f\ ^™ F^ A
> c PA
United States
Proleclton
Air Quality Criteria for
Ozone and Related
Photochemical Oxidants
(First External Review Draft)
Volume II of
-------
EPA/600/R-05/004bA
January 2005
Air Quality Criteria for Ozone and Related
Photochemical Oxidants
Volume II
National Center for Environmental Assessment-RTF Office
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, NC
-------
DISCLAIMER
This document is an external review draft for review purposes only and does not
constitute U.S. Environmental Protection Agency policy. Mention of trade names or
commercial products does not constitute endorsement or recommendation for use.
PREFACE
National Ambient Air Quality Standards (NAAQS) are promulgated by the United States
Environmental Protection Agency (EPA) to meet requirements set forth in Sections 108 and 109
of the U.S. Clean Air Act (CAA). Sections 108 and 109 require the EPA Administrator (1) to
list widespread air pollutants that reasonably may be expected to endanger public health or
welfare; (2) to issue air quality criteria for them that assess the latest available scientific
information on nature and effects of ambient exposure to them; (3) to set "primary" NAAQS to
protect human health with adequate margin of safety and to set "secondary" NAAQS to protect
against welfare effects (e.g., effects on vegetation, ecosystems, visibility, climate, manmade
materials, etc); and (5) to periodically review and revise, as appropriate, the criteria and NAAQS
for a given listed pollutant or class of pollutants.
In 1971, the U.S. Environmental Protection Agency (EPA) promulgated National Ambient
Air Quality Standards (NAAQS) to protect the public health and welfare from adverse effects of
photochemical oxidants. The EPA promulgates the NAAQS on the basis of scientific
information contained in air quality criteria issued under Section 108 of the Clean Air Act.
Following the review of criteria as contained in the EPA document, Air Quality Criteria for
Ozone and other Photochemical Oxidants published in 1978, the chemical designation of the
standards was changed from photochemical oxidants to ozone (O3) in 1979 and a 1-hour O3
NAAQS was set. The 1978 document focused primarily on the scientific air quality criteria for
O3 and, to a lesser extent, on those for other photochemical oxidants such as hydrogen peroxide
and the peroxyacyl nitrates, as have subsequent revised versions of the ozone document.
To meet Clean Air Act requirements noted above for periodic review of criteria and
NAAQS, the O3 criteria document, Air Quality Criteria for Ozone and Other Photochemical
Il-ii
-------
Oxidants, was next revised and then released in August 1986; and a supplement, Summary of
Selected New Information on Effects of Ozone on Health and Vegetation, was issued in January
1992. These documents were the basis for a March 1993 decision by EPA that revision of the
existing 1-h NAAQS for O3 was not appropriate at that time. That decision, however, did not
take into account some of the newer scientific data that became available after completion of the
1986 criteria document. Such literature was assessed in the next periodic revision of the O3 air
quality criteria document, which was completed in 1996 and provided scientific bases supporting
the setting by EPA in 1997 of an 8-h O3 NAAQS that is currently in force together with the 1-h
O3 standard.
The purpose of this revised air quality criteria document for O3 and related photochemical
oxidants is to critically evaluate and assess the latest scientific information published since that
assessed in the above 1996 Ozone Air Quality Criteria Document (O3 AQCD), with the main
focus being on pertinent new information useful in evaluating health and environmental effects
data associated with ambient air O3 exposures. However, some other scientific data are also
presented and evaluated in order to provide a better understanding of the nature, sources,
distribution, measurement, and concentrations of O3 and related photochemical oxidants and
their precursors in the environment. The document assesses pertinent literature available
through 2004.
The present draft document (dated January 2005) is being released for public comment and
review by the Clean Air Scientific Advisory Committee (CASAC) to obtain comments on the
organization and structure of the document, the issues addressed, the approaches employed in
assessing and interpreting the newly available information on O3 exposures and effects, and the
key findings and conclusions arrived at as a consequence of this assessment. Public comments
and recommendations will be taken into account making any appropriate further revisions to this
document for incorporation into a Second External Review Draft. That draft will be released for
further public comment and CASAC review before last revisions are made in response and
incorporated into a final version to be completed by early 2006. Evaluations contained in the
present document will be drawn on to provide inputs to associated PM Staff Paper analyses
prepared by EPA's Office of Air Quality Planning and Standards (OAQPS) to pose options for
consideration by the EPA Administrator with regard to proposal and, ultimately, promulgation of
decisions on potential retention or revision, as appropriate, of the current O3 NAAQS.
-------
Preparation of this document was coordinated by staff of EPA's National Center for
Environmental Assessment in Research Triangle Park (NCEA-RTP). NCEA-RTP scientific
staff, together with experts from other EPA/ORD laboratories and academia, contributed to
writing of document chapters. Earlier drafts of document materials were reviewed by non-EPA
experts in peer consultation workshops held by EPA. The document describes the nature,
sources, distribution, measurement, and concentrations of O3 in outdoor (ambient) and indoor
environments. It also evaluates the latest data on human exposures to ambient O3 and
consequent health effects in exposed human populations, to support decision making regarding
the primary, health-related O3 NAAQS. The document also evaluates ambient O3 environmental
effects on vegetation and ecosystems, man-made materials, and surface level solar UV radiation
flux and global climate change, to support decision making on secondary O3 NAAQS.
NCEA acknowledges the valuable contributions provided by authors, contributors, and
reviewers and the diligence of its staff and contractors in the preparation of this draft document.
Il-iv
-------
Air Quality Criteria for Ozone and Related
Photochemical Oxidants
(First External Review Draft)
VOLUME I
Executive Summary E-l
1. INTRODUCTION 1-1
2. PHYSICS AND CHEMISTRY OF OZONE IN THE ATMOSPHERE 2-1
CHAPTER 2 ANNEX (ATMOSPHERIC PHYSICS/CHEMISTRY) AX2-1
3. ENVIRONMENTAL CONCENTRATIONS, PATTERNS, AND
EXPOSURE ESTIMATES 3-1
CHAPTER 3 ANNEX (AIR QUALITY AND EXPOSURE) AX3-1
II-v
-------
Air Quality Criteria for Ozone and Related
Photochemical Oxidants
(First External Review Draft)
(cont'd)
VOLUME II
4. DOSIMETRY, SPECIES HOMOLOGY, SENSITIVITY, AND
ANIMAL-TO-HUMAN EXTRAPOLATION 4-1
CHAPTER 4 ANNEX (DOSIMETRY) AX4-1
5. TOXICOLOGICAL EFFECTS OF OZONE AND RELATED
PHOTOCHEMICAL OXIDANTS IN LABORATORY ANIMALS
AND IN VITRO TEST SYSTEMS 5-1
CHAPTER 5 ANNEX (ANIMAL TOXICOLOGY) AX5-1
6. CONTROLLED HUMAN EXPOSURE STUDIES OF OZONE AND
RELATED PHOTOCHEMICAL OXIDANTS 6-1
CHAPTER 6 ANNEX (CONTROLLED HUMAN EXPOSURE) AX6-1
7. EPIDEMIOLOGICAL STUDIES OF HUMAN HEALTH EFFECTS
ASSOCIATED WITH AMBIENT OZONE EXPOSURE 7-1
CHAPTER 7 ANNEX (EPIDEMIOLOGY) AX7-1
8. INTEGRATIVE SYNTHESIS 8-1
Il-vi
-------
Air Quality Criteria for Ozone and Related
Photochemical Oxidants
(First External Review Draft)
(cont'd)
VOLUME
9. ENVIRONMENTAL EFFECTS: OZONE EFFECTS ON
VEGETATION AND ECOSYSTEMS 9-1
10. TROPOSPHERIC OZONE EFFECTS ON UV-B FLUX AND
CLIMATE CHANGE PROCESSES 10-1
11. EFFECT OF OZONE ON MAN-MADE MATERIALS 11-1
-------
Table of Contents
Page
List of Tables I-xviii
List of Figures I-xxi
Authors, Contributors, and Reviewers I-xxv
U.S. Environmental Protection Agency Project Team for Development of Air
Quality Criteria for Ozone and Related Photochemical Oxidants I-xxvi
U.S. Environmental Protection Agency Science Advisory Board (SAB)
Staff Office Clean Air Scientific Advisory Committee (CASAC)
Ozone Review Panel I-xxvii
Abbreviations and Acronyms I-xxviii
4. DOSIMETRY, SPECIES HOMOLOGY, SENSITIVITY, AND
ANIMAL-TO-HUMAN EXTRAPOLATION 4-1
4.1 INTRODUCTION 4-1
4.2 DOSIMETRY OF OZONE IN THE RESPIRATORY TRACT 4-2
4.2.1 Bolus-Response Studies 4-3
4.2.2 General Uptake Studies 4-5
4.2.3 Dosimetry Modeling 4-6
4.2.4 Summary and Conclusions - Dosimetry 4-9
4.3 SPECIES HOMOLOGY, SENSITIVITY, AND ANIMAL-TO-
HUMAN EXTRAPOLATION 4-11
4.3.1 Summary and Conclusions: Species Homology, Sensitivity,
and Animal-to-Human Extrapolation 4-17
REFERENCES 4-18
AX4. DOSIMETRY OF OZONE IN THE RESPIRATORY TRACT AX4-1
AX4.1 INTRODUCTION AX4-1
AX4.2 EXPERIMENTAL OZONE DOSIMETRY INVESTIGATIONS AX4-3
AX4.2.1 Bolus-Response Studies AX4-3
AX4.2.2 General Uptake Studies AX4-9
AX4.3 DOSIMETRY MODELING AX4-10
AX4.4 SPECIES HOMOLOGY, SENSITIVITY AND ANIMAL-TO-
HUMAN EXTRAPOLATION AX4-16
REFERENCES AX4-18
5. TOXICOLOGICAL EFFECTS OF OZONE AND RELATED PHOTOCHEMICAL
OXIDANTS IN LABORATORY ANIMALS AND IN VITRO TEST SYSTEMS 5-1
5.1 INTRODUCTION 5-1
5.2 RESPIRATORY TRACT EFFECTS OF OZONE 5-2
5.2.1 Biochemical Effects 5-2
5.2.1.1 Cellular Targets of O3 Interaction 5-2
5.2.1.2 Monooxygenases 5-3
5.2.1.3 Antioxidants, Antioxidant Metabolism, and
Mitochondrial Oxygen Consumption 5-3
5.2.1.4 Lipid Metabolism and Content of the Lung 5-5
5.2.1.5 Protein Synthesis 5-6
-------
Table of Contents
(cont'd)
Page
5.2.1.6 Gene Expression 5-7
5.2.1.7 Summary and Conclusions - Biochemical
Effects 5-7
5.2.2 Lung Host Defenses 5-9
5.2.2.1 Clearance 5-9
5.2.2.2 Alveolar Macrophages 5-10
5.2.2.3 Immune System 5-11
5.2.2.4 Interactions with Infectious Microorganisms 5-14
5.2.2.5 Summary and Conclusions - Lung Host
Defenses 5-15
5.2.3 Inflammation and Lung Permeability Changes 5-16
5.2.3.1 Time Course of Inflammation and Lung
Permeability Changes 5-16
5.2.3.2 Concentration and Time of Exposure 5-17
5.2.3.3 Susceptibility Factors 5-17
5.2.3.4 Mediators of Inflammatory Response
and Injury 5-20
5.2.3.5 The Role of Nitric Oxide Synthase and
Reactive Nitrogen in Inflammation 5-23
5.2.3.6 Summary and Conclusions - Inflammation
and Permeability Changes 5-24
5.2.4 Morphological Effects 5-26
5.2.4.1 Short Term Exposure Effects 5-27
5.2.4.2 Summary of Short-Term Morphological Effects .. 5-30
5.2.4.3 Long Term Exposure Effects 5-30
5.2.4.4 Summary and Conclusions - Long-Term
Morphological Effects 5-34
5.2.5 Effects on Pulmonary Function 5-35
5.2.5.1 Acute and Short-Term Exposure Effects on
Pulmonary Function 5-35
5.2.5.2 Summary and Conclusions - Short- and
Long-Term Effects on Pulmonary Function 5-36
5.2.5.3 Ozone Effects on Airway Responsiveness 5-37
5.2.5.4 Summary and Conclusions - Effects on
Airway Responsiveness 5-46
5.2.6 Genotoxicity Potential of Ozone 5-47
5.2.6.1 Summary and Conclusions - Genotoxicity
Potential of Ozone 5-48
5.3 SYSTEMIC EFFECTS OF OZONE EXPOSURE 5-48
5.3.1 Neurobehavioral Effects 5-48
5.3.2 NeuroendocrineEffects 5-49
5.3.3 Cardiovascular Effects 5-50
5.3.4 Reproductive and Developmental Effects 5-52
5.3.5 Effects on the Liver, Spleen, and Thymus 5-54
Il-ix
-------
Table of Contents
(cont'd)
Page
5.3.6 Effects on Cutaneous and Ocular Tissues 5-54
5.3.7 Summary and Conclusions - Systemic Effects of Ozone 5-5 5
5.4 INTERACTIONS OF OZONE WITH OTHER CO-OCCURRING
POLLUTANTS 5-56
5.4.1 Ozone and Nitrogen Oxides 5-5 6
5.4.2 Ozone and Other Copollutants 5-57
5.4.3 Complex (Multicomponent) Mixtures Containing Ozone 5-60
5.4.4 Summary and Conclusions - Interactions of Ozone with
other Co-occurring Pollutants 5-67
5.5 EFFECTS OF OTHER PHOTOCHEMICAL OXIDANTS 5-68
5.5.1 Summary and Conclusions - Effects of Other Photochemical
Oxidants 5-69
REFERENCES 5-70
AX5. ANNEX TO CHAPTER 5 OF OZONE AQCD AX5-1
AX5.1 INTRODUCTION AX5-1
AX5.2 RESPIRATORY TRACT EFFECTS OF OZONE AX5-1
AX5.2.1 Biochemical Effects AX5-1
AX5.2.1.1 Cellular Targets of Ozone Interaction AX5-1
AX5.2.1.2 Monooxygenases AX5-3
AX5.2.1.3 Antioxidants, Antioxidant Metabolism,
and Mitochondrial Oxygen Consumption AX5-5
AX5.2.1.4 Lipid Metabolism and Content of
the Lung AX5-7
AX5.2.1.5 Protein Synthesis AX5-8
AX5.2.1.6 Gene Expression AX5-8
AX5.2.2 Lung Host Defenses AX5-9
AX5.2.2.1 Clearance AX5-9
AX5.2.2.2 Alveolar Macrophages AX5-18
AX5.2.2.3 Immune System AX5-20
AX5.2.2.4 Interaction with Infectious Microorganisms .. AX5-24
AX5.2.3 Inflammation and Lung Permeability Changes AX5-24
AX5.2.3.1 Time Course of Inflammation and Lung
Permeability Changes AX5-25
AX5.2.3.2 Concentration and Time of Exposure AX5-34
AX5.3.3.3 Susceptibility Factors AX5-34
AX5.2.3.4 Mediators of Inflammatory Response
and Injury AX5-38
AX5.2.3.5 Role of Nitric Oxide Synthase and
Reactive Nitrogen in Inflammation AX5-41
AX5.2.4 Morphological Effects AX5-42
AX5.2.4.1 Introduction AX5-42
AX5.2.4.2 Short-Term Exposure Effects on
Morphology AX5-43
II-x
-------
Table of Contents
(cont'd)
Page
AX5.2.4.3 Long-Term Exposure Effects on
Morphology AX5-49
AX5.2.5 Effects on Pulmonary Function AX5-55
AX5.2.5.1 Introduction AX5-55
AX5.2.5.2 Acute and Short-Term Exposure Effects
on Pulmonary Function AX5-55
AX5.2.5.3 Long-Term Exposure Effects on
Pulmonary Function AX5-57
AX5.3.5.4 Acute and Chronic Exposure Effects on
Airway Responsiveness AX5-57
AX5.2.6 Genotoxicity Potential of Ozone AX5-67
AX5.3 SYSTEMIC EFFECTS OF OZONE EXPOSURE AX5-68
AX5.3.1 Neurobehavional Effects AX5-69
AX5.3.2 Neuroendocrine Effects AX5-77
AX5.3.3 Cardiovascular Effects AX5-78
AX5.3.4 Reproductive and Developmental Effects AX5-79
AX5.3.5 Effects on the Liver, Spleen, and Thymus AX5-80
AX5.3.6 Ozone Effects on Cutaneous and Ocular Tissues AX5-81
AX5.4 INTERACTIONS OF OZONE WITH OTHER CO-OCCURRING
POLLUTANTS AX5-82
AX5.4.1 Ozone and Nitrogen Oxides AX5-83
AX5.4.2 Ozone and Other Copollutants AX5-85
AX5.4.3 Complex (Multicomponent) Mixtures Containing Ozone .... AX5-90
AX5.6 EFFECTS OF OTHER PHOTOCHEMICAL OXIDANTS AX5-96
REFERENCES AX5-98
6. CONTROLLED HUMAN EXPOSURE STUDIES OF OZONE AND
RELATED PHOTOCHEMICAL OXIDANTS 6-1
6.1 INTRODUCTION 6-1
6.2 PULMONARY FUNCTION EFFECTS OF OZONE EXPOSURE
IN HEALTHY SUBJECTS 6-3
6.2.2 Acute Exposure for Up to 2 h 6-3
6.2.3 Prolonged Ozone Exposures 6-5
6.2.3.1 Effect of Exercise Ventilation Rate on FEVj
Response to 6.6 h Ozone Exposure 6-6
6.2.3.2 Exercise Ventilation Rate as a Function of
Body/Lung Size on FEV] Response to 6.6 h
Ozone Exposure 6-6
6.2.3.3 Comparison of 2 h IE to 6.6 h O3 Exposure
Effects on Pulmonary Function 6-7
6.2.4 Triangular Ozone Exposures 6-7
6.2.5 Mechanisms of Pulmonary Function Responses 6-8
6.2.5.1 Pathophysiologic Mechanisms 6-10
6.2.5.2 Mechanisms at a Cellular and Molecular Level ... 6-12
Il-xi
-------
Table of Contents
(cont'd)
Pas
6.3 SUBJECTS WITH PREEXISTING DISEASE 6-14
6.3.1 Subjects with Chronic Obstructive Pulmonary Disease 6-14
6.3.2 Subjects with Asthma 6-14
6.3.3 Subjects with Allergic Rhinitis 6-16
6.3.4 Subjects with Cardiovascular Disease 6-18
6.4 INTERSUBJECT VARIABILITY AND REPRODUCIBILITY
OF RESPONSE 6-18
6.5 FACTORS MODIFYING RESPONSIVENESS TO OZONE 6-20
6.5.1 Influence of Age 6-20
6.5.2 Gender and Hormonal Influences 6-21
6.5.3 Racial, Ethnic and Socioeconomic Status Factors 6-22
6.5.4 Influence of Physical Activity 6-22
6.5.5 Environmental Factors 6-23
6.5.6 Oxidant-Antioxidant Balance 6-24
6.5.7 Genetic and Other Factors 6-24
6.6 REPEATED O3 EXPOSURE EFFECTS 6-25
6.7 EFFECTS ONEXERCISEPERFORMANCE 6-25
6.8 EFFECTS ON AIRWAY RESPONSIVENESS 6-26
6.9 INFLAMMATION AND HOST DEFENSE EFFECTS 6-27
6.9.1 Introduction 6-27
6.9.2 Inflammatory Response in the Upper Respiratory Tract 6-28
6.9.3 Inflammatory Response in the Lower Respiratory Tract 6-29
6.9.4 Effects of Repeated Exposures and Adaptation of Responses ... 6-34
6.9.5 Effect of Anti-Inflammatory and other Mitigating Agents 6-36
6.9.6 Changes in Host Defense Capability Following
Ozone Exposures 6-36
6.10 EXTRAPULMONARY EFFECTS OF OZONE 6-38
6.11 EFFECTS OF OZONE MIXED WITH OTHER POLLUTANTS 6-38
6.12 CONTROLLED STUDIES OF AMBIENT AIR EXPOSURES 6-39
6.12.1 Mobile Laboratory Studies 6-40
6.12.2 Aircraft Cabin Studies 6-40
6.13 SUMMARY 6-41
REFERENCES 6-43
AX6. CONTROLLED HUMAN EXPOSURE STUDIES OF OZONE AND
RELATED PHOTOCHEMICAL OXIDANTS AX6-1
AX6.1 INTRODUCTION AX6-1
AX6.2 PULMONARY FUNCTION EFFECTS OF OZONE EXPOSURE
IN HEALTHY SUBJECTS AX6-2
AX6.2.1 Introduction AX6-2
AX6.2.2 Acute Ozone Exposures for Up to 2 Hours AX6-3
AX6.2.3 Prolonged Ozone Exposures AX6-8
-------
Table of Contents
(cont'd)
Page
AX6.2.3.1 Effect of Exercise Ventilation Rate on
FEVj Response to 6.6 h Ozone Exposure .... AX6-11
AX6.2.3.2 Exercise Ventilation Rate as a Function of
Body/Lung Size on FEVj Response to 6.6 h
Ozone Exposure AX6-12
AX6.2.3.3 Comparison of 6.6 h Ozone Exposure
Pulmonary Responses to Those Observed in
2 h Intermittent Exercise Ozone Exposures . . . AX6-14
AX6.2.4 Triangular Ozone Exposures AX6-15
AX6.2.5 Mechanisms of Pulmonary Function Responses AX6-18
AX6.2.5.1 Pathophysiologic Mechanisms AX6-19
AX6.2.5.2 Mechanisms at a Cellular and Molecular
Level AX6-25
AX6.3 PULMONARY FUNCTION EFFECTS OF OZONE EXPOSURE
IN SUBJECTS WITH PREEXISTING DISEASE AX6-26
AX6.3.1 Subjects with Chronic Obstructive Pulmonary Disease AX6-26
AX6.3.2 Subjects with Asthma AX6-31
AX6.3.3 Subjects with Allergic Rhinitis AX6-35
AX6.3.4 Subjects with Cardiovascular Disease AX6-37
AX6.4 INTERSUBJECT VARIABILITY AND REPRODUCIBILITY
OF RESPONSE AX6-38
AX6.5 INFLUENCE OF AGE, GENDER, ETHNIC, ENVIRONMENTAL
AND OTHER FACTORS AX6-42
AX6.5.1 Influence of Age AX6-42
AX6.5.2 Gender and Hormonal Influences AX6-48
AX6.5.3 Racial, Ethnic and Socioeconomic Status Factors AX6-55
AX6.5.4 Influence of Physical Activity AX6-56
AX6.5.5 Environmental Factors AX6-57
AX6.5.6 Oxidant-Antioxidant Balance AX6-62
AX6.5.7 Genetic and Other Factors AX6-65
AX6.6 REPEATED EXPOSURES TO OZONE AX6-66
AX6.7 EFFECTS ON EXERCISE PERFORMANCE AX6-78
AX6.7.1 Introduction AX6-78
AX6.7.2 Effect on Maximal Oxygen Uptake AX6-78
AX6.7.3 Effect on Endurance Exercise Performance AX6-80
AX6.8 EFFECTS ON AIRWAY RESPONSIVENESS AX6-82
AX6.9 EFFECTS ON INFLAMMATION AND HOST DEFENSE AX6-93
AX6.9.1 Introduction AX6-93
AX6.9.2 Inflammatory Responses in the Upper Respiratory Tract ... AX6-108
AX6.9.3 Inflammatory Responses in the Lower Respiratory Tract ... AX6-110
AX6.9.4 Effects of Repeated Exposures and Adaptation
of Responses AX6-115
AX6.9.5 Effect of Anti-Inflammatory and other Mitigating Agents .. AX6-117
-------
Table of Contents
(cont'd)
Pas
AX6.9.6 Changes in Host Defense Capability Following
Ozone Exposure AX6-118
AX6.10 EXTRAPULMONARY EFFECTS OF OZONE AX6-121
AX6.11 OZONE MIXED WITH OTHER POLLUTANTS AX6-122
AX6.11.1 Ozone and Sulfur Oxides AX6-122
AX6.11.2 Ozone and Nitrogen-Containing Pollutants AX6-128
AX6.11.3 Ozone and Other Pollutant Mixtures Including
Particulate Matter AX6-130
AX6.12 CONTROLLED STUDIES OF AMBIENT AIR EXPOSURES AX6-131
AX6.12.1 Mobile Laboratory Studies AX6-131
AX6.12.2 Aircraft Cabin Studies AX6-132
REFERENCES AX6-135
7. EPIDEMIOLOGICAL STUDIES OF HUMAN HEALTH EFFECTS
ASSOCIATED WITH AMBIENT OZONE EXPOSURE 7-1
7.1 INTRODUCTION 7-1
7.1.1 Approach to Identifying O3 Epidemiologic Studies 7-2
7.1.2 Approach to Assessing Epidemiologic Evidence 7-2
7.1.3 Study Designs and Analysis Methods Used to Assess
O3 Health Effects 7-6
7.1.3.1 Exposure Assessment in Epidemiologic Studies ... 7-6
7.1.3.2 O3 Exposure Indices Used 7-7
7.1.3.3 Lag of O3 Exposure Used 7-8
7.1.3.4 Model Specification Issues 7-9
7.1.3.5 Controlling for Temporal Trends and
Meteorologic Effects 7-10
7.1.3.6 Confounding Effects of Copollutants 7-12
7.1.3.7 Model Uncertainty and Multiple Testing 7-13
7.1.3.8 Impact of GAM Convergence Issue on O3
Risk Estimates 7-14
7.1.4 Approach to Presenting O3 Epidemiologic Evidence 7-16
7.2 FIELD STUDIES ADDRESSING ACUTE EFFECTS OF OZONE 7-17
7.2.1 Summary of Key Findings on Field Studies of Acute Effects
From the 1996 O3 AQCD 7-17
7.2.2 Introduction to Recent Field Studies of Acute O3 Effects 7-18
7.2.3 Acute O3 Exposure and Lung Function 7-19
7.2.3.1 Acute O3 Studies with Spirometry (FEVj) 7-20
7.2.3.2 Acute O3 Studies of PEF 7-28
7.2.4 Respiratory Symptoms 7-35
7.2.5 Acute Airway Inflammation 7-41
7.2.6 Acute O3 Exposure and School Absences 7-43
7.2.7 Cardiac Physiologic Endpoints 7-44
7.2.8 Summary of Field Studies Assessing Acute O3 Effects 7-45
-------
Table of Contents
(cont'd)
Page
7.3 EFFECTS OF OZONE ON DAILY EMERGENCY DEPARTMENT
VISITS AND HOSPITAL ADMISSIONS 7-46
7.3.1 Summary of Key Findings on Studies of Emergency
Department Visits and Hospital Admissions from the
1996 O3 AQCD 7-46
7.3.2 Review of Recent Studies of Emergency Department Visits
for Respiratory Diseases 7-47
7.3.3 Studies of Hospital Admissions for Respiratory Diseases 7-51
7.3.4 Association of O3 with Hospital Admissions for
Cardiovascular Disease 7-58
7.3.5 Summary of Acute O3 Effects on Daily Emergency
Department Visits and Hospital Admissions 7-58
7.4 ACUTE EFFECTS OF OZONE ON MORTALITY 7-59
7.4.1 Summary of Key Findings on Acute Effects of O3 on
Mortality From the 1996 O3 AQCD 7-59
7.4.2 Introduction to Assessment of Current O3-Mortality Studies .... 7-60
7.4.3 Single-Pollutant Model O3-Mortality Risk Estimates 7-61
7.4.4 Seasonal Variation in O3-Mortality Risk Estimates 7-65
7.4.5 O3-Mortality Risk Estimates Adjusting for PM Exposure 7-67
7.4.6 O3 Risk Estimates for Specific Causes of Mortality 7-71
7.4.7 O3-Mortality Risk Estimates for Specific Subpopulations 7-72
7.4.8 Summary of Acute O3 Effects on Mortality 7-74
7.5 CHRONIC EFFECTS OF OZONE 7-75
7.5.1 Summary of Key Findings on Studies of Health Effects and
Chronic O3 Exposure from the 1996 O3 AQCD 7-75
7.5.2 Introduction to Morbidity Effects of Chronic O3 Exposure 7-75
7.5.3 Seasonal O3 Effects on Lung Function 7-76
7.5.4 Chronic O3 Effects on Lung Function 7-78
7.5.5 Chronic O3 Exposure and Respiratory Inflammation 7-81
7.5.6 Risk of Asthma Development 7-82
7.5.7 Respiratory Effects of Chronic O3 Exposure on Susceptible
Populations 7-84
7.5.8 Mortality Effects of Chronic O3 Exposure 7-85
7.5.9 Summary of Chronic O3 Effects on Morbidity and Mortality .... 7-87
7.6 INTERPRETATIVE ASSESSMENT OF THE EVIDENCE IN
EPIDEMIOLOGIC STUDIES OF OZONE HEALTH EFFECTS 7-87
7.6.1 Introduction 7-87
7.6.2 Exposure Assessment 7-88
7.6.2.1 Relationship between Ambient Concentrations
and Personal Exposure to O3 7-89
7.6.2.2 Factors Affecting the Relationship between
Ambient Concentrations and Personal Exposures
to O3 7-90
II-xv
-------
Table of Contents
(cont'd)
Page
7.6.2.3 Assessing Chronic Exposure to O3 7-92
7.6.3 O3 Exposure Indices 7-92
7.6.4 Lag Time: Period between O3 Exposure and Observed
Health Effect 7-94
7.6.5 Confounding by Temporal Trends and Meteorologic Effects .... 7-99
7.6.5.1 Assessment of O3 Effects after Adjusting for
Temporal Trends and Meteorologic Effects 7-99
7.6.5.2 Importance of Season-Specific Estimates of
O3 Health Effects 7-103
7.6.6 Assessment of Confounding by Copollutants 7-110
7.6.6.1 Relationship between Personal Exposure to
O3 and Copollutants 7-111
7.6.6.2 Assessment of Confounding Using
Multipollutant Regression Models 7-112
7.6.7 Issues of Model Uncertainty and Multiple Hypothesis Testing . 7-116
7.6.8 Concentration-Response Function and Threshold 7-118
7.6.9 Spatial Variability in O3 Effect 7-122
7.6.10 Health Effects of O3 in Susceptible Populations 7-124
7.6.10.1 Health Effects Associated with Ambient
O3 Exposure in Asthmatics 7-125
7.6.10.2 Age-Related Differences in O3 Effects 7-128
7.6.11 Summary of Key Findings and Conclusions Derived From
O3 Epidemiologic Studies 7-130
REFERENCES 7-134
AX7. STUDIES OF HUMAN HEALTH EFFECTS ASSOCIATED WITH
AMBIENT OZONE EXPOSURE AX7-1
8. INTEGRATIVE SYNTHESIS 8-1
8.1 INTRODUCTION 8-1
8.1.1 Chapter Organization 8-2
8.1.2 Current Standards 8-3
8.2 TRENDS IN UNITED STATES OZONE AIR QUALITY 8-4
8.2.1 Ozone Concentrations, Patterns 8-4
8.2.2 Seasonal Variations 8-6
8.2.3 Long-Term Trends 8-7
8.2.4 Ozone Interactions with Other Ambient Pollutants 8-7
8.3 AMBIENT OZONE EXPOSURE ASSESSMENTS 8-8
8.3.1 Personal Exposure 8-8
8.3.2 Indoor Concentrations 8-8
8.4 SYNTHESIS OF AVAILABLE INFORMATION ON OZONE-
RELATED HEALTH EFFECTS 8-9
8.4.1 Assessment of Epidemiological Evidence 8-10
8.4.2 Strength of Epidemiological Associations 8-11
-------
Table of Contents
(cont'd)
Page
8.4.3 Acute Exposure Studies 8-12
8.4.3.1 Panel Studies 8-12
8.4.3.2 Asthma Panels 8-13
8.4.3.3 School Absences 8-14
8.4.3.4 Field Studies on Cardiovascular Effects 8-15
8.4.4 Emergency Department Visits and Hospital Admissions 8-15
8.4.5 Acute Effects of Ozone on Mortality 8-16
8.4.6 Chronic Ozone Exposure Studies 8-20
8.4.7 Robustness of Epidemiological Associations 8-21
8.4.7.1 Exposure Issues: Ambient versus Personal 8-21
8.4.7.2 Confounding by Temporal Trends and
Meteorologic Effects 8-22
8.4.7.3 Assessment of Confounding by Copollutants .... 8-22
8.4.7.4 Lag Period between Ozone Exposure and
Health Response 8-23
8.4.7.5 Concentration-Response Functions and
Threshold 8-24
8.4.7.6 Summary and Conclusions for Epidemiology
Findings 8-24
8.4.8 Integration of Experimental and Epidemiologic Evidence 8-25
8.4.8.1 Background on Cross-Cutting Issues 8-27
8.4.8.2 Approaches to Experimental Evaluation of
Ozone Health Effects 8-27
8.4.8.3 Interspecies Comparison of Experimental
Effects — Dosimetry Considerations 8-28
8.4.8.4 Integrated Critical Analysis of Physiological,
Biochemical Effects 8-29
8.4.9 Preexisting Disease as a Potential Risk Factor 8-37
8.4.10 Biological Plausibility and Coherence of Evidence for
Adverse Respiratory Health Effects 8-38
8.4.10.1 Pulmonary Function 8-39
8.4.10.2 Lung Injury, Inflammation, and Host Defense .... 8-42
8.4.11 Coherence Between Epidemiological and Experimental
Evidence for Respiratory Health Effects 8-44
8.4.12 Summary and Conclusions for Ozone Health Effects 8-45
REFERENCES 8-47
_I_\J-J_L J-j_l_VJ-j_l_ >l v_xJ—'k-J OT^/
-------
List of Tables
Number Page
AX4-1 New Experimental Human Studies on Ozone Dosimetry AX4-4
AX4-2 New Ozone Dosimetry Model Investigations AX4-12
AX5-1 Effects of Ozone on Lung Monooxygenases AX5-4
AX5-2 Effects of Ozone on Lung Host Defenses AX5-10
AX5-3 Effects of Ozone on Lung Permeability and Inflammation AX5-26
AX5-4 Inbred Mouse Strain Susceptibility AX5-37
AX5-5 Effects of Ozone on Lung Structure: Short-Term Exposures AX5-44
AX5-6 Effects of Ozone on Lung Structure: Long-Term Exposures AX5-50
AX5-7 Effects of Ozone on Airway Responsiveness AX5-61
AX5-8 Systemic Effects of Ozone AX5-70
AX5-9 Interactions of Ozone With Nitrogen Dioxide AX5-84
AX5-10 Interactions Of Ozone With Particles AX5-86
AX6-1 Controlled Exposure of Healthy Humans to Ozone for 1 to 2 Hours
during Exercise AX6-5
AX6-2 Pulmonary Function Effects after Prolonged Exposures to Ozone AX6-9
AX6-3 Ozone Exposure in Subjects with Preexisting Disease AX6-27
AX6-4 Classification of Asthma Severity AX6-34
AX6-5 Age Differences in Pulmonary Function Responses to Ozone AX6-43
AX6-6 Gender and Hormonal Differences in Pulmonary Function Responses to Ozone AX6-49
AX6-7 Influence of Ethnic, Environmental, and Other Factors AX6-59
AX6-8 Changes in Forced Expiratory Volume in One Second After Repeated Daily
Exposure to Ozone AX6-67
AX6-9 Pulmonary Function Effects with Repeated Exposures to Ozone AX6-69
-------
List of Tables
(cont'd)
Number Page
AX6-10 Ozone Effects on Exercise Performance AX6-79
AX6-11 Airway Responsiveness Following Ozone Exposures AX6-84
AX6-12 Studies of Respiratory Tract Inflammatory Effects from Controlled
Human Exposure to Ozone AX6-95
AX6-13 Studies of Effects on Host Defense, on Drug Effects and Supportive
In Vitro Studies Relating to Controlled Human Exposure to Ozone AX6-102
AX6-14 Ozone Mixed with Other Pollutants AX6-123
AX6-15 Acute Effects of Ozone in Ambient Air in Field Studies With a
Mobile Laboratory AX6-133
7-la Field Studies that Investigated the Association between Acute Ambient
O3 Exposure and Changes in FEV] 7-20
7-lb Changes in FEVj (95% CI) Associated with Acute Ambient O3 Exposures,
Ordered by Size of the Estimate 7-21
7-lc Cross-day Changes in FEVj Associated with Acute Ambient O3 Exposures,
Ordered by Size of the Estimate 7-23
7-2 Estimated O3 External Doses and Changes in FEVj Associated with Acute
Ambient O3 Exposures in Outdoor Workers and Athletes 7-26
AX7-1 Effects of Acute O3 Exposure on Lung Function and Respiratory Symptoms
in Field Studies AX7-2
AX7-2 Effects of Acute O3 Exposure on Cardiovascular Outcomes in
Field Studies AX7-24
AX7-3 Effects of O3 on Daily Emergency Department Visits AX7-28
AX7-4 Effects of O3 on Daily Hospital Admissions AX7-36
AX7-5 Effects of Acute O3 Exposure on Mortality AX7-46
AX7-6 Effects of Chronic O3 Exposure on Respiratory Health AX7-64
AX7-7 Effects of Chronic O3 Exposure on Mortality AX7-78
-------
List of Tables
(cont'd)
Number Page
8-1 Current National Ambient Air Quality Standards (NAAQS) in the
United States 8-3
8-2 Gradation of Individual Responses to Short-Term Ozone Exposure in
Healthy Persons 8-4
8-3 Gradation of Individual Responses to Short-Term Ozone Exposure in
Persons with Impaired Respiratory Systems 8-5
II-xx
-------
List of Figures
Number Page
AX5-1 Major secondary products of ozone interaction with lung cells AX5-2
6-1 Predicted O3-induced decrements in FEVj as a function of exposure
duration and O3 concentration in young healthy adults (20 yrs of age)
during moderate IE (VE = 30 L/min) 6-5
6-2 Recovery of FEVj responses following a 2 h exposure to 0.4 ppm O3 with IE ... 6-10
6-3 Predicted O3-induced decrements in FEV] as a function of exposure duration
and level of IE (line labels are VE levels) in young healthy adults (20 yrs of
age) exposed to 0.3 ppm O3 6-22
AX6-1 FEV] decrements as a function of O3 concentration following a 2 h exposure
with incremental exercise (15 min intervals) or rest AX6-7
AX6-2 Average FEVj decrements (±SE) for prolonged 6.6 h exposures to 0.12 ppm
O3 as a function of exercise VE AX6-13
AX6-3 The forced expiratory volume in 1 s (FEVj) is shown in relation to exposure
duration (hours) under three exposure conditions AX6-16
AX6-4a,b Recovery of spirometric responses following a 2 h exposure to 0.4 ppm
O3 with IE AX6-20
AX6-5 Plot of the mean FEV] (% baseline) vs. time for ozone exposed cohorts AX6-23
AX6-6 Frequency distributions of percent decrements in FEVj for 6.6-h exposure
to four concentrations of ozone AX6-39
AX6-7 Effect of O3 exposure (0.42 ppm for 1.5 h with IE) on FEV] as a function
of subject age AX6-47
AX6-8 Regression curves were fitted to day-by-day postexposure FEV] values
obtained after repeated daily acute exposures to O3 for 2 to 3 h with
intermittent exercise at a VE of 24 to 43 L/min (adaptation studies) AX6-77
7-la Percent change (95% CI) in morning PEF in children per 40 ppb increase
in 1-h max O3 or equivalent, arranged by size of the effect estimate 7-29
7-lb Percent change (95% CI) in afternoon PEF in children per 40 ppb increase
in 1-h max O3 or equivalent, arranged by size of the effect estimate 7-30
7-2 Density curves of the % change in PEF per 30 ppb increase in 8-h avg O3
with a cumulative lag of 1 to 5 days for the individual eight NCICAS cities
and the pooled average of all cities 7-32
-------
List of Figures
Number Page
7-3 Density curves of the odds ratios for the prevalence of cough among
asthmatic children 7-36
7-4 Density curves of the odds ratios for the prevalence of extra bronchodilator
use among asthmatic children 7-37
7-5 Density curves of the odds ratios for the incidence of symptoms per 30 ppb
increase in 8-h avg O3 with a cumulative lag of 1 to 4 days for the individual
eight cities and the pooled average of all cities 7-38
7-6 Ozone-associated % change (95% CI) in emergency department visits for
asthma per 40 ppb increase in 1-h max O3 or equivalent 7-48
7-7 Ozone-associated % change (95% CI) in total respiratory hospitalizations
for all year analyses per 40 ppb increase in 1-h max O3 or equivalent 7-52
7-8 Ozone-associated % change (95% CI) in total respiratory hospitalizations
by season per 40 ppb increase in 1-h max O3 or equivalent 7-53
7-9 Ozone-associated % change (95% CI) in total respiratory hospitalizations with
adjustment for PM indices per 40 ppb increase in 1-h max O3 or equivalent 7-54
7-10 All cause (nonaccidental) O3 excess mortality risk estimates (95% CI) for all year
analyses per 40 ppb increase in 1-h max O3 or equivalent 7-62
7-11 Bayesian city-specific and national average estimates for the % change
(95% CI) in daily mortality per 10 ppb increase in 24-h avg O3 in the previous
week using a constrained distributed lag model for 95 U.S. communities
(NMMAPS), arranged by size of the effect estimate 7-64
7-12 All cause (nonaccidental) O3 excess mortality risk estimates (95% CI) by
season per 40 ppb increase in 1-h max O3 or equivalent 7-66
7-13 All cause (nonaccidental) O3 excess mortality risk estimates (95% CI) with
adjustment for PM indices for all year analyses per 40 ppb increase in 1-h max
O3 or equivalent 7-69
7-14 All cause (nonaccidental) O3 excess mortality risk estimates (95% CI)
with adjustment for PM indices by season per 40 ppb increase in 1-h max
O3 or equivalent 7-70
7-15 Comparison of single-day lags (0-, 1-, 2-, and 3-day) to a cumulative multiday
lag (0- to 6-day) for % changes in all cause mortality per 20 ppb increase in
24-h avg O3 in all ages 7-96
-------
List of Figures
Number Page
7-16 Comparison of single-day lags (0-, 1-, 2-, 3-, 4-, and 5-day) to a cumulative
multiday lag (0- to 4-day) for % changes in total respiratory hospitalizations
per 40 ppb increase in 1-h max O3 in children less than two years of age 7-97
7-17 Comparison of single-day lags (1-, 2-, 3-, 4-, 5-, and 6-day) to a cumulative
multiday lag (1- to 5-day) for % changes in PEF per 30 ppb increase in 8-h
avg O3 in urban children 7-98
7-18 Daily nonaccidental mortality in Montreal, Canada as a function of mean
temperature, using natural splines with two degrees of freedom 7-100
7-19 Scatterplots of daily levels of O3 with TSP and SO2 in Philadelphia, PA
by season 7-104
7-20 Estimated total (nonaccidental) mortality relative risk per 100 ppb increase in
24-h avg O3 of reach seasonal data set 7-105
7-21 Marginal posterior distributions of the national average estimates of O3
effects on total mortality per 10 ppb increase in 24-h avg O3 at a 1-day lag
for all year, summer (June-August), and winter (December-February)
analyses in 90 U.S. cities 7-106
7-22 Summary density curves of the % change in all cause mortality for all year
data and by season per 40 ppb increase in 1-h max O3 or equivalent 7-108
7-23 Summary density curves of the % change in total respiratory hospital
admissions for all year data and by season per 40 ppb increase in 1-h
max O3 or equivalent 7-109
7-24 Posterior means and 95% posterior intervals of the national average estimate
of PM10 effects on total mortality from non-external causes per 10 ^ig/m3
increase in 24-h avg PM10 at a 1-day lag within sets of 90 U.S. cities with
pollutant data available 7-113
7-25 Posterior means and 95% posterior intervals of the national average
estimate of O3 effects on total mortality from non-external causes per
10 ppb increase in 24-h avg O3 at 0-, 1-, and 2-day lags within sets of
90 U.S. cities with pollutant data available 7-113
7-26 Maximum likelihood estimates of O3-mortality for 95 U.S. communities,
determined using a constrained distributed lag model for lags 0 through 6 days . 7-114
8-1 Ozone-associated percent change (95% CI) in emergency room visits
for asthma 8-16
-------
List of Figures
Number Page
8-2 Ozone associated percent change (95% CI) in total respiratory
hospitalizations (95% CI) for all year and for by season 8-17
8-3 All cause (nonaccidental) O3 excess mortality risk estimates (95% CI)
for all year and for by season per 40 ppb increase in 1-h maximum O3
or equivalent 8-18
8-4
All cause (nonaccidental) O3 excess mortality risk estimates (95% CI) with
adjustment for PM indices for all analyses
8-19
II-xxiv
-------
Authors, Contributors, and Reviewers
To be inserted in Second External Review draft.
II-xxv
-------
U.S. ENVIRONMENTAL PROTECTION AGENCY
PROJECT TEAM FOR DEVELOPMENT OF AIR QUALITY CRITERIA
FOR OZONE
To be inserted in Second External Review draft.
II-xxvi
-------
U.S. ENVIRONMENTAL PROTECTION AGENCY
SCIENCE ADVISORY BOARD (SAB) STAFF OFFICE
CLEAN AIR SCIENTIFIC ADVISORY COMMITTEE (CASAC) OZONE
REVIEW PANEL
To be inserted in Second External Review draft.
II-xxvii
-------
Abbreviations and Acronyms
AA
ascorbic acid
ACh
acetylcholine
ADs
alveolar entrance rings
ADSS
aged and diluted sidestream cigarette smoke
AED
aerodynamic diameter
AER
air exchange rate
AF
adsorbed fraction
AH9
ascorbic acid
AHR
airway hyperreactivity
AHSMOG
ALI
Adventist Health Study on Smog
AIRS Aerometric Information Retrieval System (U.S. Environmental Protection
Agency)
air-liquid interface
AM
alveolar macrophage
A,,
cros-sectional area of peripheral lung
AP
alkaline phosphatase
APHEA
Air Pollution on Health: European Approach
AQCD
Air Quality Criteria Document
ATR
atrial natriuretic factor
BAL
bronchoalveolar lavage
BALF
bronchoalveolar lavage fluid
BALT
bronchus-associated lymphoid tissue
B[a]P
benzo[a]pyrene
BHR
bronchial hyperresponsiveness
b.i.d.
twice a day
BMZ
basement membrane zone
BP
blood pressure
BrdU
bromodeoxyuridine
-------
BS
BSA
C
CxT
C3a
CAPs
CAR
CARB
CAT
CC16
CCh
CCSP
Cdyn
CE
CFD
CHO
CI
CINC
CIU, CBU
CMD
CO
C02
ConA
COPD
cw
Cyt
A
DD
DHBA
black smoke
body surface area
concentration
concentration times duration of exposure
complement protein fragment
concentrated ambient particles
centriacinar region
California Air Resources Board
cell antioxidant capacity
Clara cell secretory protein
carbachol
Clara cell secretory protein
dynamic lung compliance
continuous exercise
computational fluid dynamics
Chinese hamster ovary
confidence interval
cytokine-induced neutrophil chemoattractant
cumulative inhalation unit
count mean diameter
carbon monoxide
carbon dioxide
concanavalin A
chronic obstructive pulmonary disease
chest wall compliance
cytochrome
mean change in a variable
doubling dose
2,3 -deny droxybenzoic acid
II-xxix
-------
DNA
deoxyribonucleic acid
DPPC
dipalmitoylglycero-3-phosphocholine
DR
disulfide reductase
convergence precision
EEG
electroencephalographic
ELF
epithelial lining fluid
EM
electron microscopy
ENA
epithelial cell-derived neutrophil-activating peptide
EPA
U.S. Environmental Protection Agency
ETS
environmental tobacco smoke
EU
frequency of breathing
female
F344
Fischer 344
FA
filtered air
FA
fatty acid
FA
fractional absorption; absorbed fraction
FAA
Federal Aviation Administration
FEF
forced expiratory flow
FEF9
25
forced expiratory flow after 25% vital capacity
FEF9
25-75
forced expiratory flow between 25 and 75% of vital capacity
FEF,
forced expiratory flow after 50% vital capacity
FEF,
6oP
FEF75
forced expiratory flow after 75% vital capacity
FEV,
0.75
forced expiratory volume in 0.75 s
FEV,
forced expiratory volume in 1 s
FFA
free fatty acid
FGF
FGFR
II-XXX
-------
FIVC
forced inspiratory vital capacity
Fn
fibronectin
FP
fluticasone propionate
FRC
functional residual capacity
FS
field stimulation
FVC
forced vital capacity
GAM
General Additive Model
GDI
glutathione-disulfide transhydrogenase
GEE
(model)
GLM
General Linear Model
GM-CSF
granulocyte-macrophage colony stimulating factor
G6PD
glucose-6-phosphate dehydrogenase
GR
glutathione reductase
GSH
glutathione
GSHPx, GPx
glutathione peroxidase
GST
glutathione-^-transferase
hydrogen ion
H2CO, HCHO formaldehyde
HDMA
house dust mite allergen
hydrogen peroxide
H2S04
sulfuric acid
HEI
Health Effects Institute
HHP-C9
1 -hy droxy-1 -hy droperoxynonane
HIST
histamine
HLA
human lymphocyte antigen
HNE
4-hydroxynonenal
HR
heart rate
HSP
heat shock protein
HSV
herpes simplex virus
II-xxxi
-------
5-HT
IAS
1C
ICAM
ICRP
ICS
IE
Ig
IL
IN
INF
inh
iNOS
5-hydroxytryptamine
interalveolar septum
inspiratory capacity
intracellular adhesion molecule
International Commission on Radiological Protection
inhaled steroids
intermittent exercise
immunoglobulin (IgA, IgE, IgG, IgM)
interleukin (IL-1, IL-6, IL-8)
intranasal
interferon
inhalation
inducible nitric oxide synthase
ip intraperitoneal
IT
IU
iv
intratracheal
International Units
intravenous
Ka intrinsic mass transfer coefficient/parameter
Kg mass transfer coefficient for gas phase
KTB
K,
tracheobronchial region overall mass transfer coefficient
mass transfer coefficient for liquid phase
Kj reaction rate constant
A
LDH
LIS
LM
LOESS
LPS
LRT
ozone uptake efficiency
lactate dehydrogenase
lateral intercellular space
light microscopy
locally estimated smoothing splines
lipopolysaccharide
lower respiratory tract
II-xxxii
-------
LT
leukotriene (LTB4, LTC4, LTD4, LTE4)
M
male
M
maximum number of iterations
MCB
monochlorobimane
MCh
methacholine
MCP
monocyte chemotactic protein
MDA
malondi aldehyde
MHC
major histocompatibility
MIP
macrophage inflammatory protein
MLN
mediastinal lymph node
MMAD
mass median aerodynamic diameter
mRNA
messenger ribonucleic acid
MSA
metropolitan statistical area
Mt
metallothionein
n,N
number
NAAQS
National Ambient Air Quality Standards
NADH
reduced nicotinamide adenine dinucleotide
NADPH
NAG
reduced nicotinamide adenine dinucleotide phosphate
NADPH-CR cytochrome c reductase pertaining to nicotinamide adenine dinucleotide
phosphate activity
N-acetyl-p-t/-glucosamine
NB-KB
NCICAS
nuclear factor kappa B
NCEA-RTP National Center for Environmental Assessment Division in Research Triangle
Park, NC
National Cooperative Inner-City Asthma Study
NHBE
cultured human bronchial epithelial (cells)
(NH4)2S04
ammonium sulfate
Nffl
National Institutes of Health
NIST
National Institute of Standards and Technology
NK
natural killer (cells)
-------
NL
NLF
NMMAPS
NO
NO2
NOX
NOS
NS
NSAID
NSBR
NTP
NTS
02
03
OAQPS
8-OHdG
OVA
P
PAF
PAN
PAR
PBPK
PC20
PC-ALF
PCI
PD
PDioo
nasal lavage
nasal lavage fluid
National Morbidity, Mortality and Air Pollution Study
nitric oxide
nitrogen dioxide
nitrogen oxides
nitric oxide synthase
nonsmoker
non-steroidal anti-inflammatory agent
nonspecific bronchial respnsiveness
National Toxicology Program
nucleus tractus solitarius
superoxide
ozone
Office of Air Quality Planning and Standards
8-hydroxy-2 '-deoxyguanosine
ovalbumin
probability
platelet-activating factor
peroxyacetyl nitrate
proximal alveolar region
physiologically based pharmacokinetic
provocative concentration that produces a 20% decrease in forced expiratory
volume in 1 s
l-palmitoyl-2-(9-oxononanoyl)-sn-glycero-3-phosphocholine
picryl chloride
potential difference
provocative dose that produces a 100% decrease in forced expiratory volume
in 1 s
II-xxxiv
-------
PD20 provocative dose that produces a 20% decrease in forced expiratory volume in
Is
PE
PEF
"enh
PG
6PGD
PGP
pH
PHA
PIF
PM
PM10
PM15
PM25
PMN
PND
PPN
PSA
PUFA
PUL
PWM
QCE
r
R
9
r
R2
rALP
RANTES
Raw
postexposure
peak expiratory flow
enhanced pause
prostaglandm (PGD2, PGE, PGE1; PGE2 PGFla, PGF2a)
6-phosphogluconate dehydrogenase
protein gene product
hydrogen ion concentration
phytohemagglutinin
peak inspiratory flow
particulate matter
particulate matter of mass median aerodynamic diameter < 10 j^m
particulate matter of mass median aerodynamic diameter < 1 5 j^m
particulate matter of mass median aerodynamic diameter < 2.5 j^m
poly morphonucl ear neutrolphil leukocyte (also called neutrophil)
post natal day
peroxypropionyl nitrate
picryl sulfonic acid
polyunsaturated fatty acid
pulmonary
pokeweed mitogen
quasi continuous exercise
linear regression correlation coefficient
intraclass correlation coefficient
correlation coefficient
multiple correlation coefficient
recombinant antileukoprotease
regulated on activation, normal T cell-expressed and -secreted
airway resistance
II-XXXV
-------
KB respiratory bronchiole
RER rough endoplasmic reticulum
RFS
RH relative humidity
RL total pulmonary resistance
ROI reactive oxygen intermediate
ROS reactive oxygen species
RT respiratory tract
RT total respiratory resistance
a geometric standard deviation
smoker
SAC Staphylococcus aureus Cowan 1 strain
SAW small airway function
sc subcutaneous
SC stratum corneum
SD, S-D Sprague-Dawley
SD standard deviation
SE standard error
SES socioeconomic status
SGaw specific airway conductance
SNPs single nucleotide polymorphisms
SO2 sulfur dioxide
SO42 sulfate ion
SO42 sulfate
SOA secondary organic aerosol
SOD superoxide dismutase
SP substance P
SP surfactant protein (SP-A, SP-D)
SRaw specific airway resistance
II-xxxvi
-------
SRBC
sheep red blood cell
temperature
time (duration of exposure)
triiodothyronine
thyroxine
TB
tracheobronchial
TEA
thiobarbituric acid
99m-
Tc-DTPA
radiolabeled diethylene triamine pentaacetic acid
T
ico
core body temperature
T
J-
cytotoxic T-lymphocytes
Th
helper T-lymphocyte
TLC
total lung capacity
TLR
Toll-like receptor
TNF
tumor necrosis factor
TNFR
tumor necrosis factor receptor
TSH
thyroid-stimulating hormone
TSP
total suspended particulate
TWA
time-weighted average
TX
thromboxane (A2, B2)
URT
upper respiratory tract
UV
ultraviolet
V
volume
VC
vital capacity
VCAM
VD
dead space
minute ventilation; expired volume per minute
V
Emax
maximum minute ventilation
VT
average inspiratory flow
V
max25%
maximum expiratory flow at 25% of the vital capacity
II-xxxvii
-------
* max50%
maximum expiratory flow at 50% of the vital capacity
'max50%TLC
maximum expiratory flow at 50% of the total lung capacity
V
max75%
maximum expiratory flow at 75% of the vital capacity
VMD
volume mean diameter
V09
oxygen uptake by the body
VO
2max
maximal oxygen uptake (maximal aerobic capacity)
VOCs
volatile organic compounds
VT
tidal volume
VT
tracheal transepithelial potential
' TB
tracheobronchial region volume
Tmax
maximum tidal volume
WT
wild-type
II-xxxviii
-------
i 4. DOSIMETRY, SPECIES HOMOLOGY, SENSITIVITY,
2 AND ANIMAL-TO-HUMAN EXTRAPOLATION
3
4
5 4.1 INTRODUCTION
6 The dosimetry of ozone (O3) in humans has been examined in a series of studies published
7 in the past decade. This important new information further characterizes the dose of O3
8 delivered to various sites in the respiratory tract (RT). Ozone, classified as a reactive gas,
9 interacts with surfactant, antioxidants, and other compounds in the epithelial lining fluid (ELF).
10 Researchers have attempted to obtain a greater understanding of how these complex interactions
11 affect uptake of O3, which eventually link to O3-induced injury. New work has also been
12 completed evaluating species differences in responses to O3 exposures, which allow more
13 accurate quantitative extrapolation from animal to human.
14 This chapter is not intended to be a complete overview of O3 dosimetry and animal-to-
15 human comparisons, but rather, it is an update of the dosimetry/extrapolation chapter from the
16 last O3 criteria document (U.S. Environmental Protection Agency, 1996), or 1996 O3 AQCD, and
17 other reviews of the earlier published literature. The framework for presenting this chapter is
18 first a discussion in Section 4.2 of general concepts of the dosimetry of O3 in the RT. Bolus-
19 response studies are then presented in Section 4.2.1 followed by general uptake studies in
20 Section 4.2.2. Dosimetry modeling is presented in Section 4.2.3 followed by the summary and
21 conclusions for the dosimetry material in Section 4.2.4. The chapter continues in Section 4.3
22 with a discussion of species comparisons and ends with a discussion of animal-to-human
23 extrapolation. More detailed discussions of the studies are presented in the supporting material
24 to this chapter (Annex AX4). The toxicological effects of O3 in laboratory animals and in vitro
25 test systems are discussed in Chapter 5 and direct effects of O3 in humans are discussed in
26 Chapter 6. The historical O3 literature is very briefly summarized in this chapter, providing a
27 very concise overview of previous work. The reader is referred to the 1996 O3 AQCD for more
28 detailed discussion of the literature prior to the early 1990s.
29
January 2005 4-1 DRAFT-DO NOT QUOTE OR CITE
-------
1 4.2 DOSIMETRY OF OZONE IN THE RESPIRATORY TRACT
2 Dosimetry refers to measuring or estimating the quantity of or rate at which a chemical is
3 absorbed by target sites within the RT. The compound most directly responsible for toxic effects
4 may be the inhaled gas O3 or one of its chemical reaction products. Complete identification of
5 the actual toxic agents and their integration into dosimetry is a complex issue that has not been
6 resolved. Thus, most dosimetry investigations are concerned with the dose of the primary
7 inhaled chemical. The actual meaning of the word 'dose' is somewhat subjective. Dahl (1990)
8 classified O3 as a reactive gas and discussed dose in terms of: inhaled O3 concentration; amount
9 of O3 inhaled as determined by minute volume, vapor concentration, and exposure duration;
10 uptake or the amount of O3 retained (i.e., not exhaled); O3 or its active metabolites delivered to
11 target cells or tissues; O3 or its reactive metabolites delivered to target biomolecules or
12 organelles; and O3 or its metabolites participating in the ultimate toxic reactions - the effective
13 dose. Understanding dosimetry as it relates to O3-induced injury is complex due to the fact that
14 O3 interacts primarily with the ELF which contains surfactant and antioxidants. In the upper
15 airways ELF is thick and highly protective against oxidant injury. In lower airways ELF is
16 thinner, has lower levels of antioxidants, and thus, allows more cellular injury. Adding to the
17 complexity is the fact that O3 can react with molecules in the ELF to create even more reactive
18 metabolites, which can then diffuse within the lung or be transported out of the lung to generate
19 systemic effects.
20 Since the 1996 ozone criteria document (U.S. Environmental Protection Agency, 1996),
21 all new experiments have been carried out in humans to obtain direct measurements of absorbed
22 O3 in the RT, the upper RT (URT) region proximal to the tracheal entrance, and in the
23 tracheobronchial (TB) region; no uptake experiments have been identified as being performed
24 using laboratory animals.
25 In vivo dosimetry studies described in the last criteria document estimate that uptake of
26 inhaled O3 in rat is 0.50 of inhaled O3 concentration and of that about half is removed in the
27 head, the just under half in the lungs, and only a small percentage removed in the larynx/trachea.
28 There was no agreement as to whether uptake was dependent on flow. The uptake of O3 in the
29 RT of humans at rest was between 0.8 and 0.95 . Studies in humans have shown that increasing
30 minute ventilation (VE) with exercise (by increasing both breathing frequency and tidal volume)
31 causes only a small decrease in uptake efficiency by the total RT. Mode of breathing had little
January 2005 4-2 DRAFT-DO NOT QUOTE OR CITE
-------
1 effect on uptake; oral breathing had approximately 10% greater uptake efficiency than nasal
2 breathing. Comparing bronchoalveolar lavage (BAL) cells from rat and human, a 0.4 ppm
3 exposure in exercising humans gave 4 to 5 times the dose of O3 (dose retained) than a rat at rest
4 exposed to the same concentration. Overall, uptake efficiency data between humans and other
5 species had a great deal of consistency. Miller (1995) reviewed the major factors influencing RT
6 uptake of O3: (1) structure of the RT region, (2) ventilation, and (3) gas transport mechanisms.
7 In comparing rats and humans, they differ greatly in URT structure, which imparts disparate
8 airflow streams. Rate, depth, and route of breathing all influence the amount of O3 inhaled.
9 However, route of breathing has little biological significance. In humans at exposures to
10 0.5 ppm (resting) or 0.2 ppm (exercising), breathing frequency increases and tidal volume
11 decreases such that minute ventilation is not altered. Miller further noted that local dose is the
12 critical link between exposure and response, and that modeling of O3 has typically discounted the
13 mucociliary layer of the URT and tracheobronchial regions.
14 In vitro dosimetry studies using isolated lung preparations showed that uptake efficiency is
15 chemical-reaction dependent, indicating the importance of reaction product formation. These
16 reaction products, created mainly by the ozonolysis of polyunsaturated fatty acids (PUFA) ,
17 included hydrogen peroxide, aldehydes, and hydroxyhydroperoxides, which are mediators of O3
18 toxicity. Other products are created by the reaction of O3 with other ELF constituents, all of
19 which must be considered in understanding the dosimetry of O3.
20 The next two sections review the available new experimental studies on O3 dosimetry, all
21 of which were conducted by Ultman and colleagues. Table AX4-1 in Annex AX4 summarizes
22 theses studies.
23
24 4.2.1 Bolus-Response Studies
25 The bolus-response method has been used by the Ultman group as an approach to explore
26 the absorption of O3 by humans. This non-invasive method consists of an injection of a known
27 volume and concentration of O3 during inspiration. Ozone uptake is the amount of O3 absorbed
28 during a single inspiration relative to the amount contained the inhaled bolus. Exposure to
29 nitrogen dioxide (NO2) or sulfur dioxide (SO2) just prior to O3 exposure has an effect on the
30 amount of O3 absorbed in conducting airways. Asplund et al. (1996) used continuous O3
31 exposure, which consisted of 2 hours of O3at 0.0, 0.12, or 0.36 ppm. Rigas et al. (1997) used
January 2005 4-3 DRAFT-DO NOT QUOTE OR CITE
-------
1 continuous 2 h exposures of 0.0, 0.36, or 0.75 ppm NO2, or 0.0 or 0.36 SO2. In both experiments
2 the continuous exposure was interrupted every 30 min by a series of bolus breaths of 2 ppm O3 at
3 250 mL/s, targeting the lower conducting airways. With continuous O3 exposure, the absorbed
4 fraction (FA) of the bolus decreased, which suggested to the authors that the continuous O3
5 exposure possibly depleted biochemical substrates from the airways, reducing their capacity to
6 absorb more O3. With NO2 and SO2 continuous exposures, the absorbed fraction of the bolus
7 increased, which the authors contributed to the these same biochemical substrates being made
8 available by the continuous exposure.
9 The bolus-response technique was also used to ascertain differences in lung anatomy and
10 gender that can alter the exposure-dose cascade (Bush et al., 1996a). Forced vital capacity
11 (FVC), total lung capacity (TLC) and anatomic dead space (VD) were determined for ten male
12 and ten female subjects, who were then exposed to a 20 ml bolus of 3 ppm O3 injected into the
13 airstream. In all subjects, dosimetry differences could be explained by differences in VD. The
14 investigators point out that the applicability of their results may be limited because of their
15 assumptions that the intrinsic mass transfer parameter (Ka) was independent of location in the
16 RT and that there was no mucous resistance. They further suggested that the dependence of Ka
17 on flowrate and VD be restricted to flowrates < 1000 mL/s until studies at higher rates have been
18 performed.
19 Nodelman and Ultman (1999a) used the bolus-response method to demonstrate that the
20 uptake distributions of O3 was sensitive to the mode of breathing and to the airflow rate.
21 As flowrates increased from 150 to 1000 mL/s, O3 penetrated deeper in to the lung and
22 penetration was further increased with oral breathing. The authors suggest that the switch from
23 nasal to oral breathing coupled with increases in respiratory flow as occurs during exercise
24 causes a shift in the O3 dose distribution, allowing O3 to penetrate deeper into the lung,
25 increasing the potential for damage to bronchiolar and alveolar tissues.
26 Very recent work by this group (Ultman et al., 2004) demonstrated that differences in
27 regional O3 uptake are due, in part, to differences in lung anatomy. Using 60 male and female
28 subjects exposed to O3 at a high minute ventilation, they measured the penetration volume at
29 which 50% of the O3 bolus was taken up. Very little O3 was taken up in the upper airway, thus
30 most of it was taken up in the lower conducting airways and peripheral airspaces. This
31 significant difference in uptake suggests to the authors that in females the smaller airways,
January 2005 4-4 DRAFT-DO NOT QUOTE OR CITE
-------
1 and associated larger surface-to-volume ratio, enhance local O3 uptake and cause reduced
2 penetration of O3 into the distal lung. Thus, these findings indicate that overall O3 uptake is not
3 related to airway size, but that the distribution of O3 shifts distally as the size of the airway is
4 increased.
5 These bolus- response studies suggest that prior continuous exposure to O3 limits uptake
6 from a bolus dose. This paradigm of exposure would have some relevance to environmental
7 exposures in which humans may receive differing day and night concentrations of O3. The lack
8 of gender differences in O3 dosimetry is an important finding that is in agreement with
9 Sarangapani et al. (2003), discussed in Section4.2.3, who reported no gender differences in O3
10 extraction. The new findings characterizing O3 uptake as inversely related to airflow are in
11 agreement with earlier animal studies. Discussions about estimating mass transfer coefficients
12 and the "accuracy" of these models are contained in Annex AX4.
13
14 4.2.2 General Uptake Studies
15 Ultman and colleagues have recently completed some general uptake studies to determine
16 the ratio of O3 uptake to the quantity of O3 inhaled. Fractional absorption (FA) was determined
17 at exposures of 0.2 or 0.4 ppm O3 while exercising at a minute volume of approximately
18 20 L/min for 60 minutes or 40 L/min for 30 minutes in both men and women (Rigas et al., 2000).
19 Fractional absorption ranged from 0.56 to 0.98 with only large changes in concentration, minute
20 volume, and exposure time having statistically significant effects on FA. Intersubject
21 differences had the largest influence on FA, resulting in a variation of approximately 10%.
22 As the quantity of O3 retained by the RT is equal to FA times the quantity of O3 inhaled,
23 relatively large changes in concentration, minute volume, or exposure time may result in
24 relatively large changes in the amount of O3 retained by the RT or absorbed locally. The authors
25 concluded that for exposure times < 2 h, inhaled dose is a reasonable predictor of actual uptake
26 as long as there are fixed concentrations of O3 and fixed levels of exercise. More importantly,
27 individuals exposed to similar doses varied in the amount of actual dose received.
28 Santiago et al. (2001) studied the effects of airflow rate (3 to 15 L/min) and O3
29 concentration (0.1, 0.2, or 0.4 ppm) on O3 uptake in nasal cavities of males and females. The FA
30 in the nose was inversely related to the flowrate and the concentration of O3 in the inlet air.
31 They computed a gas-phase diffusion resistance of < 24% of overall diffusion resistance which
January 2005 4-5 DRAFT-DO NOT QUOTE OR CITE
-------
1 suggested to them that simultaneously occurring diffusion and chemical reactions in the mucous
2 layer were the limiting factors in O3 uptake. Difference in O3 uptake ranged from 0.63 to 0.97 at
3 flowrates of 3 L/min and 0.25 to 0.50 at 15 L/min. The effect of flowrate, concentration, and
4 subject on FA were statistically significant, but subject variability accounted for approximately
5 half of the total variation in FA. Both these general uptake studies, done at environmentally
6 relevant O3 concentrations, indicate that inter-individual differences in fractional uptake are
7 extremely important in O3 dose-response relationships.
8 In the research mentioned above, Ultman et al. (2004) also completed continuous exposure
9 studies. The same 60 subjects were exposed continuously for 1 h to either clean air or 0.25 ppm
10 ozone while exercising at a target minute ventilation of 30 L/min. This is the first study to assess
11 ventilatory and dosimetric parameters for an entire hour of exposure. Additionally they
12 measured bronchial cross-sectional area available for gas diffusion in addition to other
13 ventilatory parameters. The mean fractional O3 uptake efficiency was 0.89 ± 0.06. They found
14 an inverse correlation between uptake and breathing frequency and a direct correlation between
15 uptake and tidal volume. The uptake efficiency decreased during the four sequential 15 minute
16 intervals of the 1 h exposure, demonstrating a general decrease in uptake efficiency with
17 increased breathing frequency and decreasing tidal volume. Ozone uptake rate correlated with
18 individual bronchial cross-sectional area, but did not correlate with individual %FEVj. Neither
19 of these parameters correlated with the penetration volume determined in the bolus studies
20 mentioned above. The authors concluded that the intersubject differences in forced respiratory
21 responses were not due to differences in O3 uptake. However, these data did partially support
22 the hypothesis that the differences in cross-sectional area available for gas diffusion induce
23 differences in O3 uptake.
24
25 4.2.3 Dosimetry Modeling
26 When all of the animal and human in vivo O3 uptake efficiency data are compared, there is
27 a good degree of consistency across data sets, which raises the level of confidence with which
28 these data sets can be used to support dosimetric model formulations. Early models predicted
29 that net O3 dose to lung lining fluid plus tissue gradually decreases distally from the trachea
30 toward the end of the TB, and then rapidly decreases in the pulmonary region. When the
31 theoretical dose of O3 to lung tissue is computed, it is low in the trachea, increases to a maximum
January 2005 4-6 DRAFT-DO NOT QUOTE OR CITE
-------
1 in the terminal bronchioles of the first generation of the pulmonary region, and then decreases
2 rapidly distally into the pulmonary region. The increased VT and flow, associated with exercise
3 in humans or CO2-stimulated ventilation increases in rats, shifts O3 dose further into the
4 periphery of the lung, causing a disproportionate increase in distal lung dose.
5 Localized damage to lung tissue has been modeled showing variation of O3 dose among
6 anatomically equivalent ventilatory units as a function of path length from the trachea with
7 shorter paths showing greater damage. More recent data indicate that the primary site of acute
8 cell injury occurs in the conducting airways (Postlethwait et al., 2000). These data must be
9 considered when developing models that attempt to predict site-specific locations of O3-induced
10 injury. The early models computed relationships between delivered regional dose and response
11 with the assumption that O3 was the active agent responsible for injury. It is now known that
12 reactive intermediates such as hydrohydroxyperoxides and aldehydes are important agents
13 mediating the response to O3 (further discussed in Section 5.3.1). Thus, models must consider
14 O3 reaction/diffusion in the ELF and ELF-derived reactions products.
15 Table AX4-2 in the annex presents a summary of new theoretical studies of the uptake of
16 O3 by the RTs (or regions) of humans and laboratory animals that have become available since
17 the 1996 review. They are discussed below.
18 Overton and Graham (1995) created a rat model combining multiple path anatomic models
19 and one-dimensional convection-dispersion equations which simulates transport and uptake of
20 O3 in airways and airspaces of the modeled TB region. Predictions from this model realistically
21 detail O3 transport and uptake of different but morphologically equivalent sites.
22 Using computational fluid dynamics (CFD), Cohen-Hubal et al. (1996) modeled the effect
23 of the mucus layer thickness in the nasal passage of a rat. Predictions of overall uptake were
24 within the range of measured uptake. Predicted regional O3 flux was correlated with measured
25 cell proliferation for the CFD simulation that incorporated two regions, each with a different
26 mucus thickness. But using bolus-response data described above, Hu et al. (1994) and Bush
27 et al. (2001) estimate a reaction rate constant that is more than 1000 times as large as that used
28 by Cohen-Hubal et al. (1996).
29 The authors acknowledge that the reaction rate constant may be underestimated in this
30 model due to the report by Pryor (1992) predicting that O3 does not penetrate a liquid lining
31 layer more than 0.1 |im thick.
January 2005 4-7 DRAFT-DO NOT QUOTE OR CITE
-------
1 With a RT dosimetry model, Overton et al. (1996) investigated the sensitivity of absorbed
2 fraction (AF), proximal alveolar region (PAR) dose, and PAR dose ratio to TB region volume
3 (VTB) and TB region expansion in humans and rats. The PAR was defined as the first generation
4 distal to terminal bronchioles and the PAR dose ratio was defined as the ratio of a rat's predicted
5 PAR dose to a human's predicted PAR dose. This ratio relates human and rat exposure
6 concentrations so that both species receive the same PAR dose. In rats, the PAR is a region of
7 major damage from O3. For each species, three values of V^ were used: a mean value from the
8 literature and the mean ± twice the SD. For both the rat and human simulations, there were
9 several general findings: (1) AF and PAR both increased with decreasing V^, e.g., using the
10 highest kTB, the PAR for Vra-2SD was five times greater than the PAR for Vra+2SD, (2) AF and
11 PAR both decreased with TB expansion relative to no expansion, 3) PAR increased with tidal
12 volume, 4) PAR increased with decreasing k^, and 5) AF increased with kTB.
13 Bush et al. (2001) modified their single-path model (Bush et al., 1996) so that simulations
14 would coincide with experimental AF data for O3 and C12 during oral and nasal breathing.
15 Relative to their original model, the Bush et al. (2001) model added lung expansion and
16 modified the mass transfer coefficients for both the gas-phase (kg) and the liquid-phase (k,).
17 Using kb a variable reaction rate constant (k,.) was defined as ^ divided by the mucus layer
18 thickness which was assumed to decrease in thickness with progression from the trachea to the
19 lower airways. Consistent with Overton et al. (1996), considering expansion of the TB airways
20 reduced AF versus no expansion. As very little inhaled O3 reaches the peripheral lung, it was not
21 surprising that alveolar expansion had minimal affect on AF. Ignoring k,., the simulations for O3
22 and C12 were nearly the same since the gas-phase diffusion coefficients of O3 and C12 are similar.
23 But for a given volumetric depth into the TB airways of the lung, experimental AF data are
24 always greater for O3 than for C12. The authors surmised that the difference between the AF for
25 these gases could be explained adequately based solely on the diffusive resistance of O3 in
26 airways surface fluid (modeled by k,.). Interestingly, k,. was lower for oral than nasal breathing,
27 implying less antioxidant capacity in the airway surface liquid of the oral versus the nasal cavity.
28 Qualitatively, model simulations also agreed well with the experimental data of Gerrity et al.
29 (1995).
30 Age- and gender-specific differences in both regional and systemic uptake in humans was
31 modeled using physiologically-based pharmacokinetic (PBPK ) approach (Sarangapani et al.,
January 2005 4-8 DRAFT-DO NOT QUOTE OR CITE
-------
1 2003). The model estimated that regional extraction of O3 is relatively insensitive to age and
2 gender and that the postnatal period is the age (in which extraction of O3 in infants is 2- to 8-fold
3 higher than is adults) at which the largest difference in pharmacokinetics exist.
4 A recent attempt was made (Mudway and Kelly, 2004) to model O3 dose-inflammatory
5 response using a meta-analysis of 23 exposures in published human chamber studies. The O3
6 concentrations ranged from 0.08 to 0.6 ppm and the exposure durations ranged from 60 to
7 396 minutes. The analysis showed linear relationships between O3 dose and neutrophilia in
8 bronchoalveolar lavage fluid (BALF). Linear relationships were also observed between O3 dose
9 and protein leakage into BALF, which suggested to the authors that a large-scale study could
10 determine a possible O3 threshold level for these inflammatory responses.
11
12 4.2.4 Summary and Conclusions - Dosimetry
13 Ozone is a highly reactive gas and powerful oxidant with a short half-life. Uptake occurs
14 in mucous membranes of the RT where O3 immediately reacts with components of the ELF.
15 Uptake efficiency is chemical-reaction dependent and the reaction products (hydrogen peroxide,
16 aldehydes, and hydroxyhydroperoxides) created by ozonolysis of polyunsaturated fatty acids
17 (PUFA) mediate of O3 toxicity. The previous literature review found that uptake of O3 in rat is
18 about 0.50 and in humans at rest is about 0.8 to 0.95. About 0.07 of the O3 is removed in the
19 larynx/trachea, about 0.50 is removed in the head, and about 0.43 is removed in the lungs, where
20 the primary site of damage is the centriacinar region (CAR). There was no agreement as to
21 whether uptake was dependent on flow. Studies in humans have shown that increasing VE with
22 exercise (by increasing both breathing frequency and tidal volume) causes only a small decrease
23 in uptake efficiency by the total RT. Mode of breathing had little effect on uptake, suggesting
24 that O3 is removed equally by both mouth and nose. Comparing BAL cells from rat and human,
25 a 0.4 ppm exposure in exercising humans gave 4 to 5 times the dose of O3 than a rat at rest
26 exposed to the same concentration.
27 New research on O3 uptake has been performed in humans, but not in laboratory animals.
28 Bolus-response studies demonstrated that a previous continuous exposure to O3 decreased the
29 absorption of a bolus of O3, probably due to depletion of compounds able to absorb O3. Previous
30 continuous exposure to NO2 and SO2 increased absorption of a bolus of O3. These data are of
31 some relevance to environmental exposures where humans may receive differing concentrations
January 2005 4-9 DRAFT-DO NOT QUOTE OR CITE
-------
1 of O3 depending on time of day. Another bolus-response study showed that absorption of O3 was
2 dependent on VD, but not height, weight, age, gender, FVC, or TLC. In contrast to earlier data,
3 the bolus-response method was used to demonstrate that the uptake distributions of O3 is
4 sensitive to the mode of breathing and to the airflow rate. As flow rates increased from 150 to
5 1000 mL/s, O3 penetrated deeper in to the lung and was further increased with oral breathing.
6 This suggests that the switch from nasal to oral breathing coupled with increases in respiratory
7 flow as occurs during exercise causes a shift in the O3 dose distribution, allowing O3 to penetrate
8 deeper into the lung, increasing the potential of damage to bronchiolar and alveolar tissues. The
9 finding that O3 uptake is inversely related to airflow agrees with earlier animal studies.
10 New general uptake study data demonstrate that exercising men and women receiving
11 0.2 or 0.4 ppm O3 at 20 L/min for 60 minutes or 40 L/min for 30 minutes absorb 0.56 to 0.98.
12 The absorbed fraction or FA is affected only by large changes in concentration, minute volume,
13 and exposure time. This suggests that for exposure times < 2 h, inhaled dose is a reasonable
14 predictor of actual uptake as long as there are fixed concentrations of O3 and fixed levels of
15 exercise. Individuals exposed to similar concentrations vary considerably in the amount of
16 actual dose received. This intersubject variability was also demonstrated in a study of O3 uptake
17 in nasal cavities of men and women. The FA in the nose was inversely related to the flowrate
18 and the concentration of O3 suggesting that simultaneously occurring diffusion and chemical
19 reactions in the mucous layer were the limiting factors in O3 uptake. Both these general uptake
20 studies, done at environmentally relevant O3 concentrations, indicate that inter-individual
21 differences in fractional uptake, which can range from 0.25 to 0.97, are extremely important in
22 O3 dose-response relationships.
23 The consistency of uptake data generated in animal and human studies allow a high level
24 of confidence in their use in dosimetry modeling. Early models predicted that net O3 dose to
25 ELF and tissue gradually decreases distally from the trachea toward the end of the TB, and then
26 rapidly decreases in the pulmonary region. Exercise-induced or CO2-stimulated increases in VT
27 and flow, shift O3 dose further into the periphery of the lung, causing a disproportionate increase
28 in distal lung dose. Localized damage to lung tissue has been modeled showing variation of O3
29 dose among anatomically equivalent ventilatory units as a function of path length from the
30 trachea with shorter paths showing greater damage.
January 2005 4-10 DRAFT-DO NOT QUOTE OR CITE
-------
1 New models have produced some refinements of earlier models such as: (1) the use of
2 mucus resistance and thickness in describing O3 dosimetry and determining the patterns of
3 O3-induced lesions; (2) the shape of the dose versus generation plot along any path from the
4 trachea to alveoli is independent of path, with the tissue dose decreasing with increasing
5 generation index; (3) simulations sensitive to conducting airway volume but relatively
6 insensitive to characteristics of the respiratory airspace; (4) the importance of TB region
7 expansion; (5) the importance of dose received in the PAR both inter-individual differences and
8 extrapolations based on dose; (6) revaluation of mass transfer coefficients for conducting
9 airways, and (7) extraction of O3 in infants is 2- to 8-fold higher than in adults, but the
10 differences leveled out by age 5. Additionally, more recent data indicate that the primary site of
11 acute cell injury occurs in the conducting airways and that reactive intermediates in the ELF,
12 rather than O3 itself, are responsible for pulmonary injury. These data must be considered when
13 developing new models.
14
15
16 4.3 SPECIES HOMOLOGY, SENSITIVITY, AND ANIMAL-TO-HUMAN
17 EXTRAPOLATION
18 Basic similarities exist across human and other animals species with regard to basic
19 anatomy, physiology, biochemistry, cell biology, and disease processes. This homology creates
20 similarities in acute O3-induced effects, especially in the respiratory tract and in lung defense
21 mechanisms. Rodents appear to have a slightly higher tachypneic response to O3, which is
22 clearly concentration-dependent in most species and shows parallel exacerbation when
23 hyperventilation (e.g., exercise or CO2) is superimposed. What is not known is whether this is
24 evidence of pulmonary irritant sensitivity, perhaps as a prelude to toxicity, or whether tachypnea
25 is a defensive action taken by the respiratory system to minimize distal lung O3 deposition.
26 Airway or lung resistance in humans is not affected appreciably by acute exposure to O3, except
27 under conditions of heavy exercise; animals appear to need high-level exposures or special
28 preparations that bypass nasal scrubbing. Dynamic lung compliance (Cdyn) has been shown to
29 have small magnitude decreases in response to O3 in some studies across species, but it is
30 thought that these changes are of little biological significant for ambient exposures. Spirometric
31 changes, the hallmark of O3 response in humans, occur in rats, but to a lesser degree. It is
January 2005 4-11 DRAFT-DO NOT QUOTE OR CITE
-------
1 unclear, however, the degree to which anesthesia (rat) and the comparability of hyperventilation
2 induced by CO2 (rat) or exercise (human) may influence this difference in responsiveness.
3 Collectively, the acute functional response of laboratory animals to O3 appears quite homologous
4 to that of the human.
5 Examination of BAL constituents show that the influx of inflammatory cells and protein
6 from the serum is influenced by species, but perhaps to less extent than by ventilation and
7 antioxidant status. Adjustment for these factors can modulate responses to approximate animal
8 responses to those of humans. Unfortunately, these influential factors are rarely measured and,
9 even less often, controlled. Increases in protein levels in BALF with O3 exposures in guinea pigs
10 are also a factor in the species' susceptibility to the effects of O3. Species comparisons of acute
11 O3 exposures to mice, guinea pigs, rats, hamsters, and rabbits found that guinea pigs were the
12 most responsive (to > 0.2 ppm); rabbits were the least responsive (2.0 ppm only); and rats,
13 hamsters, and mice were intermediate (effects at > 1.0 ppm). Rats and humans have subtle
14 species-specific differences in inflammatory responses to O3 in terms of the timing of PMN
15 influx in the nasal and bronchoalveolar regions.
16 When humans are exposed to O3 repeatedly for several consecutive days, lung function
17 decrements subside, and normal spirometric parameters are regained. This phenomenon of
18 functional attenuation also has been demonstrated in rats, not only in terms of spirometry, but
19 also in terms of the classic tachypneic ventilatory response. Full or partial attenuation of the
20 BAL parameters also appears to occur in both rats and humans, but exposure scenario appears to
21 play a role; other cellular changes in animals do not attenuate. Existing epidemiologic studies
22 provide only suggestive evidence that persistent or progressive deterioration in lung function is
23 associated with long-term oxidant-pollutant exposure (See Chapter 7). These long-term effects
24 are thought to be expressed in the form of maximum airflow or spirometric abnormalities, but
25 the foundation for this conclusion remains weak and hypothetical. Animal study data, although
26 suggesting that O3 has effects on lung function at near-ambient levels, present a variable picture
27 of response that may or may not relate to technical conditions of exposure or some other, yet
28 undiscovered variable of response. Thus, a cogent interpretation of the animal findings as
29 definitive evidence of chronic deterioration of lung function would be difficult at this time.
30 However, the subtle functional defects apparent after 12 to 18 mo of exposure and the detailed
31 morphometric assessments of the O3-induced lesions do appear consistent with the modicum of
January 2005 4-12 DRAFT-DO NOT QUOTE OR CITE
-------
1 studies focusing on long-term effects in human populations. Based on the apparent homology of
2 these responses between humans and laboratory animals, animal studies may potentially provide
3 a means to more directly assess such chronic health concerns.
4 A species' susceptibility to the effects of O3 exposure may be due, in part, to biochemical
5 differences among species. Evidence for this is provided by differences in activity of SD rat and
6 rhesus monkey CYP moonoxygenases elicited by O3 exposure (Lee et al., 1998). Additional
7 characterization of species- and region-specific CYP enzymes will create a better understanding
8 of the differences in response to O3. This will allow more accurate extrapolation from animal
9 exposures to human exposures and toxic effects.
10 Antioxidant metabolism varies widely among species, which can greatly influence the
11 effects of O3 (discussed in greater detail in 5.2.1.3). The guinea pig appears to be the species
12 most susceptible to ozone. Early studies ranked mice > rats > guinea pigs in order of antioxidant
13 responsiveness to O3 challenge. Guinea pigs have been shown to have lower basal levels of
14 GSH transferase activity, lower activity of GSH peroxidases, and lower levels of vitamin E
15 compared to rats. These lower levels of antioxidants combined with increases in protein levels
16 in BALF (discussed above) with O3 exposures likely explain, at least in part, the species'
17 susceptibility to the effects of O3.
18 Because cytokine and chemokine responses are so important in an animal's defense against
19 O3 exposure, comparisons of differences in species expression and activity of these
20 inflammatory mediators is necessary. Arsalane et al. (1995) compared guinea pig and human
21 AM recovered in BALF and subsequently exposed in vitro to 0.1 to 1 ppm for 60 minutes.
22 Measurement of inflammatory cytokines showed a peak at 0.4 ppm in both species. Guinea pig
23 AM had an increase in IL-6 and TNFa while human AM had increases in TNFa, IL-lb, IL-6 and
24 IL-8. This exposure also caused an increase in mRNA expression for TNFa, IL-lb, IL-6 and
25 IL-8 in human cells. At 0.1 ppm exposures, only TNFa secretion was increased. These data
26 suggest similar cytokine responses in guinea pigs and humans, both qualitatively and
27 quantitatively.
28 Species differences in morphological responses to O3 exposure have been characterized by
29 Dormans et al. (1999), as discussed in previous sections. Dormans et al. (1999) continuously
30 exposed rats, mice, and male guinea pigs to filtered air, 0.2, or 0.4 ppm O3 for 3, 7, 28, and
31 56 days. The animals exposed for 28 days were examined at 3, 7, or 28 days PE. Depending
January 2005 4-13 DRAFT-DO NOT QUOTE OR CITE
-------
1 on the endpoint studied, the species varied in sensitivity. Greater sensitivity was shown in the
2 mouse as determined by biochemical endpoints, persistence of bronchiolar epithelial
3 hypertrophy, and recovery time. Guinea pigs were more sensitive in terms of the inflammatory
4 response though all three species had increases in the inflammatory response after three days that
5 did not decrease with exposure. These data on inflammation are in general agreement with
6 Hatch et al., (1986), discussed above. In all species the longest exposure to the highest dose
7 caused increased collagen in ductal septa and large lamellar bodies in Type II cells, but that
8 response also occurred in rats and guinea pigs at 0.2 ppm. No fibrosis was seen at the shorter
9 exposure times and the authors question whether fibrosis occurs in healthy humans after
10 continuous exposure. The authors do not rule out the possibility that some of these differences
11 may be attributable to differences in total inhaled dose or dose actually reaching a target site.
12 Overall, the authors rated mice as most susceptible, followed by guinea pigs and rats.
13 Comparisons of airway effects in rats, monkeys and ferrets resulting from exposures of
14 1.0 ppm O3 for 8 h (Sterner-Kock et al. 2000) demonstrated that monkeys and ferrets had a
15 similar inflammatory responses and epithelial necrosis. The response of these two species was
16 more severe than that seen in rats. These data suggest that ferrets are a good animal model for
17 O3-induced airway effects due to the similarities in pulmonary structure between primates and
18 ferrets.
19 A number of species, including nonhuman primates, dogs, cats, rabbits, and rodents, have
20 been used to study the effects of O3 exposure on airway bronchoconstriction. A commonly used
21 model of bronchospasm utilizes guinea pigs acutely exposed to high O3 concentrations (2 to
22 3 ppm) to induce airway hyperreactivity (AHR). As mentioned earlier, the model is helpful for
23 determining mechanistic aspects of AHR, but is not really relevant for extrapolation to potential
24 airway responses in humans exposed to ambient levels of O3. Additionally, guinea pigs have
25 been shown to have AHR in other studies that is very similar to human asthmatics, but the utility
26 of guinea pig data is somewhat limited by their disparity from other animal models.
27 The rat is a key species used in O3 toxicological studies, but the rat has both behavioral
28 and physiological mechanisms that can lower core temperature in response to acute exposures,
29 thus limiting extrapolation of rat data to humans. Iwasaki et al. (1998) evaluated cardiovascular
30 and thermoregulatory responses to O3 at exposure of 0.1, 0.3, and 0.5 ppm O3 8 hrs/day for
31 4 consecutive days. A dose-dependent disruption of HR and Tco were seen on the first and
January 2005 4-14 DRAFT-DO NOT QUOTE OR CITE
-------
1 second days of exposure, which then recovered to control values. Watkinson et al. (2003)
2 exposed rats to 0.5 ppm O3 and observed this hypothermic response, which included lowered
3 HR, lowered Tco, and increased inflammatory components in BALF. The authors suggested that
4 the response is an inherent reflexive pattern that can possibly attenuate O3 toxicity in rodents.
5 They discuss the cascade of effects created by decreases in Tco, which include: (1) lowered
6 metabolic rate, (2) altered enzyme kinetics, (3) altered membrane function, (4) decreased oxygen
7 consumption and demand, (5) reductions in minute ventilation, which would act to limit the dose
8 of O3 delivered to the lungs. These effects are concurrent with changes in HR which lead to:
9 (1) decreased CO, (2) lowered BP, and (3) decreased tissue perfusion, all of which may lead to
10 functional deficits. They hypothermic response has not been observed in humans except at very
11 high exposures, which complicates extrapolation of results in rat studies to humans.
12 The importance of animal studies derives from their utilization in determining cause-effect
13 relationships between exposure and health outcome, but the animal data must be integrated with
14 epidemiological studies and controlled human clinical studies. Animal studies can corroborate
15 both clinical and epidemiology studies and further provide important data that is impossible to
16 collect in human studies. Toxic pulmonary and extrapulmonary effects following ozone
17 exposure have been well-studied in rodents, nonhuman primates, and a few other species;
18 so, extrapolation, both qualitative and quantitative, to human exposures and consequent health
19 effects is possible. Quantitative extrapolation, required to determine what specific exposure is
20 likely to cause an effect in humans, is theoretically founded on the equivalency of mechanisms
21 across species. At the molecular level, O3 acts on the carbon-carbon double bond in
22 polyunsaturated fatty acids and on sulfhydryl groups in proteins, both of which are found within
23 cell membranes in animals and humans. At higher levels of cellular organization, cells affected
24 in animals (e.g., AMs, Type 1 cells) have similar functions in humans, and organ systems (e.g.,
25 respiratory system) have major interspecies similarities. However, interspecies differences do
26 occur and confound extrapolation.
27 Quantitative extrapolation, which involves a combination of dosimetry and species
28 sensitivity, still requires more research before it can be fully realized. Knowledge of dosimetric
29 animal-to-human extrapolation is more advanced than that of species-sensitivity, but
30 extrapolation models have not been completely validated, and therefore, significant uncertainties
31 remain. Mathematical modeling of O3 deposition in the lower respiratory tract (i.e., from the
January 2005 4-15 DRAFT-DO NOT QUOTE OR CITE
-------
1 trachea to alveoli) of several animal species and humans shows that the pattern of regional dose
2 is similar, but that absolute values differ. In spite of structural and ventilatory differences
3 between species, the greatest predicted tissue dose is to the CAR. Even though the CAR of rats
4 has very rudimentary respiratory bronchioles, compared to well-developed ones in primates, the
5 CAR of both rats and nonhuman primates respond similarly to O3.
6 Experimental measurement of delivered O3 doses estimate that total respiratory uptake is
7 -47% in laboratory animals and -87% in exercising humans, while nasopharyngeal removal is
8 -17% in rats and -40% in humans. The previous O3 AQCD (U.S. Environmental Protection
9 Agency, 1996) provided the first quantitative animal-to-human extrapolation of morphological
10 changes in the proximal alveolar region using rat and monkey studies. The extrapolation
11 predicted that a 9-year-old child would have a 20% or 75% increase in PAR tissue thickness if
12 their sensitivity to O3 was equal to that of a rat or monkey, respectively. Adults would have
13 15 or 70% increase, suggesting the potential for chronic effects in humans. In spite of the
14 significant uncertainties, this extrapolation raises concern about the potential for chronic effects
15 in humans
16 Experiments using 0.4 ppm 18O3 suggested that exercising humans received a 4- to 5-fold
17 higher 18O3 concentrations in BAL than resting rats (Hatch et al., 1994). That level of exposure
18 increased BAL protein and PMNs in humans, while a concentration of 2.0 ppm in rats was
19 necessary for similar effects. Caveats in the interpretation of 18O3 studies include: (1) only a
20 very small portion of the labeled compound is recoverable to assess incorporation; and
21 (2) if species being compared differ in physiocochemical factors controlling mass transfer
22 and downstream O3 metabolism, it could cause significant differences in the amount of inhaled
23 18O3 that is detected during subsequent tissue analysis. Further, species differences in pulmonary
24 anatomy, ventilation, antioxidants, and susceptibility all influence dose, repair processes, and
25 tolerance to subsequent O3 exposure. Important differences between exercising humans and
26 resting rats that can affect tissue O3 dose include: (1) increased ventilation and O3 delivery with
27 exercise; (2) decreased pulmonary ventilation and body temperature during O3 exposure in rats;
28 (3) diminished dose received in rats due to their burying their noses in their fur during exposure;
29 and (4) increased concentration of antioxidants in ELF in rats compared to humans. These
30 antioxidants are important for converting O3 to inactive products before toxicity occurs (Kari
31 et al., 1997; Gunnison and Hatch, 1999; Plopper et al., 1998), though this quenching is not
January 2005 4-16 DRAFT-DO NOT QUOTE OR CITE
-------
1 quantitative. These and possibly other differences between rats and humans suggest that a
2 2 ppm exposure in nonexercising rats approximates a 0.4 ppm exposure in exercising humans.
3 Further comparisons of exercising human exposure to 0.1 ppm for 6 hours (Devlin et al., 1991)
4 and resting rat exposure to 0.3 ppm show inflammatory and permeability changes in humans but
5 not rats.
6
7 4.3.1 Summary and Conclusions: Species Homology, Sensitivity, and
8 Animal-to-Human Extrapolation
9 Comparisons of acute exposures in rat and human suggest that, though both species have
10 similar qualitative responses to O3 exposure, there are interspecies mechanistic disparities that
11 necessitate careful comparisons of dose-response relationships. There is no perfect nonhuman
12 species with which to model O3 toxicity. All have limitations that must be considered when
13 attempting to extrapolate to human exposures. Awareness of these limitations, even at the level
14 of subtle strain differences within a test species, is extremely important. The currently available
15 data suggest that LOELs in resting rats are approximately 4- to-5 fold higher than for exercising
16 humans for toxicological endpoints including BAL protein and BAL PMNs. Studies comparing
17 species-specific differences in O3-induced effects showed that guinea pigs were the most
18 susceptible, rabbits the least susceptible, and rodents intermediate in susceptibility. The recent
19 work being done utilizing various mouse strains with differing sensitivities to O3 will help us to
20 understand the extremely complex inter-individual differences in human sensitivity to O3.
21
January 2005 4-17 DRAFT-DO NOT QUOTE OR CITE
-------
1 REFERENCES
2 Arsalane, K.; Cosset, P.; Vanhee, D.; Voisin, C.; Hamid, Q.; Tonnel, A.-B.; Wallaert, B. (1995) Ozone stimulates
3 synthesis of inflammatory cytokines by alveolar macrophages in vitro. Am. J. Respir. Cell Mol. Biol.
4 13:60-68.
5 Asplund, P. T.; Ben-Jebria, A.; Rigas, M. L.; Ultman, J. S. (1996) Longitudinal distribution of ozone absorption in
6 the lung: effect of continuous inhalation exposure. Arch. Environ. Health 51: 431-438.
7 Bush, M. L.; Asplund, P. T.; Miles, K. A.; Ben-Jebria, A.; Ultman, J. S. (1996a) Longitudinal distribution of O3
8 absorption in the lung: gender differences and intersubject variability. J. Appl. Physiol. 81: 1651-1657.
9 Bush, M. L.; Raybold, T.; Abeles, S.; Hu, S.-C; Ben-Jebria, A.; Ultman, J. S. (1996b) Longitudinal distribution of
10 ozone absorption in the lung: simulation with a single-path model. Toxicol. Appl. Pharmacol. 140: 219-226.
11 Bush, M. L.; Zhang, W.; Ben-Jebria, A.; Ultman, J. S. (2001) Longitudinal distribution of ozone and chlorine in the
12 human respiratory tract: simulation of nasal and oral breathing with the single-path diffusion model. Toxicol.
13 Appl. Pharmacol. 173: 137-145.
14 Cohen-Hubal, E. A.; Kimbell, J. S.; Fedkiw, P. S. (1996) Incorporation of nasal-lining mass-transfer resistance into a
15 CFD model for prediction of ozone dosimetry in the upper respiratory tract. Inhalation Toxicol. 8: 831-857.
16 Dahl, A. R. (1990) Dose concepts for inhaled vapors and gases. Toxicol. Appl. Pharmacol. 103: 185-197.
17 Devlin, R. B.; McDonnell, W. F.; Mann, R.; Becker, S.; House, D. E.; Schreinemachers, D.; Koren, H. S. (1991)
18 Exposure of humans to ambient levels of ozone for 6.6 hours causes cellular and biochemical changes in the
19 lung. Am. J. Respir. Cell Mol. Biol. 4: 72-81.
20 Dormans, J. A. M. A.; VanBree, L.; Boere, A. J. F.; Marra, M.; Rombout, P. J. A. (1999) Interspecies differences in
21 time course of pulmonary toxicity following repeated exposure to ozone. Inhalation Toxicol. 11: 309-329.
22 Gerrity, T. R.; Biscardi, F.; Strong, A.; Garlington, A. R.; Brown, J. S.; Bromberg, P. A. (1995) Bronchoscopic
23 determination of ozone uptake in humans. J. Appl. Physiol. 79: 852-860.
24 Gunnison, A. F.; Hatch, G. E. (1999) O3-induced inflammation in prepregnant, pregnant, and lactating rats correlates
25 with O3 dose estimated by 18O. Am. J. Physiol. 276: L332-L340.
26 Hatch, G. E.; Slade, R.; Stead, A. G.; Graham, J. A. (1986) Species comparison of acute inhalation toxicity of ozone
27 and phosgene. J. Toxicol. Environ. Health 19: 43-53.
28 Hu, S.-C.; Ben-Jebria, A.; Ultman, J. S. (1994) Longitudinal distribution of ozone absorption in the lung: effects of
29 respiratory flow. J. Appl. Physiol. 77: 574-583.
30 Iwasaki, T.; Takahashi, M.; Saito, H.; Arito, H. (1998) Adaptation of extrapulmonary responses to ozone exposure in
31 conscious rats. Ind. Health 36: 57-60.
32 Kari, F.; Hatch, G.; Slade, R.; Crissman, K.; Simeonova, P. P.; Luster, M. (1997) Dietary restriction mitigates
33 ozone-induced lung inflammation in rats: a role for endogenous antioxidants. Am. J. Respir. Cell Mol. Biol.
34 17:740-747.
35 Lee, C.; Watt, K. C.; Chang, A. M.; Plopper, C. G.; Buckpitt, A. R.; Pinkerton, K. E. (1998) Site-selective
3 6 differences in cytochrome P450 isoform activities: comparison of expression in rat and rhesus monkey lung
37 and induction in rats. Drug Metab. Dispos. 26: 396-400.
3 8 Miller, F. J. (1995) Uptake and fate of ozone in the respiratory tract. Toxicol. Lett. 82/83: 277-285.
39 Mudway, I. S.; Kelly, F. J. (2004) An investigation of inhaled ozone dose and the magnitude of airway inflammation
40 in healthy adults. Am. J. Respir. Crit. Care Med. 169: 1089-1095.
41 Nodelman, V.; Ultman, J. S. (1999a) Longitudinal distribution of chlorine absorption in human airways: a
42 comparison to ozone absorption. J. Appl. Physiol. 87: 2073-2080.
43 Overton, J. H.; Graham, R. C. (1995) Simulation of the uptake of a reactive gas in a rat respiratory tract model with
44 an asymmetric tracheobronchial region patterned on complete conducting airway cast data. Comput. Biomed.
45 Res. 28: 171-190.
46 Overton, J. H.; Graham, R. C.; Menache, M. G.; Mercer, R. R.; Miller, F. J. (1996) Influence of tracheobronchial
47 region expansion and volume on reactive gas uptake and interspecies dose extrapolations. Inhalation Toxicol.
48 8:723-745.
49 Plopper, C. G.; Hatch, G. E.; Wong, V.; Duan, X.; Weir, A. J.; Tarkington, B. K.; Devlin, R. B.; Becker, S.;
50 Buckpitt, A. R. (1998) Relationship of inhaled ozone concentration to acute tracheobronchial epithelial injury,
51 site-specific ozone dose and glutathione depletion in rhesus monkeys. Am. J. Respir. Cell Mol. Biol.
52 19:387-399.
January 2005 4-18 DRAFT-DO NOT QUOTE OR CITE
-------
1 Postlethwait, E. M; Joad, J. P.; Hyde, D. M; Schelegle, E. S.; Brie, J. M; Weir, A. J.; Putney, L. F.; Wong, V. J.;
2 Velsor, L. W.; Plopper, C. G. (2000) Three-dimensional mapping of ozone-induced acute cytotoxicity in
3 tracheobronchial airways of isolated perfused rat lung. Am. J. Respir. Cell Mol. Biol. 22: 191-199.
4 Pryor, W. A. (1992) How far does ozone penetrate into the pulmonary air/tissue boundary before it reacts? Free
5 Radic. Biol. Med. 12: 83-88.
6 Rigas, M. L.; Ben-Jebria, A.; Ultman, J. S. (1997) Longitudinal distribution of ozone absorption in the lung: effects
7 of nitrogen dioxide, sulfur dioxide, and ozone exposures. Arch. Environ. Health 52: 173-178.
8 Rigas, M. L.; Catlin, S. N.; Ben-Jebria, A.; Ultman, J. S. (2000) Ozone uptake in the intact human respiratory tract:
9 relationship between inhaled dose and actual dose. J. Appl. Physiol. 88: 2015-2022.
10 Santiago, L. Y.; Harm, M. C.; Ben-Jebria, A.; Ultman, J. S. (2001) Ozone adsorption in the human nose during
11 unidirectional airflow. J. Appl. Physiol. 91: 725-732.
12 Sarangapani, R.; Gentry, P. R.; Covington, T. R.; Teeguarden, J. G.; Clewell, H. J., III. (2003) Evaluation of the
13 potential impact of age- and gender-specific lung morphology and ventilation rate on the dosimetry of vapors.
14 Inhalation Toxicol. 15: 987-1016.
15 Sterner-Kock, A.; Kock, M.; Braun, R.; Hyde, D. M. (2000) Ozone-induced epithelial injury in the ferret is similar
16 to nonhuman primates. Am. J. Respir. Crit. Care Med. 162: 1152-1156.
17 U.S. Environmental Protection Agency. (1996) Air quality criteria for ozone and related photochemical oxidants.
18 Research Triangle Park, NC: Office of Research and Development; report nos. EPA/600/AP-93/004aF-cF. 3v.
19 Available from: NTIS, Springfield, VA; PB96-185582, PB96-185590, and PB96-185608. Available online at:
20 www.epa.gov/ncea/ozone.htm.
21 Ultman, J. S.; Ben-Jebria, A.; Arnold, S. F. (2004) Uptake distribution of ozone in human lungs: intersubject
22 variability in physiologic response. Boston, MA: Health Effects Institute.
23
January 2005 4-19 DRAFT-DO NOT QUOTE OR CITE
-------
i CHAPTER 4 ANNEX. DOSIMETRY OF OZONE IN
2 THE RESPIRATORY TRACT
3
4
5 AX4.1 INTRODUCTION
6 This annex serves to provide supporting material for Chapter 4 - Dosimetry, Species,
7 Homology, Sensitivity, and Animal-to-Human Extrapolation. It includes tables that summarize
8 new literature published since the last O3 criteria documents (U.S. Environmental Protection
9 Agency, 1996). In addition, it provides descriptions of those new findings, in many cases, with
10 more detail than is provided in the chapter.
11 Dosimetry refers to measuring or estimating the quantity of or rate at which a chemical is
12 absorbed by target sites within the respiratory tract (RT). The compound most directly
13 responsible for toxic effects may be the inhaled gas O3 or one of its chemical reaction products.
14 Complete identification of the actual toxic agents and their integration into dosimetry is a
15 complex issue that has not been resolved. Thus, most dosimetry investigations are concerned
16 with the dose of the primary inhaled chemical. In this context, a further confounding aspect can
17 be the units of dose (e.g., mass retained per breath, mass retained per breath per body weight,
18 mass retained per breath per respiratory tract surface area). That is, when comparing dose
19 between species, what is the relevant measure of dose? This question has not been answered;
20 units are often dictated by the type of experiment or by a choice made by the investigators.
21 There is also some lack of agreement as to what constitutes "dose." Dahl's seminal paper (1990)
22 classified O3 as a reactive gas and discussed the characterization of dose measurement by
23 parameters including: (1) inhaled O3 concentration; (2) amount of O3 inhaled as determined by
24 minute volume, vapor concentration, and exposure duration; (3) uptake or the amount of O3
25 retained (i.e., not exhaled); (4) O3 or its active metabolites delivered to target cells or tissues;
26 (5) O3 or its reactive metabolites delivered to target biomolecules or organelles; and (6) O3 or its
27 metabolites participating in the ultimate toxic reactions - the effective dose. This
28 characterization goes from least complex to greatest, culminating in measurement of the fraction
29 of the inhaled O3 that participates in the effects of cellular perturbation and/or injury.
30 Understanding dosimetry as it relates to O3-induced injury is complex due to the fact that O3
31 interacts primarily with the epithelial lining fluid (ELF) which contains surfactant and
January 2005 AX4-1 DRAFT-DO NOT QUOTE OR CITE
-------
1 antioxidants. In the upper airways ELF is thick and highly protective against oxidant injury.
2 In lower airways ELF is thinner, has lower levels of antioxidants, and thus, allows more cellular
3 injury. Adding to the complexity is the fact that O3 can react with molecules in the ELF to create
4 even more reactive metabolites, which can then diffuse within the lung or be transported out of
5 the lung to generate systemic effects. Section 5.3.1 contains further information on the cellular
6 targets of O3 interactions and antioxidants.
7 Experimental dosimetry studies in laboratory animals and humans, and theoretical
8 (dosimetry modeling) studies, have been used to obtain information on O3 dose. Since the last
9 ozone criteria document (U.S. Environmental Protection Agency, 1996), all new experiments
10 have been carried out in humans to obtain direct measurements of absorbed O3 in the RT, the
11 upper RT (URT) region proximal to the tracheal entrance, and in the tracheobronchial (TB)
12 region; no uptake experiments have been performed using laboratory animals. Experimentally
13 obtaining dosimetry data is extremely difficult in smaller regions or locations, such as specific
14 airways or the centriacinar region (CAR; junction of conducting airways and gas exchange
15 region), where lesions caused by O3 occur. Nevertheless, experimentation is important for
16 determining dose, making dose comparisons between subpopulations and between different
17 species, assessing hypotheses and concepts, and validating mathematical models that can be used
18 to predict dose at specific respiratory tract sites and under more general conditions.
19 Theoretical studies are based on the use of mathematical models developed for the
20 purposes of simulating the uptake and distribution of absorbed gases in the tissues and fluids of
21 the RT. Because the factors affecting the transport and absorption of gases are applicable to all
22 mammals, a model that uses appropriate species or disease-specific anatomical and ventilatory
23 parameters can be used to describe absorption in the species and in different-sized, aged, or
24 diseased members of the same species. More importantly, models also may be used to make
25 interspecies and intraspecies dose comparisons, to compare and reconcile data from different
26 experiments, to predict dose in conditions not possible or feasible experimentally, and to better
27 understand the processes involved in toxicity.
28 A review (Miller, 1995) of the factors influencing RT uptake of O3 stated that structure of
29 the RT region, ventilation, and gas transport mechanisms were important. Additionally, local
30 dose is the critical link between exposure and response. A criticism of previous models of O3
January 2005 AX4-2 DRAFT-DO NOT QUOTE OR CITE
-------
1 uptake is that they have typically discounted the mucociliary layer of the URT and
2 tracheobronchial regions.
3 For a more detailed discussion on experimental and theoretical dosimetry studies the reader
4 is referred to the previous O3 criteria document (Volume III, Chapter 8, U.S. Environmental
5 Protection Agency, 1996).
6
7
8 AX4.2 EXPERIMENTAL OZONE DOSIMETRY INVESTIGATIONS
9 There have been some advances in understanding human O3 dosimetry that better enable
10 quantitative extrapolation from laboratory animal data. The next two sections review the
11 available new experimental studies on O3 dosimetry, which involve only human subjects and are
12 all from the same laboratory. Of the studies considered in the following discussion, five
13 involved the use of the bolus response method as a probe to obtain information about the
14 mechanism of O3 uptake in the URT and TB regions. Of the remaining two investigations, one
15 focused on total uptake by the RT and the other on uptake by the nasal cavities. Table AX4-1
16 provides a summary of the newer studies.
17
18 AX4.2.1 Bolus-Response Studies
19 The bolus-response method has been used as a probe to explore the effects of physiological
20 and anatomical differences or changes on the uptake of O3 by human beings.
21 Asplund et al. (1996) studied the effects of continuous O3 inhalation on O3 uptake and
22 Rigas et al. (1997) investigated the potential effects of continuous coexposure to O3, nitrogen
23 dioxide (NO2), or sulfur dioxide (SO2) on O3 absorption. In both of these studies, subjects were
24 exposed "continuously" to a gas for 2 h. Every 30 min, breathing at 250 mL/s, a series of bolus
25 test breaths was performed targeted at the lower conducting airways. Differences in bolus-
26 response absorbed fraction from an established baseline indicated the degree to which the
27 "continuous" gas exposure affected the absorption of O3. Depending on the gas and
28 concentration, changes in absorbed fraction ranged from -3 to +7 % (see Table 5-1).
29 Continuous O3 exposure lowered the uptake of O3, whereas NO2 and SO2 increased the uptake of
30 O3. The investigators concluded that in the tested airways, NO2 and SO2 increased the capacity
31 to absorb O3 because more of the compounds oxidized by O3 were made available. On the other
January 2005 AX4-3 DRAFT-DO NOT QUOTE OR CITE
-------
3
to
o
Table AX4-1. New Experimental Human Studies on Ozone Dosimetrya
X
H
6
o
o
H
O
Purpose/Objective
Determine the effect
of continuous O3
inhalation on O3
uptake
Evaluate the
influence of VD on
intersubject variation
of O3 dose.
Investigate the effect
of continuous
exposure to O3,
nitrogen dioxide and
sulfur dioxide on O3
absorption.
Subject
Characteristics
8 male,
3 female,
22-3 1 years old,
166-186 cm,
64-93 kg
10 male,
22-30 years old,
163-186 cm,
64-92 kg;
10 female,
22-35 years old,
149-177 cm,
48-81 kg
6 male, 2 1-29 years
old, 165 185 cm,
60-92 kg;
6 female,
19-33 years old,
152-173 cm,
48-61 kg
Region of
Interest
Central
conducting
airways
(70-120 mL
from lips)
Conducting
airways
Lower
conducting
airways
(70-120 mL
from lips)
Breathing
Patterns/Exposure
2 h of continuous exposure at
rest: 0.0, 0.12, and 0.36 ppm
O3. Spontaneous breathing.
Bolus test breaths every
30 minutes using 250 mL/sec
constant flow rate.
Bolus-response test
(VT = 500ml at 250 mL/sec
constant flow rate). Fowler
single-breath N2 washout
method to determine VD.
2 h of continuous exposure
at rest: O3 (0, 0.36 ppm),
SO2(0, 0.36 ppm), orNO2
(0, 0.36, 0.75 ppm).
5-min Bolus test every
30 minutes: VT = 500 ml;
250 mL/sec constant flow
rate.
Results Reference
Averaged over all subjects and the Asplund et al.
4 measurement intervals, the absorbed fraction (1996)
(AF) changed +0.04, -0.005, and -0.03 for the
0, 0. 12, and 0.36 ppm continuous exposures,
respectively. These changes are
approximately +6, -1, and -4 % based on an
average AF value of 0.7 in the range
70 -120 ml.b Both non zero exposures were
significantly different than the air exposure.
On average, for the same VP , women had a Bush et al.
larger AF than men; women had a smaller VD ( 1 996a)
than men. However, for the same value of
VP/VD, AF for men and women were
indistinguishable. Further analysis indicated
"that previously reported gender differences
may be due to a failure in properly accounting
for tissue surface within the conducting
airways".
Averaging over all subjects or by gender, all Rigas et al.
exposures except O3 resulted in an increase of (1997)
AF. Based on an AF reference valueb, the
change in AF ranged from -3 to +7 %. Only
the O3 and the NO2 (0.36 ppm) exposures
were significantly different from the air
exposures.
o
HH
H
W
-------
Table AX4-1 (cont'd). New Experimental Human Studies on Ozone Dosimetry'
1
to
o
o
Purpose/Objective
Compare the
absorption of
chlorine and O3.
Determine how the
physical-chemical
properties of these
compounds affect
their uptake
distribution in the RT
Subject
Characteristics
5 male, 2 1-26 years
old, 168-198 cm,
64-95 kg; 5 female,
18-28 years old,
162-178 cm,
55-68 kgc
Region of
Interest
Conducting
airways.
Nasal & oral
routes
Breathing
Patterns/Exposure
Bolus-response technique;
VT = 500 ml; 3 constant flow
rates: 150, 250, and
1000 mL/sec.
Results
Ozone dose to the URT was sensitive to the
mode of breathing and to the respiratory rate.
With increased airflow rate, O3 retained by the
upper airways decreased from 95 to 50%.
TB region dose ranged from 0 to 35%.
Mass transfer theory indicated that the
diffusion resistance of the tissue phase is
important for O3. The gas phase resistances
were the same for O3 and C12.
Reference
Nodelman
and Ultman
(1999a)c
>
To determine O3
uptake relative to
inhaled O3 dose.
5 male,
5 female,
18-35 years old,
175 ± 13 (SD) cm,
72 ± 13 (SD) kg
Respiratory Breath-by-breath calculation
tract; oral of O3 retention based on data
breathing from fast response analyzers
for O3 and airflow rates. Oral
breathing: 0.2 or 0.4 ppm O3
at VE of approximately 20
L/min for 60 min or 40 L/min
for 30 min.
The FA for all breaths was 0.86.
Concentration, minute volume, and time have
small but statistically significant effects on AF
when compared to intersubject variability.
The investigators concluded: for a given
subject, constant O3 exposure, a given exercise
level, and time < 2 h, inhaled dose is a
reasonable surrogate for the actual uptake of
O3. However, the actual doses may vary
considerably among individuals who are
exposed to similar inhaled doses.
Rigas et al.
(2000)
H
6
o
o
H
O
o
HH
H
W
-------
Table AX4-1 (cont'd). New Experimental Human Studies on Ozone Dosimetry!
to
o
o
Purpose/Objective
Subject
Characteristics
Region of
Interest
Breathing
Patterns/Exposure
Results
Reference
Study the effect of
gas flow rate and O3
concentration on O3
uptake in the nose.
7 male, 3 female,
26 ± years,
170 ± 11 (SD)cm,
75 ± 20 (SD) kg
Nasal For a given flow rate and
cavities exposure concentration, the
subjects inhaled through one
nostril and exhale through the
other. For two 1-h sessions,
a series of 9-12 measurements
of AF were carried out for
10 s each: (1) O3 exposure
concentration = 0.4 ppm; flow
rates = 3, 5, 8, and 15 L/min.
(2) O3 exposure = 0.1, 0.2, and
0.4 ppm; flow rate =
15 L/min. (3) O3 exposure
= 0.4 ppm, flow rate
= 15 L/min; AF was measured
every 5 min for 1 h.
(1) With the exposure concentration at
0.4 ppm O3, AF decreased from 0.80 to
0.33 when the flow rate was increased from
3 to 15 L/min. (2) At a flow rate of 15 L/min,
the AF changed from 0.36 to 0.32 when the
exposure concentration increased from 0.1 to
0.4 ppm O3. (3) Statistical analysis indicated
that the AF was not associated with the time at
which the measurement was taken.
Santiago
etal. (2001)
H
6
o
o
H
O
Evaluate intersubject
variability in O3
uptake; correlate
differences in
breathing pattern and
lung anatomy with O3
uptake
nonsmokers,
32 male, 22.9 ± 0.8
years old,178±l cm,
80.6 ± 2.5kg;
28 female,
22.4 ± 0.9 years old,
166 ± 1 cm,
62.1 ±2.2 kg
Respiratory Continuous: 1 h exposure to
tract; oral 0.25 ppm, exercising at
breathing 30L/min.
Bolus: breath-by- breath
calculation of O3 retention.
Timing of bolus varied to
create penetration volumes of
10 to 250 ml. Peak inhaled
bolus of ~1 ppm.
Continuous: Fractional O3 uptake efficiency
ranged from 0.70 to 0.98 (mean 0.89 ± 0.06).
Inverse correlation between uptake and
breathing frequency. Direct correlation
between uptake and tidal volume. Intersubject
differences in forced respiratory responses not
due to differences in O3 uptake.
Bolus: The penetration volume at which 50%
of the bolus was taken up was 90.4 ml in
females and 107 ml in males. Distribution of
O3 shifts distally as the size of the airway
increases.
Ultman et al.
(2004)
o
HH
H
W
a See Appendix A for abbreviations and acronyms.
b Fig. 4, Hu et al. (1994), for the 250 mL/s curve and penetration volume range of 70 - 120 ml; the average AF is approximately 0.7.
0 Subject characteristics are from Nodelman and Ultman (1999b).
-------
1 hand, they conjectured that continuous O3 exposure depleted these compounds, thereby reducing
2 O3 uptake.
3 Bush et al. (1996a) investigated the effect of lung anatomy and gender on O3 absorption in
4 the conducting airways during oral breathing using the bolus-response technique. Absorption
5 was measured using this technique applied to 10 men and 10 women. Anatomy was defined in
6 terms of forced vital capacity (FVC), total lung capacity (TLC), and dead space (VD). The
7 absorbed fraction data were analyzed in terms of a function of penetration volume, airflow rate,
8 and an "intrinsic mass transfer parameter (Ka)", which was determined for each subject and
9 found to be highly correlated with VD, but not with height, weight, age, gender, FVC, or TLC.
10 That is, in all subjects, whether men or women, dosimetry differences could be explained by
11 differences in VD. Based on Hu et al. (1994), where absorbed fraction was determined for
12 several flow rates, Bush et al. (1996a) inferred that Ka was proportional to flow rate/VD. The
13 investigators point out that the applicability of their results may be limited because of their
14 assumptions that Ka was independent of location in the RT and that there was no mucous
15 resistance. They also suggested that the dependence of Ka on flow rate and VD be restricted to
16 flow rates < 1000 mL/s until studies at higher rates have been performed.
17 With flow rates of 150, 250, and 1000 mL/s, Nodelman and Ultman (1999b) used the
18 bolus-response technique to compare the uptake distributions of O3 and chlorine gas (C12), and to
19 investigate how their uptakes were affected by their physical and chemical properties. Ozone
20 dose to the URT was found to be sensitive to the mode of breathing and to the airflow rate. With
21 increased rate, O3 retained by the upper airways decreased from 95 to 50% and TB region dose
22 increased from 0 to 35%. At the highest flow rate only 10% of the O3 reached the pulmonary
23 region. Mass transfer theory indicated that the diffusion resistance of the tissue phase is
24 important for O3. The gas phase resistances were found to be the same for O3 and C12 as
25 expected. These resistances were inversely related to the volumes of the oral and nasal cavities
26 during oral and nasal breathing, respectively.
27 Ultman et al. (2004) used both bolus and continuous exposures to test the hypotheses that
28 differences in O3 uptake in lungs are responsible for variation in O3-induced changes in lung
29 function parameters and that differences in O3 uptake are due to variations in breathing patterns
30 and lung anatomy. Thirty-two males and 28 female nonsmokers were exposed to bolus
31 penetration volumes ranging from 10 to 250 ml, which was determined by the timing of the
January 2005 AX4-7 DRAFT-DO NOT QUOTE OR CITE
-------
1 bolus injection. The subjects controlled their breathing to generate a target respired flow of
2 lOOOml/sec. At this high minute ventilation, there was very little uptake in the upper airway and
3 most of the O3 reached areas where gas exchange takes place. To quantify intersubject
4 differences in O3 bolus uptake, they measured the penetration volume at which 50% of the O3
5 was taken up. Values for penetration volume ranged from 69 to 134 ml and were directly
6 correlated with the subjects' values for anatomic dead space volume. A better correlation was
7 seen when the volume of the upper airways was subtracted. The penetration volume at which
8 50% of the bolus was taken up was 90.4 ml in females and 107 ml in males. This significant
9 difference in uptake suggests to the authors that in females the smaller airways, and associated
10 larger surface-to-volume ratio, enhance local O3 uptake and cause reduced penetration of O3 into
11 the distal lung. Thus, these findings indicate that overall O3 uptake is not related to airway size,
12 but that the distribution of O3 shifts distally as the size of the airway in increased.
13
14 General comment on estimating mass transfer coefficients. Bush et al. (1996b) and Nodelman
15 and Ultman (1999a) used a simple model to analyze their bolus-response data. This model
16 presented by Hu et al. (1992, 1994) assumed steady-state mass transfer by convection (but no
17 dispersion) and the mass transfer of O3 to the walls of a tube of uniform cross-sectional area.
18 These assumptions led to an analytical solution (for the absorbed fraction) which was a function
19 of an "overall mass transfer coefficient," penetration volume, and airflow rate. As the
20 investigators have shown, the model is very useful for statistical analysis and hypothesis testing.
21 Given the absorbed fraction data, the model overall mass transfer coefficients were estimated for
22 each flow rate. In those bolus-response studies that used this method to analyze data, there was
23 no discussion of the models' "accuracy" in representing mass transfer in the human respiratory
24 tract with respect to omitting dispersion. In addition, the formulation of the gas phase mass
25 transfer coefficient does not take into account that it has a theoretical lower limit greater than
26 zero as the airflow rate goes to zero (Miller et al., 1985; Bush et al., 2001). As a consequence,
27 there is no way to judge the usefulness of the values of the estimated mass transfer coefficients
28 for dosimetry simulations that are based on convection-dispersion equations, or whether or not
29 the simple model's mass transfer coefficients, as well as other parameters derived using these
30 coefficients, are the same as actual physiological parameters.
31
January 2005 AX4-8 DRAFT-DO NOT QUOTE OR CITE
-------
1 AX4.2.2 General Uptake Studies
2 Rigas et al. (2000) performed an experiment to determine the ratio of O3 uptake to the
3 quantity of O3 inhaled (fractional absorption, FA). Five men and five women were exposed
4 orally to 0.2 or 0.4 ppm O3 while exercising at a minute volume of approximately 20 L/min for
5 60 minutes or 40 L/min for 30 minutes. Ozone retention was calculated from breath-by-breath
6 data taken from fast response analyzers of O3 and airflow rates. The FA was statistically
7 analyzed in terms of subject, exposure concentration, minute volume, and exposure time.
8 Fractional absorption ranged from 0.56 to 0.98 with a mean ± SD of 0.85 ± 0.06 for all
9 recorded breaths. Intersubject differences had the largest influence on FA, resulting in a
10 variation of approximately 10%. Statistical analysis indicated that concentration, minute
11 volume, and exposure time had statistically significant effects on FA. However, relatively large
12 changes in these variables were estimated to result in relatively small changes in FA. Note: the
13 quantity of O3 retained by the RT is equal to FA times the quantity of O3 inhaled; thus, relatively
14 large changes in concentration, minute volume, or exposure time may result in relatively large
15 changes in the amount of O3 retained by the RT or absorbed locally. Also, according to Overton
16 et al. (1996), difference in PAR dose due to anatomical variability may be considerably larger
17 than corresponding small changes in FA would indicate.
18 Santiago et al. (2001) studied the effects of airflow rate and O3 concentration on O3 uptake
19 in the nasal cavities of three women and seven men. Air was supplied at a constant flow rate to
20 one nostril and exited from the other nostril while the subject kept the velopharyngeal aperture
21 closed by raising the soft palate. Thus, a constant unidirectional flow of air plus O3 was
22 restricted to the nasal cavities. The fraction of O3 absorbed was calculated using the inlet and
23 outlet concentrations. Inlet concentration and airflow rate were varied in order to determine their
24 effect on O3 uptake.
25 The mean FA decreased from 0.80 to 0.33 with an increase in flow rate from 3 to
26 15 L/min. The effect of both flow rate and subject on FA was statistically significant. Further
27 analysis indicated that the overall mass transfer coefficient was highly correlated with the flow
28 rate and that the gas phase resistance contributed from 6.3% (15 L/min) to 23% (3 L/min) of the
29 total resistance to O3 transfer to the nasal cavity surface. Concentration had a small, but
30 statistically significant effect on FA, when the inlet concentration was increased from 0.1 to
31 0.4 ppm O3, FA decreased from 0.36 to 0.32. The investigators observed that differences in FA
January 2005 AX4-9 DRAFT-DO NOT QUOTE OR CITE
-------
1 among subjects were important; generally, subject variability accounted for approximately half
2 of the total variation in FA.
3 As mentioned above Ultman et al. (2004) tested hypotheses that differences in O3 uptake in
4 lungs are responsible for variation in O3-induced changes in lung function parameters and that
5 differences in O3 uptake are due to variations in breathing patterns and lung anatomy. Thirty-
6 two males and 28 female nonsmokers were exposed continuously for 1 h to either clean air or
7 0.25 ppm ozone while exercising at a target minute ventilation of 30 L/min. They first
8 determined the forced expiratory response to clean air, then evaluated O3 uptake measuring dead
9 space volume, cross-sectional area of peripheral lung (Ap) for CO2 diffusion, FEVb FVC, and
10 FEF25o/0.75o/0. The fractional O3 uptake efficiency ranged from 0.70 to 0.98, with a mean of
11 0.89 ± 0.06. They found an inverse correlation between uptake and breathing frequency and a
12 direct correlation between uptake and tidal volume. The uptake efficiency decreased during the
13 four sequential 15 minute intervals of the 1 h exposure (0.906 ± 0.058 to 0.873 ± 0.088, first and
14 last, respectively), demonstrating a general decrease in uptake efficiency with increased
15 breathing frequency and decreasing tidal volume. Ozone uptake rate correlated with individual
16 %Ap, but did not correlate with individual "/oFEVj. Neither of these parameters correlated with
17 the penetration volume determined in the bolus studies mentioned above. The authors concluded
18 that the intersubject differences in forced respiratory responses were not due to differences in O3
19 uptake. However, these data did partially support the second hypothesis, i.e., that the differences
20 in cross-sectional area available for gas diffusion induce differences in O3 uptake.
21
22
23 AX4.3 DOSIMETRY MODELING
24 When all of the animal and human in vivo O3 uptake efficiency data are compared, there is
25 a good degree of consistency across data sets (U.S. Environmental Protection Agency, 1996).
26 This agreement raises the level of confidence with which these data sets can be used to support
27 dosimetric model formulations.
28 Recent data indicate that the primary site of acute cell injury occurs in the conducting
29 airways (Postlethwait et al., 2000). These data must be considered when developing models that
30 attempt to predict site-specific locations of O3-induced injury. The early models computed
31 relationships between delivered regional dose and response with the assumption that O3 was the
January 2005 AX4-10 DRAFT-DO NOT QUOTE OR CITE
-------
1 active agent responsible for injury. It is now known that reactive intermediates such as
2 hydrohydroxyperoxides and aldehydes are important agents mediating the response to O3
3 (further discussed in Section 5.3.1). Thus, models must consider O3 reaction/diffusion in the
4 epithelial lining fluid (ELF) and ELF-derived reactions products.
5 Table AX4-2 presents a summary of new theoretical studies on the uptake of O3 by the RTs
6 (or regions) of humans and laboratory animals that have become available since the 1996 review.
7 They are discussed below.
8 Overton and Graham (1995) described the development and simulation results of a
9 dosimetry model that was applied to a TB region anatomical model that had branching airways,
10 but which had identical single-path pulmonary units distal to each terminal bronchiole. The
11 anatomical model of the TB region was based on Raabe et al. (1976), which reported lung cast
12 data for the TB region of a 330 g rat.
13 Rat effects data (from the PAR) are available that are identified with the lobe and the
14 generation in the lobe from which tissue samples were obtained (Pinkerton et al., 1995, 1998).
15 Models, like Overton et al. (1995), can be helpful in understanding the distribution of the
16 magnitude of such effects as well as suggesting sampling sites for future experiments.
17 Using computational fluid dynamics (CFD), Cohen-Hubal et al. (1996) explored the effect
18 of the mucus layer thickness in the nasal passage of a rat. The nasal lining was composed of
19 mucus and tissue layers in which mass transfer was by molecular diffusion with first order
20 chemical reaction. Physicochemical parameters for O3 were obtained from the literature. Three
21 scenarios were considered: 10 jim thick mucus layer, no mucus layer, and two nasal passage
22 regions each with a different mucus layer thickness. Predictions of overall uptake were within
23 the range of measured uptake. Predicted regional O3 flux was correlated with measured cell
24 proliferation for the CFD simulation that incorporated two regions, each with a different mucus
25 thickness.
26 The reaction rate constant used by Cohen-Hubal and co-workers may be too low. Using
27 bolus-response data, Hu et al. (1994) and Bush et al. (2001) estimated a reaction rate constant
28 that is more than a 1000 times as large as that used by Cohen-Hubal et al. (1996). A rate
29 constant this large could result in a conclusion different than those based on the smaller constant.
30 With an RT dosimetry model, Overton et al. (1996) investigated the sensitivity of absorbed
31 fraction (AF), proximal alveolar region (PAR) dose, and PAR dose ratio to TB region volume
January 2005 AX4-11 DRAFT-DO NOT QUOTE OR CITE
-------
Table AX4-2. New Ozone Dosimetry Model Investigations"
to
o
o
Purpose/Objective
Type of mass transport
model/Anatomical modelb
Species/ RT region
of interest/Regional
anatomical models
Ventilation and
Exposure
Results
Reference
to
H
6
o
o
H
O
To describe an RT
dosimetry model that
uses a branching TB
region anatomical
model and to
illustrate the results
of its application to a
rat exposed to O3.
To incorporate into
the CFD model of
Kimbelletal. (1993)
resistance to mass
transfer in the nasal
lining and to
investigate the
effects of this lining
on O3 uptake.
To determine if the
single-path model is
able to simulate
bolus inhalation data
recorded during oral
breathing at quiet
respiratory flow.
One-dimensional (along axis
of airflow), time-dependent,
convection-dispersion
equation of mass transport
applied to each airway or
model segment. URT: single
path; TB: asymmetric
branching airways. PUL:
single path anatomical model
distal to each terminal
bronchiole.
Three dimensional steady-
state Navier-Stokes equations
for solving air velocity flow
field. Three dimensional
steady-state convection-
diffusion equation for O3
transport.
Three-dimensional CFD
model of the nasal passages
of a rat.
Single-path, one-dimensional
(along axis of airflow),
time-dependent, convection-
dispersion equation of mass
transport. Single-path
anatomical model
Rat/ RT/URT:
Patraetal. (1987).
TB: multiple path
model of Raabe et al.
(1976). PUL:
Mercer etal. (1991).
Rat/nasal passages
Nasal passages:
Kimbelletal. (1993).
Human/ RT/URT
(oral): Olson et al.
(1973).
LRT: Weibel(1963).
f= ISObpm;
VT=1.5,2.0,
2.5 mL.
One constant
concentration.
Steady-state
unidirectional
("inhalation")
flow rate
= 576mLl/min.
One constant
concentration.
VT = 500 mL,
f= 15bpm,
constant flow
rate = 250 mL/s.
Bolus-response
simulations.
(protocol used is
described by Hu
etal., 1992).
(1) For VT = 2.0 mL, f = 150 bpm: The general shape
of the dose versus generation plot along any path from
the trachea to a sac is independent of path: generally
the tissue dose decreases with increasing generation
index. In the TB region, the coefficient of variation for
dose ranges from 0 to 34 %, depending on generation.
The maximum ratio of the largest to smallest dose in
the same generation is 7; the average ratio being 3. In
the first PUL region model segment, the coefficient of
variation for the dose is 29 %. (2) The average dose to
the first PUL region model segment increases with
increasing VT.
Predictions of overall uptake were within the range of
measured uptake. Results suggest that mucus resistance
is important for describing O3 dosimetry and this
thickness may play a role for determining patterns of
O3-induced lesions in the rat nasal passage.
Simulations are sensitive to conducting airway volume
but are relative insensitive to characteristics of the
respiratory airspace. Although the gas-phase resistance
to lateral diffusion limits O3 absorption during quiet
breathing, diffusion through mucus may become
important at the large respiratory flows that are
normally associated with exercise. The single-path
convection-diffusion model was a reasonable approach
/to simulate the bolus-response data.
Overton and
Graham
(1995)
Cohen-
Hubal et al.
(1996)
Bush et al.
(1996b)
o
HH
H
W
-------
Table AX4-2 (cont'd). New Ozone Dosimetry Model Investigations a
c
1
to
o
o
Purpose/Objective
To assess age- and
gender-specific
differences in
regional and
systemic uptake.
Type of mass transport
model/Anatomical model b
PBPK, at ages 1, 3, 6, months
and 1,5, 10, 15,25, 50, and
75 years
Species/ RT region
of interest/Regional
anatomical models
Human/ET/TB
Ventilation and
Exposure
Pulmonary
ventilation ranged
from 34 mL/s (in
1 -month-old) to
190 mL/s (in
15-year-old). VT
varied with age
Results
Regional extraction is insensitive to age. Extraction per
unit surface area is 2- to 8-fold higher in infants
compared to adults. PU and ET regions have a large
increase in unit extraction with increasing age. Early
postnatal period is time of largest differences in PK, due
to immaturity of metabolic enzymes.
Reference
Sarangapani
et al. (2003)
H
6
o
o
H
O
To examine the
impact on predictions
due to the value used
for the TB region
volume at FRC and
due to TB region
volume change
during respiration.
To make parameter
modifications so that
a single-path model
would simulate AF
from bolus-response
experiments
involving O3
(and C12).
Single-path, one-dimensional
(along axis of airflow),
time-dependent, convection-
dispersion equation of mass
transport. Single-path
anatomical model
Single-path, one-dimensional
(along axis of airflow), time-
dependent, convection-
dispersion equation of mass
transport. Single-path
anatomical model
Human /RT/
URT: Nunnetal.
(1959)
LRT: Weibel(1963)
Rat/RT/
URT: Patraetal.
(1987)
TB: Yehetal. (1979)
PUL: Mercer etal.
(1991)
Human/ RT/URT
(oral): Olson et al.
(1973).
URT (nasal):
Olson etal. (1973)
and Guilmette et al.
(1989)
LRT: Weibel(1963).
Human: VT = 500,
2250mL;f=15,
30 bpm.
Rat: VT=1.4,
2.4 ml; f= 96,
157 bpm.
One constant
concentration.
Oral & nasal
breathing. Flow
rates = 150,250,
1000 mL/s,
VT = 500 mLl.
Bolus-response
simulations
(1) A better understanding and characterization of the
role of TB region expansion (mainly the rat) and
volume is important for an improved understanding of
respiratory-tract dosimetry modeling of reactive gases.
(2) Extrapolations based on dose in the PAR can differ
significantly from those based on exposure
concentration or total uptake.
(3) Human subjects who appear similar outwardly may
have very different PAR doses and potentially different
responses to the same exposure.(Uptake by the URT
was not considered.)
(Simulation results for O3 only) (1) Using parameter
values from the literature and assuming that absorption
was gas-phase controlled, the simulations of O3 data
were realistic at flow rate = 250 mL/s, but not realistic
at 1000 mL/s.(2) Accurate simulations at 250 mL/s
required modification of mass transfer coefficients
reported in the literature for the conducting airways.(3)
It was necessary to include a diffusion resistance for the
epithelial lining fluid based on an assumed O3 reaction
rate constant that was much greater than in in vitro
estimates.(4) Partial validation of the final parameters
(determined at 250 mL/s) was obtained by simulations
of bolus-response data at flow rates of 150 and 1000
mL/s. Validation was obtained also by simulating
internal measurements of O3 in subjects exposed during
quiet breathing.
Overton
etal. (1996)
Bush et al.
(2001)
o
HH
H
W
"See Appendix A for abbreviations and acronyms.
bThe anatomical models used in an investigation generally differ from those described in the references, e.g., dimensions are often scaled to dimensions appropriate to the
dosimetry investigation; or the original structure may be simplified, keeping or scaling the original dimensions.
-------
1 (VTB) and TB region expansion in human beings and rats. The PAR was defined as the first
2 generation distal to terminal bronchioles and the PAR dose ratio was defined as the ratio of a
3 rat's predicted PAR dose to a human's predicted PAR dose. This ratio relates human and rat
4 exposure concentrations so that both species receive the same PAR dose. In rats the PAR is a
5 region of major damage from O3. For each species, three literature values of V^ were used:
6 a mean value and the mean ± twice the SD. The following predictions were obtained:
7 (1) The sensitivity of AF and PAR dose to V^ depends on species, ventilation, TB region
8 overall mass transfer coefficient (!%,), and expansion. Depending on these latter four
9 parameters, AF was predicted to be 1 to 25% smaller and 1 to 40% larger than the values
10 predicted for the mean V^, given the range of V^. However, AF can be insensitive to Vra and
11 PAR dose very sensitive. For kTB = 0.26 cm/s and quiet breathing, AF was predicted to vary by
12 less than 3% for the ±2 SD range of VTB; in contrast, the PAR dose predicted using the smallest
13 VXB is five times larger than the PAR dose predicted with the largest V^. The effect of VTB is
14 much less during heavy exercise: the ratio of maximum to minimum PAR dose was
15 approximately 1.5. In any case, the simulations predicted that fractional changes in AF due to
16 different V^ are not, in general, a good predictor of the fractional changes in PAR doses.
17 (2) Relative to no expansion in the TB region, expansion decreases both AF and PAR
18 dose. The largest effect of including expansion in the human simulations was to decrease the AF
19 by «8%; in rats, the maximum decrease was -45%. The PAR doses decreased relatively more,
20 25 and 65% in human beings and rat, respectively.
21 (3) The authors attempted to obtain an understanding as to uncertainty or variability in
22 estimates of exposure concentrations (that give the same PAR dose in both species) if the
23 literature mean value of VTB was used. For various values off, VT, kra, and expansion, the PAR
24 dose ratios at upper and lower values of VTB deviated in absolute values from the PAR dose ratio
25 calculated at the mean values of VTB by as little as 10% to as large as 310%. The smallest
26 deviation occurred at the largest VT and smallest kra for both species; whereas, the largest
27 deviation occurred at the smallest VT and largest kra for both species.
28 Bush et al. (2001) modified the single-path model of Bush et al. (1996b) in order to be able
29 to simulate absorbed fraction data for O3 (and C12, which is not considered) for three airflow
30 rates and for oral and nasal breathing. By adjusting several parameters a reasonable agreement
31 between predicted and experimental values was obtained. On the other hand, the O3 plots of the
January 2005 AX4-14 DRAFT-DO NOT QUOTE OR CITE
-------
1 experimental and predicted values of absorbed fraction versus penetration volume (e.g., Figures
2 4 and 5 of Bush et al., 2001) show sequential groups composed of only positive or only negative
3 residuals, indicating a lack of fit. Possibly adjusting other parameters would eliminate this.
4 To obtain an independent validation of the model, Bush et al. (2001) simulated measurements of
5 O3 concentrations made by Gerrity et al. (1995) during both inhalation and exhalation at four
6 locations between the mouth and the bronchus intermedius of human subjects. Simulated and
7 experimental values obtained are in close agreement. Note, however, that Bush et al. made no
8 quantitative assessment of how well their simulations agreed with the experimental data;
9 assessments were made on the basis of visual inspection of experimental and simulated values
10 plotted on the same figure. Thus, evaluation of the model was, or is, subjective.
11 Recently Sarangapani et al. (2003) used physiologically based pharmacokinetic (PBPK)
12 modeling to characterize age- and gender-specific differences in both regional and systemic
13 uptake of O3 in humans. This model indicated that regional extraction of O3 is relatively
14 insensitive to age, but extraction per unit surface area is 2- to 8-fold higher in infants compared
15 to adults, due to the region-specific mass transfer coefficient not varying with age. The PU and
16 ET regions have a large increase in unit extraction with increasing age because both regions
17 increase in surface area. Males and females in this model have similar trends in regional
18 extraction and regional unit extraction. In early childhood, dose metrics were as much as
19 12 times higher than adult levels, but these differences leveled out with age, such that inhalation
20 exposures varied little after age 5. These data suggest that the early postnatal period is the time
21 of the largest difference in pharmacokinetics observed, and this difference is primarily due to the
22 immaturity of the metabolic enzymes used to clear O3 from the respiratory tract.
23 Mudway and Kelly (2004) attempted to model O3 dose-inflammatory response using a
24 metaanalysis of 23 exposures in published human chamber studies. The O3 concentrations
25 ranged from 0.08 to 0.6 ppm and the exposure durations ranged from 60 to 396 minutes. The
26 analysis showed linear relationships between O3 dose and neutrophilia in bronchoalveolar lavage
27 fluid (BALF). Linear relationships were also observed between O3 dose and protein leakage
28 into BALF.
29
30
January 2005 AX4-15 DRAFT-DO NOT QUOTE OR CITE
-------
1 AX4.4 SPECIES HOMOLOGY, SENSITIVITY AND ANIMAL-TO-
2 HUMAN EXTRAPOLATION
3 Biochemical differences among species are becoming increasingly apparent and these
4 differences may factor into a species' susceptibility to the effects of O3 exposure. Lee et al.
5 (1998) compared SD rats and rhesus monkeys to ascertain species differences in the various
6 isoforms of CYP moonoxygenases in response to O3 exposure (discussed in more detail in
7 Section 5.3.1.2). Differences in activities between rat and monkey were 2- to 10-fold, depending
8 on the isoform and the specific lung region assayed. This study supports the view that
9 differential expression of CYPs is a key factor in determining the toxicity of O3. As further
10 characterization of species- and region-specific CYP enzymes occurs, a greater understanding of
11 the differences in response may allow more accurate extrapolation from animal exposures to
12 human exposures and toxic effects.
13 Arsalane et al. (1995) compared guinea pig and human AM recovered in BALF and
14 subsequently exposed in vitro to 0.1 to 1 ppm for 60 minutes. Measurement of inflammatory
15 cytokines showed a peak at 0.4 ppm in both species. Guinea pig AM had an increase in IL-6 and
16 TNF-a while human AM had increases in TNF-a, IL-lb, IL-6 and IL-8. This exposure also
17 caused an increase in mRNA expression for TNF-a, IL-lb, IL-6 and IL-8 in human cells.
18 At 0.1 ppm exposures, only TNF-a secretion was increased. These data suggest similar cytokine
19 responses in guinea pigs and humans, both qualitatively and quantitatively.
20 Dormans et al. (1999) continuously exposed rats, mice, male guinea pigs to filtered air, 0.2,
21 or 0.4 ppm for 3, to 56 days or to 28 days with 3, 7, and 28 days PE. Depending on the endpoint
22 studied, the species varied in sensitivity. Greater sensitivity was shown in the mouse as
23 determined by biochemical endpoints, persistence of bronchiolar epithelial hypertrophy, and
24 recovery time. Guinea pigs were more sensitive in terms of the inflammatory response though
25 all three species had increases in the inflammatory response after three days that did not decrease
26 with exposure. In all species the longest exposure to the highest dose caused increased collagen
27 in ductal septa and large lamellar bodies in Type II cells, but that response also occurred in rats
28 and guinea pigs at 0.2 ppm. No fibrosis was seen at the shorter exposure times and the authors
29 question whether fibrosis occurs in healthy humans after continuous exposure. The authors do
30 not rule out the possibility that some of these differences may be attributable to differences in
January 2005 AX4-16 DRAFT-DO NOT QUOTE OR CITE
-------
1 total inhaled dose or dose actually reaching a target site. Overall, the authors rated mice as most
2 susceptible, followed by guinea pigs and rats.
3 Comparisons of airway effects in rats, monkeys and ferrets resulting from exposures of
4 1.0 ppm O3 for 8 h (Sterner-Kock et al. 2000) demonstrated that monkeys and ferrets had a
5 similar inflammatory responses and epithelial necrosis. The response of these two species was
6 more severe than that seen in rats. These data suggest that ferrets are a good animal model for
7 O3-induced airway effects due to the similarities in pulmonary structure between primates and
8 ferrets.
9 The rat is a key species used in O3 toxicological studies, but Watkinson and Gordon,
10 (1993) suggest that, because the rat has both behavioral and physiological mechanisms that can
11 lower core temperature in response to acute exposures, extrapolation of these exposure data to
12 humans may be limited. Another laboratory (Iwasaki et al., 1998) has demonstrated both
13 cardiovascular and thermoregulatory responses to O3 at exposure to 0.1, 0.3, and 0.5 ppm O3
14 8 h/day for 4 consecutive days. A dose-dependent disruption of HR and Tco were seen on the
15 first and second days of exposure, which then recovered to control values. Watkinson et al.
16 (2003) exposed rats to 0.5 ppm O3 and observed this hypothermic response which included
17 lowered HR, lowered Tco, and increased inflammatory components in BALF. The authors
18 suggest that the response is an inherent reflexive pattern that can possibly attenuate O3 toxicity in
19 rodents. They discuss the cascade of effects created by decreases in Tco, which include:
20 (1) lowered metabolic rate, (2) altered enzyme kinetics, (3) altered membrane function,
21 (4) decreased oxygen consumption and demand, (5) reductions in minute ventilation, which
22 would act to limit the dose of O3 delivered to the lungs. These effects are concurrent with
23 changes in HR which lead to: (1) decreased CO, (2) lowered BP, (3) decreased tissue perfusion,
24 all of which may lead to functional deficits. The hypothermic response has not been observed in
25 humans except at very high exposures.
26
January 2005 AX4-17 DRAFT-DO NOT QUOTE OR CITE
-------
1 REFERENCES
2 Arsalane, K.; Cosset, P.; Vanhee, D.; Voisin, C.; Hamid, Q.; Tonnel, A.-B.; Wallaert, B. (1995) Ozone stimulates
3 synthesis of inflammatory cytokines by alveolar macrophages in vitro. Am. J. Respir. Cell Mol. Biol.
4 13:60-68.
5 Asplund, P. T.; Ben-Jebria, A.; Rigas, M. L.; Ultman, J. S. (1996) Longitudinal distribution of ozone absorption in
6 the lung: effect of continuous inhalation exposure. Arch. Environ. Health 51: 431-438.
7 Bush, M. L.; Asplund, P. T.; Miles, K. A.; Ben-Jebria, A.; Ultman, J. S. (1996a) Longitudinal distribution of O3
8 absorption in the lung: gender differences and intersubject variability. J. Appl. Physiol. 81: 1651-1657.
9 Bush, M. L.; Raybold, T.; Abeles, S.; Hu, S.-C; Ben-Jebria, A.; Ultman, J. S. (1996b) Longitudinal distribution of
10 ozone absorption in the lung: simulation with a single-path model. Toxicol. Appl. Pharmacol. 140: 219-226.
11 Bush, M. L.; Zhang, W.; Ben-Jebria, A.; Ultman, J. S. (2001) Longitudinal distribution of ozone and chlorine in the
12 human respiratory tract: simulation of nasal and oral breathing with the single-path diffusion model. Toxicol.
13 Appl. Pharmacol. 173: 137-145.
14 Cohen-Hubal, E. A.; Kimbell, J. S.; Fedkiw, P. S. (1996) Incorporation of nasal-lining mass-transfer resistance into a
15 CFD model for prediction of ozone dosimetry in the upper respiratory tract. Inhalation Toxicol. 8: 831-857.
16 Dahl, A. R. (1990) Dose concepts for inhaled vapors and gases. Toxicol. Appl. Pharmacol. 103: 185-197.
17 Dormans, J. A. M. A.; VanBree, L.; Boere, A. J. F.; Marra, M.; Rombout, P. J. A. (1999) Interspecies differences in
18 time course of pulmonary toxicity following repeated exposure to ozone. Inhalation Toxicol. 11: 309-329.
19 Gerrity, T. R.; Biscardi, F.; Strong, A.; Garlington, A. R.; Brown, J. S.; Bromberg, P. A. (1995) Bronchoscopic
20 determination of ozone uptake in humans. J. Appl. Physiol. 79: 852-860.
21 Hu, S. C.; Ben-Jebria, A.; Ultman, J. S. (1992) Longitudinal distribution of ozone absorption in the lung: quiet
22 respiration in healthy subjects. J. Appl. Physiol. 73: 1655-1667.
23 Hu, S.-C.; Ben-Jebria, A.; Ultman, J. S. (1994) Longitudinal distribution of ozone absorption in the lung: effects of
24 respiratory flow. J. Appl. Physiol. 77: 574-583.
25 Iwasaki, T.; Takahashi, M.; Saito, H.; Arito, H. (1998) Adaptation of extrapulmonary responses to ozone exposure in
26 conscious rats. Ind. Health 36: 57-60.
27 Lee, C.; Watt, K. C.; Chang, A. M.; Plopper, C. G.; Buckpitt, A. R.; Pinkerton, K. E. (1998) Site-selective
28 differences in cytochrome P450 isoform activities: comparison of expression in rat and rhesus monkey lung
29 and induction in rats. Drug Metab. Dispos. 26: 396-400.
3 0 Miller, F. J. (1995) Uptake and fate of ozone in the respiratory tract. Toxicol. Lett. 82/83: 277-285.
31 Miller, F. J.; Overton, J. H., Jr.; Jaskot, R. H.; Menzel, D. B. (1985) A model of the regional uptake of gaseous
3 2 pollutants in the lung: I. the sensitivity of the uptake of ozone in the human lung to lower respiratory tract
33 secretions and exercise. Toxicol. Appl. Pharmacol. 79: 11-27.
34 Mudway, I. S.; Kelly, F. J. (2004) An investigation of inhaled ozone dose and the magnitude of airway inflammation
35 in healthy adults. Am. J. Respir. Crit. Care Med. 169: 1089-1095.
36 Nodelman, V.; Ultman, J. S. (1999) Longitudinal distribution of chlorine absorption in human airways: comparison
37 of nasal and oral quiet breathing. J. Appl. Physiol. 86: 1984-1993.
38 Overton, J. H.; Graham, R. C. (1995) Simulation of the uptake of a reactive gas in a rat respiratory tract model with
39 an asymmetric tracheobronchial region patterned on complete conducting airway cast data. Comput. Biomed.
40 Res. 28: 171-190.
41 Overton, J. H.; Graham, R. C.; Menache, M. G.; Mercer, R. R.; Miller, F. J. (1996) Influence of tracheobronchial
42 region expansion and volume on reactive gas uptake and interspecies dose extrapolations. Inhalation Toxicol.
43 8:723-745.
44 Pinkerton, K. E.; Menache, M. G.; Plopper, C. G. (1995) Consequences of prolonged inhalation of ozone on F344/N
45 rats: collaborative studies. Part IX. Changes in the tracheobronchial epithelium, pulmonary acinus, and lung
46 antioxidant enzyme activity. Cambridge, MA: Health Effects Institute; pp. 41-98; research report no. 65.
47 Available from: NTIS, Springfield, VA; PB95-261996.
48 Pinkerton, K. E.; Weller, B. L.; Menache, M. G.; Plopper, C. G. (1998) Consequences of prolonged inhalation of
49 ozone on F344/N rats: collaborative studies. Part XIII. A comparison of changes in the tracheobronchial
50 epithelium and pulmonary acinus in male rats at 3 and 20 months. Cambridge, MA: Health Effects Institute;
51 research report no. 65.
52 Postlethwait, E. M.; Joad, J. P.; Hyde, D. M.; Schelegle, E. S.; Brie, J. M.; Weir, A. J.; Putney, L. F.; Wong, V. J.;
53 Velsor, L. W.; Plopper, C. G. (2000) Three-dimensional mapping of ozone-induced acute cytotoxicity in
54 tracheobronchial airways of isolated perfused rat lung. Am. J. Respir. Cell Mol. Biol. 22: 191-199.
January 2005 AX4-18 DRAFT-DO NOT QUOTE OR CITE
-------
1 Raabe, O. G.; Yeh, H. C.; Schum, G. M; Phalen, R. F. (1976) Tracheobronchial geometry: human, dog, rat,
2 hamster. Albuquerque, NM: Lovelace Foundation; report no. LF-53.
3 Rigas, M. L.; Ben-Jebria, A.; Ultman, J. S. (1997) Longitudinal distribution of ozone absorption in the lung: effects
4 of nitrogen dioxide, sulfur dioxide, and ozone exposures. Arch. Environ. Health 52: 173-178.
5 Rigas, M. L.; Catlin, S. N.; Ben-Jebria, A.; Ultman, J. S. (2000) Ozone uptake in the intact human respiratory tract:
6 relationship between inhaled dose and actual dose. J. Appl. Physiol. 88: 2015-2022.
7 Santiago, L. Y.; Harm, M. C.; Ben-Jebria, A.; Ultman, J. S. (2001) Ozone adsorption in the human nose during
8 unidirectional airflow. J. Appl. Physiol. 91: 725-732.
9 Sarangapani, R.; Gentry, P. R.; Covington, T. R.; Teeguarden, J. G.; Clewell, H. J., III. (2003) Evaluation of the
10 potential impact of age- and gender-specific lung morphology and ventilation rate on the dosimetry of vapors.
11 Inhalation Toxicol. 15: 987-1016.
12 Sterner-Kock, A.; Kock, M.; Braun, R.; Hyde, D. M. (2000) Ozone-induced epithelial injury in the ferret is similar
13 to nonhuman primates. Am. J. Respir. Crit. Care Med. 162: 1152-1156.
14 U.S. Environmental Protection Agency. (1996) Air quality criteria for ozone and related photochemical oxidants.
15 Research Triangle Park, NC: Office of Research and Development; report nos. EPA/600/AP-93/004aF-cF. 3v.
16 Available from: NTIS, Springfield, VA; PB96-185582, PB96-185590, and PB96-185608. Available online at:
17 www.epa.gov/ncea/ozone.htm.
18 Watkinson, W. P.; Aileru, A. A.; Dowd, S. M.; Doerfler, D. L.; Tepper, J. S.; Costa, D. L. (1993) Acute effects of
19 ozone on heart rate and body temperature in the unanesthetized, unrestrained rat maintained at different
20 ambient temperatures. Inhalation Toxicol. 5: 129-147.
21 Watkinson, W. P.; Campen, M. J.; Wichers, L. B.; Nolan, J. P.; Costa, D. L. (2003) Cardiac and thermoregulatory
22 responses to inhaled pollutants in healthy and compromised rodents: modulation via interaction with
23 environmental factors. Environ. Res. 92: 35-47.
24
25
January 2005 AX4-19 DRAFT-DO NOT QUOTE OR CITE
-------
i 5. TOXICOLOGICAL EFFECTS OF OZONE AND
2 RELATED PHOTOCHEMICAL OXIDANTS IN
3 LABORATORY ANIMALS AND IN VITRO
4 TEST SYSTEMS
5
6
7 5.1 INTRODUCTION
8 A wide range of effects of ozone (O3) has been demonstrated in laboratory animals. The
9 major research findings are that environmentally relevant levels of O3 cause lung inflammation;
10 decreases in host defenses against infectious lung disease; acute changes in lung function,
11 structure, and metabolism; chronic lung disease, some elements of which are irreversible; and
12 systemic effects on target organs (e.g., brain, heart, liver, immune system) distant from the lung.
13 The research also has served to expand the understanding of mechanisms of O3 toxicity and the
14 relationships between concentration and duration of exposure.
15 The framework for presenting the health effects of O3 in animals begins with a presentation
16 of respiratory tract effects, followed by systemic effects, and then interactions of O3 with other
17 common co-occurring pollutants. The information discussed in this chapter is founded on a very
18 wide body of literature on studies in laboratory animals and on in vitro test systems of animal
19 cell lines and organ systems that may mimic responses in intact animals. The direct effects of O3
20 in humans are discussed in the following chapter (Chapter 6).
21 This chapter is not intended to be a compendium of all that is known about O3; rather, it is
22 an update of the toxicology chapter from the last O3 criteria document (U.S. Environmental
23 Protection Agency, 1996), or 1996 O3 CD, and other reviews of the earlier published literature.
24 The historical O3 literature is very briefly summarized in an opening paragraph of each section or
25 subsection. This paragraph is intended as a very concise overview of previous work, and the
26 reader is referred to the 1996 O3 CD for more detailed discussion of the literature prior to the
27 early 1990's. Each section then continues with brief discussions of the key new studies (or
28 somewhat older studies that were not included in the previous CD). Longer discussions of new
29 studies are included where warranted. Sections are ended with comparisons of data from the
December 2005 5-1 DRAFT-DO NOT QUOTE OR CITE
-------
1 previous CD with new data and basic conclusions are drawn. More detailed descriptive
2 summaries of new studies and results are provided in text and tables in Annex AX5.
3 Except for nitrogen dioxide (NO2), the subject of another criteria document (U.S.
4 Environmental Protection Agency, 1993), there is very little relevant information on other
5 photochemical oxidants in the published literature. What is known about the effects of these
6 other oxidants is also summarized briefly in this chapter.
7
8
9 5.2 RESPIRATORY TRACT EFFECTS OF OZONE
10 5.2.1 Biochemical Effects
11 Biochemically detected effects of O3 are integrally involved in effects on both structure and
12 function (respiratory and nonrespiratory) of the respiratory tract. Changes in xenobiotic
13 metabolism, antioxidant metabolism and oxygen consumption, lipids and arachidonic acid
14 metabolism, and collagen metabolism are all observed with O3 exposure, though the mechanisms
15 and associations are not fully understood.
16
17 5.2.1.1 Cellular Targets of O3 Interaction
18 Ozone has the potential to interact with a wide range of different cellular components that
19 include polyunsaturated fatty acids (PUFAs); some protein amino acid residues (cysteine,
20 histidine, methionine, and tryptophan); and some low-molecular-weight compounds that include
21 glutathione (GSH), urate, vitamins C and E, and free amino acids. Early work demonstrated that
22 O3 being a highly reactive compound, does not penetrate much beyond the epithelial lining fluid
23 (ELF). Ozone-induced cell damage most likely results from its reactions with PUFAs to form
24 stable but less reactive ozonide, aldehyde, and hydroperoxide reaction products (the reactions
25 are summarized in Figure 5-1 of Annex AX5). These reaction products (Crigee ozonides and
26 hydroxyhydroperoxides) may act as signal transduction molecules involved in signaling of
27 cellular responses such as inflammation, and thus mediate O3 toxicity.
28 These recent reports combined with observations reported in the previous O3 CD (US EPA,
29 1996) suggest that interactions of O3 with cellular components and ELF generate toxic ozonation
30 products and mediate toxic effects through these products.
31
December 2005 5-2 DRAFT-DO NOT QUOTE OR CITE
-------
1 5.2.1.2 Monooxygenases
2 Both short- and long-term exposures to O3 have been shown to enhance lung xenobiotic
3 metabolism, possibly as a result of changes in the number and function of bronchiolar epithelial
4 Clara cells and alveolar epithelial Type 2 cells. Studies of the effects of O3 on lung
5 monooxygenases are listed in Table 5-1. Early studies showed that exposure to O3 increased
6 CYP 2B1 (the major CYP isoform in rat lung) content and activity in rat lung. Ozone exposures
7 also caused hypertrophy and hyperplasia of CYP 2B 1-immunoreactive Clara cells. Comparisons
8 of rat and rhesus monkey CYP isoforms demonstrated species-specific and region-specific (e.g.,
9 trachea, parenchyma, differences in the activities of P450 isoforms (Lee et al., 1998)
10 Watt et al. (1998) found that 1 ppm O3 in both short and long-term exposures in rat
11 increased CYP 2E1 in a region-specific manner. Paige et al. (2000) showed that a long term
12 exposure ( 0.8 ppm 8h/day for 90 days) increased the activity of CYP 2B in distal lung, but not
13 trachea or intrapulmonary airways. Studies have focused on P450 gene expression to examine
14 possible genetic mechanisms that may explain differential O3-sensitivity (Mango et al., 1998).
15 Mice (129 strain) deficient in Clara cell secretory protein (CCSP-/-), which are oxidant-sensitive,
16 were exposed to 1 ppm O3 for 2 hours. The CCSP null mice demonstrated increases in IL-6 and
17 metallothionein (Mt) mRNA that preceded decreases in Clara cell CYP2F2 mRNA (normally
18 expressed at high levels in mouse lung) levels. In 129 strain wildtype (WT) mice, RNA levels
19 changed similarly, to a lesser degree. These data suggest a protective role against oxidant
20 damage for CSPP, and further, that genetic susceptibility to oxidant stress may be mediated, in
21 part, by the gene coding for CSPP.
22
23 5.2.1.3 Antioxidants, Antioxidant Metabolism, and Mitochondrial Oxygen Consumption
24 Ozone also undergoes reactions with AA, GSH, and uric acid, all antioxidants present in
25 ELF. This is a protective interaction, but even with environmentally relevant exposures to O3,
26 the reactivity of O3 is not quantitatively quenched. Antioxidants offer some protection from O3
27 exposure, but often do not maintain sufficiently high concentration to fully protect the lung.
28 Thus, O3- induced cell injury occurs in both the lower and upper respiratory tract. Early work
29 has shown that short-term exposures to < 1 ppm O3 increase antioxidant metabolism, including
30 levels of cytosolic enzymes G6PD, 6PGD, GR, and GSHPx. Re-exposure after a recovery
December 2005 5-3 DRAFT-DO NOT QUOTE OR CITE
-------
1 period causes increases equivalent to first-time exposures, thus previous exposure appears to not
2 be protective.
3 Increases in enzyme activity appear to increase as a function of age, suggesting that O3
4 exposure can cause greater lung injury in the older animal. Species differences exist in
5 antioxidant metabolism, with guinea pigs being very sensitive due to their diminished increases
6 in antioxidants and antioxidant enzymes. Long term exposures of rats to urban patterns of O3
7 (daily peaks of 0.25 ppm) caused increases in GSHPx and GR, but not superoxide dismutase
8 (SOD). The enzyme changes could be accounted for by changes in the steady-state cell
9 population or in cellular antioxidant capacity.
10 Ozone induced both site- and cell-specific changes in copper-zinc (Cu-Zn) and manganese
11 (Mn) SOD in rats exposed to 1.0 ppm O3 for up to 3 months (Weller et al., 1997). Cu-Zn SOD
12 labeling was decreased in epithelial cells in airways and parenchyma. Mn SOD labeling was
13 increased in both AM and epithelial type II cells of the centriacinar region (CAR), which the
14 authors suggest may allow these cells to tolerate further O3 exposure.
15 In a recent report Freed et al. (1999) evaluated the role of antioxidants in O3- induced
16 oxidant stress in dogs (exposed to 0.2 ppm in a 6-h exposure) by inhibiting the antioxidant
17 transport using probenencid (an ani on-transport inhibitor). Blocking antioxidant transport
18 caused heterogeneously distributed increases in peripheral airway resistance and reactivity,
19 supporting the hypothesis that in the lung periphery, endogenous antioxidants moderate the
20 effects of O3 and that this exposure is a subthreshold stimulus for producing effects on peripheral
21 airway resistance and reactivity in dogs. The authors further found that treatment with
22 probenecid also inhibited O3-induced neutrophilic inflammation, providing evidence for a
23 dissociation between airway function and inflammation, and suggesting that O3-induced
24 inflammation and airway hyperreactivity (AHR) are independent phenomena.
25 Mudway and Kelly (1998) modeled the interactions of O3 with ELF antioxidants using a
26 continually mixed, interfacial exposure set up with O3 concentrations of 0 to 1.5 ppm. Uric acid
27 was ranked the most O3-reactive, AA the second most reactive, and GSH the least reactive. Thus,
28 they concluded that GSH is not an important substrate for O3, while uric acid appeared to be the
29 most important substrate which confers protection from O3 by removing it from inhaled air and
30 limiting the amount that reaches the distal lung. The authors acknowledge limitations in
31 extrapolating these data to in vivo O3 exposures.
December 2005 5-4 DRAFT-DO NOT QUOTE OR CITE
-------
1 5.2.1.4 Lipid Metabolism and Content of the Lung
2 One of the major postulated molecular mechanisms of action of O3 is peroxidation of
3 unsaturated fatty acids in the lung. In both acute and short-term studies, a variety of lung lipid
4 changes occur, including an increase in arachidonic acid. Metabolism of arachidonic acid
5 produces a variety of biologically active mediators that can, in turn, affect host defenses, lung
6 function, the immune system, and other functions. The protein A component of surfactant is also
7 a primary target of O3 interaction. During the first few days of O3 exposure, the changes in lung
8 lipid biosynthesis can be accounted for by the alveolar epithelial proliferative repair. With
9 longer exposures (eg 0.12 ppm for 90 days ), an increases in PUFAs and a decrease in
10 cholesterol-esters are seen, indicative of long-term alterations of surfactant lipid composition.
11 Several new studies examined the effects of O3 exposure on phospholipids in lung tissue.
12 Ozonation of PUFAs is has been shown to generate other aldehydes such as nonanal and hexanal
13 in rat (Pryor et al., 1996; Frampton et al., 1999). These aldehydes are short-lived and found to
14 not affect lung function (Frampton et al., 1999). These observations suggest that levels of these
15 aldehydes are dependent on a dynamic relationship between their production and their
16 di sappearance from the ELF.
17 Pryor et al. (1995) proposed a cascade mechanism whereby ozonation products cause
18 activation of specific lipases, which then trigger the activation of second messenger pathways
19 (e.g., phospholipase A2 or phospholipase C). This group (Kafoury et al., 1999) showed that
20 exposure of cultured human bronchial epithelial cells to the lipid ozonation product 1 -palmitoyl-
21 2-(9-oxononanoyl)-sn-glycero-3-phosphocholine elicited release of platelet-activating factor
22 (PAF) and prostaglandin E2, but not IL-6. The lipid ozonation product 1-hydroxy-l-
23 hydroperoxynonane caused release of PAF and IL-6 in these cells, but not prostaglandin E2.
24 These results suggest to the authors that O3-induced production of lipid ozonation products
25 causes release of proinflammatory mediators that then generate an early inflammatory response.
26 Long et al. (2001) exposed hamsters to 0.12, 1.0 or 3.0 ppm O3 to evaluate lipid peroxidation and
27 antioxidant depletion. Six hour exposures to the two higher levels resulted in increased BALF
28 neutrophil numbers and F2-isoprostanes. Exposures to 1.0 ppm O3 with 1 h of exercise caused
29 increased levels of F2-isoprostanes.
30 Postlethwait et al. (1998) utilized three biologically relevant models, isolated epithelial
31 lining fluid, intact lung, and liposome suspensions to determine the O3-induced production of
December 2005 5-5 DRAFT-DO NOT QUOTE OR CITE
-------
1 heptanal, nonanal and hexanal. Data obtained from these studies suggested that PUFAs directly
2 react with O3 and the amount of bioactive lipids produced is inversely related to ascorbic acid
3 (AA) availability. The authors caution that there are limitations to the use of measurements of
4 these reactions products in determining O3 dose-response relationships due to the heterogenous
5 nature of O3 reactions in the ELF. Connor et al. (2004) have recently reported that interfacial
6 phospholipids may modulate the distribution of inhaled O3 and the extent of site-specific cell
7 injury. They utilized interfacial films composed of dipalmitoylglycero-3-phosphocholine
8 (DPPC) with rat lung lavage fluid and human fibroblast cell culture systems.
9 Hamilton et al. (1998) reported increased protein adducts in human AM exposed to
10 0.4 ppm O3 for 1 h with exercise. These adducts were found to be created due to the formation
11 of one of the most toxic ozonation products, 4-hydroxynonenal (FINE). Using human AM in
12 vitro cultures treated with FINE they also demonstrated a potential role of FINE in acute cellular
13 toxic effect of O3.
14 Uhlson et al. (2002) reacted O3 with calf lung surfactant which resulted in the production
15 of l-palmitoyl-2-(9'-oxo-nonanoyl)-glycerophosphocholine (16:Oa/9-al-GPCho). The biological
16 activity of this oxidized phospholipid included: (1) decreased macrophage viability,
17 (2) induction of apoptosis in pulmonary epithelial-like A549 cells, (3) and release of IL-8 from
18 A549 cells. Exposures levels of 0.125 ppm O3 in this in vivo system were capable of generating
19 biologically active phospholipids that were capable of mediating toxic effects of O3.
20 Thus, new work has attempted to elucidate the mechanisms by which reactions of O3 with
21 lipids create phospholipids that then mediate downstream toxic effects. However, it is uncertain
22 whether these described changes in lipid content and/or metabolism lead to significant changes
23 in surface tension or compliance properties of the lung, and thus are biologically relevant and
24 affect human health.
25
26 5.2.1.5 Protein Synthesis
27 Collagen, a structural protein involved in fibrosis, increases with O3 exposure. Some
28 studies have shown that this increase persists after exposure stops. The increased collagen has
29 been correlated with structural changes in the lung. Rats exposed to an urban pattern of O3 with
30 daily peaks of 0.25 ppm for 38 weeks displayed extracellular matrix thickening. Increased levels
31 of collagen in CAR were demonstrated in female rats exposed to 0.5 to 1.0 ppm O3 for 6 h/day
December 2005 5-6 DRAFT-DO NOT QUOTE OR CITE
-------
1 for 20 months and in monkeys exposed to 0.61 ppm for 1 year. Both increased age and health
2 status (e.g., emphysemic) were implicated in the increased collagen formation in response to O3
3 exposure. A recent time-course study (van Bree et al., 2001) evaluating the lung injury and
4 changes in collagen content in rats exposed acutely or subchronically to 0.4 ppm O3
5 demonstrated CAR thickening of septa which progressed from 7 through 56 days of exposure.
6 Though collagen content decreased with PE recovery, the structural fibrotic changes in ductular
7 septa and respiratory bronchioles persisted, suggesting that subchronic O3 exposures in rats
8 creates a progression of structural lung injury that can evolve to a more chronic form, which
9 included fibrosis. As with changes in lung lipids, the biological relevance and adverse health
10 effects of altered protein synthesis and collagen accumulation are uncertain.
11
12 5.2.1.6 Gene Expression
13 Gohil et al. (2003) examined differential gene expression in C57BL/6 mice exposed to
14 1 ppm O3 for three consecutive nights for 8 hours. Ozone exposure induced changes in
15 expression of 260 genes ( 80% repressed and 20% induced). These included genes involved in
16 progression of the cell cycle, several NF-KB-activated genes, and genes invoved in xenobiotic
17 and major histocompatibility complex, suggesting that O3 exposure suppresses immune function
18 and xenobiotic metabolism and enhances cellular proliferation.
19
20 5.2.1.7 Summary and Conclusions - Biochemical Effects
21 Ozone has been shown to interact with a wide range of different cellular components
22 including PUFAs, amino acid residues, and some low-molecular-weight compounds (GSH,
23 urate, vitamins C and E). As O3 does not penetrate much beyond the ELF, damage likely results
24 from its PUFA ozonation products (mostly hydroxyhydroperoxides) involvement in signaling of
25 cellular responses such as inflammation. New work has shown that ozonation of PUFA also
26 forms the aldehydes nonanal, heptanal, and hexanal, the production of which is dependent on AA
27 availability. Saturated phospholipids are thought to reduce the local dose and limit site-specific
28 cell injury from O3 exposure. Another ozonation product FINE creates protein adducts that have
29 been linked to apoptosis and heat shock proteins in vitro.
30 Both short- and long-term exposures to O3 have been shown to enhance lung xenobiotic
31 metabolism, possibly as a result of changes in the number and function of bronchiolar epithelial
December 2005 5-7 DRAFT-DO NOT QUOTE OR CITE
-------
1 Clara cells and alveolar epithelial Type 2 cells. This modulation is both species- and region-
2 specific and includes the isoforms CYP 2B1, CYP 2E1.CCSP is also involved in inflammatory
3 responses to O3 exposure. Mice strains with differing sensitivities to O3 show that responses in
4 protein, LDH and inflammatory cell influx are due to CCSP levels and changes in lung
5 epithelium permability.
6 Reactions of O3 with AA, GSH, and uric acid (all antioxidants present in ELF) are a
7 protective mechanism. But even with environmentally relevant exposures, the reactivity of O3 is
8 not quantitatively quenched and cell injury occurs in both the lower and upper respiratory tract.
9 Early work has shown that short-term exposures to < 1 ppm O3 increase antioxidant metabolism.
10 Re-exposure after a recovery period causes increases equivalent to first-time exposures,
11 suggesting that previous exposure is not protective. Increases in enzyme activity appear to
12 increase as a function of age, suggesting that O3 exposure can cause greater lung injury in the
13 older animal. Long-term urban patterns of exposure to O3 (daily peaks of 0.25 ppm) caused
14 increases in GSHPx and GR, but not SOD. Recent work has suggested that endogenous
15 antioxidants moderate the effects of O3 and that this exposure is a subthreshold stimulus for
16 producing effects on peripheral airway resistance and reactivity, thus indicating a dissociation
17 between airway function and inflammation.
18 In both acute and short-term studies, a variety of lung lipid changes occur with O3
19 exposure, including an increase in AA. With longer exposures (e.g., 0.12 ppm for 90 days),
20 an increase in PUFAs and a decrease in cholesterol-esters are seen, indicative of long-term
21 alterations of surfactant lipid composition. Whether these changes in lipid content and/or
22 metabolism lead to significant changes in surface tension or compliance properties of the lung
23 remains unknown. New studies evaluating O3-induced alterations in lipid metabolism have not
24 been completed.
25 Collagen, a structural protein involved in fibrosis, increases with O3 exposure, and some
26 studies have shown that this increase persists after exposure stops. Urban patterns of exposure
27 (daily peaks of 0.25 ppm for 38 weeks) created extracellular matrix thickening. Increases in
28 centriacinar collagen were demonstrated in female rats exposed to 0.5 to 1.0 ppm O3 for 6 h/day
29 for 20 months and in monkeys exposed to 0.61 ppm for 1 year. New work examining the time
30 course of lung injury and changes in collagen content in rats exposed acutely or subchronically
31 to 0.4 ppm O3 showed centriacinar thickening of septa. Collagen content decreased with PE
December 2005 5-8 DRAFT-DO NOT QUOTE OR CITE
-------
1 recovery but not the structural fibrotic changes in ductular septa and respiratory bronchioles,
2 which suggests that subchronic O3 exposures in rats creates a progression of structural lung
3 injury that can evolve to a more chronic form, which includes fibrosis.
4
5 5.2.2 Lung Host Defenses
6 Defense mechanisms, including the mucociliary clearance system, AMs, and humoral- and
7 cell-mediated immune system, exist in the lung to protect it from infectious and neoplastic
8 disease and inhaled particles. Summaries of key new animal studies examining the effects of O3
9 on lung host defenses are presented in Table AX5-2 of Annex AX5. Acute human exposures to
10 O3 result in similar effects on AMs (see Chapter 6).
11
12 5.2.2.1 Clearance
13 Early studies of the effect of O3 on the mucociliary escalator showed morphological
14 damage to ciliated epithelial cells of the tracheobronchial tree at doses of < 1 ppm. Functionally,
15 O3 slowed particle clearance in rats at doses of 0.8 ppm for 4 h and in rabbits at 0.6 ppm for 2 h
16 exposures. Acute exposures at 0.5 ppm O3 in sheep caused increased basal secretion of
17 glycoproteins, while longer exposures reduced tracheal glycoprotein secretions, both of which
18 can alter the effectiveness of the mucociliary escalator. Early postnatal exposures of sheep to
19 1 ppm O3 caused retardation of normal morphologic development of the tracheal epithelium,
20 decreased epithelial mucosia density, decreased tracheal mucous velocity, and delayed
21 development of carbohydrate composition. Conversely , alveolar clearance in rabbits after acute
22 exposure (0.1 ppm, 2 h/day, for 1 to 4 days) is increased. Longer exposures showed no effect
23 and increased O3 (1.2 ppm) slowed clearance. This pattern of clearance occurs in rats also.
24 A study using rat tracheal explants exposed to O3 (Churg et al., 1996) showed that uptake of
25 TiO2 and asbestos was enhanced at 0.01 and 0.1 ppm, respectively. The authors attribute the
26 increased uptake as a direct effect of O3, suggesting mediation by H2O2 or hydroxyl radical.
27 Studies of the clearance of the radiolabled chelate 99mTc diethylenetriamine pentaacetic acid
28 (Tc-DTPA) have shown that clearance is significantly increased following a 3h exposure to
29 0.8 ppm O3 in SD rats (Pearson and Bhalla,1997). Examination of regional clearance of
30 99mTc-DTPA in dogs following a 6 h isolated sublobar exposure to 0.4 ppm O3 or air showed that
31 O3 decreased the clearance halftime by 50% at 1 day following exposure (Foster and Freed,
December 2005 5-9 DRAFT-DO NOT QUOTE OR CITE
-------
1 1999). Clearance was still elevated at 7 d PE but had recovered by 14 d. So, a single local
2 exposure to O3 increases transepithelial clearance but without any influence on contralateral
3 segements, i.e., only for epithelia directly exposed to O3.
4 Alveolar clearance is slower than tracheobronchial clearance and involves particle
5 movement through interstitial pathways to the lymphatic system or movement of particle-laden
6 AMs to the bottom of the mucociliary escalator. Exposures of rabbits to 0.1 ppm accelerated
7 clearance while 1.2 ppm slowed clearance. A chronic exposure has been shown to slow
8 clearance. New evaluations of the effects of O3 on alveolar clearance have not been performed.
9
10 5.2.2.2 Alveolar Macrophages
11 A primary function of AMs is to clear the lung of infectious and non-infectious particles by
12 phagocytosis, detoxification, and removal. Further, AMs secrete cellular mediators that recruit
13 and activate inflammatory cells in the lungs. Ozone has been shown to inhibit phagocytosis at
14 0.1 ppm for 2 h in rabbits. This inhibition returns to control levels if exposures are repeated for
15 several days. The production of superoxide anion radicals and the activity AM lysosomal
16 enzymes (both involved in bactericidal activity) are inhibited by 3 h exposures to 0.4 and
17 0.25 ppm O3 in rodents and rabbits, respectively. Production of IFNy was decreased in rabbit
18 AM by 1 ppm O3 for 3 h.
19 New studies have shown that O3 affects AM chemotaxis, cell adhesion, and surface
20 expression of cell adhesion molecules (Bhalla, 1996). AM from SD rats exposed to 0.8 ppm O3
21 for 3 h showed greater mobility and greater adhesion than air exposed controls. This increased
22 mobility and adhesion were attenuated by CD16b and ICAM-1 antibodies, suggesting these
23 adhesion molecules modulate O3-induced inflammation. Antibodies to TNFa and ILla also
24 mitigated AM adherence, suggesting further that the inflammatory response to O3 is mediated by
25 these cytokines (Pearson and Bhalla, 1997). Cohen et al. (1996) showed that O3 reduces binding
26 of INF y to AM in WEHI-3 cells, and additionally reduces phagocytic activity, production of
27 reactive oxygen intermediates, and elevation of intracellular Ca++. Glutathione content in AM is
28 reduced by a 2 ppm 3 h exposure to O3, possibly due to its interaction with ozonation products
29 from O3-induced lipid peroxidation (Pendino et al., 1996).
30 Cohen et al. (2001, 2002) exposed male F-344 rats to either 0.1 or 0.3 ppm O3 for 4 h/day,
31 5 days/week or either 1 or 3 weeks. In this study, superoxide anion production was increased at
December 2005 5-10 DRAFT-DO NOT QUOTE OR CITE
-------
1 1 week. Hydrogen peroxide production was reduced at both exposure concentrations and
2 durations and was further reduced with INFy stimulation, suggesting that one effect of O3 is
3 compromised killing of bacteria by AM due to the reduction in hydrogen peroxide production.
4 Ozone treatment (2 ppm O3, 3 h in female SD rats) caused a time-dependent increase in NO
5 levels in both AM and type II epithelial cells that was correlated with increased expression of
6 iNOS mRNA and protein (Laskin et al., 1998). Inhibition of NF-KB, caused a dose-dependent
7 inhibition of NO and iNOS production. Additionally, O3 caused a time-dependent increase in
8 NF-KB binding activity in the nucleus of both cell types. The authors hypothesize that O3
9 exposure causes the cytokines TNFa and IL-1 Pa to bind to surface receptors and initiate
10 intracellular signaling pathways in AM leading to activation of NF-KB, its entry into the nucleus,
11 and its binding to the regulatory sequences of genes such as iNOS to allow their transcription.
12 Additional studies (Laskin et al., 2002) using AM isolated from C57B16xl29 mice with a
13 targeted disruption of the gene for iNOS showed no toxicity to 0.8 ppm O3 for 3h, as measured
14 by BALF protein levels and nitrotyrosine staining of the lung. Additionally, mice
15 overexpressing human Cu, Zn superoxide dismutase (SOD) and mice with a targeted disruption
16 of p50 NF-KB were also resistant to O3 toxicity. WT mice exposed to O3 showed an increase in
17 expression of STAT-1, a protein that binds to the regulatory region of iNOS. Taken together,
18 these results suggest to the authors that a number of proteins including NF-KB, phosphoinoside
19 3-kinase, and STAT-1 that bind to and regulate expression of iNOS are modulated by O3
20 exposure. The same iNOS knockout mice strain exposed to O3 (Fakhrzadeh et al., 2002) showed
21 no increase in AM superoxide anion and prostaglandin. These data provide further evidence the
22 NO and its reactive oxidative product peroxynitrite are important in O3-induced lung injury.
23 Further discussions of the role of nitric oxide synthase/reactive nitrogen and
24 cytokines/chemokines in O3-induced inflammation are provided in Section 5.2.3.
25
26 5.2.2.3 Immune System
27 The effects of O3 on the immune system are complex and depend on the exposure
28 parameters and observation periods. T-cell-dependent functions appear to be more affected than
29 B-cell-dependent functions. Generally, there is an early immunosuppressive effect that can, with
30 continued exposure, either return to normal or actually enhance immunity. Changes in immune
31 cell population occur with O3 exposure including T:B-cell ratios in the MLN. Natural killer
December 2005 5-11 DRAFT-DO NOT QUOTE OR CITE
-------
1 (NK) cell activity increases with 1 week exposures of 0.2 to 0.4 ppm O3 but decreases with
2 exposures to 0.82 ppm. Ozone exposure has also shown to be responsible for enhancement of
3 allergic sensitization at levels of 0.5 to 0.8 ppm for 3 days. Studies of the effects of O3 on the
4 immune system are summarized in Table AX5-2.
5 Recent work examining immune system responses to O3, Garssen et al. (1997) have shown
6 that BALB/c mice sensitized with PCI are hyperreactive to carbachol after a PSA challenge, but
7 not if exposed to 0.4 to 1.6 mg/m3 O3. The sensitized mice also demonstrated a suppressed
8 inflammatory reaction (PMN) with 1.6 mg O3 exposure. These results are opposite to the effect
9 on type I (IgE-mediated) allergic reactions, which the authors suggest is due to activation of Th-
10 2 cell-dependent reactions that are possibly potentiated by O3 or to a direct effect by O3 on Th-1
11 cells or other cells that are crucial for the tracheal hyperreactivity and inflammation seen in this
12 mouse model.
13 Kleeberger et al. (2000, 2001b) have recently demonstrated a potential interaction between
14 the innate and acquired immune system with O3 exposure. Using O3-susceptible (C57BL/6J) and
15 O3-resistant (C3H/HeJ) mice, they identified a candidate gene on chromosome 4, Toll-like
16 receptor 4 (Tlr4). Ozone exposure (0.3ppm for 24 to 72 hours) of C3H/HeJ and C3H/HeOuJ
17 mice, the latter differing from the O3-resistant strain by a polymorphism in the coding region of
18 77r¥, then demonstrated greater protein concentrations in the OuJ strain. The two strains
19 exhibited differential expression of Tlr4 mRNA with O3 exposure. Thus, a quantitative trait
20 locus on chromosome 4 appears to be responsible for a significant portion of the genetic variance
21 in O3 -induced lung hyperpermeability. In these mouse strains lavageable protein concentration
22 was lowered by inhibition of inducible nitric oxide synthase (iNOS) and by targeted disruption
23 of Nos2. Comparisons of C3H/HeJ and C3H/HeOuJ O3 exposures demonstrated reduced Nos2
24 and Tlr4 mRNA levels in the O3-resistant C3H/HeJ mice. These data are consistent with the
25 hypothesis that O3 -induced lung hyperpermeability is mediated by iNOS. These studies suggest
26 a role for TLR4 in the host response to O3 similar to the role it has demonstrated in
27 lipopolysaccharide (LPS) sensitivity (Schwartz 2002; Wells et al. 2003). TLR4 signaling is
28 thought to be critical to linking the innate and acquired immune system through antigen
29 presenting cells and Thl/Th2 differentiation.
30 Ozone exposure has been shown to affect IgE responses in both in vitro and in mice.
31 Becker et al. (1991) demonstrated changes in IgG production in cultured human lymphocytes
December 2005 5-12 DRAFT-DO NOT QUOTE OR CITE
-------
1 with O3 exposures of 1.0, 0.5, and 0.1 ppm for 2 h. Subsequent to O3 exposure, cells were
2 stimulated with pokeweed mitogen (PWM, a T-cell-dependent stimulus) or Staphylococcus
3 aureus Cowan 1 strain (SAC, a T-cell-independent stimulus). Both B and T cells were affected
4 by O3. T cells also demonstrated an increase in IL-6 and a decrease in IL-2, which suggested to
5 the authors that O3 may have direct effects on IgG producing cells and concurrently an effect that
6 is mediated by altered production of T cell immunoregulatory molecules. Responses to repeated
7 O3 (0.08 - 0.25 ppm) and OVA (1%) exposures were compared in "IgE-high responder"
8 (BALB/c) and "IgE-low responder" (C57BL/6) mice (Neuhaus-Steinmetz et al., 2000). Ozone
9 appeared to shift the immune response toward a Th2-like pattern in the two mouse strains with
10 differing potentials for developing allergic reactions.
11 Surfactant protein A and D (SP-A and SP-D) were shown to create an inflammatory
12 feedback loop with perturbations in lung immune defenses (reviewed in Hawgood and Poulain,
13 2001). Earlier studies suggested that SP-A is a target for O3 toxicity by causing inhibition of SP-
14 A self-association and SP-A-mediated lipid vesicle aggregation. Further, O3 reduced the ability
15 of SP-A to inhibit phospholipid secretion by alveolar type II cells O3 and reduced the capacity of
16 SP-A to induce superoxide anion production and enhance phagocytosis of herpes simplex virus.
17 Bridges et al. (2000) reported that both SP-A and SP-D directly protect surfactant phospholipids
18 and macrophages from oxidative damage by blocking accumulation of TEARS and conjugated
19 dienes.
20 Eight human variants of PS-A in CHO cells exposed to O3 (Ippm for 4 hr) showed
21 decreased ability to stimulate cytokine (TNF- and IL-8) production in THP-1 cells, a
22 macrophage-like cell line (Wang et al., 2002). Each variant had a unique time- and dose-
23 dependent pattern of stimulation of cytokine production with O3 exposure which the authors
24 attribute to possible differences in susceptibility to O3 oxidation. Targeted disruption of mouse
25 SP- A and SP-D (Hawgood et al, 2002) caused increases in BAL phospholipid, macrophage, and
26 protein through 24 weeks of age. Further, the deficient mice developed patchy lung
27 inflammation and air space enlargement consistent with emphysema. Future experiments using
28 these null mice will help to establish the role of SP-A and SP-D in pulmonary host defense to O3
29 exposure.
30
December 2005 5-13 DRAFT-DO NOT QUOTE OR CITE
-------
1 5.2.2.4 Interactions with Infectious Microorganisms
2 Ozone-induced dysfunction of host defense systems results in enhanced susceptibility to
3 bacterial lung infections. Acute exposures of 0.08 ppm (3h) O3 can overcome the ability of mice
4 to resist infection (by decreasing lung bactericidal activity) with Streptococcal bacteria, resulting
5 in mortality. Changes in antibacterial defenses are dependent on exposure regimens, species and
6 strain of test animal, species of bacteria, and age of animal, with young mice more susceptible to
7 the effects of O3. The effect of O3 exposure on antibacterial host defenses appears to be
8 concentration- and time-dependent. Early studies using the mouse "infectivity model," consisting
9 of exposure to clean air or O3 followed by exposure to an aerosolized microorganism, showed
10 that the difference in mortality between O3-exposed groups and controls is concentration-related.
11 Chronic exposures (weeks, months) of 0.1 ppm do not cause greater effects on infectivity than
12 short exposures, due to defense parameters becoming reestablished with prolonged exposures.
13 More recent studies of O3-induced modulation of cell-mediated immune responses showed
14 effects on the onset and persistence of infection. Cohen et al. (2001,2002) exposed male F-344
15 rats subchronically to either 0.1 or 0.3 ppm O3. Subsequent exposure with viable Listeria
16 monocytogenes demonstrated no observed effect on cumulative mortality, but did show a
17 concentration-related effect on morbidity onset and persistence. These data suggest that O3 may
18 cause a possible imbalance between Th-1 and Th-2 cells, which can subsequently lead to
19 suppression of the resistance to intracellular pathogens.
20 Effects of O3 on viral infections are dependent on the temporal relationship between O3
21 exposure and viral infection. Only high concentrations (1.0 ppm O3, 3 h/day, 5 days, mice)
22 increased viral-induced mortality. No detrimental effects were seen with a 120-day exposure to
23 0.5 ppm O3 on acute lung injury from influenza virus administered immediately before O3
24 exposure started. But there were O3-enhanced postifluenzal alveolitis and lung parenchymal
25 changes. As O3 does not affect lung influenza viral liters, it apparently does not impact antiviral
26 clearance mechanisms. In general, the evidence suggests that O3 can enhance both bacterial and
27 viral lung infections, but the key mechanisms have not yet been identified. New studies on the
28 interactions of O3 and viral infections have not been published.
29
December 2005 5-14 DRAFT-DO NOT QUOTE OR CITE
-------
1 5.2.2.5 Summary and Conclusions - Lung Host Defenses
2 New data on lung host defenses support earlier work which suggests that mucociliary
3 clearance is affected in most test species at just under Ippm, with lower levels (~0.1 ppm)
4 increasing clearance and somewhat higher levels decreasing clearance. These data also propose
5 mechanisms whereby O3 affects clearance, which include uptake being a direct effect of O3, but
6 modulated by ROS and hydroxyl radicals.
7 Alveolar macrophage function is disrupted by O3 as shown by a number of studies
8 demonstrating inhibition of phagocytosis at concentrations ranging from 0.1 to 1.2 ppm. This
9 inhibition returns to control levels if exposures are repeated for several days. Two new studies
10 corroborate earlier findings of increases in AM number in that same exposure range. In this
11 environmentally relevant exposure range, new studies support older findings of decreased
12 resistance to microbial pathogens as shown by the endpoints examining superoxide radical
13 formation, altered chemotaxis/motility, decreased INFy, levels, decreased lysosomal activity,
14 increased PGE levels, and increased NO mRNA and protein.
15 New research evaluating the effects of O3 on immune function advances previous work that
16 has shown that exposures can enhance or suppress immune responsiveness depending on the
17 species studied, concentration of O3, route of exposure of allergen, and timing of exposure.
18 Continuous exposure to O3 impairs immune responses for the first several days of exposure,
19 followed by an adaptation to O3 that allows a return of normal immune responses. Most species
20 show little effect of O3 exposures prior to immunization, but a suppression of responses to
21 antigen in O3 exposures post-immunization. The use of mouse strains with genetically
22 determined sensitivity or resistance to O3 indicated a possible interaction between the innate and
23 acquired immune system, and further, that O3 may shift the immune response towards a Th-2-
24 like pattern. Work has also focused the deleterious effects of O3 exposure on SP-A and SP-D
25 and their immunomodulatory function in protecting against oxidative stress.
26 Several new studies evaluating the effects of O3 exposures on infectious microorganisms
27 are in concurrence with previous studies which showed, in general, increased mortality and
28 morbidity, decreased clearance, increased bacterial growth, and increased severity of infection at
29 exposure levels of 0.1 to 1 ppm O3 for 1 week.
30
December 2005 5-15 DRAFT-DO NOT QUOTE OR CITE
-------
1 5.2.3 Inflammation and Lung Permeability Changes
2 Lung inflammation and increased permeability, which are distinct events controlled by
3 independent mechanisms, are two well-characterized effects of O3 exposure. Disruption of the
4 lung barrier leads to leakage of serum proteins, influx of polymorphonuclear leukocytes (PMNs),
5 release of bioactive mediators, and movement of compounds from the airspaces into the blood.
6 Increases in permeability and inflammation have been observed at levels as low as 0.1 ppm O3
7 for 2 h/day for 6 days in rabbit and 0.12 ppm in mice (24-h exposure) and rats (6-h exposure).
8 After acute exposures, the influence of the time of exposure increases as the concentration of O3
9 increases. The exact role of inflammation in causation of lung disease is not known, nor is the
10 relationship between inflammation and changes in lung function. Table AX5-3 in Annex AX5
11 summarizes new key studies describing the potential for O3 exposure affect lung permeability
12 and inflammation. Controlled human exposure studies discussed in Chapter 6 indicate that the
13 majority of acute responses in humans are similar to those observed in animals.
14
15 5.2.3.1 Time Course of Inflammation and Lung Permeability Changes
16 The maximal increase in BALF protein, albumin, and PMN occurs in 8 to 18 h after the
17 cessation of acute exposure. A recent study of OVA-sensitized male Dunkin-Hartley guinea pigs
18 exposed to 1.0 ppm O3 for 3 h showed that PMN levels were significantly increased at 3 h PE,
19 but BAL protein levels were not, suggesting a lack of correlation between the two endpoints
20 (Sun et al., 1997). Increased PMN without a concordant increase in BAL protein levels were
21 found when the guinea pigs were exposed to 1.0 ppm O3 for 1 h and evaluated 24-h PE. The first
22 group also had an increase in AHR, but not the second group, which suggests a dissociation
23 between PMN levels and AHR.
24 Earlier work demonstrated that O3 exposures of 0.8 to 1 ppm transiently increase the
25 permeability from the air to the blood compartment. This permeability is greatest in trachea and
26 bronchoalveolar zone, and may allow increased entry of antigens and other bioactive compounds
27 (e.g., bronchoconstrictors) into lung tissues. The time course of the influx of PMNs into the lung
28 and the BALF fluid levels of macrophage inflammatory protein-2 (MIP-2) were found to be
29 roughly similar to that for proteins (Bhalla and Gupta, 2000). Adherence of neutrophils to
30 pulmonary vascular endothelium is maximal within 2 h after exposure and returns to control
31 levels by 12 h PE (Lavnikova et al., 1998). In an in vitro system utilizing rat alveolar type II cell
December 2005 5-16 DRAFT-DO NOT QUOTE OR CITE
-------
1 monolayers, O3 produced a dose-dependent increase in permeability (Cheek et al., 1995).
2 At higher O3 levels, neutrophils exacerbated the injury, but their presence after the exposure
3 expedited restoration of epithelial barrier. Vesely et al. (1999) have demonstrated that
4 neutrophils contribute to the repair process in O3 -injured airway epithelium.
5 Subchronic exposures (3 to 7 days) have been found to cause increases in BALF protein
6 and PMNs that typically peak after a few days (depending upon species tested and exposures)
7 and return towards control even with continuing exposure. Van Bree et al. (2002) observed
8 lower BALF levels of protein, fibronectin, IL-6 and inflammatory cells in rats exposed for 5 days
9 than in rats exposed for 1 day, suggesting adaptation to O3 exposure. Postexposure challenge
10 with single O3 exposures at different time points showed recovery of susceptibility to O3.
11 McKinney et al. (1998) observed differences in IL-6 levels due to repetitive exposures and
12 demonstarted a role of IL-6 in the adaptive response induced by repeated O3 exposures.
13
14 5.2.3.2 Concentration and Time of Exposure
15 Analysis of the relative influence of concentration and duration of exposure (i.e., CXT) of
16 O3 has shown that concentration generally dominates the response. The impact of T was C-
17 dependent (at higher Cs, the impact of T was greater); at the lowest C and T values, this
18 dependence appeared to be lost. New studies evaluating CXT relationships in animal models
19 have not been found.
20
21 5.2.3.3 Susceptibility Factors
22 Factors that have been studied for potential impact on the effects of O3 exposure include
23 age, gender, nutritional status, exposure to co-pollutants, exercise, and genetic variability. A full
24 characterization of the effects of age on O3 responses has not been completed. Data available
25 indicate that effects of age on O3 responses are endpoint-dependent, with young mice, rats and
26 rabbits having greater prostaglandin levels with exposure and senescent rats having greater IL-6
27 and N-acetyly-p-D-glucosaminidase levels with exposure.
28 A new study (Johnston et al., 2000b) compared gene expression of chemokines and
29 cytokine in newborn and 8-week-old C57B1/6J mice exposed to 1.0 or 2.5 ppm for 4, 20, or 24 h.
30 The newborn mice displayed increased levels of Mt mRNA only, while the 8-week-old mice had
31 increases in MIP-1 a, MIP-2, IL-6, and Mt mRNA. Comparisons were made with mice of the
December 2005 5-17 DRAFT-DO NOT QUOTE OR CITE
-------
1 same age groups with exposures to endotoxin (10 min). Both age groups displayed similar
2 cytokine/chemokine profiles with endotoxin exposure. This suggested to the authors that the
3 responses to endotoxin, which does not cause epithelial injury, and the responses to O3, which
4 does, demonstrate that differences in inflammatory control between newborn and adult mice is
5 secondary to epithelial injury.
6 Pregnancy and lactation increased the susceptibility of rats to acute O3, but no clear effects
7 of gender have been identified. The effects of vitamin C deficiency on O3 responses are unclear.
8 Ascorbate-deficient guinea pigs exposed to O3 demonstrated only minimal effects on injury and
9 inflammation (Kodavanti et al., 1995). Utilizing a diet-restricted (20% of the freely-fed diet) rat
10 model, Elsayed (2001) demonstrated higher survivability on exposure to higher O3 (0.8 ppm
11 continuously for 3 d) compared to freely-fed rats. Pre-exposure to sidestream cigarette smoke
12 had been found to cause increased lung injury (Yu et al., 2002). In vitro studies on the
13 macrophages from smoke + O3- exposed animals responded by a greater release of TNF-a
14 following LPS stimulation when compared to macrophages exposed to air, smoke or O3 alone.
15 Recent lines of evidence illustrate that genetic background is an extremely important
16 determinant of susceptibility to O3. In earlier studies using inflammation-prone (susceptible)
17 C57BL/6J (B6) and inflammation-resistant C3H/HeJ (C3) mouse strains and high doses of O3
18 (2ppm for 3 hours) identified Inf-2 as a locus controlling susceptibility. Further studies in these
19 two strains of mice identified that the acute and subacute exposures are controlled by two
20 distinct genes, referred to as Inf-l and Inf-2, respectively (Tankersley and Kleeberger, 1994).
21 Kleeberger et al. (1997) also identified another potential susceptibility gene, tumor necrosis
22 factor (Tnf, which codes for the pro-inflammatory cytokine TNF-a ) on a qualitative trait locus
23 on mouse chromosome 17. By neutralizing the function of TNF-a with a specific antibody, they
24 were able to confer protection against O3 injury in susceptible mice. The group then
25 demonstrated a role for TNF receptor 1 and 2 (TNFR1 and TNFR2, respectively) signaling in
26 subacute ( 0.3 ppm for 48 hrs) O3-induced pulmonary epithelial injury and inflammation (Cho
27 et al., (2001). TNFR1 and TNFR2 knockouts were less sensitive to subacute O3 exposure than
28 WT C57BL/6J mice. Further studies using these knockouts by Shore et al. (2001) indicated a
29 role of TNF-a in AUR but not in O3-induced infiltration of PMN, and provided evidence for the
30 mechanistic separation of hyperresponsiveness and PMN infiltration.
December 2005 5-18 DRAFT-DO NOT QUOTE OR CITE
-------
1 In studies to evaluate differences in susceptibility to death from O3 exposure, Prows et al.
2 (1997; 1999) exposed A/J and C57BL/6J mice exposed to 10 ppm O3. The A/J strain is more
3 sensitive to O3-induced death, while the C57BL/6J strains is resistant to O3-induced death. They
4 identified two loci (acute lung injury-1 and -3, Ali-1 and Ali-3, respectively) on chromosome
5 11 that appear to control susceptibility to death after O3 exposure.
6 An integrated and more comprehensive effort to identify the genetic basis for the
7 susceptibility to O3-induced lung injury was reported by Savov et al. (2004). In this report,
8 summarized in Table AX5-4 of Annex AX5, acute lung injury to high dose of O3 (2 ppm for 3h)
9 was assessed and integrated with physiological, biochemical, and genetic observations using 9
10 inbred mouse strains. This work indicated the presence of genetic loci on chromosomes 1,7, 15
11 associated with phenotypic characteristics for resistance to acute O3-induced lung injury. They
12 identified C3H/HeJ and A/J as consistently O3-resistant,C57BL/6J and 129/SvIm as consistently
13 O3-vulnerable, and CAST/Ei, BTBR, DBA/2J, FVB/NJ, and BALB/cJ as intermediate in
14 response to O3.
15 A similar comparative study by Wattiez et al. (2003) using five inbred mouse strains to
16 characterize the molecular basis for O3-induced lung injury identified differential expression of
17 CCSP isoform CC16a in BALF of C57BL/6J (O3-sensitive) and C3H/HeJ (O3-resistant) strains.
18 Ozone-induced changes in CCSP expression were evaluated in five inbred mouse strains:
19 C57BL/6J and CBA both considered sensitive to acute O3-induced inflammation, C3H/HeJ and
20 AKR/J both considered resistant, and SJL/J considered intermediate (Broeckaert et al., 2003).
21 Two exposure paradigms (1.8 ppm O3 for 3 h or 0.11 ppm O3, 24/h day for up to 3 days) were
22 used, and BALF and serum were assayed immediately after exposure or at 6 h. Both exposure
23 levels caused a transient increase in CC16 in serum that correlated with BALF changes in
24 protein, LDH, and inflammatory cells. There was an inverse relationship between preexposure
25 levels of CC16 in BALF and epithelial damage based on serum CC16 levels and BALF markers
26 of inflammation. There was also an inverse relationship between preexposure levels of albumin
27 in BALF and lung epithelium damage. These results suggest to the authors that a major
28 determinant of susceptibility to O3 is basal lung epithelium permeability. I.e., the leakiness of
29 the epithelium allows CC16 to enter the blood and protein and inflammatory cells to enter the
30 lung. As all of the mouse strains had similar levels of preexposure CC16 mRNA, they conclude
31 that strain differences in the basal permeability of the airway epithelium is responsible for lung
December 2005 5-19 DRAFT-DO NOT QUOTE OR CITE
-------
1 differences in basal CC16 among strains. This group also used 2-dimensional protein
2 electrophoresis to examine differences in BALF protein between C57BL/6J and C3H/HeJ mice
3 before and after O3 exposure. They found the CC16 monomer, a 7kD protein, in two isoforms
4 with differing p/values, CC16a (4.9) and CC16b (5.2). C57BL/6J mice had lower levels of
5 CC16a (the more acidic form) than the C3H/HeJ mice, but both strain had similar levels of
6 CC16b. They hypothesize that the C57BL/6J strain has greater epithelial permeability, and thus
7 allows more leakage of CC16a. These data taken together suggest a protective role against
8 oxidant damage for CSPP, and further that genetic susceptibility to oxidant stress may be
9 moderated, in part, by the gene coding for CSPP.
10
11 5.2.3.4 Mediators of Inflammatory Response and Injury
12 Ozone reacts with lipids in the ELF or epithelial cell membranes, creating ozonation
13 products that then stimulate airway epithelial cells, AMs, and PMNs to release a host of pro-
14 inflammatory mediators which include cytokines, chemokines, reactive oxygen species,
15 eicosanoids, and platelet activating factors. At O3 exposures of > 1 ppm, these mediators recruit
16 PMN and increase expression of MIP-2 mRNA or BALF levels of MIP-2 (Driscoll et al., 1993;
17 Haddad et al., 1995; Bhalla and Gupta, 2000). The increased mRNA expression was associated
18 with an increased neutrophilia in the lung. Zhao et al. (1998) showed that O3 exposure in mice
19 and rats causes an increase in monocyte chemotactic protein-1 (MCP-1).
20 Fibronectin, an extracellular matrix glycoprotein, is thought to have a role in lung
21 inflammation and inflammatory disorders, and has shown to be increased with exposures of
22 1 ppm for 14 days. Gupta et al. (1998) observed an increase in both fibronectin protein and
23 mRNA expression in the lung of rats exposed to 0.8 ppm O3. A mechanistic role of fibronectin
24 in O3-induced inflammation and injury was suggested on the basis of comparability of temporal
25 changes in BALF protein, fibronectin and alkaline phosphatase activity (Bhalla et al., 1999).
26 Studies have reported an effect of O3 on other cytokines and inflammatory mediators.
27 An increase occurred for cytokine-induced neutrophil chemoattractant (CINC) and NF-icB
28 expression in vivo (Haddad et al., 1996; Koto et al., 1997), for IL-8 in vivo and in vitro (Chang
29 et al., 1998), TNFa, fibronectin, IL-1 and CINC release by macrophages ex vivo (Pendino et al.,
30 1994; Ishii et al., 1997), and NF-KB and TNFa (Nichols et al., 2001; see 6.9.2). An increase in
31 lung CINC mRNA occurred within 2 hrs after the end of a 3 hr exposure of rats to 1 ppm O3.
December 2005 5-20 DRAFT-DO NOT QUOTE OR CITE
-------
1 The CINC mRNA expression was associated with neutrophilia at 24 hrs PE. Exposure of guinea
2 pig AMs recovered in BALF and exposed in vitro to 0.4 ppm O3 for 1- h produced a significant
3 increase in IL-6 and TNFa (Arsalane et al., 1995). An exposure of human AMs to an identical
4 O3 concentration increased TNFa, IL-lb, IL-6 and IL-8 and their mRNAs. Ozone exposure of
5 mice caused an increase in IL-6, MlP-la, MIP-2, eotaxin and Mt abundance (Johnston et al.,
6 1999a). The IL-6 and MT increase was enhanced in mice deficient in CCSP, suggesting a
7 protective role of Clara cells and their secretions (Mango et al., 1998). CCSP deficiency also
8 increased sensitivity of mice to O3, as determined by an increase in abundance of MlP-la and
9 MIP-2 following a 4 hr exposure (Johnston et al., 1999b).
10 Mast cells, which are located below the epithelium, release proinflammatory mediators and
11 have been shown to contribute to O3-induced epithelial damage. Mast cell-deficient mice
12 exposed to 2 ppm O3 showed no inflammation or epithelial injury which was observed in WT
13 mice (Longphre et al., 1996). Greater increases in lavageable macrophages, epithelial cells and
14 PMNs were observed in mast cell-sufficient mice than in mast cell-deficient mice exposed to
15 0.26 ppm (Kleeberger et., 2001a). Increases in inflammatory cells were also observed in mast
16 cell-deficient mice repleted of mast cells, but O3-induced permeability increase was not different
17 in genotypic groups exposed to 0.26 ppm. When a mast cell line was exposed to varying O3
18 concentrations, spontaneous release of serotonin and modest generation of PGD2 occurred only
19 under conditions that caused cytotoxicity (Peden and Dailey, 1995). Additionally, O3 inhibited
20 IgE- and A23187- induced degranulation. Mast cells recovered from O3-exposed peripheral
21 airways of ascaris sensitive dogs released significantly less histamine and PGD2 following in
22 vitro challenge with ascaris antigen or calcium ionophore (Spannhake, 1996). Ozone exposure
23 also promoted eosinophil recruitment in the nose and airways in response to instillation of OVA
24 or OVA-pulsed dendritic cells and aggravated allergy like symptoms in guinea pigs (lijima et al.,
25 2001).
26 The role of PMNs and cellular mediators in lung injury and epithelial permeability has
27 been investigated using antibodies and inhibitors of known specificity to block inflammatory cell
28 functions and cytokine activity. Treatment of rats with cyclophosphamide prior to O3 exposure
29 resulted in a decreased recovery of PMNs in the BALF and attenuated permeability induced by
30 O3 (Bassett et al., 2001). Pretreatment of animals with antiserum against rat neutrophils
31 abrogated PMN accumulation in the lung, but did not alter permeability increase produced by
December 2005 5-21 DRAFT-DO NOT QUOTE OR CITE
-------
1 O3. DeLorme et al. (2002) showed a relationship between neutrophilic inflammation and AHR.
2 Treatment of rats with anti-neutrophil serum protected the animals from O3-induced AHR.
3 Studies utilizing antibodies to selected pro- or anti-inflammatory cytokines suggest a role of
4 TNFa, IL-10, and IL-lb in O3-induced changes in permeability, inflammation and cytokine
5 release (Ishii et al., 1997; Reinhart et al., 1999; Bhalla et al., 2002). An attenuation of O3-
6 induced increase in permeability and inflammation was also observed in mice treated, either
7 before or after exposure, with UK-74505, a platelet-activating factor (PAF) receptor antagonist
8 (Longphre et al., 1999). These results were interpreted to indicate that O3-induced epithelial and
9 inflammatory changes are mediated in part by activation of PAF receptors.
10 Ozone exposure stimulates macrophage motility towards a chemotactic gradient, and
11 macrophages from rats exposed to 0.8 ppm O3 adhered to epithelial cells (ARL-14) in culture to
12 a greater extent than macrophages from air-exposed controls (Bhalla, 1996). Both macrophage
13 motility and chemotaxis were attenuated by antibodies to cell adhesion molecules CD-I Ib and
14 ICAM-1, suggesting a role for cell adhesion molecules in O3-induced cellular interactions. This
15 may also explain the increased tissue localization and reduced recovery of macrophages in
16 BALF (Pearson and Bhalla, 1997) following O3 exposure. Studies investigating the mechanisms
17 of PMN recruitment in the lung have explored the role of cell adhesion molecules that mediate
18 PMN-endothelial interactions. An exposure of female rats to O3 had an attenuating effect on
19 CD-18 expression on AMs and vascular PMNs, but the expression of CD62L, a member of
20 selectin family, on vascular PMNs was not affected (Hoffer et al., 1999). In monkeys, O3-
21 induced inflammation was blocked by treatment with a monoclonal antibody to CD 18,
22 suggesting dependence of PMN recruitment on this adhesion molecule (Hyde et al., 1999).
23 Treatment of monkeys with CD 18 antibody also reduced tracheal expression of the P6 integrin
24 (Miller et al., 2001). A single 3 hr exposure of rats to O3 caused an elevation in concentration of
25 ICAM-1, but not CD-18, in the BALF (Bhalla and Gupta, 2000). Takahashi et al. (1995a) found
26 an increase in tissue expression of ICAM-1 in mice exposed to 2 ppm O3, noting a temporal
27 correlation of inflammatory activity and ICAM-1 expression which varied in different regions of
28 the lung. A comparable pattern of time-related changes in total protein, fibronectin and alkaline
29 phosphatase activity in the BALF of rats exposed to 0.8 ppm O3 was also noted by Bhalla et al.
30 (1999). Together, these studies support the role of extracellular matrix protein and cell adhesion
31 molecules in the induction of lung inflammation and injury.
December 2005 5-22 DRAFT-DO NOT QUOTE OR CITE
-------
1 5.2.3.5 The Role of Nitric Oxide Synthase and Reactive Nitrogen in Inflammation
2 Nitric oxide (NO) is a messenger molecule involved in many biological processes,
3 including inflammation. Cells in the respiratory tract (including mast cells, neutrophils,
4 epithelial cells, neurons, and macrophages) produce three differing forms of nitric oxide synthase
5 (NOS), the enzyme that catalyzes the formation of NO. NOS-l(neuronal) and NOS-3
6 (endothelial) are constitutively expressed, whereas NOS-2 (also referred to as iNOS) is
7 inducible, commonly by pro-inflammatory cytokines. An acute exposure of rats to 2 ppm O3
8 caused an increase in BALF macrophage number and total protein, in iNOS expression, in
9 fibronectin, and in TNFa production by AMs (Pendino et al., 1995). All of these effects of O3
10 were reduced by pretreatment with gadolinium chloride, a macrophage inhibitor. Macrophages
11 isolated from O3-exposed mice produced increased amounts of NO, superoxide anion, and PGE2,
12 but production of these mediators by macrophages from NOS knockout mice was not elevated
13 (Fakhrzadeh et al., 2002). Additionally, mice deficient in NOS or mice treated with
14 NG-monomethyl-L-arginine, an inhibitor of total NOS, were protected from O3-induced
15 permeability, inflammation, and injury, suggesting a role of NO in the production of O3 effects
16 (Kleeberger et al., 2001b; Fakhrzadeh et al., 2002). These results contrast with a study showing
17 that O3 exposure produced greater injury, as determined by measurement of MIP-2, matrix
18 metalloproteinases, total protein, cell content and tyrosine nitration of whole lung protein, in
19 iNOS knockout mice than in wild type mice (Kenyon et al., 2002). This group suggests that
20 protein nitration is related to inflammation and is not dependent on iNOS-derived NO. They
21 point out the possible experimental differences, such as O3 concentration, for inconsistency
22 between their results and those of Kleeberger et al. (2001b).
23 Rats pretreated with ebselen, a potent anti-inflammatory, immunomodulator, and
24 NO/peroxynitrite scavenger, and then exposed to 2 ppm O3 for 4 -h had decreased numbers of
25 neutrophils, lowered albumin levels, and inhibited nitration of tyrosine residues in BALF 18 h
26 PE, though macrophage iNOS expression was not changed (Ishii et al., 2000a). These results
27 suggest an iNOS-independent mechanism for O3-induced inflammation. Inoue et al. (2000)
28 demonstrated in human transformed bronchial epithelial cells that NO-generating compounds
29 (TNFa, IL-1P, and INF-y) induce IL-8 production and that NOS inhibitors inhibit IL-8
30 production. In vivo experiments in the same study using male Hartley-strain guinea pigs
31 exposed to 3 ppm O3 for 2-h showed that NOS inhibitor pretreatment attenuated-O3 induced
December 2005 5-23 DRAFT-DO NOT QUOTE OR CITE
-------
1 neutrophil recruitment and AHR at 5 hours after exposure. The NOS inhibitors also blunted the
2 increase in nitrate/nitrite levels and in IL-8 mRNA, at the 5 hours PE. The authors hypothesize
3 that NO, or its derivatives, facilitate AHR and inflammation after O3 exposure, possibly
4 mediated by IL-8. Jang et al. (2002) showed a dose-dependent increases in nitrate (indicative of
5 in vivo NO generation) with O3 exposure (0.12, 0.5, 1, or 2 ppm for 3 h). Functional studies of
6 enhanced pause (Penh) demonstrated increases with O3 which were also dose-dependent. Western
7 blot analysis of lung tissue showed increases in NOS-1, but not in NOS -3 or iNOS isoforms.
8 These results suggest that in mice NOS-1 may induce airway responsiveness by a neutrophilic
9 airway inflammation. The literature regarding the effects of O3 exposure on NOS activity is
10 complex and conflicting. Similarly, the issue of protein nitration as it relates to cell injury due to
11 O3 exposure is somewhat controversial.
12
13 5.2.3.6 Summary and Conclusions - Inflammation and Permeability Changes
14 Airway mucosa in the normal lung serves as an effective barrier that controls bidirectional
15 flow of fluids and cells between the air and blood compartments. Ozone disrupts this function,
16 resulting in an increase in serum proteins, bioactive mediators, and PMNs in the interstitium and
17 air spaces of the lung. Generally, the initiation of inflammation is an important component of
18 the defense process; however, its persistence and/or repeated occurrence can result in adverse
19 health effects.
20 The relative influence of concentration and duration of exposure (i.e., C x T) has been
21 investigated extensively in rats, using BALF protein as an endpoint. Although, the interaction
22 between C and T is complex, C generally dominated the response. The impact of T was
23 C-dependent (at higher Cs, the impact of T was greater); at the lowest C and T values, this
24 dependence appeared to be lost.
25 In rats, a single 3hr exposure to 0.5 ppm O3 produced a significant increase in both
26 permeability and inflammation, but a comparable exposure to 0.3 or 0.15 ppm did not produce
27 an effect. In a study comparing the responses of five species exposed to several concentrations
28 of O3, ranging from 0.2 to 2.0 ppm for 4 h, BAL was performed 18 h PE. Guinea pigs were the
29 most responsive (increased BALF protein at > 0.2 ppm); rabbits were the least responsive (effect
30 at 2.0 ppm only); and rats, hamsters, and mice were intermediate (effects at > 1.0 ppm). Among
31 rat strains, an acute exposure to O3 resulted in a greater injury, inflammation BALF levels of IL-
December 2005 5-24 DRAFT-DO NOT QUOTE OR CITE
-------
1 6 in Wistar than in SD or F344 rats. As exposures continue for 3 to 7 days, the increases in
2 BALF protein and PMNs typically peak after a few days (depending upon species tested and
3 exposures) and return towards control even with continuing exposure.
4 Other factors that have been studied for potential impact on the effects of O3 include age,
5 gender, nutritional status, genetic variability, exercise and exposure to co-pollutants. The effects
6 of age on lung inflammation are not well known. After an acute exposure to 0.8 or 1 ppm, young
7 mice, rats, and rabbits had greater changes in prostaglandins in BALF, but there were no age-
8 dependent effects on BALF protein or cell number. Comparisons of male and female animals,
9 and vitamin C or ascorbate deficiency did not reveal significant differences in the effects of O3,
10 but exercise during exposure increased susceptibility.
11 Ozone also increases the permeability from the air to the blood compartment. Ozone
12 (0.8 ppm; 2 h) caused a 2-fold increase of the transport of labeled DTPA from the rat tracheal
13 lumen to the blood. This coincided with a 2-fold increase in the number of endocytic vesicles in
14 epithelial cells that contained intraluminally instilled HRP as a tracer. These studies also suggest
15 an uneven disruption of tight junctions and alternate transport through endocytotic mechanisms.
16 In studies aimed at detecting the effects of O3 exposure on regional permeability, O3 increased
17 the transmucosal transport of DTPA and BSA more in the trachea and bronchoalveolar zone than
18 in the nose. These changes in barrier integrity may allow increased entry of antigens and other
19 bioactive compounds (e.g., bronchoconstrictors) into lung tissue. Data from analyses at regular
20 intervals PR indicate that maximal increases in BALF protein, albumin and number of PMNs
21 occur 8 to 18 h (depending on the study) after an acute exposure ceases.
22 Recent studies have placed a major focus on mediators released from inflammatory cells to
23 understand the mechanisms of O3-induced inflammation and injury. Cytokines and chemokines
24 have been shown to be released as a result of stimulation or injury of macrophages, epithelial
25 cells and PMNs. Exposure of guinea pig AMs recovered in BALF and exposed in vitro to
26 0.4 ppm O3 produced a significant increase in IL-6 and TNFa. An exposure of human AMs to
27 an identical O3 concentration increased TNFa, IL-lp, IL-6 and IL-8. The expression of MIP-2
28 mRNA or BALF levels of MIP-2 increased in mice and rats exposed to O3 concentrations
29 > 1 ppm. An increase after O3 exposure has also been reported for other cytokines and
30 inflammatory mediators, including CINC and fibronectin. The CINC mRNA expression was
31 associated with neutrophilia at 24 hrs PE. Ozone exposure of mice also caused an increase in
December 2005 5-25 DRAFT-DO NOT QUOTE OR CITE
-------
1 IL-6, MlP-la and eotaxin in mice. Further understanding of the role of mediators has come
2 from studies utilizing antibodies and inhibitors of known specificity. In these studies treatment
3 of rats with an anti IL-6 receptor antibody prior to a nighttime exposure to O3 abolished
4 O3-induced cellular adaptive response following a subsequent exposure. Studies utilizing
5 antibodies to selected pro- or anti-inflammatory cytokines suggest a role of TNFa, interleukin-10
6 (IL-10) and IL-lp in O3-induced changes in permeability, inflammation and cytokine release.
7 Studies investigating the mechanisms of PMN recruitment in the lung have explored the
8 role of cell adhesion molecules that mediate PMN-endothelial cell interactions. An increase in
9 tissue expression of 1C AM-1 occurred in mice exposed to 2 ppm O3. A comparable pattern of
10 time-related changes in total protein, fibronectin and alkaline phosphatase activity in the BALF
11 was observed in rats exposed to 1 ppm O3. In monkeys, the O3-induced inflammation was
12 blocked by treatment with a monocolonal antibody to CD 18, suggesting dependence of PMN
13 recruitment on this adhesion molecule. Together, these studies support the role of extracellular
14 matrix protein and cell adhesion molecules in lung inflammation and injury.
15 Ozone exposure also affects macrophage functions, and consequently their role in lung
16 inflammation. Macrophages isolated from O3-exposed mice produced increased amounts of NO,
17 superoxide anion and PGE2, but production of these mediators by macrophages from NOS
18 knockout mice was not elevated. Additionally, mice deficient in NOS or mice treated with NG-
19 monomethyl-L-arginine, an inhibitor of total NOS, were protected from O3-induced
20 permeability, inflammation and injury, suggesting a role of NO in the production of O3 effects.
21
22 5.2.4 Morphological Effects
23 Most mammalian species show generally similar morphological responses to < 1 ppm O3
24 which differ only by region, cell type, exposure parameters, and length of time between exposure
25 and examination. Constant low exposures to O3 create an early bronchoalveolar exudation,
26 which declines with continued exposure and drops in the PE period. Epithelial hyperplasia also
27 starts early, increases in magnitude for several weeks, plateaus with continuing exposure, and
28 declines slowly during PE. Interstitial fibrosis has a later onset, continues to increase throughout
29 the exposure, and can continue to increase after the exposure ends. Nonhuman primates respond
30 more than rats at this concentration, due to differences in antioxidants, the CAR (predicted to
31 receive the highest dose of O3), the presence of respiratory bronchioles, acinar volume, and
December 2005 5-26 DRAFT-DO NOT QUOTE OR CITE
-------
1 differences in the nasal cavity's ability to "scrub" the O3. Ciliated epithelial cells of the airway,
2 Type 1 epithelial cells of the gas-exchange region, and ciliated cells in the nasal cavity are the
3 cells most affected by O3. Ciliated cells are replaced by nonciliated cells (which are unable to
4 provide clearance function) and Type 1 cells are replaced by Type 2 cells, which are thicker and
5 produce more lipids. Inflammation also occurs, especially in the CAR, wherein the tissue is
6 thickened as collagen accumulates. At exposures of 0.25 ppm O3 (8 h/day, 18 mo) in monkeys,
7 the distal airway is remodeled as bronchiolar epithelium replaces the cells present in alveolar
8 ducts. In both rodents and monkeys, it appears that the natural seasonal patterns of O3 exposure
9 alters morphology more than continuous exposures, thus long-term animal studies with
10 uninterrupted exposures may underestimate morphological effects.
11
12 5.2.4.1 Short Term Exposure Effects
13 Morphological effects of key exposure studies generally lasting less than 1 week are
14 summarized in Table AX5-5. Harkema et al. (1997a) reviewed toxicological studies of the nasal
15 epithelial response to short-term O3. New information regarding the effects of O3 in this region
16 include demonstrations that the topical anti-inflammatory corticosteriod fluticasone propionate
17 prevents inflammation and mucous cell metaplasia in rats after cumulative O3 exposure (0.5 ppm
18 O3, 8h/day, for 3 or 5 days) (Hotchkiss et al., 1998). Exposure to bacterial endotoxin, a common
19 ambient air toxicant, can potentiate mucous cell metaplasia in the nasal transitional epithelium of
20 rats caused by a previous O3 exposure (Fanucchi et al., 1998). Male F344/N Hsd rats were
21 intranasally instilled with endotoxin after exposure to filtered air (FA) or 0.5 ppm O3, (8 h/d for 3
22 d). Mucous cell metaplasia was not found in the air/endotoxin group, but was found in the
23 O3/saline group and was most severe in the O3/endotoxin group. A similar synergistic effect was
24 demonstrated by Wagner et al. (2001a,b) with exposure of Fischer rats to O3 and endotoxin.
25 Ozone alone created epithelial lesions in the nasal transitional epithelium, while endotoxin alone
26 caused lesions in the respiratory epithelium of the nose and conducting airways. The enhanced
27 O3-induced mucous cell metaplasia was related to neutrophilic inflammation.
28 Pre-metaplastic responses, such as mucin mRNA upregulation, neutrophilic inflammation,
29 and epithelial proliferation, were shown to be responsible for O3-induced mucous cell metaplasia
30 in the transitional epithelium of rats (Cho et al., 1999a, 2000). Male F344/N rats exposed to O3,
31 (0.5 ppm,8 h/d for 1, 2, or 3 d) demonstrated a rapid increase in an airway-specific mucin gene
December 2005 5-27 DRAFT-DO NOT QUOTE OR CITE
-------
1 mRNA rapidly after exposure to O3, both before and during the onset of mucous cell metaplasia.
2 Neutrophilic inflammation coincided with epithelial DNA synthesis and upregulation of rMuc-
3 SAC, but was resolved before the development of epithelial metaplasia. The mucous cell
4 metaplasia was neutrophil-dependent, whereas O3-induced epithelial cell proliferation and mucin
5 gene upregulation were neutrophil-independent.
6 Dormans et al. (1999) compared the extent and time course of fibrotic changes in mice,
7 rats, and guinea pigs exposed to 0.2 and 0.4 ppm O3 for 3, 7, 28, and 56 days. They found a
8 concentration-related centriacinar inflammation in all three species, with a maximum after 3 days
9 of exposure and total recovery within 3 days after exposure. Repair of O3 damage by removal of
10 injured epithelial cells is enhanced by the influx of neutrophils (Hyde et al., 1999; Veseley et al.,
11 1999b; Miller et al., 2001; see Section 5.2.3). Labeling indices for rat nasal transitional
12 epithelial cell DNA were greatest 20 to 24 h after O3 (0.5 ppm for 8h) exposure, but still greater
13 than control by 36 h PE (Hotchkiss et al., 1997).
14 Very few published studies have explicitly explored susceptibility factors such as species,
15 gender, age, antioxidant defense, acute and chronic airway disease, and exercise. Most typical
16 laboratory species studied have qualitatively similar effects associated with O3 exposure.
17 Dormans et al. (1999) compared morphological, histological, and biochemical effects in the rat,
18 mouse, and guinea pig after O3 exposure and after recovery in clean air. Wistar RIV:Tox male
19 rats, NIH male mice, and Hartley Crl:(HA)BR male guinea pigs were continuously exposed to
20 FA, 0.2, or 0.4 ppm for 3, 7, 28, and 56 days. Recovery from 28 days of exposure was studied at
21 intervals of 3, 7, and 28 days PE. The mouse was the most sensitive as shown by a concentration
22 and exposure-time dependent persistence of bronchiolar epithelial hypertrophy, elevated lung
23 enzymes, and slow recovery from exposure. Exposure to the high dose for 56 d in both rats and
24 guinea pigs caused increased amounts of collagen in ductal septa and large lamellar bodies in
25 Type II cells. The inflammatory response was greater in the guinea pig. Overall, the authors
26 rated mice as most susceptible, followed by guinea pigs and rats.
27 Ferrets, monkeys and rats were exposed to O3 (1.0 ppm, 8 h) to compare airway effects
28 Sterner-Kock et al. (2000). The ferrets and monkeys had similar epithelial necrosis and
29 inflammation that was more severe than that found in rats. Because ferrets have a similar
30 pulmonary structure as humans (e.g., well-developed respiratory bronchioles and submucosal
31 glands), the authors concluded that the ferret would be a better model than rodents for O3-
December 2005 5-28 DRAFT-DO NOT QUOTE OR CITE
-------
1 induced airway effects. Age susceptibility is dependent on the endpoint examined. One new
2 study (Dormans et al., 1996) demonstrated that O3- induced centriacinar lesions are larger in
3 younger rats than in older rats.
4 Some new studies have examined O3-induced morphological effects in compromised
5 laboratory animals. Rats with endotoxin-induced rhinitis were more susceptible to mucous cell
6 metaplasia in the nasal transitional epithelium caused by a 3-day exposure to 0.5 ppm O3 (Cho
7 et al.,1999b). Wagner et al. (2002) reported a similar O3-induced enhancement of inflammatory
8 and epithelial responses associated with allergic rhinitis. Brown Norway rats were exposed to
9 0.5 ppm O3, 8h/day for 1 day or 3 consecutive days and then immediately challenged intranasally
10 with either saline or ovalbumin (OVA). Multiple exposures to O3 caused greater increases in
11 mucosubstances produced in the nose by allergen challenge.
12 Recent research has focused on the concept of O3 susceptible and non-susceptible sites
13 within the respiratory tract, including in situ antioxidant status and metabolic activity. Plopper
14 et al. (1998) examined whether the variability of acute epithelial injury to short-term O3 exposure
15 within the tracheobronchial tree is related to local tissue doses of O3 or to local concentrations of
16 reduced glutathione (GSH). Adult male rhesus monkeys exposed to O3 (0.4 or 1.0 ppm for 2 h)
17 demonstrated significant cellular injury at all sites, but the most damage, along with increased
18 inflammatory cells, occurred in the proximal respiratory bronchiole. A significant reduction in
19 GSH was found in the proximal bronchus at 0.4 ppm O3 and in the respiratory bronchiole at
20 1.0 ppm O3. A significant decrease in the percent of macrophages, along with significant
21 increases in the percent of neutrophils and eosinophils, and a doubling of total lavage protein,
22 were found after exposure to 1.0 ppm O3 only. The authors concluded that the variability of
23 local O3 dose in the respiratory tract was related to inhaled O3 concentration and was closely
24 associated with local GSH depletion and with the degree of epithelial injury.
25 Plopper and colleagues (e.g., Watt et al.,1998; Paige et al., 2000) explored the site-specific
26 relationship between epithelial effects of O3 exposure and the metabolism of bioactivated
27 compounds within the respiratory tract of rats. The distribution of CYP2E1-dependent activity,
28 measured with a selective substrate (p-nitrocatechol), was found to be highest in the distal
29 bronchioles and minor daughter airways, and lower in the lobar bronchi and major daughter
30 airways. Short-term O3 exposure (1 ppm for 8 h) increased CYP2E1 activity in the lobar
31 bronchi/major daughter airways only; however, long-term O3 exposure (1 ppm for 90 days)
December 2005 5-29 DRAFT-DO NOT QUOTE OR CITE
-------
1 decreased CYP2E1 activity in the major and minor airways, further complicating the
2 interpretation of O3 effects based on concentration and duration of exposure and recovery. Rats
3 treated i.p. with 1-nitronaphthalene, a pulmonary toxicant requiring metabolic activation, and
4 exposed to 0.8 ppm O3, 8h/day for 90 days showed greater histopathologic and morphometric
5 effects in the CAR of the lung (Paige et al., 2000). Despite reported tolerance to oxidant stress
6 after long-term O3 exposure, there was increased severity of ciliated cell toxicity.
7
8 5.2.4.2 Summary of Short-Term Morphological Effects
9 Short-term exposures to O3 cause similar alterations in lung structure in a variety of
10 laboratory animal species at concentrations of 0.15 ppm in rats and lower concentrations in
11 primates. Cells in the CAR are the primary targets of O3, but ciliated epithelial cells in the nasal
12 cavity and airways and Type 1 epithelial cells in the gas exchange region are also targets. New
13 work has shown that a topical anti-inflammatory corticosteroid can prevent these effects in nasal
14 epithelia, while exposure to bacterial endotoxin can potentiate the effects. Ozone-induced
15 fibrotic changes in the CAR are maximal at 3 d of exposure and recover 3 d PE with exposures
16 of 0.2 ppm in rodents. New studies of susceptibility factors demonstrated that ferrets and
17 monkeys have similar inflammatory and necrotic responses to 1 ppm O3, which differs from
18 lesser injury seen in rats. Rats with induced allergic rhinitis are more susceptible to 0.5 ppm
19 than are controls. Important new work has demonstrated variability of local O3 dose and
20 subsequent injury in the RT due to depletion of GSH. The proximal respiratory bronchiole
21 receives the most acute epithelial injury from exposures < 1 ppm, while metabolic effects were
22 greatest in the distal bronchioles and minor daughter airways.
23
24 5.2.4.3 Long Term Exposure Effects
25 Summaries of new studies of morphological effects of exposure lasting longer than 1 week
26 are listed in Table AX5-6 in Annex AX5. In general, as the duration of exposure lengthens,
27 there is not a concomitant linear increase in the intensity of effect of a given endpoint. Rather, as
28 exposure proceeds past 1 week to 1 year, Type 1 cell necrosis and inflammatory responses
29 generally decrease to near control values, and hyperpalstic and fibrotic changes remain elevated.
30 After long-term exposure ended, some indicies of fibrosis persisted and in some cases became
31 more severe during PE periods in clean air.
December 2005 5-30 DRAFT-DO NOT QUOTE OR CITE
-------
1 Effects of O3 on the upper respiratory tract of F344 rats exposed to O3 (0.12, 0.5, or
2 1.0 ppm for 20 months) included marked mucous cell metaplasia in the rats exposed to 0.5 and
3 1.0 pm O3, but not at 0.12 ppm O3 (Harkema et al., 1997a) . In a follow-up study, hyperplasia
4 was found in the nasal epithelium of rats exposed to 0.25 and 0.5 ppm, 8h/day, 7 days/week, for
5 13 weeks (Harkema et al., 1999). The mucous cell metaplasia, and associated intraepithelial
6 mucosubstances, induced by 0.5 ppm O3 persisted for 13 weeks after exposure. An acute (8-h)
7 exposure to 0.5 ppm O3 13 weeks after the chronic exposure induced an additional increase of
8 mucosubstances in the nasal epithelium of rats but not in rats chronically exposed to 0 or
9 0.25 ppm O3. The persistent nature of the O3-induced mucous cell metaplasia in rats reported in
10 this study suggests that O3 exposure may have the potential to induce similar long-lasting
11 alterations in the airways of humans.
12 No significant changes in nasal tissue were seen in rats continuously exposed for 49 days
13 to the ambient air of Mexico City, Mexico (Moss et al., 2001), which is in contrast to two rat
14 studies which did demonstrate development of secretory hyperplasia in rats exposed to ambient
15 air of Sao Paulo (Saldiva et al., 1992; Lemos et al., 1994). Because of the persistent nature of
16 these changes in the controlled studies with rats, and the fact that the upper airways of humans
17 are probably more sensitive, like the monkey, the authors suggested that long-term exposure to
18 ambient levels of O3 could induce significant nasal epithelial lesions that may compromise the
19 upper respiratory tract defense mechanisms of exposed human populations.
20 Rats exposed to 0.5 ppm O3 for 1 month exhibited Bcl-2 in protein extracts of nasal
21 epithelium (Tesfaigzi et al., 1998). Further, after 3 and 6 months of exposure, the number of
22 metaplastic mucous cells in the transitional epithelium was indirectly related to the percentage of
23 cells that were Bcl-2 positive . Cells from rats exposed to FA did not express any Bcl-2. This
24 study suggests that apoptosis regulators like Bcl-2 may play a role in the development and
25 resolution of mucous cell metaplasia in the nasal airway
26 A spectrum of lesions was reported (Herbert et al., 1996) in the nasal cavity and
27 centriacinar lung of male and female mice exposed to 0.5 or 1.0 ppm of O3 for 2 years, which
28 persisted with continued exposure for 30 months. These lesions included bone loss in the
29 maxilloturbinates, mucosal inflammation, mucous cell metaplasia in the nasal transitional
30 epithelium and increased interstitial and epithelial thickening in the proximal alveolar region.
31 In the CAR, there were increased numbers of nonciliated cells. However, changes in other
December 2005 5-31 DRAFT-DO NOT QUOTE OR CITE
-------
1 endpoints including lung function and lung biochemistry were not evident. The investigators'
2 interpretation of the entire study is that rodents exposed to the two higher O3 concentrations had
3 some structural hallmarks of chronic airway disease in humans. This interpretation is
4 strengthened further by a comparative pathology study of young (11 to 30 years old) accident
5 victims from Los Angeles, CA, and Miami, FL, (Sherwin et al., 2000) showing increased scores
6 for extent and severity of chronic inflammation in the CAR of Los Angeles residents. (See
1 Chapter 7 for a more detailed discussion of the population-based studies on O3.)
8 A long-term study using a simulated, seasonal O3-exposure pattern in infant monkeys was
9 reported by Plopper and colleagues (Evans et al., 2003; Schelegle et al., 2003a; Chen et al., 2003
10 Plopper and Fanucchi, 2000; Fanucchi et al., 2000) using. Infant rhesus monkeys (30 days old)
11 were exposed to FA, house dust mite allergen aerosol (HDMA), or O3 + HDMA. The 0.5 ppm
12 O3 exposures were 8 h/day for 5 days, every 14 days for a total of 11 O3 episodes. Half of the
13 monkeys were sensitized to house dust mite allergen (Dermatophagoides farinae) at 14 and
14 28 days of age. The sensitized monkeys were exposed to HDMA for 2h/day on Days 3-5 of the
15 FA or O3 exposures. The lungs were removed during the last FA exposure and the right and left
16 cranial and right middle lobes were separately inflation fixed. Microdisection and morphometric
17 analyses were performed on the conducting airways to the level of the most proximal respiratory
18 bronchiole. Repeated exposures to O3 or O3 + HDMA over a 6-month period resulted in an
19 atypical development of the basement membrane zone of airways in nonsensitized developing
20 monkeys. Remodeling in the distal conducting airways was found in the sensitized monkeys as a
21 result of the damage and repair processes occurring with repeated exposure (Evans et al., 2003;
22 Schelegle et al., 2003a; Fanucchi et al., 2000). Lung function changes in these monkeys
23 (Schelegle et al., 2003b), and associated adaptation of the respiratory motor responses (Chen
24 et al., 2003), are described in Section 5.2.5.2. Collectively, these findings provide a
25 pathophysiologic basis for changes in airway function described in children growing up in
26 polluted metropolitan areas (e.g., Tager, 1999) (See Chapter 7).
27 Necropsy of the left caudal lobe of these infant monkeys showed accumulation of
28 eosinophils and mucous cells within the combined epithelium and interstitium compartments in
29 the conducting airways and in the terminal/respiratory bronchioles (Schelegle et al., 2003a) .
30 House dust mite sensitization and HDMA challenge alone, or combined with O3 exposure,
31 resulted in significantly greater eosinophil accumulation in the conducting airways when
December 2005 5-32 DRAFT-DO NOT QUOTE OR CITE
-------
1 compared to FA and O3 only exposures. A significant accumulation of eosinophils was found in
2 the terminal/respiratory bronchioles of the sensitized monkeys challenged with HDMA when
3 compared to monkeys exposed to FA, O3, and FIDMA + O3. The mean mass of mucous cells
4 increased in the fifth generation conducting airways of sensitized animals challenged with
5 FIDMA alone and when combined with O3 exposure, and in the terminal bronchioles of
6 sensitized animals exposed to FIDMA + O3. The tracheal basement membrane of HDMA-
7 sensitized monkeys exposed to FIDMA or to HDMA + O3 was significantly increased over
8 controls; however, there were no significant changes in the airway diameter of proximal and
9 mid-level airways. The authors interpreted these findings to indicate that the combination of
10 cyclic O3 exposure and HDMA challenge in HDMA-sensitized infant monkeys act
11 synergistically to produce an allergic-reactive airway phenotype characterized by significant
12 eosinophilia of midlevel conducting airways, transmigration of eosinophils into the lumen, and
13 an altered structural development of conducting airways that is associated with increased airway
14 resistance and nonspecific airway reactivity (see Section 5.2.5). Exposures of sensitized young
15 monkeys to HDMA alone, or to O3 alone, resulted in eosinophilia of the mid-level conducting
16 airways and the terminal/respiratory bronchioles, but without alterations in airway structure or
17 function.
18 Examination of development of the tracheal basement membrane zone (BMZ) in these
19 monkeys (Evans et al., 2003) showed that with exposures to either O3 or HDMA + O3 BMZ
20 development was affected. Abnormalities in the BMZ included: (1) irregular and thin collagen
21 throughout the BMZ; (2) perclecan depeleted or severely reduced; (3) FGFR-1
22 immunoreactivity was reduced; (4) FGF-2 immunoreactivity was absent in perlecan-deficient
23 BMZ, but was present in the lateral intercelluar space (LIS), in basal cells, and in attenuated
24 fibroblasts; (5) syndecan-4 immumoreactivity was increased in basal cells. The authors interpret
25 these data to suggest that O3 targets cells associated with synthesis of epithelial BMZ perlecan.
26 The absence of FGF-2, normally stored in the BMZ, could affect downstream signaling in
27 airway epithelium and could be responsible for the abnormal development of the airway seen in
28 this study, and thus be an important mechanism modulating O3-induced injury. Midlevel bronchi
29 and bronchioles from these monkeys (Larson et al., 2004) demonstrated decrements in the
30 density of epithelial nerves in the axial path between the sixth and seventh airway generations in
31 exposures to O3. Combined O3+HDMA exposures exacerbated this reduction. They attribute
December 2005 5-33 DRAFT-DO NOT QUOTE OR CITE
-------
1 this loss of nerve plexuses to neural regression or stunted nerve development, the latter
2 corroborated by the Evans et al. (2003) finding of decreased growth factors following O3
3 exposure. Additionally, they found streaks or clusters of cells immunoreactive for protein gene
4 product 9.5 (PGP 9.5, a pan-neuronal marker) and negative for calcitonin gene-related peptide.
5 The functional significance of this is unknown but suggests to the authors a possible injury-
6 repair process induced by O3.
7 Remodeling of the distal airways and CAR is one of the most disturbing aspects of the
8 morphological changes occurring after long-term exposure to O3. Recently, bronchiolization
9 was reported in rats exposed to 0.4 ppm O3 for only 56 days (van Bree et al., 2001). They also
10 found collagen formation progressively increased with increasing O3 exposure and remained
11 increased into PE recovery. In addition to centriacinar remodeling, Pinkerton et al. (1998)
12 reported thickening of tracheal, bronchial, and bronchiolar epithelium after 3 or 20 months
13 exposure to 1 ppm O3, but not to 0.12 ppm. Although some older literature had reported that
14 chronic exposures to < 1.0 ppm O3 cause emphysema, no current literature supports this
15 hypothesis.
16
17 5.2.4.4 Summary and Conclusions - Long-Term Morphological Effects
18 The progression of effects during and after a chronic exposure at a range of 0.5 to 1.0 ppm
19 is complex, with inflammation peaking over the first few days of exposure, then dropping, then
20 plateauing, and finally, largely disappearing. Epithelial hyperplasia follows a somewhat similar
21 pattern. In contrast, fibrotic changes in the tissue increase very slowly over months of exposure,
22 and, after exposure ceases, the changes sometimes persist or increase. Pattern of exposure in this
23 same concentration range determines effects, with 18 mo of daily exposure causing less
24 morphologic damage than exposures on alternating months. This is important as environmental
25 O3 exposure is typically seasonal. Plopper and colleagues' long term study of infant rhesus
26 monkeys exposed to simulated, seasonal O3 (0.5 ppm 8h/day for 5 days, every 14 days for 11
27 episodes) demonstrated : 1) remodeling in the distal airways; 2) abnormalities in tracheal
28 basement membrane; 3) eosinophil accumulation in conducting airways; 4) decrements in airway
29 innervation. These findings advance earlier information regarding possible injury-repair
30 processes occurring with seasonal O3 exposures.
31
December 2005 5-34 DRAFT-DO NOT QUOTE OR CITE
-------
1 5.2.5 Effects on Pulmonary Function
2 5.2.5.1 Acute and Short-Term Exposure Effects on Pulmonary Function
3 Numerous pulmonary function studies of the effects of short-term O3 exposure (defined
4 here as < 1 week of exposure) in several animal species have been conducted and generally show
5 responses similar to those of humans (e.g., increased breathing frequency, decreased tidal
6 volume, increased resistance, decreased forced vital capacity [FVC] and changes in the
7 expiratory flow-volume curve). These effects are seen at 0.25 to 0.4 ppm O3 for several h in a
8 number of species. At concentrations of > 1 ppm, breathing mechanics (compliance and
9 resistance) are affected. The breathing pattern returns to normal after O3 exposure. In rats
10 exposed to 0.35 to 1 ppm O3 for 2 h/day for 5 days, there is a pattern of attenuation of pulmonary
11 function responses similar to that observed in humans. Concurrently, there was no attenuation of
12 biochemical indicators of lung injury or of morphological changes.
13 New work demonstrating attenuation of pulmonary functions was completed by Wiester
14 et al. (1996) who exposed male Fischer 344 rats to 0.5 ppm O3 for either 6 or 23 h/day over 5
15 days. Ozone-induced changes in lung volume were attenuated during the 5 exposure days and
16 returned to control levels after 7 days recovery. The responses to repeated O3 exposure in rats
17 were exacerbated by reduced ambient temperature, presumably as a result of increased metabolic
18 activity.
19 Researchers have utilized inbred mouse strains with varying ventilatory responses to O3 to
20 attempt to model susceptible populations. As differences were seen in inflammatory responses
21 to acute O3 exposures in C57BL/6J and C3FI/HeJ mice, comparisons were made of their
22 ventilatory responses also (Tankersley et al., 1993). Following an exposure of 2 ppm O3 for 3 h,
23 breathing frequency (f), tidal volume (VT), and minute ventilation were measured 1 and 24 h in
24 both normocapnia (or air at -0% CO2) and hypercapnia (5 or 8% CO2). They demonstrated that
25 acute O3 exposures caused altered hypercapnic ventilatory control, which varied between strains.
26 This suggested to the authors that O3-induced alterations in ventilation are determined, at least in
27 part, by genetic factors.
28 Paquette et al. (1994) measured ventilatory responses in C57BL/6J and C3H/HeJ mice
29 given repeated subacute exposures. The two strains had differing responses to both normocapnia
30 and hypercapnia. Normocapnic VE was greater following subacute O3 exposure in C57BL/6J
31 mice than in C3H/HeJ mice, due to increased and reduced VT, respectively. This suggests that
December 2005 5-35 DRAFT-DO NOT QUOTE OR CITE
-------
1 the increased VT in C57BL/6J mice may contribute to the increased susceptibility to lung injury
2 due to a greater dose of O3 reaching the lower lung. Hypercapnic ventilatory responses
3 following subacute O3 exposures demonstrated reduced VE (due to decreased VT) in C57BL/6J
4 only. Evaluations of O3 dosimetry were performed in these two strains using 18O3-labeled ozone
5 (Slade et al., 1997). Immediately after exposures of 2 ppm 18O3 for 2-3 h, C3H/HeJ mice had
6 46% less 18O in lungs and 61% less in trachea, than C57BL/6J. Additionally, C3H/HeJ mice had
7 a greater body temperature decrease following O3 exposure than C57BL/6J mice, suggesting that
8 the differences in susceptibility to O3 are due to differences the ability to decrease body
9 temperature and, consequently decrease the dose of O3 to the lung.
10 Tracheal transepithelial potential (VT) has also been shown to differ in eight mouse strains
11 6 h after exposure to 2 ppm O3 for 3 h (Takahashi et al., 1995b). AKR/J, C3H/HeJ, and CBA/J
12 were identified as resistant strains and 129/J, A/J, C57BL/6J, C3HeB/FeJ and SJL/J were
13 identified as susceptible strains. The authors noted that strains' responses to this parameter did
14 not show concordance with inflammatory responses, suggesting to the authors that the two
15 phenotypes are not controlled by the same genetic factors.
16 Savov et al. (2004) characterized ventilatory responses in nine mouse strains exposed to O3
17 (2.0 ppm O3 for 3 h). Table AX5-4 in Annex AX5 lists the baseline Penh, the Penh following O3,
18 and the Penh response to methacholine (MCh) following O3. C57BL/6J was hyporeactive to MCh
19 prior to O3, but was very responsive to MCh following O3. Conversely, C3H/HeJ had an
20 intermediate baseline Penh and a small response to MCh following O3 exposure. This study
21 corroborates the evidence of no consistent relationship between respiratory Penh and
22 inflammation.
23
24 5.2.5.2 Summary and Conclusions - Short- and Long-Term Effects on
25 Pulmonary Function
26 Early work has demonstrated that during acute exposure of -0.2 ppm O3 in rats, the most
27 commonly observed alterations are increased frequency of breathing and decreased tidal volume
28 (i.e., rapid, shallow breathing). Exposures of-1.0 ppm O3 affect breathing mechanics
29 (compliance and resistance). Additionally, decreased lung volumes are observed in rats with
30 acute exposures at levels of 0.5 ppm. New work utilizing inbred mouse strains with varying
31 ventilatory responses to O3 has suggested that: (1) control of the ventilatory response is
December 2005 5-36 DRAFT-DO NOT QUOTE OR CITE
-------
1 determined, at least in part, by genetic factors; (2) increased VT in some strains may contribute
2 to lung injury due to a greater dose of O3 reaching the lower lung; (3) some strains' ability to
3 reduce body temperature may account for their decreased O3-induce lung injury; (4) tracheal
4 transepithelial potential is determined, in part, by genetic factors. Importantly, the genetic loci
5 that appear to be modulating various aspects of pulmonary responses to O3 differ from each other
6 and from loci controlling inflammatory responses.
7 Exposures of 2 h/day for 5 days create a pattern of attenuation of pulmonary function in
8 both rats and humans without concurrent attenuation of lung injury and morphological changes,
9 indicating that the attenuation did not result in protection against all the effects of O3. Long-term
10 O3 exposure studies evaluating pulmonary function are not available. Earlier work has
11 demonstrated that repeated daily exposure of rats to an episodic profile of O3 caused small, but
12 significant decrements in lung function that were consistent with early indicators of focal
13 fibrogenesis in the proximal alveolar region, without overt fibrosis.
14
15 5.2.5.3 Ozone Effects on Airway Responsiveness
16 Effects of O3 on airway reactivity have been observed in a variety of species at an exposure
17 range of 0.5 to 1 ppm. Many of the new studies on pulmonary function in laboratory animals
18 allow a better prediction of the effects of O3 exposure on the exacerbation of asthma symptoms
19 and the risk of developing asthma in humans. However, it is necessary to understand the factors
20 that determine airway responsiveness across different mammalian species as discussed in
21 Chapter 4.
22 The physiological characteristics of asthma include intermittent airway obstruction and
23 increased airway responsiveness to various chemical and physical stimuli. Methods used to
24 assess airway responsiveness in humans include airway challenge with nonspecific
25 bronchoconstrictors (e.g., inhaled methacholine or histamine) and with indirect (e.g., inhalation
26 of adenosine monophosphate, hypertonic saline, mannitol) stimuli to bronchoconstriction
27 (Anderson, 1996). Laboratory animal studies employ intravenous agonist challenges as well as
28 inhalation challenges, though inhaled agonist challenges are preferred in humans. Sommer et al.
29 (2001) reported some differences in the two routes for bronchoconstrictor administration.
30 Traditional studies of airway responsiveness require sedation in both infants and laboratory
31 animals. Exercise testing is not possible with sedation unless exercise is "simulated" by
December 2005 5-37 DRAFT-DO NOT QUOTE OR CITE
-------
1 increasing ventilation using elevated F;CO2 and the need for artificial ventilation in laboratory
2 animal studies may cause breathing patterns that affect O3 deposition. load et al. (2000) reported
3 that when 1 ppm O3 for 90 min is administered to isolated rat lung at either 2.4 ml/40 bpm or
4 1.2 m/80 bpm, the more rapid breathing pattern elicits less epithelial cell injury than the slower
5 breathing pattern. Though this study design does not really model rapid shallow breathing
6 elicited in the intact animal, it shows greater reduction in injury in the proximal axial airway
7 compared to its adjacent airway branch and terminal bronchiole. The rapid, shallow breathing
8 pattern protects the large conducting airways of rats, but causes a more even distribution of
9 epithelial cell injury to the terminal bronchioles (Schelegle et al., 2001). Postlethwait et al.
10 (2000) demonstrated that the conducting airways are the primary site of acute cytotoxicity from
11 O3 exposure. Three-dimensional mapping of the airway tree in SD rat isolated lung exposed to
12 0, 0.25, 0.5, or 1.0 ppm O3 showed a concentration-dependent increase in injured cells. Injury
13 was evident in proximal and distal conduction airways, lowest in terminal bronchioles, and
14 highest in the small side branches downstream of bifurcations. These exposure levels did not
15 concurrently elicit changes in LDH activity or total protein in BALF, suggesting that the
16 mapping technique is a more sensitive measure of injury and is useful in dosimetry studies.
17 Whole-body plethysmography of unanesthetized, unrestrained rodents has been used to
18 indirectly measure pulmonary resistance (Shore et al., 2001, 2002; Goldsmith et al., 2002; Jang
19 et al., 2002). However, these indices of inspiratory/expiratory pressure differences, including
20 enhanced pause (Penh) may be less sensitive than direct measurements of lung airflow resistance
21 (Murphy, 2002). Sommer et al. (1998) demonstrated that unrestrained guinea pigs have a daily
22 variability in pulmonary resistance that is similar to that occurring in humans, indicating that
23 circadian rhythms of airway caliber must be considered when performing airway challenge tests
24 in any species. Changes in airway structure caused by viral infections also must be considered
25 when evaluating laboratory animal studies. Animals with acute viral illness have morphological
26 evidence of inflammatory cell infiltration, bronchiolar wall edema, epithelial hyperplasia, and
27 increased airway mucous plugs that can cause airway narrowing, air trapping, and serious
28 functional changes in the lung (Folkerts et al., 1998).
29 Exercise-induced bronchoconstriction in humans appears to be mediated by changes in the
30 tonicity of the airway lining fluid (Anderson and Daviskas, 2000). Brannan et al. (1998) suggest
31 that a test in laboratory animals based on the inhalation of mannitol aerosol (hyperosmolar)
December 2005 5-38 DRAFT-DO NOT QUOTE OR CITE
-------
1 might be feasible and provide information similar to that from exercise challenges in cooperative
2 children and adults. Unfortunately, there have been few reports of mannitol or adenosine
3 monophosphate challenges in laboratory animals; most studies have utilized histamine,
4 methacholine, acetylcholine, or carbachol to determine outcome. In active humans with asthma,
5 adenosine monophosphate challenges appear to better reflect ongoing airway inflammation than
6 histamine or methacholine challenges (Polosa and Holgate, 1997; Avital et al, 1995a,b), and
7 might be useful in identifying mechanisms of asthma in laboratory animals and their
8 responsiveness to environmental pollutants.
9
10 Airway Responsiveness in Asthma
11 The increased responsiveness to bronchoconstrictor challenge in asthma is thought to
12 result from a combination of structural and physiological factors that include increased
13 inner-wall thickness, increased smooth-muscle responsiveness, and mucus secretion. These
14 factors also are likely to determine a level of innate airway responsiveness that is genetically
15 influenced. This baseline responsiveness is thought to be modulated in asthma by chronic
16 inflammation and airway remodeling (Stick, 2002). For example, about 90% of children with
17 asthma symptoms in the previous year will exhibit increased airway responsiveness to one or
18 more challenge tests (Sears et al., 1986); however, 10% of healthy children also will respond to
19 one or other of the challenge tests. Longitudinal studies in adults have shown that the
20 development of airway responsiveness is associated with persistence of symptoms (O'Conner
21 et al., 1995) suggesting airway remodeling. This hypothesis is in agreement with the
22 inconsistent relationship reported in the literature between airway responsiveness and markers of
23 inflammation.
24
25 Airway Responsiveness in Infants
26 The age at which nonspecific AHR is first seen in humans is unknown, but it is known that
27 infants show increased responsiveness compared to older children, probably due to differences in
28 dose received. When correction is made for this dose effect, infants and older children appear to
29 have a similar response to inhaled histamine (Stick et al., 1990; Stick, 2002). The importance of
30 this observation is that absolute values of airway responsiveness cannot be used to compare
31 airway responsiveness at different ages, and they certainly cannot be used to compare
December 2005 5-39 DRAFT-DO NOT QUOTE OR CITE
-------
1 responsiveness across different species. However, airway responsiveness can be tracked over
2 time within given populations, or alternatively, by using non-parametric analyses based on
3 ranking subjects at each time point. Such analyses have been used in birth-cohort studies to
4 investigate the role of airway responsiveness in the early genesis of asthma. One unique
5 birth-cohort study has shown that airway responsiveness at one month is a predictor of lung
6 function at six years (Palmer et al., 2001). Data from this study also show that the genetic
7 determinants of atopy and airway responsiveness are independent (Palmer et al., 2000).
8 In another study of infants with wheeze, persistence of AHR was associated with persistence of
9 symptoms, although airway responsiveness at one month of age was neither a sensitive nor a
10 specific predictor of outcome (Delacourt et al., 2001).
11 The human studies imply that airway responsiveness is a key factor in asthma, but it is not
12 clear if the factors that are important for airway responsiveness in early life are related to
13 inflammation, structure or physiology of the airways, or the combination of all three.
14 Furthermore, it is not clear how viruses, allergens and irritants in the environment modify innate
15 airway responses (Holt et al., 1999), but they are known to be important. Laboratory animal
16 studies have tried to answer some of these key questions.
17
18 Airway Responsiveness in Laboratory Animals
19 A large data base of laboratory animal research has been collected on the role of O3 in
20 producing an increase in AHR. Exposure levels (> 1 ppm for > 30 min) in many of these studies
21 are not environmentally relevant, but information can be obtained regarding the mechanisms of
22 action of O3 concerning: O3 concentration and peak response time, inhaled versus intravenous
23 challenge with nonspecific bronchoconstrictors, neurogenic mediation, neutrophilic
24 inflammation, and interactions with specific biological agents (e.g., antigens and viruses).
25 Many species of laboratory animals have been used to study the effects of O3 on airway
26 bronchoconstriction. Ozone-induced AHR in guinea pigs has been used to model human
27 bronchospasm (Kudo et al., 1996; van Hoof et al., 1996; 1997a; Matsubara et al., 1997a,b;
28 Sun and Chung, 1997; Aizawa et al., 1999a,b; Tsai et al., 1998; Nakano et al., 2000). Because
29 these studies were done at 2 to 3 ppm O3, these results are not directly relevant for extrapolation
30 to potential airway responses in humans exposed to ambient levels of O3. Humans with reactive
31 airway disease (e.g., asthma) appear to be sensitive to ambient levels of O3 (see Chapters 6
December 2005 5-40 DRAFT-DO NOT QUOTE OR CITE
-------
1 and 7) and the current understanding is that O3 exacerbates airway responsiveness to specific
2 allergens, presumably by nonspecifically increasing AHR.
3 Shore et al. (2000, 2003) have shown that O3-induced AHR is reduced in immature rats and
4 mice. SD rats exposed to 2 ppm O3 at ages 2, 4, 6, .8, or 12 weeks and A/J mice exposed to
5 0.3 to 3 ppm for 3 h at age 2, 4, 8, or 12 weeks had similar concentration-related decreases in
6 VE except at the youngest ages. This smaller decrement in VE suggested a delivered dose that
1 was much greater in the younger animals. This group (Shore et al., 2003) has also recently
8 shown that obese mice have greater ventilatory responses to O3. Exposures of 2.0 ppm O3 for 3
9 h to lean, WT C57BL/6J and ob/ob mice (mice with a genetic defect in the coding for leptin, the
10 satiety hormone) showed that the ob/ob mice had enhanced AHR and inflammation compared to
11 the WT mice. These data correlate with epidemiological data showing increased incidence of
12 asthma in overweight children.
13 Increased AHR to various nonspecific bronchoconstrictive agents (e.g., ACh,
14 methacholine, histamine, carbachol) given by inhalation or intravenous routes has been
15 previously shown in laboratory animals exposed to O3 concentrations < 1.0 ppm (Table 5-7).
16 Recently, Dye et al. (1999) showed hyperresponsiveness to methacholine in rats 2 h after
17 exposure to 2 ppm O3 for 2 h. AHR can be induced by specific antigens as well as O3. The most
18 commonly used laboratory animal model is the OVA sensitized guinea pig. Animals sensitized
19 with OVA have been shown to have similar responses to nonspecific bronchoconstrictors as
20 control animals; however, OVA-sensitized guinea pigs exposed to O3 showed increased AHR to
21 histamine (Vargas et al., 1994), which was further enhanced by an antigen challenge. When
22 exposed to O3 before sensitization, repeated exposures to very high levels (5.0 ppm) decreased
23 the OVA sensitization threshold; however, in already sensitized animals, a 2-h exposure to > 1.0
24 ppm enhanced airway responsiveness to OVA, suggesting that O3 exposure does not modify the
25 development of antigen-induced AHR and, in fact, may enhance AHR at high levels of exposure.
26 OVA-sensitized guinea pigs (Sun et al., 1997) and mice (Yamauchi et al., 2002) were used
27 to determine the enhancement of antigen-induced bronchoconstriction by acute, high-level O3
28 (1.0 ppm O3 for 1 h). Male Dunkin-Hartley guinea pigs were sensitized by i.p. injection of OVA
29 and exposed to O3 alone, OVA aerosol, or O3 + OVA. Ozone exposure alone increased
30 bronchial responsiveness to ACh at 3 h, but not 24 h, while OVA alone had no effect. Combined
31 exposure to O3 and OVA (1 ppm for 1 h, then 3 min OVA) increased bronchial responsiveness to
December 2005 5-41 DRAFT-DO NOT QUOTE OR CITE
-------
1 ACh 3 h after O3 exposure. At 24 h following O3 exposure, AHR increased when OVA
2 challenge was performed at 21 h, suggesting that O3 pre-exposure can potentiate OVA-induced
3 AHR. Neutrophil counts in the BALF increased at 3 and 24 h after O3 exposure alone but were
4 not further increased when O3 exposure was combined with OVA airway challenge; however
5 protein content of the BALF did increase at 3 and 24 h in the O3 and OVA groups. Thus, this
6 study also indicates that high-ambient O3 exposure can augment antigen (OVA)-induced AHR in
7 guinea pigs.
8 Yamauchi et al. (2002) sensitized male C57BL/6 mice by i.p. injection of OVA and then
9 exposed them to O3. The sensitized mice had AHR to methacholine. Ozone exposure caused
10 significant decreases in dynamic lung compliance, minute ventilation, and PaO2 in OVA-
11 sensitized mice, but not in controls. A marker of inflammation (soluble intercellular adhesion
12 molecule-1 [sICAM-1]) was elevated in the BAL fluid of OVA-sensitized mice, but sICAM-1
13 levels were not significantly changed by O3 exposure, indicating that the O3-induced AHR to
14 methacholine was not caused by O3-induced inflammation.
15 Ozone-induced AHR may be temporally associated with neutrophils (DeLorme et al.,
16 2002) and other inflammatory cells stimulated by leukotrienes (Stevens et al., 1995a), cytokines
17 (Koto et al., 1997), mast cells (Igarashi et al., 1998; Noviski et al., 1999), or by oxygen radicals
18 (Takahashi et al., 1993; Stevens et al., 1995b; Tsukagoshi et al., 1995; Kudo et al., 1996). Two
19 new studies, however, have shown that inflammation is not a prerequisite of AHR (Stevens et al.,
20 1994; Koto et al., 1997), and some investigators have suggested that O3-induced AHR may be
21 epithelium dependent (Takata et al., 1995; Matsubara et al., 1995; McGraw et al., 2000).
22 For example, neonatal rats pretreated with capsaicin, which will permanently destroy C-fibers
23 and prevent O3-induced release of neuropeptides (Vesely et al., 1999a), and then exposed to O3
24 when adults, showed a marked increase in airway responsiveness to inhaled aerosolized
25 methacholine (Jimba et al., 1995). Some investigators (Matsumoto et al., 1999; DeLorme et al.,
26 2002) have shown that respective intravenous pretreatment with neutrophil elastase inhibitor or
27 PMN antiserum can block O3-induced AHR; other investigators (Koto et al., 1995; Aizawa et al.,
28 1997; Takebayashi et al., 1998) have shown that depletion of tachykinins by capsaicin treatment,
29 or by a specific tachykinin receptor antagonist, can block the induction of AHR by O3. The
30 seemingly disparate responses in laboratory animals may be due to species- or strain-specific
31 differences in inherent reactivity to bronchoconstrictors, or to inherent differences in
December 2005 5-42 DRAFT-DO NOT QUOTE OR CITE
-------
1 susceptibility to O3-induced inflammation (Zhang et al., 1995; Depuydt et al., 1999; Dye et al.,
2 1999).
3 Studies that may be potentially relevant to ambient levels of O3 were conducted in vivo, in
4 an isolated perfused lung model, and in ex vivo lung segments using multihour and repeated
5 multihour exposures with ambient levels of O3. A study on the relationship between O3-induced
6 AHR and tracheal epithelial function was conducted in New Zealand white rabbits by Freed
7 et al. (1996). Rabbits exposed to O3 (0.2 ppm for 7 h) demonstrated significantly decreased PD
8 but no changes in lung resistance. Changes in the compartmentalized lung resistance, measured
9 in response to ACh challenge before and after bilateral vagotomy, were not significantly
10 different in air-exposed rabbits; however, bilateral vagotomy did enhance peripheral lung
11 reactivity in O3-exposed rabbits. The ACh-induced increase in lung resistance with O3 exposure
12 (140%) was two times higher than with air exposure, indicating that ambient-level O3 exposure
13 affects tracheal epithelial function in rabbits and increases central airway reactivity, possibly
14 through vagally-mediated mechanisms.
15 Pulmonary mechanics and hemodynamics were studied in the New Zealand white rabbit
16 isolated perfused lung model that allowed partitioning of the total pressure gradient into arterial,
17 pre- and post-capillary, and venous components ( Delaunois et al., 1998). Exposures to O3 (0.4
18 ppm for 4 h) were followed by evaluation of airway responsiveness to ACh, substance P (SP), or
19 histamine immediately or 48 h later. Ozone inhibited pulmonary mechanical reactivity to all
20 three bronchoconstrictors that persisted for 48 h and modified vasoreactivity of the vascular bed,
21 but only at 48 h PE. Arterial segmental pressure, normally insensitive to ACh and SP, was
22 significantly elevated by O3; precapillary segmental pressure decreased in response to Ach,
23 suggesting that O3 can induce direct vascular constriction, but the vascular responses are variable
24 and depend on the agonist used and on the species studied.
25 Airway responsiveness to the same three compounds was evaluated by Segura et al. (1997)
26 in guinea pigs exposed to O3 (0.15, 0.3, 0.6, or 1.2 ppm for 4 h). Ozone did not cause AHR to
27 ACh or histamine, except at the highest concentration (1.2 ppm O3) for histamine. However, O3
28 did cause AHR to SP at > 0.3 ppm, suggesting that O3 destroys neutral endopeptidases
29 (responsible for SP inactivation) in airway epithelial cells. Vargas et al. (1998), in a follow-up
30 study, demonstrated that guinea pigs chronically exposed to 0.3 ppm O3 for 4 h/day became
31 adapted to SP-induced AHR. Ozone caused increased sensitivity to SP after 1, 3, 6, 12, and
December 2005 5-43 DRAFT-DO NOT QUOTE OR CITE
-------
1 24 days of exposure that was associated with airway inflammation; however, after 48 days of
2 exposure, the increased sensitivity to SP was lost.
3 This study is in accordance with Szarek et al. (1995) who demonstrated that AHR
4 associated with acute O3 exposures does not persist during long-term exposure to ambient-levels
5 of O3 (< 1 ppm). Fischer 344 rats, exposed to 0.0, 0.12, 0.5, or 1.0 ppm O3, 6h/day, 5 days/week
6 for 20 months, demonstrated significantly reduced responses to bethanechol, ACh, and electrical
7 field stimulation in eighth generation airway segments. This suggests that some adaption had
8 taken place during long-term exposure, possibly increased inner wall thickness.
9 It is well known that the changes in breathing pattern and lung function caused by O3 are
10 attenuated with repeated daily exposures for at least 3 to 5 days. But guinea pigs exposed to
11 0.5 ppm O3, 8 h/day for 7 days showed enhancement of responsiveness of rapidly adapting
12 airway receptors (load et al., 1998). Repeated exposure increased receptor activity to SP,
13 methacholine, and hyperinflation; there were no significant effects on baseline or SP- and
14 methacholine-induced changes in lung compliance and resistance, suggesting that the
15 responsiveness of rapidly adapting receptors was enhanced.
16 Male and female Hartley guinea pigs exposed to O3 (0.1 and 0.3 ppm, 4 h/day, 4 days/week
17 for 24 weeks) were evaluated for airway responsiveness following ACh or OVA inhalation
18 challenges (Schlesinger et al., 2002a,b). Ozone exposure did not cause AHR in nonsensitized
19 animals but did exacerbate AHR to both ACh and OVA in sensitized animals that persisted for
20 4 weeks after exposure. The effects of O3 on airway responsiveness were gender independent
21 and were concentration-related for the ACh challenges.
22 Schelegle et al. (2003a) evaluated airway responsiveness in infant rhesus monkeys exposed
23 to a 5 day O3 episode repeated every 14 days over a 6-month period. Half of the monkeys were
24 sensitized to house dust mite allergen (HDMA; Dermatophagoid.es farinae) at 14 and 28 days of
25 age before exposure to a total of 11 episodes of O3 (0.5 ppm, 8 h/day for 5 days followed by
26 9 days of FA), HDMA, or O3 + HDMA. Baseline Raw was significantly elevated after 10
27 exposure episodes in the HDMA + O3 group compared to the FA, HDMA, and O3 exposure
28 groups. Aerosol challenge with HDMA at the end of the 10th episode did not significantly affect
29 Raw, VT, fB, or SaO2. Aerosol challenge with histamine was not significantly different after
30 6 episodes; however, the EC150 Raw for the HDMA + O3 group was significantly reduced after
31 10 episodes when compared to the FA, HDMA, and O3 exposure groups, indicating the
December 2005 5-44 DRAFT-DO NOT QUOTE OR CITE
-------
1 development of AHR in this group sometime between episodes 6 and 10. The results are
2 consistent with altered structural development of the conducting airways.
3 During repeated episodic exposures to O3, respiratory responses are first altered to a rapid,
4 shallow breathing pattern, which has long been considered protective, especially to the deep
5 lung. This dogma has been discounted recently as discussed above (Schelegle et al., 2001).
6 Alfaro et al. (2004) examined the site-specific deposition of 18O (1 ppm 2 h) at breathing
7 frequencies of 80, 120, 160, or 200 breaths/minute. At all frequencies, parenchymal areas had a
8 lower content of 18O than trachea and bronchi. As breathing frequency increased from 80 to 160
9 bpm, the deposition showed a reduction in midlevel trachea and an increase in both mainstream
10 bronchi. At this frequency there was also an increase in deposition in parenchyma supplied by
11 short (cranial) airway paths, consistent with results seen by Schelegle et al., (2001). At 200 bpm
12 18O deposition in trachea increased, concurrent with increases in right cranial and caudal bronchi
13 regions. Right cranial parenchymal content decreased at 200 bpm, whereas right caudal
14 parenchymal levels did not change at any breathing frequency. These two studies provide
15 evidence that O3-induced rapid, shallow breathing creates a more evenly distributed injury
16 pattern, with possibly greater protection from focal injury to the large conducting airways
17 including the trachea and the left mainstem bronchus.
18 Another study of the adaptive phenomena in SD rats used an exposure paradigm consisting
19 of 5 days of daily 8 h 1 ppm O3 exposures followed by 9 days of recovery in FA (Schelegle
20 et al., 2003b). This O3/FA pattern was repeated for 4 cycles and demonstrated that the O3 -
21 induced rapid shallow breathing pattern was followed by adaptation occurred with each cycle.
22 But the release of SP from the trachea , the neutrophil content, and cell proliferation became
23 attenuated after the first cycle, suggesting a disconnect from the rapid shallow breathing
24 response. Hypercellularity of the CAR epithelium and thickening of the CAR interstitium, not
25 linked to changes in cell proliferation, were also found. The authors suggest mechanism(s) of
26 injury from repeated O3 exposures consisting of diminished neutrophilic inflammation/and or
27 release of mitogenic neuropeptides, depressed cell proliferative response, and cumulative distal
28 airway lesion.
29 Following the initial response of a rapid, shallow breathing pattern, animals eventually
30 adapt with continued episodic exposure despite the continued presence of epithelial damage,
31 altered structural development, and inflammation of the airways. Chen et al. (2003) used a
December 2005 5-45 DRAFT-DO NOT QUOTE OR CITE
-------
1 subset of the monkeys from the Schlegele et al. (2003a) study to demonstrate that attenuation of
2 O3-induced rapid shallow breathing and lung function changes typically seen with repeated O3
3 exposure may be caused by the adaptation of the respiratory motor responses. This episodic O3
4 exposure appeared to create neuroplasticity of the nucleus tractus solitarius (NTS; a region of the
5 brainstem which controls respiration), including increased nonspecific excitability of the NTS
6 neurons, an increased input resistance, and an increased spiking response to intracellular
7 injections of depolarizing current.
8
9 5.2.5.4 Summary and Conclusions - Effects on Airway Responsiveness
10 Ozone-induced AHR has been reported in a number of laboratory species at an exposure
11 range of 0.5 to 1.0 ppm and in human asthmatics at ambient levels. In asthmatics, O3 is thought
12 to exacerbate AHR to specific allergens by nonspecifically increasing AHR. New studies have
13 demonstrated that AHR in asthmatics is due in part to chronic inflammation and airway
14 remodeling. Animal studies have shown that O3 exposure can augment OVA-induced AHR.
15 Importantly, there is a temporal relationship between inflammatory cell influx and O3-induced
16 AHR, but inflammation is not a prerequisite of AHR. Repeated O3 exposures enhance AHR,
17 possibly by modulating rapidly adapting airway receptors or by altering the structure of
18 conducting airways.
19 Currently reported investigations on AHR with repeated O3 exposure to nonsensitized
20 laboratory animals have shown equivocal results, especially at the most relevant ambient O3
21 concentrations of < 0.3 ppm. The few available studies in sensitized laboratory animals are
22 consistent with the O3-induced exacerbation AHR reported in atopic humans with asthma (see
23 Chapter 6) but the results are difficult to extrapolate because of interindividual and interspecies
24 differences in responsiveness to bronchoprovocation and possible adaptation of airway
25 responsiveness with long-term, repeated O3 exposures. Therefore, further studies in laboratory
26 animals are needed to investigate responses to the different challenges in relation to
27 measurements of airway inflammation and the other physiological and structural factors known
28 to contribute to airway responsiveness in human subjects.
29 Important new information indicates that rapid shallow breathing in response to O3 causes
30 a more evenly distributed injury pattern rather than protects from injury. New insights into the
31 mechanisms of O3-induced AHR suggest that: (1) exercise-induced bronchoconstriction may be
December 2005 5-46 DRAFT-DO NOT QUOTE OR CITE
-------
1 mediated by changes in tonicity of the bronchial smooth muscles; (2) vagally-mediated
2 mechanisms may affect trachela epithelial function and increase central airway reactivity;
3 (3) O3 may induce direct vascular constriction; (4) O3 may destroy neural endopeptidases in
4 airway epithelial cells, thus preventing the inactivation of SP; and (5) repeated O3 exposures may
5 diminish neutrophilic inflammation, depress cell proliferation, and cause cumulative distal
6 airway lesions.
7
8 5.2.6 Genotoxicity Potential of Ozone
9 There has been an historical interest in the ability of ground-level pollution to cause cancer,
10 especially lung cancer. This interest has been amplified in recent years by a report of increased
11 risks of incident lung cancer that were associated with elevated long-term ambient
12 concentrations of O3, PM10, and SO2 in nonsmoking California males (Beeson et al., 1998;
13 Abbey et al., 1999). The nationwide American Cancer Society study (Pope et al., 2002) showed
14 no significant effect of O3 on mortality risk, but a positive association of July-September O3
15 concentrations and cardiopulmonary mortality. Studies on children and young adults of
16 Southwest metropolitan Mexico City, repeatedly exposed to high levels of O3, PM, NOX,
17 aldehydes, metals, and other components in a complex ambient mixture, also report DNA
18 damage in blood leukocytes and nasal epithelial cells (Valverde et al., 1997; Calderon-
19 Garciduenas et al., 1999), and abnormal nasal biopsies (Calderon-Garciduenas et al., 2001). (See
20 Chapter 6 for a discussion of the human studies}
21 Many experimental studies have been conducted to explore the mutagenic and
22 carcinogenic potential of O3. In vitro studies are difficult to interpret due to the high exposure
23 levels and culture systems that allowed the potential formation of artifacts. More recently
24 published in vivo exposure studies found increased DNA strand breaks in respiratory cells from
25 guinea pigs (Ferng et al., 1997) and mice (Bornholdt et al., 2002) but, again, only on exposures
26 to high doses of O3 (1 ppm for 72 h and 1 or 2 ppm for 90 min, respectively). Exposing the A/J
27 mouse strain (known to have a high spontaneous incidence pulmonary adenomas) to 0.12, 0.50,
28 and 1.0 ppm O3 for 6 h/day, 5 days/week for up to 9 months, Witschi et al. (1999) did not find O3
29 exposure-related differences in lung tumor multiplicity or incidence.
30 Similarly, in a sub-chronic exposure study (B6C3FJ mice to 0.5 ppm O3 for 6 h/day,
31 5 days/week for 12 weeks) Kim et al. (2001) did not find statistically significant increases in the
December 2005 5-47 DRAFT-DO NOT QUOTE OR CITE
-------
1 incidence of lung tumors. Significant differences in mean body weight as well as mean absolute
2 and relative weights of several organs (e.g., liver, spleen, kidney, testes, and ovary) were
3 observed between O3-exposed and air-exposed mice. Histopathologic examination of major
4 organs revealed oviductal carcinomas in 3/10 O3-exposed female mice.
5
6 5.2.6.1 Summary and Conclusions - Genotoxicity Potential of Ozone
7 The weight of evidence from new experimental studies does not appear to support ambient
8 O3 as a pulmonary carcinogen in laboratory animal models. These new data are in agreement
9 with a definitive study of carcinogenicity of O3 from the NTP study (National Toxicology
10 Program, 1994; Boorman et al., 1994), which was negative in male and female rats, ambiguous
11 in male mice, and positive only in female mice at high concentrations of O3 (i.e., 1.0 ppm).
12 However, the new animal studies are not in agreement with epidemiologic studies discussed in
13 Chapter 7, which may suggest significant species differences in this health endpoint.
14
15
16 5.3 SYSTEMIC EFFECTS OF OZONE EXPOSURE
17 Ozone indirectly affects organs beyond the respiratory system due to O3 reaction products
18 entering the bloodstream and being transported to target sites. Extra-pulmonary effects could
19 also be due to the exposure-related production of mediators, metabolic products and cell
20 trafficking. Although systemic effects are of interest and indicate a very broad array of
21 O3 effects, they are of limited influence and difficult to interpret. By protecting from respiratory
22 tract effects, these systemic effects will likely be protected against also. Systemic effects are
23 only summarized briefly here and in Table AX5-8.
24
25 5.3.1 Neurobehavioral Effects
26 Animal behavior, both motor activity and operant behavior, has been shown to be
27 suppressed by acute 3 exposures of 0.12 ppm. There is a dose dependent decrease in activity
28 with increasing exposure levels. Additionally, these lowered activity levels tend to attenuate
29 with longer exposure periods. New studies in adult laboratory animals confirm that
30 environmentally- relevant O3 concentrations from 0.2 to 1.0 ppm can decrease motor activity and
31 affect short- and long-term memory, as tested by passive avoidance conditioning in rats (Rivas-
December 2005 5-48 DRAFT-DO NOT QUOTE OR CITE
-------
1 Arancibia et al., 1998; Avila-Costa et al., 1999; Dorado-Martinez et al., 2001), or water-maze
2 learning tasks in mice (Sorace et al., 2001). The effects have been attributed to reactive
3 oxygen/nitrogen species and/or ozonation products. The memory deficits could be blocked by
4 administration of vitamin E (Guerrero et al, 1999) or taurine (Rivas-Arancibia et al., 2000).
5 Increased freezing and decreased exploratory behaviors were accompanied by decreased
6 serotonin levels and increased levels of NO, glutamate, dopamine and striatal lipoperoxidation in
7 rats exposed to 1 ppm of O3 for 4 h (Rivas-Arancibia et al., 2003). The O3-exposed animals also
8 demonstrated neuronal cytoplasm and dendrite vacuolation and dilation of RER cisterns, which
9 the authors interpret as a neurodegenerative process resulting from the oxidative stress of acute
10 O3 exposure. Nino-Cabrera et al. (2002) demonstrated that a 0.7 ppm O3 exposure for 4 h can
11 induce ultrastructural alterations in the hippocampus and prefrontal cortex in aged rats. These
12 are areas of the brain where degenerative age-related changes in learning and memory functions
13 have been reported (Bimonte et al., 2003).
14 Paz (1997) reviewed a series of studies that demonstrated significant alterations of
15 electroencephalographic (EEG) patterns during sleep in animals acutely exposed to O3 (0.35 to
16 1.0 ppm). Rats and cats both showed loss of paradoxical sleep time after 2 to 8 h of O3 exposure
17 (Paz and Bazan-Perkins, 1992; Paz and Huitron-Resendiz, 1996). Increased total wakefulness,
18 alterations in circadian rhythm, and a permanent 50% loss of paradoxical sleep time were shown
19 in rat pups born to dams exposed to 1.0 ppm O3 during gestation (Haro and Paz, 1993). Effects
20 on sleep patterns were associated with alterations in brain neurotransmitter levels (Huitron-
21 Resendiz et al., 1994; Gonzalez-Pina and Paz, 1997) thought to be caused by O3 reaction
22 products or prostaglandins (Koyama and Hayaishi, 1994). The permanent effects in pups caused
23 by high O3 exposure during gestation were attributed to the diminished antioxidant capability of
24 fetal tissue (Gunther et al., 1993).
25
26 5.3.2 Neuroendocrine Effects
27 Early studies suggested an interaction of O3 with the pituitary-thyroid-adrenal axis because
28 thyroidectomy, hypophysectomy, and adrenalectomy protected against the lethal effects of high
29 concentrations of O3 Concentrations of 0.7 to 1.0 ppm O3 caused morphological changes in the
30 parathyroid; thymic atrophy; decreased serum levels of thyroid stimulating hormone,
31 triiodothyronine (T3), thyroxine (T4), free T4, and protein binding; and increased prolactin.
December 2005 5-49 DRAFT-DO NOT QUOTE OR CITE
-------
1 In more recent studies, increased toxicity to O3 was reported in hyperthyroid rats by Huffman
2 et al. (2001) and T3 supplementation was shown to increase metabolic rate and pulmonary injury
3 in the lungs of O3-treated animals (Sen et al., 1993).
4 The mechanisms by which O3 affects neuroendocrine function are not well understood.
5 Cottet-Emard et al. (1997) examined catecholamine activity in rat sympathetic efferents and
6 brain areas of prime importance to adaptation to environmental stressors. Exposures of 0.5 ppm
7 O3 for 5 days caused inhibition of norepinephrine turnover in heart (-48% of the control level)
8 but not in lungs and failed to modify the tyrosine hydroxylase activity in superior cervical
9 ganglia, and the catecholamine content in the adrenal glands. In the CNS, O3 inhibited tyrosine
10 hydroxylase activity in noradrenergic brainstem cell groups and decreased catecholamine
11 turnover was in the cortex (- 49%) and striatum (-18%) but not in the hypothalamus. This
12 suggests that high ambient levels of O3 can produce marked neural disturbances in structures
13 involved in the integration of chemosensory inputs, arousal, and motor control, effects that may
14 be responsible for some of the behavioral effects seen with O3 exposure.
15 High, non-ambient levels of O3 (e.g., > 1.0 ppm) affect visual and olfactory neural
16 pathways in the rat. For example, Custodio-Ramierez and Paz (1997) reported a significant
17 delay in visual evoked potentials recorded in the visual cortex and the lateral geniculate nucleus
18 of male Wistar rats acutely exposed to high levels of O3 (1.5, and 3.0 ppm for 4 h). Colin-
19 Barenque et al. (1999), using the same strain, reported cytological and ultrastructural changes in
20 the granule layer of the olfactory bulb after a 4-h exposure to 1 to 1.5 ppm O3. Although these
21 neural effects are thought to be caused by O3 reaction products, especially free radicals, the
22 studies do not add much to an understanding of the underlying mechanisms.
23
24 5.3.3 Cardiovascular Effects
25 Studies of the effects on hematological parameters and blood chemistry have shown that
26 erythrocytes are a target of O3. Exposures to 1.0 ppm O3 for 3 h have been found to decrease
27 HR, MAP, and core temperature and to induce arrhythmias in some exposures. These effects are
28 more pronounced in adult and awake rats than in younger or sleeping animals. Exposures of
29 0.2 ppm for 48 h have been shown to cause bradycardia, while exposures of 0.1 ppm for 24 have
30 been shown to cause bradyarrhythmia in rats only.
December 2005 5-50 DRAFT-DO NOT QUOTE OR CITE
-------
1 A more recent study of rats exposed to FA for 6 h, followed 2 days later by a 5-hr exposure
2 to 0.1 ppm O3, 5 days later by a 5-hr exposure to 0.3 ppm O3, and 10 days later by a 5-hr
3 exposure to 0.5 ppm O3 used the the head-out plethysmograph for continuous measurements
4 (Arito et al., 1997). Each of the O3 exposures was preceded by a 1-hr exposure to FA. Transient
5 rapid shallow breathing with slightly increased HR appeared 1-2 min after the start of O3
6 exposures and was attributed to an olfactory response. Persistent rapid shallow breathing with a
7 progressive decrease in HR occurred with a latent period of 1-2 hr. During the last 90-min of
8 exposure, averaged values for relative minute ventilation tended to decrease with the increase in
9 O3 concentration for young (4-6 mo) but not old (20-22 mo) rats.
10 New studies utilizing radiotelemetry transmitters in unanesthetized and unrestrained rats,
11 Watkinson et al. (1995; 2001) and Highfill and Watkinson (1996) demonstrated that when HR
12 was reduced during O3 exposure, the Tco and activity levels also decreased. The decreases in Tco
13 and blood pressure reported by in these studies and by Arito et al., (1997) suggest that the
14 changes in ventilation and HR are mediated through physiological and behavioral defense
15 mechanisms in an attempt to minimize the irritant effects of O3 inhalation. Decreased activity
16 was previously reported in laboratory animals during exposure to O3 (see above).
17 Similar cardiovascular and thermoregulatory responses in rats to O3 were reported by
18 Iwasaki et al. (1998). Repeated exposure to 0.1, 0.3, and 0.5 ppm O3 8 hrs/day for 4 consecutive
19 days caused disruption of circadian rhythms of HR and Tco on the first and second exposure days
20 that was concentration-dependent. The decreased HR and Tco recovered to control values on the
21 third and fourth days of O3 exposure.
22 The thermoregulatory response to O3 was further characterized by Watkinson et al. (2003).
23 Male Fischer-344 rats were exposed to 0.0 ppm* 24 h/day (air), 0.5 ppm x 6h/day (intermittent)
24 or 0.5 ppm x 23 h/day (continuous) at 3 temperatures, 10° C (cold), 22° C (room), or 34° C
25 (warm). Another protocol examined the effects of O3 exposure (0.5 ppm) and exercise described
26 as rest, moderate, heavy or CO2-stimulated ventilation. Both intermittent and continuous O3
27 exposure caused decreases in HR and Tco and increases in BALF inflammatory markers.
28 Exercise in FA caused increases in HR and Tco while exercise in O3 caused decreases in those
29 parameters. Carbon dioxide and O3 induced the greatest deficits in HR and Tco. Several factors
30 were suggested that may modulate the hypothermic response, including dose, animal mass, and
31 environmental stress).
December 2005 5-51 DRAFT-DO NOT QUOTE OR CITE
-------
1 Laboratory animals exposed to relatively high ambient O3 concentrations (> 0.5 ppm)
2 demonstrate tissue edema in the heart and lungs. This may be due to increased circulating levels
3 of atrial natriuretic factor (ANF), which is known to mediate capillary permeability,
4 vasodilation, and blood pressure (Daly et al., 2002). Increased levels of ANF were reported in
5 the heart, lungs, and circulation of rats exposed to 0.5 ppm O3 for 8 h (Vesely et al., 1994a,b,c)
6
7 5.3.4 Reproductive and Developmental Effects
8 Early studies of pre- and postnatal exposure to O3 were performed at relatively high
9 concentrations. Teratogenic effects were not observed with intermittent exposures of 0.44 to
10 1.97 ppm O3 during any part of gestation. Continuous exposure during mid-gestation increased
11 the resorption of embryos while exposures during late gestation delayed some behavioral
12 developments (e.g., righting, eye opening). There were no effects on neonatal mortality up to
13 1.5 ppm O3, whereas some transient effects on weight gain were observed at exposures of
14 0.6 ppm O3.
15 Recent studies tend to confirm previous conclusions that prenatal exposures to O3
16 concentrations < 1.0 ppm do not cause major or widespread somatic or neurobehavioral effects
17 in the offspring of laboratory animals. These studies generally add some weight toward a
18 negative interpretation of the importance of contributions of low, ambient O3 to lower birth
19 weights and gross development defects reported in neonates born to women exposed to typical
20 ambient pollution (e.g., Renner, 2002; Chen et al., 2002; Ritz and Yu, 1999). Some postnatal O3
21 exposure studies continue to find a few, subtle or borderline somatic and behavioral deficits that
22 will require further research to better assess potential risk to developing humans.
23 Recent studies of somatic and neurobehavioral development in female CD-I mice exposed
24 during pregnancy (days 7 to 17) to O3 (0, 0.4, 0.8, or 1.2 ppm) failed to show any O3 effects on
25 reproductive or behavioral performance (Bignami et al., 1994). The study did find significant
26 decreases in body weight gain and delayed eye opening in pups in the 1.2 ppm exposure group.
27 The lack of effect on behavioral performance contrasts with earlier findings, which may be due
28 to the use of different species, differing exposure durations, cross-fostering used in the latter
29 study different species and exposure durations during pregnancy. A second study using CD-I
30 mice exposed in utero from conception through day 17 of pregnancy to 0, 0.2, 0.4, and 0.6 ppm
31 O3 found no significant deficits in reproductive performance, postnatal somatic and
December 2005 5-52 DRAFT-DO NOT QUOTE OR CITE
-------
1 neurobehavioral development, or adult motor activity (Petruzzi et al., 1995). A third study by
2 the same group (Petruzzi et al., 1999), using O3 exposures (0.3, 0.6, or 0.9 ppm) which continued
3 postnatally until weaning, showed subtle changes in handedness and morphine reactivity.
4 Exposures to 0.6 ppm O3 caused a reduced preference for the right paw in adulthood. Exposures
5 to 0.9 ppm O3 altered hot plate avoidance after IP treatment with morphine in adulthood.
6 CD-I mice exposed to 0.6 ppm O3 from birth through weaning demonstrated no
7 impairment of navigational performance during acquisition and only subtle changes during
8 reversal (Dell'Omo et al., 1995a). Additionally, there were no O3-induced effects on
9 reproductive performance, but offspring showed a significant reduction in body weight. Effects
10 on neurobehavioral development with this exposure were minor, with some attenuation of
11 activity responses and impairment of passive avoidance acquisition (DeH'Omo et al. (1995b).
12 The offspring of CD-I mice continuously exposed from 30 days prior to the formation of
13 breeding pairs until PND 17 to 0.0, 0.3, or 0.6 ppm O3 showed only small and selective effects
14 on somatic and sensorimotor development (Sorace et al., 2001).
15 Morphological changes were found in the anterior cerebellar lobe of rat pups born to dams
16 exposed during the entire gestation period to very high (1.0 ppm) O3 concentrations for 12 h/day.
17 (Rivas-Manzano and Paz, 1999). Additionally, the dams displayed significantly fewer
18 implantations, increased rate of reabsorptions, a high incidence of spontaneous abortion, and
19 offspring with low birth weight, as noted by previous investigators.
20
21 5.3.5 Effects on the Liver, Spleen, and Thymus
22 Early investigations of the effects of O3 on liver centered on xenobiotic metabolism, and
23 the prolongation of sleeping time, which was observed at 0.1 ppm O3. In some species, only
24 adults and especially females were affected. In rats, high (1.0 to 2.0 ppm) acute O3 exposures
25 caused increased production of NO by hepatocytes and enhanced protein synthesis (Laskin et al.,
26 1994; 1996). The O3-associated effects shown in the liver are thought to be mediated by
27 inflammatory cytokines or other cytotoxic mediators released by activated macrophages in the
28 lungs (Vincent et al., 1996; Laskin et al., 1998; Laskin and Laskin, 2001). Except for the earlier
29 work on xenobiotic metabolism, the responses occurred only after very high acute O3 exposures.
30 Examinations of the effects of O3 on spleen and thymus have shown that O3 primarily
31 affects T-cell mediated systemic immunity. As with the O3-associated effects shown in the liver,
December 2005 5-53 DRAFT-DO NOT QUOTE OR CITE
-------
1 most of the statistically significant changes occurred after acute exposures to very high O3
2 concentrations and relate to systemic oxidative stress. Using more relevant ambient urban O3
3 exposure patterns, effects were not found on systemic immune function of rats.
4
5 5.3.6 Effects on Cutaneous and Ocular Tissues
6 Ozone exposure not only affects various organ systems, when inhaled, but also has direct
7 effects on the exposed skin and eyes. The outermost layer of the skin (stratum corneum; SC)
8 may be oxidized, which can lead to compromise of the skin barrier and an epidermal
9 proinflammatory response ( Weber et al., 2001; Cotovio et al., 2001; Thiele, 2001). These
10 effects are found only at very high concentrations (>l-5 ppm) and have not been shown at more
11 relevant ambient levels of exposure. The skin possesses a well-developed defense system
12 against oxidative stress, utilizing nonenzymatic (e.g., vitamin C and E, glutathione, uric acid,
13 a-tocopherol) and enzymatic (e.g., superoxide dismutase, catalase, glutathione reductase and
14 peroxidase) antioxidants (Cross et al., 1998). Ocular tissues have similar antioxidant protective
15 function as the skin but are not as well studied (Mucke, 1996; Rose et al., 1998). Effects of
16 ground-level smog on the eyes have been reported but generally are attributed to related
17 photochemical oxidants like peroxyacetyl nitrate (Vyskocil et al., 1998) or possibly to
18 atmospheric O3 precursors or reaction products like aldehydes.
19 Hairless mice (SKH-1) were used to evaluate the cutaneous effects of O3 (1, 5, and 10 ppm
20 for 2 h or 1 ppm for 2 h on six consecutive days) by Thiele et al. (1997a,b,c,d). In the upper
21 epidermis decreased antioxidant levels were observed at > 1.0 ppm O3, while in both upper and
22 lower epidermal layers increased malondialdehyde (MDA) was found with exposures > 5 ppm.
23 Dermal application of vitamin E prevented MDA accumulation.
24 The same mouse model exposed to O3 (0.8 to 10 ppm for 2 h) was used to demonstrate that
25 O3 depletes the low molecular weight antioxidants (e.g., a-tocopherol, vitamin C, glutathione,
26 uric acid) in the SC at > 1.0 ppm and causes increased MDA at > 5 ppm (Weber et al, 1999,
27 2000,2001). Valacchi et al. (2000) demonstrated that preexposure to O3 followed by low-dose
28 ultraviolet (UV) radiation (0.33 MED) caused depletion of a-tocopherol at an exposure level of
29 0.5 ppm O3. This suggests that combined low doses of UV radiation and near-ambient levels of
30 O3 may cause oxidative stress on the SC. Additional studies demonstrated that stress-inducible
31 proteins (e.g., heme oxygenase-1) and other heat shock proteins (e.g., HSP27 and HSP70) were
December 2005 5-54 DRAFT-DO NOT QUOTE OR CITE
-------
1 induced in deeper cellular layers of the epidermis following O3 exposure (8.0 ppm for 2 h)
2 (Valacchi et al., 2002). Prolonged exposure to lower concentrations of O3 (0.8 ppm) for 6 h also
3 induces cellular stress responses that included the formation of HNE protein adducts, HSP27,
4 and heme-oxygenase-1 in the deeper cellular layers of the skin that continued for up to 18 h after
5 O3 exposure, followed by repair processes (Valacchi et al., 2003).
6 The importance of O3 and UV-induced cellular protein oxidation found in murine skin
7 models to possibly similar environmentally-induced changes in human SC keratins was
8 identified by Thiele et al. (1998, 1999) and Thiele (2001). Using the presence of carbonyl
9 groups in proteins as a marker of reactive oxygen mediated protein oxidation, they reported
10 higher carbonyl levels in the upper SC from the tanned skin of humans and in the skin of healthy
11 human volunteers exposed to model chemical oxidants (e.g., hypochlorite, benzoyl peroxide)
12 that were inversely correlated with vitamin E levels. The environmentally-induced oxidative
13 damage identified in human SC represents an early pathophysiological stage in the development
14 of barrier disruption and inflammation, and possibly has implications for the process of
15 desquamation. The relevance of potentiation of environmental oxidative stress by O3 exposure
16 of human skin needs further study.
17
18 5.3.7 Summary and Conclusions - Systemic Effects of Ozone
19 Neurobehavioral effects of O3 at concentrations of 0.2 to 1.0 ppm include decreased motor
20 activity, short- and long-term memory deficits, increased freezing behavior, decreased
21 exploratory behaviors. These effects have been associated with reactive oxygen/nitrogen
22 species, ozonation products, altered neurotransmitter levels, morphological changes in several
23 brain regions, and altered EEG patterns during sleep. Neuroendocrine effects of O3 include
24 morphological and hormonal changes in the pituitary-thyroid-adrenal axis at concentrations of
25 -0.75 ppm and alterations of visual and olfactory neural pathways at concentrations > 1 ppm.
26 Mechanisms underlying these effects are not understood at this time. Cardiovascular effects of
27 O3 at concentrations of 0.3 to 0.5 ppm include decreased HR, Tco, and BP, which have been
28 termed a hypothermic response. Concentrations of O3 >0.5 ppm cause tissue edema (possibly
29 mediated by ANF). These data are in accordance with O3-associated cardiac defects found in
30 neonates and fetuses delivered in southern California during 1987 to 1993, suggesting that high
December 2005 5-55 DRAFT-DO NOT QUOTE OR CITE
-------
1 urban exposures of O3 and its related co-pollutants can adversely affect the cardiovascular
2 system.
3 Prenatal exposures to O3 concentrations < 1.0 ppm did not cause noticeable somatic or
4 neurobehavioral effects in offspring, while concentrations of 1.0 to 1.5 ppm caused varying
5 effects on neonatal mortality. Some studies have shown an effect of O3 on liver xenobiotic
6 enzymes at concentrations as low as 0.1 ppm, while other studies have shown no alterations in
7 metabolic enzymes at even 1 ppm, with the effects appearing to be highly-species specific.
8 Effects on spleen and thymus appear to only occur at high O3 concentrations (> 1.0 ppm), while
9 relevant ambient, urban exposures have no effect on systemic immune function in rats. Effects
10 of O3 on cutaneous and ocular tissue are only seen at high, non-relevant concentrations.
11
12
13 5.4 INTERACTIONS OF OZONE WITH OTHER CO-OCCURRING
14 POLLUTANTS
15 Animal studies of the effects of O3 in combination with other air pollutants show that
16 antagonism, addititivity, and synergism can result, depending on the animal species, exposure
17 regimen, and health endpoint. These studies are of three types, ambient air mixtures, laboratory -
18 generated complex mixtures, and binary mixtures. Tables AX5-9 and AX5-10 list binary studies
19 of coexposures to nitrogen dioxide and PM, respectively.
20
21 5.4.1 Ozone and Nitrogen Oxides
22 The most commonly studied copollutant in binary mixtures with O3 is NO2. Both early
23 studies and more recent ones indicate that, although interaction may occur between these two
24 pollutants, in general, O3 often masked the effects of the NO2 or accounted for most of the
25 response, due to the greater toxicity of O3. Very generally, additivity occurred after acute
26 exposure and synergism occurred with prolonged exposure. Interpreting the mixture studies is
27 challenging because laboratory exposure patterns rarely simulate real-world exposure patterns.
28 In the case of NO2 and O3, NO2 typically peaks before O3, with a mixture occurring between the
29 peaks, but most laboratory exposures used mixtures only. Also, most studies of O3 and NO2
30 mixtures used ambient levels of O3 and levels of NO2 high above ambient.
December 2005 5-56 DRAFT-DO NOT QUOTE OR CITE
-------
1 Recent work has shown that chronic exposures of rats to O3 (0.8 ppm) and NO2 (14.4 ppm)
2 for 6 h/day caused development of respiratory insufficiency and severe weight loss. Half of
3 these animals died after 55 to 78 days of exposure due to severe fibrosis (Farman et al., 1997).
4 Increased total lung collagen and elastin was observed, with loss of mature collagen, suggesting
5 breakdown and remodeling of the lung parenchyma. Morphological examination of these
6 coexposures demonstrate a sequence of events starting with increasing inflammatory and mild
7 fibrotic changes for the first 3 weeks of exposure and stabilized or even reduced changes after
8 4 to 6 weeks, and severe increases over 7 to 9 weeks of exposure (Farman et al., 1999). This
9 suggests a repair processes occurring during the middle 4 to 6 weeks of exposure become
10 overwhelmed, leading to progressive fibrosis after 7 to 8 weeks of exposure. When the
11 coexposure was extended for 90 days, lesions were shown to extended far into the acinus, but the
12 extent of tissue involvement was the same after 7, 78, and 90 days of exposure. At the end of
13 exposure, high levels of procollagen types I and III mRNA were observed within central acini in
14 the lungs from the combined exposure group but not in lungs from the rats exposed to O3 or NO2
15 alone.
16 Sprague-Dawley rats exposed to 0.3 ppm O3 and the combined exposure of O3 and 1.2 ppm
17 NO2 for 3 d demonstrated significant DNA single-strand breaks in AMs ( Bermudez et al., 1999).
18 No changes were caused by NO2-only exposure. The same exposures stimulated the activity of
19 polyADPR synthetase, suggesting a response to lung cellular DNA repair caused by oxidant-
20 induced lung injury (Bermudez, 2001). The laboratory animal model of progressive pulmonary
21 fibrosis, utilizing long-term, combined O3 (0.4 to 0.8 ppm) and high-level NO2 (7 to 14 ppm)
22 exposure, causes an initial acute pulmonary inflammation, followed by adaptation and repair,
23 and eventually causing pulmonary fibrosis after 6 to 13 weeks of exposure (Ishii et al., 2000b;
24 Weller et al., 2000). Unfortunately, this model is not very useful for understanding potential
25 interactive effects of ambient concentrations of O3 and NO2.
26
27 5.4.2 Ozone and Other Copollutants
28 Ozone and Formaldehyde
29 Early studies with combined exposures to O3 and formaldehyde (HCHO) found evidence of
30 both synergistic and non-interactive effects. New work includes studies of biochemical and
31 histopathological endpoints in rats exposed to 0.4 ppm O3 and 3.6 ppm HCHO, alone and
December 2005 5-57 DRAFT-DO NOT QUOTE OR CITE
-------
1 combined, for 8 h/day for 3 days (Cassee and Feron, 1994). They demonstrated no interactive
2 effects in the nasal respiratory epithelium, despite the high levels of HCHO when compared to
3 typical ambient levels of 1 to 10 ppb (e.g., Rehle et al., 2001). Mautz (2003) studied changes in
4 breathing pattern and epithelial cell proliferation using exposures of 0.6 ppm O3 and 10 ppm
5 HCHO alone and in combination for 3 h with exercise at two times resting ventilation. Even
6 with exercise, HCHO does not substantially penetrate to the lower respiratory tract to interact
7 with O3, and does not alter breathing patterns to modify local O3 dose. Parenchymal injury was,
8 therefore, due to O3 alone. In the nasal transitional epithelium and in the trachea, however,
9 combined exposure produced additive effects due to the increased volume of toxicants during
10 exercise. No other combined pollutant studies have been published in the peer-reviewed
11 literature, although two studies compared the respiratory effects of O3 to HCHO. Nielson et al.,
12 (1999) compared upper airway sensory irritation caused by HCHO concentrations up to 4 ppm to
13 the lower airway irritation caused by O3. Using BALB/c mice, they continuously measured fB,
14 Vt, expiratory flow, T;, Te, and respiratory patterns during acute, 30-min exposures. The NOEL
15 for HCHO was 0.3 ppm, compared to 1.0 ppm for O3.
16 Thus, O3 and HCHO do not appear to have additive effects, except during exercise, and
17 that is due to increased volume of gas reaching the tissue. Any possible synergism occurs in the
18 nasal epithelium. HCHO exerts its effects primarily in the upper respiratory tract, whereas the
19 primary site of acute cell injury from O3 occurs in the conducting airways. EPA is currently
20 completing a toxicological and epidemiological review and risk characterization for
21 formaldehyde.
22
23 Ozone and Tobacco Smoke
24 Early studies of combined exposures of O3 (Ippm) and tobacco smoke demonstrated
25 altered airway responsiveness to inhaled bronchoconstrictor challenge and tracheal vascular
26 permeability in guinea pigs. A more recent study (Wu et al., 1997) reported that inhalation of
27 cigarette smoke evokes a transient bronchoconstrictive effect in anesthetized guinea pigs. Total
28 pulmonary resistance (RL) and dynamic lung compliance (Cdyn) were compared before and after
29 acute exposure to 1.5 ppm O3 for 1 h. Cigarette smoke alone (7 ml) at a low concentration (33%)
30 induced a mild and reproducible bronchoconstriction that slowly developed and reached its peak
31 after a delay of > 1 min. After O3, the same cigarette smoke inhalation challenge evoked an
December 2005 5-58 DRAFT-DO NOT QUOTE OR CITE
-------
1 intense bronchoconstriction that occurred more rapidly, reaching its peak within 20 s, and was
2 sustained for > 2 min. Pretreatment with selective antagonists of neurokinin type 1 and 2
3 receptors completely blocked the enhanced airway responsiveness suggesting that O3 exposure
4 induced AHR to inhaled cigarette smoke, which resulted primarily from the bronchoconstrictive
5 effect of endogenous tachykinins.
6 The above studies were conducted with undiluted tobacco smoke and high O3
7 concentrations. To determine the effects of aged and diluted sidestream cigarette smoke (ADSS)
8 as a surrogate of environmental tobacco smoke (ETS) on O3-induced lung injury, Yu et al.
9 (2002) exposed male B6C3F1 mice to (1) FA, (2) ADSS, (3) O3, or (4) ADSS followed by O3
10 (ADSS/O3). Exposure to 30 mg/m3 ADSS, 6 h/day for 3 days, followed by exposure to 0.5 ppm
11 O3 for 24 h was associated with a significant increase in the number of cells recovered by B AL
12 compared with exposure to ADSS alone or O3 alone. Neutrophils, lymphocytes, and total
13 protein levels in BAL were increased following the combined exposure when compared with all
14 other groups. Within the CAR, the percentage of proliferating cells was unchanged from control
15 following exposure to ADSS alone but was significantly elevated following exposure to O3 and
16 further augmented in a statistically significant manner in mice exposed to ADSS/O3. Following
17 exposure to O3 alone or ADSS/O3, the ability of AMs to release IL-6 under LPS stimulation was
18 significantly decreased, while exposure to ADSS alone or ADSS/O3 caused a significantly
19 increased release of TNFa from AMs under LPS stimulation. These data suggest that ADSS
20 exposure enhances the sensitivity of animals to O3-induced lung injury.
21 Acute exposure to ETS also may make a healthy person more susceptible to sequential O3
22 exposure by affecting lung barrier function or the underlying epithelium. Toxicological studies
23 with components of ETS (e.g., nicotine receptor agonists, acrolein, and oxidants) have shown
24 that the vagal bronchopulmonary C-fibers are stimulated by acute exposures that initiate both
25 central and local responses (Bonham et al., 2001; Mutoh et al., 2000). The central responses
26 (e.g., tachypnea, cough, bronchoconstriction, increased mucous secretion) are more protective of
27 the lungs; however, local responses may include increased sensitization of the C-fibers to other
28 irritants, including O3. Active tobacco smokers should not be similarly affected because they
29 already have significant chronic airway inflammation and increased mucus production. In fact,
30 chronic smokers appear to have diminished lung function responses to O3 (see Chapter 6).
31
December 2005 5-59 DRAFT-DO NOT QUOTE OR CITE
-------
1 5.4.3 Complex (Multicomponent) Mixtures Containing Ozone
2 Studies using complex (multicomponent) mixtures containing O3 are difficult to interpret
3 because of chemical interactions between the components, as well as the resultant production of
4 variable amounts of numerous secondary reaction products, and a lack of precise control over the
5 ultimate composition of the exposure environment. Irradiated automobile exhaust mixtures
6 containing total oxdant concentrations (expressed as O3) in the range of 0.2 to 1.0 ppm have been
7 shown to cause pulmonary function changes in several species.
8 A more recent attempt has been made to examine multicomponent mixtures resulting from
9 the reaction of O3 with unsaturated hydrocarbons [e.g., isoprene (C5H8) and terpene (C10H16)],
10 producing HCHO, formic acid, acetone, acrolein, acetic acid, and other oxidation products, many
11 of which are strong airway irritants. Wilkins et al. (2001) evaluated sensory irritation by
12 measuring mean fB in the mouse bioassay and found a 50% reduction after 30 min of exposure to
13 reaction products of O3 and isoprene. The mixture at this time period contained < 0.2 ppm O3, so
14 the authors attributed the observed effects to the oxidation products. Clausen et al. (2001), using
15 the same mouse model, evaluated the reaction products of O3 and limonene. A 33% reduction in
16 mean fB was produced after 30 min of exposure to the mixture containing < 0.3 ppm O3, again
17 implicating the effects of strong irritant products. Further work needs to be done with these
18 complex reaction mixtures because of their potential impact on the respiratory tract. The results
19 would be particularly important, however, to the reaction of O3 indoors (see Chapter 3).
20 Pollutant mixtures containing acid aerosols comprise another type of commonly examined
21 exposure atmosphere (studies summarized in Table AX5-10). Earlier studies that employed
22 simultaneous single, repeated, or continuous exposures of various animal species to mixtures of
23 acid sulfates and O3 found responses for several endpoints, including tracheobronchial
24 mucociliary clearance, alveolar clearance, pulmonary mechanics, and lung morphology, to be
25 due solely to O3 Some synergism was noted for bacterial infectivity, response to antigen, and
26 effects on lung protein content and the rate of collagen synthesis.
27 More recent studies found some differences in airway responses to inhaled acid particle-O3
28 mixtures that may have been partly due to airway dosimetry. Various physical and chemical
29 mechanisms may be responsible (see Schlesinger, 1995). For example, physical adsorption or
30 absorption of O3 or its reaction products on a particle could result in transport to more sensitive
31 sites, or to sites where O3, by itself, would not normally be reactive. Chemical reactions on the
December 2005 5-60 DRAFT-DO NOT QUOTE OR CITE
-------
1 surface of particles can form secondary products that are more lexicologically active, or
2 chemical characteristics of the particle may change the residence time or reactivity of oxidation
3 products at the site of deposition. The hypothesis that synergism between O3 and sulfates is due
4 to decreased pH changing the residence time or reactivity of reactants, such as free radicals, was
5 tested by Chen et al. (1995) and El-Fawal et al. (1995). Male New Zealand white rabbits were
6 exposed for 3 h to 125 |_ig/m3 H2SO4, 0.1, 0.3, or 0.6 ppm O3, and to combinations. Chen et al.
7 (1995) demonstrated that decreased pH following exposure to acid aerosol was correlated with
8 phagocytic activity and capacity of harvested macrophages and that exposure to O3/ H2SO4 the
9 removed this relationship. El-Fawal et al. (1995) showed that responsiveness of rabbit harvested
10 bronchial rings to ACh was increased following O3 exposure, but that O3/ H2SO4 combinations
11 resulted in antagonism.
12 Using rat tracheal explant cultures, Churg et al. (1996) demonstrated increased uptake of
13 asbestos or TiO2 in response to 10 min O3 (up to 1.0 ppm) pre-exposure suggesting that low
14 concentrations of O3 may increase the penetration of some types of PM into epithelial cells.
15 Using human epithelial cell cultures, Madden et al. (2000) demonstrated a greater potency for
16 ozonized diesel PM to induce prostaglandin E2 production. This suggests that O3 can modify the
17 biological activity of PM derived from diesel exhaust.
18 Effects of combined exposures of O3 and resuspended urban particles on cell proliferation
19 in epithelial cells of the terminal bronchioles and the alveolar ducts were examined by Vincent
20 et al. (1997) and Adamson et al. (1999). Rats exposed to 0.8 ppm O3 in combination with 5 or
21 50 mg/m3 particles for 4 h demonstrated greatly potentiated proliferative effects compared to O3
22 exposure alone. These findings using resuspended dusts, although at high concentrations, are
23 consistent with the studies demonstrating interaction between H2SO4 aerosols and O3. Effects of
24 acute coexposure to O3 and fine or ultrafine H2SO4 aerosols on lung morphology were examined
25 by Kimmel et al. (1997). They demonstrated that alveolar septal volume was increased in
26 animals co-exposed to O3 and ultrafine, but not fine, H2SO4. Interestingly, cell proliferation was
27 increased only in animals co-exposed to fine H2SO4 and O3, as compared to animals exposed to
28 O3 alone. Subchronic exposure to acid aerosols (20 to 150 |ig/m3 H2SO4) had no interactive
29 effect on the biochemical and morphometric changes produced by either intermittent or
30 continuous exposure to 0.12 to 0.2 ppm O3, which suggests that the interactive effects of O3 and
31 acid aerosol coexposure in the lung disappeared during the long-term exposure (Last and
December 2005 5-61 DRAFT-DO NOT QUOTE OR CITE
-------
1 Pinkerton, 1997). Sindhu et al. (1998) observed an increase in rat lung putrescine levels after
2 repeated, combined exposures to O3 and a nitric acid vapor.
3 Other studies have examined interactions between carbon particles and O3. The
4 interactions of intratracheally instilled carbon particles by followed by either a 7-day or 60-day
5 exposure to 0.5 ppm O3 in rats was evaulated by Creutzenberg et al. (1995). The carbon
6 particles caused diminished phagocytotic capacity and chemotactic migration capability of AMs
7 that was stimulated by the subsequent O3 exposure. Inflammatory responses following
8 coexposure of O3 plus fine, H2SO4-coated, carbon particles (MMAD = 0.26 jim) for 1 or 5 days
9 was examined by Kleinman et al. (1999). The response with the O3-particle mixture was greater
10 after 5 days (4 h/day) than after Day 1. This contrasted with O3 exposure alone (0.4 ppm), which
11 caused marked inflammation on acute exposure, but no inflammation after 5 consecutive days of
12 exposure.
13 The effects of a mixture of elemental carbon particles, O3, and ammonium bisulfate on rat
14 lung collagen content and macrophage activity was examined by Kleinman et al. (2000).
15 Decreases in lung collagen, and increases in macrophage respiratory burst and phagocytosis
16 were observed relative to other pollutant combinations. Mautz et al. (2001) used a similar
17 mixture (i.e., elemental carbon particles, O3, ammonium bisulfate, but with NO2 also) and
18 exposure regimen as Kleinman et al. (2000). Also observed were were decreases in pulmonary
19 macrophage Fc-receptor binding and phagocytosis and increases in acid phosphatase staining.
20 Bronchoalveolar epithelial permeability and cell proliferation were increased. Altered breathing-
21 patterns also were observed, with some adaptations occurring.
22 Bolarin et al. (1997) exposed rats to 50 or 100 |ig/m3 carbon particles in combination with
23 ammonium bisulfate and O3. Despite 4 weeks of exposure, they observed no changes in protein
24 concentration in lavage fluid or blood prolyl 4-hydroxylase, an enzyme involved in collagen
25 metabolism. Slight decreases in plasma fibronectin were present in animals exposed to the
26 combined pollutants versus O3 alone. Thus, the potential for adverse effects in the lungs of
27 animals challenged with a combined exposure to particles and gaseous pollutants is dependent
28 on numerous factors, including the gaseous co-pollutant, concentration, and time.
29 In a complex series of studies, Oberdorster and colleagues examined the interaction of
30 several pulmonary oxidative stress pollutants. Elder et al. (2000a,b) reported the results of
31 combined exposure to ultrafine carbon particles (100 |ig/m3) and O3 (1 ppm) in young and old
December 2005 5-62 DRAFT-DO NOT QUOTE OR CITE
-------
1 Fischer 344 rats that were pretreated with aerosolized endotoxin. In old rats, exposure to carbon
2 and O3 produced an interaction that resulted in a greater influx in neutrophils than that produced
3 by either agent alone. This interaction was not seen in young rats. Oxidant release from lavage
4 fluid cells also was assessed and the combination of endotoxin, carbon particles, and O3
5 produced an increase in oxidant release in old rats. This mixture produced the opposite response
6 in the cells recovered from the lungs of the young rats, indicating that the lungs of the aged
7 animals underwent greater oxidative stress in response to a complex pollutant mix of particles,
8 O3, and a biogenic agent. Johnston et al. (2000; 2002) reported the results of combined exposure
9 to O3 (1.0 and 2.5 ppm for 4, 20, or 24 h) and low-dose endotoxin, or to O3 and endotoxin
10 separately, in newborn and adult C57BL/6J mice. In the first study, adult (8 wk old) mice
11 showed greater sensitivity to O3 than newborn (36 h old) mice on the basis of mRNAs encoding
12 for various chemokines and cytokines. In contrast, adult and newborn mice responded similarly
13 2 h after endotoxin exposure (10 ng for 10 min), suggesting that age differences in O3-generated
14 inflammation is secondary to epithelial cell injury. In the second study, 8 wk old mice exposed
15 to O3 (1 ppm for 24 h) followed by endotoxin (37.5 EU for 10 min) showed increased
16 responsiveness over either exposure alone, on the basis of increased expression of chemokine
17 and cytokine messages and increased BAL fluid levels of protein and PMNs.
18 Fanucchi et al. (1998) and Wagner et al. (2001a,b) examined the synergistic effect of
19 coexposure to O3 and endotoxin on the nasal transitional epithelium of rats that also was
20 mediated, in part, by neutrophils. Fisher 344 rats intranasally instilled with endotoxin and
21 exposed to 0.5 ppm O3, 8 h per day, for 3 days developed mucous cell metaplasia in the nasal
22 transitional epithelium, an area normally devoid of mucous cells; whereas, intratracheal
23 instillation of endotoxin (20 jig) caused mucous cell metaplasia rapidly in the respiratory
24 epithelium of the conducting airways. A synergistic increase of intraepithelial mucosubstances
25 and morphological evidence of mucous cell metaplasia were found in rat maxilloturbinates upon
26 exposure to both O3 and endotoxin, compared to each pollutant alone. A similar response was
27 reported in OVA-sensitized Brown Norway rats exposed to 0.5 ppm O3, 8 h/day for 3 days
28 (Wagner et al., 2002), indicating that coexposure to O3 and inflammatory biogenic substances
29 like allergens (e.g., OVA) or bacterial endotoxin can augment epithelial and inflammatory
30 responses in rat nasal passages.
December 2005 5-63 DRAFT-DO NOT QUOTE OR CITE
-------
1 In follow-up studies, Wagner et al. (2003) reported that coexposure of rats to O3 and
2 endotoxin also enhanced epithelial and neutrophilic inflammatory responses in the pulmonary
3 airways. Fisher 344 rats were intranasally instilled with endotoxin and exposed to 1.0 ppm O3
4 for 8 h, which was repeated 24 h later. Three days after the last exposure, BALF was analyzed
5 for inflammatory cells and secreted mucosubstances (mucin SAC), and lung tissue was processed
6 for morphometric analysis. Endotoxin instillation alone caused a dose-dependent increase in
7 BALF neutrophils that was further increased 2-fold in O3-exposed rats given 20 [ig endotoxin,
8 the highest dose. Mucin glycoprotein SAC also was increased in the BALF at this dose but not
9 at lower endotoxin doses. Ozone exposure alone did not cause mucus hypersecretion, but it did
10 potentiate mucus secretion in rats given both 2 and 20 [ig endotoxin and increased intraepithelial
11 mucosub stances 2-fold, which was further substantiated by significant increases in mucin gene
12 (rMucSAC) mRNA levels in the conducting airways.
13 The effect of O3 modifying the biological potency of PM (diesel PM and carbon black) was
14 examined by Madden et al. (2000) in rats. Reaction of NIST Standard Reference Material
15 # 2975 diesel PM with 0.1 ppm O3 for 48 hr increased the potency (compared to unexposed or
16 air-exposed diesel PM) to induce neutrophil influx, total protein, and LDH in lung lavage fluid in
17 response to intratracheal instillation. Exposure of the diesel PM to high, non-ambient O3
18 concentration (1.0 ppm) attenuated the increased potency, suggesting destruction of the bioactive
19 reaction products. Unlike the diesel particles, carbon black particles exposed to 0.1 ppm O3 did
20 not exhibit an increase in biological potency, which suggested that the reaction of organic
21 components of the diesel PM with O3 were responsible for the increased potency.
22 Ulrich et al. (2002) investigated the effect of ambient PM from Ottawa Canada (EHC-93)
23 on O3-induced inflammation. Male Wistar rats were exposed to 0.8 ppm O3 for 8 h and allowed
24 to recover before intratracheal instillation of 0.5, 1.5, and 5 mg of EHC-93 in 0.3 ml of saline.
25 The high concentrations of PM used were sufficient to induce pulmonary inflammation, which
26 was not exacerbated by pre-exposure to O3. Rats from the combined exposure group did have
27 higher and more persistent lung lavage protein and albumin levels, as well as increased plasma
28 fibrinogen levels when compared to PM exposure alone.
29 The interaction of PM and O3 was further examined in a murine model of OVA-induced
30 asthma. Kobzik et al. (2001) investigated whether coexposure to inhaled, concentrated ambient
31 particles (CAPs) from Boston, MA and to O3 could exacerbate asthma-like symptoms. On days
December 2005 5-64 DRAFT-DO NOT QUOTE OR CITE
-------
1 7 and 14 of life, half of the BALB/c mice used in this study were sensitized by intraperitoneal
2 (ip) injection of OVA and then exposed to OVA aerosol on three successive days to create the
3 asthma phenotype. The other half received the ip OVA but were exposed to a phosphate-
4 buffered saline aerosol (controls). The mice were further subdivided (n > 6I/group) and exposed
5 for 5 h to CAPs, ranging from 63 to 1,569 |ig/m3, 0.3 ppm O3, CAPs + O3, or to FA. Pulmonary
6 resistance and airway responsiveness to an aerosolized MCh challenge were measured after
7 exposures. A small, statistically significant increase in pulmonary resistance and airway
8 responsiveness, respectively, was found in both normal and asthmatic mice immediately after
9 exposure to CAPs alone and to CAPs + O3 but not to O3 alone or to FA. By 24 h after exposure,
10 the responses returned to baseline levels. There were no significant increases in airway
11 inflammation after any of the pollutant exposures. In this well-designed study of a small-animal
12 model of asthma, O3 and CAPs did not appear to be synergistic. In further analysis of the data
13 using specific elemental groupings of the CAPs, the acutely increased pulmonary resistance was
14 found to be associated with the AISi fraction of PM. Thus, some components of concentrated
15 PM2 5 may affect airway caliber in sensitized animals, but the results are difficult to extrapolate
16 to p eopl e with asthma.
17 Animal studies have examined the adverse cardiopulmonary effects of complex mixtures in
18 urban and rural environments of Italy (Gulisano et al., 1997), Spain (Lorz and Lopez, 1997), and
19 Mexico (Vanda et al., 1998; Moss et al., 2001). Some of these studies have taken advantage of
20 the differences in pollutant mixtures of urban and rural environments to report primarily
21 morphological changes in the nasopharynx and lower respiratory tract (Gulisano et al., 1997;
22 Lorz and Lopez, 1997) of lambs and pigeons, respectively, after natural, continuous exposures to
23 ambient pollution. Each study has provided evidence that animals living in urban air pollutants
24 have greater pulmonary changes than would occur in a rural and presumably cleaner,
25 environment. The study by Moss et al. (2001) examined the nasal and lung tissue of rats
26 exposed (23 h/day) to Mexico City air for up to 7 weeks and compared them to controls similarly
27 exposed to FA. No inflammatory or epithelial lesions were found using quantitative
28 morphological techniques; however, the concentrations of pollutants were low. Extrapolation of
29 these results to humans is restricted, however, by uncontrolled exposure conditions, small
30 sample sizes, and other unknown exposure and nutritional factors in the studies in mammals and
31 birds, and the negative studies in rodents. They also bring up the issue of which species of
December 2005 5-65 DRAFT-DO NOT QUOTE OR CITE
-------
1 "sentinel" animals is more useful for predicting urban pollutant effects in humans. Thus, in these
2 field studies, it is difficult to assign a specific role to any specific component of the mixture for
3 the significant cardiopulmonary effects reported.
4 Similar morphological changes (Calderon-Garciduefias et al., 2000a; 2001) and chest X-ray
5 evidence of mild lung hyperinflation (Calderon-Garciduefias et al., 2000b) have been reported in
6 children residing in urban and rural areas of Mexico City. (See Chapter 7 for details of these
1 studies^) The ambient air in urban areas, particularly in Southwestern Mexico City, is a complex
8 mixture of particles and gases, including high concentrations of O3 and aldehydes that previously
9 have been shown to cause airway inflammation and epithelial lesions in humans (e.g., Calderon-
10 Garciduefias et al., 1992, 1994, 1996) and laboratory animals (Morgan et al., 1986; Heck et al.,
11 1990; Harkema et al., 1994, 1997a,b). The described effects demonstrate a persistent, ongoing
12 upper and lower airway inflammatory process and chest X-ray abnormalities in children residing
13 predominantly in highly polluted areas. Again, extrapolation of these results to urban
14 populations of the United States is difficult because of the unique complex of urban air in
15 Mexico City, uncontrolled exposure conditions, and other unknown exposure and nutritional
16 factors.
17
18 5.4.4 Summary and Conclusions - Interactions of Ozone with other
19 Co-occurring Pollutants
20 It is difficult to summarize the role that O3 plays in exposure responses to binary mixtures,
21 and even harder to determine its role in responses to multicomponent, complex atmospheres.
22 Though the specific mechanisms of action of the individual pollutants within a mixture may be
23 known, the exact bases for toxic interactions have not been elucidated clearly. Certain generic
24 mechanisms that may underlie pollutant interactions: (1) physical, involving adsorption of one
25 pollutant onto another and subsequent transport to more or less sensitive sites or to sites where
26 one of the components of the mixture normally would not deposit in concentrated amounts
27 (probably not involved in O3-related interactions; (2) production of secondary products that may
28 be more lexicologically active than the primary materials, demonstrated or suggested in a
29 number of studies as a basis for interaction between O3 and NO2 and between O3 and PM;
30 (3) biological or chemical alterations at target sites that affect response to O3 or the copollutant,
31 which which has been suggested to underlie interactions with mixtures of O3 and acid sulfates; 4)
December 2005 5-66 DRAFT-DO NOT QUOTE OR CITE
-------
1 O3- or copollutant-induced physiological change, such as alteration in ventilation pattern,
2 resulting in changes in the penetration or deposition of one pollutant when another is present.
3 This has been implicated in enhanced responses to various O3-containing mixtures with exercise.
4 Evaluation of interactions between O3 and copollutants is a complex procedure. Responses
5 are dependent on a number of host and environmental factors, such that different studies using
6 the same copollutants may show different types or magnitudes of interactions. The occurrence
7 and nature of any interaction is dependent on the endpoint being examined and is also highly
8 related to the specific conditions of each study, such as animal species, health status, exposure
9 method, dose, exposure sequence, and the physicochemical characteristics of the copollutants.
10 Because of this, it is difficult to compare studies, even those examining similar endpoints, that
11 were performed under different exposure conditions. Thus, any description of interactions is
12 really valid only for the specific conditions of the study in question and cannot be generalized to
13 all conditions of exposure to a particular chemical mixture. Furthermore, it is generally not
14 possible to extrapolate the effect of pollutant mixtures from studies on the effects of each
15 component when given separately. In any case, what can be concluded from the database is that
16 interactions of O3-containing mixtures are generally synergistic (antagonism has been noted in a
17 few studies), depending on the various factors noted above, and that O3 may produce more
18 significant biological responses as a component of a mixture than when inhaled alone.
19 Furthermore, although most studies have shown that interaction occurs only at higher than
20 ambient concentrations with acute exposure, some have demonstrated interaction at more
21 environmentally relevant levels (e.g., 0.05 to 0.1 ppm O3 with NO2) and with repeated exposures.
22
23
24 5.5 EFFECTS OF OTHER PHOTOCHEMICAL OXIDANTS
25 Peroxyacetyl nitrate (PAN) and peroxypropionyl nitrate (PPN) are the most abundant
26 non-O3 oxidants in ambient air of industrialized areas, other than the inorganic nitrogenous
27 oxidants such as NO2, and possibly HNO3. Ambient levels of PAN and PPN were reported to be
28 decreasing over the 1990's and available air quality data (Grosjean et al., 2001; Grosjean, 2003;
29 Jakobi and Fabian, 1997) indicate that present peak concentrations of PAN and PPN in ambient
30 air from urban areas are in the low ppb range (e.g., < 1 to 10 ppb). The levels found in nonurban
31 areas are considerably lower (Gaffney et al., 1993).
December 2005 5-67 DRAFT-DO NOT QUOTE OR CITE
-------
1 Reactions occur in the troposphere between O3 and hydrocarbons (e.g., d-limonene) to
2 produce epoxides, hydroperoxides, and peroxides. The majority of the measured ambient
3 hydroperoxides produced is hydrogen peroxide (H2O2), although a small amount of organic
4 hydroperoxides (ROOH) also may be formed. Friedlander and Yeh (1998) have estimated that
5 atmospheric aerosols can carry as high as 1 mM of H2O2 and organic hydroperoxides (e.g.,
6 hydroxymethylhydroperoxide) in the associated water. In vitro cell and tissue damage are
7 induced by high concentrations of liquid phase H2O2 (50 jiM to 1 mM). Morio et al. (2001)
8 demonstrated that 10 and 20 ppb of inhaled H2O2 vapor can penetrate the lower lung where it
9 causes inflammation. It is likely that hygroscopic components of PM transport ambient H2O2
10 into the lower lung and induce tissue injury as well. Exposure of rats to a H2O2-fine particle
11 mixture (215 or 429 |ig/m3 ammonium sulfate) resulted in increased neutrophil influx, and
12 production of inflammatory mediators by AMs (Morio et al., 2001). Hygroscopic secondary
13 organic aerosols generated by the O3/hydrocarbon reactions and their co-occurrence with H2O2
14 also provides another possible mechanism, yet to be validated, whereby H2O2 can be transported
15 into the lower respiratory tract (e.g., Friedlander and Yeh, 1998). Interaction of inhaled O3 with
16 unsaturated fatty acids on cell membranes and mucus in the airways generates epoxides,
17 hydroperoxides, and secondary ozonation products such as 4-hydroxynonenal (see Section 5.2.1)
18 Inhalation toxicological information on the effects of the non-O3 oxidants has been limited
19 to a few studies on PAN, but at concentrations much higher (approximately 100- to 1,000 fold)
20 than levels typically found in ambient air. Such high acute levels cause changes in lung
21 morphology, behavioral modifications, weight loss, and susceptibility to pulmonary infections.
22 Therefore, acute toxicity of PAN is much lower than O3, and it is unlikely that present ambient
23 PAN levels would affect pulmonary function responses to O3 (Vyskocil et al., 1998).
24 Cytogenetic studies indicate that PAN is not a potent mutagen, clastogen, or DNA damaging
25 agent in mammalian cells in vivo or in vitro at concentrations several orders of magnitude higher
26 than the generally encountered ambient air levels in most cities (Vyskocil et al., 1998; Kligerman
27 et al., 1995; Heddle et al., 1993). Some studies suggest that PAN may be a weak bacterial
28 mutagen at concentrations much higher than exist in present urban atmospheres (DeMarini et al.,
29 2000; Kleindienst et al., 1990).
30
December 2005 5-68 DRAFT-DO NOT QUOTE OR CITE
-------
1 5.5.1 Summary and Conclusions - Effects of Other Photochemical Oxidants
2 Concentrations of PAN and PPN (<1 to 10 ppb) in ambient air are unlikely to affect
3 pulmonary function or cause DNA damage. Levels of 10-20 ppm H2O2 can penetrate to the
4 lower lung directly or be transported there by PM, where inflammation can result; however,
5 ambient levels of are typically < ~5 ppb. As toxicology studies of other photochemical oxidants
6 are rare, quantitative scientific evaluations of possible health effects of environmental exposures
7 cannot be completed at this time.
December 2005 5-69 DRAFT-DO NOT QUOTE OR CITE
-------
1 REFERENCES
2 Adamson, I. Y. R.; Vincent, R.; Bjarnason, S. G. (1999) Cell injury and interstitial inflammation in rat lung after
3 inhalation of ozone and urban particulates. Am. J. Respir. Cell Mol. Biol. 20: 1067-1072.
4 Aizawa, H.; Koto, H.; Nakano, H.; Inoue, H.; Matsumoto, K.; Takata, S.; Shigyo, M.; Hara, N. (1997) The effects of
5 a specific tachykinin receptor antagonist FK-224 on ozone-induced airway hyperresponsiveness and
6 inflammation. Respirology 2: 261-265.
7 Aizawa, H.; Shigyo, M.; Nakano, H.; Matsumoto, K.; Inoue, H.; Hara, N. (1999a) Effect of the Chinese herbal
8 medicine, Bakumondo-to, on airway hyperresponsiveness induced by ozone exposure in guinea-pigs.
9 Respirology 4: 349-354.
10 Aizawa, H.; Shigyo, M.; Matsumoto, K.; Inoue, H.; Koto, H.; Hara, N. (1999b) PACAP reverses airway
11 hyperresponsiveness induced by ozone exposure in guinea pigs. Respiration 66: 538-542.
12 Alfaro, M. F.; Putney, L.; Tarkington, B. K.; Hatch, G. E.; Hyde, D. M.; Schelegle, E. S. (2004) Effect of rapid
13 shallow breathing on the distribution of 18O-labeled ozone reaction product in the respiratory tract of the rat.
14 Inhalation Toxicol.: in press.
15 Anderson, S. D. (1996) Challenge tests to assess airway hyperresponsiveness and efficacy of drugs used in the
16 treatment of asthma. J. Aerosol Med. 9: 95-109.
17 Anderson, S. D.; Daviskas, E. (2000) The mechanism of exercise-induced asthma is ... J. Allergy Clin. Immunol.
18 106:453-459.
19 Arito, H.; Takahashi, M.; Iwasaki, T.; Uchiyama, I. (1997) Age-related changes in ventilatory and heart rate
20 responses to acute ozone exposure in the conscious rat. Ind. Health 35: 78-86.
21 Arsalane, K.; Gosset, P.; Vanhee, D.; Voisin, C.; Hamid, Q.; Tonnel, A.-B.; Wallaert, B. (1995) Ozone stimulates
22 synthesis of inflammatory cytokines by alveolar macrophages in vitro. Am. J. Respir. Cell Mol. Biol.
23 13:60-68.
24 Avila-Costa, M. R.; Colin-Barenque, L.; Fortoul, T. L; Machado-Salas, J. P.; Espinosa-Villanueva, J.;
25 Rugerio-Vargas, C.; Rivas-Arancibia, S. (1999) Memory deterioration in an oxidative stress model and its
26 correlation with cytological changes on rat hippocampus CA1. Neurosci. Lett. 270: 107-109.
27 Avital, A.; Springer, C.; Bar-Yishay, E.; Godfrey, S. (1995a) Adenosine, methacholine, and exercise challenges in
28 children with asthma or paediatric chronic obstructive pulmonary disease. Thorax 50: 511-516.
29 Avital, A.; Picard, E.; Uwyyed, K.; Springer, C. (1995b) Comparison of adenosine 5'-monophosphate and
3 0 methacoline for the differentiation of asthma from chronic airway diseases with the use of the auscultative
31 method in very young children. J. Pediatr. 127: 438-440.
32 Bassett, D.; Elbon-Copp, C.; Otterbein, S.; Barraclough-Mitchell, H.; DeLorme, M.; Yang, H. (2001) Inflammatory
33 cell availability affects ozone-induced lung damage. J. Toxicol. Environ. Health A 64: 547-565.
34 Becker, S.; Quay, J.; Koren, H. S. (1991) Effect of ozone on immunoglobulin production by humanB cells in vitro.
35 J. Toxicol. Environ. Health 34: 353-366.
36 Beeson, W. L.; Abbey, D. E.; Knutsen, S. F. (1998) Long-term concentrations of ambient air pollutants and incident
37 lung cancer in California adults: results from the AHSMOG study. Environ. Health Perspect. 106: 813-823.
3 8 Bermudez, E. (2001) Detection of poly(ADP-ribose) synthetase activity in alveolar macrophages of rats exposed to
39 nitrogen dioxide and ozone. Inhalation Toxicol. 13: 69-84.
40 Bermudez, E.; Ferng, S. F.; Castro, C. E.; Mustafa, M. G. (1999) DNA strand breaks caused by exposure to ozone
41 and nitrogen dioxide. Environ. Res. 81: 72-80.
42 Bhalla, D. K. (1996) Alteration of alveolar macrophage chemotaxis, cell adhesion, and cell adhesion molecules
43 following ozone exposure of rats. J. Cell. Physiol. 169: 429-438.
44 Bhalla, D. K.; Gupta, S. K. (2000) Lung injury, inflammation, and inflammatory stimuli in rats exposed to ozone.
45 J. Toxicol. Environ. Health 59: 211-228.
46 Bhalla, D. K.; Gupta, S. K.; Reinhart, P. G. (1999) Alteration of epithelial integrity, alkaline phosphatase activity,
47 and fibronectin expression in lungs of rats exposed to ozone. J. Toxicol. Environ. Health A 56: 329-343.
48 Bhalla, D. K.; Reinhart, P. G.; Bai, C.; Gupta, S. K. (2002) Amelioration of ozone-induced lung injury by anti-tumor
49 necrosis factor-"alpha". Toxicol. Sci. 69: 400-408.
50 Bignami, G.; Musi, B.; DeH'Omo, G.; Laviola, G.; Alleva, E. (1994) Limited effects of ozone exposure during
51 pregnancy on physical and neurobehavioral development of CD-I mice. Toxicol. Appl. Pharmacol.
52 129:264-271.
53 Bimonte, H. A.; Nelson, M. E.; Granholm, A. C. (2003) Age-related deficits as working memory load increases:
54 relationships with growth factors. Neurobiol. Aging 24: 37-48.
December 2005 5-70 DRAFT-DO NOT QUOTE OR CITE
-------
1 Bolarin, D. M; Bhalla, D. K.; Kleinman, M. T. (1997) Effects of repeated exposures of geriatric rats to ozone and
2 particle-containing atmospheres: an analysis of bronchoalveolar lavage and plasma proteins. Inhalation
3 Toxicol. 9: 423-434.
4 Bonham, A. C.; Chen, C. Y.; Mutoh, T.; Joad, J. P. (2001) Lung C-fiber CNS reflex: role in the respiratory
5 consequences of extended environmental tobacco smoke exposure ini young guinea pigs. Environ. Health
6 Perspect. 109(suppl. 4): 573-578.
7 Boorman, G. A.; Hailey, R.; Grumbein, S.; Chou, B. J.; Herbert, R. A.; Goehl, T.; Mellick, P. W.; Roycroft, J. H.;
8 Haseman, J. K.; Sills, R. (1994) Toxicology and carcinogenesis studies of ozone and ozone
9 4-(A^-nitrosomethylamino)-l-(3-pyridyl)-l-butanone in Fischer-344/N rats. Toxicol. Pathol. 22: 545-554.
10 Bornholdt, J.; Dybdahl, M.; Vogel, U.; Hansen, M.; Loft, S.; Wallin, H. (2002) Inhalation of ozone induces DNA
11 strand breaks and inflammation in mice. Mutat. Res. 520: 63-71.
12 Brannan, J. D.; Koskela, H.; Anderson, S. D.; Chew, N. (1998) Responsiveness to Mannitol in asthmatic subjects
13 with exercise- and hyperventiliation-induced asthma. Am. J. Respir. Crit. Care Med. 158: 1120-1126.
14 Bridges, J. P.; Davis, H. W.; Demodarasamy, M.; Kuroki, Y.; Howies, G.; Hui, D. Y.; McCormack, F. X. (2000)
15 Pulmonary surfactant proteins A and D are potent endogenous inhibitors of lipid peroxidation and oxidative
16 cellular injury. J. Biol. Chem. 275: 38848-38855.
17 Broeckaert, F.; Clippe, A.; Wattiez, R.; Falmagne, P.; Bernard, A. (2003) Lung hyperpermeability, Clara-cell
18 secretory potein (CC16), and susceptibility to ozone of five inbred strains of mice. Inhalation Toxicol.
19 15: 1209-1230.
20 Calderon-Garciduenas, L.; Osorno-Velaquez, A.; Bravo-Alvarez, H.; Delgado-Chavez, R.; Barrios-Marquez, R.
21 (1992) Histopathologic changes of the nasal mucosa in Southwest metropolitan Mexico City inhabitants.
22 Am. J. Pathol. 140: 225-232.
23 Calderon-Garciduenas, L.; Rodriguez-Alcaraz, A.; Garcia, R.; Sanchez, G.; Barragan, G.; Camacho, R.; Ramirez,
24 L. (1994) Human nasal mucosal changes after exposure to urban pollution. Environ. Health Perspect.
25 102: 1074-1080.
26 Calderon-Garciduenas, L.; Osnaya-Brizuela, N.; Ramirez-Martinez, L.; Villarreal-Calderon, A. (1996) DNA strand
27 breaks in human nasal respiratory epithelium are induced upon exposure to urban pollution. Environ. Health
28 Perspect. 104: 160-168.
29 Calderon-Garciduenas, L.; Wen-Wang, L.; Zhang, Y.-J.; Rodriguez-Alcaraz, A.; Osnaya, N.; Villarreal-Calderon,
30 A.; Santella, R. M. (1999) 8-hydroxy-2'-deoxyguanosine, a major mutagenic oxidative DNA lesion, and DNA
31 strand breaks in nasal respiratory epithelium of children exposed to urban pollution. Environ. Health Perspect.
32 107:469-474.
33 Calderon-Garciduenas, L.; Devlin, R. B.; Miller, F. J. (2000a) Respiratory tract pathology and cytokine imbalance in
34 clinically healthy children chronically and sequentially exposed to air pollutants. Med. Hypotheses 55:
35 373-378.
36 Calderon-Garciduenas, L.; Mora-Tiscareno, A.; Chung, C. J.; Valencia, G.; Fordham, L. A.; Garcia, R.; Osnaya, N.;
37 Romero, L.; Acuna, H.; Villarreal-Calderon, A. (2000b) Exposure to air pollution is associated with lung
3 8 hyperinflation in healthy children and adolescents in southwest Mexico City: a pilot study. Inhalation Toxicol.
39 12:537-561.
40 Calderon-Garciduenas, L.; Gambling, T. M.; Acuna, H.; Garcia, R.; Osnaya, N.; Monroy, S.; Villarreal-Calderon,
41 A.; Carson, J.; Koren, H. S.; Devlin, R. B. (2001a) Canines as sentinel species for assessing chronic
42 exposures to air pollutants: part 2. cardiac pathology. Toxicol. Sci. 61: 356-367.
43 Calderon-Garciduenas, L.; Rodrigues-Alcaraz, A.; Valencia-Salazar, G.; Mora-Tascareno, A.; Garcia, R.;
44 Osnaya, N.; Villarreal-Calderon, A.; Devlin, R. B.; Van Dyke, T. L. (2001b) Nasal biopsies of children
45 exposed to air pollutants. Toxicol. Pathol. 29: 558-564.
46 Campos-Bedolla, P.; Vargas, M. H.; Montano, L. M. (2002) Effect of acute ozone exposure on pregnant rat uterus
47 contractile responses. Reprod. Toxicol. 16: 269-273.
48 Cassee, F. R.; Feron, V. J. (1994) Biochemical and histopathological changes in nasal epithelium of rats after 3-day
49 intermittent exposure to formaldehyde and ozone alone or in combination. Toxicol. Lett. 72: 257-268.
50 Chang, M. M.-J.; Wu, R.; Plopper, C. G.; Hyde, D. M. (1998) IL-8 is one of the major chemokines produced by
51 monkey airway epithelium after ozone-induced injury. Am. J. Physiol. 275: L524-L532.
52 Cheek, J. M.; McDonald, R. J.; Rapalyea, L.; Tarkington, B. K.; Hyde, D. M. (1995) Neutrophils enhance removal
53 of ozone-injured alveolar epithelial cells in vitro. Am. J. Physiol. 269: L527-L535.
54 Chen, L. C.; Qu, Q.; Amdur, M. O.; Schlesinger, R. B. (1995) Alteration of pulmonary macrophage intracellular pH
55 following inhalation exposure to sulfuric acid/ozone mixtures. Exp. Lung Res. 21: 113-128.
December 2005 5-71 DRAFT-DO NOT QUOTE OR CITE
-------
1 Chen, L.; Yang, W.; Jennison, B. L.; Goodrich, A.; Omaye, S. T. (2002) Air pollution and birth weight in northern
2 Nevada, 1991-1999. Inhalation Toxicol. 14: 141-157.
3 Chen, C.-Y.; Bonham, A. C.; Plopper, C. G.; Joad, J. P. (2003) Plasticity in respiratory motor control: selected
4 contribution: neuroplasticity in nucleus tractus solitarius neurons following episodic ozone exposure in infant
5 primates. J. Appl. Physiol. 94: 819-827.
6 Cho, H. Y.; Hotchkiss, J. A.; Harkema, J. R. (1999a) Inflammatory and epithelial responses during the development
7 of ozone-induced mucous cell metaplasia in the nasal epithelium of rats. Toxicol. Sci. 51: 135-145.
8 Cho, H. Y.; Hotchkiss, J. A.; Bennett, C. B.; Harkema, J. R. (1999b) Effects of pre-existing rhinitis on
9 ozone-induced mucous cell metaplasia in rat nasal epithelium. Toxicol. Appl. Pharmacol. 158: 92-102.
10 Cho, H. Y.; Hotchkiss, J. A.; Bennett, C. B.; Harkema, J. R. (2000) Neutrophil-dependent and
11 neutrophil-independent alterations in the nasal epithelium of ozone-exposed rats. Am. J. Respir. Crit. Care
12 Med. 162: 629-636.
13 Cho, H.-Y.; Zhang, L.-Y.; Kleeberger, S. R. (2001) Ozone-induced lung inflammation and hyperreactivity are
14 mediated via tumor necrosis factor-a receptors. Am. J. Physiol. 280: L537-L546.
15 Churg, A.; Brauer, M; Keeling, B. (1996) Ozone enhances the uptake of mineral particles by tracheobronchial
16 epithelial cells in organ culture. Am. J. Respir. Crit. Care Med. 153: 1230-1233.
17 Clausen, P. A.; Wilkins, C. K.; Wolkoff, P.; Nielsen, G. D. (2001) Chemical and biological evaluation of a reaction
18 mixture of R-(+)-limonene/ozone: formation of strong airway irritants. Environ. Int. 26: 511-522.
19 Cohen, M. D.; Zelikoff, J. T.; Qu, Q.; Schlesinger, R. B. (1996) Effects of ozone upon macrophage-interferon
20 interactions. Toxicology 114: 243-252.
21 Cohen, M. D.; Sisco, M.; Li, Y.; Zelikoff, J. T.; Schlesinger, R. B. (2001) Ozone-induced modulation of
22 cell-mediated immune responses in the lungs. Toxicol. Appl. Pharmacol. 171: 71-84.
23 Cohen, M. D.; Sisco, M.; Baker, K.; Li, Y.; Lawrence, D.; VanLoveren, H.; Zelikoff, J. T.; Schlesinger, R. B.
24 (2002) Effects of inhaled ozone on pulmonary immune cells critical to antibacterial responses in situ.
25 Inhalation Toxicol. 14: 599-619.
26 Colin-Barenque, L.; Avila-Costa, M. R.; Fortoul, T.; Rugerio-Vargas, C.; Machado-Salas, J. P.;
27 Espinosa-Villanueva, J.; Rivas-Arancibia, S. (1999) Morphologic alteration of the olfactory bulb after acute
28 ozone exposure in rats. Neurosci. Lett. 274: 1-4.
29 Connor, L. M.; Ballinger, C. A.; Albrecht, T. B.; Postlethwait, E. M. (2004) Interfacial phospholipids inhibit ozone
30 reactive absorption-mediated cytotoxicity in vitro. Am. J. Physiol.: 10.1152/ajplung.00397.2003.
31 Cotovio, J.; Onno, L.; Justine, P.; Lamure, S.; Catroux, P. (2001) Generation of oxidative stress in human cutaneous
32 models following in vitro ozone exposure. Toxicol. in Vitro 15: 357-362.
33 Cottet-Emard, J.-M.; Dalmaz, Y.; Pequignot, J.; Peyrin, L.; Pequignot, J.-M. (1997) Long-term exposure to ozone
34 alters peripheral and central catecholamine activity in rats. Pfluegers Arch. 433: 744-749.
35 Creutzenberg, O.; Bellmann, B.; Klingebiel, R.; Heinrich, U.; Muhle, H. (1995) Phagocytosis and chemotaxis of rat
36 alveolar macrophages after a combined or separate exposure to ozone and carbon black. Exp. Toxicol. Pathol.
37 47:202-206.
38 Cross, C. E.; Van Der Vliet, A.; Louie, S.; Thiele, J. J.; Halliwell, B. (1998) Oxidative stress and antioxidants at
39 biosurfaces: plants, skin, and respiratory tract surfaces. Environ. Health Perspect. 106(suppl. 5): 1241-1251.
40 Custodio-Ramirez, V.; Paz, C. (1997) Ozone produces functional deficits in the rat visual pathway.
41 Electroencephalogr. Clin. Neurophysiol. 104: 269-273.
42 Daly, C.; Fox, K.; Henein, M. (2002) Natriuretic peptides in the diagnosis of heart disease-first amongst equals?
43 Int. J. Cardiol. 84: 107-113.
44 Delacourt, C.; Benoist, M. R.; Waernessyckle, S. (2001) Relationship between bronchial responsiveness and clinical
45 evolution in infants who wheeze. A four-year prospective study. Am. J. Respir. Crit. Care Med.
46 164: 1382-1386.
47 Delaunois, A.; Segura, P.; Montano, L. M.; Vargas, M. H.; Ansay, M.; Gustin, P. (1998) Comparison of
48 ozone-induced effects on lung mechanics and hemodynamics in the rabbit. Toxicol. Appl. Pharmacol.
49 150:58-67.
50 DeH'Omo, G.; Fiore, M.; Petruzzi, S.; Alleva, E.; Bignami, G. (1995a) Neurobehavioral development of CD-I mice
51 after combined gestational and postnatal exposure to ozone. Arch. Toxicol. 69: 608-616.
52 DeH'Omo, G.; Wolfer, D.; Alleva, E.; Lipp, H.-P. (1995b) Developmental exposure to ozone induces subtle changes
53 in swimming navigation of adult mice. Toxicol. Lett. 81: 91-99.
54 DeLorme, M. P.; Yang, H.; Elbon-Copp, C.; Gao, X.; Barraclough-Mitchell, H.; Bassett, D. J. P. (2002)
55 Hyperresponsive airways correlate with lung tissue inflammatory cell changes in ozone-exposed rats.
56 J. Toxicol. Environ. Health Part A 65: 1453-1470.
December 2005 5-72 DRAFT-DO NOT QUOTE OR CITE
-------
1 DeMarini, D. M; Shelton, M. L.; Kohan, M. J.; Hudgens, E. E.; Kleindienst, T. E.; Ball, L. M.; Walsh, D.; de Boer,
2 J. G.; Lewis-Bevan, L.; Rabinowitz, J. R.; Claxton, L. D.; Lewtas, J. (2000) Mutagenicity in lung of Big
3 Blue(R) mice and induction of tandem-base substitutions in Salmonella by the air pollutant peroxyacetyl
4 nitrate (PAN): predicted formation of intrastrand cross-links. Mutat. Res. 457: 41-55.
5 Depuydt, P.; Joos, G. F.; Pauwels, R. A. (1999) Ambient ozone concentrations induce airway hyperresponsiveness
6 in some rat strains. Eur. Respir. J. 14: 125-131.
7 Dorado-Martinez, C.; Parades-Carbajal, C.; Mascher, D.; Borgonio-Perez, G.; Rivas-Arancibia, S. (2001) Effects of
8 different ozone doses on memory, motor activity and lipid peroxidation levels, in rats. Int. J. Neurosci.
9 108: 149-161.
10 Dormans, J. A. M. A.; Boere, A. J. F.; vanLoveren, H.; Rombout, P. J. A.; Marra, M.; vanBree, L. (1996)
11 Age-related toxicity in rat lungs following acute and repeated ozone exposure. Inhalation Toxicol. 8: 903-925.
12 Dormans, J. A. M. A.; VanBree, L.; Boere, A. J. F.; Marra, M.; Rombout, P. J. A. (1999) Interspecies differences in
13 time course of pulmonary toxicity following repeated exposure to ozone. Inhalation Toxicol. 11: 309-329.
14 Driscoll, K. E.; Simpson, L.; Carter, J.; Hassenbein, D.; Leikauf, G. D. (1993) Ozone inhalation stimulates
15 expression of a neutrophil chemotactic protein, macrophage inflammatory protein 2. Toxicol. Appl.
16 Pharmacol. 119: 306-309.
17 Dye, J. A.; Madden, M. C.; Richards, J. H.; Lehmann, J. R.; Devlin, R. B.; Costa, D. L. (1999) Ozone effects on
18 airway responsiveness, lung injury, and inflammation. Comparative rat strain and in vivo/in vitro
19 investigations. Inhalation Toxicol. 11: 1015-1040.
20 El-Fawal, H. A. N.; McGovern, T.; Schlesinger, R. B. (1995) Nonspecific bronchial responsiveness assessed in vitro
21 following acute inhalation exposure to ozone and ozone/sulfuric acid mixtures. Exp. Lung Res. 21: 129-139.
22 Elder, A. C. P.; Gelein, R.; Finkelstein J. N.; Cox, C.; Oberdorster, G. (2000a) Endotoxin priming affects the lung
23 response to ultrafine particles and ozone in young and old rats. In: Phalen, R. F., ed. Inhalation toxicology:
24 proceedings of the third colloquium on paniculate air pollution and human health (first special issue); June,
25 1999; Durham, NC. Inhalation Toxicol. 12(suppl. 1): 85-98.
26 Elder, A. C. P.; Gelein, R.; Finkelstein, J. N.; Cox, C.; Oberdorster, G. (2000b) Pulmonary inflammatory response to
27 inhaled ultrafine particles is modified by age, ozone exposure, and bacterial toxin. In: Grant, L. D., ed.
28 PM2000: paniculate matter and health. Inhalation Toxicol. 12(suppl. 4): 227-246.
29 Elsayed, N. M. (2001) Diet restriction modulates lung response and survivability of rats exposed to ozone.
30 Toxicology 159: 171-182.
31 Evans, M. J.; Johnson, L. V.; Stephens, R. J.; Freeman, G. (1976) Renewal of the terminal bronchiolar epithelium in
32 the rat following exposure to NO2 or O3. Lab. Invest. 35: 246-257.
33 Evans, M. J.; Fanucchi, M. V.; Baker, G. L.; Van Winkle, L. S.; Pantle, L. M.; Nishio, S. J.; Schelegle, E. S.;
34 Gershwhin, L. J.; Miller, L. A.; Hyde, D. M.; Sannes, P. L.; Plopper, C. G. (2003) Atypical development of
3 5 the trachea! basement membrane zone of infant rhesus monkeys exposed to ozone and allergen. Am. J.
36 Physiol. 285:L931-L939.
37 Fakhrzadeh, L.; Laskin, J. D.; Laskin, D. L. (2002) Deficiency in inducible nitric oxide synthase protects mice from
38 ozone-induced lung inflammation and tissue injury. Am. J. Respir. Cell Mol. Biol. 26: 413-419.
39 Fanucchi, M. V.; Hotchkiss, J. A.; Harkema, J. R. (1998) Endotoxin potentiates ozone-induced mucous cell
40 metaplasia in rat nasal epithelium. Toxicol. Appl. Pharmacol. 152: 1-9.
41 Fanucchi, M. V; Wong, V J.; Hinds, D.; Tarkington, B. K.; Van Winkle, L. S.; Evans, M. J.; Plopper, C. G. (2000)
42 Repeated episodes of exposure to ozone alters postnatal development of distal conducting airways in infant
43 rhesus monkeys. Am. J. Respir. Crit. Care Med. 161: A615.
44 Farman, C. A.; Pinkerton, K. E.; Rajini, P.; Witschi, H.; Last, J. A. (1997) Evolution of lung lesions in rats exposed
45 to mixtures of ozone and nitrogen dioxide. Inhalation Toxicol. 9: 647-677'.
46 Farman, C. A.; Watkins, K.; Van Hoozen, B.; Last, J. A.; Witschi, H.; Pinkerton, K. E. (1999) Centriacinar
47 remodeling and sustained procollagen gene expression after exposure to ozone and nitrogen dioxide. Am. J.
48 Respir. Cell Mol. Biol. 20: 303-311.
49 Ferng, S.-F.; Castro, C. E.; Afifi, A. A.; Bermudez, E.; Mustafa, M. G. (1997) Ozone-induced DNA strand breaks in
50 guinea pig tracheobronchial epithelial cells. J. Toxicol. Environ. Health 51: 353-367.
51 Folkerts, G.; Busse, W. W.; Nijkamp, F. P.; Sorkness, R.; Gern, J. E. (1998) Virus-induced airway
52 hyperresponsiveness and asthma. Am. J. Respir. Crit. Care Med. 157: 1708-1720.
53 Foster, W. M.; Freed, A. N. (1999) Regional clearance of solute from peripheral airway epithelia: recovery after
54 sublobar exposure to ozone. J. Appl. Physiol. 86: 641-646.
December 2005 5-73 DRAFT-DO NOT QUOTE OR CITE
-------
1 Frampton, M. W.; Pryor, W. A.; Cueto, R.; Cox, C.; Morrow, P. E.; Utell, M. J. (1999) Aldehydes (nonanal and
2 hexanal) in rat and human bronchoalveolar lavage fluid after ozone exposure. Cambridge, MA: Health Effects
3 Institute; research report no. 90. Available: www.healtheffects.org/Pubs/Frampton-C.pdf [2000, February 9].
4 Freed, A. N.; Chou, C. L.; Fuller, S. D.; Croxton, T. L. (1996) Ozone-induced vagal reflex modulates airways
5 reactivity in rabbits. Respir. Physiol. 105: 95-102.
6 Freed, A. N.; Cueto, R.; Pryor, W. A. (1999) Antioxidant transport modulates peripheral airway reactivity and
7 inflammation during ozone exposure. J. Appl. Physiol. 87: 1595-1603.
8 Friedlander, S. K.; Yeh, E. K. (1998) The submicron atmospheric aerosol as a carrier of reactive chemical species:
9 case of peroxides. Appl. Occup. Environ. Hyg. 13: 416-420.
10 Gaffney, J. S.; Marley, N. A.; Prestbo, E. W. (1993) Measurements of peroxyacetyl nitrate at a remote site in the
11 southwestern United States: tropospheric implications. Environ. Sci. Technol. 27: 1905-1910.
12 Garssen, J.; Van Bree, L.; Van Der Vliet, H.; Van Loveren, H. (1997) Ozone-induced impairment of pulmonary type
13 IV hypersensitivity and airway hyperresponsiveness in mice. Inhalation Toxicol. 9: 581-599.
14 Gohil, K.; Cross, C. E.; Last, J. A. (2003) Ozone-induced disruptions of lung transcriptomes. Biochem. Biophys.
15 Res. Commun. 305: 719-728.
16 Goldsmith, C.-A. W.; Ning, Y.-Y.; Qin, G.; Imrich, A.; Lawrence, J.; Murthy, G. G., K.; Catalano, P. J.; Kobzik, L.
17 (2002) Combined air pollution particle and ozone exposure increases airway responsiveness in mice.
18 Inhalation Toxicol. 14: 325-347.
19 Grosjean, D. (2003) Ambient PAN and PPN in southern California from 1960 to the SCOS97-NARSTO. Atmos.
20 Environ. 37(suppl. 2): S221-S238.
21 Grosjean, E.; Grosjean, D.; Woodhouse, L. F. (2001) Peroxyacetyl nitrate and peroxypropionyl nitrate during SCOS
22 97-NARSTO. Environ. Sci. Technol. 35: 4007-4014.
23 Guerrero, A. L.; Dorado-Martinez, C.; Rodriguez, A.; Pedroza-Rios, K.; Borgonio-Perez, G.; Rivas-Arancibia, S.
24 (1999) Effects of vitamin E on ozone-induced memory deficits and lipid peroxidation in rats. NeuroReport 10:
25 1689-1692.
26 Gulisano, M.; Marceddu, S.; Barbara, A.; Pacini, A.; Buiatti, E.; Martini, A.; Pacini, P. (1997) Damage to the
27 nasopharyngeal mucosa induced by current levels of urban air pollution: a field study in lambs. Eur. Respir. J.
28 10:567-572.
29 Gtinther, T.; Hollriegl, V.; Vormann, J. (1993) Perinatal development of iron and antioxidant defence systems.
30 J. Trace Elem. Electrolytes Health Dis. 7: 47-52.
31 Gupta, S. K.; Reinhart, P. G.; Bhalla, D. K. (1998) Enhancement of fibronectin expression in rat lung by ozone and
32 an inflammatory stimulus. Am. J. Physiol. 275: L330-L335.
33 Haddad, E.-B.; Liu, S. F.; Salmon, M.; Robichaud, A.; Barnes, P. J.; Chung, K. F. (1995) Expression of inducible
34 nitric oxide synthase mRNA in Brown Norway rats exposed to ozone: effect of dexamethasone. Eur. J.
35 Pharmacol. Environ. Toxicol. Pharmacol. Sect. 293: 287-290.
36 Haddad, E.-B.; Salmon, M.; Koto, H.; Barnes, P. J.; Adcock, I.; Chung, K. F. (1996) Ozone induction of
3 7 cytokine-induced neutrophil chemoattractant (CINC) and nuclear factor-kappa b in rat lung: inhibition by
38 corticosteroids. FEES Lett. 379: 265-268.
39 Hamilton, R. F.; Li, L.; Eschenbacher, W. L.; Szweda, L.; Holian, A. (1998) Potential involvement of
40 4-hydroxynonenal in the response of human lung cells to ozone. Am. J. Physiol. 274: L8-L16.
41 Harkema, J. R.; Morgan, K. T.; Gross, E. A.; Catalano, P. J.; Griffith, W. C. (1994) Consequences of prolonged
42 inhalation of ozone on F344/N rats: collaborative studies. Part VII: effects on the nasal mucociliary apparatus.
43 Cambridge, MA: Health Effects Institute; research report no. 65.
44 Harkema, J. R.; Catalano, P. J.; Hotchkiss, J. A. (1997a) Consequences of prolonged inhalation of ozone on F344/N
45 rats: collaborative studies. Part XII. Atrophy of bone in nasal turbinates. Cambridge, MA: Health Effects
46 Institute; research report no. 65.
47 Harkema, J. R.; Hotchkiss, J. A.; Griffith, W. C. (1997b) Mucous cell metaplasia in rat nasal epithelium after a
48 20-month exposure to ozone: a morphometric study of epithelial differentiation. Am. J. Respir. Cell Mol.
49 Biol. 16: 521-530.
50 Harkema, J. R.; Hotchkiss, J. A.; Barr, E. B.; Bennett, C. B.; Gallup, M.; Lee, J. K.; Basbaum, C. (1999)
51 Long-lasting effects of chronic ozone exposure on rat nasal epithelium. Am. J. Respir. Cell Mol. Biol.
52 20:517-529.
53 Kara, R.; Paz, C. (1993) Effects of ozone exposure during pregnancy on ontogeny of sleep in rats. Neurosci. Lett.
54 164:67-70.
55 Hawgood, S.; Poulain, F. R. (2001) The pulmonary collectins and surfactant metabolism. Annu. Rev. Physiol.
56 63:495-519.
December 2005 5-74 DRAFT-DO NOT QUOTE OR CITE
-------
1 Hawgood, S.; Ochs, M.; Jung, A.; Akiyama, J.; Allen, L.; Brown, C.; Edmondson, J.; Levitt, S.; Carlson, E.;
2 Gillespie, A. M.; Villar, A.; Epstein, C. J.; Poulain, F. R. (2002) Sequential targeted deficiency of SP-A and
3 -D leads to progressive alveolar lipoproteinosis and emphysema. Am. J. Physiol. 283: L1002-L1010.
4 Heck, H. d'A.; Casanova, M.; Starr, T. B. (1990) Formaldehyde toxicity-new understanding. Crit. Rev. Toxicol.
5 20:397-426.
6 Heddle, J. A.; Shepson, P. B.; Gingerich, J. D.; So, K. W. (1993) Mutagenicity of peroxyacetyl nitrate (PAN) in
7 vivo: tests for somatic mutations and chromosomal aberrations. Environ. Mol. Mutagen. 21: 58-66.
8 Herbert, R. A.; Hailey, J. R.; Grumbein, S.; Chou, B. J.; Sills, R. C.; Haseman, J. K.; Goehl, T.; Miller, R. A.;
9 Roycroft, J. H.; Boorman, G. A. (1996) Two-year and lifetime toxicity and carcinogenicity studies of ozone in
10 B6C3F1 mice. Toxicol. Pathol. 24: 539-548.
11 Highfill, J. W.; Watkinson, W. P. (1996) Ozone toxicity in the rat. II. Modeling changes due to ambient temperatures
12 and duration. J. Appl. Physiol. 80: 1811-1818.
13 Hoffer, E.; Baum, Y.; Tabak, A.; Frevert, C. (1999) Adhesion molecules of blood polymorphonuclear leukocytes
14 and alveolar macrophages in rats: modulation by exposure to ozone. Hum. Exp. Toxicol. 18: 547-551.
15 Holt, P. G.; Macaubas, C.; Stumbles, P. A.; Sly, P. D. (1999) The role of allergy in the development of asthma.
16 Nature (London, U. K.) 402(suppl. 25): B12-B16.
17 Hotchkiss, J. A.; Harkema, J. R.; Johnson, N. F. (1997) Kinetics of nasal epithelial cell loss and proliferation in F344
18 rats following a single exposure to 0.5 ppm ozone. Toxicol. Appl. Pharmacol. 143: 75-82.
19 Hotchkiss, J. A.; Hilaski, R.; Cho, H.; Regan, K.; Spencer, P.; Slack, K.; Harkema, J. R. (1998) Fluticasone
20 propionate attenuates ozone-induced rhinitis and mucous cell metaplasia in rat nasal airway epithelium.
21 Am. J. Respir. Cell Mol. Biol. 18: 91-99.
22 Huffman, L. J.; Judy, D. J.; Brumbaugh, K.; Frazer, D. G.; Reynolds, J. S.; McKinney, W. G.; Goldsmith, W. T.
23 (2001) Hyperthyroidism increases the risk of ozone-induced lung toxicity in rats. Toxicol. Appl. Pharmacol.
24 173: 18-26.
25 Huitron-Resendiz, S.; Custodio-Ramirez, V.; Escalante-Membrillo, C.; Gonzalez-Pina, R.; Paz, C. (1994) Sleep
26 alterations and brain regional changes of serotonin and its metabolite in rats exposed to ozone. Neurosci. Lett.
27 177: 119-122.
28 Hyde, D. M.; Miller, L. A.; McDonald, R. J.; Stovall, M. Y.; Wong, V.; Pinkerton, K. E.; Wegner, C. D.; Rothlein,
29 R.; Plopper, C. G. (1999) Neutrophils enhance clearance of necrotic epithelial cells in ozone-induced lung
30 injury in rhesus monkeys. Am. J. Physiol. 277: LI 190-L1198.
31 Igarashi, A.; lijima, H.; Tamura, G.; Shirato, K. (1998) Tazanolast inhibits ozone-induced airway
32 hyperresponsiveness in guinea pigs. Am. J. Respir. Crit. Care Med. 157: 1531-1535.
33 lijima, M. K.; Kobayashi, T.; Kamada, H.; Shimojo, N. (2001) Exposure to ozone aggravates nasal allergy-like
34 symptoms in guinea pigs. Toxicol. Lett. 123: 77-85.
35 Inoue, H.; Aizawa, H.; Nakano, H.; Matsumoto, K.; Kuwano, K.; Nadel, J. A.; Kara, N. (2000) Nitric oxide synthase
36 inhibitors attenuate ozone-induced airway inflammation in guinea pigs: possible role of interleukin-8. Am. J.
37 Respir. Crit. Care Med. 161: 249-256.
38 Ishii, Y.; Yang, H.; Sakamoto, T.; Nomura, A.; Hasegawa, S.; Hirata, F.; Bassett, D. J. P. (1997) Rat alveolar
3 9 macrophage cytokine production and regulation of neutrophil recruitment following acute ozone exposure.
40 Toxicol. Appl. Pharmacol. 147: 214-223.
41 Ishii, Y.; Hirano, K.; Morishima, Y.; Masuyama, K.; Goto, Y.; Nomura, A.; Sakamoto, T.; Uchida, Y.; Sagai, M.;
42 Sekizawa, K. (2000) Early molecular and cellular events of oxidant-induced pulmonary fibrosis in rats.
43 Toxicol. Appl. Pharmacol. 167: 173-181.
44 Iwasaki, T.; Takahashi, M.; Saito, H.; Arito, H. (1998) Adaptation of extrapulmonary responses to ozone exposure in
45 conscious rats. Ind. Health 36: 57-60.
46 Jakobi, G.; Fabian, P. (1997) Indoor/outdoor concentrations of ozone and peroxyacetyl nitrate (PAN). Int. J.
47 Biometeorol. 40: 162-165.
48 Jang, A.-S.; Choi, I.-S.; Koh, Y.-I.; Park, C.-S.; Lee, J.-S. (2002) The relationship between alveolar epithelial
49 proliferation and airway obstruction after ozone exposure. Allergy 57: 737-740.
50 Jimba, M.; Skornik, W. A.; Killingsworth, C. R.; Long, N. C.; Brain, J. D.; Shore, S. A. (1995) Role of C fibers in
51 physiological responses to ozone in rats. J. Appl. Physiol. 78: 1757-1763.
52 Joad, J. P.; Kott, K. S.; Bonham, A. C. (1998) Exposing guinea pigs to ozone for 1 wk enhances responsiveness of
53 rapidly adapting receptors. J. Appl. Physiol. 84: 1190-1197.
54 Joad, J. P.; Brie, J. M.; Weir, A. J.; Putney, L.; Hyde, D. M.; Postlewait, E. M.; Plopper, C. G. (2000) Effect of
55 respiratory pattern on ozone injury to the airways of isolated rat lungs. Toxicol. Appl. Pharmacol. 169: 26-32.
December 2005 5-75 DRAFT-DO NOT QUOTE OR CITE
-------
1 Johnston, C. I; Stripp, B. R.; Reynolds, S. D.; Avissar, N. E.; Reed, C. K.; Finkelstein, J. N. (1999) Inflammatory
2 and antioxidant gene expression in C57BL/6J mice after lethal and sublethal ozone exposures. Exp. Lung Res.
3 25:81-97.
4 Johnston, C. J.; Reed, C. K.; Avissar, N. E.; Gelein, R.; Finkelstein, J. N. (2000a) Antioxidant and inflammatory
5 response after acute nitrogen dioxide and ozone exposures in C57B1/6 mice. Inhalation Toxicol. 12: 187-203.
6 Johnston, C. J.; Oberdorster, G.; Gelein, R.; Finkelstein, J. N. (2000b) Newborn mice differ from adult mice in
7 chemokine and cytokine expression to ozone, but not to endotoxin. Inhalation Toxicol. 12: 205-224.
8 Johnston, C. J.; Oberdorster, G.; Gelein, R.; Finkelstein, J. N. (2002) Endotoxin potentiates ozone-induced
9 pulmonary chemokine and inflammatory responses. Exp. Lung Res. 28: 419-433.
10 Kafoury, R. M; Pryor, W. A.; Squadrito, G. L.; Salgo, M. G.; Zou, X.; Friedman, M. (1999) Induction of
11 inflammatory mediators in human airway epithelial cells by lipid ozonation products. Am. J. Respir. Crit.
12 Care Med. 160: 1934-1942.
13 Kenyon, N. J.; Van Der Vliet, A.; Schock, B.C.; Okamoto, T.; McGrew, G. M.; Last, J. A. (2002) Susceptibility to
14 ozone-induced acute lung injury in iNOS-deficient mice. Am. J. Physiol. 282: L540-L545.
15 Kim, M. Y.; Son, J. W.; Cho, M. H.; Choi, C. S.; Chae, C. H.; Lee, M. H. (2001) Oviductural carcinoma in B6C3F1
16 female mice exposed to 0.5 ppm ozone. Vet. Hum. Toxicol. 43: 370-372.
17 Kimmel, T. A.; Chen, L. C.; Bosland, M. C.; Nadziejko, C. (1997) Influence of acid aerosol droplet size on
18 structural changes in the rat lung caused by acute exposure to sulfuric acid and ozone. Toxicol. Appl.
19 Pharmacol. 144: 348-355.
20 Kleeberger, S. R.; Levitt, R. C.; Zhang, L.-Y.; Longphre, M.; Harkema, J.; Jedlicka, A.; Eleff, S. M.; DiSilvestre, D.;
21 Holroyd, K. J. (1997) Linkage analysis of susceptibility to ozone-induced lung inflammation in inbred mice.
22 Nat. Genet. 17: 475-478.
23 Kleeberger, S. R.; Reddy, S.; Zhang, L.-Y.; Jedlicka, A. E. (2000) Genetic susceptibility to ozone-induced lung
24 hyperpermeability: role of toll-like receptor 4. Am. J. Respir. Cell Mol. Biol. 22: 620-627.
25 Kleeberger, S. R.; Reddy, S. P.; Zhang, L.-Y.; Cho, H.-Y.; Jedlicka, A. E. (2001) Toll-like receptor 4 mediates
26 ozone-induced murine lung hyperpermeability via inducible nitric oxide synthase. Am. J. Physiol.
27 280: L326-L333.
28 Kleindienst, T. E.; Shepson, P. B.; Smith, D. F.; Hudgens, E. E.; Nero, C. M.; Cupitt, L. T.; Bufalini, J. J.; Claxton,
29 L. D. (1990) Comparison of mutagenic activities of several peroxyacyl nitrates. Environ. Mol. Mutagen.
30 16:70-80.
31 Kleinman, M. T.; Mautz, W. J.; Bjarnason, S. (1999) Adaptive and non-adaptive responses in rats exposed to ozone,
32 alone and in mixtures, with acidic aerosols. Inhalation Toxicol. 11: 249-264.
33 Kleinman, M. T.; Bufalino, C.; Rasmussen, R.; Hyde, D.; Bhalla, D. K.; Mautz, W. J. (2000) Toxicity of chemical
34 components of ambient fine paniculate matter (PM2 5) inhaled by aged rats. J. Appl. Toxicol. 20: 357-364.
35 Kligerman, A. D.; Mottus, K.; Erexson, G. L. (1995) Cytogenetic analyses of the in vitro and in vivo responses of
36 murine cells to peroxyacetyl nitrate (PAN). Mutat. Res. 341: 199-206.
37 Kobzik, L.; Goldsmith, C.-A. W.; Ning, Y. Y.; Qin, G.; Morgan, B.; Imrich, A.; Lawrence J.; Murthy, G. G. K.;
3 8 Catalano, P. J. (2001) Effects of combined ozone and air pollution particle exposure in mice. Boston, MA:
39 Health Effects Institute; research report no. 106. Available: http://www.healtheffects.org/Pubs/Kobzik.pdf
40 [27 January 2003].
41 Kodavanti, U. P.; Hatch, G. E.; Starcher, B.; Giri, S. N.; Winsett, D.; Costa, D. L. (1995) Ozone-induced pulmonary
42 functional, pathological, and biochemical changes in normal and vitamin C-deficient guinea pigs. Fundam.
43 Appl. Toxicol. 24: 154-164.
44 Koto, H.; Aizawa, H.; Takata, S.; Inoue, H.; Kara, N. (1995) An important role of tachykinins in ozone-induced
45 airway hyperresponsiveness. Am. J. Respir. Crit. Care Med. 151: 1763-1769.
46 Koto, H.; Salmon, M.; Haddad el-B.; Huang, T.-J.; Zagorski, J.; Chung, K. F. (1997) Role of cytokine-induced
47 neutrophil chemoattractant (CINC) in ozone-induced airway inflammation and hyperresponsiveness. Am. J.
48 Respir. Crit. Care Med. 156: 234-239.
49 Koyama, Y.; Hayaishi, O. (1994) Modulation by prostaglandins of activity of sleep-related neurons in the
50 preoptic/anterior hypothalamic areas in rats. Brain Res. Bull. 33: 367-372.
51 Kudo, M.; Nishikawa, M.; Ikeda, H.; Okubo, T. (1996) Involvement of superoxide anions in ozone-induced airway
52 hyperresponsiveness in unanesthetized guinea pigs. Environ. Toxicol. Pharmacol. 2: 25-30.
53 Larson, S. D.; Schelegle, E. S.; Walby, W. F.; Gershwin, L. J.; Fanuccihi, M. V; Evans, M. J.; Joad, J. P.;
54 Tarkington, B. K.; Hyde, D. M.; Plopper, C. G. (2004) Postnatal remodeling of the neural components of the
5 5 epithelial-mesenchymal trophic unit in the proximal airways of infant rhesus monkeys exposed to ozone and
56 allergen. Toxicol. Appl. Pharmacol. 194: 211-220.
December 2005 5-76 DRAFT-DO NOT QUOTE OR CITE
-------
1 Laskin, D. L.; Laskin, J. D. (2001) Role of macrophages and inflammatory mediators in chemically induced toxicity.
2 Toxicology (Ireland) 160: 111-118.
3 Laskin, D. L.; Pendino, K. J.; Punjabi, C. J.; del Valle, M. R.; Laskin, J. D. (1994) Pulmonary and hepatic effects of
4 inhaled ozone in rats. Environ. Health Perspect. 102(suppl. 10): 61-64.
5 Laskin, J. D.; Heck, D. E.; Laskin, D. L. (1996) Nitric oxide production in the lung and liver following inhalation of
6 the pulmonary irritant ozone. In: Snyder, R.; Kocsis, J. J.; Sipes, I. G.; Kalf, G. F.; Jollow, D. J.; Greim, H.;
7 Monks, T. J.; Witmer, C. M., eds. Biological Reactive Intermediates V: Basic Mechanistic Research in
8 Toxicology and Human Risk Assessment: proceedings of the Fifth International Symposium; January 1995;
9 Munich, Germany. Adv. Exp. Med. Biol. 387: 141-146.
10 Laskin, D. L.; Sunil, V.; Guo, Y.; Heck, D. E.; Laskin, J. D. (1998) Increased nitric oxide synthase in the lung after
11 ozone inhalation is associated with activation of NF-KB. Environ. Health Perspect. 106(suppl. 5): 1175-1178.
12 Laskin, D. L.; Fakhrzadeh, L.; Heck, D. E.; Gerecke, D.; Laskin, J. D. (2002) Upregulation of phosphoinositide
13 3 -kinase and protein kinase B in alveolar macrophages following ozone inhalation. Role of NF-"kappa"B and
14 STAT-1 in ozone-induced nitric oxide production and toxicity. Mol. Cell. Biochem. 234-235: 91-98.
15 Last, J. A.; Pinkerton, K. E. (1997) Chronic exposure of rats to ozone and sulfuric acid aerosol: biochemical and
16 structural responses. Toxicology 116: 133-146.
17 Lavnikova, N.; Prokhorova, S.; Lakhotia, A. V.; Gordon, R.; Laskin, D. L. (1998) Distinct inflammatory responses
18 of adherent vascular lung neutrophils to pulmonary irritants. J. Inflammation 48: 56-66
19 Lee, C.; Watt, K. C.; Chang, A. M.; Plopper, C. G.; Buckpitt, A. R.; Pinkerton, K. E. (1998) Site-selective
20 differences in cytochrome P450 isoform activities: comparison of expression in rat and rhesus monkey lung
21 and induction in rats. Drug Metab. Dispos. 26: 396-400.
22 Lemos, M.; Lichtenfels, A. J. F. C.; Amaro, E., Jr.; Macchione, M.; Martins, M. A.; King, M.; Bohm, G. M.;
23 Saldiva, P. H. N. (1994) Quantitative pathology of nasal passages in rats exposed to urban levels of air
24 pollution. Environ. Res. 66: 87-95.
25 Long, N. C.; Suh, J.; Morrow, J. D.; Schiestl, R. H.; Krishna Murthy, G. G.; Brain, J. D.; Frei, B. (2001) Ozone
26 causes lipid peroxidation but little antioxidant depletion in exercising and nonexercising hamsters. J. Appl.
27 Physiol. 91: 1694-1700.
28 Longphre, M.; Zhang, L.-Y.; Harkema, J. R.; Kleeberger, S. R. (1996) Mast cells contribute to O3-induced epithelial
29 damage and proliferation in nasal and bronchial airways of mice. J. Appl. Physiol. 80: 1322-1330.
30 Longphre, M.; Zhang, L.-Y.; Harkema, J. R.; Kleeberger, S. R. (1999) Ozone-induced pulmonary inflammation and
31 epithelial proliferation are partially mediated by PAF. J. Appl. Physiol. 86: 341-349.
32 Lorz, C.; Lopez, J. (1997) Incidence of air pollution in the pulmonary surfactant system of the pigeon (Columbia
33 livid). Anat. Rec. 249: 206-212.
34 Madden, M. C.; Richards, J. H.; Dailey, L. A.; Hatch, G. E.; Ohio, A. J. (2000) Effect of ozone on diesel exhaust
35 particle toxicity in rat lung. Toxicol. Appl. Pharmacol. 168: 140-148.
36 Mango, G. W.; Johnston, C. J.; Reynolds, S. D.; Finkelstein, J. N.; Plopper, C. G.; Stripp, B. R. (1998) Clara cell
37 secretory protein deficiency increases oxidant stress response in conducting airways. Am. J. Physiol.
38 275: L348-L356.
39 Matsubara, S.; Fushimi, K.; Kaminuma, O.; Kikkawa, H.; Shimazu, N.; Iwasaki, H.; Ikezawa, K. (1995) Importance
40 of impairment of the airway epithelium for ozone-induced airway hyperresponsiveness in guinea pigs. Jpn. J.
41 Pharmacol. 67: 375-382.
42 Matsubara, S.; Kikkawa, H.; Kaminuma, O.; Ikezawa, K. (1997a) Angiotensin-converting enzyme inhibitors can
43 potentiate ozone-induced airway hyperresponsiveness. Eur. J. Pharmacol. 337: 259-265.
44 Matsubara, S.; Fushimi, K.; Kaminuma, O.; Kikkawa, H.; Ikezawa, K.; Naito, K. (1997b) Prevention of
45 ozone-induced airway hyperresponsiveness and epithelial injury by phosphodiesterase inhibitors in guinea
46 pigs. Environ. Toxicol. Pharmacol. 3: 201-209.
47 Matsumoto, K.; Aizawa, H.; Inoue, H.; Koto, H.; Nakano, H.; Kara, N. (1999) Role of neutrophil elastase in
48 ozone-induced airway responses in guinea-pigs. Eur. Respir. J. 14: 1088-1094.
49 Mautz, W. J. (2003) Exercising animal models in inhalation toxicology: interactions with ozone and formaldehyde.
50 Environ. Res. 92: 14-26.
51 Mautz, W. J.; Kleinman, M. T.; Bhalla, D. K.; Phalen, R. F. (2001) Respiratory tract responses to repeated inhalation
52 of an oxidant and acid gas-particle air pollutant mixture. Toxicol. Sci. 61: 331-341.
53 McGraw, D. W.; Forbes, S. L.; Mak, J. C. W.; Witte, D. P.; Carrigan, P. E.; Leikauf, G. D.; Liggett, S. B. (2000)
54 Transgenic overexpression of beta(2)-adrenergic receptors in airway epithelial cells decreases
55 bronchoconstriction. Am. J. Physiol. 279: L379-L389.
December 2005 5-77 DRAFT-DO NOT QUOTE OR CITE
-------
1 McKinney, W. I; Jaskot, R. H.; Richards, J. H.; Costa, D. L.; Dreher, K. L. (1998) Cytokine mediation of
2 ozone-induced pulmonary adaptation. Am. J. Respir. Cell. Mol. Biol. 18: 696-705.
3 Miller, L. A.; Barnett, N. L.; Sheppard, D.; Hyde, D. M. (2001) Expression of the P6 integrin subunit is associated
4 with sites of neutrophil influx in lung epithelium. J. Histochem. Cytochem. 49: 41-48.
5 Morgan, M. S.; Meyer, P.; Holub, R.; Frank, R. (1986) Overall and regional lung function in dogs exposed acutely to
6 ozone. Environ. Res. 41: 546-557.
7 Morio, L. A.; Hooper, K. A.; Brittingham, J.; Li, T.-H.; Gordon, R. E.; Turpin, B. J.; Laskin, D. L. (2001) Tissue
8 injury following inhalation of fine paniculate matter and hydrogen perioxide is associated with altered
9 production of inflammatory mediators and antioxidants by alveolar macrophages. Toxicol. Appl. Pharmacol.
10 177: 188-199.
11 Moss, O. R.; Gross, E. A.; James, R. A.; Janszen, D. B.; Ross, P. W.; Roberts, K. C.; Howard, A. M.; Harkema,
12 J. R.; Calderon-Garciduenas, L.; Morgan, K. T. (2001) Respiratory tract toxicity in rats exposed to Mexico
13 City air. Cambridge, MA: Health Effects Institute; research report no. 100. Available:
14 http://www.healtheffects.org/pubs-research.htm [15 May, 2003].
15 Mucke, W. (1996) The environment and the eye. Topics of ophthalmic toxicology. Leban. Med. J. 44: 146-150.
16 Murphy, D. J. (2002) Assessment of respiratory function in safety pharmacology. Fundam. Clin. Pharmacol.
17 16: 183-196.
18 Mutoh, T.; Joad, J. P.; Bonham, A. C. (2000) Chronic passive cigarette smoke exposure augments
19 bronchopulmonary C-fibre inputs to nucleus tractus solitarii neurones and reflex output in young guinea-pigs.
20 J. Physiol. (London) 523: 223-23 3.
21 Nakano, H.; Aizawa, H.; Matsumoto, K.; Fukuyama, S.; Inoue, H.; Kara, N. (2000) Cyclooxygenase-2 participates
22 in the late phase of airway hyperresponsiveness after ozone exposure in guinea pigs. Eur. J. Pharmacol.
23 403:267-275.
24 National Toxicology Program. (1994) NTP technical report on the toxicology and carcinogenesis studies of ozone
25 (CAS no. 10028-15-6) and ozone/NNK (CAS no. 10028-15-6/64091-91-4) in F344/N rats and B6C3FJ mice
26 (inhalation studies). Research Triangle Park, NC: U.S. Department of Health and Human Services, National
27 Institutes of Health; publication no. 95-3371. (National Toxicology Program technical report series: no. 440).
28 Neuhaus-Steinmetz, U.; Uffhausen, F.; Herz, U.; Renz, H. (2000) Priming of allergic immune responses by repeated
29 ozone exposure in mice. Am. J. Respir. Cell Mol. Biol. 23: 228-233.
30 Nichols, B. G.; Woods, J. S.; Luchtel, D. L.; Corral, J.; Koenig, J. Q. (2001) Effects of ozone exposure on nuclear
31 factor-KB activation and tumor necrosis factor-a expression in human nasal epithelial cells. Toxicol. Sci.
32 60:356-362.
33 Nielsen, G. D.; Hougaard, K. S.; Larsen, S. T.; Hammer, M.; Wolkoff, P.; Clausen, P. A.; Wilkins, C. K.; Alarie, Y.
34 (1999) Acute airway effects of formaldehyde and ozone inBALB/c mice. Hum. Exp. Toxicol. 18: 400-409.
35 Nino-Cabrera, H. G.; Colin-Barenque, L.; Avila-Costa, M. R.; Espinosa-Villanueva, J.; Fortoul, T. I.;
36 Rivas-Arancibia, S. (2002) Differences between hippocampus and cerebral cortex in aged rats in an oxidative
37 stress model. Int. J. Neurosci. 112: 373-381.
38 Noviski, N.; Brewer, J. P.; Skornik, W. A.; Galli, S. J.; Drazen, J. M.; Martin, T. R. (1999) Mast cell activation is
39 not required for induction of airway hyperresponsiveness by ozone in mice. J. Appl. Physiol. 86: 202-210.
40 O'Connor, G. T.; Sparrow, D.; Weiss, S. T. (1995) A prospective longitudinal study of methacholine airway
41 responsiveness as a predictor of pulmonary-function decline: the Normative Aging Study. Am. J. Respir. Crit.
42 Care Med. 152: 87-92.
43 Paige, R. C.; Wong, V.; Plopper, C. G. (2000) Long-term exposure to ozone increases acute pulmonary centriacinar
44 injury by 1-nitronaphthalene: II. Quantitative histopathology. J. Pharmacol. Exp. Ther. 295: 942-950.
45 Palmer, L. J.; Burton, P. R.; Faux, J. A. (2000) Independent inheritance of serum immunoglobulin E concentrations
46 and airway responsiveness. Am. J. Respir. Crit. Care Med. 161: 1836-1842.
47 Palmer, L. J.; Rye, P. J.; Gibson, N. A. (2001) Airway responsiveness in early infancy predicts asthma, lung
48 function, and respiratory symptoms by school age. Am. J. Respir. Crit. Care Med. 163: 37-42.
49 Paquette, N. C.; Tankersley, C. G.; Zhang, L.-Y.; Kleeberger, S. R. (1994) Repeated subacute ozone exposure of
50 inbred mice: airway inflammation and ventilation. Exp. Lung Res. 20: 579-594.
51 Paz, C.; Bazan-Perkins, B. (1992) Sleep-wake disorganization in cats exposed to ozone. Neurosci. Lett.
52 140:270-272.
53 Paz, C.; Huitron-Resendiz, S. (1996) The effects of ozone exposure on the sleep-wake cycle and serotonin contents
54 in the pons of the rat. Neurosci. Lett. 204: 49-52.
55 Paz, C. (1997) Some consequences of ozone exposure on health. Arch. Med. Res. 28: 163-170.
December 2005 5-78 DRAFT-DO NOT QUOTE OR CITE
-------
1 Pearson, A. C.; Bhalla, D. K. (1997) Effects of ozone onmacrophage adhesion in vitro and epithelial and
2 inflammatory responses in vivo: the role of cytokines. J. Toxicol. Environ. Health 50: 143-157.
3 Peden, D. B.; Dailey, L. (1995) Modulation of mast cell functions by in vitro ozone exposure. Am. J. Physiol.
4 268: L902-L910.
5 Pendino, K. J.; Shuler, R. L.; Laskin, J. D.; Laskin, D. L. (1994) Enhanced production of interleukin-1, tumor
6 necrosis factor-"alpha", and fibronectin by rat lung phagocytes following inhalation of a pulmonary irritant.
7 Am. J. Respir. Cell Mol. Biol. 11: 279-286.
8 Pendino, K. J.; Meidhof, T. M.; Heck, D. E.; Laskin, J. D.; Laskin, D. L. (1995) Inhibition of macrophages with
9 gadolinium chloride abrogates ozone-induced pulmonary injury and inflammatory mediator production. Am.
10 J. Respir. Cell Mol. Biol. 13: 125-132.
11 Pendino, K. J.; Gardner, C. R.; Shuler, R. L.; Laskin, J. D.; Durham, S. K.; Barton, D. S.; Ohnishi, S. T.; Ohnishi, T.;
12 Laskin, D. L. (1996) Inhibition of ozone-induced nitric oxide synthase expression in the lung by endotoxin.
13 Am. J. Respir. Cell Mol. Biol. 14: 516-525.
14 Petruzzi, S.; Fiore, M.; Dell'Omo, G.; Bignami, G.; Alleva, E. (1995) Medium and long-term behavioral effects in
15 mice of extended gestational exposure to ozone. Neurotoxicol. Teratol. 17: 463-470.
16 Petruzzi, S.; De Acetis, L.; Chiarotti, F.; Sorace, A.; Alleva, E. (1999) Limited changes in handedness and morphine
17 reactivity in CD-I mice after pre- and postnatal ozone exposure. Acta Neurobiol. Exp. 59: 115-122.
18 Pinkerton, K. E.; Weller, B. L.; Menache, M. G.; Plopper, C. G. (1998) Consequences of prolonged inhalation of
19 ozone on F344/N rats: collaborative studies. Part XIII. A comparison of changes in the tracheobronchial
20 epithelium and pulmonary acinus in male rats at 3 and 20 months. Cambridge, MA: Health Effects Institute;
21 research report no. 65.
22 Plopper, C. G.; Fanucchi, M. V. (2000) Do urban environmental pollutants exacerbate childhood lung diseases?
23 Environ. Health Perspect. 108: A252-A253.
24 Plopper, C. G.; Hatch, G. E.; Wong, V.; Duan, X.; Weir, A. J.; Tarkington, B. K.; Devlin, R. B.; Becker, S.;
25 Buckpitt, A. R. (1998) Relationship of inhaled ozone concentration to acute tracheobronchial epithelial injury,
26 site-specific ozone dose and glutathione depletion in rhesus monkeys. Am. J. Respir. Cell Mol. Biol.
27 19:387-399.
28 Polosa, R.; Holgate, S. T. (1997) Adenosine bronchoprovocation: a promising marker of allergic inflammation in
29 asthma? Thorax 52: 919-923.
30 Postlethwait, E. M.; Cueto, R.; Velsor, L. W.; Pryor, W. A. (1998) Cyinduced formation of bioactive lipids:
31 estimated surface concentrations and lining layer effects. Am. J. Physiol. 274: L1006-L1016.
32 Postlethwait, E. M.; Joad, J. P.; Hyde, D. M.; Schelegle, E. S.; Brie, J. M.; Weir, A. J.; Putney, L. F.; Wong, V. J.;
33 Velsor, L. W.; Plopper, C. G. (2000) Three-dimensional mapping of ozone-induced acute cytotoxicity in
34 tracheobronchial airways of isolated perfused rat lung. Am. J. Respir. Cell Mol. Biol. 22: 191-199.
35 Prows, D. R.; Shertzer, H. G.; Daly, M. J.; Sidman, C. L.; Leikauf, G. D. (1997) Genetic analysis of ozone-induced
36 acute lung injury in sensitive and resistant strains of mice. Nat. Genet. 17: 471-474.
37 Prows, D. R.; Daly, M. J.; Shertzer, H. G.; Leikauf, G. D. (1999) Ozone-induced acute lung injury: genetic analysis
38 of F2 mice generated from A/J and C57BL/6J strains. Am. J. Physiol. 277: L372-L380.
39 Pryor, W. A.; Squadrito, G. L.; Friedman, M. (1995) A new mechanism for the toxicity of ozone. Toxicol. Lett.
40 82/83:287-293.
41 Pryor, W. A.; Bermudez, E.; Cueto, R.; Squadrito, G. L. (1996) Detection of aldehydes in bronchoalveolar lavage of
42 rats exposed to ozone. Fundam. Appl. Toxicol. 34: 148-156.
43 Rehle, D.; Leleux, D.; Erdelyi, M.; Tittel, F.; Fraser, M.; Friedfeld, S.; et al. (2001) Ambient formaldehyde detection
44 with a laser spectrometer based on difference-frequency generation in PPLN. Appl. Phys. B: Lasers Opt.
45 72: 947-952.
46 Reinhart, P. G.; Gupta, S. K.; Bhalla, D. K. (1999) Attenuation of ozone-induced lung injury by interleukin-10.
47 Toxicol. Lett. 110: 35-42.
48 Renner, R. (2002) Bad air and birth defects. Environ. Health Perspect. 110: A291.
49 Ritz, B.; Yu, F. (1999) The effect of ambient carbon monoxide on low birth weight among children born in southern
50 California between 1989 and 1993. Environ. Health Perspect. 107: 17-25.
51 Ritz, B.; Yu, F.; Fruin, S.; Chapa, G.; Shaw, G. M.; Harris, J. A. (2002) Ambient air pollution and risk of birth
52 defects in Southern California. Am. J. Epidemiol. 155: 17-25.
53 Rivas-Arancibia, S.; Vazquez-Sandoval, R.; Gonzalez-Kladiano, D.; Schneider-Rivas, S.; Lechuga-Guerrero, A.
54 (1998) Effects of ozone exposure in rats on memory and levels of brain and pulmonary superoxide dismutase.
55 Environ. Res. 76: 33-39.
December 2005 5-79 DRAFT-DO NOT QUOTE OR CITE
-------
1 Rivas-Arancibia, S.; Dorado-Martinez, C.; Borgonio-Perez, G.; Hiriart-Urdanivia, M.; Verdugo-Diaz, L.;
2 Duran-Vazquez, A.; Colin-Baranque, L.; Avila-Costa, M. R. (2000) Effects of taurine on ozone-induced
3 memory deficits and lipid peroxidation levels in brains of young, mature, and old rats. Environ. Res. 82: 7-17.
4 Rivas-Arancibia, S.; Dorado-Martinez, C.; Colin-Barenque, L.; Kendrick, K. M.; De la Riva, C.; Guevara-Guzman,
5 R. (2003) Effect of acute ozone exposure on locomotor behavior and striatal function. Pharmacol. Biochem.
6 Behav. 74: 891-900.
7 Rivas-Manzano, P.; Paz, C. (1999) Cerebellar morphological alterations in rats induced by prenatal ozone exposure.
8 Neurosci. Lett. 276: 37-40.
9 Rose, R. C.; Richer, S. P.; Bode, A. M. (1998) Ocular oxidants and antioxidant protection. Proc. Soc. Exp. Biol.
10 Med. 217: 397-407.
11 Saldiva, P. H. N.; King, M.; Delmonte, V. L. C.; Macchione, M.; Parada, M. A. C.; Daliberto, M. L.; Sakae, R. S.;
12 Criado, P. M. P.; Silveira, P. L. P.; Zin, W. A.; Bohm, G. M. (1992) Respiratory alterations due to urban air
13 pollution: an experimental study in rats. Environ. Res. 57: 19-33.
14 Savov, J. D.; Whitehead, G. S.; Wang, I; Liao, G.; Usuka, I; Peltz, G.; Foster, W. M.; Schwartz, D. A. (2004)
15 Ozone-induced acute pulmonary injury in inbred mouse strains. Am. J. Respir. Cell Mol. Biol. 31: 69-77'.
16 Schelegle, E. S.; Alfaro, M. F.; Putney, L.; Stovall, M.; Tyler, N.; Hyde, D. M. (2001) Effect of C-fiber-mediated,
17 ozone-induced rapid shallow breathing on airway epithelial injury in rats. J. Appl. Physiol. 91: 1611-1618.
18 Schelegle, E. S.; Miller, L. A.; Gershwin, L. J.; Fanucchi, M. V.; Van Winkle, L. S.; Gerriets, J. E.; Walby, W. F.;
19 Mitchell, V; Tarkington, B. K.; Wong, V J.; Baker, G. L.; Pantle, L. M.; Joad, J. P.; Pinkerton, K. E.; Wu,
20 R.; Evans, M. J.; Hyde, D. M.; Plopper, C. G. (2003a) Repeated episodes of ozone inhalation amplifies the
21 effects of allergen sensitization and inhalation on airway immune and structural development in Rhesus
22 monkeys. Toxicol. Appl. Pharmacol. 191: 74-85.
23 Schelegle, E. S.; Walby, W. F.; Alfaro, M. F.; Wong, V. J.; Putney, L.; Stovall, M. Y.; Sterner-Kock, A.; Hyde,
24 D. M.; Plopper, C. G. (2003b) Repeated episodes of ozone inhalation attenuates airway injury/repair and
25 release of substance P, but not adaptation. Toxicol. Appl. Pharmacol. 186: 127-142.
26 Schlesinger, R. B. (1995) Interaction of gaseous and paniculate pollutants in the respiratory tract: mechanisms and
27 modulators. Toxicology 105: 315-325.
28 Schlesinger, R. B.; Cohen, M. D.; Gordon, T.; Nadziejko, C.; Zelikoff, J. T.; Sisco, M.; Regal, J. F.; Menache, M. G.
29 (2002a) Ozone differentially modulates airway responsiveness in atopic versus nonatopic guinea pigs.
30 Inhalation Toxicol. 14:431-457.
31 Schlesinger, R. B.; Cohen, M.; Gordon, T.; Nadziejko, C.; Zelikoff, J. T.; Sisco, M.; Regal, J. F.; Menache, M. G.
32 (2002b) Ozone-induced modulation of airway hyperresponsiveness in guinea pigs. Boston, MA: Health
33 Effects Institute; research report no. 109.
34 Schwartz, D. A. (2002) TLR4 and LPS hyporesponsiveness in humans. Int. J. Hyg. Environ. Health 205: 221-227.
35 Sears, M. R.; Jones, D. T.; Holdaway, M. D.; Hewitt, C. J.; Flannery, E. M.; Herbison, G. P.; Silva, P. A. (1986)
36 Prevalence of bronchial reactivity to inhaled methacholine in New Zealand children. Thorax 41: 283-289.
37 Segura, P.; Montano, L. M.; Bazan-Perkins, B.; Gustin, P.; Vargas, M. H. (1997) Ozone at high-pollution urban
3 8 levels causes airway hyperresponsiveness to substance P but not to other agonists. Environ. Toxicol.
39 Pharmacol. 3: 91-95.
40 Sen, S.; Dulchavsky, S. A.; Dutta, S. (1993) Effects of triiodothyronine (T3) supplementation upon ozone-induced
41 lung injury. Free Radic. Res. Commun. 18: 299-308.
42 Sherwin, R. P.; Richters, V.; Kraft, P.; Richters, A. (2000) Centriacinar region inflammatory disease in young
43 individuals: a comparative study of Miami and Los Angeles residents. Virchows Arch. 437: 422-428.
44 Shore, S. A.; Abraham, J. H.; Schwartzman, I. N.; Murthy, G. G.; Laporte, J. D. (2000) Ventilatory responses to
45 ozone are reduced in immature rats. J. Appl. Physiol. 88: 2023-2030.
46 Shore, S. A.; Schwartzman, I. N.; Le Blanc, B.; Krishna Murthy, G. G.; Doerschuk, C. M. (2001) Tumor necrosis
47 factor receptor 2 contributes to ozone-induced airway hyperresponsiveness in mice. Am. J. Respir. Crit. Care
48 Med. 164: 602-607.
49 Shore, S. A.; Johnston, R. A.; Schwartzman, I. N.; Chism, D.; Krishna Murthy, G. G. (2002) Ozone-induced airway
50 hyperresponsiveness is reduced in immature mice. J. Appl. Physiol. 92: 1019-1028.
51 Shore, S. A.; Rivera-Sanchez, Y. M.; Schwartzman, I. N.; Johnston, R. A. (2003) Responses to ozone are increased
52 in obese mice. J. Appl. Physiol. 95: 938-945.
53 Sindhu, R. K.; Mautz, W. J.; Kikkawa, Y. (1998) Chronic exposure to ozone and nitric acid vapor results in
54 increased levels of rat pulmonary putrescine. Arch. Toxicol. 72: 445-449.
55 Slade, R.; Watkinson, W. P.; Hatch, G. E. (1997) Mouse strain differences in ozone dosimetry and body temperature
56 changes. Am. J. Physiol. 272: L73-L77.
December 2005 5-80 DRAFT-DO NOT QUOTE OR CITE
-------
1 Sommer, B.; Montano, L. M; Chavez, J.; Gustin, P.; Vargas, M. H. (1998) Guinea pig lung resistance shows
2 circadian rhythmicity not influenced by ozone. Respir. Physiol. 113: 223-229.
3 Sommer, B.; Vargas, M. H.; Chavez, J.; Carbajal, V; Segura, P.; Montano, L. M. (2001) Differences between
4 inhaled and intravenous bronchial challenge to detect O3-induced hyperresponsiveness. J. Appl. Physiol.
5 91:2595-2601.
6 Sorace, A.; De Acetis, L.; Alleva, E.; Santucci, D. (2001) Prolonged exposure to low doses of ozone: short- and
7 long-term changes to behavioral performance in mice. Environ. Res. 85: 122-134.
8 Spannhake, E. W. (1996) Down-regulation of canine airway mast cell function following exposure to ozone in vivo.
9 Exp. Lung Res. 22: 163-178.
10 Sterner-Kock, A.; Kock, M.; Braun, R.; Hyde, D. M. (2000) Ozone-induced epithelial injury in the ferret is similar to
11 nonhuman primates. Am. J. Respir. Crit. Care Med. 162: 1152-1156.
12 Stevens, W. H. M.; Adelroth, E.; Wattie, J.; Woolley, M. J.; Ellis, R.; Dahlback, M.; O'Byrne, P. M. (1994) Effect of
13 inhaled budesonide on ozone-induced airway hyperresponsiveness and bronchoalveolar lavage cells in dogs.
14 J. Appl. Physiol. 77: 2578-2583.
15 Stevens, W. H. M.; VanderHeyden, C.; Wattie, J.; Lane, C. G.; Smith, W.; O'Byrne, P. M. (1995a) Effect of a
16 leukotriene B4 receptor antagonist SC-53228 on ozone-induced airway hyperresponsiveness and inflammation
17 in dogs. Am. J. Respir. Crit. Care Med. 152: 1443-1448.
18 Stevens, W. H. M.; Conlon, P. D.; O'Byrne, P. M. (1995b) Ozone-induced oxygen radical release from
19 bronchoalveolar lavage cells and airway hyper-responsiveness in dogs. J. Physiol. (London) 486: 257-265.
20 Stick, S. M. (2002) Pulmonary physiology, airway responsiveness and asthma. Med. J. Aust. 177: S55-S56.
21 Stick, S. M.; Turnbull, S.; Chua, H. L.; Landau, L. L; LeSouef, P. N. (1990) Bronchial responsiveness to histamine
22 in infants and older children. Am. Rev. Respir. Dis. 142: 1143-1146.
23 Sun, J.; Chung, K. F. (1997) Interaction of ozone exposure with airway hyperresponsiveness and inflammation
24 induced by trimellitic anhydride in sensitized guinea pigs. J. Toxicol. Environ. Health 51: 77-87.
25 Sun, J.; Koto, H.; Chung, K. F. (1997) Interaction of ozone and allergen challenges on bronchial responsiveness and
26 inflammation in sensitised guinea pigs. Int. Arch. Allergy Immunol. 112: 191-195.
27 Szarek, J. L.; Stewart, N. L.; Zhang, J. Z.; Webb, J. A.; Valentovic, M. A.; Catalano, P. (1995) Contractile responses
28 and structure of small bronchi isolated from rats after 20 months' exposure to ozone. Fundam. Appl. Toxicol.
29 28: 199-208.
30 Tager, I. B. (1999) Air pollution and lung function growth. Is it ozone? [editorial]. Am. J. Respir. Crit. Care Med.
31 160:387-389.
32 Takahashi, T.; Miura, M.; Katsumata, U.; Ichinose, M.; Kimura, K.; Inoue, H.; Takishima, T.; Shirato, K. (1993)
33 Involvement of superoxide in ozone-induced airway hyperresponsiveness in anesthetized cats. Am. Rev.
34 Respir. Dis. 148: 103-106.
35 Takahashi, N.; Yu, X.-Y.; Schofield, B. H.; Kleeberger, S. R.; Scott, A. L.; Hasegawa, S.; Spannhake, E. W. (1995a)
3 6 Expression of ICAM-1 in airway epithelium after acute ozone exposure in the mouse. J. Appl. Physiol.
37 79: 1753-1761.
38 Takahashi, M.; Kleeberger, S. R.; Croxton, T. L. (1995b) Genetic control of susceptibility to ozone-induced changes
39 in mouse trachea! electrophysiology. Am. J. Physiol. 269: L6-L10.
40 Takata, S.; Aizawa, H.; Inoue, H.; Koto, H.; Kara, N. (1995) Ozone exposure suppresses epithelium-dependent
41 relaxation in feline airway. Lung 173: 47-56.
42 Takebayashi, T.; Abraham, J.; Murthy, G. G. K.; Lilly, C.; Rodger, I.; Shore, S. A. (1998) Role of tachykinins in
43 airway responses to ozone in rats. J. Appl. Physiol. 85: 442-450.
44 Tankersley, C. G.; Kleeberger, S. R. (1994) Ozone-induced inflammation and altered ventilation in genetically
45 susceptible mice: a comparison of acute and subacute exposures. Toxicol Lett. 72: 279-289.
46 Tankersley, C. G.; Fitzgerald, R. S.; Mitzner, W. A.; Kleeberger, S. R. (1993) Hypercapnic ventilatory responses in
47 mice differentially susceptible to acute ozone exposure. J. Appl. Physiol. 75: 2613-2619.
48 Tesfaigzi, J.; Hotchkiss, J. A.; Harkema, J. R. (1998) Expression of the Bcl-2 protein in nasal epithelia of F344/N
49 rats during mucous cell metaplasia and remodeling. Am. J. Respir. Cell Mol. Biol. 18: 794-799.
50 Thiele, J. J. (2001) Oxidative targets in the stratum corneum. A new basis for antioxidative strategies. Skin
51 Pharmacol. Appl. Skin Physiol. 14(suppl. 1): 87-91.
52 Thiele, J. J.; Traber, M. G.; Podda, M.; Tsang, K.; Cross, C. E.; Packer, L. (1997a) Ozone depletes tocopherols and
53 tocotrienols topically applied to murine skin. FEES Lett. 401: 167-170.
54 Thiele, J. J.; Podda, M.; Packer, L. (1997b) Tropospheric ozone: an emerging environmental stress to skin. Biol.
55 Chem. 378: 1299-1305.
December 2005 5-81 DRAFT-DO NOT QUOTE OR CITE
-------
1 Thiele, J. I; Traber, M. G.; Polefka, T. G.; Cross, C. E.; Packer, L. (1997c) Ozone-exposure depletes vitamin E and
2 induces lipid peroxidation in murine stratum corneum. J. Invest. Dermatol. 108: 753-757.
3 Thiele, J. J.; Traber, M. G.; Tsang, K.; Cross, C. E.; Packer, L. (1997d) In vivo exposure to ozone depletes vitamins
4 C and E and induces lipid peroxidation in epidermal layers of murine skin. Free Radical Biol. Med.
5 23:385-391.
6 Thiele, J. J.; Traber, M. G.; Packer, L. (1998) Depletion of human stratum corneum vitamin E: an early and sensitive
7 in vivo marker of UV induced photo-oxidation. J. Invest. Dermatol. 110: 756-761.
8 Thiele, J. J.; Hsieh, S. N.; Briviba, K.; Sies, H. (1999) Protein oxidation in human stratum corneum: susceptibility of
9 keratins to oxidation in vitro and presence of a keratin oxidation gradient in vivo. J. Invest. Dermatol.
10 113:335-339.
11 Tsai, J.-J.; Lin, Y.-C.; Kwan, Z.-H.; Kao, H.-L. (1998) Effects of ozone on ovalbumin sensitization in guinea pigs.
12 J. Microbiol. Immunol. Infect. 31: 225-232.
13 Tsukagoshi, H.; Haddad, E.-B.; Sun, J.; Barnes, P. J.; Chung, K. F. (1995) Ozone-induced airway
14 hyperresponsiveness: role of superoxide anions, NEP, and BK receptors. J. Appl. Physiol. 78: 1015-1022.
15 U.S. Environmental Protection Agency. (1993) Air quality criteria for oxides of nitrogen. Research Triangle Park,
16 NC: Office of Health and Environmental Assessment, Environmental Criteria and Assessment Office; report
17 nos. EPA/600/8-9!/049aF-cF. 3v. Available from: NTIS, Springfield, VA; PB95-124533, PB95-124525, and
18 PB95-124517.
19 U.S. Environmental Protection Agency. (1996) Air quality criteria for ozone and related photochemical oxidants.
20 Research Triangle Park, NC: Office of Research and Development; report nos. EP A/600/AP-93/004aF-cF. 3v.
21 Available from: NTIS, Springfield, VA; PB96-185582, PB96-185590, and PB96-185608. Available online at:
22 www.epa.gov/ncea/ozone.htm.
23 Uhlson, C.; Harrison, K.; Allen, C. B.; Ahmad, S.; White, C. W.; Murphy, R. C. (2002) Oxidized phospholipids
24 derived from ozone-treated lung surfactant extract reduce macrophage and epithelial cell viability. Chem.
25 Res. Toxicol. 15: 896-906.
26 Ulrich, M. M. W.; Alink, G. M.; Kumarathasan, P.; Vincent, R.; Boere, A. J.; Cassee, F. R. (2002) Health effects
27 and time course of paniculate matter on the cardiopulmonary system in rats with lung inflammation.
28 J. Toxicol. Environ. Health Part A 65: 1571-1595.
29 Valacchi, G.; Van der Vliet, A.; Schock, B.C.; Okamoto, T.; Obermuller-Jevic, U.; Cross, C. E.; Packer, L. (2002)
30 Ozone exposure activates oxidative stress responses in murine skin. Toxicology 179: 163-170.
31 Valacchi, G.; Pagnin, E.; Okamoto, T.; Corbacho, A. M.; Olano, E.; Davis, P. A.; Van der Vliet, A.; Packer, L.;
32 Cross, C. E. (2003) Induction of stress proteins and MMP-9 by 0.8 ppm of ozone in murine skin. Biochem.
33 Biophys. Res. Commun. 305: 741-746.
34 Valverde, M.; del Carmen Lopez, M.; Lopez, I.; Sanchez, I.; Fortoul, T. I.; Ostrosky-Wegman, P.; Rojas, E. (1997)
3 5 DNA damage in leukocytes and buccal and nasal epithelial cells of individuals exposed to air pollution in
36 Mexico City. Environ. Mol. Mutagen. 30: 147-152.
37 Van Bree, L.; Dormans, J. A. M. A.; Boere, A. J. F.; Rombout, P. J. A. (2001) Time study on development and
38 repair of lung injury following ozone exposure in rats. Inhalation Toxicol. 13: 703-717.
39 Van Bree, L.; Dormans, J. A. M. A.; Koren, H. S.; Devlin, R. B.; Rombout, P. J. A. (2002) Attenuation and recovery
40 of pulmonary injury in rats following short-term, repeated daily exposure to ozone. Inhalation Toxicol.
41 14:883-900.
42 Van Hoof, I. H. J. M.; Van Bree, L.; Bast, A. (1996) Changes in receptor function by oxidative stress in guinea pig
43 trachea! smooth muscle. Cent. Eur. J. Public Health 4(suppl.): 3-5.
44 Van Hoof, H. J. M.; Van Acker, F. A. A.; Voss, H.-P.; Van Bree, L.; Bast, A. (1997) Acute exposure to ozone does
45 not influence neuroreceptor density and sensitivity in guinea pig lung. Toxicol. Lett. 90: 53-60.
46 Vanda, B.; de Buen, N.; Jasso, R.; Valero, G.; Vargas, M. H.; Olmos, R.; Arreola, J. L.; Santillan, P.; Alonso, P.
47 (1998) Inflammatory cells and ferruginous bodies inbronchoalveolar lavage in urban dogs. Acta Cytol.
48 42: 939-944.
49 Vargas, M. H.; Segura, P.; Campos, M. G.; Hong, E.; Montano, L. M. (1994) Effect of ozone exposure on
50 antigen-induced airway hyperresponsiveness in guinea pigs. J. Toxicol. Environ. Health 42: 435-442.
51 Vargas, M. H.; Romero, L.; Sommer, B.; Zamudio, P.; Gustin, P.; Montano, L. M. (1998) Chronic exposure to ozone
52 causes tolerance to airway hyperresponsiveness in guiea pigs: lack of SOD role. J. Appl. Physiol.
53 84: 1749-1755.
54 Vesely, D. L. (1999) Atrial natriuretic peptides in the diagnosis and treatment of congestive heart failure. Congest.
55 Heart Fail. 5: 171-179.
December 2005 5-82 DRAFT-DO NOT QUOTE OR CITE
-------
1 Vesely, K. R.; Schelegle, E. S.; Stovall, M. Y.; Harkema, J. R.; Green, J. F.; Hyde, D. M. (1999a) Breathing pattern
2 response and epithelial labeling in ozone-induced airway injury in neutrophil-depleted rats. Am. J. Respir.
3 Cell Mol. Biol. 20:699-709.
4 Vesely, K. R.; Hyde, D. M.; Stovall, M. Y.; Harkema, J. R.; Green, J. F.; Schelegle, E. S. (1999b)
5 Capsaicin-sensitive C-fiber-mediated protective responses in ozone inhalation in rats. J. Appl. Physiol.
6 86:951-962.
7 Vincent, R.; Janzen, E. G.; Chen, G.; Kumarathasan, P.; Haire, D. L.; Guenette, J.; Chen, J. Z.; Bray, T. M. (1996)
8 Spin trapping study in the lungs and liver of F344 rats after exposure to ozone. Free Radical Res. 25: 475-488.
9 Vincent, R.; Bjarnason, S. G.; Adamson, I. Y. R.; Hedgecock, C.; Kumarathasan, P.; Guenette, J.; Potvin, M.;
10 Goegan, P.; Bouthillier, L. (1997) Acute pulmonary toxicity of urban paniculate matter and ozone. Am. J.
11 Pathol. 151: 1563-1570.
12 Vyskocil, A.; Viau, C.; Lamy, S. (1998) Peroxyacetyl nitrate: review of toxicity. Hum. Exp. Toxicol. 17: 212-220.
13 Wagner, J. G.; Hotchkiss, J. A.; Harkema, J. R. (200la) Effects of ozone and endotoxin coexposure on rat airway
14 epithelium: potentiation of toxicant-induced alterations. Environ. Health Perspect. 109(suppl. 4): 591-598.
15 Wagner, J. G.; VanDyken, S. J.; Hotchkiss, J. A.; Harkema, J. R. (200Ib) Endotoxin enhancement of ozone-induced
16 mucous cell metaplasia is neutrophil-dependent in rat nasal epithelium. Toxicol. Sci. 60: 338-347.
17 Wagner, J. G.; Hotchkiss, J. A.; Harkema, J. R. (2002) Enhancement of nasal inflammatory and epithelial responses
18 after ozone and allergen coexposure in brown Norway rats. Toxicol. Sci. 67: 284-294.
19 Wagner, J. G.; VanDyken, S. J.; Wierenga, J. R.; Hotchkiss, J. A.; Harkema, J. R. (2003) Ozone exposure enhances
20 endotoxin-induced mucous cell metaplasia in rat pulmonary airways. Toxicol. Sci. 74: 437-446.
21 Wang, G.; Umstead, T. M.; Phelps, D. S.; Al-Mondhiry, H.; Floras, J. (2002) The effect of ozone exposure on the
22 ability of human surfactant protein A variants to stimulate cytokine production. Environ. Health Perspect.
23 110:79-84.
24 Watkinson, W. P.; Wiester, M. J.; Highfill, J. W. (1995) Ozone toxicity in the rat. I. Effect of changes in ambient
25 temperature on extrapulmonary physiological parameters. J. Appl. Physiol. 78: 1108-1120.
26 Watkinson, W. P.; Campen, M. J.; Nolan, J. P.; Costa, D. L. (2001) Cardiovascular and systemic responses to
27 inhaled pollutants in rodents: effects of ozone and paniculate matter. Environ. Health Perspect.
28 109(suppl. 4): 539-546.
29 Watkinson, W. P.; Campen, M. J.; Wichers, L. B.; Nolan, J. P.; Costa, D. L. (2003) Cardiac and thermoregulatory
3 0 responses to inhaled pollutants in healthy and compromised rodents: modulation via interaction with
31 environmental factors. Environ. Res. 92: 35-47.
32 Watt, K. C.; Plopper, C. G.; Weir, A. J.; Tarkington, B.; Buckpitt, A. R. (1998) Cytochrome P450 2E1 in rat
33 tracheobronchial airways: response to ozone exposure. Toxicol. Appl. Pharmacol. 149: 195-202.
34 Wattiez, R.; Noel-Georis, L; Cruyt, C.; Broeckaert, F.; Bernard, A.; Falmagne, P. (2003) Susceptibility to oxidative
3 5 stress: proteomic analysis of bronchoalveolar lavage from ozone-sensitive and ozone-resistant strains of mice.
36 Proteomics 3: 658-665.
37 Weber, S. U.; Thiele, J. J.; Cross, C. E.; Packer, L. (1999) Vitamin C, uric acid, and glutathione gradients in murine
3 8 stratum corneum and their susceptibility to ozone exposure. J. Invest. Dermatol. 113:1128-1132.
39 Weber, S. U.; Jothi, S.; Thiele, J. J. (2000) High-pressure liquid chromatography analysis of ozone-induced
40 depletion of hydrophilic and lipophilic antioxidants in murine skin. Methods Enzymol. 319: 536-546.
41 Weber, S. U.; Han, N.; Packer, L. (2001) Ozone: an emerging oxidative stressorto skin. Curr. Probl. in Dermatol.
42 29:52-61.
43 Weller, B. L.; Crapo, J. D.; Slot, J.; Posthuma, G.; Plopper, C. G.; Pinkerton, K. E. (1997) Site- and cell-specific
44 alteration of lung copper/zinc and manganese superoxide dismutases by chronic ozone exposure. Am. J.
45 Respir. Cell Mol. Biol. 17: 552-560.
46 Weller, B. L.; Witschi, H.; Pinkerton, K. E. (2000) Quantitation and localization of pulmonary manganese
47 superoxide dismutase and tumor necrosis factor a following exposure to ozone and nitrogen dioxide. Toxicol.
48 Sci. 54: 452-461.
49 Wells, C. A.; Ravasi, T.; Faulkner, G. J.; Carninci, P.; Okazaki, Y.; Hayashizaki, Y.; Sweet, M.; Wainwright, B. J.;
50 Hume, D. A. (2003) Genetic control of the innate immune response. BMC Immunol. 4: 5. Available:
51 http://www.biomedcentral.eom/1471-2172/4/5 [18 February, 2003]
52 Wiester, M. J.; Watkinson, W. P.; Costa, D. L.; Crissman, K. M.; Richards, J. H.; Winsett, D. W.; Highfill, J. W.
53 (1996) Ozone toxicity in the rat. III. Effect of changes in ambient temperature on pulmonary parameters. J.
54 Appl. Physiol. 81: 1691-1700.
December 2005 5-83 DRAFT-DO NOT QUOTE OR CITE
-------
1 Wilkins, C. K.; Clausen, P. A.; Wolkoff, P.; Larsen, S. T.; Hammer, M.; Larsen, K.; Hansen, V.; Nielsen, G. D.
2 (2001) Formation of strong irritants in mixtures of isoprene/ozone and isoprene/ozone/nitrogen dioxide.
3 Environ. Health Perspect. 109: 937-941.
4 Witschi, H.; Espiritu, I.; Pinkerton, K. E.; Murphy, K.; Maronpot, R. R. (1999) Ozone carcinogenesis revisited.
5 Toxicol. Sci. 52: 162-167.
6 Wu, Z.-X.; Morton, R. F.; Lee, L.-Y. (1997) Role of tachykinins in ozone-induced airway hyperresponsiveness to
7 cigarette smoke in guinea pigs. J. Appl. Physiol. 83: 958-965.
8 Yamauchi, T.; Shima, M.; Kuwaki, T.; Ando, M.; Ohmichi, M.; Fukuda, Y.; Adachi, M. (2002) Acute effects of
9 ozone exposure on lung function in mice sensitized to ovalbumin. Toxicology (Ireland) 172: 69-78.
10 Yu, M.; Pinkerton, K. E.; Witschi, H. (2002) Short-term exposure to aged and diluted sidestream cigarette smoke
11 enhances ozone-induced lung injury in B6C3F1 mice. Toxicol. Sci. 65: 99-106.
12 Zhang, L.-Y.; Levitt, R. C.; Kleeberger, S. R. (1995) Differential susceptibility to ozone-induced airways
13 hyperreactivity in inbred strains of mice. Exp. Lung Res. 21: 503-518.
14 Zhao, Q.; Simpson, L. G.; Driscoll, K. E.; Leikauf, G. D. (1998) Chemokine regulation of ozone-induced neutrophil
15 and monocyte inflammation. Am. J. Physiol. 274: L39-L46.
16
17
December 2005 5-84 DRAFT-DO NOT QUOTE OR CITE
-------
i AX5. ANNEX TO CHAPTER 5 OF OZONE AQCD
2
3
4 AX5.1 INTRODUCTION
5 This annex serves to provide supporting material for Chapter 5, Toxicological Effects of
6 Ozone and Related Photochemical Oxidants in Laboratory Animals and In Vitro Test Systems.
7 It includes tables that summarize new toxicological literature published since the last O3 criteria
8 document (U.S. Environmental Protect!on Agency, 1996). In addition, it provides descriptions
9 of those new finding, in many cases, with more detail than is provided in the chapter.
10
11
12 AX5.2 RESPIRATORY TRACT EFFECTS OF OZONE
13 AX5.2.1 Biochemical Effects
14 Biochemically detected effects of O3 are integrally involved in effects on both structure
15 and function (respiratory and nonrespiratory) of the respiratory tract. However, even the few
16 relatively clear associations (e.g., increases in collagen metabolism/collagen and lung fibrosis)
17 are not fully understood, leading to difficulty in interpretation. For presentation, this section
18 summarizes studies designed to identify biochemical targets of O3, as well as biochemical
19 measurements of O3-induced changes in xenobiotic metabolism, antioxidant metabolism and
20 oxygen consumption, lipids and arachidonic acid metabolism, and collagen metabolism.
21 Only descriptions of new studies, published since the previous O3 CD are included. Detailed
22 discussions of older O3 literature are found in U.S. Environmental Protection Agency (1986,
23 1996).
24
25 AX5.2.1.1 Cellular Targets of Ozone Interaction
26 New studies characterizing the cellular targets of O3 interaction include the following.
27 Figure AX5-1 details the major secondary products of ozone interaction with lung cells.
28 Frampton et al. (1999) demonstrated the ozonation of PUFA to form nonanal and hexanal
29 in rat BAL after exposures to 0.22 ppm O3 for 4 h with exercise. Increases in nonanal were not
30 accompanied by significant changes in lung function, in epithelial permeability, or in airway
December 2005 AX5-1 DRAFT-DO NOT QUOTE OR CITE
-------
RHC = CH 4
PUFA
either In j
the », RHC
absence N
ofH2O
Criegi
O O
\ I
• O3 > RHC — CH— -
ozone trioxolane
0~OX or in the
CH — presence >
xo/ of H2O
se ozonide hyc
> RHC = O — O + RHC = O
carbonyl oxide aldehyde
RHC > RHC = O + H2O2
OOH aldehyde hydrogen
iroxyhydroperoxy cpd. peroxide
Figure AX5-1. Major secondary products of ozone interaction with lung cells.
1 inflammation. Hexanal levels did not increase significantly and levels of both aldehydes
2 returned to baseline by 18 h PE. Pryor et al. (1996) exposed rats to 0.5 to 10 ppm O3 both with
3 and without 5% CO2 to measure the amount of aldehyde generated in BAL, and also the rate of
4 disappearance of aldehydes from the ELF following the O3 exposure. Ozone exposure with CO2
5 increased the tidal volume and the yield of aldehydes with a maximal aldehyde yield at 2.5 ppm
6 for 1 h. Absolute yields were impossible to ascertain in this system because deposition of O3 is
7 unknown and aldehyde recovery is not complete due to loss of aldehyde by volatization and by
8 diffusion into underlying tissue. The data showed that at 0.5 ppm O3 with 5% CO2, levels of
9 hexanal and nonanal increased at 30 minutes, decreased slightly from that level at 60 minutes,
10 was maximal at 90 minutes and then dropped to 60 minutes levels at 120 minutes. Levels of
11 heptanal did not change appreciably during this time course. Levels of these aldehydes were
12 dependent on a dynamic relationship between their production and the disappearance from the
13 ELF. The authors stated that O3 is the limiting reagent in this process because the amount of
14 PUFA far exceed the amount of O3 on a molar basis. Because of the limitations of measuring
15 aldehydes in this study paradigm, it is not useful for quantitative dosimetry; however, the authors
16 suggest the study does serve to demonstrate the use of aldehydes as biomarkers of O3 exposure
17 since nonanol is produced in an O3-specific pathway.
18 Postlethwait et al. (1998) utilized three biologically relevant models (isolated epithelial
19 lining fluid, intact lung, and liposome suspensions) to determine the O3-induced production of
20 heptanal, nonanal and hexanal in an attempt to estimate formation of lipid-derived bioactive
December 2005
AX5-2
DRAFT-DO NOT QUOTE OR CITE
-------
1 compounds. Data suggest that PUFAs directly react with O3 and the amount of bioactive lipids
2 produced is inversely related to ascorbic acid availability. The authors caution that there are
3 limitations to the use of measurements of these reactions products in determining O3 dose-
4 response relationships due to the heterogenous nature of O3 reactions in the epithelial lining
5 fluid. Connor et al, (2004) have recently examined the reactive absorption of O3 within ELF
6 using interfacial films composed of dipalmitoylglycero-3-phosphocholine (DPPC) and rat lung
7 lavage fluid. The films reduced O3 reactive absorption by antioxidants. Further experiments
8 using a human lung fibroblast cell line exposed to O3 demonstrated that ascorbic acid (AA)
9 produced cell injury, that high levels of O3 and AA were needed to induce cell injury, and the
10 DPPC films reduced the amount of cell injury. From these data the authors suggest that O3
11 reactions with ELF substrates cause cell injury, that films of active, saturated phospholipids
12 reduce the local dose of O3-derived reaction products, and that these interfacial phospholipids
13 modulate the distribution of inhaled O3 and the extent of site-specific cell injury.
14 Recent studies have examined the formation of ozonation products such as
15 4-hydroxynonenal (FINE), a toxic aldehyde that reacts with cysteine, histamine, and lysine
16 amino acid residues and creates protein adducts. Hamilton et al. (1998) demonstrated (see
17 Chapter 6) using human AM exposed to 0.4 ppm O3 for 1 h that exposure caused apoptosis, an
18 increase in a 32-kDa protein adduct, and an increase in ferritin and a 72-kDa heat shock protein.
19 By exposing AM to FINE in vitro, all of these effects are replicated, which the authors interpret
20 to mean that creation of protein adducts and apoptotic cell death are cellular toxic effect of acute
21 O3 exposure and that it is mediated, at least in part by FINE.
22
23 AX5.2.1.2 Monooxygenases
24 Monooxygenases constitute a class of enzymes, including the cytochrome P-450 (C-P450)
25 or CYP system, that metabolize both endogenous and exogenous substances. Such metabolism
26 can result in detoxification (e.g., drugs like pentobarbital) or activation to more potent
27 metabolites (e.g., carcinogenic metabolites of benzo[a]pyrene, B[a]P). Although the liver has
28 the greatest capacity for xenobiotic metabolism, the lung also has its complement. The effects of
29 O3 on this metabolic system are summarized in Table AX5-1.
30 Lee et al. (1998) characterized the activities of various isoforms of CYPs in both rat and
31 rhesus monkey lung using microdissection techniques and found regiospecific and species-
December 2005 AX5-3 DRAFT-DO NOT QUOTE OR CITE
-------
December 2005
X
-k
Table AX5-1.
Effects of Ozone on Lung Monooxygenases
Concentration
ppm
1.0
1.0
0.8
1
aCYP =
WT =
MT =
CCSP
ug/m3 Duration
1,960 8h
1,960 90 days
1,600 8 h/ day for
90 days
1,960 2h
= Cytochrome P-450
wild-type
metallothionein
= Clara Cell Secretary Protein
Species
Rat
male, SD
350-600 g
Rat,
male SD
275-300 g
Mice,
Clara cell secretory
protein deficient,
WT strain 129
Effects"
Increases CYP2E1 activity in lobar-bronchi and major daughter airway with 8 h
exposure. Decreased CYP2E1 activities in both major and minor daughter
airways with 90 day exposure. O3 does not result in consistent dramatic
alterations in CYP2E1 activities.
C YP2B activity increased. Linked to Clara cells in distal lung only — not in
trachea or intrapulmonary airway.
CCSP-1-mice had increases in IL-6 and MT mRNA that preceded decreases in
Clara cell C YP2F2 mRNA. WT mice had levels change but to a lesser degree.
Reference
Watt etal. (1998)
Paige et al. (2000a)
Mango etal. (1998)
H
6
o
o
H
O
O
H
W
O
O
HH
H
W
-------
1 specific differences. In rat parenchyma both CYP 1 Al and CYP 2B were highest, whereas, in
2 rat airways, CYP 2E1 was highest. In rat airways and parenchyma P450 reductase activities
3 were high, and conversely, low in trachea. Monkeys did not exhibit such site-selective
4 differences in CYP 2B1, CYP 1A1, and P450 reductase; however, they had high CYP2E1
5 activity in parenchyma and distal bronchioles.
6 Watt et al. (1998) found that 1 ppm O3 in both short and long-term exposures in rat
7 increased CYP 2E1 in a region-specific manner (a more detailed discussion is in the Morphology
8 section of this chapter). This group (Paige et al., 2000a) further characterized CYP 2B
9 expression and activity in a long-term O3 exposure (0.8 ppm 8 h/day for 90 days). Activity of
10 CYP 2B increased 3-fold following this exposure, while CYP 2B like-immunoreactivity
11 increased 2-fold in microsomes prepared from distal lung. Changes in immunodetectable CYP
12 2B protein and activity were limited to Clara cells in distal lung and not present in trachea or
13 intrapulmonary airways.
14 Studies have focused on P450 gene expression to examine possible genetic mechanisms
15 that may explain differential O3-sensitivity (Mango et al., 1998). Mice (129 strain) deficient in
16 Clara cell secretory protein (CCSP-/-), which are oxidant-sensitive, were exposed to 1 ppm O3
17 for 2 h. The CCSP null mice demonstrated increases in IL-6 and metallothionein (MT) mRNA
18 that preceded decreases in Clara cell CYP2F2 mRNA (normally expressed at high levels in
19 mouse lung) levels. In 129 strain wild-type mice, RNA levels changed similarly, to a lesser
20 degree.
21
22 AX5.2.1.3 Antioxidants, Antioxidant Metabolism, and Mitochondrial
23 Oxygen Consumption
24 Ozone is an oxidant that produces reactive oxygen species (ROS) involved in the
25 molecular mechanism(s) of toxicity. Antioxidant chemical defenses are important in modulating
26 O3 toxicity.
27 Weller et al. (1997) studied the changes in distribution and abundance of copper-zinc
28 (Cu-Zn) and manganese (Mn) SOD in Fischer 344 rats exposed to 1.0 ppm O3 for up to
29 3 months. Using immunohistochemistry and immunogold labeling, they identified Cu-Zn SOD
30 positive and Mn SOD positive cells in exposed animals and compared them to controls.
31 In epithelial cells in airways and parenchyma they found reduced Cu-Zn SOD labeling in
December 2005 AX5-5 DRAFT-DO NOT QUOTE OR CITE
-------
1 O3-exposed rats. In the CAR regions in both AMs and type II epithelial cells they found
2 significantly increased levels of Mn SOD. Mn SOD levels were not increased in type I epithelial
3 cells, fibroblasts, or Clara cells. The authors suggest that the increased levels of Mn SOD in
4 type II cells in the proximal alveolar duct confer tolerance and protection from further
5 O3-induced injury.
6 To look at the effect of antioxidants on modulating O3-induced oxidant stress, Freed et al.
7 (1999) inhibited antioxidant transport using probenecid (an anion-transport inhibitor) in dogs
8 exposed to 0.2 ppm in a 6h exposure. Blocking antioxidant transport caused heterogeneously
9 distributed increases in peripheral airway resistance and reactivity, supporting the hypothesis
10 that in the lung periphery, endogenous antioxidants moderate the effects of O3 and that this
11 exposure is a subthreshold stimulus for producing effects on peripheral airway resistance and
12 reactivity in dogs. Treatment with probenecid also inhibited O3-induced neutrophilic
13 inflammation, which was present in untreated animals exposed to O3. This finding provides
14 evidence of a dissociation between airway function and inflammation, suggests that O3-induced
15 inflammation and AHR are independent phenomena, and further, that O3-induced neutrophilic
16 influx is dependent on a probenecid-sensitive transport process. The authors postulated that
17 probenecid has either a direct or indirect effect on either cytokine or leukotriene transport.
18 Probenecid treatment also caused a 50-60% decrease in plasma urate, a decrease in ascorbate,
19 and a decrease in BALF protein.
20 Mudway and Kelly (1998) modeled the interactions of O3 with three ELF antioxidants,
21 AA, uric acid and GSH. They used a continually mixed, interfacial exposure set up in perpex
22 chambers with O3 concentrations of 0, 0.1, 0.25, 0.5, 1.0 or 1.5 ppm. Exposures were carried out
23 with each oxidant individually, with the antioxidants as a mixture, and with and without human
24 albumin. In all three exposure conditions the ranking of reactivity with O3 was uric
25 acid > AA > GSH. The reactions did not cause changes in sample pH and no protein carbonyl
26 formations was observed with the antioxidants. They also observed consumption of the
27 antioxidants occurring in a linear fashion with time and a positive relationship to O3
28 concentration. They concluded that GSH is not an important substrate for O3, while uric acid
29 appeared to be the most important substrate which confers protection from O3 by removing it
30 from inhaled air and limiting the amount that reaches the distal lung. The authors acknowledge
December 2005 AX5-6 DRAFT-DO NOT QUOTE OR CITE
-------
1 limitations in extrapolating these data to in vivo O3 exposures due to the absence of surfactant
2 lipids and airway mucus in the model system.
3
4 AX5.2.1.4 Lipid Metabolism and Content of the Lung
5 One of the major postulated molecular mechanisms of action of O3 is peroxidation of
6 unsaturated fatty acids in the lung, prompting interest in measurement of lipids and lipid
7 metabolism. Several new studies have examined the effects of O3 exposure on phospholipid in
8 lung tissue.
9 A new mechanism for the toxicity of O3 was proposed by Pryor et al. (1995). This
10 mechanism suggests that the biochemical changes due to O3 exposure are relayed to deeper
11 tissue in the lung by a cascade of ozonation products. They consider lipid oxidation products as
12 the most likely molecules to do this because lipids are present in high concentrations in the ELF
13 and they react with O3 to form stable molecules. These lipid oxidation products cause activation
14 of specific Upases, which then trigger the activation of second messenger pathways (e.g.,
15 phospholipase A2 or phospholipase C). Experiments were completed by this group (Kafoury et
16 al., 1999) using exposures of cultured human bronchial epithelial (NHBE) cells to the lipid
17 ozonation product 1 -palmitoyl-2-(9-oxononanoyl)-sn-glycero-3-phosphocholine (PC-ALF) and
18 1-hydroxy-l-hydroperoxynonane (HHP-C9). Measurements of PAF, PGE2, IL-6 and IL-8 were
19 completed . PC-ALF elicited release of platelet-activating factor (PAF) and prostaglandin E2,
20 but not IL-6. HHP-C9 caused release of PAF and IL-6 in these cells, but not prostaglandin E2.
21 These results suggest to the authors that O3-induced production of lipid ozonation products
22 causes release of proinflammatory mediators that then generate an early inflammatory response.
23 Long et al. (2001) exposed hamsters to 0.12, 1.0 or 3.0 ppm O3 to evaluate lipid
24 peroxidation and antioxidant depletion. After 6 h exposures to the two higher levels resulted in
25 increased BALF neutrophil numbers and F2-isoprostanes. The highest exposure only caused
26 increased levels of BALF urate and decreased plasma levels of ascorbate. Exposures to the
27 0.12 ppm had no effect on BALF neutrophils of F2-isoprostanes or on plasma antioxidants.
28 Exposures to 1.0 ppm O3 with 1 h of exercise caused increased levels of F2-isoprostanes.
29 Uhlson et al. (2002) evaluated the formation of oxidized phosphlipids by reacting O3 with
30 calf lung surfactant. Low levels of ozone (0.06, 0.125, and 0.25 ppm) for exposures of 2 to 48 h
31 created a dose- and time-dependent increase in the formation of l-palmitoyl-2-(9'-oxo-
December 2005 AX5-7 DRAFT-DO NOT QUOTE OR CITE
-------
1 nonanoyl)-glycerophosphocholine (16:Oa/9-al-GPCho), an oxidized phospholipid, which
2 possessed biological activity in three assays. The 16:Oa/9-al-GPCho: 1) decreased macrophage
3 viability by necrosis at 6 |iM, 2) induced apoptosis in pulmonary epithelial-like A549 cells at
4 100 to 200 |iM , and 3) elicited release of IL-8 from A549 cells at 50-100 |iM.
5
6 AX5.2.1.5 Protein Synthesis
7 One new study of the effects of O3 on protein synthesis involved an examination of the
8 time course of lung injury and changes in collagen content in rats exposed acutely or
9 subchronically to 0.4 ppm O3 (van Bree et al., 2001). They observed centriacinar thickening of
10 septa after 7 days of exposure. This progressed at 28 and 56 days of exposure. After 28 days of
11 O3, the increase in collagen content in ductular was apparent and it increased progressively until
12 the 56 daytime point. While collagen content decreased with PE recovery, the structural fibrotic
13 changes in ductular septa did not return to control levels. Additionally, they observed the
14 presence of respiratory bronchioles after O3 exposure, which persisted after an 80-day recovery
15 period. These data suggest that subchronic O3 exposures in rats creates a progression of
16 structural lung injury that can evolve to a more chronic form, which included fibrosis.
17
18 AX5.2.1.6 Gene Expression
19 Gohil et al. (2003) have used genomic technology to examine differential gene expression
20 in C57BL/6 mice exposed to 1 ppm O3 for three consecutive nights for 8 h. Utilizing the
21 Affymetrix GeneChip, found O3-induced changes in the expression of 260 genes, of which 80%
22 were repressed and 20% induced. A number of genes involved in progression of the cell cycle
23 were increased, including ribonucleotide reductase and S-adenosyl methionine decarboxylaseS.
24 Several NF-KB-activated genes were induced including inhibitor of apoptosis, platelet-derived
25 growth factor receptor a, monocyte chemoattractant protein 1, topoisomerase (DNA) II-a, and
26 serum amyloid 3. These genes are causatively linked with inflammation and in concert with the
27 induced cell cycle genes, may account for increased proliferation of Clara cells and Type II
28 pneumocytes. Ozone caused suppression in the expression of several genes involved in
29 xenobiotic metabolism and in genes coding for major histocompatibility complex. These data
30 suggest O3 exposure suppresses immune function and xenobiotic metabolism and enhances
31 cellular proliferation.
December 2005 AX5-8 DRAFT-DO NOT QUOTE OR CITE
-------
1 AX5.2.2 Lung Host Defenses
2 A number of defense mechanisms operate in the respiratory tract to protect the host from
3 infectious and neoplastic disease. In humans and in animals, the conducting airways of the lungs
4 are primarily protected by the mucociliary escalator. The mucus layer acts to entrap many
5 gaseous and particulate agents and they are cleared from the tract before they have the
6 opportunity to reach underlying tissues. Defects in mucociliary transport can be caused by
7 changing the chemical nature of the mucous secretions, by paralyzing the cilia, or by producing
8 focal lesions in the ciliated epithelium, making it more susceptible to toxic inhalants. Within the
9 gaseous exchange region of the lung, the first line of cellular defense against microbes and
10 nonviable particles is the alveolar macrophage (AM). Impairment of AM function would alter
11 the lung's capabilities to maintain sterility, to clear the lungs of inhaled particles phagocytosized
12 by these cells, to mount an immune response, and to release immunologically-active soluble
13 mediators. Such effects would reduce the host's ability to resist infection and may be involved
14 in the pathogenesis of other chronic diseases. In addition to AM, other local humoral and
15 cell-mediated immune responses are active in protecting the host against such infectious insults,
16 as well as tumor cells. Animal studies have shown that each one of the above defense systems
17 can be altered following exposure to O3. New studies are summarized in Table AX5-2 are
18 discussed below.
19
20 AX5.2.2.1 Clearance
21 Mucociliary Clearance
22 To ascertain the mechanism(s) by which O3 modulates uptake of particles, Churg et al.
23 (1996) prepared 2 mm SD rat tracheal explants and exposed them to either room air or 0.01,
24 0.05, 0.10, or 1.0 ppm O3 for 10 minutes. After the O3 exposure, explants were then submerged
25 in either 5 mg/ml amosite asbestos or 4 mg/ml titanium dioxide for 1 h. Uptake of particles
26 assayed 7 days later indicated a dose-dependent uptake of TiO2 starting at 0.01 ppm and uptake
27 of asbestos at the two highest doses. To understand the potential role of oxidative stress in this
28 uptake, in some experiments at doses of 0.1 ppm O3, the reactive oxygen species scavengers
29 catalase or superoxide dismutase, or the iron chelator deferoxamine were added. Uptake of both
30 particles was inhibited by deferoxamine or catalase, but not by superoxide dismutase. Based on
31 these results, the authors concluded that: (1) uptake of particles in the trachea is a direct effect of
December 2005 AX5-9 DRAFT-DO NOT QUOTE OR CITE
-------
o
O
to
O
O
Table AX5-2. Effects of Ozone on Lung Host Defenses
Concentration3 Duration
Species
Effects"
Reference
Microbiologic Endpoints
0.1, 0.3ppm
0.8 ppm
4 h/day,
5 days/week,
1 or 3 weeks
Rat
(F344)
3h
Rat
(F344)
No effect on cumulative mortality from subsequent lung infection with Cohen et al.
4-8 x 106 Listeria monocytogenes, but concentration-related effects on (2001, 2002)
morbidity onset and persistence. One-week exposed rats: listeric
burdens trended higher than in controls; 0.3 ppm rats displayed
continual burden increases and no onset of resolution; in situ IL-lct,
TNF-oc, and IFNy levels 48 and 96 h post-infection (4 x 106 level)
higher than controls. Three-week exposed rats: no O3-related change
in bacterial clearance; IL-la, TNFa, and IFNg levels higher than control
only at 48 h post-infection (4 * 106) and only with 0.3 ppm rats
Single exposure to S. zooepidemicus led to differential clearance Dong et al.
patterns in exposed rats maintained on ad libitum or O3-mitigating (1998)
calorie-restricted diets
X
0.128 to 1.0
24 h/day for
1 week
Rat
Rats infected with 3.8 x 108 Listeria monocytogenes (IT). Decreased clearance
of Listeria at highest exposure (1.0 ppm) and impaired cellular immune
response: decreased T/B cell ratios in lymph nodes; delayed-type
hypersensitivity response to Listeria-antigen; and depressed lympho-
proliferation response in spleen and lymph nodes. Increased formation of
lung granulomas.
Steerenberg
etal. (1996)
H
6
o
o
H
O
Clearance Endpoints (Non-Microbial)
0.01-1.0 ppm lOmin
0.4 ppm
6h
Rat
(SD)
Dog
Single 10 min exposure of trachea! explants, followed by 1 h incubation Churg et al.
with particles, led to dose-related increases in uptake of amosite (1996)
asbestos and titanium dioxide particles. Effect inhibited by added
catalase or desferoxamine, but not by superoxide dismutase.
Increased trachea! permeability to 99mTc-DTPA after direct sublobar Foster and
exposure to O3. Clearance halftimes remained significantly lower for Freed (1999)
1-7 d PE, but recovered by 14 days PE.
o
HH
H
W
-------
o
O
to
O
O
Table AX5-2 (cont'd). Effects of Ozone on Lung Host Defenses
>
X
H
6
o
o
H
O
Concentration3
Alveolar Macrophage
0.8 ppm
0.8 ppm
1.0 ppm
2.0 ppm
Alveolar Macrophage
0.1, 0.3 ppm
0.1, 0.3 ppm
0.3 ppm
0.8 ppm
0.8 ppm
Duration
Endpoints (General)
3h
3h
Species
Rat (SD)
Rat (SD)
4 h Mouse cell line
3h
Endpoints (Functional)
4 h/day,
5 days/week,
1 or 3 weeks
4 h/day,
5 days/week,
1 or 3 weeks
5 h/day,
5 days/week,
4 weeks
3h
3h
Rat (SD)
Rat (F344)
Rat (F344)
Rat (F344)
Rat (SD)
Rat (F344)
Effects"
Increased ex vivo AM adherence to epithelial cultures mitigated by cell
pretreatment with anti-CD lib or anti-ICAM-1 antibodies.
Increased ex vivo AM adherence to epithelial cell cultures mitigated by
cell pretreatment with anti-TNF-cc/IL-la antibodies.
Increased intracellular calcium resting levels in WEHI-3 cells.
Decreased rates of calcium influx due to digitonin.
Decreased AM reduced glutathione content. Effect blocked by
pretreatment with bacterial endotoxin.
Superoxide anion: increased AM production (1 week; 0.1,0.3 ppm);
no intergroup differences noted after IFNy stimulation. H2O2: reduced
production (1 week; 0. 1, 0.3 ppm); further reduced production after
treatment with IFNy (0.1, 0.3 ppm, 1 and 3 weeks).
Increased AM superoxide anion production (1 week; 0.1, 0.3 ppm),
Lower H2O2 production (1 week; 0.1,0.3 ppm). Reduced production
after treatment with IFNy - superoxide (0.3 ppm, 1 week) and H2O2
(0.1 ppm, 1 week) - relative to cells without IFNy treatment. No effects
from 3 -week exposures.
No effect on AM endotoxin-stimulated IL-loc, IL-6, or
TNF-oc production. Decrease in stimulated, but not spontaneous,
superoxide formation; variable effects on H2O2 formation. No effect
on AM spontaneous, endotoxin-, or IFNy -stimulated, NO formation.
Increased AM motility in response to chemotaxin; effect mitigated by
cell pretreatment with anti-CD 1 Ib or anti-ICAM-1 antibodies.
Decrease in AM phagocytic activity
Reference
Bhalla (1996)
Pearson and
Bhalla (1997)
Cohen et al.
(1996)
Pendino et al.
(1996)
Cohen et al.
(2001)
Cohen et al.
(2002)
Cohen et al.
(1998)
Bhalla (1996)
Dong et al.
(1998)
O
HH
H
W
-------
O
O
1
1-1
to
O
O
:x!
rN
1
K>
o
c
H
6
o
0
H
O
Concentration3 Duration
Table AX5-2 (cont'd).
Species
Effects of Ozone on Lung Host Defenses
Effects"
Reference
Alveolar Macrophage Endpoints (Functional) (cont'd)
0.8 ppm 3 h
1.0 ppm 24 h/day,
3 days
Cytokines, Chemokines: Production,
0.1, 0.3 ppm 4 h/day,
5 days/week,
1 or 3 weeks
0.1, 0.3 ppm 4 h/day,
5 days/week,
1 or 3 weeks
0.3 ppm 5 h/day,
5 days/week,
4 weeks
0.3 ppm 24 or 96 h
1.0 ppm 1, 2, or4h, 2,
2.5 ppm 4, or 24 h
Mouse (B6J129SV)
(C57/BL6X 129 NOS"'")
Rat (Wistar)
Increased AM spontaneous and IFNy+LPS-induced NOS expression
and NO production and PGE2 release. Initial decrease in ROI
production, with eventual rebound. Knockout (NOS"'") mice AM
incapable of similar response to O3 - no inducible NO or PGE2 above
control levels and consistent decreased ROI production
Lavage fluid from exposed rats subsequently inhibited IFNy -induced
AM NO production.
Fakhrzadeh
et al. (2002)
Koike et al.
(1998, 1999)
Binding, and Inducible Endpoints
Rat (F344)
Rat (F344)
Rat (F344)
Mouse (C57B1/6J)
Superoxide anion: no intergroup differences noted after IFNy
stimulation. H2O2: reduced production after treatment with IFNy.
Decreased expression of CDS among lung lymphocytes (0.1 ppm only;
3 weeks); effect exacerbated by stimulation with IFNcc (but not with
IL-lcc). Decreased expression of CD25 (IL-2R) on CD3+ lymphocytes
(0.3 ppm only; 3 weeks); effect worsened by treatment with IL-lcc
(0.1, 0.3 ppm; 3 weeks). No effects on IL-2 -inducible lympho-
proliferation. Reduced AM production of ROIs after treatment with
IFNy; superoxide (0.3 ppm, 1 week) and H2O2 (0.1 ppm, 1 week) -
relative to untreated cells
No effect on AM endotoxin-stimulated IL-lcc, IL-6, or
TNF-cc production.
0.3 ppm: Increased lung: MIP-2, MCP-1, and eotaxin mRNA
expression.
1.0 ppm: After 4 h, increased lung: MIP-2, MCP-1, eotaxin, and
TL-6 mRNA exnression.
Cohen et al.
(2001)
Cohen et al.
(2002)
Cohen et al.
(1998)
Johnston et al.
(1999a)
2.5 ppm: After 2 h, increased lung: MIP-2, MCP-1, eotaxin, and
IL-6 mRNA expression.
No exposure-related increases in lung IL-lcc, IL-lp, IL-lRcc, IL-10,
IL-12, or IFNy mRNA expression.
O
HH
H
W
-------
Table AX5-2 (cont'd). Effects of Ozone on Lung Host Defenses
V 1
(J '
CD
to
O
o
X
OJ
O
H
6
o
0
H
O
Concentration3 Duration Species
Cytokines, Chemokines: Production, Binding, and Inducible
1.0 ppm 4h Mouse cell line
1.0 ppm 6h Rat(SD)
1.0 ppm 24h/day, Rat(Wistar)
3 days
1.0 ppm 24 h Mouse (C57B1/6J)
1.0 ppm 24 h Mouse (C57B1/6J)
1.0 ppm 8 h/day, Mouse (C57B1/6)
3 days (C57Bl/6Ai' NOS^)
1.0, 2.5 ppm 4 or 24 h Mouse (C57B1/6J)
Effects"
Endpoints (cont'd)
Decreased binding of IFNy by WEHI-3 cells. Decreased superoxide
production by IFNy -treated cells; no similar effect on H2O2 production.
Decreased IFNy-stimulatable phagocytic activity. No effect on IFNy-
inducible la (MHC Class II) antigen expression.
Increased AM MIP-lcc, CINC, TNF-cc, and IL-lp mRNA expression.
Induced increase in MIP-lcc and CINC mRNA temporally inhibited by
cell treatment with anti-TNF-cc/IL-lp antibodies.
Lavage fluid from exposed rats subsequently inhibited: ConA-
stimulated lymphocyte IFNy production, but had no effect on IL-2
production; IL-2 -induced lymphopro-liferation; and, IFNy -induced
AM NO production.
Increased lung: MIP-2 (4 h PE) and MCP-1 (4 and 24 h PE) mRNA
expression.
Increased lung MIP-2 and MCP-1 mRNA expression (4 and 24 h PE);
no effects on mRNA levels of IL-lcc, IL-lp, IL-lRcc, IL-6, MIF,
MIP-lcc, MIP-lp, eotaxin, or RANTES at either timepoint in recovery
period. Enhanced expressions of some cytokines/chemokines were
maintained longer than normal by coexposure to endotoxin.
Knockout (NOS"'") mice have more lavageable MIP-2 after exposure
than wild-type; both greater than control.
Dose-related increases in cytokine/chemokine induction. Increased
lung MIP-lcc, MIP-2, eotaxin (4 and 24 h), IL-6 (4 h only), and iNOS
mRNA expression.
Reference
Cohen et al.
(1996)
Ishii et al.
(1997)
Koike et al.
(1998, 1999)
Johnston et al.
(2001)
Johnston et al.
(2002)
Kenyon et al.
(2002)
Johnston et al.
(2000)
O
HH
H
W
-------
Table AX5-2 (cont'd). Effects of Ozone on Lung Host Defenses
(J '
CD
to
o
o
X
O
H
6
o
0
H
O
o
s
0
o
H
W
Concentration3 Duration Species
Effects"
Cytokines, Chemokines: Production, Binding, and Inducible Endpoints (cont'd)
0.6, 2.0 ppm 3h Mouse (C57BL/6) Increased lung MIP-2 (4 hPE) and MCP-1 mRNA expression (24 h
and Rat PE); PMN and monocyte increased accumulation in lungs consistent
with sequential expression of the chemokines. NF-kB activation also
increased 20-24 h PE.
0.8 ppm 3 h Rat (SD)
0.8 ppm 3 h Mouse (B6J129SV)
(C57B1/6X 129 NOS"'")
0.8 ppm 3 h Mouse (B6J129SV)
(C57B1/6X 129 NOS"'")
2.0 ppm 3 h Rat (SD)
180-500 ug/m3 and BALB/C
1% OVA C57BL/6
Alveolar Macrophage/Lung NO- and iNOS-Related Endpoints
0.3 ppm 5 h/day, Rat (F344)
5d/week,
4 weeks
0.8 ppm 3 h Mouse (B6J129SV)
(C57B1/6X 129 NOS"'")
Increased ex vivo AM adherence to epithelial cells mitigated by cell
treatment with anti-TNF-a or IL-la antibodies.
Increased AM IFNy+LPS-induced NOS expression and NO production,
as well as induced PGE2 release. Knockout (NOS"'") mice AM incapable
of similar response to O3 - no inducible NO or PGE2 above control
levels.
Increased AM IFNy+LPS-induced NOS expression and NO production.
Increased AM spontaneous and IFNy+LPS-induced NOS expression
and NO production. AM from exposed rats showed rapid
onset/prolonged activation of NF-KB.
O3 - dose-dependent increases in IgE, IL-4, IL-5; recruitment of
eosinophils and lymphocytes in BALB/c; O3 + OVA - increased IgG,
antibody liters, leukotrienes, airway responsiveness, immediate
culaneous hypersensitivity reactions in BALB/c. In C58BL/6 only
O3 + OVA caused culaneous hypersensitivity and altered IgG responses.
No effecl on AM spontaneous, endotoxin-, or IFNy -stimulated,
NO formation.
Increased AM IFNy+LPS-induced NOS expression and NO production
and PGE2 release. Knockoul (NOS"'") mice AM incapable of similar
response to O3 - no inducible NO or PGE2 above conlrol levels
Reference
Zhao el al.
(1998)
Pearson and
Bhalla (1997)
Fakhrzadeh
el al. (2002)
Laskin el al.
(2002)
Laskin el al.
(1998a)
Neuhaus-
Sleinmelz el al.
(2000)
Cohen el al.
(1998)
Fakhrzadeh
el al. (2002)
-------
Table AX5-2 (cont'd). Effects of Ozone on Lung Host Defenses
V 1
(J '
CD
to
o
o
X
O
H
6
o
0
H
O
O
0
o
H
W
Concentration3 Duration Species
Alveolar Macrophage/Lung NO- and iNOS-Related Endpoints
0.8 ppm 3 h Mouse (B6J129SV)
(C57B1/6X 129 NOS"'")
1.0 ppm 8 h/day, Mouse (C57B1/6)
3 days (C57Bl/6Ai" NOS"'")
1.0 ppm 24 h/day, Rat(Wistar)
3 days
1.0, 2.5 ppm 4 or 24 h Mouse (C57B1/6J)
2.0 ppm 3 h Rat (SD)
2.0 ppm 3 h Rat (SD)
3.0 ppm 6h Rat (Brown Norway)
0.12, 0.5, or 2 ppm 3 h Mice BALB/c
Effects"
(cont'd)
Increased AM spontaneous and IFNy+LPS-induced NOS expression
and NO production. AM from exposed mice showed rapid and
prolonged activation of NF-KB, STAT-1 (expression, activity),
phosphoinositide 3-kinase, and protein kinase B.
Knockout (NOS"'") mice have more lavageable PMN, MIP-2, and
protein in lungs after exposure than wild-type.
Lavage fluid from exposed rats subsequently inhibited IFNy -induced
AM NO production.
Dose-related increase in lung iNOS mRNA expression
Increased AM spontaneous, IFNy, and LPS-induced NO production,
as well as spontaneous and LPS-induced NOS expression. Effect
somewhat ameliorated by pretreatment with bacterial endotoxin.
Increased AM spontaneous and IFNy+LPS-induced NOS expression
and NO production. AM from exposed rats showed rapid
onset/prolonged activation of NF-KB.
Increased lung iNOS mRNA expression. Effect blocked by
pretreatment with dexamethasone.
Dose-dependent increases in nitrate and Penh; increases in nNOS but
not iNOS or eNOS.
Reference
Laskin et al.
(2002)
Kenyon et al.
(2002)
Koike et al.
(1998, 1999)
Johnston et al.
(2000)
Pendino et al.
(1996)
Laskin et al.
(1998b)
Haddad et al.
(1995)
Jang et al.
(2002)
-------
Table AX5-2 (cont'd). Effects of Ozone on Lung Host Defenses
(J '
CD
to
O
O
X
Oi
Concentration3
Duration
Surface Marker-Related Endpoints
0.8 ppm 3 h
1.0 ppm
1.0 ppm
1.0 ppm
0.1, 0.3 ppm
4h
2h
3 days
4 h/day,
5 days/week,
1 or 3 weeks
Species
Rat (SD)
Mouse cell line
Rat (SD)
Rat (Wistar)
Rat (F344)
Effects"
Increased expression of AM CD lib, but no effect on ICAM-1
No effect on IFNy-inducible la (MHC Class II) antigen expression on
WEHI-3 cells.
Decreased expression of integrins CD 18 on AM and CD1 Ib on PMN.
No effect on PMN CD62L selectin.
Increased expression of surface markers associated with antigen
presentation: la (MHC Class II) antigen, B7.1, B7.2, and CDllb/c on
BAL cells. Effect attributed to influx of monocytes.
Decreased expression of CDS among lung lymphocytes (0.1 ppm only;
3 weeks); effect exacerbated by stimulation of cells with IFNcc (but not
with IL-lcc). Decreased expression of CD25 (IL-2R) on CD3+
lymphocytes (0.3 ppm only; 3 weeks); effect worsened by treatment of
cells with IL-lcc (0.1 and 0.3 ppm; 3 weeks).
Reference
Bhalla (1996)
Cohen et al.
(1996)
Hoffer et al.
(1999)
Koike et al.
(2001)
Cohen et al.
(2002)
NK- and Lymphocyte-Related Endpoints
0.1, 0.3 ppm
4 h/day,
5 days/week,
1 or 3 weeks
Rat (F344)
Decreased expression of CDS among lung lymphocytes (0.1 ppm only;
3 weeks); effect exacerbated by stimulation of cells with IFNcc (but not
with IL-lcc). Decreased expression of CD25 (IL-2R) on CD3+
Cohen et al.
(2002)
lymphocytes (0.3 ppm only; 3 weeks); effect worsened by treatment of
cells with IL-lcc (0.1 and 0.3 ppm; 3 weeks). Lymphoproliferation: no
H
O
§ 0.4, 0.8, 1.6 ppm 12 h
H
O
C 1.0 ppm 24 h/day,
2 3 days
W
O
H
W
VAAWL \JAA OJJWAALCIAAVW LIO \JA AA_/-ZJ-AAAUUX-AU'AV_< AWAAAAO, \J . A JJJJAAA AAAVAVO.OVU.
response to ConA mitogen (1 week only); 0.3 ppm - decreased response
to ConA (1 week only).
Mouse (Balb/c) Decreased pulmonary delayed-type hypersensitivity reactions to low
MW agents, likely via activation of TH2 -dependent pathways.
Rat (Wistar) Lavage fluid from exposed rats subsequently inhibited ConA-stimulated
lymphocyte IFNy production, but had no effect on IL-2 production;
material also inhibited IL-2 -induced lymphoproliferation.
Garssen et al.
(1997)
Koike et al.
(1999)
-------
Table AX5-2 (cont'd). Effects of Ozone on Lung Host Defenses
(J '
CD
to
O
O
Concentration3 Duration
Susceptibility Factors
0.3 ppm 24 to 72 h
1 ppm 4 h
10 ppm until death
Species
Mice
C57BL/6J
C3H/HeJ
C3H/HeOuJ
CHO-K1 cell line SP-A
Mice
A/J (O3 sensitive)
and C57BL/6J
(O3-resistant)
Effects"
Lavageable protein concentration lowered by inhibition of iNOS and by
targeted disruption of Nos2; reduced Nos2 and Tlr4 mRNA levels in the
O3-resistant C3H/HeJ mice.
Differences exist biochemically and functionally in SP-A variants.
O3 exposure affects the ability of variants to stimulate TNF-cc and IL-8.
No differences in histology or wet-to-dry lung weights; two loci- acute
lung injury -1 or -2 confer susceptibility.
Reference
Kleeberget et al.
(200 Ib)
Wang et al.
(2002)
Prows et al.
(1999) and
Prows et al.
(1997)
X
Conversion to ug/m3 ~ ppm value x 1960
bCommon abbreviations used:
AM = Alveolar macrophage; PE = Postexposure (i.e., time after O3 exposure ceased); MIP = macrophage inflammatory protein;
PMN = Polymorphonuclear leukocyte; MLN = Mediastinal lymph node; CINC = cytokine-induce neutrophil chemoattractant;
BAL = Bronchoalveolar lavage; DTPA = diethylenetriaminepentaacetic acid; ROI = reactive oxygen intermediate/superoxide anion;
IFN = Interferon; BALT = bronchus-associated lymphoid tissue; MCP = monocyte chemoattractant protein; CON A = Concanavalin A
H
6
O
O
H
O
O
HH
H
W
-------
1 O3, as inflammatory cells are not present in the explants, (2) reactive oxygen species are
2 important mediators of particle uptake, with hydrogen peroxide having a primary role in the
3 process, and (3) due to the protective effect of deferoxamine, hydroxyl radical is probably
4 involved in the uptake also.
5 Pearson and Bhalla (1997) have utilized the radiolabled chelate 99mTc diethylenetriamine
6 pentaacetic acid (Tc-DTPA) to assess the effect of O3 exposure on clearance across epithelial
7 surfaces. 99mTc-DTPA clearance has been found to be significantly increased following a 3-h
8 exposure to 0.8 ppm O3 in SD rats. Pretreatment with anti-IL-la and anti-TNF-a did not affect
9 the permeability, suggesting that these soluble mediators are not involved in this process. Foster
10 and Freed (1999) also used 99mTc-DTPA to examine regional clearance in dogs following a 6-h
11 isolated sublobar exposure to 0.4 ppm O3 or air. Ozone decreased the clearance halftime of
12 99mTc-DTPA by 50% at 1 day following exposure. Seven days PE the halftime was still reduced
13 by 29% and by 14 days PE, clearance had recovered to normal levels. These data provide
14 evidence that a single local exposure to O3 increases transepithelial clearance, but without any
15 influence on contralateral segements, i.e., only for epithelia directly exposed to O3, and the
16 altered permeability changes recover to normal levels in 2 weeks.
17
18 Alveolar Clearance
19 New evaluations of the effects of O3 on alveolar clearance have not been performed.
20
21 AX5.2.2.2 Alveolar Macrophages
22 Effects of O3 on other AM functions are summarized on Table AX5-2 and new studies are
23 discussed here.
24 Dong et al. (1998) reported that caloric restriction enhanced phagocytic function in
25 O3-exposed rats. Whereas ad-libitum fed rats had a prolonged infection and pulmonary
26 inflammation from a Streptococcus challenge, calorie restricted rats had no infection and no
27 inflammation.
28 Ozone exposure has been implicated in altered chemotaxis and cell adhesion properties of
29 AM. Bhalla (1996) reported that macrophages isolated from O3-exposed SD rats (0.8 ppm O3 or
30 air, nose-only, for 3 h and then examined AM from BALF at 12 h PE) showed greater mobility
31 and greater adhesion than AM isolated from air-exposed rats. This increased mobility and
December 2005 AX5-18 DRAFT-DO NOT QUOTE OR CITE
-------
1 adhesion were attenuated when AMs were incubated with monoclonal antibodies to CD1 Ib
2 (leucocyte adhesion molecules) or ICAM-1 (epithelial cell adhesion molecules). The authors
3 suggest that these observed changes in basic cell surface-associated macrophage properties are
4 relevant to subsequent O3-induced lung inflammatory responses. Using the same O3-exposure
5 parameters, Pearson and Bhalla (1997) also observed that the increased AM adherence to
6 epithelial cell cultures induced by O3 exposure was found to be mitigated by pretreatment with
7 antibodies to TNF-a and IL-la, suggesting that the early inflammatory response to O3, in part,
8 may be mediated by IL-la and/or TNF-a.
9 Additional studies have been carried out to characterize the mechanisms by which O3
10 induces decreased lung resistance against microbial pathogens. Cohen et al. (1996) have
11 exposed the WEHI-3 cell line, a BALB/c myelomonocytic AM-like cell, to 1 ppm O3 for 4 h to
12 determine the effects of O3 on AM activation by interferon-y (INF y). Ozone at this
13 concentration reduced binding of INF y to AM and affected the AM functional parameters of
14 phagocytic activity, production of reactive oxygen intermediates, and elevation of intracellular
15 calcium. Further,O3 increased intracellular calcium resting levels and decreased the rates of
16 calcium influx due to digitonin. The authors concluded that this O3-induced modulation of AM
17 function could be responsible for the increased microbial pathogen survival following O3
18 exposure. Pendino et al. (1996) studied the role of glutathione content in AM functions. When
19 BALF recovered from female SD rats exposed to 2 ppm or air for 3 h was assayed for
20 intracellular glutathione (with the fluorescent indicator dye monochlorobimane [MCB]),
21 indicated that AM from O3-exposed rats had reduced levels of intracellular glutathione compared
22 to air-exposed rats. This reduction in glutathione levels may be due to its interaction with
23 ozonation products from O3-induced lipid peroxidation.
24 Bactericidal activity of AM is mediated by hydrogen peroxide production. To study the
25 effect of O3 on this response function of AM, Cohen et al. (2001,2002) exposed male F-344 rats
26 to either 0.1 or 0.3 ppm O3 for 4 h/day, 5 days/week or either 1 or 3 weeks and assessed
27 superoxide anion and hydrogen peroxide production in AM recovered from BAL 24 h PE. They
28 found increased superoxide anion production at 1 week 0.1 and 0.3 ppm exposure and did not
29 observe any intergroup differences when stimulated by INFy. Conversely, hydrogen peroxide
30 production was reduced at both exposure concentrations and durations and was further reduced
December 2005 AX5-19 DRAFT-DO NOT QUOTE OR CITE
-------
1 with INFy stimulation. The authors suggest that the compromised killing of bacteria by AM in
2 O3-exposed rats may be due to the reduction in hydrogen peroxide production.
3 Laskin et al. (1998a) have examined the activation of AM and type II epithelial cells in
4 female SD rats exposed to 2 ppm O3 for 3 h. Ozone treatment caused a time-dependent increase
5 in NO levels in both cell types that was correlated with increased expression of iNOS mRNA
6 and protein. Laskin et al (1998b) hypothesized that the inflammatory mediators such as TNF-a
7 and IL-lb may mediate the increase in NO release by activating the expression of iNOS through
8 NF-KB signaling. They demonstrated it by treating the cells with pyrrolidine dithiocarbamate,
9 an inhibitor of NF-KB that caused a dose dependent inhibition of NO production and iNOS
10 expression. This group (Laskin et al., 2002) further investigated the mechanisms by which O3
11 activates AM using C57B16xl29 mice with a targeted disruption of the gene for iNOS. These
12 mice exposed to 0.8 ppm O3 for 3 h showed no toxicity as measured by BALF protein levels and
13 nitrotyrosine staining of the lung. Additionally, mice overexpressing human Cu, Zn superoxide
14 dismutase (SOD) and mice with a targeted disruption of p50 NF-KB were also resistant to O3
15 toxicity. Wild-type mice exposed to O3 showed an increase in expression of STAT-1, a protein
16 that binds to the regulatory region of iNOS. Taken together, these results suggest to the authors
17 that a number of proteins including NF-KB, phosphoinoside 3-kinase, and STAT-1 that bind to
18 and regulate expression of iNOS are modulated by O3 exposure. Another study by this group
19 (Fakhrzadeh et al., 2002) used the same iNOS knockout mice strain to further characterize O3
20 toxicity. In wild-type mice O3 exposure causes an increase in AM superoxide anion and
21 prostaglandin (PG)E2, but in the knockouts, the reactive nitrogen intermediates were not
22 produced and (PG)E2 was at control levels. Further discussions of the role of nitric oxide
23 synthase and reactive nitrogen in O3-induced inflammation are contained in Section 5.2.3.5.
24 Additionally, cytokines and chemokines are very important components of the AM response to
25 O3 and are discussed in detail in Section 5.2.3.4.
26
27 AX5.2.2.3 Immune System
28 Other than by natural protection (e.g., opsonizing antibody, nonspecific phagocytosis by
29 AM), the immune system defends the lung by mounting three major waves of response: natural
30 killer (NK) cells (nonspecific lymphocytes that kill viruses, bacteria, and tumor cells), followed
31 by cytotoxic T-lymphocytes (TCTL- lymphocytes that lyse specifically recognized microbial and
December 2005 AX5-20 DRAFT-DO NOT QUOTE OR CITE
-------
1 tumor-cell targets), followed by antigen-specific antibodies. These T-cell types are involved
2 with other immunologically active cells (e.g., B-cells and AM), which in a complex manner,
3 interact in immunological defense. To date, only a few of these mechanisms have been
4 investigated in the context of their role in O3 susceptibility. Effects on the systemic immune
5 system can be different from those in the lung (see Section 5.3). New studies reporting
6 O3-induced effects on the immune system in the lung are described here.
7 Garssen et al. (1997) have studied the effects of O3 on non-IgE-mediated pulmonary hyper-
8 immune reactions induced by picryl chloride (PCI). BALB/c mice sensitized with PCI, both
9 actively and passively (by adoptive transfer of lymphoid cells from pre-sensitized mice), were
10 then challenged with picryl sulfonic acid (PSA). The mice were exposed to!2 h of 0.4, 0.8, or
11 1.6 mg/m3 O3 during one night, at 4 days or 7 days after skin sensitization (which was either just
12 before or just after PSA challenge, i.e., during the induction or effector phase). Non-sensitized
13 mice showed no changes in tracheal reactivity to carbacol with O3 exposure. Sensitized mice
14 were hyperreactive to carbachol 48 h after PSA challenge, whereas sensitized mice exposed to
15 all concentrations of O3 showed no significant tracheal hyperreactivity to carbachol. The
16 sensitized mice also demonstrated a suppressed inflammatory reaction (PMN) with 1.6 mg O3
17 exposure. O3 exposure following PSA challenge also caused a suppression of tracheal
18 hyperresponsiveness. In a separate experiment wherein mice were exposed to O3 before
19 sensitization and then lymphoid cells from these mice were injected into non-exposed mice, the
20 recipients also demonstrated an inhibition of the induction of hyperreactivity. These results are
21 opposite to the effect on type I (IgE-mediated) allergic reactions, which the authors suggest is
22 due to activation of Th-2 cell-dependent reactions that are possibly potentiated by O3 or to a
23 direct effect by O3 on Th-1 cells or other cells that are crucial for the tracheal hyperreactivity and
24 inflammation seen in this mouse model.
25 Recent evidence also point towards the potential interaction between the innate and
26 acquired immune system with O3 exposure. Kleeberger et al. (2000) performed a genome screen
27 on O3-susceptible (C57BL/6J) and O3-resistant (C3H/HeJ) mice and identified a candidate gene
28 on chromosome 4, Toll-like receptor 4 (Tlr4), a gene implicated in endotoxin susceptibility and
29 innate immunity. When O3-resistant strain C3H/HeJ and C3H/HeOuJ (differing from the O3-
30 resistant strain by a polymorphism in the coding region of Tlr4) were exposed to 0.3 ppm for 24
31 to 72 h, greater protein concentrations were demonstrated in the OuJ strain. The two strains
December 2005 AX5-21 DRAFT-DO NOT QUOTE OR CITE
-------
1 exhibited differential expression of Tlr4 mRNA with O3 exposure. These data point to a
2 quantitative trait locus on chromosome 4 as being responsible for a significant portion of the
3 genetic variance in O3 -induced lung hyperpermeability. Further investigation by this laboratory
4 (Kleeberger et al., 2001a) using these mouse strains showed lavageable protein concentration
5 was lowered by inhibition of inducible nitric oxide synthase (iNOS) and by targeted disruption
6 ofNos2. Further comparisons on O3 exposure in these two strains (C3FI/HeJ and C3H/HeOuJ)
7 demonstrated reduced Nos2 and Tlr4 mRNA levels in the O3-resistant C3FI/HeJ mice. These
8 data are consistent with the hypothesis that O3 -induced lung hyperpermeability is mediated by
9 iNOS . These studies also suggested a role for toll-like receptor 4 (TLR4) in the host response to
10 O3 similar to the role it has demonstrated in LPS sensitivity (Schwartz, 2002; Wells et al., 2003).
11 TLR4 signaling is thought to be critical to linking the innate and acquired immune system
12 through antigen presenting cells and Thl/Th2 differentiation.
13 Neuhaus-Steinmetz et al. (2000) compared the response to repeated O3 (180-500 |ig/m3)
14 and OVA (1%) exposure in "IgE-high responder" (BALB/c) and "IgE-low responder "
15 (C57BL/6) mice. In BALB/c mice exposed to O3, a T-helper (Th)2-like response consisting of
16 dose-dependent increases in IgE, IL-4, IL-5, and recruitment of eosinophils and lymphocytes
17 into airways was generated. Concurrent O3/OVA exposures in BALB/c mice increased IgG,
18 antibody liters, leukotrienes, airway responsiveness, and immediate cutaneous hypersensitivity
19 reactions. In C57BL/6 mice only the combined O3/OVA exposure caused immediate cutaneous
20 hypersensitivity and altered IgG responses, thus demonstrating that O3 has the potential for
21 shifting the immune response toward a Th2-like pattern in two mouse strains with differing
22 potentials for developing allergic reactions. Becker et al. (1991) have demonstrated changes in
23 IgG production with O3 exposures of 1.0, 0.5, and 0.1 ppm for 2 h in vitro with human
24 lymphocytes. Subsequent to O3 exposure, when lymphocytes were stimulated with pokeweed
25 mitogen (PWM, a T-cell-dependent stimulus) or Staphylococcus aureus Cowan 1 strain (SAC, a
26 T-cell-independent stimulus), both B and T cells were found to be affected by O3 preexposure.
27 T cells also demonstrated an increase in IL-6 and a decrease in IL-2, suggesting that O3 may
28 have direct effects on IgG-producing cells and concurrently an effect that is mediated by altered
29 production of T cell immunoregulatory molecules.
30
December 2005 AX5-22 DRAFT-DO NOT QUOTE OR CITE
-------
1 Surfactant protein A and D (SP-A and SP-D) are members of the collectin family, named
2 for their composition of both collagens and lectins. The surfactant proteins are secreted by
3 airway epithelial cells and are a part of the innate immune response with important
4 immunomodulatory function. Perturbations in lung immune defenses lead to a feedback loop
5 between inflammation and SP-A and SP-D levels (reviewed in Hawgood and Poulain, 2001).
6 Costing et al. (1991) investigated whether surfactant protein A (SP-A) is a target of O3 toxicity
7 by exposing human and canine SP-A to 0.75 ppm for 4 h. Functionally the in vitro exposure
8 inhibited SP-A self-association and SP-A-mediated lipid vesicle aggregation. Additionally SP-A
9 decreased binding of SP-A to mannose. Structurally, O3 oxidized tryptophan and methionine
10 residues on the protein. Additional work by Costing et al. (1992) further examined the effect of
11 O3 on the function of SP-A from dog and human. In vitro O3 exposure at 0.4 or 0.75 ppm for 4 h
12 reduced the ability of SP-A to inhibit phospholipid secretion by alveolar type II cells and
13 reduced the capacity of SP-A to induce superoxide anion production and enhance phagocytosis
14 of herpes simplex virus (HSV). In vivo exposures of rats (0.4 ppm for 12 h) generated SP-A less
15 capable of stimulating superoxide anion production by AM. These data suggested that inhibition
16 of interactions between SP-A and alveolar cells may be one mode of toxicity of O3. Bridges
17 et al. (2000) reported that both SP-A and SP-D directly protect surfactant phospholipids and
18 macrophages from oxidative damage. Both proteins were found to block accumulation of
19 TEARS and conjugated dienes generated during oxidation of surfactant lipids or low density
20 lipoprotein particles by a mechanism that does not involve metal chelation or oxidative
21 modification of the proteins. Wang et al. (2002) expressed human variants of PS-A in CHO cells
22 and then exposed the expressed protein to O3 (Ippm for 4 h). All of the eight SP-A variants
23 studied showed decreased ability to stimulate cytokine (TNF- and IL-8) production in THP-1
24 cells, a macrophage-like cell line. Each variant exhibited a unique time- and dose-dependent
25 pattern of stimulation of cytokine production with O3 exposure, suggesting their potential role in
26 underlying susceptibility to O3 toxicity. Targeted disruption of mouse SP- A and SP-D
27 (Hawgood et al, 2002) caused increases in bronchoalveolar lavage phospholipid, macrophage,
28 and protein through 24 weeks of age. Further, the deficient mice developed patchy lung
29 inflammation and air space enlargement consistent with emphysema.
30
December 2005 AX5-23 DRAFT-DO NOT QUOTE OR CITE
-------
1 AX5.2.2.4 Interaction with Infectious Microorganisms
2 Numerous investigations have sought to understand the effects of O3 on overall functioning
3 of host defense systems by challenging animals with infectious agents before, during, or after
4 exposure and observing the outcome, typically mortality. This work (summarized in
5 Table AX5-2) shows that the results are dependent on microbial species, animal species, and O3
6 exposure and its temporal relationship to infectious challenge.
7 Recent studies of O3-induced modulation of cell-mediated immune responses showed
8 effects on the onset and persistence of infection. Cohen et al. (2001,2002) exposed male F-344
9 rats to either 0.1 or 0.3 ppm O3 for 4 h/day, 5 days/week or either 1 or 3 weeks. One day later
10 the rats were instilled with viable Listeria monocytogenes (4 x 106 for 8 x 106) and then tested at
11 1, 58, 72, or 96 h postinfection. There was no observed effect on cumulative mortality, but there
12 was a concentration-related effect on morbidity onset and persistence. In the one week-exposed
13 rats the listeric burdens trended higher than in controls and the high dose rats showed continual
14 burden increased and without resolution. Levels of IL-la, TNF-a, and IFNy were higher than
15 controls at the 48 and 96 h time period. In the 3-week-exposed rats there were no changes in
16 bacterial clearance. Levels of IL-la, TNF-a, and IFNy were higher than controls only at the 48
17 h time period only in the 0.3 ppm-exposed rats. These observations suggest that exposure to O3
18 may be causing a possible imbalance between Th-1 and Th-2 cells, which can subsequently lead
19 to suppression of the resistance to intracellular pathogens. In this case, the host defense to
20 Listeria monocytogenes, which is predominantly a Th-1-type response, is adversely affected.
21
22 AX5.2.3 Inflammation and Lung Permeability Changes
23 Ozone has long been recognized to cause lung inflammation and increased permeability,
24 which are distinct events under control by independent mechanisms. The normal lung has an
25 effective barrier function that controls bidirectional flow of fluids and cells between the air and
26 blood compartments. Ozone disrupts this function, resulting in an increase in serum proteins,
27 bioactive mediators and polymorphonuclear leukocytes (PMNs) in the air spaces of the lung.
28 These inflammatory changes have been detected microscopically in tissue preparations and by
29 analyses of bronchoalveolar lavage fluid (BALF). Generally, the initiation of inflammation is
30 part of a defense process; however, its persistence and/or repeated occurrence can result in health
December 2005 AX5-24 DRAFT-DO NOT QUOTE OR CITE
-------
1 effects. For example, at the early stage, the increase in PMNs may enable more phagocytosis of
2 microbes, and increased edema fluids or protein associated with the PMN influx could provide
3 an enhanced medium for microbial growth, however, these are untested hypotheses. Mediators
4 in the influxing fluid and secreted from the increased number of PMNs can recruit other cells
5 (e.g., fibroblasts and AMs). If released, the large stores of proteolytic enzymes in these
6 phagocytic cells could damage lung tissue over the longer term. Cheek et al. (1994) suggest that
7 PMNs may also play a role in removal of O3-injured cells. The exact role of inflammation in
8 causation of lung disease is not known, nor is the relationship between inflammation and
9 changes in lung function. However, it is associated with acute changes in pulmonary function
10 and chronic diseases such as asthma, chronic bronchitis and emphysema. Table AX5-3
11 summarizes key new studies on the potential for O3 to increase lung permeability and to cause
12 inflammation.
13
14 AX5.2.3.1 Time Course of Inflammation and Lung Permeability Changes
15 Studies on the time course of the inflammatory response (e.g., Cheng et al., 1995) indicated
16 that the maximal increases in BALF protein, albumin and number of PMNs occur 8 to 18 h
17 (depending on the study) post acute exposure. In rats, a single 3-h exposure to 0.5 ppm O3
18 produced a significant increase in both permeability and inflammation, but a comparable
19 exposure to 0.3 or 0.15 ppm did not produce an effect (Bhalla and Hoffman, 1997). Dye et al.
20 (1999) reported that an acute exposure to 0.5 ppm O3 results in a significantly greater lung
21 injury, inflammation and BALF levels of IL-6 in Wistar rats than in Sprague Dawley or F344
22 rats. BALF cell count, PGE and IL-6 levels were consistently lower in F344 rats compared to
23 other strains. Sun et al. (1997) exposed OVA-sensitized male Dunkin-Hartley guinea pigs to 1
24 ppm O3 for 3 h and examined protein levels and PMN levels at 3 h PE. PMN levels were
25 significantly increased, without any change in BAL protein levels, suggesting a lack of
26 correlation between the two endpoints. When guinea pigs exposed to 1 ppm O3 for 1 h were
27 evaluated 24 h PE, they exhibited the same trend, increase in PMN without a concordant
28 increase in BAL protein levels. Increased AHR observed in the first group but not in the second
29 group suggest a dissociation between PMN levels and AHR.
30 Depending upon species tested and exposure regimens, continuous exposure for 3 to 7 days
31 resulted in an increase in BALF protein and PMNs that typically peak after a few days and return
December 2005 AX5-25 DRAFT-DO NOT QUOTE OR CITE
-------
December 2
o
o
.
X
to
Oi
O
^
fe
H
6
0
n
s.**
H
O
0
H
W
O
/^
Table AX5-3. Effects of Ozone on Lung Permeability and Inflammation
Concentration
ppm
0.1
0.2
0.5
0.1
0.2
0.4
1.0
0.2
0.4
0.8
0.26
0.3
0.1
0.3
1.0
Hg/m3
196
392
980
196
322
784
1,960
392
784
1,568
510
588
196
588
1,960
Duration Species
0.5 h Primary rat alveolar
in vitro type II cells
1 h, in vitro Guinea pig (Hartley) and
human alveolar
macrophages
23 h/day for 1 week Guinea pigs, F (Hartley),
260-330 g
8 h/day, 5 days/week Mice, M (mast cell-
for 1-90 days deficient and
-sufficient),
6-8 weeks old
48 hand 72 h. Mice, M (C57BL/6J and
Exposures repeated C3H/HeJ) 6-8 weeks old
after 14 days
60 min Rat basophilic leukemia
cell line (RBL-2H3)
Effects3
Decreased resistance (Rt ) after 0.5 ppm from 2 to 24 h PE
and at 48 h in monolayers subjected to PMNs. Significantly
lower Rt after PMN treatment at 0.2 and 0.5 ppm.
Exposure of guinea pig alveolar macrophages to 0.4 ppm for
60 minutes produced a significant increase in IL-6 and TNF-
cc, and an exposure of human alveolar macrophages to
identical O3 concentration increased TNF-oc, IL-lb, IL-6 and
IL-8 protein and mRNA expression.
Increase in BALF protein and albumin immediately after 0.8
ppm exposure, with no effect of ascorbate deficiency in diet.
O3 -induced increase in BALF PMN number was only
slightly augmented by ascorbate deficiency.
Greater increases in lavageable macrophages, epithelial cells
and PMNs in mast cell -sufficient and mast cell-deficient
mice repleted of mast cells than in mast cell-deficient mice.
Cyinduced permeability increase was not different in
genotypic groups.
Greater BALF protein, inflammatory cell and LDH response
in C57BL/6J than in C3H/HeJ after initial exposure.
Repeated exposure caused a smaller increase in BALF
protein and number of macrophages, lymphocytes and
epithelial cells in both strains, but PMN number was greater
in both strains of mice compared to initial exposure.
O3 inhibited IgE- and A23 187 - indued degranulation.
Spontaneous release of serotonin and modest generation of
PGD2 occurred only under conditions that caused
cytotoxicity.
Reference
Cheek et al.
(1995)
Arsalane et al.
(1995)
Kodavanti
etal. (1995)
Kleeberger
etal. (200 Ib)
Paquette et al.
(1994)
Peden and
Dailey (1995)
H
W
0.3
2.0
588
3,920
72 h
3h
Mice (C57BL/6J and
C3H/HeJ)
Greater PMN response in C57BL/6J than in C3H/HeJ after
acute and subacute exposures. Responses of recombinant
mice were discordant and suggested two distinct genes
controlling acute and subacute responses.
Tankersley and
Kleeberger,
(1994)
-------
o
O
to
O
O
Table AX5-3 (cont'd). Effects of Ozone on Lung Permeability and Inflammation
X
-------
December 1
o
o
^
X
to
oo
M
\^
%
H
1
O
0
!2
O
H
O
0
H
W
o
o
HH
—1
1 3
W
Table AX5-3 (cont'd). Effects of Ozone
Concentration
ppm Hg/m3 Duration Species
1.0-2.0 0-3,920 3h Mice (C57BL/6),
6-8 weeks old and
rats (Wistar),
14-16 weeks old
0.5 980 24 h following a 3 -day Mice, M (B6C3F1) 25 ±
(6 h/day) exposure to 2 g
cigarette smoke
0.5 980 8 h during nighttime Rat, M (Wistar, SD and
F344) 90 days old
0.8 1,568 2hand6h Rats, M (Fisher),
Juvenile (2 months;
180-250 g), Adult
(9 months; 370-420 g),
Old (18 months;
375-425 g), Senescent
(24 months; 400-450 g)
0.8 1,568 3h Rat, M (SD)
6-8 weeks old
0.8 1,568 3h Rat, M (SD)
6-8 weeks old
0.8 1,568 3h Rat, M (SD)
200-225 g
on Lung Permeability and Inflammation
Effects3
Steady state MCP-1 mRNA increase after 0.6 ppm, with
maximal increase after 2 ppm. After 2 ppm, MIP-2 mRNA
peaked at 4 h PE and MCP-1 mRNA peaked at 24 h PE.
BALF neutrophils and monocytes peaked at 24 and 72 h PE,
respectively. BALF MCP-1 activity induced by O3 was
inhibited by ananti-MCP-1 antibody.
BALF protein, neutrophils and lymphocytes were increased
in animals exposed to smoke and then to O3. Macrophages
from this group also responded with greater release of TNF-
a upon LPS stimulation
Exposure resulted in a significantly greater injury,
inflammation and BALF levels of IL-6 in Wistar than in SD
orF344 rats.
Comparable effect on the leakage of alveolar protein in rats
of different age groups, but a greater increase occurred in
interleukin-6 and N-acetyl-beta-D-glucosaminidase in
senescent animals than in juvenile and adult rats.
Increased adhesion of macrophages from exposed animals to
rat alveolar type II epithelial cells in culture. Treatment with
anti-TNF-a + anti-IL-la antibody decreased adhesion in
vitro, but not permeability in vivo
Increase in fibronectin protein in BALF and lung tissue, and
fibronectin mRNA in lung tissue. The increase produced by
O3 was amplified in animals pre-treated intra-tracheally with
rabbit serum to induce inflammation.
Treatment of animals with IL-10 prior to O3 exposure caused
a reduction in O3 induced BALF protein, albumin and
fibronectin and tissue fibronectin mRNA
Reference
Zhao et al.
(1998)
Yu et al.
(2002)
Dye et al.
(1999)
Vincent et al.
(1996).
Pearson and
Bhalla (1997)
Gupta et al.
(1998)
Reinhart et al.
(1999)
-------
o
O
to
O
O
Table AX5-3 (cont'd). Effects of Ozone on Lung Permeability and Inflammation
Concentration
ppm
Duration
Species
Effects3
Reference
X
to
VO
fe
H
6
o
o
H
O
O
H
W
O
O
HH
H
W
0.8
0.8
0.8
1.0
1.0
0.2
0.5
1.0
1,568
8h
0.8 1,568 48 h
1.0-2.0 1,960-3,920 3 h
1,568 8h
1,568 3 h
1,960
1,968
392
980
1,960
5 min exposure of
airway segments
following
bronchoscopy
8h
In vitro at liquid/air
interface
Monkey (Rhesus),
3 years 8 months-3 years
10 month old
Rat, M (SD)
6-8 weeks old
Monkeys, M (Rhesus),
3 years 8 months-3 years
10 months old
(5.1-7.6 kg)
Mice, F
(C57BL6X129NOSII
knockout and wild-type
B6J129SVF2)
Dogs, M (Mongrel)
Adult
Monkeys (Rhesus)
Primary TBE, BEAS-2b
SandHBEl
Pretreatment of monkeys with a monoclonal anti-CD 18 Hyde et al.
antibody resulted in a significant inhibition of O3-induced (1999)
neutrophil emigration and accumulation of necrotic airway
epithelial cells.
Cyclophosphamide treatment ameliorated O3-induced BALF Bassett et al.
neutrophils and albumin after short term and 1-day (2001)
exposure. Anti-neutrophil serum reduced lavageable
neutrophils but did not affect permeability.
Trachea! epithelium of exposed animals expressed b6 Hyde et al.
integrin. The integrin expression was reduced or (1999)
undetectable in animals treated with CD-18 antibody.
Alveolar macrophages from O3 exposed wild-type mice Fakhrzadeh
produced increased amounts of NO, peroxynitrite, et al. (2002)
superoxide anion, and PGE2. Nitrogen intermediates were
not produced and PGE2 was at control level in exposed
NOSH knockout mice. These mice were also protected from
O3-induced inflammation and injury.
Mast cells from O3-exposed airways of ascaris sensitive Spannhake
dogs released significantly less histamine and PGD2 (1996)
following in vitro challenge with ascaris antigen or calcium
ionophore.
Increase in steady state IL-8 mRNA in airway epithelium. Chang et al.
Increase in IL-8 protein staining declined at 24 h after (1998)
exposure.
Dose related increase in IL-8 release in the conditioned
media. Ozone produced greater toxicity in cell lines than in
primary cultures.
-------
o
O
to
O
O
Table AX5-3 (cont'd). Effects of Ozone on Lung Permeability and Inflammation
X
-------
o
O
to
O
O
Table AX5-3 (cont'd). Effects of Ozone on Lung Permeability and Inflammation
X
fe
H
6
o
o
H
O
O
H
W
O
O
HH
H
W
Concentration
ppm Hg/m3 Duration Species
1.0 1,960 8h/nightforthree Mice, (C57B1/6 wild-
nights type and iNOS
knockout)
1.0 1,960 4h Mice, M (129 strain,
wild-type and clara cell
secretory protein-
deficient), 2-3 mo old
1.2 2,352 6h Rat, M (BN), 200-250 g
2.0 3,920 3h Mice, M (C57BL/6J)
2.0 3,920 3h Rat, F (SD)
6-8 weeks old
2.0 3,920 3h Rat, F (SD)
6-8 weeks old
2.0 3,920 4h Rat, M (Wistar),
200-225 g
Effects3
O3 exposure produced greater injury, as determined by
measurement of MIP-2, matrix metalloproteinases, total
protein, cell content and tyrosine nitration of whole lung
protein, in iNOS knockout mice than in wild-type mice.
Increases in abundance of mRNAs encoding eotaxin, MIP-
la and MIP-2 in CCSP-/-, but no change in wild-type mice.
Eotaxin mRNA expression in the lungs increased 1.6-fold
immediately after and 4-fold at 20 h. Number of lavageable
eosinophils increased 3- and 15-fold respectively at these
time points. Alveolar macrophages and bronchial epithelial
cells stained positively for eotaxin.
O3 -induced increase in protein and PMNs in BALF, and
pulmonary epithelial cell proliferation were significantly
reduced in animals pre-treated with UK-74505, a platelet
activating factor-receptor antagonist.
BALF cells from exposed animals released 2 to 3 times
greater IL-1 and TNF-cc, and greater fibronectin.
Immunocytochemistry showed greater staining of these
mediators in lung tissue from exposed rats.
Increase in BALF macrophage number and total protein.
Increase in iNOS expression, and increase in Fibronectin
and TNF-cc production by alveolar macrophages. O3 effects
were reduced by pretreatment with gadolinium chloride, a
macrophage inhibitor.
A transient increase in tissue neutrophils correlated with an
elevation and subsequent decline in airway
hyperresponsiveness. Pretreatment of rats with anti-
neutrophil serum protected the animals from O3-induced
airway hyperresponsiveness
Reference
Kenyon et al.
(2002)
(Johnston et al.
(1999b)
Ishii et al.
(1998)
Longphre et al.
(1999)
Pendino et al.
(1994)
Pendino et al.
(1995)
DeLorme et al.
(2002)
-------
o
O
to
O
O
Table AX5-3 (cont'd). Effects of Ozone on Lung Permeability and Inflammation
X
-------
1 towards control even with continuing exposure. Van Bree et al. (2002) reported adaptation of
2 rats to O3 following 5 days of exposure. Animals exposed for 5 days had lower BALF proteins,
3 fibronectin, IL-6, and inflammatory cells than animals exposed for 1 day. Postexposure
4 challenge with single O3 exposures at different time points showed that a recovery of
5 susceptibility to O3 (as measured by BALF levels of albumin, IL-6, and the number of
6 macrophages and neutrophils) occurred at -15-20 days, but total protein and fibronectin levels
7 remained attenuated even at 20 days post-5-day exposure. The recovery with regards to BrdU
8 labeling occurred in 5-10 days after the 5 day exposure. McKinney et al. (1998) investigated the
9 role of IL-6 in the adaptive response induced by repeated O3 exposures and observed a
10 significant increases in IL-6 levels following a nighttime exposure of rats to O.Sppm O3 as
11 compared to a daytime exposure. The kinetics of inflammation were similar following these
12 exposures, but a second exposure subsequent to the nighttime exposure resulted in lesser
13 inflammation than an exposure subsequent to a daytime exposure. Pretreatment of rats with an
14 anti-IL-6 receptor antibody prior to the nighttime exposure abolished O3-induced adaptation with
15 regards to IL-6.
16 The time course of the influx of PMNs into the lung and the BALF fluid levels of
17 macrophage inflammatory protein-2 were found to be roughly similar to that for proteins (Bhalla
18 and Gupta, 2000). Adherence of neutrophils to pulmonary vascular endothelium is maximal
19 within 2 h after exposure and returns to control levels by 12 h PE (Lavnikova et al., 1998). In an
20 in vitro system utilizing rat alveolar type II cell monolayers, O3 produced a dose-dependent
21 increase in permeability (Cheek et al., 1995). At higher O3 levels, neutrophils exacerbated the
22 injury, but their presence after the exposure expedited restoration of epithelial barrier.
23 Vesely et al. (1999a) have demonstrated that neutrophils contribute to the repair process in
24 O3 -injured airway epithelium. When rats were depleted of neutrophils by rabbit anti-rat
25 neutrophil serum, and exposed to 1 ppm O3 for 8 h, epithelial necrosis in the nasal cavity,
26 bronchi, and distal airways were observed, suggesting a role for neutrophils in repair processes.
27 Proliferation of terminal bronchiolar epithelial cells, as assessed by BrdU-incorporation, was
28 also decreased by O3 exposure, suggesting a role for neutrophils in this process.
29
December 2005 AX5-33 DRAFT-DO NOT QUOTE OR CITE
-------
1 AX5.2.3.2 Concentration and Time of Exposure
2 The relative influence of concentration and duration of exposure (i.e., C x T) has been
3 investigated extensively in rats, using BALF protein as an endpoint. Though the interaction
4 between C and T is complex, concentration generally dominated the response. The impact of
5 T was C-dependent (at higher Cs, the impact of T was greater); at the lowest C and T values, this
6 dependence appeared to be lost.
7 New studies evaluating C x T relationships have not been found.
8
9 AX5.3.3.3 Susceptibility Factors
10 Factors that have been studied for potential impact on the effects of O3 include age, gender,
11 nutritional status, genetic variability, exercise and exposure to co-pollutants.
12 The effects of age on lung inflammation are not well known. Vincent et al. (1996) found
13 O3 did not differentially affect the leakage of alveolar protein in rats of different age groups, but
14 an O3-induced increase in IL-6 and N-acetyl-beta-D-glucosaminidase (NAG) was observed in
15 senescent animals compared to juvenile and adult rats. Johnston et al. (2000b) compared gene
16 expression of chemokines and cytokine in newborn and 8-week-old C57B1/6J mice exposed to
17 1.0 or 2.5 ppm O3 for 4, 20, or 24 h. The animals were killed immediately after exposure, total
18 RNA was isolated from lung tissue, ribonuclease protection assays were completed for a number
19 of cytokines/chemokines including IL-12, IL-10, IL-la, IL-lp, IL-IRa, MIF, IFN-y, MlP-la,
20 MIP-2,IL-6, and Mt. The newborn mice displayed increased levels of Mt mRNA only, while the
21 8-week-old mice had increases in MlP-la, MIP-2, IL-6, and Mt mRNA. Comparisons were
22 made with mice of the same age groups with exposures to endotoxin (10 min). Both age groups
23 displayed similar cytokine/chemokine profiles with endotoxin exposure. This suggested to the
24 authors that the responses to endotoxin, which does not cause epithelial injury, and the responses
25 to O3, which does, demonstrate that differences in inflammatory control between newborn and
26 adult mice is secondary to epithelial injury.
27 Ascorbate deficiency had been found to have only minimal effect on injury and
28 inflammation in guinea pigs exposed to O3 (Kodavanti et al., 1995). Elsayed (2001)
29 demonstrated that a general dietary restriction to 20% of the freely-fed diet for 60 days caused an
30 extreme reduction in body weight in 1-month-old SD rats. These rats, exposed to 0.8 ppm
31 continuously for 3 days, had levels of antioxidants and detoxifying enzymes that were increased
December 2005 AX5-34 DRAFT-DO NOT QUOTE OR CITE
-------
1 less than in freely fed animals. Lung injury, as detected by cell proliferation in CAR and BALF
2 levels of protein, number of neutrophils, was increased in sidestream cigarette smoke exposed
3 mice that were subsequently exposed to O3 (Yu et al., 2002). Macrophages from smoke + O3
4 exposed animals also responded by a greater release of TNF-a following LPS stimulation when
5 compared to macrophages exposed to air, smoke or O3 alone.
6 Recent lines of evidence illustrate the importance of genetic susceptibility in O3 health
7 effects. The effects of acute and subacute exposures were studied by Tankersley and Kleeberger
8 (1994) in inflammation-prone (susceptible) C57BL/6J(B6) and inflammation-resistant
9 C3H/HeJ(C3) strains of mice. Based on the neutrophilic response to O3 in these two strains and
10 in recombinant mice, the authors concluded that the acute and subacute exposures are controlled
11 by two distinct genes, referred to as Inf-land Inf-2 respectively. Exposures, when repeated
12 fourteen days after the initial exposures, caused a smaller increase in BALF protein and number
13 of macrophages, lymphocytes and epithelial cells in both strains, but PMN number was greater
14 in both strains compared to initial exposure (Paquette et al., 1994).
15 Further studies by Kleeberger et al. (1997) identified another potential susceptibility gene,
16 tumor necrosis factor (TnJ), on a qualitative trait locus on mouse chromosome 17. Tnf codes for
17 the pro-inflammatory cytokine TNF-a. By neutralizing the function of TNF-a with a specific
18 antibody, they were able to confer protection against O3 injury in susceptible mice. Cho et al.
19 (2001) demonstrated a role for tumor necrosis factor receptor 1 and 2 (TNFR1 and TNFR2,
20 respectively) signaling in subacute (0.3 ppm for 48 h) O3-induced pulmonary epithelial injury
21 and inflammation. TNFR1 and TNFR2 knockouts were less sensitive to subacute O3 exposure
22 than wild-type C57BL/6J mice. With acute exposures to O3 (2 pm for 3 h), airway
23 hyperreactivity was diminished in knockout mice compared to wild-type mice, but lung
24 inflammation and permeability were increased. Based on these studies, it has been
25 hyposthesized that in subacute O3 exposures, TNF-a is a susceptibility gene in mice, and further,
26 that independent mechanisms control lung inflammation and permeability. Further evidence for
27 the mechanistic separation of hyperresponsiveness and PMN infiltration was provided by Shore
28 et al. (2001) from studies using wild-type and TNFR knockout mice exposed to 2 ppm O3 for
29 3 h. Numbers of PMN in BAL collected 21 h PE were not changed in air exposures, but were
30 increased to the same extent in both wild-type and TNFR knockout mice, whereas
31 hyperresponsiveness was increased only in wild-type mice, but not in the knockouts.
December 2005 AX5-35 DRAFT-DO NOT QUOTE OR CITE
-------
1 Prows et al (1999) characterized the differences between A/J (O3-sensitive) and C57BL/6J
2 (O3-resistant) mice exposed to 10 ppm O3, which continued until the animals died. Sensitive
3 animals survived < 13 h and resistant mice survived > 13 h. Histological examination of the
4 lungs and wet-to-dry lung weight ratios did not differ between the two strains. Though the dose
5 given was not environmentally relevant, the genome wide scans confirmed earlier findings
6 (Prows et al. 1997) that two loci (acute lung injury-1 and -3, Alii and AH3, respectively) on
7 chromosome 11 control susceptibility to death after O3 exposure.
8 A recent comprehensive characterization of lung injury in nine inbred mouse strains (Savov
9 et al., 2004) has been summarized in Table AX5-4. The exposure consisted of 3 h of 2.0 ppm O3,
10 followed by room air, for 6 h or 24 h PE and analyzed for plethysmography, MCh challenge,
11 BALF, histology, and single nucleotide polymorphisms (SNPs). This group identified two strains
12 (C3H/HeJ and A/J) as consistently O3-resistant, and two strains (C57BL/6J and 129/SvIm) as
13 consistently O3-vulnerable. Five strains were characterized as having inconsistent phenotypes
14 with intermediate responses to O3 (CAST/Ei, BTBR, DBA/2J, FVB/NJ, and BALB/cJ). Their
15 in silico genome scan identified on chromosome 1, a 170- to 189-Mb region associated with the
16 6-h airway hyperreactivity response and the 24-h inflammatory response; on chromosome 7, a 30-
17 to 40-Mb region associated with the 6-h inflammatory response and the 6-h protein increase; and
18 on chromosome 17, a 30- to 40-Mb region associated with the 24-h airway hyperreactivity
19 response. They found no consistent correlation between the concentration of total protein in
20 BALF and influx of inflammatory cells (PMN), which they attribute to be regulated by different
21 genes. They did find a correlation between O3-induced increases in IL-6 and PMN concentration.
22 Ozone-induced changes in CCSP (called CC16 by this group) expression were evaluated in
23 five inbred mouse strains: C57BL/6J and CBA both considered sensitive to acute O3-induced
24 inflammation, C3H/HeJ and AKR/J both considered resistant, and SJL/J considered intermediate
25 (Broeckaert et al., 2003). Two exposures paradigms were used, 1.8 ppm O3 for 3 h or 0.11 ppm
26 O3, 24/h day for up to 3 days, and BALF and serum were assayed immediately after exposure or
27 at 6 h PE. Both exposure levels caused a transient increase in CC16 in serum that correlated with
28 BALF changes in protein, LDH, and inflammatory cells. There was an inverse relationship
29 between preexposure levels of CC16 in BALF and epithelial damage based on serum CC16 levels
30 and BALF markers of inflammation. There was also an inverse relationship between
December 2005 AX5-36 DRAFT-DO NOT QUOTE OR CITE
-------
Table AX5-4. Inbred Mouse Strain Susceptibility
Mouse Strain
C57BL/6J
129/SvIm
BTBR
BALB/cJ
DBA/2J
A/J
FVB/NJ
CAST/Ei
C3H/HeJ
Baseline
hyporeactive
hyperreactive
hyperreactive
intermediate
intermediate
very hyperreactive
intermediate
intermediate
intermediate
Penh
O3 only
susceptible
susceptible
susceptible
susceptible
resistant
resistant
resistant
resistant
resistant
PMN Protein
OjthenMCh 6h 24h 6h 24h
much more t 1 11 ns ns
responsive
more responsive 1 1 1 1 t ns
more responsive 1 1 1 1 1 1 1 1
more responsive t ns t ns ns t
less responsive 11 1 ns 1
much less 1 1 1 1 ns ns
responsive
less responsive 11 It t 1
less responsive 1 1 1 1 1 1 t 1 1
less responsive t I ns I
IL-6 % PCNA
6h 24h 24h
II 1 >4
It t >2
H t t < 1
t ns <1
1 1 >1
It ns > 4
1 ns > 1
H t t < 1
II ns < 1
Overall
Response
toO3
highly
sensitive
highly
sensitive
intermediate
intermediate
intermediate
highly
resistant
intermediate
intermediate
highly
resistant
-------
1 preexposure levels of albumin in BALF and lung epithelium damage. Based on these results, the
2 authors conclude that a major determinant of susceptibility to O3 is basal lung epithelial
3 permeability. As all of the mouse strains had similar levels of preexposure CC16 mRNA, they
4 explored the possible role of CC16 isozymes in differences among strains. The CC16 monomer a
5 7kD protein exist in two isoforms with differing pi values, CC16a (4.9) and CC16b (5.2).
6 To evaluate the role of CC16 isoform profiles in permeability differences between C57BL/6J and
7 C3H/HeJ, this group evaluated the CC16 protein profiles in BALF of the strains before and after
8 O3 exposure following two-dimensional protein electrophoresis analysis. C57BL/6J mice had
9 lower levels of CC16a (the more acidic form) than C3FI/HeJ. But both the strains had similar
10 levels of CC16b. Based on these observations Broeckaert et al (2003) conclude that greater
11 epithelial permeability observed in C57BL/6J may be due to difference in the expression of
12 CC16a and possibly other antioxidant/inflammatory proteins.
13 Wattiez et al. (2003) examined BALF protein from C57BL/6J (O3-sensitive) and C3H/HeJ
14 (O3-resistant) mice exposed to filtered air using a two-dimensional polyacrylamide gel approach
15 to analyze the protein profiles. C3H/HeJ mice expressed 1.3 times more Clara cell protein 16
16 (CC16) than C57BL/6J mice, and further, expressed more of the acidic isoform of CC16. Strain-
17 specific differential expression of isoforms of the antioxidant protein 2 (AOP2), the isoelectric
18 point 5.7 isoform in C3Ft/HeJ and isoelectric point 6.0 isoform in C57BL/6J were observed.
19 These studies suggested a potential role for the strain-specific differential expression in their
20 protein toward differential susceptibility to oxidative stress.
21
22 AX5.2.3.4 Mediators of Inflammatory Response and Injury
23 While neutrophils in the lung characterize an inflammatory response to O3, the release of
24 chemotactic mediators by inflammatory cells indicates their state of activation and their role in
25 continued inflammation and injury. Studies in recent years have placed a greater focus on these
26 mediators to understand the mechanisms implicated in O3-induced inflammation and injury.
27 Cytokines and chemokines have been shown to be released as a result of stimulation or injury of
28 macrophages, epithelial cells and PMNs. Many of these mediators have been implicated in PMN
29 recruitment in the lung following O3 exposure. The expression of macrophage inflammatory
30 protein 2 (MIP-2) mRNA or BALF levels of MIP-2 increased in mice and rats exposed to O3
31 concentrations equal to or greater than 1 ppm (Driscoll et al. 1993; Haddad et al., 1995; Bhalla
December 2005 AX5-38 DRAFT-DO NOT QUOTE OR CITE
-------
1 and Gupta, 2000). The increased mRNA expression was associated with an increased
2 neutrophilia in the lung. Ozone exposure also caused an increase in monocyte chemotactic
3 protein-1 (MCP-1) mRNA in mice and rats (Zhao et al., 1998). These studies implicate MCP-1 in
4 O3-induced monocyte accumulation in the lung and suggest a role of NFkB in MCP-1 gene
5 expression. Fibronectin, an extracellular matrix glycoprotein, has been studied for its role in lung
6 inflammation and inflammatory disorders. Gupta et al. (1998) observed an increase in both
7 fibronectin protein and mRNA expression in the lung of rats exposed to 0.8 ppm O3. A
8 mechanistic role of fibronectin in O3-induced inflammation and injury was suggested on the basis
9 of comparability of temporal changes in B ALF protein, fibronectin and alkaline phosphatase
10 activity (Bhalla et al., 1999). Numerous studies have reported O3-induced differential expression
11 of various cytokines and inflammatory mediators both in vivo and ex vivo: increased expression
12 of cytokine-induced neutrophil chemoattractant (CINC) and NF-KB expression in vivo (Haddad
13 et al., 1996; Koto et al., 1997); IL-8 both in vivo and in vitro (Chang et al., 1998); tumor necrosis
14 factor (TNF-a), fibronectin, interleukin-1 (IL-1), and CINC release by macrophages ex vivo
15 (Pendino et al., 1994; Ishii et al., 1997), and NF-aB and TNF-a (Nichols et al., 2001). An
16 increase in lung CINC mRNA occurred within 2 h after the end of a 3 h exposure of rats to 1 ppm
17 O3. The CINC mRNA expression was associated with neutrophilia at 24 h post-O3 exposure.
18 Exposure of guinea pig alveolar macrophages recovered in BALF and exposed in vitro to 0.4 ppm
19 O3 for 60 minutes produced a significant increase in IL-6 and TNF-a (Arsalane et al., 1995). An
20 exposure of human AMs to an identical O3 concentration increased TNF-a, IL-lb, IL-6 and IL-8.
21 This exposure also caused an increase in mRNA expression for TNF-a, IL-lb, IL-6 and IL-8 in
22 human cells. Ozone exposure caused an increase in IL-6, MlP-la, MIP-2, eotaxin and
23 metallothionein (MT) expression in mice (Johnston et al., 1999a). The IL-6 and MT increase was
24 enhanced in mice deficient in Clara cell secretory protein (CCSP), suggesting a protective role of
25 Clara cells and their secretions (Mango et al., 1998). CCSP deficiency also increased sensitivity
26 of mice to O3, as determined by an increase in abundance of MlP-la and MIP-2 following a 4 h
27 exposure (Johnston et al., 1999b).
28 A role for mast cells in airway responses is proposed on the basis of chronic exposure
29 studies demonstrating greater increases in lavageable macrophages, epithelial cells and PMNs in
30 mast cell-sufficient mice compared to mast cell-deficient mice exposed to 0.26 ppm O3
31 (Kleeberger et al., 2001b). Similar results were earlier reported by the same group using a higher
December 2005 AX5-39 DRAFT-DO NOT QUOTE OR CITE
-------
1 O3 concentration (Longphre et al., 1996). Increases in inflammatory cells were also observed in
2 mast cell-deficient mice repleted with mast cells, but O3-induced permeability increase was not
3 different in genotypic groups exposed to 0.26 ppm. When a mast cell line was exposed to varying
4 O3 concentrations, spontaneous release of serotonin and modest generation of PGD2 occurred
5 only under conditions that caused cytotoxicity (Peden and Dailey, 1995). Additionally, O3
6 inhibited IgE- and A23187-induced degranulation. Mast cells recovered from O3-exposed
7 peripheral airways of ascaris sensitive dogs released significantly less histamine and PGD2
8 following in vitro challenge with ascaris antigen or calcium ionophore (Spannhake, 1996). Ozone
9 exposure also promoted eosinophil recruitment in the nose and airways in response to instillation
10 of ovalbumin or ovalbumin-pulsed dendritic cells and aggravated allergy like symptoms in guinea
11 pigs (lijima et al., 2001).
12 Treatment of rats with cyclophosphamide prior to O3 exposure resulted in a decreased
13 recovery of PMNs in the BALF and attenuated permeability induced by O3 (Bassett et al., 2001).
14 Additionally, they found that pretreatment of animals with antiserum against rat neutrophils
15 abrogated PMN accumulation in the lung, but did not alter permeability changes produced by O3.
16 DeLorme et al. (2002) showed a relationship between neutrophilic inflammation and airway
17 hyperresponsiveness. Treatment of rats with anti-neutrophil serum protected the animals from
18 O3-induced airway hyperresponsiveness. Studies utilizing antibodies to selected pro- or anti-
19 inflammatory cytokines suggest a role of TNF-a, interleukin-10 (IL-10) and IL-lb in O3-induced
20 changes in permeability, inflammation and cytokine release (Ishii et al., 1997; Reinhart et al.,
21 1999; Bhalla et al., 2002). An attenuation of O3-induced increase in permeability and
22 inflammation was also observed in mice treated, either before or after exposure, with UK-74505,
23 a platelet-activating factor receptor antagonist (Longphre et al., 1999), suggesting that O3-induced
24 epithelial and inflammatory changes are mediated in part by activation of PAF receptors.
25 Ozone exposure stimulates macrophage motility towards a chemotactic gradient, and
26 macrophages from rats exposed to 0.8 ppm O3 adhered to epithelial cells (ARL-14) in culture to a
27 greater extent than macrophages from air-exposed controls (Bhalla, 1996). Both macrophage
28 motility and chemotaxis were attenuated by antibodies to cell adhesion molecules CD-I Ib and
29 ICAM-1. An exposure of female rats to O3 had an attenuating effect on CD-18 expression on
30 alveolar macrophages and vascular PMNs, but the expression of CD62L, a member of selectin
31 family, on vascular PMNs was not affected (Hoffer et al., 1999). In monkeys, the O3-induced
December 2005 AX5-40 DRAFT-DO NOT QUOTE OR CITE
-------
1 inflammation was blocked by treatment with a monocolonal antibody to CD 18, suggesting
2 dependence of PMN recruitment on this adhesion molecule (Hyde et al., 1999). Treatment of
3 monkeys with CD 18 antibody also reduced tracheal expression of the beta6 integrin (Miller
4 et al., 2001). A single 3 h exposure of rats to O3 resulted in an increase in neutrophil adhesion to
5 epithelial cells in culture (Bhalla and Young, 1992) and caused an elevation in concentration of
6 ICAM-1, but not CD-18, in the BALF (Bhalla and Gupta, 2000). Takahashi et al. (1995a) found
7 an increase in tissue expression of ICAM-1 in mice exposed to 2 ppm O3. They noted a temporal
8 correlation of inflammatory activity and ICAM-1 expression which varied in different regions of
9 the lung. A comparable pattern of time-related changes in total protein, fibronectin and alkaline
10 phosphatase activity in the BALF of rats exposed to 0.8 ppm O3 was also observed by Bhalla
11 etal. (1999).
12
13 AX5.2.3.5 Role of Nitric Oxide Synthase and Reactive Nitrogen in Inflammation
14 An acute exposure of rats to 2 ppm O3 caused an increase in the expression of iNOS activity
15 with an increase in BALF macrophage number and total protein and increase in fibronectin and
16 TNF-a production by AMs (Pendino et al., 1995). All of these effects of O3 were reduced by
17 pretreatment with gadolinium chloride, a macrophage inhibitor. Macrophages isolated from O3-
18 exposed mice produced increased amounts of nitric oxide, superoxide anion and PGE2, but
19 production of these mediators by macrophages from NOS knockout mice was not elevated
s~<,
20 (Fakhrzadeh et al., 2002). Additionally, mice deficient in NOS or mice treated with N -
21 monomethyl-L-arginine, an inhibitor of total NOS, were protected from O3-induced permeability,
22 inflammation and injury, suggesting a role of nitric oxide in the production of O3 effects
23 (Kleeberger et al., 2001b; Fakhrzadeh et al., 2002). Another study demonstrated greater injury
24 (as determined by measurement of MIP-2, matrix metalloproteinases, total protein, cell content
25 and tyrosine nitration of whole lung protein) in iNOS knockout mice than in wild-type mice on O3
26 exposure (Kenyon et al., 2002). They proposed that protein nitration differences are related to
27 inflammation and may not be dependent on iNOS-derived NO.
28 Ishii et al. (2000a) performed studies using pretreated rats with ebselen (a potent anti-
29 inflammatory, immunomodulator and NO/peroxynitrite scavenger) and then exposed to 2 ppm O3
30 for 4 h. The pretreated rats had decreased numbers of neutrophils, lowered albumin levels, and
31 inhibited nitration of tyrosine residues in BALF 18 h PE, without changes in macrophage iNOS
December 2005 AX5-41 DRAFT-DO NOT QUOTE OR CITE
-------
1 expression. These results suggest that an iNOS-independent mechanism may be involved in O3-
2 induced inflammation. Inoue et al. (2000) demonstrated in human transformed bronchial
3 epithelial cells that NO-generating compounds (TNF-a, IL-lp, and INF-y) induce IL-8
4 production and that NOS inhibitors inhibit IL-8 production. In vivo experiments in the same
5 study using male Hartley-strain guinea pigs exposed to 3 ppm O3 for 2 h showed that NOS
6 inhibitor pretreatment attenuated-O3 induced neutrophil recruitment and airway
7 hyperresponsiveness at 5 h after exposure. The NOS inhibitors also blunted the increase in
8 nitrate/nitrite levels and in IL-8 mRNA, at the 5 h PE. The authors hypothesize that NO, or its
9 derivatives, facilitate airway hyperresponsiveness and inflammation after O3 exposure, possibly
10 mediated by IL-8. Jang et al. (2002) have attempted to characterize the mechanism by which
11 short-term O3 exposures (0.12, 0.5, 1, or 2 ppm for 3 h) cause airway inflammation and
12 responsiveness in BALB/c mice. Using a modified Griess reaction, measurement of nitrate and
13 nitrite in BAL fluid after O3 exposure showed dose-dependent increases in nitrate, which is
14 indicative of in vivo NO generation. Functional studies of enhanced pause (Penh) demonstrated
15 increases with O3 which were also dose-dependent. Western blot analysis of lung tissue showed
16 increases in NOS-1, but not in NOS -3 or iNOS isoforms. The authors conclude that in mice
17 NOS-1 may induce airway responsiveness by a neutrophilic airway inflammation.
18
19 AX5.2.4 Morphological Effects
20 AX5.2.4.1 Introduction
21 All laboratory animal species studied to date show generally similar morphological
22 responses to < 1 ppm O3. The precise characteristics of the structural changes due to O3 are
23 dependent on the exposure regimen, time of examination, distribution of sensitive cells, and the
24 type of centriacinar region (i.e., junction between the end of the terminal bronchioles and the first
25 few generations of either respiratory bronchioles or alveolar ducts, depending on the species).
26 The presentation of this morphology section will begin with effects of short-term exposure
27 (< 1 week). The subsequent discussion of long-term exposure effects (>1 week) is an artificial
28 separation, but consistent with the division of studies in other effects sections. The long-term
29 studies that included evaluation of various durations of exposures are presented totally in this
30 subsection to illustrate the effects of exposure duration and to provide a better understanding of
31 possible chronic effects of multi-seasonal, ambient O3 exposures in the population (Chapter 7).
December 2005 AX5-42 DRAFT-DO NOT QUOTE OR CITE
-------
1 AX5.2.4.2 Short-Term Exposure Effects on Morphology
2 Morphological effects of key new exposure studies generally lasting less than 1 week are
3 summarized in Table AX5-5. The following discussion is by region of the respiratory tract first,
4 followed by exposure and susceptibility factors.
5 Hotchkiss et al. (1998) explored the efficacy of a topical anti-inflammatory corticosteriod,
6 fluticasone propionate (FP), in preventing the inflammation and mucous cell metaplasia in rats
7 after cumulative O3 exposure. Male F-344 rats were exposed to filtered air or 0.5 ppm O3,
8 8 h/day, for 3 or 5 days. Immediately before and after exposure, the rats were given FP (25 jig)
9 by intranasal instillation (50 jiL/nasal passage) or an equivalent amount of vehicle only. Nasal
10 tissues were processed for light microscopy 2 h and 3 days after the 3- and 5-day exposures,
11 respectively. Rats treated with FP had 30 to 60 % less nasal inflammation after 3 and 5 days of
12 O3 exposure and 85% less mucous cell metaplasia after the 5-day exposure compared with
13 vehicle-instilled, O3-exposed controls.
14 Fanucchi et al. (1998) reported that exposure to bacterial endotoxin, a common ambient air
15 toxin, can potentiate mucous cell metaplasia in the nasal transitional epithelium of rats caused by
16 a previous O3 exposure. Male F344/N Hsd rats were intranasally instilled with saline or
17 100 |ig/ml endotoxin after exposure to filtered air or 0.5 ppm O3, 8 h/day for 3 days, and
18 evaluated 6 h and 3 days PE. Mucous cell metaplasia was not found in the air/endotoxin group,
19 but was found in the O3/saline group and was most severe in the O3/endotoxin group. A similar
20 synergistic effect of O3 and endotoxin on the nasal epithelium was reported in Fischer rats by
21 Wagner et al. (2001a,b). When exposed to O3 alone (0.5 ppm, 8 h/day for 3 days), rats developed
22 epithelial lesions in the nasal transitional epithelium. Exposure to endotoxin alone (20 jig) caused
23 lesions in the respiratory epithelium of the nose and conducting airways. Endotoxin enhancement
24 of the O3-induced mucous cell metaplasia was related to neutrophilic inflammation.
25 Cho et al. (1999a, 2000) reported that O3-induced mucous cell metaplasia in the transitional
26 epithelium of rats may be dependent on pre-metaplastic responses, such as mucin mRNA
27 upregulation, neutrophilic inflammation, and epithelial proliferation. Male F344/N rats were
28 exposed to 0.5 ppm O3, 8 h/day for 1, 2, or 3 days and 2 h, or 1, 2, or 4 days PE were assayed for
29 the parameters listed above. A rapid increase in an airway-specific mucin gene (rMuc-5AC
30 mRNA) occurred rapidly after exposure to O3, both before and during the onset of mucous cell
31 metaplasia. Neutrophilic inflammation coincided with epithelial DNA synthesis and
December 2005 AX5-43 DRAFT-DO NOT QUOTE OR CITE
-------
Table AX5-5. Effects of Ozone on Lung Structure: Short-Term Exposures
to
O
o
X
Concentration
ppm
0.1
0.5
1.0
0.2
0.4
0.2
0.4
0.8
0.4
0.4
1.0
Hg/m3
196
980
1,960
392
784
392
784
1,568
784
784
1,960
Duration
8 h/ day x 1 day
8 h/day x 1 day
8 h/day x 1, 10,
75, and 90 days
3, 7, 28, and
56 days; 3-, 7-,
and 28-day
recovery from
28 days of
exposure
23 h/day for
7 days
12 h/day;
1- or 7-day
exposure
2h
Species
Rat;
male;
Sprague-Dawley
Mouse;
male; Nffl;
Rat;
male; Wistar
RIV:Tox
Guinea pig;
male; Hartley
Crl:(HA)BR
Guinea pig;
female; Hartley;
±AH2 diet
Rat;
(Wistar RiV:TOX;
M&F; 1,3,9, &
18 months of age
Monkey;
adult male rhesus
Effects3
No dose-related response on CYP2E1, one of six P450 enzymes identified
in respiratory tissue. CYP2E1 activity was elevated (250% and 280%) in the
lobar bronchi / major daughters airways immediately after 1.0 ppm O3 exposure
for Iday and 10 days, respectively, but not in the trachea or distal bronchioles;
CYPE1 activity was unchanged and decreased after 1.0 ppm O3 exposure for 75
and 90 days, respectively.
Concentration-related centriacinar inflammation, with a maximum after 3 days
of exposure; number of alveolar macrophages and pulmonary cell density
increased progressively until 56 days of exposure, with the guinea pig the most
sensitive species. Concentration and exposure-time dependent hypertrophy of
bronchiolar epithelium in mouse only. Exposure to 0.2 ppm for 3 and 7 days
caused significant histological and morphometric changes in all 3 species;
exposure for 56 days caused alveolar duct fibrosis in rat and guinea pigs. Total
recovery in rats after 28-day exposure, but not in guinea pigs or mice .
Treatment-related lesions were observed after exposure to 0.4 and 0.8 ppm O3;
lesions were primarily seen in the terminal bronchioles and consisted
of mononuclear cell and neutrophilic infiltrate and thickening of the
peribronchiolar interstitium. Effects were only marginally exacerbated by the
AH2 (ascorbic acid) deficient diet and lesions were resolved after 1 week in FA.
Centriacinar inflammation (increased alveolar macrophages and PMNs;
increased proximal and ductular septal density) was greatest in young rats
(1 month and 3 months for 1-and 7-day exposures, respectively) and decreased
with age. No major gender differences were noted.
Reduced glutathione (GSH) increased in the proximal intrapulmonary bronchus
after 0.4 ppm O3 and in the respiratory bronchiole after 1.0 ppm O3. Local O3
Reference
Watt et al.
(1998)
Dormans
etal. (1999)
Kodavanti
etal. (1995)
Dormans
etal. (1996)
Plopper et al.
(1998)
dose (measured as excess 18O) varied by as much as a factor of three in different
airways of monkeys exposed to 1.0 ppm, with respiratory bronchioles having
the highest concentration and the parenchyma the lowest concentration. After
exposure to 0.4 ppm, the O3 dose was 60% to 70% less and epithelial injury was
minimal, except in the respiratory bronchiole, where cell loss and necrosis
occurred, but was 50% less than found at 1.0 ppm.
-------
o
O
to
O
O
Table AX5-5 (cont'd). Effects of Ozone on Lung Structure: Short-Term Exposures
X
Concentration
ppm fig/m3
0.5 980
0.5 980
0.5 980
0.5 980
Duration Species
8 h Rat;
+ BrdU to label male; F344
epithelial cells
8 h/day, Rat;
3 or 5 days; male; F344
+ fluticasone
propionate (FP)
intranasally
8 h/day, Rat;
3 days male; F344/N Hsd
+ endotoxin
(100 ug/mL )
intranasally
8 h/day, Rat;
1, 2, or 3 days male; F344/N
+ BrdU to label
epithelial cells
+ antirat
neutrophil
antiserum
Effects3
O3 exposure induced a transient influx of neutrophils and a significant (17%)
loss of NTE cells 2-4 h after exposure. Increased epithelial DNA synthesis was
first detected 12 h PE. LI and ULLI indices of epithelial cell DNA synthesis
were greatest 20-24 h and still elevated 36 h PE; numeric density of NTE cells
returned to control levels 20-24 h PE.
No significant difference of FP on morphometry of the maxilloturbinates;
O3 exposure caused neutrophilic rhinitis with 3.3- and 1.6-fold more
intraepithelial neutrophils (3-day and 5-day exposure, respectively) and marked
mucous cell metaplasia (5 -day exposure only) with numerous mucous cells and
approximately 60 times more IM in the nasal transitional epithelium; FP-treated
rats exposed to O3 had minimal nasal inflammation and mucous cell metaplasia.
Endotoxin-induced neutrophilia in nasal mucosa with NTE; mucous cell
metaplasia was not detected in air/endotoxin-exposed rats, was observed
in O3/saline-exposed rats, and was most severe in O3/endotoxin-exposed rats.
Acute O3 exposure induced a rapid increase in rMuc-5AC mRNA levels prior to
the onset of mucous cell metaplasia; neutrophilic inflammation coincided with
epithelial DNA synthesis and upregulation, but was resolved when mucous cell
metaplasis first appeared in the NTE.
Maxilloturbinates lined with NTE determined the epithelial labeling
index, numeric densities of neutrophils, total epithelial and mucous secretory
cells, amount of stored intraepithelial mucosubstances, and steady-state
ratMUC-5AC (mucin) mRNA levels. Four days after a 3-d exposure,
antiserum-treated, O3-exposed rats had 66% less stored intraepithelial
mucosubstances and 58% fewer mucous cells in their NTE than did controls.
Antiserum treatment had no effects on O3-induced epithelial cell proliferation or
mucin mRNA upregulation.
Reference
Hotchkiss
etal. (1997)
Hotchkiss
etal. (1998)
Fanucchi
etal. (1998)
Cho et al.
(1999a,
2000)
-------
December
to
o
o
X
v\
|V
4^
ON
Table AX5-5 (cont'd). Effects of Ozone on Lung Structure: Short-Term Exposures
Concentration
ppm fig/m3 Duration
0.5 980 8 h/day,
3 days
+ endotoxin
0.5 980 8 h/day,
1 and 3 days
+ OVA
(1%,
50 uL/nasal
passage)
1 1,960 8 h
Species
Rat;
F-344
Rat;
Brown Norway
Rat;
Sprague-Dawley
Ferret; young male
Monkey; young male
rhesus
Effects3
Enhanced epithelial lesions in the NTE and respiratory epithelium of the nose
and conducting airways by endotoxin and O3 exposures, respectively;
synergistic effects of coexposure mediated by neutrophils. Endotoxin increased
rMuc-5 AC mRNA levels in the NTE of O3 -exposed rats; neutrophil depletion,
however, had no effect on endotoxin-induced upregulation of mucin gene
mRNA levels. Endotoxin enhanced the O3-induced increase in stored
mucosubstances (4-fold increase), but only in neutrophil-sufficient rats
O3 enhanced the appearance of eosinophils in the maxilloturbinates of
OVA-challenged rats but did not increase inflammation in other nasal tissues;
O3/OVA coexposures for 3 days increased the number of epithelial cells as well
as the appearance of mucus-containing cells in the NTE lining the
maxilloturbinates.
Severe, acute infiltration of neutrophils along with necrotic bronchiolar
epithelium in all lung regions, especially in the centriacinar region; necrosis
and inflammation was more severe in ferrets and monkeys than in rats.
Reference
Wagner et al.
(2001a,b)
Wagner et al.
(2002)
Sterner-Kock
et al. (2000)
aAM = Alveolar macrophage.
PE = Postexposure (i.e., time after O3 exposure ceased).
LM = Light microscopy.
EM = Electron microscopy.
RB = Respiratory bronchiole.
TB = Terminal bronchiole.
IAS = Interalveolar septum.
PMN = Polymorphonuclear leukocyte.
-------
1 upregulation of rMuc-5AC, but was resolved before the development of epithelial metaplasia. In
2 the follow-up study, the investigators found that only the mucous cell metaplasia was neutrophil-
3 dependent, whereas O3-induced epithelial cell proliferation and mucin gene upregulation were
4 neutrophil-independent.
5 In the centriacinar region., different species have similar responses to low levels of O3
6 (> 0.2 ppm for 1 week; Dormans et al., 1999). Dormans et al. (1999) compared the extent and
7 time course of fibrotic changes in mice, rats, and guinea pigs exposed to 0.2 and 0.4 ppm O3 for 3,
8 7, 28, and 56 days. They found a concentration-related centriacinar inflammation in all three
9 species, with a maximum occurring after 3 days of exposure and total recovery within 3 days.
10 The effects of exposure duration are complex and are likely responsible for the similar
11 patterns of biochemical responses (see Section 5.2.1). Repair of the damage by removal of
12 injured epithelial cells is enhanced by the influx of neutrophils (Hyde et al., 1999; Veseley et al.,
13 1999a; Miller et al., 2001; see Section 5.2.3).
14 Hotchkiss et al. (1997) reported that labeling indices for rat nasal transitional epithelial cell
15 DNA were greatest 20 to 24 h after an 8-h exposure to 0.5 ppm O3, but still greater than control
16 by36hPE.
17 Exploring the role of susceptibility factors on morphological changes, Dormans et al.
18 (1999) compared morphological, histological, and biochemical effects in the rat, mouse, and
19 guinea pig after O3 exposure and after recovery in clean air. Wistar RIV:Tox male rats, NIH male
20 mice, and Hartley Crl:(HA)BR male guinea pigs were continuously exposed to filtered air, 0.2, or
21 0.4 ppm for 3, 7, 28, and 56 days. Recovery from 28 days of exposure was studied at intervals of
22 3, 7, and 28 days PE. Morphometric analysis was performed only on lung parenchyma with
23 proximal alveoli and smaller alveolar ducts and no distinct species-specific differences were
24 noted. The mouse was the most sensitive as shown by a concentration and exposure-time
25 dependent persistence of bronchiolar epithelial hypertrophy, elevated lung enzymes, and slow
26 recovery from exposure. In both rats and guinea pigs, 56 days of exposure to 0.4 ppm O3 caused
27 increased amounts of collagen in ductal septa and large lamellar bodies in Type II cells; however,
28 the inflammatory response was greater in the guinea pig. Overall, the authors rated mice as most
29 susceptible, followed by guinea pigs and rats.
30 In another comparative study of airway effects, Sterner-Kock et al. (2000) exposed ferrets,
31 monkeys and rats to 1.0 ppm O3 for 8 h. The ferrets developed epithelial necrosis and
December 2005 AX5-47 DRAFT-DO NOT QUOTE OR CITE
-------
1 inflammation that was similar to the monkey, and more severe than that found in rats. Because
2 ferrets have a similar pulmonary structure as humans (e.g., well-developed respiratory
3 bronchioles and submucosal glands), the authors concluded that the ferret would be a better
4 model than rodents for O3-induced airway effects.
5 Younger rats were found to have larger centriacinar lesions than older rats after a 1- or 7-
6 day exposure to 0.4 ppm O3 (Dormans et al., 1996). Thus, age susceptibility is dependent on the
7 endpoint examined.
8 Rats with endotoxin-induced rhinitis were more susceptible to mucous cell metaplasia in the
9 nasal transitional epithelium caused by a 3-day exposure to 0.5 ppm O3 (Cho et al.,1999b).
10 Wagner et al. (2002) reported a similar O3-induced enhancement of inflammatory and epithelial
11 responses associated with allergic rhinitis. Brown Norway rats were exposed to 0.5 ppm O3,
12 8 h/day for 1 day or 3 consecutive days and then immediately challenged intranasally with either
13 saline or ovalbumin.
14 More recent research has focused on the concept of O3 susceptible and non-susceptible sites
15 within the respiratory tract, including in situ antioxidant status and metabolic activity. Plopper
16 et al. (1998) examined whether the variability of acute epithelial injury to short-term O3 exposure
17 within the tracheobronchial tree is related to local tissue doses of O3 or to local concentrations of
18 reduced glutathione (GSH). Adult male rhesus monkeys were exposed for 2 h to filtered air or O3
19 (0.4 or 1.0 ppm). The O3 was generated by 18 O2 for determination of local O3 dose in the trachea,
20 proximal bronchi, distal bronchi, and proximal respiratory bronchioles. Analyses of GSH and
21 extracellular components (BAL) also were performed. Significant cellular injury was found at all
22 sites, but the most damage, along with increased inflammatory cells, occurred in the proximal
23 respiratory bronchiole. A significant reduction in GSH was found in the proximal bronchus at 0.4
24 ppm O3, and in the respiratory bronchiole at 1.0 ppm O3. A significant decrease in the percent of
25 macrophages, along with significant increases in the percent of neutrophils and eosinophils, and a
26 doubling of total lavage protein, were found after exposure to 1.0 ppm O3 only. The authors
27 concluded that the variability of local O3 dose in the respiratory tract was related to inhaled O3
28 concentration and was closely associated with local GSH depletion and with the degree of
29 epithelial injury.
30 Plopper et al. (e.g., Watt et al.,1998; Paige et al., 2000) explored the site-specific
31 relationship between epithelial effects of O3 exposure and the metabolism of bioactivated
December 2005 AX5-48 DRAFT-DO NOT QUOTE OR CITE
-------
1 compounds within the respiratory tract of rats. The distribution of cytochrome P450 (CYP2E1)-
2 dependent activity, measured with a selective substrate (p-nitrocatechol), was found to be highest
3 in the distal bronchioles and minor daughter airways, and lower in the lobar bronchi and major
4 daughter airways. Short-term O3 exposure (1 ppm for 8 h) increased CYP2E1 activity in the
5 lobar bronchi/major daughter airways only; however, long-term O3 exposure (1 ppm for 90 days)
6 decreased CYP2E1 activity in the major and minor airways, further complicating the
7 interpretation of O3 effects based on concentration and duration of exposure and recovery. Rats
8 treated i.p. with 1-nitronaphthalene, a pulmonary toxicant requiring metabolic activation, and
9 exposed to 0.8 ppm O3, 8 h/day for 90 days showed greater histopathologic and morphometric
10 effects in the centriacinar region of the lung (Paige et al., 2000b). Despite reported tolerance to
11 oxidant stress after long-term O3 exposure, there was increased severity of ciliated cell toxicity.
12
13 AX5.2.4.3 Long-Term Exposure Effects on Morphology
14 Key new exposure studies describing the morphological effects of O3 exposures lasting
15 longer than 1 week are summarized in Table AX5-6.
16 Marked mucous cell metaplasia was found in F344 rats exposed to 0.5 and 1.0 ppm O3, but
17 not 0.12 ppm for 20 months (Harkema et al., 1997a). In a follow-up study, hyperplasia was found
18 in the nasal epithelium of rats exposed to 0.25 and 0.5 ppm, 8 h/day, 7 days/week, for 13 weeks
19 (Harkema et al., 1999). The mucous cell metaplasia, and associated intraepithelial
20 mucosubstances, induced by 0.5 ppm O3 persisted for 13 weeks after exposure. An acute (8-h)
21 exposure to 0.5 ppm O3 13 weeks after the chronic exposure induced an additional increase of
22 mucosubstances in the nasal epithelium of rats, but not in rats chronically exposed to 0 or
23 0.25 ppm O3.
24 Rats continuously exposed for 6 months to the ambient air of Sao Paulo, Brazil (11 ppb O3;
25 1.25 ppm CO; 35 |ig/m3 PM; 29 |ig/m3 SO2) also developed secretory hyperplasia in the upper
26 airways (Lemos et al., 1994). No significant changes in nasal tissue, however, were seen in rats
27 continuously exposed for 49 days to the ambient air of Mexico City, Mexico (Moss et al., 2001).
28 Apoptosis regulators like Bcl-2 may play a role in the development and resolution of
29 mucous cell metaplasia in the nasal airway (Tesfaigzi et al., 1998). In rats exposed to 0.5 ppm O3
30 for 1 month, Bcl-2 was found in protein extracts of nasal epithelium. After 3 and 6 months of
31 exposure, the number of metaplastic mucous cells in the transitional epithelium was indirectly
December 2005 AX5-49 DRAFT-DO NOT QUOTE OR CITE
-------
Table AX5-6. Effects of Ozone on Lung Structure: Long-Term Exposures
1, J
tt>
<^ '
CD
>-l
to
O
o
^
X
Y1
o
O
i-rj
H
6
o
0
H
O
o
d
Concentration
ppm
Mexico
City
Ambient:
0.018
(> 0.12
for 18 1-h
intervals)
0.12
0.5
1.0
0.12
0.50
1.0
0.12
1.0
jig/m3 Duration
23 h/day
for 7 weeks
35.3
235 6 h/day,
980 5 days/week
1,960 for
20 months
235 6 h/day,
980 5 days/week
1.960 for 24 and
30 months
235 6 h/day,
1,960 5 days/week,
for 2 or 3
months
Species
Rat;
male;
F344;
8 weeks old
Rat;
male;
F344;
6-8 weeks old
Mouse; male
and female;
B6C3F1;
6-7 weeks old
Rat;
F344/N
Effects3
No inflammatory or epithelial lesions in nasal airways or respiratory tract.
LM morphometry of CAR remodeling. Thickened tips of alveolar septa lining
ADs (alveolar entrance rings) 0.2 mm from TB in rats exposed to 0. 12 ppm and
to 0.6 mm in rats exposed to 1.0 ppm. At 0.5 and 1.0 ppm, atrophy of nasal
turbinates, mucous cell metaplasia in NTE, increased volume of interstitium and
epithelium along ADs due to epithelial metaplasia, and bronchiolar epithelial
hyperplasia. At 1.0 ppm, increased AMs and mild fibrotic response (increase in
interstitial matrix and cellular interstitium; the latter due to increase in volume in
interstitial fibroblasts). More effects in PAR than in terminal bronchioles.
Effects not influenced by gender or by aging. Effects similar to, or model of,
early fibrotic human disease (e.g., idiopathic pulmonary fibrosis).
Effects in the nose and centriacinar region of the lung at 0.5 and 1.0 ppm. Nasal
lesions were mild: hyaline degeneration, hyperplasia, squamous metaplasia,
fibrosis, suppurative inflammation of transitional and respiratory epithelium; and
atrophy of olfactory epithelium. Lung lesions: alveolar/bronchiolar epithelial
metaplasia and histiocytosis in terminal bronchioles, alveolar ducts, and
proximal alveoli. Severity was greatest in mice exposed to 1.0 ppm O3, but there
was minimal interstitial fibrosis.
Morphometric changes (epithelial thickening, bronchiolarization) occurred after
2 or 3 months exposure to 1.0 ppm O3; effects were similar to those found with
20 months exposure (see Pinkerton et al., 1995)
Reference
Moss et al.
(2001)
Catalano et al.
(1995a,b);
Chang et al.
(1995);
Harkema et al.
(1994, 1997a,b)
Pinkerton et al.
(1995);
Plopper et al.
(1994a);
Stockstill et al.
(1995)
Herbert et al.
(1996)
Pinkerton et al.
(1998)
O
HH
H
W
-------
o
O
to
O
O
Table AX5-6 (cont'd). Effects of Ozone on Lung Structure: Long-Term Exposures
X
Concentration
ppm
Duration
Species
Effects3
Reference
0.25 490 8 h/day,
0.5 980 7 days/week
for
13 weeks
0.4 784 23.5 h/day
for 1, 3, 7, 28,
or 56 days
0.5 980 8 h/day for 1,
3, and 6
months
Rat;
male;
F344/NHSD;
10-14 weeks
old
Rat;
Wistar
7 weeks old
Rat;
male;
F344/N
Mucous cell hyperplasia in nasal epithelium after exposure to 0.25 and 0.5 ppm
O3; still evident after 13 weeks recovery from 0.5 ppm O3 exposure. Mucous cell
metaplasia found only after 0.5 ppm O3, but still detectable 13 weeks PE.
Acute inflammatory response (increased PMNs and plasma protein in B ALF)
reached a maximum at day 1 and resolved within 6 days during exposure;
AMs in B ALF increased progressively up to day 56, and slowly returned to near
control levels with PE recovery. Histological examination and morphometry of
the lungs revealed CAR inflammatory responses throughout O3 exposure;
thickening of septa was observed at day 7. Ductular septa thickened
progressively at days 7, 28, and 56 of exposure; showed increased collagen at
day 28, which was further enhanced at day 56. Increased RBs with continuous
exposure. Collagen and bronchiolization remained present after a recovery
period.
Increased Bcl-2, a regulator of apotosis, after 1 month, decreasing somewhat
thereafter, returning to baseline by 13 weeks PE; increased number of
metaplastic mucous cells in NTE after 3 and 6 months.
Harkema
(1999)
Van Bree
(2002)
Tesfaigzi
(1998)
etal.
etal.
etal.
0.5 980 8 h/day for 1,
3, and 6
months
0.5 980 8 h/day for
5 days,
every 5 days
for a total of
1 1 episodes
Rat;
male;
F344/N
Monkey;
bonnet;
30-day-old
infants
Increased Bcl-2, a regulator of apotosis, after 1 month, decreasing somewhat
thereafter, returning to baseline by 13 weeks PE; increased number of
metaplastic mucous cells in NTE after 3 and 6 months.
Increased density and distribution of goblet cells in RB whole mounts stained
with AB/PAS; extensive remodeling of distal airway with O3 and O3 + HDMA
challenge; increased airways resistance and reactivity, and respiratory motor
adaptation also occurred. Authors conclude that periodic cycles of acute injury
and repair associated with the episodic nature of environmental patterns of O3
exposure alters postnatal morphogenesis and epithelial differentiation in the
distal lung of infant primates.
Tesfaigzi et al.
(1998)
Evans et al.
(2003);
Schelegle et al.
(2003a);
Chen et al.
(2003);
Fanucchi et al.
(2000);
Plopper and
Fanucchi (2000)
-------
o
O
to
O
O
Table AX5-6 (cont'd). Effects of Ozone on Lung Structure: Long-Term Exposures
>
X
(Si
I
to
Concentration
ppm
0.8
0.5
0.5
jig/m3 Duration
1,568 8 h/day for
90 days
+ 1-NN
(100 mg/kg)
980 11 episodes of
5 days each,
8 h/day
followed by
9 days of
recovery
980 1 1 episodes of
5 days each,
8 h/day
followed by
9 days of
recovery
Species
Rat;
male;
Sprague-
Dawley
Monkey;
Macaca
mulatto;
30 days old
Monkey;
Macaca
mulatto;
30 days old
Effects3
Increased O3-induced centriacinar toxicity (histopathology, TEM, morphometry)
of 1-Nitronaphthalene (1-NN), a pulmonary cytotoxicant requiring metabolic
activation, especially to ciliated cells.
In small conducting airways O3 caused decrements in density of airway epithelial
nerves. Reduction greater with HDMA + O3. O3 or HDMA+O3 caused increase
in number of PGP 9.5 (pan-neuronal marker) in airway. CGRP-IR nerves were
in close contact with the PGP9.5 positive cells. Appearance of clusters of
PGP9.5+/CGRP cells. Suggests episodic O3 alters developmental pattern of
neural innervation of epithelial compartment.
Abnormalities in the BMZ included: (1) irregular and thin collagen throughout
the BMZ; (2) perclecan depeleted or severely reduced; (3) FGFR-1
immunoreactivity was reduced; (4) FGF-2 immunoreactivity was absent in
perlecan-deficient BMZ, but was present in the lateral intercelluar space (LIS),
in basal cells, and in attenuated fibroblasts; (5) syndecan-4 immunoreactivity
was increased in basal cells.
Reference
Paige et al.
(2000b)
Larson et al.
(2004)
Evans et al.
(2003)
H
6
o
O
H
/O
aTB = Terminal bronchiole.
PE = Postexposure (i.e., time after O3 exposure ceased).
AM = Alveolar macrophage.
LM = Light microscopy.
EM = Electron microscopy.
RB = Respiratory bronchiole.
IAS = Interalveolar septum.
C x T = Product of concentration and time.
O
HH
H
W
-------
1 related to the percentage of cells that were Bcl-2 positive. Cells from rats exposed to filtered air
2 did not express any Bcl-2.
3 A similar spectrum of lesions also was reported (Herbert et al., 1996) in the nasal cavity and
4 centriacinar lung of male and female mice exposed to 0.5 or 1.0 ppm of O3 for 2 years, which
5 persisted with continued exposure for 30 months. Few changes, however, were found in other
6 endpoints (e.g., lung function or lung biochemistry) examined in these rats. The investigators'
7 interpretation of the entire study is that rodents exposed to the two higher O3 concentrations had
8 some structural hallmarks of chronic airway disease in humans.
9 A fifth long-term study was reported in infant monkeys by Plopper et al. (Evans et al., 2003;
10 Schelegle et al., 2003a, 2003b; Chen et al., 2003; Plopper and Fanucchi, 2000; Fanucchi et al.,
11 2000) using a shorter simulated, seasonal O3-exposure pattern, but at a higher O3 concentration
12 (0.5 ppm) than the protocol used by Tyler et al. (1988, 1991a). Infant rhesus monkeys (30 days
13 old) were exposed to filtered air, house dust mite allergen aerosol (HDMA), or O3 + HDMA. The
14 O3 exposures were 8 h/day for 5 days, every 14 days for a total of 11 O3 episodes. Half of the
15 monkeys were sensitized to house dust mite allergen (Dermatophagoides farinae) at 14 and 28
16 days of age. The sensitized monkeys were exposed to HDMA for 2h/day on days 3-5 of the FA
17 or O3 exposures. The lungs were removed during the last filtered air exposure and the right and
18 left cranial and right middle lobes were separately inflation fixed. Microdisection and
19 morphometric analyses were performed on the conducting airways to the level of the most
20 proximal respiratory bronchiole. Repeated exposures to O3 or O3 + HDMA over a 6-month
21 period resulted in an atypical development of the basement membrane zone of airways in
22 nonsensitized developing monkeys. A profound remodeling in the distal conducting airways was
23 found in the sensitized monkeys as a result of the damage and repair processes occurring with
24 repeated exposure (Evans et al., 2003; Schelegle et al., 2003a; Fanucchi et al., 2000).
25 Schelegle et al. (2003a) reported the lung histopathology results from the O3 exposures to
26 infant monkeys. At necropsy, cross sections of the left caudal lobe were prepared from each
27 animal. The accumulation of eosinophils and mucous cells within the combined epithelium and
28 interstitium compartments was determined in the conducting airways and in the
29 terminal/respiratory bronchioles. House dust mite sensitization and HDMA challenge alone, or
30 combined with O3 exposure, resulted in significantly greater eosinophil accumulation in the
31 conducting airways when compared to FA and O3 only exposures. A significant accumulation of
December 2005 AX5-53 DRAFT-DO NOT QUOTE OR CITE
-------
1 eosinophils was found in the terminal/respiratory bronchioles of the sensitized monkeys
2 challenged with HDMA when compared to monkeys exposed to FA, O3, and HDMA + O3. The
3 mean mass of mucous cells increased in the fifth generation conducting airways of sensitized
4 animals challenged with HDMA alone and when combined with O3 exposure, and in the terminal
5 bronchioles of sensitized animals exposed to HDMA + O3. The tracheal basement membrane of
6 house dust mite-sensitized monkeys exposed to HDMA or to HDMA + O3 was significantly
7 increased over controls; however, there were no significant changes in the airway diameter of
8 proximal and mid-level airways. The authors interpreted these findings to indicate that the
9 combination of cyclic O3 exposure and HDMA challenge in house dust mite-sensitized infant
10 monkeys act synergistically to produce an allergic-reactive airway phenotype characterized by
11 significant eosinophilia of midlevel conducting airways, transmigration of eosinophils into the
12 lumen, and an altered structural development of conducting airways. Exposures of sensitized
13 young monkeys to HDMA alone, or to O3 alone, resulted in eosinophilia of the mid-level
14 conducting airways and the terminal/respiratory bronchioles, but without alterations in airway
15 structure or function. Evans et al. (2003) examined development of the tracheal basement
16 membrane zone (BMZ) in these monkeys and found that with exposures to either O3 or HDMA +
17 O3 BMZ development was affected. Abnormalities in the BMZ included: (1) irregular and thin
18 collagen throughout the BMZ; (2) perclecan depeleted or severely reduced; (3) FGFR-1
19 immunoreactivity was reduced; (4) FGF-2 immunoreactivity was absent in perlecan-deficient
20 BMZ, but was present in the lateral intercelluar space (LIS), in basal cells, and in attenuated
21 fibroblasts; (5) syndecan-4 immunoreactivity was increased in basal cells. The authors interpret
22 these data to suggest that O3 targets cells associated with synthesis of epithelial BMZ perlecan.
23 The absence of FGF-2, normally stored in the BMZ, could affect downstream signaling in airway
24 epithelium and could be responsible for the abnormal development of the airway seen in this
25 study, and thus be an important mechanism modulating O3-induced injury.
26 Mid-level bronchi and bronchioles from these monkeys were examined for alterations in
27 airway innervation (Larson et al., 2004). They found decrements in the density of epithelial
28 nerves in the axial path between the sixth and seventh airway generations in exposures to O3.
29 Combined O3+HDMA exposures exacerbated this reduction. They attribute this loss of nerve
30 plexuses to neural regression or stunted nerve development, the latter corroborated by the Evans
31 et al. (2003) finding of decreased growth factors following O3 exposure. Additionally, they found
December 2005 AX5-54 DRAFT-DO NOT QUOTE OR CITE
-------
1 streaks or clusters of cells immunoreactive for protein gene product 9.5 (PGP 9.5, a pan-neuronal
2 marker) and negative for calcitonin gene-related peptide. The functional significance of this is
3 unknown, and presumed to be implicated in injury-repair process induced by O3.
4 Recently, bronchiolization was reported in rats exposed to 0.4 ppm O3 for only 56 days (van
5 Bree et al., 2001). Collagen formation progressively increased with increasing O3 exposure, and
6 remained increased into PE recovery. In addition to centriacinar remodeling, Pinkerton et al.
7 (1998) reported thickening of tracheal, bronchial, and bronchiolar epithelium after 3 or 20 months
8 exposure to 1 ppm. No such responses were observed at either time point for exposures to 0.12
9 ppm O3.
10
11 AX5.2.5 Effects on Pulmonary Function
12 AX5.2.5.1 Introduction
13 Numerous pulmonary function studies of the effects of short-term O3 exposure (defined here
14 as < 1 week of exposure) in several animal species have been conducted and generally show
15 responses similar to those of humans (e.g., increased breathing frequency, decreased tidal volume,
16 increased resistance, decreased forced vital capacity [FVC] and changes in the expiratory flow-
17 volume curve). The breathing pattern returns to normal after O3 exposure.
18 This section will provide a brief overview of acute and short-term exposure effects and then
19 focus on functional changes observed after long-term exposure to O3 (defined here as >1 week of
20 exposure) and on the new studies on O3-induced airway hyperresponsiveness (AHR).
21
22 AX5.2.5.2 Acute and Short-Term Exposure Effects on Pulmonary Function
23 Wiester et al. (1996) exposed male Fischer 344 rats to 0.5 ppm O3 for either 6 or 23 h/day
24 over 5 days. Ozone-induced changes in lung volume were attenuated during the 5 exposure days
25 and returned to control levels after 7 days recovery. The responses to repeated O3 exposure in
26 rats were exacerbated by reduced ambient temperature, presumably as a result of increased
27 metabolic activity.
28 Recent work has utilized inbred mouse strains with varying ventilatory responses to O3 to
29 attempt to model susceptible populations. As differences were seen in inflammatory responses to
30 acute O3 exposures in C57BL/6J and C3H/HeJ mice, comparisons were made of their ventilatory
31 responses also (Tankersley et al., 1993). Following an exposure of 2 ppm O3 for 3 h, breathing
December 2005 AX5-55 DRAFT-DO NOT QUOTE OR CITE
-------
1 frequency (f), tidal volume (VT), and minute ventilation were measured 1 and 24 h in both
2 normocapnia (or air at -0% CO2) and hypercapnia (5 or 8% CO2). They demonstrated that acute
3 O3 exposures caused altered hypercapnic ventilatory control, which varied between strains. The
4 observations from this study indicate that control of ventilation is at least in part regulated by
5 other genetic factors.
6 The Paquette et al. (1994) study discussed in 5.3.3.3 also measured ventilatory responses in
7 C57BL/6J and C3H/HeJ mice on repeated subacute exposures to O3. C57BL/6J and C3H/HeJ
8 had differing responses to both normocapnia and hypercapnia. Normocapnic VE was greater
9 following subacute O3 exposure in C57BL/6J mice than in C3H/HeJ mice, due to increased and
10 reduced VT, respectively. The authors speculated that this increased VT in C57BL/6J mice may
11 contribute to the increased susceptibility to lung injury due to a greater dose of O3 reaching the
12 lower lung. Hypercapnic ventilatory responses following subacute O3 exposures demonstrated
13 reduced VE (due to decreased VT) in C57BL/6J only. Evaluations of O3 dosimetry were
14 performed in these two strains using 18O-labeled O3 (Slade et al., 1997). Immediately after
15 exposures to 2 ppm 18O3 for 2-3 h, C3H/HeJ mice had 46% less 18O in lungs and 61% less in
16 trachea, than C57BL/6J. Additionally, C3H/HeJ mice had a greater body temperature decrease
17 following O3 exposure than C57BL/6J mice. The authors suggested that the differences in
18 susceptibility to O3 are due to differences the ability to decrease body temperature and,
19 consequently decrease the dose of O3 to the lung.
20 Takahashi et al. (1995b) measured tracheal transepithelial potential (VT) in eight mouse
21 strains 6 h after exposure to 2 ppm O3 for 3 h. AKR/J, C3H/HeJ, and CBA/J were identified as
22 resistant strains and 129/J, NJ, C57BL/6J, C3HeB/FeJ and SJL/J were identified as susceptible
23 strains. The pattern of inheritance for this trait suggested an autosomal recessive pattern. The
24 authors noted that strains' responses to this parameter did not show concordance with
25 inflammatory responses, suggesting to the authors that the two phenotypes are not controlled by
26 the same genetic factors.
27 The Savov et al. (2004) study discussed in 5.2.3.3 (2.0 ppm O3 for 3 h) characterized
28 ventilatory responses using whole body plethysmography and enhanced pause index (Penh) in the
29 nine mouse strains evaluated. Table AX5-4 lists the baseline Penh, the Penh following O3, and the
30 Penh response to methacholine (MCh) following O3. C57BL/6J was hyporeactive to MCh prior to
December 2005 AX5-56 DRAFT-DO NOT QUOTE OR CITE
-------
1 O3, but was very responsive to MCh following O3. Conversely, C3H/HeJ had an intermediate
2 baseline Penh and a small response to MCh following O3 exposure.
3
4 AX5.2.5.3 Long-Term Exposure Effects on Pulmonary Function
5 New long-term O3 exposure studies evaluating pulmonary function are not available.
6
7 AX5.3.5.4 Acute and Chronic Exposure Effects on Airway Responsiveness
8 New studies in laboratory animals allow possible ways of predicting, with increased
9 specificity, the effects of O3 exposure on the exacerbation of asthma symptoms and the risk of
10 developing asthma in humans. A variety of methods have been used to assess airway
11 responsiveness in humans, including airway challenge with nonspecific bronchoconstrictors (e.g.,
12 inhaled methacholine or histamine) and with indirect (e.g., inhalation of adenosine
13 monophosphate, hypertonic saline, mannitol) stimuli to bronchoconstriction (Anderson, 1996).
14 Although inhaled agonist challenges are preferred in humans, laboratory animals studies have
15 employed intravenous (i.v.) agonist challenges as well as inhalation challenges. The
16 comparability of these two routes for bronchoconstrictor administration has not been well studied,
17 however differences have been reported (e.g., Sommer et al., 2001). Most challenge tests require
18 an outcome measure that reflects airway function, such as pulmonary resistance, dynamic lung
19 compliance, or deceased forced expiratory flow and volume. In a generalized sense, resistance is
20 a measure of large airway function, and dynamic compliance is a measure of small airway
21 function. In regards to pulmonary mechanics, human infants are much like laboratory rodents
22 because both have very compliant chest walls. Therefore, this discussion will make comparisons
23 from the published literature on airway responsiveness in human infants as recently reviewed, for
24 example, by Stick, 2002.
25 As with infants, a limitation of the traditional types of studies in laboratory rodents is the
26 requirement for sedation. The need for artificial ventilation in laboratory animal studies may
27 cause breathing patterns that affect O3 deposition (see Section 5.2). load et al. (2000) reported
28 that when 1 ppm O3 for 90 min is administered to isolated rat lung at either 2.4 ml/40 bpm or
29 1.2 m/80 bpm, the more rapid breathing pattern elicits less epithelial cell injury than the slower
30 breathing pattern. They further showed greater reduction in injury in the proximal axial airway
31 compared to its adjacent airway branch and terminal bronchiole. For rats, normal respiration is
December 2005 AX5-57 DRAFT-DO NOT QUOTE OR CITE
-------
1 approximately 100 bpm, so this paradigm does not really model rapid shallow breathing elicited
2 in the intact animal. Schelegle et al. (2001) showed that the large conducting airways of rats are
3 protected by rapid, shallow breathing, but there is a more even distribution of epithelial cell injury
4 to the terminal bronchioles. Recent observations (Postlethwait et al., 2000) demonstrate that the
5 conducting airways are the primary site of acute cytotoxicity from O3 exposure. By utilizing a
6 new analytic approach of three-dimensional mapping of the airway tree in SD rat isolated lung
7 exposed to 0, 0.25, 0.5, or 1.0 ppm O3 for 20 to 90 minutes, they showed a concentration-
8 dependent increase in injured cells. Injury was evident in proximal and distal conduction
9 airways, lowest in terminal bronchioles, and highest in the small side branches downstream of
10 bifurcations. These exposure levels did not concurrently elicit changes in LDH activity or total
11 protein in BALF.
12 More recent methods of studying laboratory animals utilize unanesthetized, unrestrained
13 rodents in a whole-body plethysmograph (e.g., Shore et al., 2001, 2002; Goldsmith et al., 2002;
14 Jang et al., 2002), but pulmonary resistance is measured indirectly using several indices of
15 inspiratory/expiratory pressure differences, including enhanced pause (Penh), that may be less
16 sensitive than direct measurements of lung airflow resistance (Murphy, 2002). Also, in another
17 study by Sommer et al. (1998), unrestrained guinea pigs were shown to have a daily variability in
18 pulmonary resistance that is similar to that occurring in humans. Therefore, circadian rhythms of
19 airway caliber must be considered when performing airway challenge tests in any species.
20 Animals with acute viral illness have morphological evidence of inflammatory cell infiltration,
21 bronchiolar wall edema, epithelial hyperplasia, and increased airway mucous plugs that can cause
22 airway narrowing, air trapping, and serious functional changes in the lung (Folkerts et al., 1998).
23 Exercise-induced bronchoconstriction in humans appears to be mediated by changes in the
24 tonicity of the airway lining fluid (Anderson and Daviskas, 2000) and, therefore, a test in
25 laboratory animals based on the inhalation of mannitol aerosol (hyperosmolar) might be feasible
26 and provide information similar to that from exercise challenges in cooperative children and
27 adults (Brannan et al., 1998). In active humans with asthma, adenosine monophosphate
28 challenges appear to better reflect ongoing airway inflammation than histamine or methacholine
29 challenges (Polosa and Holgate, 1997; Avital et al, 1995a,b), and might be useful in identifying
30 mechanisms of asthma in laboratory animals and their responsiveness to environmental
31 pollutants.
December 2005 AX5-58 DRAFT-DO NOT QUOTE OR CITE
-------
1 Airway responsiveness in asthma
2 The increased responsiveness to bronchoconstrictor challenge in asthma is thought to result
3 from a combination of structural and physiological factors that include increased inner-wall
4 thickness, increased smooth-muscle responsiveness, and mucus secretion. This baseline
5 responsiveness is thought to be modulated in asthma by chronic inflammation and airway
6 remodeling (Stick, 2002). Longitudinal studies in adults have shown that the development of
7 airway responsiveness is associated with persistence of symptoms (O'Conner et al., 1995).
8
9 Airway responsiveness in infants
10 The age at which nonspecific airway hyperresponsiveness first appears in humans is
11 unknown, although both genetic and environmental factors are most likely to play a role.
12 Although underlying physiological or structural factors may determine this relative increase in
13 responsiveness in infants compared with older children, the most likely explanation is that infants
14 receive a relatively larger dose of inhaled challenge agent than older children. Thus, when a
15 correction is made for this dose effect, infants and older children appear to have a similar
16 response to inhaled histamine (Stick et al., 1990; Stick, 2002). Airway responsiveness at one
17 month is a predictor of lung function at six years (Palmer et al., 2001). Data from this study also
18 show that the genetic determinants of atopy and airway responsiveness are independent (Palmer
19 et al., 2000). In another study of infants with wheeze, persistence of airway hyperresponsiveness
20 was associated with persistence of symptoms, although airway responsiveness at one month of
21 age was neither a sensitive nor a specific predictor of outcome (Delacourt et al., 2001).
22 The human studies imply that airway responsiveness is a key factor in asthma, but it is not
23 clear if the factors that are important for airway responsiveness in early life are related to
24 inflammation, structure or physiology of the airways, or the combination of all three.
25 Furthermore, it is not clear how viruses, allergens and irritants in the environment modify innate
26 airway responses (Holt et al., 1999), but they are known to be important.
27
28 Airway responsiveness in laboratory animals
29 Laboratory animals, including rodents (mice, rats, guinea pigs), rabbits, cats, dogs, and
30 nonhuman primates have been used to study the effects of O3 exposure on airway
31 bronchoconstriction. New studies examining airway responsiveness in laboratory animals are
December 2005 AX5-59 DRAFT-DO NOT QUOTE OR CITE
-------
1 listed in Table AX5-7. Ozone-induced AHR in guinea pigs has been used as a model of
2 bronchospasm (e.g., Kudo et al., 1996; van Hoof et al., 1996; 1997a,b; Matsubara et al., 1997a,b;
3 Sun and Chung, 1997; Aizawa et al., 1999a,b; Tsai et al., 1998; Nakano et al., 2000). In this
4 model, the guinea pigs are acutely exposed for 1 or 2 h to high O3 concentrations (2 to 3 ppm).
5 The model is useful for understanding mechanisms of bronchospasm, but are not directly relevant
6 for extrapolation to potential airway responses in humans exposed to ambient levels of O3. Dye
7 et al. (1999) showed hyperresponsiveness to methacholine in rats 2 h after exposure to 2 ppm
8 O3 for 2 h.
9 Shore et al. (2000) have shown that O3-induced AHR is reduced in immature SD rats. The
10 animals were exposed to 2 ppm O3 for 3 h in nose-only-exposure plethysmographs and baseline
11 VE was normalized for body weight. VE was reduced, primarily as a result of decreased VT, with
12 O3 exposure. Adults rats had 40-50% decreases in VE, 6-wk-old rats had smaller decreases, and
13 2- and 4 week-old rats had no significant changes in VE.. This suggested to the authors that the
14 higher baseline VE in the young rats, combined with the smaller decreases in VE with O3
15 exposure, created a much larger dose in the immature rats. Shore et al. (2002) completed
16 complementary studies in A/J mice at ages 2, 4, 8, or 12 weeks using exposures of 0.3 to 3 ppm
17 for 3 h. Ozone caused a similar concentration-related decreases in VE except in the 2- and 4-week
18 old mice. This suggested that the young mice are less sensitive than adult mice to O3 in terms of
19 AHR. Lean and obese mice were also compared for differences in AHR response to O3 exposure
20 (2.0 ppm O3 for 3 h). Shore et al. (2003) exposed lean, WT C57BL/6J mice and mice with a
21 genetic defect in the gene that codes for leptin, the satiety hormone. These ob/ob mice had
22 enhanced AHR and inflammation compared to the WT mice.
23 Airway hyperresponsiveness can be induced by specific antigens as well as O3. The most
24 commonly used laboratory animal model is the ovalbumin (OVA) sensitized guinea pig.
25 Animals sensitized with OVA have been shown to have similar responses to nonspecific
26 bronchoconstrictors (e.g., carbachol) as control animals; however, OVA-sensitized guinea pigs
27 exposed to O3 showed increased AHR to histamine (Vargas et al., 1994). Guinea pigs were
28 sensitized by inhalation exposure to ovalbumin and subsequently challenged with histamine; the
29 main endpoint was specific airway resistance. When exposed to O3 before sensitization, repeated
30 exposures to very high levels (5.0 ppm) decreased the OVA sensitization threshold; however, in
31 already sensitized animals, a 2-h exposure to > 1.0 ppm enhanced airway responsiveness to
December 2005 AX5-60 DRAFT-DO NOT QUOTE OR CITE
-------
Table AX5-7. Effects of Ozone on Airway Responsiveness
1, J
i
cr
^
to
o
o
^
X
ON
^
O
^
"Tl
H
6
o
2|
0
H
O
Cj
O
H
Ozone
Concentration"
ppm ug/m
0.1 196
0.3 588
0.15 294
0.30 588
0.60 1,176
1.2 2,352
0.3 588
0.5 980
1 1,960
1 1,960
2 3,920
3 5,880
Exposure
Duration
4 h/day,
4 days/week
for 24 weeks
4h
4 h/day for 1,
3, 6, 12, 24,
or 38 days
8 h/day for 5 days,
repeated every
14 days for
6 months
Ih
Ih
2h
Ih
Challenge"
Agent Route
ACh inh
OVA inh
ACh iv
Hist iv
SP iv
SP iv
Hist inh
Ach inh
OVA inh
Mch inh
OVA inh
Mch inh
Hist iv
OVA iv
Species, Sex, Strain,
Drugs and Agec
none Guinea pig, M & F,
Hartley
Na Guinea pig, M
Pentobarbital Hartley,
500-600g
Na Guinea pig, M
Pentobarbital Hartley,
500-600 g
Ketamine + Rhesus monkey,
Diprivan M, 30 days old
Urethane Guinea pig, M,
Dunkin-Hartley,
250-300 g
Urethane Mouse, M,
C57BL/6,
6 weeks old
Ketamine + Rat, M,
Zylazine F344
14 months old
Guinea pig, M,
Hartley,
500-700 g
Observed Effect(s)
O3 exposure did not produce airway hyperresponsiveness to
ACh in nonsensitized animals; in OVA-sensitized animals,
there was increased responsiveness to both nonspecific (ACh)
and specific (OVA) airway challenge that persisted for 4 weeks
after exposure 0. 1 and 0.3 ppm O3. Effects were not gender
specific and were not associated with BALF inflammatory
indicators, but were associated with antigen-specific antibodies
in blood.
Increased airway responsiveness to Hist, but not Ach, 16-18 h
after 1.2 ppm O3 exposure only. Increased responsiveness to
SP occurred after exposure to > 0.3 ppm O3
Increased airway responsiveness to SP occurred 16-18 h after
exposure to 0.3 ppm O3 for 1, 3, 6, 12, and 24 days; but not
after 48 days. Highly significant correlation between airway
responsiveness and BALF total cells, Ams, neutrophils, and
eosinophils, suggesting that airway inflammation is involved.
Increased airway responsiveness to Hist after 10 episodes of
exposure to O3 + HDMA in sensitized infant monkeys.
Increased bronchial responsiveness at 3 h, but not 24 h after O3;
OVA had no effect on baseline, but enhanced airway
responsiveness 24 h after O3
Ozone caused increased Cdyn and VE, and decreased PaO2 in
OVA- sensitized mice
Increased airway responsiveness to MCh 2 h PE
Increased airway responsiveness to histamine after O3 exposure
in OVA sensitized guinea pigs, with enhanced responsiveness
after OVA challenge
Reference
Schlesinger
et al. (2002a,b)
Segura et al.
(1997)
Vargas et al.
(1998
Schelegle et al.
(2003b)
Sun et al.
(1997)
Yamauchi et al.
(2002)
Dye et al.
(1999)
Vargas et al.
(1994)
O
HH
H
W
"Table ordered according to ozone concentration.
bMCh = methylcholine, ACh = acetylcholine, Hist = histamine, 5-HT = 5-hydroxytryptamine, SP = substance P, FS = field stimulation, CCh = carbachol, TX = thromboxane,
KC1 = potassium chloride, Pt = platinum; Route: iv = intravenous, inh = inhalation., sc = subcutaneous, ip = intraperitoneal
°Age or body weight at start of exposure.
-------
1 ovalbumin. Thus, O3 exposure does not modify the development of antigen-induced AHR and, in
2 fact, may enhance AHR at high levels of exposure.
3 The enhancement of antigen-induced bronchoconstriction by acute, high-level O3 was
4 further explored in OVA-sensitized guinea pigs (Sun et al., 1997) and mice (Yamauchi et al.,
5 2002). Male Dunkin-Hartley guinea pigs were sensitized by i.p. injection of OVA (1 mL 20 jig)
6 and exposed to filtered air or 1 ppm O3 for 1 h (Sun et al., 1997). Airway responsiveness to
7 inhaled acetylcholine was measured 3 h and 24 h after air and O3 exposures and BAL was
8 performed. Four other groups of OVA-sensitized animals were exposed to OVA aerosol or to O3
9 alone, or in combination, and airway responsiveness to acetylcholine was measured 3 h and
10 24 h PE. In the combined exposure groups, OVA aerosol exposure was initiated either
11 immediately after, or 21 h after O3 exposure. In this study, O3 exposure increased bronchial
12 responsiveness to acetylcholine at 3 h, but not 24 h, while OVA alone had no effect. Combined
13 exposure to O3 and OVA (1 ppm for 1 h, then 3 min OVA) increased bronchial responsiveness to
14 acetylcholine 3 h after O3 exposure. At 24 h following O3 exposure, AHR increased when OVA
15 challenge was performed at 21 h, suggesting to authors that O3 pre-exposure can potentiate OVA-
16 induced AHR. Neutrophil counts in the BALF increased at 3 and 24 h after O3 exposure alone,
17 but were not further increased when O3 exposure was combined with OVA airway challenge;
18 however protein content of the BALF did increase at 3 and 24 h in the O3 and OVA groups.
19 Thus, this study also indicates that high-ambient O3 exposure can augment antigen(OVA)-
20 induced AHR in guinea pigs.
21 Male C57BL/6 mice were sensitized by i.p. injection of OVA (50 jig) and exposed to
22 filtered air or 1 ppm O3 for 1 h (Yamauchi et al., 2002). Airway responsiveness to methacholine
23 was measured 24 h after an inhalation challenge to OVA (10 mg/mL) in OVA-sensitized and
24 control groups, and 3 h after the OVA inhalation challenge in animals exposed to O3 to maximize
25 the susceptible time period for AHR (Yamaguchi et al., 1994). Pulmonary function was
26 measured by plethysmography before and at 10-min intervals during the 1-h air and O3 exposures.
27 Blood gases and BAL fluid were monitored immediately after exposure. Mice sensitized to OVA
28 had AHR to methacholine. Ozone exposure caused significant decreases in dynamic lung
29 compliance, minute ventilation, and PaO2 in OVA-sensitized mice, but not in controls. A marker
30 of inflammation (soluble intercellular adhesion molecule-1 [sICAM-1]) was elevated in the BAL
31 fluid of OVA-sensitized mice, but sICAM-1 levels were not significantly changed by O3
December 2005 AX5-62 DRAFT-DO NOT QUOTE OR CITE
-------
1 exposure, indicating that the O3-induced AHR to methacholine was not caused by O3-induced
2 inflammation.
3 Ozone-induced AHR may be temporally associated with neutrophils (DeLorme et al., 2002)
4 and other inflammatory cells stimulated by leukotrienes (Stevens et al., 1995a), cytokines (Koto
5 et al., 1997), mast cells (Igarashi et al., 1998; Noviski et al., 1999), or by oxygen radicals
6 (Takahashi et al., 1993; Stevens et al., 1995b; Tsukagoshi et al., 1995; Kudo et al., 1996). Two
7 new studies have shown that inflammation is not a prerequisite of AHR (Stevens et al., 1994;
8 Koto et al., 1997), and some investigators have suggested that O3-induced AHR may be
9 epithelium dependent (Takata et al., 1995; Matsubara et al., 1995; McGraw et al., 2000). For
10 example, neonatal rats pretreated with capsaicin, which will permanently destroy C-fibers and
11 prevent O3-induced release of neuropeptides (Vesely et al., 1999b), and then exposed to O3 when
12 adults, showed a marked increase in airway responsiveness to inhaled aerosolized methacholine
13 (Jimba et al., 1995). Some investigators (Matsumoto et al., 1999; DeLorme et al., 2002) have
14 shown that respective intravenous pretreatment with neutrophil elastase inhibitor or PMN
15 antiserum can block O3-induced AHR; other investigators (Koto et al., 1995; Aizawa et al., 1997;
16 Takebayashi et al., 1998) have shown that depletion of tachykinins by capsaicin treatment, or by a
17 specific tachykinin receptor antagonist, can block the induction of AHR by O3. The seemingly
18 disparate responses in laboratory animals may be due to species- or strain-specific differences in
19 inherent reactivity to bronchoconstrictors, or to inherent differences in susceptibility to O3-
20 induced inflammation (Zhang et al., 1995; Depuydt et al., 1999; Dye et al., 1999).
21 The studies referenced above are useful for gaining an understanding of how acute
22 O3 exposure modulates specific airway responsiveness to allergen challenge, but the
23 O3 concentrations used in these studies are not typical of ambient exposures in the population.
24 More recently published studies that may be potentially relevant to ambient levels of O3 were
25 conducted in vivo, in an isolated perfused lung model, and in ex vivo lung segments using
26 multihour and repeated multihour exposures with ambient levels of O3. A study on the
27 relationship between O3-induced AHR and tracheal epithelial function was conducted in New
28 Zealand white rabbits by Freed et al. (1996). Rabbits were exposed to filtered air or to 0.2 ppm
29 O3 for 7 h. Tracheal transepithelial potential difference (PD) was measured 3 h after exposure
30 and lung resistance and reactivity were partitioned into central and peripheral components using
31 forced oscillation. Exposure to O3 significantly decreased PD, but did not change lung resistance.
December 2005 AX5-63 DRAFT-DO NOT QUOTE OR CITE
-------
1 Changes in the compartmentalized lung resistance, measured in response to bronchoconstrictor
2 aerosol challenge (acetylcholine) before and after bilateral vagotomy, were not significantly
3 different in air-exposed rabbits; however, bilateral vagotomy did enhance peripheral lung
4 reactivity in O3-exposed rabbits. The acetylcholine-induced increase in lung resistance with O3
5 exposure (140%) was two times higher than with air exposure, indicating that ambient-level O3
6 exposure affects tracheal epithelial function in rabbits and increases central airway reactivity,
7 possibly through vagally-mediated mechanisms.
8 Delaunois et al. (1998) studied pulmonary mechanics and hemodynamics in the isolated
9 perfused lung model that allowed partitioning of the total pressure gradient into arterial, pre- and
10 post-capillary, and venous components. New Zealand white rabbits were exposed to filtered air
11 or to 0.4 ppm O3 for 4 h and evaluated for airway responsiveness to acetylcholine, substance P, or
12 histamine immediately or 48 h later. Ozone exposure did not significantly change baseline values
13 of pulmonary resistance and dynamic compliance, but inhibited pulmonary mechanical reactivity
14 to all three bronchoconstrictors that persisted for 48 h. Ozone also modified vasoreactivity of the
15 vascular bed, but only at 48 h PE. Arterial segmental pressure, normally insensitive to
16 acetylcholine and substance P, was significantly elevated by O3; precapillary segmental pressure
17 decreased in response to acetylcholine. The authors concluded that O3 can induce direct vascular
18 constriction, but the vascular responses are variable and depend on the agonist used and on the
19 species studied.
20 Guinea pigs were exposed to filtered air, 0.15, 0.3, 0.6, or 1.2 ppm O3 for 4 h and evaluated
21 for airway responsiveness to acetylcholine, substance P, or histamine 16 to 18 h later (Segura
22 et al., 1997). Ozone did not cause airway hyperresponsiveness to acetylcholine or histamine,
23 except at the highest concentration (1.2 ppm O3) for histamine. However, O3 did cause
24 hyperresponsiveness to substance P at > 0.3 ppm. The authors speculated that O3 destroys neutral
25 endopeptidases, responsible for substance P inactivation, that are located in airway epithelial
26 cells. In a follow-up study at the same laboratory, Vargas et al. (1998) reported that guinea pigs
27 chronically exposed to 0.3 ppm O3 for 4 h/day became adapted to substance P-induced AHR.
28 Ozone caused increased sensitivity to substance P after 1, 3, 6, 12, and 24 days of exposure that
29 was associated with airway inflammation; however, after 48 days of exposure, the increased
30 sensitivity to substance P was lost.
December 2005 AX5-64 DRAFT-DO NOT QUOTE OR CITE
-------
1 The effects of repeated short-term exposure or long-term exposure to O3 on airway
2 responsiveness have been investigated in several laboratory animal studies. Both Vargas et al.
3 (1998), discussed above, and Szarek et al. (1995) reported that the AHR associated with acute O3
4 exposures does not persist during long-term exposure to ambient-levels of O3 (< 1 ppm). In the
5 Szarek et al. (1995) study, Fischer 344 rats were exposed to 0.0, 0.12, 0.5, or 1.0 ppm O3, 6h/day,
6 5 days/week for 20 months. Eighth generation airway segments were isolated from the exposed
7 rats and circumferential tension development was measured in response to bethanechol,
8 acetylcholine, and electrical field stimulation and normalized to smooth muscle area. Maximum
9 responses of the small bronchi of male rats were significantly reduced after exposure to 0.12 and
10 0.5 ppm O3, suggesting some adaption had taken place during long-term exposure, possibly
11 increased inner wall thickness.
12 Changes in breathing pattern and lung function caused by O3 are attenuated with repeated
13 daily exposures for at least 3 to 5 days. load et al. (1998), however, reported that repeated daily
14 O3 exposure enhances, rather than diminishes, the responsiveness of rapidly adapting airway
15 receptors. Guinea pigs were exposed to 0.5 ppm O3, 8 h/day for 7 days and then studied for
16 measurement of impulse activity of the rapidly adapting receptors, dynamic lung compliance, and
17 lung resistance at baseline and in response to substance P, methacholine, hyperinflation, and
18 removal of end-expired airway pressure. Repeated exposure increased receptor activity to
19 substance P, methacholine, and hyperinflation; there were no significant effects on baseline or
20 substance P- and methacholine-induced changes in lung compliance and resistance. Because
21 agonist-induced changes in receptor activity precede lung function changes, the authors
22 concluded that the responsiveness of rapidly adapting receptors was enhanced.
23 Schlesinger et al. (2002a,b) evaluated airway responsiveness following acetylcholine or
24 OVA inhalation challenges in male and female Hartley guinea pigs exposed to 0.1 and 0.3 ppm
25 O3, 4 h/day, 4 days/week for 24 weeks. The bronchoprovocation tests were performed at 4 week
26 intervals during exposure and at 4- to 8-week intervals during the PE period in nonsensitized
27 animals and in animals sensitized to allergen (OVA) prior to, or concurrent with, O3 exposure.
28 Ozone exposure did not cause AHR in nonsensitized animals, but did exacerbate AHR to both
29 acetylcholine and OVA in sensitized animals that persisted for 4 weeks after exposure. The
30 effects of O3 on airway responsiveness were gender independent and they were concentration-
31 related for the acetylcholine challenges. The study did not show any evidence of adaptation.
December 2005 AX5-65 DRAFT-DO NOT QUOTE OR CITE
-------
1 Schelegle et al. (2003a) evaluated airway responsiveness in infant rhesus monkeys exposed
2 to a 5 day O3 episode repeated every 14 days over a 6-month period. Half of the monkeys were
3 sensitized to house dust mite allergen (Dermatophagoides farinae) at 14 and 28 days of age
4 before exposure to a total of 11 episodes of O3 (0.5 ppm, 8 h/day for 5 days followed by 9 days of
5 FA), house dust mite allergen aerosol (HDMA), or O3 + HDMA. Monkeys were sedated for
6 measurement of airway responsiveness and then anesthetized for measurement of pulmonary
7 mechanics (e.g., Raw, R,.s) using a head-out body plethysmograph. A necropsy was performed
8 after all pulmonary function measurement were taken (see the previous Section 5.3.4 for results).
9 The HDMA and histamine aerosol challenges were administered until Raw doubled. Data were
10 expressed as the concentration increasing Raw by 150% (EC 150 Raw). Other measurements
11 included VT, fB, and SaO2 (estimated by pulse oximeter). Baseline Raw was significantly elevated
12 after 10 exposure episodes in the HDMA + O3 group compared to the FA, HDMA, and O3
13 exposure groups. Aerosol challenge with HDMA at the end of the 10th episode did not
14 significantly affect Raw, VT, fB, or SaO2. Aerosol challenge with histamine was not significantly
15 different after 6 episodes; however, the EC 150 Raw for the HDMA + O3 group was significantly
16 reduced after 10 episodes when compared to the FA, HDMA, and O3 exposure groups, indicating
17 the development of airway hyperresponsiveness in this group sometime between episodes 6
18 and 10.
19 Using 18O exposures at 1 ppm for 2 h and breathing frequencies of 80, 120, 160, or
20 200 breaths/minute, Alfaro et al., (2004) examined the site-specific deposition of 18O. At all
21 frequencies, parenchymal areas had a lower content of 18O than trachea and bronchi.
22 As breathing frequency increased from 80 to 160 bpm, the deposition showed a reduction in
23 midlevel trachea and an increase in both mainstream bronchi. At this frequency there was also an
24 increase in deposition in parenchyma supplied by short (cranial) airway paths. At 200 bpm 18O
25 deposition in trachea increased, concurrent with increases in right cranial and caudal bronchi
26 regions. Right cranial parenchymal content decreased at 200 bpm, whereas right caudal
27 parenchymal levels did not change at any breathing frequency. The authors list some limitations
28 of this study, such as the possible effect on regional distribution of ventilation by the use of the
29 negative-pressure ventilator, the effect of paralysis on airway geometry, and possible
30 translocation of 18O during the 2 h exposure period. But the evidence provided by these studies
31 strongly suggests that the effect of rapid, shallow breathing is to create a more evenly distributed
December 2005 AX5-66 DRAFT-DO NOT QUOTE OR CITE
-------
1 injury pattern, with possibly greater protection from focal injury to the large conducting airways
2 including the trachea and the left mainstem bronchus.
3 Schelegle et al. (2003b) examined adaptive phenomena in SD rats using an exposure
4 paradigm consisting of 5 days of daily 8 h 1 ppm O3 exposures followed by 9 days of recovery in
5 filtered air. This O3/FA pattern was repeated for 4 cycles and animals were analyzed on day 1
6 and day 5 of each exposure and at the end of the filtered air period. The O3 -induced rapid
7 shallow breathing pattern followed by adaptation occurred with each cycle, however, the release
8 of substance P from the trachea , the neutrophil content, and cell proliferation, as visualized by
9 BrdU labeling, became attenuated after the first cycle, thus displaying a disconnect from the rapid
10 shallow breathing response. The repeated cycles of O3 also created hypercellularity of the CAR
11 epithelium and thickening of the CAR interstitium, not linked to changes in cell proliferation.
12 The authors hypothesize a mechanism of injury from repeated O3 exposures that consists of: (1)
13 diminished neutrophilic inflammation/and or release of mitogenic neuropeptides, (2) depressed
14 cell proliferative response, and (3) cumulative distal airway lesion.
15 Using a subset of monkeys from the same study (Schelegle et al., 2003b) reported above,
16 Chen et al. (2003) reported that attenuation of O3-induced rapid shallow breathing and lung
17 function changes typically seen with repeated O3 exposure may be caused by the adaptation of the
18 respiratory motor responses. The monkeys were killed 3 to 5 days after exposure to
19 11 "episodes" of O3 and brain stem coronal slices were prepared. Whole cell recordings were
20 performed on neurons from the nucleus tractus solitarius (NTS), the brain stem region that
21 processes lung sensory signals. Episodic O3 exposure resulted in neuroplasticity of the NTS
22 including increased nonspecific excitability of the NTS neurons, an increased input resistance,
23 and an increased spiking response to intracellular injections of depolarizing current.
24
25 AX5.2.6 Genotoxicity Potential of Ozone
26 Many experimental studies have been conducted to explore the mutagenic and carcinogenic
27 potential of O3. Recently published in vivo exposure studies found increased DNA strand breaks
28 in respiratory cells from guinea pigs (Ferng et al., 1997) and mice (Bornholdt et al., 2002) but
29 only after high O3 exposures (1 ppm for 72 h and 1 or 2 ppm for 90 min, respectively).
30 Witschi et al. (1999) exposed female strain A/J mice to 0.12, 0.50, and 1.0 ppm O3 for 6
31 h/day, 5 days/week for up to 9 months. After 5 months, one-third of the O3-exposed mice were
December 2005 AX5-67 DRAFT-DO NOT QUOTE OR CITE
-------
1 compared to one-half of the controls exposed to filtered air. There was no statistically significant
2 difference in lung tumor multiplicity or incidence. The remaining O3-exposed mice were split
3 into two groups. Ozone exposure continued in one group for an additional 4 months and the mice
4 in the second group were allowed to recover in filtered air. Again, there were no statistically
5 significant differences in lung tumor multiplicity between control mice and mice exposed to any
6 concentration of O3 for 9 months. The highest, and only statistically significant lung tumor
7 incidence, was found in the mice exposed to 0.5 ppm O3. In the O3-exposed mice allowed to
8 recover in filtered air, only the mice exposed to 0.12 ppm O3 had statistically significant increases
9 in lung tumor incidence and multiplicity. These results were considered by the authors to be
10 spurious and of no significance for data interpretation.
11 Kim et al. (2001) evaluated the effects of O3 inhalation exposure in B6C3FJ mice.
12 No increased incidence of lung tumors was found after exposure to 0.5 ppm O3 for 6 h/day,
13 5 days/week for 12 weeks. There were statistically significant differences in mean body weight
14 between O3-exposed mice and air-exposed controls, as well as significant differences in the mean
15 absolute and relative weights of several organs (e.g., liver, spleen, kidney, testes, and ovary).
16 Histopathologic examination of major organs revealed oviductal carcinomas in 3/10 O3-exposed
17 female mice.
18
19
20 AX5.3 SYSTEMIC EFFECTS OF OZONE EXPOSURE
21 Mathematical models of O3 dosimetry predict that essentially no O3 penetrates to the blood
22 of the alveolar capillaries, and thus is unlikely to enter the bloodstream (see Section 5.2).
23 However, numerous studies have indicated that inhalation of O3 can produce effects in
24 lymphocytes, erythrocytes, and serum, as well as several organ systems. The mechanism by
25 which O3 causes such systemic changes is unknown but it seems most likely that some reaction
26 product of O3, which then penetrates to the blood and is transported to some target site, is a
27 probable mechanism. Extra-pulmonary effects could also be due to the exposure-related
28 production of mediators, metabolic products and cell trafficking. Although systemic effects are
29 of interest and indicate a very broad array of O3 effects, they are of limited influence and difficult
30 to interpret. By protecting from respiratory tract effects, these systemic effects will likely be
December 2005 AX5-68 DRAFT-DO NOT QUOTE OR CITE
-------
1 protected against also. New studies of systemic effects are discussed here and summarized in
2 Table AX5-8.
3
4 AX5.3.1 Neurobehaviorial Effects
5 Acute exposures to increasing O3 concentrations affect animal behavior (see Table AX5-8
6 for details of studies). Recently reported studies in adult laboratory animals confirm that relevant
7 O3 concentrations from 0.2 to 1.0 ppm can decrease motor activity and affect short- and long-term
8 memory. This has been shown in passive avoidance conditioning studies in rats (Rivas-Arancibia
9 et al., 1998; Avila-Costa et al., 1999; Dorado-Martinez et al., 2001), and in water-maze learning
10 tasks in mice (Sorace et al., 2001). The effects have been attributed to reactive oxygen/nitrogen
11 species and/or ozonation products. The memory deficits could be blocked by administration of
12 vitamin E (Guerrero et al, 1999) or taurine (Rivas-Arancibia et al., 2000). Rivas-Arancibia et al.
13 (2003) demonstrated in rats that 1 ppm of O3 for 4 h caused increased freezing and decreased
14 exploratory behaviors that were accompanied by decreased serotonin levels and increased levels
15 of NO, glutamate, dopamine and striatal lipoperoxidation. Morphological changes were also
16 observed in O3-exposed animals, including neuronal cytoplasm and dendrite vacuolation and
17 dilation of rough endoplasmic reticulum cisterns, which the authors interpret as a
18 neurodegenerative process resulting from the oxidative stress of acute O3 exposure. A recent
19 study by Nino-Cabrera et al. (2002) reports that 0.7 ppm O3 exposure for 4 h can induce
20 ultrastructural alterations in the hippocampus and prefrontal cortex in aged rats, areas of the brain
21 where degenerative age-related changes in learning and memory functions have been reported
22 (Bimonte et al., 2003).
23 In a series of studies reviewed by Paz (1997), animals acutely exposed to O3 concentrations
24 from 0.35 to 1.0 ppm demonstrated significant alterations of electroencephalographic (EEG)
25 patterns during sleep. For example, rats and cats both showed loss of paradoxical sleep time after
26 2 to 8 h of O3 exposure (Paz and Bazan-Perkins, 1992; Paz and Huitron-Resendiz, 1996).
27 A permanent 50% loss of paradoxical sleep time, as well as increased total wakefulness and
28 alterations in circadian rhythm, were shown in rat pups born to dams exposed to 1.0 ppm O3
29 during the entire period of gestation (Haro and Paz, 1993). The sleep pattern effects were
30 associated with alterations in brain neurotransmitter levels (Huitron-Resendiz et al., 1994;
31 Gonzalez-Pina and Paz, 1997) and most likely caused by O3 reaction products or prostaglandins
December 2005 AX5-69 DRAFT-DO NOT QUOTE OR CITE
-------
o
O
to
O
O
Table AX5-8. Systemic Effects of Ozone
X
Ozone Concentration
ppm
jig/m3 Duration
Species
Effects3
Reference
NEUROBEHAVIORAL EFFECTS
0.1
0.2
0.5
1.0
0.1
0.4
0.7
1.1
1.5
0.3
0.6
196 4h
392
980
1,960
196 4h
784
1,372
2,156
2,940
588 30 days
1,176
Rat
Wistar
male
Rat
Wistar
male
Mouse
CD-I
M, F
Rats exposed for 4 h to 0.2, 0.5, and 1 ppm O3 showed long-term memory
deterioration and decreased motor activity, which was reversed 24 h later.
Brain and pulmonary Cu/Zn SOD levels were increased in animals
exposed to 0.1, 0.2, and 0.5 ppm O3, but decreased in animals exposed
to 1 ppm O3.
O3 caused memory impairment at > 0.7 ppm (one trial passive avoidance
test), decreased motor activity at > 1.1 ppm, and increased lipid
peroxidation at > 0.4 ppm. Lipid perioxidation levels from the frontal
cortex, hippocampus, striatum and cerebellum increased with increasing
O3 concentration.
O3 exposure slightly but selectively affected neurobehavioral performance
in male mice assessed with a 5-min open-field test on exposure days 4 and
19 and on day 3 after the end of the exposure. O3 exposure, however, did
Rivas-Arancibia
etal. (1998)
Dorado-Martinez
etal. (2001)
Sorace et al.
(2001)
not grossly affect neurobehavioral development. Reversal learning in the
Morris water maze test was consistently impaired in both prenatally and
adult exposed mice. In addition, longer latency to step-through in the first
trial of the passive avoidance test and a decrease in wall rearing in the hot-
plate test were recorded in O3 prenatally exposed mice. Except for the
first open-field test, altered responses were observed only in animals
exposed to 0.3 ppm O3.
0.35
0.75
1.5
0.7
686 12 h
1,470
2,940
1,372 4 h
Rat
Wistar
male
Rat
O3 exposure decreased paradoxical sleep after 2 h of exposure, and
increased slow wave sleep after 12 h of exposure at all O3 concentrations;
5-HT concentrations in the pons increased with increasing O3
concentration.
Vitamin E administered before or after O3 exposure blocked memory
deterioration (passive avoidance)and increases in lipid peroxidation levels
in the striatum, hippocampus and frontal cortex that were associated with
oxidative stress.
Paz and Huitron-
Resendiz (1996)
Guerrero et al.
(1999)
-------
o
O
to
O
O
Table AX5-8 (cont'd). Systemic Effects of Ozone
X
Ozone Concentration
ppm Hg/m3 Duration Species
NEUROBEHAVIORAL EFFECTS (cont'd)
0.7 1,372 4h Rat
Wistar
male
27 months old
0.7 1,372 4h Rat
0.8 1,568
1 1,960 12h/day Rat
during dark
period
1 1,960 4h Rat
Wistar
male
1 1,960 3h Rat
1.5 2,940 24 h Rat
Wistar
male
Effects3
O3 exposure increased ultrastructural alterations in the hippocampus and
prefrontal cortex in aged rats compared with controls. These areas are
related to learning and memory functions, which are the first degenerative
aging changes observed.
Taurine (43 mg/kg) given before or after O3 exposure improved memory
deterioration in an age-specific manner. Old rats showed peroxidation in
all control groups and an improvement in memory with taurine. When
taurine was applied before O3, peroxidation levels were high in the frontal
cortex of old rats and the hippocampus of young rats; in the striatum,
peroxidation caused by O3 was blocked when taurine was applied either
before or after exposure.
O3 exposure during pregnancy affects the neural regulation of paradoxical
sleep and circadian rhythm of rat pups 30, 60, and 90 days after birth.
O3 caused alterations in long-term memory and a significant reduction of
dendritic spines. Results provide evidence that deterioration in memory is
probably due to the reduction in spine density in the pyramidal neurons of
the hippocampus.
O3 or its reaction products affect the metabolism of major neurotransmitter
systems as rapidly as after 1 h of exposure. There were significant
increases in dopamine (DA), and its metabolites noradrenaline (NA) and
3,4 dihydroxyphenylacetic acid (DOPAC), and 5-hydroxyindolacetic acid
(5-HIAA) in the midbrain and the striatum.
Adult rats exposed to O3 spend decreased time in wakefulness and
paradoxical sleep and a significant increase in time in slow-wave sleep.
Neurochemical changes include increased metabolism of serotonin in
the medulla oblongata, pons, and midbrain.
Reference
Nino-Cabrera
et al. (2002)
Rivas-Arancibia
et al. ( 2000)
Haro and Paz
(1993)
Avila-Costa et al.
(1999)
Gonzalez-Pina
and Paz (1997)
Huitron-Resendiz
etal. (1994)
-------
o
O
to
O
O
Table AX5-8 (cont'd). Systemic Effects of Ozone
X
to
Ozone
ppm
Concentration
jig/m3 Duration
Species
Effects3
Reference
NEUROENDOCRINE EFFECTS
0.5
0.5 to 3
1.0
980 20 h/day
for 5 days
.0 980 to 3 h
5,880
1,960 24 h
Rat
Rat
Sprague-
Dawley
male
Rat
Sprague-
Dawley
male
O3 produced marked neural disturbances in structures involved in the
integration of chemosensory inputs, arousal, and motor control. O3
inhibited tyrosine hydroxylase activity in noradrenergic brainstem cell
groups, including the locus ceruleus (-62%) and the caudal A2 subset
(- 57%). Catecholamine turnover was decreased by O3 in the cortex
(- 49%) and striatum (- 18%) but not in the hypothalamus.
Hyperthyroid, T4-treated rats (0.1 - 1.0 mg/kg/day for 7 days) had
increased pulmonary injury (BALF LDH, albumin, PMNs) at 18 h PE
compared to control rats.
Hyperthyroid, T3 -treated rats had increased metabolic activity and
Oj-induced pulmonary injury, but lipid peroxidation, as assessed by
alkane generation, was not affected.
Cottet-Emard
etal. (1997)
Huffman et al.
(2001)
Sen etal. (1993)
CARDIOVASCULAR EFFECTS
0.1
0.3
0.5
196 5h
588
980
Rat
Wistar young
(4-6 month)
and old
(22-24 month)
Transient rapid shallow breathing with slightly increased HR appeared
1-2 min after the start of O3 exposure, possibly due to olfactory sensation;
persistent rapid shallow breathing with a progressive decrease in HR
occurred with a latent period of 1-2 h. The last 90-min averaged values
for relative minute ventilation tended to decrease with the increase in the
Arito etal. (1997)
level of exposure to O3 and these values for young rats were significantly
lower than those for old rats. An exposure of young rats to 0.1 ppm O3 for
shorter than 5 h significantly decreased the tidal volume and HR and
increased breathing frequency, but no significant changes were observed
in old rats. There were no differences between young and old rats in
non-observable-adverse-effect-levels (NOAELs) for the O3-induced
persistent ventilatory and HR responses, when the NOAELs were
determined by exposure to 0.3 and 0.5 ppm O3.
-------
o
O
to
O
O
Table AX5-8 (cont'd). Systemic Effects of Ozone
X
Ozone Concentration
ppm Hg/m3 Duration
CARDIOVASCULAR EFFECTS (cont'd)
0.1 196 8h/day
0.3 588 for 4 days
0.5 980
0.25 to 2.0 490 to 2 h to 5 days
3920
0.5 588 6 h/day
23 h/day
for 5 days
0.5 588 8h
Species
Rat
Wistar
male
Rat
Mouse
Guinea pig
Rat
F-344
male
Rat
F-344
male
Effects3
Orcadian rhythms of HR and core body temperature were significantly
decreased on the first and second O3 exposure days in a concentration
dependent manner, and returned to control levels on the third and fourth
days.
Robust and consistent decreases in HR and core body temperature; smaller
decreases in metabolism, minute ventilation, blood pressure, and cardiac
output that vary inversely with ambient temperature and body mass.
Minimal extrapulmonary effects were observed at a core body temperature
of 34 °C; O3 exposures at 22 and 10 °C produced significant decreases in
heart rate (160 and 210 beats/min, respectively), core body temperature
(2.0 and 3.5 °C, respectively), and body weight (15 and 40 g,
respectively). Decreases in these functional parameters reached their
maxima over the first 2 exposure days and returned to control levels after
the 3rd day of exposure.
O3 exposure increased atrial natriuretic peptides in the heart, lung, and
circulation, suggesting they mediate the decreased BP and pulmonary
edema observed with similar O3 exposures.
Reference
Iwasaki et al.
(1998)
Watkinson et al.
(2001)
Watkinson et al.
(1995);
Highfill and
Watkinson (1996)
Vesely et al.
(1994a,b,c)
-------
o
O
to
O
O
Table AX5-8 (cont'd). Systemic Effects of Ozone
X
Ozone Concentration
ppm
Hg/m3
REPRODUCTIVE AND
0.2
0.4
0.6
0.3
0.6
0.9
0.3
0.6
0.4
0.8
1.2
392
784
1,176
588
1,176
1,764
588
1,176
784
1,568
2,352
Duration
Species
Effects3
Reference
DEVELOPMENTAL EFFECTS
Continuous up
to day 17 of
pregnancy
Continuous up
to postnatal
day 26
Continuous
until
gestational
day 17
Continuous
during
gestation
days 7-17
Mouse
CD-I
Mouse
CD-I
Mouse
CD-I
Mouse
CD-I
No significant effects on either reproductive performance, postnatal
somatic and neurobehavioral development (as assessed by a Fox test
battery) or adult motor activity (including within-session habituation);
some subtle or borderline behavioral deficits were noted, however.
O3 caused subtle CNS effects but did not affect the animals' capability to
learn a reflexive response (limb withdrawal); females exposed to 0.6 ppm
O3 showed a reduced preference for the right paw than both their same-sex
controls and 0.6 ppm males. The effect was more robust in the case of an
organised avoidance response (wall-rearing).
Exposure to O3 did not grossly affect neurobehavioral development,
as assessed by somatic and sensorimotor development (postnatal day
(PND) 2-20), homing performance (PND 12), motor activity (PND 21),
passive avoidance (PND 22-23), water maze performances (PND 70-74),
and response to a nociceptive stimulus (PND 100).
No effect of O3 on reproductive performance; no significant somatic
developmental effects in Cyexposed pups except for a delay in eye
opening that was not concentration dependent.
Petruzzi et
(1995)
Petruzzi et
(1999)
al.
al.
Sorace et al.
(2001)
Bignami et
(1994)
al.
-------
o
O
to
O
O
Table AX5-8 (cont'd). Systemic Effects of Ozone
Ozone Concentration
ppm
jig/m3 Duration
Species
Effects3
Reference
X
REPRODUCTIVE AND DEVELOPMENTAL EFFECTS (cont'd)
0.6
1,176
Continuous
from birth to
weaning
Mouse
CD-I
1.0
2.0
1,960
3,920
3h
Rat
Sprague-
Dawley
female
DeH'Omo et al.
(1995a,b)
Exposure to O3 did not produce any significant impairment of the
acquisition phase during swimming navigation, a sensitive indicator
for hippocampal damage; however, O3 slightly increased the swimming
paths during the last day of the reversal phase. Mice exposed to O3
showed a slightly but significantly higher swimming speed during all
the days, which was unrelated to differences in body weight and to
navigational performances. Moreover, mice exposed to O3 (with the
exception of one animal) had a strong tendency to make turns to the left
while the controls, independent of sex, preferred clockwise turns.
High O3 exposure stimulates hepatocytes to produce increased amounts of Laskin et al.
nitric oxide as well as protein, possibly mediated by cytokines such (1994, 1996,
as TNF-oc produced by alveolar macrophages. When macrophage function 1998); Laskin and
is blocked, hepatic injury induced by O3 is prevented. Laskin (2001)
2.0
3,920
2h
Rat
F-344
Utilizing electron paramagnetic resonance (EPR) spectroscopy of
chloroform extracts of liver homogenates, a significant flux of hydrogen
peroxide produced from the reaction of O3 with lipids of the extracellular
lining could be a source of biologically relevant amounts of hydroxyl
radical. EPR signals for carbon-centred alkoxyl and alkyl adducts were
detected with C-phenyl N-tert-butyl nitrone (PEN) in the liver of animals
exposed to O3.
Vincent et al.
(1996)
-------
o
O
to
O
O
Table AX5-8 (cont'd). Systemic Effects of Ozone
X
Ozone Concentration
ppm
EFFECTS
0.5
0.8
0.8
1.0
10.0
1.0
5.0
10.0
1.0
5.0
10.0
jig/m3 Duration
ON CUTANEOUS TISSUE
980 2h
1,568 6 h
1,568 2 h
1,960
19,600
1,960 2 h
9,800
19,600
1,960 2 h
9,800
19,600
Species
Mouse
hairless
female
Mouse
SKH-1
hairless
Mouse
SKH-1
hairless
Mouse
SKH-1 hairless
Mouse
Effects3
a tocopherol levels in the stratum corneum (SC) were not affected by O3
exposure (0.5 ppm) alone, but were significantly depleted by combined
exposure to UV and O3.
Increased lipid peroxidation in the skin epidermis and dermis activated
stress proteins HSP27 and HO-1, and activated a proteolytic enzyme
system (MMP-9) related to matrix injury and repair processes.
High O3 depletes hydrophilic antioxidants in the SC: Vit. C decreased to
80%, GSH decreased to 41%, and uric acid decreased to 44% of control
levels after exposure to > 1.0 ppm O3
Vit. E levels decreased and malondialdehyde levels increased in the SC
with increasing O3 concentration.
High O3 exerts an oxidizing effect on the outermost layer of the skin (SC);
depletes low-molecular-weight antioxidants (a tocopherol, vit. C,
glutathione, uric acid) in a concentration dependent manner; increases
malondialdehyde levels associated with lipid peroxidation
Reference
Valacchi et al.
(2000)
Valacchi et al.
(2003)
Weber et al.
(2000)
Thiele et al.
(1997a)
Weber et al.
(2000)
aRER = Rough endoplasmic reticulum.
PE = Postexposure (i.e., time after O3 exposure ceased).
TSH = Thyroid stimulating hormone.
T3 = Triiodothyronine.
T4 = Thyroxine.
cyt. = Cytochrome.
NADPH = Reduced nucotinamide adenine dinucleotide phosphate.
NADH =
B[a]P = Benzo[a]pyrene.
NK = Natural killer.
PHA = Phytohemagglutin.
ConA = Concanavalin A.
LPS = Lipopolysaccharide.
SRBC = Sheep red blood cell.
TEA = Thiobarbituric acid.
ONP =
IgE = Immunoglobulin E.
-------
1 (Koyama and Hayaishi, 1994). The permanent effects in pups caused by high O3 exposure during
2 gestation were attributed to the diminished antioxidant capability of fetal tissue (Giinther et al.,
3 1993).
4
5 AX5.3.2 Neuroendocrine Effects
6 Several new studies have examined the interaction of O3 with the pituitary-thyroid-adrenal
7 axis. Sen et al. (1993) found that T3 supplementation increased metabolic rate and pulmonary
8 injury in the lungs of O3-treated animals. Increased toxicity to O3 was later reported in
9 hyperthyroid rats by Huffman et al. (2001).
10 Mechanisms involved in the interaction of O3 and the neuroendocrine system are still are
11 not well understood. Cottet-Emard et al. (1997) studied the effects of exposure to 0.5 ppm O3 for
12 5 days on catecholamine activity in rat sympathetic efferents and brain areas of prime importance
13 to adaptation to environmental stressors. Catecholamine activity was assessed by estimating the
14 turnover rate of catecholamines and in vivo tyrosine hydroxylase activity in peripheral and central
15 structures (i.e., heart, lungs, superior cervical ganglia, cerebral cortex, hypothalamus and
16 striatum), and in A2 cell groups within the nucleus tractus solitarius (NTS) and locus ceruleus
17 (A6). Ozone inhibited norepinephrine turnover in heart (-48% of the control level) but not in
18 lungs and failed to modify the tyrosine hydroxylase activity in superior cervical ganglia, and the
19 catecholamine content in the adrenal glands. In the central nervous system, O3 inhibited tyrosine
20 hydroxylase activity in noradrenergic brainstem cell groups, including the locus ceruleus (-62%)
21 and the caudal A2 subset (- 57%). Catecholamine turnover was decreased by O3 in the cortex
22 (-49%) and striatum (-18%) but not in the hypothalamus. These data show that high ambient
23 levels of O3 can produce marked neural disturbances in structures involved in the integration of
24 chemosensory inputs, arousal, and motor control, effects that may be responsible for some of the
25 behavioral effects described in Section 5.2.1.
26 High, non-ambient levels of O3 (e.g., > 1.0 ppm) have been shown to affect visual and
27 olfactory neural pathways in the rat. Custodio-Ramierez and Paz (1997) reported a significant
28 delay in visual evoked potentials recorded in the visual cortex and the lateral geniculate nucleus
29 of male Wistar rats acutely exposed for 4 h to high levels of O3 (1.5, and 3.0 ppm). Also using
30 Wistar rats, Colin-Barenque et al. (1999) reported cytological and ultrastructural changes in the
31 granule layer of the olfactory bulb after a 4-h exposure to 1 to 1.5 ppm O3.
December 2005 AX5-77 DRAFT-DO NOT QUOTE OR CITE
-------
1 AX5.3.3 Cardiovascular Effects
2 New studies evaluating the cardiovascular effects of O3 have monitored continuous
3 cardiovascular and ventilatory measurements in unanesthetized rats exposed to O3. Using the
4 head-out plethysmograph for continuous measurements, Arito et al. (1997) exposed rats to
5 filtered air for 6 h, followed 2 days later by a 5-h exposure to 0.1 ppm O3, 5 days later by a 5-h
6 exposure to 0.3 ppm O3, and 10 days later by a 5-h exposure to 0.5 ppm O3. Each of the O3
7 exposures was preceded by a 1-h exposure to filtered air. Transient rapid shallow breathing with
8 slightly increased HR appeared 1-2 min after the start of O3 exposures and was attributed to an
9 olfactory response. Persistent rapid shallow breathing with a progressive decrease in HR
10 occurred with a latent period of 1-2 h. During the last 90-min of exposure, averaged values for
11 relative minute ventilation tended to decrease with the increase in O3 concentration for young
12 (4-6 months) but not old (20-22 months) rats.
13 In a series of studies utilizing radiotelemetry transmitters for monitoring ECG, HR, core
14 body temperature (Tco), and motor activity in unanesthetized and unrestrained rats, Watkinson
15 et al. (1995; 2001) and Highfill and Watkinson (1996) demonstrated that when HR was reduced
16 during O3 exposure, the Tco decreased in association with reduced activity. The decreases in body
17 temperature and associated decreases in blood pressure reported by Watkinson et al., and also by
18 Arito et al. (1997), suggested that the pattern and magnitude of the ventilation and HR responses
19 were mediated through some physiological and behavioral defense mechanisms acting to
20 minimize the irritant effects of O3 inhalation.
21 The adaptation of cardiovascular and thermoregulatory responses to O3 also was reported by
22 Iwasaki et al. (1998) in ECG electrode- and thermistor sensor implanted rats after repeated
23 exposure to 0.1, 0.3, and 0.5 ppm O3 8 h/day for 4 consecutive days. Circadian rhythms of HR
24 and Tco were disrupted on the first and second O3 exposure days in a concentration dependent
25 manner. The 8-h and 12-h averaged values of HR and Tco decreased significantly on the first and
26 second exposure days and recovered to control values after small but significant rebound
27 increases on the third and fourth days of O3 exposure.
28 More recent reports by Watkinson et al. (2003) further examined the thermoregulatory
29 response to O3 exposures. Male Fischer-344 rats were exposed to one of three possible O3 levels (
30 0.0 ppm x 24 h/day [air], 0.5 ppm x 6h/day [intermittent] or 0.5 ppm x 23 h/day [continuous]) at
31 one of three temperatures (10° C [cold], 22° C [room], or 34° C [warm]) for a total of 9 treatment
December 2005 AX5-78 DRAFT-DO NOT QUOTE OR CITE
-------
1 groups. Another protocol examined the effects of O3 exposure (0.5 ppm) and exercise described
2 as rest, moderate, heavy or CO2-stimulated ventilation. Both intermittent and continuous O3
3 exposure caused decreases in HR and Tco and increases in BALF inflammatory markers. Exercise
4 in filtered air caused increases in HR and Tco while exercise in O3 caused decreases in those
5 parameters. Carbon dioxide and O3 induced the greatest deficits in HR and Tco. The authors
6 discuss several factors which may modulate the hypothermic response, including: 1) dose, 2)
7 animal mass (i.e., smaller animals show a greater response), and 3) environmental stress (e.g.,
8 restraint, exercise). The authors further discuss possible problems with extrapolation to humans
9 that may be caused by this response (discussed in Chapter 4).
10 The tissue edema reported in the heart and lungs of laboratory animals exposed to relatively
11 high ambient O3 concentrations (> 0.5 ppm) may be caused by increased circulating levels of
12 atrial natriuretic factor (ANF) which is known to be a possible mediator of increased capillary
13 permeability, vasodilation, and decreased blood pressure (Daly et al., 2002; Vesely et al.
14 (1994a,b,c) reported increased levels of ANF in the heart, lungs, and circulation of rats exposed to
15 0.5 ppm O3 for 8 h.
16
17 AX5.3.4 Reproductive and Developmental Effects
18 New studies on the developmental effects of O3 demonstrate that prenatal exposures to O3
19 concentrations < 1.0 ppm do not cause major or widespread somatic or neurobehavioral effects in
20 the offspring of laboratory animals. Animal studies evaluating O3-induced reproductive effects
21 have not been completed.
22
23 Developmental Effects
24 A study of somatic and neurobehavioral development was reported by Bignami et al. (1994)
25 in female CD-I mice exposed during pregnancy (days 7 to 17) to O3 concentrations of 0, 0.4, 0.8,
26 or 1.2 ppm. They did not find any O3 effects on reproductive or behavioral performance, or on
27 neonatal mortality, but found a significant decrease in body weight gain and delayed eye opening
28 in pups in the 1.2-ppm exposure group. A follow-up study by Petruzzi et al. (1995) did not find
29 any significant deficits in reproductive performance, postnatal somatic and neurobehavioral
30 development, or adult motor activity in CD-I mice exposed in utero from conception through day
31 17 of pregnancy to 0, 0.2, 0.4, and 0.6 ppm O3. In a subsequent study by Petruzzi et al. (1999),
December 2005 AX5-79 DRAFT-DO NOT QUOTE OR CITE
-------
1 subtle changes in handedness and morphine reactivity were found when the O3 exposures (0.3,
2 0.6, or 0.9 ppm) continued postnatally until weaning [post natal day (PND) 26]. Female mice
3 exposed to 0.6 ppm O3 showed a reduced preference for the right paw at PND 70, and 0.9 ppm O3
4 altered hot plate avoidance after IP treatment with morphine (10 mg/kg) at PND 100.
5 Dell'Omo et al. (1995a,b) exposed CD-I mice to 0.6 ppm O3 from birth through weaning
6 (PND 22 or 26). Swimming navigation was tested (DeH'Omo et al., 1995a) at 12 to 13 weeks of
7 age using acquisition and reversal trials. Exposure to O3 did not produce any significant
8 impairment of navigational performance during acquisition and only subtle changes during
9 reversal. As in previous studies, there were no significant effects of O3 on reproductive
10 performance but O3 offspring showed a significant reduction in body weight. Ozone effects on
11 neurobehavioral development were not large and very selective, with some attenuation of activity
12 responses and impairment of passive avoidance acquisition (DeH'Omo et al. (1995b). Similarly,
13 only small and selective effects on somatic and sensorimotor development were found in the
14 offspring of CD-I mice continuously exposed from 30 days prior to the formation of breeding
15 pairs until PND 17 to 0.0, 0.3, or 0.6 ppm O3 (Sorace et al., 2001).
16
17 AX5.3.5 Effects on the Liver, Spleen, and Thymus
18 Liver
19 New studies on the effects of O3 on liver showed that, in rats, high (1 to 2 ppm) acute O3
20 exposures caused increased production of NO by hepatocytes and enhanced protein synthesis
21 (Laskin et al., 1994; 1996).
22 The O3-associated effects shown in the liver are most likely mediated by inflammatory
23 cytokines (e.g., TNF alpha) or other cytotoxic mediators (e.g., hydroxyl radicals) released by
24 activated macrophages in the lungs (Vincent et al., 1996; Laskin et al., 1998; Laskin and Laskin,
25 2001). Except for the earlier work on xenobiotic metabolism, the responses occurred after very
26 high acute O3 exposures.
27
28 Spleen and Thymus
29 Ozone has been shown to primarily affect T-cell mediated systemic immunity. New studies
30 evaluating O3-induced effects on spleen and thymus have not been completed.
31
December 2005 AX5-80 DRAFT-DO NOT QUOTE OR CITE
-------
1 AX5.3.6 Ozone Effects on Cutaneous and Ocular Tissues
2 Ground-level smog exposure not only affects various organ systems, when inhaled, but may
3 potentially have direct effects on exposed skin and eyes. Several new studies have examined the
4 effects of O3 on skin. Ozone can have an oxidizing effect on the outermost layer of the skin,
5 called the stratum corneum (SC) where it may compromise skin barrier function and possibly
6 induce an epidermal proinflammatory response (e.g., Weber et al., 2001; Cotovio et al., 2001;
7 Thiele, 2001); however, these cutaneous effects of O3 are found only at very high concentrations
8 used in experimental studies and have not been shown at more relevant ambient- or near-ambient
9 levels (< 0.5 ppm) of O3 exposure. The lack of ambient-level O3 effects on the skin is most likely
10 due to a well-developed defense against oxidative stress, utilizing nonenzymatic (e.g., vitamin C
11 and E, glutathione, uric acid, a-tocopherol) and enzymatic (e.g., superoxide dismutase, catalase,
12 glutathione reductase and peroxidase) antioxidants found in many living organisms (Cross et al.,
13 1998).
14 Effects of ground-level smog on the eyes have also been reported, but generally are
15 attributed to related photochemical oxidants like peroxyacetyl nitrate (Vyskocil et al., 1998) or
16 possibly to atmospheric O3 precursors or reaction products like aldehydes. Ocular tissues also
17 have similar antioxidant protective function as in the skin, but are not well studied (Mucke, 1996;
18 Rose etal., 1998).
19 The cutaneous effects of O3 exposure were first reported by Thiele et al. (1997a,b,c,d).
20 Hairless mice (SKH-1) were exposed to 1, 5, and 10 ppm O3 for 2 h or to 1 ppm O3 for 2 h on six
21 consecutive days and skin layers from the epidermis and dermis were separated for analysis of
22 antioxidants and lipid peroxidation products. Decreased antioxidant levels (a-tocopherol;
23 ascorbic acid) were found in the upper epidermis and increased malondialdehyde (MDA), a lipid
24 peroxidation product, was found in both the upper and lower epidermal layers. Effects were dose
25 dependent and became significant (p < 0.05) for a-tocopherol and MDA after single exposures to
26 >1.0 and >5.0 ppm O3, respectively. Repeated O3 exposures caused significantly higher MDA
27 concentrations. The effects of MDA accumulation could be prevented by enriching the skin with
28 vitamin E.
29 In the SKH-1 hairless mouse model, Weber et al. (1999, 2000, 2001) demonstrated that
30 0.8 to 10 ppm O3 exposure for 2 h depletes the low molecular weight antioxidants (e.g.,
31 a-tocopherol, vitamin C, glutathione, uric acid) in the SC in a dose-response manner and causes
December 2005 AX5-81 DRAFT-DO NOT QUOTE OR CITE
-------
1 increased MDA. The effects were statistically significant (p < 0.05) at > 1 ppm O3 and > 5 ppm
2 for antioxidant depletion and increased MDA, respectively. Reactive aldehydes were found in the
3 epidermis at the highest O3 concentration tested, though this level of exposure is of limited
4 relevance. Preexposure to O3 followed by low-dose ultraviolet (UV) radiation (0.33 MED)
5 decreased the significance level to 0.5 ppm O3, probably through combined oxidative stress on the
6 SC (Valacchi et al., 2000). Stress-inducible proteins (e.g., heme oxygenase-1) and other heat
7 shock proteins (e.g., HSP27 and HSP70) were found in deeper cellular layers of the epidermis
8 after 2 h of exposure to 8.0 ppm O3 (Valacchi et al., 2002). Prolonged exposure to lower
9 concentrations of O3 (0.8 ppm) for 6 h also induces cellular stress responses that included the
10 formation of HNE protein adducts, HSP27, and heme-oxygenase-1 in the deeper cellular layers of
11 the skin that continued for up to 18 h after O3 exposure, followed by repair processes (Valacchi
12 et al., 2003).
13
14
15 AX5.4 INTERACTIONS OF OZONE WITH OTHER CO-OCCURRING
16 POLLUTANTS
17 Ozone is part of a complex mixture of air pollutants with a composition and pattern that
18 varies geographically and temporally (by hour of the day, day of the week, and season). Health
19 effects caused by the complex mixture are undoubtedly different (either subtly or significantly)
20 from the additive effects of a few of the hundreds of compounds present. The only disciplinary
21 approach that can evaluate a "real-world" complex mixture is epidemiology (Chapter 7).
22 However, because of the difficulty in evaluation of causative factors and quantitative
23 relationships in epidemiology studies, it is useful to consider animal toxicological studies of
24 mixtures. Such studies can be divided into three categories: (1) ambient air mixtures,
25 (2) laboratory-generated complex mixtures (e.g., gasoline combustion mixtures having
26 ultraviolet-irradiation, other reaction mixtures with O3 and several other components), and
27 (3) binary mixtures. In most cases, experimental designs in the first two classes did not have an
28 O3-only group, making it difficult to impossible to discern the influence of O3. The more recent
29 mixture studies that are discussed here typically have been with NO2, sulfuric acid (H2SO4), or
30 ammonium sulfate ([NH4]2SO4).
December 2005 AX5-82 DRAFT-DO NOT QUOTE OR CITE
-------
1 Interpreting the mixture studies in terms of real-world risk is difficult because laboratory
2 exposure patterns do not always represent real-world exposure patterns. For example, in the real
3 world, NO2 often peaks before O3 peaks, with a mixture occurring between the peaks, but most
4 laboratory exposures used mixtures only. Also, most studies of O3 and NO2 mixtures used
5 ambient levels of O3 and levels of NO2 high above ambient. As shall be seen, all interaction
6 possibilities have occurred, depending upon the composition of the mixture, the endpoint
7 examined, and the exposure regimen. In some cases, no interaction was found. Most often,
8 additivity (the effects of the mixture are equal to the sum of the effects of the individual
9 components) or synergism (the effects of the mixture are greater than the sum of the effects of the
10 individual components) was observed. Antagonism (the effects of the mixture are less than the
11 sum of the individual components) was rarely found.
12
13 AX5.4.1 Ozone and Nitrogen Oxides
14 The most commonly studied copollutant in binary mixtures with O3 is NO2. New studies
15 evaluating the effects of combined O3/NO2 exposures are listed in Table AX5-9.
16 Recent work has demonstrated that chronic exposures of rats to 0.8 ppm O3 and 14.4 ppm
17 NO2 for 6 h/day caused the rats to develop insufficiency and severe weight loss (Farman et al.,
18 1997). Half of these animals died after 55 to 78 days of exposure due to severe fibrosis.
19 Biochemical analysis of lung tissue demonstrated increased total lung collagen and elastin, with
20 loss of mature collagen, indicating there was breakdown and remodeling of the lung parenchyma.
21 In follow-up rat studies, Farman et al. (1999) reported a sequence of events starting with
22 increasing inflammatory and mild fibrotic changes for the first 3 weeks of exposure to 0.8 ppm O3
23 and 14.4 ppm NO2, stabilized or even reduced changes after 4 to 6 weeks, and severe increases
24 over 7 to 9 weeks of exposure. The authors speculated that repair processes occurring during the
25 middle 4 to 6 weeks of exposure become overwhelmed, leading to progressive fibrosis after 7 to 8
26 weeks of exposure. When combined exposures to 0.8 ppm O3 and 14.4 ppm NO2 for 6 h/day,
27 7 days/week were extended for 90 days, Farman et al. (1999) found that the lesion extended far
28 into the acinus, but the extent of tissue involvement was the same after 7, 78, and 90 days of
29 exposure. At the end of exposure, in situ hybridization for procollagen types I and III
30 demonstrated high levels of messenger RNA within central acini in the lungs from the combined
31 exposure group, but not in lungs from the rats exposed to O3 and NO2 alone.
December 2005 AX5-83 DRAFT-DO NOT QUOTE OR CITE
-------
O
O
i
{_} '
CD
>-l
to
O
O
v\
X
oo
O
i
H
6
l>
-^H
0
H
O
c
O
s
0
&
O
H
W
Table AX5-9.
Concentration
O3 NO2
ppm fig/m3 Hg/m3 ppm Duration
MORPHOLOGY
0.8 1,568 27,072 14.4 6 h/day ,
7 days/week
for 90 days
0.3 588 2,256 1.2 Continuous for
3 days
BIOCHEMISTRY (cont'd)
Interactions of Ozone With Nitrogen Dioxide
Species Endpoints" Interaction
Rat, M Morphometry of lung syngeristic; more
(S-D) parenchyma; DNA probes for peripheral centriacinar
10-12 weeks procollagen; in situ lesion, but same after
old mRNA hybridization 7, 78, and 90 days of
exposure.
Rat, M DNA single strand breaks; None; effect due to O3
(S-D) polyADPR synthetase of
3 months old AMs; total cells, protein, and
LDHinBALF
0.8 1,568 27,072 14.4 6 h/day, 7 days/week Rat, M (S- Lung hydroxyproline, Synergistic; fibrosis
for 9 weeks
BAL = Bronchoalveolar lavage.
PG = Prostaglandin.
G-6-PD = Glucose-6-phosphate dehydrogenase.
GOT = GSH-disulfide transhydrogenase.
GSHPX = GSH peroxidase.
D) 10-12 hydrooxypyridinium, DNA, after 7-8 weeks of
weeks old and protein content of whole exposure
lung; morphology and
labeling index
SOD = Superoxide dismutase.
DR = Disulfide reductase.
Reference
Farman et al.
(1999)
Bermudez et al.
(1999,
Bermudez
(2001)
Farman et al.
(1997)
NADPH-CR = Reduced nicotinamide adenine dinucleotide phosphate-cytochrome c reductase.
GSH = Glutathione.
6-PG-D = 6-phosphogluconate dehydrogenase.
-------
1 Bermudez et al. (1999) reported that a 3-day exposure to 0.3 ppm O3, and the combined
2 exposure of O3 and 1.2 ppm NO2, caused significant DNA single-strand breaks in the alveolar
3 macrophages of Sprague-Dawley rats. No changes were caused by NO2-only exposure. In a
4 follow-up study, Bermudez (2001) showed that the same exposures stimulated the activity of
5 polyADPR synthetase, an indicator of response to lung cellular DNA repair caused by oxidant-
6 induced lung injury.
7 Other published reports (Ishii et al., 2000b; Weller et al., 2000) indicate that the laboratory
8 animal model of progressive pulmonary fibrosis, utilizing long-term, combined O3 (0.4 to
9 0.8 ppm) and high-level NO2 (7 to 14 ppm) exposure, causes an initial acute pulmonary
10 inflammation, followed by adaptation and repair, and eventually causing pulmonary fibrosis after
11 6 to 13 weeks of exposure.
12
13 AX5.4.2 Ozone and Other Copollutants
14 Although the bulk of the toxicological database for binary mixtures of O3 involves NO2 or
15 acidic sulfate and nitrate aerosols (see Section 5.4.3), a few studies have examined responses
16 to combinations of O3 with other single pollutants, such as formaldehyde (HCHO); or with
17 surrogates of pollutants treated as a single pollutant, such as tobacco smoke. New studies
18 evaluating coexposures of O3 with acid aerosols and particle mixtures are listed in Table AX5-10.
19
20 Interactive effects of ozone and formaldehyde.
21 A study by Cassee and Feron (1994) focused on biochemical and histopathological changes
22 in the nasal respiratory epithelium of rats exposed 8 h/day for 3 days to 0.4 ppm O3 and 3.6 ppm
23 HCHO, alone and combined. No interactive effects were found, however, despite the high levels
24 of HCHO when compared to typical ambient levels of 1 to 10 ppb (e.g., Rehle et al., 2001). In a
25 follow-up to their previous combined exposure study in rats, Mautz (2003) elaborated on changes
26 in breathing pattern and epithelial cell proliferation attributed to O3 and HCHO. Rats were
27 exposed to 0.6 ppm O3 and 10 ppm HCHO alone and in combination for 3 h with exercise at two
28 times resting ventilation. Even with exercise, HCHO does not penetrate to the lower respiratory
29 tract to interact with O3, and does not alter breathing patterns to modify local O3 dose.
30 Parenchymal injury was, therefore, due to O3 alone. In the nasal transitional epithelium and in the
31 trachea, however, combined exposure produced additive effects due to the increased volume of
December 2005 AX5-85 DRAFT-DO NOT QUOTE OR CITE
-------
December 2005
Table AX5-10. Interactions Of Ozone With Particles
Concentration
03
ppm
Hg/m3
PM
mg/m3a
Durationb
Species
Endpoints0 Interaction
Reference
SULFURIC ACID
X
oo
O\
0.1
0.2
0.1
0.3
0.6
0.1
0.3
0.6
0.6
196
392
196
588
1,176
196
588
1,176
1,176
0.02-0.15
(0.4 - 0.8 urn)
0.50 (0.3 urn)
0.125(0.3 urn)
0.50 (0.3 urn)
0.125(0.3 urn)
0.5 (0.06 and
0.3 umMMD)
23. 5 h/day or
intermittent 12
h/day for up to
90 days
3h
3h
4 h/day for
2 days
Rat
S-D
male
Rabbit NZW
male
Rabbit NZW
male
Rat
Morphology No interaction
Biochemistry
AM intracellular pH Antagonism
homeostasis and H+
extrusion
Airway responsiveness Antagonism
(in vitro bronchial rings
+ ACh)
Morphology: volume Synergism: ultrafme + O3, but
percentage of total not fine
Last and
Pinkerton
(1997)
Chen et al.
(1995)
El-Fawal et al.
(1995)
Kimmel et al.
(1997)
parenchyma containing
injured alveolar septae;
bromode oxyuridine Synergism: fine + O3
cell labeling index in
the periacinar region
PARTICLE MIXTURES
H ai
O
o
2
o
H 0.16
O 0.30
g 0.59
S
W
&
O
H
W
196
314
588
1,156
Diesel PM (NIST
#2975) reacted with
O3 for 48 h
0.05 - 0.22 mg/m3
ammonium bisulfate
0.03 -0.10 mg/m3 C
0.11- 0.39 pmNO2
0.02 -0.11 mg/m3
HNO3
(0.3 umMMAD)
24 h (IT)
4 h/day,
3 days/week
for 4 weeks
Rat
S-D
Rat
F344N
male
Inflammation
breathing pattern,
morphology, lavagable
protein, and clearance
Synergism
Complex interactions, but
possible loss of typical
attenuation seen with O3 only
exposure, reflecting persistence
of inflammation
Madden et al.
(2000)
Mautz et al.
(2001)
-------
December 2005
X
oo
O
H
6
O
0
H
O
O
0
O
H
W
Table AX5-10 (cont'd). Interactions Of Ozone With
Concentration
O3 PM
ppm fig/m3 mg/m3a
PARTICLE MIXTURES (cont'd)
0.2 392 0.07and 0.14 mg/m3
ammonium bisulfate
(0.45 urn MMMD);
0.05 and 0.10
mg/m3 carbon
0.2 392 0.50 mg/m3
ammonium bisulfate
(0.45 urn MMMD)
and elemental
carbon)
0.3 588 0.063 to 1.57 mg/m3
CAPs (Boston) + ip
OVA sensitization
0.4 784 0.20 and 0.50
mg/m3 fine, H2O2-
coated carbon
(0.26 urn MMMD)
0.5 980 Endotoxin (IN)
100 ug 24 h and
48 h after the 3rd
O3 exposure
0.5 980 OVA (IN)
50 ul (1%)
Durationb
4 h/day,
3 days/week
for 4 weeks
4 h/day,
3 days/week
for 4 weeks;
nose only
5h
4 h/day for 1 or
5 days
8 h/day for
3 days
8 h/day for
1 day or 3
consecutive
days
Species
Rat
F344
male
22-24 months
old
Rat
F344N-NIA
22-24 months
old
Mouse
BALB/c
Rat
S-D
300 g
Rat
F344
10-12 weeks
old
Rat
Brown
Norway 10-12
weeks old
Endpoints0
BAL protein and
albumin; plasma
hydroxylase and
fibronectin
DNA labeling of
dividing lung epithelial
and interstitial cells by
5-bromo-2-
deoxyuridine
Airway function
Inflammation
Nasal morphology
Nasal morphology
Particles
Interaction
Questionable interaction
Synergism
Interaction: increased RL and
airway responsiveness
Synergism for effect on day 5
Synergism: increased
intraepithelial mucosubstances
and mucous cell metaplasia
Synergism: increased
intraepithelial mucosubstances
and mucous all metaplasia
Reference
Bolarin et al.
(1997)
Kleinman
et al. (2000)
Kobzik et al.
(2001)
Kleinman
etal. (1999)
Fanucchi et al.
(1998)
Wagner et al.
(2001a,b)
Wagner et al.
(2002)
-------
o
O
to
O
O
Table AX5-10 (cont'd). Interactions Of Ozone With Particles
X
oo
oo
Concentration
O3 PM
ppm Hg/m3 mg/m3a
PARTICLE MIXTURES (cont'd)
0.8 1,600 0.5mg. 1.5 rag or
5 mg of PM from
Ottawa Canada
(EHC-93)
1 1,960 O.llmg/m3 ultra
fine carbon ( 25 nm
CMD) + endotoxin
OH))
1 1,960 Endotoxin (37.5
EU) for 10 minutes
1 1,960 Endotoxin (IN) 0, 2,
or 20 ug in 120 uL
Duration* Species Endpoints0
2, 4, and Rat Inflammation
7 days after
IT instillation
6 h Rat Inflammation
F344
male,
10 weeks and
22 months old;
Mouse
TSK
male, 14-
17 months old
4, 20, or 24 h Mouse Inflammation
C57BL/6J
8 weeks old
8 h, repeated Rat Lung morphometric
after 24 h F344 analysis and
inflammation
Interaction
Interaction: increased TNF-oc in
BALF
Interaction: increased PMNs and
ROS release from BALF cells for
old rats and mice primed with
endotoxin; depressed in young
rats
Synergism: increased BALF
protein and PMNs
Synergism: increased BALF
PMNs and mucin glycoprotein;
increased intraepithelial
mucosubstances and mucingene
mRNA
Reference
Ulrich et al.
(2002)
Elder et al.
(2000a,b)
Johnston et al.
(2000, 2002)
Wagner et al.
(2003)
aVMD = Volume median diameter.
og = Geometric standard deviation.
MMAD = Mass median aerodynamic diameter.
CMD = Count median diameter.
PM = Paniculate Matter.
OVA= Ovalbumin.
AED = Aerodynamic diameter.
bUnless indicated otherwise, whole-body exposures used.
°BAL = Bronchoalveolar lavage.
AM = Alveolar macrophage.
IN = Intranasal
IT = Intratracheal.
-------
1 toxicants during exercise. No other combined pollutant studies have been published in the peer-
2 reviewed literature, although two studies compared the respiratory effects of O3 to HCHO.
3 Nielson et al. (1999) compared upper airway sensory irritation caused by HCHO concentrations
4 up to 4 ppm to the lower airway irritation caused by O3. Using BALB/c mice, they continuously
5 measured fB, Vt, expiratory flow, T;, Te, and respiratory patterns during acute, 30-min exposures.
6 The NOEL for HCHO was 0.3 ppm, compared to 1.0 ppm for O3.
7
8 Interactive effects of ozone and tobacco smoke.
9 Wu et al. (1997) also reported that inhalation of cigarette smoke evokes a transient
10 bronchoconstrictive effect in anesthetized guinea pigs. To examine whether O3 increases airway
11 responsiveness to cigarette smoke, effects of smoke inhalation challenge on total pulmonary
12 resistance (RL) and dynamic lung compliance (Cdyn) were compared before and after acute
13 exposure to 1.5 ppm O3for 1 h. Before O3 exposure, inhalation of two breaths of cigarette smoke
14 (7 ml) at a low concentration (33%) induced a mild and reproducible bronchoconstriction that
15 slowly developed and reached its peak (ARL = 67 ± 19%, ACdyn = -29 ± 6%) after a delay of
16 > 1 min. After exposure to O3, the same cigarette smoke inhalation challenge evoked an intense
17 bronchoconstriction that occurred more rapidly, reaching its peak (ARL = 620 ± 224%, ACdyn =
18 -35 ± 7%) within 20 s, and was sustained for > 2 min. By contrast, sham exposure to room air
19 did not alter the bronchomotor response to cigarette smoke challenge. Pretreatment with selective
20 antagonists of neurokinin type 1 and 2 receptors completely blocked the enhanced airway
21 responsiveness. The authors concluded that O3 exposure induced airway hyperresponsiveness to
22 inhaled cigarette smoke, which resulted primarily from the bronchoconstrictive effect of
23 endogenous tachykinins.
24 To determine the effects of aged and diluted sidestream cigarette smoke (ADSS) as a
25 surrogate of environmental tobacco smoke (ETS) on O3-induced lung injury, Yu et al. (2002)
26 exposed male B6C3F1 mice to (1) filtered air, (2) ADSS, (3) O3, or (4) ADSS followed by O3
27 (ADSS/O3). Exposure to 30 mg/m3 ADSS, 6 h/day for 3 days, followed by exposure to 0.5 ppm
28 O3 for 24 h was associated with a significant increase in the number of cells recovered by B AL
29 compared with exposure to ADSS alone or O3 alone. The proportion of neutrophils and
30 lymphocytes, as well as total protein level in BAL, also was significantly elevated following the
31 combined exposure when compared with all other groups. Within the centriacinar regions of the
December 2005 AX5-89 DRAFT-DO NOT QUOTE OR CITE
-------
1 lungs, the percentage of proliferating cells identified by bromodeoxyuridine (BrdU) labeling was
2 unchanged from control following exposure to ADSS alone, but was significantly elevated
3 following exposure to O3 (280% of control), and further augmented in a statistically significant
4 manner in mice exposed to ADSS/O3 (402% of control). Following exposure to O3 alone or
5 combined with ADSS, the ability of AMs to release interleukin (IL)-6 under lipopolysaccharide
6 (LPS) stimulation was significantly decreased, while exposure to ADSS alone or ADSS/O3
7 caused a significantly increased release of TNF-a from AMs under LPS stimulation. The authors
8 concluded that ADSS exposure enhances the sensitivity of animals to O3-induced lung injury.
9 Toxicological studies with components of ETS (e.g., nicotine receptor agonists, acrolein,
10 and oxidants) have shown that the vagal bronchopulmonary C-fibers are stimulated by acute
11 exposures that initiate both central and local responses (Bonham et al., 2001; Mutoh et al., 2000).
12 The central responses (e.g., tachypnea, cough, bronchoconstriction, increased mucous secretion)
13 are more protective of the lung; however, local responses may include increased sensitization of
14 the C-fibers to other irritants, including O3.
15
16 AX5.4.3 Complex (Multicomponent) Mixtures Containing Ozone
17 Ambient pollution in most areas is a complex mix of more than two chemicals, and a
18 number of new studies have examined the effects of exposure to multicomponent atmospheres
19 containing O3. Some of these studies attempted to simulate photochemical reaction products
20 occurring under actual atmospheric conditions. However, the results of these studies are often
21 difficult to interpret because of chemical interactions between the components, as well as the
22 resultant production of variable amounts of numerous secondary reaction products, and a lack of
23 precise control over the ultimate composition of the exposure environment. In addition, the role
24 of O3 in the observed biological responses is often obscure.
25 A recent attempt has been made to examine multicomponent mixtures resulting from the
26 reaction of O3 with unsaturated hydrocarbons [e.g., isoprene (C5H8) and terpene (C10H16)],
27 producing HCHO, formic acid, acetone, acrolein, acetic acid, and other oxidation products, many
28 of which are strong airway irritants. Wilkins et al. (2001) evaluated sensory irritation by
29 measuring mean fB in the mouse bioassay and found a 50% reduction after 30 min of exposure to
30 reaction products of O3 and isoprene. The mixture at this time period contained < 0.2 ppm O3, so
31 the authors attributed the observed effects to the oxidation products. Clausen et al. (2001), using
December 2005 AX5-90 DRAFT-DO NOT QUOTE OR CITE
-------
1 the same mouse model, evaluated the reaction products of O3 and limonene. A 33% reduction in
2 mean fB was produced after 30 min of exposure to the mixture containing < 0.3 ppm O3, again
3 implicating the effects of strong irritant products.
4 Pollutant mixtures containing acid aerosols comprise another type of commonly examined
5 exposure atmosphere. Most of these mixtures included acidic sulfate aerosols as the copollutant.
6 Peak (1-h) ambient levels of sulfuric acid (H2SO4) are estimated at 75 |ig/m3, with longer (12-h)
7 averages about one-third of this concentration.
8 More recent studies found some differences in airway responses to inhaled acid particle-O3
9 mixtures that may have been partly due to airway dosimetry. Various physical and chemical
10 mechanisms may be responsible (see Schlesinger, 1995). For example, physical adsorption or
11 absorption of O3 or its reaction products on a particle could result in transport to more sensitive
12 sites, or to sites where O3, by itself, would not normally be reactive. Chemical reactions on the
13 surface of particles can form secondary products that are more lexicologically active, or chemical
14 characteristics of the particle may change the residence time or reactivity of oxidation products at
15 the site of deposition. Chen et al. (1995) and El-Fawal et al. (1995) tested this hypothesis on
16 rabbit pulmonary macrophages and on airway reactivity, respectively. Male New Zealand white
17 rabbits were exposed for 3 h to 125 |ig/m3 H2SO4, 0.1, 0.3, or 0.6 ppm O3, and to combinations of
18 O3 and H2SO4. Decreased pH following exposure to acid aerosol was correlated with phagocytic
19 activity and capacity of harvested macrophages; however, exposure to the mixture did not show
20 this relationship (Chen et al., 1995). Responsiveness of harvested bronchial rings to acetylcholine
21 was increased following O3 exposure, but the combination of O3 and H2SO4 resulted in
22 antagonism (El-Fawal et al., 1995).
23 Churg et al. (1996) demonstrated increased uptake of asbestos or TiO2 into rat tracheal
24 explant cultures in response to 10 min O3 (up to 1.0 ppm) pre-exposure. These data suggest that
25 low concentrations of O3 may increase the penetration of some types of PM into epithelial cells.
26 More recently, Madden et al. (2000) demonstrated a greater potency for ozonized diesel PM to
27 induce prostaglandin E2 production from human epithelial cell cultures, suggesting that O3 can
28 modify the biological activity of PM derived from diesel exhaust (see details below).
29 Vincent et al. (1997) and Adamson et al. (1999) exposed rats to 0.8 ppm O3 in combination
30 with 5 or 50 mg/m3 of resuspended urban particles for 4 h. Although PM alone caused no change
31 in cell proliferation (3H-thymidine labeling), co-exposure to either concentration of resuspended
December 2005 AX5-91 DRAFT-DO NOT QUOTE OR CITE
-------
1 PM with O3 greatly potentiated the proliferative effects of exposure to O3 alone. These
2 interactive changes occurred in epithelial cells of the terminal bronchioles and the alveolar ducts.
3 Kimmel et al. (1997) examined the effect of acute coexposure to O3 and fine or ultrafine H2SO4
4 aerosols on rat lung morphology. They determined morphometrically that alveolar septal volume
5 was increased in animals coexposed to O3 and ultrafine, but not fine, H2SO4. Interestingly, cell
6 labeling, an index of proliferative cell changes, was increased only in animals co-exposed to fine
7 H2SO4 and O3, as compared to animals exposed to O3 alone. Last and Pinkerton (1997) found that
8 subchronic exposure to acid aerosols (20 to 150 |ig/m3 H2SO4) had no interactive effect on the
9 biochemical and morphometric changes produced by either intermittent or continuous O3
10 exposure (0.12 to 0.2 ppm). Thus, the interactive effects of O3 and acid aerosol coexposure in the
11 lung disappeared during the long-term exposure. Sindhu et al. (1998) observed an increase in rat
12 lung putrescine levels after repeated, combined exposures to O3 and a nitric acid vapor.
13 Other studies have examined interactions between carbon particles and O3. Creutzenberg
14 et al. (1995) treated rats with a high concentration of carbon particles by intratracheal instillation
15 followed by either a 7-day or 60-day exposure to 0.5 ppm O3. The phagocytotic capacity and
16 chemotactic migration capability of alveolar macrophages was impaired by carbon black, but was
17 stimulated by O3. Kleinman et al. (1999) examined the effects of O3 plus fine, H2SO4-coated,
18 carbon particles (MMAD = 0.26 jim) for 1 or 5 days. They found that the inflammatory response
19 with the O3-particle mixture was greater after 5 days (4 h/day) than after day 1. This contrasted
20 with O3 exposure alone (0.4 ppm), which caused marked inflammation on acute exposure, but no
21 inflammation after 5 consecutive days of exposure.
22 Kleinman et al. (2000) examined the effects of a mixture of elemental carbon particles, O3,
23 and ammonium bisulfate on rat lung collagen content and macrophage activity. Decreases in lung
24 collagen, and increases in macrophage respiratory burst and phagocytosis were observed relative
25 to other pollutant combinations. Mautz et al. (2001) used a similar mixture (i.e., elemental carbon
26 particles, O3, ammonium bisulfate, but with NO2 also) and exposure regimen as Kleinman et al.
27 (2000). There were decreases in pulmonary macrophage Fc-receptor binding and phagocytosis
28 and increases in acid phosphatase staining. Bronchoalveolar epithelial permeability and cell
29 proliferation were increased. Altered breathing-patterns also were observed, with some
30 adaptations occurring.
December 2005 AX5-92 DRAFT-DO NOT QUOTE OR CITE
-------
1 Bolarin et al. (1997) exposed rats to 50 or 100 |ig/m3 carbon particles in combination with
2 ammonium bisulfate and O3. Despite 4 weeks of exposure, they observed no changes in protein
3 concentration in lavage fluid or blood prolyl 4-hydroxylase, an enzyme involved in collagen
4 metabolism. Slight decreases in plasma fibronectin were present in animals exposed to the
5 combined pollutants versus O3 alone. Thus, the potential for adverse effects in the lungs of
6 animals challenged with a combined exposure to particles and gaseous pollutants is dependent on
7 numerous factors, including the gaseous co-pollutant, concentration, and time.
8 In a complex series of studies, Oberdorster et al. examined the interaction of several
9 pulmonary oxidative stress pollutants. Elder et al. (2000a,b) reported the results of combined
10 exposure to ultrafme carbon particles (100 |ig/m3) and O3 (1 ppm) in young and old Fischer 344
11 rats that were pretreated with aerosolized endotoxin. In old rats, exposure to carbon and O3
12 produced an interaction that resulted in a greater influx in neutrophils than that produced by either
13 agent alone. This interaction was not seen in young rats. Oxidant release from lavage fluid cells
14 also was assessed and the combination of endotoxin, carbon particles, and O3 produced an
15 increase in oxidant release in old rats. This mixture produced the opposite response in the cells
16 recovered from the lungs of the young rats, indicating that the lungs of the aged animals
17 underwent greater oxidative stress in response to a complex pollutant mix of particles, O3, and a
18 biogenic agent. Johnston et al. (2000; 2002) reported the results of combined exposure to O3 (1
19 and 2.5 ppm for 4, 20, or 24 h) and low-dose endotoxin, or to O3 and endotoxin separately, in
20 newborn and adult C57BL/6J mice. In the first study, adult (8 week old) mice showed greater
21 sensitivity to O3 than newborn (36 h old) mice on the basis of mRNAs encoding for various
22 chemokines and cytokines. In contrast, adult and newborn mice responded similarly 2 h after
23 endotoxin exposure (10 ng for 10 min), suggesting that age differences in O3-generated
24 inflammation is secondary to epithelial cell injury. In the second study, 8 week old mice exposed
25 to O3 (1 ppm for 24 h) followed by endotoxin (37.5 EU for 10 min) showed increased
26 responsiveness over either exposure alone, on the basis of increased expression of chemokine and
27 cytokine messages and increased BAL fluid levels of protein and PMNs.
28 Fanucchi et al. (1998) and Wagner et al. (2001a,b) examined the synergistic effect of
29 co-exposure to O3 and endotoxin on the nasal transitional epithelium of rats that also was
30 mediated, in part, by neutrophils. Fisher 344 rats intranasally instilled with endotoxin and
31 exposed to 0.5 ppm O3, 8 h per day, for 3 days developed mucous cell metaplasia in the nasal
December 2005 AX5-93 DRAFT-DO NOT QUOTE OR CITE
-------
1 transitional epithelium, an area normally devoid of mucous cells; whereas, intratracheal
2 instillation of endotoxin (20 jig) caused mucous cell metaplasia rapidly in the respiratory
3 epithelium of the conducting airways. A synergistic increase of intraepithelial mucosubstances
4 and morphological evidence of mucous cell metaplasia were found in rat maxilloturbinates upon
5 exposure to both O3 and endotoxin, compared to each pollutant alone. A similar response was
6 reported in OVA-sensitized Brown Norway rats exposed to 0.5 ppm O3, 8 h/day for 3 days
7 (Wagner et al., 2002), indicating that coexposure to O3 and inflammatory biogenic substances like
8 allergens (e.g., OVA) or bacterial endotoxin can augment epithelial and inflammatory responses
9 in rat nasal passages.
10 In follow-up studies, Wagner et al. (2003) reported that coexposure of rats to O3 and
11 endotoxin also enhanced epithelial and neutrophilic inflammatory responses in the pulmonary
12 airways. Fisher 344 rats were intranasally instilled with endotoxin and exposed to 1.0 ppm O3 for
13 8 h, which was repeated 24 h later. Three days after the last exposure, BALF was analyzed for
14 inflammatory cells and secreted mucosub stances (mucin SAC), and lung tissue was processed for
15 morphometric analysis. Endotoxin instillation alone caused a dose-dependent increase in BALF
16 neutrophils that was further increased 2-fold in O3-exposed rats given 20 jig endotoxin, the
17 highest dose. Mucin glycoprotein SAC also was increased in the BALF at this dose, but not at
18 lower endotoxin doses. Ozone exposure alone did not cause mucus hypersecretion, but it did
19 potentiate mucus secretion in rats given both 2 and 20 jig endotoxin and increased intraepithelial
20 mucosub stances 2-fold, which was further substantiated by significant increases in mucin gene
21 (rMucSAC) mRNA levels in the conducting airways.
22 The effect of O3 modifying the biological potency of PM (diesel PM and carbon black) was
23 examined by Madden et al. (2000) in rats. Reaction of NIST Standard Reference Material # 2975
24 diesel PM with 0.1 ppm O3 for 48 h increased the potency (compared to unexposed or air-exposed
25 diesel PM) to induce neutrophil influx, total protein, and LDH in lung lavage fluid in response to
26 intratracheal instillation. Exposure of the diesel PM to high, non-ambient O3 concentration (1.0
27 ppm) attenuated the increased potency, suggesting destruction of the bioactive reaction products.
28 Unlike the diesel particles, carbon black particles exposed to 0.1 ppm O3 did not exhibit an
29 increase in biological potency, which suggested that the reaction of organic components of the
30 diesel PM with O3 were responsible for the increased potency. Reaction of particle components
31 with O3 was ascertained by chemical determination of specific classes of organic compounds.
December 2005 AX5-94 DRAFT-DO NOT QUOTE OR CITE
-------
1 Ulrich et al. (2002) investigated the effect of ambient PM from Ottawa Canada (EHC-93)
2 on O3-induced inflammation. Male Wistar rats were exposed to 0.8 ppm O3 for 8 h and allowed
3 to recover before intratracheal instillation of 0.5, 1.5, and 5 mg of EHC-93 in 0.3 ml of saline.
4 The high concentrations of PM used were sufficient to induce pulmonary inflammation, which
5 was not exacerbated by pre-exposure to O3. Rats from the combined exposure group did have
6 higher and more persistent lung lavage protein and albumin levels, as well as increased plasma
7 fibrinogen levels when compared to PM exposure alone.
8 The interaction of PM and O3 was further examined in a murine model of ovalbumin
9 (OVA)-induced asthma. Kobzik et al. (2001) investigated whether coexposure to inhaled,
10 concentrated ambient particles (CAPs) from Boston, MA and to O3 could exacerbate asthma-like
11 symptoms. On days 7 and 14 of life, half of the BALB/c mice used in this study were sensitized
12 by intraperitoneal (ip) injection of OVA and then exposed to OVA aerosol on three successive
13 days to create the asthma phenotype. The other half received the ip OVA, but were exposed to a
14 phosphate-buffered saline aerosol (controls). The mice were further subdivided (n > 61/group)
15 and exposed for 5 h to CAPs, ranging from 63 to 1,569 |ig/m3, 0.3 ppm O3, CAPs + O3, or to
16 filtered air. Pulmonary resistance and airway responsiveness to an aerosolized MCh challenge
17 were measured after exposures. A small, statistically significant increase in pulmonary resistance
18 and airway responsiveness, respectively, was found in both normal and asthmatic mice
19 immediately after exposure to CAPs alone and to CAPs + O3, but not to O3 alone or to filtered air.
20 By 24 h after exposure, the responses returned to baseline levels. There were no significant
21 increases in airway inflammation after any of the pollutant exposures. In this well-designed study
22 of a small-animal model of asthma, O3 and CAPs did not appear to be synergistic. In further
23 analysis of the data using specific elemental groupings of the CAPs, the acutely increased
24 pulmonary resistance was found to be associated with the AISi fraction of PM. Thus, some
25 components of concentrated PM25 may affect airway caliber in sensitized animals, but the results
26 are difficult to extrapolate to people with asthma.
27 Six unique animal studies have examined the adverse cardiopulmonary effects of complex
28 mixtures in urban and rural environments of Italy (Gulisano et al., 1997), Spain (Lorz and Lopez,
29 1997), and Mexico (Vanda et al., 1998; Calderon-Garciduefias et al., 2001; Moss et al., 2001).
30 Five of these studies have taken advantage of the differences in pollutant mixtures of urban and
31 rural environments to report primarily morphological changes in the nasopharynx and lower
December 2005 AX5-95 DRAFT-DO NOT QUOTE OR CITE
-------
1 respiratory tract (Gulisano et al., 1997; Lorz and Lopez, 1997; Calderon-Garciduefias et al.,
2 200la) of lambs, pigeons, and dogs, respectively, after natural, continuous exposures to ambient
3 pollution. Each study has provided evidence that animals living in urban air pollutants have
4 greater pulmonary changes than that would occur in a rural and presumably cleaner, environment.
5 The study by Moss et al. (2001) examined the nasal and lung tissue of rats exposed (23 h/day) to
6 Mexico City air for up to 7 weeks and compared them to controls similarly exposed to filtered air.
7 No inflammatory or epithelial lesions were found using quantitative morphological techniques;
8 however, the concentrations of pollutants were low.
9 The ambient air in urban areas, particularly in Southwestern Mexico City, is a complex
10 mixture of particles and gases, including high concentrations of O3 and aldehydes that previously
11 have been shown to cause airway inflammation and epithelial lesions in laboratory animals
12 (Harkema et al., 1994, 1997a,b).
13
14
15 AX5.6 EFFECTS OF OTHER PHOTOCHEMICAL OXIDANTS
16 Complex atmospheric physical and chemical processes involving two classes of precursor
17 pollutants, volatile organic compounds and nitrogen oxides (NOX), lead to the formation of O3 and
18 other photochemical oxidants found in ambient air, such as peroxyacyl nitrates, nitric acid
19 (HNO3), and sulfuric acid, and to the formation of other compounds, such as PM, formaldehyde
20 (HCHO), and other carbonyl compounds (see Chapter 2). Peroxyacetyl nitrate (PAN) and
21 peroxypropionyl nitrate (PPN) are the most abundant of the non-ozone oxidants in ambient air of
22 industrialized areas, other than the inorganic nitrogenous oxidants such as NO2, and possibly
23 HNO3.
24 Tropospheric reactions between O3 and hydrocarbons (e.g., d-limonene) produce epoxides,
25 hydroperoxides, and peroxides. Hydrogen peroxide (H2O2) presumably constitutes the majority
26 of the measured ambient hydroperoxides (< 5 ppb), although a small amount of organic
27 hydroperoxides (ROOH) also may be formed. On the basis of equilibrium calculations and
28 limited data, Friedlander and Yeh (1998) estimated that atmospheric aerosols can carry as high as
29 1 mM of H2O2 and organic hydroperoxides (e.g., hydroxymethylhydroperoxide) in the associated
30 water. High concentrations of liquid phase H2O2 (50 jiM to 1 mM) are known to induce in vitro
31 cell and tissue damage, and recent in vivo studies indicate that 10 and 20 ppb of inhaled H2O2
December 2005 AX5-96 DRAFT-DO NOT QUOTE OR CITE
-------
1 vapor can penetrate the lower lung where it causes inflammation (Morio et al., 2001). It is likely
2 that hygroscopic components of PM transport ambient H2O2 into the lower lung and induce tissue
3 injury as well. Exposure of rats to a H2O2-fine particle mixture (215 or 429 |ig/m3 ammonium
4 sulfate) resulted in increased neutrophil influx, and production of inflammatory mediators by
5 alveolar macrophages (Morio et al., 2001). Hygroscopic secondary organic aerosols (SOA)
6 generated by the O3/hydrocarbon reactions and their co-occurrence with H2O2 also provides
7 another possible mechanism whereby H2O2 can be transported into the lower respiratory tract
8 (e.g., Friedlander and Yeh, 1998).
9 Therefore, acute toxicity of PAN is much lower than O3 and it is unlikely that present
10 ambient PAN levels would affect pulmonary function responses to O3 (Vyskocil et al., 1998).
11 Cytogenetic studies indicate that PAN is not a potent mutagen, clastogen, or DNA damaging
12 agent in mammalian cells in vivo or in vitro at concentrations several orders of magnitude higher
13 than the generally encountered ambient air levels in most cities (Vyskocil et al., 1998; Kligerman
14 et al., 1995; Heddle et al., 1993). Some studies suggest that PAN may be a weak bacterial
15 mutagen at concentrations much higher than exist in present urban atmospheres (DeMarini et al.,
16 2000; Kleindienst et al., 1990).
17
December 2005 AX5-97 DRAFT-DO NOT QUOTE OR CITE
-------
1 REFERENCES
2 Adamson, I. Y. R.; Vincent, R.; Bjarnason, S. G. (1999) Cell injury and interstitial inflammation in rat lung after
3 inhalation of ozone and urban particulates. Am. J. Respir. Cell Mol. Biol. 20: 1067-1072.
4 Aizawa, H.; Koto, H.; Nakano, H.; Inoue, H.; Matsumoto, K.; Takata, S.; Shigyo, M.; Hara, N. (1997) The effects of
5 a specific tachykinin receptor antagonist FK-224 on ozone-induced airway hyperresponsiveness and
6 inflammation. Respirology 2: 261-265.
7 Aizawa, H.; Shigyo, M.; Nakano, H.; Matsumoto, K.; Inoue, H.; Hara, N. (1999a) Effect of the Chinese herbal
8 medicine, Bakumondo-to, on airway hyperresponsiveness induced by ozone exposure in guinea-pigs.
9 Respirology 4: 349-354.
10 Aizawa, H.; Shigyo, M.; Matsumoto, K.; Inoue, H.; Koto, H.; Hara, N. (1999b) PACAP reverses airway
11 hyperresponsiveness induced by ozone exposure in guinea pigs. Respiration 66: 538-542.
12 Alfaro, M. F.; Putney, L.; Tarkington, B. K.; Hatch, G. E.; Hyde, D. M.; Schelegle, E. S. (2004) Effect of rapid
13 shallow breathing on the distribution of 18O-labeled ozone reaction product in the respiratory tract of the rat.
14 Inhalation Toxicol. 16: 77-85.
15 Anderson, S. D. (1996) Challenge tests to assess airway hyperresponsiveness and efficacy of drugs used in the
16 treatment of asthma. J. Aerosol Med. 9: 95-109.
17 Anderson, S. D.; Daviskas, E. (2000) The mechanism of exercise-induced asthma is ... J. Allergy Clin. Immunol. 106:
18 453-459.
19 Arito, H.; Takahashi, M.; Iwasaki, T.; Uchiyama, I. (1997) Age-related changes in ventilatory and heart rate
20 responses to acute ozone exposure in the conscious rat. Ind. Health 35: 78-86.
21 Arsalane, K.; Gosset, P.; Vanhee, D.; Voisin, C.; Hamid, Q.; Tonnel, A.-B.; Wallaert, B. (1995) Ozone stimulates
22 synthesis of inflammatory cytokines by alveolar macrophages in vitro. Am. J. Respir. Cell Mol. Biol.
23 13:60-68.
24 Asplund, P. T.; Ben-Jebria, A.; Rigas, M. L.; Ultman, J. S. (1996) Longitudinal distribution of ozone absorption in the
25 lung: effect of continuous inhalation exposure. Arch. Environ. Health 51: 431-438.
26 Atwal, O. S.; Pemsingh, R. S. (1981) Morphology of microvascular changes and endothelial regeneration in
27 experimental ozone-induced parathyroiditis: III. some pathologic considerations. Am. J. Pathol. 102: 297-307.
28 Atwal, O. S.; Wilson, T. (1974) Parathyroid gland changes following ozone inhalation: a morphologic study.
29 Arch. Environ. Health 28: 91-100.
30 Atwal, O. S.; Samagh, B. S.; Bhatnagar, M. K. (1975) A possible autoimmune parathyroiditis following ozone
31 inhalation: II. a histopathologic, ultrastructural, and immunofluorescent study. Am. J. Pathol. 79: 53-68.
32 Avila-Costa, M. R.; Colin-Barenque, L.; Fortoul, T. L; Machado-Salas, J. P.; Espinosa-Villanueva, J.;
33 Rugerio-Vargas, C.; Rivas-Arancibia, S. (1999) Memory deterioration in an oxidative stress model and its
34 correlation with cytological changes on rat hippocampus CA1. Neurosci. Lett. 270: 107-109.
35 Avital, A.; Springer, C.; Bar-Yishay, E.; Godfrey, S. (1995a) Adenosine, methacholine, and exercise challenges in
36 children with asthma or paediatric chronic obstructive pulmonary disease. Thorax 50: 511-516.
37 Avital, A.; Picard, E.; Uwyyed, K.; Springer, C. (1995b) Comparison of adenosine 5'-monophosphate and
3 8 methacoline for the differentiation of asthma from chronic airway diseases with the use of the auscultative
39 method in very young children. J. Pediatr. 127: 438-440.
40 Bassett, D.; Elbon-Copp, C.; Otterbein, S.; Barraclough-Mitchell, H.; DeLorme, M.; Yang, H. (2001) Inflammatory
41 cell availability affects ozone-induced lung damage. J. Toxicol. Environ. Health A 64: 547-565.
42 Becker, S.; Quay, J.; Koren, H. S. (1991) Effect of ozone on immunoglobulin production by humanB cells in vitro. J.
43 Toxicol. Environ. Health 34: 353-366.
44 Bermudez, E. (2001) Detection of poly(ADP-ribose) synthetase activity in alveolar macrophages of rats exposed to
45 nitrogen dioxide and ozone. Inhalation Toxicol. 13: 69-84.
46 Bermudez, E.; Ferng, S. F.; Castro, C. E.; Mustafa, M. G. (1999) DNA strand breaks caused by exposure to ozone
47 and nitrogen dioxide. Environ. Res. 81: 72-80.
48 Bhalla, D. K. (1996) Alteration of alveolar macrophage chemotaxis, cell adhesion, and cell adhesion molecules
49 following ozone exposure of rats. J. Cell. Physiol. 169: 429-438.
50 Bhalla, D. K. (2002) Interactive effects of cigarette smoke and ozone in the induction of lung injury. Toxicol. Sci. 65:
51 1-3.
52 Bhalla, D. K.; Gupta, S. K. (2000) Lung injury, inflammation, and inflammatory stimuli in rats exposed to ozone.
53 J. Toxicol. Environ. Health 59: 211-228.
54 Bhalla, D. K.; Hoffman, L. (1997) Time course of airway epithelial and inflammatory changes in rats exposed to
55 moderate levels of ozone. Inhalation Toxicol. 9: 829-842.
December 2005 AX5-98 DRAFT-DO NOT QUOTE OR CITE
-------
1 Bhalla, D. K.; Young, C. (1992) Effects of acute exposure to O3 on rats: sequence of epithelial and inflammatory
2 changes in the distal airways. Inhalation Toxicol. 4: 17-31.
3 Bhalla, D. K.; Gupta, S. K.; Reinhart, P. G. (1999) Alteration of epithelial integrity, alkaline phosphatase activity, and
4 fibronectin expression in lungs of rats exposed to ozone. J. Toxicol. Environ. Health A 56: 329-343.
5 Bhalla, D. K.; Reinhart, P. G.; Bai, C.; Gupta, S. K. (2002) Amelioration of ozone-induced lung injury by anti-tumor
6 necrosis factor-a. Toxicol. Sci. 69: 400-408.
7 Bignami, G.; Musi, B.; Dell'Omo, G.; Laviola, G.; Alleva, E. (1994) Limited effects of ozone exposure during
8 pregnancy on physical and neurobehavioral development of CD-I mice. Toxicol. Appl. Pharmacol.
9 129:264-271.
10 Bimonte, H. A.; Nelson, M. E.; Granholm, A. C. (2003) Age-related deficits as working memory load increases:
11 relationships with growth factors. Neurobiol. Aging 24: 37-48.
12 Bolarin, D. M.; Bhalla, D. K.; Kleinman, M. T. (1997) Effects of repeated exposures of geriatric rats to ozone and
13 particle-containing atmospheres: an analysis of bronchoalveolar lavage and plasma proteins. Inhalation
14 Toxicol. 9: 423-434.
15 Bonham, A. C.; Chen, C. Y.; Mutoh, T.; Joad, J. P. (2001) Lung C-fiber CNS reflex: role in the respiratory
16 consequences of extended environmental tobacco smoke exposure ini young guinea pigs. Environ. Health
17 Perspect. 109(suppl. 4): 573-578.
18 Bornholdt, J.; Dybdahl, M.; Vogel, U.; Hansen, M.; Loft, S.; Wallin, H. (2002) Inhalation of ozone inducesDNA
19 strand breaks and inflammation in mice. Mutat. Res. 520: 63-71.
20 Brannan, J. D.; Koskela, H.; Anderson, S. D.; Chew, N. (1998) Responsiveness to Mannitol in asthmatic subjects
21 with exercise- and hyperventiliation-induced asthma. Am. J. Respir. Crit. Care Med. 158: 1120-1126.
22 Bridges, J. P.; Davis, H. W.; Demodarasamy, M.; Kuroki, Y.; Howies, G.; Hui, D. Y.; McCormack, F. X. (2000)
23 Pulmonary surfactant proteins A and D are potent endogenous inhibitors of lipid peroxidation and oxidative
24 cellular injury. J. Biol. Chem. 275: 38848-38855.
25 Broeckaert, F.; Clippe, A.; Wattiez, R.; Falmagne, P.; Bernard, A. (2003) Lung hyperpermeability, Clara-cell
26 secretory potein (CC16), and susceptibility to ozone of five inbred strains of mice. Inhalation Toxicol. 15:
27 1209-1230.
28 Calderon-Garciduenas, L.; Valencia-Salazar, G.; Rodriguez-Alcaraz, A.; Gambling, T. M.; Garcia, R.; Osnaya, N.;
29 Villarreal-Calderon, A.; Devlin, R. B.; Carson, J. L. (2001) Ultrastructural nasal pathology in children
30 chronically and sequentially exposed to air pollutants. Am. J. Respir. Cell Mol. Biol. 24: 132-138.
31 Campos-Bedolla, P.; Vargas, M. H.; Montano, L. M. (2002) Effect of acute ozone exposure on pregnant rat uterus
32 contractile responses. Reprod. Toxicol. 16: 269-273.
33 Cassee, F. R.; Feron, V. J. (1994) Biochemical and histopathological changes in nasal epithelium of rats after 3-day
34 intermittent exposure to formaldehyde and ozone alone or in combination. Toxicol. Lett. 72: 257-268.
35 Catalano, P. J.; Rogus, J.; Ryan, L. M. (1995a) Consequences of prolonged inhalation of ozone onF344/Nrats:
36 collaborative studies. Part X: robust composite scores based on median polish analysis. Cambridge, MA:
37 Health Effects Institute; research report no. 65.
38 Catalano, P. J.; Chang, L.-Y. L.; Harkema, J. R.; Kaden, D. A.; Last, J. A.; Mellick, P. W.; Parks, W. C.; Pinkerton,
39 K. E.; Radhakrishnamurthy, B.; Ryan, L. M.; Szarek, J. L. (1995b) Consequences of prolonged inhalation of
40 ozone on F344/N rats: collaborative studies. Part XI: Integrative summary. Cambridge, MA: Health Effects
41 Institute; research report no. 65.
42 Chang, L.-Y.; Stockstill, B. L.; Menache, M. G.; Mercer, R. R.; Crapo, J. D. (1995) Consequences of prolonged
43 inhalation of ozone on F344/N rats: collaborative studies. Part VIII: morphometric analysis of structural
44 alterations in alveolar regions. Cambridge, MA: Health Effects Institute; pp. 3-39; research report no. 65.
45 Chang, M. M.-J.; Wu, R.; Plopper, C. G.; Hyde, D. M. (1998) IL-8 is one of the major chemokines produced by
46 monkey airway epithelium after ozone-induced injury. Am. J. Physiol. 275: L524-L532.
47 Cheek, J. M.; Buckpitt, A. R.; Li, C.; Tarkington, B. K.; Plopper, C. G. (1994) Ozone injury to alveolar epithelium in
48 vitro does not reflect loss of antioxidant defenses. Toxicol. Appl. Pharmacol. 125: 59-69.
49 Cheek, J. M.; McDonald, R. J.; Rapalyea, L.; Tarkington, B. K.; Hyde, D. M. (1995) Neutrophils enhance removal of
50 ozone-injured alveolar epithelial cells in vitro. Am. J. Physiol. 269: L527-L535.
51 Chen, L. C.; Qu, Q.; Amdur, M. O.; Schlesinger, R. B. (1995) Alteration of pulmonary macrophage intracellular pH
52 following inhalation exposure to sulfuric acid/ozone mixtures. Exp. Lung Res. 21: 113-128.
53 Chen, C.-Y.; Bonham, A. C.; Plopper, C. G.; Joad, J. P. (2003) Plasticity in respiratory motor control: selected
54 contribution: neuroplasticity in nucleus tractus solitarius neurons following episodic ozone exposure in infant
55 primates. J. Appl. Physiol. 94: 819-827.
December 2005 AX5-99 DRAFT-DO NOT QUOTE OR CITE
-------
1 Cheng, P.-W.; Boat, T. F.; Shaikh, S.; Wang, O.-L.; Hu, P.-C.; Costa, D. L. (1995) Differential effects of ozone on
2 lung epithelial lining fluid volume and protein content. Exp. Lung Res. 21: 351-365.
3 Cho, H. Y.; Hotchkiss, J. A.; Harkema, J. R. (1999a) Inflammatory and epithelial responses during the development
4 of ozone-induced mucous cell metaplasia in the nasal epithelium of rats. Toxicol. Sci. 51: 135-145.
5 Cho, H. Y.; Hotchkiss, J. A.; Bennett, C. B.; Harkema, J. R. (1999b) Effects of pre-existing rhinitis on ozone-induced
6 mucous cell metaplasia in rat nasal epithelium. Toxicol. Appl. Pharmacol. 158: 92-102.
7 Cho, H. Y.; Hotchkiss, J. A.; Bennett, C. B.; Harkema, J. R. (2000) Neutrophil-dependent and neutrophil-independent
8 alterations in the nasal epithelium of ozone-exposed rats. Am. J. Respir. Crit. Care Med. 162: 629-636.
9 Cho, H.-Y.; Zhang, L.-Y.; Kleeberger, S. R. (2001) Ozone-induced lung inflammation and hyperreactivity are
10 mediated via tumor necrosis factor-a receptors. Am. J. Physiol. 280: L537-L546.
11 Churg, A.; Brauer, M; Keeling, B. (1996) Ozone enhances the uptake of mineral particles by tracheobronchial
12 epithelial cells in organ culture. Am. J. Respir. Crit. Care Med. 153: 1230-1233.
13 Clausen, P. A.; Wilkins, C. K.; Wolkoff, P.; Nielsen, G. D. (2001) Chemical and biological evaluation of a reaction
14 mixture of R-(+)-limonene/ozone: formation of strong airway irritants. Environ. Int. 26: 511-522.
15 Cohen, M. D.; Zelikoff, J. T.; Qu, Q.; Schlesinger, R. B. (1996) Effects of ozone upon macrophage-interferon
16 interactions. Toxicology 114: 243-252.
17 Cohen, M. D.; Zelikoff, J. T.; Chen, L.-C.; Schlesinger, R. B. (1998) Immunotoxicologic effects of inhaled
18 chromium: role of particle solubility and co-exposure to ozone. Toxicol. Appl. Pharmacol. 152: 30-40.
19 Cohen, M. D.; Sisco, M.; Li, Y.; Zelikoff, J. T.; Schlesinger, R. B. (2001) Ozone-induced modulation of
20 cell-mediated immune responses in the lungs. Toxicol. Appl. Pharmacol. 171: 71-84.
21 Cohen, M. D.; Sisco, M.; Baker, K.; Li, Y.; Lawrence, D.; VanLoveren, H.; Zelikoff, J. T.; Schlesinger, R. B. (2002)
22 Effects of inhaled ozone on pulmonary immune cells critical to antibacterial responses in situ. Inhalation
23 Toxicol. 14: 599-619.
24 Colin-Barenque, L.; Avila-Costa, M. R.; Fortoul, T.; Rugerio-Vargas, C.; Machado-Salas, J. P.; Espinosa-Villanueva,
25 J.; Rivas-Arancibia, S. (1999) Morphologic alteration of the olfactory bulb after acute ozone exposure in rats.
26 Neurosci. Lett. 274: 1-4.
27 Connor, L. M.; Ballinger, C. A.; Albrecht, T. B.; Postlethwait, E. M. (2004) Interfacial phospholipids inhibit ozone
28 reactive absorption-mediated cytotoxicity in vitro. Am. J. Physiol.: 10.1152/ajplung.00397.2003.
29 Cotovio, J.; Onno, L.; Justine, P.; Lamure, S.; Catroux, P. (2001) Generation of oxidative stress in human cutaneous
30 models following in vitro ozone exposure. Toxicol. in Vitro 15: 357-362.
31 Cottet-Emard, J.-M.; Dalmaz, Y.; Pequignot, J.; Peyrin, L.; Pequignot, J.-M. (1997) Long-term exposure to ozone
32 alters peripheral and central catecholamine activity in rats. Pfluegers Arch. 433: 744-749.
33 Creutzenberg, O.; Bellmann, B.; Klingebiel, R.; Heinrich, U.; Muhle, H. (1995) Phagocytosis and chemotaxis of rat
34 alveolar macrophages after a combined or separate exposure to ozone and carbon black. Exp. Toxicol. Pathol.
35 47:202-206.
36 Cross, C. E.; VanDer Vliet, A.; Louie, S.; Thiele, J. J.; Halliwell, B. (1998) Oxidative stress and antioxidants at
37 biosurfaces: plants, skin, and respiratory tract surfaces. Environ. Health Perspect. 106(suppl. 5): 1241-1251.
38 Custodio-Ramirez, V.; Paz, C. (1997) Ozone produces functional deficits in the rat visual pathway.
39 Electroencephalogr. Clin. Neurophysiol. 104: 269-273.
40 Daly, C.; Fox, K.; Henein, M. (2002) Natriuretic peptides in the diagnosis of heart disease-first amongst equals?
41 Int. J. Cardiol. 84: 107-113.
42 DeLorme, M. P.; Yang, H.; Elbon-Copp, C.; Gao, X.; Barraclough-Mitchell, H.; Bassett, D. J. P. (2002)
43 Hyperresponsive airways correlate with lung tissue inflammatory cell changes in ozone-exposed rats. J.
44 Toxicol. Environ. Health Part A 65: 1453-1470.
45 DeMarini, D. M.; Shelton, M. L.; Kohan, M. J.; Hudgens, E. E.; Kleindienst, T. E.; Ball, L. M.; Walsh, D.; de Boer, J.
46 G.; Lewis-Bevan, L.; Rabinowitz, J. R.; Claxton, L. D.; Lewtas, J. (2000) Mutagenicity in lung of Big Blue(R)
47 mice and induction of tandem-base substitutions in Salmonella by the air pollutant peroxyacetyl nitrate (PAN):
48 predicted formation of intrastrand cross-links. Mutat. Res. 457: 41-55.
49 Delacourt, C.; Benoist, M. R.; Waernessyckle, S. (2001) Relationship between bronchial responsiveness and clinical
50 evolution in infants who wheeze. A four-year prospective study. Am. J. Respir. Crit. Care Med.
51 164:1382-1386.
52 Delaunois, A.; Segura, P.; Montano, L. M.; Vargas, M. H.; Ansay, M.; Gustin, P. (1998) Comparison of
53 ozone-induced effects on lung mechanics and hemodynamics in the rabbit. Toxicol. Appl. Pharmacol.
54 150:58-67.
55 DeH'Omo, G.; Fiore, M.; Petruzzi, S.; Alleva, E.; Bignami, G. (1995a) Neurobehavioral development of CD-I mice
56 after combined gestational and postnatal exposure to ozone. Arch. Toxicol. 69: 608-616.
December 2005 AX5-100 DRAFT-DO NOT QUOTE OR CITE
-------
1 Dell'Omo, G.; Wolfer, D.; Alleva, E.; Lipp, H.-P. (1995b) Developmental exposure to ozone induces subtle changes
2 in swimming navigation of adult mice. Toxicol. Lett. 81: 91-99.
3 Depuydt, P.; Joos, G. F.; Pauwels, R. A. (1999) Ambient ozone concentrations induce airway hyperresponsiveness in
4 some rat strains. Eur. Respir. J. 14: 125-131.
5 Dong, W.; Selgrade, M. K.; Gilmour,M. I.; Lange, R. W.; Park, P.; Luster, M. I.; Kari, F. W. (1998) Altered alveolar
6 macrophage function in calorie-restricted rats. Am. J. Respir. Cell Mol. Biol. 19: 462-469
7 Dorado-Martinez, C.; Parades-Carbajal, C.; Mascher, D.; Borgonio-Perez, G.; Rivas-Arancibia, S. (2001) Effects of
8 different ozone doses on memory, motor activity and lipid peroxidation levels, in rats. Int. J. Neurosci.
9 108: 149-161.
10 Dormans, J. A. M. A.; Boere, A. J. F.; vanLoveren, H.; Rombout, P. J. A.; Marra, M.; vanBree, L. (1996)
11 Age-related toxicity in rat lungs following acute and repeated ozone exposure. Inhalation Toxicol. 8: 903-925.
12 Dormans, J. A. M. A.; VanBree, L.; Boere, A. J. F.; Marra, M.; Rombout, P. J. A. (1999) Interspecies differences in
13 time course of pulmonary toxicity following repeated exposure to ozone. Inhalation Toxicol. 11: 309-329.
14 Driscoll, K. E.; Simpson, L.; Carter, J.; Hassenbein, D.; Leikauf, G. D. (1993) Ozone inhalation stimulates expression
15 of a neutrophil chemotactic protein, macrophage inflammatory protein 2. Toxicol. Appl. Pharmacol. 119:
16 306-309.
17 Dye, J. A.; Madden, M. C.; Richards, J. H.; Lehmann, J. R.; Devlin, R. B.; Costa, D. L. (1999) Ozone effects on
18 airway responsiveness, lung injury, and inflammation. Comparative rat strain and in vivo/in vitro
19 investigations. Inhalation Toxicol. 11: 1015-1040.
20 El-Fawal, H. A. N.; McGovern, T.; Schlesinger, R. B. (1995) Nonspecific bronchial responsiveness assessed in vitro
21 following acute inhalation exposure to ozone and ozone/sulfuric acid mixtures. Exp. Lung Res. 21: 129-139.
22 Elder, A. C. P.; Gelein, R.; Finkelstein J. N.; Cox, C.; Oberdorster, G. (2000a) Endotoxin priming affects the lung
23 response to ultrafine particles and ozone in young and old rats. In: Phalen, R. F., ed. Inhalation toxicology:
24 proceedings of the third colloquium on paniculate air pollution and human health (first special issue); June,
25 1999; Durham, NC. Inhalation Toxicol. 12(suppl. 1): 85-98.
26 Elder, A. C. P.; Gelein, R.; Finkelstein, J. N.; Cox, C.; Oberdorster, G. (2000b) Pulmonary inflammatory response to
27 inhaled ultrafine particles is modified by age, ozone exposure, and bacterial toxin. In: Grant, L. D., ed.
28 PM2000: paniculate matter and health. Inhalation Toxicol. 12(suppl. 4): 227-246.
29 Elsayed, N. M. (2001) Diet restriction modulates lung response and survivability of rats exposed to ozone.
30 Toxicology 159: 171-182.
31 Evans, M. J.; Fanucchi, M. V.; Baker, G. L.; Van Winkle, L. S.; Pantle, L. M.; Nishio, S. J.; Schelegle, E. S.;
32 Gershwhin, L. J.; Miller, L. A.; Hyde, D. M.; Sannes, P. L.; Plopper, C. G. (2003) Atypical development of the
33 trachea! basement membrane zone of infant rhesus monkeys exposed to ozone and allergen. Am. J. Physiol.
34 285:L931-L939.
35 Fakhrzadeh, L.; Laskin, J. D.; Laskin, D. L. (2002) Deficiency in inducible nitric oxide synthase protects mice from
36 ozone-induced lung inflammation and tissue injury. Am. J. Respir. Cell Mol. Biol. 26: 413-419.
37 Fanucchi, M. V.; Hotchkiss, J. A.; Harkema, J. R. (1998) Endotoxin potentiates ozone-induced mucous cell
38 metaplasia in rat nasal epithelium. Toxicol. Appl. Pharmacol. 152: 1-9.
39 Fanucchi, M. V; Wong, V J.; Hinds, D.; Tarkington, B. K.; Van Winkle, L. S.; Evans, M. J.; Plopper, C. G. (2000)
40 Repeated episodes of exposure to ozone alters postnatal development of distal conducting airways in infant
41 rhesus monkeys. Am. J. Respir. Crit. Care Med. 161: A615.
42 Farman, C. A.; Pinkerton, K. E.; Rajini, P.; Witschi, H.; Last, J. A. (1997) Evolution of lung lesions in rats exposed to
43 mixtures of ozone and nitrogen dioxide. Inhalation Toxicol. 9: 647-677'.
44 Farman, C. A.; Watkins, K.; Van Hoozen, B.; Last, J. A.; Witschi, H.; Pinkerton, K. E. (1999) Centriacinar
45 remodeling and sustained procollagen gene expression after exposure to ozone and nitrogen dioxide. Am. J.
46 Respir. Cell Mol. Biol. 20: 303-311.
47 Ferng, S.-F.; Castro, C. E.; Afifi, A. A.; Bermudez, E.; Mustafa, M. G. (1997) Ozone-induced DNA strand breaks in
48 guinea pig tracheobronchial epithelial cells. J. Toxicol. Environ. Health 51: 353-367.
49 Folkerts, G.; Busse, W. W.; Nijkamp, F. P.; Sorkness, R.; Gern, J. E. (1998) Virus-induced airway
50 hyperresponsiveness and asthma. Am. J. Respir. Crit. Care Med. 157: 1708-1720.
51 Foster, W. M.; Freed, A. N. (1999) Regional clearance of solute from peripheral airway epithelia: recovery after
52 sublobar exposure to ozone. J. Appl. Physiol. 86: 641-646.
53 Frampton, M. W.; Pryor, W. A.; Cueto, R.; Cox, C.; Morrow, P. E.; Utell, M. J. (1999) Aldehydes (nonanal and
54 hexanal) in rat and human bronchoalveolar lavage fluid after ozone exposure. Cambridge, MA: Health Effects
55 Institute; research report no. 90. Available: www.healtheffects.org/Pubs/Frampton-C.pdf [2000, February 9].
December 2005 AX5-101 DRAFT-DO NOT QUOTE OR CITE
-------
1 Friedlander, S. K.; Yeh, E. K. (1998) The submicron atmospheric aerosol as a carrier of reactive chemical species:
2 case of peroxides. Appl. Occup. Environ. Hyg. 13: 416-420.
3 Garssen, I; Van Bree, L.; Van Der Vliet, H.; Van Loveren, H. (1997) Ozone-induced impairment of pulmonary type
4 IV hypersensitivity and airway hyperresponsiveness in mice. Inhalation Toxicol. 9: 581-599.
5 Gohil, K.; Cross, C. E.; Last, J. A. (2003) Ozone-induced disruptions of lung transcriptomes. Biochem. Biophys. Res.
6 Commun. 305: 719-728.
7 Goldsmith, C.-A. W.; Ning, Y.-Y.; Qin, G.; Imrich, A.; Lawrence, J.; Murthy, G. G., K.; Catalano, P. J.; Kobzik, L.
8 (2002) Combined air pollution particle and ozone exposure increases airway responsiveness in mice. Inhalation
9 Toxicol. 14: 325-347.
10 Gonzalez-Pina, R.; Paz, C. (1997) Brain monoamine changes in rats after short periods of ozone exposure.
11 Neurochem. Res. 22: 63-66.
12 Guerrero, A. L.; Dorado-Martinez, C.; Rodriguez, A.; Pedroza-Rios, K.; Borgonio-Perez, G.; Rivas-Arancibia, S.
13 (1999) Effects of vitamin E on ozone-induced memory deficits and lipid peroxidation in rats. NeuroReport 10:
14 1689-1692.
15 Gulisano, M; Marceddu, S.; Barbara, A.; Pacini, A.; Buiatti, E.; Martini, A.; Pacini, P. (1997) Damage to the
16 nasopharyngeal mucosa induced by current levels of urban air pollution: a field study in lambs. Eur. Respir. J.
17 10:567-572.
18 Gtinther, T.; Hollriegl, V; Vormann, J. (1993) Perinatal development of iron and antioxidant defence systems.
19 J. Trace Elem. Electrolytes Health Dis. 7: 47-52.
20 Gupta, S. K.; Reinhart, P. G.; Bhalla, D. K. (1998) Enhancement of fibronectin expression in rat lung by ozone and an
21 inflammatory stimulus. Am. J. Physiol. 275: L330-L335.
22 Haddad, E.-B.; Liu, S. F.; Salmon, M.; Robichaud, A.; Barnes, P. J.; Chung, K. F. (1995) Expression of inducible
23 nitric oxide synthase mRNA in Brown Norway rats exposed to ozone: effect of dexamethasone. Eur. J.
24 Pharmacol. Environ. Toxicol. Pharmacol. Sect. 293: 287-290.
25 Haddad, E.-B.; Salmon, M.; Koto, H.; Barnes, P. J.; Adcock, I.; Chung, K. F. (1996) Ozone induction of
26 cytokine-induced neutrophil chemoattractant (CINC) and nuclear factor-kappa b in rat lung: inhibition by
27 corticosteroids. FEES Lett. 379: 265-268.
28 Hamilton, R. F.; Li, L.; Eschenbacher, W. L.; Szweda, L.; Holian, A. (1998) Potential involvement of
29 4-hydroxynonenal in the response of human lung cells to ozone. Am. J. Physiol. 274: L8-L16.
30 Hanna, L. M.; Frank, R.; Scherer, P. W. (1989) Absorption of soluble gases and vapors in the respiratory system.
31 In: Chang, H. K.; Paiva, M., eds. Respiratory physiology: an analytical approach. New York, NY: Marcel
32 Dekker, Inc.; pp. 277-316. (Lenfant, C., ed. Lung biology in health and disease: v. 40).
33 Harkema, J. R.; Morgan, K. T.; Gross, E. A.; Catalano, P. J.; Griffith, W. C. (1994) Consequences of prolonged
34 inhalation of ozone on F344/N rats: collaborative studies. Part VII: effects on the nasal mucociliary apparatus.
3 5 Cambridge, MA: Health Effects Institute; research report no. 65.
36 Harkema, J. R.; Catalano, P. J.; Hotchkiss, J. A. (1997a) Consequences of prolonged inhalation of ozone on F344/N
37 rats: collaborative studies. Part XII. Atrophy of bone in nasal turbinates. Cambridge, MA: Health Effects
3 8 Institute; research report no. 65.
39 Harkema, J. R.; Hotchkiss, J. A.; Griffith, W. C. (1997b) Mucous cell metaplasia in rat nasal epithelium after a
40 20-month exposure to ozone: a morphometric study of epithelial differentiation. Am. J. Respir. Cell Mol. Biol.
41 16:521-530.
42 Harkema, J. R.; Hotchkiss, J. A.; Barr, E. B.; Bennett, C. B.; Gallup, M.; Lee, J. K.; Basbaum, C. (1999) Long-lasting
43 effects of chronic ozone exposure on rat nasal epithelium. Am. J. Respir. Cell Mol. Biol. 20: 517-529.
44 Kara, R.; Paz, C. (1993) Effects of ozone exposure during pregnancy on ontogeny of sleep in rats. Neurosci. Lett.
45 164:67-70.
46 Hawgood, S.; Poulain, F. R. (2001) The pulmonary collectins and surfactant metabolism. Annu. Rev. Physiol.
47 63:495-519.
48 Hawgood, S.; Ochs, M.; Jung, A.; Akiyama, J.; Allen, L.; Brown, C.; Edmondson, J.; Levitt, S.; Carlson, E.;
49 Gillespie, A. M.; Villar, A.; Epstein, C. J.; Poulain, F. R. (2002) Sequential targeted deficiency of SP-A and -D
50 leads to progressive alveolar lipoproteinosis and emphysema. Am. J. Physiol. 283: L1002-L1010.
51 Heddle, J. A.; Shepson, P. B.; Gingerich, J. D.; So, K. W. (1993) Mutagenicity of peroxyacetyl nitrate (PAN) in vivo:
52 tests for somatic mutations and chromosomal aberrations. Environ. Mol. Mutagen. 21: 58-66.
53 Herbert, R. A.; Hailey, J. R.; Grumbein, S.; Chou, B. J.; Sills, R. C.; Haseman, J. K.; Goehl, T.; Miller, R. A.;
54 Roycroft, J. H.; Boorman, G. A. (1996) Two-year and lifetime toxicity and carcinogenicity studies of ozone in
55 B6C3F1 mice. Toxicol. Pathol. 24: 539-548.
December 2005 AX5-102 DRAFT-DO NOT QUOTE OR CITE
-------
1 Highfill, J. W.; Watkinson, W. P. (1996) Ozone toxicity in the rat. II. Modeling changes due to ambient temperatures
2 and duration. J. Appl. Physiol. 80: 1811-1818.
3 Hoffer, E.; Baum, Y.; Tabak, A.; Frevert, C. (1999) Adhesion molecules of blood polymorphonuclear leukocytes and
4 alveolar macrophages in rats: modulation by exposure to ozone. Hum. Exp. Toxicol. 18: 547-551.
5 Holt, P. G.; Macaubas, C.; Stumbles, P. A.; Sly, P. D. (1999) The role of allergy in the development of asthma.
6 Nature (London, U. K.) 402(suppl. 25): B12-B16.
7 Hotchkiss, J. A.; Harkema, J. R.; Johnson, N. F. (1997) Kinetics of nasal epithelial cell loss and proliferation in F344
8 rats following a single exposure to 0.5 ppm ozone. Toxicol. Appl. Pharmacol. 143: 75-82.
9 Hotchkiss, J. A.; Hilaski, R.; Cho, H.; Regan, K.; Spencer, P.; Slack, K.; Harkema, J. R. (1998) Fluticasone
10 propionate attenuates ozone-induced rhinitis and mucous cell metaplasia in rat nasal airway epithelium. Am. J.
11 Respir. Cell Mol. Biol. 18: 91-99.
12 Huffman, L. J.; Judy, D. J.; Brumbaugh, K.; Frazer, D. G.; Reynolds, J. S.; McKinney, W. G.; Goldsmith, W. T.
13 (2001) Hyperthyroidism increases the risk of ozone-induced lung toxicity in rats. Toxicol. Appl. Pharmacol.
14 173: 18-26.
15 Huitron-Resendiz, S.; Custodio-Ramirez, V.; Escalante-Membrillo, C.; Gonzalez-Pina, R.; Paz, C. (1994) Sleep
16 alterations and brain regional changes of serotonin and its metabolite in rats exposed to ozone. Neurosci. Lett.
17 177: 119-122.
18 Hyde, D. M; Miller, L. A.; McDonald, R. J.; Stovall, M. Y.; Wong, V.; Pinkerton, K. E.; Wegner, C. D.; Rothlein,
19 R.; Plopper, C. G. (1999) Neutrophils enhance clearance of necrotic epithelial cells in ozone-induced lung
20 injury in rhesus monkeys. Am. J. Physiol. 277: LI 190-L1198.
21 Igarashi, A.; lijima, H.; Tamura, G.; Shirato, K. (1998) Tazanolast inhibits ozone-induced airway
22 hyperresponsiveness in guinea pigs. Am. J. Respir. Crit. Care Med. 157: 1531-1535.
23 lijima, M. K.; Kobayashi, T.; Kamada, H.; Shimojo, N. (2001) Exposure to ozone aggravates nasal allergy-like
24 symptoms in guinea pigs. Toxicol. Lett. 123: 77-85.
25 Inoue, H.; Aizawa, H.; Nakano, H.; Matsumoto, K.; Kuwano, K.; Nadel, J. A.; Kara, N. (2000) Nitric oxide synthase
26 inhibitors attenuate ozone-induced airway inflammation in guinea pigs: possible role of interleukin-8. Am. J.
27 Respir. Crit. Care Med. 161: 249-256.
28 Ishii, Y.; Yang, H.; Sakamoto, T.; Nomura, A.; Hasegawa, S.; Hirata, F.; Bassett, D. J. P. (1997) Rat alveolar
29 macrophage cytokine production and regulation of neutrophil recruitment following acute ozone exposure.
30 Toxicol. Appl. Pharmacol. 147: 214-223.
31 Ishii, Y.; Shirato, M.; Nomura, A.; Sakamoto, T.; Uchida, Y.; Ohtsuka, M.; Sagai, M.; Hasegawa, S. (1998) Cloning
32 of rat eotaxin: ozone inhalation increases mRNA and protein expression in lungs of brown Norway rats. Am. J.
33 Physiol. 274: L171-L176.
34 Ishii, Y.; Hashimoto, K.; Hirano, K.; Morishima, Y.; Mochizuki, M.; Masuyama, K.; Nomura, A.; Sakamoto, T.;
35 Uchida, Y.; Sagai, M.; Sekizawa, K. (2000a) Ebselen decreases ozone-induced pulmonary inflammation in
36 rats. Lung 178: 225-234.
37 Ishii, Y.; Hirano, K.; Morishima, Y.; Masuyama, K.; Goto, Y.; Nomura, A.; Sakamoto, T.; Uchida, Y.; Sagai, M.;
3 8 Sekizawa, K. (2000b) Early molecular and cellular events of oxidant-induced pulmonary fibrosis in rats.
39 Toxicol. Appl. Pharmacol. 167: 173-181.
40 Iwasaki, T.; Takahashi, M.; Saito, H.; Arito, H. (1998) Adaptation of extrapulmonary responses to ozone exposure in
41 conscious rats. Ind. Health 36: 57-60.
42 Jang, A.-S.; Choi, I.-S.; Koh, Y.-I.; Park, C.-S.; Lee, J.-S. (2002) The relationship between alveolar epithelial
43 proliferation and airway obstruction after ozone exposure. Allergy 57: 737-740.
44 Jimba, M.; Skornik, W. A.; Killingsworth, C. R.; Long, N. C.; Brain, J. D.; Shore, S. A. (1995) Role of C fibers in
45 physiological responses to ozone in rats. J. Appl. Physiol. 78: 1757-1763.
46 Joad, J. P.; Kott, K. S.; Bonham, A. C. (1998) Exposing guinea pigs to ozone for 1 wk enhances responsiveness of
47 rapidly adapting receptors. J. Appl. Physiol. 84: 1190-1197.
48 Joad, J. P.; Brie, J. M.; Weir, A. J.; Putney, L.; Hyde, D. M.; Postlewait, E. M.; Plopper, C. G. (2000) Effect of
49 respiratory pattern on ozone injury to the airways of isolated rat lungs. Toxicol. Appl. Pharmacol. 169: 26-32.
50 Johnston, C. J.; Stripp, B. R.; Reynolds, S. D.; Avissar, N. E.; Reed, C. K.; Finkelstein, J. N. (1999a) Inflammatory
51 and antioxidant gene expression in C57BL/6J mice after lethal and sublethal ozone exposures. Exp. Lung Res.
52 25:81-97.
53 Johnston, C. J.; Finkelstein, J. N.; Oberdorster, G.; Reynolds, S. D.; Stripp, B. R. (1999b) Clara cell secretory
54 protein-deficient mice differ from wild-type mice in inflammatory chemokine expression to oxygen and ozone,
55 but not to endotoxin. Exp. Lung Res. 25: 7-21.
December 2005 AX5-103 DRAFT-DO NOT QUOTE OR CITE
-------
1 Johnston, C. I; Reed, C. K.; Avissar, N. E.; Gelein, R.; Finkelstein, J. N. (2000a) Antioxidant and inflammatory
2 response after acute nitrogen dioxide and ozone exposures in C57B1/6 mice. Inhalation Toxicol. 12: 187-203.
3 Johnston, C. J.; Oberdorster, G.; Gelein, R.; Finkelstein, J. N. (2000b) Newborn mice differ from adult mice in
4 chemokine and cytokine expression to ozone, but not to endotoxin. Inhalation Toxicol. 12: 205-224.
5 Johnston, C. J.; Oberdorster, G.; Finkelstein, J. N. (2001) Recovery from oxidant-mediated lung injury: response of
6 metallothionein, MIP-2, and MCP-1 to nitrogen dioxide, oxygen, and ozone exposures. Inhalation Toxicol. 13:
7 689-702.
8 Johnston, C. J.; Oberdorster, G.; Gelein, R.; Finkelstein, J. N. (2002) Endotoxin potentiates ozone-induced pulmonary
9 chemokine and inflammatory responses. Exp. Lung Res. 28: 419-433.
10 Kafoury, R. M; Pryor, W. A.; Squadrito, G. L.; Salgo, M. G.; Zou, X.; Friedman, M. (1999) Induction of
11 inflammatory mediators in human airway epithelial cells by lipid ozonation products. Am. J. Respir. Crit. Care
12 Med. 160: 1934-1942.
13 Kenyon, N. J.; Van Der Vliet, A.; Schock, B.C.; Okamoto, T.; McGrew, G. M.; Last, J. A. (2002) Susceptibility to
14 ozone-induced acute lung injury in iNOS-deficient mice. Am. J. Physiol. 282: L540-L545.
15 Kim, M. Y.; Son, J. W.; Cho, M. H.; Choi, C. S.; Chae, C. H.; Lee, M. H. (2001) Oviductural carcinoma in B6C3F1
16 female mice exposed to 0.5 ppm ozone. Vet. Hum. Toxicol. 43: 370-372.
17 Kimmel, T. A.; Chen, L. C.; Bosland, M. C.; Nadziejko, C. (1997) Influence of acid aerosol droplet size on structural
18 changes in the rat lung caused by acute exposure to sulfuric acid and ozone. Toxicol. Appl. Pharmacol. 144:
19 348-355.
20 Kleeberger, S. R.; Levitt, R. C.; Zhang, L.-Y.; Longphre, M.; Harkema, J.; Jedlicka, A.; Eleff, S. M.; DiSilvestre, D.;
21 Holroyd, K. J. (1997) Linkage analysis of susceptibility to ozone-induced lung inflammation in inbred mice.
22 Nat. Genet. 17: 475-478.
23 Kleeberger, S. R.; Reddy, S.; Zhang, L.-Y.; Jedlicka, A. E. (2000) Genetic susceptibility to ozone-induced lung
24 hyperpermeability: role of toll-like receptor 4. Am. J. Respir. Cell Mol. Biol. 22: 620-627.
25 Kleeberger, S. R.; Reddy, S. P.; Zhang, L.-Y.; Cho, H.-Y.; Jedlicka, A. E. (2001a) Toll-like receptor 4 mediates
26 ozone-induced murine lung hyperpermeability via inducible nitric oxide synthase. Am. J. Physiol. 280:
27 L326-L333.
28 Kleeberger, S. R.; Ohtsuka, Y.; Ahang, L.-Y.; Longphre, M. (200Ib) Airway responses to chronic ozone exposure are
29 partially mediated through mast cells. J. Appl. Physiol. 90: 713-723.
30 Kleindienst, T. E.; Shepson, P. B.; Smith, D. F.; Hudgens, E. E.; Nero, C. M.; Cupitt, L. T.; Bufalini, J. J.; Claxton, L.
31 D. (1990) Comparison of mutagenic activities of several peroxyacyl nitrates. Environ. Mol. Mutagen.
32 16:70-80.
33 Kleinman, M. T.; Mautz, W. J.; Bjarnason, S. (1999) Adaptive and non-adaptive responses in rats exposed to ozone,
34 alone and in mixtures, with acidic aerosols. Inhalation Toxicol. 11: 249-264.
35 Kleinman, M. T.; Bufalino, C.; Rasmussen, R.; Hyde, D.; Bhalla, D. K.; Mautz, W. J. (2000) Toxicity of chemical
36 components of ambient fine paniculate matter (PM2 5) inhaled by aged rats. J. Appl. Toxicol. 20: 357-364.
37 Kligerman, A. D.; Mottus, K.; Erexson, G. L. (1995) Cytogenetic analyses of the in vitro and in vivo responses of
38 murine cells to peroxyacetyl nitrate (PAN). Mutat. Res. 341: 199-206.
39 Kobzik, L.; Goldsmith, C.-A. W.; Ning, Y. Y.; Qin, G.; Morgan, B.; Imrich, A.; Lawrence J.; Murthy, G. G. K.;
40 Catalano, P. J. (2001) Effects of combined ozone and air pollution particle exposure in mice. Boston, MA:
41 Health Effects Institute; research report no. 106. Available: http://www.healtheffects.org/Pubs/Kobzik.pdf
42 [27 January 2003].
43 Kodavanti, U. P.; Hatch, G. E.; Starcher, B.; Giri, S. N.; Winsett, D.; Costa, D. L. (1995) Ozone-induced pulmonary
44 functional, pathological, and biochemical changes in normal and vitamin C-deficient guinea pigs. Fundam.
45 Appl. Toxicol. 24: 154-164.
46 Koike, E.; Kobayashi, T.; Nelson, D. J.; Me William, A. S.; Holt, P. G. (1998) Effect of ozone exposure on alveolar
47 macrophage-mediated immunosuppressive activity in rats. Toxicol. Sci. 41: 217-223.
48 Koike, E.; Kobayashi, T.; Murakami, M.; McWilliam, A. S.; Holt, P. G. (1999) Mechanisms of ozone-induced
49 inhibitory effect of bronchoalveolar lavage fluid on alveolar macrophage-mediated immunosuppressive
50 activity in rats. J. Leukoc. Biol. 66: 75-82.
51 Koike, E.; Kobayashi, T.; Shimojo, N. (2001) Ozone exposure enhances expression of cell-surface molecules
52 associated with antigen-presenting activity on bronchoalveolar lavage cells in rats. Toxicol. Sci. 63: 115-124.
53 Koto, H.; Aizawa, H.; Takata, S.; Inoue, H.; Kara, N. (1995) An important role of tachykinins in ozone-induced
54 airway hyperresponsiveness. Am. J. Respir. Crit. Care Med. 151: 1763-1769.
December 2005 AX5-104 DRAFT-DO NOT QUOTE OR CITE
-------
1 Koto, H.; Salmon, M; Haddad el-B.; Huang, T.-J.; Zagorski, I; Chung, K. F. (1997) Role of cytokine-induced
2 neutrophil chemoattractant (CINC) in ozone-induced airway inflammation and hyperresponsiveness. Am. J.
3 Respir. Crit. Care Med. 156: 234-239.
4 Koyama, Y.; Hayaishi, O. (1994) Modulation by prostaglandins of activity of sleep-related neurons in the
5 preoptic/anterior hypothalamic areas in rats. Brain Res. Bull. 33: 367-372.
6 Kudo, M.; Nishikawa, M.; Ikeda, H.; Okubo, T. (1996) Involvement of superoxide anions in ozone-induced airway
7 hyperresponsiveness in unanesthetized guinea pigs. Environ. Toxicol. Pharmacol. 2: 25-30.
8 Larson, S. D.; Schelegle, E. S.; Walby, W. F.; Gershwin, L. J.; Fanuccihi, M. V.; Evans, M. J.; Joad, J. P.;
9 Tarkington, B. K.; Hyde, D. M.; Plopper, C. G. (2004) Postnatal remodeling of the neural components of the
10 epithelial-mesenchymal trophic unit in the proximal airways of infant rhesus monkeys exposed to ozone and
11 allergen. Toxicol. Appl. Pharmacol. 194: 211-220.
12 Laskin, D. L.; Laskin, J. D. (2001) Role of macrophages and inflammatory mediators in chemically induced toxicity.
13 Toxicology (Ireland) 160: 111-118.
14 Laskin, D. L.; Pendino, K. J.; Punjabi, C. J.; del Valle, M. R.; Laskin, J. D. (1994) Pulmonary and hepatic effects of
15 inhaled ozone in rats. Environ. Health Perspect. 102(suppl. 10): 61-64.
16 Laskin, J. D.; Heck, D. E.; Laskin, D. L. (1996) Nitric oxide production in the lung and liver following inhalation of
17 the pulmonary irritant ozone. In: Snyder, R.; Kocsis, J. J.; Sipes, I. G.; Kalf, G. F.; Jollow, D. J.; Greim, H.;
18 Monks, T. J.; Witmer, C. M., eds. Biological Reactive Intermediates V: Basic Mechanistic Research in
19 Toxicology and Human Risk Assessment: proceedings of the Fifth International Symposium; January 1995;
20 Munich, Germany. Adv. Exp. Med. Biol. 387: 141-146.
21 Laskin, D. L.; Heck, D. E.; Laskin, J. D. (1998a) Role of inflammatory cytokines and nitric oxide in hepatic and
22 pulmonary toxicity. Toxicol. Lett. 102-103: 289-293.
23 Laskin, D. L.; Sunil, V.; Guo, Y.; Heck, D. E.; Laskin, J. D. (1998b) Increased nitric oxide synthase in the lung after
24 ozone inhalation is associated with activation of NF-KB. Environ. Health Perspect. 106(suppl. 5): 1175-1178.
25 Laskin, D. L.; Fakhrzadeh, L.; Heck, D. E.; Gerecke, D.; Laskin, J. D. (2002) Upregulation of phosphoinositide
26 3-kinase and protein kinase B in alveolar macrophages following ozone inhalation. Role of NF-KB and
27 STAT-1 in ozone-induced nitric oxide production and toxicity. Mol. Cell. Biochem. 234-235: 91-98.
28 Last, J. A.; Pinkerton, K. E. (1997) Chronic exposure of rats to ozone and sulfuric acid aerosol: biochemical and
29 structural responses. Toxicology 116: 133-146.
30 Lavnikova, N.; Prokhorova, S.; Lakhotia, A. V.; Gordon, R.; Laskin, D. L. (1998) Distinct inflammatory responses of
31 adherent vascular lung neutrophils to pulmonary irritants. J. Inflammation 48: 56-66
32 Lee, C.; Watt, K. C.; Chang, A. M.; Plopper, C. G.; Buckpitt, A. R.; Pinkerton, K. E. (1998) Site-selective differences
33 in cytochrome P450 isoform activities: comparison of expression in rat and rhesus monkey lung and induction
34 in rats. Drug Metab. Dispos. 26: 396-400.
35 Lemos, M.; Lichtenfels, A. J. F. C.; Amaro, E., Jr.; Macchione, M.; Martins, M. A.; King, M.; Bohm, G. M.; Saldiva,
36 P. H. N. (1994) Quantitative pathology of nasal passages in rats exposed to urban levels of air pollution.
37 Environ. Res. 66: 87-95.
38 Long, N. C.; Suh, J.; Morrow, J. D.; Schiestl, R. H.; KrishnaMurthy, G. G.; Brain, J. D.; Frei, B. (2001) Ozone
3 9 causes lipid peroxidation but little antioxidant depletion in exercising and nonexercising hamsters. J. Appl.
40 Physiol. 91: 1694-1700.
41 Longphre, M.; Zhang, L.-Y.; Harkema, J. R.; Kleeberger, S. R. (1996) Mast cells contribute to O3-induced epithelial
42 damage and proliferation in nasal and bronchial airways of mice. J. Appl. Physiol. 80: 1322-1330.
43 Longphre, M.; Zhang, L.-Y.; Harkema, J. R.; Kleeberger, S. R. (1999) Ozone-induced pulmonary inflammation and
44 epithelial proliferation are partially mediated by PAF. J. Appl. Physiol. 86: 341-349.
45 Lorz, C.; Lopez, J. (1997) Incidence of air pollution in the pulmonary surfactant system of the pigeon (Columbia
46 livid). Anat. Rec. 249: 206-212.
47 Lum, H.; Tyler, W. S.; Hyde, D. M.; Plopper, C. G. (1983) Morphometry of in situ and lavaged pulmonary alveolar
48 macrophages from control and ozone-exposed rats. Exp. Lung Res. 5: 61-78.
49 Madden, M. C.; Richards, J. H.; Dailey, L. A.; Hatch, G. E.; Ohio, A. J. (2000) Effect of ozone on diesel exhaust
50 particle toxicity in rat lung. Toxicol. Appl. Pharmacol. 168: 140-148.
51 Mango, G. W.; Johnston, C. J.; Reynolds, S. D.; Finkelstein, J. N.; Plopper, C. G.; Stripp, B. R. (1998) Clara cell
52 secretory protein deficiency increases oxidant stress response in conducting airways. Am. J. Physiol. 275:
53 L348-L356.
54 Matsubara, S.; Fushimi, K.; Kaminuma, O.; Kikkawa, H.; Shimazu, N.; Iwasaki, H.; Ikezawa, K. (1995) Importance
55 of impairment of the airway epithelium for ozone-induced airway hyperresponsiveness in guinea pigs. Jpn. J.
56 Pharmacol. 67: 375-382.
December 2005 AX5-105 DRAFT-DO NOT QUOTE OR CITE
-------
1 Matsubara, S.; Kikkawa, H.; Kaminuma, O.; Ikezawa, K. (1997a) Angiotensin-converting enzyme inhibitors can
2 potentiate ozone-induced airway hyperresponsiveness. Eur. J. Pharmacol. 337: 259-265.
3 Matsubara, S.; Fushimi, K.; Kaminuma, O.; Kikkawa, H.; Ikezawa, K.; Naito, K. (1997b) Prevention of
4 ozone-induced airway hyperresponsiveness and epithelial injury by phosphodiesterase inhibitors in guinea
5 pigs. Environ. Toxicol. Pharmacol. 3: 201-209.
6 Matsumoto, K.; Aizawa, H.; Inoue, H.; Koto, H.; Nakano, H.; Hara, N. (1999) Role of neutrophil elastase in
7 ozone-induced airway responses in guinea-pigs. Eur. Respir. J. 14: 1088-1094.
8 Mautz, W. J. (2003) Exercising animal models in inhalation toxicology: interactions with ozone and formaldehyde.
9 Environ. Res. 92: 14-26.
10 Mautz, W. J.; Kleinman, M. T.; Bhalla, D. K.; Phalen, R. F. (2001) Respiratory tract responses to repeated inhalation
11 of an oxidant and acid gas-particle air pollutant mixture. Toxicol. Sci. 61: 331-341.
12 McGraw, D. W.; Forbes, S. L.; Mak, J. C. W.; Witte, D. P.; Carrigan, P. E.; Leikauf, G. D.; Liggett, S. B. (2000)
13 Transgenic overexpression of beta(2)-adrenergic receptors in airway epithelial cells decreases
14 bronchoconstriction. Am. J. Physiol. 279: L379-L389.
15 McKinney, W. J.; Jaskot, R. H.; Richards, J. H.; Costa, D. L.; Dreher, K. L. (1998) Cytokine mediation of
16 ozone-induced pulmonary adaptation. Am. J. Respir. Cell. Mol. Biol. 18: 696-705.
17 Miller, L. A.; Barnett, N. L.; Sheppard, D.; Hyde, D. M. (2001) Expression of the P6 integrin subunit is associated
18 with sites of neutrophil influx in lung epithelium. J. Histochem. Cytochem. 49: 41-48.
19 Morio, L. A.; Hooper, K. A.; Brittingham, J.; Li, T.-H.; Gordon, R. E.; Turpin, B. J.; Laskin, D. L. (2001) Tissue
20 injury following inhalation of fine paniculate matter and hydrogen perioxide is associated with altered
21 production of inflammatory mediators and antioxidants by alveolar macrophages. Toxicol. Appl. Pharmacol.
22 177: 188-199.
23 Moss, O. R.; Gross, E. A.; James, R. A.; Janszen, D. B.; Ross, P. W.; Roberts, K. C.; Howard, A. M.; Harkema, J. R.;
24 Calderon-Garciduenas, L.; Morgan, K. T. (2001) Respiratory tract toxicity in rats exposed to Mexico City air.
25 Cambridge, MA: Health Effects Institute; research report no. 100. Available:
26 http://www.healtheffects.org/pubs-research.htm [15 May, 2003].
27 Mticke, W. (1996) The environment and the eye. Topics of ophthalmic toxicology. Leban. Med. J. 44: 146-150.
28 Murphy, D. J. (2002) Assessment of respiratory function in safety pharmacology. Fundam. Clin. Pharmacol.
29 16: 183-196.
30 Mutoh, T.; Joad, J. P.; Bonham, A. C. (2000) Chronic passive cigarette smoke exposure augments bronchopulmonary
31 C-fibre inputs to nucleus tractus solitarii neurones and reflex output in young guinea-pigs. J. Physiol. (London)
32 523:223-233.
33 Nakano, H.; Aizawa, H.; Matsumoto, K.; Fukuyama, S.; Inoue, H.; Hara, N. (2000) Cyclooxygenase-2 participates in
34 the late phase of airway hyperresponsiveness after ozone exposure in guinea pigs. Eur. J. Pharmacol.
35 403:267-275.
36 Neuhaus-Steinmetz, U.; Uffhausen, F.; Herz, U.; Renz, H. (2000) Priming of allergic immune responses by repeated
37 ozone exposure in mice. Am. J. Respir. Cell Mol. Biol. 23: 228-233.
38 Nichols, B. G.; Woods, J. S.; Luchtel, D. L.; Corral, J.; Koenig, J. Q. (2001) Effects of ozone exposure on nuclear
39 factor-KB activation and tumor necrosis factor-a expression in human nasal epithelial cells. Toxicol. Sci. 60:
40 356-362.
41 Nielsen, G. D.; Hougaard, K. S.; Larsen, S. T.; Hammer, M.; Wolkoff, P.; Clausen, P. A.; Wilkins, C. K.; Alarie, Y.
42 (1999) Acute airway effects of formaldehyde and ozone inBALB/c mice. Hum. Exp. Toxicol. 18: 400-409.
43 Nino-Cabrera, H. G.; Colin-Barenque, L.; Avila-Costa, M. R.; Espinosa-Villanueva, J.; Fortoul, T. I.;
44 Rivas-Arancibia, S. (2002) Differences between hippocampus and cerebral cortex in aged rats in an oxidative
45 stress model. Int. J. Neurosci. 112: 373-381.
46 Noviski, N.; Brewer, J. P.; Skornik, W. A.; Galli, S. J.; Drazen, J. M.; Martin, T. R. (1999) Mast cell activation is not
47 required for induction of airway hyperresponsiveness by ozone in mice. J. Appl. Physiol. 86: 202-210.
48 O'Connor, G. T.; Sparrow, D.; Weiss, S. T. (1995) A prospective longitudinal study of methacholine airway
49 responsiveness as a predictor of pulmonary-function decline: the Normative Aging Study. Am. J. Respir. Crit.
50 Care Med. 152: 87-92.
51 Costing, R. S.; Van Iwaarden, J. F.; Van Bree, L.; Verhoef, J.; Van Golde, L. M. G.; Haagsman, H. P. (1992)
52 Exposure of surfactant protein A to ozone in vitro and in vivo impairs its interactions with alveolar cells. Am.
53 J. Physiol. 262: L63-L68.
54 Paige, R. C.; Royce, F. H.; Plopper, C. G.; Buckpitt, A. R. (2000a) Long-term exposure to ozone increases acute
55 pulmonary centriacinar injury by 1-nitronaphthalene: I. Region-specific enzyme activity. J. Pharmacol. Exp.
56 Ther. 295: 934-941.
December 2005 AX5-106 DRAFT-DO NOT QUOTE OR CITE
-------
1 Paige, R. C.; Wong, V.; Plopper, C. G. (2000b) Long-term exposure to ozone increases acute pulmonary centriacinar
2 injury by l-nitronaphthalene: II. Quantitative histopathology. J. Pharmacol. Exp. Ther. 295: 942-950.
3 Palmer, L. J.; Burton, P. R.; Faux, J. A. (2000) Independent inheritance of serum immunoglobulin E concentrations
4 and airway responsiveness. Am. J. Respir. Crit. Care Med. 161: 1836-1842.
5 Palmer, L. J.; Rye, P. J.; Gibson, N. A. (2001) Airway responsiveness in early infancy predicts asthma, lung function,
6 and respiratory symptoms by school age. Am. J. Respir. Crit. Care Med. 163: 37-42.
7 Paquette, N. C.; Tankersley, C. G.; Zhang, L.-Y.; Kleeberger, S. R. (1994) Repeated subacute ozone exposure of
8 inbred mice: airway inflammation and ventilation. Exp. Lung Res. 20: 579-594.
9 Paz, C. (1997) Some consequences of ozone exposure on health. Arch. Med. Res. 28: 163-170.
10 Paz, C.; Bazan-Perkins, B. (1992) Sleep-wake disorganization in cats exposed to ozone. Neurosci. Lett. 140: 270-272.
11 Paz, C.; Huitron-Resendiz, S. (1996) The effects of ozone exposure on the sleep-wake cycle and serotonin contents in
12 the pons of the rat. Neurosci. Lett. 204: 49-52.
13 Pearson, A. C.; Bhalla, D. K. (1997) Effects of ozone onmacrophage adhesion in vitro and epithelial and
14 inflammatory responses in vivo: the role of cytokines. J. Toxicol. Environ. Health 50: 143-157.
15 Peden, D. B.; Dailey, L. (1995) Modulation of mast cell functions by in vitro ozone exposure. Am. J. Physiol.
16 268: L902-L910.
17 Pendino, K. J.; Shuler, R. L.; Laskin, J. D.; Laskin, D. L. (1994) Enhanced production of interleukin-1, tumor
18 necrosis factor-a, and fibronectin by rat lung phagocytes following inhalation of a pulmonary irritant. Am. J.
19 Respir. Cell Mol. Biol. 11: 279-286.
20 Pendino, K. J.; Meidhof, T. M.; Heck, D. E.; Laskin, J. D.; Laskin, D. L. (1995) Inhibition of macrophages with
21 gadolinium chloride abrogates ozone-induced pulmonary injury and inflammatory mediator production. Am. J.
22 Respir. Cell Mol. Biol. 13: 125-132.
23 Pendino, K. J.; Gardner, C. R.; Shuler, R. L.; Laskin, J. D.; Durham, S. K.; Barton, D. S.; Ohnishi, S. T.; Ohnishi, T.;
24 Laskin, D. L. (1996) Inhibition of ozone-induced nitric oxide synthase expression in the lung by endotoxin.
25 Am. J. Respir. Cell Mol. Biol. 14: 516-525.
26 Petruzzi, S.; Fiore, M.; DeH'Omo, G.; Bignami, G.; Alleva, E. (1995) Medium and long-term behavioral effects in
27 mice of extended gestational exposure to ozone. Neurotoxicol. Teratol. 17: 463-470.
28 Petruzzi, S.; De Acetis, L.; Chiarotti, F.; Sorace, A.; Alleva, E. (1999) Limited changes in handedness and morphine
29 reactivity in CD-I mice after pre- and postnatal ozone exposure. Acta Neurobiol. Exp. 59: 115-122.
30 Pinkerton, K. E.; Menache, M. G.; Plopper, C. G. (1995) Consequences of prolonged inhalation of ozone on F344/N
31 rats: collaborative studies. Part IX. Changes in the tracheobronchial epithelium, pulmonary acinus, and lung
32 antioxidant enzyme activity. Cambridge, MA: Health Effects Institute; pp. 41-98; research report no. 65.
33 Available from: NTIS, Springfield, VA; PB95-261996.
34 Pinkerton, K. E.; Weller, B. L.; Menache, M. G.; Plopper, C. G. (1998) Consequences of prolonged inhalation of
3 5 ozone on F344/N rats: collaborative studies. Part XIII. A comparison of changes in the tracheobronchial
36 epithelium and pulmonary acinus in male rats at 3 and 20 months. Cambridge, MA: Health Effects Institute;
37 research report no. 65.
38 Plopper, C. G.; Fanucchi, M. V. (2000) Do urban environmental pollutants exacerbate childhood lung diseases?
39 Environ. HealthPerspect. 108: A252-A253.
40 Plopper, C. G.; Chu, F.-P.; Haselton, C. J.; Peake, J.; Wu, J.; Pinkerton, K. E. (1994) Dose-dependent tolerance to
41 ozone: I. tracheobronchial epithelial reorganization in rats after 20 months' exposure. Am. J. Pathol.
42 144:404-420.
43 Plopper, C. G.; Hatch, G. E.; Wong, V.; Duan, X.; Weir, A. J.; Tarkington, B. K.; Devlin, R. B.; Becker, S.; Buckpitt,
44 A. R. (1998) Relationship of inhaled ozone concentration to acute tracheobronchial epithelial injury,
45 site-specific ozone dose and glutathione depletion in rhesus monkeys. Am. J. Respir. Cell Mol. Biol.
46 19:387-399.
47 Polosa, R.; Holgate, S. T. (1997) Adenosine bronchoprovocation: a promising marker of allergic inflammation in
48 asthma? Thorax 52: 919-923.
49 Postlethwait, E. M.; Cueto, R.; Velsor, L. W.; Pryor, W. A. (1998) Cyinduced formation of bioactive lipids:
50 estimated surface concentrations and lining layer effects. Am. J. Physiol. 274: L1006-L1016.
51 Postlethwait, E. M.; Joad, J. P.; Hyde, D. M.; Schelegle, E. S.; Brie, J. M.; Weir, A. J.; Putney, L. F.; Wong, V. J.;
52 Velsor, L. W.; Plopper, C. G. (2000) Three-dimensional mapping of ozone-induced acute cytotoxicity in
53 tracheobronchial airways of isolated perfused rat lung. Am. J. Respir. Cell Mol. Biol. 22: 191-199.
54 Prows, D. R.; Shertzer, H. G.; Daly, M. J.; Sidman, C. L.; Leikauf, G. D. (1997) Genetic analysis of ozone-induced
55 acute lung injury in sensitive and resistant strains of mice. Nat. Genet. 17: 471-474.
December 2005 AX5-107 DRAFT-DO NOT QUOTE OR CITE
-------
1 Prows, D. R.; Daly, M. I; Shertzer, H. G.; Leikauf, G. D. (1999) Ozone-induced acute lung injury: genetic analysis of
2 F2 mice generated from A/J and C57BL/6J strains. Am. J. Physiol. 277: L372-L380.
3 Pryor, W. A.; Squadrito, G. L.; Friedman, M. (1995) A new mechanism for the toxicity of ozone. Toxicol. Lett.
4 82/83:287-293.
5 Pryor, W. A.; Bermudez, E.; Cueto, R.; Squadrito, G. L. (1996) Detection of aldehydes in bronchoalveolar lavage of
6 rats exposed to ozone. Fundam. Appl. Toxicol. 34: 148-156.
7 Rehle, D.; Leleux, D.; Erdelyi, M.; Tittel, F.; Fraser, M.; Friedfeld, S.; et al. (2001) Ambient formaldehyde detection
8 with a laser spectrometer based on difference-frequency generation in PPLN. Appl. Phys. B: Lasers Opt.
9 72: 947-952.
10 Reinhart, P. G.; Gupta, S. K.; Bhalla, D. K. (1999) Attenuation of ozone-induced lung injury by interleukin-10.
11 Toxicol. Lett. 110:35-42.
12 Rivas-Arancibia, S.; Vazquez-Sandoval, R.; Gonzalez-Kladiano, D.; Schneider-Rivas, S.; Lechuga-Guerrero, A.
13 (1998) Effects of ozone exposure in rats on memory and levels of brain and pulmonary superoxide dismutase.
14 Environ. Res. 76: 33-39.
15 Rivas-Arancibia, S.; Dorado-Martinez, C.; Borgonio-Perez, G.; Hiriart-Urdanivia, M.; Verdugo-Diaz, L.;
16 Duran-Vazquez, A.; Colin-Baranque, L.; Avila-Costa, M. R. (2000) Effects of taurine on ozone-induced
17 memory deficits and lipid peroxidation levels in brains of young, mature, and old rats. Environ. Res. 82: 7-17.
18 Rivas-Arancibia, S.; Dorado-Martinez, C.; Colin-Barenque, L.; Kendrick, K. M.; De la Riva, C.; Guevara-Guzman,
19 R. (2003) Effect of acute ozone exposure on locomotor behavior and striatal function. Pharmacol. Biochem.
20 Behav. 74: 891-900.
21 Rivas-Manzano, P.; Paz, C. (1999) Cerebellar morphological alterations in rats induced by prenatal ozone exposure.
22 Neurosci. Lett. 276: 37-40.
23 Rose, R. C.; Richer, S. P.; Bode, A. M. (1998) Ocular oxidants and antioxidant protection. Proc. Soc. Exp. Biol. Med.
24 217:397-407.
25 Savov, J. D.; Whitehead, G. S.; Wang, J.; Liao, G.; Usuka, J.; Peltz, G.; Foster, W. M.; Schwartz, D. A. (2004)
26 Ozone-induced acute pulmonary injury in inbred mouse strains. Am. J. Respir. Cell Mol. Biol. 31: 69-11.
27 Schelegle, E. S.; Alfaro, M. F.; Putney, L.; Stovall, M.; Tyler, N.; Hyde, D. M. (2001) Effect of C-fiber-mediated,
28 ozone-induced rapid shallow breathing on airway epithelial injury in rats. J. Appl. Physiol. 91: 1611-1618.
29 Schelegle, E. S.; Miller, L. A.; Gershwin, L. J.; Fanucchi, M. V.; Van Winkle, L. S.; Gerriets, J. E.; Walby, W. F.;
30 Mitchell, V; Tarkington, B. K.; Wong, V J.; Baker, G. L.; Pantle, L. M.; Joad, J. P.; Pinkerton, K. E.; Wu, R.;
31 Evans, M. J.; Hyde, D. M.; Plopper, C. G. (2003a) Repeated episodes of ozone inhalation amplifies the effects
32 of allergen sensitization and inhalation on airway immune and structural development in Rhesus monkeys.
33 Toxicol. Appl. Pharmacol. 191: 74-85.
34 Schelegle, E. S.; Walby, W. F.; Alfaro, M. F.; Wong, V. J.; Putney, L.; Stovall, M. Y.; Sterner-Kock, A.; Hyde,
35 D. M.; Plopper, C. G. (2003b) Repeated episodes of ozone inhalation attenuates airway injury/repair and
36 release of substance P, but not adaptation. Toxicol. Appl. Pharmacol. 186: 127-142.
3 7 Schlesinger, R. B. (1995) Interaction of gaseous and paniculate pollutants in the respiratory tract: mechanisms and
38 modulators. Toxicology 105: 315-325.
39 Schlesinger, R. B.; Cohen, M. D.; Gordon, T.; Nadziejko, C.; Zelikoff, J. T.; Sisco, M.; Regal, J. F.; Menache, M. G.
40 (2002a) Ozone differentially modulates airway responsiveness in atopic versus nonatopic guinea pigs.
41 Inhalation Toxicol. 14:431-457.
42 Schlesinger, R. B.; Cohen, M.; Gordon, T.; Nadziejko, C.; Zelikoff, J. T.; Sisco, M.; Regal, J. F.; Menache, M. G.
43 (2002b) Ozone-induced modulation of airway hyperresponsiveness in guinea pigs. Boston, MA: Health Effects
44 Institute; research report no. 109.
45 Schwartz, D. A. (2002) TLR4 and LPS hyporesponsiveness in humans. Int. J. Hyg. Environ. Health 205: 221-227.
46 Segura, P.; Montano, L. M.; Bazan-Perkins, B.; Gustin, P.; Vargas, M. H. (1997) Ozone at high-pollution urban levels
47 causes airway hyperresponsiveness to substance P but not to other agonists. Environ. Toxicol. Pharmacol. 3:
48 91-95.
49 Sen, S.; Dulchavsky, S. A.; Dutta, S. (1993) Effects of triiodothyronine (T3) supplementation upon ozone-induced
50 lung injury. Free Radic. Res. Commun. 18: 299-308.
51 Shore, S. A.; Abraham, J. H.; Schwartzman, I. N.; Murthy, G. G.; Laporte, J. D. (2000) Ventilatory responses to
52 ozone are reduced in immature rats. J. Appl. Physiol. 88: 2023-2030.
53 Shore, S. A.; Schwartzman, I. N.; Le Blanc, B.; Krishna Murthy, G. G.; Doerschuk, C. M. (2001) Tumor necrosis
54 factor receptor 2 contributes to ozone-induced airway hyperresponsiveness in mice. Am. J. Respir. Crit. Care
55 Med. 164: 602-607.
December 2005 AX5-108 DRAFT-DO NOT QUOTE OR CITE
-------
1 Shore, S. A.; Johnston, R. A.; Schwartzman, I. N.; Chism, D.; Krishna Murthy, G. G. (2002) Ozone-induced airway
2 hyperresponsiveness is reduced in immature mice. J. Appl. Physiol. 92: 1019-1028.
3 Shore, S. A.; Rivera-Sanchez, Y. M; Schwartzman, I. N.; Johnston, R. A. (2003) Responses to ozone are increased in
4 obese mice. J. Appl. Physiol. 95: 938-945.
5 Sindhu, R. K.; Mautz, W. J.; Kikkawa, Y. (1998) Chronic exposure to ozone and nitric acid vapor results in increased
6 levels of rat pulmonary putrescine. Arch. Toxicol. 72: 445-449.
7 Slade, R.; Watkinson, W. P.; Hatch, G. E. (1997) Mouse strain differences in ozone dosimetry and body temperature
8 changes. Am. J. Physiol. 272: L73-L77.
9 Sommer, B.; Montano, L. M.; Chavez, J.; Gustin, P.; Vargas, M. H. (1998) Guinea pig lung resistance shows
10 circadian rhythmicity not influenced by ozone. Respir. Physiol. 113: 223-229.
11 Sommer, B.; Vargas, M. H.; Chavez, J.; Carbajal, V; Segura, P.; Montano, L. M. (2001) Differences between inhaled
12 and intravenous bronchial challenge to detect O3-induced hyperresponsiveness. J. Appl. Physiol.
13 91:2595-2601.
14 Sorace, A.; De Acetis, L.; Alleva, E.; Santucci, D. (2001) Prolonged exposure to low doses of ozone: short- and
15 long-term changes to behavioral performance in mice. Environ. Res. 85: 122-134.
16 Spannhake, E. W. (1996) Down-regulation of canine airway mast cell function following exposure to ozone in vivo.
17 Exp. Lung Res. 22: 163-178.
18 Steerenberg, P. A.; Garssen, J.; van Bree, L.; van Loveren, H. (1996) Ozone alters T-helper cell mediated bronchial
19 hyperreactivity and resistance to bacterial infection. Exp. Toxicol. Pathol. 48: 497-499.
20 Sterner-Kock, A.; Kock, M.; Braun, R.; Hyde, D. M. (2000) Ozone-induced epithelial injury in the ferret is similar to
21 nonhuman primates. Am. J. Respir. Crit. Care Med. 162: 1152-1156.
22 Stevens, W. H. M.; Adelroth, E.; Wattie, J.; Woolley, M. J.; Ellis, R.; Dahlback, M.; O'Byrne, P. M. (1994) Effect of
23 inhaled budesonide on ozone-induced airway hyperresponsiveness and bronchoalveolar lavage cells in dogs. J.
24 Appl. Physiol. 77: 2578-2583.
25 Stevens, W. H. M.; VanderHeyden, C; Wattie, J.; Lane, C. G.; Smith, W.; O'Byrne, P. M. (1995a) Effect of a
26 leukotriene B4 receptor antagonist SC-53228 on ozone-induced airway hyperresponsiveness and inflammation
27 in dogs. Am. J. Respir. Crit. Care Med. 152: 1443-1448.
28 Stevens, W. H. M.; Conlon, P. D.; O'Byrne, P. M. (1995b) Ozone-induced oxygen radical release from
29 bronchoalveolar lavage cells and airway hyper-responsiveness in dogs. J. Physiol. (London) 486: 257-265.
30 Stick, S. M. (2002) Pulmonary physiology, airway responsiveness and asthma. Med. J. Aust. 177: S55-S56.
31 Stick, S. M.; Turnbull, S.; Chua, H. L.; Landau, L. L; LeSouef, P. N. (1990) Bronchial responsiveness to histamine in
32 infants and older children. Am. Rev. Respir. Dis. 142: 1143-1146.
33 Stockstill, B. L.; Chang, L.-Y.; Menache, M. G.; Mellick, P. W.; Mercer, R. R.; Crapo, J. D. (1995) Bronchiolarized
34 metaplasia and interstitial fibrosis in rat lungs chronically exposed to high ambient levels of ozone. Toxicol.
35 Appl. Pharmacol. 134: 251-263.
36 Sun, J.; Chung, K. F. (1997) Interaction of ozone exposure with airway hyperresponsiveness and inflammation
3 7 induced by trimellitic anhydride in sensitized guinea pigs. J. Toxicol. Environ. Health 51: 77-87.
3 8 Sun, J.; Koto, H.; Chung, K. F. (1997) Interaction of ozone and allergen challenges on bronchial responsiveness and
39 inflammation in sensitised guinea pigs. Int. Arch. Allergy Immunol. 112: 191-195.
40 Szarek, J. L.; Stewart, N. L.; Zhang, J. Z.; Webb, J. A.; Valentovic, M. A.; Catalano, P. (1995) Contractile responses
41 and structure of small bronchi isolated from rats after 20 months' exposure to ozone. Fundam. Appl. Toxicol.
42 28: 199-208.
43 Takahashi, T.; Miura, M.; Katsumata, U.; Ichinose, M.; Kimura, K.; Inoue, H.; Takishima, T.; Shirato, K. (1993)
44 Involvement of superoxide in ozone-induced airway hyperresponsiveness in anesthetized cats. Am. Rev.
45 Respir. Dis. 148: 103-106.
46 Takahashi, M.; Kleeberger, S. R.; Croxton, T. L. (1995) Genetic control of susceptibility to ozone-induced changes in
47 mouse trachea! electrophysiology. Am. J. Physiol. 269: L6-L10.
48 Takata, S.; Aizawa, H.; Inoue, H.; Koto, H.; Kara, N. (1995) Ozone exposure suppresses epithelium-dependent
49 relaxation in feline airway. Lung 173: 47-56.
50 Takebayashi, T.; Abraham, J.; Murthy, G. G. K.; Lilly, C.; Rodger, I.; Shore, S. A. (1998) Role of tachykinins in
51 airway responses to ozone in rats. J. Appl. Physiol. 85: 442-450.
52 Tankersley, C. G.; Kleeberger, S. R. (1994) Ozone-induced inflammation and altered ventilation in genetically
53 susceptible mice: a comparison of acute and subacute exposures. Toxicol Lett. 72: 279-289.
54 Tankersley, C. G.; Fitzgerald, R. S.; Mitzner, W. A.; Kleeberger, S. R. (1993) Hypercapnic ventilatory responses in
55 mice differentially susceptible to acute ozone exposure. J. Appl. Physiol. 75: 2613-2619.
December 2005 AX5-109 DRAFT-DO NOT QUOTE OR CITE
-------
1 Tesfaigzi, I; Hotchkiss, J. A.; Harkema, J. R. (1998) Expression of the Bcl-2 protein in nasal epithelia of F344/N rats
2 during mucous cell metaplasia and remodeling. Am. J. Respir. Cell Mol. Biol. 18: 794-799.
3 Thiele, J. J. (2001) Oxidative targets in the stratum corneum. A new basis for antioxidative strategies. Skin
4 Pharmacol. Appl. SkinPhysiol. 14(suppl. 1): 87-91.
5 Thiele, J. J.; Traber, M. G.; Podda, M; Tsang, K.; Cross, C. E.; Packer, L. (1997a) Ozone depletes tocopherols and
6 tocotrienols topically applied to murine skin. FEES Lett. 401: 167-170.
7 Thiele, J. J.; Podda, M.; Packer, L. (1997b) Tropospheric ozone: an emerging environmental stress to skin. Biol.
8 Chem. 378: 1299-1305.
9 Thiele, J. J.; Traber, M. G.; Polefka, T. G.; Cross, C. E.; Packer, L. (1997c) Ozone-exposure depletes vitamin E and
10 induces lipid peroxidation in murine stratum corneum. J. Invest. Dermatol. 108: 753-757.
11 Thiele, J. J.; Traber, M. G.; Tsang, K.; Cross, C. E.; Packer, L. (1997d) In vivo exposure to ozone depletes vitamins C
12 and E and induces lipid peroxidation in epidermal layers of murine skin. Free Radical Biol. Med. 23: 385-391.
13 Tsai, J.-J.; Lin, Y.-C.; Kwan, Z.-H.; Kao, H.-L. (1998) Effects of ozone on ovalbumin sensitization in guinea pigs.
14 J. Microbiol. Immunol. Infect. 31: 225-232.
15 Tsukagoshi, H.; Haddad, E.-B.; Sun, J.; Barnes, P. J.; Chung, K. F. (1995) Ozone-induced airway
16 hyperresponsiveness: role of superoxide anions, NEP, and BK receptors. J. Appl. Physiol. 78: 1015-1022.
17 Tyler, W. S.; Tyler, N. K.; Last, J. A.; Gillespie, M. J.; Barstow, T. J. (1988) Comparison of daily and seasonal
18 exposures of young monkeys to ozone. Toxicology 50: 131-144.
19 Tyler, W. S.; Tyler, N. K.; Magliano, D. J.; Hinds, D. M.; Tarkington, B.; Julian, M. D.; Hyde, D. M.; Plopper, C. G.;
20 Dungworth, D. L. (1991) Effects of ozone inhalation on lungs of juvenile monkeys. Morphometry after a 12
21 month exposure and following a 6 month post-exposure period. In: Berglund, R. L.; Lawson, D. R.; McKee, D.
22 J., eds. Tropospheric ozone and the environment: papers from an international conference; March 1990; Los
23 Angeles, CA. Pittsburgh, PA: Air & Waste Management Association; pp. 151-160. (A&WMA transactions
24 series no. TR-19).
25 U.S. Environmental Protection Agency. (1986) Air quality criteria for ozone and other photochemical oxidants.
26 Research Triangle Park, NC: Office of Health and Environmental Assessment, Environmental Criteria and
27 Assessment Office; report nos. EPA-600/8-84-020aF-eF. 5v. Available from: NTIS, Springfield, VA;
28 PB87-142949.
29 U.S. Environmental Protection Agency. (1996) Air quality criteria for ozone and related photochemical oxidants.
3 0 Research Triangle Park, NC: Office of Research and Development; report nos. EPA/600/AP-93/004aF-cF. 3v.
31 Available from: NTIS, Springfield, VA; PB96-185582, PB96-185590, and PB96-185608. Available online at:
32 www.epa.gov/ncea/ozone.htm.
33 Uhlson, C.; Harrison, K.; Allen, C. B.; Ahmad, S.; White, C. W.; Murphy, R. C. (2002) Oxidized phospholipids
34 derived from ozone-treated lung surfactant extract reduce macrophage and epithelial cell viability. Chem. Res.
35 Toxicol. 15: 896-906.
36 Ulrich, M. M. W.; Alink, G. M.; Kumarathasan, P.; Vincent, R.; Boere, A. J.; Cassee, F. R. (2002) Health effects and
3 7 time course of paniculate matter on the cardiopulmonary system in rats with lung inflammation. J. Toxicol.
38 Environ. HealthPart A 65: 1571-1595.
39 Valacchi, G.; Bocci, V (2000) Studies on the biological effects of ozone: 11. Release of factors from human
40 endothelial cells. Mediators Inflammation 9: 271-276.
41 Valacchi, G.; Weber, S. U.; Luu, C.; Cross, C. E.; Packer, L. (2000) Ozone potentiates vitamin E depletion by
42 ultraviolet radiation in the murine stratum corneum. FEES Lett. 466: 165-168.
43 Valacchi, G.; Van der Vliet, A.; Schock, B.C.; Okamoto, T.; Obermuller-Jevic, U.; Cross, C. E.; Packer, L. (2002)
44 Ozone exposure activates oxidative stress responses in murine skin. Toxicology 179: 163-170.
45 Valacchi, G.; Pagnin, E.; Okamoto, T.; Corbacho, A. M.; Olano, E.; Davis, P. A.; Van der Vliet, A.; Packer, L.;
46 Cross, C. E. (2003) Induction of stress proteins and MMP-9 by 0.8 ppm of ozone in murine skin. Biochem.
47 Biophys. Res. Commun. 305: 741-746.
48 Van Bree, L.; Dormans, J. A. M. A.; Boere, A. J. F.; Rombout, P. J. A. (2001) Time study on development and repair
49 of lung injury following ozone exposure in rats. Inhalation Toxicol. 13: 703-717.
50 Van Bree, L.; Dormans, J. A. M. A.; Koren, H. S.; Devlin, R. B.; Rombout, P. J. A. (2002) Attenuation and recovery
51 of pulmonary injury in rats following short-term, repeated daily exposure to ozone. Inhalation Toxicol.
52 14:883-900.
53 Van Hoof, I. H. J. M.; Van Bree, L.; Bast, A. (1996) Changes in receptor function by oxidative stress in guinea pig
54 trachea! smooth muscle. Cent. Eur. J. Public Health 4(suppl.): 3-5.
55 Van Hoof, H. J. M.; Van Acker, F. A. A.; Voss, H.-P.; Van Bree, L.; Bast, A. (1997a) Acute exposure to ozone does
56 not influence neuroreceptor density and sensitivity in guinea pig lung. Toxicol. Lett. 90: 53-60.
December 2005 AX5-110 DRAFT-DO NOT QUOTE OR CITE
-------
1 Van Hoof, H. J. M; Voss, H.-P.; Kramer, K.; Boere, A. J. F.; Dormans, J. A. M. A.; Van Bree, L.; Bast, A. (1997b)
2 Changes in neuroreceptor function of tracheal smooth muscle following acute ozone exposure of guinea pigs.
3 Toxicology 120: 159-169.
4 Vanda, B.; de Buen, N.; Jasso, R.; Valero, G.; Vargas, M. H.; Olmos, R.; Arreola, J. L.; Santillan, P.; Alonso, P.
5 (1998) Inflammatory cells and ferruginous bodies inbronchoalveolar lavage in urban dogs. Acta Cytol.
6 42: 939-944.
7 Vargas, M. H.; Segura, P.; Campos, M. G.; Hong, E.; Montano, L. M. (1994) Effect of ozone exposure on
8 antigen-induced airway hyperresponsiveness in guinea pigs. J. Toxicol. Environ. Health 42: 435-442.
9 Vargas, M. H.; Romero, L.; Sommer, B.; Zamudio, P.; Gustin, P.; Montano, L. M. (1998) Chronic exposure to ozone
10 causes tolerance to airway hyperresponsiveness in guiea pigs: lack of SOD role. J. Appl. Physiol.
11 84:1749-1755.
12 Vesely, D. L.; Giordano, A. T.; Raska-Emery, P.; Montgomery, M. R. (1994a) Ozone increases amino- and
13 carboxy-terminal atrial natriuretic factor prohormone peptides in lung, heart, and circulation. J. Biochem.
14 Toxicol. 9: 107-112.
15 Vesely, D. L.; Giordano, A. T.; Raska-Emery, P.; Montgomery, M. R. (1994b) Increase in atrial natriuretic factor in
16 the lungs, heart, and circulatory system owing to ozone. Chest 105: 1551-1554.
17 Vesely, D. L.; Giordano, A. T.; Raska-Emery, P.; Montgomery, M. R. (1994c) Ozone increases atrial natriuretic
18 peptides in heart, lung and circulation of aged vs. adult animals. Gerontology (Basel) 40: 227-236.
19 Vesely, K. R.; Schelegle, E. S.; Stovall, M. Y.; Harkema, J. R.; Green, J. F.; Hyde, D. M. (1999a) Breathing pattern
20 response and epithelial labeling in ozone-induced airway injury in neutrophil-depleted rats. Am. J. Respir. Cell
21 Mol. Biol. 20: 699-709.
22 Vesely, K. R.; Hyde, D. M.; Stovall, M. Y.; Harkema, J. R.; Green, J. F.; Schelegle, E. S. (1999b) Capsaicin-sensitive
23 C-fiber-mediated protective responses in ozone inhalation in rats. J. Appl. Physiol. 86: 951-962.
24 Vincent, R.; Janzen, E. G.; Chen, G.; Kumarathasan, P.; Haire, D. L.; Guenette, J.; Chen, J. Z.; Bray, T. M. (1996)
25 Spin trapping study in the lungs and liver of F344 rats after exposure to ozone. Free Radical Res. 25: 475-488.
26 Vincent, R.; Bjarnason, S. G.; Adamson, I. Y. R.; Hedgecock, C.; Kumarathasan, P.; Guenette, J.; Potvin, M.;
27 Goegan, P.; Bouthillier, L. (1997) Acute pulmonary toxicity of urban paniculate matter and ozone. Am. J.
28 Pathol. 151: 1563-1570.
29 Vyskocil, A.; Viau, C.; Lamy, S. (1998) Peroxyacetyl nitrate: review of toxicity. Hum. Exp. Toxicol. 17: 212-220.
30 Wagner, J. G.; Hotchkiss, J. A.; Harkema, J. R. (200la) Effects of ozone and endotoxin coexposure on rat airway
31 epithelium: potentiation of toxicant-induced alterations. Environ. Health Perspect. 109(suppl. 4): 591-598.
32 Wagner, J. G.; VanDyken, S. J.; Hotchkiss, J. A.; Harkema, J. R. (200Ib) Endotoxin enhancement of ozone-induced
33 mucous cell metaplasia is neutrophil-dependent in rat nasal epithelium. Toxicol. Sci. 60: 338-347.
34 Wagner, J. G.; Hotchkiss, J. A.; Harkema, J. R. (2002) Enhancement of nasal inflammatory and epithelial responses
3 5 after ozone and allergen coexposure in brown Norway rats. Toxicol. Sci. 67: 284-294.
36 Wagner, J. G.; VanDyken, S. J.; Wierenga, J. R.; Hotchkiss, J. A.; Harkema, J. R. (2003) Ozone exposure enhances
37 endotoxin-induced mucous cell metaplasia in rat pulmonary airways. Toxicol. Sci. 74: 437-446.
38 Wang, G.; Umstead, T. M.; Phelps, D. S.; Al-Mondhiry, H.; Floras, J. (2002) The effect of ozone exposure on the
3 9 ability of human surfactant protein A variants to stimulate cytokine production. Environ. Health Perspect. 110:
40 79-84.
41 Watkinson, W. P.; Wiester, M. J.; Highfill, J. W. (1995) Ozone toxicity in the rat. I. Effect of changes in ambient
42 temperature on extrapulmonary physiological parameters. J. Appl. Physiol. 78: 1108-1120.
43 Watkinson, W. P.; Campen, M. J.; Nolan, J. P.; Costa, D. L. (2001) Cardiovascular and systemic responses to inhaled
44 pollutants in rodents: effects of ozone and paniculate matter. Environ. Health Perspect. 109(suppl. 4): 539-546.
45 Watkinson, W. P.; Campen, M. J.; Wichers, L. B.; Nolan, J. P.; Costa, D. L. (2003) Cardiac and thermoregulatory
46 responses to inhaled pollutants in healthy and compromised rodents: modulation via interaction with
47 environmental factors. Environ. Res. 92: 35-47.
48 Watt, K. C.; Plopper, C. G.; Weir, A. J.; Tarkington, B.; Buckpitt, A. R. (1998) Cytochrome P450 2E1 in rat
49 tracheobronchial airways: response to ozone exposure. Toxicol. Appl. Pharmacol. 149: 195-202.
50 Wattiez, R.; Noel-Georis, L; Cruyt, C.; Broeckaert, F.; Bernard, A.; Falmagne, P. (2003) Susceptibility to oxidative
51 stress: proteomic analysis of bronchoalveolar lavage from ozone-sensitive and ozone-resistant strains of mice.
52 Proteomics 3: 658-665.
53 Weber, S. U.; Thiele, J. J.; Cross, C. E.; Packer, L. (1999) Vitamin C, uric acid, and glutathione gradients in murine
54 stratum corneum and their susceptibility to ozone exposure. J. Invest. Dermatol. 113:1128-1132.
55 Weber, S. U.; Jothi, S.; Thiele, J. J. (2000) High-pressure liquid chromatography analysis of ozone-induced depletion
56 of hydrophilic and lipophilic antioxidants in murine skin. Methods Enzymol. 319: 536-546.
December 2005 AX5-111 DRAFT-DO NOT QUOTE OR CITE
-------
1 Weber, S. U.; Han, N.; Packer, L. (2001) Ozone: an emerging oxidative stressor to skin. Curr. Probl. in Dermatol. 29:
2 52-61.
3 Weller, B. L.; Crapo, J. D.; Slot, I; Posthuma, G.; Plopper, C. G.; Pinkerton, K. E. (1997) Site- and cell-specific
4 alteration of lung copper/zinc and manganese superoxide dismutases by chronic ozone exposure. Am. J.
5 Respir. Cell Mol. Biol. 17: 552-560.
6 Weller, B. L.; Witschi, H.; Pinkerton, K. E. (2000) Quantitation and localization of pulmonary manganese superoxide
7 dismutase and tumor necrosis factor a following exposure to ozone and nitrogen dioxide. Toxicol. Sci. 54:
8 452-461.
9 Wells, C. A.; Ravasi, T.; Faulkner, G. J.; Carninci, P.; Okazaki, Y.; Hayashizaki, Y.; Sweet, M; Wainwright, B. J.;
10 Hume, D. A. (2003) Genetic control of the innate immune response. BMC Immunol. 4: 5. Available:
11 http://www.biomedcentral.eom/1471-2172/4/5 [18 February, 2003]
12 Wiester, M. J.; Watkinson, W. P.; Costa, D. L.; Crissman, K. M.; Richards, J. H.; Winsett, D. W.; Highfill, J. W.
13 (1996) Ozone toxicity in the rat. III. Effect of changes in ambient temperature on pulmonary parameters. J.
14 Appl. Physiol. 81: 1691-1700.
15 Wilkins, C. K.; Clausen, P. A.; Wolkoff, P.; Larsen, S. T.; Hammer, M.; Larsen, K.; Hansen, V.; Nielsen, G. D.
16 (2001) Formation of strong irritants in mixtures of isoprene/ozone and isoprene/ozone/nitrogen dioxide.
17 Environ. Health Perspect. 109: 937-941.
18 Witschi, H.; Espiritu, L; Pinkerton, K. E.; Murphy, K.; Maronpot, R. R. (1999) Ozone carcinogenesis revisited.
19 Toxicol. Sci. 52: 162-167.
20 Wu, Z.-X.; Morton, R. F.; Lee, L.-Y. (1997) Role of tachykinins in ozone-induced airway hyperresponsiveness to
21 cigarette smoke in guinea pigs. J. Appl. Physiol. 83: 958-965.
22 Yamaguchi, S.; Nagai, H.; Tanaka, H.; Tsujimoto, M.; Tsuruoka, N. (1994) Time course study for antigen-induced
23 airway hyperreactivity and the effect of soluble IL-5 receptor. Life Sci. 54(Pharmacol. Letters): PL 471-475.
24 Yamauchi, T.; Shima, M.; Kuwaki, T.; Ando, M.; Ohmichi, M.; Fukuda, Y.; Adachi, M. (2002) Acute effects of
25 ozone exposure on lung function in mice sensitized to ovalbumin. Toxicology (Ireland) 172: 69-78.
26 Yu, M.; Pinkerton, K. E.; Witschi, H. (2002) Short-term exposure to aged and diluted sidestream cigarette smoke
27 enhances ozone-induced lung injury inB6C3Fl mice. Toxicol. Sci. 65: 99-106.
28 Zhang, L.-Y.; Levitt, R. C.; Kleeberger, S. R. (1995) Differential susceptibility to ozone-induced airways
29 hyperreactivity in inbred strains of mice. Exp. Lung Res. 21: 503-518.
30 Zhao, Q.; Simpson, L. G.; Driscoll, K. E.; Leikauf, G. D. (1998) Chemokine regulation of ozone-induced neutrophil
31 and monocyte inflammation. Am. J. Physiol. 274: L39-L46.
32
December 2005 AX5-112 DRAFT-DO NOT QUOTE OR CITE
-------
i 6. CONTROLLED HUMAN EXPOSURE STUDIES
2 OF OZONE AND RELATED
3 PHOTOCHEMICAL OXIDANTS
4
5
6 6.1 INTRODUCTION
7 In the previous chapter, results of ozone (O3) studies in laboratory animals and in vitro test
8 systems were presented. The extrapolation of results from animal studies is one mechanism by
9 which information on potential adverse human health effects from exposure to O3 is obtained.
10 More direct evidence of human health effects due to O3 exposure can be obtained through
11 controlled human exposure studies of volunteers or through field and epidemiologic studies of
12 populations exposed to ambient O3 (see Chapter 7). Controlled human exposure studies
13 typically use fixed concentrations of O3 under carefully regulated environmental conditions and
14 subject activity levels. This chapter discusses studies in which volunteers were exposed for up
15 to 8 h to between 0.08 to 0.75 ppm O3 while at rest or during varying intensities of exercise.
16 The majority of controlled human studies have investigated the effects of exposure to O3 in
17 young non-smoking healthy adults (18 to 35 years of age) performing continuous exercise (CE)
18 or intermittent exercise (IE). Varied combinations of O3 concentration, exercise routine, and
19 exposure duration have been used in these studies. The responses to ambient O3 concentrations
20 include decreased inspiratory capacity; mild bronchoconstriction; rapid, shallow breathing
21 pattern during exercise; and symptoms of cough and pain on deep inspiration. Reflex inhibition
22 of inspiration results in a decrease in forced vital capacity (F VC) and total lung capacity (TLC)
23 and, in combination with mild bronchoconstriction, contributes to a decrease in the forced
24 expiratory volume in 1 s (FEVj). In addition to physiological pulmonary responses and
25 respiratory symptoms, O3 has been shown to result in airway hyperresponsiveness and
26 inflammation.
27 The most salient observations from studies reviewed in the 1996 EPA Ozone Air Quality
28 Criteria Document or O3 AQCD (U.S. Environmental Protection Agency, 1996) were that:
29 (1) young healthy adults exposed to O3 concentrations > 0.08 ppm develop significant reversible,
30 transient decrements in pulmonary function if minute ventilation (VE ) or duration of exposure
31 are increased sufficiently, (2) children experience similar spirometric responses but lesser
January 2005 6-1 DRAFT-DO NOT QUOTE OR CITE
-------
1 symptoms from O3 exposure relative to young adults, (3) O3-induced spirometric responses are
2 decreased in the elderly relative to young adults, (4) there is a large degree of intersubject
3 variability in physiologic and symptomatic responses to O3 but responses tend to be reproducible
4 within a given individual over a period of several months, and (5) subjects exposed repeated to
5 O3 over several days develop a tolerance to successive exposures, as demonstrated by an
6 attenuation of responses, which is lost after about a week without exposure.
7 There are several important limitations associated with these clinical studies: (1) the
8 ability to study only short-term, acute effects; (2) difficulties in trying to link short-term effects
9 with long-term consequences; (3) the use of a small number of volunteers that may not be
10 representative of the general population; and (4) the statistical limitations associated with the
11 small sample size. Sample size affects the power of a study, and having a small number of
12 samples causes a risk of Type II error, i.e., the incorrect conclusion that no difference exists
13 between treatments or groups when comparisons are not significantly different. This affects the
14 confidence in estimates of a minimum O3 concentration at which some degree of pulmonary
15 impairment will occur in both the general population and susceptible subpopulations. As a
16 result, the conclusions drawn from many of the studies cited in this chapter may underestimate
17 the presence of responses at low O3 concentrations and low activity levels.
18 Most of the scientific information summarized in this chapter comes from the literature
19 published since the 1996 O3 AQCD (U.S. Environmental Protection Agency, 1996). In addition
20 to further study of physiological pulmonary responses and respiratory symptoms, much of this
21 literature has focused on mechanisms of inflammation and cellular responses to injury induced
22 by O3 inhalation. A more thorough discussion and review of this literature appears in Annex
23 AX6 of this document. In summarizing the literature, effects are described if they are
24 statistically significant at a probability (p-value) of less than 0.05, otherwise trends are noted
25 as such.
26 As spirometry typically improves in healthy young adults with exercise exposures to filter
27 air (FA), the term "O3-induced" is used herein and in the annex to designate effects that have
28 been corrected for responses during FA exposures. For healthy adults, an O3-induced change in
29 lung function is the difference between the decrement experienced with O3 exposure and the
30 improvement observed with FA exposure. However, the distinction between an O3-induced
31 change and a post- versus preexposure change is particularly important in individuals with
January 2005 6-2 DRAFT-DO NOT QUOTE OR CITE
-------
1 respiratory disease who may experience exercise-induced decrements in pulmonary function
2 during both FA and O3 exposures. Hence, in subjects with respiratory disease, exercise-induced
3 responses could be mistaken for O3-induced responses in the absence of a correction for FA
4 responses.
5
6
7 6.2 PULMONARY FUNCTION EFFECTS OF OZONE EXPOSURE
8 IN HEALTHY SUBJECTS
9 6.2.1 Introduction
10 As reviewed in the 1986 and 1996 O3 AQCD's (U.S. Environmental Protection Agency,
11 1986, 1996), 0.5 ppm is the lowest O3 concentration at which statistically significant reductions
12 in FVC and FEVj have been reported in sedentary subjects. On average, young adults (n = 23;
13 mean age, 22 yrs) exposed at rest for 2 h to 0.5 ppm O3 had O3-induced decrements of-4% in
14 FVC and -7% in FEVj (Folinsbee et al., 1978; Horvath et al., 1979). During exercise,
15 spirometric and symptoms responses are observed at lower O3 concentrations. For acute
16 exposures of 2 h or less to ^0.12 ppm O3, if VE is sufficiently increased by exercise, healthy
17 human subjects generally experience decreases in TLC, inspiratory capacity (1C), FVC, FEVb
18 mean forced expiratory flow from 25% to 75% of FVC (FEF25.75), and tidal volume (VT) and
19 increases in specific airways resistance (sRaw), breathing frequency (fB), and airway
20 responsiveness. These exposure also cause symptoms of cough, pain on deep inspiration,
21 shortness of breath, throat irritation, and wheezing. With exposures of 4- to 8-h in duration,
22 statistically significant pulmonary function and symptoms responses are observed at lower
23 O3 concentrations and lower VE than in shorter duration studies.
24
25 6.2.2 Acute Exposure for Up to 2 h
26 With heavy CE (VE = 89 L/min), an O3-induced decrement of 9.7% in FEVj has been
27 reported for healthy young adults (n = 17; age, 24 ±3 yrs) exposed for only 1 h to 0.12 ppm O3
28 (Gong et al., 1986). With moderate-to-heavy IE (15 min intervals of rest and exercise
29 [VE = 68 L/min]), McDonnell et al. (1983) reported a physiologically small, but significant,
30 O3-induced decrement of 3.4% in FEVj for young healthy adults (n = 22, age, 22 ±3 yrs)
January 2005 6-3 DRAFT-DO NOT QUOTE OR CITE
-------
1 exposed for 2 h to 0.12 ppm O3. Using the same 2 h exposure protocol, Linn et al. (1986) found
2 no statistically significant spirometic responses at O3 concentrations of 0.16 ppm and lower.
3 However, the subjects in the Linn et al. (1986) study were potentially exposed concurrently in
4 Los Angeles to ambient O3 levels of between 0.12 and 0.16 ppm and were on average 3 yrs older
5 than the subjects in the McDonnell et al. (1983) study. (The attenuating effects of increasing age
6 and repeated O3 exposures are discussed in Sections 6.5.1 and 6.6, respectively.) The disparities
7 between the Linn et al. (1986) and McDonnell et al. (1983) studies demonstrate the difficulty in
8 determining a no-effect-level for O3 based on relatively small study populations.
9 Studies analyzing large data sets (>300 subjects) provide better predictive ability of acute
10 changes in FEVj at low levels of O3 and VE than possible via comparisons between smaller
11 studies. Such an analysis was performed by McDonnell et al. (1997), who examined FEVj
12 responses in 485 healthy white males (18 to 36 years of age; subjects recruited from the area
13 around Chapel Hill, NC) exposed once for 2 h to O3 concentrations of up to 0.40 ppm at rest or
14 with IE. Decrements in FEVj were modeled by sigmoid-shaped curve as a function of subject
15 age, O3 concentration, VE, and duration of exposure. Figure 6-1 illustrates the predicted
16 O3-induced decrements in FEVj for young healthy adults (20 yrs of age) exposed for up to 2 h
17 to O3 during moderate IE (VE = 30 L/min). The responses illustrated for 0.1 ppm in Figure 6-1
18 are approximately the same as responses predicted for an exposure to 0.3 ppm at rest. Although
19 not illustrated in the figure, the predicted FEVj decrements increase with VE . Regarding
20 applicability to the general population, the McDonnell et al. (1997) model has an apparent
21 limitation of considering only data for white males. However, two other large studies (n = 372;
22 18 to 35 yrs of age; subjects recruited from the area around Chapel Hill, NC) found no
23 significant gender nor race effects on spirometric responses to O3 exposure (Seal et al., 1993,
24 1996).
25 Ultman et al. (2004) recently reported pulmonary responses in 60 young heathy non-
26 smoking adults (32 M, 28 F) exposed to 0.25 ppm O3 for 1 h with CE at a target VE of 30 L/min.
27 Consistent with findings reported in the 1996 O3 criteria document, considerable intersubject
28 variability in FEVj decrements was reported by Ultman et al. (2004) with responses ranging
29 from -4 to 56%. One-third of the subjects had FEVj decrements of > 15% and 7% of the
30 subjects had decrements of > 40%. It should be pointed out that the McDonnell et al. (1997)
January 2005 6-4 DRAFT-DO NOT QUOTE OR CITE
-------
Time (h)
Figure 6-1. Predicted O3-induced decrements in FEVt as a function of exposure duration
and O3 concentration in young healthy adults (20 yrs of age) during moderate
•
IE (VE = 30 L/min). Predictions are for Model 1 coefficients in Table 3 of
McDonnell et al. (1997).
1 model predicts only average responses. In a more recent study, McDonnell et al. (1999) also
2 reported a model predicting average symptom responses from O3 exposure. Unfortunately,
3 neither of these papers (McDonnell et al., 1997, 1999) provide predictions of intersubject
4 variability in response. (Section 6.4 of this Chapter discusses intersubject variability in response
5 to O3 exposure).
6
7 6.2.3 Prolonged Ozone Exposures
8 In the exposure range of 0.08 to 0.16 ppm O3, a number of studies using moderate quasi
9 continuous exercise (QCE; 50 min exercise and 10 min rest per h) for 4 to 8 h have shown
10 significant responses under the following conditions: 0.16 ppm for 4 h with QCE at
11 VE « 40 L/min (Folinsbee et al., 1994), 0.08 to 0.12 ppm for 6.6 h with QCE at VE « 35 to
12 40 L/min (Adams, 2002; Adams, 2003a; Folinsbee et al., 1988; Horstman et al., 1990), and
13 0.12 ppm for 8 h of IE (30 min per h) at VE « 40 L/min (Hazucha et al., 1992). Symptoms and
14 spirometric responses increased with duration of exposure, O3 concentration, and total VE.
15 Airway resistance is only modestly affected with moderate or even heavy exercise combined
16 with O3 exposure (Folinsbee et al., 1978; McDonnell et al., 1983; Seal et al., 1993).
January 2005
6-5
DRAFT-DO NOT QUOTE OR CITE
-------
1 6.2.3.1 Effect of Exercise Ventilation Rate on FEVt Response to 6.6 h Ozone Exposure
2 It is well established that response to O3 exposure is a function of VE in studies of 2 h or
3 less in duration (See Section AX6.2.2). It is reasonable to expect that response to a prolonged
4 6.6-h O3 exposure is also function of VE, although quantitative analyzes are lacking. Data from
5 five similar prolonged exposure studies are available for evaluation of FEVj responses as a
6 function of exercise VE (Adams, 2000; Adams and Ollison, 1997; Folinsbee et al., 1988, 1994;
7 Horstman et al., 1990). Each of these studies exposed similarly aged subjects (mean ages 22 to
8 25 yrs) to 0.12 ppm O3 for 6.6 h. In total, ten sets of mean FEVj decrements were available for
9 exercise VE ranging from 20 to 43 L/min, although no data were available for VE between 20
10 and 30 L/min (data illustrated in Figure AX6-2). As in 2 h exposure studies, FEVj decrements
11 are a function VE in prolonged 6.6-h exposure studies as demonstrated by a significant
12 correlation between these variables (Pearson, r = 0.95, p < 0.001; Spearman, r = 0.84, p < 0.01).
13
14 6.2.3.2 Exercise Ventilation Rate as a Function of Body/Lung Size on FEVt Response
15 to 6.6 h Ozone Exposure
16 Based on the assumption that the total inhaled O3 dose (product of O3 concentration,
17 exposure duration, and VE) is proportional to the lung size, exercise VE are typically selected to
18 be a multiple of body surface area (BSA) or FVC. Data from several recent studies do not
19 support the contention that VE should be normalized. In an analysis of data from 485 young
20 adults, McDonnell et al. (1997) found that any effect of BSA, height, or baseline FVC on percent
21 decrement in FEVj was small to nonexistent. This is consistent with Messineo and Adams
22 (1990), who compared pulmonary function responses in young adult women having small
23 (n = 14) or large (n = 14) lung sizes (mean FVC of 3.74 and 5.11 L, respectively) and found no
24 significant group difference in FEVj decrements. For 30 subjects exposed to 0.12 ppm O3 for
25 6.6 h, Adams (2000) also reported that FEVj responses were more closely related to VE than to
26 VE normalized to BSA. The O3 dosimetry study of Bush et al. (1996) suggested that
27 normalization of the O3 dose might more appropriately be a function of anatomic dead space.
28 Ozone penetrates deeper into the lungs of individuals with larger conducting airway volumes,
29 however, FEVj responses in subjects exposed for 2 h to 0.25 ppm O3 do not appear to be
30 associated with O3 uptake (Ultman et al., 2004).
January 2005 6-6 DRAFT-DO NOT QUOTE OR CITE
-------
1 6.2.3.3 Comparison of 2 h IE to 6.6 h O3 Exposure Effects on Pulmonary Function
2 Adams (2003b) examined whether prolonged 6.6-h QCE exposure to a relatively low O3
3 concentration (0.08 ppm) and the 2-h IE exposure at a relatively high O3 concentration (0.30
4 ppm) elicited consistent individual subject FEVj responses. Individual subject O3 exposure
5 reproducibility was first examined via a regression plot of the postexposure FEVj response to the
6 6.6-h chamber exposure as a function of postexposure FEVj response to the 2-h IE chamber
7 exposure. The R2 of 0.40, although statistically significant, was substantially less than that
8 observed in a comparison of individual FEVj response to the two 2-h IE exposures by chamber
9 and face mask, respectively (R2 = 0.83). The Spearman rank order correlation for the chamber
10 6.6-h and chamber 2-h exposure comparison was also substantially less (0.49) than that obtained
11 for the two 2-h IE exposures (0.85). The primary reason for the greater variability in the
12 chamber 6.6-h exposure FEVj response as a function of that observed for the two 2-h IE
13 exposures is very likely related to the increased variability in response upon repeated exposure to
14 O3 concentrations lower than 0.18 ppm (R = 0.57, compared to a mean R of 0.82 at higher
15 concentrations) reported by McDonnell et al. (1985a). This rationale is supported by the lower R
16 (0.60) observed by Adams (2003b) for the FEVj responses found in 6.6 h chamber and face
17 mask exposures to 0.08 ppm O3, compared to an R of 0.91 observed for responses found at 0.30
18 ppm O3.
19
20 6.2.4 Triangular Ozone Exposures
21 To further explore the factors that determine responsiveness to O3, Hazucha et al. (1992)
22 designed a protocol to examine the effect of varying, rather than constant, O3 concentrations.
23 Subjects were exposed to an O3 level that increased linearly from 0 to 0.24 ppm for the first 4 h
24 and then decreased linearly from 0.24 to 0 ppm over the second 4 h of the 8 h exposure
25 (triangular concentration profile) and to a constant level exposure of 0.12 ppm O3 for 8 h. While
26 total inhaled O3 doses for the constant and the triangular concentration profile were almost
27 identical, the FEVj response was dissimilar. For the constant 0.12 ppm O3 exposure, FEVj
28 declined -5% by the fifth hour and then remained at that level. With the triangular O3
29 concentration profile, there was minimal FEVj response over the first 3 h followed by a rapid
30 decrease in FEVj (-10.3%) over next 3 h. During the seventh and eighth hours, FEVj improved
January 2005 6-7 DRAFT-DO NOT QUOTE OR CITE
-------
1 to -6.3% despite continued exposure to a lower O3 concentration (0.12 to 0.00 ppm, mean =
2 0.06 ppm).
3 More recently, Adams (2003a) used a less abrupt triangular O3 exposure profile at
4 concentrations assumed to be typical of outdoor ambient conditions (beginning at 0.03 ppm,
5 increasing steadily to 0.15 ppm in the fourth hour and decreasing steadily to 0.05 ppm at 6.6 h
6 (mean = 0.08 ppm). Postexposure values for FEVj and symptoms were not significantly
7 different between the 6.6 h triangular and a square-wave 0.08 ppm O3 exposure. There was no
8 evidence of FEVj response recovery with the triangular exposure as observed by Hazucha et al.
9 (1992). Rather, FEVj responses observed by Adams (2003a) for the triangular exposure seemed
10 to plateau during the last 2 h, i.e., -5.46% at 4.6 h, -6.27% at 5.6 h, and -5.77% at 6.6 h.
11 With square-wave O3 exposures between 0.08 to 0.12 ppm, FEVj decrements may increase
12 with time of exposure (and O3 dose) or reach plateau (Horstman et al., 1990; McDonnell et al.,
13 1991). For the triangular exposures used by Hazucha et al. (1992) and Adams (2003a), maximal
14 FEVj responses occurred 1 h to 2 h after peak O3 concentration and 1 h to 2 h before the
15 maximal O3 dose occurred (at the end of the O3 exposure).
16
17 6.2.5 Mechanisms of Pulmonary Function Responses
18 Inhalation of O3 for several hours while physically active elicits both subjective respiratory
19 tract symptoms and acute pathophysiologic changes. The typical symptomatic response
20 consistently reported in studies is that of tracheobronchial airway irritation. This is accompanied
21 by decrements in lung capacities and volumes, bronchoconstriction, airway hyperresponsiveness,
22 airway inflammation, immune system activation, and epithelial injury. The severity of
23 symptoms and the magnitude of response depend on inhaled dose, O3 sensitivity of an individual
24 and the extent of tolerance resulting from previous exposures. The development of effects is
25 time- dependent during both exposure and recovery periods with considerable overlap of
26 evolving and receding effects. The time sequence, magnitude and the type of responses of this
27 complex series of events, both in terms of development and recovery, indicate that several
28 mechanisms, activated at different time of exposure must contribute to the overall lung function
29 response (U.S. Environmental Protection Agency, 1996).
30 Available information on recovery from O3 exposure indicates that an initial phase of
31 recovery proceeds relatively rapidly, and some 40 to 65% of the acute spirometric and symptom
January 2005 6-8 DRAFT-DO NOT QUOTE OR CITE
-------
1 response appears to occur within about 2 h (Folinsbee and Hazucha, 1989). Following a 2 h
2 exposure to 0.4 ppm O3 with IE, Nightingale et al. (2000) observed a 13.5% decrement in FEVj.
3 By 3 h postexposure, however, only a 2.7% FEVj decrement persisted as illustrated in
4 Figure 6-2. A similar postexposure recovery in FVC was also observed. Gerrity et al. (1993)
5 suggested that for healthy young adults transient increases in mucus clearance (mediated by
6 cholinergic receptors) due to O3 exposure may be coincident to pulmonary function responses,
7 i.e., the transient increases in clearance and decrements in lung function return to baseline values
8 within 2 to 3 h postexposure. However, there is some indication that the spirometric responses,
9 especially at higher O3 concentrations, are not fully recovered within 24 h (Folinsbee and
10 Horvath, 1986; Folinsbee et al., 1998). Small residual lung function effects are almost
11 completely resolved within 24 hours. In hyperresponsive individuals, the recovery takes longer,
12 as much as 48 hours, to return to baseline values. Collectively, these observations suggest that
13 there is a rapid recovery of O3-induced spirometric responses and symptoms, which may occur
14 during resting exposure to O3 (Folinsbee et al., 1977) or as O3 concentration is reduced during
15 exposure (Hazucha et al., 1992), and a slower phase, which may take at least 24 h to complete
16 (Folinsbee and Hazucha, 2000). Repeated exposure studies at higher concentrations typically
17 show that FEVj response to O3 is enhanced on the second of several days of exposure (Table
18 AX6-8). This enhanced response suggests a residual effect of the previous exposure, about 22 h
19 earlier, even though the preexposure spirometry may be the same as on the previous day. The
20 absence of the enhanced response with repeated exposure at lower O3 concentrations may be the
21 result of a more complete recovery or less damage to pulmonary tissues (Folinsbee et al., 1994).
22 As the exposure to O3 progresses, airway inflammation begins to develop and the immune
23 response at both cellular and subcellular level is activated. Airway hyperreactivity develops
24 more slowly than pulmonary function effects, while neutrophilic inflammation of the airways
25 develops even more slowly and reaches the maximum 3 to 6 h postexposure. The cellular
26 responses (e.g., release of immuno-modulatory cytokines) appear to still be active as late as 20 h
27 postexposure (Torres et al., 2000). Following cessation of exposure, the recovery in terms of
28 breathing pattern, pulmonary function and airway hyperreactivity progresses rapidly and is
29 almost complete within 4 to 6 hours in moderately responsive individuals. More slowly
30 developing inflammatory and cellular changes may persist for up to 48 hours.
31
January 2005 6-9 DRAFT-DO NOT QUOTE OR CITE
-------
4.0-,
LU
2.0-
Pre 0 2 4
Time from Exposure (hours)
24
Figure 6-2. Recovery of FEVt responses following a 2 h exposure to 0.4 ppm O3 with IE.
Immediately postexposure, FEVt was significantly (**p<0.001) decreased.
At 3 h postexposure, FEVt was at 97% of the preexposure value.
Adapted from Nightingale et al. (2000).
1 6.2.5.1 Pathophysiologic Mechanisms
2 Breathing pattern changes
3 Human studies consistently report that inhalation of O3 alters the breathing pattern without
4 significantly affecting minute ventilation. A progressive decrease in tidal volume and a
5 "compensatory" increase in frequency of breathing to maintain steady minute ventilation during
6 exposure suggests a direct modulation of ventilatory control. These changes parallel a response
7 of many animal species exposed to O3 and other lower airway irritants (Tepper et al., 1990).
8 Bronchial C-fibers and rapidly adapting receptors appear to be the primary vagal afferents
9 responsible for O3-induced changes in ventilatory rate and depth in both humans (Folinsbee and
10 Hazucha, 2000) and animals (Coleridge et al., 1993; Hazucha and Sant'Ambrogio, 1993;
11 Schelegle et al., 1993).
12 The potential modulation of breathing pattern by activation of sensory afferents located in
13 extrathoracic airways by O3 has not yet been studied in humans. Nasal only O3 exposure of rats
14 produces changes in breathing pattern that are similar to changes observed in humans (Kleinman
15 etal., 1999).
January 2005
6-10
DRAFT-DO NOT QUOTE OR CITE
-------
1 Symptoms and lung function changes
2 As discussed, in addition to changes in ventilatory control, O3 inhalation by humans will
3 also induce a variety of symptoms, reduce vital capacity (VC) and related functional measures,
4 and increase airway resistance.
5 Schelegle et al. (2001) recently demonstrated that the reduction in VC due to O3 exposure
6 is a reflex action and not a voluntary termination of inspiration as result of discomfort. They
7 reported that O3-induced symptom responses (mediated in part by bronchial C-fibers) are
8 substantially reduced by inhaled topical anesthetic. However, the anesthetic had a minor and
9 irregular effect on pulmonary function decrements and tachypnea. Since respiratory symptom
10 response were largely abolished, these findings support reflex inhibition of VC due to
11 stimulation of both bronchial and pulmonary C-fibers.
12 The involvement of nociceptive bronchial C-fibers modulated by opioid receptors in
13 limiting maximal inspiration and eliciting subjective symptoms in humans was studied by
14 Passannante et al. (1998). Sufentanil (an opioid agonist and analgesic) rapidly reversed
15 O3-induced symptom responses and reduced spirometric decrements in "strong" responders. The
16 incomplete recovery in FEVj following sufentanil administration, however, suggests
17 involvement of non-opioid receptor modulated mechanisms as well. Interestingly, naloxone
18 (opioid receptor antagonist) had no significant effect on FEVj decrements in "weak" responders.
19 Plasma levels of p-endorphin (a potent pain suppressor) were not related with O3 responses.
20
21 Airway hyperreactivity
22 In addition to limitation of maximal inspiration and its effects on other spirometric
23 endpoints, activation of airway sensory afferents also plays a role in receptor-mediated
24 bronchoconstriction and an increase in airway resistance. Despite this common mechanism,
25 post-O3 pulmonary function changes and either early or late bronchial hyperresponsiveness
26 (BHR) to inhaled aerosolized methacholine or histamine are poorly correlated either in time or
27 magnitude. Fentanyl and indomethacin, the drugs that have been shown to attenuate O3-induced
28 lung function decrements in humans, did not prevent induction of BHR when administered to
29 guinea pigs prior to O3 exposure (Yeadon et al., 1992). Neither does post-O3 BHR seem to be
30 related to airway baseline reactivity. These findings imply that the mechanisms are either not
January 2005 6-11 DRAFT-DO NOT QUOTE OR CITE
-------
1 related or are activated independently in time. Animal studies (with limited support from human
2 studies) have suggested that an early post-O3 BHR is, at least in part, vagally mediated (Freed,
3 1996) and that stimulation of C-fibers can lead to increased responsiveness of bronchial smooth
4 muscle independently of systemic and inflammatory changes which may be even absent (load
5 et al., 1996). In vitro study of isolated human bronchi have reported that O3-induced airway
6 sensitization involves changes in smooth muscle excitation-contraction coupling (Marthan,
7 1996). Characteristic O3-induced inflammatory airway neutrophilia which at one time was
8 considered a leading BHR mechanism, has been found in a murine model, to be only
9 coincidentally associated with BHR and there was no cause and effect relationship (Zhang et al.,
10 1995). However, this observation does not rule out involvement of other cells such as
11 eosinophils or T-helper cells in BHR modulation. There is some evidence that release of
12 inflammatory mediators by these cells can sustain BHR and bronchoconstriction. In vitro and
13 animal studies have also suggested that airway neutral endopeptidase activity can be a strong
14 modulator of BHR (Marthan et al., 1996; Yeadon et al., 1992). Late BHR observed in some
15 studies is plausibly due to a sustained damage of airway epithelium and continual release of
16 inflammatory mediators (Foster et al., 2000). Thus, O3-induced BHR appears to be a product of
17 many mechanisms acting at different time periods and levels of the bronchial smooth muscle
18 signaling pathways (The effects ofO3 on BHR are described in Section 6.8}.
19
20 6.2.5.2 Mechanisms at a Cellular and Molecular Level
21 Stimulation of vagal afferents by O3 and reactive products, the primary mechanism of lung
22 function impairment is enhanced and sustained by what can be considered in this context to be
23 secondary mechanisms activated at a cellular and molecular level. The complexity of these
24 mechanisms is beyond the scope of this section and the reader is directed to Section 6.9 of this
25 chapter for greater detail. A comprehensive review by Mudway and Kelly (2000) discusses the
26 cellular and molecular mechanisms of O3-induced pulmonary response in great detail.
27 Stimulation of bronchial C-fibers by O3 not only inhibits maximal inspiration but, through
28 local axon reflexes, induces neurogenic inflammation. This pathophysiologic process is
29 characterized by release of tachykinins and other proinflammatory neuropeptides. Ozone
30 exposure has been shown to elevate C-fiber-associated tachykinin substance P in human
31 bronchial lavage fluid (Hazbun et al. 1993) and to deplete neuropeptides synthesized and
January 2005 6-12 DRAFT-DO NOT QUOTE OR CITE
-------
1 released from C-fibers in human airway epithelium rich in substance P-immunoreactive axons.
2 Substance P and other transmitters are known to induce granulocyte adhesion and subsequent
3 transposition into the airways, increase vascular permeability and plasma protein extravasation,
4 cause bronchoconstriction, and promote mucus secretion (Solway and Leff, 1991). Although the
5 initial pathways of neurogenic, antigen-induced, and generally immune-mediated inflammation
6 are not the same, they eventually converge leading to further amplification of airway
7 inflammatory processes by subsequent release of cytokines, eicosanoids, and other mediators.
8 Significantly negative correlations between O3-induced leukotriene (LTC4/D4/E4) production and
9 spirometric decrements (Hazucha et al., 1996), and an increased level of postexposure PGE2, a
10 mediator known to stimulate bronchial C-fibers, show that these mediators play an important
11 role in attenuation of lung function due to O3 exposure (Mohammed et al., 1993; Hazucha et al.,
12 1996). Moreover, because the density of bronchial C-fibers is much lower in the small than
13 large airways, the reported post O3 dysfunction of small airways assessed by decrement in
14 FEF25.75 (Weinman et al., 1995; Frank et al., 2001) may be due in part to inflammation. Also,
15 because of the relative slowness of inflammatory responses as compared to reflex effects, O3-
16 triggered inflammatory mechanisms are unlikely to initially contribute to progressive lung
17 function reduction. It is plausible, however, that when fully activated, they sustain and possibly
18 further aggravate already impaired lung function. Indeed, a prolonged recovery of residual
19 spirometric decrements following the initial rapid improvement after exposure termination could
20 be due to slowly resolving airway inflammation. Bronchial biopsies performed 6 h postexposure
21 have shown that O3 caused a significant decrease in immunoreactivity to substance P in the
22 submucosa (Krishna et al., 1997). A strong negative correlation with FEVj also suggests that the
23 release of substance P may be a contributing mechanism to persistent post-O3
24 bronchoconstriction (Krishna et al., 1997). Persistent spirometry changes observed for up to
25 48 h postexposure could plausibly be sustained by the inflammatory mediators, many of which
26 have bronchoconstrictive properties (Blomberg et al., 1999).
27
28
January 2005 6-13 DRAFT-DO NOT QUOTE OR CITE
-------
1 6.3 SUBJECTS WITH PREEXISTING DISEASE
2 Individuals with respiratory disease are of primary concern in evaluating the health effects
3 of O3 because even a small change in function is likely to have more impact on a person with
4 reduced reserve, i.e., O3-induced effects are superimposed on preexisting pulmonary impairment.
5
6 6.3.1 Subjects with Chronic Obstructive Pulmonary Disease
7 For patients with COPD performing light to moderate IE, no decrements in pulmonary
8 function were observed after 1- and 2-h exposures to < 0.30 ppm O3 (Kehrl et al., 1985; Linn
9 et al., 1982a, 1983a; Solic et al., 1982) and only small decreases in forced expiratory volume
10 were observed for 3-h exposures of chronic bronchitics to 0.41 ppm O3 (Kulle et al., 1984).
11 More recently, Gong et al. (1997a) found no significant difference in response between age-
12 matched controls and COPD patients to a 4 h exposure to 0.24 ppm O3 with IE. Although the
13 clinical significance is uncertain, small transient decreases in arterial blood oxygen saturation
14 have also been observed in some of these studies.
15
16 6.3.2 Subjects with Asthma
17 Based on studies reviewed in the 1996 criteria document (U.S. Environmental Protection
18 Agency, 1996), asthmatics appear to be at least as sensitive to acute effects of O3 as healthy
19 nonasthmatic subjects.
20 Several recent studies support a tendency for slightly increased spirometric responses in
21 mild asthmatics versus healthy subjects. Alexis et al. (2000) reported reductions in FVC (12%,
22 10%) and FEVj (13%, 11%) for 13 mild asthmatic and 9 healthy subjects, respectively, exposed
23 to 0.4 ppm O3 for 2 h with IE ( VE = 30 L/min). The FVC and FEVj responses were attenuated
24 by indomethacin in the healthy subjects but not the asthmatics. As assessed by the magnitude of
25 reductions in mid-flows (viz. FEF25, FEF50, FEF60p, FEF75) following O3 exposure, the small
26 airways tended to be more affected in asthmatics than healthy subjects. In a larger study, Torres
27 et al. (1996) exposed 24 asthmatics, 12 allergic rhinitics, and 10 healthy subjects to 0.25 ppm O3
28 for 3 h with IE. The O3-induced FEVj decrements tended to be greater in the diseased
29 populations (allergic rhinitis, 14.1%; asthmatics, 12.5%; healthy controls, 10.2%).
30 Scannell et al. (1996) exposed 18 asthmatics to 0.2 ppm O3 for 4 h with IE (VE -25 L/min/m2
January 2005 6-14 DRAFT-DO NOT QUOTE OR CITE
-------
1 BSA). An O3-induced increase in sRaw tended to be greater in the asthmatics compared to 81
2 healthy subjects who underwent similar experimental protocols (Aris et al., 1995; Balmes et al.,
3 1996).
4 Similar O3-induced spirometric responses are suggested by some studies. The Scannell
5 et al. (1996) study of 18 asthmatics reported FEVj and FVC decrements that were similar to 81
6 healthy subjects (Aris et al., 1995; Balmes et al., 1996). Similar group decrements in FEVj and
7 FVC were reported by Hiltermann et al. (1995), who exposed 6 asthmatics and 6 healthy
8 subjects to 0.4 ppm O3 for 2 h with light IE. Basha et al. (1994) also reported similar spirometric
9 responses between 5 asthmatic and 5 healthy subjects exposed to 0.2 ppm O3 for 6 h with IE.
10 The lack of significant differences in the Hilltermann et al. (1995) and Basha et al. (1994)
11 studies is not compelling given the extremely small sample sizes and corresponding lack of
12 statistical power. The Basha et al. (1994) study was also confounded by the asthmatics having
13 an average preexposure FEVj that was about 430 mL lower (a 12% difference) on the O3-day
14 relative to the air-day. Hence, only the Scannell et al. (1996) study supports similar O3-induced
15 spirometric responses in asthmatics versus healthy subjects.
16 One study has reported that asthmatics tend to have smaller O3-induced FEVj decrements
17 relative healthy subjects (3% versus 8%, respectively) when exposed to 0.2 ppm O3 for 2 h with
18 IE (Mudway et al., 2001). However, the asthmatics in the Mudway et al. (2001) study also
19 tended to be older than the healthy subjects, which could partially explain their lesser response.
20 In a longer exposure duration (7.6 h) study, Horstman et al. (1995) exposed 17 mild-to-
21 moderate asthmatics and 13 healthy controls to 0.16 ppm O3 or FA with QCE (VE «30 L/min).
22 The FEVj decrement observed in the asthmatics was significantly greater than in the healthy
23 subjects (19% versus 10%, respectively). There was also tendency for a greater O3-induced
24 decrease in FEF25_75 in asthmatics relative to the healthy subjects (24% versus 15%,
25 respectively). A significant positive correlation in asthmatics was also reported between O3-
26 induced spirometric responses and baseline lung function, i.e., responses increased with severity
27 of disease.
28 With repeated O3 exposures asthmatics, like healthy subjects (see Section 6.6} develop
29 tolerance. Gong et al. (1997b) exposed 10 asthmatics to 0.4 ppm O3, 3 h per day with IE
30 (VE -32 L/min), for 5 consecutive days. Symptom and spirometric responses were greatest on
31 the first (-35 % FEVj) and second (-34 % FEVj) exposure days, and progressively diminished
January 2005 6-15 DRAFT-DO NOT QUOTE OR CITE
-------
1 toward baseline levels (-6 % FEVj) by the fifth exposure day. Similar to healthy subjects,
2 asthmatics lost their tolerance 4 and 7 days later.
3 Other studies have reported that asthmatics have a somewhat exaggerated inflammatory
4 response to acute O3 exposure relative to healthy controls (e.g., McBride et al., 1994; Basha
5 et al., 1994; Peden et al., 1995, 1997; Peden, 2001a; Scannell et al., 1996; Hiltermann et al.,
6 1997, 1999; Michelson et al., 1999; Vagaggini et al., 1999; Newson et al., 2000; Holz et al.,
7 2002) also (see Section 6.9). Inflammatory responses do not appear to be correlated with lung
8 function responses in either asthmatic or healthy subjects (Balmes et al., 1996, 1997; Holz et al.,
9 1999). This lack of correlations between inflammatory and spirometric responses may be due to
10 differences in the time kinetics of these responses (Stenfors et al., 2002). In addition, airway
11 responsiveness to inhaled allergens is increased by O3 exposure in subjects with allergic asthma
12 for up to 24 h (see Section 6.8).
13
14 6.3.3 Subjects with Allergic Rhinitis
15 Allergic rhinitis is a condition defined by inflammation of the nasal membranes. Nayak
16 (2003) recently reviewed the commonalities between asthma and allergic rhinitis. Clinically,
17 greater than 60% of asthmatics have allergic rhinitis and slightly less than 40% of allergic
18 rhinitics have asthma. Leukotrienes and histamine are well-recognized mediators of responses
19 (viz., inflammation, hyperresponsiveness, and bronchoconstriction) in both asthma and allergic
20 rhinitis. Although, rhinitis and asthma are distinguished as affecting the upper and lower
21 airways, respectively, it has been suggested that these diseases are manifestations of the same
22 disease entity.
23 Given the prevalence of concomitant asthma and rhinitis and their common response
24 mediators, it should be expected that allergic rhinitics might respond more similarly to
25 asthmatics than healthy individuals. Regarding spirometric responses, Torres et al. (1996)
26 provide the only data demonstrating a trend in support of this supposition.
27 Studies demonstrating the interaction between air pollutants and allergic processes in the
28 human nasal airways and rhinoconjunctival tissue have been reviewed by Peden (200Ib) and
29 Riediker et al. (2001), respectively. Ozone exposure of subjects with allergic rhinitis has been
30 shown to induce nasal inflammation and increase airway responsiveness to nonspecific
31 bronchoconstrictors.
January 2005 6-16 DRAFT-DO NOT QUOTE OR CITE
-------
1 Peden et al. (1995), who studied allergic asthmatics exposed to O3 found that O3 causes an
2 increased response to nasal allergen challenge in addition to nasal inflammatory responses.
3 Their data suggested that allergic subjects have an increased immediate response to allergen after
4 O3 exposure. In a follow-up study, Michelson et al. (1999) reported that 0.4 ppm O3 did not
5 promote early-phase-response mediator release or enhance the response to allergen challenge in
6 the nasal airways of mild, asymptomatic dust mite-sensitive asthmatic subjects. Ozone did,
7 however, promote an inflammatory cell influx, which helps induce a more significant late-phase
8 response in this population.
9 Torres et al. (1996) found that O3 causes an increased response to bronchial allergen
10 challenge in subjects with allergic rhinitis. This study also measured responses in healthy
11 subjects and mildly allergic asthmatics (see Sections AX6.3.2 andAX6.8). All subjects were
12 exposed to 0.25 ppm O3 for 3 h with IE. Statistically significant O3-induced decrements in FEVj
13 occurred in rhinitics (14.1%), asthmatics (12.5%), and the healthy controls (10.2%), but these
14 responses did not differ statistically between groups. Methacholine responsiveness was
15 significantly increased in asthmatics, but not in subjects with allergic rhinitis. Airway
16 responsiveness to an individual's historical allergen (either grass and birch pollen, house dust
17 mite, or animal dander) was significantly increased after O3 exposure when compared to FA
18 exposure. The authors concluded that subjects with allergic rhinitis, but without asthma, could
19 be at risk if a high O3 exposure is followed by a high dose of allergen.
20 Holz et al. (2002) extended the results of Torres et al. (1996) by demonstrating that
21 repeated daily exposure to lower concentrations of O3 (0.125 ppm for 4 days) causes an
22 increased response to bronchial allergen challenge in subjects with preexisting allergic airway
23 disease, with or without asthma. These investigators observed no major difference in the pattern
24 of bronchial allergen response between asthmatics or rhinitics, except for a 10-fold increase in
25 the dose of allergen required to elicit a similar response (> 20% decrease in FEVj) in the
26 asthmatic subjects. Early phase responses were more consistent in subjects with rhinitis and
27 late-phase responses were more pronounced in subjects with asthma. There also was a tendency
28 towards a greater effect of O3 in subjects with greater baseline response to specific allergens
29 (chosen on the basis of skin prick test and history, viz., grass, rye, birch, or alder pollen, house
30 dust mite, or animal dander). These data suggest that the presence of allergic bronchial
31 sensitization, but not a history of asthma, may be a key determinant of increased airway allergen
Tanuary 2005 6-17 DRAFT-DO NOT QUOTE OR CITE
-------
1 responsiveness following exposure to O3 (for a more complete discussion of airway
2 responsiveness) see Section AX6.8.
3
4 6.3.4 Subjects with Cardiovascular Disease
5 Possibly due to the age of subjects studied, O3 exposure does not appear to result in
6 significant pulmonary function impairment or evidence of cardiovascular strain in patients with
7 cardiovascular disease relative to healthy controls. Gong et al. (1998) exposed 10 hypertensive
8 and 6 healthy adult males, 41 to 78 years of age, to 0.3 ppm O3 for 3 h with IE at 30 L/min. For
9 all subjects combined (no significant group differences), there was an O3-induced decrement of
10 7% in FEVj and an 70% increase in the alveolar-arterial oxygen tension gradient. The overall
11 results did not indicate any major acute cardiovascular effects of O3 in either the hypertensive or
12 normal subjects.
13
14
15 6.4 INTERSUBJECT VARIABILITY AND REPRODUCIBILITY
16 OF RESPONSE
17 Analysis of factors that contribute to intersubj ect variability is important for the
18 understanding of individual responses, mechanisms of response, and health risks associated with
19 acute O3 exposures. A large intersubj ect variability in response to O3 has been reported by
20 numerous investigators (Adams et al., 1981; Aris et al., 1995; Folinsbee et al., 1978; Kulle et al.,
21 1985; McDonnell et al., 1983). The magnitude of individual variability in FEVj response in 2 h
22 IE exposures increases at higher O3 concentrations (Kulle et al., 1985: McDonnell et al., 1983).
23 McDonnell (1996) examined the FEVj response data from three 6.6-h exposure studies
24 conducted at the EPA Health Effects Research Laboratory, and showed that the FEVj responses
25 in FA were small with most tightly grouped around zero. With increasing O3 concentrations
26 between 0.08 and 0.12 ppm, the mean response became asymmetrical with a few individuals
27 experiencing quite large decrements in FEVj {Intersubject variability observed in O3 dosimetry
28 studies is discussed in Chapter 4.2).
29 As an example of the variation in spirometric responses to O3 exposure, Hazucha et al.
30 (2003) analyzed the distribution of O3 responsiveness in 240 subjects (18 to 60 years of age)
31 exposed to 0.42 ppm O3 (on 3 occasions) for 1.5 h with IE at VE = 20 L/min/m2 BSA. Across all
January 2005 6-18 DRAFT-DO NOT QUOTE OR CITE
-------
1 ages, 18% of subjects were weak responders (< 5% FEVj decrement), 39% were moderate
2 responders, and 43% were strong responders (> 15% FEVj decrement). Younger subjects
3 (< 35 years of age) were predominately strong responders, whereas, older subjects (> 35 years of
4 age) were mainly weak responders. The influence of age on intersubject variability was also
5 noted by Passannante et al. (1998) who found that subjects under 35 years of age were more like
6 to be strong responders than older individuals. For repeated exposures, Hazucha et al. (2003)
7 reported that the reproducibility of FEVj responses was related to the length of time between
8 exposures. The Spearman correlation coefficient of 0.54 was found between responses for
9 exposures separated by 105 days (median), whereas, a correlation coefficient of 0.85 was found
10 between responses for exposures separated by only 7 days (median).
11 The more reproducible the subject's response, the more precisely it indicates his/her
12 intrinsic responsiveness. In 2 h IE O3 exposures, McDonnell et al. (1985b) found a relatively
13 poor FEVj reproducibility (R = 0.58) at the lowest concentration, 0.12 ppm, due, in part, to a
14 lack of specific O3 response or a uniformly small response in the majority of subjects. It was
15 concluded that for 2 h IE O3 exposures equal to or greater than 0.18 ppm, the intersubject
16 differences in magnitude of change in FVC and FEVj are quite reproducible over time (21 to
17 385 days; mean = 33 days) and are due primarily to differences in intrinsic responsiveness of
18 individual subjects to O3 exposure.
19 Intersubject variability, mechanisms of response, and health risks associated with acute O3
20 exposures are complicated by a poor association between various O3-induced responses. In a
21 retrospective study of 485 male subjects (ages 18 to 36 yrs) exposed to one of six O3
22 concentrations at one of three activity levels for 2 h, McDonnell et al. (1999) observed
23 significant, but low, Spearman rank order correlations between FEVj response and symptoms of
24 cough (R = 0.39), shortness of breath (R = 0.41), and pain on deep inspiration (R = 0.30). These
25 authors concluded that these responses are related mechanistically to some degree, but indicates
26 that there is not a single factor which is responsible for the observed individual differences in O3
27 responsiveness across the spectrum of symptom and lung function responses.
28 The effect of large intersubject variability on the ability to predict individual
29 responsiveness to O3 was demonstrated by McDonnell et al. (1993). These investigators
30 analyzed the data of 290 male subjects (18 to 32 years of age) who underwent repeat 2 h IE
31 exposures to one or more O3 concentrations ranging from 0.12 to 0.40 ppm. They attempted to
January 2005 6-19 DRAFT-DO NOT QUOTE OR CITE
-------
1 identify personal characteristics (i.e., age, height, baseline pulmonary functions, presence of
2 allergies, and past smoking history) that might predict individual differences in FEVj response.
3 Only age contributed significantly to intersubject responsiveness (younger subjects were more
4 responsive), accounting for just 4% of the observed variance. Interestingly, O3 concentration
5 accounted for only 31% of the variance, strongly suggesting the importance of as yet undefined
6 individual characteristics that determine FEVj responsiveness to O3. The authors concluded that
7 much individual variability in FEVj response to O3 remains unexplained.
8
9
10 6.5 FACTORS MODIFYING RESPONSIVENESS TO OZONE
11 6.5.1 Influence of Age
12 Beyond the age of 18 to 20 yrs, spirometric and symptom responses to O3 exposure begin
13 to decline with increasing age. In healthy individuals, the rate of decline in O3 responsiveness
14 appears to be greater in younger (18 to 35 yrs) versus middle aged (35 to 55 yrs) individuals
15 (Passannante et al., 1998; Hazucha et al., 2003). Beyond this age (> 55 yrs), acute O3 exposure
16 elicits minimal spirometric changes. An average FEVj decrement of-3% has been reported by
17 Gong et al. (1997a) for this older population under a "worst case" exposure scenario (0.24 ppm
18 O3 with 4 h IE). Although Gong et al. (1997a) and others have examined responses to O3
19 exposure in subjects of various ages, the exposure conditions differ between most studies so that
20 age effects remain uncertain.
21 Three recent studies, which analyzed large data sets (>240 subjects) of similarly exposed
22 subjects, show clearly discernable changes in FEVj responses to O3 as a function of age. Seal
23 et al. (1996) analyzed O3-induced spirometric responses in 371 young nonsmokers (18 to
24 35 years of age) exposed for 2.3 h during IE at a VE of 25 L/min/m2 BSA. On average, for the
25 same O3 concentration (C), the response of 25, 30, and 35 year old individuals are predicted to
26 be 83, 65, and 48%, respectively, of the response in 20 year olds. For example, a 5.4%
27 decrement in FEVj is predicted for 20 year old exposed to 0.12 ppm O3 for 2.3 h IE (VE = 25
28 L/min/m2 BSA), whereas, a similarly exposed 35 yr old is predicted to have only a 2.6%
29 decrement.
30 McDonnell et al. (1997) examined FEVj responses in 485 healthy white males (18 to
31 36 years of age) exposed once for 2 h to an O3 concentration of 0.0, 0.12, 0.18, 0.24, 0.30, or
January 2005 6-20 DRAFT-DO NOT QUOTE OR CITE
-------
1 0.40 ppm at rest or one of two levels of IE (VE of 25 and 35 L/min/m2 BSA). For the same
2 exposure conditions (C, VE, and duration), the average responses of 25, 30, and 35 year old
3 individuals are predicted to be 69, 48, and 33%, respectively, of the response in 20 year olds.
4 Hazucha et al. (2003) analyzed the distribution of O3 responsiveness in 240 subjects (18 to
5 60 years of age) exposed to 0.42 ppm O3 for 1.5 h with IE at VE = 20 L/min/m2 BSA. In males,
6 the FEVj responses of 25, 35, and 50 year olds are predicted to be 94, 83, and 50% ,
7 respectively, of the average response in 20 year old males. In females, the FEVj responses of 25,
8 35, and 50 year olds are predicted to be 82, 46, and 18%, respectively, of the average response in
9 20 year old females.
10 For subjects aged 18 to 36 yrs, McDonnell et al. (1999) recently reported that symptom
11 responses from O3 exposure also decrease with increasing age. Whether the same age-dependent
12 pattern of O3 sensitivity decline also holds for airway reactivity or inflammatory endpoints has
13 not been determined.
14
15 6.5.2 Gender and Hormonal Influences
16 Several studies have suggested that physiological differences between the genders may
17 predispose females to a greater susceptibility to O3. Lower plasma and nasal lavage fluid levels
18 of uric acid (the most prevalent antioxidant) in females relative to males may be a contributing
19 factor (Housley et al., 1996). Consequently, reduced absorption of O3 in the upper airways may
20 promote its deeper penetration. Dosimetric measurements have shown that the absorption
21 distribution of O3 is independent of gender when absorption is normalized to anatomical dead
22 space (Bush et al., 1996). More recently, Ultman et al. (2004) reported that the whole lung
23 uptake fraction of O3 was significantly greater in males (91.4%) than females (87.1%). But, this
24 increase in O3 uptake in the males was consistent with their larger VT and smaller fB relative to
25 the females. Furthermore, O3 uptake was not correlated with spirometric responses. Thus, a
26 differential removal of O3 by uric acid seems to be minimal. In general, the physiologic
27 response of young healthy females to O3 exposure appears comparable to the response of young
28 males (Hazucha et al., 2003). Although, during the follicular phase of the menstrual cycle, lung
29 function response to O3 is enhanced (Fox et al., 1993).
30
January 2005 6-21 DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
6.5.3 Racial, Ethnic and Socioeconomic Status Factors
A few epidemiologic studies have implied that minorities are more responsive to O3 than
then Caucasians. However, this may be more of a consequence of the overall quality of health
care and SES then an innate sensitivity to oxidants (Gwynn and Thurston, 2001; Seal et al,
1996). The paucity of data has prevented making any definitive conclusions on the influence of
race, ethnic or other related factors on the responsiveness to O3.
6.5.4 Influence of Physical Activity
Any physical activity will increase minute ventilation and therefore the dose of inhaled O3.
Consequently, the intensity of physiological response following an acute exposure will be
strongly associated with minute ventilation (see Figure 6-3).
Time (h)
Figure 6-3. Predicted O3-induced decrements in FEVt as a function of exposure duration
and level of IE (line labels are VE levels) in young healthy adults (20 yrs of
age) exposed to 0.3 ppm O3. The illustrated activity levels range from rest
(VE = 10 L/min) to moderate exercise (VE = 40 L/min). Predictions are for
Model 1 coefficients in Table 3 of McDonnell et al. (1997).
Source: Based on McDonnell et al. (1997).
January 2005
6-22
DRAFT-DO NOT QUOTE OR CITE
-------
1 6.5.5 Environmental Factors
2 Since the 1996 O3 AQCD (U.S. Environmental Protection Agency, 1996) few human
3 laboratory studies have examined the potential influence of environmental factors such as rural
4 versus urban environment, passive cigarette smoke exposure, and bioactive admixtures such as
5 endotoxin on healthy individual's pulmonary function changes due to O3.
6 New controlled human exposure studies have confirmed that smokers are less responsive
7 to O3 than nonsmokers. Spirometric and plethysmographic pulmonary function decline,
8 nonspecific airway hyperreactivity, and inflammatory response of smokers to O3 were all weaker
9 than those reported for nonsmokers. Although all these responses are intrinsically related, the
10 functional association between them, as in nonsmokers, has been weak. Similarly, the time
11 course of development and recovery of these effects, as well their reproducibility, was not
12 different from nonsmokers. Chronic airway inflammation with desensitization of bronchial
13 nerve endings and an increased production of mucus may plausibly explain the pseudo-
14 protective effect of smoking (Frampton et al., 1997; Torres et al., 1997).
15 The effect on environment tobacco smoke (ETS) on O3 responses has received very little
16 attention. In one study, preexposure of mice to sidestream cigarette smoke (ETS surrogate),
17 elicited no immediate effects, but potentiated subsequent O3-induced inflammatory responses
18 (Yu et al., 2002) (See Chapter 5.4.2 for additional ETS details}.
19 The influence of ambient temperature on pulmonary effects induced by O3 exposure in
20 humans has been studied infrequently under controlled laboratory conditions. Several
21 experimental human studies have reported additive effects of heat and O3 exposure (see U.S.
22 Environmental Protection Agency, 1986, 1996). Foster et al. (2000) exposed 9 young healthy
23 subjects for 130 min (IE 10 min at 36 to 39 1/min) to filtered air and to ramp profile O3 at 22°
24 and 30 °C, 45-55% RH. The O3 exposure started at 0.12 ppm, reached the peak of 0.24 ppm
25 mid-way through and subsequently declined to 0.12 ppm at the end of exposure. At the end of
26 exposure FEVj decreased significantly (p < 0.5) by -8% at 22°C and -6.5% at 30 °C. One day
27 (19 h) later, the decline of 2.3% from baseline was still significant (p < 0.05). FVC decrements
28 were smaller and significant only for the 22 °C condition immediately postexposure. There was
29 a decline in specific airway conductance (sGaw; p < 0.05) at 30°C but not at 22 °C. The
30 nonspecific bronchial responsiveness to methacloline assessed as PC50 sGaw was significantly
January 2005 6-23 DRAFT-DO NOT QUOTE OR CITE
-------
1 (p < 0.05) higher one day following O3 exposure at both temperatures but more so at 30 °C.
2 Thus, these findings suggest that elevated temperature may partially attenuate spirometric
3 responses but enhance airway reactivity.
4
5 6.5.6 Oxidant-Antioxidant Balance
6 The first line of defense against oxidative stress is antioxidant present in epithelial lining
7 fluid (ELF) which scavenge free radicals and limit lipid peroxidation. Exposure to O3 depletes
8 the antioxidant level in nasal ELF probably due to scrubbing of O3 (Mudway et al., 1999),
9 however, the concentration and the activity of antioxidant enzymes either in ELF or plasma do
10 not appear to be related to O3 responsiveness (Avissar et al., 2000; Blomberg et al., 1999; Samet
11 et al., 2001). Carefully controlled studies of dietary antioxidant supplementation have
12 demonstrated some protective effects of a-tocopherol and ascorbate on spirometric lung function
13 from O3 but not on the intensity of subjective symptoms and inflammatory response including
14 cell recruitment, activation and a release of mediators (Samet et al., 2001; Trenga et al., 2001).
15 Dietary antioxidants have also afforded partial protection to asthmatics by attenuating post-
16 exposure bronchial hyperresponsiveness (Trenga et al., 2001). Animal studies (described in
17 Chapter 5.2.1.3) have also demonstrated the protective effects of ELF antioxidants during O3
18 exposures.
19
20 6.5.7 Genetic and Other Factors
21 Several recent studies (Bergamaschi et al., 2001) have reported that genetic polymorphism
22 of antioxidant enzymes may modulate pulmonary function and inflammatory response to O3
23 challenge. It appears that healthy carriers of NQO1 wild type in combination with GSTM1 null
24 genotype are more responsive to O3. Adults with GSTM1 null only genotype did not show O3
25 hyperresponsiveness. In contrast, asthmatic children with GSTM1 null genotype (Romieu et al,
26 2004) were reported to be more responsive to O3. In general, the findings between studies are
27 inconsistent and additional, better controlled studies, are needed to clarify influence of genetic
28 polymorphism on O3 responsiveness.
29
30
January 2005 6-24 DRAFT-DO NOT QUOTE OR CITE
-------
1 6.6 REPEATED O3 EXPOSURE EFFECTS
2 Based on studies reviewed here and in the previous O3 criteria documents (U.S.
3 Environmental Protection Agency, 1986, 1996), several conclusions can be drawn about
4 repeated 1 to 2-h O3 exposures. Repeated exposures to O3 can cause an enhanced (i.e., greater)
5 pulmonary function response on the second day of exposure (see Tables AX6-8 andAX6-9for
6 added detail). This enhancement appears to be dependent on the interval between the exposures
7 (24 h is associated with the greatest increase) and is absent with intervals > 3 days (Bedi et al.,
8 1985; Folinsbee and Horvath, 1986; Schonfeld et al., 1989). An enhanced response also appears
9 to depend to some extent on the magnitude of the initial response (Horvath et al., 1981). Small
10 responses to the first O3 exposure are less likely to result in an enhanced response on the second
11 day of O3 exposure (Folinsbee et al., 1994). With continued daily exposures (i.e., beyond the
12 second day) there is an attenuation of pulmonary function responses, typically after 3 to 5 days
13 of repeated exposure. This attenuated response persists for less than 1 week (Kulle et al., 1982;
14 Linn et al., 1982b) or as long as 2 weeks (Horvath et al., 1981). In temporal conjunction with
15 pulmonary function changes, symptoms induced by O3, such as cough, PDI, and chest
16 discomfort, are increased on the second exposure day and attenuated with repeated exposure
17 thereafter (Folinsbee et al., 1980, 1998; Foxcroft and Adams, 1986; Linn et al., 1982b).
18 O3-induced changes in airway responsiveness persist longer and attenuate more slowly than
19 pulmonary function and symptoms responses (Dimeo et al., 1981; Kulle et al., 1982), although
20 this has been studied only on a limited basis (Folinsbee et al., 1994). In longer-duration (4 h to
21 6.6 h), lower-concentration studies that do not cause an enhanced second-day response, the
22 attenuation of response to O3 appears to proceed more rapidly (Folinsbee et al., 1994) [Effects of
23 repeated exposures on inflammatory responses are discussed in Section 6.9.4).
24
25
26 6.7 EFFECTS ON EXERCISE PERFORMANCE
27 The effects of acute O3 inhalation on endurance exercise performance have been examined
28 in numerous controlled laboratory studies. These studies were discussed in the 1996 O3 AQCD
29 (U.S. Environmental Protection Agency, 1996) and can be divided into two categories: (1) those
30 that examined the effects of acute O3 inhalation on maximal oxygen uptake (VO2max) and
January 2005 6-25 DRAFT-DO NOT QUOTE OR CITE
-------
1 (2) those that examined the effects of acute O3 inhalation on the ability to complete strenuous
2 continuous exercise protocols of up to 1 h in duration.
3 In brief, endurance exercise performance and VO2max may be limited by acute exposure to
4 O3 (Adams and Schelegle, 1983; Schelegle and Adams, 1986; Gong et al., 1986; Foxcroft and
5 Adams, 1986; Folinsbee et al., 1977; Linder et al., 1988). Gong et al. (1986) and Schelegle and
6 Adams (1986) found that significant reductions in maximal endurance exercise performance may
7 occur in well-conditioned athletes while they perform CE (VE > 80 L/min) for 1 h at O3
8 concentrations > 0.18 ppm. Reports from studies of exposure to O3 during high-intensity
9 exercise indicate that breathing discomfort associated with maximal ventilation may be an
10 important factor in limiting exercise performance in some, but not all, subjects.
11
12
13 6.8 EFFECTS ON AIRWAY RESPONSIVENESS
14 Airway or bronchial hyperresponsiveness (BHR) refers to a condition in which the
15 propensity for the airways to bronchoconstrict due to a variety of stimuli becomes augmented.
16 Airway responsiveness is typically quantified by measuring the decrement in pulmonary
17 function (i.e., spirometry or plethysmography) following the inhalation of small amounts of an
18 aerosolized bronchoconstrictor agent (specific [antigen, allergen] or nonspecific [methacholine,
19 histamine]) or a measured stimulus (e.g., exercise, cold air).
20 Ozone exposure causes an increase in nonspecific airway responsiveness as indicated by a
21 reduction in the concentration of methacholine or histamine required to produce a given
22 reduction in FEVj or increase in SRaw. Increased airway responsiveness is an important
23 consequence of exposure to O3 because its presence means that the airways are predisposed to
24 narrowing on inhalation of a variety of stimuli (e.g., specific allergens, SO2, cold air).
25 Ozone exposure of asthmatic subjects, who characteristically have increased airway
26 responsiveness at baseline, can cause further increases in responsiveness (Kreit et al., 1989).
27 Similar relative changes in airway responsiveness are seen in asthmatics exposed to O3 despite
28 their markedly different baseline airway responsiveness. Several studies (Torres et al., 1996;
29 Kehrl et al., 1999; Molfino et al., 1991) have been published suggesting an increase in specific
30 (i.e., allergen-induced) airway reactivity. An important aspect of increased airway
January 2005 6-26 DRAFT-DO NOT QUOTE OR CITE
-------
1 responsiveness after O3 exposure is that this represents a plausible link between ambient O3
2 exposure and increased hospital admissions for asthma.
3 Changes in airway responsiveness after O3 exposure appear to be resolved more slowly
4 than changes in FEVj or respiratory symptoms (Folinsbee and Hazucha, 2000). Furthermore, in
5 studies of repeated exposure to O3, changes in airway responsiveness tend to be somewhat less
6 susceptible to attenuation with consecutive exposures than changes in FEVj (Dimeo et al., 1981;
7 Folinsbee et al., 1994; Gong et al., 1997b; Kulle et al., 1982). Increases in airway
8 responsiveness do not appear to be strongly associated with decrements in lung function or
9 increases in symptoms.
10 The mechanism of O3-induced increases in airway responsiveness is only partially
11 understood, but it appears to be associated with a number of cellular and biochemical changes in
12 airway tissue. Although inflammation could play a role in the increase in airway responsiveness,
13 cyclooxygenase inhibitors have not been effective at blocking the O3-induced influx of PMNs
14 into BAL fluid (Hazucha et al., 1996; Ying et al., 1990). Therefore, O3-induced airway
15 responsiveness may not be due to the presence of PMNs in the airway or to the release of
16 arachidonic acid metabolites. Rather, it seems likely that the mechanism for this response is
17 multifactorial, possibly involving the presence of cytokines, prostanoids, or neuropeptides;
18 activation of macrophages, eosinophils, or mast cells; and epithelial damage that increases direct
19 access of mediators to the smooth muscle or receptors in the airways that are responsible for
20 reflex bronchoconstriction.
21
22
23 6.9 INFLAMMATION AND HOST DEFENSE EFFECTS
24 6.9.1 Introduction
25 Short-term exposure of humans to O3 can cause acute inflammation and that long-term
26 exposure of laboratory animals results in a chronic inflammatory state (see Chapter 5). The
27 relationship between repetitive bouts of acute inflammation in humans caused by O3 and the
28 development of chronic respiratory disease is unknown.
29 The presence of neutrophils (PMNs) in the lung has long been accepted as a hallmark of
30 inflammation and is an important indicator that O3 causes inflammation in the lungs. It is
31 apparent, however, that inflammation within airway tissues may persist beyond the point that
January 2005 6-27 DRAFT-DO NOT QUOTE OR CITE
-------
1 inflammatory cells are found in BAL fluid. Soluble mediators of inflammation such as the
2 cytokines (IL-6, IL-8) and arachidonic acid metabolites (e.g., PGE2, PGF2a, thromboxane, and
3 leukotrienes [LTs] such as LTB4) have been measured in the BAL fluid of humans exposed to
4 O3. In addition to their role in inflammation, many of these compounds have
5 bronchoconstrictive properties and may be involved in increased airway responsiveness
6 following O3 exposure.
7 Some recent evidence suggests that changes in small airways function may provide a
8 sensitive indicator of O3 exposure and effect, despite the fact that inherent variability in their
9 measurement by standard spirometric approaches make their assessment difficult. Observations
10 of increased functional responsiveness of these areas relative to the more central airways, and of
11 persistent effects following repeated exposure, may indicate that further investigation of
12 inflammatory processes in these regions is warranted.
13
14 6.9.2 Inflammatory Response in the Upper Respiratory Tract
15 The nasal passages constitute the primary portal for inspired air at rest and, therefore, the
16 first region of the respiratory tract to come in contact with airborne pollutants. Nikasinovic et al.
17 (2003) recently reviewed the literature of laboratory-based nasal inflammatory studies published
18 since 1985. Nasal lavage (NL) has provided a useful tool for assessing O3-induced inflammation
19 in the nasopharynx. Increased levels of PMNs in the NL fluid of humans exposed to 0.5 ppm
20 O3 at rest for 4 h has been reported (Graham et al., 1988; Bascom et al., 1990).
21 Graham and Koren (1990) compared inflammatory mediators present in both the NL and
22 BAL fluids of humans exposed to 0.4 ppm O3 for 2 h. Similar increases in PMN were observed
23 in NL and BAL, suggesting a qualitative correlation between inflammatory changes in the lower
24 airways (BAL) and the upper respiratory tract (NL). Torres et al. (1997) compared NL and BAL
25 in smokers and nonsmokers exposed to 0.22 ppm O3 for 4 h. In contrast to Graham and Koren
26 (1990), they did not find a relationship between numbers or percentages of PMNs in the nose
27 and the lung, perhaps in part due to the variability observed in their NL recoveries. Albumin, a
28 marker of epithelial cell permeability, was increased 18 h later, but not immediately after
29 exposure, as seen by Bascom et al. (1990).
30 McBride et al. (1994) reported that asthmatic subjects were more sensitive than non-
31 asthmatics to upper airway inflammation at an O3 concentration (0.24 ppm (1.5 h)) that did not
January 2005 6-28 DRAFT-DO NOT QUOTE OR CITE
-------
1 affect lung or nasal function or biochemical mediators. A significant increase in the number of
2 PMNs in NL fluid was detected in the asthmatic subjects both immediately and 24 h after
3 exposure. Peden et al. (1995) also found that O3 at a concentration of 0.4 ppm had a direct nasal
4 inflammatory effect, and reported a priming effect on the response to nasal allergen challenge, as
5 well. A subsequent study in dust mite-sensitive asthmatic subjects indicated that O3 at this
6 concentration enhanced eosinophil influx in response to allergen, but did not promote early
7 mediator release or enhance the nasal response to allergen (Michelson et al., 1999). Similar to
8 observations made in the lower airways, the presence of O3 molecular "targets" in nasal lining
9 fluid is likely to provide some level of local protection against exposure. In a study of healthy
10 subjects exposed to 0.2 ppm O3 for 2 h, Mudway and colleagues (1999) observed a significant
11 depletion of uric acid in NL fluid at 1.5 h following exposure.
12
13 6.9.3 Inflammatory Response in the Lower Respiratory Tract
14 Seltzer et al. (1986) were the first to demonstrate that exposure of humans to O3 resulted in
15 inflammation in the lung. Bronchoalveolar lavage fluid (3 h post-exposure) from subjects
16 exposed to O3 contained increased PMNs as well as increased levels of PGE2, PGF2a, and TXB2
17 compared to fluid from air-exposed subjects. Koren et al. (1989a,b) described inflammatory
18 changes 18 h after O3 exposure. In addition to an eightfold increase in PMNs, Koren et al.
19 reported a two-fold increase in BAL fluid protein, albumin, and immunoglobulin G (IgG) levels,
20 suggestive of increased epithelial cell permeability. There was a 12-fold increase in IL-6 levels,
21 a two-fold increase in PGE2, and a two-fold increase in the complement component, C3a.
22 Evidence for stimulation of fibrogenic processes in the lung was shown by significant increases
23 in coagulation components, Tissue Factor and Factor VII (McGee et al., 1990), urokinase
24 plasminogen activator and fibronectin (Koren et al., 1989a). Subsequent studies by Lang et al.,
25 (1998), using co-cultures of cells of the BEAS-2B bronchial epithelial line and of the HFL-1
26 lung fibroblast line, provided additional information about O3-induced fibrogenic processes.
27 They demonstrated that steady-state mRNA levels of both alpha 1 and procollagens type I and
28 III in the fibroblasts were increased following O3 exposure and that this effect was mediated by
29 the O3-exposed epithelial cells. This group of studies demonstrated that exposure to O3 results in
30 an inflammatory reaction in the lung, as evidenced by increases in PMNs and proinflammatory
31 compounds. Furthermore, they demonstrated that cells and mediators capable of damaging
January 2005 6-29 DRAFT-DO NOT QUOTE OR CITE
-------
1 pulmonary tissue are increased after O3 exposure, and provided early suggestion of the potential
2 importance of the epithelial cell-myofibroblast "axis" in modulating fibrotic and fibrinolytic
3 processes in the airways.
4 Isolated lavage of the mainstream bronchus using balloon catheters or BAL using small
5 volumes of saline have been used to assess O3-induced changes in the large airways. Studies
6 collecting lavage fluid from isolated airway segments after O3 exposure indicate increased
7 neutrophils in the airways (Aris et al., 1993; Balmes et al., 1996; Scannell et al., 1996). Other
8 evidence of airway neutrophil increase comes from studies in which the initial lavage fraction
9 ("bronchial fraction") showed increased levels of neutrophils (Schelegle et al., 1991; Peden
10 et al., 1997; Balmes et al., 1996; Torres et al., 1997). Bronchial biopsies show increased PMNs
11 in airway tissue (Aris et al.,1993) and, in sputum collected after O3 exposure, neutrophil numbers
12 are elevated (Fahy et al., 1995).
13 Increased BAL protein, suggesting O3-induced changes in epithelial permeability (Koren
14 et al., 1989a, 1991 and Devlin et al., 1991) supports earlier work in which increased epithelial
15 permeability, as measured by increased clearance of radiolabled diethylene triamine pentaacetic
16 acid (99mTc-DTPA) from the lungs of humans exposed to O3, was demonstrated (Kehrl et al.,
17 1987). In addition, Foster and Stetkiewicz (1996) have shown that increased permeability
18 persists for at least 18-20 h and the effect is greater at the lung apices than at the base. In a study
19 of mild atopic asthmatics exposed to 0.2 ppm O3 for 2 h, Newson, et al. (2000) observed a 2-fold
20 increase in the percentage of PMNs present at 6 hours post exposure, with no change in markers
21 of increased permeability as assessed by sputum induction. By 24 h, the neutrophilia was seen
22 to subside while levels of albumin, total protein, myeloperoxidase, and eosinophil cationic
23 protein increased significantly. It was concluded that the transient PMN influx induced by acute
24 exposure of these asthmatic subjects was followed by plasma extravasation and the activation of
25 both PMNs and eosinophils within the airway tissues. Such changes in permeability associated
26 with acute inflammation may provide better access of inhaled antigens, particulates, and other
27 substances to the submucosal region.
28 Devlin et al. (1991) reported an inflammatory response in subjects exposed to 0.08 and
29 0.10 ppm O3 for 6.6 h. Increased numbers of PMNs and levels of IL-6 were found at both
30 O3 concentrations, suggesting that lung inflammation from O3 can occur as a consequence of
31 prolonged exposure to ambient levels while exercising. Interestingly, those individuals who had
January 2005 6-30 DRAFT-DO NOT QUOTE OR CITE
-------
1 the largest increases in inflammatory mediators in this study did not necessarily have the largest
2 decrements in pulmonary function, suggesting that separate mechanisms underlie these two
3 responses. The absence of a relationship between spirometric responses and inflammatory cells
4 and markers has been reported in several studies (Balmes et al., 1996; Schelegle et al., 1991;
5 Torres et al., 1997; Hazucha et al., 1996; Blomberg et al., 1999). These observations relate
6 largely to disparities in the times of onset and duration following single exposures.
7 As indicated above, a variety of potent proinflammatory mediators have been reported to
8 be released into the airway lumen following O3 exposure. Studies of human alveolar
9 macrophages (AM) and airway epithelial cells exposed to O3 in vitro suggest that most mediators
10 found in the BAL fluid of O3-exposed humans are produced by epithelial cells. Macrophages
11 exposed to O3 in vitro showed only small increases in PGE2 (Becker et al., 1991). In contrast,
12 airway epithelial cells exposed in vitro to O3 showed large concentration-dependent increases in
13 PGE2, TXB2, LTB4, LTC4, and LTD4 (McKinnon et al., 1993) and increases in IL-6, IL-8, and
14 fibronectin at O3 concentrations as low as 0.1 ppm (Devlin et al., 1994). Macrophages lavaged
15 from subjects exposed to 0.4 ppm (Koren et al., 1989a) showed changes in the rate of synthesis
16 of 123 different proteins, whereas AMs exposed to O3 in vitro showed changes in only six
17 proteins, suggesting that macrophage function was altered by mediators released from other
18 cells. Furthermore, recent evidence suggests that the release of mediators from AMs may be
19 modulated by the products of O3-induced oxidation of airway lining fluid components, such as
20 human surfactant protein A (Wang et al., 2002).
21 Although the release of mediators has been demonstrated to occur at exposure
22 concentrations and times that are minimally cytotoxic to airway cells, potentially detrimental
23 latent effects have been demonstrated in the absence of cytotoxicity. These include the
24 generation of DNA single strand breaks (Kozumbo et al., 1996) and the loss of cellular
25 replicative activity (Gabrielson et al., 1994) in bronchial epithelial cells exposed in vitro, and the
26 formation of protein and DNA adducts. A highly toxic aldehyde formed during O3-induced lipid
27 peroxidation is 4-hydroxynonenal (HNE). Healthy human subjects exposed to 0.4 ppm O3 for 1
28 h underwent BAL 6 h later. Analysis of lavaged alveolar macrophages by Western blot
29 indicated increased levels of a 32-kDa HNE-protein adduct, as well as 72-kDa heat shock protein
30 and ferritin, in O3- versus air-exposed subjects (Hamilton et al., 1998). In a recent study of
31 healthy subjects exposed to 0.1 ppm O3 for 2 h (Corradi et al., 2002), formation of 8-hydroxy-2'-
January 2005 6-31 DRAFT-DO NOT QUOTE OR CITE
-------
1 deoxyguanosine (8-OHdG), a biomarker of reactive oxidant species (ROS)-DNA interaction,
2 was measured in peripheral blood lymphocytes. At 18 h post exposure, 8-OHdG was
3 significantly increased in cells compared to pre-exposure levels, presumably linked to concurrent
4 increases in chemical markers of ROS. Of interest, the increase in 8-OHdG was only significant
5 in a subgroup of subjects with the wild genotype for NAD(P)H:quinone oxidoreductase and the
6 null genotype for glutathione-S-transferase Ml, suggesting that polymorphisms in redox
7 enzymes may confer "susceptibility' to O3 in some individuals. The generation of ROS
8 following exposure to O3 has been shown to be associated with a wide range of responses. In a
9 recent study, ROS production by alveolar macrophages lavaged from subjects exposed to
10 0.22 ppm for 4 h was assessed by flow cytometry (Voter et al., 2001). Levels were found to be
11 significantly elevated 18 h post exposure and associated with several markers of increased
12 permeability. An in vitro study of human tracheal epithelial cells exposed to O3 indicated that
13 generation of ROS resulted in decrease in synthesis of the bronchodilatory prostaglandin, PGE2,
14 as a result of inactivation of prostaglandin endoperoxide G/H synthase 2 (Alpert et al., 1997).
15 These and similar studies indicate that the responses to products of O3 exposure in the airways
16 encompass a broad range of both stimulatory and inhibitory activities, many of which may be
17 modulated by susceptibility factors upstream in the exposure process, at the level of
18 compensating for the imposed oxidant stress.
19 The inflammatory responses to O3 exposure also have been studied in asthmatic subjects
20 (Basha et al., 1994; Scannell et al., 1996; Peden et al., 1997). In these studies, asthmatics
21 showed significantly more neutrophils in the BAL (18 h post-exposure) than similarly exposed
22 healthy individuals. In one of these studies (Peden et al., 1997), which included only allergic
23 asthmatics who tested positive for Dematophagoides farinae antigen, there was an eosinophilic
24 inflammation (2-fold increase), as well as neutrophilic inflammation (3-fold increase). In a
25 study of subjects with intermittent asthma that utilized a 2-fold higher concentration of O3 (0.4
26 ppm) for 2 h, increases in eosinophil cationic protein, neutrophil elastase and IL-8 were found to
27 be significantly increased 16 h post-exposure and comparable in induced sputum and BAL fluid
28 (Hiltermann et al, 1999). In two studies (Basha et al., 1994; Scannell et al., 1996), IL-8 was
29 significantly higher in post-O3 exposure BAL in asthmatics compared to non-asthmatics,
30 suggesting a possible mediator for the increased neutrophilic inflammation in those subjects.
31 In a recent study comparing the neutrophil response to O3 at a concentration and exposure time
January 2005 6-32 DRAFT-DO NOT QUOTE OR CITE
-------
1 similar to those of the latter three studies, Stenfors and colleagues (2002) were unable to detect a
2 difference in the increased neutrophil numbers between 15 mild asthmatic and 15 healthy
3 subjects by bronchial wash at the 6 h post-exposure time point. These results suggest that, at
4 least with regard to neutrophil influx, differences between healthy and asthmatic individuals
5 develop gradually following exposure and may not become evident until later in the process.
6 In another study, mild asthmatics who exhibited a late phase underwent allergen challenge 24 hrs
7 before a 2 h exposure to 0.27 ppm O3 or filtered air in a cross-over design (Vagaggini et al.,
8 2002). At 6 h post-exposure, eosinophil numbers in induced sputum were found to be
9 significantly greater after O3 than after air. Studies such as these suggest that the time course of
10 eosinophil and neutrophil influx following O3 exposure can occur to levels detectable within the
11 airway lumen by as early as 6 h. They also suggest that the previous or concurrent activation of
12 proinflammatory pathways within the airway epithelium may enhance the inflammatory effects
13 of O3. For example, in an in vitro study of epithelial cells from the upper and lower respiratory
14 tract, cytokine production induced by rhinovirus infection was enhanced synergistically by
15 concurrent exposure to O3 at 0.2 ppm for 3 h (Spannhake et al, 2002). The use of bronchial
16 mucosal biopsies has also provided important insight into the modulation by O3 of existing
17 inflammatory processes within asthmatics. In a study of healthy and allergic asthmatic subjects
18 exposed to 0.2 ppm O3 or filtered air for 2 h, biopsies were performed 6 hr following exposure
19 (Bosson et al., 2003). Monoclonal antibodies were used to assess epithelial expression of a
20 variety of cytokines and chemokines. At baseline (air exposure), asthmatic subjects showed
21 significantly higher expression of interleukins (IL)-4 and -5. Following O3 exposure, the
22 epithelial expression of IL-5, IL-8, granulocyte-macrophage colony-stimulating factor (GM-
23 CSF) and epithelial cell-derived neutrophil-activating peptide 78 (ENA-78) was significantly
24 greater in asthmatic subjects, as compared to healthy subjects. In vitro studies of bronchial
25 epithelial cells derived by biopsy from nonatopic, nonasthmatic subjects and asthmatic subjects
26 also demonstrated the preferential release of GM-CSF and also of regulated on activation,
27 normal T cell-expressed and -secreted (RANTES) from asthmatic cells following O3 exposure.
28 The time course of the inflammatory response to O3 in humans has not been explored fully.
29 Nevertheless, studies in which BAL was performed 1-3 h (Devlin et al., 1990; Koren et al.,
30 1991; Seltzer et al., 1986) after exposure to 0.4 ppm O3 demonstrated that the inflammatory
31 response is quickly initiated, and other studies (Koren et al., 1989a,b; Torres et al., 1997;
January 2005 6-33 DRAFT-DO NOT QUOTE OR CITE
-------
1 Scannell et al., 1996; Balmes et al., 1996) indicated that, even 18 h after exposure, inflammatory
2 mediators such as IL-6 and PMNs were still elevated. However, different markers show peak
3 responses at different times. Ozone-induced increases in IL-8, IL-6, and PGE2 are greater
4 immediately after O3 exposure, whereas BAL levels of fibronectin and plasminogen activator are
5 greater after 18 h. PMNs and some products (protein, Tissue Factor) are similarly elevated both
6 1 and 18 h after O3 exposure (Devlin et al., 1996; Torres et al., 1997). Schelegle et al. (1991)
7 found increased PMNs in the "proximal airway" lavage at 1, 6, and 24 h after O3 exposure, with
8 a peak response at 6 h. In a typical BAL sample, PMNs were elevated only at the later time
9 points. This is consistent with the greater increase 18 h after exposure seen by Torres et al.
10 (1997). In addition to the influx of PMNs and (in allergic asthmatics) eosinophils, lymphocyte
11 numbers in BAL were also seen to be elevated significantly at 6 h following exposure of healthy
12 subjects to 0.2 ppm O3 for 2 h (Blomberg et al., 1997). Analysis of these cells by flow cytometry
13 indicated the increased presence of CD3+, CD4+ and CD8+ T cell subsets. This same laboratory
14 later demonstrated that within 1.5 h following exposure of healthy subjects to the same O3
15 regimen, expression of human leukocyte antigen (HLA)-DR on lavaged macrophages underwent
16 a significant, 2.5-fold increase (Blomberg et al., 1999). The significance of these alterations in
17 immune system components and those in IL-4 and IL-5 expression described above in the
18 studies of Bosson et al. (2003) has not been fully explored and may suggest a role for O3 in the
19 modulation of immune inflammatory processes.
20
21 6.9.4 Effects of Repeated Exposures and Adaptation of Responses
22 Physiologic and symptomatic responses in humans following repeated exposure to O3 were
23 discussed in Section 6.6. Inflammatory responses upon repeated O3 exposures are discussed in
24 this section. Animal studies suggest that while inflammation may be diminished with repeated
25 exposure, underlying damage to lung epithelial cells continues (Tepper et al., 1989). Markers
26 from BALF following both 2-h (Devlin et al., 1997) and 4-h (Christian et al., 1998; Torres et al.,
27 2000) repeated O3 exposures (up to 5 days) indicate that there is ongoing cellular damage
28 irrespective of the attenuation of some cellular inflammatory responses of the airways,
29 pulmonary function, and symptom responses.
30 Devlin et al. (1997) examined the inflammatory responces of humans repeatedly exposed
31 to 0.4 ppm O3 for 5 consecutive days. Several indicators of inflammation (e.g., PMN influx, IL-
January 2005 6-34 DRAFT-DO NOT QUOTE OR CITE
-------
1 6, PGE2, BAL protein, fibronectin) were attenuated after 5 days of exposure (i.e., values were
2 not different from FA). Several markers (LDH, IL-8, total protein, epithelial cells) did not show
3 attenuation, indicating that tissue damage probably continues to occur during repeated exposure.
4 The recovery of the inflammatory response occurred for some markers after 10 days, but some
5 responses were not normalized even after 20 days. The continued presence of markers of
6 cellular injury indicates a persistent but not necessarily recognized due to attention of
7 spirometric and symptom responses to O3.
8 Christian et al. (1998) randomly subjected heathy subjects to a single exposure and to 4
9 consecutive days of exposure to 0.2 ppm O3 for 4 h. Both "bronchial" and "alveolar" fractions
10 of the BAL showed decreased numbers of PMNs and fibronectin concentration at day 4 versus
11 the single exposure, and a decrease in IL-6 levels in the alveolar fraction.
12 Following a similar study design and exposure parameters, Torres et al. (2000) found both
13 functional and BAL cellular responses to O3 were abolished at 24 h postexposure following the
14 fourth exposure day. However, levels of total protein, IL-6, IL-8, reduced glutathione and ortho-
15 tyrosine were still increased significantly. In addition, visual scores for bronchitis, erythema and
16 the numbers of neutrophils in the mucosal biopsies were increased. Their results indicate that,
17 despite reduction of some markers of inflammation in BAL and measures of large airway
18 function, inflammation within the airways persists following repeated exposure to O3.
19 Holz, et al. (2002) made a comparison of early and late responses to allergen challenge
20 following O3 in subjects with allergic rhinitis or allergic asthma. With some variation, both early
21 and late FEVj and cellular responses in the two subject groups were significantly enhanced by 4
22 consecutive days of exposure to 0.125 ppm O3 for 3 h.
23 In another study, Frank and colleagues (2001) exposed healthy subjects to FA and to O3
24 (0.25 ppm, 2 h) on 4 consecutive days each, with pulmonary function measurements being made
25 prior to and following each exposure. BAL was performed on day 5, 24 h following the last
26 exposure. On day 5, PMN numbers remained significantly higher following O3 compared to FA.
27 Of particular note in this study was the observation that small airway function, assessed by
28 grouping values for isovolumetric FEF25_75, VmaxSO and Vmax75 into a single value, showed
29 persistent reduction from day 2 through day 5. Following exposure of and asthmatic and healthy
30 subjects for one day to 0.4 ppm O3 for 2 h, Alexis et al. (2000) have also reported that variables
31 representing small airways function (viz., FEF25, FEF50, FEF 60P, FEF75) demonstrated the
January 2005 6-3 5 DRAFT-DO NOT QUOTE OR CITE
-------
1 greatest O3-induced decline in the asthmatic subjects. These data suggest that techniques
2 monitoring the function in the small peripheral airway regions, the primary sites of O3 uptake in
3 the lung, may provide important information regarding both acute and cumulative effects of O3
4 exposure.
5
6 6.9.5 Effect of Anti-Inflammatory and other Mitigating Agents
7 Pretreatment of healthy subjects with non-steroidal anti-inflammatory drugs (ibuprofen,
8 etc.) has been found to partially suppress development of airway inflammation and pulmonary
9 function changes (U.S. Environmental Protection Agency, 1996). Although atropine blocked the
10 increase in Raw in response to O3 exposure, it did not alter the spirometric or symptom
11 responses (Beckett et al., 1985). Similarly, albuterol and salbutamol, which had no effect on O3-
12 induced changes in spirometry, also had no effect of symptom responses (McKenzie et al., 1987;
13 Gongetal., 1988). The anti-inflammatory medications indomethacin and ibuprofen, which
14 partially inhibit the spirometric responses to O3 exposure, also cause a reduction in respiratory
15 symptoms (Schelegle et al., 1987; Hazucha et al., 1994). Indomethacin attenuates decrements in
16 FEVj and FVC in healthy subjects, but not asthmatics (Alexis et al., 2000). In contrast,
17 inhalation of the corticosteroid budesonide does not prevent or even attenuate O3-induced
18 responses in healthy subjects as assessed by measurements of lung function, bronchial reactivity
19 and airway inflammation (Nightingale et al., 2000). In asthmatic subjects, budesonide decreases
20 airway neutrophil influx following O3 exposure (Vagaggini et al., 2001). This suggests that
21 corticosteroids may be effective only when the inflammation is already present, such as in
22 asthmatics.
23
24 6.9.6 Changes in Host Defense Capability Following Ozone Exposures
25 A number of studies clearly show that a single acute exposure (1 to 4 h) of humans to
26 moderate concentrations of O3 (0.2 to 0.6 ppm) while exercising at moderate to heavy levels
27 results in a number of cellular and biochemical changes in the lung including an inflammatory
28 response characterized by increased numbers of PMNs, increased permeability of the epithelial
29 cells lining the respiratory tract, cell damage, and production of proinflammatory cytokines and
30 prostaglandins. This response can be detected as early as 1 h after exposure (Koren et al., 1991;
January 2005 6-36 DRAFT-DO NOT QUOTE OR CITE
-------
1 Schelegle et al., 1991) and persists for at least 18 h (Aris et al., 1993; Koren et al., 1989a). The
2 response profile of these mediators is not defined adequately, although it is clear that the time
3 course of response varies for different mediators and cells (Devlin et al., 1997; Schelegle et al.,
4 1991). These changes also occur in humans exposed to 0.08 and 0.10 ppm O3 for 6 to 8 h
5 (Devlin et al., 1991; Peden et al., 1997). Decrements in the ability of AMs to phagocytose
6 microorganisms have also been reported. Ozone also causes inflammatory changes in the nose,
7 as indicated by increased levels of PMNs and albumin, a marker for increased epithelial cell
8 permeability. Nasal lavage analyses, however, are not necessarily parallel to BAL analyses.
9 There appears to be no strong correlation between any of the measured cellular and
10 biochemical changes and changes in lung function measurements, suggesting that different
11 mechanisms may be responsible for these processes (Balmes et al., 1996; Devlin et al., 1991).
12 The idea of different mechanisms is supported by a study in which ibuprofen, a cyclooxygenase
13 inhibitor, blunted the O3-induced decrements in lung function without altering the O3-induced
14 increase in PMNs or epithelial cell permeability (Hazucha et al., 1996).
15 In vitro studies suggest that epithelial cells are the primary target of O3 in the lung and that
16 O3 induces them to produce many of the mediators found in the BAL fluid of humans exposed to
17 O3. Although O3 does not induce AMs to produce these compounds in large quantities, it does
18 directly impair the ability of AMs to phagocytose and kill microorganisms.
19 Only two studies (Foster et al., 1987; Gerrity et al., 1993) have investigated the effect of
20 O3 exposure on mucociliary particle clearance in humans. Foster et al. (1987) measured
21 clearance during and after a 2 h exposure to 0.4 ppm O3. Gerrity et al. (1993) measured
22 clearance at 2 h postexposure (0.4 ppm O3), by which time, sRaw had returned to baseline and
23 FVC was within 5% of baseline (versus an 11% decrement immediately postexposure). Foster
24 et al. (1987) found a stimulatory effect of acute O3 exposure on mucociliary clearance. Gerrity
25 et al. (1993), who observed no effect on clearance, suggested that transient clearance increases
26 are coincident to pulmonary function responses. Investigators in both studies suggested that
27 O3-induced increases in mucociliary clearance could be mediated by cholinergic receptors.
28
29
January 2005 6-37 DRAFT-DO NOT QUOTE OR CITE
-------
1 6.10 EXTRAPULMONARY EFFECTS OF OZONE
2 Ozone reacts rapidly on contact with respiratory system tissue and is not absorbed or
3 transported to extrapulmonary sites to any significant degree as such. Human exposure studies
4 discussed in the previous criteria documents (U.S. Environmental Protection Agency, 1986,
5 1996) failed to demonstrate any consistent extrapulmonary effects. More recently, some human
6 exposure studies have attempted to identify specific markers of exposure to O3 in blood. Foster
7 et al. (1996) found a reduction in the serum levels of the free radical scavenger a-tocopherol
8 after O3 exposure. Liu et al. (1997, 1999) used a salicylate metabolite, 2,3, dehydroxybenzoic
9 acid (DHBA), to indicate increased levels of hydroxyl radical which hydroxylates salicylate to
10 DHBA. Increased DHBA levels after exposure to 0.12 and 0.40 ppm suggest that O3 increases
11 production of hydroxyl radical. The levels of DHBA were correlated with changes in
12 spirometry.
13 Gong et al. (1998) monitored ECG, HR, cardiac output, blood pressure, oxygen saturation,
14 and chemistries, as well as calculating other hemodynamic variables (e.g., stroke volume,
15 vascular resistance, rate-pressure products) in both healthy and hypertensive adult males, 41 to
16 78 years of age. No major acute cardiovascular effects were found in either the normal or
17 hypertensive subjects after exposure to 0.3 ppm O3 for 3 h with intermittent exercise at 30 L/min.
18 Statistically significant O3 effects for both groups combined were a decrease in FEVl3 and
19 increases in HR, rate-pressure product, and the alveolar-to-arterial PO2 gradient, which might be
20 more important in some patients with severe cardiovascular disease.
21
22
23 6.11 EFFECTS OF OZONE MIXED WITH OTHER POLLUTANTS
24 Over the past 10 years only a handful of human controlled studies have examined the
25 effects of pollutant mixtures containing O3. The results of a controlled study on children (Linn
26 et al., 1997), designed to approximate exposure conditions of an epidemiologic study (Neas
27 et al., 1995) by matching the population and exposure atmosphere (0.1 ppm O3, 0.1 ppm SO2 and
28 101 |ig/m2 H2SO4), did not support the findings of this epidemiologic study. The study points
29 out the difficulties in attempting to link the outcomes of epidemiologic and controlled studies.
30 Another vulnerable population, asthmatics, demonstrated enhanced airway reactivity to house
31 dust mite following exposures to O3, NO2, and the combination of the two gases. Spirometric
January 2005 6-3 8 DRAFT-DO NOT QUOTE OR CITE
-------
1 response, however, was impaired only by O3, and O3+NO2 at higher concentrations (Jenkins
2 et al., 1999). It is plausible that uneven longitudinal absorption of SO2, NO2, and O3 in the
3 conducting airways may influence a response. Ozone has been found to be scrubbed more
4 efficiently in proximal airways and to penetrate less into the distal airways than either SO2 and
5 NO2 (Rigas et al., 1997). Inhalation of a mixture of PM25 and O3 by healthy subjects increased
6 brachial artery tone and reactivity (Brook et al., 2002). Since no other cardiovascular endpoints
7 were affected by the exposure, the pathophysiological importance of this observation remains
8 unclear.
9 All in all, the contention that air pollutant mixtures elicit stronger pathophysiologic effects
10 than individual pollutants of the mix is only weakly supported by human studies of either healthy
11 or at-risk population. The studies summarized in this section complement the studies reviewed
12 in the 1996 O3 AQCD (U.S. Environmental Protection Agency, 1996). Regarding the latter, the
13 mobile laboratory comparative studies of exercising athletes (Avol et al., 1984, 1985) with
14 chamber exposures to oxidant-polluted ambient air (mean O3 concentration of 0.153 ppm) and
15 purified air containing a controlled concentration of generated O3 at 0.16 ppm showed similar
16 pulmonary function responses and symptoms. These results strongly suggest that acute
17 exposures of coexisting ambient pollutants had minimal contribution to these responses under
18 the typical summer ambient conditions in Southern California. However, no unifying
19 conclusions can be reached since each study employed different mixtures and examined different
20 aspects of a response [The complexities ofO3 and co-pollutant exposures in animal studies are
21 discussed in Chapter 5.4.4}.
22
23
24 6.12 CONTROLLED STUDIES OF AMBIENT AIR EXPOSURES
25 A large amount of informative O3 exposure-effects data has been obtained in controlled
26 laboratory exposure studies under a variety of different experimental conditions. However,
27 laboratory simulation of the variable pollutant mixtures present in ambient air is not practical.
28 Thus, the exposure effects of one or several artificially generated pollutants (i.e., a simple
29 mixture) on pulmonary function and symptoms may not explain responses to ambient air where
30 complex pollutant mixtures exist.
31
January 2005 6-39 DRAFT-DO NOT QUOTE OR CITE
-------
1 6.12.1 Mobile Laboratory Studies
2 Quantitatively useful information on the effects of acute exposure to photochemical
3 oxidants on pulmonary function and symptoms responses from field studies using a mobile
4 laboratory were presented in prior criteria documents (U.S. Environmental Protection Agency,
5 1986, 1996). Relative to controlled exposure studies, mobile laboratory ambient air studies
6 suffer the additional limation of a dependence on ambient outdoor conditions. Consistent with
7 controlled exposure studies, mobile studies in California demonstrated that pulmonary effects
8 from exposure to ambient air in Los Angeles are related to O3 concentration and level of
9 exercise. Healthy subjects with a history of allergy also appeared to be more responsive to O3
10 than "nonallergic" subjects (Linn et al., 1980, 1983b), although a standardized evaluation of
11 atopic status was not performed.
12
13 6.12.2 Aircraft Cabin Studies
14 Respiratory symptoms and pulmonary function effects resulting from exposure to O3 in
15 commercial aircraft flying at high altitudes, and in altitude-simulation studies, have been
16 assessed previously (U.S. Environmental Protection Agency, 1986, 1996). Commercial aircraft
17 cabin O3 levels were reported to be very low (average concentration 0.01 to 0.02 ppm) during
18 92 randomly selected smoking and nonsmoking flights in 1989 (Nagda et al., 1989). None of
19 these flights recorded O3 concentrations exceeding the 3-h time-weighted average (TWA)
20 standard of 0.10 ppm promulgated by the U.S. Federal Aviation Administration (FAA, 1980),
21 probably due to the use of O3-scrubbing catalytic filters (Melton, 1990).
22 Ozone contamination aboard high-altitude aircraft also has been an interest to the U.S. Air
23 Force because of complaints by crew members of frequent symptoms of dryness and irritation of
24 the eyes, nose, and throat and an occasional cough (Hetrick et al., 2000). Despite the lack of
25 ventilation system modifications as used in commercial aircraft, the O3 concentrations never
26 exceeded the FAA ceiling limit of 0.25 ppm and exceeded the 3-h TWA of 0.10 ppm only 7% of
27 the total monitored flight time (43 h). The authors concluded that extremely low average
28 relative humidity (12%) during flight operations was most likely responsible for the reported
29 symptoms.
30
31
January 2005 6-40 DRAFT-DO NOT QUOTE OR CITE
-------
1 6.13 SUMMARY
2 Responses in humans exposed to ambient O3 concentrations include decreased inspiratory
3 capacity; mild bronchoconstriction; rapid, shallow breathing pattern during exercise; and
4 symptoms of cough and pain on deep inspiration. Reflex inhibition of inspiration results in a
5 decrease in forced vital capacity (FVC) and, in combination with mild bronchoconstriction,
6 contributes to a decrease in the forced expiratory volume in 1 s (FEVj). In addition to
7 physiological pulmonary responses and respiratory symptoms, O3 exposure also results in airway
8 hyperresponsiveness, inflammation, immune system activation, and epithelial injury. With
9 repeated O3 exposures over several days, spirometric and symptom responses become attenuated,
10 but this tolerance is lost after about a week without exposure. Airway responsiveness also
11 appears to be attenuated with repeated O3 exposures, but less than FEVj.
12 Young healthy adults exposed to O3 concentrations > 0.08 ppm develop significant
13 reversible, transient decrements in pulmonary function if minute ventilation ( VE ) or duration of
14 exposure are increased sufficiently. O3-induced decrements in FEVj do not appear to depend on
15 gender, race, body surface area, height, lung size, or baseline FVC in young healthy adults.
16 Healthy children experience similar spirometric responses but lesser symptoms from O3
17 exposure relative to young adults. Beyond the age of 18 to 20 yrs, spirometric and symptom
18 responses to O3 exposure begin to decline with increasing age. There is a large degree of
19 intersubject variability in physiologic and symptomatic responses of heathy adults exposed to
20 O3. However, responses tend to be reproducible within a given individual over a period of
21 several months. With increasing O3 concentration, the distribution FEVj decrements becomes
22 asymmetrical with a few individuals experiencing large decrements.
23 There is a tendency for slightly increased spirometric responses in mild asthmatics and
24 allergic rhinitics relative to healthy subjects. Spirometric responses in asthmatics appear to be
25 affected by baseline lung function, i.e., responses increase with disease severity. With repeated
26 daily O3 exposures, spirometric responses of asthmatics become attenuated, however, airway
27 responsiveness becomes increased in subjects with preexisting allergic airway disease (with or
28 without asthma). Possibly due to patient age, O3 exposure does not appear to cause significant
29 pulmonary function impairment or evidence of cardiovascular strain in patients with
30 cardiovascular disease or chronic obstructive pulmonary disease relative to healthy subjects.
January 2005 6-41 DRAFT-DO NOT QUOTE OR CITE
-------
1 Available information on recovery from O3 exposure indicates that an initial phase of
2 recovery in healthy individuals proceeds relatively rapidly, with acute spirometric and symptom
3 response appears to occur within about 2 to 4 h. Small residual lung function effects are almost
4 completely resolved within 24 hours. Effects of O3 on the small airways, assessed by decrement
5 in FEF25.75, may be due in part to inflammation. Indeed, a prolonged recovery of residual
6 spirometric decrements following the initial rapid recovery could be due to slowly resolving
7 airway inflammation. In hyperresponsive individuals, this recovery takes longer, as much as 48
8 hours, to return to baseline values. Persistent spirometry changes observed for up to 48 h
9 postexposure could plausibly be sustained by the inflammatory mediators. Cellular responses
10 (e.g., release of immuno-modulatory cytokines) appear to still be active as late as 20 h
11 postexposure. More slowly developing inflammatory and cellular changes may persist for up to
12 48 h, but the time course in humans has not been explored fully.
13 Soluble mediators of inflammation such as the cytokines (IL-6, IL-8) and arachidonic acid
14 metabolites (e.g., PGE2, PGF2a, thromboxane, and leukotrienes [LTs] such as LTB4) have been
15 measured in the BAL fluid of humans exposed to O3. Many of these compounds have
16 bronchoconstrictive properties and may be involved in increased airway responsiveness
17 following O3 exposure. Some indicators of inflammation (e.g., PMN influx, IL-6, PGE2, BAL
18 protein, fibronectin) are attenuated with repeated O3 exposures. Indicating that tissue damage
19 probably continues to occur during repeated O3 exposure, however, other markers (LDH, IL-8,
20 total protein, epithelial cells) did not show attenuation. There appears to be no strong correlation
21 between any of the measured cellular and biochemical changes and changes in lung function
22 measurements. Whether airway reactivity or inflammatory responses to O3 are dependent on the
23 age of the exposed individual, such as spirometric responses, has not been determined.
24 Dietary antioxidant supplementation attenuates O3-induced spirometric responses but not
25 the intensity of subjective symptoms nor inflammatory responses. Dietary antioxidants also
26 afforded partial protection to asthmatics by attenuating postexposure bronchial
27 hyperresponsiveness.
28
January 2005 6-42 DRAFT-DO NOT QUOTE OR CITE
-------
1 REFERENCES
2 Adams, W. C. (2000) Ozone dose-response effects of varied equivalent minute ventilation rates.
3 J. Exposure Anal. Environ. Epidemiol. 10: 217-226.
4 Adams, W. C. (2002) Comparison of chamber and face-mask 6.6-hour exposures to ozone on
5 pulmonary function and symptoms responses. Inhalation Toxicol. 14: 745-764.
6 Adams, W. C. (2003a) Comparison of chamber and face mask 6.6-hour exposure to 0.08 ppm
7 ozone via square-wave and triangular profiles on pulmonary responses. Inhalation Toxicol.
8 15:265-281.
9 Adams, W. C. (2003b) Relation of pulmonary responses induced by 6.6-h exposures to 0.08 ppm
10 ozone and 2-h exposures to 0.30 ppm ozone via chamber and face-mask inhalation.
11 Inhalation Toxicol. 15: 745-759.
12 Adams, W. C.; Ollison, W. M. (1997) Effects of prolonged simulated ambient ozone dosing
13 patterns on human pulmonary function and symptomatology. Presented at: 90th annual
14 meeting of the Air & Waste Management Association; June; Toronto, Ontario, Canada.
15 Pittsburgh, PA: Air & Waste Management Association; paper no. 97-MP9.02.
16 Adams, W. C.; Schelegle, E. S. (1983) Ozone and high ventilation effects on pulmonary function
17 and endurance performance. J. Appl. Physiol.: Respir. Environ. Exercise Physiol. 55:
18 805-812.
19 Adams, W. C.; Savin, W. M.; Christo, A. E. (1981) Detection of ozone toxicity during
20 continuous exercise via the effective dose concept. J. Appl. Physiol.: Respir. Environ.
21 Exercise Physiol. 51: 415-422.
22 Alexis, N.; Urch, B.; Tarlo, S.; Corey, P.; Pengelly, D.; O'Byrne, P.; Silverman, F. (2000)
23 Cyclooxygenase metabolites play a different role in ozone-induced pulmonary function
24 decline in asthmatics compared to normals. Inhalation Toxicol. 12: 1205-1224.
25 Alpert, S. E.; Walenga, R. W.; Jaspers, I; Qu, Q.; Chen, L. C. (1997) Ozone inactivates
26 Cyclooxygenase in human tracheal epithelial cells without altering PGHS-2 mRNA or
27 protein. Am. J. Physiol. 272: L879-L887.
28 Aris, R. M.; Christian, D.; Hearne, P. Q.; Kerr, K.; Finkbeiner, W. E.; Balmes, J. R. (1993)
29 Ozone-induced airway inflammation in human subjects as determined by airway lavage
30 and biopsy. Am. Rev. Respir. Dis. 148: 1363-1372.
31 Aris, R. M.; Tager, L; Christian, D.; Kelly, T.; Balmes, J. R. (1995) Methacholine
32 responsiveness is not associated with O3-induced decreases in FEVj. Chest 107: 621-628.
33 Avissar, N. E.; Reed, C. K.; Cox, C.; Frampton, M. W.; Finkelstein, J. N. (2000) Ozone, but not
34 nitrogen dioxide, exposure decreases glutathione peroxidases in epithelial lining fluid of
35 human lung. Am. J. Respir. Crit. Care Med. 162: 1342-1347.
36 Avol, E. L.; Linn, W. S.; Venet, T. G.; Shamoo, D. A.; Hackney, J. D. (1984) Comparative
37 respiratory effects of ozone and ambient oxidant pollution exposure during heavy exercise.
38 J. Air Pollut. Control Assoc. 34: 804-809.
39 Avol, E. L.; Linn, W. S.; Venet, T. G.; Shamoo, D. A.; Spier, C. E.; Hackney, J. D. (1985)
40 Comparative effects of laboratory generated ozone and ambient oxidant exposure in
41 continuously exercising subjects. In: Lee, S. D., ed. Evaluation of the scientific basis for
42 ozone/oxidants standards: proceedings of an APCA international specialty conference;
43 November 1984; Houston, TX. Pittsburgh, PA: Air Pollution Control Association;
44 pp. 216-225. (APCA international specialty conference transactions: TR-4).
January 2005 6-43 DRAFT-DO NOT QUOTE OR CITE
-------
1 Balmes, J. R.; Chen, L. L.; Scannell, C.; Tager, I.; Christian, D.; Hearne, P. Q.; Kelly, T.; Aris,
2 R. M. (1996) Ozone-induced decrements in FEVj and FVC do not correlate with measures
3 of inflammation. Am. J. Respir. Crit. Care Med. 153: 904-909.
4 Balmes, J. R.; Aris, R. M.; Chen, L. L.; Scannell, C.; Tager, I. B.; Finkbeiner, W.; Christian, D.;
5 Kelly, T.; Hearne, P. Q.; Ferrando, R.; Welch, B. (1997) Effects of ozone on normal and
6 potentially sensitive human subjects, part I: airway inflammation and responsiveness to
7 ozone in normal and asthmatic subjects. Cambridge, MA: Health Effects Institute.
8 Research report no. 78; pp 1-37, 81-99.
9 Bascom, R.; Naclerio, R. M.; Fitzgerald, T. K.; Kagey-Sobotka, A.; Proud, D. (1990) Effect of
10 ozone inhalation on the response to nasal challenge with antigen of allergic subjects. Am.
11 Rev. Respir. Dis. 142: 594-601.
12 Basha, M. A.; Gross, K. B.; Gwizdala, C. J.; Haidar, A. H.; Popovich, J., Jr. (1994)
13 Bronchoalveolar lavage neutrophilia in asthmatic and healthy volunteers after controlled
14 exposure to ozone and filtered purified air. Chest 106: 1757-1765.
15 Becker, S.; Madden, M. C.; Newman, S. L.; Devlin, R. B.; Koren, H. S. (1991) Modulation of
16 human alveolar macrophage properties by ozone exposure in vitro. Toxicol. Appl.
17 Pharmacol. 110:403-415.
18 Bedi, J. F.; Drechsler-Parks, D. M.; Horvath, S. M. (1985) Duration of increased pulmonary
19 function sensitivity to an initial ozone exposure. Am. Ind. Hyg. Assoc. J. 46: 731-734.
20 Bergamaschi, E.; De Palma, G.; Mozzoni, P.; Vanni, S.; Vettori, M. V.; Broeckaert, F.; Bernard,
21 A.; Mutti, A. (2001) Polymorphism of quinone-metabolizing enzymes and susceptibility to
22 ozone-induced acute effects. Am. J. Respir. Crit. Care Med. 163: 1426-1431.
23 Blomberg, A.; Helleday, R.; Pourazar, J.; Stenfors, N.; Kelly, F. J.; Frew, A. J.; Holgate, S. T.;
24 Sandstrom, T. (1997) Early airway and peripheral blood cell responses to 0.20 ppm ozone
25 in healthy human subjects. Eur. Respir. J. 10(suppl 25): 274S.
26 Blomberg, A.; Mudway, IS.; Nordenhall, C.; Hedenstrom, H.; Kelly, F. J.; Frew, A. J.; Holgate,
27 S. T.; Sandstrom, T. (1999) Ozone-induced lung function decrements do not correlate with
28 early airway inflammatory or antioxidant responses. Eur. Respir. J. 13: 1418-1428.
29 Bosson, J.; Stenfors, N.; Bucht, A.; Helleday, R.; Pourazar, J.; Holgate, S. T.; Kelly, f J.;
30 Sandstrom, T.; Wilson, S.; Frew, A. J.; Blomberg, A. (2003) Ozone-induced bronchial
31 epithelial cytokine expression differs between healthy and asthmatic subjects. Clin. Exp.
32 Allergy 33: 777-782.
33 Brook, R. D.; Brook, J. R.; Urch, B.; Vincent, R.; Rajagopalan, S.; Silverman, F. (2002)
34 Inhalation of fine particulate air pollution and ozone causes acute arterial vasoconstriction
35 in healthy adults. Circulation 105: 1534-1536.
36 Christian, D. L.; Chen, L. L.; Scannell, C. H.; Ferrando, R. E.; Welch, B. S.; Balmes, J. R.
37 (1998) Ozone-induced inflammation is attenuated with multiday exposure. Am. J. Respir.
38 Crit. Care Med. 158: 532-537.
39 Coleridge, J. C. G.; Coleridge, H. M.; Schelegle, E. S.; Green, J. F. (1993) Acute inhalation of
40 ozone stimulates bronchial C-fibers and rapidly adapting receptors in dogs. J. Appl.
41 Physiol. 74: 2345-2352.
42 Corradi, M.; Alinovi, R.; Goldoni, M.; Vettori, M.; Folesani, G.; Mozzoni, P.; Cavazzini, S.;
43 Bergamaschi, E.; Rossi, L.; Mutti, A. (2002) Biomarkers of oxidative stress after
44 controlled human exposure to ozone. Toxicol. Lett. 134: 219-225.
45 Devlin, R. B.; McDonnell, W. F.; Koren, H. S.; Becker, S. (1990) Prolonged exposure of humans
46 to 0.10 and 0.08 ppm ozone results in inflammation in the lung. Presented at: 83rd annual
January 2005 6-44 DRAFT-DO NOT QUOTE OR CITE
-------
1 meeting of the Air & Waste Management Association; June; Pittsburgh, PA. Pittsburgh,
2 PA: Air & Waste Management Association; paper no. 90-150.2.
3 Devlin, R. B.; McDonnell, W. F.; Mann, R.; Becker, S.; House, D. E.; Schreinemachers, D.;
4 Koren, H. S. (1991) Exposure of humans to ambient levels of ozone for 6.6 hours causes
5 cellular and biochemical changes in the lung. Am. J. Respir. Cell Mol. Biol. 4: 72-81.
6 Devlin, R. B.; McKinnon, K. P.; Noah, T.; Becker, S.; Koren, H. S. (1994) Ozone-induced
7 release of cytokines and fibronectin by alveolar macrophages and airway epithelial cells.
8 Am. J. Physiol. 266: L612-L619.
9 Devlin, R. B.; McDonnell, W. F.; Becker, S.; Madden, M. C.; McGee, M. P.; Perez, R.; Hatch,
10 G.; House, D. E.; Koren, H. S. (1996) Time-dependent changes of inflammatory mediators
11 in the lungs of humans exposed to 0.4 ppm ozone for 2 hr: a comparison of mediators
12 found in bronchoalveolar lavage fluid 1 and 18 hr after exposure. Toxicol. Appl.
13 Pharmacol. 138: 176-185.
14 Devlin, R. B.; Folinsbee, L. J.; Biscardi, F.; Hatch, G.; Becker, S.; Madden, M. C.; Robbins, M.;
15 Koren, H. S. (1997) Inflammation and cell damage induced by repeated exposure of
16 humans to ozone. Inhalation Toxicol. 9: 211-235.
17 Dimeo, M. J.; Glenn, M. G; Holtzman, M. J.; Shelter, J. R.; Nadel, J. A.; Boushey, H. A. (1981)
18 Threshold concentration of ozone causing an increase in bronchial reactivity in humans
19 and adaptation with repeated exposures. Am. Rev. Respir. Dis. 124: 245-248.
20 Fahy, J. V.; Wong, H. H.; Liu, J. T.; Boushey, H. A. (1995) Analysis of induced sputum after air
21 and ozone exposures in healthy subjects. Environ. Res. 70: 77-83.
22 Federal Aviation Administration. (1980) Airplane cabin ozone contamination. F. R. (January 21)
23 45: 3880-3883.
24 Folinsbee, L. J.; Hazucha, M. J. (1989) Persistence of ozone-induced changes in lung function
25 and airway responsiveness. In: Schneider, T.; Lee, S. D.; Wolters, G. J. R.; Grant, L. D.,
26 eds. Atmospheric ozone research and its policy implications: proceedings of the 3rd
27 US-Dutch international symposium; May 1988; Nijmegen, The Netherlands. Amsterdam,
28 The Netherlands: Elsevier Science Publishers; pp. 483-492. (Studies in environmental
29 science 35).
30 Folinsbee, L. J.; Hazucha, M. J. (2000) Time course of response to ozone exposure in healthy
31 adult females. Inhalation Toxicol. 12: 151-167.
32 Folinsbee, L. J.; Horvath, S. M. (1986) Persistence of the acute effects of ozone exposure. Aviat.
33 Space Environ. Med. 57: 1136-1143.
34 Folinsbee, L. J.; Silverman, F.; Shephard, R. J. (1977) Decrease of maximum work performance
35 following ozone exposure. J. Appl. Physiol.: Respir. Environ. Exercise Physiol. 42:
36 531-536.
37 Folinsbee, L. J.; Drinkwater, B. L.; Bedi, J. F.; Horvath, S. M. (1978) The influence of exercise
38 on the pulmonary function changes due to exposure to low concentrations of ozone. In:
39 Folinsbee, L. J.; Wagner, J. A.; Borgia, J. F.; Drinkwater, B. L.; Gliner, J. A.; Bedi, J. F.,
40 eds. Environmental stress: individual human adaptations. New York, NY: Academic Press;
41 pp. 125-145.
42 Folinsbee, L. J.; Bedi, J. F.; Horvath, S. M. (1980) Respiratory responses in humans repeatedly
43 exposed to low concentrations of ozone. Am. Rev. Respir. Dis. 121: 431-439.
44 Folinsbee, L. J.; McDonnell, W. F.; Horstman, D. H. (1988) Pulmonary function and symptom
45 responses after 6.6-hour exposure to 0.12 ppm ozone with moderate exercise. JAPCA 38:
46 28-35.
January 2005 6-45 DRAFT-DO NOT QUOTE OR CITE
-------
1 Folinsbee, L. I; Horstman, D. H.; Kehrl, H. R.; Harder, S.; Abdul-Salaam, S.; Ives, P. J. (1994)
2 Respiratory responses to repeated prolonged exposure to 0.12 ppm ozone. Am. J. Respir.
3 Crit. Care Med. 149: 98-105.
4 Folinsbee, L. J.; Devlin, R. B.; Robbins, M. K.; Biscardi, F. H.; Abdul-Salaam, S.; Koren, H. S.
5 (1998) Repeated exposure of humans to ozone: pulmonary function and symptom
6 responses. Research Triangle Park, NC: U.S. Environmental Protection Agency; National
7 Center for Environmental Assessment; unpublished data.
8 Foster, W. M.; Stetkiewicz, P. T. (1996) Regional clearance of solute from the respiratory
9 epithelia: 18-20 h postexposure to ozone. J. Appl. Physiol. 81: 1143-1149.
10 Foster, W. M.; Costa, D. L.; Langenback, E. G. (1987) Ozone exposure alters tracheobronchial
11 mucociliary function in humans. J. Appl. Physiol. 63: 996-1002.
12 Foster, W. M.; Wills-Karp, M.; Tankersley, C. G.; Chen, X.; Paquette, N. C. (1996) Bloodborne
13 markers in humans during multiday exposure to ozone. J. Appl. Physiol. 81: 794-800.
14 Foster, W. M.; Brown, R. H.; Macri, K.; Mitchell, C. S. (2000) Bronchial reactivity of healthy
15 subjects: 18-20 h postexposure to ozone. J. Appl. Physiol. 89: 1804-1810.
16 Fox, S. D.; Adams, W. C.; Brookes, K. A.; Lasley, B. L. (1993) Enhanced response to ozone
17 exposure during the follicular phase of the menstrual cycle. Environ. Health Perspect. 101:
18 242-244.
19 Foxcroft, W. J.; Adams, W. C. (1986) Effects of ozone exposure on four consecutive days on
•
20 work performance and VO2max. J. Appl. Physiol. 61: 960-966.
21 Frampton, M. W.; Morrow, P. E.; Torres, A.; Cox, C.; Voter, K. Z.; Utell, M. J.; Gibb, F. R.;
22 Speers, D. M. (1997) Ozone responsiveness in smokers and nonsmokers. Am. J. Respir.
23 Crit. Care Med. 155: 116-121.
24 Frank, R.; Liu, M. C.; Spannhake, E. W.; Mlynarek, S.; Macri, K.; Weinmann, G. G. (2001)
25 Repetitive ozone exposure of young adults: evidence of persistent small airway
26 dysfunction. Am. J. Respir. Crit. Care Med. 164: 1253-1260.
27 Freed, A. N.; Chou, C. L.; Fuller, S. D.; Croxton, T. L. (1996) Ozone-induced vagal reflex
28 modulates airways reactivity in rabbits. Respir. Physiol. 105: 95-102.
29 Gabrielson, E. W.; Yu, X.-Y.; Spannhake, E. W. (1994) Comparison of the toxic effects of
30 hydrogen peroxide and ozone on cultured human bronchial epithelial cells. Environ. Health
31 Perspect. 102: 972-974.
32 Gerrity, T. R.; Bennett, W. D.; Kehrl, H.; DeWitt, P. J. (1993) Mucociliary clearance of inhaled
33 particles measured at 2 h after ozone exposure in humans. J. Appl. Physiol. 74: 2984-2989.
34 Gong, H., Jr.; Bradley, P. W.; Simmons, M. S.; Tashkin, D. P. (1986) Impaired exercise
35 performance and pulmonary function in elite cyclists during low-level ozone exposure in a
36 hot environment. Am. Rev. Respir. Dis. 134: 726-733.
37 Gong, H., Jr.; Bedi, J. F.; Horvath, S. M. (1988) Inhaled albuterol does not protect against ozone
38 toxicity in nonasthmatic athletes. Arch. Environ. Health 43: 46-53.
39 Gong, H., Jr.; Shamoo, D. A.; Anderson, K. R.; Linn, W. S. (1997a) Responses of older men
40 with and without chronic obstructive pulmonary disease to prolonged ozone exposure.
41 Arch. Environ. Health 52: 18-25.
42 Gong, H., Jr.; McManus, M. S.; Linn, W. S. (1997b) Attenuated response to repeated daily
43 ozone exposures in asthmatic subjects. Arch. Environ. Health 52: 34-41.
44 Gong, H., Jr.; Wong, R.; Sarma, R. J.; Linn, W. S.; Sullivan, E. D.; Shamoo, D. A.; Anderson, K.
45 R.; Prasad, S. B. (1998) Cardiovascular effects of ozone exposure in human volunteers.
46 Am. J. Respir. Crit. Care Med. 158: 538-546.
January 2005 6-46 DRAFT-DO NOT QUOTE OR CITE
-------
1 Graham, D. E.; Koren, H. S. (1990) Biomarkers of inflammation in ozone-exposed humans:
2 comparison of the nasal and bronchoalveolar lavage. Am. Rev. Respir. Dis. 142: 152-156.
3 Graham, D.; Henderson, F.; House, D. (1988) Neutrophil influx measured in nasal lavages of
4 humans exposed to ozone. Arch. Environ. Health 43: 228-233.
5 Gwynn, R. C.; Thurston, G. D. (2001) The burden of air pollution: impacts among racial
6 minorities. Environ. Health Perspect. 109(suppl. 4): 501-506.
7 Hamilton, R. F.; Li, L.; Eschenbacher, W. L.; Szweda, L.; Holian, A. (1998) Potential
8 involvement of 4-hydroxynonenal in the response of human lung cells to ozone. Am. J.
9 Physiol. 274: L8-L16.
10 Hazbun, M. E.; Hamilton, R.; Holian, A.; Eschenbacher, W. L. (1993) Ozone-induced increases
11 in substance P and 8-epi-prostaglandin F2(X in the airways of human subjects. Am. J. Respir.
12 Cell Mol. Biol. 9: 568-572.
13 Hazucha, M. J.; Folinsbee, L. J.; Seal, E., Jr. (1992) Effects of steady-state and variable ozone
14 concentration profiles on pulmonary function. Am. Rev. Respir. Dis. 146: 1487-1493.
15 Hazucha, M. J.; Folinsbee, L. J.; Seal, E.; Bromberg, P. A. (1994) Lung function response of
16 healthy women after sequential exposures to NO2 and O3. Am. J. Respir. Crit. Care Med.
17 150: 642-647.
18 Hazucha, M. J.; Madden, M.; Pape, G.; Becker, S.; Devlin, R.; Koren, H. S.; Kehrl, H.;
19 Bromberg, P. A. (1996) Effects of cyclo-oxygenase inhibition on ozone-induced
20 respiratory inflammation and lung function changes. Eur. J. Appl. Physiol. Occup. Med.
21 73: 17-27.
22 Hazucha, M. J.; Folinsbee, L. J.; Bromberg, P. A. (2003) Distribution and reproducibility of
23 spirometric response to ozone by gender and age. J. Appl. Physiol. 95: 1917-1925.
24 Hetrick, S. M.; Gould, W. D.; Christensen, D. E. (2000) Inflight cabin ozone aboard long
25 duration C-5 airlift missions: a historical issue revisited. Aviat. Space Environ. Med. 71:
26 408-414.
27 Hiltermann, T. J. N.; Stolk, J.; Hiemstra, P. S.; Fokkens, P. H. B.; Rombout, P. J. A.; Sont, J. K.;
28 Sterk, P. J.; Dijkman, J. H. (1995) Effect of ozone exposure on maximal airway narrowing
29 in non-asthmatic and asthmatic subjects. Clin. Sci. 89: 619-624.
30 Hiltermann, T. J. N.; de Bruijne, C. R.; Stolk, J.; Zwinderman, A. H.; Spieksma, F. Th. M.;
31 Roemer, W.; Steerenberg, P. A.; Fischer, P. H.; van Bree, L.; Hiemstra, P. S. (1997)
32 Effects of photochemical air pollution and allergen exposure on upper respiratory tract
33 inflammation in asthmatics. Am. J. Respir. Crit. Care Med. 156: 1765-1772.
34 Hiltermann, J. T. N.; Lapperre, T. S.; Van Bree, L.; Steerenberg, P. A.; Brahim, J. J.; Sont, J. K.;
35 Sterk, P. J.; Hiemstra, P. S.; Stolk, J. (1999) Ozone-induced inflammation assessed in
36 sputum and bronchial lavage fluid from asthmatics: a new noninvasive tool in
37 epidemiologic studies on air pollution and asthma. Free Radical Biol. Med. 27: 1448-1454.
38 Holz, O.; Torres, R. A.; Timm, P.; Mucke, M.; Richter, K.; Koschyk, S.; Magnussen, H. (1999)
39 Ozone-induced airway inflammatory changes differ between individuals and are
40 reproducible. Am. J. Respir. Crit. Care Med. 159: 776-784.
41 Holz, O.; Mucke, M.; Paasch, K.; Bohme, S.; Timm, P.; Richter, K.; Magnussen, H.; Torres, R.
42 A. (2002) Repeated ozone exposures enhance bronchial allergen responses in subjects with
43 rhinitis or asthma. Clin. Exp. Allergy. 32: 681-689.
44 Horstman, D. H.; Folinsbee, L. J.; Ives, P. J.; Abdul-Salaam, S.; McDonnell, W. F. (1990) Ozone
45 concentration and pulmonary response relationships for 6.6-hour exposures with five hours
46 of moderate exercise to 0.08, 0.10, and0.12ppm. Am. Rev. Respir. Dis. 142: 1158-1163.
January 2005 6-47 DRAFT-DO NOT QUOTE OR CITE
-------
1 Horstman, D. H.; Ball, B. A.; Brown, J.; Gerrity, T.; Folinsbee, L. J. (1995) Comparison of
2 pulmonary responses of asthmatic and nonasthmatic subjects performing light exercise
3 while exposed to a low level of ozone. Toxicol. Ind. Health 11: 369-385.
4 Horvath, S. M.; Gliner, J. A.; Matsen-Twisdale, J. A. (1979) Pulmonary function and maximum
5 exercise responses following acute ozone exposure. Aviat. Space Environ. Med. 50:
6 901-905.
7 Horvath, S. M.; Gliner, J. A.; Folinsbee, L. J. (1981) Adaptation to ozone: duration of effect.
8 Am. Rev. Respir. Dis. 123: 496-499.
9 Housley, D. G.; Eccles, R.; Richards, R. J. (1996) Gender difference in the concentration of the
10 antioxidant uric acid in human nasal lavage. Acta Oto-Laryngol. 116: 751-754.
11 Jenkins, H. S.; Devalia, J. L.; Mister, R. L.; Bevan, A. M.; Rusznak, C.; Davies, R. J. (1999) The
12 effect of exposure to ozone and nitrogen dioxide on the airway response of atopic
13 asthmatics to inhaled allergen: dose- and time-dependent effects. Am. J. Respir. Crit. Care
14 Med. 160: 33-39.
15 load, J. P.; Kott, K. S.; Brie, J. M. (1996) The local C-fiber contribution to ozone-induced effects
16 on the isolated guinea pig lung. Toxicol. Appl. Pharmacol. 141: 561-567.
17 Torres, R.; Nowak, D.; Magnussen, H.; Speckin, P.; Koschyk, S. (1996) The effect of ozone
18 exposure on allergen responsiveness in subjects with asthma or rhinitis. Am. J. Respir.
19 Crit. Care Med. 153: 56-64.
20 Torres, R. A.; Holz, O.; Zachgo, W.; Timm, P.; Koschyk, S.; Miiller, B.; Grimminger, F.; Seeger,
21 W.; Kelly, F. J.; Dunster, C.; Frischer, T.; Lubec, G.; Waschewski, M.; Niendorf, A.;
22 Magnussen, H. (2000) The effect of repeated ozone exposures on inflammatory markers in
23 bronchoalveolar lavage fluid and mucosal biopsies. Am. J. Respir. Crit. Care Med. 161:
24 1855-1861.
25
26 Kehrl, H. R.; Hazucha, M. J.; Solic, J. J.; Bromberg, P. A. (1985) Responses of subjects with
27 chronic obstructive pulmonary disease after exposures to 0.3 ppm ozone. Am. Rev. Respir.
28 Dis. 131:719-724.
29 Kehrl, H. R.; Vincent, L. M.; Kowalsky, R. J.; Horstman, D. H.; O'Neil, J. J.; McCartney, W. H.;
30 Bromberg, P. A. (1987) Ozone exposure increases respiratory epithelial permeability in
31 humans. Am. Rev. Respir. Dis. 135: 1124-1128.
32 Kehrl, H. R.; Peden, D. B.; Ball, B. A.; Folinsbee, L. J.; Horstman, D. H. (1999) Increased
33 specific airway reactivity of persons with mild allergic asthma after 7.6 hours of exposure
34 to 0.16 ppm ozone. J. Allergy. Clin. Immunol. 104: 1198-1204.
35 Kleinman, M. T.; Mautz, W. J.; Bjarnason, S. (1999) Adaptive and non-adaptive responses in
36 rats exposed to ozone, alone and in mixtures, with acidic aerosols. Inhalation Toxicol. 11:
37 249-264.
38 Koren, H. S.; Devlin, R. B.; Graham, D. E.; Mann, R.; McGee, M. P.; Horstman, D. H.;
39 Kozumbo, W. J.; Becker, S.; House, D. E.; McDonnell, W. F.; Bromberg, P. A. (1989a)
40 Ozone-induced inflammation in the lower airways of human subjects. Am. Rev. Respir.
41 Dis. 139: 407-415.
42 Koren, H. S.; Devlin, R. B.; Graham, D. E.; Mann, R.; McDonnell, W. F. (1989b) The
43 inflammatory response in human lung exposed to ambient levels of ozone. In: Schneider,
44 T.; Lee, S. D.; Wolters, G. J. R.; Grant, L. D., eds. Atmospheric ozone research and its
45 policy implications: proceedings of the 3rd US-Dutch international symposium; May 1988;
46 Nijmegen, The Netherlands. Amsterdam, The Netherlands: Elsevier Science Publishers;
47 pp. 745-753. (Studies in environmental science 35).
January 2005 6-48 DRAFT-DO NOT QUOTE OR CITE
-------
1 Koren, H. S.; Devlin, R. B.; Becker, S.; Perez, R.; McDonnell, W. F. (1991) Time-dependent
2 changes of markers associated with inflammation in the lungs of humans exposed to
3 ambient levels of ozone. Toxicol. Pathol. 19: 406-411.
4 Kozumbo, W. 1; Hanley, N. M.; Agarwas, S.; Thomas, M. I; Madden, M. C. (1996) Products of
5 ozonized arachidonic acid potentiate the formation of DNA single strand breaks in cultured
6 human lung cells. Environ. Mol. Mutagen. 27: 185-195.
7 Kreit, J. W.; Gross, K. B.; Moore, T. B.; Lorenzen, T. I; D'Arcy, I; Eschenbacher, W. L. (1989)
8 Ozone-induced changes in pulmonary function and bronchial responsiveness in asthmatics.
9 J. Appl. Physiol. 66: 217-222.
10 Krishna, M. T.; Springall, D.; Meng, Q.-H.; Withers, N.; Macleod, D.; Biscione, G.; Frew, A.;
11 Polak, J.; Holgate, S. (1997a) Effects of ozone on epithelium and sensory nerves in the
12 bronchial mucosa of healthy humans. Am. J. Respir. Crit. Care Med. 156: 943-950.
13 Krishna, M. T.; Blomberg, A.; Biscione, G. L.; Kelly, F.; Sandstrom, T.; Frew, A.; Holgate, S.
14 (1997b) Short-term ozone exposure upregulates P-selectin in normal human airways. Am.
15 J. Respir. Crit. Care Med. 155: 1798-1803.
16 Kulle, T. J.; Sauder, L. R.; Kerr, H. D.; Farrell, B. P.; Bermel, M. S.; Smith, D. M. (1982)
17 Duration of pulmonary function adaptation to ozone in humans. Am. Ind. Hyg. Assoc. J.
18 43: 832-837.
19 Kulle, T. J.; Milman, J. H.; Sauder, L. R.; Kerr, H. D.; Farrell, B. P.; Miller, W. R. (1984)
20 Pulmonary function adaptation to ozone in subjects with chronic bronchitis. Environ. Res.
21 34: 55-63.
22 Kulle, T. J.; Sauder, L. R.; Hebel, J. R.; Chatham, M. D. (1985) Ozone response relationships in
23 healthy nonsmokers. Am. Rev. Respir. Dis. 132: 36-41.
24 Lang, D. S.; Torres, R. A.; Miicke, M.; Siegfried, W.; Magnussen, H. (1998) Interactions
25 between human bronchoepithelial cells and lung fibroblasts after ozone exposure in vitro.
26 Toxicol. Lett. 96-97: 13-24.
27 Linder, J.; Herren, D.; Monn, C.; Wanner, H.-U. (1988) Die Wirkung von Ozon auf die
28 korperliche Leistungsfahigkeit [The effect of ozone on physical activity]. Schweiz Z.
29 Sportmed. 36: 5-10.
30 Linn, W. S.; Jones, M. P.; Bachmayer, E. A.; Spier, C. E.; Mazur, S. F.; Avol, E. L.; Hackney, J.
31 D. (1980) Short-term respiratory effects of polluted ambient air: a laboratory study of
32 volunteers in a high-oxidant community. Am. Rev. Respir. Dis. 121: 243-252.
33 Linn, W. S.; Fischer, D. A.; Medway, D. A.; Anzar, U. T.; Spier, C. E.; Valencia, L. M.; Venet,
34 T. G.; Hackney, J. D. (1982a) Short-term respiratory effects of 0.12 ppm ozone exposure in
35 volunteers with chronic obstructive pulmonary disease. Am. Rev. Respir. Dis. 125:
36 658-663.
37 Linn, W. S.; Medway, D. A.; Anzar, U. T.; Valencia, L. M.; Spier, C. E.; Tsao, F. S.-D.; Fischer,
38 D. A.; Hackney, J. D. (1982b) Persistence of adaptation to ozone in volunteers exposed
39 repeatedly for six weeks. Am. Rev. Respir. Dis. 125: 491-495.
40 Linn, W. S.; Shamoo, D. A.; Venet, T. G.; Spier, C. E.; Valencia, L. M.; Anzar, U. T.; Hackney,
41 J. D. (1983a) Response to ozone in volunteers with chronic obstructive pulmonary disease.
42 Arch. Environ. Health 38: 278-283.
43
44 Linn, W. S.; Avol, E. L.; Hackney, J. D. (1983b) Effects of ambient oxidant pollutants on
45 humans: a movable environmental chamber study. In: Lee, S. D.; Mustafa, M. G.;
46 Mehlman, M. A., eds. International symposium on the biomedical effects of ozone and
47 related photochemical oxidants; March 1982; Pinehurst, NC. Princeton, NJ: Princeton
January 2005 6-49 DRAFT-DO NOT QUOTE OR CITE
-------
1 Scientific Publishers, Inc.; pp. 125-137. (Advances in modern environmental toxicology: v.
2 5).
3 Linn, W. S.; Avol, E. L.; Shamoo, D. A.; Spier, C. E.; Valencia, L. M.; Venet, T. G.; Fischer, D.
4 A.; Hackney, J. D. (1986) A dose-response study of healthy, heavily exercising men
5 exposed to ozone at concentrations near the ambient air quality standard. Toxicol. Ind.
6 Health 2: 99-112.
7 Linn, W. S.; Gong, H., Jr.; Shamoo, D. A.; Anderson, K. R.; Avol, E. L. (1997) Chamber
8 exposures of children to mixed ozone, sulfur dioxide, and sulfuric acid. Arch. Environ.
9 Health 52: 179-187.
10 Liu, L.; Leech, J. A.; Urch, R. B.; Silverman, F. S. (1997) In vivo salicylate hyroxylation: a
11 potential biomarker for assessing acute ozone exposure and effects in humans. Am. J.
12 Respir. Crit. Care Med. 156: 1405-1412.
13 Liu, L.; Leech, J. A.; Urch, R. B.; Poon, R.; Zimmerman, B.; Kubay, J. M.; Silverman, F. S.
14 (1999) A comparison of biomarkers of ozone exposure in human plasma, nasal lavage, and
15 sputum. Inhalation Toxicol. 11: 657-674.
16 Marthan, R.; Roux, E.; Savineau, J.-P. (1996) Human bronchial smooth muscle responsiveness
17 after in vitro exposure to oxidizing pollutants. Cell Biol. Toxicol. 12: 245-249.
18 McBride, D. E.; Koenig, J. Q.; Luchtel, D. L.; Williams, P. V.; Henderson, W. R., Jr. (1994)
19 Inflammatory effects of ozone in the upper airways of subjects with asthma. Am. J. Respir.
20 Crit. Care Med. 149: 1192-1197.
21 McDonnell, W. F. (1996) Individual variability in human lung function responses to ozone
22 exposure. Environ. Toxicol. Pharmacol. 2: 171-175.
23 McDonnell, W. F.; Horstman, D. H.; Hazucha, M. J.; Seal, E., Jr.; Haak, E. D.; Salaam, S. A.;
24 House, D. E. (1983) Pulmonary effects of ozone exposure during exercise: dose-response
25 characteristics. J. Appl. Physiol.: Respir. Environ. Exercise Physiol. 54: 1345-1352.
26 McDonnell, W. F., Ill; Horstman, D. H.; Abdul-Salaam, S.; House, D. E. (1985a)
27 Reproducibility of individual responses to ozone exposure. Am. Rev. Respir. Dis. 131:
28 36-40.
29 McDonnell, W. F., Ill; Chapman, R. S.; Leigh, M. W.; Strope, G. L.; Collier, A. M. (1985b)
30 Respiratory responses of vigorously exercising children to 0.12 ppm ozone exposure. Am.
31 Rev. Respir. Dis. 132: 875-879.
32 McDonnell, W. F.; Kehrl, H. R.; Abdul-Salaam, S.; Ives, P. J.; Folinsbee, L. J.; Devlin, R. B.;
33 O'Neil, J. J.; Horstman, D. H. (1991) Respiratory response of humans exposed to low
34 levels of ozone for 6.6 hours. Arch. Environ. Health 46: 145-150.
35 McDonnell, W. F.; Muller, K. E.; Bromberg, P. A.; Shy, C. M. (1993) Predictors of individual
36 differences in acute response to ozone exposure. Am. Rev. Respir. Dis. 147: 818-825.
37 McDonnell, W. F.; Stewart, P. W.; Andreoni, S.; Seal, E., Jr.; Kehrl, H. R.; Horstman, D. H.;
38 Folinsbee, L. J.; Smith, M. V. (1997) Prediction of ozone-induced FEVj changes: effects of
39 concentration, duration, and ventilation. Am. J. Respir. Crit. Care Med. 156: 715-722.
40 McDonnell, W. F.; Stewart, P. W.; Smith, M. V.; Pan, W. K.; Pan, J. (1999) Ozone-induced
41 respiratory symptoms: exposure-response models and association with lung function. Eur.
42 Respir. J. 14: 845-853.
43 McGee, M. P.; Devlin, R.; Saluta, G.; Koren, H. (1990) Tissue factor and factor VII messenger
44 RNAs in human alveolar macrophages: effects of breathing ozone. Blood 75: 122-127.
45 McKenzie, D. C.; Stirling, D. R.; Fadl, S.; Allen, M. (1987) The effects of salbutamol on
46 pulmonary function in cyclists exposed to ozone: a pilot study. Can. J. Sport Sci. 12: 46-48.
January 2005 6-50 DRAFT-DO NOT QUOTE OR CITE
-------
1 McKinnon, K. P.; Madden, M. C.; Noah, T. L.; Devlin, R. B. (1993) In vitro ozone exposure
2 increases release of arachidonic acid products from a human bronchial epithelial cell line.
3 Toxicol. Appl. Pharmacol. 118: 215-223.
4 Melton, C. E. (1990) Airliner cabin ozone: an updated review. Washington, DC: Federal
5 Aviation Administration, Office of Aviation Medicine; report no. DOT/FAA/AM-89/13.
6 Available from: NTIS, Springfield, VA; AD-A219 264.
7 Messineo, T. D.; Adams, W. C. (1990) Ozone inhalation effects in females varying widely in
8 lung size: comparison with males. J. Appl. Physiol. 69: 96-103.
9 Michelson, P. H.; Dailey, L.; Devlin, R. B.; Peden, D. B. (1999) Ozone effects on the
10 immediate-phase response to allergen in the nasal airways of allergic asthmatic subjects.
11 Otolaryngol. Head Neck Surg. 120: 225-232.
12 Mohammed, S. P.; Higenbottam, T. W.; Adcock, J. J. (1993) Effects of aerosol-applied
13 capsaicin, histamine and prostaglandin-E2 on airway sensory receptors of anaesthetized
14 cats. J. Physiol. Lond. 46: 951-966.
15 Molfmo, N. A.; Wright, S. C.; Katz, L; Tarlo, S.; Silverman, F.; McClean, P. A.; Szalai, J. P.;
16 Raizenne, M.; Slutsky, A. S.; Zamel, N. (1991) Effect of low concentrations of ozone on
17 inhaled allergen responses in asthmatic subjects. Lancet 338(8761): 199-203.
18 Mudway, I. S.; Kelly, F. J. (2000) Ozone and the lung: a sensitive issue. Mol. Aspects. Med. 21:
19 1-48.
20 Mudway, I. S.; Blomberg, A.; Frew, A. J.; Holgate, S. T.; Sandstrom, T.; Kelly, F. J. (1999)
21 Antioxidant consumption and repletion kinetics in nasal lavage fluid following exposure of
22 healthy human volunteers to ozone. Eur. Respir. J. 13: 1429-1438.
23 Mudway, I. S.; Stenfors, N.; Blomberg, A.; Helleday, R.; Dunster, C.; Marklund, S. L.; Frew, A.
24 J.; Sandstrom, T.; Kelly, F. J. (2001) Differences in basal airway antioxidant
25 concentrations are not predictive of individual responsiveness to ozone: a comparison of
26 healthy and mild asthmatic subjects. Free Radical Biol. Med. 31: 962-974.
27 Nagda, N. L.; Fortmann, R. C.; Koontz, M. D.; Baker, S. R.; Ginevan, M. E. (1989) Airliner
28 cabin environment: contaminant measurements, health risks, and mitigation options.
29 Washington, DC: U.S. Department of Transportation, Office of the Secretary. Available
30 from: NTIS, Springfield, VA; PB91-159384.
31 Nayak, A. S. (2003) The asthma and allergic rhinitis link. Allergy and Asthma Proc.
32 24: 395-403.
33 Neas, L. M.; Dockery, D. W.; Koutrakis, P.; Tollerud, D. J.; Speizer, F. E. (1995) The
34 association of ambient air pollution with twice daily peak expiratory flow rate
35 measurements in children. Am. J. Epidemiol. 141: 111-122.
36 Newson, E. J.; Krishna, M. T.; Lau, L. C. K.; Howarth, P. H.; Holgate, S. T.; Frew, A. J. (2000)
37 Effects of short-term exposure to 0.2 ppm ozone on biomarkers of inflammation in sputum,
38 exhaled nitric oxide, and lung function in subjects with mild atopic asthma. J. Occup.
39 Environ. Med. 42: 270-277.
40 Nightingale, J. A.; Rogers, D. F.; Chung, K. F.; Barnes, P. J. (2000) No effect of inhaled
41 budesonide on the response to inhaled ozone in normal subjects. Am. J. Respir. Crit. Care
42 Med. 161:479-486.
43 Nikasinovic, L.; Momas, L; Seta, N. (2003) Nasal epithelial and inflammatory response to ozone
44 exposure: a review of laboratory-based studies published since 1985. J. Toxicol. Environ.
45 HealthB 6: 521-568.
January 2005 6-51 DRAFT-DO NOT QUOTE OR CITE
-------
1 Passannante, A. N.; Hazucha, M. J.; Bromberg, P. A.; Seal, E.; Folinsbee, L.; Koch, G. (1998)
2 Nociceptive mechanisms modulate ozone-induced human lung function decrements. J.
3 Appl. Physiol. 85: 1863-1870.
4 Peden, D. B. (2001a) Air pollution in asthma: effect of pollutants on airway inflammation. Ann.
5 Allergy Asthma Immunol. 87(suppl. 3): 12-17.
6 Peden, D. B. (200Ib) Effect of pollutants in rhinitis. Curr. Allergy Asthma Rep. 1: 242-246.
7 Peden, D. B.; Setzer, R. W., Jr.; Devlin, R. B. (1995) Ozone exposure has both a priming effect
8 on allergen-induced responses and an intrinsic inflammatory action in the nasal airways of
9 perennially allergic asthmatics. Am. J. Respir. Crit. CareMed. 151: 1336-1345.
10 Peden, D. B.; Boehlecke, B.; Horstman, D.; Devlin, R. (1997) Prolonged acute exposure to 0.16
11 ppm ozone induces eosinophilic airway inflammation in asthmatic subjects with allergies.
12 J. Allergy Clin. Immunol. 100: 802-808.
13 Riediker, M.; Monn, C.; Roller, T.; Stahel, W. A.; Wuthrich, B. (2001) Air pollutants enhance
14 rhinoconjunctivitis symptoms in pollen-allergic individuals. Ann. Allergy Asthma
15 Immunol. 87: 311-318.
16 Rigas, M. L.; Ben-Jebria, A.; Ultman, J. S. (1997) Longitudinal distribution of ozone absorption
17 in the lung: effects of nitrogen dioxide, sulfur dioxide, and ozone exposures. Arch.
18 Environ. Health 52: 173-178.
19 Romieu, L; Sienra-Monge, J. J.; Ramirez-Aguilar, M.; Moreno-Macias, H.; Reyes-Ruiz, N. L;
20 Estela del Rio-Navarro, B.; Hernandez-Avila, M.; London, S. J. (2004) Genetic
21 polymorphism ofGSTMl and antioxidant supplementation influence lung function in
22 relation to ozone exposure in asthmatic children in Mexico City. Thorax 59: 8-10.
23 Samet, J. M.; Hatch, G. E.; Horstman, D.; Steck-Scott, S.; Arab, L.; Bromberg, P. A.; Levine,
24 M.; McDonnell, W. F.; Devlin, R. B. (2001) Effect of antioxidant supplementation on
25 ozone-induced lung injury in human subjects. Am. J. Respir. Crit. Care Med. 164: 819-825.
26 Scannell, C.; Chen, L.; Aris, R. M.; Tager, L; Christian, D.; Ferrando, R.; Welch, B.; Kelly, T.;
27 Balmes, J. R. (1996) Greater ozone-induced inflammatory responses in subjects with
28 asthma. Am. J. Respir. Crit. Care Med. 154: 24-29.
29 Schelegle, E. S.; Adams, W. C. (1986) Reduced exercise time in competitive simulations
30 consequent to low level ozone exposure. Med. Sci. Sports Exercise 18: 408-414.
31 Schelegle, E. S.; Adams, W. C.; Siefkin, A. D. (1987) Indomethacin pretreatment reduces
32 ozone-induced pulmonary function decrements in human subjects. Am. Rev. Respir. Dis.
33 136: 1350-1354.
34 Schelegle, E. S.; Siefkin, A. D.; McDonald, R. J. (1991) Time course of ozone-induced
35 neutrophilia in normal humans. Am. Rev. Respir. Dis. 143: 1353-1358.
36 Schelegle, E. S.; Carl, M. L.; Coleridge, H. M.; Coleridge, J. C. G.; Green, J. F. (1993)
37 Contribution of vagal afferents to respiratory reflexes evoked by acute inhalation of ozone
38 in dogs. J. Appl. Physiol. 74: 2338-2344.
39 Schelegle, E. S.; Eldridge, M. W.; Cross, C. E.; Walby, W. F.; Adams, W. C. (2001) Differential
40 effects of airway anesthesia on ozone-induced pulmonary responses in human subjects.
41 Am. J. Respir. Crit. CareMed. 163: 1121-1127.
42 Schonfeld, B. R.; Adams, W. C.; Schelegle, E. S. (1989) Duration of enhanced responsiveness
43 upon re-exposure to ozone. Arch. Environ. Health 44: 229-236.
44 Schwartz, L. W.; Dungworth, D. L.; Mustafa, M. G.; Tarkington, B. K.; Tyler, W. S. (1976)
45 Pulmonary responses of rats to ambient levels of ozone: effects of 7-day intermittent or
46 continuous exposure. Lab. Invest. 34: 565-578.
January 2005 6-52 DRAFT-DO NOT QUOTE OR CITE
-------
1 Seal, E., Jr.; McDonnell, W. F.; House, D. E.; Salaam, S. A.; Dewitt, P. I; Butler, S. O.; Green,
2 J.; Raggio, L. (1993) The pulmonary response of white and black adults to six
3 concentrations of ozone. Am. Rev. Respir. Dis. 147: 804-810.
4 Seal, E., Jr.; McDonnell, W. F.; House, D. E. (1996) Effects of age, socioeconomic status, and
5 menstrual cycle on pulmonary response to ozone. Arch. Environ. Health 51: 132-137.
6 Seltzer, J.; Bigby, B. G.; Stulbarg, M.; Holtzman, M. J.; Nadel, J. A.; Ueki, I. F.; Leikauf, G D.;
7 Goetzl, E. J.; Boushey, H. A. (1986) O3-induced change in bronchial reactivity to
8 methacholine and airway inflammation in humans. J. Appl. Physiol. 60: 1321-1326.
9 Solic, J. J.; Hazucha, M. J.; Bromberg, P. A. (1982) The acute effects of 0.2 ppm ozone in
10 patients with chronic obstructive pulmonary disease. Am. Rev. Respir. Dis. 125: 664-669.
11 Solway, J.; Left, A. R. (1991) Sensory neuropeptides and airway function. J. Appl. Physiol. 71:
12 2077-2087.
13 Spannhake, E. W.; Reddy, S. P. M.; Jacoby, D. B.; Yu, X.-Y.; Saatian, B.; Tian, J. (2002)
14 Synergism between rhinovirus infection and oxidant pollutant exposure enhances airway
15 epithelial cell cytokine production. Environ. Health Perspect. 110: 665-670.
16 Stenfors, N.; Pourazar, J.; Blomberg, A.; Krishna, M. T.; Mudway, I; Helleday, R.; Kelly, F. J.;
17 Frew, A. J.; Sandstrom, T. (2002) Effect of ozone on bronchial mucosal inflammation in
18 asthmatic and healthy subjects. Respir. Med. 96: 352-358.
19 Tepper, J. S.; Costa, D. L.; Lehmann, J. R.; Weber, M. F.; Hatch, G. E. (1989) Unattenuated
20 structural and biochemical alterations in the rat lung during functional adaptation to ozone.
21 Am. Rev. Respir. Dis. 140: 493-501.
22 Tepper, J. S.; Wiester, M. J.; Weber, M. F.; Menache, M. G. (1990) Measurements of
23 cardiopulmonary response in awake rats during acute exposure to near-ambient
24 concentrations of ozone. Fundam. Appl. Toxicol. 10: 7-15.
25 Torres, A.; Utell, M. J.; Morow, P. E.; Voter, K. Z.; Whitin, J. C.; Cox, C.; Looney, R. J.; Speers,
26 D. M.; Tsai, Y.; Frampton, M. W. (1997) Airway inflammation in smokers and
27 nonsmokers with varying responsiveness to ozone. Am. J. Respir. Crit. Care Med. 156:
28 728-736.
29 Trenga, C. A.; Koenig, J. Q.; Williams, P. V. (2001) Dietary antioxidants and ozone-induced
30 bronchial hyperresponsiveness in adults with asthma. Arch. Environ. Health 56: 242-249.
31 U.S. Environmental Protection Agency. (1986) Air quality criteria for ozone and other
32 photochemical oxidants. Research Triangle Park, NC: Office of Health and Environmental
33 Assessment, Environmental Criteria and Assessment Office; report nos.
34 EPA-600/8-84-020aF-eF. 5v. Available from: NTIS, Springfield, VA; PB87-142949.
35 U.S. Environmental Protection Agency. (1996) Air quality criteria for ozone and related
36 photochemical oxidants. Research Triangle Park, NC: Office of Research and
37 Development; report nos. EPA/600/AP-93/004aF-cF. 3v. Available from: NTIS,
38 Springfield, VA; PB96-185582, PB96-185590, and PB96-185608. Available:
39 www.epa.gov/ncea/ozone.htm.
40 Ultman, J. S.; Ben-Jebria, A.; Arnold, S. F. (2004) Uptake distribution of ozone in human lungs:
41 intersubject variability in physiologic response. Boston, MA: Health Effects Institute.
42 Vagaggini, B.; Carnevali, S.; Macchioni, P.; Taccola, M.; Fornai, E.; Bacci, E.; Bartoli, M. L.;
43 Cianchetti, S.; Dente, F. L.; Di Franco, A.; Giannini, D.; Paggiaro. P. L. (1999) Airway
44 inflammatory response to ozone in subjects with different asthma severity. Eur. Respir. J.
45 13:274-280.
46 Vagaggini, B.; Taccola, M.; Conti, L; Carnevali, S.; Cianchetti, S.; Bartoli, M. L.; Bacci, E.;
47 Dente, F. L.; Di Franco, A.; Giannini, D.; Paggiaro, P. L. (2001) Budesonide reduces
January 2005 6-53 DRAFT-DO NOT QUOTE OR CITE
-------
1 neutrophilic but not functional airway response to ozone in mild asthmatics. Am. J. Respir.
2 Crit. Care Med. 164: 2172-2176.
3 Vagaggini, B.; Taccola, M.; Clanchetti, S.; Carnevali, S.; Bartoli, M. L.; Bacci, E.; Dente, F. L.;
4 Di Franco, A.; Giannini, D.; Paggiaro, P. L. (2002) Ozone exposure increases eosinophilic
5 airway response induced by previous allergen challenge. Am. J. Respir. Crit. Care Med.
6 166: 1073-1077.
7 Voter, K. Z.; Whitin, J. C.; Torres, A.; Morrow, P. E.; Cox, C.; Tsai, Y.; Utell, M. J.; Frampton,
8 M. W. (2001) Ozone exposure and the production of reactive oxygen species by
9 bronchoalveolar cells in humans. Inhalation Toxicol. 13: 465-483.
10 Wang, G.; Umstead, T. M.; Phelps, D. S.; Al-Mondhiry, H.; Floros, J. (2002) The effect of ozone
11 exposure on the ability of human surfactant protein A variants to stimulate cytokine
12 production. Environ. Health Perspect. 110: 79-84.
13 Weinmann, G. G.; Weidenbach-Gerbase, M.; Foster, W. M.; Zacur, H.; Frank, R. (1995)
14 Evidence for ozone-induced small-airway dysfunction: lack of menstrual-cycle and gender
15 effects. Am. J. Respir. Crit. Care Med. 152: 988-996.
16 Weymer, A. R.; Gong, H., Jr.; Lyness, A.; Linn, W. S. (1994) Pre-exposure to ozone does not
17 enhance or produce exercise-induced asthma. Am. J. Respir. Crit. Care Med. 149:
18 1413-1419.
19 Yeadon, M.; Wilkinson, D.; Darley-Usmar, V.; O'Leary, V. J.; Payne, A. N. (1992) Mechanisms
20 contributing to ozone-induced bronchial hyperreactivity in guinea-pigs. Pulm. Pharmacol.
21 5:39-50.
22 Ying, R. L.; Gross, K. B.; Terzo, T. S.; Eschenbacher, W. L. (1990) Indomethacin does not
23 inhibit the ozone-induced increase in bronchial responsiveness in human subjects. Am.
24 Rev. Respir. Dis. 142: 817-821.
25 Yu, M.; Pinkerton, K. E.; Witschi, H. (2002) Short-term exposure to aged and diluted sidestream
26 cigarette smoke enhances ozone-induced lung injury in B6C3F1 mice. Toxicol. Sci. 65:
27 99-106.
28 Zhang, L.-Y.; Levitt, R. C.; Kleeberger, S. R. (1995) Differential susceptibility to ozone-induced
29 airways hyperreactivity in inbred strains of mice. Exp. Lung Res. 21: 503-518.
30
31
January 2005 6-54 DRAFT-DO NOT QUOTE OR CITE
-------
i ANNEX AX6. CONTROLLED HUMAN EXPOSURE
2 STUDIES OF OZONE AND RELATED
3 PHOTOCHEMICAL OXIDANTS
4
5
6 AX6.1 INTRODUCTION
7 In the previous chapter, results of ozone (O3) studies in laboratory animals and in vitro test
8 systems were presented. The extrapolation of results from animal studies is one mechanism by
9 which information on potential adverse human health effects from exposure to O3 is obtained.
10 More direct evidence of human health effects due to O3 exposure can be obtained through
11 controlled human exposure studies of volunteer subjects or through field and epidemiologic
12 studies of populations exposed to ambient O3. Controlled human exposure studies, discussed in
13 this chapter, typically use fixed concentrations of O3 under carefully regulated environmental
14 conditions and subject activity levels.
15 Most of the scientific information selected for review and evaluation in this chapter comes
16 from the literature published since 1996 which, in addition to further study of physiological
17 pulmonary responses and respiratory symptoms, has focused on mechanisms of inflammation
18 and cellular responses to injury induced by O3 inhalation. Older studies are discussed where
19 only limited new data are available and where new and old data are conflicting. The reader is
20 referred to both the 1986 and 1996 Air Quality Criteria documents (U.S. Environmental
21 Protection Agency, 1986, 1996) for a more extensive discussion of older studies. Summary
22 tables of the relevant O3 literature are included for each of the major subsections.
23 In summarizing the human health effects literature, changes from control are described if
24 statistically significant at a probability (p) value less than 0.05, otherwise trends are noted
25 as such.
26
January 2005 AX6-1 DRAFT-DO NOT QUOTE OR CITE
-------
1 AX6.2 PULMONARY FUNCTION EFFECTS OF OZONE EXPOSURE IN
2 HEALTHY SUBJECTS
3 AX6.2.1 Introduction
4 The responses observed in young healthy non-smoking human adults exposed to ambient
5 O3 concentrations include decreased inspiratory capacity; mild bronchoconstriction; rapid,
6 shallow breathing pattern during exercise; and symptoms of cough and pain on deep inspiration.
7 In addition, O3 has been shown to result in airway hyperresponsiveness as demonstrated by an
8 increased physiological response to a nonspecific bronchoconstrictor, as well as airway injury
9 and inflammation assessed via bronchoalveolar lavage and biopsy. Reflex inhibition of
10 inspiration and consequent decrease in inspiratory capacity results in a decrease in forced vital
11 capacity (FVC) and total lung capacity (TLC) and, in combination with mild
12 bronchoconstriction, contributes to a decrease in the forced expiratory volume in 1 s (FEVj).
13 Given that both FEVj and FVC are subject to decrease with O3 exposures, changes in the ratio
14 (FEVj/FVC) become difficult to interpret and so are not discussed.
15 The majority of controlled human studies have investigated the effects of exposure to
16 variable O3 concentrations in healthy subjects performing continuous exercise (CE) or
17 intermittent exercise (IE) for variable periods of time. These studies have several important
18 limitations: (1) the ability to study only short-term, acute effects; (2) the inability to link short-
19 term effects with long-term consequences; (3) the use of a small number of volunteers that may
20 not be representative of the general population; and (4) the statistical limitations associated with
21 the small sample size. Nonetheless, studies reviewed in the 1996 EPA criteria document (U.S.
22 Environmental Protection Agency, 1996) provided a large body of data describing the effects
23 and dose-response characteristics of O3 as function of O3 concentration (C), minute ventilation
24 (VE), and duration or time (T) of exposure. In most of these studies, subjects were exposed to
25 O3 and to filtered air (FA [reported as 0 ppm O3]) as a control. The most salient observations
26 from these studies were: (1) healthy subjects exposed to O3 concentrations > 0.08 ppm develop
27 significant reversible, transient decrements in pulmonary function if VE or T are increased
28 sufficiently, (2) there is a large degree of intersubject variability in physiologic and symptomatic
29 responses to O3 and these responses tend to be reproducible within a given individual over a
30 several months period, and (3) subjects exposed repeated to O3 over several days develop a
January 2005 AX6-2 DRAFT-DO NOT QUOTE OR CITE
-------
1 tolerance to successive exposures, as demonstrated by an attention of responses, which is lost
2 after about a week without exposure.
3 In this section, the effects of single O3 exposures of 1- to 8-h in duration on pulmonary
4 function in healthy nonsmoking subjects are examined by reviewing studies that investigate:
5 (1) the O3 exposure-response relationship; (2) intersubject variability, individual sensitivity, and
6 the association between responses; and (3) mechanisms of pulmonary function responses and the
7 relationship between tissue-level events and functional responses. Discussion will largely be
8 limited to studies published subsequent to the 1996 EPA criteria document (U.S. Environmental
9 Protection Agency, 1996)
10
11 AX6.2.2 Acute Ozone Exposures for Up to 2 Hours
12 At-Rest Exposures. Exposure studies investigating the effects of O3 exposures on sedentary
13 subjects were discussed in the 1986 EPA criteria document (U.S. Environmental Protection
14 Agency, 1986). The lowest O3 concentration at which significant reductions in FVC and FEVj
15 were reported was 0.5 ppm (Folinsbee et al., 1978; Horvath et al., 1979). Averaging there results
16 of these two studies and correcting for FA responses, exposing resting young adults (n=23,
17 age=22) to 0.5 ppm O3 results in an -4% reduction in FVC and an -7% reduction FEVj. At
18 lower O3 concentrations of 0.25 to 0.3 ppm, resting exposures did not significantly affect lung
19 function.
20 Exposures with Exercise. Collectively, the studies reviewed in the 1996 EPA criteria
21 document (U.S. Environmental Protection Agency, 1996) demonstrated that healthy young
22 adults performing moderate to heavy IE or CE of 1 to 2.5 h duration, exposed to 0.12 to
23 0.18 ppm O3 experienced statistically significant decrements in pulmonary function and
24 respiratory symptoms. As an example, 2 hr exposures to 0.12 and 0.18 ppm O3 during heavy IE
25 (exercise VE = 65 L/min) have resulted in FEVj decrements of 2.0 ± 0.8% (mean ± SE; n = 40)
26 and 9.5 ± 1.1% (n = 89), respectively (McDonnell and Smith, 1994). Significant decrements in
27 pulmonary function have been reported in heavily exercising healthy adults exposed for 1 h with
28 CE at O3 concentrations of 0.12 ppm (Gong et al., 1986), 0.16 ppm (Avol et al., 1984), and
29 0.2 ppm (Adams and Schelegle, 1983; Folinsbee et al., 1984).
30 In an attempt to describe O3 dose-response characteristics, many investigators modeled
31 acute responses as a function of total inhaled O3 dose (C * T x VE), which was found to be a
January 2005 AX6-3 DRAFT-DO NOT QUOTE OR CITE
-------
1 better predictor of response than O3 concentration, VE, or T of exposure, alone. In an analysis of
2 6 studies with 1 to 2 h exposures to between 0.12 and 0.18 ppm O3 with exercise, Folinsbee et al.
3 (1988) reported a good correlation (r = 0.81) between total inhaled O3 dose and FEVj
4 decrements. For a given exposure duration, total inhaled O3 dose can be increased by increases
5 in C and/or VE . In exposures of fixed duration, results of several studies suggested that O3
6 concentration was a more important predictor of response or explained more of the variability in
7 response than VE (Adams et al., 1981; Folinsbee et al, 1978; Hazucha, 1987). Based on a review
8 of previously published studies, Hazucha (1987) noted that relative to the FEVj decrement
9 occurring at a given C and VE, doubling C (e.g., from 0.1 to 0.2 ppm) would increase the FEVj
10 decrement by 400%, whereas doubling the VE (e.g., from an exercise VE of 20 to 40 L/min)
11 which would only increase the FEVj decrement by 190%. Thus, C appears to have a greater
12 affect than VE on FEVj responses even when total inhaled O3 doses are equivalent.
13 New studies (i.e., not reviewed in the 1996 EPA criteria document) that provide
14 spirometric responses for up to 2 h exposures are summarized in Table AX6-1. Most of these
15 newer studies have investigated mechanisms affecting responses, inflammation, and/or effects in
16 diseased groups versus healthy adults, accordingly their findings may be summarized differently
17 in several sections of this chapter. Rather than a FA exposure, some of these studies use O3
18 exposures with placebo as a control. Studies appearing in Table 1, but not discussed in this
19 section, are discussed in other sections of this chapter as indicated within the table.
20 McDonnell et al. (1997) pooled the results of eight studies entailing 485 healthy male
21 subjects exposed for 2 h on one occasion to one of six O3 concentrations (0.0, 0.12, 0.18, 0.24,
22 0.30, or 0.40 ppm) at rest or one of two levels of IE (VE of 25 and 35 L/min/m2 BSA). FEVj
23 was measured preexposure, after 1 h of exposure, and immediately postexposure. Decrements in
24 FEVj were modeled by sigmoid-shaped curve as a function of subject age, O3 concentration, VE,
25 and T. The modeled decrements reach a plateau with increasing T and dose rate (C * VE). That
26 is, for a given O3 concentration, exercise VE level, and after a certain length of exposure, the
27 FEVj response tends not to increase further with increasing duration of exposure. The modeled
28 FEVj responses increased with C x VE and T, decreased with subject age, but were only
29 minimally affected by body size corrections to VE. Fitted and experimental FEVj decrements
January 2005 AX6-4 DRAFT-DO NOT QUOTE OR CITE
-------
Table AX6-1. Controlled Exposure of Healthy Humans to Ozone for 1 to 2 Hours during Exercise3
to
o
o
X
H
6
o
o
H
O
Ozone
Concentration11
ppm
0.0
0.4
0.0
0.2
0.0
0.2
0.0
0.12
0.18
0.24
0.30
0.40
0.4
ug/m3
0
784
0
392
0
392
0
235
353
471
589
784
784
Exposure Duration
and Activity
2hIE
4x15 min
on bicycle,
VE = 30 L/min
2hIE
4x15 min
at VE = 20
L/min/m2 BSA
2hIE
4x15 min
at VE = 20
L/min/m2 BSA
2 h rest or IE
(4x15 min
at VE = 25 or 35
L/min/m2 BSA)
2hIE
20 min mild-mod.
exercise,
1 0 min rest
Number and
Exposure Gender of Subject
Conditions' Subjects Characteristics
NA 5 M, 4 F Healthy adults
25 ± 2 years old
6 M, 7 F Mild atopic asthmatics
22 ± 0.7 years old
20 °C 8 M, 5 F Healthy NS
50% RH median age 23 years
20 °C 10 M, 12 F Healthy NS
50% RH mean age 24 years
22 °C 485 M (each Healthy NS
40% RH subject exposed 1 8 to 36 years old
at one activity mean age 24 years
level to one O3
concentration)
NA 4 M, 5 F Healthy NS
30 ± 3 years old
Observed Effect(s)
O3-induced reductions in FVC (12%, 10%) and FEVj (13%, 11%)
for asthmatic and healthy subjects. Significant reductions in mid-
flows in both asthmatics and healthy subjects. Indomethacin
pretreatment significantly decreased FVC and FEV; responses to
O3 in healthy but not asthmatic subjects. See Section AX6.3. 2 and
Tables AX6-3 andAX6-13.
Median O3-induced decrements of 70 mL, 190 mL, and 400 mL/s
in FVC, FEVj, and FEF25.75, respectively. Spirometric responses
not predicted of inflammatory responses. See Sections AX6.2. 5. 2,
AX6.5.6, andAX6.9.3 andTableAX6-12.
Significant O3-induced decrement in FEVj immediately post-
exposure but not significantly different from baseline 2 h later.
No correlation between Clara cell protein (CC16) and FEV;
decrement. CC16 levels, elevated by O3 exposure, remained
high at 6 h post-exposure, but returned to baseline by 18 h
postexposure. See Table AX6-12
Statistical analysis of 8 experimental chamber studies conducted
between 1980 and 1993 by the U.S. EPA in Chapel Hill, NC.
Decrement in FEVj described by sigmoid-shaped curve as a
function of subject age, O3 concentration, VB and time. Response
decreased with age, was minimally affected by body size
corrections, and was not more sensitive to O3 concentration
than VE. Also see Section AX6.5
Subjects previously in Nightingale et al. (2000) study. Placebo-
control: Immediately postexposure decrements in FVC (9%) and
FEVj (14%) relative to pre-exposure values. FEVj decrement
only 9% at 1 hr postexposure. By 3 h postexposure, recovery in
FVC to 97% and FEV; to 98% of preexposure values. Significant
increases in 8-isoprostane at 4 h postexposure. Budesonide for
2 wk prior to exposure did not affect responses.
Reference
Alexis et al.
(2000)
Blomberg
etal. (1999)
Blomberg
et al. (2003)
McDonnell
etal. (1997)
Montuschi
et al. (2002)
o
HH
H
W
-------
Table AX6-1 (con't). Controlled Exposure of Healthy Humans to Ozone for 1 to 2 Hours during Exercise"
to
o
o
X
Oi
H
6
o
o
H
O
Ozone
Concentration11
ppm ug/m3 and Activity Conditions'
0.0 392 2hIE 20 °C
0.2 4xl5min 50% RH
at VE = 20
L/min/m2 BSA
0.4 784 2hIE NA
20 min mild-mod.
exercise,
10 min rest
0.0 784 2hIE 21 °C
0.4 4x15 min at 40% RH
VE =18 L/min/m2
BSA
2 exposures:
25% subjects
exposed to air-air,
75% to O3-O3
0.0 784 2hIE 20 °C
0.4 4x15 min 40% RH
at VE = 20
L/min/m2 BSA
0.0 490 1 h CE NA
0.25 VE = 30 L/min Face mask
exposure
Number and
Gender of
Subjects
6M, 9F
9M, 6F
6M, 9F
Weak
re spenders
7 M, 13F
Strong
re spenders
21 M, 21 F
Placebo group
15 M, 1 F
Antioxidant
group
13 M, 2F
32 M, 28 F
Subject
Characteristics
Healthy adults
24 years old
Mild asthmatics
29 years old
Healthy NS
mean age ~3 1 years
Healthy NS
20 to 59 years old
Healthy NS
mean age 27 years
Healthy NS
22.6 ± 0.6 years old
Observed Effect(s)
O3-induced FEVl decrement (8%, healthy adults; 3% asthmatics)
and PMN increase (20.6%, healthy adults; 15.2% asthmatics).
Primary goal was to investigate relationship between antioxidant
defenses and O3 responses in asthmatics and healthy adults.
See Tables AX6-3 andAX6-13.
Placebo-control: O3 caused significant decrements in FEV;
(13.5%) and FVC (10%) immediately following exposure,
a small increase in Mch-reactivity, and increased PMNs and
myeloperoxidase in induced sputum at 4 h postexposure. FEVj at
96% and FVC at 97% preexposure values at 3 h postexposure.
Budesonide for 2 wk prior to exposure did not affect spirometric
responses. See Section AX6. 2. 5 and Table AX6-13
Significant O3-induced decrements in spirometric lung function.
Young adults (< 35 years) were significantly more responsive
than older individuals (> 35 years). Sufentanil, a narcotic
analgesic, largely abolished symptom responses and improved
FEVj in strong responders. Naloxone, an opioid antagonist,
did not affect O3 effects in weak responders. See Section
AX6.2.5.1
Placebo and antioxidant groups had O3-induced decrements in
FEVj (20 and 14%) and FVC (13 and 10%), respectively.
Percent neutrophils and IL-6 levels in BAL fluid obtained 1 h
post exposure were not different in the two treatment groups.
See Table AX6-13.
Mean O3-induced FEVj decrements of 15.9% in males and 9.4%
in females (gender differences not significant). FEV; decrements
ranged from -4 to 56%; decrements >15% in 20 subjects and
>40% in 4 subjects. Uptake of O3 greater in males than females,
but uptake not correlated with spirometric responses.
Reference
Mudway etal.
(2001)
Stenfors etal.
(2002)
Nightingale
et al. (2000)
Passannante
etal. (1998)
Samet et al.
(2001)
Steck-Scott
et al. (2004)
Ultman etal.
(2004)
o
HH
H
W
"See Appendix A for abbreviations and acronyms.
bListed from lowest to highest O3 concentration.
'Studies conducted in exposure chamber unless otherwise indicated.
-------
1
2
3
4
5
6
7
following a 2 h exposure at three nominal levels ofVE are illustrated in Figure AX6-1 as a
function of O3 concentration. Their analysis indicated that C was marginally, but not
significantly more important than VE in predicting FEVj response. Additionally, the McDonnell
et al. (1997) analysis revealed that some prior analyzes of IE protocols may have over estimated
the relative importance of C over VE in predicting FEVj responses by considering only the VE
during exercise and ignoring the VE during periods of rest.
24-
20-
•£ 16-
E
« 12-|
o
I 8H
HI
4-
0-
-4-
0.0
T = 2h
0.1
0.2 0.3
O3 (ppm)
Ve = 40 L/min (n)
Ve = 31 L/min («)
Ve = 10 L/min (A)
0.4
0.5
Figure AX6-1. FEVt decrements as a function of O3 concentration following a 2 h
exposure with incremental exercise (15 min intervals) or rest. Points are
experimental data (mean ± SE) and lines are model predictions for each
•
activity level. Minute ventilation (VE) represent average across intervals
of rest and exercise.
Source: McDonnell et al. (1997).
January 2005
AX6-7
DRAFT-DO NOT QUOTE OR CITE
-------
1 Ultman et al. (2004) measured O3 uptake and pulmonary responses in 60 young heathy
2 non-smoking adults (32 M, 28 F). A bolus technique was used to quantify the uptake of O3 as a
3 function of the volume into the lung which the bolus penetrated. From these measurements, the
4 volumetric depth at which 50% uptake occurred was calculated. This volumetric lung depth was
5 correlated with conducting airways volume, i.e., a greater fraction of O3 penetrated to deeper into
6 the lungs of individuals have larger conducting airways volumes. Two weeks after the bolus
7 measurements, subjects were exposed via a face mask to FA and subsequently two weeks later to
8 0.25 ppm O3 for 1 h with CE at a target VE of 30 L/min. The breath-by-breath uptake of O3 was
9 measured. There was a small but significant reduction in the breath-by-breath uptake of O3 from
10 90.6% on average for the first 15 minutes to 87.3% on average for the last 15 minutes of
11 exposure. The uptake fraction was significantly greater in males (91.4%) than females (87.1%),
12 which is consistent with the larger fB and smaller VT of the females than males. Uptake was not
13 correlated with spirometric responses. However, there was tendency for males to have greater
14 O3-induced FEVj decrements than females, 15.9% versus 9.4%, respectively. There was
15 considerable intersubject variability in FEVj decrements which ranged from -4 to 56% with 20
16 subjects having decrements of >15% and 4 subjects with >40% decrements (see Section AX6.4
17 for additional discussion regarding intersubject variability).
18
19 AX6.2.3 Prolonged Ozone Exposures
20 Between 1988 and 1994, a number studies were completed that described the responses of
21 subjects exposed to relatively low (0.08 to 0.16 ppm) O3 concentrations for exposure durations
22 of 4 to 8 h. These studies were discussed in the 1996 criteria document (U.S. Environmental
23 Protection Agency, 1996) and only a select few are briefly discussed here. Table AX6-2 details
24 newer studies of healthy subjects undergoing prolonged exposures at O3 concentrations ranging
25 from 0.06 to 0.20 ppm. In most of these studies, statistically significant changes in pulmonary
26 function, symptoms, and airway responsiveness have been observed during and after exposures
27 to O3 concentrations of 0.08 ppm and higher. As with studies conducted at higher O3
28 concentrations for shorter periods of time, there is considerable intersubject variability in
29 response (see Section AX6.4).
30 Folinsbee et al. (1988) first reported the effects of a 6.6 h exposure to 0.12 ppm O3 in ten
31 young healthy adults (25 ± 4 yr) with quasi continuous exercise that was intended to simulate a
January 2005 AX6-8 DRAFT-DO NOT QUOTE OR CITE
-------
Table AX6-2. Pulmonary Function Effects after Prolonged Exposures to Ozone"
to
o
o
X
H
6
o
o
H
O
Ozone Concentration11 Exposure
ppm
Studies with
0.18
0.0
0.20
0.2
0.0
0.24
Studies with
0.0
0.04
0.08
0.12
0.12
ug/m3 and Activity
4 hr Exposures
353 4hIE
(4 x 50 min)
VE = 35L/min
0 4hIE
392 (4 x 50 min cycle
ergometry or
treadmill running
[VE = 40L/min])
392 4 h IE
(4 x 50 min)
VE = 25 L/min/m2
BSA
0 4hIE
470 (4x15 min)
VE =20 L/min
>6 hr Exposures
0 6.6 h
78 IE (6 x 50min)
157 VE = 20 L/min/m2
235 BSA
235 3 day-6.6h/day
IE (6 x 50 min)
VE = 17 L/min/m2,
20 L/min/m2
BSA, and 23
L/min/m2 BSA
Number and
Exposure Gender of Subject
Conditions Subjects Characteristics
23 °C 2 M, 2 F Adults NS, 21
50% RH to 33 years old
20 °C FA: 1 1 M, 3 Adult NS, 19 to
50% RH F 41 years old
03: 9M, 3F
20 °C 42 M, 24 F Adults NS,
50% RH 18 to 50 years
old
24°C 10 M Healthy NS,
40% RH 60 to 69 years
9 M COPD
59 to 71 years
23 °C 15 M, 15 F Healthy NS,
50% RH 22.4 ± 2.4 yrs
old
23 °C 1 5 M, 1 5 F Healthy NS, 1 8
50% RH to 31 years old
Observed Effect(s)
FVC decreased 19% and FEVj decreased 29% in these four
pre-screened sensitive subjects.
Decrease in FVC, FEVj, VT, and SRaw and increase in fB with
O3 exposure compared with FA; total cell count and LDH
increased in isolated left main bronchus lavage and inflammatory
cell influx occurred with O3 exposure compared to FA exposure.
FEVj decreased by 18.6%; Pre-exposure methacholine
responsiveness was weakly correlated with the functional response
to O3 exposure. Symptoms were also weakly correlated with the
FEVj response (r = -0.31 to -0.37)
Healthy: small, 3.3%, decline in FEVj (p=0.03 [not reported in
paper], paired-t on O3 versus FA pre-post FEVj). COPD: 8%
decline in FEV; (p=ns, O3 versus FA). Adjusted for exercise,
ozone effects did not differ significantly between COPD patients
and healthy subjects.
See Section AX6. 5.1.
FEVj and total symptoms at 6.6 h exposure to 0.04 ppm not
significantly different from FA. FEVj (-6.4%) and total symptoms
significant at 6.6 h exposure to 0.08 ppm. FEV; (-15.4%) at 6.6 h
not significantly different between chamber and face mask
exposure to 0.12 ppm.
FEVj at 6.6 h decreased significantly by 9.3%, 11.7%, and 13.9%,
respectively at three different exercise VE rates, but were not
significantly different from each other. Total symptoms at the
highest VE protocol were significantly greater than for the lowest
VE protocol beginning at 4.6 h. Largest subjects (2.2 m2 BSA)
had significantly greater average FEVj decrement for the three
protocols, 18.5% compared to the smallest subjects (1.4 m2
BSA), 6.5%.
Reference
Adams
(2000a)
Aris et al.
(1993)
Aris et al.
(1995)
Gong et al.
(1997a)
Adams (2002)
Adams
(2000b)
o
HH
H
W
-------
Table AX6-2 (cont'd). Pulmonary Function Effects after Prolonged Exposures to Ozonea
to
o
o
X
I
o
Ozone Concentration11
ppm
(a) 0.08
(b) 0.08
(mean) varied
from 0.03 to
0.15
(a) 0.08
(b) 0.30
(a) 0.12
(b)0.12
(mean) varied
from 0.07 to
0.16
ug/m3
235
235
(mean)
157
588
235
235
(mean)
Exposure
Duration
and Activity
6.6 h
IE (6 x 50 min)
VE = 20 L/min/m2
BSA
6.6 h
IE (6 x 50 min)
VE = 20 L/min/m2
BSA
2h
IE (4 x 15 min)
VE = 35 L/min/m2
BSA
6.6 h IE
(6 x 50 min)
(a,b,c) VE = 20
L/min/m2 BSA
(d)VE=12
L/min/m2 BSA
Number and
Exposure Gender of Subject
Conditions Subjects Characteristics
23 °C 15 M Healthy NS,
50% RH 15 F 18 to 25 years
old
23 °C 15 M Healthy NS,
50% RH 15 F 18 to 25 years
old
23 °C 6 M, 6 F Healthy NS,
50% RH 19 to 25 years
old
Observed Effect(s)
(a) FEVj decreased 6.2% after 6.6 h in square-wave exposures.
Total symptoms significantly increased at 5.6 and 6.6 h.
(b) FEV; decreased 5.6 to 6.2% after 4.6 to 6.6 h, respectively,
in varied exposure; total symptoms significantly increased also
after 4.6 to 6.6 h. No significant difference between face mask
and chamber exposures.
Significantly greater FEV; decrement (12.4%) for 2-h, 0.30 ppm
exposure than for 6.6-h, 0.08 ppm exposure (3.6%).
(a) FEVj decreased 11% at 6.6 h in square-wave exposure.
Total symptoms significant from 4.6 to 6.6 h.
(b) FEV; decreased 13% at 6.6 h; not significantly different from
square-wave exposure. Total symptoms significant from 4.6 to
6.6 h.
Reference
Adams
(2003a)
Adams
(2003b)
Adams and
Ollison(1997)
H
6
o
o
H
O
(c)0.12 235
(mean) varied (mean)
from 0.11 to
0.13
(d)0.12
235
(c) FEVj decreased 10.3% at 6.6 h; not significantly different from
square-wave exposure. Total symptoms significant from 4.6 to
6.6 h.
(d) FEV; decreased 3.6% at 6.6 h; significantly less than for
20 L/min/m2 BSA protocols.
"See Appendix A for abbreviations and acronyms.
bListed from lowest to highest O3 concentration.
o
HH
H
W
-------
1 full workday of heavy physical labor. Except for a 35-min lunch break after 3 h, the subjects
2 exercised at a moderate level ( VE « 40 L/min) for 50 min of each hour. Ignoring the lunch
3 break during which lung function did not change appreciably, approximately linear decreases
4 were observed in FVC, FEVl3 and FEV25_75 with duration of O3 exposure. Correcting for FA
5 responses, decrements of 8.2, 14.9, and 26.8% in FVC, FEVl5 and FEV25.75 occurred as a result
6 of the O3 exposure. Using the same 6.6 h protocol, but a lower O3 concentration of 0.08 ppm,
7 Horstman et al. (1990) and McDonnell et al. (1991) observed decrements corrected for FA (and
8 averaged across studies) of 5, 8, and 11% in FVC, FEVb and FEV25_75, respectively, in 60 young
9 adults (25 ± 5 years old). Horvath et al. (1991) observed a 4% (p = 0.03)1 decrement in FEVj
10 using the forementioned protocol (i.e., 6.6 h and 0.08 ppm O3) in 11 healthy adults (37 ± 4 yr).
11 The smaller decrement observed by Horvath et al. (1991) versus Horstman et al. (1990) and
12 McDonnell et al. (1991) is consistent with response decreasing as subject age increases (see
13 Section AX6.5.1}.
14
15 AX6.2.3.1 Effect of Exercise Ventilation Rate on FEVt Response to 6.6 h Ozone Exposure
16 It is well known that response to O3 exposure is a function of VE in studies of 2 h or less in
17 duration (See Section AX6.2.2). It is reasonable to expect that response to a prolonged 6.6-h O3
18 exposure is also function of VE, although quantitative analyzes are lacking.
19 In an attempt to quantify this effect, Adams and Ollison (1997) exposed 12 young adults to
20 an average O3 concentration of 0.12 ppm for 6.6 h at varied exercise VE . They observed a mean
21 FEVj decrements of 10 to 11% in two protocols having a mean exercise VE of 33 L/min and a
22 14% decrement in a protocol with a mean exercise VE of 36 L/min. These FEVj decrements
23 were significantly greater than the average decrement of 3.6% (not significantly different from
24 FA response) observed at an exercise VE of only 20 L/min. In a subsequent study of 30 healthy
25 adults (Adams, 2000b), the effect of smaller exercise VE differences on pulmonary function and
26 symptoms responses to 6.6 h exposure to 0.12 ppm O3 was examined. FEVj decrements of 9.3,
27 11.7, and 13.9% were observed for the exercise VE of 30.2, 35.5, and 40.8 L/min, respectively.
28 Along with the tendency for FEVj responses to increase with VE, total symptoms severity was
'Based on two-tailed paired t-test of data in Table III of Horvath et al. (1991).
January 2005 AX6-11 DRAFT-DO NOT QUOTE OR CITE
-------
1 found to be significantly greater at the end of the highest VE protocol relative to the lowest VE
2 protocol. Although the FEVj responses were not significantly different from each other, the
3 power of the study to detect differences between the three VE was not reported and no analysis
4 was performed using all of the data (e.g., a mixed effects model). Data from the Adams and
5 Ollison (1997) and Adams (2000b) studies are illustrated in Figure AX6-2 with data from three
6 older studies. There are a paucity of data below an exercise VE of 30 L/min. Existing data for
7 exposure to 0.12 ppm O3 suggests that FEVj responses increase with increasing exercise VE until
8 at least 35 L/min.
9
10 AX6.2.3.2 Exercise Ventilation Rate as a Function of Body/Lung Size on FEVt Response
11 to 6.6 h Ozone Exposure
12 Typically, with the assumption that the total inhaled O3 dose should be proportional to the
13 lung size of each individual, exercise VE in 6.6 h exposures has been set as a multiple of body
14 surface area (BSA) (McDonnell et al., 1991) or as a product of eight times FVC (Folinsbee et al.,
15 1988; Frank et al., 2001; Horstman et al., 1990). Utilizing previously published data, McDonnell
16 et al. (1997) developed a statistical model analyzing the effects of O3 concentration, VE, duration
17 of exposure, age, and body and lung size on FEVj response. They concluded that any effect of
18 BSA, height, or baseline FVC on percent decrement in FEVj in this population of 485 young
19 adults was small if it exists at all. This is consistent with Messineo and Adams (1990), who
20 examined pulmonary function responses in young adult women having small (n = 14) or large
21 (n = 14) lung sizes (mean FVC of 3.74 and 5.11 L, respectively). Subject were exposed to
22 0.30 ppm O3 for 1 h with CE (VE = 47 L/min). There was no significant difference between the
23 group FEVj decrements (22.1 and 25.6% for small and large lung, respectively). In addition,
24 Messineo and Adams (1990) also did a retrospective analysis of 36 young adult males who each
25 had completed similar 1 h exposures to 0.30 ppm O3 with CE (VE « 70 L/min) and found lung
26 size was not real ted with FEVj response.
27 Adams (2000b) studied a group of 30 young adult men and women exposed to 0.12 ppm
28 O3 for 6.6 h on three occasions while exercising 50 min of each hour at one of three different VE
29 levels (viz., 17, 20, and 23 l/min/m2BSA). Their postexposure FEVj responses were regressed
30 as a function of BSA (which was directly related to the absolute amount of VE during exercise
January 2005 AX6-12 DRAFT-DO NOT QUOTE OR CITE
-------
25
& 20
0)
E
S> 15
o
0)
Q
£10
LU
LL
5-
• Adams (2000b)
D Adams and Ollison (1997)
A Folinsbee et al. (1988)
A Folinsbee et al. (1994)
• Horstman et al. (1990)
15 20 25 30 35
Minute Ventilation (L/min)
40
45
Figure AX6-2. Average FEVt decrements (±SE) for prolonged 6.6 h exposures to 0.12 ppm
•
O3 as a function of exercise VE. Since age affects response to O3 exposure,
selected studies had subjects with mean ages between 22 and 25 years.
FEVj decrements were calculated as mean O3 responses minus mean air
responses. Unless provided in papers, SE were estimated from variability
in post O3 exposure responses. In one case, the SE for VE of 33 L/Min
(10.3% decrement) was taken as the SE of data from protocol with VE of
33 L/min (11% decrement). All studies used a constant 0.12 ppm O3
exposure except two (*) which used 0.115 ppm O3 for hours 1-2 and 5-6
and 0.13 ppm O3 for hours 3-4 of exposure.
1 and, thus, primarily responsible for individual differences in total inhaled O3 dose). The slope
2 was significantly different from zero (p = 0.01), meaning that the smallest subjects, who had the
3 lowest exercise VE (~ 26 L/min), had a lower FEVj decrement (-5%) than the largest subjects
4 (-17%), whose exercise VE was -44 L/min. This relationship was not a gender-based
5 difference, as the mean female's FEVj decrement was -11.2%, which was not significantly
6 different from the male's -12.2% mean value. Similarly, when total symptoms severity
7 response was regressed against BSA, the slope was significantly different than zero (p = 0.0001),
8 with lower values for smaller subjects than for larger subjects. Results of this study suggest that
January 2005
AX6-13
DRAFT-DO NOT QUOTE OR CITE
-------
1 for the O3 concentration and exposure duration used, responses are more closely related to VE
2 than VE normalized to BSA. Further, this observation is in agreement with McDonnell et al.
3 (1997), who observed no evidence that measurements of lung or body size were significantly
4 related to FEVj response in 2 h IE exposures. These authors state that the absence of an
5 observed relationship between FEVj response and BSA, height, or FVC may be due to the poor
6 correlation between these variables and airway caliber (Collins et al., 1986; Martin et al., 1987).
7 Also, the O3 dosimetry study of Bush et al. (1996) indicated that normalization of the O3 dose
8 would be more appropriately applied as a function of anatomic dead space.
9
10 AX6.2.3.3 Comparison of 6.6 h Ozone Exposure Pulmonary Responses to Those Observed
11 in 2 h Intermittent Exercise Ozone Exposures
12 It has been shown that greater O3 concentration (Horstman et al., 1990) and higher VE
13 (Adams, 2000b) each elicit greater FEVj response in prolonged, 6.6-h exposures, but data on the
14 relative effect of O3 concentration, VE, and T in prolonged exposures are very limited and have
15 not been systematically compared to data from shorter (< 2-h) exposures. In a recent study
16 (Adams, 2003b), the group mean FEVj response for a 2-h IE exposure to 0.30 ppm O3 was
17 -12.4%, while that for a 6.6-h exposure to 0.08 ppm O3 was -3.5%. The total inhaled O3 dose
18 (as the simple product of C XT x VE ) was 1358 ppm-L for the 2-h exposure and 946 ppm-L for
19 the 6.6-h exposure. Thus, the FEVj decrement was 3.5 times greater and the total inhaled O3
20 dose was 1.44 times greater for the 2-h exposure compared to the 6.6-h exposure. This
21 difference illustrates the limitations of utilizing the concept of total O3 dose for comparisons
22 between studies of vastly different exposure durations.
23 Adams (2003b) also examined whether prolonged 6.6 h exposure to a relatively low O3
24 concentration (0.08 ppm) and the 2-h IE exposure at a relatively high O3 concentration (0.30
25 ppm) elicited consistent individual subject effects, i.e, were those most or least affected in one
26 exposure also similarly affected in the other? Individual subject O3 exposure reproducibility was
27 first examined via a regression plot of the postexposure FEVj response to the 6.6-h chamber
28 exposure as a function of postexposure FEVj response to the 2-h chamber exposure. The R2 of
29 0.40, although statistically significant, was substantially less than that observed in a comparison
30 of individual FEVj response to two 2-h IE exposures by chamber and face mask, respectively
31 (R2 = 0.83). The Spearman rank order correlation for the chamber 6.6-h and chamber 2-h
January 2005 AX6-14 DRAFT-DO NOT QUOTE OR CITE
-------
1 exposure comparison was also substantially less (0.49) than that obtained for the two 2-h
2 exposures (0.85). The primary reason for the greater variability in the chamber 6.6-h exposure
3 FEVj response as a function of that observed for the two 2-h IE exposures is very likely related
4 to the increased variability in response upon repeated exposure to O3 concentrations lower than
5 0.18 ppm (R = 0.57, compared to a mean R of 0.82 at higher concentrations) reported by
6 McDonnell et al. (1985a). This rationale is supported by the lower R (0.60) observed by Adams
7 (2003b) for the FEVj responses found in 6.6 h chamber and face mask exposures to 0.08 ppm
8 O3, compared to an R of 0.91 observed for responses found at 0.30 ppm O3.
9
10 AX6.2.4 Triangular Ozone Exposures
11 To further explore the factors that determine responsiveness to O3, Hazucha et al. (1992)
12 designed a protocol to examine the effect of varying, rather than constant, O3 concentrations.
13 In this study, subjects were exposed to a constant level of 0.12 ppm O3 for 8 h and to an O3 level
14 that increased linearly from 0 to 0.24 ppm for the first 4 h and then decreased linearly from
15 0.24 to 0 over the second 4 h of the 8 h exposure (triangular concentration profile). Subjects
16 performed moderate exercise (VE -40 L/min) during the first 30 minutes of each hour. The total
17 inhaled O3 dose (i.e., C x T x VE ) for the constant versus the triangular concentration profile was
18 almost identical. FEVj responses are illustrated in Figure AX6-3. With exposure to the constant
19 0.12 ppm O3, FEVj declined approximately 5% by the fifth hour of exposure and then remained
20 at that level. This observation clearly indicates a response plateau as suggested in other
21 prolonged exposure studies (Horstman et al., 1990; McDonnell et al., 1991). However, with the
22 triangular O3 concentration profile after a minimal initial response over the first 3 h, Hazucha
23 et al. (1992) observed a substantial decrease in FEVj corresponding to the higher average O3
24 concentration that reached a nadir after 6 h (-10.3%). Despite 2 h of continued exposure to a
25 lower O3 concentration (0.12 to 0.00 ppm, mean = 0.06 ppm), FEVj improved and was only
26 reduced by 6.3% (relative to the preexposure FEVj) at the end of the 8-h exposure. The authors
27 concluded that total inhaled O3 dose (C x VE x T) was not a sufficient index of O3 exposure and
28 that, as observed by others (Adams et al., 1981; Folinsbee et al., 1978; Hazucha, 1987; Larsen
29 et al., 1991), O3 concentration appears to be more important in determining exposure effects than
30 is either duration or the volume of air breathed during the exposure. However, it should be noted
31 that the mean O3 concentration for Hazucha et al.'s triangular exposure profile was 0.12 ppm
January 2005 AX6-15 DRAFT-DO NOT QUOTE OR CITE
-------
4.5
4.4
4.3
HI
(0
4.2
4.1
4.0
3.9
••••••• 0 ppm
—H— 0.12 ppm
Constant
—A— Variable
0246
Exposure Duration (h)
Figure AX6-3. The forced expiratory volume in 1 s (FEVt) is shown in relation to exposure
duration (hours) under three exposure conditions. Subjects exercised
(minute ventilation « 40 L/min) for 30 min during each hour; FEVt was
measured at the end of the intervening rest period. Standard error of the
mean for these FEVt averages (not shown) ranged from 120 to 150 mL.
Source: Hazuchaetal. (1992).
1 at 4 h, 0.138 ppm at 5 h, 0.14 ppm at 6 h, and 0.133 ppm at 7 h, before falling to 0.12 ppm at 8 h.
2 The FEVj responses of the last 4 hours (Figure AX6-3) follow a closely similar pattern as the
3 total mean O3 concentration over the same time period.
4 It has become increasingly well realized that laboratory simulations of air-pollution risk-
5 assessment need to employ O3 concentration profiles that more accurately mimic those
6 encountered during summer daylight ambient air pollution episodes (Adams and Ollison, 1997;
7 Lefohn and Foley, 1993; Rombout et al., 1986). Neither square-wave O3 exposures or the one
8 8-h study by Hazucha et al. (1992) that utilized a triangular shaped varied O3 exposure described
9 above closely resembles the variable diurnal daylight O3 concentration pattern observed in many
10 urban areas experiencing air-pollution episodes (Lefohn and Foley, 1993). Recently, 6.6 h less
January 2005
AX6-16
DRAFT-DO NOT QUOTE OR CITE
-------
1 abrupt triangular O3 exposure profiles at lower concentrations more typical of outdoor ambient
2 conditions have been examined (Adams 2003a; Adams and Ollison, 1997).
3 Using a face-mask inhalation system, Adams and Ollison (1997) observed no significant
4 differences in postexposure pulmonary function responses or symptoms between the 6.6-h,
5 0.12 ppm O3 square-wave exposure; and those observed for a triangular O3 profile in which
6 concentration was increased steadily from 0.068 ppm to 0.159 ppm at 3.5 h and then decreased
7 steadily to 0.097 ppm at end exposure. Further, no attenuation in FEVj response during the last
8 2 h was observed in either the 6.6 h square-wave or the triangular exposures. In a subsequent
9 study (Adams, 2003a), no significant difference was observed in pulmonary function responses
10 or symptoms between face-mask and chamber exposure systems either for a 6.6-h, 0.08 ppm O3
11 square-wave profile or for the triangular O3 exposure beginning at 0.03 ppm, increasing steadily
12 to 0.15 ppm in the fourth hour, and decreasing steadily to 0.05 ppm at 6.6 h (mean = 0.08 ppm).
13 For the chamber-exposure comparison, postexposure values for FEVj and symptoms were not
14 significantly different from the responses for the square-wave 0.08 ppm O3 exposure. However,
15 analysis showed that FEVj response for the square-wave protocol did not become statistically
16 significant until the 6.6-h postexposure value, while that for the triangular exposure protocol was
17 significant at 4.6 h (when O3 concentration was 0.15 ppm). Earlier significant FEVj responses
18 for the triangular protocol were accompanied by significant increases in symptoms at 4.6 h,
19 continuing on through the fifth and sixth hours when the mean O3 concentration was 0.065 ppm.
20 Symptoms for the square-wave 0.08 ppm exposure did not become statistically significant until
21 5.6 h. The rate of FEVj response to the triangular exposure did not decrease as was observed by
22 Hazucha et al. (1992) during the last two hours of their 8-h triangular exposure (Figure AX6-3).
23 Rather, FEVj responses for the triangular exposure showed clear signs of plateauing during the
24 last 2 h; i.e., -5.46% at 4.6 h, -6.27% at 5.6 h, and -5.77% at 6.6 h. The most probable reason
25 for differences in the triangular O3 profile observations of Hazucha et al. (1992) and those of
26 Adams (2003a) is that the increase and decrease in Hazucha et al.'s study (i.e., 0 to 0.24 ppm and
27 back to 0) encompassed a much greater range of O3 concentrations than those used by Adams
28 (2003a), viz., 0.03 ppm to 0.15 ppm from 3.6 to 4.6 h, then decreasing to 0.05 ppm at 6.6 h.
29 Nonetheless, the greatest FEVj decrement was observed at 6 h of Hazucha et al.'s 8 h triangular
30 exposure (Figure AX6-3) corresponding to the time when total mean O3 concentration was
31 highest (0.14 ppm), with a very similar response at 7 h when total mean O3 concentration was
January 2005 AX6-17 DRAFT-DO NOT QUOTE OR CITE
-------
1 0.138 ppm. Adams (2003a) observed the greatest FEVj decrement at 5.6 h (-6.27% with total
2 mean O3 concentration of 0.086 ppm), which was not significantly different than the 4.6-h value
3 of -5.46% (total mean O3 concentration = 0.0875 ppm).
4 Whereas FEVj decrements during square-wave O3 exposures between 0.08 to 0.12 ppm
5 tend to increase with time of exposure (i.e., with steadily increasing total inhaled dose), FEVj
6 decrements during triangular exposures (Hazucha et al., 1992; Adams, 2003a) occurred 1 to 2 h
7 after the peak O3 concentration and 1 h to 2 h before the maximal total O3 inhaled dose occurred
8 at end exposure. This difference, especially because O3 concentration profiles during summer
9 daylight air-pollution episodes rarely mimic a square-wave, implies that triangular O3 exposure
10 profiles most frequently observed during summer daylight hours merit further investigation.
11
12 AX6.2.5 Mechanisms of Pulmonary Function Responses
13 Inhalation of O3 for several hours while physically active elicits both subjective respiratory
14 tract symptoms and acute pathophysiologic changes. The typical symptomatic response
15 consistently reported in studies is that of tracheobronchial airway irritation. This is accompanied
16 by decrements in lung capacities and volumes, bronchoconstriction, airway hyperresponsiveness,
17 airway inflammation, immune system activation, and epithelial injury. The severity of
18 symptoms and the magnitude of response depend on inhaled dose, O3 sensitivity of an individual
19 and the extent of tolerance resulting from previous exposures. The development of effects is
20 time dependent during both exposure and recovery periods with considerable overlap of evolving
21 and receding effects.
22 Exposure to O3 initiates reflex responses manifested as a decline in spirometric lung
23 function parameters (1FVC, IFEVl3 1FEF25.75), bronchoconstriction (t SRaw) and altered
24 breathing pattern (1VT, t fB), which becomes more pronounced as exposure progresses and
25 symptoms of throat irritation, cough, substernal soreness and pain on deep inspiration develop.
26 The spirometric lung function decline and the severity of symptoms during a variable (ramp)
27 exposure profile seem to peak a short time (about 1 to 2 h) following the highest concentration of
28 O3 (Hazucha et al., 1992; Adams, 2003a). Exposure to a uniform O3 concentration profile elicits
29 the maximum spirometric response at the end of exposure (Hazucha et al., 1992; Adams, 2003a).
30 Regardless of exposure concentration profile, as the exposure to O3 progresses airway
31 inflammation begins to develop and the immune response at both cellular and subcellular level is
January 2005 AX6-18 DRAFT-DO NOT QUOTE OR CITE
-------
1 activated. Airway hyperreactivity develops slower than pulmonary function effects, while
2 neutrophilic inflammation of the airways develops even more slowly and reaches the maximum
3 3 to 6 h postexposure. The cellular responses (e.g., release of immunoregulatory cytokines)
4 appear to still be active as late as 20 h postexposure (Torres et al., 2000). Following cessation of
5 exposure, the recovery in terms of breathing pattern, pulmonary function and airway
6 hyperreactivity progresses rapidly and is almost complete within 4 to 6 hours in moderately
7 responsive individuals. Persisting small residual lung function effects are almost completely
8 resolved within 24 hours. Following a 2 h exposure to 0.4 ppm O3 with IE, Nightingale et al.
9 (2000) observed a 13.5% decrement in FEVj. By 3 h postexposure, however, only a 2.7% FEVj
10 decrement persisted. As illustrated in Figure AX6-4, a similar postexposure recovery in FVC
11 was observed. In hyperresponsive individuals, the recovery takes longer and as much as
12 48 hours to return to baseline values. More slowly developing inflammatory and cellular
13 changes persist for up to 48 hours. The time sequence, magnitude and the type of responses of
14 this complex series of events, both in terms of development and recovery, indicate that several
15 mechanisms, activated at different time of exposure must contribute to the overall lung function
16 response (U.S. Environmental Protection Agency, 1996).
17
18 AX6.2.5.1 Pathophysiologic Mechanisms
19 Breathing pattern changes
20 Human studies consistently report that inhalation of O3 alters the breathing pattern without
21 significantly affecting minute ventilation. A progressive decrease in tidal volume and a
22 "compensatory" increase in frequency of breathing to maintain steady minute ventilation during
23 exposure suggests a direct modulation of ventilatory control. These changes parallel a response
24 of many animal species exposed to O3 and other lower airway irritants (Tepper et al., 1990).
25 Although alteration of a breathing pattern could be to some degree voluntary, the presence of the
26 response in animals and the absence of perception of the pattern change by subjects, even before
27 appearance of the first subjective symptoms of irritation, suggests an involuntary reflex
28 mechanism.
29 Direct recording from single afferent vagal fibers in animals convincingly demonstrated
30 that bronchial C-fibers and rapidly adapting receptors are the primary vagal afferents responsible
31 for O3-induced changes in ventilatory rate and depth (Coleridge et al., 1993; Hazucha and
January 2005 AX6-19 DRAFT-DO NOT QUOTE OR CITE
-------
a.
LJ
b.
o
4.0-,
3.5-
2.5-
2.0-
I I I I I I I ' >\
Pre 0 1 2 3 4 24
Time from Exposure (hours)
4.5-I
4.0-
3.5-
3.0-
Pre 01234
Time from Exposure (hours)
24
Figure AX6-4a,b. Recovery of spirometric responses following a 2 h exposure to 0.4 ppm
O3 with IE. Immediately postexposure, there were significant
decrements (**p < 0.001, ***p < 0.0005) in FVC (10%) and FEVt
(13.5%) compared to preexposure values. At 3 h postexposure, FVC
and FEVt were at 96 and 97% of preexposure values, respectively.
Adapted from Nightingale et al. (2000).
1 Sant'Ambrogio, 1993). In spontaneously breathing dogs, an increase in VT/T; (T; decreased more
2 than VT) was attributed to an increased inspiratory drive due to stimulation of rapidly adapting
3 receptors and bronchial C-fibers by O3 (Schelegle et al., 1993). Folinsbee and Hazucha (2000)
4 also observed similar changes in VT/T; and other breath-timing parameters in humans exposed to
January 2005
AX6-20
DRAFT-DO NOT QUOTE OR CITE
-------
1 O3 implying activation of the same mechanisms. They also reported that Pm0A (pressure at
2 mouth at 0.1 sec of inspiration against a transiently occluded mouthpiece which is considered an
3 index of inspiratory drive) increased during controlled hypercapnia without a change in the slope
4 of Pm0 j versus pCO2 relation suggesting that the primary mechanism is an increased inspiratory
5 drive. Since no significant within-individual differences in ventilatory response to CO2 between
6 air exposure and O3 exposure were found, the CO2 chemoreceptors did not modulate the
7 response. Therefore, the principal peripheral mechanism modulating changes in breathing
8 pattern appears to be direct and indirect stimulation of lung receptors and bronchial C-fibers by
9 O3 and/or its oxidative products. The activity of these afferents, centrally integrated with input
10 from other sensory pathways, drives the ventilatory controller, which determines the depth and
11 the frequency of breathing.
12 The potential modulation of breathing pattern by activation of sensory afferents located in
13 extrathoracic airways by O3 has not yet been studied in humans. Laboratory animal studies have
14 shown that the larynx, pharynx, and nasal mucosa are densely populated by free-ending,
15 unmyelinated sensory afferents resembling nociceptive C-fibers (Spit et al., 1993; Sekizawa and
16 Tsubone, 1994). They are almost certainly stimulated by O3 and likely contribute to overall
17 ventilatory and symptomatic responses. Nasal only exposure of rats produced O3-induced
18 changes in breathing pattern that are similar to changes found in humans (Kleinman et al., 1999).
19
20 Symptoms and lung function changes
21 As already discussed, in addition to changes in ventilatory control, O3 inhalation by
22 humans will also induce a variety of symptoms, reduce vital capacity (VC) and related functional
23 measures, and increase airway resistance. Hazucha et al. (1989) postulated that a reduction of
24 VC by O3 is due to a reflex inhibition of inspiration and not due to a voluntary reduction of
25 inspiratory effort. Recently, Schelegle et al. (2001) convincingly demonstrated that a reduction
26 of VC due to O3 is indeed reflex in origin and not a result of subjective discomfort and
27 consequent premature voluntary termination of inspiration. They reported that inhalation of an
28 aerosolized topical anesthetic tetracaine substantially reduced if not abolished O3-induced
29 symptoms that are known to be mediated in part by bronchial C-fibers. Yet, such local
30 anesthesia of the upper airway mucosa had a minor and irregular effect on pulmonary function
31 decrements and tachypnea, strongly supporting neural mediation, i.e., stimulation of both
January 2005 AX6-21 DRAFT-DO NOT QUOTE OR CITE
-------
1 bronchial and pulmonary C-fibers, and not voluntary inhibition of inspiration (due to pain) as the
2 key mechanism.
3 The involvement of nociceptive bronchial C-fibers modulated by opioid receptors in
4 limiting maximal inspiration and eliciting subjective symptoms in humans was studied by
5 Passannante et al. (1998). The authors hypothesized that highly variable responses among
6 individuals might reflect the individual's inability or unwillingness to take a full inspiration.
7 Moreover, development of symptoms of pain on deep inspiration, cough and substernal soreness
8 suggested that nociceptive mechanism(s) might be involved in O3-induced inhibition of maximal
9 inspiration. If this were so, pain suppression or inhibition by opioid receptor agonists should
10 partially or fully reverse symptoms and lung functional impairment. Subjects for this study were
11 pre-screened with exposure to 0.42 ppm O3 and classified either as "weak" (FEVj > 95% of
12 preexposure value), "strong" (FEVj < 85% of preexposure value), or "moderate" responders.
13 Sixty two (28 M, 34 F) healthy volunteers (18 to 59 yrs old), known from the previous screening
14 to be "weak" (n = 20) or "strong" (n = 42) O3-responders, participated in this double-blind
15 crossover study. Subjects underwent either two 2 h exposures to air, or two 2 h exposures to
16 0.42 ppm O3, with 15 min IE at 17.5 1/min/m2 BSA. Immediately following postexposure
17 spirometry the "weak" responders were given (in random order) either the potent opioid receptor
18 antagonist naloxone (0.15 mg/kg) or saline, while "strong" responders received (in random
19 order) either the potent, rapid-acting opioid agonist and analgesic sufentanil (0.2 jig/kg), or
20 physiologic saline administered through an indwelling catheter. Administration of saline or
21 naloxone had no significant effect on the relatively small decrements in FEVj observed in
22 "weak" responders. However, as hypothesized, sufentanil rapidly reversed both the O3-induced
23 symptomatic effects and spirometric decrements (FEV^ p < 0.0001) in "strong" responders
24 (Figure AX6-5). All the same, the reversal was not complete and the average post-sufentanil
25 FEVj remained significantly below (-7.3%) the preexposure value suggesting involvement of
26 non-opioid receptor modulated mechanisms as well. Uneven suppression of symptoms has
27 implied involvement of both A-6 and bronchial C-fibers. The plasma p-endorphin (a potent
28 pain suppressor) levels, though substantially elevated immediately postexposure and post-drug
29 administration, were not related to individuals' O3 responsiveness. These observations have
30 demonstrated that nociceptive mechanisms play a key role in modulating O3-induced inhibition
31 of inspiration. Moreover, these findings are consistent with and further support the concept that
January 2005 AX6-22 DRAFT-DO NOT QUOTE OR CITE
-------
LJJ
0.5
1 1.5
Time (hours)
2.5
Figure AX6-5. Plot of the mean FE Vt (% baseline) vs. time for ozone exposed cohorts.
Solid lines represent data for "strong" males (n = 14; solid squares) and
females (n = 15; solid circles) that received sufentanil and dotted lines
represent data for the same cohorts after receiving saline. Dashed lines
represent data for "weak" males (n = 5; open squares) and females (n = 10;
open circles) that received naloxone and dot-dash lines represent data for
the same cohorts after receiving saline. The arrow denotes the time of drug
administration (~2.1 hrs). Vertical bars associated with the symbols are
one-sided SEM.
Source: Adapted from Passannante et al. (1998).
1 the primary mechanism of O3-induced reduction in inspiratory lung function, is an inhibition of
2 inspiration elicited by stimulation of the C-fibers. The absence of effect of naloxone in "weak"
3 responders shows that the weak response is not due to excessive endorphin production in those
4 individuals. However, other neurogenic mechanisms not modulated by opioid receptors may
5 have some though limited role in inspiratory inhibition.
January 2005
AX6-23
DRAFT-DO NOT QUOTE OR CITE
-------
1 Airway hyperreactivity
2 In addition to limitation of maximal inspiration and its effects on other spirometric
3 endopoints, activation of airway sensory afferents also plays a role in receptor-mediated
4 bronchoconstriction and an increase in airway resistance. Despite this common mechanism,
5 post-O3 pulmonary function changes and either early or late bronchial hyperresponsiveness
6 (BHR) to inhaled aerosolized methacholine or histamine are poorly correlated either in time or
7 magnitude. Fentanyl and indomethacin, the drugs that have been shown to attenuate O3-induced
8 lung function decrements in humans, did not prevent induction of BHR when administered to
9 guinea pigs prior to O3 exposure (Yeadon et al., 1992). Neither does post-O3 BHR seem to be
10 related to airway baseline reactivity. These findings imply that the mechanisms are either not
11 related or are activated independently in time. Animal studies (with limited support from human
12 studies) have suggested that an early post-O3 BHR is, at least in part, vagally mediated (Freed,
13 1996) and that stimulation of C-fibers can lead to increased responsiveness of bronchial smooth
14 muscle independently of systemic and inflammatory changes which may be even absent (load
15 et al., 1996). In vitro study of isolated human bronchi have reported that O3-induced airway
16 sensitization involves changes in smooth muscle excitation-contraction coupling (Marthan,
17 1996). Characteristic O3-induced inflammatory airway neutrophilia which at one time was
18 considered a leading BHR mechanism, has been found in a murine model, to be only
19 coincidentally associated with BHR and there was no cause and effect relationship (Zhang et al.,
20 1995). However, this observation does not rule out involvement of other cells such as
21 eosinophils or T-helper cells in BHR modulation. There is some evidence that release of
22 inflammatory mediators by these cells can sustain BHR and bronchoconstriction. In vitro and
23 animal studies have also suggested that airway neutral endopeptidase activity can be a strong
24 modulator of BHR (Marthan et al., 1996; Yeadon et al., 1992). Late BHR observed in some
25 studies is plausibly due to a sustained damage of airway epithelium and continual release of
26 inflammatory mediators (Foster et al., 2000). Thus, O3-induced BHR appears to be a product of
27 many mechanisms acting at different time periods and levels of the bronchial smooth muscle
28 signaling pathways. [The effects ofO3 on BHR are described in Section AX6.8. ]
29
January 2005 AX6-24 DRAFT-DO NOT QUOTE OR CITE
-------
1 AX6.2.5.2 Mechanisms at a Cellular and Molecular Level
2 Stimulation of vagal afferents by O3 and reactive products, the primary mechanism of lung
3 function impairment is enhanced and sustained by what can be considered in this context to be
4 secondary mechanisms activated at a cellular and molecular level. The complexity of these
5 mechanisms is beyond the scope of this section and the reader is directed to Section AX6.9 of
6 this chapter for greater details. A comprehensive review by Mudway and Kelly (2000) discusses
7 the cellular and molecular mechanisms of O3-induced pulmonary response in great detail.
8
9 Neurogenic airway inflammation
10 Stimulation of bronchial C-fibers by O3 not only inhibits maximal inspiration but, through
11 local axon reflexes, induces neurogenic inflammation. This pathophysiologic process is
12 characterized by release of tachykinins and other proinflammatory neuropeptides. Ozone
13 exposure has been shown to elevate C-fiber-associated tachykinin substance P in human
14 bronchial lavage fluid (Hazbun et al. 1993) and to deplete neuropeptides synthesized and
15 released from C-fibers in human airway epithelium rich in substance P-immunoreactive axons.
16 Substance P and other transmitters are known to induce granulocyte adhesion and subsequent
17 transposition into the airways, increase vascular permeability and plasma protein extravasation,
18 cause bronchoconstriction, and promote mucus secretion (Solway and Leff, 1991). Although the
19 initial pathways of neurogenic, antigen-induced, and generally immune-mediated inflammation
20 are not the same, they eventually converge leading to further amplification of airway
21 inflammatory processes by subsequent release of cytokines, eicosanoids, and other mediators.
22 Significantly negative correlations between O3-induced leukotriene (LTC4/D4/E4) production and
23 spirometric decrements (Hazucha et al., 1996), and an increased level of postexposure PGE2, a
24 mediator known to stimulate bronchial C-fibers, show that these mediators play an important
25 role in attenuation of lung function due to O3 exposure (Mohammed et al., 1993; Hazucha et al.,
26 1996). Moreover, because the density of bronchial C-fibers is much lower in the small than
27 large airways, the reported post O3 dysfunction of small airways assessed by decrement in
28 FEF25.75 (Weinman et al., 1995; Frank et al., 2001) may be due in part to inflammation. Also,
29 because of the relative slowness of inflammatory responses as compared to reflex effects, O3-
30 triggered inflammatory mechanisms are unlikely to initially contribute to progressive lung
31 function reduction. It is plausible, however, that when fully activated, they sustain and possibly
January 2005 AX6-25 DRAFT-DO NOT QUOTE OR CITE
-------
1 further aggravate already impaired lung function. Indeed, a prolonged recovery of residual
2 spirometric decrements following the initial rapid improvement after exposure termination could
3 be due to slowly resolving airway inflammation. Bronchial biopsies performed 6 h postexposure
4 have shown that O3 caused a significant decrease in immunoreactivity to substance P in the
5 submucosa (Krishna et al., 1997a). A strong negative correlation with FEVj also suggests that
6 the release of substance P may be a contributing mechanism to persistent post-O3
7 bronchoconstriction (Krishna et al., 1997a). Persistent spirometry changes observed for up to
8 48 h postexposure could plausibly be sustained by the inflammatory mediators, many of which
9 have bronchoconstrictive properties (Blomberg et al., 1999).
10
11
12 AX6.3 PULMONARY FUNCTION EFFECTS OF OZONE EXPOSURE
13 IN SUBJECTS WITH PREEXISTING DISEASE
14 This section examines the effects of O3 exposure on pulmonary function in subjects with
15 preexisting disease by reviewing O3 exposure studies that utilized subjects with (1) chronic
16 obstructive pulmonary disease (COPD), (2) asthma, (3) allergic rhinitis, and (4) ischemic heart
17 disease. Studies of subjects with preexisting disease exposed to O3, published subsequent to or
18 not included in the 1996 Air Quality Criteria Document (U.S. Environmental Protection Agency,
19 1996), are summarized in Table AX6-3. Studies examining increased airway responsiveness
20 after O3 exposure are discussed in Section AX6.8.
21
22 AX6.3.1 Subjects with Chronic Obstructive Pulmonary Disease
23 Five studies of O3-induced responses in COPD patients were available for inclusion in the
24 1996 criteria document (U.S. Environmental Protection Agency, 1996). The COPD patients in
25 these studies were exposed during light IE (4 studies) or at rest (1 study) for 1 to 2 hours to O3
26 concentrations between 0.1 and 0.3 ppm. None of theses studies found significant O3-induced
27 changes in pulmonary function. Of the four studies examining arterial oxygen saturation, two
28 reported small but statistically significant O3-induced decreases in the COPD patients. These
29 limited data suggest COPD patients experience minimal O3-induced effects for 0.3 ppm O3
30 exposures less than 2 hours in duration. These findings are also consistent decreasing O3 effects
31 with increasing age (see Section AX6.5.1).
January 2005 AX6-26 DRAFT-DO NOT QUOTE OR CITE
-------
o
O
to
O
O
Table AX6-3. Ozone Exposure in Subjects with Preexisting Disease"
X
ON
to
H
6
o
o
H
O
O
HH
H
W
Ozone
Concentration'
ppm Hg/ni3
Exposure Duration Exposure Number and
and Activity Condition Gender of Subjects
Subject Characteristics
Observed Effect(s)
Reference
Subjects with Chronic Obstructive Pulmonary or Heart Disease
0.0 0
0.24 472
0.3 588
4 h IE 24 °C 9 M
15 min exercise 40% RH 10 M
15 min rest
VE ~ 20 L/min
3 h IE 22 °C 10 M
VE = 30 L/min 50% RH 6 M
COPD patients
Age-matched healthy NS
All subjects 59-71 years old
Hypertension
42-61 years old
Healthy
41-49 years old
No significant changes in FEVj, FVC, or SRaw due to
ozone in COPD patients. Equivocal SaO2 decrement
during 2nd and 3rd hours of ozone exposure in COPD
patients. Adjusted for exercise, ozone effects did not
differ significantly between COPD patients and healthy
subjects.
No major cardiovascular effects in either healthy or
hypertensive subjects.
Gong et al.
(1997a)
Gong and
Tierney (1995)
Gong et al.
(1998)
Subjects with Allergic Rhinitis
0.0 0
0.2 392
0.125 245
0.250 490
0.125 245
0.0 0
0.25 490
1 h CE 20 °C 13 M, 1 F
at VE = 25 L/min/m2 50% RH
BSA
3h IE 27 °C 5 F, 6 M
(10 min rest, 15 min 50 % RH
exercise on bicycle) 6 F, 16 M
VE = 30 L/min
3h IE x 4 days
3hIE, 27 °C 13 M, 11 F
VE = 30 L/min 54% RH
15 min ex/10 min mouthpiece 6 M, 6 F
rest/5 min no O3; exposure
every 30 min.
5M, 5F
Dust mite sensitized asthmatics
mean age 29 ± 5 years
Mild bronchial asthma
20-53 years old
Allergic rhinitis
19-48 years old
Atopic mild asthma
Positive allergen and IgE tests
Healthy NS
FEVj decrement following O3 of 10% not significantly
different from the 4% decrement following FA. Subjects
received dust mite antigen challenge at 0.5 h FA and O3
postexposures and were lavaged 6 h post-challenge.
Amount of allergen producing 1 5% FEVj decrement was
decreased by O3 compared to FA in 9 of 14 subjects.
PMN in proximal airway lavage tended to be greater after
O3 than FA (p=0.06).
Mean early-phase FEVj response and number of > 20%
reductions in FEVj were significantly greater after
0.25 ppm O3 or 4 x 0.125 ppm O3. Most of the > 15%
late-phase FEVj responses occurred after 4 days of
exposure to 0.125 ppm O3, as well as significant
inflammatory effects, as indicated by increased sputum
eosinophils (asthma and allergic rhinitis) and increased
sputum lymphocytes, mast cell tryptase, histamine, and
LDH (asthma only).
O3-induced FEVj decrements of 12.5, 14. 1, and 10.2%
in asthmatics, allergic rhinitics and healthy subjects,
respectively (group differences not significant).
Methacholine responsiveness increased in asthmatics.
Allergen responsiveness: increased significantly after O3
exposure in asthmatics (~ 2 dose shift) and a smaller shift
is rhinitics. No change in healthy. Neither allergen or
methacholine response correlated with lung function and
were not correlated with each.
Chen et al.
(2004)
Holz et al.
(2002)
Jorres et al.
(1996)
-------
o
O
to
O
O
Table AX6-3 (cont'd). Ozone Exposure in Subjects with Preexisting Disease"
X
to
oo
H
6
o
o
H
O
O
HH
H
W
Ozone
Concentration'
ppni fig/m3
Exposure Duration Exposure
and Activity Conditions
Number and
Gender of Subjects Subject Characteristics
Observed Effect(s)
Reference
Subjects with Asthma
0.4 784
0.0 0
0.4 784
0.12 236
0.0 0
0.2 392
0.4 784
0.0 0
0.12 235
0.4 784
2h IE NA
(15 min rest, 15 min exercise
on bicycle)
VE = 30 L/min
2h IE NA
4x15 min on bicycle,
VE = 40 L/min
Rest 22 °C
40% RH
6 h 22 °C
30 min rest/30 min exercise 50% RH
VE ~ 25 L/min
3h 6x15 min cycle ergometer 31 °C
VE » 32L/min 35% RH
5 consecutive days
1 h rest NA
air-antigen
O3-antigen
2 h IE Head mask
15 min exercise exposure
15 min rest »18°C
VE » 20L/min 60% RH
4 F, 5 M Healthy
(25 ± 2 years old)
7 F, 6 M Mild atopic asthma;
beta agonists only
(22 ± 0.7 years old)
15 Healthy adults
18-40 years old
9 Mild atopic asthmatics
18-40 years old
10 M, 5 F atopic asthma
5 M Healthy NS
5 M Asthmatics, physician
diagnosed,
All 18-45 years
8 M , 2 F Asthmatic NS adults
beta-agonist use only
19-48 years old
ATS criteria for asthma
9 M, 6 F Mild allergic asthma; 18 to
49 years of age.
5 M, 1 F Healthy adults
6 M Atopic asthmatics
Significant reductions in FVC (12%, 10%) and
FEV[ (13%, 1 1%) for asthmatic and healthy
subjects, respectively; attenuated by indomethacin
in healthy subjects only. Significant reductions in
mid-flows which tended to be greater in asthmatics
than healthy subjects. Indomethacin treatment
attenuated mid-flow-reductions somewhat more
in asthmatics than healthy subjects.
Sputum collected 24 h before and 4-6 h post O3
exposure. Baseline GDI Ib expression positively
correlated with O3-induced PMN. Increased
expression of mCD14 on macrophages following
O3 compared to FA. Asthmatic PMN response
similar to healthy subjects (also see Table AX6-3).
No spirometric data available.
No effect of O3 on airway response to grass
allergen.
Similar spirometric responses in asthmatic and
healthy. However, preexposure FEV1 and FVC
were both -0.4 L lower on O3-day than FA day.
More PMN's in asthmatics. IL-8 and IL-6 higher
in asthmatics exposed to O3. No relationship of
spirometry and symptoms to inflammation.
FEVj decreased 35% on first exposure day.
Methacholine reactivity increased about ten-fold.
Also see Table AX6- 7 for repeated exposure results.
No effect of O3 on airway response to grass or
ragweed allergen.
FEVj responses of healthy and asthmatic similar
(~ 15% decrease). Maximal FEVj response to
methacholine increased similarly in both groups
(12 h postexposure). Larger increase in PC20 in
healthy subjects. Both groups had increased
PMN's in sputum no correlation of PMN's
and lung function.
Alexis et al.
(2000)
Alexis et al.
(2004)
Ball et al.
(1996)
Basha et al.
(1994)
Gong et al.
(1997b)
Hanania
etal. (1998)
Hiltermann
etal. (1995)
-------
o
O
to
O
O
Table AX6-3 (cont'd). Ozone Exposure in Subjects with Preexisting Disease"
X
to
VO
H
6
o
o
H
O
Ozone
Concentration'
ppm ug/m
Number and
Exposure Duration and Exposure Gender of
Activity Conditions Subjects
Subject
Characteristics
Observed Effect(s)
Reference
Adult Subjects with Asthma (cont'd)
0.0 0
0.16 314
0.0 0
0.25 490
0.16 314
0.25 490
0.40 784
0.0 0
0.2 392
0.2 396
7.6 h 18° C 13 M
25 min treadmill, 40% RH
25 min cycle/10 min rest 7 M, 10 F
per hour.
VE =27-32 L/min
3hIE, 27 °C 13 M, 11 F
VE = 30 L/min 54% RH
15 min ex/10 min rest/5 mouthpiece 6 M, 6 F
min no O3; every 30 min. exposure
5M, 5F
7.6 h 22 °C 4 M, 5 F
25 min treadmill, 25 min 40 % RH
cycle/10 min rest per
hour.
VE = 25 L/min
VE = 25-45 L/min NA 8 M, 4 F
8 M, 10 F
22 M, 16 F
2 h IE 20 °C 6 M, 9 F
4x15 min 50% RH
at VE = 20
L/min/m2 BSA 9 M, 6 F
2h IE 22 °C 5 F, 4 M
(1 5 min rest, 1 5 min 40 % RH
exercise on bicycle)
VE = 20 L/min/m2 BSA
Healthy NS,
age 19-32 years.
Moderate Asthmatics,
physician diagnosed,
beta agonist users,
age 19-32 years.
Atopic mild asthma
Positive allergen and
IgE tests
Healthy NS
Mild atopic asthma;
no meds 12 hpre-
exposure
20-35 years old
Asthmatics
Allergic rhinitics
Healthy adults
All < 26 years old
Healthy adults
24 years old
Mild asthmatics
29 years old
Mild atopic asthma; no
meds 8 h pre-exposure
21-42 years old
FEVj decreased 19% in asthmatics and only 10% in
non-asthmatics. High responders had worse baseline airway
status. More wheeze in asthmatics after O3.
O3-induced FEVj decrements of 12.5, 14.1, and 10.2% in
asthmatics, allergic rhinitics and healthy subjects, respectively
(group differences not significant). Methacholine
responsiveness increased in asthmatics. Allergen responsiveness
increased after O3 exposure in asthmatics (= 2 dose shift),
a smaller shift occurred in rhinitics, no change occurred in
healthy subjects. Neither allergen nor methacholine responses
were correlated with each other or with lung function.
Significant FEVj decrease of 9.1 % following O3 exposure;
marked individual variability with responses ranging from 2
% to 26 %.
Healthy 12.2% decrease in FEVj, Rhinitics 10.1%,
asthmatics 12.4%
O3-induced FEV; decrement (8%, healthy adults; 3% asthmatics)
and PMN increase (20.6%, healthy adults; 15.2% asthmatics).
Primary goal was to investigate relationship between antioxidant
defenses and O3 responses in asthmatics and healthy adults
(see Tables AX6-3 andAX6 -13).
Significant decrease in FEVj and a trend toward decreases in
mean inspiratory flow, FEF25, and FEF75 after O3 exposure.
No significant differences in FEF50, FVC, TLC, Raw, or sRaw.
No correlation between sputum neutrophils at 6 h postexposure
and FEVj immediately after exposure.
Horstman
etal. (1995)
Jorres et al.
(1996)
Kehrl et al.
(1999)
Magnussen
etal. (1994)
Mudway
etal. (2001)
Stenfors
et al. (2002)
Newson
et al. (2000)
O
HH
H
W
-------
O
o
i
{_} '
CD
>-l
(^
to
O
O
•^
J>
1
O
Table AX6-3 (cont
Ozone
Concentration'
Exposure Duration Exposure
ppm Hg/ni3 and Activity Conditions
Adult Subjects with Asthma (cont'd)
0.4 784 2 h rest 21 "C
40% RH
0.16 314 7.6 h 18° C
25 min treadmill, 40% RH
25 min cycle/ every
hour.
0.0 0 4h 21 °C
0.2 392 50 min exercise, 50% RH
10 min rest each hour.
VE » 45-50 L/min
"See Appendix A for abbreviations and acronyms.
f~li-r*ii«fiH hw rf^ct 3iiH f^vfMvicf*' Axjitliiii rmr>f*ntr3tir*ii
with Preexisting Disease"
Observed Effect(s)
Ozone resulted in nasal inflammation (increased PMN's)
and caused augmented response to nasal allergen challenge.
Increased eosinophils and PMN's after O3 exposure more
in initial (bronchial) fraction. No correlation of eosinophils
and PMN's, FEVj & FVC decreased 14% and 9%
respectively.
FVC, FEVj decreased 17.6% and 25% respectively.
Trend for larger increase in SRaw in asthmatics. Larger
increase in PMN's and protein in asthmatics indicating
more inflammation. No increase in eosinophils.
Spirometry changes in asthmatics similar to healthy
subjects (Aris et al., 1995; Balmes et al., 1997).
Reference
Peden et al.
(1995)
Peden et al.
(1997)
Scannell et al.
(1996)
H
6
o
o
H
O
O
HH
H
W
-------
1 More recently, Gong et al. (1997a) exposed 9 COPD patients (age range, 59 to 71 years;
2 mean age 66 ± 4 years) and 10 healthy NS (age range, 60 to 69 years; mean age 65 ± 3 years)
3 to 0.24 ppm for 4 h with interment light exercise («20 L/min). COPD patients had decreases in
4 FEVj following both clean air (-11%, p = 0.06) and O3 (-19%, p < 0.01) exposures. These
5 FEVj decrements, presumably due to exercise, were primarily attributable to four of the patients
6 who lost greater than 14% of their FEVj following both the air and O3 exposures. Relative to
7 clean air, O3 caused a statistically insignificant FEVj decrement of - 8% in COPD patients which
8 was not statistically different from the decrement of -3% in healthy subjects. Ozone-induced
9 symptoms, sRaw, SaO2, and postexposure bronchial activity also exhibited little or no difference
10 between the COPD patients and the healthy subjects.
11
12 AX6.3.2 Subjects with Asthma
13 Based on studies reviewed in the 1996 criteria document (U.S. Environmental Protection
14 Agency, 1996) asthmatics appear to be at least as sensitive to acute effects of O3 as healthy
15 nonasthmatic subjects. At rest, neither adolescent asthmatics nor healthy controls had significant
16 responses as a result of an hour exposure to 0.12 ppm O3. Exposure of adult asthmatics to 0.25
17 ppm O3 for 2 h at rest also caused no significant responses. Preexposure to between 0.10 and
18 0.25 ppm O3 for 1 hr with light IE does not appear to exacerbate exercise-induced asthma
19 (Fernandes et al., 1994; Weymer et al., 1994). At higher exposures (0.4 ppm O3 with heavy IE
20 for 2 h), Kreit et al. (1989) and Eschenbacher et al. (1989) demonstrated significantly greater
21 FEVj and FEF25.75 decrements in asthmatics than in healthy controls. With longer duration
22 exposures to lower O3 levels (0.12 ppm with moderate IE for 6.5 h), asthmatics have also shown
23 a tendency for greater FEVj decrements than healthy non-asthmatics (Linn et al., 1994). Newer
24 studies (see Table AX6-3) continue to suggest that asthmatics are at least as sensitive as healthy
25 controls to O3-induced responses.
26 Studies of less than 3 h duration have reported similar or tendencies for increased O3-
27 induced spirometric responses up to O3 concentrations of 0.4 ppm. Similar group decrements in
28 FEVj and FVC were reported by Hiltermann et al. (1995), who exposed 6 asthmatics and
29 6 healthy subjects to 0.4 ppm O3 for 2 h with light IE. Alexis et al. (2000) exposed 13 mild
30 atopic asthmatics and 9 healthy subjects for 2 h to 0.4 ppm O3 with IE ( VE = 30 L/min). Similar
31 O3-induced group decrements in FEVj and FVC were also reported by these investigators.
January 2005 AX6-31 DRAFT-DO NOT QUOTE OR CITE
-------
1 A tendency, however, for an increased O3-induced reduction in mid-flows (viz., FEF25, FEF50,
2 FEF60p, FEF75) was reported for the asthmatics relative to the healthy subjects. In a larger study,
3 Torres et al. (1996) exposed 24 asthmatics, 12 allergic rhinitis, and 10 healthy subjects to
4 0.25 ppm O3 for 3 h with IE. Statistically significant O3-induced decreases in FEVj occurred in
5 all groups, but tended to be lower in healthy controls (allergic rhinitis, -14.1%; asthmatics,
6 -12.5%; healthy controls, -10.2%). One study reported that asthmatics tended to have less of an
7 FEVj response to O3 than healthy controls (Mudway et al., 2001). In that study, however, the
8 asthmatics also tended to be older than the healthy subjects which could partially explain their
9 lesser response.
10 Studies between 4 and 8 h duration, with O3 concentrations of 0.2 ppm or less, also suggest
11 a tendency for increased O3-induced pulmonary function responses in asthmatics relative to
12 healthy subjects. Scannell et al. (1996) exposed 18 asthmatics to 0.2 ppm O3 for 4 h with IE
13 (VE -25 L/min/m2 BSA). Baseline and hourly pulmonary function measurements of FEVl3
14 FVC, and sRaw were obtained. Asthmatic responses were compared to 81 healthy subjects who
15 underwent similar experimental protocols (Aris et al., 1995; Balmes et al., 1996). Asthmatic
16 subjects experienced a significant O3-induced increase in sRaw, FEVj and FVC. The O3-induced
17 increase in sRaw tended to be greater in asthmatics than the healthy subjects, whereas similar
18 group decrements in FEVj and FVC were observed. Basha et al. (1994) also reported similar
19 spirometric responses between 5 asthmatic and 5 healthy subjects exposed to 0.2 ppm O3 for 6 h
20 with IE. However, the mean preexposure FEVj in the asthmatics was about 430 mL less (i.e.,
21 ~ 12% decreased) on the O3-day relative to the air-day. In a longer exposure duration (7.6 h)
22 study, Horstman et al. (1995) exposed 17 asthmatics and 13 healthy controls to 0.16 ppm O3 or
23 FA with alternating periods of exercise (50 min, VE -30 L/min) and rest (10 min). Both groups
24 had significant O3-induced decrements in FEVl3 FVC, and FEV25_75. The asthmatic and healthy
25 subjects had similar O3-induced reductions in FVC. The FEVj decrement experienced by the
26 asthmatics was significantly greater in the healthy controls (19% versus 10%, respectively).
27 There was also tendency for a greater O3-induced decrease in FEF25.75 in asthmatics relative to
28 the healthy subjects (24% versus 15%, respectively).
29 With repeated O3 exposures asthmatics, like healthy subjects (see Section AX6.6) develop
30 tolerance. Gong et al. (1997b) exposed 10 asthmatics to 0.4 ppm O3, 3 h per day with IE
31 ( VE -32 L/min), for 5 consecutive days. Symptom and spirometric responses were greatest on
January 2005 AX6-32 DRAFT-DO NOT QUOTE OR CITE
-------
1 the first (-35 % FEVj) and second (-34 % FEVj) exposure days, and progressively diminished
2 toward baseline levels (-6 % FEVj) by the fifth exposure day. Similar to healthy subjects,
3 asthmatics lost their tolerance 4 and 7 days later.
4 Other published studies with similar results (e.g., McBride et al., 1994; Basha et al., 1994;
5 Peden et al., 1995, 1997; Peden, 2001a; Scannell et al., 1996; Hiltermann et al., 1997, 1999;
6 Michelson et al., 1999; Vagaggini et al., 1999; Newson et al., 2000; Holz et al., 2002) also
7 reported that asthmatics have a reproducible and somewhat exaggerated inflammatory response
8 to acute O3 exposure (see Section AX6.9}. For instance, Scannell et al. (1996) performed lavages
9 at 18 h post O3 exposure to assess inflammatory responses in asthmatics. Asthmatic responses
10 were compared to healthy subjects who underwent a similar experimental protocol (Balmes
11 et al., 1996). Ozone-induced increases in BAL neutrophils and total protein were significantly
12 greater in asthmatics than healthy subjects. There was also a trend for an ozone related increased
13 IL-8 in the asthmatics relative to healthy subjects. Inflammatory responses do not appear to be
14 correlated with lung function responses in either asthmatic or healthy subjects (Balmes et al.,
15 1996, 1997; Holz et al., 1999). This lack of correlations between inflammatory and spirometric
16 responses may be due to differences in the time kinetics of these responses (Stenfors et al.,
17 2002). In addition, airway responsiveness to inhaled allergens is increased by O3 exposure in
18 subjects with allergic asthma for up to 24 h (see Section AX6.8).
19 One of the difficulties in comparing O3-induced spirometric responses of healthy subjects
20 versus asthmatics is the variability in responsiveness of asthmatics. Most of the asthma studies
21 were conducted on subjects with a clinical history of mild disease. However, classification
22 asthma severity is not only based on functional assessment (e.g., percent predicted FEVj), but
23 also on clinical symptoms, signs, and medication use (Table AX6-4). Although "mild atopic
24 asthmatics" are frequently targeted as an experimental group, the criteria for classification has
25 varied considerably within and across the available published studies. Although the magnitude
26 of group mean changes in spirometry may not be significantly different between healthy and
27 asthmatic subjects, many of the studies have reported clinically significant changes in some
28 individuals.
29 Alexis et al. (2000) explored the possibility that the mechanisms of O3-induced spirometric
30 responses may differ between asthmatics and healthy subjects. Physician-diagnosed mild atopic
31 asthmatics and healthy subjects were pretreated with 75 mg/day of indomethacin (a COX
January 2005 AX6-33 DRAFT-DO NOT QUOTE OR CITE
-------
Table AX6-4. Classification of Asthma Severity1
to
o
o
X
Classification
Severe
persistent
Moderate
persistent
Days with
Step symptoms
4 Continual
3 Daily
Nights with
symptoms
Frequent
> I/week
Lung Function2
FEVlorPEF PEF
% predicted variability
oral (%)
< 60 > 30
between > 30
60 and 80
Medications3
Daily
High-dose inhaled steroids (ICS)
and long-acting inhaled p2-agonist
If needed, add oral steroids
Low-to-medium-dose ICS and
long-acting p2-agonist (preferred)
Quick relief
Short-acting inhaled
p2-agonist, as needed; oral
steroids may be required
Short-acting inhaled
p2-agonist, as needed;
Mild
persistent
> 2/week,
but< 1
time/day
> 2/week
Mild
intermittent
< 2/week < 2/month
Or
Medium-dose ICS (another preferred
option for children ages < 5 years)
Or
Low-to-medium-dose ICS and either
leukotriene modifier or theophylline
80 20-30 Low-dose inhaled steroids (preferred)
Or
Cromolyn leukotriene modifier,
or (except for children aged < 5 years)
nedocromil or sustained release
theophylline to serum concentration
of 5-15 ug/mL
80 < 20 No daily medicine needed
oral steroids may be
required
Short-acting inhaled
p2-agonist, as needed;
oral steroids may be
required
Short-acting inhaled
p2-agonist, as needed;
oral steroids may be
required
1 Sources: Centers for Disease Control (2003); National Heart, Lung, and Blood Institute (1997, 2003).
2For adults and children aged > 5 years who can use a spirometer or peak flow meter.
3The medications listed here are appropriate for treating asthma at different levels of severity. The preferred treatments, dosage, and type of medication
recommended vary for adults and children and are detailed in the EPR-Update 2002 stepwise approach to therapy. The stepwise approach emphasizes that
therapy should be stepped up as necessary and stepped down when possible to identify the least amount of medication required to achieve goals of therapy.
The stepwise approach to care is intended to assist, not replace, the clinical decision-making required to meet individual patient needs.
-------
1 inhibitor) or placebo and then exposed for 2 h to 0.4 ppm O3 or to FA during mild IE
2 (VE =30L/m). The number and severity of O3-induced symptoms were significantly increased
3 in both asthmatics and healthy subjects. These symptom responses were similar between the
4 subject groups and unaffected by indomethacin pretreatment. Asthmatics and healthy subjects
5 also had similar O3-induced reductions in FVC and FEVj. These restrictive-type responses,
6 occurring due to the combined effects of bronchoconstriction and reflex inhibition of inspiration
7 (see Section AX6.2.7), were attenuated by indomethacin in the healthy subj ects but not the
8 asthmatics. Thus, in healthy subjects but not asthmatics, COX metabolites may contribute to
9 O3-induced reductions in FVC and FEVj. As assessed by the magnitude of reductions in
10 mid-flows (viz. FEF25, FEF50, FEF60p, FEF75), the small airways of the asthmatics tended to be
11 more affected than the healthy subjects. This suggests asthmatics may be more sensitive to
12 small airway effects of O3, which is consistent with the observed increases in inflammation and
13 airway responsiveness. Indomethacin pretreatment attenuated some of these O3-induced small
14 airways effects (FEF50 in healthy subjects, FEF60p in asthmatics).
15
16 AX6.3.3 Subjects with Allergic Rhinitis
17 Most O3 exposure studies in humans with existing respiratory disease have focused on lung
18 diseases like COPD and asthma. However, chronic inflammatory disorders of the nasal airway,
19 especially allergic rhinitis, are very common in the population. People with allergic rhinitis have
20 genetic risk factors for the development of atopy that predispose them to increased upper airway
21 responsiveness to specific allergens as well as nonspecific air pollutants like O3. Studies
22 demonstrating the interaction between air pollutants and allergic processes in the human nasal
23 airways and rhinoconjunctival tissue have been reviewed by Peden (2001b) and Riediker et al.
24 (2001), respectively. Ozone exposure of subjects with allergic rhinitis has been shown to induce
25 nasal inflammation and increase airway responsiveness to nonspecific bronchoconstrictors,
26 although to a lesser degree than experienced by asthmatics.
27 McDonnell et al. (1987) exposed non-asthmatic adults with allergic rhinitis to 0.18 ppm
28 O3. The allergic rhinitics were no more responsive to O3 than healthy controls, based on
29 symptoms, spirometry, or airway reactivity to histamine although they had a small but
30 significantly greater increase in SRaw. The data on subjects with allergic rhinitis and asthmatic
January 2005 AX6-3 5 DRAFT-DO NOT QUOTE OR CITE
-------
1 subjects suggest that both of these groups have a greater rise in Raw to O3 with a relative order
2 of airway responsiveness to O3 being normal < allergic < asthmatic.
3 Bascom et al. (1990) studied the upper respiratory response to acute O3 inhalation, nasal
4 challenge with antigen, and the combination of O3 plus antigen in subjects with allergic rhinitis.
5 Exposure to O3 caused significant increases in upper and lower airway symptoms, a mixed
6 inflammatory cell influx with a seven-fold increase in nasal lavage PMNs, a 20-fold increase in
7 eosinophils, and a 10-fold increase in mononuclear cells, as well as an apparent sloughing of
8 epithelial cells. McBride et al. (1994) also observed increased nasal PMN's after O3 exposure in
9 atopic asthmatics. Peden et al. (1995), who studied allergic asthmatics exposed to O3 found that
10 O3 causes an increased response to nasal allergen challenge in addition to nasal inflammatory
11 responses. Their data suggested that allergic subjects have an increased immediate response to
12 allergen after O3 exposure. In a follow-up study, Michelson et al. (1999) reported that 0.4 ppm
13 O3 did not promote early-phase-response mediator release or enhance the response to allergen
14 challenge in the nasal airways of mild, asymptomatic dust mite-sensitive asthmatic subjects.
15 Ozone did, however, promote an inflammatory cell influx, which helps induce a more significant
16 late-phase response in this population.
17 Torres et al. (1996) found that O3 causes an increased response to bronchial allergen
18 challenge in subjects with allergic rhinitis. This study also compared responses in subjects with
19 mild allergic asthma (see Sections AX6.3.2 andAX6.8). The subjects were exposed to 0.25 ppm
20 O3 for 3 h with IE. Airway responsiveness to methacholine was determined 1 h before and after
21 exposure; responsiveness to allergen was determined 3 h after exposure. Statistically significant
22 decreases in FEVj occurred in subjects with allergic rhinitis (13.8%) and allergic asthma
23 (10.6%), and in healthy controls (7.3%). Methacholine responsiveness was statistically
24 increased in asthmatics, but not in subjects with allergic rhinitis. Airway responsiveness to an
25 individual's historical allergen (either grass and birch pollen, house dust mite, or animal dander)
26 was significantly increased after O3 exposure when compared to FA exposure. In subjects with
27 asthma and allergic rhinitis, a maximum percent fall in FEVj of 27.9 % and 7.8%, respectively,
28 occurred 3 days after O3 exposure when they were challenged with of the highest common dose
29 of allergen. The authors concluded that subjects with allergic rhinitis, but without asthma, could
30 be at risk if a high O3 exposure is followed by a high dose of allergen.
January 2005 AX6-36 DRAFT-DO NOT QUOTE OR CITE
-------
1 Holz et al. (2002) extended the results of Torres et al. (1996) by demonstrating that
2 repeated daily exposure to lower concentrations of O3 (0.125 ppm for 4 days) causes an
3 increased response to bronchial allergen challenge in subjects with preexisting allergic airway
4 disease, with or without asthma. There was no major difference in the pattern of bronchial
5 allergen response between subjects with asthma or rhinitis, except for a 10-fold increase in the
6 dose of allergen required to elicit a similar response (> 20% decrease in FEVj) in the asthmatic
7 subjects. Early phase responses were more consistent in subjects with rhinitis and late-phase
8 responses were more pronounced in subjects with asthma. There also was a tendency towards a
9 greater effect of O3 in subjects with greater baseline response to specific allergens chosen on the
10 basis of skin prick test and history (viz., grass, rye, birch, or alder pollen, house dust mite, or
11 animal dander). These data suggest that the presence of allergic bronchial sensitization, but not a
12 history of asthma, is a key determinant of increased airway allergen responsiveness with O3.
13 [A more complete discussion of airway responsiveness is found in Section AX6.8]
14
15 AX6.3.4 Subjects with Cardiovascular Disease
16 Superko et al. (1984) exposed six middle-aged males with angina-symptom-limited
17 exercise tolerance for 40 min to FA and to 0.2 and 0.3 O3 while they were exercising
18 continuously according to a protocol simulating their angina-symptom-limited exercise training
19 prescription (meanVE= 35 L/min). No significant pulmonary function impairment or evidence
20 of cardiovascular strain induced by O3 inhalation was observed. Gong et al. (1998) exposed
21 hypertensive and healthy adult males, 41 to 78 years of age, to 0.3 ppm O3 for 3 h with IE at
22 30 L/min. The ECG was monitored by telemetry, blood pressure by cuff measurement, and a
23 venous catheter was inserted for measurement of routing blood chemistries and cardiac ezymes.
24 Pulmonary artery and radial artery catheters were placed percutaneously for additional blood
25 sampling and for measurement of hemodynamic pressures, cardiac output, and SaO2. Other
26 hemodynamic variables were calculated, including cardiac index, stroke volume, pulmonary and
27 systemic vascular resistance, left and right ventricular stroke-work indices, and rate-pressure
28 product. Spirometric volumes (FVC, FEVj) and respiratory symptoms were measured before
29 and after the O3 exposures. The overall results did not indicate any major acute cardiovascular
30 effects of O3 in either the hypertensive or normal subjects. Statistically significant O3 effects for
31 both groups combined were a decrease in FEVj and increases in FIR, rate-pressure product, and
January 2005 AX6-37 DRAFT-DO NOT QUOTE OR CITE
-------
1 the alveolar-to-arterial PO2 gradient, suggesting that impaired gas exchange was being
2 compensated for by increased myocardial work. These effects might be more important in some
3 patients with severe cardiovascular disease.
4
5
6 AX6.4 INTERSUBJECT VARIABILITY AND REPRODUCIBILITY
7 OF RESPONSE
8 Analysis of the factors that contribute to intersubject variability is important for the
9 understanding of individual responses, mechanisms of response, and health risks associated with
10 acute O3 exposures. Bates et al. (1972) noted that variation between individuals in sensitivity
11 and response was evident in respiratory symptoms and pulmonary function following O3
12 exposure. A large degree of intersubject variability in response to O3 has been consistently
13 reported in the literature (Adams et al., 1981; Aris et al., 1995; Folinsbee et al., 1978; Kulle
14 et al., 1985; McDonnell et al., 1983). Kulle et al. (1985) noted that the magnitude of variability
15 between individuals in FEVj responses increases with O3 concentration. Similarly, McDonnell
16 et al. (1983) observed FEVj decrements ranging from 3 to 48% (mean 18%) in 29 young adult
17 males exposed to 0.40 ppm O3 for 2 h during heavy IE. At a lower O3 concentration of
18 0.18 ppm, 20 similarly exposed subjects had FEVj decrements ranging from 0 o 23%
19 (mean = 6%), while those exposed to FA (n = 20) had decrements ranging from -2% to 6%
20 (mean = 1%) (McDonnell et al., 1983). All of the subjects in these studies were young adult
21 males. (Intersubject variability related to age and gender is discussed in Sections AX6.5.1 and
22 AX6.5.2, respectively.)
23 More recently, McDonnell (1996) examined the FEVj response data from three 6.6 h
24 exposure studies of young adult males conducted at the EPA Health Effects Research Laboratory
25 in Chapel Hill, NC (Folinsbee et al., 1988; Horstman et al., 1990; McDonnell et al., 1991).
26 The response distributions for subjects at each of four O3 concentrations (0.0, 0.08, 0.10, and
27 0.12 ppm) are illustrated in Figure AX6-6. It is apparent that the FEVj responses in FA are
28 small with most tightly grouped around zero. With increasing O3 concentration, the mean
29 response increases as does the variability about the mean. At higher O3 concentrations, the
30 distribution of response becomes asymmetric with a few individuals experiencing large FEVj
31 decrements. The response distribution in Figure AX6-6 allows estimates of the number or
January 2005 AX6-3 8 DRAFT-DO NOT QUOTE OR CITE
-------
tn
+j
o
O
+J
c
o
40-
30-
0.12 ppm
-15-10 -5 0 5 10 15 20 25 30 35 40 45 50
50-
30-
0.10 ppm
i . . . n
-15-10 -5 0 5 10 15 20 25 30 35 40 45 50
40-
30-
20-
10-
0.08 ppm
II I ni—in
-15-10 -5 0 5 10 15 20 25 30 35 40 45 50
40-
30-
20-
10-
n-
PI
~ll
I — I
0 ppm
n
I
-15-10 -5 0 5 10 15 20 25 30 35 40 45 50
FEV-, (% Decrement)
Figure AX6-6. Frequency distributions of percent decrements in FEVt for 6.6-h exposure
to four concentrations of ozone.
Source: McDonnell (1996).
January 2005
AX6-39 DRAFT-DO NOT QUOTE OR CITE
-------
1 percentage of subjects responding in excess of a certain level. With FA exposure, none of 87
2 subjects had a FEVj decrement in excess of 10%; however, 26%, 31%, and 46% exceeded a 10%
3 decrement at 0.08, 0.10, and 0.12 ppm, respectively. FEVj decrements as large as 30 to 50%
4 were even observed in some individuals. In 6.6-h face mask exposures of young adults (half
5 women) to 0.08 ppm O3, Adams (2002) found that 6 of 30 subjects (20%) had > 10% decrements
6 in FEVj. The response distributions in Figure AX6-6 underlines the wide range of response to
7 O3 under prolonged exposure conditions and reinforces the observations by others consequent to
8 2 h IE exposures at higher O3 concentrations (Horvath et al.,1981; McDonnell et al., 1983).
9 Some of the intersubject variability in response to O3 inhalation may be due to intrasubject
10 variability, i.e., how reproducible the measured responses are in an individual between several
11 O3 exposures. The more reproducible the subject's response, the more precisely it indicates
12 his/her intrinsic responsiveness. McDonnell et al. (1985a) examined the reproducibility of
13 individual responses to O3 in healthy human subjects (n = 32) who underwent repeated
14 exposures within a period of 21 to 385 days (mean = 88 days; no median reported) at one of five
15 O3 concentrations ranging from 0.12 to 0.40 ppm. Reproducibility was assessed using the
16 intraclass correlation coefficient (R). The most reproducible responses studied were FVC
17 (R = 0.92) and FEVj (R = 0.91). However, at the lowest concentration, 0.12 ppm, relatively
18 poor FEVj reproducibility was observed (R = 0.58) due, in part, to a lack of specific O3 response
19 or a uniformly small response in the majority of subjects. McDonnell et al. (1985a) concluded
20 that for 2 h IE O3 exposures equal to or greater than 0.18 ppm, the intersubject differences in
21 magnitude of change in FVC and FEVj are quite reproducible over time and likely due to
22 differences in intrinsic responsiveness of individual subjects. Hazucha et al. (2003) exposed
23 47 subjects on three occasions for 1.5 h, with moderate intensity IE, to 0.40 to 0.42 ppm O3.
24 Reproducibility of FEVj responses was related to the length of time between re-exposures,
25 with a Spearman correlation R of 0.54 obtained between responses for exposures 1 and
26 2 (median = 105 days), and an R of 0.85 between responses for exposures 2 and 3
27 (median = 7 days).
28 Identification of mechanisms of response and health risks associated with acute O3
29 exposures are complicated by a poor association between various O3-induced responses.
30 For example, McDonnell et al. (1983) observed a very low correlation between changes in sRaw
31 and FVC (r = -0.16) for 135 subjects exposed to O3 concentrations ranging from 0.12 to
January 2005 AX6-40 DRAFT-DO NOT QUOTE OR CITE
-------
1 0.40 ppm for 2.5 h with IE. In a retrospective study of 485 male subjects (ages 18 to 36 yrs)
2 exposed for 2 h to one of six O3 concentrations at one of three activity levels, McDonnell et al.
3 (1999) observed significant, but low, Spearman rank order correlations between FEVj response
4 and symptoms of cough (R = 0.39), shortness of breath (R = 0.41), and pain on deep inspiration
5 (R = 0.30). The authors concluded from their data that the O3-induced responses are related
6 mechanistically to some degree, but that there is not a single factor which is responsible for the
7 observed individual differences in O3 responsiveness across the spectrum of symptom and lung
8 function responses. This conclusion is supported by differences in reproducibility observed by
9 McDonnell et al., (1985a). Compared to the intraclass correlation coefficient for FEVj (R =
10 0.91), relatively low but statistically significant R values for symptoms ranged from 0.37 to 0.77,
11 with that for sRaw being 0.54. The reproducibility correlations for fB (R = -0.20) and VT
12 (R = -0.03) were not statistically significant.
13 The effect of this large intersubject variability on the ability to predict individual
14 responsiveness to O3 was demonstrated by McDonnell et al. (1993). These investigators
15 analyzed the data of 290 male subjects (18 to 32 years of age) who underwent repeat 2 h IE
16 exposures to one or more O3 concentrations ranging from 0.12 to 0.40 ppm in order to identify
17 personal characteristics (i.e., age, height, baseline pulmonary functions, presence of allergies,
18 and past smoking history) that might predict individual differences in FEVj response. Only age
19 contributed significantly to intersubject responsiveness (younger subjects were more
20 responsive), accounting for just 4% of the observed variance. Interestingly, O3 concentration
21 accounted for only 31% of the variance, strongly suggesting the importance of as yet undefined
22 individual characteristics that determine FEVj responsiveness to O3. A more general form of
23 this model was developed to investigate the O3 exposure FEVj response relationship (McDonnell
24 et al., 1997). These authors used data from 485 male subjects (age = 18 to 36 years) exposed
25 once for 2 h to one of six O3 concentrations (ranging from 0.0 to 0.40 ppm) at one of 3 activity
26 levels (rest, n= 78; moderate IE, n = 92; or heavy IE, n = 314). In addition to investigating the
27 influence of subject's age, the model focused on determining whether FEVj response was more
28 sensitive to changes in C than to changes in VE, and whether the magnitude of responses is
29 independent of differences in lung size. It was found that the unweighted proportion of the
30 variability in individual responses explained by C, VE, T, and age was 41%, with no evidence
31 that the sensitivity of FEVj response to VE was different than changes in C, and no evidence that
January 2005 AX6-41 DRAFT-DO NOT QUOTE OR CITE
-------
1 magnitude of response was related to measures of body or lung size. The authors concluded that
2 much inter-individual variability in FEVj response to O3 remains unexplained.
3
4
5 AX6.5 INFLUENCE OF AGE, GENDER, ETHNIC, ENVIRONMENTAL
6 AND OTHER FACTORS
7 AX6.5.1 Influence of Age
8 On the basis of results reported from epidemiologic studies, children and adolescents are
9 considered to be at increased risk, but not necessarily more responsive, to ambient oxidants than
10 adults. However, findings of controlled laboratory studies that have examined the acute effects
11 of O3 on children and adolescents do not completely support this assertion (Table AX6-5).
12 Children experience about the same decrements in spirometric endpoints as young adults
13 exposed to comparable O3 doses (McDonnell et al., 1985b; Avol et al., 1987). In contrast to
14 young adults, however, they had no symptomatic response, which may put them at an increased
15 risk for continued exposure. Similarly, young adults (Linn et al., 1986; Avol et al., 1984) have
16 shown comparable spirometric function response when exposed to low O3 dose under similar
17 conditions. Among adults, however, it has been repeatedly demonstrated that older individuals
18 respond to O3 inhalation with less intense lung function changes than younger adults. Thus,
19 children, adolescents, and young adults appear to be about equally responsive to O3, but more
20 responsive than middle-aged and older adults when exposed to a comparable dose of O3 (U.S.
21 Environmental Protection Agency, 1996).
22 Gong et al. (1997a) studied ten healthy men (60 to 69 years old) and nine COPD patients
23 (59 to 71 years old) from the Los Angeles area who were exposed to 0.24 ppm O3 while
24 intermittently exercising every 15 min at a light load (-20 L/min) for 4 h. 2Healthy subjects
25 showed a small but significant O3-induced FEVj decrement of 3.3% (p = 0.03 [not reported in
26 paper] paired-t on O3 versus FA pre-post FEVj). Small but statistically nonsignificant changes
27 were also observed for respiratory symptoms, airway resistance and arterial O2 saturation. In the
28 COPD patients, there was an 8% FEVj decrement due to O3 exposure which was not
2Personal communication from authors, correction to Table 2 in Gong et al. (1997a), the %FEV{ change at
the end of the ozone exposure for subject ID 2195 should read 4.9 and not the published value of -4.3, the mean and
standard deviation reported in the table are correct.
January 2005 AX6-42 DRAFT-DO NOT QUOTE OR CITE
-------
Table AX6-5. Age Differences in Pulmonary Function Responses to Ozonea
to
o
o
X
H
6
o
o
H
O
o
HH
H
W
Ozone
Concentration1"
ppm
0.40
0.42
0.0
0.40
0.0
0.24
0.0
0.12
0.18
0.24
0.30
0.40
0.0
0.12
0.18
0.24
0.30
0.40
ug/m3
784
823
0
784
0
470
0
235
353
471
589
784
0
235
353
471
589
784
Exposure Duration
and Activity
2 h IE (15' ex/1 5' rest)
VE - 33-45 L/min
(47 subjects only)
1. 5 h IE (20' ex/10' rest)
VE * 33-45 L/min
(All subjects)
2 h, IE (15' ex/1 5' rest)
VE- 18 L/min/m2 BSA
2 exposures: 25% of subj .
exposed to air-air,
75% exposed to O3-O3
4 h, IE (15' ex/ 15' rest)
VE =20 L/min
2 h rest or IE
(4x15 min
at VE = 25 or 35
L/min/m2 BSA)
2.33h IE
(4x15 min
atVE = 25
L/min/m2 BSA)
Exposure
Conditions
-22 °C
40% RH
treadmill
21 °C
40% RH
treadmill
24°C
40% RH
22 °C
40% RH
22 °C
40% RH
Number and
Gender of
Subjects
146 M
94 F
28 M
34 F
10 M
9M
485 WM
(each subject
exposed at one
activity level
to one O3
concentration)
371 (WM,
BM, WF, BF;
-25% per
group) each
subject
exposed to one
03
concentration
Subject
Characteristics
Healthy NS
18 to 60 years old
Healthy NS
18 to 57 years old
Healthy NS
18 to 59 years old
Healthy NS
60 to 69 years old
COPD
59 to 71 years old
Healthy NS
18 to 36 years old
mean age 24 years
Healthy NS
18 to 35 years old
mean age 24 years
Observed Effect(s)
Young individuals of both gender (< 35 years)
significantly more responsive than older
subjects. Strong responses are less common
over the age of 35 years, especially in women.
The variability of an individual's responsiveness
to repeated exposures to O3 decreases with age.
Significant decrements in spirometric lung
function in all groups. Young males and females
(< 35 years) were significantly more responsive
than older individuals (> 35 years).
Healthy: small, 3.3%, decline in FEVj (p=0.03 [not
reported in paper], paired-t on O3 versus FA pre-post
FEVj). COPD: 8% decline in FEVj (p=ns, O3 versus
FA). Adjusted for exercise, ozone effects did not differ
significantly between COPD patients and healthy
subjects.
Statistical analysis of 8 experimental chamber
studies conducted between 1980 and 1993 by the
U.S. EPA in Chapel Hill, NC. O3-induced
decrement in FEV[ predicted to decrease with
age. FEV; response of a 30 year old predicted to
be 50% the response of a 20 year old. Also see
Table 6-1
Statistical analysis of experimental data
collected between 1983 and 1990 by the U.S.
EPA in Chapel Hill, NC. Cyinduced decrement
in FEV; predicted to decrease with age. FEV;
response of a 30 year old predicted to be 65%
the response of a 20 year old. No effect of
menstrual cycle phase on FEV; response.
Inconsistent effect of social economic status on
FEV[ response.
Reference
Hazucha et al.
(2003)
Passannante
etal. (1998)
Gong et al.
(1997a)
McDonnell
etal. (1997)
Seal et al.
(1996)
-------
3
to
o
o
X
rv
j\
O
>
H
6
o
2?
O
H
O
O
m
rn
O
Ozone
Concentration1"
ppm
0.18
0.24
0.30
0.40
0.45
0.45
0.45
0.12
0.20
0.30
0.113C +
other
ambient
pollutants
ug/m3
353
470
588
784
882
882
882
235
392
588
221
xauii; r^^w-«j ^'
Exposure Duration
and Activity
2.33h
IE
VE = 20 L/min/m2 BSA
Ih, CE
VE * 26 L/min
2h,IE
VE * 26 L/min
2 h, IE (20' ex/20' rest)
Male: VE = 28.5L/min
Female: VE = 26.1 L/min
2 h, IE (20' ex/20' rest)
VE - 26 L/min
1 h IE (mouthpiece)
VE = 4 to 5 x resting
1 h (mouthpiece)
50' rest/10' ex for first 7
males, 20' rest/10' ex for
remaining subjects
Male: VE - 29 L/min
Female: VE - 23 L/min
1 h CE (bicycle)
VE - 22 L/min
k^uui uy* -f»gc i^inci (;iiv^9
Number and
Exposure Gender of
Conditions Subjects
NA 48 WF, 55 BF
-23 °C 7M
58% RH 5 F
cycle/treadmill
23 °C 10 M,
46% RH 6 F
cycle/treadmill
-24 °C 8 M
63% RH
cycle
8F
22 °C 5 M, 7 F
75% RH
treadmill
-22 °C 9M, 10 F
> 75% RH
treadmill
32.7 °C 33 M, 33 F
-43% RH
cycle
111 i uiiiiirii
-------
Table AX6-5 (cont'd). Age Differences in Pulmonary Function Responses to Ozone"
to
o
o
Ozone
Concentration1"
ppm
0.45
0.12
0.18
exposure Liuranon
u,g/m3 and Activity
882 2 h, IE (20' ex/20' rest)
VE « 26 L/min
235 40 min (mouthpiece)
IE, 10 min exercise at
VE = 32.6 L/min
353 40 min (mouthpiece)
IE, 10 min exercise at
VE = 41.3L/min
Number and
Exposure Gender of
Conditions Subjects
-23 °C 8 M,
53% RH 8 F
cycle
NA 3 M, 7 F
treadmill
4M, 6F
Subject
Characteristics Observed Effect(s)
Healthy NS, Mean decrement in FEV; = 5 .6 ± 1 3%; range of
51 to 76 years old decrements = 0 to 12%.
Healthy NS, No significant change in FEV^ increased RT
14 to 19 years old with exposure to 0.18 ppm O3. Some subjects
responded to 5 to 10 mg/mL methacholine after
0.18-ppm O3 exposure, whereas none responded
to 25 mg/mL methacholine at baseline
bronchochallenge.
Reference
Drechsler-
Parks et al.
(1987a,b)
Koenig et al.
(1987)
X
"See Appendix A for abbreviations and acronyms.
bListed from lowest to highest O3 concentration.
GOzone concentration is the mean of a range of ambient concentrations.
H
6
o
o
H
O
o
HH
H
W
-------
1 significantly different from the response in the healthy subjects. The authors have concluded
2 that typical ambient concentrations of O3 are unlikely to induce "a clinically significant acute
3 lung dysfunction" in exposed older men. However, they also acknowledged that the "worst
4 case" scenario of O3 exposure used in their study causes acute spirometric responses.
5 Although Gong et al. (1997a) and others (see Table 6-5) have examined responses to O3
6 exposure in subjects of various ages, the exposure conditions differ between most studies
7 so that age effects remain uncertain. Three recent studies, which analyzed large data sets
8 (> 240 subjects) of similarly exposed subjects, show clearly discernable changes in FEVj
9 responses to O3 as a function of age.
10 Seal et al. (1996) analyzed O3-induced spirometric responses in 371 young nonsmokers
11 (18 to 35 years of age). The subject population was approximately 25% white males, 25% white
12 females, 25% black males, and 25% black females. Each subject was exposed once to 0.0, 0.12,
13 0.18, 0.24, 0.30, or 0.40 ppm ozone for 2.3 h during IE at a VE of 25 L/min/m2 BSA. A logistic
14 function was used to model and test the significance of age, socioeconomic status (SES), and
15 menstrual cycle phase as predictors of FEVj response to O3 exposure. Menstrual cycle phase
16 was not a significant. SES was inconsistent with the greatest response observed in the medium
17 SES and the lowest response in high SES. FEVj responses decreased with subject age. On
18 average, regardless of the O3 concentration, the response of 25, 30, and 35 year old individuals
19 are predicted to be 83, 65, and 48% (respectively) of the response in 20 year olds. For example,
20 in 20 year old exposed to 0.12 ppm ozone (2.3 h IE, VE = 25 L/min/m2 BSA) a 5.4% decrement
21 in FEVj is predicted, whereas, a similarly exposed 35 yr old is only predicted to have a 2.6%
22 decrement. The Seal et al. (1996) model is limited to predicting FEVj responses immediately
23 postexposure in individuals exposed for 2.3 h during IE at a VE of 25 L/min/m2 BSA.
24 McDonnell et al. (1997) examined FEVj responses in 485 healthy white males (18 to
25 36 years of age) exposed once for 2 h to an O3 concentration of 0.0, 0.12, 0.18, 0.24, 0.30, or
26 0.40 ppm at rest or one of two levels of IE (VE of 25 and 35 L/min/m2 BSA). FEVj was
27 measured preexposure, after 1 h of exposure, and immediately postexposure. Decrements in
28 FEVj were modeled by sigmoid-shaped curve as a function of subject age, O3 concentration, VE,
29 and duration of exposure. Regardless of the O3 concentration or duration of exposure, the
30 average responses of 25, 30, and 35 year old individuals are predicted to be 69, 48, and 33%
January 2005 AX6-46 DRAFT-DO NOT QUOTE OR CITE
-------
1 (respectively) of the response in 20 year olds. The McDonnell et al. (1997) model is best suited
2 to predicting FEVj responses in while males exposed to O3 for 2 h or less under IE conditions.
3 Hazucha et al. (2003) analyzed the distribution of O3 responsiveness in subjects (146 M,
4 94 F) between 18 and 60 years of age. Subjects were exposed to 0.42 ppm O3 for 1.5 h with IE
5 at VE = 20 L/min/m2 BSA. Figure AX6-7 illustrates FEVj responses to O3 exposure as a
6 function of subject age. Consistent with the discussion in Section 6.4, a large degree of
7 intersubject variability is evident in Figure AX6-7. Across all ages, 18% of subjects were weak
8 responders (<5% FEVj decrement), 39% were moderate responders, and 43% were strong
9 responders (> 15% FEVj decrement). Younger subjects (<35 years of age) were predominately
10 strong responders, whereas, older subjects (>35 years of age) were mainly weak responders.
11 In males, the FEVj responses of 25, 35, and 50 year olds are predicted to be 94, 83, and 50%
12 (respectively) of the average response in 20 year olds. In females, the FEVj responses of 25, 35,
13 and 50 year olds are predicted to be 82, 46, and 18% (respectively) of the average response in
14 20 year olds. The Hazucha et al. (1996) model is limited to predicting FEVj responses
15 immediately postexposure in individuals exposed to 0.42 ppm O3 for 1.5 h during IE at a VE of
16 20 L/min/m2 BSA.
110-
100-
-5T 9°H
-------
1 The pathophysiologic mechanisms behind the pronounced age-dependent, gender-
2 differential rate of loss of O3 responsiveness are unclear. Passannante et al. (1998) have
3 previously demonstrated that O3-induced spirometric decrements (FEVj) in healthy young and
4 middle-aged adults are principally neural in origin, involving opioid-modulated sensory
5 bronchial C-fibers. (The methodological details of this study are presented in Section AX6.2.3 of
6 this chapter} The peripheral afferents are most likely the primary site of action, which would be
7 compatible with a reflex action as well as a cortical mechanism. The pattern of progressive
8 decline, as well as the subsequent rate of recovery of spirometric lung function, suggest
9 involvement of both direct and indirect (possibly by PGE2a) stimulation and/or sensitization of
10 vagal sensory fibers. (For details, see Section AX6.2.3.1 of this chapter }
11 The additional pulmonary function data published since the release of last O3 criteria
12 document (U.S. Environmental Protection Agency, 1996) and reviewed in this section reinforce
13 the conclusions reached in that document. Children and adolescents are not more responsive to
14 O3 than young adults when exposed under controlled laboratory conditions. However, they are
15 more responsive than middle-aged and older individuals. Young individuals between the age of
16 18 and 25 years appear to be the most sensitive to O3. With progressing age, the sensitivity to O3
17 declines and at an older age (> 60 yrs) appears to be minimal except for some very responsive
18 individuals. Endpoints other than FEVj may show a different age-related pattern of
19 responsiveness.
20
21 AX6.5.2 Gender and Hormonal Influences
22 The few late 1970 and early 1980 studies specifically designed to determine symptomatic
23 and lung function responses of females to O3 were inconsistent. Some studies have concluded
24 that females might be more sensitive to O3 than males, while others found no gender differences
25 (U.S. Environmental Protection Agency, 1996). During the subsequent decade, seven studies
26 designed to systematically explore gender-based differences in lung function following O3
27 exposure were completed (Table AX6-6). Protocols included mouthpiece and chamber
28 exposures, young and old individuals, normalization of ventilation to BSA or FVC, continuous
29 and intermittent exercise, control for menstrual cycle phase, and the use of equivalent effective
30 dose of O3 during exposures. These studies have generally reported no statistically significant
31 differences in pulmonary function between males and females (Adams et al., 1987; Drechsler-
January 2005 AX6-48 DRAFT-DO NOT QUOTE OR CITE
-------
to
o
o
Table AX6-6. Gender and Hormonal Differences in Pulmonary Function Responses to Ozone"
Ozone
Concentration11
ppm ug/m3 and Activity Conditions'
0.0 490 1 h CE NA
0.25 VE = 30 L/min Face mask
exposure
Number and Gender Subject
of Subjects Characteristics
32 M, 28 F Healthy NS
22.6 ±0.6 years old
Observed Effect(s)
Mean O3-induced FEVj decrements of 15.9%
in males and 9.4% in females (gender
differences not significant). FEV; decrements
ranged from -4 to 56%; decrements >15% in
20 subjects and >40% in 4 subjects. Uptake
of O3 greater in males than females, but
uptake not correlated with spirometric
responses.
Reference
Ultman et al.
(2004)
X
VO
H
6
o
o
H
O
o
HH
H
W
0.40
0.42
0.0
0.35
0.0
0.4
0
0.12
0.24
0.30
0.40
0.0
0.35
784
823
0
686
0
784
0
235
470
588
784
0
686
2 h, IE (15' ex/1 5' rest)
VE = 33-45 L/min
1. 5 h IE (20' ex/10' rest)
VE = 33-45 L/min
1.25h, IE
(30' ex/1 5' rest/30' ex)
VE = 40 L/min
2 h, IE (15' ex/1 5' rest)
VE» 18L/min/m2BSA
2 exposures: 25% of
subj. exposed to air-air,
75% exposed to O3-O3
2.33 h IE (15' ex/15' rest)
VE = 20 L/min/m2 BSA
one exposure per subject
2.15 h, IE
(30' ex/30' rest)
22 °C 146 M
40% RH 94 F
treadmill
22 °C 19 F
40% RH
treadmill
21 °C 28 M
40% RH 34 F
treadmill
22 °C 48 WF, 55 BF
40% RH
treadmill
19-24 °C 12 M
48-55% RH 12 F
treadmill
Healthy NS,
1 8 to 60 years old
O3 responders
22.1 ±2. 7 years old
Healthy NS, 20-59 years old
Healthy NS,
18 to 35 years old
Healthy NS,
5 F follicular and 7 luteal
phase exposure,
regular menstrual cycles, 18
to 35 years old
No significant gender differences in FEVj
among young (< 35 years) and older
individuals. Strong responses are less
common over the age of 35 years, especially
in women.
FVC and FIVC changes about the same,
-13%, FEVj -20%. Increased airway
responsiveness to methacholine. Persistence
of small effects on both inspired and expired
spirometry past 18 h. Chemoreceptors not
activated but ventilatory drive was
accelerated.
Significant decrements in spirometric lung
function. No significant differences in FEV;
between young females and males and older
females and males either in responders or
non-responders subgroups.
Significant menstrual cycle phase * race
interaction for FEVj. No significant
menstrual cycle phase effect when blacks and
whites were analyzed separately.
No significant menstrual phase effects for
SRaw or cough score.
Changes in FVC, FEV,, FEF25.75, Vmaj[50%, and
Vmax25% were similar during both the follicular
and luteal phases. No significant difference
between males and females.
Hazucha
et al. (2003)
Folinsbee
and Hazucha
(2000)
Passannante
etal. (1998)
Seal et al.
(1996)
Weinmann
etal. (1995)
-------
to
o
o
Table AX6-6 (cont'd). Gender and Hormonal Differences in Pulmonary Function Responses to Ozone"
Ozone
Concentration11
ppm ug/m3
0.3 588
and Activity
1 hCE
VE = 50 L/min
Exposure Number and Gender Subject
Conditions' of Subjects Characteristics
NA 9 F Healthy NS, regular
menstrual cycles,
20 to 34 years old
Observed Effect(s)
FEVj decreased 13.1% during the mid-luteal
phase and 18.1% during the follicular phase.
Decrement in FEF25.75 was significantly larger
Reference
Fox et al.
(1993)
during the follicular phase than the mid-luteal
phase. Changes in FVC were similar in
both phases.
X
ON
(^
O
H
6
o
o
H
O
o
0.12
0.18
0.24
0.30
0.40
0.0
0.45
0.3
0
235
353
470
588
784
0 0
0.18 353
0.30 588
0
882
0 0
0.20 392
0.30 588
588
2.33 h (15'ex/15'rest)
VE = 25 L/min/m2 BSA
(one exposure/subject)
1 h (mouthpiece), CE
VE = 47 L/min
exposures > 4 days apart
2 h, IE (20' ex/20' rest)
VE = 28.5 L/min for M
VE = 26.1 L/min for F
repeated O3 exposures
22 °C
40% RH
treadmill
21 to 25 °C
45 to 60% RH
cycle
23.1 °C
46.1%RH
cycle/treadmill
30 to 33 F and 30 to Healthy NS, 18 to 35 years Decrements in FEV1; increases in SRaw and Seal et al.
1 h (mouthpiece) IE -22 °C
(50'rest/10'ex first 7 M) > 75% RH
(20' rest/10' ex all others) treadmill
33 M in each
concentration group;
total of
372 individuals
participated
14 F
14 F
10M
6F
9M, 10 F
1 h (mouthpiece), CE
21 to 25 °C
45 to 60% RH
cycle
20 M
20 F
old, blacks and whites
FVC = 5.11 ± 0.53 L,
NS, 20 to 24 years old
FVC = 3.74 ± 0.30 L,
NS, 19 to 23 years old
Healthy NS,
60 to 89 years old
Healthy NS,
64 to 71 years old
Healthy NS,
55 to 74 years old
NS, 18 to 30 years old
NS, 19 to 25 years old
cough, correlated with O3 concentration. (1993)
There were no significant differences between
the responses of males and females.
Small lung group, FVC = 3.74 ± 0.30 L. Messineo
Large lung group, FVC = 5.11 ± 0.53 L. and Adams
Significant concentration-response effect on (1990)
FVC and FEVjj lung size had no effect on
percentage decrements in FVC or FEVj.
Mean decrement in FEVj = 5.7%. Bedi et al.
Decrements in FVC and FEV; were the only (1989)
pulmonary functions significantly altered by
O3 exposure. No significant differences
between responses of men and women.
No change in any spirometic measure for Reisenauer
either group. Females had 13% increase in RT et al. (1988)
after 0.30-ppm exposure. Gender differences
not evaluated.
Significant decrements in FVC, FEV1; and
FEF25.75 following O3 exposure.
No significant differences between men and
women for spirometry or SRaw.
Adams et al.
(1987)
o
HH
H
W
-------
Table AX6-6 (cont'd). Gender and Hormonal Differences in Pulmonary Function Responses to Ozone"
c
!-J
-<
to
o
o
Ozone
Concentration11
ppm ug/m3 and Activity
0.0 0 2 h, IE (20' rest/ 20' ex)
0.45 882 VE = 27.9 L/min for M
VE»25.4L/minforF
repeated O3 exposures
0.48 941 2h, IE
VE = 25 L/min
Exposure
Conditions'
24 °C
58% RH
cycle
21 °C
(WBGT)
cycle
Number and Gender Subject
of Subjects Characteristics
8M Healthy NS,
5 1 to 69 years old
8F Healthy NS,
56 to 76 years old
10 F Healthy NS,
19 to 36 years old
Observed Effect(s)
Range of responses in FEVj:
0 to - 12% (mean = -5.6%).
No significant difference in responses of men
and women.
Tendency for women to have greater effects.
Mean decrement in FEV; = 22.4%.
Significant decrements in all spirometric
measurements. Results not significantly
different from a similar study on males
(Drechsler-Parks et al., 1984).
Reference
Drechsler-
Parks et al.
(1987a,b)
Horvath
etal. (1986)
X
H
6
o
o
H
O
" See Appendix A for abbreviations and acronyms.
b Listed from lowest to highest O3 concentration.
'WBGT = 0.7 Twetbulb + 0.3 T.
o
HH
H
W
-------
1 Parks, et al., 1987a; Messineo and Adams, 1990; Seal et al., 1993; Weinmann et al., 1995)
2 although in some studies females appeared to experience a slightly greater decline then males
3 (Drechsler-Parks et al., 1987a; Messineo and Adams, 1990). The comparative evaluations were
4 based on responses that included spirometry, airway resistance, non-specific bronchial
5 responsiveness (NSBR) determinations, and changes in frequency and severity of respiratory
6 symptoms. However, depending on how the O3 dose was calculated and normalized, the
7 findings of at least three studies may be interpreted as showing that females are more sensitive to
8 O3 than males. The findings of the seven studies are presented in detail in Section 7.2.1.3 of the
9 previous O3 criteria document (U.S. Environmental Protection Agency, 1996).
10 Some support for a possible increased sensitivity of females to O3 comes from a study of
11 uric acid concentration in nasal lavage fluid (NLF). Housley et al. (1996) found that the NLF of
12 females contains smaller amounts of uric acid than the NLF of males. The primary source of
13 uric acid is plasma; therefore, lower nasal concentrations would reflect lower plasma
14 concentrations of this antioxidant. The authors have speculated that in females, both lower
15 plasma and NLF levels (of uric acid) can plausibly make them more susceptible to oxidant
16 injury, since local antioxidant protection may not be as effective as with higher levels of uric
17 acid, and consequently more free O3 can penetrate deeper into the lung.
18 Several studies also have suggested that anatomical differences in the lung size and the
19 airways between males and females, and subsequent differences in O3 distribution and
20 absorption, may influence O3 sensitivity and potentially differential O3 response. The study of
21 Messinio and Adams (1990) have, however, convincingly demonstrated that the effective dose to
22 the lung, and not the lung size, determines the magnitude of (FEVj) response. Furthermore, the
23 O3 dosimetry experiments of Bush et al. (1996) have shown that despite gender differences in
24 longitudinal distribution of O3, the absorption distribution in conducting airways was the same
25 for both sexes when expressed as a ratio of penetration to anatomic dead space volume. This
26 implies that gender differences, if any, are not due to differences in (normal) lung anatomy.
27 The data also have shown that routine adjustment of O3 dose for body size and gender
28 differences would be more important if normalized to anatomic dead space rather than the usual
29 FVCorBSA.
30 One of the secondary objectives of a study designed to examine the role of neural
31 mechanisms involved in limiting maximal inspiration following O3 exposure has been to
January 2005 AX6-52 DRAFT-DO NOT QUOTE OR CITE
-------
1 determine if gender differences occur. A group of healthy males (n = 28) and females (n = 34)
2 were exposed to 0.42 ppm O3 for 2 h with IE. The methodological details of the study are
3 presented in Section AX6.2.5.1 of this document. As Figure AX6-4 shows, the differences
4 between males and females were, at any condition, measurement point, and O3 sensitivity status
5 only minimal and not significant (Passannante et al., 1998).
6 In another investigation, Folinsbee and Hazucha (2000) exposed a group of
7 19 O3-responsive young females (average age of 22 years, prescreened for O3 responsiveness by
8 earlier exposure) to air and 0.35 ppm O3. The randomized 75-min exposures included two
9 30-min exercise periods at a VE of 40 L/min. In addition to standard pulmonary function tests,
10 they employed several techniques used for the first time in human air pollution studies
11 assessment of O3 effects. The average lung function decline from a pre-exposure value was 13%
12 for FVC, 19.9 % for FEVj, and 30% for FEF25.75. The infrequently measured forced inspiratory
13 vital capacity (FIVC) was the same as FVC suggesting that the lung volume limiting
14 mechanisms are the same. The reduction in peak inspiratory flow (PIF) most likely reflects an
15 overall reduction in inspiratory effort associated with neurally mediated inhibition of inspiration.
16 Persistence of small inspiratory and expiratory spirometric effects, airway resistance, and airway
17 responsiveness to methacholine for up to 18 h postexposure suggests that recovery of pulmonary
18 function after O3 exposure involves more than the simple removal of an irritant. Incomplete
19 repair of damaged epithelium and still unresolved airway inflammation are the likely causes of
20 the residual effects that in some individuals persisted beyond 24 h postexposure. However, by
21 42 hours no residual effects were detected. No significant changes were found in ventilatory
22 response to CO2 between air and O3 exposures, suggesting that chemoreceptors were not affected
23 by O3. However, O3 inhalation did result in accelerated timing of breathing and a modest
24 increase in inspiratory drive. These observations are consistent with, and further supportive of,
25 the primary mechanisms of O3-induced reduction in inspiratory lung function, namely an
26 inhibition of inspiration elicited by stimulation of the C-fibers and other pulmonary receptors.
27 Because the measures of inspiratory and chemical drive to assess O3 effects were not reported in
28 any previous human study, no comparisons are possible. Because no male subjects were
29 recruited for the study, it is not possible to compare gender effects. Despite being O3-responsive,
30 however, the average post-O3 decline in expiratory lung function from preexposure (13% for
31 FVC; 19.9% for FEV^ 30% for FEF25.75) was similar to that seen in female cohorts studied by
January 2005 AX6-53 DRAFT-DO NOT QUOTE OR CITE
-------
1 other investigators under similar conditions of exposure. These were the same studies that found
2 no gender differences in O3 sensitivity (Adams et al., 1987; Messineo and Adams, 1990).
3 The study by Hazucha et al. (2003), discussed in the previous section, has in addition to
4 aging also examined gender differences in O3 responsiveness. The male (n = 146) and female
5 (n = 94) cohorts were classified into young (19 to 35 year-old) and middle-aged (35 to 60 year-
6 old) groups. This classification was selected in order to facilitate comparison with data reported
7 previously by other laboratories. Using a linear regression spline model (with a break point at
8 35 years), the authors reported that the rate of loss of sensitivity is about three times as high in
9 young females as in young males (p< 0.003). In young females, the average estimated decline in
10 FEVj response is 0.71% per year, while in young males it is 0.19% per year. Middle-aged
11 groups of both genders show about the same rate of decline (0.36 to 0.39%, respectively).
12 At 60 years of age, the model estimates about a 5% post-O3 exposure decline in FEVj for males,
13 but only a 1.3% decline for females. These observations suggest that young females lose O3
14 sensitivity faster than young males, but by middle age, the rate is about the same for both
15 genders. Descriptive statistics show that there were practically no differences in the mean value,
16 standard error of the mean, and coefficient of variation for % FEVj decrement between the group
17 of young males (n = 125; 83.7 ± 1.1%; CV = 13.5%) and young females (n = 73; 83.4 ± 1.25%;
18 CV = 12.8%). A straight linear regression model of these data was illustrated in Figure AX6-7.
19 The slopes, significant in both males (r = 0.242; p = 0.003) and females (r = 0.488; p = 0.001),
20 represent the decline in responsiveness of 0.29% and 0.55% per year respectively, as assessed by
21 FEVi.
22 Two earlier studies of the effects of the menstrual cycle phase on O3 responsiveness have
23 reported conflicting results (U.S. Environmental Protection Agency, 1996). Weinmann et al.
24 (1995) found no significant lung function effects related to menstrual cycle, although during the
25 luteal phase the effects were slightly more pronounced than during the follicular phase; while
26 Fox et al., (1993) reported that follicular phase enhanced O3 responsiveness. In a more recent
27 investigation of possible modulatory effects of hormonal changes during menstrual cycle on O3
28 response, young women (n = 150) 18 to 35 years old were exposed once to one of multiple O3
29 concentrations (0.0, 0.12, 0.18, 0.24, 0.30, 0.40 ppm) for 140 min with IE at 35 L/min/m2 BSA.
30 The women's menstrual cycle phase was determined immediately prior to O3 exposure. Post-O3,
31 no significant differences in % predicted FEVj changes that could be related to the menstrual
January 2005 AX6-54 DRAFT-DO NOT QUOTE OR CITE
-------
1 cycle phase were found. Admittedly, a less precise method of determining menstrual cycle
2 phase used in this study could have weakened the statistical power. Unfortunately, the direction
3 and magnitude of O3 response as related to the menstrual cycle phases were not reported (Seal
4 et al., 1996). Considering the inconclusiveness of findings of this study and the inconsistency of
5 results between the two earlier studies, it is not possible to make any firm conclusions about the
6 influence of the menstrual cycle on responses to O3 exposure.
7 Additional studies presented in this section clarify an open-ended conclusion reached in the
8 previous O3 criteria document (U.S Environmental Protection Agency, 1996) regarding the
9 influence of age on O3 responsiveness. Healthy young males and females are about equally
10 responsive to O3, although the rate of loss of sensitivity is higher in females than in males.
11 Middle-aged men and women are generally much less responsive to O3 than younger individuals.
12 Within this range, males appear to be slightly more responsive than females, but the rate of age-
13 related loss in FEVj is about the same. The O3 sensitivity may vary during the menstrual cycle;
14 however, this variability appears to be minimal.
15
16 AX6.5.3 Racial, Ethnic and Socioeconomic Status Factors
17 In the only laboratory study designed to compare spirometric responses of whites and
18 blacks exposed to a range of O3 concentrations (0 to 0.4 ppm), Seal et al. (1993) reported
19 inconsistent and statistically insignificant FEVj differences between white and black males and
20 females within various exposure levels. Perhaps, with larger cohorts the tendency for greater
21 responses of black than white males may become significant. Thus, based on this study it is still
22 unclear if race is a modifier of O3 sensitivity, although the findings of epidemiologic studies
23 reported in the previous criteria document "can be considered suggestive of an ethnic difference"
24 (U.S. Environmental Protection Agency, 1996). However, as Gwynn and Thurston (2001)
25 pointed out, it appears that it is more the socioeconomic status (SES) and overall quality of
26 healthcare that drives PM10- and O3-related hospital admissions than an innate or acquired
27 sensitivity to pollutants.
28 This assertion is somewhat supported by the study of Seal et al. (1996) who employed a
29 family history questionnaire to examine the influence of SES on the O3 responsiveness of
30 352 healthy, 18- to 35-year-old black and white subjects. Each subject was exposed once under
31 controlled laboratory conditions to either air or 0.12, 0.18, 0.24, 0.30, 0.40 ppm O3 for 140 min
January 2005 AX6-55 DRAFT-DO NOT QUOTE OR CITE
-------
1 with 15 min IE at 35 L/min/m2 BSA. An answer to the "Education of the father" question was
2 selected as a surrogate variable for SES status. No other qualifying indices of SES were used or
3 potential bial examined. Of the three SES categories, individuals in the middle SES category
4 showed greater concentration-dependent decline in % predicted FEVj (4-5% @ 0.4 ppm O3) than
5 low and high SES groups. The authors did not have an "immediately clear" explanation for this
6 finding. The SES to %predicted FEVj relationship by gender-race group was apparently
7 examined as well; however, these results were not presented. Perhaps a more comprehensive
8 and quantitative evaluation of SES status would have identified the key factors and clarified the
9 interpretation of these findings. With such a paucity of data it is not possible to discern the
10 influence of racial or other related factors on O3 sensitivity.
11
12 AX6.5.4 Influence of Physical Activity
13 Apart from the importance of increased minute ventilation on the inhaled dose of O3 during
14 increased physical activity, including work, recreational exercise, and more structured exercise
15 like sports, no systematic effort has been made to study other potential physical factors that may
16 modulate O3 response. The typical physiologic response of the body to exercise is to increase
17 both the rate and depth of breathing, as well as increase other responses such as heart rate, blood
18 pressure, oxygen uptake, and lung diffusion capacity.
19 Physical activity increases minute ventilation in proportion to work load. At rest, and
20 during light exercise, the dominant route of breathing is through the nose. The nose not only
21 humidifies air, among other physiologic functions, but also absorbs O3 thus decreasing the
22 overall dose. As the intensity of exercise increases, the minute ventilation increases and the
23 breathing switches from nasal to oronasal mode. There is considerable individual variation in
24 the onset of oronasal breathing, which ranges from 24 to 46 L/min (Niinimaa et al., 1980).
25 During heavy exercise, ventilation is dominated by oral breathing. Consequently, the residence
26 time of inhaled air in the nose and the airways is shorter, reducing the uptake of O3 (Kabel et al.,
27 1994). Moreover, increasing inspiratory flow and tidal volume shifts the longitudinal
28 distribution of O3 to the peripheral airways, which are more sensitive to injury than the larger,
29 proximal airways. Ozone uptake studies of human lung showed that at simulated quiet
30 breathing, 50% of O3 was absorbed in the upper airways, 50% in the conducting airways, and
January 2005 AX6-56 DRAFT-DO NOT QUOTE OR CITE
-------
1 none reached the small airways (Hu et al. 1994). With ventilation simulating heavy exercise
2 (60 L/min), the respective O3 uptakes were 10% (upper airways), 65% (conducting airways), and
3 25% (small airways). These observations imply that equal O3 dose (C x T x VE) will have a
4 greater effect on pulmonary function and inflammatory responses when inhaled during heavy
5 physical activity than when inhaled during lighter activity. Although, Ultman et al. (2004)
6 recently reported that spirometric response are not correlated with O3 uptake. (See Chapter 4 of
1 this document for more information on the dosimetry ofO3)
8 Other physiologic factors activated in response to physical activity are unlikely to have as
9 much impact on O3 responsiveness as does minute ventilation; however, their potential influence
10 has not been investigated.
11
12 AX6.5.5 Environmental Factors
13 Since the 1996 O3 criteria document not a single human laboratory study has examined the
14 potential influence of environmental factors such as rural versus urban environment, passive
15 cigarette smoke exposure, and bioactive admixtures such as endotoxin on healthy individual's
16 pulmonary function changes due to O3 (U.S. Environmental Protection Agency, 1996).
17 Some of the unresolved issues, e.g., health effects of ETS and O3 interaction, which need to
18 be examined in human studies were explored very recently in laboratory animal studies
19 (presented in a greater detail in Chapter 5). In one study on mice, preexposure of animals to
20 sidestream cigarette smoke (ETS surrogate), which elicited no immediate effects, resulted in a
21 potentiation of subsequent O3-induced inflammatory response. This finding suggests that typical
22 adverse effects of ETS do not necessarily have to elicit an immediate response to ETS, but may
23 in fact potentiate the effects of a subsequent exposure to another pollutant like O3 (Yu et al.,
24 2002). The key mechanism by which smoke inhalation may potentiate subsequent oxidant injury
25 appears to be damage to cell membranes and the resulting increase in epithelial permeability.
26 Disruption of this protective layer may facilitate as well as accelerate injury to subepithelial
27 structures when subsequently exposed to other pollutants (Bhalla, 2002). Although this may be a
28 plausible mechanism in nonsmokers and acute smokers exposed to ETS and other pollutants,
29 studies involving chronic smokers who most likely already have chronic airway inflammation do
30 not seems to show exaggerated response with exposure to O3.
31
January 2005 AX6-57 DRAFT-DO NOT QUOTE OR CITE
-------
1 More than 25 years ago, Hazucha et al. (1973) reported that the spirometric lung function
2 of smokers declined significantly less than that of nonsmokers when exposed to 0.37 ppm O3.
3 The findings of this study have been confirmed and expanded (Table AX6-7). Frampton et al.
4 (1997a) found that exposure of current smokers (n=34) and never smokers (n=56) to 0.22 ppm
5 O3 for 4 h with IE for 20 min of each 30 min period at 40 to 46 L/min, induced a substantially
6 smaller decline in FVC, FEVj and SGaw of smokers than never smokers. Smokers also
7 demonstrated a much narrower distribution of spirometric endpoints than never smokers.
8 Similarly, nonspecific airway responsiveness to methacholine was decreased in smokers.
9 However, both groups showed the consistency of response from exposure to exposure. It should
10 be noted, that despite seemingly lesser response, the smokers were more symptomatic post air
11 exposure than never smokers but the opposite was true for O3 exposure. This would suggest that
12 underlying chronic airway inflammation present in smokers has blunted stimulation of bronchial
13 C-fibers and other pulmonary receptors, the receptors substantially responsible for post O3 lung
14 function decrements. In addition to desensitization, the other "protective" mechanisms active in
15 smokers may be an increase in the mucus layer conferring not only a mechanical protection, but
16 also acting as an O3 scavenger. Another plausible explanation of a diminished responsiveness of
17 smokers may be related to elevated levels of reduced glutathione (GSH), tissue antioxidant,
18 found in epithelial lining fluid of chronic but not acute smokers (MacNee et al., 1996).
19 Despite some differences in a release of proinflammatory cytokines and subsequent
20 recruitment of inflammatory cells, both smokers and nonsmokers developed airway
21 inflammation following O3 exposure. This was demonstrated by the Torres et al. (1997) study
22 that involved exposures of about equal size cohorts of otherwise healthy young smokers,
23 nonsmoker O3 nonresponders (< 5% FEVj post O3 decrement) and nonsmoker O3 responders
24 (> 15% FEVj post O3 decrement) to air and two 0.22 ppm O3 atmospheres for 4 hours,
25 alternating 20 min of moderate exercise (25 L/min/m2 BSA) with 10 min of rest. Both O3
26 exposures were followed by nasal lavage (NL) and bronchoalveolar lavage (BAL) performed
27 immediately post one of exposures and 18 hr later following the other exposure. Neither O3
28 responsiveness nor smoking alters the magnitude or the time course of O3-induced airway
29 inflammation. The overall cell recovery was lower immediately post exposure but higher,
30 particularly in nonsmokers, 18 h post O3 exposure when compared to control (air) in all groups.
31 Recovery of lymphocytes, PMNs and AMs in both alveolar and bronchial lavage fluid showed
January 2005 AX6-58 DRAFT-DO NOT QUOTE OR CITE
-------
Table AX6-7. Influence of Ethnic, Environmental, and Other Factors
to
o
o
X
ON
(^
VO
H
6
o
o
H
O
o
HH
H
W
Ozone
Concentration
ppm Hg/m3
0.0 0
0.4 780
0.0 0
0.12 235
Exposure
Duration Exposure
and Activity Conditions
2 h 20°C
IE, 15' ex/15' rest 40% RH
VE = 20 L/min/m2
BSA
0.75 h 60% RH
IE, 15' ex/15' rest
VE = 40-46 L/min
Number
and Gender
of Subjects
15 M, 1 F
13 M, 2 F
5 M, 12 F
Subject
Characteristics
Placebo group:
healthy NS
avg. age 27 yrs.
Antiox. Suppl.
Gr: Healthy NS
avg. age 27 yrs.
Asthmatics
sensitive to SO2
19 to 38 yrs old
Observed Effect(s)
PF decrements in the supplementation
group were signif. smaller for FVC
(p < 0.046) and near significant for FEV{
(p < 0.055). The inflammatory response
(BAL) showed no significant differences
between the two groups either in the
recovery of cellular components or the
concentrations and types of inflammatory
cytokines.
No significant differences due to O3
between placebo and antioxidant
supplement cohort in either PF or
Reference
Samet et al.
(2001)
Trenga et al.
(2001)
bronchial hyperresponsiveness to 0.1 ppm
SO2. The overall results interpreted as a
demonstration of protective effect of
antioxidants from O3, particularly in
"severe" asthmatics.
0.0
0.22
0.22
0.0
0.12-
0.24a
0 4h 21°C 25(M/F)
43 1 IE, 20' ex/10' rest 37% RH
431 VE = 40-46 L/min
0 2.17 h 22°Cor30°C 5 M, 4F
235-470a IE, 10' ex/ 10' rest 45-55% RH
VE = 36-39 L/min
Healthy NS
O3 responders and
nonresponders
18 to 40 yrs old
Healthy NS
24 to 32 yrs old
Glutathione peroxidase (GPx) activity and
eGPx protein level were significantly (p =
0.0001) depleted in ELF for at least 18 h
post-exp. In BAL both endpoints were
elevated (ns). No association between
cell injury, PF, or GPx activity.
FEVj decreased signif. (p < 0.5) by -8%
at22°Cand~6.5%at30°C. 19 h post-
exp decline of 2.3% still signif.
(p<0.05). SGaw signif. (p<0.05)
declined at 30°C but not at 22°C. The
BHR assessed 19 h post-exp. as PC50
sGaw methacholine signif. (p < 0.05)
higher at both temperatures.
Avissar et al.
(2000)b
Foster et al.
(2000)
-------
Table AX6-7 (cont'd). Influence of Ethnic, Environmental, and Other Factors
to
o
o
X
Oi
H
6
o
o
H
O
Ozone
Concentration
ppm Hg/ni3
0.0 0
0.40 780
Exposure Number
Duration Exposure and Gender Subject
and Activity Conditions of Subjects Characteristics
2 h 6 M, 9 F Healthy NS
IE, 20' ex/ 10' rest avg. age 3 1 yrs.
Observed Effect(s)
Corticosteroid pretreatment had no effects
on post O3 decline in PF, PMN response,
Reference
Nightingale
et al. (2000)
VE = mild to mod.
and sputum cell count under both the
placebo and treatment conditions.
Methacholine PC20 FEV{ was equally
decreased in both cond. 4 h after
exposure. No changes in exhaled NO and
CO
0.0 0
0.22 431
0.0 0
0.22 431
0.22 431
4h 21°C
IE 37% RH
20' ex/10' rest
VE = 40-46 L/min
4h 21°C
IE, 20' ex/10' rest 37% RH
VE = 25 L/min/m2
BSA
90 M 56 never smokers
34 current
smokers
18 to 40 yrs. old
10 M, 2 F NS, O3 nonresp.,
avg. age 25 yrs.;
10 M, 3 F NS, O3 resp.,
avg. age 25 yrs;
1 1 M, 2 F smokers avg.
age 28 yrs
Smokers are less responsive to O3 as
assessed by spirometric and
plethysmographic variables. Neither age,
gender, nor methacholine responsiveness
were predictive of O3 response.
Neither O3 responsiveness nor smoking
has altered the magnitude and the time
course of O3-induced airway
inflammation. Inflammation involved all
types of cells accessible by B AL. The
recovery profile of these cells overtime
was very similar for all groups showing
highest values 18 hpostexposure.
Frampton et al.
(1997a,c)b
Torres et al.
(1997)b
Frampton et al.
(1997a,c)b
aRamp exposure from 0.12 ppm to 0.24 ppm and back to 0.12 ppm at the end of exposure.
bRelated studies, sharing of some subjects .
o
HH
H
W
-------
1 the largest increase in response to O3 in all groups, with nonsmokers showing greater relative
2 increases than smokers. Of the two cytokines, IL-6 and IL-8, IL-6 was substantially and
3 significantly (p < 0.0002) elevated immediately post exposure but returned back to control 18 h
4 later in all groups; but only nonsmokers' effects were significantly higher (p < 0.024). IL-8
5 showed a similar pattern of response but the increase in all groups, though still significant
6 (p < 0.0001), was not as high as for IL-6. Between group differences were not significant. This
7 inflammatory response involved all types of cells present in BAL fluid and the recovery profile
8 of these cells over time was very similar for all groups. In contrast to BAL, NL did not prove to
9 be a reliable marker of airway inflammation. The lack of association between lung function
10 changes (spirometry) and airway inflammation for all three groups confirms similar observations
11 reported from other laboratories. This divergence of mechanisms is further enhanced by an
12 observation that a substantially different spirometric response between O3 responders and
13 nonresponders, the airway inflammatory response of the two groups was very similar, both in
14 terms of magnitude and pattern (Torres et al., 1997).
15 The influence of ambient temperature on pulmonary effects induced by O3 exposure in
16 humans has been studied infrequently under controlled laboratory conditions. Several
17 experimental human studies published more than 20 years ago reported additive effects of heat
18 and O3 exposure (see U.S. Environmental Protection Agency, 1986, 1996). In the study of
19 Foster et al. (2000) 9 young (mean age 27 years) healthy subjects (4F/5M) were exposed for
20 130 min (IE 10 min @ 36 to 39 1/min) to filtered air and to ramp profile O3 at 22° and 30 °C,
21 45-55% RH. The order of exposures was randomized. The O3 exposure started at 0.12 ppm,
22 reached the peak of 0.24 ppm mid-way through and subsequently declined to 0.12 ppm at the
23 end of exposure. Ozone inhalation decreased VT and increased fB as compared to baseline at
24 both temperatures. At the end of exposure FEVj decreased significantly (p < 0.5) by -8% at
25 22°C and -6.5% at 30 °C. One day (19 h) later, the decline of 2.3% from baseline was still
26 significant (p < 0.05). FVC decrements were smaller and significant only at 22 °C immediately
27 postexposure. SGaw significantly (p < 0.05) declined at 30°C but not at 22 °C. A day later,
28 sGaw was elevated above the baseline for all conditions. The nonspecific bronchial
29 responsiveness (NSBR) to methacloline assessed as PC50 sGaw was significantly (p < 0.05)
30 higher one day following O3 exposure at both temperatures but more so at 30 °C. Thus, these
31 findings indicate that elevated temperature has partially attenuated spirometric response but
January 2005 AX6-61 DRAFT-DO NOT QUOTE OR CITE
-------
1 enhanced airway reactivity. Numerous studies have reported an increase in NSBR immediately
2 after exposure to O3. Whether the late NSBR reported in this study is a persistent residual effect
3 of an earlier increase in airway responsiveness, or is a true one day lag effect cannot be
4 determined from this study. Whatever the origin, however, a delayed increase in airway
5 responsiveness raises a question of potentially increased susceptibility of an individual to
6 respiratory impairment, particularly if the suggested mechanism of disrupted epithelial
7 membrane holds true.
8
9 AX6.5.6 Oxidant-Antioxidant Balance
10 Oxidant-antioxidant balance has been considered as one of the determinants of O3
11 responsiveness. Amateur cyclists who took antioxidant supplements (vitamins C, E,and
12 p-carotene) for three months showed no decrements in spirometric lung function when cycling
13 on days with high O3 levels. In contrast, matched control group of cyclists not pretreated with
14 vitamin supplements experienced an almost 2% decline in FVC and FEVj and > 5% reduction in
15 PEF during the same activity period. Adjustment of data for confounders such as PM10 and NO2
16 did not change the findings. Apparently, substantially elevated levels of plasma antioxidants
17 may afford some protection against lung function impairment (Grievink et al.,1998, 1999).
18 Both laboratory animal and human studies have repeatedly demonstrated that antioxidant
19 compounds present the first line of defense against the oxidative stress. Thus, upregulation of
20 both enzymatic and nonenzymatic antioxidant systems is critical to airway epithelial protection
21 from exposure to oxidants such as O3 and NO2 (see Table AX6-7). As an extension of an earlier
22 study focused on pulmonary function changes (Frampton et al., 1997a), Avissar et al. (2000)
23 hypothesized that concentration of glutathione peroxidase (GPx), one of the antioxidants in
24 epithelial lining fluid (ELF), is related to O3 and NO2 responsiveness. They exposed healthy
25 young nonsmokers (n=25), O3-responders, and non-responders to filtered air and twice to
26 0.22 ppm O3 for 4 h (IE, 20' ex 710' rest, @ VE 40 to 46 L/min). In the NO2 part of the study,
27 subjects were exposed to air and twice to NO2 (0.6 and 1.5 ppm) for 3 h, with IE of 10 min of
28 each 30 min @ VE of 40 L/min. Ozone exposure elicited a typical pulmonary function response
29 with neutrophilic airway inflammation in both responders and non-responders. The GPx activity
30 was significantly reduced (p = 0.0001) and eGPx protein significantly depleted (p = 0.0001) in
31 epithelial lining fluid (ELF) for at least 18 h postexposure. In contrast, both GPx and eGPx were
January 2005 AX6-62 DRAFT-DO NOT QUOTE OR CITE
-------
1 slightly elevated in bronchoalveolar lavage fluid (BALF). However, neither of the two NO2
2 exposures had a significant effect on pulmonary function, airway neutrophilia, epithelial
3 permeability, GPx activity, or eGPx protein level in either ELF or BALF. The lack of a
4 significant response to NO2 has been attributed to the weak oxidative properties of this gas.
5 No association has been observed between cell injury, assessed by ELF albumin, or pulmonary
6 function and GPx activity for O3 exposure. Thus, it is unclear what role antioxidants may have
7 in modulation of O3-induced lung function and inflammatory responses. The authors found a
8 negative association between lower baseline eGPx protein concentration in ELF and post-O3
9 neutrophilia to be an important predictor of O3-induced inflammation; however, the causal
10 relationship has not been established.
11 The effects of dietary antioxidant supplementation on O3-induced pulmonary and
12 inflammatory response of young healthy individuals has been investigated by Samet et al.
13 (2001). Under controlled conditions, subjects received ascorbate restricted diet for three weeks.
14 After the first week of prescribed diet, subjects were randomly assigned into two groups, and
15 exposed to air (2 h, IE every 15 min at 20 L/min/m2 BSA). Thereafter, one group received daily
16 placebo pills and the other a daily supplement of ascorbate, a-tocopherol and a vegetable juice
17 for the next two weeks. At the end of a two week period subjects were exposed to 0.4 ppm O3
18 under otherwise similar conditions as in sham exposures. Serum concentration of antioxidants
19 determined prior to O3 exposure showed that subjects receiving supplements had substantially
20 higher concentrations of ascorbate, tocopherol and carotenoid in blood than the control group.
21 Plasma levels of glutathione and uric acid (cellular antioxidants) remained essentially the same.
22 Ozone exposure reduced spirometric lung function in both groups; however, the average
23 decrements in the supplementation group were smaller for FVC (p = 0.046) and FEVj
24 (p = 0.055) when compared to the placebo group. There was no significant correlation between
25 individual lung function changes and respective plasma levels of antioxidants. Individuals in
26 both groups experienced typical post O3 subjective symptoms of equal severity. Similarly, the
27 inflammatory response as assessed by BALF showed no significant differences between the two
28 groups either in the recovery of cellular components or the types and concentrations of
29 inflammatory cytokines. Because of the complexity of protocol, the study was not designed as a
30 cross-over type. However, it is unlikely that the fixed air-O3 sequence of exposures influenced
31 the findings in any substantial way. Although the study did not elucidate the protective
January 2005 AX6-63 DRAFT-DO NOT QUOTE OR CITE
-------
1 mechanisms, it has demonstrated the value of dietary antioxidants in attenuating lung function
2 effects of O3. This observation may appear to contradict the findings of Avissar's and colleagues
3 study (2000) discussed above; however, neither study found association between lung function
4 changes and glutathione levels. The lack of such association suggests that activation of
5 antioxidant protective mechanisms is seemingly independent of mechanisms eliciting lung-
6 function changes and that dietary antioxidants afford protection via a different pathway than
7 tissue-dependent antioxidant enzymes. Moreover, the findings of this study have provided
8 additional evidence that symptomatic, functional, inflammatory, and antioxidant responses are
9 operating through substantially independent mechanisms.
10 Further evidence that the levels and activity of antioxidant enzymes in ELF may not be
11 predictive or indicative of O3-induced lung function or inflammatory effects has been provided
12 by a study of Blomberg et al. (1999). No association was found between the respiratory tract
13 lining fluid redox potential level, an indicator of antioxidants balance, and either spirometric or
14 inflammatory changes induced by a moderate exposure of young individuals to O3 (0.2 ppm/2 h,
15 intermittent exercise at 20 L/min/m2 BSA). However, O3 exposure caused a partial depletion of
16 antioxidants (uric acid, GSH, EC-SOD) in nasal ELF and a compensatory increase in plasma uric
17 acid, affording at least some local protection (Mudway et al. 1999). More recently, Mudway
18 et al. (2001) investigated the effect of baseline antioxidant levels on response to a 2-h exposure
19 to 0.2 ppm O3 in 15 asthmatic and 15 healthy subjects. In the BALF of 15 healthy subjects,
20 significant O3-induced reductions in ascorbate and increases in glutathione disulphide and
21 EC-SOD were observed, whereas, levels were unaffected by O3 exposure in the asthmatics.
22 In both groups, BALF levels of uric acid and a-Tocopherol were unaffected by O3.
23 Trenga et al. (2001) studied the potential protective effects of dietary antioxidants (500 mg
24 vitamin C and 400 IU of vitamin E) on bronchial responsiveness of young to middle-aged
25 asthmatics who were prescreened for their hyperreactivity to SO2. Prior to the 1st exposure,
26 subjects took either two supplements or two placebo pills at breakfast time for 4 weeks. They
27 continued taking respective pills for another week when the 2nd exposure took place. The 45-min
28 exposures to air and 0.12 ppm O3 (15 min IE @ 3 x resting VE ) via mouthpiece were
29 randomized. Each exposure was followed by two 10-min challenges to 0.10 and 0.25 ppm SO2
30 with exercise to determine bronchial hyperresponsiveness. Due to variability of baseline lung
31 function at different test sessions, and the way the data have been presented, it is difficult to
January 2005 AX6-64 DRAFT-DO NOT QUOTE OR CITE
-------
1 interpret the results. The authors have reported no significant differences due to O3 between the
2 placebo and supplemented cohort in either lung function or bronchial hyperresponsiveness to
3 0.1 ppm SO2. However, post hoc classification of subjects into "mild" and "severe" asthmatics
4 (based on responses to SO2 during screening session) produced an unusual finding. In "severe"
5 asthmatics, the challenge with 0.25 ppm SO2 completely reversed O3-induced whereas 0.1 ppm
6 SO2-enhanced decrements for PEF and FEF25.75. The overall results of the study were interpreted
7 by the investigators as a demonstration of the protective effect of antioxidants from O3 exposure,
8 particularly in "severe" asthmatics. It has been repeatedly demonstrated that pulmonary function
9 response to O3 is reflex in origin, involving stimulation of bronchial C-fibers (see Section
10 AX6.2.5.1 for more information). With a relatively low O3 dose used in this study the reflex
11 response may be a dominant mechanism. Numerous animal and a few human studies used high
12 concentration of SO2 to "knock off the lung receptors. It is plausible, therefore, that the
13 reversal of O3-induced spirometric decrements are due to a suppression of bronchial C-fibers
14 activation and a subsequent reversal of a reflex response. Thus, the observed recovery of
15 "severe" asthmatics following the second SO2 challenge reported by Trenga and colleagues
16 (2001) may not be related to antioxidant protection.
17
18 AX6.5.7 Genetic and Other Factors
19 It has been repeatedly postulated that genetic factors may play an important role in
20 individual responsiveness to ozone. Recent studies (Bergamaschi et al., 2001; Corradi et al,
21 2002; Romieu et al, 2004) have indeed found that genetic polymorphisms of various enzymes,
22 namely NAD(P)H:quinone oxidoreductase (NQO1) and glutathione-S-trasferase Ml (GSTM1),
23 may play an important role in attenuating oxidative stress of airway epithelium. Bergamaschi
24 and colleagues (2001) studied young nonsmokers (15 F, 9 M; mean age 28.5 years) who cycled
25 for two hours on a cycling circuit in a city park on days with the average ozone concentration
26 ranging from 32 to 103 ppb. There was no control study group nor the intensity of bicycling has
27 been reported . Since spirometry was done within 30 min post-ride, it is difficult to gage how
28 much of the statistically significant (p = 0.026) mean decrement of 160 ml in FEVj of 8/24
29 individuals with NQO1 wild type and GSTM1 null genotypes was due to ozone. Individuals
30 with other genotype combinations including GSTM1 null had a mean post-ride decrement of
31 FEVj of only 40 mL. The post-ride serum level of Clara cell protein (CC16), a biomarker of
January 2005 AX6-65 DRAFT-DO NOT QUOTE OR CITE
-------
1 airway permeability, has been elevated in both subgroups. Only a "susceptible" subgroup
2 carrying NQO1 wild type in combination with GSTM1 null genotype, serum concentration of
3 CC16 showed positive correlation with ambient concentration of ozone and negative correlation
4 with FEVj changes. Despite some interesting observations, the study results should be
5 interpreted cautiously.
6 Pretreatment of healthy young subjects with inhaled corticosteroids (2 x 800 |ig/day
7 budesonide, a maximal clinical dose) for 2 weeks prior to O3 exposure (0.4 ppm/2 h, alternating
8 20 min exercise at SOW with 10 min rest) had no apparent effect on a typical lung function
9 decline or inflammatory response to exposure. Because of the complexity of the protocol, the
10 study was not a cross-over design and no control air exposures were conducted. Both the
11 placebo and treatment conditions caused the same magnitude of changes. Similarly, nonspecific
12 bronchial reactivity to methacholine (PC20 FEVj) was increased about the same 4 h after
13 exposure. Neither absolute nor relative sputum cell counts were affected by budesonide
14 treatment and O3 induced a typical neutrophilic response in both groups. Upregulation of pro-
15 inflammatory mediators measured in sputum was not different between the groups either. The
16 markers of inflammation and oxidative stress, exhaled NO and CO, as well as the reactive
17 product nitrite measured in exhaled breath condensate, respectively, were not significantly
18 influenced by budesonide. However, considering all these findings as a whole, budesonide
19 seemed to have a moderating, although not statistically significant, effect on O3-induced
20 response (Nightingale et al., 2000). Budesonide is an antiinflammatory drug that in laboratory
21 animal studies partially suppressed neutrophilic inflammation caused by O3 (Stevens et al.,
22 1994). Because the dose of budesonide was at therapeutic maximal levels, the pharmacologic
23 action of this drug and the site of action of O3 do not apparently coincide.
24
25
26 AX6.6 REPEATED EXPOSURES TO OZONE
27 Repeated daily exposure to O3 in the laboratory for 4 or 5 days leads to attenuated changes
28 in pulmonary function responses and symptoms (Hackney et al., 1977a; U.S. Environmental
29 Protection Agency, 1986, 1996). A summary of studies investigating FEVj responses to
30 repeated daily exposure for up to 5 days is given in Table AX6-8. The FEVj responses to
31 repeated O3 exposure typically have shown an increased response on the second exposure day
January 2005 AX6-66 DRAFT-DO NOT QUOTE OR CITE
-------
fa
to
O
o
X
ON
ON
O
l>
H
6
o
0
H
O
Table AX6-8. Changes in Forced Expiratory Volume in One Second ,
Ozone
Concentration1"
ppm
0.12
0.20
0.20
0.20
0.20
0.25
0.35
0.35
0.35
0.35
0.40
0.40
0.4
0.4
0.42
0.45
0.45
0.47
0.5
0.5
ug/m3
235
392
392
392
392
490
686
686
686
686
784
784
784
784
823
882
882
921
980
980
Exposure Duration
and Activity0
6.6 h, IE (40)
2 h, IE (30)
2 h, IE (18 and 30)
2 h, IE (18 and 30)
1 h, CE (60)
1 h, CE (63)
2 h, IE (30)
1 h, CE (60)
1 h, CE (60)
1 h, CE (60)
3 h, IE (4-5 x resting)
3 h, IE (4-5 x resting)
2 h, IE (65)
3 h, IE (32)
2 h, IE (30)
2 h, IE (27)
2 h, IE (27)
2 h, IE (3 x resting)
2 h, IE (30)
2. 5 h, IE (2 x resting)
After Repeated Daily Exposure to Ozone"
Number and Gender Percent Change in FEV; on Consecutive
of Subjects Exposure Days References'1
17M
10M
8M, 13 F
9
15M
4M,2F
5M,2F
10M
8M
10M
10M
15M
13M5
HFf
8M
8M, 2Fh
24 M
1M, 5F
10M,6F
8M, 2F8
8M
6
First
-12.79
+1.4
-3.0
-8.7
-5.02
-20.2
-18.8
-5.3
-31.0
-16.1
-14.4
-15.9
-9.2
-8.8
-18.0
-34.7
-21.1
-13.3
-5.8
-11.4
-8.7
-2.7
Second
-8.73
+2.7
-4.5
-10.1
-7.8
-34.8
—
-5.0
-41.0
-30.4
—
-24.6
-10.8
-12.9
-29.9
-31.1
-26.4
—
-5.6
-22.9
-16.5
-4.9
Third
-2.54
-1.6
-1.1
-3.2
—
—
-22.3
-2.2
-33.0
—
-20.6
-5.3
-4.1
-21.1
-18.5
-18.0
-22.8
-1.9
-11.9
-3.5
-2.4
Fourth Fifth
-0.6 0.2 Folinsbeeetal. (1994)
— — Folinsbeeetal. (1980)
— — Glineretal. (1983)
— — Glineretal. (1983)
— — Brookes etal. (1989)
— — Folinsbee and Horvath (1986)
— —
— — Folinsbeeetal. (1980)
-25.0 — Foxcroft and Adams (1986)
— — Schonfeld etal. (1989)
— —
— — Brookes etal. (1989)
-0.7 -1.0 Kulle etal. (1982)
-3.0 -1.6 Kulle etal. (1982)
-7.0 -4.4 Folinsbeeetal. (1998)
-12.0 -6.2 Gong etal. (1997b)
-6.3 -2.3 Horvath etal. (1981)
— — Bedi etal. (1985)
— — Bedi etal. (1989)
-4.3 — Linn etal. (1982)
— — Folinsbeeetal. (1980)
-0.7 — Hackney etal. (1977a)
"See Appendix A for abbreviations and acronyms.
bT,isteH from lowest to highest O, concentration
O
HH
H
W
'Exposure duration and intensity of IE or CE were variable; VE (number in parentheses) given in liters per minute or as a multiple of resting ventilation.
*For a more complete discussion of these studies, see Table AX6-9 and U.S. Environmental Protection Agency (1986).
"Subjects were especially sensitive on prior exposure to 0.42 ppm O3 as evidenced by a decrease in FEV[ of more than 20%. These nine subjects are a subset of the total group of
21 individuals used in this study.
'Bronchial reactivity to a methacholine challenge also was studied.
8Seven subjects completed entire experiment.
hSubjects had mild asthma.
-------
1 (Day 2) compared to the initial (Day 1) exposure response. This is readily apparent in repeated
2 exposures to a range of concentrations from 0.4 to 0.5 ppm O3 accompanied by moderate
3 exercise (Folinsbee et al., 1980; Horvath et al., 1981; Linn et al., 1982), and at lower
4 concentrations, 0.20 to 0.35 ppm, when accompanied by heavy exercise (Brookes et al., 1989;
5 Folinsbee and Horvath, 1986; Foxcroft and Adams, 1986; Schonfeld et al., 1989). Mechanisms
6 for enhanced pulmonary function responses on Day 2 have not been established, although
7 persistence of acute O3-induced damage for greater than 24 h may be important (Folinsbee et al.,
8 1993). An enhanced Day 2 FEVj response was less obvious or absent in exposures at lower
9 concentrations or those that caused relatively small group mean O3-induced decrements.
10 For example, Bedi et al. (1988) found no enhancement of the relatively small pulmonary
11 function responses in older subjects (median age, 65 years) exposed repeatedly to O3. Three
12 reports (Bedi et al., 1985; Folinsbee and Horvath, 1986; Schonfeld et al., 1989) demonstrated
13 that enhanced pulmonary function responsiveness was present within 12 h, lasted for at least
14 24 h and possibly 48 h, but was absent after 72 h.
15 After 3 to 5 days of consecutive daily exposures to O3, FEVj responses are markedly
16 diminished or absent. One study (Horvath et al., 1981) suggested that the rapidity of this decline
17 in FEVj response was related to the magnitude of the subjects' initial responses to O3 or their
18 "sensitivity." A summary of studies examining the effects of repeated exposures to O3 on FEVj
19 and other pulmonary function, symptoms, and airway inflammation is given in Table AX6-9.
20 Studies examining persistence of the attenuation of pulmonary function responses following
21 4 days of repeated exposure (Horvath et al., 1981; Kulle et al., 1982; Linn et al., 1982) indicate
22 that attenuation is relatively short-lived, being partially reversed within 3 to 7 days and typically
23 abolished within 1 to 2 weeks. Repeated exposures separated by 1 week (for up to 6 weeks)
24 apparently do not induce attenuation of the pulmonary function response (Linn et al., 1982).
25 Gong et al. (1997b) studied the effects of repeated exposure to 0.4 ppm O3 in a group of mild
26 asthmatics and observed a similar pattern of responses as those seen previously in healthy
27 subjects. The attenuation of pulmonary responses reached after 5 days of consecutive O3
28 exposure was partially lost at 4 and 7 days post exposure.
29 In addition to the significant attenuation or absence of pulmonary function responses after
30 several consecutive daily O3 exposures, symptoms of cough and chest discomfort usually
31 associated with O3 exposure generally are substantially reduced or absent (Folinsbee et al., 1980,
January 2005 AX6-68 DRAFT-DO NOT QUOTE OR CITE
-------
Table AX6-9. Pulmonary Function Effects with Repeated Exposures to Ozone"
to
o
o
X
Oi
ON
VO
H
6
o
o
H
O
Ozone
Concentration11 Exposure
ppm ug/m3 Activity
0.25 490 2 h IE, (30 min
rest, 30 min
exercise), VE =
39 L/min
0.2 392 4 h IE
(4 x 30 min
exercise), VE =
14.8 L/min/m2
BSA
0.2 392 4 h IE
(4 x 30 min
exercise), VE =
25 L/min/m2
BSA
0.4 784 3h/day for 5
days
IE ( 1 5 min rest,
15 min exercise)
VE = 32 L/min
0.12 235 6.6 h
50 min
exercise/10 min
rest, 30 min
lunch
VE = 38.8L/min
0.4 784 2 h IE (15 min
rest, 15 min
exercise
VE ~ 60 L/min
Number and
Gender of Subject
Exposure Conditions Subjects Characteristics
21.4°C 5M, 3F Healthy, NS
43.9%RH
4 days consecutive FA
exposure; 4 days
consecutive O3
exposure
1 day FA, 1 day, O3; 15 M, 8 F Healthy, NS
4 days consecutive 21 to 35 years old
exposure to O3
20 °C 50% RH (1 day, 9 M, 6 F Healthy, NS
O3; 4 days consecutive 23 to 37 years old
exposure to O3
31 °C 8M, 2F Mild asthma
35% RH adult
5 consecutive days
plus follow up @ 4 or
7 days
18°C 17 M Healthy NS
40% RH
five consecutive
daily exposures
5 days consecutive O3 16 M Healthy NS
exposure
Observed Effect(s)
FVC and FEVj decrements were significantly
attenuated on Day 4 of O3 exposure compared to day 1
of O3 exposure. Significant small airway function
depression accompanied by significant reutrophilia in
BALE one day following the end of O3 exposure.
FEVj decrement and symptoms significantly reduced
on Day 4 of O3 exposure compared to Day 1 of O3
exposure. Airway inflammation of mucosa persisted on
Day 4 although some inflammatory markers in BALE
attenuated significantly.
Significant decrease in FVC, FEVj, SRaw, and
symptoms on Day 4 of O3 exposure compared to a
single day of O3 exposure. Number of PMNs,
fibronectin, and IL6 in BALE were significantly
decreased on Day 4 compared to a single day of O3
exposure.
FEVj decreased 35% on day 1 and only 6% on day 5.
Bronchial reactivity increased after day 1 and remained
elevated. Adaptation of asthmatics is similar to healthy
subjects but may be slower and less complete.
FEVj responses were maximal on first day of exposure
(- 13%), less on second day (-9%), absent thereafter.
Symptoms only the first 2 days. Methacholine airway
responsiveness was at least doubled on all exposure
days, but was highest on the second day of O3. Airway
responsiveness was still higher than air control after 5
days of O3 exposure. Trend to lessened response, but it
was not achieved after 5 days.
O3-exposure FEVj decrement was greater on day 2,
29.9%, than day 1, 18.0%, then decreased on day 3,
21 .1%, day 4, 7% and day 5, 4.4%
Reference
Frank et al.
(2001)
Jorres et al.
(2000)
Christian et al.
(1998)
Gong et al.
(1997b)
Folinsbee et al.
(1994)
Folinsbee et al.
(1998)
Devlin et al.
(1997)
o
HH
H
W
-------
to
o
o
Table AX6-9 (cont'd). Pulmonary Function Effects with Repeated Exposures to Ozonea
X
Ozone
Concentration11
ppm ug/m3 and Activity
0.45 882 2 h
IE
Exposure Conditions
23.3 °C
63% RH
Number
and Gender
of Subjects
10 M, 6F
Subject
Characteristics
Healthy NS
60 to 89 years old
Observed Effect(s)
Overall increase in symptoms, but no single
symptom increased significantly. FVC
Reference
Bedi et al.
(1989)
(3 x 20 min exercise)
0.20/0.20 392/392 1 h
0.35/0.20 686/392 CEat60L/min
0.35/0.35 686/686
0.35
686
0.45
882
60 min CE
VB = 60 L/min
2hIE
(3 x 20 min exercise)
Exposed for 3 consecutive
days, not exposed for 2 days,
then exposed to 0.45 ppm
again for 1 day
21 to 25 °C
40 to 60% RH
(three 2-day sets
of exposures)
21 to 25 °C
40 to 60% RH
(two exposures for each
subject separated by 24, 48,
72, or 120 h)
23.3 °C
62.5% RH
(three exposures with
a minimum 1-week interval)
15M
40 M
(4 groups
of 10)
8M, 8F
median 65 years old;
mean FVC = 3.99 L;
meanFEV^S.Ol L;
FEV/FVC range =
61 to 85%
Healthy aerobically
trained NS,
FVC = 4.24 to 6.98 L
NS; nonallergic,
non-Los Angeles
residents for > 6 mo;
=25 years old
Healthy NS, 61 years
old for M and
65 years old for F
(FVC = 4.97 L for
M and 3.11 LforF)
decreased 111 mL and 104 mL on Days 1 and 2,
respectively. FEVl fell by 171 and 164 mL, and
FEV3 fell by 185 and 172 mL. No significant
changes on Days 3 and 4 or with FA. FEVl
changes were -5.8, -5.6, -1.9, and -1.7% on
the four O3 days.
Consecutive days of exposure to 0.20 ppm Brookes et al.
produced similar FEVj responses on each day (1989)
(-5.02, -7.80); 0.35/0.20 ppm pair caused
increased response to 0.20 ppm on second day
(-8.74); 0.35/0.35 ppm caused much increased
response on Day 2 (-15.9, -24.6). Symptoms
were worse on the second exposure to 0.35 ppm,
but not with second exposure to 0.20 ppm.
No differences between responses to exposures Schonfeld et al.
separated by 72 or 120 h. Enhanced YEVl (1989)
response at24 h (-16.1% vs. -30.4%). Possible
enhanced response at 48 h (-14.4% vs.
-20.6%). Similar trends observed for breathing
pattern and SRaw.
Spirometric changes were not reproducible from Bedi et al.
time to time after O3 exposure R < 0.50). (1988)
Repeat exposures to air yielded consistent
responses.
0.18 353 2h
IE (heavy)
VE = 60 to "
(35 L/min/m2 BSA)
31°C 59 adult
35% RH Los Angeles
(screen exposures residents
in spring 1986; 12
second exposures in responsive
summer/fall 1986 and winter 13
1987 and spring 1987 for nonresponsi
responders and ve
nonresponders only)
Responders:
19 to 40 years old
6 atopic,
2 asthmatic,
4 normal
Nonresponders:
18 to 39 years old,
13 normal
Responders had AFEVj = -12.4% after initial
screening; nonresponders had no change.
Responders had nonsignificant response in late
summer or early winter, but were responsive
again in early spring (spring 1986, -385 mL;
Autumn 1986, -17 mL; winter 1987, +16 mL;
spring 1987, -347 mL). Nonresponders did not
change with season. Suggests that responders
responses may vary with ambient exposure, but
nonresponders generally remain nonresponsive.
Linn et al.
(1988)
(also see
Hackney et al..
1989)
-------
to
o
o
Table AX6-9 (cont'd). Pulmonary Function Effects with Repeated Exposures to Ozonea
Ozone
Concentration11
ppm ug/m3 and Activity
Number
and Gender Subject
Exposure Conditions of Subjects Characteristics
Observed Effect(s) Reference
0.45
(+ 0.30
PAN)
882
2h
IE (20 min rest,
20 min exercise)
VB = 27 L/min
22 °C
60% RH
5 days consecutive
exposure to PAN + O3
3 M, 5 F Healthy NS,
Mean age = 24 years
FEVj decreased ~ 19% with O3 alone, ~ 15% on
Day 1 of O3 + PAN, =5% on Day 5 of O3 +
PAN, =7% 3 days after 5 days of O3 + PAN,
= 15% after 5 days of O3 + PAN. Similar to
other repeated O3 exposure studies, O3 responses
peaked after 2 days, were depressed 3 days later,
and responses returned 7 days later. PAN
probably had no effect on repeated to O3
exposure responses.
Drechsler-Parks
etal. (1987b)
(also see
Table AX6-14)
X
H
6
o
o
H
O
0.35 686 =lhCE 22 to 25 °C 8M
(see paper for details) 35 to 50% RH
(1 day FA; 1 day O3; 4 days
consecutive exposure to O3)
Aerobically trained
healthy NS (some
were known O3
sensitive),
22.4 ±2.2 years old
Largest FEV; decrease on second of 4 days O3
exposure (-40% mean decrease). Trend for
attenuation of pulmonary function response not
complete in 4 days. VOj^ decreased with
single acute O3 exposure (-6%) but was not
significant after 4 days of O3 exposure (-4%).
Performance time was less after acute O3 (21 1 s)
exposure than after FA (253 s).
Foxcroft and
Adams (1986)
"See Appendix A for abbreviations and acronyms.
bListed from lowest to highest O, concentration.
o
HH
H
W
-------
1 1994; Foxcroft and Adams, 1986; Linn et al., 1982). Airway responsiveness to methacholine is
2 increased with an initial O3 exposure (Holtzman et al., 1979; Folinsbee et al., 1988), may be
3 further increased with subsequent exposures (Folinsbee et al.,1994), and shows a tendency for
4 the increased response to diminish with repeated exposure (Dimeo et al., 1981; Kulle et al.,
5 1982). The initially enhanced and then lessened response may be related to changes that occur
6 during the repair of pulmonary epithelia damaged as a consequence of O3 exposure.
7 Inflammatory responses (Koren et al., 1989a), epithelial damage, and changes in permeability
8 (Kehrl et al., 1987) might explain a portion of these responses. By blocking pulmonary function
9 responses and symptoms with indomethacin pretreatment, Schonfeld et al. (1989) demonstrated
10 that in the absence of an initial response, pulmonary function and symptoms effects were not
11 enhanced on Day 2 by repeated exposure to 0.35 ppm O3. These results suggest that airway
12 inflammation and the release of cyclooxygenase products of arachidonic acid play a role in the
13 enhanced pulmonary function responses and symptoms observed upon reexposure to O3 within
14 48 h.
15 Response to laboratory O3 exposure as a function of the season of the year in the South
16 Coast Air Basin of Los Angeles, CA, has been examined in several studies (Avol et al., 1988;
17 Hackney et al., 1989; Linn et al., 1988). Their primary purpose was to determine whether O3
18 responsive subjects would remain responsive after regular ambient exposure during the "smog
19 season". The subjects were exposed to 0.18 ppm O3 for 2 h with heavy IE on four occasions,
20 spring, fall, winter, and the following spring. The marked difference in FEVj response between
21 responsive and nonresponsive subjects seen initially (-12.4% versus +1%) no longer was present
22 after the summer smog season (fall test) or 3 to 5 months later (winter test). However, when the
23 subjects were exposed to O3 during the following spring, the responsive subjects again had
24 significantly larger changes in FEVl5 suggesting a seasonal variation in response.
25 Brookes et al. (1989) and Gliner et al. (1983) tested whether initial exposure to one O3
26 concentration could alter response to subsequent exposure to a different O3 concentration.
27 Gliner et al. (1983) showed that FEVj response to 0.40 ppm O3 was not influenced by previously
28 being exposed to 0.20 ppm O3 for 2 h on 3 consecutive days. Brookes et al. (1989) found
29 enhanced FEVj and symptoms upon exposure to 0.20 ppm after previous exposure to 0.35 ppm
30 O3. These observations suggest that, although preexposure to low concentrations of O3 may not
January 2005 AX6-72 DRAFT-DO NOT QUOTE OR CITE
-------
1 influence responses to higher concentrations, preexposure to a high concentration of O3 can
2 significantly increase responses to a lower concentration on the following day.
3 Foxcroft and Adams (1986) demonstrated that decrements in exercise performance seen
4 after 1 h of exposure to 0.35 ppm O3 with heavy CE were significantly less after 4 consecutive
5 days exposure than they were after a single acute exposure. Further, exercise performance,
6 VO2max, VEmax and HR^ were not significantly different after 4 days of O3 exposure compared to
7 those observed in a FA exposure. Despite the change in exercise performance, Foxcroft and
8 Adams (1986) did not observe a significant attenuation of FEVj response, although symptoms
9 were significantly reduced. However, these investigators selected known O3-sensitive subjects
10 whose FEVj decrements exceeded 30% on the first 3 days of exposure. The large magnitude of
11 these responses, the trend for the responses to decrease on the third and fourth day, the decreased
12 symptoms, and the observations by Horvath et al. (1981) that O3-sensitive subjects adapt slowly,
13 suggest that attenuation of response would have occurred if the exposure series had been
14 continued for another 1 or 2 days. These observations support the contention advanced by
15 Horvath et al. (1981) that the progression of attenuation of response is a function of initial "O3
16 sensitivity."
17 Drechsler-Parks et al. (1987b) examined the response to repeated exposures to 0.45 ppm O3
18 plus 0.30 ppm peroxyacetyl nitrate (PAN) in 8 healthy subjects and found similar FEVj
19 responses to exposures to O3 (-19%) and to O3 plus PAN (-15%). Thus, PAN did not increase
20 responses to O3. Further, repeated exposure to the PAN plus O3 mixture resulted in similar
21 changes to those seen with repeated O3 exposure alone. The FEVj responses fell to less than
22 - 5% after the fifth day, with the attenuation of response persisting 3 days after the repeated
23 exposures, but being absent after 7 days. These observations suggest that PAN does not
24 influence the attenuation of response to repeated O3 exposure. If the PAN responses are
25 considered negligible, this study confirms the observation that the attenuation of O3 responses
26 with chamber exposures lasts no longer than 1 week. [More discussion on the interaction ofO3
27 with other pollutants can be found in Section AX6.11.]
28 Folinsbee et al. (1993) exposed a group of 16 healthy males to 0.4 ppm O3 for 2 h/day on
29 5 consecutive days. Subjects performed heavy IE (VE = 60 to 70 L/min). Decrements in FEVj
30 averaged 18.0, 29.9, 21.1, 7.0, and 4.4% on the 5 exposure days. However, baseline preexposure
31 FEVj decreased from the first day's preexposure measurement and was depressed by an average
January 2005 AX6-73 DRAFT-DO NOT QUOTE OR CITE
-------
1 of about 5% by the third day. This study illustrates that, with high-concentration and heavy-
2 exercise exposures, pulmonary function responses may not be completely recovered within 24 h.
3 During this study, BALF also was obtained immediately after the Day 5 exposure, with results
4 reported by Devlin et al. (1997). These authors found that some inflammation and cellular
5 responses associated with acute O3 exposure were also attenuated after 5 consecutive days of O3
6 exposure (compared to historical data for responses after a single-day exposure), although
7 indicators of epithelial cell damage—not seen immediately after acute exposure—were present
8 in BALF after the fifth day of exposure. When reexposed again 2 weeks later, changes in BALF
9 indicated that epithelial cells appeared fully repaired (Devlin et al., 1997).
10 Frank et al. (2001) exposed 8 healthy young adults to 0.25 ppm O3 for 2 h with moderate
11 IE (exercise VE = 40 L/min) on 4 consecutive days. In addition to standard pulmonary function
12 measures, isovolumetric FEF25.75, Vmax50 and Vmax75 were grouped into a single value representing
13 small airway function (SAWgrp). Exercise ventilatory pattern was also monitored each day,
14 while peripheral airway resistance was measured by bronchoscopy followed by lavage on Day 5.
15 The authors observed two patterns of functional response in their subjects— attenuation and
16 persistent. Values of FVC and FEVj showed significant attenuation by Day 4 compared to Day
17 1 values. However, SAWgrp and rapid shallow breathing during exercise persisted on Day 4
18 compared to Day 1, and were accompanied by significant neutrophilia in BALF 1 day following
19 the end of O3 exposure. Frank et al. (2001) suggested that both types of functional response (i.e.,
20 attenuation and persistence) are linked causally to inflammation. They contend that the
21 attenuation component is attributable at least in part to a reduction in local tissue dose during
22 repetitive exposure that is likely to result from the biochemical, mechanical, and morphological
23 changes set in motion by inflammation. They speculated that the persistent component
24 represents the inefficiencies incurred through inflammation. Whether the persistent small airway
25 dysfunction is a forerunner of more permanent change in the event that oxidant stress is extended
26 over lengthy periods of time is unknown.
27 Early repeated multihour (6 to 8 h) exposures focused on exposures to low concentrations
28 of O3 between 0.08 and 0.12 ppm (Folinsbee et al., 1994; Horvath et al., 1991; Linn et al., 1994).
29 Horvath et al. (1991) exposed subjects for 2 consecutive days to 0.08 ppm using the 6.6-h
30 prolonged exposure protocol (see Table AX6-2). They observed small pre- to postexposure
31 changes in FEVj (-2.5%) on the first day, but no change on the second day. Linn et al. (1994)
January 2005 AX6-74 DRAFT-DO NOT QUOTE OR CITE
-------
1 observed a 1.7% decrease in FEVj in healthy subjects after 6.6 h exposure to 0.12 ppm O3.
2 A second consecutive day exposure to O3 yielded even smaller (< 1%) responses. In a group of
3 asthmatics exposed under similar conditions (Linn et al., 1994), the FEVj response on the first
4 day was -8.6% which was reduced to -6.7% on day 2, both significantly greater than those
5 observed for the nonasthmatics group. The observations of Horvath et al. (1991) and Linn et al.
6 (1994) elicited a somewhat different pattern of response (no enhancement of response after the
7 first exposure) than that seen at higher concentrations in 2 h exposures with heavy exercise
8 (Tables AX6-8 and AX6-9). However, the subjects studied by Horvath et al. (1991) were
9 exposed only to 0.08 ppm O3 and were somewhat older (30 to 43 yrs) than the subjects studied
10 by Folinsbee et al. (1994), mean age of 25 yrs, while the nonasthmatic subjects studied by Linn
11 et al. (1994) were also older (mean = 32 yrs), had lower exercise VE (-20%) and were residents
12 of Los Angeles who often encountered ambient levels of O3 at or above 0.12 ppm.
13 Folinsbee et al. (1994) exposed 17 subjects to 0.12 ppm O3 for 6.6 h, with 50 min of
14 moderately heavy exercise ( VE = 39 L/min) each hour, on 5 consecutive days. Compared with
15 FA, the percentage changes in FEVi over the five days were -12.8%, -8.7%, -2.5%, -0.06%,
16 and +0.18%. A parallel attenuation of symptoms was observed, but the effect of O3 in enhancing
17 airway responsiveness (measured by increase in SRaw upon methacholine challenge) over
18 5 days was not attenuated (3.67, 4.55, 3.99, 3.24, and 3.74, compared to 2.22 in FA control).
19 Nasal lavage revealed no increases in neutrophils except on the first O3 exposure day.
20 Christian et al. (1998) exposed 15 adults (6 females and 9 males; mean age = 29.1 yrs) to
21 4 consecutive days at 0.20 ppm O3 for 4 h, with 30 mm of IE (exercise VE = 25 L/min/m2) each
22 hour. Measures of FEVb FVC, and symptoms were all significantly reduced on Day 1, further
23 decreased on Day 2, and then attenuated to near FA control values on Day 4. The pattern of
24 SRaw response was similar, being greatest on Day 2 and no different from FA control on Day 4.
25 BAL was done on Day 5 and showed that neutrophil recruitment to the respiratory tract was
26 attenuated with repeated short-term exposures, compared to Day 1 control O3 exposure, while
27 airway epithelial injury appeared to continue as reflected by no attenuation of IL-6, IL-8, total
28 protein, and LDH. The authors concluded that such injury might lead to airway remodeling,
29 which has been observed in several animal studies (Brummer et al., 1977; Schwartz et al., 1976;
30 Tepper et al., 1989; Van Bree et al., 1989). In a similar study to that of Christian et al. (1998),
31 Torres et al. (2000) exposed 23 adults (8 females and 15 males; mean age = 27.9 yrs) on
January 2005 AX6-75 DRAFT-DO NOT QUOTE OR CITE
-------
1 4 consecutive days to 0.20 ppm O3 for 4 h, with 30 min of IE (exercise VE = 26 L/min) each
2 hour. The authors observed that FEVj was significantly reduced and symptoms were
3 significantly increased on Day 1. On Day 2, FEVj was further decreased, while symptoms
4 remained unchanged. By Day 4, both FEVj and symptoms were attenuated to near FA, control
5 values. Twenty hours after the Day 4 exposure, BAL and bronchial mucosal biopsies were
6 performed. These authors found via bronchial mucosal biopsies that inflammation of the
7 bronchial mucosa persisted after repeated O3 exposure, despite attenuation of some inflammatory
8 markers in BALF and attenuation of lung function responses and symptoms. Further, Torres
9 et al. (2000) observed persistent although small decrease in baseline FEVj measured before
10 exposure, thereby suggesting that there are different time scales of the functional responses to
11 O3, which may reflect different mechanisms. The levels of protein remaining elevated after
12 repeated exposures confirms the findings of others (Christian et al., 1998; Devlin et al., 1997),
13 and suggests that there is ongoing cellular damage irrespective of the attenuation of cellular
14 inflammatory responses with the airways. [Further discussion on the inflammatory responses to
15 O3 can be found in Section AX6.9. ]
16 Based on studies cited here and in the previous O3 criteria documents (U.S. Environmental
17 Protection Agency, 1986, 1996), several conclusions can be drawn about repeated 1- to 2-h O3
18 exposures. Repeated exposures to O3 can cause an enhanced (i.e., greater) lung function
19 response on the second day of exposure. This enhancement appears to be dependent on the
20 interval between the exposures (24 h is associated with the greatest increase) and is absent with
21 intervals > 3 days. As shown in Figure AX6-8, an enhanced response also appears to depend on
22 O3 concentration and to some extent on the magnitude of the initial response. Small responses to
23 the first O3 exposure are less likely to result in an enhanced response on the second day of O3
24 exposure. Repeated daily exposure also results in attenuation of pulmonary function responses,
25 typically after 3 to 5 days of exposure. This attenuated response persists for less than 1 week or
26 as long as 2 weeks. In temporal conjunction with the pulmonary function changes, symptoms
27 induced by O3, such as cough and chest discomfort, also are attenuated with repeated exposure.
28 Ozone-induced changes in airway responsiveness attenuate more slowly than pulmonary
29 function responses and symptoms. Attenuation of the changes in airway responsiveness appear
30 to persist longer than changes in pulmonary function, although this has been studied only on a
31 limited basis. In longer-duration (6.6 h), lower-concentration studies that do not cause an
January 2005 AX6-76 DRAFT-DO NOT QUOTE OR CITE
-------
c
o
o
•s
HI
100-
95-
90-
85-
80-
75-
70-
65-
T I I I I I
0.0 0.5 1.0 1.5 2.0 2.5
Time (days)
i
3.0
I
3.5
4.0
Figure AX6-8. Regression curves were fitted to day-by-day postexposure FEVt values
obtained after repeated daily acute exposures to O3 for 2 to 3 h with
intermittent exercise at a VE of 24 to 43 L/min (adaptation studies).
Symbols represent the results from individual studies conducted at 0.2 ppm
for 2 h (+), 0.35 ppm for 2 h (•), 0.4 ppm for 2 h (+), 0.5 ppm for 2 h (#),
and 0.54 ppm for 3 h (A). Also shown for comparison are the FEVt values
obtained after exposure to 0.12 ppm O3 for 10 h (•).
Source: Modified from Hazucha (1993).
1 enhanced second-day response, the attenuation of response to O3 appears to proceed more
2 rapidly. Inflammatory markers from BALF on the day following both 2 h (Devlin et al., 1997)
3 and 4 h (Christian et al., 1998; Torres et al., 2000) repeated O3 exposure for 4 days indicate that
4 there is ongoing cellular damage irrespective of the attenuation of some cellular inflammatory
5 responses of the airways, lung function responses and symptoms.
January 2005
AX6-77
DRAFT-DO NOT QUOTE OR CITE
-------
1 AX6.7 EFFECTS ON EXERCISE PERFORMANCE
2 AX6.7.1 Introduction
3 In an early epidemiologic study examining race performances in Los Angeles area high
4 school cross-country runners, Wayne et al. (1967) observed that endurance exercise performance
5 was depressed by inhalation of ambient oxidant air pollutants. The authors concluded that the
6 detrimental effects of oxidant air pollutants on race performance might have been related to the
7 associated discomfort in breathing, thus limiting the runners' motivation to perform at high
8 levels, although physiologic effects limiting O2 availability could not be ruled out.
9 Subsequently, the effects of acute O3 inhalation on endurance exercise performance have been
10 examined in numerous controlled laboratory studies. These studies were discussed in the
11 previous O3 criteria document (U.S. Environmental Protection Agency, 1996) in two categories:
12 (1) those that examined the effects of acute O3 inhalation on maximal oxygen uptake (VO2max)
13 and (2) those that examined the effects of acute O3 inhalation on the ability to complete
14 strenuous continuous exercise protocols of up to 1 h in duration. In this section, major
15 observations in these studies are briefly reviewed with emphasis on reexamining the primary
16 mechanisms causing decrements in VO2max and endurance exercise performance consequent to
17 O3 inhalation. A summary of major studies of O3 inhalation effects on endurance exercise
18 performance, together with observed pulmonary function and symptoms of breathing discomfort
19 responses, is given in Table AX6-10.
20
21 AX6.7.2 Effect on Maximal Oxygen Uptake
22 Three early studies (Folinsbee et al., 1977; Horvath et al., 1979; Savin and Adams, 1979)
23 examining the effects of acute O3 exposures on VO2max were reviewed in an earlier O3 criteria
24 document (U.S. Environmental Protection Agency, 1986). Briefly, Folinsbee et al. (1977)
25 observed that VO2max was significantly decreased (10.5%) following a 2-h exposure to 0.75 ppm
26 O3 with light (50 Watts) IE. Reduction in VO2max was accompanied by a decrease in maximal
27 ventilation, maximal heart rate, and a large decrease in maximal tidal volume. In addition, the
28 2-h IE O3 exposure resulted in a 22.3% decrease in FEVj and significant symptoms of cough and
29 chest discomfort. In contrast, Horvath et al. (1979) did not observe a change in VO2max or other
30 maximal cardiopulmonary endpoints in subjects exposed for 2 h at rest to either 0.50 or
January 2005 AX6-78 DRAFT-DO NOT QUOTE OR CITE
-------
Table AX6-10. Ozone Effects on Exercise Performance"
3
C
to
o
o
^>
X
ON
i
VO
O
!>
!-rj
H
6
o
2!
-*— 1
o
H
O
o
NW'
H
W
0
o
HH
H
W
Ozone
Concentration b
ppm
0.06-0.07
0.12-0.13
0.18
0.35
0.12
0.20
0.12
0.18
0.24
0.21
0.20
0.35
0.25
0.50
0.75
0.15
0.30
0.75
Hg/m3
120-140
245-260
353
686
235
392
235
353
470
412
392
686
490
980
1,470
294588
1,470
xauii; r^^vu
Exposure Duration Exposure
and Activity Conditions
CE 23 to 24.5 °C
(VE = 30 to 120 L/min) 50 to 53% RH
16 to 28 min progressive
maximum exercise
protocol
1 h CE or competitive NA
simulation at mean
VE = 94 L/min
50 min CE 22 to 25 °C
VE = 60 L/min 35 to 50% RH
IhCE 31 °C
VE = 89 L/min
1 h competitive 23 to 26 °C
simulation exposures at 45 to 60% RH
mean
VE = 87 L/min
1 h CE at 75% V02max 19 to 21 °C
60 to 70% RH
1 h CE or competitive 23 to 26 °C
simulation at mean 45 to 60% RH
VE = 77.5 L/min
2 h rest NA
—30 min, progressively 23 °C
incremented exercise to 50% RH
voluntary exhaustion
2 h IE NA
(4x 15 min light [50 W]
bicycle ergometry)
'-AV. V^i/LFllC i^llCVia LFll iliACl
Number and
Gender of Subject
Subjects Characteristics
12 M, 12 F Athletic
# not given; Well-trained
all males distance runners
8 M Trained
nonathletes
15M, 2F Highly trained
competitive
cyclists
10 M Highly trained
competitive
cyclists
6 M, 1 F Well-trained
cyclists
10 M Well-trained
distance runners
8M, 5F
9 M Healthy, NS
21 to 44 years old
13 M 4 light S,
9NS
viac i ci iiri UKIIIVI;
Observed Effect(s)
Reduced maximum performance time and increased symptoms
of breathing discomfort during O3 exposure.
Maximal treadmill run time reduced from 71.7 min in FA to
66.2 min during O3 exposure with no decease in arterial O2
saturation.
VT decreased, fB increased with 50-min O3 exposures; decrease
in FVC, FEVj, FEF25.75, performance time, VO2max, V^^, and
HR,,,,,, from FA to 0.35-ppm O3 exposure.
Decrease in VEmax, VO2max, VTmax, workload, ride time, FVC,
and FEVj with 0.20 ppm O3 exposure, but not significant with
0. 12-ppm O3 exposure, as compared to FA exposure.
Decrease in exercise time of 7.7 min and 10. 1 min for subjects
unable to complete the competitive simulation at 0.18 and 0.24
ppm O3, respectively; decrease in FVC and FEVj for 0.18-and
0.24-ppm O3 exposure compared with FA exposure.
Decrease in FVC, FEVj, FEF25.75, and MVV with 0.21 ppm O3
compared with FA exposure.
VT decreased and fB increased with continuous 50-min O3
exposures; decrease in FVC, FEVj, and FEF25.75 from FA to
0.20 ppm and FA to 0.35-ppm O3 exposure in all conditions;
three subjects unable to complete continuous and competitive
protocols at 0.35 ppm O3.
FVC decreased with 0.50- and 0.75-ppm O3 exposure
compared with FA; 4% nonsignificant decrease in mean
VO2max following 0.75 ppm O3 compared with FA exposure.
Exposure to 0. 15 and 0.30 ppm O3 did not decrease maximal
exercise performance or VO2max compared to FA.
No significant pulmonary function or symptom responses were
observed, although a trend (P < .10) was evident.
Decrease in FVC, FEVj, ERV, 1C, and FEF50% after 1-h
0.75-ppm O3 exposure; decrease in VO2max, VTmax, VEmax,
maximal workload, and HRmax following 0.75-ppm O3
exposure compared with FA.
Reference
Linder et al.
(1988)
Folinsbee et al.
(1986)
Foxcroft and
Adams (1986)
Gong et al.
(1986)
Schelegle and
Adams (1986)
Folinsbee et al.
(1984)
Adams and
Schelegle (1983)
Horvath et al.
(1979)
Savin and
Adams (1979)
Folinsbee et al.
(1977)
"See Appendix A for abbreviations and acronyms.
""Listed from
lowest to highest O3 concentration.
-------
1 0.75 ppm, although FVC was significantly decreased 10% following the latter exposure. Without
2 preliminary exposure to O3, Savin and Adams (1979) examined the effects of a 30-min exposure
3 to 0.15 and 0.30 ppm O3 while performing a progressively incremented exercise test to volitional
4 fatigue (mean = 31.5 min in FA). No significant effect on maximal work time or VO2max was
5 observed compared to that observed upon FA exposure. Further, no significant effect on
6 pulmonary function, maximal heart rate, and maximal tidal volume was observed, although
7 maximal VE was significantly reduced 7% in the 0.30 ppm O3 exposure. Results of these early
8 studies suggest that VO2max is reduced if the incremented maximal exercise test is preceded by an
9 O3 exposure of sufficient total inhaled dose of O3 to result in significant pulmonary function
10 decrements and symptoms of breathing discomfort.
11 Using trained nonathletes, Foxcroft and Adams (1986) observed significant (p < 0.05)
12 reductions in rapidly incremented VO2max exercise performance time (-16.7%), VO2max (-6.0%),
13 maximal VE (-15.0%), and maximal heart rate (-5.6%) immediately following an initial 50-min
14 exposure to 0.35 ppm O3 during heavy CE (VE = 60 L/min). These decrements were
15 accompanied by a significant reduction in FEVj (-23%) and the occurrence of marked
16 symptoms of breathing discomfort. Similarly, Gong et al. (1986) found significant reductions in
17 rapidly incremented VO2max exercise performance time (-29.7%), VO2max (-16.4%), maximal VE
18 (-18.5%), and maximal workload (-7.8%) in endurance cyclists immediately following a 1-h
19 exposure to 0.20 ppm O3 with very heavy exercise ( VE 89 L/min), but not following exposure to
20 0.12 ppm. Gong et al. (1986) observed only a 5.6% FEVj decrement and mild symptoms
21 following exposure to 0.12 ppm, but a large decrement in FEVj (-21.6%) and substantial
22 symptoms of breathing discomfort following the 0.20 ppm exposure, which the authors
23 contended probably limited maximal performance and VO2max.
24
25 AX6.7.3 Effect on Endurance Exercise Performance
26 A number of studies of well trained endurance athletes exposed to O3 have consistently
27 observed an impairment of 1-h continuous heavy exercise performance of some individuals
28 (Adams and Schelegle, 1983; Avol et al., 1984; Folinsbee et al., 1984; Gong et al., 1986). The
29 performance impairment is indicated by an inability to complete the prescribed O3 exposures
30 (even at concentrations as low as 0.16 ppm) that subjects were able to complete in FA (Avol
January 2005 AX6-80 DRAFT-DO NOT QUOTE OR CITE
-------
1 et al., 1984). Other indications of impaired endurance exercise performance upon exposure to O3
2 include a -7.7% reduced endurance treadmill running time when exposed to 0.18 ppm O3
3 (Folinsbee et al., 1986), which was accompanied by significantly decreased FEVj and
4 significantly elevated symptoms of breathing discomfort. Another study (Schelegle and Adams,
5 1986) observed the failure of some trained endurance athletes to complete a 1-h competitive
6 simulation protocol upon exposure to O3 (30 min warm-up, followed immediately by 30 min at
7 the maximal workload that each subject could just maintain in FA; mean VE = 120 L/min).
8 In this study, all subjects (n = 10) completed the FA exposure, whereas one, five, and seven
9 subjects could not complete the 0.12, 0.18, and 0.24 ppm O3 exposures, respectively. Following
10 the 0.18 ppm and 0.24 ppm O3 exposures, but not the 0.12 ppm exposure, FEVj was reduced
11 significantly and symptoms were significantly increased. Linder et al. (1988) also observed
12 small decrements in performance time (1 to 2 min) during a progressive maximal exercise test
13 (mean = 21.8 min) at O3 concentrations of 0.065 and 0.125 ppm. These small effects were
14 accompanied by a significant increase in subjective perception of overall effort at 0.125 ppm, but
15 with no significant reduction in FEVj at either O3 concentration. Collectively, reduced
16 endurance exercise performance and associated pulmonary responses are clearly related to the
17 total inhaled dose of O3 (Adams and Schelegle, 1983; Avol et al., 1984; Schelegle and Adams,
18 1986).
19 Mechanisms limiting VO2max and maximal exercise performance upon O3 exposure have
20 not been precisely identified. Schelegle and Adams (1986) observed no significant effect of O3
21 on cardiorespiratory responses, and there was no indirect indication that arterial O2 saturation
22 was affected. The latter is consistent with the observation that measured arterial O2 saturation at
23 the end of a maximal endurance treadmill run was not affected by O3 (Folinsbee et al., 1986).
24 In studies in which O3 inhalation resulted in a significant decrease in VO2max, and/or maximal
25 exercise performance, significantly decreased FEVj and marked symptoms of breathing
26 discomfort were observed (Adams and Schelegle, 1983; Avol et al., 1984; Folinsbee et al., 1977,
27 1984, 1986; Foxcroft and Adams, 1986; Gong et al., 1986; Schelegle and Adams, 1986).
28 However, Gong et al. (1986) observed rather weak correlations between FEVj impairment and
29 physiological variable responses during maximal exercise (R = 0.26 to 0.44). Rather, these
30 authors concluded that substantial symptoms of breathing discomfort consequent to 1 h of very
31 heavy exercise while exposed to 0.20 ppm O3, probably limited maximal performance and
January 2005 AX6-81 DRAFT-DO NOT QUOTE OR CITE
-------
1 VO2max either voluntarily or involuntarily (Gong et al., 1986). Strong support for this contention
2 is provided by the observation of significant increases in VO2max (4.7%) and maximal
3 performance time (8.8%) following four consecutive days of 1 h exposure to 0.35 ppm O3 with
4 heavy exercise (VE =60 L/min) compared to initial O3 exposure (Foxcroft and Adams, 1986).
5 These improvements, which were not significantly different from those for FA, were
6 accompanied by a significant reduction in symptoms of breathing discomfort with no significant
7 attenuation of FEVj and other pulmonary function responses. In this regard, Schelegle et al.
8 (1987) observed a disparate effect of indomethacin pretreatment (an inhibitor of the cyclo-
9 oxygenation of arachidonic acid to prostaglandins associated with inflammatory responses) on
10 O3-induced pulmonary function response (significant reduction) and an overall rating of
11 perceived exertion and symptoms of pain on deep inspiration and shortness of breath (no
12 significant effect).
13
14
15 AX6.8 EFFECTS ON AIRWAY RESPONSIVENESS
16 Increased airway responsiveness, also called airway hyperresponsiveness (AHR) or
17 bronchial hyperreactivity, indicates that the airways are more reactive to bronchoconstriction
18 induced by a variety of stimuli (e.g., specific allergens, exercise, SO2, cold air) than they would
19 be when normoreactive. In order to determine the level of airway responsiveness, airway
20 function (usually assessed by spirometry or plethysmography) is measured after the inhalation of
21 small amounts of an aerosolized specific (e.g., antigen, allergen) or nonspecific (e.g.,
22 methacholine, histamine) bronchoconstrictor agent or measured stimulus (e.g., exercise, cold
23 air). The dose or concentration of the agent or stimulus is increased from a control, baseline
24 level (placebo) until a predetermined degree of airway response, such as a 20% drop in FEVj or
25 a 100% increase in Raw, has occurred (Cropp et al., 1980; Sterk et al., 1993). The dose or
26 concentration of the bronchoconstrictor agent that produced the increased responsiveness often is
27 referred to as the "PD^FEVj" or "PC^FEVj" (i.e., the provocative dose or concentration that
28 produced a 20% drop in FEVj) or the "PD100SRaw" (i.e., the provocative dose that produced a
29 100% increase in SRaw). A high level of bronchial responsiveness is a hallmark of asthma. The
30 range of nonspecific bronchial responsiveness, as expressed by the PD20 for example, is at least
31 1,000-fold from the most sensitive asthmatics to the least sensitive healthy subjects.
January 2005 AX6-82 DRAFT-DO NOT QUOTE OR CITE
-------
1 Unfortunately, it is difficult to compare the PD2QFEVJ or PD100SRaw across studies because of
2 the many different ways of presenting dose response to bronchoconstrictor drugs, for example,
3 by mg/mL, units/mL, and molar solution; or by cumulative dose (CIU or CBU) and doubling
4 dose (DD). Other typical bronchial challenge tests with nonspecific bronchoconstrictor stimuli
5 are based on exercise intensity or temperature of inhaled cold air.
6 Increases in nonspecific airway responsiveness were previously reported as an important
7 consequence of exposure to O3 (e.g., Golden et al., 1978; Table AX6-11). Konig et al. (1980)
8 and Holtzman et al. (1979) found the increased airway responsiveness after O3 exposure in
9 healthy subjects appeared to be resolved after 24 h. Because atopic subjects had similar
10 increases in responsiveness to histamine after O3 exposure as non-atopic subjects, Holtzman
11 et al. (1979) concluded that the increased nonspecific bronchial responsiveness after O3 exposure
12 was not related to atopy. Folinsbee and Hazucha (1989) showed increased airway
13 responsiveness in 18 female subjects 1 and 18 h after exposure to 0.35 ppm O3. Taken together,
14 these studies suggest that O3-induced increases in airway responsiveness usually resolve 18 to
15 24 h after exposure, but may persist in some individuals for longer periods.
16 Gong et al. (1986) found increased nonspecific airway responsiveness in elite cyclists
17 exercising at competitive levels with O3 concentrations as low as 0.12 ppm. Folinsbee et al.
18 (1988) found an approximate doubling of the mean methacholine responsiveness in a group of
19 healthy volunteers exposed for 6.6 h to 0.12 ppm O3. Horstman et al. (1990) demonstrated
20 significant decreases in the PD100SRaw in 22 healthy subjects immediately after a 6.6-h exposure
21 to concentrations of O3 as low as 0.08 ppm. No relationship was found between O3-induced
22 changes in airway responsiveness and changes in FVC or FEVj (Folinsbee et al., 1988; Aris
23 et al., 1995), suggesting that changes in airway responsiveness and spirometric volumes occur by
24 different mechanisms.
25 Dimeo et al. (1981) were the first to investigate attenuation of the O3-induced increases in
26 nonspecific airway responsiveness after repeated O3 exposure. Over 3 days of a 2 h/day
27 exposure to 0.40 ppm O3, they found progressive attenuation of the increases in airway
28 responsiveness such that, after the third day of O3 exposure, histamine airway responsiveness
29 was no longer different from the sham exposure levels. Kulle et al. (1982) found that there was
30 a significantly enhanced response to methacholine after the first 3 days of exposure, but this
31 response slowly normalized by the end of the fifth day. Folinsbee et al. (1994) found a more
January 2005 AX6-83 DRAFT-DO NOT QUOTE OR CITE
-------
TABLE AX6-11. Airway Responsiveness Following Ozone Exposures"
to
o
o
Ozone
Concentrationb
ppm
ug/m3
Exposure Duration Exposure Number and Subject
and Activity Conditions Gender of Subjects Characteristics
Observed Effect(s)
Reference
0.125
0.250
0.125
245 3hIE
490 (10 min rest, 15 min
exercise on bicycle)
VE = 30 L/min
245 3hIEx4days
27 °C 5 F, 6 M Mild bronchial
50 % RH 20-53 years old asthma
6F, 16 M
1 9-48 years old Allergic rhinitis
Mean early-phase FEV; response and number
of > 20% reductions in FEV! were
significantly greater after 0.25 ppm O3 or
4 x 0.125 ppm O3. Most of the > 15% late-
phase FEV; responses occurred after
exposure to 4 x 0.125 ppm O3, as well as
significant inflammatory effects, as indicated
by increased sputum eosinophils (asthma and
allergic rhinitis) and increased sputum
lymphocytes, mast cell tryptase, histamine,
and LDH (asthma only).
Holz et al.
(2002)
X
ON
oo
H
6
o
o
H
O
o
HH
H
W
0.4
0.12
0.2
0.4
0.16
784 2 h IE
VE = 20 L/min/m2
BSA
235 45 min IE
1 5 min exercise
VE = 3 x resting
392 4 h IE
VE = 25 L/min/m2
BSA
784 2 h IE
40 min/h @ 50 W
314 7.6 h IE
50 min/h
VE s25 1/min
NA
60% RH for
test
atmospheres
20° C
62% RH
NA
22°C
40% RH
6F
1M
19-26 years old
12 F
5M
19-3 8 years old
4F
8M
23-47 years old
15 healthy subjects ;
9F,6M;31.1±2.1
years old
5F
4M
Stable mild asthma;
no meds 8 h
preexposure
Physician diagnosed
asthma; SO2-induced
airway
hyperreactivity
Healthy nonsmokers
Healthy; nonatopic
Mild atopic
asthma, HMD
sensitive, 20-35 years
old
Increased bronchial responsiveness to
methacholine 16 h after exposure; inhaled
apocynin treatment significantly reduced O3-
induced airway responsiveness
Dietary supplementation with 400 IU Vit E +
500 mg Vit C reduced airway responsiveness
to 0.10 and 0.25 ppm SO2 challenge
Increased sputum total cells, % neurtophils,
IL-6, and IL-8 at 18 h after exposure;
increased airway responsiveness to
methacholine 2 h after postexposure FEV;
returned to 5% of base-line; no anti-
inflammatory effect of azithromycin
Decreased FEV; and FVC; increased
bronchial reactivity to methacholine
4 h post-exposure; no protection from
inhaled corticosteroid, budesonide
Mean 9.1% FEV; decrease 18 h after O3
exposure; provocative dose of dust mite
allergen decreased from 10.3 to 9.7 dose
units.
Peters et al.
(2001)
Trenga et al.
(2001)
Criqui et al.
(2000)
Nightingale
et al. (2000)
Kehrl et al.
(1999)
-------
TABLE AX6-11 (cont'd). Airway Responsiveness Following Ozone Exposures3
to
o
o
X
ON
oo
H
6
o
o
H
O
o
HH
H
W
Ozone
Concentration1"
ppm
0.2
0.12
Air-antigen
0.4
0.2
0.4
0.12
0.25
ug/m3
392
235
784
392
784
236
490
Exposure Duration
and Activity
4hIE
40 min/h @ 50 W
1 h rest
2hIE
VE= 20 L/min/m2
BSA
4hIE
50 min/h
VE = 25 L/min/m2
BSA
3 h/d for 5 days;
alternating 15 min of rest
and exercise at
VE = 32 L/min
Rest
3hIE
VE = 30 L/min
1 5 min ex/
10 min rest/
5 min no O3; every
30 min.
Exposure
Conditions
NA
NA
NA
20° C
50% RH
31°C
35% RH
22 °C
40% RH
27 °C
54% RH
mouthpiece
exposure
Number and
Gender of Subjects
1 0 asthmatic (6 F, 4
M),
26.6 ± 2.3 years old;
1 0 healthy (4 F, 6
M),
27.3 ±1.4 years old.
6F
9M
5F
1M
18-27 years old
6F
12 M
18-36 years old
2F
8M
19-48 years old
5F
10M
24 mild asthmatics
11F/13M
12 allergic rhinitics
6M/6F
Subject
Characteristics
Mild atopic asthma;
non-atopic healthy
subjects; no meds
8 weeks pre-exposure
Mild allergic asthma;
1 8 to 49 years if age
Stable mild asthma;
no meds 8 h
preexposure
Phy sician-diagno sed
mild asthma; no meds
prior to exposure
Mild asthma
requiring only
occasional
bronchodilator
therapy
atopic
asthma
atopic mild asthmatic
NS
Observed Effect(s)
Decreased FEV; in asthmatic (9.3%)
and healthy (6.7%) subjects; increased
sputum neutrophils in both groups (NS);
no change in methacholine airway reactivity
24 h post-exposure
No effect of O3 on airway response to grass
or ragweed allergen.
Increased airway responsiveness to
methacholine 16 h postexposure; no effect
of proteinase inhibitor, rALP
Decreased FEV; and FVC, increased SRaw;
lower respiratory Sx; increased
% neutrophils, total protein, LDH,
fibronectin, IL-8, GM-CSF, and MPO
in BAL. Correlation between pre-exposure
methacholine challenge and O3-induced
SRaw increase.
Significant FEV[ and Sx response on 1 st and
2nd O3 exposure days, then diminishing with
continued exposure; tolerance partially lost
4 and 7 days postexposure; bronchial
reactivity to methacholine peaked on 1 st O3
exposure day, but remained elevated with
continued exposure
No effect of O3 on airway response to grass
allergen.
Increased allergen responsiveness afer O3
exposure.
Reference
Nightingale
etal. (1999)
Hanania
etal. (1998)
Hiltermann
etal.
(1998)
Balmes et al.
(1997);
Scannell
etal. (1996)
Gong et al.
(1997b)
Ball et al.
(1996)
Jorres et al.
(1996)
-------
TABLE AX6-11 (cont'd). Airway Responsiveness Following Ozone Exposures3
X
Oi
oo
Oi
H
6
o
o
H
O
Ozone
Concentrationb
ppm
0.2
0.12
0.12
0.10
0.25
0.40
Air-antigen
0.12 ppm
O3-antigen
0.08
0.10
0.12
ug/m3
392
235
235
196
490
785
157
196
235
Exposure Duration
and Activity
4hIE
50min/10min
exercise/rest each hour
IhR
6.6 h, IE x 5 days
50 min exercise/10 min
rest, 30 min lunch
VE=38.8L/min
1 h light IE
2x15 min on treadmill
VE = 27 L/min
1 h at rest
6.6 h
IE at « 39 L/min
Exposure
Conditions
22 °C
50% RH
Ambient T
&RHfor
exposure;
23°C&
50% RH for
exercise
challenge
18 °C
40% RH
21 °C
40% RH
NA
18 °C
40% RH
Number and Subject
Gender of Subj ects Characteristics
42M/24F 18-50 years
NS healthy
8 F Mild stable asthma
7M
19-45 years old
17 M Healthy nonsmokers
25 ± 4 years old
9 F Stable mild
12 M asthmatics with FEV;
1 9-40 years old > 70% and
methacholine
responsiveness
4 M, 3 F Asthmatic,
21 to 64 years old
22 M Healthy NS,
18 to 32 years old
Observed Effect(s)
FEV^- 18.6%), FVC (- 14.6%), decreased
after O3. Baseline PC100 for methacholine
was not related to changes in FVC, FEVl5
a weak association was seen for PC100 and
increased SRaw.
No significant difference in % fall FEV[ or
V40p; no increase in bronchial responsiveness
to exercise challenge
FEV[ responses were maximal on 1st day of
exposure (-13%), less on second day (-9%),
absent thereafter. Sx responses only the first
2 days. Methacholine airway responsiveness
was at least doubled on all exposure days,
but was highest on the second day of O3.
No significant differences in FEV[ or FVC
were observed for 0.10 and 0.25 ppm O3-FA
exposures or postexposure exercise challenge;
12 subjects exposed to 0.40 ppm O3 showed
significant reduction in FEV[.
Increased bronchoconstrictor response to
inhaled ragweed or grass after O3 exposure
compared to air.
33, 47, and 55% decreases in cumulative dose
of methacholine required to produce a 100%
increase in SRaw after exposure to O3 at 0.08,
0.10, and 0.12 ppm, respectively.
Reference
Aris et al.
(1995)
Fernandes
etal. (1994)
Folinsbee
etal. (1994)
Weymer
etal.
(1994)
Molfino
etal.
(1991)
Horstman
etal. (1990)
o
HH
H
W
-------
3
to
o
o
TABLE AX6-11 (cont'd). Airway Responsiveness Following Ozone Exposures'1
X
Oi
oo
H
6
O
o
H
O
Ozone
Concentrationb
ppm Mg/m3
0.12 ppm
CylOOppb
S02
0.12 ppm
O3-0.12ppm
03
Air-100ppbSO2
0.35 686
0.40 784
0.12 235
0.12 235
0.20 392
0.40 784
0.20 392
0.40 784
0.40 784
— Exposure Duration Exposure
and Activity Conditions
45 min in first atmosphere 22 °C
and 1 5 min in second IE 75% RH
70 min with IE at NA
40 L/min
2 h with IE at 22 °C
VE = 53 to 55 L/min 50% RH
6. 6 h with IE at NA
«25 L/min/m2 BSA
lhat VE= 89 L/min 31 °C
followed by 3 to 4 min 35% RH
at -150 L/min
3 h/day for 5 days in a
row
2 h with IE at 2 x resting 22 °C
2 h with IE at 2 x resting 55% RH
2 h/day for 3 days
Number and Subject
Gender of Subj ects Characteristics
8 M, 5 F Asthmatic,
12 to 18 years old
18 F Healthy NS,
19 to 28 years old
8M, 10 F 9 asthmatics (5 F,
4M),
9 healthy (5 F, 4 M),
18 to 34 years old
10 M Healthy NS,
18 to 33 years old
15M,2F Elite cyclists,
19 to 30 years old
13M, 11 F Healthy NS,
1 9 to 46 years old
12 M, 7 F Healthy NS,
21 to 32 years old
Observed Effect(s)
Greater declines in FEV[ and Vmaj,50o/0 and
greater increase in respiratory resistance after
O3-SO2 than after O3-O3 or air-SO2.
PD100 decreased from 59 CIU after air
exposure to 41 CIU and 45 CIU, 1 and
18 h after O3 exposure, respectively.
Decreased PC100SRsw from 33 mg/mL to
8.5 mg/mL in healthy subjects after O3.
PC100SRllw fell from 0.52 mg/mL to
0.19 mg/mL in asthmatic subjects after
exposure to O3 and from 0.48 mg/mL
to 0.27 mg/mL after exposure to air.
Approximate doubling of mean
methacholine responsiveness after exposure.
On an individual basis, no relationship
between O3-induced changes in airway
responsiveness and FEV; or FVC.
Greater than 20% increase in histamine
responsiveness in one subject at 0.12 ppm
O3 and in nine subjects at 0.20 ppm O3.
Enhanced response to methacholine after
first 3 days, but this response normalized by
Day5.
110% increase in ASRaw to a 10-breath
histamine (1.6%) aerosol challenge after
exposure to O3 at 0.40 ppm, but no change
at 0.20 ppm. Progressive adaptation of this
effect over 3-day exposure.
Reference
Koenig et al.
(1990)
Folinsbee
and Hazucha
(1989)
Kreit et al.
(1989)
Folinsbee
etal. (1988)
Gong et al.
(1986)
Kulle et al.
(1982)
Dimeo et al.
(1981)
o
HH
H
W
-------
TABLE AX6-11 (cont'd). Airway Responsiveness Following Ozone Exposures3
1
to
o
o
1
oo
oo
Ozone
Concentration1"
ppm Mg/m3 and Activity Conditions Gender of Subjects Characteristics
0.10 196 2h NA 14 HealthNS,
0.32 627 24 ± 2 years old
1.00 1,960
0.60 1,176 2 h with IE at 2 x resting 22 °C 1 1 M, 5 F 9 atopic,
5 5 % RH 7 nonatopic ,
NS, 21 to 35 years
old
0.6 1,176 2 hat rest NA 5 M, 3 F Healthy NS,
22 to 30 years old
Observed Effect(s)
Increased airway responsiveness
to methacholine immediately after exposure
at the two highest concentrations of O3.
Ten-breath methacholine or histamine
challenge increased SRaw > 150% in
16 nonasthmatics after O3. On average, the
atopic subjects had greater responses than
the nonatopic subjects. The increased
responsiveness resolved after 24 h. Atropine
premedication blocked the O3-induced
increase in airway responsiveness.
300% increase in histamine-induced ARaw
5 min after O3 exposure; 84 and 50%
increases 24 h and 1 week after exposure
(p > 0.05), respectively. Two subjects had
an increased response to histamine 1 week
after exposure.
Reference
Konig et al.
(1980)
Holtzman
etal. (1979)
Golden et al.
(1978)
H
6
o
o
H
O
aSee Appendix A for abbreviations and acronyms.
bListed from lowest to highest O3 concentration.
o
HH
H
W
-------
1 persistent effect of O3 on airway responsiveness which was only partially attenuated after
2 5 consecutive days of O3 exposure.
3 The occurrence and duration of increased nonspecific airway responsiveness following O3
4 exposure could have important clinical implications for asthmatics. Kreit et al. (1989)
5 investigated changes in airway responsiveness to methacholine that occur after O3 exposure in
6 mild asthmatics. They found that the baseline PC100SRaw declined from 0.52 to 0.19 mg/mL
7 after a 2-h exposure to 0.40 ppm O3 as compared to a decline from 0.48 to 0.27 mg/mL after air
8 exposure; however, because of the large variability in responses of the asthmatics, the percent
9 decrease from baseline in mean PC100SRaw was not statistically different between healthy and
10 asthmatic subjects (74.2 and 63.5%, respectively).
11 Two studies examined the effects of preexposure to O3 on exacerbation of exercise-induced
12 bronchoconstriction (Fernandes et al., 1994; Weymer et al., 1994). Fernandes et al. (1994)
13 preexposed subjects with stable mild asthma and a history of > 15% decline in FEVj after
14 exercise to 0.12 ppm O3 for 1 h at rest followed by a 6-min exercise challenge test and found no
15 significant effect on either the magnitude or time course of exercise-induced
16 bronchoconstriction. Similarly, Weymer et al. (1994) observed that preexposure to either 0.10 or
17 0.25 ppm O3 for 60 min while performing light IE did not enhance or produce exercise-induced
18 bronchoconstriction in otherwise healthy adult subjects with stable mild asthma. Although the
19 results suggested that preexposure to O3 neither enhances nor produces exercise-induced asthma
20 in asthmatic subjects, the relatively low total inhaled doses of O3 used in these studies limit the
21 ability to draw any definitive conclusions.
22 Gong et al. (1997b) found that subjects with asthma developed tolerance to repeated
23 O3 exposures in a manner similar to normal subjects; however, there were more persistent effects
24 of O3 on airway responsiveness, which only partially attenuated when compared to filtered air
25 controls. Volunteer subjects with mild asthma requiring no more than bronchodilator therapy
26 were exposed to filtered air or 0.4 ppm O3, 3 h/d for 5 consecutive days, and follow-up
27 exposures 4 and 7 days later. Symptom and FEVj responses were large on the 1st and 2nd
28 exposure days, and diminished progressively toward filtered air responses by the 5th exposure
29 day. A methacholine challenge was performed when postexposure FEVj returned to within 10%
30 of preexposure baseline levels. The first O3 exposure significantly decreased PD^FEVj by an
31 order of magnitude and subsequent exposures resulted in smaller decreases, but they were still
January 2005 AX6-89 DRAFT-DO NOT QUOTE OR CITE
-------
1 significantly different from air control levels. Thus, the effects of consecutive O3 exposures on
2 bronchial reactivity differ somewhat from the effects on lung function. The same conclusion
3 was drawn by Folinsbee et al. (1994) after consecutive 5-day O3 exposures in healthy subjects,
4 despite a much lower bronchial reactivity both before and after O3 exposure.
5 A larger number of studies examined the effects of O3 on exacerbation of antigen-induced
6 asthma. Molfmo et al. (1991) were the first to report the effects of a 1-h resting exposure to
7 0.12 ppm O3 on the response of subjects with mild, stable atopic asthma to a ragweed or grass
8 allergen inhalation challenge. Allergen challenges were performed 24 h after air and
9 O3 exposure. Their findings suggested that allergen-specific airway responsiveness of mild
10 asthmatics is increased after O3 exposure. However, Ball et al. (1996) and Hanania et al. (1998)
11 were unable to confirm the findings of Molfmo et al. (1991) in a group of grass-sensitive mild
12 allergic asthmatics exposed to 0.12 ppm O3 for 1 h. The differences between Hanania et al.
13 (1998) and Molfmo et al. (1991), both conducted in the same laboratory, were due to better, less
14 variable control of the 1 h 0.12 ppm O3 exposure and better study design by Hanania and
15 colleagues. In the original, Molfmo et al. (1991) study, the control (air) and experimental (O3)
16 exposures were not randomized after the second subject because of long-lasting (3 months),
17 O3-induced potentiation of airway reactivity in that subject. For safety reasons, therefore, the air
18 exposures were performed prior to the O3 exposures for the remaining 5 of 7 subjects being
19 evaluated. It is possible that the first antigen challenge caused the significant increase in the
20 second (post-O3) antigen challenge.
21 Torres et al. (1996) later confirmed that higher O3 concentrations cause increased airway
22 reactivity to specific antigens in subjects with mild allergic asthma, and to a lesser extent in
23 subjects with allergic rhinitis, after exposure to 0.25 ppm O3 for 3 h. The same laboratory
24 repeated this study in separate groups of subjects with asthma and rhinitis and found similar
25 enhancement of allergen responsiveness after O3 exposure (Holz et al., 2002); however, the
26 effects of a 3-h exposure to 0.25 ppm O3 were more variable, most likely due to performing the
27 allergen challenges 20 h after exposure, rather than the 3 h used in the first study.
28 The timing of allergen challenges in O3-exposed subjects with allergic asthma is important.
29 Bronchial provocation with allergen, and subsequent binding with IgE antibodies on mast cells
30 in the lungs, triggers the release of histamine and leukotrienes and a prompt early-phase
31 contraction of the smooth muscle cells of the bronchi, causing a narrowing of the lumen of the
January 2005 AX6-90 DRAFT-DO NOT QUOTE OR CITE
-------
1 bronchi and a decrease in bronchial airflow (i.e., decreased FEVj). In many asthma patients,
2 however, the release of histamine and leukotrienes from the mast cells also attracts an
3 accumulation of inflammatory cells, especially eosinophils, followed by the production of mucus
4 and a late-phase decrease in bronchial airflow for 4 to 8 h.
5 A significant finding from the study by Holz et al. (2002) was that clinically relevant
6 decreases in FEVj (> 20%) occurred during the early-phase allergen response in subjects with
7 rhinitis after a consecutive 4-day exposure to 0.125 ppm O3. Kehrl et al. (1999) previously
8 found an increased reactivity to house dust mite antigen in asthmatics 16 to 18 h after exposure
9 to 0.16 ppm O3 for 7.6 hours. These important observations indicate that O3 not only causes
10 immediate increases in airway-antigen reactivity, but that this effect may persist for at least 18 to
11 20 h. Ozone exposure, therefore, may be a clinically important co-factor in the response to
12 airborne bronchoconstrictor substances in individuals with pre-existing allergic asthma. It is
13 plausible that this phenomenon could contribute to increased symptom exacerbations and, even,
14 consequent increased physician or ER visits, and possible hospital admissions (see Chapter 7).
15 A number of human studies, especially more recent ones, have been undertaken to
16 determine various aspects of O3-induced increases in nonspecific airway responsiveness, but
17 most studies have been conducted in laboratory animals (See the toxicology chapter, Section
18 5.3.4.4.). In humans, increased airway permeability (Kehrl et al., 1987; Molfmo et al., 1992)
19 could play a role in increased airway responsiveness. Inflammatory cells and mediators also
20 could affect changes in airway responsiveness. The results of a multiphase study (Scannell
21 et al., 1996; Balmes et al., 1997) showed a correlation between preexposure methacholine
22 responsiveness in healthy subjects and increased SRaw caused by a 4 h exposure to 0.2 ppm O3,
23 but not with O3-induced decreases in FEVj and FVC. The O3-induced increase in SRaw, in turn,
24 was correlated with O3-induced increases in neutrophils and total protein concentration in BAL
25 fluid. Subjects with asthma had a significantly greater inflammatory response to the same O3
26 exposures, but it was not correlated with increased SRaw, and nonspecific airway provocation
27 was not measured. Therefore, it is difficult to determine from this series of studies if underlying
28 airway inflammation plays a role in increased airway responsiveness to nonspecific
29 bronchoconstrictors. The study, however, confirmed an earlier observation (e.g., Balmes et al.,
30 1996) that O3-induced changes in airway inflammation and lung volume measurements are not
31 correlated.
January 2005 AX6-91 DRAFT-DO NOT QUOTE OR CITE
-------
1 Hiltermann et al. (1998) reported that neutrophil-derived serine proteinases associated with
2 O3-induced inflammation are not important mediators for O3-induced nonspecific airway
3 hyperresponsiveness. Subjects with mild asthma, prescreened for O3-induced airway
4 responsiveness to methacholine, were administered an aerosol of recombinant antileukoprotease
5 (rALP) or placebo at hourly intervals two times before and six times after exposure to filtered air
6 or 0.4 ppm O3 for 2 h. Methacholine challenges were performed 16 h after exposure. Treatment
7 with rALP had no effect on the O3-induced decrease in FEVj or PC^FEVj in response to
8 methacholine challenge. The authors speculated that proteinase-mediated tissue injury caused
9 by O3 may not be important in the development of airway hyperresponsiveness of asthmatics to
10 O3. In a subsequent study using a similar protocol (Peters et al., 2001), subjects with mild
11 asthma were administered an aerosol of apocynin, an inhibitor of NADPH oxidase present in
12 inflammatory cells such as eosinophils and neutrophils, or a placebo. In this study, methacholine
13 challenge performed 16 h after O3 exposure showed treatment-related effects on PC2QFEVJ,
14 without an effect on FEVj. The authors concluded that apocynin could prevent O3-induced
15 bronchial hyperresponsiveness in subjects with asthma, possibly by preventing superoxide
16 formation by eosinophils and neutrophils in the larger airways.
17 Nightingale et al. (1999) reported that exposures of healthy subjects and subjects with mild
18 atopic asthma to a lower O3 concentration (0.2 ppm) for 4 h caused a similar neutrophilic lung
19 inflammation in both groups but no changes in airway responsiveness to methacholine measured
20 24 h after O3 exposure in either group. There were, however, significant decreases in FEVj of
21 6.7 and 9.3% immediately after O3 exposure in both healthy and asthmatic subjects, respectively.
22 In a subsequent study, a significant increase in bronchoresponsiveness to methacholine was
23 reported 4 h after healthy subjects were exposed to 0.4 ppm O3 for 2 h (Nightingale et al., 2000).
24 In the latter study, preexposure treatment with inhaled budesonide (a corticosteroid) did not
25 protect against O3-induced effects on spirometry, methacholine challenge, or sputum neutrophils.
26 These studies also confirm the earlier reported findings that O3-induced increases in airway
27 responsiveness usually resolve by 24 h after exposure.
28 Ozone-induced airway inflammation and hyperresponsiveness were used by Criqui et al.
29 (2000) to evaluate anti-inflammatory properties of the macrolide antibiotic, azithromycin. In a
30 double-blind, cross-over study, healthy volunteers were exposed to 0.2 ppm O3 for 4 h after
31 pretreatment with azithromycin or a placebo. Sputum induction 18 h postexposure resulted in
January 2005 AX6-92 DRAFT-DO NOT QUOTE OR CITE
-------
1 significantly increased total cells, percent neutrophils, IL-6, and IL-8 in both azithromycin- and
2 placebo-treated subjects. Significant pre- to post-exposure decreases in FEVj and FVC also
3 were found in both subject groups. Airway responsiveness to methacholine was not significantly
4 different between azithromycin-treated and placebo-treated subjects when they were challenged
5 2 h after postexposure FEVj decrements returned to within 5 % of baseline. Thus, azithromycin
6 did not have anti-inflammatory effects in this study.
7 The effects of dietary antioxidants on O3-induced bronchial hyperresponsiveness were
8 evaluated in adult subjects with asthma by Trenga et al. (2001). Recruited subjects were
9 pretested for responsiveness to a provocative SO2 challenge (0.10 and 0.25 ppm) while
10 exercising on a treadmill and selected for study if they experienced a > 8% decrease in FVC.
11 The rationale for this environmental challenge approach is based on previously published work
12 by this laboratory (Koenig et al., 1990). In a placebo-controlled, double-blind crossover study,
13 subjects took two vitamin supplements (400 IU vitamin E and 500 mg vitamin C) or two
14 placebos once a day for 4 weeks, and were exposed to filtered air and to 0.12 ppm O3 for 45 min
15 during intermittent exercise at three times resting ventilation. Blood samples were used to verify
16 placebo and vitamin treatment levels. Provocative airway challenges with SO2 were performed
17 immediately after O3 exposure. Ozone exposure potentiated the SO2 challenge in asthmatics,
18 and subjects given antioxidant supplementation responded less severely to the airway challenge
19 than subjects given the placebo. The protective effect of antioxidants was even more
20 pronounced among subjects with more severe asthma and higher sensitivity to SO2.
21
22
23 AX6.9 EFFECTS ON INFLAMMATION AND HOST DEFENSE
24 AX6.9.1 Introduction
25 In general, inflammation can be considered as the host response to injury, and the
26 induction of inflammation can be accepted as evidence that injury has occurred. Several
27 outcomes are possible: (1) inflammation can resolve entirely; (2) continued acute inflammation
28 can evolve into a chronic inflammatory state; (3) continued inflammation can alter the structure
29 or function of other pulmonary tissue, leading to diseases such as fibrosis or emphysema;
30 (4) inflammation can alter the body's host defense response to inhaled microorganisms; and
31 (5) inflammation can alter the lung's response to other agents such as allergens or toxins.
January 2005 AX6-93 DRAFT-DO NOT QUOTE OR CITE
-------
1 At present, it is known that short-term exposure of humans to O3 can cause acute inflammation
2 and that long-term exposure of laboratory animals results in a chronic inflammatory state (see
3 Chapter 5). However, the relationship between repetitive bouts of acute inflammation in humans
4 caused by O3 and the development of chronic respiratory disease is unknown.
5 Bronchoalveolar lavage (BAL) using fiberoptic bronchoscopy has been utilized to sample
6 cells and fluids lining the respiratory tract primarily from the alveolar region, although the use of
7 small volume lavages or balloon catheters permits sampling of the airways. Cells and fluid can
8 be retrieved from the nasal passages using nasal lavage (NL) and brush or scrape biopsy.
9 Several studies have analyzed BAL and NL fluid and cells from O3-exposed humans for
10 markers of inflammation and lung damage (see Tables AX6-12 and AX6-13). The presence of
11 neutrophils (PMNs) in the lung has long been accepted as a hallmark of inflammation and is an
12 important indicator that O3 causes inflammation in the lungs. It is apparent, however, that
13 inflammation within airway tissues may persist beyond the point that inflammatory cells are
14 found in BAL fluid. Soluble mediators of inflammation such as the cytokines (IL-6, IL-8) and
15 arachidonic acid metabolites (e.g., PGE2, PGF2a, thromboxane, and leukotrienes [LTs] such as
16 LTB4) have been measured in the BAL fluid of humans exposed to O3. In addition to their role
17 in inflammation, many of these compounds have bronchoconstrictive properties and may be
18 involved in increased airway responsiveness following O3 exposure.
19 Some recent evidence suggests that changes in small airways function may provide a
20 sensitive indicator of O3 exposure and effect (see Section AX6.2.5), despite the fact that inherent
21 variability in their measurement by standard spirometric approaches make their assessment
22 difficult. Observations of increased functional responsiveness of these areas relative to the more
23 central airways, and of persistent effects following repeated exposure, may indicate that further
24 investigation of inflammatory processes in these regions is warranted.
25 Under normal circumstances, the epithelia lining the large and small airways develop tight
26 junctions and restrict the penetration of exogenous particles and macromolecules from the
27 airway lumen into the interstitium and blood, as well as restrict the flow of plasma components
28 into the airway lumen. O3 disrupts the integrity of the epithelial cell barrier in human airways, as
29 measured by markers of plasma influx such as albumin, immunoglobulin, and other proteins into
30 the airways. Markers of epithelial cell damage such as lactate dehydrogenase (LDH) also have
31 been measured in the BAL fluid of humans exposed to O3. Other soluble factors that have been
January 2005 AX6-94 DRAFT-DO NOT QUOTE OR CITE
-------
Table AX6-12. Studies of Respiratory Tract Inflammatory Effects from Controlled Human Exposure to Ozone"
to
o
o
X
Ozone Concentration11
ppm |ig/m3 Duration
Activity Level
(VE)
Number and Gender
of Subjects
Observed Effect(s)
Reference
Upper Airway Studies
0.4 784 2 h
0.2 392 2h
0.4 980 2h
0.12 235 1.5 h
0.24 470
0.5 980 4h
0.4 784 2 h
0.5 980 4hon
2 consecutive days
At rest
IE
(15 min/30 min);
(VE) « 20
L/min/m2 BSA
At rest
IE
(20 L/min) at
15 -min intervals
Resting
IE
(70 L/min) at
15 -min intervals
Resting
12 mild,
asymptomatic dust
mite-sensitive
asthmatics;
18-35 years of age
8 M, 5 F healthy NS
20-3 1 years of age
10 mild NS
asthmatics
18-3 5 years old
5 M, 5 F, asthmatic;
4 M, 4 F,
nonasthmatic;
18 to 41 years old
6 M, 6 F,
allergic rhinitics,
31.4 ± 2.0 (SD)
years old
11 M,
18 to 35 years old
41 M
(21 O3-exposed,
20 air-exposed),
18 to 35 years old
Release of early-onset mast cell-derived mediators into NL in
response to allergen not enhanced following O3 exposure.
Neutrophil and eosinophil inflammatory mediators were not
increased after O3 exposure or enhanced after allergen
challenge. O3 increased eosinophil influx following allergen
exposure.
No neutrophilia in NL samples by 1 .5 h post exposure.
Depletion of uric acid in NL fluid by 30% during h 2 of
exposure with increase in plasma uric acid levels.
No depletion of ascorbic acid, reduced glutathione,
extracellular superoxide dismutase.
Response to allergen increased (NS). PMN and
eosinophils increased after O3 plus allergen challenge.
Ozone alone increased inflammation in the nose.
NL done immediately and 24 h after exposure.
Increased number of PMNs at both times in asthmatic subjects
exposed to 0.24 ppm O3; no change in nonasthmatic subjects.
No change in lung or nasal function.
NL done immediately after exposure. Increased upper
and lower respiratory symptoms and increased levels of PMNs,
eosinophils, and albumin in NL fluid.
NL done immediately before, immediately after, and 22 h after
exposure. Increased numbers of PMNs at both times after
exposure; increased levels of tryptase, a marker of mast cell
degranulation, immediately after exposure; increased levels
of albumin 22 h after exposure.
NL done immediately before and after each exposure
and 22 h after the second exposure. Increased levels of PMNs
at all times after the first exposure, with peak values occurring
immediately prior to the second exposure.
Michelson
etal. (1999)
Mudway et al.
(1999)
Peden et al.
(1995)
McBride et al.
(1994)
Bascom et al.
(1990)
Graham and
Koren(1990)
Koren et al.
(1990)
Graham et al.
(1988)
-------
Table AX6-12 (cont'd). Studies of Respiratory Tract Inflammatory Effects from Controlled Human Exposure to Ozonea
to
o
o
X
Oi
Ozone Concentrationb
ppm ug/m3 Duration (VE)
Number and Gender
of Subjects
Observed Effect(s)
Reference
Lower Airway Studies
0.2 392 2h IE
(15 min/30 min);
(VE) - 20
L/min/m2 BSA
0.1 196 2h mild IE
0.2 392 2h IE
(15 min/30 min);
(VE) - 20
L/min/m2 body
surface area
0.27 529 2h IE
(20 min/ 60 min);
(VE) - 25
L/min/m2 BSA
0.22 431 4 h IE
(15 min/30 min);
(VE) = 25
L/min/m2 BSA
6M, 6F healthy,
nonatopic and
9 M, 6F mild
asthmatic subjects,
19-48 years of age
12 M, 10 F
healthy subjects
mean age -30
years
6M, 9F healthy
subjects and
9 M, 6F mild
asthmatics
12 subjects with
intermittent-mild
asthma exhibiting a
dual response; 18-37
years of age
12 nonsmoker,
nonresponders;
1 3 nonsmoker,
responders;
13 smokers;
1 8-40 years of age
Significantly higher baseline expression of IL-4 and IL-5 in
bronchial mucosal biopsies from asthmatic vs. healthy subjects
6 h post-exposure. Following O3 exposure, epithelial
expression of IL-5, GM-CSF, ENA-78, and IL-8 increased
significantly in asthmatics, as compared to healthy subjects.
Markers of exposure in exhaled breath condensate, including
increased 8-isoprostane, TEARS and LTB-4, and a marker of
ROS-DNA interaction in peripheral blood leukocytes
(8-OHdG), were increased in a sub-set of subjects bearing the
wild genotype for NAD(P)H:quinone oxidoreductase and the
null genotype for glutathione-S-transferase Ml .
No evidence seen for increased responsiveness to the
inflammatory effects of O3 in mild asthmatics versus healthy
subjects at 6 h following exposure. Used neutrophil
recruitment and exacerbation of pre-existing inflammation.
Exposure to O3 24 h following allergen challenge resulted in a
significant decrease in FEV1, FVC and VC and increase in
symptom scores compared to air exposure. The percentage of
eosinophils, but not neutrophils, in induced sputum was higher
6 h after O3 than after air.
Recovery of AM was approximately 3-fold higher in BAL
from smokers versus nonsmokers. Unstimulated AM from
smokers released ~2-fold greater amounts of superoxide anion
than from nonsmokers at 30 min and 18 h post-exposure, but
release was not further enhanced by stimulation of the cells.
ROS generation by AM from non-smokers decreased following
exposure at 18 h; markers of epithelial permeability increased.
No relationship was found between measures of ROS
production and lung function responsiveness to O3.
Bosson et al.
(2003)
Corradi et al.
(2002)
Stenfors et al.
(2002)
Vagaggini
et al. (2002)
Voter et al.
(2001)
-------
Table AX6-12 (cont'd). Studies of Respiratory Tract Inflammatory Effects from Controlled Human Exposure to Ozonea
to
o
o
X
Ozone Concentration11
ppm ug/m3 Duration
Activity Level
(VE)
Number and Gender
of Subjects
Observed Effect(s)
Reference
Lower Airway Studies (cont'd)
0.2 392 2h
0.4 784 2 h
0.4 784 2 h
0.4 784 1 h
0.2 392 2h
0.4 784 2 h/day for 5 days,
2 h either 10 or
20 days later
IE
(15 min/30 min);
(VE) - 20
L/min/m2 BSA
IE
(15 min/30 min);
(VE) - 20
L/min/m2 BSA
At rest
Continuous
exercise;
(VE) - 30
L/min/m2 BSA
IE
(40 L/min) at
15 -min intervals
8M, 5F healthy
nonsmokers;
20-3 1 years of age
10M, 6F subjects
with intermittent
asthma;
19-35 years of age
12 mild,
asymptomatic dust
mite-sensitive
asthmatics;
18-35 years of age
4 healthy subjects
15 healthy
nonsmokers
16 M; 18 to 35 years
of age
Early (1 .5 h post-exposure) increase in adhesion molecule
expression, submucosal mast cell numbers and alterations in
lining fluid redox status. No clear relationship between early
markers of response and lung function deficits. 2.5-fold
increase in % human leukocyte antigen (HLA)-DR+ alveolar
macrophages in BAL.
In a cross-over study, levels of eosinophil cationic protein, IL-8
and percentage eosinophils were found to be highly correlated
in induced sputum and BAL 16 h following O3 exposure.
Release of early-onset mast cell-derived mediators into NL in
response to allergen not enhanced following O3 exposure.
Neutrophil and eosinophil inflammatory mediators were not
increased after O3 exposure or enhanced after allergen
challenge. O3 increased eosinophil influx following allergen
exposure.
Apoptosis was observed in cells obtained by airway lavage 6 h
following exposure. AM obtained by BAL showed the
presence of a 4-hydroxynonenal (HNE) protein adduct and the
stress proteins, 72-kD heat shock protein and ferritin. These
effects were replicated by in vitro exposure of AM to HNE.
Increased numbers of CD3+, CD4+, and CD8+ T lymphocyte
subsets, in addition to neutrophils, in BAL 6 h post-exposure.
BAL done immediately after fifth day of exposure and again
after exposure 10 or 20 days later. Most markers of
inflammation (PMNs, IL-6, PGE2, LDH, elastase, fibronectin)
showed complete attenuation; markers of damage did not.
Reversal of attenuation was not complete for some markers,
even after 20 days.
Blomberg
etal. (1999)
Hiltermann
etal. (1999)
Michelson
etal. (1999)
Hamilton et al.
(1998)
Blomberg
etal. (1997)
Devlin et al.
(1997)
-------
Table AX6-12 (cont'd). Studies of Respiratory Tract Inflammatory Effects from Controlled Human Exposure to Ozone"
to
o
o
X
oo
Ozone Concentrationb
ppm |ig/m3
Exposure
Duration
Activity Level
(VE)
Number and Gender
of Subjects
Observed Effect(s)
Reference
Lower Airway Studies (cont'd)
0.22 431
0.12 235
0.16 314
0.2 392
0.4 784
0.4 784
0.2 392
4h
2h
7.6 h
4h
T = 20 °C
RH = 50%
2h
T = 22 °C
RH = 50%
2h
1 5 min,
ex/ 15 min, rest
4h
T = 20 °C
RH = 50%
IE
20 min ex/1 9 min
rest
(VE) * 39-45 1/min
IE
(15 min/30 min);
(VE) - 20
L/min/m2 BSA
IE
50 min/hr
(VE) = 25L/min
IE
(50 min/60 min);
(VE)«44L/min
1 5 min rest
1 5 min exercise
cycle ergometer
(VE) « 55 1/min
(VE) = 66 1/min
IE
(50 min/60 min);
(VE)«44L/min
31M,7F
smokers and
nonsmokers
9M, 3F healthy
nonsmokers; mean
age ~ 28 years
8 asthmatics sensitive
to dust mites
14 M, 6 F
healthy NS
1 1 healthy
nonsmokers;
18-3 5 years
8M
healthy nonsmokers
17 M, 6F
mild asthmatics
Post O3 exposure FEV; in 3 groups: Smokers (-13.9%);
non-responders (- 1 .4%) and responders (-28.5%). PMN's
increased immediately and at 18 h in all groups. Eosinophils
and lymphocytes increased after O3. IL-6 increased more in
non-smokers. No relationship of symptoms with inflammation,
lung function changes not related to inflammation. Nasal
lavage indicators did not predict bronchial or alveolar
inflammation.
Increase in the percentage of vessels expressing P-selectin in
bronchial biopsies at 1 .5 h post-exposure. No changes in
FEVj, FVC, inflammatory cells or markers in BAL, or vessels
expressing VCAM-1, E-selectin or ICAM-1 in biopsies.
Increased numbers of eosinophils in BAL after O3 exposure.
Ozone increased PMN, protein, IL-8, for all subjects.
No relationship of inflammation with spirometric responses.
Mean FEVl5 change = - 10%. BAL occurred at 0, 2, or 4 h
post-exposure. Small n limits statistical inference. Trend for
PMN's to be highest at 4 h. LTC4 increased at all time points.
No change in PGE2 or thromboxane.
Comparison of BAL at 1 h post-exposure vs. 18 h post-
exposure. At 1 h, PMN's, total protein, LDH, al-antitrypsin,
fibronectin, PGE2, thromboxane B2, C3a, tissue factor, and
clotting factor VII were increased. IL-6 and PGE2 were higher
after 1 h than 18 h. Fibronectin and tissue plasminogen
activator higher after 1 8 h. No time differences for PMN and
protein.
Increased PMN, protein, IL-8, LDH, in BAL. Inflammatory
responses were greater than a group of non-asthmatics
(Balmesetal., 1996)
Frampton
etal. (1997a)
Torres et al.
(1997)
Krishna et al.
(1997b)
Peden et al.
(1997)
Balmes et al.
(1996)
Coffey et al.
(1996)
Devlin et al.
(1996)
(compare with
Koren et al.
(1989a)
Scannell et al.
(1996)
-------
Table AX6-12 (cont'd). Studies of Respiratory Tract Inflammatory Effects from Controlled Human Exposure to Ozone"
to
o
o
X
Ozone Concentrationb
ppm ug/m3
Exposure
Duration
Activity Level
(VE)
Number and Gender
of Subjects
Observed Effect(s)
Reference
Lower Airway Studies (cont'd)
0.4 784
0.2 392
0.08 157
0.10 196
0.4 784
0.3 588
0.40 784
0.40 784
0.4 784
2 h mouthpiece
exposure
20 °C
42% RH
4h
6.6 h
2h
Ih
(mouth-piece)
2h
2h
2h
1 5 min exercise
1 5 min rest
(VE) « 40 1/min
IE
(50 min/60 min);
(VE) = 40 L/min
IE
(50 min/60 min) +
35 min lunch;
(VE) = 40 L/min
IE
(15 min/30min);
(VE) = 70 L/min
CE (60 L/min)
IE
(15 min/30 min);
(VE) = 70 L/min
IE
(15 min/30 min);
(VE) = 70 L/min
IE
(15 min/30 min);
(VE) = 70 L/min
5M, 5F
healthy; age ~ 30
15 M, 13 F,
21 to 39 years old
18 M,
18 to 35 years of age
10 M,
18 to 35 years old
5M
11M, 18 to
35 years old
11 M,
18 to 35 years old
11 M,
18 to 35 years old
Sputum induction 4 h after O3 exposure 3-fold increase in
neutrophils and a decrease in macrophages after O3 exposure.
IL-6, IL-8, and myeleperoxidase increased after O3. Possible
relationship of IL-8 and PMN levels.
Bronchial lavage, bronchial biopsies, and BAL done 1 8 h
after exposure. BAL shows changes similar to other studies.
Airway lavage shows increased cells, LDH, IL-8. Biopsies
show increased number of PMNs.
BAL fluid 18 h after exposure to 0.1 ppm O3 had significant
increases in PMNs, protein, PGE2, fibronectin, IL-6, lactate
dehydrogenase, and a-1 antitrypsin compared with the same
subjects exposed to FA. Similar but smaller increases in all
mediators after exposure to 0.08 ppm O3 except for protein and
fibronectin. Decreased phagocytosis of yeast by alveolar
macrophages was noted at both concentrations.
BAL fluid 1 h after exposure to 0.4 ppm O3 had significant
increases in PMNs, protein, PGE2, TXB2, IL-6, LDH, a-1
antitrypsin, and tissue factor compared with the same subjects
exposed to FA. Decreased phagocytosis of yeast by alveolar
macrophages.
Significantly elevated PMNs in the BAL fluid 1, 6, and 24 h
after exposure, with peak increases at 6 h.
Macrophages removed 18 h after exposure had changes in the
rate of synthesis of 123 different proteins as assayed by
computerized densitometry of two-dimensional gel protein
profiles.
BAL fluid 18 h after exposure contained increased levels of the
coagulation factors, tissue factor, and factor VII. Macrophages
in the BAL fluid had elevated tissue factor mRNA.
BAL fluid 18 h after exposure had significant increases in
PMNs, protein, albumin, IgG, PGE2, plasminogen activator,
elastase, complement C3a, and fibronectin.
Fahy et al.
(1995)
Aris et al.
(1993)
Devlin et al.
(1990,1991)
Koren et al.
(1991)
Koren et al.
(1991)
Schelegle
etal. (1991)
Devlin and
Koren (1990)
McGee et al.
(1990)
Koren et al.
(1989a,b)
-------
Table AX6-12 (cont'd). Studies of Respiratory Tract Inflammatory Effects from Controlled Human Exposure to Ozonea
to
o
o
X
O
O
Ozone Concentrationb
ppm ug/m3 Duration
Activity Level
(VE)
Number and Gender
of Subjects
Observed Effect(s)
Reference
Repeated Exposure Studies
0.125 245 3 h exposures to
0.25 490 both O3 cones, and
to FA;
3 h on four
consecutive days to
0.125; study arms
separated by
>4 wks
0.25 490 2 h on four
consecutive days;
O3 and FA exposure
study arms
separated by >3
wks
0.2 392 single,
4 h exposures to O3
and to FA;
4 h on four
consecutive days to
O3; study arms
separated by
>4 wks
0.2 392 single, 4 h
exposure;
4 h exposures on
four consecutive
days; study arms
separated by
>4 wks
IE
(15 min/30 min)
IE
(30 min/60 min);
(VE) ~ 8 times the
FVC/min
IE
(15 min/30 min);
(MeanVE) =
14.8 L/min/m2
BSA
IE
(30 min/60 min);
(MeanVE) = 25
L/min/m2 BSA
5M, 6F allergic
asthmatic and
16M, 6F allergic
rhinitic subjects;
19-53 years of age
5M, 3F healthy
subjects;
25-31 years of age
15M, 8F healthy
subjects;
21-35 years of age
9M, 6F healthy NS
23-37 years of age
All subjects underwent 4 exposure arms and were challenged
with allergen 20 h following the last exposure in each. Sputum
was induced 6-7 h later. In rhinitics, but not asthmatics, the
incidence and magnitude of early phase FEV; decrements to
Ag were greater after 0.25 and 4x 0. 125 ppm O3. Repeated
exposure caused increases in neutrophil and eosinophil
numbers in both subject groups, as well as increased
percentage and number of lymphocytes in the asthmatics.
Maximal mean reductions in FEV; and FVC were observed on
day 2, and became negligible by day 4. FEF25.75, VmaxSO, and
Vmax75 were combined into a single value representing small
airway function (SAWgrp). This variable was the only one to
show persistent depression of the 24 h post-exposure baseline
from day 2 to day 5 measurements. Numbers of PMNs in BAL
fluid on day 5 were significantly higher in subjects following
O3, compared to air, exposures.
All subjects underwent 3 exposure arms with BAL and
bronchial mucosal biopsies performed 20 h following the last
exposure in each. After repeated exposure, functional and
BAL cellular responses were not different from those after FA,
whereas total protein, IL-6, IL-8, reduced glutathione and
ortho-tyrosine remained elevated. Also at this time,
macroscopic scores of inflammation and tissue neutrophils
were increased in mucosal biopsies. IL-10 was detected only
in BAL fluid following repeated O3 exposure.
Subjects were randomly assigned to each of the exposure
regimens in a crossover design. Compared to single exposure,
repeated exposure resulted in an initial progression followed by
an attenuation of decrements in FEVj, FVC and specific
airways resistance by day 4. Bronchial and BAL washings
showed decreases in the numbers of PMNs and fibronectin
levels and IL-6 was decreased in BAL fluid on day 4.
Holz et al,
(2002)
Frank et al.
(2001)
Jorres et al.
(2000)
Christian et al.
(1998)
-------
to
o
o
Table AX6-12 (cont'd). Studies of Respiratory Tract Inflammatory Effects from Controlled Human Exposure to Ozonea
Ozone Concentration11
ppm |ig/m3 Duration
Repeated Exposure Studies (cont'd)
0.4 784 2 h/day for 5 days,
2 h either 10 or
20 days later
Activity Level
(VE)
IE
(60 L/min) at
1 5 -min intervals
Number and Gender
of Subjects
16 M;
18 to 35 years of age
Observed Effect(s)
BAL done immediately after fifth day of exposure and again
after exposure 10 or 20 days later. Most markers of
inflammation (PMNs, IL-6, PGE2, fibronectin) showed
Reference
Devlin et al.
(1997)
complete attenuation; markers of damage (LDH, IL-8, protein,
al-antitrypsin, elastase) did not. Reversal of attenuation was
not complete for some markers, even after 20 days.
X
0.40 784 2 h
0.60
IE
(83 W for women,
1 00 W for men)
at 1 5-min intervals
7M, 3F
23 to 41 years of age
BAL fluid 3 h after exposure had significant increases
in PMNs, PGE2, TXB2, and PGF2ct at both O3 concentrations.
Seltzer et al.
(1986)
* See Appendix A for abbreviations and acronyms.
b Listed from lowest to highest O3 concentration.
H
b
o
o
H
O
O
H
W
O
O
HH
H
W
-------
Table AX6-13. Studies of Effects on Host Defense, on Drug Effects and Supportive In Vitro Studies Relating to Controlled
Human Exposure to Ozone"
to
o
o
X
O
to
H
6
o
o
H
O
Ozone Concentration13
ppm |ig/m3 Duration
Activity Level Number and
(VE) Gender of Subjects
Observed Effect(s)
Reference
Host Defense
0.2 392 2 h
0.3 588 6 h/day for
5
consecutive
days
0.2 382 2 h
IE(15min/30 4M, 5F
min); mild atopic
(VE) ~ 20 asthmatics;
L/min/m2 BSA 2 1 -42 years of age
IE (light treadmill) 24 M
(12 O3, 12 air)
IE 10M,2F
(15 min/30 min); healthy NS
( VE) ~ 30 L/min mean ~ 28 years
of age
A significant decline in FEV! and VC immediately following
exposure. A 2-fold increase in percent PMNs, with no changes in
other biomarkers, was observed at 6 h post exposure. By 24 h post-
exposure, PMNs had decreased, but albumin, total protein,
myeloperoxidase and eosinophil cationic protein had increased.
Subjects inoculated with type 39 rhinovirus prior to exposure.
NL was performed on the morning of Days 1 to 5, 8, 15, and 30.
No difference in virus liters in NL fluid of air and O3-exposed
subjects at any time tested. No difference in PMNs or interferon
gamma in NL fluid, or in blood lymphocyte proliferative response
to viral antigen.
Subjects were exposed to O3 and FA in a cross-over design and
underwent BAL 6 h post-exposure. O3 exposure induced a 3-fold
increase in % PMNs and epithelial cells, and increased IL-8, Gro-a,
and total protein in BAL fluid. % PMNs correlated positively with
chemokine levels. Exposure also resulted in a significant decrease in
the CD4+/CD8+ ratio and the % of activated CD4+ and CD8+ T
cells in BAL fluid.
Newson et al.
(2000)
Henderson et al.
(1988)
Krishna et al.
(1998)
Host Defense - Mucous Clearance
0.4 784 1 h
0.20 392 2 h
0.40 784
CE (40 L/min) 1 5 healthy NS
18 to 35 years old
IE (light treadmill) 7 M,
27.2 ± 6.0 (SD)
years old
Subjects inhaled radiolabeled iron oxide particles 2 h after exposure.
No significant O3 -induced effect on clearance of particles during the
next 3 h or the following morning.
Subjects inhaled radiolabeled iron oxide particles immediately
before exposure. Concentration-dependent increase in rate of
particle clearance 2 h after exposure, although clearance was
confined primarily to the peripheral airways at the lower O3
concentration.
Gerrity et al.
(1993)
Foster et al.
(1987)
o
HH
H
W
-------
to
o
o
Table AX6-13 (cont'd). Studies of Effects on Host Defense, on Drug Effects and Supportive In Vitro Studies Relating to
Controlled Human Exposure to Ozone"
Ozone Concentration11
Exposure Number and Gender
ppm Hg/m3 Duration Activity Level (VE) of Subjects
Observed Effect(s)
Reference
Host Defense - Epithelial Permeability
0.15
0.35
0.5
0.4
294 130 min IE
686 10 exercise/
10 rest
(VE) = 8xFVC
784 2.25 h IE
(70 L/min) at
15-min intervals
784 2 h IE
(70 L/min) at
15-min intervals
8M,1F
NS
16 M,
20 to 30 years old
8M,
20 to 30 years old
Subjects inhaled 99mTc-DTPA 19 h after exposure to O3. Clearance was
increased in the lung periphery. Clearance was not related to spirometry.
Similar design and results as earlier study (Kehrl et al., 1987). For the
combined studies the average rate of clearance was 60% faster in O3-
exposed subjects.
Subjects inhaled 99mTc-DTPA 75 min after exposure. Significantly
increased clearance of 99mTc-DTPA from the lung in O3-exposed
subjects. Subjects had expected changes in FVC and SRaw.
Foster and
Stetkiewicz
(1996)
Kehrl etal. (1989)
Kehrl etal. (1987)
Drug Effects on Inflammation
X
I
o
H
6
o
o
H
O
o
HH
H
W
0.4
784
2h
23 healthy adults Subjects were exposed to O3 following random selection for a 2 wk daily
regimen of antioxidants, including vegetable juice high in the carotenoid,
lycopene, or placebo. Concentrations of lycopene in the lungs of
supplemented subjects increased by 12% following treatment.
Supplemented subjects showed a 20% decrease in epithelial cell DNA
damage as assessed by the Comet Assay. Effects attributable to lycopene
could not be separated from those of other antioxidants.
Arab etal. (2002)
0.0 0 2 h IE 4 M, 5 F
0.4 784 20 min
mild-mod.
exercise,
10 min rest
0.2 392 2h IE
All exposures (15 min/30 min);
separated by (VE) ~ 20 L/min/m2
at least 2 wks BSA
(mean ~ 30d)
Healthy NS
30 ± 3 years old
Healthy (6 M, 9 F)
and mild asthmatic
(9 M, 6 F) subjects
Subjects previously in Nightingale et al. (2000) study. Placebo-control:
Immediately postexposure decrements in FVC (9%) and FEVj (14%)
relative to pre-exposure values. FEVj decrement only 9% at 1 hr
postexposure. By 3 h postexposure, recovery in FVC to 97% and FEVj
to 98% of preexposure values. Significant increases in 8-isoprostane at 4
h postexposure. Budesonide for 2 wk prior to exposure did not affect
responses.
Comparison was made of responses in healthy subjects, who had higher
basal ascorbate (ASC) levels and lower glutathione disulfide (GSSG)
levels than those of asthmatics. 6 h after exposure, ASC levels were
decreased and GSSG levels were increased in BAL fluid of normals, but
not asthmatics. Despite these differences in basal antioxidant levels and
response to O3, decrements in FEVj and neutrophil influx did not differ
in the two subject groups.
Montuschi et al.
(2002)
Mud way et al.
(2001)
-------
Table AX6-13 (cont'd). Studies of Effects on Host Defense, on Drug Effects and Supportive In Vitro Studies Relating to
Controlled Human Exposure to Ozone"
to
o
o
X
I
o
H
6
o
o
H
O
o
HH
H
W
Oz one Concentration11
ppm
ug/m3 Duration
Activity Level ( VE)
Number and Gender
of Subjects
Observed Effect(s)
Reference
Drug Effects on Inflammation (cont'd)
0.4
0.27
0.4
0.4
0.0
0.4
784 2h
529 2h
All exposures
separated by
at least 1 wk
(mean ~ 14 d)
784 2h
784 2h
784 2 h IE
4x15 min at
VE=18
L/min/m2
BSA
2 exposures:
25% subjects
exposed to
air-air, 75%
to O3-O3
IE
15-min intervals;
VE ~ 20 L/min/m2
BSA
Continuous
exercise;
(VE) ~ 25 L/min/m2
BSA
IE
15-min intervals
VE min ~ 30 L/min
IE
(20 min/30 min);
workload @
50 watts
21 °C
40% RH
Placebo group
15 M, 1 F
Antioxidant group
13M,2F
Mean age 27 years
7 M, F subjects with
mild asthma; 20-50
years of age
5 M, 4 F healthy
6 M, 7 F asthmatics
6M, 9 F healthy NS
mean ~ 3 1 years of
age
Weak responders
7 M, 13F
Strong responders
21 M, 21 F
Healthy NS
20 to 59 years old
All subjects were exposed to FA and then entered a 2 wk regimen
of placebo or 250 mg Vit C, 50IU a-tocopherol, and 12 oz veg.
cocktail/day prior to O3 exposure. O3-induced decrements in FEVj and
FVC were 30% and 24% less, respectively, in supplemented subjects.
Percent neutrophils and IL-6 levels in BAL fluid obtained 1 h post
exposure were not different in the two treatment groups.
Subjects were randomly exposed to FA and to O3 before and after 4 wks
of treatment with 400 ug budesonide, b.i.d. Budesonide did not inhibit
the decrement in FEV; or increase in symptom scores, but significantly
reduced the increase in % neutrophils and IL-8 in sputum induced 6 h
post-exposure.
Subjects were pretreated for 3 days prior to exposure with indomethacin
(75 mg/day) or placebo. Similar reductions in FEV; and FVC were seen
in both groups following placebo, whereas mid-flows showed greater
decline in asthmatics than normals. Indomethacin attenuated decrements
in FEVj and FVC in normals, but not asthamtics. Attenuation of
decrements was seen for FEF60p in asthmatics and for FEF50 in normals.
Subjects were randomly exposed to FA and to O3 before and after 2 wks
of treatment with 800 ug budesonide, b.i.d. O3 caused significant
decrements in FEVj and FVC immediately following exposure, and a
small increase in Mch-reactivity and increases in neutrophils and
myeloperoxidase in sputum induced at 4 h post-exposure. No differences
were detected between responses in the two treatment groups.
Significant Cyinduced decrements in spirometric lung function.
Young adults (< 35 years) were significantly more responsive than older
individuals (> 35 years). Sufentanil, a narcotic analgesic, largely
abolished symptom responses and improved FEVj in strong responders.
Naloxone, an opioid antagonist, did not affect O3 effects in weak
responders. See Section AX6.2. 5.1
Sametetal. (2001)
Stech-Scott
et al. (2004)
Vagaggini, et al.
(2001)
Alexis et al. (2000)
Nightingale etal.
(2000)
Passannante etal.
(1998)
-------
Table AX6-13 (cont'd). Studies of Effects on Host Defense, on Drug Effects and Supportive In Vitro Studies Relating to
Controlled Human Exposure to Ozone"
to
o
o
X
I
o
H
6
o
o
H
O
o
HH
H
W
Ozone Concentration11
ppm Hg/m3 Duration
Activity Level ( VE)
Number and Gender
of Subjects
Observed Effect(s)
Reference
Drug Effects on Inflammation (cont'd)
0.4 784 2 h
0.4 784 2 h
0.35 686 1 h
IE
(60 L/min) at
15-min intervals
IE
(15 min/ 30 min);
( VE) = 30 L/min/m2
BSA
Continuous
exercise;
(VE) = 60 L/min
10 M
1 3 healthy male
subjects
14 healthy college-
age males
Subjects given 800 mg ibuprofen or placebo 90 min before exposure.
Subjects given ibuprofen had less of a decrease in FEVj after O3
exposure. BAL fluid 1 h after exposure contained similar levels of
PMNs, protein, fibronectin, LDH, a-1 antitrypsin, LTB4, and C3a in both
ibuprofen and placebo groups. However, subjects given ibuprofen had
decreased levels of IL-6, TXB2, and PGE2.
Four days prior to O3 exposure, subjects received either no treatment,
placebo or 150 mg indomethacin/day. Indomethacin treatment
attenuated the O3-induced decrease in FEVj, but had no effect on the O3-
induced increase in Mch responsiveness.
In a placebo- and air-controlled random design, subjects were treated
with 75 mg indomethacin every 12 h for 5 days prior to exposure.
Indomethacin significantly reduced O3-induced decrements in FEVj and
FVC.
Hazuchaetal.
(1996)
Yingetal. (1990)
Schelegle et al.
(1987)
Supportive In Vitro Studies
0.01 to 19.6 to 6h
0.10 196
0.1 196 24 h
0.2 392 3 h
1 1,690 4h
bronchial
epithelial cells
Nasal mucosa
Nasal epithelial
cells and airway
epithelial cell line
Macrophage-like
THP-1 cells
Nonatopic,
nonasthmatic and
atopic, mild
asthmatic bronchial
biopsy samples
Allergic and
nonallergic patients
Exposure to 0.01-0.10 ppm O3 significantly decreased the electrical
resistance of cells from asthmatic sources, compared to nonasthmatic
sources. This range of O3 concentrations also increased the
movement of 14C-BSA across the confluent cultures of "asthmatic"
cells to an extent that was greater than that in "nonasthmatic" cells.
Increased concentrations of neurokinin A and substance P in
medium following O3 exposure. Levels of release of both
neuropeptides were higher from tissues derived from allergic
compared to nonallergic patients.
Synergistic effect of O3 exposure on rhino virus- induced release
of IL-8 at 24 h through mechanisms abrogated by antioxidant
pretreatment. Additive enhancement of 1C AM- 1 expression.
THP-1 cells were treated with samples of human surfactant protein
A (SP-A) genetic variants (SP-A1 and SP-A2) that had been
previously exposed to O3. O3-exposed variants differed in then-
ability to stimulate the production of TNFa and IL-8 by these cells.
Bayram et al.
(2002)
Schierhorn et al.
(2002)
Spannhake et al.
(2002)
Wang et al.
(2002)
-------
Table AX6-13 (cont'd). Studies of Effects on Host Defense, on Drug Effects and Supportive In Vitro Studies Relating to
Controlled Human Exposure to Ozone"
to
o
o
X
Oi
O
Oi
H
6
o
o
H
O
Ozone Concentration5
ppm
ug/m3 Duration
Activity Level
(VE)
Number and
Gender of Subjects Observed Effect(s)
Reference
Supportive In Vitro Studies (cont'd)
0.01 to
0.10
0.12
0.24
0.50
0.06 to
0.20
0.5
0.5
0.4
19.6 to 6h
196
235 3h
470
980
118 to 24 h
392
980 Ih
980 Ih
784 Ih
bronchial
epithelial cells
Nasal epithelial
cells
Nasal mucosa
Lung fibroblast
cell line with an
airway epithelial
cell line
tracheal epithelial
cells
Lung fibroblasts;
airway epithelial
cell line
Nonatopic, No difference in constitutive release of IL-8, GM-CSF, si CAM- 1
nonasthmatic and and RANTES from cells from nonasthmatic and asthmatic sources,
atopic, mild except for detection of RANTES in latter cells only. Increased
asthmatic bronchial release of all mediators 24 h after 0.05 to 0.10 ppm O3 in
biopsy samples "asthmatic" cells, but only IL-8 and si CAM- 1 in "nonasthmatic"
cells.
Small dose-response activation of NF-KB coinciding with
O3-induced production of free radicals assessed by electron spin
resonance. Increased TNFa at two higher concentrations of O3 at
16 h post-expo sure.
105 surgical Increased histamine release correlated with mast cell degranulation.
samples from Increased release of IL-1, IL-6, IL-8 and TNFa following O3
atopic and exposure at 0. 10 ppm. Release of IL-4, IL-6, IL-8 and TNFa at
nonatopic patients this concentration was significantly greater from tissues from atopic
versus nonatopic patients.
BEAS-2B cells in the presence or absence of HFL-1 cells were
exposed and incubated for 11 or 23 h. Steady-state mRNA levels
of alpha 1 procollagens type I and II, as well as TGF(31, were
increased in O3-exposed co-cultured fibroblasts compared to air
controls. Data support interactions between the cell types in the
presence and the absence of O3 -expo sure.
O3 exposure caused an increase in ROS formation and a decline in
PGE2 production. No differences in mRNA and protein levels of
prostaglandin endoperoxide G/H synthase 2 (PGHS-2) or the rate
of its synthesis were detected, suggesting a direct effect of O3-
generated oxidants on PGHS-2 activity.
Cells incubated with O3-exposed arachidonic acid (AA) were found
to contain DNA single strand breaks. Pretreatment of the exposed
AA solution with catalase eliminated the effect on DNA, indicating
its dependence on H2O2 production. The effect was potentiated by
the non-carbony 1 component of ozonized AA.
Bayram et al.
(2001)
Nichols et al.
(2001)
Schierhorn et al.
(1999)
Langetal. (1998)
Alpert et al.
(1997)
Kozumbo et al.
(1996)
o
HH
H
W
-------
3
Table AX6-13 (cont'd). Studies of Effects on Host Defense, on Drug Effects and Supportive In Vitro Studies Relating to
Controlled Human Exposure to Ozone"
<<
to
o
o
X
Oi
i
O
O
H
6
o
0
H
O
O
0
o
H
W
Ozone Concentration13
ppm |ig/m3 Duration
Supportive In Vitro Studies (cont'd)
0.25 490 6 h
0.50 980
0.25 490 1 h
0.50 980
1.00 1,960
0.20 to 392 to 2hor4h
1.0 1960
0.25 490 1 h
0.50 980
1.00 1,960
0.30 588 1 h
1.00 1,960
Activity Level
(VE)
Human nasal
epithelial cells
Airway epithelial
cell line and
alveolar
macrophages
Airway epithelial
cell line
Airway epithelial
cell line
Alveolar
macrophages
Number and
Gender of Subjects Observed Effect(s)
Increased in ICAM-1, IL-6, IL-1, and TNF expression at 0.5 ppm.
No increase in IL-8 expression. No increases at 0.25 ppm.
Increased secretion of IL-6, IL-8, and fibronectin by epithelial cells,
even at lowest O3 concentration. No O3-induced secretion of these
compounds by macrophages.
O3 caused a dose-related loss in cellular replicative activity at
exposure levels that caused minimal cytotoxicity. DNAsingle
strand breaks were not detected. These effects were different from
those of H2O2 and, thus, not likely related to production of this
oxidant within the cells.
Concentration-dependent increased secretion of PGE2, TXB2,
PGF2ct, LTB4, and LTD4. More secretion basolaterally than apically.
Concentration-dependent increases in PGE2 production, and
decreases in phagocytosis of sheep erythrocytes. No O3-induced
secretion of IL-1, TNF, or IL-6.
Reference
Beck etal. (1994)
Devlin et al.
(1994)
Gabrielson et al.
(1994).
McKinnon et al.
(1993)
Becker et al.
(1991)
"See Appendix A for abbreviations and acronyms.
bListed from lowest to highest O3 concentration.
-------
1 studied include those involved with fibrin deposition and degradation (Tissue Factor, Factor VII,
2 and plasminogen activator), potential markers of fibrogenesis (fibronectin, platelet derived
3 growth factor), and components of the complement cascade (C3a).
4 Inflammatory cells of the lung such as alveolar macrophages (AMs), monocytes, and
5 PMNs also constitute an important component of the pulmonary host defense system. Upon
6 activation, they are capable of generating free radicals and enzymes with microbicidal
7 capabilities, but they also have the potential to damage nearby cells. More recently published
8 studies since the last literature review (U.S. Environmental Protection Agency, 1996) observed
9 changes in T lymphocyte subsets in the airways following exposure to O3 that suggest
10 components of the immune host defense also may be affected.
11
12 AX6.9.2 Inflammatory Responses in the Upper Respiratory Tract
13 The nasal passages constitute the primary portal for inspired air at rest and, therefore, the
14 first region of the respiratory tract to come in contact with airborne pollutants. Nikasinovic et al.
15 (2003) recently reviewed the literature of laboratory-based nasal inflammatory studies published
16 since 1985. Nasal lavage (NL) has provided a useful tool for assessing O3-induced inflammation
17 in the nasopharynx. Nasal lavage is simple and rapid to perform, is noninvasive, and allows
18 collection of multiple sequential samples. Graham et al. (1988) reported increased levels of
19 PMNs in the NL fluid of humans exposed to 0.5 ppm O3 at rest for 4 h on 2 consecutive days,
20 with NL performed immediately before and after each exposure, as well as 22 h after the second
21 exposure. Nasal lavage fluid contained elevated numbers of PMNs at all postexposure times
22 tested, with peak values occurring immediately prior to the second day of exposure. Bascom
23 et al. (1990) exposed subjects with allergic rhinitis to 0.5 ppm O3 at rest for 4 h, and found
24 increases in PMNs, eosinophils, and mononuclear cells following O3 exposure. Graham and
25 Koren (1990) compared inflammatory mediators present in both the NL and BAL fluids of
26 humans exposed to 0.4 ppm O3 for 2 h. Increases in NL and BAL PMNs were similar (6.6- and
27 eightfold, respectively), suggesting a qualitative correlation between inflammatory changes in
28 the lower airways (BAL) and the upper respiratory tract (NL), although the PMN increase in NL
29 could not quantitatively predict the PMN increase in BAL. Torres et al. (1997) compared NL
30 and BAL in smokers and nonsmokers exposed to 0.22 ppm O3 for 4 h. In contrast to Graham
31 and Koren (1990), they did not find a relationship between numbers or percentages of
January 2005 AX6-108 DRAFT-DO NOT QUOTE OR CITE
-------
1 inflammatory cells (PMNs) in the nose and the lung, perhaps in part due to the variability
2 observed in their NL recoveries. Albumin, a marker of epithelial cell permeability, was
3 increased 18 h later, but not immediately after exposure, as seen by Bascom et al. (1990).
4 Tryptase, a constituent of mast cells, was also elevated after O3 exposure at 0.4 ppm for 2 h
5 (Koren et al., 1990). McBride et al. (1994) reported that asthmatic subjects were more sensitive
6 than non-asthmatics to upper airway inflammation at an O3 concentration (0.24 ppm (1.5 h)) that
7 did not affect lung or nasal function or biochemical mediators. A significant increase in the
8 number of PMNs in NL fluid was detected in the asthmatic subjects both immediately and 24 h
9 after exposure. Peden et al. (1995) also found that O3 at a concentration of 0.4 ppm had a direct
10 nasal inflammatory effect, and reported a priming effect on the response to nasal allergen
11 challenge, as well. A subsequent study in dust mite-sensitive asthmatic subjects indicated that
12 O3 at this concentration enhanced eosinophil influx in response to allergen, but did not promote
13 early mediator release or enhance the nasal response to allergen (Michelson et al., 1999).
14 Similar to observations made in the lower airways, the presence of O3 molecular "targets" in
15 nasal lining fluid is likely to provide some level of local protection against exposure. In a study
16 of healthy subjects exposed to 0.2 ppm O3 for 2 h, Mudway and colleagues (1999) observed a
17 significant depletion of uric acid in NL fluid at 1.5 h following exposure.
18 An increasing number of studies have taken advantage of advances in cell and tissue
19 culture techniques to examine the role of upper and lower airway epithelial cells and mucosa in
20 transducing the effects of O3 exposure. Many of these studies have provided important insight
21 into the basis of observations made in vivo. One of the methods used enables the cells or tissue
22 samples to be cultured at the air-liquid interface (ALI), allowing cells to establish apical and
23 basal polarity, and both cells and tissue samples to undergo exposure to O3 at the apical surfaces
24 as would occur in vivo. Nichols and colleagues (2001) examined human nasal epithelial cells
25 grown at the ALI for changes in free radical production, based on electron spin resonance, and
26 activation of the NF-KB transcription factor following exposure to O3 at 0.12 to 0.5 ppm for 3 h.
27 They found a dose-related activation of NF-KB within the cells that coincided with O3-induced
28 free radical production and increased release of TNFa at levels above 0.24 ppm. These data
29 confirm the importance of this oxidant stress-associated pathway in transducing the O3 signal
30 within nasal epithelial cells, and suggest its role in directing the inflammatory response. In a
31 study of nasal mucosal biopsy plugs, Schierhorn, et al. (1999) found that tissues exposed to O3 at
January 2005 AX6-109 DRAFT-DO NOT QUOTE OR CITE
-------
1 a concentration of 0.1 ppm induced release of IL-4, IL-6, IL-8, and TNFa that was significantly
2 greater from tissues from atopic patients compared to nonatopic controls. In a subsequent study,
3 this same exposure regimen caused the release of significantly greater amounts of the
4 neuropeptides, neurokinin A and substance P, from allergic patients, compared to nonallergic
5 controls, suggesting increased activation of sensory nerves by O3 in the allergic tissues
6 (Schierhorn et al., 2002).
7
8 AX6.9.3 Inflammatory Responses in the Lower Respiratory Tract
9 Seltzer et al. (1986) were the first to demonstrate that exposure of humans to O3 resulted in
10 inflammation in the lung. Bronchoalveolar lavage fluid (3 h post-exposure) from subjects
11 exposed to O3 contained increased PMNs as well as increased levels of PGE2, PGF2a, and TXB2
12 compared to fluid from air-exposed subjects. Koren et al. (1989a,b) described inflammatory
13 changes 18 h after O3 exposure. In addition to an eightfold increase in PMNs, Koren et al.
14 reported a two-fold increase in BAL fluid protein, albumin, and immunoglobulin G (IgG) levels,
15 suggestive of increased epithelial cell permeability. There was a 12-fold increase in IL-6 levels,
16 a two-fold increase in PGE2, and a two-fold increase in the complement component, C3a.
17 Evidence for stimulation of fibrogenic processes in the lung was shown by significant increases
18 in coagulation components, Tissue Factor and Factor VII (McGee et al., 1990), urokinase
19 plasminogen activator and fibronectin (Koren et al., 1989a). Subsequent studies by Lang, et al.
20 (1998), using co-cultures of cells of the BEAS-2B bronchial epithelial line and of the HFL-1
21 lung fibroblast line, provided additional information about O3-induced fibrogenic processes.
22 They demonstrated that steady-state mRNA levels of both alpha 1 and procollagens type I and
23 III in the fibroblasts were increased following O3 exposure and that this effect was mediated by
24 the O3-exposed epithelial cells. This group of studies demonstrated that exposure to O3 results in
25 an inflammatory reaction in the lung, as evidenced by increases in PMNs and proinflammatory
26 compounds. Furthermore, they demonstrated that cells and mediators capable of damaging
27 pulmonary tissue are increased after O3 exposure, and provided early suggestion of the potential
28 importance of the epithelial cell-myofibroblast "axis" in modulating fibrotic and fibrinolytic
29 processes in the airways.
30 Isolated lavage of the mainstream bronchus using balloon catheters or BAL using small
31 volumes of saline have been used to assess O3-induced changes in the large airways. Studies
January 2005 AX6-110 DRAFT-DO NOT QUOTE OR CITE
-------
1 collecting lavage fluid from isolated airway segments after O3 exposure indicate increased
2 neutrophils in the airways (Aris et al., 1993; Balmes et al., 1996; Scannell et al., 1996). Other
3 evidence of airway neutrophil increase comes from studies in which the initial lavage fraction
4 ("bronchial fraction") showed increased levels of neutrophils (Schelegle et al., 1991; Peden
5 et al., 1997; Balmes et al., 1996; Torres et al., 1997). Bronchial biopsies show increased PMNs
6 in airway tissue (Aris et al.,1993) and, in sputum collected after O3 exposure, neutrophil numbers
7 are elevated (Fahy et al., 1995).
8 Increased BAL protein, suggesting O3-induced changes in epithelial permeability (Koren
9 et al., 1989a, 1991 and Devlin et al., 1991) supports earlier work in which increased epithelial
10 permeability, as measured by increased clearance of radiolabled diethylene triamine pentaacetic
11 acid (99mTc-DTPA) from the lungs of humans exposed to O3, was demonstrated (Kehrl et al.,
12 1987). In addition, Foster and Stetkiewicz (1996) have shown that increased permeability
13 persists for at least 18-20 h and the effect is greater at the lung apices than at the base. In a study
14 of mild atopic asthmatics exposed to 0.2 ppm O3 for 2 h, Newson, et al. (2000) observed a 2-fold
15 increase in the percentage of PMNs present at 6 hours post exposure, with no change in markers
16 of increased permeability as assessed by sputum induction. By 24 h, the neutrophilia was seen
17 to subside while levels of albumin, total protein, myeloperoxidase, and eosinophil cationic
18 protein increased significantly. It was concluded that the transient PMN influx induced by acute
19 exposure of these asthmatic subjects was followed by plasma extravasation and the activation of
20 both PMNs and eosinophils within the airway tissues. Such changes in permeability associated
21 with acute inflammation may provide better access of inhaled antigens, particulates, and other
22 substances to the submucosal region.
23 Devlin et al. (1991) reported an inflammatory response in subjects exposed to 0.08 and
24 0.10 ppm O3 for 6.6 h. Increased numbers of PMNs and levels of IL-6 were found at both
25 O3 concentrations, suggesting that lung inflammation from O3 can occur as a consequence of
26 prolonged exposure to ambient levels while exercising. Interestingly, those individuals who had
27 the largest increases in inflammatory mediators in this study did not necessarily have the largest
28 decrements in pulmonary function, suggesting that separate mechanisms underlie these two
29 responses. The absence of a relationship between spirometric responses and inflammatory cells
30 and markers has been reported in several studies, including Balmes et al., 1996; Schelegle et al.,
31 1991; Torres et al., 1997; Hazucha et al., 1996; Blomberg et al., 1999. These observations relate
January 2005 AX6-111 DRAFT-DO NOT QUOTE OR CITE
-------
1 largely to disparities in the times of onset and duration following single exposures. The
2 relationship between inflammatory and residual functional responses following repeated or
3 chronic exposures may represent a somewhat different case (see Section AX6.9.4).
4 As indicated above, a variety of potent proinflammatory mediators have been reported to
5 be released into the airway lumen following O3 exposure. Studies of human alveolar
6 macrophages (AM) and airway epithelial cells exposed to O3 in vitro suggest that most mediators
7 found in the BAL fluid of O3-exposed humans are produced by epithelial cells. Macrophages
8 exposed to O3 in vitro showed only small increases in PGE2 (Becker et al., 1991). In contrast,
9 airway epithelial cells exposed in vitro to O3 showed large concentration-dependent increases in
10 PGE2, TXB2, LTB4, LTC4, and LTD4 (McKinnon et al., 1993) and increases in IL-6, IL-8, and
11 fibronectin at O3 concentrations as low as 0.1 ppm (Devlin et al., 1994). Macrophages lavaged
12 from subjects exposed to 0.4 ppm (Koren et al., 1989a) showed changes in the rate of synthesis
13 of 123 different proteins, whereas AMs exposed to O3 in vitro showed changes in only six
14 proteins, suggesting that macrophage function was altered by mediators released from other
15 cells. Furthermore, recent evidence suggests that the release of mediators from AMs may be
16 modulated by the products of O3-induced oxidation of airway lining fluid components, such as
17 human surfactant protein A (Wang et al., 2002).
18 Although the release of mediators has been demonstrated to occur at exposure
19 concentrations and times that are minimally cytotoxic to airway cells, potentially detrimental
20 latent effects have been demonstrated in the absence of cytotoxicity. These include the
21 generation of DNA single strand breaks (Kozumbo et al., 1996) and the loss of cellular
22 replicative activity (Gabrielson et al., 1994) in bronchial epithelial cells exposed in vitro, and the
23 formation of protein and DNA adducts. A highly toxic aldehyde formed during O3-induced lipid
24 peroxidation is 4-hydroxynonenal (HNE). Healthy human subjects exposed to 0.4 ppm O3 for 1
25 h underwent BAL 6 h later. Analysis of lavaged alveolar macrophages by Western blot
26 indicated increased levels of a 32-kDa HNE-protein adduct, as well as 72-kDa heat shock protein
27 and ferritin, in O3- versus air-exposed subjects (Hamilton et al., 1998). In a recent study of
28 healthy subjects exposed to 0.1 ppm O3 for 2 h (Corradi et al., 2002), formation of 8-hydroxy-2'-
29 deoxyguanosine (8-OHdG), a biomarker of reactive oxidant species (ROS)-DNA interaction,
30 was measured in peripheral blood lymphocytes. At 18 h post exposure, 8-OHdG was
31 significantly increased in cells compared to pre-exposure levels, presumably linked to concurrent
January 2005 AX6-112 DRAFT-DO NOT QUOTE OR CITE
-------
1 increases in chemical markers of ROS. Of interest, the increase in 8-OHdG was only significant
2 in a subgroup of subjects with the wild genotype for NAD(P)H:quinone oxidoreductase and the
3 null genotype for glutathione-S-transferase Ml, suggesting that polymorphisms in redox
4 enzymes may confer "susceptibility' to O3 in some individuals. The generation of ROS
5 following exposure to O3 has been shown to be associated with a wide range of responses. In a
6 recent study, ROS production by alveolar macrophages lavaged from subjects exposed to
7 0.22 ppm for 4 h was assessed by flow cytometry (Voter et al., 2001). Levels were found to be
8 significantly elevated 18 h post exposure and associated with several markers of increased
9 permeability. An in vitro study of human tracheal epithelial cells exposed to O3 indicated that
10 generation of ROS resulted in decrease in synthesis of the bronchodilatory prostaglandin, PGE2,
11 as a result of inactivation of prostaglandin endoperoxide G/H synthase 2 (Alpert et al., 1997).
12 These and similar studies indicate that the responses to products of O3 exposure in the airways
13 encompass a broad range of both stimulatory and inhibitory activities, many of which may be
14 modulated by susceptibility factors upstream in the exposure process, at the level of
15 compensating for the imposed oxidant stress.
16 The inflammatory responses to O3 exposure also have been studied in asthmatic subjects
17 (Basha et al., 1994; Scannell et al., 1996; Peden et al., 1997). In these studies, asthmatics
18 showed significantly more neutrophils in the BAL (18 h post-exposure) than similarly exposed
19 healthy individuals. In one of these studies (Peden et al., 1997), which included only allergic
20 asthmatics who tested positive for Dematophagoides farinae antigen, there was an eosinophilic
21 inflammation (2-fold increase), as well as neutrophilic inflammation (3-fold increase). In a
22 study of subjects with intermittent asthma that utilized a 2-fold higher concentration of O3 (0.4
23 ppm) for 2 h, increases in eosinophil cationic protein, neutrophil elastase and IL-8 were found to
24 be significantly increased 16 h post-exposure and comparable in induced sputum and BAL fluid
25 (Hiltermann et al, 1999). In two studies (Basha et al., 1994; Scannell et al., 1996), IL-8 was
26 significantly higher in post-O3 exposure BAL in asthmatics compared to non-asthmatics,
27 suggesting a possible mediator for the increased neutrophilic inflammation in those subjects.
28 In a recent study comparing the neutrophil response to O3 at a concentration and exposure time
29 similar to those of the latter three studies, Stenfors and colleagues (2002) were unable to detect a
30 difference in the increased neutrophil numbers between 15 mild asthmatic and 15 healthy
31 subjects by bronchial wash at the 6 h post-exposure time point. These results suggest that, at
January 2005 AX6-113 DRAFT-DO NOT QUOTE OR CITE
-------
1 least with regard to neutrophil influx, differences between healthy and asthmatic individuals
2 develop gradually following exposure and may not become evident until later in the process.
3 In another study, mild asthmatics who exhibited a late phase underwent allergen challenge 24 hrs
4 before a 2 h exposure to 0.27 ppm O3 or filtered air in a cross-over design (Vagaggini et al.,
5 2002). At 6 h post-exposure, eosinophil numbers in induced sputum were found to be
6 significantly greater after O3 than after air. Studies such as these suggest that the time course of
7 eosinophil and neutrophil influx following O3 exposure can occur to levels detectable within the
8 airway lumen by as early as 6 h. They also suggest that the previous or concurrent activation of
9 proinflammatory pathways within the airway epithelium may enhance the inflammatory effects
10 of O3. For example, in an in vitro study of epithelial cells from the upper and lower respiratory
11 tract, cytokine production induced by rhinovirus infection was enhanced synergistically by
12 concurrent exposure to O3 at 0.2 ppm for 3 h (Spannhake et al, 2002). The use of bronchial
13 mucosal biopsies has also provided important insight into the modulation by O3 of existing
14 inflammatory processes within asthmatics. In a study of healthy and allergic asthmatic subjects
15 exposed to 0.2 ppm O3 or filtered air for 2 h, biopsies were performed 6 hr following exposure
16 (Bosson et al., 2003). Monoclonal antibodies were used to assess epithelial expression of a
17 variety of cytokines and chemokines. At baseline (air exposure), asthmatic subjects showed
18 significantly higher expression of interleukins (IL)-4 and -5. Following O3 exposure, the
19 epithelial expression of IL-5, IL-8, granulocyte-macrophage colony-stimulating factor (GM-
20 CSF) and epithelial cell-derived neutrophil-activating peptide 78 (ENA-78) was significantly
21 greater in asthmatic subjects, as compared to healthy subjects. In vitro studies of bronchial
22 epithelial cells derived by biopsy from nonatopic, nonasthmatic subjects and asthmatic subjects
23 also demonstrated the preferential release of GM-CSF and also of regulated on activation,
24 normal T cell-expressed and -secreted (RANTES) from asthmatic cells following O3 exposure.
25 The time course of the inflammatory response to O3 in humans has not been explored fully.
26 Nevertheless, studies in which BAL was performed 1-3 h (Devlin et al., 1990; Koren et al.,
27 1991; Seltzer et al., 1986) after exposure to 0.4 ppm O3 demonstrated that the inflammatory
28 response is quickly initiated, and other studies (Koren et al., 1989a,b; Torres et al., 1997;
29 Scannell et al., 1996; Balmes et al., 1996) indicated that, even 18 h after exposure, inflammatory
30 mediators such as IL-6 and PMNs were still elevated. However, different markers show peak
31 responses at different times. Ozone-induced increases in IL-8, IL-6, and PGE2 are greater
January 2005 AX6-114 DRAFT-DO NOT QUOTE OR CITE
-------
1 immediately after O3 exposure, whereas BAL levels of fibronectin and plasminogen activator are
2 greater after 18 h. PMNs and some products (protein, Tissue Factor) are similarly elevated both
3 1 and 18 h after O3 exposure (Devlin et al., 1996; Torres et al., 1997). Schelegle et al. (1991)
4 found increased PMNs in the "proximal airway" lavage at 1, 6, and 24 h after O3 exposure, with
5 a peak response at 6 h. In a typical BAL sample, PMNs were elevated only at the later time
6 points. This is consistent with the greater increase 18 h after exposure seen by Torres et al.
7 (1997). In addition to the influx of PMNs and (in allergic asthmatics) eosinophils, lymphocyte
8 numbers in BAL were also seen to be elevated significantly at 6 h following exposure of healthy
9 subjects to 0.2 ppm O3 for 2 h (Blomberg et al., 1997). Analysis of these cells by flow cytometry
10 indicated the increased presence of CD3+, CD4+ and CD8+ T cell subsets. This same laboratory
11 later demonstrated that within 1.5 h following exposure of healthy subjects to the same O3
12 regimen, expression of human leukocyte antigen (HLA)-DR on lavaged macrophages underwent
13 a significant, 2.5-fold increase (Blomberg et al., 1999). The significance of these alterations in
14 immune system components and those in IL-4 and IL-5 expression described above in the
15 studies of Bosson et al. (2003) has not been fully explored and may suggest a role for O3 in the
16 modulation of immune inflammatory processes.
17
18 AX6.9.4 Effects of Repeated Exposures and Adaptation of Responses
19 Residents of areas with high oxidant concentrations tend to have somewhat blunted
20 pulmonary function responses and symptoms to O3 exposure (Hackney et al., 1976, 1977b, 1989;
21 Avol et al., 1988; Linn et al., 1988). Animal studies suggest that while inflammation may be
22 diminished with repeated exposure, underlying damage to lung epithelial cells continues (Tepper
23 et al., 1989). Devlin et al. (1997) examined the inflammatory responces of humans repeatedly
24 exposed to 0.4 ppm O3 for 5 consecutive days. Several indicators of inflammation (e.g., PMN
25 influx, IL-6, PGE2, fibronectin, macrophage phagocytosis) were attenuated after 5 days of
26 exposure (i.e., values were not different from FA). Several markers (LDH, IL-8, total protein,
27 epithelial cells) did not show attenuation, indicating that tissue damage probably continues to
28 occur during repeated exposure. The recovery of the inflammatory response occurred for some
29 markers after 10 days, but some responses were not normalized even after 20 days. The
30 continued presence of markers of cellular injury indicates a persistent but not necessarily
31 perceived response to O3. Christian et al. (1998) randomly subjected heathy subjects to a single
January 2005 AX6-115 DRAFT-DO NOT QUOTE OR CITE
-------
1 exposure and to 4 consecutive days of exposure to 0.2 ppm O3 for 4 h. As reported by others,
2 they found an attenuation of FEVl3 FVC and specific airway resistance when comparing the
3 single exposure with day 4 of the multiday exposure regimen. Similarly, both "bronchial" and
4 "alveolar" fractions of the BAL showed decreased numbers of PMNs and fibronectin
5 concentration at day 4 versus the single exposure, and a decrease in IL-6 levels in the alveolar
6 fraction. Following a similar study design and exposure parameters, but with single day filtered
7 air controls, Torres et al. (2000) found a decrease in FEVj and increases in the percentages of
8 neutrophils and lymphocytes, in concentrations of total protein, IL-6, IL-8, reduced glutathione,
9 ortho-tyrosine and urate in BAL fluid, but no changes in bronchial biopsy histology following
10 the single exposure. Twenty hours after the day 4 exposure, both functional and BAL cellular
11 responses to O3 were abolished. However, levels of total protein, IL-6, IL-8, reduced glutathione
12 and ortho-tyrosine were still increased significantly. In addition, following the day 4 exposure,
13 visual scores for bronchitis, erythema and the numbers of neutrophils in the mucosal biopsies
14 were increased. Their results indicate that, despite reduction of some markers of inflammation
15 in BAL and measures of large airway function, inflammation within the airways persists
16 following repeated exposure to O3. In another study, Frank and colleagues (2001) exposed
17 healthy subjects to filtered air and to O3 (0.25 ppm, 2 h) on 4 consecutive days each, with
18 pulmonary function measurements being made prior to and following each exposure. BAL was
19 performed on day 5, 24 h following the last exposure. On day 5, PMN numbers remained
20 significantly higher in the O3 arm compared to air control. Of particular note in this study was
21 the observation that small airway function, assessed by grouping values for isovolumetric
22 FEF25.75, VmaxSO and Vmax75 into a single value, showed persistent reduction from day 2
23 through day 5. These data suggest that methods to more effectively monitor function in the most
24 peripheral airway regions, which are known to be the primary sites of O3 deposition in the lung,
25 may provide important information regarding the cumulative effects of O3 exposure. It is
26 interesting to note that Alexis et al. (2000) reported that, following exposure of normals and
27 asthmatics to 0.4 ppm O3 for 2 h, variables representing small airways function (viz., FEF25, FEF
28 50, FEF 60P, FEF75) demonstrated the greatest O3-induced decline in the asthmatic subjects. Holz,
29 et al. (2002) made a comparison of early and late responses to allergen challenge following O3 in
30 subjects with allergic rhinitis or allergic asthma. With some variation, both early and late FEVj
January 2005 AX6-116 DRAFT-DO NOT QUOTE OR CITE
-------
1 and cellular responses in the two subject groups were significantly enhanced by 4 consecutive
2 days of exposure to 0.125 ppm O3 for 3 h.
3
4 AX6.9.5 Effect of Anti-Inflammatory and other Mitigating Agents
5 Studies have shown that indomethacin, a non-steroidal anti-inflammatory agent (NSAID)
6 that inhibits the production of cyclooxygenase products of arachidonic acid metabolism, is
7 capable of blunting the well-documented decrements in pulmonary function observed in humans
8 exposed to O3 (Schelegle et al., 1987; Ying et al., 1990). In the latter study, indomethacin did
9 not alter the O3-induced increase in bronchial responsiveness to methacholine. Pretreatment of
10 healthy subjects and asthmatics with indomethacin prior to exposure to 0.4 ppm for 2 h
11 significantly attenuated decreases in FVC and FEVj in normals, but not asthmatics (Alexis et al.,
12 2000). Subjects have also been given ibuprofen, another NSAID agent that blocks
13 cyclooxygenase metabolism, prior to O3 exposure. Ibuprofen blunted decrements in lung
14 function following O3 exposure (Hazucha et al., 1996). Subjects given ibuprofen also had
15 reduced BAL levels of the cyclooxygenase product PGE2 and thromboxane B2, as well as IL-6,
16 but no decreases were observed in PMNs, fibronectin, permeability, LDH activity, or
17 macrophage phagocytic function. These studies suggest that NSAIDs can blunt O3-induced
18 decrements in FEVj with selective (perhaps drug-specific) affects on mediator release and other
19 markers of inflammation.
20 At least two studies have looked at the effects of the inhaled corticosteroid, budesonide, on
21 the effects of O3, with differing outcome perhaps associated with the presence of preexistent
22 disease. Nightingale and colleagues (2000) exposed healthy nonsmokers to 0.4 ppm O3 for 2 h
23 following 2 wk of treatment with budesonide (800 micrograms, twice daily) or placebo in a
24 blinded, randomized cross-over study. This relatively high exposure resulted in significant
25 decreases in spirometric measures and increases in methacholine reactivity and neutrophils and
26 myeloperoxidase in induced sputum. No significant differences were observed in any of these
27 endpoints following budesonide treatment versus placebo. In contrast, Vagaggini et al. (2001)
28 compared the effects of treatment with budesonide (400 micrograms, twice daily) for 4 wk on
29 the responses of mild asthmatic subjects to exposure to 0.27 ppm O3for 2 h. Prior to exposure,
30 at the midpoint and end of exposure, and at 6 h post exposure, FEVj was measured and a
31 symptom questionnaire was administered; at 6 h post exposure, sputum was induced.
January 2005 AX6-117 DRAFT-DO NOT QUOTE OR CITE
-------
1 Budesonide treatment did not inhibit the decrement in FEVj or alter symptom score, but
2 significantly blunted the increase in percent PMNs and concentration of IL-8 in the sputum. The
3 difference in subject health status between the two studies (healthy versus mild asthmatic) may
4 suggest a basis for the differing outcomes; however, because of differences in the corticosteroid
5 dosage and O3 exposure levels, that basis remains unclear.
6 Because the O3 exerts its actions in the respiratory tract by virtue of its strong oxidant
7 activity, it is reasonable to assume that molecules that can act as surrogate targets in the airways,
8 as constituents of either extracellular fluids or the intracellular milieus, could abrogate the effects
9 of O3. Some studies have examined the ability of dietary "antioxidant" supplements to reduce
10 the risk of exposure of the lung to oxidant exposure. In a study of healthy, nonsmoking adults,
11 Samet and colleagues (2001) restricted dietary ascorbate and randomly treated subjects for
12 2 weeks with a mixture of vitamin C, a-tocopherol and vegetable cocktail high in carrot and
13 tomato juices or placebo. Responses to 0.4 ppm O3 for 2 h were assessed in both groups at the
14 end of treatment. O3 -induced decrements in FEVj and FVC were significantly reduced in the
15 supplemented group, whereas the inflammatory response, as assessed by percentage neutrophils
16 and levels of IL-6 in BAL fluid, were unaffected by antioxidant supplementation. In a study that
17 focused on supplementation with a commercial vegetable cocktail high in the carotenoid,
18 lycopene, healthy subjects were exposed for 2 h to 0.4 ppm O3 after 2 wk of antioxidant
19 supplementation or placebo (Arab et al., 2002). These investigators observed that lung epithelial
20 cell DNA damage, as measured by the Comet Assay, decreased by 20 % in supplemented
21 subjects. However, the relationships between the types and levels of antioxidants in airway
22 lining fluid and responsiveness to O3 exposure is likely to be complex. In a study in which
23 differences in ascorbate and glutathione concentrations between healthy and mild asthmatic
24 subjects were exploited, no relationship between antioxidant levels and spirometric or cellular
25 responses could be detected (Mudway, et al., 2001).
26
27 AX6.9.6 Changes in Host Defense Capability Following Ozone Exposure
28 Concern about the effect of O3 on human host defense capability derives from numerous
29 animal studies demonstrating that acute exposure to as little as 0.08 ppm O3 causes decrements
30 in antibacterial host defenses (see Chapter 5). A study of experimental rhinovirus infection in
31 susceptible human volunteers failed to show any effect of 5 consecutive days of O3 exposure on
January 2005 AX6-118 DRAFT-DO NOT QUOTE OR CITE
-------
1 the clinical evolution of, or host response to, a viral challenge (Henderson et al., 1988). Healthy
2 men were nasally inoculated with type 39 rhinovirus (103 TCID50). There was no difference
3 between the O3-exposed and control groups in rhinovirus liters in nasal secretions, in levels of
4 interferon gamma or PMNs in NL fluid, or in blood lymphocyte proliferative response to
5 rhinovirus antigen. However, subsequent findings that rhinovirus can attach to the intracellular
6 adhesion molecule (ICAM)-l receptor on respiratory tract epithelial cells (Greve et al., 1989)
7 and that O3 can up-regulate the ICAM-1 receptor on nasal epithelial cells (Beck et al., 1994)
8 suggest that more studies are needed to explore the possibility that prior O3 exposure can
9 enhance rhinovirus binding to, and infection of, the nasal epithelium.
10 In a single study, human AM host defense capacity was measured in vitro in AMs removed
11 from subjects exposed to 0.08 and 0.10 ppm O3 for 6.6 h while undergoing moderate exercise.
12 Alveolar macrophages from O3-exposed subjects had significant decrements in complement-
13 receptor-mediated phagocytosis of Candida albicans (Devlin et al., 1991). The impairment of
14 AM host defense capability could potentially result in decreased ability to phagocytose and kill
15 inhaled microorganisms in vivo. A concentration-dependent decrease in phagocytosis of AMs
16 exposed to 0.1 to 1.0 ppm O3 in vitro has also been shown Becker et al. (1991). Although the
17 evidence is inconclusive at present, there is a concern that O3 may render humans and animals
18 more susceptible to a subsequent bacterial challenge.
19 Only two studies (Foster et al., 1987; Gerrity et al., 1993) have investigated the effect of
20 O3 exposure on mucociliary particle clearance in humans. Foster et al. (1987) had seven healthy
21 subjects inhale radiolabeled particles (5 jim MMAD) and then exposed these subjects to FA or
22 O3 (0.2 and 0.4 ppm) during light IE for 2 h. Gerrity et al. (1993) exposed 15 healthy subjects to
23 FA or 0.4 ppm O3 during CE (40 L/min) for 1 h; at 2 h post O3 exposure, subjects then inhaled
24 radiolabeled particles (5 jim MMAD). Subjects in both studies had similar pulmonary function
25 responses (average FVC decrease of 11 to 12%) immediately post exposure to 0.4 ppm O3. The
26 Foster et al. (1987) study suggested there is a stimulatory affect of O3 on mucociliary clearance;
27 whereas, Gerrity et al. (1993) found that in the recovery period following O3 exposure, mucus
28 clearance is similar to control, i.e., following a FA exposure. The clearance findings in these
29 studies are complementary not conflicting. Investigators in both studies suggested that
30 O3-induced increases in mucociliary clearance could be mediated by cholinergic receptors.
31 Gerrity et al. (1993) further suggested that transient clearance increases might be coincident to
January 2005 AX6-119 DRAFT-DO NOT QUOTE OR CITE
-------
1 pulmonary function responses; this supposition based on the return of sRaw to baseline and the
2 recovery of FVC to within 5% of baseline (versus an 11% decrement immediately postexposure)
3 prior to clearance measurements.
4 Insofar as the airway epithelial surface provides a barrier to entry of biological, chemical
5 and particulate contaminants into the submucosal region, the maintenance of barrier integrity
6 represents a component of host defense. Many of the studies of upper and lower respiratory
7 responses to O3 exposure previously cited above have reported increases in markers of airway
8 permeability after both acute exposures and repeated exposures. These findings suggest that O3
9 may increase access of airborne agents. In a study of bronchial epithelial cells obtained from
10 nonatopic and mild atopic asthmatic subjects (Bayram et al., 2002), cells were grown to
11 confluence and transferred to porous membranes. When the cultures again reached confluence,
12 they were exposed to 0.01-0.1 ppm O3 or air and their permeability was assessed by measuring
13 the paracellular flux of 14C-BSA. The increase in permeability 24 h following O3 exposure was
14 observed to be significantly greater in cultures of cells derived from asthmatics, compared to
15 healthy subjects. Thus, the late increase in airway permeability following exposure of asthmatic
16 subjects to O3, of the sort described by Newson et al. (2000), may be related to an inherent
17 susceptibility of 'asthmatic' cells to the barrier-reducing effects of O3.
18 As referenced in Section 6.9.3, the O3-induced increase in the numbers of CD8+ T
19 lymphocytes in the airways of healthy subjects reported by Blomberg, et al. (1997) poses several
20 interesting questions regarding possible alterations in immune surveillance processes following
21 exposure. In a subsequent study from the same group, Krishna et al. (1998) exposed healthy
22 subjects to 0.2 ppm O3 or filtered air for 2 h followed by BAL at 6 h. In addition to increased
23 PMNs and other typical markers of inflammation, they found a significant decrease in the
24 CD4+/CD8+ T lymphocyte ratio and in the proportion of activated CD4+ and CD8+ cells.
25 Studies relating to the effects of low-level O3 exposure on the influx and activity of immuno-
26 competent cells in the upper and lower respiratory tracts may shed additional light on
27 modulation of this important area of host defense.
28
29
January 2005 AX6-120 DRAFT-DO NOT QUOTE OR CITE
-------
1 AX6.10 EXTRAPULMONARY EFFECTS OF OZONE
2 Ozone reacts rapidly on contact with respiratory system tissue and is not absorbed or
3 transported to extrapulmonary sites to any significant degree as such. Laboratory animal studies
4 suggest that reaction products formed by the interaction of O3 with respiratory system fluids or
5 tissues may produce effects measured outside the respiratory tract—either in the blood, as
6 changes in circulating blood lymphocytes, erythrocytes, and serum, or as changes in the structure
7 or function of other organs, such as the parathyroid gland, the heart, the liver, and the central
8 nervous system. Very little is known, however, about the mechanisms by which O3 could cause
9 these extrapulmonary effects. (See Section 5.4 for a discussion of the systemic effects of
10 O3 observed in laboratory animals^)
11 The results from human exposure studies discussed in the previous criteria documents
12 (U.S. Environmental Protection Agency, 1986, 1996) failed to demonstrate any consistent
13 extrapulmonary effects. Early studies on peripheral blood lymphocytes collected from human
14 volunteers did not find any significant genotoxic or functional changes at O3 exposures of 0.4 to
15 0.6 ppm for up to 4 h/day. Limited data on human subjects indicated that 0.5 ppm O3 exposure
16 for over 2 h caused transient changes in blood erythrocytes and sera (e.g., erythrocyte fragility
17 and enzyme activities), but the physiological significance of these studies remains questionable.
18 The conclusions drawn from these early studies raise doubt that cellular damage or altered
19 function is occurring to circulating cells at O3 exposures under 0.5 ppm.
20 Other human exposure studies have attempted to identify specific markers of exposure to
21 O3 in blood. For example, Schelegle et al. (1989) showed that PGF2a was elevated after O3
22 exposure (0.35 ppm); however, no increase in a-1 protease inhibitor was observed by Johnson
23 et al. (1986). Foster et al. (1996) found a reduction in the serum levels of the free radical
24 scavenger a-tocopherol after O3 exposure. Vender et al. (1994) failed to find any changes in
25 indices of red blood cell antioxidant capacity (GSH, CAT) in healthy male subjects exposed to
26 0.16 ppm O3 for 7.5 h while intermittently exercising. Liu et al. (1997, 1999) used a salicylate
27 metabolite, 2,3, dehydroxybenzoic acid (DFffiA), to indicate increased levels of hydroxyl radical
28 which hydroxylates salicylate to DHBA. Increased DHBA levels after exposure to 0.12 and
29 0.40 ppm suggest that O3 increases production of hydroxyl radical. The levels of DHBA were
30 correlated with changes in spirometry.
January 2005 AX6-121 DRAFT-DO NOT QUOTE OR CITE
-------
1 Only a few experimental human studies have examined O3 effects in other non-pulmonary
2 organ systems besides blood. Early studies on the central nervous system (Gliner et al., 1979,
3 1980) were not able to find significant effects on motor activity or behavior (vigilance and
4 psychomotor performance) from O3 exposures at rest up to 0.75 ppm (U.S. Environmental
5 Protection Agency, 1986). Drechsler-Parks et al. (1995) monitored ECG, HR, cardiac output,
6 stroke volume, and systolic time intervals in healthy, older subjects (56 to 85 years of age)
7 exposed to 0.45 ppm O3 using a noninvasive impedance cardiographic method. No changes
8 were found at this high O3 concentration after 2 h of exposure while the subjects exercised
9 intermittently at 25 L/min. Gong et al. (1998) monitored ECG, HR, cardiac output, blood
10 pressure, oxygen saturation, and chemistries, as well as calculating other hemodynamic variables
11 (e.g., stroke volume, vascular resistance, rate-pressure products) in both healthy and
12 hypertensive adult males, 41 to 78 years of age. No major acute cardiovascular effects were
13 found in either the normal or hypertensive subjects after exposure to 0.3 ppm O3 for 3 h with
14 intermittent exercise at 30 L/min. Statistically significant O3 effects for both groups combined
15 were a decrease in FEVl3 and increases in HR, rate-pressure product, and the alveolar-to-arterial
16 PO2 gradient, which might be more important in some patients with severe cardiovascular
17 disease. [See Section AX6.3 for a more detailed discussion of the effects ofO3 exposure in
18 subjects with preexisting disease. ]
19
20
21 AX6.11 OZONE MIXED WITH OTHER POLLUTANTS
22 Controlled laboratory studies simulating conditions of ambient exposures have failed for
23 the most part to demonstrate significant adverse effects either in healthy subjects, atopic
24 individuals, or in young and middle-aged asthmatics.
25
26 AX6.11.1 Ozone and Sulfur Oxides
27 The difference in solubilities and other chemical properties of O3 and SOX seems to limit
28 chemical interaction and formation of related species in the mixture of these pollutants either in
29 liquid or gaseous phase. Laboratory studies reviewed in the previous O3 criteria document
30 (Table AX6-14) reported, except for one study (Linn et al., 1994), no significant effects on
31 healthy individuals exposed to mixtures of O3 and SO2 or H2SO4 aerosol. In the study of Linn
January 2005 AX6-122 DRAFT-DO NOT QUOTE OR CITE
-------
Table AX6-14. Ozone Mixed with Other Pollutants3
to
o
o
X
I
to
H
6
o
o
H
/O
o
HH
H
W
Concentration11
ppm Hg/in3
Pollutant
Exposure Duration Exposure
and Activity Conditions'
Number and
Gender of Subject
Subjects Characteristics
Observed Effect(s)
Reference
Sulfur-Containing Pollutants
0.0
0.1 +
0.1 +
0.2
0.3
0.12
0.30
0.05
0.12
0.08
0.12
0.18
0
196" +
262" +
101"
392
564
235
564
70
235
100
157
235
353
100
100
Air
03 +
S02 +
H2S04
03
NO2
H2SO4
03
N02
H2SO4
HNO3
03
H2SO4
03
03
03
NaCl
H2S04
4 h 25 °C
IE 15' ex/ 15' rest 50% RH
VE=22L/min
90 min. 21 °C
VE ~ 32 L/min 50% RH
IE 3 x 15 min
1. 5 h with IE for 22 °C
2 consecutive days; 65% RH
VE = 23.2L/min
6.5 h 21 °C
2 consecutive days 50% RH
50 min exercise/h
VE = 29 L/min
3-h exposure to aerosol, 21 °C
followed 24 h later by a -40% RH
3-h exposure to O3. IE
(10 min per half hour)
VE = 4 times resting
(30 to 364 min)
8 M, 7 F Healthy
1 M, 4 F Asthmatic
10M, 11F Allergic
All NS,
9 to 12 yrs. old
24 Asthmatic NS,
(17 M, 7F) 11 to 18 years old
22 completed Asthmatic NS,
study; adolescents; NS, 12 to
15M, 7F 19 years old
8 M, 7 F Nonasthmatic NS,
22 to 4 1 years old
13M, 17 F Asthmatic NS,
18 to 50 years old
16 M, 14 F Nonasthmatic NS,
18 to 45 years old
10 M, 20 F Asthmatic NS,
21 to 42 years old
Spirometry, PEER and subjective
symptoms score showed no meaningful
changes between any condition for a
total study population. The symptoms
score reported by a subset of
asthmatics/allergies were positively
associated with inhaled concentration
ofH2SO4(p = 0.01).
H2SO4/O3/NO2, O3/NO2 and clean air
produced similar responses
No significant pulmonary function
changes following any exposure
compared to response to clean air.
Six additional subjects started the
study, but dropped out due to
uncomfortable symptoms.
Exposure to O3 or O3 + H2SO4 induced
significant decrements in
forced expiratory function. Differences
between O3 and O3 + H2SO4 were, at
best, marginally significant. O3 is the
more important pollutant for inducing
respiratory effects. A few asthmatic
and nonasthmatic subjects were more
responsive to O3 + H2SO4 than to
O3 alone.
No significant changes in symptoms or
lung function with any aerosol/O3
combination in the healthy group.
In asthmatics, H2SO4 preexposure
enhanced the small decrements in FVC
that occurred following exposure to
0.18ppmO3. Asthmatics had no
significant changes on FEV; with any
O3 exposures, but symptoms were
greater.
Linn et al.
(1997)
Linnetal.
(1995)
Koenig et al.
(1994)
Linn et al.
(1994)
Utell et al.
(1994)
Frampton
etal. (1995)
-------
Table AX6-14 (cont'd). Ozone Mixed with Other Pollutants3
to
o
o
X
I
to
H
6
o
o
H
/O
Concentration11
ppm
Hg/m3
Pollutant
Number and
Exposure Duration Exposure Gender of
and Activity Conditions' Subjects
Subject
Characteristics
Observed Effect(s)
Reference
Sulfur-Containing Pollutants (cont'd)
0.12
0.10
0.25
235
262
490
1,200
to
1,600
03
SO2
03
H2SO4
1 h (mouthpiece) 22 °C 8 M, 5 F
IE VE ~ 30 L/min 75% RH
45-min exposure to air or
O3, followed by 1 5-min
exposure to O3 or SO2
2h 35 °C 9M
IE 83% RH
VE = 30 to 32 L/min
Allergic asthmatics,
12 to 18 years old,
medications withheld
for at least 4 h before
exposures
Healthy NS,
19 to 29 years old
Prior exposure to O3 potentiated
pulmonary function responses to SO2;
decrements in FEVj were -3, -2, and
-8% for the air/O3, O3/O3, and O3/SO2
exposures, respectively.
No significant effects of exposure to O3
alone or combined with H2SO4 aerosol.
Koenig et al.
(1990)
Horvath
etal. (1987)
Nitrogen-Containing Pollutants
0.0
0.2
0.4
0.2+0.4
0.1
0.2
0.1+0.2
0.0
0.36
0.36
0.75
0.36
0.45
0.60
0
392
752
196
376
0
706
677
1,411
943
883
1,129
Air
03
NO2
O3+NO2
03
N02
03+N02
Air
03
NO2
NO2
SO2
03
NO2
3 h 9 M, 2 F
IE 10' ex/ 20' rest
VE =32 L/min
6h
IE 10' ex/ 20' rest
VE =32 L/min
T = 25°C
RH = 50%
2 h Head only 6 M, 6 F
rest exposure
2-h random exposures to 23.6 °C 6 M
FA, O3, NO2, and O3 + 62% RH 2 F
NO2; IE; VE = 26-29
L/min
Atopic asthmatics
22 to 41 yrs. old
Healthy NS
19 to 33 yrs. old
Healthy, NS, 56 to
85 years old
Exposure to NO2 alone had minimal
effects on FEVj. However, O3 alone or
in combination elicited significantly
greater decline in FEVj in a short (3 h)
exposure (higher concentrations) than a
long (6 h) exposure where the effects
were nonsignificant. Allergen
challenge inhalation significantly
reduced PD20 FEV; in all short but not
the long exposures. No additive or
potentiating effects have been
observed.
For NO2 and SO2 the absorbed fraction
of O3 increased relative (to baseline)
whereas after O3 exposure it decreased.
The differences explained by an
increased production of O3-reactive
substrate in ELF due to inflammation.
Exercise-induced cardiac output was
smaller with O3 + NO2 exposure
compared to FA or O3 alone.
Jenkins et al.
(1999)
Rigas et al.
(1997)
Drechsler-
Parks et al.
(1995)
o
HH
H
W
-------
Table AX6-14 (cont'd). Ozone Mixed with Other Pollutants3
to
o
o
X
I
to
H
6
o
o
H
/O
o
HH
H
W
Concentration11
ppm
(ig/m3 Pollutant
Exposure Duration
and Activity
Exposure
Conditions'
Number and
Gender of
Subjects
Subject
Characteristics
Observed Effect(s)
Reference
Nitrogen-Containing Pollutants (cont'd)
0.30
0.60
0.2
0.0
0.12
0.30
0.12+0.30
589 O3
1,129 N02
392 O3
500 HNO3
H20
0 Air
235 03
564 N02
03 + N02
2-h exposure to NO2 or 21 °C
FA, followed 3 h later by 40% RH
2-h exposure to O3, IE
VE = 20 L/min/m2 BSA
5 h 20 °C
IE (50 min/h exercise) 5% RH
VE ~ 40 L/min
2 h HNO3 or H2O fog or
air, followed by 1-h break,
followed by 3 h O3
1 h (mouthpiece) 22 °C
IE 75% RH
VE = 33 L/min
VE = 35 L/min
21 F Healthy NS,
18 to 34 years old
6 M, 4 F Healthy NS,
minimum of 1 0%
decrement in FEV;
after 3 h exposure to
0.20 ppm O3 with
50 min exercise/h
5 M, 7 F Healthy NS,
12 to 17 years old
9 M, 3 F Asthmatic
13 to 18 years old
No significant effect of NO2 exposures
on any measured parameter. Sequential
exposure of NO2 followed by O3
induced small but significantly larger
decrements in FEVj and FEF25.75 than
FA/O3 sequence. Subjects had
increased airway responsiveness to
methacholine after both exposures, with
significantly greater responsiveness
after the NO2/O3 sequences than after
the FA/O3 sequence.
Exposure to HNO3 or H2O fog followed
by O3 induced smaller pulmonary
function decrements than air followed
by03.
Findings inconsistent across cohorts
and atmospheres. No significant
differences in FEV; and RT between
asthmatics and healthy, or between
atmospheres and cohorts.
Hazucha
etal. (1994)
Aris etal.
(1991)
Koenig et al.
(1988)
0.30 589 O3 1 h (mouthpiece)
0.60 1,129 N02 CE
VE ~ 70 L/min for men
VB ~ 50 L/min for women
20 M, 20 F
Healthy NS,
21.4±1.5(SD)
years old for F,
22.7 ± 3.3 (SD) years
old for M
No differences between responses to
O3 and NO2 + O3 for spirometric
parameters. Increase in SRaw with
NO2 + O3 was significantly less than for
O, alone.
Adams etal.
(1987)
0.30
0.30
0.15
0.15
0.15
0.15
0.15
589
564
200
294
284
200
294
282
393
200
03
N02
H2S04
03
NO2
H2SO4
03
N02
S02
H2S04
2 hCE for 20 min 28 to 29 °C 6M
V = 25 L/min 50 to 60% RH
2 h, 60 min 6 M
total exercise
V ~ 25 L/min
2 h, 60 min 3 M
total exercise
V = 25 L/min
Healthy subjects, Possible small decrease in SGaw
some smokers
Possible small decrease in SGaw
Possible small decrease in FEV;
Kagawa
(1986)
-------
3
C
^
to
o
o
>
X
ON
i
to
ON
O
^>
H
1
O
o
0
H
O
xauii; r^^w
Concentration11
Exposure Duration
ppm Hg/in3 Pollutant and Activity
Peroxyacetyl Nitrate
0.45 883 O3 2h
0.60 1,129 N02 IE
0.13 644 PAN VE = 25 L/min
0.45 883 O3 2h
0.30 1,485 PAN IE
VE = 27 L/min
0.485 952 O3 2h
0.27 1,337 PAN IE
VE = 25 L/min
Particle-Containing Pollutants
0.0 0 Air 2-2. 5 h
0.12 235b+ O3+ rest
153b PM25
"See Appendix A for abbreviations and acronyms.
'Grouped by pollutant mixture.
CWBGT = 0.7 Twetbult + 0.3 T^hulhagl
-------
1 et al. (1994), which was a repeated 6.5 h exposure protocol, O3 alone and O3 + H2SO4 induced
2 significant spirometric decrements in healthy adults and asthmatics, but the magnitude of effects
3 between exposure atmospheres was not significant. Asthmatic and atopic subjects showed
4 somewhat enhanced or potentiated response to mixtures or sequential exposure, respectively;
5 however, the observed effects were almost entirely attributable to O3 (U.S. Environmental
6 Protection Agency, 1996). Thus, in both healthy and asthmatic subjects, the interactive effects
7 of O3 and other pollutants were marginal and the response was dominated by O3.
8 Since 1994, the only laboratory study that examined the health effects of a mixture of O3
9 and sulfur oxides (SO2 and H2SO4) has been that of Linn et al. (1997). In this study, the
10 investigators closely simulated ambient summer haze air pollution conditions in Uniontown, PA
11 as well as controlled the selection of study subjects with the objective to corroborate earlier
12 reported findings of an epidemiologic study of Neas et al. (1995). The subjects were 41 children
13 (22F/19M) 9 to 12 yrs old. Of these, 26 children had history of asthma or allergy. During a
14 14-day study period, children were exposed on the 4th and 11th day for 4 hrs (IE, 15 min @ avg.
15 VE 22 L/min) in random order to air and a mixture of 0.10 ppm O3, 0.10 ppm SO2 and 42 to
16 198 mg/m3 H2SO4 (mean cone. 101 mg/m3, 0.6 mm MMAD). The effects of controlled
17 exposures were assessed by spirometry. Except for exposure days, children used diaries to
18 record activity, respiratory symptoms, location, and PEFR. Thus, every exposure day was
19 bracketed by 3 days of monitoring. Spirometry, PEFR, and respiratory symptoms score showed
20 no meaningful changes between any condition for a total study population. The symptoms score
21 reported by a subset of asthmatic/allergic subjects was positively associated with the inhaled
22 concentration of H2SO4 (p = 0.01). However, the reported symptoms were different from the
23 ones reported in the Uniontown study (Neas et al., 1995). Although retrospective statistical
24 power calculations using these study observations for the symptoms score, PEFR, and
25 spirometric endpoints were sufficient to detect with > 80% probability the same magnitude of
26 changes as observed in Uniontown, the effects were minimal and not significant. The divergent
27 observations of the two studies have been explained by the presence of an unidentified
28 environmental factor in Uniontown, differences in physico-chemical properties of acid,
29 differences in time course of exposure and history of previous exposure of children to pollutants,
30 psychological and physiological factors related to chamber exposures, and by other conjectures.
31
January 2005 AX6-127 DRAFT-DO NOT QUOTE OR CITE
-------
1 AX6.11.2 Ozone and Nitrogen-Containing Pollutants
2 Nitrogen dioxide is a key component of the photooxidation cycle and formation of O3.
3 Both gases are almost invariably present in ambient atmosphere. Compared to O3, NOX species
4 have limited solubility and moderate oxidizing capability. Both O3 and NO2 are irritants and
5 tissue oxidants and exert their toxic actions through many common mechanisms. The regional
6 dosimetry and the primary sites of action of O3 and NO2 overlap but are not the same. Since
7 these gases are relatively insoluble in water, they will likely penetrate into the peripheral airways
8 that are more sensitive to damage than better protected conducting airways. The controlled
9 studies reviewed in the previous O3 criteria document (Table AX6-14) generally reported only
10 small pulmonary function changes after combined exposures of NO2 or nitric acid (HNO3) with
11 O3, regardless if the interactive effects were potentiating or additive. In two of these studies, the
12 effects reached statistical significance, but they were not coherent. Preexposure with NO2
13 potentiated both spirometric and nonspecific airway reactivity response following subsequent O3
14 exposure (Hazucha et al., 1994); however, exposure to NO2 + O3 mixture blunted SRaw increase
15 as compared to O3 alone (Adams et al.,1987). As with O3 and SOX mixtures, the effects have
16 been dominated by O3 (U.S. Environmental Protection Agency, 1996).
17 Combined exposure to O3 and NO2 also blunted the exercise-induced increase in cardiac
18 output found with FA and O3 exposures alone (Drechsler-Parks, 1995). Eight healthy older
19 subjects (56 to 85 years of age) were exposed for 2 h to FA, 0.60 ppm NO2, 0.45 ppm O3, and to
20 0.60 ppm NO2 + 0.45 ppm O3 while alternating 20-min periods of rest and exercise. Cardiac
21 output, HR, stroke volume, and systolic time intervals were measured by noninvasive impedance
22 cardiography at the beginning of each exposure, while the subjects were at rest, and again during
23 the last 5 min of exercise. Metabolic exercise data (VE, VO2, fB) also were measured. There
24 were no statistically significant differences between exposures for FIR, VE, VO2, fB, stroke
25 volume, or systolic time intervals. Exercise increased cardiac output after all exposures;
26 however, the incremental increase over rest was significantly smaller for the combined O3 and
27 NO2 exposures. The authors speculated that nitrate and nitrite reaction products from the
28 interaction of O3 and NO2 cross the air/blood interface in the lungs, causing peripheral
29 vasodilation and a subsequent drop in cardiac output. No major cardiovascular effects of O3
30 only exposures have been reported in human subjects (see Section AX6.10).
January 2005 AX6-128 DRAFT-DO NOT QUOTE OR CITE
-------
1 Despite suggested potentiation of O3 response by NO2 in healthy subjects, it is unclear
2 what response, and at what dose, either sequential or combined gas exposures will induce in
3 asthmatics. Jenkins et al. (1999) exposed 11 atopic asthmatics in random order to air, 0.1 ppm
4 O3, 0.2 ppm NO2, and 0.1 ppm O3 + 0.2 ppm NO2 for 6 h (IE for 10 min @ 32 L/min every
5 40 min). Two weeks later, 10 of these subjects were exposed for 3 h to doubled concentrations
6 of these gases (i.e., 0.2 ppm O3, 0.4 ppm NO2, and 0.2 ppm O3 + 0.4 ppm NO2) employing the
7 same exercise regimen. Immediately following each exposure, subjects were challenged with
8 allergen (D. pteronyssinus) and PD20 FEVj was determined. Exposure to NO2 alone had
9 minimal effects on FEVj or airway responsiveness. However, O3 alone or in combination with
10 NO2 elicited a significantly (p < 0.05) greater decline in FEVj in a short (3 h) exposure (higher
11 concentrations) than the long (6 h) exposure, where the effects were not significant. Allergen
12 challenge inhalation significantly (p = 0.018 to 0.002) reduced PD20 FEVj in all short, but not the
13 long, exposures. No associations were observed between pollutant concentrations and
14 physiologic endpoints. The statistical analyses of these data suggest that the combined effect
15 (O3 + NO2) on lung function (FVC, FEVj) was not significantly greater than the effect of
16 individual gases for 6-h exposures, thus no additive or potentiating effects have been observed.
17 Shorter 3-h exposures using twice as high NO2 concentrations, however, showed significant
18 FEVj decrements following exposures to atmospheres containing O3. The analysis also suggests
19 that it is the inhaled concentration, rather than total dose, that determines lung airway
20 responsiveness to allergen.
21 The potential for interaction between O3 and other gas mixtures was studied by Rigas et al.
22 (1997). They used an O3 bolus absorption technique to determine how exposures to O3, NO2,
23 and SO2 will affect distribution of O3 adsorption by airway mucosa. The selected O3 bolus
24 volume was set to reach lower conducting airways. Healthy young nonsmokers (6F/6M) were
25 exposed on separate days at rest in a head dome to 0.36 ppm O3, 0.36 ppm NO2, 0.75 ppm NO2
26 and 0.75 ppm SO2 for 2 h. The rationale for the selection of these gases was their differential
27 absorption. Because O3 and NO2 are much less soluble in liquid (i.e., ELF) than SO2, they are
28 expected to penetrate deeper into the lung than SO2 which is absorbed more quickly in the
29 epithelial lining fluid of the upper airways. The actual experimental measurements have shown
30 that during continuous NO2 and SO2 exposure the absorbed fraction of an O3 bolus in lower
31 conducting airways increased relative to baseline, whereas during continuous O3 exposure the O3
January 2005 AX6-129 DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
bolus fraction in lower conducting airways decreased. The authors attempted to explain the
differences by suggesting that there may be increased production of an O3-reactive substrate in
epithelial lining fluid due to airway inflammation. As interpreted by the investigators, during
NO2 and SO2 exposures the substrate was not depleted by these gases and so could react with the
O3 bolus, whereas during O3 exposure the substrate was depleted, causing the fractional
absorption of the O3 bolus to decrease. Greater absorption in males than females for all gases
was attributed to anatomical differences in the bronchial tree.
9 AX6.11.3 Ozone and Other Pollutant Mixtures Including Particulate Matter
10 Almost all of the studies published over the last twenty years investigating the health
11 effects of mixtures of O3 with other air pollutants involved peroxyacetyl nitrate (PAN). These
12 studies on healthy individuals exposed under laboratory conditions came from the Horvath
13 laboratory at UC Santa Barbara (Table AX6-13). In the last of this series of studies, Drechsler-
14 Parks and colleagues (1989) found the same equivocal interaction of O3 and PAN as in previous
15 studies, which is attributable to O3 exposure alone (U.S. Environmental Protection Agency,
16 1996). Subsequently, only a couple of studies have investigated the effects of more complex air
17 pollutant mixtures on human pathophysiology under controlled conditions.
18 It is not only the interaction between air pollutants in ambient air; but, as Rigas et al.
19 (1997) has found, an uneven distribution of O3, SO2, and NO2 absorption in the lower conducting
20 airways of young healthy subjects may modulate pathophysiologic response as well. Exposure
21 to SO2 and NO2 increased, while exposure to O3 decreased, the absorbing capacity of the airways
22 for O3. The authors have suggested that SO2 or NO2 -inflamed airways release additional
23 substrates into the epithelial lining fluid that react with O3, thus progressively removing O3 from
24 the airway lumen. This mechanism may explain findings of antagonistic response (e.g., Adams
25 et al., 1987; Dreschler-Parks, 1995) when the two gases are combined in an exposure
26 atmosphere.
27 The mechanisms by which inhalation exposure to other complex ambient atmospheres
28 containing particulate matter (PM) and O3 induce cardiac events frequently reported in
29 epidemiologic studies are rarely studied in human subjects under laboratory conditions.
30 Recently, Brook et al. (2002) have reported changes in brachial artery tone and reactivity in
31 healthy nonsmokers following 2-h exposures to a mixture of 0.12 ppm O3 and 153 |ig/m3 of
January 2005 AX6-130 DRAFT-DO NOT QUOTE OR CITE
-------
1 concentrated ambient PM25, and a control atmosphere of filtered air with a trace of O3
2 administered in random order. Neither systolic nor diastolic pressure was affected by pollutant
3 exposure despite a significant brachial artery constriction and a reduction in arterial diameter
4 when compared to filtered air (p = 0.03). The authors postulate that changes in arterial tone may
5 be a plausible mechanism of air pollution-induced cardiac events. However, the observations of
6 no changes in blood pressure, and an absence of flow- and nitroglycerin- mediated brachial
7 artery dilatation, cast some doubt on the plausibility of this mechanism. A number of other
8 proposed mechanisms advanced to establish a link between cardiac events due to pollution and
9 changes in vasomotor tone based on the findings of this study are purely speculative.
10
11
12 AX6.12 CONTROLLED STUDIES OF AMBIENT AIR EXPOSURES
13 A large amount of informative O3 exposure-effects data has been obtained in controlled
14 laboratory exposure studies under a variety of different experimental conditions. However,
15 laboratory simulation of the variable pollutant mixtures present in ambient air is not practical.
16 Thus, the exposure effects of one or several artificially generated pollutants (i.e., a simple
17 mixture) on pulmonary function and symptoms may not explain responses to ambient air where
18 complex pollutant mixtures exist. Epidemiologic studies, which do investigate ambient air
19 exposures, do not typically provide the level of control and monitoring necessary to adequately
20 characterize short term responses. Thus, controlled exposures to ambient air using limited
21 numbers of volunteers have been used to try and bridge the gap between laboratory and
22 community exposures.
23
24 AX6.12.1 Mobile Laboratory Studies
25 As presented in previous criteria documents (U.S. Environmental Protection Agency, 1986;
26 1996), quantitatively useful information on the effects of acute exposure to photochemical
27 oxidants on pulmonary function and symptoms responses originated from field studies using a
28 mobile laboratory. These field studies involved subjects exposed to ambient air, FA without
29 pollutants, or FA containing artificially generated concentrations of O3 that are comparable to
30 those measured in the ambient environment. As a result, measured pulmonary responses in
31 ambient air can be directly compared to those found in more artificial or controlled conditions.
January 2005 AX6-131 DRAFT-DO NOT QUOTE OR CITE
-------
1 However, the mobile laboratory shares some of the same limitations of stationary exposure
2 laboratories (e.g., limited number of both subjects and artificially generated pollutants for
3 testing). Further, mobile laboratory ambient air studies are dependent on ambient outdoor
4 conditions which can be unpredictable, uncontrollable, and not completely characterizable.
5 As summarized in Table AX6-15, investigators in California used a mobile laboratory and
6 demonstrated that pulmonary effects of ambient air in Los Angeles residents are related to O3
7 concentration and level of exercise (Avol et al., 1983, 1984, 1985a,b,c, 1987; Linn et al., 1980,
8 1983). Avol et al. (1987) observed no significant pulmonary function or symptoms responses in
9 children (8 to 11 years) engaged in moderate continuous exercise for 1 h while breathing
10 ambient air with an O3 concentration of 0.113 ppm. However, significant pulmonary function
11 decrements and increased symptoms of breathing discomfort were observed in healthy
12 exercising (1 h continuous) adolescents (Avol et al., 1985a,b), athletes, (Avol et al., 1984, 1985c)
13 and lightly exercising asthmatic subjects (Linn et al., 1980, 1983) at O3 concentrations averaging
14 from 0.144 to 0.174 ppm. Many of the healthy subjects with a history of allergy appeared to be
15 more responsive to O3 than "nonallergic" subjects (Linn et al., 1980, 1983), although a
16 standardized evaluation of atopic status was not performed. Comparative studies of exercising
17 athletes (Avol et al., 1984, 1985c) with chamber exposures to oxidant-polluted ambient air
18 (mean O3 concentration of 0.153 ppm) and purified air containing a controlled concentration of
19 generated O3 at 0.16 ppm showed similar pulmonary function responses and symptoms,
20 suggesting that acute exposures to coexisting ambient pollutants had minimal contribution to
21 these responses under the typical summer ambient conditions in Southern California. This
22 contention is similar to, but extends, the laboratory finding of no significant difference in
23 pulmonary function effects between O3 and O3 plus PAN exposures (Drechsler-Parks, 1987b).
24 Additional supporting evidence is provided in Section AX6.11.
25
26 AX6.12.2 Aircraft Cabin Studies
27 Respiratory symptoms and pulmonary function effects resulting from exposure to O3 in
28 commercial aircraft flying at high altitudes, and in altitude-simulation studies, have been
29 reviewed elsewhere (U.S. Environmental Protection Agency, 1986, 1996). Flight attendants,
30 because of their physical activities at altitude, tend to receive higher exposures. In a series of
31 hypobaric chamber studies of nonsmoking subjects exposed to 1,829 m (6,000 ft) and O3 at
January 2005 AX6-132 DRAFT-DO NOT QUOTE OR CITE
-------
fa
to
O
o
X
ON
OJ
OJ
O
3>
'•Tj
H
1
O
o
2|
0
h-j
o
Table AX6-15. Acute Effects of Ozone in Ambient Air in Field S
Mean Ozone
Concentration" Ambient
ppm MS/m3 (°C) Duration (VE) Number of Subjects
0.1 13. ±.033 221 ±65 33 ± 1 Ih CE (22 L/min) 66 healthy children,
8 to 1 1 years old
0.144 ±.043 282 ±84 32 ± 1 Ih CE (32 L/min) 59 healthy
adolescents,
12 to 15 years old
0.153 ±.025 300 ±49 32 ±2 Ih CE (53 L/min) 50 healthy adults
(competitive
bicyclists)
0.156 ±.055 306 ±107 33 ±4 Ih CE (38 L/min) 48 healthy adults,
50 asthmatic adults
0.165 ±.059 323 ±115 33 ± 3 Ih CE (42 L/min) 60 "healthy" adults
(7 were asthmatic)
0.174 ±.068 341 ±133 33 ±2 2h IE (2 times resting) 34 "healthy" adults,
at 15-min intervals 30 asthmatic adults
studies With a Mobile Laboratory"
Observed Effect(s)
No significant changes in forced expiratory
function and symptoms of breathing
discomfort after exposure to 0. 1 1 3 ppm O3 in
ambient air.
Small significant decreases in FVC (-2.1%),
FEV075 (-4.0%), FEV; (-4.2%), andPEFR
(-4.4%) relative to control with no recovery
during a 1-h postexposure rest; no significant
increases in symptoms.
Mild increases in symptoms scores and
significant decreases in FEV; (-5.3%) and
FVC; mean changes in ambient air were not
statistically different from those in purified air
containing 0.16 ppm O3.
No significant changes for total symptom
score or forced expiratory performance in
normals or asthmatics; however, FEV;
remained low or decreased further (-3%) 3 h
after ambient air exposure in asthmatics.
Small significant decreases in FEV[ (-3.3%)
and FVC with no recovery during a 1-h
postexposure rest; TLC decreased and AN2
increased slightly.
Increased symptom scores and small
significant decreases in FEV[ (-2.4%), FVC,
PEFR, and TLC in both asthmatic and healthy
subjects; however, 25/34 healthy subjects were
allergic and "atypically" reactive to polluted
ambient air.
Reference
Avol et al.
(1987)
Avol et al.
(1985a,b)
Avol et al.
(1984, 1985c)
Linn et al.
(1983)
Avol et al.
(1983)
Linn et al.
(1983)
Avol et al.
(1983)
Linn et al.
(1980, 1983)
O
HH
H
W
aSee Appendix A for abbreviations and acronyms.
bRanked by lowest level of O3 in ambient air, presented as the mean ± SD.
cMean±SD.
-------
1 concentrations of 0.2 and 0.3 ppm for 3 or 4 h (Lategola et al., 1980a,b), increased symptoms
2 and pulmonary function decrements occurred at 0.3 ppm but not at 0.2 ppm.
3 Commercial aircraft cabin O3 levels were reported to be very low (average concentration
4 0.01 to 0.02 ppm) during 92 randomly selected smoking and nonsmoking flights in 1989 (Nagda
5 et al., 1989). None of these flights recorded O3 concentrations exceeding the 3-h time-weighted
6 average (TWA) standard of 0.10 ppm promulgated by the Federal Aviation Administration
7 (FAA, 1980), probably due to the use of O3-scrubbing catalytic filters (Melton, 1990). However,
8 in-flight O3 exposure can still occur because catalytic filters are not necessarily in continuous use
9 during flight. Other factors to consider in aircraft cabins, however, are erratic temperature
10 changes, lower barometric pressure and oxygen pressure, and lower humidity, often reaching
11 levels between 4 and 17% (Rayman, 2002).
12 Ozone contamination aboard high-altitude aircraft also has been an interest to the U.S. Air
13 Force because of complaints by crew members of frequent symptoms of dryness and irritation of
14 the eyes, nose, and throat and an occasional cough (Hetrick et al., 2000). Despite the lack of
15 ventilation system modifications as used in commercial aircraft, the O3 concentrations never
16 exceeded the FAA ceiling limit of 0.25 ppm and exceeded the 3-h TWA of 0.10 ppm only 7% of
17 the total monitored flight time (43 h). The authors concluded that extremely low average
18 relative humidity (12%) during flight operations was most likely responsible for the reported
19 symptoms.
20
21
January 2005 AX6-134 DRAFT-DO NOT QUOTE OR CITE
-------
1 REFERENCES
2 Adams, W. C. (2000a) Feasibility study of prolonged ozone inhalation exposure via face mask. Inhalation Toxicol.
3 12:299-313.
4 Adams, W. C. (2000b) Ozone dose-response effects of varied equivalent minute ventilation rates. J. Exposure Anal.
5 Environ. Epidemiol. 10: 217-226.
6 Adams, W. C. (2002) Comparison of chamber and face-mask 6.6-hour exposures to ozone on pulmonary function
7 and symptoms responses. Inhalation Toxicol. 14: 745-764.
8 Adams, W. C. (2003a) Comparison of chamber and face mask 6.6-hour exposure to 0.08 ppm ozone via
9 square-wave and triangular profiles on pulmonary responses. Inhalation Toxicol. 15: 265-281.
10 Adams, W. C. (2003b) Relation of pulmonary responses induced by 6.6-h exposures to 0.08 ppm ozone and 2-h
11 exposures to 0.30 ppm ozone via chamber and face-mask inhalation. Inhalation Toxicol. 15: 745-759.
12 Adams, W. C.; Ollison, W. M. (1997) Effects of prolonged simulated ambient ozone dosing patterns on human
13 pulmonary function and symptomatology. Presented at: 90th annual meeting of the Air & Waste Management
14 Association; June; Toronto, Ontario, Canada. Pittsburgh, PA: Air & Waste Management Association; paper
15 no. 97-MP9.02.
16 Adams, W. C.; Schelegle, E. S. (1983) Ozone and high ventilation effects on pulmonary function and endurance
17 performance. J. Appl. Physiol.: Respir. Environ. Exercise Physiol. 55: 805-812.
18 Adams, W. C.; Savin, W. M.; Christo, A. E. (1981) Detection of ozone toxicity during continuous exercise via the
19 effective dose concept. J. Appl. Physiol.: Respir. Environ. Exercise Physiol. 51: 415-422.
20 Adams, W. C.; Brookes, K. A.; Schelegle, E. S. (1987) Effects of NO2 alone and in combination with O3 on young
21 men and women. J. Appl. Physiol. 62: 1698-1704.
22 Alexis, N.; Urch, B.; Tarlo, S.; Corey, P.; Pengelly, D.; O'Byrne, P.; Silverman, F. (2000) Cyclooxygenase
23 metabolites play a different role in ozone-induced pulmonary function decline in asthmatics compared to
24 normals. Inhalation Toxicol. 12: 1205-1224.
25 Alexis, N. E.; Becker, S.; Bromberg, P. A.; Devlin, R.; Peden, D. B. (2004) Circulating GDI Ib expression correlates
26 with the neutrophil response and airway mCD14 expression is enhanced following ozone exposure in
27 humans. Clin. Immunol. Ill: 126-131.
28 Alpert, S. E.; Walenga, R. W.; Jaspers, I.; Qu, Q.; Chen, L. C. (1997) Ozone inactivates Cyclooxygenase in human
29 tracheal epithelial cells without altering PGHS-2 mRNA or protein. Am. J. Physiol. 272: L879-L887.
30 Arab, L.; Steck-Scott, S.; Fleishauer, A. T. (2002) Lycopene and the lung. Exp. Biol. Med. 227: 894-899.
31 Aris, R.; Christian, D.; Sheppard, D.; Balmes, J. R. (1991) The effects of sequential exposure to acidic fog and
32 ozone on pulmonary function in exercising subjects. Am. Rev. Respir. Dis. 143: 85-91.
33 Aris, R. M.; Christian, D.; Hearne, P. Q.; Kerr, K.; Finkbeiner, W. E.; Balmes, J. R. (1993) Ozone-induced airway
34 inflammation in human subjects as determined by airway lavage and biopsy. Am. Rev. Respir. Dis. 148:
35 1363-1372.
36 Aris, R. M.; Tager, L; Christian, D.; Kelly, T.; Balmes, J. R. (1995) Methacholine responsiveness is not associated
37 with O3-induced decreases inFEVj. Chest 107: 621-628.
38 Avissar, N. E.; Reed, C. K.; Cox, C.; Frampton, M. W.; Finkelstein, J. N. (2000) Ozone, but not nitrogen dioxide,
39 exposure decreases glutathione peroxidases in epithelial lining fluid of human lung. Am. J. Respir. Crit. Care
40 Med. 162: 1342-1347.
41 Avol, E. L.; Linn, W. S.; Shamoo, D. A.; Venet, T. G.; Hackney, J. D. (1983) Acute respiratory effects of
42 Los Angeles smog in continuously exercising adults. J. Air Pollut. Control Assoc. 33: 1055-1060.
43 Avol, E. L.; Linn, W. S.; Venet, T. G.; Shamoo, D. A.; Hackney, J. D. (1984) Comparative respiratory effects of
44 ozone and ambient oxidant pollution exposure during heavy exercise. J. Air Pollut. Control Assoc.
45 34:804-809.
46 Avol, E. L.; Linn, W. S.; Shamoo, D. A.; Valencia, L. M.; Anzar, U. T.; Venet, T. G.; Hackney, J. D. (1985a)
47 Respiratory effects of photochemical oxidant air pollution in exercising adolescents. Am. Rev. Respir. Dis.
48 132:619-622.
49 Avol, E. L.; Linn, W. S.; Shamoo, D. A.; Valencia, L. M.; Anzar, U. T.; Hackney, J. D. (1985b) Short-term health
50 effects of ambient air pollution in adolescents. In: Lee, S. D., ed. Evaluation of the scientific basis for
51 ozone/oxidants standards: transactions of an APCA international specialty conference; November 1984;
52 Houston, TX. Pittsburgh, PA: Air Pollution Control Association; pp. 329-336. (APCA international specialty
53 conference transactions: TR-4).
January 2005 AX6-13 5 DRAFT-DO NOT QUOTE OR CITE
-------
1 Avol, E. L.; Linn, W. S.; Venet, T. G.; Shamoo, D. A.; Spier, C. E.; Hackney, J. D. (1985c) Comparative effects of
2 laboratory generated ozone and ambient oxidant exposure in continuously exercising subjects. In: Lee, S. D.,
3 ed. Evaluation of the scientific basis for ozone/oxidants standards: proceedings of an APCA international
4 specialty conference; November 1984; Houston, TX. Pittsburgh, PA: Air Pollution Control Association; pp.
5 216-225. (APCA international specialty conference transactions: TR-4).
6 Avol, E. L.; Linn, W. S.; Shamoo, D. A.; Spier, C. E.; Valencia, L. M; Venet, T. G.; Trim, S. C.; Hackney, J. D.
7 (1987) Short-term respiratory effects of photochemical oxidant exposure in exercising children. JAPCA
8 37: 158-162.
9 Avol, E. L.; Linn, W. S.; Shamoo, D. A.; Hackney, J. D. (1988) Seasonal ozone reactivity in Los Angeles residents.
10 Presented at: 81st annual meeting of the Air Pollution Control Association; June; Dallas, TX. Pittsburgh, PA:
11 Air Pollution Control Association; paper no. 88-122.6.
12 Ball, B. A.; Folinsbee, L. J.; Peden, D. B.; Kehrl, H. R. (1996) Allergen bronchoprovocation of patients with mild
13 allergic asthma after ozone exposure. J. Allergy Clin. Immunol. 98: 563-572.
14 Balmes, J. R.; Chen, L. L.; Scannell, C.; Tager, I.; Christian, D.; Hearne, P. Q.; Kelly, T.; Aris, R. M. (1996)
15 Ozone-induced decrements in FEVj and FVC do not correlate with measures of inflammation. Am. J. Respir.
16 Crit. Care Med. 153: 904-909.
17 Balmes, J. R.; Aris, R. M.; Chen, L. L.; Scannell, C.; Tager, I. B.; Finkbeiner, W.; Christian, D.; Kelly, T.; Hearne,
18 P. Q.; Ferrando, R.; Welch, B. (1997) Effects of ozone on normal and potentially sensitive human subjects.
19 part I: airway inflammation and responsiveness to ozone in normal and asthmatic subjects. Cambridge, MA:
20 Health Effects Institute. Research report no. 78; pp 1-37, 81-99.
21 Bascom, R.; Naclerio, R. M.; Fitzgerald, T. K.; Kagey-Sobotka, A.; Proud, D. (1990) Effect of ozone inhalation on
22 the response to nasal challenge with antigen of allergic subjects. Am. Rev. Respir. Dis. 142: 594-601.
23 Basha, M. A.; Gross, K. B.; Gwizdala, C. J.; Haidar, A. H.; Popovich, J., Jr. (1994) Bronchoalveolar lavage
24 neutrophilia in asthmatic and healthy volunteers after controlled exposure to ozone and filtered purified air.
25 Chest 106: 1757-1765.
26 Bates, D. V; Bell, G. M.; Burnham, C. D.; Hazucha, M.; Mantha, J.; Pengelly, L. D.; Silverman, F. (1972)
27 Short-term effects of ozone on the lung. J. Appl. Physiol. 32: 176-181.
28 Bayram, H.; Sapsford, R. J.; Abdelaziz, M. M.; Khair, O. A. (2001) Effect of ozone and nitrogen dioxide on the
29 release of proinflammatory mediators from bronchial epithelial cells of nonatopic nonasthmatic subjects and
30 atopic asthmatic patients in vitro. J. Allergy Clin. Immunol. 107: 287-294.
31 Bayram, H.; Rusznak, C.; Khair, O. A.; Sapsford, R. J.; Abdelaziz, M. M. (2002) Effect of ozone and nitrogen
32 dioxide on the permeability of bronchial epithelial cell cultures of non-asthmatic and asthmatic subjects.
33 Clin. Exp. Allergy 32: 1285-1292.
34 Beck, N.; Koenig, J. Q.; Luchtel, D. L.; Altaian, L. C.; Orsborn, M. T.; Kenney, J. S. (1994) Ozone can increase the
3 5 expression of intercellular adhesion molecule-1 and the synthesis of cytokines by human nasal epithelial cells.
36 InhalationToxicol. 6: 345-357.
37 Becker, S.; Madden, M. C.; Newman, S. L.; Devlin, R. B.; Koren, H. S. (1991) Modulation of human alveolar
38 macrophage properties by ozone exposure in vitro. Toxicol. Appl. Pharmacol. 110: 403-415.
39 Bedi, J. F.; Drechsler-Parks, D. M.; Horvath, S. M. (1985) Duration of increased pulmonary function sensitivity to
40 an initial ozone exposure. Am. Ind. Hyg. Assoc. J. 46: 731-734.
41 Bedi, J. F.; Horvath, S. M.; Drechsler-Parks, D. M. (1988) Reproducibility of the pulmonary function response of
42 older men and women to a 2-hour ozone exposure. JAPCA 38: 1016-1019.
43 Bedi, J. F.; Horvath, S. M.; Drechsler-Parks, D. M. (1989) Adaptation by older individuals repeatedly exposed to
44 0.45 parts per million ozone for two hours. JAPCA 39: 194-199.
45 Bergamaschi, E.; De Palma, G.; Mozzoni, P.; Vanni, S.; Vettori, M. V.; Broeckaert, F.; Bernard, A.; Mutti, A.
46 (2001) Polymorphism of quinone-metabolizing enzymes and susceptibility to ozone-induced acute effects.
47 Am. J. Respir. Crit. Care Med. 163: 1426-1431.
48 Bhalla, D. K. (2002) Interactive effects of cigarette smoke and ozone in the induction of lung injury. Toxicol. Sci.
49 65: 1-3.
50 Blomberg, A.; Helleday, R.; Pourazar, J.; Stenfors, N.; Kelly, F. J.; Frew, A. J.; Holgate, S. T.; Sandstrom, T. (1997)
51 Early airway and peripheral blood cell responses to 0.20 ppm ozone in healthy human subjects. Eur. Respir. J.
52 10(suppl 25): 274S.
53 Blomberg, A.; Mudway, I. S.; Nordenhall, C.; Hedenstrom, H.; Kelly, F. J.; Frew, A. J.; Holgate, S. T.;
54 Sandstrom, T. (1999) Ozone-induced lung function decrements do not correlate with early airway
55 inflammatory or antioxidant responses. Eur. Respir. J. 13: 1418-1428.
January 2005 AX6-136 DRAFT-DO NOT QUOTE OR CITE
-------
1 Blomberg, A.; Mudway, I.; Svensson, M; Hagenbjork-Gustafsson, A.; Thomasson, L.; Helleday, R.; Dumont, X.;
2 Forsberg, B.; Nordberg, G.; Bernard, A. (2003) Clara cell protein as a biomarker for ozone-induced lung
3 injury in humans. Eur. Respir. J. 22: 883-888.
4 Bosson, J.; Stenfors, N.; Bucht, A.; Helleday, R.; Pourazar, J.; Holgate, S. T.; Kelly, f. J.; Sandstrom, T.; Wilson, S.;
5 Frew, A. J.; Blomberg, A. (2003) Ozone-induced bronchial epithelial cytokine expression differs between
6 healthy and asthmatic subjects. Clin. Exp. Allergy 33: 777-782.
7 Brook, R. D.; Brook, J. R.; Urch, B.; Vincent, R.; Rajagopalan, S.; Silverman, F. (2002) Inhalation of fine paniculate
8 air pollution and ozone causes acute arterial vasoconstriction in healthy adults. Circulation 105: 1534-1536.
9 Brookes, K. A.; Adams, W. C.; Schelegle, E. S. (1989) 0.35 ppm O3 exposure induces hyperresponsiveness on 24-h
10 reexposure to 0.20 ppm O3. J. Appl. Physiol. 66: 2756-2762.
11 Brummer, M. E. G.; Schwartz, L. W.; McQuillen, N. K. (1977) A quantitative study of lung damage by scanning
12 electron microscopy. Inflammatory cell response to high-ambient levels of ozone. In: Johari, O.; Becker,
13 R. P., eds. Scanning electron microscopy/1977/11 - biological applications of the SEM: proceedings of the
14 workshops on advances inbiomedical applications of the SEM & STEM; March-April; Chicago, IL. Chicago,
15 IL: IIT Research Institute; pp. 513-518.
16 Bush, M. L.; Asplund, P. T.; Miles, K. A.; Ben-Jebria, A.; Ultman, J. S. (1996) Longitudinal distribution of O3
17 absorption in the lung: gender differences and intersubject variability. J. Appl. Physiol. 81: 1651-1657.
18 Centers for Disease Control and Prevention. (2003) Key clinical activities for quality asthma care: recommendations
19 of the National Asthma Education and Prevention Program. Morb. Mortal. Wkly. Rep. 52(RR-6): 1-8.
20 Available: http://www.cdc.gov/mmwr/PDF/rr/rr5206.pdf (11 April 2003).
21 Chen, L. L.; Tager, I. B.; Peden, D. B.; Christian, D. L.; Ferrando, R. E.; Welch, B. S.; Balmes, J. R. (2004) Effect
22 of ozone exposure on airway responses to inhaled allergen in asthmatic subjects. Chest 125: 2328-2335.
23 Christian, D. L.; Chen, L. L.; Scannell, C. H.; Ferrando, R. E.; Welch, B. S.; Balmes, J. R. (1998) Ozone-induced
24 inflammation is attenuated with multiday exposure. Am. J. Respir. Crit. Care Med. 158: 532-537.
25 Coffey, M. J.; Wheeler, C. S.; Gross, K. B.; Eschenbacher, W. L.; Sporn, P. H. S.; Peters-Golden, M. (1996)
26 Increased 5-lipoxygenase metabolism in the lungs of human subjects exposed to ozone. Toxicology
27 114: 187-197.
28 Coleridge, J. C. G.; Coleridge, H. M.; Schelegle, E. S.; Green, J. F. (1993) Acute inhalation of ozone stimulates
29 bronchial C-fibers and rapidly adapting receptors in dogs. J. Appl. Physiol. 74: 2345-2352.
30 Collins, D. V.; Cutillo, A. G.; Armstrong, J. D.; Crapo, R. O.; Kanner, R. E.; Tocino, I.; Renzetti, A. D., Jr. (1986)
31 Large airway size, lung size, and maximal expiratory flow in healthy nonsmokers. Am. Res. Respir. Dis.
32 134:951-955.
33 Corradi, M.; Alinovi, R.; Goldoni, M.; Vettori, M.; Folesani, G.; Mozzoni, P.; Cavazzini, S.; Bergamaschi, E.; Rossi,
34 L.; Mutti, A. (2002) Biomarkers of oxidative stress after controlled human exposure to ozone. Toxicol. Lett.
35 134:219-225.
36 Criqui, G. I.; Solomon, C.; Welch, B. S.; Ferrando, R. E.; Boushey, H. A.; Balmes, J. R. (2000) Effects of
37 azithromycin on ozone-induced airway neutrophilia and cytokine release. Eur. Respir. J. 15: 856-862.
38 Cropp, G. J. A.; Bernstein, I. L.; Boushey, H. A., Jr.; Hyde, R. W.; Rosenthal, R. R.; Specter, S. L.; Townley, R. G.
39 (1980) Guidelines for bronchial inhalation challenges with pharmacologic and antigenic agents. ATS News
40 (Spring): 11-19.
41 Devlin, R. B.; Koren, H. S. (1990) The use of quantitative two-dimensional gel electrophoresis to analyze changes in
42 alveolar macrophage proteins in humans exposed to ozone. Am. J. Respir. Cell Mol. Biol. 2: 281-288.
43 Devlin, R. B.; McDonnell, W. F.; Koren, H. S.; Becker, S. (1990) Prolonged exposure of humans to 0.10 and
44 0.08 ppm ozone results in inflammation in the lung. Presented at: 83rd annual meeting of the Air & Waste
45 Management Association; June; Pittsburgh, PA. Pittsburgh, PA: Air & Waste Management Association;
46 paper no. 90-150.2.
47 Devlin, R. B.; McDonnell, W. F.; Mann, R.; Becker, S.; House, D. E.; Schreinemachers, D.; Koren, H. S. (1991)
48 Exposure of humans to ambient levels of ozone for 6.6 hours causes cellular and biochemical changes in the
49 lung. Am. J. Respir. Cell Mol. Biol. 4: 72-81.
50 Devlin, R. B.; McKinnon, K. P.; Noah, T.; Becker, S.; Koren, H. S. (1994) Ozone-induced release of cytokines and
51 fibronectinby alveolar macrophages and airway epithelial cells. Am. J. Physiol. 266: L612-L619.
52 Devlin, R. B.; McDonnell, W. F.; Becker, S.; Madden, M. C.; McGee, M. P.; Perez, R.; Hatch, G.; House, D. E.;
53 Koren, H. S. (1996) Time-dependent changes of inflammatory mediators in the lungs of humans exposed to
54 0.4 ppm ozone for 2 hr: a comparison of mediators found in bronchoalveolar lavage fluid 1 and 18 hr after
55 exposure. Toxicol. Appl. Pharmacol. 138: 176-185.
January 2005 AX6-137 DRAFT-DO NOT QUOTE OR CITE
-------
1 Devlin, R. B.; Folinsbee, L. I; Biscardi, F.; Hatch, G.; Becker, S.; Madden, M. C.; Robbins, M.; Koren, H. S. (1997)
2 Inflammation and cell damage induced by repeated exposure of humans to ozone. Inhalation Toxicol.
3 9:211-235.
4 Dimeo, M. I; Glenn, M. G.; Holtzman, M. I; Sheller, J. R.; Nadel, J. A.; Boushey, H. A. (1981) Threshold
5 concentration of ozone causing an increase in bronchial reactivity in humans and adaptation with repeated
6 exposures. Am. Rev. Respir. Dis. 124: 245-248.
7 Drechsler-Parks, D. M. (1995) Cardiac output effects of O3 and NO2 exposure in healthy older adults. Toxicol. Ind.
8 Health 11: 99-109.
9 Drechsler-Parks, D. M.; Bedi, J. F.; Horvath, S. M. (1984) Interaction of peroxyacetyl nitrate and ozone on
10 pulmonary functions. Am. Rev. Respir. Dis. 130: 1033-1037.
11 Drechsler-Parks, D. M.; Bedi, J. F.; Horvath, S. M. (1987a) Pulmonary function responses of older men and women
12 to ozone exposure. Exp. Gerontol. 22: 91-101.
13 Drechsler-Parks, D. M.; Bedi, J. F.; Horvath, S. M. (1987b) Pulmonary function desensitization on repeated
14 exposures to the combination of peroxyacetyl nitrate and ozone. JAPCA 37: 1199-1201.
15 Drechsler-Parks, D. M.; Bedi, J. F.; Horvath, S. M. (1989) Pulmonary function responses of young and older adults
16 to mixtures of O3, NO2 and PAN. Toxicol. Ind. Health 5: 505-517.
17 Drechsler-Parks, D. M.; Horvath, S. M.; Bedi, J. F. (1990) The "effective dose" concept in older adults exposed to
18 ozone. Exp. Gerontol. 25: 107-115.
19 Eschenbacher, W. L.; Ying, R. L.; Kreit, J. W.; Gross, K. B. (1989) Ozone-induced lung function changes in normal
20 and asthmatic subjects and the effect of indomethacin. In: Schneider, T.; Lee, S. D.; Wolters, G. J. R.;
21 Grant, L. D., eds. Atmospheric ozone research and its policy implications: proceedings of the 3rd US-Dutch
22 international symposium; May 1988; Nijmegen, The Netherlands. Amsterdam, The Netherlands: Elsevier
23 Science Publishers; pp. 493-499. (Studies in environmental science 35).
24 Fahy, J. V.; Wong, H. H.; Liu, J. T.; Boushey, H. A. (1995) Analysis of induced sputum after air and ozone
25 exposures in healthy subjects. Environ. Res. 70: 77-83.
26 Federal Aviation Administration. (1980) Airplane cabin ozone contamination. F. R. (January 21) 45: 3880-3883.
27 Fernandes, A. L. G.; Molfino, N. A.; McClean, P. A.; Silverman, F.; Tarlo, S.; Raizenne, M.; Slutsky, A. S.;
28 Zamel, N. (1994) The effect of pre-exposure to 0.12 ppm of ozone on exercise-induced asthma. Chest
29 106: 1077-1082.
30 Folinsbee, L. J.; Hazucha, M. J. (1989) Persistence of ozone-induced changes in lung function and airway
31 responsiveness. In: Schneider, T.; Lee, S. D.; Wolters, G. J. R.; Grant, L. D., eds. Atmospheric ozone research
32 and its policy implications: proceedings of the 3rd US-Dutch international symposium; May 1988; Nijmegen,
33 The Netherlands. Amsterdam, The Netherlands: Elsevier Science Publishers; pp. 483-492. (Studies in
34 environmental science 35).
35 Folinsbee, L. J.; Hazucha, M. J. (2000) Time course of response to ozone exposure in healthy adult females.
36 Inhalation Toxicol. 12: 151-167.
37 Folinsbee, L. J.; Horvath, S. M. (1986) Persistence of the acute effects of ozone exposure. Aviat. Space Environ.
38 Med. 57: 1136-1143.
39 Folinsbee, L. J.; Silverman, F.; Shephard, R. J. (1975) Exercise responses following ozone exposure. J. Appl.
40 Physiol. 38: 996-1001.
41 Folinsbee, L. J.; Silverman, F.; Shephard, R. J. (1977) Decrease of maximum work performance following ozone
42 exposure. J. Appl. Physiol.: Respir. Environ. Exercise Physiol. 42: 531-536.
43 Folinsbee, L. J.; Drinkwater, B. L.; Bedi, J. F.; Horvath, S. M. (1978) The influence of exercise on the pulmonary
44 function changes due to exposure to low concentrations of ozone. In: Folinsbee, L. J.; Wagner, J. A.; Borgia,
45 J. F.; Drinkwater, B. L.; Gliner, J. A.; Bedi, J. F., eds. Environmental stress: individual human adaptations.
46 New York, NY: Academic Press; pp. 125-145.
47 Folinsbee, L. J.; Bedi, J. F.; Horvath, S. M. (1980) Respiratory responses in humans repeatedly exposed to low
48 concentrations of ozone. Am. Rev. Respir. Dis. 121: 431-439.
49 Folinsbee, L. J.; Bedi, J. F.; Horvath, S. M. (1984) Pulmonary function changes after 1 h continuous heavy exercise
50 in 0.21 ppm ozone. J. Appl. Physiol.: Respir. Environ. Exercise Physiol. 57: 984-988.
51 Folinsbee, L. J.; Horstman, D. H.; Vorona, R. D.; Prince, J. M.; Berry, M. (1986) Determinants of endurance
52 performance during ozone inhalation. Presented at: 30th International Union of Physiological Sciences
53 congress; July; Vancouver, BC, Canada. Proc. Int. Union Physiol. Sci. 16: 176.
54 Folinsbee, L. J.; McDonnell, W. F.; Horstman, D. H. (1988) Pulmonary function and symptom responses after
55 6.6-hour exposure to 0.12 ppm ozone with moderate exercise. JAPCA 38: 28-35.
January 2005 AX6-13 8 DRAFT-DO NOT QUOTE OR CITE
-------
1 Folinsbee, L. I; Devlin, R. B.; Abdul-Salaam, S.; Koren, H. S. (1993) Repeated severe ozone exposure causes
2 depressed baseline spirometry. Am. Rev. Respir. Dis. 147: A638.
3 Folinsbee, L. I; Horstman, D. H.; Kehrl, H. R.; Harder, S.; Abdul-Salaam, S.; Ives, P. J. (1994) Respiratory
4 responses to repeated prolonged exposure to 0.12 ppm ozone. Am. J. Respir. Crit. Care Med. 149: 98-105.
5 Folinsbee, L. J.; Devlin, R. B.; Robbins, M. K.; Biscardi, F. H.; Abdul-Salaam, S.; Koren, H. S. (1998) Repeated
6 exposure of humans to ozone: pulmonary function and symptom responses. Research Triangle Park, NC:
7 U.S. Environmental Protection Agency; National Center for Environmental Assessment; unpublished data.
8 Foster, W. M.; Stetkiewicz, P. T. (1996) Regional clearance of solute from the respiratory epithelia: 18-20 h
9 postexposure to ozone. J. Appl. Physiol. 81: 1143-1149.
10 Foster, W. M.; Costa, D. L.; Langenback, E. G. (1987) Ozone exposure alters tracheobronchial mucociliary function
11 in humans. J. Appl. Physiol. 63: 996-1002.
12 Foster, W. M.; Wills-Karp, M.; Tankersley, C. G.; Chen, X.; Paquette, N. C. (1996) Bloodborne markers in humans
13 during multiday exposure to ozone. J. Appl. Physiol. 81: 794-800.
14 Foster, W. M.; Brown, R. H.; Macri, K.; Mitchell, C. S. (2000) Bronchial reactivity of healthy subjects: 18-20 h
15 postexposure to ozone. J. Appl. Physiol. 89: 1804-1810.
16 Fox, S. D.; Adams, W. C.; Brookes, K. A.; Lasley, B. L. (1993) Enhanced response to ozone exposure during the
17 follicular phase of the menstrual cycle. Environ. Health Perspect. 101: 242-244.
18 Foxcroft, W. J.; Adams, W. C. (1986) Effects of ozone exposure on four consecutive days on work performance and
19 VO2max. J. Appl. Physiol. 61: 960-966.
20 Frampton, M. W.; Morrow, P. E.; Cox, C.; Levy, P. C.; Condemi, J. J.; Speers, D.; Gibb, F. R.; Utell, M. J. (1995)
21 Sulfuric acid aerosol followed by ozone exposure in healthy and asthmatic subjects. Environ. Res. 69: 1-14.
22 Frampton, M. W.; Morrow, P. E.; Torres, A.; Cox, C.; Voter, K. Z.; Utell, M. J.; Gibb, F. R.; Speers, D. M. (1997a)
23 Ozone responsiveness in smokers and nonsmokers. Am. J. Respir. Crit. Care Med. 155: 116-121.
24 Frampton, M. W.; Balmes, J. R.; Cox, C.; Krein, P. M.; Speers, D. M.; Tsai, Y.; Utell, M. J. (1997b) Effects of
25 ozone on normal and potentially sensitive human subjects. Part III: mediators of inflammation in
26 bronchoalveolar lavage fluid from nonsmokers, smokers, and asthmatic subjects exposed to ozone: a
27 collaborative study. Cambridge, MA: Health Effects Institute. Research report 78; pp. 73-99, 81-99.
28 Frank, R.; Liu, M. C.; Spannhake, E. W.; Mlynarek, S.; Macri, K.; Weinmann, G. G. (2001) Repetitive ozone
29 exposure of young adults: evidence of persistent small airway dysfunction. Am. J. Respir. Crit. Care Med.
30 164: 1253-1260.
31 Freed, A. N.; Chou, C. L.; Fuller, S. D.; Croxton, T. L. (1996) Ozone-induced vagal reflex modulates airways
32 reactivity in rabbits. Respir. Physiol. 105: 95-102.
33 Gabrielson, E. W.; Yu, X.-Y.; Spannhake, E. W. (1994) Comparison of the toxic effects of hydrogen peroxide and
34 ozone on cultured human bronchial epithelial cells. Environ. Health Perspect. 102: 972-974.
35 Gerrity, T. R.; Bennett, W. D.; Kehrl, H.; DeWitt, P. J. (1993) Mucociliary clearance of inhaled particles measured
36 at 2 h after ozone exposure in humans. J. Appl. Physiol. 74: 2984-2989.
37 Gliner, J. A.; Matsen-Twisdale, J. A.; Horvath, S. M. (1979) Auditory and visual sustained attention during ozone
38 exposure. Aviat. Space Environ. Med. 50: 906-910.
39 Gliner, J. A.; Horvath, S. M.; Sorich, R. A.; Hanley, J. (1980) Psychomotor performance during ozone exposure:
40 spectral and discriminant function analysis of EEG. Aviat. Space Environ. Med. 51: 344-351.
41 Gliner, J. A.; Horvath, S. M.; Folinsbee, L. J. (1983) Preexposure to low ozone concentrations does not diminish the
42 pulmonary function response on exposure to higher ozone concentrations. Am. Rev. Respir. Dis. 127: 51-55.
43 Golden, J. A.; Nadel, J. A.; Boushey, H. A. (1978) Bronchial hyperirritability in healthy subjects after exposure to
44 ozone. Am. Rev. Respir. Dis. 118: 287-294.
45 Gomez, F. P.; Rodriguez-Roisin, R. (2002) Global initiative for chronic obstructive lung disease (GOLD) guidelines
46 for chronic obstructive pulmonary disease. Curr. Opin. Pulm. Med. 8: 81-86.
47 Gong, H., Jr.; Tierney, D. F. (1995) Evaluation of chronic obstructive pulmonary disease patients for ozone
48 sensitivity: validation of health advisories. Sacramento, CA.: California State Air Resources Board, Research
49 Division; CARB-A133-123. Available from: NTIS, Springfield, VA; PB98-153620.
50 Gong, H., Jr.; Bradley, P. W.; Simmons, M. S.; Tashkin, D. P. (1986) Impaired exercise performance and pulmonary
51 function in elite cyclists during low-level ozone exposure in a hot environment. Am. Rev. Respir. Dis.
52 134:726-733.
53 Gong, H., Jr.; Shamoo, D. A.; Anderson, K. R.; Linn, W. S. (1997a) Responses of older men with and without
54 chronic obstructive pulmonary disease to prolonged ozone exposure. Arch. Environ. Health 52: 18-25.
55 Gong, H., Jr.; McManus, M. S.; Linn, W. S. (1997b) Attenuated response to repeated daily ozone exposures in
56 asthmatic subjects. Arch. Environ. Health 52: 34-41.
January 2005 AX6-139 DRAFT-DO NOT QUOTE OR CITE
-------
1 Gong, H., Jr.; Wong, R.; Sarma, R. J.; Linn, W. S.; Sullivan, E. D.; Shamoo, D. A.; Anderson, K. R.; Prasad, S. B.
2 (1998) Cardiovascular effects of ozone exposure in human volunteers. Am. J. Respir. Crit. Care Med.
3 158: 538-546.
4 Graham, D. E.; Koren, H. S. (1990) Biomarkers of inflammation in ozone-exposed humans: comparison of the nasal
5 and bronchoalveolar lavage. Am. Rev. Respir. Dis. 142: 152-156.
6 Graham, D.; Henderson, F.; House, D. (1988) Neutrophil influx measured in nasal lavages of humans exposed to
7 ozone. Arch. Environ. Health 43: 228-233.
8 Greve, J. M; Davis, G.; Meyer, A. M; Forte, C. P.; Yost, S. C; Marlor, C. W.; Kamarck, M. E.; McClelland, A.
9 (1989) The major human rhinovirus receptor is ICAM-1. Cell 56: 839-847.
10 Grievink, L.; Jansen, S. M. A.; Van't Veer, P.; Brunekreef, B. (1998) Acute effects of ozone on pulmonary function
11 of cyclists receiving antioxidant supplements. Occup. Environ. Med. 55: 13-17.
12 Grievink, L.; Zijlstra, A. G.; Ke, X.; Brunekreef, B. (1999) Double-blind intervention trial on modulation of ozone
13 effects on pulmonary function by antioxidant supplements. Am. J. Epidemiol. 149: 306-314.
14 Gwynn, R. C.; Thurston, G. D. (2001) The burden of air pollution: impacts among racial minorities. Environ. Health
15 Perspect. 109(suppl. 4): 501-506.
16 Hackney, J. D.; Linn, W. S.; Buckley, R. D.; Hislop, H. J. (1976) Studies in adaptation to ambient oxidant air
17 pollution: effects of ozone exposure in Los Angeles residents vs. new arrivals. Environ. Health Perspect.
18 18: 141-146.
19 Hackney, J. D.; Linn, W. S.; Mohler, J. G.; Collier, C. R. (1977a) Adaptation to short-term respiratory effects of
20 ozone in men exposed repeatedly. J. Appl. Physiol.: Respir. Environ. Exercise Physiol. 43: 82-85.
21 Hackney, J. D.; Linn, W. S.; Karuza, S. K.; Buckley, R. D.; Law, D. C.; Bates, D. V.; Hazucha, M.; Pengelly, L. D.;
22 Silverman, F. (1977b) Effects of ozone exposure in Canadians and southern Californians: evidence for
23 adaptation? Arch. Environ. Health 32: 110-116.
24 Hackney, J. D.; Linn, W. S.; Shamoo, D. A.; Avol, E. L. (1989) Responses of selected reactive and nonreactive
25 volunteers to ozone exposure in high and low pollution seasons. In: Schneider, T.; Lee, S. D.; Welters, G. J.
26 R.; Grant, L. D., eds. Atmospheric ozone research and its policy implications: proceedings of the 3rd
27 US-Dutch international symposium; May 1988; Nijmegen, The Netherlands. Amsterdam, The Netherlands:
28 Elsevier Science Publishers; pp. 311-318. (Studies in environmental science 35).
29 Hamilton, R. F.; Li, L.; Eschenbacher, W. L.; Szweda, L.; Holian, A. (1998) Potential involvement of
30 4-hydroxynonenal in the response of human lung cells to ozone. Am. J. Physiol. 274: L8-L16.
31 Hanania, N. A.; Tarlo, S. M.; Silverman, F.; Urch, B.; Senathirajah, N.; Zamel, N.; Corey, P. (1998) Effect of
32 exposure to low levels of ozone on the response to inhaled allergen in allergic asthmatic patients. Chest
33 114:752-756.
34 Hazbun, M. E.; Hamilton, R.; Holian, A.; Eschenbacher, W. L. (1993) Ozone-induced increases in substance P and
3 5 8-epi-prostaglandin F2a in the airways of human subjects. Am. J. Respir. Cell Mol. Biol. 9: 568-572.
36 Hazucha, M. J. (1987) Relationship between ozone exposure and pulmonary function changes. J. Appl. Physiol.
37 62: 1671-1680.
38 Hazucha, M. J. (1993) Meta-analysis and "effective dose" revisited. In: Mohr, U., ed. Proceedings of the 3rd
39 International Inhalation Symposium on Advances in Controlled Clinical Inhalation Studies; October, 1991.
40 Berlin, Germany: Springer-Verlag; pp. 247-256.
41 Hazucha, M. J.; Sant'Ambrogio, G. (1993) Effects of ozone on the activity of slowly (SAR) and rapidly adapting
42 (RAR) receptors in cats. FASEB J. 7: 407A.
43 Hazucha, M.; Silverman, F.; Parent, C.; Field, S.; Bates, D. V. (1973) Pulmonary function in man after short-term
44 exposure to ozone. Arch. Environ. Health 27: 183-188.
45 Hazucha, M. J.; Bates, D. V.; Bromberg, P. A. (1989) Mechanism of action of ozone on the human lung. J. Appl.
46 Physiol. 67: 1535-1541.
47 Hazucha, M. J.; Folinsbee, L. J.; Seal, E., Jr. (1992) Effects of steady-state and variable ozone concentration profiles
48 on pulmonary function. Am. Rev. Respir. Dis. 146: 1487-1493.
49 Hazucha, M. J.; Folinsbee, L. J.; Seal, E.; Bromberg, P. A. (1994) Lung function response of healthy women after
50 sequential exposures to NO2 and O3. Am. J. Respir. Crit. Care Med. 150: 642-647.
51 Hazucha, M. J.; Madden, M.; Pape, G.; Becker, S.; Devlin, R.; Koren, H. S.; Kehrl, H.; Bromberg, P. A. (1996)
52 Effects of cyclo-oxygenase inhibition on ozone-induced respiratory inflammation and lung function changes.
53 Eur. J. Appl. Physiol. Occup. Med. 73: 17-27.
54 Hazucha, M. J.; Folinsbee, L. J.; Bromberg, P. A. (2003) Distribution and reproducibility of spirometric response to
55 ozone by gender and age. J. Appl. Physiol. 95: 1917-1925.
January 2005 AX6-140 DRAFT-DO NOT QUOTE OR CITE
-------
1 Henderson, F. W.; Dubovi, E. I; Harder, S.; Seal, E., Jr.; Graham, D. (1988) Experimental rhinovirus infection in
2 human volunteers exposed to ozone. Am. Rev. Respir. Dis. 137: 1124-1128.
3 Hetrick, S. M; Gould, W. D.; Christensen, D. E. (2000) Inflight cabin ozone aboard long duration C-5 airlift
4 missions: a historical issue revisited. Aviat. Space Environ. Med. 71: 408-414.
5 Hiltermann, T. J. N.; Stolk, I; Hiemstra, P. S.; Fokkens, P. H. B.; Rombout, P. J. A.; Sont, J. K.; Sterk, P. I;
6 Dijkman, J. H. (1995) Effect of ozone exposure on maximal airway narrowing in non-asthmatic and asthmatic
7 subjects. Clin. Sci. 89: 619-624.
8 Hiltermann, T. J. N.; de Bruijne, C. R.; Stolk, J.; Zwinderman, A. H.; Spieksma, F. Th. M.; Roemer, W.;
9 Steerenberg, P. A.; Fischer, P. H.; van Bree, L.; Hiemstra, P. S. (1997) Effects of photochemical air pollution
10 and allergen exposure on upper respiratory tract inflammation in asthmatics. Am. J. Respir. Crit. Care Med.
11 156:1765-1772.
12 Hiltermann, T. J. N.; Peters, E. A.; Alberts, B.; Kwikkers, K.; Borggreven, P. A.; Hiemstra, P. S.; Dijkman, J. H.;
13 van Bree, L. A.; Stolk, J. (1998) Ozone-induced airway hyperresponsiveness in patients with asthma: role of
14 neutrophil-derived serine proteinases. Free Radical Biol. Med. 24: 952-958.
15 Hiltermann, J. T. N.; Lapperre, T. S.; Van Bree, L.; Steerenberg, P. A.; Brahim, J. J.; Sont, J. K.; Sterk, P. J.;
16 Hiemstra, P. S.; Stolk, J. (1999) Ozone-induced inflammation assessed in sputum and bronchial lavage fluid
17 from asthmatics: a new noninvasive tool in epidemiologic studies on air pollution and asthma. Free Radical
18 Biol. Med. 27: 1448-1454.
19 Holtzman, M. J.; Cunningham, J. H.; Sheller, J. R.; Irsigler, G. B.; Nadel, J. A.; Boushey, H. A. (1979) Effect of
20 ozone on bronchial reactivity in atopic and nonatopic subjects. Am. Rev. Respir. Dis. 120: 1059-1067.
21 Holz, O.; Torres, R. A.; Timm, P.; Mticke, M.; Richter, K.; Koschyk, S.; Magnussen, H. (1999) Ozone-induced
22 airway inflammatory changes differ between individuals and are reproducible. Am. J. Respir. Crit. Care Med.
23 159:776-784.
24 Holz, O.; Mticke, M.; Paasch, K.; Bohme, S.; Timm, P.; Richter, K.; Magnussen, H.; Torres, R. A. (2002) Repeated
25 ozone exposures enhance bronchial allergen responses in subjects with rhinitis or asthma. Clin. Exp. Allergy.
26 32:681-689.
27 Horstman, D. H.; Folinsbee, L. J.; Ives, P. J.; Abdul-Salaam, S.; McDonnell, W. F. (1990) Ozone concentration and
28 pulmonary response relationships for 6.6-hour exposures with five hours of moderate exercise to 0.08, 0.10,
29 and 0.12 ppm. Am. Rev. Respir. Dis. 142: 1158-1163.
30 Horstman, D. H.; Ball, B. A.; Brown, J.; Gerrity, T.; Folinsbee, L. J. (1995) Comparison of pulmonary responses of
31 asthmatic and nonasthmatic subjects performing light exercise while exposed to a low level of ozone.
32 Toxicol. Ind. Health 11: 369-385.
33 Horvath, S. M.; Gliner, J. A.; Matsen-Twisdale, J. A. (1979) Pulmonary function and maximum exercise responses
34 following acute ozone exposure. Aviat. Space Environ. Med. 50: 901-905.
35 Horvath, S. M.; Gliner, J. A.; Folinsbee, L. J. (1981) Adaptation to ozone: duration of effect. Am. Rev. Respir. Dis.
36 123:496-499.
37 Horvath, S. M.; Bedi, J. F.; Drechsler-Parks, D. M. (1986) Effects of peroxyacetyl nitrate alone and in combination
3 8 with ozone in healthy young women. J. Air Pollut. Control Assoc. 36: 265-270.
39 Horvath, S. M.; Folinsbee, L. J.; Bedi, J. F. (1987) Combined effect of ozone and sulfuric acid on pulmonary
40 function in man. Am. Ind. Hyg. Assoc. J. 48: 94-98.
41 Horvath, S. M.; Bedi, J. F.; Drechsler-Parks, D. M.; Williams, R. E. (1991) Alterations in pulmonary function
42 parameters during exposure to 80 ppb ozone for 6.6 hours in healthy middle aged individuals. In: Berglund,
43 R. L.; Lawson, D. R.; McKee, D. J., eds. Tropospheric ozone and the environment: papers from an
44 international conference; March 1990; Los Angeles, CA. Pittsburgh, PA: Air & Waste Management
45 Association; pp. 59-70. (A&WMA transactions series: no. TR-19).
46 Housley, D. G.; Eccles, R.; Richards, R. J. (1996) Gender difference in the concentration of the antioxidant uric acid
47 in human nasal lavage. Acta Oto-Laryngol. 116: 751-754.
48 Hu, S.-C.; Ben-Jebria, A.; Ultman, J. S. (1994) Longitudinal distribution of ozone absorption in the lung: effects of
49 respiratory flow. J. Appl. Physiol. 77: 574-583.
50 Jenkins, H. S.; Devalia, J. L.; Mister, R. L.; Bevan, A. M.; Rusznak, C.; Davies, R. J. (1999) The effect of exposure
51 to ozone and nitrogen dioxide on the airway response of atopic asthmatics to inhaled allergen: dose- and
52 time-dependent effects. Am. J. Respir. Crit. Care Med. 160: 33-39.
53 Joad, J. P.; Kott, K. S.; Brie, J. M. (1996) The local C-fiber contribution to ozone-induced effects on the isolated
54 guinea pig lung. Toxicol. Appl. Pharmacol. 141: 561-567.
January 2005 AX6-141 DRAFT-DO NOT QUOTE OR CITE
-------
1 Johnson, D. A.; Winters, R. S.; Woolley, T.; Graham, D.; Henderson, F. W. (1986) Ozone effects on
2 alpha-1-proteinase inhibitor in vivo: blood plasma inhibitory activity is unchanged. Exp. Lung Res.
3 11:95-103.
4 Torres, R.; Nowak, D.; Magnussen, H.; Speckin, P.; Koschyk, S. (1996) The effect of ozone exposure on allergen
5 responsiveness in subjects with asthma or rhinitis. Am. J. Respir. Crit. Care Med. 153: 56-64.
6 Torres, R. A.; Holz, O.; Zachgo, W.; Timm, P.; Koschyk, S.; Miiller, B.; Grimminger, F.; Seeger, W.; Kelly, F. J.;
7 Dunster, C.; Frischer, T.; Lubec, G.; Waschewski, M; Niendorf, A.; Magnussen, H. (2000) The effect of
8 repeated ozone exposures on inflammatory markers in bronchoalveolar lavage fluid and mucosal biopsies.
9 Am. J. Respir. Crit. Care Med. 161: 1855-1861.
10 Kabel, J. R.; Ben-Jebria, A.; Ultman, J. S. (1994) Longitudinal distribution of ozone absorption in the lung:
11 comparison of nasal and oral quiet breathing. J. Appl. Physiol. 77: 2584-2592.
12 Kagawa, J. (1986) Experimental studies on human health effects of aerosol and gaseous pollutants. In: Lee, S. D.;
13 Schneider, T.; Grant, L. D.; Verkerk, P. J., eds. Aerosols: research, risk assessment and control strategies -
14 proceedings of the second U.S.-Dutch international symposium; May 1985; Williamsburg, VA. Chelsea, MI:
15 Lewis Publishers, Inc.; pp. 683-697.
16 Kehrl, H. R.; Vincent, L. M.; Kowalsky, R. J.; Horstman, D. H.; O'Neil, J. J.; McCartney, W. H.; Bromberg, P. A.
17 (1987) Ozone exposure increases respiratory epithelial permeability in humans. Am. Rev. Respir. Dis.
18 135: 1124-1128.
19 Kehrl, H. R.; Vincent, L. M.; Kowalsky, R. J.; Horstman, D. H.; O'Neil, J. J.; McCartney, W. H.; Bromberg, P. A.
20 (1989) Ozone-induced changes in the pulmonary clearance of 99mTc-DTPA in man. In: Schneider, T.; Lee,
21 S. D.; Wolters, G. J. R.; Grant, L. D., eds. Atmospheric ozone research and its policy implications:
22 proceedings of the 3rd US-Dutch international symposium; May 1988; Nijmegen, The Netherlands.
23 Amsterdam, The Netherlands: Elsevier Science Publishers; pp. 343-351. (Studies in environmental science
24 35).
25 Kehrl, H. R.; Peden, D. B.; Ball, B. A.; Folinsbee, L. J.; Horstman, D. H. (1999) Increased specific airway reactivity
26 of persons with mild allergic asthma after 7.6 hours of exposure to 0.16 ppm ozone. J. Allergy. Clin.
27 Immunol. 104: 1198-1204.
28 Kleinman, M. T.; Mautz, W. J.; Bjarnason, S. (1999) Adaptive and non-adaptive responses in rats exposed to ozone,
29 alone and in mixtures, with acidic aerosols. Inhalation Toxicol. 11: 249-264.
30 Koenig, J. Q.; Covert, D. S.; Marshall, S. G.; Van Belle, G.; Pierson, W. E. (1987) The effects of ozone and nitrogen
31 dioxide on pulmonary function in healthy and in asthmatic adolescents. Am. Rev. Respir. Dis.
32 136: 1152-1157.
33 Koenig, J. Q.; Covert, D. S.; Smith, M. S.; Van Belle, G.; Pierson, W. E. (1988) The pulmonary effects of ozone and
34 nitrogen dioxide alone and combined in healthy and asthmatic adolescent subjects. Toxicol. Ind. Health
35 4:521-532.
36 Koenig, J. Q.; Covert, D. S.; Hanley, Q. S.; Van Belle, G.; Pierson, W. E. (1990) Prior exposure to ozone potentiates
37 subsequent response to sulfur dioxide in adolescent asthmatic subjects. Am. Rev. Respir. Dis. 141: 377-380.
38 Koenig, J. Q.; Covert, D. S.; Pierson, W. E.; Hanley, Q. S.; Rebolledo, V; Dumler, K.; McKinney, S. E. (1994)
39 Oxidant and acid aerosol exposure in healthy subjects and subjects with asthma. Part I: effects of oxidants,
40 combined with sulfuric or nitric acid, on the pulmonary function of adolescents with asthma. Cambridge, MA:
41 Health Effects Institute; pp. 1-36; research report no. 70.
42 Konig, G.; Rommelt, H.; Kienle, H.; Dirnagl, K.; Polke, H.; Fruhmann, G. (1980) Anderung der
43 bronchomotorischen Reagibilitat des Menschen durch Einwirkung von Ozon [Changes in the bronchial
44 reactivity of humans caused by the influence of ozone]. Arbeitsmed. Sozialmed. Praeventivmed.
45 151:261-263.
46 Koren, H. S.; Devlin, R. B.; Graham, D. E.; Mann, R.; McGee, M. P.; Horstman, D. H.; Kozumbo, W. J.; Becker, S.;
47 House, D. E.; McDonnell, W. F.; Bromberg, P. A. (1989a) Ozone-induced inflammation in the lower airways
48 of human subjects. Am. Rev. Respir. Dis. 139: 407-415.
49 Koren, H. S.; Devlin, R. B.; Graham, D. E.; Mann, R.; McDonnell, W. F. (1989b) The inflammatory response in
50 human lung exposed to ambient levels of ozone. In: Schneider, T.; Lee, S. D.; Wolters, G. J. R.; Grant, L. D.,
51 eds. Atmospheric ozone research and its policy implications: proceedings of the 3rd US-Dutch international
52 symposium; May 1988; Nijmegen, The Netherlands. Amsterdam, The Netherlands: Elsevier Science
53 Publishers; pp. 745-753. (Studies in environmental science 35).
54 Koren, H. S.; Hatch, G. E.; Graham, D. E. (1990) Nasal lavage as a tool in assessing acute inflammation in response
55 to inhaled pollutants. Toxicology 60: 15-25.
January 2005 AX6-142 DRAFT-DO NOT QUOTE OR CITE
-------
1 Koren, H. S.; Devlin, R. B.; Becker, S.; Perez, R.; McDonnell, W. F. (1991) Time-dependent changes of markers
2 associated with inflammation in the lungs of humans exposed to ambient levels of ozone. Toxicol. Pathol.
3 19:406-411.
4 Kozumbo, W. I; Hanley, N. M; Agarwas, S.; Thomas, M. I; Madden, M. C. (1996) Products of ozonized
5 arachidonic acid potentiate the formation of DNA single strand breaks in cultured human lung cells. Environ.
6 Mol. Mutagen. 27: 185-195.
7 Kreit, J. W.; Gross, K. B.; Moore, T. B.; Lorenzen, T. I; D'Arcy, I; Eschenbacher, W. L. (1989) Ozone-induced
8 changes in pulmonary function and bronchial responsiveness in asthmatics. J. Appl. Physiol. 66: 217-222.
9 Krishna, M. T.; Springall, D.; Meng, Q.-H.; Withers, N.; Macleod, D.; Biscione, G.; Frew, A.; Polak, J.; Holgate, S.
10 (1997a) Effects of ozone on epithelium and sensory nerves in the bronchial mucosa of healthy humans. Am. J.
11 Respir. Crit. Care Med. 156: 943-950.
12 Krishna, M. T.; Blomberg, A.; Biscione, G. L.; Kelly, F.; Sandstrom, T.; Frew, A.; Holgate, S. (1997b) Short-term
13 ozone exposure upregulates P-selectin in normal human airways. Am. J. Respir. Crit. Care Med.
14 155: 1798-1803.
15 Krishna, M. T.; Madden, J.; Teran, L. M.; Biscione, G. L.; Lau, L. C. K.; Withers, N. J.; Sandstrom, T.; Mudway, L;
16 Kelly, F. J.; Walls, A.; Frew, A. J.; Holgate, S. T. (1998) Effects of 0.2 ppm ozone onbiomarkers of
17 inflammation in bronchoalveolar lavage fluid and bronchial mucosa of healthy subjects. Eur. Respir. J.
18 11: 1294-1300
19 Kulle, T. J.; Sauder, L. R.; Kerr, H. D.; Farrell, B. P.; Bermel, M. S.; Smith, D. M. (1982) Duration of pulmonary
20 function adaptation to ozone in humans. Am. Ind. Hyg. Assoc. J. 43: 832-837.
21 Kulle, T. J.; Sauder, L. R.; Hebel, J. R.; Chatham, M. D. (1985) Ozone response relationships in healthy
22 nonsmokers. Am. Rev. Respir. Dis. 132: 36-41.
23 Lang, D. S.; Torres, R. A.; Mticke, M.; Siegfried, W.; Magnussen, H. (1998) Interactions between human
24 bronchoepithelial cells and lung fibroblasts after ozone exposure in vitro. Toxicol. Lett. 96-97: 13-24.
25 Larsen, R. I.; McDonnell, W. F.; Horstman, D. H.; Folinsbee, L. J. (1991) An air quality data analysis system for
26 interrelating effects, standards, and needed source reductions: part 11. a lognormal model relating human lung
27 function decrease to O3 exposure. J. Air Waste Manage. Assoc. 41: 455-459.
28 Lategola, M. T.; Melton, C. E.; Higgins, E. A. (1980) Effects of ozone on symptoms and cardiopulmonary function
29 in a flight attendant surrogate population. Aviat. Space Environ. Med. 51: 237-246.
30 Lategola, M. T.; Melton, C. E.; Higgins, E. A. (1980) Pulmonary and symptom threshold effects of ozone in airline
31 passenger and cockpit crew surrogates. Aviat. Space Environ. Med. 51: 878-884.
32
33 Lefohn, A. S.; Foley, J. K. (1993) Establishing relevant ozone standards to protect vegetation and human health:
34 exposure/dose-response considerations. Air Waste 43: 106-112.
35 Linder, J.; Herren, D.; Monn, C.; Wanner, H.-U. (1988) Die Wirkung von Ozon auf die korperliche
36 Leistungsfa'higkeit [The effect of ozone on physical activity]. Schweiz Z. Sportmed. 36: 5-10.
37 Linn, W. S.; Buckley, R. D.; Spier, C. E.; Blessey, R. L.; Jones, M. P.; Fischer, D. A.; Hackney, J. D. (1978) Health
38 effects of ozone exposure in asthmatics. Am. Rev. Respir. Dis. 117: 835-843.
39 Linn, W. S.; Jones, M. P.; Bachmayer, E. A.; Spier, C. E.; Mazur, S. F.; Avol, E. L.; Hackney, J. D. (1980)
40 Short-term respiratory effects of polluted ambient air: a laboratory study of volunteers in a high-oxidant
41 community. Am. Rev. Respir. Dis. 121: 243-252.
42 Linn, W. S.; Fischer, D. A.; Medway, D. A.; Anzar, U. T.; Spier, C. E.; Valencia, L. M.; Venet, T. G.; Hackney,
43 J. D. (1982) Short-term respiratory effects of 0.12 ppm ozone exposure in volunteers with chronic obstructive
44 pulmonary disease. Am. Rev. Respir. Dis. 125: 658-663.
45 Linn, W. S.; Avol, E. L.; Hackney, J. D. (1983) Effects of ambient oxidant pollutants on humans: a movable
46 environmental chamber study. In: Lee, S. D.; Mustafa, M. G.; Mehlman, M. A., eds. International symposium
47 on the biomedical effects of ozone and related photochemical oxidants; March 1982; Pinehurst, NC.
48 Princeton, NJ: Princeton Scientific Publishers, Inc.; pp. 125-137. (Advances in modern environmental
49 toxicology: v. 5).
50 Linn, W. S.; Avol, E. L.; Shamoo, D. A.; Spier, C. E.; Valencia, L. M.; Venet, T. G.; Fischer, D. A.; Hackney, J. D.
51 (1986) A dose-response study of healthy, heavily exercising men exposed to ozone at concentrations near the
52 ambient air quality standard. Toxicol. Ind. Health 2: 99-112.
53 Linn, W. S.; Avol, E. L.; Shamoo, D. A.; Peng, R.-C.; Valencia, L. M.; Little, D. E.; Hackney, J. D. (1988) Repeated
54 laboratory ozone exposures of volunteer Los Angeles residents: an apparent seasonal variation in response.
55 Toxicol. Ind. Health 4: 505-520.
January 2005 AX6-143 DRAFT-DO NOT QUOTE OR CITE
-------
1 Linn, W. S.; Shamoo, D. A.; Anderson, K. R.; Peng, R.-C.; Avol, E. L.; Hackney, J. D. (1994) Effects of prolonged,
2 repeated exposure to ozone, sulfuric acid, and their combination in healthy and asthmatic volunteers. Am. J.
3 Respir. Crit. Care Med. 150: 431-440.
4 Linn, W. S.; Anderson, K. R.; Shamoo, D. A.; Edwards, S. A.; Webb, T. L.; Hackney, J. D.; Gong, H., Jr. (1995)
5 Controlled exposures of young asthmatics to mixed oxidant gases and acid aerosol. Am. J. Respir. Crit. Care
6 Med. 152: 885-891.
7 Linn, W. S.; Gong, H., Jr.; Shamoo, D. A.; Anderson, K. R.; Avol, E. L. (1997) Chamber exposures of children to
8 mixed ozone, sulfur dioxide, and sulfuric acid. Arch. Environ. Health 52: 179-187.
9 Liu, L.; Leech, J. A.; Urch, R. B.; Silverman, F. S. (1997) In vivo salicylate hyroxylation: a potential biomarker for
10 assessing acute ozone exposure and effects in humans. Am. J. Respir. Crit. Care Med. 156: 1405-1412.
11 Liu, L.; Leech, J. A.; Urch, R. B.; Poon, R.; Zimmerman, B.; Kubay, J. M.; Silverman, F. S. (1999) A comparison of
12 biomarkers of ozone exposure in human plasma, nasal lavage, and sputum. Inhalation Toxicol. 11: 657-674.
13 MacNee, W.; Morrison, D.; Rahman, I.; Li, X. Y.; Donaldson, K. (1996) Cigarette smoke and ozone-induced
14 epithelial perturbation in vivo and in vitro: the role of glutathione. Chest 109(suppl.): 39S.
15 Magnussen, H.; Nowak, D.; Jorres, R. (1994) Experimentelle Schadstoffeffekte bei Gesundenund Asthmatikern
16 [Experimental effects of air pollutants in healthy and asthmatic subjects]. Pneumologie 48: 85-88.
17 Marthan, R.; Roux, E.; Savineau, J.-P. (1996) Human bronchial smooth muscle responsiveness after in vitro
18 exposure to oxidizing pollutants. Cell Biol. Toxicol. 12: 245-249.
19 Martin, T. R.; Castile, R. G.; Fredberg, J. J.; Wohl, M. E. B.; Mead, J. (1987) Airway size is related to sex but not
20 lung size in normal adults. J. Appl. Physiol. 63: 2042-2047.
21 McBride, D. E.; Koenig, J. Q.; Luchtel, D. L.; Williams, P. V.; Henderson, W. R., Jr. (1994) Inflammatory effects of
22 ozone in the upper airways of subjects with asthma. Am. J. Respir. Crit. Care Med. 149: 1192-1197.
23 McDonnell, W. F. (1996) Individual variability in human lung function responses to ozone exposure. Environ.
24 Toxicol. Pharmacol. 2: 171-175.
25 McDonnell, W. F.; Smith, M. V. (1994) Description of acute ozone response as a function of exposure rate and total
26 inhaled dose. J. Appl. Physiol. 76: 2776-2784.
27 McDonnell, W. F.; Horstman, D. H.; Hazucha, M. J.; Seal, E., Jr.; Haak, E. D.; Salaam, S. A.; House, D. E. (1983)
28 Pulmonary effects of ozone exposure during exercise: dose-response characteristics. J. Appl. Physiol.: Respir.
29 Environ. Exercise Physiol. 54: 1345-1352.
30 McDonnell, W. F., Ill; Horstman, D. H.; Abdul-Salaam, S.; House, D. E. (1985a) Reproducibility of individual
31 responses to ozone exposure. Am. Rev. Respir. Dis. 131: 36-40.
32 McDonnell, W. F., Ill; Chapman, R. S.; Leigh, M. W.; Strope, G. L.; Collier, A. M. (1985b) Respiratory responses
33 of vigorously exercising children to 0.12 ppm ozone exposure. Am. Rev. Respir. Dis. 132: 875-879.
34 McDonnell, W. F.; Horstman, D. H.; Abdul-Salaam, S.; Raggio, L. J.; Green, J. A. (1987) The respiratory responses
35 of subjects with allergic rhinitis to ozone exposure and their relationship to nonspecific airway reactivity.
36 Toxicol. Ind. Health 3: 507-517.
37 McDonnell, W. F.; Kehrl, H. R.; Abdul-Salaam, S.; Ives, P. J.; Folinsbee, L. J.; Devlin, R. B.; O'Neil, J. J.;
38 Horstman, D. H. (1991) Respiratory response of humans exposed to low levels of ozone for 6.6 hours. Arch.
39 Environ. Health 46: 145-150.
40 McDonnell, W. F.; Muller, K. E.; Bromberg, P. A.; Shy, C. M. (1993) Predictors of individual differences in acute
41 response to ozone exposure. Am. Rev. Respir. Dis. 147: 818-825.
42 McDonnell, W. F.; Stewart, P. W.; Andreoni, S.; Seal, E., Jr.; Kehrl, H. R.; Horstman, D. H.; Folinsbee, L. J.; Smith,
43 M. V. (1997) Prediction of ozone-induced FEVj changes: effects of concentration, duration, and ventilation.
44 Am. J. Respir. Crit. Care Med. 156: 715-722.
45 McDonnell, W. F.; Stewart, P. W.; Smith, M. V.; Pan, W. K.; Pan, J. (1999) Ozone-induced respiratory symptoms:
46 exposure-response models and association with lung function. Eur. Respir. J. 14: 845-853.
47 McGee, M. P.; Devlin, R.; Saluta, G.; Koren, H. (1990) Tissue factor and factor VII messenger RNAs in human
48 alveolar macrophages: effects of breathing ozone. Blood 75: 122-127.
49 McKinnon, K. P.; Madden, M. C.; Noah, T. L.; Devlin, R. B. (1993) In vitro ozone exposure increases release of
50 arachidonic acid products from a human bronchial epithelial cell line. Toxicol. Appl. Pharmacol.
51 118:215-223.
52 Melton, C. E. (1990) Airliner cabin ozone: an updated review. Washington, DC: Federal Aviation Administration,
53 Office of Aviation Medicine; report no. DOT/FAA/AM-89/13. Available from: NTIS, Springfield, VA;
54 AD-A219 264.
55 Messineo, T. D.; Adams, W. C. (1990) Ozone inhalation effects in females varying widely in lung size: comparison
56 with males. J. Appl. Physiol. 69: 96-103.
January 2005 AX6-144 DRAFT-DO NOT QUOTE OR CITE
-------
1 Michelson, P. H.; Dailey, L.; Devlin, R. B.; Peden, D. B. (1999) Ozone effects on the immediate-phase response to
2 allergen in the nasal airways of allergic asthmatic subjects. Otolaryngol. Head Neck Surg. 120: 225-232.
3 Mohammed, S. P.; Higenbottam, T. W.; Adcock, J. J. (1993) Effects of aerosol-applied capsaicin, histamine and
4 prostaglandin-E2 on airway sensory receptors of anaesthetized cats. J. Physiol. Lond. 46: 951-966.
5 Molfino, N. A.; Wright, S. C.; Katz, L; Tarlo, S.; Silverman, F.; McClean, P. A.; Szalai, J. P.; Raizenne, M; Slutsky,
6 A. S.; Zamel, N. (1991) Effect of low concentrations of ozone on inhaled allergen responses in asthmatic
7 subjects. Lancet 338(8761): 199-203.
8 Molfino, N. A.; Slutsky, A. S.; Zamel, N. (1992) The effects of air pollution on allergic bronchial responsiveness.
9 Clin. Exp. Allergy 22: 667-672.
10 Montuschi, P.; Nightingale, J. A.; Kharitonov, S. A.; Barnes, P. J. (2002) Ozone-induced increase in exhaled
11 8-isoprostane in healthy subjects is resistant to inhaled budesonide. Free Radical Biol. Med. 33: 1403-1408.
12 Mudway, I. S.; Kelly, F. J. (2000) Ozone and the lung: a sensitive issue. Mol. Aspects. Med. 21: 1-48.
13 Mudway, I. S.; Blomberg, A.; Frew, A. J.; Holgate, S. T.; Sandstrom, T.; Kelly, F. J. (1999) Antioxidant
14 consumption and repletion kinetics in nasal lavage fluid following exposure of healthy human volunteers to
15 ozone. Eur. Respir. J. 13: 1429-1438.
16 Mudway, I. S.; Stenfors, N.; Blomberg, A.; Helleday, R.; Dunster, C.; Marklund, S. L.; Frew, A. J.; Sandstrom, T.;
17 Kelly, F. J. (2001) Differences in basal airway antioxidant concentrations are not predictive of individual
18 responsiveness to ozone: a comparison of healthy and mild asthmatic subjects. Free Radical Biol. Med.
19 31:962-974.
20 Nagda, N. L.; Fortmann, R. C.; Koontz, M. D.; Baker, S. R.; Ginevan, M. E. (1989) Airliner cabin environment:
21 contaminant measurements, health risks, and mitigation options. Washington, DC: U.S. Department of
22 Transportation, Office of the Secretary. Available from: NTIS, Springfield, VA; PB91 -1593 84.
23 National Heart, Lung, and Blood Institute (NHLBI). (1997) Guidelines for the diagnosis and management of asthma:
24 expert panel report 2. Bethesda, MD: U.S. Department of Health and Human Services, National Institutes of
25 Health; publication no. 97-4051. Available: http://www.nhlbi.nih.gov/guidelines/asthma/asthgdln.pdf
26 (11 April 2003).
27 National Heart, Lung, and Blood Institute (NHLBI). (2003) Expert panel report: guidelines for the diagnosis and
28 management of asthma. Update on selected topics 2002. Bethesda, MD: U.S. Department of Health and
29 Human Services, National Institutes of Health; NIH publication no. 02-5074. Available:
30 http://www.nhlbi.nih.gov/guidelines/asthma/asthmafullrpt.pdf [26 August, 2004].
31 Neas, L. M.; Dockery, D. W.; Koutrakis, P.; Tollerud, D. J.; Speizer, F. E. (1995) The association of ambient air
32 pollution with twice daily peak expiratory flow rate measurements in children. Am. J. Epidemiol.
33 141: 111-122.
34 Newson, E. J.; Krishna, M. T.; Lau, L. C. K.; Howarth, P. H.; Holgate, S. T.; Frew, A. J. (2000) Effects of
3 5 short-term exposure to 0.2 ppm ozone on biomarkers of inflammation in sputum, exhaled nitric oxide, and
36 lung function in subjects with mild atopic asthma. J. Occup. Environ. Med. 42: 270-277.
37 Nichols, B. G.; Woods, J. S.; Luchtel, D. L.; Corral, J.; Koenig, J. Q. (2001) Effects of ozone exposure on nuclear
3 8 factor-KB activation and tumor necrosis factor-a expression in human nasal epithelial cells. Toxicol. Sci.
39 60:356-362.
40 Nightingale, J. A.; Rogers, D. F.; Barnes, P. J. (1999) Effect of inhaled ozone on exhaled nitric oxide, pulmonary
41 function, and induced sputum in normal and asthmatic subjects. Thorax 54: 1061-1069.
42 Nightingale, J. A.; Rogers, D. F.; Chung, K. F.; Barnes, P. J. (2000) No effect of inhaled budesonide on the response
43 to inhaled ozone in normal subjects. Am. J. Respir. Crit. Care Med. 161: 479-486.
44 Niinimaa, V.; Cole, P.; Mintz, S.; Shephard, R. J. (1980) The switching point from nasal to oronasal breathing.
45 Respir. Physiol. 42: 61-71.
46 Nikasinovic, L.; Momas, I.; Seta, N. (2003) Nasal epithelial and inflammatory response to ozone exposure: a review
47 of laboratory-based studies published since 1985. J. Toxicol. Environ. Health B 6: 521-568.
48 Passannante, A. N.; Hazucha, M. J.; Bromberg, P. A.; Seal, E.; Folinsbee, L.; Koch, G. (1998) Nociceptive
49 mechanisms modulate ozone-induced human lung function decrements. J. Appl. Physiol. 85: 1863-1870.
50 Peden, D. B. (2001a) Air pollution in asthma: effect of pollutants on airway inflammation. Ann. Allergy Asthma
51 Immunol. 87(suppl. 3): 12-17.
52 Peden, D. B. (200 Ib) Effect of pollutants in rhinitis. Curr. Allergy Asthma Rep. 1: 242-246.
53 Peden, D. B.; Setzer, R. W., Jr.; Devlin, R. B. (1995) Ozone exposure has both a priming effect on allergen-induced
54 responses and an intrinsic inflammatory action in the nasal airways of perennially allergic asthmatics. Am. J.
55 Respir. Crit. Care Med. 151: 1336-1345.
January 2005 AX6-145 DRAFT-DO NOT QUOTE OR CITE
-------
1 Peden, D. B.; Boehlecke, B.; Horstman, D.; Devlin, R. (1997) Prolonged acute exposure to 0.16 ppm ozone induces
2 eosinophilic airway inflammation in asthmatic subjects with allergies. J. Allergy Clin. Immunol.
3 100:802-808.
4 Peters, E. A.; Hiltermann, J. T.; Stolk, J. (2001) Effect of apocynin on ozone-induced airway hyperresponsiveness to
5 methacholine in asthmatics. Free Radical Biol. Med. 31: 1442-1447.
6 Rayman, R. B. (2002) Cabin air quality: an overview. Aviat. Space Environ. Med. 73: 211-215.
7 Reisenauer, C. S.; Koenig, J. Q.; McManus, M. S.; Smith, M. S.; Kusic, G.; Pierson, W. E. (1988) Pulmonary
8 response to ozone exposures in healthy individuals aged 55 years or greater. JAPCA 38: 51-55.
9 Riediker, M.; Monn, C.; Koller, T.; Stahel, W. A.; Wuthrich, B. (2001) Air pollutants enhance rhinoconjunctivitis
10 symptoms in pollen-allergic individuals. Ann. Allergy Asthma Immunol. 87: 311-318.
11 Rigas, M. L.; Ben-Jebria, A.; Ultman, J. S. (1997) Longitudinal distribution of ozone absorption in the lung: effects
12 of nitrogen dioxide, sulfur dioxide, and ozone exposures. Arch. Environ. Health 52: 173-178.
13 Rombout, P. J. A.; Dormans, J. A. M. A.; Marra, M.; VanEsch, G. J. (1986) Influence of exposure regimen on
14 nitrogen dioxide-induced morphological changes in the rat lung. Environ. Res. 41: 466-480.
15 Romieu, L; Sienra-Monge, J. J.; Ramirez-Aguilar, M.; Moreno-Macias, H.; Reyes-Ruiz, N. L; Estela del
16 Rio-Navarro, B.; Hernandez-Avila, M.; London, S. J. (2004) Genetic polymorphism of GSTM1 and
17 antioxidant supplementation influence lung function in relation to ozone exposure in asthmatic children in
18 Mexico City. Thorax 59: 8-10.
19 Samet, J. M.; Hatch, G. E.; Horstman, D.; Steck-Scott, S.; Arab, L.; Bromberg, P. A.; Levine, M.; McDonnell,
20 W. F.; Devlin, R. B. (2001) Effect of antioxidant supplementation on ozone-induced lung injury in human
21 subjects. Am. J. Respir. Crit. Care Med. 164: 819-825.
22 Savin, W. M.; Adams, W. C. (1979) Effects of ozone inhalation on work performance and VO2max. J. Appl. Physiol.:
23 Respir. Environ. Exercise Physiol. 46: 309-314.
24 Scannell, C.; Chen, L.; Aris, R. M.; Tager, I.; Christian, D.; Ferrando, R.; Welch, B.; Kelly, T.; Balmes, J. R. (1996)
25 Greater ozone-induced inflammatory responses in subjects with asthma. Am. J. Respir. Crit. Care Med.
26 154:24-29.
27 Schelegle, E. S.; Adams, W. C. (1986) Reduced exercise time in competitive simulations consequent to low level
28 ozone exposure. Med. Sci. Sports Exercise 18: 408-414.
29 Schelegle, E. S.; Adams, W. C.; Siefkin, A. D. (1987) Indomethacin pretreatment reduces ozone-induced pulmonary
30 function decrements in human subjects. Am. Rev. Respir. Dis. 136: 1350-1354.
31 Schelegle, E. S.; Adams, W. C.; Giri, S. N.; Siefkin, A. D. (1989) Acute ozone exposure increases plasma
32 prostaglandin F2a in ozone-sensitive human subjects. Am. Rev. Respir. Dis. 140: 211-216.
33 Schelegle, E. S.; Siefkin, A. D.; McDonald, R. J. (1991) Time course of ozone-induced neutrophilia in normal
34 humans. Am. Rev. Respir. Dis. 143: 1353-1358.
35 Schelegle, E. S.; Carl, M. L.; Coleridge, H. M.; Coleridge, J. C. G.; Green, J. F. (1993) Contribution of vagal
36 afferents to respiratory reflexes evoked by acute inhalation of ozone in dogs. J. Appl. Physiol. 74: 2338-2344.
37 Schelegle, E. S.; Eldridge, M. W.; Cross, C. E.; Walby, W. F.; Adams, W. C. (2001) Differential effects of airway
38 anesthesia on ozone-induced pulmonary responses in human subjects. Am. J. Respir. Crit. Care Med.
39 163: 1121-1127.
40 Schierhorn, K.; Zhang, M.; Matthias, C.; Kunkel, G. (1999) Influence of ozone and nitrogen dioxide on histamine
41 and interleukin formation in a human nasal mucosa culture system. Am. J. Respir. Cell Mol. Biol.
42 20: 1013-1019.
43 Schierhorn, K.; Hanf, G.; Fischer, A.; Umland, B.; Olze, H.; Kunkel, G. (2002) Ozone-induced release of
44 neuropeptides from human nasal mucosa cells. Int. Arch. Allergy Immunol. 129: 145-151.
45 Schonfeld, B. R.; Adams, W. C.; Schelegle, E. S. (1989) Duration of enhanced responsiveness upon re-exposure to
46 ozone. Arch. Environ. Health 44: 229-236.
47 Schwartz, L. W.; Dungworth, D. L.; Mustafa, M. G.; Tarkington, B. K.; Tyler, W. S. (1976) Pulmonary responses of
48 rats to ambient levels of ozone: effects of 7-day intermittent or continuous exposure. Lab. Invest. 34: 565-578.
49 Seal, E., Jr.; McDonnell, W. F.; House, D. E.; Salaam, S. A.; Dewitt, P. J.; Butler, S. O.; Green, J.; Raggio, L.
50 (1993) The pulmonary response of white and black adults to six concentrations of ozone. Am. Rev. Respir.
51 Dis. 147: 804-810.
52 Seal, E., Jr.; McDonnell, W. F.; House, D. E. (1996) Effects of age, socioeconomic status, and menstrual cycle on
53 pulmonary response to ozone. Arch. Environ. Health 51: 132-137.
54 Sekizawa, S. I.; Tsubone, H. (1994) Nasal receptors responding to noxious chemical irritants. Respir. Physiol.
55 96:37-48.
January 2005 AX6-146 DRAFT-DO NOT QUOTE OR CITE
-------
1 Seltzer, I; Bigby, B. G.; Stulbarg, M; Holtzman, M. I; Nadel, J. A.; Ueki, I. F.; Leikauf, G. D.; Goetzl, E. J.;
2 Boushey, H. A. (1986) O3-induced change in bronchial reactivity to methacholine and airway inflammation in
3 humans. J. Appl. Physiol. 60: 1321-1326.
4 Solway, J.; Leff, A. R. (1991) Sensory neuropeptides and airway function. J. Appl. Physiol. 71: 2077-2087.
5 Spannhake, E. W.; Reddy, S. P. M.; Jacoby, D. B.; Yu, X.-Y.; Saatian, B.; Tian, J. (2002) Synergismbetween
6 rhinovirus infection and oxidant pollutant exposure enhances airway epithelial cell cytokine production.
7 Environ. Health Perspect. 110: 665-670.
8 Spit, B. J.; Bretschneider, F.; Hendriksen, E. G. J.; Kuper, C. F. (1993) Ultrastructure of free nerve endings in
9 respiratory and squamous epithelium on the rat nasal septum. Cell Tissue Res. 274: 329-335.
10 Steck-Scott, S.; Arab, L.; Craft, N. E.; Samet, J. M. (2004) Plasma and lung macrophage responsiveness to
11 carotenoid supplementation and ozone exposure in humans. Eur. J. Clin. Nutr. 1-9.
12 Stenfors, N.; Pourazar, J.; Blomberg, A.; Krishna, M. T.; Mudway, L; Helleday, R.; Kelly, F. J.; Frew, A. J.;
13 Sandstrom, T. (2002) Effect of ozone on bronchial mucosal inflammation in asthmatic and healthy subjects.
14 Respir. Med. 96: 352-358.
15 Sterk, P. J.; Fabbri, L. M.; Quanjer, P. H.; Cockcroft, D. W.; O'Byrne, P. M.; Anderson, S. D.; Juniper, E. F.; Malo,
16 J.-L. (1993) Airway responsiveness. Standardized challenge testing with pharmacological, physical and
17 sensitizing stimuli in adults. Report Working Party Standardization of Lung Function Tests, European
18 Community for Steel and Coal. Eur. Respir. J. 6(suppl. 16): 53-83.
19 Stevens, W. H. M.; Adelroth, E.; Wattie, J.; Woolley, M. J.; Ellis, R.; Dahlback, M.; O'Byrne, P. M. (1994) Effect of
20 inhaled budesonide on ozone-induced airway hyperresponsiveness and bronchoalveolar lavage cells in dogs.
21 J. Appl. Physiol. 77: 2578-2583.
22 Superko, H. R.; Adams, W. C.; Daly, P. W. (1984) Effects of ozone inhalation during exercise in selected patients
23 with heart disease. Am. J. Med. 77: 463-470.
24 Tepper, J. S.; Costa, D. L.; Lehmann, J. R.; Weber, M. F.; Hatch, G. E. (1989) Unattenuated structural and
25 biochemical alterations in the rat lung during functional adaptation to ozone. Am. Rev. Respir. Dis.
26 140:493-501.
27 Tepper, J. S.; Wiester, M. J.; Weber, M. F.; Menache, M. G. (1990) Measurements of cardiopulmonary response in
28 awake rats during acute exposure to near-ambient concentrations of ozone. Fundam. Appl. Toxicol. 10: 7-15.
29 Torres, A.; Utell, M. J.; Morow, P. E.; Voter, K. Z.; Whitin, J. C.; Cox, C.; Looney, R. J.; Speers, D. M.; Tsai, Y.;
30 Frampton, M. W. (1997) Airway inflammation in smokers and nonsmokers with varying responsiveness to
31 ozone. Am. J. Respir. Crit. Care Med. 156: 728-736.
32 Trenga, C. A.; Koenig, J. Q.; Williams, P. V. (2001) Dietary antioxidants and ozone-induced bronchial
33 hyperresponsiveness in adults with asthma. Arch. Environ. Health 56: 242-249.
34 U.S. Environmental Protection Agency. (1986) Air quality criteria for ozone and other photochemical oxidants.
3 5 Research Triangle Park, NC: Office of Health and Environmental Assessment, Environmental Criteria and
36 Assessment Office; report nos. EPA-600/8-84-020aF-eF. 5v. Available from: NTIS, Springfield, VA;
37 PB87-142949.
38 U.S. Environmental Protection Agency. (1996) Air quality criteria for ozone and related photochemical oxidants.
3 9 Research Triangle Park, NC: Office of Research and Development; report nos. EPA/600/AP-93/004aF-cF. 3v.
40 Available from: NTIS, Springfield, VA; PB96-185582, PB96-185590, and PB96-185608. Available:
41 www.epa.gov/ncea/ozone.htm.
42 Ultman, J. S.; Ben-Jebria, A.; Arnold, S. F. (2004) Uptake distribution of ozone in human lungs: intersubject
43 variability in physiologic response. Boston, MA: Health Effects Institute.
44 Utell, M. J.; Frampton, M. W.; Morrow, P. E.; Cox, C.; Levy, P. C.; Speers, D. M.; Gibb, F. R. (1994) Oxidant and
45 acid aerosol exposure in healthy subjects and subjects with asthma. Part II: effects of sequential sulfuric acid
46 and ozone exposures on the pulmonary function of healthy subjects and subjects with asthma. Cambridge,
47 MA: Health Effects Institute; pp. 37-93; research report no. 70.
48 Vagaggini, B.; Carnevali, S.; Macchioni, P.; Taccola, M.; Fornai, E.; Bacci, E.; Bartoli, M. L.; Cianchetti, S.; Dente,
49 F. L.; Di Franco, A.; Giannini, D.; Paggiaro. P. L. (1999) Airway inflammatory response to ozone in subjects
50 with different asthma severity. Eur. Respir. J. 13: 274-280.
51 Vagaggini, B.; Taccola, M.; Conti, I.; Carnevali, S.; Cianchetti, S.; Bartoli, M. L.; Bacci, E.; Dente, F. L.;
52 Di Franco, A.; Giannini, D.; Paggiaro, P. L. (2001) Budesonide reduces neutrophilic but not functional airway
53 response to ozone in mild asthmatics. Am. J. Respir. Crit. Care Med. 164: 2172-2176.
54 Vagaggini, B.; Taccola, M.; Cianchetti, S.; Carnevali, S.; Bartoli, M. L.; Bacci, E.; Dente, F. L.; Di Franco, A.;
55 Giannini, D.; Paggiaro, P. L. (2002) Ozone exposure increases eosinophilic airway response induced by
56 previous allergen challenge. Am. J. Respir. Crit. Care Med. 166: 1073-1077.
January 2005 AX6-147 DRAFT-DO NOT QUOTE OR CITE
-------
1 Van Bree, L.; Rombout, P. J. A.; Rietjens, I. M. C. M; Dormans, J. A. M. A.; Marra, M. (1989) Pathobiochemical
2 effects in rat lung related to episodic ozone exposure. In: Schneider, T.; Lee, S. D.; Wolters, G. J. R.; Grant,
3 L. D., eds. Atmospheric ozone research and its policy implications: proceedings of the 3rd US-Dutch
4 international symposium; May 1988; Nijmegen, The Netherlands. Amsterdam, The Netherlands: Elsevier
5 Science Publishers; pp. 723-732. (Studies in environmental science 35).
6 Vender, R. L.; Horstman, D. H.; Mangione, S. (1994) Red blood cell antioxidants in human volunteers exposed to
7 ozone. Toxicol. Ind. Health 10: 53-58.
8 Voter, K. Z.; Whitin, J. C.; Torres, A.; Morrow, P. E.; Cox, C.; Tsai, Y.; Utell, M. J.; Frampton, M. W. (2001)
9 Ozone exposure and the production of reactive oxygen species by bronchoalveolar cells in humans. Inhalation
10 Toxicol. 13:465-483.
11 Wang, G.; Umstead, T. M.; Phelps, D. S.; Al-Mondhiry, H.; Floras, J. (2002) The effect of ozone exposure on the
12 ability of human surfactant protein A variants to stimulate cytokine production. Environ. Health Perspect.
13 110:79-84.
14 Wayne, W. S.; Wehrle, P. F.; Carroll, R. E. (1967) Oxidant air pollution and athletic performance. JAMA J. Am.
15 Med. Assoc. 199: 151-154.
16 Weinmann, G. G.; Weidenbach-Gerbase, M.; Foster, W. M.; Zacur, H.; Frank, R. (1995) Evidence for
17 ozone-induced small-airway dysfunction: lack of menstrual-cycle and gender effects. Am. J. Respir. Crit.
18 Care Med. 152: 988-996.
19 Weymer, A. R.; Gong, H., Jr.; Lyness, A.; Linn, W. S. (1994) Pre-exposure to ozone does not enhance orproduce
20 exercise-induced asthma. Am. J. Respir. Crit. Care Med. 149: 1413-1419.
21 Yeadon, M.; Wilkinson, D.; Darley-Usmar, V.; O'Leary, V. J.; Payne, A. N. (1992) Mechanisms contributing to
22 ozone-induced bronchial hyperreactivity in guinea-pigs. Pulm. Pharmacol. 5: 39-50.
23 Ying, R. L.; Gross, K. B.; Terzo, T. S.; Eschenbacher, W. L. (1990) Indomethacin does not inhibit the
24 ozone-induced increase in bronchial responsiveness in human subjects. Am. Rev. Respir. Dis. 142: 817-821.
25 Yu, M.; Pinkerton, K. E.; Witschi, H. (2002) Short-term exposure to aged and diluted sidestream cigarette smoke
26 enhances ozone-induced lung injury inB6C3Fl mice. Toxicol. Sci. 65: 99-106.
27 Zhang, L.-Y.; Levitt, R. C.; Kleeberger, S. R. (1995) Differential susceptibility to ozone-induced airways
28 hyperreactivity in inbred strains of mice. Exp. Lung Res. 21: 503-518.
29
January 2005 AX6-148 DRAFT-DO NOT QUOTE OR CITE
-------
i 7. EPIDEMIOLOGICAL STUDIES OF
2 HUMAN HEALTH EFFECTS ASSOCIATED
3 WITH AMBIENT OZONE EXPOSURE
4
5
6 7.1 INTRODUCTION
7 This chapter evaluates current epidemiologic literature on health and physiological effects
8 of ambient O3 exposure. Epidemiologic studies linking community ambient O3 concentrations to
9 health effects were reported in the 1996 Ozone Air Quality Criteria Document (O3 AQCD; U.S.
10 Environmental Protection Agency, 1996a). Many of those studies reported that pulmonary
11 function decrements, hospital and emergency department admissions, and respiratory symptoms
12 in human populations were associated with ambient levels of O3. Numerous more recent
13 epidemiologic studies discussed in this chapter evaluate the relationship of ambient O3 to
14 morbidity and mortality, and thereby provide an expanded basis for assessment of health effects
15 associated with exposures to O3 at concentrations currently encountered in the U.S.
16 As discussed elsewhere in this document (Chapters 5 and 6), a substantial amount of
17 experimental evidence links O3 exposure unequivocally with respiratory effects in laboratory
18 animals and humans. These include structural changes in the bronchiolar-alveolar transition
19 (centriacinar) region of the lung, biochemical evidence of acute cellular/tissue injury,
20 inflammation, increased frequency and severity of experimental bacterial infection, and
21 temporary reductions in mechanical lung function. These effects have been observed with
22 exposure to O3 at ambient or near-ambient concentrations. Thus, many of the reported
23 epidemiologic associations of ambient O3 with respiratory health effects have considerable
24 biological credibility. Accordingly, the new epidemiologic studies of ambient O3 assessed here
25 are best considered in combination with information from the other chapters on ambient O3
26 concentration and exposure (Chapter 3), and toxicological effects of O3 in animals and humans
27 (Chapters 5 and 6, respectively). The epidemiologic studies constitute important information on
28 associations between health effects and exposures of human populations to "real-world" O3 and
29 also help to identify susceptible subgroups and associated risk factors.
30
January 2005 7-1 DRAFT-DO NOT QUOTE OR CITE
-------
1 7.1.1 Approach to Identifying O3 Epidemiologic Studies
2 Numerous O3 epidemiologic papers have been published since completion of the 1996 O3
3 AQCD. The U.S. Environmental Protection Agency (NCEA-RTP) has implemented a
4 systematic approach to identify relevant epidemiologic studies for consideration in this chapter.
5 In general, an ongoing search has been employed in conjunction with other strategies to identify
6 O3 epidemiology literature pertinent to developing criteria for O3 National Ambient Air Quality
7 Standards (NAAQS). A publication base was established using Medline, Pascal, BIOSIS, NTIS,
8 and Embase, and a set of search terms proven by prior use to identify pertinent literature. The
9 search strategy was reexamined and modified to enhance identification of published papers.
10 PubMed was added to the search regime.
11 While the above search regime provided good coverage of the relevant literature,
12 additional approaches augmented the traditional search methods. First, a Federal Register
13 Notice was issued requesting information and published papers from the public at large. Next,
14 non-EPA chapter authors, expert in this field, identified literature on their own. NCEA-RTP
15 staff also identified publications as an element of their assessment and interpretation of the
16 literature. Finally, additional potentially relevant publications will be included following
17 external review as a result of comments from both the public and CAS AC. The combination of
18 these approaches is believed to produce a comprehensive collection of studies appropriate for
19 review and assessment here. The principal objective criteria used for selecting literature for the
20 present assessment is to include all identified studies that evaluated the relationship between
21 measured ambient O3 levels and a human health outcome. All new studies published through
22 October 2004, as identified using the search, have been included in this AQCD and additional
23 efforts have been made to assess more recent studies.
24
25 7.1.2 Approach to Assessing Epidemiologic Evidence
26 Definitions of the various types of epidemiologic studies assessed have been provided in an
27 earlier PM AQCD (U.S. Environmental Protection Agency, 1996b). Briefly, epidemiologic
28 studies are generally divided into two groups, morbidity studies and mortality studies. Morbidity
29 studies evaluate O3 effects on a wide range of health endpoints, including the following:
30 changes in pulmonary function, respiratory symptoms, and self-medication in asthmatics;
31 respiratory- and cardiovascular-related emergency department visits and hospital admissions;
January 2005 7-2 DRAFT-DO NOT QUOTE OR CITE
-------
1 and changes in cardiovascular physiology/functions and airway inflammation. Mortality studies
2 investigate O3 effects on total (nonaccidental) mortality and cause-specific mortality, providing
3 evidence related to a clearly adverse endpoint. The epidemiologic strategies most commonly
4 used in O3 health studies are of four types: (1) ecologic studies; (2) time-series semi-ecologic
5 studies; (3) prospective cohort studies; and (4) case-control and crossover studies. All of these
6 are observational studies rather than experimental studies.
7 The approach to assessing epidemiologic evidence has been eloquently stated most
8 recently in the 2004 PM AQCD (U.S. Environmental Protection Agency, 2004a) and is adapted
9 here. The critical assessment of epidemiologic evidence presented in this chapter is conceptually
10 based upon consideration of salient aspects of the evidence of associations so as to reach
11 fundamental judgments as to the likely causal significance of the observed associations. In so
12 doing, it is appropriate to draw from those aspects initially presented in Hill's classic monograph
13 (Hill, 1965) and widely used by the scientific community in conducting such evidence-based
14 reviews. A number of these aspects are judged to be particularly salient in evaluating the body
15 of evidence available in this review, including the aspects described by Hill as strength,
16 experiment, consistency, plausibility, and coherence. Other aspects identified by Hill, including
17 temporality and biological gradient, are also relevant and considered here (e.g., in characterizing
18 lag structures and concentration-response relationships), but are more directly addressed in the
19 design and analyses of the individual epidemiologic studies included in this assessment.
20 As discussed below, these salient aspects are interrelated and considered throughout the
21 evaluation of the epidemiologic evidence presented in this chapter, and are more generally
22 reflected in the integrative synthesis presented in Chapter 8 of this AQCD.
23 In the following sections, the general evaluation of the strength of the epidemiological
24 evidence reflects consideration not only of the magnitude of reported O3 effect estimates and
25 their statistical significance, but also of the precision of the effect estimates and the robustness of
26 the effects associations. Consideration of the robustness of the associations takes into account a
27 number of factors, including in particular the impact of alternative models and model
28 specifications and potential confounding by copollutants, as well as issues related to the
29 consequences of measurement error.
30 Consideration of the consistency of the effects associations, as discussed in the following
31 sections, involves looking across the results of multi- and single-city studies conducted by
January 2005 7-3 DRAFT-DO NOT QUOTE OR CITE
-------
1 different investigators in different places and times. Relevant factors are known to exhibit much
2 variation across studies, including, for example, the presence and levels of copollutants, the
3 relationships between central measures of O3 and exposure-related factors, relevant demographic
4 factors related to sensitive subpopulations, and climatic and meteorological conditions. Thus, in
5 this case, consideration of consistency and the related heterogeneity of effects are appropriately
6 understood as an evaluation of the similarity or general concordance of results, rather than an
7 expectation of finding quantitative results within a very narrow range.
8 Looking beyond the epidemiological evidence, evaluation of the biological plausibility of
9 the O3-health effects associations observed in epidemiologic studies reflects consideration of
10 both exposure-related factors and dosimetric/toxicologic evidence relevant to identification of
11 potential biological mechanisms. Similarly, coherence of health effects associations reported in
12 the epidemiologic literature reflects consideration of information pertaining to the nature of the
13 various respiratory- and cardiac-related mortality and morbidity effects and biological markers
14 evaluated in toxicologic and human clinical studies. These broader aspects of the assessment are
15 only touched upon in this chapter but are more fully integrated in the discussion presented in
16 Chapters.
17 In identifying these aspects as being particularly salient in this assessment, it is also
18 important to recognize that no one aspect is either necessary or sufficient for drawing inferences
19 of causality. As Hill (1965) emphasized:
20
21 None of my nine viewpoints can bring indisputable evidence for or against the cause-
22 and-effect hypothesis and none can be required as a sine qua non. What they can do,
23 with greater or less strength, is to help us to make up our minds on the fundamental
24 question — is there any other way of explaining the set of facts before us, is there
25 any other answer equally, or more, likely than cause and effect?
26
27 Thus, while these aspects frame considerations weighed in assessing the epidemiologic evidence,
28 they do not lend themselves to being considered in terms of simple formulas or hard-and-fast
29 rules of evidence leading to answers about causality (Hill, 1965). One, for example, cannot
30 simply count up the numbers of studies reporting statistically significant results for the various
31 O3 indicators and health endpoints evaluated in this assessment and reach credible conclusions
32 about the relative strength of the evidence and the likelihood of causality. Rather, these
33 important considerations are taken into account and discussed throughout this assessment with
January 2005 7-4 DRAFT-DO NOT QUOTE OR CITE
-------
1 the goal of producing an objective appraisal of the evidence, informed by peer and public
2 comment and advice, including weighing of alternative views on controversial issues, leading to
3 conclusions and inferences that reflect the best judgements of the scientists engaged in this
4 review.
5 In assessing the relative scientific quality of epidemiologic studies reviewed here and to
6 assist in interpreting their findings, the following considerations were taken into account:
7 (1) To what extent are the aerometric data/exposure metrics used of adequate quality and
sufficiently representative to serve as credible exposure indicators, well-reflecting
geographic or temporal differences in study population pollutant exposures in the
range(s) of pollutant concentrations evaluated?
8 (2) Were the study populations well-defined and adequately selected so as to allow for
meaningful comparisons between study groups or meaningful temporal analyses of
health effects results?
9 (3) Were the health endpoint measurements meaningful and reliable, including clear
definition of diagnostic criteria utilized and consistency in obtaining dependent
variable measurements?
10 (4) Were the statistical analyses used appropriate, and properly performed and interpreted,
including accurate data handling and transfer during analyses?
11 (5) Were likely important covariates (e.g., potential confounders or effect modifiers)
adequately controlled for or taken into account in the study design and statistical
analyses?
12 (6) Were the reported findings internally consistent, biologically plausible, and coherent
in terms of consistency with other known facts?
13 These guidelines provide benchmarks for judging the relative quality of various studies and
14 in assessing the body of epidemiologic evidence. Detailed critical analysis of all epidemiologic
15 studies on O3 health effects, especially in relation to all of the above questions, is beyond the
16 scope of this document. Of most importance for present purposes are those studies which
17 provide useful qualitative or quantitative information on exposure-response relationships for
18 health effects associated with ambient air levels of O3 likely to be encountered in the U. S. among
19 healthy and susceptible populations.
20
21
January 2005 7-5 DRAFT-DO NOT QUOTE OR CITE
-------
1 7.1.3 Study Designs and Analysis Methods Used to Assess O3 Health Effects
2 Prior to discussing results from the recent O3 studies, issues and questions arising from the
3 study designs and analysis methods used in the assessment of O3 effect estimates will be briefly
4 presented. Air pollution time-series studies in particular have design and analysis aspects that
5 complicate the interpretation of O3 health effects. Analyses using administrative data (e.g.,
6 numbers of deaths and emergency hospital admissions) have inherent limitations as well as
7 strengths (Virnig and McBean, 2001), however in this section we focus mainly on the topics of
8 exposure assessment and model specification in time-series or longitudinal studies. Potential
9 biases that may result from O3 exposure measurement error, and choice of exposure index and
10 lag period are first presented. A discussion of model specification issues and potential
11 confounding by temporal factors, meteorological effects, seasonal trends, and copollutants
12 follow. Integrative discussion of these topics is presented later in Section 7.6.
13
14 7.1.3.1 Exposure Assessment in Epidemiologic Studies
15 In general, the exposure of the participant is not directly observed, and the concentration of
16 O3 and other air pollutants at one or more stationary air monitors is used as a proxy for
17 individual exposure to ambient air pollution. In an ideal situation, studies of air pollution health
18 effects would be conducted at the individual level, with information on personal exposure to the
19 various pollutants. However, determining accurate personal exposure information is difficult
20 and often impractical. In many epidemiologic studies, especially time-series studies with
21 administrative data on mortality and hospitalization outcomes, data from central ambient
22 monitoring sites often are used as the estimate of exposure. Routinely collected ambient data,
23 though readily available and convenient, may not represent true personal exposure. The use of
24 ambient data tends to underestimate the effect of the air pollutant on health (Krzyzanowski,
25 1997). As discussed thoroughly in the 2004 PM AQCD (Section 8.4.5), the resulting exposure
26 measurement error and its effect on the estimates of relative risk must be considered. In theory,
27 there are three components to exposure measurement error: (1) the use of average population
28 rather than individual exposure data; (2) the difference between the average personal ambient
29 exposure and the ambient concentrations; and (3) the difference between the true and measured
30 ambient concentrations. Zeger et al. (2000) indicated that the first and third error components
31 are largely Berksonian errors and would not significantly bias the risk estimate. However, the
January 2005 7-6 DRAFT-DO NOT QUOTE OR CITE
-------
1 second error component resulting from the difference between the average personal ambient
2 exposure and ambient concentration levels might introduce bias, especially if indoor sources are
3 associated with ambient levels.
4 Several studies measured O3 concentrations in a variety of indoor environments, including
5 homes (Lee et al., 2004), schools (Linn et al., 1996), and the workplace (Liu et al., 1995).
6 Indoor O3 concentrations were, in general, approximately one-tenth of the outdoor
7 concentrations in these studies. However, the specific contribution of indoor sources to indoor
8 O3 levels has not been investigated. Few indoor sources of O3 exist, possible sources being
9 office equipment (e.g., photocopiers, laser printers) and air cleaners. As described in Section 3.9
10 of this document, indoor O3 exposure primarily results from infiltration of O3 from the outdoors
11 through ventilation and is noted as minimal.
12 The impact of measurement error on O3 effect estimates was demonstrated in a study by
13 Navidi et al. (1999). In this study, a simulation was conducted using data from the University of
14 Southern California Children's Health Study of the long-term effects of air pollutants on
15 children. The effect estimate from computed "true" O3 exposure was compared to effect
16 estimates from exposure determined using several methods: (1) ambient stationary monitors;
17 (2) the microenvironmental approach (multiply concentrations in various microenvironments by
18 time present in each microenvironment); and (3) personal sampling. Effect estimates based on
19 all three exposure measures were biased towards the null. The bias that results when using the
20 microenvironmental and personal sampling approach is due to nondifferential measurement
21 error. Use of ambient monitors to determine exposure will tend to overestimate true personal O3
22 exposure (assumes that subjects are outdoors 100% of their time), thus generally their use will
23 result in effect estimates that are biased towards the null.
24
25 7.1.3.2 O3 Exposure Indices Used
26 The results of studies of mortality and morbidity health outcomes from exposure to O3 are
27 usually presented in this document as a relative risk, or risk rate relative to a baseline mortality
28 or morbidity rate. These relative risks are based on an incremental change in exposure. To
29 enhance comparability between studies, presenting these relative risks by a uniform exposure
30 increment is needed. However, determining a standard increment is complicated by the use of
31 different O3 exposure indices in the existing health studies. The three daily O3 exposure indices
January 2005 7-7 DRAFT-DO NOT QUOTE OR CITE
-------
1 that most often appear in the literature are 1-h average maximum (1-h max), 8-h average
2 maximum (8-h max), and 24-h average (24-h avg). As concentrations are lower and less
3 variable for the longer averaging times, relative risks of adverse health outcomes for a specific
4 numeric concentration range are not directly comparable across metrics. Using the nationwide
5 distributional data for O3 monitors in U.S. Metropolitan Statistical Areas, increments
6 representative of a low-to-high change in O3 concentrations were developed based on mean
7 and upper percentile values in the dataset (Langstaff, 2003):
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
Daily Exposure Index Exposure Increment (ppb)
1-h max O3
8-h max O3
24-h avg O3
40
30
20
In the following discussion sections, efforts were made to standardize the O3 excess risks using
these increments, except as noted, so that the risk estimates could be compared across studies.
7.1.3.3 Lag of O3 Exposure Used
Lags of exposure may reflect the distribution of effects across time in a population and the
potential mechanisms of effects. However, simply choosing the most significant exposure lag
may bias the air pollution risk estimates away from the null, as shown by a simulation by
Lumley and Sheppard (2000) that used PM2 5 as an example. This is especially true when the
choice is made from a large number of lags. Most of the O3 time-series studies examined
relatively small numbers of lagged days, typically 0 through 3 days, and/or cumulative lags
thereof (e.g., cumulative lag of 0 and 1 day). An examination of the "most significant" lags
suggests that the majority of the single-day associations were immediate (0-day lag), not a
random pattern in which associations could be observed on any of the lags examined with equal
probabilities. However, the lags may vary by health outcome, as some effects are delayed (e.g.,
airway inflammation) and are captured in longer lag periods. Note that when associations are
found at multiple days, presenting selected risk estimates from single-day lags may result in bias
(i.e., typically toward underestimation of magnitude of overall risk).
January 2005
DRAFT-DO NOT QUOTE OR CITE
-------
1 7.1.3.4 Model Specification Issues
2 The relationships between daily numbers of deaths and hospital admissions, and levels of
3 O3 and related environmental factors have been analyzed widely over the past decade, yielding
4 insights into the possible effects of O3 on acute exacerbations of respiratory and cardiovascular
5 diseases, and related mortality. These daily time-series studies exploit the high degree of day-to-
6 day variability in ambient air pollution concentrations to develop quantitative estimates of
7 impacts on daily health outcomes. The basic analytical approach used to estimate the effects of
8 O3 in this type of study is multiple regression. Because a given location is followed over time,
9 many factors that might confound a multicity cross-sectional study do not affect time-series
10 studies. Cross-sectional confounders include cigarette smoking, diet, occupation, and other risk
11 factors that may vary across cities in ways that correlate with variations in air pollution levels.
12 In contrast, these factors are unlikely to vary over time in a way that correlates with day-to-day
13 variations in air pollution, thus confounding by these factors is minimized in a time-series study.
14 Longer-term secular time trends, such as changes in morbidity due to improved clinical
15 management of disease, also generally do not present a confounder problem in time-series
16 studies because these trends are removed analytically. Other advantages of the daily time-series
17 study design include the relatively large sample sizes in terms of person-days and the readily
18 available data, making such studies convenient and economical to conduct in a wide variety of
19 locations.
20 However, several challenges present themselves with respect to designing and interpreting
21 time-series studies. The principal challenge facing the analyst in the daily time-series context is
22 avoiding bias due to confounding by short-term temporal factors operating over time scales from
23 days to seasons. In the current regression models used to estimate short-term effects of air
24 pollution, there are two major potential confounders that need to be considered: (1) seasonal
25 trend and other "long-wave" temporal trends; and (2) weather effects. Both of these variables
26 tend to predict a significant fraction of fluctuations in time-series. Unfortunately, both of these
27 terms are also highly correlated with O3, as O3 has strong seasonal cycles and is formed more at
28 higher temperatures. The correlation of O3 with these confounding terms tends to be higher than
29 that for PM or other gaseous pollutants. In the U.S., the mass concentration of PM25 generally
30 does not have strong seasonal cycles like O3 because PM2 5 tends to reflect both primary
31 emissions (throughout the year, but often higher in winter in most U.S. cities) and secondary
January 2005 7-9 DRAFT-DO NOT QUOTE OR CITE
-------
1 aerosols (higher in summer). Therefore, PM2 5 and O3 effect estimates from studies primarily
2 designed to examine PM2 5 health effects may not be comparable as model specifications that
3 may be appropriate for PM2 5 may not necessarily be adequate for O3. The following section
4 reviews the current methodologies used to control for potential confounding by temporal trends
5 and weather effects.
6
7 7.1.3.5 Controlling for Temporal Trends and Meteorologic Effects
8 An examination of recent time-series studies indicates that several types of fitting
9 approaches have been used to adjust for temporal trends and weather effects. The use of
10 parametric and nonparametric smoothers with varying degrees of freedom per year has emerged
11 as the prevailing approach. The use of larger degrees of freedom to adjust for potential
12 confounding by time-varying factors may inadvertently result in ascribing more effects to these
13 unmeasured potential confounders and take away the air pollution effect. Often smaller
14 pollution effect estimates are observed when more degrees of freedom are used. Currently, the
15 degrees of freedom used to adjust for temporal trends in time-series studies generally range from
16 4 to 12 degrees of freedom per year using either nonparametric or parametric smoothers.
17 Statistical diagnostics such as Akaike's Information Criteria, residual autocorrelation, or
18 dispersion of the regression model often are used to choose or evaluate the adequacy of the
19 degrees of freedom for temporal trend, but these diagnostics do not provide epidemiological
20 justification or interpretation of the fitted model.
21 The issue of model specifications to adjust for temporal trends and weather variables in
22 time-series studies was a consideration of several researchers that conducted sensitivity analyses
23 of PM data (HEI, 2003). The sensitivity of O3 coefficients to model specifications for temporal
24 trend adjustment has not been as well-studied. Only one recent multicity study examined the
25 sensitivity of O3 coefficients to the extent of smoothing for adjustment of temporal trends and
26 meteorologic factors (Bell et al., 2004). Most, if not all, O3 studies used the same model
27 specifications to estimate the excess risks for PM and other gaseous pollutants. The relationship
28 between a pollutant and the temporal trend or weather effect being fitted differs for each
29 pollutant, and interpretation of the excess risk estimates needs to take into consideration this
30 varying concurvity (nonlinear analogue of multiple correlation) across pollutants. As noted
31 above, O3 is expected to have the strongest correlation with both temporal (seasonal) trend and
January 2005 7-10 DRAFT-DO NOT QUOTE OR CITE
-------
1 weather effects. The strong annual cycle in O3 concentrations presents a unique problem in
2 time-series analyses where time trends are fitted simultaneously with pollution and other model
3 terms (i.e., co-adjustment). In this setting, the annual O3 cycle itself may compete with the
4 smooth function of time to explain some of the annual, cyclic behavior in the health outcome,
5 which can result in biased effect estimates for O3 when data for all seasons are analyzed
6 together.
7 Current weather models used in time-series analyses can be classified into: (1) quantile
8 (e.g., quartile, quintile) indicators; (2) parametric functional forms such as V- or U-shape
9 functions; and (3) parametric (e.g., natural splines) or nonparametric (e.g., locally estimated
10 smoothing splines [LOESS]) smoothing functions. More recent studies tend to use smoothing
11 functions. While these methods provide flexible ways to fit health outcomes as a function of
12 temperature and other weather variables, there are two major issues that need further
13 examination to enable more meaningful interpretation of O3 morbidity and mortality effects.
14 The first issue is the interpretation of weather or temperature effects. Most researchers
15 agree about the morbidity and mortality effects of extreme temperatures (i.e., heat waves or cold
16 spells). However, as extreme hot or cold temperatures, by definition, happen rarely, much of the
17 health effects occur in the mild or moderate temperature range. Given the significant correlation
18 between O3 and temperature, ascribing the association between temperature and health outcomes
19 solely to temperature effects may underestimate the effect of O3.
20 The second issue is that in most studies weather model specifications are fitted for year-
21 round data. Such models will ignore the correlation structure that can change across seasons,
22 resulting in inefficiency and model mis-specification. This is particularly important for O3,
23 which appears to change its relationship with temperature as well as with other pollutants across
24 seasons. Ambient O3 levels are typically higher in the summer or warm season, often referred to
25 as the O3 season. In the winter or colder months, O3 levels tend to be much lower compared to
26 the summer months. During the winter in some urban locations, O3 mainly comes from the free
27 troposphere and can be considered a tracer for relatively clean air (i.e., cold, clear air coming
28 down from the upper atmosphere), as discussed in Chapter 3 of this AQCD. The clean air is
29 associated with the passage of cold fronts and the onset of high-pressure conditions, which occur
30 with colder temperatures. Thus, sunny clear winter days following a high-pressure system are
31 the days when air pollution levels from primary emissions (e.g., NO2, SO2, and PM from local
January 2005 7-11 DRAFT-DO NOT QUOTE OR CITE
-------
1 sources) tend to be lower and O3 is relatively higher. This can lead to negative correlations
2 between O3 and the primary pollutants in the winter. As shown in Figure 3-6 in the Chapter 3
3 Annex, the relationship between O3 and PM2 5 was U-shaped for the year-round data in Fort
4 Meade, MD. The negative PM2 5/O3 slope was in the range of O3 concentrations less than 30
5 ppb, providing supporting evidence of the aforementioned winter phenomenon.
6 This changing relationship between O3 and temperature, as well as O3 and other pollutants
7 across seasons, and its potential implications to health effects modeling has not been examined
8 thoroughly in the time-series literature. Even the flexible smoother-based adjustments for
9 seasonal and other time-varying variables cannot fully take into account these complex
10 relationships. One obvious way to alleviate or avoid this complication is to analyze data by
11 season. While this practice reduces sample size, its extent would not be as serious as PM (which
12 is collected only every 6th day in most locations) because O3 is collected daily, though only in
13 warm seasons in some states. An alternative approach is to include separate O3 concentration
14 variables for each season (by multiplying O3 concentrations by a season indicator variable).
15 In locations where seasonal variability may be a factor, O3 effect estimates calculated using
16 year-round data can be misleading, as the changing relationship between O3, temperature, and
17 other pollutants across seasons may have a significant influence on the estimates. Analyses have
18 indicated that confounding from seasonal variability may be controlled effectively by stratifying
19 the data by season.
20
21 7.1.3.6 Confounding Effects of Copollutants
22 Extensive discussions on the issues related to confounding effects among air pollutants in
23 time-series study design are provided in Section 8.4.3 of the 2004 PM AQCD. Since the general
24 issues discussed in that document are applicable to all pollutants, such discussions are not
25 repeated here. What was not discussed in the 2004 PM AQCD was the issue of changing
26 relationships among air pollutants across seasons. For O3, the confounder of main interest is
27 PM, especially fine particles or sulfates that are high in summer, as other copollutants (e.g., CO,
28 NO2, SO2) tend to be elevated in the colder season. As mentioned in the previous section, PM
29 indices in some urban locations may be positively correlated with O3 in the summer and
30 negatively correlated in the winter. Thus the correlation between O3 and PM for year-round data
31 may be misleading. The high reactivity of O3 with certain copollutants further complicates the
January 2005 7-12 DRAFT-DO NOT QUOTE OR CITE
-------
1 analysis. For example, the reaction between NOX, emitted from motor vehicles, and O3 results in
2 reduced O3 levels but increased NO2 levels during high traffic periods.
3 Multipollutant models often are used to assess potential confounding by copollutants. The
4 limitations of multipollutant regression models, including the potential transfer of "effects" from
5 causal pollutant to noncausal pollutant in the presence of unequal measurement errors, are
6 discussed in the 2004 PM AQCD (Sections 8.4.3 and 8.4.5). In addition, uncertainty remains as
7 to the use of multipollutant regression models to assess the independent health effects of
8 pollutants that are correlated. Particularly in the case of O3, there remains concern as to whether
9 multipollutant regression models for year-round data can adjust for potential confounding
10 adequately due to the changing relationship between O3 and other pollutants. Despite these
11 limitations, multipollutant models are still the prevailing approach in most, if not all, studies of
12 O3 health effects and serve as an important tool in addressing the issue of confounding by
13 copollutants, especially in season-stratified analyses.
14
15 7.1.3.7 Model Uncertainty and Multiple Testing
16 In the analyses of air pollution health effects, there is often uncertainty as to which model
17 is most appropriate. In the case of PM, there were concerns that the positive associations found
18 with mortality were the result of multiple testing and selection. Testing many models to identify
19 the best fit can lead to an underestimation of uncertainty, thus there is a need for statistical
20 methods that can identify the best model while properly accounting for model uncertainty.
21 Standard methods of variable selection include Akaike Information Criterion and Bayes
22 Information Criterion. Bayesian model averaging is a recent family of methods used to address
23 these issues. In Bayesian model averaging, predictions and inferences are based on a set of
24 models, rather than a single model, and each model contributes proportionally to the support it
25 receives from the observed data (Clyde, 1999). While there remains concerns about the large
26 enumeration of models, Bayesian model averaging is a computationally efficient way to
27 incorporate model uncertainty into decision making.
28 Some researchers have used other methods to address the issue of multiple testing.
29 Dominici et al. (2003) used a minimum number of tests in the U.S. 90 cities study, which
30 minimized the uncertainty associated with multiple testing, but at the cost of possibly not
31 identifying the best model. Another method used by Dominici et al. (2003) to evaluate the
January 2005 7-13 DRAFT-DO NOT QUOTE OR CITE
-------
1 model was sensitivity testing. Lumley and Sheppard (2000) used different control variables to
2 check the bias in model identification. They found that the bias was small, but of the same
3 magnitude as the estimated health impacts. Another approach is to use one set of data for model
4 identification, and a second set of data for model fitting. Cross validation also sheds light on this
5 issue.
6 With the currently available knowledge, multiple testing is unavoidable in air pollution
7 health effects analyses. To address the issues of model uncertainty and multiple testing, further
8 research leading to the development of standard methodology may be necessary.
9
10 7.1.3.8 Impact of GAM Convergence Issue on O3 Risk Estimates
11 Generalized Additive Models (GAM) have been widely utilized for epidemiologic analysis
12 of the health effects attributable to air pollution. The impact of the GAM convergence issue was
13 thoroughly discussed in Section 8.4.2 of the 2004 PM AQCD. Reports have indicated that using
14 the default convergence criteria in the Splus software package for the GAM function can lead to
15 suboptimal regression estimates for PM and an underestimation of the standard error of that
16 effect estimate (Dominici et al., 2002; Ramsay et al., 2003). GAM default convergence criterion
17 has a convergence precision of 10~3 and a maximum number of 10 iterations. The more stringent
18 convergence criterion refers to increased stringency of both the convergence precision and
19 number of iterations. The default convergence criteria was found to be a problem when the
20 estimated relative risks were small, and two or more nonparametric smoothing curves were
21 included in the GAM (Dominici et al., 2002). The magnitude and direction of the bias depend in
22 part on the concurvity of the independent variables in the GAM and the magnitude of the risk
23 estimate. Most attention has been focused on the influence of the GAM function on effect
24 estimates for PM. However, because O3 covaries more strongly with both weather and time
25 factors than does PM, the issue of GAM convergence criteria for O3 needs to be considered.
26 A recent meta-analysis by Stieb et al. (2003) found a difference in O3-mortality risk
27 estimates between the GAM studies and non-GAM studies. In the single-pollutant models, the
28 O3-mortality risk estimates for the non-GAM studies and GAM studies were 1.8% (95% CI: 0.5,
29 3.1) and 2.2% (95% CI: 1.4, 2.8), respectively, per 40 ppb daily 1-h max O3. In the
30 multipollutant models, the pooled risk estimate was 1.0% (95% CI: -0.5, 2.6) for non-GAM
31 studies and 0.5% (95% CI: -1.0, 1.9) for GAM studies.
January 2005 7-14 DRAFT-DO NOT QUOTE OR CITE
-------
1 A few GAM studies reanalyzed the O3 risk estimates using more stringent convergence
2 criteria or general linear models (GLM). Reanalysis of an asthma hospital admissions study in
3 Seattle, WA (Sheppard et al., 1999; reanalysis Sheppard, 2003) indicated that there were only
4 slight changes in the risk estimates when using more stringent convergence precision (10~8) in
5 GAM. The original GAM analysis indicated an excess risk of 9% (95% CI: 3, 17) whereas the
6 stringent GAM analysis found an excess risk of 11% (95% CI: 3, 19) per 30 ppb increase in 8-h
7 max O3. Similar results were found using GLM with natural splines, 11% (95% CI: 2, 20).
8 In the reanalysis of Santa Clara County, CA data, Fairley (1999; reanalysis Fairley, 2003) used
9 the same methods as the original analysis except the convergence precision (e) was increased
10 from 10~4 to 10~12 and the maximum number of iterations (M) were increased from 10 to 107.
11 The O3 nonaccidental mortality risk estimates slightly increased from 2.8% using default GAM
12 parameters to 2.9% using stringent GAM parameters per 30 ppb increase in 8-h max O3. The
13 O3-mortality risk estimates further increased to 3.0% using GLM with natural cubic splines.
14 In the reanalysis of the Netherlands data by Hoek et al. (2000; reanalysis Hoek, 2003), the O3
15 nonaccidental mortality risk estimates increased from 1.3% (default GAM) to 1.5% (stringent
16 GAM, e = 10~8, M = 103) and 1.6% (GLM with natural splines) per 30 ppb increase in 8-h avg
17 O3 (12 p.m.-8 p.m.). In the analysis of the large 90 U.S. cities (Samet et al., 2000; reanalysis
18 Dominici et al., 2003), the year-round combined estimate of O3 nonaccidental mortality risk
19 changed from a nonsignificant negative value of approximately -0.2% (default GAM) per
20 20 ppb change in 24-h avg O3 to a significantly positive excess risk of 0.8% (stringent GAM,
21 e= 10~15, M=103).
22 Most analyses comparing results using default GAM convergence criteria to results from
23 stringent GAM convergence criteria and GLM have found little difference among the O3 effect
24 estimates. However, one study by Cifuentes et al. (2000) in Santiago, Chile observed a large
25 difference in the O3-mortality excess risks calculated using default GAM (2.4%) and GLM
26 (0.3%). Therefore, the impact of the GAM convergence problem appears to vary depending on
27 data sets, and likely depends upon the intercorrelation among covariates and the magnitude of
28 the risk estimate. However, in the limited number of studies that have reanalyzed O3 risk
29 estimates, there is little evidence that default GAM analyses resulted in positively biased
30 estimates as observed for PM. Generally it appears that the use of default convergence criteria
31 in GAM tends to bias risk estimates towards the null, in addition to underestimating the standard
January 2005 7-15 DRAFT-DO NOT QUOTE OR CITE
-------
1 errors. In uniformity with the approach used in the 2004 PM AQCD, the results from studies
2 that analyzed data using GAM with default convergence criteria and at least two nonparametric
3 smoothing terms are generally not considered in this chapter, with a few exceptions as noted.
4
5 7.1.4 Approach to Presenting O3 Epidemiologic Evidence
6 To produce a thorough appraisal of the evidence, we first concisely highlight key points
7 derived from the 1996 O3 AQCD assessment. Then pivotal information, including
8 methodological features and results, from important new studies that have become available
9 since the prior O3 AQCD are presented in summary tables in Chapter 7 of the Annex. In the
10 main body of the chapter, particular emphasis is focused on those studies and analyses
11 considered to provide information most directly applicable for development of criteria. Not all
12 studies should be accorded equal weight in the overall interpretive assessment of evidence
13 regarding O3-association health effects. Among well-conducted studies with adequate control
14 for confounding, increasing scientific weight should be accorded in proportion to the precision
15 of their effect estimates. Small-scale studies without a wide range of exposures generally
16 produce less precise estimates compared to larger studies with an adequate exposure gradient.
17 Therefore, the range of exposures, the size of the study as indicated by the length of the study
18 period and total number of events, and the inverse variance of the principal effect estimate are all
19 important indices useful in determining the likely precision of the health effect estimates and in
20 according relative scientific importance to the findings of a given study. In any case, emphasis
21 should be accorded to estimates from studies with narrow confidence bands.
22 Emphasis is placed on text discussion of (1) new multicity studies that employ
23 standardized methodological analyses for evaluating O3 effects across several or numerous cities
24 and often provide overall effect estimates based on combined analyses of information pooled
25 across multiple cities; (2) studies that consider O3 as a component of a complex mixture of air
26 pollutants, including in particular the gaseous criteria pollutants (CO, NO2, SO2) and PM; and
27 (3) North American studies conducted in the U.S. or Canada. Multicity studies are of particular
28 interest and value due to their evaluation of a wider range of O3 exposures and large numbers of
29 observations, thus possibly providing more precise effect estimates than most smaller scale
30 independent studies of single cities. Another potential advantage of the multicity studies,
31 compared to meta-analyses of multiple "independent" studies, is consistency in data handling
January 2005 7-16 DRAFT-DO NOT QUOTE OR CITE
-------
1 and model specifications that eliminates variation due to study design. Also, unlike regular
2 meta-analyses, they do not suffer from potential omission of nonsignificant results due to
3 "publication bias." Furthermore, geographic patterns of air pollution effects have the potential
4 to provide especially valuable evidence regarding relative homogeneity and/or heterogeneity of
5 O3 health effects relationships across geographic locations. In accordance to the emphasis
6 placed on the O3 epidemiology studies in this chapter, the tables in the Chapter 7 Annex were
7 organized by region with multicity studies in each region presented first.
8 In the coming sections, field/panel studies and studies of emergency department visits and
9 hospital admissions, which contributed to the establishment of the revised 1997 NAAQS for O3,
10 are presented first. This is followed by a discussion of O3-related mortality and chronic effects.
11 The chapter ends with an integrative discussion providing a summary and conclusions.
12
13
14 7.2 FIELD STUDIES ADDRESSING ACUTE EFFECTS OF OZONE
15 7.2.1 Summary of Key Findings on Field Studies of Acute Effects
16 From the 1996 O3 AQCD
17 In the 1996 O3 AQCD, individual-level camp and exercise studies provided useful
18 quantitative information on the exposure-response relationships linking human lung function
19 declines with O3 exposure occurring in ambient air. The available body of evidence supported a
20 dominant role of O3 in the observed lung function decrements. Extensive epidemiologic
21 evidence of pulmonary function responses to ambient O3 came from camp studies. Six studies
22 from three separate research groups provided a combined database on individual exposure-
23 response relationships for 616 children (mostly healthy, nonasthmatic) ranging in age from 7 to
24 17 years, each with at least six sequential measurements of FEVj (forced expiratory volume in
25 1 second) while attending summer camps (Avol et al., 1990; Higgins et al., 1990; Raizenne
26 et al., 1987, 1989; Spektor et al., 1988a, 1991). When analyzed together using consistent
27 analytical methods, these data yielded an average relationship between afternoon FEVj and
28 concurrent-hour O3 concentration of -0.50 mL/ppb (p < 0.0001), with study-specific slopes
29 ranging from -0.19 to -1.29 mL/ppb. Exposure in camp studies usually extended for multiple
30 hours. Although the regression results noted above were based on one-hour O3 levels, single-
31 and multiple-hour averages were observed to be highly correlated, thus these results might
January 2005 7-17 DRAFT-DO NOT QUOTE OR CITE
-------
1 represent, to some extent, the influence of multihour exposures. In addition to the camp study
2 results, two studies involving lung function measurements before and after well-defined exercise
3 events in adults yielded exposure-response slopes of -0.4 mL/ppb (Selwyn et al., 1985) and
4 -1.35 mL/ppb (Spektor et al., 1988b). Ozone concentrations during exercise events of
5 approximately !/2-hour duration ranged from 4 to 135 ppb in these studies.
6 Results from other field panel studies also supported a consistent relationship between
7 ambient O3/oxidant exposure and acute respiratory morbidity in the population. Respiratory
8 symptoms (or exacerbation of asthma) and decrements in peak expiratory flow (PEF) occurred
9 with increased ambient O3 concentrations, especially in asthmatic children (Lebowitz et al.,
10 1991; Krzyzanowski et al., 1992). The aggregate results showed greater responses in asthmatic
11 individuals than in nonasthmatics (Lebowitz et al., 1991; Krzyzanowski et al., 1992), indicating
12 that asthmatics might constitute a sensitive group in epidemiologic studies of oxidant air
13 pollution. Since the 1996 O3 AQCD, new research has examined a broad scope of field studies
14 which are presented next.
15
16 7.2.2 Introduction to Recent Field Studies of Acute O3 Effects
17 Numerous field studies carried out over the past decade have tested for and, in many cases,
18 observed acute associations between measures of respiratory ill-health and O3 concentrations in
19 groups of subjects (Table AX7-1 in Chapter 7 Annex). Acute field studies are distinguished
20 from other acute epidemiologic study designs in that they recruit and collect data from
21 individual human subjects instead of utilizing administrative data on aggregate health outcomes
22 such as daily mortality, hospital admissions, or emergency department visits. Because of the
23 logistical burden associated with direct data collection from individual subjects, field/panel
24 studies tend to be small in both numbers of subjects and in duration of follow-up. While this
25 may limit the statistical power of field studies, it is compensated for by the ability to determine
26 individual-level information on health outcomes, exposure levels, and other potentially
27 confounding factors.
28 The most common outcomes measured in acute field studies on the effects of air pollution
29 exposure are lung function and various respiratory symptoms. Other respiratory outcomes
30 examined on a limited basis include inflammation and generation of hydroxyl radicals in the
31 upper airways, and school absences. Several studies examined cardiovascular outcomes
January 2005 7-18 DRAFT-DO NOT QUOTE OR CITE
-------
1 including heart rate variability and risk of myocardial infractions. The first group of studies
2 provides varying degrees of evidence supporting the conclusion that elevated O3 levels can have
3 negative impacts on lung function and symptoms, confirming and adding to the body of
4 knowledge that is presented in the 1996 O3 AQCD. Some emphasis has been placed in
5 examining the independent role of O3 in the presence of PM and other pollutants. The other new
6 studies contribute information on cardiopulmonary outcomes which had not been as well-
7 documented previously.
8
9 7.2.3 Acute O3 Exposure and Lung Function
10 As discussed in the 1996 O3 AQCD and in the earlier chapter of this document on
11 controlled human exposure studies (Chapter 6), a large body of literature from clinical and field
12 studies has clearly and consistently demonstrated reversible decrements in pulmonary function
13 following acute O3 exposure. Significant O3-induced spirometric and symptom responses have
14 been observed in clinical studies of exercising healthy young adults (see Section 6.2) and in
15 some potentially susceptible subpopulations, namely asthmatics and children (see Sections 6.3.2
16 and 6.5.1). Field studies of acute O3 exposure that examine pulmonary function fall into two
17 distinct groupings, those that conduct spirometry (FEVj and FVC [forced vital capacity]) and
18 those that measure PEF. Results from the previous O3 AQCD and Chapter 6 of this document
19 support the conclusion that the spirometric parameter FEVj is the stronger and more consistent
20 measure of lung function. PEF is a useful clinical measure that is more feasible to perform in
21 field studies, however its measurements are more variable and possibly less reliable than FEVj
22 (Fuhlbriggeetal., 2001).
23 Studies of FEVj will be presented first, followed by a discussion of PEF studies. Other
24 dividing aspects between these two major types of lung function studies include health status of
25 subjects (e.g., healthy, mildly asthmatic, severely asthmatic), time spent outdoors, and exertion
26 levels. Several studies brought these factors together to produce informative data. Some FEVj
27 studies involved both increased outdoor O3 exposure and higher exertion levels. The results
28 from this group of subjects are comparable to those from the exercising subjects in the clinical
29 studies discussed in Chapter 6.
30
31
January 2005 7-19 DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
7.2.3.1 Acute O3 Studies with Spirometry (FEVt)
Studies published over the past decade have provided some new insights on the acute
effects of O3 on FEVj. Tables 7-la, 7-lb, and 7-lc summarize the results of all studies that
investigated quantitative O3-related effects on FEVj. Four studies of spirometry were not
included in the tables; three studies did not provide quantitative O3 data (Cuijpers et al., 1994;
Delfino et al., 2004; Frischer et al., 1997) and one measured FEV075 (forced expiratory volume
in 0.75 seconds) (Scarlett et al., 1996). With few exceptions, the O3 effect estimates showed
decrements for FEVj across studies and several were statistically significant. These studies are
discussed in further detail, starting with the O3 effect on individuals with elevated exertion
levels.
Exercise and outdoor worker panels
The current 8-hour NAAQS for O3 has its original basis in the controlled human exposure
studies, as discussed in Chapter 6. These field studies with subjects at elevated exertion levels
Table 7-la. Field Studies that Investigated the Association between Acute Ambient O3
Exposure and Changes in FEVt
Reference
Study Location
Study Period
n"
Linn etal. (1996)
Korricketal. (1998)
Braueretal. (1996)
Hoppeetal. (1995a)
Ulmeretal. (1997)
Castillejos etal. (1995)
Romieuetal. (1998)
Romieu et al. (2002)
Chen etal. (1999)
Rubidoux, Upland, and Torrance, CA
Mount Washington, NH
Fraser Valley, British Columbia,
Canada
Munich, Germany
Freudenstadt and Villingen, Germany
SW Mexico City
Mexico City
Mexico City
Sanchun, Taihsi, and Linyuan, Taiwan
Fall-spring 1992-1993, 1993-1994 269
Summers 1991, 1992 530
Jun-Aug 1993 58
Apr-Sep 1992-1994 208
Mar-Oct 1994 135
Aug 1990-Oct 1991 40
Mar-May 1996; Jun-Aug 1996 47
Octl998-Apr2000 158
May 1995-Jan 1996 895
aThe number of the total study population is presented. Some of the effect estimates presented in Tables 7-lb and
7-lc were based on a subset of the total population.
January 2005
7-20
DRAFT-DO NOT QUOTE OR CITE
-------
Table 7-lb. Changes in FEVt (95% CI) Associated with Acute Ambient O3 Exposures,
Ordered by Size of the Estimate
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
Reference
Braueretal. (1996)
Braueretal. (1996)
Romieuetal. (1998)
Ulmeretal. (1997)
Hoppeetal. (1995a)
Romieuetal. (1998)
Hoppeetal. (1995a)
Ulmeretal. (1997)
Hoppeetal. (1995a)
Ulmeretal. (1997)
Hoppeetal. (1995a)
Ulmeretal. (1997)
Romieuetal. (1998)
Chen etal. (1999)
Chen etal. (1999)
Romieu et al. (2002)
Romieu et al. (2002)
Romieuetal. (1998)
Chen etal. (1999)
Study Population/Analysis
Berry pickers, next morning
Berry pickers, afternoon
Street workers on placebo
(1st phase, lag 0-1)
School children in
Freudenstadt
Juvenile asthmatics
Street workers on placebo
(1st phase, lagO)
Clerks
School boys in Freudenstadt
and Villingen
Athletes
School children in
Freudenstadt and Villingen
Forestry workers
School girls in Freudenstadt
and Villingen
Street workers on
supplement (1st phase,
lag 0-1)
Children, with NO2 in model
(lag 1)
Children (lag 1)
Moderate to severe asthmatic
children on placebo (lag 1)
Moderate to severe asthmatic
children on placebo, with
NO2 and PM10 in model
(lag 1)
Street workers on
supplement
(1st phase, lagO)
Children (lag 2)
Mean O3
Level (ppb)
40.3
40.3
123
50.6
74 b
123
68 b
41.4
71b
41.4
64 b
41.4
123
19.7-110.3°
19.7-110.3°
102
102
123
N/A
Exposure
Index
1-hmax
1-hmax
1-hmax
!/2-h max
!/2-h max
1-hmax
i/2-h max
!/2-h max
!/2-h max
i/2-h max
i/2-h max
i/i-h max
1-hmax
1-h max
1-h max
1-hmax
1-hmax
1-hmax
24-h avg
Change in FEV/
(mL)
-180.0
-152.0
-117.2
-87.5
-84.0
-71.6
-63.2
-61.4
-60.8
-56.6
-56.0
-44.2
-41.2
-34.0
-25.6
-18.8
-18.4
-17.6
-17.4
(-227.0
(-183.4
(-207.4
(-143.2
(-196.4
(-113.9
(-108.8
(-122.9
(-115.2
(-101.3
(-118.4
(-105.0
(-143.9
(-60.7,
(-49.1,
(-34.2,
(-35.5,
(-68.6,
(-41.7,
,-133.0)
, - 120.6)
,-27.0)
,-31.7)
, 28.4)
,-29.3)
,-17.6)
,0.0)
,-6.4)
,-12.0)
,6.4)
, 16.7)
,61.5)
-7.3)
-2.1)
-3.4)
-1.3)
33.4)
6.9)
January 2005
7-21
DRAFT-DO NOT QUOTE OR CITE
-------
Table 7-lb (cont'd). Changes in FEVt (95% CI) Associated with Acute Ambient O3
Exposures, Ordered by Size of the Estimate
Reference
20
21
22
23
24
25
26
27
28
29
30
31
Chen et al.
(1999)
Ulmeretal. (1997)
Chen et al.
Romieu
et
Chen et al.
Romieu
Chen et
et
al.
Linn et al.
Linn et al.
Romieu
Romieu
Romieu
et
et
et
(1999)
al. (1998)
(1999)
al. (1998)
(1999)
(1996)
(1996)
al. (2002)
al. (2002)
al. (2002)
Study Population/Analysis
Children (lag 2)
School children in Villingen
Children (lag 1)
Street workers on placebo
(2nd phase, lag 0)
Children (lag 7)
Street workers on placebo
(2nd phase, lag 0-1)
Children (lag 7)
School children, next
morning
School children, afternoon
All asthmatic children on
placebo (lag 1)
Moderate to severe asthmatic
on supplement (lag 1)
Moderate to severe asthmatic
Mean O3
Level (ppb)
19.7-110.3°
32.1
N/A
123
19.7-110.3°
123
N/A
23
23
102
102
102
Exposure
Index
1-h max
Vi-h max
24-h avg
1-h max
1-h max
1-h max
24-h avg
24-h avg
24-h avg
1-h max
1-h max
1-h max
Change in FEVj3
(mL)
-16.0
-15.0
-13.6
-13.2
-12.4
-12.0
-6
-5
-3
-3
-0
-0
.0
.2
.6
.6
.7
.2
(-44.2,
(-74.
(-33.
(-64.
(-31.
6,
2,
2,
2,
(-96.7,
(-20.
(-15.
(-13.
(-13.
(-15.
(-15.
9,
o,
8,
5,
1,
6,
12.2)
44.5)
6.0)
37.8)
6.4)
72.7)
8.9)
4.6)
6.6)
6.3)
13.7)
15.1)
32 Romieu et al. (2002)
33 Romieu etal. (1998)
on supplement, with NO2
and PM10 in model (lag 1)
All asthmatic children on
supplement (lag 1)
Street workers on
supplement
(2nd phase, lag 0)
102 1-h max 0.8 (-9.8,11.3)
123 1-h max 6.0 (-23.8,35.8)
34
35
Hoppeetal. (1995a)
Romieu etal. (1998)
Seniors
Street workers on
supplement
(2nd phase, lag 0-1)
70 b
123
!/2-h max
1-h max
13.6
27.2
(-26.
(-25.
8, 54.0)
3,79.7)
a Change in FEVl is per standard unit ppb O3 (40 ppb for i/2-h max O3 and 1-h max O3, 30 ppb for 8-h max O3, and
20 ppb for 24-hr avg O3).
bMean O3 concentration on high O3 days.
0 Range of O3 concentrations.
January 2005
7-22
DRAFT-DO NOT QUOTE OR CITE
-------
Table 7-lc. Cross-day Changes in FEVt Associated with Acute Ambient O3 Exposures,
Ordered by Size of the Estimate
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
Reference
Korrick et al. (1998)
Korrick et al. (1998)
Korrick et al. (1998)
Korrick et al. (1998)
Korrick et al. (1998)
Korrick et al. (1998)
Korrick et al. (1998)
Korrick et al. (1998)
Korrick et al. (1998)
Korrick et al. (1998)
Korrick et al. (1998)
Korrick et al. (1998)
Korrick et al. (1998)
Korrick et al. (1998)
Linnetal. (1996)
Castillejos et al. (1995)
Braueretal. (1996)
Study Population/
Analysis
Hikers with wheeze or
asthma (post-pre-hike)
Hikers who hiked 8-12
hours (post-pre-hike)
Hikers age 28-37 years
(post-pre-hike)
Hikers who never smoked
(post-pre-hike)
Hikers male
(post-pre-hike)
Hikers age 38-47 years
(post-pre-hike)
All hikers
(post-pre-hike)
All hikers, with PM2 5
and acidity in model
(post-pre-hike)
Hikers age 18-27 years
(post-pre-hike)
Hikers female
(post-pre-hike)
Hikers age 48-64 years
(post-pre-hike)
Hikers without wheeze or
asthma (post-pre-hike)
Hikers who hiked 2-8 hours
(post-pre-hike)
Hikers who formerly
smoked (post-pre-hike)
School children
(p.m. -a.m.)
Private primary school
(post- pre -exercise)
Berry pickers
(post- pre-workshift)
Mean O3
Level (ppb)
40
40
40
40
40
40
40
40
40
40
40
40
40
40
23
112.3
40.3
Exposure
Index
8-h avg
8-h avg
8-h avg
8-h avg
8-h avg
8-h avg
8-h avg
8-h avg
8-h avg
8-h avg
8-h avg
8-h avg
8-h avg
8-h avg
24-h avg
1-hmax
1-hmax
Cross-day Change
in FEVj a (mL)
-182.5b
-84.5b
-82.1b
-72. 3 b
-67.4b
-64.9b
-62.5b
-58.8b
-52.7b
-47.8b
-46.5b
-44.1b
-40.4b
-29.4b
-11.6
-9.1C
0
(-312.2, -52.9)
(-154.1, -14.9)
(-139.7, -24.4)
(-132.3, -12.2)
(-127.4, -7.3)
(-127.3, -2.5)
(-115.3, -9.7)
(-135.6, 18.0)
(-117.5, 12.2)
(-141.4,45.9)
(-125.8,32.7)
(-101.7, 13.5)
(-110.0,29.2)
(-125.4,66.6)
(-20.6, -2.6)
(-13.6, -4.7)
(-47.0,47.0)
aCross-day change in FEVj is per standard unit ppb O3 (40 ppb for 1-h max O3, 30 ppb for 8-h avg O3, and 20 ppb
for 24-h avg O3).
b Korrick et al. presented % change in FEVj. The data was transformed to FEVj units of mL by multiplying by the
total population average FEVj of 4,083 mL.
0 Castillejos et al. presented % change in FEVj. The data was transformed to FEVj units of mL by multiplying by
the total population average FEVj of 1,900 mL.
January 2005
7-23
DRAFT-DO NOT QUOTE OR CITE
-------
1 are of particular interest due to their similarities to the human chamber studies. The majority of
2 human chamber studies have examined the effects of O3 exposure in subjects performing
3 continuous or intermittent exercise for variable periods of time (see Chapter 6 of this O3 AQCD).
4 A study by Brauer and colleagues (1996) reported unusually large O3 effects on lung
5 function among outdoor workers. This study presented O3 effects during an extended outdoor
6 exposure period combined with elevated levels of exertion. The investigators repeatedly
7 measured spirometric lung function before and after outdoor summer work shifts over 59 days
8 on a group of 58 berry pickers in Fraser Valley, British Columbia, Canada. Subjects, both male
9 and female, ranged from 10 to 69 years old, with a mean age of 44 years. Outdoor work shifts
10 averaged 11 hours in duration. The mean 1-h max O3 concentration was 40.3 ppb. Exertion
11 levels were estimated using portable heart rate monitors carried over a period of four or more
12 hours by a representative subset of subjects during 16 work shifts. Heart rates were essentially
13 constant over the work shift, averaging 36% higher than resting levels. The authors estimated
14 that minute ventilations may have averaged roughly 30 L/min during work. Post-shift FEVj and
15 FVC showed large decreases as a function of O3 concentration and those effects remained
16 significant when PM2 5 was included in the analysis. Significant declines in lung function also
17 were observed on the morning following high O3 exposure. The effects seen in this study are
18 larger than have been reported previously. For example, afternoon FEVj was 3.8 mL lower per
19 1 ppb increase in O3 concentrations, compared to the decline of 0.4 mL/ppb and 1.35 mL/ppb
20 observed in the earlier adult exercise studies (Spektor et al., 1988b; Selwyn et al., 1985).
21 Further, when data were restricted to days with 1-h max O3 concentrations under 40 ppb, the O3
22 effects on afternoon FEVj did not change in magnitude and remained significant.
23 In a Mexico City study of 47 outdoor street workers (Romieu et al., 1998), spirometry was
24 performed repeatedly at the end of the workshift over a two month period. Subjects were
25 exposed to outdoor ambient O3 levels for a mean of 7.4 hours during the workday. Among those
26 who had never taken an antioxidant supplement (subjects who received a placebo during the 1st
27 phase of the study), same day O3 concentrations were significantly associated with decreases in
28 FEVj. A mean decline of 71.6 mL (95% CI: 29.3, 113.9) was observed per 40 ppb increase in
29 1-h max O3. The results from this study, in addition to those from the Canadian study of berry
30 pickers (Brauer et al., 1996), indicate that outdoor workers are a potentially susceptible
31 population that may need protection from O3 exposures.
January 2005 7-24 DRAFT-DO NOT QUOTE OR CITE
-------
1 Hoppe et al. (1995a) examined forestry workers (n = 41) for changes in pulmonary
2 function attributable to O3 exposure in Munich, Germany. In addition, athletes (n = 43) were
3 monitored in the afternoon after a two-hour outdoor training period. Pulmonary function tests
4 were conducted on days of both "high" (mean lA>-h max O3 of 64 to 74 ppb) and "low" (mean
5 !/2-h max O3 of 32 to 34 ppb) ambient O3 concentrations. From the average activity levels,
6 ventilation rates were estimated. Athletes, who had a fairly high ventilation rate of 80 L/min,
7 experienced a significant decrease of 60.8 mL (95% CI: 6.4, 115.2) in FEVj per 40 ppb increase
8 in !/2-h max O3. Among the forestry workers, an O3-related decline in FEVj also was observed
9 (- 56.0 mL), however the change was not statistically significant.
10 For the above studies that examine outdoor workers and athletes who train outdoors,
11 Table 7-2 presents the estimated O3 external doses and compares them to changes in FEVj
12 associated with acute ambient O3 exposures. The use of estimated O3 external doses offers
13 another potential for insight into studies that examine subjects with elevated ventilation rates and
14 prolonged outdoor exposures at varying ambient O3 concentrations. No consistent relationship
15 between estimated O3 external doses and changes in FEVj can be derived from the limited
16 evidence.
17 One FEVj study appears to mirror the sort of outcome typically seen in clinical studies. In
18 a study by Korrick et al. (1998), adult hikers (n = 530) of Mount Washington, NH performed
19 spirometry before and after hiking for a mean of 8 hours (range: 2 - 12). The mean hourly O3
20 concentration ranged from 21 to 74 ppb. After their hike, subjects experienced a mean decline
21 of 1.5% (95% CI: 0.2, 28) in FEVi and 1.3% (95% CI: 0.5, 2.1) in FVC per 30 ppb increase in
22 the mean of the hourly O3 concentration during the hike. In addition, Korrick et al. (1998)
23 compared hikers who hiked 8 to 12 hours to those who hiked 2 to 8 hours. Among those who
24 hiked longer, the % change in FEVj was more than two-fold greater per ppb exposure compared
25 to those who hiked only for 2 to 8 hours. Each hour hiked, which may reflect dose, was
26 associated with a decline of 0.3% (p = 0.05) in FEVj, after adjusting for O3.
27 In a Mexico City study, the O3 effect attributable to exercise was determined using a group
28 of school children (n = 40) chronically exposed to moderate to high levels of O3 (Castillejos
29 et al., 1995). Children were tested up to 8 times between August 1990 and October 1991.
30 Spirometry was performed by the children before and after a one-hour intermittent exercise
31 session outdoors. Outdoor O3 levels ranged up to 365 ppb, with a mean of 112.3 ppb. Linear
January 2005 7-25 DRAFT-DO NOT QUOTE OR CITE
-------
Table 7-2. Estimated O3 External Doses and Changes in FEVt Associated with Acute
Ambient O3 Exposures in Outdoor Workers and Athletes
1
2
3
4
Reference
Brauer et al.
Hoppe et al.
Hoppe et al.
Romieu et al
Study Location
( 1 996) Fraser Valley, Canada
(1995a) Munich,
(1995a) Munich,
. (1998) Mexico
Germany
Germany
City
Study Period
Jun-Aug 1993
Apr-Sep 1994
Apr-Sep 1993
Mar-May 1996
N Study
58 Berry
Population
pickers
43 Athletes
41 Forestry Workers
13 Street
Workers3
Age
(years)
10-69
13-38
20-60
18-58
Mean O3 Exposure
Level (ppb) Duration (h)
1
2
3
4
26.0
71.0C
64.0 c
67.3
11
2
7
9
Ventilation
Rate (L/min)
30
80
40
28 d
Exposure
(mg/h)
91
660
298
219
Dose
(mg)
997
1320
2083
1971
Change in FEV/
(mL)
-152.0 (-183.4
-60.8 (-115.2
-56.0 (-118.4
-71.6 (-113.9
,-120.6)
, -6.4)
,6.4)
,-29.3)
aFor the street workers in Romieu et al. (1998), only results from subjects who had never taken the antioxidant
supplement (on placebo during 1st phase of study) are presented here.
b Change in FEV{ is per 40 ppb increase in 1-h max O3 or equivalent.
°Mean O3 concentration during exposure period was not presented in Hoppe et al. (1995). The !/2-h max
O3 concentrations are shown here.
d Ventilation rate was not presented. The ventilation rate of 28 L/min was calculated for a male performing
a heavy workload for 1/8 of the workday in Table B. 17 of the International Commission on Radiological
Protection (ICRP) Publication 66 (ICRP, 1994).
1 trend analyses of the relationship between quintiles of O3 and % change in lung function were
2 significant. However, stratified analyses indicated that statistically significant changes were
3 observed only with higher quintiles of O3 exposure (72-125 ppb and 183-365 ppb). Therefore,
4 when exercising at higher O3 levels, children experienced significant declines in pulmonary
5 function despite the repeated daily exposure to moderate and high levels of O3 in Mexico City.
6 Collectively, the above studies confirm and extend clinical observations that prolonged
7 exposure periods, combined with elevated levels of exertion or exercise, may magnify the effect
8 of O3 on lung function. The most representative data comes from the Korrick et al. (1998) hiker
9 study. This U.S. study provided outcome measures stratified by several factors (e.g., gender,
10 age, smoking status, presence of asthma) within a population capable of more than normal
11 exertion.
January 2005
7-26
DRAFT-DO NOT QUOTE OR CITE
-------
1 Panels of other risk groups
2 Hoppe et al. (1995a,b) examined several potentially susceptible populations for changes in
3 pulmonary function attributable to O3 exposure in Munich, Germany. The forestry workers and
4 athletes were discussed in the previous section. Senior citizens (n = 41) and juvenile asthmatics
5 (n = 43) were also monitored on "low" O3 and "high" O3 days. Subjects were requested to stay
6 outdoors for at least 2 hours just before the afternoon pulmonary function test. Clerks (n = 40)
7 were considered the nonrisk control group. Although clerks spent the majority of their time
8 indoors, their outdoor exposures on the "high" O3 days were similar to that of the four other risk
9 groups. The results showed no significant O3 effects on the senior citizens, who had the lowest
10 ventilation rate. Asthmatics and clerks experienced slight reductions in FEVj on high O3 days.
11 Among all risk groups, juvenile asthmatics experienced the largest O3-related decline in FEVl3
12 though not statistically significant.
13 Several other panel studies performed spirometry in children, another potentially
14 susceptible group (Chen et al., 1999; Cuijpers et al., 1994; Frischer et al., 1997; Linn et al., 1996;
15 Romieu et al., 2002; Scarlett et al., 1996; Ulmer et al., 1997). All studies, with the exception of
16 Scarlett et al. (1996), observed a statistically significant decrease in FEVj associated with O3
17 exposure. One large study measured spirometric lung function in 895 school children in three
18 towns in Taiwan (Chen et al., 1999). Lung function was measured only once for each subject.
19 The authors reported statistically significant associations between diminished FEVj and FVC
20 with a 1-day lag of O3 concentrations. Effect sizes were typical of those observed in past studies,
21 i.e., 0.5 to 1.0 mL decline in FEVj per ppb increase in O3 concentration. Ozone was the only
22 significant air pollutant in multipollutant models including SO2, CO, PM10, and NO2. The O3
23 associations became nonsignificant when days with O3 above 60 ppb were excluded from the
24 analysis, implying a practical threshold of around 60 ppb in this individual study.
25 Linn et al. (1996) repeatedly measured spirometric lung function among 269 school
26 children in three southern California communities (Rubidoux, Upland, and Torrance). Lung
27 function was measured over five consecutive days, once in each of three seasons over two school
28 years. Between-week variability was effectively removed from the analysis by seasonal terms in
29 the model. Statistical power was limited by the narrow range of exposures that were
30 experienced within each week. In addition, the study was restricted to the school year,
31 eliminating most of the "high" O3 season from consideration. During the study period, 24-h
January 2005 7-27 DRAFT-DO NOT QUOTE OR CITE
-------
1 avg O3 levels at the central monitoring site ranged up to 53 ppb (mean 23 ppb) while personal
2 measurements ranged up to 16 ppb (mean 5 ppb). The difference between morning (tested near
3 the beginning of the school day) and afternoon (tested following lunch) FEVj was significantly
4 associated with same-day O3 concentrations. Other associations (involving individual morning
5 or afternoon FVC and FEVj measurements) went in the plausible direction but were not
6 statistically significant.
7 Ulmer et al. (1997) examined 135 children aged 8 to 11 years in two towns in Germany
8 from March to October 1994 for O3 effects on pulmonary function at four time periods. The
9 cross-sectional results at each of the four time points showed limited FVC and no FEVj
10 associations. However, the longitudinal analysis, which combined data from all four periods,
11 obtained a statistically significant negative association between O3 exposure and both FVC and
12 FEVj for the town with the higher O3 levels (median lA>-h max of 50.6 ppb versus 32.1 ppb).
13 In the cross-sectional analysis, between-person variability could not be distinguished from
14 within-person variability, limiting the statistical power. The longitudinal study design, in which
15 subjects provided multiple days of measurements, had greater power as it provided information
16 about both between- and within-subject responses.
17
18 7.2.3.2 Acute O3 Studies of PEF
19 Many studies of the acute effect of O3 on PEF examined PEF levels daily, both in the
20 morning and afternoon. PEF follows a circadian rhythm with the highest values found during
21 the late afternoon and lowest values during the night and early morning. Due to the diurnal
22 variation in PEF, most studies analyzed their data after stratifying by time of day. The peak flow
23 studies examined both asthmatic panels and healthy individuals. The asthma panels are
24 discussed first.
25
26 Asthma panels
27 Asthmatics were examined in several panel studies. Several aspects of these studies affect
28 the outcomes. For example, large panels have a greater opportunity to test the hypothesis of an
29 O3 effect on PEF. In addition, the severity of asthma in the panel subjects and the medications
30 that they take may affect the results of the study. Figures 7-la and 7-lb present % changes in
31 morning and evening PEF outcomes from six panel studies of mostly asthmatic children. The
January 2005 7-28 DRAFT-DO NOT QUOTE OR CITE
-------
10
U- 5
LJJ
Q_
C
d> n
OJ u
C
re
O
-10
-
-
-
-
-
-
1
Asthmatic
n
X
1
no
yes
1
t
2
—
E
i i l i I I
„ t t I t H I ' -
-
-
3 4 5 6 7 8 9 10 11 12
Code
Figure 7-la. Percent change (95% CI) in morning PEF in children per 40 ppb increase in
1-h max O3 or equivalent, arranged by size of the effect estimate. Study
codes are explained in the tables below. Information on study location and
period, study population, and O3 exposure is presented.
Code Reference
Study Location
Study Period
N
1
2
3
4
5
6
7
8
9
10
11
12
Code
1
2
3
4
5
6
7
8
9
10
11
12
Ross etal. (2002) a
Gold etal. (1999)
Neasetal. (1999)
Neasetal. (1999)
Romieu etal. (1996)
Gielen etal. (1997)
Romieu etal. (1996)
Romieu etal. (1997)
Romieu etal. (1997)
Gielen etal. (1997)
Romieu etal. (1997)
Romieu etal. (1996)
Population
asthmatic
nonasthmatic
nonasthmatic
nonasthmatic
mildly asthmatic
asthmatic
mildly asthmatic
mildly asthmatic
mildly asthmatic
asthmatic
mildly asthmatic
mildly asthmatic
East Mo line, IL
SW Mexico City
Philadelphia, PA
Philadelphia, PA
N Mexico City
May-Oct 1994
Amsterdam, the Netherlands
N Mexico City
SW Mexico City
SW Mexico City
Amsterdam, the Netherlands
SW Mexico City
N Mexico City
Age (years) Mean O3 Level
5-49
8-11
6-11 57.5
6-11 57.5
5-13
7-13
5-13
5-13
5-13
7-13
5-13
5-13
41.5
52
(SW); 55.9
(SW); 55.9
190
35
190
196
196
35
196
190
Jan-Feb
Jul-Sep
Jul-Sep
Apr-Jul
Apr-Jul
Apr-Jul
Apr-Jul
Apr-Jul
Apr-Jul
Apr-Jul
Apr-Jul
(Ppb)
(ME)
(ME)
, Apr-May, Oct-Nov
1993
1993
1991
1995
1991
1991
1991
1995
1991
1991
,Novl991-Feb
,Novl991-Feb
,Novl991-Feb
,Novl991-Feb
,Novl991-Feb
,Novl991-Feb
O3 Exposure Index
8-h max
24-h avg
12-havgb
12-havg"
1-h max
8-h max
1-h max
1-h max
1-h max
8-h max
1-h max
1-h max
1991
1992
1992
1992
1992
1992
1992
40
40
156
156
71
61
71
65
65
61
65
71
Exposure Lag Days
0-1
1-10
1-5
0
0
2
2
0
2
1
1
1
a Study population also includes adults.
b Percent PEF change is presented per 25 ppb increase in 12-h avg O3. The standard units are used for other O3 indices.
January 2005
7-29
DRAFT-DO NOT QUOTE OR CITE
-------
10
Uj 5
Q.
_C
0) 0
re
6
SS -5
-10
. Asthmatic
n no
x yes
.
I ill"
- , | t I ! " f H -
- f I -
1 2 3 4 5 6 7 8 9 10 11 12
Code
Figure 7-lb. Percent change (95% CI) in afternoon PEF in children per 40 ppb increase
in 1-h max O3 or equivalent, arranged by size of the effect estimate. Study
codes are explained in the tables below. Information on study location and
period, study population, and O3 exposure is presented.
Code Reference Study Location Study Period N
1
2
3
4
5
6
1
8
9
10
11
12
Code
1
2
3
4
5
6
1
8
9
10
11
12
Ross etal. (2002) "
Gold etal. (1999)
Romieuetal. (1997)
Romieu etal. (1996)
Romieuetal. (1997)
Neas etal. (1999)
Gielen etal. (1997)
Romieuetal. (1996)
Neas etal. (1999)
Romieuetal. (1996)
Gielen etal. (1997)
Romieuetal. (1997)
Population
asthmatic
nonasthmatic
mildly asthmatic
mildly asthmatic
mildly asthmatic
nonasthmatic
asthmatic
mildly asthmatic
nonasthmatic
mildly asthmatic
asthmatic
mildly asthmatic
East Moline, IL
SW Mexico City
SW Mexico City
N Mexico City
SW Mexico City
Philadelphia, PA
May-Oct 1994
Jan-Feb, Apr-May, Oct-Nov
Amsterdam, the Netherlands
N Mexico City
Philadelphia, PA
N Mexico City
Amsterdam, the Netherlands
SW Mexico City
Age (years) Mean O3 Level
5-49
8-11
5-13
5-13
5-13
6-11 57.
7-13
5-13
6-11 57.
5-13
7-13
5-13
41.5
52
196
190
196
,5 (SW); 55.9
35
190
,5 (SW); 55.9
190
35
196
Apr-Jul
Apr-Jul
Apr-Jul
Jul-Sep
Apr-Jul
Apr-Jul
Jul-Sep
Apr-Jul
Apr-Jul
Apr-Jul
(Ppb)
(ME)
(ME)
1991
1991
1991
1993
1995
1991
1993
1991
1995
1991
,Novl991-Feb
,Novl991-Feb
,Novl991-Feb
,Novl991-Feb
,Novl991-Feb
,Novl991-Feb
O3 Exposure Index
8-h max
24-h avg
1-h max
1-h max
1-h max
12-havgb
8-h max
1-h max
12-havgb
1-h max
8-h max
1-h max
1991
1992
1992
1992
1992
1992
1992
40
40
65
71
65
156
61
6
156
71
61
65
Exposure Lag Days
0
0-9
1
2
0
0
2
1
1-5
0
1
2
1 Study population also includes adults.
b Percent PEF change is presented per 25 ppb increase in 12-h avg O3. The standard units are used for other O3 indices.
January 2005
7-30
DRAFT-DO NOT QUOTE OR CITE
-------
1 individual effect estimates are identified by study codes which are linked to the associated tables
2 that provide study details. The tables present general information on the study location and
3 period as well as specifics on the study population and O3 exposure. Only single city studies that
4 performed analyses stratified by time of day are included in the figure. Studies that examined
5 cross-day changes and daily variability in PEF were excluded from the figure (e.g., Just et al.,
6 2002; Thurston et al., 1997). Collectively, all of the studies indicated decrements of peak flow
7 but most of the individual estimates were not statistically significant.
8 Mortimer et al. (2000, 2002) examined 846 asthmatic children from the National
9 Cooperative Inner-City Asthma Study (NCICAS) for O3-related changes in PEF. Children from
10 eight urban areas in the U.S. (St. Louis, MO; Chicago, IL; Detroit, MI; Cleveland, OH;
11 Washington, DC; Baltimore, MD; East Harlem, NY; and Bronx NY) were monitored from June
12 through August 1993. This multicities study provides representative data for the U.S. Asthmatic
13 children from urban areas are an important subgroup of asthmatics. Study children either had
14 physician-diagnosed asthma and symptoms in the past 12 months or respiratory symptoms
15 consistent with asthma that lasted > 6 weeks during the previous year. In a focused analysis,
16 Mortimer et al. (2000) observed that the subpopulation of low birth weight and premature
17 asthmatic children had significantly greater O3-associated declines in PEF than normal birth
18 weight children.
19 In the main study, Mortimer et al. (2002) further investigated changes in morning PEF
20 associated with O3 concentrations in the eight urban areas. The reductions in morning PEF
21 were not significant in each individual city, however when the data from all eight cities were
22 combined, a statistically significant change of-1.18% (95% CI: -2.10, -0.26) per 30 ppb
23 increase in 8-h avg O3 (10 a.m.-6 p.m.) was observed with a cumulative lag of 1 to 5 days.
24 Figure 7-2 illustrates the probability density curves (or density curves) of the results from the
25 city-stratified analysis and that from the pooled analysis of all eight cities. Summary density
26 curves serve as a descriptive aid to the understanding of multiple effect estimates. These curves
27 can be viewed as smoothed histograms. However, unlike a histogram, summary density curves
28 account for varying standard errors of the individual mean effect estimates. Normal distribution
29 functions can be calculated for each effect estimate and standard error. The density curve for the
30 all cities analysis was calculated by taking the derivative of the normal distribution function
31 from the analysis that pooled data from all eight cities. This density curve is a graphical
January 2005 7-31 DRAFT-DO NOT QUOTE OR CITE
-------
1
0.8
~ 0.6
W
Q °-4
0.2
0
City-Stratified Analysis
All City Analysis
-3
-2 -1
% Change in PEF
Figure 7-2. Density curves of the % change in PEF per 30 ppb increase in 8-h avg O3
with a cumulative lag of 1 to 5 days for the individual eight NCICAS cities and
the pooled average of all cities. Note that 99% and 78% of the areas under
the curves are less than zero for the pooled cities analysis and individual
cities analysis, respectively.
Source: Derived from Mortimer et al. (2002).
1 presentation of the all cities regression analysis presented by Mortimer et al. (2002) and only
2 represents one effect estimate and corresponding standard error. The summary density curve for
3 the city-stratified analysis was calculated by summing together the normal distribution functions
4 for each of the eight cities, then taking the derivative of the summed function. The individual
5 city estimates presented by Mortimer et al. (2002) were used to calculate the summary density
6 curve. Both density curves graphically depict the mean and distribution of the % change in PEF
7 per 30 ppb increase in 8-h avg O3. The area under the density curve and to the left of a value on
8 the x-axis is an estimate of the probability that the effect estimate will be less than or equal to
9 that value. For example, the area under the density curve to the left of 0% change in PEF is 99%
10 in the all cities analysis. As a statistically significant decline in PEF was observed in the all
11 cities regression analysis, it is expected that more than 97.5% (p < 0.05) of this area would be
12 less than zero. A wider distribution was observed in the city-stratified analysis, with only 78%
13 of the area less than zero. The all cities analysis likely had a smaller standard error compared to
14 the city-specific analysis as it was based upon more subjects and considered differences between
15 cities to vary about the same mean effect. The regression analysis by Mortimer et al. (2002)
January 2005
7-32
DRAFT-DO NOT QUOTE OR CITE
-------
1 suggested a lack of heterogeneity by city, as indicated by the nonsignificant interaction term
2 between O3 effect and city. As shown in Figure 7-2, the summary density curve of the city-
3 stratified analysis has a peak at about the same value as the curve of the all cities analysis,
4 suggesting a common O3 effect for all eight cities and small variation among them. The
5 unimodal shape of the density curve of the city-stratified analysis also indicates the absence of
6 outlying cities.
7 Mortimer et al. (2002) further noted that small declines in PEF may be of uncertain clinical
8 significance, thus they calculated the incidence of > 10% declines in PEF. A 5 to 15% change in
9 FEVj has been expressed as having clinical importance to asthma morbidity (American Thoracic
10 Society, 1991; Lebowitz et al., 1987; Lippmann, 1988). Although greater variability is expected
11 in PEF measurements, a > 10% change in PEF also likely has clinical significance. In Mortimer
12 et al. (2002), the incidence of > 10% declines in PEF was statistically significant, indicating that
13 O3 exposure may be associated with clinically significant changes in PEF in asthmatic children.
14 This study also observed that excluding days when 8-h avg O3 levels were less than 80 ppb
15 provided effect estimates which were similar to those when all days were included in the
16 analysis.
17 In Mexico City, two studies of asthmatic school children were carried out simultaneously
18 in the northern (Romieu et al., 1996) and southwestern sections of the city (Romieu et al., 1997).
19 In the northern study, 71 mildly asthmatic school children aged 5 to 13 years old, were followed
20 over time for daily morning (before breakfast) and afternoon (bedtime) PEF. In single-pollutant
21 models, O3 at 0-, 1-, and 2-day lags was associated with diminished morning and afternoon PEF,
22 but only the 0-day lag morning effect was statistically significant. The O3 effect became
23 nonsignificant when PM2 5 was added to the model. In the southwestern study, 65 mildly
24 asthmatic children aged 5 to 13 years old were followed during the summer and winter for daily
25 morning and afternoon PEF. Significant effects at 0- and 1-day lag O3 were observed on
26 afternoon PEF, with effects larger with a 1-day lag. Associations involving O3 were stronger
27 than those involving PM10. Several additional studies, both in the U.S. and in other countries,
28 reported statistically significant associations between O3 exposure and decrements in PEF among
29 asthmatics (Gielen et al., 1997; Jalaludin et al., 2000; Just et al., 2002; Ross et al., 2002;
30 Thurston et al., 1997).
January 2005 7-33 DRAFT-DO NOT QUOTE OR CITE
-------
1 Other epidemiologic studies did not find a significant O3 effect on the lung function of
2 asthmatics. Delfino et al. (1997a) examined morning and evening PEF among 22 asthmatics
3 ranging in age from 9 to 46 years, living in Alpine, CA. Daily ambient 12-h avg O3 (8 a.m.-
4 8 p.m.) concentrations ranged from 34 to 103 ppb, with a mean value of 64 ppb. Unique to this
5 study, personal O3 exposures were measured using 12-h passive O3 samplers that were worn by
6 the subjects. The personal 12-h avg O3 (8 a.m.-8 p.m.) concentrations, which had a mean value
7 of 18 ppb, were much lower than the fixed-site ambient levels. Quantitative O3 results were not
8 reported but researchers stated that no significant O3 effects were observed on morning and
9 evening PEF. In Hiltermann et al. (1998), 60 nonsmoking adults aged 18 to 55 years in
10 Bilthoven, the Netherlands, were followed between July and October 1995 with morning and
11 afternoon PEF measurements. Ozone was associated with declines in PEF, but statistical
12 significance was not achieved.
13 Results from the multicities study by Mortimer et al. (2002), as well as those from several
14 regional studies provide evidence of a significant relationship between O3 concentrations and
15 PEF among asthmatics. Collectively, these studies indicate that O3 may be associated with
16 declines in lung function in this potentially susceptible population.
17
18 Panels of healthy subjects
19 The effect of O3 on PEF in healthy subjects also was investigated in several studies.
20 During the summer of 1990, Neas et al. (1995) examined 83 children in Uniontown, PA and
21 reported twice daily PEF measurements. Researchers found that evening PEF was associated
22 with O3 levels weighted by hours spent outdoors. Using a similar repeated measures design,
23 Neas et al. (1999) saw evidence for effects due to ambient O3 exposure among 156 children
24 attending two summer day camps in the Philadelphia, PA area. Associations were found
25 between afternoon PEF (recorded before leaving camp) and same-day O3 concentrations, and
26 between morning PEF (recorded upon arrival at camp) and previous-day O3 concentrations.
27 However, the relationship between PEF and O3 was statistically significant only when a
28 cumulative lag period of 1 to 5 days was considered. Similarly, Naeher et al. (1999), in a sample
29 of 473 nonsmoking women (age 19 to 43 years) living in Vinton, VA, also showed the largest,
30 significant O3-related decrease in evening PEF with a 5-day cumulative lag exposure.
January 2005 7-34 DRAFT-DO NOT QUOTE OR CITE
-------
1 Another study in southwestern Mexico City analyzed morning and afternoon PEF data
2 collected from 40 school children aged 8 to 11 years (Gold et al., 1999). Subjects provided
3 measurements upon arriving and before departing from school each day. Diminished PEF was
4 associated with 1-day lag O3 concentrations, but the only statistically significant findings were
5 obtained for PEF regressed on O3 concentrations with a cumulative 10-day lag period. This may
6 imply a cumulative effect of O3, or it may reflect confounding by other time-varying factors.
7 These results, however, are in accord with controlled human exposure studies that have shown
8 an attenuated decline in pulmonary function with repeated days of O3 exposure (see Section 6.6,
9 Repeated Exposure to O3), and with epidemiologic studies that have assessed lung function over
10 the course of the O3 season (Brauer et al., 1996; Kinney and Lippmann 2000).
11
12 7.2.4 Respiratory Symptoms
13 Studies published over the past decade represent an improved new body of data on the
14 symptom effects of O3. Respiratory symptoms are usually measured in the context of acute air
15 pollution field studies using questionnaire forms, or "daily diaries," that are filled out by study
16 subjects, usually without the direct supervision of research staff. Questions address the daily
17 experience of coughing, wheezing, shortness of breath (or difficulty breathing), production of
18 phlegm, and others. While convenient and potentially useful in identifying acute episodes of
19 morbidity, measurements of daily symptoms are prone to a variety of errors. These include
20 misunderstanding of the meaning of symptoms, variability in individual interpretation of
21 symptoms, inability to remember symptoms if not recorded soon after their occurrence, reporting
22 bias if days of high air pollution levels are identifiable by subjects, and the possibility of falsified
23 data. In spite of these potential problems, the ease of data collection has made daily symptom
24 assessment a common feature of field studies. Many of the studies reviewed above for lung
25 function results also included measurements of daily symptoms. Pearce et al. (1998) reports that
26 one advantage in the case of asthma panels is that the population is usually already familiar with
27 the symptom terms such as wheezing and cough. Delfino et al. (1998a) further states that the use
28 of repeated daily symptom diaries has additional advantages of reducing recall bias given the
29 proximity of events and allowing health effects to be modeled with each subject serving as their
30 own control over time. Also, study design can blind the participants from the air pollution
January 2005 7-3 5 DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
aspect of the study. Careful efforts by study staff can help ensure that the symptom diaries will
provide information that is less affected by the potential problems noted.
Similar to studies of lung function, respiratory symptom studies can be divided into two
groups, asthma panels or healthy subjects. Asthma panel studies are presented first.
Asthma panels
Most studies examining respiratory symptoms related to O3 exposure focused on asthmatic
children. Among the health outcomes, of particular interest were those associated with asthma,
including cough, wheeze, shortness of breath, and increased medication use. Figures 7-3 and 7-4
present the probability density curves for O3-related cough and medication use among asthmatic
children from six studies (Gielen et al., 1997; Jalaludin et al., 2004; Just et al., 2002; Ostro et al.,
2001; Romieu et al., 1996, 1997). Only single city/region studies that present odds ratios are
included in the figure for consistency. Studies that present change in severity of symptoms,
another informative health outcome, are excluded from the figure since this expression differs
from indicating simple presence of symptoms. The study by Gent et al. (2003) also is excluded
from this figure as odds ratios for cough and mediation use were analyzed for quintiles of O3
concentrations using the lowest quintile as the reference.
0.8
0.9 1 1.1 1.2
Odds Ratio for Cough
1.3
1.4
Figure 7-3. Density curves of the odds ratios for the prevalence of cough among asthmatic
children. Fifteen odds ratios from six studies (Gielen et al., 1997; Jalaludin
et al., 2004; Just et al., 2002; Ostro et al., 2001; Romieu et al., 1996,1997)
are standardized per 40 ppb increase in 1-h max O3 or equivalent.
January 2005
7-36
DRAFT-DO NOT QUOTE OR CITE
-------
40 F^
0.9
1 1.1 1.2
Odds Ratio for Bronchodilator Use
1.3
1.4
Figure 7-4. Density curves of the odds ratios for the prevalence of extra bronchodilator
use among asthmatic children. Nine odds ratios from six studies (Gielen
et al., 1997; Jalaludin et al., 2004; Just et al., 2002; Ostro et al., 2001; Romieu
et al., 1996,1997) are standardized per 40 ppb increase in 1-h max O3 or
equivalent.
1 The various effect estimates for the association between O3 concentrations and cough are
2 depicted as probability density curves (or density curves) in Figure 7-3. Each density curve
3 represents an individual effect estimate and corresponding standard error. The use of density
4 curves allows one to visualize better the distribution of O3-related effects on cough. Despite the
5 variability in the individual effect estimates, there is consistency in the O3 effects as indicated by
6 the considerable overlap in distributions. In general, the majority of the area under the density
7 curves appears to be greater than an odds ratio of one, suggesting a positive association between
8 O3 concentration and cough among asthmatic children. Figure 7-4 presents the density curves
9 for the various effect estimates for O3-associated bronchodilator use. Similar to cough, the
10 majority of the area under the density curves in this figure is greater than an odds ratio of one,
11 indicating that O3 may be associated with increased bronchodilator use in children with mild to
12 severe asthma.
13 Among the studies reporting results for daily symptoms and asthma medication use,
14 several observed associations with O3 concentrations that appeared fairly robust (Delfino et al.,
15 2003; Desqueyroux et al., 2002ab; Gent et al., 2003; Hilterman et al., 1998; Just et al., 2002;
16 Mortimer et al., 2000, 2002; Newhouse et al., 2004; Romieu et al., 1996, 1997; Ross et al., 2002;
17 Thurston et al., 1997).
January 2005
7-37
DRAFT-DO NOT QUOTE OR CITE
-------
1 Mortimer et al. (2002) reported morning symptoms in 846 asthmatic children from eight
2 urban areas of the U.S. to be most strongly associated with a cumulative 4-day lag of O3
3 concentrations in the NCICAS. The NCICAS used standard protocols which included
4 instructing caretakers of the subjects to record symptoms in the daily diary by observing or
5 asking the child (Mitchell et al., 1997). Symptoms reported included cough, chest tightness, and
6 wheeze. In the analysis pooling data from all eight cities, the odds ratio for the incidence of
7 symptoms was 1.35 (95% CI: 1.04, 1.69) per 30 ppb increase in 8-h avg O3. Excluding days
8 when 8-h avg O3 (10 a.m.-6 p.m.) was greater than 80 ppb, the odds ratio was 1.37 (95% CI:
9 1.02, 1.82) for incidence of morning symptoms. Figure 7-5 presents the density curves of the
10 odds ratios for the incidence of symptoms from the city-stratified analysis and that from the all
11 cities analysis. This figure confirms the regression results that there is a significant increase in
12 odds for incidence of symptoms, as the area under the density curve with an odds ratio greater
13 than one is 99%. The unimodal distribution of the city-stratified summary density curve
14 indicates a lack of significant heterogeneity among the eight cities.
15
City-Stratified Analysis
All City Analysis
1 1.5
Odds Ratio for Incidence of Symptoms
Figure 7-5. Density curves of the odds ratios for the incidence of symptoms per 30 ppb
increase in 8-h avg O3 with a cumulative lag of 1 to 4 days for the individual
eight cities and the pooled average of all cities. Note that 99% and 76% of the
areas under the curves are greater than one for the pooled cities and
individual cities analyses, respectively.
Source: Derived from Mortimer et al. (2002).
January 2005
7-38
DRAFT-DO NOT QUOTE OR CITE
-------
1 Another one of the larger studies was that of Gent and colleagues (2003), where
2 271 asthmatic children under age 12 and living in southern New England were followed over
3 6 months (April through September) for daily symptoms. The data were analyzed for two
4 separate groups of subjects, 130 who used maintenance asthma medications during the follow-up
5 period and 141 who did not. The need for regular medication was considered to be a proxy for
6 more severe asthma. Not taking any medication on a regular basis and not needing to use a
7 bronchodilator would suggest the presence of very mild asthma. Significant effects of 1-day lag
8 O3 were observed on a variety of respiratory symptoms only in the medication user group. Both
9 daily 1-h max and 8-h max O3 concentrations were similarly related to symptoms such as chest
10 tightness and shortness of breath. Effects of O3, but not PM25, remained significant and even
11 increased in magnitude in two-pollutant models. Some of the significant associations were noted
12 at 1-h max O3 levels below 60 ppb. In contrast, no significant effects were observed among
13 asthmatics not using maintenance medication. In terms of person-days of follow-up, this is one
14 of the larger studies currently available that address symptom outcomes in relation to O3, and
15 provides supportive evidence for effects of O3 independent of PM2 5.
16 Some international studies have reported significant symptoms associations with O3. The
17 incidence of asthma attacks was significantly associated with O3 concentrations in a group of
18 60 severe asthmatics (mean age 55 years) followed over a 13-month period in Paris
19 (Desqueyroux et al., 2002a). In a similar study, Desqueyroux et al. (2002b) observed significant
20 O3-associated exacerbation of symptoms in 39 adult patients (mean age 67 years) with chronic
21 obstructive pulmonary disease (COPD). Interestingly, in contrast to the controlled human
22 studies (see Section 6.3.1, Subjects with COPD), the O3 effect appeared larger among subjects
23 who smoked and those with more severe COPD. However, the low O3 concentrations
24 experienced during this study (summer mean 8-h max O3 of 41 |ig/m3 or approximately 21 ppb)
25 raise plausibility questions. In a study of 60 nonsmoking asthmatic adults (aged 18 to 55 years)
26 in Bilthoven, the Netherlands, Hilterman and colleagues (1998) reported significant associations
27 between O3 and daily symptoms of shortness of breath and pain upon deep inspiration. The O3
28 associations were stronger than those of PM10, NO2, SO2, and black smoke (BS). No differences
29 in response were evident between subgroups of subjects defined on the basis of steroid use or
30 airway hyperresponsiveness. Daily use of bronchodilators or steroid inhalers was not found to
31 be associated with O3 in this study.
January 2005 7-39 DRAFT-DO NOT QUOTE OR CITE
-------
1 Other studies showed only limited or a lack of evidence for symptom increases associated
2 with O3 exposure (Avol et al., 1998; Chen et al., 1998; Delfmo et al., 1996, 1997a, 1998a;
3 Gielen et al., 1997; Jalaludin et al., 2004; Ostro et al., 2001; Taggart et al., 1996). Ostro et al.
4 (2001) reported no associations between daily symptoms and ambient O3 concentrations in a
5 cohort of 138 African-American children with asthma followed over 3 months (August to
6 October) in Central Los Angeles and Pasadena, CA. However, the use of extra asthma
7 medication was associated with 1-h max O3 concentrations at a 1-day lag. Delfmo and
8 colleagues (1996) followed 12 asthmatic teens living in San Diego, CA for respiratory symptoms
9 over a two-month period and saw no relationship with central site ambient O3. Personal O3
10 exposures measured with passive diffusion monitors were associated with the composite
11 symptom score and p2-agonist inhaler use, but the relationship with symptom score disappeared
12 when weekday/weekend differences were controlled in the statistical analysis. Study power was
13 likely compromised by the small sample size. This observation of stronger effects based on
14 personal monitoring is intriguing; it suggests that substantial gains in power may be achieved if
15 exposure misclassification is reduced through the use of personal exposure measurements rather
16 than central site O3 concentrations. A similar study of 22 asthmatics in Alpine, CA observed no
17 effects of O3 on symptoms when personal O3 exposure was used as the exposure metric (Delfmo
18 et al., 1997a). However, a later study in the same location involving 24 subjects (Delfmo et al.,
19 1998a) did find an association between respiratory symptoms and ambient O3 exposure, with
20 stronger O3 effects experienced by asthmatics not on anti-inflammatory medication. In this
21 study, a binary symptom score was used, whereas the earlier study used a linear symptom score
22 of 0 through 6.
23 In conclusion, the various studies seem to indicate a positive association between O3
24 concentrations, and respiratory symptoms and increased medication use in asthmatics. The
25 multicities study by Mortimer et al. (2002) provides an asthmatic population most representative
26 of the U.S., but several single city studies also add to the knowledge base.
27
28 Panels of healthy subjects
29 Fewer studies examined the effect of O3 on respiratory symptoms in healthy individuals.
30 Neas et al. (1995) reported evening cough was associated with O3 levels weighted by hours spent
31 outdoors in school children. The study by Linn and colleagues (1996) of 269 school children in
January 2005 7-40 DRAFT-DO NOT QUOTE OR CITE
-------
1 southern California reported no associations between respiratory symptoms and O3, but subjects
2 were exposed to fairly low O3 concentrations as determined using personal monitors. Gold et al.
3 (1999) examined symptoms in 40 healthy children in southwest Mexico City. Pollutant
4 exposures were associated with increased production of phlegm in the morning, although the
5 effects of the air pollutants (PM2 5, PM10, and O3) could not be separated in multipollutant
6 models. Hoek and Brunekreef (1995) did not find a consistent association between ambient O3
7 levels, and prevalence and incidence of respiratory symptoms in children living in two rural
8 towns in the Netherlands. Collectively, these studies indicate that there is no consistent evidence
9 of an association between O3 and respiratory symptoms among healthy children.
10
11 7.2.5 Acute Airway Inflammation
12 Acute airway inflammation has been shown to occur among adults exposed to 80 ppb O3
13 over 6.6 hours with exercise in controlled chamber studies (Devlin et al., 1991). Kopp and
14 colleagues (1999) attempted to document inflammation of the upper airways in response to
15 summer season O3 exposures by following a group of 170 school children in two towns in the
16 German Black Forest from March to October of 1994. To assess inflammation, the investigators
17 collected nasal lavage samples at 11 time points spanning the follow-up period. The nasal
18 lavage samples were analyzed for markers of inflammation, including eosinophil cationic
19 protein, albumin, and leukocyte counts. Subjects who were sensitized to inhaled allergens were
20 excluded. When analyzed across the entire follow-up period, no association was detected
21 between upper airway inflammation and O3 concentrations. More detailed analysis showed that
22 the first significant O3 episode of the summer was followed by a rise in eosinophil cationic
23 protein levels, however, subsequent and even higher O3 episodes had no effect. These findings
24 suggest an adaptive response of inflammation in the nasal airways that is consistent with
25 controlled human studies (see Section 6.9, Effects of Inflammation and Host Defense).
26 Frischer and colleagues (1993) collected nasal lavage samples from 44 school children in
27 Umkirch, Germany the morning after "low" O3 days (< 140 |ig/m3 or approximately 72 ppb) and
28 "high" O3 days (> 180 |ig/m3 or approximately 93 ppb) to measure levels of biochemical markers
29 of inflammation. The researchers found that higher O3 levels were significantly associated with
30 increased polymorphonuclear leukocyte counts in all children, and increases in
31 myeloperoxydases and eosinophilic cation proteins among children without symptoms of rhinitis
January 2005 7-41 DRAFT-DO NOT QUOTE OR CITE
-------
1 (n = 30). These results indicated that O3 was associated with inflammation in the upper airways.
2 Frischer et al. (1997) further investigated whether hydroxyl radical attacks played a role in
3 mediating the O3-associated inflammatory response of the airways. Ortho- and/>ara-tyrosine
4 levels were measured in the nasal lavage samples and the ortholpara radical ratio was used to
5 determine the generation of hydroxyl radicals. Significant increases in the ortholpara ratio were
6 observed on days following high ambient O3 levels. However, the ortholpara ratio was not
7 related to polymorphonuclear leukocyte counts, suggesting that there was no detectable
8 relationship between hydroxyl radical attacks and the inflammatory response seen in these
9 children. Similar to the study by Kopp et al. (1999), the ortholpara ratio decreased at the end of
10 the summer although O3 concentrations were still high, providing additional evidence for a
11 possible adaptive response. These findings, however, do not preclude the possibility that other
12 unmeasured effects, including cell damage or lower airway responses, may have occurred with
13 ongoing summer season exposures. In fact, a study of joggers repeatedly exposed to O3 while
14 exercising over the summer in New York City suggested that cell damage may occur in the
15 absence of ongoing inflammation (Kinney et al., 1996).
16 In two Mexico City studies by Romieu et al. (1998, 2002), the effect of antioxidant
17 supplements on the association between O3 and lung function in outdoor workers and asthmatic
18 children was investigated. Romieu and colleagues (1998) observed significant inverse
19 associations between O3 and lung function parameters, including FVC, FEVl3 and FEF25.75
20 (forced expiratory flow at 25 to 75% of FVC), among outdoor workers who were on the placebo,
21 but not among those taking the antioxidant supplement during the 1st phase of testing. Likewise,
22 O3 concentrations were associated with declines in lung function among children with moderate
23 to severe asthma who were on the placebo, but no associations were found among those who
24 were taking the vitamin C and E supplement (Romieu et al., 2002). These results indicate that
25 supplementation with antioxidants may modulate the impact of O3 exposure on the small airways
26 of two potentially susceptible populations, outdoor workers and children with moderate to severe
27 asthma. In a further analysis, genetic factors were found to contribute to the variability between
28 individuals in the effects of O3 on lung function (Romieu et al., 2004). Individuals with
29 polymorphism of the glutathione S-transferase gene (GSTM1 null genotype) lack glutathione
30 transferase enzyme activity, which plays an important role in protecting cells against oxidative
31 damage. Results from this analysis indicate that asthmatic children with GSTM1 null genotype
January 2005 7-42 DRAFT-DO NOT QUOTE OR CITE
-------
1 were found to be more susceptible to the impact of O3 exposure on small airways function.
2 Romieu et al. (2004) noted that supplementation with the antioxidant vitamins C and E above the
3 minimum daily requirement might compensate for the genetic susceptibility.
4
5 7.2.6 Acute O3 Exposure and School Absences
6 The association between school absenteeism and ambient air pollution was assessed in two
7 studies (Chen et al., 2000; Gilliland et al., 2001). In the study by Chen and colleagues (2000),
8 daily school absenteeism was examined in 27,793 students (kindergarten to 6th grade) from
9 57 elementary school students in Washoe County, NV over a two-year period. One major
10 limitation of this study was that the percent of total daily absences was the outcome of interest,
11 not illness-related absences, as reasons for absences were not noted in all schools. In models
12 adjusting for PM10 and CO concentrations, ambient O3 levels were significantly associated with
13 school absenteeism. With a distributed lag of 1 to 14 days, O3 concentrations were associated
14 with a 10.41% (95% CI: 2.73, 18.09) excess rate of school absences per 40 ppb increase in 1-h
15 max O3. PM10 and CO concentrations also were significantly associated with school
16 absenteeism, however, the effect estimate for PM10 was negative. The inverse relationship
17 between O3 and PM10 may have partially attributed to the negative association observed between
18 PM10 and school absenteeism.
19 Ozone-related school absences also were examined in a study of 1,933 4th grade students
20 from 12 southern California communities participating in the Children's Health Study (Gilliland
21 et al., 2001). Due to its size and comprehensive characterization of health outcomes, this study
22 is especially valuable in assessing the effect of O3 on illness-related school absenteeism in
23 children. The study spanned a period, January through June 1996, that captured a wide range of
24 exposures while staying mostly below the highest levels observed in the summer season.
25 All school absences that occurred during this period were followed up with phone calls to
26 determine whether they were illness-related. For illness-related absences, further questions
27 assessed whether the illness was respiratory or gastrointestinal, with respiratory symptoms
28 including runny nose/sneeze, sore throat, cough, earache, wheezing, or asthma attacks. Multiple
29 pollutants were measured at a central site in each of the 12 communities. The statistical analysis
30 controlled for temporal cycles, day of week, and temperature, and expressed exposure as a
31 distributed lag out to 30 days. Some concern exists regarding the possibility of residual seasonal
January 2005 7-43 DRAFT-DO NOT QUOTE OR CITE
-------
1 confounding given the six-month time span of the monitoring period. Significant associations
2 were found between the 30-day distributed lag of 8-h avg O3 (10 a.m.-6 p.m.) and all absence
3 categories. Larger O3 effects were seen for respiratory causes (147% increase per 30 ppb
4 increase in 8-h avg O3) than for nonrespiratory causes (61% increase). Among the respiratory
5 absences, larger effects were seen for lower respiratory diseases with wet cough than for upper
6 respiratory diseases. PM10 was only associated with upper respiratory disease absences.
7 Results from Chen et al. (2000) and, more notably, Gilliland et al. (2001) indicate that
8 ambient O3 concentrations, lagged over two to four weeks, are significantly associated with
9 school absenteeism, particularly respiratory illness-related absences. These two studies on
10 school absenteeism were conducted in communities in Nevada and southern California,
11 however the results are most likely representative of U.S. populations.
12
13 7.2.7 Cardiac Physiologic Endpoints
14 Limited pathophysiologic air pollution studies have examined cardiac physiologic
15 endpoints (Table AX7-2 in Chapter 7 Annex). Several studies examined associations between
16 PM exposures and gaseous pollutants, and various measures of heart beat rhythms in panels of
17 elderly subjects as discussed in the 2004 PM AQCD (Section 8.3.1). Decreased heart rate
18 variability has been identified as a predictor of increased cardiovascular morbidity and mortality.
19 One study examined the increased risk of myocardial infarction and pollutants (Peters et al.,
20 2001). Lack of consistency in the limited studies argued for caution in regards to drawing
21 conclusions on the relationship between cardiovascular outcomes and PM exposure. Among
22 these studies, Gold et al. (2000; reanalysis Gold et al., 2003) and Peters et al. (2000a, 2001)
23 discussed limited evaluation of a potential role for O3 exposure. In addition, two recent studies
24 provided limited evidence for an association between O3 concentrations and heart rate variability
25 in primarily elderly populations (Holguin et al., 2003; Park et al., 2004). Two related studies by
26 Rich et al. (2004) and Vedal et al. (2004) examined the relationship between various air
27 pollutants, including O3, and cardiac arrhythmias using two different study designs, but both did
28 not find any consistent evidence that exposure to air pollution affected the risk of arrhythmias.
29 Two limited controlled human exposure studies with cardiovascular outcomes (Gong et al.,
30 1998a; Superko et al., 1984), described in Chapter 6, Section 6.3.4, provide no supporting data.
January 2005 7-44 DRAFT-DO NOT QUOTE OR CITE
-------
1 The above panel studies with small numbers of subjects had limited ability to adequately
2 test the hypothesis of heart rate variability. A recent large population-based study, the first in
3 this field, examined PM10, O3, and other gaseous air pollutants, and their potentially adverse
4 effects on cardiac autonomic control (Liao et al., 2004). Liao et al. investigated short-term
5 associations between ambient pollutants and cardiac autonomic control from the 4th cohort
6 examination (1996-1998) of the population-based Atherosclerosis Risk in Communities Study.
7 PM10 (24-h avg) and O3 exposure (8-h avg, 10 a.m.-6 p.m.) one day prior to the randomly
8 allocated examination date were used. They calculated 5-minute heart rate variability indices
9 between 8:30 a.m. and 12:30 p.m. and used logarithmically-transformed data on high-frequency
10 (0.15 to 0.40 Hz) and low-frequency (0.04 to 0.15 Hz) power, standard deviation of normal
11 R-R intervals, and mean heart rate. The effective sample sizes for PM10 and O3 were 4,899 and
12 5,431, respectively, from three U.S. study centers in North Carolina, Minnesota, and Mississippi.
13 PM10 concentrations measured one day prior to the heart rate variability measurements were
14 inversely associated with both frequency and time domain heart rate variability indices.
15 Ambient O3 concentrations were inversely associated with high-frequency power among whites.
16 Consistently more pronounced associations were suggested between PM10 and heart rate
17 variability among persons with a history of hypertension. These findings were cross-sectionally
18 derived from a population-based sample and reflect only the short-term effects of air pollution
19 on heart rate variability. When the regression coefficients for each individual pollutant model
20 were compared, the effects for PM10 was considerably larger than the effects for gaseous
21 pollutants such as O3. While these data are supportive of the hypothesized air pollution-heart
22 rate variability-cardiovascular disease pathway at the population level, replication of these
23 interactions in other studies is needed before any conclusions can be made.
24
25 7.2.8 Summary of Field Studies Assessing Acute O3 Effects
26 • Results from recent field/panel studies support the evidence from clinical studies that
acute O3 exposure is associated with a significant effect on lung function, as indicated
by decrements in FEVj, FVC, and PEF. The declines in lung function were noted
particularly in children and asthmatics.
27 • Limited evidence suggests that more time spent outdoors, higher levels of exertion,
and the related increase in O3 exposure may potentiate the risk of respiratory effects.
In addition to children and asthmatics, adults who work or exercise outdoors may be
particularly susceptible to O3-associated health effects.
January 2005 7-45 DRAFT-DO NOT QUOTE OR CITE
-------
4
5
Many new studies have examined the association between O3 concentrations and a
wide variety of respiratory symptoms (e.g., cough, wheeze, production of phlegm,
and shortness of breath). Collectively, the results indicate that acute exposure to O3
is associated with increased respiratory symptoms and increased as-needed medication
use in children and asthmatics.
Additional panel studies investigated the effect of O3 on other health outcomes,
including school absences, and markers of inflammation and oxidative damage.
Ozone exposure was associated with significant increases in respiratory-related school
absences, as well as increased inflammation and generation of hydroxyl radicals in the
upper airways. Use of antioxidant supplements was found to diminish the O3 effect on
lung function.
Few field studies have examined the association between O3 and cardiac physiologic
outcomes. The current evidence is rather limited but supportive of a potential effect on
heart rate variability. Additional studies need to be performed before any conclusions
can be made regarding an O3 effect on cardiovascular outcomes.
6 7.3 EFFECTS OF OZONE ON DAILY EMERGENCY DEPARTMENT
7 VISITS AND HOSPITAL ADMISSIONS
8 7.3.1 Summary of Key Findings on Studies of Emergency Department Visits
9 and Hospital Admissions from the 1996 O3 AQCD
10 In the 1996 O3 AQCD, aggregate population time-series studies of O3-related health effects
11 provided relevant evidence of acute responses, even below a 1-h max O3 of 0.12 ppm.
12 Emergency room visits and hospital admissions were examined as possible outcomes following
13 exposure to O3. In the case of emergency room visits, the evidence was limited (Bates et al.,
14 1990; Cody et al., 1992; Weisel et al., 1995; White et al., 1994), but results generally indicated
15 an O3 effect on morbidity. The strongest and most consistent evidence of O3 effects, at levels
16 both above and below 0.12 ppm 1-h max O3, was provided by the multiple studies that had been
17 conducted on summertime daily hospital admissions for respiratory causes in various locales in
18 eastern North America (Bates and Sizto, 1983, 1987, 1989; Burnett et al., 1994; Lipfert and
19 Hammerstrom, 1992; Thurston et al., 1992, 1994). These studies consistently demonstrated that
20 O3 air pollution was associated with increased hospital admissions, accounting for roughly one to
21 three excess respiratory hospital admissions per million persons with each 100 ppb increase in
22 1-h max O3. This association had been shown to remain even after statistically controlling for
23 the possible confounding effects of temperature and copollutants (e.g., FT, SO4 2, PM10), as well
January 2005 7-46 DRAFT-DO NOT QUOTE OR CITE
-------
1 as when considering only days with 1-h max O3 concentrations below 0.12 ppm. Furthermore,
2 these results implied that O3 air pollution could account for a substantial portion of summertime
3 hospital admissions for respiratory causes on the most polluted days. Overall, the aggregate
4 population time-series studies considered in the 1996 O3 AQCD provided strong evidence that
5 ambient exposures to O3 can cause significant exacerbations of preexisting respiratory disease in
6 the general public at concentrations below 0.12 ppm.
7
8 7.3.2 Review of Recent Studies of Emergency Department Visits for
9 Respiratory Diseases
10 Emergency department visits represent an important acute outcome that may be affected by
11 O3 exposures. Morbidities that result in emergency department visits are closely related to, but
12 are generally less severe than, those that result in unscheduled hospital admissions. In many
13 cases, acute health problems are successfully treated in the emergency department; a subset of
14 more severe cases that present initially to the emergency department may require admission to
15 the hospital.
16 Several studies have been published in the past decade examining the temporal
17 associations between O3 exposures and emergency department visits for respiratory diseases
18 (Table AX7-3 in Chapter 7 Annex). Total respiratory causes for emergency room visits may
19 include asthma, pneumonia, bronchitis, emphysema, other upper and lower respiratory infections
20 such as influenza, and a few other minor categories. Asthma visits typically dominate the daily
21 incidence counts. Chronic bronchitis and emphysema often are combined to define COPD,
22 which is a prominent diagnosis among older adults with lung disease. Figure 7-6 presents %
23 changes in emergency department visits for asthma, with results expressed in standardized
24 increments. Results from all lags presented are included in the figure. Weisel et al. (2002) was
25 excluded as relative risks were not presented and could not be estimated. Among the U.S.
26 studies, there was one multicity study which examined three cities in Ohio (Jaffe et al., 2003).
27 Several presented Atlanta, GA data. In general, O3 effect estimates from summer only analyses
28 tended to be positive and larger compared to results from cool season or all year analyses.
29 Among studies with adequate controls for seasonal patterns, many reported at least one
30 significant positive association involving O3. These studies examined emergency department
31 visits for total respiratory complaints (Delfmo et al., 1997b, 1998b; Hernandez-Garduno et al.,
January 2005 7-47 DRAFT-DO NOT QUOTE OR CITE
-------
% Change in Emergency Department Visits for Asthma
to
o
o
oo
Jaffe etal. (2003):
Cleveland OH (aqe 5-34)
Columbus OH (aqe 5-34)
Cincinnati, Cleveland, —
and Columbus, OH (age 5-34)
Friedman etal. (2001):
Atlanta, GA (age 1-1 6)
Peel etal. (2004):
Atlanta, GA (all ages)
Tolbert etal. (2000):
Atlanta, GA (age 0-1 6)
Stieb etal. (1996):
St. John, Canada (all ages)
Atkinson et al. (1999a):
London, England (all ages)
Atkinson etal. (1999a):
London, England (age 0-14)
Atkinson etal. (1999a):
London, England (agel 5-64)
Thompson etal. (2001):
Belfast, N. Ireland (children)
Castellsagueetal. (1995):
Barcelona, Spain (age 15-65)
Tobias etal. (1999):
Barcelona, Spain (age >14)
Teniasetal. (1998):
Valencia, Spain (age > 14)
-20 0 20 40 60 80 100 120 140 160 180 20C
i i i i i i i i i i i i i i i i i i i i i i i
| U.S. and Canada
X All year
£ bg ° 9 Warm
• I -in "3
lag J O Cool
• Hn n
0 lag 0-1
lag u-z
-X- lag 0-2
• lag 0-2 (SE not given)
— •— lag 1
• Inn °
| Europe |
X — lagO
—X— Iag1
^ laa 0
x" Ian D-1
X bg 0 °
X lay 0-3
« lin 0 1
lag 0-1
A lin n
v |uy u
— X lagO
V inn 1
©hn 1
Figure 7-6. Ozone-associated % change (95% CI) in emergency department visits for asthma per 40 ppb increase in 1-h max
O3 or equivalent.
-------
1 1997; Ilabaca et al., 1999; Jones et al., 1995; Lin et al., 1999), asthma (Friedman et al., 2001;
2 Jaffe et al., 2003; Stieb et al., 1996; Tenias et al., 1998; Tobias et al., 1999; Tolbert et al., 2000;
3 Weisel et al., 2002), and COPD (Tenias et al., 2002).
4 One recent study examined emergency department visits for total and cause-specific
5 respiratory diseases in Atlanta, GA over an 8-year period (Peel et al., 2004). A distributed lag of
6 0 to 2 days was specified a priori. Ozone concentrations were significantly associated with
7 emergency department visits for total respiratory diseases and upper respiratory infections in all
8 ages. A marginally significant association was observed with asthma visits (2.6% excess risk
9 per 30 ppb increase in 8-h max O3), which became stronger when analysis was restricted to the
10 warm months (3.1% excess risk). In multipollutant models adjusting for PM10, NO2 and CO, O3
11 was the only pollutant that remained significantly associated with upper respiratory infections.
12 Another large asthma emergency department study was carried out during the months of May
13 through September from 1984 to 1992 in St. John, New Brunswick, Canada (Stieb et al., 1996).
14 Effects were examined separately among children aged less than 15 years and in persons aged
15 15 years and older. A significant effect of O3 on emergency department visits was reported
16 among persons 15 years and older. There was evidence of a threshold somewhere in the range
17 below a 1-h max O3 of 75 ppb. A study in Valencia, Spain from 1994 to 1995 observed that
18 emergency room visits for asthma among persons over 14 years old were robustly associated
19 with relatively low O3 levels (median 1-h max O3 of 62.8 |ig/m3 or approximately 32.4 ppb)
20 (Tenias et al., 1998). The excess risk of asthma emergency room visits was larger in the warm
21 season (May to October), 85% excess risk per 40 ppb increase in 1-h max O3, compared to the
22 cool season (November-April), 31% excess risk (Tenias et al., 1998).
23 Among the studies that observed a statistically significant association between O3 and
24 emergency department visits for respiratory outcomes, O3 effects were found to be robust to
25 adjustment for PM10, NO2, SO2, and black smoke (Lin et al., 1999; Peel et al., 2004; Tenias et al.,
26 1998). One study by Tolbert and colleagues (2000) observed that the significant univariate
27 effects of both O3 and PM10 on pediatric asthma emergency department visits in Atlanta, GA
28 became non-significant in two-pollutant regressions, reflecting the high correlation between the
29 two pollutants (r = 0.75).
30 For several other "positive" studies with total respiratory and asthma outcomes,
31 inconsistencies confound an interpretation of likely causal effects. For example, in a Montreal,
January 2005 7-49 DRAFT-DO NOT QUOTE OR CITE
-------
1 Canada study, O3 effects on total respiratory emergency department visits were seen in a short
2 data series from the summer of 1993 but not in a similar data series from the summer of 1992
3 (Delfino et al., 1997b). The significant 1993 results were seen only for persons older than
4 64 years, in spite of greater asthma prevalence among children. A very similar analysis of two
5 additional summers (1989 and 1990) revealed an O3 association only for 1989 and again only in
6 persons over 64 years old (Delfino et al., 1998b). An analysis of data on respiratory emergency
7 department visits from June to August of 1990 in Baton Rouge, LA reported O3 effects in adults,
8 but not in children or among the elderly (Jones et al., 1995).
9 Tobias and colleagues (1999) showed that regression results for asthma emergency
10 department visits could be quite sensitive to methods used to control for asthma epidemics.
11 Ozone was associated with the outcome variable in only one of eight models tested. An Atlanta,
12 GA study by Zhu et al. (2003) examined asthma emergency department visits in children during
13 three summers using Bayesian hierarchical modeling to address model variability. Data was
14 analyzed at the zip code level to account for spatially misaligned longitudinal data. Results
15 indicated a positive, but nonsignificant relationship between O3 and emergency room visits
16 for asthma.
17 Other studies also reported nonsignificant findings for O3 (Atkinson et al., 1999a;
18 Castellsague et al., 1995; Chew et al., 1999). One study by Thompson and colleagues (2001)
19 in Belfast, Northern Ireland observed a significant 21% decrease in risk of childhood asthma
20 admissions per 20 ppb increase in 24-h avg O3 in the cold season (November-April). After
21 adjusting for benzene levels, O3 was no longer associated with asthma emergency department
22 visits. The inverse relationship of O3 with benzene concentrations (r = -0.65), and perhaps with
23 other pollutants, might have produced the apparent protective effect of O3. No significant O3
24 effect was found in the warm season (May-October). The O3 levels were low in both seasons,
25 with a mean 24-h avg O3 concentration of 18.7 ppb in the warm season and 17.1 ppb in the cold
26 season. Atkinson et al. (1999a) in London, England also did not find an association between O3
27 and emergency department visits at a mean 8-h max O3 concentration of 17.5 ppb. Several other
28 emergency department studies looking at O3 are more difficult to interpret due to inadequate
29 control for seasonal patterns, very low O3 levels, or because no quantitative results were shown
30 for O3 (Buchdahl et al., 1996, 2000; Garty et al., 1998; Holmen et al., 1997; Lierl and Hornung,
31 2003; Lipsett et al., 1997; Nutman et al., 1998).
January 2005 7-50 DRAFT-DO NOT QUOTE OR CITE
-------
1 Although several studies found a significant association between O3 concentrations and
2 emergency department visits for respiratory causes, some inconsistencies were observed. The
3 inconsistencies may be attributable, at least partially, to differences in study design and model
4 specifications among the various studies. For example, ambient O3 concentrations, length of the
5 study period, and statistical methods used to control confounding by seasonal patterns and
6 copollutants appear to affect the observed O3 effect on emergency department visits. In general,
7 an excess risk of emergency department visits was observed during the summer season when O3
8 concentrations were higher.
9
10 7.3.3 Studies of Hospital Admissions for Respiratory Diseases
11 Hospital admissions represent a medical response to a serious degree of morbidity for a
12 particular disease. Scheduled hospitalizations are planned in advance when a particular clinical
13 treatment is needed. However, unscheduled admissions are ones that occur in response to
14 unanticipated disease exacerbations and are more likely to be affected by environmental factors,
15 such as air pollution. As such, the hospital admissions studies reviewed here focused
16 specifically on unscheduled admissions. Study details and results from hospital admissions
17 studies published over the past decade are summarized in Table AX7-4 (in the Chapter 7
18 Annex). As a group, these hospitalization studies tend to be larger in terms of geographic and
19 temporal coverage, and indicate results that are generally more consistent than those reviewed
20 above for emergency department visits. The following aspects of these studies should be
21 considered in comparing results: (1) difference in type of respiratory diseases for hospital
22 admission; (2) analysis by season versus all year; (3) O3 only versus multipollutant models;
23 (4) age of study population; (5) number of exposure lag days; (6) single-city versus multicity
24 studies; (7) mean level of O3 during study; (8) length of study (e.g., < 5 years versus > 5 years);
25 and (9) type of study (e.g., case-crossover versus time-series).
26 Figures 7-7 through 7-9 present risk estimates from all total respiratory hospital admission
27 studies. Burnett et al. (1995), which did not present quantitative results for O3, and Yang et al.
28 (2003), which only presented odd ratios, were excluded from the figure. In cases where multiple
29 lags were presented, the multiday lag was selected to represent the cumulative effect from all
30 days examined. For Luginaah et al. (2004), cumulative lags are not analyzed, thus the effect
31 estimates from a 1-day lag are included in this figure. For studies that presented risk estimates
January 2005 7-51 DRAFT-DO NOT QUOTE OR CITE
-------
Linn et al. (2000): Los Angeles,CA (age 30+)
Gwynn et al. (2000): Buffalo, NY (all ages)
Burnett et al. (2001): Toronto, Canada (age <2)
Luginaah et al. (2004): Windsor, Canada (all ages, males only)
Luginaahetal. (2004): Windsor, Canada (all ages,females only)
Atkinson et al. (1999b): London, England (all ages)
Atkinson et al. (1999b): London, England (age 0-14)
Atkinson etal.(1999b):London,England (age 15-64)
Atkinson etal.(1999b): London, England (age 65+)
Poncede Leon etal.(1996):London,England (all ages)
Ponce de Leon et al. (1996): London, England (age 0-14)
Ponce de Leon et al. (1996): London, England (age 15-64)
Ponce de Leon et al. (1996): London, England (age 65+)
Prescott et al. (1998): Edinburgh, Scotland (age <65)
Prescott et al. (1998): Edinburgh, Scotland (age 65+)
Schouten etal. (1996): Amsterdam, the Netherlands (age 15-64)
Schouten et al. (1996): Amsterdam, the Netherlands (age 65+)
Schouten etal. (1996): Rotterdam, the Netherlands (age 15-64)
Hagen et al. (2000): Drammen, Norway (all ages)
Oftedal et al. (2003): Drammen, Norway (all ages)
Gouveia and Fletcher (2000a):Sao Paulo, Brazil (age<5) —
Petroeschevskyetal.(2001):Brisbane,Austrailia (all ages)
Petroeschevsky et al. (2001): Brisbane, Austrailia (age 0-4)
Petroeschevsky et al. (2001): Brisbane, Austrailia (age 5-14)
Petroeschevsky etal. (2001): Brisbane, Austrailia (age 15-64)
Petroeschevsky et al. (2001): Brisbane, Austrailia (age 65+)
Wongetal.(1999a): Hong Kong (all ages)
Wong etal.(1999a): Hong Kong (age 0-4)
Wong etal.(1999a): Hong Kong (age 5-64)
Wong et al. (1999a): Hong Kong (age 65+)
% Change in Respiratory Hospitalization
-20 -10 0 10 20 30
I I I I I I I I I I I
U.S. and Canada
Europe
lagO
Asia
lagO
9- Iag1
lag 0-4
Iag1
Iag1
lagO
-• lag 0-2
lag 0-3
lag 0-3
lag 0-2
lag 0-1
lag 0-5
- lagO
lagO
lagO
lag 2
lagO
lag 2
lag 3-
—• lag 0-3
-• lag 0-3
• lag 0-3
lag 0-3
Figure 7-7. Ozone-associated % change (95% CI) in total respiratory hospitalizations
for all year analyses per 40 ppb increase in 1-h max O3 or equivalent.
January 2005
7-52
DRAFT-DO NOT QUOTE OR CITE
-------
% Change in Respiratory Hospitalizations
to
o
o
H
6
o
o
H
O
O
H
W
O
O
HH
H
W
Linn etal. (2000):
Los Angeles, CA (age 30+)
Schwartz etal. (1996):
Cleveland, OH (age 65+)
Burnett etal.(1997a):
16 Canadian Cities (all ages)
Burnett etal.(1997b):
Toronto, Canada (all ages)
Burnett etal. (2001):
Toronto, Canada (age < 2)
Ponce de Leon etal.(1996):
London, England (all ages)
Ponce de Leon etal.(1996):
London, England (age 0-14)
Ponce de Leon etal.(1996):
London, England (age 15-64)
Ponce de Leon etal.(1996):
London, England (age 65+)
Schouten etal.(1996):
Amsterdam,the Netherlands (age 15-64)
Schouten etal.(1996):
Amsterdam, the Netherlands (age 65+)
Schouten etal.(1996):
Rotterdam, the Netherlands (age 15-64)
Schouten etal.(1996):
Amsterdam and Rotterdam, the Netherlands (all ages)
Wongetal.(1999a):
Hong Kong (all ages)
0
|u.S. and Canada
— €
— €
PI
| Europe]
-e
|Asia
0 20 40 60 80 10C
i i i i i i i i i i
-K-
£«- lagO
• lin 1 °
g * lag 1
• I--, H o
*• lag 0-2
A Ian 0-2
0 lag 0 °
• inn n °
K Inn n °
vv
• Ian 0 1
3
• Ian O-^
X All year
h -•" lag 2
h • Warm
x O Cool
n lag 0-3
Figure 7-8. Ozone-associated % change (95% CI) in total respiratory hospitalizations by season per 40 ppb increase in 1-h
max O3 or equivalent.
-------
-20
% Change in Respiratory Hospitalization
0 10 20
Burnett etal.(1997b):
Toronto, Canada
(all ages, warm season only) —
with PM2 5
withPM10
with PM10_25
Burnett etal. (2001):
Toronto, Canada
(age <2, warm season only) -
with PM2 5
With PM10_25-
Ponce de Leon etal. (1996):
London, England (all ages) -
with BS-
Hagen etal.(2000):
Dram men, Norway (all ages)-
withPM10-
Gouveia and Fletcher (2000a):
Sao Paulo, Brazil (age <5) -
withPM10-
Petroeschevsky etal. (2001):
Brisbane, Australia (all ages)-
with nephelometer -
Petroeschevsky etal. (2001):
Brisbane, Australia (ages 15-64)-
with nephelometer -
Wongetal.(1999a):
Hong Kong (all ages)-
withPM10-
-10
I
10
_l
30
I
| Europe
-e-
Latin America
Australia
Asia
-e-
lag 1-3
-e-
Iag1
lagO
lagO
-e-
Iag2
-e-
Iag2
-e-
lag 0-3
40
I
50
Figure 7-9. Ozone-associated % change (95% CI) in total respiratory hospitalizations
with adjustment for PM indices per 40 ppb increase in 1-h max O3 or
equivalent. Analyses performed using all year data unless noted otherwise.
1 from all four seasons, only the summer and winter estimates are presented. Figure 7-7 plots the
2 relative risk estimates and 95% CIs from 13 studies that analyzed all year data. The
3 preponderance of positive risk estimates, with some that are statistically significant, is readily
4 apparent. The impact of seasonal and multipollutant analyses were further examined in Figures
5 7-8 and 7-9. In Figure 7-8, it appears that the warm season estimates, collectively, tend to be
January 2005
7-54
DRAFT-DO NOT QUOTE OR CITE
-------
1 larger, positive values compared to all year and cool season estimates. Most of the negative
2 estimates were from analyses using cool season data only, which might reflect the inverse
3 correlation between O3 and copollutants, namely PM, during that season. Figure 7-9 compares
4 the risk estimates from models with and without adjustment for PM indices. This figure
5 indicates that O3 risk estimates are fairly robust to PM adjustment in the all year and warm
6 season only data. None of the studies examined PM-adjusted O3 risk estimates in cool season
7 only data.
8 The most robust and informative results on the effects of O3 on respiratory hospital
9 admissions are those from studies carried out using a consistent analytical methodology across a
10 broad geographic area (Anderson et al., 1997; Burnett et al., 1995, 1997a). These studies have
11 all reported a significant O3 effect on respiratory hospital admissions. The largest such study
12 to-date was carried out using data on all-age respiratory hospital admissions from 16 Canadian
13 cities with populations exceeding 100,000 covering the period 1981 to 1991 (Burnett et al.,
14 1997a). In addition to O3, the authors evaluated health effects of SO2, NO2, CO, and coefficient
15 of haze (a surrogate for black carbon particle concentrations). Pooling the 16 cities, a significant
16 positive association was observed between respiratory hospital admissions and the 1-day lag O3
17 concentration in the spring (5.6% excess risk per 40 ppb increase in 1-h max O3) and summer
18 (6.7%). The results for fall were also positive, though of smaller magnitude (3.8%). There was
19 no evidence for an O3 effect in the winter season (-0.8%). Control outcomes related to blood,
20 nervous system, digestive system, and genitourinary system disorders were not associated with
21 O3. In a previous study focused mainly on evaluating health impacts of sulfate particles, Burnett
22 and colleagues (1995) reported results from a time-series analysis of all-age respiratory hospital
23 admissions to 168 hospitals in Ontario, Canada over the 6-year period 1983 to 1988. The
24 outcome data were prefiltered to remove seasonal variations using a weighted 19-day moving
25 average. The authors reported that O3 was associated with respiratory hospital admissions;
26 however no quantitative results for O3 were presented.
27 Results from an analysis of five European cities indicated strong and consistent O3 effects
28 on unscheduled hospital admissions for COPD (Anderson et al., 1997). The five cities examined
29 - London, Paris, Amsterdam, Rotterdam, and Barcelona - were among those included in the
30 multicity APHEA (Air Pollution on Health: European Approach) study. The number of years of
31 available data varied from 5 to 13 years among the cities. In addition to O3, the study considered
January 2005 7-55 DRAFT-DO NOT QUOTE OR CITE
-------
1 health impacts of BS, TSP, NO2, and SO2. City-specific effect estimates were pooled across
2 cities using weighted means. Significant effects were seen for O3, BS, TSP, and NO2. Ozone
3 effects were statistically significant in full year analyses and were larger in the warm season
4 (April-September), 4.6% excess risk per 40 ppb increase in 1-h max O3, compared to the cool
5 season (October-March), 1.5% excess risk. The authors reported that among all pollutants
6 examined, the most consistent and significant findings were for O3. In addition, there was no
7 significant heterogeneity in O3 effects among the cities. No two-pollutant model results were
8 reported.
9 Several additional studies carried out in one or two cities over a span of five or more years
10 provided substantial additional evidence regarding O3 effects on respiratory hospital admissions
11 (Anderson et al., 1998; Burnett et al., 1999, 2001; Moolgavkar et al.,1997; Petroeschevsky et al.,
12 2001; Ponce de Leon et al., 1996; Sheppard et al., 1999 [reanalysis Sheppard, 2003]; Yang et al.,
13 2003). Two separate analyses of a large dataset from Toronto, Canada spanning the years 1980
14 to 1994 reported significant O3 effects on respiratory hospitalizations for all ages (Burnett et al.,
15 1999) and for persons under the age of 2 years (Burnett et al., 2001). Analysis was performed
16 using Poisson GAM (default convergence criteria) with a nonparametric LOESS prefilter applied
17 to the pollution and hospitalization data. Both studies demonstrated that O3 effects were robust
18 when PM measures were added to the regression, whereas PM effects from univariate
19 regressions were markedly attenuated when O3 was added to the regression. These results imply
20 more robust associations with respiratory hospitalizations for O3 than PM.
21 Moolgavkar and colleagues (1997) reported significant and robust O3 effects on respiratory
22 hospital admissions in adults 65 years and older in Minneapolis and St. Paul, MN, but not in
23 Birmingham, AL. The absence of effects in the southern city may reflect less penetration of O3
24 into the indoor environment due to greater use of air conditioning, and thus less correlation
25 between central site O3 monitoring and actual exposures of the urban populace. In Brisbane,
26 Australia during 1987 to 1994, significant O3 effects on all-age and age-stratified asthma and
27 total respiratory hospital admissions were observed (Petroeschevsky et al., 2001). The O3 effects
28 were robust to inclusion of PM (based on light scattering) and SO2 in copollutant regression
29 models. Effect sizes appeared consistent in the warm and cool seasons, possibly reflecting the
30 relatively small degree of seasonal variation in O3 levels observed in Brisbane.
January 2005 7-56 DRAFT-DO NOT QUOTE OR CITE
-------
1 Less consistent effects of O3 were seen in other respiratory hospitalization studies
2 (Schouten et al., 1996; Lin et al., 2003; Lin et al., 2004; Morgan et al., 1998a; Oftedal et al.,
3 2003). In a study conducted in Amsterdam and Rotterdam, the Netherlands, significant
4 associations were observed, however results were difficult to interpret due to the large number of
5 statistical tests performed (Schouten et al., 1996). Using a different analytical approach
6 (case-crossover analysis), Lin and colleagues (2003) found no evidence for O3 effects on asthma
7 admissions in 6- to 12-year-olds over the period 1981 to 1993 in Toronto, Canada. In a
8 California study by Niedell (2004), a negative association was observed between hospitalizations
9 for asthma and naturally occurring seasonal variations in O3 within zip codes in children aged
10 0 to 18 years. However, the O3 effect was found to be influenced by socioeconomic status.
11 Among children of low socioeconomic status, O3 generally was associated with increased
12 hospitalizations, with statistical significance reached in certain age groups. Niedell further stated
13 that avoidance behavior on high O3 days may have attributed to the negative relationship
14 observed in children of higher socioeconomic status.
15 Another set of studies have examined associations between O3 and respiratory
16 hospitalizations in single cities over shorter (< 5 years) time spans. Positive and significant O3
17 effects were reported in Cleveland, OH (Schwartz et al., 1996); Northern New Jersey (Weisel
18 et al., 2002); Toronto, Canada (Burnett et al., 1997b); Helsinki, Finland (Ponka and Virtanen,
19 1996); Sao Paulo, Brazil (Gouveia and Fletcher, 2000a); and Hong Kong (Wong et al., 1999a).
20 The Helsinki study reported significant effects of O3 on both asthma and on digestive disorders
21 in a setting of very low O3 concentrations (Ponka and Virtanen, 1996), which raises questions
22 of plausibility.
23 No significant association with O3 was seen in studies from Los Angeles, CA (Linn et al.,
24 2000; Mann et al., 2002; Nauenberg and Basu, 1999); Vancouver, Canada (Lin et al., 2004);
25 London, England (Atkinson et al., 1999b); Edinburgh, Scotland (Prescott et al., 1998); and
26 Drammen, Norway (Hagen et al., 2000). Several of the studies reporting non-significant O3
27 effects were carried out in locations with low O3 levels, suggestive of a nonlinear exposure-
28 response relationship (Lin et al., 2004; Prescott et al., 1998). The non-significant findings in the
29 South Coast air basin, CA area are surprising given the elevated O3 concentrations observed
30 there (Mann et al., 2002). Inadequate control of seasonal confounding may underlie some of the
January 2005 7-57 DRAFT-DO NOT QUOTE OR CITE
-------
1 non-significant and negative findings. An additional factor likely contributing to the variability
2 of results is the relatively small sample sizes included in some of these studies.
3 In conclusion, while some inconsistencies are noted across studies, the evidence supports
4 the findings of significant and robust effects of O3 on various respiratory disease hospitalization
5 outcomes. Large multicity studies, as well as many studies from individual cities have reported
6 significant O3 associations with total respiratory hospitalizations, asthma, and COPD, especially
7 in studies analyzing the O3 effect during the summer or warm season.
8
9 7.3.4 Association of O3 with Hospital Admissions for Cardiovascular Disease
10 Among the subset of hospital admissions studies that have examined associations of O3
11 with cardiovascular outcomes, most have found no consistent positive associations (Ballester
12 et al., 2001; Burnett et al., 1995, 1999; Linn et al., 2000; Mann et al., 2002; Petroeschevsky
13 et al., 2001; Prescott et al., 1998). The exceptions are one study in Toronto, Canada, which
14 reported robust associations with both total respiratory and cardiovascular hospital admissions
15 (Burnett et al., 1997b), and one in Hong Kong, in which circulatory, ischemic heart, and heart
16 failure were all significantly associated with O3 in the cool but not the warm season (Wong et al.,
17 1999b). In the Hong Kong study, O3 concentrations were similar in both seasons, with warm
18 season levels slightly lower, mean 31.2 |ig/m3, compared to the cool season, mean 34.8 |ig/m3.
19 The authors speculated that differing activity patterns and home ventilation factors may have
20 contributed to the seasonal differences in O3 effects. Weather in Hong Kong is mild throughout
21 the year, but less humid and cloudy in the cool season. Thus, during the cool season people are
22 more likely to open windows or stay outdoors, resulting in higher personal exposures even with
23 similar ambient concentrations. Based on this small set of studies, current evidence does not
24 support a conclusion that O3 has independent effects on cardiovascular hospitalizations.
25
26 7.3.5 Summary of Acute O3 Effects on Daily Emergency Department Visits
27 and Hospital Admissions
28 • The vast majority of hospitalization studies conducted over the past decade have
looked at effects of O3 on either total respiratory diseases and/or asthma. Significant
associations with O3 were observed with both outcomes in many cases. Studies of
emergency department visits for respiratory conditions also reported significant
O3 effects, but the results tend to be less consistent across studies.
January 2005 7-58 DRAFT-DO NOT QUOTE OR CITE
-------
1 • Many of the daily emergency department visits and hospitalization studies analyzed
O3 risk estimates using year-round data. Given the strong seasonal variations in
O3 concentrations and the changing relationship between O3 and other copollutants
by seasons, inadequate adjustment for seasonal effects might have masked or
underestimated the association between O3 and the respiratory disease outcomes.
Season stratified analyses typically yield more reliable O3 effect estimates.
2 • Numerous studies have reported O3 effects while controlling for copollutants, including
PM, in the analytical model. The evidence is supportive of independent O3 effects on
respiratory admissions and emergency department visits. In most studies, O3 effects
have been reported to be at least as robust as PM, and in some cases more so.
3 • A subset of hospital admission studies examined the effect of O3 on cardiovascular
outcomes. The limited evidence is inconclusive regarding the association between
O3 exposure and cardiovascular hospitalizations.
4
5
6 7.4 ACUTE EFFECTS OF OZONE ON MORTALITY
7 7.4.1 Summary of Key Findings on Acute Effects of O3 on Mortality From
8 the 1996 O3 AQCD
9 A limited number of studies examined O3-mortality associations at the time of the previous
10 O3 AQCD, most of which were from the 1950s and 1960s. The 1996 O3 AQCD considered these
11 historical studies to be flawed because of either inadequate adjustment for seasonal trend or
12 temperature, or because of the use of questionable exposure indices. There were only a few
13 time-series studies that examined O3-mortality associations between the 1980s and mid-1990s.
14 These studies used more sophisticated approaches in addressing seasonal confounding and
15 weather models. One of these studies (Shumway et al., 1988) focused on the associations with
16 long-term fluctuations in Los Angeles, CA but did not examine short-term associations. A study
17 that reanalyzed the Los Angeles, CA data with a focus on the short-term associations (Kinney
18 and Ozkaynak, 1991) did find that, of the PM and gaseous criteria pollutants, O3 (reported as
19 total oxidants) was most strongly associated with total nonaccidental mortality. Then two
20 studies, one using Detroit, MI data (Schwartz, 1991) and the other using St. Louis, MO and
21 Kingston-Harriman, TN data (Dockery et al., 1992), reported that PM but not O3 was
22 significantly associated with mortality. However, the 1996 O3 AQCD discussed that, without
23 sufficient presentation of model diagnosis regarding the relationship between O3 and the weather
January 2005 7-59 DRAFT-DO NOT QUOTE OR CITE
-------
1 models used, it was difficult to evaluate whether the lack of O3-mortality associations was
2 possibly due to overspecification of the weather model. In summary, due to the insufficient
3 number of studies that examined O3-mortality associations and the uncertainties regarding
4 weather model specifications, the 1996 O3 AQCD was unable to quantitatively assess O3-
5 mortality excess risk estimates, or even provide qualitative assessment of the likelihood of
6 O3-mortality associations.
7
8 7.4.2 Introduction to Assessment of Current O3-Mortality Studies
9 Introductory discussions of the PM mortality effects often cite historical air pollution
10 incidents such as the 1952 London, England smog episode in which thousands of deaths were
11 attributed to the air pollution from coal burning. There is no counterpart "historical episode" for
12 O3-mortality effects. Instead, the early recognition of the adverse health effects of summer
13 oxidant air pollution, mainly from Los Angeles and other major cities with a high density of
14 automobiles, were based on symptoms such as eye and throat irritations. Thus, the focus of PM
15 epidemiology and that of O3 epidemiology have been historically different.
16 As shown in Table AX7-5 in the Chapter 7 Annex, the number of short-term mortality
17 studies that analyzed O3 has increased markedly since the last publication of the O3 AQCD in
18 1996. The increased attention to PM-mortality associations in the early 1990s lead to the
19 increase in studies that also examined O3, most often as a potential confounder for PM.
20 Although many of these PM studies also reported O3 estimates, they often lacked specific
21 hypotheses regarding mortality effects of O3 as the focus of these studies was to examine the
22 PM-mortality effect. This is in contrast to the O3-morbidity studies, most of which were
23 specifically designed to examine effects of "summer haze" and O3 (or oxidants) on respiratory
24 and other symptoms, lung functions, and emergency department visits, etc. However, new
25 studies with hypotheses developed specifically for O3 effects on mortality have become
26 available, such as the large U.S. 95 communities study by Bell et al. (2004), the U.S. 14 cities
27 study by Schwartz (2004), and the 23 European cities study by Gryparis et al. (2004) discussed
28 in the next section.
29
January 2005 7-60 DRAFT-DO NOT QUOTE OR CITE
-------
1 7.4.3 Single-Pollutant Model O3-Mortality Risk Estimates
2 To facilitate a quantitative overview of the O3-mortality effect estimates and their
3 corresponding uncertainties, the percent excess risks of total nonaccidental mortality calculated
4 using all year data are plotted in Figure 7-10. Studies that only conducted seasonal analyses will
5 be presented in the next section. This figure does not include studies that only examined cause-
6 specific mortality. In studies where multiple lags were presented, the multiday lag was selected
7 to represent the cumulative effect from all days examined. If cumulative lags were not analyzed,
8 the effect estimate from the 0- or 1-day lag was selected for presentation. All effect estimates
9 are from single-pollutant models and include all age groups unless noted otherwise. The
10 majority of the estimates are positive with a few exceptions. Four multicity studies showed
11 positive and significant O3 effect estimates for all cause (nonaccidental) mortality. An excess
12 mortality risk of 4.5% per 40 ppb increase in 1-h max O3 was estimated from the four European
13 cities of the APHEA project (Touloumi et al., 1997). The European effect estimate was larger
14 than those from the large U.S. National Morbidity, Mortality and Air Pollution Study
15 (NMMAPS). An excess risk of 1.0% and 0.8% per 20 ppb increase in 24-h avg O3 was observed
16 from the U.S. 95 communities study (Bell et al., 2004) and U.S. 90 cities study (Samet et al.,
17 2000; reanalysis Dominici et al., 2003), respectively. Similarly, the U.S. 14 cities study by
18 Schwartz (2004) observed an 0.8% excess risk per 40 ppb increase in 1-h max O3.
19 Only one multicity study did not observe a statistically significant O3 effect on mortality.
20 As an extension of the four European cities study, researchers of the APHEA project
21 investigated the effect of O3 on total, cardiovascular, and respiratory mortality in 23 cities
22 throughout Europe (Gryparis et al., 2004). A cumulative lag of 0 to 1 days was hypothesized
23 a priori. A two-stage hierarchical model, which accounted for statistical variance and
24 heterogeneity among cities, was used to estimate the pooled regression coefficients. Due to
25 substantial heterogeneity among cities, random effects regression models were applied. The
26 pooled effect estimate for the 23 European cities was a positive, but nonsignificant value of
27 0.23% (95% CI: -0.85, 1.95) per 40 ppb increase in 1-h max O3 for all seasons. The researchers
28 noted that there was a considerable seasonal difference in the O3 effect on mortality, thus the
29 nonsignificant effect for the all year data might be attributable to inadequate adjustment for
30 confounding by season. This seasonal effect will be discussed further in the next section.
31
January 2005 7-61 DRAFT-DO NOT QUOTE OR CITE
-------
Bell et al. (2004): U.S.95 communities -
Sametetal.(2000;reanalysis Dominic! etal.,2003): U.S.90cities -
Schwartz(2004): U.S.14dties-
Kinney and Ozkaynak (1991): Los Angeles County, CA —
Kinneyetal.(1995): Los Angeles County, CA -
Fairley (2003): Santa Clara County,CA -
Gamble (1998): Dallas,TX-
Dockeryetal.(1992): St. Louis, MO -
ItoandThurston (1996): Cook County, IL -
Lippmannetal. (2000; reanalysis Ito, 2004): Detroit, Ml -
Lipfertetal.(2000a): Philadelphia,PA -
Lipfert et al.(2000a): 4 Counties in PA -
Lipfert et al. (2000a): 7 Counties in PA and NJ -
Lipfert et al. (2000a): 7 Counties in PA and NJ (age 0-65) -
Lipfert et al. (2000a): 7 Counties in PA and NJ (age 65+) -
Chock etal. (2000): Pittsburgh, PA (age 0-74) -
Chocketal.(2000): Pittsburgh,PA (age 75+) -
Dockery et al. (1992): Eastern Tennessee —
Klemm and Mason (2000): Atlanta, GA -
Klemm etal. (2004): Atlanta, G A (age 65+) -
Gryparis etal.(2004): 23 European cities —
Touloumi etal. (1997): 4 European cities (Athens,Barcelona,London, Paris) —
Anderson etal (1996): London,England —
Bremneretal.(1999): London,England —
Bremner etal. (1999): London, England (age 0-64) —
Bremner etal.(1999): London, England (age 65+) —
Bremner etal. (1999): London, England (age 65-74) -
Bremner etal.(1999): London, England (age 75+) —
Prescottetal.(1998): Edinburgh,Scotland -
Zmirouetal.(1996): Lyon,France —
Hoek etal. (2000; reanalysis Hoek,2003): The Netherlands -
RoemerandvanWijnen (2001): Amsterdam, the Netherlands —
Verhoeffetal.(1995): Amsterdam,the Netherlands -
Peters et al. (2000b): Coal basin in Czech Republic -
Peters et al.(2000b): NE Bavaria,Germany -
Garcia-Aymerichetal.(2000): Barcelona,Spain —
Sunyeretal.(1996): Barcelona,Spain —
Sunyeretal.(1996): Barcelona,Spain (age70+) —
Borja-Aburto et al. (1997): Mexico City —
Borja-Aburto et al (1997): Mexico City (age <5) -
Borja-Aburto et al. (1997): Mexico City (age 65+) -
Borja-Aburto et al. (1998): SW Mexico City -
Borja-Aburto et al. (1998): SW Mexico City (age 65+) -
O'Neill etal. (2004): Mexico City-
O'Neill et al. (2004): Mexico City (age 65+) -
Gouveia and Fletcher (2000b): Sao Paulo, Brazil —
Gouveia and Fletcher (2000b): Sao Paulo, Brazil (age 65+) -
Saldiva etal.(1995): Sao Paulo,Brazil (age65+) -
Ostroetal.(1996): Santiago,Chile -
Morgan etal. (1998b): Sydney, Australia —
Simpson etal.(1997): Brisbane,Australia —
Simpson etal.(1997): Brisbane,Australia (age 0-64) -
Simpson etal. (1997): Brisbane, Australia (age 65+) —
Kim etal. (2004): Seoul, Korea -
Lee eta I. (1999): Seoul, Korea -
Lee and Schwartz (1999): Seoul, Korea —
Lee etal. (1999): Ulsan, Korea -
% Change in Mortality
-10
10
I
20
lag 0-1 •
lag 0-1 •
lag 0-1 •
lag 0-1 •
lag 0-1 •
Latin America
> lag 0-6
lagO
lagO
Iag1
Iag1
© Multicity combined
• Single city
lag 0-1
I Confidence bands not provided but
noted as significant at p=0.055
• lagO
- Iag1
lag 0-1
lag 0-1
lagO
lag 0-6
- lagO
-•
- lags
- lagO
Iag1
lagO
lag 0-1
• lagO
•-
• lagO
• lagO
lagO
Figure 7-10. All cause (nonaccidental) O3 excess mortality risk estimates (95% CI) for all
year analyses per 40 ppb increase in 1-h max O3 or equivalent. Analyses
include all ages unless otherwise noted.
January 2005
7-62
DRAFT-DO NOT QUOTE OR CITE
-------
1 The U.S. 95 communities study by Bell et al. (2004) extended an earlier analysis (Samet
2 et al., 2000; reanalysis Dominici et al., 2003) on short-term effects of O3 on total and
3 cardiopulmonary mortality using the NMMAPS data base from 1987 to 2002. The results of this
4 study are discussed in detail here because of the study's emphasis on U.S. data and the inclusion
5 of 95 large communities across the country, making this mortality study most representative of
6 the U.S. population. In addition, this study is one of few that have focused specifically on O3
7 hypotheses testing. Within-community results first were calculated using single-day lags of 0, 1,
8 and 2 days, and a 7-day distributed lag in O3 exposure. A two-stage hierarchical model was used
9 to determine a national average effect estimate. Figure 7-11 presents community-specific and
10 national average O3 risk estimates for total mortality per 10 ppb increase in 24-h avg O3 from a
11 constrained 7-day distributed lag model. For total mortality in the single-day lag models, the
12 national estimates of the O3 effect on mortality were statistically significant for 0-, 1-, and 2-day
13 lags, with the largest effect observed from O3 concentrations at a 0-day lag. The 7-day
14 distributed lag model estimated the cumulative risk of mortality associated with O3
15 concentrations on the same day and six previous days. Results from the constrained distributed
16 lag model indicated that a 10 ppb increase in 24-h avg O3 in the previous week was associated
17 with a statistically significant increase of 0.52% (95% CI: 0.27, 0.77%) in daily mortality. The
18 mean 24-h avg O3 concentration was approximately 26 ppb for the 95 communities. Figure 7-11
19 illustrates the preponderance of cities with a positive yet nonsignificant relationship between O3
20 and mortality. However, the national average O3 mortality estimate, which includes results from
21 all 95 U.S. communities, is positive and statistically significant. These results further support
22 the data presented in Figure 7-10 from nearly 40 studies in locations both in the U.S. and in other
23 countries.
24 Several studies conducted meta-analyses of O3-mortality associations (Stieb et al., 2002,
25 2003; Thurston and Ito, 2001; World Health Organization, 2004). Most of these studies included
26 GAM studies using default convergence criteria except Stieb et al. (2003), which compared
27 effect estimates from GAM-affected studies to non-GAM studies. All of these meta-analyses
28 reported fairly consistent and positive combined estimates, approximately 2% excess total
29 non-accidental mortality per 40 ppb increase in 1-h max O3. However, most of these studies
30 were not analytical in design in that they did not attempt to examine the source of heterogeneity,
January 2005 7-63 DRAFT-DO NOT QUOTE OR CITE
-------
% Change in Mortality
Orlando
_
_ , . .
Las Veaas
Birmingham
— npino
«Lt P^tpr«hnrn
„ P .
^anta Ana 1 Anahipm
1 nkr Phnrlr-
Tucson
P Phri"Ti
^T 1 nni"
Mn^lfrnnn
!__.._,, pjttf
— j'l .
...
Baltimore
-
.. _ i
Jackson
Portland OR
_ '.
1 h "t
Austin
San Rprnarlino
P 'rl
Cleveland
flrlinntnn
' M ri t
1 inrnln
Washington
° UmMad!son
,, .
Phrirniv
Knoxville
1 nni«:villp
Hnnnlnln
Rochester
....
.,, P
Colorado Sennas
„ P ~
Phiririn
Shreveoort
NATIONAL —
-1 C
I
*
1123
I I I
—
5
5
™
^
^
^
5
*
5
*
*
~
^
2
5
*
5
^
~
*
~
~
^
5
"
' _
' ^
— . —
Figure 7-11. Bayesian city-specific and national average estimates for the % change (95%
CI) in daily mortality per 10 ppb increase in 24-h avg O3 in the previous
week using a constrained distributed lag model for 95 U.S. communities
(NMMAPS), arranged by size of the effect estimate.
Source: Derived from Bell et al. (2004).
January 2005
7-64
DRAFT-DO NOT QUOTE OR CITE
-------
1 although some suggested an influence of weather model specification (Thurston and Ito, 2001)
2 and another reported evidence of publication bias (World Health Organization, 2004) in the past
3 literature. None of these studies addressed the issue of season-specific estimates, and therefore,
4 interpreting these combined estimates requires caution. The estimates from these meta-analyses
5 appear to be larger than the national average estimate of 1.0% excess risk per 20 ppb increase in
6 24-h avg O3 from the largest U.S. 95 communities study (Bell et al., 2004). There are a few new
7 meta-analyses and multicity studies currently being conducted specifically to address such issues
8 as season-specific analyses, publication bias, weather model specification, potential confounding
9 by fine particles, distributed lag effects, and the potential influence of air conditioning. These
10 studies are expected to provide new information that will shed light on the outstanding questions.
11 Collectively, the above studies suggest an excess risk of total nonaccidental mortality
12 associated with acute O3 exposure. Despite the different analytical approaches and alternative
13 model specifications used in the various studies, overall, the range of estimates were relatively
14 narrow, with the positive estimates from 0 to 7% per 40 ppb increase in 1-h max O3 or
15 equivalent.
16
17 7.4.4 Seasonal Variation in O3-Mortality Risk Estimates
18 Since the seasonal cycle of O3 follows the seasonal cycle of temperature (which is
19 inversely related to the mortality seasonal cycle), inadequate adjustment of temporal trends in
20 the regression model may lead to negative O3-mortality risk estimates. In addition, as discussed
21 in Section 7.1.3.5, in some cities low-level O3 during winter may be negatively correlated with
22 PM and other primary pollutants, resulting in negative correlations between O3 and mortality
23 even in short-term relationships. The confounding effect by season could be substantially
24 reduced by conducting season-stratified analyses.
25 A fewer number of O3 mortality studies performed seasonal analyses. Figure 7-12 presents
26 the studies that reported O3 risk estimates for all cause mortality by season. For those studies
27 that obtained O3 risk estimates for each of the four seasons, only summer and winter results are
28 shown. The estimates for year-round data analyses, when available, also are shown for
29 comparisons. In all the studies, the O3 risk estimates are larger during the warm season than the
30 cool season, with the all year estimates generally in between the two seasonal estimates.
31
January 2005 7-65 DRAFT-DO NOT QUOTE OR CITE
-------
% Change in Mortality
Bell etal.(2004): U.S.95 communities -
Samet et al.(2000; reanalysis
Dominici et al.2003):U.S.90 cities"
Schwartz (2004): U.S. 14 cities -
Ostro et al. (1995): San Bernadino County
and Riverside County, CA
Gamble (1998):Dallas,TX-
Moolgavkaretal.(1995):Philadelphia,PA -
Chock etal.(2000): Pittsburgh, PA (age 0-74) -
Chock etal. (2000): Pittsburgh, PA (age 75+) -
Vedaletal.(2003):Vancouver,Canada -
Gryparis et al. (2004): 21 European cities -
Anderson etal.(1996): London,England -
Hoeketal.(2000;reanalysis
Hoek,2003):The Netherlands"
Sunyer et al. (1995): Barcelona, Spain -
Sunyeretal.(1996):Barcelona,Spain (age 70+) -
Borja-Aburto et al. (1997): Mexico City -
Cifuentes et al.(2000):Santiago,Chile -
Ostro etal. (1996): Santiago,Chile -
Simpson etal.(1997):Brisbane,Australia -
Kim etal.(2004):Seoul,Korea -
-10
I
10
I
20
30
I
|u.S. and Canada]
-e-
*
• lag 0-6
<
• Iag1
<*r
-0-
lag 1 -2
lagO
Latin America
Australia
-e-
-e-
-e-
lagO
lagO
• (SE not given; significant at 0.05 level)
O (SE not given; not significant at 0.05 level)
~ Iag1
lagO
lagO
-G-
-B-
lag 0-1
lagO
,ag1
-e-
-9-
lagO
Iag1
lagO
lag 1-2
Iag1
-e-
Iag1
lagO
Figure 7-12. All cause (nonaccidental) O3 excess mortality risk estimates (95% CI)
by season per 40 ppb increase in 1-h max O3 or equivalent. Analyses include
all ages unless otherwise noted.
January 2005
7-66
DRAFT-DO NOT QUOTE OR CITE
-------
1 In three U.S. and European multicity studies (Gryparis et al., 2004; Samet et al., 2000
2 [reanalysis Dominici et al., 2003]; Schwartz, 2004), season-stratified analyses indicated that the
3 O3-mortality effect estimates were significant and positive in the warm season, with larger
4 effects observed compared to the year-round analyses. The effect estimates from the cool season
5 were notably smaller and less or nonsignificant. In the case of the U.S. 90 cities study, the cool
6 season mortality estimate was negative, which was most likely attributable to the inverse
7 relationship between O3 and PM in the winter.
8 In the U.S. 95 communities study by Bell et al. (2004), the warm season (April-October)
9 effect estimate was 0.78% (95% CI: 0.26, 1.30) excess risk per 20 ppb increase in 24-h avg O3,
10 compared to 1.04% (95% CI: 0.54, 1.55) calculated using all available data. The small
11 difference in the size of the effect estimates between the analysis using all available data and that
12 using warm season only data might be attributable, at least partially, to the varying peak O3
13 seasons and the difference in O3 concentrations by community or region. In addition, the
14 varying seasonal relationship between O3 and PM by community also might have contributed to
15 the difference, as these mortality effect estimates were not adjusted for confounding by PM.
16 Studies that conducted analysis by season indicate that O3 mortality risk estimates are often
17 larger in the warm season compared to the colder season. The larger effects observed in the
18 warm season when O3 levels are higher are consistent with causal association. The seasonal
19 dependence of O3-mortality effects complicates interpretation of O3 risk estimates calculated
20 from year-round data without adequate adjustment of temporal trends.
21
22 7.4.5 O3-Mortality Risk Estimates Adjusting for PM Exposure
23 As previously mentioned, the confounding between "winter type" pollution (e.g., CO, SO2,
24 and NO2) and O3 is not of great concern because the peaks of these pollutants do not strongly
25 coincide. The main confounders of interest for O3, especially for the northeast U.S., are
26 "summer haze" type pollutants such as acid aerosols and sulfates. Since very few studies had
27 these chemical measurements, PM (especially PM2 5), may serve as surrogates. However, due to
28 the expected high correlation among the constituents of the "summer haze mix," multipollutant
29 models including these pollutants may result in unstable coefficients, and therefore, an
30 interpretation of such results requires some caution.
31
January 2005 7-67 DRAFT-DO NOT QUOTE OR CITE
-------
1 Figure 7-13 shows the O3 risk estimates with and without adjustment for PM indices using
2 all year data in studies that conducted two-pollutant analyses. Approximately half of the O3 risk
3 estimates slightly increased while the other half slightly decreased in value with the inclusion of
4 PM indices in the model. In general, the O3-mortality risk estimates were robust to adjustment
5 for PM in the models, with the exception of Los Angeles, CA data with PM10 (Kinney et al.,
6 1995) and Mexico City data with TSP (Borja-Aburto et al., 1997).
7 The U.S. 95 communities study by Bell et al. (2004) examined the sensitivity of acute
8 O3-mortality effects to potential confounding by PM10. Restricting analysis to days when both
9 O3 and PM10 data were available, the community-specific O3-mortality effect estimates as well as
10 the national average results indicated that O3 was robust to adjustment for PM10 (Bell et al.,
11 2004). One study (Lipfert et al., 2000a) reported O3 risk estimates with and without sulfate
12 adjustment. Lipfert et al. (2000a) calculated O3 risk estimates based on mean (45 ppb) less
13 background (not stated) levels of 1-h max O3 in seven counties in Pennsylvania and New Jersey.
14 The O3 risk estimate was not substantially affected by the addition of sulfate in the model (3.2%
15 versus 3.0% with sulfate) and remained statistically significant.
16 Several O3-mortality studies examined the effect of confounding by PM indices in different
17 seasons (Figure 7-14). In analyses using all year data and warm season only data, O3 risk
18 estimates were once again fairly robust to adjustment for PM indices, with values showing both
19 slight increases and decreases with the inclusion of PM in the model. In contrast, in the analyses
20 using cool season data only, the O3 risk estimates all increased with the adjustment of PM
21 indices, although none reached statistical significance. For example, in the European study of
22 21 cities (two cities that did not have 8-h max O3 data were excluded from the analysis), the
23 summer O3-mortality estimate was relatively robust to adjustment for PM10, slightly decreasing
24 from 1.82% (95% CI: 0.99, 3.06) to 1.58% (95% CI: 0.47, 2.88) excess risk per 30 ppb increase
25 in 8-h max O3 (Gryparis et al., 2004). In contrast, the winter effect estimate increased from
26 0.70% (95% CI: -0.70, 2.17) to 1.29% (95% CI: -0.46, 3.00) per 30 ppb increase in 8-h max O3
27 after adjusting for PM10. These results indicate that the confounding effect by PM may vary by
28 season. Although PM does not appear to significantly confound the association between O3 and
29 mortality in the analysis of warm season data, during the cool season, the inverse relationship
30 between O3 and PM10 may influence the effect estimate for O3-related mortality.
31
January 2005 7-68 DRAFT-DO NOT QUOTE OR CITE
-------
% Change in Mortality
Samet et al, (2000, reanalysis
Dominid et al., 2003): U.S. 90 cities -
with PM10 -
Schwartz (2004): U.S. 14 cities -
with PM10 -
Kinney and Ozkaynak (1991): Los Angeles County, CA -
with KM -
Kinney etal. (1995): Los Angeles County CA -
with PM10 -
with PM10
Lipfertetal. (2000a): Philadelphia, PA -
withPM25 -
with PM10 -
Lipfert et al. (2000a): 4 Counties in PA -
withPM25 -
with PM10 -
Lipfert et al. (2000a): 7 Counties in PA and NJ -
withPM26 -
with PM10 -
Lipfert et al. (2000a): 7 Counties in PA and NJ (age 0-65) -
withPM25 -
with PM10 ~
Lipfertetal. (2000a): 7 Counties in PA and NJ (age 65+) -
withPM25 -
with PM10 -
with PM10 -
Touloumietal. (1997): 4 European cities
with BS
Anderson et al.(1996): London, England -
with BS -
Hoek(2000, reanalysis Hoek, 2003): The Netherlands -
with PM10 -
Borja-Aburto etal. (1997): Mexico City -
with TSP -
Borja-Aburto et al. (1997): Mexico City (age <5) -
Borja-Aburto et al. (1997): Mexico City (age 65+) -
with TSP
Ostroetal. (1996): Santiago, Chile -
are 7-13. All cause (nonaccid
with adjustment foi
10 -5 0 5 10 15
I ill
I 1
\¥±] ^_ lag 0 o O3 only
I§I lag 0 • Q3 with PM
"^ lagi
^~ Iag1
f~\
lag 0 ° -
9 laa 1-2
9 laa 1-2
lag 0-1 0° A
Q / Confidence bands not
lag °-1 • /*• provided but noted as
* \ significant at p=0.055
O )
lag 0-1 • _J Noted as nonsignificant
^ at p=0.055
0 ~N
lag 0-1 • 1
9 / Confidence bands not
>• provided but noted as
laq 0 1 2 \ significant at p=0.055
0° lag 0
•° laaO
| Europe |
ff laa varied by city
^ lagO
~^" Iag1
^ laa 2
Latin America ^ ^ lag 0
§ laqO
- ~e~ laq 0
^ lag 1-2
t lag 1-7
2 laq 1
1 1 c~\
|Asia| ~ laa 1
W
ental) O3 excess mortality risk estimates (95% CI)
r PM indices for all year analyses per 40 ppb increas
1-h max O3 or equivalent. Analyses include all ages unless otherwise noted.
January 2005
7-69
DRAFT-DO NOT QUOTE OR CITE
-------
% Change in Mortality
-15
Ostro et al. (1995): San Bernadino County
and Riverside County, CA —
with PM2.5 -
Moolgavkar et al. (1995): Philadelphia, PA -
with TSP -
with TSP -
Chock et al. (2000): Pittsburgh, PA (age 0-74) -
with PM10 —
with PM10 -
Chock et al. (2000): Pittsburgh, PA (age 75+) -
with PM10 •
with PM10 -
with PM,0 -
Gryparisetal.(2004):21 European cities —
with PM,n -
10 "
with PM1(
Anderson etal.( 1996):London, England -
with BS -
with BS -
with BS •
Ostro et al. (1996): Santiago, Chile -
with PM10 •
with PM,n •
Kim etal.(2004): Seoul, Korea -
with PM10 -
with PM10 -
U.S.
Europe
-10
J
Latin America
-5
I
5
I
-e-
-e-
-x-
-e-
-e-
-e-
-e-
-x-
X All year • Warm O Cool
lagO
lag 1
lagO
lag 0
lag 0-1
-e-
lag 0
-e-
-e-
Iag1
-X-
10
Iag1
Figure 7-14. All cause (nonaccidental) O3 excess mortality risk estimates (95% CI)
with adjustment for PM indices by season per 40 ppb increase in 1-h max
O3 or equivalent. Analyses include all ages unless otherwise noted.
January 2005
7-70
DRAFT-DO NOT QUOTE OR CITE
-------
1 7.4.6 O3 Risk Estimates for Specific Causes of Mortality
2 Many of the time-series mortality studies examined broad underlying causes of mortality,
3 such as cardiovascular and respiratory causes. The U.S. 95 communities study (Bell et al., 2004)
4 analyzed O3 effect estimates from cardiovascular and respiratory mortality. Significant effects
5 were seen at 0- and 2-day lags with results similar to total mortality. The national average
6 estimate from the constrained distributed lag model was slightly greater for cardiopulmonary
7 deaths, with an excess risk of 1.28% (95% CI: 0.62, 1.97) per 20 ppb increase in 24-h avg O3 in
8 the preceding week. In a related study, Huang et al. (2004) examined O3 effects on
9 cardiopulmonary mortality during the summers of 1987 to 1994 in 19 large U.S. cities from the
10 NMMAPS database. In the 7-day distributed lag model, the O3 effect estimate was 2.52% (95%
11 CI: 0.94, 4.10) excess risk in cardiopulmonary mortality per 20 ppb increase in 24-h avg O3
12 (Huang et al., 2004). Several studies observed that the risk estimates for the respiratory category
13 were larger than the cardiovascular and/or total nonaccidental categories (e.g., Anderson et al.,
14 1996; Gouveia and Fletcher, 2000b; Gryparis et al., 2004; Zmirou et al., 1998). In the European
15 21 multicities study (Gryparis et al., 2004), the warm season effect estimate for respiratory
16 mortality was 6.75% excess risk per 30 ppb increase in 8-h max O3, compared to 2.70% for
17 cardiovascular mortality and 1.82% for total mortality. In contrast, other studies have found that
18 the risk estimates for the respiratory category were essentially zero or even negative while the
19 risk estimates for total or cardiovascular categories were positive (e.g., Borja-Aburto et al., 1998;
20 Bremner et al., 1999; Lipfert et al., 2000a; Morgan et al., 1998b). These apparent
21 inconsistencies across studies may be due in part to the difference in model specifications, but
22 they may also reflect the lower statistical power associated with the smaller daily counts of the
23 respiratory category (usually accounting for less than 10% of total deaths) compared to the larger
24 daily counts for the cardiovascular category (approximately 40 to 50% of total deaths). Thus, an
25 examination of the differences in risk estimates across specific causes requires a large population
26 and/or a long period of data collection.
27 The analyses of a 9-year data set for the whole population of the Netherlands
28 (population = 14.8 million) provided O3 (and other pollutants) risk estimates for more specific
29 causes of mortality, including COPD, pneumonia, and subcategories of cardiovascular causes
30 (Hoek et al., 2000, 2001; reanalysis Hoek, 2003). The excess risk estimate for COPD was small
January 2005 7-71 DRAFT-DO NOT QUOTE OR CITE
-------
1 and not significant (0.8% [95% CI: -2.4, 4.2] per 30 ppb 8-h avg O3), while the excess risk
2 estimate for pneumonia (5.6% [95% CI: 1.8, 9.5]) was much larger than that for total
3 nonaccidental mortality (1.6% [95% CI: 0.9, 2.4]). The excess risk estimates for some of the
4 cardiovascular subcategories, including heart failure (3.8% [95% CI: 0.5, 7.3]) and thrombosis-
5 related disease (6.0% [95% CI: 1.1, 10.8]), showed greater risk estimates than that for total
6 mortality. However, these elevated relative risks were not specific to O3. For example, most of
7 the pollutants examined, including PM10, BS, SO2, NO2, CO and NO3 , were significantly
8 associated with pneumonia. Therefore, it is difficult to make a causal inference specific to O3
9 based on these results.
10 De Leon et al. (2003) examined the role of contributing respiratory causes in the
11 associations between air pollution and nonrespiratory mortality (circulatory and cancer) in
12 New York City during the period of 1985 to 1994. The main finding of this study was that, for
13 the older age group (75+ years), the estimated excess mortality risks for PM10 were higher for the
14 nonrespiratory deaths that had contributing respiratory causes, compared to the nonrespiratory
15 deaths without contributing respiratory causes. This pattern was also seen for CO and SO2, but
16 not for O3. Therefore, this study did not suggest a role of contributing respiratory causes in the
17 association between O3 and nonrespiratory causes of deaths.
18 In summary, these studies examining specific respiratory or cardiovascular causes of death
19 often found risk estimates that were higher than those for the total or broader death cause
20 categories, but their lower statistical power in the smaller subcategories often made it difficult to
21 distinguish the contrasts in estimates.
22
23 7.4.7 O3-Mortality Risk Estimates for Specific Subpopulations
24 Some studies examined O3-mortality risk estimates in potentially susceptible
25 subpopulations, such as those with underlying cardiopulmonary disease. Sunyer et al. (2002)
26 examined the associations between air pollution and deaths in a cohort of patients (467 men and
27 611 women) with severe asthma in Barcelona, Spain during the period of 1986 to 1995. A case-
28 crossover study design was used to estimate excess odds of mortality adjusting for weather and
29 epidemics in three groups: (1) those who had only one asthma emergency department visit;
30 (2) those who had more than one asthma emergency department visit; and (3) those who had
31 more than one asthma and COPD emergency department visit. Those with more than one
January 2005 7-72 DRAFT-DO NOT QUOTE OR CITE
-------
1 asthma emergency department visit showed the strongest associations with the examined air
2 pollutants, with NO2 being the most significant predictor, followed by O3. Sunyer et al. (2002)
3 reported a significant association between O3 and all cause deaths for this group during the warm
4 season, with an odds ratio of 1.90 (95% CI: 1.09, 3.30) per 48 |ig/m3 increase in 1-h max O3,
5 compared to an odds ratio of 1.02 (95% CI: 0.73, 1.43) for those with only one asthma
6 emergency department visit and 1.05 (95% CI: 0.73, 1.50) for the group with a concomitant
7 diagnosis of COPD. The magnitude of the effect size estimates reported for patients with more
8 than one asthma emergency department visit was large compared to the total mortality risk
9 estimate (relative risk of 1.03 per 48 |ig/m3 increase in 1-h max O3) observed in the related study
10 by Sunyer et al. (1996). In another Barcelona study, Saez et al. (1999) examined asthma
11 mortality death among persons aged 2 to 45 years. Once again, O3 and NO2 were the only air
12 pollutants that were significantly associated with asthma mortality death. While the similarity of
13 the patterns of associations between O3 and NO2 makes it difficult to speculate on the specific
14 causal role of O3, the results of these studies suggest that individuals with severe asthma may
15 make up a subpopulation that is sensitive to these pollutants.
16 Sunyer and Basagna (2001) also performed an analysis of emergency department visits by
17 a cohort with COPD. The results from this study suggested that PM10, but not gases were
18 associated with mortality risks for the COPD cohort. However, a Mexico City study by Tellez-
19 Rojo et al. (2000) observed a significant association between COPD mortality and O3, along with
20 PM10, among patients living outside a medical unit. For a cumulative 5-day lag, an excess risk of
21 15.6% (95% CI: 4.0, 28.4) per 1-h max O3 was observed for COPD mortality.
22 Goldberg et al. (2003) investigated the association between air pollution and daily
23 mortality with congestive heart failure as the underlying cause of death in patients aged 65 years
24 or more in Montreal, Quebec, Canada during the period of 1984 to 1993. Analysis was stratified
25 into two groups, those whose underlying cause of death was congestive heart failure and those
26 with a diagnosis of congestive heart failure one year before their death. They found no
27 association between daily mortality for congestive heart failure and any pollutants. However,
28 they did find significant associations between daily mortality among those who were classified
29 as having congestive heart failure before death and coefficient of haze, SO2, and NO2. Ozone
30 was not significantly associated but showed positive risk estimates for year-round and warm
January 2005 7-73 DRAFT-DO NOT QUOTE OR CITE
-------
1 season data and a negative risk estimate for cool season data. While the 10-year study period for
2 this data was long, the daily mean death counts for the specific subcategory chosen was
3 relatively small (0.7/day for mortality with congestive heart failure as underlying cause of death
4 and 4.0/day for total mortality in patients previously diagnosed with congestive heart failure),
5 limiting the power of the study.
6 Few studies have examined O3-mortality effects for specific subpopulations. Among those
7 that investigated the effect of air pollution in populations with underlying cardiopulmonary
8 diseases, associations were not unique to O3 but were shared with other pollutants. The results
9 from Spain (Saez et al., 1999; Sunyer et al., 2002) suggest that severe asthmatics may be
10 susceptible to the mortality effects associated with NO2 and O3.
11
12 7.4.8 Summary of Acute O3 Effects on Mortality
13 • A substantial body of new data on acute mortality effects of O3 has emerged since the
previous O3 AQCD. While uncertainties remain in some areas, it can be concluded
that robust associations have been identified between various measures of daily O3
concentrations and increased risk of mortality. The fairly small but consistent
associations cannot be readily explained by confounding due to time, weather,
nor copollutants.
14 • The majority of the available O3-mortality risk estimates were computed using all
year data. The results from the studies that conducted analysis by season suggest that
the O3 risk estimates were larger in the warm season. Some of the risk estimates in
the cool season were negative, possibly reflecting the negative correlation between
low-level O3 and PM (and other primary pollutants) during that season. Thus,
without adequate adjustment for temporal trends, the O3 risk estimates obtained for
year-round data may be misleading and likely underestimate the effects during the
warm season.
15 • Some studies examined specific subcategories of mortality, but most of these studies
had limited statistical power to detect associations due to the small daily mortality
counts. The large U.S. 95 communities study indicated that there was a slightly
greater risk of cardiopulmonary mortality compared to total mortality.
16 • Few studies examined the effect of O3 on mortality in subpopulations with underlying
cardiopulmonary diseases. Similar to cause-specific mortality, these population-
specific studies had limited statistical power to detect associations. The evidence
suggests that individuals with severe asthma may be at increased risk of O3-related
mortality, however, similar results were seen with other pollutants.
17
January 2005 7-74 DRAFT-DO NOT QUOTE OR CITE
-------
1 7.5 CHRONIC EFFECTS OF OZONE
2 7.5.1 Summary of Key Findings on Studies of Health Effects and Chronic
3 O3 Exposure from the 1996 O3 AQCD
4 The 1996 O3 AQCD concluded that there was insufficient evidence from the limited
5 number of studies to determine whether long-term ambient O3 exposures resulted in chronic
6 health effects. However, the aggregate evidence suggested that chronic O3 exposure, along with
7 other environmental factors, could be responsible for health effects in exposed populations.
8
9 7.5.2 Introduction to Morbidity Effects of Chronic O3 Exposure
10 Several new longitudinal epidemiologic investigations have yielded information on health
11 effects of long-term O3 exposures. Epidemiologic interest in investigating long-term effects has
12 been motivated by several considerations. Animal toxicology studies carried out from the late
13 1980s onward demonstrated that long-term exposures can result in permanent changes in the
14 small airways of the lung, including remodeling of the airway architecture (specifically the distal
15 airways and centriacinar region) and deposition of collagen, as discussed earlier in Chapter 5.
16 These changes result from the damage and repair processes that occur with repeated exposure.
17 Indices of fibrosis also were found to persist after exposure in some of the studies. Collectively,
18 these findings provide a potential pathophysiologic basis for the changes in airway function
19 observed in children in longitudinal studies. Seasonal ambient patterns of exposure may be of
20 greater concern than continuous daily exposure. In the classical study by Tyler et al. (1988),
21 seasonal exposure was associated with greater increases in total lung collagen and pulmonary
22 function changes suggestive of a delay in lung maturation in animals.
23 Controlled human exposure studies clearly demonstrated acute inflammation in the lung at
24 ambient exposure levels. Epidemiologic studies could examine whether repeated exposures over
25 multiple episode periods and/or multiple years would lead to persistent inflammation and result
26 in damage to the human lung, especially in the small, terminal bronchiolar regions where
27 vulnerability is greatest. However, the challenges to addressing these issues in epidemiology
28 studies are formidable, and as a result there exists relatively limited literature in this area. Long-
29 term O3 concentrations tend to be correlated with long-term concentrations of other pollutants,
30 making specific attribution difficult. Subtle pulmonary effects require health outcome measures
31 that are sensitive, and must usually be directly collected from individual human subjects, rather
January 2005 7-75 DRAFT-DO NOT QUOTE OR CITE
-------
1 than from administrative data bases. Although these factors make chronic studies difficult and
2 expensive to conduct, efforts must be made to design studies with adequate power to examine
3 the hypothesis being tested. Epidemiologic studies are the only approach to investigate a
4 possible link between chronic exposure to ozone and the occurrence of human health effects.
5 Here we review studies published from 1996 onward in which health effects were tested in
6 relation to O3 exposures extending from several weeks to many years (Table AX7-6 in the
7 Chapter 7 Annex). The available literature falls into four general categories: (1) studies
8 examining seasonal changes in lung function and lung function growth as related to O3
9 exposures in peak season; (2) studies addressing lung function growth or decline of lung
10 function over several years in relation to long-term O3 exposures; (3) studies addressing
11 respiratory inflammation in high versus low exposure groups or time periods; and (4) studies
12 addressing longitudinal and cross-sectional associations between long-term O3 exposures and
13 asthma development and prevalence.
14
15 7.5.3 Seasonal O3 Effects on Lung Function
16 While it has been well-documented in both chamber and field studies that daily, multihour
17 exposures to O3 result in transient declines in lung function, much less is known about the effects
18 of repeated exposures to O3 over extended periods on lung function. Several new studies
19 reported over the past decade have examined lung function changes over seasonal time periods
20 with differing levels of O3 exposures (Frischer et al., 1999; Horak et al., 2002a,b; Ihorst et al.,
21 2004; Kinney and Lippmann, 2000; Kopp et al., 2000). These seasonal effects of O3 are
22 examined first in this section. In the next section is a discussion of effects over years, as
23 opposed to over seasons, in addition to multiyear analyses of seasonal studies.
24 In a large study, Frischer and colleagues collected repeated lung function measurements in
25 1,150 Austrian school children (mean age 7.8 years) from nine towns that differed in mean
26 O3 levels. Mean summertime O3 exposure ranged from 32.4 to 37.3 ppb during the three
27 summers. Lung function was measured in the spring and fall over a three-year period from 1994
28 to 1996, yielding six measurements per child. The seasonal change in lung function was
29 significantly and inversely associated with seasonal mean O3 levels. FEVj increase was lower
30 by 156.6 mL (-0.029 mL/day/ppb x 90 days/year x 3 years x 20 ppb) for each 20 ppb increase
31 in mean 24-h avg O3 concentrations over the three summers and 129.6 mL over the three
January 2005 7-76 DRAFT-DO NOT QUOTE OR CITE
-------
1 winters. When analyses were restricted to children who had spent the whole summer period in
2 their community, the changes were greater, with a greater O3-related reduction of 183.6 mL in
3 FEVj growth. Other pollutants (PM10, SO2, and NO2) had less consistent associations with lung
4 function growth. Horak et al. (2002a,b) extended the study of Frischer et al. (1999) with an
5 additional year of data and indicated that seasonal mean O3 was associated with a negative effect
6 on lung function growth, confirming results from the previous three-year study. In an editorial,
7 Tager (1999) stated that the Frischer et al. (1999) data provided the first prospective evidence of
8 an association between exposure to ambient air pollution and alterations in lung growth in
9 children. Tager further noted that the prospective study design represented a substantial
10 improvement over data derived from cross-sectional studies and should be emulated.
11 Kopp et al. (2000), in a cohort of 797 children in Austria and southwestern Germany,
12 reported lower lung function growth in children exposed to high (44 to 52 ppb O3) levels of
13 ambient O3. Children residing in low O3 (24 to 33 ppb) areas experienced a 43 mL increase
14 in FEVj whereas those in high O3 areas only experienced a 16 mL increase during the summer of
15 1994. Similar results were found in data from the summer of 1995. In another Austrian study,
16 Ihorst et al. (2004) examined 2,153 children with a median age of 7.6 years and reported summer
17 pulmonary function results revealing a significantly lower FVC and FEVj increase associated
18 with higher O3 exposures in the summer, but not in the winter.
19 In a pilot study (Kinney and Lippmann, 2000), 72 nonsmoking adults (mean age 20 years)
20 from the 2nd year class of students at the U.S. Military Academy at West Point, NY provided
21 two lung function measurements, one before and one after a five-week long summer training
22 program at four locations. There was a greater decline in FEVj among students at the Fort Dix
23 location (78 mL) as compared to students at the other locations (31 mL). Ozone levels at Fort
24 Dix averaged 71 ppb (mean of daily 1-h max O3) over the summer training period versus mean
25 values of 55 to 62 ppb at the other three locations. In addition to the higher mean O3 level, Fort
26 Dix had greater peak O3 values (23 hours > 100 ppb) compared to the other locations (1 hour
27 > 100 ppb). Ambient levels of other pollutants, PM10 and SO2, were relatively low during the
28 study and did not vary across the four sites. Though conclusions are limited by the small size of
29 the study, results are consistent with a seasonal decline in lung function that may be due, in part,
30 to O3 exposures. Another interesting observation from this study was that a larger decline was
31 observed in subjects with post-summer measurements in the first two weeks after returning from
January 2005 7-77 DRAFT-DO NOT QUOTE OR CITE
-------
1 training compared to those measured in the 3rd and 4th weeks, indicating that O3-related lung
2 function declines might be reversible.
3 Collectively, the above studies indicate that seasonal O3 exposure is associated with
4 declines in lung function growth in children. The study by Kinney and Lippman (2000) provide
5 limited evidence that seasonal O3 also may affect lung function in adults, though the effect may
6 be somewhat transient.
7
8 7.5.4 Chronic O3 Effects on Lung Function
9 Lung capacity grows during childhood and adolescence as body size increases, reaches
10 a maximum during the 20s, and then begins to decline steadily and progressively with age.
11 There has long been concern that long-term exposure to air pollution might lead to slower
12 growth in lung capacity, diminished maximally attained capacity, and/or more rapid decline
13 in capacity with age. The concern arises by analogy with cigarette smoking, where it is well-
14 documented that lung function declines more rapidly with age in a dose-dependent manner
15 among adults who smoke cigarettes. Adults who stop smoking return to a normal rate of decline
16 in capacity, although there is no evidence that they regain the capacity previously lost due to
17 smoking (Burchfiel et al., 1995). Because O3 is a strong respiratory irritant, and is associated
18 with acute lung function declines as well as inflammation and re-structuring of the respiratory
19 airways, it seems plausible that there might be a negative impact of long-term O3 exposures on
20 lung function. Exposures that affect lung function growth during childhood, in particular, might
21 cause greater long-term risks. Thus, studies of effects on diminished rate of lung function
22 growth in children are especially important.
23 Several studies published over the past decade have examined the relationship between
24 lung function and long-term O3 exposure. The most extensive and robust recent study of
25 respiratory effects in relation to long-term air pollution exposures among children has been the
26 Children's Health Study carried out in 12 communities of southern California starting in 1993
27 (Peters et al., 1999a,b; Gauderman et al., 2000, 2002, 2004a,b). No significant associations were
28 observed between long-term O3 exposures and self-reports of respiratory symptoms or asthma
29 (Peters et al., 1999a). In the initial report examining the relationship between lung function at
30 enrollment and levels of air pollution in the community, there was evidence that annual mean O3
31 levels were associated with decreased FVC, FEVj, PEF, and FEF25.75 (the latter two being
January 2005 7-78 DRAFT-DO NOT QUOTE OR CITE
-------
1 statistically significant) among females but not males (Peters et al., 1999b). Among the 4th
2 graders, a longitudinal analysis of lung function growth over eight years indicated decrements
3 were associated significantly with PM and NO2, but not with O3 (Guaderman et al., 2000,
4 2004a,b). A 2nd cohort of 4th graders were recruited in 1996 and followed over four years
5 (Gauderman et al., 2002). Stratified analyses by time spent outdoors indicated a
6 significant association between decreased PEF growth and O3 exposure only in children who
7 spent more than 1.3 hours outdoors (Guaderman et al., 2002).
8 Ihorst et al. (2004) found that there were no associations between lung function growth rate
9 and mean summer O3 levels for FVC and FEVj over a 3.5-year period, in contrast to the
10 significant seasonal effects discussed earlier. Unlike the smaller increase in lung function
11 parameters over the 1st two summers among children in high O3 areas, a greater increase was
12 observed during the 3rd summer and no difference in increase was observed during the 4th
13 summer. The authors then concluded that medium-term effects on schoolchildren lung growth
14 are possibly present but are not detected over a 3- to 5-year period due to partial reversibility.
15 The study by Frischer et al. (1999) showed results similar to the Ihorst et al. (2004) study.
16 Although a significant O3-related reduction in lung function growth was observed when three
17 years were analyzed collectively, smaller changes were observed throughout the years. FEVj
18 increase was significantly lower by 34.0 mL for each 20 ppb increase in mean 24-h avg O3 in the
19 1st year compared to a nonsignificant but greater increase of 7.3 mL in the 3rd year (Frischer
20 et al., 1999). Results from Horak et al. (2002a) indicated that the four-year cumulative reduction
21 in FEVj was 151.2 mL with O3 levels of 20 ppb, which was less than the cumulative estimate of
22 156.6 mL from the 1 st three years, indicating that there was little if any O3-related changes in
23 lung function growth during the 4th year.
24 Evidence for a relationship between long-term O3 exposures and decrements in maximally
25 attained lung function was observed in a nationwide cohort of 520 1st year students at Yale
26 College in New Haven, CT (Galizia and Kinney 1999; Kinney et al., 1998). Each student
27 performed one lung function test in the spring of their 1st year at college. Ozone exposures were
28 estimated by linking 10-year mean summer season 1-h max O3 levels at the nearest monitoring
29 station to the residential locations reported each year from birth to the time of measurement.
30 Students who had lived four or more years in areas with long-term mean O3 levels above 80 ppb
31 had significantly lower FEVj (-3.07% [95% CI: -0.22, -5.92]) and FEF25.75 (-8.11% [95% CI:
January 2005 7-79 DRAFT-DO NOT QUOTE OR CITE
-------
1 -2.32, -13.90]) compared to their classmates with lower long-term O3 exposures. Stratification
2 by gender indicated that males had much larger effect estimates than females, which might
3 reflect higher outdoor activity levels and corresponding higher O3 exposure during childhood.
4 A similar study of 130 1st year college freshmen at the University of California at Berkeley
5 also reported significant effects of O3 on lung function (Kiinzli et al., 1997; Tager et al., 1998).
6 Enrollment was limited to students from either the San Francisco or Los Angeles, CA
7 metropolitan areas. After controlling for city of origin, long-term O3 exposures were associated
8 with declines in FEF25.75 and FEF75 (forced expiratory flow after 75% of FVC has been exhaled).
9 No effects were seen for PM10 and NO2. Kiinzli and colleagues noted that significant changes in
10 these mid- and end-expiratory flow measures could be considered early indicators for pathologic
11 changes that might ultimately progress to COPD, as evidenced by animal studies that show that
12 the primary site of O3 injury in the lung is the centriacinar region (Chapter 5). In another
13 California-based study (Gong et al., 1998b), there was no relationship between long-term
14 changes in lung function (over an approximately 10-year period) and acute responsiveness to O3
15 exposure (over a two-hour period in a controlled chamber environment) among persons living in
16 high O3 communities.
17 An autopsy pathologic study examining centriacinar region inflammatory disease was part
18 of a discussion of long-term O3 effects in the animal toxicology studies in Chapter 5. Sherwin
19 et al. (2000) examined subjects for the above pathologic outcome in Miami, FL and Los
20 Angeles, CA residents. A trend towards greater degrees of centriacinar region alterations was
21 observed in the lungs of Los Angeles residents compared to Miami residents, independent of a
22 smoking effect. The results suggest that the greater extent and severity of centriacinar region
23 alterations might be related to the higher O3 levels in Los Angeles. Beyond the challenge of
24 differentiating the lifetime of exposure for subjects in the two cities, various confounding factors
25 also can impact this study. The pathogenesis of centriacinar region alteration is undoubtedly
26 multifactorial with respiratory infection and adverse environmental influences being two major
27 considerations. In addition, Sherwin et al. (2000) noted that the study was limited due to the
28 relatively small number of cases available. Nonetheless, as observed by Tager (1993), the use of
29 human postmortem specimens is of interest in future epidemiology studies.
30
31
January 2005 7-80 DRAFT-DO NOT QUOTE OR CITE
-------
1 7.5.5 Chronic O3 Exposure and Respiratory Inflammation
2 As noted in Chapter 6, human chamber studies have demonstrated that brief (2 to
3 6.6 hours) exposures to O3 while exercising result in inflammation in the lung, including the
4 alveolar region where gas exchange takes place. This acute effect is potentially important for
5 chronic effects because repeated inflammation can result in the release of substances from
6 inflammatory cells that can damage the sensitive cells lining the lung. Over extended periods,
7 repeated insults of this kind could lead to permanent damage to and re-structuring of the small
8 airways and alveoli. In addition, since inflammation is a fundamental feature of asthma, there is
9 concern that O3-induced inflammation can exacerbate existing asthma or perhaps promote the
10 development of asthma among genetically pre-disposed individuals. Several studies are
11 discussed next, examining different outcomes related to inflammation.
12 In a study by Kinney et al. (1996), bronchoalveolar lavage fluids were collected in the
13 summer and winter from a group of 19 adult joggers living and working on an island in
14 New York harbor. The mean 1-h max O3 for the three-month periods were 58 ppb in the
15 summer and 32 ppb in the winter. PM10 and NO2 concentrations were similar across the two
16 seasons. There was little evidence for acute inflammation in bronchoalveolar lavage fluids
17 collected during the summer as compared to that collected from the same subjects in the winter.
18 However, there was evidence of enhanced cell damage, as measured by lactate dehydrogenase,
19 in the summer lavage fluids. These results indicate that acute inflammatory responses may
20 diminish with repeated exposures over the course of a summer (which have been demonstrated
21 in multiday chamber exposures, Chapter 6, Section 6.9) but cell damage may be ongoing.
22 Pollution effects in the nose can be viewed as a potential surrogate measure for effects that
23 may occur in the lungs, though doses to nasal tissues are usually higher for a given pollutant
24 concentration. In Chapter 5, morphological effects of O3 on the upper respiratory tract indicated
25 quantitative changes in the nasal transitional respiratory epithelium. The persistent nature of the
26 O3-induced mucous cell metaplasia in rats, as discussed in Chapter 5, suggests that O3 exposure
27 may have the potential to induce similar long-lasting alterations in the airways of humans.
28 A series of interesting studies in Mexico City have demonstrated inflammation and genetic
29 damage to cells in the nasal passages of children chronically exposed to O3 and other air
30 pollutants (Calderon-Garciduefias et al., 1995, 1997, 1999). Nasal lavage samples and nasal
31 biopsies from children living in Mexico City were compared to those from children living in a
January 2005 7-81 DRAFT-DO NOT QUOTE OR CITE
-------
1 clean coastal town with no detectable air pollutants. In the first study, urban children (n = 38)
2 from Mexico City were found to have significantly higher polymorphonuclear leukocyte counts
3 and abnormal nasal cytologies compared to nonurban children (n = 28) (Calderon-Garciduefias
4 et al., 1995). A later study observed that cells collected from the lining of the nose had
5 significantly higher amounts of DNA damage in the urban children in Mexico City (n = 129)
6 versus nonurban children (n = 19) (Calderon-Garciduefias et al., 1997). Among exposed
7 children, DNA damage was greater with increasing age, suggesting an accumulation of damage
8 over time with ongoing pollution exposures. Another study of 86 urban and 12 nonurban
9 children reported similar findings, in addition to increased levels of specific DNA mutations
10 (Calderon-Garciduefias et al., 1999). They also noted far higher respiratory symptom prevalence
11 in the urban children. Fortoul et al. (2003) examined DNA strand breaks in nasal epithelial cells
12 from asthmatic and nonasthmatic medical students in Mexico City and noted greater genotoxic
13 damage in asthmatics. These results indicate that asthmatics may have a greater vulnerability for
14 DNA damage, or a decreased ability to repair it, compared to nonasthmatic subjects.
15 Another outcome of inflammation was examined in a study by Frischer et al. (2001).
16 In this cross-sectional study, urinary eosinophil protein was analyzed as a marker of eosinophil
17 activation in 877 school children living in nine Austrian communities with varying O3 exposure.
18 The results indicated that O3 exposure was significantly associated with eosinophil
19 inflammation.
20 In the Mexico City studies, specific attribution of these adverse respiratory and genotoxic
21 effects to O3 is difficult given the complex mixture of pollutants present in the ambient air.
22 In particular, the DNA effects seem more plausibly related to other components of urban air,
23 such as semi-volatile organic compounds. However, the inflammatory changes such as
24 increased eosinophil levels observed in the Austrian study would be consistent with known
25 effects of O3.
26
27 7.5.6 Risk of Asthma Development
28 Recent reports from longitudinal cohort studies have reported associations between
29 the onset of asthma and long-term O3 exposures (McConnell et al., 2002; McDonnell et al.,
30 1999). Significant associations between new cases of asthma among adult males and long-term
31 O3 exposure were observed in a cohort of nonsmoking adults in California (Greer et al., 1993;
January 2005 7-82 DRAFT-DO NOT QUOTE OR CITE
-------
1 McDonnell et al., 1999). The Advent!st Health and Smog (AHSMOG) study cohort of 3,914
2 (age 27-87 years, 36% male) was drawn from nonsmoking, non-Hispanic white California
3 seventh day Adventists. Subjects were surveyed in 1977, 1987, and 1992. To be eligible,
4 subjects had to have lived 10 or more years within 5 miles of their current residence in 1977.
5 Residences from 1977 onward were followed and linked in time and space to interpolated
6 concentrations of O3, PM10, SO2, and NO2. New asthma cases were defined as self-reported
7 doctor-diagnosed asthma at either the 1987 or 1992 follow-up questionnaire among those who
8 had not reported having asthma upon enrollment in 1977. During the 10 year follow-up (1977-
9 1987), the incidence of new asthma was 2.1% for males and 2.2% for females (Greer et al.,
10 1993). A relative risk of 3.12 per 10 ppb increase in annual mean O3 was observed in males,
11 compared to a non-significant relative risk of 0.94 in females. In the 15-year follow-up study
12 (1977-1992), 3.2% of the eligible males and a slightly greater 4.3% of the eligible females
13 developed adult asthma (McDonnell et al., 1999). For males, the relative risk of developing
14 asthma was 2.27 per 30 ppb increase in 8-h avg O3 (9 a.m.-5 p.m.). Once again, there was no
15 evidence of an association between O3 and new-onset asthma in females (relative risk of 0.85).
16 The lack of an association does not necessarily indicate no effect of O3 on the development of
17 asthma among females. For example, differences in time-activity patterns in females and males
18 may influence relative exposures to O3, leading to greater misclassification of exposure in
19 females. The consistency of the results in the two studies with different follow-up times and
20 indices of O3 exposure provides evidence that long-term O3 exposure may be associated with
21 asthma incidence in adult males. However, as the AHSMOG cohort was drawn from a narrow
22 subject definition, the representativeness of this cohort to the general U.S. population may be
23 limited.
24 A similar study of incident asthma cases in relation to O3 among children was carried out
25 in the Children's Health Study (McConnell et al., 2002). Annual surveys of 3,535 initially
26 nonasthmatic children (ages 9 to 16 years at enrollment) enabled identification of new-onset
27 asthma cases through 1998. Communities were stratified by pollution levels, with six high-O3
28 communities (mean 1-h max O3 of 75.4 ppb over four years) and six low-O3 communities
29 (50.1 ppb). Asthma risk was not higher for residents of the six high-O3 communities versus
30 residents of the six low-O3 communities. However, within the high-O3 communities, asthma risk
31 was 3.3 times greater for children who played three or more sports as compared with children
January 2005 7-83 DRAFT-DO NOT QUOTE OR CITE
-------
1 who played no sports. This association was absent in the low-O3 communities (relative risk of
2 0.8). No associations with asthma were seen for PM10, PM2 5, NO2, or inorganic acid vapors.
3 These results suggest effect modification of the impacts of O3 on asthma risk by physical
4 activity. Playing sports may indicate outdoor activity when O3 levels are higher and an
5 increased ventilation rate, which may lead to increased O3 exposure. Replication of these
6 findings in other cohorts would lend greater weight to a causal interpretation.
7 Recent cross-sectional surveys have detected no associations between long-term O3
8 exposures and asthma prevalence, asthma-related symptoms, or allergy to common aeroallergens
9 in children after controlling for covariates (Charpin et al., 1999; Kuo et al., 2002; Ramadour
10 et al., 2000). However, reported O3 levels were quite low in all cases, with a range of 16 to
11 27 ppb for 8-h max O3. In addition, compared to the longitudinal study design, which observes
12 new onset of asthma prospectively, the cross-sectional study design is inherently weaker.
13 Longitudinal studies provide the strongest evidence on the question of asthma development and
14 is the preferred approach for future research.
15
16 7.5.7 Respiratory Effects of Chronic O3 Exposure on Susceptible
17 Populations
18 Studies on the effect of long-term O3 exposure on respiratory health has mostly focused on
19 potentially susceptible populations, including children and individuals who exercise outdoors, as
20 discussed in this section. Ozone exposure was associated significantly with declines in lung
21 function or reduced lung function growth, and respiratory inflammation in these susceptible
22 populations.
23 Other studies have investigated additional symptoms and groups of potentially susceptible
24 individuals. McConnell et al. (1999) examined the association between O3 levels and the
25 prevalence of chronic lower respiratory tract symptoms in southern California children with
26 asthma (n = 3,676). In this questionnaire-based study, bronchitis, phlegm, and cough were not
27 associated with annual mean O3 concentrations in children with asthma or wheeze. All other
28 pollutants examined, PM10, PM2 5, NO2, and gaseous acid, was associated with an increase in
29 phlegm, but not cough.
30 One new study examined a susceptible group not examined before. Goss et al. (2004)
31 investigated the effect of O3 on pulmonary exacerbations and lung function in individuals with
January 2005 7-84 DRAFT-DO NOT QUOTE OR CITE
-------
1 cystic fibrosis over the age of 6 years (n = 11,484). The study included patients enrolled in the
2 Cystic Fibrosis Foundation National Patient Registry. The registry contained demographic and
3 clinical data collected annually at accredited centers for cystic fibrosis. In 1999 and 2000, the
4 annual mean O3 concentration from 616 monitors in the U.S. EPA Aerometric Information
5 Retrieval System (AIRS) was 51.0 ppb (SD 7.3). Exposure was assessed by linking air pollution
6 values from AIRS with the patient's home zip code. No clear association was found between
7 annual mean O3 and lung function parameters. However, a 10 ppb increase in annual mean O3
8 was associated with a 10% (95% CI: 3, 17) increase in the odds of two or more pulmonary
9 exacerbations. Significant excess odds of pulmonary exacerbations also were observed with
10 increased annual mean PM10 and PM2 5 concentrations.
11 In summary, several studies have identified and investigated potentially susceptible
12 populations. Although effects are not specific to O3 exposure, the results suggest that O3 likely
13 contributes to the adverse respiratory health responses observed in these populations.
14
15 7.5.8 Mortality Effects of Chronic O3 Exposure
16 There is inconsistent and inconclusive evidence for a relationship between long-term O3
17 exposure and increased mortality risk (Table AX7-7 in the Chapter 7 Annex). A long-term
18 prospective cohort study (AHSMOG; 1977-1992) of 6,338 nonsmoking, non-Hispanic white
19 subjects living in California found a significant association between long-term O3 exposure and
20 increased lung cancer risk among males only (Beeson et al., 1998). The relative risk for lung
21 cancer incident among males was 3.56 (95% CI: 1.35, 9.42) per 556 hours/year when O3 levels
22 exceeded 100 ppb (Beeson et al., 1998). A stronger association was observed in males who
23 never smoked (4.48 [95% CI: 1.25, 16.04]) compared to those who smoked in the past (2.15
24 [95% CI: 0.42, 10.89]) (Beeson et al., 1998). An expanded study by Abbey et al. (1999)
25 examining mortality effects of long-term O3 exposure in the same study population confirmed
26 the results of the previous study by Beeson and colleagues. The association between lung cancer
27 mortality and chronic O3 exposure was significant in males only, with a relative risk of 4.19
28 (95% CI: 1.81, 9.69) (Abbey et al., 1999). However, the very small numbers of lung cancer
29 deaths (12 for males and 18 for females) raise concerns in regards to the precision of the effect
30 estimate, as evidenced by the wide confidence intervals. No other mortality outcomes were
31 found to be associated with chronic O3 exposure. A particular strength of this study was the
January 2005 7-85 DRAFT-DO NOT QUOTE OR CITE
-------
1 extensive effort devoted to assessing long-term air pollution exposures, including interpolation
2 to residential and work locations from monitoring sites over time and space. However, the
3 observation of a lung cancer effect but no effect on cardiopulmonary mortality raises concerns.
4 The gender-specific O3 effects may be partially attributable to the differences in activity and time
5 spent outdoors by gender. The questionnaires indicated that males spent approximately twice as
6 much time outdoors and performed more vigorous outdoor exercises, especially during the
7 summer, compared to the females.
8 Lipfert et al. (2000b, 2003) reported positive effects on mortality for peak O3 exposures
9 (95th percentile levels) in the U.S. Veterans Cohort study of approximately 50,000 male middle-
10 aged men recruited with a diagnosis of hypertension. The actual analysis involved smaller
11 subcohorts based on exposure and mortality follow-up periods. Four separate exposure periods
12 were defined as follows: 1960-1974; 1975-1981; 1982-1988; and 1989-1996. Three mortality
13 follow-up periods were considered: 1976-1981; 1982-1988; and 1989-1996. In a preliminary
14 screening of regression results, Lipfert et al. (2000b) compared univariate and multivariate
15 models by mean and peak (95th percentile) O3 concentrations. For mean O3, a significant
16 negative relationship was reported in the univariate model and a nonsignificant negative
17 relationship was found in the multivariate model. For peak O3 concentration, the univariate
18 model indicated a nonsignificant positive relationship and the multivariate model resulted in a
19 significant positive relationship. Peak O3 was used in subsequent analyses. The mean of the
20 peak values ranged from 85 to 140 ppb over the four exposure periods. For concurrent exposure
21 periods, the mortality risk was significant, with a 6.1% excess risk per mean 95th percentile O3
22 less estimated background level (not stated). When exposure periods preceding death were
23 considered, the excess mortality risk was nonsignificant (-0.2%). In a further analysis, Lipfert
24 et al. (2003) reported the strongest positive association for concurrent exposure to peak O3 for
25 the subset with low diastolic blood pressure during the period of 1982-1988. Once again, the O3
26 effect was found to be reduced when the exposure (1982-1988) preceded mortality (1989-1996).
27 No effect of long-term O3 concentrations on mortality risk was observed in a larger
28 prospective cohort study of approximately 500,000 U.S. adults (Pope et al., 2002). Strong and
29 consistent effects of PM2 5 were observed in this study for both lung cancer and cardiopulmonary
30 mortality.
31
January 2005 7-86 DRAFT-DO NOT QUOTE OR CITE
-------
1 7.5.9 Summary of Chronic O3 Effects on Morbidity and Mortality
2 • In the past decade, important new longitudinal studies have examined the chronic
effect of O3 exposure on respiratory health outcomes. Evidence from recent long-term
morbidity studies have indicated that chronic exposure to O3 may be associated with
declines in lung function, inflammation, and development of asthma in children and
adults. Seasonal decrements or reduced growth in lung function measures have been
reported in several studies, however changes appear to be transient. Studies of lung
function declines with longer-term or annual data are not as conclusive.
3 • Few studies have investigated the effect of long-term O3 exposure on mortality.
Uncertainties regarding the exposure period of relevance, and inconsistencies across
mortality outcomes and gender raise concerns regarding plausibility. The current
evidence is inconclusive for a relationship between chronic O3 exposure and increased
risk of mortality.
4
5
6 7.6 INTERPRETATIVE ASSESSMENT OF THE EVIDENCE IN
7 EPIDEMIOLOGIC STUDIES OF OZONE HEALTH EFFECTS
8 7.6.1 Introduction
9 In the 1996 O3 AQCD, the epidemiology section focused primarily on individual-level
10 camp and exercise studies. These field studies indicated exposure-response relationships
11 between O3 exposure from the ambient air, and declines in pulmonary function, increases in
12 respiratory symptoms, and exacerbation of asthma, especially in children. Numerous new
13 studies provide additional evidence for evaluating associations between O3 exposure and the
14 above respiratory health outcomes. The 1996 O3 AQCD review of aggregate population time-
15 series studies suggested an association between ambient O3 concentrations and increased
16 hospitalizations. Limited studies examined the O3-mortality relationship. The current O3 AQCD
17 further presents results from time-series studies that have addressed previously unresolved
18 issues regarding potential linkages between ambient O3 concentrations and health outcomes,
19 particularly mortality. Daily time-series studies minimize confounding by population
20 characteristics (e.g., cigarette smoking, diet, occupation) by following the same population from
21 day to day. However, confounders operating over shorter time scales can affect O3 risk estimates
22 in time-series studies.
January 2005 7-87 DRAFT-DO NOT QUOTE OR CITE
-------
1 In this section, the issues and attendant uncertainties that affect the interpretation of O3
2 health effects will be discussed. The consequences of using stationary ambient monitors as an
3 estimate of personal exposure in epidemiology studies will be examined first and a discussion of
4 the temporal relationship between O3 exposure and the occurrence of health effects will follow.
5 Of particular interest are the issues arising from model specifications in time-series studies to
6 adjust for confounding by temporal factors, meteorological effects, and copollutants. The shape
7 of the concentration-response relationship and heterogeneity of O3 health effects also will be
8 discussed briefly. All of these issues are of much importance for characterizing and interpreting
9 ambient O3-health effects associations.
10
11 7.6.2 Exposure Assessment
12 Various methods have been used to assess exposure in air pollution epidemiology studies.
13 Navidi et al. (1999) describes the two methods commonly used to ascertain personal exposure:
14 (1) the microenvironmental (indirect) approach; and (2) the personal sampling (direct) approach.
15 Both methods are associated with measurement error. To determine personal exposure using the
16 microenvironmental approach, the concentrations of the various microenvironments are
17 multiplied by the time spent in each microenvironment. Both the concentration and time
18 component contribute to the measurement error. Although there is no time component to the
19 measurement error in the personal sampling approach, the estimation of exposure using personal
20 monitoring devices do contribute to the error, especially in the case of O3. The passive badges
21 commonly used for personal sampling of O3 provide integrated personal exposure information.
22 Their sensitivity to wind velocity, badge placement, and interference with other copollutants
23 may result in measurement error. Results from the error analysis models developed by Navidi
24 et al. (1999) indicated that neither the microenvironmental or personal sampling approach gave
25 reliable health effect estimates when measurement errors were uncorrected. The nondifferential
26 measurement error biased the effect estimates toward zero under the model assumptions.
27 However, if the measurement error was correlated with the health response, a bias away from the
28 null could result. The use of central ambient monitors to estimate exposure also biased the
29 estimates towards the null. Most people spend the majority of their time indoors, where O3
30 levels tend to be much lower than outdoor ambient levels. Using ambient concentrations to
January 2005 7-88 DRAFT-DO NOT QUOTE OR CITE
-------
1 determine exposure generally overestimates true personal O3 exposure, resulting in effect
2 estimates biased towards the null.
3
4 7.6.2.1 Relationship between Ambient Concentrations and Personal Exposure to O3
5 Several studies have examined the relationship between ambient O3 concentrations from a
6 central monitoring site and personal O3 exposure (Avol et al., 1998; Brauer and Brook, 1997;
7 Chang et al., 2000; Delfmo et al., 1996; Lee et al., 2004; Liard et al., 1999; Linn et al., 1996; Liu
8 et al., 1995, 1997; O'Neill et al., 2003; Sarnat et al., 2001). In a Baltimore, MD study of older
9 adults, individuals with COPD, and children, 24-h avg ambient O3 concentrations from a
10 monitoring site were not found to be significantly associated with personal O3 exposure (Sarnat
11 et al., 2001). The mixed regression effect estimates were P = 0.01 (t = 1.21) and P = 0.00
12 (t = 0.03), for summer and winter, respectively. A study by O'Neill et al. (2003), in contrast,
13 found a statistically significant association between personal and ambient O3 concentrations in
14 Mexico City outdoor workers (P = 0.56, t = 8.52). The subjects in the Sarnat et al. (2001) study
15 spent less than 6% of their time outdoors, whereas the personal exposure data from O'Neill et al.
16 (2003) were from subjects who spent the entire measurement period outdoors. A scripted
17 exposure study by Chang et al. (2000) provided supportive evidence for the conflicting results in
18 the Sarnat et al. (2001) and O'Neill et al. (2003) studies. In this study, activities were scripted to
19 simulate activities typical of older adults living in Baltimore, MD. Their activities were derived
20 from the U.S. EPA-sponsored National Human Activity Pattern Survey study (Klepeis, 1999).
21 Chang et al. (2000) compared one-hour personal and ambient O3 measurements in several
22 microenvironments. There was no correlation between personal and ambient O3 concentrations
23 in the indoor residence (r = 0.09 and r = 0.05, for summer and winter, respectively), although a
24 moderate correlation was found in other indoor environments such as restaurants, hospitals, and
25 shopping malls (r = 0.34 in summer, r = 0.46 in winter). In comparison, the correlation in
26 outdoor environments (near and away from roads) was moderate to high (0.68 < r < 0.91) and
27 statistically significant.
28 Brauer and Brook (1997) observed that the daily averaged personal O3 measurements and
29 ambient concentrations were well-correlated after stratifying groups by time spent outdoors.
30 The clinic workers (n = 25) spent 9% of their time outdoors (24-hour samples) whereas the farm
31 workers (n = 15) spent 100% of their monitored time outdoors (6-14 hour workshift samples).
January 2005 7-89 DRAFT-DO NOT QUOTE OR CITE
-------
1 The personal to ambient O3 concentration ratios were significantly different for the clinic
2 workers (0.28) and farm workers (0.96). However, the Spearman correlation coefficients were
3 similar in the two groups, 0.60 and 0.64 for the clinic workers and farm workers, respectively.
4 In a Paris, France study by Liard et al. (1999), adults (n = 55) and children (n = 39) wore passive
5 O3 monitors for 4 consecutive days during three periods. For each period, all adults wore the O3
6 monitors over the same 4 days. Likewise, all children wore monitors over the same 4 days for
7 each of the three periods, but on different days from the adults. The ambient O3 concentrations
8 from the stationary monitoring sites did not explain a high percentage of the variance of personal
9 O3 exposure (non-significant [value not stated] in adults and 21% in children). However, when
10 personal measurements from all subjects were aggregated for each of the six periods, the 4-day
11 mean personal O3 exposure was found to be highly correlated with the corresponding mean
12 ambient concentration (r = 0.83, p < 0.05). Similarly, a study of Los Angeles school children by
13 Linn et al. (1996) found that daily 24-h avg ambient O3 concentrations from a central site were
14 well-correlated (r = 0.61) with daily averaged personal O3 concentrations (8-10 children/day, 132
15 total monitoring days).
16 The low correlation observed between personal O3 exposures and ambient O3
17 concentrations in the study by Sarnat et al. (2001) suggests that O3 concentrations measured at
18 central ambient monitors do not explain the variance of individual personal exposures.
19 However, in time-series studies, daily averaged personal exposures from the aggregate
20 population is of greater relevance than exposures from specific individuals. Although
21 unresolved issues do remain, the limited evidence indicates that ambient O3 concentrations from
22 central monitors may serve as valid surrogate measures for aggregate personal O3 exposures in
23 population time-series studies investigating mortality and hospitalization outcomes.
24
25 7.6.2.2 Factors Affecting the Relationship between Ambient Concentrations and
26 Personal Exposures to O3
27 In cohort studies investigating acute and chronic morbidity outcomes, O3 exposure
28 assessment may be improved by accounting for the distance between home and the monitoring
29 site, time-activity patterns (e.g., percentage of time spent outdoors, type of outdoor activity, time
30 of day during outdoor activity), and elements that affect indoor air exchange rates (e.g.,
31 ventilation conditions, home characteristics) in conjunction with the ambient O3 data from
32 stationary monitor sites. Several studies (Geyh et al., 2000; Lee et al., 2004; Liu et al., 1995,
January 2005 7-90 DRAFT-DO NOT QUOTE OR CITE
-------
1 1997) demonstrated that the association between personal O3 exposure and ambient O3
2 concentrations is affected by these factors. A study by Geyh et al. (2000) observed higher indoor
3 and personal O3 concentrations in a southern California community with 2% air conditioned
4 homes compared to a community with 93% air conditioned homes during the summer (high O3)
5 months, but showed no difference in O3 levels during the winter (low O3) months. Lee et al.
6 (2004) observed that personal O3 exposure was positively correlated with outdoor time (r = 0.19,
7 p < 0.01) and negatively correlated with indoor time (r = -0.17, p < 0.01). Additional factors that
8 affected indoor O3 levels were air conditioning, window fans, and window opening. The O3
9 exposure assessment study by Liu et al. (1995) found that after adjusting for time spent in
10 various indoor and outdoor microenvironments (e.g., car with windows open, car with windows
11 closed, school, work, home, outdoors near home, outdoors other than near home), mean 12-hour
12 ambient O3 concentrations explained 32% of the variance in personal exposure in the summer.
13 Other factors, including age, gender, and occupation, also may affect general exposure
14 patterns to O3 by influencing time-activity patterns. In a southern California study by Avol et al.
15 (1998), boys were found to spend more time outdoors and be more physically active than girls
16 (Avol et al., 1998). Another southern California study found that boys were outdoors 30 minutes
17 longer than girls, and had higher personal O3 exposure during both high and low O3 months
18 (Geyh et al., 2000). Outdoors workers also tend to be exposed to higher levels of O3 (Brauer and
19 Brook, 1997; O'Neill et al., 2003).
20 The announcement of air quality indices also may influence personal exposures to O3.
21 Niedell (2004) examined the effect of smog alerts on the relationship between O3 and hospital
22 admissions for asthma in California children aged 0 to 18 years. Air quality episodes, or smog
23 alerts, are issued in California on days when O3 concentrations exceed 200 ppb. Smog alerts
24 were found to have a significant negative effect on asthma admissions for children aged 1 to
25 12 years, providing supportive evidence that avoidance behavior might be present on days of
26 high O3 concentrations. Avoidance behavior may include staying indoors and exercising less on
27 days when a smog alert is announced, resulting in reduced exposure to O3. Therefore air quality
28 indices may affect the relationship between ambient and personal O3 concentrations.
29 In summary, results indicate that the relationship between ambient and personal O3
30 concentrations varies depending on factors such as time spent outdoors, ventilation conditions,
31 personal factors, and air quality indices. The use of questionnaires to obtain information on
January 2005 7-91 DRAFT-DO NOT QUOTE OR CITE
-------
1 personal characteristics, time-activity patterns, and home characteristics may improve further the
2 accuracy of the personal O3 exposure estimates from ambient O3 data.
3
4 7.6.2.3 Assessing Chronic Exposure to O3
5 Several studies examined methods of estimating chronic exposure to O3. A pilot study
6 (n = 14) by Gonzales et al. (2003) indicated that the use of retrospective questionnaires might be
7 a reliable method for reconstructing past time-activity and location pattern information. The
8 7-day, 24-h avg O3 exposures were estimated using data from an ambient monitor (mean
9 29.5 ppb), and information from prospective diaries and questionnaires completed one year after
10 the monitoring period. The O3 estimates from both prospective diaries (mean 10.6 ppb) and
11 retrospective questionnaires (mean 11.8 ppb) differed only slightly, although both estimates were
12 greater than the personal exposure measurement (mean 5.7 ppb). A study by Kiinzli et al. (1997)
13 compared the retrospective assessments of outdoor time-activity patterns using three formats, a
14 questionnaire, a table, and a 24-hour log. College freshmen (n = 44) noted activity patterns at
15 their last residence using two or three methods and then were retested 5-7 days later. The
16 within-subject variance in reporting moderate to heavy activity was 13% for both the activity
17 questionnaire and activity table, and 32% for the 24-hour log. The data from the activity tables
18 also were similar to data published from the California Air Resources Board (CARB) 24-hour
19 recall diary study (Jenkins et al., 1992). Results from the above studies suggest that the use of
20 activity questionnaires or activity tables to determine time-activity patterns may be suitable in
21 large retrospective epidemiologic studies of health effects requiring estimates of chronic
22 exposure to ambient O3.
23
24 7.6.3 O3 Exposure Indices
25 Three O3 indices were used often to indicate daily O3 exposure: maximum 1-h average
26 (1-h max); maximum 8-h average (8-h max); and 24-h average (24-h avg). The 8-h max O3 is a
27 frequently used index in newer epidemiologic studies, as it best reflects the current U.S. EPA
28 standard. These O3 exposure indices are highly correlated as indicated in the following studies.
29 In the 21 European multicities acute mortality study (Gryparis et al., 2004), 1-h max O3 was
30 found to be highly correlated with 8-h max O3, with a median correlation coefficient of 0.98
31 (range: 0.91 - 0.99). Among single city studies, the 1-h max O3 and 8-h max O3 also were found
January 2005 7-92 DRAFT-DO NOT QUOTE OR CITE
-------
1 to have correlation coefficients ranging from 0.91 to 0.99 in various cities such as Atlanta, GA
2 (Tolbert et al., 2000; White et al., 1994); southern New England (Gent et al., 2003); Ontario,
3 Canada (Burnett et al., 1994); and Mexico City (Loomis et al., 1996; Romieu et al., 1995).
4 In addition, 1-h max O3 was found to be highly correlated with 24-h avg O3, as observed in the
5 Mexico City study by Loomis et al. (1996) (r = 0.77) and in the Ontario, Canada study by
6 Burnett et al. (1994) (r = 0.87).
7 All studies discussed in Sections 7.2 to 7.5 were examined for presentation of the three O3
8 exposure indices. Several presented the concentration data and correlations among 1-h max, 8-h
9 max, and 24-h avg O3 ambient measures. Some presented the associated risk estimates of
10 comparable analyses for the three exposure indices. No papers provided an analysis statistically
11 comparing the indices. Summary of the available data is provided below starting with two
12 multeity mortality studies.
13 In the large U.S. 95 communities study by Bell et al. (2004), increases in O3-associated
14 daily mortality were estimated using all three O3 indices. The increments used in this document
15 to standardize expressions of excess risks are 40 ppb for 1-h max O3, 30 ppb for 8-h max O3, and
16 20 ppb for 24-h avg O3, as discussed in Section 7.1.3.2. For these increments, the effect
17 estimates calculated by Bell et al. (2004) were 1.34%, 1.28%, and 1.04% excess risk in mortality
18 for 1-h max O3, 8-h max O3, and 24-avg O3, respectively. A statistical test examining
19 differences among these risk estimates indicated that there were no significant differences by
20 exposure index. In the European study of 21 cities (of the 23 cities, two did not have 8-h max O3
21 data), the O3-mortality effect estimate for the summer season was slightly smaller for 8-h max
22 O3, 1.82% excess risk, compared to 1-h max O3, 2.59% excess risk, but both were statistically
23 significant (Gryparis et al., 2004). Once again, a statistical test between the two risk estimates
24 yielded no significant difference between the indices.
25 Several single city mortality studies examined multiple O3 exposure indices (Anderson
26 et al., 1996; Dab et al., 1996; Sunyer et al., 2002; Zmirou et al., 1996; Borja-Aburto et al., 1997).
27 These studies did not differentiate risk estimates by exposure index as the results were
28 considered similar. Hospital admission studies also provided limited data for index
29 comparisons. Schouten et al. (1996) showed the same positive nonsignificant association
30 between total respiratory hospitalizations and O3 using both 8-h max O3 and 1-h max O3 in the
31 summer (4.0% excess risk per standardized increment). For emergency department visits, the
January 2005 7-93 DRAFT-DO NOT QUOTE OR CITE
-------
1 examples of Delfmo et al. (1998) and Weisel et al. (2002) provided data not indicative of
2 differences between the indices. A further example, Tolbert et al. (2000) noted an increase in
3 emergency room visits of 4.0% per standard deviation increase (approximately 20 ppb) for both
4 1-h max O3 and 8-h max O3 as being expected since the correlation between the indices was
5 0.99.
6 Peak flow asthma panel studies generally only used one index in these studies, thus no
7 comparison data is available. One respiratory symptom study (Gent et al., 2003) did examine
8 both 1-h max O3 and 8-h max O3 but noted no differences in the results. Only one FEVj panel
9 study examined more than one exposure index. Chen et al. (1999) examined 1-h max O3 and
10 24-h avg O3 and reported at a 1-day lag for children a decrement in FEVj of -25.6 mL (-49.1,
11 -2.1) for 1-h max O3 and -13.6 mL (-33.2, 6.0) for 24-h avg O3. For 2- and 7-day lags, smaller
12 differences were observed between the two indices. None of these FEVj outcomes, including
13 those for a 1-day lag, upon testing were significantly different by index.
14 Limited information is available to reach conclusions for comparison of the three indices
15 1-h max O3, 8-h max O3, and 24-h avg O3. In general, for the same distributional increment
16 (e.g., interquartile range), the excess health risk estimates and significance of associations were
17 comparable for all three daily O3 indices. The similarity in effects for the three exposure indices
18 was consistent for all outcomes. This is expected due to the high correlation among the indices.
19 The relationship between the various health endpoints and the three O3 exposure indices,
20 and the associated study designs and analyses are such that the high correlation among the
21 indices presents a significant challenge in distinguishing the most appropriate measure for
22 epidemiologic studies. The commonly used 8-h max O3 or 8-h avg O3 index continues to be an
23 appropriate choice as no other exposure index has been demonstrated to offer a better advantage.
24
25 7.6.4 Lag Time: Period between O3 Exposure and Observed Health Effect
26 The choice of lag days for the relationship between exposure and health effects depends on
27 the hypothesis being tested and the mechanism involved in the expression of the outcome.
28 Effects can occur acutely with exposure on the same or previous day, cumulatively over several
29 days, or after a delayed period of a few days. With knowledge of the mechanism of effect, the
30 choice of lag days can be determined prior to analysis. For example, one can expect cough to
31 occur acutely after exposure with a lag of 0 or 1 day, as O3 can act as an immediate irritant.
January 2005 7-94 DRAFT-DO NOT QUOTE OR CITE
-------
1 However, an O3-related inflammatory response may not lead to asthma exacerbation until
2 several days later. An asthmatic may be impacted by O3 on the first day of exposure, have
3 effects triggered further on the second day, then report to the emergency room for an asthmatic
4 attack three days after exposure. Further, within a population of asthmatics, exacerbation of
5 asthma symptoms may be observed over a period of several days, since each asthmatic has
6 individual aspects of the disease and may be affected by the exposure differently depending on
7 his/her sensitivity and disease severity. Therefore, it may be necessary to examine longer lag
8 periods to fully understand the relationship between O3 exposure and effect. When the
9 mechanism of the health effect is unknown, investigating the association between outcome and
10 exposure over several days also may be informative.
11 Most of the O3 time-series studies investigated a small number of lagged days, typically
12 0 through 3 days, and/or cumulative lag periods. For outcomes of mortality and hospitalizations,
13 the largest, most significant associations with O3 concentrations were observed when using short
14 lag periods, in particular a 0-day lag (exposure on same day) and a 1-day lag (exposure on
15 previous day). In the U.S. 95 communities study by Bell et al. (2004), the largest risk estimate
16 for O3-mortality was obtained with a 0-day lag, followed by diminishing risk estimates with 1-,
17 2-, and 3-day lags. In a study of 16 Canadian cities by Burnett et al. (1997a), the strongest
18 association between O3 and respiratory hospitalizations was found at a 1-day lag. Once again,
19 there was a decline in the magnitude and significance of the effect with increasing days lagged
20 for O3. These results suggest that O3 has a rapid effect on these respiratory health outcomes.
21 Less explored is the issue of multiday effects of O3. When associations are found at
22 multiple lag days, results from a single-day model may underestimate the cumulative effect of
23 O3 on health outcomes. In the large U.S. 95 communities study (Bell et al., 2004), distributed
24 lag models were used to estimate the effect of O3 levels on mortality over a lag of 0 to 6 days.
25 Results indicated that when accounting for multiple days, the effect estimates were twice as large
26 as those from single-day analyses, as shown in Figure 7-15. Bell et al. (2004) estimated a
27 cumulative excess risk of 1.04% on daily mortality per 20 ppb increase in 24-h avg O3 during the
28 previous week, compared to a 0.50% excess risk associated with O3 exposure on a single day, in
29 this case a 0-day lag. In a related U.S. study of the 19 largest cities by Huang et al. (2004), the
30 O3 estimate for the summer season was 1.50% excess risk of cardiopulmonary mortality with
31 current-day exposure and 2.52% for a 7-day cumulative lag.
January 2005 7-95 DRAFT-DO NOT QUOTE OR CITE
-------
0.5 1
% Change in Mortality
1.5
Figure 7-15. Comparison of single-day lags (0-, 1-, 2-, and 3-day) to a cumulative multiday
lag (0- to 6-day) for % changes in all cause mortality per 20 ppb increase in
24-h avg O3 in all ages.
Source: Derived from Bell et al. (2004).
1 Burnett et al. (2001) investigated the association between respiratory hospitalizations and
2 O3 in children less than 2 years of age. Lags up to five days were examined after stratifying by
3 season (Figure 7-16). In the summer season, significant associations between O3 and daily
4 admissions were found in several of the lags, with the largest risk estimate of 12.5% excess risk
5 per 40 ppb increase in 1-h max O3 at a 1-day lag. In comparison, the O3-related risk estimate
6 was 30.2% using a cumulative lag period of 5 days. In another study, Anderson et al. (1997)
7 investigated the association between O3 and daily hospital admissions for COPD in five
8 European cities. Lags up to 5 days were examined, and the largest risk estimates were found
9 using 0- and 1-day lags. Anderson et al. observed a 4.5% excess risk per 40 ppb increase in 1-h
10 max O3 using a single-day lag compared to a 7.7% excess risk using a 5-day cumulative lag.
11 Among the field studies, Mortimer et al. (2002) reported O3-related changes in PEF for
12 single-day lags from 1 to 6 days and a multiday lag period of 5 days. No associations were seen
13 between evening outcome measures and either single-day or multiday exposure lags. Small,
14 nonsignificant morning effects were observed at 1- and 2-day lags. The effect of O3 on morning
15 outcomes increased over several days, with the strongest association seen using multiday lag
January 2005
7-96
DRAFT-DO NOT QUOTE OR CITE
-------
0.20 R
0 10 20 30 40
% Change in Respiratory Hospitalization
50
Figure 7-16. Comparison of single-day lags (0-, 1-, 2-, 3-, 4-, and 5-day) to a cumulative
multiday lag (0- to 4-day) for % changes in total respiratory hospitalizations
per 40 ppb increase in 1-h max O3 in children less than two years of age.
Source: Derived from Burnett et al. (2001).
1 periods (Figure 7-17). Unrestricted lag models suggested that the O3 exposure from 3 to 5 days
2 prior had a greater impact on morning % PEF than more immediate exposures. Results were
3 similar when comparing single- and multiple-day exposure lags on the incidence of respiratory
4 symptoms in the morning (Mortimer et al., 2002).
5 Weisel et al. (2002) stated that a lag period of 1 to 3 days between exposure to O3 and
6 hospital admissions or emergency department visits for asthma was plausible because it might
7 take time for the disease to progress to the most serious responses following exposure.
8 In addition, taking medication could delay further the progression of the adverse effect.
9 Mortimer et al. (2002) discussed biological mechanisms for delayed effects on pulmonary
10 function, which included increased bronchial reactivity secondary to airway inflammation
11 associated with irritant exposure. Animal toxicology and human chamber studies (see
12 Chapters 5 and 6) provide further evidence that exposure to O3 may augment cellular infiltration
13 and cellular activation, enhance release of cytotoxic inflammatory mediators, and alter
14 membrane permeability. Examining longer lag periods allows studies to investigate the
15 cumulative O3-related effects over several days rather than one day only. The use of longer lag
16 periods also allows for delayed effects at 3 to 6 days to be observed. However, interpretation of
January 2005
7-97
DRAFT-DO NOT QUOTE OR CITE
-------
2.0
1.5
g 1.0
0.5
0.0
Single-day Lags
2
/ xX
Multiday Lag 1-5 [ / \
3/ /
1
\
\\
\ \
\ \
c\ \
6 \ \
-3 -2.5 -2 -1.5 -1 -0.5 0
% Change in PEF
0.5 1
1.5
Figure 7-17. Comparison of single-day lags (1-, 2-, 3-, 4-, 5-, and 6-day) to a cumulative
multiday lag (1- to 5-day) for % changes in PEF per 30 ppb increase in 8-h
avg O3 in urban children.
Source: Derived from Mortimer et al. (2002).
1 the results from multiday lags may not be as straightforward as that from single-day lag
2 analyses. Few field studies examined or presented exposure lags of more than 3 days. In a study
3 of asthma symptoms, Delfino et al. (1998) stated that no long-term lag effects were seen, but did
4 not provide the lags examined.
5 Bias resulting from the selection of lags has not been examined specifically for O3 effects.
6 However, the issue of lags has been investigated for PM and the results of this analysis are most
7 likely of relevance for O3. Lumley and Sheppard (2000) performed a simulation study to
8 examine model selection bias in air pollution epidemiology. Sheppard et al. (1999; reanalysis
9 Sheppard, 2003) had investigated the association between asthma hospital admissions and
10 ambient PM25 concentrations over a eight-year period in Seattle, WA. Note that the results from
11 Lumley and Sheppard (2000) and Sheppard et al. (1999) were based on GAM using default
12 convergence criteria. A negative control analysis, using simulated data with no association
13 between PM exposure and the health outcome, and a positive control analysis, in which a
14 specified non-zero excess risk is added to the simulation, were performed for comparison. The
15 bias from selection of best of seven lags (0 to 6 days) and residual seasonal confounding in the
16 negative control analysis (median log relative risk of 0.0013) was approximately half the log
January 2005
7-98
DRAFT-DO NOT QUOTE OR CITE
-------
1 relative risk estimated from the observed data (0.0027), after adjusting for season and
2 temperature. In the positive control model (true log relative risk of 0.0083), the bias was small
3 (median log relative risk of 0.0080). Results from these simulations indicate that bias from
4 selection of lags may be negligible when the true association is moderately large. However, if
5 the relative risk is relatively small, as in the case of air pollution epidemiology, bias may be of
6 issue. Selection of the largest risk estimate from a series of lags potentially can lead to positive
7 bias towards finding a significant association.
8 Selection of lag periods should depend on the hypothesis of the study and the potential
9 mechanism of the effect. Bias can result from the reporting of only the largest and most
10 significant risk estimate, as well as the reporting of single-day lag results when significant
11 relationship are found on multiple lag days. Most studies have shown that O3 has a fairly
12 consistent, immediate effect on respiratory health with the majority finding significant
13 relationships at 0- and 1-day lags, especially for acute mortality and hospitalization outcomes.
14 Some studies indicated a greater cumulative O3 effect observed over longer lag periods,
15 suggesting that in addition to single-day lags, multiday lags should be investigated to fully
16 capture a delayed O3 effect on health outcomes. The issue of lags warrants further investigation.
17
18 7.6.5 Confounding by Temporal Trends and Meteorologic Effects
19 The challenge of analyzing acute O3 effects in time-series studies is to avoid bias due to
20 confounding by daily to seasonal temporal factors. On a seasonal scale, the analysis must
21 remove the influence of the strong seasonal cycles that usually exist in both health outcomes and
22 O3. On a daily scale, weather factors and other air pollutants also may confound the association
23 of interest. This section discusses the interpretation of effect estimates after adjusting for
24 temporal trends and meteorologic effects.
25
26 7.6.5.1 Assessment of O3 Effects after Adjusting for Temporal Trends and
27 Meteorologic Effects
28 The relationship between O3 and health outcomes are significantly affected by temporal
29 trends and meterological factors, namely temperature. Analyses of the association between
30 health outcomes and O3 concentrations using raw data, therefore, can be misleading. In an
31 analysis of Madrid, Spain data by Diaz et al. (1999), a U-shaped relationship was observed
32 between mortality and O3 concentrations, with an associated minimum at 35 |ig/m3
January 2005 7-99 DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
(approximately 18 ppb) for 24-h avg O3. The negative portion of the slope is likely due to the
opposing seasonal cycles in mortality (high in winter) and temperature (low in winter).
However, little is discussed about the interpretation of the fitted morbidity and mortality
"effects" of temperature. See, for example, Figure 7-18, which shows the fitted nonaccidental
mortality as a function of natural spline smoothing of mean temperature in Montreal, Quebec
(Goldberg and Burnett, 2003). The positive slope of the temperature-mortality relationship is
fitted most tightly in the mild temperature range in which we do not expect mortality effects of
temperature. It is possible that temperature has mortality effects in the mild temperature range,
however because daily fluctuations of air pollution, especially O3, are strongly influenced by
weather conditions, ascribing the association between temperature and mortality entirely to
temperature effects may underestimate the effects of air pollution.
0.20-
o
i 0.10
o
v>
o
o.
If)
0.05-
0.00-
-0.05 -
-30 -20 -10 0 10 20
Mean Temperature (°C)
30
Figure 7-18. Daily nonaccidental mortality in Montreal, Canada as a function of mean
temperature, using natural splines with two degrees of freedom.
Source: Goldberg and Burnett (2003).
January 2005
7-100
DRAFT-DO NOT QUOTE OR CITE
-------
1 A 2003 HEI report investigated the impact of the selection of GAM convergence criteria to
2 adjust for temporal trends and weather variables in PM time-series studies (HEI, 2003). These
3 sensitivity analyses included the use of varying degrees of freedom for smoothing terms to adjust
4 for temporal trends in the Poisson regression model. Sensitivity analyses specifically for O3
5 effects have not been performed, with the exception of one new study. In the U.S. 95
6 communities data, Bell et al. (2004) performed a sensitivity analysis of the O3 excess mortality
7 risk estimates to tripling the degrees of freedom for smoothing terms used to adjust for temporal
8 trends. They found that varying the degrees of freedom from 7 to 21 per year did not
9 significantly affect the O3-mortality estimates, with effect estimates ranging from 0.82 to 1.08%
10 excess risk per 20 ppb increase in 24-h avg O3 during the previous week. Using more degrees of
11 freedom in temporal trend fitting (i.e., controlling shorter temporal fluctuations) means ascribing
12 more details of daily health outcomes to unmeasured potential confounders and possibly taking
13 away real weather and air pollution effects. However, results from this large multicity study
14 indicated that O3 effects were robust to aggressive smoothing of temporary trends.
15 Temporal cycles in daily hospital admissions or emergency department visits are often
16 considerably more episodic and variable than is usually the case for daily mortality. As a result,
17 smoothing functions that have been developed and tuned for analyses of daily mortality data
18 may not work as well at removing cyclic patterns from morbidity counts. Two methods are
19 commonly used for season adjustment, and an important distinction exists in the manner in
20 which these adjustments are applied in the analysis. The pre-adjustment method involves
21 applying the adjustment to both outcome and air pollution variables prior to the regression
22 analysis. In this case, the regression is done on the residuals following subtraction of smooth
23 functions for each variable. The second method, co-adjustment involves applying the
24 adjustment as part of the regression analysis, by fitting a function of time while simultaneously
25 fitting the regression effect of air pollution and weather factors. These two approaches have
26 been viewed as largely interchangeable. However, the co-adjustment approach may lead to
27 biased air pollution effect estimates in cases where both outcome and pollution variables exhibit
28 strong seasonal cycles. This was demonstrated using a 15-year time-series data of daily hospital
29 admissions for acute respiratory diseases in children under 2 years old (Burnett et al., 2001).
30 Note that this analysis used Poisson GAM with default convergence criteria. Pre-adjustment
31 followed by regression analysis yielded a statistically significant estimate of 14.1% increase in
January 2005 7-101 DRAFT-DO NOT QUOTE OR CITE
-------
1 admissions per 40 ppb increase in 1-h max O3 using year-round data. However, when the co-
2 adjustment method was applied, there was a statistically significant 7.2% decrease in
3 admissions. The authors suggested that the co-adjustment method allows O3 to compete with the
4 smoothing variable to explain some of the seasonal variability in the outcome, whereas pre-
5 adjustment eliminates the seasonal variability prior to analysis of O3 effects. Interestingly, when
6 the authors limited the analysis to the warm season (May-August), both methods yielded similar
7 results (32.3% versus 30.0% increase for co-adjustment and pre-adjustment, respectively)
8 implying that stratification by season can remove a significant amount of the confounding
9 seasonality. This finding may be important to consider in reviewing the acute O3 mortality and
10 morbidity literature since the vast majority of studies published over the past decade have used
11 the co-adjustment method. However, the use of pre-adjustment versus co-adjustment in time-
12 series studies is an unresolved issue. More empirical research in different locales is needed to
13 evaluate the merits of these two methods as far as O3 is concerned, and to determine what
14 endpoints may be affected.
15 An interesting study by Schwartz (2004) examined the sensitivity of the O3-mortality
16 relationship to methods used to control for temperature. Using a case-crossover analysis, the
17 effect of O3 on mortality was examined in 14 cities across the U.S. from 1986 to 1993. Control
18 days for an event were selected to be all other days from the same month of the same year.
19 Initially, temperature lagged 0 and 1 day was controlled using nonlinear regression splines with
20 3 degrees of freedom each. In a comparison analysis, control days were restricted to a subset
21 that was matched on temperature. The effect estimate for all year data was a 0.8% excess risk
22 per 40 ppb increase in 1-h max O3 in the analysis using nonlinear regression splines, compared
23 to a 0.9% excess risk using temperature matched controls. The effect estimates from the two
24 analyses were not significantly different. Results were similar when restricting analysis to warm
25 season only data.
26 More sensitivity analysis of O3 effect estimates to the extent of adjustment for temporal
27 trends and meteorological factors is needed, but perhaps it is equally as important to evaluate the
28 epidemiological adequacy of a given adjustment. For example, do the fitted mortality series
29 sufficiently depict influenza epidemics? Or, when larger degrees of freedom (e.g., 12 degrees of
30 freedom per year) are used, what "unmeasured" confounders, other than weather and pollution,
31 are the investigators trying to adjust? Even in PM studies that conducted sensitivity analyses,
January 2005 7-102 DRAFT-DO NOT QUOTE OR CITE
-------
1 investigators rarely stated assumptions clearly, and not enough discussions were provided as to
2 potential reasons for the sensitivity of results.
3 Given their relationship to health outcomes and O3 exposure, adjusting for temporal trends
4 and meteorologic factors is critical to obtain meaningful O3 effect estimates. While the
5 prevailing analytical approaches fit the data flexibly, the estimated effects of meteorologic
6 variables and their impact on the adjusted O3 effects are not adequately discussed. More work is
7 needed in this area to reduce the uncertainty involved in the epidemiologic interpretation of O3
8 effect estimates.
9
10 7.6.5.2 Importance of Season-Specific Estimates of O3 Health Effects
11 Analysis of O3 health effects is further complicated as the relationship of O3 with
12 temperature and with other pollutants appears to change across seasons. Such relationships can
13 be observed in Figure 7-19 from a study by Moolgavkar et al. (1995). In this study, Moolgavkar
14 et al. examined the relationship between daily mortality and air pollution (TSP, SO2, and O3) by
15 season in Philadelphia, PA for the period of 1973 to 1988. During the summer, there was a
16 positive relationship between O3 and TSP, as well as O3 and SO2. In contrast, the relationship
17 between O3 and TSP, and O3 and SO2 inversed during the winter. Note that a greater range of O3
18 concentrations was observed during the summer. The analyses indicated that while both TSP
19 and SO2 showed positive and significant associations with mortality in all four seasons in single-
20 pollutant models, O3 showed positive and significant associations only in the summer when the
21 mean O3 concentration was the highest (Figure 7-20). The O3-mortality association was negative
22 (though not significantly) in the winter when the mean O3 concentration was low. The addition
23 of TSP or SO2 in the regression did not attenuate the O3 effect estimate in the summer, and in the
24 three-pollutant model in the summer, only O3 remained significant. In another Philadelphia
25 study by Moolgavkar and Luebeck (1996), analyzed using GAM with default convergence
26 criteria but with only one nonparametric smoothing term, O3 was also positively and
27 significantly associated with mortality in the summer. A negative association between O3 and
28 mortality was observed during the winter in the single-pollutant model. With all pollutants
29 (TSP, SO2, NO2, and O3) included in the model, the O3 effect in the summer remained
30 significant. Both studies did not analyze the data for the year-round data set, therefore the
31 relationship between the excess risk estimates for each season and the year-round data could not
January 2005 7-103 DRAFT-DO NOT QUOTE OR CITE
-------
Summer
100-
^ 80-
Q. 60-
Q.
8 40^
20-
0-
0
30-
n 20-|
Q.
a.
S 1H
o-
50 100
TSP
150
50 100 150
TSP (|jg/m3)
loo-
se-
s'
Q. 60-
Q.
cp 40 -
O
20-
0-
200
Winter
Q.
a.
200
30-
20-
CD 10^
0-
0 20 40 60 80 100
S02 (ppb)
0 20 40 60 80 100
SO2 (ppb)
Figure 7-19. Scatterplots of daily levels of O3 with TSP and SO2 in Philadelphia, PA
by season.
Source: Derived from Moolgavkar et al. (1995).
1 be compared. The results from these studies, however, suggest that year-round analyses may
2 mask the positive (or negative) associations that may exist in particular seasons.
3 In the analyses of the U.S. 90 cities data by Samet et al. (2000; reanalysis Dominici et al.,
4 2003), the focus of the study was PM10, but O3 and other gaseous pollutants also were analyzed
5 in single- and multiple-pollutant models. In the reanalysis (Dominici et al., 2003), the combined
January 2005
7-104
DRAFT-DO NOT QUOTE OR CITE
-------
1.4-
*
j£ 1.2 -
_>
"•5
— 1 0 -
0.8-
+S
r7 fft
ii ii
•
Spring Summer
+
+T+S
I
{
i <
1
(
)
i
+
T
3
Fall Winter
26 ppb 36 ppb 16
ppb 12 ppb
Figure 7-20. Estimated total (nonaccidental) mortality relative risk per 100 ppb increase
in 24-h avg O3 of reach seasonal data set. Within each season, the left-hand
estimate is for O3 alone; the estimate in the middle (+T) is with TSP, the
right-hand estimate (+S) is with SO2 in the model. Seasonal mean O3
concentrations are noted.
1 O3-mortality estimate for all seasons, summer only, and winter only analyses were all
2 statistically significant at a lag of 1 day (see Figure 7-21). However, while an excess risk in
3 mortality was observed for all seasons and summer only analyses, a negative estimate was
4 obtained for the winter only analysis. It should be noted that, Samet et al. and Dominici et al.'s
5 analyses used a weather model specification that is more detailed than other studies in that it had
6 multiple terms for temperature and dewpoint (these two variables are generally highly
7 correlated). Thus, it is possible that the high concurvity of O3 with these weather covariates may
8 have produced these conflicting results. Another possibility is that, as mentioned previously, the
9 negative correlation between O3 and PM and other primary pollutants may have produced the
10 apparent negative relationship between O3 and mortality in the winter (note that PM and
11 mortality were positively associated). In the similar U.S. 95 communities study by Bell et al.
12 (2004), analyses with only winter data were not performed, however, both the all year and
13 summer only analyses indicated statistically significant positive risk estimates (1.04% and
14 0.78% excess risk per 20 ppb increase in 24-h avg O3, respectively, using a constrained
15 distributed 7-day lag model).
16 Anderson et al. (1996) examined the relationship between air pollution (O3, NO2, BS, and
17 SO2) and daily mortality (all cause, cardiovascular, and respiratory) in London, England for the
January 2005
7-105
DRAFT-DO NOT QUOTE OR CITE
-------
0)
o
'•p
.Q
Q
O
£
in
o
o.
All year
Summer
Winter
-1.5 -1.0 -0.5 0.0 0.5 1.0
% Change in Mortality
Figure 1-21. Marginal posterior distributions of the national average estimates of O3
effects on total mortality per 10 ppb increase in 24-h avg O3 at a 1-day lag
for all year, summer (June-August), and winter (December-February)
analyses in 90 U.S. cities.
Source: Derived from Dominici et al. (2003).
1 study period of 1987 to 1992 using a Poisson GLM model. They examined the associations
2 using data from all year, as well as the warm season (April-September) and the cool season
3 (October-March) separately. Their results indicated that the estimated O3 relative risks were
4 larger in the warm season than in the cool season for all cause mortality. The percent excess risk
5 estimated per 30 ppb increase in 8-h avg O3 (9 a.m.-5 p.m.) was 3.05% (95% CI: 1.39, 4.7),
6 4.37% (95% CI: 2.17, 6.62), and 0.96% (95% CI: -1.10, 3.06) for all year, warm season, and
7 cool season, respectively. A similar pattern was seen for cardiovascular mortality, but the
8 estimated risk was negative (not significantly) for the cool season. For respiratory mortality, the
9 estimated excess risks were similar between the cool and warm seasons. Many other studies also
10 reported larger excess mortality risks in the warm (or summer) season than in the cool (or
11 winter) season (see Figure 7-12 in Section 7.4.4). These studies showed cool season risk
12 estimates that were either smaller compared to warm season estimates or slightly negative (but
13 not significant). Of the studies that analyzed data by season, only one study in Pittsburgh, PA
14 (Chock et al., 2000) showed negative risk estimates in the summer.
January 2005
7-106
DRAFT-DO NOT QUOTE OR CITE
-------
1 The studies that observed larger (positive) associations between O3 and mortality in warm
2 seasons are consistent with the expectation that O3, if harmful, should have a stronger association
3 with health outcomes in the summer when concentrations are higher. However, the negative
4 O3-mortality associations seen in the winter in the U.S. 90 cities study (Samet et al., 2000;
5 reanalysis Dominici et al., 2003) and Philadelphia, PA data (Moolgavkar et al., 1995) suggest
6 that further examination of this issue is required. Specifically, if the O3 level in the winter is
7 shown to be negatively associated with factors (e.g., PM) that are positively associated with
8 mortality, then these potentially spurious negative O3-mortality associations can be explained.
9 Several examples of this phenomenon also exist in morbidity studies investigating the effect of
10 O3 on daily hospital admissions and emergency department visits (Anderson et al., 1998; Burnett
11 et al., 2001; Prescott et al., 1998; Thompson et al., 2001). A study by Thompson and colleagues
12 (2001) in Belfast, Northern Ireland observed a significant decrease in emergency department
13 visits for childhood asthma in the cold season (November-April), but not in the warm season.
14 Ozone concentrations were found to be inversely related to benzene levels (r = -0.65). After
15 adjusting for benzene levels, there was no significant association between O3 and asthma
16 emergency department visits.
17 Unlike the time-series studies examining outcomes of mortality, hospital admissions, and
18 emergency department visits, most acute field studies did not perform year-round analyses.
19 These acute field studies that examined the relationship between O3 and lung function,
20 respiratory symptoms, and inflammation focused primarily on the O3 effect during the warm
21 season, when O3 levels were expected to be high.
22 The potential influence of season on O3 effect estimates was examined using summary
23 density curves. The O3 effect observed in all year data was compared to effects from warm
24 season and cool season only data (Figures 7-22 and 7-23). Summary probability density curves
25 (or summary density curves) were calculated to review the effect estimates from the various
26 studies. To calculate the summary density curve, the normal distribution function first was
27 determined for each effect estimate and corresponding standard error. Then the individual
28 normal distribution functions were summed together to obtain the pooled normal distribution
29 function. The summary density curve is calculated by taking the derivative of the pooled normal
30 distribution function. Unlike a single normal density curve, the summary density curve is
31 distribution-free and may be multimodal. The summary density curves shown in Figures 7-22
January 2005 7-107 DRAFT-DO NOT QUOTE OR CITE
-------
0 bi
-12 -10 -8-6-4-20 2 4 6
% Change in Mortality
8 10 12
All year
Warm
season
Cool season
% area under the density curve and > 0 84% 89% 73%
Mean (SD) effect estimates 1.7% (4.4%) 2.4% (4.3%) -0.0% (4.4%)
Mode effect estimates 1.3% 1.6% 0.2%
Figure 7-22. Summary density curves of the % change in all cause mortality for all year
data and by season per 40 ppb increase in 1-h max O3 or equivalent. Effect
estimates from 14 studies have been included in the summary density curves
(see Figure 7-12 in Section 7.4.4 for the effect estimates).
1 and 7-23 were smoothed by adding a constant to the standard error of each effect estimate in the
2 calculation of the individual distribution functions. The constant is a default for normal
3 distribution densities and is larger when the number of effect estimates is smaller, as presented
4 by Silverman (1986). Since the normal distribution is unimodal, this constant will oversmooth
5 when the density is multimodal. In Figure 7-22, the summary density curves representing the %
6 all cause (nonaccidental) mortality associated with O3 concentrations are presented (see Figure
7 7-12 in Section 7.4.4 for the effect estimates). The summary density curves were calculated
8 using results from 14 studies that reported at least two of the three estimates. This figure
9 indicates that 84% of the area under the density curve has a value greater than zero for all year
10 data compared to 89% for warm season data and 73% for cool season data. Therefore, both all
11 year and warm season data generally indicates a significant, positive O3 effect. The mean effect
January 2005
7-108
DRAFT-DO NOT QUOTE OR CITE
-------
0.08
0.06
0.04
0.02
0.00
-15
0 15 30
% Change in Respiratory Hospital Admissions
45
All year
Warm season
Cool season
% area under the density curve and > 0
Mean (SD) effect estimates
Mode effect estimates
84% 88% 53%
7.7% (8.8%) 9.4% (11.3%) 1.6% (7.3%)
1.9% 4.3% -0.8%
Figure 7-23. Summary density curves of the % change in total respiratory hospital
admissions for all year data and by season per 40 ppb increase in 1-h max
O3 or equivalent. Effect estimates from six studies have been included in
the summary density curves (see Figure 7-8 in Section 7.3.3 for the effect
estimates).
1 estimate is a 1.7% excess risk in mortality per 40 ppb increase in 1-h max O3 using all year data,
2 compared to a slightly larger 2.4% excess risk using warm season data. The cool season only
3 data indicates that there is no excess risk associated with O3 concentrations.
4 Similar observations are made when examining the O3 effect on total respiratory hospital
5 admissions (Figure 7-23). Six studies provided season-specific estimates as well as all year
6 results (see Figure 7-8 in Section 7.3.3 for the effect estimates). Once again, a large % of the
7 area under the summary density curve is greater than zero when using all year and warm season
8 data, 84% and 88%, respectively, compared to cool season data, 53%. The mean O3 effect
9 estimate also is slightly larger for warm season data only, 9.4% excess risk per 40 ppb increase
10 in 1-h max O3, compared to all year analyses, 7.7% excess risk. A small O3 effect (1.6% excess
11 risk) is observed when using cool season data only.
January 2005
7-109
DRAFT-DO NOT QUOTE OR CITE
-------
1 Integrating seasonal influences across the various health outcomes supports the view that
2 O3 effects are different in the cool and warm seasons, with greater effects observed during the
3 warm season. As this relates to higher O3 levels produced during the warm season, the larger
4 effects are an appropriate conclusion. Therefore, these results indicate that warm season data
5 should be used to derive quantitative relationships for the effect of O3 on health outcomes. This
6 conclusion is supported by epidemiologic researchers who focus on warm season data as an
7 a priori design for the studies. The results also support a rationale to view the cool season as
8 inappropriate to derive information from in regards to the level of effect. However, studying
9 summer data only when all year data is available weakens the power of the study since less data
10 is analyzed. In addition, increased adverse health outcomes are observed in the winter, some of
11 which may be attributable to O3. The O3 effect in the winter time may be masked by the effects
12 of PM due to the negative correlation between these variables (see Section 7.6.6.2 for further
13 discussion). Therefore, analysis of all year data may be improved by adjusting for PM indices in
14 addition to adequate adjustment of meteorological factors and temporal trends. The methods of
15 statistical analysis, in addition to various other factors, need to be considered in the study design
16 stage to choose the proper time period to examine the health effects of O3. The data presented
17 here may aid in making this choice.
18 Seasonality influences the relationship between O3 and health outcomes as it may serve as
19 an indicator for changing meteorologic factors, namely temperature, and copollutant
20 concentrations. Given the potentially significant effect of season, O3 effect estimates computed
21 for year-round data need to be interpreted with caution. Small or no effects may simply reflect
22 the cancellation of positive associations in the summer and negative associations in the winter, or
23 the presence of confounding due to the strong seasonal character of O3 concentrations.
24
25 7.6.6 Assessment of Confounding by Copollutants
26 Potential confounding by daily variations in copollutants is another analytical issue to be
27 considered. With respect to copollutants, daily variations in O3 tend to not correlate highly with
28 most other criteria pollutants (e.g., CO, NO2, SO2, PM10), but may be more correlated with
29 secondary fine PM (e.g., PM2 5, sulfates) measured during the summer months. Assessing the
30 independent health effects of two pollutants that are somewhat correlated over time is
31 problematic. If high correlations between O3 and PM or other gaseous pollutants exist in a given
January 2005 7-110 DRAFT-DO NOT QUOTE OR CITE
-------
1 area, then disentangling their relative individual partial contributions to observed health effects
2 associations becomes very difficult. The changing relationship between O3 and other
3 copollutants also is of issue. In some urban locations, the correlation between PM indices and
4 O3 is positive in the summer and negative in the winter. This section will further discuss the
5 correlation between O3 and copollutants and confounding of the O3 effect by copollutants.
6
7 7.6.6.1 Relationship between Personal Exposure to O3 and Copollutants
8 To be confounders of the association between O3 and adverse health effects, copollutants
9 must be associated with both O3 exposure and the health outcome (Rothman and Greenland,
10 1998, p. 121). Many studies have shown that copollutants of O3, namely PM, NO2, SO2, and
11 CO, are associated with respiratory and, in some cases, cardiovascular health outcomes.
12 In addition, ambient levels of these copollutants, measured at central monitoring sites, have been
13 found to be highly correlated to ambient O3 concentrations. However, few studies have
14 examined the association between personal O3 concentrations and personal exposures to other
15 copollutants. In a scripted exposure study discussed earlier, Chang et al. (2000) examined the
16 relationship between 1-hour personal O3 and personal PM25 levels in several microenvironments,
17 including indoors, outdoors, and in vehicles. Chang et al. (2000) did not find a significant
18 correlation between personal O3 and PM25 concentrations in any of the microenvironments, even
19 after stratifying the data by season. In a Baltimore, MD study of susceptible populations (older
20 adults, individuals with COPD, and children), Sarnat et al. (2001) found that ambient 24-h avg
21 O3 concentrations and ambient 24-h avg PM25 levels were positively correlated (r = 0.67,
22 p < 0.05) in the summer and negatively correlated (r = -0.72, p < 0.05) in the winter. However,
23 no relationship was found between 24-h avg personal O3 and personal PM2 5 concentrations.
24 Interestingly, a significant correlation also was observed between ambient O3 and personal
25 PM2 5, with a mixed regression effect estimate of P = 0.28 (t = 4.00) in the summer and P =
26 -0.29 (t = -4.68) in the winter. In contrast to the results from Sarnat et al. (2001), a study by
27 Delfino et al. (2004) did not find a significant correlation between ambient 8-h max O3 and
28 ambient 24-h avg PM2 5 (r = 0.24) concentrations during two warm sampling periods, August-
29 October and April-June, in Alpine, CA. Personal PM measurements were taken using a
30 nephelometer which responds mainly to PM in the 0.1 to 10 |im range, with the highest response
31 in the fine PM range. Ambient 8-h max O3 was not found to be correlated with personal 8-h max
January 2005 7-111 DRAFT-DO NOT QUOTE OR CITE
-------
1 PM (r = 0.03) or personal 24-h avg PM (r = 0.01). Personal O3 exposure data was not available
2 in this study. Another study by Delfino et al. (1996) found that personal 12-h avg O3 levels were
3 not associated with ambient PM2 5 levels (r = 0.03) in a study of asthmatics (aged 9-16 years old)
4 in San Diego, C A. These studies provide limited evidence for a lack of correlation between
5 personal O3 levels and personal exposure to PM.
6
7 7.6.6.2 Assessment of Confounding Using Multipollutant Regression Models
8 The multipollutant regression model often is used to determine whether the pollutant-
9 specific effect is robust. However, due to the multicolinearity among O3 and pollutants, and the
10 changing correlation by seasons, multipollutant regression models may not adjust for potential
11 confounding adequately, especially when using year-round data. Results from the U.S. 90 cities
12 study (Samet et al., 2000; reanalysis Dominici et al., 2003), as depicted in Figure 7-24, indicated
13 that PM10 risk estimates were robust to including O3 and other gaseous pollutants in
14 multipollutant models. In a similar analysis, the effect of copollutants on O3-mortality risk
15 estimates also were investigated in this dataset. While the addition of PM10 in the model did not
16 substantially change the O3-mortality risk estimate, a slight decline in the O3 effect was observed
17 (Figure 7-25). In the extended U.S. 95 communities study (Bell et al., 2004), the city-specific
18 O3-mortality effects were found to be robust to the adjustment for PM10, as indicated by the
19 nearly 1:1 ratio between estimates with and without PM10 adjustment shown in Figure 7-26.
20 These results indicated that PM10 did not confound the association between O3 and mortality in
21 this large study. Limited data was available to examine the potential confounding effect of PM2 5
22 on the O3-mortality relationship. A weighted second-stage linear regression indicated that there
23 was no association between long-term PM2 5 average and the community-specific O3-mortality
24 effect estimate. Several other mortality and morbidity studies have investigated confounding of
25 O3 risk estimates using multipollutant models with year-round data, and most have reported that
26 O3 effects were robust to adjustment for copollutants (see Figures 7-9 and 7-13 in Sections 7.3.3
27 and 7.4.5, respectively).
28 Since the pollutant most correlated with O3 in the summer is sulfate (which is in the fine
29 particle size range), especially in the eastern U.S., the potential confounder of main interest for
30 O3 is PM25 and sulfate in the summer. However, the results from two-pollutant regression
31 models with O3 and sulfate (or PM2 5) should be interpreted with caution because both of these
January 2005 7-112 DRAFT-DO NOT QUOTE OR CITE
-------
1.0-
t o.sj
o
0
O)
c
o
-0.5-
-1.0-
Lag1
ABODE
Models
Figure 7-24. Posterior means and 95% posterior intervals of the national average estimate
of PM10 effects on total mortality from non-external causes per 10 ug/m3
increase in 24-h avg PM10 at a 1-day lag within sets of 90 U.S. cities with
pollutant data available. Models A = PM10 only; B = PM10 + O3; C = PM10 +
O3 + NO2; D = PM10 + O3 + SO2; E = PM10 + O3 + CO.
Source: Derived from Dominici et al. (2003).
1.0-
•4-1
.|S 0.5-
o
•- 0
U
0)
G)
C
re
g -0.5-
-1.0-
La
•
t .
4 «
i
gO
i
<
•
I
i
i i
j
La
i . -
i
31
i
> '
i
1 ,
i
•
i
•
t *
f
L<
1
> <
i
1
ig2
l
> 4
1
I
•
> <
1
1
ABCDEABCDEABCDE
Models
Figure 7-25. Posterior means and 95% posterior intervals of the national average estimate
of O3 effects on total mortality from non-external causes per 10 ppb increase
in 24-h avg O3 at 0-, 1-, and 2-day lags within sets of 90 U.S. cities with
pollutant data available. Models A = O3 only; B = O3 + PM10; C = O3 + PM10
+ NO2; D = O3 +PM10 +SO2; E = O3 + PM10 + CO.
Source: Derived from Dominici at al. (2003).
January 2005
7-113
DRAFT-DO NOT QUOTE OR CITE
-------
O)
re
re
Q
o
S
Q.
O
I
15 -
10 -
5 -
0 -
-5 -
= National Average Effect
-5
10
15
Without PM10 Adjustment
(using only days with PM10 data)
Figure 7-26. Maximum likelihood estimates of O3-mortality for 95 U.S. communities,
determined using a constrained distributed lag model for lags 0 through
6 days. Same dataset was used for O3 estimates with and without adjustment
for PM10.
Source: Derived from Bell et al. (2004).
1 pollutants are formed under the same atmospheric condition and are both part of the "summer
2 haze" pollution mix. A simple two-pollutant regression model does not address their possible
3 synergistic effects, and the high correlation between the two pollutants may lead to unstable and
4 possibly misleading results. In any case, most studies that analyzed O3 with PM indices did not
5 have PM2 5 data and very few examined sulfate data. The studies that did have PM2 5 data,
6 including Santa Clara County, CA (Fairley, 1999; reanalysis Fairley, 2003), Philadelphia, PA
7 (Lipfert et al., 2000a), and Detroit, MI (Lippmann et al., 2000; reanalysis Ito, 2003), examined
8 copollutant models for year-round data only, but O3 mortality risk estimates were not
January 2005
7-114
DRAFT-DO NOT QUOTE OR CITE
-------
1 substantially affected by the addition of PM25. A mortality study by Lipfert et al. (2000a) also
2 found that O3 risk estimates were not affected by the addition of sulfate. Amongst the morbidity
3 studies, the two summertime studies in Toronto, Canada by Burnett et al. (1997b, 2001) found
4 that the O3 effect was only slightly attenuated after including PM25 in the model. In one of these
5 studies (Burnett et al., 1997b), the effect of O3 also was adjusted for sulfate. With the addition of
6 sulfate in the model, the risk estimate for O3 remained relatively stable, from an 11% excess risk
7 to a 9% excess risk per 20 ppb increase in 12-h avg O3 at a 1-day lag.
8 Other studies have estimated O3 health risks with copollutants in the model by season.
9 Amongst the mortality studies (see Figure 7-14 in Section 7.4.5), the O3 risk estimates in the
10 warm season were mostly positive and significant, with the exception of the Pittsburgh, PA
11 analysis by Chock et al. (2000). Adjusting for copollutants, in particular PM indices, did not
12 substantially change the O3-mortality effect estimates, with both slight reductions and increases
13 observed in the adjusted estimates. In the analysis using cool season data only, the O3 effect
14 estimates were generally negative, but none were statistically significant. In contrast to the
15 analyses of warm season data, the O3 risk estimates all increased slightly with the adjustment of
16 PM indices. The inverse relationship between O3 and PM during the cool season most likely
17 influenced the O3-mortality effect estimates in the single-pollutant models. Thus, the
18 confounding effect by PM indices appears to vary by season, most likely due to the changing
19 relationship between O3 and PM by season. These results indicate that although PM does not
20 seem to influence significantly the association between O3 and mortality during the warm
21 season, PM may be a confounder of the O3-mortality relationship in the cool season.
22 A study of respiratory hospitalizations in 16 Canadian cites by Burnett et al. (1997a) also
23 stratified O3 risk estimates by season. A preliminary analysis studying O3 effects found that a
24 positive association between O3 and respiratory hospitalizations was observed in the spring,
25 summer, and fall, but not in the winter. In an analysis restricted to warmer months (April-
26 December), the pooled O3 risk estimates for all cities were significant but attenuated with the
27 addition of copollutants and dewpoint temperature into the model. Of the 16 Canadian cities,
28 Montreal and Vancouver showed no association between O3 and respiratory hospitalizations
29 after adjusting for dewpoint temperature. After the exclusion of these two cities, the O3 risk
30 estimates were found to be robust with the addition of copollutants in the models. Two
31 additional respiratory hospitalization studies in the metropolitan Toronto, Canada area by
January 2005 7-115 DRAFT-DO NOT QUOTE OR CITE
-------
1 Burnett et al. (1997b, 2001) also observed consistent O3 risk estimates with the inclusion of
2 copollutants. The analyses in both studies were restricted to warm months (May-September).
3 In field studies, power to assess independent O3 effects may be limited by small sample
4 sizes and short follow-up times. Among the field studies, the O3 effect also was found to be
5 robust to the addition of copollutants in multipollutant models, with a few exceptions.
6 For example, the effect of O3 on PEF was not robust to adjustments for PM2 5 and sulfate, in
7 studies by Romieu et al. (1996) and Neas et al. (1999). In general, however, O3 effects on
8 respiratory symptoms (Romieu et al., 1996), lung function parameters (Brauer et al., 1996, Gold
9 et al., 1999), and asthma medication use (Gent et al., 2003) were robust to inclusion of PM25.
10 Often, the effects for O3 were found to be stronger than those for PM.
11 Multipollutant regression analyses indicated that O3 risk estimates were not sensitive to the
12 inclusion of copollutants, including PM25 and sulfate, in both year-round and warm season data.
13 These results suggest that the effect of O3 on respiratory health outcomes appears to be robust
14 and independent of the effects of other copollutants. However, there is concern as to whether
15 analysis of the O3 effect on health outcomes is confounded by PM indices in the cool season.
16 In addition, uncertainty remains as to the use of multipollutant regression models to assess the
17 independent health effects of pollutants that are correlated.
18
19 7.6.7 Issues of Model Uncertainty and Multiple Hypothesis Testing
20 Epidemiologic studies that investigated the association between O3 and various health
21 outcomes often found a significant effect. A major concern is whether these significant
22 associations are an artifact of model selection resulting from multiple hypothesis testing.
23 Testing multiple hypotheses may, at times, be appropriate. For example, developing several
24 hypotheses a priori allows researchers to explore more thoroughly potential mechanisms for an
25 O3-related health effect. Sensitivity analyses, which are critical for model validation, also use
26 multiple hypothesis testing. The basic issue with multiple hypothesis testing is that an extremely
27 large number of models are possible, any of which may turn out to give the best statistical "fit"
28 of a given set of data. Including all potentially confounding variables into the model is not
29 practical as this may result in overfitting the model and inflated standard errors. On the other
30 hand, selection of one "best" model ignores the uncertainty involved in model selection and
31 leads to an underestimation of the error. Akaike Information Criterion and Bayes Information
January 2005 7-116 DRAFT-DO NOT QUOTE OR CITE
-------
1 Criterion are some of the statistical methods used to assist in model variable selection. Recent
2 attention has focused on Bayesian model averaging as an efficient method to incorporate model
3 uncertainty into decision-making.
4 A few authors have applied Bayesian model averaging to study the effect of air pollution
5 on mortality. Clyde et al. (2000) and Clyde (2000) used Bayesian model averaging to analyze
6 the relationship between mortality and PM concentrations from Phoenix, AZ and Birmingham,
7 AL, respectively. In addition to the uncertainty of effect estimation, Bayesian model averaging
8 incorporated uncertainty regarding the choice of confounding variables, pollutants, and lags.
9 In the Phoenix, AZ study, Clyde et al. (2000) did not observe a PM2 5 effect on mortality, but did
10 find that coarse PM (PM10_2 5) was significantly associated with increased mortality. In a
11 reanalysis of the Birmingham, AL study (original analysis Schwartz, 1993), Clyde (2000)
12 observed that the PM10 effect originally estimated by Schwartz was plausible but Bayesian
13 model averaging results supported a smaller risk estimate. However, Clyde (2000) noted that
14 her analysis of the Birmingham data did not take into consideration factors that might bias the
15 estimated effect toward the null. For example, measurement error in the exposure variables were
16 not considered. In addition, the Poisson model (similar to many other regression models)
17 assumed that all individuals in a population had equal risks, including potentially susceptible
18 populations such as those with respiratory illnesses and outdoor workers.
19 Only one study using Bayesian model averaging reported a coefficient for O3-related
20 mortality. Koop and Tole (2004) used Bayesian model averaging to analyze the effect of various
21 air pollutants, including O3, SO2, CO, NO, NO2, PM10_2 5, and PM25, on mortality in Toronto,
22 Canada. Current values and up to 3-day lags were considered. In addition, a comprehensive set
23 of meteorological variables were included in the models. The 50+ explanatory variables
24 required the fitting of an enormous number of potential models. Although the point estimates
25 for all pollutants were positive, very small effects were found. Sixty-six percent of the PM data
26 used to calculate these effect estimates were imputed. Ozone data was collected daily,
27 eliminating the need for imputation. For O3, the cumulative effect on non-accidental deaths was
28 an excess of 0.054 deaths (posterior SD 0.159) per one standard deviation (9.15 ppb, 24-h avg
29 O3) increase in O3 levels. The most probable O3 model estimated the same-day effect of O3 on
30 mortality to be a statistically significant 0.526 (posterior SD 0.176) excess deaths. However, this
31 most probably model received only 0.23% of the probability. Koop and Tole concluded that the
January 2005 7-117 DRAFT-DO NOT QUOTE OR CITE
-------
1 standard error of the cumulative effect was much too large to base policy advice. However, in
2 the context of the many interaction terms, meteorological variables, smoothing surfaces, and the
3 relatively loose posterior distribution, it is likely that Koop and Tole have overestimated the
4 variance of their pollution coefficients.
5 Model diagnostics may be a way to reduce model uncertainty (George, 1999 in comments
6 to Hoeting et al., 1999). However, Hoeting et al. state that model diagnostics is often based upon
7 methods that use multiple testing. They believe that diagnostics should be used first to suggest
8 better models and Bayesian model averaging should be used later to compare all models. The
9 models considered should have "appreciable likelihood" or be excluded from Bayesian model
10 averaging (George, 1999). A problem with Bayesian model averaging occurs when variables are
11 highly correlated. When this occurs, the estimated posterior effects may be diluted, resulting in
12 smaller coefficients (George, 1999). George believed that the issue of dilution could be
13 addressed by altering the prior probabilities used. In their reply to George, Hoeting et al. stated
14 that dilution is a problem when the highly correlated variables act by the same mechanism and
15 may serve as surrogates for the same variable, which may be the case for air pollutants.
16 While Bayesian model averaging can theoretically be used to take into account uncertainty,
17 claims of causality based on observational studies may be highly sensitive to the choice of prior
18 distributions and class of models under consideration (Clyde et al., 2000). Additional research in
19 this area may provide new and interesting insights into the issues of model uncertainty and
20 multiple hypothesis testing.
21
22 7.6.8 Concentration-Response Function and Threshold
23 An important consideration in determining whether a safe level of O3 can be identified is
24 whether the concentration-response relationship is linear across the full concentration range or
25 instead shows evidence of a threshold. Of particular interest is the shape of the concentration-
26 response curve in the vicinity of the current 8-h NAAQS for O3 of 80 ppb. The O3
27 concentration-response relationship has been explored in several studies.
28 To examine the shape of the concentration-response relationship between O3 and mortality,
29 Gryparis et al. (2004) used meta-smoothing to combine smooth curves across the 23 European
30 cities in a hierarchical model. For the summer period, the estimated concentration-response
January 2005 7-118 DRAFT-DO NOT QUOTE OR CITE
-------
1 curve did not appear to deviate significantly from linearity within the range of O3 concentrations
2 commonly observed in European cities.
3 In the U.S. 95 communities study (Bell et al., 2004), effect estimates calculated using only
4 days with 24-h avg O3 levels less than 60 ppb were compared to those using all data. At a lag of
5 1 day, O3 was associated with an excess risk of 0.36% per 20 ppb increase in 24-h avg O3 using
6 data from all days and only a slightly smaller risk of 0.30% when data was limited to days less
7 than 60 ppb. These results suggest that if there is a threshold, it is present at 24-h avg O3 levels
8 below 60 ppb. Fairley (2003) reanalyzed the Santa Clara County mortality data using GAM
9 with stringent convergence criteria and examined a new exposure index for O3. He noted O3
10 concentrations exceeding 60 ppb each hour and calculated a daily sum of these exceedances.
11 Fairley's index incorporates measures of concentration and exposure duration. This type of
12 index is called a linear time-integrated concentration, also known as dosage. The O3 index with
13 the 60 ppb threshold level was found to be significantly associated with mortality in single-
14 pollutant models as well as in multi-pollutant models. Two other threshold levels were
15 examined, 40 ppb and 80 ppb. Both produced statistically significant results in single-pollutant
16 models. These results indicate that the threshold for O3-mortality effects is less than 40 ppb.
17 The implication for thresholds in terms of the three standard indices (i.e., 1-h max, 8-h max, and
18 24-h avg) is unclear, but there may be an empirical relationship.
19 Vedal et al. (2003) observed that the annual mean 1-h max O3 concentration of 27.3 ppb in
20 Vancouver, Canada, was lower than that in any of the 90 NMMAPS cities (Samet et al., 2000),
21 thus a study in this city may be able to better focus on the shape of the concentration-response
22 curve at lower levels. In this Vancouver study, a statistically significant O3 effect was observed
23 on total mortality at a 0-day lag during the summer. Statistically significant effects on
24 respiratory mortality at a 2-day lag and marginally significant effects on cardiovascular mortality
25 at a 0-day lag also were observed for O3 in the summer. The O3 effect on mortality was found to
26 be robust in two-pollutant models. Vedal et al. (2003) questioned if O3, other gaseous pollutants,
27 and PM were acting as surrogate markers of pollutant sources that contain more toxic
28 compounds, as the low measured concentrations were unlikely in their opinion to cause the
29 observed effects. They further stated that measurement error and interference by meteorological
30 factors might have contributed to the inability to detect a threshold. Vedal et al. (2003)
31 concluded that O3 concentrations were associated with adverse effects on mortality even at low
January 2005 7-119 DRAFT-DO NOT QUOTE OR CITE
-------
1 levels. Although this study supports the argument that there is no threshold concentrations
2 below which adverse effects cannot be detected, the results must be interpreted with caution as
3 concerns remain.
4 Kim et al. (2004) investigated the presence of a threshold in O3-mortality effects in Seoul,
5 Korea by analyzing data using a log linear GAM (linear model), a cubic natural spline model
6 (nonlinear model), and a B-mode splined model (threshold model). Models were stratified by
7 season and adjusted for PM10, long-term time trend, and meteorological variables. An estimated
8 threshold value of 47 ppb was observed for 1-h max O3. None of the other pollutants examined,
9 including PM10, SO2, NO2, and CO, had a nonlinear association with mortality. Using summer
10 data only, the B-spline model resulted in an excess mortality risk of 7.1% (95% CI: 3.1, 11.2)
11 per 40 ppb increase in 1-h max O3, compared to an excess risk of 3.6% (95% CI, 0.5, 6.8)
12 calculated using the log linear model. If a threshold truly exists, results from the Kim et al.
13 (2004) study suggest that the use of log-linear models may underestimate the O3 effect on
14 mortality at levels above the threshold.
15 In London, England data (Anderson et al., 1996), an adjusted O3-mortality plot indicated a
16 possible threshold level around 50 ppb for 8-h avg O3. A study by Simpson et al. (1997) in
17 Brisbane, Australia observed a significant excess risk in mortality only in the highest quintile of
18 O3 exposure, which had a mean concentration of 42 ppb for 1-h max O3. One study by Lipfert
19 et al. (2000b) examined the presence of a threshold in the effect of chronic O3 exposure on
20 mortality in U.S. veterans. A simple concentration-response plot comparing the risk estimate in
21 the upper two tertiles to that from the lowest tertile seemed to indicate a threshold level of
22 approximately 140 ppb of 1-h max O3 during the period 1975 to 1981 for both concurrent
23 mortality (1975-1981) and delayed mortality (1982-1988).
24 Among several studies with morbidity outcomes, examination of the shape of the
25 concentration-response function indicated evidence of an effect threshold. In a study of all-age
26 respiratory hospital admissions in Toronto, Canada, effects of O3 appeared to become apparent
27 only above approximately 30 ppb daily 1-h max O3 (Burnett et al., 1997b). In London, England,
28 Ponce de Leon et al. (1996) observed an indication of a threshold in the O3 effect on
29 hospitalizations at 40 to 50 ppb for 8-h max O3 and 50 to 60 ppb for 1-h max O3. In a study of
30 emergency department visits for asthma in St. John, Canada, effects observed in the over 15
31 years age group were apparent only when data above the 95th percentile (75 ppb daily 1-h max
January 2005 7-120 DRAFT-DO NOT QUOTE OR CITE
-------
1 O3) were included (Stieb et al., 1996). However, other morbidity studies observed a monotonic
2 increase in the concentration-response function, suggesting that there was no threshold in O3
3 effects on hospitalizations and emergency department visits (Burnett et al., 1997a; Jaffe et al.,
4 2003; Petroeschevsky et al., 2001; Tenias et al., 1998).
5 In a field study by Mortimer et al. (2002), the association of ambient O3 levels with PEF
6 and asthma symptoms was investigated in eight urban cities in the U.S. The mean 8-h avg O3
7 was 48 ppb, with less than 5% of days exceeding 80 ppb. Analysis performed using all data
8 indicated that a 15 ppb change in 8-h avg O3 was associated with statistically significant
9 decrements in PEF (-0.59% [95% CI: -1.05, -0.13]) and increased incidence of respiratory
10 symptoms (odds ratio of 1.16 [95% CI: 1.02, 1.30]) over multiday lag periods. When data was
11 restricted to days when ambient O3 concentrations were less than 80 ppb, the O3 effects
12 persisted, with a significant PEF decline (-0.70% [95% CI: -1.29, -0.12]) and incidence of
13 morning symptoms (odds ratio of 1.17 [95% CI: 1.01, 1.35]). A study by Chen et al. (1999) also
14 found that there was no clear threshold in the O3 effect on FEVj and FVC in Taiwanese school
15 children.
16 Note that adjusting for seasonal cycles does not address the issue of the changing
17 relationship between O3 concentrations and personal exposure across seasons. The ambient O3
18 levels are lower in the cold season, but people are likely to be exposed to even lower levels of
19 O3 in cold seasons due to the shorter time spent outdoors and the longer time spent indoors with
20 closed windows. This is in contrast to what occurs with fine particles, which can effectively
21 penetrate the indoors. Thus, a more "accurate" concentration-response relationship may need to
22 be examined in a summer-only data set (which may suffer low data density in the low
23 concentration range). Even for summer data, however, an interpretation of the relationship is
24 not straightforward because of the possible influence of the use of air conditioning (an effective
25 remover of O3). Greater use of air conditioning would be expected on hot days when the
26 O3 level is higher, but the use of air conditioning may also vary from city to city and across
27 social class within a city. These complications make it difficult to examine the existence of a
28 threshold of O3 health effects in the observational data.
29 Limited studies have examined the issue of thresholds in O3 health effects studies. Some
30 studies have found a low level threshold while others have found no threshold in O3 effects.
31 An absence of a detectable threshold in population studies does not indicate an absence of
January 2005 7-121 DRAFT-DO NOT QUOTE OR CITE
-------
1 individual thresholds. For a further discussion on thresholds in air pollutant health effects, see
2 Section 8.4.7 in the 2004 PM AQCD. While no conclusion can be made regarding the threshold
3 issue, the limited evidence shows that the possible threshold level may be well below the current
4 standards. The distribution of thresholds, particularly around the NAAQS value of 80 ppb for
5 8-h max O3, needs to be further investigated.
6
7 7.6.9 Spatial Variability in O3 Effect
8 As described in Chapter 3 of this AQCD, O3 concentrations tend to be more spatially
9 variable than PM2 5 concentrations in urban areas. In addition, relative personal exposures to O3
10 likely vary by region. This spatial variability in O3 concentrations and personal exposures may
11 contribute to the heterogeneity in observed O3 health effects. More than 80% of the O3-mortality
12 estimates from the various studies conducted in North America, South America, Europe, and
13 Australia were between 0 and 7% excess risk per 40 ppb increase in 1-h max O3 using year-
14 round data. In general, the O3-mortality estimates were greater when using summer only data
15 compared to year-round data. Though not all statistically significant, most of the O3-mortality
16 estimates were greater than zero, indicating a positive relationship between O3 exposure and
17 mortality. The O3 risk estimates from the numerous hospitalization and emergency department
18 visit studies were generally larger in magnitude and more variable from study to study compared
19 to the mortality studies. These differences in the O3 effect estimates may be attributable to the
20 greater variability in the outcome measure, such as more subcategories of outcome and varying
21 degrees of severity, in hospitalization studies compared to mortality studies.
22 As differences in study design, population, and data analysis may affect risk estimates,
23 studies that were conducted in multiple cities using standardized methods were further examined
24 to investigate the spatial heterogeneity of O3 effects. Bell et al. (2004) conducted a time-series
25 analysis of O3 and mortality in 95 U.S. communities from 1987 to 2000. A 10 ppb increase in
26 O3 in the previous week was associated with a statistically significant increase of 0.52% excess
27 risk of mortality in the pooled analysis of 95 communities. Although some heterogeneity was
28 observed among the communities (previously shown in Figure 7-11 of Section 7.4.3), the range
29 of the community-specific effect estimates were fairly narrow. Of the 95 U.S. communities, 93
30 had positive O3-mortality risk estimates. Only 5 had risk estimates greater than 1% per 10 ppb
January 2005 7-122 DRAFT-DO NOT QUOTE OR CITE
-------
1 increase in 24-h avg O3 during the previous week, with all communities indicating an excess
2 mortality risk less than 2%.
3 Greater heterogeneity was observed in the European study of 23 cities in 14 countries
4 (Gryparis et al., 2004). In the year-round analyses, only 8 of the 23 cities had positive O3-
5 mortality effect estimates. However, in the analyses using summer data only, the risk estimates
6 were positive in 19 of the 23 cities, with a range of 0.8 to 8% excess risk per 40 ppb increase in
7 1-h max O3. The heterogeneity may be attributable to the considerable variability among
8 countries in factors that may influence the relationship between ambient O3 concentrations and
9 personal exposure to O3, such as climate, use of air conditioning, personal activity patterns, and
10 socioeconomic factors. In addition, the variability in the concentration and composition of co-
11 existing pollutants by cities or countries may contribute to the heterogeneity in the O3-mortality
12 effects. For example, concentrations of NO2 may vary widely by region, depending on the
13 differences in traffic density.
14 Among the hospitalization studies, Burnett et al. (1997a) conducted the largest multicity
15 study of 16 Canadian cities. The mean daily 1-h max O3 was 31 ppb in the 16 cities. The pooled
16 O3 estimate was 5.6% (95% CI: 3.4, 7.9) excess risk in respiratory hospitalization per 40 ppb
17 increase in 1-h max O3 using warm season data (April to December). The risk estimates were
18 fairly homogenous across the 16 Canadian cities, ranging from 3.1% for Vancouver to 7.7% for
19 Quebec City.
20 Anderson et al. (1997) investigated the association between O3 and hospital admissions for
21 COPD in five European cities, London, Paris, Amsterdam, Rotterdam, and Barcelona. The
22 pooled risk estimate was 5.0%, 4.7%, and 3.5% excess per 30 ppb increase in 8-h max O3 for
23 year-round, warm season, and cool season data, respectively. Results from the APHEA study
24 showed similar variability to that from the Burnett et al. (1997a) study. The year-round excess
25 risk estimates were lower in the two Dutch cities, 3.0%, compared to that in Paris, 9.8%.
26 In general, however, there was no significant evidence of heterogeneity in the O3 effects among
27 the five European cities.
28 Among the field studies, various respiratory health outcomes were examined, including
29 PEF, spirometric parameters, respiratory symptoms, and medication use. Only one field study
30 investigated the O3 effect in several locations (Mortimer et al., 2002). Mortimer et al. (2002)
31 investigated the association between ambient O3 concentrations, and PEF and asthma symptoms
January 2005 7-123 DRAFT-DO NOT QUOTE OR CITE
-------
1 in asthmatic children living in eight urban cities in the U.S. - St. Louis, MO; Chicago, IL;
2 Detroit, MI; Cleveland, OH; Washington, DC; Baltimore, MD; East Harlem, NY; and Bronx,
3 NY. In the analysis pooling data from all eight cities, a 15 ppb increase in 8-h avg O3 was
4 associated with a significant decrement of -0.59% in morning PEF for a 5-day cumulative lag
5 period. The % changes in PEF were negative in all cities except for Baltimore, 0.24%. Among
6 the other seven cities, the % changes in PEF were quite homogenous, with values ranging from
7 -0.54% for Washington, DC to -0.86% for St. Louis. A 15 ppb increase in 8-h avg O3 also was
8 associated with an increased incidence of morning symptoms in the pooled analysis (odds ratio
9 of 1.16 for a 4-day cumulative lag period). In all cities except for St. Louis, there was an
10 increase in the incidence of morning symptoms. The odds ratios for incidence of morning
11 symptoms varied more by city compared to the PEF measurements, ranging from 1.09 for
12 Chicago to 1.72 for Detroit. The greater variance in incidence of symptoms may indicate the
13 lack of standardization in the use of symptoms as a health outcome measure.
14 Most of the multicity studies found consistent O3 effect estimates for mortality,
15 hospitalizations, and other respiratory health outcomes. The slight heterogeneity of O3 effects
16 may be partially attributable to the use of centrally located ambient monitors to assess exposure.
17 There may be differences in relative personal exposures to O3 due to varying factors, namely use
18 of air conditioning and activity patterns, that affect the relationship between personal exposure
19 and ambient concentrations. The variability in the concentration and composition of
20 copollutants present also may contribute to the heterogeneity of the effect of O3 on health
21 outcomes as confounding by copollutants may vary by region.
22
23 7.6.10 Health Effects of O3 in Susceptible Populations
24 In this section, the effects of O3 on morbidity and mortality in potentially susceptible
25 populations will be examined. In epidemiology studies of O3 health effects, the most widely
26 studied subpopulation was asthmatics. Also of interest were the observed health effects of O3 on
27 different age groups, particularly children and the elderly. This section begins with a discussion
28 of the O3-related health effects in asthmatics.
29
30
January 2005 7-124 DRAFT-DO NOT QUOTE OR CITE
-------
1 7.6.10.1 Health Effects Associated with Ambient O3 Exposure in Asthmatics
2 Epidemiological studies of health effects from acute O3 exposure in asthmatics have
3 examined a range of outcomes: pulmonary function, respiratory symptoms, inflammation,
4 emergency room visits, hospital admissions, and mortality. Chronic O3 exposures have been
5 associated with similar outcomes, with the exception of emergency room visits and
6 hospitalizations. Both are discussed in the earlier text. This subsection draws together this
7 information to examine whether the evidence indicates that O3 exposure impacts asthmatics.
8 In Germany and Mexico City, O3 exposure was associated with a decline in FEVj in
9 asthmatic adults and children (Hoppe et al., 1995a; Romieu et al., 2002). Change in FEVj also
10 was examined in a group of asthmatic hikers in Mount Washington, NH (Korrick et al., 1998).
11 Compared to the healthy subjects, the asthmatic subjects experienced a four-fold greater decline
12 in FEVj with the same exposure to O3. The results from the hiker study are consistent with those
13 observed in controlled human exposure studies (discussed in Chapter 6), which also indicate
14 significantly greater decrements in FEVj among mild asthmatics versus nonasthmatic subjects
15 with heavy intermittent exercise.
16 PEF was examined in panels of asthmatics in several field studies (see Figures 7-la and
17 7-lb). Collectively, all the studies indicated decrements of morning peak flow but most of the
18 estimates were not statistically significant. One multicity study of eight urban areas in the U.S.
19 observed reductions in morning PEF that were not significant in each individual city (Mortimer
20 et al., 2002). However, the analysis combining data from all eight cities indicated a statistically
21 significant change with a cumulative lag of 1 to 5 days. Further analysis showed that the
22 incidence of > 10% decline in morning PEF was statistically significant, which was discussed by
23 the author as an indication that O3 exposure may be associated with clinically significant changes
24 in PEF in asthmatic children. The study examined 846 asthmatic children, the largest asthma
25 panel study reported.
26 Respiratory symptom increases in asthma panels were examined in several field studies,
27 some of which also examined PEF as discussed above. The health indicators examined varied
28 among these studies and the analyses results were both negative and positive with a few being
29 statistically significant. Collectively, they are suggestive of a potential effect on respiratory
30 symptoms but the evidence in the available studies is not strong. Two U.S. studies examining
31 larger panels may be better studies from which to draw inferences as the large sample size
January 2005 7-125 DRAFT-DO NOT QUOTE OR CITE
-------
1 provides greater power to examine the effect of O3 on respiratory symptoms. The eight U.S.
2 urban cities study mentioned above reported morning symptoms in the 846 asthmatic children to
3 be most strongly associated with a 4-day cumulative lag period of O3 concentrations (Mortimer
4 et al., 2002). A New England study examined 271 asthmatic children and observed a significant
5 O3 effect on a variety of respiratory symptoms at a lag of 1 day among the 130 subjects who used
6 maintenance asthma medications (Gent et al., 2003).
7 Few epidemiological studies have examined airway inflammation in asthmatics. A Mexico
8 City study indicated that supplementation with antioxidants may modulate the impact of O3
9 exposure on the small airways of children with moderate to severe asthma (Romieu et al., 2002).
10 A related study indicated that asthmatic children with GSTM1 null genotype were found to be
11 more susceptible to the impact of O3 exposure on small airways (Romieu et al, 2004).
12 An additional study in Mexico City examined DNA strand breaks in nasal epithelial cells in
13 asthmatic and nonasthmatics medical students and noted greater genotoxic damage in asthmatics
14 (Fortoul et al., 2003).
15 Emergency department visits for asthmatics have been examined in several studies and
16 range from negative to positive results with limited analyses providing significant results (see
17 Figure 7-6 in Section 7.3.2). Examination of the studies indicated that seasonal summer studies
18 tended to yield positive outcomes, as expected based on earlier discussions. Two studies in
19 Atlanta, GA (Tolbert et al., 2000) and Valencia, Spain (Tenias et al., 1998) indicated significant,
20 positive effects in warm season analyses. Further, a Canadian study, one of the larger studies
21 conducted in the summer season, reported a large significant increase in asthma emergency
22 department visits when the daily 1-h max O3 concentration exceeded 75 ppb (Stieb et al., 1996).
23 A three-city study in Ohio also indicated a positive result during the summer (Jaffe et al., 2003).
24 Other studies of mostly year-long data tended to produce nonsignificant results, which in some
25 cases were negative (Atkinson et al., 1999a; Castellsague et al., 1995; Thompson et al., 2001;
26 Tobias etal., 1999).
27 Hospital admission studies that specifically examined asthmatics were fewer in number
28 than those that examined total respiratory diseases. Significant effects were noted in all age
29 groups in studies conducted in Seattle, WA (Sheppard et al., 2003), New Jersey (Weisel et al.,
30 2002), Toronto, Canada (Burnett et al., 1999), London, England (Anderson et al., 1998),
31 Brisbane, Australia (Petroeschevsky et al., 2001), and Hong Kong (Wong et al., 1999a).
January 2005 7-126 DRAFT-DO NOT QUOTE OR CITE
-------
1 However, several other studies, mostly examining the effect on asthmatic children, did not
2 observe a significant relationship (Gouveia and Fletcher, 2000a; Lin et al., 2003; Morgan et al.,
3 1998; Nauenberg and Basu, 1999; Schouten et al., 1996).
4 Acute mortality related to asthma was examined in Barcelona, Spain (Saez et al., 1999;
5 Sunyer et al., 2002). Severe asthmatics with more than one asthma emergency visit showed the
6 strongest mortality associations with air pollutants, NO2 being the most significant predictor
7 followed by O3 (Sunyer et al., 2002).
8 Recent reports from longitudinal cohort studies in California have reported associations
9 between the onset of asthma and long-term O3 exposures (Greer et al., 1993; McConnell et al.,
10 2002; McDonnell et al., 1999). Significant associations were seen in males but not females
11 (Greer et al., 1993; McDonnell et al., 1999). In six high O3 communities, asthma risk was
12 elevated for children who played three or more sports as compared with children who played no
13 sports (McConnell et al., 2002). Playing sports may indicate outdoor activity and an increased
14 ventilation rate which may lead to increased exposure. These outcomes would benefit from
15 replication in other cohorts in regards to indicating weight of a causal interpretation.
16 A few studies provide limited discussion of concentration-response functions and
17 thresholds. In the eight urban areas U.S. study, the odds ratios for incidence of > 10% decline in
18 morning PEF and incidence of morning symptoms when excluding days with 8-h avg O3 greater
19 than 80 ppb were nearly identical to those including data from all days (Mortimer et al., 2002)
20 In the New England asthma panel study (Gent et al., 2003), some of the significant associations
21 for symptoms occurred at 1-h max O3 levels below 60 ppb. In the St. John, Canada study (Stieb
22 et al., 2003), a significant effect of O3 on emergency department visits was reported with
23 evidence of a threshold somewhere in the range below a 1 -h max O3 of 75 ppb in the 15 years
24 and over age group.
25 Overall, asthma subjects have been examined across most health endpoints of interest. The
26 results reported in these studies range from negative to positive estimates with some indicating a
27 significant positive excess risk associated with O3. While no endpoint in itself seems to indicate
28 an unquestionable demonstration of an association, studies with adequate sample size and
29 understandable power consistently provide strong positive and significant results, especially
30 during the summer months when higher O3 levels occur. This view is strengthened as positive
31 results are obtained cohesively across the varied outcomes. Therefore, based on the evidence it
January 2005 7-127 DRAFT-DO NOT QUOTE OR CITE
-------
1 seems prudent to consider asthmatics as a potentially susceptible group that requires protection
2 from O3 exposures.
3 A study by Niedell (2004) examined the relationship between air pollutants and asthma
4 hospitalizations in California. The most recent EPA O3 report (U.S. Environmental Protection
5 Agency, 2004b) indicated that O3 levels in the pacific southwest region had decreased by 9%
6 from 1990 to 2003. This downward trend in O3 levels was mostly influenced by the
7 improvements in Los Angeles and other southern California metropolitan areas. As shown in
8 Figure AX3-60 of the Chapter 3 Annex, O3 concentrations decreased by over 30% in Los
9 Angeles from 1992 to 1998. Results from this study noted declines in levels of air pollutants
10 since 1992 and decreased asthma admissions in 1998 for children aged 1 to 18 years ranging
11 from 5 to 14%, depending on the age group. The greatest decline (> 10%) in air pollution-
12 related asthma admissions was observed among 3 to 12 year old children. Although this benefit
13 analysis was not specific to O3, it provides evidence of decreased morbidity resulting from
14 reduced air pollutant concentrations, including O3. Many studies have reported short-term
15 associations between O3 and morbidity outcomes, yet a largely unaddressed question remains as
16 to the extent to which reductions in ambient O3 actually lead to reductions in adverse health
17 outcomes attributable to O3. This question is not only important in terms of "accountability"
18 from the regulatory point of view, but it is also a scientific question that challenges the predictive
19 validity of statistical models and their underlying assumptions used thus far to estimate excess
20 health effects due to ambient O3.
21
22 7.6.10.2 Age-Related Differences in O3 Effects
23 Several mortality studies have investigated age-related differences in O3 effects. Among
24 the studies that observed a significant association between O3 and mortality, a comparison of all
25 age or younger age (< 65 years of age) O3-mortality risk estimates to that of the elderly
26 population (> 65 years) indicates that, in general, the elderly population is more susceptible to
27 O3 effects (Borja-Aburto et al. 1997; Bremner et al., 1999; Gouveia and Fletcher 2000b; O'Neill
28 et al., 2004; Simpson et al., 1997; Sartor et al., 1995; Sunyer et al., 2002). For example, a study
29 by Gouveia and Fletcher (2000b) examined the O3-mortality effect by age in Sao Paulo, Brazil.
30 There were 151,756 deaths for all non-violent causes over the period of 1991 to 1993, of which
31 49% occurred in the elderly. Among all ages, O3 was associated with a non-significant 0.6%
January 2005 7-128 DRAFT-DO NOT QUOTE OR CITE
-------
1 (95% CI: -0.8, 2.0) excess risk in all cause mortality per 40 ppb increase in 1-h max O3.
2 In comparison, in the elderly population, the O3-mortality risk estimate was nearly three-fold
3 greater, 1.7% (95% CI: 0.0, 3.3). Similarly, a Mexico City study found that O3-mortality risk
4 estimates were 1.3% and 2.8% per 20 ppb increase in 24-h avg O3 concentration in all ages and
5 the elderly, respectively (O'Neill et al., 2004).
6 The large U.S. 95 communities study (Bell et al., 2004) did not find evidence of significant
7 heterogeneity in risk across three age groups, < 65 years, 65 to 74 years, and > 75 years of age.
8 Effect estimates were only slightly higher for those 65 to 74 years, 1.40% excess risk per 20 ppb
9 increase in 24-h avg O3, compared to individuals less than 65 years and 75 years or greater,
10 1.00% and 1.04%, respectively. However, Bell et al. (2004) noted that despite similar effect
11 estimates, the absolute effect of O3 is substantially greater in the elderly population due to the
12 higher underlying mortality rates, which leads to a larger number of extra deaths for the elderly
13 compared to the general population.
14 Few mortality studies examined another potentially susceptible age group, young children
15 under the age of 5 years. The results were mixed, with one Mexico City study showing a lower
16 risk of O3-related all cause mortality in young children compared to all ages and the elderly
17 (Borja-Aburto et al., 1997) and another study showing a greater risk in respiratory mortality in
18 young children compared to the elderly (Gouveia and Fletcher, 2000b). It should be noted that
19 approximately 10% of mortality occurred in young children, thus the statistical power to study
20 the O3 effect in this age group was limited.
21 With respect to age-specificity of associations between O3 and acute respiratory
22 hospitalizations or emergency department visits, no clear pattern emerges from recent studies.
23 Significant associations have been reported for all ages (Anderson et al., 1997; Burnett et al.,
24 1995, 1997b, 1999; Weisel et al., 2002), adults or elderly (Burnett et al., 1997a; Delfino et al.,
25 1997, 1998; Moolgavkar et al., 1997; Schwartz et al., 1996; Yang et al., 2003), and children
26 (Burnett et al., 2001; Gouveia and Fletcher, 2000a; Lin et al., 1999; Ponka and Virtanen, 1996;
27 Tolbert et al., 2000; Yang et al., 2003). Interestingly, studies that have examined effects in
28 multiple age strata often have seen effects only in non-pediatric strata (Delfino et al., 1997,
29 1998; Stieb et al., 1996; Jones et al., 1995). Several studies that focused on children did not
30 report significant O3 effects, though in some cases these studies are limited by small size,
31 inadequate control of seasonal patterns, or very low O3 levels (Lierl and Hornung, 2003; Lin
January 2005 7-129 DRAFT-DO NOT QUOTE OR CITE
-------
1 et al., 2003; Thompson et al., 2001). If O3 is causally related to exacerbations of respiratory
2 diseases leading to hospital usage, one would expect to see effects most prominently among
3 children, for whom asthma is most prevalent and exposures may be greater.
4 Many of the field studies focused on the effect of O3 on the respiratory health of school
5 children, however, none have compared the results from children to that in other age groups.
6 In general, children experienced significant decrements in pulmonary function parameters,
7 including PEF, FEVl9 and FVC (Castillejos et al., 1995; Chen et al., 1999; Gielen et al., 1997;
8 Gold et al., 1999; Jalaludin et al., 2000; Mortimer et al., 2002; Romieu et al., 1996; Thurston
9 et al., 1997), and some experienced increases in respiratory symptoms (Delfmo et al., 2003;
10 Gold et al., 1999; Neas et al., 1995; Romieu et al., 1996, 1997; Thurston et al., 1997) and asthma
11 medication use (Delfmo et al., 1996; Just et al., 2002; Ostro et al., 2001). These respiratory
12 heath effects were observed in both healthy and asthmatic children.
13 Collectively, there is supporting evidence of age-related differences in susceptibility to O3
14 health effects. The elderly population (> 65 years of age) appear to be at increased risk of
15 O3-related mortality and hospitalizations, and children (< 18 years of age) experience other
16 potentially adverse respiratory health outcomes with increased O3 exposure.
17
18 7.6.11 Summary of Key Findings and Conclusions Derived From O3
19 Epidemiologic Studies
20 In the previous 1996 O3 AQCD, there was considerable evidence of O3-related respiratory
21 health effects from individual-level camp and exercise studies, as well as some consistent
22 evidence from time-series studies of emergency room visits and hospitalizations. Since the 1996
23 document, more field studies have been conducted, with some emphasis on additional outcome
24 markers such as respiratory symptoms and asthma medication use. Another significant addition
25 to the current O3 AQCD is the substantial number of short-term O3 mortality studies, which is in
26 part due to the increase in the number of studies that examined PM-mortality associations.
27 Considering the wide variability in possible study designs and statistical model specification
28 choices, the reported O3 risk estimates for the various health outcomes are in reasonably good
29 agreement. In the case of O3-mortality time-series studies, combinations of choices in model
30 specifications (the number of weather terms and degrees of freedom for smoothing of mortality-
31 temporal trends) alone may explain the extent of the difference in O3 risk estimates across
January 2005 7-130 DRAFT-DO NOT QUOTE OR CITE
-------
1 studies. As use of time-series studies to investigate air pollution effects has become more
2 common, there has been a great effort to evaluate the issues surrounding these studies.
3 In this section, conclusions regarding O3 health effects from the epidemiologic evidence
4 and the issues that may affect the interpretation of the effect estimates are briefly summarized.
5 A more integrative synthesis of all relevant information will be presented in Chapter 8 of this
6 AQCD.
7 (1) Field/panel studies of acute O? effects. Results from recent field/panel studies
continue to confirm that short-term O3 exposure is associated with acute
decrements in lung function, increased respiratory symptoms, and increased
medication use, particularly in children and asthmatics. Taken together with the
evidence from controlled human exposure studies, O3 is likely causally related
to the various respiratory health outcomes.
8 (2) O? effects on emergency department visits and hospitalizations. Large multicity
studies, as well as many studies from individual cities have reported a significant
O3 effect on total respiratory, asthma, and COPD hospital visits and admissions.
Studies using year-round data noted some inconsistencies in the O3 effect on daily
emergency department visits and hospitalizations. However, studies with data
restricted to the summer or warm season, in general, indicated positive and robust
associations between ambient O3 concentrations and respiratory morbidity.
9 (3) Acute O? effects on mortality. The majority of the studies suggest an elevated risk
of mortality associated with acute exposure to O3, especially in the summer or
warm season when O3 levels are expected to be high. However, as the magnitude
of the O3-mortality risk estimates are generally small, bias due to the uncertainties
regarding model specification and adjustment for confounding may be of concern.
10 (4) Chronic O? effects on morbidity and mortality. Few studies have investigated the
effect of chronic O3 exposure on morbidity and mortality. The strongest evidence
is for the association between O3 exposure and seasonal decrements or reduced
growth in lung function measures in adults and children. Less conclusive are
longitudinal studies investigating the association of chronic O3 exposure on yearly
lung function, asthma incidence, and respiratory symptoms. Chronic O3-mortality
studies observed inconsistencies across exposure periods, cause-specific mortality
outcomes, and gender. Based on the current evidence, the chronic effect of O3
exposure on morbidity and mortality outcomes is still inconclusive.
11 (5) Exposure assessement. Exposure misclassification may result from the use of
stationary ambient monitors to determine exposure in population studies.
Although central ambient monitors do not explain the variance of individual
personal exposures, significant correlations are found between aggregate personal
O3 measurements and O3 concentrations from ambient monitors. A simulation
study indicated that the use of ambient monitor data will tend to bias effect
estimates towards the null.
January 2005 7-131 DRAFT-DO NOT QUOTE OR CITE
-------
(6) (X exposure indices. The three most commonly used daily O3 exposure indices,
1-h max O3, 8-max O3, and 24-h avg O3, were found to be highly correlated in
studies conducted in various regions. In addition, the effect estimates and
significance of associations across all health outcomes were comparable when
using the same distributional increment for all three indices. The commonly
used 8-h max O3 index, which is also reflective of the new 8-h NAAQS for O3,
continues to be an appropriate choice.
(7) Selection of exposure lag structure. Most studies did not hypothesize a priori the
temporal relationship between O3 exposure and the occurrence of health effects.
Bias can result from the selection of the largest, most significant effect estimates.
However, the majority of the studies found an immediate O3 effect, with health
effects having the strongest associations with exposure on the same day and/or
previous day. Some studies found greater cumulative effects of O3 over longer
lag periods, indicating that multiday lags also should be investigated.
(8) Sensitivity to model specifications for temporal trends. Ozone effect estimates
that were reported in studies whose main focus was PM often were calculated
using the same model specifications as PM. The sensitivity of the O3 risk
estimates to alternative model specifications has not been throughly investigated.
Uncertainty remains regarding the extent of confounding on estimates of O3 health
risks, however limited evidence indicates that O3 effects were robust to various
model specifications for temporal trend adjustment.
(9) Influence of seasonal trends. An evaluation of the confounding effects of
meteorologic factors and copollutants on O3 risk estimates is complicated by their
changing relationship with O3 across seasons. Mortality and morbidity effect
estimates calculated using all year or cool season data are generally smaller than
those from warm season only data. In locations where seasonal variability may be
considerable, efforts should be made to determine season-specific risk estimates.
(10) Confounding by copollutants. Multipollutant models most often are used to
adjust for confounding by copollutants. Results from these analyses indicate that
copollutants generally do not appear to confound the association between O3 and
acute health effects. However, due to the varying concurvity across pollutants,
multipollutant models may not be adequate to determine the independent effects
of individual pollutants. Given the limitations, results generally suggest that the
inclusion of copollutants into the models do not substantially affect O3 risk
estimates.
(11) Model uncertainty and multiple testing. Various statistical methods have been
used to assist model selection. While Bayesian model averaging is a useful tool
that incorporates model uncertainty into the effect estimates, its use may be limited
due to the large number of variables typically considered in air pollution health
effects and the high degree of correlation between the various air pollutants.
January 2005 7-132 DRAFT-DO NOT QUOTE OR CITE
-------
(12) Concentration-response function. Supporting evidence for an effect threshold
is provided by the numerous studies where O3 effects are seen only in the warm
months when O3 levels are higher and more variable. However, in the few
mortality and morbidity studies that have specifically examined the O3
concentration-response relationship, there is conflicting evidence regarding the
presence of an effect threshold. Lack of evidence for a population-level threshold
does not preclude the existence of individual thresholds.
(13) Spatial variability in O? effects. Consistent O3 effect estimates were observed
overall for mortality, hospitalizations, and other respiratory health outcomes in
multicity studies, indicating little heterogeneity of O3 effects by location. The
slight heterogeneity observed may be partially attributable to the differences in
relative personal exposure to O3 and the varying concentration and composition
of copollutants present by region.
(14) O3 health effects in asthmatics. The effect of O3 on asthmatics has been examined
widely in both time-series studies and field panel studies. Across various
respiratory health outcomes, results were consistently positive and, at times,
statistically significant, indicating that asthmatics may be a potentially susceptible
population that requires protection from O3 exposures.
(15) Age-related differences in (X health effects. Supporting evidence exists for
heterogeneity in the effects of O3 by age. The elderly population (> 65 years
of age) appear to be at greater risk of O3-related mortality and hospitalizations
compared to all age or younger populations. In addition, negative respiratory
health outcomes were associated with O3 exposure in children (< 18 years of age).
5
6
January 2005 7-133 DRAFT-DO NOT QUOTE OR CITE
-------
1 REFERENCES
2 Abbey, D. E.; Nishino, N.; McDonnell, W. F.; Burchette, R. J.; Knutsen, S. F.; Beeson, W. L.; Yang, J. X. (1999)
3 Long-term inhalable particles and other air pollutants related to mortality in nonsmokers. Am. J. Respir. Crit.
4 Care Med. 159: 373-382.
5 American Thoracic Society. (1991) Lung function testing: selection of reference values and interpretative strategies
6 Am. Rev. Respir. Dis. 144: 1202-1218.
7 Anderson, H. R.; Ponce de Leon, A.; Bland, J. M; Bower, J. S.; Strachan, D. P. (1996) Air pollution and daily
8 mortality in London: 1987-92. Br. Med. J. 312: 665-669.
9 Anderson, H. R.; Spix, C.; Medina, S.; Schouten, J. P.; Castellsague, J.; Rossi, G.; Zmirou, D.; Touloumi, G.;
10 Wojtyniak, B.; Ponka, A.; Bacharova, L.; Schwartz, J.; Katsouyanni, K. (1997) Air pollution and daily
11 admissions for chronic obstructive pulmonary disease in 6 European cities: results from the APHEA project.
12 Eur. Respir. J. 10: 1064-1071.
13 Anderson, H. R.; Ponce de Leon, A.; Bland, J. M.; Bower, J. S.; Emberlin, J.; Strachen, D. P. (1998) Air pollution,
14 pollens, and daily admissions for asthma in London 1987-92. Thorax 53: 842-848.
15 Atkinson, R. W.; Anderson, H. R.; Strachan, D. P.; Bland, J. M.; Bremner, S. A.; Ponce de Leon, A. (1999a)
16 Short-term associations between outdoor air pollution and visits to accident and emergency departments in
17 London for respiratory complaints. Eur. Respir. J. 13: 257-265.
18 Atkinson, R. W.; Bremner, S. A.; Anderson, H. R.; Strachan, D. P.; Bland, J. M.; Ponce de Leon, A. (1999b)
19 Short-term associations between emergency hospital admissions for respiratory and cardiovascular disease
20 and outdoor air pollution in London. Arch. Environ. Health 54: 3 98-411.
21 Avol, E. L.; Trim, S. C.; Little, D. E.; Spier, C. E.; Smith, M. N.; Peng, R.-C.; Linn, W. S.; Hackney, J. D.; Gross,
22 K. B.; D'Arcy, J. B.; Gibbons, D.; Higgins, I. T. T. (1990) Ozone exposure and lung function in children
23 attending a southern California summer camp. Presented at: 83rd annual meeting and exhibition of the Air &
24 Waste Management Association; June; Pittsburgh, PA. Pittsburgh, PA: Air & Waste Management
25 Association; paper no. 90-150.3.
26 Avol, E. L.; Navidi, W. C.; Rappaport, E. B.; Peters, J. M. (1998) Acute effects of ambient ozone on asthmatic,
27 wheezy, and healthy children. Cambridge, MA: Health Effects Institute; research report no. 82.
28 Ballester, F.; Tenias, J. M.; Perez-Hoyos, S. (2001) Air pollution and emergency hospital admissions for
29 cardiovascular diseases in Valencia, Spain. J. Epidemiol. Community Health 55: 57-65.
30 Bates, D. V.; Sizto, R. (1983) Relationship between air pollutant levels and hospital admissions in Southern Ontario.
31 Can. J. Public Health 74: 117-122.
32 Bates, D. V.; Sizto, R. (1987) Air pollution and hospital admissions in southern Ontario: the acid summer haze
33 effect. Environ. Res. 43: 317-331.
34 Bates, D. V.; Sizto, R. (1989) The Ontario Air Pollution study: identification of the causative agent. Environ. Health
35 Perspect. 79: 69-72.
36 Bates, D. V.; Baker-Anderson, M.; Sizto, R. (1990) Asthma attack periodicity: a study of hospital emergency visits
37 in Vancouver. Environ. Res. 51: 51-70.
38 Beeson, W. L.; Abbey, D. E.; Knutsen, S. F. (1998) Long-term concentrations of ambient air pollutants and incident
39 lung cancer in California adults: results from the AHSMOG study. Environ. Health Perspect. 106: 813-823.
40 Bell, M. L.; McDermott, A.; Zeger, S. L.; Samet, J. M.; Dominici, F. (2004) Ozone and short-term mortality in
41 95 US urban communities, 1987-2000. JAMA J. Am. Med. Assoc. 292: 2372-2378.
42 Borja-Aburto, V H.; Loomis, D. P.; Bangdiwala, S. I.; Shy, C. M.; Rascon-Pacheco, R. A. (1997) Ozone, suspended
43 particulates, and daily mortality in Mexico City. Am. J. Epidemiol. 145: 258-268.
44 Borja-Aburto, V H.; Castillejos, M.; Gold, D. R.; Bierzwinski, S.; Loomis, D. (1998) Mortality and ambient fine
45 particles in southwest Mexico City, 1993-1995. Environ. Health Perspect. 106: 849-855.
46 Bourcier, T.; Viboud, C.; Cohen, J.-C.; Thomas, F.; Bury, T.; Cadiot, L.; Mestre, O.; Flahault, A.; Borderie, V;
47 Laroche, L. (2003) Effects of air pollution and climatic conditions on the frequency of ophthalmological
48 emergency examinations. Br. J. Ophthalmol. 87: 809-811.
49 Brauer, M.; Brook, J. R. (1997) Ozone personal exposures and health effects for selected groups residing in the
50 Fraser Valley. In: Steyn, D. G.; Bottenheim, J. W., eds. The Lower Fraser Valley Oxidants/Pacific '93 Field
51 Study. Atmos. Environ. 31: 2113-2121.
52 Brauer, M.; Blair, J.; Vedal, S. (1996) Effect of ambient ozone exposure on lung function in farm workers. Am. J.
53 Respir. Crit. Care Med. 154: 981-987.
January 2005 7-134 DRAFT-DO NOT QUOTE OR CITE
-------
1 Bremner, S. A.; Anderson, H. R.; Atkinson, R. W.; McMichael, A. J.; Strachan, D. P.; Bland, J. M; Bower, J. S.
2 (1999) Short term associations between outdoor air pollution and mortality in London 1992-4. Occup.
3 Environ. Med. 56: 237-244.
4 Buchdahl, R.; Parker, A.; Stebbings, T.; Babiker, A. (1996) Association between air pollution and acute childhood
5 wheezy episodes: prospective observational study. Br. Med. J. 312: 661-664.
6 Buchdahl, R.; Willems, C. D.; Vander, M.; Babiker, A. (2000) Associations between ambient ozone, hydrocarbons,
7 and childhood wheezy episodes: a prospective observational study in south east London. Occup. Environ.
8 Med. 57: 86-93.
9 Burchfiel, C. M.; Marcus, E. B.; Curb, J. D.; Maclean, C. J.; Vollmer, W. M.; Johnson, L. R.; Fong, K.; Rodriguez,
10 B. L.; Masaki, K. H.; Buist, A. S. (1995) Effects of smoking and smoking cessation on longitudinal decline in
11 pulmonary function. Am. J. Respir. Crit. Care Med. 151: 1778-1785.
12 Burnett, R. T.; Dales, R. E.; Raizenne, M. E.; Krewski, D.; Summers, P. W.; Roberts, G. R.; Raad-Young, M.;
13 Dann, T.; Brook, J. (1994) Effects of low ambient levels of ozone and sulfates on the frequency of respiratory
14 admissions to Ontario hospitals. Environ. Res. 65: 172-194.
15 Burnett, R. T.; Dales, R.; Krewski, D.; Vincent, R.; Dann, T.; Brook, J. R. (1995) Associations between ambient
16 paniculate sulfate and admissions to Ontario hospitals for cardiac and respiratory diseases. Am. J. Epidemiol.
17 142: 15-22.
18 Burnett, R. T.; Brook, J. R.; Yung, W. T.; Dales, R. E.; Krewski, D. (1997a) Association between ozone and
19 hospitalization for respiratory diseases in 16 Canadian cities. Environ. Res. 72: 24-31.
20 Burnett, R. T.; Cakmak, S.; Brook, J. R.; Krewski, D. (1997b) The role of paniculate size and chemistry in the
21 association between summertime ambient air pollution and hospitalization for cardiorespiratory diseases.
22 Environ. Health Perspect. 105: 614-620.
23 Burnett, R. T.; Smith-Doiron, M.; Stieb, D.; Cakmak, S.; Brook, J. R. (1999) Effects of paniculate and gaseous air
24 pollution on cardiorespiratory hospitalizations. Arch. Environ. Health 54: 130-139.
25 Burnett, R. T.; Smith-Doiron, M.; Stieb, D.; Raizenne, M. E.; Brook, J. R.; Dales, R. E.; Leech, J. A.; Cakmak, S.;
26 Krewski, D. (2001) Association between ozone and hospitalization for acute respiratory diseases in children
27 less than 2 years of age. Am. J. Epidemiol. 153: 444-452.
28 Calderon-Garciduenas, L.; Rodriguez-Alcaraz, A.; Garcia, R.; Ramirez, L.; Barragan, G. (1995) Nasal inflammatory
29 responses in children exposed to a polluted urban atmosphere. J. Toxicol. Environ. Health 45: 427-437.
30 Calderon-Garciduenas, L.; Osnaya, N.; Rodriguez-Alcaraz, A.; Villarreal-Calderon, A. (1997) DNA damage in nasal
31 respiratory epithelium from children exposed to urban pollution. Environ. Mol. Mutagen. 30: 11-20.
32 Calderon-Garciduenas, L.; Wen-Wang, L.; Zhang, Y.-J.; Rodriguez-Alcaraz, A.; Osnaya, N.; Villarreal-Calderon,
33 A.; Santella, R. M. (1999) 8-hydroxy-2'-deoxyguanosine, a major mutagenic oxidative DNA lesion, and DNA
34 strand breaks in nasal respiratory epithelium of children exposed to urban pollution. Environ. Health Perspect.
35 107:469-474.
36 Castellsague, J.; Sunyer, J.; Saez, M.; Anto, J. M. (1995) Short-term association between air pollution and
37 emergency room visits for asthma in Barcelona. Thorax 50: 1051-1056.
38 Castillejos, M.; Gold, D. R.; Damokosh, A. I.; Serrano, P.; Allen, G.; McDonnell, W. F.; Dockery, D.; Velasco,
39 S. R.; Hernandez, M.; Hayes, C. (1995) Acute effects of ozone on the pulmonary function of exercising
40 schoolchildren from Mexico City. Am. J. Respir. Crit. Care Med. 152: 1501-1507.
41 Chang, L.-T.; Koutrakis, P.; Catalano, P. J.; Suh, H. H. (2000) Hourly personal exposures to fine particles and
42 gaseous pollutants—results from Baltimore, Maryland. J. Air Waste Manage. Assoc. 50: 1223-1235.
43 Charpin, D.; Pascal, L.; Birnbaum, J.; Armengaud, A.; Sambuc, R.; Lanteaume, A.; Vervloet, D. (1999) Gaseous air
44 pollution and atopy. Clin. Exp. Allergy 29: 1474-1480.
45 Chen, P.-C; Lai, Y.-M.; Wang, J.-D.; Yang, C.-Y.; Hwang, J.-S.; Kuo, H.-W.; Huang, S.-L.; Chan, C.-C. (1998)
46 Adverse effect of air pollution on respiratory health of primary school children in Taiwan. Environ. Health
47 Perspect. 106:331-335.
48 Chen, P.-C.; Lai, Y.-M.; Chan, C.-C.; Hwang, J.-S.; Yang, C.-Y.; Wang, J.-D. (1999) Short-term effect of ozone on
49 the pulmonary function of children in primary school. Environ. Health Perspect. 107: 921-925.
50 Chen, L.; Jennison, B. L.; Yang, W.; Omaye, S. T. (2000) Elementary school absenteeism and air pollution.
51 Inhalation Toxicol. 12: 997-1016.
52 Chen, L.; Yang, W.; Jennison, B. L.; Goodrich, A.; Omaye, S. T. (2002) Air pollution and birth weight in northern
53 Nevada, 1991-1999. Inhalation Toxicol. 14: 141-157.
54 Chew, F. T.; Goh, D. Y. T.; Ooi, B. C.; Saharom, R.; Hui, J. K. S.; Lee, B. W. (1999) Association of ambient
55 air-pollution levels with acute asthma exacerbation among children in Singapore. Allergy (Copenhagen)
56 54:320-329.
January 2005 7-135 DRAFT-DO NOT QUOTE OR CITE
-------
1 Chock, D. P.; Winkler, S. L.; Chen, C. (2000) A study of the association between daily mortality and ambient air
2 pollutant concentrations in Pittsburgh, Pennsylvania. J. Air Waste Manage. Assoc. 50: 1481-1500.
3 Cifuentes, L. A.; Vega, J.; Kopfer, K.; Lave, L. B. (2000) Effect of the fine fraction of paniculate matter versus the
4 coarse mass and other pollutants on daily mortality in Santiago, Chile. J. Air Waste Manage. Assoc. 50:
5 1287-1298.
6 Clyde, M. A. (1999) Bayesian model averaging and model search strategies. In: Bernardo, J. M.; Berger, J. O.;
7 Dawid, A. P.; Smith, A. F. M., eds. Bayesian Statistics 6: proceedings of the Sixth Valencia International
8 Meeting, June; pp. 157-185. Oxford, UK. Oxford, UK: Clarendon Press.
9 Clyde, M. (2000) Model uncertainty and health effect studies for paniculate matter. Environmetrics 11: 745-763.
10 Clyde, M. A.; Guttorp, P.; Sullivan, E. (2000) Effects of ambient fine and coarse particles on mortality in Phoenix,
11 Arizona. Seattle, WA: University of Washington, National Research Center for Statistics and the
12 Environment; NRCSE technical report series, NRCSE-TRS no. 040. Available:
13 http://www.nrcse.washington.edu/pdf/trs40_pm.pdf [18 October, 2004].
14 Cody, R. P.; Weisel, C. P.; Birnbaum, G.; Lioy, P. J. (1992) The effect of ozone associated with summertime
15 photochemical smog on the frequency of asthma visits to hospital emergency departments. Environ. Res.
16 58: 184-194.
17 Cuijpers, C. E. J.; Swaen, G. M. H.; Wesseling, G.; Wouters, E. F. M. (1994) Acute respiratory effects of summer
18 smog in primary school children. Toxicol. Lett. 72: 227-235.
19 Dab, W.; Medina, S.; Quenel, P.; Le Moullec, Y.; Le Tertre, A.; Thelot, B.; Monteil, C.; Lameloise, P.; Pirard, P.;
20 Momas, I.; Ferry, R.; Festy, B. (1996) Short term respiratory health effects of ambient air pollution: results of
21 the APHEA project in Paris. In: St Leger, S., ed. The APHEA project. Short term effects of air pollution on
22 health: a European approach using epidemiological time series data. J. Epidemiol. Commun. Health
23 50(suppl. 1): S42-S46.
24 De Leon, S. F.; Thurston, G. D.; Ito, K. (2003) Contribution of respiratory disease to nonrespiratory mortality
25 associations with air pollution. Am. J. Respir. Crit. CareMed. 167: 1117-1123.
26 Delfino, R. J.; Coate, B. D.; Zeiger, R. S.; Seltzer, J. M.; Street, D. H.; Koutrakis, P. (1996) Daily asthma severity in
27 relation to personal ozone exposure and outdoor fungal spores. Am. J. Respir. Crit. Care Med. 154: 633-641.
28 Delfino, R. J.; Zeiger, R. S.; Seltzer, J. M.; Street, D. H.; Matteucci, R. M.; Anderson, P. R.; Koutrakis, P. (1997a)
29 The effect of outdoor fungal spore concentrations on daily asthma severity. Environ. Health Perspect.
30 105:622-635.
31 Delfino, R. J.; Murphy-Moulton, A. M.; Burnett, R. T.; Brook, J. R.; Becklake, M. R. (1997b) Effects of air
32 pollution on emergency room visits for respiratory illnesses in Montreal, Quebec. Am. J. Respir. Crit. Care
33 Med. 155: 568-576.
34 Delfino, R. J.; Zeiger, R. S.; Seltzer, J. M.; Street, D. H. (1998a) Symptoms inpediatric asthmatics and air pollution:
3 5 differences in effects by symptom severity, anti-inflammatory medication use and paniculate averaging time.
36 Environ. Health Perspect. 106: 751-761.
37 Delfino, R. J.; Murphy-Moulton, A. M.; Becklake, M. R. (1998b) Emergency room visits for respiratory illnesses
3 8 among the elderly in Montreal: association with low level ozone exposure. Environ. Res. 76: 67-77.
39 Delfino, R. J.; Gone, H.; Linn, W. S.; Pellizzari, E. D.; Hu, Y. (2003) Asthma symptoms in Hispanic children and
40 daily ambient exposures to toxic and criteria air pollutants. Environ. Health Perspect. Ill: 647-656.
41 Delfino, R. J., Quintana, P. J. E.; Flora, J.; Gastanaga, V M.; Samimi, B. S.; Kleinman, M. T.; Liu, L.-J. S.;
42 Bufalino, C.; Wu, C.-F.; McLaren, C. E. (2004) Association of FEVj in asthmatic children with personal and
43 microenvironmental exposure to airborne particulate matter. Environ. Health Perspect. 112: 932-941.
44 Desqueyroux, H.; Pujet, J.-C.; Prosper, M.; Squinazi, F.; Momas, I. (2002a) Short-term effects of low-level air
45 pollution on respiratory health of adults suffering from moderate to severe asthma. Environ. Res. A 89: 29-37.
46 Desqueyroux, H.; Pujet, J.-C.; Prosper, M.; Le Moullec, Y.; Momas, I. (2002b) Effects of air pollution on adults
47 with chronic obstructive pulmonary disease. Arch. Environ. Health 57: 554-560.
48 Detels, R.; Tashkin, D. P.; Sayre, J. W.; Rokaw, S. N.; Coulson, A. H.; Massey, F. J., Jr.; Wegman, D. H. (1987)
49 The UCLA population studies of chronic obstructive respiratory disease: 9. lung function changes associated
50 with chronic exposure to photochemical oxidants; a cohort study among never-smokers. Chest 92: 594-603.
51 Devlin, R. B.; McDonnell, W. F.; Mann, R.; Becker, S.; House, D. E.; Schreinemachers, D.; Koren, H. S. (1991)
52 Exposure of humans to ambient levels of ozone for 6.6 hours causes cellular and biochemical changes in the
53 lung. Am. J. Respir. Cell Mol. Biol. 4: 72-81.
54 Diaz, J.; Garcia, R.; Ribera, P.; Alberdi, J. C.; Hernandez, E.; Pajares, M. S.; Otero, A. (1999) Modeling of air
55 pollution and its relationship with mortality and morbidity in Madrid, Spain. Int. Arch. Occup. Environ.
56 Health 72: 366-376.
January 2005 7-136 DRAFT-DO NOT QUOTE OR CITE
-------
1 Dockery, D. W.; Schwartz, I; Spengler, J. D. (1992) Air pollution and daily mortality: associations with particulates
2 and acid aerosols. Environ. Res. 59: 362-373.
3 Dominici, F.; McDermott, A.; Zeger, S. L.; Samet, J. M. (2002) On the use of generalized additive models in
4 time-series studies of air pollution and health. Am. J. Epidemiol. 156: 193-203.
5 Dominici, F.; McDermott, A.; Daniels, M.; Zeger, S. L.; Samet, J. M. (2003) Mortality among residents of 90 cities.
6 In: Revised analyses of time-series studies of air pollution and health. Special report. Boston, MA: Health
7 Effects Institute; pp. 9-24. Available: http://www.healtheffects.org/Pubs/TimeSeries.pdf [12 May, 2004].
8 Fairley, D. (1999) Daily mortality and air pollution in Santa Clara County, California: 1989-1996. Environ. Health
9 Perspect. 107: 637-641.
10 Fairley, D. (2003) Mortality and air pollution for Santa Clara County, California, 1989-1996. In: Revised analyses of
11 time-series studies of air pollution and health. Special report. Boston, MA: Health Effects Institute;
12 pp. 97-106. Available: http://www.healtheffects.org/news.htm [16 May, 2003].
13 Fortoul, T. L; Valverde, M.; Lopez M. D. C.; Bizarro, P.; Lopez, L; Sanchez, I.; Colin-Barenque, L.; Avila-Costa,
14 M. R.; Rojas, E.; Ostrosky-Shejet, P. (2003) Single-cell gel electrophoresis assay of nasal epithelium and
15 leukocytes from asthmatic and nonasthmatic subjects in Mexico City. Arch. Environ. Health 58: 348-352.
16 Friedman, M. S.; Powell, K. E.; Hutwagner, L.; Graham, L. M.; Teague, W. G. (2001) Impact of changes in
17 transportation and commuting behaviors during the 1996 summer Olympic games in Atlanta on air quality and
18 childhood asthma. JAMA J. Am. Med. Assoc. 285: 897-905.
19 Frischer, T. M.; Ktihr, J.; Pullwitt, A.; Meinert, R.; Forster, J.; Studnicka, M.; Koren, H. (1993) Ambient ozone
20 causes upper airways inflammation in children. Am. Rev. Respir. Dis. 148: 961-964.
21 Frischer, T.; Pullwitt, A.; Kuehr, K.; Meinert, R.; Haschke, N.; Studnicka, M.; Lubec, G. (1997) Aromatic
22 hydroxylation in nasal lavage fluid following ambient ozone exposure. Free Radical Biol. Med. 22: 201-207.
23 Frischer, T.; Studnicka, M.; Gartner, C.; Tauber, E.; Horak, F.; Veiter, A.; Spengler, J.; Ktihr, J.; Urbanek, R. (1999)
24 Lung function growth and ambient ozone: a three-year population study in school children. Am. J. Respir.
25 Crit. Care Med. 160: 390-396.
26 Frischer, T.; Studnicka, M.; Halmerbauer, G.; Horak, F.; Gartner, C.; Tauber, E.; Koller, D. Y. (2001) Ambient
27 ozone exposure is associated with eosinophil activation in healthy children. Clin. Exp. Allergy 31: 1213-1219.
28 Fuhlbrigge, A.; Kitch, B.; Paltiel, A. D.; et al. (2001) FEV{ is associated with risk of asthma attacks in a pediatric
29 population. J. Allergy Clin. Immunol. 107: 61-67.
30 Galizia, A.; Kinney, P. L. (1999) Long-term residence in areas of high ozone: associations with respiratory health in
31 a nationwide sample of nonsmoking young adults. Environ. Health Perspect. 107: 675-679.
32 Gamble, J. L. (1998) Effects of ambient air pollution on daily mortality: a time series analysis of Dallas, Texas,
33 1990-1994. Presented at: 91st annual meeting and exhibition of the Air & Waste Management Association;
34 June; San Diego, CA. Pittsburgh, PA: Air & Waste Management Association; paper no. 98-MP26.03.
35 Garcia-Aymerich, J.; Tobias, A.; Anto, J. M.; Sunyer, J. (2000) Air pollution and mortality in a cohort of patients
36 with chronic obstructive pulmonary disease: a time series analysis. J. Epidemiol. Community Health
37 54:73-74.
38 Garty, B. Z.; Kosman, E.; Ganor, E.; Berger, V.; Garty, L.; Wietzen, T.; Waisman, Y.; Mimouni, M.; Waisel, Y.
39 (1998) Emergency room visits of asthmatic children, relation to air pollution, weather, and airborne allergens.
40 Ann. Allergy Asthma Immunol. 81: 563-570.
41 Gauderman, W. J.; McConnell, R.; Gilliland, F.; London, S.; Thomas, D.; Avol, E.; Vora, H.; Berhane, K.;
42 Rappaport, E. B.; Lurmann, F.; Margolis, H. G.; Peters, J. (2000) Association between air pollution and lung
43 function growth in southern California children. Am. J. Respir. Crit. Care Med. 162: 1383-1390.
44 Gauderman, W. J.; Gilliland, G. F.; Vora, H.; Avol, E.; Stram, D.; McConnell, R.; Thomas, D.; Lurmann, F.;
45 Margolis, H. G.; Rappaport, E. B.; Berhane, K.; Peters, J. M. (2002) Association between air pollution and
46 lung function growth in southern California children: results from a second cohort. Am. J. Respir. Crit. Care
47 Med. 166: 76-84.
48 Gauderman, W. J.; Avol, E.; Gilliland, F.; Vora, H.; Thomas, D.; Berhane, K.; McConnell, R.; Kuenzli, N.;
49 Lurmann, F.; Rappaport, E.; Margolis, H.; Bates, D.; Peters, J. (2004a) The effect of air pollution on lung
50 development from 10 to 18 years of age. N. Engl. J. Med. 351: 1057-1067.
51 Gauderman, W. J.; Avol, E.; Gilliland, F. (2004b) Air pollution and lung function [reply letter]. N. Engl. J. Med.
52 351:2653.
53 Gent, J. F.; Triche, E. W.; Holford, T. R.; Belanger, K.; Bracken, M. B.; Beckett, W. S.; Leaderer, B. P. (2003)
54 Association of low-level ozone and fine particles with respiratory symptoms in children with asthma. JAMA
55 J. Am. Med. Assoc. 290: 1859-1867.
56 George, E. I. (1999) Comment on Hoeting et al., 1999, "Bayesian model averaging: a tutorial." Stat. Sci.
57 14:409-412.
January 2005 7-137 DRAFT-DO NOT QUOTE OR CITE
-------
1 Geyh, A. S.; Xue, I; Ozkaynak, H.; Spengler, J. D. (2000) The Harvard Southern California chronic ozone exposure
2 study: assessing ozone exposure of grade-school-age children in two southern California communities.
3 Environ. Health Perspect. 108: 265-270.
4 Gielen, M. H.; VanDer Zee, S. C.; Van Wijnen, J. H.; Van Steen, C. J.; Brunekreef, B. (1997) Acute effects of
5 summer air pollution on respiratory health of asthmatic children. Am. J. Respir. Crit. Care Med.
6 155:2105-2108.
7 Gilliland, F. D.; Berhane, K.; Rappaport, E. B.; Thomas, D. C.; Avol, E.; Gauderman, W. J.; London, S. J.;
8 Margolis, H. G.; McConnell, R.; Islam, K. T.; Peters, J. M. (2001) The effects of ambient air pollution on
9 school absenteeism due to respiratory illnesses. Epidemiology 12: 43-54.
10 Gold, D. R.; Damokosh, A. I.; Pope, C. A., Ill; Dockery, D. W.; McDonnell, W. F.; Serrano, P.; Retama, A.;
11 Castillejos, M. (1999) Paniculate and ozone pollutant effects on the respiratory function of children in
12 southwest Mexico City. Epidemiology 10: 8-16.
13 Gold, D. R.; Litonjua, A.; Schwartz, J.; Lovett, E.; Larson, A.; Nearing, B.; Allen, G.; Verrier, M.; Cherry, R.;
14 Verrier, R. (2000) Ambient pollution and heart rate variability. Circulation 101: 1267-1273.
15 Gold, D. R.; Schwartz, J.; Litonjua, A.; Verrier, R.; Zanobetti, A. (2003) Ambient pollution and reduced heart rate
16 variability. In: Revised analyses of time-series studies of air pollution and health. Special report. Boston, MA:
17 Health Effects Institute; pp. 107-112. Available: http://www.healtheffects.org/Pubs/TimeSeries.pdf
18 [18 October, 2004].
19 Goldberg, M. S.; Burnett, R. T. (2003) Revised analysis of the Montreal time-series study. In: Revised analyses of
20 time-series studies of air pollution and health. Special report. Boston, MA: Health Effects Institute;
21 pp. 113-132. Available: http://www.healtheffects.org/Pubs/TimeSeries.pdf [13 August 2003).
22 Goldberg, M. S.; Burnett, R. T.; Valois, M.-F.; Flegel, K.; Bailar, J. C., Ill; Brook, J.; Vincent, R.; Radon, K. (2003)
23 Associations between ambient air pollution and daily mortality among persons with congestive heart failure.
24 Environ. Res. 91:8-20.
25 Gong, H., Jr.; Wong, R.; Sarma, R. J.; Linn, W. S.; Sullivan, E. D.; Shamoo, D. A.; Anderson, K. R.; Prasad, S. B.
26 (1998a) Cardiovascular effects of ozone exposure in human volunteers. Am. J. Respir. Crit. Care Med.
27 158: 538-546.
28 Gong, H., Jr.; Simmons, M. S.; Linn, W. S.; McDonnell, W. F.; Westerdahl, D. (1998b) Relationship between acute
29 ozone responsiveness and chronic loss of lung function in residents of a high-ozone community.
30 Arch. Environ. Health 53: 313-319.
31 Gonzales, M.; Ngo, L.; Hammond, S. K.; Tager, I. (2003) Validation of a questionnaire and microenvironmental
32 model for estimating past exposures to ozone. Int. J. Environ. Health Res. 13: 249-260.
33 Goss, C. H.; Newsom, S. A.; Schildcrout, J. S.; Sheppard, L.; Kaufman, J. D. (2004) Effect of ambient air pollution
34 on pulmonary exacerbations and lung function in cystic fibrosis. Am. J. Respir. Crit. Care Med. 169: 816-821.
3 5 Gouveia, N.; Fletcher, T. (2000a) Respiratory diseases in children and outdoor air pollution in Sao Paulo, Brazil:
36 a time series analysis. Occup. Environ. Med. 57: 477-483.
37 Gouveia, N.; Fletcher, T. (2000b) Time series analysis of air pollution and mortality: effects by cause, age and
38 socioeconomic status. J. Epidemiol. Community Health 54: 750-755.
39 Gouveia, N.; Bremner, S. A.; Novaes, H. M. D. (2004) Association between ambient air pollution and birth weight
40 in Sao Paulo, Brazil. J. Epidemiol. Community Health 58: 11-17.
41 Greer, J. R.; Abbey, D. E.; Burchette, R. J. (1993) Asthma related to occupational and ambient air pollutants in
42 nonsmokers. J. Occup. Med. 35: 909-915.
43 Gryparis, A.; Forsberg, B.; Katsouyanni, K.; Analitis, A.; Touloumi, G.; Schwartz, J.; Samoli, E.; Medina, S.;
44 Anderson, H. R.; Niciu, E. M.; Wichmann, H.-E.; Kriz, B.; Kosnik, M.; Skorkovsky, J.; Vonk, J. M.;
45 Dortbudak, Z. (2004) Acute effects of ozone on mortality from the "air pollution and health: a European
46 approach" project. Am. J. Respir. Crit. Care Med. 170: 1080-1087.
47 Hagen, J. A.; Nafstad, P.; Skrondal, A.; Bjerkly, S.; Magnus, P. (2000) Associations between outdoor air pollutants
48 and hospitalization for respiratory diseases. Epidemiology 11: 136-140.
49 Health Effects Institute. (2003) Revised analyses of time-series studies of air pollution and health. Boston, MA:
50 Health Effects Institute; special report. Available: Available:
51 http://www.healtheffects.org/Pubs/TimeSeries.pdf [27 June 2003].
52 Hernandez-Garduno, E.; Perez-Neria, J.; Paccagnella, A. M.; Pina-Garcia, M.; Munguia-Castro, M.;
53 Catalan-Vazquez, M.; Rojas-Ramos, M. (1997) Air pollution and respiratory health in Mexico City. J. Occup.
54 Environ. Med. 39: 299-307.
55 Higgins, I. T. T.; D'Arcy, J. B.; Gibbons, D. I.; Avol, E. L.; Gross, K. B. (1990) Effect of exposures to ambient
56 ozone on ventilatory lung function in children. Am. Rev. Respir. Dis. 141: 1136-1146.
57 Hill, A. B. (1965) The environment and disease: association or causation? Proc. R. Soc. Med. 58: 295-300.
January 2005 7-138 DRAFT-DO NOT QUOTE OR CITE
-------
1 Hiltermann, T. J. N.; Stolk, J.; Van der Zee, S. C.; Branekreef, B.; De Bruijne, C. R.; Fischer, P. H.; Ameling, C. B.;
2 Sterk, P. J.; Hiemstra, P. S.; Van Bree, L. (1998) Asthma severity and susceptibility to air pollution. Eur.
3 Respir. J. 11:686-693.
4 Hoek, G. (2003) Daily mortality and air pollution in The Netherlands. In: Revised analyses of time-series studies of
5 air pollution and health. Special report. Boston, MA: Health Effects Institute; pp. 133-142. Available:
6 http://www.healtheffects.org/Pubs/TimeSeries.pdf [12 May, 2004].
7 Hoek, G.; Branekreef, B. (1995) Effect of photochemical air pollution on acute respiratory symptoms in children.
8 Am. J. Respir. Crit. Care Med. 151: 27-32.
9 Hoek, G.; Branekreef, B.; Verhoeff, A.; Van Wijnen, J.; Fischer, P. (2000) Daily mortality and air pollution in the
10 Netherlands. J. Air Waste Manage. Assoc. 50: 1380-1389.
11 Hoek, G.; Branekreef, B.; Fischer, P.; Van Wijnen, J. (2001) The association between air pollution and heart failure,
12 arrhythmia, embolism, thrombosis, and other cardiovascular causes of death in a time series study.
13 Epidemiology 12: 355-357.
14 Hoeting, J. A.; Madigan, D.; Raftery, A. E.; Volinsky, C. T. (1999) Bayesian model averaging: a tutorial. Stat. Sci.
15 14:382-417.
16 Holguin, F.; Tellez-Rojo, M. M.; Hernandez, M.; Cortez, M.; Chow, J. C.; Watson, J. G.; Mannino, D.; Romieu, I.
17 (2003) Air pollution and heart rate variability among the elderly in Mexico City. Epidemiology 14: 521-527.
18 Holmen, A.; Blomqvist, J.; Frindberg, H.; Johnelius, Y.; Eriksson, N. E.; Henricson, K. A.; Herrstrom, P.; Hogstedt,
19 B. (1997) Frequency of patients with acute asthma in relation to ozone, nitrogen dioxide, other pollutants of
20 ambient air and meteorological observations. Int. Arch. Occup. Environ. Health 69: 317-322.
21 Hoppe, P.; Praml, G.; Rabe, G.; Lindner, J.; Frahmann, G.; Kessel, R. (1995a) Environmental ozone field study on
22 pulmonary and subjective responses of assumed risk groups. Environ. Res. 71: 109-121.
23 Hoppe, P.; Lindner, J.; Praml, G.; Bronner, N. (1995b) Effects of environmental ozone on the lung function of senior
24 citizens. Int. J. Biometeorol. 38: 122-125.
25 Horak, F., Jr.; Studnicka, M.; Gartner, C.; Spengler, J. D.; Tauber, E.; Urbanek, R.; Veiter, A.; Frischer, T. (2002a)
26 Paniculate matter and lung function growth in children: a 3-yr follow-up study in Austrian schoolchildren.
27 Eur. Respir. J. 19: 838-845.
28 Horak, F., Jr.; Studnicka, M.; Gartner, C.; Spengler, J. D.; Tauber, E.; Urbanek, R.; Veiter, A.; Frischer, T. (2002b)
29 Paniculate matter and lung function growth in children: a 3-yr follow-up study in Austrian schoolchildren
30 [author response]. Eur. Respir. J. 20: 1355.
31 Huang, Y.; Dominici, F.; Bell, M. L. (2004) Bayesian hierarchical distributed lag models for summer ozone
32 exposure and cardio-respiratory mortality. Environmetrics: in press.
33 Ihorst, G.; Frischer, T.; Horak, F.; Schumacher, M.; Kopp, M.; Forster, J.; Mattes, J.; Kuehr, J. (2004) Long- and
34 medium-term ozone effects on lung growth including a broad spectrum of exposure. Eur. Respir. J.
35 23:292-299.
36 Ilabaca, M.; Olaeta, I.; Campos, E.; Villaire, J.; Tellez-Rojo, M. M.; Romieu, I. (1999) Association between levels
37 of fine paniculate and emergency visits for pneumonia and other respiratory illnesses among children in
38 Santiago, Chile. J. Air Waste Manage. Assoc. 49: 154-163.
39 International Commission on Radiological Protection (ICRP). (1994) Human respiratory tract model for radiological
40 protection: a report of a task group of the International Commission on Radiological Protection. Oxford,
41 United Kingdom: Elsevier Science Ltd. (ICRP publication 66; Annals of the ICRP: v. 24, pp. 1-482).
42 Ito, K. (2003) Associations of paniculate matter components with daily mortality and morbidity in Detroit,
43 Michigan. In: Revised analyses of time-series studies of air pollution and health. Special report. Boston, MA:
44 Health Effects Institute; pp. 143-156. Available: http://www.healtheffects.org/Pubs/TimeSeries.pdf [12 May,
45 2004].
46 Ito, K. (2004) Revised ozone risk estimates for daily mortality in Detroit, Michigan [personal communication with
47 attachments to Jee Young Kim]. New York, NY: New York University School of Medicine, Nelson Institute
48 of Environmental Medicine; October 31.
49 Ito, K.; Thurston, G. D. (1996) Daily PM10/mortality associations: an investigation of at-risk subpopulations.
50 J. Exposure Anal. Environ. Epidemiol. 6: 79-95.
51 Jaffe, D. H.; Singer, M. E.; Rimm, A. A. (2003) Air pollution and emergency department visits for asthma among
52 Ohio Medicaid recipients, 1991-1996. Environ. Res. 91: 21-28.
53 Jalaludin, B. B.; Chey, T.; O'Toole, B. I.; Smith, W. T.; Capon, A. G.; Leeder, S. R. (2000) Acute effects of low
54 levels of ambient ozone on peak expiratory flow rate in a cohort of Australian children. Int. J. Epidemiol.
55 29: 549-557.
January 2005 7-139 DRAFT-DO NOT QUOTE OR CITE
-------
1 Jalaludin, B. B.; O'Toole, B. I.; Leader, S. R. (2004) Acute effects of urban ambient air pollution on respiratory
2 symptoms, asthma medication use, and doctor visits for asthma in a cohort of Australian children. Environ
3 Res. 95: 32-42.
4 Jenkins, P. L.; Phillips, T. I; Mulberg, E. I; Hui, S. P. (1992) Activity patterns of Californians: use of and proximity
5 to indoor pollutant sources. Atmos. Environ. Part A 26: 2141-2148.
6 Jones, G. N.; Sletten, C.; Mandry, C.; Brantley, P. J. (1995) Ozone level effect on respiratory illness: an
7 investigation of emergency department visits. South. Med. J. 88: 1049-1056.
8 Just, J.; Segala, C.; Sahraoui, F.; Priol, G.; Grimfeld, A.; Neukirch, F. (2002) Short-term health effects of paniculate
9 and photochemical air pollution in asthmatic children. Eur. Respir. J. 20: 899-906.
10 Kim, S.-Y.; Lee, J.-T.; Hong, Y.-C; Ahn, K.-J.; Kim, H. (2004) Determining the threshold effect of ozone on daily
11 mortality: an analysis of ozone and mortality in Seoul, Korea, 1995-1999. Environ. Res. 94: 113-119.
12 Kinney, P. L.; Lippmann, M. (2000) Respiratory effects of seasonal exposures to ozone and particles. Arch. Environ.
13 Health 55: 210-216.
14 Kinney, P. L.; Ozkaynak, H. (1991) Associations of daily mortality and air pollution in Los Angeles County.
15 Environ. Res. 54: 99-120.
16 Kinney, P. L.; Ito, K.; Thurston, G. D. (1995) A sensitivity analysis of mortality/PM10 associations in Los Angeles.
17 In: Phalen, R. F.; Bates, D. V., eds. Proceedings of the colloquium on paniculate air pollution and human
18 mortality and morbidity; January 1994; Irvine, CA. Inhalation Toxicol. 7: 59-69.
19 Kinney, P. L.; Nilsen, D. M.; Lippmann, M.; Brescia, M.; Gordon, T.; McGovern, T.; El Fawal, H.; Devlin, R. B.;
20 Rom, W. N. (1996) Biomarkers of lung inflammation in recreational joggers exposed to ozone. Am. J. Respir.
21 Crit. Care Med. 154: 1430-1435.
22 Kinney, P. L.; Aggarwal, M.; Nikiforov, S. V.; Nadas, A. (1998) Methods development for epidemiologic
23 investigations of the health effects of prolonged ozone exposure. Part III: an approach to retrospective
24 estimation of lifetime ozone exposure using a questionnaire and ambient monitoring data (U.S. sites).
25 Cambridge, MA: Health Effects Institute; research report no. 81; pp. 79-108.
26 Klemm, R. J.; Mason, R. M., Jr. (2000) Aerosol Research and Inhalation Epidemiological Study (ARIES):
27 air quality and daily mortality statistical modeling—interim results. J. Air. Waste Manage. Assoc.
28 50: 1433-1439.
29 Klemm, R. J.; Lipfert, F. W.; Wyzga, R. E.; Gust, C. (2004) Daily mortality and air pollution in Atlanta: two years
30 of data from ARIES. Inhalation Toxicol. 16(suppl. 1): 131-141.
31 Klepeis, N. E. (1999) An introduction to the indirect exposure assessment approach: modeling human exposure
32 using microenvironmental measurements and the recent national human activity pattern survey. Environ.
33 HealthPerspect. Suppl. 107(2): 365-374.
3 4 Koop, G.; Tole, L. (2004) Measuring the health effects of air pollution: to what extent can we really say that people
35 are dying from bad air? J. Environ. Econ. Manage. 47: 30-54.
36 Kopp, M. V.; Ulmer, C.; Ihorst, G.; Seydewitz, H. H.; Frischer, T.; Forster, J.; Kuehr, J. (1999) Upper airway
3 7 inflammation in children exposed to ambient ozone and potential signs of adaptation. Eur. Respir. J.
38 14:854-861.
39 Kopp, M. V.; Bohnet, W.; Frischer, T.; Ulmer, C.; Studnicka, M.; Ihorst, G.; Gardner, C.; Forster, J.; Urbanek, R.;
40 Kuehr, J. (2000) Effects of ambient ozone on lung function in children over a two-summer period. Eur.
41 Respir. J. 16: 893-900.
42 Koren, H. S.; Hatch, G. E.; Graham, D. E. (1990) Nasal lavage as a tool in assessing acute inflammation in response
43 to inhaled pollutants. Toxicology 60: 15-25.
44 Korrick, S. A.; Neas, L. M.; Dockery, D. W.; Gold, D. R.; Allen, G. A.; Hill, L. B.; Kimball, K. D.; Rosner, B. A.;
45 Speizer, F. E. (1998) Effects of ozone and other pollutants on the pulmonary function of adult hikers.
46 Environ. Health Perspect. 106: 93-99.
47 Krzyzanowski, M. (1997) Methods for assessing the extent of exposure and effects of air pollution. Occup. Environ.
48 Med. 54: 145-151.
49 Krzyzanowski, M.; Quackenboss, J. J.; Lebowitz, M. D. (1992) Relation of peak expiratory flow rates and symptoms
50 to ambient ozone. Arch. Environ. Health 47: 107-115.
51 Ktinzli, N.; Lurmann, F.; Segal, M.; Ngo, L.; Balmes, J.; Tager, I. B. (1997) Association between lifetime ambient
52 ozone exposure and pulmonary function in college freshmen—results of a pilot study. Environ. Res. 72: 8-23.
53 Kuo, H. W.; Lai, J. S.; Lee, M. C.; Tai, R. C.; Lee, M. C. (2002) Respiratory effects of air pollutants among
54 asthmatics in central Taiwan. Arch. Environ. Health 57: 194-200.
January 2005 7-140 DRAFT-DO NOT QUOTE OR CITE
-------
1 Lagerkvist, B. I; Bernard, A.; Blomberg, A.; Bergstrom, E.; Forsberg, B.; Holmstrom, K.; Karp, K.; Lundstrom,
2 N.-G.; Segerstedt, B.; Svensson, M; Nordberg, G. (2004) Pulmonary epithelial integrity in children:
3 relationship to ambient ozone exposure and swimming pool attendance. Environ. Health Perspect.
4 112: 1768-1771.
5 Langstaff, J. (2003) Percentiles of 1996-2000 ozone concentrations [memorandum to Joe Pinto]. Research Triangle
6 Park, NC: U.S. Environmental Protection Agency, Office of Air Quality Planning and Standards; September
7 17.
8 Lebowitz, M. D.; Camilli, A. E.; Bronnimann, D.; Quackenboss, J. (1987) The significance and meaningfulness of
9 intraindividual changes in objective test results as responses to air contaminants. Presented at: 80th annual
10 meeting of the Air Pollution Control Association; June; New York, NY. Pittsburgh, PA: Air Pollution Control
11 Association; paper no. 87-32.1.
12 Lebowitz, M. D.; Quackenboss, J. J.; Krzyzanowski, M. (1991) Acute respiratory effects of prolonged ambient
13 ozone. In: Berglund, R. L.; Lawson, D. R.; McKee, D. J., eds. Tropospheric ozone and the environment:
14 papers from an international conference; March 1990; Los Angeles, CA. Pittsburgh, PA: Air & Waste
15 Management Association; pp. 111-119. (A&WMA transactions series no. TR-19).
16 Lee, J. -T.; Schwartz, J. (1999) Reanaly sis of the effects of air pollution on daily mortality in Seoul, Korea:
17 a case-crossover design. Environ. Health Perspect. 107: 633-636.
18 Lee, J.-T.; Shin, D.; Chung, Y. (1999) Air pollution and daily mortality in Seoul and Ulsan, Korea. Environ. Health
19 Perspect. 107: 149-154.
20 Lee, K.; Parkhurst, W. J.; Xue, J.; Ozkaynak, H.; Neuberg, D.; Spengler, J. D. (2004) Outdoor/indoor/presonal ozone
21 exposures of children in Nashville, Tennessee. J. Air Waste Manage. Assoc. 54: 352-359.
22 Liao, D.; Duan, Y.; Whitsel, E. A.; Zheng, Z.-J.; Heiss, G.; Chinchilli, V. M.; Lin, H.-M. (2004) Association of
23 higher levels of ambient criteria pollutants with impaired cardiac autonomic control: a population-based
24 study. Am. J. Epidemiol. 159: 768-777.
25 Liard, R.; Zureik, M.; Le Moullec, Y.; Soussan, D.; Glorian, M.; Grimfeld, A.; Neukirch, F. (1999) Use of personal
26 passive samplers for measurement of NO2, NO, and O3 levels in panel studies. Environ. Res. 81: 339-348.
27 Lierl, M. B.; Hornung, R. W. (2003) Relationship of outdoor air quality to pediatric asthma exacerbations. Ann.
28 Allergy Asthma Immunol. 90: 28-33.
29 Lin, C. A.; Martins, M. A.; Farhat, S. C. L.; Pope, C. A., Ill; Conceicao, G. M. S.; Anastacio, V. M.; Hatanaka, M.;
30 Andrade, W. C.; Hamaue, W. R.; Bohm, G. M.; Saldiva, P. H. N. (1999) Air pollution and respiratory illness
31 of children in Sao Paulo, Brazil. Paediatr. Perinat. Epidemiol. 13: 475-488.
32 Lin, M.; Chen, Y.; Burnett, R. T.; Villeneuve, P. J.; Krewski, D. (2003) Effect of short-term exposure to gaseous
3 3 pollution on asthma hospitalisation in children: a bi-directional case-crossover analysis. J. Epidemiol.
34 Community Health 57: 50-55.
35 Lin, M.; Chen, Y.; Villeneuve, P. J.; Burnett, R. T.; Lemyre, L.; Hertzman, C.; McGrail, K. M.; Krewski, D. (2004)
3 6 Gaseous air pollutants and asthma hospitalization of children with low household income in Vancouver,
37 British Columbia, Canada. Am. J. Epidemiol. 159: 294-303.
38 Linn, W. S.; Shamoo, D. A.; Anderson, K. R.; Peng, R.-C.; Avol, E. L.; Hackney, J. D.; Gong, H., Jr. (1996)
39 Short-term air pollution exposures and responses in Los Angeles area schoolchildren. J. Exposure Anal.
40 Environ. Epidemiol. 6: 449-472.
41 Linn, W. S.; Szlachcic, Y.; Gong, H., Jr.; Kinney, P. L.; Berhane, K. T. (2000) Air pollution and daily hospital
42 admissions in metropolitan Los Angeles. Environ. Health Perspect. 108: 427-434.
43 Lipfert, F. W.; Hammerstrom, T. (1992) Temporal patterns in air pollution and hospital admissions. Environ. Res.
44 59: 374-399.
45 Lipfert, F. W.; Morris, S. C.; Wyzga, R. E. (2000a) Daily mortality in the Philadelphia metropolitan area and
46 size-classified paniculate matter. J. Air Waste Manage. Assoc. 50: 1501-1513.
47 Lipfert, F. W.; Perry, H. M., Jr.; Miller, J. P.; Baty, J. D.; Wyzga, R. E.; Carmody, S. E. (2000b) The Washington
48 University-EPRI veterans' cohort mortality study: preliminary results. In: Grant, L. D., ed. PM2000:
49 paniculate matter and health. Inhalation Toxicol. 12(suppl. 4): 41-73.
50 Lipfert, F. W.; Perry, H. M., Jr.; Miller, J. P.; Baty, J. D.; Wyzga, R. E.; Carmody, S. E. (2003) Air pollution, blood
51 pressure, and their long-term associations with mortality. Inhalation Toxicol. 15: 493-512.
52 Lippmann, M. (1988) Health significance of pulmonary function responses to airborne irritants. JAPCA
53 38:881-887.
54 Lippmann, M.; Ito, K.; Nadas, A.; Burnett, R. T. (2000) Association of paniculate matter components with daily
55 mortality and morbidity in urban populations. Cambridge, MA: Health Effects Institute; research report
56 no. 95.
January 2005 7-141 DRAFT-DO NOT QUOTE OR CITE
-------
1 Lipsett, M; Hurley, S.; Ostro, B. (1997) Air pollution and emergency room visits for asthma in Santa Clara County,
2 California. Environ. Health Perspect. 105: 216-222.
3 Liu, L.-J. S.; Koutrakis, P.; Leech, I; Broder, I. (1995) Assessment of ozone exposures in the greater metropolitan
4 Toronto area. J. Air Waste Manage. Assoc. 45: 223-234.
5 Liu, L.-J. S.; Delfino, R.; Koutrakis, P. (1997) Ozone exposure assessment in a southern California community.
6 Environ. Health Perspect. 105: 58-65.
7 Loomis, D. P.; Borja-Aburto, V. H.; Bangdiwala, S. I.; Shy, C. M. (1996) Ozone exposure and daily mortality in
8 Mexico City: a time-series analysis. Cambridge, MA: Health Effects Institute; research report no. 75.
9 Luginaah, I. N.; Fung, K. Y.; Gorey, K. M.; Webster, G.; Wills, C. (2004) Association of Ambient Air Pollution
10 with Respiratory Hospitalization in a Government Designated 'Area of Concern': The Case of Windsor,
11 Ontario. Environ. Health Perspect. 10.1289/ehp.7300. Available: http://dx.doi.org/ (14 December 2004).
12 Lumley, T.; Sheppard, L. (2000) Assessing seasonal confounding and model selection bias in air pollution
13 epidemiology using positive and negative control analyses. Environmetrics 11: 705-717.
14 Mann, J. K.; Tager, I. B.; Lurmann, F.; Segal, M.; Quesenberry, C. P., Jr.; Lugg, M. M.; Shan, J.; Van denEeden,
15 S. K. (2002) Air pollution and hospital admissions for ischemic heart disease in persons with congestive heart
16 failure or arrhythmia. Environ. Health Perspect. 110: 1247-1252.
17 McConnell, R.; Berhane, K.; Gilliland, F.; London, S. J.; Vora, H.; Avol, E.; Gauderman, W. J.; Margolis, H. G.;
18 Lurmann, F.; Thomas, D. C.; Peters, J. M. (1999) Air pollution and bronchitic symptoms in southern
19 California children with asthma. Environ. Health Perspect. 107: 757-760.
20 McConnell, R.; Berhane, K.; Gilliland, F.; London, S. J.; Islam, T.; Gauderman, W. J.; Avol, E.; Margolis, H. G.;
21 Peters, J. M. (2002) Asthma in exercising children exposed to ozone: a cohort study. Lancet 359: 386-391.
22 McDonnell, W. F.; Abbey, D. E.; Nishino, N.; Lebowitz, M. D. (1999) Long-term ambient ozone concentration and
23 the incidence of asthma in nonsmoking adults: the ahsmog study. Environ. Res. 80: 110-121.
24 Metzger, K. B.; Tolbert, P. E.; Klein, M.; Peel, J. L.; Flanders, W. D.; Todd, K. H.; Mulholland, J. A.; Ryan, P. B.;
25 Frumkin, H. (2004) Ambient air pollution and cardiovascular emergency department visits. Epidemiology
26 15:46-56.
27 Mitchell, H.; Senturia, Y.; Gergen, P.; Baker, D.; Joseph, C.; McNiff-Mortimer, K.; Wedner, H. J.; Crain, E.;
28 Eggleston, P.; Evans, R., Ill; Kattan, M.; Kercsmar, C.; Leickly, F.; Malveaux, F.; Smartt, E.; Weiss, K.
29 (1997) Design and methods of the National Cooperative Inner-City Asthma Study. Pediatr. Pulmonol.
30 24:237-252.
31 Moolgavkar, S. H.; Luebeck, E. G. (1996) A critical review of the evidence on paniculate air pollution and
32 mortality. Epidemiology 7: 420-428.
33 Moolgavkar, S. H.; Luebeck, E. G.; Hall, T. A.; Anderson, E. L. (1995) Air pollution and daily mortality in
34 Philadelphia. Epidemiology 6: 476-484.
35 Moolgavkar, S. H.; Luebeck, E. G.; Anderson, E. L. (1997) Air pollution and hospital admissions for respiratory
36 causes in Minneapolis-St. Paul and Birmingham. Epidemiology 8: 364-370.
37 Morgan, G.; Corbett, S.; Wlodarczyk, J. (1998a) Air pollution and hospital admissions in Sydney, Australia, 1990 to
38 1994. Am. J. Public Health 88: 1761-1766.
39 Morgan, G.; Corbett, S.; Wlodarczyk, J.; Lewis, P. (1998b) Air pollution and daily mortality in Sydney, Australia,
40 1989 through 1993. Am. J. Public Health 88: 759-764.
41 Mortimer, K. M.; Tager, I. B.; Dockery, D. W.; Neas, L. M.; Redline, S. (2000) The effect of ozone on inner-city
42 children with asthma: identification of susceptible subgroups. Am. J. Respir. Crit. Care Med. 162: 1838-1845.
43 Mortimer, K. M.; Neas, L. M.; Dockery, D. W.; Redline, S.; Tager, I. B. (2002) The effect of air pollution on
44 inner-city children with asthma. Eur. Respir. J. 19: 699-705.
45 Naeher, L. P.; Holford, T. R.; Beckett, W. S.; Belanger, K.; Triche, E. W.; Bracken, M. B.; Leaderer, B. P. (1999)
46 Healthy women's PEF variations with ambient summer concentrations of PM10, PN2 5, SO42", H+, and O3.
47 Am. J. Respir. Crit. Care Med. 160: 117-125.
48 Nauenberg, E.; Basu, K. (1999) Effect of insurance coverage on the relationship between asthma hospitalizations
49 and exposure to air pollution. Public Health Rep. 114: 135-148.
50 Navidi, W.; Thomas, D.; Langholz, B.; Stram, D. (1999) Statistical methods for epidemiologic studies of the health
51 effects of air pollution. Cambridge, MA: Health Effects Institute; research report no. 86.
52 Neas, L. M.; Dockery, D. W.; Koutrakis, P.; Tollerud, D. J.; Speizer, F. E. (1995) The association of ambient air
53 pollution with twice daily peak expiratory flow rate measurements in children. Am. J. Epidemiol.
54 141: 111-122.
55 Neas, L. M.; Dockery, D. W.; Koutrakis, P.; Speizer, F. E. (1999) Fine particles and peak flow in children: acidity
56 versus mass. Epidemiology 10: 550-553.
January 2005 7-142 DRAFT-DO NOT QUOTE OR CITE
-------
1 Neidell, M. J. (2004) Air pollution, health, and socio-economic status: the effect of outdoor air quality on childhood
2 asthma. J. Health Econ. 23: 1209-1236.
3 Newhouse, C. P.; Levetin, B. S.; Levetin, E. (2004) Correlation of environmental factors with asthma and rhinitis
4 symptoms in Tulsa, OK. Ann. Allergy Asthma Immunol. 92: 356-366.
5 Nutman, A.; Solomon, Y.; Mendel, S.; Nutman, J.; Hines, E.; Topilsky, M.; Kivity, S. (1998) The use of a neural
6 network for studying the relationship between air pollution and asthma-related emergency room visits. Respir.
7 Med. 92: 1199-1202.
8 Oftedal, B.; Nafstad, P.; Magnus, P.; Bjerkly, S.; Skrondal, A. (2003) Traffic related air pollution and acute hospital
9 admission for respiratory diseases inDrammen, Norway 1995-2000. Eur. J. Epidemiol. 18: 671-675.
10 O'Neill, M. S.; Ramirez-Aguilar, M.; Meneses-Gonzalez, F.; Hernandez-Avila, M.; Geyh, A. S.; Sienra-Monge, J. J.;
11 Romieu, I. (2003) Ozone exposure among Mexico City outdoor workers. J. Air Waste Manage. Assoc.
12 53: 339-346.
13 O'Neill, M. S.; Loomis, D.; Borja-Aburto, V. H. (2004) Ozone, area social conditions, and mortality in Mexico City.
14 Environ. Res. 94: 234-242.
15 Ostro, B. (1995) Fine paniculate air pollution and mortality in two Southern California counties. Environ. Res.
16 70:98-104.
17 Ostro, B.; Sanchez, J. M.; Aranda, C.; Eskeland, G. S. (1996) Air pollution and mortality: results from a study of
18 Santiago, Chile. In: Lippmann, M., ed. Papers from the ISEA-ISEE annual meeting; September 1994;
19 Research Triangle Park, NC. J. Exposure Anal. Environ. Epidemiol. 6: 97-114.
20 Ostro, B.; Lipsett, M.; Mann, J.; Braxton-Owens, H.; White, M. (2001) Air pollution and exacerbation of asthma in
21 African-American children in Los Angeles. Epidemiology 12: 200-208.
22 Palli, D.; Saieva, C.; Grechi, D.; Masala, G.; Zanna, I.; Barbara, A.; Decarli, A.; Munnia, A.; Peluso, M. (2004)
23 DNA bulky adducts in a Mediterranean population correlate with environmental ozone concentration, an
24 indicator of photochemical smog. Int. J. Cancer 109: 17-23.
25 Park, S. K.; O'Neill, M. S.; Vokonas, P. S.; Sparrow, D.; Schwartz, J. (2004) Effects of air pollution on heart rate
26 variability: the VA normative aging study. Environ. Health Perspect: in press, 10.1289/ehp.7447. Available:
27 http://dx.doi.org/ [6 December, 2004].
28 Pearce, N.; Beasley, R.; Burgess, C.; Crane, J. (1998) Asthma epidemiology: principles and methods. New York,
29 NY: Oxford University Press.
30 Peel, J. L.; Tolbert, P. E.; Klein, M.; Metzger, K. B.; Planers, W. D.; Knox, T.; Mulholland, J. A.; Ryan, P. B.;
31 Frumkin, H. (2004) Ambient air pollution and respiratory emergency department visits. Epidemiology: in
32 press.
33 Pereira, L. A. A.; Loomis, D.; Conceicao, G. M. S.; Braga, A. L. F.; Areas, R. M.; Kishi, H. S.; Singer, J. M.; Bohm,
34 G. M.; Saldiva, P. H. N. (1998) Association between air pollution and intrauterine mortality in Sao Paulo,
35 Brazil. Environ. Health Perspect. 106: 325-329.
36 Peters, J. M.; Avol, E.; Navidi, W.; London, S. J.; Gauderman, W. J.; Lurmann, F.; Linn, W. S.; Margolis, H.;
37 Rappaport, E.; Gong, H., Jr.; Thomas, D. C. (1999a) A study of twelve southern California communities with
3 8 differing levels and types of air pollution. I. Prevalence of respiratory morbidity. Am. J. Respir. Crit. Care
39 Med. 159: 760-767.
40 Peters, J. M.; Avol, E.; Gauderman, W. J.; Linn, W. S.; Navidi, W.; London, S. J.; Margolis, H.; Rappaport, E.;
41 Vora, H.; Gong, H., Jr.; Thomas, D. C. (1999b) A study of twelve southern California communities with
42 differing levels and types of air pollution. II. Effects on pulmonary function. Am. J. Respir. Crit. Care Med.
43 159:768-775.
44 Peters, A.; Liu, E.; Verrier, R. L.; Schwartz, J.; Gold, D. R.; Mittleman, M.; Baliff, J.; Oh, J. A.; Allen, G.;
45 Monahan, K.; Dockery, D. W. (2000a) Air pollution and incidence of cardiac arrhythmia. Epidemiology
46 11: 11-17.
47 Peters, A.; Skorkovsky, J.; Kotesovec, F.; Brynda, J.; Spix, C.; Wichmann, H. E.; Heinrich, J. (2000b) Associations
48 between mortality and air pollution in central Europe. Environ. Health Perspect. 108: 283-287.
49 Peters, A.; Dockery, D. W.; Muller, J. E.; Mittleman, M. A. (2001) Increased paniculate air pollution and the
50 triggering of myocardial infarction. Circulation 103: 2810-2815.
51 Petroeschevsky, A.; Simpson, R. W.; Thalib, L.; Rutherford, S. (2001) Associations between outdoor air pollution
52 and hospital admissions in Brisbane, Australia. Arch. Environ. Health 56: 37-52.
53 Ponce de Leon, A.; Anderson, H. R.; Bland, J. M.; Strachan, D. P.; Bower, J. (1996) Effects of air pollution on daily
54 hospital admissions for respiratory disease in London between 1987-88 and 1991-92. In: StLeger, S., ed.
5 5 The APHEA project. Short term effects of air pollution on health: a European approach using epidemiological
56 time series data. J. Epidemiol. Commun. Health 50(suppl. 1): S63-S70.
January 2005 7-143 DRAFT-DO NOT QUOTE OR CITE
-------
1 Ponka, A.; Virtanen, M. (1996) Asthma and ambient air pollution in Helsinki. In: St Leger, S., ed. The APHEA
2 project. Short term effects of air pollution on health: a European approach using epidemiological time series
3 data. J. Epidemiol. Community Health 50(suppl. 1): S59-S62.
4 Ponka, A.; Savela, M.; Virtanen, M. (1998) Mortality and air pollution in Helsinki. Arch. Environ. Health
5 53:281-286.
6 Pope, C. A., Ill; Burnett, R. T.; Thun, M. J.; Calle, E. E.; Krewski, D.; Ito, K.; Thurston, G. D. (2002) Lung cancer,
7 cardiopulmonary mortality, and long-term exposure to fine paniculate air pollution. JAMA J. Am. Med.
8 Assoc. 287: 1132-1141.
9 Prescott, G. J.; Cohen, G. R.; Elton, R. A.; Fowkes, F. G. R.; Agius, R. M. (1998) Urban air pollution and
10 cardiopulmonary ill health: a 14.5 year time series study. Occup. Environ. Med. 55: 697-704.
11 Raizenne, M.; Stern, B.; Burnett, R.; Spengler, J. (1987) Acute respiratory function and transported air pollutants:
12 observational studies. Presented at: 80th annual meeting of the Air Pollution Control Association; June;
13 New York, NY. Pittsburgh, PA: Air Pollution Control Association; paper no. 87-32.6.
14 Raizenne, M. E.; Burnett, R. T.; Stern, B.; Franklin, C. A.; Spengler, J. D. (1989) Acute lung function responses to
15 ambient acid aerosol exposures in children. Environ. Health Perspect. 79: 179-185.
16 Ramadour, M.; Burel, C.; Lanteaume, A.; Vervloet, D.; Charpin, D.; Brisse, F.; Dutau, H.; Charpin, D. (2000)
17 Prevalence of asthma and rhinitis in relation to long-term exposure to gaseous air pollutants. Allergy
18 (Copenhagen) 55: 1163-1169.
19 Ramsay, T. O.; Burnett, R. T.; Krewski, D. (2003) The effect of concurvity in generalized additive models linking
20 mortality to ambient paniculate matter. Epidemiology 14: 18-23.
21 Rich, K. E.; Petkau, J.; Vedal, S.;Brauer,M. (2004) A case-crossover analysis of paniculate air pollution and
22 cardiac arrhythmia in patients with implantable cardioverter defibrillators. Inhalation Toxicol. 16: 363-372.
23 Ritz, B.; Yu, F.; Chapa, G.; Fruin, S. (2000) Effect of air pollution on preterm birth among children born in Southern
24 California between 1989 and 1993. Epidemiology 11: 502-511.
25 Roemer, W. H.; Van Wijnen, J. H . (2001) Daily mortality and air pollution along busy streets in Amsterdam,
26 1987-1998. Epidemiology 12: 649-653.
27 Romieu, I.; Meneses, F.; Sienra-Monge, J. J. L.; Huerta, J.; Velasco, S. R.; White, M. C.; Etzel, R. A.;
28 Hernandez-Avila, M. (1995) Effects of urban air pollutants on emergency visits for childhood asthma in
29 Mexico City. Am. J. Epidemiol. 141: 546-553.
30 Romieu, L; Meneses, F.; Ruiz, S.; Sienra, J. J.; Huerta, J.; White, M. C.; Etzel, R. A. (1996) Effects of air pollution
31 on the respiratory health of asthmatic children living in Mexico City. Am. J. Respir. Crit. Care Med.
32 154:300-307.
33 Romieu, L; Meneses, F.; Ruiz, S.; Huerta, J.; Sienra, J. J.; White, M.; Etzel, R.; Hernandez, M. (1997) Effects of
34 intermittent ozone exposure on peak expiratory flow and respiratory symptoms among asthmatic children in
35 Mexico City. Arch. Environ. Health 52: 368-376.
36 Romieu, L; Meneses, F.; Ramirez, M.; Ruiz, S.; Padilla, R. P.; Sienra, J. J.; Gerber, M.; Grievink, L.; Dekker, R.;
37 Walda, L; Brunekreef, B. (1998) Antioxidant supplementation and respiratory functions among workers
3 8 exposed to high levels of ozone. Am. J. Respir. Crit. Care Med. 158: 226-232.
39 Romieu, L; Sienra-Monge, J. J.; Ramirez-Aguilar, M.; Tellez-Rojo, M. M.; Moreno-Macias, H.; Reyes-Ruiz, N. L;
40 Del Rio-Navarro, B. E.; Ruiz-Navarro, M. X.; Hatch, G.; Slade, R.; Hernandez-Avila, M. (2002) Antioxidant
41 supplementation and lung functions among children with asthma exposed to high levels of air pollutants. Am.
42 J. Respir. Crit. Care Med. 166: 703-709.
43 Romieu, L; Sienra-Monge, J. J.; Ramirez-Aguilar, M.; Moreno-Macias, H.; Reyes-Ruiz, N. L; Estela del
44 Rio-Navarro, B.; Hernandez-Avila, M.; London, S. J. (2004) Genetic polymorphism of GSTM1 and
45 antioxidant supplementation influence lung function in relation to ozone exposure in asthmatic children in
46 Mexico City. Thorax 59: 8-10.
47 Ross, M. A.; Persky, V. W.; Scheff, P. A.; Chung, J.; Curtis, L.; Ramakrishnan, V.; Wadden, R. A.; Hryhorczuk,
48 D. O. (2002) Effect of ozone and aeroallergens on the respiratory health of asthmatics. Arch. Environ. Health
49 57: 568-578.
50 Rothman, K. J.; Greenland, S., eds. (1998) Modern epidemiology. 2nd ed. Philadelphia, PA: Lippincott-Raven
51 Publishers.
52 Saez, M.; Tobias, A.; Munoz, P.; Campbell, M. J. (1999) A GEE moving average analysis of the relationship
53 between air pollution and mortality for asthma in Barcelona, Spain. Stat. Med. 18: 2077-2086.
54 Saldiva, P. H. N.; Lichtenfels, A. J. F. C.; Paiva, P. S. O.; Barone, I. A.; Martins, M. A.; Massad, E.; Pereira, J. C.
55 R.; Xavier, V. P.; Singer, J. M.; Bohm, G. M. (1994) Association between air pollution and mortality due to
56 respiratory diseases in children in Sao Paulo, Brazil: a preliminary report. Environ. Res. 65: 218-225.
January 2005 7-144 DRAFT-DO NOT QUOTE OR CITE
-------
1 Saldiva, P. H. N.; Pope, C. A., Ill; Schwartz, I; Dockery, D. W.; Lichtenfels, A. I; Salge, J. M; Barone, I.; Bohm,
2 G. M. (1995) Air pollution and mortality in elderly people: a time-series study in Sao Paulo, Brazil. Arch.
3 Environ. Health 50: 159-163.
4 Samet, J. M.; Zeger, S. L.; Dominici, F.; Curriero, F.; Coursac, I.; Dockery, D. W.; Schwartz, J.; Zanobetti, A.
5 (2000) The national morbidity, mortality, and air pollution study. Part II: morbidity, mortality, and air
6 pollution in the United States. Cambridge, MA: Health Effects Institute; research report no. 94, part II.
7 Sarnat, J. A.; Schwartz, J.; Catalano, P. J.; Suh, H. H. (2001) Gaseous pollutants in paniculate matter epidemiology:
8 confounders or surrogates? Environ. Health Perspect. 109: 1053-1061.
9 Sartor, F.; Snacken, R.; Demuth, C.; Walckiers, D. (1995) Temperature, ambient ozone levels, and mortality during
10 summer, 1994, in Belgium. Environ. Res. 70: 105-113.
11 Scarlett, J. F.; Abbott, K. J.; Peacock, J. L.; Strachan, D. P.; Anderson, H. R. (1996) Acute effects of summer air
12 pollution on respiratory function in primary school children in southern England. Thorax 51:1109-1114.
13 Schouten, J. P.; Vonk, J. M.; de Graaf, A. (1996) Short term effects of air pollution on emergency hospital
14 admissions for respiratory disease: results of the APHEA project in two major cities in The Netherlands,
15 1977-89. In: St Leger, S., ed. The APHEA project. Short term effects of air pollution on health: a European
16 approach using epidemiological time series data. J. Epidemiol. Community Health 50(suppl. 1): S22-S29.
17 Schwartz, J. (1991) Paniculate air pollution and daily mortality in Detroit. Environ. Res. 56: 204-213.
18 Schwartz, J. (1993) Air pollution and daily mortality in Birmingham, Alabama. Am. J. Epidemiol. 137: 1136-1147.
19 Schwartz, J. (2004) How sensitive is the association between ozone and daily deaths to control for temperature?
20 Am. J. Respir. Crit. Care Med.: in press, 10.1164/rccm.200407-933OC.
21 Schwartz, J.; Spix, C.; Touloumi, G.; Bacharova, L.; Barumamdzadeh, T.; le Tertre, A.; Piekarksi, T.; Ponce de
22 Leon, A.; Ponka, A.; Rossi, G.; Saez, M.; Schouten, J. P. (1996) Methodological issues in studies of air
23 pollution and daily counts of deaths or hospital admissions. In: St Leger, S., ed. The APHEA project. Short
24 term effects of air pollution on health: a European approach using epidemiological time series data.
25 J. Epidemiol. Commun. Health 50(suppl. 1): S3-S11.
26 Selwyn, B. J.; Stock, T. H.; Hardy, R. J.; Chan, F. A.; Jenkins, D. E.; Kotchmar, D. J.; Chapman, R. S. (1985)
27 Health effects of ambient ozone exposure in vigorously exercising adults. In: Lee, S. D., ed. Evaluation of the
28 scientific basis for ozone/oxidants standards: proceedings of an APCA international specialty conference;
29 November 1984; Houston, TX. Pittsburgh, PA: Air Pollution Control Association; pp. 281-296. (APCA
30 international specialty conference transactions: TR-4).
31 Sheppard, L. (2003) Ambient air pollution and nonelderly asthma hospital admissions in Seattle, Washington,
32 1987-1994. In: Revised analyses of time-series studies of air pollution and health. Special report. Boston,
33 MA: Health Effects Institute; pp. 227-230. Available: http://www.healtheffects.org/Pubs/TimeSeries.pdf
34 [18 October, 2004].
35 Sheppard, L.; Levy, D.; Norris, G.; Larson, T. V.; Koenig, J. Q. (1999) Effects of ambient air pollution on
36 nonelderly asthma hospital admissions in Seattle, Washington, 1987-1994. Epidemiology 10: 23-30.
37 Sherwin, R. P.; Richters, V.; Kraft, P.; Richters, A. (2000) Centriacinar region inflammatory disease in young
38 individuals: a comparative study of Miami and Los Angeles residents. Virchows Arch. 437: 422-428.
39 Shumway, R. H.; Azari, A. S.; Pawitan, Y. (1988) Modeling mortality fluctuations in Los Angeles as functions of
40 pollution and weather effects. Environ. Res. 45: 224-241.
41 Silverman, B. W. (1986) Density estimation. London, England: Chapman & Hall.
42 Simpson, R. W.; Williams, G.; Petroeschevsky, A.; Morgan, G.; Rutherford, S. (1997) Associations between outdoor
43 air pollution and daily mortality in Brisbane, Australia. Arch. Environ. Health 52: 442-454.
44 Spektor, D. M.; Lippmann, M.; Lioy, P. J.; Thurston, G. D.; Citak, K.; James, D. J.; Bock, N.; Speizer, F. E.;
45 Hayes, C. (1988a) Effects of ambient ozone on respiratory function in active, normal children. Am. Rev.
46 Respir. Dis. 137:313-320.
47 Spektor, D. M.; Lippmann, M.; Thurston, G. D.; Lioy, P. J.; Stecko, J.; O'Connor, G.; Garshick, E.; Speizer, F. E.;
48 Hayes, C. (1988b) Effects of ambient ozone on respiratory function in healthy adults exercising outdoors.
49 Am. Rev. Respir. Dis. 138: 821-828.
50 Spektor, D. M.; Lippmann, M. (1991) Health effects of ambient ozone on healthy children at a summer camp. In:
51 Berglund, R. L.; Lawson, D. R.; McKee, D. J., eds. Tropospheric ozone and the environment: papers from an
52 international conference; March 1990; Los Angeles, CA. Pittsburgh, PA: Air & Waste Management
53 Association; pp. 83-89. (A&WMA transactions series no. TR-19).
54 Stieb, D. M.; Burnett, R. T.; Beveridge, R. C.; Brook, J. R. (1996) Association between ozone and asthma
55 emergency department visits in Saint John, New Brunswick, Canada. Environ. Health Perspect.
56 104: 1354-1360.
January 2005 7-145 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 and mortality:
2 effects of gases and particles and the influence of cause of death, age, and season. J. Air Waste Manage.
3 Assoc. 52: 470-484.
4 Stieb, D. M.; Judek, S.; Burnett, R. T. (2003) Meta-analysis of time-series studies of air pollution and mortality:
5 update in relation to the use of generalized additive models. J. Air Waste Manage. 53: 258-261.
6 Sunyer, J.; Basagana, X. (2001) Particles, and not gases, are associated with the risk of death in patients with chronic
7 obstructive pulmonary disease. Int. J. Epidemiol. 30: 1138-1140.
8 Sunyer, J.; Castellsague, J.; Saez, M.; Tobias, A.; Anto, J. M. (1996) Air pollution and mortality in Barcelona. In:
9 St Leger, S., ed. The APHEA project. Short term effects of air pollution on health: a European approach using
10 epidemiological time series data. J. Epidemiol. Community Health 50(suppl. 1): S76-S80.
11 Sunyer, J.; Basagana, X.; Belmonte, J.; Anto, J. M. (2002) Effect of nitrogen dioxide and ozone on the risk of dying
12 in patients with severe asthma. Thorax 57: 687-693.
13 Superko, H. R.; Adams, W. C.; Daly, P. W. (1984) Effects of ozone inhalation during exercise in selected patients
14 with heart disease. Am. J. Med. 77: 463-470.
15 Tager, I. B. (1993) Introduction to working group on tropospheric ozone, Health Effects Institute Environmental
16 Epidemiology Planning Project. Environ. Health Perspect. 101(suppl. 4): 205-207.
17 Tager, I. B. (1999) Air pollution and lung function growth. Is it ozone? [editorial]. Am. J. Respir. Crit. Care Med.
18 160:387-389.
19 Tager, I. B.; Kunzli, N.; Lurmann, F.; Ngo, L.; Segal, M.; Balmes, J. (1998) Methods development for
20 epidemiologic investigations of the health effects of prolonged ozone exposure. Part II: an approach to
21 retrospective estimation of lifetime ozone exposure using a questionnaire and ambient monitoring data
22 (California sites). Cambridge, MA: Health Effects Institute; research report no. 81; pp. 27-78.
23 Taggart, S. C. O.; Custovic, A.; Francis, H. C.; Faragher, E. B.; Yates, C. J.; Higgins, B. G.; Woodcock, A. (1996)
24 Asthmatic bronchial hyperresponsiveness varies with ambient levels of summertime air pollution. Eur.
25 Respir. J. 9: 1146-1154.
26 Tellez-Rojo, M. M.; Romieu, L; Ruiz-Velasco, S.; Lezana, M.-A.; Hernandez-Avila, M. M. (2000) Daily respiratory
27 mortality and PMio pollution in Mexico City: importance of considering place of death. Eur. Respir. J.
28 16:391-396.
29 Tenias, J. M.; Ballester, F.; Rivera, M. L. (1998) Association between hospital emergency visits for asthma and air
30 pollution in Valencia, Spain. Occup. Environ. Med. 55: 541-547.
31 Tenias, J. M.; Ballester, F.; Perez-Hoyos, S.; Rivera, M. L. (2002) Air pollution and hospital emergency room
32 admissions for chronic obstructive pulmonary disease in Valencia, Spain. Arch. Environ. Health 57: 41-47.
33 Thompson, A. J.; Shields, M. D.; Patterson, C. C. (2001) Acute asthma exacerbations and air pollutants in children
34 living in Belfast, Northern Ireland. Arch. Environ. Health 56: 234-241.
35 Thurston, G. D.; Ito, K. (2001) Epidemiological studies of acute ozone exposures and mortality. J. Exposure Anal.
36 Environ. Epidemiol. 11: 286-294.
37 Thurston, G. D.; Ito, K.; Kinney, P. L.; Lippmann, M. (1992) A multi-year study of air pollution and respiratory
3 8 hospital admissions in three New York State metropolitan areas: results for 1988 and 1989 summers.
39 J. Exposure Anal. Environ. Epidemiol. 2: 429-450.
40 Thurston, G. D.; Ito, K.; Hayes, C. G.; Bates, D. V.; Lippmann, M. (1994) Respiratory hospital admissions and
41 summertime haze air pollution in Toronto, Ontario: consideration of the role of acid aerosols. Environ. Res.
42 65:271-290.
43 Thurston, G. D.; Lippmann, M.; Scott, M. B.; Fine, J. M. (1997) Summertime haze air pollution and children with
44 asthma. Am. J. Respir. Crit. Care Med. 155: 654-660.
45 Tobias, A.; Campbell, M. J.; Saez, M. (1999) Modelling asthma epidemics on the relationship between air pollution
46 and asthma emergency visits in Barcelona, Spain. Eur. J. Epidemiol. 15: 799-803.
47 Tolbert, P. E.; Mulholland, J. A.; Macintosh, D. L.; Xu, F.; Daniels, D.; Devine, O. J.; Carlin, B. P.; Klein, M.;
48 Dorley, J.; Butler, A. J.; Nordenberg, D. F.; Frumkin, H.; Ryan, P. B.; White, M. C. (2000) Air quality and
49 pediatric emergency room visits for asthma in Atlanta, Georgia. Am. J. Epidemiol. 151: 798-810.
50 Touloumi, G.; Katsouyanni, K.; Zmirou, D.; Schwartz, J.; Spix, C.; Ponce de Leon, A.; Tobias, A.; Quennel, P.;
51 Rabczenko, D.; Bacharova, L.; Bisanti, L.; Vonk, J. M.; Ponka, A. (1997) Short-term effects of ambient
52 oxidant exposure on mortality: a combined analysis within the APHEA project. Am. J. Epidemiol.
53 146: 177-185.
54 Tsai, S.-S.; Huang, C.-H.; Goggins, W. B.; Wu, T.-N.; Yang, C.-Y. (2003) Relationship between air pollution and
55 daily mortality in a tropical city: Kaohsiung, Taiwan. J. Toxicol. Environ. Health Part A 66: 1341-1349.
56 Tyler, W. S.; Tyler, N. K.; Last, J. A.; Gillespie, M. J.; Barstow, T. J. (1988) Comparison of daily and seasonal
57 exposures of young monkeys to ozone. Toxicology 50: 131-144.
January 2005 7-146 DRAFT-DO NOT QUOTE OR CITE
-------
1 U.S. Environmental Protection Agency. (1996a) Air quality criteria for ozone and related photochemical oxidants.
2 Research Triangle Park, NC: Office of Research and Development; report nos. EPA/600/AP-93/004aF-cF. 3v.
3 Available from: NTIS, Springfield, VA; PB96-185582, PB96-185590, and PB96-185608. Available:
4 www.epa.gov/ncea/ozone.htm.
5 U.S. Environmental Protection Agency. (1996b) Air quality criteria for paniculate matter. Research Triangle Park,
6 NC: National Center for Environmental Assessment-RTF Office; report nos. EPA/600/P-95/001aF-cF. 3v.
7 U.S. Environmental Protection Agency. (2004a) Air quality criteria for paniculate matter. Research Triangle Park,
8 NC: National Center for Environmental Assessment; report no. EPA/600/P-99/002aF-bF. 2v. Available:
9 http://cfpub.epa.gov/ncea/ [9 November, 2004].
10 U.S. Environmental Protection Agency. (2004b) The ozone report: measuring progress through 2003. Cincinnati,
11 OH: U.S. Environmental Protection Agency; report no. EPA454/K-04-001.
12 Ulmer, C.; Kopp, M; Ihorst, G.; Frischer, T.; Forster, I; Kuehr, J. (1997) Effects of ambient ozone exposures during
13 the spring and summer of 1994 on pulmonary function of schoolchildren. Pediatr. Pulmonol. 23: 344-353.
14 Vedal, S.; Brauer, M.; White, R.; Petkau, J. (2003) Air pollution and daily mortality in a city with low levels of
15 pollution. Environ. Health Perspect. 111:45-51.
16 Vedal, S.; Rich, K.; Brauer, M.; White, R.; Petkau, J. (2004) Air pollution and cardiac arrhythmias in patients with
17 implantable cardiovascular defibrillators. Inhalation Toxicol. 16: 353-362.
18 Verhoeff, A. P.; Hoek, G.; Schwartz, J.; Van Wijnen, J. H. (1996) Air pollution and daily mortality in Amsterdam.
19 Epidemiology 7: 225-230.
20 Virnig, B. A.; McBean, M. (2001) Administrative data for public health surveillance and planning. Annu. Rev.
21 Public Health 22: 213-230.
22 Weisel, C. P.; Cody, R. P.; Lioy, P. J. (1995) Relationship between summertime ambient ozone levels and
23 emergency department visits for asthma in central New Jersey. Environ. Health Perspect.
24 103(suppl. 2): 97-102.
25 Weisel, C. P.; Cody, R. P.; Georgopoulos, P. G.; Purushothaman, V; Weiss, S. H.; Bielory, L.; Gregory, P.; Stern,
26 A. H. (2002) Concepts in developing health-based indicators for ozone. Int. Arch. Occup. Environ. Health
27 75:415-422.
28 White, M. C.; Etzel, R. A.; Wilcox, W. D.; Lloyd, C. (1994) Exacerbations of childhood asthma and ozone pollution
29 in Atlanta. Environ. Res. 65: 56-68.
30 Wong, T. W.; Lau, T. S.; Yu, T. S.; Neller, A.; Wong, S. L.; Tarn, W.; Pang, S. W. (1999a) Air pollution and
31 hospital admissions for respiratory and cardiovascular diseases in Hong Kong. Occup. Environ. Med.
32 56:679-683.
33 Wong, C.-M; Ma, S.; Hedley, A. J. Lam, T.-H. (1999b) Does ozone have any effect on daily hospital admissions for
34 circulatory diseases? J. Epidemiol. Community Health 53: 580-581.
3 5 World Health Organization. (2004) Meta-analysis of time-series studies and panel studies of paniculate matter (PM)
36 and ozone (O3): report of a WHO task group. Copenhagen, Denmark: WHO Regional Office for Europe;
37 document no. EUR/04/5042688. Available: http://www.euro.who.int/document/E82792.pdf
38 [18 November, 2004].
39 Yang, Q.; Chen, Y.; Shi, Y.; Burnett, R. T.; McGrail, K. M.; Krewski, D. (2003) Association between ozone and
40 respiratory admissions among children and the elderly in Vancouver, Canada. Inhalation Toxicol.
41 15: 1297-1308.
42 Yang, C.-Y.; Chang, C.-C.; Chuang, H.-Y.; Tsai, S.-S.; Wu, T.-N.; Ho, C.-K. (2004) Relationship between air
43 pollution and daily mortality in a subtropical city: Taipei, Taiwan. Environ. Int. 30: 519-523.
44 Zeger, S. L.; Thomas, D.; Dominici, F.; Samet, J. M.; Schwartz, J.; Dockery, D.; Cohen, A. (2000) Exposure
45 measurement error in time-series studies of air pollution: concepts and consequences. Environ. Health
46 Perspect. 108: 419-426.
47 Zhu, L.; Carlin, B. P.; Gelfand, A. E. (2003) Hierarchical regression with misaligned spatial data: relating ambient
48 ozone and pediatric asthma ER visits in Atlanta. Environmetrics 14: 537-557.
49 Zmirou, D.; Barumandzadeh, T.; Balducci, F.; Ritter, P.; Laham, G.; Ghilardi, J.-P. (1996) Short term effects of air
50 pollution on mortality in the city of Lyon, France, 1985-90. In: St Leger, S., ed. The APHEA project. Short
51 term effects of air pollution on health: a European approach using epidemiological time series data.
52 J. Epidemiol. Community Health 50(suppl. 1): S30-S35.
53
January 2005 7-147 DRAFT-DO NOT QUOTE OR CITE
-------
CHAPTER 7 ANNEX
EPIDEMIOLOGICAL STUDIES OF HUMAN
HEALTH EFFECTS ASSOCIATED WITH AMBIENT
OZONE EXPOSURE
January 2005 AX7-1 DRAFT-DO NOT QUOTE OR CITE
-------
Table AX7-1. Effects of Acute O3 Exposure on Lung Function and Respiratory Symptoms in Field Studies
to
o
o
Reference, Study Copollutants
Location and Period Outcomes and Methods Mean O3 Levels Considered
Findings, Interpretation
Effects
X
to
United States
Mortimer et al. (2002)
Eight urban areas in
the U.S.:
St. Louis, MO;
Chicago, IL;
Detroit, MI;
Cleveland, OH;
Washington, DC;
Baltimore, MD;
East Harlem, NY;
Bronx, NY
Jun-Aug 1993
Examined 846 asthmatic
children aged 4-9 years for
O3 exposure effects on PEF
and morning symptoms
using linear mixed effect
models and GEE.
8-h avg O3
(10 a.m.-6 p.m.):
44ppb
PM10,N02,
S02
No associations were seen between single
or multiday O3 measures and any evening
outcome measure. The effects of O3 on
morning outcomes increased over several
days with the strongest associations seen
for multiday lags. Joint modeling of O3
with NO2 or SO2 resulted in slightly
reduced estimates for each pollutant.
8-h avg O3 (per 15 ppb):
% change in morning PEF:
Lag 1-5:
All areas: -0.59% (-1.05,-0.13)
St. Louis: -0.86% (-2.10, 0.38)
Chicago: -0.62% (-2.41,1.16)
Detroit: -0.75% (-2.36, 0.86)
Cleveland: -0.62% (-2.23, 0.99)
Washington, DC: -0.54% (-2.02,
0.93)
Baltimore: 0.24% (-0.95, 1.43)
EastHarlem: -0.73% (-1.63, 0.17)
Bronx: -0.69% (-1.54, 0.15)
Odds ratios:
Morning symptoms:
Lag 1-4:
All areas: 1.16(1.02,1.30)
St. Louis: 0.82 (0.59, 1.14)
Chicago: 1.09(0.69,1.72)
Detroit: 1.72(1.12,2.64)
Cleveland: 1.20(0.81,1.79)
Washington, DC: 1.11 (0.72, 1.72)
Baltimore: 1.19(0.89,1.60)
EastHarlem: 1.22(0.97,1.53)
Bronx: 1.23(0.98,1.54)
-------
Table AX7-1 (cont'd). Effects of Acute O3 Exposure on Lung Function and Respiratory Symptoms in Field Studies
to
o
o
Reference, Study
Location and Period
Copollutants
Outcomes and Methods Mean O3 Levels Considered
Findings, Interpretation
Effects
X
United States (cont'd)
Mortimer et al. (2000)
Eight urban areas in the
U.S.:
St. Louis, MO;
Chicago, IL;
Detroit, MI;
Cleveland, OH;
Washington, DC;
Baltimore, MD;
East Harlem, NY;
Bronx, NY
Jun-Aug 1993
Avoletal. (1998)
Southern California
communities
Spring-summer 1994
A cohort of 846 asthmatic
children aged 4-9 years
examined for effects of
summer O3 exposure on
PEF and morning
symptoms. Two subgroups
were compared: (l)low
birth weight or premature
and (2) normal birth weight
or full-term. Analysis using
GEE and linear mixed
models.
Three panels of children
(age 10-12 years):
(1) asthmatic (n= 53);
(2) wheezy (n = 54); and
(3) healthy (n = 103).
Examined for symptoms,
medication use, outdoor
time, physical activity, and
pulmonary function
measures in relation to O3
exposure, via logistic
regression and GLM.
8-h avg O3
(10 a.m.-6 p.m.):
48ppb
None
Stratified
analysis of low
and high 24-h
avg O3:
Fixed site O3:
Low: < 100 ppb
High: > 100 ppb
Personal O3:
Low: < 15.6 ppb
High: > 32.4 ppb
None
Low birth weight and premature
asthmatic children had greater declines in
PEF and higher incidence of morning
symptoms than normal birth weight and
full-term asthmatic children.
The three groups responded similarly.
Few pulmonary function or symptom
associations. Asthmatic children had
the most trouble breathing, the most
wheezing, and the most inhaler use on
high O3 day sin the spring. Ozone levels
were considered too low during the
period of the study. Noncompliance
by subjects may have been a problem.
Other analysis methods may have been
more appropriate.
8-h avg O3 (per 15 ppb):
% change in morning PEF:
Low birth weight:
Lag 1-5: -1.83%(-2.65,-1.01)
Normal birth weight:
Lag 1-5: -0.30% (-0.79, 0.19)
Interaction term for birth weight,
p<0.05
Odds ratios:
Morning symptoms:
Low birth weight:
Lag 1-4: 1.42(1.10,1.82)
Normal birth weight:
Lag 1-4: 1.09(0.95,1.24)
Interaction term for birth weight,
p<0.05
Multiple endpoints analyzed. Few
consistent or statistically significant
responses to O3 exposure reported.
-------
Table AX7-1 (cont'd). Effects of Acute O3 Exposure on Lung Function and Respiratory Symptoms in Field Studies
to
o
o
Reference, Study
Location and Period
Copollutants
Outcomes and Methods Mean O3 Levels Considered
Findings, Interpretation
Effects
United States (cont'd)
X
Gillilandetal. (2001)
12 Southern California
communities
Jan-Jun 1996
Linn etal. (1996)
Three towns in
California: Rubidoux,
Upland, Torrance
Fall-spring 1992-1993
and 1993-1994
1,933 4th grade children
(age 9-10 years) followed
for school absences. Each
absence classified as illness-
related or not. Among
former, classified into
respiratory
or gastrointestinal.
Respiratory absences further
classified into upper or
lower. Pollution measured
in central site in each town.
Analysis of distributed lag
effects controlling for time,
day of week, and
temperature in a Poisson
model.
269 school children (age
unspecified), each followed
for moming/aftemoon lung
function and symptoms for
one week in fall, winter, and
spring over 2 school years.
Personal exposure
monitoring in a subset.
Analyzed afternoon
symptoms versus same day
pollution and morning
symptoms versus 1-day
lag pollution.
8-h avg O3
(10 a.m.-6 p.m.):
Levels not
reported.
PM10,N02
24-h avg O3:
Personal:
5ppb
SD3
Central site:
23ppb
SD12
PM2,,N02
O3 strongly associated with illness-related
and respiratory absences. PM10 only
associated with upper respiratory
absences. Long distributed lag effects
for O3 raise questions about adequacy
of control for seasonal changes.
Central site O3 correlated with personal
exposures, r= 0.61. Ozone effects
observed on lung function but only
significant for FEV[ in one analysis.
No effects on symptoms. Ozone effects
were not robust to NO2 or PM2 5. Power
may have been limited by short followup
within seasons (limiting both person-days
and variability in exposures).
8-h avg O3 (per 20 ppb):
% change in absences:
All illness:
62.9% (18.4, 124.1)
Nonrespiratory illnesses:
37.3% (5.7, 78.3)
Respiratory illnesses:
82.9% (3.9, 222.0)
Upper respiratory:
45.1% (21.3, 73.7)
Lower respiratory with wet cough:
173.9% (91.3, 292.3)
Change in lung function (per ppb):
FEV[ next morning:
-0.26 mL (SE 0.25), p = 0.30
FEV; afternoon:
-0.18 mL(SE 0.26), p = 0.49
FEV[ crossday difference:
-0.58 mL(SE 0.23), p = 0.01
FVC next morning:
-0.21 mL (SE 0.22),p = 0.34
FVC afternoon:
-0.20 mL(SE 0.29), p = 0.48
FVC crossday difference:
-0.25 mL (SE 0.25), p = 0.32
-------
Table AX7-1 (cont'd). Effects of Acute O3 Exposure on Lung Function and Respiratory Symptoms in Field Studies
to
o
o
Reference, Study
Location and Period
Copollutants
Outcomes and Methods Mean O3 Levels Considered
Findings, Interpretation
Effects
X
United States (cont'd)
Ostroetal. (2001)
Central Los Angeles
and Pasadena, CA
Aug-Octl993
138 African-American
children aged 8-13 years
with doctor diagnosed
asthma requiring
medication in past year
followed for daily
respiratory symptoms and
medication use. Lags of
0 to 3 days examined.
l-hmaxO3: PM10,NO2, Correlation between PM10 and O3 was
pollen, mold r = 0.35. Significant O3 effect seen for
Los Angeles: extra medication use (above normal use).
59.5 ppb No O3 effect on symptoms in expected
SD 31.4 direction observed. Inverse association
seen for cough. PM10 effects seen at a lag
Pasadena: of 3 days. Time factors not explicitly
95.8 ppb controlled in analysis; may have led to
SD 49.0 confounding of O3 effects.
1-h max O3 (per 40 ppb):
Odds ratios:
Extra medication use:
Lagl: 1.15(1.12,1.19)
Respiratory symptoms:
Shortness of breath:
Lag 3: 1.01 (0.92,1.10)
Wheeze:
Lag 3: 0.94(0.88,1.00)
Cough:
Lag 3: 0.93(0.87,0.99)
Delfino et al. (2003)
Los Angeles, CA
Nov 1999- Jan 2000
Delfino etal. (1997a)
Alpine, CA
May-Augl994
A panel study of
22 Hispanic children with
asthma aged 10-16 years.
Filled out symptom diaries
in relation to pollutant
levels. Analysis using GEE
model.
22 asthmatics aged
9-46 years followed for
respiratory symptoms,
morning-afternoon PEF,
and P2 agonist inhaler use.
Personal O3 measured for
12 hours/day using passive
monitors. GLM mixed
model.
l-hmaxO3:
25.4 ppb
SD9.6
Ambient:
12-h avg O3
(8 a.m.-8 p.m.):
64 ppb
SD17
Personal:
12-h avg O3:
(8 a.m.-8 p.m.)
18 ppb
SD14
NO2, SO2, Support the view that air toxics in
CO, volatile the pollutant mix from traffic may
organic have adverse effects on asthma in
compounds, children.
PM10
PM10, pollen, No O3 effects observed.
fungi
1-h max O3 (per 14.0 ppb):
Odds ratio:
Symptoms interfering with
daily activities:
LagO: 1.99(1.06,3.72)
No quantitative results for O3.
-------
Table AX7-1 (cont'd). Effects of Acute O3 Exposure on Lung Function and Respiratory Symptoms in Field Studies
to
o
o
Reference, Study
Location and Period
Copollutants
Outcomes and Methods Mean O3 Levels Considered
Findings, Interpretation
Effects
X
United States (cont'd)
Delfinoetal. (1998a)
Alpine, CA
Aug-Octl995
Delfino et al. (2004)
Alpine, CA
Aug-Oct 1999, Apr-Jun
2000
A panel of 24 asthmatics
aged 9-17 years followed
for daily symptoms.
Analysis using GEE model.
1 9 asthmatic children (age
9- 1 7 years) followed daily
for 2 weeks to determine
relationship between air
pollutants, namely PM, and
FEVj. Linear mixed model
used for analysis.
l-hmaxO3:
90 ppb
SD 1 8
8-h max O3 :
62. 8 ppb
SD 15.1
IQR 22.0
PM10
PM2 5, PM10,
NO2
Asthma symptoms were significantly
associated with both ambient O3 and
PM10 in single-pollutant models. Ozone
effects generally robust to PM10. Current
day O3 effects strongest in asthmatics not
on anti-inflammatory medication. Effects
of O3 and PM10 were largely independent.
The largest effects for PM10 were seen for
a 5-day distributed lag. For O3 effects,
there were no lag day effects; current day
results showed the greatest effect.
Significant declines in FEV; associated
with various PM indices (personal, indoor
home, etc.), but not ambient O3 levels.
1-h max O3 (per 58 ppb):
Odds ratios:
O3 only model:
LagO: 1.54(1.02,2.33)
O3 with PM10 model:
LagO: 1.46(0.93,2.29)
No quantitative results for O3.
Delfinoetal. (1996)
San Diego, CA
Sep-Octl993
12 well-characterized
moderate asthmatics aged
9-16 years (7 males,
5 females) followed over
6 weeks for medication use
and respiratory symptoms.
Allergy measured at
baseline with skin prick
tests. Personal O3
measured with passive
badge. Analysis with
GLM mixed model.
Ambient: PM25, SO42~, No effect of ambient O3 on symptom
1-h max O3: FT, HNO3, score. Personal O3 significant for
68 ppb pollen, fungal symptoms, but effect disappeared when
SD 30 spores confounding day of week effect was
controlled with weekend dummy
Ambient: variable. (32 inhaler used among
12-h avg O3: 7 subjects was significantly related to
43 ppb personal O3. Results of this small study
SD 17 suggest the value of personal exposure
data in providing more accurate estimates
Personal: of exposures. However, nearly 50% of
12-h avg O3: personal O3 measurements were below
11.6 ppb limits of detection, diminishing value of
SD 11.2 these data. Pollen and fine particulate
(low levels) were not associated with
any of the outcomes.
Change in (32-agonist inhaler use
(per ppb personal O3):
0.0152 puffs/day (SE 0.0075),
p = 0.04
-------
Table AX7-1 (cont'd). Effects of Acute O3 Exposure on Lung Function and Respiratory Symptoms in Field Studies
to
o
o
Reference, Study
Location and Period
Copollutants
Outcomes and Methods Mean O3 Levels Considered
Findings, Interpretation
Effects
X
United States (cont'd)
Chen et al. (2000)
Washoe County, NV
1996-1998
School absenteeism
examined among 27,793
students (kindergarten to
6th grade) from 57
elementary schools.
First-order autoregression
models used to assess
relationship between O3 and
school absenteeism after
adjusting for weather, day
of week, month, holidays,
and time trends. Ozone
levels from the current day,
and cumulative lags of
1-14 days, 1-21 days, and
1-28 days examined.
l-hmaxO3: PM10, CO Multipollutant models were examined.
37.45 ppb Ozone concentrations in the preceding
SD 13.37 14 days were significantly associated
with school absenteeism for students
in grades 1 through 6, but not those in
kindergarten. Both PM10 and CO
concentrations on the concurrent day
were associated with school absenteeism,
but the estimate for PM10 was a negative
value.
1-h max O3 (per 50 ppb):
Total absence rate:
O3 with PM10 and CO model:
Lag 1-14: 3.79% (1.04, 6.55)
Newhouse et al. (2004)
Tulsa, OK
Sep-Oct2000
24 subjects aged 9-64 years
with physician diagnosis of
asthma. Performed PEF
twice daily (morning and
afternoon), and reported
daily respiratory symptoms
and medication use.
Forward stepwise multiple
regression models and
Pearson correlation
analyses.
24-h avg O3: PM2 5, CO, Among ambient air pollutants, O3 seemed
30 ppb SO2, pollen, to be most significant factor. Morning
Range 10-70 fungal spores PEF values significantly associated with
average and maximum O3 levels on the
previous day. Individual symptoms,
including wheezing, headache, and
fatigue, also significantly related to
average and maximum daily O3. Multiple
regression analyses produced complex
models with different predictor variables
for each symptom.
Pearson correlation coefficient:
Morning PEF:
Mean O3 levels:
Lagl: -0.274, p< 0.05
Maximum O3 levels:
Lagl: -0.2W,p<0.05
-------
Table AX7-1 (cont'd). Effects of Acute O3 Exposure on Lung Function and Respiratory Symptoms in Field Studies
to
o
o
Reference, Study
Location and Period
Copollutants
Outcomes and Methods Mean O3 Levels Considered
Findings, Interpretation
Effects
X
oo
United States (cont'd)
Ross et al. (2002)
East Moline, IL and
nearby communities
May-Octl994
Neasetal. (1995)
Uniontown, PA
Summer 1990
59 asthmatics aged 5-49
years recruited. 19 lost to
follow-up, yielding
study population of 40.
Assessment of PEF
and respiratory symptoms.
Analytical methods unclear
in terms of control for
time factors.
83 4th and 5th grade
children reported twice
daily PEF and the presence
of cold, cough, or wheeze.
Relationship to pollutants
was analyzed by an
autoregressive linear
regression model/GEE.
The number of hours each
child spent outdoors during
the preceding 12-h period
was evaluated.
8-hmaxO3:
41.5 ppb
SD 14.2
IQR20
PM10, S02,
NO2, pollen,
fungi
12-h avg O3:
Daytime
(8 a.m.-8p.m.):
50.0 ppb
Overnight
(8 p.m.-8 a.m.):
24.5 ppb
SO2, PM10, FT
Saw significant associations between
O3 and both PEF declines and
symptom increases. Most but not all
effects remained after controlling for
temperature, pollen and fungi. The O3
effect on morning PEF disappeared after
adjusting for temperature. No PM10
effects observed.
Evening cough was associated with O3
levels weighted by hours spent outdoors
during the prior 12 hours. A decrease in
PEF was associated with O3 levels
weighted by hours spent outdoors. When
particle-strong acidity was added to the
model, the decrement was decreased and
no longer significant.
8-h max O3 (per 20 ppb):
Change in PEF (L/min):
Morning:
Lag 0-1: -2.29 (-4.26,-0.33)
Afternoon:
LagO: -2.58 (-4.26,-0.89)
Symptom score (on scale of 0-3):
Morning:
Lag 1-3: 0.08(0.03,0.13)
Afternoon:
Lag 1-3: 0.08(0.04,0.12)
12-h avg O3 (per 30 ppb increment
weighted by proportion of time
spent outdoors during prior 12
hours):
Evening PEF:
-2.79 L/min (-6.7,-1.1)
Odds ratio:
Evening cough:
2.20(1.02,4.75)
Neasetal. (1999)
Philadelphia, PA
Jul-Sep 1993
156 children aged 6-11
years at two summer camps
followed for twice-daily
PEF. Analysis using mixed
effects models adjusting for
autocorrelated errors.
Daytime
12-h avg O3
(9 a.m-9 p.m.):
SW camp:
57.5 ppb
IQR 19.8
NE camp:
55.9 ppb
IQR 21.9
FT, S042-,
PM2,,PM1
Some O3 effects detected as well as PM
effects. Similar O3-related decrements
observed in both morning and afternoon
PEF. Ozone effects not robust to SO42~ in
two-pollutant models, whereas SO42~
effects relatively robust to O3.
12-h avg O3 (per 20 ppb):
Morning and evening PEF:
O3 only models:
LagO:
-1.38 L/min (-2.81, 0.04)
Lag 1-5:
-2.58 L/min (-4.81,-0.35)
O3 with SO42~ model:
Lag not specified:
-0.04 L/min
-------
Table AX7-1 (cont'd). Effects of Acute O3 Exposure on Lung Function and Respiratory Symptoms in Field Studies
to
o
o
Reference, Study
Location and Period
Outcomes and Methods Mean O3 Levels
Copollutants
Considered
Findings, Interpretation
Effects
X
United States (cont'd)
Gent et al. (2003)
Southern New England
Apr-Sep2001
Korricketal. (1998)
Mount Washington,
NH
Summers 1991, 1992
271 children (age
<\2 years) with active,
doctor-diagnosed asthma
followed over 183 days for
respiratory symptoms. For
analysis, cohort split into
two groups: 1 30 who used
maintenance medication
during follow-up and
141 who did not, on
assumption that medication
users had more severe
asthma. Logistic regression
analyses performed.
Evaluated the acute effects
of ambient O3 on pulmonary
function of exercising
adults. 530 hikers (age
15-64 years) were
examined. Analysis using a
general linear regression
model.
l-hmaxO3:
58.6 ppb
SD 19.0
8-hmaxO3:
51.3 ppb
SD15.5
PM25
Mean O3 per
hour of hiking:
40 ppb
Range 21-74
PM2 5, smoke,
acidity
Correlation between 1-h max O3 and
daily PM2 5 was 0.77 during this warm-
season study. Large numbers of
statistical tests performed. Significant
associations between symptoms and O3
seen only in medication users, a subgroup
considered to be more sensitive. PM2 5
significant for some symptoms, but not
in two-pollutant models. Ozone effects
generally robust to PM2 5.
With prolonged outdoor exercise low-
level exposures to O3 were associated
with significant effects on pulmonary
function. Hikers with asthma had a
4-fold greater responsiveness to exposure
toO3.
1-h max O3 (per 50 ppb):
Odds ratios:
Regular medication users (n = 130):
Chest tightness:
O3 only model:
Lagl: 1.26(1.00,1.48)
O3 with PM2 5 model:
Lagl: 1.42(1.14,1.78)
Shortness of breath:
O3 only model:
Lagl: 1.22(1.02,1.45)
% change in lung function
(per 50 ppb O3):
j: -2.6% (-4.7, -0.4)
FVC: -2.2% (-3. 5, -0.8)
Thurston et al. (1997)
Connecticut River
Valley, CT
June 1991, 1992, 1993
Children (age 7-13 years)
with moderate-to- severe
asthma followed for
medication use, lung
function, and medical
symptoms at a summer
asthma camp for one week
in 1991 (n = 52), 1992 (n =
58), and 1993 (n = 56).
Analysis was conducted
using both Poisson
modeling and GLM.
l-hmaxO3:
1991: 114.0 ppb
1992: 52.2 ppb
1993: 84.6 ppb
1991-1993: 83.6
ppb
FT, S042-
O3 was most consistently associated
with acute asthma exacerbation, chest
symptoms, and lung function decrements.
Pollen was poorly associated with any
adverse effect. Consistent results were
obtained between the aggregate and
individual analyses.
1-h max O3 (per 83.6 ppb):
Relative risks:
P2-agonist use: 1.46, p< 0.05
Chest symptoms: 1.50, p< 0.05
Change in PEF (per ppb):
-0.096 L/min,p< 0.05
-------
Table AX7-1 (cont'd). Effects of Acute O3 Exposure on Lung Function and Respiratory Symptoms in Field Studies
to
o
o
Reference, Study
Location and Period
Outcomes and Methods Mean O3 Levels
Copollutants
Considered
Findings, Interpretation
Effects
United States (cont'd)
Naeheretal. (1999)
Vinton, VA
Summers 1995, 1996
Relationship between O3 8-h max O3:
and daily change in PEF 53.69 ppb
studied in a sample of 473 Range
nonsmoking women aged 17.00-87.63
19-43 years who recently
delivered babies. PEF 24-havgO3:
performed twice daily for a 34.87 ppb
2-week period. Mixed Range
linear random coefficient 8.74-56.63
model.
PM2 5, PM10, O3 was the only exposure related to
SO42~, FT evening PEF with 5-day cumulative lag
exposure showing the greatest effect.
24-h avg O3 (per 30 ppb):
Evening PEF:
Lag 1-5:
-7.65 L/min(-13.0,-2.25)
Canada
X Braueretal. (1996)
7"1 Fraser Valley, British
O Columbia, Canada
Jun-Aug 1993
58 berry pickers aged 10-69
years had lung function
measured before and after a
series of outdoor work shifts
(average duration =11
hours) over 59 days.
Analysis using pooled
regression with subject-
specific intercepts, with and
without temperature control.
l-hmaxO3:
40.3 ppb
SD 15.2
Work shift 03:
26.0 ppb
SD11.8
PM25, S042-.
N(V, NH4+,
FT
End shift FEV; and FVC significantly
diminished in relation to O3 levels.
PM2 5 also related to lung function
declines, but O3 remained significant in
2-pollutant models. Next morning lung
function remained diminished following
high O3 days. Ozone effects still evident
at or below 40 ppb. There was an overall
decline of lung function of roughly
10% over course of study, suggesting
subchronic effect. Levels of other
pollutants low during study.
Change in lung function (per ppb
l-hmaxO3):
Endshift lung function:
FEVI: -3.8mL(SE0.4)
FVC: -5.4mL(SE0.6)
Next morning function:
FEVI: -4.5 mL (SE 0.6)
FVC: -5.2mL(SE0.7)
-------
Table AX7-1 (cont'd). Effects of Acute O3 Exposure on Lung Function and Respiratory Symptoms in Field Studies
to
o
o
Reference, Study
Location and Period
Outcomes and Methods Mean O3 Levels
Copollutants
Considered
Findings, Interpretation
Effects
X
Canada (cont'd)
Brauer and Brook
(1997)
Fraser Valley, British
Columbia, Canada
Jun-Aug 1993
Additional analysis of
Brauer etal., 1996 with
personal exposure presented
for three groups, stratified
by time spent outdoors.
Group 1: 25 individuals
who spent most of the day
indoors.
Group 2: 25 individuals
who spent much of the day
indoors, but still spent
several daylight hours
outdoors.
Group 3: 15 individuals
who spent the entire work
day outdoors.
l-hmaxO3:
Ambient:
40ppb
SD15
Range 13-84
PM2 5, SO42~, Group 1: 9.0% sampling time (24-h)
NO3~, NH4+, outdoors. Personal to ambient O3 ratio
FT was 0.28.
Group 2: 25.8% sampling time (24-h)
outdoors. Personal to ambient O3 ratio
was 0.48.
Group 3: 100% sampling time (11-h
workshift) outdoors. Personal to ambient
O3 ratio was 0.96.
One of the first direct demonstrations that
magnitude of personal exposure to O3 is
related to amount of time spent outdoors.
Further showed that, on average, outdoor
fixed O3 monitors were representative of
day-to-day changes in O3 exposure
experienced by the study population.
Same outcomes as reported in
Brauer etal., 1996.
Europe
Scarlett etal. (1996)
Surrey, England
Jun-Jul 1994
Examined 154 children
aged 7 years in a primary
school next to a major
motorway for O3 exposure
effects on PEF0 75, FVC, and
FEV[ using autoregression
for % change in function.
8-h max O3: PM10, NO2, No significant association was seen
50.7ppb pollen between pulmonary function measures
SD 24.48 and O3 levels. No pollen effects.
Change in lung function (per ppb O3
weighted by inverse of variance):
FEV075:
Lagl: 0.01 mL (-0.12, 0.13)
FVC:
Lagl: 0.07 mL (-0.09, 0.23)
FEV075/FVC:
Lagl: -0.1%(-5.1,4.8)
-------
Table AX7-1 (cont'd). Effects of Acute O3 Exposure on Lung Function and Respiratory Symptoms in Field Studies
to
o
o
Reference, Study
Location and Period
Outcomes and Methods Mean O3 Levels
Copollutants
Considered
Findings, Interpretation
Effects
X
to
Europe (cont'd)
Taggartetal. (1996)
Runcom and Widnes
in NW England
M-Sep 1993
Desqueyroux et al.
(2002a)
Paris, France
Nov 1995-Nov 1996
Desqueyroux et al.
(2002b)
Paris, France
Oct 1995-Nov 1996
Investigated the relationship
of asthmatic bronchial
hyperresponsiveness and
pulmonary function to
ambient levels of
summertime air pollution
among 38 adult nonsmoking
asthmatics (age 18-70
years) using log-linear
models. Analysis limited
to investigation of within
subject variance of the
dependent variables.
60 severe asthmatics (mean
age 55 years) were
monitored by their
physicians for asthma
attacks. Asthma attacks
were based on medical data
collected by a pulmonary
physician at time of clinical
examination. Analysis
using GEE.
39 adult patients with
severe COPD (mean age
67 years) followed over
14 months by physicians
for exacerbations. Logistic
regression with GEE,
examining exposure lags of
0 to 5 days.
1-havg O3:
Maximum 61
24-h avg O3:
Maximum 24. 5
SO2, NO2,
smoke
8-h avg O3
(10 a.m.-6 p.m.):
Summer:
41 |ig/m3
SD18
Winter:
11 ng/m3
SD10
8-h avg O3
(10 a.m.-6 p.m.):
Summer:
41 |ig/m3
SD18
Winter:
11 ng/m3
SD10
PMU
PM10, S02,
NO,
No association found for O3. Changes
in bronchial hyperresponsiveness were
found to correlate significantly with
change in the levels of 24-h mean SO2,
NO2, and smoke.
Significant associations between PM10,
O3, and incident asthma attacks were
found. Low O3 levels raise plausibility
50 COPD exacerbations observed over
follow-up period. 1-, 2-, and 3-day lag
O3 significantly related to exacerbations.
No other pollutants significant. Low O3
levels raise plausibility and confounding
concerns.
24-h avg O3 (per 10 |ig/m3):
% change in bronchial
hyperresponsiveness:
Lagl: 0.3% (-16.6,20.6)
Lag 2: 2.6% (-22.1, 34.9)
8-havgO3(perlO|ig/m3):
Odds ratio:
Lag 2: 1.20(1.03,1.41)
8-h avg O3 (per 10 ng/m3):
Odds ratio:
Lagl: 1.56(1.05,2.32)
Effects appeared larger among
smokers and those with worse gas
exchange lung function.
-------
Table AX7-1 (cont'd). Effects of Acute O3 Exposure on Lung Function and Respiratory Symptoms in Field Studies
to
o
o
Reference, Study
Location and Period
Outcomes and Methods Mean O3 Levels
Copollutants
Considered
Findings, Interpretation
Effects
X
Europe (cont'd)
Just et al. (2002)
Paris, France
Apr- Jim 1996
Lagerkvist et al. (2004)
Brussels, Belgium
May 2002
82 medically diagnosed
asthmatic children (mean
age 10.9 years) followed for
O3 exposure and PEF,
asthmatic attacks, cough,
supplementary use of
P2-agonists, and symptoms
of airway irritation.
Analysis by GEE.
57 children (mean age 10.8
years) stratified by
swimming pool attendance.
Pulmonary function test
performed and Clara cell
protein levels measured in
blood before and after light
exercise outdoors for two
hours. Analysis using
student's t-test and Pearson
correlation test. For dose
calculations, O3 levels
indoors assumed to be 50%
of the mean outdoor O3
concentration.
24-h avg O3:
58.9 |ig/m3
SD 24.5
Range
10.0-121.0
PM10, NO2
Daytime outdoor None
Range 77-1 16
Exposure dose:
Range 352-914
|ig/m3-hour
In asthmatic children, O3 exposure was
related to the occurrence of asthma
attacks and additional bronchodilator use.
O3 was the only pollutant associated with
changes in lung function, as shown by an
increase in PEF variability and decrease
in PEF.
Ozone levels did not have any adverse
effect on FEV[ after 2 hours of outdoor
exercise. In addition, no significant
differences were observed between Clara
cell protein levels before and after
exercise. A marginally significant
positive correlation between ambient O3
dose and Clara cell protein levels
observed among the nonswimmers,
indicating increased antioxidant activity
following O3 exposure in this group. The
lack of a clear relationship between Clara
cell protein levels and O3 dose may be
attributable to the short period of time
between measurements and diurnal
variability of the protein levels.
24-havgO3(perlO|ig/m3):
% change in daily PEF variability:
Lag 0-2: 2.6%, p = 0.05
Odds ratio:
Supplementary use of p2-agonist
on days on which no steroids
were used:
LagO: 1.41 (1.05,1.89)
Pearson correlation:
O3 exposure dose and Clara cell
protein levels in serum:
All subjects (n = 54):
r = 0.17, p = 0.21
Nonswimmers (n = 33):
r = 0.34, p = 0.06
Swimmers (n = 21):
r =-0.08, p = 0.74
-------
Table AX7-1 (cont'd). Effects of Acute O3 Exposure on Lung Function and Respiratory Symptoms in Field Studies
to
o
o
Reference, Study
Location and Period
Copollutants
Outcomes and Methods Mean O3 Levels Considered
Findings, Interpretation
Effects
X
Europe (cont'd)
Frischeretal. (1993)
Umkirch, Germany
May-Octl991
Nasal lavage repeatedly
performed on 44 school
children (age 9-11 years)
according to protocol
published by Koren et al.
(1990). Samples collected
morning after "low" and
"high" O3 days. Nasal
lavage samples analyzed for
polymorphonuclear
leukocyte counts, albumin,
tryptase, eosinophil cationic
protein, and
myeloperoxidase. Analysis
using individual regression
methods.
Stratified
analysis of half
hour avg O3 at
3 p.m.:
Low:
< 140 ng/m3
High:
> 180 ng/m3
None
Significant higher polymorphonuclear
leukocyte counts after high O3 days.
In children without symptoms of rhinitis,
significantly elevated myeloperoxydase
and eosinophil cationic protein
concentrations detected. Results suggest
that ambient O3 produces an
inflammatory response in the upper
airways of healthy children.
Children without symptoms of
rhinitis (n = 30):
Myeloperoxydase:
LowO3: median 77.3 9 |ig/L
HighO3: median 138.60 ng/L
p < 0.05; Wilcoxon sign rank test
Eosinophilic cationic protein:
LowO3: median 3.49 ng/L
High O3: median 5.39 ng/L
p < 0.05; Wilcoxon sign rank test
Frischeretal. (1997)
Umkirch, Germany
May-Octl991
Examined 44 school
children aged 9-11 years for
ratio of ort/zo-tyrosine to
para- tyro sine in nasal
lavage as a marker of
hydroxyl radical attack.
Nasal lavage performed
according to protocol
published by Koren et al.
(1990). Concomitant lung
function tests performed.
Analysis using individual
regression methods.
Stratified
analysis of 54-h
avg O3 at 3 p.m.:
Low:
< 140 ng/m3
High:
> 180 ng/m3
None
Ambient O3 was associated with the
generation of hydroxyl radicals in the
upper airways of healthy children and
significant lung function decrements.
However, the ortho/para ratio was not
related to polymorphonuclear leukocyte
counts. Passive smoking was not related
to outcomes.
FEV; (% predicted):
Low: 105.4 (SD 15.6)
High: 103.9 (SD 15.0)
A: 1.5, p = 0.031
Ortho/para ratio:
Low: 0.02 (SD 0.07)
High:0.18(SD0.16)
A: 0.17, p = 0.0001
-------
Table AX7-1 (cont'd). Effects of Acute O3 Exposure on Lung Function and Respiratory Symptoms in Field Studies
to
o
o
Reference, Study
Location and Period
Outcomes and Methods Mean O3 Levels
Copollutants
Considered
Findings, Interpretation
Effects
X
Europe (cont'd)
Hoppeetal. (1995a,b)
Munich, Germany
Apr-Sep 1992-1994
Five study groups (age
12-95 years): (1) senior
citizens (n = 41);
(2) juvenile asthmatics
(n = 43); (3) forestry
workers (n = 41);
(4) athletes (n = 43); and
(5) clerks (n = 40) as a
control group. Examined
for lung function (FVC,
FEVj, PEF) and questions
on irritated airways. Each
subject tested 8 days, 4 days
with elevated or high O3 and
4 days with low O3.
Analysis using Wilcoxon
matched pairs signed rank
test and linear regression.
!/2-h max O3
(1 p.m.-4 p.m.):
Seniors:
High: 70 ppb
Low: 31 ppb
Asthmatics:
High: 74 ppb
Low: 34 ppb
Forestry
workers:
High: 64 ppb
Low: 32 ppb
Athletes:
High: 71 ppb
Low: 28 ppb
Clerks:
High: 68 ppb
Low: 15 ppb
None
No indication that senior citizens
represent a risk group in this study.
Senior citizens had the lowest ventilation
rate (mean 10 L/min). Athletes and
clerks experienced significant decrements
in lung function parameters. Well-
medicated juvenile asthmatics have a
trend towards large pulmonary
decrements. Forestry workers were
exposed to motor tool exhaust, which
might be a potential promoting factor.
!/2-h max O3 (per 100 ppb):
Change in lung function:
Seniors:
FEVI: 0.034 L(SD 0.101)
PEF: 0.006 L/s (SD 0.578)
Asthmatics:
FEVj: -0.210 L (SD 0.281)
PEF: -0.712 L/s (SD 0.134)*
Forestry workers:
FEVj: -0.140 L (SD 0.156)
PEF: -1.154 L/s (SD 0.885)*
Athletes:
FEVI: -0.152 L(SD 0.136)*
PEF: -0.622 L/s (SD 0.589)*
Clerks:
FEVI: -0.158 L(SD 0.114)*
PEF: -0.520 L/s (SD 0.486)*
Koppetal. (1999)
Two towns in Black
Forest, Germany
Mar-Octl994
170 school children (median
age 9.1 years) followed over
11 time points with nasal
lavage sampling. Subjects
were not sensitive to
inhaled allergens. Nasal
lavage samples analyzed for
eosinophil cationic protein,
albumen, and leukocytes.
Analysis using GEE.
!/2-h max O3:
Villingen:
64 ug/m3
5%-95% 1-140
Freudenstadt:
105 ug/m3
5%-95% 45-179
PM10, NO2, Eosinophil cationic protein and leukocyte
SO2, TSP levels peaked soon after first major
O3 episode of summer, but did not
show response to later, even higher,
O3 episodes. These observations are
consistent with an adaptive response
in terms of nasal inflammation.
Change in log eosinophil cationic
protein concentration (per ug/m3
03):
Early summer:
0.97(0.03,1.92)
Late summer:
-0.43 (-1.34, 0.47)
-------
Table AX7-1 (cont'd). Effects of Acute O3 Exposure on Lung Function and Respiratory Symptoms in Field Studies
Reference, Study
Location and Period
to
o
o
Copollutants
Outcomes and Methods Mean O3 Levels Considered
Findings, Interpretation
Effects
X
Europe (cont'd)
Ulmeretal. (1997)
Freudenstadt and
Villingen, Germany
Mar-Oct 1994
Cuijpers et al. (1994)
Maastricht, the
Netherlands
Nov-Dec 1990
(baseline),
Aug 8-16 1991 (smog
episode)
135 children aged 8-11
years in two towns were
evaluated. Pulmonary
function was associated
with the highest O3
concentration in the
previous 24 hours.
An initial cross-sectional
analysis was followed by
a longitudinal analysis
using GEE with the data
at four time periods
(Apr, Jun, Aug, Sep).
During episode, 212
children (age unspecified)
randomly chosen from
535 reexamined for lung
function and symptoms.
Corrected baseline lung
function compared by
paired t-test. Difference in
prevalence of respiratory
symptoms examined.
!/2-h max O3:
Freudenstadt:
Median 50.6 ppb
90th% interval
22.5-89.7
Villingen:
Median 32.1 ppb
90th% interval
0.5-70.1
Baseline:
8-h avg O3:
Range 2-56
ug/m3
Smog episode:
l-hmaxO3:
Exceeded
160 ug/m3 on
11 days
None
PM10, S02,
N02
In the cross-sectional analysis,
a significant negative association between
O3 exposure and FVC was only shown
at the June testing. ForFEVl5
no significant associations were detected.
In contrast, the longitudinal analysis
obtained a statistically significant
negative correlation between O3
exposure, and FVC and FEV[ for the
subpopulation living in the town with
higher O3 levels, Freudenstadt. The
associations were more pronounced
in males than females.
Small decrements in FEV[ and FEF25.75
were found in the 212 children.
However, significant decreases in
resistance parameters also were noted.
Each day a different group of 30 children
were measured. The results of the lung
function are contradictory in that
spirometry suggest airflow obstruction
while impedance measurement suggest
otherwise. Respiratory symptoms
impacted by low response rate of 122 of
212 children due to summer holidays.
No increase was observed.
Change in lung function (per ug/m3
Vi-hmaxO;,):
Freudenstadt:
-1. 13 mL,p = 0.002
Villingen:
-0.19mL,p = 0.62
FVC:
Freudenstadt:
- 1. 23 mL,p = 0.002
Villingen:
0.02 mL,p = 0.96
Change in lung function and
impedance between baseline and
smog episode:
FEV •
-0.032L(SD0.226\p<0.05
FEF25.75:
-0.086 L/s (SD 0.415), p < 0.01
Resistence at 8 Hz:
-0.47 cmH20/(L/s) (SD 1.17),
p < 0.05
Gielenetal. (1997)
Amsterdam, the
Netherlands
Apr-Jul 1995
61 children aged 7-13 years
from two special schools for
chronically ill children,
followed for twice-daily
PEF, symptoms, and
medication usage. 77% of
cohort had doctor-diagnosed
asthma.
l-hmaxO3:
77.3 ug/m3
SD15.7
8-h max O3:
67.0 ug/m3
SD 14.9
PM10, BS, Morning PEF significantly associated
pollen with 8-h max O3 at a lag of 2 days.
BS also associated with PEF. Among
14 symptom models tested, only one
yielded a significant O3 finding (for
upper respiratory symptoms). PM10
and BS, but not O3, were related to
P2-agonist inhaler use.
8-h max O3 (per 83.2 ug/m3):
% change in PEF:
Morning:
Lag 2: -1.86% (-3.58,-0.14)
Afternoon:
Lag 2: -1.88% (-3.94, 0.18)
-------
Table AX7-1 (cont'd). Effects of Acute O3 Exposure on Lung Function and Respiratory Symptoms in Field Studies
to
o
o
Reference, Study
Location and Period
Outcomes and Methods Mean O3 Levels
Copollutants
Considered
Findings, Interpretation
Effects
X
Europe (cont'd)
Hilterman et al. (1998)
Bilthoven, the
Netherlands
M-Octl995
Hoek and Brunekreef
(1995)
Deume and Enkhuizen,
the Netherlands
Mar-M 1989
60 adult nonsmoking
intermittent to severe
asthmatics (age 18-55
years) followed over
96 days. Measured morning
and afternoon PEF,
respiratory symptoms, and
medication use. Analysis
controlled for time trends,
aeroallergens,
environmental tobacco
smoke exposures, day of
week, temperature. Lags of
0 to 2 days examined.
The occurrence of acute
respiratory symptoms
investigated in children
aged 7-11 years (Deume
n = 241; Enkhuizen n = 59).
Symptoms included cough,
shortness of breath, upper
and lower respiratory
symptoms, throat and eye
irritation, headache and
nausea. Ozone-related
symptom prevalence and
incidence were examined.
Lags of 0 and 1 day, and
mean O3 concentration from
previous week were
investigated. Analyses
using Ist-order
autoregressive models and
logistic regression models.
8-hmaxO3:
80.1 ug/m3
Range 6-94
PM10,N02,
S02, BS
l-hmaxO3:
Deurne:
57ppb
SD20
Range 22-107
Enkhuizen:
59ppb
SD14
Range 14-114
PM10,N02.
SO,
O3 had strongest association with
symptoms of any pollutant analyzed.
PEF lower with O3 but not statistically
significant. No effect on medication use.
No effect modification by steroid use or
hyperresponsiveness.
No consistent association between
ambient O3 concentrations and the
prevalence or incidence of symptoms in
either city. The one significant positive
coefficient in Enkhuizen for prevalence
of upper respiratory symptoms was not
confirmed by the Deurne results.
No associations of daily symptom
prevalence or incidence found with any
of the other copollutants examined.
8-h max O3 (per 100 ug/m3):
Odds ratios:
Respiratory symptoms:
Shortness of breath:
LagO: 1.18(1.02,1.36)
Sleep disturbed by breathing:
LagO: 1.14(0.90,1.45)
Pain on deep inspiration:
LagO: 1.44(1.10,1.88)
Cough of phlegm:
LagO: 0.94(0.83,1.07)
Bronchodilator use:
LagO: 1.05(0.94,1.19)
1-h max O3 (per 50 ppb):
Prevalence of symptoms:
Deume:
Any respiratory symptom:
LagO: -0.06 (SE 0.04)
Cough:
LagO: -0.07 (SE 0.07)
Upper respiratory symptoms:
LagO: -0.06 (SE 0.05)
Enkhuizen:
Any respiratory symptom:
LagO: 0.12(SE0.07)
Cough:
LagO: -0.07 (SE 0.18)
Upper respiratory symptoms:
LagO: 0.18(SE0.09)*
-------
Table AX7-1 (cont'd). Effects of Acute O3 Exposure on Lung Function and Respiratory Symptoms in Field Studies
to
o
o
Reference, Study
Location and Period
Copollutants
Outcomes and Methods Mean O3 Levels Considered
Findings, Interpretation
Effects
Latin America
Castillejos et al. (1995) Children aged 754-11 years 1-h max O3:
SW Mexico City
Augl990-Octl991
(22 males, 18 females)
tested up to 8 times for
FEV; and FVC, before and
after exercise. Target
minute ventilation was 35
L/min/m2. Analysis using
GEE models.
PM,f
112.3ppb
Range 0-365
5th quintile mean
229.1 ppb
The mean % decrements in lung function
were significantly greater than zero only
in the fifth quintile of O3 exposure
(183-365 ppb).
% change with exercise in
5th quintile of O3 exposure
(183-365 ppb):
j: -2. 85% (-4.40, -1.31)
FVC: -1.43% (-2.81, -0.06)
X
oo
Gold etal. (1999)
SW Mexico City
1991
40 school children aged
8-11 years in polluted
community followed for
twice-daily PEF and
respiratory symptoms. PEF
deviations in
morning/afternoon from
child-specific means
analyzed in relation to
pollution, temperature,
season, and time trend.
Morning symptoms
analyzed by Poission
regression.
24-h avg O3: PM2 5, PM10 Reported significant declines in PEF
52.0 ppb in relation to 24-h avg O3 levels.
IQR 25 Effects did not vary by baseline symptom
history. Lags chosen to maximize effects
and varied by outcome. Ozone generally
robust to PM2 5. Morning phlegm
significantly related to 24-h avg O3 at
a 1-day lag.
24-h avg O3 (per 25 ppb):
% change in PEF:
Morning:
Lag 1-10: -3.8%(-5.8,-1.8)
Afternoon:
Lag 0-9: -4.6% (-7.0,-2.1)
% change in phlegm:
Morning:
Lagl: 1.1% (1.0, 1.3)
-------
Table AX7-1 (cont'd). Effects of Acute O3 Exposure on Lung Function and Respiratory Symptoms in Field Studies
to
o
o
Reference, Study
Location and Period
Copollutants
Outcomes and Methods Mean O3 Levels Considered
Findings, Interpretation
Effects
X
VO
Latin America (cont'd)
Romieuetal. (1996)
N Mexico City
Apr-Jul 1991,
Nov 1991-Feb 1992
71 mildly asthmatic
children aged 5-13 years
followed for PEF and
respiratory symptoms.
Lower respiratory
symptoms included cough,
phlegm, wheeze and/or
difficulty breathing.
1-h max O3: PM25,PM10, O3 effects observed on both PEF and
190 ppb NO2, SO2 symptoms. Symptom, but not PEF,
SD 80 effects robust to PM10 in two-pollutant
models. Symptoms related to O3
included cough and difficulty breathing.
1-h max O3 (per 50 ppb):
Change in PEF (L/min):
Morning:
Lag 0: -2.44 (-4.40, -0.49)
Lagl: -0.23 (-0.41, 1.62)
Lag 2: -1.49 (-3.80, 0.80)
Afternoon:
LagO: -0.56 (-2.70, 1.60)
Lagl: -1.27 (-3.20, 0.62)
Lag 2: -1.92 (-4.50, 0.66)
Odds ratios:
Lower respiratory symptoms:
LagO: 1.09(1.03,1.15)
Lagl: 1.10(1.04,1.17)
Lag 2: 1.04(0.97,1.12)
Romieuetal. (1997)
SW Mexico City
Apr-Jul 1991,
Nov 1991-Feb 1992
Same period as Romieu
et al., 1996, but in different
section of city. 65 mildly
asthmatic children aged
5-13 years followed for
twice-daily PEF, and
respiratory symptoms.
Up to 2 months follow-up
per child. Analysis
included temperature and
looked at 0- to 2-day lags.
No time controls. Lower
respiratory symptoms
included cough, phlegm,
wheeze and/or difficulty
breathing.
l-hmaxO3:
196 ppb
SD78
PM,f
O3 had significant effects on PEF and
symptoms, with largest effects at lags 0
and 1 day. Symptoms related to O3
included cough and phlegm. Ozone
effects stronger than those for PM, n.
1-h max O3 (per 50 ppb):
Change in PEF (L/min):
Morning:
LagO: -1.32 (-3.21, 0.57)
Lagl: -0.39 (-2.24, 1.47)
Lag 2: -0.97 (-2.94, 0.99)
Afternoon:
LagO: -1.81 (-3.60,-0.01)
Lagl: -2.32 (-4.17,-0.47)
Lag 2: -0.21 (-2.44,2.02)
Odds ratios:
Lower respiratory symptoms:
LagO: 1.11 (1.05,1.19)
Lagl: 1.08(1.01,1.15)
Lag 2: 1.07(1.02,1.13)
-------
Table AX7-1 (cont'd). Effects of Acute O3 Exposure on Lung Function and Respiratory Symptoms in Field Studies
Reference, Study
Location and Period
to
o
o
Copollutants
Outcomes and Methods Mean O3 Levels Considered
Findings, Interpretation
Effects
X
to
o
Latin America (cont'd)
Romieuetal. (1998)
Mexico City
Mar-May 1996
(1st phase)
Jun-Aug 1996
(2nd phase)
Romieu et al. (2002)
Mexico City
Octl998-Apr2000
47 street workers aged
18-58 years randomly
selected to take a daily
supplement (vitamin C,
vitamin E, and beta
carotene) or placebo during
1 st phase of study.
Following washout period,
the use of supplements and
placebos was reversed
during 2nd phase.
Pulmonary function test
performed twice a week at
end of workday. Plasma
concentrations of beta
carotene and a-tocopherol
measured. Analysis using
GEE models.
158 asthmatic children aged 1 -h max O3:
6-16 years randomly given a 102ppb
vitamin (C and E) SD 47
supplement or placebo
followed for 12 weeks.
Peak flow was measured
twice a day and spirometry
was performed twice per
week in the morning.
Double blind study. Plasma
concentration of vitamin E
levels measured. Analysis
using GEE models.
l-hmaxO3: PM10,NO2 During the 1 st phase, O3 levels were
55% of days significantly associated with declines in
exceeded 110 lung function parameters. No
ppb associations were observed in the daily
supplement group. A significant
Workday hourly supplement effect was observed. Ozone-
average during related decrements also were observed
workday prior to during the 2nd phase, however the
pulmonary associations were not significant.
function test: Supplementation with antioxidants during
67.3 ppb the 1 st phase may have had a residual
SD 24 protective effect on the lung.
PM10, NO2 Ozone levels were significantly
correlated with decrements in FEF25.75
in the placebo group, but not in the
supplement group. When analysis was
restricted to children with moderate-to-
severe asthma, amplitudes of decrements
were larger and significant for FEV1;
FEF25.75, and PEF in the placebo group.
Supplementation with antioxidants may
modulate the impact of O3 exposure on
the small airways of children with
moderate to severe asthma.
1-h max O3 (per 10 ppb):
Placebo group:
1st phase:
FEV •
LagO: -17.9 mL (SE 5.4)*
FVC:
LagO: -14.8 mL(SE 7.1)*
2nd phase:
FEVI:
LagO: -3.3mL(SE6.5)
FVC:
LagO: -0.27 mL (SE 7.8)
No significant associations with O3
observed when taking supplements.
1-h max O3 (per 10 ppb):
Children with moderate to severe
asthma:
Placebo group:
O3 with PM10 and NO2 models:
FEVI:
Lagl: -4.59 mL, p = 0.04
FEF25.75:
Lagl: -13.32mL/s,p<0.01
PEF:
Lagl: -15.01 mL/s, p = 0.04
No association observed in the
vitamin supplement group.
-------
Table AX7-1 (cont'd). Effects of Acute O3 Exposure on Lung Function and Respiratory Symptoms in Field Studies
to
o
o
Reference, Study
Location and Period
Copollutants
Outcomes and Methods Mean O3 Levels Considered
Findings, Interpretation
Effects
X
to
Latin America (cont'd)
Romieu et al. (2004)
Mexico City
Octl998-Apr2000
Additional analysis of
Romieu et al., 2002 with
data on glutathion
S-transferase Ml
polymorphism (GSTM1
null genotype) in
158 asthmatic children.
Analysis performed using
GEE models, stratified by
GSTM1 genotype (null
versus positive) within the
two treatment groups
(placebo and antioxidant
supplemented).
l-hmaxO3:
102 ppb
SD47
None
In the placebo group, O3 exposure was
significantly and inversely associated
with FEF2 5_75 in children who had the
GSTM1 null genotype, with larger effects
seen in children with moderate-to-severe
asthma. No significant decrements were
seen in the GSTM1 positive children.
These results provide preliminary
evidence that asthmatic children who
may be genetically impaired to handle
oxidative stress are most susceptible to
the effect of O3 on small airways
function.
1-h max O3 (per 50 ppb):
FEF2 5_75 in children with moderate
to severe asthma:
Placebo group:
GSTM1 null:
Lagl: -80.8mL/s,p = 0.002
GSTM1 positive:
Lag 1: -34.4 mL/s, p> 0.10
Supplement group:
GSTM1 null:
Lag 1: -7.3 mL/s, p> 0.10
GSTM1 positive:
Lag 1: 2.0 mL/s, p> 0.10
Australia
Jalaludin et al. (2000)
Sydney, Australia
Feb-Dec 1994
Three groups of children
(mean age 9.6 years):
(1) n = 45 with history of
wheeze 12 months, positive
histamine challenge, and
doctor-diagnosed asthma;
(2) n = 60 with history of
wheeze and doctor-
diagnosed asthma;
(3) n = 20 with only history
of wheeze. Examined for
evening PEF and daily O3
using GEE model and
population regression
models.
Mean daytime PM10, NO2 A significant negative association was
O3 (6 a.m.- found between daily mean deviation in
9 p.m.): PEF and same-day mean daytime O3
12 ppb concentration after adjusting for
SD 6.8 copollutants, time trend, meteorological
variables, pollen count, w&Alternaria
Maximum count. The association was stronger in
daytime O3 a subgroup of children with bronchial
(6 a.m.-9 p.m.): hyper-reactivity and doctor-diagnosed
26 ppb asthma. In contrast, the same-day
SD 14.4 maximum O3 concentration was not
statistically associated.
Change in PEF (per ppb mean
daytime O3):
All children (n = 125):
O3 only model:
-0.9178 (SE 0.4192), p = 0.03
O3 with PM10 model:
-0.9195 (SE 0.4199), p = 0.03
O3 with PM10 and NO2 model:
-0.8823 (SE 0.4225), p = 0.04
-------
Table AX7-1 (cont'd). Effects of Acute O3 Exposure on Lung Function and Respiratory Symptoms in Field Studies
to
o
o
Reference, Study
Location and Period
Copollutants
Outcomes and Methods Mean O3 Levels Considered
Findings, Interpretation
Effects
X
to
to
Australia (cont'd)
Jalaludin et al. (2004)
Sydney, Australia
Feb-Dec 1994
Same three groups of
children as studied in
Jalaludin et al., 2000.
Examined relationship
between O3 and evening
respiratory symptoms
(wheeze, dry cough, and
wet cough), evening asthma
medication use (inhaled
(32-agonist and inhaled
corticosteroids), and doctor
visits for asthma. Analysis
using GEE logistic
regression models.
Mean daytime PM10, NO2
O3 (6 a.m.-
9p.m.):
12ppb
SD6.8
Maximum
daytime O3
(6 a.m.-9p.m.):
26 ppb
SD 14.4
No significant O3 effects observed on
evening symptoms, evening asthma
medication use, and doctors visits.
Also, no differences in the response of
children in the three groups. A potential
limitation is that the use of evening
outcome measures rather than morning
values may have obscured the effect of
ambient air pollutants. Only consistent
relationship was found between mean
daytime PM10 concentrations and doctor
visits for asthma.
Mean daytime O3 (per 8.3 ppb):
Odds ratios:
All children (n = 125):
Wheeze:
Lagl: 1.00(0.93,1.08)
Dry cough:
Lagl: 1.03(0.96,1.11)
Wet cough:
Lagl: 0.97(0.92,1.03)
Inhaled p2-agonist use:
Lagl: 1.02(0.97,1.07)
Inhaled corticosteroid use:
Lagl: 1.02(0.99,1.04)
Doctor visit for asthma:
Lagl: 1.05(0.77,1.43)
Asia
Chen etal. (1998)
Six communities in
Taiwan
1994-1995
4,697 school children (age
unspecified) from a rural
area (Taihsi), urban areas
(Keelung and Sanchung),
and petrochemical industrial
areas (Jenwu, Linyuan, and
Toufen) cross-sectionally
examined for respiratory
symptoms and diseases
using parent-completed
questionnaires. Multiple
logistic regression models
used to compare odds of
symptoms and diseases in
urban or petrochemical
areas to the rural area after
controlling for potential
confounding factors.
24-h avg O3:
Rural area:
52.56 ppb
Urban area:
Mean range
38.34-41.90 ppb
Petrochemical
industrial area:
Mean range
52.12-60.64 ppb
SO2, CO,
PM10,N02
School children in urban communities,
but not in petrochemical industrial areas,
had significantly more respiratory
symptoms and diseases compared to
those living in the rural community.
However, mean O3 levels in the urban
communities were lower than that of the
rural community. No causal relationship
could be derived between O3 and
respiratory symptoms and diseases in this
cross-sectional study.
Urban areas compared to rural area:
Odds ratios:
Respiratory symptoms:
Morning cough:
1.33(0.98,1.80)
Day or night cough:
1.67(1.21,2.29)
Shortness of breath:
1.40(1.04,1.91)
Wheezing or asthma:
1.68(1.11,2.54)
-------
^ Table AX7-1 (cont'd). Effects of Acute O3 Exposure on Lung Function and Respiratory Symptoms in Field Studies
c ~
Reference, Study Copollutants
^ Location and Period Outcomes and Methods Mean O3 Levels Considered Findings, Interpretation Effects
to
O Asia (cont'd)
Chen etal. (1999) Valid lung function data l-hmaxO3: SO2, CO, FEV; and FVC significantly associated Change in lung function:
Three towns in Taiwan: obtained once on each of Range PM10,NO2 with 1-day lag O3. FVC also related to
Sanchun, Taihsi, 895 children (age 8-13 19.7-110.3 ppb NO2, SO2, and CO. No PM10 effects O3 only models:
Linyuan years) in three towns. observed. Only O3 remained significant Lag 1:
May 1995-Jan 1996 Examined relation between in multipollutant models. No PM10 FEVj! -0.64 mL/ppb (SE 0.34)*
lung function and pollution effects. A significant O3 effect was not FVC: -0.79 mL/ppb (SE 0.32)*
concentrations on same day evident at O3 levels below 60 ppb.
and over previous 1, 2, and O3 with NO2 models:
7 days. Multipollutant Lag 1:
models examined. FEV^ -0.85 mL/ppb (SE 0.34)*
FVC: -0.91 mL/ppb (SE 0.37)*
*p<0.05
X ^^^^^^^^^^^^^^^^^^^^^^^^^^=^^^^^^^^^^^^^^^^^^^^^^^^^^=
to
-------
Table AX7-2. Effects of Acute O3 Exposure on Cardiovascular Outcomes in Field Studies
to
o
o
Reference, Study
Location and Period
Outcomes and Methods Mean O3 Levels
Copollutants
Considered
Findings, Interpretation
Effects
X
to
United States
Liao et al. (2004)
Three locations in U.S.:
Minneapolis, MN;
Jackson, MS; Forsyth
County, NC
1996-1998
Peters et al. (2000a)
Eastern Massachusetts
1995-1997
5,431 cohort members of
the Atherosclerosis Risk in
Communities study, men
and women aged 45-64
years at entry in 1987.
Association between O3 and
cardiac autonomic control
assessed using 5-minute
heart rate variability indices
collected over a 4-hour
period. Analysis using
multivariable linear
regression models, adjusting
for individual
cardiovascular disease risk
factors and meteorological
factors.
Records of detected
arrhythmias and therapeutic
interventions were
downloaded from
defibrillators implanted in
cardiac clinic patients aged
22-85 years (n = 100).
Analysis was restricted to
defibrillator discharges
precipitated by ventricular
tachycardias or fibrillation.
Data was analyzed by
logistic regression models
using fixed effect models
with individual intercepts.
8-h avg O3
(10 a.m.-6 p.m.):
41 ppb
SD16
PM10, CO,
S02, N02
24-h avg O3:
18.6 ppb
IQR 14.0
PM25,PM10,
BC, CO, NO2,
S02
Significant interaction between O3 and
ethnicity in relation to high-frequency
power (p < 0.05). Ambient O3
significantly associated with high-
frequency power among whites, but not
blacks. No significant O3 effect on other
heart rate variability indices, including
low-frequency power and SD of normal
R-R intervals. More consistent
relationships observed between PM10
and heart rate variability indices.
No significant O3 effects observed for
defibrillator discharge interventions.
For patients with ten or more
interventions, increased arrhythmias
were associated significantly with
PM2 5, CO, and NO2 at various lag
periods, but not O3.
8-h avg O3 (per 16 ppb):
Log-transformed high-frequency
power:
White race:
Lagl: -0.069 (SE 0.019)*
Black race:
Lagl: 0.047 (SE 0.034)
Log-transformed high-frequency
power:
Lagl: -0.010 (SE 0.016)
SD of normal R-R intervals:
Lagl: -0.336 (SE 0.290)
*p<0.05
24-h avg O3 (per 32 ppb):
Odds ratios:
Defibrillator discharges:
Patients with at least one event:
LagO: 0.96(0.47,1.98)
Patients with at least ten events:
LagO: 1.23(0.53,2.87)
-------
Table AX7-2 (cont'd). Effects of Acute O3 Exposure on Cardiovascular Outcomes in Field Studies
to
o
o
Reference, Study
Location and Period
Outcomes and Methods Mean O3 Levels
Copollutants
Considered
Findings, Interpretation
Effects
United States (cont'd)
Peters etal. (2001)
Greater Boston area,
MA
Janl995-May 1996
Case-crossover study design
used to investigate
association between air
pollution and triggering of
myocardial infarction in
772 patients (mean age
6 1.6 years). For each
subject, one case period
was matched to three
control periods 24 hours
apart. Conditional logistic
regression used for analysis.
l-hmaxO3:
19.8ppb
24-h avg O3:
19.9ppb
PM25,PM1
PM1
CO, NO,, SO,
None of the gaseous pollutants, including Odds ratios:
O3, were significantly associated with
the triggering of myocardial infarctions.
Significant associations reported for
Myocardial infarctions:
2-h avg O3 (per 45 ppb):
Lag 1 hour:
1.31 (0.85,2.03)
24-h avg O3 (per 30 ppb):
Lag 24 hours:
0.94(0.60,1.49)
X
to
Park et al. (2004)
Greater Boston area,
MA
Nov 2000-Oct 2003
Effect of O3 on heart rate
variability was examined in
497 adult males (mean age
72.7 years). Subjects were
monitored during a 4-
minute rest period between
8 a.m. and 1 p.m. Ozone
levels measured at central
site 1 km from study site.
Exposure averaging times
of 4-hours, 24-hours, and
48-hours investigated.
Modifying effects of
hypertension, ischemic
heart disease, diabetes,
and use of cardiac/
antihypertensive
medications also examined.
24-h avg O3: PM2 5, particle Of the pollutants examined, only PM2 5
23.0 ppb number and O3 showed significant associations
SD 13.0 concentration, with heart rate variability outcomes.
BC, NO2, SO2, The 4-hour averaging period was most
CO strongly associated with heart rate
variability indices. The O3 effect was
robust in models including PM2 5. The
associations between O3 and heart rate
variability indices were stronger in
subjects with hypertension (n = 335) and
ischemic heart disease (n = 142).
In addition, calcium-channel blockers
significantly influenced the effect of O3
on low frequency power. Major
limitations of this study are the use of a
short 4-minute period to monitor heart
rate variability and the lack of repeated
measurements for each subject.
4-h avg O3 (per 13 ppb):
Change in low frequency power:
All subjects:
-11.5% (-21.3, -0.4)
Subjects with hypertension:
-12.6% (-25.0, 1.9)
Subjects without hypertension:
-5.4% (-21.6, 14.1)
Subjects with ischemic heart
disease:
-25.8% (-41.9,-5.3)
Subjects without ischemic heart
disease:
-4.8% (-16.7, 8.8)
-------
Table AX7-2 (cont'd). Effects of Acute O3 Exposure on Cardiovascular Outcomes in Field Studies
to
o
o
Reference, Study
Location and Period
Outcomes and Methods Mean O3 Levels
Copollutants
Considered
Findings, Interpretation
Effects
X
to
United States (cont'd)
Gold et al. (2000;
reanalysis Gold et al.,
2003)
Boston, MA
Jun-Sep 1997
Repeated measurements
of heart rate variability in
subjects aged 53-87 years
(n = 21,163 observations).
Twenty-five minute
protocol included 5 minutes
each of rest, standing,
exercise outdoors, recovery,
and 20 cycles of slow
breathing. Ozone levels
measured at central site
4.8 miles from study site.
Analyses using random
effects models and GAM
with stringent convergence
criteria.
l-hmaxO3: PM25, NO2, Increased levels of O3 were associated
25.7ppb SO2 with reduced r-MSSD (square root of the
IQR 23.0 mean of the squared differences between
adjacent normal RR intervals) during the
slow breathing period after exercise
outdoors. The estimated O3 effects were
similar to those of PM2 5. Results suggest
that O3 exposure may decrease vagal
tone, leading to reduced heart rate
variability.
1-h max O3 (per 23.0 ppb):
Change in r-MSSD:
During first rest period:
O3 only model:
-3.0 ms(SE 1.9), p = 0.12
During slow breathing period:
O3 only model:
-5.8ms(SE2.4),p = 0.02
O3 with PM2 5 model:
-5.4 ms(SE 2.5), p = 0.03
Canada
Rich et al. (2004)
Vancouver, British
Columbia, Canada
Feb-Dec 2000
Case-crossover study
design used to investigate
association between air
pollution and cardiac
arrhythmia in patients aged
15-85 years (n = 34) with
implantable cardioverter
defribillators. Controls
periods were selected 7 days
before and after each case
day. Analysis using
conditional logistic
regression.
1-h max O3: PM25, PM10, No consistent association between any
27.5 ppb EC, OC, of the air pollutants, including O3, and
SD 9.7 SO42~, CO, implantable cardioverter defribillators
IQR 13.4 NO2, SO2 discharges. No significant association
observed in all year data, however,
significant relationship found in winter
months at a 3-day lag. Overall, little
evidence that air pollutants affect risk
of cardiac arrhythmias, however, power
was limited to study subtle effects.
No quantitative results for O3.
-------
Table AX7-2 (cont'd). Effects of Acute O3 Exposure on Cardiovascular Outcomes in Field Studies
to
o
o
Reference, Study
Location and Period
Outcomes and Methods Mean O3 Levels
Copollutants
Considered
Findings, Interpretation
Effects
Canada (cont'd)
Vedal et al. (2004)
Vancouver, British
Columbia, Canada
1997-2000
Retrospective, longitudinal
panel study of 50 patients
(age 12-77 years) with
implantable cardioverter
defribillators. Occurrence
of cardiac arrhythmia was
associated with air
pollutants over four-year
period. GEE used for
analysis.
l-hmaxO3:
28.2 ppb
SD 10.2
IQR 13.8
PM10, CO,
N02, S02
No consistent association between any
of the air pollutants and % change in
arrhythmia. Among patients with at least
2 arrhythmia event-days per year, a
significant negative relationship between
O3 and arrhythmias was observed at lag
3-day during the summer, but no
associations were found during the
winter. These results do not provide
evidence for an O3 effect on cardiac
arrhythmias in susceptible patients.
No quantitative results for O3.
X
to
Latin America
Holguin et al. (2003)
Mexico City
Feb-Apr 2000
Association between O3
and heart rate variability
examined in 34 elderly
subjects (mean age
79 years), in a nursing
home. Subjects were
monitored during a
5-minute rest period
between 8 a.m. and 1 p.m.
every other day for a
3-month period. A total of
595 observations were
collected. Ambient O3
levels obtained from central
site 3 km upwind from
study site. Analysis
performed using GEE
models adjusting for
potential confounding
factors including age and
average heart rate.
1-h max O3: PM2 5 (indoor, Only PM2 5 and O3 were significantly
149 ppb outdoor, associated with heart rate variability
SD40 total), NO2, outcomes. A significant effect of O3 on
SO2, CO heart rate variability was limited to
subjects with hypertension (n = 21).
In two-pollutant models, the magnitude
of the PM2 5 effect decreased slightly but
remained significant, whereas O3 was
no longer associated with heart rate
variability indices.
1-h max O3 (per 10 ppb):
Log10 high frequency
power/100,000 ms2:
All subjects:
-0.010 (-0.022, 0.001)
Subjects with hypertension:
-0.031 (-0.051,-0.012)
Subjects without hypertension:
0.002 (-0.012, 0.016)
Log10 low frequency
power/100,000 ms2:
All subjects:
-0.010 (-0.021, 0.001)
Subjects with hypertension:
-0.029 (-0.046,-0.011)
Subjects without hypertension:
0.001 (-0.012, 0.015)
-------
Table AX7-3. Effects of O3 on Daily Emergency Department Visits
1
to
o
o
J>
X
~I~J
to
oo
t-1
Reference, Study
Location and Period Outcomes and Design
United States
Jaffe et al. (2003) Daily time series study
Cincinnati, Cleveland, of emergency
and Columbus, OH department visits for
Jun-Aug 1991-1996 asthma among
Medicaid recipients
aged 5-34 years.
Mean O3
Levels
8-h max O3:
Cincinnati:
60 ppb
SD20
Cleveland:
50 ppb
SD17
Columbus:
57 ppb
SD16
Copollutants Lag Structure Method, Findings,
Considered Examined Interpretation
PM10, NO2, 1 , 2, 3 Poisson regression with control
SO2 for city, day of week, week,
year, minimum temperature,
overall trend, and a dispersion
parameter. No specific effort
to control cycles, but
regression residuals were
uncorrelated, presumably due
to seasonal restriction. Results
shown for individual cities and
overall. PM10 available only
every 6th day. Positive
relationships between
emergency department visits
for asthma and 8-h max O3
levels lagged 2 to 3 days.
Results of borderline statistical
significance. Other pollutants
also related to asthma
emergency department visits
in single-pollutant models.
Effects
(Relative Risk and 95% CI)
8-h max O3 (per 30 ppb):
Cincinnati:
Lag 2: 1.16(1.00,1.37)
Cleveland:
Lag 2: 1.03(0.92,1.16)
Columbus:
Lag 3: 1.16(0.98,1.37)
Three cities: 1.09(1.00,
1.19)
Jones etal. (1995)
Baton Rouge, LA
Jun-Aug 1990
Daily emergency
department visits for
respiratory complaints
over a 3-month period
in pediatric (age 0-17
years), adult (age 18-60
years), and geriatric
(age > 60 years)
subgroups.
l-hmaxO3:
69.1 ppb
SD28.7
24-havgO3:
28.2 ppb
SD11.7
Mold, pollen Not specified.
Relatively simple statistical
approach using ordinary least
squares regression to model
effects of O3 by itself and of O3
along with pollen counts, mold
counts, temperature, and
relative humidity. No direct
control of cycles but authors
reported non-significant
autocorrelations among model
residuals. Data restriction to
3-month period may have
reduced any cyclic behavior.
Significant associations
between respiratory emergency
department visits and O3
observed for adult age group
only in multiple regression
models.
24-h avg O3 (per 20 ppb):
Pediatric: 0.87(0.65,1.09)
Adult: 1.20(1.01,1.39)
Geriatric: 1.27(0.93,1.61)
-------
Table AX7-3 (cont'd). Effects of O3 on Daily Emergency Department Visits
X
to
VO
Reference, Study
Location and Period
United States (cont'd)
Weisel et al. (2002)
New Jersey
May-Aug 1995
Friedman et al. (2001)
Atlanta, GA
M-Aug 1996
Outcomes and Design
Daily asthma
emergency department
visits for all ages.
Emergency department
visits and hospital
admissions for asthma
Mean O3 Copollutants
Levels Considered
1-h max O3; Pollen, spores
5-h avg O3
(10a.m.-
3 p.m.); and 8-h
avg O3 (2 p.m.-
10p.m.)
analyzed.
Levels not
reported.
1-h max O3: NO2, SO2, CO,
Levels not PM10, mold
reported.
Lag Structure Method, Findings,
Examined Interpretation
0, 1 , 2, 3 No control for time, but
authors report no
autocorrelation, which
alleviates concerns about lack
of control. Significant O3
effects reported, even after
adjusting for potential
confounding by pollen. All
three O3 indices gave
essentially same results.
0, 0-1, 0-2 Analysis using Poisson GEE
models addressing serial
autocorrelation. An overall
Effects
(Relative Risk and 95% CI)
Slope estimate
(visits/day/ppb):
Excluding data from May
1995 when pollen levels are
high:
O3 only model:
LagO: 0.039, p = 0.049
O3 with pollen model:
LagO: 0.040, p = 0.014
1-h max O3 (per 50 ppb):
Pediatric emergency
in children aged 1-16
years. Outcomes
measures during 1996
Summer Olympics
were compared to a
baseline period of 4
weeks before and after
the Olympic Games.
decrease was observed when
comparing the number of visits
and hospitalizations during the
Olympic Games to the baseline
period. However, significant
associations between O3 and
asthma events were found
during the Olympic Games.
departments:
LagO: 1.2(0.99,1.56)
Lag 0-1: 1.4(1.04,1.79)
Lag 0-2: 1.4(1.03,1.86)
Metzger et al. (2004)
Atlanta, GA
Jan 1993-Aug 2000
Emergency department
visits for total and
cause- specific
cardiovascular diseases
by age groups 19+
years and 65+ years.
8-h max O3:
Median 53.9
ppb
1 Oth % to 90th
% range 13.2 -
44.7
NO2, SO2, CO,
PM25,PM10,
PM10.2,,
ultrafine PM
count, SO42~,
LT, EC, OC,
metals,
oxygenated
hydrocarbons
0-2 Poisson GLM regression used
for analysis. A priori models
specified a lag of 0 to 2 days.
Secondary analyses performed
to assess alternative pollutant
lag structures, seasonal
influences, and age effects.
Cardiovascular visits were
significantly associated with
several pollutants, including
N02, CO, and PM2 5, but
notO3.
8-h max O3 (per 25 ppb):
All ages:
Total cardiovascular:
1.008(0.987,1.030)
Dysrhythmia:
1.008(0.967,1.051)
Congestive heart failure:
0.965(0.918,1.014)
Ischemic heart disease:
1.019(0.981,1.059)
Peripheral vascular and
cerebro vascular disease:
1.028(0.985,1.073)
-------
Table AX7-3 (cont'd). Effects of O3 on Daily Emergency Department Visits
X
Reference, Study
Location and Period
United States (cont'd)
Peel et al. (2004)
Atlanta, GA
Jan 1993-Aug 2000
Tolbert et al. (2000)
Atlanta, GA
Jun-Aug 1993-1995
Outcomes and Design
Emergency department
visits for total and
cause- specific
respiratory diseases by
age groups 0-1, 2-18,
19+, and 65+ years.
Pediatric (aged 0-16
years) asthma
emergency department
visits over three
summers in Atlanta.
A unique feature of the
study was assignment
of O3 exposures to
zip code centroids
based on spatial
interpolation from nine
O3 monitoring stations.
Mean O3
Levels
8-h max O3:
55.6ppb
SD23.8
l-hmaxO3:
68.8 ppb
SD21.1
8-h max O3:
59.3 ppb
SD19.1
Copollutants
Considered
NO2, SO2, CO,
PM25,PM10,
PM10.2,,
ultrafine PM
count, SO42~,
H+, EC, OC,
metals,
oxygenated
hydrocarbons
PM10,N02,
mold, pollen
Lag Structure Method, Findings,
Examined Interpretation
0-2 Poisson GEE and GLM
regression used for analysis.
A priori models specified a lag
of 0 to 2 days. Also performed
secondary analyses estimating
the overall effect of pollution
over the previous two weeks.
Seasonal analyses indicated
stronger associations with
asthma in the warm months.
Quantitative results not
presented for multipollutant,
age-specific, and seasonal
analyses.
1 A priori specification of
model, including a lag of 1 day
for all pollutants and
meteorological variables.
Secondary analysis using
logistic regression of the
probability of daily asthma
visits, referenced to total visits
(asthma and non-asthma).
Significant association with
O3 and PM10 in 1-, but not in
2-pollutant models (correlation
between O3 and PM10:
r = 0.75). Secondary analysis
indicated nonlinearity, with O3
effects clearly significant only
for days > 100 ppb versus days
< 50 ppb.
Effects
(Relative Risk and 95% CI)
8-h max O3 (per 25 ppb):
All ages:
Total respiratory:
1.024(1.008,1.039)
Upper respiratory infections:
1.027(1.009,1.045)
Asthma:
1.022(0.996,1.049)
Pneumonia:
1.015(0.981,1.050)
COPD:
1.029(0.977,1.084)
8-h max O3 (per 20 ppb):
Poisson regression:
O3 only model:
1.040(1.008,1.074)
Logistic regression:
O3 only model:
1.042(1.017,1.068)
O3 with PM10 model:
1.024(0.982,1.069)
-------
Table AX7-3 (cont'd). Effects of O3 on Daily Emergency Department Visits
to
o
o
X
Reference, Study
Location and Period
United States (cont'd)
Zhu et al. (2003)
Atlanta, GA
Jun-Aug 1993-1995
Outcomes and Design
Asthma emergency
department visits in
children (age 0-16
Mean O3
Levels
8-h max O3:
Levels not
reported.
Copollutants
Considered
None
Lag Structure
Examined
1
Method, Findings,
Interpretation
Used Bayseian hierarchical
modeling to address model
variability and spatial
Effects
(Relative Risk and 95% CI)
8-h max O3 (per 20 ppb):
Posterior median:
years) over three
summers in Atlanta.
associations. Data was
analyzed at the zip code level
to account for spatially
misaligned longitudinal data.
A positive, but nonsignificant
relationship between O3 and
emergency room visits for
asthma.
1.016(0.984,1.049)
Canada
Delfinoetal. (1997b)
Montreal, Quebec,
Canada
Jun-Sep 1992-1993
Daily time series
ecologic study of
emergency department
visits for respiratory
complaints within five
age strata (< 2, 2-1&,
19-34, 35-64, > 64
years).
l-hmaxO3: PM10, PM25,
1992: 33.2 ppb SO42-, H+
SD 12.6
1993: 36.2 ppb
SD 13.8
8-h max O3:
1992: 28. 8 ppb
SDH. 3
1993: 30.7 ppb
SDH. 5
0, 1, 2 Used ordinary least squares,
with 1 9-day weighted moving
average pre-filter to control
cycles; weather also controlled.
Significant effects reported for
1-day lag O3 in 1993 only for
age > 64 years. This O3 effect
reported to be robust in two-
pollutant models. LowO3
levels raise plausibility
concerns. Short data series,
multiple tests performed, and
inconsistent results across
years and age groups raise
possibility of chance findings.
1993 (age > 64 years):
1-h max O3 (per 36.2 ppb):
Lag 1: 1.214(1.084,1.343)
8-h max O3 (per 30.7 ppb):
Lagl: 1.222(1.091,1.354)
-------
Table AX7-3 (cont'd). Effects of O3 on Daily Emergency Department Visits
X
to
Reference, Study
Location and Period
Canada (cont'd)
Delfinoetal. (1998b)
Montreal, Quebec,
Canada
Jun-Aug 1989-1990
Stiebetal. (1996)
Saint John, New
Brunswick, Canada
May-Sep 1984-1992
Outcomes and Design
Daily time series
ecologic study of
emergency department
visits for respiratory
complaints across all
ages and within four
age strata (<2, 2-34,
35-64, >64 years).
Daily emergency
department visits for
asthma in all ages,
age<15 years and
15+ years.
Mean O3 Copollutants
Levels Considered
1 -h max O3 : Estimated
1989: 44.1ppb PM25
SD 18.3
1990: 35.4 ppb
SD 12.9
8-h max O3:
1989: 37. 5 ppb
SD 15.5
1990: 29. 9 ppb
SD11.2
l-hmaxO3: SO2,NO2,
4 1.6 ppb SO42% TSP
Range 0-1 60
95th % 75
Lag Structure Method, Findings,
Examined Interpretation
0, 1, 2 Same analytical approach used
in Delfino et al, 1997. Results
presented only for 1989.
Significant effects reported for
1-day lag O3 in 1989 only for
age > 64 years. This O3 effect
reported to be robust in
2-pollutant models.
0, 1, 2, 3 Poisson analysis with control
of time based on 19-day
moving average filter. Also
controlled day of week and
weather variables. Ozone was
Effects
(Relative Risk and 95% CI)
1 989 (age > 64 years):
1-h max O3 (per 44.1 ppb):
Lag 1: 1.187(0.969,1.281)
8-h max O3 (per 37.5 ppb):
Lagl: 1.218(1.097,1.338)
No significant O3 effects in
other age groups or for 1990.
1-h max O3 > 75 ppb:
Lag 2: 1.33(1.10,1.56)
only pollutant consistently
associated with emergency
department visits for asthma,
but effect appeared nonlinear,
with health impacts evident
only above 75 ppb O3.
-------
3
Table AX7-3 (cont'd). Effects of O3 on Daily Emergency Department Visits
to
o
o
X
Reference, Study
Location and Period
Europe
Atkinson et al.
(1999a)
London, England
1992-1994
Thompson et al.
(2001)
Belfast, N Ireland
1993-1995
Outcomes and Design
Emergency department
visits for respiratory
complaints, asthma for
all ages and age 0-14,
15-64, and 65+ years.
Asthma emergency
department admissions
in children (age
unspecified)
Mean O3
Levels
8-h max O3:
17.5 ppb
SD11.5
24-h avg O3:
Warm season:
18.7 ppb
IQR9
Cold season:
17.1 ppb
IQR12
Copollutants Lag Structure Method, Findings,
Considered Examined Interpretation
NO2, SO2, CO, 0, 1 , 2, 3 Poisson GLM regression used
PM10 0-1,0-2,0-3 for analysis. No warm season
analysis attempted. PM10
positively associated.
PM10, SO2, 0,0-1, 0-2, 0-3 GLM with sinusoids.
NO2, CO, Pre-adjustment. Very low O3
benzene levels in both seasons. No O3
effect in warm season.
Significant inverse O3
associations in full-year and
cold-season models. After
adjusting for benzene in model
O3 was no longer negatively
associated with asthma visits.
Effects
(Relative Risk and 95% CI)
8-h max O3 (per 25.7 ppb):
All ages:
Total respiratory:
Lag 1: 1.017(0.991,1.043)
Asthma:
Lagl: 1.027(0.983,1.072)
24-h avg O3 (per 10 ppb):
All year:
O3 only model:
Lag 0-1: 0.93(0.87,1.00)
O3 with benzene model:
Lag 0-1: 1.08(0.97,1.21)
Warm season:
O3 only model:
Lag 0-1: 0.99(0.89,1.10)
Cold season:
H
6
o
2!
•£—\
O
H
0
H
O
O
H
W
Bourcier et al. (2003) Ophthalmological 24-h avg O3: PM10, SO2, 0,1,2,3 Logistic Regression
Paris, France emergency 35.7 |ig/m3 NO2
Jan 1999-Dec 1999 examination; Range 1-97
conjunctivitis and
related ocular surface
problems.
Lag 0-1: 0.89(0.82,0.97)
Results indicate a strong
relation to NO2 and NO.
24-h avg O3 (per 69 |ig/m3):
Conjunctivitis:
LagO: 1.13(0.90,1.42)
-------
Table AX7-3 (cont'd). Effects of O3 on Daily Emergency Department Visits
to
o
o
X
Reference, Study
Location and Period
Europe (cont'd)
Castellsague et al.
(1995)
Barcelona, Spain
1985-1989
Tobias etal. (1999)
Barcelona, Spain
Outcomes and Design
Daily emergency
department visits for
asthma in persons
aged > 14 years.
Daily asthma
emergency department
Mean O3
Levels
l-hmaxO3:
Summer:
43ppb
IQR22
Winter:
29ppb
IQR16
Levels not
reported.
Copollutants Lag Structure Method, Findings,
Considered Examined Interpretation
BS, SO2, NO2 Not specified. Poisson regression with year
and month dummy variables
and extensive control for
weather factors (minimum,
maximum, mean temperature,
relative humidity, dewpoint
temperature; continuous and
categorical parameterizations)
BS, NO2, SO2 Not specified. Poisson analysis using
APHEA methodology.
Effects
(Relative Risk and 95% CI)
1-h max O3 (per 12.7 ppb):
Summer:
0.991 (0.939, 1.045)
Winter:
1.055(0.998,1.116)
O3 results were sensitive to
method used to control
1986-1989
visits. Investigated
sensitivity of results
to four alternative
methods for controlling
asthma epidemics.
Asthma epidemics either
not controlled, or controlled
with one, six, or individual
epidemic dummy variables.
asthma epidemics, with
regression coefficients
ranging over 5-fold
depending on the model.
Only 1 of 8 models reported
had a significant O3 effect.
O
5>
-n
H
6
o
!2
O
— ]
O
r^t
0
H
O
^—i
Tenias et al.
(1998,2002)
Valencia, Spain
1994-1995
Daily emergency
department visits for
asthma and COPD
in persons aged
> 14 years.
l-hmaxO3:
All year:
62.8 ug/m3
Warm season:
74.0 ug/m3
Cool season:
5 1.4 ug/m3
BS, NO2, SO2, 0, 1 , 2, 3, 4, 5 Poisson analysis using APHEA
CO methodology. Compared
warm and cold season effects.
GAM explored in sensitivity
analysis. For asthma, both O3
and NO2 significant in 1- and
2-pollutant models, and O3
effect larger in warm season.
For COPD, both O3 and CO
significant in both 1- and
2-pollutant models and no
difference in O3 effects by
season.
1-h max O3 (per 10
Asthma:
All year:
Lagl: 1.06(1.01,
Warm season:
Lag 1: 1.08(1.02,
Cold season:
Lagl: 1.04(0.97,
COPD:
All year:
Lag 5: 1.06(1.02,
ug/m3):
1.11)
1.05)
1.11)
1.10)
O
HH
H
W
-------
Table AX7-3 (cont'd). Effects of O3 on Daily Emergency Department Visits
to
o
o
X
Reference, Study
Location and Period
Latin America
Hemandez-Garduno
etal. (1997)
Mexico City
May 1992-Jan 1993
Lin etal. (1999)
Sao Paulo, Brazil
May 1991 -Apr 1993
Ilabacaetal. (1999)
Santiago, Chile
Feb 1995-Aug 1996
Asia
Chew etal. (1999)
Singapore
Jan 1990-Dec 1994
Outcomes and Design
Visits to clinics for
respiratory diseases in
persons aged 1 month
to 92 years.
Daily pediatric (age
unspecified) respiratory
emergency department
visits.
The association
between pollutant
levels and emergency
visits for pneumonia
and other respiratory
illnesses among
children.
Emergency department
visits for asthma
in persons aged
3-21 years.
Mean O3
Levels
% time
exceeding 1-h
max O3 of 120
ppb:
6.1-13.2% by
location
l-hmaxO3:
34 ppb
SD22
O3 1-h max:
Warm season:
66.6 ug/m3
SD25.2
Cold season:
27.6 ug/m3
SD 20.2
l-hmaxO3:
23 ppb
SD15
Copollutants Lag Structure Method, Findings,
Considered Examined Interpretation
SO2, NO2, CO 0,1,2,3,4,5 GLM with pre-adjustment.
Ozone at lags 0 and 5 days
significantly associated with
daily visits for all ages,
age < 14 years, and 15+ years.
Neither O3 nor NO2 significant
in 2-pollutant model.
SO2, CO, 0,1,0-1,0-2, Seasonal control using month
PM10,NO2 0-3,0-4,0-5 dummy variables. Also
controlled day of week,
temperature. Both O3 and
PM10 associated with outcome
alone and together.
PM10, PM2 5, 1,2, 3, 1-7 Poisson regression analysis.
SO2,NO2
TSP, PM10, 0,1,2 Simplistic but probably
SO2, NO2 adequate control for time by
including 1-day lagged
outcome as covariate.
In adjusted models that
included covariates, O3
had no significant effect.
Effects
(Relative Risk and 95% CI)
1-h max O3 (per maximum
less average, value not given)
LagO: 1.19(SE0.08),
p<0.05
Lag 5: 1.19 (SE 0.08),
p<0.05
1-h max O3 (per 5 ppb):
O3 only model:
Lag 0-4: 1.022(1.016,1.028)
O3 with PM10 model:
Lag 0-4: 1.015(1.009,1.021)
Warm season:
1-h max O3 (per 30 ug/m3):
Lag 2: 1.019(1.003,1.035)
Cold season:
1-h max O3 (per 24 ug/m3):
Lag 2: 0.995(0.978,1.011)
No quantitative results
presented for O3.
-------
Table AX7-4. Effects of O3 on Daily Hospital Admissions
t—t
1
to
o
o
^
X
OJ
ON
O
|>
6
0
2;
o
H
O
0
w
o
o
HH
H
W
Reference, Study
Location and Period
United States
Niedell (2004)
California
1992-1998
Mann et al. (2002)
South Coast air basin,
CA
1988-1995
Linn et al. (2000)
Los Angeles, CA
1992-1995
Outcomes and Design
Emergency department
visits for asthma within
five age strata (0-1, 1-3,
3-6, 6-12, and 12-1 8
years).
Ischemic heart disease
admissions for age
40+ years.
Total respiratory and
total cardiovascular
admissions for age
30+ years.
Copollutants
Mean O3 Levels Considered
O3 (index not CO, NO2, PM10;
specified): multipollutant
38.9 ppb models
SD 17.8
Low SES:
40.1 ppb
High SES:
38.3 ppb
8-h max O3: PM10, CO, NO2
50.3 ppb
SD30.1
IQR 39.6
24-havg03: PM10,CO,NO2
Winter: 14 ppb
SD7
Spring: 32 ppb
SD10
Summer: 36 ppb
SD8
Fall: 15 ppb
SD9
Lag Structure Method, Findings,
Examined Interpretation
Not specified. Statistical analysis using
naturally occurring seasonal
variations in pollutant
concentrations within zip
codes. Methodology not
clearly stated. Consistent
significant positive effects
only observed for CO.
Negative O3 effect observed
in all age groups. Number of
smog alerts was negatively
associated with asthma
hospitalizations, indicating
avoidance behavior on high
O3 days. Interaction term
with indicator variable for
low SES was positive in all
age groups and statistically
significant in age 3-6 years
and 12-18 years, after
adjusting for number of smog
alerts.
0, 1 , 2, 3, 4, 5, Poisson GAM with cubic
0-1, 0-2, 0-3, B-splines; co-adjustment.
0-4 No significant O3 effects
observed overall or in warm
season. CO and NO2
significant in full-year
analyses.
0 Poisson GLM;
co-adjustment. Only
significant O3 effects
observed were inverse
associations with total
cardiac admission in full-year
and winter season, suggesting
residual confounding.
No significant effects of O3
on respiratory admissions.
Effects
(Relative Risk and 95% CI)
Slope estimate (adjusting for
number of smog alerts):
O3 with CO, NO2, and PM10
models:
Age 3-6 years:
-0.038 (SE 0.014)
Age 6-12 years:
-0.044 (SE 0.013)
Age 12- 18 years:
-0.022 (SE 0.011)
O3 x low SES interaction
term:
Age 3-6 years:
0.092 (SE 0.026)
Age 6-12 years:
0.024 (SE 0.024)
Age 12- 18 years:
0.042 (SE 0.01 9)
O3 coefficients all negative,
but no consistent, significant
effect.
24-h avg O3 (per 10 ppb):
All year:
Respiratory:
1.008(1.000,1.016)
Cardiovascular:
0.993(0.987,0.999)
-------
^**
to
o
o
X
OJ
O
E>
'•Tj
H
6
0
o
H
0
C|
o
w
o
H
w
Reference, Study
Location and Period
United States (cont'd)
Nauenberg and Basu
(1999)
Los Angeles, CA
1991-1994
Sheppard et al.
(1999; reanalysis
Sheppard, 2003)
Seattle, WA
1987-1994
Moolgavkar et al.
(1997)
Minneapolis/St. Paul,
MN and Birmingham,
AL
1986-1991
Schwartz et al. (1996)
Cleveland, OH
Apr-Oct 1988-1990
i 64 years.
Total respiratory
admissions for age
65+ years.
LF1C J"V^V/-t. 1^1 ICV. IS Ul \_/j Ull J
Copollutants
Mean O3 Levels Considered
24-h avg O3: PM10
19. 88 ppb
SDH. 13
8-hmaxO3: PM25,PM10,
30.4 PM10.25, SO2,
IQR 20 CO
24-h avg 03: PM10, SO2,NO2
Minnesota:
26.2 ppb
IQR 15.3
Alabama:
25.1 ppb
IQR 12.7
l-hmaxO3: PM10, SO2
56 ppb
IQR 28
uauy nus|jii
-------
Table AX7-4 (cont'd). Effects of O3 on Daily Hospital Admissions
to
o
o
X
oo
Reference, Study
Location and Period Outcomes and Design
United States (cont'd)
Weisel et al. (2002) Asthma admissions for
New Jersey all ages.
May-Aug 1995
Canada
Burnett et al. (1 997a) Total respiratory
16 Canadian cities admissions for all ages,
1981-1991 age <65 years and
65+ years.
Burnett et al. (1995) Respiratory and
168 Hospitals in cardiovascular
Copollutants
Mean O3 Levels Considered
1-h max O3; Pollen, spores
5-h avg O3
(10 a.m.-3 p.m.);
and 8-h avg O3
(2 p.m.-lO p.m.)
analyzed.
Levels not
reported.
1-h max O3: SO2,NO2,
CO, coefficient
All cities: ofhaze
A
All year;
31 ppb
95th % 60
Apr-Dec only:
33 ppb
95th % 64
1-h max O3: SO42~
36.3 ppb
Lag Structure Method, Findings,
Examined Interpretation
0,1,2,3 No control for time, but
authors report no
autocorrelation, which
alleviates concerns about
lack of control. Significant
O3 effects reported after
adjusting for potential
confounding by pollen.
0, 1, 2, 0-1, Poisson GLM with co-
0-2, 1-2 adjustment. Results stratified
by season. Significant O3
effect observed in warm
season only. No O3 effects
on control outcomes. Results
consistent across cities.
1 GLM with pre-adjustment of
outcome variables. Results
Effects
(Relative Risk and 95% CI)
Slope estimate
(admissions/day /ppb):
5-h avg O3 and 8-h avg O3:
O3 only model:
Lag 2: 0.099, p = 0.057
All three O3 indices:
O3 with pollen model:
Lag 2: 0.1 l,p = 0.033
1-h max O3 (per 30 ppb):
All ages:
Jan-Mar:
Lag 1: 0.994(0.964,1.025)
Apr-Jun:
Lagl: 1.042(1.012,1.073)
Jul-Sep:
Lag 1: 1.050(1.026,1.074)
Oct-Dec:
Lagl: 1.028(0.998,1.059)
No quantitative results
presented for O3.
Ontario, Canada
1983-1988
admissions for all
ages and within age
strata. Study focused
mainly on testing for
sulfate effects.
stratified by season. Authors
report that O3 associated with
respiratory admission in
warm season only.
-------
Table AX7-4 (cont'd). Effects of O3 on Daily Hospital Admissions
to
o
o
Reference, Study
Location and Period
Outcomes and Design Mean O3 Levels
Copollutants
Considered
Lag Structure
Examined
Method, Findings,
Interpretation
Effects
(Relative Risk and 95% CI)
X
VO
Canada (cont'd)
Burnett etal. (1997b)
Toronto, Ontario,
Canada
Summers 1992-1994
Burnett etal. (1999)
Toronto, Ontario,
Canada
1980-1994
Unscheduled
respiratory and
cardiovascular
admissions for all ages.
Cause-specific
respiratory and
cardiovascular
admissions for all ages.
Cause categories
included asthma,
COPD, respiratory
infections, heart failure,
ischemic heart disease,
and cerebrovascular
disease.
l-hmaxO3
41.2ppb
IQR22
24-h avg O3 :
19. 5 ppb
IQR19
PM25,PM10,
S042-, S02,
N02, CO,
coefficient
of haze
H+
Estimated
PM25,PM10,
PM10.25, CO,
NO2, SO2
0,1,2,3,4,
2 to 5 multiday
periods lagged
1 to 4 days
0,1,2,0-1,
0-2,1-2,1-3,
2-3,2-4
Poisson GLM with co-
adjustment. Results stratified
by season. Ozone and
coefficient of haze strongest
predictors of outcomes.
Ozone effects on both
outcomes were robust to PM.
PM effects were not robust to
Poisson GAM with LOESS
pre-filter applied to pollution
and hospitalization data.
Ozone effects seen for
respiratory outcomes only.
Ozone effect robust to PM;
not vice versa. No seasonal
stratification.
12-h avg O3 (8 a.m.-8 p.m.)
(per 11.5 ppb):
Models adjusted for
temperature and dewpoint:
Respiratory :
Lag 1-3: 1.064(1.039,1.090)
Cardiovascular:
Lag 2-4: 1.074(1.035,1.115)
24-h avg O3 (per 19.5 ppb):
1.063(1.036,1.091)
Asthma:
Lag 1-3:
COPD:
Lag 2-4: 1.073(1.038,1.107)
Respiratory infection:
Lag 1-2: 1.044(1.024,1.065)
Burnett etal. (2001)
Toronto, Ontario,
Canada
1980-1994
Acute respiratory
disease admissions for
age <2 years.
l-hmaxO3
45.2 ppb
IQR25
Estimated
PM25,PM10,
PM10.25,CO,
NO2, SO2
1,2,3,4, Poisson GAM with LOESS
5, 0-4 pre-filter applied to pollution
and hospitalization data.
Sensitivity analyses using co-
adjustment. Results stratified
by season. Ozone effects
significant only in summer.
Ozone effect robust to PM;
not vice versa.
1-h max O3 (per 45.2 ppb):
Summer:
O3 only model:
Lag 0-4: 1.348(1.193,1.524)
O3 with PM2 5 model:
Lag 0-4: 1.330(1.131,1.565)
-------
Table AX7-4 (cont'd). Effects of O3 on Daily Hospital Admissions
1
to
o
o
Reference, Study
Location and Period
Canada (cont'd)
Copollutants
Outcomes and Design Mean O3 Levels Considered
Lag Structure
Examined
Method, Findings,
Interpretation
Effects
(Relative Risk and 95% CI)
X
-------
Table AX7-4 (cont'd). Effects of O3 on Daily Hospital Admissions
to
o
o
Reference, Study
Location and Period
Outcomes and Design Mean O3 Levels
Copollutants
Considered
Lag Structure
Examined
Method, Findings,
Interpretation
Effects
(Relative Risk and 95% CI)
X
Canada (cont'd)
Yang et al. (2003)
Vancouver, British
Columbia, Canada
Jan 1986-Dec 1998
Daily respiratory
admissions in children
aged < 3 years and
adults aged 65+ years.
24-havgO3:
13.41 ppb
SD 66.61
IQR 9.74
CO, NO2, SO2, 1,2,3,4,5 Used bidirectional case-
coefficient crossover analysis,
of haze comparing air pollution on
day of admission to levels
one week prior and after.
SES evaluated. O3 was
positively associated with
respiratory hospital
admissions among young
children and the elderly.
24-h avg O3 (per 9.74 ppb):
Odds ratios:
Age < 3 years:
Lag 4: 1.22(1.15,1.30)
Age 65+ years:
Lag 4: 1.13(1.09,1.18)
Europe
Anderson etal. (1997)
Five European cities:
London, Paris,
Amsterdam,
Rotterdam, Barcelona
Study periods vary by
city, ranging from
1977-1992
Anderson et al. (1998)
London, England
1987-1992
Emergency COPD
admissions for all ages.
Each city analyzed
previously by
individual teams.
Results combined here
via meta-analysis.
Admissions for asthma
in all ages and age
0-14, 15-64, and
65+ years.
Range across
five cities:
l-hmaxO3
(median):
All year:
36-77 ug/m3
Warm season:
48-91 ug/m3
Cool season:
20-64 ug/m3
8-h max O3:
15. 5 ppb
IQR 13
l-hmaxO3:
20.6 ppb
IQR 16
TSP, SO2, NO2, 0, 1, 2, 3, 4, 5 Poisson GLM using APHEA
BS methodology. Results
stratified by season.
Ozone most consistent and
significant predictor of
admissions. Warm season
effect larger.
SO2,NO2, 0,1,2,0-1, Poisson GLM using APHEA
BS, pollens 0-2 method; co-adjustment.
Ozone significantly
associated with asthma
admissions in the warm
season for all ages and for
age 15-64 years. Warm
season O3 effect robust in
2-pollutant models. Inverse
associations observed in the
cool season for some age
groups.
1-h max O3 (per 50 ug/m3):
Weighted mean effect across
five cities (best lag selected
for each city):
All year:
1.03(1.01,1.05)
Warm season:
1.03(1.01,1.05)
Cool season:
1.01(0.98,1.05)
8-h max O3 (per 10 ppb):
All ages:
Warm season:
Lag 1: 1.022(1.006,1.038)
Cool season:
Lagl: 0.968(0.946,0.992)
-------
1
to
o
o
Reference, Study
Location and Period
1 ilUIC .M.yv/-t ^lUlll U^. ijllCtlS Ul VJ
Copollutants
Outcomes and Design Mean O3 Levels Considered
'3 Ull LWliy AAUS|JI
Lag Structure
Examined
Hill j-VUIIIISSIUIIS
Method, Findings,
Interpretation
Effects
(Relative Risk and 95% CI)
Atkinson et al.
(1999b)
London, England
1992-1994
Total and cause-
specific respiratory and
cardiovascular
admissions in all ages
and in all ages and
age 0-14, 15-64, and
65+ years.
17.5 ppb
SD11.5
NO2, SO2, CO,
PM10, BS
0,1,2,3,0-1, Poisson GLM using APHEA
0-2, 0-3 methodology. No significant
associations seen between O3
and respiratory admissions.
Ozone was positively
associated with total
cardiovascular admissions in
age 65+ years. Seasonal
analyses were not conducted.
8-h max O3 (per 25.7 ppb):
All ages:
Total respiratory:
Lag 1: 1.012(0.990,1.035)
Total cardiovascular:
Lag 2: 1.023(1.002,1.046)
X
-------
Table AX7-4 (cont'd). Effects of O3 on Daily Hospital Admissions
t—t
1
o
o
X
T]
OJ
o
H
O'
O
O
H
0
0
w
o
o
HH
W
Reference, Study
Location and Period
Europe (cont'd)
Schouten et al. (1996)
Amsterdam and
Rotterdam,
the Netherlands
1977-1989
Hagen et al. (2000)
Drammen, Norway
Nov 1994-Dec 1997
Oftedal et al. (2003)
Drammen, Norway
1995-2000
Ponka and Virtanen
(1996)
Helsinki, Finland
1987-1989
Ballester et al. (2001)
Valencia, Spain
1994-1996
Outcomes and Design
Unscheduled total
respiratory, asthma, and
COPD admissions in
all ages.
Total respiratory
admissions for all ages.
Admissions for
respiratory disease.
Asthma admissions for
age 0-14 years and
15-64 years.
Emergency total
cardiovascular
admissions for all ages.
Mean O3 Levels
l-hmaxO3:
Amsterdam:
Summer:
97 ng/m3
Winter:
62 ng/m3
Rotterdam:
Summer:
96 |ig/m3
Winter:
54 |ig/m3
24-h avg O3:
44.48 ng/m3
SD 18.40
IQR 26.29
24-h avg O3:
44.6 |ig/m3
SD 19.2
IQR 26. 9
O3 (index not
specified):
22 |ig/m3
8-h max O3:
23ppb
Range 5-64
Copollutants
Considered
SO2, NO2, BS
PM10, NO2, SO2,
benzene,
toluene,
formaldehyde
Benzene,
formaldehyde,
toluene, PM10,
NO2, SO2
TSP, S02, N02
SO2, NO2,
CO,BS
Lag Structure Method, Findings,
Examined Interpretation
0,1,2,0-1, Poisson GLM using APHEA
0-2, 0-3, 0-4, methodology; co-adjustment.
0-5 No consistent O3 effects.
Concern regarding multiple
comparisons.
0 Poisson GAM with partial
splines; co-adjustment.
Single and multipollutant
models evaluated. No O3
effects. Ozone levels low
and cycles may not have
been adequately controlled.
0 Benzene had the strongest
association.
0, 1, 2, 3, 4, 5 Poisson GLM using APHEA
methodology. Reported
significant O3 effect for age
0-14 years, but also for
control (digestive disease)
conditions. Ozone levels
very low.
0, 1, 2, 3, 4, 5 Poisson GLM using APHEA
methodology. Results
stratified by season. No O3
Effects
(Relative Risk and 95% CI)
1-h max O3 (per 100 ng/m3):
Amsterdam and Rotterdam:
Total respiratory, all ages:
Summer:
Lag 2: 1.051(1.029,1.073)
Winter:
Lag 2: 0.976(0.951,1.002)
24-h avg O3
(per 26.29 |ig/m3):
LagO: 0.964(0.899-1.033)
24-h avg O3 (per 26.9 |ig/m3):
0.996(0.942,1.053)
Not quantitatively useful.
8-h max O3 (per 5 ppb):
Lag 2: 0.99(0.97-1.01)
effects.
-------
to
o
o
Table AX7-4 (cont'd). Effects of O3 on Daily Hospital Admissions
Reference, Study
Location and Period
Outcomes and Design
Mean O3
Levels
Copollutants
Considered
Lag Structure
Examined
Method, Findings,
Interpretation
Effects
(Relative Risk and 95% CI)
Latin America
Gouveia and Fletcher
(2000a)
Sao Paulo, Brazil
Nov 1992-Sep 1994
Total respiratory, l-hmaxO3:
pneumonia, and asthma 63.4 |ig/m3
admissions for age SD 38.1
< 5 years. IQR 50.3
PM10,N02,
S02, CO
0,1,2 Poisson GLM with co-
adjustment using sine/cosine
waves. Significant O3 effects
on total respiratory and
pneumonia admissions. Ozone
effects fairky robust to NO2
andPM,n.
l-hmaxO3 (per
119.6|ig/m3):
Total respiratory:
LagO: 1.054(1.003,1.107)
Pneumonia:
LagO: 1.076(1.014,1.142)
Asthma:
Lag 2: 1.011(0.899,1.136)
X
Australia
Morgan etal. (1998a)
Sydney, Australia
1990-1994
Petroeschevsky et al.
(2001)
Brisbane, Australia
1987-1994
Admissions for asthma
(age 1-14 years, 15-64
years), COPD (age
65+ years), and heart
disease (all ages, 0-64
years, 65+ years).
Unscheduled asthma,
total respiratory and
total cardiovascular
admissions in several
age strata: all ages, 0-4,
5-14,15-64, 65+ years.
l-hmaxO3:
25 ppb
SD13
IQR 11
Bscatter, NO2
0,1,2,0-1,
0-2
l-hmaxO3:
25.3 ppb
Range 2.5-107.3
8-h avg O3
(10a.m.-
6 p.m.):
19.0 ppb
Range 1.7-64.7
Better, S°2,
NO,
,1,2,3,0-2,
0-4
Poisson with GEE.
No significant effects of O3
in single or multipollutant
models.
Poisson GLM using APHEA
co-adjustment methodology.
Results stratified by season.
Ozone significantly related to
asthma and total respiratory
admissions, not for cardiac
admissions. Effects varied by
age group. Ozone effects
robust to copollutants.
1-h max O3 (per 28 ppb):
Asthma, age 1-14 years:
Lag 1: 0.975(0.932,1.019)
Asthma, age 15-64 years:
LagO: 1.025(0.975,1.078)
COPD, age 65+ years:
LagO: 1.010(0.960,1.062)
Heart disease, all ages:
LagO: 1.012(0.990,1.035)
8-h avg O3 (per 10 ppb):
All ages:
Total respiratory:
Lag 2: 1.023(1.003,1.043)
Asthma:
Lag 0-4: 1.090(1.042,1.141)
Total cardiovascular:
Lag 3: 0.987(0.971,1.002)
-------
1
to
o
o
1 ilLMC ±\-t
Reference, Study
Location and Period Outcomes and Design
-V/-t ^lUlll U
Mean O3
Levels
If. ijllCtlS Ul V
Copollutants
Considered
J3 Ull ifail^ AAUS|
Lag Structure
Examined
Illtll j-VUIIIISSIUIIS
Method, Findings,
Interpretation
Effects
(Relative Risk and 95% CI)
Wongetal. (1999a)
Hong Kong
1994-1995
Total and cause-
specific respiratory and
cardiovascular
admissions in several
age strata: all ages, 0-4,
5-64, 65+ years.
8-h max O3:
20.2 ng/m3
Median 24.15
IQR31.63
N02, S02,
PM,n
0,1,2,3,4,5,
0-1,0-2,0-3,
0-4, 0-5
Poisson GLM using APHEA
methodology. Ozone
significantly associated with
total and cause specific
respiratory and cardiac
outcomes. Ozone results
robust to adjustment for high
PM10, but not high NO2.
Effects of O3 persisted in
cold season.
8-hmaxO3(perlO|ig/m3):
All ages:
Total respiratory:
Lag 0-3: 1.022(1.015,1.029)
Total cardiovascular:
Lag 0-5: 1.013(1.005,1.021)
X
Wongetal. (1999b)
Hong Kong
Jan 1995-Jun 1997
Total and cause-
specific cardiovascular
admissions in all ages.
O3 (index not NO2, SO2, 0, 1, 2, 3, 4, 5, GLM with sinusoids;
specified): respirablePM 0-1,0-2,0-3, co-adjustment. Ozone
0-4, 0-5 significantly associated with
Warm season: total and cause-specific
31.2 |ig/m3 cardiovascular admissions in
Cool season: cool season only, when O3
34.8 |ig/m3 levels are higher in Hong
Kong. Details missing in brief
report.
O3 (per 50 ng/m3):
O3 with NO2 models:
Total cardiovascular:
All year:
Lag 0-1: 1.03(1.00,1.07)
Warm season:
Lag 0-1: 1.01 (0.95,1.07)
Cool season:
Lag 0-1: 1.08(1.02,1.14)
-------
1
to
o
o
Reference, Study
Location and Period
i
O
E>
Tj
H
6
o
21
O
H
O
0
H
W
O
O
H
W
Samet et al. (2000; All cause; 24-h avg O3: PM10, NO2, SO2,
reanalysis Dominici et al., cardiopulmonary Mean range: CO; 2-pollutant
2003) Approximately 12 ppb models
90 U.S. cities (Des Moines, IA) to 36
1 987-1 994 ppb (San Bernardino,
CA)
0,1,2 Poisson GAM 24-h avg O3 (per 1 0 ppb):
(reanalyzed with
stringent Posterior means:
convergence
criteria); Poisson All cause (results given in
GLM graphic format):
All year:
Lag 0: approximately 0.4%
Lag 1: approximately 0.2%
Summer:
Lag 1: approximately 0.5%
Winter:
Lag 1: approximately -0.5%
-------
Table AX7-5 (cont'd). Effects of Acute O3 Exposure on Mortality
to
o
o
Reference, Study
Location and Period
Outcome
Measure
Mean O, Levels
Copollutants
Considered
Lag Structure
Reported
Method
Effect Estimates
X
United States (cont'd)
Huang et al. (2004)
19 U.S. cities
Jun-Sept 1987-1994
Schwartz (2004)
14 U.S. cities
1986-1993
Cardiopulmonary
24-havgO3:
26ppb
PM10, PM2.5; 2-
pollutant models
0,1,2,0-6
Poisson GLM;
Bayesian
hierarchical
model
All cause
l-hmaxO3:
Median range:
35.1 ppb (Chicago, IL)
to 60.0 ppb (Provo, UT)
PM10; 2-pollutant
models
Case-crossover
analysis;
controlled for
temperature
using nonlinear
regression splines
and matching
24-h avg O3 (per 10 ppb):
Posterior means:
O3 only model:
LagO: 0.73% (0.27, 1.19)
O3 with PM10 model:
LagO: 0.74% (-0.33, 1.72)
O3 only model:
Lag 0-6: 1.25% (0.47, 2.03)
Model adjusted for heat
waves:
Lag 0-6: 1.11%(0.38, 1.86)
l-hmaxO3(10ppb):
Analysis with temperature
regression splines:
All year:
0.19% (0.03, 0.35)
Warm season:
H
6
o
o
H
O
0
H
W
o
O
l— '
H
W
0.26% (0.07, 0.44)
Cold season:
0% (-0.27, 0.27)
Analysis with temperature
matched controls:
All year:
0.23% (0.01, 0.44)
Warm season:
0.37% (0.1 1,0.62)
Cold season:
-0.13% (-0.53, 0.28)
-------
Table AX7-5 (cont'd). Effects of Acute O3 Exposure on Mortality
e
to
o
o
>
£j
4^
oo
O
^
'•Tj
H
6
H
0
0
H
W
O
O
H
W
Reference, Study Outcome
Location and Period Measure
United States (cont'd)
Kinney and Ozkaynak All cause;
(1991) respiratory;
Los Angeles County, CA circulatory
1970-1979
Kinney et al. (1995) All cause
Los Angeles County, CA
1985-1990
Ostro(1995) All cause
San Bernardino County and
Riverside County, CA
1980-1986
Fairley (1999; reanalysis All cause;
Fairley, 2003) respiratory;
Santa Clara County, CA cardiovascular
1989-1996
Mean O3 Levels
1-h max total oxidants
(Ox):
75 ppb
l-hmaxO3:
70 ppb
l-hmaxO3:
140 ppb
8-h max O3:
29 ppb
24-havgO3:
16 ppb
O3 ppb-hours > 60 ppb:
Levels not reported.
Copollutants
Considered
KM (particle
optical reflectance),
N02, S02, CO;
multipollutant
models
PM10, N02, S02,
CO; 2-pollutant
models
PM25
PM10,PM25,
PM10.2,, S042-,
coefficient of haze,
NO3", NO2, SO2; 2-
pollutant models
Lag Structure
Reported Method
1 OLS (ordinary
least squares) on
high-pass filtered
variables
1 Linear, log-
linear, and
Poisson
0 Autoregressive
linear; Poisson
0 Poisson GAM
(reanalyzed with
stringent
convergence
criteria);
Poisson GLM
Effect Estimates
All cause:
Multipollutant model:
Slope estimate:
0.03 deaths/ppb (SE 0.009),
p = 0.0005
1-h max O3 (per 143 ppb):
O3 only model:
2% (0, 5)
O3 with PM10 model:
0%(-6, 6)
1 -h max O3 (per 100 ppb):
Warm season:
2.0% (0.0, 5.0)
GLM:
All cause:
8-h max O3 (per 33 ppb):
3.3% (-0.3, 7.0)
O3 ppb-hours > 60 ppb
(increment not reported):
4.1% (1.5, 6.7)
-------
r**
1
to
O
o
Reference, Study
Location and Period
United States (cont'd)
Gamble (1998)
Dallas, TX
1990-1994
1 ilLMC
Outcome
Measure
All cause;
respiratory;
cardiovascular;
cancer; other
J"V^V/-J ^lUlll U^. Hill
Mean O3 Levels
24-h avg O3:
All year:
22 ppb
Summer:
30 ppb
Winter:
12 ppb
CtlS Ul .fY^UlC V^3 1
Copollutants
Considered
PM10, NO2, SO2,
CO; 2-pollutant
models
^A|JUSUI C Ull 1TAUI lillll V
Lag Structure
Reported Method Effect
1-2 Poisson GLM All cause:
All year:
24-h avg O3
2.7% (0.6, 4
Summer:
24-h avg O3
Estimates
(per 14.7 ppb):
.8)
(per 13.1 ppb):
X
3.5% (p< 0.05)
Winter:
24-h avg O3 (per 7.7 ppb):
2.4% (p > 0.05)
k
4^
VO
M
^^
fe
o
2!
•£—\
O
H
0
C
0
H
W
O
O
H
W
Dockery etal. (1992)
St. Louis, MO and Eastern
Tennessee
1985-1986
Ito and Thurston (1996)
Cook County, IL
1985-1990
All cause
All cause;
respiratory;
circulatory;
cancer;
race/gender
subcategories
24-h avg O3 : PM10, PM2 5, SO42',
H+, NO2, SO2
St. Louis, MO:
22.5 ppb
Eastern Tennessee:
23.0 ppb
l-hmax03: PM10, NO2, SO2,
3 8.1 ppb CO; 2-pollutant
models
1 Poisson with 24-h avg O3 (per 20
GEE
St. Louis, MO:
0.6% (t = 0.38)
Eastern Tennessee:
-1.3%(t=-0.37)
0-1 Poisson GLM 1-h max O3 (per 100
All cause:
O3 only model:
10% (6, 15)
O3 with PM10 model:
7% (1,12)
|ig/m3):
ppb):
-------
3
Table AX7-5 (cont'd). Effects of Acute O3 Exposure on Mortality
to
o
o
Reference, Study
Location and Period
Outcome
Measure
Mean O3 Levels
Copollutants
Considered
Lag Structure
Reported
Method
Effect Estimates
United States (cont'd)
Lippmann et al. (2000;
reanalysis Ito, 2003)
Detroit, MI
1985-1990
1992-1994
All cause;
respiratory;
cardiovascular;
cause-specific
24-havgO3:
25ppb
PM10,PM25,
PM10.25,S042-,H+,
N02, S02, CO;
2-pollutant models
0,1,2,3,0-1,
0-2, 0-3
Poisson GAM
(reanalyzed with
stringent
convergence
criteria); Poisson
GLM
24-h avg O3 (per 5th to
95th % increment):
GAM with stringent
convergence criteria:
For all lags and outcomes
during both study periods
(n= 140):
Median 1.6%
Range-1.8-2.6
X
Lipfert et al. (2000a)
Seven counties in
Philadelphia, PA area
1991-1995
Moolgavkar et al. (1995)
Philadelphia, PA
1973-1988
All cause;
respiratory;
cardiovascular;
all ages; age 65+
years; age < 65
years; various
subregional
boundaries
All cause
l-hmaxO3:
44.76 ppb
24-h avg O3:
23. 44 ppb
24-h avg O3:
Spring:
25. 9 ppb
Summer:
35.5 ppb
Fall:
16.2 ppb
Winter:
11. 9 ppb
PM10,PM25,
PM10.2.5, S042-,
NCV, other PM
indices, NO2, SO2,
CO; 2-pollutant
models
TSP, SO2;
multipollutant
models
0-1 Linear with 19-
day weighted
average
Shumway filters
1 Poisson; GEE
and
nonparametric
bootstrap
methods
1-h max O3 (per 45 ppb less
background, not reported):
All cause, all ages:
O3 only model:
3.19%, p< 0.055
O3 with PM2 5 model:
3. 34%, p< 0.055
24-h avg O3 (per 100 ppb):
O3 with TSP and SO2
models:
Spring:
2.0% (-6.7, 11.5)
Summer:
14.9% (6.8, 23.6)
Fall:
-4. 5% (-13. 9, 5.9)
Winter:
0.4% (-15. 6, 19.4)
-------
1
to
o
o
Reference, Study
Location and Period
1 ilLMC t
Outcome
Measure
l.yv/-j ^lUlll U). ijllCtlS Ul .fY^UlC V^3 1
Copollutants
Mean O3 Levels Considered
^A|JUSUI C Ull 1TAUI lillll V
Lag Structure
Reported Method
Effect Estimates
Chock et al. (2000)
Pittsburgh, PA
1989-1991
All cause; age
< 74 years;
age 75+ years
l-hmaxO3:
Levels not reported.
PM10, NO2, SO2,
CO; 2-, 5-, and
6-pollutant models
Poisson GLM
1-h max O3 (per 40 ppb):
Age <74 years:
O3 only model:
-1.5% (t=-0.68)
O3 with PM10 model:
-2.0% (t=-0.93)
Age 75+ years:
O3 only model:
-1.8% (t=-0.82)
O3 with PM10 model:
-2.2% (t=-0.98)
X
De Leon et al. (2003)
New York City
1985-1994
Circulatory and
cancer with and
without
contributing
respiratory causes
24-havgO3:
21.59ppb
PM10, NO2, SO2,
CO; 2-pollutant
models
Oorl
Poisson GAM;
Poisson GLM
Quanititative results not
given.
Circulatory deaths:
Larger O3 effect estimates
with contributing respiratory
causes than without (RS.
non-significant).
Cancer deaths:
Smaller O3 effect estimates
with contributing respiratory
causes than without (PJl
non-significant).
-------
Table AX7-5 (cont'd). Effects of Acute O3 Exposure on Mortality
to
o
o
X
to
H
6
o
o
H
O
O
H
W
O
O
HH
H
W
Reference, Study
Location and Period
United States (cont'd)
Klemm and Mason (2000);
Klemm et al. (2004)
Atlanta, GA
Aug 1998- July 2000
Canada
Vedal et al. (2003)
Vancouver, British
Columbia, Canada
1994-1996
Goldberg et al. (2003)
Montreal, Quebec, Canada
1984-1993
Outcome
Measure Mean O3 Levels
All cause; 8-h max O3:
respiratory; 47.03 ppb
cardiovascular;
cancer; other; age
< 65 years; age
65+ years
All cause; 1 -h max O3 :
respiratory; 27.3 ppb
cardiovascular
Congestive heart 24-havgO3:
failure as 29 |ig/m3
underlying cause
of death versus
those classified as
having congestive
heart failure one
year prior to death
Copollutants Lag Structure
Considered Reported Method Effect Estimates
PM25,PM10.25,EC, 0-1 PoissonGLM All cause, age 65+ years:
OC, NO2, SO42~, using quarterly,
NO3', SO2, CO monthly, or Quarterly knots:
biweekly knots Slope estimate:
for temporal 0.00079 (SE 0.00099),
smoothing t = 0.80
Monthly knots:
Slope estimate:
0.00136 (SE 0.001 11),
t=1.22
PM10, NO2, SO2, 0,1,2 PoissonGAM 1-h max O3 (per 10.2 ppb):
CO
All cause:
Summer:
LagO: 4.2% (1.4, 7.0)
Winter:
LagO: 0.5% (-1.9, 3.0)
TSP, coefficient of 0-2 PoissonGLM 24-h avg O3 (per 2 1.3
haze, PM10, SO42~, Hg/m3):
S02,N02,CO
Congestive heart failure as
underlying cause of death:
4.54% (-5.64, 15.81)
Having congestive heart
failure one year prior to
death:
2.34% (-1.78, 6.63)
-------
1
to
o
o
Reference, Study
Location and Period
1 ilLMC t
Outcome
Measure
l.yv/-j ^lUlll U). ijllCtlS Ul .fY^UlC V^3 1
Copollutants
Mean O3 Levels Considered
^A|JUSUI C Ull 1TAUI lillll V
Lag Structure
Reported Method
Effect Estimates
X
Gryparis et al. (2004)
23 European cities
Study periods vary by city,
ranging from 1990-1997
All cause;
respiratory;
cardiovascular
l-hmaxO3:
Median range:
Summer:
44 ppb (Tel Aviv, Israel)
to 117 ppb (Torino,
Italy)
Winter:
11 ppb (Milan, Italy) to
57 ppb (Athens, Greece)
8-h max O3:
Median range:
Summer:
30 ppb (Rome, Italy) to
99 ppb (Torino, Italy)
Winter:
8 ppb (Milan, Italy) to
49 ppb (Budapest,
Hungary)
PM10, NO2, SO2,
CO; 2-pollutant
models
0-1
Poisson GAM;
Bayesian
hierarchical
model
8-h max O3 (per 10 ug/m3):
Weighted mean effect
across 21 cities with 8-h
max O3 concentrations:
Random effects model:
All cause:
All year:
0.03% (-0.18, 0.21)
Summer:
O3 only model:
0.31% (0.17, 0.52)
O3 with PM10 model:
0.27% (0.08, 0.49)
Winter:
O3 only model:
0.12% (-0.12, 0.37)
O3 with PM10 model:
0.22% (-0.08, 0.51)
Touloumi et al. ( 1 997) All cause
Four European cities:
London, Paris
Barcelona, Athens
Study periods vary by city,
ranging from 1986-1992
l-hmaxO3:
London:
4 1.2 ug/m3
Paris:
46.1 ug/m3
Barcelona:
72.4 ug/m3
Athens:
93.8 ug/m3
BS,NO2; 0,1,2,3,0-1, Poisson l-hmaxO3 (50 ug/m3):
2-pollutant models 0-2, 0-3 autoregressive
Weighted mean effect
across four cities (best lag
selected for each city):
Random effects model:
O3 only model:
2. 9% (1.0, 4.9)
O3 with BS model:
2. 8% (0.5, 5.0)
-------
^**
to
o
o
X
r N
L/l
[\
O
£5
H
6
0
2;
o
H
0
0
w
o
o
H
w
1 ilLMC
Reference, Study Outcome
Location and Period Measure
Europe (cont'd)
Zmirouetal. (1998) Respiratory;
Four European cities: cardiovascular
London, Paris, Lyon,
Barcelona
Study periods vary by city,
ranging from 1985-1992
Anderson et al. ( 1 996) All cause;
London, England respiratory;
1987-1992 cardiovascular
J"V^V/-J ^lUlll U^. Hill
Mean O3 Levels
8-h avg O3 (9 a.m.-
5 p.m.):
London:
Cold: 2 1.0 ug/m3
Warm: 40.8 ug/m3
Paris:
Cold: 11.5 ug/m3
Warm: 42.7 ug/m3
Lyon:
Cold: 2 1.0 ug/m3
Warm: 40.8 ug/m3
Barcelona:
Cold: 51.5 ug/m3
Warm: 89.7 ug/m3
l-hmaxO3:
20.6 ppb
8-h avg O3 (9 a.m.-
5p.m.): 15. 5 ppb
ctis ui .rvtuic v^3 tiA|ju»uic uii IVAUI liiiuy
Copollutants Lag Structure
Considered Reported Method Effect Estimates
BS, TSP, SO2, NO2 0,1,2,3,0-1, Poisson GLM 8-h avg O3 (per 50 ug/m3):
0-2, 0-3
Weighted mean effect
across four cities (best lag
selected for each city):
Random effects model:
Respiratory:
5% (2, 8)
Cardiovascular:
2% (0, 3)
BS,NO2, SO2; 0 Poisson GLM All cause:
2-pollutant models
All year:
1-h max O3 (per 31 ppb):
2.59% (1.30, 3.89)
Warm season:
1-h max O3 (per 34 ppb):
3.49% (1.81, 5.20)
Cool season:
1-h max O3 (per 26 ppb):
0.99% (-0.80, 2.81)
-------
^**
to
o
o
^
X
^
£
O
>
H
1
O
0
2|
O
H
O
H
W
O
O
H
W
Reference, Study
Location and Period
Europe (cont'd)
Bremneretal. (1999)
London, England
1992-1994
Prescott etal. (1998)
Edinburgh, Scotland
1992-1995
Dab etal. (1996)
Paris, France
1987-1992
Zmirou etal. (1996)
Lyon, France
1985-1990
1 ilLMC /
Outcome
Measure
All cause;
respiratory;
cardiovascular; all
cancer; all others;
all ages; age
specific (0-64,
65+, 65-74,
75+ years)
All cause;
respiratory;
cardiovascular; all
ages, age < 65
years, age 65+
years
Respiratory
All cause;
respiratory;
cardiovascular;
digestive
^^V/-J ^lUlll U^. Hill
Mean O3 Levels
l-hmaxO3:
22.6 ppb
8-h max O3:
17.5 ppb
24-havgO3:
14.5 ppb
l-hmaxO3:
23.2 ng/m3
8-h max O3:
11.5 |ig/m3
l-hmaxO3:
15.23 ng/m3
8-h avg O3
(9 a.m.-5 p.m.):
9.94 ng/m3
CtlS Ul .fY^UlC V^3 HiA|JU»UIC Ull 1VAUI lillllV
Copollutants Lag Structure
Considered Reported Method Effect Estimates
BS, PM10,NO2, Selected best Poisson GLM 8-h max O3 (per 26 ppb):
S02,CO; fromO, 1,2, 3,
2-pollutant models (all cause); All ages:
0,1,2,3,0-1, All cause:
0-2,0-3 Lag 2: -0.7% (-2. 3, 0.9)
(respiratory, Respiratory:
cardiovascular) Lag 2: -3. 6% (-7.7, 0.8)
Cardiovascular:
Lag 2: 3. 5% (0.5, 6.7)
BS,PM10,NO2, 0 Poisson GLM 24-h avg O3 (per 10 ppb):
S02, CO; 2-
pollutant models All cause, all ages:
-4.2% (-8.1, -0.1)
BS,PM13,NO2, 0 Poisson l-hmaxO3 (per 100 |ig/m3):
SO2, CO autoregressive 1 .074 (0.934, 1 .235)
8-h max O3 (per 100 |ig/m3):
1.040(0.934,1.157)
PM13, SO2,NO2 Selected best Poisson GLM 8-h avg O3 (per 50 |ig/m3):
fromO, 1,2, 3
All cause:
LagO: 3% (-5, 12)
Respiratory:
Lagl: l%(-8, 10)
Cardiovascular:
Lagl: 0%(-11,12)
-------
Table AX7-5 (cont'd). Effects of Acute O3 Exposure on Mortality
to
o
o
Reference, Study
Location and Period
Outcome
Measure
Mean O, Levels
Copollutants
Considered
Lag Structure
Reported
Method
Effect Estimates
X
ON
o
o
2
o
H
O
o
H
W
O
O
HH
H
W
Europe (cont'd)
Sartor etal. (1995)
Belgium
Summer 1994
Hoek et al, (2000;
reanalysis Hoek, 2003)
The Netherlands: entire
country, four urban areas
1986-1994
All cause; age
< 65 years;
age 65+ years
All cause; COPD;
pneumonia;
cardiovascular
24-havgO3:
During heat wave (42
day period): 72.4 |ig/m3
Before heat wave (43
day period): 52.4 |ig/m3
After heat wave (39 day
period): 38.6 |ig/m3
8-h avg O3 (12 p.m.-
8 p.m.):
Median: 47 |ig/m3
TSP, NO, NO2, SO2
PM10, BS, SO42%
N03-, N02, S02,
CO; 2-pollutant
models
0,1,2
1,0-6
Log-linear
regression
Poisson GAM
(reanalyzed
with stringent
convergence
criteria);
Poisson GLM
No individual regression
coefficient for O3 alone;
interaction with temperature
24-havgO3(froml8.8to
111.5 |ig/m3)and
temperature (from 10.0 to
27.5°C):
Age < 65 years:
Lag 1: 16% increase in
mortality (5.3% expected)
Age 65+ years:
Lag 1: 36.5% increase in
mortality (4% expected)
GLM:
All cause:
8-h avg O3 (per 150 |ig/m3):
Lagl: 4.3% (2.4, 6.2)
8-h avg O3 (per 120 |ig/m3):
Lag 0-6: 5.9% (3.1, 8.7)
-------
r**
1
to
O
o
Reference, Study
Location and Period
Europe (cont'd)
Hoeketal. (2001;
reanalysis Hoek, 2003)
The Netherlands
1986-1994
1 ilLMC t
Outcome
Measure
Total
cardiovascular;
myocardial
infarction;
arrhythmia; heart
failure;
cerebro vascular;
thrombosis-related
Y^V/-J ^Ulll U^. ijllCtlS Ul .fY^UlC \J^ I
Copollutants
Mean O3 Levels Considered
8-havgO3(12p.m.- PM10, NO2, SO2,
8 p.m.): CO
Median: 47 |ig/m3
^A|JUSUI C Ull 1TAUI lillll V
Lag Structure
Reported Method
1 Poisson GAM
(reanalyzed with
stringent
convergence
criteria); Poisson
GLM
Effect Estimates
8-h avg O3 (per 150 |ig/m3):
GLM:
Total cardiovascular:
6.2% (3.3, 9.2)
Myocardial infarction:
4. 3% (0.1, 8.6)
X
Arrhythmia:
11.4% (-1.2, 25.5)
Heart failure:
10.2% (1.2, 19.9)
Cerebrovascular:
9.1% (2.9, 15.7)
Thrombosis-related:
16.6% (2.8, 32.2)
L
>
H
O'
O
O
H
0
"
O
x^
H
W
O
O
H
W
Roemer and van Wijinen
(2001)
Amsterdam, the
Netherlands
1987-1998
Verhoeffetal. (1996)
Amsterdam, the
Netherlands
1986-1992
All cause 8-h max O3:
Background sites:
43 lig/m3
Traffic sites:
36 |ig/m3
All cause; all ages; 1-h max O3:
age 65+ years 43 |ig/m3
BS, PM10, NO2, 1 , 2, 0-6 Poisson GAM
SO2, CO (default
convergence
criteria but with
only one
smoother)
PM10, NO2, SO2, 0,1,2 Poisson
CO; multipollutant
models
8-h max O3 (per 100 |ig/m3):
Total population using
background sites:
Lagl: -0.3% (-4. 1,3.7)
Total population using
traffic sites:
Lagl: 0.2% (-3.6, 4.2)
1-h max O3 (per 100 |ig/m3)
All ages:
LagO: 1.8% (-3. 8, 7.8)
Lagl: 0.1% (-4.7, 5.1)
Lag 2: 4.9% (0.1, 10.0)
-------
Table AX7-5 (cont'd). Effects of Acute O3 Exposure on Mortality
t—t
1
to
O
C2
v\
X
(!/i
00
O
>
H
M
^^
0
O
H
0
crj
O
H
W
O
O
HH
H
W
Reference, Study
Location and Period
Europe (cont'd)
Peters et al. (2000b)
NE Bavaria, Germany and
coal basin in Czech
Republic
1982-1994
Ponkaetal. (1998)
Helsinki, Finland
1987-1993
Garcia-Aymerich et al.
(2000)
Barcelona, Spain
1985-1989
Saezetal. (1999)
Barcelona, Spain
1986-1989
Sunyeretal. (1996)
Barcelona, Spain
1985-1991
Outcome
Measure
All cause for
Czech Republic;
all cause and
cardiovascular for
Bavaria, Germany
All cause;
cardiovascular;
age < 65 years,
age 65+ years
All cause;
respiratory;
cardiovascular;
general
population;
patients with
COPD
Asthma mortality;
age 2-45 years
All cause;
respiratory;
cardiovascular;
all ages;
age 70+ years
Mean O3 Levels
24-h avg O3:
Czech Republic:
40.3 ng/m3
Bavaria, Germany:
38.2 ng/m3
24-h avg O3:
Median: 18 |ig/m3
l-hmaxO3:
Levels not reported.
l-hmaxO3:
Levels not reported.
l-hmaxO3:
Summer:
86.5 ng/m3
Winter:
55.2 |ig/m3
Copollutants Lag Structure
Considered Reported Method Effect Estimates
TSP, PM10, NO2, 0, 1, 2, 3 Poisson GLM 24-h avg O3 (per
SO2,CO 100ng/m3):
All cause:
Czech Republic:
Lag 2: 7.8% (-1.8, 18.4)
Bavaria, Germany:
LagO: 8.2% (0.4, 16.7)
TSP, PM10, NO2, 0, 1 , 2, 3, 4, 5, Poisson GLM 24-h avg O3 (per 20 |ig/m3):
S02 6, 7
All cause, age < 65 years:
Not significant, values not
reported.
Cardiovascular, age < 65
years:
Lag 5: -11.7% (-18.9, -3.9)
Lag 6: 9.9% (1.1, 19.5)
BS,NO2, SO2, 5 (general Poisson GLM l-hmaxO3 (per 50 |ig/m3):
population);
3 (COPD cohort) General population:
2.4% (0.6, 4.2)
COPD patients:
4.0% (-4.7, 13.4)
BS,NO2, SO2, 0-2 Poisson with Slope estimate:
GEE 0.021 (SE 0.01 l),p = 0.054
BS, SO2, NO2 0, 1, 5 Autoregressive 1-h max O3 (per 100 |ig/m3):
Poisson
All cause, all ages:
All year:
LagO: 4.8% (1.2, 8.6)
Summer:
LagO: 5. 8% (1.7, 10.1)
Winter:
LagO: 2.6% (-3. 5, 9.1)
-------
X
VO
Diaz etal. (1999)
Madrid, Spain
1990-1992
Table AX7-5 (cont'd). Effects of Acute O3 Exposure on Mortality
t—t
1
to
o
o
Reference, Study
Location and Period
Europe (cont'd)
Sunyer and Basagana
(2001)
Barcelona, Spain
1990-1995
Sunyer et al. (2002)
Barcelona, Spain
1986-1995
Outcome
Measure
Mortality in a
cohort of patients
with COPD
Mortality in a
cohort of patients
with severe
asthma
Mean O3 Levels
l-hmaxO3:
Mean not reported
IQR21 ng/m3
l-hmaxO3:
Median: 69.3 |ig/m3
8-h max O3:
Copollutants
Considered
PM10, NO2, CO
PM10, BS, SO2,
N02, CO, pollen
Lag Structure
Reported Method
0-2 Conditional
logistic (case-
crossover)
0-2 Conditional
logistic (case-
crossover)
Effect Estimates
1-hmax O3(per21 jig/m3):
Odds ratio:
0.979(0.919,1.065)
1-hmax O3 (per 48 jig/m3):
Odds ratios:
Patients with only one
All cause;
respiratory;
cardiovascular
Median: 54.4 |ig/m3
24-havgO3:
Levels not reported.
TSP, NO,, SO,, CO
1,4,10
Autoregressive
linear
admission:
1.096(0.820,1.466)
Patients with more than
one admission:
1.688(0.978,2.643)
24-h avg O3 (per 25 |ig/m3):
For O3 levels higher than 35
H
6
o
o
H
O
O
H
W
O
O
HH
H
W
Latin America
Borja-Aburto et al. (1997)
Mexico City
1990-1992
All cause; all ages;
age < 5 years; age
> 65 years
l-hmaxO3:
Median 155 ppb
8-h max O3:
Median 94 ppb
10-h avg O3 (8a.m.-
6 p.m.):
Median 87 ppb
24-h avg O3:
Median 54 ppb
TSP, SO2, CO;
2-pollutant models
0,1,2
Poisson
iteratively
weighted and
filtered least-
squares method
All cause:
Lag 4: 12% (p < 0.01)
U-shaped (quadratic) O3-
mortality relationship with a
minimum of 35 |ig/m3.
1-h max O3 (per 100 ppb):
All ages:
O3 only model:
LagO: 2.4% (1.1, 3.9)
O3 with TSP model:
LagO: -1.8% (-10.0, 6.4)
-------
Table AX7-5 (cont'd). Effects of Acute O3 Exposure on Mortality
t—t
1
to
o
o
Reference, Study
Location and Period
Latin America (cont'd)
Borja-Aburto et al. (1998)
SW Mexico City
1993-1995
Outcome
Measure
All cause;
respiratory;
cardiovascular;
other; all ages;
age > 65 years
Mean O3 Levels
l-hmaxO3:
163 ppb
24-h avg O3:
44 ppb
Copollutants Lag Structure
Considered Reported Method
PM25, NO2, SO2; 0, 1, 2, 3, 4, 5, Poisson GAM
2-pollutant models 1-2 (default
convergence
criteria but with
only one
smoother)
Effect Estimates
24-h avg O3 (per 10 ppb):
All cause, all ages:
Lag 1-2: 0.6% (-0.3, 1.5)
All cause, age > 65 years:
Lag 1-2: 0.8% (-0.4, 2.0)
Respiratory, all ages:
Lag 1-2: -0.7% (-3.6, 2.1)
Cardiovascular, all ages:
Lag 1-2: 1.8% (0.1, 3.5)
Other noninjury, all ages:
Lag 1-2: 0.3% (-0.9, 1.4)
^ O'Neill et al. (2004)
Jx; Mexico City
T-" 1996-1998
All cause; all ages;
age 65+ years;
SES gradient
24-h avg O3:
35.3 ppb
PM,,
0-1
Poisson GAM
24-h avg O3 (per 10 ppb):
All ages:
0.65% (0.02, 1.28)
Age 65+ years:
1.39% (0.51, 2.28)
SES gradient did not show
any consistent pattern.
H
o
^^
o
H
0
c|
o
H
W
O
^
O
H
W
Tellez-Rojo et al. (2000)
Mexico City
1994
Gouveia and Fletcher
(2000b)
Sao Paulo, Brazil
1991-1993
Respiratory ; 1 -h max O3 :
COPD mortality; 1 34.5 ppb
age 65+ years;
within medical
unit; outside of
medical unit
All ages (all l-hmaxO3:
cause); age 67.9 |ig/m3
< 5 years (all
cause, respiratory,
pneumonia); age
65+ years (all
cause, respiratory,
cardiovascular)
PM10, N02, S02 1, 2, 3, 4, 5, 1-3, Poisson,
1-5, 1-7 iteratively
weighted and
filtered least-
squares method
PM10, NO2, SO2, 0, 1, 2 Poisson GLM
CO
1-h max O3 (per 40 ppb):
Outside medical unit:
Respiratory:
Lag 1-5: 14.0% (4.1, 24.9)
COPD mortality:
Lag 1-5: 15. 6% (4.0, 28.4)
1-h max O3 (per 106 |ig/m3):
All cause, all ages:
LagO: 0.8%(-1. 1,2.7)
All cause, age 65+ years:
LagO: 2.3% (0,4.6)
-------
Table AX7-5 (cont'd). Effects of Acute O3 Exposure on Mortality
t—t
1
to
o
o
X
7"1
^
o
>
H
o
0
o
H
0
H
W
O
^
O
w
Reference, Study
Location and Period
Latin America (cont'd)
Pereiraetal. (1998)
Sao Paulo, Brazil
1991-1992
Saldivaetal. (1994)
Sao Paulo, Brazil
1990-1991
Saldivaetal. (1995)
Sao Paulo, Brazil
1990-1991
Cifuentes et al. (2000)
Santiago, Chile
1988-1966
Ostroetal. (1996)
Santiago, Chile
1989-1991
Outcome
Measure Mean O3 Levels
Intrauterine 1 -h max O3 :
mortality 67.5 \igfm3
Respiratory; age 24-havgO3:
< 5 years 12. 14 ppb
All cause; age l-hmaxO3:
65+ years 38.3 ppb
24-havgO3:
12.5 ppb
All cause l-hmaxO3:
Summer:
108.2 ppb
All cause l-hmaxO3:
52.8 ppb
Copollutants
Considered
PM10, N02, S02,
CO
PM10, N02, S02,
CO; multipollutant
models
PM10, NO2, SO2,
CO; 2-pollutant
models
PM25,PM10.25, CO,
SO2,NO2
PM10, NO2, SO2; 2-
pollutant models
Lag Structure
Reported Method
0 Poisson,
linear with
M-estimation
0-2 OLS of
transformed data
0-1 OLS; Poisson
with GEE
0, 1, 2, 3, 4, 5, Poisson GAM
1-2, 1-3, 1-4, 1-5 (default
convergence
criteira);
Poisson GLM
1 OLS, several
other methods
Effect Estimates
Slope estimate:
0.0000 (SE 0.0004)
Slope estimate:
0.01048 deaths/day/ppb
(SE 0.02481 ),p = 0.673
Slope estimate:
l-hmaxO3:
0.0280 deaths/day/ppb
(SE 0.0213), p> 0.05
24-havgO3:
0.0093 deaths/day/ppb
(SE 0.0813), p> 0.05
1-h max O3 per (108.2 ppb):
GLM:
Summer:
O3 only model:
Lag 1-2: 0.3%(t=0.3)
Multipollutant model:
Lag 1-2: -0.1% (t= -0.1)
All year:
1-h max O3 (per 52.8 ppb):
-3% (-4, -2)
Summer:
1 -h max O3 (per 100 ppb):
4% (0, 9)
-------
Table AX7-5 (cont'd). Effects of Acute O3 Exposure on Mortality
to
o
o
X
to
Reference, Study
Location and Period
Australia
Morgan etal. (1998b)
Sydney, Australia
1989-1993
Simpson etal. (1997)
Brisbane, Australia
1987-1993
Asia
Kim et al. (2004)
Seoul, Korea
1995-1999
Outcome
Measure Mean O3 Levels
All cause; 1 -h max O3 :
respiratory; 24 ppb
cardiovascular
All cause; 8-h avg O3
respiratory; (10 a.m.-6 p.m.):
cardiovascular;
all ages; age All year:
< 65 years; 18.1 ppb
age 65+ years Summer:
20.2 ppb
Winter:
16.1 ppb
All cause l-hmaxO3:
All year:
35. 16 ppb
Summer:
46.87 ppb
Winter:
2 1.26 ppb
Copollutants
Considered
PMby
nephelometer, NO2;
multipollutant
models
PM10,PMby
nephelometer, NO2,
SO2, CO
PM10, NO2, SO2,
CO; 2-pollutant
models
Lag Structure
Reported Method
0 Poisson with
GEE
0 Autoregressive
Poisson with
GEE
1 Poisson GAM
(linear model);
GLM with cubic
natural spline;
GLM with B-
mode spline
(threshold model)
Effect Estimates
l-hmaxO3(per28ppb):
All cause:
2.04% (0.37, 3.73)
Respiratory:
-0.84% (-7.16, 5.91)
Cardiovascular:
2.52% (-0.25, 5.38)
8-h avg O3 (per 10 ppb):
All cause, all ages:
All year:
2.4% (0.8, 4.0)
Summer:
3.0% (1.0, 5.0)
Winter:
1.3% (-1.4, 4.1)
1-h max O3 (per 21 .5 ppb):
All year:
Linear model:
2.6% (1.7, 3.5)
Threshold model:
3.4% (2.3, 4.4)
Summer:
Linear model:
1.9% (0.5, 3.3)
Threshold model:
3. 8% (2.0, 5.7)
-------
Table AX7-5 (cont'd). Effects of Acute O3 Exposure on Mortality
t—t
1
to
o
o
>
X
a\
OJ
O
!>
-n
H
6
0
2|
O
H
O
0
H
W
o
O
HH
H
W
Reference, Study Outcome
Location and Period Measure
Asia (cont'd)
Lee et al. (1999) All cause
Seoul and Ulsan, Korea
1991-1995
Lee and Schwartz (1999) All cause
Seoul, Korea
1991-1995
Tsai et al. (2003) All cause;
Kaohsiung, Taiwan respiratory;
1994-2000 cardiovascular;
tropical area
Yang et al. (2004) All cause;
Taipei, Taiwan respiratory;
1994-1998 cardiovascular;
subtropical area
Copollutants
Mean O3 Levels Considered
l-hmaxO3: TSP, SO2
Seoul:
32.4 ppb
Ulsan:
26.0 ppb
l-hmaxO3: TSP, SO2
Seoul:
32.4 ppb
24-h avg O3: PM10, SO2, NO2,
23. 6 ppb CO
24-h avg O3: PM10, SO2, NO2,
17.18 ppb CO
Lag Structure
Reported Method Effect Estimates
0 Poissonwith 1-h max O3 (per 50 ppb):
GEE
Seoul:
1.5% (0.5, 2.5)
Ulsan:
2.0% (-11. 1,17.0)
0 Conditional 1-h max O3 (per 50 ppb):
logistic (case-
crossover with Two controls, plus and
bidirectional minus one week:
control sampling) 1.5% (-1.2,4.2)
Four controls, plus and
minus two weeks:
2. 3% (-0.1, 4. 8)
0-2 Case-crossover 24-h avg O3 (per 19.2 ppb):
analysis
Odds ratios:
All cause:
0.994(0.995,1.035)
Respiratory:
0.996(0.848,1.169)
Cardiovascular:
1.005(0.919,1.098)
0-2 Case-crossover 24-h avg O3 (per 9.34 ppb):
analysis
Odds ratios:
All cause:
0.999(0.972-1.026)
Respiratory:
0.991(0.897-1.094)
Cardiovascular:
1.004(0.952-1.058)
-------
Table AX7-6. Effects of Chronic O3 Exposure on Respiratory Health
to
o
o
Reference, Study
Location, and Period
Mean O3 Levels
Study Description
Results and Comments
X
United States
Galizia and Kinney
(1999; expo sure data
Kinney et al., 1998)
U.S. nationwide
1995
l-hmaxO3:
10-year mean Jun-Aug:
61.2 ppb
SD15.5
Range 13-185
Goss et al. (2004)
U.S. nationwide
1999-2000
l-hmaxO3
Sl.Oppb
SD7.3
Nationwide sample of 520 young adults. Subjects
were nonsmokers, aged 17-21 years, 50% males.
Each subject provided one spirometric lung function
measurement in the spring of their 1st year at Yale
College in New Haven, CT, and completed a
questionnaire addressing residential history,
respiratory diseases, and activity patterns. Long-
term O3 exposure was treated as a high/low
dichotomous variable, with subjects assigned to the
high O3 category if they lived for 4+ years in
counties with 10-year summer mean O3 levels
greater than 80 ppb. Four lung function variables
(FVC, FEY^ FEF25.75, FEF75) were regressed on
O3 exposure, controlling for age, height, height
squared, sex, race, parental education, and maternal
smoking history. Respiratory symptom histories
(cough, phlegm, wheeze apart from colds,
and composite index for any of the three symptoms)
were logistically regressed on O3 exposure,
controlling for sex, race, parental education, and
maternal smoking.
11,484 cystic fibrosis patients over the age of
6 years. Exposure to O3, PM2 5, PM10, NO2, SO2,
and CO assessed by linking Aerometric Information
Retrieval System with patients' home zip code.
Studied exacerbation and lung function. Mortality
was also of interest, but study was underpowered to
examine this outcome. Logistic regression models
were used to analyze the exacerbations and multiple
linear regression was used to study lung function.
O3 monitoring season and regional effects also were
examined.
Significant decrements in FEV[ and FEF25.75 in relation to O3
exposure were observed for all subjects and for males alone, but
not for females alone. Similar patterns observed for FVC and
FEF75, but not with statistical significance.
% difference in lung function for high versus low O3 exposure
groups:
All subjects: -3.07% (-0.22, -5.92)
Females: -0.26% (3.79, -4.31)
Males: -4.71% (-0.66, -8.76)
FEF25.75:
All subjects: -8.11% (-2.32, -13.90)
Females: -1.96% (6. 39, -10.30)
Males: -13.02% (-4.87, -21.17)
Wheeze and respiratory symptom index were significantly elevated
for high O3 exposure group.
Odds ratios for symptoms:
Wheeze: 1.97(1.06,3.66)
Respiratory symptom index: 2.00 (1.15, 3.46)
Ozone may increase the risk for pulmonary exacerbations in cystic
fibrosis patients.
Odds ratios for two or more exacerbations (per 10 ppb increase in
1-hmax O3):
O3 only model: 1.10 (1.03, 1.17)
O3 withPM25 model: 1.08 (1.01, 1.15)
PM2 5, but not O3, was significantly associated with declines in
lung function in these patients.
-------
Table AX7-6 (cont'd). Effects of Chronic O3 Exposure on Respiratory Health
to
o
o
Reference, Study
Location, and Period
Mean O3 Levels
Study Description
Results and Comments
X
United States (cont'd)
Kinney and Lippmann
(2000)
Fort Sill, OK;
Fort Leonard Wood,
MO; Fort Dix, NJ;
Fort Benning, GA;
West Point, NY
Apr-Sep 1990
Greeretal. (1993)
California
1973-1987
l-hmaxO3:
Mean during 5-week
summer training
period:
Fort Benning, GA:
55.6 ppb
(0 hours O3 > 100 ppb)
Fort Dix, NJ:
71.3 ppb
(23 hours O3> 100
ppb)
Fort Leonard Wood,
MO:
55.4 ppb
(1 hours O3 > 100 ppb)
Fort Sill, OK:
61.7 ppb
(1 hours O3 > 100 ppb)
Annual mean O3:
Levels not reported.
72 nonsmoking students (mean age 20.25 years) at
the U.S. Military Academy at West Point, NY were
measured for lung function and respiratory
symptoms before (Apr) and after (Aug-Sep) taking
part in an intensive, largely outdoor, summer
training over five weeks (Jul 11-Aug 15) at four
U.S. military bases. Ozone levels in the Fort Dix,
NJ area were consistently higher than at the three
remaining three locations. Analysis assessed
change in lung function and respiratory symptoms
measured before and soon after the summer
training, and examined whether adverse trends
would be more pronounced in students exposed to
higher O3 levels during summer training.
3,914 nonsmoking adults aged 25+ years at
enrollment in 1977 completed questionnaires at two
time points, 1977 and 1987. To be eligible, subjects
had to have lived 10 or more years within 5 miles of
current residence. Residential histories used to
interpolate air pollution levels to zip centroids over
a 20-year period (1966-1987). New asthma cases
defined as answering yes to doctor diagnosed
asthma at 1987 followup among those answering no
at enrollment in 1977. Multiple logistic regression
used to test for associations between new-onset
asthma and long-term exposures to air pollution,
controlling for age, education, pneumonia or
bronchitis before age 16 years, and years worked
with a smoker through 1987. All models stratified
by gender.
Mean FEV! declined significantly over the two measurement
points for all subjects combined, which may reflect combined
effects of O3 with exposures to dust, vehicle exhaust, and
environmental tobacco smoke as reported by subjects from all four
locations in the post-summer questionnaire. However, a larger
mean decline was seen at the higher O3 site, Fort Dix, than at the
remaining three sites, suggesting an influence of O3 exposures.
A larger decline was observed in subjects with post-summer
measurements in the 1 st two weeks after returning from training
compared to those measured in the 3rd and 4th weeks, which is
consistent with the lung function effects being somewhat transient.
Change in lung function over the summer:
All locations: -44 mL (SE 21), p = 0.035
Fort Dix: -78 mL (SE 41), p = 0.07
Forts Sill, Leonard Wood, and Benning combined:
-31mL(SE24),p = 0.21
There were 27 incident cases of asthma among 1,305 males and
51 incident cases among 2,272 females. In logistic regression
analyses, long-term O3 exposures were associated with increased
risk of incident asthma among males but not females.
Relative risks for incident cases of asthma (per 10 ppb increase
in annual mean O3):
Males: 3.12 (1.61, 5.85)
Females: 0.94 (0.65, 1.34)
-------
Table AX7-6 (cont'd). Effects of Chronic O3 Exposure on Respiratory Health
to
o
o
Reference, Study
Location, and Period
Mean O3 Levels
Study Description
Results and Comments
X
Oi
United States (cont'd)
McDonnell et al.
(1999)
California
1973-1992
Peters etal. (1999a,b)
12 Southern
California
communities
1993-1994
8-h avg O3
(9 a.m.-5 p.m):
20-year mean:
46.5 ppb
SD15.3
l-hmaxO3:
Mean range:
1986-1990:
30.2 ppb (Santa Maria)
to 109.2 ppb
(San Dimas)
1994:
35.5 ppb (Santa Maria)
to 97.5 ppb
(Lake Gregory)
This study continued the work of Greer et al.
(1993). 3,091 nonsmoking adults completed
questionnaires at one additional time point, 1992.
Residential histories used to interpolate air pollution
levels to zip centroids over the period 1973-1992,
yielding annual mean exposure estimates for O3,
PM10, SO2, and NO2. New asthma cases defined as
answering yes to doctor diagnosed asthma at either
1987 or 1992. Multiple logistic regression used to
test for associations between new-onset asthma and
long-term exposures to air pollution, controlling for
age, education, pneumonia or bronchitis before age
16, and ever smoking. All models run separately
for males and females.
3,676 children aged 9-16 years enrolled into
the 1st cohort of the Children's Health Study in
1993. Subjects provided questionnaire data on
respiratory disease histories and symptoms.
3,293 subjects also underwent pulmonary function
testing, of which 2,781 were used in air pollution
regressions. Air pollution data for O3, PM10, PM2 5,
NO2, and inorganic acid vapors analyzed from
1986-1990 and 1994. For cross-sectional analysis
of respiratory diseases, individual pollutants were
tested for associations with ever asthma, current
asthma, bronchitis, cough, and wheeze after
controlling for covariates. For analysis of lung
function, individual pollutants and pairs of
pollutants were regressed with FVC, FEV1; FEF25.75,
and PEF, controlling for usual demographic and
anthropometric covariates.
There were 32 incident cases of asthma among 972 males and
79 incident cases among 1,786 females. In logistic regression
analyses, long-term O3 exposures were associated with increased
risk of incident asthma among males but not females. Other
pollutants were neither associated with asthma incidence nor
were confounders of the O3 association in males.
Relative risks for incident cases of asthma (per 27 ppb increase
in annual mean 8-h avg O3):
Males: 2.09 (1.03, 4.16)
Females: 0.86 (0.58, 1.26)
Acids and NO2, but not O3, were associated significantly with
prevalence of wheeze. No associations of O3 with any of the
respiratory diseases or symptoms.
Decreased lung function was associated with multiple pollutants
among females but not males. For O3 exposure in females, all four
lung function variables declined with increasing exposure.
Associations were stronger for current (1994) exposure compared
to previous (1986-1990) exposure. In males who spent more time
outdoors, FVC and FEV[ declined significantly with higher current
exposure to O3.
Change in lung function (per 40 ppb 1-h max O3 from 1986-1990):
Females:
PEF: -187.2mL/s(SE50.1),p<0.005
FEF25.75: -102.2 mL/s (SE 28.8), p<0.0\
Males:
PEF: 31.1 mL/s (SE 48.8), p> 0.05
FEF25.75: 11.7 mL/s (SE 26.7), p> 0.05
-------
Table AX7-6 (cont'd). Effects of Chronic O3 Exposure on Respiratory Health
to
o
o
Reference, Study
Location, and Period
Mean O3 Levels
Study Description
Results and Comments
X
United States (cont'd)
Gauderman et al.
(2000, 2004a,b)
12 Southern
California
communities
1993-2001
Gauderman et al.
(2002)
12 Southern
California
communities
1996-1999
McConnell et al.
(1999)
12 Southern
California
communities
1993
8-h avg O3
(10 a.m.-6 p.m.):
Mean range:
Approximately 28 ppb
(Long Beach) to 65 ppb
(Lake Arrowhead)
8-h avg O3
(10 a.m.-6 p.m.):
Mean range:
Approximately 27 ppb
(Long Beach) to 67 ppb
(Lake Gregory)
l-hmaxO3:
Estimated annual daily
mean:
65.5 ppb
Range 35.5-97.5
Analysis of longitudinal lung function change
in relation to long-term air pollution levels in
the Children's Health Study. Children from 4th
(n = 1,498), 7th (n = 802), and 10th (n = 735) grade
enrolled in 1993. Children enrolled in 7th and
10th grade were followed until 1997; 4th graders
were followed until 2001. Baseline questionnaires
completed and annual pulmonary function tests
(FVC, FEVj, FEF25.75, FEF75) performed. Air
pollution monitoring stations established in the
12 study communities beginning in 1994 to measure
O3, NO2, PM10, PM2 5, and inorganic acid. Analysis
using adjusted linear regression models.
Second cohort of the Children's Health Study.
2,081 4th graders (mean age 9.9 years) enrolled in
1996. Baseline questionnaires were completed and
annual pulmonary function tests (FVC, FEV1; FEF25.
75, FEF25.75/FVC, PEF) were performed. 1,672
children had at least two pulmonary function test
data. Air pollutants examined include O3, NO2,
PM10, PM25, inorganic acid, elemental carbon, and
organic carbon. Adjusted linear regression model
was used.
First cohort of the Children's Health Study.
Association of O3 with prevalence of chronic lower
respiratory tract symptoms among children with a
history of asthma was examined in a cross-sectional
study in 12 communities. Questionnaires were
completed by parents of 3,676 4th, 7th, and 10th
graders, of which 493 had asthma. Exposure data
(O3 NO2, PM10, PM2 5, and inorganic acid vapors)
collected in 1994 used to estimate exposure.
Analysis using logistic regression method.
In the 7th and 10th grade cohorts, difference in lung function
growth from the least to the most polluted community was not
associated with any of the air pollutants, including O3. Among the
4th graders, decreased lung growth was associated with exposures
to PM and NO2, but not with O3.
In this cohort, a significant association between O3 and PEF and
FVC was noted in children spending more time outdoors.
% difference in lung function growth from least to most polluted
community (per 36.6 ppb increase in annual mean 8-h avg O3):
PEF:
All children: -1.21% (-2.06, -0.36)
Children more outdoors: -1.62% (-2.93, -0.29)
Children less outdoors: -0.87% (-2.09, 0.37)
Children with asthma were much more likely to have bronchitis or
related symptoms than children without such history. Among the
asthmatic children, significant relationship were observed between
phlegm and all pollutants studied, with the exception of O3.
-------
Table AX7-6 (cont'd). Effects of Chronic O3 Exposure on Respiratory Health
to
o
o
Reference, Study
Location, and Period
Mean O3 Levels
Study Description
Results and Comments
X
OO
United States (cont'd)
McConnell et al.
(2002)
12 Southern
California
communities
1993-1998
Rite et al. (2000)
Southern California
1989-1993
l-hmaxO3:
Four-year mean
(1994-1997):
Low pollution
communities (n = 6):
50.1 ppb
Range 37.7-67.9
High pollution
communities (n = 6):
75.4 ppb
Range 69.3-87.2
8-h avg O3
(9 a.m.-5 p.m.):
Six weeks before birth:
36.9 ppb
SD 19.4
Range 3.3-117 ppb
3,535 children (age 9-16 years) without a history of
asthma recruited in 1993 and 1996, and followed
with annual surveys through 1998 to determine
incidence of new onset asthma. Participation
in sports assessed at baseline. Copollutants included
PM10, PM2 5, NO2, and inorganic acid vapors.
Asthma incidence was examined as a function of
number of sports played in high and low pollution
communities, controlling for age, sex, and ethnic
origin.
Data on 97,158 singleton births over period
1989-1993 linked to air monitoring data during
different periods of pregnancy to determine
associations between pollution exposures and
preterm birth. Besides O3, pollutants of interest
included PM10, NO2, and CO. Multiple regression
analysis used, controlling for maternal age, race,
education, parity, and other factors.
Asthma incidence was not higher in the high pollution
communities as compared with the low pollution communities,
regardless of the pollutant used to define high/low. In fact,
the high O3 communities had generally lower asthma incidence.
However, in high O3 communities, there was an increased risk of
asthma in children playing three or more sports compared to those
playing no sports; no such increase was observed in the low O3
communities. No other pollutant showed this association. These
results suggest that high levels of physical activity is associated
with risk of new asthma development for children living
in communities with high O3 levels.
Relative risk of developing asthma in children playing three or
more sports compared to those playing no sports:
Low pollution communities: 0.8 (0.4, 1.6)
High pollution communities: 3.3 (1.9, 5.9)
Both PM10 and CO during early or late pregnancy were associated
with increased risk for preterm birth. No associations observed
withO3.
-------
Table AX7-6 (cont'd). Effects of Chronic O3 Exposure on Respiratory Health
to
o
o
Reference, Study
Location, and Period
Mean O3 Levels
Study Description
Results and Comments
X
vo
United States (cont'd)
Kilnzlietal. (1997);
Tageretal. (1998)
Los Angeles and San
Francisco, CA;
Berkeley, CA
1995
Sherwin et al. (2000)
Los Angeles, CA and
Miami, FL
1995-1997
8-h avg O3
(10 a.m.-6 p.m.):
Range of lifetime
mean:
Los Angeles:
25-74 ppb
San Francisco:
16-33 ppb
Levels not reported.
In a pilot study, 130 freshman students (age 17-21
years) at the University of California at Berkeley
measured for lung function and histories of
residential locations and indoor/outdoor activity
patterns and levels. By design, students had
previously resided in one of two metropolitan areas
that differed greatly in O3 concentrations,
San Francisco or Los Angeles. A key goal was to
test whether measures of small airways function
(e.g., nitrogen washout, FEF25.75, FEF75) were
sensitive measures of long-term O3 impacts.
Lifetime exposures to O3, PM10 and NO2 assigned
by interpolation to sequence of residence locations
from available monitoring stations. Multiple
exposure measures were derived with varying
degrees of incorporation of time-activity
information, going from ecological concentration
to individual time-activity weighted exposure.
Performed linear regression analysis of lung
function on O3 exposures, controlling for height,
ethnicity, gender, and region.
Lungs obtained from autopsies of young residents
(age 11-30 years) of Miami (n = 20) and Los
Angeles (n = 18) who died suddenly from homicide,
vehicular accident, or other violence.
Semiquantitative measurements of centriacinar
region alterations were compared between the
two cities.
Decreased FEF25.75 and FEF75 were associated with long-term O3
exposures. Results were similar whether O3 exposure was purely
ecologic or incorporated time-activity information. FVC, FEV^
and nitrogen washout were generally not associated with O3 levels.
No evidence for PM10 or NO2 main effects or confounding of O3.
Similar patterns results using O3 hours > 60 ppb as exposure metric
instead of daily 8-h avg O3 (10 a.m.-6 p.m.). Surprisingly, region
of residence was a major negative confounder as lung function was
lower on average among students from the low O3 city, San
Francisco, than among those who had lived in Los Angeles.
Ozone exposures were significant predictors only after controlling
the regional effect.
Change in lung function (per 20 ppb increase in lifetime mean 8-h
avg O3):
FEF25.75: -420 mL/s (-886, 46); 7.2% of population mean
FEF75: -334 mL/s (-657, -11); 14% decline of population mean
A greater extent (p < 0.02) and severity (p < 0.02) of centriacinar
region alterations were observed in lungs of the Los Angeles
residents than the Miami residents. These differences could not be
attributed to smoking history. The higher O3 levels in Los Angeles
might be responsible for the greater centriacinar region alterations,
however correlations could not be performed due to the relatively
small number of cases available.
-------
Table AX7-6 (cont'd). Effects of Chronic O3 Exposure on Respiratory Health
to
o
o
Reference, Study
Location, and Period
Mean O3 Levels
Study Description
Results and Comments
X
United States (cont'd)
Gongetal. (1998b)
Glendora, CA
1977-1987
Chen et al. (2002)
Washoe County, NV
1991-1999
Kinneyetal. (1996)
New York City
1992-1993
l-hmaxO3:
Annual means range
(1983-1987):
109 ppb to 134 ppb
8-hmaxO3:
27.23 ppb
SD 10.62
Range 2.76-62.44
l-hmaxO3:
Summer (Jul-Sep
1992):
58 ppb
Maximum 100
Winter (Jan-Mar 1992):
32 ppb
Maximum 64
Summer (Jul-Sep
1993):
69 ppb
Maximum 142
164 adults (mean age 45 years; 34% males) from a
high O3 community underwent lung function testing
in 1986-1987 (T3). Subjects were recruited from a
cohort of 208 nonsmoking adults who had been
tested on two previous occasions: 1977-1978 (Tl)
and 1982-1983 (T2). Analyzed changes in lung
function at three time points. Subjects were also
asked to undergo controlled exposures to 0.40 ppm
O3 over 2 hours with intermittent exercise.
45 subjects agreed to participate. Investigators
hypothesized that acutely responsive subjects would
show more rapid declines in function over the study
period.
Birth weight for 36,305 single births analyzed
in relation to mean PM10, O3, and CO levels in
trimesters 1, 2, and 3.
19 healthy adult joggers (age 23-38 years; 18 males)
from the Governors Island U.S. Coast Guard facility
in New York harbor underwent a series of two
bronchoalveolar lavages, first in the summer of
1992 and then again in the winter of 1992. Because
the summer of 1992 had lower than average O3
levels, six subjects underwent a third
bronchoalveolar lavage in the summer of 1993.
Study tested whether inflammatory markers in
bronchoalveolar lavage fluid were elevated during
the summer O3 season among adults who regularly
exercised outdoors. Outcomes included cell
differentials, release of interleukin-8 (IL-8) and
tumor necrosis factor-alpha (TNF-a) in
bronchoalveolar lavage cells supernatants, release of
reactive oxygen species by macrophages, and
concentrations of protein, lactate dehydrogenase,
IL-8, fibronectin, al-antitrypsin (al-AT),
complement fragments (C3a), and prostaglandin E2
(PGE2) in bronchoalveolar lavage fluids.
Mean FVC and FEV! showed nonsignificant increase from T2 to
T3, whereas an earlier analysis of the Tl to T2 change had found a
significant decline in function (Betels et al., 1987). There was
evidence for 'regression to the mean,' in that subject with larger
declines in function from Tl to T2 tended to have larger increases
in function from T2 to T3. A consistent decline in FEVj/FVC ratio
was observed at all three time points (p < 0.0001 by ANOVA).
Acute changes in lung function, determined using controlled O3
exposures, were not associated with chronic lung function changes.
PM10 was the only air pollutant associated with decreased birth
weights. Ozone levels quite low throughout study.
There was no evidence of acute inflammation in the summer of
1992 compared to the winter; i.e., neutrophil differentials, IL-8 and
TNF-a showed no significant differences. However, a measure of
cell damage, lactate dehydrogenase, was elevated in the summer,
suggesting possible O3-mediated damage to the lung epithelium
with repeated exposures to O3 while exercising. O3 levels during
the summer of 1992 were atypically low for New York City.
Among six subjects who agreed to undergo a third bronchoalveolar
lavage test in the summer of 1993, lactate dehydrogenase was
again elevated compared to winter. In addition, IL-8 was elevated
in the summer of 1993, suggesting acute inflammation.
-------
Table AX7-6 (cont'd). Effects of Chronic O3 Exposure on Respiratory Health
to
o
o
Reference, Study
Location, and Period
Study Description
Results and Comments
X
Europe
Charpinetal. (1999)
Seven towns in
SE France
Jan-Feb 1993
Ramadour et al.
(2000)
Seven towns in
SE France
Jan-Feb 1993
8-h max O3: 2,073 children (age 10-11 years) from 7 towns
Range of means: tested for atopy based on skin prick testing
30.2-52.1 ug/m3 (house dust mite, cat dander, grass pollen,
cypress pollen, and Alternaria). Towns represented
24-h avg O3: a range of ambient O3 and other pollutant (NO2 and
Range of means: SO2) levels. Tested hypothesis that atopy is greater
20.1-42.1 ug/m3 in towns with higher photochemical pollution levels.
To be eligible, subjects must have resided in current
town for at least 3 years. Authors stated that Jan to
Feb pollution levels correlated with levels observed
throughout the year, though no data was given to
support this.
8-h max O3: 2,445 children (age 13-14 years) who had lived at
Range of means: their current residence for at least three years were
30.2-52.1 ug/m3 recruited from schools in seven towns in SE France.
This region has highest O3 levels in France.
Subjects completed ISAAC survey of asthma and
respiratory symptoms. In addition to O3 also
collected data on SO2 and NO2. Analyzed
relationships between asthma and other respiratory
conditions with mean air pollution levels across the
seven towns using logistic regression, controlling
for family history of asthma, personal history of
early-life respiratory diseases, and SES. Also did
simple univariate linear regressions.
In this cross-sectional analysis, no differences in atopy levels were
seen across the seven towns. Authors concluded that long-term
exposures to oxidant pollution do not favor increased allergy to
common allergens. The very low winter O3 levels monitored and
lack of long-term exposure data make it impossible to reach this
conclusion in a definitive manner.
In logistic regressions, no significant associations seen between O3
and 12-month history of wheezing, history of asthma attack,
exercise induced asthma and/or dry cough in last 12 months.
In simple bivariate scatterplots of respiratory outcomes versus
mean O3 levels in the seven towns, there appeared to be strong
positive relationships (r = 0.71 for wheezing in last 12 months
and r = 0.96 for asthma attacks). No data on slope estimates given.
Concerns about potential confounding across towns
limits the interpretation of this study.
-------
Table AX7-6 (cont'd). Effects of Chronic O3 Exposure on Respiratory Health
to
o
o
Reference, Study
Location, and Period
Study Description
Results and Comments
X
to
Europe (cont'd)
Ihorst et al. (2004)
Nine communities in
Lower Austria
Apr 1994-Oct 1997
Six communities in
Germany
Febl996-0ctl999
Kopp et al. (2000)
Ten communities in
Austria and SW
Germany
Mar 1994-Nov 1995
!/2-h avg O3:
Quartile ranges:
Summer:
1st quartile: 22-30 ppb
2nd quartile: 30-38 ppb
3rd quartile: 38-46 ppb
4th quartile: 46-54 ppb
Winter:
1st quartile: 4-12 ppb
2nd quartile: 12-20 ppb
3rd quartile: 20-28 ppb
4th quartile: 28-36 ppb
!/2-h avg O3:
Stratified by low,
medium, high
exposure:
Low: 24-33 ppb
Medium: 35-38 ppb
High: 44-52 ppb
2,153 children (median age 7.6 years) were studied
for the effects of semi-annual and 3'/2-year mean O3
concentrations on FVC and FEVj. As a measure of
lung growth, the difference between two
consecutive values for each child was divided by the
number of days between tests. The effect of O3
exposure on lung growth was analyzed by linear
regression models, after adjusting for sex, age,
height at start of the time period, and passive
smoking exposure.
797 children with a mean age of 8.2 years. Four
pulmonary function tests (FVC, FEV[) performed
on each child over two summers. Examined
association between average daily lung function
growth and exposure to O3, PM10, NO2, and SO2.
Analysis using linear regression models.
Higher semi-annual mean O3 levels were associated with
diminished lung function growth during the summer, but
increased lung function growth in the winter.
Change in lung function (4th quartile compared to 1 st quartile
semi-annual O3 mean):
Summer:
FEVj (mL/100 days): -18.5 (-27.1,-9.8)
FVC (mL/100 days): -19.2 (-27.8, -10.6)
Winter:
FEV; (mL/100 days): 10.9 (2.1, 19.7)
FVC (mL/100 days): 16.4 (8.3, 24.6)
No associations between longer term O3 exposure (mean summer
O3 over a 3!/2-year period) and lung function growth was found.
Lower FVC and FEV[ increases were observed in children exposed
to high ambient O3 levels compared to those exposed to lower O3
levels during the summer. During the winter, children in higher O3
areas showed a slightly greater increase in FVC and FEV; than
those in the low O3 areas, which might reflect that children catch
up in lung function deficits during the winter season.
Change in lung function for high versus low O3 exposure groups
(per ppb O3):
FEVI:
Summer of 1994: -0.303 mL/day, p = 0.007
Winter of 1994/1995: 0.158 mL/day, p = 0.006
Summer of 1995: -0.322 mL/day, p = 0.001
-------
Table AX7-6 (cont'd). Effects of Chronic O3 Exposure on Respiratory Health
to
o
o
Reference, Study
Location, and Period
Study Description
Results and Comments
X
Europe (cont'd)
Frischeretal. (1999)
Nine communities in
Austria
1994-1996
Frischeretal. (2001)
Nine communities in
Austria
Sep-Octl997
24-h avg O3:
Summer:
34.8ppb
SD8.7
Winter:
23.1ppb
SD7.7
!/2-h avg O3:
30-day mean:
31.57ppb
IQR 20.61
Communities from two counties chosen to represent
a broad range of O3 concentrations; a two-fold range
in mean levels was observed. 1,150 children (mean
age 7.8 years; 52% males) from grades 1 and 2
performed spirometry in spring and fall over three
years (total of six measurements per child) to
determine if seasonal exposure to O3 would be
associated with diminished lung function growth,
especially over the summer seasons. Ozone levels
were low during lung function testing periods.
Participation rates were high. At baseline,
respiratory histories were collected and subjects
were tested for allergy by skin prick. Examined
association between O3 levels and change in lung
function (FVC, FEVj, and MEF50 [maximal
expiratory flow at 50% of vital capacity]) over each
season, controlling for baseline function, atopy,
gender, site, environmental tobacco
smoke exposure, season, and change in height.
Other pollutants studied included PM10, SO2, and
NO,.
A cross-sectional study of 877 school children
(mean age 11.2 years). Analyzed for urinary
eosinophil protein as a marker of eosinophil
activation determined from a single spot urine
sample using linear regression models.
Seasonal mean O3 exposures were associated with reductions in
growth in all three lung function measures. Inconsistent results
seen for other pollutants. Summer season lung function growth
decrements per unit O3 were larger when data restricted to children
who spent whole summer in their community. No evidence for
nonlinear O3 effect. No confounding of O3 effect by temperature,
ETS, or acute respiratory illnesses.
Change in lung function (per ppb O3):
FEV; (mL/day):
All subjects:
Summer: -0.029 (SE 0.005), p< 0.001
Winter: -0.024 (SE 0.006), p< 0.001
Restricted to subjects who stayed in community:
Summer: -0.034 (SE 0.009), p < 0.001
FVC (mL/day):
All subjects:
Summer: -0.018 (SE 0.005), p< 0.001
Winter: -0.010 (SE 0.006), p = 0.08
Restricted to subjects who stayed in community:
Summer: -0.033 (SE 0.007), p < 0.001
Log-transformed urinary eosinophil protein-X concentrations were
found to be significantly associated with O3 levels, after adjusting
for gender, site, and atopy.
Change in log urinary eosinophil protein-X (per ppb O3):
0.007 ug/mmol creatinine (SE 0.02), p < 0.001
-------
Table AX7-6 (cont'd). Effects of Chronic O3 Exposure on Respiratory Health
to
o
o
Reference, Study
Location, and Period
Study Description
Results and Comments
X
Europe (cont'd)
Horak et al. (2002a,b)
Eight communities in
Austria
1994-1997
Palli et al. (2004)
Florence, Italy
1993-1998
Seasonal mean O3:
Summer:
31.8ppb
Range 18.7-49.3
Winter:
19.8ppb
Range 12.7-35.9
24-h avg O3:
Range of monthly
means from 1993-1998:
Approximately
25-125 ppb
This study continued the work of Frischer et al.,
1999 by including one additional year of data, 1997.
The major hypothesis considered PM10. For this
study, 80.6% of the 975 children (mean age 8.11
years) performed all six lung function tests. A total
of 860 children were included in the GEE analysis.
Multipollutant analysis for PM10, SO2, and NO2.
320 residents (age 35-64 years) in the metropolitan
area of Florence enrolled in a study investigating the
correlation between levels of DNA bulky adducts
and cumulative O3 exposure. One blood sample was
collected for each subject. Various time windows
of exposure were examined, ranging from 0-15 days
to 0-90 days prior to the blood draw. Simple
Spearman correlations between DNA adduct levels
and different O3 exposure time windows were
calculated after stratifying by smoking history,
area of residence, and population type (random
sample or exposed workers).
Seasonal mean O3 showed a negative effect on lung function
growth, confirming the previous shorter study. Ozone effects were
robust to inclusion of PM10 into the model. However, for FEV; in
winter, the O3 effect slightly diminished after including PM10.
Taking into account only children who stayed at home the whole
summer period did not affect the results.
Change in lung function (per ppb O3):
FEV; (mL/day):
O3 only models:
Summer: -0.021, p<0.00\
Winter: -0.020, p<0.00\
O3 with PM10 models:
Summer: -0.020,p<0.00\
Winter: -0.012, p = 0.04
Consistent relationships between O3 exposure and DNA adduct
levels were observed only among never smokers. Correlations
were highest among never smokers who resided in the urban area
and were not occupationally exposed to vehicle traffic pollution.
Associations were significant up to a time window of 0-60 days
prior to the blood draw in the subgroup of never smokers, with
strongest relationships observed between 30-45 days prior.
-------
Table AX7-6 (cont'd). Effects of Chronic O3 Exposure on Respiratory Health
to
o
o
Reference, Study
Location, and Period
Study Description
Results and Comments
X
Latin America
Calderon-Garciduenas
etal. (1995)
SW Mexico City
Nov 1993
Manzanillo, Mexico
Jan 1994
Calderon-Garciduenas
etal. (1997)
SW Mexico City
Sep-Nov 1995
Manzanillo, Mexico
Jan 1995
SW Mexico City
(urban):
l-havgO3>120ppb:
4.4 hours/day
Maximum 307 ppb
Manzanillo, Pacific
port (control):
No detectable air
pollutants.
SW Mexico City
(urban):
l-havgO3> 120 ppb:
82 hours/month
Maximum 286 ppb
Manzanillo, Pacific
port (control):
No detectable air
pollutants.
Nasal lavage samples collected from 38 urban
(mean age 12.2 years) and 28 control (mean age
11.7 years) children. Samples analyzed for
polymorphonuclear leukocyte counts, expression
of human complement receptor type 3 (GDI Ib) on
nasal polymorphonuclear leukocytes, and nasal
cytologies.
129 urban and 19 control children aged 6-12 years
old with no history of smoking or environmental
tobacco smoke exposure and no current medication
use for atopy or asthma. Three nasal biopsies
obtained at 4-week intervals and analyzed for DNA
damage based on the presence of DNA fragments.
Nasal cytologies revealed that children from Mexico City had
abnormal nasal mucosae, including mucosal atrophy, marked
decreases in the numbers of ciliated-type cells and goblet cells,
and squamous metaplasia.
Exposed children had significantly higher nasal
polymorphonuclear leukocyte counts (p < 0.001) and nasal GDI Ib
expression (p < 0.001) compared to controls. However, the
inflammatory response did not seem to correlate with the previous
day's O3 exposure in a dose-dependent manner, suggesting that
there might be a competing inflammatory mechanism at the
brochioalveolar level. Overall, these results suggest that ambient
O3 produces an inflammatory response in chronically exposed
children.
Urban children had significantly more DNA fragments than did
control children (p < 0.0001). Percentage of damaged cells was
82.2% (SE 6.4) in urban children and 17.0% (SE 6.1) in control
children. Among urban children, more upper respiratory
symptoms and DNA damage was seen with increasing age.
Older children spent more time outdoors and engaged in physical
activities (p<0.001).
Urban children were exposed to a complex pollution mix, making
it difficult to attribute effects to O3 specifically. However, authors
noted that O3 was the pollutant with most exceedences of air
quality standard.
-------
Table AX7-6 (cont'd). Effects of Chronic O3 Exposure on Respiratory Health
to
o
o
Reference, Study
Location, and Period
Study Description
Results and Comments
X
Latin America (cont'd)
Calderon-Garciduenas SW Mexico City
etal. (1999)
SW Mexico City
May-June 1996
Manzanillo, Mexico
May 1996
Fortoul et al. (2003)
Mexico City
May 1997
(urban):
l-havgO3>80ppb:
May:
161 hours/month
Maximum 232 ppb
June:
98 hours/month
Maximum 261 ppb
Manzanillo, Pacific
port (control):
Mean< 10 ppb
9-h avg O3
(9 a.m.-6 p.m.):
South:
121 ppb
North:
89 ppb
86 urban and 12 control children aged 6-13 years
old with no history of smoking or environmental
tobacco smoke exposure and no use of medication
for atopy or asthma. Urban children stratified into
five groups by school grade level (1st through 5th).
Nasal epithelial biopsies obtained from inferior
nasal turbinates, and analyzed for single strand
DNA breaks and for 8-OHdG (8-hydroxy-2'-
deoxyguanosine), a mutagenic lesion produced by
G^T mutations. These outcomes relate to possible
carcinogenic effects of air pollution exposures.
Multiple air pollutants monitored in SW Mexico
City within 3 miles of urban subject residences.
Estimated DNA strand breaks on nasal epithelial
cells and leucocytes sampled from asthmatic
(n = 15) and nonasthmatic (n = 224) medical
students aged 18-28 years using a single-cell
gel electrophoresis assay.
No respiratory symptoms reported by control children; urban
children reported multiple nasal and lung symptoms, including
cough and chest discomfort among 46% of urban children, with
higher rates for 5th versus 1 st graders. 8-OHdG was
approximately 3-fold higher in biopsies from urban children
(p < 0.05), however, no differences by school grade. Single strand
DNA breaks were more common in urban versus control children,
with an age-dependent increase in the urban children (p < 0.05).
These results suggest that DNA damage is present in the nasal
epithelial cells of children living in highly polluted SW Mexico
City and may reflect enhanced risk of cancer later in life.
Though O3 represents an important component of the pollution
mix, it is not possible to attribute effects solely to O3.
Greater genotoxic damage in asthmatics' nasal epithelial cells
(p < 0.05) may reflect their higher vulnerability for DNA damage,
or a decreased ability to repair it, compared with nonasthmatic
subjects.
Gouveia et al. (2004)
Sao Paulo, Brazil
1997
l-hmaxO3:
63.0 ppb
SD33.5
Birth weight for 179,460 single births analyzed
in relation to PM10, SO2, CO, NO2, and O3 levels in
trimester 1, 2, and 3. GAM and logistic regression
models used for analysis.
Exposures to PM10 and CO during 1 st trimester were found to
have significant negative associations with birth weight.
No associations observed for the other air pollutants, including O3.
-------
Table AX7-6 (cont'd). Effects of Chronic O3 Exposure on Respiratory Health
to
o
o
Reference, Study
Location, and Period Mean O3 Levels
Study Description
Results and Comments
X
Asia
Kuo et al. (2002)
Central Taiwan
1996
l-hmaxO3:
Annual mean range
across 7 of 8 schools:
18.6-27.3 ppb
Respiratory questionaire administered to
12,926 children aged 13-16 years at eight junior
high schools in central Taiwan, to determine asthma
prevalence. The association between asthma
prevalence and air pollution exposure analyzed by
simple Pearson correlations of prevalence with
annual mean air pollution levels (O3, SO2, PM10,
and NO2), and by multiple logistic regression.
The 775 asthmatics who were identified then
provided follow-up data on symptoms and
exacerbations over a one-year period. Simple
Pearson correlations were computed between
monthly hospital admissions and air pollution
levels, not controlling for covariates such as season
or weather.
Asthma prevalence ranged from 5.5% to 14.5% across the 8
schools. Based on simple Pearson's correlations, mean O3
(r = 0.51) and NO2 (r = 0.63) levels were correlated with variations
in asthma prevalence. However, only NO2 remained significant
in multiple logistic regression analyses after adjusting for various
potential confounding factors.
Longitudinal hospital admissions results are inconclusive due
to analytical limitations. Monthly correlations of hospital
admissions for asthmatics yielded variable results, all of which
would be confounded by temporal factors.
-------
Table AX7-7. Effects of Chronic O3 Exposure on Mortality
to
o
o
Reference, Location,
Study Period
Mean O3 Levels
Study Description
Results and Comments
United States
Pope et al. (2002)
U.S. nationwide
1982-1998
l-hmaxO3:
59.7 ppb
SD 12.8
24-h avg O3:
45.5 ppb
SD7.3
Approximately 500,000 members of American Cancer
Society cohort enrolled in 1982 and followed through
1998 for all cause, cardiopulmonary, lung cancer, and
all other cause mortality. Age at enrollment was
30+ years. Air pollution concentrations in urban area of
residence at time of enrollment assessed from 1982
through 1998. Other pollutants considered include TSP,
PM15 , PM10, PM2
, PM15.2 5, S042-, S02, N02, and CO.
No significant effect of O3 on mortality risk, though the
association of Jul-Sep O3 concentrations with all cause
and cardiopulmonary mortality were positive and nearly
significant.
Residential location was known only at enrollment to
study in 1982. Thus, exposure misclassification is
likely to be high.
0.05
Analyses were robust to the deletion of diastolic blood
pressure in the models, indicating that the association
between mortality and O3 was not mediated through
blood pressure.
-------
Table AX7-7 (cont'd). Effects of Chronic O3 Exposure on Mortality
to
o
o
Reference, Location,
Study Period
Mean O3 Levels
Study Description
Results and Comments
lOOppb:
330 h/year
SD295
IQR 551
Prospective cohort study of 6,338 nonsmoking non-
Hispanic white adult members of the Adventist Health
Study followed for all cause, cardiopulmonary,
nonmalignant respiratory, and lung cancer mortality.
Participants were aged 27-95 years at enrollment in
1977. 1,628 (989 females, 639 males) mortality events
followed through 1992. All results were stratified by
gender. Used Cox proportional hazards analysis,
adjusting for age at enrollment, past smoking,
environmental tobacco smoke exposure, alcohol use,
education, occupation, and body mass index. Analyzed
mortality from all natural causes, cardiopulmonary,
nonmalignant respiratory, and lung cancer. Ozone
results were presented for both metrics.
Of 16 regressions involving O3 exposures (two genders;
four mortality causes; two O3 metrics), 11 were positive
and one was statistically significant, for lung cancer in
males for O3 h/year > 100 ppb.
Relative risks for lung cancer mortality in males:
24-h avg O3 (per 12.0 ppb):
2.10(0.99,4.44)
O3 h/year > 100 ppb (per 551 hours/year):
4.19(1.81,9.69)
Inconsistency across outcomes and genders raises
possibility of spurious finding. The lack of
cardiopulmonary effects raises plausibility concerns.
Beesonetal. (1998)
Three California air
basins: San Francisco,
South Coast (Los Angeles
and eastward), San Diego
1977-1992
Annual mean O3
26.2 ppb
SD7.7
O3h/year> 100 ppb:
333 h/year
SD297
6,338 nonsmoking non-Hispanic white adult members
of the Adventist Health Study aged 27-95 years at time
of enrollment. 36 (20 females, 16 males) histologically
confirmed lung cancers were diagnosed through 1992.
Extensive exposure assessment, with assignment of
individual long-term exposures to O3, PM10, SO42", and
SO2, was a unique strength of this study. All results
were stratified by gender. Used Cox proportional
hazards analysis, adjusting for age at enrollment, past
smoking, education, and alcohol use.
Males, but not females, showed moderate association
for O3 and incident lung cancer risk.
Relative risks for lung cancer incident in males:
O3 h/year > 100 ppb (per 556 hours/year):
All males: 3.56(1.35,9.42)
Never smokers: 4.48 (1.25, 16.04)
Past smokers: 2.15 (0.42, 10.89)
-------
i 8. INTEGRATIVE SYNTHESIS
2
3
4 8.1 INTRODUCTION
5 This integrative synthesis (Chapter 8) aims to provide a coherent framework for the
6 assessment of health risks associated with human exposures to ambient ozone (O3) in the United
7 States. The main goal of this chapter is to integrate newly available scientific information with
8 that discussed in the 1996 O3 AQCD, to address issues central to the EPA's assessment of
9 scientific information needed to support the current review of the primary O3 NAAQS. Other
10 scientific information concerning ambient O3 welfare effects (i.e., effects on vegetation
11 /ecosystems, surface-level solar UV flux/climate changes, and man-made materials) and
12 pertinent to review of secondary O3 standards is assessed in ensuing Chapters 9, 10, and 11. The
13 integrated assessment of scientific findings provided here and elsewhere in this document will be
14 used and their policy implications considered in an Ozone Staff Paper to be prepared by EPA's
15 Office of Air Quality Planning and Standards (OAQPS). The scientific and technical
16 assessments provided in that Staff Paper will "bridge the gap" between scientific assessments in
17 this criteria document and judgments required of the EPA administrator in evaluating whether to
18 retain or, possibly, to revise the current O3 NAAQS.
19 Ozone found in the earth's atmosphere generally originates from photochemical reactions
20 that are predominantly catalyzed by the interaction of sunlight with other pollutants, especially
21 nitrogen oxides (NOX) and hydrocarbons such as volatile organic compounds (VOCs). Other
22 photochemical oxidants, such as peroxyacetyl nitrate (PAN) and hydrogen peroxide (H2O2), are
23 also generated along with O3 by such atmospheric interactions. In addition to the tropospheric
24 O3 generated by these interactions, some O3 is found near the earth's surface as the result of its
25 downward transport from the stratosphere, even in the absence of photochemical reactions in the
26 troposphere. However, in contrast to stratospheric O3, which plays an important role in
27 maintaining the habitability of the planet by shielding the surface from harmful solar ultraviolet
28 (UV) radiation, tropospheric O3 at the surface can exert adverse effects on humans, animals, and
29 vegetation. This criteria document is mainly focused on assessment of health and welfare effects
30 resulting from exposures to surface level concentrations of tropospheric O3, with only relatively
31 limited attention begin accorded to other photochemical oxidants such as PAN or H2O2.
January 2005 8-1 DRAFT-DO NOT QUOTE OR CITE
-------
1 8.1.1 Chapter Organization
2 This integrative synthesis chapter is divided into several major sections. This first section
3 (Introduction) not only aims to orient the reader to the organization and content of the chapter,
4 but also provides background information on the current O3 NAAQS and important types of
5 human responses to O3 exposure that were considered as key bases for the 1997 EPA revision of
6 the O3 NAAQS. The next section (Section 8.2) focuses on air quality trends and current ambient
7 O3 concentrations to provide context for ensuing discussions of ambient O3 exposures and its
8 effects on human health and welfare.
9 The subsequent sections (8.3, 8.4, and 8.5) then build upon the integrative synthesis
10 presented in Chapter 9 of the 1996 O3 AQCD (U.S. Environmental Protection Agency, 1996) to
11 integrate newly available key scientific information assessed in Chapters 4 through 7 of this
12 document. This includes integration of information on dosimetry, as well as lexicological,
13 human clinical, and epidemiological studies.
14 These sections collectively address the following key issues: (1) ambient exposures,
15 personal exposures, and dosimetric considerations; (2) experimental studies on toxicological
16 responses to acute O3 exposures in humans (clinical studies) and both acute and chronic effects
17 in animals; (3) assessment of epidemiological evidence for associations between O3 exposure in
18 human populations and health effects and the robustness of these associations; (4) integration of
19 the experimental data with epidemiological assessments; (5) biological mechanisms and other
20 evidence useful in judging the plausibility of adverse health effects being associated with human
21 exposures to ambient O3 levels encountered in the United States; and (6) identification of
22 susceptible and vulnerable populations potentially at increased risk for O3-related health effects
23 and potential public health impacts of human exposure to ambient O3 in the United States.
24 The present chapter mainly focuses on discussion of new scientific information that has
25 become available since the 1996 O3 criteria review that supported EPA's revision of the O3
26 NAAQS in 1987. However, it also highlights important data gaps and uncertainties that still
27 exist with regard to various key issues and notes important research needs in a number of key
28 areas. Detailed evaluation of such research needs is beyond the scope of this document, but will
29 be undertaken as part of later EPA efforts focused on identification of O3 research needs and
30 development of research planning documents.
31
January 2005 8-2 DRAFT-DO NOT QUOTE OR CITE
-------
1 8.1.2 Current Standards
2 The NAAQS for ambient O3 were revised in 1997 by adding an 8-h standard (Table 8-1) in
3 addition to the 1979 1-h standard, which is met if the fourth highest daily maximum 1-h O3 over
4 a 3-year period is < 0.12 ppm. The 8-h standard is met when the 3-year average of the annual
5 fourth highest daily maximum 8-h average concentration is < 0.08 ppm. The 1997 standards
6 were based on various scientific supportive data from human exposure and epidemiological
7 studies as assessed in the 1996 O3 AQCD. The gradations of individual responses observed with
8 short-term exposure to O3 in healthy persons (Table 8-2) and in persons with impaired
9 respiratory systems (Table 8-3) are representative of the critical information used in these
10 evaluations, as summarized in Tables 9-1 and 9-2 of the 1996 O3 AQCD and reproduced herein
11 Tables 8-2 and 8-3, respectively. Detailed assessments of the scientific information and
12 supportive data used in generating these tables can be found in the 1996 O3 AQCD (U.S.
13 Environmental Protection Agency, 1996). Key findings from health studies that have become
14 newly available since the 1996 criteria review are discussed below in later sections of this
15 chapter and any important consequent reaffirmations or modifications of findings of the types
16 summarized in Tables 8-2 and 8-3 are highlighted.
17
18
Table 8-1. Current National Ambient Air Quality Standards (NAAQS) in the
United States
Pollutant
Ozone
Date of
Promulgation
7/18/97
(62FR38856)
3/9/94
(58FR52852)
Primary NAAQS
0.08 ppm
(157 Mg/m3)
0.12 ppm
(235 Mg/m3)
Averaging Time
8-ha
l-hb
Secondary
NAAQS
Same as primary
Same as primary
"Based on the 3-year average of the annual fourth-highest daily maximum 8-h average O3 concentration
measured at each monitor within an area.
b The standard is attained when the expected number of days per calendar year with maximum hourly average
concentrations above 0.12 ppm is < 1.
January 2005 8-3 DRAFT-DO NOT QUOTE OR CITE
-------
Table 8-2. Gradation of Individual Responses to Short-Term Ozone Exposure in
Healthy Personsa*
Functional Response
FEVj
Nonspecific
bronchial responsiveness'5
Duration of
response
Symptomatic Response
Cough
Chest pain
Duration of response
Impact of Responses
Interference with normal
activity
None
Within normal
range (±3%)
Within normal
range
None
Normal
Infrequent
cough
None
None
Normal
None
Small
Decrements of
3 to < 10%
Increases of
< 100%
< 4 hours
Mild
Cough with deep
breath
Discomfort just
noticeable on
exercise or deep
breath
< 4 hours
Normal
None
Moderate
Decrements of
> 10 but < 20%
Increases of
< 300%
> 4 hours but
< 24 hours
Moderate
Frequent
spontaneous cough
Marked discomfort
on exercise or deep
breath
> 4 hours but
< 24 hours
Mild
A few sensitive
individuals choose
to limit activity
Large
Decrements of
> 20%
Increases of
> 300%
> 24 hours
Severe
Persistent
uncontrollable
cough
Severe discomfort
on exercise or
deep breath
> 24 hours
Moderate
Many sensitive
individuals choose
to limit activity
a See text for discussion; see Appendix A for abbreviations and acronyms.
b An increase in nonspecific bronchial responsiveness of 100% is equivalent to a 50% decrease in PD20 or PD100
(see Chapter 7, Section 7.2.3).
This table is reproduced from the 1996 O3 AQCD (Table 9-1, page 9-24) (U.S. Environmental Protection
Agency, 1996).
1 8.2 TRENDS IN UNITED STATES OZONE AIR QUALITY
2 8.2.1 Ozone Concentrations, Patterns
3 Ozone is monitored in the United States during "O3 seasons," which vary in length from
4 geographic region to region. The O3 season extends all year in the Southwest, but in most other
5 areas of the country, O3 is typically monitored from April to October. However, O3 is present
6 year-round, not only in polluted areas, but in clean areas as well. The median O3 concentration
7 in the United States from 1996 to 2000, averaged over the appropriate O3 season, was 33 ppb for
8 "urban" monitors located in Metropolitan Statistical Areas (MSAs); and it was 37 ppb for
January 2005
8-4
DRAFT-DO NOT QUOTE OR CITE
-------
Table 8-3. Gradation of Individual Responses to Short-Term Ozone Exposure in
Persons with Impaired Respiratory Systemsa*
Functional Response
FEVj change
Nonspecific
bronchial responsiveness'5
Airway resistance
(SRaw)
Duration of response
Symptomatic Response
Wheeze
Cough
Chest pain
Duration of response
Impact of Responses
Interference with normal
activity
None
Decrements of
<3%
Within normal
range
Within normal
range (±20%)
None
Normal
None
Infrequent
cough
None
None
Normal
None
Small
Decrements of
3 to < 10%
Increases of < 100%
SRaw increased
< 100%
< 4 hours
Mild
With otherwise
normal breathing
Cough with deep
breath
Discomfort just
noticeable on
exercise or deep
breath
< 4 hours
Mild
Few individuals
choose to limit
activity
Moderate
Decrements of
> 10 but < 20%
Increases of
< 300%
SRaw increased up
to 200% or up to
15cmH2O/s
> 4 hours but
< 24 hours
Moderate
With shortness of
breath
Frequent
spontaneous
cough
Marked
discomfort on
exercise or deep
breath
> 4 hours, but
< 24 hours
Moderate
Many individuals
choose to limit
activity
Large
Decrements of
> 20%
Increases of
> 300%
SRaw increased
> 200% or more
thanl5cmH2O/s
> 24 hours
Severe
Persistent with
shortness of
breath
Persistent
uncontrollable
cough
Severe discomfort
on exercise or
deep breath
> 24 hours
Severe
Most individuals
choose to limit
activity
Medical treatment
No change
Normal medication
as needed
Increased Physician or
frequency of emergency room
medication use or visit
additional
medication
a See text for discussion; see Appendix A for abbreviations and acronyms.
b An increase in nonspecific bronchial responsiveness of 100% is equivalent to a 50% decrease in PD20 or PD100
(see Chapter 7, Section 7.2.3).
This table is reproduced from the 1996 O3 AQCD (Table 9-2, page 9-25) (U.S. Environmental Protection Agency,
1996).
January 2005
$-5
DRAFT-DO NOT QUOTE OR CITE
-------
1 monitors located outside MSAs. Median daily maximum 8-h concentrations between 10:00 a.m.
2 and 6:00 p.m. were 46 and 47 ppb for monitors located in and outside MSAs, respectively.
3 Median daily maximum 1-h concentrations were 56 ppb for monitors located in MSAs and 55
4 ppb for monitors located outside of them. The daily maximum 1-h concentrations tended to be
5 much higher in some large urban areas or in areas downwind of them, e.g., they were 202 ppb in
6 Houston, TX in 1999 and 161 ppb in 2000. Daily 1-h maximum O3 concentrations were lower
7 in nonurban areas of the country but still above 120 ppb in many locations. Eight-hour daily
8 maximum concentrations were not as high as 1-h daily maxima, but they also tended to be highly
9 correlated with the 1-h maxima.
10 Within individual MSAs, O3 concentrations tend to be well correlated across monitoring
11 sites, although variations in concentrations can be substantial. In many city centers, O3
12 concentrations tend to be lower than in either upwind or downwind areas, largely due to NO
13 emitted by motor vehicles. Thus, although emissions of nitrogen oxides and VOCs from motor
14 vehicles contribute to O3 formation, the relationship to O3 concentrations is not straightforward
15 in terms of proximity to mobile sources. In urban areas with high traffic density or near
16 highways, emissions of NO from traffic react with ozone, thereby reducing its concentration.
17 For example, much lower ozone concentrations overall are found in downtown Los Angeles
18 (e.g., in Lynwood) than at sites located further downwind (e.g., in San Bernadino). The much
19 higher levels are formed from photochemical reactions involving the urban emissions, including
20 products produced as the result of reactions titrating ozone in the urban core. Thus, ozone
21 concentrations tend to be higher downwind of urban centers, and they decrease again in going to
22 areas that are remote from precursor sources.
23
24 8.2.2 Seasonal Variations
25 Ozone concentrations tend to peak in early to mid-afternoon in areas where there is strong
26 photochemical activity and to peak later in the afternoon or early evening in areas where
27 transport is more important in determining the O3 abundance. Summertime maxima in O3
28 concentrations occur in U.S. areas where substantial photochemical activity acts on O3
29 precursors emitted as the result of human activities. Monthly maxima can occur anytime from
30 June through August. However, springtime maxima are observed in National Parks, mainly in
31 the western United States and at a number of other relatively unpolluted monitoring sites
January 2005 8-6 DRAFT-DO NOT QUOTE OR CITE
-------
1 throughout the Northern Hemisphere. For example, the highest O3 concentrations at
2 Yellowstone National Park tend to occur during April and May. Typically, monthly minima
3 tend to occur from November through February at polluted sites and during the fall at relatively
4 remote sites.
5
6 8.2.3 Long-Term Trends
7 Nationwide, 1-h O3 concentrations decreased -29% from 1980 to 2003 and by -16% from
8 1990 to 2003; and, for the 8-h standard, O3 levels decreased -21% since 1980 and -9% from
9 1990 to 2003. Note that 1-h and 8-h O3 levels continue to decrease nationwide, but the rate of
10 decrease has slowed since 1990. These trends have not been uniform across the United States.
11 In general, O3 reductions have been largest in New England and in states along the West Coast
12 and smallest in the Midwest. Downward trends in O3 in California have been driven mainly by
13 reductions in Southern California, with reductions in other areas not being as large.
14
15 8.2.4 Ozone Interactions with Other Ambient Pollutants
16 Data for other oxidants (e.g., H2O2, PAN) and oxidation products (e.g., HNO3, H2SO4)
17 in the atmosphere are not as abundant as they are for O3. Because data for these species are
18 usually obtained only as part of specialized field studies, it is difficult to relate O3 concentrations
19 to ambient levels of other species. In general, these secondary species are expected to be at least
20 moderately positively correlated with O3. On the other hand, primary species are expected to be
21 more highly correlated with each other than with secondary species, provided that the primary
22 species originate from common source areas. Relationships between ambient O3 and PM2 5
23 concentrations are complex, because particulate matter (PM) is not a single distinct chemical
24 species, but rather a mix of primary and secondary species. As an example of the subject
25 complexity, PM2 5 concentrations were positively correlated with O3 during the summer, but
26 negatively correlated with O3 during the winter at Ft. Meade, MD. More data are needed before
27 this result can be applied to other areas; and the degree of positive or negative correlation
28 between O3 and PM or other pollutants may vary to a greater or lesser extent by season.
29
30
January 2005 8-7 DRAFT-DO NOT QUOTE OR CITE
-------
1 8.3 AMBIENT OZONE EXPOSURE ASSESSMENTS
2 Exposure to O3 and related photochemical oxidants varies with time due to changes in
3 ambient concentration and because people move between locations with different O3
4 concentrations. The amount of O3 delivered to the lung is not only influenced by the ambient
5 concentration but also by the individual's breathing rate. Thus, activity level is an important
6 consideration in determining the potential exposure and dose received.
7 The use of ambient air monitoring stations is still the most common surrogate for assigning
8 exposure estimates in epidemiological studies. Since the primary source of O3 exposure is the
9 ambient air, monitoring concentration data should provide a relative assignment of exposure
10 with time if: concentrations were uniform across the region; time-activitys pattern were the
11 same across the population; and housing characteristics, such as ventilation rates and O3 sinks
12 contributing to its indoor decay rates, were constant for the study area. Because these factors
13 vary by population and location, there tend to be errors not only in estimating the magnitude of
14 the exposure but also in relative exposure assignments based solely on ambient monitoring data.
15 Still, such data can be used to evaluate health outcomes associated with chronic O3 exposure.
16
17 8.3.1 Personal Exposure
18 Personal O3 concentrations have been measured for children, outdoor workers, and
19 individuals with COPD, populations potentially susceptible to respiratory irritants. Children and
20 outdoor workers have somewhat higher exposures than other individuals, because they spend
21 more time outdoors engaged in moderate and heavy exertion. Children are also more active
22 outside and, therefore, have a higher breathing rate than most adults. However, the available
23 exposure studies are not sufficient to allow for confident generalization of differences in
24 exposure between the general population and potentially susceptible subpopulations.
25
26 8.3.2 Indoor Concentrations
27 There are few indoor sources of O3. Generally, O3 enters indoor environments through
28 infiltration from outdoors and through building components such as windows, doors, and
29 ventilation systems. The concentration of O3 in indoor environments is primarily dependent on
30 the outdoor O3 concentration and the air exchange rate (AER) or outdoor infiltration. Ozone
31 concentrations indoors are higher during the outdoor O3 season.
January 2005 8-8 DRAFT-DO NOT QUOTE OR CITE
-------
1 Ozone reacts indoors with other contaminants, possibly producing compounds with greater
2 toxicity. Ozone concentrations are typically lower indoors than outdoors, in part due to gas
3 phase reactions that produce other oxidants analogous to the production of photochemical smog.
4 The production of these species indoors is a function of the indoor O3 concentration and the
5 presence of the other necessary precursors, volatile organic compounds (VOCs) and nitrogen
6 dioxide (NO2 ), along with an optimal AER.
7 Several studies have measured O3 concentrations in residences, schools, office buildings
8 and museums, and concentrations varied at all locations. However, indoor concentrations were
9 generally associated with the AER in the indoor environment (increasing with higher AER) and
10 generally tend to be notably lower than outdoor ambient O3 levels. For example, one study
11 examining the relationship between O3 concentrations indoors and outside of a school in
12 New England reported averaged O3 concentrations of 20 ppb indoors and 40 ppb outdoors. With
13 regard to mobile source microenvironments, as is the case for other enclosed environments,
14 ozone exposures depend on the extend of mixing of outdoor air into the vehicle cabin.
15 If windows are kept open, ozone in the vehicle may be expected to approach outdoor values;
16 however, if windows are kept closed and there is air conditioning, then interior values could be
17 much lower than those outside, especially if recirculated air is used. For example, in one N.C.
18 study involving police cars with air conditioning and recirculated air, O3 concentrations in the
19 vehicle cabin (11.7 ppb average) were less than half those outside (28.3 ppb average at outdoor
20 monitoring sites in the area).
21
22
23 8.4 SYNTHESIS OF AVAILABLE INFORMATION ON
24 OZONE-RELATED HEALTH EFFECTS
25 The integrated synthesis of the latest available information on O3-related health effects
26 poses large challenges, especially in view of the emergence of important new information
27 generated since the 1996 O3 AQCD, which adds greatly to the complexity of any integrative
28 assessment. Such information includes new findings from:
29 • Epidemiological studies, reflecting progress in addressing many research recommendations
from the last review as well as raising new issues and reevaluating previously addressed
issues that remain important in interpreting the body of epidemiological evidence and
characterization of its strengths and limitations;
January 2005 8-9 DRAFT-DO NOT QUOTE OR CITE
-------
1 • Experimental toxicological studies using laboratory animals and controlled human
exposures aimed at understanding the potential biochemical mechanisms underlying toxic
effects, pathology, and susceptibility.
2 Thus despite substantial progress, challenges remain in integrating the new scientific information
3 on O3 health effects, including newly reported epidemiological evidence for associations
4 between ambient O3 exposures and increased mortality risks among human populations.
5
6 8.4.1 Assessment of Epidemiological Evidence
7 Based on the O3 epidemiological evidence available at the time, the 1996 O3 AQCD
8 arrived at the following conclusions:
9
10 An association between daily mortality and O3 concentration for areas with high O3
11 levels (e.g., Los Angeles) has been suggested, although the magnitude of such an effect
12 is unclear. Increased O3 levels are associated with increased hospital admissions and
13 emergency department visits for respiratory causes. Analyses from data in the
14 northeastern United States suggest that O3 air pollution is associated with a substantial
15 portion (on the order of 10 to 20%) of all summertime respiratory hospital visits and
16 admissions. Pulmonary function in children at summer camps in southern Ontario,
17 Canada, in the northeastern United States, and in Southern California is associated
18 with O3 concentration." (U. S. EPA, 1996, pl-29).
19
20 The 1996 O3 AQCD further stated that only suggestive epidemiologic evidence existed for health
21 effects of chronic ambient O3 exposure in the population, and this was partly due to an inability
22 to isolate potential effects related to O3 from those of other pollutants, especially PM (U.S.
23 Environmental Protection Agency, 1996).
24 The scientific strength and limitations of the growing body of epidemiological evidence for
25 associations between exposure to O3 and health effects discussed in this section is based
26 primarily on Chapter 7 evaluations. The following criteria were considered in assessing the
27 relative scientific quality of the epidemiologic studies: (1) quality of exposure metrics to
28 evaluate credible exposure indicators; (2) quality and size of the study groups/population to
29 arrive at meaningful analysis of health effects; (3) robustness of reported associations (based on
30 defined health endpoint criteria), potential confounding by copollutants; (4) the strength of
31 reported associations, in terms of magnitude, statistical significance and statistical power of
32 effects estimates; (5) temporality, in terms of lag periods between exposure and observed effects;
33 and (6) biological plausibility, consistency and coherency of the reported findings. The body of
34 epidemiological evidence is further considered in terms of its coherence within itself and in
January 2005 8-10 DRAFT-DO NOT QUOTE OR CITE
-------
1 relation to findings derived from controlled human exposure studies which, overall, provide
2 insights into the plausibility of reported O3 human health effects reflecting causal relationships.
3 Many newly available epidemiological studies have provided additional evidence for
4 O3-related health effects beyond that which was known previously. Significant statistical
5 associations have been observed by various investigators between acute O3 exposure and several
6 respiratory health endpoints, including: mortality; hospital admissions; emergency department
7 visits; respiratory illness and symptoms; and changes in pulmonary function. Similarly, long-
8 term exposure to O3 has been associated with: increased morbidity; development of respiratory
9 disease; and declines in lung function and lung function growth. The epidemiological studies
10 that have been conducted in areas across the United States and Canada, as well as in Europe,
11 Latin America, Australia and Asia, are summarized in Annex 7. Based on evidence extracted
12 from the full body of epidemiologic studies that have been carried out and reviewed since the
13 1996 O3 AQCD (U.S. Environmental Protection Agency, 1996), it has been well demonstrated
14 that deleterious human health outcomes are positively associated with ambient O3 concentrations
15 currently encountered in the United States and elsewhere.
16
17 8.4.2 Strength of Epidemiological Associations
18 As quoted above, assessments in the 1996 O3 AQCD supported a consistent relationship
19 between O3 concentration and respiratory illness, hospital visits and reduced lung function.
20 However, due to insufficient evidence examining O3-mortality associations and uncertainties
21 regarding weather model specification, the 1996 O3 AQCD was limited to only a very qualitative
22 assessment of O3-mortality associations. Since then, generalized Additive Models (GAMs) have
23 become widely utilized for epidemiologic analysis of health effects attributable to air pollution,
24 making quantitative estimation of O3-mortality risks much more meaningful. On the other hand,
25 certain statistical issues raised with regard to use of default convergence criteria in applications
26 of commercially available software employed for GAM analyses in many newly available air
27 pollution epidemiologic studies led to a reanalyses of previously published studies and revised
28 estimation of reported PM — mortality/morbidity risks. The impacts of the GAM-related
29 statistical issues were thoroughly discussed in the 2004 PM AQCD (U.S. Environmental
30 Protection Agency, 2004). Of most importance here, the reanalyses of a number of studies,
31 comparing results using default GAM convergence criteria to results from analyses using
January 2005 8-11 DRAFT-DO NOT QUOTE OR CITE
-------
1 stringent GAM convergence criteria and/or from GLM analyses, found little difference among
2 the O3 effect estimates obtained (as discussed in detail in Chapter 7 of this document).
3 Furthermore, the magnitude of the effect-size estimates observed from O3-mortality relationships
4 tend to be relatively consistent across the newly available studies and to compare well with those
5 obtained for O3-morbidity endpoints.
6
7 8.4.3 Acute Exposure Studies
8 Numerous epidemiological studies carried out over the past decade have added evidence to
9 the knowledge base that was assessed in the 1996 O3 AQCD, which included both (a) individual-
10 level camp and exercise studies that established a relationship between human lung function
11 decline with ambient O3 exposure and (b) aggregate time-series studies that suggested positive
12 relationships for O3-related respiratory morbidity. The new studies reviewed in Chapter 7 in this
13 document included numerous field/panel studies and time-series studies from various regions.
14 In field studies on the effects of air pollution exposure, the most common health outcomes
15 measured were lung function and respiratory symptoms. Time-series studies examined daily
16 hospital admissions, emergency department visits, and mortality data.
17
18 8.4.3.1 Panel Studies
19 Many of the new field/panel studies reviewed in Chapter 7 and the controlled human
20 exposure studies reviewed in Chapter 6 of this document provide additional data supporting two
21 major findings reported in the 1996 O3 AQCD, i.e.: (1) O3-related lung function decrements and
22 (2) respiratory symptoms in exercising healthy subjects and asthmatic subjects. Pulmonary
23 function was determined by either spirometry (forced expiratory volume in 1 s [FEVJ and
24 forced vital capacity [FVC]) or by peak expiratory flow (PEF) meters. While the spirometric
25 parameter, FEVj is a stronger and more consistent measure of lung function, PEF is more
26 feasibly performed in field studies.
27 In a number of newly available field/panel studies, FEVj was measured in panels of
28 exercising children, outdoor workers, and adult hikers exposed to ambient O3 while experiencing
29 elevated exertion levels. Collectively, the results of the new studies (discussed in Section
30 7.2.3.1) confirm and extend those from analogues field/panel studies assessed in the 1996 O3
31 AQCD and findings from experimental controlled human exposure studies indicating that acute
January 2005 8-12 DRAFT-DO NOT QUOTE OR CITE
-------
1 O3 exposures prolonged over several hours and combined with elevated levels of exertion or
2 exercise magnify O3 effects on lung function, as evaluated in terms of FEVj.
3 For example, six field studies by three different research groups of 7- to 17-year-old,
4 healthy (nonasthmatic) children exposed for several hours to ambient O3 during increased
5 physical exertion in summer camp activities were assessed in the 1996 O3 AQCD. When
6 analyzed together by consistent statistical methods, the data from those studies showed an
7 average relationship between afternoon FEVj and concurrent 1-h O3 concentrations of
8 -0.50 mL/ppb, with individual slopes ranging from -0.19 to -1.29 mL/ppb (likely reflecting, in
9 part, the multi-hour O3 exposures preceding the pulmonary function tests). Four new filed/panel
10 studies assessed in Section 7.2.3.1 of this document that evaluated pulmonary function in healthy
11 school-aged children exposed to mean 1-h O3 concentrations ranging from -20 to 120 ppb found
12 exposure-response functions of approximately -0.23 to -1.42 mL/ppb. Also, two other studies
13 assessed in the 1996 document that measured lung function before and after well-defined
14 exercise events (0.5-h long) in adults during exposures to ambient O3 across 4 to 135 ppb found
15 exposure-response slopes of -0.4 mL/ppb. In comparison, four new studies of healthy adult
16 workers (street workers, berry pickers) and hikers engaged in prolonged (> 6 to 8 h) strenuous
17 physical exertion while exposed to mean ambient O3 concentrations of-26 to 70 ppb (1-h
18 maximum) or 40 ppb (8-h average) reported exposure-response slopes of-1.13 to -3.8 mL/ppb
19 (as assessed in Chapter 7 of this document). The most representative data is that of Korrick et al.
20 (1998) from a U.S. study of adult hikers that provided outcome measures stratified by gender,
21 age, smoking-status, and presence of asthma within a population capable of above-normal
22 exertion.
23
24 8.4.3.2 Asthma Panels
25 Several studies assessed in the 1996 O3 AQCD that evaluated elevated respiratory
26 symptoms and/or pulmonary function decrements in asthmatic children showed greater
27 responses in asthmatic than nonasthmatic subjects, suggesting that asthmatic individuals might
28 constitute a sensitive population group in oxidant epidemiologic studies.
29 Additional panel studies carried out over the past decade to understand the effect of acute
30 exposure to O3 in asthmatics evaluated either (a) lung function by PEF and/or (b) respiratory
31 symptoms (i.e. cough, wheeze, shortness of breath and medication use) ascertained by
January 2005 8-13 DRAFT-DO NOT QUOTE OR CITE
-------
1 questionnaire. Several regional studies, typically consisting of children with asthma, collectively
2 tend to confirm O3-induced decrements in pulmonary function both in the United States and in
3 other countries (see Section 7.2.3.2). One U.S. multicity study (Mortimer et al., 2002), featuring
4 the largest panel of asthmatic children from eight urban areas, observed a statistically significant
5 decrement in PEF with a cumulative lag of 1 to 5 days (-1.18% per 30 ppb increase in 8-h
6 average O3). Overall, these studies suggest that ambient O3 exposures may be associated with
7 enhanced decreases in lung function in asthmatics.
8 Most studies evaluating respiratory symptoms (i.e. cough, shortness of breath, and wheeze)
9 and the increased use of asthma medications related to O3 exposure also focused on asthmatic
10 children. Several U.S. studies observed significant associations between O3 ambient
11 concentrations and increased symptoms or asthma medication use that appeared to be fairly
12 robust to adjustment for copollutants. Analyses by Mortimer et al. (2002), conducted in eight
13 U.S. urban areas, and Gent et al. (2003), conducted in the New England region, have used data
14 from large sample populations likely to be representative of U.S. data. Odds ratios from six new
15 studies for prevalence of cough among asthmatic children mainly varied from ~1.05 to 1.5
16 standardized per 40 ppb increase in 1-h max O3 (or equivalent) or -1.35 for all symptoms per
17 30 ppb increase in 8-h O3 concentration.
18
19 8.4.3.3 School Absences
20 Two new U.S. studies (Chen et al., 2000; Gilliland et al., 2001) investigated the
21 relationship between ambient O3 concentrations and school absenteeism. Both studies were
22 carried out during a period when O3 levels were mostly below the highest levels typically
23 observed in the summer season. In the Chen et al. study, with a distributed lag of 1 to 14 days, a
24 10.4% excess rate of school absences was found per 40 ppb increase in daily 1-h max O3
25 concentrations. The study by Gilliland et al. (2001), which was able to distinguish specific
26 illness-related absence, found significant O3 effects on school absences due to respiratory causes
27 with a lag period ranging from 2 to 4 weeks. A notably higher respiratory-related absence rate
28 increase (147% increase per 30 ppb increase in 8-h O3) was seen versus that seen for
29 non-respiratory causes (61% increase per 30 ppb).
30
January 2005 8-14 DRAFT-DO NOT QUOTE OR CITE
-------
1 8.4.3.4 Field Studies on Cardiovascular Effects
2 A limited number of air pollution studies have examined cardiac physiologic endpoints,
3 including heart rate variability, arrhythmia, and risk of myocardial infarction. One large U.S.
4 study (Liao et al., 2004) found that ambient O3 concentrations were inversely associated with
5 ECG high-frequency power readings among whites. However, consistently more pronounced
6 associations were suggested between PM10 and heart rate variability among persons with a
7 history of hypertension. While these results may somewhat be supportive of hypothesized air
8 pollution-heart rate variability-cardiovascular disease pathways at the population level, a lack of
9 consistency within or across the limited available studies indicates that additional studies are
10 needed before any clear conclusions can be made.
11
12 8.4.4 Emergency Department Visits and Hospital Admissions
13 Many time-series studies reviewed in the 1996 O3 AQCD indicated positive associations
14 between O3 air pollution and increased hospital admissions. Strong evidence establishing a
15 correlation between O3 exposure and increased exacerbations of preexisting respiratory disease
16 in the general public were reported at 1 h-maximum O3 concentrations < 0.12 ppm. Several
17 studies have been published over the past decade examining the temporal associations between
18 O3 exposures and emergency department visits for respiratory diseases (see Table AX7-2 in
19 Annex 7). Among studies with adequate controls for seasonal patterns, many reported at least
20 one significant positive association involving O3. Overall, the analyses of data for asthma-
21 related emergency room visits clearly indicate increased respiratory morbidity during warm
22 seasons when ambient O3 concentrations are high. These studies are summarized in Figure 8-1,
23 showing both yearly and seasonal results from U.S. and Canadian studies.
24 Studies reviewed in Chapter 7 (Section 7.3.3) reported a significant O3 effect on respiratory
25 hospital admissions. While some inconsistencies are noted across studies, the evidence is
26 supportive of significant and robust O3 effects on hospitalizations for various respiratory
27 diseases. Large multeity studies, as well as many studies from individual cities, have reported
28 significant O3 associations with total respiratory asthma and chronic obstructive pulmonary
29 disease (COPD) hospitalizations, especially in studies analyzing the O3 effects during the
30 summer or warm season (Figure 8-2). The most robust and informative results on the effects of
31 O3 on respiratory hospital admissions are from muticity studies that used a consistent analytical
January 2005 8-15 DRAFT-DO NOT QUOTE OR CITE
-------
Percentage Change in Emergency Department Visits for Asthma
Jaffeetal. (2003):
Cincinnati,OH (age 5-34)
Cleveland OH (age 5-34)
Columbus,OH (age 5-34)
Jaffe etal. (2003):
Cincinnati, Cleveland,
and Columbus,OH (age 5-34)
Friedman etal. (2001):
Atlanta, GA (age 1-16)
Peel etal. (2004):
Atlanta,GA (all ages)
Tolbert etal. (2000):
Atlanta, GA (age 0-16)
Stieb etal. (1996):
St.John,Canada (all ages)
-20 0 20 40
i i 1 i i i i i
• Inn °
• ho °
• laa 3
• hn 0
^X- lag 0-2
• lag 0-2 (SE not given)
X Iag1
60 80 1(
i i i
X All year
• Warm
lag 0 1
ho 0 ">
Figure 8-1. Ozone-associated percent change (95% CI) in emergency room visits for
asthma. Percent change effects are per 40 ppb increase in 1-h maximum O3 or
equivalent. Analysis includes all age unless otherwise noted, and only studies
conducted in the United States and Canada are presented.
1 methodology across a broad geographic area, such as the 16 Canadian cities study by Burnett
2 et al. (1997a). Of the few studies that examined the relationship between O3 and hospital
3 admissions for cardiovascular diseases, most did not find any consistent positive associations.
4
5 8.4.5 Acute Effects of Ozone on Mortality
6 Due to the limited number of studies and uncertainties regarding weather model
7 specifications, no meaningful quantitative assessment of O3-mortality associations were possible
8 in the 1996 O3 AQCD. However, newly available large multicity studies designed specifically to
9 examine the effect of O3 on mortality have provided much more robust and credible information.
10 The results from two key studies carried out in 95 U.S. communities (U.S. National Morbidity,
11 Mortality Air Pollution Study [NMMAPS]; Bell et al., 2004) and in 23 European cities (Air
12 Pollution on Health: European Approach [APHEA]; Gryparis et al., 2004) showed positive and
13 significant O3 effect estimates for all cause (nonaccidental) mortality (Figure 8-3; see Section 7.4
14 of Chapter 7 for complete discussion). The influence of season on O3-mortality risk estimates
15 from various U.S. and Canadian time-series studies is also shown in Figure 8-3. In the APEHA
January 2005
8-16
DRAFT-DO NOT QUOTE OR CITE
-------
Linn etal. (2000):
Los Angeles, CA (age 30+)
Schwartz etal. (1996):
Cleveland, OH (age 65+)
Burnett etal.(1997a):
16 Canadian Cities (all ages)
Burnett etal.(1997b): .
Toronto,Canada (all ages)
Burnett etal. (2001):
Toronto, Canada (age< 2)
Luginaah etal. (2004):
Windsor,Canada •
(all ages, males only)
Luginaah etal. (2004):
Windsor,Canada •
(all ages,females only)
Percentage Change in Respiratory Hospitalization
-20 0 20 40 60 80 100
i i i i i i i i i i i i
-x-
-e-
lagO
- lag 1-2
Iag1
lag 1-3
Iag1
lag 0-4
Figure 8-2. Ozone associated percent change (95% CI) in total respiratory hospitalizations
(95% CI) for all year and for by season. Percent change effects are per
40 ppb increase in 1-h maximum O3 or equivalent. Analysis includes all age
unless otherwise noted, and only studies conducted in the United States and
Canada are presented.
1 study (Gryparis et al., 2004), significant O3 effects were observed only during the warm season.
2 With the exception of one study (Chock et al., 2000), all risk estimates from warm-season-only
3 analyses were positive, with the majority indicating statistical significance at p < 0.05.
4 The effect of PM on mortality was thoroughly discussed in the 2004 PM AQCD. Because
5 PM indices correlate highly with O3 levels in some areas, confounding of the O3-mortality
6 association by PM is of great concern. Figure 8-4 shows O3-mortality risk estimates with and
7 without adjustment for PM indices. Collectively, the results indicate that the O3 risk estimates
8 were not substantially affected with the addition of PM in the various reported analyses.
9 The effect estimates presented in Figures 8-3 and 8-4 lead to the following findings:
10 (1) O3-mortality associations from several U.S. and Canadian studies reported fairly consistent
11 and positive combined estimates of 0.4 to 4.8% excess risk of total nonaccidental mortality per
12 40 ppb increase in 1-h maximum O3 (excluding the Vedal et al., 2003 study, which examined the
January 2005
8-17
DRAFT-DO NOT QUOTE OR CITE
-------
Bell etal. (2004): U.S. 95 communities -
Sametetal.(2000;reanalysis _
Dominici et al. 2003): U.S. 90 cities
Schwartz (2004): U.S. 14 cities -
Kinney and Ozkaynak (1991): Los Angeles County, CA —
Kinneyetal.fi 995): Los Angeles County, CA —
Ostro (1995): San Bernadino County
and Riverside County, CA
Fairley (2003): Santa Clara County,CA -
Gamble (1998):Dallas,TX -
Dockeryetal.fi 992): St. Louis, MO -
ItoandThurston (1996): Cook County, IL -
Lippmannetal.(2000;reanalysisIto,2004): Detroit,Ml -
Lipfertetal.(2000a): Philadelphia, PA -
Lipfert et al. (2000a): 4 Counties in PA -
Lipfert et al. (2000a): 7 Counties in PA and NJ -
Lipfert et al. (2000a): 7 Counties in PA and NJ (age 0-65) -
Lipfert etal.(2000a): 7 Counties in PA and NJ (age 65+) -
Moolgavkar et al. (1995): Philadelphia, PA -
Chock etal. (2000): Pittsburgh, PA (age 0-74) -
Chock etal. (2000): Pittsburgh, PA (age 75+) -
Dockery et al. (1992): Eastern Tennessee -
Klemm and Mason (2000): Atlanta, GA -
Klemm etal. (2004): Atlanta, GA (age 65+) -
Vedaletal.(2003):Vancouver,Canada -
Percentage Change in Mortality
-10
0
I
10
20
_l
30
_l
Multiple Cities/Regions
rX-
Single City/Region
x Iag1
•X- Iag1
lag 1-2
lag 0-1
lag 0-1
lag 0-1
lag 0-1
lag 0-1
-e-
-e-
lagO
lag 0-6
Iag1
lagO
• (SE not given; significant at 0.05 level)
O (SE not given; not significant at 0.05 level)
lag 0-1
lag 0-3
' Confidence bands not provided but
noted as significant at p=0.055
Iag1
lagO
Iag1
lag 0-1
Figure 8-3. All cause (nonaccidental ) O3 excess mortality risk estimates (95% CI)
for all year and for by season per 40 ppb increase in 1-h maximum O3
or equivalent. Analysis includes all age unless otherwise noted, and
only studies conducted in the United States and Canada are presented.
January 2005
8-18
DRAFT-DO NOT QUOTE OR CITE
-------
Percentage Change in Mortality
Samet et al. (2000, reanalysis
Dominicietal.,2003): U.S. 90 cities —
with PM-|o ~
Schwartz (2004): U.S. 14 cities -
with PM10 -
Kinney and Ozkaynak (1991): Los Angeles County, CA —
with KM -
Kinney etal.(1995): Los Angeles County, CA —
with PM10 —
Ostro et al. (1995): San Bernadino County
and Riverside County, CA —
with PM2 5 —
Fairley (2003): Santa Clara County,CA -
withPM25 -
Gamble (1998): Dallas,TX -
with PM10 -
ItoandThurston (1996): CookCounty.lL —
with PM-|o ~~
Lipfert etal.(2000a): Philadelphia, PA -
with PM2 5 —
with PM10 -
Lipfert et al. (2000a): 4 Counties in PA —
with PM2 5 —
with PM-|o —
Lipfert et al. (2000a): 7 Counties in PA and NJ -
with PM2 5 —
with PM-|o —
Lipfert et al.(2000a): 7 Counties in PA and NJ
(age 0-65) -
with PM2 5 -
with PM10 —
Lipfert et al.(2000a): 7 Counties in PA and NJ
(age 65+) —
with PM2 5 -
Moolgavkar et al. (1995):Philadelphia, PA —
with TSP -
Moolgavkar et al. (1995):Philadelphia, PA —
with TSP -
Chock et al. (2000): Pittsburgh, PA (age 0-74) —
with PM-|o —
Chock et al. (2000): Pittsburgh, PA (age 0-74) -
withPMjQ —
Chock et al. (2000): Pittsburgh, PA (age 0-74) -
withPM10 -
Chock et al. (2000): Pittsburgh, PA (age 75+) -
with PM10 -
Chock et al. (2000): Pittsburgh, PA (age 75+) -
withPM10 -
Chock et al. (2000): Pittsburgh, PA (age 75+) -
withPM10 -
-10
I
-5
I
lagO
0
|
10
I
15
I
lag 0-1 •
lag 0-1
lag 0
lag 0-1
lag 0-
-e-
lagO
lagO
Iag1
Iag1
lagO
All Year
O 03 only
• 03 with PM
Warm Season
A 03 only
A 03 with PM
Cool Season
D 03 only
• O3withPM
lag1-2
O
O
O
-1 •
0
0
-1 •
•
lag 1-2
Confidence bands not
provided but noted as
significant at p=0.055
Noted as nonsignificant
at p=0.055
Confidence bands not
provided but noted as
significant at p=0.055
^~lag1
Iag1
"" lagO
lagO
1 lagO
lag 0
lagO
lagO
Figure 8-4. All cause (nonaccidental) O3 excess mortality risk estimates (95% CI)
with adjustment for PM indices for all analyses. Percent change effects
are per 40 ppb increase in 1-h maximum O3 or equivalent, and adjusted
for PM indices. Only U.S. and Canada studies are presented.
January 2005
8-19
DRAFT-DO NOT QUOTE OR CITE
-------
1 O3-mortality association in a region with very low O3 concentrations); (2) season-stratified
2 analyses indicated that the O3-mortality effect estimates were significant and positive in the
3 warm season, with larger effects observed compared to the year-round and cool-season analyses;
4 (3) the risk estimates were robust to adjustment for PM indices, indicating that O3 effect on
5 mortality is independent of PM.
6 The results from the U.S. studies are generally consistent with those from other regions of
7 the world. The results from all available studies indicate substantial strength in the
8 epidemiological evidence for the association between exposure to O3 and excess risk of total
9 nonaccidental mortality. The overall range of estimates were relatively narrow with the positive
10 estimates between 0 and 7% per 40 ppb increase in 1-h maximum O3 or equivalent.
11
12 8.4.6 Chronic Ozone Exposure Studies
13 There were a limited number of studies reported in the 1996 O3 AQCD that provided
14 insufficient evidence to consider potential health effects of long-term ambient O3 exposures.
15 Several longitudinal epidemiological studies carried out in the past decade evaluated the
16 potential effects of chronic exposure (several weeks to many years) to O3 on lung function,
17 respiratory symptoms, lung inflammation, asthma prevalence, and mortality.
18 Evidence from recent long-term morbidity studies indicates that chronic exposure
19 to O3 may have negative effects on inflammation, respiratory symptoms, and development of
20 asthma; however, the evidence is limited and, at times, lacks consistency. The strongest
21 evidence for an effect from chronic O3 exposure is derived from studies examining lung function
22 measurements. Seasonal decrements or reduced growth in lung function measures have been
23 reported in several studies; however, the changes appear to be transient. Studies of lung function
24 decrements with longer-term or annual data are not as conclusive.
25 Very few studies have investigated the effect of long-term O3 exposure on mortality.
26 Uncertainties regarding the exposure period of relevance and inconsistencies across mortality
27 outcomes and gender raise concerns regarding plausibility. The most representative U.S. study
28 by Pope et al. (2002) observed positive but non-significant associations between O3 exposure
29 and all cause mortality. Thus, the current evidence is inconclusive for a relationship between
30 chronic O3 exposure and increased mortality risk.
31
January 2005 8-20 DRAFT-DO NOT QUOTE OR CITE
-------
1 8.4.7 Robustness of Epidemiological Associations
2 In evaluating the strength of the epidemiological evidence, the magnitude of observed O3
3 effect estimates and their statistical significance is important; however, consideration must be
4 given to the precision of the effect estimates and the robustness of the effects associations.
5 Examining the robustness of the associations includes the impact of alternative models, model
6 specifications for temporal trends and meteorological factors, and potential confounding by
7 copollutants. Also of interest are issues related to exposure assessment and measurement error.
8 A detailed discussion on each of these topics can be found in Chapter 7 (Section 7-6). The
9 following sections focus on the extent to which the current epidemiological findings can be
10 considered robust.
11
12 8.4.7.1 Exposure Issues: Ambient versus Personal
13 In time-series studies, or other large-scale epidemiology studies of long duration, it is often
14 impractical and unfeasible to monitor the personal exposure of each subject. Thus, the ambient
15 concentrations of O3 and other air pollutants at central monitoring sites are often used as a proxy
16 for individual exposure measurements. The relationship between ambient O3 concentrations and
17 personal O3 exposure levels varies depending on factors such as time spent outdoors, ventilation
18 conditions, personal factors, and air quality indices. Because ambient concentrations often
19 overestimate true personal O3 exposures, the use of ambient data likely tends to underestimate
20 the effect of the air pollutant on health.
21 Comparisons between ambient concentrations and personal exposures to O3 have indicated
22 that ambient concentrations do not reflect the variability of individual exposures. However,
23 daily ambient O3 concentrations have been shown to be well-correlated to daily-averaged
24 personal exposures obtained by aggregating the personal measurements from all subjects.
25 Therefore, though unresolved issues remain, the evidence suggests that ambient O3 levels
26 measured at central monitors may serve as valid surrogate measures for aggregate personal
27 exposures in time-series studies investigating mortality and hospitalization outcomes.
28 Ambient O3 measurements from the three main exposure indices (1-h maximum O3, 8-h
29 maximum O3, and 24-h average O3) used were highly correlated. As such, the excess health risk
30 estimates and significance of associations appear to be comparable for the same distributional
31 increment. The commonly used 8-h maximum O3 or 8-h average O3 index continue to be
January 2005 8-21 DRAFT-DO NOT QUOTE OR CITE
-------
1 appropriate choices, as no other exposure index has been demonstrated to offer a better
2 advantage.
3
4 8.4.7.2 Confounding by Temporal Trends and Meteorologic Effects
5 The effect of seasonal differences in the health outcomes and O3 exposure levels were
6 recognized in the 1996 O3 AQCD; and this issue is discussed in detail in Section 7.6.5 of this
7 document. Two important factors, i.e., temporal trends and meteorological factors must be
8 considered in evaluating O3 health effects estimates. In the U.S. 95 communities study (Bell
9 et al., 2004), sensitivity analyses indicated that the O3 risk estimates were robust to tripling the
10 degrees of freedom for smoothing terms used to control for temporal trends. In a case-crossover
11 study by Schwartz (2004), the O3-mortality risk estimates from an analysis using nonlinear
12 regression splines to control for temperature were similar to those from an analysis that matched
13 on temperature, indicating that the effect estimates were not sensitive to methods used to control
14 confounding by temperature.
15 Analysis of O3 health effects is further complicated in view of the fact that the relationship
16 of O3 with temperature and with other pollutants appears to change across seasons. As shown in
17 Figures 8-4, the O3 effect estimates from warm season data were consistently larger compared to
18 those calculated using all-year data and cool-season data. In a study of daily hospital admissions
19 (Burnett et al., 2001), season-stratified analyses appeared to effectively control confounding by
20 season.
21 In summary, adjusting for temporal trends and meteorological factors is critical to
22 obtaining meaningful O3-effect estimates. Seasonal analyses indicate that mortality and
23 morbidity data computed using year-round data need to be interpreted with caution.
24 Air pollution epidemiological studies that integrate sensitivity analyses for seasonal
25 stratification, meteorological factors, and multipollutant models may provide a better and
26 more comprehensive understanding of the health effects estimates.
27
28 8.4.7.3 Assessment of Confounding by Copollutants
29 The presence and influence of PM and other gaseous copollutants have to be considered in
30 assessing O3-health effects associations found by observational studies. The potential for
31 copollutant confounding in the epidemiological time-series studies was assessed in some detail
January 2005 8-22 DRAFT-DO NOT QUOTE OR CITE
-------
1 in Section 7.6.6. Multipollutant modeling is the most common method used to test for potential
2 confounding in epidemiological studies; however, interpretation of the results is often
3 complicated by the high degree of correlation among air pollutants. The O3 mortality risk
4 estimates from two-pollutant models adjusted for PM are presented in Figure 8-4 (U.S. and
5 Canadian studies only). In the two multicity studies analyzed here, the addition of PM10 did not
6 substantially change the risk estimates (Samet et al., 2000; Dominci et al., 2003; Schwart, 2004).
7 The O3-mortality effects in single-city studies also were robust after adjusting for PM10 indices,
8 both in all-year and season-stratified analyses data.
9 In summary, assessing the health effects attributable to O3 is very challenging, even with
10 well-designed studies. Definitive partitioning out of the individual pollutant-specific health
11 outcomes from among an ambient mixture of multiple components is very difficult due to the
12 dynamic nature of their interactions over time. However, the new limited time-series studies that
13 made an exhaustive survey using populations from multiple U.S. cities do provide substantial
14 epidemiological evidence indicating that associations for O3 with mortality and morbidity are
15 robust to confounding by copollutants.
16
17 8.4.7.4 Lag Period between Ozone Exposure and Health Response
18 The lag times between causes and effects depend on underlying biological mechanism
19 involved in the process as well as the hypotheses tested. Different lag periods are appropriate for
20 assessing different health outcomes. As discussed in Section 7.6.4, examining longer lag periods
21 may be needed to understand more fully the O3-related health outcomes. The most significant
22 associations between O3 concentrations and mortality and respiratory hospitalization were
23 observed with 0-day and 1-day lags. These associations generally diminished with increased lag
24 days. In the 95 U.S. communities studies (Bell et al., 2004), the mortality risk estimated over
25 multiple days (cumulative lag of 0 to 6 days) using distributed lag models indicated an effect of
26 O3 that was twice as large as the effect estimated using 1-day lags. It should be noted that when
27 there is a pattern of effects across lag periods, selecting the 1-day lag effect estimate is likely to
28 underestimate the overall effect size and does not fully capture the risk distributed over adjacent
29 days. Longer averaging periods may aid in characterizing cumulative O3-related effects over
30 several days; however, interpreting these results may not be straightforward.
31
January 2005 8-23 DRAFT-DO NOT QUOTE OR CITE
-------
1 8.4.7.5 Concentration-Response Functions and Threshold
2 Ozone concentration-response relationships have been explored in several studies with
3 various health outcomes, including mortality, hospitalizations, emergency department visits,
4 lung function, and respiratory symptoms. While some studies found no threshold for O3 health
5 effects, others have found that a low-level threshold may be present. Note that an absence of a
6 detectable threshold in population studies does not necessarily indicate an absence of individual
7 thresholds and, conversely, evidence of a threshold for individuals does not necessarily indicate
8 a population threshold due to variability in response among individuals in the population. With
9 the current evidence, no definitive conclusion can be made regarding the threshold issue;
10 however, the limited evidence suggests that the possible threshold level may be well below the
11 current O3 standard level. The distribution of potential thresholds, particularly around the
12 NAAQS value of 80 ppb for 8-h maximum O3, needs to be further investigated.
13
14 8.4.7.6 Summary and Conclusions for Epidemiology Findings
15 Discussions presented in the previous sections evaluated the merits of the epidemiological
16 studies to derive judgments about the potential causal relationship between O3 exposures and
17 health outcomes. These evaluations were carried out in the context of the criteria listed in
18 Section 8.2.1. Information with regard to one of the criteria, i.e., coherence and biological
19 plausibility, is discussed in the section following the next one, which undertakes to provide an
20 integrated analysis of the biological evidence from human and animal toxicology studies with
21 the epidemiological evidence.
22 The results from the new field/panel studies evaluated in this document provided additional
23 evidence for likely causal relationships being reflected by significant associations between acute
24 O3 exposure and (a) decrements in lung function, (b) respiratory symptoms, and (c) increased use
25 of asthma medication in children and, in some cases, adults. Similarly, significant positive
26 associations can be inferred between acute O3 exposure and respiratory morbidity indexed by
27 hospital admissions and emergency visits, especially based on season-stratified data. The results
28 from large multicity studies suggest an elevated risk of mortality for acute exposure to O3;
29 however, the magnitudes of these estimates are small. Analysis of the data from chronic
30 mortality and morbidity studies indicate some significant associations between O3 and seasonal
January 2005 8-24 DRAFT-DO NOT QUOTE OR CITE
-------
1 changes in lung function; but, overall, the strength of the data does not allow establishment of a
2 conclusive relationship to O3 as a causal factor for other observed health outcomes.
3 Issues regarding strengths of models used in air pollution epidemiology were carefully
4 considered. There have been improvements in the modeling to adjust for potential confounding
5 variables, including temporal trends, meteorological factors, and copollutants. However, more
6 sensitivity analyses would still be useful to examine the extent of adequate adjustment for
7 confounding by these factors. Results from multipollutant models indicate that copollutants,
8 e.g., PM, generally do not confound the association between O3 and acute health outcomes,
9 suggesting an independent effect of O3.
10 In conclusion, the epidemiological evidence continues to support likely causal associations
11 between O3 and acute respiratory morbidity and mortality, based on the assessment of strength,
12 robustness, and consistency of results reported from numerous studies reviewed in Chapter 7.
13 Substantial evidence is lacking, however, by which to convincingly establish a positive
14 association between chronic O3 exposure and respiratory morbidity and mortality. Additional
15 investigations are needed to further understand the health effects resulting from long-term O3
16 exposure.
17 The positive associations for increased morbidity and mortality risk estimates during
18 warmer seasons (when O3 concentrations tend to be high) support a causal role for O3 in
19 affecting human health. Though seasonality in O3-related health effects was observed in both
20 time-series and longitudinal cohort studies, no clear evidence of a threshold for O3 effects has yet
21 been found.
22
23 8.4.8 Integration of Experimental and Epidemiologic Evidence
24 In this section, effects are made to integrate the epidemiological evidence discussed above
25 with results of human and animal experimental studies carried out in vivo and in vitro, to
26 understand O3-induced alterations at the physiological, pathological, and biochemical levels of
27 importance for the assessment of human health effects due to ambient O3 exposure. Also, the
28 influence of O3-induced changes at cellular and molecular levels are integrated to elucidate
29 scientific bases for the observed physiological and pathological alterations. These research
30 reports will be evaluated to assess (1) the scientific merit pertaining to the biological plausibility
31 of the health outcome associations observed in the epidemiological studies and (2) the coherence
January 2005 8-25 DRAFT-DO NOT QUOTE OR CITE
-------
1 of the overall body of evidence relevant to O3-related health outcomes supporting conclusions
2 regarding the attribution of observed effects of ambient O3 exposure.
3 The 1996 O3 AQCD, based on the limited number of controlled human exposure studies
4 and the animal toxicology data available to that date, arrived at the following conclusions
5 regarding potential health effects of ambient O3 exposure:
6 • Human studies have shown decreases in pulmonary function responsiveness to O3
exposure as a function of increasing age, although symptom rates remain similar across
age groups.
7 • Toxicological studies are not easily interpreted, but tend to suggest that young animals are
not more responsive to O3 than adults.
8 • Available toxicological and human data have not conclusively demonstrated that males
and females respond differently to O3. If gender differences exist for lung function
responsiveness to O3, they are not based on differences in baseline pulmonary function.
9 • Data are not adequate to determine whether any ethnic or racial group has a different
distribution of responsiveness to O3. In particular, the responses of nonwhite asthmatics
have not been investigated.
10 • Information derived from O3 exposure of smokers is limited. The general trend is that
smokers are less responsive than nonsmokers, but this reduced responsiveness may wane
after cessation of smoking.
11 • Nutritional status (e.g., vitamin E deficiency) makes laboratory rats more susceptible to
O3-induced effects, but it is not clear if vitamin E supplementation has an effect in human
populations. Such supplementation has no or minimal effects in animals. The role of such
antioxidant vitamins, especially their deficiency, in O3 responsiveness has not been well
studied.
12 As mentioned above, many questions remained unanswered, and the 1996 O3 AQCD
13 neither identified clear biological bases nor provided convincing experimental evidence
14 supporting the biological plausibility of reported O3 effects and/or the mechanisms of action
15 underlying potential O3 toxicity.
16 The available new epidemiological research reports on the health effects of O3 and
17 controlled human exposure studies using novel and refined models indicate certain positive
18 O3-health effect associations. This section focuses on interpreting the overall meaning of the
19 epidemiological findings and evaluates their bearing in the context of obtaining evidence for the
20 biological plausibility and possible mechanism(s) of action. This section also addresses the
21 complexities involved in extrapolating the extent of coherence observed from epidemiological
January 2005 8-26 DRAFT-DO NOT QUOTE OR CITE
-------
1 studies to specific health outcomes and related toxicological and biochemical mechanisms for
2 observed pathological and physiological changes in controlled human, animal, and in vitro
3 exposure studies. Experimental in vitro and in vivo studies using novel molecular technologies
4 that predict new hypotheses regarding the mechanisms of action are also included and discussed
5 as appropriate.
6 Several criteria listed in Section 8.2.1 are used in evaluating the available scientific support
7 for conclusions regarding potential causal relationships between O3 exposure and specific types
8 of health outcomes. In addition to those criteria addressed in the preceding discussion of
9 epidemiological evidence, certain other critical evaluation measures must be considered to
10 ensure that these observations are biologically relevant and consistent with experimentally
11 demonstrated biological mechanisms of action. For this assessment, the ensuing discussion on
12 biological plausibility and coherence considers (a) the extent to which available epidemiological
13 evidence shows associations with a range of logically linked health endpoints and (b) whether
14 available toxicological and biochemical evidence provides support for the observed
15 epidemiological associations reflecting causal relationships.
16
17 8.4.8.1 Background on Cross-Cutting Issues
18 Discussion of several cross-cutting issues that will facilitate a clear understanding of the
19 ensuing assessment is provided here to enhance an integrated and comprehensive understanding
20 of the experimental and epidemiological studies on O3 health effects. An important issue to
21 be considered is the extrapolation of observed effects from the perspective of dosimetry and
22 animal-to-human extrapolation models used and their strength. The most challenging issue is
23 the interpretation of the epidemiological observations in light of physiological and toxicological
24 endpoints derived from the experimental studies, wherein the data derived for O3-specific effects
25 have to be used to interpret and evaluate possible causative roles for O3 in contributing to health
26 outcomes observed in the air pollution epidemiology studies.
27
28 8.4.8.2 Approaches to Experimental Evaluation of Ozone Health Effects
29 Three chapters in the current document provide detailed discussion of various experimental
30 approaches utilized to evaluate O3-related health effects. Chapter 4 discusses dosimetry issues in
31 both animal and human exposure scenarios. Chapter 5 discusses the experimental studies of
January 2005 8-27 DRAFT-DO NOT QUOTE OR CITE
-------
1 physiological, biochemical (cellular and molecular changes) and pathological observations in
2 laboratory animals (including nonhuman primates, dogs, and rodent species) and in vitro studies
3 using cell culture systems (in certain cases, on humans cells recovered from BALF postexposure
4 to O3). Chapter 6 evaluated controlled human experimental studies that investigated various
5 physiological and biochemical endpoints. Many of the experimental animal toxicology studies
6 have been carried out using relatively high O3 exposure concentrations/doses that do not reflect
7 "real-world" exposure scenarios. These approaches have been used mainly to test hypotheses to
8 understand potential mechanism(s) of action implicated with the health outcomes identified in
9 epidemiological studies. In interpreting the results from the experimental approaches, one must
10 consider the following three issues: (1) controlled animal exposures studies use high
11 concentration to elicit biochemical/ physiological changes in healthy animals; (2) the roles of
12 other confounding pollutants that commonly occur with ambient exposures cannot be fully
13 reflected in the controlled exposure studies, and (3) the differences between human and rodents
14 with regard to O3 inhalability, deposition, clearance, and retention profiles (see Chapter 4 and 5
15 for details). Note that most of the in vitro toxicological studies were aimed at hypothesis
16 generation to predict mechanism(s) of action based on cellular and molecular endpoints and,
17 therefore, also generally used high O3 concentrations. The following discussion attempts to
18 integrate the experimental and epidemiological evidence to develop a holistic understanding of
19 O3 health effects in keeping with the points discussed above.
20
21 8.4.8.3 Interspecies Comparison of Experimental Effects — Dosimetry Considerations
22 As discussed in the previous section, the most important factor to consider while
23 attempting to integrate experimental studies across the species is the exposure-dose relationship.
24 Animal studies, particularly rats, have been valuable in developing mathematical dosimetry
25 models and useful for dosimetric extrapolation to humans in the strict sense of dose-response
26 basis. For example, dose-dependent increases in breathing frequency and decrease in tidal
27 volume were observed in both animals and humans with little effect on uptake. The O3-uptake
28 efficiency data from human and animals were consistent without much bearing on the mode
29 of breathing.
30
January 2005 8-28 DRAFT-DO NOT QUOTE OR CITE
-------
1 Experimental O3 dosimetry studies include response to bolus dose and general uptake.
2 Although the former approach is of limited relevance to environmental exposures, it has
3 indicated that prior exposure to O3 limits uptake of bolus dose. New uptake studies (Ultman
4 et al., 2004) carried out in controlled human clinical studies observed gender-specific differences
5 in the uptake of O3, but these differences do not correlate well with spirometric responses. Bolus
6 dose studies demonstrated that the uptake and regional respiratory tract distribution of O3 is
7 sensitive to the mode of breathing (nasal or oral) and to air flow rate. The change in breathing
8 due to exercise can cause a shift in distribution, allowing deeper penetration with resultant
9 damage to bronchiolar and alveolar tissue. This observation of an inverse relationship between
10 uptake and air flow is in agreement with animal studies. Additional uptake studies carried out in
11 humans using environmentally relevant O3 concentrations demonstrated the significance of
12 incorporating intersubject variability in dose-response relationship predictions and extrapolation.
13 The high degree of consistency observed in O3 uptake in animal and human experimental
14 exposure studies provided increased confidence in the use of theoretical dosimetry modeling
15 (see Chapter 4 for detailed discussion). The early models computed dose-response relationships
16 based on the assumption of O3 as the only active toxicant responsible for the observed
17 respiratory injury. Newer models have taken into consideration various factors such as age, as
18 well as anatomical, physiological, and biochemical alterations. The identification of conducting
19 airways as the primary site of acute cell injury, the site of O3 reaction/diffusion in the epithelial
20 lining fluid, the roles of intermediate reactive oxygen species (ROS) and lipid-ozonation
21 products in oxidative injury, and the roles of metabolic enzyme profiles in developing lung
22 tissue, when incorporated, will lead to refined novel models. The PBPK models were developed
23 using some of the refinements listed above and indicated the difference in dose metrics between
24 adult and infant with no difference projected after the age of five.
25
26 8.4.8.4 Integrated Critical Analysis of Physiological, Biochemical Effects
27 In the following subsections, research results generated from experimental studies on
28 humans and animals during the past decade are assessed (keeping in view the interspecies
29 differences discussed in the preceding section) in evaluating experimental evidence for
30 epidemiological studies to lead to the discussion of the biological plausibility and coherence for
31 O3 health effects in the later sections.
January 2005 8-29 DRAFT-DO NOT QUOTE OR CITE
-------
1 8.4.8.4.1 Pulmonary Function
2 The 1996 O3 AQCD reported decreases in FVC, FEVl3 decreased tidal volume, increased
3 breathing frequency, and increased resistance on short-term exposure to O3, based on research
4 reviewed from epidemiological, controlled human exposure, and animal studies. Inhalation of
5 O3 for several hours while physically active elicits both subjective respiratory tract symptoms
6 and acute pathophysiologic changes. The typical symptomatic response consistently reported is
7 that of tracheobronchial airway irritation. This is accompanied by decrements in lung capacities
8 and volumes, bronchoconstriction, airway hyperresponsiveness, airway inflammation, immune
9 system activation, and epithelial injury. The severity of symptoms and the magnitude of
10 response depend on inhaled dose, individual O3 sensitivity, and the extent of tolerance resulting
11 from previous exposures.
12 The development of effects is time-dependent during both exposure and recovery periods,
13 with considerable overlap of evolving and receding effects. In healthy human subjects exposed
14 to typical ambient concentrations (i.e., < 0.2 ppm O3), spirometric responses largely resolve
15 within a few hours (4 to 6 h) postexposure, however cellular effects persist for longer periods
16 (-24 h). Persisting small residual lung function effects are almost completely resolved within
17 24 hours. In hyperresponsive individuals, the recovery takes longer, as much as 48 h, to return
18 to baseline values.
19 It has also been observed that there is a large amount of intersubject variability and,
20 furthermore, the majority of these symptoms are attenuated after repeated exposure, but such
21 tolerance to O3 is lost within a week postexposure. Recent controlled human exposure studies
22 on prolonged O3 exposure using healthy subjects (reviewed in Chapter 6) found statistically
23 significant changes in pulmonary function during or after exposure to O3 concentrations
24 > 0.08 ppm.
25 New controlled human exposure studies (reviewed in Chapter 6) clearly indicate that FEVj
26 decrements and symptom responses decrease with age beyond young adulthood (18 to 20 years).
27 Hazucha et al. (2003) also examined gender differences along with age in O3 responsiveness and
28 observed that young females lose O3 sensitivity faster than young males, but that the rate is about
29 the same for both genders by middle age.
30 Human studies consistently report that inhalation of O3 alters the breathing pattern without
31 significantly affecting minute ventilation. A progressive decrease in tidal volume and a
January 2005 8-30 DRAFT-DO NOT QUOTE OR CITE
-------
1 "compensatory" increase in frequency of breathing to maintain steady minute ventilation during
2 exposure suggests a direct modulation of ventilatory control. These changes parallel a response
3 of many animal species exposed to O3 and other lower airway irritants (Tepper et al., 1990).
4 Earlier human studies (Coleridge et al., 1993; Hazucha and Sant'Ambrogio, 1993) reported a
5 role for bronchial C-fibers and rapidly adapting receptors are the primary vagal afferents
6 responsible for O3-induced changes in ventilatory rate and depth. A new study by Passannante
7 et al. (1998) observed that the primary mechanism of O3-induced reduction in inspiratory lung
8 function is an inhibition of inspiration elicited by stimulation of the C-fibers and suggest a role
9 for nociceptive mechanisms in modulating O3-induced inhibition of inspiration. This neurogenic
10 mechanism also has an effect on airway responsiveness and lung inflammation.
11 Lung function changes evaluated in patients with preexisting respiratory diseases, under
12 controlled experimental exposure regimens with or without physical exertion in the form of
13 intermittent exercise, indicated minimal O3-induced effects in COPD patients. However, newer
14 studies (see Chapter 6) indicate that pulmonary function deficiencies detected by spirometric
15 analyses in asthmatics augment the observations made in the 1996 O3 AQCD. More specifically,
16 Gong et al. (1997a) exposed nine COPD patients (0.24 ppm O3 for 4 h with intermittent exercise)
17 and observed a nonsignificant FEVj decrement of - 8% in COPD patients, which was not
18 statistically different from the decrement of -3% in healthy subjects. In contract, studies of
19 between 4- and 8-h duration with O3 concentrations of < 0.2 ppm, suggest a tendency for
20 increased O3-induced pulmonary function responses in asthmatics relative to healthy subjects
21 (Scannell et al., 1996). Similarly, Alexis et al. (2000) observed statistically significant
22 O3-induced decreases in FEVj in mild atopic asthmatics compared to healthy controls. Though
23 controlled human exposure studies may not provide the required statistical power (due to the
24 limited number of subjects compared to panel or field studies), they do suggest that asthmatics
25 are at least, if not more, sensitive than healthy subjects.
26
27 8.4.8.4.2 Airway Responsiveness
28 Increased responsiveness of the pulmonary airways, or "airway hyperresponsiveness"
29 (AHR), is usually analyzed in response to a bronchoconstrictor challenge. An extensive animal
30 studies database (using rats, mice, guinea pigs, and rabbits) exploring the effects of acute, long-
31 term, and repeated exposures to O3, indicates that induction of AHR occurs at high O3
January 2005 8-31 DRAFT-DO NOT QUOTE OR CITE
-------
1 concentrations. These studies provide clues to the roles of physiological and biochemical
2 components involved in this process. Experimental human exposure studies also reported that
3 acute O3-induced AHR, independent of pulmonary function changes or lung inflammation,
4 resolves to normalcy within 24 h. Airway hyperresponsiveness, as well as O3-induced airway-
5 antigen reactivity, were observed in asthmatics in several controlled exposure studies and found
6 to persist several hours postexposure (Chapter 6). Gong et al. (1997b) found that subjects with
7 asthma developed tolerance to repeated O3 exposures in a manner similar to normal subjects;
8 however, subjects with asthma had more persistent effects of O3 on airway responsiveness,
9 which was only partially attenuated when compared to filtered-air (FA) control subjects. These
10 observations suggest that O3 may act as a cofactor in response to airborne allergens or other
11 bronchoconstrictor agents in people with allergic asthma. Ozone-mediated modulation of airway
12 responsiveness may be a plausible link between ambient O3-exposure- related increased use of
13 asthma medication and the increased hospital admissions and emergency department visits
14 observed in epidemiological studies. Biochemical alterations observed in humans and animals
15 with exposure to O3 and discussed in the following sections may provide additional insights into
16 their roles in mechanistic aspects underlying the observed AHR.
17
18 8.2.8.4.3 Morphological and Biochemical Abnormalities
19 Most of the research results alluded to the ensuing discussion come from toxicology
20 studies using various laboratory animal species that were usually exposed to higher, non-ambient
21 concentrations of O3. However, these exploratory and mechanistic studies may provide
22 important and useful hypotheses to consider in integrating various health outcomes observed or
23 predicted by epidemiological studies. Controlled human exposure studies evaluated a few
24 cellular and biochemical parameters, mostly from BALF analyses. These studies have yielded
25 some limited evidence supporting the observations made in animal toxicology studies. Again,
26 caution should be exercised in extrapolating these observations to humans, due to species-
27 specific differences, as outlined earlier (see Section 8.2.7.3).
28
29 Lung Injury and Morphological Changes
30 Most animal species tested, including primates, exhibit similar morphological alterations
31 dependent on the exposure dose and the regional specificity. Differences in the distribution of
January 2005 8-32 DRAFT-DO NOT QUOTE OR CITE
-------
1 antioxidants in the centriacinar region (CAR) of the lung were responsible for the differences in
2 injury and morphological changes observed between nonhuman primates and rodents. Cells in
3 the CAR are the primary targets of O3, but ciliated cells in the nasal cavity, airways, and Type I
4 epithelial cells in the gas-exchange region are also targets. Though acute O3 exposure induces
5 structural changes such as fibrosis in CAR, these structural alterations appear to be transient with
6 recovery shortly postexposure, but the time for recovery is dependent on species and the dose of
7 O3. The remodeling of lung tissue indicated in a simulated seasonal-exposure scenario in
8 primates suggests the development of possible stable structural alterations as compared to
9 continuous-exposure scenarios. In an autopsy pathologic study, a significantly greater extent
10 and severity of centriacinar region alterations were observed in lungs of Los Angeles residents
11 than Miami residents, independent of a smoking effect (Sherwin et al., 2000). The results
12 suggest that the greater extent and severity of centriacinar region alterations may be related to
13 the higher O3 levels in Los Angeles. Similar observations of CAR thickening and deposition of
14 collagen in the rat suggests that long-term O3 exposure may cause a progressive structural lung
15 injury that can evolve into a more chronic form, such as fibrosis. Ozone-induced mucous
16 membrane cell metaplasia observed in rodents appears to be mediated by inflammation. Again,
17 one must be cautious in extrapolating these observations in animals to humans, given the
18 exposure regimens and doses used.
19
20 Lung Inflammation and Permeability
21 Ozone has long been recognized to cause lung inflammation and increased permeability in
22 the rat lung. These distinct, independent biological events have been observed in all species
23 studied, including humans, in response to acute exposure to O3. Increased lung inflammation
24 and permeability have been observed at levels as low as 0.12 ppm exposure for 6 h in rats and
25 24 h in mice. Both the inflammatory response and increased lung permeability have been
26 observed as early as 1 h and found to persist for at least 18 h in humans on exposure to O3 at
27 0.2 to 0.6 ppm. Subchronic exposures in animals suggest that permeability changes are transient
28 (and species-dependent) and return to control levels even with continuing exposure. Repeated
29 exposures in humans also indicate ongoing cellular damage irrespective of attenuation of the
30 inflammatory responses and lung function. Several studies have analyzed bronchioalveolar
January 2005 8-33 DRAFT-DO NOT QUOTE OR CITE
-------
1 lavage (BAL) and nasal lavage (NL) fluid and cells from O3-exposed humans for markers of
2 inflammation and lung damage (see Tables AX6-12 and AX6-13 in the Annex to Chapter 6).
3 The presence of neutrophils (PMNs) in the lung has long been accepted as a hallmark of
4 inflammation and is an important indicator that O3 causes inflammation in the lungs. It is
5 apparent, however, that inflammation within airway tissues may persist beyond the point that
6 inflammatory cells are found in BAL fluid. Soluble mediators of inflammation such as the
7 cytokines IL-6 and IL-8 as well as arachidonic acid metabolites (e.g., PGE2, PGF2a,
8 thromboxane, and leukotrienes [LTs] such as LTB4) have been measured in the BAL fluid of
9 humans exposed to O3. In addition to their role in inflammation, many of these compounds have
10 bronchoconstrictive properties and may be involved in increased airway responsiveness
11 following O3 exposure. Inflammation and cellular responses associated with acute O3 exposure
12 were also attenuated after 5 consecutive days of O3 exposure (compared to historical data for
13 responses after a single-day exposure). Even though indicators of epithelial cell damage were
14 not seen immediately after acute exposure, they were present in BALF after the fifth day of
15 exposure. When reexposed 2 weeks later, changes in BALF indicated that epithelial cells
16 appeared to be fully repaired (Devlin et al., 1997). Similar adaptive response was also observed
17 in an epidemiological study by Kopp et al. (1999). The analysis of BALF in human subjects
18 after first O3 peak in summer indicated increased levels of protein and leukocytes and no such
19 increase was observed later in summer even after exposure to higher levels of O3.
20 Interaction of O3 with the constituents of the extracellular lining fluid and the induction of
21 oxidative stress is implicated in injury and inflammation. Animal toxicological and a few
22 in vitro studies analyzed cells recovered from BALF for many biochemical mediators implicated
23 in injury and inflammation and found alterations in the expression of cytokines, chemokines, and
24 adhesion molecules, indicative of an active stress response as well as injury repair and
25 regeneration processes. Both animal and human studies indicate cellular and biochemical
26 changes associated with inflammation and increased permeability, but the relationship between
27 these changes and their role in lung function and airway responses is not known.
28
29 Host Defense
30 A number of closely integrated defense mechanisms exist that offer protection to
31 respiratory tract cells from the adverse effects of inhaled pollutants and microbes. Acute O3
January 2005 8-34 DRAFT-DO NOT QUOTE OR CITE
-------
1 exposure had been found to impair host defense capabilities both in humans and animals by
2 depressing alveolar macrophage functions and by decreasing the mucocilliary clearance of
3 inhaled particles and microbes. Interference by O3 exposure with the clearance process has been
4 found to be dose-dependent, whereby low doses accelerate clearance, but high doses slow
5 clearance. Some respiratory tract regional- and species-specific differences have also been
6 observed. Though acute O3 exposures were found to suppress alveolar phagocytosis and
7 immune functions, these alterations appeared to be transient and were attenuated with continuous
8 or repeated exposures. Continuous exposures to O3 impairs immune responses, followed by
9 adaptation and recovery to normalcy. Studies using various genetically sensitive or susceptible
10 strains indicated a possible interaction between innate and acquired immune system components,
11 possibly through Toll-like receptor-4 and downstream pathways; but these studies were carried
12 out using high O3 doses that may not be relevant to ambient exposures.
13
14 Biochemical A Iterations
15 An extensive experimental database, including the research presented in 1996 O3 AQCD,
16 suggests that potential biochemical alterations in various intermediary metabolic pathways are
17 involved in lung injury, inflammation, and functional alterations. Recent experimental evidence
18 still points to the importance of initial interactions of O3 with the lipid constituents of the ELF
19 and to the generation of ozonation products and secondary redox mediators in the initiation of
20 site-specific cell-injury response cascades. One such ozonation product, 4-hydroxynonenal, has
21 been found to bind to proteins and increase protein adducts in human alveolar macrophages,
22 suggesting a possible role in acute cell toxicity. Species- and region-specific increases in lung
23 xenobiotic metabolism has been observed in response to both short- and long-term O3 exposure.
24 Antioxidants in the ELF react with O3 and confer protection from toxicity, and even with
25 environmentally relevant exposures, the reactivity of O3 was not quenched. Species-specific and
26 age-dependent changes in the antioxidant metabolism add another dimension to their role in this
27 process. Carefully controlled studies of dietary antioxidant supplementation (Samet et al., 2001;
28 Trenga et al., 2001) have found some protective effects of a-tocopherol and ascorbate for
29 O3-induced spirometric lung function decrements but not for the intensity of subjective
30 symptoms and inflammatory responses (including cell recruitment, activation, and a release of
31 mediators). Dietary antioxidants have also afforded partial protection to asthmatics by
January 2005 8-35 DRAFT-DO NOT QUOTE OR CITE
-------
1 attenuating postexposure bronchial hyperresponsiveness (Trenga et al., 2001). Also, two
2 epidemiologic studies of street workers and asthmatic children in Mexico City found that
3 subjects taking antioxidant supplements containing vitamins E and C were protected from
4 O3-induced changes in lung function (Romieu et al., 1998, 2002).
5 Based on the above, it is evident that the extensive experimental database accumulated
6 from animal toxicology studies (including nonhuman primate studies) and limited controlled
7 human exposure studies, has provided insights into various biochemical, cellular, and molecular
8 alterations in lung tissue exposed to O3. The majority of these studies used acute exposure
9 regimens and high concentrations, and provide hypotheses regarding potential molecular
10 mechanisms implicated in O3 toxicity. Utilizing this information in relevant rodent-to-human
11 extrapolation models with appropriate species-specific adjustments may well provide useful
12 information on initial biochemical alterations that may aid in the development of suitable
13 biomarkers for O3 exposures/effects.
14
15 Systemic Effects
16 A number of rodent toxicology studies that investigated the effects of acute O3 exposure on
17 extrapulmonary systems have reported neurobehavioral, neuroendocrine, developmental, and
18 skin effects, albeit typically at much higher than ambient O3 concentrations. Ultrastructural,
19 cellular, and biochemical parameters evaluated in these studies indicate a role for O3 in
20 mediating biochemical and functional alterations through interactions with the redox systems.
21 An increasing body of animal toxicology evidence suggests that thermoregulatory and
22 hematological alterations (in heart rate variability and/or core body temperature) may mediate
23 acute cardiovascular effects. Limited human exposure studies have also explored O3-induced
24 cardiovascular effects, but did not observe acute cardiovascular effects in normal and
25 hypertensive subjects.
26
27 Susceptibility Factors
28 Many factors such as age, gender, disease, nutritional status, smoking, and genetic
29 variability may contribute to the differential effects of environmental pollutants, including O3.
30 Genetic factors, such as single nucleotide polymorphisms (SNPs) and developmental defects,
31 can contribute to innate susceptibility, while acquired susceptibility may develop due to personal
January 2005 8-36 DRAFT-DO NOT QUOTE OR CITE
-------
1 habits (smoking, diet, exercise) and other risk factors such as age, gender, pregnancy, and
2 copollutants. However, the available information from animal toxicologic and epidemiologic
3 studies does provide clear scientific evidence by which to identify and/or associate any specific
4 factor as contributing to adverse health effects of O3 (U.S. Environmental Protection Agency,
5 1996).
6 New animal toxicology studies using various strains of mice and rat have identified O3-
7 sensitive and resistant strains and illustrate the importance of genetic background in determining
8 O3 susceptibility. Biochemical and molecular parameters extensively evaluated in these
9 experiments were used to identify specific loci on the chromosomes and, in some cases, to relate
10 the differential expression of specific genes to biochemical and physiological differences
11 observed among these species. Utilizing O3-sensitive and O3-resistant species, it has been
12 possible to identify the involvement of AHR and inflammation processes in O3 susceptibility.
13 However, most of these studies were carried out using high doses of O3, making the relevance of
14 these studies questionable in human health effects assessment. No doubt, the molecular
15 parameters identified in these studies may serve as useful biomarkers with the availability of
16 suitable technologies and, ultimately, can likely be integrated with epidemiological studies.
17 Interindividual differences in O3 responsiveness have been observed across a spectrum of
18 symptoms and lung function responses but do no yet allow identification of important underlying
19 factors, except a significant role for age.
20
21 8.4.9 Preexisting Disease as a Potential Risk Factor
22 People with preexisting pulmonary disease may be at increased risk from O3 exposure.
23 Altered physiological, morphological and biochemical states typical of respiratory diseases like
24 asthma, COPD and chronic bronchitis may render people sensitive to additional oxidative burden
25 induced by O3 exposure. Based on studies assessed in the 1996 criteria document (U.S.
26 Environmental Protection Agency, 1996), asthmatics appear to be at least as, or more, sensitive
27 to acute effects of O3 as healthy nonasthmatic subjects. The new results reviewed in Chapters 6
28 and 7 from controlled exposure and epidemiological studies also suggest that asthmatics are a
29 potentially sensitive subpopulation for O3 health effects.
30 A number of time-series epidemiological studies have reported increased risk in study
31 subsets of individuals with preexisting lung diseases and tend to implicate asthmatics as
January 2005 8-37 DRAFT-DO NOT QUOTE OR CITE
-------
1 potentially susceptible individuals. The epidemiological studies of acute exposure to O3
2 discussed in Section 8.2.3 indicate increased risk for exacerbation of disease symptoms during
3 the warm season.
4 Newly available human exposure studies by Stenfors and coworkers have shown
5 differences regarding PMN influx in BALF between asthmatics and healthy human subjects.
6 In vitro studies (Stenfors et al., 2002) using nasal mucosal biopsies from atopic and nonatopic
7 subjects exposed to 0.1 ppm O3 found significant difference in the induced release of IL-4, IL-6,
8 IL-8, and TNF-a. A subsequent study by the same group (Schierhorn et al., 2002) found a
9 significant difference in the O3-induced release of the neuropeptides neurokinin A and substance
10 P from allergic patients, compared to nonallergic controls, suggesting increased activation of
11 sensory nerves by O3 in the allergic tissues. Another report from Bayram et al. (2002) using in
12 vitro culture of bronchial epithelial cells recovered from atopic and nonatopic asthmatics
13 indicated the existence of a significant difference in permeability by measuring the paracellular
14 flux of 14C-BSA. Additional controlled O3 exposure studies in human subjects with intermittent
15 asthma (Hiltermann et al., 1999), and asthmatics (Basha et al., 1994; Scannell et al., 1996)
16 reported increased secretion of IL-8 suggesting increased neutrophilic inflammation in those
17 subjects.
18 The observation of increased pathological symptoms in long-term animal exposure studies
19 in the absence of observable physiological changes also suggests that chronic exposure may
20 increase susceptibility to adverse health effects, but this needs to be validated via long-term
21 epidemiological studies.
22
23 8.4.10 Biological Plausibility and Coherence of Evidence for Adverse
24 Respiratory Health Effects
25 The research evidence discussed in the preceding section indicates that injury to lung tissue
26 is the initial step in mediating deleterious health effects of O3, and, in turn, activates a cascade of
27 events starting with inflammation, altered permeability of the epithelial barrier, impaired
28 clearance mechanisms (including host defense), and pulmonary structural alterations that can
29 potentially exacerbate a preexisting disease status. Although many or all of the above proposed
30 mechanisms are hypothesized to be implicated in O3 toxicity, scientific evidence is still lacking
31 for clearly establishing a role for one or a group of mechanistic pathways underlying O3 health
January 2005 8-38 DRAFT-DO NOT QUOTE OR CITE
-------
1 effects observed in epidemiologic studies. Most of these mechanisms of action were predicted
2 based on animal toxicology studies, with some support from human exposure studies.
3 In this section, the new scientific information reviewed on animal toxicology studies
4 (Chapter 5) and human exposure studies (Chapter 6) are used to evaluate plausible biological
5 bases for the health effects observed in epidemiological (Chapter 7) studies. The interpretations
6 provided in this section are, in the majority of situations, based on theoretical extrapolations,
7 while being mindful (based on our understanding in the post-genome era) of the existence of
8 common biochemical and molecular pathways that operate or function across different species.
9 In order to help interpret health endpoints (hospital admissions, mortality, and disease
10 exacerbations) purported to be associated epidemiologically with either acute or long-term
11 ambient exposure to O3, this section is organized into two subsections, based on the
12 physiological observations presented in the first section of (a) O3 effects on pulmonary function
13 as supported by cellular and molecular biological observations discussed in the second section,
14 and (b) O3-related lung injury, inflammation and host defense effects.
15 As exposure to O3 progresses, lung injury and inflammation begin to develop and initiate
16 cellular and subcellular changes. Airway hyperreactivity develops more slowly than pulmonary
17 function effects, while inflammation develops even more slowly and reaches its maximum 3 to
18 6 h postexposure. Cellular responses, such as release of inflammatory mediators or cytokines,
19 appear to be active as late as 20-h postexposure (Torres et al., 2000). Although the following
20 discussion is divided into two subsection, there may be cross-references between the sections to
21 better establish meaningful biological plausibility, as physiological and biochemical changes
22 overlap. Each subsection summarizes pertinent key information and then arrives at conclusions
23 as to the plausibility of effects attributable to O3 exposure.
24
25 8.4.10.1 Pulmonary Function
26 Ozone-induced critical respiratory functional deficiencies were monitored by measuring
27 changes in pulmonary function. Studies detailing alterations in pulmonary function discussed in
28 Chapters 5, 6, and 7 from animal, human toxicology, and epidemiology studies are summarized
29 in Sections 8.2.2, 8.2.3, and 8.2.7.4. Evaluation of pulmonary function on acute O3 exposure in
30 animals show a positive association with increased breathing frequency, decreased tidal volume
31 (rapid and shallow breathing), increased resistance, and altered breathing mechanics (compliance
January 2005 8-39 DRAFT-DO NOT QUOTE OR CITE
-------
1 and resistance). A similar increased breathing frequency observed in human subjects suggests
2 modulation of ventilatory control on acute O3 exposure. Direct or indirect stimulation of lung
3 receptors and bronchial C-fibers by O3 and/or its oxidative products have been implicated in
4 modulating breathing pattern changes. Acute O3-induced biochemical changes suggest potential
5 interactions of O3 with the extracellular lining fluid and the generation of lipid ozonation
6 products and reactive oxygen species, ultimately leading to lung injury and/or inflammation.
7 These reactive species cause inhibition or decrease in the maximal inspiration capacity by
8 neurogenic mechanisms acting via the C-fiber afferents. Recent work by Passannante et al.
9 (1998) implicates stimulation of nociceptive receptors on bronchial C-fibers as the primary
10 mechanism for O3-induced inhibition of inspiration. Alternately, inhibition in maximal
11 inspiration may also be mediated by mediators such as prostaglandin E2 released due to lung
12 epithelial injury. This hypothesis gains strength from the observation of the blocking of
13 spirometric response in human subjects who were pretreated with nonsteroidal anti-
14 inflammatory agents such as indomethacin and ibuprofen. While recovery from pulmonary
15 function decline and airway hyperreactivity had been found to be rapid (4 to 6 h) in moderately
16 responsive individuals, persistent small residual lung function effects were found to take more
17 than 48 h to return to baseline values in hyperresponsive individuals (Nightingale et al., 2000).
18 Such an extended recovery from lung function decline, airway hyperresponsiveness, and
19 increased O3-induced pulmonary function decline in asthmatics (Scannell et al., 1996) compared
20 to normal subjects may be responsible for the increased emergency room visits or hospital
21 admission and the increased use of asthma medication in asthmatics reported in recent time-
22 series epidemiological studies. The contribution of morphological alterations in the decline of
23 lung function on chronic exposure is not known.
24 Airway hyperresponsiveness (AHR) due to O3 exposure is another important factor
25 involved in the observed decline in pulmonary function. Intermittent airway obstruction and
26 increased airway responsiveness to physical or chemical stimuli is also the hallmark of asthma.
27 Asthma-related AHR includes both physiological and morphological components such as
28 inner-wall thickening and mucus secretion. Airway hyperresponsiveness in response to
29 chemical challenge is found to be predominant in children compared to adults and older children.
30 Several controlled human O3 exposure studies reported increased airway responsiveness at
31 baseline both in normal and asthmatic subjects. The mechanisms mediating AHR are not yet
January 2005 8-40 DRAFT-DO NOT QUOTE OR CITE
-------
1 fully understood, but it appears to be mediated by multiple pathways. Involvement of AHR in
2 pulmonary function decline appears to be mediated by neurogenic mechanisms, as pretreatment
3 of neonatal rats with capsaicin prevented O3-induced release of neuropetides, suggesting a role
4 for C-fibers in AHR. Significant reduction in the immunoreactivity for substance P in the
5 bronchial biopsies from human subjects 6 h postexposure and its negative correlation with FEVj
6 decrements suggests involvement of similar neurogenic mechanisms to be persistent in
7 O3-induced bronchoconstriction (Krishna et al., 1997). Studies carried out in human subjects
8 using cyclooxygenase inhibitors to block the influx of PMNs and the inhibition of neutrophilic
9 inflammation by probenecid in dogs (Freed et al., 1999) indicated that O3-induced inflammation
10 and AFIR occur as two independent events. Many of the inflammatory mediators that exhibit
11 bronchoconstrictive properties may also play a significant role in the persistent spirometry
12 changes observed on exposure to O3 (Blomberg et al., 1999). Either the inflammation or AFIR
13 may be the underlying biological mechanism responsible for the observed lung function decline
14 in children and male human subjects reported in epidemiological studies.
15 Repeated-exposure studies in monkeys with a house dust mite antigen-sensitization
16 regimen (Schelegle et al., 2003; Chen et al., 2003) associated lung function changes to the
17 adaptation of respiratory motor responses. Similarly, controlled human exposure studies using
18 asthmatic subjects exposed to house dust mite indicated an immediate increase in airway-antigen
19 reactivity that persisted longer (18 to 20 h) in asthmatics than in normal subjects ( Kehrl et al.,
20 1999). The enhanced decline in pulmonary function in subjects with allergic rhinitis (Torres
21 et al., 1996) suggests slow recovery from O3-induced pulmonary function declines. These
22 observations suggest that O3 exposure, therefore, may be a clinically important cofactor in the
23 response to airborne bronchoconstrictor substances in individuals with preexisting allergic
24 asthma. This phenomenon could plausibly contribute to increased symptom exacerbations and,
25 even increased consequent physician or ER visits and possible hospital admissions. However,
26 even a small decline in lung function and its persistence in sensitive populations such as
27 asthmatics will have substantial effects and can lead to increased frequencies in emergency room
28 hospital visits or in medication use.
29
January 2005 8-41 DRAFT-DO NOT QUOTE OR CITE
-------
1 8.4.10.2 Lung Injury, Inflammation, and Host Defense
2 An extensive biochemical database collected from animal toxicology studies using varied
3 species indicates that the interaction of O3 with the ELF in the lung leads to the generation of a
4 series of reactive radical species, including lipid ozonation products and hydroperoxy radicals,
5 which contribute to the increased injury and inflammatory response. Using in vitro and ex vivo
6 studies on isolated ELF, it has been increasingly recognized that ozonation of polyunsaturated
7 fatty acids (PUFA) and its distribution profiles in the respiratory tract plays a critical role in the
8 extent of site-specific injury (Postlewait et.al., 1998; Connor et al., 2004). Ozone-induced lung
9 injury was also found to cause persistent fibrotic changes with the accumulation of collagen and
10 the thickening of CAR postexposure, suggesting progressive structural changes in the lung
11 tissue. Bronchial mucosal biopsies after repeated O3 exposure over 5 days in human subjects
12 indicated that inflammation of the bronchial mucosa persisted after repeated O3 exposure,
13 despite attenuation of some inflammatory markers in BALF and attenuation of lung function
14 responses and symptoms. Along with this, the persistent, although small, decrease in baseline
15 FEVj observed by Torres et al. (2000) suggested a difference in timescales among the functional
16 responses to O3. Elevated protein levels remaining after repeated exposures confirm the findings
17 of others (Christian et al., 1998; Devlin et al., 1997) and suggest that cellular damage is ongoing
18 irrespective of the attenuation of cellular inflammatory responses. The results of chronic-
19 exposure studies in animal toxicology evaluating fibrotic changes are inconsistent. Simulated
20 seasonal-exposure studies in infant rhesus monkey (0.5 ppm O3) indicate possible injury-repair
21 processes as observed with chronic exposure studies described in the 1996 O3 AQCD (U.S.
22 Environmental Protection Agency, 1996).
23 Lung inflammation is a host response to injury, which in turn triggers various biochemical
24 and physiological responses such as epithelial permeability, PMN influx, and the release of
25 pro- and anti-inflammatory mediators. The inflammatory response is a transient phenomenon
26 and resolves entirely postexposure in all the species studied, including humans. Subchronic
27 exposure to O3 has been found to induce inflammation after a few days (depending upon species
28 studied and exposure dose), which resolves to a normal state, even with continuing exposure;
29 adaptation on repetitive exposures has also been observed. Biochemical and molecular analysis
30 of BALF from various animal species and human subjects exposed to O3 clearly suggest the
31 participation of various inflammatory mediators, which in turn can activate multiple cascades of
January 2005 8-42 DRAFT-DO NOT QUOTE OR CITE
-------
1 physiological events. The adaptive response to repetitive exposure to O3 indicates resolution of
2 many of these inflammatory markers to different extents, suggesting that persistent mild
3 inflammation may compromise the abilities of the lung tissue to handle various host defense
4 functions and allergic responses.
5 Ozone-induced dysfunction in the barrier function of lung epithelium and subsequent
6 permeability changes also can impact lung host defense functions. Altered mucocilliary and
7 alveolar clearance of inhaled particles or microorganisms observed on acute and subchronic
8 exposure to O3 in animal toxicology studies indicate compromise in this important host defense
9 function. Similar compensation observed in the in vitro studies using cells recovered in BALF
10 from normal and asthmatic human subjects (Newson et al., 2000; Bosson et al., 2003; Bayram
11 et al., 2002) on acute O3 exposure suggest an additional burden on the host defense functions.
12 The new information obtained in the past decade on the morphological, biochemical,
13 cellular, and molecular aspects of O3 toxicology have increased our understanding of the
14 intricate biochemical and molecular mechanisms involved in lung tissue pathology. Combining
15 basic toxicology approaches with the sophisticated molecular technologies and using various O3-
16 sensitive and -resistant animal strains in these investigations have provided additional knowledge
17 in understanding the possible biochemical bases for the adverse health effects. Newer studies
18 that examined various inflammatory parameters, such as PMN influx (Stenfors et al., 2002) and
19 molecular changes in the nasal or bronchial biopsies from atopic asthmatics and normal human
20 subjects, indicated significant differences. Nichols et al., (2001) observed a role for oxidative
21 stress in O3-induced inflammation and increased release of TNF-a from nasal epithelial cells.
22 Increased neurogenic involvement in the O3-induced inflammatory response was observed by
23 Schierhorn et al. (1999). Taken together, the new science gathered from an extensive animal
24 toxicology database and limited human controlled exposure studies on pulmonary function, lung
25 defense, and biochemistry provide evidence consistent and coherent with health outcomes
26 endpoints observed in human subjects in clinical or field studies. These biological and
27 toxicological observations will gain additional value and support with future research efforts,
28 particularly those using controlled human exposure studies including subjects with preexisting
29 disease. Such research inquiries will aid in reducing the data gaps and in the development of
30 better extrapolation models that can be used in interpreting the health endpoints monitored in
31 epidemiological studies.
January 2005 8-43 DRAFT-DO NOT QUOTE OR CITE
-------
1 8.4.11 Coherence Between Epidemiological and Experimental Evidence for
2 Respiratory Health Effects
3 Recent epidemiological studies (collected from various metropolitan cities in the United
4 States, Canada, and Europe) reported associations between short-term O3 exposure for various
5 indices such as respiratory-related mortality, hospital admissions, and emergency department
6 visits for respiratory diseases. Three U.S. multicity studies of 95 communities and 90 cities
7 reported positive associations between acute O3 exposure and daily mortality. Data from four
8 European multicity studies have also reported similar positive associations to daily mortality on
9 acute O3 exposure. Several epidemiological individual and multicity studies reported significant
10 O3 effects on hospital admission. Some of the epidemiological studies from Europe reported a
11 strong relationship between unscheduled hospitalizations for COPD and O3 exposure.
12 In addition, new evidence exists for O3 effects on lung function decline/respiratory symptoms
13 from controlled human exposure (exercising) studies and recent exercise panel (field) studies.
14 Recent time-series studies reported excess risk for emergency department visits, particularly in
15 summer seasons with relatively high O3 concentrations. Epidemiological studies also reported
16 positive associations between the onset of asthma and asthma prevalence due to long-term O3
17 exposure.
18 The respiratory health effects observed in epidemiological studies gain relevance from
19 controlled human exposure studies. The health responses observed in these studies were
20 indicative of deleterious health effects due to exposure to O3. The studies indicate an acute
21 O3-induced decline in lung function with inflammatory changes in the lung such as increased
22 levels of infiltrating PMNs, release of inflammatory mediators, lung injury, permeability
23 changes, and altered host defense mechanisms. These observations gain further support from
24 similar biochemical, morphological, and immunological changes in animal toxicology studies
25 that suggest perturbations in lung tissue physiology. This body of evidence provides coherent
26 links between the results of large multicity epidemiological studies reporting increases in
27 hospitalizations and unscheduled emergency department visits with toxicologic evidence of
28 acute O3-induced lung tissue injury and inflammation in humans and animals.
29
January 2005 8-44 DRAFT-DO NOT QUOTE OR CITE
-------
1 8.4.12 Summary and Conclusions for Ozone Health Effects
2 This section discussed the development of a coherent understanding of O3 health effects
3 through considering the plausibility and coherence of information derived from epidemiological
4 evidence and human and animal toxicology studies. As discussed in Sections 8.4.2 to 8.4.5,
5 considerations related to the epidemiologic evidence alone appear to support likely causality for
6 observed associations between O3 and health outcomes. This section further described evidence
7 from both epidemiologic and toxicologic studies for health effects that are logically linked
8 together.
9 Epidemiological studies have reported positive associations between O3 exposure and
10 health effects across a range of endpoints, from respiratory-related mortality, increased
11 respiratory-related hospital admissions, and emergency department visits, to more subtle effects
12 such as decrements in lung function, lung inflammation, airway responsiveness, and altered
13 mucocilliary clearance. The new toxicologic and physiological evidence suggest links to
14 potential molecular pathways that may provide reasonable explanations for the observed
15 epidemiological findings, but as described earlier, caveats must be considered in interpreting
16 these studies. The new toxicological information confirms the earlier findings and presents
17 important new evidence supporting the plausibility of associations between O3 and adverse
18 respiratory health effects. While many research questions remain, the convergence of
19 epidemiologic and toxicologic evidence related to respiratory health effects for ambient O3
20 exposure argues for coherence and plausibility for this body of evidence.
21 Controlled human exposure studies have provided new information indicating that age at
22 time of exposure is a major susceptibility factor for observed decrements in lung function.
23 Epidemiological studies and some preliminary supportive data from toxicological studies
24 suggest that asthmatics are a potential sensitive subpopulation although additional scientific
25 evidence is needed. However, little experimental evidence is available by which to judge the
26 plausibility of any chronic O3 exposure effects observed in epidemiologic studies. Thus, further
27 study is required on the potential toxicologic or pathologic mechanisms that may underly chronic
28 effects of ambient O3 exposure to relate to observed respiratory health effects in epidemiological
29 studies.
30 Analysis of the body of toxicologic studies suggests plausible mechanisms for
31 epidemiologic findings. The newly available epidemiological studies on positive associations
January 2005 8-45 DRAFT-DO NOT QUOTE OR CITE
-------
1 between acute exposure to O3 and a range of health outcomes support the general conclusion that
2 O3 is causally related to respiratory-related mortality and morbidity. A very limited database
3 (epidemiologic and toxicologic) is available on the long-term effects of O3 on respiratory-related
4 mortality and morbidity, but given our understanding of the plausible biological mechanisms
5 implicated in acute response and differential recovery, O3 may also be causally related to long-
6 term respiratory-related health risks. Additional scientific information to support the predicted
7 sensitivity of asthmatics reported in epidemiological studies is still needed. Substantial scientific
8 evidence gathered in the past decade provides additional support for the conclusions stated in the
9 1996 O3 AQCD with regard to health effects shown to be associated with ambient O3 exposure.
10 The recent epidemiological studies provide further strong evidence supporting potential
11 morbidity health risks associated with exposure to ambient O3. Furthermore, newly available
12 epidemiological data linking acute O3 exposure to respiratory-related mortality and to the
13 exacerbation of respiratory-related disease symptoms suggest even larger health impacts and
14 costs to society than previously demonstrated, warranting additional research.
15
January 2005 8-46 DRAFT-DO NOT QUOTE OR CITE
-------
1 REFERENCES
2 Alexis, N.; Urch, B.; Tarlo, S.; Corey, P.; Pengelly, D.; O'Byrne, P.; Silverman, F. (2000) Cyclooxygenase
3 metabolites play a different role in ozone-induced pulmonary function decline in asthmatics compared to
4 normals. Inhalation Toxicol. 12: 1205-1224.
5 Basha, M. A.; Gross, K. B.; Gwizdala, C. J.; Haidar, A. H.; Popovich, I, Jr. (1994) Bronchoalveolar lavage
6 neutrophilia in asthmatic and healthy volunteers after controlled exposure to ozone and filtered purified air.
7 Chest 106: 1757-1765.
8 Bayram, H.; Rusznak, C.; Khair, O. A.; Sapsford, R. J.; Abdelaziz, M. M. (2002) Effect of ozone and nitrogen
9 dioxide on the permeability of bronchial epithelial cell cultures of non-asthmatic and asthmatic subjects.
10 Clin. Exp. Allergy 32: 1285-1292.
11 Bell, M. L.; McDermott, A.; Zeger, S. L.; Samet, J. M.; Dominici, F. (2004) Ozone and short-term mortality in
12 95 US urban communities, 1987-2000. JAMA J. Am. Med. Assoc. 292: 2372-2378.
13 Blomberg, A.; Mudway, I. S.; Nordenhall, C.; Hedenstrom, H.; Kelly, F. J.; Frew, A. J.; Holgate, S. T.; Sandstrom,
14 T. (1999) Ozone-induced lung function decrements do not correlate with early airway inflammatory or
15 antioxidant responses. Eur. Respir. J. 13: 1418-1428.
16 Bosson, J.; Stenfors, N.; Bucht, A.; Helleday, R.; Pourazar, J.; Holgate, S. T.; Kelly, F. J.; Sandstrom, T.; Wilson, S.;
17 Frew, A. J.; Blomberg, A. (2003) Ozone-induced bronchial epithelial cytokine expression differs between
18 healthy and asthmatic subjects. Clin. Exp. Allergy 33: 777-782.
19 Burnett, R. T.; Brook, J. R.; Yung, W. T.; Dales, R. E.; Krewski, D. (1997) Association between ozone and
20 hospitalization for respiratory diseases in 16 Canadian cities. Environ. Res. 72: 24-31.
21 Burnett, R. T.; Cakmak, S.; Brook, J. R.; Krewski, D. (1997) The role of paniculate size and chemistry in the
22 association between summertime ambient air pollution and hospitalization for cardiorespiratory diseases.
23 Environ. Health Perspect. 105: 614-620.
24 Burnett, R. T.; Smith-Doiron, M.; Stieb, D.; Raizenne, M. E.; Brook, J. R.; Dales, R. E.; Leech, J. A.; Cakmak, S.;
25 Krewski, D. (2001) Association between ozone and hospitalization for acute respiratory diseases in children
26 less than 2 years of age. Am. J. Epidemiol. 153: 444-452.
27 Chen, C.-Y.; Bonham, A. C.; Plopper, C. G.; Joad, J. P. (2003) Plasticity in respiratory motor control: selected
28 contribution: neuroplasticity in nucleus tractus solitarius neurons following episodic ozone exposure in infant
29 primates. J. Appl. Physiol. 94: 819-827.
30 Chen, L.; Jennison, B. L.; Yang, W.; Omaye, S. T. (2000) Elementary school absenteeism and air pollution.
31 Inhalation Toxicol. 12: 997-1016.
32 Chock, D. P.; Winkler, S. L.; Chen, C. (2000) A study of the association between daily mortality and ambient air
33 pollutant concentrations in Pittsburgh, Pennsylvania. J. Air Waste Manage. Assoc. 50: 1481-1500.
34 Christian, D. L.; Chen, L. L.; Scannell, C. H.; Ferrando, R. E.; Welch, B. S.; Balmes, J. R. (1998) Ozone-induced
35 inflammation is attenuated with multiday exposure. Am. J. Respir. Crit. Care Med. 158: 532-537.
36 Coleridge, J. C. G.; Coleridge, H. M.; Schelegle, E. S.; Green, J. F. (1993) Acute inhalation of ozone stimulates
37 bronchial C-fibers and rapidly adapting receptors in dogs. J. Appl. Physiol. 74: 2345-2352.
38 Connor, L. M.; Ballinger, C. A.; Albrecht, T. B.; Postlethwait, E. M. (2004) Interfacial phospholipids inhibit ozone
39 reactive absorption-mediated cytotoxicity in vitro. Am. J. Physiol. 286: L1169-L1178.
40 Devlin, R. B.; Folinsbee, L. J.; Biscardi, F.; Hatch, G.; Becker, S.; Madden, M. C.; Robbins, M.; Koren, H. S. (1997)
41 Inflammation and cell damage induced by repeated exposure of humans to ozone. Inhalation Toxicol.
42 9:211-235.
43 Dockery, D. W.; Schwartz, J.; Spengler, J. D. (1992) Air pollution and daily mortality: associations with particulates
44 and acid aerosols. Environ. Res. 59: 362-373.
45 Dominici, F.; McDermott, A.; Daniels, M.; Zeger, S. L.; Samet, J. M. (2003) Mortality among residents of 90 cities.
46 In: Revised analyses of time-series studies of air pollution and health. Special report. Boston, MA: Health
47 Effects Institute; pp. 9-24. Available: http://www.healtheffects.org/Pubs/TimeSeries.pdf [12 May, 2004].
48 Fairley, D. (2003) Mortality and air pollution for Santa Clara County, California, 1989-1996. In: Revised analyses of
49 time-series studies of air pollution and health. Special report. Boston, MA: Health Effects Institute; pp.
50 97-106. Available: http://www.healtheffects.org/Pubs/TimeSeries.pdf [18 October, 2004].
51 Freed, A. N.; Cueto, R.; Pryor, W. A. (1999) Antioxidant transport modulates peripheral airway reactivity and
52 inflammation during ozone exposure. J. Appl. Physiol. 87: 1595-1603.
53 Friedman, M. S.; Powell, K. E.; Hutwagner, L.; Graham, L. M.; Teague, W. G. (2001) Impact of changes in
54 transportation and commuting behaviors during the 1996 summer Olympic games in Atlanta on air quality and
55 childhood asthma. JAMA J. Am. Med. Assoc. 285: 897-905.
January 2005 8-47 DRAFT-DO NOT QUOTE OR CITE
-------
1 Gamble, J. L. (1998) Effects of ambient air pollution on daily mortality: a time series analysis of Dallas, Texas,
2 1990-1994. Presented at: 91st annual meeting and exhibition of the Air & Waste Management Association;
3 June; San Diego, CA. Pittsburgh, PA: Air & Waste Management Association; paper no. 98-MP26.03.
4 Gent, J. F.; Triche, E. W.; Holford, T. R.; Belanger, K.; Bracken, M. B.; Beckett, W. S.; Leaderer, B. P. (2003)
5 Association of low-level ozone and fine particles with respiratory symptoms in children with asthma. JAMA
6 J. Am. Med. Assoc. 290: 1859-1867.
7 Gilliland, F. D.; Berhane, K.; Rappaport, E. B.; Thomas, D. C.; Avol, E.; Gauderman, W. J.; London, S. J.;
8 Margolis, H. G.; McConnell, R.; Islam, K. T.; Peters, J. M. (2001) The effects of ambient air pollution on
9 school absenteeism due to respiratory illnesses. Epidemiology 12: 43-54.
10 Gong, H., Jr.; McManus, M. S.; Linn, W. S. (1997) Attenuated response to repeated daily ozone exposures in
11 asthmatic subjects. Arch. Environ. Health 52: 34-41.
12 Gong, H., Jr.; Shamoo, D. A.; Anderson, K. R.; Linn, W. S. (1997) Responses of older men with and without
13 chronic obstructive pulmonary disease to prolonged ozone exposure. Arch. Environ. Health 52: 18-25.
14 Gryparis, A.; Forsberg, B.; Katsouyanni, K.; Analitis, A.; Touloumi, G.; Schwartz, J.; Samoli, E.; Medina, S.;
15 Anderson, H. R.; Niciu, E. M.; Wichmann, H.-E.; Kriz, B.; Kosnik, M.; Skorkovsky, J.; Vonk, J. M.;
16 Dortbudak, Z. (2004) Acute effects of ozone on mortality from the "air pollution and health: a European
17 approach" project. Am. J. Respir. Crit. Care Med. 170: 1080-1087.
18 Hazucha, M. J.; Folinsbee, L. J.; Bromberg, P. A. (2003) Distribution and reproducibility of spirometric response to
19 ozone by gender and age. J. Appl. Physiol. 95: 1917-1925.
20 Hazucha, M. J.; Sant'Ambrogio, G. (1993) Effects of ozone on the activity of slowly (SAR) and rapidly adapting
21 (RAR) receptors in cats. FASEB J. 7: 407A.
22 Hiltermann, J. T. N.; Lapperre, T. S.; Van Bree, L.; Steerenberg, P. A.; Brahim, J. J.; Sont, J. K.; Sterk, P. J.;
23 Hiemstra, P. S.; Stolk, J. (1999) Ozone-induced inflammation assessed in sputum and bronchial lavage fluid
24 from asthmatics: a new noninvasive tool in epidemiologic studies on air pollution and asthma. Free Radical
25 Biol. Med. 27: 1448-1454.
26 Ito, K. (2003) Associations of paniculate matter components with daily mortality and morbidity in Detroit,
27 Michigan. In: Revised analyses of time-series studies of air pollution and health. Special report. Boston, MA:
28 Health Effects Institute; pp. 143-156. Available: http://www.healtheffects.org/Pubs/TimeSeries.pdf
29 [12 May, 2004].
30 Ito, K.; Thurston, G. D. (1996) Daily PM10/mortality associations: an investigation of at-risk subpopulations.
31 J. Exposure Anal. Environ. Epidemiol. 6: 79-95.
32 Jaffe, D. H.; Singer, M. E.; Rimm, A. A. (2003) Air pollution and emergency department visits for asthma among
33 Ohio Medicaid recipients, 1991-1996. Environ. Res. 91: 21-28.
34 Jorres, R. A.; Holz, O.; Zachgo, W.; Timm, P.; Koschyk, S.; Mtiller, B.; Grimminger, F.; Seeger, W.; Kelly, F. J.;
35 Dunster, C.; Frischer, T.; Lubec, G.; Waschewski, M.; Niendorf, A.; Magnussen, H. (2000) The effect of
3 6 repeated ozone exposures on inflammatory markers in bronchoalveolar lavage fluid and mucosal biopsies.
37 Am. J. Respir. Crit. Care Med. 161: 1855-1861.
38 Jorres, R.; Nowak,D.; Magnussen, H.; Speckin, P.; Koschyk, S. (1996) The effect of ozone exposure on allergen
39 responsiveness in subjects with asthma or rhinitis. Am. J. Respir. Crit. Care Med. 153: 56-64.
40 Kehrl, H. R.; Peden, D. B.; Ball, B. A.; Folinsbee, L. J.; Horstman, D. H. (1999) Increased specific airway reactivity
41 of persons with mild allergic asthma after 7.6 hours of exposure to 0.16 ppm ozone. J. Allergy. Clin.
42 Immunol. 104: 1198-1204.
43 Kinney, P. L.; Ito, K.; Thurston, G. D. (1995) A sensitivity analysis of mortality/PM10 associations in Los Angeles.
44 In: Phalen, R. F.; Bates, D. V., eds. Proceedings of the colloquium on paniculate air pollution and human
45 mortality and morbidity; January 1994; Irvine, CA. Inhalation Toxicol. 7: 59-69.
46 Kinney, P. L.; Ozkaynak, H. (1991) Associations of daily mortality and air pollution in Los Angeles County.
47 Environ. Res. 54: 99-120.
48 Klemm, R. J.; Lipfert, F. W.; Wyzga, R. E.; Gust, C. (2004) Daily mortality and air pollution in Atlanta: two years
49 of data from ARIES. Inhalation Toxicol. 16(suppl. 1): 131-141.
50 Klemm, R. J.; Mason, R. M., Jr. (2000) Aerosol Research and Inhalation Epidemiological Study (ARIES): air
51 quality and daily mortality statistical modeling-interim results. J. Air. Waste Manage. Assoc. 50: 1433-1439.
52 Kopp, M. V.; Ulmer, C.; Ihorst, G.; Seydewitz, H. H.; Frischer, T.; Forster, J.; Kuehr, J. (1999) Upper airway
53 inflammation in children exposed to ambient ozone and potential signs of adaptation. Eur. Respir. J.
54 14:854-861.
55 Korrick, S. A.; Neas, L. M.; Dockery, D. W.; Gold, D. R.; Allen, G. A.; Hill, L. B.; Kimball, K. D.; Rosner, B. A.;
56 Speizer, F. E. (1998) Effects of ozone and other pollutants on the pulmonary function of adult hikers.
57 Environ. Health Perspect. 106: 93-99.
January 2005 8-48 DRAFT-DO NOT QUOTE OR CITE
-------
1 Krishna, M. T.; Springall, D.; Meng, Q.-H.; Withers, N.; Macleod, D.; Biscione, G.; Frew, A.; Polak, J.; Holgate, S.
2 (1997) Effects of ozone on epithelium and sensory nerves in the bronchial mucosa of healthy humans. Am. J.
3 Respir. Crit. Care Med. 156: 943-950.
4 Liao, D.; Duan, Y.; Whitsel, E. A.; Zheng, Z.-J.; Heiss, G.; Chinchilli, V. M.; Lin, H.-M. (2004) Association of
5 higher levels of ambient criteria pollutants with impaired cardiac autonomic control: a population-based
6 study. Am. J. Epidemiol. 159: 768-777.
7 Linn, W. S.; Szlachcic, Y.; Gong, H., Jr.; Kinney, P. L.; Berhane, K. T. (2000) Air pollution and daily hospital
8 admissions in metropolitan Los Angeles. Environ. Health Perspect. 108: 427-434.
9 Lipfert, F. W.; Morris, S. C.; Wyzga, R. E. (2000) Daily mortality in the Philadelphia metropolitan area and
10 size-classified paniculate matter. J. Air Waste Manage. Assoc. 50: 1501-1513.
11 Lippmann, M.; Ito, K.; Nadas, A.; Burnett, R. T. (2000) Association of paniculate matter components with daily
12 mortality and morbidity in urban populations. Cambridge, MA: Health Effects Institute; research report no.
13 95.
14 Luginaah, I. N.; Fung, K. Y.; Gorey, K. M.; Webster, G.; Wills, C. (2004) Association of Ambient Air Pollution
15 with Respiratory Hospitalization in a Government Designated sArea of Concern^: The Case of Windsor,
16 Ontario. Environ. Health Perspect. 10.1289/ehp.7300. Available: http://dx.doi.org/ (14 December 2004).
17 Moolgavkar, S. H.; Luebeck, E. G.; Hall, T. A.; Anderson, E. L. (1995) Air pollution and daily mortality in
18 Philadelphia. Epidemiology 6: 476-484.
19 Mortimer, K. M.; Neas, L. M.; Dockery, D. W.; Redline, S.; Tager, I. B. (2002) The effect of air pollution on
20 inner-city children with asthma. Eur. Respir. J. 19: 699-705.
21 Newson, E. J.; Krishna, M. T.; Lau, L. C. K.; Howarth, P. H.; Holgate, S. T.; Frew, A. J. (2000) Effects of
22 short-term exposure to 0.2 ppm ozone on biomarkers of inflammation in sputum, exhaled nitric oxide, and
23 lung function in subjects with mild atopic asthma. J. Occup. Environ. Med. 42: 270-277.
24 Nichols, B. G.; Woods, J. S.; Luchtel, D. L.; Corral, J.; Koenig, J. Q. (2001) Effects of ozone exposure on nuclear
25 factor-KB activation and tumor necrosis factor-a expression in human nasal epithelial cells. Toxicol. Sci.
26 60:356-362.
27 Nightingale, J. A.; Rogers, D. F.; Chung, K. F.; Barnes, P. J. (2000) No effect of inhaled budesonide on the response
28 to inhaled ozone in normal subjects. Am. J. Respir. Crit. Care Med. 161: 479-486.
29 Ostro, B. (1995) Fine paniculate air pollution and mortality in two Southern California counties. Environ. Res.
30 70:98-104.
31 Passannante, A. N.; Hazucha, M. J.; Bromberg, P. A.; Seal, E.; Folinsbee, L.; Koch, G. (1998) Nociceptive
32 mechanisms modulate ozone-induced human lung function decrements. J. Appl. Physiol. 85: 1863-1870.
33 Peel, J. L.; Tolbert, P. E.; Klein, M.; Metzger, K. B.; Planers, W. D.; Knox, T.; Mulholland, J. A.; Ryan, P. B.;
34 Frumkin, H. (2004) Ambient air pollution and respiratory emergency department visits. Epidemiology: in
35 press.
36 Pope, C. A., Ill; Burnett, R. T.; Thun, M. J.; Calle, E. E.; Krewski, D.; Ito, K.; Thurston, G. D. (2002) Lung cancer,
37 cardiopulmonary mortality, and long-term exposure to fine paniculate air pollution. JAMA J. Am. Med.
38 Assoc. 287: 1132-1141.
39 Postlethwait, E. M.; Cueto, R.; Velsor, L. W.; Pryor, W. A. (1998) Cyinduced formation of bioactive lipids:
40 estimated surface concentrations and lining layer effects. Am. J. Physiol. 274: L1006-L1016.
41 Romieu, I.; Meneses, F.; Ramirez, M.; Ruiz, S.; Padilla, R. P.; Sienra, J. J.; Gerber, M.; Grievink, L.; Dekker, R.;
42 Walda, I.; Brunekreef, B. (1998) Antioxidant supplementation and respiratory functions among workers
43 exposed to high levels of ozone. Am. J. Respir. Crit. Care Med. 158: 226-232.
44 Romieu, I.; Sienra-Monge, J. J.; Ramirez-Aguilar, M.; Tellez-Rojo, M. M.; Moreno-Macias, H.; Reyes-Ruiz, N. I.;
45 Del Rio-Navarro, B. E.; Ruiz-Navarro, M. X.; Hatch, G.; Slade, R.; Hernandez-Avila, M. (2002) Antioxidant
46 supplementation and lung functions among children with asthma exposed to high levels of air pollutants.
47 Am. J. Respir. Crit. Care Med. 166: 703-709.
48 Samet, J. M.; Zeger, S. L.; Dominici, F.; Curriero, F.; Coursac, L; Dockery, D. W.; Schwartz, J.; Zanobetti, A.
49 (2000) The national morbidity, mortality, and air pollution study. Part II: morbidity, mortality, and air
50 pollution in the United States. Cambridge, MA: Health Effects Institute; research report no. 94, part II.
51 Samet, J. M.; Hatch, G. E.; Horstman, D.; Steck-Scott, S.; Arab, L.; Bromberg, P. A.; Levine, M.; McDonnell,
52 W. F.; Devlin, R. B. (2001) Effect of antioxidant supplementation on ozone-induced lung injury in human
53 subjects. Am. J. Respir. Crit. Care Med. 164: 819-825.
54 Scannell, C.; Chen, L.; Aris, R. M.; Tager, I.; Christian, D.; Ferrando, R.; Welch, B.; Kelly, T.; Balmes, J. R. (1996)
55 Greater ozone-induced inflammatory responses in subjects with asthma. Am. J. Respir. Crit. Care Med.
56 154:24-29.
January 2005 8-49 DRAFT-DO NOT QUOTE OR CITE
-------
1 Schelegle, E. S.; Miller, L. A.; Gershwin, L. I; Fanucchi, M. V.; Van Winkle, L. S.; Gerriets, J. E.; Walby, W. F.;
2 Mitchell, V.; Tarkington, B. K.; Wong, V. J.; Baker, G. L.; Pantle, L. M.; Joad, J. P.; Pinkerton, K. E.;
3 Wu, R.; Evans, M. J.; Hyde, D. M.; Plopper, C. G. (2003) Repeated episodes of ozone inhalation amplifies
4 the effects of allergen sensitization and inhalation on airway immune and structural development in Rhesus
5 monkeys. Toxicol. Appl. Pharmacol. 191: 74-85.
6 Schierhorn, K.; Hanf, G.; Fischer, A.; Umland, B.; Olze, H.; Kunkel, G. (2002) Ozone-induced release of
7 neuropeptides from human nasal mucosa cells. Int. Arch. Allergy Immunol. 129: 145-151.
8 Schierhorn, K.; Zhang, M.; Matthias, C.; Kunkel, G. (1999) Influence of ozone and nitrogen dioxide on histamine
9 and interleukin formation in a human nasal mucosa culture system. Am. J. Respir. Cell Mol. Biol.
10 20: 1013-1019.
11 Schwartz, J. (2004) How sensitive is the association between ozone and daily deaths to control for temperature?
12 Am. J. Respir. Crit. Care Med.: in press, 10.1164/rccm.200407-933OC.
13 Schwartz, J.; Spix, C.; Touloumi, G.; Bacharova, L.; Barumamdzadeh, T.; le Tertre, A.; Piekarksi, T.; Ponce de
14 Leon, A.; Ponka, A.; Rossi, G.; Saez, M.; Schouten, J. P. (1996) Methodological issues in studies of air
15 pollution and daily counts of deaths or hospital admissions. In: St Leger, S., ed. The APHEA project.
16 Short term effects of air pollution on health: a European approach using epidemiological time series data.
17 J. Epidemiol. Commun. Health 50(suppl. 1): S3-SI 1.
18 Sherwin, R. P.; Richters, V.; Kraft, P.; Richters, A. (2000) Centriacinar region inflammatory disease in young
19 individuals: a comparative study of Miami and Los Angeles residents. Virchows Arch. 437: 422-428.
20 Stenfors, N.; Pourazar, J.; Blomberg, A.; Krishna, M. T.; Mudway, I.; Helleday, R.; Kelly, F. J.; Frew, A. J.;
21 Sandstrom, T. (2002) Effect of ozone on bronchial mucosal inflammation in asthmatic and healthy subjects.
22 Respir. Med. 96: 352-358.
23 Stieb, D. M.; Burnett, R. T.; Beveridge, R. C.; Brook, J. R. (1996) Association between ozone and asthma
24 emergency department visits in Saint John, New Brunswick, Canada. Environ. Health Perspect.
25 104: 1354-1360.
26 Tepper, J. S.; Wiester, M. J.; Weber, M. F.; Menache, M. G. (1990) Measurements of cardiopulmonary response in
27 awake rats during acute exposure to near-ambient concentrations of ozone. J. Appl. Toxicol. 10:7-15.
28 Tolbert, P. E.; Mulholland, J. A.; Macintosh, D. L.; Xu, F.; Daniels, D.; Devine, O. J.; Carlin, B. P.; Klein, M.;
29 Dorley, J.; Butler, A. J.; Nordenberg, D. F.; Frumkin, H.; Ryan, P. B.; White, M. C. (2000) Air quality and
30 pediatric emergency room visits for asthma in Atlanta, Georgia. Am. J. Epidemiol. 151: 798-810.
31 Trenga, C. A.; Koenig, J. Q.; Williams, P. V. (2001) Dietary antioxidants and ozone-induced bronchial
32 hyperresponsiveness in adults with asthma. Arch. Environ. Health 56: 242-249.
33 U.S. Environmental Protection Agency. (1996) Air quality criteria for ozone and related photochemical oxidants.
34 Research Triangle Park, NC: Office of Research and Development; report nos. EPA/600/AP-93/004aF-cF. 3v.
35 Available from: NTIS, Springfield, VA; PB96-185582, PB96-185590, and PB96-185608. Available:
36 http://cfpub2.epa.gov/ncea/.
37 U.S. Environmental Protection Agency. (2004) Air quality criteria for paniculate matter. Research Triangle Park,
3 8 NC: National Center for Environmental Assessment; report no. EPA/600/P-99/002aF-bF. 2v. Available:
39 http://cfpub.epa.gov/ncea/ [9 November, 2004].
40 Ultman, J. S.; Ben-Jebria, A.; Arnold, S. F. (2004) Uptake distribution of ozone in human lungs: intersubject
41 variability in physiologic response. Boston, MA: Health Effects Institute.
42 Vedal, S.; Brauer, M.; White, R.; Petkau, J. (2003) Air pollution and daily mortality in a city with low levels of
43 pollution. Environ. Health Perspect. Ill: 45-51.
44
January 2005 8-50 DRAFT-DO NOT QUOTE OR CITE
------- |