)
Mean: 59 ppb
IQR: 26 ppb
CO
PM2.5
Increment: 26 ppb (IQR)
Acute bronchiolitis
Period of Study:
1995-2000
Study Design: Case-crossover
N: 19,109
Statistical Analyses: Conditional logistic
regression
Covariates: Day of wk, temperature,
humidity
Seasons: Nov-Maronly
Lag: 0-4 days
Number of Stations: 34
OR 0.96 [0.94, 0.99] lag 4
OR 0.97 [0.95, 0.99] lag 1
Stratified by Gestational Age at Birth:
37-44 wks
0.98 [0.95, 1.00] lag 1; 0.97 [0.94, 0.99] lag 4
34-36 wks
0.90 [0.84, 0.97] lag 1; 0.94 [0.88, 1.02] lag 4
29-33 wks
1.01 [0.91, 1.13] lag 1; 0.90 [0.80, 1.01] lag 4
25-28 wks
0.94 [0.78, 1.13] lag 1; 0.90 [0.73, 1.11] lag 4
Linn et al. (2000)
Los Angeles, United
States
Period of Study:
1992-1995
Outcomes (ICD 9 codes): Asthma (493),
COPD (APR-DRG 88), Pulmonary
diagnoses
(APR-DRG 75-101)
Age groups analyzed: >30
Study Design: Time-series
N: 302,600
Statistical Analyses: Poisson regression,
GAM, OLS regression
Covariates: day of wk, holiday, max
temperature, min temperature, rain days,
mean temperature, barometric pressure,
season
Seasons: Winter (Jan-Mar),
Spring (Apr-Jun), Summer (Jul-Sep),
Fall (Oct-Dec)
Statistical Package: SPSS and SAS
Lag: 0, 1 days
All concentrations are in
pphm.
Winter: 3. 4 ±1.3
Spring: 2. 8 ±0.9
Summer: 3. 4 ± 1.0
Autumn: 4.1 ± 1.4
Overall: 3. 4 ±1.3
Winter:
CO; r= 0.89
PM10;r=0.88
O3;r=-0.23
Spring:
CO; r= 0.92
PM10; r = 0.67
03;r=0.35
Summer
CO; r= 0.94
PM10;r=0.80
O3;r=0.11
Winter
CO; r= 0.84
PM10;r=0.80
03;r=-0.00
Increment: 10 ppb
All pulmonary
All seasons: 0.7% ±0.3%
Winter: 1.1% ±0.5%
Spring: 0.7% ±0.1%
Summer: 0.4% ±0.8%
Autumn: 1.2% ±0.4%
Asthma
All season: 1.4% ±0.5%
Winter: 2. 8% ±0.1%
Spring: NR
Summer: NR
Autumn: 1.9% ±0.8%
COPD
All season: 0.8% ±0.4%
Winter: NR
Spring: NR
Summer: NR
Autumn: 1.6% ±0.6%
-------
O
to
O
O
oo
TABLE AX6.3-3 (cont'd). RESPIRATORY HEALTH EFFECTS OF OXIDES OF NITROGEN:
HOSPITAL ADMISSIONS
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels &
Monitoring Stations
Copollutants & Effects: Relative Risk or Percent Change
Correlations & Confidence Intervals (95%)
UNITED STATES (cont'd)
X
oo
H
6
o
o
H
O
O
H
W
O
HH
H
W
Magas et al. (2007)
Oklahoma City, OK
Period of Study: 2001-
2003
Gwynn*et al. (2000)
Buffalo, NY
United States
Period of Study:
1988-1990
Days: 1,090
Zanobetti and Schwartz
(2006)
Boston, MA, United
States
Period of Study:
1995-1999
Outcomes (ICD 9 codes): Asthma (493)
Age groups analyzed: 0-14
Study Design: Time-series
N: 1,270
Statistical Analyses: negative binomial
regression
Covariates: gender, day of wk, holiday,
Lag:
Outcomes (ICD 9 codes): Respiratory
admissions: Acute bronchitis/bronchiolitis (466);
Pneumonia (480-4860); COPD and Asthma (490-
493, 496)
Age groups analyzed: 6
Study Design: Time-series
N: 24,
Statistical Analyses: Poisson regression with
GLM and GAM
Covariates: Season, day of wk, holiday,
temperature, relative humidity
Lag: 0-3 days
Outcomes (ICD 9 codes): Pneumonia (480-7)
Age groups analyzed: 65+
Special Population: Medicare patients only
Study Design: Case-crossover
N: 24,857
Statistical Analyses: Conditional logisitic
regression
Covariates: Apparent temperature, day ofwk
Seasons: Warm (Apr-Sep), Cool (Oct-Mar)
Statistical Package: SAS
Lag: 0, 1 days, 0-1 avg
24-havg: 11.7ppb
Number of monitors: 10
24-h avg NO2 (ppb):
Min: 4.0
25th: 15.5
Mean: 20.5
75th: 24.5
Max: 47.5
NO2 median 23.20 ppb;
90-10%: 20.41 ppb;
For lag 0-12 day avg
90-10% =16.8 ppb;
IQR= 10.83
Number of Stations: 5
03
PM2.5
H+; r = 0.22
SO42~;r=0.36
PM10; r = 0.44
O3;r=0.06
SO2;r = 0.36
CO; r= 0.65
COH; r = 0.72
PM2.5; r = 0.55
BC;r=0.70
CO; r= 0.67
O3;r=-0.14
Qualitative results: ambient concentrations
of NO2 increased pediatric asthma
hospitalizations
Increment: 27 9 ppb (Max-Mean; IQR)
NO2 alone:
Max-Mean RR 1.033 (t = 1.32) lag 1
IQRRR1.01(t= 1.32) lag 1
Increment: 20.41 ppb (90-10%)
Pneumonia
-0.16% [-4.73, 4.42] lag 0
Increment: 16.78 ppb (90-10%)
Pneumonia
2.26% [-2.55, 7.01] lag 0-1
-------
O
to
O
O
oo
TABLE AX6.3-3 (cont'd). RESPIRATORY HEALTH EFFECTS OF OXIDES OF NITROGEN:
HOSPITAL ADMISSIONS
X
ON
OJ
vo
H
6
o
o
H
O
O
H
W
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels &
Monitoring Stations
Copollutants & Effects: Relative Risk or Percent Change &
Correlations Confidence Intervals (95%)
CANADA
Burnett etal. (1997a)
16 cities
Canada
Period of Study:
4/1981-12/1991
Days: 3,927
Yang et al. (2003)
Vancouver, Canada
Period of Study: 1986-
1998
Days: 4748
Outcomes (ICD 9 codes): All respiratory
admissions
(466, 480-6, 490-4, 496)
Study Design: Time-series
N: 720,519
# of hospitals: 134
Statistical Analyses: Random effects
relative risk regression model
Covariates: Long-term trend, season, day of
wk, hospital,
Statistical Package: NR
Lag: 0, 1,2 day
Outcomes (ICD 9 codes): All respiratory
admissions (460-519)
Study Design: Case-crossover
Age groups analyzed: <3, >65
Statistical Analyses: Conditional logistic
regression
Statistical Package: NR
Lag: 0-5 days
1 -h max NO2 (ppb)
Mean: 35.5
SD=16.5
25th: 25
50th: 33
75th: 43
95th: 62
99th: 87
24-h avg NO2 (ppb):
Mean: 18.74
SD = 5.66
5th: 11.35
25th: 14.88
50th: 17.80
75th: 21.45
100th: 49.00
IQR: 5.57
Number of stations: 30
O3;r=0.20 Increment: 10 ppb
CO
SO2 Single pollutant
COH NO2 and respiratory admissions, p = 0.772
Multipollutant model (adjusted for CO, O3, SO2,
COH, dew point):
RR 0.999 [0.9922, 1.0059] lag 0
CO Increment: 5. 57 ppb (IQR)
SO2
O3; r = -0.32 All Respiratory Admissions <3 yrs:
COH NO2 alone: OR 1.05 [1.02, 1.09] lag 1
NO2 + O3: OR 1.05 [1.02, 1.09] lag 1
N02 + 03 + CO + COH + S02: OR 1 .05
[0.99, 1.11] lag 1
All Respiratory Admissions >65 yrs:
NO2 alone: OR 1.05 [1.03, 1.07] lag 1
N02 + 03: OR 1.04 [1.02, 1.07] lag 1
NO2 + O3 + CO + COH + SO2: OR 1 .05
[1.01, 1.08] lag 1
O
HH
H
W
-------
O
to
O
O
oo
X
H
6
o
o
H
O
O
H
W
TABLE AX6.3-3 (cont'd). RESPIRATORY HEALTH EFFECTS OF OXIDES OF NITROGEN:
HOSPITAL ADMISSIONS
Reference, Study Mean Levels &
Location, & Period Outcomes, Design, & Methods Monitoring Stations
Copollutants & Effects:
Correlations
Relative Risk or Percent Change &
Confidence Intervals (95%)
CANADA (cont'd)
Fung et al. (2006) Outcomes (ICD 9 codes): All respiratory NO224-havg:
Vancouver, BC, Canada hospitalizations (460-519) Mean: 16. 83 ppb,
Age groups analyzed: 65+ SD = 4.34;
Period of Study: Study Design: (1) Time-series, (2) Case- !QR: 5.43 ppb;
6/1/95-3/31/99 crossover, (3) DM-models (Dewanji and range: 7.22,33.89
Moolgavkar 2000, 2002)
N: 40,974
Statistical Analyses: (1 ) Poisson,
(2) conditional logistic regression, (3) DM
method - analyze recurrent data in which
the occurrence of events at the individual
level over time is available
Covariates: Day of wk
Statistical Package: S-Plus and R
Lag: Current day, 3 and 5 day lag
CO; r= 0.74 Increment: 5.43 ppb. (IQR)
COH; r = 0.72
S02; r = 0.57 NO2
PM10;r=0.54 RR 1
PM2.5;r = 0.35 RR 1
PM10-2.5;r = 0.52 RR 1
O3;r=-0.32 RR 1
N02
RR1
RR1
RR1
RR1
NO2
RR1
RR1
RR1
RR1
Time-series
.018
.024
.025
.027
[1.
[1.
[1.
ro.
,003, 1
,004, 1
,000, 1
,998, 1
.034]
.044]
.050]
.058]
lag
lag
lag
las
;0
;0-3
;0-5
;0-7
Case-crossover
.028
.035
.032
.028
ri
L
[1.
[1.
[0.
,010, 1
,012,1
,006, 1
,997, 1
047]
J
.059]
.060]
.060]
las
c
lag
lag
lag
;0
;0-3
;0-5
;0-7
DM model
.012
.018
.007
.002
[0.
[1.
[0.
[0.
,997, 1
,000, 1
,988, 1
,981,1
.027]
.037]
.026]
.023]
lag
lag
lag
lag
;0
;0-3
;0-5
;0-7
DM method produced slightly higher RR
estimates on O3, SO2, and PM2 5 compared to
Time-series and case-crossover, and slightly
lower RR estimates on COH, NO2, and PM10,
though the results were not significantly different
from one another.
O
HH
H
W
-------
O
to
O
O
oo
X
-k
H
6
o
o
H
O
O
H
W
TABLE AX6.3-3 (cont'd). RESPIRATORY HEALTH EFFECTS OF OXIDES OF NITROGEN:
HOSPITAL ADMISSIONS
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels & Copollutants &
Monitoring Stations Correlations
Effects: Relative Risk or Percent Change &
Confidence Intervals (95%)
CANADA (cont'd)
Yang (2005)
Vancouver, BC, Canada
Period of Study:
1994-1998
Days: 1826
Outcomes (ICD 9 codes): COPD excluding
asthma (490-2, 494, 496)
Age groups analyzed: 65+
Study Design: Time-series
N: 6,027
Statistical Analyses: Poisson regression
with GAM (with more stringent criteria)
Covariates: Temperature, relative humidity,
day of wk, temporal trends, season
Statistical Package: S-Plus
Lag: 0-6 days, moving avgs
24-havg: 17.03 ppb, PM10;r=0.61
SD = 4.48; SO2;r = 0.61
IQR: 5.47 ppb; CO; r= 0.73
Range: 4.28,33.89 O3;r=-0.10
Winter: 19.20(4.86)
Spring: 15.36(3.72)
Summer: 16.33(4.57)
Fall: 17.27(3.77)
Number of Stations: 31
Increment: 5. 5 ppb (IQR)
COPD
>65 yrs, yr round
RR 1.05 [1.01, 1.09] lag 0
RR 1.04 [1.00, 1.10] lag 0-1
RR 1.07 [1.01, 1.13] lag 0-2
RR 1.08 [1.02, 1.15] lag 0-3
RR 1.10 [1.03, 1.17] lag 0-4
RR1.11 [1.04, 1.19] lag 0-5
RR1.11 [1.04, 1.20] lag 0-6
Two-pollutant model
PM10: RR 1.03 [0.90, 1.17] lag 0
CO: RR 1.07 [0.96, 1.20] lag 0-6
03: RR 1.12 [1.04, 1.20] lag 0-6
Multipollutant models
N02, CO, S02, 03, PM10: RR 1.01 [0.88, 1.16]
N02, CO, S02, 03: RR 1.06 [0.95,1.19]
NO2 was strongest predictor of hospital admission
for COPD among all gaseous pollutants in single-
pollutant models.
O
HH
H
W
-------
O
to
O
O
oo
TABLE AX6.3-3 (cont'd). RESPIRATORY HEALTH EFFECTS OF OXIDES OF NITROGEN:
HOSPITAL ADMISSIONS
X
to
H
6
o
o
H
O
O
H
W
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels &
Monitoring Stations
Copollutants & Effects: Relative Risk or Percent Change &
Correlations Confidence Intervals (95%)
CANADA (cont'd)
Lin* et al. (2004)
Vancouver, BC
Canada
Period of Study:
1987-1991
Outcomes (ICD 9 codes): Asthma (493)
Age groups analyzed: 6-12
Study Design: Time-series
N: 3,754 (2,331 male, 1,423 female)
Statistical Analyses: Semi-parametric
Poisson regression with GAM (with default
and more stringent criteria)
Covariates: Trend, day of wk,
Statistical package: S-Plus
Lag: Cumulative 1-7 day
24-h avg NO2 (ppb)
Mean: 18.65
SD = 5.59
Min: 4.28
25th: 14.82
50th: 17.75
75th: 21.36
Max: 45.36
Number of stations: 30
CO; r = 0.73 Increment: 6.54 ppb (IQR)
SO2;r = 0.67
O3;r=-0.03 Boys 6-12 yrs by SES status: Low; High
PM2.5; r = 0.37 Lag 1 RR 1.13 [1.04, 1.23]; 1.04 [0.95, 1.14]
PM10;r=0.55 Lag 2 RR 1.13 [1.02, 1.24]; 1.06 [0.95, 1.18]
Lag 3 RR 1.14 [1.02, 1.27]; 1.06 [0.94, 1.19]
Lag 4 RR 1.14 [1.02, 1.28]; 1.05 [0.92, 1.19]
Lag 5 RR 1.12 [0.99, 1.27]; 1.10 [0.96, 1.26]
Lag 6 RR 1.12 [0.98, 1.28]; 1.07 [0.93, 1.23]
Lag 7 RR 1.11 [0.97,1.28]; 1.09 [0.94, 1.27]
Girls 6-12 yrs by SES status: Low; High
Lag 1 RR 1.07 [0.96, 1.19]; 1.01 [0.90, 1.13]
Lag 2 RR 1.03 [0.91, 1.17]; 0.98 [0.85, 1.12]
Lag 3 RR 1.04 [0.91, 1.20]; 0.98 [0.84, 1.13]
Lag 4 RR 1.11 [0.95,1.29]; 1.01 [0.86,1.19]
Lag 5 RR 1.11 [0.94, 1.30]; 0.99 [0.83, 1.17]
Lag 6 RR 1.08 [0.91, 1.28]; 1.03 [0.86, 1.24]
Lag 7 RR 1.07 [0.90, 1.28]; 1.09 [0.90, 1.32]
Multipollutant model (adjusted for SO2)
Boys, Low SES:
1.16 [1.06, 1.28] lag 1
1.18 [1.03, 1.34] lag 4
Results presented are default GAM, but authors
state that use of natural cubic splines with a more
stringent convergence rate produced similar
results.
O
HH
H
W
-------
O
to
O
O
oo
TABLE AX6.3-3 (cont'd). RESPIRATORY HEALTH EFFECTS OF OXIDES OF NITROGEN:
HOSPITAL ADMISSIONS
X
H
6
o
o
H
O
O
H
W
Reference, Study
Location, & Period Outcomes, Design, & Methods
Mean Levels &
Monitoring Stations
Copollutants &
Correlations
Effects: Relative Risk or Percent Change &
Confidence Intervals (95%)
CANADA (cont'd)
Chen etal. (2005)
Vancouver, BC
Period of Study:
6/1995-3/1999
Lin etal. (2003)
Toronto, ON
Period of Study:
1981-1993
Outcomes (ICD 9 codes): All Respiratory
(460-519)
Age groups analyzed: 65+
Study Design: Time-series
N: 12,869 overall admissions
Statistical Analyses: Poisson regression
with GLM
Covariates: Trend, day of wk, weather
Statistical package: S-Plus
Lag: 1-7
Outcomes (ICD 9 codes): Asthma (493)
Age groups analyzed: 6-12
Study Design: Bi-directional case-crossover
N: 7,319
Statistical Analyses: Conditional logistic
regression
Covariates: Daily maximum and minimum
temperatures and avg relative humidity
Lag: Cumulative lag of 1-7 days
24-havg: 16.8 (4. 3) ppb
Range: 7.2-33.9
IQR: 5.4
NO2 24-havg: 25.24 ppb,
SD = 9.04;
IQR: 11 ppb;
Range: 3.00, 82.00
Number of Stations: 4
PM10;r=0.54
PM10_2.5;r = 0.54
PM2.5; r = 0.36
CO; r= 0.74
SO2;r = 0.57
03;r=-0.32
C0;r:
SO2;r
PM10;
O3;r =
PM2.5;
= 0.55
= 0.54
r=0.52
:Q.03
r = 0.50
PM10-2.5;r = 0.38
No analyses forNO2
Increment: 11
Boys 6-12 yrs;
ppb (IQR)
Girls 6-12
yrs
LagO: OR 1.04 [0.99, 1.10]; 0.
Lag 0-1: OR1
Lag 0-2: OR 1
Lag 0-3: OR 1
Lag 0-4: OR 1
Lag 0-5: OR 1
Lag 0-6: OR 1
.07 [1.00,
.09 [1.01,
.10 [1.01,
.10 [1.00,
.12 [1.01,
.11 [1.00,
1.14];
1.17];
1.20];
1.20];
1.23];
1.24];
99 [0
1.03
1.07
1.09
1.14
1.16
1.16
.92,1.
[0.
[0.
[0.
[1.
[1.
[1.
,94,
,96,
,97,
,02,
,02,
,02,
06]
1.12]
1.18]
1.21]
1.28]
1.31]
1.32]
O
HH
H
W
-------
O
to
O
O
oo
TABLE AX6.3-3 (cont'd). RESPIRATORY HEALTH EFFECTS OF OXIDES OF NITROGEN:
HOSPITAL ADMISSIONS
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels &
Monitoring Stations
Copollutants & Effects: Relative Risk or Percent Change
Correlations & Confidence Intervals (95%)
CANADA (cont'd)
X
Burnett etal. (1997b)
Toronto, Canada
Period of Study:
1992-1994
Outcomes (ICD 9 codes): Respiratory
tracheobronchitis (480-6), COPD (491-4, 496)
Study Design: Time-series
Statistical Analyses: Poisson regression, GEE,
GAM
Covariates: Temperature, dew point
temperature, long-term trend, season, influenza,
day of wk
Seasons: summers only
Lag: 0,1,2,3,4 days
MeanNO2: 38.5 ppb
IQRNO2: 5.75 ppb
Range: 12, 81
Number of Stations:
6-11
PM10;r=0.61
CO; r= 0.25
H+; r = 0.25
SO4; r = 0.34
TP;r=0.61
FP; r = 0.45
CP;r=0.57
COH;r = 0.61
O3;r=0.07
SO2; r = 0.46
Increment: 5.75 ppb (IQR)
Respiratory - Percent increase
4.4% [CI 2.4, 6.4], lag 0
Copollutant and multipollutant models RR
(t-statistic):
N02,COH: 1.018(1.36)
NO2,H+: 1.037(3.61)
NO2, SO4: 1.033(3.05)
N02,PM10: 1.039(2.85)
NO2,PM2.5: 1.037(3.13)
N02, PM10-2.5: 1.037(2.96)
NO2, O3, SO2: 1.028(2.45)
N02, 03, S02, COH: 1.010(0.71)
N02, 03, S02, FT: 1.027(2.39)
NO2, O3, SO2, SO4: 1.027(2.36)
N02, 03, S02, PM10: 1.028 (1.77)
NO2, O3, SO2, PM2 5: 1.028 (2.26)
N02, 03, S02, PM10-2.5: 1.022(1.71)
H
6
o
o
H
O
O
H
W
O
HH
H
W
-------
O
to
O
O
oo
TABLE AX6.3-3 (cont'd). RESPIRATORY HEALTH EFFECTS OF OXIDES OF NITROGEN:
HOSPITAL ADMISSIONS
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels &
Monitoring Stations
Copollutants & Effects: Relative Risk or Percent Change &
Correlations Confidence Intervals (95%)
CANADA (cont'd)
X
Burnett etal. (1999)
Metro Toronto, Canada
Period of Study:
1980-1994
Outcomes (ICD 9 codes): Asthma (493);
Obstructive lung disease (490-2,496);
Respiratory Infection (464, 466, 480-7, 494)
Study Design: Time-series
Statistical Analyses: Poisson regression
model with stepwise analysis
Covariates: Long-term trends, season, day of
wk, daily maximum temperature, daily
minimum temperature, daily avg dew point
temperature, daily avg relative humidity
Statistical Package: S-Plus, SAS
Lag: 0,1,2 day s, cumulative
24-h mean: 25.2 ppb,
SD = 9.1,
CV = 36;
IQR = 23
Number of stations: 4
COH;
PM2.5;
PM10-
PM10;
CO; r
SO2;r
O3;r=
r = NR
r = 0.50
.5;r = 0.3
r=0.52
= 0.55
= 0.54
-0.03
Increment: 25.2 ppb (Mean)
7.72 excess daily admissions due to pollution of
all sorts. 40.4% increase; or 3 excess daily
admissions traced to NO2.
Single-pollutant model percent increase
(t statistic)
Asthma: 3.33% (2.37) lag 0
OLD 2.21% (1.07) lag 1
Respiratory infection: 6.89% (5.53), lag 2
Multipollutant model percent increase (SE)
Respiratory infection:
NO2 alone: 4.64 (SE>3)
N02 + S02 + 03 + PM2.5: 4.04 (SE >2)
NO2 + SO2 + O3 + PM10-2.5: 4.56 (SE >3)
NO2 + SO2 + O3 + PM10: 4.16(SE>3)
NO2 + O3 + PM2.5: 4.44 (SE >2)
H
6
o
o
H
O
O
H
W
Lin etal. (2005) Outcomes (ICD 9 codes): Respiratory
Toronto, Canada infections (464,466, 480-487)
Age groups analyzed: 0-14
Period of Study: N: 6,782
1998-2001 Study Design: Bidirectional Case-Crossover
Statistical Analyses: Conditional logistic
regression
Covariates: temperature, dew point
temperature,
Statistical Package: SAS v 8.2
Lag: 1-7 day exposure averages
24-h avg: 24.54
(7.56) ppb
Range: 9.2-53.75
25th: 18.75
50th: 24.00
75th: 29.33
Number of monitors: 7
CO, r = 0.20 Increment: 10.6 ppb (IQR)
SO2,r = 0.61 All children:
O3,r = 0 NO2 alone: 1.20 [1.08, 1.34] lag 0-5
PM10,r=0.54 N02 + PM2.5 + PM10-2.5: 1.13 [0.97, 1.31] lag 0-5
PM2.5, r = 0.48 Boys:
PMio-2.5, r = 0.40 NO2alone: 1.13 [0.98, 1.29] lag 0-5
NO2 + PM2.5 + PM10-2.5: 1.00 [0.83, 1.21] lag 0-5
Girls:
NO2 alone: 1.28 [1.09-1.50] lag 0-5
N02 + PM2.5 + PM10-2.5: 1.31 [1.05, 1.63] lag 0-5
O
HH
H
W
-------
O
to
O
O
oo
TABLE AX6.3-3 (cont'd). RESPIRATORY HEALTH EFFECTS OF OXIDES OF NITROGEN:
HOSPITAL ADMISSIONS
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels &
Monitoring Stations
Copollutants & Effects: Relative Risk or Percent Change
Correlations & Confidence Intervals (95%)
CANADA (cont'd)
X
Oi
-k
Oi
H
6
o
o
H
O
O
H
W
Burnett* et al. (2001)
Toronto, Canada
Period of Study:
1980-1994
Fung et al. (2007)
Ontario, Canada
Period of Study:
1996-2000
Outcomes (ICD 9 codes): Croup (464.4),
pneumonia (480-486), asthma (493), acute
bronchitis/bronchiolitis (466)
Age groups analyzed: <2 yrs
Study Design: Time-series
Statistical Analyses: Poisson regression with
GAM
Covariates: temporal trend, day of wk,
temperature, relative humidity
Statistical Package: S-Plus
Lag: 0-5 days
Outcomes (ICD 9 codes): All respiratory
(460-519)
Age groups analyzed: 0-4, 5-14, 15-19, 20-24,
25-54, 55-64, 65-74, 75+
Study Design:
Statistical Analyses:
Covariates:
Statistical Package:
Lag:
1-h max NO2 (ppb)
Mean: 44.1
CV: 33
5th: 25
25th: 35
50th: 42
75th: 52
95th: 70
99th: 86
100th: 146
Number of stations: 4
London SO2
Mean: 18.10 ppb (7.86) O3
Range: 0-53 CO
Windsor
Mean: 23.50 ppb (7.59)
Range: 6-50
Sarnia
Mean: 16.85 ppb (8.13)
Range: 0-52
O3;r=0.52
SO2
CO
PM2.5
PMio-2.5
Increment: NR
All respiratory admissions:
Single-pollutant:
Percent increase: 20.2 (t = 3.43) lag 0-1
Multipollutant (adjusted for O3):
Percent increase: 7.1 (t = 1.09) lag 0-1
Not reported
O
HH
H
W
-------
O
to
O
O
oo
TABLE AX6.3-3 (cont'd). RESPIRATORY HEALTH EFFECTS OF OXIDES OF NITROGEN:
HOSPITAL ADMISSIONS
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels &
Monitoring Stations
Copollutants & Effects: Relative Risk or Percent Change &
Correlations Confidence Intervals (95%)
CANADA (cont'd)
X
Luginaah et al. (2005)
Windsor, ON, Canada
Period of Study:
4/1/95-12/31/00
Outcomes (ICD 9 codes): Respiratory
admissions (460-519)
Age groups analyzed: 0-14,15-64, 65+, all
ages
Study Design: (1) Time-series and (2) case-
crossover
N: 4,214
# of Hospitals: 4
Statistical Analyses: (1) Poisson regression,
GAM with natural splines (stricter criteria),
(2) conditional logistic regression with Cox
proportional hazards model
Covariates: Temperature, humidity, change in
barometric pressure, day of wk
Statistical Package: S-Plus
Lag: 1,2,3 days
NO2 mean 1-h max:
38.9 ppb,
SD = 12.3;
IQR: 16
Number of stations: 4
SO2; r = 0.22
CO; r= 0.38
PM10;r=0.33
COH; r = 0.49
03; r = 0.26
TRS;r=0.06
Increment: 16 ppb (IQR)
Time-series, females; males
All ages, lag 1 1.035 [0.971, 1.104]; 0.944
[0.886,1.006]
0-14 yrs, lag 2 1.114 [0.994, 1.248]; 0.955
[0.866,1.004]
15-65 yr, lag 3 1.121 [0.978,1.285], 1.012
[0.841,1.216]
65+yr, lag 1 1.020 [0.930, 1.119]; 0.9196
[0.832,1.016]
Case-crossover, females; males
All ages, lag 1 1.078 [0.995, 1.168]; 0.957
[0.883,1.036]
0-14 yrs, lag 2 1.189 [1.002, 1.411]; 0.933
[0.810,1.074]
15-65 yr, lag 3 1.114 [0.915,1.356]; 0.972
[0.744,1.268]
65+yr,lag 1 1.081 [0.964, 1.212]; 0.915
[0.810,1.034]
H
6
o
o
H
O
O
H
W
O
HH
H
W
-------
O
to
O
O
oo
TABLE AX6.3-3 (cont'd). RESPIRATORY HEALTH EFFECTS OF OXIDES OF NITROGEN:
HOSPITAL ADMISSIONS
X
OO
H
6
o
o
H
O
O
H
W
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels &
Monitoring Stations
Copollutants &
Correlations
Effects: Relative Risk or Percent Change &
Confidence Intervals (95%)
AUSTRALIA/NEW ZEALAND
Simpson et at (2005a)
Multi-city study,
Australia (Sydney,
Melbourne, Brisbane,
Perth)
Period of Study:
1996-1999
Outcomes (ICD 9/ICD 10): All respiratory
(460-5 19/JOO-J99 excluding J95.4-J95.9,
R09.1, R09.8), asthma (493/J45, J46, J44.8),
COPD (490-492, 494-496/J40-J44, J47, J67),
pneumonia with bronchitis (466, 480-486/J12-
17, J18.0, J18.1, J18.8, J18.9, J20, J21)
Age groups analyzed: 15-64 (asthma), 65+ (all
respiratory, COPD, asthma, pneumonia with
bronchitis)
Study Design: Time-series
Statistical Analyses: Followed APHEA2
protocol: (1) Single city: (a) GAM with default
and more stringent criteria, (b) GLM with
default and more stringent criteria, (c) penalized
spline models. (2) Multicity meta analysis:
random effects meta-analysis
Covariates: Temperature, relative humidity, day
of wk, holiday, influenza epidemic,
brushfire/controlled bum
Statistical Package: S-Plus, R
Lag: 0,1, 2 days
1 h max NO2 ppb
(range)
Brisbane: 24.1
(2.1,63.3)
Sydney: 23.7
(6.5, 59.4)
Melbourne: 23.7
(4.4, 66.7)
Perth: 16.3(1.9,41.0)
Brisbane:
O3;r=0.15
BSP;r=0.50
Melbourne:
03;r=0.30
BSP; r = 0.29
Sidney ;
O3; r = 0.24
BSP; r= 0.54
Perth:
03; r = 0.28
BSP; r = 0.62
Increment: 1 ppb
Respiratory
>65 yrs 1.0027 [1.0015, 1.0039] lag 0-1
COPD and Asthma
>65 yrs 1.0020 [1.0003, 1.0037] lag 0-1
Pneumonia and Acute Bronchitis
>65 yrs 1.0030 [1.0011, 1.0048] lag 0-1
Multipollutant Model
Respiratory >65 yrs
N02 Alone: 1.0027 [1.0015,1.0039] lagO-1
NO2 + BSP: 1.0023 [1.0009, 1.0038] lag 0-1
N02 + 03: 1.0028 [1.0016, 1.0040] lag 0-1
GAM results from S-Plus and R similar to one
another, but different than results from GLM.
GAM results from S-Plus presented.
O
HH
H
W
-------
O
to
O
O
oo
TABLE AX6.3-3 (cont'd). RESPIRATORY HEALTH EFFECTS OF OXIDES OF NITROGEN:
HOSPITAL ADMISSIONS
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels &
Monitoring Stations
Copollutants & Effects: Relative Risk or Percent Change
Correlations & Confidence Intervals (95%)
AUSTRALIA/NEW ZEALAND (cont'd)
X
vo
H
6
o
o
H
O
O
H
W
Barnett et al. (2005)
Multicity, Australia/
New Zealand;
(Auckland, Brisbane,
Canberra, Christchurch,
Melbourne, Perth,
Sydney)
Period of Study:
1998-2001
Outcomes (ICD 9/ICD 10): All respiratory
(460-519/JOO-J99 excluding J95.4-J95.9,
RO9.1, RO9.8), asthma (493/J45, J46,
J44.8), COPD (490-492, 494-496/J40-J44,
J47, J67), pneumonia with bronchitis (466,
480-486/J12-17, J18.0, J18.1, J18.8, J18.9,
J20,J21)
Age groups analyzed: 0, 1-4, 5-14
Study Design: Case-crossover
Statistical Analyses: Conditional logistic
regression, random effects meta-analysis
Covariates: Temperature, current-previous
day temperature, relative humidity,
pressure, extremes of hot and cold, day of
wk, holiday, day after holiday
Season: Cool, May-Oct; Warm, Nov-Apr
Statistical Package: SAS
Lag: 0-1 days
24-h avg (ppb) (range):
Auckland 10.2
(1.7,28.9)
Brisbane 7.6 (1.4, 19.1)
Canberra 7.0 (0, 22.5)
Christchurch 7.1
(0.2,24.5)
Melbourne 11.7 (2, 29.5)
Perth 9.0 (2, 23.3)
Sydney 11.5(2.5,24.5)
IQR: 5.1 ppb
Daily 1-h max (range):
Auckland 19.1
(4.2, 86.3)
Brisbane 17.3 (4, 44.1)
Canberra 17.9 (0, 53.7)
Christchurch 15.7
(1.2,54.6)
Melbourne 23.2
(4.4, 62.5)
Perth 21.3 (4.4,48)
Sydney 22.6 (5.2, 51.4)
IQR: 9.0 ppb
BS;r= 0.39, 0.63
PM2.5;r = 0.34, 0.68
PM10;r= 0.21,0.57
CO; r = 0.53, 0.73
SO2;r = 0.15, 0.58
O3;r=-0.15, 0.28
Increment: 5.1 ppb (24 h) or per 9 ppb (1-h
max). (IQR)
24-h avg NO2 (5.1 ppb change)
Pneumonia and acute bronchitis
Oyrs 3.2% [-1.8, 8.4] lag 0-1
1-4 yrs 4.8% [-1.0, 11.0] lag 0-1
5-14 yrs (sample size too small)
Respiratory
Oyrs 3.1% [-1.0, 7.3] lag 0-1
1-4 yrs 2.4% [-0.8, 5.7] lag 0-1
5-14 yrs 5.8% [1.7, 10.1] lag 0-1
Asthma
0 yrs No analysis (poor diagnosis)
1-4 yrs 2.6% [-1.3, 6.6] lag 0-1
5-14 yrs 6.0% [0.2, 12.1] lag 0-1
1 h NO2 maximum (9.0 ppb change)
Pneumonia and acute bronchitis
Oyrs2..8% [-1.8, 7.7] lag 0-1
l-4yrs4.1%[-2.4,11.0]lagO-l
5-14 yrs (sample size too small)
Respiratory
Oyrs2.2% [-1.6, 6.1] lag 0-1
1-4 yrs 2.8% [0.7, 4.9] lag 0-1
5-14 yrs 4.7% [1.6, 7.9] lag 0-1
Asthma
0 yrs No analysis (poor diagnosis)
1-4 yrs 2.5% [-0.2, 5.2] lag 0-1
5-14 yrs 2.6% [-2.2, 7.6] lag 0-1
O
HH
H
W
-------
O
to
O
O
oo
TABLE AX6.3-3 (cont'd). RESPIRATORY HEALTH EFFECTS OF OXIDES OF NITROGEN:
HOSPITAL ADMISSIONS
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels &
Monitoring Stations
Copollutants & Effects: Relative Risk or Percent Change
Correlations & Confidence Intervals (95%)
AUSTRALIA/NEW ZEALAND (cont'd)
X
ON
t^ft
O
H
6
o
o
H
O
O
H
W
O
HH
H
W
Erbas et al. (2005)
Melbourne, Australia
Period of Study : 2000-2001
Hinwood et al. (2006)
Perth, Australia
Period of Study:
1992-1998
Outcomes (ICD 10): Asthma (J45, J46)
Age groups analyzed: 1-15
Study Design: Time-series
N: 8,955
# of Hospitals: 6
Statistical Analyses: Poisson
regression, GAM and GEE
Covariates: Day of wk
Dose-response investigated?: Yes
Statistical Package: NR
Lag: 0,1,2 days
Outcomes (ICD 9): COPD (490-496,
excluding 493); Pneumonia (480-
489.99); Asthma (493)
Age groups analyzed: <15, 65+, all
ages
Study Design: Case-crossover, time-
stratified
Statistical Analyses: Conditional
logistic regression
Covariates: Temperature, change in
temperature, maximum humidity,
holiday, day of wk
Statistical Package: NR
Lag: 0,1,2,3 days or cumulative 0-2
and
0-3 days
l-hmeanNO2: 16.80 ppb,
SD = 8.61;
Range: 2.43,63.00
24-h mean [Std. Dev]
(10th and 90th percentile)
All yr 10.3 [5.0]
(4.4,17.1)
Summer 9.6 [4.8]
(4.3, 15.7)
Winter 11.1 [5.1]
(4.8, 18.0)
Daily 1-hmax
Mean [Std. Dev]
All yr 24.8 [10.1]
(13.3,37.5)
Summer 24.9 [8.9]
(12.4, 39.2)
Winter 24.7 [11.1]
(14.4,35.7)
Number of stations: 3
PM10
03
03;r=-0.06
CO; r= 0.57
BS;r=0.39
PM10
PM2.5
Increment: 90th-1 Oth percentile
Inner Melbourne; increment = 25.54 ppb
RR 0.83 [0.68, 0.98] lag 0
Western Melbourne; increment = 28.86 ppb
RR 1.15 [1.03, 1.27] lag 2
Eastern Melbourne; increment = 17.67 ppb
RR 1.07 [0.93, 1.22] lag 0
South/Southeastern; increment = 17.74 ppb
RR 0.98 [0.79, 1.18] lag 1
Increment: 1 ppb (all values were estimated
from the graphs)
All respiratory NO2 (24 h)
>65yrsOR: 1.005 [1.001, 1.011] lag 1
All ages OR: 1.002 [0.998, 1.004] lag 1
Pneumonia NO2 (24 h)
>65 yrs OR: 1.006 [0.999, 1.014] lag 1
All ages OR: 1.002 [0.998, 1.010] lag 1
COPD N02 (24 h)
>65 yrs OR: 1.004 [0.990, 1.012] lag 2
All ages OR: 1.001 [0.995, 1.010] lag 2
Asthma NO2 (24 h)
0-14 yrs OR: 1.002 [0.998, 1.004] lag 0
>65 yrs OR: 0.996 [0.988, 1.002] lag 0
All ages OR: 1.001 [0.999, 1.003] lag 0
-------
O
to
O
O
oo
TABLE AX6.3-3 (cont'd). RESPIRATORY HEALTH EFFECTS OF OXIDES OF NITROGEN:
HOSPITAL ADMISSIONS
Reference, Study
Location, & Period Outcomes, Design, & Methods
Mean Levels &
Monitoring Stations
Copollutants &
Correlations
Effects: Relative Risk or Percent Change
& Confidence Intervals (95%)
AUSTRALIA/NEW ZEALAND
Morgan etal. (1998a)
Sydney, Australia
Period of Study:
1990-1994
Outcomes (ICD 9): COPD (490-492, 494,
496); Asthma (493)
Age groups analyzed: 1-14, 15-64, 65+, all
ages
Study Design: Time-series
# of hospitals: 27
Statistical Analyses: APHEA protocol,
Poisson regression, GEE
Covariates: Long-term trend, temperature,
dew point, day of wk, holiday
Statistical Package: SAS
Lag: 0,1,2 days and cumulative
24-h daily mean: 15ppb,
SD = 6,Range: 0,52,
IQR: 11, 90-10th
percentile: 17
Mean daily 1-h max: 29 ppb,
SD = 3,
Range: 0, 139,
IQR: 15, 90-10th
percentile: 29
# of stations: 3-14, r = 0.52
24-h avg NO2:
PM(24h);r=0.53
PM(lh);r=0.51
O3;r=-0.9
l-hmaxNO2:
PM(24 h); r = 0.45
PM(lh);r=0.44
O3;r=0.13
X
Increment: 90-1 Oth percentile
24-h avg (17 ppb)
Asthma: 1-14 yrs 3.28% [-1.72, 8.54] lag 0
15-64 yrs 2.29% [-2.97, 7.83] lag 0
COPD: >65 yrs 4.30% [-0.75, 9.61] lag 1
Daily 1-h maximum (29 ppb)
Asthma: 1-14 yrs 5.29% [1.07, 9.68] lag 0
15-64 yrs. 3.18% [-1.53, 8.11] lag 0
COPD: 65+ yrs. 4.60% [-0.17, 9.61] lag 1
Multipollutant model (29 ppb)
Asthma: 1-14 yrs. 5.95% [1.11, 11.02] lag 0
COPD: 65+ yrs. 3.70% [-1.03, 8.66] lag 1
H
6
o
o
H
O
O
H
W
O
HH
H
W
Petroeschevsky et al.
(2001)
Brisbane, Australia
Period of Study:
1987-1994
Days: 2922
Outcomes (ICD 9): All respiratory (460-519);
Asthma (493)
Age groups analyzed: 0-4, 5-14, 15-64, 65+,
all ages
Study Design: Timeseries
N: 33,710 (13,246 = asthma)
Statistical Analyses: APHEA protocol,
Poisson regression, GEE
Covariates: Temperature, humidity, season,
infectious disease, day of wk, holiday
Season: Summer, Autumn, Winter, Spring,
Allyr
Dose-response investigated?: Yes
Statistical Package: SAS
Lag: Single: 1,2,3 day
Mean (range) 24-h avg:
Cumulative: 0-2,0-4
Overall: 139(12,497)
Summer: 97(20,331)
Autumn: 129(33,319)
Winter: 179(12,454)
Spring: 153(35,497)
Mean (range) 1-h max
Overall: 282 (35, 1558)
Summer: 206(35,580)
Autumn: 256(70,585)
Winter: 354(35,805)
Spring: 321 (35, 1558)
# of stations: 3,
r = 0.43, 0.53
BSP
03
S02
Increment: 10 ppb
Respiratory (1-h max):
0-4 yrs 1.015 [0.996, 1.035] lag 3
5-14 yrs 0.985 [0.950, 1.021] lag 0
All ages 0.989 [0.977, 1.002] lag 1
Respiratory (24-h avg):
15-64 yrs 1.027 [0.984, 1.071] lag 0
>65 yrs 0.903 [0.851, 0.959] lag 5
Asthma (1-h max):
0-4 yrs 0.975 [0.947, 1.004] lagO
5-64 yrs 0.983 [0.949, 1.018] lag 1
All ages 0.962 [0.936, 0.989] lag 0-2
-------
O
to
O
O
oo
X
ON
t^ft
to
TABLE AX6.3-3 (cont'd). RESPIRATORY HEALTH EFFECTS OF OXIDES OF NITROGEN:
HOSPITAL ADMISSIONS
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels & Copollutants & Effects: Relative Risk or Percent Change
Monitoring Stations Correlations & Confidence Intervals (95%)
EUROPE
Anderson et al. (1997)
Multicity, Europe
(Amsterdam,
Barcelona, London,
Paris, Rotterdam)
Period of study:
1977- 1989 for
Amsterdam and
Rotterdam
1986- 1992 for
Barcelona
1987-1991 for London
1980-1989 for Milan
1987-1 992 for Paris
Outcomes (ICD 9): COPD - unspecified
bronchitis (490), chronic bronchitis (491),
emphysema (492), chronic airways
obstruction (496)
Study Design: Time-series
Statistical Analyses: APHEA protocol,
Poisson regression, meta-analysis
Covariates: Trend, season, day of wk,
holiday, influenza, temperature, humidity
Season: Cool, Oct-Mar; Warm, Apr-Sep
Statistical Package: NR
Lag: 0,1,2 days and 0-3 cumulative
24-h all yr avg: (ug/m3) SO2
Amsterdam: 50 BS
Barcelona: 53 TSP
London: 67 O3
Paris: 42
Rotterdam: 52
1-hmax:
Amsterdam: 75
Barcelona: 93
London: 67
Paris' 64
Rotterdam: 78
Increment: 50 ug/m
Meta-analytic results - Weighted mean
values from 6 cities
COPD- Warm season
24 h 1.03 [1.00, 1.06] lag 1
1 h 1.02 [1.00, 1.05] lag 1
COPD-Cool season
24 h 1.01 [0.99,1.03]
Ih 1.02 [0.99, 1.05]
COPD-A11 Year
24 hr 1.019 [1.002, 1.047] lag
o A i i no,c n nn/i i n*>,cn i —
; 1
. n •?
cumulative
1 hr 1.013 [1.003, 1.022] lag 1
1 hr 1.014 [0.976, 1.054] lag 0-3, cumulative
H
6
o
o
H
O
O
H
W
O
HH
H
W
Atkinson etal. (2001)
Multicity, Europe
(Barcelona,
Birmingham, London,
Milan, Netherlands,
Paris, Rome,
Stockholm)
Period of study:
1998-1997
Outcomes (ICD 9): Asthma (493), COPD
(490-496), All respiratory (460-519)
Study Design: Time-series
Statistical Analyses: APHEA protocol,
Poisson regression, meta-analysis
Covariates: Season, temperature, humidity,
holiday, influenza
Statistical Package: NR
Lag: NR
1-hmax of NO2 (ug/m3)
Barcelona: 94.4
Birmingham: 75.8
London: 95.9
Milan: 147.0
Netherlands: 50.1
Paris: 87.2
Rome: 139.7
Stockholm: 35.6
S02, 03, CO, BS
PM10;r =
Barcelona: 0.48
B'gham: 0.68
London: 0.70
Milan: 0.72
Netherlands: 0.64
Paris: 0.44
Rome: 0.32
Stockholm: 0.30
Increment: 10 ug/m3 for PM10; change in
NO2 not described.
Asthma, 0 to 14 yrs:
ForPM10: 1.2% [0.2, 2.3]
ForPM10 + N02: 0.1 [-0.8,1.0]
Asthma, 15 to 64 yrs:
ForPM10: 1.1% [0.3, 1.8]
ForPM10 + NO2: 0.4 [-0.5, 1.3]
COPD + Asthma, >65 yrs
ForPM10: 1.0% [0.4, 1.5]
ForPM10 + NO2: 0.8 [-0.6, 2.1]
All Respiratory, >65 yrs of age
ForPM10: 0.9% [0.6, 1.3]
ForPM10 + NO2: 0.7 [-0.3, 1.7]
-------
O
to
O
O
oo
TABLE AX6.3-3 (cont'd). RESPIRATORY HEALTH EFFECTS OF OXIDES OF NITROGEN:
HOSPITAL ADMISSIONS
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels &
Monitoring Stations
Copollutants & Effects: Relative Risk or Percent Change
Correlations & Confidence Intervals (95%)
EUROPE (cont'd)
Sunyeretal. (1997)
Multicity, Europe
(Barcelona, Helsinki,
Paris, London)
Period of Study:
1986-1992
Outcomes (ICD 9): Asthma (493)
Age groups analyzed: <15,15-64
Study Design: Time-series
Statistical Analyses: APHEA protocol,
Poisson regression, GEE; meta-analysis
Covariates: Humidity, temperature, influenza,
soybean, long-term trend, season, day of wk
Season: Cool, Oct-Mar; Warm:
Apr-Sep
Statistical Package: NR
Lag: 0,1,2,3 and cumulative 1-3
24-h median (range) (ug/m )
Barcelona: 53 (5, 142)
Helsinki: 35(9,78)
London: 69(27,347)
Paris: 42 (12, 157)
# of stations:
Barcelona: 3
London: 2
Paris: 4
Helsinki: 8
SO2
black smoke
03
Increment: 50 ug/m3 of 24-h avg for all
cities combined
Asthma
15-64yrs
1.029 [1.003, 1.055] lag 0-1
1.038 [1.008-1.068] lag 0-3, cumulative
<15 yrs
1.026 [1.006, 1.049] lag 2
1.037 [1.004, 1.067] lag 0-3, cumulative
1.080 [1.025,1.140] -Winteronly
X
Two-pollutant models:
NO2/Black smoke
15-64 yrs
1.055 [1.005, 1.109] lag 0-1
15-64 yrs 1.088 [1.025,1.155] cumulative 0-
3
<15 yrs
1.036 [0.956,1.122]
H
6
o
o
H
O
O
H
W
N02/S02
<15yrs
1.034 [0.988,1.082]
O
HH
H
W
-------
O
to
O
O
oo
TABLE AX6.3-3 (cont'd). RESPIRATORY HEALTH EFFECTS OF OXIDES OF NITROGEN:
HOSPITAL ADMISSIONS
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels &
Monitoring Stations
Copollutants & Effects: Relative Risk or Percent Change &
Correlations Confidence Intervals (95%)
EUROPE (cont'd)
X
Schouten et al.
(1996)
Multicity, The
Netherlands
(Amsterdam,
Rotterdam)
Period of Study:
04/01/77-09/30/89
Outcomes (ICD 9): All respiratory (460-519),
COPD (490-2, 494, 496), Asthma (493)
Age groups analyzed: 15-64, 65+, all ages
Study Design: Time-series
Statistical Analyses: APHEA protocol, Poisson
regression
Covariates: Long-term trend, season, influenza,
day of wk, holiday, temperature, humidity
Season: Cool, Nov-Apr; Warm: May-Oct
Statistical Package: NR
Lag: 0,1,2 days; and cumulative 0-1 and 0-3
day lags
24-h avg NO2
Amsterdam
Mean/Med: 50/50 ug/m3
Rotterdam
Mean: 54/52 ug/m3
Daily max 1 h
Amsterdam
Mean/Med: 75/75 ug/m3
Rotterdam
Mean/Med: 82/78 ug/m3
# of stations: 1 per city
SO2
BS
03
Increment: 100 ug/m increment
All respiratory, Amsterdam 24 h mean; 1-h max
15-64 yrs RR 0.890 [0.783, 1.012]; 0.894
[0.821, 0.973] lag 1
>65 yrs RR 1.023 [0.907, 1.154]; 0.996 [0.918,
1.080] lag 2
All respiratory, Rotterdam 24 h mean; 1-h max
(1985-89)
15-64 yrs RR 0.965 [0.833, 1.118]; 1.036
[0.951,1.129] lagl
>65 yrs RR 1.172 [0.990, 1.387]; 1.073 [0.970,
1.186] lag 0
COPD, Amsterdam, 24-h mean,
All ages RR 0.937 [0.818, 1.079] lag 1
Asthma Amsterdam, 24-h mean,
All ages RR 1.062 [0.887, 1.271] lag 2
COPD, Rotterdam 24-h mean
All ages RR 1.051 [0.903, 1.223] lag 2
H
6
o
o
H
O
O
H
W
O
HH
H
W
-------
O
to
O
O
oo
TABLE AX6.3-3 (cont'd). RESPIRATORY HEALTH EFFECTS OF OXIDES OF NITROGEN:
HOSPITAL ADMISSIONS
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels &
Monitoring Stations
Copollutants & Effects: Relative Risk or Percent Change
Correlations & Confidence Intervals (95%)
EUROPE (cont'd)
X
Ponce de Leon et al.
(1996)
London, England
Period of Study:
04/1987-1988;
1991-02/1992
Outcomes (ICD 9): All respiratory (460-519)
Age groups analyzed: 0-14, 15-64, 65+, all
ages
Study Design: Timeseries
N: 19,901
Statistical Analyses: APHEA protocol, Poisson
regression GAM
Covariates: Long-term trend, season, influenza,
day of wk, holiday, temperature, humidity
Season: Cool, Oct-Mar; Warm:
Apr-Sep
Dose-Response Investigated: Yes
Statistical Package: SAS
Lag: 0,1,2 days, 0-3 cumulative avg.
NO224-havg: 37.3 ppb,
Med: 35
SD=13.8
IQR: 14 ppb
1-hmax: 57.4 ppb,
Med: 51
SD = 26.4
IQR: 21 ppb
# of stations: 2
SO2; r = 0.45
BS;r = 0.44
03
Increment: 90th-1 Oth percentile
(24-havg: 27 ppb)
Allyr
All ages 1.0114 [1.006, 1.0222] lag 2
0-14 yrs 1.0104 [0.9943, 1.0267] Iag2
15-64 yr 1.0113 [0.9920, 1.0309] lag 1
>65 yr 1.0216 [1.0049, 1.0386] lag 2
Warm season
All ages 1.0276 [1.0042,1.0515] lag 2
0-14 yrs 1.038 [1.0009, 1.0765] lag 2
15-64 yr 1.0040 [0.9651, 1.0445] lag 1
>65 yr 1.0326 [0.9965, 1.0699] lag 2
Cool season
All ages 1.0060 [0.9943,10177] Iag2
0-14 yrs 1.0027 [0.9855, 1.0202] Iag2
15-64 yr 1.0136 [0.9920, 1.0357] lag 1
>65 yr 1.0174 [0.9994, 1.0358] lag 2
H
6
o
o
H
O
O
H
W
O
HH
H
W
-------
O
to
O
O
oo
TABLE AX6.3-3 (cont'd). RESPIRATORY HEALTH EFFECTS OF OXIDES OF NITROGEN:
HOSPITAL ADMISSIONS
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels &
Monitoring Stations
Copollutants & Effects: Relative Risk or Percent Change
Correlations & Confidence Intervals (95%)
EUROPE (cont'd)
X
Oi
t^ft
Oi
H
6
o
o
H
O
O
H
W
O
HH
H
W
Atkinson et al.
(1999a)
London, England
Period of Study:
1992 to 1994
Days: 1096
Spixetal. (1998)
Multi-city (London,
Amsterdam,
Rotterdam, Paris),
Europe
Period of Study:
1977 and 1991
Outcomes (ICD 9): All respiratory (460-519),
Asthma (493), Asthma + COPD (490-6), Lower
respiratory disease (466, 480-6)
Age groups analyzed: 0-14, 15-64, 65+, all
ages
Study Design: Time-series
N: 165,032
Statistical Analyses: APHEA protocol, Poisson
regression
Covariates: Long-term trend, season, influenza,
day of wk, holiday, temperature, humidity
Season: Cool, Oct-Mar; Warm: Apr-Sep
Dose-Response Investigated?: Yes
Statistical Package: SAS
Lag: 0,1,2 days, 0-1, 0-2, 0-3 cum. avg.
Hospital Admissions
Outcomes (ICD 9 codes): All respiratory
(460-519); Asthma (493)
Age groups analyzed: 15-64,65+
Study Design: Time-series
# of Hospitals:
Statistical Analyses: Poisson regression
following APHEA protocol. Pooled meta-
analysis adjusted for heterogeneity
Covariates: Trend, seasonality, day of wk,
holiday, temperature, humidity, unusual events
(strikes, etc.)
Statistical Package:
Lag: 1 to 3 days
NO2 1-hmean: 50.3ppb,
SD = 17.0,
Range: 22.0, 224.3 ppb, 10th
centile: 34.3, 90th percentile:
70.3
# of stations: 3; r= 0.7, 0.96
03,
CO,
PMio
BS,
S02
NO2 daily mean (ug/m )
London 35
Amsterdam 50
Rotterdam 53
Paris 42
Increment: 36 ppb (90th-10th centile)
All ages
Respiratory 1.64% [0.14, 3.15] lag 1
Asthma
1.80% [-0.77,4.44] lag 0
0-14yrs
Respiratory 1.94% [-0.39, 4.32] lag 2
Asthma
1%[-1.42, 5.77] lag 3
15-64yrs
Respiratory 1.61% [-0.82, 4.09] lag 1
Asthma
5.08% [0.81, 9.53] lag 1
65+ yrs
Respiratory 2.53% [0.58, 4.52] lag 3
Asthma 4.53% [-2.36, 11.91] lag 3
COPD 3.53% [0.64, 6.50] lag 3
Lower Resp. 3.47% [0.08, 6.97] lag 3
SO2, O3, BS, TSP Increment: 50 ug/m3.
All cities, yr round
15-64 yrs RR 1.010 [0.985,1.036]
Warm RR 1.00 [0.96, 1.04]
Cold RR 1.01 [0.98,1.04]
>65 yrs RR 1.019 [0.982, 1.060]
Warm RR 1.02 [0.99, 1.06]
Cold RR 1.00 [0.98,1.03]
-------
O
to
O
O
oo
TABLE AX6.3-3 (cont'd). RESPIRATORY HEALTH EFFECTS OF OXIDES OF NITROGEN:
HOSPITAL ADMISSIONS
X
H
6
o
o
H
O
O
H
W
O
HH
H
W
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels &
Monitoring Stations
Copollutants & Effects: Relative Risk or Percent Change
Correlations & Confidence Intervals (95%)
EUROPE (cont'd)
Wong *etal. (2002)
London England and
Hong Kong
Period of Study:
London: 1992-1994
Hong Kong:
1995-1997
Days: 1,096
Outcomes (ICD 9): All respiratory admissions
(460-5 19); asthma (493)
Age groups analyzed: 15-64, 65+, all ages
Study Design: Time-series
Statistical Analyses: APHEA protocol, Poisson
regression with GAM
Covariates: Long-term trend, season, influenza,
day of wk, holiday, temperature, humidity,
thunderstorms
Season: Cool, Oct-Mar; Warm: Apr-Sep
Dose-Response Investigated?: Yes
Statistical Package: S-Plus
Lag: 0,1,2,3,4 days, 0-1 cum. avg
24hNO2ug/m3
Hong Kong
Mean: 55.9
Warm: 48.1
Cool: 63.8
SD=19.4
Range: 15.3,151.5
10th: 31.8
50th: 53.5
90th: 81.8
London
Mean: 64.3
Warm: 62.6
Cool: 66.1
SD = 20.4
Range: 23.7,255.8
10th: 42.3
50th: 61.2
90th: 88.8
# of stations:
Hong Kong: 7; r = 0.65,
0.90
London: 3;r=0.80
Hong Kong Increment: 10 ug/m
PM10;r=0.82
SO2; r = 0.37 Asthma, 15-64 yrs
O3; r = 0.43 Hong Kong
ER -0.6 [-2. 1,1.0] lag 0-1
London ER -1.3 [-2.6, 0.1] lag 1
PM10; r= 0.68 Warm: ER -0.5 [-2.7, 1.6] lag 0-1
SO2;r=0.71 Cool: ER -0.6 [-2.8, 1.6] lag 0-1
O3;r=-0.29 London
ER1.0 [0.0, 2.1] lag 0-1
ER 1.1 [0.2,2.0] lag 2
Warm: ER 0.6 [-0.8, 2.0] lag 0-1
Cool: ER 1.3 [-0.1, 2.8] lag 0-1
Respiratory 65+ yrs
Hong Kong
ER 1.8 [1.2, 2.4] lag 0-1
ER 1.3 [0.8, 1.8] lag 0
Warm: ER 0.8 [0.1, 1.6] lag 0-1
Cool: ER 3.0 [2. 1,3. 9] lag 0-1
+03: ER 1.6 [1.0, 2.3] lag 0-1
+PM10: ER 1.7 [0.8, 2. 7] lag 0-1
+SO2: ER 1.6 [0.8, 2.4] lag 0-1
T .onHon
ER-0.1 [-0.6, 0.5] lag 0-1
ER 0.9 [0.5, 1.3] lag 3
Warm: ER 0.6 [-0.2, 1.4] lag 0-1
Cool: ER-0.7 [-1.4, 0.0] lag 0-1
+O3: ER -0.1 [-0.5, 0.6] lag 0-1
+PM10: ER-0.4 [-1.2, 0.4] lag 0-1
+SO2: ER-0.2 [-0.9, 0.5] lag 0-1
-------
O
to
O
O
oo
TABLE AX6.3-3 (cont'd). RESPIRATORY HEALTH EFFECTS OF OXIDES OF NITROGEN:
HOSPITAL ADMISSIONS
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels &
Monitoring Stations
Copollutants & Effects: Relative Risk or Percent Change
Correlations & Confidence Intervals (95%)
EUROPE (cont'd)
X
OO
H
6
o
o
H
O
O
H
W
Anderson et al.
(1998)
London, England
Period of Study: Apr
1987-February 1992
Days: 1,782
Outcomes (ICD 9): Asthma (493)
Age groups analyzed: <15, 15-64, 65+
Study Design: Ti
N: 16
Statistical Analyses: APHEA protocol, Poisson
regression
Covariates: Time trends, seasonal cycles, day
of wk, public holidays, influenza epidemics,
temperature, humidity
Season: Cool (Oct-Mar); Warm (Apr-Sep)
Dose-Response Investigated?: Yes
Statistical Package: S
Lag: 0,1,2 days
24-h avg NO2 (ppb)
Mean: 37.2
SD=12.3
Range: 14,182
5th: 22
10th: 25
25th: 30
50th: 36
75th: 42
90th: 50
95th: 58
1 -h max NO2 (ppb)
Mean: 57.2
SD = 23.0
Range: 21,370
5th: 35
10th: 38
25th: 44
50th: 52
75th: 64
90th: 81
95th: 98
Number of stations: 2
03
SO2
BS
Increment: 10 ppb in 24-h NO2
0-14 yrs
Whole yrRR 1.25 [0.3, 2.2] lag 2; RR1.77
[0.39, 3.18] lag 0-3
+ 03 RR 1.13 [-0.10, 2.36] lag 2
+ SO2 RR 0.97 [-0.05, 1.99] lag 2
+ BS RR 2.26 [0.83, 3.71] lag 2
Warm season RR 1.42 [-0.3, 3.17] lag 2;
RR 3.01 [3.8, 5.72] lag 0-3
Cool season RR 1.18 [0.02, 2.35] lag 2;
RR 1.22 [-0.48, 2.96] lag 0-3
15-64 yrs
Whole yrRR 0.95 [-0.26, 2.17] lag 0;
RR 0.99 [-0.36, 3.36] lag 0-1
Warm RR 0.46 [-1.70, 2.67] lag 0;
RR0.05[-2.45,2.61]lagO-l
Cool season RR 1.21 [-0.22, 2.5] lag 0;
RR1.43
[-0.18, 3.06] lag 0-1
65+ yrs
Whole yrRR 2.96 [0.67, 5.31] lag 2;
RR3.14
[-0.04, 6.42] lag 0-3
+ 03 RR 4.51 [1.43, 7.69] lag 2
+ SO2 RR 2.49 [-0.25, 5.31] lag 2
+ BS RR 1.88 [-1.49, 5.36] lag 2
Warm RR 1.89 [-2.41, 6.38] lag 2;
RR-1.76 [-7.27, 4.07] lag 0-3
Cool season RR 3.52 [0.81, 6.30] lag 2;
RR 5.57 [1.85, 9.43] lag 0-3
O
HH
H
W
-------
O
to
O
O
oo
TABLE AX6.3-3 (cont'd). RESPIRATORY HEALTH EFFECTS OF OXIDES OF NITROGEN:
HOSPITAL ADMISSIONS
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels &
Monitoring Stations
Copollutants & Effects: Relative Risk or Percent Change
Correlations & Confidence Intervals (95%)
EUROPE (cont'd)
X
Anderson et al.
(1998) (cont'd)
+ O3 RR 5.14 [0.69, 9.79] lag 2
+ SO2 RR 2.10 [-1.08, 5.39] lag 2
+ BS RR 4.47 [-0.04, 9.19] lag 2
All ages
Whole yrRR 1.25 [0.49,2.02] lag 2;
RR 2.05 [0.96, 3.15] lag 0-3
+ O3 RR 1.08 [0.12, 2.05] lag 2
+ S02 RR 0.99 [0.18, 1.81] lag 2
+ BS RR 1.23 [0.47, 2.00] lag 2
Warm RR 1.15 [-0.25, 2.57] lag 2;
RR 1.54 [-0.54, 3.67] lag 0-3
Cool season RR 1.30 [0.38, 2.23] lag 2;
RR 2.26 [0.94, 3.59] lag 0-3
+ O3 RR 0.50 [-0.79, 1.81] lag 2
+ SO2 RR 1.10 [0.12, 2.08] lag 2
+ BS RR 1.29 [0.37, 2.22] lag 2
H
6
o
o
H
O
O
H
W
Anderson et al.
(2001)
West Midlands
conurbation, United
Kingdom
Period of Study:
10/1994-12/1996
Hospital Admissions:
Outcomes (ICD 9 codes): All respiratory (460-
519), Asthma (493), COPD (490-496,
excluding 493)
Age groups analyzed: 0-14, 15-64, 65+
Study Design: Time-series
Statistical Analyses: Followed APHEA 2
protocol, GAM
Covariates: Season, temperature, humidity,
epidemics, day of wk, holidays
Statistical Package: S-Plus 4.5 Pro
Lag: 0,1,2,3,0-1,0-2,0-3
1-h max avg: 37.2 ppb,
15.1 (SD)
Min: 10.7 ppb
Max: 176.1 ppb
10th: 22.9 ppb
90th: 51.7 ppb
# of monitors: 5
PM10; r = 0.62
PM2.5;r=0.61
PM2.5_10;r=0.25
BS;r = 0.65
SO4;r=0.30
SO2;r=0.52
O3;r=0.08
CO; r= 0.73
Increment: 25.5 ppb (90th- 10th)
All respiratory
All ages 1.7% [-0.2, 3.7] lag 0-1
0-14 yrs 2.3% [-0.6, 5.3] lag 0-1
15-64 yrs 0.0% [-3.7, 3.8] lag 0-1
>65 yrs 1.0% [-1.8, 3.9] lag 0-1
COPD with asthma
0-14 yrs 4.0% [-2.0, 10.2] lag 0-1
15-64 yrs -3.3% [-10.4, 4.4] lag 0-1
>65 yrs 2.5% [-2.1, 7.3] lag 0-1
O
HH
H
W
-------
O
to
O
O
oo
TABLE AX6.3-3 (cont'd). RESPIRATORY HEALTH EFFECTS OF OXIDES OF NITROGEN:
HOSPITAL ADMISSIONS
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels &
Monitoring Stations
Copollutants & Effects: Relative Risk or Percent Change
Correlations & Confidence Intervals (95%)
EUROPE (cont'd)
Prescott et al. (1998)
Edinburgh, United
Kingdom
Period of Study:
10/92-6/95
Outcomes (ICD 9): Pneumonia (480-7), COPD
+ Asthma (490-496)
Age groups analyzed: <65, 65+
Study Design: Time-series
Statistical Analyses: Poisson log linear
regression
Covariates: Trend, seasonal and wkly variation,
temperature, wind speed, day of wk
Lag: 0,1 or 3 day rolling avg
NO2: 26.4 ± 7.0 ppb
Min: 9 ppb
Max: 58 ppb
IQR: 10 ppb
# of Stations: 1
CO
PM10
S02
03
BS
Increment: 10 ppb
Respiratory admissions
>65 yrs
3.1 [-4.6, 11.5] rolling 3-day avg
<65 yrs
-0.2% [-7.5, 7.7] rolling 3-day avg
X
Oi
H
6
o
Thompson et al.
(2001)
Belfast, Northern
Ireland
Period of Study:
1993-1995
Outcomes: Asthma
ICD9: NR
Age groups analyzed: 0-14
Study Design: Time-series
N: 1,095
Number of hospitals: 1
Statistical Analyses: Poisson regression
Covariates: Season, long-term trend,
temperature, day of wk, holidays
Season: Warm (May-Oct), Cold
(Nov-Apr)
Statistical Package: Stata
Lag: 0,1,2,3 days
24-h mean:
Warm: 19.2 (7.9) ppb;
Range: 13-23
Cold: 23.3 (9.0) ppb;
Range: 18-28
S02; r = 0.82
PM10; r = 0.77
CO; r= 0.69
O3;r=-0.62
NOX; r = 0.93
log (NO); r = 0.84
log (CO); r = 0.69
Increment: 10 ppb
All seasons
RR 1.08 [1.03, 1.13] lag 0
RR1.11 [1.05, 1.17] lag 0-1
RR 1.10 [1.04, 1.17] lag 0-2
RR 1.12 [1.03, 1.02] lag 0-3
Warm season
RR 1.14 [1.04, 1.26] lag 0-1
Cold season
RR 1.10 [1.03, 1.17] lag 0-1
NO2 + Benzene
RR 0.99 [0.87, 1.13] lag 0-1
*Model made no allowance for possible
autocorrelation in the data or for extra-
Poisson variation.
O
H
O
O
c
H
W
O
HH
H
W
-------
O
to
O
O
oo
TABLE AX6.3-3 (cont'd). RESPIRATORY HEALTH EFFECTS OF OXIDES OF NITROGEN:
HOSPITAL ADMISSIONS
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels &
Monitoring Stations
Copollutants & Effects: Relative Risk or Percent
Correlations Change & Confidence Intervals (95%)
EUROPE (cont'd)
X
Oi
Hagen et al. (2000)
Drammen, Norway
Period of Study:
1994-1997
Outcomes (ICD 9): All respiratory admissions
(460-519)
Age groups analyzed: All ages
Study Design: Time-series
Number of hospitals: 1
Statistical Analyses: Poisson regression with
GAM (adhered to HEI phase l.B report)
Covariates: Time trends, day of wk, holiday,
influenza, temperature, humidity
Lag: 0,1,2,3 days
NO2 24-h avg (ng/m3): 36.15,
SD=16
IQR: 16.92 ng/m3
# of Stations: 2
PM10;r=0.61 Increment: NO2: 16.92 ng/mj (IQR);
S02;r = 0.58 NO: 29|ig/m3 (IQR)
Benzene; r = 0.31
NO; r = 0.70 Single-pollutant model
O3; r = -0.47 Respiratory disease only
Formaldehyde; NO2: RR 1.058 [0.994, 1.127]
r = 0.68 NO: 1.048 [1.013, 1.084]
Toluene; r = 0.65 All disease
NO2: RR 1.011 [0.988,1.035]
Two-pollutant model with PM10
NO2: 1.044 [0.966, 1.127]
NO: 1.045 [1.007, 1.084]
Three-pollutant model with PM10 + Benzene
N02: 1.015 [0.939, 1.097]
NO: 1.031 [0.986,1.077]
Oftedal et al. (2003) Outcomes (ICD 10): All respiratory admissions Mean: 33.8 |ig/m3
H
6
o
o
H
O
O
H
W
Drammen, Norway
Period of Study:
1994-2000
(JOO-J99)
Age groups analyzed: All ages
Study Design: Time-series
Statistical Analyses: Semi-parametric Poisson
regression, GAM with more stringent criteria
Covariates: Temperature, humidity, influenza
Lag: 2,3 days
SD = 16.2
IQR: 20.8 |ig/m3
PM10
S02
03
Benzene
Formaldehyde
Toluene
Increment: 20.8 ng/m3 (IQR)
All respiratory disease
Single-pollutant model
RR 1.060 [1.017,1.105] lag 3
Two-pollutant model
Adjusted for PM10
RR 1.063 [1.008,1.120]
Adjusted for benzene
RR 1.046 [1.002,1.091]
O
HH
H
W
-------
O
to
O
O
oo
TABLE AX6.3-3 (cont'd). RESPIRATORY HEALTH EFFECTS OF OXIDES OF NITROGEN:
HOSPITAL ADMISSIONS
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels &
Monitoring Stations
Copollutants & Effects: Relative Risk or Percent Change
Correlations & Confidence Intervals (95%)
EUROPE (cont'd)
X
Oi
ON
to
Ponka(1991) Outcomes (ICD 9 codes): Asthma (493)
Helsinki, Finland Age groups analyzed: 0-14; 15-64; >65 yrs
Study Design: Time-series
Period of Study: N: 4,209
1987-1989 Statistical Analyses: Correlations and partial
correlations
Covariates: Minimum temperature
Statistical Package:
Lag: 0-1
24-havg: 38.6 (16.3) ug/m3
Range: 4.0-169.6
Number of Monitors: 4
SO2;r = 0.4516
NO; r = 0.6664
O3;r=-0.2582
TSP;r=0.1962
CO
Correlations between hospital admissions
(HA) for asthma and pollutants and
temperature by ages.
0-14 vrs
HA: -0.0166
Emergency HA: 0.0061
15-64 vrs
HA: 0.1648 p< 0.0001
Emergency HA: 0.1189 p< 0.0001
>65 vrs
HA: 0.1501 p< 0.0001
Emergency HA: 0.1392 p< 0.0001
Partial correlations between admissions for
asthma and SO2 were standardized for
temperature.
HA: 0.1830 p< 0.0001
Emergency HA: 0.1137 p = 0.0004
H
6
o
o
H
O
O
H
W
O
HH
H
W
Ponka and Virtanen
(1994)
Helsinki, Finland
Period of Study:
1987-1989
Days: 1096
Outcomes (ICD 9): Chronic bronchitis and
emphysema (491-492)
Age groups analyzed: <65, >65
Study Design: Time-series
Statistical Analyses: Poisson regression
Covariates: Season, day of wk, yr, influenza,
humidity, temperature
Season: Summer (Jun-Aug), Autumn (Sep-
Nov), Winter (Dec-Feb), Spring (Mar-May)
Lag: 0-7 days
24-hmean: 39 ug/m3 SO2
SD = 16.2; O3
Range: 4, 170 TSP
# of stations: 2
Increment: NR
Chronic bronchitis and emphysema
>65 yrs
RR 0.87 [0.71, 1.07] lag 0
RR 1.07 [0.86, 1.33] lag 1
RR 1.16 [0.93, 1.46] lag 2
RR 1.08 [0.86, 1.35] lag 3
RR 0.94 [0.76, 1.18] lag 4
RR 0.90 [0.72, 1.12] lag 5
RR1.31 [1.03, 1.66] lag 6
RR 0.82 [0.67, 1.01] lag 7
<65 yrs
NR
-------
O
to
O
O
oo
TABLE AX6.3-3 (cont'd). RESPIRATORY HEALTH EFFECTS OF OXIDES OF NITROGEN:
HOSPITAL ADMISSIONS
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels &
Monitoring Stations
Copollutants & Effects: Relative Risk or Percent Change
Correlations & Confidence Intervals (95%)
EUROPE (cont'd)
Ponka and Virtanen,
(1996)
Helsinki, Finland
Period of study:
1987-1989
Hospital Admissions 24-h avg (ug/m ):
Outcomes (ICD 9 codes): Asthma (493) Winter: 38
Age groups analyzed: 0-14,15-64,65+ Spring: 44
Study Design: Time-series Summer: 39
Statistical Analyses: Fall: 34
Covariates: Long-term trend, season, epidemics,
day of wk, holidays, temperature, relative
humidity
Statistical Package:
Lag: 0-2
SO2
03
TSP
No results presented for NO2 because they
were not statistically significant
X
Oi
H
6
o
o
H
O
O
H
W
Rossi etal. (1993)
Oulu, Finland
Period of Study:
10/1/1985-9/30/1986
ED Visits
Outcome(s) (ICD 9): Asthma (493)
Age groups analyzed: 15-85
Study Design: Time-series
N: 232
Statistical Analyses: Pearson's and partial
correlation coefficients and multiple regression
with stepwise discriminate analysis
Covariates: Temperature, humidity
Statistical Package: BMDP software
Lag: 0,1,2,3
24-h mean: 13.4 ug/m
Range: 0-69
1-hrmax:
38.5 ug/m3
Range: 0-154
# of Monitoring Stations: 4
NO2; r = 0.48 TSP; Pearson correlation coefficients
H2S
ED asthma visits and same day
SO2: r = 0.20 p< 0.001 lag 0
Weekly ED asthma visits and same wk SO2:
r = 0.42 p< 0.001
Weekly ED asthma visits and previous wk
SO2: r = 0.58 p< 0.001
Multi-pollutant (NO2; TSP; H2S)
Regression coefficient:
Allyr: p = 0.209, p = 0.034
Winter: p = 0.201, p = 0.014
Summer: p = 0.041, p = 0.714
O
HH
H
W
-------
O
to
O
O
oo
TABLE AX6.3-3 (cont'd). RESPIRATORY HEALTH EFFECTS OF OXIDES OF NITROGEN: HOSPITAL
ADMISSIONS
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels &
Monitoring Stations
Copollutants & Effects: Relative Risk or Percent Change
Correlations & Confidence Intervals (95%)
EUROPE (cont'd)
Andersen et al. Outcomes (ICD 10): chronic bronchitis (J41-
(2007a) 42), emphysema (J43), COPD (J44), asthma
Copenhagen, (J45), status asthmaticus (J46)
Denmark Age groups analyzed: 5-18,65+
Number of hospitals: 9
Period of Study: Study Design: Time-series
1999-2004 Statistical Analyses: Poisson regression with
GAM
Covariates: Temperature, long-term trend,
seasonality, influenza, day of wk, public
holidays, school holidays,
Lag: 0, 1, 2, 3, 4, 5, 0-4, 0-5 days
24-h avg: 12 (5) ppb
Statistical package: R
IQR: 7
25th: 8
75th: 15
PM10; r = 0.42
PM10-biomass; r = 0.41
PM10-Secondary; r =
0.43
PMio-Oil; r = 0.42
PM10-Crustal; r = 0.24
PM10-Sea salt; r=-0.19
PM10-Vehicle; r = 0.65
CO; r= 0.74
Increment: 7 ppb (IQR):
All respiratory disease (65+):
NO2: 1.040 [1.009, 1.072] lag 5 day moving
avg
NO2 + PM10: 1.014 [0.978, 1.051] lag 5
day ma
Asthma (5-18 yrs):
NO2: 1.128 [1.029,1.235] lag 6 day ma
N02+PM10: 1.032 [0.917-1.162] lag 6
day ma
X
Oi
H
6
o
o
H
O
O
H
W
Andersen et al. Outcomes (ICD 10): chronic bronchitis (J41-
(2007b) 42), emphysema (J43), COPD (J44), asthma
Copenhagen, (J45), status asthmaticus (J46)
Denmark Age groups analyzed: 5-18,65+
Number of hospitals: 9
Period of Study: Study Design: Time-series
5/15/2001- Statistical Analyses: Poisson regression with
12/31/2004 GAM
Covariates: Temperature, long-term trend,
seasonality, influenza, day of wk, public
holidays, school holidays,
Lag: 0, 1, 2, 3, 4, 5, 0-4, 0-5 days
24-h avg: 11 (5) ppb
Statistical package: R
IQR: 6
25th: 8
50th: 11
75th: 14
99th: 28
PM10
PM2.5
CO
03
Increment: 6 ppb (IQR):
All respiratory disease (65+):
NO2: 1.06 [1.01, 1.12] lag 0-4 moving avg
N02 + NCtot: 1.06 [0.99, 1.13] lag 0-4 ma
Asthma (5-18 yrs):
N02: 1.04 [0.92, 1.18] lag 0-5 ma
N02+NCtot: 0.97 [0.83-1.14] lag 0-5 ma
O
HH
H
W
-------
O
to
O
O
oo
TABLE AX6.3-3 (cont'd). RESPIRATORY HEALTH EFFECTS OF OXIDES OF NITROGEN:
HOSPITAL ADMISSIONS
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels &
Monitoring Stations
Copollutants & Effects: Relative Risk or Percent Change
Correlations & Confidence Intervals (95%)
EUROPE (cont'd)
X
Oi
H
6
o
o
H
O
O
H
W
Dab* etal. (1996)
Paris, France
Period of Study:
1/1/87-9/30/92
Linares et al. (2006)
Madrid, Spain
Period of Study:
1995-2000
Outcomes (ICD 9): All respiratory (460-519),
Asthma (493), COPD (490-496)
Age groups analyzed: All ages
Study Design: Time-series
Number of hospitals: 27
Statistical Analyses: Poisson regression,
followed APHEA protocol
Covariates: Temperature, relative humidity,
influenza, long-term trend, season, holiday,
medical worker strike
Lag: 0,1,2 days, 0-3 cumulative
Outcomes (ICD 9): All respiratory (460-519),
bronchitis (460-496), pneumonia (480-487)
Age groups analyzed: <10
Study Design: Time-series
Number of hospitals: 1
Statistical Analyses: Poisson regression
Covariates: Temperature, pressure, relative
humidity
Statistical Package: S Plus 2000
Lag:
NO2 24-h avg: 45 ug/m3
5th: 22,99th: 108.3
Daily maximum 1-h
concentration: 73.8 ug/m
5th: 37.5,99th: 202.7
SO2
03
PM13
BS
24-h avg: 64.8 (17.1) ug/m3
Range: 23-144
Number of monitors: 24
PM10;r=0.71
03;r=-0.41
SO2;r = 0.63
Increment: 100 ug/m
All respiratory (1987-1990)
24-h avg NO2: RR 1.043 [0.997, 1.090] lag 0
l-hmaxN02: RR 1.015 [0.993, 1.037] lag 0
Asthma (1987-1992)
24-h avg: RR 1.175 [1.059, 1.304] lag 0-1
1-hmax: RR 1.081 [1.019, 1.148] lag 0-1
COPD
24-h avg: RR 0.974 [0.898, 1.058] lag 2
1-hmax: RR0.961 [0.919, 1.014] lag 2
Qualitative results suggest linear relationship
without threshold for NO2 concentration and
respiratory hospital admissions.
O
HH
H
W
-------
O
to
O
O
oo
TABLE AX6.3-3 (cont'd). RESPIRATORY HEALTH EFFECTS OF OXIDES OF NITROGEN:
HOSPITAL ADMISSIONS
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels &
Monitoring Stations
Copollutants & Effects: Relative Risk or Percent Change &
Correlations Confidence Intervals (95%)
EUROPE (cont'd)
X
ON
ON
Llorcaetal. (2005)
Torrelavega, Spain
Period of Study:
1992-1995
Days: 1,461
Outcomes (ICD 9): All respiratory admissions
(460-519)
Age groups analyzed: All ages
Study Design: Time-series
Number of hospitals: 1
Statistical Analyses: Poisson regression
Covariates: Short and long-term trends
Statistical Package: Stata
Lag: NR
24-havgNO2: 21.3 ug/m3,
SD=16.5
24-havgNO: 12.2 ug/m3,
SD= 15.2
# of Stations: 3
SO2;r = 0.588
NO; r = 0.855
TSP;r=-0.12
SH2;r = 0.545
Increment: 100 ug/m
Single-pollutant model
All cardio-respiratory admissions
N02: RR 1.37 [1.26, 1.49]
NO: RR 1.33 [1.22, 1.46]
Respiratory admissions
NO2: RR 1.54 [1.34, 1.76]
NO: RR1.35 [1.17,1.56]
5-pollutant model
All cardio-respiratory admissions
N02: RR 1.20 [1.05, 1.39]
NO: RR0.93 [0.79,1.09]
Respiratory admissions
N02: RR 1.69 [1.34, 2.13]
NO: RR 0.87 [0.67, 1.13]
H
6
o
o
H
O
O
H
W
Migliaretti and Outcome(s) (ICD 9): Asthma (493)
Cavallo (2004) Age groups analyzed: <4,4-15
Turin, Italy Study Design: Case-Control
Controls: Age matched with other respiratory
Period of Study: disease (ICD9: 460-7, 490-2, 494-6, 500-19)
1997-1999 N: Cases = 734, controls = 25,523
Statistical Analyses: Logistic regression
Covariates: Seasonality, temperature, humidity,
solar radiation
Seasons: Cold: Oct-Mar; Warm: Apr-Sep
Statistical Package: SPSS
Lag: 0-3 days and cumulative
Controls:
Mean: 113.3 ug/m ,
SD=30.5
Cases:
Mean: 117.4 ug/m3,
SD = 29.7
TSP
Increment: 10 ug/m3
<4 yrs 2.8% [0.03, 5.03] lag 1-3 cumulative
4-15 yrs 2.7% [-0.01, 6.06] lag 1-3 cumulative
All ages 2.8% [0.07, 4.09] lag 1-3 cumulative
Two-pollutant model adjusted for TSP
NO2 2.1% [-0.1, 5.6]
O
HH
H
W
-------
O
to
O
O
oo
TABLE AX6.3-3 (cont'd). RESPIRATORY HEALTH EFFECTS OF OXIDES OF NITROGEN:
HOSPITAL ADMISSIONS
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels &
Monitoring Stations
Copollutants & Effects: Relative Risk or Percent Change
Correlations & Confidence Intervals (95%)
EUROPE (cont'd)
X
Oi
H
6
o
o
H
O
O
H
W
Farchi et al. (2006)
Rome, Italy
Period of Study:
11/94-2/95
Outcome(s) (ICD 9): All respiratory conditions
(381-382,460-466, 480-493); acute upper
respiratory tract infections (380-382, 460-465);
lower respiratory tract conditions including
asthma (466,480-493)
Age groups analyzed: 6-7
Study Design: Cohort (SIDRIA)
N: 2,947
Statistical Analyses: Cox regression models,
GAM
Covariates: Gender, paternal education,
paternal smoking
Statistical Package: STATA 8.0
Mean: 46.9 ug/m3 (10.2)
IQR: 17
Range: 24-66
Traffic
Increment: 10 ug/m
All respiratory conditions:
HR: 1.28 [0.98-1.68]
1 st Quartile (24-35 ug/m3): 1.00
2nd Quartile (35-47 ug/m3): 1.06 [0.45-2.53]
3rd Quartile (47-52 ug/m3): 1.57 [0.59-4.13]
4th quartile (52-66 ug/m3): 1.95 [0.81-4.71]
Acute URT infections:
HR: 1.56 [0.96-2.56]
1st Quartile (24-35 ug/m3): 1.00
2nd Quartile (35-47 ug/m3): 0.55 [0.08-3.61]
3rd Quartile (47-52 ug/m3): 1.25 [0.25-6.24]
4th quartile (52-66 ug/m3): 3.04 [0.67-13.79]
Acute LRT infections and asthma:
HR: 1.10 [0.80-1.51]
1st Quartile (24-35 ug/m3): 1.00
2nd Quartile (35-47 ug/m3): 1.34 [0.51-3.21]
3rd Quartile (47-52 ug/m3): 1.58 [0.35-4.10]
4th quartile (52-66 ug/m3): 1.24 [0.64-3.08]
O
HH
H
W
-------
O
to
O
O
oo
TABLE AX6.3-3 (cont'd). RESPIRATORY HEALTH EFFECTS OF OXIDES OF NITROGEN:
HOSPITAL ADMISSIONS
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels &
Monitoring Stations
Copollutants & Effects: Relative Risk or Percent Change
Correlations & Confidence Intervals (95%)
EUROPE (cont'd)
X
Oi
ON
OO
H
6
o
o
H
O
O
H
W
Fusco* etal. (2001)
Rome, Italy
Period of Study:
1/1/95-10/31/97
Outcomes (ICD 9): All respiratory (460-519
excluding 470-478), Asthma (493), COPD
(490-492, 494-496), Respiratory infections
(460-466, 480-486)
Age groups analyzed: 0-14, all ages
Study Design: Time-series
Statistical Analyses: Semi-parametric Poisson
regression with GAM
Covariates: Influenza, day, temperature,
humidity, day of wk, holiday
Season: Warm (Apr-Sep), Cold (Oct-Mar)
Statistical Package: S-Plus 4
Lag: 0-4 days
NO2 24-h avg (ug/m3): 86.7, PM10:
SD=16.2 All yr;r = 0.35
Cold; r = 0.50
IQR: 22.3 ug/m3 Warm; r = 0.25
S02:
# of stations: 5; r= 0.66-0.79 All yr;r = 0.33
Cold; r = 0.40
Warm; r = 0.68
CO:
All yr;r = 0.31
Cold; r= 0.41
Warm; r= 0.59
03:
All yr;r = 0.19
Cold; r= 0.19
Warm; r= 0.13
Increment: 22.3 ug/m (IQR)
All respiratory
All ages: 2.5% [0.9, 4.2] lag 0
0-14 yrs: 4.0% [0.6, 7.5] lag 0
Respiratory infections
All ages: 4.0% [1.6, 6.5] lag 0
0-14 yrs: 4.0% [0.2, 8.0] lag 0
Asthma
All ages: 4.6% [-0.5,10.0] lag 0
0-14 yrs: 10.7% [3.0, 19.0] lag 1
COPD
>65yrs: 2.2% [-0.7, 5.2] lag 0
Multipollutant models
All respiratory (NO2 + CO)
All ages: 0.9% [-0.8,2.8] lag 0
0-14 yrs: 3.3% [-0.2, 6.9] lag 0
Acute infections (NO2 + CO)
All ages: 3.9% [1.3, 6.7] lag 0
0-14 yrs: 2.9% [-1.0, 7.0] lag 0
Asthma (NO2 + CO)
All ages: 1.4% [-3.9, 7.1] lag 0
0-14 yrs: 8.3% [-0.1,17.4] lag 1
COPD (NO2 + CO)
>65yrs: -1.0%[-4.1, 2.2] lag 0
O
HH
H
W
-------
O
to
O
O
oo
TABLE AX6.3-3 (cont'd). RESPIRATORY HEALTH EFFECTS OF OXIDES OF NITROGEN:
HOSPITAL ADMISSIONS
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels &
Monitoring Stations
Copollutants & Effects: Relative Risk or Percent Change
Correlations & Confidence Intervals (95%)
EUROPE (cont'd)
Pantazopoulou et al.
(1995)
Athens, Greece
Period of Study:
1988
Outcomes: All respiratory admissions
ICD9: NR
Age groups analyzed: All ages
Study Design: Time-series
N: 15,236
Number of hospitals: 14
Statistical Analyses: Multiple linear regression
Covariates: Season, day of wk, holiday,
temperature, relative humidity
Season: Warm (3/22-9/21), Cold (1/1-3/21 and
9/22-12/31)
Lag: NR
NO2 24-h avg
Winter: 94 ug/m ,
SD = 25
5th: 59,50th: 93,95th: 135
Summer: 111 ug/m3,
SD=32
5th: 65,50th: 108,95th: 173
# of stations: 2
CO
BS
Increment: 76 ug/m in winter and 108
ug/m3 in summer (95th-5th)
Respiratory disease admissions
Winter: Percent increase: 24% [6.4,43.5]
Summer: Percent increase: 9.3%
[-14.1,24.4]
LATIN AMERICA
X
Oi
ON
VO
H
6
o
o
H
O
O
H
W
O
HH
H
W
Gouveiaand Outcomes (ICD 9): All respiratory; Pneumonia
Fletcher, (2000a) (480-486); asthma or bronchitis (466, 490, 491,
Sao Paulo, Brazil 493)
Age groups analyzed:
-------
O
to
O
O
oo
X
TABLE AX6.3-3 (cont'd). RESPIRATORY HEALTH EFFECTS OF OXIDES OF NITROGEN:
HOSPITAL ADMISSIONS
Reference, Study
Location, & Period Outcomes, Design, & Methods
Mean Levels &
Monitoring Stations
Copollutants &
Correlations
Effects: Relative Risk or Percent Change
& Confidence Intervals (95%)
LATIN AMERICA (cont'd)
Braga* et al. (1999) Hospital Admissions
Sao Paulo, Brazil Outcomes (ICD 9 codes): All respiratory
(466,480-486,491-492,496)
Period of Study: Age groups analyzed: <1 3 yrs
10/1992-10/1993 Study Design: Time-series
24-havg 174.84(101.38)
ug/m3
Min: 26.0
Max: 668.3
PM10;r=0.53
CO; r= 0.42
S02;r = 0.53
03;r =
Due to problems with NO2 monitors, this
pollutant could not be included in the
analysis.
N: 68,918
# of Hospitals: 112
Statistical Analyses: Multiple linear regression
models (least squares). Also used Poisson
regression techniques. GLM and GAM using
LOESS for smoothing.
Covariates: Season, temperature, humidity, day
ofwk,
Statistical Package: SPSS, S-Plus
Lag: 1,2,3,4,5,6,7 moving avgs
# of monitors: 13
H
6
o
o
H
O
O
H
W
Braga* etal. (2001)
Sao Paulo, Brazil
Period of Study:
1/93-11/97
Outcomes (ICD 9): All respiratory admissions
(460-519)
Age groups analyzed: 0-19, <2, 3-5,
6-13, 14-19
Study Design: Time-series
Statistical Analyses: Poisson regression with
GAM
Covariates: Long-term trend, season,
temperature, relative humidity, day ofwk,
holiday
Statistical Package: S-Plus 4.5
Lag: 0-6 moving avg
NO2mean: 141.4 ug/m,
SD = 71.2
IQR: 80.5 ug/m3
Range: 25,652.1
# of stations: 5-6
PM10; r = 0.62
SO2;r = 0.54
CO; r= 0.58
03;r=0.34
Increment: 80.5 ug/m3 (IQR)
All Respiratory admissions
<2 yrs 9.4% [6.2, 12.6] lag 5
3-5 yrs 1.6% [-6.4, 9.6]
6-13 yrs 2.3% [-5.9, 10.4]
14-19 yrs-3.0% [-15.7, 9.7]
All ages 6.5% [3.3, 9.7]
O
HH
H
W
-------
O
to
O
O
oo
TABLE AX6.3-3 (cont'd). RESPIRATORY HEALTH EFFECTS OF OXIDES OF NITROGEN:
HOSPITAL ADMISSIONS
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels &
Monitoring Stations
Copollutants & Effects: Relative Risk or Percent Change
Correlations & Confidence Intervals (95%)
LATIN AMERICA (cont'd)
Braga*etal. (2001)
Sao Paulo, Brazil
Period of Study:
1/93-11/97
Outcomes (ICD 9): All respiratory admissions
(460-519)
Age groups analyzed: 0-19, <2, 3-5,
6-13, 14-19
Study Design: Time-series
Statistical Analyses: Poisson regression with
GAM
Covariates: Long-term trend, season,
temperature, relative humidity, day of wk,
holiday
Statistical Package: S-Plus 4.5
Lag: 0-6 moving avg
NO2 mean: 141.4 ug/m ,
SD = 71.2
IQR: 80.5 ug/m3
Range: 25,652.1
# of stations: 5-6
PM10; r = 0.62
SO2;r = 0.54
CO; r= 0.58
O3;r=0.34
Increment: 80.5 ug/m3 (IQR)
All Respiratory admissions
<2 yrs 9.4% [6.2, 12.6] lag 5
3-5 yrs 1.6% [-6.4, 9.6]
6-13 yrs 2.3% [-5.9, 10.4]
14-19 yrs-3.0% [-15.7, 9.7]
All ages 6.5% [3.3, 9.7]
X
H
6
o
o
H
O
O
H
W
O
HH
H
W
-------
O
to
O
O
oo
TABLE AX6.3-3 (cont'd). RESPIRATORY HEALTH EFFECTS OF OXIDES OF NITROGEN:
HOSPITAL ADMISSIONS
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels &
Monitoring Stations
Copollutants & Effects: Relative Risk or Percent Change
Correlations & Confidence Intervals (95%)
LATIN AMERICA (cont'd)
X
to
H
6
o
o
H
O
O
H
W
Farhat* et al. (2005)
Sao Paulo, Brazil
Period of Study:
8/96-8/97
Days: 396
Outcomes (ICD 9):
Pneumonia/bronchiopheumonia (480-6), asthma
(493), bronchiolitis (466), Obstructive disease
493,466)
Age groups analyzed: <13
Study Design: Time-series
N: 1,021
Number of hospitals: 1
Statistical Analyses: Poisson regression with
GAM
Covariates: Time, temperature, humidity, day
of wk, season
Statistical package: S-Plus
Lag: 0-7 days, 2,3,4 day moving avg
Mean: 125.3 ug/m3
SD = 5 1 .7
IQR: 65. 04 ug/m3
Range' 42 5 369 5
PM10;r=0.83
SO2;r = 0.66
CO; r= 0.59
O3; r = 0.47
Increment: 65.04 ug/m3 (IQR)
Single-pollutant models (estimated from
graphs)
Lower respiratory tract disease:
N02 alone: ~ 18% [ 13, 24] lag 0-3
NO2 + PM10 16.1% [5.4, 26.8] lag 0-2
N02 + S02 24.7% [18.2, 31.3] lag 0-2
N02 + CO 19.2% [11.8,26.6] lag 0-2
NO2 + O3 16.1% [9.5, 22.7] lag 0-2
Multipollutant model (PM10, SO2, CO, O3)
18.4% [3.4, 33.5] 2 day avg
Pneumonia:
NO2 alone: -17.5% [3, 32.5] lag 0-2
N02 + PM10 8.1% [-11.4, 27.6] lag 0-2
N02 + S02 13.1% [-3.4, 29.7] lag 0-2
NO2 + CO 14.6% [-4.9, 34.1] lag 0-2
NO2 + O3 12.4% [-5.6, 30.4] lag 0-2
Multipollutant model (PM10, SO2, CO, O3)
1.8% [-23.9, 27.6] 2 day avg
Asthma or Bronchiolitis
NO2 alone: 30.5% [9, 56] lag 0-1
NO2 + PM10 47.7% [1.15, 94.2] lag 0-2
N02 + S02 33.1% [5.7, 60.5] lag 0-2
NO2 + CO 28.8% [-0.2, 57.9] lag 0-2
NO2 + O3 28.0% [-1.0, 57.0] lag 0-2
Multipollutant model (PM10, SO2, CO, O3)
39.3% [-14.9, 93.5] 2 day avg
O
HH
H
W
-------
O
to
O
O
oo
TABLE AX6.3-3 (cont'd). RESPIRATORY HEALTH EFFECTS OF OXIDES OF NITROGEN:
HOSPITAL ADMISSIONS
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels &
Monitoring Stations
Copollutants & Effects: Relative Risk or Percent Change
Correlations & Confidence Intervals (95%)
LATIN AMERICA (cont'd)
Arbex et al., 2007
Araraquara, Brazil
Period of Study:
3/2003-7/2004
Outcomes (ICD 10): asthma (J15)
Age groups analyzed: <13
Study Design: Ecological Time-series
N: 1,021
Number of hospitals: 1
Statistical Analyses: Poisson regression with
GLM
Covariates: Long-term trend, weather
Statistical package: S-Plus
Lag: 0-7 days, 2,3,4 day moving avg
TSP (ug/m3)
Increment: 10 ug/m
Asthma hospital admissions:
11.6% [5.4, 17.7] lag 1-5
X
H
6
o
o
H
O
O
H
W
O
HH
H
W
-------
O
to
O
O
oo
X
H
6
o
o
H
O
O
H
W
TABLE AX6.3-3 (cont'd). RESPIRATORY HEALTH EFFECTS OF OXIDES OF NITROGEN:
HOSPITAL ADMISSIONS
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels &
Monitoring Stations
Copollutants &
Correlations
Effects: Relative Risk or Percent
Change & Confidence Intervals
(95%)
ASIA
Lee et al. (2006)
Hong Kong, China
Period of Study:
1997-2002
Outcomes (ICD 9): Asthma (493)
Age groups analyzed: < 18
Study Design: Time-series
N: 26,663
Statistical Analyses: Semi-parametric Poisson
NO2 24-h mean: 64.7 ug/m3,
SD = 20.9
IQR: 27.1 ug/m3
25th: 49.7,75th: 76.8
PM10;r=0.78
PM2.5; r = 0.75
SO2; r = 0.49
03;r=0.35
Increment: 27. 1 ug/m3 (IQR)
Asthma
Single-pollutant model
4.37% [2.5 1,6.271 lag 0
Days: 2,191
Chew etal. (1999)
Singapore
Period of Study:
1990-1994
regression with GAM (similar to APHEA 2)
Covariates: Long-term trend, temperature,
relative humidity, influenza, day of wk, holiday
Statistical package: SAS 8.02
Lag: 0-5 days
Outcome(s) (ICD 9): Asthma (493)
Age groups analyzed: 3-12,13-21
Study Design: Time-series
N: 23,000
# of Hospitals: 2
Statistical Analyses: Linear regression, GLM
Covariates: Variables that were significantly
associated with ER visits were retained in the
model
Statistical Package: SAS/STAT, SAS/ETS 6.0
Lag: l,2daysavgs
# of stations: 9-10, r= 0.53, 0.94,
Mean = 0.78
24-havg: 18.9 ug/m3,
SD=15.0,
Max < 40
# of Stations: 15
5.88% [4.00, 7.70] lag 1
7.19% [5.37, 9.04] lag 2
9.08% [7.26, 10.93] lag 3
7.64% [5.84, 9.48] lag 4
6.40% [4.60, 8.22] lag 5
Multipollutant model - including PM,
SO2, and O3
5.64% [3.21, 8.14] lag 3
Other lags NR
SO2; r = -0.22 Categorical analysis (via ANOVA) p-
O3; r = 0.17 value and Pearson correlation
TSP' r = 0 23 coefficient (r) using continuous data
comparing daily air pollutant levels and
daily number of hospital admissions.
Age Group:
3-12 13-21
LagO r=0.13 r=0.05
p = 0.013p<0.18
Lagl r=0.13 r=0.02
P = 0.02 p = 0.75
Lag 2 r=0.13 r=0.07
p = 0.35 p = 0.012
O
HH
H
W
-------
O
to
O
O
oo
TABLE AX6.3-3 (cont'd). RESPIRATORY HEALTH EFFECTS OF OXIDES OF NITROGEN:
HOSPITAL ADMISSIONS
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels &
Monitoring Stations
Copollutants & Effects: Relative Risk or Percent Change
Correlations & Confidence Intervals (95%)
ASIA (cont'd)
X
Tsai et al. (2006)
Kaohsiung, Taiwan
Period of Study:
1996-2003
Days: 2922
Outcomes (ICD 9): Asthma (493)
Study Design: Case-crossover
N: 17,682
Statistical Analyses: Conditional logistic
regression
Covariates: Temperature, humidity
Season: Warm (>25°C); Cool (<25°C)
Statistical package: SAS
Lag: 0-2 days cumulative
NO224-hmean: 27.20 ppb
IQR: 17 ppb
Range: 4.83,63.40
# of stations: 6
PM10
SO2
03
CO
Increment: 17 ppb (IQR)
Seasonality
Single-pollutant model
>25°C 1.259 [1.111, 1.427] lag 0-2
<25°C 2.119 [1.875, 2.394] lag 0-2
Dual-pollutant model
Adjusted for PM10
>25°C 1.082 [0.913, 1.283] lag 0-2
<25°C 2.105 [1.791, 2.474] lag 0-2
Adjusted for CO
>25°C 0.949 [0.792, 1.137] lag 0-2
<25°C 2.30 [1.915, 2.762] lag 0-2
Adjusted for SO2
>25°C 1.294 [1.128, 1.485] lag 0-2
<25°C 2.627 [2.256, 3.058] lag 0-2
Adjusted for O3
>25°C 1.081 [0.945, 1.238] lag 0-2
<25°C 2.096 [1.851, 2.373] lag 0-2
H
6
o
o
H
O
O
H
W
Chen et al. (2006)
Taiwan
Period of Study:
1/1998-12/2001
Outcomes (ICD 9): Asthma (493)
Age Groups: 0-4, 5-14, 15-44, 45-64, 65+
Study Design: Time-series
N: 126,671
Statistical Analyses: Spearman Rank
Correlations
Covariates:
Season:
Statistical package: SPSS
Lag:
Mean monthly NO2 averaged
across 55 monitors: 37.64
(4.89) ppb
Minimum: 29.52
25th: 33.72
50th: 37.07
75th: 40.63
Maximum: 47.65
PM10; r =
SO2; r =
C0;r =
03;r =
Spearman rank correlations show that
seasonal variations in adult asthma
admissions are significantly correlated with
levels of NO2 (r = 0.423, p = 0.003).
O
HH
H
W
-------
O
to
O
O
oo
TABLE AX6.3-3 (cont'd). RESPIRATORY HEALTH EFFECTS OF OXIDES OF NITROGEN:
HOSPITAL ADMISSIONS
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels &
Monitoring Stations
Copollutants & Effects: Relative Risk or Percent
Correlations Change & Confidence Intervals (95%)
ASIA (cont'd)
X
Oi
-!j
Oi
Lee* et al. (2002)
Seoul, Korea
Period of Study:
12/1/97-12/31/99
Days: 822
Yang et al. (2007)
Taipei, Taiwan
Period of Study:
1996-2003
Outcomes (ICD 10): Asthma (J45 - J46)
Age groups analyzed: <15
Study Design: Time-series
N: 6,436
Statistical Analyses: Poisson regression, log link
with GAM
Covariates: Time, day of wk, temperature, humidity
Season: Spring (Mar-May), Summer (Jun-Aug), Fall
(Sep-Nov), Winter (Dec-Feb)
Statistical package: NR
Lag: 0-2 days cumulative
Outcomes (ICD 9): Asthma (493)
Age groups analyzed: All
Study Design: Case-crossover
N: 25,602
Number of hospitals: 47
Statistical Analyses: Conditional logistic regression
Covariates: Temperature, humidity
Statistical package: SAS
Lag: 0-2
24-h NO2 (ppb)
Mean: 31.5
SD=10.3
5th: 16.0
25th: 23.7
50th: 30.7
75th: 38.3
95th: 48.6
# of stations: 27
24-havg: 30.77 ppb
Range: 3.84-77.97
25th: 25.55
50th: 30.31
75th: 35.60
Number of monitors: 6
SO2; r = 0.72
O3;r=-0.07
CO; r= 0.79
PM10; r = 0.74
SO2
PM10
CO
03
Increment: 14.6 ppb (IQR)
Asthma
NO2: RR 1.15 [1.10, 1.20] lag 0-2
N02 + PM10: RR 1.13 [1.07, 1.19] lag 0-2
N02 + S02: RR 1.20 [1.11, 1.29] lag 0-2
NO2 + O3: RR 1.14 [1.09, 1.20] lag 0-2
N02 + C0: RR 1.12 [1.03, 1.22] lag 0-2
NO2 + O3 + CO + PM10 + SO2: RR 1.098
[1.002,1.202]
Increment: 10.05 ppb (IQR)
NO2 alone:
>25C: 1.178 [1.113, 1.247] lag 0-2
<25C: 1.128, 1.076, 1.1 82] lag 0-2
N02 + PM10:
>25C: 1.328 [1.224, 1.441] lag 0-2
<25C: 1.144 [1.077, 1.215] lag 0-2
H
6
o
o
H
O
O
H
W
O
HH
H
W
NO2 + SO2:
>25C: 1.224 [1.140, 1.314] lag 0-2
<25C: 1.219 [1.150, 1.291] lag 0-2
NO2 + CO:
>25C: 1.084 [0.999, 1.176] lag 0-2
<25C: 1.198 [1.111, 1.291] lag 0-2
NO2 + O3:
>25C: 1.219 [1.142, 1.301] lag 0-2
<25C: 1.156 [1.102, 1.212] lag 0-2
-------
O
to
O
O
oo
TABLE AX6.3-3 (cont'd). RESPIRATORY HEALTH EFFECTS OF OXIDES OF NITROGEN:
HOSPITAL ADMISSIONS
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels &
Monitoring Stations
Copollutants & Effects: Relative Risk or Percent
Correlations Change & Confidence Intervals (95%)
ASIA (cont'd)
X
Yang and Chen
(2007)
Taipei, Taiwan
Period of Study:
1996-2003
Outcomes (ICD 9): COPD (493)
Age groups analyzed:
Study Design: Case-crossover
N: 25,602
Number of hospitals: 47
Statistical Analyses: Conditional logistic regression
Covariates: Temperature, humidity
Statistical package: SAS
Lag: 0-2
24-h avg: 30.77 ppb
Range: 3.84-77.97
25th: 25.55
50th: 30.31
75th: 35.60
Number of monitors: 6
SO2
PM10
CO
03
Increment: 10.05 ppb (IQR)
NO2 alone:
>20C: 1.193 [1.158, 1.230] lag 0-2
<20C: 0.972 [0.922, 1.024] lag 0-2
N02 + PM10:
>20C: 1.183 [1.137, 1.231] lag 0-2
<20 C: 0.920 [0.862, 0.982] lag 0-2
NO2 + SO2:
>20C: 1.302 [1.254, 1.351] lag 0-2
<20 C: 0.895 [0.837, 0.956] lag 0-2
NO2 + CO:
>20C: 1.154 [1.102, 1.208] lag 0-2
<20C: 0.972 [0.892, 1.059] lag 0-2
H
6
o
o
H
O
O
H
W
NO2 + O3:
>20C: 1.163 [1.126, 1.200] lag 0-2
<20C: 0.952 [0.901, 1.006] lag 0-2
O
HH
H
W
-------
g TABLE AX6.3-3 (cont'd). RESPIRATORY HEALTH EFFECTS OF OXIDES OF NITROGEN:
| HOSPITAL ADMISSIONS
.j Reference, Study Outcomes, Design, & Methods Mean Levels & Copollutants & Effects: Relative Risk or Percent
O Location, & Period Monitoring Stations Correlations Change & Confidence Intervals (95%)
00 ASIA (cont'd)
Ko et al. (2007a) Outcomes (ICD 9):
Hong Kong Age groups analyzed:
Period of Study # of hospitals:
Study Design:
Statistical Analyses:
Covariates:
Statistical package:
Lag:
X
oo
H
6
o
o
H
O
O
H
W
O
HH
H
W
-------
O
to
O
O
oo
TABLE AX6.3-3 (cont'd). RESPIRATORY HEALTH EFFECTS OF OXIDES OF NITROGEN:
HOSPITAL ADMISSIONS
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels &
Monitoring Stations
Copollutants & Effects: Relative Risk or Percent
Correlations Change & Confidence Intervals (95%)
ASIA (cont'd)
X
H
6
o
o
H
O
O
H
W
O
HH
H
W
Lee et al. (2006)
Hong Kong, China
Period of Study:
1997-2002
Lee et al. (2007)
Kaohsiung, Taiwan
Period of Study:
1996-2003
Outcomes (ICD 9): asthma (493)
Age groups analyzed: <18
N: 26,663
Study Design: Time-series
Statistical Analyses: Poisson regression with
GAM
Covariates: Temperature, humidity, influenza,
day of wk, holidays
Statistical package: SAS v 8.02
Lag: 0,1,2,3,4,5
Outcomes (ICD 9): COPD (490-492, 494,
496)
Age groups analyzed: All
# of hospitals: 63
N: 25,108
Study Design: Case-crossover
Statistical Analyses: Conditional logistic
regression
Covariates: temperature, humidity
Season: Warm: >25 C, cool: <25 C
Statistical package: SAS v 8.2
Lag: 0-2 cumulative avg
24-havg: 64.7 (20.9) ug/m3
25th: 49.7
50th: 63.5
75th: 76.8
IQR: 27.1
Number of monitors: 10
24-h avg: 27.2 ppb
Range: 4.83-63.40
25th: 18.4
50th: 27.17
75th: 35.40
# of monitors: 6
SO2;r=0.49
PM10; r = 0.78
PM2.5; r = 0.75
O3;r = 0.35
SO2
PM10
CO
03
Increment: 27.1 ug/m (IQR)
LagO: 4.37% [2.51, 6.27]
Lagl: 5.88% [4.00, 7.70]
Lag 2: 7.19% [5.37, 9.04]
Lag 3: 9.08% [7.26, 10.93]
Lag 4: 7.64% [5.84, 9.48]
Lag 5: 6.40% [4.60, 8.22]
NO2 alone: 9.08% [7.26, 10.93] lag 3
NO2 + SO2 + PM10 + PM2 5 + O3: 5.64%
[3.21, 8.14] lag 3
Increment: 17 ppb (IQR)
NO2 alone:
>25C: 1.241 [1.117, 1.379] lag 0-2
<25C: 1.975 [1.785, 2.186] lag 0-2
N02 + PM10:
>25C: 1.083 [0.939, 1.249] lag 0-2
<25C: 1.957 [1.709, 2.241] lag 0-2
NO2 + SO2:
>25C: 1.264 [1.127, 1.418] lag 0-2
<25 C: 2.378 [2.095, 2.700] lag 0-2
NO2 + CO:
>25C: 0.984 [0.848, 1.141] lag 0-2
<25 C: 2.035 [1.746, 2.373] lag 0-2
N02 + 03:
>25C: 1.076 [0.961, 1.205] lag 0-2
<25C: 1.946 [1.755, 2.157] lag 0-2
-------
O
to
O
O
oo
TABLE AX6.3-3 (cont'd). RESPIRATORY HEALTH EFFECTS OF OXIDES OF NITROGEN:
HOSPITAL ADMISSIONS
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels & Monitoring
Stations
Copollutants & Effects: Relative Risk or Percent
Correlations Change & Confidence Intervals (95%)
ASIA (cont'd)
Wong etal. (1999)
Hong Kong, China
Period of Study:
1994-1995
Outcomes (ICD 9): All respiratory admissions
(460-6,471-8, 480-7, 490-6); Asthma (493),
COPD (490-496), Pneumonia (480-7)
Age groups analyzed: 0-4, 5-64, >65, all ages
# of hospitals: 12
Study Design: Time-series
Statistical Analyses: Poisson regression
(followed APHEA protocol)
Covariates: Trend, season, day of wk, holiday,
temperature, humidity
Statistical package: SAS 8.02
Lag: days 0-3 cumulative
Median 24-h NO2: 51.39 ug/m3
Range: 16.41, 122.44
25th: 39.93,75th: 66.50
# of stations: 7,
r = 0.68, 0.89
03
SO2
PM10;r=0.79
Increment =10 ug/m
Overall increase in admissions:
1.020 [1.013, 1.028] lag 0-3
Respiratory Relative Risks (RR)
0-4 yrs: 1.020 [1.010, 1.030] lag 0-3
5-64yrs: 1.023 [1.011, 1.034] lag 0-3
>65yrs: 1.024 [1.014, 1.035] lag 0-3
Cold Season: 1.004 [0.988, 1.020]
N02 + highPM10: 1.009 [0.993, 1.025]
NO2 + highO3: 1.013 [0.999, 1.026]
X
OO
o
H
6
o
o
H
O
O
H
W
O
HH
H
W
Wong etal. (200la)
Hong Kong, China
Period of Study:
1993-1994
Outcomes (ICD 9): Asthma (493)
Age groups analyzed: < 15
N: 1,217
# of hospitals: 1
Study Design: Time-series
Statistical Analyses: Poisson regression
(followed APHEA protocol)
Covariates: Season, temperature, humidity
Season: Summer (Jun-Aug), Autumn
(Sep-Nov), Winter (Dec-Feb), Spring
(Mar-May)
Lag: 0,1,2,3,4,5 days; and cumulative 0-2 and
0-3 days.
24-h avg
NO2 mean: 43.3 ug/m ,
SD = 16.6
Range: 9, 106 ug/m3
Autumn: 51.7(17.6)
Winter: 46.6(15.5)
Spring: 40.7(11.8)
Summer: 32.6(13.7)
# of stations: 9
PM10
SO,
Asthma: 1.026 [1.01, 1.042] lag 0-3
COPD: 1.029 [1.019, 1.040] lag 0-3
Pneumonia: 1.028 [1.015, 1.041] lag 0-3
Increment: 10 ug/m3
Asthma
Allyr: 1. 08 p = 0.001
Autumn: 1. 08 p = 0.017
Winter: NR
Spring: NR
Summer: NR
*Default GAM
+Did not report correction for over-dispersion
NR: Not Reported
APHEA: Air Pollution and Health: A European Approach
-------
O
to
O
O
oo
TABLE AX6.3-4. RESPIRATORY HEALTH EFFECTS OF OXIDES OF NITROGEN:
EMERGENCY DEPARTMENT VISITS
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels of NO2 &
Monitoring Stations
Copollutant Effects and Interpretation:
Correlations Relative Risk & Confidence Intervals (95%)
UNITED STATES
X
ON
OO
Jaffe et al. (2003)
3 cities, Ohio, United
States (Cleveland,
Columbus, Cincinnati)
Period of Study:
7/91-6/96
Outcome (ICD-9): Asthma (493)
Age Groups Analyzed: 5-34
Study Design: Time-series
N: 4,416
Statistical Analyses: Poisson regression
using a standard GAM approach
Covariates: City, day of wk, wk, yr,
minimum temperature, overall trend,
dispersion parameter
Season: Jun to Aug only
Dose-response investigated: Yes
Statistical Package: NR
Lag: 0-3 days
Cincinnati
24-havg: 50 ppb, SD = 15
Cleveland
24-havg: 48 ppb, SD = 16
NO2 was not monitored in
Columbus due to relatively
low levels
Cincinnati:
PM10;r=0.36
SO2;r = 0.07
O3;r=0.60
Cleveland:
PM10; 0.34
SO2;r = 0.28
O3; r = 0.42
No
multipollutant
models were
utilized.
Increment: 10 ppb
Cincinnati: 6% [-1.0, 13] lag 1
Cleveland: 4% [-1,8] lag 1
All cities: 3% [-1.0, 7]
Attributable risk from NO2 increment:
Cincinnati 0.72 (RR 1.06)
Cleveland 0.44 (RR 1.04)
Regression diagnostics for Cincinnati showed
significant linear trend during entire study period
and for each wk (6/1-8/31). No trends observed
for Cleveland.
Regression Models assessing exposure thresholds
showed a possible dose-response for NO2
(percent increase after 40 ppb). No increased risk
until minimum concentration of 40 ppb was
reached.
H
6
o
o
H
O
O
H
W
O
HH
H
W
Norris* etal. (1999)
Seattle, WA, United States
Period of Study:
1995-1996
Outcome (ICD-9): Asthma (493)
Age groups analyzed: <18yrs
Study Design: Time-series
N: 900 ER visits
Statistical Analyses: Semi parametric
Poisson regression using GAM.
Covariates: day of wk, time trends,
temperature, dew point temperature
Dose-response investigated: Yes
Statistical Package: NR
Lag: 0,2 days
24 h: 20.2 ppb, SD = 7.1
IQR: 9 ppb
1-hmax: 34.0 ppb,
SD=11.3
IQR: 12 ppb
CO; r= 0.66
PM; r = 0.66
SO2;r = 0.25
Increment: IQR
24-h avg (9-ppb increment)
RR 0.99 [0.90, 1.08] lag 2
1-h max (12-ppb increment)
RR 1.05 [0.99, 1.12] lag 0
Age and hospital utilization (high and low)
segregation (<5, 5-11, and 12-17 yrs) did not
figure significantly in the association between
emergency room visits and asthma.
-------
O
to
O
O
oo
TABLE AX6.3-4 (cont'd). RESPIRATORY HEALTH EFFECTS OF OXIDES OF NITROGEN:
EMERGENCY DEPARTMENT VISITS
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels of NO2 &
Monitoring Stations
Copollutant Effects and Interpretation:
Correlations Relative Risk & Confidence Intervals (95%)
UNITED STATES (cont'd)
X
OO
to
H
6
o
o
H
O
O
H
W
Lipsettetal. (1997)
Santa Clara County,
California, United States
Period of Study:
1988-1992
Outcome(s): Asthma
ICD-9 Code(s): NR
Age groups analyzed: All
Study Design: Time-series
Statistical Analyses: Poisson
Regression; GEE repeated with GAM
Covariates: Minimum temperature,
day of study, precipitation, hospital,
day of wk, yr, overdispersion
parameter
Season: Winters only
Statistical Package: SAS, S Plus,
Stata
Lag: 0-5 days
NO2 1-hmean: 69 ppb,
SD = 28
Range: 29, 150 ppb
PM10; r = 0.82
COH;r = 0.8
No multipollutant
model due to high
correlation
between pollutants
Same day NO2 was associated with ER visits for
asthma ((3 = 0.013, p = 0.024)
Absence of association between lagged or
multiday specifications of NO2 and asthma ER
visits (data not shown) suggest that same day
association may be artifact of covariation with
PM10.
O
HH
H
W
-------
O
to
O
O
oo
TABLE AX6.3-4 (cont'd). RESPIRATORY HEALTH EFFECTS OF OXIDES OF NITROGEN:
EMERGENCY DEPARTMENT VISITS
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels of NO2 &
Monitoring Stations
Copollutant Effects and Interpretation:
Correlations Relative Risk & Confidence Intervals (95%)
UNITED STATES (cont'd)
X
ON
OO
H
6
o
o
H
O
O
H
W
Peel et al. (2005)
Atlanta, GA, United States
Period of Study:
1/93-8/2000
Tolbert et al. (2000)
Atlanta, GA, United States
Period of Study:
1993-1995
Outcome(s) (ICD-9): All respiratory
(460-6, 477, 480-6, 480-6, 490-3, 496);
Asthma (493); COPD (491-2, 496);
Pneumonia (480-486); Upper Respiratory
Infection (460-6,477)
Age groups analyzed: All, 2-18
Study Design: Time-series
N: 484,830
# of Hospitals: 31
Statistical Analyses: Poisson Regression,
GEE, GLM, and GAM (data not shown
for GAM) Covariates: day of wk, hospital
entry/exit, holidays, time trend; season,
temperature, dew point temperature
Statistical Package: SAS, S-Plus
Lag: 0 to 7 days. 3-day moving avgs.
Outcome(s) (ICD-9): Asthma (493),
wheezing (786.09), Reactive airways
disease (RADS) (519.1)
Age groups analyzed: 0-16; 2-5, 610,
11-16
Study Design: Case-Control
N: 5,934
Statistical Analyses: Ecological GEE
analysis (Poisson model with logit link)
and logistic regression
Covariates: Day of wk, day of summer,
yr, interaction of day of summer and yr
Season: Summers only
Statistical Package: SAS
Lag: 1 day (a priori)
1-hmax: 45.9ppb,
SD= 17.3
NOX 1-h max continuous
Mean: 81.7ppb,
SD = 53.8
Range = 5.35, 306
Number of stations: 2
O3; r = 0.42
SO2;r = 0.34
CO; r= 0.68
PM10; r = 0.46
Evaluated
multipollutant
models (data not
shown)
PM10; r = 0.44
O3;r=0.51
Increment: 20 ppb
All respiratory
RR 1.016 [1.006, 1.027] lag 0-2, 3-day moving avg
Upper Respiratory Infection (URI)
RR 1.019 [1.006, 1.031] lag 0-2, 3-day moving avg
Asthma
All: 1.014 [0.997, 1.030] lag 0-2, 3-day moving avg
2-18: 1.027 [1.005, 1.050] lag 0-2, 3-day moving avg
Pneumonia
RR 1.000 [0.983, 1.019] lag 0-2, 3-day moving avg
COPD
RR 1.035 [1.006, 1.065] lag 0-2, 3-day moving avg
Increment: 50 ppb
Age 0-16:
RR 1.012 [0.987,1.039] lag 1
O
HH
H
W
-------
O
to
O
O
oo
TABLE AX6.3-4 (cont'd). RESPIRATORY HEALTH EFFECTS OF OXIDES OF NITROGEN:
EMERGENCY DEPARTMENT VISITS
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels of NO2 &
Monitoring Stations
Copollutant Effects and Interpretation:
Correlations Relative Risk & Confidence Intervals (95%)
UNITED STATES (cont'd)
X
ON
OO
H
6
o
o
H
O
O
H
W
Tolbert et al. (2007)
Atlanta, GA
Period of Study:
1993-2004
Cassino* et al. (1999)
New York City, NY
United States
Period of Study:
1/1989-12/1993
Outcome(s) (ICD-9): Combined
respiratory diseases (493, 786.07, 786.09,
491, 492, 496, 460-465, 477, 480-486,
466.1,466.11,466.19)
Age groups analyzed: All
Study Design: Time-series
N: 1,072,429
Number of hospitals: 41
Statistical Analyses: Poisson regression
with GLM
Covariates: Day of wk, season, hospital,
holiday, temperature, dew point)
Statistical Package: SASv. 9.1
Lag: 0-2 (a priori)
Outcome(s) (ICD-9): Asthma (493);
COPD (496), bronchitis (490),
emphysema (492), bronchiectasis (494)
Study Design: Time-series
N: 1,115
# of Hospitals: 11
Statistical Analyses: Time-series
regression, Poisson regression with GLM
and GAM; Linear regression, Logistic
regression with GEE
Covariates: Season, trend, day of wk,
temperature, humidity
Statistical Package: S Plus and SAS
Lag: 0-3 days
1-hmax: 43.2ppb
Range: 1.0-181.0
10th: 22.0
25th: 31.0
50th: 41.0
75th: 54.0
90th: 66.0
24-h avg NO2:
Mean: 45.0ppb
Median: 43 ppb
10%31ppb
25% 37 ppb
75% 53 ppb
90% 63 ppb
PM10;r=0.53
O3; r = 0.44
S02;r = 0.36
CO; r= 0.70
PM2.5; r = 0.47
Increment: 23 ppb (IQR)
RR 1.015 [1.004, 1.025] lag 0-2
10-2.5 .
PM2.5sulfate;r = 0.14
PM2 5EC; r = 0.64
PM2.5OC; r = 0.62
OHC; r = 0.24
03
CO
S02
Increment: 15 ppb (IQR)
RR 0.97 [0.85, 1.09] lag 0
RR 1.04 [0.92, 1.18] lag 1
RR 1.06 [0.94, 1.2] lag 2
RR 0.97 [0.86, 1.09] lag 3
O
HH
H
W
-------
O
to
O
O
oo
TABLE AX6.3-4 (cont'd). RESPIRATORY HEALTH EFFECTS OF OXIDES OF NITROGEN:
EMERGENCY DEPARTMENT VISITS
X
ON
OO
H
6
o
o
H
O
O
H
W
O
HH
H
W
Reference, Study Mean Levels of NO2 & Copollutant
Location, & Period Outcomes, Design, & Methods Monitoring Stations Correlations
Effects and Interpretation:
Relative Risk & Confidence Intervals (95%)
CANADA
Bates etal. (1990) Outcome(s) (ICD 9): Asthma (493); May-Oct May-Oct.
Vancouver Region, Pneumonia (480-486); SO2l-hmax: O3;r=0.35
BC, Canada Chronic bronchitis (491,492,496); Range: 0.0337-0.0458 ppm SO2;r = 0.67
Other respiratory (466) CoH r = 0 53
Period of Study: 7/1/1984- Age groups analyzed: All; 15-60 Nov-Apr SO4;r = 0.50
10/31/1986 Study Design: Range: 0.0364-0.0455 ppm
# of Hospitals: 9 Nov-Apr
Statistical Analyses: Pearson correlation Number of stations: 11 O3;r=0.31
coefficients were calculated between §Q . r = Q g j
asthma visits and environmental variables CoR- r = 0 69
Season: Warm (May-Oct); SO''r = 049
Cool (Nov-Apr) 4'
Covariates: NR
Statistical Package: NR
Lag: 0,1,2
Correlation Coefficients:
Warm Season (May-Oct)
Asthma (1-14 yrs)
NS lag 0
NS lag 1
NS lag 2
Respiratory (1-14)
NS lag 0
NS lag 1
NS lag 2
Total (1-14)
NS lag 0
NS lag 1
NS lag 2
Asthma (15-60 yrs)
NS lag 0
NS lag 1
NS lag 2
Respiratory (15-60 yrs)
r = 0.120 lag Op< 0.01
NS lag 1
NS lag 2
Total (15-60 yrs)
NS lag 0
NS lag 1
NS lag 2
Asthma (61+ yrs)
NS lag 0
NS lag 1
NS lag 2
Respiratory (61+ yrs)
NS lag 0
NS lag 1
NS lag 2
-------
O
to
O
O
oo
TABLE AX6.3-4 (cont'd). RESPIRATORY HEALTH EFFECTS OF OXIDES OF NITROGEN:
EMERGENCY DEPARTMENT VISITS
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels of NO2 &
Monitoring Stations
Copollutant Effects and Interpretation:
Correlations Relative Risk & Confidence Intervals (95%)
CANADA (cont'd)
Bates etal. (1990)
(cont'd)
Total(61+yrs)
NS lag 0
NS lag 1
NS lag 2
X
Oi
OO
Oi
H
6
o
o
H
O
O
H
W
Cool Season (Nov - Apr)
Asthma (1-14 yrs)
NS lag 0
NS lag 1
NS lag 2
Respiratory (1-14)
NS lag 0
NS lag 1
NS lag 2
Total (1-14)
NS lag 0
NS lag 1
NS lag 2
Asthma (15-60 yrs)
NS lag 0
NS lag 1
NS lag 2
Respiratory (15-60 yrs)
r = 0.120 lag Op< 0.01
NS lag 1
NS lag 2
Total (15-60 yrs)
NS lag 0
NS lag 1
NS lag 2
Asthma (61+yrs)
NS lag 0
O
HH
H
W
-------
O
to
O
O
oo
TABLE AX6.3-4 (cont'd). RESPIRATORY HEALTH EFFECTS OF OXIDES OF NITROGEN:
EMERGENCY DEPARTMENT VISITS
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels of NO2 &
Monitoring Stations
Copollutant Effects and Interpretation:
Correlations Relative Risk & Confidence Intervals (95%)
CANADA (cont'd)
X
ON
OO
H
6
o
o
H
O
O
H
W
O
HH
H
W
Bates etal. (1990)
(cont'd)
Kestenetal. (1995)
Toronto, ON
Period of Study:
7/1/1991-6/30/1992
Stieb etal. (1996)
St. John, New Brunswick,
Canada
Period of Study: 1984-
1992
Outcome(s): Asthma
Age groups analyzed: All
Study Design: Time-series
N: 854
# of Hospitals: 1
Statistical analysis: autoregressive
technique
Statistical Package: SAS v 6.04
Lag: 0,1
Outcome(s): Asthma
(May-Sept only)
ICD-9 Codes: NR
Age groups analyzed: 0-15, >15
Study Design: Time-series
N: 1,163
# of Hospitals: 2
Statistical Analyses: SAS NLIN
(Equivalent to Poisson GEE)
Covariates: day of wk, long-term trends,
Season: Summers only (May-Sep)
Dose-response investigated: Yes
Statistical Package: SAS
Lag: 0-3 days
24-h avg NO2: SO2
Range: 2.20-3.75 x 0.01 ppm O3
l-hmaxNO2 (ppb)
Mean: 25.2
Range: 0, 120
95th: 60
O3;r=0.16
S02;r=-0.03
SO42~;r=0.16
TSP;r=0.15
NS lag 1
NS lag 2
Respiratory (61+ yrs)
r=0.\76\ag\p<0.00\
r=Q.\78\ag2p
-------
O
to
O
O
oo
X
ON
OO
oo
TABLE AX6.3-4 (cont'd). RESPIRATORY HEALTH EFFECTS OF OXIDES OF NITROGEN:
EMERGENCY DEPARTMENT VISITS
Reference, Study
Location, & Period
Mean Levels of NO2 &
Outcomes, Design, & Methods Monitoring Stations
Copollutant
Correlations
Effects and Interpretation:
Relative Risk & Confidence Intervals (95%)
CANADA (cont'd)
Stieb* et al. (2000)
Saint John, New
Brunswick, Canada
Period of Study:
Retrospective:
7/92-6/94
Prospective:
7/94-3/96
Outcome(s): Asthma; COPD; Annual mean: 8.9 ppb
Respiratory infection (bronchitis, spring/fall mean: 10.0 ppb
bronchiolitis, croup, pneumonia); Max: 82
All respiratory
ICD-9 Codes: NR
Age groups analyzed: All
Study Design: Time-series
N: 19,821
Statistical Analyses: Poisson regression,
GAM
Covariates: Day ofwk, selected weather
variables in each model
Seasons: All yr, summer only
Dose-response investigated: Yes
Statistical Package: S-Plus
Lag: all yr = 0; summer
only = 8
03;r=-0.02
SO2;r = 0.41
TRS;r=0.16
PM10;r=0.35
PM2.5;r = 0.35
11^=0.25
SO42~;r=0.33
COH; r = 0.49
Assessed
multipollutant
models
Increment: 8.9ppb(IQR)
Respiratory visits: - 3.8%, p = 0.070 lag 0
May to Sept: 11.5%, p = 0.17 lag 8
Multipollutant model (NO2, O3, SO2)
-3.6% [-7.5, 0.5] lag 0
Multipollutant model (ln(NO2), O3, SO2 COH)
May to Sept: 4.7% [0.8 to 8.6] lag 8
Non- linear effect of NO2 on summertime
respiratory visits observed and log
transformation strengthened the association.
H
6
o
o
H
O
O
H
W
O
HH
H
W
-------
O
to
O
O
oo
TABLE AX6.3-4 (cont'd). RESPIRATORY HEALTH EFFECTS OF OXIDES OF NITROGEN:
EMERGENCY DEPARTMENT VISITS
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels of NO2 &
Monitoring Stations
Copollutant Effects and Interpretation:
Correlations Relative Risk & Confidence Intervals (95%)
EUROPE and MIDDLE-EAST
Sunyeretal. (1997)
Multi-city, Europe
(Barcelona, Helsinki,
Paris, London)
Period of Study: 1986-
1992
Outcomes (ICD-9): Asthma (493)
Age groups analyzed: <15,15-64
Study Design: Time-series
Statistical Analyses: APHEA protocol,
Poisson regression, GEE; meta-analysis
Covariates: Humidity, temperature,
influenza, soybean, long-term trend,
season, day of wk
Season: Cool, Oct-Mar; Warm: Apr-Sep
Statistical Package: NR
Lag: 0,1,2,3 and cumulative 1-3
24-h median (range)
(ug/m3)
Barcelona: 53 (5, 142)
Helsinki: 35(9,78)
London: 69(27,347)
Paris: 42 (12, 157)
# of stations:
Barcelona: 3
London: 2
Paris: 4
Helsinki: 8
SO2
black smoke
03
Increment: 50 ug/m3 of 24-h avg for all cities
combined
Asthma
15-64yrs
1.029 [1.003, 1.055] lag 0-1
1.038 [1.008, 1.068] lag 0-3, cumulative
<15 yrs
1.026 [1.006, 1.049] lag 2
1.037 [1.004, 1.067] lag 0-3, cumulative
1.080 [1.025, 1.140] -Winter only
X
OO
VO
Two-pollutant models:
NO2/Black smoke
15-64 yrs 1.055 [1.005, 1.109] lag 0-1
15-64 yrs 1.088 [1.025, 1.155] cumulative 0-3
<15yrs 1.036 [0.956, 1.122]
H
6
o
o
H
O
O
H
W
N02/S02
<15yrs 1.034 [0.988, 1.082]
O
HH
H
W
-------
O
to
O
O
oo
TABLE AX.3-4 (cont'd). RESPIRATORY HEALTH EFFECTS OF OXIDES OF NITROGEN:
EMERGENCY DEPARTMENT VISITS
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels of NO2 &
Monitoring Stations
Copollutant Effects and Interpretation:
Correlations Relative Risk & Confidence Intervals (95%)
EUROPE and MIDDLE-EAST (cont'd)
X
H
6
o
o
H
O
O
H
W
O
HH
H
W
Atkinson et al. (1999b)
London, United Kingdom
Period of Study:
1/92-1294
Buchdahl et al. (1996)
London, United Kingdom
Period of Study: 3/1/92-
2/28/93
Outcome(s) (ICD-9): Respiratory
ailments (490-496), including asthma,
wheezing, inhaler request, chest
infection, COPD, difficulty in breathing,
cough, croup, pleurisy, noisy breathing
Age groups analyzed: 0-14; 15-64; >65;
All ages
Study Design: Time-series
N: 98,685
# of Hospitals: 12
Statistical Analyses: Poisson regression,
APHEA protocol
Covariates: Long-term trend, season, day
of wk, influenza, temperature, humidity
Statistical Package: SAS
Lag: 0,1,0-2, and 0-3 days
Outcomes: Daily acute wheezy episodes
ICD-9: NR
Age groups analyzed: > 16
Study Design: Case-control
N: 1,025 cases, 4,285 controls
Number of hospitals: 1
Statistical Analyses: Poisson regression
Covariates: Season, temperature, wind
speed
Season: Spring (Apr-Jun), Summer
(Jul-Sep), Autumn (Oct Dec),
Winter (Jan-Mar)
Statistical Package: Stata
Lag: 0-7 days
1-hmax: 50.3ppb,
SD = 17.0
# of Stations: 3;
r= 0.70, 0.96
NO2, O3
(8 h), SO2
(24 h), CO (24 h),
PM10 (24 h), BS
NO2 24-h yr round mean:
60 ug/m3, SD = 17
IQR: 17 ug/m3
Spring: 59(19)
Summer: 55(18)
Fall: 66(13)
Winter: 61 (17)
SO2 r = 0.62
O3r=-0.18
Increment: 36 ppb in 1-hmax
Single-pollutant model
Asthma Only
0-14 yrs 8.97% [4.39, 13.74] lag 1
15-64 yrs 4.44% [0.14, 8.92] lag 1
All ages 4.37% [1.32, 7.52] lag 0
All Respiratory
0-14 yrs 2.17% [-0.49, 4.91] lag 1
15-64 yrs 1.87% [-0.69, 4.49] lag 2
>65 yrs 3.97% [0.51, 7.55] lag 0
All Ages 1.20% [-0.57, 3.00]
Two-pollutant model Asthma Only 0-14 yrs:
SO2: 5.75% [0.39, 11.40] lag 1
CO: 8.34% [3.61, 13.29] lag 0
PM10: 6.95% [1.96,12.19] lag 2
BS: 8.32% [3.56, 13.30] lag 2
03: 9.68% [5.02, 14.54] lag 0
Increment: 17 ug/m3 (IQR)
No adjustments to model
RR 1.07 [1.01, 1.14] lag not specified
Adjusted for temperature and season.
RR 1.02 [0.96, 1.09] lag not specified
-------
O
to
O
O
oo
TABLE AX6.3-4 (cont'd). RESPIRATORY HEALTH EFFECTS OF OXIDES OF NITROGEN:
EMERGENCY DEPARTMENT VISITS
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels of NO2 &
Monitoring Stations
Copollutant Effects and Interpretation:
Correlations Relative Risk & Confidence Intervals (95%)
EUROPE and MIDDLE-EAST (cont'd)
X
H
6
o
o
H
O
O
H
W
O
HH
H
W
Thompson et al. (2001)
Belfast, Northern Ireland
Period of Study:
1993-1995
Outcome(s): Asthma
ICD-9 Code(s): NR
Age groups analyzed: Children
Study Design: Time-series
N: 1,044
Statistical Analyses: Followed APHEA
protocol, Poisson regression analysis
Covariates: Season, long-term trend,
temperature, day of wk, holiday
Season: Warm (May-Oct);
Cold (Nov-Apr)
Statistical Package: Stata
Lag: 0-3
Warm Season
NO2(ppb): Mean: 19.20;
SD = 7.90;
IQR: 13.0,23.0
NOx(ppb): Mean: 35.50;
SD = 25.50;
IQR: 21.0,40.0
NO(ppb): Mean: 16.4;
SD = 19.70;
IQR: 7.0,17.0
Cold Season
NO2(ppb): Mean: 23.30;
SD = 9.00;
IQR: 18.0,28.0
NOx(ppb): Mean: 50.50;
SD = 50.50;
IQR: 26.0,56.0
NO(ppb): Mean: 27.30;
SD = 43.10;
IQR: 9.0,28.0
NO2:
PM10; r = 0.77
S02;r = 0.82
NOX; r = 0.93
NO; r = 0.84
O3;r=-0.62
CO; r= 0.69
Benzene; r = 0.83
NOX:
PM10; r = 0.73
S02;r = 0.83
N02;r=0.92
NO; r = 0.97
O3;r=-0.73
CO; r= 0.74
Benzene; r = 0.86
NO:
PM10;r=0.65
SO2; r = 0.76
NOX; r = 0.97
N02;r=0.84
O3;r=-0.76
CO; r= 0.71
Benzene; r = 0.82
NO2 Increment: lOppb
NOX Increment: per doubling
NO Increment: per doubling
N02
Lag ORR 1.08 [1.03, 1.13]
Lag 0-1 RR 1.11 [1.05,1.17]
Lag 0-2 RR 1.10 [1.04, 1.17]
Lag 0-3 RR 1.12 [1.03, 1.20]
Warm only Lag 0-1 RR 1.14 [1.04, 1.26]
Cold only Lag 0-1 RR1.10 [1.03, 1.17]
Adjusted for Benzene Lag 0-1 RR 0.99
[0.87,1.13]
NOX
Lag ORR 1.07 [1.02, 1.12]
Lag 0-1 RR 1.10 [1.05, 1.16]
Lag 0-2 RR 1.10 [1.03, 1.17]
Lag 0-3 RR 1.11 [1.04,1.20]
Warm only Lag 0-1 RR 1.13 [1.03, 1.24]
Cold only Lag 0-1 RR 1.09 [1.02, 1.16]
Adjusted for Benzene Lag 0-1 RR 0.89
[0.77, 1.03]
NO
Lag ORR 1.04 [1.01, 1.07]
Lag 0-1 RR 1.07 [1.03, 1.11]
Lag 0-2 RR 1.06 [1.02, 1.11]
Lag 0-3 RR 1.08 [1.02, 1.14]
Warm only Lag 0-1 RR 1.08 [1.01, 1.16]
Cold only Lag 0-1 RR 1.06 [1.01, 1.11]
Adjusted for Benzene Lag 0-1 RR 0.93
[0.85, 1.01]
-------
O
to
O
O
oo
TABLE AX6.3-4 (cont'd). RESPIRATORY HEALTH EFFECTS OF OXIDES OF NITROGEN:
EMERGENCY DEPARTMENT VISITS
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels of NO2 &
Monitoring Stations
Copollutant
Correlations
Effects and Interpretation:
Relative Risk & Confidence Intervals (95%)
EUROPE and MIDDLE-EAST (cont'd)
Boutin-Forzano et al.
(2004)
Marseille, France
Outcome(s): Asthma
ICD-9 Code(s): NR
Age groups analyzed: 3-49
MeanNO2: 34.9 ug/m3
Range: 3.0, 85
SO2;r = 0.56
O3;r=0.58
Increment: 10 ug/m3
Increased ER visits
Period of Study:
4/97-3/98
Study Design: Case-Crossover
N: 549
Statistical Analyses: Logistic regression
Covariates: Minimal daily temperature,
maximum daily temperature, minimum
daily relative humidity, maximum daily
relative humidity, day of wk
Statistical Package: NR
Lag: 0-4 days
OR 1.0067 [0.9960, 1.0176] lag 0
X
to
H
6
o
o
H
O
O
H
W
Castellsague et al. (1995)
Barcelona, Spain
Period of Study:
1986-1989
Outcome(s): Asthma
ICD-9 Code(s): NR
Age groups analyzed: 15-64
Study Design: Time-series
# of Hospitals: 4
Statistical Analyses: Poisson regression
Covariates: long-time trend, day of wk,
temperature, relative humidity, dew point
temperature
Seasons: Winter: Jan-Mar;
Summer: Jul-Sep
Dose-Response investigated: Yes
Statistical Package: NR
Lag: 0, 1-5 days and cumulative.
Summer: lag 2 days
Winter: lag 1 day
Mean NO2 (ug/m )
Summer: 104.0
Winter: 100.8
IQR (ug/m3):
Summer: 48
Winter: 37
S02; r = NR
O3; r = NR
# of Stations:
3 automatic
15 manual,
Increment: 25 ug/m3
Seasonal differences
Summer:
1.071 [1.101, 1.130] lag 0-5 cumulative
1.045 [1.009, 1.081] lag 0
Winter:
1.072 [1.010, 1.137] lag 0-2 cumulative
1.056 [1.011, 1.104] lag 0
Asthma visits increased across quartiles of NO2
in summer; a positive but less consistent
increase across quartiles was observed in winter.
O
HH
H
W
-------
O
to
O
O
oo
X
H
6
o
o
H
O
O
H
W
O
HH
H
W
TABLE AX6.3-4 (cont'd). RESPIRATORY HEALTH EFFECTS OF OXIDES OF NITROGEN:
EMERGENCY DEPARTMENT VISITS
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels of NO2 &
Monitoring Stations
Copollutant
Correlations
Effects and Interpretation:
Relative Risk & Confidence Intervals (95%)
EUROPE and MIDDLE-EAST (cont'd)
Tobias etal. (1999)
Barcelona, Spain
Outcome(s): Asthma
ICD-9: NR
Age groups analyzed: >14
24-h avg NO2 ug/m
Non-epidemic days: 54.7
BS
SO2
03
P x 104 (SE x 104 ) using Std Poisson
Without modeling asthma epidemics: 1 1 .25 (1 1 .79) p > 0. 1
Modeling epidemics with 1 dummy variable: 1 . 1 8 (7. 59) p
Period of Study:
1986-1989
Galan et al. (2003)
Madrid, Spain
Period of Study:
1995-1998
Study Design: Time-series
Statistical Analyses: Poisson
regression, followed APHEA
protocol
Covariates: temperature, humidity,
long-term trend, season, day of wk
Statistical Package: NR
Lag: NR
Outcome(s) (ICD-9): Asthma (493)
Age groups analyzed: All
Study Design: Time-series
N: 4,827
Statistical Analyses: Poisson
regression, (1) classic APHEA
protocol and (2) GAM with stringent
criteria
Covariates: trend, yr, season, day of
wk, holidays, temperature, humidity,
influenza, acute respiratory
infections, pollen
Statistical Package: NR
Lag: 0-4 days
(20.8)
Epidemic days:
58.9 (26.7)
24-h mean: 67.1 ug/m3
SD=18.0
IQR: 20.5
Max: 147.5
# of Stations: 15
PM10;r= 0.717
SO2;r= 0.610
03;r= -0.209
Modeling epidemics with 6 dummy variables: 13.60 (7.79)
p<0.1
Modeling each epidemic with dummy variable: 14.40
(7.44) p< 0.1
P x 104 (SE x 104) using Autoregressive Poisson
Without modeling asthma epidemics:
13.65(11.81)p>0.1
Modeling epidemics with 1 dummy variable: 3.28 (7.77) p
>0.1
Modeling epidemics with 6 dummy variables: 16.49 (8.01)
p<0.05
Modeling each epidemic with dummy variable: 18.18
(8.01)p<0.1
Increment: 10 ug/m3
Asthma:
RR 1.013 [0.991, 1.035] lag 0
RR 1.011 [0.989, 1.032] lag 1
RR 1.013 [0.992, 1.034] lag 2
RR 1.033 [1.013, 1.054] lag 3
RR 1.026 [1.006, 1.047] lag 4
Multipollutant model:
NO2/SO2 1.031 [1.004,1.059] lag 3
N02/PM10 1.001 [0.971, 1.031] lag 3
NO2/Pollen 1.024 [1.004, 1.044] lag 3
NO2/Pollen/O3 1.024 [1.005, 1.045] Poisson
N02/Pollen/03 1.022 [1.005, 1.040] GAM
-------
O
to
O
O
oo
TABLE AX6.3-4 (cont'd). RESPIRATORY HEALTH EFFECTS OF OXIDES OF NITROGEN:
EMERGENCY DEPARTMENT VISITS
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels of NO2 &
Monitoring Stations
Copollutant
Correlations
Effects and Interpretation:
Relative Risk & Confidence Intervals (95%)
EUROPE and MIDDLE-EAST (cont'd)
X
H
6
o
o
H
O
O
H
W
Teniasetal. (1998)
Valencia, Spain
Period of Study:
1993-1995
Seasons:
Cold: Nov-Apr
Warm: May-Oct
Tenias et al. (2002)
Valencia, Spain
Period of Study: 1994-
1995
Outcome(s): Asthma
ICD-9 Code(s): NR
Age groups analyzed: >14
Study Design: Time-series
N: 734
Statistical Analyses: Poisson
regression, APHEA protocol
Covariates: seasonality,
temperature, humidity, long-term
trend, day of wk, holidays, influenza
Seasons: Cold: Nov-Apr;
Warm: May-Oct
Dose-Response Investigated: Yes
Statistical Package: NR
Lag: 0-3 days
Outcome(s): COPD
ICD-9 Code(s): NR
Age groups analyzed: >14
Study Design: Time-series
N: 1,298
# of Hospitals: 1
Statistical Analyses: Poisson
regression, APHEA protocol; basal
models and GAM
Covariates: Seasonality, annual
cycles, temperature, humidity, day
ofwk, feast days
Seasons: Cold: Nov-Apr;
Warm: May-Oct
Dose-Response Investigated: Yes
Statistical Package: NR
Lag: 0-3 days
24 h:
57.7 ug/m3
Cold: 55.9
Warm: 59.4
1-hmax:
101.1 ug/m3
Cold: 97.3
Warm: 102.8
# of Stations: 2
NO224-havg: 57.7
ug/m ; Range: 12, 135
1-hmax: 100.1 ug/m3;
Range: 31,305
# of Stations: 6 manual
and 5 automatic; r = 0.87
24 h:
O3;r=-0.304
SO2 (24 h); r = 0.265
S02(lh);r= 0.261
Ih:
O3;r=-0.192
SO2(24h);r=0.199
S02(lh);r= 0.201
BS;r = 0.246
SO2;r=0.194
CO; r= 0.180
O3;r=-0.192
Increment: 10 ug/m
NO2 24-h avg
Allyr 1.076 [1.020, 1.134] lag 0
Cold: 1.083 [1.022, 1.148] lag 0
Warm: 1.066 [0.989, 1.149] lag 0
NO2 1-hmax
Allyr 1.037 [1.008, 1.066] lag 0
Cold: 1.034 [1.004, 1.066] lag 0
Warm: 1.044 [1.002, 1.088] lag 0
Increment: 10 ug/m3
24-h avg NO2
All yrRR 0.979 [0.943, 1.042] lag 0
Cold: 24-h avg: RR 0.991 [0.953, 1.030] lag 0
Warm: 24-h avg: RR 0.961 [0.900, 1.023] lag 0
l-hmaxNO2
All yrRR 0.986 [0.966, 1.007] lag 0
Cold: 24-h avg: RR 0.996 [0.975, 1.018] lag 0
Warm: 24-h avg: RR 0.968 [0.935, 1.003] lag 0
Possibility of a linear relationship between pollution
and risk of emergency cases could not be ruled out.
O
HH
H
W
-------
O
to
O
O
oo
TABLE AX6.3-4 (cont'd). RESPIRATORY HEALTH EFFECTS OF OXIDES OF NITROGEN:
EMERGENCY DEPARTMENT VISITS
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels of NO2 &
Monitoring Stations
Copollutant Effects and Interpretation:
Correlations Relative Risk & Confidence Intervals (95%)
EUROPE and MIDDLE-EAST (cont'd)
X
Migliaretti et al. (2005)
Turin, Italy
Period of Study: 1997-
1999
Outcome (ICD-9): Asthma (493)
Age groups analyzed:
<15, 15-64,>64
Study Design: Case-Control
Controls: age matched with other
respiratory disease (ICD-9: 460 487,
490-2,494-6, 500-19) or heart disease
(ICD-9: 390-405,410-429)
N: cases = 1,401
controls = 201,071
Statistical Analyses: Logistic regression
Covariates: Seasonality, temperature,
humidity, solar radiation, wind velocity,
day of wk, holiday, gender, age,
education level
Seasons: Cold: Oct-Mar; Warm: Apr-
Sep
Statistical Package: NR
Lag: 0-3 days and cumulative
All Participants:
24-h mean: 112.7 ug/m3,
SD = 30.2, Two-pollutant
Median = 107.7
Cases:
24-h mean: 117.1 ug/m3,
SD = 30.0,
Median = 113.0
Controls:
24-h mean: 112.7 ug/m3,
SD = 30.2,
Median = 107.7
# of Stations: 10; r = 0.79
model adjusted for
TSP
Increment: 10 ug/m
Single Pollutant (NO2):
<15yrs2.3%[0.3,4.40]
15-64 yrs 3.10% [-0.01, 7.70]
>64yrs 7.70% [0.20, 15.20]
All ages 2.40% [0.5, 4.30]
Copollutant (NO2 and TSP)
<15 yrs 1.71% [-0.02, 5.00]
15-64 yrs 1.20% [-0.06, 6.50]
>64 yrs 0.91% [-0.08, 5.91]
All ages 1.10% [-0.02, 3.82]
H
6
o
o
H
O
O
H
W
O
HH
H
W
-------
O
to
O
O
oo
TABLE AX6.3-4 (cont'd). RESPIRATORY HEALTH EFFECTS OF OXIDES OF NITROGEN:
EMERGENCY DEPARTMENT VISITS
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels of NO2 &
Monitoring Stations
Effects and Interpretation:
Copollutant Relative Risk & Confidence Intervals
Correlations (95%)
EUROPE and MIDDLE-EAST (cont'd)
X
Oi
OD
Oi
H
6
o
o
H
O
O
H
W
Bedeschi et al. (2007)
Reggio Emilia, Italy
Period of Study:
03/2001-03/2002
Vigotti et al. (2007)
Pisa, Italy
Period of Study: 2000
Outcome (ICD-9): Respiratory disorders,
asthma or asthma-like disorders, other
respiratory disorders
Age groups analyzed: <15
Study Design: Time-series
N: 854 children, 1051 visits
Statistical Analyses: Poisson regression
with GAM
Covariates: Weekday, festivity day,
humidity, precipitation, temperature, flu
epidemic, pollen concentrations
Statistical Package: R software
Lag: 0-5 days and cumulative
Outcome (ICD-9): Respiratory
complaints (493,468, 466)
Age groups analyzed: <10, >65
Study Design: Ecologic
N: 966
Number of hospitals: 1
Statistical Analyses: Robust poisson
regression in GAM model
Covariates: Day of study, temperature,
humidity, rain, influenza epidemics, day
of wk, holidays
Statistical Package: NR
Lag: up to 5 d
24-havg: NO2 (ug/mj)
Mean: 49(13.8)
Range: 21.6-107.5
Median: 47.5
SO2;r = 0.56
CO;r = 0.77
TSP;r=0.58
PM10;r=0.57
O3;r=-0.50
Increment: 10 ug/m
All Respiratory Disorders:
Italians: 9% [1.0, 17.6] lag 4
Foreigners: 17.6% [3.9, 33.0] lag 4
All: 11.0% [3.6, 18.8] lag 4
24-havg: 45.6 (11.0) ug/m3
Range: 21.3-74.0
50th: 44.8
Number of monitors: 3
PM10;r=0.58
CO; r= 0.62
Increment: 10 ug/m
Children: 1.118 [1.014, 1.233] lag 0-2
65+: 1.06 [0.967, 1.162] lag 0-2
O
HH
H
W
-------
O
to
O
O
oo
TABLE AX6.3-4 (cont'd). RESPIRATORY HEALTH EFFECTS OF OXIDES OF NITROGEN:
EMERGENCY DEPARTMENT VISITS
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels of NO2 &
Monitoring Stations
Copollutant Effects and Interpretation:
Correlations Relative Risk & Confidence Intervals (95%)
EUROPE and MIDDLE-EAST (cont'd)
X
H
6
o
o
H
O
O
H
W
Pantazopoulou et al.
(1995)
Athens, Greece
Period of Study : 1988
Garty etal. (1998)
Tel Aviv, Israel
1993
Outcomes: All respiratory visits
ICD-9: NR
Age groups analyzed: All ages
Study Design: Time-series
N: 213,316
Number of hospitals: 14
Statistical Analyses: Multiple linear
regression
Covariates: Season, day of wk, holiday,
temperature, relative humidity
Season: Warm (3/22-9/21), Cold (1/1-
3/2 land 9/22-12/31)
Lag: NR
Outcome(s): Asthma
ICD-9 Code(s): NR
Age groups analyzed: 1-18
Study Design: Descriptive study with
correlations
N: 1,076
Statistical Analyses: Pearson correlation
and partial correlation coefficients
Covariates: maximum and minimum
ambient temperatures, relative humidity
and barometric pressure
Statistical Package: Statistix
NO2 24-h avg
Winter: 94 ug/m3,
SD = 25
5th: 59,50th: 93,
95th: 135
Summer: 111 ug/m3,
SD = 32
5th: 65,50th: 108,
95th: 173
# of stations: 2
24-h mean of NOX
(estimated from
histogram): 60 ug/m3;
Range: 50,250
CO
BS
Increment: 76 ug/m3 in winter and 108 ug/m3 in
summer (95th-5th)
Respiratory disease admissions
Winter: Percent increase: p = 66.8 [19.6, 113.9]
Summer: Percent increase: p = 21.2 [-35.1, 77.5]
Correlation between NOX and ER visits for asthma:
All Yr:
Daily data r= 0.30
Running mean for 7 days r = 0.62
Excluding Sept:
Daily data r= 0.37
Running mean for 7 days r = 0.74
38% of variance in number of ER visits explained
by fluctuations in NOX. Increases to 55% when
Sept. is omitted from analyses.
O
HH
H
W
-------
O
to
O
O
oo
TABLE AX6.3-4 (cont'd). RESPIRATORY HEALTH EFFECTS OF OXIDES OF NITROGEN:
EMERGENCY DEPARTMENT VISITS
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels of NO2 &
Monitoring Stations
Copollutant Effects and Interpretation:
Correlations Relative Risk & Confidence Intervals (95%)
LATIN AMERICA
X
oo
H
6
o
o
H
O
O
H
W
Farhat* et al. (2005)
Sao Paulo, Brazil
Period of Study:
1996-1997
Martins* et al. (2002)
Sao Paulo, Brazil
Period of Study:
5/96-9/98
Outcome(s) (ICD-9): Lower Respiratory
Disease (466, 480-5)
Age groups analyzed: <13
Study Design: Time-series
N: 4,534
# of Hospitals: 1
Statistical Analyses: (l)Poisson
regression and (2) GAM-no mention of
more stringent criteria
Covariates: Long-term trends,
seasonality, temperature, humidity
Statistical Package: S-Plus
Lag: 0-7 days, 2,3,4 day moving avg
Outcome(s)(ICD-10): Chronic Lower
Respiratory Disease (CLRD) (J40-J47);
includes chronic bronchitis, emphysema,
other COPDs, asthma, bronchiectasia
Age groups analyzed: >64
Study Design: Time-series
N: 712
# of Hospitals: 1
Catchment area: 13,163 total ER visits
Statistical Analyses: Poisson regression
and GAM - no mention of more stringent
criteria
Covariates: weekdays, time, minimum
temperature, relative humidity, daily
number of non-respiratory emergency
room visits made by elderly
Statistical Package: S-Plus
Lag: 2-7 days and 3 day moving avgs
Mean: 125.3 ug/mj
SD = 51.7
IQR: 65.04 ug/m3
# of Stations: 6
PM10;r=0.83
SO2;r = 0.66
CO; r = 0.59
NO2 max 1-h avg (ug/m ):
117.6,
SD = 53.0,
Range: 32.1,421.6
IQR: 62.2 ug/m3
# of Stations: 4
O3; r = 0.44
SO2; r = 0.67
PM10;r=0.83
CO; r= 0.62
Increment: IQR of 65.04 ug/m
Single-pollutant models (estimated from
graphs):
LRD-17.5% [12.5, 24]
Multipollutant models:
Adjusted for:
PM10 16.1% [5.4, 26.8] 4 day avg
S02 24.7% [18.2, 31.3] 4 day avg
CO 19.2% [11.8, 26.6] 4 day avg
Multipollutant model
18.4% [3.4, 33.5] 4 day avg
Increment: IQR of 62.2 ug/m3
Percent increase: 4.5% [-6.5, 15] lag 3 day
moving avg (estimated from graph)
O
HH
H
W
-------
O
to
O
O
oo
TABLE AX6.3-4 (cont'd). RESPIRATORY HEALTH EFFECTS OF OXIDES OF NITROGEN:
EMERGENCY DEPARTMENT VISITS
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels of NO2 &
Monitoring Stations
Copollutant Effects and Interpretation:
Correlations Relative Risk & Confidence Intervals (95%)
LATIN AMERICA (cont'd)
X
Ilabacaetal. (1999)
Santiago, Chile
Period of Study:
2/1/95-8/31/96
Outcome(s) (ICD-9): Upper respiratory
illness (460 465, 487);
Days: 578
Lower respiratory illness
(466, 480-486, 490-494, 496, 519.1,
033.9);
Pneumonia (480 486)
Age groups analyzed: <15
Study Design: Time-series
# of Hospitals: 1
Statistical Analyses: Poisson regression
Covariates: Long-term trend, season, day
of wk, temperature, humidity, influenza
epidemic
Season: Warm (Sep-Apr),
Cool (May-Aug)
Statistical Package: NR
Lag: 0-3 days
24-h avg NO2:
Warm:
Mean: 97.0
Median: 91.5
SD = 34.6
Range: 37.2,246
5th: 54.3
95th: 163.0
Cool:
Mean: 160.2
Median: 154.4
SD = 59.5
Range: 60.1,397.5
5th: 74.4
95th: 266.0
# of stations: 4, r = 0.70,0.:
Warm: Increment: IQR
SO2;r = 0.66
O3; r = 0.15 All respiratory
PM10; r = 0.71 Cool
PM2.5;r = 0.70 Lag 2 IQR: 56.4 RR 1.0378 [1.0211, 1.0549]
Lag 3 IQR: 56.4 RR 1.0294 [1.0131, 1.0460]
Cool: Lagavg7IQR: 33.84 RR 1.0161 [1.0000, 1.0325]
SO2; r = 0.74 Warm
03;r=0.22 Lag 2 IQR: 30.08 RR 1.0208 [0.9992, 1.0428]
PM10;r=0.82 Lag 3 IQR: 30.08 RR 1.0395 [1.0181, 1.0612]
PM2.5;r = 0.80 Lag avg 7 IQR: 22.56 RR 1.0251 [0.9964, 1.0548]
Upper respiratory
Cool
Lag 2 IQR: 56.4 RR 1.0569 [1.0339, 1.0803]
Lag 3 IQR: 56.4 RR 1.0318 [1.0095, 1.0545]
Lag avg 7 IQR: 33.84 RR 1.0177 [0.9960, 1.0399]
Warm
Lag 2 IQR: 30.08 RR 1.0150 [0.9881, 1.0426]
Lag 3 IQR: 30.08 RR 1.0425 [1.0157, 1.0699]
Lag avg 7 IQR: 22.56 RR 0.9944 [0.9591, 1.0311]
H
6
o
o
H
O
O
H
W
Pneumonia
Cool
Lag 2 IQR: 56.4 RR 1.0824 [1.0300, 1.1374]
Lag 3 IQR: 56.4 RR 1.0768 [1.0273, 1.1287]
Lag avg 7 IQR: 33.84 RR 1.0564 [1.0062, 1.1092]
Warm
Lag 2 IQR: 30.08 RR 1.1232 [1.0450, 1.2072]
Lag 3 IQR: 30.08 RR 1.0029 [0.9332, 1.0779]
Lag avg 7 IQR: 22.56 RR 1.1084 [1.0071, 1.2200]
O
HH
H
W
-------
O
to
O
O
oo
TABLE AX6.3-4 (cont'd). RESPIRATORY HEALTH EFFECTS OF OXIDES OF NITROGEN:
EMERGENCY DEPARTMENT VISITS
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels of NO2 &
Monitoring Stations
Copollutant Effects and Interpretation:
Correlations Relative Risk & Confidence Intervals (95%)
LATIN AMERICA (cont'd)
X
O
O
Lin etal. (1999)
Sao Paulo, Brazil
Period of Study: May
1991 -Apr 1993
Days: 621
Outcome(s): Respiratory disease, Upper
respiratory illness, Lower respiratory
illness, Wheezing
ICD-9 Code(s): NR
Age groups analyzed: <13
Study Design: Time-series
# of Hospitals: 1
Statistical Analyses: Gaussian and
Poisson regression
Covariates: Long-term trend,
seasonality, day of wk, temperature,
humidity
Statistical Package: NR
Lag: 5-day lagged moving avgs
NO2 ug/m :
Mean: 163
SD = 85
Range: 2,688
Number of stations: 3
SO2;r = 0.38
CO; r= 0.35
PM10; r = 0.40
O3;r=0.15
Increment: NR
All respiratory illness
NO2 alone: RR 1.003 [1.001, 1.005] 5-day
moving avg
NO2 + PM10 + O3 + SO2 + CO: RR 0.996
[0.994, 0.998]
Lower respiratory illness
NO2 alone: RR 0.999 [0.991, 1.007] 5-day
moving avg
NO2 + PM10 + O3 + SO2 + CO: RR 0.990
[0.982, 0.998]
Upper respiratory illness
N02 alone: RR 1.003 [0.999, 1.007] 5-day
moving avg
N02 + PM10 + 03 + S02 + CO: RR 0.996
[0.992,1.000]
H
6
o
o
H
O
O
H
W
Wheezing
NO2 alone: RR 0.996 [0.990, 1.002] 5-day
moving avg
N02 + PM10 + 03 + S02 + CO: RR 0.991
[0.983,0.999]
O
HH
H
W
-------
O
to
O
O
oo
TABLE AX6.3-4 (cont'd). RESPIRATORY HEALTH EFFECTS OF OXIDES OF NITROGEN:
EMERGENCY DEPARTMENT VISITS
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels of NO2 &
Monitoring Stations
Copollutant Effects and Interpretation:
Correlations Relative Risk & Confidence Intervals (95%)
ASIA
X
I
o
H
6
o
o
H
O
o
H
W
Kim et al. (2007)
Seoul, Korea
Period of Study: 2002
Outcome(s) (ICD-10): Asthma (J45, J46)
Age groups analyzed: All
Study Design: Case-crossover
N: 92,535
Statistical Analyses: Conditional logistic
regression
Covariates: Time trend, weather conditions,
seasonality
Statistical Package: NR
Lag: 0, 1,2, 3,4, 2-4 ma
24-havg: 36.0 (14.7) ppb PM10
Range: 2.3-108.0 CO
50th: 34.3 SO2
IQR: 20.1 O3
Increment: 20.1 ppb
Stratified by individual SES:
Highest SES quintile: 1.06 [1.02, 1.10] lag 2-4
2nd Quintile: 1.06 [1.02, 1.09] lag 2-4
3rd Quintile: 1.03 [0.99, 1.06] lag 2-4
4th Quintile: 1.06 [ 1.02, 1.10] lag 2-4
Lowest SES quintile: 1.05 [1.00, 1.10] lag 2-4
Stratified by Regional SES:
Highest SES quintile: 0.96 [0.90, 1.02] lag 2-4
2nd Quintile: 1.08 [1.04, 1.13] lag 2-4
3rd Quintile: 1.03 [1.00, 1.07] lag 2-4
4th Quintile: 1.06 [ 1.02, 1.10] lag 2-4
Lowest SES quintile: 1.06 [1.02, 1.09] lag 2-4
Overall: 1.05 [1.03, 1.06]
Relative Effect Modification for Interaction
Terms:
Stratified by individual SES x air pollution:
Highest SES quintile: 1.00 [referent]
2nd Quintile: 0.99 [0.95, 1.04]
3rd Quintile: 0.96 [0.92, 1.01]
4th Quintile: 1.00 [0.95, 1.05]
Lowest SES quintile: 0.99 [0.93, 1.05]
Stratified by Regional SES x air pollution:
O
HH
H
W
-------
O
to
O
O
oo
TABLE AX6.3-4 (cont'd). RESPIRATORY HEALTH EFFECTS OF OXIDES OF NITROGEN:
EMERGENCY DEPARTMENT VISITS
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels of NO2 &
Monitoring Stations
Copollutant Effects and Interpretation:
Correlations Relative Risk & Confidence Intervals (95%)
ASIA (cont'd)
X
O
to
H
6
o
o
H
O
O
H
W
Kim et al. (2007)
(cont'd)
Sun et al. (2006)
Central Taiwan
Period of Study: 2004
Chew etal. (1999)
Singapore
Period of Study:
1990-1994
Outcome(s) (ICD-9): asthma (493)
Age groups analyzed: <16,16-55
Study Design: Cross-sectional
Number of hospitals: 4
Statistical Analyses: Multiple correlation
coefficients/multiple regression analysis
Covariates:
Statistical Package: SPSS v 12.0
Lag:
Outcome(s) (ICD-9): Asthma (493)
Age groups analyzed: 3-12,13-21
Study Design: Time-series
N: 23,000
# of Hospitals: 2
Statistical Analyses: Linear regression, GLM
Covariates: Variables that were significantly
associated with ER visits were retained in the
model
Statistical Package: SAS/STAT, SAS/ETS 6.0
Lag: l,2daysavgs
Number of monitors: 11
24-havg: 18.9 ug/m3,
SD= 15.0,Max<40
# of Stations: 15
Highest SES quintile: 1.00 [referent]
2ndQuintile: 1.11 [1.03,1.20]
3rd Quintile: 1.07 [1.00, 1.15]
4th Quintile: 1.09 [1.02, 1.17]
Lowest SES quintile: 1.09 [1.02, 1.16]
SO2 Children:
O3 r = 0.72, p = 0.004
CO Adults:
PM10 r = 0.428, p = 0.083
Emergency visits for asthma increased with
increased levels of NO2 for children but not for
adults.
SO2; r = -0.22 Categorical analysis (via ANOVA) p-value and
O3; r = 0.17 Pearson correlation coefficient (r) using
TSP' r = 0 23 continuous data comparing daily air pollutant
levels and daily number of ER visits
AgeGroup:3-12 13-21
LagO r=0.10 r=0.09
p = 0.0019 p< 0.001
Lagl r=0.12 r=0.04
p < 0.001 p = 0.0014
Lag 2 r=0.14 r=0.03
p < 0.001 p = 0.0066
O
HH
H
W
-------
O
to
O
O
oo
TABLE AX6.3-4 (cont'd). RESPIRATORY HEALTH EFFECTS OF OXIDES OF NITROGEN:
EMERGENCY DEPARTMENT VISITS
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels of NO2 &
Monitoring Stations
Copollutant Effects and Interpretation:
Correlations Relative Risk & Confidence Intervals (95%)
ASIA (cont'd)
X
I
o
H
6
o
o
H
O
o
H
W
Hwang and Chan (2002)
Taiwan
Period of Study:
1998
Tanakaetal. (1998)
Kushiro, Japan
Period of Study: 1992-
1993
Outcome(s) (ICD-9): Lower Respiratory
Disease (LRD) (466, 480-6) including acute
bronchitis, acute bronchiolits, pneumonia
Age groups analyzed: 0-14, 15-64, >65, all
ages
Study Design: Time-series
Catchment area: Clinic records from
50 communities
Statistical Analyses: Linear regression,
GLM
Covariates: Temperature, dew point
temperature, season, day of wk, holiday
Statistical Package: NR
Lag: 0,1,2 days and avgs
Outcome(s): Asthma
ICD-9 Code(s): NR
Age groups analyzed: 15-79
Study Design: Time-series
N: 102
# of Hospitals: 1
Statistical Analyses: Poisson regression
Covariates: Temperature, vapor pressure,
barometric pressure, relative humidity, wind
velocity, wind direction at maximal velocity
Statistical Package: NR
24-hr avg: 23.6 ppb,
SD = 5.4, Range: 13.0,34.1
NO2 24-h avg
9.2 ± 4.6 ppb in fog
11.5 ± 5.7 in fog free days
Max NO2 24-h avg <30 ppb
SO2
PM10
03
CO
No correlations
for individual
pollutants.
Increment: 10% change in NO2 (natural avg)
which is equivalent to 2.4 ppb. NOTE: The
percent change is for the rate of clinic use NOT
for relative risk for adverse effect.
Increased clinic visits for lower respiratory
disease (LRD) by age group
Colinearity of 0-14yrs
pollutants 1.3% [1.0,1.6] lag 0
prevented use of 15-64 yrs
multipollutant 1.5% [1.3,1.8] lag 0
models >65 yrs
1.8% [1.4,2.2] lag 0
All ages
1.4% [1.2,1.6] lag 0
Increment: 15 ppb
N02; r = NR
S02; r = NR
SPM (TSP);
r=0.70
O3; r = NR
Nonatopic
OR 0.62 [0.45, 0.84]
Atopic
OR 0.81 [0.67,0.97]
* Default GAM
+Did not report correction for over-dispersion
NR: Not Reported
APHEA: Air Pollution and Health: a European Approach
O
HH
H
W
-------
O
to
O
O
oo
TABLE AX6.3-5. RESPIRATORY HEALTH EFFECTS OF OXIDES OF NITROGEN: GENERAL
PRACTITIONER/CLINIC VISITS
X
I
o
H
6
o
o
H
O
o
H
W
Reference, Study
Location, & Period
Mean Levels
Outcomes, Design, & Methods & Monitoring Stations
Copollutants
Correlations
Effects:
Relative Risk & Confidence Intervals (95%)
NORTH AMERICA
Sinclair and Tolsma
(2004)
Atlanta, GA
Period of Study:
8/1998-8/2000
Hernandez-Garduno
etal. (1997)
Mexico City, Mexico
Period of Study: May
15, 1992 -January 31,
1993
Outcome(s) (ICD9): Asthma (493), UPJ 1-h max: 51.22 (18.54) ppb
(460-466, 477), LPJ (466.1, 480-486)
Age groups analyzed: <18, >18 Monitors: 1 ARIES monitor in
Study Design: Time-series downtown Atlanta
N: 232,350
Statistical Analyses: Poisson regression
with GLM
Covariates: Season, day of wk, federal
holiday, study month, long-term trend
Statistical Package: SAS v 8.02
Lag: 0-2,3-5,6-8
Outcome(s): Respiratory illness Number of Stations: 5
ICD9: NR
Age groups analyzed: <15, 15+,
all ages (0-92)
Study Design: Time-series
N: 24,113
Number of Clinics: 5
Statistical Analyses: Cross-correlation,
linear regression and autoregressive error
model analyses
Covariates: Long-term trend, day of wk,
temperature, humidity
Statistical Package: SAS
Lag: 0-6
PM2.5,
PM10-2.5
PM10
PM2.5
components
PMuf
Polar VOCs
03
SO2
S02
03
CO
NOX
No NO2 results presented because they were not
statistically significant for any lag periods
examined.
Increment: Maximum NO2 concentration of all
days-MeanNO2 concentration of all days
<14 yrs:
NO2lagO: RR 1.29 ±0.09 (p< 0.01)
NO2lag6: RR 1.18 ± 0.09 (p> 0.05)
15+ yrs:
NO2lagO: RR 1.14 ± 0.07 (p< 0.05)
NO2lag6: RR 1.10 ± 0.06 (p> 0.05)
All ages:
NO2lagO: RR 1.43 ± 0.15 (p< 0.01)
NO2lag6: RR 1.29 ± 0.15 (p> 0.05)
O
HH
H
W
-------
O
to
O
O
oo
X
I
O
TABLE AX6.3-5 (cont'd). RESPIRATORY HEALTH EFFECTS OF OXIDES OF NITROGEN: GENERAL
PRACTITIONER/CLINIC VISITS
Reference, Study
Location, & Period Outcomes, Design, & Methods
Mean Levels
& Monitoring Stations
Copollutants
Correlations
Effects:
Relative Risk & Confidence Intervals
(95%)
NORTH AMERICA (cont'd)
Villeneuve et al. Outcome(s) (ICD9): Allergic Rhinitis
(2006) (177)
Toronto, ON, Canada Age groups analyzed: >65
Study Design: Time-series
Period of Study: N: 52,691
1995-2000 Statistical Analyses: GLM, using natural
24-h avg: 25.4 ppb,
SD = 7.7
IQR: 10.3 ppb,
Range 9.2, 71.7
Number of stations: 9
SO2
03
CO
PM2.5
PMiQ-2 5
PM10
Increment: 10.3 ppb (IQR)
All results estimated from Stick Graph:
All Yr:
Mean Increase: 1.9% [-0.2, 3.8] lag 0
Days: 2,190
splines (more stringent criteria than
default)
Covariates: Day of wk, holiday,
temperature, relative humidity,
aero-allergens
Season: All yr; Warm, May-Oct;
Cool, Nov-Apr
Statistical Package: S-Plus
Lag: 0-6
Warm:
Mean Increase: 0.1% [-3.2, 3.8] lag 0
Cool:
Mean Increase: 1.4% [0.0, 5.9] lag 0
H
6
o
o
H
O
O
H
W
O
HH
H
W
-------
O
to
O
O
oo
TABLE AX6.3-5 (cont'd). RESPIRATORY HEALTH EFFECTS OF OXIDES OF NITROGEN: GENERAL
PRACTITIONER/CLINIC VISITS
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels
& Monitoring Stations
Copollutants
Correlations
Effects:
Relative Risk & Confidence Intervals (95%)
EUROPE
X
Oi
O
Oi
H
6
o
o
H
O
O
H
W
O
HH
H
W
Hajatetal. (1999)
London, United
Kingdom
Period of Study: 1992-
1994
Outcome (ICD9): Asthma (493); Lower
respiratory disease (464, 466,476, 480-3,
490-2, 485-7, 4994-6, 500, 503-5, 510-5)
Age groups analyzed: 0-14; 15-64; 65+;
all ages
Study Design: Time-series analysis
Statistical Analysis: Poisson regression,
APHEA protocol
Covariates: Long-term trends, seasonality,
day of wk, temperature, humidity
Seasons: Warm, Apr-Sep; Cool, Oct-Mar;
Allyr
Dose-response investigated?: Yes
Statistical package: SAS
Lag: 0-3 days, cumulative
All yr 24-h avg: 33.6ppb,
SD= 10.5
Warm: 32.8(19.8)
Cool: 34.5(10.1)
10th-90th all yr percentile:
24ppb
SO2;r = 0.61 Increment: 24 ppb (90th-10th percentile)
BS; r = 0.70 Asthma
CO; r = 0.72 All ages 2.1% [-0.7, 4.9] lag 0; 3.1%
PM10;r=0.73 [-0.4, 6.7] lag 0-1
O3;r= -0.10 0-14 yrs 6.1% [1.2, 11.3] lag 1;6.9%
[1.7, 12.4] lag 0-1
Warm: 13.2% [5.6, 21.3] lag 1
Cool: -0.1% [-6.3, 6.5] lag 1
15-64 yrs 3.0% [-0.7, 6.7] lag 0; 3.1%
[-1.6, 7.9] lag 0-3
Warm: 3.3% [-2.0, 8.9] lag 0
Cool: 2.6% [-2.3, 7.7] lag 0
65+ yrs 9.9% [1.6, 18.7] lag 2; 5.3%
[-3, 14.3] lag 0-3
Warm: 18.6% [6.3, 32.4] lag 2
Cool: -0.5% [-9.6,11.8] lag 2
Lower Respiratory disease
All ages 1.3% [-0.4, 3.0] lag 1; 1.2%
[-0.7, 3.1] lag 0-2
0-14 yrs 4.8% [1.3, 8.3] lag 2; 4.5%
[0.4, 8.7] lag 0-3
Warm: 1.4% [-3.8, 6.9] lag 2
Cool: 7.2% [2.8, 11.6] lag 2
15-64 yrs 1.1% [-1.1, 3.4] lag 2; 0.8% [-1.8, 3.5] lag 0-2
Warm: 2.3% [-1.2, 5.9] lag 2
Cool: 0.2% [-2.6, 3.1] lag 2
65+-1.7% [-4.3, 1.1] lag 0
Warm: -1.7% [-5.9, 2.6] lag 0
Cool: -1.6% [-4.8,1.8]lagO
Two-pollutant model-Asthma
NO2 alone: 5.2% [0.8, 9.8]
N02/03: 6.7% [2.2, 11.4]
NO2/SO2: 3.9% [-1.2, 9.2]
NO2/PM10: 5.3% [-0.6, 11.6]
-------
O
to
O
O
oo
TABLE AX6.3-5 (cont'd). RESPIRATORY HEALTH EFFECTS OF OXIDES OF NITROGEN: GENERAL
PRACTITIONER/CLINIC VISITS
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels
& Monitoring Stations
Copollutants
Correlations
Effects:
Relative Risk & Confidence Intervals (95%)
EUROPE (cont'd)
Hajatetal. (1999)
(cont'd)
Two-pollutant model - Lower Respiratory disease
NO2 alone 4.2% [1.1, 7.3]
N02/034.9%[1.8,8.2]
N02/S022.5%[-1.1,6.2]
N02/PM103.5%[0.1,6.9]
X
I
o
H
6
o
o
H
O
o
H
W
Hajat*etal. (2001)
London, United
Kingdom
Period of Study:
1992-1994
Outcome (ICD9): Allergic Rhinitis (477)
Age groups analyzed: 0-14; 15-64; 65+;
all ages
Study Design: Time-series analysis
N: 4,214
Statistical Analysis: Poisson regression,
GAM
Covariates: long-term trends, seasonality,
day of wk, temperature, humidity,
variation in practice population, counts for
lagged allergic pollen measures, daily
number of consultations for influenza
Dose-response investigated? Yes
Statistical package: S-Plus
Lag: 0-6 days, cumulative
NO224-havg: 33.6ppb,
SD=10.5
# of Stations: 3,
r= 0.7-0.96
SO2;r = 0.61
BS;r=0.70
CO; r= 0.72
PM10;r=0.73
O3;r=-0.10
Increment: 24 ppb (90th-10th percentile)
Single-pollutant model
<\ to 14 yrs
11.0% [3.8, 18.8] lag 4
12.6% [4.6, 21.3] lag 0-4
15 to 64 yrs
5.5% [2.0, 9.1] lag 6
11.1% [6.8, 15.6] lag 0-6
>64 yrs - too small for analysis
Two-pollutant models
-------
O
to
O
O
oo
TABLE AX6.3-5 (cont'd). RESPIRATORY HEALTH EFFECTS OF OXIDES OF NITROGEN: GENERAL
PRACTITIONER/CLINIC VISITS
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels
& Monitoring Stations
Copollutants
Correlations
Effects:
Relative Risk & Confidence Intervals (95%)
EUROPE (cont'd)
X
O
oo
Hajat*etal. (2002)
London, United
Kingdom
Period of Study:
1992-1994
Outcome (ICD9): Upper Respiratory
Disease, excluding Rhinitis (460-3, 465,
470-5, 478)
Age groups analyzed: 0-14; 15-64; 65+;
all ages
Study Design: Time-series analysis
Statistical Analysis: Poisson regression,
GAM
Covariates: Long-term trends, seasonality,
day of wk, holidays, temperature,
humidity, variation in practice population,
counts for lagged allergic pollen measures,
daily number of consultations for influenza
Seasons: Warm: Apr-Sep;
Cool: Oct-Mar
Dose-response investigated?: Yes
Statistical package: S-Plus
Lag: 0,1,2,3 days
NO224-havg: 33.6ppb,
SD= 10.5
Warm: (Apr-Sep)
Mean: 32.8ppb,
SD=10.1
Cool: (Oct-Mar)
Mean: 34.5 ppb,
SD= 10.1
# of Stations: 3
SO2;r = 0.61
BS;r=0.70
CO; r= 0.72
PM10;r=0.73
03;r=-0.10
Increment (90th-10th percentile): All yr: 24 ppb;
Warm season: 25.8 ppb;
Cool season: 22.1 ppb
Single-pollutant model
Allyr
0-14 yr 2.0% [-0.3, 4.3] lag 3
15-64 yrs 5.1% [2.0, 8.3] lag 2
>65 yrs 8.7% [3.8, 13.8] lag 2
Warm
0-14 yrs 2.5% [-0.9, 6.1] lag 3
15-64 yrs 6.7% [3.7, 9.8] lag 2
>65 yrs 6.6% [-1.1,14.9] lag 2
Cool
0-14 yrs 1.7% [-1.1, 4.6] lag 3
15-64 yrs 1.2% [-1.3, 3.9] lag 2
>65 yrs 9.4% [2.8, 16.4] lag 2
H
6
o
o
H
O
O
H
W
O
HH
H
W
Two-pollutant models
0-14 yrs
NO2&O3: 1.7% [-0.6, 3.9]
N02 & S02: 2.2% [-0.4, 5.0]
N02&PM10: 1.5% [-1.7, 4.8]
For 15-64 yrs
NO2&O3: 4.4% [2.2, 6.8]
N02&S02: 4.4% [1.6, 7.2]
NO2 & PM10: 2.7% [-0.5, 5.9]
For >65 yrs
NO2&O3: 8.1% [3.0, 13.6]
NO2&SO2: 8.6% [2.1,15.4]
NO2&PM10: 4.3% [-2.8, 11.8]
-------
O
to
O
O
oo
TABLE AX6.3-5 (cont'd). RESPIRATORY HEALTH EFFECTS OF OXIDES OF NITROGEN: GENERAL
PRACTITIONER/CLINIC VISITS
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels
& Monitoring Stations
Copollutants
Correlations
Effects:
Relative Risk & Confidence Intervals (95%)
EUROPE (cont'd)
X
O
VO
Chardon et al. (2007)
Greater Paris, France
Period of Study:
2000-2003
Outcome (ICD9): Asthma, URD, LRD
Age groups analyzed: 0-14; 15-64; 65+;
all ages
Study Design: Time-series analysis
Statistical Analysis: Poisson regression,
GAM
Covariates: Long-term trends, seasonality,
day of wk, holidays, temperature,
humidity, counts for lagged allergic pollen
measures, daily number of consultations
for influenza
Seasons: Warm: Apr-Sep;
Cool: Oct-Mar
Dose-response investigated?: Yes
Statistical package: R software
Lag: 0,1,2, 3,0-1,0-2, 0-3 days
24-h avg: 44.4 (14.92) ug/m3
Median: 43.6
IQR: 33.7-53.2
Range: 12.3-132.8
Number of monitors: 12-15
PM10;r = 0.68
PM2.5; r = 0.68
Increment: 10 ug/m
URD: 0.7% [-0.9,2.3] lag 0-3
LRD: 1.1% [-0.7, 2.9], lag 0-3
Asthma: -0.3% [-3.3, 2.7], lag 0-3
* Default GAM
+ Did not report correction for over-dispersion
NR: Not Reported
APHEA: Air Pollution and Health: a European Approach
H
6
o
o
H
O
O
H
W
O
HH
H
W
-------
O
to
O
O
oo
TABLE AX6.3-6. HUMAN HEALTH EFFECTS OF OXIDES OF NITROGEN: CVD HOSPITAL ADMISSIONS
AND VISITS: UNITED STATES AND CANADA
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels
& Monitoring Stations
Copollutants Effects: Relative Risk or Percent Change &
(Correlations) Confidence Intervals ([95% Lower, Upper])
X
Burnett etal. (1997a)
Metropolitan Toronto
(Toronto, North York,
East York, Etobicoke,
Scarborough, York),
Canada
Study period:
1992-1994, 388 days,
summers only
Outcome(s) (ICD9): IHD
410-414; Cardiac Dysrhythmias 427;
Heart failure 428. All Cardiac 410-414,
427,428. Obtained from hospital
discharge data.
Population: 2.6 Million residents
Study design: Time-series
Age groups analyzed: all
# Hospitals: NR
Statistical analysis: relative risk
regression models, GEEs.
Covariates: Adjusted for long-term
trends, seasonal and subseasonal
variation, day of the wk, temperature,
dew point
Seasons: Summer only
Dose response: Figures presented
Statistical package: NR
Lag: 1-4 days
NO2 daily 1 -h max (ppb): H+ (0.25)
Mean: 38.5 SO4(0.34)
CV: 29 TP(0.61)
Min: 12 FP (0.45)
25th percentile: 31 CP(0.61)
50th percentile: 38 COH(0.61)
75th percentile: 45 O3 (0.07)
Max: 81 SO2 (0.46)
CO (0.25)
# of Stations: 6-11
(Results are reported for
additional metrics including
24-h avg and daytime avg (day))
Results reported for RR for an IQR increment
increase in NO2. T ratio in parentheses.
All Cardiac Disease
Single-pollutant model
1.049 (3.13), daily avg over4 days, lag 0
Multipollutant model
1.30 (1.68), w/N02,03, S02,
Objective of study was to evaluate the role of
particle size and chemistry on cardio and
respiratory diseases. NO2 attenuated the effect
of particulate in this study.
H
6
o
o
H
O
O
H
W
O
HH
H
W
-------
O
to
O
O
oo
TABLE AX6.3-6 (cont'd). HUMAN HEALTH EFFECTS OF OXIDES OF NITROGEN: CVD HOSPITAL ADMISSIONS
AND VISITS: UNITED STATES AND CANADA
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels
& Monitoring Stations
Copollutants Effects: Relative Risk or Percent Change &
(Correlations) Confidence Intervals ([95% Lower, Upper])
X
H
6
o
o
H
O
O
H
W
Burnett etal. (1999)*
Metropolitan Toronto
(Toronto, North York,
East York, Etobicoke,
Scarborough, York),
Canada
Study Period:
1980-1995, 15 yr
Outcome(s) (ICD9): IHD
410-414; Cardiac Dysrhythmias 427;
Heart failure 428; All Cardiac 410-414,
427, 428; Cerebrovascular Disease
Obtained from hospital discharge data
430-438; Peripheral Circulation Disease
440-459.
Population: 2.13-2.42 million residents
Study Design: Time-series
Statistical Analysis: GAMs with LOESS
smoothers to remove temporal and
weather related trends, stepwise
regression used to select minimum
number of air pollutants in multipollutant
models.
Covariates: Long-term trends, seasonal
variation, day of wk, temperature, and
humidity.
Statistical Package: SPLUS, SAS
Lag(s): 0-2 day
NO2 daily avg (ppb)
Mean: 25.2
5th percentile: 13
25th percentile: 19
50th percentile: 24
75th percentile: 30
95th percentile: 42
Max: 82
Multiple day avgs used in
models
PM2.5 (0.50)
PM10-2.5 (0.38)
PM10 (0.52)
CO (0.55)
SO2 (0.55)
O3(-0.04)
Results reported for % increase in hospital
admissions for an increment increase in NO2
equal to the mean value.
Single-Pollutant Models:
Dysrhythmias: 5.33 (1.73) 3-day avg, lag 0
Heart Failure: 9.48 (6.33), 1 day, lag 0
IHD: 9.73 (8.4) 2-day avg, lag 0
Cerebrovascular disease: 1.98 (1.34), 1 day,
lagO
Peripheral circulation: 3.57(1.78), 1-day,
lagO
Multipollutant Models:
Heart failure
6.89 (w/CO)
6.68 (w/ CO, PM2.5)
6.33(w/CO,PM2.5,PM10-2.5)
6.45 (w/ CO, PM2.5, PMio-2.5, PM10)
IHD
8.34 (w/CO, SO2)
7.76 (w/ CO, S02, PM2.5)
8.41 (w/ CO, SO2, PM2 5, PM10-2.5)
8.52 (w/ CO, SO2, PM2.5, PMio-2.5, PM10)
In multipollutant models, gaseous pollutants
were selected by stepwise regression. PM
variables were then added to the model.
O
HH
H
W
-------
O
to
O
O
oo
TABLE AX6.3-6 (cont'd). HUMAN HEALTH EFFECTS OF OXIDES OF NITROGEN: CVD HOSPITAL ADMISSIONS
AND VISITS: UNITED STATES AND CANADA
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels
& Monitoring Stations
Copollutants Effects: Relative Risk or Percent Change &
(Correlations) Confidence Intervals ([95% Lower, Upper])
Morris etal. (1995)
US (Chicago, Detroit,
LA, Milwaukee, NYC,
Philadelphia)
Study Period:
1986-1989, 4 yr
Outcome(s) (ICD9): CHF 428.
Daily Medicare hospital admission
records.
Study Design: Time-series
Statistical Analyses: GLM, negative
binomial distribution
Age groups analyzed: >65 yrs
Covariates: Temperature, indicator
variables for mo to adjust for weather
effects and seasonal trends, day of wk, yr
Statistical Software: S-PLUS
Lag(s): 0-7 day
NO2 1 h-max (ppm)
Mean: (SD)
LA: 0.077(0.028)
Chicago: 0.045(0.013)
Philadelphia: 0.054 (0.017)
New York: 0.064(0.022)
Detroit: 0.041 (0.015)
Houston: 0.041 (0.017)
Milwaukee: 0.040(0.014)
SO2 1 -h max Results reported for RR of admission for CHF
O3 1-h max associated with an incremental increase in NO2
COl-hmax oflOppb.
Correlations of CHF:
N02 with other LA: 1.15(1.10,1.19)
pollutants Chicago: 1.17(1.07,1.27)
strong. Philadelphia: 1.03(0.95,1.12)
New York: 1.07(1.02,1.13)
Multipollutant Detroit: 1.04(0.92,1.18)
models run. Houston: 0.99(0.88,1.10)
Milwaukee: 1.05 (0.89, 1.23)
X
RR diminished in multipollutant models
(4 copollutants) for all cities with the exception
of New York.
H
6
o
o
H
O
O
H
W
Dales et al (2006)
Canada (11 largest
cities)
Study period: January 1,
1986-December 31,
2000.
Outcomes (ICD-9): Asphyxia (799.0),
respiratory failure (799.1), dyspnea and
respiratory abnormalities (786.0),
respiratory distress syndrome (769),
unspecificed birth asphyxia in live-born
infant (768.9), other respiratory problems
after birth (770.8) and pneumonia (486)
Data from: Canaidian Institutue for
Health Information:
9542 records for patients aged birth to 27
days
Study Design: Time series
Statistical analysis: Random-effects
regression model
Statitsitical Software: S-PLUS
NO2 24-hour mean levels (ppb):
Calgary: 25.6
Edmonton: 24.6
Halifax: 15.1
Hamilton: 20.8
London: 20.0
Ottawa: 21.2
Saint John: 9.2
Toronto: 25.1
Vancouver: 19.0
Windsor: 24.9
Winnipeg: 15.2
Population weighted avg: 21.8
Range of
Pearson
pairwise
correlations
PM10: -0.26
to 0.69
O3: -0.55 to
0.05
SO2: 0.20 to
0.67
CO: 0.13 to
0.76
Pooled estimate of % increase in neonatal
resipartory hospital admissions (95% CI):
Interquartile range: 10.0
Single pollutant model: 2.94 (1 .93 to 3.95)
Multi pollutant model: 2.85 (1.68 to 4.02)
Multipollutant model restricted to days with
PM10 measures: 2.48 (1.18 to 3.80)
O
HH
H
W
-------
O
to
O
O
oo
TABLE AX6.3-6 (cont'd). HUMAN HEALTH EFFECTS OF OXIDES OF NITROGEN: CVD HOSPITAL ADMISSIONS
AND VISITS: UNITED STATES AND CANADA
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels
& Monitoring Stations
Copollutants Effects: Relative Risk or Percent Change &
(Correlations) Confidence Intervals ([95% Lower, Upper])
X
H
6
o
o
H
O
O
H
W
Wellenius et al. (2005b)
Birmingham, Chicago,
Cleveland, Detroit,
Minneapolis, New
Haven, Pittsburgh,
Seattle
Study Period:
Jan 1986-Nov 1999
(varies slightly
depending on city)
Outcome(s) IS, primary diagnosis of
acute but ill-defined cerebrovascular
disease or occlusion of the cerebral
arteries; HS, primary diagnosis of
intracerebral hemorrhage. ICD codes not
provided. Hospital admissions
ascertained from the Centers for
Medicare and Medicaid Services. Cases
determined from discharge data were
admitted from the ER to the hospital.
NIS: 155,503
NHS: 19,314
Study Design: Time-stratified Case-
crossover. Control days chosen such that
they fell in same mo and same day of wk.
Design controls for seasonality, time
trends, chronic and other slowly varying
potential confounders.
Statistical Analysis: 2-stage hierarchical
model (random effects), conditional
logistic regression for city effects in the
first stage.
Software package: SAS
Covariates:
Lag(s): 0-2, unconstrained distributed
lags
NO2 24-h (ppb)
10th: 13.71
25th: 18.05
Median: 23.54
75th: 29.98
90th: 36.54
NO2 data not available for
Birmingham, Salt Lake, and
Seattle.
PM10 (0.53)
CO, S02
Correlation
only provided
forPM
because study
hypothesis
involves PM
Results reported for percent increase in stroke
admissions for an incremental increase in NO2
equivalent to one IQR (11.93).
Ischemic Stroke: 2.94 (1.78, 4.12), lag 0
Hemorrhagic Stroke: 0.38 (-2.66, 3.51), lag 0
Multipollutant models not run.
O
HH
H
W
-------
O
to
O
O
oo
TABLE AX6.3-6 (cont'd). HUMAN HEALTH EFFECTS OF OXIDES OF NITROGEN: CVD HOSPITAL ADMISSIONS
AND VISITS: UNITED STATES AND CANADA
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels
& Monitoring Stations
Copollutants Effects: Relative Risk or Percent Change &
(Correlations) Confidence Intervals ([95% Lower, Upper])
Fung etal. (2005)
Windsor, Ontario,
Canada
Study Period:
Apr 1995-Jan 2000
Outcome(s) (ICD9): CHF 428;
IHD 410-414; dysrhythmias 427.
Hospital admissions from Ontario Health
Insurance Plan records.
Study Design: Time-series
Statistical Analysis: GLM
N: 11,632 cardiac admission, 4.4/day for
65+ age group
Age groups analyzed: 65+, <65 yr
Statistical Software: SPLUS
Lag(s): lag 0, 2, 3 day avg
NO2 l-hmax(ppb):
Mean(SD): 38.9(12.3)
Min: 0
Max: 117
SO2 (0.22)
CO (0.38)
O3 (0.26)
COH(0.39)
PM10 (0.33)
Results expressed as percent change associated
with an incremental increase in NO2 equivalent
totheIQR(16ppb)
Cardiac:
<65 age group:
0.7 (-5.5, 6.6)
2.1 (-3.7, 8.2)
3.7 (-2.9, 10.7)
65+ age group:
0.8 (-2.2, 3.9), lag 0
0.9 (-2.7, 4.6), 2-day avg (lag 0-1)
0.8 (-3.3, 5.0), 3-day avg (lag 0-2)
X
Effect for NO2 not observed in these data.
Association of SO2 with cardiac admissions
observed.
H
6
o
o
H
O
O
H
W
O
HH
H
W
-------
O
to
O
O
oo
TABLE AX6.3-6 (cont'd). HUMAN HEALTH EFFECTS OF OXIDES OF NITROGEN: CVD HOSPITAL ADMISSIONS
AND VISITS: UNITED STATES AND CANADA
X
H
6
o
o
H
O
O
H
W
Reference, Study
Location, & Period
Linn et at (2000)
Metropolitan Los
Angeles, USA
Outcomes, Design, & Methods
Outcome(s): All Patient Refined
Diagnosis Related Groups (based on
medicare diagnosis related groups).
Mean Levels
& Monitoring Stations
NO2 24-h (ppm)
Winter
Copollutants
(Correlations)
CO (0.84, 0.92)
O3 (-0.23, 0.11)
PM10 (-0.67, 0.8)
Effects: Relative Risk or Percent Change &
Confidence Intervals ([95% Lower, Upper])
Results reported as increase % increase in
admission for a 10 ppb increase in NO2. SD in
parentheses.
Study Period:
1992-1995
CVD APR-DRG 103-144;
Cerebrovascular APR-DRG 14-17 and
22; CHF APR-DRG 127; MI APR-DRG
111, 115,121; cardiac ARR APR-DRG
138; Occlusive Stroke APR-DRG 14.
Hospital admission records used to
ascertain cases.
Study Design: Time-series
Statistical Analyses: Poisson regression,
cubic spline smooth on time, indicator
variables to adjust for temperature and
rain.
Covariates: Day of wk, holidays, long-
term trend, seasonal variation,
temperature, humidity
Lag(s): 0-1
Seasons: Winter, Spring, Summer,
Autumn
Statistical Software: SPSS, SAS
Mean: (SD) 3.4 (1.3)
Range: 1.1,9.1
Spring
Mean(SD): 2.8(0.9)
Range: 1.1,6.1
Summer
Mean(SD): 3.4(1.0)
Range: 0.7,6.7
Autumn
Mean(SD): 4.1(1.4)
Range: 1.6,8.4
Range in correlations
depends on the season,
independent effects of
pollutants could not be
distinguished.
# Stations: 6+
CVD, lag 0
All Seasons: 1.4(0.2)
Winter: 1.6(0.4)
Spring: 0.1 (0.6)
Summer: 1.1 (0.5)
Autumn: 1.4(0.3)
Cerebrovascular, lag 0
All Seasons: 0.4(0.4)
Winter: -1.3(0.7)
Spring: 4.2(1.2)
Summer: 0.9(1.2)
Autumn: 0.7(0.6)
MI, lag 0 (yr round)
1.1(0.5)
CHF, lag 0 (yr round)
1.0 (0.5), winter 1.9 (0.9)
Cardiac Arrhythmia, lag 0
(yr round)
0.6(0.5)
Occlusive stroke, lag 0
(yr round)
2.0 (0.5), winter 2.7 (1.0), autumn 0.1 (0.05)
Significant effects observed in fall for occlusive
stroke and winter for CHF.
O
HH
H
W
-------
O
to
O
O
oo
TABLE AX6.3-6 (cont'd). HUMAN HEALTH EFFECTS OF OXIDES OF NITROGEN: CVD HOSPITAL ADMISSIONS
AND VISITS: UNITED STATES AND CANADA
X
Oi
H
6
o
o
H
O
O
H
W
O
HH
H
W
Reference, Study
Location, & Period
Lippmann et al. (2000*;
reanalysis Ito, 2003,
2004)
Windsor Ontario
(near Detroit MI)
Study period:
1992- 1994 (hospital
admissions - mortality
study spanned longer
period)
Outcomes, Design, & Methods
Outcome(s): IHD 410-414; dysrhythmias
427; heart failure 428; stroke 431-437.
Study Design: Time-series
Statistical Analysis: Poisson regression
GAM. Results of reanalysis by Ito 2003,
2004 with GLM are presented.
Lag(s): 0-3 day
Mean Levels Copollutants
& Monitoring Stations (Correlations)
NO2 24-h avg (ppb)
5th %:
25th %:
50th %:
75th %:
95th %:
Mean:
11
: 16
: 21
: 26
: 36
21.3
PM10 (0.49)
PM2.5 (0.48)
PM10-2.5 (0.32)
1^(0.14)
SO4(0.35)
03(0.14)
SO2 (0.53)
CO (0.68)
Effects: Relative Risk or Percent Change &
Confidence Intervals ([95% Lower, Upper])
Results reported for RR for incremental increase
in NO2 of 5th to 95th percentile (28 ppb).
IHD
(0.94, 1.10), lag 0
0.98(0.90, 1.06), lag 1
Dysrhythmias
0.98(0.86, 1.12), lag 0
1.03(0.90, 1.06), lag 1
Heart Failure
1. 00 (0.91, 1.09), lag 0
1.07(0.98, 1.17), lag 1
Stroke
0.99(0.90, 1.09), lag 0
0.98(0.89, 1.08), lag 1
Mann et al. (2002)
South coast air basin of
CA, U.S.
Study Period:
1988-1995, Syr
Outcome(s) IHD 410-414; or IHD with
accompanying diagnosis of CHF 428; or
Arrhythmia 426, 427; Ascertained
through health insurance records.
Study Design: Time-series
N: 54,863 IHD admissions
Age groups analyzed: <40; 40-59; >60.
Statistical Analysis: Poisson regression,
GAM with cubic splines, results pooled
across air basins using inverse variance
weighting as no evidence of
heterogeneity was observed.
Covariates: study day, temperature,
relative humidity, day of wk.
Lag(s): 0-2, 2-4 day moving avg
Software: SPLUS
Seasons: Some analyses restricted to
Apr-Oct
NO2 24-h avg (ppb):
Exposure assigned for each
air basin based on health
insurance participant's zip
code.
Mean(SD): 37.2(15.7)
Range: 3.69, 138
Median: 34.8
# Stations: 25-35
O3 8 h-max
(-0.16,0.54)
CO 8-h max
(0.64, 0.86)
PM10 24-h avg
(0.36, 0.60)
Range depends on
air basin.
No multipollutant
models run.
Traffic pollution
generally
implicated in
findings.
Results reported for percent increase in
admissions for a 10-ppb incremental increase in
NO2.
All IHD: 1.68(1.08,2.28), lag 0
MI: 1.04(1.05, 3.02), lag 0
Other acute IHD: 1.75 (0.72, 2.78), lag 0
IHD w/ secondary diagnosis of Arrhythmia:
1.81 (0.78, 2.85), lag 0
IHD w/ secondary diagnosis of CHF:
2.32(0.69, 3.98), lag 0
IHD w/ no secondary diagnosis:
0.46 (-0.81, 1.74), lag 0
Effect of secondary diagnosis strongest in the
40-59 age group.
Group with secondary CHF may be sensitive
subpopulation or their vulnerability may be due
to greater prevalence of MI as the primary
diagnosis.
-------
O
to
O
O
oo
TABLE AX6.3-6 (cont'd). HUMAN HEALTH EFFECTS OF OXIDES OF NITROGEN: CVD HOSPITAL ADMISSIONS
AND VISITS: UNITED STATES AND CANADA
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels
& Monitoring Stations
Copollutants Effects: Relative Risk or Percent Change &
(Correlations) Confidence Intervals ([95% Lower, Upper])
X
Metzger et al. (2004)
Atlanta, GA
Period of Study: Jan
1993-Aug312000,4yr
Outcome(s): IHD 410-414; AMI 410;
Dysrhythmias 427; cardiac arrest 427.5;
congestive heart failure 428; peripheral
and cerebrovascular disease 433-437,
440, 443-444,451-453; atherosclerosis
440; stroke 436. ED visits from billing
records.
N: 4,407,535 visits, 37 CVD visits/day
# Hospitals: 31
Age groups analyzed: Adults >19,
elderly 56+
Statistical Analysis: Poisson regression,
GLM. Sensitivity analyses using GEE
and GAM (strict convergence criteria)
Covariates: long-term trends, mean and
dew point temp, relative humidity (cubic
splines)
Statistical Software: SAS
Season: Warm: Apr 15-Oct 14;
Cool: Oct 15-Apr 14.
Lag(s): 0-3 day
NO2 1-h max (ppb):
Median: 26.3
10th-90thpercentile
Range: 25,68
PM10 24 h (0.49)
O3 8-h max (0.42)
SO2 (0.34)
CO Ih (0.68)
1998-2000 Only
PM2.5 (0.46)
Course PM (.46)
Ultrafme PM (.26)
Water-soluble
metals (.32)
Sulfates(.17)
OC (0.63)
EC (.37)
OHC (0.3)
Multipollutant
models used. All
models specified
a priori.
Results presented for RR of an incremental
increase in 1-h max NO2 equivalent to 20 ppb
(1 SD).
All CVD: 1.025(1.012,1.039)
Dysrhythmia: 1.019 (0.994, 1.044)
CHF: 1.010(0.981,1.040)
IHD: 1.029(1.005,1.053)
PERI: 1.041(1.013,1.069)
Finger wounds: 1.010(0.993,1.027)
Lag is 3-day moving avg for results above.
NO2 effect was generally attenuated in two
pollutant models. The attenuation was strongest
in the period after 1998.
H
6
o
o
H
O
O
H
W
O
HH
H
W
-------
O
to
O
O
oo
TABLE AX6.3-6 (cont'd). HUMAN HEALTH EFFECTS OF OXIDES OF NITROGEN: CVD HOSPITAL ADMISSIONS
AND VISITS: UNITED STATES AND CANADA
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels
& Monitoring Stations
Copollutants Effects: Relative Risk or Percent Change &
(Correlations) Confidence Intervals ([95% Lower, Upper])
X
OO
H
6
o
o
H
O
O
H
W
Moolgavkar
(2000a,b,c)*
Cook County, IL,
Los Angeles County,
CA,
Maricopa County, AZ
1987-1995
Outcome(s) (ICD9): CVD
390-429; Cerebrovascular disease
430-448. Hospital admissions from CA
department of health database.
Age groups analyzed: 20-64, 65+ yrs
Study Design: Time-series
N: 118 CVD admissions/day
# Hospitals: NR
Statistical Analysis: Poisson regression,
GAM
Covariates: Adjustment for day of wk,
long-term temporal trends, relative
humidity, temperature
Statistical Package: SPLUS
Lag: 0-5 days
NO2 24-h avg (ppb)
Cook County:
Min: 7
Ql: 20
Median: 25
Q3: 30
Max: 58
N02 24-h avg (ppb) LA
County:
Min: 10
Ql: 30
Median: 38
Q3: 48
Max: 102
NO2 24-h avg (ppb)
Maricopa County:
Min: 2
Ql: 14
Median: 19
Q3: 26
Max: 56
PM10 (0.22-0.70)
PM2.5 (0.73)
(LA only)
CO (0.63-0.80)
SO2 (0.02- 0.74)
O3 (-0.23-0.02)
Two-pollutant
models (see results)
Results reported for percent change in hospital
admissions per 10-ppb increase in 24-h avg
NO2. T statistic in parentheses.
CVD, 65+:
Cook County
2.9 (10.2), lag 0
2.3 (6.7), lag 0, two-pollutant model (PM10)
2.9 (8.1), lag 0, two-pollutant model (CO)
2.8 (8.8), lag 0, two-pollutant model (SO2)
LA County
2.3 (16.7), lag 0
-0.1 (-0.5), lag 0, two-pollutant model (CO)
1.7 (8.0), lag 0, two-pollutant model (SO2)
Maricopa County
2.9(4.1), lag 0
-0.3 (-0.3), lag 0, two-pollutant model (CO)
2.6 (3.6), lag 0, two-pollutant model (SO2)
Cerebrovascular Disease, 65+:
Cook County
1.6(3.6)
LA County
1.1(5.7)
Effect size generally diminished with increasing
lag time. Increase in hospital admissions
(1.3 for CVD and 1.9 for Cerebrovascular) also
observed for the 20-64 age group.
O
HH
H
W
-------
O
to
O
O
oo
TABLE AX6.3-6 (cont'd). HUMAN HEALTH EFFECTS OF OXIDES OF NITROGEN: CVD HOSPITAL ADMISSIONS
AND VISITS: UNITED STATES AND CANADA
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels
& Monitoring Stations
Copollutants Effects: Relative Risk or Percent Change &
(Correlations) Confidence Intervals ([95% Lower, Upper])
X
Moolgavkar (2003)*
Cook County, IL,
Los Angeles County, CA,
Maricopa County, AZ
1987-1995
Outcome(s) (ICD9): CVD
390-429; Cerebrovascular disease
430-448 was not considered in the
reanalysis. Hospital admissions from CA
department of health database.
Age groups analyzed: 20-64, 65+ yrs
Study Design: Time-series
N: 118 CVD admissions/day
# Hospitals: NR
Statistical Analysis: Poisson regression,
GAM with strict convergence criteria
(10-8), GLM using natural splines
Covariates: Adjustment for day of wk,
long-term temporal trends, relative
humidity, temperature
Statistical Package: SPLUS
Lag: 0-5 days
NO2 24-h avg (ppb)
Cook County:
Min: 7
Ql: 20
Median: 25
Q3: 30
Max: 58
NO2 24-h avg (ppb)
LA County:
Min: 10
Ql: 30
Median: 38
Q3: 48
Max: 102
NO2 24-h avg (ppb)
Maricopa County:
Min: 2
Ql: 14
Median: 19
Q3: 26
Max: 56
PM10 (0.22-0.70)
PM2.5 (0.73)
(LA only)
CO (0.63-0.80)
S02 (0.02-0.74)
O3 (-0.23-0.02)
Two-pollutant
models (see results)
Results for CVD not shown but use of stringent
criteria in GAM did not alter results
substantially. However, increased smoothing of
temporal trends attenuated results for all gases.
H
6
o
o
H
O
O
H
W
O
HH
H
W
-------
O
to
O
O
oo
X
to
O
H
6
o
o
H
O
O
H
W
TABLE AX6.3-6 (cont'd). HUMAN HEALTH EFFECTS OF OXIDES OF NITROGEN: CVD HOSPITAL ADMISSIONS
AND VISITS: UNITED STATES AND CANADA
Reference, Study
Location, & Period
Peel et al. (2007)
Atlanta, GA
Study Period:
Outcomes, Design, & Methods
Outcome(s) (ICD9): IHD 410-414;
dysrhythmia 427; CHF 428; peripheral
vascular and cerebrovascular disease
433-437, 440, 443, 444, 451-453.
Mean Levels
& Monitoring Stations
NO2 l-hmax(ppb):
Mean(SD): 45.9(17.3)
10th: 25.0
90th: 68.0
Copollutants
(Correlations)
PM10 24-h avg
O3 8-h max
SO2 1-hmax
CO 1-hmax
Effects: Relative Risk or Percent Change &
Confidence Intervals ([95% Lower, Upper])
Results expressed as OR for association of CVD
admissions with a 20-ppb incremental increase
inN02.
Jan 1993-Aug 2000
Computerized billing records for ED
visits.
Comorbid conditions: Hypertension 401 -
405; diabetes 250; dysrhythmia 427, CHF
428; atherosclerosis 440; COPD 491,
492, 496; pneumonia 480-486; upper
respiratory infection 460-465, 466.0;
asthma 493, 786.09.
# Hospitals: 31
N: 4,407,535 visits
Study Design: Case-crossover. CVD
outcomes among susceptible groups with
Comorbid conditions.
Statistical Analyses: Conditional logistic
regression.
Covariates: Cubic splines for
temperature and humidity included in
models. Time independent variables
controlled through design.
Statistical Software: SAS
Lag(s): 3-day avg, lagged 0-2 day
Correlations not
reported
Comorbid Hypertension:
IHD: 1.036(0.997,1.076)
Dysrhythmia: 1.095(1.030,1.165)
PERI: 1.031(0.987,1.076)
CHF: 1.037(0.985,1.090)
Comorbid Diabetes:
IHD: 1.003(0.95,1.059)
Dysrhythmia: 1.158 (1.046, 1.282)
PERI: 1.012(0.947,1.082)
CHF: 1.017(0.959,1.078)
O
HH
H
W
-------
O
to
O
O
oo
TABLE AX6.3-6 (cont'd). HUMAN HEALTH EFFECTS OF OXIDES OF NITROGEN: CVD HOSPITAL ADMISSIONS
AND VISITS: UNITED STATES AND CANADA
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels &
Monitoring Stations
Copollutants Effects: Relative Risk or Percent Change &
(Correlations) Confidence Intervals ([95% Lower, Upper])
X
ON
to
Schwartz, (1997)*
Tucson, AZ
Study Period:
Jan 1988-Dec 1990.
Stieb et al. (2000) *
Saint John, New
Brunswick Canada
Study Period:
July 1992-Mar 1996
Outcome(s) (ICD9): CVD 390-429.
Ascertained from hospital discharge
records.
Study Design: Time-series
Statistical Analysis: Poisson regression,
GAM
Age groups analyzed: 65+
Covariates: Long-term and seasonal
trends, day of the wk, temperature, dew
point,
Statistical Software: SPLUS
Outcome(s): Angina pectoris; MI;
dysrhythmia/conduction disturbance;
CHF; All Cardiac. ED Visits collected
prospectively.
Study Design: Time-series
Statistical Analyses: Poisson regression,
GAM, LOESS smooth for temporal and
weather related variables
N: 19,821 ER visits
# Hospitals: 2
Lag(s): 1-8 days
NO2 24-h avg (ppb):
Mean: 19.3
10th: 9.9
25th: 13.2
50th: 19
75th: 24.6
90th: 29.8
NO2 24-h avg (ppb)
Mean(SD): 8.9(5.5)
95th: 19
Max: 35
NO2 max (ppb)
Mean(SD): 20.2
95th: 39
Max: 82
PM10 (0.326)
03 (-0.456)
SO2 (0.482)
CO (0.673)
CO (0.68)
H2S.(0.07)
03(-0.02)
S02(0.41)
PM10 (0.35)
PM2.5 (0.35)
H+(-0.25)
S04(0.33)
COH (0.49)
Results reported as a percent increase in admission for
an increment in NO2 equivalent to the IQR (11.4 ppb).
CVD 0.69% (-2.3, 3.8)
Tucson selected to minimize correlations between
pollutants. Since there was no association between
NO2 and admissions, author suggests results for CO
not confounded by NO2.
Results reported for percent change in admissions
based on a single-pollutant model for incremental
increase in NO2 equivalent to 1 IQR (8.9 ppb)
Cardiac visits:
-3.9, p-value = 0.136, lag 2, allyr
10.1, p-value = 0.051, lag 5, May-Sept
For specific CVD diagnoses, ARR and CHF
approached significance. NO2 was not a focus of this
paper.
H
6
o
o
H
O
O
H
W
O
HH
H
W
Tolbert et al. (2007)
Atlanta, GA
Study Period:
1993-2004
Outcome(s) (ICD9): All CVD including
IHD 410-414; cardiac dysrhythmias 427;
CHF 428; peripheral vascular and
cerebrovascular disease 433-437, 440
443-445,451-453. Emergency visits
primary and secondary diagnostic codes.
Study Design: Time-series
Statistical Analysis: GLM, cubic splines
with monthly knots, indicators for
season, day of wk, holiday, excluded
days with missing exposure data
Statistical Software: SAS
Lag(s): 3 day moving avg (0-2 d)
NO2 l-hmax(ppb):
Mean: 43.2
Minimum: 1.0
10th: 22
25th: 31
Median: 41
75th: 54
90th: 66
Maximum: 181
PM10 (0.53)
03 (0.44)
CO (0.70)
SO2 (0.36)
Course PM (0.70)
PM2.5 (0.47)
PM25SO4(0.14)
PM2.5 EC (0.64)
PM2.5 OC (0.62)
PM2.5 TC (0.65)
PM2 5 water sol
metals (0.32)
OHC (0.24)
Results reported for RR based on incremental increase
of NO2 equivalent to 1 IQR (23 ppb):
Single pollutant model results:
CVD 1.015 (1.004, 1.025), lag 0-2
NO2 effect diminished in multipollutant models
containing CO and PM2 5TC (shown in figure).
-------
O
to
O
O
oo
TABLE AX6.3-6 (cont'd). HUMAN HEALTH EFFECTS OF OXIDES OF NITROGEN: CVD HOSPITAL ADMISSIONS
AND VISITS: UNITED STATES AND CANADA
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels &
Monitoring Stations
Copollutants Effects: Relative Risk or Percent Change &
(Correlations) Confidence Intervals ([95% Lower, Upper])
X
to
to
Villeneuve et al. (2006)
Edmonton, Canada
Study Period:
Apr 1992-Mar 2002
Outcome(s) (ICD9): Acute ischemic
stroke 434, 436; hemorrhagic stroke 430,
432; transient ischemic attack (TIA) 435;
Other 433, 437,438. ED visits supplied
by Capital Health.
N: 12,422 Stroke Visits
Catchment area: 1.5 million people
Study Design: Case-crossover, exposure
index time compared to referent time.
Time-independent variables controlled in
the design. Index and referent day
matched by day of wk.
Statistical Analysis: Conditional logistic
regression, stratified by season and
gender.
Covariates: Temperature and humidity
Statistical Software: SAS
Season: Warm: Apr-Sept;
Cool: Oct-Mar.
Lag(s): 0, 1, 3-day avg
NO2 24 h ppb:
Allyr
Mean(SD): 24(9.8)
Median: 22.0
25th: 16.5
75th: 30.0
IQR: 13.5
Summer
Mean(SD): 18.6(6.4)
Median: 17.5
25th: 14.0
75th: 22.0
IQR: 8
Winter
Mean(SD): 29.4(9.6)
Median: 28.5
25th: 22.5
75th: 35.5
IQR: 13.0
O3 24-h max
(-0.33)
O3 24-h avg
(-0.51)
SO2 25-h avg
(0.42)
CO 24-h avg
(0.74)
PM10 24-h avg
(0.34)
PM2524-havg
(0.41)
All yr correlations
summarized.
Results reported for an incremental increase in NO2
equivalent to one IQR NO2.
Ischemic Stroke, Summer
1.17(1.05, 1.31), lag 0
1.18(1.05, 1.31), lag 1
1.26(1.09, 1.46), 3-day avg
Hemorrhagic stroke, Summer
1.16(0.99,1.37)
1.14(0.97,1.35)
1.18(0.95,1.46)
TIA not associated with increase in NO2. Above
results are strongest effects, which were observed
during summer.
Authors attribute NO2 effect to vehicular traffic since
NO2 and CO are highly correlated.
H
6
o
o
H
O
O
H
W
O
HH
H
W
-------
O
to
O
O
oo
TABLE AX6.3-6 (cont'd). HUMAN HEALTH EFFECTS OF OXIDES OF NITROGEN: CVD HOSPITAL ADMISSIONS
AND VISITS: UNITED STATES AND CANADA
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels &
Monitoring Stations
Copollutants Effects: Relative Risk or Percent Change &
(Correlations) Confidence Intervals ([95% Lower, Upper])
X
I
to
Wellenius et al. (2005a)
Allegheny County, PA
(near Pittsburgh)
Study Period:
Jan 1987-Nov 1999
Outcome(s): CHF 428. Cases are
Medicare patients admitted from ER with
discharge of CHF.
Study Design: Case-crossover, control
exposures same mo and day of wk,
controlling for season by design.
Statistical Analysis: Conditional logistic
regression
N: 55,019 admissions, including repeat
admissions, 86% admitted < 5 times
Age groups analyzed: 65+ yrs (Medicare
recipients)
Covariates: Temperature and pressure.
Effect modification by age, gender,
secondary diagnosis arrhythmias, atrial
fibrillation, COPD, hypertension, type 2
diabetes, AMI within 30 days, angina
pectoris, IHD, acute respiratory infection.
Statistical Software: SAS
Lag(s): 0-3
NO2 24-h avg (ppb):
Mean(SD): 26.48(8.02)
5th: 15.10
25th: 20.61
Median: 25.70
75th: 31.30
95th: 4102
# Stations: 2
PM10 (0.64)
CO (0.70)
O3(-0.04)
SO2 (0.52)
Results reported for the percent increase in
admissions for an increment of NO2 equivalent
to one IQR (11 ppb).
CHF, single-pollutant model
4.22 (2.61, 5.85), lag 0
CHF, two-pollutant model
4.05 (1.83, 6.31), adjusted for PM10
-0.37 (-2.59, 1.89), adjusted for CO
3.73 (2.10, 5.39), adjusted for O3
3.79 (1.93, 5.67), adjusted for SO2
CHF admission was 3x higher among those
with history of MI.
H
6
o
O
H
/O
o
H
W
O
HH
H
W
-------
O
to
O
O
oo
TABLE AX6.3-6 (cont'd). HUMAN HEALTH EFFECTS OF OXIDES OF NITROGEN: CVD HOSPITAL ADMISSIONS
AND VISITS: UNITED STATES AND CANADA
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels
& Monitoring Stations
Copollutants Effects: Relative Risk or Percent Change &
(Correlations) Confidence Intervals ([95% Lower, Upper])
Zanobetti and Schwartz Outcome(s) (ICD9): MI 410.
(2006)
Boston, MA
1995-1999
Admissions through the emergency room
from Medicare claims.
Age group analyzed: 65+ yrs
Study Design: Case-crossover, control
days matched yr, mos and temperature
Statistical Analysis: Conditional logistic
regression
N: 15,578
Covariates: Temperature (regression
spline), day of wk
Seasons: Hot (Apr-Sept) and cold
Software: SAS
Lags: 0, 0-1 previous day avg
NO2 24-h avg ppb
5th: 12.59
25th: 18.30
Median: 23.20
75th: 29.13
95th:
90th-10th: 20.41
# Stations: 4
O3(-0.14)
BC (0.70)
CO (0.67)
PM2.5 (0.55)
PM non-traffic
(0.14)
(residuals from
model of PM25
regressed on BC)
Results reported for percent increase in
admissions for incremental increase in NO2
equivalent to the 90th-1 Oth percentiles (20.41 or
16.80 for 0-1, previous day avg).
MI
10.21(3.82, 15.61), lag 0
12.67 (5.82, 18.04), lag 0-1, previous day avg
Results suggest traffic exposure is responsible
for the observed effect. Effects more
pronounced in the summer season.
X
I
to
H
6
o
O
H
/O
o
H
W
*Default GAM
AMI Acute Myocardial
Infarction
ARR Arrhythmia
BC Black Carbon
COH coefficient of haze
CP Course Particulate
CVD Cardiovascular Disease
EC Elemental Carbon
FP Fine Particulate
HS Hemorrhagic Stroke
ICD9 International Classification of Disease,
9th Revision
IHD Ischemic Heart Disease
IS ischemic stroke
MI Myocardial Infarction
OC Organic Carbon
OHC Oxygenated
Hydrocarbons
PERI Peripheral Vascular and
Cerebrovascular Disease
PM Particulate Matter
PIH primary intracerebral hemorrhage
PNC Particle Number Concentration
SHS Subarachnoid hemorrhagic stroke
TP Total Particulate
UBRE Unbiased Risk Estimator
O
HH
H
W
-------
O
to
O
O
oo
TABLE AX6.3-7. HUMAN HEALTH EFFECTS OF OXIDES OF NITROGEN: CVD HOSPITAL ADMISSIONS
AND VISITS: AUSTRALIA AND NEW ZEALAND
X
I
to
H
6
o
O
H
/O
o
H
W
Reference, Study
Location, & Period
Barnett et al. (2006)
Australia and New
Zealand: Brisbane,
Canberra, Melbourne,
Perth, Sydney
Period of Study:
1998-2001
Outcomes, Design, & Methods
Outcome(s) (ICD9): All CVD 390-459;
ARR427; Cardiac disease 390-429;
Cardiac failure 428; IHD 410-413; MI
410; Stroke 430-438.
Ages groups analyzed: 15-64 yrs, >65yrs
Study Design: Time stratified, case-
crossover, multicity study
# of Hospitals: All ER admissions from
state government health departments
Statistical Analyses: Random effects
meta analysis, heterogeneity assessed
using 12 statistic.
Covariates: Matched analysis controlling
for long-term trend, seasonal variation,
and respiratory epidemics. Temperature
(current-previous day) and relative
humidity, pressure, extremes of hot and
cold, days of wk, holidays, day after
holiday, rainfall in some models.
Matched on copollutants.
Statistical Package: SAS
Lag: 0-3
Mean Levels &
Monitoring Stations
NO2 (ppb), Mean (range)
Auckland
1-hmax: 19.1(4.2-86.3)
24-havg: 10.2(1.7-28.9)
Brisbane
1-hmax: 17.3(4-44.1)
24-havg: 7.6(1.4-19.1)
Canberra
1-hmax: 17.9(0-53.7)
24-havg: 7.0 (0-22. 5)
Christchurch
1-hmax: 15.7(1.2-54.6)
24-havg: 7.1 (0.2-24.5)
Melbourne
1-hmax: 23.2(4.4-48)
24-havg: 11.7(2-29.5)
Perth
1-hmax: 21.3(4.4-48)
24-havg: 9.0(2-23.3)
Sydney
1-hmax: 22.6(5.2-51.4)
24-havg: 11.5(2.5-24.5)
24havgIQR: 5.1
# of Stations: 1-13
depending on the city
Copollutants
(Correlations)
PM10 24 h
C024h
S02 24 h
O38h
BS24h
Matched analysis
conducted to
control for
copollutants
Effects: Relative Risk or Percent Change &
Confidence Intervals ([95% Lower, Upper])
Results reported for % change in hospital
admissions associated with one IQR increase in
24-havg NO2, lag 0-1.
Arrhythmia
>65: 0.4 (-1.8, 2.6)
15-64: 5.1(2.2,8.1)
Cardiac
>65: 3.4(1.9,4.9)
15-64: 2.2(0.9,3.4)
Cardiac failure
>65: 6.9(2.2,11.8)
15-64: 4.6(0.1,6.1)
IHD
>65: 2.5(1.0,4.1)
15-64: 0.7 (-1.0, 2.4)
MI
>65: 4.4(1.0,8.0)
15-64: 1.7 (-1.1, 2.4)
All CVD
>65: 3.0(2.1,3.9)
15-64: 1.7(0.6,2.8)
NO2 association became smaller when matched
with CO. Authors hypothesize that NO2 is a
good surrogate for PM, which may explain
these associations.
O
HH
H
W
-------
O
to
O
O
oo
TABLE AX6.3-7 (cont'd). HUMAN HEALTH EFFECTS OF OXIDES OF NITROGEN: CVD HOSPITAL ADMISSIONS
AND VISITS: AUSTRALIA AND NEW ZEALAND
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels &
Monitoring Stations
Copollutants Effects: Relative Risk or Percent Change &
(Correlations) Confidence Intervals ([95% Lower, Upper])
X
Oi
to
Oi
Simpson et al. (2005a)
Australia (Brisbane,
Melbourne, Perth,
Sydney).
Study Period:
Jan 1996-Dec 1999
Outcome(s) (ICD9): Cardiac disease 390-
429; IHD 410-413; stroke 430-438.
Study Design: Time-series.
Statistical analysis: APHEA2 protocol,
GAM (did not indicate use of stringent
convergence criteria), GLM with natural
splines, penalized splines. Random effects
meta-analysis with tests for homogeneity.
Age groups analyzed: All, 15-64, 65+
Covariates: long-term trend, temperature,
humidity, day of wk, holidays, influenza
epidemics
Software package: SPLUS, R
Lag(s): 1-3 days
NO l-hmax(ppb):
Mean (range):
Brisbane: 21.4(2.1,63.3)
Sydney: 23.7(6.5,59.4)
Melbourne: 23.7 (4.4, 66.7)
Perth: 16.3(1.9,41.0)
PM10 24 h
PM2.5
BS24h
(0.29, 0.62)
03lh
CO8h
Not all correlations
reported. NO2
affect attenuated
slightly when
modeled with BS
but not with O3.
May be
confounding of
NO2 effect by
particulate.
Single-city results reported for percent increase
for an increment in 1-h max NO2 equivalent to
one IQR. Pooled results reported for an
increment of 1 ppb NO2.
Cardiac
All ages: 1.0023 (1.0016, 1.0030), lag 0-1
15-64: 1.0015 (1.0006,1.0025), lag 0
>65: 1.0018 (1.0011, 1.0025), lag 0-1
IHD
All ages: 1.0019(1.0010,1.0027)
>65: 1.0017(1.0007,1.0027)
No effect observed/reported for stroke.
Multipollutant results (65+ age group):
Cardiac:
1.0032 (1.0006, 1.0022), w/BS, lag 0-1
1.0032 (1.0024, 1.0039), w/O3, lag 0-1
Heterogeneity in CVD results among cities,
probably due to different pollutant mixtures,
may have affected the results.
H
6
o
o
H
O
O
H
W
Hinwood et al. (2006)
Perth, Australia
Study Period:
1992-1998
Outcome(s): All CVD unscheduled
admissions. Obtained from discharge
records using ICD9 Codes.
Age groups analyzed: All ages, 65+
Study design: Case-crossover, time-
stratified with 3-4 controls within same mo
Statistical Analysis: conditional logistic
regression
N: 26.5 daily CVD admissions
Seasons: Nov-Apr, May-Oct
NO2 24-h (ppb)
Mean: 10.3
SD = 5
10th percentile: 4.4
90th percentile: 17.1
NO2 1-h max (ppb)
Mean: 24.8
SD=10.1
10th percentile: 13.3
90th percentile: 37.5
O3 1 h, 8 h (- .06)
CO 8 h (.57)
BSP24h(.39)
Results reported for OR per incremental
increase of 1 ppb NO2.
All CVD (estimated from graph)
NO2 24 h65+: 1.005 (1.001, 1.006), lag 1
NO2 24 h all ages: 1.003(1.001, 1.007), lag 1
O
HH
H
W
# of Stations: 3
-------
O
to
O
O
oo
X
I
to
H
6
o
O
H
/O
o
H
W
TABLE AX6.3-7 (cont'd). HUMAN HEALTH EFFECTS OF OXIDES OF NITROGEN: CVD HOSPITAL ADMISSIONS
AND VISITS: AUSTRALIA AND NEW ZEALAND
Reference, Study
Location, & Period
Jalaludin et al. (2006)
Sydney, Australia
Period of Study:
Jan 1997-Dec 2001
Outcomes, Design, & Methods
Outcome(s) (ICD9): All CVD 390-459;
cardiac disease 390-429; IHD 410-413;
and cerebro vascular disease or stroke
430-438; Emergency room attendances
obtained from health department data.
Age groups included: 65+
Study Design: Time-series, multi-city
APHEA2 Protocol.
Statistical Analysis: GAM (with
appropriate convergence criteria) and
GLM Models. Only GLM presented.
T r\ "$
Lag: 0-3
Covariates: Daily avg temperature and
daily relative, humidity, long-term trends,
seasonality, weather, day of wk, public
school holidays, outliers and influenza
epidemics.
Dose response: Quartile analysis
Season: Separate analyses for warm
(Nov-Apr) and cool periods (May-Oct).
Mean Levels &
Monitoring Stations
NO2 daily 1-h avg
Mean: 32.2
SD = 7.4
Min: 5.2
Ql: 18.2
Median: 23
Q3: 27.5
Max: 59.4
# of Stations: 14
Copollutants
(Correlations)
BS 24-h avg (0.35)
PM10 24-h avg
(0.44)
PM2524-havg
(0.45)
CO 8-h avg (0.55)
O3 1-h avg (0.45)
SO2 24-h avg (0.56)
Two-pollutant
models to adjust for
copollutants
Effects: Relative Risk or Percent Change &
Confidence Intervals ([95% Lower, Upper])
Results reported for % change in hospital
admissions associated with one IQR increase in
l-havgNO2.
All CVD
2.32(1.45, 3.19), lag 0
0.45 (-0.52, 1.42), lag 1
1.31 (0.28, 2.35), lag 0-1
Cardiac Disease
2.00 (0.81, 3.20), lag 0
0.91 (-0.26,2.09), lag 1
1.78(0.54, 3. 04), lag 0-1
IHD
2.11 (0.34, 3. 91), lag 0
0.76 (-0.97, 2. 52), lag 1
1.73 (-0.10, 3.59), lag 0-1
Stroke
-1.66 (-3.80, 0.51) lag 0
-1.11 (-3.19, 1.02), lag 1
-1.68 (-3.90, 0.60), lag 0-1
Effect of NO2 attenuated when CO was included
in the model. NO2 effect most prominent
during the cool season.
O
HH
H
W
-------
O
to
O
O
oo
TABLE AX6.3-7 (cont'd). HUMAN HEALTH EFFECTS OF OXIDES OF NITROGEN: CVD HOSPITAL ADMISSIONS
AND VISITS: AUSTRALIA AND NEW ZEALAND
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels &
Monitoring Stations
Copollutants Effects: Relative Risk or Percent Change &
(Correlations) Confidence Intervals ([95% Lower, Upper])
X
to
oo
Morgan etal. (1998a)
Sydney, Australia
Study Period:
Jan 1990-Dec 1994
Outcome(s) (ICD9): Heart Disease 410,
413,427,428. Inpatient statistics
database for New South Wales Health
Department.
Study Design: Time-series
Statistical Analysis: Poisson regression,
GEE
# Hospitals: 27
Covariates: Daily mean temperature,
dew point temperature
Lag(s): 0-2 days, cumulative
Statistical Software: SAS
NO2 24-h avg (ppb):
Mean(SD): 15(6)
IQR: 11 ppb
10th-90th: 17
NO2 l-hmax(ppb):
Mean(SD): 29(3)
10-90th: 29 ppb
NO2 24-h max: 52
NO2l-hmax: 139
# Stations: 3-14
(1990-1994)
O3 1-hmax
(-0.086)
PM (0.533, 0.506)
Correlations for
24-h avg NO2
concentrations
Multipollutant
models
Results reported as percent increase in
admissions associated with an incremental
increase in 1-h max NO2 and 24-h avg
equivalent to the 10th-90th percentile.
Heart Disease:
24-h avg, lag 0
All ages: 7.52(5.21,9.88)
65+: 8.39(5.41,11.46)
0-64: 5.81(1.63,10.17)
1-hmax, lag 0
All ages: 6.08(3.63,8.59)
65+: 6.71(4.25,9.23)
0-64: 4.79(1.18,8.53)
65+: 6.68(3.61, 9.84) Particulate,O3
Results lost precision but did not change
substantially when stratified by age or when
24-h averaging time was used.
H
6
o
o
H
O
O
H
W
O
HH
H
W
-------
O
to
O
O
oo
TABLE AX6.3-7 (cont'd). HUMAN HEALTH EFFECTS OF OXIDES OF NITROGEN: CVD HOSPITAL ADMISSIONS
AND VISITS: AUSTRALIA AND NEW ZEALAND
X
to
VO
H
6
o
o
H
O
O
H
W
Reference, Study
Location, & Period
Petroeschevsky et al.
(2001)
Brisbane, Australia
Study Period:
Jan 1987-Dec 1994,
2,922 days
Outcomes, Design, & Methods
Outcome(s) (ICD9): CVD 390-459.
Hospital admissions, non-residents
excluded.
Study Design: Time-series
Statistical Analyses: Poisson regression,
APHEA protocol, linear regression and
GEEs
Age groups analyzed: 15-64,65+
Covariates: Temperature, humidity,
rainfall. Long-term trends, season, flu,
day of wk, holidays.
Statistical Software: SAS
Lag(s): lag 0-4, 3-day avg, 5 day avg
Mean Levels &
Monitoring Stations
NO2 1-h max (pphm)
Summer
Mean: 206
Min: 0.35
Max: 5.8
Fall
Mean: 2.56
Min: 0.70
Max' 5 85
Winter
Mean: 3.54
Min: 0.35
Max: 8.05
Spring
Mean: 3.12
Min: 0.55
Max: 15.58
Overall
Mean: 2.82
Min: 0.35
Max: 15.58
Copollutants Effects: Relative Risk or Percent Change &
(Correlations) Confidence Intervals ([95% Lower, Upper])
BSP Results reported for RR for CVD emergency
O3 admissions associated with a one-unit increase
SQ2 in NO2 1 -h max.
Correlation
between pollutants CVD 1 5-64 yrs
not reported. 0.986 (0.968, 1.005), lag 3
CVD 65+ yrs
0.990(0.977,1.003)
CVD all ages
0.987(0.976,0.998)
*Default GAM
AMI Acute Myocardial
Infarction
ARR Arrhythmia
BC Black Carbon
COH coefficient of haze
CP Course Particulate
CVD Cardiovascular Disease
EC Elemental Carbon
FP Fine Particulate
HS Hemorrhagic Stroke
ICD9 International Classification of Disease,
9th Revision
IHD Ischemic Heart Disease
IS ischemic stroke
MI Myocardial Infarction
OC Organic Carbon
OHC Oxygenated
Hydrocarbons
PERI Peripheral Vascular and
Cerebrovascular Disease
PM Particulate Matter
PIH primary intracerebral hemorrhage
PNC Particle Number Concentration
SHS Subarachnoid hemorrhagic stroke
TP Total Particulate
UBRE Unbiased Risk Estimator
O
HH
H
W
-------
O
to
O
O
oo
TABLE AX6.3-8. HUMAN HEALTH EFFECTS OF OXIDES OF NITROGEN: CVD HOSPITAL ADMISSIONS
AND VISITS: EUROPE
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels &
Monitoring Stations
Copollutants Effects: Relative Risk or Percent Change &
(Correlations) Confidence Intervals ([95% Lower, Upper])
X
H
6
o
o
H
O
O
H
W
Ballester et al. (2006)
Multi-city, Spain:
Barcelona, Bilbao,
Castellon, Gijon,
Huelva, Madrid,
Granada, Oviedo,
Seville, Valencia,
Zaragoza
Period of Study:
1995/1996-1999,
N= 1,096 day
Outcome(s) (ICD9): All CVD 390-459;
Heart diseases 410-414,427,428.
Emergency admission from hospital
records. Discharge data used.
Study Design: Time-series, meta-
analysis to pool cities
N: daily mean admissions reported by
city
Statistical Analyses: Poisson regression
and GAM, with stringent convergence
criteria, meta-analysis with fixed effect
model. Tested linearity by modeling
pollutant in linear and non-linear way
(spline smoothing). Linear model
provided best results 55% of time but
used in all cases to facilitate
comparability.
Covariates: temperature, humidity and
influenza, day of wk unusual events,
seasonal variation and trend of the series
Seasons: Hot: May to Oct;
Cold: NovtoApr
Statistical Package: SPLUS
Lag: 0-3
NO2 24-h avg (ug/m ):
Mean: 51.5
10th percentile: 29.5
90th percentile: 74.4
# of Stations: Depends on
the city
Correlation among stations:
NR
CO 8-h max (0.58)
O38-h max .(0.03)
S02 24 h (0.46)
BS24h(0.48)
TSP 24 h (0.48)
PM10 24 h (0.40)
Two-pollutant
models used to
adjust for
copollutants.
Results reported for % change in hospital
admissions associated with 10 ug/m2 increase in
N02.
All CVD
0.38% (0.07%, 0.69%), lag 0-1
Heart Diseases:
0.86% (0.44%, 1.28%)
Effect of NO2 was diminished in two-pollutant
models.
O
HH
H
W
-------
O
to
O
O
oo
TABLE AX6.3-8 (cont'd). HUMAN HEALTH EFFECTS OF OXIDES OF NITROGEN: CVD HOSPITAL ADMISSIONS
AND VISITS: EUROPE
X
H
6
o
o
H
O
O
H
W
Reference, Study
Location, & Period
Lanki et al. (2006)
Europe (Augsburg,
Helsinki, Rome,
Stockholm)
Study period:
1992-2000
Outcomes, Design, & Methods
Outcome(s) (ICD9): AMI 410.
Ascertained from discharge records or
AMI registry data depending on the city.
Study Design: Time-series
Statistical Analysis: Poisson regression,
for non-linear confounders - penalized
splines in GAM chosen to minimize
UBRE score. Random-effects model for
pooled estimates.
N: 26,854 hospitalizations
Statistical Software: R package
Covariates: Barometric pressure,
temperature, humidity.
Lag(s): 0-3 day
Mean Levels & Copollutants
Monitoring Stations (Correlations)
NO2 (|ig/m3)
Augsburg
25th: 40.2
50th: 49.2
75th: 58.9
98th: 88.7
Barcelona
25th: 34.8
50th: 45.0
75th: 60.0
98th: 86.0
Helsinki
25th: 21.8
50th: 28.7
75th: 37.6
98th: 64.7
Rome
25th: 61.9
50th: 70.6
75th: 80.4
98th: 102.5
Stockholm
25th: 16.3
50th: 22.2
75th: 28.6
98th: 45.9
PM10 (0.29, 0.64)
C0(0.43,0.75)
03 (0.17, 0.38)
Range in
correlations
depends on the city.
Two-pollutant
models for PNC
withO3 andPM10
only.
Effects: Relative Risk or Percent Change &
Confidence Intervals ([95% Lower, Upper])
Results reported as RR associated with an
incremental increase in NO2 equivalent to the
IQR (8 ng/m2).
Pooled results for 5 Cities:
First MI:
0.996(0.988, 1.015), lag 0
0.998(0.986, 1.010), lag 1
1.003(0.994, 1.011), lag 2
1.001(0.989, 1.014), lag 3
No significant results observed for analyses
stratified by age or season for lag 0/1 .
O
HH
H
W
-------
O
to
O
O
oo
TABLE AX6.3-8 (cont'd). HUMAN HEALTH EFFECTS OF OXIDES OF NITROGEN: CVD HOSPITAL ADMISSIONS
AND VISITS: EUROPE
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels &
Monitoring Stations
Copollutants Effects: Relative Risk or Percent Change &
(Correlations) Confidence Intervals ([95% Lower, Upper])
X
to
H
6
o
o
H
O
O
H
W
Von Klot et al. (2005)
Europe (Augsburg,
Barcelona, Helsinki,
Rome, Stockholm)
Study Period:
1992-2000
Outcome(s) (ICD9): Re-admission for
AMI 410; angina pectoris 411 and 413;
Cardiac diseases including AMI angina
pectoris, dysrhythmia (427), heart failure
(428). Hospital admissions database used
to identify cases.
Population: Incident cases of MI during
1992-2000 among those >35 yrs old.
N Augsburg: 1560
N Barcelona: 1134
N Helsinki: 4026
NRome: 7384
N Stockholm: 7902
Study Design: Prospective Cohort
Statistical Analyses: Poisson regression,
at risk period from the 29th day after the
index event until the event of interest,
death, migration, or loss to follow-up.
GLM models, penalized spline functions
for continuous confounders. City results
pooled using random-effects model.
Heterogeneity assessed. Sensitivity
analyses conducted varying the smooth
functions, convergence criteria, and how
confounders were specified.
Statistical Software: R package
Covariates: Daily mean temperature,
dew point temperature, barometric
pressure, relative humidity, vacations or
holidays.
Lag: 0-3 days
NO2 24-h avg (ug/m ):
Augsburg
Mean: 49.6
5th: 30
25th: 39.7
75th: 57.2
95th: 75.3
Barcelona
Mean: 47.7
5th: 18
25th: 34.0
75th: 60
95th: 83
Helsinki
Mean: 30.1
5th: 13
25th: 21.2
75th: 36.7
95th: 52.9
Rome
Mean: 15.8
5th: 5.4
25th: 10.1
75th: 21.7
95th: 25.9
Stockholm
Mean: 22.8
5th: 10.3
25th: 16
75th: 28
95th: 39.4
C024h
(0.44, 0.75)
038h
(-0.2,-0.13)
PM10 (.29, .66)
PNC (.44, .83)
Two-pollutant
models but NO2,
CO, and PNC not
modeled together
because they were
too highly
correlated.
Results reported for RR for incremental
increases in same day NO2 equivalent to the
mean of the city specific IQR's multiplied by
0.05 (8 ug/m ). Pooled results are below:
MI
1.028(0.997, 1.060), lag 0
Angina Pectoris
1.032(1.006, 1.058), lag 0
Cardiac Diseases
1.032(1.014, 1.051), lag 0
Two-pollutant models show that the effect of
NO2 independent of PM10 and O3. Traffic
exhaust may be associated with cardiac
re-admission.
O
HH
H
W
# Stations: 1-5
-------
O
to
O
O
oo
TABLE AX6.3-8 (cont'd). HUMAN HEALTH EFFECTS OF OXIDES OF NITROGEN: CVD HOSPITAL ADMISSIONS
AND VISITS: EUROPE
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels &
Monitoring Stations
Copollutants Effects: Relative Risk or Percent Change &
(Correlations) Confidence Intervals ([95% Lower, Upper])
X
Andersen et al. (2007a)
Copenhagen, Denmark
Study Period:
1999-2004
Outcome(s) (ICD10): angina pectoris
120; acute and subsequent MI 121-22;
other acute IHD124; chronic IHD125;
pulmonary embolism 126; cardiac arrest
146; cardiac arrhythmias 148-49; hear
failure 150. Hospital admissions from
Danish Hospital Register.
# Hospitals: 9 (within 15 km of
monitoring station)
Study Population: 65 +
Catchment area: 1.5 million
Study Design: Time-series
Statistical Analysis: Principal
components analysis for source
apportionment. Poisson GAM,
smoothing splines for weather, long-term
trends/seasonality, indicator variables for
influenza, holidays
Software: mgcv package, R
Lag(s): 0-5 d, 0-3davg
24-h avg NO2 (ppb)
Mean(SD): 12(5)
25th: 8
75th: 15
IQR: 7
PM10 (0.45)
CO (0.74)
Source Specific
PM10
Biomass(0.41)
Secondary (0.43)
Oil (0.42)
Crustal (0.24)
Sea Salt (-0.19)
Vehicle (0.65)
Results reported for RR associated with an
incremental increase in NO2 equivalent to one
IQR (7 ppb).
Single pollutant
1.013 (0.993, 1.033), lag 0-3 avg
2-pollutant, NO2 with PM10
1.000(0.975,1.026)
H
6
o
o
H
O
O
H
W
O
HH
H
W
-------
O
to
O
O
oo
TABLE AX6.3-8 (cont'd). HUMAN HEALTH EFFECTS OF OXIDES OF NITROGEN: CVD HOSPITAL ADMISSIONS
AND VISITS: EUROPE
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels & Monitoring
Stations
Copollutants Effects: Relative Risk or Percent Change &
(Correlations) Confidence Intervals ([95% Lower, Upper])
X
H
6
o
o
H
O
O
H
W
Andersen et al. (2007b)
Copenhagen, Denmark
Study Period:
2001-2004
Outcome(s) (ICD10): Angina pectoris
120; acute and subsequent MI 121-22;
other acute IHD124; chronic IHD125;
pulmonary embolism 126; cardiac arrest
146; cardiac arrhythmias 148-49; heart
failure 150. Hospital admissions from
Danish Hospital Register.
# Hospitals: 9 (within 15 km of
monitoring station)
Study Population: 65 +
Catchment area: 1.5 million
Study Design: Time-series
Statistical Analysis: Poisson GAM,
smoothing splines for weather, long-term
trends/seasonality, indicator variables for
influenza, holidays
Software: mgcv package, R
Lag(s): 0-5 d, 0-3 d avg
24-h avg NO2 (ppb)
Mean(SD): 11(5)
25th: 8
50th: 11
75th: 14
99th: 28
IQR: 6
24-h avg NOX (ppb)
Mean(SD): 15(8)
25th: 9
50th: 12
75th: 18
99th: 46
IQR: 9
24-h avg NOX curbside (ppb)
Mean(SD): 83(36)
25th: 58
50th: 76
75th: 103
99th: 207
IQR: 45
NCtotw/NO2(0.68)
NCtotw/NOx(0.66)
NCa57w/N02(0.57)
Results reported for associations of a 6-ppb
increase equivalent to one IQR of NO2 with all
CVD.
One-pollutant model:
1.0(0.98, 1.03), lag 0-3
2-pollutant model with NCtot
1.0(0.96,1.03)
O
HH
H
W
-------
O
to
O
O
oo
TABLE AX6.3-8 (cont'd). HUMAN HEALTH EFFECTS OF OXIDES OF NITROGEN: CVD HOSPITAL ADMISSIONS
AND VISITS: EUROPE
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels &
Monitoring Stations
Copollutants Effects: Relative Risk or Percent Change &
(Correlations) Confidence Intervals ([95% Lower, Upper])
X
H
6
o
o
H
O
O
H
W
Atkinson et al. (1999b)
London, England
Period of Study:
1992-1994,
N= 1,096 day
Outcome(s) (ICD9): All CVD 390-459;
IHD 410-414. Emergency admissions
obtained from the Hospital Episode
Statistics (HES) database.
Ages groups analyzed: 0-14 yr, 15-64 yr,
0-64 yr, 65+ yr, 65 74 yr, 75+ yr
Study Design: Time-series, hospital
admission counts
N: 189, 109 CVD admissions
Catchment area: 7 million residing in
1600 Km2 area of Thames basin.
Statistical Analyses: APHEA protocol,
Poisson regression
Covariates: Adjusted long-term seasonal
patterns, day of wk, influenza,
temperature, humidity (compared
alternative methods for modeling
meteorological including linear,
quadratic, piece-wise, spline)
Seasons: Warm season Apr-Sept, cool
season remaining mos, interactions
between season investigated
Dose response investigated: Yes, bubble
charts presented
Statistical Package: SAS
Lag: 0-3
l-hmax(ppb)
Mean: 50.3
SD = 17.0
Min: 22.0
Max: 224.3
10th-90thpercentile: 36
# of Stations: 3, results
averaged across stations
Correlation among stations:
0.7-0.96
PM10 24 h
C024h
S02 24 h
O38h
BS24h
Correlations of NO2
with CO, SO2, O3,
BS ranged from
0.6-0.7
Correlation of NO2
with O3 negative
Two-pollutant
models used adjust
for copollutants
Results reported for % change in hospital
admissions associated with 10th-90th percentile
increase in NO2 (36 ppb)
All CVD
Ages 0-64: 1.20% (-0.62%, 3.05%), lag 0
Ages 65+: 1.68% (0.32%, 3.06%), lag 0
IHD
Ages 0-64: 1.53% (-1.22%, 4.37%), lag 0
Ages 65+: 3.03% (0.87%, 5.24), lag 0
NO2 was associated with increased CVD
admissions for all ages but this association was
stronger among those 65+ yrs old. Similar
increase associated with IHD among those 65+
yrs old.
Monitors close to roadways were not used in the
study. Correlations for NO2 between urban
monitoring sites were high. Authors suggest
that the pollution levels are uniform across the
study area. Authors did not investigate the
interaction between meteorological variables
and air pollution. In two-pollutant models, O3
had little impact on NO2. BS moderated the
association of NO2 with CVD among the 65+
age group. Suggestion that NO2 associations
were non-linear.
O
HH
H
W
-------
O
to
O
O
oo
TABLE AX6.3-8 (cont'd). HUMAN HEALTH EFFECTS OF OXIDES OF NITROGEN: CVD HOSPITAL ADMISSIONS
AND VISITS: EUROPE
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels &
Monitoring Stations
Copollutants Effects: Relative Risk or Percent Change &
(Correlations) Confidence Intervals ([95% Lower, Upper])
X
Oi
Oi
Ballester et al. (2001)
Valencia, Spain
Period of Study:
1992-1996
Outcome(s) (ICD9): All CVD 390-459;
heart diseases 390-459; cerebrovascular
diseases 430-438. Admissions from city
registry - discharge codes used.
Study Design: Time-series
N: 1080 CVD admissions
# of Hospitals: 2
Catchment area: 376,681 inhabitants of
Urban Valencia
Statistical Analyses: Poisson regression,
GAM with parametric smoothers,
APHEA/ Spanish EMECAM protocol.
Both Linear and non parametric model,
including a loess term was fitted,
departure from linearity assess by
comparing deviance of both models.
Covariates: Long-term trend and
seasonality, temperature and humidity,
wk days, flu, special events, air pollution.
Seasons: Hot season May to Oct;
Cold season Nov to Apr
Statistical Package: SAS
Lag: 0-4
1-hmax (ug/m2)
Mean: 116.1
SD = NR
Min: 21.1
Max: 469.0
Median: 113.2
# of Stations:
5 automatic
14 manual,
CO 24 h (0.03)
S02 24 h (0.33)
O38h(-0.26)
BS(0.33)
Two-pollutant
models used to
adjust for
copollutants.
Correlation among stations:
0.3-0.62 for BS, 0.46-0.78
for gaseous pollutants
Results reported for RR corresponding to a
10 ug/m2 increase inNO2
All CVD
1.0302(1.0042,1.0568), lag 0
Heart Disease
1.0085(0.9984,1.0188), lag 2
Cerebrovascular Disease
1.0362(1.0066,1.0667), lag 4
Clear association of NO2 with cerebrovascular
disease observed. Association persisted after
Inclusion of BS and SO2 in two-pollutant
models with NO2.
Cases of digestive disorders served as a control
group - null association with NO2 observed.
H
6
o
o
H
O
O
H
W
O
HH
H
W
-------
O
to
O
O
oo
TABLE AX6.3-8 (cont'd). HUMAN HEALTH EFFECTS OF OXIDES OF NITROGEN: CVD HOSPITAL ADMISSIONS
AND VISITS: EUROPE
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels &
Monitoring Stations
Copollutants Effects: Relative Risk or Percent Change &
(Correlations) Confidence Intervals ([95% Lower, Upper])
X
D'Ippolitietal. (2003)
Rome, Italy
Study Period: Jan 1995-
Junl997
Outcome(s) (ICD): AMI 410 (first
episode). Computerized hospital
admission data.
Study Design: Case-crossover, time-
stratified, control days within same mo
falling on the same day.
Statistical Analyses: Conditional logistic
regression, examined homogeneity across
co-morbidity categories
N: 6531 cases
Age groups analyzed: 18-64yrs,
65-74 yrs, > 75
Season: Cool: Oct-Mar;
Warm: Apr-Sept
Lag(s): 0-4 day, 0-2 day cum avg
Dose Response: OR for increasing
quartiles presented and p- value for trend.
NO2 24
Mean(SD):
25th: 74.9
50th: 86.0
75th: 96.9
IQR: 22
# Stations: 5
86.4(15.8)
TSP 24 h (0.37)
S0224h(0.31)
CO 24 h (0.03)
No multipollutant
models
Results presented for OR associated with
incremental increase in NO2 equivalent to one
IQR.
AMI
1.026(1.002, 1.052), lag 0
1.026(0.997, 1.057), lag 0-2
Association observed for NO2 but TSP
association more consistent. Authors think that
TSP, CO, and NO2 cannot be distinguished
from traffic-related pollution in general.
H
6
o
o
H
O
O
H
W
Llorca et al. (2005)
Torrelavega, Spain
Study period:
1992-1995
Outcome(s) (ICD): CVD (called cardiac
in paper) 390-459. Emergency
admissions, excluding non residents.
Obtained admissions records from
hospital admin office.
Study design: Time-series
Statistical analyses: Poisson regression,
APHEA protocol
Covariates: rainfall, temperature, wind
speed direction
N: 1 8, 1 37 admissions
Statistical software: STATA
Lag(s): not reported
NO2 24 h ng/m3
Mean(SD): 21.3(16.5)
TSP (-0.12)
SO2 (0.588)
SH2 (0.545)
NO (0.855)
Multipollutant
models
Results reported for RR of hospital admissions
for 100 ng/m3 increase inNO2.
CVD admissions:
1.27 (1.14, 1.42), 1-pollutant model
1.10 (0.92, 1.32), 5-pollutant model
Effect of NO2 diminished in multipollutant
model.
O
HH
H
W
-------
O
to
O
O
oo
TABLE AX6.3-8 (cont'd). HUMAN HEALTH EFFECTS OF OXIDES OF NITROGEN: CVD HOSPITAL ADMISSIONS
AND VISITS: EUROPE
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels &
Monitoring Stations
Copollutants Effects: Relative Risk or Percent Change &
(Correlations) Confidence Intervals ([95% Lower, Upper])
X
oo
H
6
o
o
H
O
O
H
W
Pantazopoulou et al.
(1995)
Athens, Greece
Study Period: 1988
(Winter and Summer)
Poloniecki et al. (1997)
London, UK
Study Period:
Apr 1987-Mar 1994, 7
yrs
Outcome(s): Cardiac Disease ICD codes
not provided. Cases ascertained from
National Center for Emergency Service
database. Cases diagnosed at time of
admission so they are ED visits and were
not necessarily admitted to the hospital.
Study design: Time-series
Statistical Analyses: Linear regression
(not well described)
Covariates: Dummy variables for winter
mos with Jan as referent. Dummy
variables for summer mos with Apr as
referent. Day of the wk, holidays,
temperature, relative humidity,
N: 25,027 cardiac admissions.
Lag(s): NR
Outcome(s): All CVD 390-459; MI 410;
Angina pectoris 413; other IHD 414;
ARR 427; congestive heart failure 428;
cerebrovascular disease 430-438.
Hospital Episode Statistics (HES) data on
emergency hospital admissions.
Study Design: Time-series
N: 373, 556 CVD admissions
Statistical Analyses: Poisson regression,
linear and quadratic terms to adjust for
long-term trends. APHEA protocol
Covariates: long-term trends, seasonal
variation, day of wk, influenza,
temperature and humidity.
Season: Warm, Apr-Sept;
Cool, Oct-Mar.
Lag: 0-1 day
NO2 l-hmax(ng/m3):
Winter
Mean(SD): 94(25)
5th: 59
50th: 93
95th: 135
Summer
Mean(SD): 111(32)
5th: 65
50th: 108
95th: 173
# Stations: 2
NO2 24 h ppb:
Min: 8
10%: 23
Median: 35
90%: 53
Max: 198
CO,BS
No correlations
provided
Black Smoke
C024h
S02 24 h
O38h
Correlations
between pollutants
high but not
specified.
Results reported for regression coefficients
based on an incremental increase in NO2 of
76 ng/m3 in winter and 108 ng/m3 in summer
(5th to 95th percentile).
Winter (regression coefficient)
11.2(3.3,19.2)
Summer (regression coefficient)
-0.06 (-6.6, 6.5)
Results expressed as a relative rate (RR) for an
incremental increase of NO2 equivalent to
30 ppb (10th-90th percentile).
AMI: 1.0274(1.0084,1.0479)
Angina Pectoris: 1.0212 (0.9950,1.0457)
Other IHD: 0.99(0.0067,1.0289)
Cardiac ARR: 1.0274 (1.0006, 1.0984)
Heart Failure: 0.9970(0.9769, 1.0194)
Cerebro vascular Disease: 0.9851
(0.9684,1.0045)
Other Circulatory: 1.0182 (1.0000,1.0398)
All CVD: 1.0243 (1.0054,1.0448)
No attenuation of NO2 association with MI in
two-pollutant model (cool season).
O
HH
H
W
-------
O
to
O
O
oo
TABLE AX6.3-8 (cont'd). HUMAN HEALTH EFFECTS OF OXIDES OF NITROGEN: CVD HOSPITAL ADMISSIONS
AND VISITS: EUROPE
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels &
Monitoring Stations
Copollutants Effects: Relative Risk or Percent Change &
(Correlations) Confidence Intervals ([95% Lower, Upper])
X
vo
Ponka and Virtanen
(1996)
Helsinki, Finland
Study Period:
1987-1989, 3 yrs
Outcome(s) (ICD9): IHD 410-414; MI
410; TIA 411; Cerebrovascular diseases
430-438; Cerebral ischemia due to
occlusion of extracerebral vessels 433;
Cerebral ischemia due to occlusion of
cerebral vessels 434; Transient ischemic
cerebral attack 435. Case ascertainment
was for both emergency admission and
hospital admissions - done via registry
system.
Study Design: Time-series
Statistical Analyses: Poisson regression,
pollutant concentrations log transformed
N: 12,664 all IHD admissions; 7005 IHD
ED admissions; 7232 cerebrovascular
hospital admissions; 3737
cerebrovascular ED admissions.
Covariates: Weather, day of wk, long-
term trends, influenza
Lag(s): 1-7 days
NO2 8 h (ug/m3)
Mean(SD): 39(16.2)
Range: 4, 170
NO 8 h ug/m3
Mean(SD): 91(61)
Range: 7,467
# Stations: 2
SO28h
N08h
TSPSh
O38h
NO2 highly
correlated with SO2
and TSP.
Results reported are regression coefficients and
standard errors (SE).
NO2 with ED admissions for transient short
term ischemic attack
-0.056 (0.105), p = 0.59, lag 1
NO2 with ED admissions for cerebrovascular
disease
-0.025 (0.057), p = 0.657, lag 1
NO with IHD, all admissions
0.097 (0.023), p< 0.001, lag 1
NO with IHD, ED admissions
0.111 (0.030), p< 0.001, lag 1
Significant increase in admissions for transient
short-term ischemic attack and cerebrovascular
diseases for lag 6 associated with NO2 exposure.
H
6
o
o
H
O
O
H
W
Prescottetal. (1998)*
Edinburgh, UK
Study period:
Oct 1992-June 1995
Outcome(s) (ICD9): Cardiac and
cerebral ischemia 410-414,426-429, 434-
440. Extracted from Scottish record
linkage system.
Study Design: Time-series
Statistical Analysis: Poisson, log linear
regression models
Age groups analyzed: <65, 65+yrs
Covariates: Seasonal and wkday
variation, temperature, and wind speed.
Lag(s): 0, 1, 3 day moving avg
NO2 24 h (ppb)
Mean(SD): 26.4(7.0)
Range: 9, 58
IQR: 10 ppb
03, 24 h
PM, 24 h
SO2, 24 h
CO, 24 h
Correlations not
reported.
Results reported for percent change in
admissions based on an incremental increase in
NO2 equivalent to the IQR of 10 ppb.
<65 yrs, CVD admissions
-0.05 (-5.2, 4.5), 3 day moving avg
65+ yrs, CVD admissions
-0.9 (-8.2, 7.0), 3 day moving avg
Data for lag 1 not presented.
O
HH
H
W
-------
O
to
O
O
oo
TABLE AX6.3-8 (cont'd). HUMAN HEALTH EFFECTS OF OXIDES OF NITROGEN: CVD HOSPITAL ADMISSIONS
AND VISITS: EUROPE
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels &
Monitoring Stations
Copollutants Effects: Relative Risk or Percent Change &
(Correlations) Confidence Intervals ([95% Lower, Upper])
X
Yallop et al. (2007)
London, England
Study Period:
Jan. 1988-Oct. 2001,
>1400 days
Outcome(s): Acute pain in Sickle Cell
Disease (HbSS, HbSC, HbS/pO,
thalassaemia, HbS/p+). Admitted to
hospital for at least one night.
Study Design: Time-series
Statistical Analyses: Cross-correlation
function
N: 1047 admissions
Covariates: No adjustment made in
analysis, discussion includes statement
that the effects of weather variables and
copollutants are inter-related.
Statistical Package: SPSS
Lag(s): 0-2 days
Dose response: Quartile analysis, graphs
presented, ANOVA comparing means
across quartiles.
NR
O3, CO, NO, NO2,
PM10:
daily avg used for
all copollutants
High O3 levels
correlate with low
NO, low CO,
increased wind
speeds and low
humidity and each
was associated with
admission for pain.
Not possible to
distinguish
associations in
analysis.
Results reported are cross-correlation
coefficients. NO inversely correlated with
admission for acute pain in SCO.
CFF: -0.063, lag 0
H
6
o
o
H
O
O
H
W
*Default GAM
AMI Acute Myocardial
Infarction
ARR Arrhythmia
BC Black Carbon
COH coefficient of haze
CP Course Particulate
CVD Cardiovascular Disease
EC Elemental Carbon
FP Fine Particulate
HS Hemorrhagic Stroke
ICD9 International Classification of Disease,
9th Revision
IHD Ischemic Heart Disease
IS ischemic stroke
MI Myocardial Infarction
OC Organic Carbon
OHC Oxygenated
Hydrocarbons
PERI Peripheral Vascular and
Cerebrovascular Disease
PM Particulate Matter
PIH primary intracerebral hemorrhage
PNC Particle Number Concentration
SHS Subarachnoid hemorrhagic stroke
TP Total Particulate
UBRE Unbiased Risk Estimator
O
HH
H
W
-------
O
to
O
O
oo
TABLE AX6.3-9. HUMAN HEALTH EFFECTS OF OXIDES OF NITROGEN: CVD HOSPITAL ADMISSIONS
AND VISITS: ASIA
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels &
Monitoring Stations
Copollutants Effects: Relative Risk or Percent Change &
(Correlations) Confidence Intervals ([95% Lower, Upper])
X
I
£
H
6
o
o
H
O
O
H
W
Chan et al. (2006)
Taipai, Taiwan
Period of Study:
Apr 1997-Dec 2002,
2090 days
Outcome(s) (ICD9): Cerebrovascular
disease 430-437; stroke 430-434;
hemorrhagic stroke 430-432; ischemic
stroke 433-434. Emergency admission
data collected from National Taiwan
University Hospital.
Ages groups analyzed: Age >50 included
in study
Study Design: Time-series
N: 7341 Cerebrovascular admissions
among those >50 yrs old
# of Hospitals:
Catchment area:
Statistical Analyses: Poisson regression,
GAMs used to adjust for non-linear
relation between confounders and ER
admissions.
Covariates: Time-trend variables: yr,
mo, and day of wk, daily temperature
difference, and dew point temperature.
Linearity: Investigated graphically by
using the LOESS smoother.
Statistical Package: NR
Lag: 0-3, cumulative lag up to 3 days
NO224-havg (ppb):
Mean: 29.9
SD = 8.4
Min: 8.3
Max: 77.1
IQR: 9.6 ppb
# of Stations: 16
Correlation among
stations: NR
PM1024h;r=0.50
PM2 5 24 h; r = 0.64
CO 8-h avg; r = 0.77
SO2 24 h; r = 0.64
O3 1-h max; r= 0.43
Two-pollutant
models to adjust for
copollutants.
Results reported for OR for association of
emergency department admissions with an IQR
increase in NO2 (9.3 ppb)
Cerebrovascular:
1.032(0.991,1.074), lag 0
Stroke:
0.994(0.914, 1.074), lag 0
Ischemic stroke:
1.025(0.956, 1.094), lag 0
Hemorrhagic stroke:
0.963(0.884, 1.042), lag 0
No significant associations for NO2 reported.
Lag 0 shown but similar null results were
obtained for lags 1-3. NO2 highly correlated
with PM and CO.
O
HH
H
W
-------
O
to
O
O
oo
TABLE AX6.3-9 (cont'd). HUMAN HEALTH EFFECTS OF OXIDES OF NITROGEN: CVD HOSPITAL ADMISSIONS
AND VISITS: ASIA
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels &
Monitoring Stations
Copollutants Effects: Relative Risk or Percent Change &
(Correlations) Confidence Intervals ([95% Lower, Upper])
X
to
H
6
o
o
H
O
O
H
W
Chang et al. (2005)
Taipei, Taiwan
Study Period:
1997-2001, 5 yrs
Hosseinpoor et al.
(2005)
Tehran, Iran
Study period:
Marl996-Mar2001,5
yrs
Outcome(s) (ICD9): CVD 410-429.
Daily clinic visits or hospital admission
from computerized records of National
Health Insurance. Discharge data.
Source Population: 2.64 Million
N: 40.8 admissions/day, 74,509/5 yrs
# Hospitals: 41
Study Design: Case-crossover, referent
day 1 wk before or after index day
Statistical Analyses: Conditional logistic
regression.
Covariates: Same day temperature and
humidity.
Season: Warm/cool (stratified by
temperature cutpoint of 20 °C)
Lag(s): 0-2 days
Outcome(s) (ICD9): Angina pectoris
413. Primary discharge diagnosis from
registry databases or records.
Study Design: Time-series
Statistical Methods: Poisson regression
# Hospitals: 25
Covariates: Long-term trends,
seasonality, temperature, humidity,
holiday, post-holiday, day of wk.
Lag(s): 0-3
NO2 24-h avg (ppb):
Mean: 31.54
Min: 8.13
25th: 26.27
50th: 31.03
75th: 36.22
Max: 77.97
# of Stations: 6
NO2 24-h avg (ug/m3)
Mean(SD): 60.01(39.69)
Min: 0.30
25th: 29.39
Median: 47.42
75th: 84.55
Max: 324.78
CO 24-h avg
O3 24-h avg
SO2 24-h avg
PM10 24-h avg
Correlations not
reported.
Two-pollutant
models to adjust for
copollutants
N02 CO 03 PMi,
Correlations not
reported
OR for the association of CVD admissions with
an incremental increase in 24-h avg NO2
equivalent to one IQR, 9.95 ppb.
Warm(>20°C)
1.177(1.150, 1.205), lag 0-2
Cool (<20 °C)
1.112(1.058, 1.168), lag 0-2
NO2 effect remained in all warm season two-
pollutant models. Effect remained in cool
season two-pollutant models with the exception
of the model that included PM10.
Results reported for relative risk in hospital
admissions per increment of 10 ug/m3 SO2.
Angina
1.00618 (1.00261, 1.00976), lag 1
In a multipollutant model only CO (lag 1) was
significantly associated with angina pectoris
related hospital admissions.
O
HH
H
W
-------
O
to
O
O
oo
TABLE AX6.3-9 (cont'd). HUMAN HEALTH EFFECTS OF OXIDES OF NITROGEN: CVD HOSPITAL ADMISSIONS
AND VISITS: ASIA
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels &
Monitoring Stations
Copollutants Effects: Relative Risk or Percent Change &
(Correlations) Confidence Intervals ([95% Lower, Upper])
X
H
6
o
o
H
O
O
H
W
O
HH
H
W
Lee et al. (2003a)*
Seoul, Korea
Study period:
Dec 1997-Dec 1999,
822 days, 184 day sin
summer
Tsai et al. (2003a)
Kaohsiung, Taiwan
Study period:
1997-2000
Outcome(s) (ICD10): IHD: Angina
pectoris 120; Acute or subsequent MI 121-
123; other acute IHD 124. Electronic
medical insurance data used.
Study Design: Time-series
Statistical Methods: Poisson regression,
GAM with default convergence criteria.
Age groups analyzed: all ages, 64+
Covariates: long-term trends LOESS
smooth, temperature, humidity, day of wk.
Season: Presented results for summer
(Jun, Jul, Aug) and entire period.
Lag(s): 0-6
Outcome(s) (ICD9): All cerebrovascular
430-438; SHS 430; PIH 431-432; IS 433-
435; Other 436-438. Ascertained from
National Health Insurance Program
computerized admissions records.
Study Design: Case-crossover
Statistical Analysis: Conditional logistic
regression.
Statistical Software: SAS
Seasons: >20°C;<20°C.
N: 23,179 stroke admissions
# Hospitals: 63
Lag(s): 0-2, cumulative lag up to 2
previous days
NO2 24 h (ppb):
5th: 16
10th: 23.7
Median: 30.7
75th: 38.3
95th: 48.6
Mean(SD): 31.5(10.3)
IQR: 14.6
24-h avg NO2 (ppb)
Min: 6.25
25th: 19.25
Median: 28.67
75th: 36.33
Max: 63.40
Mean: 28.17
PM10;r= 0.73, 0.74
S02;r = 0.72, 0.79
O3;r= -0.07, 0.63
CO; r = 0.67, 0.79
Range depends on
summer vs. entire
period.
Two-pollutant
models
PM10
S02
CO
03
Results reported for RR of IHD hospital
admission for an incremental increase in NO2
equivalent to one IQR.
64+, entire study period:
1.08(1.03, 1.14), lag 5
64+, summer only:
1.25 (1.11, 1.41), lag 5
Results for lag 5 presented above. Lag 0 or 1
results largely null - presented graphically.
Confounding by PM10 was not observed in these
data using two-pollutant models.
Results reported as OR for the association of
admissions with an incremental increase of NO2
equivalent to the IQR of 17.1 ppb
PIH admissions
Warm: 1.56 (1.32, 1.84), lag 0-2
Cool: 0.81(0.0, 1.31), lag 0-2
IS admissions:
Warm: 1.55 (1.40, 1.71), lag 0-2
Cool: 1.16(0.81,1.68), lag 0-2
Effects persisted after adjustment for PM10,
SO2, CO, and O3.
PIH:
1.31(1.03, 1.66) N02w/PM10
1.66 (1.38, 2.00), NO2 w/ SO2
1.60(1.25,2.05)NO2w/CO
1.51(1.26, 1.80)N02w/03
IS:
1.39(1.20, 1.60) N02w/PM10
1.62(1.45, 1.81),N02w/S02
1.54(1.33, 1.79),NO2w/CO
1.53(1.37, 1.71),N02w/03
-------
O
to
O
O
oo
TABLE AX6.3-9 (cont'd). HUMAN HEALTH EFFECTS OF OXIDES OF NITROGEN: CVD HOSPITAL ADMISSIONS
AND VISITS: ASIA
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels &
Monitoring Stations
Copollutants Effects: Relative Risk or Percent Change &
(Correlations) Confidence Intervals ([95% Lower, Upper])
X
H
6
o
o
H
O
O
H
W
O
HH
H
W
Wong etal. (1999)
Hong Kong, China
Study Period:
1994-1995
Wong et al. (2002)*
Hong Kong
London
Study Period:
1995-1997
Outcome(s) (ICD9): CVD: 410-417,
420-438,440-444; CHF 428; IHD 410-
414; Cerebrovascular Disease 430-438.
Hospital admissions through ER
departments via Hospital Authority
(discharge data).
Study Design: Time-series
Statistical Analyses: Poisson regression,
linear and quadradic terms for long-term
trends, APHEA protocol
# Hospitals: 12
Covariates: Daily temperature, relative
humidity day of wk, holidays, influenza,
long-term trends (yr and seasonality
variables). Interaction of pollutants with
cold season examined.
Season: Cold (Dec-Mar)
Lag(s): 0-3 days
Outcome(s) (ICD9): Cardiac disease
396-429; IHD 410-414. Admissions
through the emergency department,
general outpatient, or direct to inpatient
wards.
Study design:
Statistical analysis: Poisson regression,
GAMs, nonparametric smooth functions
(LOESS)
Covariates:
Statistical Software: SPlus
NO2 24-h avg (ug/m )
Minimum: 16.41
25th: 39.93
Median: 51.39
75th: 51.39
Maximum: 122.44
24-h avg NO2
Hong Kong
Mean (warm/cool): 55.9
(48.1/63.8)
Minimum: 15.3
10th: 31.8
50th: 53.5
90th: 81.8
Max: 151.5
Hong Kong
Mean (warm/cool): 64.3
(62.6.1/66.1)
Minimum: 23.7
10th: 42.3
50th: 61.2
90th: 88.8
Max: 255.8
PM10;r=0.79
S02
03
Range for other pollutants:
r = 0.68, 0.89.
Hong Kong
S02;r = 0.37
PM10;r=0.82
O3;r=0.43
London
SO2;r = 0.71
PM10;r=0.68
O3;r=-0.29
Results reported for RR associated with
incremental increase inNO2 equal to 10 ug/m .
CVD
5-64 yrs: 1.008 (0.998, 1.018), lag 0-1
65+yrs: 1.016 (1.009, 1.023), lag 0-1
All ages: 1.013(1.007, 1.020), lag 0-1
CHF
1.044(1.25, 1.063), lag 0-3
IHD
1.010(0.999, 1.020), lag 0-1
Cerebrovascular Disease
1.008(0.998, 1.018), lag 0-1
Interaction of NO2 with O3 observed
Results reported for excess risk associated with
a 10 ug/m3 change in mean concentration
Single-pollutant model.
Hong Kong: 1.8 (1.2, 2.4), lag 0-1
London: -0.1 (-0.6, 0.5), lag 0-1
Multipollutant results
Hong Kong:
1.6 (1.0, 1.3), lag 0-1, adjusted for Ozone
1.7 (0.8, 2.7), lag 0-1, adjusted for PM10
1.6 (0.8, 2.4), lag 0-1, adjusted for SO2
London:
0.1 (-0.5, 0.6), lag 0-1, adjusted for Ozone
-0.4 (-1.2, 0.4), lag 0-1, adjusted for PM10
-0.2 (-0.9, 0.5), lag 0-1, adjusted for SO2
-------
O
to
O
O
oo
TABLE AX6.3-9 (cont'd). HUMAN HEALTH EFFECTS OF OXIDES OF NITROGEN: CVD HOSPITAL ADMISSIONS
AND VISITS: ASIA
Reference, Study
Location, & Period
Outcomes, Design, & Methods
Mean Levels &
Monitoring Stations
Copollutants Effects: Relative Risk or Percent Change &
(Correlations) Confidence Intervals ([95% Lower, Upper])
X
Yang et al. (2004b)
Kaohsiung, Taiwan
Period of Study:
1997-2000
Outcome(s) (ICD9): All CVD: 410-429
"(All CVD typically defined to include
ICD9 codes 390-459)
N: 29,661
Study Design: Case-crossover
Statistical Analysis: Poisson time-series
regression models, APHEA protocol
# of Hospitals: 63
Seasons: Authors indicate not considered
because the Taiwanese climate is tropical
with no apparent seasonal cycle
Covariates: Stratified by warm (>25°)
and cold (<25°) days, temperature and
humidity measurements included in the
model
Statistical Package: SAS
Lag: 0-2 days
24-h avg (ppb)
Min: 6.25
25%: 19.25
50%: 28.67
75%: 36.33
Max: 63.40
Mean: 28.17
# of Stations: 6
Correlation among stations:
NR
PM10
CO
S02
O38
Two-pollutant
models used to
adjust for
copollutants
Correlations NR
OR's for the association of one IQR (17.08 ppb)
increase in NO2 with daily counts of CVD
hospital admissions are reported.
All CVD (ICD9: 410-429), one-pollutant model
>25°: 1.380(1.246,1.508)
<25°: 2.215(2.014,2.437)
All CVD (ICD9: 410-429), two-pollutant
models
Adjusted for PM10:
>25°: 1.380(1.246,1.508)
<25°: 2.215(2.014,2.437)
Adjusted for SO2:
>25°: 1.149(1.017,1.299)
<25°: 2.362(2.081,2.682)
Adjusted for CO:
>25°: 1.039(0.919,1.176)
<25°: 2.472(2.138,2.858)
Adjusted for O3:
>25°: 1.159(1.051,1.277)
<25°: 2.243(2.037,2.471)
H
6
o
o
H
O
O
H
W
Association of CVD admissions with NO2
attenuated on warm days after adjustment for
copollutants. Association persisted on cool
days. Kaohsiung is the center of Taiwan's
heavy industry.
O
HH
H
W
-------
o
to
o
o
oo
X
ON
ON
TABLE AX6.3-9 (cont'd). HUMAN HEALTH EFFECTS OF OXIDES OF NITROGEN: CVD HOSPITAL ADMISSIONS
AND VISITS: ASIA
Reference, Study
Location, & Period
Yeetal. (2001)
Tokyo, Japan
Study Period:
Jul-Aug,
1980-1995
Outcomes, Design, & Methods
Outcome(s): Angina 4 13; Cardiac
insufficiency 428; Hypertension 401-405;
MI 4 10. Diagnosis made by attending
physician for hospital emergency
transports.
Age groups analyzed: 65+ yrs male and
female
Statistical analysis: GLM
Covariates: Maximum temperature,
confounding by season minimal since
only 2 summer mos included in analysis
Statistical Software: SAS
Lag(s): 1-4 days
Mean Levels &
Monitoring Stations
NO2 24-h avg (ppb)
Min: 5.3
Max: 72.2
Mean(SD): 25.4(11.4)
Copollutants
(Correlations)
O3;r= 0.183
PM10; r = 0.643
S02;r = 0.333
CO; r= 0.759
Effects: Relative Risk or Percent Change &
Confidence Intervals ([95% Lower, Upper])
Results reported for model coefficient and
95% CI.
Angina:
0.007 (0.004, 0.009)
Cardiac insufficiency:
0.006(0.003,0.01)
MI:
0.006(0.003,0.01)
* Default GAM
AMI Acute Myocardial
Infarction
ARR Arrhythmia
BC Black Carbon
COH coefficient of haze
CP Course Particulate
CVD Cardiovascular Disease
EC Elemental Carbon
FP Fine Particulate
HS Hemorrhagic Stroke
ICD9 International Classification of Disease,
9th Revision
IHD Ischemic Heart Disease
IS ischemic stroke
MI Myocardial Infarction
OC Organic Carbon
OHC Oxygenated
Hydrocarbons
PERI Peripheral Vascular and
Cerebrovascular Disease
PM Particulate Matter
PIH primary intracerebral hemorrhage
PNC Particle Number Concentration
SHS Subarachnoid hemorrhagic stroke
TP Total Particulate
UBRE Unbiased Risk Estimator
H
6
o
o
H
O
O
H
W
O
HH
H
W
-------
TABLE AX6.3-10. STUDIES EXAMINING EXPOSURE TO AMBIENT NO2 AND HEART RATE VARIABILITY AS
MEASURED BY STANDARD DEVIATION OF NORMAL-TO-NORMAL INTERVALS (SDNN)
to
o
H
6
o
z;
0
H
/O
o
H
W
O
n
HH
H
W
Author, Year,
Location
Liao et al.
(2004)
US, ARIC study
Chan et al.
(2005)
Taiwan
Wheeler et al.
(2006)
Atlanta
Luttmann-
Gibson et al.
(2006)
Steubenville
Schwartz et al.
(2005)
Boston
Avg
Study Design Time
Subjects: 4,390 adults
Analysis Method: multivariable
linear regression 24 h
Subjects: 83 adults recruited from
cardiology
Analysis Method: linear mixed
effects regression 1 h
Subjects: 30 adults (12 MI + 22
COPD)
Analysis Method: linear mixed
models 4 h
Subjects: 32 adults (>50 yrs)
Analysis Method: mixed models 24 h
Subjects:
28 elderly adults
Analysis Method: hierarchical
models 24 h
NO2 Cone (ppb)
Copollutant
Mean (sd) Range Correlation Outcome % Change (95% CI)
21(8) none lag 1 -5.0% (-9.2, -7)
PM10: 0.4
03: -0.4
SO2: 0.5
33(15) 1,110 CO: 0.7 4-hlag -4.5% (-8.1, -30)
8-hlag -6.9% (-12.0, -1.8)
MI patients
18 (no sd plO-p20, PM25: 0.4 [N = 12]
given) 7,30 CO: 0.5 4-hlag -26.0% (-42.1, -8.6)
COPD patients
[N = 22] 4-h lag 16.6% (0.2, 34.3)
PM25: 0.4
O3: -0.3
10(nosd p25-p75, SO2: 0.3
given) 6,13 lag 1 0.3% (-6.0, 6.6)
PM25: :0.3
p25-p75, 03: 0.02
medlS 14,23 CO: 0.6 lag 1 -1.6% (-7.8, 5.1)
-------
TABLE AX6.3-11. STUDIES EXAMINING EXPOSURE TO AMBIENT NO2 AND HEART RATE VARIABILITY AS
MEASURED BY VARIABLES RECORDED ON IMPLANTABLE CARDIOVERTER DEFIBRILLATORS (ICDS)
to
o
o
oo
>
X
ON
i
oo
o
[>
H
6
o
0
H
O
o
H
W
O
O
HH
H
W
Author, Year,
Location
Peters et al. (2000a)
Eastern MA
Rich et al. (2005)
Boston
Dockery etal. (2005)
Boston
Pekkanen et al. (2002)
Finland
Ruidavets et al. (2005)
France
NO2 Cone (ppb)
Subjects Analysis Method Mean (sd) Range
100 cardiac logistic regression, 23 (no sd
outpatients fixed effects given) 11,65
203 cardiac p25-max,
outpatients case-crossover med22 18,62
307 cardiac p25-p95,
outpatients logistic regression, GEE med23 19,34
45 cardiac p25-max,
patients linear regression, GAM med 16 12,36
polytomous logistic
863 adults regression 16(6) 2,48
Copollutant
Correlation Outcome
PM2.5: 0.6
Os: -0.3 RiskoflCD-recorded
SO2: 0.3 ventricular
CO: 0.7 arrhythmias
lagl
lag 0-4
all patients lag 0-1
patients with recent
arrhythmia
(<3 days) lag 0-1
PM2.5 > 0.4
O3 < - 0.4 patients with recent
SO2 > 0.4 arrhythmia
CO: 0.6 (<3 days) lag 0-1
PM2.5: 0.4
CO: 0.3 lag 2
03: -0.3
S02: 0.7 lagSh
OR (95% CI)
1.55(1.05,2.29)
1.88(1.01,3.49)
1.54(1.11,2.18)
2.09(1.26,3.51)
2.14(1.14,4.03)
14.1 (3.0,65.4)
2.7(1.2,5.4)
All results given for 20-ppb increase in NO2 with 24-h averaging time.
-------
TABLE AX6.3-12. BIRTH WEIGHT AND LONG-TERM NO2 EXPOSURE STUDIES
O
tr
to
O
O
oo
X
i
VO
O
§
H
6
o
^^
0
H
O
O
H
W
O
O
HH
H
W
Cone Range (ppb) Correlation
Author, Year, Mid- with Other
Location Study Details Low range High Pollutants Outcomes
Lin et al. (2004) Subjects: 92,288 birth cert <26.1 26.1,32.9 >32.9 Pregnancy
Taiwan Years: 1995-1997
Group: Term LEW
Analysis method: Logistic regression
Distance 3 km
Medium NO2
HighN02
<24.3 24.3,34.7 >34.7 Trimester 1
Medium NO2
HighN02
<24.0 24.0,34.4 >34.4 Trimester 2
Medium NO2
HighNO2
<23.8 23.8,34.2 >34.2 Trimester 3
Medium NO2
HighNO2
Lee et al. (2003b) Subjects: 388,105 birth cert 25 31.4 39.7 PM10: 0.66 Pregnancy
Seoul, Korea Years: 1996-1998 SO2: 0.75
Group: Term LEW model (GAM) , Interquartile CO: 0.77
Averaging time: 24h
Analysis method: Generalized additive
PM10: 0.81 Trimester 1
S02: 0.77
CO: 0.78
PM10: 0.8 Trimester 2
S02: 0.76
CO: 0.82
Trimester 3
Odds Ratio
1.06(0.93,1
1.06(0.89,1
1.10(0.96,1
1.09(0.89,1
0.87 (0.76, 1
0.93 (0.77, 1
1.01 (0.88, 1
0.86(0.71,1
1.04(1.00,1
1.02(0.99,1
1.03(1.01,1
0.98 (0.96, 1
(95% CI)
.22)
.26)
.27)
.32)
.00)
.12)
.16)
.03)
.08)
.04)
.06)
.00)
-------
TABLE AX6.3-12 (cont'd). BIRTH WEIGHT AND LONG-TERM NO2 EXPOSURE STUDIES
O
to
O
O
oo
X
H
6
o
o
H
O
O
H
W
Cone Range (ppb)
Author, Year, Mid- Correlation with
Location Study Details Low range High Other Pollutants Outcomes
Bobak M. (2000) Subjects: 69,935 birth cert SO2: 0.53 Trimester 1
Czech Year: 1991
Group: LEW adjusted for GA
Averaging time: 24 h
Analysis method: Logistic regression, 50 jig
increase 12.2 20 31.1
SO2: 0.62 Trimester 2
SO2: 0.63 Trimesters
Gouveia et al. (2004) Subjects: 179,460 live births First Trimester
Sao Paulo city, Brazil Group: Ministry of Health, Brazil
Year: 1997
Analysis method: GAM models 43.5 117.9 399.6
1Q
2Q
3Q
4Q
Second
Trimester
1Q
2Q
3Q
4Q
Third trimester
1Q
2Q
3Q
4Q
Odds Ratio (95% CI)
0.
0.
0.
1
1.
1.
1.
1
0.
1.
1.
1
0.
1.
1.
,98(0.81,1.18)
,99(0.80,1.23)
,97(0.80,1.18)
,060
,197
126
,986
,008
,034
,992
,041
,046
(0
(0.
(0.
(0.
(0.
(0.
(0.
(0.
(0.
.971-1.157)
885-1.619)
812-1.560)
902-1.076)
871-1.167
861-1.243)
913-1.078)
927-1.169
889-1.231)
O
HH
H
W
-------
TABLE AX6.3-12 (cont'd). BIRTH WEIGHT AND LONG-TERM NO2 EXPOSURE STUDIES
O
to
O
O
oo
X
H
6
o
o
H
O
O
H
W
Author, Year,
Location
Maroziene and
Grazuleviciene (2002)
Kaunas, Lithuania
Liu etal. (2003)
Vancouver
Salam et al. (2005)
Southern CA
Study Details
Subjects: 3,988 birth cert
Group: LEW adjusted for GA
Year: 1998
Analysis method: Logistic regression,
10 jig increase
Subjects: 229,085 birth cert
Years: 1986-1998
Group: LEW adjusted for GA
Averaging time: 24 h
Analysis method: Logistic regression,
10 ppb increase
Subjects: 3,901 birth cert
Group Term LEW, CHS:
Years: 1975-1987
Analysis method: Logistic regression
Distance: 5 km or 3 within 50 km, within
county
Cone Range (ppb)
Mid- Correlation with
Low range High Other Pollutants Outcome
Pregnancy
6.2(5.7) Medium NO2
HighNO2
Trimester 1
Trimester 2
Trimester 3
O3: -0.25 First mo
SO2: 0.61
CO: 0.72
15.1 18.1 22.3
Last mo
PM10: 0.55 Pregnancy
03: -0.1
CO: 0.41
36.1
(15.4)
IQR 25 Trimester 1
Trimester 2
Trimester 3
Odds Ratio (95%
CI)
1.28(0.97,1.68)
0.96(0.47,1.96)
1.54(0.80,2.96)
0.91(0.53,1.56)
0.93(0.61,1.41)
1.34(0.94,1.92)
0.98(0.90,1.07)
0.94(0.85,1.04)
0.8(0.4,1.4)
0.9(0.5,1.5)
1.0(0.6,1.6)
0.6(0.4,1.1)
O
HH
H
W
-------
TABLE AX6.3-12 (cont'd). BIRTH WEIGHT AND LONG-TERM NO2 EXPOSURE STUDIES
O
to
O
O
oo
X
to
Author, Year,
Location
Bell et al. (2007)
CT and MA
Slama et al.
(2007)
Munich
Study Details
Subjects: 358,504 birth cert
Group: LEW adjusted for GA
Years: 1999-2002
Analysis method: logistic regression,
interquartile linear regression, difference in
gms per IQR
Subj ects : 1016 non-premature births
Group: LISA
Analysis method: Poisson Regression
Cone Range (ppb)
Mid- Correlation with
Low range High Other Pollutants
PM2.5: 0.64
PM10: 0.55
17.4
(5.0)
IQR 4.8
0.52 0.75 0.90
Outcome
pregnancy
black mothers
white mothers
Adjusted 1Q
Adjusted 2Q
Adjusted 3Q
Adjusted 4Q
Continuous
coding
Odds Ratio (95%
CI)
1.027(1.002,1.051)
-12.7 (-18.0, -7.5)
-8.3 (-10.4, -6.3)
1
0.80(0.52-1.28)
1.32(0.86-2.09)
1.16(0.71-1.71)
1.21(0.86-1.68)
H
6
o
o
H
O
O
H
W
O
HH
H
W
-------
TABLE AX6.3-13. PRETERM DELIVERY AND LONG-TERM NO2 EXPOSURE STUDIES
O
to
O
O
oo
X
H
6
o
o
H
O
O
H
W
O
HH
H
W
Author, Year,
Location
Bobak (2000)
Czech
Liu S. et al. (2003)
Vancouver
Maroziene and
Grazuleviciene R.
(2002)
Kaunas, Lithuania
Rite et al. (2000)
Southern CA
Study Details
Subjects: 69,935 birth cert
Group: Preterm
Years: 1991
Avg time: 24 h
Analysis Method: Logistic regression, 50 jig
increase
Subjects: 229,085 birth cert
Group: Preterm
Years: 1986-1998
Avg time: 24 h
Distance: 13 monitors
Analysis Method: 10 ppb increase
Subjects: 3,988 birth cert
Group: Preterm
Analysis Method: Logistic regression
Subjects: 97,158 birth cert
Group: Preterm
Years: 1989-1993
Avg time: 24 h
Analysis Method: Logistic regression
Distance: Zip code within 2 miles
Cone Range (ppb)
Mid- Correlation with
Low range High Other Pollutants Outcome
12.2 20 31.1 SO2: 0.62 trimester 1
trimester 2
trimester 3
15.1 18.1 22.3 O3: -0.25 first mo
S02: 0.61
CO: 0.72
last mo
pregnancy
6.2(5.7) medium NO2
highNO2
trimester 1
trimester 2
trimester 3
32 40.9 50.4 PM10: 0.74 first mo
O3: -0.12
CO: 0.64
6 wks before
birth
Odds Ratio
(95% CI)
1.10(1.00,1.21)
1.08(0.98,1.19)
1.11(1.00,1.23)
1.01(0.94,1.07)
1.08(0.99,1.17)
1.25(1.07,1.46)
1.14(0.77,1.68)
1.68(1.15,2.46)
1.67(1.28,2.18)
1.13(0.90,1.40)
1.19(0.96,1.47)
No effects for
any preg period
No effects for
any preg period
-------
O
to
O
O
oo
TABLE AX6.3-13 (cont'd). PRETERM DELIVERY AND LONG-TERM NO2 EXPOSURE STUDIES
Cone Range (ppb)
Author, Year, Mid- Correlation with Odds Ratio
Location Study Details Low range High Other Pollutants Outcome (95% CI)
Leem et al. (2006)
Inchon, Korea
Subjects: 52,113 birth cert
Group: Preterm
Years: 2001-2002
Analysis Method: Log binomial regression
15.78
22.93
29.9
PM10: 0.37
SO2: 0.54
CO: 0.63
Trimester 1 Q2 1.13(0.99,1.27)
Trimester 1 Q3
Trimester 1 Q4
Trimester 3 Q2
Trimester 3 Q3
Trimester 3 Q4
1.07(0.94,1.21)
1.24(1.09,1.41)
Trend .02
1.06(0.93,1.20)
1.14(1.01,1.29)
1.21(1.07,1.37)
Trend <.001
X
Hansen et al. (2006)
Brisbane
Subjects: 28,200 birth cert
Group: Preterm
Years: 2000-2003
Avg time: 24 h
Analysis Method: Logistic regression
PM10: 0.32
O3: 0.13
trimester 1
90 days before
birth
0.93(0.78,1.12)
1.03 (0.86, 1.23)
H
6
o
o
H
O
O
H
W
O
HH
H
W
-------
TABLE AX6.3-14. FETAL GROWTH AND LONG-TERM NO2 EXPOSURE STUDIES
O
to
O
O
oo
Author, Year,
Location
Salam et al. (2005)
Southern CA, CHS
Study Details
Subjects: 3,901 birth cert
Group: Term SGA, <1 5% of data
Years: 1975-1987
Avg time: 24 h
Analysis Methods: Linear mixed model, IQR = 25
Distance: 5 km or 3 monitors within 50 km
Cone Range (ppb) Correlation
Mid- with Other
Low range High Pollutants
36.1 PM10: 0.55
(15.4) Q3: -0.1
CO: 0.69
Outcome
Pregnancy
Trimester 1
Trimester 2
Trimester 3
Odds Ratio
(95% CI)
1.1(0.9,1.3)
1.2(1.0,1.4)
1.0(0.8,1.2)
1.0(0.8,1.2)
X
Marines et al. (2005)
Sydney
Subjects: 51,460 birth cert
Group: SGA, >2sd below national data
Years: 1998-2000
Avg time: 1-hmax
Analysis Methods: Logistic regression, 1 ppb
Distance: 5 km
18
23
27.5
23.2
(7.4)
PM2.5: 0.66
PM10: 0.47
03: 0.29
CO: 0.57
Trimester 1
Trimester 2
Trimester 3
1 mo before
birth
1.06(0.99,1.14)
1.14(1.07,1.22)
1.13(1.05,1.21)
1.07(1.00,1.14)
H
6
o
o
H
O
O
H
W
Liu et al. (2003) Subjects: 229,085 birth cert
Vancouver Group: Term SGA, < 10% national
Years: 1986-1998
Avg time: 24 h
Analysis Methods: Logistic regression, 10 ppb
Distance: 13 monitors Avg
15.1
18.1
22.3
SO2: 0.61
O3: -0.25
CO: 0.72
Trimester 1
Trimester 2
Trimester 3
First mo
Last mo
1.03(0.98,1.10)
0.94(0.88,1.00)
0.98(0.92,1.06)
1.05(1.01,1.10)
0.98(0.92,1.03)
O
HH
H
W
-------
TABLE AX6.3-15. LUNG FUNCTION AND LONG-TERM NO2 EXPOSURE
O
to
O
O
oo
X
Oi
Oi
H
6
o
o
H
O
O
H
W
Author, Year
Gauderman
(2004)
Southern CA
Moseler et al.
(1994)
Frieberg,
Germany
Ackermann-
Liebrich et al.
(1997)
Switzerland
Study Details
Subjects: 1757 children age 10-18, CHS
Group: Lung function, Longitudinal
Avg Time: 24-h annual
Anlaysis Method: 2-stage linear Regression, 34.6 ppb
Distance: Study monitors in 12 towns
Subjects: 467 children age 9-16
Group: Lung function
Avg Time: Median wkly
Anlaysis Method: Linear regression, Parameter estimates
Subjects: 3,1 15 adults, 3-yr residents, nonsmokers,
SAPALDIA
Group: Lung function
Avg Time: 24-h annual
Anlaysis Method: 2-stage linear Regression
Distance: Monitors in 8 Study areas
Cone Range (ppb) Correlation
with Other
Low Mid-range High Pollutants Outcome
PM25: 0.79 Difference
PM10: 0.67 in lung
O3: -0.11 growth
FVC
FEVj
MMEF
2 1 .28 with asthma
threshold symp
FEVj
lnMEF75%
lnMEF50%
lnMEF25%
no asthma
symp
FEVj
lnMEF75%
lnMEF50%
lnMEF25%
18.9(8.5) PM10: 0.91 FVC
O3: -0.78
SO2: 0.86
FEVj
Odds Ratio (95% CI)
-95 (-183.4, -0.6)
-101.4 (-164. 5, -38.4)
-211 (-377.6, -44.4)
0.437
-0.011
-0.022
-0.029
-0.049
0.003
0.004
0.003
-0.0123 (-0.0152,
-0.0094)
-0.0070 (-0.0099,
-0.0041)
O
HH
H
W
-------
TABLE AX6.3-15 (cont'd). LUNG FUNCTION AND LONG-TERM NO2 EXPOSURE
O
to
O
O
oo
X
H
6
o
o
H
O
O
H
W
Author, Year,
Location
Schindler et al.
(1998)
Switzerland
Peters et al.
(1999a)
Southern CA
Tager et al.
(2005)
Southern &
Northern CA
Cone Range (ppb)
Analysis Mid-
Study Details Method LOW range High
Subjects: 560 adults, 3-yr residents, SAPALDIA
Group: Lung function
Avg Time: Wkly avg
Anlaysis Method: Linear regression
Distance: Personal and Home monitors
Subjects: 3,293 children, CHS
Group: Lung function
Avg Time: 24 h
Anlaysis Method: Linear regression
Distance: Study monitors in 12 towns
Subjects: 255 students UC Berkeley
Group: Lung function
Avg Time:
Anlaysis Method: Linear regression (men): 22 30 40
(women): 21 27 40
Correlation
with Other
Pollutants Outcome
FVC home
FVC personal
FEV home
FEV personal
FVC all: 1986-
1990
FVC girls:
1986-1990
FEViall: 1986-
1990
FEVi girls:
1986-1990
FVC all: 1994
FVC girls: 1994
FEViall: 1994
FEV! girls:
1994
lnFEF75 men
03: 0.57
lnFEF75 women
Odds Ratio (95% CI)
% change- 0.59 (-1)
-42.6(13.5)
-58.5(15.4)
-23.2(12.5)
-39.9(13.9)
-46.2(16.0)
-56.7(19.8)
-22.3(14.8)
-44.1(16.1)
-0.029 (0.003)
-0.032 (0.002)
O
HH
H
W
-------
O
to
O
O
oo
X
OO
TABLE AX6.3-16 ASTHMA AND LONG-TERM NO2 EXPOSURE
Author, Year,
Location Study Design
Garrett et al. (1 999) Subjects: 148 children ages 7-14
Latrobe Valley, Years: 1994-1995
Australia Distance: In home
Study Group: Asthma, Monash Q
Correlation Conc Ran§e (PPb)
with Other LOW Mid- High Odds Ratio
Analysis Method Pollutants range Study Factor (95% CI)
Logistic regression 6
10 ug
Bedroom NO2 1.01
(0.75,1.37)
Indoor mean 1 .00
(.075,1.31)
winter 0.99
(0.84,1.16)
summer 2.52
(0.99, 6.42)
Hirschetal. (1999)
Dresden, Germany
Subjects: 5,421 children ages 5-7, 9-11
Years: 1995-1996, 12 mo residence
Distance: 4 monitors within 1 km
Study Group: Asthma, ISAAC
Logistic regression
10 ug
29.3
37.8
Home address 1.16
(0.94,1.42)
Home & school 1.14
(0.86,1.51)
H
6
o
o
H
O
O
H
W
O
HH
H
W
Peters etal. (1999b)
Southern CA, CHS
Subjects: 3,676 children Age 9-16
Years: 1994
Avg time: 24 h
Distance: Study monitors in 12 towns
Study Groups: Asthma, Questionnaire
Logistic regression
IQR = 25 ppb
21.5
mean
all children
boys
girls
1.21
(0.850,1.71)
1.25
(0.90,1.75)
1.07
(0.57,2.02)
-------
TABLE AX6.3-16 (cont'd). ASTHMA AND LONG-TERM NO2 EXPOSURE
O
to
O
O
oo
X
H
6
o
o
H
O
O
H
W
Author, Year
Millstein et al. (2004)
Southern CA, CHS
Penard-Morand et al.
(2005)
France 6 towns
Studnicka et al.
(1997)
8 communities,
Lower Austria
Correlation
with Other
Study Design Analysis Method Pollutants
Subjects: 2,034 children age 9-11 Mixed effects PM2.5: 0.28
Years: 1995 model PM10: 0.39
Distance: Study monitors in 12 towns
Study Groups: Asthma, Medication use IQR = 5.74 ppb
Subjects: 4,901 children Age 9-11 Logistic regression PMi0: 0.46
Years: 1999-2000, 3 yr residence O3: 0.76
Avgtime: 3 yrs 10 ug SO2: 0.35
Distance: monitoring sites, school
address
Study Groups: Asthma, ISAAC
Subjects: 843 children Logistic regression
Distance: monitor in each community
Avg time: 3 yrs <.05
Study Group: A sthma, ISAAC
Cone Range (ppb)
Mid-
Low range High Study Factor
annual
Mar-August
Sept-Feb
8.7, 16.1, lifetime asthma
16.0 25.7
current asthma
8.0, 11.7, 14.7, Ever asthma low
8.7 13.3 17.0
Ever asthma
medium
Ever asthma
high
Current asthma
low
Currrent asthma
medium
Current asthma
high
Odds Ratio
(95% CI)
0.94
(0.71,
0.96
(0.68,
0.90
(0.66,
0.94
(0.83,
0.92
(0.77,
1.28
2.14
5.81
1.7
1.47
8.78
1.22)
1.37)
1.24)
1.07)
1.10)
O
HH
H
W
-------
TABLE AX6.3-16 (cont'd). ASTHMA AND LONG-TERM NO2 EXPOSURE
O
to
O
O
oo
X
Oi
Oi
O
H
6
o
o
H
O
O
H
W
Author, Year
Wang etal. (1999)
Taiwan
Ramadour et al.
(2000)
7 communities,
France
Shima and Adachi
et al. (2000)
7 communities, Japan
Study Design
Subjects: 1 17,080 students age 11-16
Distance: 24 district monitors
Study Group: Asthma
Subjects: 2,445 children
Years: 3-yr residence
age 13-14
Distance: Monitors in each community
Study Group: Asthma, ISAAC
Subjects: 905 children age 9-10
Distance: In home measurements
Monitors near schools
Study Groups: Asthma, Prevalence,
Incidence
Correlation Conc Ran§e (PPb)
with Other Mid-
Analysis Method Pollutants LOW range High Study Factor
Logistic regression 28 Current
median asthma
Above/below
median
Logisitic regression 11-27
mean
Logistic regression 20-29 30-39 >40 Outdoor 4th
grade girls
10-ppb increase 7-25
mean
Outdoors
Outdoor 5th
grade girls
Outdoor 6th
grade girls
Indoor 4th
grade girls
Indoor 5th
grade girls
Indoor 6th
grade girls
Outdoor
Indoor
Odds Ratio
(95% CI)
1.08
(1.04,1.13)
Nonsignificant
Results
1.14
(0.65,2.09)
1.14
(0.63,2.13)
0.95
(0.45,2.05)
1.63
(1.06,2.54)
1.67
(1.06,2.66)
1.18
(0.62,2.18)
2.10
(1.10,4.75)
0.87
(0.51,1.43)
O
HH
H
W
-------
TABLE AX6.3-16 (cont'd). ASTHMA AND LONG-TERM NO2 EXPOSURE
O
to
O
O
oo
X
Oi
Oi
H
6
o
o
H
O
O
H
W
O
HH
H
W
Author, Year
Kim et al. (2004a)
San Francisco Bay
area
Gauderman et al.
(2005)
Southern CA CHS
Hwang et al. (2005)
Taiwan,
National study
Study Subjects
Subjects: 1,109 children Age 9-11
Distance: 10 school sites
Study Group: Asthma
Subjects: 208 children
Avg time: 4 wk
Distance: Outside home
Study Group: Asthma
Subjects: 32,672 children
Distance: Schools within
1 km of monitors
Study Group: Asthma, ISAAC
Correlation Conc Ran§e (PPb)
with Other Mid-
Analysis Method Pollutants LOW range High Study Factor
2-stage Hierarchical PM25: "low" 24 mean All children
model O3: "low"
IQR = 3.6N02
IQR = 14.9 NOX
All 1 yr
residents
1 yr resident
girls
1 yr resident
boys
All children
All 1 yr
residents
1 yr resident
girls
1 yr resident
boys
Logistic regression 13-51 Lifetime
asthma
IQR =5. 7
Asthma med
use
2-stage Hierarchical PM10: 0.34 21.5 29.6 33.1 Parental atopy
model O3: -0.39
SO2: 0.5
lOppbNOx
No parental
atopy
Odds Ratio
(95% CI)
1.02
(0.97, 1.07)
1.04
(0.98,1.10)
1.09
(1.03,1.15)
1.00
(0.94, 1.07)
1.04
(0.97,1.11)
1.07
(1.00,1.14)
1.17
(1.06,1.29)
1.02
(0.93,1.11)
1.83
(1.04,3.21)
2.19
(1.20,4.01)
0.99
(0.92, 1.07)
1.02
(0.95,1.10)
-------
TABLE AX6.3-17. RESPIRATORY SYMPTOMS AND LONG-TERM NO2 EXPOSURE
O
to
O
O
oo
X
Oi
Oi
to
H
6
o
o
H
O
O
H
W
Cone Range (ppb) Correlation
Author, Year
Karr et al.
(2006)
Karr et al.
(2006)
Garrett et al.
(1999)
wheeze
cough
short of breath
chest tightness
any symptoms
any symptoms
any symptoms
Hirsch et al.
(1999)
wheeze home
wheeze school
cough home
cough school
cough non-
atopic child
Peters et al.
(1999b)
wheeze
cough
wheeze boys
wheeze girls
Study
Location
Southern
California
Southern
California
Latrobe
Valley
Australia
Dresden
Germany
Southern CA
CHS
Study Group Study Subjects
18,595 cases;
Infant 169,472 controls
Bronchiolitis ages 3 wks to 1 yr
18,595 cases;
Infant 169,472 controls
Bronchiolitis ages 3 wks to 1 yr
Symptoms 148 children
MonashQ Age 7- 14
1994-1995
Symptoms 5,421 children
ISAAC Age 5-7, 9 1 1
1995-1996
12 mo residence
Symptoms 3,676 children
Questionnaire Age 9- 16
1994
Odds Ration
(95% CI)
1.03 [0.99, 1.07]
per 16 ppb
1.04 [1.00, 1.08]
per 15 ppb
1.15(0.85, 1.54)
1.47(0.99,2.18)
1.23 (0.92, 1.64)
1.12(0.81, 1.56)
1.24(0.91, 1.68)
1.12(0.93, 1.35)
2.71(1.11,6.59)
1.13(0.93, 1.37)
0.95 (0.72, 1.26)
1.22(1.94, 1.44)
1.21 (0.96, 1.52)
1.42(1.10, 1.84)
1.12(0.86, 1.45)
1.14(0.94, 1.39)
1.54(1.04,2.29)
0.86(0.57, 1.29)
Analysis Mid- with Other
Method Unit of Averaging Time LOW range High Pollutants Distance
Conditional
logistic Chronic (lifetime avg of
regression 1-h daily max) (ppb) 12 58 204
Conditional Subchronic (avg of 1-h
logistic daily max 1 mo prior to
regression hospitalization) (ppb) 12 57 152
Logistic
regression 6
10 ug
10 ug mean
10 ug winter
10 ug summer
Logistic
regression 29.3 33.8 37.8
10 ug
Logistic 21.5
regression 24 h mean
IQR = 25 ppb
34 monitors
34 monitors
in home
4 monitors
Within 1 km
Study
monitors
in 12 towns
O
HH
H
W
-------
TABLE AX6.3-17 (cont'd). RESPIRATORY SYMPTOMS AND LONG-TERM NO2 EXPOSURE
fa
o
tr
to
o
o
oo
^
X
ON
ON
OJ
M
\^
C
uJ
1
o
o
o
H
O
o
cj
H
W
O
O
HH
H
W
Author, Year Study Location
Millstein et al.
(2004) Southern CA
wheeze CHS
wheeze
Mar-Aug
wheeze
Sept-Feb
Penard-
Morand et al.
(2005) France 6 towns
wheeze past
12mos.
Wageningen and
Roemer et al. Bennekom,
(1993) Netherlands
Mukala et al. Helsinki
(1999)
cough Finland
cough
nasal symp
winter
nasal symp
winter
nasal symp
spring
nasal symp
spring
Study Group Study Subjects
Symptoms 2,034 children
Age 9- 11
1995
Symptoms 4,901 children
IS SAC Age 9- 11
1999-2000
3-yr residence
73 children
grades 3-8
Symptoms Dec 1990-
Questionnaire Mar 1991
Symptoms 163 children
Age 3-6
1991
Odds Ration
(95% CI)
0.93(0.77, 1.12)
0.79(0.40, 1.53)
0.85(0.64, 1.14)
0.87(0.75, 1.01)
No associations
1.23 (0.89, 1.70)
1.52(1.00,2.31)
0.99(0.58, 1.68)
0.89(0.44. 1.82)
0.76(0.56, 1.02)
0.68(0.46, 1.01)
Unit of Conc Ran§e (PPb)
Averaging Mid-
Analysis Method Time Low range High
Mixed effects
model Moly
IQR = 5.74ppb
Logistic regression 3 yrs
8.7, 16.1,
10 ug 16.0 25.7
Time series using
Yule-Walker
estimation method 24 hr avg 127
GEE WklyAvg <8.6 8.6, >14.5
14.5
2nd tertile
3rd tertile
2nd tertile
3rd tertile
2nd tertile
3rd tertile
Correlation
with Other
Pollutants Distance
PM25: 0.28 Study monitors
PMi0: 0.39 in 12 towns
29 monitoring
sites,
school address
03: 0.76
SO2: 0.35
PMi0: 0.46
National Air
PM10: 0.57 Quality
SO2: 0.26 Monitoring
BS: 0.65 Network
Palmes tubes
On outer
garment
-------
TABLE AX6.3-17 (cont'd). RESPIRATORY SYMPTOMS AND LONG-TERM NO2 EXPOSURE
O
to
O
O
oo
X
ON
ON
H
6
o
o
H
O
O
H
W
Study
Author, Year Location
Pikhart et al. Prague
(2000)
wheeze Czech
wheeze
wheeze
Setiani, (1996) 6 cities in
Japan
Lacrimacy
Eye itch
Runny nose
Sore throat
Cough
Plegm
SOB
Sum of cough
with phlegm and
SOB
Van Strien
(2004) CT and MA
wheeze
wheeze
wheeze
cough
cough
cough
short of breath
short of breath
short of breath
Odds Ration
Study Group Study Subjects (95% CI)
Symptoms 3,045 children
SAVIAH Age 7-10 1.16(0.95,1.42)
1993-1994 1.07(0.86,1.33)
1.08(0.86, 1.36)
Hiroshima 13,836 adult non- Logistic regression
Community smoking women coefficient
Health Study aged 40-59 (standard error)
0.047 (0.046)
0.036 (0.046
-0.018(0.076)
0.059 (0.042)
-0.046(0.044)
-0.088(0.049)
-0.056(0.058)
-0.035 (0.030)
Symptoms 849 children
Agel2mos 1.15(0.79,1.67)
1.03 (0.69, 1.53)
1.45 (0.92, 2.27)
0.96(0.69, 1.36)
1.33 (0.94, 1.88)
1.52(1.00,2.31)
1.59(0.96,2.62)
1.95(1.17.3.27)
2.38(1.31,4.34)
Unit of Conc Ran§e (PPb> Correlation
Averaging Mid- with Other
Analysis Method Time Low range High Pollutants Distance
Multi-level model 14.8 19 24.1
Individual covariates
Ecological covariates
Both covariates
Individual multiple 24 h graph graph graph SPM: 0.606
linear regression ox: -0.337
analysis
Poisson regression 10-14 day 5.1 9.9 17.4 In home
Q2 Avg
Q3
Q4
Q2
Q3
Q4
Q2
Q3
Q4
O
HH
H
W
-------
TABLE AX6.3-17 (cont'd). RESPIRATORY SYMPTOMS AND LONG-TERM NO2 EXPOSURE
O
to
O
O
oo
:x!
rN
ON
i
Study Study
Author, Year Location Group
Nitschke et al.
(2006) Adelaide Symptoms
Wheeze school Australia
Wheeze home
Cough school
Cough home
Difficult breath
school
Difficult breath
home
Chest tight
school
Chest tight home
Hoek and
Brukekreef Wageningen, Primary
(1993) Netherlands school
Study Odds Ration
Subjects (95% CI)
174 asthmatic
Children, age
5-13 0.99(0.93,1.06)
2000 1.00(0.90,1.11)
1.01 (0.98, 1.04)
0.99(0.96, 1.02)
1.11 (1.05, 1.18)
1.03(1.01, 1.05)
1.12(1.07, 1.17)
1.02(0.95, 1.09)
1 12 children
grades 4-7 No association
Unit of Conc Ran§e (PPb)
Averaging Mid-
Analysis Method Time Low range High
School 117
Zero-inflated negative 34 (28) max
Home 147
binomial regression 20 (22) max
10-ppb increase
Individual linear
regression analysis and
distribution of individual
regression slopes 24-h 127
Correlation
with Other
Pollutants Distance
9 days in class
3 days at home
PMi»: 0.55
S02: 0.28
BS: 0.65
H
6
o
o
H
O
O
H
W
O
HH
H
W
-------
TABLE AX6.3-17 (cont'd). RESPIRATORY SYMPTOMS AND LONG-TERM NO2 EXPOSURE
O
to
O
O
oo
X
ON
ON
H
6
o
Study Study
Author, Year Location Group
Delfmo et al. Southern Asthmatic
(2006) California children
Personal NO2
Not taking anti-
inflammatory
meds
Taking anti-
inflammatory
meds
Inhaled
corticosteroids
Antileukotrienes
± inhaled
corticosteroes
Central site NO2
Not taking anti-
inflammatory
meds
Taking anti-
inflammatory
meds
Inhaled
corticosteroids
Antileukotrienes
± inhaled
corticosteroes
Study Odds Ration
Subjects (95% CI) Analysis Method
Linear mixed effects
45 children models (Verbeke and
ages 9- 1 8 Molenberghs 200 1 )
0.80 (-3.01 to 4.61)
1.67 (0.55 to 2.79)
1.22 (0.04 to 2.40
1.73 (-0.70 to 4.16)
0.96 (-1.34 to 3.26)
1.48 (0.47 to 2.50)
1.32 (0.33 to 2.32)
-7.5 (-2.83 to 1.32)
Unit of Cone Range (ppb) Correlation
Averaging Mid- with Other
Time Low range High Pollutants
Personal NO2,
personal PM2.5:
24-h 0.33
Central NO2,
personal PM25:
0.22
Central NO2,
central PM25:
0.25
Distance
Backpack monitor,
active sampling system,
central site exposure
O
H
O
O
c
H
W
O
HH
H
W
-------
TABLE AX6.3-17 (cont'd). RESPIRATORY SYMPTOMS AND LONG-TERM NO2 EXPOSURE
fa
o
tr
to
o
o
oo
X
ON
ON
^
O
^
TJ
H
6
o
2|
0
H
O
o
H
W
Study
Author, Year Study Location Group
Salome et al.
(1996) Australia asthmatic
Day of exposure-
room air
Change in
symptom score:
adult
Change in
symptom score:
Child
Week following
exposure- room air
Severity score:
adult
Severity score:
child
Russia, Austria,
Pattenden et al. Italy, Switzerland,
(2006) Netherlands PATY
Wheeze
Asthma
Bronchitis
Phlegm
Nocturnal cough
Morning cough
Sensitivity to
inhaled allergens
Hay fever
Itchy rash
Woken by wheeze
Allergy to pets
Unit of Conc Ran§e (PPb) Correlation
Odds Ration Averaging Mid- with Other
Study Subjects (95% CI) Analysis Method Time Low range High Pollutants Distance
20 (9 adults and 0.02 1.12
11 children) ANOVA ppm ppm
0.01 (0.38)
-0.02(0.26)
4.38(1.5)
4.20(1.3)
23,955 children Logistic regression,
ages 6-12 Cochran y? 12.45 50.00 variable
1993-1999
1.01(0.93-1.10)
1.02(0.94-1.09)
0.99(0.88-1.12)
1.05(0.95-1.17)
1.13(0.94-1.35)
1.15(1.01-1.30)
1.13(1.01-1.26)
1.04(0.98-1.11)
1.05(0.98-1.12)
1.06(0.89-1.26)
1.14(0.99-1.31)
O
HH
H
W
-------
TABLE AX6.3-17 (cont'd). RESPIRATORY SYMPTOMS AND LONG-TERM NO2 EXPOSURE
O
to
O
O
oo
X
Oi
Oi
OO
H
6
o
o
H
O
O
H
W
Author, Year
Chen etal. (1999)
Day average
1 day lag- FVC
1 day lag- FEVio
2 day lag- FVC
2 day lag FEVio
7 day lag- FVC
7 day lag- FEV10
Daytime Peak
1 day lag- FVC
1 day lag- FEVio
2 day lag- FVC
2 day lag FEVio
7 day lag- FVC
7 day lag- FEVio
Unit of Conc RanSe (PPb> Correlation
Study Study Odds Ration Analysis Averaging Mid- with Other
Location Study Group Subjects (95% CI) Method Time Low range High Pollutants Distance
Taiwan Study on Air 941 children Coefficient models 9.2 141.6
Pollution and Health ages 8-13 (standard error)
-2.66(1.23)
-0.46(1.16)
-3.32(1.53)
-0.93(1.45)
1.39(1.71)
2.52(1.61)
-0.59(0.40)
-0.16(0.37)
-1.33(0.72)
-0.36(0.68)
-0.13(0.87)
0.43 (0.82
O
HH
H
W
-------
TABLE AX6.3-18. LUNG CANCER
O
to
O
O
oo
X
ON
ON
H
6
o
o
H
O
O
H
W
Cone Range (ppb)
Mid-
Author, Year, Location Exposure Study Subjects Low range High Analysis
Nyberg et al. (2000) From addresses 1 ,042 cases, 2,364 controls
Stockholm, Sweeden and traffic men age 40-75 8.1 10.6 13.3 logistic regression
30-yr estimated exposure
10 ng
Q2
Q3
Q4
10-yr estimated exposure
10 jig
Q2
Q3
Q4
90th percentile
Nafstad (2004) Home address 1 6,209 men age 40-49 at
Norway 1972-1974 entry followed 1972-1998 5.32 10.6 16 Cox proportional
lung cancer incidence
10 ng
Q2
Q3
Q4
non-lung cancer
10 jig
Q2
Q3
Q4
Odds Ratio (95% CI)
1
1
0
1
1
1
1
1
1
1
0
1
1
1
0
1
1
.05 (0.
.18(0.
.90 (0.
.05 (0.
.10(0.
.15(0.
.01 (0.
.07 (0.
.44(1.
.08(1.
.90 (0.
.06 (0.
.36(1.
.02 (0.
.98 (0.
.05 (0.
.04 (0.
,93,
,93,
,71,
,79,
,97,
,91,
,79,
,81,
,05,
,02,
,70,
,81,
,01,
,99,
,88,
,94,
,91,
1.18)
1.49)
1.14)
1.40)
1.23)
1.46)
1.29)
1.42)
1.99)
1.15)
1.15)
1.38)
1.83)
1.06
1.08)
1.18)
1.18)
O
HH
H
W
-------
O
to
O
O
oo
X
TABLE AX6.3-19. EFFECTS OF ACUTE NOX EXPOSURE ON MORTALITY. RISK ESTIMATES ARE
STANDARDIZED FOR PER 20 ppb 24-h AVG NO2 INCREMENT
Reference, Study
Location, and Period
Outcome
Measure
Mean NO2 Levels
Copollutants
Considered
Lag Structure
Reported
Method/Design
Effect Estimates
META ANALYSIS
Stieb et al. (2002),
re-analysis (2003)
meta-analysis of
estimates from
multiple countries.
All cause
24-h avg ranged from
13 ppb (Brisbane,
Australia) to 38 ppb
(Santiago, Chile).
"Representative"
concentration: 24 ppb
PMio, O3, SO2, CO
The lags and
multiday
averaging used in
these estimates
varied
Meta-analysis of
Time-series study
results
Single-pollutant model (11
estimates): 0.8%(95%CI:
0.2,1.5);
Multipollutant model
estimates (3 estimates):
0.4%(95%CI: -0.2,1.1)
UNITED STATES
Samet et al. (2000a,b)
(reanalysis Dominici
et al., 2003) 90 U.S.
All cause;
cardiopulmonary
Ranged from 9 ppb
(Kansas City) to 39
ppb (Los Angeles),
PMio, O3, SO2,
CO; two-pollutant
models
0,1,2
Poisson GAM,
reanalyzed with
stringent convergence
24-h avg NO2 (per 20 ppb):
Posterior means:
All cause:
cities (58 U.S. cities
with NO2 data)
1987-1994
24-h avg
criteria; Poisson GLM.
Time-series study.
Lagl: 0.50% (0.09, 0.90)
Lag 1 with PM10 and SO2:
0.48% (-0.54, 1.51)
H
6
o
o
H
O
O
H
W
Kinney and Ozkaynak
(1991)
Los Angeles County,
CA
1970-1979
Kelsalletal. (1997)
Philadelphia, PA,
1974-1988
All cause; 69 ppb, 24-h avg
respiratory;
circulatory
All cause; 39.6 ppb, 24-h avg
respiratory;
cardiovascular,
KM (particle optical 1
reflectance), NO2,
S02, CO;
multipollutant models
TSP, CO, SO2, O3 0 (AIC presented
for 0 through 5)
OLS (ordinary least
squares) on high-pass
filtered variables.
Time-series study.
Poisson GAM
All cause:
Exhaustive multipollutant
model: 0.5% (-0.1, 1.2);
Two pollutant with Ox:
0.7% (0.5, 1.0)
All cause:
Single pollutant: 0.3%
(-0.6,1.1);
With TSP: -1.2%
(-2.2, -0.2)
O
HH
H
W
-------
O
to
O
O
oo
TABLE AX6.3-19 (cont'd). EFFECTS OF ACUTE NOX EXPOSURE ON MORTALITY. RISK ESTIMATES ARE
STANDARDIZED FOR PER 20 ppb 24-h AVG NO2 INCREMENT
X
H
6
o
o
H
O
O
H
W
Reference, Study
Location, and Period
Outcome
Measure
Copollutants
Mean NO2 Levels Considered
Lag Structure
Reported Method/Design
Effect Estimates
UNITED STATES (cont'd)
Ostro et al. (2000)
Coachella Valley, CA
1989-1998
Fairley (1999;
reanalysis
Fairley, 2003)
Santa Clara County,
CA
1989-1996
Gamble (1998)
Dallas, TX
1990-1994
Dockeryetal. (1992)
St. Louis, MO and
Eastern Tennessee
1985-1986
All cause;
respiratory;
cardiovascular;
cancer; other
All cause;
respiratory;
circulatory
All cause;
cardiopulmonary
All cause
20 ppb, 24-h avg PM10, PM2.5,
PM10-2.5, O3, CO
28 ppb, 24-h avg PM10, PM2.5,
PM10-2.5, S042\
coefficient of
haze, NQ3, O3,
S02;
1 5 ppb, 24-h avg PM10, O3, SO2,
CO; two-pollutant
models
St. Louis: 20 ppb; PM10, PM2.5, SO4,
Eastern Tennessee: LT", O3, SO2
12.6 ppb, 24-h avg
0-4 Poisson GAM with
default convergence
criteria. Time-series
study.
0, 1 Poisson GAM,
reanalyzed with
stringent convergence
criteria; Poisson GLM.
Time-series study.
Avg 4-5 Poisson GLM. Time-
series study.
Lag 1 Poisson with GEE.
Time-series study.
Lag 0 day:
All cause: 5. 5% (1.0, 10.3)
Respiratory: 1.8%(-10.3, 15.5)
Cardiovascular: 3.7% (- 1 .7, 9.3)
Lagl:
All cause: 1.9% (0.2, 3.7);
Cardiovascular: 1 .4% (- 1 .7, 4.5);
Respiratory: 4. 8% (-0.3, 10.2)
All cause: 4.4% (0.0, 9.0)
Cardiovascular: 1.9% (-4.6, 9.0)
Respiratory: 13. 7% (-2.0, 32.0)
All cause:
St. Louis,MO: 0.7% (-3.5, 5.1)
Eastern Tennessee: 3. 9% (-8.7, 18.2)
O
HH
H
W
-------
O
to
O
O
oo
X
to
H
6
o
o
H
O
O
H
W
O
HH
H
W
TABLE AX6.3-19 (cont'd). EFFECTS OF ACUTE NOX EXPOSURE ON MORTALITY. RISK ESTIMATES ARE
STANDARDIZED FOR PER 20 ppb 24-h AVG NO2 INCREMENT
Reference, Study
Location, and Period Outcome Measur
UNITED STATES (cont'd)
Moolgavkar (2003) All cause;
Cook County, IL and cardiovascular
Los Angeles County,
CA, 1987 1995
e Mean NO2 Levels
Cook County: 25 ppb;
Los Angeles: 38 ppb,
24-h avg
Copollutants Lag Structure
Considered Reported
PM2 5, PMio, O3, 0,1,2,3,4,5
SO2, CO; two
pollutant models
Method/Design
Poisson GAM with
default convergence
criteria. Time-series
study.
Effect Estimates
All cause:
Lagl:
Cook County:
Single pollutant:
Moolgavkar
(2000a,b,c);
re-analysis (2003).
Cook County, IL; Los
Angeles County, CA,
and Maricopa County,
AZ,
1987-1995
Cardiovascular;
cerebro vascular;
COPD
Cook County: 25 ppb;
Los Angeles: 38 ppb;
Maricopa County: 19
ppb, 24-h avg
PM2.5, PMio, 03,
SO2, CO; two- and
three-pollutant
models
0,1,2,3,4,5
Poisson GAM with
default convergence
criteria in the original
Moolgavkar (2000);
GAM with stringent
convergence criteria
and GLM with natural
splines in the 2003
re-analysis. The 2000
analysis presented
total death risk
estimates only in
figures.
2.2% (1.3, 3.1); with
PM10: 1.8% (0.7,
3.0);
Los Angeles: Single
pollutant: 2.0% (1.6,
2.5); with PM25:
1.8% (0.1, 3.6).
GAM, Lagl:
Cardiovascular:
Cook County: 1.1%
(-0.5,2.8); Los
Angeles: 2.8% (2.0,
3.6); Maricopa Co.:
4.6% (0.5, 9.0);
Re-analysis, GLM:
Total deaths: 2.5%
(1.5,3.6)
Lippmann et al.
(2000); reanalysis Ito,
(2003, 2004)
Detroit, MI
1985-1990
1992-1994
All cause;
respiratory;
circulatory;
cause-specific
1985-1990: 23.3 ppb,
24-h avg
1992-1994: 21.3 ppb,
24-h avg
PM10,PM2.5, 0,1,2,3,0-1,
PMio-2.5, S042\ Lf, 0-2, 0-3
03, S02, CO;
two-pollutant
models
Poisson GAM,
reanalyzed with
stringent convergence
criteria; Poisson GLM.
Numerical NO2 risk
estimates were not
presented in the re-
analysis. Time-series
study.
Poisson GAM:
All cause:
Lagl:
1985-1990:
0.9% (-1.2, 3.0)
1992-1994:
1.3% (-1.5, 4.2)
-------
O
to
O
O
oo
TABLE AX6.3-19 (cont'd). EFFECTS OF ACUTE NOX EXPOSURE ON MORTALITY. RISK ESTIMATES ARE
STANDARDIZED FOR PER 20 ppb 24-h AVG NO2 INCREMENT
X
H
6
o
o
H
O
O
H
W
Reference, Study
Location, and Period
Outcome Measure Mean NO2 Levels
Copollutants
Considered
Lag Structure
Reported Method/Design
Effect Estimates
UNITED STATES (cont'd)
Lipfert et al. (2000)
Seven counties in
Philadelphia, PA area
1991-1995
Chock et al. (2000)
Pittsburgh, PA
1989-1991
De Leon etal. (2003)
New York City, NY
1985-1994
Klemm and Mason
(2000); Klemm et al.
(2004)
Atlanta, GA
Aug 1998-M2000
All cause; 20.4 ppb, 24-h avg
respiratory;
cardiovascular;
all ages; age
65+ yrs; age
<65 yrs; various
subregional
boundaries
All cause; age Not reported.
<74 yrs;
age 75+ yrs
Circulatory and 40.6 ppb, 24-h avg
cancer with and
without contributing
respiratory causes
All cause; 51.3 ppb, max 1-h.
respiratory;
cardiovascular;
cancer; other; age
<65 yrs; age
65+ yrs
PMio, PM2.5,
PM10_2.5, S04
O3, other PM
indices, NO2, SO2,
CO; two-pollutant
models
PMio, NO2, SO2,
CO; two-, live-,
and six-pollutant
models
PMio, O3, SO2,
CO; two-pollutant
models
PM2.5, PMio-2.5,
EC, OC, 03,
SO42~,
N03, S02, CO
0-1 Linear with 1 9-day
weighted avg
Shumway filters.
Time-series study.
Numerous results.
0, plus minus 3 days. Poisson GLM. Time-
series study.
Numerous results
0 or 1 Poisson GAM with
stringent convergence
criteria; Poisson GLM.
Time-series study.
0-1 Poisson GLM using
quarterly, moly, or
biweekly knots for
temporal smoothing.
Time-series study.
All-cause, avg of 0- and
1-day lags,
Philadelphia:
2.2% (p > 0.05)
All cause, lag 0, age
0-74: 0.5% (-2.4, 3.5);
age 75+: 1.0% (-1.9, 4.0)
Gaseous pollutants results
were given only in
figures. Circulatory:
Age < 75: -1%
Age 75+: -2%
All cause, age 65+ yrs:
avg 0-1 days
Quarterly knots:
1.0% (-4.2,6.6);
Moly knots:
3.1% (-3.0, 9.7);
Bi-wkly knots:
0.9% (-5.9, 8.2)
O
HH
H
W
-------
O
to
O
O
oo
X
H
6
o
o
H
O
O
H
W
O
HH
H
W
TABLE AX6.3-19 (cont'd). EFFECTS OF ACUTE NOX EXPOSURE ON MORTALITY. RISK ESTIMATES ARE
STANDARDIZED FOR PER 20 ppb 24-h AVG NO2 INCREMENT
Reference, Study
Location, and Period
Outcome Measure
Mean NO2 Levels
Copollutants
Considered
Lag Structure
Reported Method/Design
Effect Estimates
UNITED STATES (cont'd)
Gwynn et al. (2000)
Buffalo, NY
Time-series study.
All cause;
respiratory;
circulatory
24-h avg 21 ppb
PMio, CoH, O3, SO2,
CO, If, SO42~
Poisson GAM with
Default convergence
criteria.
All cause (lag 3):
2.1% (-0.3, 4.6);
Circulatory (lag 2):
1.3% (-2. 9, 5.6);
Respiratory (lag 1):
6.4% (-2. 5, 16.2)
CANADA
Burnett et al. (2004)
12 Canadian cities
1981-1999
All cause
24-h avg ranged from
10 (Saint John) to 26
(Calgary) ppb.
PM25,PM10-25, O3,
S02, CO
1,0-2 Poisson GLM. Time-
series study.
Lag 0-2, single
pollutant: 2.0%
(1.1, 2.9); with O3:
1.8% (0.9, 2.7)
Burnett et al. (2000);
re-analysis (2003)
8 Canadian cities
1986-1996
All cause
24-h avg ranged from
15 (Winnipeg) to 26
(Calgary) ppb.
PM2.5,PM10,
PM2.5_10, S02,
03,CO
0, 1, 0-2
Poisson GAM with
default convergence
criteria. Time-series
study. The 2003 re-
analysis did not
consider gaseous
pollutants.
Days when PM indices
available, lag 1, single
pollutant: 2.4% (0.7,
4.1);withPM25: 3.1%
(1.2,5.1)
Days when PM indices
available, lag 1, single
pollutant: 3.6%
(1.6, 5.7); with PM25:
2.8% (0.5, 5.2)
Burnett et al. (1 998a), All cause
1 1 Canadian cities
1980-1991
24-h avg ranged from SO2, O3, CO
14 (Winnipeg) to 28
(Calgary) ppb.
0,1,2,0-1,0-2
examined but the
best lag/averaging
for each city
chosen
Poisson GAM with
default convergence
criteria. Time-series
study.
Single pollutant: 4.5%
(3.0, 6.0); with all
gaseous pollutants:
3. 5% (1.7, 5.3)
-------
O
to
O
O
oo
TABLE AX6.3-19 (cont'd). EFFECTS OF ACUTE NOX EXPOSURE ON MORTALITY. RISK ESTIMATES ARE
STANDARDIZED FOR PER 20 ppb 24-h AVG NO2 INCREMENT
Reference, Study
Location, and Period Outcome Measure Mean NO2 Levels
Copollutants
Considered
Lag Structure
Reported
Method/Design
Effect Estimates
CANADA (cont'd)
Burnett etal. (1998b),
Toronto,
1980-1994
All cause
24-h avg 25 ppb.
SO2, O3, CO, TSP,
COH, estimated PM10,
estimated PM2 s
0,1,0-1
Poisson GAM with
default convergence
criteria. Time-series
study.
Single pollutant (lag
0): 1.7% (0.7, 2.7);
with CO: 0.4%
(-0.6,1.5)
X
H
6
o
o
H
O
O
H
W
Vedal et al. (2003)
Vancouver, British
Columbia, Canada
1994-1996
Villeneuve et al.
(2003)
Vancouver, British
Columbia, Canada
1986-1999
Goldberg et al. (2003)
Montreal, Quebec,
Canada
1984-1993
All cause; 17 ppb, 24-h avg
respiratory;
cardiovascular
All cause; 1 9 ppb, 24-h avg
respiratory;
cardiovascular;
cancer;
socioeconomic
status
Congestive heart 22 ppb, 24-h avg
Failure (CHF) as
underlying cause of
death vs. those
classified as having
congestive heart
failure 1 yr prior to
death
PMio, O3, SO2, 0,1,2
CO
PM2.5,PM10, 0,1,0-2
PMio-2.5, TSP,
coefficient of
haze, SO42~, SO2,
O3, CO
PM2.5, coefficient 0,1,0-2
of haze, SO42~,
S02, 03, CO
Poisson GAM with
stringent convergence
criteria. Time-series
study. By season.
Poisson GLM with
natural splines.
Time-series study.
Poisson GLM with
natural splines.
Time-series study.
Results presented in
figures only. NO2
showed associations
in winter but not in
summer.
All yr:
All cause
Lag 1: 4.0%
(0.9, 7.2)
Respiratory:
LagO: 2.1%
(-3.0,7.4)
Cardiovascular:
LagO: 4.3%
(-4.2, 13.4)
CHF as underlying
cause of death:
Lag 1: 1.0%
(-5.1,7.5)
Having CHF 1 yr
prior to death:
Lag 1: 3.4%
(0.9,6.0)
O
HH
H
W
-------
O
to
O
O
oo
TABLE AX6.3-19 (cont'd). EFFECTS OF ACUTE NOX EXPOSURE ON MORTALITY. RISK ESTIMATES ARE
STANDARDIZED FOR PER 20 ppb 24-h AVG NO2 INCREMENT
X
Oi
Oi
H
6
o
o
H
O
O
H
W
O
HH
H
W
Reference, Study
Location, and Period Outcome Measure
Copollutants
Mean NO2 Levels Considered
Lag Structure
Reported Method/Design
Effect Estimates
EUROPE
Samoli et al. (2006) All cause,
30 APHEA2 cities. respiratory;
Study periods vary cardiovascular
by city, ranging from
1990 to 1997
Samoli et al. (2005) All-cause
9 APHEA2 cities.
Period not reported.
Touloumi et al. (1997) All cause
Six European cities:
London, Paris, Lyon,
Barcelona, Athens,
Koln.
Study periods vary by
city, ranging from
1977 to 1992
1-h max ranged from BS, PM10, SO2, O3
24 (Wroclaw) to 81
(Milan) ppb
The selected cities had None
1 -h max medians
above 58 ppb and the
third quartiles above
68.
Ranged from 37 BS, O3; two-pollutant
(Paris) to 70 (Athens) models
ppb, 1-h max
01 Poisson model with
penalized splines.
0 1 Poisson model with
either non-parametric
or cubic spline smooth
function in each city,
and combined across
cities.
0,1,2,3,0-1,0-2, Poisson
0-3 (best lag autoregressive. Time-
selected for each series study.
city)
All-cause: single:
1.8% (1.3, 2.2); with
SO2: 1.5% (1.0, 2.0)
Cardiovascular: single:
2.3%
(1.7, 3.0); with SO2:
1.9% (1.1, 2.7)
Respiratory: single:
2.2% (1.0, 3.4); with
SO2: 1.1%
(-0.4,2.6)
No numeric estimate
presented. The
concentration-response
was approximately
linear.
All-cause: Single-
pollutant model:
1.0% (0.6, 1.3);
WithBS:
0.5% (0.0, 0.9).
Zmirouetal. (1998)
Four European cities:
London, Paris, Lyon,
Barcelona
Study periods vary by
city, ranging from
1985-1992
Respiratory;
cardiovascular
Ranged from 24 BS, TSP, SO2, O3
(Paris) to 37 (Athens)
ppb in cold season and
23 (Paris) to 37
(Athens) ppb in warm
season, 24-h avg
0,1,2,3,0-1,0-2,
0-3 (best lag
selected for each
city)
Poisson GLM. Time-
series study.
Western Europe:
Respiratory:
0.0% (-1.1, 1.1)
Cardiovascular:
0.8% (0.0, 1.5)
-------
O
to
O
O
oo
X
TABLE AX6.3-19 (cont'd). EFFECTS OF ACUTE NOX EXPOSURE ON MORTALITY. RISK ESTIMATES ARE
STANDARDIZED FOR PER 20 ppb 24-h AVG NO2 INCREMENT
Reference, Study
Location, and Period
Outcome Measure
Mean NO2 Levels
Copollutants Lag Structure
Considered Reported
Method/Design
Effect Estimates
EUROPE (cont'd)
Biggeri et al. (2005)
8 Italian cities,
Period variable
between
1990-1999
Anderson etal. (1996)
London, England
1987-1992
All cause;
respiratory;
cardiovascular
All cause;
respiratory;
cardiovascular
24-h avg ranged from
30 (Verona) to 51
(Rome) ppb
37 ppb, 24-h avg
Only single-pollutant 0-1
models; O3, SO2, CO,
PM10
BS, O3, SO2; 0, 1
two-pollutant models
Poisson GLM. Time-
series study.
Poisson GLM. Time-
series study.
All cause:
3. 6% (2.3, 5.0);
Respiratory:
5. 6% (0.2, 11.2)
Cardiovascular:
5.1% (3.0, 7.3)
All cause (Lag 1 ):
0.6% (-0.1, 1.2);
Respiratory (lag 1):
-0.7% (-2.3, 1.0)
Cardiovascular:
0.5% (-0.4, 1.4)
Bremneretal. (1999)
London, England
1992-1994
All cause;
respiratory;
cardiovascular; all
cancer; all others; all
ages; age specific
(0-64, 65+, 65-74,
75+ yrs)
34 ppb, 24-h avg
BS, PMio, 03, S02,
CO; two-pollutant
models
Selected best from
0, 1,2, 3, (all cause);
0,1,2,3,0-1,0-2,
0-3 (respiratory,
cardiovascular)
Poisson GLM. Time- All cause (lag 1):
series study. 0.9% (0.0, 1.9)
Respiratory (lag 3):
1.9% (-0.3, 4.2)
Cardiovascular (lag 1):
1.9% (0.6, 3.2)
H
6
o
o
H
O
O
H
W
Anderson etal. (2001)
West Midlands region,
England
1994-1996
All cause; 37 ppb, 1-hmax
respiratory;
cardiovascular.
PMio, PM2.5, 0-1
PM25_10,BS, SO42~,
03, S02, CO
Poisson GAM with
default convergence
criteria. Time-series
study.
All cause:
1.7% (-0.5, 3.8)
Respiratory:
3. 3% (-1.9, 8.8)
Cardiovascular:
3.1% (-0.2, 6.4)
O
HH
H
W
-------
O
to
O
O
oo
TABLE AX6.3-19 (cont'd). EFFECTS OF ACUTE NOX EXPOSURE ON MORTALITY. RISK ESTIMATES ARE
STANDARDIZED FOR PER 20 ppb 24-h AVG NO2 INCREMENT
X
OO
H
6
o
o
H
O
O
H
W
Reference, Study
Location, and Period
Outcome Measure
Copollutants
Mean NO2 Levels Considered
Lag Structure
Reported Method/Design
Effect Estimates
EUROPE (cont'd)
Prescottetal. (1998)
Edinburgh, Scotland
1992-1995
Le Tertre et al. (2002)
Le Havre, Lyon, Paris,
Rouen, Strasbourg, and
Toulouse, France
Study periods vary by
city, ranging from
1990-1995
Zeghnoun et al. (2001)
Rouen and Le Havre,
France 1990-1995
Dab etal. (1996)
Paris, France
1987-1992
All cause;
respiratory;
cardiovascular; all
ages; age <65 yrs;
age >65 yrs
All cause;
respiratory;
cardiovascular
All cause;
respiratory;
cardiovascular
Respiratory
26 ppb, 24-h avg BS, PM10, O3, SO2,
CO; two-pollutant
models
Ranged from BS, O3, SO2
1 5 (Toulouse) to
28 (Paris) ppb,
24-h avg
24-h avg 1 8 ppb in SO2, BS, PM13, O3
Rouen; 20 ppb in
Le Havre
24 ppb, 24-h avg BS, PM13, O3, SO2,
CO
0 Poisson GLM.
Time-series study.
0-1 Poisson GAM with
default convergence
criteria. Time-series
study.
0, 1 , 2, 3, 0-3, Poisson GAM with
default convergence
criteria. Time-series
study.
0 Poisson
autoregressive.
Time-series study.
Results presented as
figures only.
Essentially no
associations in all
categories. Very wide
confidence intervals.
Six-city pooled
estimates:
All cause:
2. 9% (1.6, 4.2)
Respiratory:
3.1% (-1.7, 8.0)
Cardiovascular:
3. 5% (1.1, 5. 9)
All cause in Rouen
(lag 1): 5.5%
(0.2, 11.1); in
Le Havre (lag 1):
2.4% (-3.4, 8.5)
Lagl:
2.1% (3. 1,7.7)
O
HH
H
W
-------
O
to
O
O
oo
X
VO
TABLE AX6.3-19 (cont'd). EFFECTS OF ACUTE NOX EXPOSURE ON MORTALITY. RISK ESTIMATES ARE
STANDARDIZED FOR PER 20 ppb 24-h AVG NO2 INCREMENT
Reference, Study
Location, and Period
Outcome Measure
Mean NO2 Levels
Copollutants Lag Structure
Considered Reported Method/Design
Effect Estimates
EUROPE (cont'd)
Zmirouetal. (1996)
Lyon, France
1985-1990
Sartor etal. (1995)
Belgium
Summer 1994
All cause;
respiratory;
cardiovascular;
digestive
All cause; age <65
yrs; age 65+ yrs
37 ppb, 24-h avg
24-h avg NO2:
Geometric mean:
During heat wave
(42 day period): 17 ppb
Before heat wave
(43 day period): 15 ppb
After heat wave
(39 day period): 13 ppb
PM13, SO2, 03 Selected best Poisson GLM.
from Time-series study.
0,1,2,3
TSP, NO, 03, S02 0,1,2 Log-linear
regression for O3
and temperature.
Time-series study.
All cause (lag 1 ):
1.5% (-1.5, 4.6)
Respiratory (lag 2):
-2. 3% (-15. 6, 13.0)
Cardiovascular (lag 1):
0.8% (-2.7, 4. 3)
Only correlation coefficients
presented for NO2. Unlike
O3, NO2 was not particularly
elevated during the heat
wave.
H
6
o
o
H
O
O
H
W
O
HH
H
W
-------
O
to
O
O
oo
TABLE AX6.3-19 (cont'd). EFFECTS OF ACUTE NOX EXPOSURE ON MORTALITY. RISK ESTIMATES ARE
STANDARDIZED FOR PER 20 ppb 24-h AVG NO2 INCREMENT
Reference, Study
Location, and Period
Outcome Measure Mean NO2 Levels
Copollutants
Considered
Lag Structure
Reported Method/Design
Effect Estimates
EUROPE (cont'd)
Hoek et al. (2000);
reanalysis Hoek,
(2003)
The Netherlands:
entire country, four
urban areas
1986-1994
All cause; COPD; 24-h avg median: 17 ppb
pneumonia; in the Netherlands; 24 ppb
cardiovascular in the four major cities
PMio, BS, S042\
N(V, 03, S02,
CO; two-pollutant
models
1,0-6 PoissonGAM,
reanalyzed with
stringent
convergence
criteria; Poisson
GLM. Time-series
study.
Poisson GLM:
All cause:
Lag 1: 1.9% (1.2, 2.7)
Lag 0-6: 2.6% (1.2, 4.0);
withBS: 1.3% (-0.9, 3.5);
Cardiovascular (lag 0 6):
2.7% (0.7, 4.7).
COPD (lag 0-6):
10.4% (4.5, 16.7).
Pneumonia (lag 0-6):
19.9% (11. 5, 29.0).
X
OO
o
H
6
o
o
H
O
O
H
W
Hoek etal. (2001);
reanalysis Hoek,
(2003)
The Netherlands
1986-1994
Total
cardiovascular;
myocardial
infarction;
arrhythmia; heart
failure;
cerebro vascular;
thrombosis-related
24-h avg median: 17 ppb PM10,
in the Netherlands; 24 ppb CO
in the four major cities
03, S02,
Poisson GAM,
reanalyzed with
stringent
convergence
criteria; Poisson
GLM. Time-series
study.
Poisson GLM:
Total cardiovascular:
2.7% (0.7,4.7)
Myocardial infarction:
0.3% (-2.6, 3.2)
Arrhythmia:
1.7% (-6.6, 10.6)
Heart failure:
7.6% (1.4,14.2)
Cerebro vascular:
5.1% (0.9, 9.6)
Thrombosis-related:
-1.2% (-9.6, 8.1)
O
HH
H
W
-------
O
to
O
O
oo
X
OO
TABLE AX6.3-19 (cont'd). EFFECTS OF ACUTE NOX EXPOSURE ON MORTALITY. RISK ESTIMATES ARE
STANDARDIZED FOR PER 20 ppb 24-h AVG NO2 INCREMENT
Reference, Study
Location, and Period Outcome Measure
Mean NO2 Levels
Copollutants Lag Structure
Considered Reported Method/Design
Effect Estimates
EUROPE (cont'd)
Roemer and van All cause
Wijinen(2001)
Amsterdam, The
Netherlands
1987-1998
Verhoeff et al. (1996) All cause; all ages;
Amsterdam, The age 65+ yrs
Netherlands
24-h avg:
Background sites:
24 ppb
Traffic sites:
34 ppb
l-hmaxO3:
43 ug/m
Maximum 301
BS, PMio, O3, SO2, 1 , 2, 0-6 Poisson GAM with
CO default convergence
criteria (only one
smoother). Time-
series study.
PMio, O3, CO; 0,1,2 Poisson. Time-
multipollutant models series study.
Total population using
background sites: Lag 1 :
3. 8% (1.7, 5.9);
Traffic pop. using background
sites: Lag 1: 5.7% (0.6, 11.0);
Total pop. using traffic sites:
Lagl:
1.7% (0.4, 3.0)
1-h max O3 (per 100 ug/m )
All ages:
1986-1992
NO NO,!!!
LagO: 1.8% (-3.8, 7.8)
Lagl: 0.1% (-4.7, 5.1)
Lag 2: 4.9% (0.1, 10.0)
H
6
o
o
H
O
O
H
W
Fischer et al. (2003)
The Netherlands,
1986-1994
All-cause,
cardiovascular,
COPD, and
pneumonia in age
groups <45, 45-64,
65-74, 75+
24-h avg median
17 ppb
o, BS, O3, SO2,
0-6
CO
Poisson GAM with
default convergence
criteria. Time-
series study.
Cardiovascular:
Age<45: -1.3% (-13.0, 12.1):
age 45-64: -0.4% (-4.8,4.3);
age 65 74: 4.4% (0.8, 8.0); age
75 and up: 3.5% (1.4, 5.6)
O
HH
H
W
-------
O
to
O
O
oo
TABLE AX6.3-19 (cont'd). EFFECTS OF ACUTE NOX EXPOSURE ON MORTALITY. RISK ESTIMATES ARE
STANDARDIZED FOR PER 20 ppb 24-h AVG NO2 INCREMENT
X
OO
to
H
6
o
o
H
O
O
H
W
O
HH
H
W
Reference, Study
Location, and Period
EUROPE (cont'd)
Spix and Wichman
(1 996) Koln, Germany
1977-1985
Peters et al. (2000b)
NE Bavaria, Germany
1982-1994
Coal basin in Czech
Republic
1993-1994
Michelozzi et al.
(1998) Rome, Italy
1992-1995
Ponkaetal. (1998)
Helsinki, Finland
1987-1993
Saez et al. (2002)
Seven Spanish cities,
variable study periods
between 1991 and
1996.
Outcome Measure
All-cause
All cause;
respiratory;
cardiovascular;
cancer
All-cause
All cause;
cardiovascular; age
<65 yrs, age 65+ yrs
All cause;
respiratory;
cardiovascular
Copollutants
Mean NO2 Levels Considered
24-h avg 24 ppb; TSP, PM7, SO2
l-hmax38ppb
24-h avg: TSP, PM10, O3, SO2,
Czech Republic: CO
17.6 ppb
Bavaria, Germany:
13. 2 ppb
24-h avg 52 ppb PM13, SO2, O3, CO
24-h avg: TSP, PM10, O3, SO2
Median 20 ppb
24-h avg mean ranged O3, PM, SO2, CO
from 17 ppb in Huelva
to 35 ppb in Valencia.
Lag Structure
Reported Method/Design
0,1,0-1 PoissonGLM.
Time-series study.
0,1,2,3 PoissonGLM.
Time-series study.
0,1,2,3,4 Poisson GAM with
default convergence
criteria. Time-series
study.
0,1,2,3,4,5, PoissonGLM.
6, 7 Time-series study.
0-3 Poisson GAM with
default convergence
criteria. Time-series
study.
Effect Estimates
Lag 1: 0.4%
(-0.4, 1.2)
Czech Republic:
All cause:
Lag 1: 2.1%
(-1.7,6.1)
Bavaria, Germany:
All cause:
Lag 1: -0.1%
(-3.6,3.6)
Lag 2: all-yr: 1.6%
(0.4,2.9);
Cold season 0.3%
(-1.2,1.8);
Warm season: 4.2%
(1.8, -6.6)
No risk estimate presented
forN02.
PM10 and O3 were reported
to have stronger
associations.
All cause:
2.6% (1.6, 3.6);
with all other poll:
1.7% (0.0, 3.3);
Respiratory:
7.1% (-14.0, 33.5)
Cardiovascular:
4.4% (-0.2, 9.2)
-------
O
to
O
O
oo
TABLE AX6.3-19 (cont'd). EFFECTS OF ACUTE NOX EXPOSURE ON MORTALITY. RISK ESTIMATES ARE
STANDARDIZED FOR PER 20 ppb 24-h AVG NO2 INCREMENT
Reference, Study
Location, and Period
Outcome Measure Mean NO2 Levels
Copollutants
Considered
Lag Structure
Reported
Method/Design
Effect Estimates
EUROPE (cont'd)
X
OO
Garcia-Ay merich et al.
(2000)
Barcelona, Spain
1985-1989
All cause;
respiratory;
cardiovascular;
general population;
patients with COPD
Levels not reported. BS, O3, SO2
Selected best avg lag Poisson GLM.
Time-series study.
All cause:
General population:
Lag 0-3: 3.3% (0.8, 5.8)
COPD patients:
Lag 0-2: 10.9% (0.4,22.6)
Respiratory:
General population:
Lag 0-1: 3.3% (-2.3, 9.2)
COPD patients:
Lag 0-2: 12.1% (-4.3, 31.4)
Cardiovascular:
General population:
Lag 0-3: 2.4% (-0.9, 5.8)
COPD patients:
Lag 0-2: 4.3% (-13.6, 25.8)
H
6
o
o
H
O
O
H
W
Saezetal. (1999)
Barcelona, Spain
1986-1989
Asthma mortality;
age 2-45 yrs
Levels not reported. BS, O3, SO2
0-2
Poisson with GEE.
Time-series study.
RR = 4.1(0.5, 35.0)
O
HH
H
W
-------
O
to
O
O
oo
TABLE AX6.3-19 (cont'd). EFFECTS OF ACUTE NOX EXPOSURE ON MORTALITY. RISK ESTIMATES ARE
STANDARDIZED FOR PER 20 ppb 24-h AVG NO2 INCREMENT
Reference, Study
Location, and Period
Outcome Measure
Mean NO2 Levels
Copollutants
Considered
Lag Structure
Reported
Method/Design
Effect Estimates
EUROPE (cont'd)
Sunyeretal. (1996)
Barcelona, Spain
1985-1991
All cause; respiratory;
cardiovascular; all
ages; age 70+ yrs
1-hmax:
Median:
Summer: 51 ppb
Winter: 46 ppb
BS, SO2, O3
Selected best single-
day lag
Autoregressive All yr, all ages:
Poisson. Time-series All cause:
study. Lagl: 1.9% (0.8, 3.1)
Respiratory:
LagO: 1.5% (-1.9, 5.0)
Cardiovascular:
Lagl: 2.2% (0.5, 3.9)
X
OO
H
6
o
o
H
O
O
H
W
O
HH
H
W
Sunyer and Basagana
(2001)
Barcelona, Spain
1990-1995
Sunyer et al. (2002)
Barcelona, Spain
1986-1995
Mortality in a cohort
of patients with
COPD
Mean not reported
IQR8.9ppb24-havg
All cause, respiratory, 1-hmax: median
and cardiovascular 47 ppb;
mortality in a cohort 24-h avg median
of patients with severe 27 ppb
asthma
o, 03, CO
0-2
PMio, BS, S02, 03, 0-2
CO, pollen
Conditional logistic
(case-crossover)
Conditional logistic
(case-crossover)
Summer risk estimates
larger than winter risk
estimates.
7.8% (-2.0, 18.6) with
PM10:
3.9% (-12.0,22.5)
Odds Ratio:
Patients with 1 asthma
admission:
All cause:
1.10(0.80,1.51)
Cardiovascular:
1.70(0.96,2.99)
Patients with more than
1 asthma admission:
All cause:
2.14(1.10,4.14)
Cardiovascular:
1.53(0.46,5.07)
-------
O
to
O
O
oo
TABLE AX6.3-19 (cont'd). EFFECTS OF ACUTE NOX EXPOSURE ON MORTALITY. RISK ESTIMATES ARE
STANDARDIZED FOR PER 20 ppb 24-h AVG NO2 INCREMENT
Reference, Study
Location, and
Period
Outcome Measure
Mean NO2 Levels
Copollutants
Considered
Lag Structure
Reported
Method/Design
Effect Estimates
EUROPE (cont'd)
Diaz etal. (1999)
Madrid, Spain
1990-1992
All cause; respiratory;
cardiovascular
24-h avg
Levels not reported.
TSP, 03, S02, CO
1,4,10
Autoregressive linear.
Time-series study.
Only significant risk estimates were
shown. For NO2, only respiratory
mortality was significantly
(p < 0.05) associated with an excess
percent risk 8.5%.
X
OO
H
6
o
o
H
O
O
H
W
LATIN AMERICA
Borja-Aburto et al.
(1997)
Mexico City
1990-1992
Borja-Aburto et al.
(1998)
SW Mexico City
1993-1995
All cause; respiratory; 1 -h max O3: TSP, SO2, CO;
cardiovascular; all Median 155 ppb two-pollutant
ages; age <5 yrs; age models
>65yrs 8-hmax03:
Median 94 ppb
10-havgO3
(8 a.m.-6 p.m.):
Median 87 ppb
24-h avg O3:
Median 54 ppb
All cause; respiratory; 37.7 ppb, 24-h avg PM2.5, O3, SO2;
cardiovascular; other; two-pollutant
all ages; age >65 yrs models
0, 1 , 2 Poisson iteratively
weighted and filtered
least-squares method.
Time-series study.
0, 1 , 2, 3, 4, 5, Poisson GAM with
and multiday default convergence
avg criteria (only one
smoother). Time-
series study.
1-h max O3 (per 100 ppb):
All ages:
Lag 1-5:
All cause: 2.3% (-1.0, 5.6);
Cardiovascular: 2. 8% (-3.2, 9.2);
Respiratory: 4.7% (-5.1, 15.5).
O
HH
H
W
-------
O
to
O
O
oo
TABLE AX6.3-19 (cont'd). EFFECTS OF ACUTE NOX EXPOSURE ON MORTALITY. RISK ESTIMATES ARE
STANDARDIZED FOR PER 20 ppb 24-h AVG NO2 INCREMENT
X
Oi
OO
Oi
H
6
o
o
H
O
O
H
W
O
HH
H
W
Reference, Study
Location, and
Period
Outcome Measure
Copollutants Lag Structure
Mean NO2 Levels Considered Reported Method/Design
Effect Estimates
LATIN AMERICA (cont'd)
Loomisetal. (1999)
Mexico City
1993-1995
Gouveia and Fletcher
(2000b)
Sao Paulo, Brazil
1991-1993
Pereiraetal. (1998)
Sao Paulo, Brazil
1991-1992
Saldivaetal. (1994)
Sao Paulo, Brazil
1990-1991
Saldivaetal. (1995)
Sao Paulo, Brazil
1990-1991
Infant mortality
All ages (all cause);
age <5 yrs (all
cause, respiratory,
pneumonia); age
65+ yrs (all cause,
respiratory,
cardiovascular)
Intrauterine
mortality
Respiratory; age
<5yrs
All cause; age
65+ yrs
24-h avg 38 ppb PM2.5, O3 0, 1 , 2, 3, 4, 5, 3-5 Poisson GAM with
default convergence
criteria. Time-series
study.
1-hmax: 84 ppb PM10, 03, SO2, CO 0,1,2 Poisson GLM.
Time-series study.
24-h avg 82 ppb PM10, 03, SO2, CO 0-4 Poisson GLM.
Time-series study.
24-h avg NOX 127 ppb PM10, O3, SO2, CO; 0-2 OLSofrawor
multipollutant transformed data.
models Time-series study.
24-h avg NOX 127 ppb PM10, O3, SO2, CO; 0-1 OLS; Poisson with
two-pollutant GEE. Time-series
models study.
Lag 3-5:
11.4% (2.2, 21.4);
withPM25:
2. 9% (-10.2, 17.8)
All ages:
All cause:
LagO: -0.1% (-0.7, 0.4)
Age 65+:
All cause:
Lagl: 0.4% (-0.2, 1.1)
Respiratory:
Lag 2: 1.0% (-0.6, 2.5)
Cardiovascular:
Lagl: -0.5% (-0.4, 1.3)
Single-pollutant model: 5.1%
(2.8, 7.5);
With other pollutants: 4.7%
(1.6,7.9)
NOX slope estimate:
0.007197 deaths/day/ppb
(SE 0.003214), p = 0.025
NOX slope estimate:
0.0341 deaths/day/ppb
(SE 0.0105)
-------
O
to
O
O
oo
TABLE AX6.3-19 (cont'd). EFFECTS OF ACUTE NOX EXPOSURE ON MORTALITY. RISK ESTIMATES ARE
STANDARDIZED FOR PER 20 ppb 24-h AVG NO2 INCREMENT
X
OO
H
6
o
o
H
O
O
H
W
O
HH
H
W
Reference, Study
Location, and
Period
Outcome Measure
Mean NO2 Levels
Copollutants Lag Structure
Considered Reported Method/Design
Effect Estimates
LATIN AMERICA (cont'd)
Cifuentes et al.
(2000) Santiago,
Chile
1988-1966
Ostroetal. (1996)
Santiago, Chile
1989-1991
All cause
All cause
8-havg41 ppb
l-hmax56ppb
PM2 5, PM10 25, 0, 1 , 2, 3, 4, 5, Poisson GAM with
CO, SO2, O3 1-2, 1 3, 1-4, 1-5 default convergence
criteria; Poisson GLM.
Time-series study.
PMio, O3, SO2; two 1 OLS, Poisson.
pollutant models Time-series study.
GLM model, lag 1-2:
Single-pollutant: 1.7% (0.7,
2.7); with other pollutants:
1.5%
(0.3,2.7)
(per 25ppb 8-h avg)
Poisson,lag 1: -0.5%(-1.1, 0)
AUSTRALIA
Simpson et al.
(2005a,b)
Brisbane, Sydney,
Melbourne, and
Perth, Australia
1996-1999
Simpson et al. (2000)
Brisbane, Australia
1991-1996
Morgan etal. (1998b)
Sydney, Australia
1989-1993
All cause,
respiratory, and
cardiovascular in all
ages; cardiovascular
in age 65+ yrs
All cause,
respiratory, and
cardiovascular in all
ages; cardiovascular
in age 65+ yrs
All cause;
respiratory;
cardiovascular
1-h max ranged from
16 to 24 ppb
24-havg: whole yr:
12 ppb; cool season:
1 3 ppb; warm season
9 ppb
24-h avg 13 ppb; 1-h
max 26 ppb
PMio, PM2.5, BSP 0, 1, 2, 3, 0-1 Poisson GLM, GAM
(nephelometer), O3, with stringent
CO convergence criteria.
Time-series study.
PMio, PM2.5, BSP, 0, 1 , 2, 3, 0-1 Poisson, GAM with
O3, CO default convergence
criteria. Time-series
study.
BSP, O3 0-1 Poisson with GEE.
Time-series study.
Lag 0-1, GAM, all-cause,
Single pollutant: 3.4% (1.1,
5.7); with BSP: 3.1% (0.3,
5.9); cardiovascular: 4.3% (0.9,
7.8); respiratory: 1 1 .4% (3.5,
19.9)
All-cause (lag 1): 9.7% (4.7,
14.8); respiratory: 18. 8% (1.2,
39.6)
Lag 0-1, single pollutant, all-
cause: 3.0% (0.1, 6.0);
cardiovascular: 2.2% (-1.7,
6.4); respiratory: 8.6% (-0.4,
18.4)
-------
O
to
O
O
oo
TABLE AX6.3-19 (cont'd). EFFECTS OF ACUTE NOX EXPOSURE ON MORTALITY. RISK ESTIMATES ARE
STANDARDIZED FOR PER 20 ppb 24-h AVG NO2 INCREMENT
X
OO
oo
H
6
o
o
H
O
O
H
W
Reference, Study
Location, and
Period Outcome Measure
Mean NO2 Levels
Copollutants Lag Structure
Considered Reported Method/Design
Effect Estimates
AUSTRALIA (cont'd)
Simpson et al. (1 997) All cause;
Brisbane, Australia respiratory;
1987-1993 cardiovascular
24-h avg 14 ppb;
l-hmax28 ppb
PMio, TSP, O3, 0 Autoregressive
SO2, CO Poisson with GEE.
Time-series study.
Lag 0-1, single pollutant, all-
cause, all-yr: -1.0% (-5.2,
3.4); summer: -3.6%(-11.2,
4.7); winter: - 1 .2% (-4.0, 6.9)
ASIA
Kim et al. (2004b) All cause
Seoul, Korea
1995-1999
Lee et al. (1 999) All cause
Seoul and Ulsan,
Korea
1991-1995
24-h avg 33 ppb.
1-hmax O3:
Seoul:
32.4 ppb
10th%-90th%
14-55
Ulsan:
26.0 ppb
10th%-90th%
16-39
PMio, O3, SO2, CO; 1 Poisson GAM with
two-pollutant stringent convergence
models criteria (linear model);
GLM with cubic
natural spline; GLM
with B mode spline
(threshold model).
Time-series study.
TSP, SO2 0 Poisson with GEE.
Time-series study.
Risk estimates for NO2 not
reported.
1-h max O3 (per 50 ppb):
Seoul:
1.5% (0.5, 2. 5)
Ulsan:
2.0%(-11. 1,17.0)
O
HH
H
W
-------
O
to
O
O
oo
TABLE AX6.3-19 (cont'd). EFFECTS OF ACUTE NOX EXPOSURE ON MORTALITY. RISK ESTIMATES ARE
STANDARDIZED FOR PER 20 ppb 24-h AVG NO2 INCREMENT
X
OO
VO
H
6
o
o
H
O
O
H
W
O
HH
H
W
Reference, Study
Location, and Period
Outcome Measure Mean NO2 Levels
Copollutants Lag Structure
Considered Reported Method/Design
Effect Estimates
ASIA (cont'd)
Lee and Schwartz (1999)
Seoul, Korea
1991-1995
Kwonetal. (2001)
Seoul, Korea
1994-1998
Ha et al. (2003)
Seoul, Korea
1995-1999
Hong et al. (2002)
Seoul, Korea
1995-1998
Tsai et al. (2003b)
Kaohsiung, Taiwan
1994-2000
All cause 1 -h max O3:
Seoul:
32.4 ppb
10th%-90th%
14-55
Mortality in a cohort 24-h avg 32 ppb
of patients with
congestive heart
failure
All cause; respiratory; 24-h avg 33 ppb
postneonatal (1 mo to
1 yr); age 2 64 yrs;
age 65+
Acute stroke mortality 24-h avg 33 ppb
All cause; respiratory; 24-h avg 29 ppb
cardiovascular;
tropical area
TSP, SO2 0 Conditional
logistic
regression.
Case-crossover
with
bidirectional
control sampling.
PMm, O3, SO2, CO 0 Poisson GAM
with default
convergence
criteria; case-
crossover
analysis using
conditional
logistic
regression.
PMm, O3, SO2, CO 0 Poisson GAM
with default
convergence
criteria. Time-
series study.
PMm, O3, SO2, CO 2 Poisson GAM
with default
convergence
criteria. Time-
series study.
PMm, SO2, 03, CO 0-2 Conditional
logistic
regression.
Case-crossover
analysis.
1-hmax O3
(per 50 ppb):
Two controls ± 1 wk:
1.5% (-1.2, 4.2)
Four controls ± 2 wks:
2.3% (-0.1, 4.8)
Odds ratio in general population:
1.1% (-0.3, 2. 5)
Congestive heart failure cohort:
15. 8% (1.8, 31.7)
All cause for postneonates: 0.8%
(-5.7, 7.7); age 65+: 3. 8% (3. 7, 3.9)
4. 3% (1.6, 7.0)
Odds ratios:
All cause:
0.1% (-5. 9, 6.6);
Respiratory:
-1.0% (-22.2, 25. 9);
Cardiovascular:
-1.8%(- 14.0, 12.1)
-------
O
to
O
O
oo
X
H
6
o
o
H
O
O
H
W
TABLE AX6.3-19 (cont'd). EFFECTS OF ACUTE NOX EXPOSURE ON MORTALITY. RISK ESTIMATES ARE
STANDARDIZED FOR PER 20 ppb 24-h AVG NO2 INCREMENT
Reference, Study
Location, and Period
Outcome Measure
Mean NO2 Levels
Copollutants
Considered
Lag Structure
Reported
Method/Design
Effect Estimates
ASIA (cont'd)
Yang et al. (2004b)
Taipei, Taiwan
1994-1998
All cause; respiratory;
cardiovascular;
subtropical area
24-h avg 3 1 ppb
PMio, SO2, O3, CO
0-2
Conditional logistic
regression. Case-
crossover analysis.
Odds ratios:
All cause: 0.6% (-3. 9, 5.2);
Respiratory: 2. 5% (-13.1, 20.8);
Wong et al. (200 Ib)
Hong Kong
1995-1997
All cause; respiratory;
cardiovascular
24-h avg 25 ppb in
warm season; 33 ppb
in cold season
PMio, O3, SO2;
two-pollutant
models
0,1,2
Cardiovascular:
-1.1% (-9.5, 8.0)
Poisson GAM with All cause (lag 1):
default convergence
criteria. Time-series
study.
2.6% (0.9, 4.4);
Respiratory (lag 0):
6.1% (-1.8, 10.5);
Cardiovascular (lag 2):
5.2% (1.8, 8.7)
Wong et al. (2002)
Hong Kong
1995-1998
Respiratory; 24-h avg 29 ppb
cardiovascular;
COPD; pneumonia
and influenza;
ischemic heart dis.;
cerebrovascular
PMio, O3, SO2; two 0,1,2,0-1,0-2
pollutant models
Poisson GLM. Time-
series study.
Respiratory (0-1):
5.1% (1.6, 8.7);
Cardiovascular (lag 0-2):
3.1% (-0.2, 6.5)
O
HH
H
W
-------
O
to
O
O
oo
TABLE AX6.3-19 (cont'd). EFFECTS OF ACUTE NOX EXPOSURE ON MORTALITY. RISK ESTIMATES ARE
STANDARDIZED FOR PER 20 ppb 24-h AVG NO2 INCREMENT
Reference, Study
Location, and
Period
Outcome Measure
Mean NO2 Levels
Copollutants
Considered
Lag Structure
Reported
Method/Design
Effect Estimates
ASIA (cont'd)
Hedley et al. (2002)
Hong Kong
1985-1995
Intervention Jul 1990
(switch to low sulfur-
content fuel)
All cause; cardiovascular;
respiratory; neoplasms and
other causes; all ages; age
15-64 yrs; age 65+ yrs
Avg moly NO2:
Baseline: 29 ppb
1 yr after intervention:
25 ppb
2-5 yrs after
intervention: 28 ppb
SO2 (main pollutant
of interest, 45%
reduction observed
5 yrs after
intervention), PM10,
SO42~,NO2
Moly avgs
considered
without lags
Poisson regression of
moly avgs to estimate
changes in the
increase in deaths
from warm to cool
season. Annual
proportional change
in death rate before
and after the
intervention was also
examined.
Declines observed in all cause
(2.1%, p = 0.001), respiratory
(3.9%, p = 0.001), and
cardiovascular
(2.0%, p = 0.020) mortality after
the intervention.
As NO2 levels did not change
before and after the intervention,
NO2 likely did not play a role in
the decline in observed
mortality.
X
H
6
o
o
H
O
O
H
W
Yang et al. (2004b)
Taipei, Taiwan
1994-1998
All cause; respiratory; 24-h avg 31 ppb PM10, SO2, O3, CO 0-2
cardiovascular; subtropical
area
Conditional logistic
regression. Case-
crossover analysis.
Odds ratios:
All cause:
0.6% (-3.9, 5.2);
Respiratory:
2.5%(-13.1,20.8);
Cardiovascular:
-1.1% (-9.5, 8.0)
O
HH
H
W
-------
TABLE AX6.3-20. NO2 EXPOSURE AFFECTS ASTHMATICS
An intervention study (Pilotto et al., 2004) of respiratory symptoms of asthmatic children in
Australia resulted in reductions in several symptoms (difficulty in breathing during the day and
at night, chest tightness during the day and at night, and asthma attacks during the day) related
to reduction in NO2 exposure from in-class heaters. Information on other heater emissions,
such as ultrafme particles, was not reported.
Birth cohort studies in the United Sates (Belanger et al., 2006; Van Strein et al., 2004) and
Europe (Brauer et al., 2007) relate NO2 concentrations to increased respiratory symptoms,
infections, and asthma in the very young.
In England, Chauhan et al. (2003) and Linaker et al. (2000) studied personal NO2 exposure and
found NO2 exposure in the week before an upper respiratory infection was associated with
either increased severity of lower-respiratory-tract symptoms, or reduction of PEF for all virus
types together, and for two of the common viruses, RSV and a picorna virus, individually.
Nitschke et al. (2006) reported difficulty breathing and chest tightness associated with 10 ppb
increases in NO2 measured in school classrooms. Lung function tests were performed at the
beginning and at the end of the study period, and the authors observed personal NO2 exposures
related in a dose-response manner for reported symptoms in asthmatics.
United States multeity studies of ambient NO2 exposure examined respiratory symptoms in
asthmatics (Mortimer et al., 2002; Schildcrout et al., 2006). In the NCICAS (Mortimer et al.,
2002) the greatest effect was seen for morning symptoms (cough, wheeze, shortness of breath)
for a 6-day-morning average. In multi-pollutant models, the NO2 effect was attenuated though
remained positive, for Os, SO2, and combined coarse and fine particulate matter (PMio). In the
CAMP study (Schildcrout et al., 2006), the strongest association between NO2 and increased
risk of cough and increased use of rescue medication was found for a 2-day lag, which was not
attenuated, in two-pollutant models for CO, PMio, or SO2. Single city panel studies in the
Los Angeles area are supportive of these associations for asthmatics (Ostro et al., 2001; Delfino
et al., 2002, 2003a,b). Segala et al. (1998) and Just et al. (2002), in Paris both found positive
relationships to NO2 exposure and symptoms in asthmatics.
Few studies of the impact of NO2 on respiratory symptoms of adult asthmatics are available.
These find positive associations for NO2 exposure and respiratory symptoms in European
studies (Hiltermann et al., 1998; Von Klot et al., 2002; and Forsberg et al., 1998).
The associations between ambient concentrations of NO2 and ER visits for asthma in the United
States are positive (Jaffe et al., 2003; Peel et al., 2005; Tolbert et al., 2000). Studies conducted
outside the United States (Castlellsague et al., 1995; Sunyer et al., 1997; Atkinson et al.,
1999a,b; Tenias et al., 1998; Erbas et al., 2005) found similar results. A concentration response
for NO2 and asthma ER visits is indicated in these studies (Jaffe et al., 2003; Tenias et al., 1998;
Castellsague et al., 1995).
March 2008 AX6-192 DRAFT-DO NOT QUOTE OR CITE
-------
TABLE AX6.3-20 (cont'd). NO2 EXPOSURE AFFECTS ASTHMATICS
In relation to long-term exposure, Moseler (1994) examined a cohort in Germany and reported
decrements in lung function parameters related to NC>2 exposure measures in a group of
physician-diagnosed asthmatic children.
The relationship between long-term NC>2 exposure and asthma prevalence and incidence has
been examined in several studies. In the CHS, Gauderman et al. (2005) report a positive
relationship. A marginally significant positive relationship was seen for NC>2 exposure with
new onset asthma, while significant associations for PM were observed (Islam et al. 2007). In
a separate cohort in the Netherlands, Brauer et al. (2007) provide confirming evidence for
this relationship.
Acute mortality related to asthma was examined in Barcelona, Spain (Saez et al., 1999; Sunyer
et al., 2002). In the study by Sunyer et al. (2002), severe asthmatics with more than one
asthma emergency visit were found to have the strongest mortality associations with NC>2.
March 2008 AX6-193 DRAFT-DO NOT QUOTE OR CITE
-------
1 AX6.4 REFERENCES
2 Ackermann-Liebrich, U.; Leuenberger, P.; Schwartz, J.; Schindler, C.; Monn, C.; Bolognini, B.;
3 Bongard, J. P.; Brandli, O.; Domenighetti, G.; Elsasser, S.; Grize, L.; Karrer, W.; Keller,
4 R.; Keller-Wossidlo, H.; Kiinzli, N.; Martin, B. W.; Medici, T. C.; Perruchoud, A. P.;
5 Schoni, M. H.; Tschopp, J. M.; Villiger, B.; Wuthrich, B.; Zellweger, J. P.; Zemp, E.
6 (1997) Lung function and long term exposure to air pollutants in Switzerland. Am. J.
7 Respir. Crit. Care Med. 155: 122-129.
8 Andersen, Z. J.; Wahlin, P.; Raaschou-Nielsen, O.; Scheike, T.; Loft, S. (2007a) Ambient
9 particle source apportionment and daily hospital admissions among children and elderly
10 in Copenhagen. J. Exposure Sci. Environ. Epidemiol. 17: 625-636.
11 Andersen, Z. J.; Wahlin, P.; Raaschou-Nielsen, O.; Ketzel, M.; Scheike, T.; Loft, S. (2007b)
12 Size distribution and total number concentration of ultrafme and accumulation mode
13 particles and hospital admissions in children and the elderly in Copenhagen, Denmark.
14 Occup. Environ. Med.: 10.1136/oem.2007.033290.
15 Anderson, H. R.; Ponce de Leon, A.; Bland, J. M.; Bower, J. S.; Strachan, D. P. (1996) Air
16 pollution and daily mortality in London: 1987-92. Br. Med. J. 312: 665-669.
17 Anderson, H. R.; Spix, C.; Medina, S.; Schouten, J. P.; Castellsague, J.; Rossi, G.; Zmirou, D.;
18 Touloumi, G.; Wojtyniak, B.; Ponka, A.; Bacharova, L.; Schwartz, J.; Katsouyanni, K.
19 (1997) Air pollution and daily admissions for chronic obstructive pulmonary disease in 6
20 European cities: results from the APHEA project. Eur. Respir. J. 10: 1064-1071.
21 Anderson, H. R.; Ponce de Leon, A.; Bland, J. M.; Bower, J. S.; Emberlin, J.; Strachen, D. P.
22 (1998) Air pollution, pollens, and daily admissions for asthma in London 1987-92.
23 Thorax 53: 842-848.
24 Anderson, H. R.; Bremner, S. A.; Atkinson, R. W.; Harrison, R. M.; Walters, S. (2001)
25 Particulate matter and daily mortality and hospital admissions in the west midlands
26 conurbation of the United Kingdom: associations with fine and coarse particles, black
27 smoke and sulphate. Occup. Environ. Med. 58: 504-510.
28 Arbex, M. A.; Martins, L. C.; De Oliveira, R. C.; Pereira, L. A.; Arbex, F. F.; Can9ado, J. E.;
29 Saldiva, P. H.; Braga, A. L. (2007) Air pollution from biomass burning and asthma
30 hospital admissions in a sugar cane plantation area in Brazil. J. Epidemiol. Community
31 Health 61: 395-400.
32 Atkinson, R. W.; Bremner, S. A.; Anderson, H. R.; Strachan, D. P.; Bland, J. M.; Ponce de Leon,
33 A. (1999a) Short-term associations between emergency hospital admissions for
34 respiratory and cardiovascular disease and outdoor air pollution in London. Arch.
35 Environ. Health 54: 398-411.
36 Atkinson, R. W.; Anderson, H. R.; Strachan, D. P.; Bland, J. M.; Bremner, S. A.; Ponce de Leon,
37 A. (1999b) Short-term associations between outdoor air pollution and visits to accident
38 and emergency departments in London for respiratory complaints. Eur. Respir. J. 13:
39 257-265.
40
41
March 2008 AX6-194 DRAFT-DO NOT QUOTE OR CITE
-------
1 Atkinson, R. W.; Anderson, H. R.; Sunyer, J.; Ayres, J.; Baccini, M.; Vonk, J. M.; Boumghar,
2 A.; Forastiere, F.; Forsberg, B.; Touloumi, G.; Schwartz, J.; Katsouyanni, K. (2001)
3 Acute effects of particulate air pollution on respiratory admissions: results from APHEA
4 2 project. Am. J. Respir. Crit. Care Med. 164: 1860-1866.
5 Baccarelli, A.; Zanobetti, A.; Martinelli, I.; Grille, P.; Hou, L.; Giacomini, S.; Bonzini, M.;
6 Lanzani, G.; Mannucci, P. M.; Bertazzi, P. A.; Schwartz, J. (2007) Effects of exposure to
7 air pollution on blood coagulation. J. Thromb. Haemostasis 5: 252-260.
8 Ballester, F.; Tenias, J. M.; Perez-Hoyos, S. (2001) Air pollution and emergency hospital
9 admissions for cardiovascular diseases in Valencia, Spain. J. Epidemiol. Community
10 Health 55: 57-65.
11 Ballester, F.; Rodriguez, P.; Iniguez, C.; Saez, M.; Daponte, A.; Galan, I; Taracido, M.; Arribas,
12 F.; Bellido, J.; Cirarda, F. B.; Canada, A.; Guillen, J. J.; Guillen-Grima, F.; Lopez, E.;
13 Perez-Hoyos, S.; Lertxundi, A.; Toro, S. (2006) Air pollution and cardiovascular
14 admisisons association in Spain: results within the EMECAS project. J. Epidemiol.
15 Community Health 60: 328-336.
16 Barnett, A. G.; Williams, G. M.; Schwartz, J.; Neller, A. H.; Best, T. L.; Petroeschevsky, A. L.;
17 Simpson, R. W. (2005) Air pollution and child respiratory health: a case-crossover study
18 in Australia and New Zealand. Am. J. Respir. Crit. Care Med. 171: 1272-1278.
19 Barnett, A. G.; Williams, G. M.; Schwartz, J.; Best, T. L.; Neller, A. H.; Petroeschevsky, A. L.;
20 Simpson, R. W. (2006) The effects of air pollution on hospitalization for cardiovascular
21 disease in elderly people in Australian and New Zealand cities. Environ. Health Perspect.
22 114:1018-1023.
23 Bates, D. V.; Baker-Anderson, M.; Sizto, R. (1990) Asthma attack periodicity: a study of
24 hospital emergency visits in Vancouver. Environ. Res. 51: 51-70.
25 Bedeschi, E.; Campari, C.; Candela, S.; Collini, G.; Caranci, N.; Frasca, G.; Galassi, C.;
26 Francesca, G.; Vigotti, M. A. (2007) Urban air pollution and respiratory emergency visits
27 at pediatric unit, Reggio Emilia, Italy. J. Toxicol. Environ. Health Part A 70: 261-265.
28 Belanger, K.; Gent, J. F.; Triche, E. W.; Bracken, M. B.; Leaderer, B. P. (2006) Association of
29 indoor nitrogen dioxide exposure with respiratory symptoms in children with asthma.
30 Am. J. Respir. Crit. Care Med. 173: 297-303.
31 Bell, M. L.; Ebisu, K.; Belanger, K. (2007) Ambient air pollution and low birth weight in
32 Connecticut and Massachusetts. Environ. Health Perspect. 115:1118-1125.
33 Biggeri, A.; Baccini, M.; Bellini, P.; Terracini, B. (2005) Meta-analysis of the Italian studies of
34 short-term effects of air pollution (MISA), 1990-1999. Int. J. Occup. Environ. Health 11:
35 107-122.
36 Bobak, M. (2000) Outdoor air pollution, low birth weight, and prematurity. Environ. Health
37 Perspect. 108: 173-176.
38 Borja-Aburto, V. H.; Loomis, D. P.; Bangdiwala, S. L; Shy, C. M.; Rascon-Pacheco, R. A.
39 (1997) Ozone, suspended particulates, and daily mortality in Mexico City. Am. J.
40 Epidemiol. 145: 258-268.
March 2008 AX6-195 DRAFT-DO NOT QUOTE OR CITE
-------
1 Borja-Aburto, V. H.; Castillejos, M.; Gold, D. R.; Bierzwinski, S.; Loomis, D. (1998) Mortality
2 and ambient fine particles in southwest Mexico City, 1993-1995. Environ. Health
3 Perspect. 106: 849-855.
4 Boutin-Forzano, S.; Adel, N.; Gratecos, L.; Jullian, H.; Gamier, J. M.; Ramadour, M.;
5 Lanteaume, A.; Hamon, M.; Lafay, V.; Charpin, D. (2004) Visits to the emergency room
6 for asthma attacks and short-term variations in air pollution. A case-crossover study.
7 Respiration 71: 134-137.
8 Braga, A. L. F.; Concei9ao, G. M. S.; Pereira, L. A. A.; Kishi, H. S.; Pereira, J. C. R.; Andrade,
9 M. F.; Gon9alves, F. L. T.; Saldiva, P. H. N.; Latorre, M. R. D. O. (1999) Air pollution
10 and pediatric respiratory hospital admissions in Sao Paulo, Brazil. J. Environ. Med. 1: 95-
11 102.
12 Braga, A. L. F.; Saldiva, P. H. N.; Pereira, L. A. A.; Menezes, J. J. C.; Concei9ao, G. M. C.; Lin,
13 C. A.; Zanobetti, A.; Schwartz, J.; Dockery, D. W. (2001) Health effects of air pollution
14 exposure on children and adolescents in Sao Paulo, Brazil. Pediatr. Pulmonol. 31: 106-
15 113.
16 Brauer, M.; Hoek, G.; Van Vliet, P.; Meliefste, K.; Fischer, P. H.; Wijga, A.; Koopman, L. P.;
17 Neijens, H. J.; Gerritsen, J.; Kerkhof, M.; Heinrich, J.; Bellander, T.; Brunekreef, B.
18 (2002) Air pollution from traffic and the development of respiratory infections and
19 asthmatic and allergic symptoms in children. Am. J. Respir. Crit. Care Med. 166: 1092-
20 1098.
21 Brauer, M.; Hoek, G.; Smit, H. A.; De Jongste, J. C.; Gerritsen, J.; Postma, D. S.; Kerkhof, M.;
22 Brunekreef, B. (2007) Air pollution and development of asthma, allergy and infections in
23 a birth cohort. Eur. Respir. J. 29: 879-888.
24 Bremner, S. A.; Anderson, H. R.; Atkinson, R. W.; McMichael, A. J.; Strachan, D. P.; Bland, J.
25 M.; Bower, J. S. (1999) Short term associations between outdoor air pollution and
26 mortality in London 1992-4. Occup. Environ. Med. 56: 237-244.
27 Buchdahl, R.; Parker, A.; Stebbings, T.; Babiker, A. (1996) Association between air pollution
28 and acute childhood wheezy episodes: prospective observational study. Br. Med. J. 312:
29 661-664.
30 Burnett, R. T.; Brook, J. R.; Yung, W. T.; Dales, R. E.; Krewski, D. (1997a) Association
31 between ozone and hospitalization for respiratory diseases in 16 Canadian cities. Environ.
32 Res. 72:24-31.
33 Burnett, R. T.; Cakmak, S.; Brook, J. R.; Krewski, D. (1997b) The role of particulate size and
34 chemistry in the association between summertime ambient air pollution and
35 hospitalization for cardiorespiratory diseases. Environ. Health Perspect. 105: 614-620.
36 Burnett, R. T.; Cakmak, S.; Brook, J. R. (1998a) The effect of the urban ambient air pollution
37 mix on daily mortality rates in 11 Canadian cities. Can. J. Public Health 89: 152-156.
38 Burnett, R. T.; Cakmak, S.; Raizenne, M. E.; Stieb, D.; Vincent, R.; Krewski, D.; Brook, J. R.;
39 Philips, O.; Ozkaynak, H. (1998b) The association between ambient carbon monoxide
40 levels and daily mortality in Toronto, Canada. J. Air Waste Manage. Assoc. 48: 689-700.
March 2008 AX6-196 DRAFT-DO NOT QUOTE OR CITE
-------
1 Burnett, R. T.; Smith-Doiron, M.; Stieb, D.; Cakmak, S.; Brook, J. R. (1999) Effects of
2 particulate and gaseous air pollution on cardiorespiratory hospitalizations. Arch. Environ.
3 Health 54: 130-139.
4 Burnett, R. T.; Brook, J.; Dann, T.; Delocla, C.; Philips, O.; Cakmak, S.; Vincent, R.; Goldberg,
5 M. S.; Krewski, D. (2000) Association between particulate- and gas-phase components of
6 urban air pollution and daily mortality in eight Canadian cities. In: Grant, L. D., ed.
7 PM2000: particulate matter and health. Inhalation Toxicol. 12(suppl. 4): 15-39.
8 Burnett, R. T.; Smith-Doiron, M.; Stieb, D.; Raizenne, M. E.; Brook, J. R.; Dales, R. E.; Leech,
9 J. A.; Cakmak, S.; Krewski, D. (2001) Association between ozone and hospitalization for
10 acute respiratory diseases in children less than 2 years of age. Am. J. Epidemiol. 153:
11 444-452.
12 Burnett, R. T.; Goldberg, M. S. (2003) Size-fractionated particulate mass and daily mortality in
13 eight Canadian cities. In: Revised analyses of time-series studies of air pollution and
14 health. Special report. Boston, MA: Health Effects Institute; pp. 85-89. Available:
15 http://www.healtheffects.org/news.htm [16 May, 2003].
16 Burnett, R. T.; Stieb, D.; Brook, J. R.; Cakmak, S.; Dales, R.; Raizenne, M.; Vincent, R.; Dann,
17 T. (2004) Associations between short-term changes in nitrogen dioxide and mortality in
18 Canadian cities. Arch. Environ. Health 59: 228-236.
19 Cassino, C.; Ito, K.; Bader, I; Ciotoli, C.; Thurston, G.; Reibman, J. (1999) Cigarette smoking
20 and ozone-associated emergency department use for asthma by adults in New York City.
21 Am. J. Respir. Crit. Care Med. 159: 1773-1779.
22 Castellsague, J.; Sunyer, J.; Saez, M.; Anto, J. M. (1995) Short-term association between air
23 pollution and emergency room visits for asthma in Barcelona. Thorax 50: 1051-1056.
24 Chan, C.-C.; Chuang, K.-J.; Su, T.-C.; Lin, L.-Y. (2005) Association between nitrogen dioxide
25 and heart rate variability in a susceptible population. Eur. J. Cardiovasc. Prev. Rehabil.
26 12: 580-586.
27 Chan, C.-C.; Chuang, K.-J.; Chien, L.-C.; Chen, W.-J.; Chang, W.-T. (2006) Urban air pollution
28 and emergency admissions for cerebrovascular diseases in Taipei, Taiwan. Eur. Heart J.
29 27: 1238-1244.
30 Chang, C.-C.; Tsai, S.-S.; Ho, S.-C.; Yang, C.-Y. (2005) Air pollution and hospital admissions
31 for cardiovascular disease in Taipei, Taiwan. Environ. Res. 98: 114-119.
32 Chardon, B.; Lefranc, A.; Granados, D.; Gremy, I. (2007) Air pollution and doctors' house calls
33 for respiratory diseases in the greater Paris area (2000-3). Occup. Environ. Med. 64: 320-
34 324.
35 Chauhan, A. J.; Inskip, H. M.; Linaker, C. H.; Smith, S.; Schreiber, J. ; Johnston, S. L.; Holgate,
36 S. T. (2003) Personal exposure to nitrogen dioxide (NC^) and the severity of virus-
37 induced asthma in children. Lancet 361: 1939-1944.
38 Chen, P.-C.; Lai, Y.-M.; Chan, C.-C.; Hwang, J.-S.; Yang, C.-Y.; Wang, J.-D. (1999) Short-term
39 effect of ozone on the pulmonary function of children in primary school. Environ. Health
40 Perspect. 107: 921-925.
March 2008 AX6-197 DRAFT-DO NOT QUOTE OR CITE
-------
1 Chen, Y.; Yang, Q.; Krewski, D.; Burnett, R. T.; Shi, Y.; McGrail, K. M. (2005) The effect of
2 coarse ambient participate matter on first, second, and overall hospital admissions for
3 respiratory disease among the elderly. Inhalation Toxicol. 17: 649-655.
4 Chen, C.-H.; Xirasagar, S.; Lin, H.-C. (2006) Seasonality in adult asthma admissions, air
5 pollutant levels, and climate: a population-based study. J. Asthma 43: 287-292.
6 Chew, F. T.; Goh, D. Y. T.; Ooi, B. C.; Saharom, R.; Hui, J. K. S.; Lee, B. W. (1999)
7 Association of ambient air-pollution levels with acute asthma exacerbation among
8 children in Singapore. Allergy (Copenhagen) 54: 320-329.
9 Chock, D. P.; Winkler, S. L.; Chen, C. (2000) A study of the association between daily mortality
10 and ambient air pollutant concentrations in Pittsburgh, Pennsylvania. J. Air Waste
11 Manage. Assoc. 50: 1481-1500.
12 Cifuentes, L. A.; Vega, J.; Kopfer, K.; Lave, L. B. (2000) Effect of the fine fraction of parti culate
13 matter versus the coarse mass and other pollutants on daily mortality in Santiago, Chile.
14 J. Air Waste Manage. Assoc. 50: 1287-1298.
15 D'Ippoliti, D.; Forastiere, F.; Ancona, C.; Agabiti, N.; Fusco, D.; Michelozzi, P.; Perucci, C. A.
16 (2003) Air pollution and myocardial infarction in Rome: a case-crossover analysis.
17 Epidemiology 14: 528-535.
18 Dab, W.; Medina, S.; Quenel, P.; Le Moullec, Y.; Le Tertre, A.; Thelot, B.; Monteil, C.;
19 Lameloise, P.; Pirard, P.; Momas, L; Ferry, R.; Festy, B. (1996) Short term respiratory
20 health effects of ambient air pollution: results of the APFIEA project in Paris. In: St
21 Leger, S., ed. The APHEA project. Short term effects of air pollution on health: a
22 European approach using epidemiological time series data. J. Epidemiol. Commun.
23 Health 50(suppl. 1): S42-S46.
24 Dales, R. E.; Cakmak, S.; Doiron, M. S. (2006) Gaseous air pollutants and hospitalization for
25 respiratory disease in the neonatal period. Environ. Health Perspect. 114: 1751-1754.
26 De Leon, S. F.; Thurston, G. D.; Ito, K. (2003) Contribution of respiratory disease to
27 nonrespiratory mortality associations with air pollution. Am. J. Respir. Crit. Care Med.
28 167:1117-1123.
29 Delfmo, R. J.; Zeiger, R. S.; Seltzer, J. M.; Street, D. H.; McLaren, C. E. (2002) Association of
30 asthma symptoms with peak parti culate air pollution and effect modification by anti-
31 inflammatory medication use. Environ. Health Perspect. 110: A607-A617.
32 Delfmo, R. J.; Gone, H.; Linn, W. S.; Pellizzari, E. D.; Hu, Y. (2003a) Asthma symptoms in
33 Hispanic children and daily ambient exposures to toxic and criteria air pollutants.
34 Environ. Health Perspect. Ill: 647-656.
35 Delfmo, R. J.; Gong, H.; Linn, W. S.; Hu, Y.; Pellizzari, E. D. (2003b) Respiratory symptoms
36 and peak expiratory flow in children with asthma in relation to volatile organic
37 compounds in exhaled breath and ambient air. J. Exposure Anal. Environ. Epidemiol. 13:
38 348-363.
39 Delfmo, R. J.; Staimer, N.; Gillen, D.; Tjoa, T.; Sioutas, C.; Fung, K.; George, S. C.; Kleinman,
40 M. T. (2006) Personal and ambient air pollution is associated with increased exhaled
41 nitric oxide in children with asthma. Environ. Health Perspect. 114: 1736-1743.
March 2008 AX6-198 DRAFT-DO NOT QUOTE OR CITE
-------
1 Dewanji, A.; Moolgavkar, S. H. (2000) A Poisson process approach for recurrent event data with
2 environmental covariates. Environmetrics 11: 665-673.
3 Dewanji, A.; Moolgavkar, S. H. (2002) Choice of stratification in Poisson process analysis of
4 recurrent event data with environmental covariates. Stat. Med. 21: 3383-3393.
5 Diaz, I; Garcia, R.; Ribera, P.; Alberdi, J. C.; Hernandez, E.; Pajares, M. S.; Otero, A. (1999)
6 Modeling of air pollution and its relationship with mortality and morbidity in Madrid,
7 Spain. Int. Arch. Occup. Environ. Health 72: 366-376.
8 Dockery, D. W.; Schwartz, J.; Spengler, J. D. (1992) Air pollution and daily mortality:
9 associations with particulates and acid aerosols. Environ. Res. 59: 362-373.
10 Dockery, D. W.; Luttmann-Gibson, H.; Rich, D. Q.; Link, M. S.; Mittleman, M. A.; Gold, D. R.;
11 Koutrakis, P.; Schwartz, J. D.; Verrier, R. L. (2005) Association of air pollution with
12 increased incidence of ventricular tachyarrhythmias recorded by implanted cardioverter
13 defibrillators. Environ. Health Perspect. 113: 670-674.
14 Dominici, F.; McDermott, A.; Zeger, S. L.; Samet, J. M. (2002) On the use of generalized
15 additive models in time-series studies of air pollution and health. Am. J. Epidemiol. 156:
16 193-203.
17 Dominici, F.; McDermott, A.; Daniels, M.; Zeger, S. L.; Samet, J. M. (2003) Mortality among
18 residents of 90 cities. In: Revised analyses of time-series studies of air pollution and
19 health. Special report. Boston, MA: Health Effects Institute; pp. 9-24. Available:
20 http://www.healtheffects.org/Pubs/TimeSeries.pdf [12 May, 2004].
21 Drechsler-Parks, D. M. (1995) Cardiac output effects of Os and NC>2 exposure in healthy older
22 adults. Toxicol. Ind. Health 11: 99-109.
23 Erbas, B.; Kelly, A.-M.; Physick, B.; Code, C.; Edwards, M. (2005) Air pollution and childhood
24 asthma emergency hospital admissions: estimating intra-city regional variations. Int. J.
25 Environ. Health Res. 15: 11-20.
26 Fairley, D. (1999) Daily mortality and air pollution in Santa Clara County, California: 1989-
27 1996. Environ. Health Perspect. 107: 637-641.
28 Fairley, D. (2003) Mortality and air pollution for Santa Clara County, California, 1989-1996. In:
29 Revised analyses of time-series studies of air pollution and health. Special report. Boston,
30 MA: Health Effects Institute; pp. 97-106. Available:
31 http://pubs.healtheffects.org/view.php?id=4 [12 February, 2008].
32 Farchi, S.; Forastiere, F.; Cesaroni, G.; Perucci, C. A. (2006) Environmental exposures and
33 hospitalisation for respiratory conditions in children: a five year follow up study in Rome,
34 Italy. Occup. Environ. Med. 63: 573-576.
35 Farhat, S. C. L.; Paulo, R. L. P.; Shimoda, T. M.; Conceicao, G. M. S.; Lin, C. A.; Braga, A. L.
36 F.; Warm, M. P. N.; Saldiva, P. H. N. (2005) Effect of air pollution on pediatric
37 respiratory emergency room visits and hospital admissions. Braz. J. Med. Biol. Res. 38:
38 227-235.
39 Fischer, P.; Hoek, G.; Brunekreef, B.; Verhoeff, A.; van Wijnen, J. (2003) Air pollution and
40 mortality in the Netherlands: are the elderly more at risk? Eur. Respir. J. 21(suppl. 40):
41 34S-38S.
March 2008 AX6-199 DRAFT-DO NOT QUOTE OR CITE
-------
1 Folinsbee, L. 1; Horvath, S. M.; Bedi, J. F.; Delehunt, J. C. (1978) Effect of 0.62 ppm NO2 on
2 cardiopulmonary function in young male nonsmokers. Environ. Res. 15: 199-205.
3 Forsberg, B.; Stjernberg, N.; Linne, R.; Segerstedt, B.; Wall, S. (1998) Daily air pollution levels
4 and acute asthma in southern Sweden. Eur. Respir. J. 12: 900-905.
5 Frampton, M. W.; Boscia, J.; Roberts, N. J., Jr.; Azadniv, M.; Torres, A.; Cox, C.; Morrow, P.
6 E.; Nichols, J.; Chalupa, D.; Frasier, L. M.; Gibb, F. R.; Speers, D. M.; Tsai, Y.; Utell, M.
7 J. (2002) Nitrogen dioxide exposure: effects on airway and blood cells. Am. J. Physiol.
8 282:L155-L165.
9 Fung, K. Y.; Luginaah, L; Gorey, K. M.; Webster, G. (2005) Air pollution and daily hospital
10 admissions for cardiovascular diseases in Windsor, Ontario. Can. J. Public Health 96: 29-
11 33.
12 Fung, K. Y.; Khan, S.; Krewski, D.; Chen, Y. (2006) Association between air pollution and
13 multiple respiratory hospitalizations among the elderly in Vancouver, Canada. Inhalation
14 Toxicol. 18: 1005-1011.
15 Fung, K. Y.; Luginaah, I. N.; Gorey, K. M. (2007) Impact of air pollution on hospital admissions
16 in Southwestern Ontario, Canada: generating hypotheses in sentinel high-exposure
17 places. Environ. Health: Global Access Sci. Source 6: 18. Available:
18 http://www.ehjournal.net/content/6/l/18 [20 February, 2008].
19 Fusco, D.; Forastiere, F.; Michelozzi, P.; Spadea, T.; Ostro, B.; Area, M.; Perucci, C. A. (2001)
20 Air pollution and hospital admissions for respiratory conditions in Rome, Italy. Eur.
21 Respir. J. 17: 1143-1150.
22 Galan, L; Tobias, A.; Banegas, J. R.; Aranguez, E. (2003) Short-term effects of air pollution on
23 daily asthma emergency room admissions. Eur. Respir. J. 22: 802-808.
24 Gamble, J. L. (1998) Effects of ambient air pollution on daily mortality: a time series analysis of
25 Dallas, Texas, 1990-1994. Presented at: 91st annual meeting and exhibition of the Air &
26 Waste Management Association; June; San Diego, CA. Pittsburgh, PA: Air & Waste
27 Management Association; paper no. 98-MP26.03.
28 Garcia-Aymerich, J.; Tobias, A.; Anto, J. M.; Sunyer, J. (2000) Air pollution and mortality in a
29 cohort of patients with chronic obstructive pulmonary disease: a time series analysis. J.
30 Epidemiol. Community Health 54: 73-74.
31 Garrett, M. H.; Hooper, M. A.; Hooper, B. M.; Abramson, M. J. (1998) Respiratory symptoms in
32 children and indoor exposure to nitrogen dioxide and gas stoves. Am. J. Respir. Crit.
33 Care Med. 158: 891-895.
34 Garrett, M. H.; Hooper, M. A.; Hooper, B. M. (1999) Nitrogen dioxide in Australian homes:
35 levels and sources. J. Air Waste Manage. Assoc. 49: 76-81.
36 Garty, B. Z.; Kosman, E.; Ganor, E.; Berger, V.; Garty, L.; Wietzen, T.; Waisman, Y.; Mimouni,
37 M.; Waisel, Y. (1998) Emergency room visits of asthmatic children, relation to air
38 pollution, weather, and airborne allergens. Ann. Allergy Asthma Immunol. 81: 563-570.
39 Gauderman, W. J.; Avol, E.; Gilliland, F.; Vora, H.; Thomas, D.; Berhane, K.; McConnell, R.;
40 Kuenzli, N.; Lurmann, F.; Rappaport, E.; Margolis, H.; Bates, D.; Peters, J. (2004) The
March 2008 AX6-200 DRAFT-DO NOT QUOTE OR CITE
-------
1 effect of air pollution on lung development from 10 to 18 years of age. N. Engl. J. Med.
2 351: 1057-1067.
3 Gauderman, W. J.; Avol, E.; Lurmann, F.; Kuenzli, N.; Gilliland, F.; Peters, J.; McConnell, R.
4 (2005) Childhood asthma and exposure to traffic and nitrogen dioxide. Epidemiology 16:
5 737-743.
6 Gilbert, N. L.; Goldberg, M. S.; Beckerman, B.; Brook, J. R.; Jerrett, M. (2005) Assessing spatial
7 variability of ambient nitrogen dioxide in Montreal, Canada, with a land-use regression
8 model. J. Air Waste Manage. Assoc. 55: 1059-1063.
9 Goldberg, M. S.; Burnett, R. T.; Valois, M.-F.; Flegel, K.; Bailar, J. C., Ill; Brook, J.; Vincent,
10 R.; Radon, K. (2003) Associations between ambient air pollution and daily mortality
11 among persons with congestive heart failure. Environ. Res. 91: 8-20.
12 Gouveia, N.; Fletcher, T. (2000a) Respiratory diseases in children and outdoor air pollution in
13 Sao Paulo, Brazil: a time series analysis. Occup. Environ. Med. 57: 477-483.
14 Gouveia, N.; Fletcher, T. (2000b) Time series analysis of air pollution and mortality: effects by
15 cause, age and socioeconomic status. J. Epidemiol. Community Health 54: 750-755.
16 Gouveia, N.; Bremner, S. A.; Novaes, H. M. D. (2004) Association between ambient air
17 pollution and birth weight in Sao Paulo, Brazil. J. Epidemiol. Community Health 58: 11-
18 17.
19 Gwynn, R. C.; Burnett, R. T.; Thurston, G. D. (2000) A time-series analysis of acidic particulate
20 matter and daily mortality and morbidity in the Buffalo, New York, region. Environ.
21 Health Perspect. 108: 125-133.
22 Ha, E.-H.; Lee, J.-T.; Kim, H.; Hong, Y.-C.; Lee, B.-E.; Park, H.-S.; Christiani, D. C. (2003)
23 Infant susceptibility of mortality to air pollution in Seoul, South Korea. Pediatrics 111:
24 284-290.
25 Hagen, J. A.; Nafstad, P.; Skrondal, A.; Bj0rkly, S.; Magnus, P. (2000) Associations between
26 outdoor air pollutants and hospitalization for respiratory diseases. Epidemiology 11: 136-
27 140.
28 Hajat, S.; Haines, A.; Goubet, S. A.; Atkinson, R. W.; Anderson, H. R. (1999) Association of air
29 pollution with daily GP consultations for asthma and other lower respiratory conditions in
30 London. Thorax 54: 597-605.
31 Hajat, S.; Haines, A.; Atkinson, R. W.; Bremner, S. A.; Anderson, H. R.; Emberlin, J. (2001)
32 Association between air pollution and daily consultations with general practitioners for
33 allergic rhinitis in London, United Kingdom. Am. J. Epidemiol. 153: 704-714.
34 Hajat, S.; Anderson, H. R.; Atkinson, R. W.; Haines, A. (2002) Effects of air pollution on
35 general practitioner consultations for upper respiratory diseases in London. Occup.
36 Environ. Med. 59: 294-299.
37 Hansen, C.; Neller, A.; Williams, G.; Simpson, R. (2006) Maternal exposure to low levels of
38 ambient air pollution and preterm birth in Brisbane, Australia. BJOG 113: 935-941.
March 2008 AX6-201 DRAFT-DO NOT QUOTE OR CITE
-------
1 Hedley, A. 1; Wong, C.-M.; Thach, T. Q.; Ma, S.; Lam, T.-H.; Anderson, H. R. (2002)
2 Cardiorespiratory and all-cause mortality after restrictions on sulphur content of fuel in
3 Hong Kong: an intervention study. Lancet 360: 1646-1652.
4 Henneberger, A.; Zareba, W.; Ibald-Mulli, A.; Riickerl, R.; Cyrys, J.; Couderc, J.-P.; Mykins, B.;
5 Woelke, G.; Wichmann, H.-E.; Peters, A. (2005) Repolarization changes induced by air
6 pollution in ischemic heart disease patients. Environ. Health Perspect. 113: 440-446.
7 Hernandez-Gardufio, E.; Perez-Neria, J.; Paccagnella, A. M.; Pifia-Garcia, M.; Munguia-Castro,
8 M.; Catalan-Vazquez, M.; Rojas-Ramos, M. (1997) Air pollution and respiratory health
9 in Mexico City. J. Occup. Environ. Med. 39: 299-307.
10 Hiltermann, T. J. N.; Stolk, J.; Van Der Zee, S. C.; Brunekreef, B.; De Bruijne, C. R.; Fischer, P.
11 H.; Ameling, C. B.; Sterk, P. J.; Hiemstra, P. S.; Van Bree, L. (1998) Asthma severity
12 and susceptibility to air pollution. Eur. Respir. J. 11: 686-693.
13 Hinwood, A. L.; De Klerk, N.; Rodriguez, C.; Jacoby, P.; Runnion, T.; Rye, P.; Landau, L.;
14 Murray, F.; Feldwick, M.; Spickett, J. (2006) The relationship between changes in daily
15 air pollution and hospitalizations in Perth, Australia 1992-1998: a case-crossover study.
16 Int. J. Environ. Health Res. 16: 27-46.
17 Hirsch, T.; Weiland, S. K.; Von Mutius, E.; Safeca, A. F.; Grafe, H.; Csaplovics, E.; Duhme, H.;
18 Keil, U.; Leupold, W. (1999) Inner city air pollution and respiratory health and atopy in
19 children. Eur. Respir. J. 14: 669-677.
20 Hoek, G. (2003) Daily mortality and air pollution in The Netherlands. In: Revised analyses of
21 time-series studies of air pollution and health. Special report. Boston, MA: Health Effects
22 Institute; pp. 133-141. Available: http://www.healtheffects.org/Pubs/TimeSeries.pdffl2
23 May, 2004].
24 Hoek, G.; Brunekreef, B. (1993) Acute effects of a winter air pollution episode on pulmonary
25 function and respiratory symptoms of children. Arch. Environ. Health 48: 328-335.
26 Hoek, G.; Brunekreef, B.; Verhoeff, A.; Van Wijnen, J.; Fischer, P. (2000) Daily mortality and
27 air pollution in the Netherlands. J. Air Waste Manage. Assoc. 50: 1380-1389.
28 Hoek, G.; Fischer, P.; Van Den Brandt, P.; Goldbohm, S.; Brunekreef, B. (2001) Estimation of
29 long-term average exposure to outdoor air pollution for a cohort study on mortality. J.
30 Exposure Anal. Environ. Epidemiol. 11: 459-469.
31 Holguin, F.; Tellez-Rojo, M. M.; Hernandez, M.; Cortez, M.; Chow, J. C.; Watson, J. G.;
32 Mannino, D.; Romieu, I. (2003) Air pollution and heart rate variability among the elderly
33 in Mexico City. Epidemiology 14: 521-527.
34 Hong, Y.-C.; Lee, J.-T.; Kim, H.; Kwon, H.-J. (2002) Air pollution: a new risk factor in ischemic
35 stroke mortality. Stroke 33: 2165-2169.
36 Hosseinpoor, A. R.; Forouzanfar, M. H.; Yunesian, M.; Asghari, F.; Naieni, K. H.; Farhood, D.
37 (2005) Air pollution and hospitalization due to angina pectoris in Tehran, Iran: a time-
38 series study. Environ. Res. 99: 126-131.
39 Hwang, J.-S.; Chan, C.-C. (2002) Effects of air pollution on daily clinic visits for lower
40 respiratory tract illness. Am. J. Epidemiol. 155: 1-10.
March 2008 AX6-202 DRAFT-DO NOT QUOTE OR CITE
-------
1 Hwang, B.-F.; Lee, Y.-L.; Lin, Y.-C.; Jaakkola, J. J. K.; Guo, Y. L. (2005) Traffic related air
2 pollution as a determinant of asthma among Taiwanese school children. Thorax 60: 467-
3 473.
4 Ilabaca, M.; Olaeta, I; Campos, E.; Villaire, J.; Tellez-Rojo, M. M.; Romieu, I. (1999)
5 Association between levels of fine particulate and emergency visits for pneumonia and
6 other respiratory illnesses among children in Santiago, Chile. J. Air Waste Manage.
7 Assoc. 49: 154-163.
8 Ito, K. (2003) Associations of parti culate matter components with daily mortality and morbidity
9 in Detroit, Michigan. In: Revised analyses of time-series studies of air pollution and
10 health. Special report. Boston, MA: Health Effects Institute; pp. 143-156. Available:
11 http://pubs.healtheffects.org/view.php?id=4 [12 February, 2008].
12 Ito, K. (2004) Revised ozone risk estimates for daily mortality and hospitalizations in Detroit,
13 Michigan [personal communication with attachments to Jee Young Kim]. New York,
14 NY: New York University School of Medicine, Nelson Institute of Environmental
15 Medicine; October 31.
16 Jaffe, D. H.; Singer, M. E.; Rimm, A. A. (2003) Air pollution and emergency department visits
17 for asthma among Ohio Medicaid recipients, 1991-1996. Environ. Res. 91: 21-28.
18 Jalaludin, B. B.; O'Toole, B. L; Leeder, S. R. (2004) Acute effects of urban ambient air pollution
19 on respiratory symptoms, asthma medication use, and doctor visits for asthma in a cohort
20 of Australian children. Environ Res. 95: 32-42.
21 Jalaludin, B.; Morgan, G.; Lincoln, D.; Sheppeard, V.; Simpson, R.; Corbett, S. (2006)
22 Associations between ambient air pollution and daily emergency department attendances
23 for cardiovascular disease in the elderly (65+ years), Sydney, Australia. J. Exposure Sci.
24 Environ. Epidemiol. 16: 225-237.
25 Just, J.; Segala, C.; Sahraoui, F.; Priol, G.; Grimfeld, A.; Neukirch, F. (2002) Short-term health
26 effects of parti culate and photochemical air pollution in asthmatic children. Eur. Respir.
27 J. 20: 899-906.
28 Karr, C.; Lumley, T.; Shepherd, K.; Davis, R.; Larson, T.; Ritz, B.; Kaufman, J. (2006) A case-
29 crossover study of wintertime ambient air pollution and infant bronchiolitis. Environ.
30 Health Perspect. 114: 277-281.
31 Kelsall, J. E.; Samet, J. M.; Zeger, S. L.; Xu, J. (1997) Air pollution and mortality in
32 Philadelphia, 1974-1988. Am. J. Epidemiol. 146: 750-762.
33 Kesten, S.; Szalai, J.; Dzyngel, B. (1995) Air quality and the frequency of emergency room visits
34 for asthma. Ann. Allergy Asthma Immunol. 74: 269-273.
35 Kim, J. J.; Smorodinsky, S.; Lipsett, M.; Singer, B. C.; Hodgson, A. T.; Ostro, B. (2004a)
36 Traffic-related air pollution near busy roads: the East Bay children's Respiratory Health
37 Study. Am. J. Respir. Crit. Care Med. 170: 520-526.
38 Kim, S.-Y.; Lee, J.-T.; Hong, Y.-C.; Ahn, K.-J.; Kim, H. (2004b) Determining the threshold
39 effect of ozone on daily mortality: an analysis of ozone and mortality in Seoul, Korea,
40 1995-1999. Environ. Res. 94: 113-119.
March 2008 AX6-203 DRAFT-DO NOT QUOTE OR CITE
-------
1 Kim, D.; Sass-Kortsak, A.; Purdham, J. T.; Dales, R. E.; Brook, J. R. (2006) Associations
2 between personal exposures and fixed-site ambient measurements of fine particulate
3 matter, nitrogen dioxide, and carbon monoxide in Toronto, Canada. J. Exposure Sci.
4 Environ. Epidemiol. 16: 172-183.
5 Kim, S-Y.; O'Neill, M. S.; Lee, J-T.; Cho, Y.; Kim, J.; Kim, H. (2007) Air pollution,
6 socioeconomic position, and emergency hospital visits for asthma in Seoul, Korea. Int.
7 Arch. Occup. Environ. Health 80: 701-710.
8 Kinney, P. L.; Ozkaynak, H. (1991) Associations of daily mortality and air pollution in Los
9 Angeles County. Environ. Res. 54: 99-120.
10 Klemm, R. J.; Mason, R. M., Jr. (2000) Aerosol Research and Inhalation Epidemiological Study
11 (ARIES): air quality and daily mortality statistical modeling—interim results. J. Air.
12 Waste Manage. Assoc. 50: 1433-1439.
13 Klemm, R. J.; Lipfert, F. W.; Wyzga, R. E.; Gust, C. (2004) Daily mortality and air pollution in
14 Atlanta: two years of data from ARIES. Inhalation Toxicol. 16(suppl. 1): 131-141.
15 Ko, F. W. S.; Tarn, W.; Wong, T. W.; Lai, C. K. W.; Wong, G. W. K.; Leung, T.-F.; Ng, S. S. S.;
16 Hui, D. S. C. (2007a) Effects of air pollution on asthma hospitalization rates in different
17 age groups in Hong Kong. Clin. Exp. Allergy 37: 1312-1319.
18 Ko, F. W. S.; Tarn, W.; Wong, T. W.; Chan, D. P. S.; Tung, A. H.; Lai, C. K. W.; Hui, D. S. C.
19 (2007b) Temporal relationship between air pollutants and hospital admissions for chronic
20 obstructive pulmonary disease in Hong Kong. Thorax 62: 780-785.
21 Korn, E. L.; Whittemore, A. S. (1979) Methods for analyzing panel studies of acute health
22 effects of air pollution. Biometrics 35: 795-804.
23 Kwon, H.-J.; Cho, S.-H.; Nyberg, F.; Pershagen, G. (2001) Effects of ambient air pollution on
24 daily mortality in a cohort of patients with congestive heart failure. Epidemiology 12:
25 413-419.
26 Langstaff, J. E. (2006) Summary of percentiles of NO2 data pooled across monitoring sites
27 (2003-2005). Personal communication.
28 Lanki, T.; Pekkanen, J.; Aalto, P.; Elosua, R.; Berglind, N.; DTppoliti, D.; Kulmala, M.; Nyberg,
29 F.; Peters, A.; Picciotto, S.; Salomaa, V.; Sunyer, J.; Tiittanen, P.; Von Klot, S.;
30 Forastiere, F.; for the HEAPSS Study Group. (2006) Associations of traffic-related air
31 pollutants with hospitalisation for first acute myocardial infarction: the HEAPSS study.
32 Occup. Environ. Med. 63: 844-851.
33 Le Tertre, A.; Quenel, P.; Eilstein, D.; Medina, S.; Prouvost, H.; Pascal, L.; Boumghar, A.;
34 Saviuc, P.; Zeghnoun, A.; Filleul, L.; Declercq, C.; Cassadou, S.; Le Goaster, C. (2002)
35 Short-term effects of air pollution on mortality in nine French cities: a quantitative
36 summary. Arch. Environ. Health 57: 311-319.
37 Lee, J.-T.; Schwartz, J. (1999) Reanalysis of the effects of air pollution on daily mortality in
38 Seoul, Korea: a case-crossover design. Environ. Health Perspect. 107: 633-636.
39 Lee, J.-T.; Shin, D.; Chung, Y. (1999) Air pollution and daily mortality in Seoul and Ulsan,
40 Korea. Environ. Health Perspect. 107: 149-154.
March 2008 AX6-204 DRAFT-DO NOT QUOTE OR CITE
-------
1 Lee, J.-T.; Kim, H.; Song, H.; Hong, Y.-C.; Cho, Y.-S.; Shin, S.-Y.; Hyun, Y.-J.; Kim, Y.-S.
2 (2002) Air pollution and asthma among children in Seoul, Korea. Epidemiology 13: 481-
3 484.
4 Lee, J.-T.; Kim, H.; Cho, Y.-S.; Hong, Y.-C.; Ha, E.-H.; Park, H. (2003a) Air pollution and
5 hospital admissions for ischemic heart diseases among individuals 64+ years of age
6 residing in Seoul, Korea. Arch. Environ. Health 58: 617-623.
7 Lee, B. E.; Ha, E. H.; Park, H. S.; Kim, Y. J.; Hong, Y. C.; Kim, H.; Lee, J. T. (2003b) Exposure
8 to air pollution during different gestational phases contributes to risks of low birth
9 weight. Hum. Reprod. 18: 638-643.
10 Lee, S. L.; Wong, W. H. S.; Lau, Y. L. (2006) Association between air pollution and asthma
11 admission among children in Hong Kong. Clin. Exp. Allergy 36: 1138-1146.
12 Lee, I.-M.; Tsai, S.-S.; Chang, C.-C.; Ho, C.-K.; Yang, C.-Y. (2007) Air pollution and hospital
13 admissions for chronic obstructive pulmonary disease in a tropical city: Kaohsiung,
14 Taiwan. Inhalation Toxicol. 19: 393-398.
15 Leem, J.-H.; Kaplan, B. M.; Shim, Y. K.; Pohl, H. R.; Gotway, C. A.; Bullard, S. M.; Rogers, J.
16 F.; Smith, M. M.; Tylenda, C. A. (2006) Exposures to air pollutants during pregnancy
17 and preterm delivery. Environ. Health Perspect. 114: 905-910.
18 Liang, K.-Y.; Zeger, S. L. (1986) Longitudinal data analysis using generalized linear models.
19 Biometrika73: 13-22.
20 Liao, D.; Duan, Y.; Whitsel, E. A.; Zheng, Z.-J.; Heiss, G.; Chinchilli, V. M.; Lin, H.-M. (2004)
21 Association of higher levels of ambient criteria pollutants with impaired cardiac
22 autonomic control: a population-based study. Am. J. Epidemiol. 159: 768-777.
23 Liao, D.; Heiss, G.; Chinchilli, V. M.; Duan, Y.; Folsom, A. R.; Lin, H. M.; Salomaa, V. (2005)
24 Association of criteria pollutants with plasma hemostatic/inflammatory markers: a
25 population-based study. J. Exposure Anal. Environ. Epidemiol. 15: 319-328.
26 Lin, C. A.; Martins, M. A.; Farhat, S. C. L.; Pope, C. A., Ill; Concei9ao, G. M. S.; Anastacio, V.
27 M.; Hatanaka, M.; Andrade, W. C.; Hamaue, W. R.; Bohm, G. M.; Saldiva, P. H. N.
28 (1999) Air pollution and respiratory illness of children in Sao Paulo, Brazil. Paediatr.
29 Perinat. Epidemiol. 13: 475-488.
30 Lin, M.; Chen, Y.; Burnett, R. T.; Villeneuve, P. J.; Krewski, D. (2003) Effect of short-term
31 exposure to gaseous pollution on asthma hospitalisation in children: a bi-directional case-
32 crossover analysis. J. Epidemiol. Community Health 57: 50-55.
33 Lin, C.-M.; Li, C.-Y.; Yang, G.-Y.; Mao, I-F. (2004) Association between maternal exposure to
34 elevated ambient sulfur dioxide during pregnancy and term low birth weight. Environ.
35 Res. 96: 41-50.
36 Lin, M.; Stieb, D. M.; Chen, Y. (2005) Coarse particulate matter and hospitalization for
37 respiratory infections in children younger than 15 years in Toronto: a case-crossover
38 analysis. Pediatrics 116: 235-240.
39 Linaker, C. H.; Chauhan, A. J.; Inskip, H. M.; Holgate, S. T.; Coggon, D. (2000) Personal
40 exposures of children to nitrogen dioxide relative to concentrations in outdoor air. Occup.
41 Environ. Med. 57: 472-476.
March 2008 AX6-205 DRAFT-DO NOT QUOTE OR CITE
-------
1 Linares, C.; Diaz, J.; Tobias, A.; De Miguel, J. M.; Otero, A. (2006) Impact of urban air
2 pollutants and noise levels over daily hospital admissions in children in Madrid: a time
3 series analysis. Int. Arch. Occup. Environ. Health 79: 143-152.
4 Linn, W. S.; Solomon, J. C.; Trim, S. C.; Spier, C. E.; Shamoo, D. A.; Venet, T. G.; Avol, E. L.;
5 Hackney, J. D. (1985) Effects of exposure to 4 ppm nitrogen dioxide in healthy and
6 asthmatic volunteers. Arch. Environ. Health 40: 234-239.
7 Linn, W. S.; Szlachcic, Y.; Gong, H., Jr.; Kinney, P. L.; Berhane, K. T. (2000) Air pollution and
8 daily hospital admissions in metropolitan Los Angeles. Environ. Health Perspect. 108:
9 427-434.
10 Lipfert, F. W.; Morris, S. C.; Wyzga, R. E. (2000) Daily mortality in the Philadelphia
11 metropolitan area and size-classified paniculate matter. J. Air Waste Manage. Assoc. 50:
12 1501-1513.
13 Lippmann, M.; Ito, K.; Nadas, A.; Burnett, R. T. (2000) Association of particulate matter
14 components with daily mortality and morbidity in urban populations. Cambridge, MA:
15 Health Effects Institute; research report no. 95.
16 Lipsett, M.; Hurley, S.; Ostro, B. (1997) Air pollution and emergency room visits for asthma in
17 Santa Clara County, California. Environ. Health Perspect. 105: 216-222.
18 Liu, S.; Krewski, D.; Shi, Y.; Chen, Y.; Burnett, R. T. (2003) Association between gaseous
19 ambient air pollutants and adverse pregnancy outcomes in Vancouver, Canada. Environ.
20 Health Perspect. Ill: 1773-1778.
21 Llorca, J.; Salas, A.; Prieto-Salceda, D.; Chinchon-Bengoechea, V.; Delgado-Rodriguez, M.
22 (2005) Nitrogen dioxide increases cardiorespiratory admissions in Torrelavega (Spain). J.
23 Environ. Health 68: 30-35.
24 Loomis, D.; Castillejos, M.; Gold, D. R.; McDonnell, W.; Borja-Aburto, V. H. (1999) Air
25 pollution and infant mortality in Mexico City. Epidemiology 10: 118-123.
26 Luginaah, I. N.; Fung, K. Y.; Gorey, K. M.; Webster, G.; Wills, C. (2005) Association of
27 ambient air pollution with respiratory hospitalization in a government designated "area of
28 concern": the case of Windsor, Ontario. Environ. Health Perspect. 113: 290-296.
29 Luttmann-Gibson, H.; Suh, H. H.; Coull, B. A.; Dockery, D. W.; Sarnet, S. E.; Schwartz, J.;
30 Stone, P. H.; Gold, D. R. (2006) Short-term effects of air pollution on heart rate
31 variability in senior adults in Steubenville, Ohio. J. Occup. Environ. Med. 48: 780-788.
32 Magas, O. K.; Gunter, J. T.; Regens, J. L. (2007) Ambient air pollution and daily pediatric
33 hospitalizations for asthma. Environ. Sci. Pollut. Res. 14: 19-23.
34 Mann, J. K.; Tager, I. B.; Lurmann, F.; Segal, M.; Quesenberry, C. P., Jr.; Lugg, M. M.; Shan, J.;
35 Van den Eeden, S. K. (2002) Air pollution and hospital admissions for ischemic heart
36 disease in persons with congestive heart failure or arrhythmia. Environ. Health Perspect.
37 110:1247-1252.
38 Mannes, T.; Jalaludin, B.; Morgan, G; Lincoln, D.; Sheppeard, V.; Corbett, S. (2005) Impact of
39 ambient air pollution on birth weight in Sydney, Australia. Occup. Environ. Med. 62:
40 524-530.
March 2008 AX6-206 DRAFT-DO NOT QUOTE OR CITE
-------
1 Maroziene, L.; Grazuleviciene, R. (2002) Maternal exposure to low-level air pollution and
2 pregnancy outcomes: a population-based study. Environ. Health: Global Access Sci.
3 Source 1: 6. Available: http://www.ehjournal.net/content/l/l/6 [10 October, 2006].
4 Martins, L. C.; Latorre, M. R. D. O.; Saldiva, P. H. N.; Braga, A. L. F. (2002) Air pollution and
5 emergency room visits due to chronic lower respiratory diseases in the elderly: an
6 ecological time-series study in Sao Paulo, Brazil. J. Occup. Environ. Med. 44: 622-627.
7 Metzger, K. B.; Tolbert, P. E.; Klein, M.; Peel, J. L.; Flanders, W. D.; Todd, K. H.; Mulholland,
8 J. A.; Ryan, P. B.; Frumkin , H. (2004) Ambient air pollution and cardiovascular
9 emergency department visits. Epidemiology 15: 46-56.
10 Metzger, K. B.; Klein, M.; Flanders, W. D.; Peel, J. L.; Mulholland, J. A.; Langberg, J. J.;
11 Tolbert, P. E. (2007) Ambient air pollution and cardiac arrhythmias in patients with
12 implantable defibrillators. Epidemiology 18: 585-592.
13 Michelozzi, P.; Forastiere, F.; Fusco, D.; Perucci, C. A.; Ostro, B.; Ancona, C.; Pallotti, G.
14 (1998) Air pollution and daily mortality in Rome, Italy. Occup. Environ. Med. 55: 605-
15 610.
16 Migliaretti, G.; Cavallo, F. (2004) Urban air pollution and asthma in children. Pediatr. Pulmonol.
17 38: 198-203.
18 Migliaretti, G.; Cadum, E.; Migliore, E.; Cavallo, F. (2005) Traffic air pollution and hospital
19 admission for asthma: a case-control approach in a Turin (Italy) population. Int. Arch.
20 Occup. Environ. Health. 78: 164-169.
21 Millstein, J.; Gilliland, F.; Berhane, K.; Gauderman, W. J.; McConnell, R.; Avol, E.; Rappaport,
22 E. B.; Peters, J. M. (2004) Effects of ambient air pollutants on asthma medication use and
23 wheezing among fourth-grade school children from 12 Southern California communities
24 enrolled in The Children's Health Study. Arch. Environ. Health 59: 505-514.
25 Moolgavkar, S. H. (2000a) Air pollution and hospital admissions for chronic obstructive
26 pulmonary disease in three metropolitan areas in the United States. In: Grant, L. D., ed.
27 PM2000: particulate matter and health. Inhalation Toxicol. 12(suppl. 4): 75-90.
28 Moolgavkar, S. H. (2000b) Air pollution and hospital admissions for diseases of the circulatory
29 system in three U.S. metropolitan areas. J. Air Waste Manage Assoc. 50: 1199-1206.
30 Moolgavkar, S. H. (2000c) Air pollution and daily mortality in three U.S. counties. Environ.
31 Health Perspect. 108: 777-784.
32 Moolgavkar, S. H. (2003) Air pollution and daily deaths and hospital admissions in Los Angeles
33 and Cook counties. In: Revised analyses of time-series studies of air pollution and health.
34 Special report. Boston, MA: Health Effects Institute; pp. 183-198. Available:
35 http://www.healtheffects.org/news.htm [16 May, 2003].
36 Moolgavkar, S. H.; Luebeck, E. G.; Anderson, E. L. (1997) Air pollution and hospital
37 admissions for respiratory causes in Minneapolis-St. Paul and Birmingham.
38 Epidemiology 8: 364-370.
39 Morgan, G.; Corbett, S.; Wlodarczyk, J. (1998a) Air pollution and hospital admissions in
40 Sydney, Australia, 1990 to 1994. Am. J. Public Health 88: 1761-1766.
March 2008 AX6-207 DRAFT-DO NOT QUOTE OR CITE
-------
1 Morgan, G.; Corbett, S.; Wlodarczyk, J.; Lewis, P. (1998b) Air pollution and daily mortality in
2 Sydney, Australia, 1989 through 1993. Am. J. Public Health 88: 759-764.
3 Morris, R. D.; Naumova, E. N.; Munasinghe, R. L. (1995) Ambient air pollution and
4 hospitalization for congestive heart failure among elderly people in seven large US cities.
5 Am. J. Public Health 85: 1361-1365.
6 Mortimer, K. M.; Neas, L. M.; Dockery, D. W.; Redline, S.; Tager, I. B. (2002) The effect of air
7 pollution on inner-city children with asthma. Eur. Respir. J. 19: 699-705.
8 Moseler, M.; Hendel-Kramer, A.; Karmaus, W.; Forster, J.; Weiss, K.; Urbanek, R.; Kuehr, J.
9 (1994) Effect of moderate NO2 air pollution on the lung function of children with
10 asthmatic symptoms. Environ. Res. 67: 109-124.
11 Mukala, K.; Pekkanen, J.; Tiittanen, P.; Aim, S.; Salonen, R. O.; Tuomisto, J. (1999) Personally
12 measured weekly exposure to NO2 and respiratory health among preschool children. Eur.
13 Respir. J. 13: 1411-1417.
14 Nafstad, P.; Haheim, L. L.; Wisloff, T.; Gram, F.; Oftedal, B.; Holme, L; Hjermann, L; Leren, P.
15 (2004) Urban air pollution and mortality in a cohort of Norwegian men. Environ. Health
16 Perspect. 112: 610-605.
17 Neidell, M. J. (2004) Air pollution, health, and socio-economic status: the effect of outdoor air
18 quality on childhood asthma. J. Health Econ. 23: 1209-1236.
19 Nerriere, E.; Zmirou-Navier, D.; Blanchard, O.; Momas, L; Ladner, J.; Le Moullec, Y.;
20 Personnaz, M.-B.; Lameloise, P.; Delmas, V.; Target, A.; Desqueyroux, H. (2005) Can
21 we use fixed ambient air monitors to estimate population long-term exposure to air
22 pollutants? The case of spatial variability in the Genotox ER study. Environ. Res. 97: 32-
23 42.
24 Nitschke, M.; Pilotto, L. S.; Attewell, R. G.; Smith, B. J.; Pisaniello, D.; Martin, J.; Ruffm, R. E.;
25 Killer, J. E. (2006) A cohort study of indoor nitrogen dioxide and house dust mite
26 exposure in asthmatic children. J. Occup. Environ. Med. 48: 462-469.
27 Norris, G; Young-Pong, S. N.; Koenig, J. Q.; Larson, T. V.; Sheppard, L.; Stout, J. W. (1999)
28 An association between fine particles and asthma emergency department visits for
29 children in Seattle. Environ. Health Perspect. 107: 489-493.
30 Nyberg, F.; Gustavsson, P.; Jarup, L.; Bellander, T.; Berglind, N.; Jakobsson, R.; Pershagen, G.
31 (2000) Urban air pollution and lung cancer in Stockholm. Epidemiology 11: 487-495.
32 Oftedal, B.; Nafstad, P.; Magnus, P.; Bj0rkly, S.; Skrondal, A. (2003) Traffic related air pollution
33 and acute hospital admission for respiratory diseases in Drammen, Norway 1995-2000.
34 Eur. J. Epidemiol. 18: 671-675.
35 Ostro, B.; Sanchez, J. M.; Aranda, C.; Eskeland, G. S. (1996) Air pollution and mortality: results
36 from a study of Santiago, Chile. In: Lippmann, M., ed. Papers from the ISEA-ISEE
37 annual meeting; September 1994; Research Triangle Park, NC. J. Exposure Anal.
38 Environ. Epidemiol. 6: 97-114.
39 Ostro, B. D.; Broadwin, R.; Lipsett, M. J. (2000) Coarse and fine particles and daily mortality in
40 the Coachella Valley, California: a follow-up study. J. Exposure Anal. Environ.
41 Epidemiol. 10: 412-419.
March 2008 AX6-208 DRAFT-DO NOT QUOTE OR CITE
-------
1 Ostro, B.; Lipsett, M.; Mann, I; Braxton-Owens, H.; White, M. (2001) Air pollution and
2 exacerbation of asthma in African-American children in Los Angeles. Epidemiology 12:
3 200-208.
4 Pantazopoulou, A.; Katsouyanni, K.; Kourea-Kremastinou, J.; Trichopoulos, D. (1995) Short-
5 term effects of air pollution on hospital emergency outpatient visits and admissions in the
6 greater Athens, Greece area. Environ. Res. 69: 31-36.
7 Park, S. K.; O'Neill, M. S.; Vokonas, P. S.; Sparrow, D.; Schwartz, J. (2005) Effects of air
8 pollution on heart rate variability: the VA normative aging study. Environ. Health
9 Perspect. 113: 304-309.
10 Pattenden, S.; Hoek, G.; Braun-Fahrlander, C.; Forastiere, F.; Kosheleva, A.; Neuberger, M.;
11 Fletcher, T. (2006) NC>2 and children's respiratory symptoms in the PATY study. Occup.
12 Environ. Med. 63: 828-835.
13 Peel, J. L.; Tolbert, P. E.; Klein, M.; Metzger, K. B.; Flanders, W. D.; Knox, T.; Mulholland, J.
14 A.; Ryan, P. B.; Frumkin, H. (2005) Ambient air pollution and respiratory emergency
15 department visits. Epidemiology 16: 164-174.
16 Peel, J. L.; Metzger, K. B.; Klein, M.; Flanders, W. D.; Mulholland, J. A.; Tolbert, P. E. (2007)
17 Ambient air pollution and cardiovascular emergency department visits in potentially
18 sensitive groups. Am. J. Epidemiol. 165: 625-633.
19 Pekkanen, J.; Brunner, E. J.; Anderson, H. R.; Tiittanen, P.; Atkinson, R. W. (2000) Daily
20 concentrations of air pollution and plasma fibrinogen in London. Occup. Environ. Med.
21 57: 818-822.
22 Pekkanen, J.; Peters, A.; Hoek, G.; Tiittanen, P.; Brunekreef, B.; de Hartog, J.; Heinrich, J.;
23 Ibald-Mulli, A.; Kreyling, W. G.; Lanki, T.; Timonen, K. L.; Vanninen, E. (2002)
24 Particulate air pollution and risk of ST-segment depression during repeated submaximal
25 exercise tests among subjects with coronary heart disease: the exposure and risk
26 assessment for fine and ultrafine particles in ambient air (ULTRA) study. Circulation
27 106: 933-938.
28 Penard-Morand, C.; Charpin, D.; Raherison, C.; Kopferschmitt, C.; Caillaud, D.; Lavaud, F.;
29 Annesi-Maesano, I. (2005) Long-term exposure to background air pollution related to
30 respiratory and allergic health in schoolchildren. Clin. Exp. Allergy 35: 1279-1287.
31 Pereira, L. A. A.; Loomis, D.; Concei9ao, G. M. S.; Braga, A. L. F.; Areas, R. M.; Kishi, H. S.;
32 Singer, J. M.; Bohm, G. M.; Saldiva, P. H. N. (1998) Association between air pollution
33 and intrauterine mortality in Sao Paulo, Brazil. Environ. Health Perspect. 106: 325-329.
34 Peters, J. M.; Avol, E.; Gauderman, W. J.; Linn, W. S.; Navidi, W.; London, S. J.; Margolis, H.;
35 Rappaport, E.; Vora, H.; Gong, H., Jr.; Thomas, D. C. (1999a) A study of twelve
36 southern California communities with differing levels and types of air pollution. II.
37 Effects on pulmonary function. Am. J. Respir. Crit. Care Med. 159: 768-775.
38 Peters, J. M.; Avol, E.; Navidi, W.; London, S. J.; Gauderman, W. J.; Lurmann, F.; Linn, W. S.;
39 Margolis, H.; Rappaport, E.; Gong, H., Jr.; Thomas, D. C. (1999b) A study of twelve
40 southern California communities with differing levels and types of air pollution. I.
41 Prevalence of respiratory morbidity. Am. J. Respir. Crit. Care Med. 159: 760-767.
March 2008 AX6-209 DRAFT-DO NOT QUOTE OR CITE
-------
1 Peters, A.; Liu, E.; Verrier, R. L.; Schwartz, J.; Gold, D. R.; Mittleman, M.; Baliff, I; Oh, J. A.;
2 Allen, G.; Monahan, K.; Dockery, D. W. (2000a) Air pollution and incidence of cardiac
3 arrhythmia. Epidemiology 11: 11-17.
4 Peters, A.; Skorkovsky, J.; Kotesovec, F.; Brynda, J.; Spix, C.; Wichmann, H. E.; Heinrich, J.
5 (2000b) Associations between mortality and air pollution in central Europe. Environ.
6 Health Perspect. 108: 283-287.
7 Petroeschevsky, A.; Simpson, R. W.; Thalib, L.; Rutherford, S. (2001) Associations between
8 outdoor air pollution and hospital admissions in Brisbane, Australia. Arch. Environ.
9 Health 56: 37-52.
10 Pikhart, H.; Bobak, M.; Kriz, B.; Danova, J.; Celko, M. A.; Prikazsky, V.; Pryl, K.; Briggs, D.;
11 Elliott, P. (2000) Outdoor air concentrations of nitrogen dioxide and sulfur dioxide and
12 prevalence of wheezing in school children. Epidemiology 11: 153-160.
13 Pilotto, L. S.; Douglas, R. M.; Attewell, R. G.; Wilson, S. R. (1997) Respiratory effects
14 associated with indoor nitrogen dioxide exposure in children. Int. J. Epidemiol. 26: 788-
15 796.
16 Pilotto, L. S.; Nitschke, M.; Smith, B. J.; Pisaniello, D.; Ruffm, R. E.; McElroy, H. J.; Martin, J.;
17 Hiller, J. E. (2004) Randomized controlled trial of unflued gas heater replacement on
18 respiratory health of asthmatic schoolchildren. Int. J. Epidemiol. 33: 208-214.
19 Pino, P.; Walter, T.; Oyarzun, M.; Villegas, R.; Romieu, I. (2004) Fine particulate matter and
20 wheezing illnesses in the first year of life. Epidemiology 15: 702-708.
21 Poloniecki, J. D.; Atkinson, R. W.; Ponce de Leon, A.; Anderson, H. R. (1997) Daily time series
22 for cardiovascular hospital admissions and previous day's air pollution in London, UK.
23 Occup. Environ. Med. 54: 535-540.
24 Ponce de Leon, A.; Anderson, H. R.; Bland, J. M.; Strachan, D. P.; Bower, J. (1996) Effects of
25 air pollution on daily hospital admissions for respiratory disease in London between
26 1987-88 and 1991-92. In: St Leger, S., ed. The APHEA project. Short term effects of air
27 pollution on health: a European approach using epidemiological time series data. J.
28 Epidemiol. Community Health 50(suppl. 1): S63-S70.
29 Ponka, A. (1991) Asthma and low level air pollution in Helsinki. Arch. Environ. Health 46: 262-
30 270.
31 Ponka, A.; Virtanen, M. (1994) Chronic bronchitis, emphysema, and low-level air pollution in
32 Helsinki, 1987-1989. Environ. Res. 65: 207-217.
33 Ponka, A.; Virtanen, M. (1996) Low-level air pollution and hospital admissions for cardiac and
34 cerebrovascular diseases in Helsinki. Am. J. Public Health 86: 1273-1280.
35 Ponka, A.; Savela, M.; Virtanen, M. (1998) Mortality and air pollution in Helsinki. Arch.
36 Environ. Health 53: 281-286.
37 Posin, C.; Clark, K.; Jones, M. P.; Patterson, J. V.; Buckley, R. D.; Hackney, J. D. (1978)
38 Nitrogen dioxide inhalation and human blood biochemistry. Arch. Environ. Health 33:
39 318-324.
March 2008 AX6-210 DRAFT-DO NOT QUOTE OR CITE
-------
1 Prescott, G. I; Cohen, G. R.; Elton, R. A.; Fowkes, F. G. R.; Agius, R. M. (1998) Urban air
2 pollution and cardiopulmonary ill health: a 14.5 year time series study. Occup. Environ.
3 Med. 55: 697-704.
4 Ramadour, M.; Burel, C.; Lanteaume, A.; Vervloet, D.; Charpin, D.; Brisse, F.; Dutau, H.;
5 Charpin, D. (2000) Prevalence of asthma and rhinitis in relation to long-term exposure to
6 gaseous air pollutants. Allergy (Copenhagen) 55: 1163-1169.
7 Ramsay, T. O.; Burnett, R. T.; Krewski, D. (2003) The effect of concurvity in generalized
8 additive models linking mortality to ambient particulate matter. Epidemiology 14: 18-23.
9 Rich, D. Q.; Schwartz, 1; Mittleman, M. A.; Link, M.; Luttmann-Gibson, H.; Catalano, P. I;
10 Speizer, F. E.; Dockery, D. W. (2005) Association of short-term ambient air pollution
11 concentrations and ventricular arrhythmias. Am. J. Epidemiol. 161:1123-1132.
12 Rich, D. Q.; Kim, M. H.; Turner, J. R.; Mittleman, M. A.; Schwartz, J.; Catalano, P. J.; Dockery,
13 D. W. (2006a) Association of ventricular arrhythmias detected by implantable
14 cardioverter defibrillator and ambient air pollutants in the St Louis, Missouri
15 metropolitan area. Occup. Environ. Med. 63: 591-596.
16 Rich, D. Q.; Mittleman, M. A.; Link, M. S.; Schwartz, J.; Luttmann-Gibson, H.; Catalano, P. J.;
17 Speizer, F. E.; Gold, D. R.; Dockery, D. W. (2006b) Increased risk of paroxysmal atrial
18 fibrillation episodes associated with acute increases in ambient air pollution. Environ.
19 Health Perspect. 114: 120-123.
20 Ritz, B.; Yu, F.; Chapa, G.; Fruin, S. (2000) Effect of air pollution on preterm birth among
21 children born in Southern California between 1989 and 1993. Epidemiology 11: 502-511.
22 Roemer, W. H.; Van Wijnen, J. H . (2001) Daily mortality and air pollution along busy streets in
23 Amsterdam, 1987-1998. Epidemiology 12: 649-653.
24 Roemer, W.; Hoek, G.; Brunekreef, B. (1993) Effect of ambient winter air pollution on
25 respiratory health of children with chronic respiratory symptoms. Am. Rev. Respir. Dis.
26 147:118-124.
27 Rossi, O. V. J.; Kinnula, V. L.; Tienari, J.; Huhti, E. (1993) Association of severe asthma attacks
28 with weather, pollen, and air pollutants. Thorax 48: 244-248.
29 Ruckerl, R.; Ibald-Mulli, A.; Koenig, W. Schneider, A.; Woelke, G.; Cyrys, J.; Heinrich, J.;
30 Marder, V.; Frampton, M.; Wichmann, H. E.; Peters, A. (2006) Air pollution and markers
31 of inflammation and coagulation in patients with coronary heart disease. Environ. Health
32 Perspect. 173:432-441.
33 Ruidavets, J.-B.; Cassadou, S.; Cournot, M.; Bataille, V.; Meybeck, M.; Ferrieres, J. (2005)
34 Increased resting heart rate with pollutants in a population based study. J. Epidemiol.
35 Community Health 59: 685-693.
36 Saez, M.; Tobias, A.; Munoz, P.; Campbell, M. J. (1999) A GEE moving average analysis of the
37 relationship between air pollution and mortality for asthma in Barcelona, Spain. Stat.
38 Med. 18: 2077-2086.
39 Saez, M.; Ballester, F.; Barcelo, M. A.; Perez-Hoyos, S.; Bellido, J.; Tenias, J. M.; Ocafia, R.;
40 Figueiras, A.; Arribas, F.; Aragones, N.; Tobias, A.; Cirera, L.; Canada, A.; on behalf of
41 the EMECAM Group. (2002) A combined analysis of the short-term effects of
March 2008 AX6-211 DRAFT-DO NOT QUOTE OR CITE
-------
1 photochemical air pollutants on mortality within the EMECAM project. Environ. Health
2 Perspect. 110:221-228.
3 Salam, M. T.; Millstein, I; Li, Y.-F.; Lurmann, F. W.; Margolis, H. G.; Gilliland, F. D. (2005)
4 Birth outcomes and prenatal exposure to ozone, carbon monoxide, and particulate matter:
5 results from the Children's Health Study. Environ. Health Perspect. 113: 1638-1644.
6 Saldiva, P. H. N.; Lichtenfels, A. J. F. C.; Paiva, P. S. O.; Barone, I. A.; Martins, M. A.; Massad,
7 E.; Pereira, J. C. R.; Xavier, V. P.; Singer, J. M.; Bohm, G. M. (1994) Association
8 between air pollution and mortality due to respiratory diseases in children in Sao Paulo,
9 Brazil: a preliminary report. Environ. Res. 65: 218-225.
10 Saldiva, P. H. N.; Pope, C. A., Ill; Schwartz, J.; Dockery, D. W.; Lichtenfels, A. J.; Salge, J. M.;
11 Barone, I; Bohm, G. M. (1995) Air pollution and mortality in elderly people: a time-
12 series study in Sao Paulo, Brazil. Arch. Environ. Health 50: 159-163.
13 Salome, C. M.; Brown, N. J.; Marks, G. B.; Woolcock, A. J.; Johnson, G. M.; Nancarrow, P. C.;
14 Quigley, S.; Tiong, J. (1996) Effect of nitrogen dioxide and other combustion products on
15 asthmatic subjects in a home-like environment. Eur. Respir. J. 9: 910-918.
16 Samet, J. M.; Dominici, F.; Zeger, S. L.; Schwartz, J.; Dockery, D. W. (2000a) National
17 morbidity, mortality, and air pollution study. Part I: methods and methodologic issues.
18 Cambridge, MA: Health Effects Institute; research report no. 94.
19 Samet, J. M.; Zeger, S. L.; Dominici, F.; Curriero, F.; Coursac, L; Dockery, D. W.; Schwartz, J.;
20 Zanobetti, A. (2000b) The national morbidity, mortality, and air pollution study. Part II:
21 morbidity, mortality, and air pollution in the United States. Cambridge, MA: Health
22 Effects Institute; research report no. 94, part II.
23 Samoli, E.; Analitis, A.; Touloumi, G.; Schwartz, J.; Anderson, H. R.; Sunyer, J.; Bisanti, L.;
24 Zmirou, D.; Vonk, J. M.; Pekkanen, J.; Goodman, P.; Paldy, A.; Schindler, C.;
25 Katsouyanni, K. (2005) Estimating the exposure-response relationships between
26 particulate matter and mortality within the APHEA multicity project. Environ. Health
27 Perspect. 113: 88-95.
28 Samoli, E.; Aga, E.; Touloumi, G.; Nisiotis, K.; Forsberg, B.; Lefranc, A.; Pekkanen, J.;
29 Wojtyniak, B.; Schindler, C.; Niciu, E.; Brunstein, R.; Dodic Fikfak, M.; Schwartz, J.;
30 Katsouyanni, K. (2006) Short-term effects of nitrogen dioxide on mortality: an analysis
31 within the APHEA project. Eur. Respir. J. 27: 1129-1137.
32 Sarnat, J. A.; Schwartz, J.; Catalano, P. J.; Suh, H. H. (2001) Gaseous pollutants in particulate
33 matter epidemiology: confounders or surrogates? Environ. Health Perspect. 109: 1053-
34 1061.
35 Sarnat, J. A.; Brown, K. W.; Schwartz, J.; Coull, B. A.; Koutrakis, P. (2005) Ambient gas
36 concentrations and personal particulate matter exposures: implications for studying the
37 health effects of particles. Epidemiology 16: 385-395.
38 Sartor, F.; Snacken, R.; Demuth, C.; Walckiers, D. (1995) Temperature, ambient ozone levels,
39 and mortality during summer, 1994, in Belgium. Environ. Res. 70: 105-113.
March 2008 AX6-212 DRAFT-DO NOT QUOTE OR CITE
-------
1 Schildcrout, J. S.; Sheppard, L.; Lumley, T.; Slaughter, J. C.; Koenig, J. Q.; Shapiro, G. G.
2 (2006) Ambient air pollution and asthma exacerbations in children: an eight-city analysis.
3 Am. J. Epidemiol. 164: 505-517.
4 Schindler, C.; Ackermann-Liebrich, U.; Leuenberger, P.; Monn, C.; Rapp, R.; Bolognini, G.;
5 Bongard, J.-P.; Brandli, O.; Domenighetti, G.; Karrer, W.; Keller, R.; Medici, T. G.;
6 Perruchoud, A. P.; Schoni, M. H.; Tschopp, J.-M.; Villiger, B.; Zellweger, J.-P.;
7 SAPALDIA Team. (1998) Associations between lung function and estimated average
8 exposure to NC>2 in eight areas of Switzerland. Epidemiology 9: 405-411.
9 Schouten, J. P.; Vonk, J. M.; de Graaf, A. (1996) Short term effects of air pollution on
10 emergency hospital admissions for respiratory disease: results of the APHEA project in
11 two major cities in The Netherlands, 1977-89. In: St Leger, S., ed. The APHEA project.
12 Short term effects of air pollution on health: a European approach using epidemiological
13 time series data. J. Epidemiol. Community Health 50(suppl. 1): S22-S29.
14 Schwartz, J. (1997) Air pollution and hospital admissions for cardiovascular disease in Tucson.
15 Epidemiology 8: 371-377.
16 Schwartz, J. (2001) Air pollution and blood markers of cardiovascular risk. Environ. Health
17 Perspect. Suppl. 109: 405-409.
18 Schwartz, J.; Dockery, D. W.; Neas, L. M.; Wypij, D.; Ware, J. H.; Spengler, J. D.; Koutrakis,
19 P.; Speizer, F. E.; Ferris, B. G., Jr. (1994) Acute effects of summer air pollution on
20 respiratory symptom reporting in children. Am. J. Respir. Crit. Care Med. 150: 1234-
21 1242.
22 Schwartz, J.; Litonjua, A.; Suh, H.; Verrier, M.; Zanobetti, A.; Syring, M.; Nearing, B.; Verrier,
23 R.; Stone, P.; MacCallum, G.; Speizer, F. E.; Gold, D. R. (2005) Traffic related pollution
24 and heart rate variability in a panel of elderly subjects. Thorax 60: 455-461.
25 Segala, C.; Fauroux, B.; Just, J.; Pascual, L.; Grimfeld, A.; Neukirch, F. (1998) Short-term effect
26 of winter air pollution on respiratory health of asthmatic children in Paris. Eur. Respir. J.
27 11:677-685.
28 Segala, C.; Poizeau, D.; Neukirch, F.; Aubier, M.; Samson, J.; Gehanno, P. (2004) Air pollution,
29 passive smoking, and respiratory symptoms in adults. Arch. Environ. Health 59: 669-676.
30 Setiani, O. (1996) Trend of air pollution and its effect on human health in Hiroshima
31 Prefecture—a retrospective study in the cities of Otake, Kure, Mihara, Takehara,
32 Fukuyana and Kaita Town 1977-1992. Hiroshima J. Med. Sci. 45: 43-50.
33 Sheppard, L. (2005) Acute air pollution effects: consequences of exposure distribution and
34 measurements. J. Toxicol. Environ. Health Part A 68: 1127-1135.
35 Sheppard, L.; Slaughter, J. C.; Schildcrout, J.; Liu, L.-J. S.; Lumley, T. (2005) Exposure and
36 measurement contributions to estimates of acute air pollution effects. J. Exposure Anal.
37 Environ. Epidemiol. 15: 366-376.
38 Shima, M.; Adachi, M. (2000) Effect of outdoor and indoor nitrogen dioxide on respiratory
39 symptoms in schoolchildren. Int. J. Epidemiol. 29: 862-870.
March 2008 AX6-213 DRAFT-DO NOT QUOTE OR CITE
-------
1 Simpson, R. W.; Williams, G.; Petroeschevsky, A.; Morgan, G.; Rutherford, S. (1997)
2 Associations between outdoor air pollution and daily mortality in Brisbane, Australia.
3 Arch. Environ. Health 52: 442-454.
4 Simpson, R.; Denison, L.; Petroeschevsky, A.; Thalib, L.; Williams, G. (2000) Effects of
5 ambient particle pollution on daily mortality in Melbourne 1991-1996. J. Exposure Anal.
6 Environ. Epidemiol. 10: 488-496.
7 Simpson, R.; Williams, G.; Petroeschevsky, A.; Best, T.; Morgan, G.; Denison, L.; Hinwood, A.;
8 Neville, G.; Neller, A. (2005a) The short-term effects of air pollution on daily mortality
9 in four Australian cities. Aust. N. Z. J. Public Health 29: 205-212.
10 Simpson, R.; Williams, G.; Petroeschevsky, A.; Best, T.; Morgan, G.; Denison, L.; Hinwood, A.;
11 Neville, G. (2005b) The short-term effects of air pollution on hospital admissions in four
12 Australian cities. Aust. N. Z. J. Public Health 29: 213-221.
13 Sinclair, A. H.; Tolsma, D. (2004) Associations and lags between air pollution and acute
14 respiratory visits in an ambulatory care setting: 25-month results from the aerosol
15 research and inhalation epidemiological study. J. Air Waste Manage. Assoc. 54: 1212-
16 1218.
17 Slama, R.; Morgenstern, V.; Cyrys, J.; Zutavern, A.; Herbarth, O.; Wichmann, H.-E.; Heinrich,
18 J.; LISA Study Group. (2007) Traffic-related atmospheric pollutants levels during
19 pregnancy and offspring's term birth weight: a study relying on a land-use regression
20 exposure model. Environ. Health Perspect. 115: 1283-1292.
21 Smith, B. J.; Nitschke, M.; Pilotto, L. S.; Ruffin, R. E.; Pisaniello, D. L.; Wilson, K. J. (2000)
22 Health effects of daily indoor nitrogen dioxide exposure in people with asthma. Eur.
23 Respir. J. 16: 879-885.
24 Spix, C.; Wichmann, H. E. (1996) Daily mortality and air pollutants: findings from Koln,
25 Germany. In: St Leger, S., ed. The APHEA project. Short term effects of air pollution on
26 health: a European approach using epidemiological time series data. J. Epidemiol.
27 Commun. Health 50(suppl. 1): S52-S58.
28 Spix, C.; Anderson, H. R.; Schwartz, J.; Vigotti, M. A.; LeTertre, A.; Vonk, J. M.; Touloumi, G.;
29 Balducci, F.; Piekarski, T.; Bacharova, L.; Tobias, A.; Ponka, A.; Katsouyanni, K. (1998)
30 Short-term effects of air pollution on hospital admissions of respiratory diseases in
31 Europe: a quantitative summary of APHEA study results. Arch. Environ. Health 53: 54-
32 64.
33 Steinvil, A.; Kordova-Biezuner, L.; Shapira, L; Berliner, S.; Rogowski, O. (2007) Short-term
34 exposure to air pollution and inflammation-sensitive biomarkers. Environ. Res. 106: 51-
35 61.
36 Stieb, D. M.; Burnett, R. T.; Beveridge, R. C.; Brook, J. R. (1996) Association between ozone
37 and asthma emergency department visits in Saint John, New Brunswick, Canada.
38 Environ. Health Perspect. 104: 1354-1360.
39 Stieb, D. M.; Beveridge, R. C.; Brook, J. R.; Smith-Doiron, M.; Burnett, R. T.; Dales, R. E.;
40 Beaulieu, S.; Judek, S.; Mamedov, A. (2000) Air pollution, aeroallergens and
41 cardiorespiratory emergency department visits in Saint John, Canada. J. Exposure Anal.
42 Environ. Epidemiol. 10: 461-477.
March 2008 AX6-214 DRAFT-DO NOT QUOTE OR CITE
-------
1 Stieb, D. M.; Judek, S.; Burnett, R. T. (2002) Meta-analysis of time-series studies of air pollution
2 and mortality: effects of gases and particles and the influence of cause of death, age, and
3 season. J. Air Waste Manage. Assoc. 52: 470-484.
4 Stieb, D. M.; Judek, S.; Burnett, R. T. (2003) Meta-analysis of time-series studies of air pollution
5 and mortality: update in relation to the use of generalized additive models. J. Air Waste
6 Manage. 53: 258-261.
7 Stiratelli, R.; Laird, N.; Ware, J. H. (1984) Random-effects models for serial observations with
8 binary response. Biometrics 40: 961-971.
9 Studnicka, M.; Hackl, E.; Pischinger, J.; Fangmeyer, C.; Haschke, N.; Kuhr, J.; Urbanek, R.;
10 Neumann, M.; Frischer, T. (1997) Traffic-related NO2 and the prevalence of asthma and
11 respiratory symptoms in seven year olds. Eur. Respir. J. 10: 2275-2278.
12 Sun, H.-L.; Chou, M.-C.; Lue, K.-H. (2006) The relationship of air pollution to ED visits for
13 asthma differ between children and adults. Am. J. Emerg. Med. 24: 709-713.
14 Sunyer, J.; Basagafia, X. (2001) Particles, and not gases, are associated with the risk of death in
15 patients with chronic obstructive pulmonary disease. Int. J. Epidemiol. 30: 1138-1140.
16 Sunyer, J.; Castellsague, J.; Saez, M.; Tobias, A.; Anto, J. M. (1996) Air pollution and mortality
17 in Barcelona. In: St Leger, S., ed. The APHEA project. Short term effects of air pollution
18 on health: a European approach using epidemiological time series data. J. Epidemiol.
19 Community Health 50(suppl. 1): S76-S80.
20 Sunyer, J.; Spix, C.; Quenel, P.; Ponce-de-Leon, A.; Ponka, A.; Barumandzadeh, T.; Touloumi,
21 G.; Bacharova, L.; Wojtyniak, B.; Vonk, J.; Bisanti, L.; Schwartz, J.; Katsouyanni, K.
22 (1997) Urban air pollution and emergency admissions for asthma in four European cities:
23 the APHEA project. Thorax 52: 760-765.
24 Sunyer, J.; Basagafia, X.; Belmonte, J.; Anto, J. M. (2002) Effect of nitrogen dioxide and ozone
25 on the risk of dying in patients with severe asthma. Thorax 57: 687-693.
26 Tager, I. B.; Balmes, J.; Lurmann, F.; Ngo, L.; Alcorn, S.; Kiinzli, N. (2005) Chronic exposure to
27 ambient ozone and lung function in young adults. Epidemiology 16: 751-759.
28 Tanaka, H.; Honma, S.; Nishi, M.; Igarashi, T.; Teramoto, S.; Nishio, F.; Abe, S. (1998) Acid
29 fog and hospital visits for asthma: an epidemiological study. Eur. Respir. J. 11: 1301-
30 1306.
31 Tenias, J. M.; Ballester, F.; Rivera, M. L. (1998) Association between hospital emergency visits
32 for asthma and air pollution in Valencia, Spain. Occup. Environ. Med. 55: 541-547.
33 Tenias, J. M.; Ballester, F.; Perez-Hoyos, S.; Rivera, M. L. (2002) Air pollution and hospital
34 emergency room admissions for chronic obstructive pulmonary disease in Valencia,
35 Spain. Arch. Environ. Health 57: 41-47.
36 Thompson, A. J.; Shields, M. D.; Patterson, C. C. (2001) Acute asthma exacerbations and air
37 pollutants in children living in Belfast, Northern Ireland. Arch. Environ. Health 56: 234-
38 241.
March 2008 AX6-215 DRAFT-DO NOT QUOTE OR CITE
-------
1 Tobias, A.; Campbell, M. J.; Saez, M. (1999) Modelling asthma epidemics on the relationship
2 between air pollution and asthma emergency visits in Barcelona, Spain. Eur. J.
3 Epidemiol. 15: 799-803.
4 Tolbert, P. E.; Klein, M.; Metzger, K. B.; Peel, J.; Flanders, W. D.; Todd, K.; Mulholland, J. A.;
5 Ryan, P. B.; Frumkin, H. (2000) Interim results of the study of parti culates and health in
6 Atlanta (SOPHIA). J. Exposure Anal. Environ. Epidemiol. 10: 446-460.
7 Tolbert, P. E.; Klein, M.; Peel, J. L.; Sarnat, S. E.; Sarnat, J. A. (2007) Multipollutant modeling
8 issues in a study of ambient air quality and emergency department visits in Atlanta. J.
9 Exposure Sci. Environ. Epidemiol. 17(Suppl. 2s): S29-S35.
10 Touloumi, G.; Katsouyanni, K.; Zmirou, D.; Schwartz, J.; Spix, C.; Ponce de Leon, A.; Tobias,
11 A.; Quennel, P.; Rabczenko, D.; Bacharova, L.; Bisanti, L.; Vonk, J. M.; Ponka, A.
12 (1997) Short-term effects of ambient oxidant exposure on mortality: a combined analysis
13 within the APHEA project. Am. J. Epidemiol. 146: 177-185.
14 Tsai, S.-S.; Goggins, W. B.; Chiu, H.-F.; Yang, C.-Y. (2003a) Evidence for an association
15 between air pollution and daily stroke admissions in Kaohsiung, Taiwan. Stroke 34:
16 2612-2616.
17 Tsai, S.-S.; Huang, C.-H.; Goggins, W. B.; Wu, T.-N.; Yang, C.-Y. (2003b) Relationship
18 between air pollution and daily mortality in a tropical city: Kaohsiung, Taiwan. J.
19 Toxicol. Environ. Health Part A 66: 1341-1349.
20 Tsai, S.-S.; Cheng, M.-H.; Chiu, H.-F.; Wu, T.-N.; Yang, C.-Y. (2006) Air pollution and hospital
21 admissions for asthma in a tropical city: Kaohsiung, Taiwan. Inhalation Toxicol. 18: 549-
22 554.
23 U.S. Environmental Protection Agency. (1993) Air quality criteria for oxides of nitrogen.
24 Research Triangle Park, NC: Office of Health and Environmental Assessment,
25 Environmental Criteria and Assessment Office; report nos. EPA/600/8-91/049aF-cF. 3v.
26 Available from: NTIS, Springfield, VA; PB95-124533, PB95-124525, and PB95-124517.
27 U.S. Environmental Protection Agency. (2004) Air quality criteria for particulate matter.
28 Research Triangle Park, NC: National Center for Environmental Assessment; report no.
29 EPA/600/P-99/002aF-bF. 2v. Available: http://cfpub.epa.gov/ncea/ [9 November, 2004].
30 U.S. Environmental Protection Agency. (2006) Air quality criteria for ozone and related
31 photochemical oxidants. Research Triangle Park, NC: National Center for Environmental
32 Assessment; report no. EPA/600/R-05/004aF-cF. 3v. Available:
33 http://cfpub.epa.gov/ncea/ [24 March, 2006].
34 Van Strien, R. T.; Gent, J. F.; Belanger, K.; Triche, E.; Bracken, M. B.; Leaderer, B. P. (2004)
35 Exposure to NO2 and nitrous acid and respiratory symptoms in the first year of life.
36 Epidemiology 15: 471-478.
37 Vedal, S.; Brauer, M.; White, R.; Petkau, J. (2003) Air pollution and daily mortality in a city
38 with low levels of pollution. Environ. Health Perspect. Ill: 45-51.
39 Verhoeff, A. P.; Hoek, G.; Schwartz, J.; Van Wijnen, J. H. (1996) Air pollution and daily
40 mortality in Amsterdam. Epidemiology 7: 225-230.
March 2008 AX6-216 DRAFT-DO NOT QUOTE OR CITE
-------
1 Vigotti, M. A.; Chiaverini, F.; Biagiola, P.; Rossi, G. (2007) Urban air pollution and emergency
2 visits for respiratory complaints in Pisa, Italy. J. Toxicol. Environ. Health Part A 70: 266-
3 269.
4 Villeneuve, P. J.; Burnett, R. T.; Shi, Y.; Krewski, D.; Goldberg, M. S.; Hertzman, C.; Chen, Y.;
5 Brook, J. (2003) A time-series study of air pollution, socioeconomic status, and mortality
6 in Vancouver, Canada. J. Exposure Anal. Environ. Epidemiol. 13: 427-435.
7 Villeneuve, P. J.; Chen, L.; Stieb, D.; Rowe, B. H. (2006) Associations between outdoor air
8 pollution and emergency department visits for stroke in Edmonton, Canada. Eur. J.
9 Epidemiol. 21: 689-700.
10 Virnig, B. A.; McBean, M. (2001) Administrative data for public health surveillance and
11 planning. Annu. Rev. Public Health 22: 213-230.
12 Von Klot, S.; Wolke, G; Tuch, T.; Heinrich, J.; Dockery, D. W.; Schwartz, J.; Kreyling, W. G;
13 Wichmann, H. E.; Peters, A. (2002) Increased asthma medication use in association with
14 ambient fine and ultrafine particles. Eur. Respir. J. 20: 691-702.
15 Von Klot, S.; Peters, A.; Aalto, P.; Bellander, T.; Berglind, N.; DTppoliti, D.; Elosua, R.;
16 Hermann, A.; Kulmala, M.; Lanki, T.; Lowel, H.; Pekkanen, J.; Picciotto, S.; Sunyer, J.;
17 Forastiere, F.; Health Effects of Particles on Susceptible Subpopulations (HEAPSS)
18 Study Group. (2005) Ambient air pollution is associated with increased risk of hospital
19 cardiac readmissions of myocardial infarction survivors in five European cities.
20 Circulation 112: 3073-3079.
21 Wang, T.-N.; Ko, Y.-C.; Chao, Y.-Y.; Huang, C.-C.; Lin, R.-S. (1999) Association between
22 indoor and outdoor air pollution and adolescent asthma from 1995 to 1996 in Taiwan.
23 Environ. Res. 81: 239-247.
24 Wellenius, G. A.; Bateson, T. F.; Mittleman, M. A.; Schwartz, J. (2005a) Particulate air pollution
25 and the rate of hospitalization for congestive heart failure among medicare beneficiaries
26 in Pittsburgh, Pennsylvania. Am. J. Epidemiol. 161: 1030-1036.
27 Wellenius, G. A.; Schwartz, J.; Mittleman, M. A. (2005b) Air pollution and hospital admissions
28 for ischemic and hemorrhagic stroke among medicare beneficiaries. Stroke 36: 2549-
29 2553.
30 Wheeler, A.; Zanobetti, A.; Gold, D. R.; Schwartz, J.; Stone, P.; Suh, H. H. (2006) The
31 relationship between ambient air pollution and heart rate variability differs for individuals
32 with heart and pulmonary disease. Environ. Health Perspect. 114: 560-566.
33 Wong, T. W.; Lau, T. S.; Yu, T. S.; Neller, A.; Wong, S. L.; Tarn, W.; Pang, S. W. (1999) Air
34 pollution and hospital admissions for respiratory and cardiovascular diseases in Hong
35 Kong. Occup. Environ. Med. 56: 679-683.
36 Wong, G. W.; Ko, F. W.; Lau, T. S.; Li, S. T.; Hui, D.; Pang, S. W.; Leung, R.; Fok, T. F.; Lai,
37 C. K. (2001a) Temporal relationship between air pollution and hospital admissions for
38 asthmatic children in Hong Kong. Clin. Exp. Allergy 31: 565-569.
39 Wong, C.-M.; Ma, S.; Hedley, A. J.; Lam, T.-H. (2001b) Effect of air pollution on daily
40 mortality in Hong Kong. Environ. Health Perspect. 109: 335-340.
March 2008 AX6-217 DRAFT-DO NOT QUOTE OR CITE
-------
1 Wong, T. W.; Tarn, W. S.; Yu, T. S.; Wong, A. H. S. (2002) Associations between daily
2 mortalities from respiratory and cardiovascular diseases and air pollution in Hong Kong,
3 China. Occup. Environ. Med. 59: 30-35.
4 Yallop, D.; Duncan, E. R.; Norris, E.; Fuller, G. W.; Thomas, N.; Walters, 1; Dick, M. C.;
5 Height, S. E.; Thein, S. L.; Rees, D. C. (2007) The associations between air quality and
6 the number of hospital admissions for acute pain and sickle-cell disease in an urban
7 environment. Br. J. Haematol. 136: 844-848.
8 Yang, C.-Y.; Chen, C.-J. (2007) Air pollution and hospital admissions for chronic obstructive
9 pulmonary disease in a subtropical city: Taipei, Taiwan. J. Toxicol. Environ. Health Part
10 A 70: 1214-1219.
11 Yang, Q.; Chen, Y.; Shi, Y.; Burnett, R. T.; McGrail, K. M.; Krewski, D. (2003) Association
12 between ozone and respiratory admissions among children and the elderly in Vancouver,
13 Canada. Inhalation Toxicol. 15: 1297-1308.
14 Yang, C.-Y.; Chen, Y.-S.; Yang, C.-H.; Ho, S.-C. (2004a) Relationship between ambient air
15 pollution and hospital admissions for cardiovascular diseases in Kaohsiung, Taiwan. J.
16 Toxicol. Environ. Health Part A 67: 483-493.
17 Yang, C.-Y.; Chang, C.-C.; Chuang, H.-Y.; Tsai, S.-S.; Wu, T.-N.; Ho, C.-K. (2004b)
18 Relationship between air pollution and daily mortality in a subtropical city: Taipei,
19 Taiwan. Environ. Int. 30: 519-523.
20 Yang, Q.; Chen, Y.; Krewski, D.; Burnett, R. T.; Shi, Y.; McGrail, K. M. (2005) Effect of short-
21 term exposure to low levels of gaseous pollutants on chronic obstructive pulmonary
22 disease hospitalizations. Environ. Res. 99: 99-105.
23 Yang, C.-Y.; Chen, C.-C.; Chen, C.-Y.; Kuo, H.-W. (2007) Air pollution and hospital admissions
24 for asthma in a subtropical city: Taipei, Taiwan. J. Toxicol. Environ. Health Part A 70:
25 111-117.
26 Ye, F.; Piver, W. T.; Ando, M.; Portier, C. J. (2001) Effects of temperature and air pollutants on
27 cardiovascular and respiratory diseases for males and females older than 65 years of age
28 in Tokyo, July and August 1980-1995. Environ. Health Perspect. 109: 355-359.
29 Zanobetti, A.; Schwartz, J. (2006) Air pollution and emergency admissions in Boston, MA. J.
30 Epidemiol. Community Health 60: 890-895.
31 Zeger, S. L.; Thomas, D.; Dominici, F.; Samet, J. M.; Schwartz, J.; Dockery, D.; Cohen, A.
32 (2000) Exposure measurement error in time-series studies of air pollution: concepts and
33 consequences. Environ. Health Perspect. 108: 419-426.
34 Zeghnoun, A.; Czernichow, P.; Beaudeau, P.; Hautemaniere, A.; Froment, L.; Le Tertre, A.;
35 Quenel, P. (2001) Short-term effects of air pollution on mortality in the cities of Rouen
36 and Le Havre, France, 1990-1995. Arch. Environ. Health 56: 327-335.
37 Zeka, A.; Schwartz, J. (2004) Estimating the independent effects of multiple pollutants in the
38 presence of measurement error: an application of a measurement-error-resistant
39 technique. Environ. Health Perspect. 112: 1686-1690.
March 2008 AX6-218 DRAFT-DO NOT QUOTE OR CITE
-------
1 Zidek, J. V. (1997) Interpolating air pollution for health impact assessment. In: Barnett, E. V.;
2 Turkman, K. F., eds. Pollution Assessment and Control. New York, NY: John Wiley &
3 Sons. (Statistics for the Environment, no. 3).
4 Zmirou, D.; Barumandzadeh, T.; Balducci, F.; Ritter, P.; Laham, G.; Ghilardi, J.-P. (1996) Short
5 term effects of air pollution on mortality in the city of Lyon, France, 1985-90. In: St
6 Leger, S., ed. The APHEA project. Short term effects of air pollution on health: a
7 European approach using epidemiological time series data. J. Epidemiol. Community
8 Health 50(suppl. 1): S30-S35.
9 Zmirou, D.; Schwartz, J.; Saez, M.; Zanobetti, A.; Wojtyniak, B.; Touloumi, G.; Spix, C.; Ponce
10 de Leon, A.; Le Moullec, Y.; Bacharova, L.; Schouten, J.; Ponka, A.; Katsouyanni, K.
11 (1998) Time-series analysis of air pollution and cause-specific mortality. Epidemiology
12 9:495-503.
March 2008 AX6-219 DRAFT-DO NOT QUOTE OR CITE
-------
4>EPA
Please make all necessary changes in the below label, PRESORTED STANDARD
detach copy or copy, and return to the address in the upper POSTAGE & FEES PAID
left-hand corner. EPA
United States DCDMIT M^ r* Q^
Environmental Protection _ PERMIT Na Q-35
Agency If you do not wish to receive these reports CHECK HERE LJ;
detach copy or copy, and return to the address in the upper
left-hand corner.
National Center for
Environmental Assessment
Research Triangle Park, NC 27711
Official Business
Penalty for Private Use
$300
EPA/600/R-07/093bB
March 2008
------- |