Disclaimer
This document has been reviewed in accordance with U.S. Environmental
Protection Agency policy and approved for publication. Mention of trade names or
commercial products does not constitute endorsement or recommendation for use.
-------
Preface
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. In 1979, the chemical designation of the standards
was changed from photochemical oxidants to ozone (O3). This document focuses 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.
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. The previous O3 criteria
document, Air Quality Criteria for Ozone and Other Photochemical Oxidants, was released in
August 1986 and a supplement, Summary of Selected New Information on Effects of Ozone on
Health and Vegetation, was released 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. The
purpose of this revised air quality criteria document for O3 and related photochemical
oxidants is to critically evaluate and assess the latest scientific data associated with exposure
to the concentrations of these pollutants found in ambient air. Emphasis is placed on the
presentation of health and environmental effects data; however, other scientific data are
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. Although the document is not intended to be an
exhaustive literature review, it is intended to cover all pertinent literature available through
1995.
This document was prepared and peer reviewed by experts from various state and
Federal governmental offices, academia, and private industry and reviewed in several public
meetings by the Clean Air Scientific Advisory Committee. The National Center for
Environmental Assessment (formerly the Environmental Criteria and Assessment Office) of
EPA's Office of Research and Development acknowledges with appreciation the contributions
provided by these authors and reviewers as well as the diligence of its staff and contractors in
the preparation of this document at the request of the Office of Air Quality Planning and
Standards.
-------
-------
Air Quality Criteria for Ozone
and Related Photochemical Oxidants
Table of Contents
Volume I
1. Executive Summary 1-1
2. Introduction 2-1
3. Tropospheric Ozone and Its Precursors 3-1
4. Environmental Concentrations, Patterns, and Exposure Estimates 4-1
Appendix A: Abbreviations and Acronyms A-l
Volume II
5. Environmental Effects of Ozone and Related Photochemical
Oxidants 5-1
Appendix A: Abbreviations and Acronyms A-l
Appendix B: Colloquial and Latin Names B-l
Volume III
6. Toxicological Effects of Ozone and Related Photochemical Oxidants ... 6-1
7. Human Health Effects of Ozone and Related Photochemical Oxidants ... 7-1
8. Extrapolation of Animal Toxicological Data to Humans 8-1
9. Integrative Summary of Ozone Health Effects 9-1
Appendix A: Abbreviations and Acronyms A-l
l-v
-------
-------
Table of Contents
Page
List of Tables III-xiv
List of Figures III-xix
Authors, Contributors, and Reviewers III-xxiii
U.S. Environmental Protection Agency Science Advisory Board,
Clean Air Scientific Advisory Committee III-xxxi
U.S. Environmental Protection Agency Project Team for Development
of Air Quality Criteria for Ozone and Related Photochemical Oxidants .... III-xxxiii
6. TOXICOLOGICAL EFFECTS OF OZONE AND RELATED
PHOTOCHEMICAL OXIDANTS 6-1
6.1 INTRODUCTION 6-1
6.2 RESPIRATORY TRACT EFFECTS OF OZONE 6-3
6.2.1 Biochemical Effects 6-3
6.2.1.1 Introduction 6-3
6.2.1.2 Cellular Targets of Ozone Interaction 6-4
6.2.1.3 Effects of Ozone Exposure on Lung
Lipid Metabolism 6-7
6.2.1.4 Effects of Ozone on Lung Antioxidant
Systems 6-8
6.2.1.5 Effects of Ozone on Lung Protein
Metabolism 6-13
6.2.1.6 Effects of Ozone Exposure on Lung Xenobiotic
Metabolism 6-17
6.2.1.7 Summary 6-19
6.2.2 Lung Inflammation and Permeability Changes 6-20
6.2.2.1 Introduction 6-20
6.2.2.2 Permeability Changes 6-21
6.2.2.3 Concomitant Changes in Permeability and
Inflammatory Cell Populations in the Lung . . . 6-29
6.2.2.4 Sensitive Populations 6-31
6.2.2.5 Repeated Exposures 6-33
6.2.2.6 Mediators of Inflammation and Permeability . . 6-33
6.2.2.7 Summary 6-35
6.2.3 Effects on Host Defense Mechanisms 6-36
6.2.3.1 Introduction 6-36
6.2.3.2 Mucociliary Clearance 6-36
6.2.3.3 Alveolobronchiolar Transport Mechanism .... 6-37
6.2.3.4 Alveolar Macrophages 6-39
I-VII
-------
Table of Contents (cont'd)
Page
6.2.3.5 Immunology 6-43
6.2.3.6 Interaction with Infectious Agents 6-48
6.2.3.7 Summary 6-52
6.2.4 Morphological Effects 6-54
6.2.4.1 Introduction 6-54
6.2.4.2 Sites Affected 6-57
6.2.4.3 Considerations of Exposure Regimens
and Methods 6-76
6.2.4.4 Considerations of Degree of Susceptibility to
Morphological Changes 6-78
6.2.4.5 Summary 6-80
6.2.5 Effects on Pulmonary Function 6-83
6.2.5.1 Introduction 6-83
6.2.5.2 Brief Ozone Exposures (Less Than
30 Minutes) 6-85
6.2.5.3 Acute Ozone Exposures (Less Than
One Day) 6-91
6.2.5.4 Repeated Acute Exposure Experiments (More
Than Three Days) 6-99
6.2.5.5 Longer Term Exposure Studies 6-100
6.2.5.6 Summary 6-102
6.2.6 Genotoxicity and Carcinogenicity of Ozone 6-103
6.2.6.1 Introduction 6-103
6.2.6.2 Ozone-Induced Deoxyribonucleic Acid
Damage 6-104
6.2.6.3 Induction of Mutation by Ozone 6-106
6.2.6.4 Induction of Cytogenetic Damage by Ozone . . . 6-108
6.2.6.5 Induction of Morphological Cell Transformation
by Ozone 6-109
6.2.6.6 Possible Direct Carcinogenic, Co-carcinogenic,
and Tumor-Promoting Effects of Ozone as
Studied in Whole Animal Carcinogenesis
Bioassays 6-111
6.2.6.7 Possible Effects of Ozone on Injected Tumor
Cells That Lodge in the Lung and Form Lung
Colonies 6-117
6.2.6.8 Summary and Conclusions 6-118
6.3 SYSTEMIC EFFECTS OF OZONE 6-121
6.3.1 Introduction 6-121
6.3.2 Central Nervous System and Behavioral Effects 6-122
6.3.3 Cardiovascular Effects 6-124
6.3.4 Hematological and Serum Chemistry Effects 6-126
I-VIII
-------
Table of Contents (cont'd)
Page
6.3.5 Other Systemic Effects 6-128
6.3.6 Summary 6-128
6.4 INTERACTIONS OF OZONE WITH OTHER
CO-OCCURRING POLLUTANTS 6-129
6.4.1 Introduction 6-129
6.4.2 Simple (Binary) Mixtures Containing Ozone 6-130
6.4.2.1 Nitrogen Dioxide as Copollutant 6-141
6.4.2.2 Acidic Compounds as Copollutants 6-146
6.4.2.3 Other Copollutants 6-148
6.4.3 Complex (Multicomponent) Mixtures Containing
Ozone 6-149
6.4.4 Summary 6-151
6.5 SUMMARY AND CONCLUSIONS 6-152
6.5.1 Introduction 6-152
6.5.2 Molecular Mechanisms of Effects 6-153
6.5.3 Respiratory Tract Effects 6-154
6.5.3.1 Effects on Host Defenses 6-154
6.5.3.2 Effects on Inflammation and Permeability .... 6-155
6.5.3.3 Effects on Structure, Function, and
Biochemistry 6-156
6.5.3.4 Genotoxicity and Carcinogenicity of Ozone . . . 6-161
6.5.3.5 Factors That Influence Ozone Exposure 6-162
6.5.4 Systemic Effects 6-163
6.5.4.1 Central Nervous System and Behavioral
Effects 6-163
6.5.4.2 Cardiovascular Effects 6-163
6.5.4.3 Reproductive and Developmental Effects 6-164
6.5.4.4 Other Systemic Effects 6-164
6.5.5 Effects of Mixtures 6-164
REFERENCES 6-166
7. HUMAN HEALTH EFFECTS OF OZONE AND RELATED
PHOTOCHEMICAL OXIDANTS 7-1
7.1 INTRODUCTION 7-1
7.2 CONTROLLED HUMAN EXPOSURE STUDIES 7-2
7.2.1 Pulmonary Function Effects of One- to Three-Hour
Ozone Exposures 7-2
7.2.1.1 Healthy Subjects 7-2
7.2.1.2 Subjects with Preexisting Disease 7-17
7.2.1.3 Influence of Gender, Age, Ethnic, and
Environmental Factors 7-25
l-ix
7.
-------
Table of Contents (cont'd)
Page
7.2.1.4 Repeated Exposures to Ozone 7-40
7.2.1.5 Effects on Exercise Performance 7-49
7.2.2 Pulmonary Function Effects of Prolonged (Multihour)
Ozone Exposures 7-53
7.2.3 Increased Airway Responsiveness 7-61
7.2.4 Inflammation and Host Defense 7-67
7.2.4.1 Introduction 7-67
7.2.4.2 Inflammation Assessed by Bronchoalveolar
Lavage 7-73
7.2.4.3 Inflammation Induced by Ambient Levels
of Ozone 7-75
7.2.4.4 Time Course of Inflammatory Response 7-76
7.2.4.5 Effect of Anti-Inflammatory Agents on
Ozone-Induced Inflammation 7-76
7.2.4.6 Use of Nasal Lavage To Assess Ozone-
Induced Inflammation in the Upper
Respiratory Tract 7-76
7.2.4.7 Changes in Host Defense Capability Following
Ozone Exposure 7-77
7.2.5 Extrapulmonary Effects of Ozone 7-78
7.2.6 Ozone Mixed with Other Pollutants 7-79
7.2.6.1 Ozone and Sulfur-Containing Pollutants 7-80
7.2.6.2 Ozone and Nitrogen-Containing Pollutants .... 7-85
7.2.6.3 Ozone, Peroxyacetyl Nitrate, and More
Complex Mixtures 7-87
7.2.6.4 Summary 7-88
7.3 SYMPTOMS AND PULMONARY FUNCTION IN
CONTROLLED STUDIES OF AMBIENT AIR EXPOSURES . . 7-88
7.3.1 Mobile Laboratory Studies 7-88
7.3.2 High-Altitude Studies 7-92
7.4 FIELD AND EPIDEMIOLOGY STUDIES 7-92
7.4.1 Acute Effects of Ozone Exposure 7-92
7.4.1.1 Introduction 7-92
7.4.1.2 Individual-Level Studies 7-93
7.4.1.3 Aggregate Population Time Series Studies .... 7-121
7.4.1.4 Summary and Conclusions 7-142
7.4.2 Chronic Effects of Ozone Exposure 7-143
7.4.2.1 Introduction 7-143
7.4.2.2 Recent Epidemiological Studies of Effects of
Chronic Exposure 7-144
7.4.2.3 Conclusions 7-158
l-x
7.
7.
-------
Table of Contents (cont'd)
Page
7.5 SUMMARY AND CONCLUSIONS 7-160
7.5.1 Controlled Human Studies of Ozone Exposure 7-160
7.5.1.1 Effects on Pulmonary Function 7-160
7.5.1.2 Symptom Responses to Ozone 7-164
7.5.1.3 Effects on Exercise Performance 7-165
7.5.1.4 Effects on Airway Responsiveness 7-165
7.5.1.5 Inflammation and Host Defense Effects 7-167
7.5.1.6 Factors Modifying Responsiveness to Ozone . . 7-167
7.5.1.7 Extrapulmonary Effects of Ozone 7-169
7.5.1.8 Effects of Ozone Mixed with Other
Pollutants 7-169
7.5.2 Field and Epidemiology Studies of Ozone Exposure ... 7-170
REFERENCES 7-173
EXTRAPOLATION OF ANIMAL TOXICOLOGICAL DATA
TO HUMANS 8-1
8.1 INTRODUCTION 8-1
8.2 OZONE DOSIMETRY 8-2
8.2.1 Introduction 8-2
8.2.2 Summary of 1986 Review of Experimental and
Theoretical Dosimetry 8-2
8.2.3 Experimental Ozone Dosimetry Data 8-3
8.2.3.1 Introduction 8-3
8.2.3.2 In Vivo Ozone Dosimetry Studies 8-3
8.2.3.3 In Vitro Ozone Dosimetry Studies 8-7
8.2.3.4 Human Ozone Dosimetry Studies 8-9
8.2.3.5 Intercomparison of Ozone Dosimetry Studies .. 8-15
8.2.4 Dosimetry Modeling 8-22
8.2.4.1 Background 8-22
8.2.4.2 Dosimetry Model Predictions 8-28
8.3 SPECIES SENSITIVITY: LUNG FUNCTION AND
INFLAMMATORY ENDPOINTS EXEMPLIFYING
AN APPROACH 8-33
8.3.1 Introduction 8-33
8.3.1.1 Dosimetry 8-34
8.3.2 Homology of Response 8-35
8.3.2.1 Lung Function Endpoints as Homologous
Indicators 8-35
8.3.2.2 Inflammatory and Antioxidant Endpoints
as Homologous Indicators 8-35
8.3.3 Studies of Lung Function 8-36
l-xi
-------
Table of Contents (cont'd)
Page
8.3.3.1 Confounding Influences in Lung Function
Studies 8-36
8.3.3.2 Acute Exposure Data 8-39
8.3.3.3 Prolonged Exposure Studies 8-52
8.4 QUANTITATIVE EXTRAPOLATION OF ACUTE OZONE
EFFECTS 8-55
8.4.1 Introduction 8-55
8.4.2 Intraspecies Delivered Dose Response 8-56
8.4.3 Interspecies Delivered Dose Response 8-56
8.5 QUANTITATIVE EXTRAPOLATION OF CHRONIC OZONE
EFFECTS 8-60
8.5.1 Introduction 8-60
8.5.2 Factors Considered in Estimating Dose 8-62
8.5.2.1 Human 8-62
8.5.2.2 Monkeys 8-68
8.5.2.3 Rats 8-68
8.5.3 Results and Discussion 8-70
8.5.3.1 Simulation Results 8-70
8.5.3.2 Interpretation of Chronic Site-Specific
Dose-Effect Estimates 8-72
8.6 SUMMARY AND CONCLUSIONS 8-76
8.6.1 Ozone Dosimetry 8-76
8.6.2 Species Homology and Sensitivity 8-78
8.6.3 Quantitative Extrapolation 8-79
REFERENCES 8-81
9. INTEGRATIVE SUMMARY OF OZONE HEALTH EFFECTS 9-1
9.1 INTRODUCTION 9-1
9.2 EXPOSURE-DOSE RELATIONSHIPS 9-2
9.3 EFFECTS OF SHORT-TERM OZONE EXPOSURES 9-4
9.3.1 Physiological Responses to Ozone Exposure 9-4
9.3.1.1 Respiratory Symptom Responses 9-4
9.3.1.2 Lung Function Responses 9-5
9.3.1.3 Changes in Airway Responsiveness 9-7
9.3.2 Exacerbation of Respiratory Disease 9-7
9.3.3 Morphological and Biochemical Abnormalities 9-9
9.3.3.1 Inflammation and Cell Damage 9-9
9.3.3.2 Host Defense 9-12
9.3.4 Quantitative Ozone Exposure-Response
Relationships 9-13
9.3.4.1 Prediction and Summary of Mean
Responses 9-15
I-XII
-------
Table of Contents (cont'd)
9.3.4.2 Prediction and Summary of Individual
Responses 9-19
9.4 EFFECTS OF LONG-TERM OZONE EXPOSURES 9-26
9.4.1 Repeated Exposures 9-26
9.4.2 Prolonged Exposures 9-27
9.4.3 Genotoxicity and Carcinogenicity of Ozone 9-31
9.5 EFFECTS OF COMBINED POLLUTANT EXPOSURES 9-31
9.6 CONCLUSIONS 9-32
REFERENCES 9-37
APPENDIX A: ABBREVIATIONS AND ACRONYMS A-l
I-XIII
-------
List of Tables
Number Page
6-1 Effects of Ozone Exposure on Lung Lipids 6-9
6-2 Effects of Ozone Exposure on Lung Antioxidants 6-11
6-3 Effects of Ozone Exposure on Lung Proteins 6-15
6-4 Effects of Ozone Exposure on Lung Xenobiotic Metabolism 6-18
6-5 Lung Inflammation and Permeability Changes Associated
with Ozone Exposure 6-22
6-6 Effects of Ozone on Host Defense Mechanisms:
Physical Clearance 6-38
6-7 Effects of Ozone on Host Defense Mechanisms:
Macrophage Alterations 6-40
6-8 Effects of Ozone on Host Defense Mechanisms:
Immunology 6-45
6-9 Effects of Ozone on Host Defense Mechanisms:
Interactions with Infectious Agents 6-50
6-10 Effects of Ozone on Conducting Airways 6-58
6-11 Effects of Ozone on Lung Structure: Short-Term
Exposures (Less Than Two Weeks) 6-61
6-12 Effects of Ozone on Lung Structure: Long-Term
Exposures (More Than Two Weeks) 6-64
6-13 Effects of Ozone on Pulmonary Function 6-86
6-14 Effects of Ozone on Airway Reactivity 6-88
6-15 Effects of Ozone on Deoxyribonucleic Acid Damage 6-105
6-16 Summary of Findings on the Mutagenicity of Ozone 6-107
6-17 Effects of Ozone on Morphological Cell Transformation 6-110
l-xiv
-------
List of Tables (cont'd)
Number Page
6-18 Summary of Results on the Possible Carcinogenicity of Ozone . . . 6-112
6-19 Alveolar/Bronchiolar Tumor Incidence in B6C3FJ Mice in the
National Toxicology Program's Chronic Ozone Study 6-116
6-20 Effects of Inhaled Ozone on the Ability of Injected Tumor
Cells to Colonize the Lungs of Mice 6-118
6-21 Summary of Data on the Genotoxicity of Ozone 6-119
6-22 Effects of Ozone on Behavior 6-123
6-23 Effects of Ozone on the Cardiovascular System 6-125
6-24 Hematology and Serum Chemistry Effects 6-127
6-25 Toxicological Interactions of Ozone and Nitrogen Dioxide 6-131
6-26 Toxicological Interactions to Binary Mixtures of Ozone with
Acids and Other Pollutants 6-136
7-1 Controlled Exposure of Healthy Human Subjects to Ozone 7-4
7-2 Ozone Exposure in Subjects with Preexisting Disease 7-18
7-3 Gender Differences in Pulmonary Function Responses to Ozone . . 7-26
7-4 Hormonal Influences on Pulmonary Function Responses to
Ozone 7-33
7-5 Age Differences in Pulmonary Function Responses to Ozone .... 7-34
7-6 Changes in Forced Expiratory Lung Volume After Repeated
Daily Exposure to Ozone 7-41
7-7 Pulmonary Function Effects with Repeated Exposures
to Ozone 7-43
7-8 Ozone Effects on Exercise Performance 7-50
7-9 Pulmonary Function Effects After Prolonged Exposures to
Ozone 7-54
l-xv
-------
List of Tables (cont'd)
Number Page
7-10 Increased Airway Responsiveness Following Ozone Exposures . . . 7-63
7-11 Bronchoalveolar Lavage Studies of Inflammatory Effects from
Controlled Human Exposure to Ozone 7-68
7-12 Additional Studies of Inflammatory and Host Defense Effects from
Controlled Human Exposure to Ozone 7-70
7-13 Ozone Mixed with Other Pollutants 7-81
7-14 Acute Effects of Ozone in Ambient Air in Field Studies with a
Mobile Laboratory 7-90
7-15 Acute Effects of Photochemical Oxidant Pollution: Lung Function
in Camp Studies 7-97
7-16 Slopes from Regressions of Forced Expiratory Volume in One
Second on Ozone for Six Camp Studies 7-103
7-17 Acute Effects of Photochemical Oxidant Pollution: Lung Function
in Exercising Subjects 7-106
7-18 Acute Effects of Photochemical Oxidant Pollution: Daily Life
Studies of Lung Function and Respiratory Symptoms 7-110
7-19 Acute Effects of Photochemical Oxidant Pollution: Symptom
Prevalance 7-114
7-20 Aggravation of Existing Respiratory Diseases by Photochemical
Oxidant Pollution 7-117
7-21 Hospital Admissions/Visits in Relation to Photochemical
Oxidant Pollution: Time Series Studies 7-124
7-22 Comparison of Regressions of Daily Summertime Respiratory
Admissions on Ozone and Temperature in Toronto, Ontario,
and Buffalo, New York, for the Summer of 1988 7-133
7-23 Summary of Effect Estimates for Ozone in Recent Studies of
Respiratory Hospital Admissions 7-136
l-xvi
-------
List of Tables (cont'd)
Number Page
7-24 Daily Mortality Associated with Exposure to Photochemical
Oxidant Pollution 7-138
7-25 Pathologic and Immunologic Changes Associated with Chronic
Ozone Exposure 7-147
7-26 Effects of Chronic Ozone Exposure on Pulmonary Function,
Respiratory Symptoms, and Chronic Respiratory Disease 7-148
7-27 Effects of Chronic Ozone Exposure on the Incidence of
Cardiovascular and Malignant Diseases 7-159
8-1 Experimental Studies on Ozone Dosimetry 8-4
8-2 Total Respiratory Tract Uptake Data 8-16
8-3 Unidirectional Upper Respiratory Tract Uptake Efficiency Data ... 8-17
8-4 Lower Respiratory Tract Uptake Efficiency Data 8-18
8-5 Theoretical Ozone Dosimetry Investigations 8-23
8-6 Assumption for Application of Dosimetry Model to Breathing
Frequency Responses to Ozone 8-32
8-7 Comparison of Total Respiratory Tract Uptake Data with
Model Predictions 8-32
8-8 Pulmonary Antioxidant Substances in Various Laboratory Animal
Species and Humans 8-37
8-9 Polymorphonuclear Leukocyte and Protein Permeability Response
to Ozone by Species 8-48
8-10 The Basis of Information for Model Parameters 8-63
8-11 Estimating Monkey Ventilatory Parameters 8-69
8-12 Summary of Simulation Results 8-71
8-13 Summaries of Study Data Used in Extrapolation of Chronic
Ozone Effects 8-74
I-XVII
-------
List of Tables (cont'd)
Number Page
9-1 Gradation of Individual Responses to Short-Term Ozone
Exposure in Healthy Persons 9-24
9-2 Gradation of Individual Responses to Short-Term Ozone
Exposure in Persons with Impaired Respiratory Systems 9-25
I-XVIII
-------
List of Figures
Number Page
6-1 Major secondary products of ozone interaction with
lung cells 6-5
6-2 Schematic representation of intrapulmonary conducting airways
and acini from animals with respiratory bronchioles (human
and nonhuman primates) and without respiratory bronchioles
(rats and most rodents) 6-55
6-3 Electron micrograph of alveolar septa in the centriacinar region
of the lungs from laboratory rats exposed to a simulated pattern
of ambient ozone for 78 weeks, showing thickened basement
membrane 6-74
6-4 Schematic comparison of the duration-response profiles for
epithelial hyperplasia, bronchoalveolar exudation, and
interstitial fibrosis in the centriacinar region of lung
exposed to a constant low concentration of ozone 6-81
7-1 Individual concentration-response curves for five separate
subjects exposed to 0.10, 0.15, 0.20, and 0.25 ppm ozone for
two hours with moderate intermittent exercise 7-13
7-2 Mean percent change in post-minus prevalues of
forced expiratory volume in one second for each
gender-race group 7-29
7-3 The forced expiratory volume in one second is shown
in relation to exposure duration at different ozone
concentrations 7-57
7-4 The distribution of response for 87 subjects exposed to
clean air and at least one of 0.08, 0.10, or 0.12 ppm ozone is
shown here 7-58
7-5 The forced expiratory volume in one second is shown in
relation to exposure duration under three exposure conditions .... 7-60
7-6 Airway function can be measured before and immediately after
the inhalation of an aerosolized bronchoconstrictor drug
like methacholine 7-62
l-xix
-------
List of Figures (cont'd)
Number Page
8-1 Total respiratory tract uptake as a function of inspiratory
flow in humans 8-19
8-2 Unidirectional uptake efficiency in the upper respiratory
tract by the nasal pathway 8-20
8-3 Unidirectional uptake efficiency in the upper respiratory
tract by the oral pathway 8-21
8-4 Uptake efficiency of the lower respiratory tract as a
function of inspiratory flow in humans 8-22
8-5 Net dose and tissue dose versus sequential segments along
anatomical model airway paths for human, rat, guinea pig,
and rabbit 8-29
8-6 Parallelogram paradigm for utilizing animal data for human
health predictions 8-34
8-7 Comparison of changes in frequency of breathing after ozone
exposure in humans and animals 8-40
8-8 Comparison of changes in resistance after ozone exposure in
humans and animals 8-43
8-9 Comparison of changes in forced vital capacity after ozone
exposure in humans and animals 8-47
8-10 Composite of data from Slade et al. (1989), Koren et al.
(1989b), and Crissman et al. (1993) comparing basal
bronchoalveolar lavage ascorbate levels to ozone-induced
changes in bronchoalveolar lavage protein 8-50
8-11 Composite of data comparing polymorphonuclear leukocytes
obtained by bronchoalveolar lavage 16 to 18 hours after
ozone exposure of humans (with exercise) to those of rats
exposed to ozone at rest or hyperventilated with carbon
dioxide 8-51
8-12 Comparison of changes in forced vital capacity in
humans and frequency of breathing in rats with up to
five consecutive days of ozone exposure 8-53
l-xx
-------
List of Figures (cont'd)
Number Page
8-13 Changes in forced expiratory volume in one second versus
pulmonary tissue dose 8-57
8-14 Relationship between change in alveolar wall thickness
and predicted ozone dose as a function of distance
from the bronchiole-alveolar duct junction 8-58
8-15 The maximum ratio of ozone-altered breathing frequency to
control breathing frequency at various ozone concentrations versus
predicted average dose rate to the proximal alveolar region 8-59
8-16 Protein in the bronchoalveolar lavage for several laboratory
animal species and humans, as related to the estimated
pulmonary dose 8-60
8-17 The variation in exposure concentration for the New York City
adult and child 8-66
8-18 Change from control of total interstitial and acellular thickness
for rats exposed to ozone in the U.S. Environmental Protection
Agency (Chang et al., 1992) and National Toxicology Program/
Health Effects Institute (Chang et al., 1994) studies 8-73
9-1 Mean predicted changes in forced expiratory volume in one second
following two-hour exposures to ozone with increasing levels of
intermittent exercise 9-16
9-2 Predicted mean decrements in forced expiratory volume in one
second for one- and two-hour exposures to ozone with
intermittent heavy exercise and 6.6-hour exposures with
moderate prolonged exercise 9-17
9-3 Derived means of bronchoalveolar lavage protein denoted by
symbols and the exponential model shown by lines as time of
exposure varies from two to eight hours 9-17
9-4 Predicted mean forced vital capacity for rats exposed to
ozone while undergoing intermittent carbon dioxide-induced
hyperpnea 9-18
l-xxi
-------
List of Figures (cont'd)
Number Page
9-5 Average number of adjusted respiratory admissions among all
168 hospitals by decile of the daily one-hour maximum ozone
level, lagged one day 9-18
9-6 Predicted mean decrements in forced expiratory volume
in one second following two-hour exposures to ozone
while undergoing heavy intermittent exercise for three ages 9-19
9-7 The distribution of response for 87 subjects exposed to
clean air and at least one of 0.08, 0.10, or 0.12 ppm
ozone 9-20
9-8 Histograms of regression slopes for forced expiratory volume in
one second versus one-hour ozone concentration in children
attending a summer camp in northwestern New Jersey 9-21
9-9 Proportion of heavily exercising individuals predicted to
experience a 10% decrement in forced expiratory volume in one
second following a one- or two-hour exposure to ozone 9-22
9-10 Proportion of heavily exercising individuals predicted
to experience mild cough following a two-hour
ozone exposure 9-22
9-11 Proportion of moderately exercising individuals exposed to ozone
for 6.6 hours predicted to experience 5, 10, or 15% decrements
in forced expiratory volume in one second as a function of
concentration times time for age equal to 24 years 9-23
9-12 A summary of the morphologic lesions found in the terminal
bronchioles and centriacinar region of the lung following
exposure of laboratory rats to filtered air or a simulated
ambient pattern of ozone for up to 78 weeks 9-28
9-13 Schematic comparison of the duration-response profiles for
epithelial hyperplasia, bronchoalveolar exudation, and
interstitial fibrosis in the centriacinar region of lung exposed
to a constant low concentration of ozone 9-29
I-XXII
-------
Authors, Contributors, and Reviewers
Chapter 6. Toxicological Effects of Ozone
and Related Photochemical Oxidants
Principal Authors
Dr. David J.P. Bassett—Department of Occupational and Environmental Health, College of
Pharmacy and Allied Health Professions, Shapero Hall, Wayne State University, Detroit, MI
48202
Dr. Deepak K. Bhalla—Department of Community and Environmental Medicine, University of
California, Irvine, CA 92717
Dr. Judith A. Graham—National Exposure Research Laboratory (MD-75), U.S. Environmental
Protection Agency, Research Triangle Park, NC 27711
Dr. George J. Jakab—Department of Environmental Health Sciences, The Johns Hopkins
University, School of Hygiene and Public Health, 615 N. Wolfe Street, Baltimore, MD 21209
Dr. Joseph R. Landolph—Department of Microbiology, USC Comprehensive Cancer Center,
USC School of Medicine, Norris Cancer Hospital and Research Institute,
1441 Eastlake Avenue, Los Angeles, CA 90033
Dr. Richard B. Schlesinger—Institute of Environmental Medicine, New York University
Medical Center, Long Meadow Road, Tuxedo, NY 10987
Dr. Jeffrey S. Tepper—Genentech, Inc., 460 Point San Bruno Blvd., South San Francisco, CA
94080
Dr. Walter S. Tyler—Department of Anatomy, School of Veterinary Medicine, University of
California, Davis, CA 95616
Contributors and Reviewers
Dr. Gary A. Boorman—National Institute of Environmental Health Sciences, P.O. Box 12233,
Research Triangle Park, NC 27709
Dr. Daniel L. Costa—National Health and Environmental Effects Research Laboratory
(MD-82), U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
I-XXIII
-------
Authors, Contributors, and Reviewers (cont'd)
Dr. Max Costa—Department of Environmental Medicine, New York University Medical
Center, 550 First Avenue, New York, NY 10016
Dr. Donald L. Dungworth—Department of Veterinary Pathology, School of Veterinary
Medicine, University of California, Davis, CA 95616
Dr. Ian Gilmour—Center for Environmental Medicine and Lung Biology, U.S. Environmental
Protection Agency (MD-92), Research Triangle Park, NC 27711
Dr. Kenneth B. Gross—Biomedical Science Department, General Motors Research Laboratory,
Warren, MI 48090
Dr. Gary E. Hatch—National Health and Environmental Effects Research Laboratory (MD-82),
U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
Dr. Rogene F. Henderson—Inhalation Toxicology Research Institute, Lovelace Biomedical and
Environmental Research Institute, P.O. Box 5890, Albuquerque, NM 87185
Dr. Dallas M. Hyde—Department of Anatomy, School of Veterinary Medicine, University of
California, Davis, CA 95616
Dr. Jerald A. Last—California Primate Research Center, University of California, Davis, CA
95616
Dr. Daniel B. Menzel—Department of Community and Environmental Medicine, College of
Medicine, University of California, Irvine, CA 92715
Dr. Robert R. Mercer—Department of Medicine, Division of Allergy and Respiratory Disease,
P.O. Box 3177, Duke University Medical Center, Durham, NC 27710
Dr. Frederick J. Miller—Chemical Industry Institute of Toxicology, P.O. Box 12137, Research
Triangle Park, NC 27709
Dr. Kathleen M. Nauss—Health Effects Institute, 141 Portland St., Suite 7300,
Cambridge, MA 02139
Dr. Steven C. Nesnow—National Health and Environmental Effects Research Laboratory
(MD-68), U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
Dr. Robert F. Phalen—Department of Community and Environmental Medicine, University of
California, Irvine, CA 92717
l-xxiv
-------
Authors, Contributors, and Reviewers (cont'd)
Dr. Charles G. Plopper—Department of Anatomy, School of Veterinary Medicine, University
of California, Davis, CA 95616
Dr. Edward M. Postlethwait—Pulmonary Division H-76, Department of Internal Medicine,
University of Texas Medical School, Galveston, TX 77550
Mr. Mark E. Raizenne—Health and Welfare Canada, Environmental Health Center, Tunney's
Pasture, Ottawa, Ontario K1A OL2, CANADA
Dr. Edward S. Schelegle—Department of Human Physiology, School of Medicine, Building 1-
A, Room 4140, University of California, Davis, CA 95616
Dr. Mary Jane Selgrade—National Health and Environmental Effects Research Laboratory
(MD-92), U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
Dr. David Warshawsky—Kettering Laboratories, University of Cincinnati, School of Medicine,
Department of Environmental Health, 3223 Eden Avenue, Cincinnati, OH 45267
Dr. M. Jean Wiester—National Health and Environmental Effects Research Laboratory
(MD-82), U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
Dr. Jeannette Wiltse—National Center for Environmental Assessment (8601), U.S.
Environmental Protection Agency, 401 M. Street, S.E., Washington, DC 20460
Dr. Hanspeter R. Witschi—Toxic Substances Research and Teaching Program, University of
California, Davis, CA 95616
Chapter 7. Human Health Effects of Ozone
and Related Photochemical Oxidants
Principal Authors
Dr. Robert M. Aris—Department of Medicine, The University of North Carolina,
724 Burnette-Womack Bldg., Chapel Hill, NC 27599
Dr. John R. Balmes—Center for Occupational and Environmental Health, San Francisco
General Hospital, 1001 Potrero Avenue, Bldg. 30, 5th Floor, San Francisco, CA 94110
Dr. Robert B. Devlin—National Health and Environmental Effects Research Laboratory
(MD-58), U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
l-xxv
-------
Authors, Contributors, and Reviewers (cont'd)
Dr. Deborah M. Drechsler-Parks—Environmental Stress Laboratory, Neuroscience Research
Institute, University of California, Santa Barbara, CA 93106
Dr. Lawrence J. Folinsbee—National Health and Environmental Effects Research Laboratory
(MD-58), U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
Dr. Henry Gong, Jr.—Rancho Los Amigos Medical Center, 7601 East Imperial Highway,
Downey, CA 90242
Dr. Patrick L. Kinney—School of Public Health, Columbia University, 60 Haven Avenue,
New York, NY 10032
Dr. Edward S. Schelegle—Department of Human Physiology, School of Medicine, Building 1-
A, Room 4140, University of California, Davis, CA 95616
Dr. Ira B. Tager—School of Public Health, 140 Earl Warren Hall, University of California,
Berkeley, CA 94720
Dr. George D. Thurston—Institute of Environmental Medicine, New York University Medical
Center, Long Meadow Road, Tuxedo, NY 10987
Contributors and Reviewers
Dr. William C. Adams—Department of Physical Education, University of California,
Davis, CA 95616
Dr. Rebecca Bascom—The University of Maryland School of Medicine, 10 S. Pine St.,
MSTF-800, Baltimore, MD 21201
Dr. David V. Bates—Department of Health Care and Epidemiology, University of British
Columbia, 5804 Fairview Ave., Vancouver, British Columbia V6T1Z3, CANADA
Dr. William S. Beckett—Yale School of Medicine, 333 Cedar St., New Haven, CT 06510
Dr. Philip A. Bromberg—Department of Medicine, The University of North Carolina,
724 Burnett-Womack Bldg., 229H, Chapel Hill, NC 27599
Dr. Douglas Dockery—Harvard School of Public Health, Environmental Epidemiology,
665 Huntington Avenue, Boston, MA 02115
Dr. William Eschenbacher—Methodist Hospital, 6565 Fannin, M.S., F980,
Houston, TX 77030
l-xxvi
-------
Authors, Contributors, and Reviewers (cont'd)
Dr. Mark Frampton—Pulmonary Disease Unit, Box 692, University of Rochester Medical
Center, 601 Elmwood Avenue, Rochester, NY 14642
Dr. Judith A. Graham—National Center for Environmental Assessment (MD-52),
U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
Dr. Milan J. Hazucha—Department of Medicine, Center for Environmental Medicine and Lung
Biology, The University of North Carolina, Trailer #4, Medical Bldg. C 224H, Chapel Hill,
NC 27599
Dr. Steven M. Horvath—Environmental Stress Laboratory, Neuroscience Research Institute,
University of California, Santa Barbara, CA 93106
Dr. Howard R. Kehrl—National Health and Environmental Effects Research Laboratory
(MD-58), U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
Dr. Jane Q. Koenig—Department of Environmental Health, School of Public Health and
Community Medicine, University of Washington, SC-34, Seattle, WA 98195
Dr. Hillel S. Koren—National Health and Environmental Effects Research Laboratory
(MD-58), U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
Dr. Debra L. Laskin—Department of Pharmacology and Toxicology, College of Pharmacy,
Rutgers University, Piscataway, NY 08855
Mr. William S. Linn—Rancho Los Amigos Medical Center, 51 Medical Science Building,
7601 East Imperial Highway, Downey, CA 90242
Dr. Fred Lipfert—23 Carll Ct, Northport, NY 11768
Dr. William F. McDonnell—National Health and Environmental Effects Research Laboratory
(MD-58), U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
Mr. Mark E. Raizenne—Health and Welfare Canada, Environmental Health Center, Tunney's
Pasture, Ottawa, Ontario K1A OL2, CANADA
Dr. Jonathan M. Samet—Department of Epidemiology, School of Hygiene and Public Health,
Johns Hopkins University, 615 N. Wolfe St., Baltimore, MD 21205
Dr. Joel Schwartz—Environmental Epidemiology Program, Harvard School of Public Health,
665 Huntington Avenue, Boston, MA 02115
I-XXVII
-------
Authors, Contributors, and Reviewers (cont'd)
Dr. Elston Seal—National Health and Environmental Effects Research Laboratory (MD-58),
U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
Dr. Dalia M. Spektor—2242 20th Street (#1), Santa Monica, CA 90405
Chapter 8. Extrapolation of Animal Toxicological Data
to Humans
Principal Authors
Dr. Daniel L. Costa—National Health and Environmental Effects Research Laboratory
(MD-82), U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
Dr. Timothy R. Gerrity—Medical Research Services, Department of Veterans Affairs,
810 Vermont Ave., N.W., Washington, DC 20420
Dr. John H. Overton—National Health and Environmental Effects Research Laboratory
(MD-82), U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
Reviewers
Dr. Daniel B. Menzel—Department of Community and Environmental Medicine, College of
Medicine, University of California, Irvine, CA 92715
Dr. Robert R. Mercer—Department of Medicine, Division of Allergy and Respiratory Disease,
P.O. Box 3177, Duke University Medical Center, Durham, NC 27710
Dr. Frederick J. Miller—Chemical Industry Institute of Toxicology, P.O. Box 12137, Research
Triangle Park, NC 27709
Dr. Kathleen M. Nauss—Health Effects Institute, 141 Portland St., Suite 7300,
Cambridge, MA 02139
Dr. Edward M. Postlethwait—Pulmonary Division H-76, Department of Internal Medicine,
University of Texas Medical School, Galveston, TX 77550
Mr. Mark E. Raizenne—Health and Welfare Canada, Environmental Health Center, Tunney's
Pasture, Ottawa, Ontario K1A OL2, CANADA
I-XXVIII
-------
Authors, Contributors, and Reviewers (cont'd)
Dr. James S. Ultman—Department of Chemical Engineering, Pennsylvania State University,
University Park, PA 16802
Dr. M. Jean Wiester—National Health and Environmental Effects Research Laboratory
(MD-82), U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
Chapter 9. Integrative Summary of Ozone Health Effects
Principal Authors
Dr. Daniel L. Costa—National Health and Environmental Effects Research Laboratory
(MD-82), U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
Dr. Robert B. Devlin—National Health and Environmental Effects Research Laboratory
(MD-58), U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
Dr. Lawrence J. Folinsbee—National Health and Environmental Effects Research Laboratory
(MD-58), U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
Dr. Timothy R. Gerrity—Medical Research Services, Department of Veterans Affairs,
810 Vermont Ave., N.W., Washington, DC 20420
Dr. Judith A. Graham—National Center for Environmental Assessment (MD-52),
U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
Dr. William F. McDonnell—National Health and Environmental Effects Research Laboratory
(MD-58), U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
Mr. James A. Raub—National Center for Environmental Assessment (MD-52),
U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
Contributors and Reviewers
Dr. Frederick J. Miller—Chemical Industry Institute of Toxicology, P.O. Box 12137, Research
Triangle Park, NC 27709
Dr. Edward S. Schelegle—Department of Human Physiology, School of Medicine, Building 1-
A, Room 4140, University of California, Davis, CA 95616
I-XXIX
-------
Authors, Contributors, and Reviewers (cont'd)
Dr. Walter S. Tyler—Department of Anatomy, School of Veterinary Medicine, University of
California, Davis, CA 95616
Dr. Jeannette Wiltse—National Center for Environmental Assessment, U.S. Environmental
Protection Agency (8601), 401 M. Street, S.E., Washington, DC 20460
l-xxx
-------
U.S. Environmental Protection Agency
Science Advisory Board
Clean Air Scientific Advisory Committee
Ozone Review
Chairman
Dr. George T. Wolff—General Motors Corporation, Environmental and Energy Staff,
General Motors Bldg., 12th Floor, 3044 West Grand Blvd., Detroit, MI 48202
Members
Dr. Stephen Ayres—Office of International Health Programs, Virginia Commonwealth
University, Medical College of Virginia, Box 980565, Richmond, VA 23298
Dr. Jay S. Jacobson—Boyce Thompson Institute, Tower Road, Cornell University, Ithaca, NY
14853
Dr. Joseph Mauderly—Inhalation Toxicology Research Institute, Lovelace Biomedical and
Environmental Research Institute, P.O. Box 5890, Albuquerque, NM 87185
Dr. Paulette Middleton—Science & Policy Associates, Inc., Western Office, Suite 140,
3445 Penrose Place, Boulder, CO 80301
Dr. James H. Price, Jr.—Research and Technology Section, Texas Natural Resources
Conservation Commission, P.O. Box 13087, Austin, TX 78711
Invited Scientific Advisory Board Members
Dr. Morton Lippmann—Institute of Environmental Medicine, New York University Medical
Center, Long Meadow Road, Tuxedo, NY 10987
Dr. Roger O. McClellan—Chemical Industry Institute of Toxicology, P.O. Box 12137,
Research Triangle Park, NC 27711
Consultants
Dr. Stephen D. Colome—Integrated Environmental Services, University Tower, Suite 280,
4199 Campus Drive, Irvine, CA 92715
Ill-xxxi
-------
U.S. Environmental Protection Agency
Science Advisory Board
Clean Air Scientific Advisory Committee
(cont'd)
Dr. A. Myrick Freeman—Department of Economics, Bowdoin College, Brunswick, ME 04011
Dr. Allan Legge—Biosphere Solutions, 1601 llth Avenue, NW, Calgary, Alberta T2N 1H1,
CANADA
Dr. William Manning—Department of Plant Pathology, University of Massachusetts, Amherst,
MA 01003
Dr. D. Warner North—Decision Focus, Inc., 650 Castro Street, Suite 300, Mountain View,
CA 94041
Dr. Frank E. Speizer—Harvard Medical School, Channing Lab, 180 Longwood Avenue,
Boston, MA 02115
Dr. George E. Taylor—Department of Environmental and Resource Sciences, 130 Fleischmann
Agriculture Bldg. 199, University of Nevada, Reno, NV 89557
Dr. Mark J. Utell—Pulmonary Disease Unit, Box 692, University of Rochester Medical
Center, 601 Elmwood Avenue, Rochester, NY 14642
Designated Federal Official
Mr. Randall C. Bond—Science Advisory Board (1400), U.S. Environmental Protection
Agency, 401 M Street, SW, Washington, DC 20460
Staff Assistant
Ms. Lori Anne Gross—Science Advisory Board (1400), U.S. Environmental Protection
Agency, 401 M Street, SW, Washington, DC 20460
I-XXXII
-------
U.S. Environmental Protection Agency
Project Team for Development of Air Quality Criteria
for Ozone and Related Photochemical Oxidants
Scientific Staff
Mr. James A. Raub—Health Scientist, National Center for Environmental Assessment
(MD-52), U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
Dr. A. Paul Altshuller—Physical Scientist, National Center for Environmental Assessment
(MD-52), U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
Mr. William G. Ewald—Health Scientist, National Center for Environmental Assessment (MD-
52), U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
Dr. J.H.B. Garner—Ecologist, National Center for Environmental Assessment (MD-52),
U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
Dr. Judith A. Graham—Associate Director, National Center for Environmental Assessment
(MD-52), U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
Ms. Ellie R. Speh—Secretary, National Center for Environmental Assessment (MD-52),
U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
Ms. Beverly E. Tilton—Physical Scientist, National Center for Environmental Assessment
(MD-52), U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
Technical Support Staff
Mr. Douglas B. Fennell—Technical Information Specialist, National Center for Environmental
Assessment (MD-52), U.S. Environmental Protection Agency, Research Triangle Park, NC
27711
Mr. Allen G. Hoyt—Technical Editor and Graphic Artist, National Center for Environmental
Assessment (MD-52), U.S. Environmental Protection Agency, Research Triangle Park, NC
27711
Ms. Diane H. Ray—Technical Information Manager (Public Comments), National Center for
Environmental Assessment (MD-52), U.S. Environmental Protection Agency, Research
Triangle Park, NC 27711
I-XXXIII
-------
U.S. Environmental Protection Agency
Project Team for Development of Air Quality Criteria
for Ozone and Related Photochemical Oxidants
(cont'd)
Mr. Richard N. Wilson—Clerk, National Center for Environmental Assessment (MD-52), U.S.
Environmental Protection Agency, Research Triangle Park, NC 27711
Document Production Staff
Ms. Marianne Barrier—Graphic Artist, ManTech Environmental Technology, Inc.,
P.O. Box 12313, Research Triangle Park, NC 27709
Mr. John R. Barton—Document Production Coordinator, ManTech Environmental Technology,
Inc., P.O. Box 12313, Research Triangle Park, NC 27709
Ms. Lynette D. Cradle—Word Processor, ManTech Environmental Technology, Inc., P.O. Box
12313, Research Triangle Park, NC 27709
Ms. Shelia H. Elliott—Word Processor, ManTech Environmental Technology, Inc.,
P.O. Box 12313, Research Triangle Park, NC 27709
Ms. Sandra K. Eltz—Word Processor, ManTech Environmental Technology, Inc.,
P.O. Box 12313, Research Triangle Park, NC 27709
Ms. Jorja R. Followill—Word Processor, ManTech Environmental Technology, Inc.,
P.O. Box 12313, Research Triangle Park, NC 27709
Ms. Sheila R. Lassiter—Word Processor, ManTech Environmental Technology, Inc.,
P.O. Box 12313, Research Triangle Park, NC 27709
Ms. Wendy B. Lloyd—Word Processor, ManTech Environmental Technology, Inc.,
P.O. Box 12313, Research Triangle Park, NC 27709
Ms. Carolyn T. Perry—Word Processor, ManTech Environmental Technology, Inc.,
P.O. Box 12313, Research Triangle Park, NC 27709
Ms. Cheryl B. Thomas—Word Processor, ManTech Environmental Technology, Inc.,
P.O. Box 12313, Research Triangle Park, NC 27709
Mr. Peter J. Winz—Technical Editor, ManTech Environmental Technology, Inc.,
P.O. Box 12313, Research Triangle Park, NC 27709
l-xxxiv
-------
U.S. Environmental Protection Agency
Project Team for Development of Air Quality Criteria
for Ozone and Related Photochemical Oxidants
(cont'd)
Technical Reference Staff
Mr. John A. Bennett—Bibliographic Editor, ManTech Environmental Technology, Inc.,
P.O. Box 12313, Research Triangle Park, NC 27709
Ms. S. Blythe Hatcher—Bibliographic Editor, Information Organizers, Inc., P.O. Box 14391,
Research Triangle Park, NC 27709
Ms. Susan L. McDonald—Bibliographic Editor, Information Organizers, Inc., P.O. Box 14391,
Research Triangle Park, NC 27709
Ms. Carol J. Rankin—Bibliographic Editor, Information Organizers, Inc., P.O. Box 14391,
Research Triangle Park, NC 27709
Ms. Deborah L. Staves—Bibliographic Editor, Information Organizers, Inc., P.O. Box 14391,
Research Triangle Park, NC 27709
Ms. Patricia R. Tierney—Bibliographic Editor, ManTech Environmental Technology, Inc.,
P.O. Box 12313, Research Triangle Park, NC 27709
I-xxxv
-------
6
Toxicological Effects of Ozone
and Related Photochemical
Oxidants
6.1 Introduction
A wide range of effects of ozone (O3) has been demonstrated in laboratory animals
(see reviews by U.S. Environmental Protection Agency [1986], Lippmann [1989, 1993], and
Graham et al. [1991]). The major research findings are that environmentally relevant levels of
O3 cause lung inflammation; decreases in host defenses against infectious lung disease; acute
changes in lung function, structure, and metabolism; chronic changes in lung structure and
lung disease, some elements of which are irreversible; and systemic effects on target organs
(e.g., liver, immune system) distant from the lung. The research also has served to expand
understanding of the mechanisms of toxicity and relationships between concentration (C) and
duration of exposure (time [T]). The framework for presenting the health studies of O3 in
animals begins with a discussion of respiratory tract effects and is followed by a presentation
of systemic effects and interaction of O3 with other common co-occurring pollutants.
Respiratory tract effects are often interrelated; however, for purposes of presentation, effects
on lung inflammation and permeability, host defenses, morphology, pulmonary function,
biochemistry, and mutagenic/carcinogenic potential are discussed separately in the main text,
drawing correlations where appropriate. This type of organization enables focus on specific
effect categories. In the few cases where one study addresses several different categories of
endpoints, cross references are made to the appropriate sections of the chapter. Each major
section on a specific effect category is followed by a summary for that section. The summary
and conclusions section for the entire chapter (Section 6.5) attempts to draw together findings
on related endpoints and to highlight key issues, such as the relative importance of exposure
concentrations and durations and the identification of potential risk factors.
A purpose of this criteria document is also to describe any key health effects of
photochemical oxidants in addition to O3. Nitrogen dioxide (NO2) and nitric oxide are the
other two primary photochemical oxidants; they have been evaluated recently in another
criteria document (U.S. Environmental Protection Agency, 1993). Formaldehyde (HCHO),
which is formed photochemically and can be toxic, also has been reviewed recently by the
U.S. Environmental Protection Agency (EPA) (Grindstaff et al., 1991). Literature searches
did not reveal any animal toxicology inhalation studies of peroxyacetyl nitrate (PAN) since the
last O3 document (U.S. Environmental Protection Agency, 1986). A myriad of other
6-1
-------
individual photochemical oxidants are formed in ambient air (Chapter 3), but they have not
been investigated by animal inhalation toxicology. The very few publications on the effects of
exposures to a mixture of oxidants are summarized in Section 6.4, which discusses pollutant
interactions. Therefore, other than in Section 6.4, this chapter does not address other
photochemical oxidants. Even so, considering the limited literature within the aforementioned
documents, the available evidence from animal toxicology studies shows that O3 is the most
potent of the oxidants for noncancer effects at environmentally relevant concentrations.
The animal toxicology database for O3 is extremely large, making it necessary to
adopt conventions for presenting succinctly the pertinent findings. Priority was placed on
analysis of research published after closure of the previous O3 criteria document (U.S.
Environmental Protection Agency, 1986); however, for the purposes of broader interpretation,
the older literature is very briefly summarized. Generally, only the highlights of the key
recent studies and their interpretation are provided here. Confirmatory recent studies are
mentioned and presented in the tables. Furthermore, studies having O3 concentrations Dl.O
ppm are highlighted with rare exception (e.g., genotoxicity studies). Genotoxicity studies at
O3 concentrations higher than 1.0 ppm were included to enable coverage of all the specific
endpoints, some of which were tested only above 1.0 ppm. In most other cases, however, the
1-ppm cut point allows portrayal of the full array of the effects of O3 that may occur from
ambient air exposure and also avoids the potential for confounding mechanisms that can occur
at very high, environmentally unrealistic concentrations. For example, very high levels of
O3 can cause severe pulmonary edema, resulting in types and magnitudes of pulmonary
function changes that would not occur in ambient air. In summarizing the literature, changes
from control are described if they were statistically significant at p < 0.05, rather than citing
the probability values for each study. Where appropriate, critique of a statistical procedure is
mentioned. A probability value is provided if it aids the understanding of trends observed in a
study (e.g., p < 0.1).
As stated above, only literature published since the last O3 criteria document is
described in detail here. The earlier findings are summarized to facilitate cross-referencing.
For example, in some cases, the older work is presented in overview in the beginning of each
main section; in other cases, the overview is at the subsection level. Generally, the newer
literature elucidates the influence of different exposure regimens and the mechanisms of several
key effects, rather than portrays undiscovered categories of effects. The newer knowledge on
molecular and biochemical interactions increases the understanding of mechanisms of effects.
For example, it is unlikely that the O3 molecule itself penetrates the lung and enters the
circulation. As another example, the relationship between inflammation measured in tissue and
lung lavage assists in the interpretation of lung lavage findings. Information on the immune
system suggests that the cell-mediated limb may be more susceptible than the humoral limb.
The ability of O3 to decrease antibacterial host defenses has long been recognized, but only
recently have viral defenses been analyzed. Much remains to be learned, but apparently
antibacterial defenses are more at risk from O3 exposure than antiviral defenses. Cellular and
interstitial changes in the lungs of O3-exposed animals were among the very early studies, with
newer work adding to a detailed understanding of morphologic lesions in the pulmonary region
identified through advanced morphometric procedures. Scientists just recently have begun to
study the effects of O3 on the nose and have discovered epithelial changes, identifying this
tissue as a significant target site of O3. One new body of information concerns the influence of
exposure concentration, duration, and pattern. For several endpoints (e.g., increased lung
6-2
-------
permeability), under acute exposure conditions, concentration has more impact than exposure
duration. The importance of exposure duration is clearly illustrated by the newer chronic
studies that show different patterns of effects (compared to acute exposures). The most
sensitive indicators appear to be morphological changes (compared to pulmonary function
changes) which is consistent with the concept that functional abnormalities follow
morphological changes and may not become apparent until a given threshold is achieved.
Studies using intermittent exposures (e.g., exposures every day versus every other week or
month for equal times) indicate that interrupted exposures can produce equal or, in some cases,
enhanced effects compared to uninterrupted exposures, suggesting a cumulative effect. Thus,
seasonal "lows" in O3 do not have benefit in these animal studies. Newer work also has
mimicked and extended human clinical studies of repeated exposures. As with humans, the
pulmonary function of rats was attenuated with several days of exposure. However, other
changes (e.g., cellular) did not attenuate in the rat, illustrating the need for comprehensive
evaluations. For the first time, a classical cancer bioassay has been performed with O3. It
helps put some of the earlier genotoxicity and carcinogenicity studies in perspective. This
brief identification of the newer additions to the O3 database is not meant to be a summary of
effects; that is the last section of the chapter. Rather, it does show the importance of
considering all the literature, not just the newer work in interpreting the effects of O3. As
mentioned, it was not feasible to repeat the 1986 O3 criteria document herein; this makes it
necessary to use both the current and the former document in evaluations.
Animal toxicological studies of O3 are of major interest because they illustrate a
fuller array of effects and exposure conditions than can be investigated in humans. Most
experts accept a qualitative animal-to-human extrapolation (i.e., O3 effects observed in several
animal species can occur in humans if causative exposure concentrations, durations, and
patterns also occur). However, there is less consensus on an approach to quantitative
extrapolation (e.g., the exposures at which effects in animals actually occur in humans).
Chapter 8, on extrapolation, provides more information on this topic.
6.2 Respiratory Tract Effects of Ozone
6.2.1 Biochemical Effects
6.2.1.1 Introduction
This section outlines studies designed to identify biochemical targets of O3, as well
as biochemical measurements of antioxidant and microsomal enzyme activities, lipids, and
proteins. It should be noted that interpretation of biochemical changes resulting from whole
lung measurements is complicated by the heterogeneity in cell type and function present in lung
tissue and the changes in cell populations that result from O3-induced inflammatory cell
infiltration and epithelial cell and fibroblast proliferations. The ability to extrapolate from in
vitro to in vivo studies and from high to low levels of O3 is further complicated by an inability
to detect biochemical changes in the whole lung when only a small proportion of the lung may
be affected by O3, especially at concentrations of O3 less than 1 ppm. Interpretation of all
biochemical measurements, therefore, needs to take into account the airway sites of O3
interaction and concomitant changes in cell populations and numbers that take place at times
other than the onset of exposure.
6.2.1.2 Cellular Targets of Ozone Interaction
6-3
-------
In vitro experiments have indicated that O3 has the potential to interact with a wide
range of different cellular components that include polyunsaturated fatty acids (PUFAs); some
protein amino acid residues (cysteine, histidine, methionine, and tryptophan); and some
low-molecular-weight compounds that include glutathione (GSH), urate, vitamins C and E, and
free amino acids (U.S. Environmental Protection Agency, 1986; Mustafa, 1990; Pryor, 1991,
1992). The mechanisms to explain the initial biochemical and physiological effects of
O3 exposure in vivo are therefore complex. Hypotheses have been developed on the direct
action of O3 with lung macromolecules, the reaction of secondary biochemical products that
could result from the generation of free radical-precursor molecules, the release of endogenous
mediators of physiological response, and the reactive oxygen intermediates and proteinases
associated with the activities of inflammatory cells that subsequently infiltrate into O3-damaged
lungs (see Section 6.2.2). Based on some theoretical calculations, Pryor (1992) hypothesized
that, because O3 is so reactive, it most likely does not penetrate beyond the surface-lining
fluids of the lung except in those terminal airway regions having minimal lining thickness
where epithelial cells may well be relatively unprotected by either mucus or surfactant. In a
review, Pryor (1991) proposed that O3-induced cell damage more likely results from the
reactions of more stable but less reactive ozonide, aldehyde, and hydroperoxide products of
O3 interaction with surface-lining fluid components than from direct interactions of O3 with
intracellular components. Although the alveolar lining fluid is relatively rich in saturated
phospholipids, it does contain some lipids with unsaturated fatty acids, cholesterol, a protein A
component, and small-molecular-weight compounds (e.g., GSH and uric acid) that have been
shown to react with O3 in both in vitro and in vivo studies (Effros et al., 1990; King and
Clements, 1985; Shelley et al., 1984).
Polyunsaturated Fatty Acids
Hitherto, the major products of O3-lipid interaction that account for cell membrane
damage have been assumed to be lipid hydroperoxides. However, evidence for the production
of hydrogen peroxide and aldehydes has been demonstrated. It has been proposed that,
although Criegee ozonation (Figure 6-1) will ultimately lead to the production of ozonides in a
lipophilic environment, in the aqueous environment of lung airways, the carbonyl oxide
intermediate can form a hydroxyhydroperoxy compound, which on elimination of hydrogen
peroxide yields another aldehyde or, in the presence of iron ions, can form an aldehyde and the
very reactive hydroxyl radical (Teige et al., 1974; Pryor, 1991). Ozonation of aqueous
emulsions of PUFAs, rat erythrocyte ghost membranes, and rat bronchoalveolar lavage (BAL)
fluid has shown hydrogen peroxide and aldehyde generation with a much smaller proportion of
ozonides and lipid hydroperoxides (Pryor et al., 1991). A mechanistic study by Santrock et al.
(1992) of the ozonation of l-palmitoyl-2-oleoyl-5«-glycero-3-phosphocholine in unilamellar
phospholipids confirmed the generation of the hydroxyperoxy compounds, which subsequently
result in the generation of hydrogen peroxide and aldehydes with further oxidation to
carboxylic acids (Figure 6-1). Similar studies conducted under nonaqueous conditions have
demonstrated the production of secondary ozonides that, under physiological conditions, would
be expected to decompose rapidly to reactive products (Lai et al., 1990). Madden et al. (1993)
have demonstrated recently production of arachidonate-derived aldehydic substances and
hydrogen peroxide from in vitro O3 exposure (0.1 and 1.0 ppm for 1 h) of arachidonate in both
a cell-free system and
6-4
-------
d°b
RHC=CH— + Q —*• RHC—CH » RHC=O—O + RHC=0
PUFA ozone trioxolane carbonyl oxide aldehyde
either in XO—Q or in the
the _^ RHC CH—Presence _^ RCH _^ RHC=O + HO
absence NOX of HO V^u
ofH O 2 OOH
2 Criegee ozonide hydroxyhydroperoxy cpd. aldehyde hydrogen peroxide
Figure 6-1. Major secondary products of ozone interaction with lung cells.
cultured human bronchial epithelial cells. Ozonides, aldehydes, hydrogen and lipid peroxides,
and related reactive oxygen intermediates, together with the phospholipid from which the
aldehyde has been removed, represent major products of O3 interaction with lung cells that all
have the potential to cause damage to membranes (see Figure 6-1).
Evidence that interaction of O3 with PUFAs takes place in vivo has not been so
easily obtained. Goheen et al. (1986) investigated the effects of fat-free diets on rats exposed
to air or to 0.96 ppm O3 for 0, 1,2, and 4 weeks and concluded that O3 does not oxidize
significant levels of the PUFAs linoleate (18:2) and arachidonate (20:4). However, cleavage
of lung fatty acid double bonds has been demonstrated in an in vivo study reported by
Rabinowitz and Bassett (1988) that involved rat exposures for 4 h to 2 ppm O3. These authors,
by using hydrogen peroxide treatment to convert ozonides and aldehydes to carboxylic acids,
were able to demonstrate O3-induced increases in glutaric and nonanoic acids that are the
ozonolysis breakdown products of lung tissue arachidonic and oleic acids, respectively. More
recent studies directed towards developing suitable biomarkers and dosimeters for O3 exposure
have analyzed rat BAL lipids after a 12-h exposure to 1.3 ppm and demonstrated the
appearance of the aldehydes nonanal and heptanal (Cueto et al., 1992). Pry or et al. (1992)
also have been able to identify cholesterol ozonation products extracted from whole lung tissue
with the same exposure of rats to 1.3 ppm O3 for 12 h.
Evidence of the role of hydrogen peroxide in O3-induced lung damage has been
described by Warren et al. (1988), who demonstrated diminished O3-induced increased BAL
protein in rats after 1 day of exposure to 0.64 ppm O3, when treated with the hydrogen
peroxide scavenger dimethylthiourea before exposure. Hitherto, the exhalation of ethane and
pentane and tissue measurements of diene-conjugates and thiobarbituric acid reactive
substances (TEARS) have been used as evidence for O3-induced free radical autoxidation of
lipids (U.S. Environmental Protection Agency, 1986). However, these measurements have
been found to be relatively insensitive for use in inhalation experiments under conditions of
low O3 concentrations (<0.5 ppm). Ichinose and Sagai (1989) were unable to demonstrate any
changes in lung TEARS as a result of exposing rats for 2 weeks to 0.4 ppm O3. As noted by
6-5
-------
Pry or (1991), malondialdehyde and other thiobarbituric-acid-reacting aldehydes can be
produced by Criegee ozonation of olefmic fatty acids that include arachidonate, as well as by
free radical peroxidative processes. In addition, malondialdehyde, being volatile as well as
highly reactive, may be lost readily from the lung or during sample preparation. However,
measurements of TEARS continue to be used for in vitro experiments designed to demonstrate
possible mechanisms by which such agents as taurine (Banks et al., 1991) and uric acid
(Meadows and Smith, 1987; Peden et al., 1993) may protect against O3-induced lipid damage.
Rietjens et al. (1987b), by preincubating rat alveolar macrophages (AMs) with either
arachidonate (20:4) or phosphatidylcholine to alter PUPA content and membrane fluidity,
respectively, demonstrated that PUPA content (not membrane fluidity) determined sensitivity
to O3 damage, measured as decreases in phagocytic activity.
Pvidence for free-radical-mediated autoxidation comes indirectly from the
demonstration that vitamin P depletion increases O3 toxicity, as reported previously
(U.S. Pnvironmental Protection Agency, 1986) and more recently (Plsayed, 1987; Plsayed
et al., 1988). More direct evidence for free radical generation has been obtained using
electron spin-trapping technology that correlated increased radical signals in isolated lung
lipids from rats exposed to increasing O3 concentrations (0 to 1.5 ppm, effect beginning at
about 0.5 ppm; 2 h) under conditions of carbon dioxide (CO^-stimulated respiration (Kennedy
et al., 1992). However, the possible contribution of activated inflammatory cell generation of
reactive oxygen intermediates to these observed free radical alterations to lung lipids needs to
be considered.
Antioxidants
Although vitamin P directly reacts with O3 at the same rate as PUP As, vitamin C
appears to react more effectively (Pryor, 1991), which, together with intracellular taurine
(Banks et al., 1991) and BAL uric acid (Meadows and Smith, 1987; Peden et al., 1993) found
in nasal and lung-lining fluids, may act as direct scavengers of O3. Ozone-induced increases in
lung polyamine metabolism in vitamin P-deficient rats suggests their possible role as
antioxidants (Plsayed, 1987). Glutathione in its reduced form (GSH) represents another
potential direct O3 scavenger. In addition to being a major intracellular antioxidant, GSH is a
component of airway-lining fluids found in BAL. Ozone would have to penetrate the cellular
membrane without reaction if it is to directly interact with intracellular GSH, an event
considered to be unlikely (Pryor, 1991, 1992). Previously observed oxidation of glutathione
and, in some cases, its loss from the lung may more likely reflect its reaction with an O3-
derived oxidant, such as a hydroperoxide or an ozonide, mediated by glutathione peroxidase
(GSHPx) and glutathione-S-transferases, respectively (Rietjens et al., 1987a), resulting in the
formation of glutathione disulphide or mixed disulphides with sulphydryl (SH)-containing
proteins. Although ozone-induced formation of glutathione sulfonate has been reported in
vitro, such irreversible oxidation of GSH has not been observed in vivo (U.S. Pnvironmental
Protection Agency, 1986; Mustafa, 1990).
Proteins
Parly studies reported that nonprotein sulfydryls (NPSHs) and the activities of
various cytosolic, microsomal, and mitochondrial enzymes are decreased immediately
following short-term exposures to relatively high levels (2 to 4 ppm) of O3 (U.S.
Pnvironmental Protection Agency, 1986; Mustafa, 1990). However, although these early
6-6
-------
biochemical effects could not be demonstrated after the first day of exposure to the lower
O3 concentration of 0.8 ppm, the methods employed may not have been sensitive enough to
detect coenzyme and enzyme changes in the centriacinar region (CAR), which is a primary
target of O3. However, together with surfactant lipids, the surfactant protein A also has been
examined as a potential target of O3 interaction (Oosting et al., 1991c, 1992). In vitro studies
by these authors have suggested that either hydrogen peroxide- or O3-induced oxidation of
methionine and tryptophan residues account for the observed changes in physicochemical
properties of canine and human surfactant protein A, measured as an impairment of self-
association and a decreased ability to aggregate phospholipid vesicles and to bind mannose
(Oosting et al., 1991c). Similar responses were found in vivo; surfactant isolated from rats
exposed for 12 h to 0.4 ppm O3 was less able to stimulate AM superoxide anion generation
than surfactant obtained from air-exposed control rats (Oosting et al., 1992). The previously
reported presence of giant lamellar bodies in O3-exposed rat lungs following exposure to
0.3 ppm for 3 h/day for 16 days is also consistent with the hypothesis that O3 reacts with
surfactant protein A (Shimura et al., 1984) and thereby interferes with its homeostatic role in
surfactant release from alveolar Type 2 cell lamellar bodies and its subsequent reuptake by
Type 2 cells and AMs.
6.2.1.3 Effects of Ozone Exposure on Lung Lipid Metabolism
Arachidonate Metabolites
Ozone-induced damage to airway epithelia (Leikauf et al., 1988) and AMs (Madden
et al., 1991) in vitro has been associated with the production of arachidonic acid metabolites by
both cyclooxygenase and lipoxygenase pathways. These metabolites have been implicated in a
variety of different physiological processes that include changes in airway permeability,
infiltration of polymorphonuclear leukocytes (PMNs) and eosinophils, and airway smooth-
muscle reactivity, discussed elsewhere in this chapter (Sections 6.2.2 and 6.2.5). Leikauf
et al. (1993) have examined the effects of fatty acid O3-degradation products on human airway
epithelial eicosanoid metabolism and concluded that the stimulating effects were increased with
product chain length, with the 3-, 6-, 9-hydroxyhydroperoxides being more potent than their
corresponding aldehydes. Madden et al. (1993) concluded that aldehydic degradation products
of arachidonate, but not hydrogen peroxide, increased in vitro polarization of leukocytes, and
decreased peripheral blood T-cell mitogenesis and natural killer (NK) cell cytotoxicity. In vivo
experiments on rabbits, guinea pigs, mice, and rats of different ages exposed to D 1.0 ppm
O3 have demonstrated increases in the products of arachidonic acid metabolism (see
Section 6.2.2).
Surfactant
Although alveolar surfactant lipids purified from lavage fluids have been shown to
be relatively enriched in saturated lipids, a varying percentage of the lipids do contain
unsaturated fatty acids depending on the species studied (King and Clements, 1985; Shelley
et al., 1984). These unsaturated lipids, together with the apoprotein as possible targets
of O3 interaction, may be expected to have an altered composition as a result of O3 inhalation.
However, surfactant-enriched material isolated by BAL from rats following an 8-h exposure to
0.8 ppm O3 retained its ability to lower surface tension in spite of an increase in protein
content (Nachtman et al., 1986). In long-term exposure studies, monkeys were exposed for 8
h/day to 0.15 and 0.3 ppm O3 for 21 and 90 days (Rao et al., 1985a,b). In contrast to
6-7
-------
measurements of total lung lipids that demonstrate a relative decrease in PUFAs after 21 days
of exposure (Rao et al., 1985a), there was a relative increase in the proportion of PUPA in the
percentage of BAL unsaturated fatty acids (increases from 34% in air controls to 41, 42, and
45% in BAL lipids recovered from monkeys exposed for 21 days to 0.15 ppm, 90 days to
0.15 ppm, and 90 days to 0.3 ppm O3, respectively) (Rao et al., 1985b). The major increases
were observed in linoleate (18:2) and arachidonate (20:4). Because these PUFAs are potential
targets for O3 interaction, their increase, rather than a decrease, in BAL fluid may best be
explained by changes in surfactant lipid production associated with alveolar Type 2 epithelial
proliferation (Section 6.2.4). Interestingly, a relative decrease in cholesterol ester with a
concomitant increase in phosphatidylcholine was observed, which supports the hypothesis that
cholesterol may represent a major target of O3 interaction (Rao et al., 1985b; Pryor et al.,
1992). The observed O3-induced changes in BAL PUFA composition were consistent with
those previously reported for rats by Roehm et al. (1972), but only for BAL lipids isolated
from vitamin E-depleted rats following 6 weeks of exposure to 0.5 ppm O3. Wright et al.
(1990) were unable to detect changes in BAL lipid and fatty acids recovered from normally fed
rats following 0.12-, 0.25-, and 0.5-ppm O3 exposures for 20 h/day for 18 mo. Results from
these studies are summarized in Table 6-1.
Tissue Lipids
In vivo pulse labeling with carbon-14-labeled acetate was used to estimate
phospholipid biosynthesis (Wright et al., 1990). Although found to be diminished at certain
time points (3 and 12 mo), no consistent trend could be demonstrated that would suggest that
O3 exposures of less than 0.5 ppm alter lung surfactant homeostasis. Bassett and Rabinowitz
(1985), using isolated perfused lungs taken from rats after 3 days of continuous exposure to
0.6 ppm O3, demonstrated an enhanced incorporation of glucose carbons into both fatty acid
and glycerol-glyceride moieties of total lung lipids by 180 and 95%, respectively. The relative
increase in carbon incorporation into free fatty acids, phosphatidic acid, phosphatidyl inositol,
and sphingosine containing lipids was consistent with the needs of a dividing cell population
for increased lipids synthesis associated with alveolar epithelial proliferative repair. It should
be noted that, in a separate study, under the same exposure conditions of 0.6 ppm O3 for
3 days, rat lungs demonstrated increased glycolytic activity and generation of reduced
nicotinamide adenine dinucleotide phosphate (NADPH) consistent with the energy and
synthetic needs of a lung undergoing repair of O3-induced damage (Bassett and Bowen-Kelly,
1986). Results from these studies are summarized in Table 6-1.
6.2.1.4 Effects of Ozone on Lung Antioxidant Systems
The O3-induced increased levels of the antioxidant NPSHs, identified mainly as
GSH in the lung, and the enzyme activities involved in GSH utilization, GSHPx and
glutathione-5-transferase (GST), and for maintaining GSH in a reduced state, glutathione
reductase (GR) and the NADPH-linked dehydrogenases of glucose-6-phosphate (G6PD) and
6-phosphogluconate (6PGD), typically have been attributed to concurrent morphological
changes rather than to any specific biochemical response (U.S. Environmental Protection
Agency, 1986). Numerous studies conducted in mice, rats, and monkeys show increases in
many of these enzyme activities at exposures as low as 0.2 ppm O3 for 1 week (rat) (U.S.
Environmental Protection Agency, 1986). The earlier research also included studies of age-
dependent responsiveness of rats (Tyson et al., 1982; Lunan et al., 1977; Elsayed et al.,
6-8
-------
1982). Rats ranging in age from 5 to 90 days old were exposed to 0.8 or 0.9 ppm O3 for 3 or
4 days or for about 20 days, depending on the experiment. Ozone altered activities of
antioxidant enzymes in an age-dependent manner. Generally, prior to weaning, enzyme
6-9
-------
Table 6-1. Effects of Ozone Exposure on Lung Lipid§
Ozone
Concentration
ppm
0.12
0.25
0.5
0.15
0.3
0.15
0.3
0.5
0.6
0.5
1.0
1.5
2.0
0.58
0.6
0.8
Qg/m
235
490
980
353
588
353
588
980
1,176
980
1,960
2,940
3,920
1,137
1,176
1,568
Exposure
3 Duration
20 h/day,
7 days/week
for 18 mo
8 h/day
for 90 days
8 h/day
for 21 and
90 days
Continuous for
0-4 weeks
2h
Continuous for
3 days
Continuous for
3 days
18 h
Species, Sex
(Strain)
Age1'
Rat, M
(F344)
28 days old
Monkey
(Bonnet)
Monkey
(Bonnet)
Rat, M
(S-D)
50 g
Rat, M
(CD)
65-85 days old
Rat, M
(Wistar)
220-250 g
Rat, M
(Wistar)
220-250 g
Rat
(F344)
260 g
Observed Effect(s)
Age-related increase in BAL and tissue phospholipids generally unaffected by O3 exposure; at 0.5 ppm, total
phospholipid increased at 6 and 12 mo.
Fraction of total lung lipid fatty acids that were PUFAs decreased from 22 to 9% and 6% following 0. 15-ppm and
0.3-ppm exposures, respectively.
BAL PUFAs (linoleate [18:2] and arachidonate [20:4]) increased, with a relative decrease in cholesterol esters.
No change in lung fatty acid content; no acceleration of essential fatty acid deficiency in rats on fat-free diet.
Extracted lung lipid EPR signal intensity proportional to O3 concentration following pretreatment with spin-trapping
agent and CO, stimulation of respiration in vivo.
Increased lipid synthesis associated with increased glucose catabolism for ATP and NADPH generation.
Increased synthesis of perfused lung glyceride-glycerol and fatty acid moieties of neutral lipids and phospholipids
from glucose carbons, with a greater proportion of shingosine and inositol synthesis.
Increase in BAL protein; no alteration in surface-tension-lowering ability of BAL.
Reference
Wright et al. (1990)
Rao et al. (1985a)
Rao et al. (1985b)
Goheen et al. (1986)
Kennedy et al. (1992)
Bassett and Bowen-Kelly
(1986)
Bassett and Rabinowitz
(1985)
Nachtman et al. (1986)
"See Appendix A for abbreviations and acronyms.
''Age or body weight at start of exposure.
-------
activities decreased, and, at older ages, they increased. The reasons for these differences are
not known, but may be due to differences in (1) dose of O3 to the lung (due to differences in
exposure concentrations in huddled neonates on bedding prior to weaning or to differences in
doses delivered to lung target sites), (2) basal levels of antioxidants and antioxidant enzymes,
or (3) cellular sensitivity. Increased lung enzyme activities can result from either increased
activity within a particular cell population or increased numbers of cells with that activity.
Age, nutritional, and species differences in O3-mediated responses must therefore be
interpreted with consideration of the underlying morphological changes (Section 6.2.4).
Relevant studies are summarized in Table 6-2.
An increase in lung alveolar Type 2 cells and in infiltrating inflammatory cells
adequately explained the observed increases in succinate oxidase, G6PD, and 6PGD activities
observed after 3 days of continuous exposure of rats to 0.75 ppm O3 when represented on a
per-milligram-deoxyribonucleic-acid (DNA) basis (Bassett et al., 1988a). These cell types are
enriched in mitochondria and in NADPH-generating capacity, needed for both lipid
biosynthesis and GSH maintenance. Similarly, no significant changes in these enzyme
activities could be detected after 3 days of exposure to the lower O3 concentration of 0.35 ppm,
further illustrating the need to take into account the concomitant changes in cell population and
number when interpreting whole-lung enzyme measurements. Increases of 150 and 108%,
respectively, were observed in the per-milligram DNA activities of the ornithine carboxylase
and 5-adenosyl-methionine decarboxylase enzymes involved in polyamine synthesis, which,
together with enhanced tritiated thymidine incorporation into DNA, have been considered to be
more sensitive measures of biochemical changes in lungs of rats exposed continuously for
3 days to 0.45 ppm O3 (Elsayed et al., 1990).
The potential role of superoxide dismutase (SOD) and catalase in protecting the lung
against O3 toxicity is not clear. Bassett et al. (1989), using a pretreatment with a phenyl-urea
compound (M2-(2-oxo-l-imidazolindinyl)ethyl]-AO-phenylurea; EDU) that increased rat lung
SOD and catalase activities, failed to demonstrate any protection against acute lung injury from
a single 3-h exposure to 2.0 ppm O3. However, Zidenberg-Cherr et al. (1991) have
demonstrated that copper (Cu)- and manganese (Mn)-deprived mice may be more susceptible to
continuous O3 exposure of 1.2 ppm for 7 days. Rahman and Massaro (1992) have
demonstrated protection against edemagenic exposures to ozone (2.5 ppm for 24 h) in rats
pretreated with endotoxin. Endotoxin pretreatment is associated with increases in lung tissue
mitochondrial Mn-SOD activity without any concomitant increases in catalase, GSHPx, and the
cytosolic Cu,Zn-SOD enzymes. Although it is difficult to conclude that mitochondrial SOD
might directly protect against O3 interactions, these results do suggest a central role of
mitochondrial SOD in the protection of the cell against oxidative stress (Rahman and Massaro,
1992).
Rahman et al. (1991) also have demonstrated that lungs from O3-exposed rats had
increased activities of Cu, Zn-SOD, Mn-SOD, catalase, and GSHPx after 5 days of exposure
to 0.7 ppm O3. These increases were attributed to enhanced gene expression, indicated by
higher messenger ribonucleic acid (mRNA) concentrations, rather than to the infiltration of
cells enriched with these enzyme activities. Chronic exposure of rats to an urban pattern of
O3 for 12 mo did not affect total SOD activity in rats, although GSHPx and GR activities per
lung were increased (Grose et al., 1989). Use of microdissection techniques following 90 days
and 20 mo of rat exposures to 0, 0.5, and 1.0 ppm O3 have shown concentration-dependent
6-11
-------
increases in SOD, GST, and GSHPx per milligram of DNA in the distal bronchioles.
In contrast, decreases in GST and GSHPx activities in major bronchi and
6-12
-------
Table 6-2. Effects of Ozone Exposure on Lung Antioxidantg
u>
Ozone
Concentration
ppm
0.06
base,
0.25
spike
0.12
0.2
0.64
0.12
0.5
1.0
0.35
0.75
0.4
0.41
0.45
0.5
Dg/m
118 base,
490 spike
235
392
1,254
235
980
1,960
686
1,470
784
800
882
980
Exposure
3 Duration
Base 13 h/day,
7 days/week;
ramped spike
9 h/day,
5 days/week for
12 mo
Continuous for
7 days
6 h/day,
5 days/week for
90 days or 20 mo
Continuous for
3 days
Continuous for
2 weeks
12 h during day or
night for 3 days or
continuous for 72 h
Continuous for
2 days
Continuous for
5 days
Species, Sex
(Strain)
Age"
Rat, M
(F344)
Rat, M
(S-D)
250-300 g
Rat, M and F
(F344)
Rat, M
(Wistar)
200-250 g
Rat, M
(Wistar)
6 weeks old
Guinea pig
(Hartley)
6 weeks old
Rat, M
(Wistar)
Guinea pig, M
(Hartley)
9 weeks old
Rat, M
(S-D)
90 days old
Rat
(Long-Evans)
10 weeks old
Observed Effect(s)
Whole lung increase in GSHPx and GSH reductase activities. SOD activity and NPSH content not affected.
Pretreatment with the H2O2 scavenger dimethylurea decreased O3 -induced tissue DNA and protein and BAL
protein, acid phosphatase, and A'-acetyl-D-D-glucosaminidase. No effect of vitamin E or D-carotene.
Using microdissection techniques and representing data as units/mg DNA, GST, GSHPx, and SOD were
increased in distal bronchioles after 90 days and 20 mo in a concentration-dependent fashion. After 90 days,
SOD and GST were lower in major daughter bronchi. After 20 mo, SOD was increased in distal trachea;
GSHPx was decreased in major bronchi but enhanced in minor bronchi; and GST decreased in major bronchi.
0.75 ppm O3-induced whole lung increases in GSHPx and GR not significant when corrected for increases in
cell number. Increases in succinate oxidase, G6PD and 6PGD activities per mg DNA were consistent with
increased Type 2 and inflammatory cell content. No increases per mg DNA at 0.35 ppm O3.
Small increases in whole rat lung levels of NPSH, vitamin C, GSHPx. Guinea pig GSHPx and GSH
transferase activities decreased.
Rats: No effect of daytime exposure. Nighttime or continuous exposure increased activities of LDH, G6PD,
GR, and GSHPx. Guinea pig: No daytime-only exposure. No effect on GR or GSHPx, G6PD increased after
nighttime or continuous exposure; lactate dehydrogenase activity increased only after continuous exposure.
Large increase in ornithine decarboxylase activity and DNA labeling reflecting polyamine metabolism and
DNA synthesis and/or repair, respectively.
Reference
Grose et al. (1989)
Warren et al. (1988)
Plopper et al. (1994b)
Bassett et al. (1988a)
Ichinose and Sagai (1989)
Van Bree et al. (1992)
Elsayed et al. (1990)
Ozone increased lung putrescine in both vitamin E-deficient or 1,000 Ill/kg groups, but increases in spermidine Elsayed (1987)
content and decarboxylase activities of ornithine and S-adenosylmethionione only in vitamin E-deficient group.
-------
Table 6-2 (cont'd). Effects of Ozone Exposure on Lung Antioxidant§
Ozone
Concentration
ppm
0.5
0.5
0.64
0.64
0.7
0.8
Dg/m
980
980
1,254
1,254
1,373
1,568
Exposure
3 Duration
Continuous for 5 days
2.25 h/day for 5 days
Continuous for 7 days
Continuous for 7 days
Continuous for
1-5 days
8 h/day for
2 mo
Species, Sex
(Strain)
Age"
Rat
(Long-Evans)
10 weeks old
Rat, M
(F344)
110 days old
Rat, M
(S-D)
3-5 weeks old
Rat, M
(S-D)
52 and 295 g
Rat, M
(S-D)
45, 80, and
300 g
Rat, M
(S-D)
2 mo old
Observed Effect(s)
Ozone increased lung vitamin E level in supplemented rats and remained unchanged in all other tissues measured.
Lung GSH initially enhanced, declining to control levels by Day 4. Lung ascorbate levels enhanced
and 5 only.
on Days 3
The whole lung O3-induced increase in ascorbate and GSH content unaffected by protein deficient diets.
Whole adult lung contents of Cu,Zn-SOD and GSHPx increased by O3 in all diet groups (ad libitum, 4-16%
protein diets); GSHPx only increased in weanling rats fed 16% protein diet. Mn-SOD only increased in lungs
from 4 and 16% protein-fed adult lungs.
By 5 days, increased lung Cu,Zn-SOD, Mn-SOD, catalase, and GSHPx per DNA in all age groups. Adult lungs:
Concomitant increases in mRNAs for Cu,Zn-SOD, catalase, and GSHPx without differences in mRNA stability.
Absence of vitamin E exacerbates O3-induced damage related to increases in whole lung levels of metabolic
enzymes. No additional amelioration by diet supplementation above 50 IU vitamin E.
Reference
Elsayed et al. (1990)
Tepper et al. (1989)
Dubick et al. (1985)
Heng et al. (1987)
Rahman et al. (1991)
Elsayed et al. (1988)
'See Appendix A for abbreviations and acronyms.
''Age or body weight at start of exposure.
-------
an increase in GSHPx were observed in minor bronchi after 20 mo of O3 exposure (Plopper
et al., 1994b). Rahman et al. (1991) concluded that changes in antioxidant enzyme activities in
some cases could be associated with alterations in cellular pathology (see Section 6.2.4),
whereas, in other cases, no correlation could be made even though the results were represented
on a per-milligram-DNA basis. The observed changes in antioxidant enzyme activities appear
to be site-specific and different at different airway locations. The response are concentration
dependent and altered by the length of O3 exposure (Plopper et al., 1994b).
Representing data on a per-gram, wet-lung basis, Ichinose and Sagai (1989)
demonstrated increases in lung NPSH, vitamin C, and GSHPx but observed no effect on
vitamin E levels, after continuous exposure of rats to 0.4 ppm O3 for 2 weeks. In contrast,
guinea pig lungs exhibited no changes in these antioxidant components when similarly
exposed. However, although using the higher concentration of 0.64 ppm for 7 days of
continuous O3 exposure, Dubick et al. (1985) demonstrated that whole lung content of
ascorbate and GSH was elevated, these changes were not significantly different when the data
were represented on a per-100-g, wet-tissue basis. Rat BAL analysis following a 12-mo
exposure to an urban pattern of O3 demonstrated decreased vitamin E and enhanced ascorbate
and protein levels (Grose et al., 1989). Because these antioxidants also have been shown to be
targets of ozone interaction, any observed increases in their steady-state level suggest an
increase in the ratio of production to degradation that could reflect either enhancement in
cellular functions in response to continued O3 exposure or alteration in the number of cells
associated with their production.
In order to demonstrate that dietary vitamin E reduces the effects of O3 exposure on
lung biochemical parameters, comparisons between vitamin E-depleted and -supplemented
diets have been used (U.S. Environmental Protection Agency, 1986) and reviewed by Pryor
(1991). Elsayed et al. (1988) fed rats a test diet containing 0 or 50 International Units (IU) of
vitamin E per kilogram for 2 mo prior to exposure to 0.8 ppm O3 for 8 h/day for 7 days.
Ozone exposure increased the whole-lung activities of mitochondrial, microsomal, and
cytosolic enzymes. Vitamin E deficiency alone had no significant effect on these lung enzyme
measurements, which were taken on a per-lung basis, but the addition of 50 IU vitamin E per
kilogram to the diet prior to O3 exposure diminished the observed O3-induced increases in
mitochondrial succinate cytochrome c reductase and GSHPx, microsomal NADPH cytochrome
c reductase, and cytosolic GSHPx and SOD observed in vitamin E-deficient rats by up to 50%.
Additional experiments using a relatively low range of vitamin E supplementation for short
time periods demonstrated that, although absence of vitamin E in the diet exacerbates the
effects of O3 on lung injury, the magnitude of a protective effect does not increase
proportionately with increased dietary vitamin E. These data support the conclusion that any
supplementation beyond the normal recommended daily allowance for vitamin E may not
necessarily provide humans with any additional protection against the effects of ambient
O3 exposure (Pryor, 1991). However, possible failure in these animal experiments to reach a
steady-state tissue level of vitamin E may have obscured protective effects.
6.2.1.5 Effects of Ozone on Lung Protein Metabolism
Exposure of rodents to DO.45 ppm O3 has been associated with increases in lung
collagen, collagen synthesis, and prolyl hydroxylase activity associated with fibrogenesis (U.S.
Environmental Protection Agency, 1986). These earlier studies showed an influence of
exposure pattern on the responses. When rats were exposed to 0.8 ppm O3 for 7 days, prolyl
6-15
-------
hydroxylase activity continued to increase, but hydroxyproline content plateaued about Day 3
of exposure and remained elevated 28 days after exposure ceased (Hussain et al., 1976a,b).
Last et al. (1984b) employed 90-day exposure regimens of rats to 0.96 ppm O3 that included
(1) a continuous 90-day exposure and (2) intermittent periods of 5 days (8 h/day) of O3 and
9 days of air, repeated seven times with a total of 35 O3 exposure days over a 90-day period.
Both groups had equivalent increases in lung collagen content. When durations were
decreased to 3 weeks, the continuous and intermittent (1 week O3, then 2 weeks air) regimens
resulted in equivalent increases in lung collagen. In nonhuman primates receiving 0.25 ppm
O3 daily or seasonally (every other month) for 18 mo, only the seasonal group had an increase
in collagen (Section 6.2.4, Tyler et al., 1988). Results from studies of lung protein
metabolism are summarized in Table 6-3.
More recently, Choi et al. (1994) examined the earliest time points from the onset
of continuous O3 exposure of rats to 1.0 ppm that caused alterations in extracellular matrix
protein gene expression. These authors demonstrated an early increase in lung fibronectin
mRNA at 2 days, which preceded an increase in Type I collagen mRNA observed at 4 days;
however, increased collagen content indicated by lung hydroxyproline content was not
significantly enhanced until after 7 days of exposure. Pickrell et al. (1987a) demonstrated
concentration-dependent decreases in antiproteinase activities in serum and lung tissue of rats
exposed to 0.5 and 1.0 ppm O3 for 48 h. Exposure to 1.0 ppm was accompanied by a
concomitant increase in inflammatory-cell-derived proteinases. A second study that examined
lung collagen metabolism and proteinolysis in rat lungs exposed to 0.57 and 1.1 ppm O3 for
19 h/day for 11 days suggested that collagen accumulation, in part, may result from decreased
collagen degradation (Pickrell et al., 1987b).
Chronic exposures of monkeys to 0.61 ppm 8 h/day for 1 year demonstrated
increased lung collagen content, even 6 mo postexposure (Last et al., 1984b). Further analysis
also has demonstrated that the collagen isolated from these O3-exposed lungs exhibited
abnormalities, as indicated by increased levels of the difunctional cross-link
dehydrodihydroxylysinonorleucine (DHLNL) and of the ratio of DHLNL to
hydroxylysinonorleucine (HLNL) (Reiser et al., 1987). Although collagen content remained
elevated, difunctional DHLNL and HLNL cross-link levels returned to normal by 6 mo
postexposure, whereas trifunctional mature cross-links (hydroxypyridinium) remained
elevated. These data suggest that structurally abnormal collagen is actively synthesized during
O3 exposure and that it becomes irreversibly deposited in the lungs.
Because O3-induced lung effects are multifocal by nature, it is reasonable that
changes in collagen content within the lung may not be easily detectable by measuring
alterations in whole lung hydroxyproline at earlier time points or in those experiments that
have used lower O3 concentrations. For example, Wright et al. (1988) calculated values for
the extent of lung collagen deposition using measured synthesis rates and concluded that 18 mo
of exposure (20 h/day) of rats to concentrations up to 0.5 ppm O3 did not change either
synthesis or accumulation of lung collagen. On the other hand, Chang et al. (1992)
demonstrated sustained thickening of rat lung extracellular matrix on long-term exposure to a
simulated urban pattern of O3 exposure (baseline of 0.06 ppm, 7 days/week, with a slow rising
peak for 9 h/day, 5 days/week to 0.25 ppm) of up to 38 weeks. More recently, Last et al.
(1993a, 1994) observed excess stainable collagen in the lung CAR of rats exposed to 0.5 and
1.0 ppm O3 for 6 h/day, 5 days a week for 20 mo. Biochemical analyses demonstrated slight
6-16
-------
but significant increases in collagen with relatively more hydroxylysine-derived cross-links in
female but not male rats, when compared with age-matched,
6-17
-------
Table 6-3. Effects of Ozone Exposure on Lung Protein!
CO
Ozone
Concentration
ppm
0.12
0.25
0.5
0.125
0.25
0.5
0.12
0.5
1.0
0.25
0.4
0.5
0.5
1.0
1.5
0.57
1.1
0.61
Dg/m
235
490
980
245
490
980
235
980
1,960
490
784
980
980
1,960
2,940
1,117
2,156
1,196
Exposure
3 Duration
20 h/day for
18 mo
1 year
6 h/day
5 days/week
20 mo
8h/day
7 days/week,
"daily" for
18 mo or
"seasonal"
O3 odd months for
18-mo period (9
mo of O3)
12 h
4 h/day for
2 days
and 6 weeks
Continuous for
48 h
19 h/day for
1 1 days
8 h/day
for 1 year
Species, Sex
(Strain)
Age1'
Rat, M
(F344)
28 days old
Rat
(F344)
6 weeks old
Rat, M and F
(F344)
4-5 weeks old
Macaco
fascicularis
6 mo old
Rat, M
(Wistar)
8 weeks old
Sheep, F
23-41 kg
Rat, F
(F344)
12-14 weeks old
Rat, F
(F344)
120-180 g
Monkey
(Cynomolgus)
6-7 mo old
Observed Effect(s)
Age-related increases in hydroxyproline content as a measure of collagen were unaffected
by O3 exposure.
No changes in collagen content, increased turnover at DO. 25 ppm after 3 or more mo of
exposure.
Excess stainable collagen in the CAR at DO. 5 ppm. Biochemical analysis demonstrated
slight but significant increases in collagen in female but not male rats exposed to DO. 5 ppm
with increased hydroxylysine-derived cross-links.
Increased collagen content in seasonal group only.
Surfactant less able to stimulate AM superoxide anion generation, confirming in vitro
results suggesting damage to surfactant protein A.
At 2 days, increased sulfated glycoproteins secretion; at 6 weeks, diminished tracheal
mucosal gland hyperplasia secretion.
Concentration-dependent decrease in antiproteinase activity at 0.5 and 1.0 ppm. Increases
in acid proteinase activity 1.0 and 1.5 ppm correlated with increased inflammatory cell
content.
After 1 1 days of 1 . 1 ppm O3, inflammatory cell infiltrate and Type 2 cell and fibroblast
proliferation, increased cathepsin D and AM elastase activity, decreased rate of intracellular
collagen degradation, and increased extracellular matrix collagen turnover (indicated by
enhanced BAL hydroxyproline). These changes preceded increased collagen content
observed 50 days PE.
Increased lung collagen content associated with elevated abnormal cross-links that were
irreversibly deposited.
Reference
Wright et al. (1988)
Filipowicz and
McCauley (1986b)
Last et al. (1993a, 1994)
Tyler et al. (1988)
Oosting et al. (1992)
Phipps et al. (1986)
Pickrell et al. (1987a)
Pickrell et al. (1987b)
Reiser etal. (1987)
-------
Table 6-3 (cont'd). Effects of Ozone Exposure on Lung ProteinS
Ozone
Concentration Species, Sex
Exposure (Strain)
ppm Dg/m 3 Duration Age1' Observed Effect(s) Reference
0.8 1,568 Continuous for Rat Increased lung dry weight, protein, and collagen synthesis greatest after 60 days of age.
3 days (S-D)
24-365 days old
0.8 1,568 6 h/night for up to Rat, M Increased lung content of collagen. No change in hydroxypyridinium or elastin. Last et al. (1993b)
90 days (S-D)
10-12 weeks old
1.0 1,960 Continuous for Rat, M Lung mRNAs for c-myc proto-oncogene and fibronectin enhanced on Day 2 and Type I collagen mRNA not Choi et al. (1994)
14 days (Wistar) increased until Day 4, preceding increases in collagen hydroxyproline observed on Day 7.
200-250 g
"See Appendix A for abbreviations and acronyms.
''Age or body weight at start of exposure.
-------
air-exposed control animals. It should be noted that no excess of mRNA for Type I procollagen
was observed by in situ hybridization in lungs of rats exposed to 1.0 ppm for 20 mo, although
increases after a 2-mo exposure under similar conditions did indicate some increase expression
of this mRNA in alveolar interstitial cells (Last et al., 1993a).
Ozone exposure also affects airway secretion of mucous glycoproteins. After
2 days of exposure of sheep to 0.5 ppm O3, with subsequent evaluation of tracheal sulfated
glycoprotein and ion fluxes in vitro, there was an increase in basal secretion that was
associated with a moderate hypertrophy of lower tracheal submucosal glands (Phipps et al.,
1986). Although 7 days of exposure resulted in hypertrophy of upper and lower tracheal
submucosal glands, glycoprotein secretion was reduced, but chloride secretion was increased,
which can be explained by a relative decrease in gland mucous content.
6.2.1.6 Effects of Ozone Exposure on Lung Xenobiotic Metabolism
Previous studies have demonstrated that exposure to 0.75 to 1.0 ppm O3 for a few
hours diminishes microsomal cytochrome P-450 content and decreases the activities of
benzo[a]pyrene hydroxylase and benzphetamine ./V-demethylase of lungs isolated from several
different experimental animal species (U.S. Environmental Protection Agency, 1986). Because
bronchiolar Clara cells and alveolar Type 2 cells are considered to be relatively enriched with
microsomal cytochrome P-450 enzyme systems, it is reasonable that damage and subsequent
proliferative repair of these cell types would be expected to change the lung's capacity to
conduct xenobiotic metabolism. In a series of rat studies, Takahashi et al. (1985) and
Takahashi and Miura (1985, 1987, 1989, 1990) have demonstrated that, although intermittent
exposure of 0.4 ppm O3 for 7 h/day for 14 days did not affect microsomal metabolism,
increasing the concentration to 0.8 ppm (Takahashi et al., 1985) or exposing the rats
continually to 0.2 and 0.4 ppm for 14 days (Takahashi and Miura, 1985) increased cytochrome
P-450 content and the activities of cytochrome P-450 reductase, benzo[a]pyrene hydroxylase,
and 7-ethoxycoumarin O-deethylase (see Table 6-4). These increased microsomal activities
were sustained in rats exposed continuously for up to 12 weeks to 0.1 to 0.4 ppm O3, with a
greater response being observed in the activity of benzphetamine ./V-demethylase, suggesting
preferential increase in the associated P-450 cytochrome isozyme (Takahashi and Miura,
1987). Ozone-induced increases in cytochrome P-450 also have been shown not to result in
concomitant increases in microsomal xenobiotic metabolism (Rietjens et al., 1988). Rat lung
microsomal benzo[a]pyrene oxidation and benzphetamine demethylation were found to be
enhanced after a 6-mo continuous exposure to 0.5 ppm O3 (Filipowicz and McCauley, 1986a).
More recent studies have explored O3-induced changes in cytochrome P-450 isozyme patterns
and correlated changes in lung xenobiotic metabolism with Clara cell enlargement and
increased numbers during a 14-day exposure of rats to 0.4 ppm O3 (Takahashi and Miura,
1990; Suzuki et al., 1992). These authors also demonstrated, by immuno-electron microscopy,
the presence of cytochrome P-450b (IIB1) in the Clara cell endoplasmic reticulum.
Changes in the extent and pattern of formation of benzo[a]pyrene products were
investigated by Bassett et al. (1988c) in lungs from rats undergoing epithelial proliferative
repair resulting from 3 days of continuous exposure to 0.6 ppm O3. Although metabolism to
all benzo[a]pyrene metabolites was enhanced 4.7-fold, the relative proportion of metabolism
involving quinone formation was enhanced from 10 to 25 %. The toxicity of other inhaled
pollutants that undergo lung xenobiotic metabolism may therefore be dependent not only on
6-20
-------
Table 6-4. Effects of Ozone Exposure on Lung Xenobiotic Metabolisnft
Ozone
Concentration
ppm
0.1
0.2
0.4
0.2
0.2
0.4
0.4
0.8
0.4
0.5
0.6
0.8
Qg/m
196
392
784
392
392
784
784
1,568
784
980
1,176
1,600
Exposure
3 Duration
Continuous for
4-12 weeks
Continuous for
2 weeks
Continuous for
7 and 14 days
7 h/day for
14 days
Continuous for
6 h, 1-14 days
Continuous for
1 year
Continuous for
3 days
Continuous for
7 days
Species, Sex
(Strain)
Age"
Rat, M
(Wistar)
19-22 weeks old
Rat, M
(Wistar)
19-22 weeks old
Rat, M
(Wistar)
22-24 weeks old
Rat, M
(Wistar)
Rat, M
(Wistar)
5 weeks old
Rat, M
(F344)
Rat, M
(Wistar)
200-220 g
Rat, M
(Wistar)
8 weeks old
Observed Effect(s)
Concentration-dependent increases in NADPH-cytochrome P-450 reductase activity and
cytochrome P-450 content during 4-12 weeks exposure to 0.2 and 0.4 ppm O3, reaching a
maximum at 12 weeks with concomitant increases in benzo[o|pyrene hydroxylase and
7-ethoxycoumarin O-deethylase activities. NADH-cytochrome b5 reductase activity unaffected.
Four weeks at 0.1 and 0.2 ppm demonstrated a preferential increase in benzphetamine N-
demethylase activity, with no alterations in coumarin hydroxylase activity.
Increases in cytochrome P-450 isozymes ascribed to constitutive types rather than induction of
other types.
By 14 days, NADPH-cytochrome P-450 reductase activity and cytochrome P-450 content
enhanced with concomitant increases in benzo[o|pyrene hydroxylase and 7-ethoxycoumarin
O-deethylase activities by Day 7; no change in NADH-cytochrome b5.
No effect at 0.4 ppm. 0.8 ppm increased NADPH-cytochrome P-450 reductase activity and
cytochrome P-450 content, with concomitant increases on Day 7 in benzo[a]pyrene
hydroxylase and 7-ethoxycoumarin O-deethylase activities that further increased by Day 14.
By 24 h, Clara cell number decreased, but by 14 days had increased. Increase in cytochrome
P-450b (IIB1) on Days 7 and 14.
Microsomal benzo[a]pyrene oxidation and benzphetamine demethylase activities enhanced after
6 mo and 1 year of exposure.
In isolated perfused lung, increase in overall benzo[a]pyrene metabolism but with a greater
proportion being metabolized to quinones.
Cytochrome P-450, cytochrome b5, and NADPH-cytochrome P-450 reductase enhanced per
lung and per gram lung but not per milligram microsomal protein. No concomitant increases
in all cytochrome P-450-dependent reactions, suggesting alterations in isozyme patterns.
Reference
Takahashi and Miura (1987)
Takahashi and Miura (1990)
Takahashi and Miura (1985)
Takahashi et al. (1985)
Suzuki et al. (1992)
Filipowicz and McCauley (1986a)
Bassettetal. (1988c)
Rietjens et al. (1988)
"See Appendix A for abbreviations and acronyms.
''Age or body weight at start of exposure.
-------
O3-induced changes in airway protective barrier function and clearance mechanisms, but also
on O3-induced changes in epithelial cell activation and detoxification reactions.
6.2.1.7 Summary
In vitro studies have provided an indication of a wide range of initial biochemical
targets of ozone interaction that include lipid PUFAs, SH-containing proteins, and small-
molecular-weight electron donors such as GSH and vitamins E and C. Demonstration that
these interactions occur in vivo and are responsible for subsequent cytotoxicity has been more
difficult to characterize and mainly has required the use of relatively high (> 1 ppm)
concentrations of O3. However, because of the high reactivity of ozone and the relatively high
abundance of PUFAs in both cell membranes and epithelial lining fluids, PUFAs are
considered to be the most likely initial target of interaction of O3 with the lung. Current
understanding of this interaction is that, in the relatively aqueous environment of the lung
airways, hydroxyhydroperoxy intermediates are formed that break down to form aldehydes and
hydrogen peroxide. Alternatively, it has been proposed that, in more hydrophobic
environments (e.g., within a cell membrane), O3 interaction with PUFAs yields ozonides and
their free radical products. Ozonides, aldehydes, hydrogen peroxide, and other lipid break-
down products and oxygen intermediates are therefore considered to be secondary products of
the initial O3 interaction with PUFAs that would account for the observed alterations of cell
lipids, SH-containing enzyme proteins, and antioxidants associated with O3-induced cell
damage. As a result of these observations, it has been hypothesized that O3 most likely does
not reach far beyond the surface lining fluids of the upper and lower airways, which are rich in
mucopolysaccharides and surfactant lipids, respectively. However, at points where coverage is
either discontinuous or thin, epithelial cell components might be expected to directly interact
with inhaled O3.
A wide array of lung biochemical measurements have been made at different times
from the onset of O3-exposure. These measurements have included lung lipids, antioxidants,
and enzyme and structural proteins that, in some cases, can be attributed to particular cell
populations. However, many of these biochemical determinations cannot be interpreted fully
without consideration of the changes in cell population that occur as a result of O3 exposure.
In addition, the sensitivity of some of these measurements has been limited by the relatively
small percentage of the whole lung affected by O3 exposure. The more recent biochemical
determinations being made on airway samples isolated by regional microdissection should help
overcome some of these limitations.
In vivo experiments have demonstrated cleavage of total lung lipid PUPA double
bonds, with arachidonate being a major target of O3 interaction, the breakdown of cholesterol,
and the production of aldehydes and hydrogen peroxide (results that are consistent with
ozonation of cell membrane and epithelial lining lipids). The protein A component of the
alveolar surfactant system also has been identified as a possible primary target of O3
interaction. Changes observed in lung lipid biosynthesis during the first few days from the
onset of O3 exposure can be accounted for by concomitant alveolar epithelial proliferative
repair. However, lavage-recovered lipids from monkeys following O3 exposures of 0.12 ppm
for 90 days have demonstrated a relative increase in PUFAs and decrease in cholesterol-esters,
suggesting some long-term alteration in surfactant lipid composition. However, age-related
changes in lavage-recovered lipids and total lung lipid biosynthesis have been shown to be
relatively unaffected in rats exposed to 0.5 ppm O3 for periods of up to 18 mo.
6-22
-------
Many studies have utilized whole-lung measurements of antioxidant enzyme
changes as indicators of biochemical responses to O3 exposure. The increased levels of the
cytosolic enzymes G6PD, 6PGD, GR, and GSHPx and mitochondrial succinate dehydrogenase
observed during the first week from the onset of exposure to O3 levels of 0.5 to 1.0 ppm are
most likely a result of the epithelial proliferation and infiltration of inflammatory cells taking
place during this period. Failure to observe similar biochemical changes at lower
O3 concentrations most likely reflects an inability to detect focal changes of altered pathology
when using whole-lung tissue samples. Longer-term exposure of rats to an urban pattern of O3
with daily peaks of 0.25 ppm has demonstrated increases in tissue GSHPx and GR but not
SOD. These enzyme changes could reflect changes in either cellular antioxidant capacity in
response to chronic O3 exposure or the steady-state cell population.
Although no long-term changes in collagen content have been observed in rats
exposed to <0.5 ppm O3 for 18 mo, extracellular matrix thickening has been observed in rats
exposed to an urban pattern of O3 with daily peaks of 0.25 ppm for 38 weeks. Exposure of
female but not male rats for 20 mo to concentrations of 0.5 and 1.0 ppm O3 for 6 h/day has
demonstrated increased centriacinar stainable collagen and collagen and difunctional
cross-links. Similar results were obtained in lungs from monkeys exposed to 0.61 ppm O3 for
1 year, providing a sensitive indicator that long-term O3 exposure does cause some fibrogenic
alterations to the lung extracellular matrix.
Ozone-induced changes in the extent and pattern of lung microsomal metabolism of
xenobiotics have provided consistent results, which may, in part, reflect changes in the
numbers and function of bronchiolar epithelial Clara cell and alveolar epithelial Type 2 cells at
different durations of O3 exposure. These cell types are relatively enriched with cytochrome
P-450-dependent enzyme systems. Changes in both lung activation and detoxification reactions
represent important effects when considering whether or not low-level O3 exposures alter the
ability of the lung to deal adequately with the co-exposure to inhaled xenobiotics found in
urban air.
6.2.2 Lung Inflammation and Permeability Changes
6.2.2.1 Introduction
The barrier functions of the airway epithelia have been investigated by isotope
tracer techniques for detecting mucosal permeability and by analysis of the BAL for total
protein and albumin concentrations. Under normal conditions, the airway epithelia restrict the
penetration of exogenous particles and macromolecules from airway lumen into airway
interstitium and blood. The integrity of the zonula occludens (tight junctions) is regarded as a
major factor in providing barrier properties to the airway epithelia so that only a small amount
of intratracheally introduced tracers finds its way across the airway epithelia into the blood.
However, disruption of the epithelial barrier creates a leak across the airway mucosa, resulting
in increased permeability of serum proteins into the air spaces and of intraluminal exogenous
tracers into the blood. Therefore, permeability is generally detected by either the tracer
transport from airway spaces to blood or measurement of total protein and albumin in the
BAL. Both of these measures are, therefore, taken into account in discussing permeability
changes in this section. Although BAL protein measurement offers a good marker for
detecting permeability changes, it is important to note that the proteins in the BAL can result
from tissue injury and secretory activity, in addition to leakage of the serum proteins across the
airway mucosa (Hatch et al., 1989; Hatch, 1992).
6-23
-------
Inflammatory cells in the lung constitute an important component of the pulmonary
defense system. In their unstimulated state, the inflammatory cells present no danger to other
cells or tissues, but, on activation, they are capable of generating proteolytic enzymes such as
elastase and reactive oxygen species such as superoxide, hydrogen peroxide (H2O2), and the
hydroxyl radical. These oxidants can cause substantial injury to cell membranes and
intracellular components by their effects on membrane lipids and proteins (biochemical effects
of O3 were described in Section 6.1). Ozone exposure also can cause the epithelial or activated
inflammatory cells to liberate arachidonic acid, which is free to enter enzymatic lipoxygenase
or cyclooxygenase pathways that lead to the production of leukotrienes (LTs) and
prostaglandins (PGs), respectively. Although some of the studies indicate a lack of change in
the production and release of cellular mediators following O3 exposure, other studies
demonstrate an elevation in the levels of arachidonic acid and its metabolites in the bronchial
washings of rats, as well as humans (see Chapter 7) exposed to O3 under controlled conditions.
The changes in the lung levels of arachidonic acid metabolites generally were observed in
animals exposed to O3 concentrations higher than 0.5 ppm. These cellular mediators can cause
a wide range of pathophysiological changes. For example, LTB4 can cause PMN aggregation
and degranulation in vitro and margination of circulating PMNs to capillary endothelium in
vivo, whereas LTC4 and LTD4 can cause contraction of vascular smooth muscle, PGE, has
bronchodilator activity, and LTD4 and PGF2n are regarded as bronchoconstrictors. Because of
the toxic potential of the products released by PMNs, AMs, mast cells, and other inflammatory
cells, it has been suggested that the recruitment of these cells into the pulmonary interstitium is
associated with lung injury and associated edema. An inflammatory response in the lung and
an elevation of transmucosal permeability are observed after O3 exposure, but the
interdependence of these two events is a topic of debate. Although AMs are involved in
cellular changes during the course of inflammation, AMs are discussed only in terms of their
primary function in the section on host defense (Section 6.2.3.4).
The previous O3 criteria document (U.S. Environmental Protection Agency, 1986)
discussed studies available at that time on the inflammatory and permeability effects of O3.
These studies recognized the increased thickness of the alveolar septa, presumably due to
increased cellularity after acute exposure to O3 and excess collagen after chronic exposure to
O3. The inflammatory cell response was reported in rats and monkeys receiving single or
repeated exposures to O3 concentrations ranging from 0.2 to 0.8 ppm (Castleman et al., 1980;
Brummer et al., 1977; Moore and Schwartz, 1981; Crapo et al., 1984). Exposures to O3 also
resulted in increased mucosal permeability, as detected by the nonspecific diffusion of phenol
red from the lung into circulation (Williams et al., 1980) or the appearance of serum proteins
in the air spaces. Increased BAL levels of total protein, albumin, and immunoglobulin (Ig) G
were detected in rats, dogs, and guinea pigs exposed acutely to O3 concentrations ranging from
0.1 to 2.5 ppm (Alpert et al., 1971; Reasor et al., 1979; Hu et al., 1982). For example, Hu
et al. (1982) found that a 72-h exposure of guinea pigs to DO.26 ppm O3 increased BAL protein
immediately after exposure and that, when the exposure duration was decreased to 3 h, protein
increased 10 to 15 h postexposure (not immediately after exposure ceased).
6.2.2.2 Permeability Changes
A number of studies have demonstrated an increase in airway mucosal permeability
following inhalation exposure to O3 concentrations of D 1.0 ppm (Table 6-5).
6-24
-------
Table 6-5. Lung Inflammation and Permeability Changes Associated with Ozone Exposurfe
NJ
Ln
Ozone
Concentration
ppm
0.1
1.2
0.1
0.3
1.0
0.1
0.3
1.2
0.1
0.3
1.0
0.1
to 10
0.2
0.8
0.1
0.3
1.2
0.1
0.2
1.2
4.0
0.12
0.3
Dg/m
196
2,352
196
588
1,960
196
588
2,352
196
588
1,960
196
to
19,600
196
392
784
1,568
196
588
2,352
196
392
2,352
7,840
235
588
Exposure
Duration
2h/day for 1,2,
6, and 13 days
2h
2h
In vitro and in vivo
2h
In vitro
2h
In vitro
2, 4, and 8 h
2h
In vitro
Continuous for
1 to 12 weeks
24 h
48 h and 72 h
Species, Sex
(Strain)
Age"
Rabbit, M
(NZW)
2-4 mo old
Rabbit, M
(NZW)
15-16 weeks old
Rabbit, M
(NZW)
2-4 mo old
Rat (S-D)
12-18 weeks old
Cow
Rat (F344)
90 days old
Guinea pig
(Hartley)
60 days old
Rabbit, M
(NZW)
2-4 mo old
Rat, M
(Wistar)
16 weeks old for
1 week exposure;
21 weeks old for
longer exposures
Mice, M
(C57BL/
6J[B6]);
(C3H/
HeJ[C3])
6-8 weeks old
Observed Effect(s) Reference
Increase in AM number at 7 days following single exposure to 0.1 ppm and increase in Driscoll et al. (1987)
number of AMs and PMNs on 1-day after cessation of 6 or 13 days of exposure. Increase
in number of PMNs at 24 h after single exposure to 1 .2 ppm.
Increase in levels of PGE, and PGF^ in BAL immediately after exposure to 1.0 ppm Schlesinger et al. (1990)
O3 only. No significant effects were observed on the levels of 6-keto-PGF,n , TXB, or LTB4
In vitro: Increase in PGE, after 0.3 ppm and increase in PGEjj after 1.2 ppm by AMs. In Driscoll et al. (1988)
vivo: Increase in the release of PGE, and PGE^ by AMs after 1.2 ppm, but no effect of
0.1 ppm.
Increased production of arachidonic acid metabolites by AMs at 1.0 ppm only. Madden et al. (1991)
Increased production of PGE and,PGF by-Jracheal epithelial cells after exposure to Leikauf et al. (1988)
0.1 and 0.3 ppm. Increased production of other arachidonic acid metabolites at > 1.0 ppm.
C x T exposure design; BAL 25 h after exposure started. PMNs measured in rats only; no Highfill et al. (1992)
C and T interaction; effect dependent on C. Exponential and polynomial response surface
model used. Similar protein responses at low C x T products; generally, the influence of T
increased as C increased. Exponential model explained 86% of the data.
Exposure of AMs to DO. 3 ppm O3 resulted in increased secretion of factors capable of Driscoll and Schlesinger (1988)
stimulating migration of inflammatory cells.
Number of AMs in BAL increased after exposure for 11 weeks to 0.2 ppm. Infiltration of Mochitate et al. (1992)
PMNs did not occur.
BAL immediately PE. Comparable increases in BAL protein, AMs, PMNs, and Kleeberger et al. (1993a)
lymphocytes in the two strains after exposure to 0. 12 ppm, but greater number of
inflammatory cells and protein concentration in B6 than in C3 mice after exposure to
0.3 ppm.
-------
Table 6-5 (cont'd). Lung Inflammation and Permeability Changes
Associated with Ozone Exposure^
Ozone
Concentration
ppm
0.12
0.8
1.5
0.12
0.8
1.5
0.12 to
0.96
0.2
0.5
1.0
2.0
0.2
0.4
0.6
0.8
0.2
0.4
0.6
0.8
0.25
peak
over a
bkgof
0.06
Dg/m
235
1,568
2,940
235
1,568
2,940
255
to
1,882
392
980
1,960
3,920
392
784
1,178
1,568
392
784
1,176
1,568
490
118
Species, Sex
Exposure (Strain)
Duration Age1'
6 h Rat, F
(F344/N)
12-18 weeks old
6 h Rat, M
(F344/N)
12-18 weeks old
6 h, 24 h, Rat, M
or 2 days (S-D)
250-300 g
4 h Mouse (Swiss Albino)
19-25 g
Guinea pig (Hartley)
3 14-522 g
Rat (S-D)
280-350 g
Rabbit (NZW)
1.7-2.5 kg
Hamster
(Golden Syrian)
94-107 g
6, 8, 12, and Rat
24 h/day for 3 days (S-D)
10-12 weeks old
7 h/day for Rat, M
1,2, or 4 days (PVG)
12-16 weeks old
13 h bkg, rose to Rat, M
peak and returned to (F344)
bkg over 9 h 60 days old
Observed Effect(s) 3 Reference
Increased number of PMNs in nasal lavage, but not in BAL at 18 h after 0. 12 ppm; increased Hotchkiss et al. (1989a)
number of PMNs in BAL, but not in nose after 1.5 ppm; number of PMNs decreased with time in
nose, with a concomitant increase in BAL PMNs after 0.8 ppm.
AMs and PMNs increased in number in BAL at various times PE at DO. 8 ppm. Hotchkiss et al. (1989b)
Total protein in BAL increased after exposure to DO. 4 ppm for 6 h and DO. 12 ppm for 1 or 2 days. Guth et al. (1986)
Transport of radiolabeled albumin from blood to the airways increased after 6- or 24-h exposure to
DO. 4 ppm and after 2 days exposure to 0.2 ppm.
Species differences in responsiveness. At 18-20 h PE, total protein in BAL increased in guinea pigs Hatch et al. (1986)
exposed to 0.2 ppm, whereas mice, hamsters, and rats responded toDl.O ppm, and rabbits
responded only to 2.0 ppm.
C and T matched such that all C x T = 14.4 ppm - h. BAL immediately after exposure ceased. Gelzleichter et al. (1992b)
Increase in PMNs equivalent in all O3 groups. Increase in protein equivalent for 6-, 8-, and 12-h
exposure groups, all of which are greater than protein in 24-h groups. Equivalent results for BAL
PMNs.
BAL approximately 17 h PE. The proportion of AMs in the BAL decreased, with a concomitant Donaldson et al. (1991 , 1993)
increase in the proportion of PMNs after 1 or 2 days exposure to DO. 6 ppm O3. No significant
effect on total number of lavageable cells or on the ability of neutrophils to injure epithelial cells.
Interstitial AMs increased in number in proximal alveolar region and TBs at one week of exposure, Chang et al. (1992)
but the effects had subsided by 3 weeks of exposure.
-------
Table 6-5 (cont'd). Lung Inflammation and Permeability Changes
Associated with Ozone Exposure^
Ozone
Concentration
ppm
0.3
2.0
0.35
0.5
1.0
0.35
0.5
0.65
0.8
0.38
0.76
1.28
2.04
0.13
0.26
0.38
0.4
0.6
0.4
Dg/m
588
3,920
686
980
1,960
686
980
1,274
1,568
750
1,500
2,500
4,000
250
500
750
784
1,176
800
Species, Sex
Exposure (Strain)
Duration Age1'
24, 48, or 72 h for Mice
0.30 ppm and 3 h for (C57BL/
2.0 ppm 6J[B6]);
(C3H/
HeJ[C3])
DBA/2J (D2), hybrids, and
recombinant inbred strains
(Rl): BXDandBXH
6-8 weeks old
2.25 h/day for 5 days Rat, M
(F344)
3-4 mo old
2, 4, and 7 h Rat, M
(F344)
1 3 weeks old
1,2, 4, and 8 h Rat
daytime (Wistar)
7 weeks old
4, 8, and 12 h
nighttime
8 h/days for 90 days Monkey, M
(Bonnet)
5.2-8 years old
12 h during day or Rat, M
night (Wistar)
Guinea pig, M
(Hartley)
9 weeks old
Observed Effect(s) Reference
Inflammatory response was greater in B6 than in C 3 or D2 mice. Fl progeny was categorized as Kleeberger et al. (1990,
resistant; F2 generation segregated into 45:16 for resistant vs. susceptible phenotypes. Among 1993b)
BXD Rl strains, 4 of 10 responded discordantly to the two exposures (0.3 and 2.0 ppm).
Among BXD Rl, 4 of 16 were discordant.
Persistent increase in BAL protein and progressive inflammation at DO. 5 ppm. Tepper et al. (1989)
C x T exposure design. All exposures included 45 min of CO, for 1 h to increase ventilation. Tepper et al. (1994, in draft)
BAL after pulmonary function tests completed. The quadratic model explained 92% of the
variance. The models suggest that C may have a more dominant influence than T.
C x T exposure design. BAL protein measured at various times PE. Rombout et al. (1989)
Daytime exposures: At 0.76 ppm, maximal increase 22 h after exposure started; after 4 and 8 h
of exposure, protein still elevated at 54 h from start of exposure.
Nighttime exposures: Temporal increase and decrease of protein more gradual, with maximal
response at 36 h after exposure started. Protein still elevated 72 h after start of 8- or 12-h
exposure to 0.26 or 0.38 ppm. Smallest tested C x T effect was with 0.13 ppm x 4 h.
Both: Multivariate regression analysis. Polynomial function shows that T has progressive
influence as C increases.
Inflammatory response in RBs at 0.64 ppm. Moffatt et al. (1987)
Nighttime exposure of rats resulted in greater increase in BAL protein, albumin, and PMNs than Van Bree et al. (1992)
the daytime exposure. A similar difference was not observed in guinea pigs.
-------
Table 6-5 (cont'd). Lung Inflammation and Permeability Changes
Associated with Ozone Exposure^
NJ
CO
Ozone
Concentration
ppm
0.5
0.5
0.5
1.0
0.75
0.8
0.6
0.8
0.8
0.8
0.8
0.8
Dg/m
980
980
980
1,960
1,410
1,568
1,176
1,568
1,568
1,568
1,568
1,568
Exposure
Duration
2h
Continuous
exposure for
1-14 days
4h
Continuous
exposure for
3 days
2h
2 h exposures
during rest or
exercise
2h
3h
3h
2h
Species, Sex
(Strain)
Age1'
Dog, M
(Mongrel)
15 ± 0.9kg,
Baboon, M
25-40 kg
Mouse, F
(Swiss)
20-25 g
Guinea pig, M
(Hartley)
300-400 g
Rat, M
(Wistar)
200-250 g
Rat, M
(S-D)
300 g
Rat, M
(S-D)
47-52 days old
Rat, M
(S-D)
50-60 days old
Rat, M
(S-D)
250-300 g
Rat, M
(S-D)
250-300 g
Rat, M
(F344)
11-12 weeks old
Observed Effect(s) 3
No effect on levels of 6-keto PGF]D, PGE,, TXA,, TXB,, or PGF^ in BAL.
PGE and total protein levels in BAL increased after the exposure, peaked at 3 days, then declined with time,
but remained higher than the controls at 7 days; protein still increased at 14 days. Total cells in BAL decreased
on Days 1 to 3 after exposure.
Depleting lungs of ascorbic acid enhanced effects of 0.5 but not 1.0 ppm on BAL protein. Depletion of lung
nonprotein sulfhydryl had no effect.
Increased number of PMNs and AMs and elevated levels of albumin in the BAL. At 4 days PE, no PMNs
were detected, but AM numbers and albumin levels were elevated.
Transient increase in tracheal and bronchoalveolar permeability, as revealed by tracer transport from airways to
blood and tracer localization in intercellular spaces.
Airway permeability increased after exposure of resting animals; trends of greater and more persistent effects in
exercising group.
Increased transport of radiolabeled tracers from blood to the air spaces following exposure.
The number of PMNs in lung parenchyma increased immediately after exposure, peaked at 8 h PE, and
returned to baseline by 16 h PE. Total protein and albumin levels in BAL increased immediately after
exposure, peaked at 8 h PE, and then declined with time, but the albumin levels were higher than the controls at
24 h PE.
Time-related changes in tracheal permeability, detected by tracer transport, and PMN influx in tracheal wall
following exposure. Increase in permeability prior to increase in PMNs.
Increased DTPA transport across the tracheal mucosa and elevated levels of protein and albumin in BAL.
Effects attenuated in leukopenic rats or rats pretreated with indomethacin or FPL 55712.
Reference
Foukeetal. (1990, 1991)
Canning et al. (1991)
Slade et al. (1989)
Bassettetal. (1988a)
Bhalla et al. (1986)
Bhalla and Crocker (1986)
Bhalla et al. (1987)
Bhalla and Crocker (1987)
Bhalla and Young (1992)
Young and Bhalla (1992)
Bhalla et al. (1992)
-------
Table 6-5 (cont'd). Lung Inflammation and Permeability Changes
Associated with Ozone Exposure^
Ozone
Concentration
ppm Qg/m
0.8 1,568
0.8 1,568
0.96 1,882
1.0 1,960
1.0 1,960
1.0 1,960
1.0 1,960
1.0 1,960
Species, Sex
Exposure (Strain)
Duration Age1'
2 h Rat, M
(F344)
250-275 g
3 h Mouse, F
(CD-I)
5 and 9 weeks old
8 h Monkey, M
(Rhesus)
2-8.5 years old
4 to 24 h Rat
(S-D)
63-70 days old
5 min Dog, M
O3 delivered to a (Mongrel)
localized area of lung 21.2 + 0.5 kg
via a Teflon catheter
fitted to bronchoscope
6 h Rat, F
(S-D)
8-9 and
13-17 weeks old
2, 4, or 6 h Rat, M
(S-D)
13, 18 days
and 8 and
16 weeks old
2h Rat
(S-D)
18 days or
14 weeks old
Rabbit
(NZW)
6, 11, 16, or
30 weeks old
Observed Effect(s) 3
PMNs isolated from blood of O3-exposed rats displayed deformation of shape, indicative of motility and
greater cell adhesion than the PMNs from air-exposed rats.
Increase in PGE, in BAL in 5- week-old mice only; effect blunted by indomethacin pretreatment.
Number of labeled PMNs into lung tissue and BAL increased immediately after exposure, peaked at 12 h PE,
and returned to baseline by 24 h PE. Total labeled and unlabeled PMNs in BAL remained elevated at 24 h,
but returned to control levels by 72 h PE. Total protein in BAL was elevated only at 24 h PE.
Total protein and PMNs in BAL and PMNs in the CAR of the lung increased with exposure duration, but the
number of AMs in BAL decreased. Treatment with anti-rat-PMN serum resulted in elimination of PMNs in
BAL, but it did not affect the O3-induced increase in BAL protein.
Number of PMNs in the subepithelial tissue increased at 1-3 h PE. Number of BAL PMNs increased at 24,
but not at 1-3 h PE.
BAL 16 h PE. Enhanced responsiveness to O3-induced inflammation and elevated protein levels in BAL
developed during pregnancy, was maintained during lactation, and disappeared following lactation.
BAL immediately after exposure. PGE, concentrations in BAL greatest in 13-day-old rats after 2 h of
exposure, but in older rats the response was seen after 6 h of exposure. In 13-day-old rats, 50% of
leukocytes in BAL were dead after 6 h of exposure; no such effect on 16-week-old adults. No age
dependence for BAL protein increase or PMN increase.
BAL immediately after exposure. In youngest animals, greater amounts of PGE, and PGRj. In youngest
rabbits, 6-keto PGF^ and TXB2 increased. No effect on LTB4. No age dependent effects on BAL protein or
cell number.
Reference
Bhalla et al. (1993)
Gilmour et al. (1993b)
Hyde et al. (1992)
Pino et al. (1992a,b)
Kleeberger et al. (1989)
Gunnison et al. (1992b)
Gunnison et al. (1992a)
Gunnison et al. (1990)
-------
Table 6-5 (cont'd). Lung Inflammation and Permeability Changes
Associated with Ozone Exposure^
Ozone
Concentration c .nri c
ppm
1.0
Exposure (Strain)
Qg/m Duration Age1'
1,960 Ih Guinea pig, M
(Hartley)
300-400 g
Observed Effect(s) 3
Appearance of horseradish peroxidase in plasma, following its intratracheal administration, was accelerated at
2 and 8 h PE, but not at 24 h PE.
Reference
Miller et al. (1986)
1.0 1,960 Ih Guinea pig, M The concentrations of PGE,, 6-keto PGF,n, and TXB, in BAL increased at various times following exposure. Miller et al. (1987)
(Hartley)
250-300 g
1.0 1,960 3h Rat
isolated perfused (S-D)
lung 350 ± 42 g
No effect on BAL protein.
load et al. (1993)
U>
O
1.8
3,528 2 or 4 h
Rat, M
(Wistar)
200-250 g
A decrease in number of AMs in BAL immediately after exposure. PMNs and albumin content of BAL
increased at 1 day PE. Increased albumin levels, but not PMNs, persisted on Day 3 PE.
Bassett et al. (1988b)
2.0 3,920 4h Guinea pig, M Interstitial PMNs increased in number immediately after exposure, but declined by 24 h PE. BAL PMNs Schultheis and Bassett (1991)
(Hartley) were maximal by 3-6 h and remained elevated by 3 days PE.
300-350 g
"See Appendix A for abbreviations and acronyms.
''Age or body weight at start of exposure.
-------
In rats exposed for 2 h to 0.8 ppm O3, labeled tracers, such as diethylenetriaminepentaacetate
(DTPA) and bovine serum albumin, introduced into the airway lumen were transferred to
blood to a greater extent than in the air-exposed rats (Bhalla et al., 1986; Bhalla and Crocker,
1986; Crocker and Bhalla, 1986). The rapidly rising concentration of the tracers in the blood
during the initial period of instillation of the tracers into the airways reflected both the
accumulation, due to slow instillation over a 5-min period, of the tracers in the airway lumen
and subsequent transfer across the respiratory epithelium. The changes in permeability
observed in this study were transient in nature, returning to the baseline value by 24 h
postexposure in the trachea and by 48 h in the distal airways. Reversible increases in airway
epithelial permeability also were observed in guinea pigs acutely exposed to 1 ppm O3 (Miller
et al., 1986). The rate of appearance of intratracheally administered horseradish peroxidase
increased in blood at 2 and 8 h after O3 exposure, as compared to rats at 24 h postexposure to
O3 and air-exposed controls. When rats were exercised at a level that increased the minute
ventilation (Vg) twofold, the effect of 0.8 ppm O3 was not only greater than in rats exposed at
rest, but the increased permeability persisted longer (Bhalla et al., 1987).
Guth et al. (1986) analyzed the permeability effects of O3 by injecting radiolabeled
albumin into the blood and measuring it in the BAL, as well as by measuring the total protein
concentration in the BAL. This study revealed a concentration-dependent increase in
permeability following a 6-h exposure of rats to DO.4 ppm or following 1 or 2 days of exposure
to DO. 12 ppm O3. For example, after a 2-day exposure to 0.12 ppm, there was a 71 % increase
in BAL protein. Tracer transport also was increased in rats exposed for 2 h to 0.8 or 2.0 ppm
O3 (Crocker and Bhalla, 1986; Bhalla and Crocker, 1987).
The relative influence of concentration and duration of O3 exposure was evaluated
by three laboratories using BAL protein as an indicator of effects. In the first study, Rombout
et al. (1989), exposed rats for 1, 2, 4, or 8 h to 0.38, 0.76, 1.28, or 2.04 ppm O3 during the
daytime (16 C x T products). A similar nighttime exposure study was conducted using
0.13 to 0.38 ppm O3 and 4, 8, or 12 h of exposure (nine C X T products). The smallest
C X T product causing an increase in protein was 0.52 ppm • h (0.13 ppm X 4 h).
A multivariate regression analysis accounted for 88.6% of the variance in the daytime data and
73.2% in the nighttime data. Animals exposed during the night were more responsive. A
quadratic polynomial function showed that the influence of T increased with increasing C and
that the influence of T was still important at the lowest O3 concentration tested (0.13 ppm).
The second study employed rats and guinea pigs, each having 12 C x T products (0.1, 0.2,
0.4, and 0.8 ppm O3; 2, 4, and 8 h) (Highfill et al., 1992). Using additional modeling
approaches, they obtained similar results to those of Rombout et al. (1989) . For example, the
exponential response surface model explained 86% of the variance in the data and showed that
the influence of T increased as C increased. However, at low C x T products, similar BAL
protein increases were observed. Further modeling of these data (Highfill and Costa, 1995)
again showed that C and T had interdependent influences. Tepper et al. (1994) performed a
similar C X T study with 12 C X T products (0.35 to 0.8 ppm O3, 2 to 7 h). However, rats
were exposed to 8 % CO2 for 45 min of each hour to increase ventilation, and BAL was
conducted on lungs that had been measured for pulmonary function. The response surface
predicted by the modeling again indicated that the influence of T increased as C increased.
Histopathological observations in the rats support the findings that C had more influence than
T. Tepper et al. (1994) compared their analysis of BAL protein to that of Highfill et al. (1992)
and found very good agreement, even though there were experimental differences. However,
6-31
-------
in Tapper et al. (1994), there were larger constants for C terms, indicating that C had a greater
influence than in the Highfill et al. (1992) study, probably because Tepper and co-workers
increased ventilation (and hence O3 dose) by using concurrent CO2 exposures.
Gelzleichter et al. (1992b) exposed rats to a single O3 C x T (14.4 ppm • h)
composed of 16 products (0.2 to 0.8 ppm O3, 6 to 24 h/day for 3 days). They found that the
24 h/day exposure groups had significantly fewer responses than the other groups, which were
all equivalent. Thus, in this study, C and T had equivalent influences on the response, except
when T was 24 h/day. This study was well conducted, but had some basic differences from
the Rombout et al. (1989) and Highfill et al. (1992) studies in that the longer exposure
durations (i.e., 24 h/day) involved a mixture of daytime and nighttime exposure that likely
altered the dose-rate of O3. Also, Gelzleichter et al. (1992b) used one C x T product,
whereas the other studies used several C x T products.
6.2.2.3 Concomitant Changes in Permeability and Inflammatory Cell Populations
in the Lung
Polymorphonuclear leukocyte infiltration in the lung following O3 exposure has
been investigated in a number of studies (Table 6-1), either by analyzing the cellular content of
the BAL or by counting PMNs in lung sections. Bassett et al. (1988a) found an increase in the
number of inflammatory cells in the BAL of rats continuously exposed for 3 days to 0.75 ppm
O3. The inflammatory response was accompanied by elevated levels of albumin and lactate
dehydrogenase, suggesting increased permeability and cellular injury. Comparable changes
were also observed in rats acutely exposed to a higher O3 concentration (Bassett et al., 1988b).
In another study, a random count of PMNs in the lung sections at 4-h intervals, following a 3-h
exposure of rats to 0.8 ppm O3, revealed a gradual increase in the number of PMNs, with a
peak at 8 h postexposure and a return to the baseline value by 16 h postexposure (Bhalla and
Young, 1992). The total protein and albumin concentrations in the BAL also increased after
the exposure, peaking at 8 h postexposure. Although the protein concentrations returned to
baseline by 16 h postexposure, the albumin levels remained above the controls after 24 h.
Alveolar changes, consisting of thickened septa, parenchymal cellularity, and increased
numbers of free cells, began to increase between 12 and 16 h postexposure and were still
increasing at 24 h postexposure.
In trachea of rats exposed for 3 h to 0.8 ppm O3, a peak of PMN infiltration at
12 h postexposure was preceded by a decline in the number of PMNs in pulmonary capillaries,
suggesting exit of PMNs from the blood vessels and their migration across the endothelial cells
into the tracheal wall (Young and Bhalla, 1992). Although a significant change in the tracheal
population of PMNs did not occur until 12 h after the end of exposure, tracheal permeability,
as detected by DTP A transport, increased immediately following O3 exposure. The results of
this study suggest that the initial changes in tracheal permeability may be independent of an
inflammatory response, but the recruited PMNs may serve to sustain the increased
permeability and to amplify O3 effects at later stages. This conclusion was based on the
observed shift of PMNs from the vascular compartment into the tracheal wall and a concurrent
peak of increased permeability. In comparable studies, Pino et al. (1992a) exposed rats to
1.0 ppm O3 for periods ranging from 4 to 24 h. Total protein and the number of PMNs in the
BAL increased with time, with the maximum increase at the end of 24 h of continuous
exposure. The number of AMs was lower in the exposed animals than in the controls.
By morphometry, the peak PMN response in the terminal bronchioles (TBs) and alveolar ducts
6-32
-------
(ADs) occurred at 4 h after an 8-h exposure. In dogs, local exposure of peripheral airways to
1 ppm O3 for a short period (5 min) produced a recognizable inflammatory response
(Kleeberger et al., 1989). An increase in the number of PMNs was detected in the
subepithelial tissue within 3 h after 5 min of exposure of the dogs, but the response had
subsided 24 h later. In BAL, on the other hand, an increase in the number of PMNs was not
observed at 3 h postexposure; the number of PMNs increased at 24 h.
Hotchkiss et al. (1989a,b) have investigated the effects on AMs and PMNs of a 6-h
O3 exposure of rats to 0.12, 0.8, or 1.5 ppm O3 and compared the inflammatory responses by
nasal lavage and BAL, as well as by morphometry in the nose and the CAR of the lung, a site
at which abnormal cellular changes generally occur following O3 exposure. Animals were
examined 3, 18, 42, or 66 h after exposure ceased. From lavage data, 0.12 ppm O3 had no
effect. At 0.8 ppm, there was an increase in the number of nasal PMNs lavaged immediately
after exposure, which tapered off (no significant change at 42 h postexposure). In contrast,
BAL PMNs increased later, beginning at 18 h postexposure and peaking at 42 h postexposure.
From morphometric data, 0.12 ppm O3 caused an increase in nasal PMNs 66 h postexposure.
At 0.8 ppm, nasal PMNs increased to their greatest extent immediately after exposure and still
were increased at later time periods. However, PMNs in the lung increased only at 18 and
66 h postexposure. The interpretation of these results was based on the presence of potential
competing mechanisms in the nose and lungs. Therefore, the attenuation of the nasal effects
are matched by simultaneous enhancement of the inflammatory response in the lung. Whether
such a balance between nasal and alveolar PMNs represents a specialization restricted to rats or
is a more general phenomenon remains to be investigated. A similar balance was not observed
in humans exposed to O3 (see Chapter 7). Subtle differences in species, O3 concentrations, and
exposure durations, however, need to be considered when making interspecies comparisons.
Hyde et al. (1992) investigated the inflammatory response in monkeys exposed to
0.96 ppm O3 for 8 h. Polymorphonuclear leukocytes were isolated from peripheral blood,
labeled with indium-lll-labeled tropolonate and infused into the cephalic vein of monkeys 4 h
before necropsy. Labeled PMNs in the lung tissue and the BAL peaked at 12 h and returned to
control values by 24 h postexposure. The total number of labeled and unlabeled PMNs in the
BAL, however, remained elevated at 24 h postexposure, but returned to baseline by 72 h.
Furthermore, the PMN peak at 24 h postexposure coincided with the maximum increase in
BAL protein at this time point. These studies suggest a strong correlation between BAL
protein concentration, epithelial necrosis, and inflammatory cells (especially eosinophils) in
bronchi, but not in the trachea or bronchioles. This observation may represent a species-
specific response. In rats, the inflammatory response in the terminal airways involved an
increase in the number of migratory cells, including PMNs, but not eosinophils (Pino et al.,
1992a). The available literature suggests that the precise time point at which the maximum
change in the number of inflammatory cells occurs is variable and may be dependent on several
factors, including animal species, concentration, duration of exposure, and mode of analysis
(i.e., BAL versus morphometry of lung parenchyma). Because the PMNs sampled by BAL
represent only a small fraction of the cells shown to be present in the air spaces by
morphometry (Downey et al., 1993), the inflammatory response detected by analyzing BAL
may or may not match the response obtained by microscopic analysis of tissue sections. Even
when the PMN response detected by the BAL analysis accurately reflects the tissue PMNs
(Hotchkiss et al., 1989a), the times at which the PMN response peaks do not necessarily
coincide when the analyses are made by the two procedures. Therefore, the mode of analysis
6-33
-------
(BAL versus morphometry) and the time at which this analysis is made need to be taken into
account when analyzing the inflammatory response. This recommendation is consistent with
the conclusion of Schultheis and Bassett (1991) that BAL does not necessarily reflect cellular
changes in the lung interstitium.
Another approach to studying the inflammatory impact of O3 and its effects on
airway permeability is based on exposure of rats to drugs that destroy leukocytes or block the
activity of chemical mediators released by these cells. To determine whether the PMNs play a
role in O3-induced increased permeability, Pino et al. (1992b) studied O3 effects in PMN-
depleted rats. Although ip injection of anti-PMN serum resulted in a nearly complete depletion
of PMNs in rats, it did not affect the increase in BAL protein following an 8-h exposure to 1.0
ppm O3. In comparable studies, rats were rendered leukopenic by ip injection of
cyclophosphamide (Bhalla et al., 1992). A 2-h exposure of untreated rats to 0.8 ppm
O3 caused a significant increase in the tracheal mucosal permeability, as measured by enhanced
trachea-to-blood transport of 99mTc-radiolabeled DTPA immediately postexposure and
accumulation of protein and albumin in BAL at 12 h postexposure. Pretreatment with
cyclophosphamide, a potent immunosuppressive agent, which can be toxic to the lung, did not
change baseline values but did eliminate the O3 response. The reasons for the discrepancy
between these results and the results of the anti-PMN serum treatment study of Pino et al.
(1992b) are not entirely clear, but it is likely that the O3 effects are dependent, in part at least,
on an interaction between different inflammatory cell types. Therefore, it is not unreasonable
to assume that, in the absence of PMNs, their role is taken up by another cell type. The
attenuation of O3 effects also was observed in rats pretreated with indomethacin, an inhibitor of
cyclooxygenase products, and FPL55712, which blocks LTD4 activity by preventing its
binding to the receptors (Bhalla et al., 1992). Based on these results, it was proposed that,
although O3 is capable of producing direct injury to cells, inflammatory cells and their products
may contribute to the injury process (Bhalla et al., 1992). This conclusion is supported by the
recent studies of load et al. (1993). In the isolated perfused rat lung, PMNs (but not O3)
increased BAL protein concentration. However, PMNs acted synergistically with O3 in the
induction of epithelial injury in the bronchioles. The recent demonstration of the effects of
O3 (0.8 ppm, 2 h) on some of the cellular activities of vascular PMNs (Bhalla et al., 1993)
further suggests potential mechanisms involved in the stimulation of PMNs and the induction
of inflammatory response. Polymorphonuclear leukocytes isolated from the blood of rats
exposed to O3 displayed shape changes, indicative of cell motility, and greater adhesion to
epithelial cells in culture than did the PMNs from rats exposed to purified air.
6.2.2.4 Sensitive Populations
In addition to investigating the inflammatory response and permeability changes in
healthy adult animals, studies in recent years have analyzed the effects of O3 on lung
inflammation and airway permeability in different animal species, in potentially susceptible
subpopulations, and under special conditions (Table 6-1). Hatch et al. (1986) performed an
interspecies comparison to determine their relative responsiveness to O3. Although the
baseline BAL protein concentration of all the species was nearly the same, there were
noticeable differences in changes in BAL protein concentration among different species
following their exposure to O3. Significant changes were observed in guinea pigs exposed to
0.2 ppm O3. Mice, hamsters, and rats responded at O3 concentrations of 1 ppm and above, but
rabbits responded only to 2 ppm O3. In the case of rats, no differences were observed in the
6-34
-------
sensitivity between males and females. When Slade et al. (1989) depleted guinea pigs of lung
ascorbic acid, they were more susceptible to an O3-induced increase in BAL protein when the
4-h exposure was to 0.5 but not 1.0 ppm. Depletion of lung nonprotein sulfhydryls did not
enhance susceptibility. Although Hu et al. (1982) report no elevation in BAL protein in O3-
exposed, vitamin C-deficient guinea pigs, when their data were statistically reanalyzed by
Slade et al. (1989), vitamin C deficiency enhanced the effects of 0.5 but not 1.0 ppm O3.
Kleeberger et al. (1990) found interstrain differences in inbred mice with regards to
inflammation and permeability changes following high-concentration (2-ppm) O3 exposure for
3 h that led the investigators to propose that the PMN response to O3 may be controlled by a
single autosomal recessive gene at a chromosomal location designated as "Inf" (inflammation)
locus. In the follow-up studies, Kleeberger et al. (1993a) exposed the "susceptible"
(C57BL/6J) and the "resistant" (C3H/HeJ) strains of mice to lower concentrations of O3.
Although changes in inflammatory response and BAL protein were observed after exposure for
24 to 72 h to 0.12 as well as 0.3 ppm O3, the elevation in response in the susceptible strain
over that in the resistant strain was observed only at 0.3 ppm O3. Further studies with
recombinant inbred strains of mice suggested that genes at different loci may be responsible for
responses to 24-h (Inf locus) and 48-h (Inf-2 locus) O3 exposures (Kleeberger et al., 1993b).
Gunnison et al. (1992a) exposed rats aged 13 and 18 days and 8 and 16 weeks old
to 1 ppm O3 for 2, 4, or 6 h. In the experiments to be discussed here, BAL was performed
immediately after exposure. Ozone exposure resulted in an increase in protein in the BAL and
a decrease in the number of leukocytes, but this decrease was not specific for a certain age
group. A weak relationship was observed between age and the number of lavageable PMNs; a
slightly greater influx of PMNs was observed in the younger rats. A strong inverse
relationship was, however, observed between age and leukocyte viability. Approximately 50%
of the total leukocytes recovered in the BAL from 13-day-old rats exposed for 6 h were dead,
as compared to about 10% dead in the 16-week-old rats. Furthermore, 13-day-old animals
were more responsive to a 2-h O3 exposure than the other age groups of rats in terms of PGE2
levels in BAL; PGE levels were enhanced more in older animals with the longer exposure
durations. The authors attribute the increase in PGE2 to an increased release of arachidonic
acid, rather than an effect on metabolism or formation on PGE2. Gunnison et al. (1990) also
have shown that levels of several eicosanoids in rabbits show a similar pattern of age-
responsiveness.
Factors such as physical activity and pregnancy, in addition to age, can modify the
airway sensitivity of rats to O3. Van Bree et al. (1992) have reported circadian variation in
response to O3. In rats exposed to 0.4 ppm O3 for 12 h, about 70% more PMNs were
recovered in the BAL after nighttime exposure than after daytime exposure. This increase was
attributed to greater physical activity and increased ventilation in the nocturnal animals. In
guinea pigs, a similar difference between daytime and nighttime exposures was not observed;
instead, the variations appeared to be related to random physical activity. The nighttime
exposures also caused a greater increase in BAL protein and albumin in rats but not in guinea
pigs. Gunnison et al. (1992b) have found that pregnant rats are more responsive to O3 (1 ppm
for 6 h) than virgin females, as measured by an enhanced inflammatory response and as
detected by the analysis of protein, PMNs, leukocytes, and enzyme activities in BAL at 18 h
postexposure. When O3 exposure occurred on Day 17 of pregnancy or Days 3,13, and 20 of
lactation, the magnitude of the increase in BAL protein and number of PMNs was greater than
the magnitude of increase in virgin rats. No such increased responsiveness was observed in
6-35
-------
rats at Day 10 to 12 of pregnancy or 14 days after lactation ceased. Enzyme changes followed
a similar pattern.
6.2.2.5 Repeated Exposures
The magnitudes of some of the effects of O3 observed after an acute exposure were
smaller following repeated exposure. This phenomenon has been referred to as tolerance,
adaptation, or attenuation. Most of the older literature is on tolerance, which classically is
defined as the phenomenon in which a previous exposure to a nonlethal concentration of
O3 provides protection against an otherwise lethal level. These studies and others at high
O3 levels are discussed in the last O3 criteria document (U.S. Environmental Protection
Agency, 1986).
Tepper et al. (1989) observed attenuation of pulmonary function changes in rats
exposed for 2.25 h/day for 5 days, but a corresponding attenuation of lung inflammation did
not occur. Histologic examination of the lung sections revealed substantially more
inflammatory cells in alveoli after 5 days of exposure to 0.5 ppm O3 than after a single
exposure to the same O3 concentration. Increased protein concentration in the BAL observed
after a single exposure also persisted after 5 days of repeated exposures. The morphologic
studies of Moffatt et al. (1987) identified an inflammatory response in the respiratory
bronchioles (RBs) of bonnet monkeys exposed for 8 h/day for 90 days to 0.64 ppm O3.
Significantly greater numbers of AMs, mast cells, and PMNs reflected persistence of
inflammation following repeated exposures. Chang et al. (1992) exposed rats to an ambient
pattern of O3. In this morphometric study, the responses (epithelial inflammation in the
proximal alveolar region and the TBs, interstitial edema, and infiltration of AMs) to 1 week of
O3 exposure had subsided after 3 and 13 weeks of exposure. Donaldson et al. (1993) did not
find a change in the total number of cells in BAL of rats exposed for 7 h/day for 4 days to
O3 concentrations ranging from 0.2 to 0.8 ppm. The number of PMNs, however, increased
after exposure to 0.6 and 0.8 ppm O3. This increase was greatest after the first day of
exposure, but it was resolved by Day 4. In the studies of Mochitate et al. (1992), the number
of BAL AMs of rats exposed to 0.2 ppm O3 for 11 weeks was about 60% greater than in the
air-exposed controls. The preferential increase in the number of small AMs was not dependent
on an enhancement of DNA synthesis. It was concluded that AMs adapt to long-term
exposures as a result of recruitment of immature AMs from an influx of monocytes. No
increase in the number of PMNs was observed in the BAL of exposed rats.
When analyzing the PMN data from different studies like the ones discussed above,
it is important to make a distinction between the PMN response in the lung interstitium versus
that observed in the BAL. It is possible that, although the inflammatory response may persist
after repeated exposures, the PMNs do not necessarily continue to migrate from pulmonary
interstitium into the air spaces. As a result, the inflammatory response is detected in the
histological sections of the lung but not in the BAL.
6.2.2.6 Mediators of Inflammation and Permeability
Although the presence of PMNs in the lung in large numbers is regarded as
evidence of a morphological response to O3, the release of chemical mediators by inflammatory
cells indicates a state of activation and represents the functional modification as a consequence
of O3 exposure. Mediators with biological and chemotactic properties have been shown to be
released as a result of stimulation or injury of AMs, epithelial cells, and PMNs. Arachidonic
6-36
-------
acid metabolites play an important role in a variety of processes, including inflammatory
response and permeability changes. Driscoll and Schlesinger (1988) found that, although AMs
isolated from rabbit lungs continually released chemoattractant factors for monocytes and
PMNs, an in vitro exposure of AMs to 0.3 and 1.3 ppm O3 resulted in the increased secretion
of factors that stimulated the migration of PMNs. Driscoll et al. (1988) also found increased
eicosanoid biosynthesis following O3 exposure. In the latter studies, elevated levels of PGE2
and PGF2Q were detected in the supernatant following in vitro exposure of rabbit AMs to 0.3
and 1.2 ppm O3 for 2 h. In a parallel in vivo study, an effect was seen only at the higher
O3 concentration. In vitro exposure in a roller-bottle system of rat AMs to 1 ppm (but not 0.1
ppm) O3 causes stimulation of both cyclooxygenase and lipoxygenase pathways of arachidonic
acid metabolism, as shown by substantial increases in the levels of 6-keto-PGFin, thromboxane
B2 (TXB2), PGE2, LTB4, LTD4, and 15-hydroxy-eicosatetraenoic acid (15-HETE) in the
supernatant of the AM culture (Madden et al., 1991). The authors attribute these effects to
both an increase in the availability of arachidonic acid and a stimulation of cycloxygenase and
lipoxygenase activities. Another in vitro study also demonstrated effects on arachidonic acid
metabolism (Leikauf et al., 1988) . Interpretation of these in vitro studies is difficult. When
Gunnison et al. (1990) compared the effect of in vitro and in vivo exposures of AM to O3 on
eicosanoid metabolism of AMs in culture, a disparity was found. Cultured AMs from
O3-exposed rabbits had a decrease in the elaboration of PGF2n; in vitro exposure caused an
increase.
Changes in the levels of eicosanoids also have been observed in vivo studies.
Schlesinger et al. (1990) found elevation of PGE2 and PGF1D in BAL of rabbits immediately
following a 2-h exposure to 1 ppm O3. Age may play a role. Five-week-old (but not 9-week-
old) mice had increased levels of PGE2 in BAL (Gilmour et al., 1993b). Lower
O3 concentrations did not affect the levels of BAL eicosanoids. Hyde et al. (1992) found an
increase in BAL concentrations of PGF2D, PGD2, and PGF^ following an 8-h exposure of
monkeys to 0.96 ppm O3. Prostaglandin concentrations in the BAL, detected using an
antibody that did not distinguish PGE, from PGE2, also increased with time following a
continuous exposure of mice to 0.5 ppm O3 (Canning et al., 1991). The peak levels of PGE at
3 days were followed by a decline with time, but the levels remained higher than the controls
after 14 days of exposure. The time course of changes in the PGE levels was matched by the
time sequence of changes in BAL protein following exposure to 0.5 ppm O3. Plasma
concentrations of 6-keto-PGFin and PGE, also were elevated in guinea pigs exposed for 1 h to
1 ppm O3 (Miller et al., 1987). Kleeberger et al. (1989) delivered 1 ppm O3 for 5 min to a
lobar bronchus of dogs using a wedged bronchoscope. An analysis of the lavage fluid
collected at 1 min postexposure revealed significant increases in the concentrations of PGD2
and histamine. Although, in this study, the concentration of TXB2 did not change after
O3 exposure, significant increases in concentrations of TXB2 in the plasma and BAL were
observed following acute exposure of guinea pigs to 1 ppm O3 (Miller et al., 1987) and humans
to lower levels of O3 (see Chapter 7). In addition, increased plasma concentrations of PGFin
and PGE, were observed in the guinea pigs exposed to O3. Fouke et al. (1990, 1991) were
unable to detect changes in the BAL concentrations of 6-keto-PGFin, PGE2, TXB2, and PGF2D
in baboons and mongrel dogs exposed to 0.5 ppm O3 for 2 h. The reasons for the lack of this
response in the baboon are not entirely clear, but the lower O3 concentration used in this study,
as compared to the exposure concentrations in dogs and guinea pigs, and species differences
offer possible explanations for the discrepancy. Prostaglandin E, and E2 have been shown to
6-37
-------
influence the inflammatory processes in the lung. Intrabronchial or iv administration of PGE2
was accompanied by increased accumulation of the inflammatory cells in the lung and elevation
of BAL protein (Downey et al., 1988). A possible mechanism involved in the
proinflammatory effects of the PGEs included arteriolar vasodilation without venodilation,
resulting in increased transfer of proteins and cells from blood into the lung by hydrostatic
pressure.
6.2.2.7 Summary
The airway epithelial lining serves as an efficient barrier against penetration of
exogenous particles and macromolecules into the lung tissue and circulation and against entry
of endogenous fluids, cells, and mediators into the air spaces. Disruption of this barrier
following O3 exposure represents a state of compromised epithelial defenses, leading to
increased transepithelial permeability. Inflammatory cells represent another important
component of pulmonary defenses. The recruitment of these cells into the lung following
O3 exposure could result in the release of mediators capable of damaging other cells in the
lung.
Toxicological studies from several laboratories demonstrate alterations in epithelial
permeability and inflammatory responses in animals exposed to O3 concentrations of 1.0 ppm
and below. In these studies, an inflammatory response, as detected by an increase in the
number of PMNs in the BAL or in lung parenchyma, was accompanied by either an increased
tracer transport across the airway mucosa or an elevation in the levels of total protein or
albumin in the BAL. These changes were observed in animals exposed to O3 concentrations as
low as 0.1 ppm in rabbits (2 h/day for 6 days of exposure [Driscoll et al., 1987]; 0.12 ppm in
mice (24 h-exposure [Kleeberger et al., 1993a]) and rats (6-h exposure [Hotchkiss et al.,
1989a] and 24-h exposure [Guth et al., 1986]); and 0.2 ppm in guinea pigs (4-h exposure
[Hatch et al., 1986]). Although monkeys also exhibit inflammatory responses, concentrations
this low have not been tested in this species. The magnitude of response and the time at which
it peaked appeared to vary with O3 concentration, exposure duration, and the mode of analysis.
Investigations of C X T relationships for BAL protein in both rats and guinea pigs showed that
T had increasing influence as C increased (Rombout et al., 1989; Highfill et al., 1992; Highfill
and Costa, 1995; Tepper et al., 1994). However, at low C x T products, similar increases
were observed (Highfill et al., 1992). The responsiveness to O3 also depended on the animal
species tested (Hatch et al., 1986) and increased under certain conditions, such as physical
activity (Van Bree et al., 1992) and pregnancy and lactation (Gunnison et al., 1992b).
To determine the impact of inflammatory cells on O3-induced airway permeability,
rats were exposed to drugs that either destroyed the inflammatory cells or blocked the activity
of their products. Treatment of rats with anti-PMN serum resulted in the depletion of PMNs
but did not affect the increase in BAL protein produced by O3 exposure (Pino et al., 1992b).
Depletion of all the leukocytes by cyclophosphamide or treatment of rats with PG and LT
antagonists resulted in an attenuation of the O3 effects on permeability (Bhalla et al., 1992).
Inflammatory cells, when activated, are capable of releasing mediators with
pathophysiologic and a variety of modulating activities. An increase in the release of
arachidonic acid metabolites following O3 exposure has been shown after both in vitro
(Driscoll and Schlesinger, 1988; Driscoll et al., 1988; Madden et al., 1991; Leikauf et al.,
1988) and in vivo exposures (Schlesinger et al., 1990; Miller et al., 1987; Canning et al.,
1991).
6-38
-------
Some of the effects seen after an acute exposure to O3 are modified on repeated
exposure. The responses following repeated exposures included persistence or an increase in
the number of PMNs or AMs on exposure of rats to 0.5 ppm (2.25 h/day) for 5 days (Tepper
et al., 1989) or exposure of monkeys to 0.64 ppm for 90 days (Moffatt et al., 1987). Other
studies report a reduced inflammatory response following repeated exposures (Chang et al.,
1992; Donaldson et al., 1993). The section on morphometry (Section 6.2.4) has an extended
discussion of microscopically evaluated inflammatory responses.
In brief, these studies show that acute exposures to O3, at concentrations of
0.12 ppm and above, are capable of producing inflammatory and permeability changes in
laboratory animals. It is clear that an assessment of the effects of O3 and interpretation of the
results requires that several factors be taken into consideration; these include O3 concentration,
duration of exposure, exposure conditions (e.g., repeated versus continuous exposure, daytime
versus nighttime exposure, rest versus exercise during exposure), animal species, method of
evaluation, sensitive populations, and time of analysis postexposure.
6.2.3 Effects on Host Defense Mechanisms
6.2.3.1 Introduction
The mammalian respiratory tract has a number of closely integrated defense
mechanisms that, when functioning normally, provide protection from the adverse effects of a
wide variety of inhaled particles and microbes (Green et al., 1977; Kelley, 1990; Schlesinger,
1989; Sibille and Reynolds, 1990). For simplicity, these interrelated defenses can be divided
into two major parts: (1) nonspecific (transport and phagocytosis) and (2) specific
(immunologic) defense mechanisms. A variety of sensitive and reliable methods have been
used to assess the effects of O3 on these components of the lung's defense system to provide a
better understanding of the health effects associated with the inhalation of this pollutant.
The previous Air Quality Criteria for Ozone and Other Photochemical Oxidants
(U.S. Environmental Protection Agency, 1986) provided a review and evaluation of the
scientific literature published up to 1986 regarding the effects of O3 on host defenses. This
section briefly summarizes the existing database through 1986; describes the data generated
since 1986; and, where appropriate, provides interpretations of the data. This section also
discusses the various components of host defenses, such as the mucociliary escalator, the
phagocytic and regulatory role of the AMs, the immune system, and integrated mechanisms
that are studied by investigating the host's response to experimental pulmonary infections.
6.2.3.2 Mucociliary Clearance
This nonspecific defense mechanism removes particles deposited on the mucous
layer of the conducting airways by ciliary action. Ciliary movement directs particles trapped
on the overlying mucous layer toward the pharynx, where it is swallowed or expectorated.
The effectiveness of the mucociliary transport system can be measured by the rate of transport
of deposited particles, the frequency of ciliary beating, and the structural integrity of the cells
that line the conducting airways. Impaired mucociliary clearance can result in an unwanted
accumulation of cellular secretions, increased infections, chronic bronchitis, and complications
associated with chronic obstructive pulmonary disease.
Studies cited in the previous criteria document (U.S. Environmental Protection
Agency, 1986) provided evidence on the effect of O3 on the morphologic integrity of the
mucociliary escalator and its ability to transport deposited particles from the respiratory tract.
6-39
-------
For example, a number of studies with various animal species reported morphologic damage to
the cells of the tracheobronchial tree from exposures to O3 (see Section 6.2.4). The cilia had
become noticeably shorter or were completely absent. Based on such morphologic
observations, related effects such as ciliostasis, increased mucous secretion, and a slowing of
mucociliary transport rates might be expected. Functional studies on mucociliary transport of
deposited particles from the respiratory tract have, in general, observed a delay in particle
clearance in early time periods following acute exposure. For example, a 4-h exposure of rats
to 0.8 ppm O3 slowed early clearance of inhaled latex spheres (Phalen et al., 1980).
Since the publication of the previous criteria document (U.S. Environmental
Protection Agency, 1986), several studies have been performed on the effects of acute
O3 exposure on the mucociliary transport apparatus (Table 6-6). Retarded mucociliary particle
clearance was observed following a 2-h exposure of rabbits to 0.6 ppm O3; extended exposures
(up to 14 days) caused no effects (Schlesinger and Driscoll, 1987). Acute exposure of adult
sheep for 4 h/day for 2 days to 0.5 ppm O3 increased basal secretion of glycoproteins in sheep
trachea, whereas a longer exposure (4 h/day, 5 days/week for 6 weeks) to 0.5 ppm O3 reduced
tracheal glycoprotein secretions (Phipps et al., 1986). In a similar manner, continuous
exposure of ferrets to 1.0 ppm O3 for 3 days increased tracheal gland secretion of
glycoproteins, which remained elevated following 7 days of exposure (McBride et al., 1991).
Because the integrity of the periciliary space is vital for efficient mucociliary action,
O3-induced hyper- or hyposecretion by the mucous glands along the conducting airways can
alter the effectiveness of the mucociliary escalator.
Mariassy et al. (1990) exposed sheep during the first week of life to 1.0 ppm O3 for
4 h/day for 5 days and observed retardation of normal morphologic development of the
tracheal epithelium and a decrease in the tracheal mucous velocity. In a similar manner,
exposure of sheep during the first week of life for 4 h/day for 5 days to 1.0 ppm O3 decreased
epithelial mucosa density and retarded the developmental decrease of tracheal mucous cells
and their carbohydrate composition (Mariassy et al., 1989). Finally, exposure of adult sheep
for either 2 h or for 5 h/day for 4 days to 1.0 ppm O3 decreased tracheal mucous velocity
(Allegraetal., 1991).
6.2.3.3 Alveolobronchiolar Transport Mechanism
In addition to the transporting of particles deposited on the mucous surface layer of
the conducting airways, particles deposited in the deep lung may be removed either up the
respiratory tract or through interstitial pathways to the lymphatic system (Green, 1973). The
pivotal mechanism of alveolobronchiolar transport involves the movement of AMs with
phagocytized particles to the bottom of the mucociliary escalator. Failure of the AMs to
phagocytize and sequester the deposited particles from the vulnerable respiratory membrane
can lead to particle entry into the interstitial spaces. Once lodged in the interstitium, particle
removal is more difficult and, depending on the toxic or infectious nature of the particle, its
interstitial location may allow the particle to set up a focus for pathologic processes. Although
Phalen et al. (1980) and Kenoyer et al. (1981) observed decreases in early (tracheobronchial)
clearance after acute O3 exposure of rats; late (alveolar) clearance was accelerated.
6-40
-------
Table 6-6. Effects of Ozone on Host Defense Mechanisms: Physical Clearancfe
ppm
0.06
base,
spike
rising
0.25
0.1
0.6
1.2
0.1
0.6
1.2
0.1
0.25
0.6
0.5
1.0
1.0
1.0
Ozone
Concentration
Dg/m
118 base,
spike rising
to 490
to
196
1,176
2,352
196
1,176
2,352
196
490
1,176
980
1,960
1,960
1,960
Exposure
Duration
13 h/day,
7 days/week base;
ramped spike
9 h/day,
5 days/week; 6 weeks
2 h/day for 1 or 13 days
2 h/day for 1 or 13 days
2 h/day for 14 days
4 h/day,
5 days/week
for 6 weeks
24 h/day for 7 days
4 h/day for 5 days
2 h and 5 h/day for
4 days
Species, Sex
(Strain)
Age"
Rat, M
(F344)
35 days old
Rabbit, M
(NZW)
2.5-3.0 kg
Rabbit, M
(NZW)
2.5-3.0 kg
Rabbit, M
(NZW)
2.5-2.7 kg
Sheep, F
23-41 kg
Ferret
Sheep
1st week of life
Sheep
26-41 kg
Observed Effect(s) Reference
Increased retention of asbestos fibers in the lung parenchyma. Pinkerton et al. (1989)
Acceleration of early alveolar particle clearance at 0.1 and 0.6 ppm for 13 days. After Driscoll et al. (1986)
single exposure, increased clearance at 0.1 ppm and decrease at 1.2 ppm.
Acceleration of early alveolar particle clearance at 0.1 and 0.6 ppm for 13 days. After Driscoll et al. (1986)
single exposure, increased clearance at 0.1 ppm and decrease at 1.2 ppm
Retarded mucociliary particle clearance at 0.6 ppm only following a single 2-h exposure; Schlesinger and Driscoll (1987)
no effect of 14-day exposure.
Increase of tracheal glycoprotein secretion following acute exposure (4 h/day for 2 days) Phipps et al. (1986)
with a decrease following longer term exposure.
Increased secretion of glycoconjugates by tracheal glands. McBride et al. (1991)
Retardation of normal morphologic development of the tracheal epithelium. Decreased Mariassy et al. (1989, 1990)
tracheal mucous velocity. Decreased tracheal mucosa epithelial density. Retardation of
developmental decrease of tracheal mucous cells and their carbohydrate composition.
Decreased tracheal mucous velocity. Allegra et al. (1991)
"See Appendix A for abbreviations and acronyms.
''Age or body weight at start of exposure.
-------
Exposure of rabbits for 2 h/day for 13 days to 0.1 and 0.6 ppm O3 resulted in
acceleration of early alveolar clearance of polystyrene latex particles (Driscoll et al., 1986).
After a single exposure to 0.1 ppm, the greatest acceleration occurred over the period shortly
after exposure ceased (1 to 4 days), although the effect was still observed at Day 14
postexposure. A single exposure to 0.6 ppm caused no effect, whereas a higher concentration
(1.2 ppm) retarded clearance. To investigate the effects of longer term O3 exposure on
alveolobronchiolar clearance, rats were exposed to an urban pattern of O3 (continuous
0.06 ppm, 7 days/week with a slow rise to a peak of 0.25 ppm and subsequent decrease to
0.06 ppm over a 9-h period for 5 days/week) for 6 weeks and were exposed 3 days later to
chrysotile asbestos, which can cause pulmonary fibrosis and neoplasia (Pinkerton et al., 1989).
Ozone did not affect the deposition of asbestos at the first AD bifurcation, the site of maximal
asbestos and O3 deposition. However, 30 days later, the lungs of the O3-exposed animals had
twice the number and mass of asbestos fibers as the air-exposed rats.
6.2.3.4 Alveolar Macrophages
Within the gaseous exchange region of the lung, the first line of defense against
microorganisms and nonviable particles that reach the alveolar surface is the AM. This
resident phagocyte is responsible for a variety of activities, including the detoxification and
removal of inhaled particles, maintenance of pulmonary sterility, and interaction with
lymphocytes for immunologic protection. Under normal conditions, AMs seek out particles
deposited on the alveolar surface and ingest them, thereby sequestering the particles from the
vulnerable respiratory membrane. If the particle is insoluble, the AMs serve as a repository
for the transport of the particle from the alveolus to the bottom of the mucociliary escalator
located at the far distal portion of the conducting airways. Degradable particles are detoxified
by powerful lysosomal enzymes, whereas microorganisms are killed by biochemical
mechanisms, such as superoxide anion radicals and lysosomal enzymes. To adequately fulfill
their defense function, the AMs must maintain active mobility, a high degree of phagocytic
activity, and an optimally functioning biochemical and enzyme system.
As discussed in the previous criteria document (U.S. Environmental Protection
Agency, 1986), short periods of O3 exposure can cause a reduction in the number of free AMs
available for pulmonary defense, and these AMs are more fragile, less phagocytic, and have
decreased lysosomal enzyme activities. The lowest O3 concentration showing AM effects in
this early work was 0.25 ppm; a 3-h exposure of rabbits decreased lysosomal enzyme activities
(Hurst etal., 1970).
Since the publication of the previous criteria document, the studies performed have
been, in general, confirmatory of previous observations (Table 6-7). Morphologic
observations showed that continuous exposure for 7 days to 0.13, 0.25, 0.5, and 0.77 ppm
O3 resulted in concentration-related increases in the number of rat AMs at 5 days postexposure,
as well as increased AM size and morphologic changes consisting of surface microvilli and
bleb formation (Dormans et al., 1990). Other morphological studies discussed in Section 6.2.4
also show increased numbers of AMs. A 2-h exposure to 0.1 ppm O3 did not affect the AM
number in the BAL when the analyses were made immediately postexposure, but, 7 days later,
the total number of AMs increased by about 70% (Driscoll et al., 1987). On repeated
exposure for 6 or 13 days, the number of AMs increased on the day after exposure. A single
exposure to a higher concentration (1.2 ppm O3) did not affect the number of AMs when
assessed immediately after exposure. It is assumed that, although
6-42
-------
Table 6-7. Effects of Ozone on Host Defense Mechanisms: Macrophage Alteration^
Ozone
Concentration
ppm
0.05
0.1
0.2
0.4
0.1
0.2
0.1
1.2
0.1
1.2
0.11
to
3.6
^ 0.12
4^ 0.8
<-° 1.5
0.12
0.27
0.8
0.12
0.25
0.5
0.13
to
0.77
0.13
0.26
0.51
0.77
0.2
Qg/m
98
196
392
784
196
392
196
2,352
196
2,352
216
to
7,056
235
1,568
2,940
235
529
1,568
235
490
980
250
to
1,500
250
500
1,000
1,500
392
Exposure
Duration
16 h
Continuous for
11 weeks
2h/day for 1, 2,
6, and 13 days
2h
3h
6h
6h
20h/day for 1,
2, 3, 7, and
14 days
Continuous for
7 days
Continuous for
7 days
Continuous for
14 days
Species, Sex
(Strain)
Age1'
Rat, M
(Wistar)
210 ± 10 (SD)g
Rat, M
(Wistar)
16-21 weeks old
Rabbit, M
(NZW)
2-4 mo old
Rabbit
(NZW)
Mouse, M
(Swiss)
Rat, F
(S-D)
Rat, M
(F344)
12-18 weeks old
Rat, F
(F344)
12-14 weeks old
Rat, M
(F433)
Rat, M
(Wistar)
8 weeks old
Rat, M
(Wistar)
8 weeks old
Rat, M
(Wistar)
10 weeks old
Observed Effect(s)
Increased adherence to nylon fibers at 0.05 and 0.1 ppm, but not at 0.2 and 0.4 ppm.
Increases in enzyme activity.
Single exposure: Decreased phagocytosis at 0. 1 and 1 .2 ppm immediately and 1 day after exposure; recovery
by 7 days after exposure in the 0.1 -ppm group. Multiple exposure to 0.1 ppm only: decreased phagocytosis
1 day after 2 and 6 days of exposure.
Increased release of PGE, and PGF^ at 1 .2 ppm; no effect at 0. 1 ppm.
Concentration-dependent decrease in superoxide anion radical production; mouse more sensitive. No effect
on murine AM phagocytosis at 0.42, 0.95, 1.0, and 1.2 ppm.
At 42 and 66 h PE, concentration-dependent increase in mitotic index beginning at 0.8 ppm;
increased size at 1.5 ppm at 18 and 42 h PE.
Increase in mitotic index and chromosome damage at 0.27 and 0.8 ppm, no effect at 0.12 ppm.
Increase in AM DNA synthesis at 2 and 3 days at 0.25 ppm O3 and at 1,2, and 3 days at
0.50 ppm O3.
Concentration-related effects on number, size, and surface morphology.
Decreased phagocytic ingestion (at all concentrations) and intracellular killing (at 00.26 ppm) ofListeria
monocytogenes.
Increases in enzyme activity at D3 days. Increased number of AMs by Day 3.
Reference
Veninga and Evelyn
(1986)
Mochitate et al. (1992)
Driscoll et al. (1987)
Driscoll et al. (1988)
Ryer-Powder et al. (1988)
Amoruso et al. (1989)
Hotchkiss et al. (1989b)
Rithidech et al. (1990)
Wright et al. (1987)
Dormans et al. (1990)
Van Loveren et al. (1988)
Mochitate and Miura
(1989)
-------
Table 6-7 (cont'd). Effects of Ozone on Host Defense Mechanisms:
Macrophage Alterations3
ppm
0.4
0.4
0.4
0.8
0.4
0.8
0.4
0.8
0.5
0.5
0.64
1.0
Ozone
Concentration
Qg/m
784
784
784
1,568
784
1,568
784
1,568
980
980
1,254
1,960
Exposure
Duration
3, 6, or 12 h;
12h/day for 1, 3, or
7 days
6 or 12 h, or
12 h/day for 3 or
7 days
3h
3h
3h
24 h/day for 14
days
24 h/day for 14
days
23 h/day for 27
days
2 h/day for 3 days
Species, Sex
(Strain)
Age1'
Mouse, M (NIH)
23-28 g
Rat, M
(Wistar)
180-200 g
Mouse, M (NIH)
25-30 g
Rat, M
(Wistar)
200-250 g
Mouse, F
(C3H/JM
C57B1/6)
30 days old
Mouse, F
(CD-I)
5 and 9 weeks old
Mouse, F
(C3H/HeJ
C57B1/6)
30 days old
Rat, F (F344)
Mouse, F
(Swiss)
20-25 g
Mouse, F
(Swiss)
20-25 g
Rat, M
(S-D)
130-150 g
Rabbit, M
(NZW)
14 weeks old
Observed Effect(s) Reference
Decreased Fc-receptor mediated phagocytosis of AMs from mice at Days 1 and 7. Decreased Costing et al. (1991a)
phagocytosis at 6 h, increased phagocytosis of AMs from rats on Day 1 . Decrease in superoxide
anion production at some exposures.
Increased ATP levels in mouse AMs only after 7 days of exposure. Oosting et al. (1991b)
Decreased phagocytosis of Streptococcus zooepidemicus (in vivo assay) and latex beads (in vitro Gilmour et al. (1993a)
assay). C57B1 mice more susceptible than C3H at 0.4 ppm only for bacteria phagocytosis,
reverse susceptibility for latex beads.
Decreased in vivo phagocytosis of Streptococcus zooepidemicus independent of age. Decrease in Gilmour et al. (1993b)
number of phagocytic cells and number of bacteria/ AM.
Decreased phagocytosis of latex beads. Gilmour and Selgrade (1993)
Decreased Fc-receptor mediated phagocytosis of AMs at Days 1-14, with trend towards maximal Canning et al. (1991)
decrease on Day 3. Effect correlated with increases in PGE. Decreased phagocytosis of
peritoneal AMs.
Decreased Fc-receptor mediated phagocytosis on Days 1, 3, and 7, but not at Day 14. Gilmour et al. (1991)
Decreased lysozyme enzyme content during chronic Pseudomonas aeruginosa bacterial infection. Sherwood et al. (1986)
Decreased cytotoxicity vs. xenogeneic tumor cells. No effect on TNF-D and H2O2 production. Zelikoff et al. (1991)
Depression of superoxide anion production immediately after exposure, with an increase at 24 h.
"See Appendix A for abbreviations and acronyms.
''Age or body weight at start of exposure.
-------
the lower O3 concentrations are stimulatory for AMs activity, the higher concentrations are
inhibitory because of their ability to produce substantial cellular injury. The number of AMs
in BAL is reduced after a single exposure to 0.8 or 1.8 ppm O3 (Bassett et al., 1988b; Bhalla
and Young, 1992; Donaldson et al., 1993). Hotchkiss et al. (1989b) also observed a slight
decrease in the number of AMs in the BAL immediately after a 6-h exposure to 0.8 and
1.5 ppm O3, but the number of AMs increased at 42 and 66 h after the exposure. The results
of the studies on AMs in general indicate that, following O3 exposure, there is a short-term
reduction in the number of lavageable AMs, but these cells increase in number over several
days following exposure.
Exposure of rats to 0.12, 0.8, and 1.5 ppm O3 for 6 h also resulted in a
concentration-dependent increase in mitotic index at 0.8 ppm at 92 h after postexposure; AM
size was only increased at the highest concentration (18 and 48 h postexposure) (Hotchkiss
et al., 1989b). In a similar study, exposure of rats for 6 h to 0.12, 0.27, and 0.8 ppm
O3 resulted in AM chromosome damage at the two higher concentrations 28 h after exposure
(Rithidechetal., 1990).
Several studies have investigated the effect of O3 exposure on AM phagocytosis.
Exposure of C3H/HeJ and C57B1/6 mice for 3 h to 0.4 and 0.8 ppm O3 decreased AM
phagocytosis of Streptococcus zooepidemicus and latex beads (Gilmour et al., 1993a). In a
similar study, a 3-h exposure of 5- and 9-week-old CD-I mice decreased AM phagocytosis of
S. zooepidemicus, but there was no effect of age (Gilmour et al., 1993b). Decreased
phagocytic ingestion of Listeria monocytogenes also was observed following continuous
exposure of rats to 0.13, 0.26, 0.51, and 0.77 ppm for 7 days; only the two lower
concentrations inhibited intracellular killing (Van Loveren et al., 1988). Although lower
O3 concentrations were not tested, rats exposed to 1.02 ppm O3 were unable to clear Listeria
from their lungs. Exposure of rabbits to 0.1 ppm O3 for 2 h/day resulted in decreased AM
phagocytosis of latex microspheres after 2 or 6 (but not 13) days of exposure (Driscoll et al.,
1987). In the same study, a single exposure to 0.1 or 1.2 ppm decreased AM phagocytosis
immediately after exposure; recovery occurred by 7 days postexposure in the 0.1-ppm group
but not in the 1.2-ppm group. That repeated exposures to O3 results in an initial suppression
of AM phagocytosis, which is followed by recovery of phagocytic potential while exposure
continues (Driscoll et al., 1987), was confirmed by the studies of Gilmour et al. (1991) and
Canning et al. (1991). Using identical exposure systems, it was observed that continuous
exposure of mice to 0.5 ppm O3 decreased AM Fc-receptor-mediated phagocytosis of sheep
erythrocytes on Days 1, 3, 5, 7, and 8 of exposure, with return to control phagocytic levels by
Day 14. This temporal trend paralleled the pattern of O3-induced reduction of lung bactericidal
activity against Staphylococcus aureus.
Interspecies comparisons of AM phagocytic potential were made by Gilmour and
Selgrade (1993), who exposed C3H/HeJ and C57B1/6 mice and Fischer 344 rats to 0.4 and
0.8 ppm O3 for 3 h. Alveolar macrophage phagocytosis of latex beads was suppressed in all
animals immediately after 0.4 ppm O3 exposure, with the percent suppression greater in both
strains of mice as compared to similarly treated rats. No differences in phagocytic suppression
were observed between 0.4- and 0.8-ppm-O3-exposed rats or the C57B1/6 mice, but
phagocytosis by AMs from 0.8-ppm-O3-exposed C3H/HeJ mice was more suppressed as
compared to the 0.4-ppm-O3-exposed group. In a similar comparative study, Oosting et al.
(199la) exposed mice and rats to 0.4 ppm O3 for single (3,6, and 12 h) and repeated (12 h/day
for 7 days) regimens. A decrease was observed in rat and mouse AM Fc-receptor mediated
6-45
-------
phagocytosis following the single O3 exposure protocol. With the repeated O3 exposure
protocol, rat AM phagocytosis was increased a day after exposure with no significant changes
on Days 3 and 7. In contrast, phagocytosis by mouse AMs was suppressed at Day 1 of
exposure and still did not recover at Day 7. In the same study, when mice were allowed to
recover for 4 days following 3 days of O3 exposure, phagocytosis by AMs was increased.
These interspecies comparisons on the effect of O3 exposure on AM phagocytic potential
indicate that mice may be more susceptible than rats.
A species comparison of superoxide anion radical production between mouse AMs
and rat AMs following a single 3-h exposure to O3 concentrations ranging from 0.11 to
3.6 ppm showed the O3 concentration that inhibits superoxide anion radical production by 50%
to be 0.41 ppm for mouse AMs and 3.0 ppm for rat AMs (Ryer-Powder et al., 1988). Oosting
et al. (199la) also found a species difference in superoxide anion production using a more
varied exposure-duration protocol; mice appeared to be more responsive than rats. This
oxygen radical is important in antibacterial activity, and both sets of authors suggest that
O3-induced impairment of pulmonary antibacterial defenses may be related to decreases in
superoxide anion radical production. Decreased lysozyme enzyme levels in rat AMs also were
observed during chronic Pseudomonas aeruginosa bacterial infection following exposure to
0.64 ppm O3 for 23 h/day for 27 days (Sherwood et al., 1986).
Exposure of rats for 16 h to 0.05, 0.1, 0.2, and 0.4 ppm O3 increased AM
adherence to nylon fibers at 0.05 and 0.1 ppm, but had no effect at 0.2 and 0.4 ppm (Veninga
and Evelyn, 1986). Increased metabolic activity of AMs retrieved from rats following
continuous O3 exposure for 14 weeks to 0.1 and 0.2 ppm was observed by Mochitate et al.
(1992). In a similar study, long-term O3 exposure of rats (continuous 0.2 ppm for 11 weeks)
continued to increase AM metabolic activity (Mochitate and Miura, 1989). Exposure of mice
and rats for 14 h/day for 7 days to 0.4 ppm O3 also increased adenosine triphosphate (ATP)
levels in the mouse AMs, but had no effect on ATP levels in rat AMs (Oosting et al., 1991b).
In addition to their phagocytic function and particle removal, AMs also play several
other roles in host defense that include (1) a regulatory role through their release of mediators
(soluble substances secreted by the AMs that produce biologic effects on other cells) such as
tumor necrosis factor, interleukin-1, and PGs; (2) activities associated with tumor surveillance;
and (3) accessory cell function in antigen presentation to lymphocytes in the initiation of the
immune response. Investigating the effect of a single 2-h exposure of rabbits to 0.1 and
1.2 ppm O3, Driscoll et al. (1988) observed an increased release of PGE2 and PGF2D by AMs
following exposure to 1.2 but not 0.1 ppm. Prostaglandin E2 can depress AM and natural
killer cell cytotoxicity to tumor cells. Perhaps this is a mechanism involved in the depression
of AM-mediated cytotoxicity toward xenogeneic tumor cells following exposure of rabbits for
2 h/day for 3 days to 1.0 ppm O3 (Zelikoff et al., 1991). No studies were found on the effects
of O3 on antigen presentation.
6.2.3.5 Immunology
In addition to the above nonspecific defense mechanisms, the respiratory system
also has specific immunologic mechanisms that can be initiated by inhaled antigens. There are
two types of immune mechanisms: antibody (humoral)-mediated and cell-mediated.
In general, humoral mechanisms neutralize viruses and microbial toxins, enhance the ingestion
of bacteria by phagocytes, and play an important role in defense of the lung against fungal and
parasitic infections. Cell-mediated mechanisms enhance the microbiocidal capacity of AMs in
6-46
-------
defense to intracellular bacteria such as Mycobacterium tuberculosis and Listeria
monocytogenes, whereas another arm of the cellular immune response generates a class of
lymphocytes that are cytotoxic for virus-infected cells. Both the humoral and cell-mediated
responses protect the respiratory tract against infectious agents and operate in three major
temporal waves: (1) natural killer cells (nonspecific lymphocytes that can destroy bacteria,
viruses, and tumor cells), (2) cytotoxic T lymphocytes (lymphocytes that lyse specifically
recognized targets), and (3) antigen-specific antibodies.
Little information was available in the previous criteria document (U.S.
Environmental Protection Agency, 1986) on the effects of O3 on immunologic defenses.
However, the data base indicated an immunotoxic effect of O3 exposure, especially on T-cell
populations. For example, Aranyi et al. (1983) found that a 90-day (5 h/day, 5 days/week)
exposure of mice to 0.1 ppm O3 suppressed blastogenesis of splenic lymphocytes to T-cell, but
not B-cell mitogens; the ability of these cells to produce antibodies was not affected either.
As can be seen from Table 6-8, this database has greatly expanded and also has been recently
reviewed (Jakab et al., 1995). Many of the studies include both the pulmonary and systemic
immune system, which, to a degree, are compartmentalized; both systems are discussed here.
Studies on the effect of O3 exposure on the immune system can be divided into three
broad categories. These are (1) measurement of lymphoid organ weights and cellular
composition, (2) determination of the functional capacity of lymphocytes in the absence of
antigenic stimulation, and (3) measurement of the immune response following antigenic
stimulation.
Dziedzic and White (1986a) observed that exposure of mice to 0.3, 0.5, and
0.7 ppm O3 for 20 h/day for 28 days resulted in a concentration-dependent initial depletion of
cells in the mediastinal lymph nodes (MLNs) (Days 1 and 2); this was followed by a T-cell
hyperplasia peaking about Days 3 and 4. There was an enhanced blastogenic response to the
T-cell mitogen concanavalin A (ConA) at 0.7 ppm O3 (only level tested). There was no effect
of O3 on cell division morphology of B cells. In a similar study, exposure of mice to 0.7 ppm
O3 for 20 h/day for 28 days resulted in an initial thymic atrophy, with return to normal thymus
weights by Day 14 of exposure (Dziedzic and White, 1986b). Exposure of rats to 0.5 ppm
O3 for 20 h/day for 14 days also increased bronchus-associated lymph node and MLN cell
proliferation at 2 and 3 days of exposure, but not at 1, 7, and 14 days of exposure (Dziedzic
et al., 1990). Bleavins and Dziedzic (1990) observed that exposure of BALB/c mice to
0.7 ppm O3 for 20 h/day for 14 days resulted in decreased spleen and thymus weights at Day
4, with recovery at Day 14. The absolute number of thymocytes decreased following exposure
of mice to 0.7 ppm O3 for 24 h/day for 7 days (Li and Richters, 1991a) and to 0.3 ppm O3 for
24 h/day for 3 weeks (Li and Richters, 1991b). Although the latter exposure protocol
(0.3 ppm for 24 h/day for 3 weeks) decreased the absolute number of thymocytes, an increase
in the percentage of thymocytes was observed in the absence of any changes in splenic T cells
(Li and Richters, 1991b). Continuous exposure of rats to DO.26 ppm O3 for 7 days increased
MLN T:B lymphocyte ratios immediately and 5 days postexposure (Van Loveren et al., 1988).
Bleavins and Dziedzic (1990) observed an increased infiltration of Thy-1.2+ lymphocytes and
IgM+ cells into the O3-induced pulmonary lesion following exposure of mice to 0.7 ppm O3 for
20 h/day for 14 days. The lesion was defined by quantitative histomorphometric analysis as
any lung area with inflammatory cell infiltration, cellular proliferation, consolidation, or
edema. Dziedzic and White (1987a) further investigated these T-cell effects by exposing
normal and athymic nude
6-47
-------
Table 6-8. Effects of Ozone on Host Defense Mechanisms: Immunolog^
Ozone
Concentration
ppm
0.06 base,
spike rising
to 0.25
0.1
0.5
1.0
0.13
0.26
0.51
0.77
1.02
0.2
0.4
0.82
0.3
0.3
0.5
0.7
0.5
0.5
0.7
0.7
0.7
Dg/m
118 base,
spike rising
to 490
196
980
1,960
250
500
1,000
1,500
2,000
400
800
1,600
588
588
980
1,372
980
980
1,372
1,372
1,372
Exposure
Duration
13 h/day, 7 days/week
base; ramped spike
9 h/day, 5 days/ week
23.5 h/day for 10 days
Continuous for 7 days
24 h/day for 7 days
24 h/day for 3 weeks
20 h/day for 28 days
24 h/day for 14 days
20 h/day for 14 days
20 h/day for 28 days
24 h/day for 7 days
20 h/day for 7 and
14 days
Species, Sex
(Strain)
Age1'
Rat, M
(F344)
60 days old
Rat, M
(F344)
8-12 weeks old
Rat, M
(Wistar)
8 weeks old
Rat, M
(Wistar)
NS
Mouse, F
(BALB/c)
3 weeks old
Mouse, F
(CD-I)
Mouse, F
(Swiss)
Rat, M
(F344)
1 1 weeks old
Mouse, F
(CD-I)
Mouse, F
(BALB/c)
5-12 weeks old
Mouse, F
(Athymic and
euthymic)
Observed Effect(s)
No effect on splenic NK cell activity or splenic lymphocyte blastogenic response to T-cell mitogens
(PHA and ConA) or B-cell mitogen (S. salmonella glycoprotein).
Decreased lung NK cell activity at 1.0 ppm on Days 1, 5, and 7, with recovery at Day 10; decreased
NK cell activity at 0.5 ppm on Day 1 (only day tested).
At DO. 26 ppm, increased T:B cell ratios in MLN. Decreased T:B cell ratios in MLN and delayed-
type hypersensitivity response following Listeria monocytogenes immunization.
Decreased lung NK cell activity at 0.82 ppm; increased lung NK cell activity at 0.2 and
0.4 ppm.
Increase in percentage of thymocytes with lower absolute numbers. No changes in splenic
T-lymphocytes.
Concentration- dependent MLN cell depletions (Days 1 and 2) followed by T-cell hyperplasia (Days
4, 7, 14, and 28). Enhanced blastogenic response to T-cell mitogen ConA at 4 and 7 days at 0.7 ppm
(other levels not tested).
Decreased antiviral serum antibody following influenza virus infection and decreased T-cells in lung
tissue.
BALT and MLN cell proliferation at 3 days, but not at Days 1, 2, 7, and 14.
MLN hyperplasia. Thymic atrophy through Day 7 with return to control level at Day 14.
Decrease in the absolute number of thymocytes.
Compared to euthymic (nu/+) mice, athymic nude (nu/nu) mice have increased lung
inflammation and lesion volume, but no MLN hyperplasia.
Reference
Selgrade et al. (1990)
Burleson et al. (1989)
Van Loveren et al. (1988)
Van Loveren et al. (1990)
Li and Richters (1991b)
Dziedzic and White (1986a)
Jakab and Hmieleski (1988)
Dziedzic et al. (1990)
Dziedzic and White (1986b)
Li and Richters (1991a)
Dziedzic and White (1987a)
-------
Table 6-8 (cont'd). Effects of Ozone on Host Defense Mechanisms: Immunolog^
Ozone
Concentration
ppm Qg/m
0.7 1,372
0.8 1,568
0.8 1,568
0.8 1,568
1.0 1,960
1.0 1,960
Exposure
Duration
20 h/day for 14 days
23 h/day for 14 days
24 h/day for 14 days
24 h/day for
4 weeks
24 h/day for
4 weeks
8 h/day for 7 days
Species, Sex
(Strain)
Age1'
Mouse, F
(BALB/c)
10-12 weeks old
Mouse, F
(B6C3F,)
6-8 weeks old
Mouse, M
(BALB/c)
2-3 mo old
Mouse, M
(BALB/c)
7 weeks old
Mouse, M
(C57B/6 x
DBA/2 Fl)
11-14 weeks old
Rat, M
(Long-Evans)
7-8 weeks old
Observed Effect(s)
Decreased spleen and thymus weight at Day 4 with recovery at Day 14. Increased
infiltration
of Thy-1.2+ lymphocytes and IgM+ cells into the O3-induced pulmonary lesion.
Decreased blastogenesis of MLN and splenic lymphocytes to PHA mitogen on Day 1 ;
no effect on Days 3, 7, and 14. Decreased splenic NK cell activity on Days 1 and 3;
no effect on Days 7 and 14. Decreased pulmonary IgG and IgA response to ovalbumin
immunization.
Suppression of delayed-type hypersensitivity response to SRBCs on Day 7; no effect on
Days 1, 3, and 14. Thymic atrophy on Day 7 with decreased Thy-1.2+ cells in thymus.
Suppression of serum IgG following ovalbumin immunization.
Decreased ability of spleen cells to generate a primary antibody response to SRBCs in vitro.
No effect on lung and splenic ADCC activity.
Enhanced blastogenesis of splenic lymphocytes to PHA, ConA, and lipopolysaccharide
mitogens at 1 .0 ppm.
Reference
Bleavins and Dziedzic (1990)
Gilmour and Jakab (1991)
Fujimaki et al. (1987)
Ozawa (1986)
Wright et al. (1989)
Eskew et al. (1986)
"See Appendix A for abbreviations and acronyms.
''Age or body weight at start of exposure.
-------
mice to 0.7 ppm O3 for 7 or 14 days (20 h/day). After 7 days of exposure, the athymic nude
mice did not have the MLN hyperplasia seen in the euthymic mice. However, the athymic
nude mice had a greater inflammatory response and an increase in lung lesion volumes
compared to the euthymic mice. Thus, it appears that T cells have some involvement in
protecting the lungs from the morphological effects of O3.
The above longitudinal studies on the effects of O3 exposure on lymphoid organ cell
numbers provide information on cellular traffic and cell numbers but provide few insights into
the functional capacity of the lymphocytes. A number of studies have investigated the effect of
O3 exposure on the blastogenic response of lymphocytes to nonspecific mitogens. These assays
measure nonspecific clonal expansion of the lymphocyte population, a critical step during the
amplification of the immune response. Exposure of mice to 0.7 ppm O3 for 20 h/day for 28
days enhanced the MLN cell blastogenic response to the T-cell mitogen ConA at 4 and 7 days
of exposure, with return to control levels by Day 14 (Dziedzic and White, 1986a). In a similar
manner, Gilmour and Jakab (1991) observed that continuous exposure of mice to 0.8 ppm
O3 decreased the MLN and splenic lymphocyte blastogenic response to the T-cell mitogen
phytohemagglutinin (PHA) on Day 1 of exposure, with the effect abrogated after prolonged
exposure. An enhanced blastogenic response of splenic lymphocytes to PHA and Con A and a
B-cell mitogen (Escherichia coli lipopolysaccharide) was observed following exposure of rats
to 1.0 ppm O3 for 8 h/day for 7 days (Eskew et al., 1986).
Natural killer cell activity also has been studied. One such study in rats observed a
decreased lung NK cell activity following 1,5, and 7 days of exposure to 1.0 ppm for
23.5 h/day, with recovery by Day 10 (Burleson et al., 1989). In a similar experiment,
Van Loveren et al. (1990) observed decreased lung NK cell activity following 7 days of
continuous exposure of rats to 0.82 ppm O3; however, exposure to 0.2 and 0.4 ppm enhanced
lung NK cell activity. Exposure of mice for 23 h/day for 14 days to 0.8 ppm O3 also
decreased splenic NK cell activity on Days 1 and 3, with a return to control values on Days 7
and 14 (Gilmour and Jakab, 1991). Finally, Selgrade et al. (1990) used an experimental
protocol designed to mimic diurnal urban O3 exposure patterns. Rats were exposed to a
background level of 0.06 ppm for 13 h (7 days/week), followed by a broad exposure spike
(5 days/week) rising from 0.06 to 0.25 ppm and returning to 0.06 ppm over 9 h, and then
followed by 2-h downtime. After 1, 3, 13, 52, or 78 weeks of exposure, spleen cells were
assessed for NK cell activity and responses to T-cell mitogens (PHA and ConA) and a B-cell
mitogen (Salmonella typhimurium glycoprotein). Ozone exposure had no effect on NK cell
activity, nor were there any O3-related changes in mitogen responses in splenic or blood
leukocytes. There were also no effects of a single 3-h exposure to 1.0 ppm O3 on spleen cell
responses to the mitogens immediately after exposure or at 24, 48, and 72 h thereafter.
Several studies also have investigated the effect of O3 exposure on the immune
response following antigenic stimulation. Fujimaki (1989) observed that exposure of mice to
0.8 ppm O3 for 24 h/day for 56 days suppressed the primary splenic antibody response to
sheep red blood cells (SRBCs; T-cell-dependent antigen) but not to DNP-Ficoll (T-cell-
independent antigen). In a similar study, exposure of mice to 0.8 ppm O3 for 24 h/day for 14
days suppressed the delayed type hypersensitivity response to SRBCs on Day 7, but not on
Days 1, 3, and 14 (Fujimaki et al., 1987). Suppression of serum IgG levels on ovalbumin
immunization was observed following exposure of mice to 0.8 ppm O3 for 24 h/day for
4 weeks (Ozawa, 1986). Decreased pulmonary IgG and IgA responses on ovalbumin
immunization were also observed in mice during a 2-week O3 exposure for 23 h/day to
6-50
-------
0.8 ppm (Gilmour and Jakab, 1991). Exposure of mice to 0.5 ppm O3 continuously for
14 days during the course of influenza virus infection also decreased the serum hemagglutinin
antiviral antibody response (Jakab and Hmieleski, 1988).
Van Loveren et al. (1988) investigated antigen-specific responses following
pulmonary Listeria infection and observed no significant changes in the delayed-type
hypersensitivity response when O3 exposure (continuous, 0.77 ppm) was for 7 days prior to
infection. However, if the O3 exposure took place when an infection with Listeria was also
present (from Days 0 to 7 or from Days 7 to 14), the delayed-type hypersensitivity response
was significantly decreased. In a similar manner, no significant changes were observed in the
splenic lymphoproliferative response to Listeria antigen when the 0.77-ppm O3 exposure
preceded the infection, whereas the response was suppressed when the 0.77-ppm O3 exposure
occurred immediately after infection or from Days 7 to 14 after infection. In the same series
of experiments, Van Loveren et al. (1988) observed that continuous exposure to 0.26 ppm
O3 impaired the increase in T:B lymphocyte ratios that occurred in response to the Listeria
infection.
6.2.3.6 Interaction with Infectious Agents
Because respiratory infections remain one of the most common public health
problems, it is important to determine whether or not exposure to air pollutants reduce
susceptibility to infectious agents. Measurement of the competence of the host's antimicrobial
mechanisms can best be tested by challenging air-pollutant-exposed animals and the clean-air-
exposed control animals to an aerosol of viable organisms. If the test substance, such as O3,
decreases the efficiency of the host's integrated protective mechanisms (i.e., physical clearance
via the mucociliary escalator, microbicidal activity of the AMs, and associated humoral and
cellular immunologic events), the microorganisms are less efficiently killed in the lungs, or the
organisms may even multiply, resulting in the demise of the host. The defensive function of
the lung is remarkably similar across animals species, and available human data suggests that
qualitative findings obtained on functional resistance mechanisms using appropriate animal
models may be extrapolated to humans (Green, 1984).
The studies detailed in the previous criteria document (U.S. Environmental
Protection Agency, 1986) primarily used the mouse "infectivity model" (Gardner, 1982).
Briefly, animals are randomly selected to be exposed to either clean air or O3. After exposure,
the animals from both groups are combined and exposed to an aerosol of microorganisms. The
vast majority of these studies have been conducted with streptococcus species. At the
termination of the infectious exposure period, the animals are housed in clean air and the
mortality rate in the two groups is determined during a 15-day holding period. In this system,
the concentrations of O3 used do not cause any mortality. The mortality in the control group
(clean air plus exposure to the microorganism) ranges from approximately 10 to 20% and
reflects the natural resistance of the host to the infectious agents. The difference in mortality
between O3-exposure groups and the controls is concentration-related (Gardner, 1982). These
studies showed that, depending on the O3 exposure protocol, a 3-h exposure to concentrations
as low as 0.08 ppm O3 can enhance the increased mortality of CD-I mice from streptococcus
infection (Coffin et al., 1967; Coffin and Gardner, 1972; Miller et al., 1978). However,
although a prolonged intermittent exposure (103 days) to 0.1 ppm O3 increased mortality in
this model system, the magnitude of the effect was not substantially greater than that after
acute exposure (Aranyi et al., 1983).
6-51
-------
Another approach to assess the effect of air pollutants on host defenses is to
quantitate rates of pulmonary bacterial inactivation following aerosol infection with
microorganisms. In this system, the animals are exposed either to clean air or to the air
pollutant and then are exposed to an aerosol of microorganisms in a manner similar to the
method used for the infectivity model. However, instead of assessing enhancement of
mortality, viable bacteria are quantitated in lung homogenates at various times after inhalation
of the microorganisms (Goldstein et al., 1971a,b). In air-exposed control animals, there is a
rapid inactivation of the inhaled microorganisms that have been deposited in the respiratory
tract. However, O3 exposure alters the ability of the microbicidal mechanisms of the lungs to
function normally and bacterial inactivation proceeds at a slower rate, indicating impairment of
host defenses. For example, Goldstein et al. (1971b) showed that a 4-h exposure of mice to
DO.6 ppm O3 after infection with S. aureus decreased lung bactericidal activity. Studies
appearing in the literature since publication of the previous criteria document (U.S.
Environmental Protection Agency, 1986) are described below (also see Table 6-9).
Gilmour et al. (1993a) observed that exposure of C3H/HeJ and C57B1/6 mice for
3 h to 0.4 and 0.8 ppm O3 resulted in decreased intrapulmonary killing of S. zooepidemicus in
both strains of mice. Although both strains were affected, the C3H/HeJ mice appeared to be
more susceptible because bactericidal activity was decreased sooner and mortality was
enhanced more. Gilmour et al. (1993b) expanded these studies to CD-I mice of different ages
(5 and 9 weeks old) exposed for 3 h to 0.4 and 0.8 ppm O3. The higher
concentration decreased intrapulmonary killing 4 h after infection with S. zooepidemicus; there
was no effect of age. However, the 5-week-old mice were more susceptible to the infection
because mortalities were 9, 41, and 61% in the air, 0.4-ppm, and 0.8-ppm exposure groups,
respectively; whereas only 4, 15, and 28% of the older animals died with analogous exposure.
Pretreatment of the mice with indomethacin reduced the O3-induced enhancement of PGE2
levels in BAL as well as the enhanced mortality in the 5-week-old mice, suggesting an
involvement of arachidonic acid metabolites in antibacterial defenses.
Gilmour and Selgrade (1993) studied the interspecies response to experimental
S. zooepidemicus infection of rats and C3H/HeJ and C57B1/5 mice following a 3-h exposure to
0.4 and 0.8 ppm O3. Exposure of rats to O3 suppressed intrapulmonary bacterial killing, with
no differences observed between the 0.4- and 0.8-ppm O3 exposure groups. Exposure of
C57B1/6 mice to 0.4 ppm O3 also resulted in a suppression of bactericidal activity, and
exposure to 0.8 ppm O3 led to bacterial proliferation in the lungs, resulting in 60% mortality at
Day 4. Exposure of C3H/HeJ mice to both 0.4 and 0.8 ppm O3 resulted in bacterial
proliferation with, respectively, 60 and 80% mortality at Day 4 after exposure. Increased
mortality from S. zooepidemicus infection following 24 h/day exposure for 5 days/week for
3 weeks also was observed following 0.3 and 0.5 but not 0.1 ppm O3 exposure (Graham et al.,
1987).
To investigate the effect of longer exposures and challenges with bacteria, Gilmour
et al. (1991) exposed mice continuously to 0.5 ppm O3 for 14 days. At 1, 3, 7, and 14 days,
intrapulmonary killing was assessed by inhalation challenge with S. aureus and Proteus
mirabilis. Ozone exposure impaired the intrapulmonary killing of S. aureus at 1 and 3 days.
However, with prolonged exposure, the bactericidal capacity of the lungs returned to normal.
In contrast to S. aureus, when P. mirabilis was the challenge organism, O3 exposure had no
suppressive effect on pulmonary bactericidal activity. The authors attribute this difference to
the defense mechanisms involved. Alveolar macrophages are
6-52
-------
Table 6-9. Effects of Ozone on Host Defense Mechanisms:
Interactions with Infectious Agents1
Ozone
Concentration
ppm
0.1
0.3
0.5
0.13
0.26
0.51
0.77
1.02
0.25
0.5
1.0
0.4
0.8
0.4
<^ 0.8
Ln
0.4
0.8
0.5
0.5
0.5
Qg/m
196
588
980
250
500
1,000
1,500
2,000
490
980
1,960
784
1,568
784
1,568
784
1,568
980
980
980
Exposure
Duration
24 h/day,
5 days/week for
3 weeks
Continuous for
7 days
3 h/day for 5 days
3h
3h
3h
24 h/day for
14 days
24 h/day for
15 days
24 h/day for
120 days
Species, Sex
(Strain)
Age1'
Mouse, F
(CD-I)
4-6 weeks old
Rat, M
(Wistar)
8 weeks old
Mouse, F
(CD-I)
3-4 weeks old
Mouse, F
(C3H/HeJ
C57B1/6)
30 days old
Mouse, F
(CD-I)
5 and 9 weeks old
Mouse, F
(C3H/HeJ
C57B1/6)
Rat
(F344)
30 days old
Mouse, F
(Swiss)
20-23 g
Mouse, F
(Swiss)
20-23 g
Mouse, F
(Swiss)
20-23 g
Observed Effect(s)
Increased mortality from Streptococcus zooepidemicus infection at 0.3 and 0.5 ppm.
Decreased bactericidal activity vs. Listeria monocytogenes at 0.77 and 1.02 ppm. Increased
mortality at 1.02 ppm.
Increased mortality from influenza virus infection and increased pulmonary virus liters at 1 .0 ppm
when infection followed 2 days of exposure, but not at other time points. Histopathologic and
pulmonary function changes more severe with this regimen. At DO. 5 ppm, increased lung wet
weight when virus given after 2 days of O3 exposure.
Decreased bactericidal activity vs. Streptococcus zooepidemicus. Increased mortality in both
strains, with greater mortality in the C3H/HeJ strain.
Decreased bactericidal activity vs. Streptococcus zooepidemicus. Increased mortality in both age
groups, with greater mortality in the 5-week-old mice.
Decreased intrapulmonary killing of Streptococcus zooepidemicus.
Decreased bactericidal activity vs. Staphylococcus aureus on Days 1 and 3; no effect on Days 7
and 14. Enhanced bactericidal activity vs. Proteus mirabilis on Days 3, 7, and 14; no effect on
Day 1.
No effect on pulmonary virus liters during influenza virus infection. O3 decreased lung
morphological injury due lo virus (Day 9).
Ozone decreased acule lung influenzal injury, bul increased pulmonary fibrosis during ihe course
of and period after influenza virus infeclion.
Reference
Graham el al. (1987)
Van Loveren el al. (1988)
Selgrade el al. (1988)
Gilmour el al. (1993a)
Gilmour el al. (1993b)
Gilmour and Selgrade (1993)
Gilmour el al. (1991)
Jakab and Hmieleski (1988)
Jakab and Bassell (1990)
"See Appendix A for abbrevialions and acronyms.
''Age or body weighl al slarl of exposure.
-------
active against the gram-positive S. aureus; AMs and PMNs defend against the gram-negative
P. mirabilis. The effects of O3 on bactericidal activity against S. aureus paralleled the effects
on AM phagocytosis (early decrease, then no change). With P. mirabilis, there was more than
a 1,000-fold increase in PMNs in the lung that was not altered by O3, enabling bactericidal
activity to occur. In a similar manner, exposure of rats for 24 h/day for 7 days to 0.13, 0.26,
0.51, 0.77, and 1.02 ppm O3 decreased pulmonary bactericidal activity against Listeria at 0.77
and 1.02 ppm, with increased mortality at 1.02 ppm (Van Loveren et al., 1988). These effects
were associated with increased pathologic lesions, characterized by multifocal infiltrates of
histiocyctic and lymphoid cells, found in the lungs and liver of O3-exposed and Listeria
infected animals as compared to Listeria infection alone.
Fewer studies of viral infectivity have been conducted. Exposure for 15 days to
0.5 ppm O3 during the course of murine influenza virus infection had no effect on pulmonary
virus titers (Jakab and Hmieleski, 1988). A 5-day exposure for 3 h/day to 1.0 ppm O3 with
influenza virus infection on the second day of exposure had no effect on pulmonary virus
titers, but did show increased mortality, increased lung wet weight, and more severe
nonsuppurative pneumonitis and epithelial metaplasia and hyperplasia, with changes in lung
function consistent with that effect (Selgrade et al., 1988). Lung wet weight also was
increased when the mice were infected after the second day of exposure to 0.5 but not
0.25 ppm. When infection occurred on other days during the 5-day O3 exposure, no such
effects were found.
Typically, influenza virus infection causes pneumonitis characterized by severe
acute lung damage that eventually resolves to persistent alveolitis and changes in the
parenchyma (focal interstitial pneumonia and collagen deposition). Jakab and Bassett (1990)
investigated the effect of long-term O3 exposure (24 h/day for 120 days to 0.5 ppm) on mice
administered influenza virus immediately before O3 exposure started. The authors observed an
increase in pulmonary fibrosis with the virus infection, as compared to O3 exposure alone.
During the course of the viral infection, O3 exposure had no effect on pulmonary virus titers
and reduced the virus-induced acute lung injury. However, from Day 30 after infection,
increased numbers of AMs, lymphocytes, and PMNs were recovered from animals exposed to
virus plus O3, as compared to virus infection alone or O3 exposure alone. This increased
alveolitis correlated with increases in morphometrically determined lung damage and lung
hydroxyproline content, a biochemical marker indicative of pulmonary fibrosis. Ozone
exposure administered 10 days after viral infection enhanced lung hydroxyproline content at
Day 30, as compared to either virus infection or O3 exposure alone. Thus, O3 enhanced
postinfluenzal aveolitis and parenchymal changes. From these data, the authors speculated that
the mechanism for the postinfluenza lung damage may be related to O3 impairing the repair
process of the viral-induced acute lung injury.
In the studies reported to date, it is clear that the temporal relationships between
O3 exposure and influenzal infection are important. This is not surprising because there are
several waves of different antiviral defense mechanisms that might be affected differently by
O3. However, they have not been studied adequately for susceptibility to O3. Apparently, O3
does not alter defenses responsible for clearing virus from the lungs, as evidenced by the lack
of effect of O3 on viral titers (Selgrade et al., 1988; Jakab and Hmieleski, 1988). The
interaction between virus and O3 on histological changes in lung tissue can be damaging
(Selgrade et al., 1988; Jakab and Bassett, 1990) or beneficial (Jakab and Bassett, 1990),
possibly depending on the time of observation relative to the stage of the infectious process.
6-54
-------
The exact reasons are not known, but perhaps the induction of interferon production by the
virus plays a role. In noninfectious studies, Dziedzic and White (1987b) observed that
interferon induction mitigates O3-induced lung lesions, defined as areas with inflammatory cell
infiltration, cellular proliferation, consolidation, or edema, and that anti-interferon treatment
exacerbates those lesions.
6.2.3.7 Summary
Exposure to O3 can result in alterations of all the defense mechanisms of the
respiratory tract, including mucociliary and alveolobronchiolar clearance, functional and
biochemical activities of AMs, immunologic competence, and susceptibility to respiratory
infections. Structural (see Section 6.2.4), functional, and biochemical alterations in the
mucociliary escalator occur after O3 exposure. Mucociliary clearance is slowed in rabbits after
a single 2-h exposure to 0.6 ppm, but repeated (up to 14-day) exposures have no such impact
(Schlesinger and Driscoll, 1987). Secretions of mucous components are affected by repeated
exposure (Phipps et al., 1986; McBride et al., 1991). When lambs were exposed (1.0 ppm O3,
4 h/day, 5 days) shortly after birth, tracheal mucous components did not develop normally
(Mariassy et al., 1989, 1990). In contrast, alveolar clearance of rabbits after acute
O3 exposure (0.1 ppm, 2 h/day, 1 to 4 days) is accelerated (Driscoll et al., 1986). In the same
study, a 14-day exposure caused no effects, and a higher concentration (1.2 ppm) slowed
alveolar clearance. A similar pattern of slowed tracheobronchial clearance and accelerated
alveolar clearance occurs in rats (Phalen et al., 1980; Kenoyer et al., 1981). A subchronic (6-
week) exposure of rats to an urban pattern of O3 increased the retention of asbestos fibers
(Pinkertonetal., 1989).
Although AMs have numerous functions, one primary role is to clear the lung of
infections and noninfectious particles. Phagocytosis of bacteria, inert particles, and antibody-
coated red blood cells (RBCs) is inhibited by acute exposure to O3. The lowest effective
concentration tested was 0.1 ppm O3 (2 h) in rabbits (Driscoll et al., 1987). If exposures are
repeated for several days, phagocytosis returns to control levels (Driscoll et al., 1987; Gilmour
et al., 1991; Canning et al., 1991). The ability of AMs to produce superoxide anion radicals
(important to bactericidal activity) is inhibited by acute exposure to O3, especially in mice as
compared to rats (Ryer-Powder et al., 1988; Oosting et al., 1991a). The effect is clearly
evident after exposure for 3 h to 0.4 ppm, as observed by dysfunction in AM phagocytosis and
enhanced susceptibility to experimental respiratory infection (Gilmour et al., 1993a, 1993b).
Thus, the evidence indicates that the AM-dependent alveolobronchiolar transport mechanisms
are impaired, as are their phagocytic and microbicidal activities, leading to decreased
resistance to respiratory infections.
The experimental database also shows that the effects of O3 on the immune system
are complex. These effects are not yet fully evaluated, and the reported effects on immune
parameters are dependent on the exposure regimen and the observation period. It appears that
the T-cell-dependent functions of the immune system are more affected than B-cell-dependent
functions (U.S. Environmental Protection Agency, 1986; Fujimaki, 1989). Generally, there is
an early immunosuppressive effect that, with continued O3 exposure, results in either return to
normal responses or immunoenhancement. For example, in mice exposed for 28 days (20
h/day) to 0.3 to 0.7 ppm O3, there was an early (Days 1 and 2) depletion of cells in the MLN,
followed by MLN T-cell hyperplasia and increased blastogenic response to a T-cell mitogen
(Dziedzic and White, 1986a). Several investigations have found an initial (Days 1 to 4)
6-55
-------
decrease in blastogenic response to T-cell mitogens in the MLN and spleen of mice exposed for
a few weeks to 0.7 or 0.8 ppm O3 that returned to control levels by the end of the exposure
(Dziedzic and White, 1986a; Gilmour and Jakab, 1991). There also are changes in cell
populations in lymphatic tissues. For example, T:B-cell ratios in the MLN increase when rats
are exposed for 7 days to DO.26 ppm O3 (Van Loveren et al., 1988). Natural killer cells in the
lung are affected under some circumstances. Van Loveren et al. (1990) showed that a 1-week
exposure to 0.2 or 0.4 ppm O3 increased NK cell activity, but a higher concentration (0.82
ppm) decreased it. Ozone also alters response to antigenic stimulation. For example, antibody
responses to a T-cell-dependent antigen were suppressed after a 56-day exposure of mice to
0.8 ppm O3, and a 14-day exposure to 0.5 ppm O3 decreased the antiviral antibody response
following influenza virus infection (Jakab and Hmieleski, 1988). The temporal relationship
between O3 exposure and antigenic stimulation is important. When O3 exposure preceded
Listeria infection, there were no effects on delayed-type hypersensitivity or splenic
lymphoproliferative responses; when O3 exposure occurred during or after Listeria infection
was initiated, these immune responses were suppressed (Van Loveren et al., 1988). With
experimental viral infections, O3 exposure decreases the T-lymphocyte responses and the
antiviral antibody response (Jakab and Hmieleski, 1988); the latter impairment may pave the
way for lowered resistance to reinfection.
The significance of O3-induced suppression of immune parameters in relation to risk
of infectious disease is an example of a generic problem that remains to be clarified. For
example, correlative studies between immune parameters and acquired immune deficiency
syndrome have provided clear insights on the magnitude of the immune dysfunction and the
progression of the disease. However, there is a paucity of such correlative studies on
alterations of immune parameters and function resistance mechanisms or disease endpoints.
Because the major functions of the immune apparatus are to protect against infectious agents
and conduct tumor surveillance, suppression of immune parameters is considered a signal of
increased susceptibility to infections and acquisition of tumors. One of the few studies that has
addressed this issue (Selgrade, 1995) illustrates how the temporal relationships between
exposure to a compound and exposure to infectious agents or tumor cells will have an impact
on the risk associated with immune suppression in experimental animal models. The types of
immune responses affected by a chemical and their importance to defense against any particular
infectious agent, the recovery time of the immune response, the length of exposure, and the
time required for mobilization of alternative defenses are among the factors that can impact the
risk of enhanced infectious disease that might be associated with an immunosuppressive event.
In addition to a suppressive effect on pulmonary immunity, O3 exposure also can
affect systemic immunity. Although these depressive effects on the systemic compartment
occur at approximately twice the O3 exposure concentrations observed for pulmonary
immunosuppression (with the exception of a study by Aranyi et al. [1983] at 0.1 ppm O3), the
observations are important because they show that the effects of O3 exposure on host resistance
are not limited to the lung alone, but may increase susceptibility to systemic infections as well
as pulmonary infections. However, Selgrade et al. (1990) found no effects on selected
systemic immune functions in rats exposed for up to 78 weeks to an urban pattern of O3.
Numerous studies have confirmed that acute or short-term exposure to O3 decreases
lung bactericidal activity and increases susceptibility to respiratory bacterial infections. The
lowest exposure showing such effects was 0.08 ppm (3 h) in the mouse streptococcal model
(Coffin et al., 1967; Coffin and Gardner; 1972; Miller et al., 1978). Further research has
6-56
-------
indicated that changes in antibacterial defenses are dependent not only on exposure regimens,
but also on species and strain of animal, species of bacteria, and age of animal (e.g., young
mice are more susceptible) (Gilmour et al., 1991, 1993a,b; Gilmour and Selgrade, 1993).
Furthermore, increasing the duration of an exposure to 0.1 ppm O3 from a few hours to
3 weeks either causes no effect or does not enhance the streptococcal-induced mortality
observed after acute exposure (Graham et al., 1987; Aranyi et al., 1983). In general, the
effect of O3 exposure on antibacterial host defenses appears to be concentration- and
time-dependent. Acute exposures result in an impairment of host defenses, whereas the
defense parameters become reestablished with more prolonged exposures.
Effects of O3 on the course of viral infections are more complex and highly
dependent on the temporal relationship between O3 exposure and viral infection. For example,
Selgrade et al. (1988) found increases in mortality and lung wet weight in mice infected with
influenza only after the second day of O3 exposure (1 ppm, 3h/day). Jakab and Bassett (1990)
found no detrimental effect of a 120-day exposure to 0.5 ppm O3 on acute lung injury from
influenza virus administered immediately before O3 exposure started. However, O3 enhanced
postinfluenzal alveolitis and lung parenchymal changes. Because O3 did not affect lung
influenza viral titers in any of these studies, it is unlikely that O3 has an impact on antiviral
clearance mechanisms.
Ozone-induced susceptibility to experimental respiratory infections has been
correlated with the immunotoxic effects of O3 by the observation that O3 exposure increases the
severity of Listeria infection while concurrently suppressing the antigen-specific immune
responses (Van Loveren et al., 1988).
6.2.4 Morphological Effects
6.2.4.1 Introduction
All mammalian species studied react to inhaled concentrations of < 1.0 ppm O3 in a
generally similar manner, with species variation in morphological responses depending on the
distribution of sensitive cells and the type of junction between gas conducting and exchange
areas of the lung (U.S. Environmental Protection Agency, 1986). The cells most damaged by
O3 are the ciliated epithelial cells in airways and Type 1 cells in gas-exchange areas. Both of
these cell types have very large surface areas (relative to volume) exposed to inhaled gases.
The many factors that influence the distribution of inhaled O3 within the respiratory system
(see Chapter 8) result in some of the largest effective doses to the epithelial cells lining the
nose and to epithelial cells located at the junction of the conducting and exchange areas, the
CAR, in lungs. The 1986 criteria document did not contain studies of morphological effects of
O3 on the nose. There are species differences both in the basic structure of the CAR and in the
epithelial cells that line the CAR (Tyler, 1983; Plopper, 1983). The CAR of humans, other
primates, dogs, cats, and a few other domesticated species consists of the last conducting
airway, the TB, several generations of RBs, alveoli that open directly into RB lumens, and
ADs that branch from RBs (Figure 6-2). In lungs from many other mammals, including those
most commonly used for inhalation toxicology (i.e., rats, mice, guinea pigs, and rabbits), RBs
are poorly developed or absent and the CAR consists of TBs that open directly into ADs.
Epithelial degenerative changes in TBs and alveoli occur early, 2 to 4 h, in an
O3 exposure (Stephens et al., 1974a; Castleman et al., 1980). Depending on the dose to
6-57
-------
Human and Nonhuman
Primates
Rats and Most
Other Rodents
Nonrespiratory
Bronchioles
Acini
Respiratory
Bronchioles
Alveolar
Ducts and
Alveoli
Alveolar
. ^ Ducts and
^ *v Alveoli
Figure 6-2. Schematic representation of intrapulmonary conducting airways and acini
from animals with respiratory bronchioles (RBs) (human and nonhuman
primates) and without RBs (rats and most rodents). For simplicity, several
generations of bronchi and nonrespiratory bronchioles are not depicted.
Source: Redrawn from Weibel (1963) using information from Tyler and Julian (1991).
individual ciliated cells, they may lose cilia; undergo degenerative changes; or become necrotic
and be sloughed into the lumen, leaving bare basement membrane until other cells replace them
(Stephens et al., 1974a; Castleman et al., 1980). In the TB of the CAR, sloughed ciliated cells
are replaced by nonciliated bronchiolar cells, which may become hyperplastic following longer
exposures. Although these changes in TB epithelial cells can be studied readily by light
microscopy (LM) or transmission electron microscopy (TEM), the surface views provided by
scanning electron microscopy (SEM) provide a more comprehensive understanding of the
three-dimensional aspects of CAR changes (Schwartz et al., 1976; Castleman et al., 1980).
Changes in mucus-secreting cells of conducting airways were considered minor in
the 1986 criteria document (U.S. Environmental Protection Agency, 1986). Schwartz et al.
(1976) did not find changes in mucous cells suggesting damage to cell organelles in rats
6-58
-------
6 weeks to 0.25 ppm O3. They reported CAR alveoli had more Type 1 and 2 epithelial cells
and more AMs. The Type 1 cells were smaller in volume, covered less surface, and were
thicker. They were aware of the results of Stephens et al. (1978) and speculated that the
changes they described may have occurred primarily during the last 3 weeks of exposure.
Boatman et al. (1983) did not find effects of O3 on lung growth following pneumonectomy.
Several studies included both an exposure and a postexposure period during which
the animals breathed air without O3. Plopper et al. (1978) reported that CAR epithelial cells
returned to normal appearance 6 days after a 72-h exposure to O3. Incomplete resolution was
reported 7 days after a 50-h O3 exposure of monkeys (Castleman et al., 1980), 10 days after a
20-day O3 exposure of mice (Ibrahim et al., 1980), and 62 days after a 180-day O3 exposure of
rats (Moore and Schwartz, 1981).
The previous criteria document (U.S. Environmental Protection Agency, 1986)
comprehensively evaluated several citations reporting emphysema following O3 exposure using
current definitions of human emphysema (Snider et al., 1985). The morphological changes
described in those earlier publications did not meet the current criteria for emphysema of the
type seen in human lungs.
6.2.4.2 Sites Affected
The sites in the respiratory system that are affected will be discussed in anatomical
sequence, beginning with the upper respiratory tract and proceeding downward. The upper or
extrathoracic conducting airways (also referred to as the nasopharyngeal region) include the
nasal cavity, pharynx, and larynx. Lower conducting airways begin with the trachea and
include the bronchi and nonterminal bronchioles (also referred to as the tracheobronchial
region). A summary of available information on these sites in the respiratory system is in
Table 6-10, Effects of Ozone on Conducting Airways. Summaries of the information available
concerning effects on the CAR, the gas exchange region (also referred to as the pulmonary
region), and other pulmonary structures are divided into effects of short-term (<2-week)
exposures in Table 6-11 and effects of long-term exposures in Table 6-12.
Nasal Cavity and Nasopharynx
The nasal cavity "conditions" inhaled air and in that process "scrubs" some reactive
pollutants from the inhaled air, thereby reducing the concentration to which other portions of
the respiratory system are exposed (Yokoyama and Frank, 1972; Miller et al., 1979).
Although this scrubbing process is protective of other portions of the respiratory system, it
results in a large dose of pollutant to the cells and tissues that line the nasal cavity.
There is a large range of variation in the structure of the nasal cavity among the
animals used for inhalation toxicology and between those animals and humans (Schreider and
Raabe, 1981). Schreider and Raabe (1981) found a striking similarity between the
nasopharyngeal cavities of monkey and humans. They proposed that, with appropriate scaling,
the monkey could serve as a model for aerosol and gas deposition in the nasopharyngeal region
of humans. Thus, the studies on monkeys by Harkema et al. (1987) provide useful information
for extrapolation to humans, as well as information concerning cellular responses in monkeys.
6-60
-------
Table 6-10. Effects of Ozone on Conducting Airway^
Ozone
Concentration
ppm
0.05
0.1
0.5
1.0
2.0
4.0
6.0
8.0
0.12
0.8
1.5
0.12
0.8
0.12
0.25
0.5
0.12
0.27
0.8
0.12
0.5
1.0
Qg/m
98
196
980
1,960
3,920
7,840
11,760
15,680
235
1,568
2,940
235
1,568
235
490
980
235
529
1,568
235
980
1,960
Exposure
Duration
30 min
(In vitro
monolayer cultures)
6h
6 h/day for
7 days
20 h/day,
7 days/week for
2 years
(4, 12, 26, 52, 78,
and 104 weeks)
6 h/day for
3 or 7 days
6 h/day,
5 days/week for
20 mo
Species, Sex
(Strain)
Age1'
Rabbit
(NZW)
Rat, F
(F344/N)
280-400 g
Rat, F
(F344/N)
12-14 weeks old
Rat, M
(F344,
CrlBR)
42 days old
Rat, F
(F344/N)
8-12 weeks old
Rat, M
(F344)
6-8 weeks old
Observed Effect(s) 3
Phase contrast LM and trypan blue exclusion of tracheal epithelium. No consistent changes at 0.05 or 0.1 ppm,
but 0.5 to 2.0 ppm resulted in cytoplasmic vacuolization without decreased viability. More evidence of cell
damage, including gaps in previously confluent cultures and loss of cell viability at 4.0 and 8.0 ppm. Lipid
peroxidation also was studied in these cultures.
LM pathology and LM morphometry of PMNs in lung and nasal epithelia. Also nasal and bronchoalveolar lavage.
See Section 6.2.2. Nasal epithelium: No necrosis, loss or attenuation of cilia, or hyperplasia at any exposure.
Increases in pavemented PMNs at most concentrations and times. Lung: No lesions due to 0.12 ppm. CAR
lesions not obvious at 0.8 or 1.5 ppm immediately or 3 h PE, but a progressive increase in CAR lesions at all other
times. CAR TBs and proximal alveolar septa thickened and increased inflammatory cells. Increased CAR tissue
PMNs in 0.8- and 1.5-ppm groups at 18 and 66 h PE and 1.5-ppm group at exposure end.
LM pathology and morphometry of LM histochemistry of nasal epithelia. No LM pathology at 0. 12 ppm, but 0.8
ppm caused hyperplasia of transitional nonciliated epithelium. In 0.8-ppm group, shortened cilia in respiratory
epithelium of nasopharynx, but not of nasal septum. Increased intraepithelial mucosubstances in all areas at 7 days
PE to 0.8 ppm.
Smaller body weights after 7 weeks exposure to 0.5 ppm. LM histopathology of nasal epithelia. Nose: At 00.25
ppm, mucous cell and respiratory epithelial hyperplasia. No lesions in mainstem or large bronchi. CAR:
Described in Table 6-12.
LM pathology and LM BrdU for DNA synthesis by nasal epithelia. No LM pathology in squamous epithelium or
in ciliated respiratory epithelium. Hyperplasia of nonciliated cuboidal/transitional epithelium at 0.8 ppm. BrdU
uptake (DNA synthesis) increased at end of 3- and 7-day exposure to 0.8 ppm. BrdU uptake decreased in
squamous epithelium only after 7 days exposure to 0.8 ppm and 7 days PE.
Nasal changes limited to nasal transitional nonciliated epithelium at 0.5 and 1.0 but not 0.12 ppm. LM
histochemistry of intraepithelial mucosubstances in the nose and bronchi. Epithelial cell hyperplasia; mild to
moderate inflammatory cell influx into the mucosa; and increased mucosubstances. Mucous flow rates decreased.
Reference
Alpert et al. (1990)
Hotchkiss et al. (1989a)
Harkema et al. (1989)
Smiler et al. (1988)
Johnson et al. (1990)
Harkema et al. (1994)
-------
Table 6-10 (cont'd). Effects of Ozone on Conducting Airway§
Ozone
Concentration
ppm
0.12
0.5
1.0
0.12
0.24
0.48
0.15
0.3
0.15
0.25
0.15
0.25
0.2
0.4
0.8
Qg/m
235
980
1,960
235
470
940
294
588
294
490
294
490
392
784
1,568
Exposure
Duration
6 h/day,
5 days/week for
20 mo
3, 6, 12, or 24 h
8 h/day for
6 days to
0.15 ppm or for
90 days to
0.15 or 0.3 ppm
8 h/day for
6 days to
0.15 ppm
or for 90 days to
0.15 or 0.3 ppm
8 h/day for
7 days
22 h/day for
3 days
Species, Sex
(Strain)
Age"
Rat, M
(F344)
6-8 weeks old
Rat, F
(F344/N)
11-13 weeks old
Macaco radiata, M, F
2-6 years old,
2.3-9.7 kg
Macaco radiata, M, F
2-6 years old,
2.3-9.7 kg
Macaco radiata
NS
Rat, M
(Wistar
RIVM [TOX])
150-190 g
Observed Effect(s) 3
LM morphometry and histochemistry of "short-" and "long-path" conducting airways and CAR.
Trachea: No changes in epithelial thickness, cell populations or stored glycoconjugate, but a dose-
dependent loss of stored glycoconjugate was found.
Bronchi: No changes in epithelial thickness or cell populations. Rats exposed to 1.0 ppm had increased
stored gylcoconjugates in cranial (short-path) and caudal (long-path) bronchi, but not in central (short-
path) bronchi.
Bronchioles: TB of rats exposed to 1.0 ppm had thicker epithelium with increased V, of nonciliated
bronchiolar cells. Vv also increased in caudal (long-path) TB of 0.5-ppm exposed group. Mass
(QrrrVQnf) (Vs) of nonciliated cells increased in caudal (long-path) TBs of all exposed rats, but not in
short-path TBs.
CAR: Increased Vs of bronchiolar epithelium in former ADs in cranial and caudal CARs of rats exposed
to 1 .0 ppm and in cranial CAR of rats exposed to 0.5 ppm.
LM pathology and semiautomatic image analysis system for DNA synthesis (Brdll uptake) of epithelium
in nasal maxilloturbinates. C x T design. No effects at 0.12 or Dl.44 ppm-h. For a given C x T,
increased DNA synthesis equal at different Cs and Ts. Nonlinear increase in DNA synthesis as C x T
increased.
LM, SEM, and TEM morphometry of nasal epithelia. Respiratory epithelium: Ciliated cell necrosis,
shortened cilia, and increased small mucous granule cells at all exposures, even at 0.15 ppm for 6 days.
Transitional epithelium: Decreased nonciliated cells without granules, increased nonciliated cells with
granules, and increased small mucous granule cells. Increased intraepithelial leukocytes only at 6 days
in both types of epithelium.
LM and TEM histochemistry for intraepithelial mucosubstances in nasal mucosa. More mucous cells
that had dilated granular endoplasmic reticulum. Also changes in mucosubstances.
The O3 concentration is not clear— the abstract states 0.64 ppm, the text mentions only 0.25 ppm. LM
morphometry of vocal fold mucosa. Disruption and hyperplasia of stratified squamous epithelium.
Epithelium thickened at 12 h and 7 days PE. Connective tissue of lamina propria thickened at 7 days
PE. Basement membrane is undulating rather than smooth. Even though thickened, epithelium
appeared normal at 7 days PE.
LM pathology and tritiated thymidine uptake by nasal epithelia. Observations on mixtures are in
Section 6.4. Respiratory epithelium: No changes at 0.2 ppm O3. Loss of cilia and disarrangement at
0.4 and 0.8 ppm. Some epithelia were hyperplastic or metaplastic or both. Thymidine uptake increased
in rostral (anterior) portions.
Reference
Plopper et al. (1994a)
Henderson et al. (1993)
Harkema et al. (1987)
Dimitriadis (1992)
Leonard et al. (1991)
Reuzel et al. (1990)
-------
Table 6-10 (cont'd). Effects of Ozone on Conducting Airway§
u>
Ozone
Concentration
ppm
0.8
0.8
0.96
0.96
1.0
1.0
1.0
Dg/m
1,568
1,568
1,882
1,882
1,960
1,960
1,960
Exposure
Duration
6 h/day for
3 or 7 days
6h
8 h/night,
7 nights/week
for 3 or
60 nights
8h
4 h/day for
5 days
(examined at
2 weeks)
4 h/day for
5 days
(examined at
2 weeks)
96 h
(in vitro tracheal
explants)
Species, Sex
(Strain)
Age1'
Rat, F
(F344/N)
12-14 weeks old
Rat, F
(F344/N)
12-16 weeks old
Rat, M
(S-D)
234-263 g
Macaco mulata, M
2.0-8.5 years old
2. 1-6.3 kg
Sheep
New-born
Sheep
New-born
Rat, M
(S-D)
250-270 g
Observed Effect(s) 3
LM morphometry and histochemistry of nasal nonciliated cuboidal epithelium. Other types of nasal
epithelia not examined. After 3 days O3 and 18 h PE, no changes in cell density or in intraepithelial
mucus. After 7 days exposure or after 3 days exposure and 4 days PE, hyperplasia and increased
intraepithelial mucosubstances with no change in the ratio of acidic to neutral mucosubstances.
LM pathology and DNA synthesis by Brdll uptake by nasal nonciliated transitional epithelium. Ozone
did not result in necrosis, exfoliation, or inflammation, but did increase DNA synthesis.
LM morphometry, histochemistry, autoradiography, and SEM and TEM morphometry of tracheal
epithelium. Neither 3 nor 60 days exposure altered the cell density of ciliated, serous, basal, brush,
migratory, or unidentified cells in tracheal epithelium. 3 days: Damage to cilia and ciliated cells,
including necrosis. Thymidine labeling index increased. Serous cell histochemistry unchanged.
60 days: Less evidence of injury than at 3 days, but more damaged ciliated cells than in controls.
Complete recovery of the epithelial changes by 42 days PE.
LM and TEM morphometry of trachea, bronchi, and RBs. Increased necrotic cells in trachea and RBs
at 1 h PE and in bronchi at 12 and 24 h PE. Decreased ciliated and basal cells in bronchi at 1, 12, and
24 h PE. Basal cells in bronchi also decreased at 72 and 168 h PE. Nonciliated bronchiolar cells in
RBs increased only at 24 h PE. In bronchi, smooth muscle increased and amorphous matrix decreased
at 24, 72, and 168 h PE. In RBs, smooth muscle increased at 24 h, fibroblasts increased at 24 and 72
h, and amorphous matrix increased at 12 h PE.
LM morphometry of tracheal epithelium. Also see Section 6.2.3. Percentage of ciliated and mucous
cells remained at newborn levels, rather than ciliated cell percent increasing and mucous cell percent
decreasing as in control lambs.
LM morphometry of mucosubstances in tracheal epithelium. No evidence of damage or inflammatory
changes. Decreased epithelial cell density, decreased ciliated and basal cells. Lectin-detectable
intraepithelial mucosubstances did not undergo the maturation changes seen in control lambs.
Tracheal organ cultures exposed in vitro. Filtered air + O3 resulted in extensive damage to cilia, and
intermediate cells were seen. Cultures exposed to 95% O, + O3 had stratified thickened epithelium
with metaplastic cells in a middle zone and no ciliated cells at the surface.
Reference
Hotchkiss et al. (1991)
Hotchkiss and Harkema (1992)
Nikula et al. (1988a)
Hyde et al. (1992)
Mariassy et al. (1990)
Mariassy et al. (1989)
Nikula and Wilson (1990)
"See Appendix A for abbreviations and acronyms.
''Age or body weight at start of exposure.
-------
Table 6-11. Effects of Ozone on Lung Structure: Short-Term Exposures (<2 Weeks)
Ozone
Concentration
ppm
0.12
0.8
1.5
0.12
0.8
1.5
0.12
0.5
0.75
0.15
0.5
0.2
0.4
0.2
0.38
0.25
0.8
0.35
0.6
0.35
0.8
Qg/m
235
1,568
2,940
235
1,568
2,940
250
1,000
1,500
393
784
393
745
490
1,568
686
1,176
686
1,568
Exposure
Duration
Continuous for
6h
6h
Continuous for
1-7 days
Continuous for
3 or 7 days
4h
3.75 h,
Rest or
exercise
29 min/day for
2 days,
Strenuous
exercise
4 h at rest,
3 h with exercise
72 h
Species, Sex
(Strain)
Age1'
Rat, M
(F344/N)
12-18 weeks old
Rat, M
(F344/N)
12-18 weeks old
Rat, M
(Wistar
RIV:TOX)
8 weeks old
Rat, M
(Wistar
RIV:TOX)
8 weeks old
Rat, M
(S-D)
7 weeks old
Rat
(S-D)
Horse
(Thoroughbred)
5-6 years old
Gelding
Rat, M
(S-D)
7 weeks old
Rat, F
(S-D)
60 or
444 days old
Observed Effect(s) 3
LM histopathology of lungs and LM morphometry of lavaged AMs: No LM histologic effect detected at 0.12 ppm.
No LM histologic effect at 0.8 and 1.5 ppm immediately or 3 h PE. At later PE times, there was mild, patchy
CAR bronchiolitis and alveolitis. Increase in AMs and PMNs from 18-66 h PE. Progressive thickening of TB
walls and CAR AD septa at 18, 42, and 66 h PE.
Observations of nose and CAR. Same rats as in Hotchkiss et al. (1989a). LM histopathology of CAR is the same;
new morphometry of PMNs in CAR and nasal mucosa. Emphasis on PMNs in nasal mucosa and nasal lavage
compared with PMNs in CAR tissues and BAL at exposure end and at 3-66 h PE.
LM: Increased AMs in CAR and parenchyma. CAR increase persisted 5 days PE. TEM and SEM: BAL
AMs had microvilli and blebs in addition to ruffles characteristic of AMs from controls. Also see Section 6.2.3.
Elastase-induced emphysema and saline control rats. LM histopathology and morphometry for alveolar size. Also
see Section 6.2.5. The incidence and severity of CAR LM lesions was the same in elastase- or saline-treated rats
exposed to O3. No change in alveolar size due to O3.
LM histopathology. Necropsy 24 h PE. Lung lesions from 0.2 ppm not reported, but 0.4 ppm resulted in
increased AMs and increased cellularity of alveolar septa with focal thickening. No increase in DNA synthesis by
nasal epithelium. See mixture effects in Section 6.4.
LM histopathology and morphometry. No effect from 0.2 ppm at rest. Increased free cells in airspaces at
0.38 ppm at rest. Exposure during exercise resulted in larger areas with free cells and in areas of septa thickened
by infiltrating cells.
LM and TEM histopathology (see also Section 6.2.5). 0.25 ppm: Lesions limited to vacuoles seen only by TEM
in TB ciliated cells. 0.8 ppm: Gross hemorrhage and edema in two of three horses. CAR lesions, visible only by
TEM, included edema, necrosis and sloughing of Type 1 cells, slight increase in AMs, shortened cilia, and
vacuoles in ciliated and nonciliated bronchiolar cells.
Examination 48 h PE. Lung: LM morphometry of lesions as a percent of parenchyma section area. No statistical
evaluation of groups exposed to O3 at rest and exercise. The lesion percent of parenchyma appears concentration-
dependent and increased by exercise similar to Mautz et al. (1985b). Nasal epithelium: Evaluated percent
thymidine labeled cells in respiratory epithelium. No change due to O3 at rest. Effects of mixtures in Section 6.4.
LM morphometry and SEM. Vv of CAR lesions. Exposed adults lost body weight. Adults exposed to 0.8 ppm
had smaller fixed lung volumes. Lesion V, larger in young than adult rats at both concentrations. Free cell (AM)
Vv increased in young rats at 0.35 ppm compared to adults. The CAR lesions were similar by SEM, but younger
rats had more AMs in the CAR. Younger rats had larger CAR lesions and more AMs, but older rats had greater
changes in body weight and fixed lung volume.
Reference
Hotchkiss et al. (1989b)
Hotchkiss et al. (1989a)
Dormans et al. (1990)
Dormans et al. (1989)
Mautz etal. (1991)
Mautz etal. (1985b)
Tyler et al. (1991c)
Mautz etal. (1988)
Stiles and Tyler (1988)
-------
Table 6-11 (cont'd). Effects of Ozone on Lung Structure: Short-Term Exposures (<2 Week^)
Ozone
Concentration
ppm Qg/m
0.35 686
0.5 980
1.0 1,960
0.4 784
0.5 980
0.64 1,254
O>
On 0.7 1,372
0.9 1,764
0.75 1,470
0.8 1,568
0.4 784
0.8 1,568
Exposure
Duration
2.25 h/day for
5 days
Continuous
up to 14 days
20 h/day for
1-14 days
Continuous,
7 days
20 h/day for
4 days
Continuous for
3 days
Continuous for
7 days
Continuous for
3h
Species, Sex
(Strain)
Age1'
Rat, M
(F344)
110-120 days old
Rat, M
(Wistar Jcl)
5 weeks old
Rat, M
(F344)
13 weeks old
Rat, M
(S-D)
250-300 g
Mouse, F
(CD-I
CrhCDl/
(CR)BR)
20-22 g
Rat, M
(Wistar)
200-250 g
Rat, M
(Wistar)
8 weeks old
Rat, M
(S-D)
250-300 g
Observed Effect(s) 3
LM histopathology and morphometry for parenchymal density and alveolar size. No significant changes in these
morphometric parameters. Histopathology: Reported only for 0.5 ppm O3 + CO, challenge group, which had
maximal CAR tissue damage on Days 4 and 5. Increased AMs on Days 2 and 3 and foci of necrotic TB
epithelium. By Day 5, hyperplasia of TB epithelium and increased AMs and other inflammatory cells, which
completely filled some CAR alveoli. Morphologic damage continued, whereas pulmonary functional changes
attenuated. Also see Section 6.2.5.
SEM morphometry, immunocytochemistry. See Section 6.2.1 for biochemistry. Number of Clara cellsfim2
increased at 14 days, but not earlier. The length of the Clara cell apical projection was increased after 6 h,
decreased at 1 day, and not different at other periods. Cytochrome P-450 was localized to agranular endoplasmic
reticulum of Clara cells.
LM morphometry of thymidine-labeled cells in bronchus-associated lymph node and MLN. Other lung changes
not described. Also see Section 6.2.3.
LM morphometry of CAR (proximal alveolar) lesions. Increased centriacinar lesions. Rats treated with
dimethylthioruea (a H2O, scavenger) had smaller lesion volumes.
LM morphometry for areal density (e.g., volume density) of lesions. No comparison of O, and air exposures.
Also see Section 6.2.3.
LM histopathology. Ozone-exposed rats gained less body weight. Increased cells in TB and CAR AD septa.
Number of cells diminished by Day 4 PE, but foci of AMs remained.
LM: Clara cell numbers/mm of TB basement membrane unchanged. Also see Section 6.2.3. Cell isolation:
Although the number of isolated Clara cells/106 cells isolated/lung was increased, the percent Clara cells in the
isolate was not changed. The percent Type 2 cells in the isolate was increased. No morphologic observations at
0.4 ppm.
By LM, PMNs in alveolar septa increased three times at 4 h PE. Number of septal PMNs peaked at 8 h PE and
then rapidly declined. Free cells, septal thickening, and cellularity increased with increasing time PE. Also see
Section 6.2.2.
Reference
Tepperetal. (1989)
Suzuki etal. (1992)
Dziedzic et al. (1990)
Warren et al. (1988)
Dziedzic and White
(1987b)
Bassettetal. (1988a)
Van Bree et al. (1989)
Bhalla and Young
(1992)
0.82 1,600 Continuous for Rat, M LM morphometry of histochemically identified Type 2 cells. Increased number of Type 2 cells than in controls. Dormans (1989)
7 days (Wistar RIV:TOX)
8 weeks old
-------
Table 6-11 (cont'd). Effects of Ozone on Lung Structure: Short-Term
Exposures (<2 Weeks/
Ozone
Concentration
ppm Dg/m
0.97 1,901
1.0 1,960
2.0 3,920
1.0 1,960
2.0 3,920
1.0 1,960
1.0 1,960
Exposure
Duration
Continuous for
7 days
4 h/day for
5 days
Continuous for
3h
(Isolated
perfused
lungs)
Continuous for
8h
4, 6, 8, and 24 h
Species, Sex
(Strain)
Age"
Rat, M
225-275 g
Dog, M, F
(Beagle)
6 weeks old
Rat, M
(S-D)
300-380 g
Rat, M
(S-D)
10 weeks old
Rat, M
(S-D)
63 days old
Observed Effect(s) 3
Smaller body weights. No other statistical comparison of controls and O3 alone (see Sections 6.2.6 and 6.4).
LM morphometry for alveolar size. Mean linear intercepts larger (indicating larger alveoli) in 1 .0- but not 2.0-ppm
group.
LM and TEM morphology. Necrosis and sloughing of airway epithelial cells of bronchi and larger bronchioles. TB
had less severe lesions, including fewer necrotic cells and less damage to cilia. Fragmentation of some Type 1 cells
with some areas of bare basal lamina.
LM and TEM morphometry. Also see Section 6.2.2. Rats received either normal rat serum or rabbit anti-rat PMN
serum before the exposure. At exposure end, both exposed groups had a smaller volume of ciliated cells per unit area
of epithelial basal lamina (Vs) compared with filtered air controls with similar serum. Ciliated cell Vs was also smaller
at 4 and 16 h PE.
TEM morphometry. Also see Section 6.2.2. Volume of necrotic cells per area basal lamina (Vs) in the TB larger than
controls at the end of 4- and 24-h exposure, but not at other exposure or PE times. With increasing exposure time,
there was a shift from necrotic cells on the basal lamina to necrotic cells free in the TB lumen. The Vs of necrotic
alveolar cells was increased after 4, 6, and 24 h of exposure. Viable undifferentiated cell Vs in TBs was increased
after 6-h exposure followed by 18 h PE, 8-h exposure followed by 16 h PE, and after a 24-h exposure. In alveoli,
viable Type 1 cell Vs was increased after a 24-h exposure. Total connective tissue Vs changes only increased in TBs
after 8-h exposure followed by 4 h PE and in alveoli at the end of 8-h exposure. The V, of migratory cells in TB
interstitium was only increased 4 h after a 6-h exposure. In alveoli, the Vs of capillaries was increased after 8-h
exposure.
Reference
Last et al. (1986)
Phalen et al. (1986)
Pino et al. (1992a)
Pino et al. (1992b)
Pino et al. (1992c)
'See Appendix A for abbreviations and acronyms.
''Age or body weight at start of exposure.
-------
Table 6-12. Effects of Ozone on Lung Structure: Long-Term Exposures (>2 Weeks)
Ozone
Concentration
ppm
Base
0.06;
spike to
0.25
A: 0.25
B: Base
0.06;
spike to
0.25
0.1
0.12
0.25
0.12
0.5
1.0
0.12
0.5
1.0
0.12
0.25
0.5
Dg/m
Base
118;
spike to
490
490;
118 to
490
196
235
490
235
980
1,960
235
980
1,960
235
490
980
Exposure
Duration
Base
13 h/day,
7 days/week;
Ramped spike
9 h/day,
5 days/week
(1, 3, 13, and
78 weeks)
6 weeks
3 weeks
13 weeks
12 h/day
22 h/day
2 h/day,
5 days/week
for 1 year
12 h/day for
6 weeks
6 h/day,
5 days/week for
20 mo
6 h/day,
5 days/week for
20 mo
20 h/week,
7 days/week for
2 years
(Examined at 4,
12, 26, 52, 78,
and 104 weeks)
Species, Sex
(Strain)
Age1'
Rat, M
(F344)
60 days old
Rat
(F344)
7 weeks old
Rabbit, M
(NZW)
3-3.5 kg
Rat, M
(F344)
1 day or 6 weeks
old
Rat, M
(F344)
6-8 weeks old
Rat, M
(F344)
6-8 weeks old
Rat, M
(F344
CrlBR)
42 days old
Observed Effect(s) 3 Reference
TEM morphometry, data expressed as volume per unit epithelial basement membrane (Vs). Acute response in the CAR Chang et al. (1992)
(proximal alveolar) region included increased AMs, interstitial edema, and interstitial cell hypertrophy. These changes
subsided by 3-weeks exposure. Changes at 13 and 78 weeks include increased Vs, total alveolar tissue, and epithelium.
Type 1 cell increases include Vs at 13 weeks; number at 13 weeks, 13 + 6 weeks PE, and 78 weeks; and decreased
surface area at 78 weeks. Type 2 cell Vs increased at 78 weeks and after 78 + 17 weeks PE, and number increased at
78 weeks. Total interstitium was increased at 78 weeks, and noncellular interstitium (collagen and basement membrane)
was increased at 13 and 78 weeks. Thickened basement membrane had crystalline deposits. No bronchiolization or
centriacinar airway remodeling. Changes in terminal bronchiolar cells include a decrease in surface area of ciliated and
nonciliated at 78 weeks. No changes in bronchiolar cell numbers. All changes diminished, except the increased volume
of Type 2 cells and the thickened basement membrane, 17 weeks PE.
Compared effects of A and B exposure regimens. Cumulative doses for A (Barry et al., 1985) were 60.5 and Chang et al. (1991)
126.0 ppm-h and, for B, were 45.3 ppm-h at 3 weeks and 196.0 ppm-h at 13 weeks. The B regimen alone (Chang et al.,
1992) is described earlier in this table. TEM morphometry of CAR. The pattern of exposure did not affect the degree of
injury.
LM morphology and morphometry of intrapulmonary conducting airways. (Also see Section 6.4). No difference in Schlesinger et al.
number of airways/area or in distribution of airway size. ESCs in the smallest conducting airways ( < 0.30 mm) (1992a)
increased at 4, 6, and 12 mo of exposure to O3 and decreased at 6 mo PE. ESCs in the next larger airways (0.31-0.49
mm) only increased after 4 mo of exposure and decreased at 6 mo PE. No effect on airways >0.50 mm.
TEM morphometry of proximal alveolar region (CAR). Type 1 epithelial cells increased in number and thickness, but Barry et al. (1985)
decreased in luminal and basal lamina surface area. Some bare basement membrane where Type 1 cells sloughed, but
not significantly increased. At 0.25 ppm, Type 2 epithelium increased in number, but not in volume, thickness, or
surface area. Interstitium increased in thickness in adults at 0.25 ppm, but not at 0.12 ppm or in juveniles. AMs
increased by 0.25 ppm at both ages, but only older rats had increased interstitial AMs. No differences due to age.
LM morphometry of CAR remodeling. Thickened tips of alveolar septa lining ADs (alveolar entrance rings) 0.2 mm Pinkerton et al.
from TB in rats exposed to 0.12 ppm and to 0.6 mm in rats exposed to 1.0 ppm. Interstitial changes accompanied these (1995)
epithelial changes.
Laser scanning confocal LM immunohistochemistry for CC10 in nonciliated bronchiolar (Clara) cells. Clara cells from Dodge et al. (1994)
rats exposed to 1.0 ppm, but not to 0.12 ppm, had increased cell volume of granule-based CC10, increased CC10
concentration within the granules, and increased number of granules per Clara cell profile.
Rats exposed to 0.5 ppm had smaller BW after 7-weeks exposure. LM histopathology. Nose: At 00.25 ppm, mucous Smiler et al. (1988)
cell respiratory epithelium hyperplasia; no lesions in mainstem or large bronchi. CAR: DO. 25 ppm, TB epithelium Wright et al. (1989,
hyperplastic and hypertrophic; bronchiolarization and airway remodeling. No changes in 0.12-ppm group after 26 weeks 1990)
of exposure. Peribronchiolar tissue and AD walls thickened by eosinophilic material after 12 weeks at 0.5 ppm and after
26-weeks at 0.25 ppm. Collagen found in these areas using special stains. Increased AMs atD0.25 ppm.
-------
Table 6-12 (cont'd). Effects of Ozone on Lung Structure: Long-Term Exposures (>2 Weekl)
CO
Ozone
Concentration
ppm Qg/m
0.15 294
0.3 588
0.25 490
0.25 490
0.25 490
0.3 588
Exposure
Duration
8 h/day for
6 or 90 days
12 h/day for
6 weeks
8 h/day,
7 days/week,
"daily" for 18 mo
or "seasonal" O3
odd months,
filtered air even
months for 18 mo
(9 mo of O3)
8 h/night,
7 nights/week,
"daily" for 18 mo
or "seasonal" O3
odd months,
filtered air even
months for 18 mo
7 h/day,
5 days/week for
6 weeks
Species, Sex
(Strain)
Age"
Macaco radiata, F, M
2-6 years old
Rat, M
(F344)
1 day or
6 weeks old
Macaco fascicularis, M
6 mo old
Rat, M
(S-D)
22 days old
Mouse, M
(Swiss-Webster)
Newborn
Observed Effect(s) 3
TEM, SEM, and LM morphometry. First generation RBs had epithelial hyperplasia, and alveoli opening
into these RBs had increased AMs. RB epithelium thickened, but no difference due to either exposure
time or concentration. RB interstitium was thickened in all exposed monkeys, but both cellular and
acellular compartments were individually thickened only after 90-days exposure to 0.3 ppm. No
differences due to age or gender. No evidence of epithelial cell necrosis nor of inflammatory cell
infiltration other than the increased AMs.
TEM morphometry of the TBs. Luminal surface area covered by cilia decreased, as did the luminal
surface of Clara (nonciliated bronchiolar) cells. Number of brush cells decreased. No differences due to
age.
LM morphometry. Also see Section 6.2.5. Low-grade respiratory bronchiolitis in both exposed groups.
Compared with controls, both groups of exposed monkeys had increased Vv of tissue other than
parenchyma and V, of RBs and their lumens. Both V, and V of RB wall increased in the "daily" group but
not in the "seasonal" group. The only significant morphometric difference between the two exposed
groups was the V, of cells, mostly AMs, free in airspace lumens. This difference and the difference in
significance of the RB wall thickness was presumed due to the difference in time after the last Oj exposure
and necropsy. Daily group necropsied the day after the last exposure, whereas seasonal group necropsied
after a month of filtered air. Seasonal group had an amount of morphological changes similar to the daily
group.
LM morphology and morphometry. Monkey data from Tyler et al. 1988 (above) compared with rats
exposed to a similar regimen. Rats: Estimated the extent of centriacinar remodeling by counting the
number of junctions of bronchioles with ADs per area of lung section (B/A J/cnf ). At the end of the
exposure, both exposed groups had more B/A J/cm2 than filtered air controls. Recovery by 30 days PE.
No difference between the two exposed groups, even though the daily group was exposed twice as many
days as the seasonal group.
LM morphometry of histochemically identified Type 2 cells. Type 2 cells tended to be larger (longer
linear intercepts), and the number per microscope field tended to be greater, but the p values were
>0.05 in final data in which the images were edited electronically. However, these values were
significant (p < 0.05) in unedited data. Exposed mice had larger (p < 0.05) body weights at both 3 and 6
weeks.
Reference
Harkema et al. (1993)
Barry et al. (1988)
Tyler et al. (1988)
Tyler et al. (1991a)
Sherwin and Richters
(1985)
-------
Table 6-12 (cont'd). Effects of Ozone on Lung Structure: Long-Term Exposures (>2 Weekl)
Ozone
Concentration
ppm
Dg/m
Exposure
Duration
Species, Sex
(Strain)
Ageb
Observed Effect(s) 3 Reference
0.35 686 4.5 h/day, Mouse, M
5 days/week for (Swiss-Webster)
4 weeks, 32 g
380 mmHg
(5,400 m)
or sea level
Automated LM morphometry of stainable elastin in alveolar walls. Simulated high altitude (5,400 m) with O, Damji and Sherwin
(SHA-X) or without O3 (SHA-C) resulted in larger lung volumes than sea-level controls (SL-C), but not (1989)
different from each other. Unlike most studies, sea-level, O3-exposed mice had the smallest lung volumes.
Alveolar wall areas, after adjustment to SL-C lung volumes, were increased only in the SHA-X group.
Alveolar wall elastin area, adjusted to the SL-C lung volumes, increased in both high-altitude groups
compared to SL-C and also were different from each other with the largest amount of elastin area in the SHA-
X group. However, if the elastin areas were not adjusted for differences in lung volumes, there were no
differences between the groups.
0.4 784 8 h/day, Macaco radiata, M LM and TEM morphometry with emphasis on RBs. Respiratory bronchiolitis and peribronchiolar
0.64 1,254 7 days/week for 5-8 years old inflammation. RB walls thicker with smaller lumens. Increased wall thickness due both to thicker epithelium
90 days (significant only at 0.64 ppm) and interstitial components. Epithelial changes in both O3 groups include
increased nonciliated bronchiolar epithelial cells and decreased Type 1 cells. Interstitial changes in both O,
groups included increased smooth muscle cells, mast cells, and fibers. Components increased at 0.64 ppm,
but not at 0.4 ppm, included interstitial AMs, PMNs, and amorphous ground substance.
Moffatt et al. (1987)
0.5 980 6 h/day, Rat, M
6 days/week for (Wistar)
2, 3, 5, and 100 g
12 mo
LM, TEM, and LM morphometry of collagen fibers. Bronchitis, peribronchitis, CAR remodeling, and
increased stainable collagen in bronchioles. Rats apparently had intercurrent respiratory disease, as 10 of
44 exposed rats died of pneumonia or pulmonary edema. In addition, it appears that only one set of 12 rats
maintained in room air served as controls for all exposed groups, even though controls and exposed rats
would be of significantly different age and size.
Hiroshima et al. (1989)
0.5
980
20 h/day,
7 days/week for
52 weeks
Rat
(F344)
LM histopathology. 6-mo exposure: Inflammation, mononuclear cells, and fibroblasts in AD walls and walls Gross and White (1987)
of adjacent CAR alveoli. TB not involved. 12-mo exposure: Similar to 6 mo with possibly a slight increase
in AMs, some increased thickening of centriacinar AD and alveolar walls, and a few foci of bronchiolization
(CAR remodeling). 12-mo exposure + 6 mo PE: Slight dilation of ADs, minimal inflammatory reaction,
slight thickening of AD and CAR alveolar walls, and a few foci of bronchiolization.
0.5
0.64
0.96
980 Continuous for
120 days
1,254 8h/night,
1,881 7 nights/week
for 42 nights,
"pair" fed
Mouse, F
(Swiss)
20-23 g
Rat, M
(S-D)
28 days old
LM morphometry and histopathology. Also see Section 6.2.3. More tissue, primarily inflammatory cells, at
Day 9. Little change from Days 10 to 120. Thickened airway walls with increased collagen. Increased
collagen in alveolar walls along the ADs.
LM morphometry and SEM. Also see Section 6.2.5. End of 42-night exposure: No difference in BW,
hemoglobin, or total serum proteins. At 0.96 but not 0.64 ppm, larger fixed and saline- filled lung volumes,
lung volume/BW ratios, and volumes of parenchyma. At both concentrations, increased V, and V of RB and
RB walls and their ratios to BW. By SEM, remodeling of CAR airways with the formation of RBs and
thickened CAR septa at both ppm. After 42 days PE: Fixed lung volume at 0.96 ppm increased. V, and V
of RB walls and ratio to BW increased at 0.96 ppm, as did ratios of volumes to BW for parenchyma, alveoli,
total RB, and RB wall. SEM revealed persistence of CAR remodeling and thickened septa.
Jakab and
(1990)
Tyler et al
Bassett
. (1987)
-------
Table 6-12 (cont'd). Effects of Ozone on Lung Structure: Long-Term Exposures (>2 Weekl)
Ozone
Concentration
ppm Qg/m
0.64 1,254
0.7 1,372
0.95 1,862
Exposure
Duration
8 h/day for
12 mo
20 h/day for
4 weeks
8 h/day for
90 days
Species, Sex
(Strain)
Age1'
Macaco faciscularis, M
6-7 mo old
Rat, M
(F344)
14 weeks old
Rat, M
(S-D)
61 days old
Observed Effect(s) 3
LM morphometry of CAR airway remodeling. Both V^, and V of RBs, and their walls and lumens,
increased at end of exposure and 6 mo PE. RB internal diameters were smaller at exposure end, but not at 6
mo PE. Vv free cells, mostly AMs, increased only at exposure end. No differences in BW or fixed lung
volumes.
LM histopathology. Also see Section 6.2.5. Exposure end: CAR inflammation of TB, AD, and CAR
(proximal) alveoli characterized by edema with mononuclear and leukocyte infiltration. 4 weeks PE: Few
inflammatory foci, edema decreased, and interstitial mast cells. Slight thickening of ADs and septa. 9
weeks PE: Inflammation cleared, TB walls slightly thickened by amorphous matrix.
LM and TEM morphology. RB: Increased volume of total RB and of RB wall and lumen. RB walls
thickened by interstitial inflammation with edema, hyperemia, fibrosis, and hypertrophied smooth muscle
and by interstitial mononuclear cells, granulocytes, and plasma cells. Epithelial and vascular basal lamina
fused. TB: Smaller internal diameter and smaller luminal volume, but no change in total TB volume or in
wall volume. Proximal AD: Most severe cell damage and inflammation at alveolar septal tips (alveolar
entrance rings). Epithelium at these tips was frequently necrotic or missing, leaving bare basement
membrane. Duct walls thicker due mainly to increased interstitial edema, fibrosis, and cellular infiltrates.
Basal lamina thickened, split or duplicated, and had granular deposits. Site of most severe injury shifted
progressively distally as new segments of RB were formed.
Reference
Tyler et al. (1991b)
Gross and White (1986)
Barr et al. (1988)
0.95 1,862 8 h/day for Rat, M
90 days or 5-day (S-D)
episodes followed 52 days old
by 9 days PE
LM and TEM morphometry. Both groups examined at exposure end. The lesions were as previously
described by Barr et al. (1988). Both groups had CAR airway remodeling with the formation of new RBs.
The only morphometric difference in RBs between the groups was the volume of RB wall, which was
greater in the daily group, but both groups were greater than controls. The volume of the total RB and of
RB lumen was increased in the daily group. RB epithelium of the daily group was more differentiated. TB
interstitium was increased in the episodic group. Alveolar duct/sac lumen volume was increased in both
groups with the increase in the episodic group significantly greater than the daily group. Alveolar volume
was decreased in the episodic group. The total amount of CAR damage was not different for both episodic
(35 exposures) and daily (90 exposures) groups.
Barr et al. (1990)
0.96 1,882 8h/night, Rat, M
7 days/week for (S-D)
3 or 234-263 g
60 nights
LM morphometry, histochemistry, autoradiography, and SEM, and TEM morphometry. Neither 3 nor
60 days of exposure altered the cell density of ciliated, serous, basal, brush, migratory, or unidentified cells
in tracheal epithelium. 3 days: Damage to cilia and ciliated cells, including necrosis. Thymidine labeling
index increased. Serous cell histochemistry unchanged. 60 days: Less evidence of injury than at 3 days,
but more damaged ciliated cells than in controls. Complete recovery of the epithelial changes by 42 days
PE.
Nikula et al. (1988a)
0.98 1,921 8 h/day, Rat, M
7 days/week for (S-D)
90 days 65 days old
LM morphometry and SEM of CAR. Remodeling of CAR. Increased thickness of septal edge (tips) of
alveoli, which form the walls of ADs (alveolar entrance rings) up to 0.6 mm from TB. Alveolar septa
thickened by replacement of Type 1 cells by Type 2 and bronchiolar cells to 0.6 mm from TB.
Pinkerton et al. (1992)
-------
Table 6-12 (cont'd). Effects of Ozone on Lung Structure: Long-term Exposures (>2 Week^)
Ozone
Concentration
ppm
1.0
Exposure
Qg/m Duration
1,960 6h/day,
5 days/week for
20 mo
Species, Sex
(Strain)
Age1'
Rat, M
(F344)
6-7 weeks old
Observed Effect(s) 3
LM morphometry, SEM, confocal microscopy, immunocytochemistry, and conventional histochemistry.
Remodeling of CAR. Former ADs were converted to RBs. Bronchiolar epithelium in these former ADs
consisted of well-differentiated ciliated and nonciliated bronchiolar (Clara) cells.
Reference
Pinkerton et al. (1993)
'See Appendix A for abbreviations and acronyms.
''Age or body weight at start of exposure.
-------
Harkema et al. (1987) exposed bonnet monkeys to 0.15 or 0.30 ppm O3, 8 h/day for
6 or 90 days. They sampled four regions of the nasal cavity and nasopharynx. Changes were
limited to the respiratory and transitional epithelium in the two most rostral (anterior) of the
four sections. No changes were reported in the caudal (posterior) two sections, the last of
which included the nasopharynx. The respiratory epithelium of the rostral nasal cavity had
both qualitative and quantitative changes. Quantitative changes included decreased density of
ciliated cells characterized qualitatively by multifocal loss of cilia, necrotic ciliated cells,
ciliated cells with attenuated cilia, and others with only microvillar surface. The respiratory
epithelium also had an increased density of SMG cells, presumably related to repair processes.
Monkeys exposed to 0.30 ppm O3 for 90 days also had increased abnormal cells with
intracytoplasmic lumens containing both cilia and microvilli. Qualitative changes were also
seen in mucous (goblet) cells, which appeared to have fewer secretory granules and dilated
endoplasmic cisternae. Ozone exposure resulted in more nonciliated cells with secretory
granules and with dilated cisternae of the endoplasmic reticulum. Like the respiratory
epithelium, the transitional epithelium had an increased density of SMG cells. In both
epithelia, inflammatory cells were increased only in the monkeys exposed to 0.15 ppm O3 for
6 days. Most of the morphometric changes in the respiratory but not the transitional
epithelium were as large after 6 days of exposure to 0.15 ppm O3 as after 90 days of exposure
to either 0.15 or 0.30 ppm. The histochemistry and cytochemistry of the nasal epithelia from
these monkeys were studied by Dimitriadis (1992). This investigator reported changes in the
intraepithelial mucosubstances and the presence of mucous cells with dilated cisternae in the
granular endoplasmic reticulum.
Acute changes in nasal epithelia from rats exposed to O3 concentrations of 0.12 to
1.0 ppm for 6 h to 7 days have been studied extensively (Table 6-10). In general, short-term
exposure to DO.2 ppm O3 results in either no changes detectable by LM or in mild hyperplasia.
Higher concentrations for up to 7 days can result in damaged cilia, hyperplasia, and increased
stored intraepithelial mucosubstances. Several studies document the hyperplasia using
morphometry or DNA synthesis and document the stored mucosubstance by histochemistry and
morphometry. In one study, the increased stored intraepithelial mucosubstances reached their
largest quantity 7 days postexposure (Harkema et al., 1989). Details of individual studies
follow.
Exposure toO.12, 0.8, or 1.5 ppm O3 for 6 h followed by postexposure periods up
to 66 h resulted in inflammatory changes characterized by increased PMNs, but without LM
evidence of necrosis, ciliary loss, or hyperplasia (Hotchkiss et al., 1989a). Hotchkiss and
Harkema (1992) reported similar LM findings in rats exposed to 0.8 ppm O3 for 6 h. They
also reported increased DNA synthesis by bromodeoxyuridine (BrdU) uptake in nasal
nonciliated transitional epithelium. Exposure to 0.8 ppm O3 6 h/day for 3 or 7 days, or for 3
days with 4 days postexposure, resulted in hyperplasia of the nasal nonciliated cuboidal
(transitional) epithelium with increased intraepithelial mucosubstances without significant
changes in histochemical staining characteristics (Hotchkiss et al., 1991). In that study, no
changes were reported for rats exposed for 3 days and examined 18 h postexposure.
Reuzel et al. (1990) exposed rats to 0.2, 0.4, or 0.8 ppm O3, 22 h/day for 3 days.
They did not report changes in rats exposed to 0.2 ppm, but those exposed to 0.4 or 0.8 ppm
had loss of cilia and disarrangement of the epithelium with hyperplasia and metaplasia. Cell
proliferation, as measured by radiolabeled thymidine, was increased at the two higher
concentrations. The influence of O3 C X T on epithelial cell proliferation in the nasal anterior
6-72
-------
maxilloturbinates was measured by BrdU uptake (Henderson et al., 1993). Rats were exposed
to 0.12, 0.24, and 0.48 ppm O3 for 3, 6, 12, and 24 h, resulting in six C X T products.
Exposure to 0.12 ppm or C X Ts of 0.72 or 1.44 ppm-h did not cause effects. For a given C
X T between 2.88 and 11.52 ppm-h, the increased DNA synthesis was similar; the response
did not increase linearly with increasing C x Ts. Generally, above 0.12 ppm O3 there was a
linear increase with increasing C but not T. Thus, exposure duration apparently was
responsible for the lack of C x T linearity. Johnson et al. (1990) also used BrdU to study
DNA synthesis in rats exposed to 0.12, 0.27, or 0.8 ppm O3, 6 h/day for 3 or 7 days, and
examined 3 or 7 days postexposure. Rats exposed to 0.8 ppm O3, but not to the lower
concentrations, had increased DNA synthesis in the nonciliated cuboidal (transitional)
epithelium at 3 and 7 days and increased numbers of labeled cells in the ciliated respiratory
epithelium and the olfactory epithelium only at 3 days. No changes were found in squamous
epithelia except a decrease in labeled cells 7 days postexposure to 0.8 ppm O3. Johnson and
co-workers reported no LM changes in the ciliated respiratory, olfactory, or squamous
epithelia, but hyperplasia occurred in the cuboidal transitional epithelium.
Epithelial mucosubstances were studied in rats exposed to 0.12 or 0.8 ppm
O3, 6 h/day for 7 days or 7 days postexposure (Harkema et al., 1989). They reported no LM
pathology in the nasal or nasopharyngeal airways from rats exposed to 0.12 ppm, with the
exception of an increase in secretory cells in ciliated epithelium. Rats exposed to 0.8 ppm had
attenuation of cilia in the lateral walls of the nasopharynx; 7 days postexposure, an increase in
stored intraepithelial mucosubstances was observed. The 0.8-ppm group also had hyperplasia
of the nonciliated transitional epithelium accompanied by an increase in PMNs in the lamina
propria. Seven days postexposure, rats in the 0.8-ppm exposure group had more stored
intraepithelial mucosubstances in some areas of ciliated respiratory and nonciliated transitional
epithelia.
In rats exposed to 0.12, 0.25, or 0.5 ppm O3, 20 h/day for 2 years, Smiler et al.
(1988) reported hyperplasia, especially of mucous cells, in the respiratory epithelium over the
rostral portion of the nasoturbinate of rats in the 0.25- and 0.5-ppm groups. The respiratory
epithelium lining other parts of the nasal cavity were less affected, and no changes were found
in the squamous and olfactory epithelia.
Harkema et al. (1994) reported no changes in the amount of mucosubstances in
conducting airways, including the nasal cavity, of rats exposed to 0.12 ppm O3, 6 h/day,
5 days/week, for up to 20 mo. After exposure to 0.5 and 1.0 ppm O3, however, mucous flow
rates were slower and mucous cell metaplasia was evident over the lateral wall and turbinates
of the proximal third of the nasal airways. Exposure to 0.5 and 1.0 ppm O3 also caused
epithelial hyperplasia in nasal transitional epithelium, an increase in eosinophilic globules in
the surface epithelium lining the distal nasal airways, and mild-to-moderate inflammatory cell
influx in the nasal mucosa of the proximal and middle nasal passages.
Larynx
Leonard et al. (1991) reported disruption and thickening of the stratified squamous
epithelium over the vocal folds of bonnet monkeys (Macaca radiatd) exposed to O3, 8 h/day
for 7 days. The basement membrane appeared undulating rather than smooth. At 7 days
postexposure, the epithelium appeared thickened, but otherwise normal. The O3 concentration
to which the monkeys were exposed is not clear because different concentrations appear in the
summary and text sections of the publication. However, these larynges were from bonnet
6-73
-------
monkeys that also were studied by Harkema et al. (1987) and Dimitriadis (1992) and,
therefore, most likely were exposed to 0.15 ppm.
Trachea and Bronchi
Several investigators studied effects of 0.96 or 1.0 ppm O3 on the tracheas of
monkeys, rats, and sheep during and after short-term (very brief) or long-term exposures.
Hyde et al. (1992) studied the trachea, bronchi, and RBs of rhesus monkeys exposed to
0.96 ppm O3 for 8 h and examined them at 1, 12, 24, 72, and 168 h postexposure. Although
the primary objective of the study concerned inflammation (see Section 6.2.2), the study also
provided much new morphometric information concerning reactions to O3 of tracheal,
bronchial, and RB epithelia and their interstitium. Both epithelial and interstitial data were
determined as volume per surface area of epithelial basal lamina (Vs). At 1 h postexposure, the
major change in the tracheal and RB epithelia was an increase in necrotic cells, whereas in the
bronchial epithelium, there were fewer ciliated and basal cells. There were no other changes
in tracheal epithelial cell Vs at any of the postexposure times examined. At 12 and
24 h postexposure, the Vs of necrotic cells was increased in bronchi but not in the trachea or
RBs. The Vs of ciliated and basal cells was smaller in the bronchial epithelium but not in the
trachea. Basal cells in bronchi also were increased at 72 and 168 h postexposure. Respiratory
bronchioles had smaller Vs of Type 1 alveolar epithelial cells at all times except 1 h
postexposure. In RBs, nonciliated bronchiolar cells were increased only at 24 h postexposure.
Epithelial cell DNA synthesis was studied in the filtered air controls and at 1 and 12 h
postexposure by radiolabeled thymidine incorporation. The only increase was observed in the
bronchial epithelium at 12 h postexposure. Changes in the interstitial components of the
trachea were minimal, with a decrease in the amorphous matrix at 24 h postexposure. Bronchi
had increased Vs of smooth muscle and decreased amorphous matrix at 24, 72, and 168 h
postexposure. Collagen fibers in the bronchial interstitium were decreased at 168 h. In RBs,
the arithmetic mean thickness was increased at 12 and 24 h, but not at other times. In RBs,
smooth muscle Vs was increased at 24 h, Vs of fibroblasts was increased at 24 and 72 h, and Vs
of the amorphous matrix was increased at 12 h postexposure.
Nikula et al. (1988a) exposed rats to 0.96 ppm O3, 8 h/night for 3 or 60 nights or
for 60 nights followed by 7 or 42 days postexposure, and examined the tracheas using LM,
TEM morphometry, SEM, LM mucosubstance histochemistry, and DNA synthesis by
radiolabeled thymidine incorporation. Ciliated cells with short or damaged cilia were
increased after 3 and 60 nights of exposure; cells with short cilia were increased after 60 nights
of exposure and 7 days postexposure. Intermediate cells, presumed to be immature ciliated
cells, were increased only after 3 nights of exposure. However, the numeric density of total
ciliated cells, basal cells, total serous cells, brush cells, and total migratory cells was not
different from controls. There were no changes in LM histochemistry for mucosubstances at
any time. The only increase in thymidine labeling occurred after 3 days of exposure.
Recovery was complete 42 days after 60 nights of exposure.
Mariassy et al. (1989, 1990) exposed newborn lambs to 1.0 ppm O3, 4 h/day for
5 days, and studied controls at birth and controls and exposed lambs at 2 weeks of age.
Tracheal mucous velocity was decreased at 2 weeks and at several additional postexposure
times (see Section 6.2.3). In control lambs, the percent of ciliated cells increased and mucous
cells decreased in the tracheas from birth to 2 weeks of age. This normal change in cell
populations did not occur in the exposed lamb tracheas. In the more detailed morphological
6-74
-------
study (Mariassy et al., 1989), epithelial cell density (cells per millimeter), rather than
differential cell counts, was reported. In tracheas from control lambs, the density of mucous
cells decreased from birth to 2 weeks of age. Ozone exposure resulted in decreased total
epithelial cell density, with decreased densities of ciliated and basal cells. Mucous cell density
remained at newborn levels. Ozone exposure also prevented the normal maturational changes
of lectin-detectable mucosubstances but not of tinctorially stained mucosubstances.
The most comprehensive study of the effects of long-term O3 exposure on
conducting airways of rats is that by Plopper et al. (1994a). They used LM morphometry and
tinctorial histochemistry to study conducting airways from the trachea to centriacinar alveoli
following two "short" and one "long" pathway by airway dissection of fixed lungs. In rats
exposed to 0.12, 0.5, or 1.0 ppm O3 for 6 h/day, 5 days/week for 20 mo, the investigators did
not find differences due to O3 exposure in tracheal or bronchial epithelial thickness, cell
populations, or stored glycoconjugates. However, they did find a concentration-dependent loss
of stored glycoconjugates in the tracheas and in the caudal long-path bronchi but not in the
cranial or central short-path bronchi. Although not significantly different from controls, there
was a concentration-dependent thinning of the epithelium in caudal long-path bronchi.
Terminal bronchioles from rats exposed to 0.5 and 1.0 ppm O3 had increased volume fraction
(Vv) of nonciliated bronchiolar (Clara) cells, and the epithelium was thicker in TBs from rats
exposed to 1.0 ppm. In all exposed rats, the mass (Vs) of nonciliated bronchiolar cells was
increased in TBs that had long pathways (caudal) but not in TBs with short pathways (cranial
and central).
Centriacinar Region
As described in the previous criteria document (U.S. Environmental Protection
Agency, 1986) and in the summary of it above, the CAR varies with the species. By common
usage (Weibel, 1963; Schreider and Raabe, 1981; Weibel, 1983; Rodriguez et al., 1987;
Haefeli-Bleuer and Weibel, 1988), the acinus consists of a TB, RBs when present, and the ADs
and alveoli supplied by that TB. In some species (e.g., humans, monkeys, dogs, and cats),
several generations of RBs are found between the TB and ADs. In other species (e.g., rats,
mice, guinea pigs, and rabbits), RBs either are absent or very poorly developed and limited to
a single, very short generation (Tyler, 1983; Tyler and Julian, 1991). The CAR consists of the
TB, RBs if present and alveoli that open directly into RBs, and the initial portions of ADs.
Acini that do not have RBs have a smaller volume than those that do. Rodriguez et al. (1987)
estimated the acinar volume in rat lungs to be 1.86 mm3, and Haefeli-Bleuer and Weibel
(1988), using the same methods, estimated the acinar volume in the human lungs at 187.0
mm3. Mercer and Crapo (1989) and Mercer et al. (1991) found that variation in acinar size
within an individual lung is an important determinant of the intensity of lesions due to inhaled
reactive gases. Thus, the intensity of CAR lesions may vary when animals with differing size
acini are compared (Plopper et al., 1991).
The CAR lesion, both in animals with small acini (e.g., rats) and animals with large
acini (e.g., monkeys) has been well described, both in the original reports and in the 1986
document (U.S. Environmental Protection Agency, 1986). Some of the reports published since
that document contain additional details concerning cellular and interstitial responses in the
CAR to short- or long-term O3 exposure and are presented in this section (Chang et al., 1992;
Pino et al., 1992c; Harkema et al., 1993). Most of these studies need TEM levels of
resolution and magnification and employ morphometric methods. In other reports,
6-75
-------
morphometric estimates of the volume of the CAR lesion were used to study factors that might
alter the intensity of the lesion or evaluate the intensity of reaction to specific exposure
regimens (Mautz et al., 1988; Warren et al., 1988; Stiles and Tyler, 1988). Due to the size
and definition of the CAR lesion, this approach can use LM morphometry to estimate lesion
volume. In other studies, the cumulative effect of O3 on the CAR is estimated by LM
morphometry of one of the components, distal airway remodeling, which results in the
formation of new RBs (Barr et al., 1988; Tyler et al., 1988; Pinkerton et al., 1993). Examples
of each type of study will be presented.
Pino et al. (1992c) exposed rats to 1.0 ppm O3 for 4, 6, 8, or 24 h, followed by
postexposure periods in filtered air for up to 20 h, so that the total exposure and postexposure
period did not exceed 24 h. Some of the rats were used for BAL (see Section 6.2.2), others
for TEM morphometry. The morphometric data are expressed as Vs values. After a 4-h
exposure, necrosis was the dominant morphologic feature, with increases in Vs of necrotic cells
in the TB epithelium (ciliated cells) and in CAR alveoli (Type 1 cells). With increasing time
of exposure or postexposure, the volume of necrotic cells in TBs shifted from the epithelium to
the lumen, with this change being significant at 24 h. In CAR alveoli, increased Vs of total
necrotic cells occurred at 4, 6, and 24 h of exposure and at 24 h in the epithelium. Healing in
the TBs, evidenced by increased Vs of undifferentiated cells, was underway 18 h after a 6-h
exposure, 16 h after an 8-h exposure, and immediately after 24 h of exposure. The only
significant change in viable alveolar cells was an increase in Vs of Type 1 cells after 24 h of
exposure. This increase appeared predominantly due to swelling of individual Type 1 cells.
Increased Vs of total TB interstitium occurred 4 h after an 8-h exposure. In CAR alveoli, total
interstitium was increased after 8 h of exposure, with much of the increase due to an increase
in capillary volume.
Chang et al. (1992) used TEM morphometry to evaluate cellular and interstitial
responses in the CAR, TB, and alveoli (proximal alveoli) of rats exposed to a 9-h peak slowly
rising to 0.25 ppm O3 superimposed on a 13-h background level of 0.06 ppm (the background
was 7 days/week, the peak was 5 days/week). Chang and co-workers examined rats after 1,3,
13, and 78 weeks of exposure; 6 weeks after a 13-week exposure; and 17 weeks after a
78-week exposure. Centriacinar region alveoli had a larger volume of total tissue and total
epithelium per area of basement membrane (Vs) only after 13 or 78 weeks of exposure, and
these values were not different after postexposure periods. Type 1 cells had a larger volume
only at 13 weeks of exposure, increased numbers at 78 weeks, and increased numbers after
13 weeks of exposure plus 6 weeks postexposure. Type 2 cell Vs was increased only after
78 weeks of exposure and 78 weeks of exposure plus 17 weeks postexposure. Macrophages in
the CAR alveoli were increased only after 1 week of exposure. In CAR alveoli, both
interstitial cells and matrix were increased after 1 week of exposure, and the matrix increased
again after 13 and 78 weeks of exposure. This difference was no longer significant after either
postexposure period. Although the data are not in the tables or figures in the article, the text
indicates that both epithelial and endothelial basement membranes were thickened after 13 and
78 weeks of exposure and after the 17-week postexposure period. Crystalline deposits in the
basement membrane are demonstrated in Figure 6-3. In TBs, the luminal surface area of Clara
cells was reduced at 1 week of exposure, and both ciliated and Clara cells had smaller luminal
surface areas after 78 weeks of exposure; these returned to control values during the
postexposure period. However, increased Vs of Type 2 cells persisted for the 17-week
postexposure period that followed the
6-76
-------
"
~" ^ ? .Tqg« ^S-
Figure 6-3. Electron micrograph of alveolar septa in the centriacinar region of the lungs
from laboratory rats exposed to a simulated pattern of ambient O3 for 78
weeks, showing thickened basement membrane (arrow heads). Inset
micrograph shows dense crystaline deposits (arrows) and cellular extensions
(*).
Source: Chang et al. (1992).
78 weeks of exposure. Chang et al. did not find significant bronchiolization of alveoli (i.e.,
distal airway remodeling). This observation may be due to the way the investigators sampled
the CAR proximal alveoli and the strict orientation of the airways to obtain exact cross
sections. This study complements those of Barry et al. (1985, 1988), who used similar TEM
morphometric methods, by extending the exposure period and adding postexposure periods.
Cellular and interstitial changes were studied in nonhuman primates exposed
8 h/day for 90 days by both Moffatt et al. (1987) and Harkema et al. (1993), using TEM and
morphometry. The study by Harkema et al. (1993) is reported here because the concentrations
used, 0.15 and 0.3 ppm O3, were lower than those used by used Moffatt et al. (1987), which
were 0.4 and 0.64 ppm. Harkema and co-workers also studied reactions after only 6 days of
exposure to 0.15 ppm. There were no major differences among the three exposed groups
6-77
-------
(i.e., 6 days to 0.15 ppm, 90 days to 0.15 ppm, and 90 days to 0.3 ppm O3). All exposed
monkeys had thicker RB epithelium and thicker RB interstitium. There were more nonciliated
cuboidal epithelial cells per millimeter of basement membrane in all exposed groups and
increased squamous cells only in the 6-day group. The thickened total interstitium was due to
increases in both acellular (matrix) and cellular components, but both compartments were
increased individually only in the 90-day O3 group. There were no differences in RB smooth
muscle. Transmission electron microscopy and SEM observations, which were not studied
quantitatively, include increased AMs in alveoli opening into RBs and increased "dome"-
shaped nonciliated bronchiolar cells, which had more apical cytoplasm, more agranular and
granular endoplasmic reticulum, more mitochondria, and more Golgi with secretory granules.
With the exception of increased AMs, there was no evidence of necrosis nor of inflammatory
cells. No differences due to age or gender were detected.
Harkema et al. (1993) speculate that their finding of a larger percent increase in RB
cuboidal cells in monkeys exposed to a lower concentration of O3 for the same time than that
reported by Moffatt et al. (1987) might be due to the difference in sampling methods.
Harkema et al. studied only the first generation RBs, whereas Moffatt et al. (1987) studied a
random sample of all generations of RBs. The first generation tends to be more damaged than
succeeding generations (Mellick et al., 1977; Eustis et al., 1981). The Harkema et al. (1993)
sampling procedure also prevented examination for the increased volume density of RBs and
decreased RB diameter reported by Fujinaka et al. (1985) and Moffatt et al. (1987).
Remodeling of Centriacinar Region Airways. This is a less well-known sequela of long-term
O3 exposure. Using SEM, Boorman et al. (1980) and, later, Moore and Schwartz (1981)
reported the development in rats of an airway with the appearance of RBs between the TB and
ADs. This new segment was longer than those occasionally seen in control rats. Respiratory
bronchioles in rats either are absent or developed to only a single, very short segment (Tyler,
1983; Tyler and Julian, 1991).
Barr et al. (1988) examined the development of this new segment using LM and
TEM morphometry on lungs from rats exposed to 0.95 ppm O3, 8 h/day for 90 days. They
reported a significant increase in the total volume of RB and of RB lumen and wall. The new
RBs reached a maximum length of four alveolar opening rings. They also noted that, in some
of these RB segments, the capillary and epithelial basal laminae were fused as they are in TBs,
rather than separate as in alveoli. Most Type 1 cell necrosis was found at the tips of alveolar
septa immediately adjacent to the RB/AD junction. Thus, the most severe epithelial damage
did not occur at the most proximal alveolus in the CAR, but rather in the alveolus immediately
distal to the newly formed RB.
Recently, Pinkerton et al. (1993) developed a new LM morphometric method to
evaluate remodeling of CAR ADs. In rats exposed to 1.0 ppm O3 intermittently for 20 mo, the
investigators reported well-differentiated ciliated and nonciliated bronchiolar epithelium lining
CAR airways that would otherwise be ADs. Some of this epithelium extended five alveoli
from the TB. Thus, the Type 1 and 2 cells characteristic of ADs were replaced by both types
of bronchiolar cells characteristic of RBs when RBs are present in control rats. Pinkerton
et al. (1995) used their new morphometric method to study rats exposed to 0.12, 0.5, or 1.0
ppm O3 for 6 h/day, 5 days/week for 20 mo. They reported significant thickening of alveolar
septal tips 200 mm from the TB in rats exposed to 0.12 ppm, which increased with
O3 concentration to 600 Dm in rats exposed to 1.0 ppm, but they did not describe the type of
6-78
-------
epithelium covering these thickened tips. Several studies that did not find CAR remodeling
also used a slightly different procedure (Barry et al., 1985; Chang et al., 1992).
Plopper et al. (1994a) examined CARs from rats exposed to the same regimen, but
studied CARs from one cranial short pathway and a caudal long pathway. They found
nonciliated bronchiolar epithelial cells in remodeled former ADs in short- and long-pathway
CARs from rats exposed to 1.0 ppm O3, but only in short-pathway CARs from rats exposed to
0.5 ppm. Central, short-pathway CARs were not examined. Using the airway dissection
method of selecting CARs to be studied, nonciliated bronchiolar cells were not found in ADs
from rats exposed to 0.12 ppm O3.
The same phenomena apparently occurs in animals with several generations of RBs
as increases in Vv and volume (V) of RBs have been reported in all O3-exposed monkeys
examined using morphometric methods to estimate Vv or V of RBs (Fujinaka et al., 1985;
Moffatt et al., 1987; Tyler et al., 1988, 1991b). Inflammatory changes and CAR remodeling
occur concomitantly, and inflammatory changes in an airway may indicate future remodeling.
Mellick et al. (1977) noted that, in monkeys exposed to 0.8 ppm, 8 h/day for 7 days, the
inflammatory process extended throughout the RBs and into ADs. Eustis et al. (1981) reported
that, in monkeys exposed to 0.8 ppm, 8 h/day for 90 days, all generations of RBs contained
aggregates of inflammatory cells. Monkeys exposed to lower concentrations for the same or
longer time have increased Vv and V of RBs (Moffatt et al., 1987).
6.2.4.3 Considerations of Exposure Regimens and Methods
Recovery During Postexposure Periods
Evidence of healing occurs soon after short-term O3 exposures cease. In the studies
of Pino et al. (1992c), evidence of healing is provided by the increased Vs of viable
undifferentiated cells in TBs detected 16 h after the end of an 8-h exposure to 1.0 ppm O3.
Chang et al. (1992) reported an increased Vs of Type 2 cells 17 weeks after a
78-week exposure to a simulated urban exposure regimen with a peak O3 concentration of
0.25 ppm; there were no changes detected 6 weeks after a 13-week exposure. Gross and
White (1987) examined rats 3 and 6 mo after a 52-week exposure (20 h/day, 7 days/week) to
0.5 ppm O3. Using LM pathology, the only changes visible 6 mo after a 12-mo exposure were
a few areas of bronchiolization, slight dilation of ADs, and slight thickening of AD walls and
adjacent alveolar septa. In an earlier study, less complete healing was reported by Gross and
White (1986), who used LM pathology to study rats 4 and 9 weeks after a 4-week exposure
(20 h/day, 7 days/week) to 0.7 ppm O3. Four weeks postexposure, Gross and White reported
a slight, unevenly distributed inflammatory reaction with condensed eosinophilic material,
presumed to be collagen, in the interstitium. Nine weeks postexposure, some AD walls and
TBs were thickened. Rats exposed to 0.96 ppm, 8 h/night for 42 nights and examined 42 days
later using LM morphometry had increased Vv and V of the RB wall and SEM evidence of
CAR remodeling (Tyler et al., 1987). Collagen content of these lungs increased during the
postexposure period (Last et al., 1984b).
Centriacinar region remodeling was more persistent in monkeys exposed to
0.64 ppm O3, 8 h/day for 12 mo followed by 6 mo postexposure (Tyler et al., 1991b). By LM
morphometry, the Vv and V of total RB, RB lumen, and RB walls were increased both at
exposure end and at 6 mo postexposure. At exposure end, but not at 6 mo postexposure,
RB internal diameters were smaller, and AMs in the CAR increased.
6-79
-------
One study concerned postexposure recovery of the trachea (Nikula et al., 1988a).
Complete recovery of the trachea (as evaluated by LM morphometry, SEM, and TEM) of rats
exposed to 60 nights (8 h/night, 7 days/week) to 0.96 ppm O3 occurred following a 42-day
postexposure period.
Effects of Episodic and Seasonal Exposure Regimens
Many investigators have noted that lesions due to O3 reach a maximum intensity in a
very few days and that, with continued exposure, the intensity of the lesion decreases. Eustis
et al. (1981) reported half the number of inflammatory cells in the CAR of monkeys exposed to
0.8 ppm for 90 days as found in monkeys exposed to the same concentration for 7 days.
Chang et al. (1992) noted that the acute reactions to the 0.06-ppm background
O3 (7 days/week), with a 9-h peak (5 days/week) slowly rising to 0.25 ppm, that they reported
at 1 week of exposure had subsided at 3 weeks of exposure. Harkema et al. (1993) reported no
difference in first generation RB epithelial thickness or cell numbers among monkeys exposed
to 0.15 ppm O3 for 6 days or to 0.15 or 0.3 ppm for 90 days.
These and other similar observations prompted Chang et al. (1991) to compare
effects of two exposure regimens, which were evaluated using the same TEM morphometric
approach. The first regimen was a "square wave", 12-h/day, 7-day/week exposure to 0.12 or
0.25 ppm O3. The second regimen simulated urban O3 exposures by exposing rats
7 days/week for 13 h to 0.06-ppm background, with a peak slowly rising to 0.25 ppm over a
9-h period (5 days/week). They calculated cumulative O3 concentration (C x T) for each
exposure regimen and concluded that increases in volume of Type 1 and 2 alveolar epithelial
cells were linearly related to increasing C X T. The relationship for Type 1 cells was more
robust.
Barr et al. (1990) used TEM and LM morphometry to compare effects of 90 days
of daily exposure of rats for 8 h/day to 0.95 ppm O3 with a regimen that modeled 5-day
episodes of O3 exposure. Each 5-day episode was followed by 9 postexposure days of filtered
air. The cycle was repeated seven times so that the "episodic" group was exposed a total of
35 days over an 89-day period, and the "daily" group was exposed for 90 days to the same
O3 concentration. Both groups had CAR remodeling with the formation of RBs. The volume
of RBs formed was not different when the two exposure groups were compared. The absolute
volume of parenchymal lesion was the same in both groups. The RB epithelial thickness was
increased in the daily group but not in the episodic group; conversely, the interstitium of both
TBs and ADs was thickened in the episodic group but not in the daily group. Thus, rats
exposed to the same concentration of O3 for 35 days over an 89-day period in an episodic
regimen had lesions as severe as those rats exposed daily for 90 days.
Effects of "seasonal" and "daily" exposure of young monkeys to 0.25 ppm O3 were
reported by Tyler et al. (1988). The daily group was exposed every day (8 h/day) for 18 mo,
whereas the seasonal group was exposed only during odd months for the 18 mo. Thus, the
daily group was exposed twice as many days to the same concentration as the seasonal group.
By LM morphometry, both groups had increased Vv of total RB and RB lumen, but RB wall
thickness was increased only in the daily group. The only significant morphometric difference
between the two groups was an increase in CAR AMs in the daily group. This difference, and
the difference in significance of the RB wall thickness in the seasonal group, was presumed due
to the daily group being exposed to O3 the day before necropsy, whereas the seasonal group
breathed filtered air for 30 days preceding necropsy. This final 30 days of filtered air
6-80
-------
apparently allowed the more acute inflammatory changes in the seasonal group to regress. The
seasonal group, but not the daily group, had increased lung collagen (Section 6.2.1) and
increased chest wall compliance (Section 6.2.5). Exposure to the same concentration of O3 for
half as many days in a seasonal regimen resulted in morphometric effects similar to daily
exposure and in physiological and lung collagen changes not found in the daily group.
Tyler et al. (199la) exposed rats to seasonal and daily regimens similar to those
used for the monkeys described above. The concentration used in both studies was 0.25 ppm
O3 and the total length of exposure was 18 mo. Rats were exposed nights, during their natural
period of activity. Both groups of rats were studied at the end of the 18-mo exposure cycle
and 30 days postexposure. The lungs were evaluated using a simplified LM morphometric
method for CAR airway remodeling, estimating the number of junctions of bronchioles (TB
and RB) with ADs per surface area of section. At exposure end, the number of junctions of
both exposure groups was increased compared to filtered-air controls; the O3 groups were not
different from each other. Neither group was different from the controls at 30 days
postexposure.
Ex Vivo and In Vitro Exposures
Results obtained from studies of isolated perfused lungs and organ culture explants
were consistent with some of the findings from in vivo studies (Pino et al., 1992a; Nikula
et al., 1988b; Nikula and Wilson, 1990).
6.2.4.4 Considerations of Degree of Susceptibility to Morphological Changes
Species Differences in Degree of Response
Plopper et al. (1991) reviewed data from nonhuman primates and rats that had been
exposed to O3 and evaluated using TEM morphometry. The data were generated in several
laboratories and the exposure and evaluation methods were somewhat different, but the data
were expressed in similar terms. In the CAR, the results were expressed as total epithelial
thickness or numbers of cells per square millimeter of basal lamina. Exposure of rats to
0.25 ppm O3, 8 h/day for 42 days, resulted in an increase of less than 100% in either
parameter compared to controls (Barry et al., 1985, 1988). Exposure of monkeys to 0.15 ppm
O3, 8 h/day for 6 days, resulted in a 230% increase in thickness and a 700% increase in cell
number compared to controls (Harkema et al., 1993). As noted earlier (Section 6.2.4.2), the
CARs of rats and monkeys are structurally different (Tyler, 1983), and the CAR cells are also
different (Plopper, 1983).
There was also a difference when Plopper et al. (1991) compared stored secretory
product per square millimeter of basal lamina in the nasal septum and lateral wall of the nasal
cavity of O3-exposed rats and monkeys. Data from the exposure of rats to 0.12 ppm, 6 h/day
for 7 days, resulted in a < 10% increase in the nasal septum and a < 100% increase in the
lateral wall. Exposure of monkeys to 0.15 ppm, 8 h/day for 6 days, resulted in a 300%
increase in the nasal septum and a 125% increase in lateral wall. As in the CAR, there are
major morphological differences in the nasal cavities of these two species (Schreider and
Raabe, 1981).
Plopper et al. (1991) also compared collagen metabolism in rats and monkeys
exposed to 1.5 ppm O3, 23 h/day for 7 days, using the uptake of tritium-labeled proline.
In rats, there was an increase of 200% above controls, whereas the increase was 800% in
monkeys.
6-81
-------
From these data, it appears that the respiratory system of monkeys is much more
responsive than that of rats to near-ambient concentrations of O3. The mechanisms responsible
for these species differences in response to O3 remain to be elucidated.
Effects of Age
Several studies published since the previous criteria document (U.S. Environmental
Protection Agency, 1986) have addressed the effects of age on the intensity of
O3 morphological changes. The study by Stephens et al. (1978) and the initial report by Barry
et al. (1983) were cited. Briefly, Stephens et al. exposed rats ranging in age from 1 to 40 days
old to 0.85 ppm O3 for 24, 48, or 72 h and examined their lungs by LM and TEM. Stephens
and co-workers reported that, prior to 20 days of age, they did not find damage to TB-ciliated
cells or to CAR Type 1 cells and that the amount of injury increased from 21 to 35 days when
a plateau in response was reached.
Barry et al. (1985, 1988) exposed 1-day-old and 6-week-old rats to 0.12 or
0.25 ppm O3, 12 h/day for 6 weeks. The 1985 study emphasized TEM morphometry of CAR
(proximal) alveoli. The investigators did not find differences in response due to age. In both
age groups, they found Type 1 cells increased in number and thickness, but decreased in both
luminal and basement membrane surface area. They found bare basement membrane where
Type 1 cells had been sloughed, but the amount was not increased in exposed groups. In the
0.25-ppm groups, but not in the 0.12-ppm groups, Type 2 cells were increased in density per
square millimeter basement membrane but not in volume. Alveolar interstitium was increased
only in adults exposed to 0.25 ppm. Macrophages in alveoli were increased in both age groups
exposed to 0.25 ppm, but not in adults exposed to 0.12 ppm. Interstitial AMs were increased
only in adults exposed to 0.25 ppm. The TEM morphometry of TBs from these rats did not
include adults exposed to 0.12 ppm (Barry et al., 1988). There were no differences due to age
at start of exposure. In both juvenile and adult rats exposed to 0.25 ppm, Barry and co-
workers found that the luminal surface covered by cilia and by nonciliated bronchiolar (Clara)
cells was reduced. The number of brush cells was also decreased.
Stiles and Tyler (1988) studied effects in a wider range of ages using LM
morphometry and SEM. They exposed 60- and 444-day-old female rats to 0.35 or 0.8 ppm O3
continuously for 72 h. Body weights of the 444-day-old rats, but not of those 60 days old,
decreased during exposure. Fixed lung volumes of 444-day-old rats exposed to 0.8 but not
0.35 ppm were smaller than same-age controls. The Vv of CAR lesions was larger in
60-day-old rats than in the 444-day-old rats exposed to either concentration. The Vv of cells
free in lumens (AMs) was increased in young rats exposed to 0.35 ppm compared to the older
rats, but was not different for rats exposed to 0.8 ppm. Young rats exposed to either
concentration had larger CAR lesions than the older rats, and young rats exposed to the lower
concentration had more AMs. Older rats had greater changes in body weight and, in those
exposed to the higher concentration, in fixed lung volume.
Effects of Exercise
Exercise increases the dose of inhaled toxicants delivered to sensitive cells (see
Chapter 8). Mautz et al. (1985b) studied the effects of 0.2 and 0.38 ppm O3 on rats at rest and
during several treadmill exercise protocols. They found increased percent of lung parenchymal
area containing free cells (AMs) in exercised rats exposed to both concentrations compared to
6-82
-------
rats exposed at rest. At the higher concentration, there was also an increase in the percent of
parenchymal area with thickened ADs and alveolar septa.
Tyler et al. (1991c) exposed thoroughbred horses (trained to a treadmill) to 0.25 or
0.8 ppm O3 for 29 min on 2 consecutive days using a protocol that included 9 min of graded
exercise (3 min at maximum speed) and 20 min of "cool out". During maximal exercise,
horses increase their rate of oxygen consumption more than other species. Two of three horses
exposed to 0.8 ppm O3 had significant areas of hemorrhage and edema, and one of them
refused the second day's exercise and exposure. By TEM, all horses exposed to 0.8 ppm had
CAR lesions including necrosis of Type 1 cells. Lesions in those exposed to 0.25 were limited
to CAR ciliated cells. No horses were exposed at rest for comparison.
Elastase-lnduced Emphysema
Rats with elastase-induced emphysema and saline-instilled controls were exposed to
0.15 or 0.5 ppm for 3 or 7 days (Dormans et al., 1989). Mean linear intercepts, a measure of
alveolar size, were determined using LM. The incidence and severity of CAR inflammatory
changes were the same in O3-exposed elastase-treated and saline-control rats. There were no
changes in mean linear intercepts due to the O3 exposure.
6.2.4.5 Summary
Research since the previous O3 criteria document (U.S. Environmental Protection
Agency, 1986) continues to support the concept that all mammalian species respond to
O3 concentrations < 1.0 ppm in a similar manner, but with significant differences in intensity
of reactions among the species studied (Plopper et al., 1991). Dungworth (1989) provided a
schematic overview of morphological reactions of the CAR from mammalian lungs to
continuous exposure to low concentrations of O3 as a series of time-response profiles
(Figure 6-4). Bronchoalveolar exudative processes are the predominate early response, but the
magnitude decreases rapidly with increasing duration of exposure and continues to decline
during postexposure periods. Epithelial hyperplasia also starts early and increases in
magnitude for several weeks, after which a plateau is reached until the exposure ends.
Epithelial hyperplasia declines slowly during postexposure periods. Interstitial fibrosis has a
later onset and may not be apparent for a month or more. The magnitude of this response,
however, continues to increase throughout the exposure and, at least in some cases (Last et al.,
1984b), continues to increase after exposure ends.
Nonhuman primates appear to respond more than rats to O3 at concentrations
< 1.0 ppm. However, the mechanisms responsible for these differences in response have not
been elucidated. Differences in cell, tissue, and circulating levels of several antioxidants are
being studied, as are differences in in vitro responses to O3 by cultures of cells from the
various species. Basic morphological differences in the structure of the most injured portion of
the lung, the CAR, and the size (volume) of the basic structural unit, the acinus, may also be
factors in the greater response of monkeys to O3. Both human lungs and lungs from nonhuman
primates have CARs characterized by several generations of RBs, whereas rats have no RBs,
or only a single poorly developed generation (Tyler and Julian, 1991). Within an individual
lung, acinar volume is directly related to the intensity of CAR lesions (Mercer and Crapo,
1989; Mercer et al., 1991). The volume of individual acini in human lungs is 100 times larger
than individual acini in rat lungs (Rodriguez et al., 1987; Haefeli-Bleuer and Weibel, 1988).
6-83
-------
Acinar volume of the monkeys used in O3 studies is not known but, on the basis of the CAR
structure, is assumed to be more like that of human lungs than rat lungs.
Another morphological factor that may be responsible in part for the greater
response to O3 of nonhuman primates than of rats may be differences in the complexity of the
nasal cavity. Schreider and Raabe (1981) studied the cross-sectional morphology of the
nasal-pharynx in rats, beagle dogs, and a rhesus monkey. They concluded that the
6-84
-------
0)
(0
c
o
Q.
(A
0)
o:
0)
T3
c
D)
(Q
Epithelial hyperplasia
Bronchioloalveolar exudate
Interstitial fibrosis
Exposure
Postexposure
1 mo
days
6 mo
Time
12 mo
Figure 6-4.
Schematic comparison of the duration-response profiles for epithelial
hyperplasia, bronchoalveolar exudation, and interstitial fibrosis in the
centriacinar region of lung exposed to a constant low concentration of
ozone.
Source: Dungworth (1989).
complexity of the nasal cavity, and therefore the "scrubbing" effect (Yokoyama and Frank,
1972; Miller et al., 1979), which reduces the concentration of inhaled O3 delivered to the
lower respiratory tract, would be greater in rats than in monkeys. Schreider and Raabe (1981)
proposed that, with appropriate scaling, the monkey could serve as a model for aerosol and gas
deposition in the nasopharyngeal region of humans. However, the sensitivity of the
nasopharyngeal epithelium may be different because changes in the nasal epithelium that follow
O3 inhalation, like those in the CAR, are more severe in monkeys than in rats (Plopper et al.,
1991).
The effect of age at start of exposure on O3-induced lung injury has not been
resolved. Barry et al. (1985, 1988) reported no differences in TEM morphometry of CAR and
TB lesions due to age at start of exposure. However, they studied a narrow range of ages, 1
and 42 days old at the start of a 42-day exposure. Barry et al. (1985, 1988) speculate that
much of the CAR response in the rats that began exposure at 1 day of age might have occurred
during exposure Days 21 to 42 because, as the earlier studies of Stephens et al. (1978) found,
6-85
-------
rats are not sensitive to morphological effects of O3 until weaning at 21 days of age. Thus, the
rats that were 1 day old at the beginning of exposure may have developed the same intensity of
lesions during the last 21 days of exposure as the older rats did in 42 days of exposure. In
studies using a wider range of ages (60- and 444-day-old rats), Stiles and Tyler (1988) reported
larger CAR lesions in the younger rats but greater changes in body weight and fixed lung
volume in the older rats following a 3-day exposure to O3.
The effects of exposure regimen and duration were evaluated in more recently
published studies. Exposure of young monkeys to 0.25 ppm O3 in a "seasonal" regimen (i.e.,
exposure in odd months and postexposure in even months) for 18 mo resulted in the same
quantity of CAR lesions as daily exposure to the same concentration for 18 mo (Tyler et al.,
1988). A similar quantity of CAR lesions was reported in rats exposed to 0.95 ppm O3 in a 35
day, "episodic" regimen (units of 5-day exposures and 9 days without exposure) for a total of
89 days, as those exposed to the same concentration each day for 90 days (Barr et al., 1990).
Chang et al. (1991) calculated the cumulative O3 concentration for a "square wave" exposure
to 0.12 or 0.25 ppm for 12 h/day, 7 days/week, with a simulated "urban" exposure regimen of
0.06 ppm for 13 h/day, 7 days/week and then raising that background 5 days/week to a peak of
0.25 ppm over a 9-h period. Using TEM morphometry of the CAR, they found no difference
due to the pattern of exposure. Thus, it appears that the pattern of daily exposure does not
influence the intensity of CAR lesions, but that episodic and seasonal patterns of exposure,
with multiple days of clean air between days of exposure, are equivalent to daily exposure.
It has become clear that remodeling of centriacinar airways is cumulative. Using a
stereological approach, Barr et al. (1988) reported an increase in the total volume of RB wall
and lumen in rats exposed to 0.95 ppm O3, 8 h/day for 90 days. Barr and co-workers also
reported continuing Type 1 cell necrosis at the tips of alveolar septa (alveolar opening rings)
immediately distal to the newly formed RB/AD junction (rather than in the TB/AD junction).
It appears that some of the necrotic Type 1 cells were replaced by bronchiolar epithelium,
rather than by Type 2 cells as previous studies indicated. This was confirmed by Pinkerton
et al. (1993), who reported fully differentiated ciliated and nonciliated bronchiolar epithelium
lining alveolar tips along a former AD up to 1 mm from the TB in lungs from rats exposed to
1.0 ppm O3 for 6 h/day, 5 days/week for 20 mo. Remodeling of centriacinar airways appears
to be a general phenomena, as increases in the Vv and V have been reported in lungs from all
exposed rats and monkeys examined using stereological or morphometric methods that could
detect this change (Fujinaka et al., 1985; Moffatt et al., 1987; Tyler et al., 1987; Tyler et al.,
1988; Barr et al., 1990; Pinkerton et al., 1992, 1993). Centriacinar region remodeling has
been demonstrated to persist in monkeys 6 mo after a 12-mo exposure to 0.64 ppm O3 (Tyler
et al., 1991b) and in rats 42 days after a 42-night exposure to 0.96 ppm O3 (Tyler et al.,
1987).
Several studies have confirmed and extended the earlier reports of epithelial
degenerative changes followed by sloughing (i.e., leaving bare basement membrane that is
recovered by other cell types), thus altering epithelial cell populations and increasing cell
density (hyperplasia) in TB and centriacinar alveoli (Barry et al., 1985, 1988; Moffatt et al.,
1987; Chang et al., 1988, 1992; Harkema et al., 1993). Epithelial replacement, a reparative
process, occurs very early (Pino et al., 1992c), even though degeneration and necrosis
continues (Barr et al., 1988). In specific airways, these processes appear to reach a maximum
early in the exposure, as reported by Harkema et al. (1993), who found no difference in the
intensity of lesions in first generation RBs, as measured by RB epithelial cell thickness and
6-86
-------
numbers, among monkeys exposed to 0.15 ppm O3 for 6 days or to 0.15 or 0.3 ppm for 90
days. However, it is important to note that there may have been differences in response if
more distal generations of RBs or random generation RBs had been selected for study. The
interstitium in the CAR also thickens by the addition of cells and matrix. Thickening of the
basement membrane and the presence of granular material in it were reported by Barr et al.
(1988) and Chang et al. (1992). Chang et al. (1992), using TEM morphometry, reported that
some changes in epithelial cell populations persist in rats for 17 weeks after a 78-week
exposure to a model urban profile with a peak of 0.25 ppm O3. At the LM level, Gross and
White (1987) reported that 6 mo following a 12-mo exposure to 0.5 ppm O3, 20 h/day, CAR
inflammation had all but disappeared, and only a slight dilation and thickening of some ADs
and adjacent alveoli remained.
The epithelia of the nasal cavity respond rapidly to O3. In ciliated regions, cilia are
attenuated, and intraepithelial mucosubstances increase. Hyperplasia and increased
intraepithelial mucosubstances are reported in areas of nonciliated transitional epithelium
(Harkema et al., 1989). These effects persisted throughout a 20-mo intermittent exposure of
rats to 0.5 or 1.0 ppm O3, but were not seen in rats exposed to 0.12 ppm for that period
(Harkema et al., 1994). After acute exposure, DNA synthesis of the epithelium of the anterior
maxilloturbinates of rats increases according to a given C x T product at D2.88 ppm-h, but the
increase is not linear with increasing C x T (Henderson et al., 1993). Changes have not been
reported in the olfactory epithelium or in the squamous epithelium of the nasal cavity.
Respiratory epithelia in other conducting airways, especially the trachea, appear to
react in a manner similar to early necrosis of ciliated cells (Hyde et al., 1992). Cell
replacement starts early (Hyde et al., 1992), and, after 60 nights of exposure of rats to
0.96 ppm O3, numeric density of specific cell types was not different from controls (Nikula
et al., 1988a). In newborn lambs exposed to 1.0 ppm O3, 4 h/day for 5 days and examined
9 days later, the normal change in epithelial cell population that occurs by 2 weeks of age did
not occur (Mariassy et al., 1990).
6.2.5 Effects on Pulmonary Function
6.2.5.1 Introduction
Numerous studies have been published on the effects of O3 exposure on pulmonary
function in animal models. This work has been reviewed by the U.S. Environmental
Protection Agency (1986) and Tepper et al. (1995). The evaluation of pulmonary function
after exposure may help provide a more integrated assessment of the severity of health effects
by indicating the magnitude, location, and duration of functional disability. In an attempt to
summarize the literature here, only key studies employing multiple concentrations or studies
demonstrating a particular functional effect, testing a different species or strain, or showing the
relationship with a unique variable such as age or sex will be discussed. Because purely
descriptive pulmonary function studies now are rarely reported, newer studies will be
discussed within the context of the study hypothesis. To enable discussion of the full range of
studies, some exposures greater than 1 ppm O3 will be discussed. For example, many of the
airway reactivity mechanism studies were conducted at higher concentrations.
This section is organized by duration of exposure (brief, acute, repeated, and long-
term). Within each of these sections, there are subsections on different types of pulmonary
function measures. These subsections include a discussion of ventilatory patterns, breathing
6-87
-------
mechanics, airway reactivity, and more extended characterizations of lung function, whenever
such data are available.
Ventilation
Evaluation of the sinusoidal breathing pattern includes the measurement of tidal
volume (VT) and frequency of breathing (f) and their product, VE. Such measurements have
proven to be sensitive indicators of O3 effects. Numerous animal and human studies have
shown that O3 exposure increases f and decreases VT (tachypnea) (U.S. Environmental
Protection Agency, 1986). Although there is evidence indicating that tachypnea may serve to
protect the deep lung from exposure, other evidence indicates that this sign of pulmonary
irritation represents deep lung toxicity and is of greater concern than breathing pattern changes
indicative of upper airway irritation.
Breathing Mechanics
Measurement of breathing mechanics (dynamic compliance [Cdyn] and total
pulmonary resistance [RJ) in animals has an advantage over simple measures of ventilation in
that these parameters can assess the mechanical effort required to breathe and can help localize
the site of dysfunction to the airways (resistance) or the parenchyma (compliance). With
sufficient O3 exposure, increases in RL and decreases in Cdyn have been observed (U.S.
Environmental Protection Agency, 1986). Changes in RL and Cdyn typically reverse rapidly
after high ambient O3 exposures; however, these alterations can signal underlying
inflammatory or lung permeability changes.
Airway Reactivity
Increased airway reactivity, an exaggerated response of the lung to an exogenously
administered bronchoconstrictor, has been observed with O3 (U.S. Environmental Protection
Agency, 1986). Typically in humans, heightened airway responsiveness is determined using
progressively increasing concentrations of aerosolized bronchoconstrictors, such as
methacholine or histamine (see Chapter 7). Although bronchoprovocation protocols employing
doubling doses of inhaled bronchoconstrictors have been relatively standardized for human
experiments, no such standardization exists for animal studies, making comparisons between
animal and human studies difficult. Increased airway reactivity is a hallmark of asthma and
occurs in many other lung diseases, yet the long-term pathological consequences of
hyperreactive airways are unknown.
Extended Functional Characterizations
A more complete assessment of the nature and magnitude of functional changes
related to O3 exposure includes an extended characterization of the lung using a battery of
human clinical pulmonary function test analogs. Such tests include measurement of static lung
volumes, volume-pressure and flow-volume relationships, as well as evaluation of
inhomogeneity of ventilation and problems associated with oxygen diffusion across the
epithelial barrier. Although these latter measurements are technically complex, they may
contribute to a more in-depth understanding of the nature and severity of the physiological
impairment and may provide in vivo evidence to suggest the anatomical localization of the
functional abnormality.
6-88
-------
6.2.5.2 Brief Ozone Exposures (Less Than 30 Minutes)
Few experiments have evaluated the effects of brief exposures (< 30 min) to O3.
Most of these brief exposure studies have examined changes in regional breathing mechanics
through exposures to the lower respiratory tract via a tracheal tube, thus eliminating any
scrubbing by the nasal or oropharynx and thereby increasing the effective dose of O3 delivered
to that region of the lung. The relevance of this method of delivering the exposure, as
compared to the typical inhalation route, is uncertain. However, positive effects have been
observed, indicating that very rapid reflex responses occur with brief, direct O3 exposure.
Ventilation and Breathing Mechanics
No studies have evaluated the effects of brief O3 exposure on ventilatory pattern;
however, breathing mechanics (Cdyn and RL) have been evaluated. The previous criteria
document (U.S. Environmental Protection Agency, 1986) described two experiments by
Gertner et al. (1983a,b,c), which demonstrated increased collateral resistance within 2 min of
exposure to 0.1 ppm O3 in anesthetized dogs exposed via a fiber-optic bronchoscope wedged
into a segmental airway. The response rapidly attenuated with exposure to 0.1 but not
1.0 ppm. Atropine or vagotomy blocked the increase in collateral flow resistance to 0.1 ppm,
indicating that vagal postganglionic stimulation was involved, but the response to the 1.0-ppm
O3 exposure was blocked only partially.
More recently, Kleeberger et al. (1988), using a technique similar to Gertner et al.
(1983a), exposed the segmental airways of mongrel dogs to 1.0 ppm O3 for 5 min through a
wedged bronchoscope (Table 6-13). As previously described, collateral resistance increased,
and this increase was reproducible even when four 5-min exposures over a 3-h period were
performed. Thus, no immediate tolerance was observed. Furthermore, this response could be
blocked partially by administration of a cyclooxygenase inhibitor (indomethacin) and a
H,-receptor blocker (chlorpheniramine), whereas a thromboxane synthetase inhibitor was
ineffective. This study suggests that histamine or cyclooxygenase products released from
resident cells directly or via the parasympathetic nervous system may mediate the increase in
collateral resistance. However, because collateral resistance probably makes up only a small
proportion of pulmonary resistance, these results may not be generalizable to more prolonged
exposures and to larger airway responses.
Airway Reactivity
Baboons were exposed via an endotracheal tube to 0.5 ppm O3 for 5 min after a
baseline methacholine inhalation challenge test (Fouke et al., 1988) (Table 6-14). Lung
resistance increased with O3 exposure, and the baboons showed an enhanced response to
methacholine. This enhanced methacholine response was due almost exclusively to the
post-O3 increase in RL and, thus, resulted in no change in the provocative dose that increased
RL by 50%. The experiment was repeated 5 to 14 days later, except that before O3 exposure,
cromolyn sodium was administered. In the presence of cromolyn, baseline RL after
O3 exposure was less (not significant), but the response to methacholine challenge was
significantly lower. In a follow-up study (Fouke et al., 1990) using a similar O3 exposure
protocol (no methacholine challenge), cromolyn partially blocked the O3-induced increase in
RL; however, post-O3 exposure analysis of BAL indicated that cromolyn did not affect the level
of several measured prostanoids (6-keto PGFin, PGEj, TXB2, or PGF2D), suggesting that these
mediators were not related to the change in RL.
6-90
-------
Table 6-13. Effects of Ozone on Pulmonary Functiori
-------
Table 6-13 (cont'd). Effects of Ozone on Pulmonary Functiofl
^
Ozone
Concentration
Exposure
ppm Dg/m Duration
1.0 1,960 5min
Through
trachea! tube
1.0 1,960 Ih
1.0 1,960 Ih
1.0 1,960 2h
Through trachea!
tube
1.0 1,960 6h/day,
5 days/week for
12 weeks
1.0 1,960 3h
2.0 3,920
1.0 1,960 6h/dayfor
2.0 3,920 7 days (1 ppm) or
3 days (2 ppm)
1.0 1,960 24 h
Drugs
Pentobarbital
Ketamine
Xylazine
Ketamine
Xylazine
Ketamine
Xylazine
In vitro
Pentobarbital
Pentobarbital
Species, Sex
(Strain)
Age"
Dog, M
(Mongrel)
20.2 ± 0.8 kg
Guinea pig, M
(Hartley)
250-300 g
Guinea pig, M
(Hartley)
250-300 g
Dog
(Mongrel)
22-25 kg
Monkey, M
(Cynomolgus)
4.5 ± O.lkg
Rat, M
(S-D)
300-380 g
Rabbit, M
(Albino)
2.5kg
Rat, M
(Wistar)
6 weeks old
3 Observed Effect(s)
Ozone-induced increase in collateral resistance blocked by indomethacin and histamine
antagonist, not by thromboxane synthetase inhibitor.
Decreased TLC, VC, FRC, RV, and RL. Indomethacin and cromolyn blocked change in FRC
and RV at 2 and 24 h PE. DLCO increased, blocked by cromolyn.
RL increased at 2 but not 8 h. Lung volumes, DL^, and alveolar ventilation increased at 8 and
24 h PE.
Tachypnea with inspiratory time and expiratory time equally shortened. No increase in
ventilatory drive 1 and 24 h PE.
No RL, Cdyn, or forced expiratory flow changes associated with exposure.
RL increased and Cdyn decreased in rat isolated perfused lung preparation.
At 17 h PE, 1 ppm increased RL; 2 ppm trapped air, decreased Cdyn and forced expiratory
flows, and increased RL.
Decreased Cdyn, Cst, YmxsosTLo increased FRC, RV; no additional effect of elastase
pretreatment in O3 exposed rats.
Reference
Kleeberger et al. (1988)
Miller et al. (1988)
Miller et al. (1987)
Sasaki et al. (1987)
Biagini et al. (1986)
Pino et al. (1992a)
Yokoyama et al.
(1989a)
Yokoyama et al. (1987)
'See Appendix A for abbreviations and acronyms.
''Age or body weight at start of exposure.
-------
Table 6-14. Effects of Ozone on Airway Reactivitf
u>
Ozone
Concentration1'
ppm
0.5
1.0
1.0
1.0
1.0
4.0
0.15
0.5
0.8
1.0
3.0
1.0
3.0
Dg/m
980
1,960
1,960
1,960
1,960
7,840
294
980
1,568
1,960
5,880
1,960
5,880
Exposure
Duratidn
2-7 h
Intermittent
8% CO,
4h
2h
8h
31 h
2h
4 h/day,
5 days/week for
18 weeks
2h
2h
0.5-2 h
2h
0.25 h
Challenge0
Agent
Ach
Mch
Mch
Ach
Mch
Ach
Ach
Hist
Ach
Sulfuric acid
Mch
Hist
Ach
Ach
Route
IV
INH
IV
INH
IV
IV
IV
INH
INH
IV
INH
INH
INH
IV
IV
Drugs
Urethane
Urethane
Urethane
Urethane
Methohexital
Pancuronium
Urethane
Propranolol
Urethane
Propranolol
Species, Sex
(Strain)
Aged
Rat, M
(F344)
70 days old
Rat
(S-D)
Rat, M
(F344)
70 days old
Rat, M
(F344,
S-D,
Wistar)
Rat, M
Rat, F
(Long-Evans)
190-210 g
Guinea pig, M
(Hartley)
4-5 weeks old
Guinea pig, M
(Hartley)
Guinea pig, F, M
(Hartley)
2 mo old
Guinea pig, F
(Hartley)
300-400 g
Guinea pig, M
(Hartley)
Guinea pig, M
(English)
550-700 g
Observed Effect(s)
No increase in airway reactivity despite use of CO, during
O3 exposure.
Reactivity in INH- but not IV-challenged rats.
Reactivity in INH- but not IV-challenged rats.
No reactivity in any of three strains.
Increased reactivity.
Reactivity increased immediately, but not 24 h PE, but not
accompanied by PMN influx or vascular protein leakage.
Increase in sensitivity, but no change in reactivity.
Increased reactivity before PMN influx.
No increased reactivity to acid; O3 alone increased gas
trapping.
Increased reactivity with 90 min at 1 ppm and with 30 min at
3 ppm. At 2 h, 3 ppm reactivity occurred at 0 and 5 but not
24 h PE.
Reactivity in INH Hist, but not IV Ach.
Reactivity blocked by lipoxygenase inhibitors but not by
indomethacin.
Reference
Tepper et al. (1995)
Uchida et al. (1992)
Tepper et al. (1995)
Tepper et al. (1995)
Tepper et al. (1995)
Evans et al. (1988)
Kagawa et al. (1989)
Tepper et al. (1990)
Silbaugh and Mauderly
(1986)
Nishikawa et al. (1990)
Tepper et al. (1990)
Lee and Murlas (1985)
-------
Table 6-14 (cont'd). Effects of Ozone on Airway Reactivitf
<*D
Ozone
Concentration1'
ppm
3.0
3.0
3.0
3.0
3.0
0.5
1.0
3.0
3.0
3.0
1.0
Dg/m
5,880
5,880
5,880
5,880
5,880
980
1,960
5,880
5,880
5,880
1,960
Exposure
Duration
0.25-2 h
2h
2h
2h
2h
2h
Trach Tube
5 min
Bronchoscope
2h
2h
Ih
2h
Trach tube
Challenge0
Age"nt
Mch
Hist
5-HT
SP
Mch
Ach
Ach
Ach
Mch
Ascaris suum
Ach
Ach
FS
Ach
Ach
Route
IV
INK
INK
IV
INK
INK
INK
INK
INK
INK
INK
INK
Drugs
Pentobarbitone
Ketamine
Xylazine
Propranolol
Chloralose
Pentobarbital
Succinylcholine
Thiopental
Chloralose
Thiopental
Chloralose
Pentothal
Chloralose
Pentobarbital
Species, Sex
(Strain)
Aged
Guinea pig, M
(Dunkin
Hartley)
450-550 g
Guinea pig, M
(Dunkin
Hartley)
494 + 39 g
Guinea pig, M
(Hartley)
550-750 g
Guinea pig, M
(Hartley)
600-700 g
Guinea pig, M
(Hartley)
600-750 g
Dog (Mongrel)
15 ± 0.9 kg
Dog, M
(Mongrel)
21.2 + 0.5kg
Dog
(Mongrel)
18-23 kg
Dog
(Mongrel)
16-25 kg
Dog
(Mongrel)
24 ± 9 kg
Dog
(Mongrel)
15-17 kg
Observed Effect(s)
Mch, Hist, 5-HT, and SP caused increase in reactivity INH,
but not IV. Ascorbic acid blocked reactivity to Hist and SP.
NEP inhibitors increased control response without changing
O3 response (i.e., O3 inhibits NEP). Atropine and vagotomy did
not or only partially reduced reactivity. Neither cyclooxygenase
or 5-lipoxygenase inhibitors affected reactivity.
Platelet activating factor antagonist did not inhibit reactivity or
eosinophils.
Reactivity occurs with leukocyte depletion using
cyclophosphamide.
With increased reactivity, increased lysosomal hydrolase observed
in BAL.
In vitro reactivity to SP and Ach, but not KC1.
Phosphoramidon blocked SP effect, but not Ach reactivity; no
increased reactivity was observed when mucosa was removed.
Increased reactivity, no change in BAL prostanoids.
Increased resistance to flow through the collateral system after
antigen challenge attenuated with O3 exposure 1-3 h and 24 h PE;
effect independent of PMNs.
Cyclo- and lipoxygenase inhibitor BW755C blocked reactivity.
In vivo but no in vitro reactivity; in vitro, trachea showed
reactivity to FS.
Reactivity blocked with Ambroxyl, but PMN increased;
ambroxyl inhibits arachidonic acid products from PMN.
Collateral resistance of small airways showed persistent reactivity
15-h PE; no effect on PMN, monocyte, or mast cell numbers at
PE time.
Reference
Yeadon et al. (1992)
Tan and Bethel (1992)
Murlas and Roum (1985b)
Lew et al. (1990)
Murlas et al. (1990)
Fouke et al. (1991)
Kleeberger et al. (1989)
Fabbri et al. (1985)
Walters et al. (1986)
Chitano et al. (1989)
Beckett et al. (1988)
-------
Table 6-14 (cont'd). Effects of Ozone on Airway Reactivitf
Ozone
Concentration'"
ppm
3.0
3.0
3.0
3.0
3.0
3.0
3.0
OD 3-°
Ln
3.0
0.5
0.6
0.5
0.6
1.0
1.0
Qg/m
5,880
5,880
5,880
5,880
5,880
5,880
5,880
5,880
5,880
980
1,176
980
1,176
1,960
1,960
Exposure
Duration
0.5 h
Trach tube
0.5 h
Trach tube
0.5 h
Trach tube
0.5 h
Trach tube
0.5 h
Trach tube
0.5 h
Trach tube
0.5 h
Trach tube
20min
Trach tube
0.5-2 h
Trach tube
5 min
5 min
6 h/day,
5 days/week for
12 weeks
2h,
I/week for
19 weeks
Challenge0
Ajfent
FS
Cch
Ach
Ach
Hist
5-HT
Ach
Hist
Cch
TX
Ach
FS
KC1
Ach
Ach
Ach
Mch
Mch
Ach/Pt
Mch
Route
INK
INK
INK
INK
INK
INK
INK
INK
INK
INH
INK
INH
Drugs
Pentobarbital
Pentobarbital
Pentobarbital
Pentobarbital
Pentobarbital
Pentobarbital
Pentobarbital
Pentobarbital
Thiopental
Chloralose
Ketamine
Fluorodiazepam
Ketamine
Fluorodiazepam
Ketamine
Xylazine
Pentobarbital
Species, Sex
(Strain)
Aged
Dog
(Mongrel)
Dog
(Mongrel)
Dog
(Mongrel)
18-30 kg
Dog
(Mongrel)
Dog
(Mongrel)
Dog
(Mongrel)
18-30 kg
Dog
(Mongrel)
21-27 kg
Dog
(Mongrel)
18-32 kg
Dog
(Mongrel)
26-32 kg
Baboon, M
25-40 kg
Baboon, M
25-40 kg
Cynomolgus,
M
4.5 ± 0.1 kg
Rhesus
Macaco, F
5-7 kg
Observed Effect(s)
In vitro increases in reactivity suggested pre- and postjunctional
inhibition (PGE,) and postjunctional excitation (TXA,).
Reactivity was not blocked by TX antagonists.
In vitro reactivity not altered by epithelial removal, indicating
O3 did not effect epithelial-derived relaxing factor.
Ganglionic blocker hexamethonium did not alter reactivity.
Airways not responsive to TX mimetic (U46619) but were to
carbachol.
In vitro hyperresponsiveness of airway smooth muscle was
observed with FS and Ach, but not KC1. FS not associated with
increased excitatory junction potentials.
CDllb/CD18 monoclonal antibody prevented PMN influx, but not
reactivity.
Allopurinol and desferoxamine inhibited reactivity without
inhibiting PMN influx.
Thromboxane synthase inhibitor blocked reactivity without
influencing PMN influx.
Brief exposure caused increased reactivity that was blocked by
cromolyn.
Reactivity partially blocked by cromolyn, but no effect on stable
prostanoids.
No increased reactivity in O3-only group, but increase with
platinum mixture.
5-lipoxygenase inhibitor blocked the development of reactivity.
Reference
Janssen et al. (1991)
Jones et al. (1990)
Jones et al. (1988b)
Jones et al. (1987)
Jones et al. (1992)
Jones et al. (1988a)
Li et al. (1992)
Matsui et al. (1991)
Aizawaetal. (1985)
Fouke et al. (1988)
Fouke et al. (1990)
Biagini et al. (1986)
Johnson et al. (1988)
"See Appendix A for abbreviations and acronyms.
'Table ordered according to animal species.
°Mch = 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.
dAge or body weight at start of exposure.
-------
The response to antigen-induced bronchoconstriction, an animal model of allergy,
also has been evaluated recently. After a 5-min exposure to 1.0 ppm O3 via a wedged
bronchoscope, collateral resistance in dogs increased for 1 to 3 h (Kleeberger et al., 1989).
After the O3-induced resistance returned to baseline, the typical increase in collateral
resistance observed in dogs challenged with Ascaris suum antigen, to which the dogs were
natively sensitive, was attenuated both 1 to 3 h and 24 h post-O3 exposure. The attenuated
antigen response appeared to be independent of PMNs in the airways. In a follow-up study,
the late-phase response to antigen (bronchoconstriction 2 to 12 h postantigen challenge) also
was blocked in allergic dogs when O3 exposure (1.0 ppm, 5 min, via a bronchoscope) preceded
antigen challenge (Turner et al., 1989). These studies suggest that, at least in the dog, brief
local administration of O3 to the airways may inhibit allergic responses.
6.2.5.3 Acute Ozone Exposures (Less Than One Day)
Ventilation
Alteration of the ventilatory pattern has long been established as a hallmark of acute
O3 exposure. Several animal studies evaluated tidal breathing changes during and after
O3 exposure (U.S. Environmental Protection Agency, 1986). For most species, a tachypneic
response (rapid and shallow breathing) has been observed. For example, Murphy et al. (1964)
studied unanesthetized guinea pigs exposed for 2-h to 0.34, 0.68, 1.08, or 1.34 ppm O3 via
nose cones, and measured tidal breathing using a constant volume plethysmograph. A similar
experimental preparation was used by Amdur et al. (1978) to evaluate the respiratory response
of guinea pigs to 0.2, 0.4, and 0.8 ppm O3. In both experiments, a monotonic increase in f
was observed. In the Amdur et al. (1978) study, decreases in VT were not observed
concomitantly.
Lee et al. (1979, 1980) showed that the tachypneic pattern observed in conscious
dogs exposed to 0.56 to 0.85 ppm was not altered by bronchodilator pretreatment or atropine
administration. These manipulations would suggest that the rapid, shallow breathing was not
caused by bronchoconstriction. The response, however, was blocked by vagal cooling, which
was interpreted by the authors to suggest that vagal sensory afferent transmission had been
blocked. Thus, the authors suggested that increased vagal afferent impulses produced
tachypnea and that the response was independent of vagal efferents (increased smooth muscle
tone).
Several new studies evaluating ventilation after acute O3 exposure have appeared in
the literature (Table 6-13). Mautz and Bufalino (1989) measured ventilation (V E) as well as
oxygen consumption and rectal temperature in awake rats exposed for 3 h to 0.2, 0.4, 0.6, and
0.8 ppm O3. Concentration-related increases in f were significantly different from controls
beginning at 0.4 ppm, with a maximal response observed up to 0.6 ppm. Tidal volume was
similarly reduced, whereas V E and rectal temperature were less sensitive to O3 exposure,
showing decreases at 0.6 and 0.8 ppm. Oxygen consumption was decreased at all
concentrations tested. The authors concluded that the O3-induced change in breathing pattern
did not cause a decrease in metabolic rate or impose a condition of hypoxia. The changes in
ventilation and O2 consumption appeared coincident or possibly preceded the irritant reflex
change in breathing pattern.
Tepper et al. (1990) exposed awake rats to 0.12, 0.25, 0.5, and 1.0 ppm O3 for
2.25 h in head-out pressure plethysmographs. During exposure, CO2-stimulated breathing was
incorporated to augment ventilation, similar to the use of exercise in human studies.
6-96
-------
Frequency increased and VT decreased monotonically between 0.25 and 1.0 ppm during a
2.25-h exposure. No decrease in V E was observed. This difference from the Mautz and
Bufalino (1989) study could be due to their restraining rats in a tightly fitting plastic flow
plethysmograph with the face sealed by an aluminum nose cone; in the Tepper et al. (1990)
study, the rats were exposed in oversized, steel, head-out plethysmographs and were
intermittently challenged with CO2, which may have overridden the metabolic depressant
effect.
Mautz et al. (1985b), using exercising dogs exposed to 0.6 ppm O3 for 140 min,
showed tachypnea, increased V E, and elevated ventilation equivalents for O2 and CO2
compared to dogs exposed to air while exercising. Pulmonary resistance fell in air-exposed,
exercising dogs, but climbed toward the end of the exposure in the O3-exposed dogs.
In a follow-up study to Lee et al. (1979, 1980), Sasaki et al. (1987) performed
similar experiments on two awake dogs that were trained to run on a treadmill. Dogs were
exposed to 1.0 ppm O3 for 2 h and evaluated before O3 exposure and at either 1 or 24 h
postexposure. In all studies with or without exercise, O3 increased f and decreased VT, without
affecting V E. Vagal blockade diminished, but did not abolish, the tachypneic response,
indicating that both vagal and nonvagal mechanisms were important. The O3-induced change
in f was due to equal reductions in inspiratory and expiratory times with no, or small,
diminution (only during CO2 rebreathing experiments) of ventilatory drive (VT/inspiratory
time). Additionally, O3 did not affect functional residual capacity (FRC) or core temperature
in resting, exercising, or vagally blockaded dogs. The authors speculate that the change in f is
due to a vagally mediated lowering of the volume threshold of the pulmonary stretch receptor
for inspiration and expiration with a concomitant increase in flow rate, thus leaving the FRC
constant. Furthermore, the authors speculated that increased sensitization of rapidly adapting
receptors, C-fiber nerve endings, and nonvagal mechanisms also may be contributors to the
tachypneic response.
Two recent studies provide further insight into the mechanism of O3-induced
changes in ventilatory patterns. In the first study, Schelegle et al. (1993) showed that
O3-induced (3 ppm O3 for 40 to 70 min) tachypnea in anesthetized, spontaneously breathing
dogs largely could be abolished by cooling the cervical vagus to 0 but not 7 DC. This would
indicate that large myelinated fibers were not involved in this reflex response, but
nonmyelinated C fibers, whose activity is decreased only at the lower temperature, are
important. In a companion study, Coleridge et al. (1993) measured the responses of five types
of single vagal nerve fibers: (1) bronchial C-fibers, (2) pulmonary C-fibers, (3) rapidly
adapting receptors, (4) slowly adapting pulmonary stretch receptors, and (5) unclassified
fibers. During exposure to O3, bronchial C-fibers were most affected. Because discharge of
these fibers was not immediate with the onset of exposure, but took time to develop, the
authors suggested that O3 may not directly stimulate these receptors and that autacoid
mediators released in the lung, which previously have been shown to stimulate these fibers,
were probably responsible for fiber activation. Rapidly adapting receptors were shown to play
a small part in this reflex response; although, surprisingly, pulmonary C-fibers and slowly
adapting receptors were found to be unimportant. Ozone also stimulated several unidentified
vagal fibers that may be responsible for residual effects not abolished by 0 DC cooling of the
vagus. Thus, early inflammatory changes, as well as direct stimulation of bronchial C-fibers,
may be responsible for the tachypnea seen with O3 exposure in animal experiments. Results
from these studies, together with the amelioration of spirometric changes found after
6-97
-------
indomethacin in humans (see Chapter 7), suggest that tachypnea, inspiratory pain, and the
reduction in forced vital capacity (FVC) could reflect early inflammatory lesions as well as
neurogenic stimulation.
Breathing Mechanics
Although changes in breathing mechanics have been observed in laboratory animals,
these changes are not observed consistently and tend to be reported more frequently at higher
exposure concentrations (U.S. Environmental Protection Agency, 1986; Table 6-13).
The previously discussed studies by Murphy et al. (1964) and Amdur et al. (1978)
evaluated breathing mechanics in unanesthetized guinea pigs. The Murphy et al. (1964) study
showed an increase in flow resistance only at concentrations > 1 ppm O3. Pulmonary
compliance was not measured. Amdur et al. (1978) observed a decrease in Cdyn after
exposure to 0.4 and 0.8 ppm O3, but no significant change in RL was noted.
In an attempt to expose unanesthetized rats using regimens analogous to human
clinical studies, Tepper et al. (1990) observed no significant changes in RL or Cdyn after a
2.25-h exposure to 0.12, 0.25, 0.5, or 1.0 ppm O3, in spite of intermittent 15-min periods of
exercise-like hyperventilation induced by CO2. Similarly, no changes in breathing mechanics
were observed by Yokoyama et al. (1987) when they evaluated anesthetized rats exposed to
1.0 ppm O3 for 24 h. However, when Cdyn was normalized for differences in FRC, the
resulting specific compliance was decreased compared to air-exposed controls. Furthermore,
when the animals were paralyzed and ventilated between 40 and 200 breaths/min, O3-treated
animals showed a frequency-dependent decrease in Cdyn as f increased above 120 breaths/min.
The authors conclude that because RL was not affected, the effect of O3 was to obstruct
peripheral airways.
Pulmonary mechanics were evaluated in anesthetized, paralyzed dogs acutely
exposed to 0.12, 0.22, and 0.45 ppm O3 for 3 h via a stainless steel tracheal tube. No changes
in RL or Cdyn were observed at any concentration (Morgan et al., 1986).
In papers by Miller et al. (1987, 1988), the effect of a 1-h exposure to 1.0 ppm
O3 was evaluated. Two hours after exposure, anesthetized, tracheostomized guinea pigs
showed a significant increase in RL that resolved by 8 h postexposure. Both indomethacin and
cromolyn sodium partially blocked the increase in RL at 2 h postexposure (Miller et al., 1988).
These results suggest that eicosanoids produced from an inflammatory response in the lung
may be responsible for the increase in RL. However, plasma levels of PGF2n and 6-keto PGFn
were not affected by O3 or drug treatment, and PGE, was not affected by O3. In an attempt to
understand the involvement of eicosanoids in the increase in RL observed with O3 exposure,
Fouke et al. (1991) showed that exposure to 0.5 ppm O3 for 2 h caused an increase in RL and a
decrease in Cdyn in anesthetized dogs. Bronchoalveolar lavage fluid from these dogs did not
have any increase in 6-keto PGF1D, PGEj, TXBj, or PGF2D, suggesting that these
cyclooxygenase products were not involved in the changes in breathing mechanics. Similar
findings were observed by these authors after brief exposures to baboons (see Section 6.2.5.2).
Gas trapping in the excised guinea pig lung was evaluated by water displacement in
guinea pigs challenged with acid aerosol exposure after a 2 h, 0.8-ppm O3 exposure (Silbaugh
and Mauderly, 1986). Ozone exposure followed by air exposure increased gas trapping to
roughly the same extent as O3 followed by a sulfuric acid challenge (1 h, 12 mg/m3) when
compared to the air-only control response. These data indicate that O3 causes an acute
6-98
-------
peripheral airway obstruction, but no additive or synergistic effect of sulfuric acid aerosol was
observed.
Airway Reactivity
Probably, the most extensive amount of laboratory animal research has been
conducted on the role of O3 in producing acute airway injury resulting in an increase in airway
reactivity (U.S. Environmental Protection Agency, 1986; Table 6-14). Much of this research
has used O3 exposures that are never encountered in the ambient environment (D3 ppm for
D30 min); thus, the relevance of these studies may be questioned. However, the studies are the
most thorough mechanistic account of such O3 effects and have shown some agreement with
human O3 exposure studies (Chapter 7); therefore, these studies are summarized briefly here.
The literature is focused around five primary issues that in recent years have been more
thoroughly evaluated.
Concentration and Peak Response Time. Easton and Murphy (1967) were the
first to demonstrate an increased responsiveness in unanesthetized guinea pigs post-O3 exposure
(2 h, 0.5 to 7 ppm). In their study, responsiveness was assessed by increased mortality due to
severe histamine-induced bronchoconstriction, as well as by increased RL and decreased Cdyn.
Lee et al. (1977) examined anesthetized dogs exposed to O3 (0.7 to 1.2 ppm, 2 h) via a tracheal
tube and determined that increased airway reactivity to inhaled histamine occurred at 24 but
not 1 h postexposure. A similar experiment, done in unanesthetized sheep by Abraham et al.
(1980), indicated that airway responsiveness was increased at 24 h, but not immediately after a
2-h exposure to 0.5 ppm O3. When the exposure was increased to 1 ppm O3, an increase in
baseline RL was reported, and reactivity increased immediately and at 24 h postexposure. In
apparent contradiction, Holtzman et al. (1983a) showed that airway reactivity increased
markedly 1 h after dogs were exposed to 2.2 ppm O3 for 2 h and was less evident at 24 h
postexposure. Gordon and Amdur (1980) also reported that airway reactivity in guinea pigs
was maximal 2 h after a 1-h exposure to 0.1, 0.2, 0.4, or 0.8 ppm O3, as defined by a
significant increase in RL or decrease in Cdyn after a single subcutaneous challenge of
histamine. The effect on RL was concentration dependent, but was significant only at 0.8 ppm.
For Cdyn, there was no concentration-related response, but all O3 exposures exacerbated the
decrease in Cdyn after histamine relative to the air-exposed group. The site of
bronchoconstriction was suggested to be the conducting airways, rather than the parenchyma,
because dynamic compliance was affected and static compliance was not (Gordon et al., 1984).
To examine the role of duration of exposure on experimental outcome, Nishikawa
et al. (1990) exposed guinea pigs to C x T products of 30 (1 ppm x 30 min), 90 (1 ppm x
90 min), 90 (3 ppm X 30 min), and 360 (3 ppm X 120 min) ppm • min O3. After exposure,
specific airway resistance (SRaw) during an inhaled methacholine challenge was measured in
unanesthetized animals at 5 min, 5 h, and 24 h. In all but the 1-ppm, 90-min exposure group,
there was an increase in baseline SRaw at 5 min, but the response was neither concentration nor
C x T dependent. At 5 min postexposure, no increase in airway responsiveness was observed
at 30 ppm • min. Airway hyperresponsiveness was observed at 90 ppm • min, using either
exposure scenario (1 ppm for 90 min or 3 ppm for 30 min), and the response to the 360 ppm •
min was greater than that observed with 90 ppm • min exposure. Significant increases in
airway responsiveness at both 5 and 24 h postexposure were observed only in the 360 ppm •
6-99
-------
min group. The authors concluded that exposure duration was an important determinant of
O3-induced airway hyperresponsiveness.
Uchida et al. (1992) reported increased airway reactivity in rats to inhaled
methacholine after a 1.0-ppm (2-h) O3 exposure. These results conflict with other published
studies in rats also using inhaled methacholine, which reported the inability to produce
consistent increases in airway reactivity after exposure to less than 4 ppm O3 (Evans et al.,
1988). Tepper et al. (1995) reported that airway hyperresponsiveness in rats challenged with
iv acetylcholine occurred only at 1 ppm O3 or higher. In these latter studies, exposure
durations ranged from 2 to 7 h, and, in some tests, CO2 was added to the exposure to increase
ventilation. Although guinea pigs are more responsive than rats, they are not as responsive as
humans to O3-induced increased airway reactivity, even under optimal conditions (Tepper
etal., 1995).
Inhaled Versus Intravenous Challenge. In a follow-up study using a similar
exposure protocol as described above, Abraham et al. (1984) observed increased
responsiveness to iv carbachol in unanesthetized sheep 24 h after a 2-h, 0.5-ppm O3 exposure;
inhaled carbachol did not produce a similar response. The authors interpreted this result to
indicate a decreased penetration of the carbachol aerosol in O3-exposed animals compared with
the direct stimulation of smooth muscle by the iv route. Roum and Murlas (1984) observed
that O3-induced hyperresponsiveness was similar for inhaled versus iv acetylcholine or
methacholine challenge through 14 h postexposure, but after that time, only iv administration
revealed a persistent O3-related response. In contrast, Yeadon et al. (1992) reported that
guinea pigs exposed to 3 ppm for 30 min were hyperresponsive to inhaled histamine,
serotonin, acetylcholine, and substance P, but were not hyperresponsive to O3 after iv
administration of the same agonists. Tepper et al. (1995) and Uchida et al. (1992) also showed
that rats were more sensitive to inhaled methacholine than to iv administration of the agonist.
Neurogenic Mediation. Lee et al. (1977) reported increased airway
responsiveness to histamine in dogs exposed to 0.7 or 1.2 ppm O3 for 2 h. Atropine and vagal
blockade were effective in reducing the O3-induced hyperresponsiveness to histamine,
suggesting that heightened vagal activity was responsible. Katsumata et al. (1990) also showed
that in the cat, airway hyperresponsiveness to histamine could be attributed to cholinergic
reflex. This is in apparent contrast to the increased O3-induced (1.0 to 1.2 ppm, 2 h)
responsiveness to histamine (subcutaneous) that was not blocked by atropine or vagotomy,
indicating minimal vagal involvement in guinea pigs (Gordon et al., 1984). In agreement,
Jones et al. (1987) found that hexamethonium, a ganglionic blocker, did not prevent
O3-induced hyperresponsiveness in dogs exposed to 3 ppm O3 for 0.5 h via an endotracheal
tube. Similarly, Yeadon et al. (1992) showed that atropine or bilateral vagotomy only partially
reduced the hyperresponsivenss in guinea pigs exposed to 3 ppm for 120 min but did not block
the response in animals exposed for only 30 min.
A role for prejunctional muscarinic receptors has been demonstrated by Schulteis
et al. (1994). The M2 receptor, which is inhibitory for acetylcholine release, was shown to be
defective immediately after a 4-h, 2-ppm O3 exposure in guinea pigs. Fourteen days after
exposure, M2 receptor function and vagally stimulated responsiveness were normal. Thus, the
role of the cholinergic system in O3-induced airway hyperresponsiveness has yet to be firmly
established.
6-100
-------
Peptidergic mediators also have recently been suggested as important modulators of
this response. Murlas et al. (1992) demonstrated that phosphoramidon, an inhibitor of neutral
endopeptidase (NEP), increased the responsiveness to substance P in air-exposed (but not
O3-exposed [3 ppm, 2 h] guinea pigs). Substance P-induced bronchoconstriction in air-exposed
animals was increased after phosphoramidon because NEP degrades substance P. The finding
was associated with a decrease in tracheal NEP in O3-exposed animals. Additionally, the
increased airway responsiveness in O3-exposed animals was reversed by inhalation of partially
purified NEP. Taken together, these results suggest that O3 inactivates NEP, thus increasing
the response to endogenous tachykinin release. A similar result was obtained by Yeadon et al.
(1992) in guinea pigs exposed to 3 ppm O3 (30 or 120 min) and challenged with aerosolized
substance P after pretreatment with the NEP inhibitors phosphoramidon, thiorphan, and
bestatin. Tepper et al. (1995) depleted guinea pigs of substance P, using multiple doses of
capsaicin, and found that airway reactivity, after a 2-h exposure to 1 ppm O3, was partially
blocked. However, although tracheal vascular permeability also was blocked by capsaicin
pretreatment, protein influx into the BAL and tachypnea were not blocked. On the other hand,
Evans et al. (1989) did not find increased tracheal vascular permeability in rats exposed to 4
ppm O3 for 2 h. These studies suggest, at least for the guinea pig, that enhancement of the
substance P response, by inhibition of NEP, may be important in O3-induced
hyper re sponsiveness.
Inflammation. Holtzman et al. (1983b) found a strong association between
increased airway responsiveness and increased PMNs present in the tracheal biopsy of dogs 1 h
after a 2-h O3 exposure to 2.1 ppm. Fabbri et al. (1984) extended these findings, showing an
association between increased airway reactivity and increased lavageable inflammatory cells
from the distal airways of dogs. Further support for this hypothesis was engendered by the
demonstration that in PMN-depleted dogs (produced by administration of hydroxyurea),
O3-induced airway hyperresponsiveness was blocked. This is in contrast to Murlas and
Roum's (1985a) findings in guinea pigs exposed to 3 ppm O3 for 2 h, which indicate that
increased airway reactivity, mucosal injury, and mast cell infiltration occur before PMN
influx. The authors speculate that PMN influx is a response to the damage, not a cause of the
increased airway reactivity. Furthermore, Murlas and Roum (1985b) showed that PMN
depletion in the guinea pig with cyclophosphamide did not prevent O3-induced airway
hyperresponsiveness. Similar results were obtained by Evans et al. (1988), who reported that
airway hyperresponsiveness was not accompanied by airway PMN influx in rats, and by load
et al. (1993), who showed that adding human PMNs to the pulmonary circulation of the rat
lung during a 3-h, 1.0-ppm exposure to O3 did not further enhance O3-induced airway
reactivity. Beckett et al. (1988) evaluated dogs exposed for 2 h to 1 ppm O3 directly to the
peripheral airways via a wedged bronchoscope. Fifteen hours postexposure, the exposed
peripheral airway segments were hyperresponsive to aerosolized acetylcholine. However, at
the site of increased responsiveness, there was no association with increased PMNs, mast cells,
or mononuclear cells. Such studies agree with perhaps the most definitive study of this
hypothesis (Li et al., 1992), which used monoclonal antibodies (CDllb/CD18) to prevent
PMN influx into the airways. When PMNs were present in the circulation, but prevented from
entering the lung, the dogs were still hyperresponsive after a 30-min exposure to 3 ppm O3.
Thus, in three species, it appears that PMN influx may be associated with O3 exposure but is
not necessary for producing airway hyperresponsiveness.
6-101
-------
Several studies have suggested that arachidonic acid metabolites may be important
in O3-induced airway hyperresponsiveness. Although the primary source of arachidonic acid
metabolites is suspected to be inflammatory cells in the lung, cells other than PMNs could be
responsible for the liberation of arachidonic acid metabolites. Only one study in dogs indicates
that blockage of cyclooxygenase products with indomethacin can protect animals from
developing airway hyperresponsiveness (O'Byrne et al., 1984). However, several, more
recent studies have found that cyclooxygenase inhibitors were ineffective in blocking this
response (Lee and Murlas, 1985; Holroyde and Norris, 1988; Yeadon et al., 1992). Two
papers indicated the importance of LTs, as demonstrated by the inhibition of
hyperresponsiveness with prior administration of 5-lipoxygenase inhibitors to guinea pigs (Lee
and Murlas, 1985; Murlas and Lee, 1985). In contrast, Yeadon et al. (1992) found that a
specific 5-lipoxygenase inhibitor did not block the response in guinea pigs. One study with
dogs showed that TX generation may be important in this phenomenon (Aizawa et al., 1985),
but, more recently, two papers from the same investigators have dispelled that notion (Jones
et al., 1990, 1992). Furthermore, exposure to 0.5 ppm O3 for 2 h caused a decrease in the
provocative dose of methacholine necessary to cause a 50% increase in RL in anesthetized dogs
(Fouke et al., 1991). Bronchoalveolar lavage on these dogs did not show any increase in 6-
keto PGFin, PGE2, TXB2, or PGF2D, suggesting that these cyclooxygenase mediators of
inflammation were not involved in the changes in airway reactivity. In summary, the initial
hypothesis of the role of PMNs or PMN-derived products in O3-induced airway
hyperresponsiveness is questionable because most newer studies, using more specific inhibitors
of PMNs, cyclooxygenase, and 5-lipoxygenase, and studies blocking TX receptors indicate the
lack of a protective effect.
Interactions with Antigen and Virus. In mice, Osebold et al. (1980) showed that
an increased number of animals became sensitized to ovalbumin after 3 to 5 days of continuous
exposures to 0.5 and 0.8 ppm O3. Matsumura (1970) and Yanai et al. (1990) made similar
findings in guinea pigs and dogs exposed acutely to higher O3 concentrations, suggesting that
O3 may enhance either sensitization or response to antigen. These results appear to agree with
recent findings in humans (see Chapter 7).
Ozone (1 ppm, 2 h) also may increase hyperreactivity associated with virus
exposure. Tepper et al. (1995) exposed rats to O3 either before or during an influenza virus
infection. Rats exposed to O3 before virus infection were more hyperresponsive to inhaled
methacholine 3 days later (at a time when there was no hyperresponsiveness to O3 alone) than
were rats exposed to only the virus. An additive effect was observed in virus-infected rats
when O3 exposure was immediately before methacholine challenge.
Extended Functional Characterizations
Extended characterizations of pulmonary function in laboratory animals indicate that
the general pattern of functional impairment reported in human studies also is observed in
animal studies of acute O3 exposure. Anesthetized and ventilated cats showed a general decline
in vital capacity (VC), static lung compliance, or diffusing capacity for carbon monoxide
(DLCO) with exposures up to 6.5 h of 0.26 to 1.0 ppm O3 (Watanabe et al., 1973). Inoue et al.
(1979) observed functional evidence of premature airway closure, as indicated by increases in
closing capacity, residual volume (RV), and closing volume, after rabbits were exposed to 0.24
or 1.1 ppm O3 for 12 h. The volume-pressure curve indicated increased lung volume at low
6-102
-------
distending pressures; additionally, nonuniform distribution of ventilation was observed. The
effects were most prominent 1 day following exposure and had mostly subsided by 7 days
postexposure.
Most studies of O3 in experimental animals make little effort to mimic human study
designs, thereby further confounding the extrapolation of their results to humans. Recently,
however, rat studies involving periods of intermittent CO2-induced hyperventilation to enhance
the delivered dose of O3 have attempted to capitalize on the qualitative similarity of the rat and
human maximum expiratory flow-volume curves as a potentially sensitive endpoint of toxicity
(Costa et al., 1988b; Tepper et al., 1989). In the rat, FVC decreases acutely with
O3 exposure, and this response has been mathematically modeled for O3 concentrations
between 0.35 and 0.8 ppm with exposure durations between 2 and 7 h (Tepper et al., 1989).
The magnitude of response is apparently less than that observed in humans (Tepper et al.,
1995), although the extent to which anesthesia mitigates the rat response or that there are
inherent species differences in dosimetry or sensitivity is not clear from these studies (see
Chapter 8).
In addition to changes in the flow-volume curve, changes in lung diffusion also are
observed. In a study that examined concentration, duration, and ventilation factors, rats were
exposed for 2 or 7 h to 0.5 or 0.8 ppm O3 with intermittent 8% CO2 to hyperventilate (D2 to
3 times resting VE) the animals as an exercise analogue to human exposures (Costa et al.,
1988a). The DLCO values were reduced by 10% at both 0.5-ppm time points and by 12% with
a 2-h exposure to 0.8 ppm. Exposure to 0.8 ppm for 7 h, however, greatly exacerbated the
alveolar effect, with a resultant 40% reduction in the DLC0. Static compliance was affected
only at this latter exposure concentration and duration. This O3-induced reduction in DLCO
appeared to correlate with the degree of lung edema in affected animals. Yokoyama et al.
(1987) found decreases in rat lung volumes (FRC and RV), static compliance (from the
volume-pressure curve), and maximal flow at 50% of VC after a 24-h exposure to 1 ppm O3.
Flow-volume curves and measurements of regional distribution of ventilation, using
a positron camera, were evaluated in anesthetized, paralyzed dogs acutely exposed via a
stainless steel tracheal tube to 0.12, 0.22, and 0.45 ppm O3 for 3 h (Morgan et al., 1986).
No changes in the flow-volume curve were observed at any concentration, but a less uniform
distribution of ventilation was noticed, with the greatest difference occurring between the
central and more peripheral regions. The authors conclude that the initial effect of O3 appears
to be obstruction of the small airways.
Miller et al. (1987, 1988) evaluated the effect of a 1-h exposure to 1.0 ppm O3 on
changes in the lung function of anesthetized, tracheostomized guinea pigs. Decreases in lung
volumes were noted at 2 h postexposure and were maximum between 8 and 24 h postexposure,
after which time they began to resolve. Alveolar ventilation (VA) and DLCO also were
decreased by exposure. The initial (2 h postexposure) reduction in DLCO may have been
caused by a bronchoconstriction-related decrease in VA. After this time, disproportionate
ratios of DLCO and VA suggest that different mechanisms were responsible for the decreased
DLC0. The authors speculate that this latter response probably involves the development of a
peripheral inflammatory response (8 to 24 h postexposure) because plasma concentrations of 6-
keto PGFin and PGE, also were elevated in guinea pigs exposed for 1 h to 1 ppm O3 (Miller
et al., 1987). Significant increases in the plasma and BAL concentrations of TXB2 also were
observed following acute exposure of guinea pigs to 1 ppm O3 (Miller et al., 1987) and humans
to 0.4 or 0.6 ppm O3 (see Chapter 7). Both indomethacin and cromolyn sodium partially
6-103
-------
blocked the reduction in lung volumes at 2 and 24 h postexposure (Miller et al., 1988).
Indomethacin was ineffective in blocking the O3-induced decrease in DLCO at either 2 or 24 h
postexposure, but cromolyn sodium blocked this O3 response. Both drugs were effective in
blocking the O3-induced decrease in VA at the same examination periods. These results
suggest that eicosanoids produced from an inflammatory response in the lung may be
responsible for the observed changes in lung function in guinea pigs. However, as noted
above the role of eicosanoid mediators in O3-induced lung injury is controversial.
6.2.5.4 Repeated Acute Exposure Experiments (More Than Three Days)
To date, few physiological studies have examined the attenuation that occurs in
humans repeatedly (3 to 7 days) exposed to O3 (U.S. Environmental Protection Agency, 1986;
Table 6-13), despite the fact that this exposure scenario most closely mimics a high oxidant
pollution episode.
Ventilation
In the only laboratory animal study using a similar exposure protocol and an
experimental design analogous to human repeated-exposure studies, Tepper et al. (1989)
showed that rats displayed an initial pulmonary irritant response (tachypnea) that attenuated
after 5 consecutive days of exposure in a manner quite similar to the response pattern of
humans (see Section 7.2). Exposures were for 2.25 h and included challenge with CO2 during
alternate 15-min periods to augment ventilation (2 to 3 times VE—equivalent to light exercise in
humans). The functional changes were largest on Day 1 or 2, depending on the parameter and
the O3 concentration (0.35, 0.5, and 1.0 ppm were evaluated). Additionally, lung biochemical
and structural consequences were examined at 0.5 ppm O3 and indicated that several indices of
lung damage increased (histopathology) or did not adapt (lavageable protein), despite the loss
of the functional response over the 5-day exposure period. Functional attenuation, however,
did not occur in the 1.0-ppm O3 group; such a nonreversing effect has not been observed in
humans. It is likely that this lack of reversal was attributable to the high concentration of
O3 and, thus, may be predictive of the human response under similar conditions.
Breathing Mechanics
In rats exposed to 1.0 ppm O3 for 6 h/day for 7 days, the only change in breathing
mechanics was an increase in RL (Yokoyama et al., 1989a). Whether attenuation occurred
cannot be ascertained because functional measurements were obtained only after the end of
exposure.
Extended Characterizations
Selgrade et al. (1988) evaluated mice exposed to 1.0 ppm O3 for 5 days (3 h/day),
with and without the inoculation of influenza virus on Day 2 of exposure. Ozone alone did not
cause an untoward effect on lung volumes, volume-pressure, and flow-volume relationships
when mice were evaluated 1,4, and 9 days postexposure. Mice exposed to the combination of
virus and O3 showed a decrease in DLCO that persisted for 9 days.
A portion of the Tepper et al. (1989) study, discussed above, was conducted using
groups of animals that were exposed between 1 to 5 days to 0.5 ppm O3. Changes in the shape
of the flow-volume curve (as indicated by the change in forced expiratory flow [FEE] at
6-104
-------
25 % of VC) were maximal on Day 2 but gradually returned to baseline with further repeated
exposures.
6.2.5.5 Longer Term Exposure Studies
The question of degenerative or irreversible lung damage when O3 exposure is
extended over periods of days to years remains paramount to the assessment of health risk.
Several new studies since the previous criteria document (U.S. Environmental Protection
Agency, 1986) have been published using more integrated approaches (structure, function, and
biochemical techniques) for understanding this problem. This is especially true for studies
evaluating near-lifetime O3 exposures in rodent species.
Ventilation and Breathing Mechanics
Tepper et al. (1991) evaluated ventilation and breathing mechanics in rats exposed
for 1, 3, 13, 52, and 78 weeks to a simulated urban profile of O3 (Table 6-13). The exposure
consisted of a 5-day/week, 9-h "ramped spike" exposure that had an integrated average of
0.19 ppm O3 and a maximum concentration of 0.25 ppm. During other periods (13 h/day,
7 days/week), the exposure remained at a 0.06-ppm O3 background level. Pulmonary function
measurements were evaluated after 1, 3, 13, 52, and 78 weeks of O3 exposure in response to a
postexposure challenge with 0, 4, and 8% CO2. Overall, there was a significant increase in
expiratory resistance, but only at 78 weeks was resistance significantly different than the time-
matched filtered-air control. At all evaluation times, VT was reduced compared to control rats;
this was especially true during challenge with CO2. Frequency of breathing was significantly
decreased when the analysis included all evaluation times, but at no single evaluation time was
the reduction significant.
Other evidence of peripheral airflow abnormalities from extended exposures to O3 is
limited. Costa et al. (1983) exposed rats to 0.2 or 0.8 ppm O3 for 6 h/day, 5 days/week for 12
weeks and did not find a concentration-related increase in pulmonary resistance measured
immediately after exposure. Yokoyama et al. (1984) measured increased central resistance in
rats exposed for 30 days to 1.0 ppm, but found increased peripheral airway resistance when
exposure was for 60 days to 0.5 ppm. Pulmonary resistance was measured at different elastic
recoil pressures. Increased RL at low distending pressures was interpreted as of peripheral
origin, whereas uniform increases across all distending pressures were described as originating
from the central airways. These changes were consistent with morphological findings of
mucus in the large bronchi of rats exposed to 1.0 ppm compared to the rats exposed to 0.5
ppm. These data also agree with a study by Wegner (1982) that suggests the occurrence of
airflow obstruction, revealed in terms of small increases in peripheral airways resistance
(as measured by oscillation harmonics) that were observed in monkeys after 1 year of exposure
to 0.64 ppm O3 (8 h/day, 7 days/week).
Airway Reactivity
No studies of airway reactivity after long-term exposures were reported before
1985. Since then, several studies have reported no increase in reactivity with daily
O3 exposure (Table 6-14). Biagini et al. (1986) observed no changes in breathing mechanics,
FEF parameters, or methacholine and platinum airway responsiveness in a group of monkeys
(cynomolgus) exposed to 1 ppm O3 for 6 h/day, 5 days/week for 12 weeks in a study designed
to examine the effects of combining O3 exposure with the respiratory sensitizer platinum.
6-105
-------
Kagawa et al. (1989) exposed guinea pigs 4 h/day, 5 days/week for 4 mo to 0.15 ppm O3.
Baseline total respiratory resistance and response to increasing concentrations of inhaled
histamine were assessed every 3 weeks, but did not change in response to O3. The only
exception is the Johnson et al. (1988) study that evaluated airway responsiveness in female
rhesus monkeys just before a 2-h single weekly exposure to 1 ppm O3 delivered via an
endotracheal tube. After 19 weeks of exposure, increased responsiveness to inhaled
methacholine was observed compared to the animal's historic control. The
hyperresponsiveness persisted approximately 15 weeks after exposures were discontinued.
Hyperresponsiveness to O3 was reinstated after a similar 7-week exposure to the same animals.
After this exposure regimen, animals recovered in approximately 9 weeks, but
hyperresponsiveness was again reinstated with four, once-per-week exposures. The
investigators described the effect of a 5-lipoxygenase inhibitor on certain portions of this
sequence; however, the descriptions of methods and results were insufficient for evaluation of
the effect of treatment on exposure.
Extended Functional Characterizations
The previous criteria document (U.S. Environmental Protection Agency, 1986)
cataloged several investigators that reported marginal increases in total lung capacity (TLC) or
its component volumes in rats after intermittent or continuous exposures to DO.25 ppm O3 for
4 to 12 weeks (Bartlett et al., 1974; Costa et al., 1983; Raub et al., 1983). In contrast to these
significant results, Yokoyama and Ichikawa (1974) previously had reported no effects on rat
static volume-pressure curves after a 6-week exposure to 0.45 ppm (6 h/day, 6 days/week).
More recent studies are summarized in Table 6-13.
Exposures of rats to 0.7 ppm O3 for 28 days (20 h/day) showed an obstructive-type
lung function abnormality characterized by a significant reduction in FEFs, lung volumes, and
DLCO, and a significant increase in FRC (Gross and White, 1986). These effects largely
reversed after an additional 9 weeks of clean air, but some airflow abnormalities persisted.
Tyler et al. (1988) exposed young monkeys to 0.25 ppm O3 for 8 h/day,
7 days/week for 18 mo or for alternate months of the 18-mo period and observed increased
chest wall compliance (Cw) and inspiratory capacity. Because Cw did not decrease with age, as
expected, the authors speculated that perhaps O3 interfered with respiratory system maturation.
This effect was greater in monkeys exposed during alternate months than in animals exposed
every month of the 18-mo period.
To address the issue of cumulative exposure over a near-lifetime, several rodent
studies have been performed using various exposure concentrations. With exposure of rats to
0.5 ppm O3 for 52 weeks (20 h/day, 7 days/week), increases in RV and FRC were apparent, as
was a fall in DLCO (Gross and White, 1987), suggesting substantial end-airway damage and
gas-trapping. After a 3-mo period in clean filtered air, these measurements were not different
than similarly treated, but air-exposed control rats. In partial contrast, 12 or 18 mo of
exposure to a daily urban profile of O3 (9-h time-weighted average of 0.19 ppm, 5 days/week;
a background of 0.06 ppm for 13 h/day, 7 days/week) resulted in small reduction in lung
volumes (RV and VC) and an enhanced nitrogen (N2) washout pattern consistent with a stiffer,
restricted lung (Costa et al., 1995). Interestingly, in spite of mural remodeling of small
airways (which was concentration dependent), no evidence of airflow obstruction was apparent
in this study. However, in a cohort group of animals exposed at the same time, RL was
increased at all time points in unanesthetized animals, as described previously (Tepper et al.,
6-106
-------
1991). Harkema and Mauderly (1994) exposed rats for 6 h/day, 5 days/week for 20 mo to
either filtered air or 0.12, 0.5, or 1.0 ppm O3. Within 3 days of the end of exposure, an
extended functional evaluation was performed; O3 caused little impact on respiratory function.
However, RV was decreased (between 21 and 36%), a finding similar, but of greater
magnitude, to the Costa et al. (1995) study. The existing morphological data in monkeys at the
higher concentrations noted above appear consistent with end-airway remodeling, but no clear
functional evidence of obstruction has been described (Eustis et al., 1981; Wegner, 1982).
6.2.5.6 Summary
Alterations in the pulmonary function of laboratory animals after exposure to
O3 have been reported by numerous investigators. These changes appear to be homologous
with the changes in pulmonary function observed in humans exposed acutely to O3 (see
Chapter 7). Although there are apparent differences in sensitivity among species, it is not
clear whether these differences are due to the use of anesthesia or restraint or variances in
tissue sensitivity or dosimetry.
Brief exposures to O3 of less than 30 min have been shown to produce reflex
responses (increased collateral resistance) and airway hyperresponsiveness. In the dog, these
changes appear to be related, in part, to parasympathetic stimulation and release of
inflammatory mediators. However, the relevance of these studies must be questioned because
O3 was delivered to a specific lung region via a bronchoscope and the contribution of collateral
resistance to total lung resistance was small.
With exposures lasting greater than an hour, a wide variety of effects has been
observed. Most notably, tachypnea (increased frequency of breathing and decreased tidal
volume) has been noted in several species at exposures as low as 0.25 to 0.4 ppm (Tepper
et al., 1990; Mautz and Bufalino, 1989). In addition to changes in breathing pattern, changes
in breathing mechanics (compliance and resistance) and increased airway reactivity have been
observed, but, generally, these effects have been reported at concentrations of 1 ppm or
greater. In dogs, RL increased and Cdyn decreased after a 2-h exposure to 0.5 or 0.6 ppm
O3 (Fouke et al., 1991; Mautz et al., 1985b). Enhanced reactivity to bronchoconstrictors has
been reported in guinea pigs at 0.5 ppm O3 (Tepper et al., 1990). The mechanisms that may
be responsible for the O3-induced increase in airway reactivity have been investigated
extensively; however, no firm conclusion can be drawn. The most consistent evidence
suggests a role for sensory afferent fibers, their associated mediators (tachykinins), and the
enzyme responsible for tachykinin degradation (NEP). However, many studies suggest that
the parasympathetic nervous system and inflammatory cells and mediators also may play a role
in O3-induced increase in airway reactivity.
Extended characterizations of pulmonary function indicate that the general pattern
of functional impairment seen in humans acutely exposed to high concentrations of
O3 (decreased lung volumes, diffusional disturbances, and inhomogeneity of ventilation) also is
observed in animals exposed to high ambient O3 (0.5 to 2.0 ppm). For example, FVC, DLCO,
and N2 slope decreased with increasing C X T products (0.5 and 0.8 ppm O3, 2 and 7 h) in
rats (Costa etal., 1989).
With daily repeated exposure to O3, Tepper et al. (1989) showed attenuation of lung
function changes (tachypnea and flow volume curve) over 5 days (2 h/day to 0.35 to 1.0 ppm,
with CO2 stimulation of breathing) similar to what is observed in repeatedly exposed humans
(Chapter 7). It is of interest that, in this study, morphological changes showed a progressive
6-107
-------
increase in severity, and other biochemical indicators of lung injury (lavageable protein and
antioxidants) did not show attenuation of the response over the same time period. The findings
from long-term exposures of O3 to laboratory animals are even more difficult to summarize.
Various findings in rats ranged from no or minimal effects (Biagini et al., 1986; Kagawa et al.,
1989; Chang et al., 1992; Harkema and Mauderly, 1994) to obstructive (Gross and White,
1986; Tepper et al., 1991) or restrictive (Costa et al., 1995) lung function abnormalities.
However, in all cases where recovery was evaluated, no severe lung injury was detected, and
the physiological alterations that were observed resolved several months after termination of
exposure.
6.2.6 Genotoxicity and Carcinogenicity of Ozone
6.2.6.1 Introduction
Ozone is a very reactive molecule and a strong oxidizing agent that can dissolve in
aqueous solutions and generate superoxide, hydrogen peroxide, and hydroxyl radicals and can
oxidize and peroxidize cellular macromolecules (reviewed in Menzel, 1970; Hoigne and Bader,
1975; U.S. Environmental Protection Agency, 1986; Mustafa, 1990; Victorin, 1992; Pryor,
1993). Early studies of the effects of O3 on purines, pyrimidines, nucleosides, nucleotides,
and nucleic acids showed that O3 rapidly degraded these compounds in vitro (Christensen and
Giese, 1954; reviewed in Menzel, 1984). Ozone-generated hydroxyl radicals can abstract
hydrogen from organic molecules, leading to further complex free-radical reactions (reviewed
in Menzel, 1970, 1984; Mustafa, 1990; Victorin, 1992). In addition, O3 initiates radical
reactions, resulting in ozonolysis of alkenes to form ozonides, which decompose on reaction
with water to form peroxyl radicals, peroxides, and aldehydes. Ozone also can oxidize amines
to amine oxides and react with PUFA to form products of lipid peroxidation (reviewed in
Menzel, 1970, 1984, 1992; Mustafa, 1990; Pryor, 1978, 1991, 1993). Ozone also has been
shown to cause a reduction in plaque formation by bacteriophage f2, to release RNA from
phage particles, to inactivate RNA, and to degrade protein (Kim et al., 1980). Hence, because
O3 generates hydroxyl radicals in aqueous solution and degrades DNA, RNA, protein, and
fatty acids in vitro, it poses a potential genotoxic hazard by virtue of its ability to generate
reactive intermediates that can oxidize nucleic acid bases (reviewed in Victorin, 1992).
However, the precise reactions that occur in living cells exposed to O3 have not yet been
defined completely. As the ensuing discussion shows, the genotoxic potential of O3 is, at
most, weak.
This section reviews the information available on the genotoxicity of O3 since the
last air quality criteria document (U.S. Environmental Protection Agency, 1986) was
published, although earlier reports are cited to create a historical and scientific perspective for
the reader. The areas covered in this review are the ability of O3 to induce DNA damage,
mutagenesis, cell transformation, carcinogenesis, co-carcinogenesis, and tumor promotion.
Although modulation of the tumorigenic response by indirect effects of O3 on the immune
system is theoretically possible, no evidence for such modulation has been reported (see
Section 6.2.3). Unfortunately, experimental data to evaluate whether O3 is genotoxic are very
limited. Hence, relevant data on genotoxic effects of O3 above 1 ppm also have been included
to ensure discussion of the full array of effects as they currently are understood. Although data
at points far above 1 ppm of O3 are not directly relevant to human health, such high-
concentration data serve to address (1) whether ozone is genotoxic at all, (2) whether
6-108
-------
concentration-response relationships exist for the specific genotoxicity endpoint studied, and
(3) what maximum sensitivity is required for discovering genotoxic effects.
A further caveat is that in many experiments utilizing in vitro systems, O3 was
added to bacteria or to mammalian cells covered by cell culture medium. In all such
experiments, the reactivity of O3 makes it highly likely that the reaction products of O3 with
culture fluid, not O3 itself, actually reach and interact with cells. This complicates
interpretation of the results, making it extremely difficult to extrapolate in vitro results to in
vivo results and making it extremely difficult, if not impossible, to extrapolate in vitro results
to potential genotoxicity in humans.
6.2.6.2 Ozone-Induced Deoxyribonucleic Acid Damage
Studies utilizing a wide range of O3 concentrations (0.1 to 20 ppm) have been
performed to determine whether O3 is genotoxic. Hamelin (1985) showed by a combination of
agarose gel electrophoresis and electron microscopy that ozonation at 5 to 20 ppm caused
single- and double-strand DNA breaks, nicking, relaxation, linearization, and then degradation
of double-stranded plasmid pAT153 DNA molecules in solution. Hamelin also showed that
ozonation of plasmid DNA reduced the transforming ability of this plasmid, and that
Escherichia coli strains with mutations in DNA repair pathways (lexA, ozrA, and recA, but
not uvrA) were less able to support the transforming ability of the ozonated plasmid. Hence,
the lexA, ozrA, and recA gene products participate in repairing O3-induced DNA breaks.
Similarly, Sawadaishi et al. (1985) showed that ozonolysis of supercoiled pBR322
DNA resulted in conversion of closed-circular DNA molecules to open-circular DNA and
caused single-strand cleavage at specific sites. The concentrations of O3 employed were not
listed. Sawadaishi et al. (1986) further explored the specificity of O3-induced damage to
supercoiled plasmid pBR322 DNA by utilizing DNA sequencing techniques. The mechanistic
data obtained, showing preferential degradation of thymine bases, are very interesting
chemically, but the O3 concentrations used were far too high (25,600 ppm) to be useful in
assessing biologically relevant effects of O3.
Exposure of naked DNA from HeLa cells to 2 ppm O3 for 24 h resulted in the
formation of hydroxymethyluracil, thymine glycol, and 8-hydroxyguanine (Cajigas et al.,
1994). These results indicate a potential mutagenicity for O3, although the question of the
penetration of O3 to the DNA of intact cells has not been explored carefully. Mura and Chung
(1990) also studied the biological consequences of ozonation of DNA. They exposed phage T7
DNA to 5 ppm O3 for periods of 5 to 15 min and found that O3 decreased the template activity
of the DNA. Both the rate of initiation of transcription and the length of the RNA chains
transcribed were reduced. They concluded that O3 induced abnormal changes in the structure
of the phage T7 DNA and that these changes interfered with the ability of the DNA to be
transcribed. In mammalian cells, Van der Zee et al. (1987) demonstrated that ozonation of
murine L929 fibroblasts caused DNA strand breaks, DNA interstrand cross-links, and DNA
protein cross-links, but the O3 concentrations used were far too high (615 ppm) to be relevant
to the ambient exposures that are the focus of this document.
Kozumbo and Agarwal (1990) conducted in vitro studies in which specific
arylamines contained in tobacco smoke (1-naphthylamine, 2-naphthylamine, aniline,
p-toluidine, o-toluidine, and m-toluidine) were exposed to 0.1 to 1.0 ppm O3 for 1 h. When
the reaction products were added to human lung fibroblasts and transformed human Type 2
cells in vitro, DNA damage occurred. This raises the possibility that smokers could incur
6-109
-------
DNA damage in their lung cells due to the interaction of O3 with arylamines contained in
tobacco smoke, but there are no data on whether such reactions occur in vivo.
A logical consequence of these findings (Table 6-15) is that O3 could inhibit DNA
replication in mammalian cells and induce cytotoxicity to these cells, and this was found by
Rasmussen (1986). Rasmussen observed that DNA replication was inhibited in a
concentration-dependent manner in Chinese hamster V79 cells by O3 concentrations from 1 to
10 ppm following a 1-h exposure. These exposure regimens also induced cytotoxicity.
Table 6-15. Effects of Ozone on Deoxyribonucleic Acid Damagfe
Ozone
Concentration
Exposure Exposure
ppm Dg/m3 Duration Conditions
Cells
Observed Effects
Reference
0.1 196 Ih 15 DM Diploid human DNA breaks
1-napthylamine lung fibro-blasts
and transformed
Type 2 cells
Kozunibo and Agarwal
(1990)
2.0 3,920 24 h Phosphate Naked DNA
buffered saline from HeLa
cells
Formation of Cajigas et al. (1994)
hydroxymethyluracil,
thymine glycol, and
8-hydroxyguanine in
DNA
1.0-10
5.0
5.0-20
1,960- 1 h
19,600
9,800 5-15 min
9,800- 5- 15 min
39,200
Culture
DNA at
50 Dg/mL; O3 at
0.5 L/min, room
temperature
lOmMTris/HCl
1 mM EDTA
Hamster
(Chinese V79
cells)
Ozonated T7
phage DNA
Plasmid
pAT153
Inhibition of DNA
replication;
cytotoxicity
Decreased in vitro
transcription
Single-/double- strand
breaks in DNA
Rasmussen (1986)
Mura and Chung
(1990)
Hamelin (1985)
aSee Appendix A for abbreviations and acronyms.
Therefore, the available data show that O3 causes single- and double-strand breaks
in plasmid DNA in vitro, damages plasmid DNA so that its ability to serve as a template for
transcription is decreased, and inhibits DNA replication and causes cytotoxicity in Chinese
hamster V79 cells (Table 6-15).
6-110
-------
6.2.6.3 Induction of Mutation by Ozone
Consistent with its ability to induce DNA damage, the very high concentration of
50 ppm O3 also induced mutation to streptomycin resistance in E. coli, via both direct
mechanisms and indirectly by the rec-lex error-prone DNA repair system, by a factor from
two- to 35-fold in an exposure-time dependent manner (Table 6-16). However, no statistical
analysis was performed on these data (Hamelin and Chung, 1975a,b; L'Herault and Chung,
1984). In assays designed to detect base substitution mutations by gases in Salmonella
typhimurium, Victorin and Stahlberg (1988a) showed that O3 alone at concentrations of 0.1 to
3.5 ppm did not induce mutation to histidine auxotrophy in Ames' strains TA100, TA102, or
TA104. Ozone concentrations D2 ppm were cytotoxic. Victorin and Stahlberg (1988b) also
showed that 0.5 and 1.0 ppm O3, in combination with 1% vinyl chloride, and 1 ppm O3, plus
0.1 or 1 % butadiene, gave rise to a slight (approximately twofold) increase in mutation
frequency. In these in vitro studies, as in many studies reviewed in this section, it must be
pointed out that because O3 is so reactive, placing vinyl chloride and butadiene in the
experiments would result in exposure of the bacteria to reactive intermediates and reaction
products resulting from the mixture, not exposure of the bacteria to O3 alone. Further, these
increases in mutation were small (no more than twofold) and not statistically analyzed, and the
authors did not test strictly for concentration-dependent effects (Table 6-16). Thus, it is not
clear that these small effects are reproducible.
More recently, Dillon et al. (1992) studied the ability of O3 to induce mutation in
the Ames' strains of Salmonella (Table 6-16). Ozone caused no mutation in Salmonella strains
TA1535, TA98, TA100, and TA104. These authors found that 0.024- and 0.039-ppm
O3 exposures caused small increases in the mutation frequency in Salmonella tester strain
TA102, which is uniquely sensitive to detecting mutation induced by oxygen radicals. These
increases in mutation frequency were significant; however, the authors did not observe
consistent concentration-dependent increases in the mutation frequency. At the higher
concentrations of O3, there appeared to be an inverse dependence for induction of mutation by
O3. The authors indicated that the cytotoxicity of O3 complicated attempts to obtain a clear
concentration response for mutagenicity. The presence of Arochlor 1254-induced rat liver S-9
metabolic activation did not affect the mutational responses in any of the strains tested. These
authors did not observe reproducible increases in mutation frequency in Salmonella strains
TA98, TA100, TA104, or TA1535. Dillon et al. (1992) therefore concluded that O3 is a weak
bacterial mutagen only under specific conditions utilizing noncytotoxic concentrations in
TA102. However, because clear concentration-dependent responses for mutation could not be
achieved, it is not clear that O3 is definitively mutagenic in these studies.
Gichner et al. (1992) investigated whether O3 could induce mutation in two
mutagenicity assays in plants. The investigators found no induction of mutation in the
Nicotiana tabacum leaf color reversion assay or in the Tradescantia stamen hair assay at 0.1 to
0.3 ppm O3 (Table 6-16).
Dubeau and Chung (1979) showed that mutants of the yeast of Saccharomyces
cerevisiae deficient in repair of single- and double-strand DNA breaks were more sensitive to
the cytotoxicity of O3 than wild-type cells, indicating that O3 kills cells partly by generating
these types of breaks. Dubeau and Chung (1982) also showed that treatment of S. cerevisiae
with 50 ppm O3 for 30 to 90 min resulted in (1) an 11- to 14-fold increased frequency of
forward mutations, (2) an increase in reversions at six different loci by two- to threefold,
6-111
-------
Table 6-16. Summary of Findings on the Mutagenicity of Ozone
Ozone
Concentration
Exposure Exposure
ppni Dg/ni3 Duration Conditions
1.0
50
50
Cells
Observed Effects
Reference
0.024
0.039
0.39
0.1-
0.3
0.1-
3.5
0.5
1.0
47
76
764
196-
588
196-
6,860
980-
1,960
35 min Culture
35 min
35 min
5 or
1 1 h/day
for
1-15 or
18 days
6 h Culture
6 h Culture 1 %
vinyl
chloride
S. typhimurium
TA102
Nicotiana,
tabacam
Tradescantia
S. typhimurium
TA100, TA102,
or TA104
S. typhimurium
TA100
At 0.024 ppni: 2.4-fold
increase in mutation frequency
At0.039ppm: 1.6-fold
increase in mutation frequency
At 0. 39 ppm: 1 . 3-fold increase
in mutation frequency;
no effects seen in
S. typhimurium TA98, TA1535,
TA100, orTA104
No mutation at color locus
No mutation with or without
metabolic activation
170% increase in mutation;
statistical analysis not conducted
Dillon et al. (1992)
Gichner et al.
(1992)
Victorin and
Stahlberg (1988a)
Victorin and
Stahlberg (1988b)
1,960 6h Culture 0.1 S. typhimurium
or 1.0% TA100
butadiene
170% increase in mutation; Victorin and
statistical analysis not conducted Stahlberg (1988b)
5,000 1-20 min Culture
98,000 30-90 min Culture
E. coli
Saccharomyces
cerevisiae
Exposure-time-dependent
mutation to streptomycin
resistance, up to 35-fold
increases in mutation frequency
Forward mutations, reversions,
gene conversion, mitotic
crossing over; no statistical
analysis conducted
L'Herault and
Chung (1984)
Dubeau and Chung
(1982)
(3) an increase in gene conversions by two- to threefold, and (4) an increase in mitotic
crossover by 1.3-fold. No statistical analysis was performed on these data, nor were
concentration-response curves generated by the authors. These authors, therefore,
demonstrated that O3 was a mutagen and a recombination-inducing agent in S. cerevisiae.
However, they also showed that its genotoxic activity was weak (20- to 200-fold less activity in
terms of the frequency of mutants or recombinants induced) compared to the known mutagens
ultraviolet light, X rays, and A/-methyl-A/Z7-nitro-A/-nitrosoguanidine (MNNG).
In this section, the inclusion of data on bacterial mutation utilizing exposures up to
50 ppm O3 can be justified in order to help determine whether O3 is genotoxic at all. The
available information shows that O3 is not mutagenic in four Salmonella tester strains and may
6-112
-------
cause weak mutagenicity in Salmonella strain TA102, but these positive results are weakened
by the lack of a concentration-response effect. The data also show O3 is mutagenic inE". coli,
is weakly mutagenic in S. cerevisiae, but is not mutagenic in N. tabacam or Tradescantia.
Hence, because O3 is mutagenic in three assays, but not in six others, and is weakly mutagenic
in assays where the results are positive compared to known strong mutagens, O3 should be
considered a weak mutagen, at most (also reviewed in Victorin, 1992).
6.2.6.4 Induction of Cytogenetic Damage by Ozone
A number of investigators have studied whether O3 induces cytogenetic damage.
The previous air quality criteria document (U.S. Environmental Protection Agency, 1986)
described a number of in vitro and in vivo studies in which O3 exposure produced cytotoxic
effects on cells and cellular components, including genetic material; very few newer studies
have been reported. Cytogenetic and mutational effects of O3 have been reported previously in
isolated cultured cell lines, human lymphocytes, and microorganisms (Fetner, 1962; Hamelin
et al., 1977a,b; Hamelin and Chung, 1975a,b; Scott and Lesher, 1963; Erdman and
Hernandez, 1982; Guerrero et al., 1979; Dubeau and Chung, 1979, 1982). One of the earliest
studies by Fetner (1962) demonstrated that in vitro exposure of human KB cells to 8 ppm
O3 for 5 and 10 min induced two- and sixfold increases in the number of chromatid deletions.
Shiraishi et al. (1986) found that treatment of Chinese hamster V79 cells with 0.1 to 1.0 ppm
O3 inhibited growth of V79 cells by 10 to 70% and also induced a concentration-dependent
increase in the number of sister chromatic exchanges (SCEs) per cell, up to a maximum of
fourfold of that in untreated, control cells. The results indicate that if cells in culture are
exposed to sufficiently high concentrations of O3 for sufficiently long periods, chromosome
damage will result.
In vivo exposure studies are of greater potential interest. Cytogenetic and
mutational effects of O3 in laboratory animals and humans are controversial. Lymphocytes
isolated from animals exposed to O3 were found to have significant increases in the numbers of
chromosome (Zelac et al., 1971a,b) and chromatid (Tice et al., 1978) aberrations, after 4- to
5-h exposures to 0.2 and 0.43 ppm O3, respectively. Single-strand breaks in DNA of mouse
peritoneal exudate cells were measurable after a 24-h exposure to 1 ppm O3 (Chancy, 1981).
Gooch et al. (1976) analyzed the bone marrow samples from Chinese hamsters exposed to 0.23
ppm O3 for 5 h and the leukocytes and spermatocytes from mice exposed for up to 2 weeks to
0.21 ppm O3. No effect was found on the frequency of chromosome aberrations, nor were
there any reciprocal translocations in the primary spermatocytes. These authors did show that
there was a slight, but significant increase in the frequency of chromatid aberrations in human
peripheral leukocytes exposed in vitro to 7.2 and 7.9 ppm O3. The small increases observed in
chromatid abnormalities in peripheral blood lymphocytes from humans exposed to 0.5 ppm
O3 for 6 to 10 h were not significant, possibly because of the small number (n = 6) of subjects
studied (Merz et al., 1975). Subsequent investigations with improved experimental design and
more human subjects, however, did not show cytogenetic effects after exposure to O3 at
various concentrations and for various times (McKenzie et al., 1977; McKenzie, 1982;
Guerrero et al., 1979). Guerrero et al. (1979) showed no elevation in the frequency of SCEs
in circulating lymphocytes of humans exposed to 0.5 ppm of O3 for 2 h. However, these
authors did find that exposure of diploid human fetal lung (WI38) cells to 0.25, 0.50, 0.75,
and 1.0 ppm O3 for 1 h in vitro led to a concentration-dependent increase in SCEs in these
cells. In addition, epidemiological studies have not shown any evidence of chromosome
6-113
-------
changes in peripheral lymphocytes of humans exposed to O3 in the ambient environment (Scott
and Burkart, 1978; Magie et al., 1982). Evidence now available, therefore, fails to
demonstrate any cytogenetic or mutagenic effects of O3 in humans when the exposure regimens
are representative of exposures that the population might actually experience.
Finally, a study conducted by Erdman and Hernandez (1982) showed that treatment
of Drosophila virilis with 30 ppm O3 for 2 to 6 h resulted in an exposure time-dependent
accrual of dominant lethals.
Therefore, O3 does induce chromosomal aberrations in cultured cells, but the results
in animals exposed to O3 for chromosomal breakage are, at most, weak and their biological
significance is controversial.
6.2.6.5 Induction of Morphological Cell Transformation by Ozone
Ozone has been studied in a number of mammalian cell culture systems to determine
whether it can induce cell transformation (Table 6-17). The cell transformation assay in
C3H/10T1/2 (10T1/2) mouse embryo cells is a standard assay that has been used by many
investigators to detect cell tranformation activity as a potential indicator of carcinogenicity and
to study molecular mechanisms of cell transformation induced by organic chemicals,
carcinogenic metals, and radiation (reviewed in Landolph, 1985, 1989, 1990, 1994). Syrian
hamster embryo (SHE) cells also are widely used to detect cell transformation by many classes
of chemical carcinogens and radiation (Borek et al., 1986, 1989a,b). Borek et al. (1986)
demonstrated that exposure of SHE cells and 10T1/2 mouse embryo cells to 5 ppm O3 for 5
min inducted morphological transformation in both cell types. Also, in both cell types, there
was a synergistic induction of morphological transformation when the cells were treated with
3 Grays of gamma radiation and 5 ppm O3. These authors therefore concluded that O3 acts as
a direct cell transforming agent and as a co-cell transforming agent in the presence of gamma
radiation. Borek et al. (1989a) also observed an additive amount of transformation when these
cell types were treated with 6 ppm O3 and 4 J/m2 of ultraviolet light. A further study by Borek
et al. (1989b) showed that exposure of 10T1/2 mouse embryo cells to 1 ppm O3 for 5 min did
not result in morphological transformation, but that increasing exposure to 5 ppm increased the
transformation frequency by a factor of 15. Ozone and gamma radiation caused a synergistic
increase in morphological transformation when O3 was added to cells after the gamma
radiation. When O3 was added to cells before the gamma radiation, the transformation was not
increased over that due to gamma radiation. These authors also showed by DNA transfection
experiments that three O3-induced transformed cell lines possessed dominantly acting
transforming genes. In these studies, cells were incubated in phosphate-buffered saline during
O3 treatment and, hence, likely were exposed to reaction products of O3 rather than O3 itself.
Thomassen et al. (1991, 1992) exposed rats by inhalation to 0.14, 0.6, or 1.2 ppm
O3 for 6 h/day, 5 days/week, for a total of 1, 2, or 4 weeks. There was no increase in the
frequency of preneoplastic transformation in cells removed from the tracheas and subsequently
cultured. Cells incubated in serum-free medium and exposed to 0.7 or 10 ppm O3 for 40 min
in vitro also did not show an increased frequency of preneoplastic transformation. When rat
tracheal epithelial cells were exposed in vitro to 0.7 ppm O3,
6-114
-------
Table 6-17. Effects of Ozone on Morphological Cell Transformation
0
0
1
0
Ozone
Concentration
ppm Dg/m3
.14, 274,
.6, 1,176,
.2 2,352
.7 1,372
0.7 1,372
10 19,600
1
5
5
6
.0 1,960
.0 9,800
.0 9,800
.0 11,760
Exposure
Duration
6 h/day,
5 days/week
for 1, 2, or
4 weeks
40 min twice
weekly for
5 weeks
(in vitro)
40 min
(in vitro)
5 min
(in vitro)
5 min
(in vitro)
10 min
(in vitro)
Species, Sex
(Strain/Cells)
Age"
Rat, M
(F344/N)
7-9 weeks old
Rat
(Tracheal
epithelial cells)
Rat
(Tracheal
epithelial cells)
Mouse
(C3H/10T1/2
embryo cells)
Hamster (Primary
diploid cells)
Mouse
(C3H/10T1/2
embryo cells)
Hamster
(Syrian primary
embryo cells)
Mouse
(C3H/10T1/2
embryo cells)
Observed Effects
No induction of preneoplastic
variants in cultured tracheal
epithelial cells
Twofold increase in frequency of
preneoplastic variants; additive
effects with MNNG
No induction of preneoplastic
variants
At 1.0 ppm: No morphological
transformation alone; increased
transformation by 0.4-Gray radiation
by 1.7-fold like a co-carcinogen.
At 5.0 ppm: 15-fold increase in
morphological transformation and
synergism with 4-Gray gamma
radiation transformation.
In both cell lines, induction of
morphological transformation and
synergism with gamma rays
In both cell lines, induction of
morphological transformation;
additive transformation with UV
light
Reference
Thomassen et al. (1991)
Thomassen et al. (1991)
Thomassen et al. (1991)
Boreketal. (1989b)
Borek et al. (1986)
Borek et al. (1989a)
aSee Appendix A for abbreviations and acronyms.
bAge at start of exposure.
two times a week for 5 weeks, there was approximately a twofold increase in the frequency of
preneoplastic variants detected. These authors also showed that treatment of rat tracheal
epithelial cells with MNNG followed by exposure of cells to 0.7 ppm O3 twice weekly for
5 weeks resulted in an approximately additive increase in the frequency of preneoplastic
variants of the cells. The results of these studies should be interpreted cautiously because the
changes are very small and because O3 exposure followed by MNNG treatment yielded
negative results (Thomassen, 1992). In addition, it is likely that the culture conditions may
significantly affect these results, particularly the volume of culture medium above the cells,
6-115
-------
and this variable has not been explored in sufficient detail. Further, the results of these
experiments are somewhat variable. Recently, a system has been developed to culture human
tracheobronchial epithelial cells and expose them to consistent and reproducible levels of
O3 (Tarkington et al., 1994).
In all these cell transformation experiments, the reactivity of O3 makes it likely that
secondary reaction products of O3 formed in the aqueous medium, not O3 itself, induced the
cell transformation. Therefore, O3 is able to induce morphological transformation in
C3H/10T1/2 mouse embryo cells and in SHE cells at high concentrations (1.0, 5.0, and
6.0 ppm) but causes no significant effects in rat tracheal epithelial cells in vitro or in vivo.
6.2.6.6 Possible Direct Carcinogenic, Co-carcinogenic, and Tumor-Promoting Effects of
Ozone as Studied in Whole Animal Carcinogenesis Bioassays
To investigate whether O3 has carcinogenic, co-carcinogenic, or tumor-promoting
effects, a number of investigators have conducted in vivo carcinogenesis bioassays with
O3 (Table 6-18). Some of the studies have used strain A mice (reviewed in Mustafa, 1990).
The advantages and disadvantages in using strain A mice as a general screen for carcinogens
by the ip route have been discussed in the literature (Stoner and Shimkin, 1985; Maronpot
et al., 1986; Stoner, 1991; Maronpot, 1991). Strain A mice only rarely have been used in
inhalation carcinogenesis assays. In addition, the A/J strain of mice has a high spontaneous
incidence of benign pulmonary tumors (adenomas). This strain of mice has been shown to be
very sensitive to tumor induction by poly cyclic aromatic hydrocarbons (PAHs), carbamates,
and aziridines and insensitive to aromatic amines, metal salts, and halogenated organic
compounds administered by the ip route (Maronpot et al., 1986). In addition, carcinogenicity
results in strain A mice did not correlate well with 2-year mouse and rat carcinogenicity results
when the results of chemical testing in strain A/St mice (59 chemicals tested) were compared
with strain A/J mice (30 chemicals tested) in a 2-year chronic bioassay (Maronpot et al.,
1986). The chemicals chosen were heavily weighted with aromatic amines. The author
concluded that "carcinogenicity test data are relevant only to the test model employed since
there is no absolute reference for carcinogenicity." Maronpot (1991) also demonstrated a poor
concordance between results of testing chemicals in the strain A assay and testing them in 2-
year rat and mouse carcinogenicity assays at the National Cancer Institute (NCI). Stoner
(1991), using the strain A mouse pulmonary assay, indicated that ip injection of PAHs,
nitrosamines, nitrosureas, carbamates, aflatoxin, metals, and hydrazines induces tumors, but
that the assay is not responsive to aromatic amines, aliphatic halides, and certain compounds
carcinogenic in rodent liver or bladder. In this assay, an increase in lung tumor multiplicity
(average number of lung tumors per mouse) caused by a chemical is considered as evidence for
the carcinogenicity of a chemical.
Hassett et al. (1985) used inbred strain A/J mice, which are very sensitive to
induction of pulmonary adenomas by chemical carcinogens. Exposure of A/J mice to
0.31 ppm O3 for 103 h per week, every other week, for 6 mo, resulted in a 1.3-fold increase
(not statistically significant) in the percent of mice with tumors (tumor incidence) and a
statistically significant 1.4-fold increase in the number of tumors per mouse (tumor
multiplicity). In this experiment, O3 did not promote the carcinogenicity of urethane when
6-116
-------
Table 6-18. Summary of Results on the Possible Carcinogenicity of Ozonfe
Ozone
Concentration
ppm Qg/m3
0.05 98
(sine wave (sine wave
from 0 to from 0 to
0.1) 96)
0.31 608
0.5 980
0.4 784
0.8 1,568
0.8 1,568
0.12
0.5
1.0
0.5
1.0
0.12
0.5
1.0
0.5
1.0
Exposure
Duration
10 h/day for
13 mo
103 h/week,
every other
week for 6 mo
102 h/first week
of each month
for 6 mo
8 h/day,
7 days/week for
18 weeks
23 h/day
7 days/week for
6 mo
6 h/day
5 days/week for
104 weeks
(2 years)
6 h/day
5 days/week for
124 weeks
(lifetime)
6 h/day
5 days/week
105 weeks
(2 years)
6 h/day
5 days/week
130 weeks
(lifetime)
Species, Sex
(Strain)
Age"
Rat, M
(Wistar)
4 weeks old
Mouse, F
(A/J)
7 weeks old
Mouse, F
(A/J)
7 weeks old
Mouse, M
(Swiss Webster and
A/J)
8 weeks old
Hamster, M
(Syrian Golden)
7-11 weeks old
Rat, M, F
(F344/N)
Rat, M, F
(F344/N)
Mouse, M, F
(B6C3F,)
Mouse, M, F
(B6C3F,)
Observed Effects
Lung tumor response increased from 0% in
BHPN- or O3-treated animals, to 8.3% in
animals treated with 0.5 g/kg BHPN +
0.05 ppm O3 (not significant).
1.33-fold increase in percent of mice with
adenomas, 1.42-fold increase in number of
tumors per mouse. No promotion of
carcinogenicity of urethane (2 mg/mouse
before O3).
2. 1 1-fold increase in percent of mice with
tumors, 3.42-fold increase in number of
tumors per mouse, interaction between
O3 and urethane (2 mg/mouse after each
O3 week).
Urethane treatment before O3 started. In
Swiss Webster mice: No increase in lung
tumor incidence; nonsignificant decrease in
tumors per lung in urethane-treated animals.
In A/J strain: No effect at 0.4 ppm.
At 0.8 ppm: Threefold increase in percent
mice with tumors, and 4.2-fold increase in
number of tumors per mouse. Both 0.4 and
0.8 ppm O3 decreased yield of tumors per
mouse in urethane-treated mice, but had no
effect on tumor incidence.
No tumors observed in animals treated with
0.8 ppm O3 only. In animals treated with
20 mg/kg DEN sc twice/week, 0.8 ppm
O3 did not increase tumors of lung, bronchus,
trachea, or nasal cavity. Tumors of lung
were decreased 50% (N.S.)
No increase in neoplasms at any
concentration tested.
No increase in neoplasms at any
concentration tested.
No effects in males. In females:
Increase in number of mice with neoplasms at
1 .0 ppm (combined alveolar/bronchiolar
adenoma or carcinoma in lung).
In males: Increase in number of mice with
carcinoma at 0.5 and 1.0 ppm, but not
significant for change in number of mice with
total neoplasms. In females: increase in
number of mice with adenomas, but not
carcinomas or total neoplasms.
Reference
Ichinose and Sagai (1992)
Hassett et al. (1985)
Hassett et al. (1985)
Last et al. (1987)
Witschi et al. (1993a,b)
National Toxicology
Program (1994)
Boorman et al. (1994)
National Toxicology
Program (1994)
Boorman et al. (1994)
National Toxicology
Program (1994)
National Toxicology
Program (1994)
"See Appendix A for abbreviations and acronyms.
'"Age at start of exposure.
O3 exposure began 1 week after a single injection of animals with a total dose of 2 mg
urethane/mouse. In a second experiment, exposure to 0.50 ppm O3 (102 h during the first
week of every month for 6 mo) caused a nonsignificant 2.1-fold increase in tumor incidence
and a 3.2-fold increase in tumor multiplicity (statistics not shown) (Hassett et al., 1985).
6-117
-------
These authors reported that exposure to 0.5 ppm O3, followed by urethane treatment (2 mg
after each O3 exposure set), resulted in an interaction between O3 and urethane such that there
were more animals with more than 16 lung tumors each. These authors concluded that
exposure to 0.31 and 0.5 ppm O3 increased the yield of pulmonary adenomas in A/J mice and
that O3 interacted with urethane to produce more lung tumors than urethane alone when O3 was
added before urethane.
The pulmonary adenomas induced in animals by chemical carcinogens are by
definition benign tumors (Stoner, 1991; Maronpot, 1991); however, they do represent
abnormal cell growth in the form of tumors and, hence, are significant biologically in that they
are early steps in the pathway toward malignancy. It is not known if these tumors can progress
to malignant tumors, because they are not as amenable to observation as are mouse skin
tumors, where adenomas can be converted into adenocarcinomas at a frequency of
approximately 8 to 10%. Similarly, it is not known with certainty whether pulmonary
adenomas in humans progress to adenocarcinomas. It is clear, however, that lung tumors in
mice are not equivalent to bronchogenic carcinomas in humans. However, because a shift is
occurring, in that fewer squamous cell bronchogenic carcinomas and more peripheral
adenocarcinomas are occurring in humans, the induction of peripheral adenomas and their
progression to peripheral adenocarcinomas in mice may be a useful area for further
mechanistic insight.
The study of Hassett et al. (1985) was reviewed extensively by scientists from EPA
and the National Institutes of Environmental Health Sciences in 1985 and 1986 (Tilton, 1986).
The consensus of these extensive reviews was that (1) the tumor yields in O3-exposed mice
were not statistically significantly different from the control animals, (2) any effects were
marginally different from control values, and (3) the strain A mouse has a high spontaneous
incidence of tumors, making it difficult to interpret the effects of O3. Chemical induction of
tumors in this assay system did not correlate well with the 2-year NCI carcinogenesis bioassay
results. In addition, because Hassett et al. (1985) did not demonstrate a concentration-response
effect in animals exposed to O3, the consensus among the reviewers was that one could not
conclude from these experiments that O3 was a significant carcinogen or tumor promoter, and
that rigorous inhalation carcinogenesis bioassays needed to be carried out with O3-exposed
animals to address this issue properly.
Last et al. (1987) also studied whether O3 exposure could influence the yield of
urethane-induced lung tumors in A/J and Swiss-Webster mice. Urethane treatment consisted
of a single ip injection (1,000 mg/kg) 1 day before O3 exposure began. In Swiss Webster
mice, exposure to 0.4 or 0.8 ppm O3 alone not only did not increase the tumor yield but
actually decreased the yield of urethane-induced lung tumors per mouse, although the decrease
was not statistically significant. In A/J mice, exposure to 0.4 ppm O3 did not increase the lung
tumor yield, but exposure to 0.8 ppm O3 caused a threefold increase in tumor incidence and a
4.2-fold increase in lung tumor multiplicity. Exposure of urethane-treated mice to 0.4 or
0.8 ppm O3 decreased lung tumor multiplicity but had no effect on tumor incidence. These
differences in the strain A mouse were significant. The authors concluded that O3 was not a
tumor promoter or tumor-enhancing agent.
Ichinose and Sagai (1992) studied the ability of O3 to interact with
./V-fe(2-hydroxypropyl) nitrosamine (BHPN) in the induction of lung tumors in Wistar rats. A
single ip injection of BHPN (0.5 g/kg) did not cause any tumors in the rats. Rats were
exposed for 10 h/day for 13 mo to a pattern of O3 consisting of a sine curve from 0 to
6-118
-------
0.1 ppm, with a mean concentration of 0.05 ppm. The 13-mo O3 exposure started the day
after BHPN injection, and the rats were examined 11 mo postexposure. No tumors were
observed in the O3-alone or control groups. When rats were exposed to 0.5 g/kg BHPN plus
0.05 ppm O3, the lung tumor incidence increased to 8.3% (3/36), but this increase was not
statistically significant. The tumors observed in this study cannot be stated definitively to have
been induced by the treatment agent.
The data available on O3 exposure and lung carcinogenesis up until 1988 have been
reviewed by Witschi (1988). The chemical reactivity of O3 and, in particular, its radiomimetic
activity (ability to mimic radiation effects, such as causing cell-cycle arrest, chromosome
breakage, etc.) also make O3 a potential risk factor for human lung cancer. Nevertheless, as of
1988, there were no experimental studies conclusively linking lifelong exposure to O3 with
lung tumor induction in any animal species, nor was there conclusive epidemiological evidence
to associate O3 exposure with the development of lung cancer in humans. As of 1988, the only
data implicating O3 as a possible tumorigenic agent were from studies carried out in mice,
where the tumors are adenomas derived from Type 2 alveolar cells or from Clara cells
(reviewed in Witschi, 1988). In the A and Swiss-Webster mouse strains used to assay the
carcinogenicity of O3, the spontaneous incidence of lung tumors is very high. Hence, results
of carcinogenicity experiments conducted on O3 to date that utilize tumor incidence as an
endpoint are not strongly positive, due to this high background. In strain A mice, in which the
spontaneous multiplicity is usually less than one tumor per mouse, the tumor multiplicity is
considered by many investigators to be a sensitive indicator of a carcinogenic effect.
However, even using this indicator, the increase in tumor multiplicity after O3 exposure is
small, raising questions about the biological significance of the effects. In addition, the assay
for inhalation carcinogenesis in strain A mice was not fully validated at this time. Whereas
Hassett et al. (1985) concluded that O3 increased the number of pulmonary adenomas in strain
A mice, Witschi (1988) concluded that O3 was not implicated unequivocally as a carcinogen in
strain A/J mice, that no classical carcinogen bioassays had been conducted on O3, and that a
definitive judgment could not be made on the carcinogenicity of O3.
A review (Witschi, 1991) of the available data on the carcinogenicity of O3 and
oxygen in mouse lungs indicated that oxidants can enhance or inhibit mouse lung
tumorigenesis, depending on the experimental protocol employed, and the carcinogenicity of
O3 in mouse lung had not been established unequivocally. Exposure of strain A mice to
O3 induces hyperplasia of Type 2 alveolar cells, leading to expansion of the target cell
population. It was speculated that this might have resulted in spontaneous transformation of
these cells (Witschi, 1991). In lungs of animals treated with a carcinogen such as urethane and
then exposed to O3 before or after carcinogen administration, it was speculated that O3 may
cause cell proliferation and result in fixation of DNA damage (Witschi, 1991). The addition of
O3 after carcinogen exposure leads to a decreased tumor incidence compared to treatment with
carcinogen alone; the reasons for this decrease with late O3 exposure are not clear.
The effects of treating male Syrian Golden hamsters with dimethylnitrosamine
(DEN) (20 mg/kg given subcutaneously twice per week) during the course of a 6-mo exposure
to 0.8 ppm O3 were studied by Witschi et al. (1993a; see Table 6-18). The rationale for this
study was to test the hypothesis that O3 acts in a manner similar to hyperoxia in enhancing
neuroendocrine lung tumors in this animal model. After exposure ceased, the animals were
maintained in air for 1 mo. Ozone exposure did not increase the incidence of lung, bronchus,
trachea, or nasal cavity tumors in the DEN-treated hamsters. There was a 50% decrease in the
6-119
-------
percent of animals with lung tumors in the DEN-plus-O3-exposed animals compared to the
DEN-plus-air-exposed animals, but this was not statistically significant. Ozone did not affect
the incidence of DEN-induced liver tumors. The authors concluded that O3 did not increase
the number of DEN-induced respiratory tumors in hamsters and that O3 exposure might have
inhibited or delayed tumor development. Although reduction in tumor incidence caused by
O3 was not significant in this study alone, overall analysis of these data with other data from a
4-mo exposure of similar design (Witschi et al., 1993b) yielded significant results.
A definitive study of the carcinogenicity of O3 and its ability to act as a
co-carcinogen or tumor promoter was conducted by the U. S. National Toxicology Program
(NTP) (National Toxicology Program, 1994; Boorman et al., 1994). Animals were exposed to
air or O3 for 6 h/day, 5 days/week for the number of weeks described below. Male and female
F344/N rats were exposed to 0.12, 0.5, and 1.0 ppm O3 for 104 weeks and to 0.5 and 1.0 ppm
for 124 weeks ("lifetime" of the animals). Similar protocols were used for B6C3F, male and
female mice, with the exception that the "2-year" study was 105 weeks and the lifetime study
was 130 weeks. This study did not find any evidence of carcinogenic activity in male or
female rats. There was a negative trend for mammary gland neoplasms in the female rats in
the 2-year study, an effect that was not seen in the lifetime study. The NTP found "equivocal
evidence"1 of carcinogenic activity in O3-exposed male mice and "some evidence" of
carcinogenic activity in female mice exposed to O3. In co-carcinogenesis experiments, male
rats were treated with a known pulmonary carcinogen, 4-(A/-methyl-A/-nitrosomino)-l-(3-
pyridyl)-l-butanone (NNK) (0.1 and 1.0 mg/kg, subcutaneous injection 3 times a week for first
20 weeks) and exposed to 0.5 ppm O3 for 6 h/day, 5 days/week for 105 weeks. The NTP
found "no evidence" that O3 enhanced the incidence of NNK-induced pulmonary neoplasms.
Table 6-19 shows the tumor incidences in mice. In the discussion to follow, all tumors
described were at lung alveolar/bronchiolar sites. There was a decrease in the number of
hepatocellular adenomas or carcinomas in female mice exposed to 1.0 ppm O3 for 2 years and
for hepatocellular carcinomas in the lifetime study. There was no statistically significant
increase in tumors at any site other than the lung.
In male mice exposed to O3 for 2 years, there were no statistically significant
increases in the mice with alveolar/bronchiolar carcinomas or a combination of adenomas or
carcinomas; at 0.5 ppm O3, there was a small, twofold increase in the incidence of adenomas.
In the lifetime studies of male mice, the incidence of mice with carcinomas increased 1.9-fold
at 0.5 ppm O3 and 2.3-fold at 1.0 ppm. The incidence of adenomas in
'The NTP evaluates the strength of the evidence for conclusions regarding each carcinogenicity study, under the
conditions of that particular study. There are five categories: two for positive results ("clear evidence" and "some
evidence"), one for uncertain findings ("equivocal evidence"), one for no observable effects ("no evidence"), and
one for experiments that cannot be judged because of major flaws ("inadequate study") (National Toxicology
Program, 1994). This approach is very different from the weight-of-evidence approach used by EPA for cancer
classification because the EPA approach considers all the available studies.
6-120
-------
Table 6-19. Alveolar/Bronchiolar Tumor Incidence in B6C3I; Mice
in the National Toxicology Program's Chronic Ozone Study
Males
ppm
2-year exposure
0.0
0.12
0.5
1.0
Lifetime exposure
0.0
0.5
1.0
Combined
0.0
0.5
1.0
Adenomas"
6/50
9/50 (p = 0
12/50 (p =
1 1/50 (p =
8/49
8/49 (p = 0
9/50 (p = 0
14/99
20/99 (p =
20/100 (p =
.3)
0.06)
0.11)
.61)"
.47)
0.16)
= 0.14)
Carcinomas"
8/50
4/50 (p = 0
8/50 (p = 0
10/50 (p =
8/49
15/49 (p =
18/50 (p =
16/99
23/99 (p =
28/100(p =
0.009)
.15)
.45)
0.27)
0.05)
0.007)
0.08)
Both"
14/50
13/50 (p =
0.44)
18/50 (p =0.12)
19/50 (p =
16/49
22/49 (p =
21/50 (p =
30/99
40/99 (p =
0.10)
0.14)
0.15)
0.06)
40/100 (p = 0.04)
Adenomas"
4/50
5/50 (p =
5/49 (p =
8/50 (p =
3/50
3/49 (p =
0.55)
0.52)
0.24)
0.63)
ll/50(p = 0.02)
7/100
8/98 (p =
19/100 (p
0.48)
= 0.01)
Females
Carcinomas*
2/50
2/50 (p = 0.65)b
5/49 (p = 0.26)
8/50 (p = 0.053)
3/50
5/49 (p = 0.33)
2/50 (p = 0.50)b
5/100
10/98 (p = 0.14)
10/100 (p = 0.14)
Both"
6/50
7/50 (p =
9/49 (p =
16/50 (p =
6/50
8/49 (p =
12/50 (p =
12/100
17/98 (p =
28/100 (p
0.57)
0.33)
= 0.02)
0.34)
= 0.10)
= 0.20)
= 0.004)
"Number of animals with neoplasm/number of animals necropsied (p [probability] value, logistic regression test).
''Lower incidence.
Source: National Toxicology Program (1994).
male mice did not change significantly. In female mice exposed for 2 years, there were no
statistically significant changes at 0.12 or 0.5 ppm O3. However, at 1.0 ppm O3, there was a
fourfold increase in the frequency of female mice with carcinoma and a 2.7-fold increase in
combined adenomas plus carcinomas. In female mice exposed for their lifetimes to O3,
0.5 ppm caused no significant effects. At 1.0 ppm O3, there was a 3.7-fold increase in the
incidence of mice bearing pulmonary adenomas, a nonsignificant change in the frequency of
mice with carcinomas, and a twofold (p = 0.1) increase in the incidence of combined
adenomas and carcinomas.
When the results of the 2-year and lifetime O3 carcinogenesis studies were
combined and analyzed, for male mice, there was no statistically significant increase in the
incidence of animals bearing adenomas, and, for carcinomas, there was a marginally
significant increase at 0.5 ppm O3 (1.4-fold increase, p = 0.08) and a significant, 1.7-fold
increase at 1.0 ppm. The incidence of male mice bearing adenomas or carcinomas showed a
marginally significant increase (1.3-fold, p = 0.06) at 0.5 ppm O3 and a 1.3-fold increase at
1.0 ppm (p = 0.045). In the combined analysis of the 2-year and lifetime exposure of female
mice, there were no statistically significant changes at 0.5 ppm O3. At 1.0 ppm O3, there was
a 2.7-fold increase in the percent of mice bearing adenomas and a 2.3-fold increase in the
frequency of mice with adenomas or carcinomas. There was no statistically significant
increase in the carcinoma incidence.
The overall conclusions of the authors of the NTP O3 inhalation carcinogenesis
study were (1) there was no increased pulmonary tumor incidence in male or female F344/N
rats exposed to 0.12, 0.5, or 1.0 ppm O3; (2) male F344/N rats treated with the tobacco
carcinogen NNK and exposed to 0.5 ppm O3 did not have an increase in the pulmonary tumor
incidence above that caused by NNK alone; (3) O3 caused a slightly increased incidence of
alveolar/bronchiolar adenoma or carcinoma that yielded equivocal evidence of carcinogenicity
6-121
-------
of O3 in male B6C3F, mice; and (4) O3 increased the incidence of alveolar/bronchiolar
adenoma or carcinoma in female B6C3F, mice, yielding some evidence of carcinogenic activity
of O3 in female mice.
Generally, in mice, adenomas appear to progress into carcinomas with time, and,
thus, the incidence of mice having both adenomas and carcinomas is probably the more useful
indicator of effects. The incidence of tumor-bearing mice was elevated significantly only in
female mice exposed for 2 years to 1.0 ppm O3. When both the 2-year and lifetime exposure
studies were combined, there was a increased incidence of tumors in the males at 0.5 and 1.0
ppm O3 and in females at 1.0 ppm. The NTP designated the data for male mice as equivocal
for carcinogenesis because the combined tumor incidence in the 2-year study was within the
historical range, and the combined incidence for the lifetime study was not significant, even
though the carcinoma incidence was significant in the lifetime study. The evaluation of female
mice resulted in NTP's finding of "some evidence of carcinogenic activity" because the
combined pulmonary adenoma/carcinoma incidence was significantly increased and outside the
range of the historical control tumor rates. When the lifetime and 2-year studies were
combined, there were 28/100 adenomas plus carcinomas in the 1.0-ppm O3 exposure group
versus 12/100 in the controls (p = 0.004).
In summary, the strongest data on carcinogenicity come from the NTP study, which
was ambiguous in male mice and positive only in female mice at high concentrations of
O3 (i.e., 1.0 ppm). This may represent a toxic or irritant effect, giving a nonspecific type of
tumor due to mitogenesis. The carcinogenicity data are weak or equivocal in male mice,
negative in F344/N male and female rats, and negative for co-carcinogenesis in male rats.
Therefore, the potential for animal carcinogenicity is uncertain at the present time.
6.2.6.7 Possible Effects of Ozone on Injected Tumor Cells That Lodge in the Lung and
Form Lung Colonies
To date, no rigorous studies have been conducted to examine the effects of O3 on
true lung tumors that would have metastasized. No studies have been conducted in which lung
tumor cells detach themselves from primary lung or other tumors growing in organs and
invade adjacent tissue, blood vessels, or lymphatics. A few studies have been conducted in
which tumor cells are injected intravenously into animals and then lodge in the lung, forming
lung colonies (Table 6-20). It must be stressed, however, that this experimental model is not
an adequate model for lung tumor cell metastasis.
Kobayashi et al. (1987) showed that exposure of C3H/He mice for 1 or 14 (but not
other) days to DO. 1 ppm O3 after mice were injected in the tail vein with the fibrosarcoma cell
line (NR-FS) increased the number of metastatic lung tumors. Animals were exposed to
O3 for 14 days, then fibrosarcoma cells were injected into the tail vein of the animals, and
pulmonary metastases were scored 14 days later. One day of exposure to 0.8 ppm O3 gave the
maximal enhancement of pulmonary metastases. This enhancement of pulmonary metastasis
was concentration-dependent, in the range from 0.4 to 0.8 ppm O3 from 1 to 14 days, but
increases were small. This effect may arise in two ways: (1) by damage to the
microvasculature and (2) by the differential sensitivity of various tumor cells to O3 cytotoxicity
(reviewed by Witschi, 1988). Richters (1988) reported that exposure of mice to 0.15 or 0.30
ppm O3 for 60 days did not increase colonization of the lungs of mice injected iv with B16
melanoma cells.
6-122
-------
Table 6-20. Effects of Inhaled Ozone on the Ability of Injected
Tumor Cells To Colonize the Lungs of Mice*
Ozone
Concentration
ppm
0.15
0.3
0.1
0.2
0.4
0.8
Dg/m3
294
588
196
392
784
1,568
Species, Sex
Exposure (Strain)
Duration Ageb
60 days Mouse, M
(C57B/6J)
5 weeks old
1-14 days Mouse, M
(C3H/He)
8- 12 weeks old
Observed Effects
No increase in lung metastases from
iv-injected B16 melanoma cells.
Pulmonary metastases from iv-injected
NR-FS fibrosarcoma cells.
After 1 and 14 days of 0.1 ppm, 1.3-fold
increases. After 5 and 7 days of 0.2
ppm, 1.3 and 2.3-fold increases.
After 1 and 5 days of 0.4 ppm, 2.3- and
2. 2- fold increases.
After 1 day of 0.8 ppm, 4.6-fold
increase.
Reference
Richters (1988)
Kobayashi et al. (1987)
aSee Appendix A for abbreviations and acronyms.
bAge at start of exposure.
6.2.6.8 Summary and Conclusions
In summary, there are some weakly positive data and some negative data on the
genotoxicity of O3 (summarized in Table 6-21). Ozone at very high concentrations (5 to
20 ppm) causes DNA strand breakage in plasmid DNA (Hamelin, 1985). Ozone is, at most,
weakly mutagenic in some assays and negative in others. Ozone is not mutagenic in
Tradescantia or N. tabacam at concentrations of 0.1 to 0.3 ppm (Gichner et al., 1992); is
weakly mutagenic in E. coli at 50 ppm, and S. cerevisiae at 50 ppm (L'Herault and Chung,
1984; Dubeau and Chung, 1982); and is nonmutagenic in three strains of Salmonella and, at
most, marginally mutagenic in Salmonella strain TA102 at concentrations of 0.024, 0.039, and
0.39 ppm (Victorin and Stahlberg, 1988a,b; Dillon et al., 1992). Despite extensive studies by
Dillon et al. (1992), the mutagenicity of O3 in Salmonella TA102 is not conclusive because
convincing concentration-dependent mutagenic effects have not yet been demonstrated,
possibly due to the strong cytotoxicity of this compound. Ozone causes cytogenetic damage in
cultured cells in vitro (e.g., Hamelin et al., 1977a,b; Dubeau and Chung, 1979, 1982), but no
effects or small and conflicting effects when animals are exposed in vivo (Zelac et al.,
1971a,b; Tice et al., 1978). Cell transformation studies have shown positive results on
exposure of cells to O3, but these studies were conducted with a fluid barrier above the cells
that may have resulted in artifacts compared to an in vivo exposure (Borek et al., 1986,
1989a,b).
The in vitro studies are mechanistically interesting, but there are difficulties in the
design of many of these studies. First, the concentrations used in these in vitro studies were
typically orders of magnitude greater than those found in ambient air. Second, extrapolation of
in vitro exposure concentrations to human exposure dose requires special methods that were
not used in these studies. Third, direct exposure of isolated cells to O3 is somewhat
6-123
-------
Table 6-21. Summary of Data on the Genotoxicity of Ozon§
Assay System in Which Ozone Was Tested Resultb Comments
Mutation to histidine prototrophy in
Salmonella TA100
Mutation to histidine prototrophy in
Salmonella TA102
Mutation to streptomycin resistance in
Eshcherichia coli
Mutation in Saccharomyces cerevisiae
Mutation in Nicotiana tabacam in a
leaf-color reversion assay
Mutation in Tradescantia in a stamen-hair
assay
Chromosomal breakage in cultured
mammalian cells
Chromosomal breakage in animals
Morphological transformation in
C3H/10T1/2 mouse embryo cells and in
Syrian hamster embryo cells
Induction of preneoplastic variants in rat
tracheal epithelial cells
D Small effects obtained, less than
twofold; concentration-response
effect was not shown at 0.024,
0.039, and0.39ppmO3.
+/D Small effects obtained, and there was
no direct exposure-response at 0.024,
0.039, and0.39ppmO3.
+ Only 50 ppm O3 was tested.
+ Ozone caused mutation and
recombination at 50 ppm, but this
was a weak response compared to
known strong mutagens (20- to
200-fold less mutagenic than UV
light, X rays, and MNNG).
D 0.1 to 0.3 ppm O3 was tested.
D 0.1 to 0.3 ppm O3 was tested.
+ 8 ppm O3 in human KB cells, and
0.1 to 1.0 ppm O3 in V79 cells.
+/D Results are, at best, weak and
controversial; results in this assay are
considered ambiguous and not
definitively positive at present.
Ozone was tested at 0.2, 0.43, 7.3,
and 7.9 ppm.
+ Experiments need to be conducted
without or with only minimal
amounts of fluid bathing the cells.
Concentrations giving positive results
are high (5 and 6 ppm O3).
D Both in vitro and in vivo exposures
give negative or, at most, only
twofold increases in cells exposed to
0.14, 0.6, 0.7, 1.2, or 10 ppm O3.
6-124
-------
(b) National Toxicology Program Studies-
Male and female F344/N rats
Table 6-21 (cont'd). Summary of Data on the Genotoxicity of Ozonfe
Assay System in Which Ozone Was Tested Resultb Comments
Lung tumor induction in whole animals
(a) Strain A/J mice, Swiss-Webster mice, +/D Positive results marginal, not
Syrian Golden hamsters, Wistar rats statistically significant; experiments
not designed to determine whether a
concentration-response exists.
Ozone was tested at 0.05, 0.31, 0.4,
0.5, 0.8, and 1.0 ppm.
D No increased incidence of pulmonary
adenomas or carcinomas in rats
exposed to 0.12, 0.5, or 1.0 ppm O3
for 2 years or with 0.5 or 1.0 ppm
O3 for animals' lifetimes.
+/D No effect at 0.12 ppm O3, slight
increases in the total pulmonary
neoplasms at 0.5 and 1.0 ppm O3,
but they were not statistically
significant.
+/D Alveolar/bronchiolar carcinoma
incidence increased twofold at
0.5 ppm O3 (p = 0.05) and 1.0 ppm
O3 (p = 0.007); no increases in total
pulmonary neoplasms.
+ Fourfold increase in
alveolar/bronchiolar carcinoma at
1.0ppmO3(p = 0.053).
+ Fourfold increase in
alveolar/bronchiolar adenomas and
carcinomas at 1.0 ppm O3
(p = 0.02).
+ Threefold increase in
alveolar/bronchiolar adenomas at
1.0ppmO3(p = 0.02).
+ Twofold increase in
bronchiolar/alveolar adenomas and
carcinomas at 1.0 ppm O3, but not
statistically significant.
Male B6C3F, mice
(2-year study)
Male B6C3F,
(Lifetime studies)
Female B6C3F, mice
(2-year study)
Female B6C3F,
(Lifetime studies)
aSee Appendix A for abbreviations and acronyms.
bD = no effect; + = effect.
6-125
-------
artifactual because it bypasses all the host defenses that would normally be functioning to
protect the individual from the inhaled dose. Direct, in vitro O3 exposure of isolated cells in
tissue culture medium also results in chemical reactions between O3 and culture media to
generate chemical species that may not be produced in vivo. Therefore, for these reasons, the
relevance and predictive value of in vitro studies to human health are questionable The most
relevant data on the genotoxicity of O3 should therefore be obtained from in vivo studies.
The earlier studies in whole animal carcinogenesis bioassays must be considered
ambiguous at this time (Witschi, 1988, 1991). The NTP study utilized an inhalation model,
assayed the carcinogenicity of O3 in male and female F344/N rats and B6C3F, mice, and also
tested whether O3 could enhance the tumorigenicity of the tobacco-specific pulmonary
carcinogen NNK (National Toxicology Program, 1994; Boorman et al., 1994). This study
clearly showed that O3 was not carcinogenic in female and male rats at 0.12, 0.5, and 1.0 ppm
O3 (6 h/day, 5 days/week for 2 years) or at 0.5 and 1.0 ppm O3 (6 h/day, 5 days/week,
lifetime). Exposure to 0.5 ppm O3 did not enhance the carcinogenicity of NNK in male rats,
leading to the conclusion that O3 does not act as a co-carcinogen or tumor promoter in these
animals. In the male mice, O3 had equivocal effects at 0.5 and 1.0 ppm O3 in the 2-year and
lifetime inhalation studies. In the female mice, there was some evidence for the
carcinogenicity of O3 at 1.0 ppm (2.7-fold increase in total pulmonary neoplasms [p = 0.02] in
the 2-year study; and twofold increase in total pulmonary neoplasms [p = 0.1] in the lifetime
study; and 2.3-fold increase in total pulmonary neoplasms when the 2-year and lifetime study
were combined [p = 0.004]).
Therefore, the earlier negative animal carcinogenesis studies, the negative
carcinogenicity results in inhalation carcinogenesis studies in F344/N male and female rats, the
ambiguous data in male B6C3F, mice, and the weak carcinogenicity of O3 in female B6C3F,
mice indicate that O3 is carcinogenic only in female B6C3F, mice at high concentrations (1.0
ppm). The weak carcinogenicity of O3 in female mice, the weak/ambiguous results in male
mice, and the negative results in male and female F344/N rats point to, at best, a weak
carcinogenicity of O3 at very high concentrations.
6.3 Systemic Effects of Ozone
6.3.1 Introduction
Ozone has long been known to cause effects in organs and tissues outside the
respiratory tract. The mechanisms are not known, but it is quite unlikely that O3 itself enters
the circulation (Pryor, 1992). Another possibility is that transported reaction products cause
distant effects. Some effects may be secondary to effects on the lung (e.g., aversive behaviors
that may result from lung irritation). The relatively few systemic studies reported since the
last O3 criteria document (U.S. Environmental Protection Agency, 1986) are discussed below.
Some classes of effects (i.e., reproduction/development, endocrine system) were studied earlier
and, hence, are cited briefly here in the introduction.
No reproductive toxicity studies of O3 were found. Only two developmental studies
provided sufficient details in the report to determine the exposures used. The only effect
observed by Kavlock et al. (1979) in pregnant rats exposed to 0.44 to 1.97 ppm O3 for the
entire period of organogenesis or the three stages of gestation was an increased resorption of
fetuses in rats exposed to 1.49 ppm in midgestation; no terata were found. A follow-up study
revealed that pups from dams exposed to 1 ppm O3 during mid- or late gestation showed lower
6-126
-------
body weights 6 days after birth (Kavlock et al., 1980). A higher concentration (1.5 ppm)
delivered during late gestation permanently runted 14% of the male pups.
Studies on the effects of O3 on the endocrine system date to 1959. Generally, the
body of work indicates that O3 can affect the pituitary-thyroid-adrenal axis
(U.S. Environmental Protection Agency, 1986). For example, a 1-day exposure to 1 ppm
O3 decreased serum levels of thyroid-stimulating hormone, thyroid hormones, and protein-
bound iodine; prolactin levels increased (demons and Garcia, 1980a,b). Structural changes
occurring in the parathyroid glands after a 4- to 8-h exposure to 0.75 ppm O3 included
hyperplasia of chief cells, but circulating hormone levels were not measured (Atwal and
Wilson, 1974).
6.3.2 Central Nervous System and Behavioral Effects
Reports of headache, dizziness, and irritation of the nose, throat, and chest are
common complaints that are associated with O3 exposure in humans (see Chapter 7).
Laboratory animal studies have been performed that demonstrate behavioral effects over a wide
range of O3 concentrations (0.08 to 1.0 ppm) and suggest that these behavioral changes may be
analogous to the symptoms reported in humans. Although these behavioral changes may be
indicative of O3-induced symptoms, they are not indicative of neurotoxicity. Most of the
studies prior to 1986 indicated that behavior could be suppressed with O3 exposure. For
example, Murphy et al. (1964) and Tepper et al. (1982) showed that running-wheel behavior
was suppressed, and Peterson and Andrews (1963) and Tepper et al. (1983) showed that mice
would alter their behavior to avoid O3 exposure. Furthermore, Weiss et al. (1981) showed that
bar-pressing responses for food reinforcement were suppressed, but greater O3 concentrations
were required to decrease this behavior than the concentrations needed to decrease running-
wheel behavior.
Since 1986, several reports have extended the previous findings (Table 6-22).
Tepper et al. (1985) compared the effects of a 6-h exposure to O3 on the suppression of
running-wheel behavior in rats and mice. The study indicated that the lowest effective
concentration was about 0.12 ppm O3 in the rat and about 0.2 ppm in the mouse. It also was
observed that, with exposure to 0.5 ppm, recovery from O3 required at least 3 h. In a follow-
up study, Tepper et al. (1985) required mice to make a response that turned off the brief
delivery (60 s) of O3 at concentrations between 0.25 to 16 ppm. Mice learned to terminate
O3 exposures at 0.5 ppm. With each of three determinations of the concentration-response
curve, mice got better at terminating O3 exposure, rather than exhibiting an adaptation to
exposure. The authors suggest that mice may have learned to use the odor of O3 as a
conditioned stimulus to initiate termination of exposure instead of responding directly to the
irritant properties of O3.
Because free-access wheel-running behavior was suppressed at 0.12 ppm
O3 (Tepper et al., 1982), and lever pressing for food reinforcement was reduced only at
0.5 ppm (Weiss et al., 1981), a series of experiments was performed to evaluate the
behavioral determinants of the O3 response (Tepper and Weiss, 1986). Food deprivation and
response contingencies (having to perform a certain response to get a reward) were found to be
relatively unimportant determinants of behavior because rats that had to run rather than press a
lever to obtain food reinforcement showed behavioral suppression of running at 0.12 ppm.
However, in another experiment, suppression of lever pressing was shown to be
6-127
-------
Table 6-22. Effects of Ozone on Behavior1
NJ
CO
Ozone
Concentration
ppm
0.08
0.12
0.25
0.5
0.08
0.12
0.25
0.5
0.08
0.12
0.25
0.5
0.12-
1.5
0.1-
0.8
0.25-
16
0.4
1.2
0.5
2.0
Exposure
Dg/m 3 Duration
157 6h
980
157 6h
980
157 6h
980
235- 6 h
2,940
196- 7 days
1,568 continuous
490- 60 s
31,360 maximum
784 13 days
2,352 continuous
980 3h
3,920
Behavioral
Conditions
Free-access wheel
running
Wheel running for food
Lever pressing for
access to the running
wheel
Nose poke response for
food
Drinking, eating
Nose poking terminated
O3 exposure
Home cage behavior
Lever pressing to avoid
electric shock
Species, Sex
(Strain)
Age"
Rat, M
(Long-Evans)
10 weeks old
Mouse, M
(Swiss-Webster)
5 weeks old
Rat, M
(Long-Evans)
300 g
Rat, M
(Long-Evans)
300 g
Rat, M
(Long-Evans)
275 g
Mice, M
(ICR)
8-26 weeks old
Mouse, M
(Swiss-Webster)
30 g
Mice
Rat
(Wistar)
300 g
Observed Effect(s)
Mice less responsive than rats. Reduction in free-access wheel running at
approximately 0.12 ppm in rats and 0.2 ppm in mice. Recovery from exposure
to 0.5 ppm did not occur by 5 h postexposure in either mice or rats.
0. 12 ppm O3 decreased wheel running for food reinforcement.
0.12 ppm O3 decreased bar press for access to the running wheel. Two of the
four animals were affected at 0.08 ppm.
0.5 ppm O3 decreased nose poking for food reinforcement. Effects were
enhanced postexposure.
Drinking, food consumption, and body weight initially decreased, but adapted
with continued exposure, starting at 0.2 ppm.
At 0.5 ppm, mice learned to terminate O3 exposure.
During first hour, rearing, grooming, sniffing, and social interactions
increased, crossings and wall climbing decreased. These behaviors did not
adapt with continued exposure.
Suppression of lever pressing began after 45 min of 2.0-ppm exposure and after
90 min of 0.5- or 1.0-ppm exposures.
Reference
Tepper et al. (1985)
Tepper and Weiss (1986)
Tepper and Weiss (1986)
Tepper and Weiss (1986)
Umezu et al. (1993)
Tepper and Wood (1985)
Musi et al. (1994)
Ichikawa et al. (1988)
'See Appendix A for abbreviations and acronyms.
''Age or body weight at start of exposure.
-------
equally sensitive to O3 exposure when pressing a lever allowed rats to have access to a running
wheel. The authors concluded that increased physical activity, either used as the response to
obtain reward, or as the reward, was an important behavioral variable in determining
sensitivity to O3 exposure. Ichikawa et al. (1988) demonstrated that behavior (lever pressing)
maintained by the avoidance of electric shock, was even less sensitive to O3 exposure (3 h, 1.0
ppm) than behaviors maintained by food reinforcement, as described above. Furthermore, the
animals recovered quickly after O3 exposure was terminated (60 to 120 min).
In mice exposed to O3 continuously for 13 days (0.4 to 1.2 ppm), both increases and
decreases in measured behaviors were observed (Musi et al., 1994). During the first hour of
exposure to 0.8 or 1.2 but not 0.4 ppm O3, increases in rearing, grooming, sniffing, and social
interactions were observed, but locomotion and bar holding declined. With continued
exposure (measurements on Days 3,7, and 10), grooming and rearing were still increased but
crossings and wall climbing remained depressed. The affected behaviors did not show
adaptation. However, drinking, food consumption, and body weight were initially depressed,
but abated with continued exposure, a finding previously reported in mice at O3 concentrations
as low as 0.2 ppm (Umezu et al., 1993).
In summary, the behavioral data indicate that transient changes in behavior occur in
rodent models that are dependent on a complex interaction of factors such as (1) the type of
behavior being measured, with some behaviors increased and others suppressed; (2) the factors
motivating that behavior (differences in reinforcement); and (3) the sensitivity of the particular
behavior (e.g., active behaviors are more affected than more sedentary behaviors).
6.3.3 Cardiovascular Effects
Several reports have demonstrated that O3 exposure causes dramatic effects to the
cardiovascular system in the rat (Table 6-23). Uchiyama et al. (1986) initially reported that
heart rate (HR) and mean arterial blood pressure (MAP) were decreased by 53 and 29%,
respectively, during a 3-h exposure to 1.0 ppm O3. Arrhythmias, including atrioventricular
block and premature atrial contractions, also were observed frequently. The effects appeared
to be age- but not sex-dependent, with 11-week-old rats showing a greater response than did
8- or 4-week-old rats. Yokoyama et al. (1989b) showed that recovery from the effects of the
3-h, 1.0-ppm O3 exposure was not complete by 5 h and that, with three consecutive daily
exposures, both the HR and MAP responses were attenuated. Further investigations by the
same group of authors (Uchiyama and Yokoyama, 1989) showed that, with exposures to
0.5 ppm O3 for 6 h, HR and MAP decreased by 32 and 18%, respectively. A 4-week
continuous exposure to 0.2 ppm initially resulted in a 12% decrease in HR, but this response
was attenuated on Day 2 and was almost eliminated by Day 3. No further effects were
observed during the rest of the 4-week exposure period. When these same animals were
subsequently challenged with 0.8 ppm O3 for 1.5 h, they also had an attenuated response when
compared to rats that were O3 naive. Additionally, some rats were instilled intratracheally
with elastase to create an animal model of emphysema. This pretreatment, however, did not
affect the outcome of either the HR or MAP responses to O3 in any of the experiments, except
in the 0.8-ppm challenge experiment. In this experiment, elastase-treated, O3-exposed rats
challenged with O3 had a similar response to O3 challenge as did O3-naive rats, suggesting that
the elastase treatment affected the ability of the rats to develop an adaptive lung response. In
contrast, Tepper et al. (1990) did not observe an alteration in
6-129
-------
Table 6-23. Effects of Ozone on the Cardiovascular Systerfl
ppm
0.1
0.2
0.2
0.5
1.0
0.25
1.0
0.5
1.0
0.5
1.0
1.0
1.0
Ozone
Concentration
Dg/m
196
392
392
980
1,960
490
1,960
980
1,960
980
1,960
1,960
1,960
Exposure
Duration and
Conditions
5 days
continuous
4 weeks
continuous
6h
3h
2h
18-20 DC
30-32 DC
6h
3h
6h
3h,
3 days
3h
135 min
Species, Sex
(Strain)
Age1'
Rat, M
(Wistar)
8 weeks old
Rat, M
(Wistar)
13 weeks old
Rat, M
(F344)
13-16 weeks old
Rat, M
(Wistar)
10 weeks old
Rat, M
(Wistar)
10-11 weeks old
Rat, M, F
(Wistar)
4, 8, and 11 weeks old
Rat, M
(F344)
90 days old
Observed Effect(s) Reference
0.1 ppm O3 caused bradyarrhythmia up to 3 days of exposure; broadycardia occurred at 0.2 Arito et al. (1990)
ppm during first 2 days of exposure. No effects on sleep-wakefulness patterns.
At 0.2 ppm a 12% decrease in HR; response attenuated by 3 days. At 0.5 ppm, HR and Uchiyama and Yokoyama (1989)
MAP decreased by 32 and 18%, respectively.
0.37 ppm O3 caused bradycardia and bradyarrhythmia, ambient temperature of 30-32 DC Watkinson et al. (1993)
blocked response.
0.5 and 1.0 ppm O3 caused bradycardia and bradyarrhythmia. 1.0-ppm response was Arito et al. (1992)
partially blocked by atropine.
1.0 ppm O3 caused bradycardia, bradyarrhythmia, and decreased MAP. The response to 1.0 Yokoyama et al. (1989b)
ppm lasted > 5 h postexposure and was attenuated with 3 consecutive daily exposures.
1.0 ppm O3 caused bradycardia, bradyarrhythmia, and decreased MAP. Older animals (1 1 Uchiyama et al. (1986)
weeks) were more affected than younger ones. No sex-related differences were noted.
Ventilation stimulated with CO2. No effect on mean blood pressure. Tepper et al. (1990)
"See Appendix A for abbreviations and acronyms.
''Age or body weight at start of exposure.
-------
blood pressure of rats exposed to 1.0 ppm O3 for 135 min, even though their ventilation was
increased by CO2.
Arito et al. (1990) demonstrated bradycardic responses at 0.2 ppm O3 during the
first 2 days of a continuous 5-day exposure; bradyarrhythmia occurred during the first 3 days
of a 0.1-ppm exposure. Simultaneously, these authors measured the sleep/wakefulness of the
rats during exposure and found that more bradyarrhythmias occurred during wakefulness than
during slow-wave or paradoxical sleep. Sleep/wakefulness patterns were not altered by this
O3 exposure. At high O3 concentrations (1 ppm for 3 h), wakefulness and paradoxical sleep
were suppressed, the amplitude of the electroencephalogram (EEG) was lowered, and slow-
wave sleep was increased (Arito et al., 1992). These EEG changes appear to be temporally
associated with the decrease in behavioral activity previously discussed (Tepper et al., 1982).
Atropine sulfate blocked the suppression of wakefulness and bradycardia in a concentration-
related manner and decreased slow-wave sleep, suggesting that some of the O3 effects are
parasympathetically mediated. The effects of O3 on paradoxical sleep and the EEG amplitude
were not affected by atropine administration. Watkinson et al. (1993) extended these findings
by showing that the core temperature of rats was also reduced when HR fell at O3 exposure
concentrations between 0.37 and 1.0 ppm (2 h). Increasing ambient temperature to 30 to
32 DC attenuated the 1.0 ppm O3-induced reduction in HR and core temperature.
In an attempt to synthesize the results from these studies, Watkinson and Gordon
(1993) questioned the relevance of these parameters in the rat as compared to the human. Rats
have different thermoregulatory responses than humans and typically respond to toxic insult by
lowering core temperature. This response has been shown to increase survival value
(Watkinson et al., 1989). Similar changes in core temperature and HR have not been reported
in humans. This may be because of the large, and thus stable, thermal mass of humans, or,
alternatively, these effects have not been observed because they were not measured, and
because most O3 exposure experiments are done using exercise, which may mask these
responses. In support of this latter idea, Coleridge et al. (1993) reported that stimulation of
bronchial C-fibers produces bradycardia. Ozone preferentially stimulates bronchial C-fibers
and, as a result, induces bradycardia and tachypnea in the anesthetized, open-chest dog model.
Furthermore, the tachypnea produced by O3 exposure is inhibited by atropine administration
(the effect on HR was not reported).
6.3.4 Hematological and Serum Chemistry Effects
Hematological effects reported in laboratory animals and humans after inhalation of
O3 indicate that the gas or, more likely, some reaction product can cross the blood-gas barrier.
The effects of in vivo O3 exposure in animals were summarized in the previous O3 criteria
document (U.S. Environmental Protection Agency, 1986). The hematologic parameters most
frequently used to evaluate O3 toxicity were morphologic and biochemical effects on
erythrocytes (RBCs). These studies reported alterations in RBC morphology, increased RBC
fragility, increased hemolysis, and decreased survival. The biochemical studies reported
variable results, depending on the O3 exposure concentration and the RBC enzyme under
investigation.
More recent studies have stressed serum effects of O3 exposure (Table 6-24).
Exposure of rats for 2 h to 0.1 ppm O3 increased plasma creatinine kinase activity, whereas no
such effect was observed when exposure was to 0.05 and 0.25 ppm O3 (Veninga and
6-131
-------
Table 6-24. Hematology and Serum Chemistry Effects
Ozone
Concentration
ppm
0.05
0.1
0.25
0.1
0.2
0.4
0.6
0.4
0.8
1.0
NJ
1.0
1.0
Exposure
Qg/m 3 Duration
100 2h
200
500
196 3h
392
784
1,176
784 Continuous
for 14 days
1,568 18 h
1,960 1 h
1,960 1 h
1,960 23 h/day for
2 weeks
1,960 4h
Species, Sex
(Strain)
Age1'
Rat, M
(Wistar)
200 g
Rabbit, F
(NZW)
2.5-3.5 years old
Rat, M
(Wistar)
20 weeks
Rat, M
(F344)
Guinea Pig, M
(Hartley)
250-300 g
Guinea Pig, M
(Hartley)
250-300 g
Rat, M
(CD)
400 ± 25 g
Mouse, M
(CD-I)
8 weeks old
Observed Effect(s)
Increased plasma creatine kinase activity at 0.1 but not 0.05 and 0.25 ppm.
No change in plasma retinol, ascorbic acid, and D-tocopherol concentrations.
Decrease in serum retinol concentration.
Decrease in plasma lactic dehydrogenase isoenzyme activity.
Increases in plasma concentrations of TXB,, 6-keto-PGF]D, and PGE, .
Increases in plasma concentrations of TXR,, 6-keto-PGF,D, and PGE, .
Heat-inactivated plasma increases DNA synthesis by lung fibroblasts and
pneumocytes.
Inhibition of RBC deformability.
Reference
Veninga and Fidler (1986)
Canada et al. (1987)
Takahashi et al. (1990)
Nachtman et al. (1988)
Miller et al. (1987)
Miller et al. (1988)
Tanswell et al. (1989, 1990)
Tanswell (1989)
Morgan et al. (1988)
'See Appendix A for abbreviations and acronyms.
''Age or body weight at start of exposure.
-------
Fidler, 1986). Decreased serum retinol concentrations were observed following continuous
exposure of rats for 14 days to 0.4 ppm O3 (Takahashi et al., 1990), but no changes in plasma
retinol, ascorbic acid, and D-tocopherol were observed following exposure of rabbits for 3 h to
O3 ranging from 0.1 to 0.6 ppm (Canada et al., 1987). In similar studies, a decrease in plasma
lactic dehydrogenase isoenzyme activity also was observed following exposure of rats for 18 h
to 0.8 ppm O3 (Nachtman et al., 1988).
Miller et al. (1987, 1988) investigated the effect of a 1-h exposure of guinea pigs to
1.0 ppm O3 on plasma eicosanoid levels and observed increases in TXE^, 6-keto-PGFin, and
PGE,. These data suggest that some of the systemic effects of O3 exposure, such as
impairment of peritoneal AM phagocytosis (Canning et al., 1991), may be mediated by the
immunosuppressive effects of the prostanoids (Oropeza-Rendon et al., 1979). Heat-inactivated
plasma from rats exposed for 23 h/day for 2 weeks to 1.0 ppm O3 also increases DNA
synthesis by lung fibroblasts (Tanswell et al., 1989) and lung pneumocytes (Tanswell et al.,
1990).
6.3.5 Other Systemic Effects
Previous studies suggest that O3 has effects on the xenobiotic metabolism of the
liver (U.S. Environmental Protection Agency, 1986). This effect has been observed in mice,
rats, and hamsters as a prolongation of pentobarbital sleeping time (Graham et al., 1981). The
effect appears to be sex dependent, with females having greater responses than males. Canada
and Calabrese (1985) performed a similar experiment in both young (3- to 4-mo-old) and older
(2-year-old) rabbits exposed for 3.75 h/day to 0.3 ppm O3 for 5 consecutive days. They
observed significant prolongation of the elimination of theophylline in older rabbits, but not in
young rabbits, and the effect was more pronounced in females than in males. In a follow-up
study, Canada et al. (1986) could not demonstrate increased pentobarbital sleeping in young
(2.5-mo-old) mice or rats of comparable age to the study by Graham et al. (1981). However,
effects were observed in older (18-mo-old) female mice and rats. Two other studies (Heng
et al., 1987; Zidenberg-Cherr et al., 1991) from the same group of investigators indicate that
liver antioxidant enzymes (Cu/Zn- and Mn-SOD and GSHPx) are decreased commensurate
with the increase in these enzymes that is observed in the lung.
6.3.6 Summary
Several reports recently have appeared that extend previous observations in
laboratory animals that indicate that ambient levels of O3 can affect animal behavior. These
effects are interpreted as analogous to O3-induced symptoms in humans, rather than as
evidence of neurotoxicity. The behavioral changes are transient but may persist several hours
after acute exposure. Different types of behaviors appear to be variably sensitive to O3
exposure, with active behaviors showing suppression at lower O3 concentrations than do more
sedentary behaviors or behaviors maintained by electric shock (Ichikawa et al., 1988; Tepper
et al., 1985; Tepper and Weiss, 1986). For example, a 6-h exposure of rats to 0.12 ppm
suppressed running-wheel behavior (Tepper et al., 1985). Furthermore, animals will respond
to terminate a 1-min exposure to 0.5 ppm O3, thus directly implicating the irritant properties of
O3 (Tepper and Wood, 1985). It appears that with additional training, animals can learn to
terminate exposure using conditioned stimuli rather than relying directly on the aversive
properties of O3 (Tepper et al., 1985).
6-133
-------
Ozone has been found to decrease HR, MAP, and core temperature profoundly in
rats (Watkinson et al., 1993; Arito et al., 1990; Uchiyama and Yokoyama, 1989). During
exposure, arrhythmias frequently occur. After a 3-h exposure to 1.0 ppm O3, these effects
appear to occur more in adult rats (11 weeks old) than in younger animals (4 and 8 weeks old),
especially when the rats were awake (as measured by EEG) during the exposure (Uchiyama
et al., 1986). The lowest exposures causing bradycardia in rats was 0.2 ppm for 48 h; 0.1
ppm for 24 h caused bradyarrhythmia (Arito et al., 1990). Similar effects have not been
observed in humans or other species. In part, this may be because they have not been
systemically examined or that human studies have been carried out during concurrent exercise,
which may mask these effects. More likely, these effects represent species differences related
to the magnitude and localization of reflex responses and differences in thermal mass.
6.4 Interactions of Ozone with Other Co-occurring
Pollutants
6.4.1 Introduction
Most of the toxicological data for O3 are derived from studies using O3 alone.
However, it is also important to evaluate responses to inhalation of typical pollutant
combinations because ambient exposures involve mixtures. Such mixtures provide a basis for
toxicological interactions, whereby combinations of chemicals may behave differently than
would be expected from consideration of the action of each separate constituent. This section
discusses toxicological studies of pollutant mixtures in which O3 is one component.
Discussions of many of these studies addressing the effect of O3 alone on various organs or
systems appear elsewhere in this chapter.
Evaluating the role of O3 in observed responses to inhaled mixtures is not easy.
In spite of the myriad of interpretative difficulties, it is essential to attempt to understand the
potential for interactions because O3 does not exist alone. One of the problems involves
definitions of terms, and the study of toxicant interaction is complicated by the dilemma of
attempting to characterize the effects from exposure to two or more chemicals. Most studies
have employed a statistical definition, but this merely provides a description, and tells nothing
about the mechanism of any interaction. But, in many cases, the mechanism of action of the
individual components may not be understood fully, and information concerning the types of
interactions may provide useful beginnings for studying mechanisms of action of the mixture
components. Furthermore, any conclusion of interaction is highly dependent on the specific
type of model used. Because the purpose of this chapter is to provide a toxicologic
background for effects of O3 in terms of public health significance, interaction will be defined
as a departure from the additivity model (i.e., interaction is considered to occur when the
response to a mixture is significantly different from the sum of the responses to the individual
components). A less than additive interaction is antagonism, whereas synergism is an
interaction that is more than additive. A subclassification of synergism, termed potentiation, is
often used to describe an interaction in which response to a mixture is greater than the sum of
the responses to the individual components, but where only one component produced a
response different from control when administered alone. In many instances, however,
potentiation and synergism have been used interchangeably. Although some synergistic
interactions actually may serve to stimulate repair processes, or otherwise reduce the harmful
6-134
-------
effects of O3, and some antagonistic interactions eventually may increase the risk of disease
development, synergism, as currently used, generally implies greater risk, and antagonism
implies lesser risk. However, such assumptions eventually may be proven to be invalid in
some instances. Also, interactions with the large number of natural air pollutants, such as
microbes, spores, and dusts, that can produce considerable responses alone are not included in
this section.
In most cases, the interaction of O3 with other pollutants has been studied using
mixtures that contained only one other copollutant (i.e., simple or binary mixtures). In such
studies, the role played by each pollutant in eliciting measured responses can be elucidated
with the appropriate experimental design, but most of the database involves exposures to the
mixture and O3 only, with no exposure to the copollutant alone. Although the O3 concentration
may have been varied among exposure groups or was present in one group and not in another
(so its relative influence could be assessed to some extent), it cannot be determined in such
cases whether the response to the mixture involved actual interaction or was merely additive.
The ambient atmosphere in most environments is generally a mixture of a number
of pollutants, and assessing effects of such multicomponent atmospheres may serve to provide
some indication of biological responses under conditions that better mimic ambient exposure.
However, very few studies have used realistic combinations of pollutant concentrations when
assessing interaction.
The ability to discern the contribution of O3 to observed responses becomes even
more difficult when such complex mixtures are studied. Even when binary mixtures are used,
they often do not mimic the ambient pattern (e.g., NO2 levels peak before O3 levels do) or
ambient concentrations (as absolute values or as ratios). Rarely are concentration-response
mixture studies performed. This raises the possibility that an unrealistic experimental design
may lead to masking the effect of a copollutant or to identifying a response that may not occur
in the real world.
Another problem in assessing responses to mixtures involves the statistical basis for
the conclusion of significant interaction. For example, a number of studies determined
interaction by comparison of the response from exposure to only one component of the mixture
with that from exposure to the complete mixture. On the other hand, some studies used
statistical approaches specifically designed to indicate interactions. As another example, it
may be relatively straightforward to study interactions when one exposure concentration of
each of two pollutants is used, but it becomes much more difficult when there are multiple
concentrations used, and even more difficult still when more than two pollutants are involved.
Because variable criteria for conclusions of interaction have been used, the available database
is one in which the statistical significance for determination of interaction varies in terms of its
robustness.
6.4.2 Simple (Binary) Mixtures Containing Ozone
Tables 6-25 and 6-26 outline studies performed since publication of the last
O3 criteria document (U.S. Environmental Protection Agency, 1986) in which experimental
animals were exposed to atmospheres containing O3 with only one other copollutant. These
tables provide the experimental details for the discussion that follows.
6-135
-------
Table 6-25. Toxicological Interactions of Ozone and Nitrogen Dioxid§
Concentration1'
ppm
0.05
0.04
0.05
0.4
0.05
0.4
0.1
0.5
1.2
4.0
0.1
0.3
1.2
2.5
0.05
0.1
0.5
1.0
0.05
0.1
0.05
0.1
0.1
0.3
Dg/m
98
75
98
752
98
752
196
980
2,256
7,520
196
588
2,256
4,700
98
196
940
1,880
98
196
94
188
196
564
Polftitant
03
NO,
O3
NO,
03
NO,
O3 (baseline)
O3 (peak)
NO, (baseline)
NO, (peak)
O3 (baseline)
O3 (peak)
NO, (baseline)
NO, (peak)
O3 (baseline)
O3 (peak)
NO, (baseline)
NO, (peak)
O3 (baseline)
O3 (peak)
NO, (baseline)
NO, (peak)
O3
NO,
Exposure Duration
NO,: 24 h/day;
O3: Intermittent during
hours 9-19/day following
sine curve from 0-0. 1 ppm
(0.05 avg); total duration:
5-22 mo
O3: Concentration ranged
from 0 to 0. 1 with sine
curve over 9-19h
(0.05 avg); NO,:
24 h/day; both 13 mo,
1 1 mo recovery
15 day, continuous
exposure to basal level;
peaks: 1 h, twice daily,
5 days/week beginning
after 64 h of continuous
exposure
NO,: 24 h/day;
O3: 8 h/day;
1, 3, 6, 18 mo
Species, Sex
(Strain)
Agec
Rat, M
(Wistar)
7 weeks old
Rat, M
(Wistar)
6 weeks old
Mouse, F
(CD-I)
4-6 weeks old
Rat, M
(F344)
5 weeks old
Endpoints
Lung protein content; lung
lipid peroxides; antioxidant
enzymes (G6PD, 6PGD, GR,
GST, GSHPx, SOD)
Development of lung tumors
from exposure (ingestion) of
carcinogen,
^V-M$(2-hydroxypropyl)
nitrosamine prior to O3 and
NO,
Bacterial infectivity
(to Streptococcus
zooepidemicus given after
pollutant exposure)
Histopathology (LM, TEM)
Time of
Endpoint Response to
Measurement Mixture
0 PE Protein: no change;
peroxide: increase
between 5 and 9 mo,
return to control at
> 9 mo (greater effect
with 0.4 ppm NO,);
enzymes: no change.
0 PE Increased tumor
incidence (compared to
air-exposed control).
Bacterial No effect at low level;
challenge given increased mortality at
0 or 18 h PE other levels.
0 PE Connective tissue
edema, Type 2 cell
hypertrophy and enlarged
lamellar
bodies.
Interaction
Reference
Protein: none (no effect Sagai and Ichinose (1991)
of O3 or NO, alone);
peroxide: synergism
(no change with O3 or
NO, alone for 9 mo);
enzyme: none (no
effect of O3 or NO,
alone).
Suggested synergism:
no increase with O3
alone (NO, alone not
done).
Synergism: at 0.05 O3
+ 0.5 (1.0) NO,, 0.1 O3
+ 1.2 (2.5) NO^;
marginal synergism at
0.1 O3 + 1.2 (4) NO,.
Both O3 and NO,
increased mortality at
two highest levels; only
NO, increased at two
lowest levels.
Changes more marked
than with O3 alone,
NO, affected response
to O3 (no quantitation
performed).
Ichinose and Sagai (1992)
Graham et al. (1987)
Terada et al. (1986)
-------
Table 6-25 (cont'd). Toxicological Interactions of Ozone and Nitrogen Dioxidfe
Concentration1'
ppm
0.15
0.35
0.2
4.0
0.2
3.6
0.4
7.2
0.6
10.8
0.8
14.4
0.2
3.6
0.4
7.2
0.6
10.8
0.2
3.6
0.4
7.2
0.6
10.8
0.8
14.4
0.3
1.2
Dg/m
294
564
392
7,520
392
6,768
784
13,536
1,176
20,304
1,568
27,072
392
6,768
784
13,536
1,176
20,304
392
6,768
784
13,536
1,176
20,304
1,568
27,072
588
2,256
Pollutant
03
NO,
03
NO,
03
NO,
03
NO,
03
NO,
03
NO,
03
NO,
03~
NO,
03~
NO,
03
NO,
03
NO,
03
NO,
03
NO,
03
NO,
Exposure Duration
7 h/day, 5 days/week
for 12 weeks
Continuous 1-2 mo
24 h/day for 3 days
12 h/day for 3 days
8 h/day for 3 days
6 h/day for 3 days
6 h/day for 3 days
24 h/day for 3 days
12 h/day for 3 days
8 h/day for 3 days
6 h/day for 3 days
Continuous for 3 days
Species, Sex
(Strain)
Agec
Mouse, M
(C57B1/6J)
5 weeks
Rat, M
(Wistar)
22 weeks old
Rat, M
(S-D)
10-12 weeks
old
Rat, M
(S-D)
10-12 weeks
old
Rat, M
(S-D)
250-275 g
Rat, M
(S-D)
3 mo old
Time of
Endpoint
Endpoints Measurement
Colonization of lung by Inject
melanoma cells (injected melanoma
after pollutant exposure) cells 0 PE
and sacrifice
3 weeks PE
Pulmonary xenobiotic 0 PE
metabolism, lung protein
(homogenate)
Protein (lavage), lavaged cells 0 PE
(epithelial PMN)
Protein (lavage), lavaged cell 0 PE
counts (epithelial, PMN,
AM), DNA content of cell
pellet
Airway labeling index 4 days PE
Lung enzymes (G6PD, 6PGD, 0 PE
ICD, GSHPx, GR, DR, GOT,
NADPH-CR)
Response to
Mixture
Increase in number of
colonies/lung
(compared to air control).
No effect on protein
content, increase in
selected enzymes.
Increased protein and
cells.
Increased protein and
cells, depending on
concentration.
Increased index in
peripheral airways (TBs
opening into ADs) and
large airways at three
highest doses, increased
alveolar index
at 0.8 + 14.4 ppm only.
Increased activity.
Interaction
Not specified: no
change with O3, but
previous study showed
effect with NO,.
Suggested antagonism
for xenobiotic enzymes:
O3 induced increase in
selected enzymes is
lowered by addition of
NO,.
Synergism: protein at
0.6 and 0.8 ppm O3
mixtures, lavaged cells
at 0.4-0.8 ppm O3;
others additive.
Synergism: cell counts
at DO. 4 ppm O3 mixture,
protein additive.
Synergism for
peripheral airways at
highest dose and large
airways at three highest
doses only.
Synergism: 6PGD,
ICD, GR, SOD;
additive: GP, DR;
others: effect same as
O3 only.
Reference
Richters (1988)
Takahashi and Miura
(1989)
Gelzleichter et al. (1992a)
Gelzleichter et al. (1992b)
Rajinietal. (1993)
Lee et al. (1990)
-------
Table 6-25 (cont'd). Toxicological Interactions of Ozone and Nitrogen Dioxidfe
u>
CO
Concentration1'
ppm Qg/m
0.3 588
3.0 5,640
0.3 588
3.0 5,640
0.4 784
0.4 752
0.4 784
0.4 752
0.4 784
0.4 752
Species, Sex
(Strain)
Polltitant Exposure Duration Agec
O3 2 h Rabbit, M
NO, Nose-only (NZW)
4.5 mo old
O3 2 h/day for 14 days Rabbit, M
NO2 Nose-only (NZW)
4.5 mo old
O3 Continuous for 2 weeks Mouse, M
NO2 (ICR);
Hamster, M
(Golden);
Rat, M
(Wistar);
Guinea pig, M
(Hartley)
all 10 weeks
old
O3 24 h/day for 2 weeks Mouse, M
NO2 (ICR);
Hamster, M
(Golden);
Rat, M
(Wistar);
Guinea pig, M
(Hartley)
all 10 weeks
old
O3 Continuous for 2 weeks Rat, M
NO2 (Wistar)
10 weeks old;
Guinea pig, M
(Hartley)
10 weeks old
Endpoints
Pulmonary eicosanoids
(lavage) (PGE,, PGF^,
6-keto-PGF]D, TXB,, LTB4)
Pulmonary eicosanoids
(lavage) (PGE,, PGE^,
6-keto-PGF,n, TXB2)
Lung lipid peroxides,
antioxidant content,
phospholipids, and fatty acids
Lung lipid peroxides,
antioxidant enzymes, total
protein (homogenate)
Lipid peroxides, lung
antioxidants, and antioxidant
enzymes
Time of
Endpoint Response to
Measurement Mixture
0 or 24 h PE Increases in PGE ,
PGFa, and TXB,
immediately
PE (compared to air
control).
0 PE after Decrease in PGE,
7 or 14 (compared to air
exposures, or control) after 7 and
24 h PE after 14 days and 24 h PE;
14 exposures 6-keto-PGF,n decreased
24 h postexposure.
0 PE Variable increases to no
effect, depending on
species.
0 PE Increases, which were
species dependent.
0 PE Increased peroxides in
guinea pig but not rat;
increased antioxidants in
rat but not guinea pig;
enzymes increased or
decreased in guinea pigs,
increased to no change in
rat.
Interaction
Synergism: PGE,,
PGF,n; TXB, effect
similar to O3 alone.
None: effects additive
or similar to NO, alone.
Not determinable:
compared mixture vs.
air control, no measure
of single pollutants
performed.
Not determinable:
compared mixtures vs.
air control, no measure
of single pollutants
performed.
Synergism for some
endpoints, additive to
no interaction for
others; species
dependent.
Reference
Schlesinger et al. (1990)
Schlesinger et al. (1991)
Sagai et al. (1987)
Ichinose et al. (1988)
Ichinose and Sagai (1989)
-------
Table 6-25 (cont'd). Toxicological Interactions of Ozone and Nitrogen Dioxidfe
u>
Concentration1'
ppm
0.4
0.4
0.45
4.8
0.6
2.5
0.35
0.6
0.6
2.5
0.8
4.0
0.8
14.4
Dg/m
784
752
882
7,520
1,176
4,700
686
1,128
1,176
4,700
1,568
7,520
1,568
27,072
Pollutant
03
NO,
03
NO,
03
NO,
03~
NO,
03
NO,
03
NO,
03
NO,
Exposure Duration
Continuous for 2 weeks
8 h/day for 7 days
4 h (rest)
3 h (exercise)
2h
(rest and exercise)
Continuous for 3-56 days
6 h/day for 3 days
(concurrent) or sequential
O3 pre-NO,; NO, pre-O3; 6
h each
Species, Sex
(Strain)
Agec
Rat, M
(Wistar)
10 weeks old;
Guinea pig, M
(Hartley)
10 weeks old
Mouse, M
(Swiss
Webster)
2 mo old
Rat, M
(S-D)
7 weeks old
Rat, M
(S-D)
47-52 days
old
Mouse, M
(BALB/c)
8-10 weeks
old
Rat, M
(S-D)
10-12 weeks
old
Endpoints
Lipid peroxides; lung
antioxidants, and antioxidant
enzymes
Lung protein, DNA,
sulfhydryl and nonsulfhydryl
content; GR, GST, G6PD,
6PGD, ICD activities
Parenchymal histopathology
Epithelial permeability
(trachea!, bronchoalveolar)
(measured 1 and 24 h PE)
Antibody response to T-cell
dependent and independent
antigens in spleen
Protein (lavage); lavaged
cell counts (epithelial, PMN,
AM), DNA content of cell
pellet
Time of
Endpoint Response to
Measurement Mixture
0 PE Increased peroxides in
guinea pig but not rat;
increased antioxidants in
rat but not guinea pig;
enzymes increased or
decreased in guinea pigs,
increased to no change in
rat.
0 PE No change in protein or
DNA; increase in activity
of ICD, G6PD, 6PGD.
2 days PE Increased focal lesions.
0, 1, 2 days PERest: increased
bronchoalveolar
Interaction
Synergism for some
endpoints, additive to
no interaction for
others; species
dependent.
Synergism: enzyme
activity.
Synergism: ascribed to
production of HNO3 in
exposure atmosphere.
Enhanced magnitude
and duration of response
Reference
Ichinose and Sagai (1989)
Mustafa et al. (1985)
Mautzetal. (1988)
Bhalla et al. (1987)
permeability at 1 and 24 h (suggested potentiation).
PE; exercise: increased
bronchoalveolar
permeability at 1 and 24 h
PE (effects greater than
with O3 alone, no effect
of NO,).
0 PE Inconsistent pattern of
increases and decreases
of lung weight, thymus
weight, or plaque
formation.
0 PE Increased protein for
concurrent or sequential,
increase in cell counts for
concurrent.
Most responses similar
to O3 only; mixture
affected some time
points not affected by
O3 alone: implied
nonadditive interaction,
but specifics not
determinable.
Synergism: protein
and cell counts for
concurrent, protein
additive or antagonistic
for sequential.
Fujimaki (1989)
Gelzleichter et al. (1992b)
-------
Table 6-25 (cont'd). Toxicological Interactions of Ozone and Nitrogen Dioxidfe
Concentration1'
ppm
0.8
14.4
Dg/m
1,568
27,072
Polltitant Exposure Duration
O3 6 h/day for 45-79 days
NO,
Species, Sex
(Strain)
Agec
Rat, M
(S-D)
10/12 weeks
old
Endpoints
Various biochemical and
histological endpoints
Time of
Endpoint
Measurement
OPE
Response to
Mixture
Increased lung DNA,
protein, collagen, elastin;
some deaths with mixture
only at D55 days;
decreased
hydroxypyridinium.
Interaction
Suggested synergism for Last et
hydroxyproline and
hydroxypyridinium.
Reference
al. (1993b)
"See Appendix A for abbreviations and acronyms.
''Grouped by pollutant mixture.
°Age or body weight at start of exposure.
-------
Table 6-26. Toxicological Interactions to Binary Mixtures of Ozone
with Acids and Other Pollutants'
Concentration1'
ppm
0.1
0.1
0.3
0.6
0.12-
0.64
0.12-
0.64
0.12-
0.64
0.15
Dg/m
196
125
196
588
1,176
50
75
125
235-
1,254
40-
1,000
235-
1,254
40-
1,000
235-
1,254
40-
1,000
294
300
Pollutant
03
H2SO4 (0.3 Dm)
03
03
03
H,SO4 (0.3 Dm)
H2SO4 (0.3 Dm)
H2SO4 (0.3 Dm)
03
H2SO4
03
H2SO4
03
H2SO4
03
H2SO4 (0.09 Dm)
Exposure Duration
2 h/day, 5 days/ week
for up to 1 year
Nose-only
3h
Nose-only
6 h for 7 days
(23.5 h/day)
23.5 h/day for 5-9 days
23.5 h/day for 7 days
1 h to H2SO4, 2 h
rest, then 1 h
toO3
Head-only (acid),
whole-body (O3)
Species, Sex
(Strain)
Agec
Rabbit, M
(NZW)
4.5 mo old
Rabbit, M
(NZW)
4.5 mo old
Rat, M
(S-D)
250-300 g
Rat, M
(S-D)
250-300 g
Rat, M
(S-D)
250-300 g
Guinea pig, M
(Hartley)
260-325 g
Endpoints
Tracheobronchial
mucociliary transport,
bronchial tree epithelial
secretory cell numbers
Lavage cell counts; lavage
LDH, PGE,, PGF^; AM
phagocytosis; superoxide
production; TNF activity
Lavageable protein
Lung tissue protein
Rate of collagen synthesis
Pulmonary function
Time of Endpoint Response to
Measurement Mixture
Mucociliary Normal to accelerated
transport during clearance, increase in
exposure; secretory secretory cell numbers at
cells 3 days after 4, early time points
8, and 12 mo of (4-mo exposure).
exposure
0 PE No effects on lavage cell
counts or LDH, PGE,,
PGF-j] or increase or
decrease in TNF and
phagocytosis depending
on exposure
concentration; no change
in superoxide.
0 PE Increase (compared to air
control).
0 PE Increase (compared to air
control).
Interaction
Clearance: No interaction;
secretory cell numbers:
synergism at 4 mo,
antagonism at 8 and 12 mo.
Antagonism:
Phagocytosis, at all
combinations; antagonism:
Superoxide at 0. 1 and
0.3 ppm O3 and 75 and
125 Dg/m3 H2SO4;
synergism: TNF at
125 Dg/m3 H2SO4 and
0.3 and 0.6 ppm O3.
Synergism at DlOO Dg/m3
H2SO4 and 0.2 ppm O3 for
3 days.
Synergism at 1 ,000 Dg/m3
H2SO4 and 0.64 ppm O3,
DlOO Dg/m3 H2SO4 and 0.20
ppm O3.
0 PE Increase (compared to air Synergism at D200 Dg/m3
control).
0 PE Acid-induced decrease in
DLCO not affected
by03.
H2SO4 and 0.64 ppm O3,
D500 Dg/m3 H2SO4 and 0.2
ppm O3 (with suggestion at
< 500 Dg/m3 H2SO4).
None: O3 did not alter acid
effect.
Reference
Schlesinger et al.
(1992a)
Schlesinger et al.
(1992b)
Warren and Last
(1987)
Warren and Last
(1987)
Warren and Last
(1987)
Chen et al. (1991)
-------
Table 6-26 (cont'd). Toxicological Interactions to Binary Mixtures of Ozone
with Acids and Other Pollutants'
^
Concentration1'
ppm
0.15
0.15
0.2
0.2
0.64
0.64
0.64
0.8
0.2
0.2
0.2
Dg/m
294
84
294
24
392
1,000
392
1,000
1,254
1,000
1,254
1,000
1,254
1,000
1,568
1,200
392
5
392
5
392
5
Pollutant 3
03
H2SO4 (layered
on ZnO)
03
H2SO4
(layered on
ZnO)
O3
H2S04
O3
H2S04
O3
H2S04
O3
H2S04
O3
H2S04
O3
H,S04
(0.63 Dm)
03
(NH4)2S04
03
(NH4)2S04
03
(NH4)2S04
Exposure Duration
1 h to H2SO4,
2 h rest, then 1 h
toO3
Head-only (acid),
whole-body (O3)
H2SO4 3 h/day for
7 days, O3 on Day 9
Head-only (acid),
whole-body (O3)
23.5 h/day for 7 days
15 or 30 days
23.5 h/day for 7 days
23.5 h/day for 7 days
23.5 h/day for 7 days
O3 for 2 h, followed
by H,SO4 for 1 h
23.5 h/day for 7 days
23.5 h/day for 2 days
23.5 h/day for 3 days
Species, Sex
(Strain)
Agec
Guinea pig, M
(Hartley)
260-325 g
Guinea pig, M
(Hartley)
260-325 g
Rat, M
(S-D)
250-300 g
Rat, M
(S-D)
250-300 g
Rat, M
(S-D)
250-300 g
Rat, M
(S-D)
250-300 g
Rat, M
(S-D)
225-275 g
Guinea pig, F, M
(Hartley)
1.5-2 mo old
Rat, M
(S-D)
250-300 g
Rat, M
(S-D)
250-300 g
Rat, M
(S-D)
250-300 g
Endpoints
Pulmonary function
Pulmonary function
Rate of collagen synthesis
Lung protein content
Protein content (lavage)
Proximal acinar lesion
volume
Lung protein and free
proline content
Airway constriction
(measured by trapped gas
volume)
Rate of collagen synthesis
Lavageable protein
Lavageable protein
Time of
Endpoint Response to
Measurement Mixture
0 PE Greater decrease in DL ,
VC after O3; no change in
alveolar volume, TLC with
mixture.
0 PE Decrease in TLC, VC,
DLCO enhanced by O3.
0 PE Increase.
0 PE Increase only at 15 days.
0 PE Increase.
0 PE Increase.
0 PE Increase with H SO -and4O
only.
0 PE Increase compared to air.
0 PE Increase.
0 PE Increase.
0 PE Increase.
Interaction
Suggested synergism
(greater than additive)
for DLCO, but not VC.
Suggested synergism
(greater than additive).
Possibly synergism:
Effect different from
O3 alone.
Suggested synergism.
None: Effect same
as O3 alone.
None: Effect same
as O3 alone.
Synergism.
No interaction:
Effect same as O3
alone; H2SO4 had no
effect.
Synergism: Effect
greater than O3; sulfate
had no effect.
Synergism: Effect
greater than O3; sulfate
had no effect.
No interaction:
Effect same as O3
alone.
Reference
Chen et al. (1991) co
Chen et al. (1991)
Warren et al. (1988)
Last (1991b)
Warren et al. (1988)
Warren et al. (1988)
Last et al. (1986)
Silbaugh and Mauderly
(1986)
Warren et al. (1986)
Warren et al. (1986)
Warren et al. (1986)
-------
Table 6-26 (cont'd). Toxicological Interactions to Binary Mixtures of Ozone
with Acids and Other Pollutants'
u>
Concentration1'
ppm
0.2-
0.64
0.64
0.96
0.3
3.0
0.15
0.1
0.6
0.4
0.4
Dg/m
392-
1,254
5
1,254
1,000
1,882
5,000
588
7,860
294
250
1,176
1,000
784
380
Pollutant
O3
(NH4)2S04
03
(NH4)2S04
0,
(NH4)2S04
0,
SO,
O3
HNO3
O3
HNO3
O3
HMSAd
(0.32 Dm)
Exposure Duration
23.5 h/day for 7 days
23.5 h/day for 7 days
23.5 h/day for 7 days
5 h/day for 3 days
Head-only
4 h/day for 4 days
Nose-only
4h
Nose-only
4h
Nose-only
Species, Sex
(Strain)
Agec
Rat, M
(S-D)
250-300 g
Rat, M
(S-D)
225-275 g
Rat, M
(S-D)
225-275 g
Sheep, F
Adult (31 kg)
Rat, M
(F344)
250 g
Rat, M
(F344)
250 g
Rat, M
(S-D)
7 weeks old
Endpoints
Lung DNA content; tissue
protein content; lavage LDH,
acid phosphatase, N-acetyl-D-
D-glucosaminidase
Lung protein and free
proline content
Lung protein content,
proline content, apparent
collagen synthesis rate,
fibroblast numbers in lesions,
lesion volume
Trachea! mucus
velocity
Lavage cell population;
protein (lavage); AM
respiratory burst, LTC4;
elastase inhibitory capacity
(lavage)
Lavage cell population;
protein (lavage); AM
respiratory burst, LTC4;
elastase inhibitory capacity
(lavage)
Breathing pattern, fatty acid
composition of surfactant,
nasal epithelium and
parenchymal lesions, lavage
protein
(24-48 h PE)
Time of
Endpoint
Measurement
OPE
OPE
OPE
Oand
24hPE
18hPE
18hPE
Pulmonary
function
during
exposure,
others 23 h
PE
Response to
Mixture
No change in DNA,
increase in tissue protein,
increases in
all enzyme levels.
No change.
Increase.
Decrease in velocity
(compared to air control) .
No change in any endpoint
(compared to air control).
Increased protein, PMN
number, elastase inhibiting
capacity; no effect on other
endpoints (compared to air
control).
Rapid breathing, increased
protein, decreased fatty acid
content, focal lesions with
thickened alveolar septa and
cellular infiltration in
parenchyma.
Interaction
No interaction: Effect
on protein and enzymes same
as O3 alone.
None.
Synergism: Effect greater than
O3 alone; H2SO4 had no effect
(previous study).
Not determinable: no measure
of single pollutants performed.
None.
Less than additive for protein,
PMN number, elastase
inhibitory capacity; no
interaction for other endpoints.
None: Effect similar to O3
alone.
Reference
Warren et al.
(1986)
Last et al. (1986)
Last et al. (1986)
Abraham et al.
(1986)
Nadziejko et al.
(1992)
Nadziejko et al.
(1992)
Mautz et al.
(1991)
-------
Table 6-26 (cont'd). Toxicological Interactions to Binary Mixtures of Ozone
with Acids and Other Pollutants'
Concentration1'
Species, Sex
(Mram) l ime ot tndpomt Response to
ppm
0.2
0.4
0.8
1.0
0.4
0.3
1.1
3.3
0.6
10
0.6
10
1.0
1.0
0.8
Dg/m
392
784
1,568
1,230
784
369
1,353
4,059
1,176
12,300
1,176
12,300
1,960
1,960
1,568
2,000-
50,000
3 Pollutant
03
O3
03
HCHO
O3
HCHO
HCHO
HCHO
O3
HCHO
03
HCHO
O3
Cigarette smoke
O3
Cigarette smoke
03
Silica
(instilled)
Exposure Duration
22 h/day for 3 days
22 h/day for 3 days
3 h (rest)
3 h (exercise)
O3: 0.5 h, then
5 puffs smoke (sequential)
O3: 1.5h, then
10 puffs smoke
(sequential)
O3: chamber, smoke:
head-only
Silica instilled on Day 1
followed by O3 for
6 h/day, 5 days/week
for 37 days beginning
Day 4
Agec Endpoints Measurement
Rat, M Nasal epithelial cell turnover, Thymidine
(Wistar) histopathology (LM) administered 2 h
150-190 g PE, sacrifice 4 h
PE
Rat, M Nasal epithelial cell 2 days PE
(S-D) turnover, parenchyma!
7 weeks old histopathology
Guinea pig, F Airway responsivity 0, 5, or 24 h PE
(Hartley) (to metacholine challenge),
350-400 g tracheal vascular permeability,
0-24 h PE
Rat, M Pulmonary fibrosis 24 h PE (last O )
(S-D)
17 weeks old
Mixture
Increased turnover
(but some conditions
produced decrease
due to change in
ventilation).
Increased focal
parenchymal lesions
with exercise, but no
effect at rest;
increased nasal cell
turnover at rest or
exercise.
Increased responsivity
and permeability
immediately PE at
both doses (magnitude,
but not duration of
effect).
No change in Jung
DNA, protein, or
hydroxyproline content;
increase in ratio of
hydroxyproline to
DNA, protein, or wet
weight (compared to air
control) at 50,000 Dg.
Interaction
Synergism for turnover
with 0.4 ppm O3 and
1-3 ppm HCHO,
depending on anatomical
site (greater than the
sum of individual
responses to O3 and
HCHO); microscopic
lesions similar to O3
and/or HCHO alone.
Parenchyma:
Synergism with
exercise, antagonism at
rest.
Nasal: Synergism.
Suggested synergism:
No effect of O3 or
smoke alone at low
dose; high O3 increased
responsivity and
permeability; high
smoke increased
responsivity.
None: No biological
significance.
Reference
Reuzel et al. (1990)
Mautz et al. (1988)
Nishikawa et al. (1992)
Shiotsuka et al. (1986)
-------
Table 6-26 (cont'd). Toxicological Interactions to Binary Mixtures of Ozone
with Acids and Other Pollutants'
Concentration1
ppm Dg/m
0.8 1,568
1.5 2,940
10,000
> Species, Sex
3 Pollutant Exposure Duration
O3 4h
O3
Carbon
black
(Strain)
Agec
Mouse, F
(Swiss)
20-23 g
Endpoints
Cell counts in lavage;
AM phagocytosis
Time of Endpoint Response to
Measurement Mixture
20 h PE Increase PMN counts
compared to O3 alone;
greater depression of
phagocytosis than O3
alone.
Interaction
At 0.8 ppm: No
intereaction; at 1.5 ppm:
Suggested synergism.
Reference
Jakab and Hemenway
(1994)
'See Appendix A for abbreviations and acronyms.
''Grouped by pollutant mixture.
cAge or body weight at start of exposure.
dHMSA = hydroxymethanesulfonate.
Ln
-------
6.4.2.1 Nitrogen Dioxide as CopoIIutant
The most commonly studied copollutant in binary mixtures with O3 is NO2. Studies
discussed in the previous O3 criteria document indicated that, although interaction may occur
between these two pollutants, in general, O3 often masked the effects of the NO2 or accounted
for most of the response. This is because, on a mole-to-mole basis, O3 is considerably more
toxic than NO2, and the relative contribution of O3 and NO2 to pulmonary injury is driven by
the exposure ratio of the two pollutants. Commonly studied endpoints for assessing effects of
these mixtures were lung morphology, biochemistry, and resistance to bacterial infection.
To put the exposure concentrations of NO2 into some perspective, short-term,
24-h averages are generally DO. 17 ppm, and 1-h averages are generally DO.4 ppm in major
metropolitan areas. However, hourly averages in most regions often exceed 0.2 ppm at least
once during the year (Schlesinger, 1992).
An earlier study noted that the morphological response of the rat lung alveolar
epithelium following 60 days of exposure to O3/NO2 mixtures (0.25 ppm O3 + 2.5 ppm NO2,
or 0.9 ppm O3 + 0.9 ppm NO2) was due to the O3 (Freeman et al., 1974). However, the
duration of exposure may affect the contributory role of the copollutant. Thus, for example,
Terada et al. (1986) exposed rats to O3 alone, or to a mixture of 0.1 ppm O3 + 0.3 ppm NO2,
with O3 administered 8 h/day and NO2 administered 24 h/day for up to 18 mo. Following 1
mo of exposure, observed lung lesions in the group exposed to the mixture were similar in
severity to those noted with exposure to O3 alone, but as the duration of exposure increased,
the morphological changes in interstitial tissue appeared more marked in those animals
exposed to the mixture. Edema of pulmonary connective tissue was more pronounced and
alveolar Type 2 cells became swollen. Although this study was not quantitative, qualitative
observations led the authors to conclude that the lesions were not due to O3 alone but that NO2
played some, albeit undefined, contributory role.
The effect of exposure duration on interaction was also noted in studies of
Schlesinger et al. (1990, 1991). Rabbits exposed to 0.3 ppm O3 + 3.0 ppm NO2 for
2 h showed synergistic increases in certain BAL eicosanoids obtained immediately after
exposure, whereas animals exposed to the same mixture for 2 h/day for 14 days showed no
interaction for the same parameters.
A number of studies examined other biochemical responses to O3/NO2 mixtures
(e.g., sulfhydryl metabolism and the activity of certain enzymes). Some of the studies
discussed in the previous O3 criteria document were found to involve synergism (e.g., Mustafa
et al., 1984). More recent studies of lung biochemistry also suggest that O3 and NO2 interact
synergistically. Ichinose and Sagai (1989) exposed rats and guinea pigs to 0.4 ppm
O3, 0.4 ppm NO2, or a mixture of the two pollutants continuously for 2 weeks. No change in
lung peroxide production was observed in rats, but the mixture synergistically increased
peroxide levels in guinea pigs. The guinea pigs showed no change in lung antioxidant content
following any exposure, whereas the mixture synergistically increased antioxidant levels in rat
lung. The conclusion of a significant interaction was based on relative changes from air
controls following exposure to the mixture, compared to changes following exposure to each
pollutant alone, using the t-test. Synergism was defined as a change greater than the sum of
the responses to individual pollutants; no specific test for interaction was performed.
Ichinose and Sagai (1989) also noted that levels of antioxidant enzymes in rat or
guinea pig lungs were variously affected by exposure to the above mixture. For example, GST
was decreased in both species exposed only to O3, but the mixture produced a reduction of this
6-146
-------
enzyme in guinea pigs and no change in rats. Thus, the occurrence of interaction was
dependent on endpoint as well as species. This latter finding likely reflected interspecies
differences in biochemical defenses against oxidant pollutants, given the results of a study by
Sagai et al. (1987) with four animal species. This study suggested that observed species
differences in lipid peroxide formation following exposure were related to the relative content
of antioxidants and the specific composition of phospholipids and their fatty acids. The guinea
pig was the most sensitive animal, and the hamster was the most resistant.
The effects of exposures to O3/NO2 mixtures on lung lipid peroxides and antioxidant
activity have been examined in a number of other studies (Ichinose et al., 1988; Lee et al.,
1990; Sagai and Ichinose, 1991), and the results generally confirm that noted above (i.e., such
mixtures tend to produce synergistic interaction). However, there is also some evidence for
antagonism. Takahashi and Miura (1989) examined effects on the pulmonary xenobiotic
system of rats exposed for 1 or 2 mo to a mixture of 0.2 ppm O3 + 4.0 ppm NO2, as well as to
each pollutant alone. Ozone induced an increase in lung cytochrome P-450 content, but the
activity of these enzymes was reduced by the addition of NO2 to the exposure atmosphere; that
is, the mixture resulted in levels intermediate between those found with O3 or NO2 alone.
However, the reduction in enzyme activity induced by NO2 was restricted to those enzymes
that had been increased by exposure to O3 alone. The authors suggested that antagonism was
due to the production of undefined secondary reaction products in the exposure atmosphere. A
similar explanation was proposed to explain observed synergism of lung antioxidant activity in
another study (Lee et al., 1990). Thus, the response to any secondary product likely depends
on the endpoint examined, assuming that the same reaction products were formed in these two
studies.
The role of exposure parameters in producing an interaction between simultaneously
inhaled O3 and NO2 was examined by Gelzleichter et al. (1992a). Rats were exposed to
various combinations of O3 and NO2 for various durations (6, 8, 12, and 24 h), such that the
C x T products were identical for each of four exposure sets. As indicated in Table 6-24, as
the exposure duration increased, the exposure concentration of each component of the mixture
decreased. Lavaged protein levels and recovered cells were the endpoints. For each exposure
combination, the additive response was predicted from the results of exposure to each pollutant
alone, and then synergism was indicated when there was deviation from additivity. Responses
to exposure to either O3 or NO2 alone for 6, 8, or 12 h showed that the product of C x T was
a constant for the observed biological effects. However, less severe changes occurred when
delivery was at the lowest dose rate (i.e., when the lowest concentration of each pollutant was
delivered over the 24-h exposure duration). Exposure at higher dose rates (i.e., 6 to 12 h)
increased the magnitude of the response. Thus, the degree of response to each pollutant alone
was not a constant function of C x T throughout the entire range of dose rates, but was
concentration driven, and was not identical at the highest and lowest rates. Responses
following exposure to the mixture did not follow C x T, even over the range of dose rates in
which C x T was constant following exposure to the pollutants individually. Thus,
interaction, in this case, synergism, appeared to be concentration dependent, in that the
response was disproportionately greater at the higher concentrations (higher dose rates) of the
constituent pollutants in the mixture. The response following exposure to the mixtures
appeared to be a function of peak concentration, rather than of cumulative dose. More
recently, Rajini et al. (1993) noted that analysis of all kinetics following similar exposure to
mixtures did not reflect a C X T relationship.
6-147
-------
All of the studies described above involved simultaneous exposure to O3 and NO2.
However, ambient exposure to these pollutants has temporal patterns, and exposure to one
agent may alter the response to another, subsequently inhaled agent. The realism of these
studies is somewhat dependent on their relationship to actual temporal patterns of pollutants in
ambient air (i.e., whether one material is the precursor of the other, as is the case for O3 and
NO2). As described in the previous O3 criteria document (U.S. Environmental Protection
Agency, 1986), Fukase et al. (1978) exposed mice for 7 days to 3 to 15 ppm NO2 for 3 h/day,
followed by 1 ppm O3 for 3 h/day, and noted an additive effect on the level of lung GSH.
Yokoyama et al. (1980) exposed rats to 5 ppm NO2 or 1 ppm O3 for 3 h/day, or to
NO2 for 3 h followed by O3 for 3 h/day, for various total durations up to 30 days, and assessed
lung mechanics in postmortem lungs, lung histology, and enzyme activity in subcellular
fractions of lung tissue. The activity of phospholipase A2 in the mitochondrial fraction was
increased in those animals exposed to O3 only or to O3 after NO2, and the response in the latter
was significantly greater than that in the former. A decrease in activity of lysolecithin
acyltransferase in the supernatent fraction was found only in those animals exposed to both
NO2 and O3. Pulmonary mechanics showed a change in pulmonary resistance (as a function of
elastic recoil pressure) in the O3- and NO2/O3-exposed animals. Histologically, the lungs of
the animals exposed to both NO2 and O3 appeared similar to those exposed to O3 alone;
however, a slight degree of epithelial necrosis in medium bronchi, not found with either NO2
or O3 alone, was seen in the animals exposed to both pollutants. In addition, damage at the
bronchoalveolar junction appeared to be somewhat more marked in animals exposed to both
gases than in those exposed to O3 alone. This study suggested that sequential exposures
produced responses that, in most cases, did not differ greatly from those due to O3 alone.
Aside from sequential exposures, simulation of ambient exposure scenarios
involving NO2 and O3 has been performed by examining the effects of a continuous baseline
exposure to one concentration of both pollutants, with superimposed short-term peaks to a
higher level of one or both gases. The endpoint generally examined in this regard has been
bacterial resistance. Studies reported in the previous criteria document (e.g., Ehrlich et al.,
1979; Ehrlich, 1983) in which mice were exposed to O3 under various scenarios of baseline
concentrations of NO2 on which were superimposed daily peak exposures to NO2 or a
combination of NO2 and O3 suggested that exposure with peaks can enhance response to
pollutant mixtures, and that the sequence of peak exposures was important in producing
reduced resistance to infection that was different from that due to exposure to the baseline
concentration only.
As a comparison, toxicologic interactions for infectivity involving simultaneous
exposure to NO2 and O3 discussed in the previous O3 criteria document were found generally to
be additive following acute exposures, with each pollutant contributing to the observed
response when its concentration reached the threshold at which the gas would have affected
bacterial resistance when administered alone (Goldstein et al., 1974). If the exposure level of
either NO2 or O3 was below this threshold, then the response was due solely to the constituent
inhaled at the more toxic concentration (Ehrlich et al., 1977).
More recently, Graham et al. (1987) examined resistance to respiratory infection
(as measured by bacterial-induced mortality) in mice continuously exposed (for 15 days,
24 h/day) to baseline levels of an NO2/O3 mixture with two daily, 1-h peaks of the mixture at
very high, high, intermediate, and low exposure concentrations (see Table 6-25 for
concentrations). Animals were also exposed to the same baseline levels of either NO2 or
6-148
-------
O3 onto which were superimposed two daily, 1-h peaks of the same single gas in
concentrations as above. At the low concentration, only NO2 increased mortality. At the
intermediate exposure level, the mixture was synergistic; NO2 alone increased mortality and
O3 had no effect. At the high exposure level, the combined exposure was again synergistic;
exposure to each gas separately increased mortality. A similar effect was seen at the very high
level, although the combined exposure just missed statistical significance for synergism. These
results are consistent with those of the earlier studies reported in the previous O3 criteria
document and support the conclusion that response depends on the specific exposure pattern.
The results of Graham et al. (1987) are also consistent with those from the earlier studies with
simultaneous exposures.
The relationship between exposure and response is very complex and seems to
depend on exposure duration, the ratio of O3 and NO2 concentrations, and other factors that
may include the production of secondary reaction products within the exposure atmosphere.
This complexity was highlighted by the study of Gelzleichter et al. (1992b), who examined
effects of combined or sequential exposures of rats to mixtures of O3 and NO2 at various
concentrations ranging from 0.2 to 0.8 ppm O3 and 3.6 to 14.4 ppm NO2. Sequential
exposures consisted of 6 h of O3 at night, followed by 6 h of NO2 during the day, or vice
versa; concurrent exposures were for 6 h/day for 3 days. Various endpoints were examined,
and it was noted that sequential and concurrent exposures did not result in the same response.
Thus, lavage protein levels were increased additively with sequential exposure (in any pollutant
order) but were found to be greater than additive with concurrent exposure. An increase in the
number of lavaged epithelial cells was additive for the O3 night/NO2 day sequence, antagonistic
for the NO2 night/O3 day sequence, and additive for concurrent exposure. An increase in the
number of lavaged PMNs was additive for both sequential conditions and was synergistic for
concurrent exposure. It was concluded that production of synergism depended on the
concentration of each pollutant within the mixture, and additivity would result for any endpoint
when the concentration of each component of the mixture fell below a certain threshold level.
However, these threshold concentrations were endpoint specific, with some endpoints being
more sensitive than others; it was speculated that the least sensitive assays were based on
changes that were reversible, whereas the most sensitive ones were irreversible. The authors
also noted that the extent of chemical reaction within the O3/NO2 mixture atmosphere was
related to the extent of toxicological interaction, suggesting that interaction was due to the
production of some secondary reaction product, which, in this case, was suggested to be
nitrogen pentoxide. This particular chemical also had been suggested in earlier studies to be
responsible for interactions following exposure to NO2/O3 atmospheres (e.g., Diggle and Gage,
1955).
The concentrations at which synergism occurred in the study of Gelzleichter et al.
(1992b) discussed above (DO.4 ppm O3 and D7.2 ppm NO2) were higher than those that
generally are found in ambient air. However, the threshold concentration for interaction was
dependent on exposure dose rate, with higher rates leading to lower threshold concentrations
for synergism.
Although interaction is clearly modulated by environmental exposure factors, such
as concentration, duration of exposure, or specific exposure regime, host factors also may play
a role. Mautz et al. (1988) examined the effect of exercise on rats exposed to mixtures of
O3 and NO2. Exercise modified the toxic interactions of combined pollutants, resulting in
synergistic interaction occurring at lower exposure concentrations of the constituent pollutants
6-149
-------
than with exposure at rest. Thus, a similar magnitude of response, an increase in the extent of
focal lesions in lung parenchyma, was noted 2 days following a 4-h exposure to 0.6 ppm
O3 + 2.5 ppm NO2 at rest or with a shorter (3-h) exposure to lower concentrations, 0.35 ppm
O3 + 0.6 ppm NO2, with exercise. In both cases, the response was different from that due to
either pollutant given alone. Furthermore, a greater response was noted with a 3-h exposure to
0.6 ppm O3 + 2.5 ppm NO2 with exercise than to the same mixture for the same exposure
duration at rest. Thus, exercise also increased the response at similar concentrations compared
to rest. The effect of exercise was ascribed to an increase in delivered dose or dose rate, due
to increased VE. The ability of exercise to enhance response to a pollutant mixture also was
noted by Bhalla et al. (1987).
The study of Mautz et al. (1988) above also provided further evidence suggesting
that chemical reactions within the exposure atmosphere may play some role in toxicologic
interaction. In this case, nitric acid (HNO3) vapor was noted at concentrations ranging from
0.02 to 0.73 ppm, depending on the concentrations of the primary constituents. As discussed
below, acids have been found to interact with O3. This study also found interaction to occur at
a concentration of one of the components, NO2, that had no effect when administered alone.
Although this appears to contrast with the conclusions of Graham et al. (1987) above, the
endpoints in these two studies were quite different.
Another aspect of pollutant interaction involves the ability of O3/NO2 mixtures to
affect the course of other lung changes (e.g., malignant tumor colonization). Richters (1988)
exposed mice to a mixture of 0.15 ppm O3 + 0.35 ppm NO2 for 7 h/day, 5 days/week for 12
weeks, following which the mice were injected (iv) with viable melanoma cells. The mice
were sacrificed 3 weeks later, and the lungs were examined for melanoma colonies. Although
exposure to O3 alone produced no change in the percentage of animals with colonies or in the
average number of colonies per lung (compared to air control), exposure to the mixture
produced an increase in the latter, suggesting to the authors that the mixture faciliated cancer
cell colonization. However, the exact role played by O3 in the mixture is not clear because a
previous study had indicated that NO2 alone facilitates blood-borne cancer cell spread to the
lungs (Richters and Kuraitis, 1981). Furthermore, the experimental model used is not
generally accepted as representing metastatic mechanisms.
Ichinose and Sagai (1992) also examined the ability of an O3/NO2 mixture to
promote primary lung tumor development. Rats were injected (ip) with BHPN and then were
exposed for 13 mo to a mixture of 0.05 ppm O3 + 0.4 ppm NO2, O3 alone, or to clean air
(chamber control). Although the NO2 exposure was continuous, the O3 exposure was
intermittent, with the concentration altered between 0 and 0.1 ppm following a sine curve from
9 to 19 h of each day (resulting in a daily mean concentration of 0.05 ppm). One other group
of rats served as a room control, maintained in a clean room for 24 mo following injection of
BHPN. After an 11-mo recovery period, all animals were autopsied. Compared to clean-air-
exposed animals, lung tumor incidence was increased in mice exposed to the mixture; O3 alone
did not increase the tumor incidence. However, the authors noted that tumor incidence in the
room control group was not different from that in the group exposed to the mixture, and
suggested that the clean air (chamber control) group should be used as a control in interpreting
the data from the pollutant-exposed animals. The enhanced incidence in mixture-exposed
animals was suggested to be due to synergistic increases in lipid peroxidation, which was noted
in other studies (see Table 6-25). A complication in interpreting this study is that a previous
study (Sagai and Ichinose, 1991) had suggested tumor development in animals exposed to an
6-150
-------
O3 and NO2 mixture without BHPN, although this latter study involved a longer exposure
duration and somewhat higher pollutant concentrations.
6.4.2.2 Acidic Compounds as Copollutants
Binary mixtures containing acids comprise another type of commonly examined
exposure atmosphere. Most of these mixtures included acidic sulfate aerosols as the
copollutant. Peak (1-h) ambient levels of sulfuric acid (H2SO4) are estimated at 75 Dg/m3, with
longer (12-h) averages about one-third of this concentration (Spengler et al., 1989).
Earlier studies that employed simultaneous single, repeated, or continuous
exposures of various animal species to mixtures of acid sulfates and O3 found responses for
several endpoints, including tracheobronchial mucociliary clearance, alveolar clearance,
pulmonary mechanics, and lung morphology, to be due solely to O3 (U.S. Environmental
Protection Agency, 1986; Cavender et al., 1977; Moore and Schwartz, 1981; Phalen et al.,
1980; Juhos et al., 1978). However, synergism was noted for bacterial infectivity in mice
(Grose et al., 1982), for response to antigen in mice (Osebold et al., 1980), and for effects on
lung protein content and the rate of collagen synthesis in rats (Last et al., 1983, 1984a; Last
and Cross, 1978). More recent studies, performed since publication of the previous O3 criteria
document, support the earlier finding of synergism between O3 and acid sulfates on lung
biochemistry, and provide possible explanations for underlying mechanisms.
Last et al. (1986) exposed rats for 7 days to O3 alone (at 0.96 ppm) and to mixtures
of O3 with one of three aerosols, sodium chloride, sodium sulfate, or ammonium sulfate
[(NH4)2SO4] (all at 5 mg/m3); only the (NH4)2SO4 was acidic. Lung protein content, proline
content, collagen synthesis rate, fibroblast numbers in parechymal lesions, and the volume of
parenchymal lesions were examined following exposure. Mixtures of O3 with sodium chloride
or sodium sulfate produced changes that did not differ from those found with O3 alone. On the
other hand, mixtures of (NH4)2SO4 with O3 resulted in increases in all of the measured
parameters, and the increases were greater in magnitude than those due to O3 alone; synergism
was concluded, although there has been some question concerning the statistical approach used
(Last, 1991a). These results suggested that acidity was necessary for synergism of the aerosols
with O3. This conclusion was further supported by demonstrating that significant interaction of
O3 with H2SO4, which is much more acidic than (NH4)2SO4, occurred at lower concentrations
than was noted for mixtures of O3 and (NH4)2SO4 (Warren and Last, 1987); interaction was
suggested with H2SO4 concentrations as low as 40 Dg/m3 (with 0.2 ppm O3) for some lung
biochemical endpoints. The studies above did not use any specific statistical test for
interaction, and conclusions of interaction were based on findings of significant differences
from responses following exposure to O3 alone and the absence of detectable responses
following exposure to H2SO4 aerosols at the same and higher concentrations.
Warren et al. (1986) found synergistic interaction with the above endpoints
following 7 days of exposure to 0.2 ppm O3 + 5 mg/m3 (NH4)2SO4. However, exposure for
only 3 days produced responses that were not different from those noted with O3 alone. This
seems to indicate that the duration of exposure is a factor affecting the occurrence of any
interaction. However, exposure duration may also affect the type of interaction. In a study by
Schlesinger et al. (1992a) in which rabbits were exposed to a mixture of 0.1 ppm
O3 + 125 Dg/m3 H2SO4 for 2 h/day, 5 days/week, a synergistic increase in bronchial epithelial
secretory cell number was noted after 4 mo of exposure, whereas antagonism was noted
following 8 mo of continued exposure.
6-151
-------
The mechanism underlying interaction between acid sulfates and O3 is not known.
Last et al. (1986) noted that similar sites of deposition for O3 and acid aerosols favored
synergism. A synergistic response of biochemical indices in rat lung with exposure to
1,000 Dg/m3 H2SO4 + 0.6 ppm O3 was found when the acid droplet diameter was 0.5 Dm,
whereas no increase compared to the O3-only response was noted when the droplet diameter
was 0.02 Dm. Apparently, the larger particles that deposited to a greater extent within the
bronchoalveolar junction, the major target site for O3, were most interactive.
Observed synergism between O3 and acid sulfates in rats also was suggested to be
due to a shift in the local microenvironmental Ph of the lung following deposition of acid,
enhancing effects of O3 by producing a change in the reactivity or residence time of reactants,
such as free radicals, involved in O3-induced tissue injury (Last et al., 1984a). If this were the
only explanation, then the effects of O3 should be enhanced consistently by the presence of acid
in an exposure atmosphere. However, in the study of Schlesinger et al. (1992b) in which
rabbits were exposed for 3 h to combinations of O3 at 0.1, 0.3, and 0.6 ppm + H2SO4 (0.3
Dm) at 50, 75, and 125 Dg/m3, antagonism was noted in the evaluation of stimulated production
of superoxide anion by AMs harvested by lavage immediately after exposure to 0.1 or 0.3 ppm
O3 in combination with 75 or 125 Dg/m3 H2SO4 and also for AM phagocytic activity at all of
the O3/acid combinations. Mixtures of O3 (0.6 ppm) and another acid, HNO3 vapor (1,000
Dg/m3), also produced antagonism for certain aspects of the function of Ams harvested from
acutely exposed rats (Nadziejko et al., 1992). Although the deposition sites of both acid and
O3 should be comparable in these two studies, perhaps the particular cellular endpoints
examined are subject to this type of interaction.
Last (1989) observed an apparent all-or-none response in rats exposed to the acid
sulfate/O3 mixtures. That is, there was no concentration-response relationship between the
concentration of acid in the mixture and the extent of change in various endpoints, compared to
effects observed with O3 alone. In the study of Schlesinger et al. (1992b), a similar
phenomenon was noted, but, in this case, the concentration of O3 in the mixture did not always
influence the response compared to that seen with acid alone. Thus, exposure-concentration-
response relationships noted with individual pollutants may not necessarily hold following
exposure to their mixtures. This is consistent with the results of Gelzleichter et al. (1992a) for
mixtures of O3 and NO2.
The above studies involved simultaneous exposures to O3 and acidic pollutants, but
some studies involving sequential exposures to O3 and acid sulfate aerosols were described in
the previous O3 criteria document. For example, Gardner et al. (1977) found an additive
increase in infectivity when mice were exposed to 0.1 ppm O3 for 3 h prior to a 2-h exposure
to 900 Dg/m3 H2SO4, whereas no difference from air control was noted when the acid was
administered prior to O3. Grose et al. (1980) noted a reduction in ciliary activity in isolated
tracheal sections obtained from hamsters exposed to 0.1 ppm O3 for 3 h, followed by exposure
to 1,090 Dg/m3 H2SO4 for 2 h, that was less in magnitude than that found with exposure to acid
alone; O3 alone had no effect.
Silbaugh and Mauderly (1986) examined the ability of O3 to increase susceptibility
to a subsequent exposure to H2SO4 in terms of producing airway constriction. Guinea pigs
were exposed to 0.8 ppm O3 for 2 h followed by H2SO4 (12 mg/m3 for 1 h). An increased
volume of trapped gas in the lungs (the metric of constriction) was seen with both O3 alone and
with the mixture, but the response to the latter did not differ from that due to the former, and
6-152
-------
acid alone had no effect. Thus, in this case, preexposure to O3 did not affect response to a
subsequent exposure to acid.
Chen et al. (1991) examined the reverse exposure scenario, whether exposure to
H2SO4 affected subsequent response to O3. Guinea pigs were exposed to H2SO4 or ultrafme
zinc oxide (ZnO) particles coated with H2SO4. A 1-h exposure to 0.15 ppm O3 following a
1-h exposure to acid (300 Dg/m3, 0.09 Dm) did not alter the response seen with acid alone, a
decline in DLC0. However, when single (1-h) or multiple (3-h/day, 7-day) exposures to acid-
coated ZnO (24 or 84 Dg/m3 equivalent H2SO4) were followed by a 1-h exposure to 0.15 ppm
O3, the effect on DLCO appeared to be greater than additive, although no specific statistical test
for interaction was performed. This study suggested that prior exposure to acid increased the
susceptibility of the guinea pig to subsequent exposure to O3, but it also showed that the
manner in which the acid was delivered affected whether or not any interaction occurred. It is
likely that the number of particles was greater in the ZnO-H2SO4 aerosol than in the H2SO4
aerosol, and the interaction may reflect this greater particle number.
6.4.2.3 Other CopoIIutants
Although the bulk of the database for binary mixtures of O3 involves NO2 or acids,
a few studies examined responses to combinations of O3 with other pollutants.
Reuzel et al. (1990) exposed rats to mixtures of O3 (0.2, 0.4, or 0.8 ppm) + HCHO
(0.3 to 3.0 ppm). Although exposure to the mixtures did not alter the nature or extent of
histological lesions, (cilia loss and epithelial hyperplasia) compared to exposure to each
pollutant alone, a site-specific synergistic increase in turnover of nasal epithelial cells was
found with all concentrations of HCHO together with 0.4 ppm O3. A lack of such response
with 0.8 ppm O3 was ascribed to an O3-induced alteration in breathing pattern, which reduced
the delivered dose. It was, however, noted that interaction occurred only when one constituent
of the mixture was administered at cytotoxic concentrations, an exposure scenario that rarely
occurs in ambient air. In any case, the authors concluded that because cell proliferation likely
plays a role in carcinogenesis, and that if mixtures potentiate cell proliferation, then exposure
to pollutant mixtures may increase cancer risk.
Mautz et al. (1988) exposed rats for 3 h, both at rest and with exercise, to a mixture
of 0.6 ppm O3 + 10 ppm HCHO. A synergistic increase in nasal epithelial cell turnover
followed exposure with exercise, whereas exposure at rest resulted in no difference from that
seen with HCHO alone. Likewise, exposure to the mixture with exercise resulted in an
increase in the number of focal lesions in lung parenchyma compared to either O3 or HCHO
alone, but exposure at rest resulted in a lower incidence of lesions than seen with O3 alone.
This latter observation was ascribed to an effect of HCHO on breathing pattern, producing a
change in inhaled dose of O3 that did not occur with exercise.
Nishikawa et al. (1992) examined the effect of sequential exposure to cigarette
smoke and O3 in altering airway responsivity to inhaled bronchoconstrictor challenge and
tracheal vascular permeability in guinea pigs. Animals were exposed to 1 ppm O3 for 0.5 h,
followed by 5 puffs of cigarette smoke, or to 1 ppm O3 for 1.5 h, followed by 10 puffs.
Exposure to O3 and five puffs increased responsivity and vascular permeability immediately
after exposure, whereas no effect on either endpoint was noted with either pollutant given
alone. Exposure to O3 and 10 puffs also increased responsivity and permeability, but to the
same extent as did the lower concentration mixture or exposure to O3 alone, whereas exposure
to 10 puffs of smoke only increased responsivity. Thus, sequential exposure to O3 and
6-153
-------
cigarette smoke enhanced the magnitude of response compared to either pollutant alone, but the
duration of response was not altered.
The potential role of O3 in enhancing fibrotic lung disease by interaction with silica
was examined by Shiotsuka et al. (1986), who exposed rats with developing silica-induced
fibrosis to O3 at 0.8 ppm for 6 h/day, 5 day/week for 37 exposure days. Silica had been
instilled (2, 12, or 50 mg) on Day 1 of the study and exposure to O3 began on
Day 3 or 4 postinstillation. There was found to be no interaction between silica and O3 in
development of fibrosis, as assessed biochemically (lung content of hydroxyproline) or
histopathologically. Although an increase was found in the ratio of hydroxyproline to total
protein in the group exposed to the mixture and instilled with the highest amount of silica, this
was not considered by the authors to be biologically significant.
6.4.3 Complex (Multicomponent) Mixtures Containing Ozone
Ambient pollution in most areas is a complex mix of more than two chemicals, and
a number of studies have examined the effects of exposure to multicomponent atmospheres
containing O3. Some of these attempted to simulate photochemical reaction products occurring
under actual atmospheric conditions. However, the results of these studies are often difficult
to interpret due to chemical interactions between the components, as well as the resultant
production of variable amounts of numerous secondary reaction products, and a lack of precise
control over the ultimate composition of the exposure environment. In addition, the role of
O3 in the observed biological responses is often obscure.
One type of experimental multicomponent atmosphere that has been examined is
ultraviolet-irradiated and nonirradiated automobile exhaust mixtures. Irradiation leads to the
formation of photochemical reaction products that are biologically more active than those in
nonirradiated mixtures. Such mixtures are characterized by total oxidant concentrations
(expressed as O3) in the range of 0.2 to 1.0 ppm. Although the effects described following
exposure were not necessarily uniquely characteristic of O3, and, although O3 could have been
responsible for some, or even most of them, in most cases, the biological effects have been
difficult to associate with any one particular component. Effects of exhaust mixtures on
different species have been discussed in the previous O3 criteria document (U.S. Environmental
Protection Agency, 1986). Pulmonary function changes were demonstrated in guinea pigs
after short-term exposures to irradiated exhaust and in dogs after long-term exposure to both
irradiated and nonirradiated exhaust mixtures.
Additional studies of complex mixtures have been performed since publication of
the previous O3 criteria document. Kleinman et al. (1985) exposed rats (Sprague-Dawley,
male, 7 weeks old, nose-only) for 4 h to atmospheres designed to represent photochemical
pollution and consisting of 0.6 ppm (1,180 Dg/m3) O3 + 2.5 ppm (4,700 Dg/m3) NO2 +
5.0 ppm (13,100 Dg/m3) sulfur dioxide (SO2) + particles. The particulate phase consisted of
1,000 Dg/m3 of either H2SO4 or (NH4)2SO4 laced with iron sulfate [Fe2(SO4)3] and manganese
sulfate (MnSO4). The metallic salts act as catalysts for the conversion of sulfur IV into sulfur
VI and for the incorporation of gases into the aerosol droplets. The respiratory region was
examined for morphological effects. A confounding factor in these studies was the production
of HNO3 vapor in atmospheres that contained O3 and NO2, a phenomenon discussed
previously, and nitrate in those that contained O3 and (NH4)2SO4, but not NO2. Nevertheless, a
significant enhancement of tissue damage was noted with exposure to atmospheres containing
H2SO4 or secondarily produced HNO3 compared to those containing (NH4)2SO4, a less acidic
6-154
-------
compound. In addition, there was some suggestion that the stronger acidic atmospheres
resulted in a greater area of the parenchyma becoming involved in lesions, which were
characterized by a thickening of alveolar walls, cellular infiltration in the interstitium, and an
increase in free cells within alveolar spaces. An increased rate of nasal epithelial cell turnover
was noted following exposure to atmospheres containing particulate acids compared with
exposure to either O3 alone or to a mixture of O3 + NO2. Furthermore, exercise seemed to
potentiate the nasal and parenchymal responses to the complex mixtures containing strong acids
(Kleinman et al., 1989), a finding similar to that with the previously discussed mixtures of
O3 and NO2 or O3 and HCHO.
Bhalla et al. (1987) examined the effects of a seven-component atmosphere (similar
to that above) on epithelial permeability of rat lungs (Sprague-Dawley, male, 47 to 52 days
old). The animals were exposed for 2 h (chambers, relative humidity [RH] = 85%) to the
following: O3 (0.6 ppm) + NO2 (2.5 ppm) + SO2 (5.0 ppm) + ferric oxide
(Fe2O3) (241 Dg/m3) + (NH4)2SO4 (308 to 364 Dg/m3) + Fe2(SO4)3 (411 to 571 Dg/m3) +
MnSO4 (7 to 9 Dg/m3). The response to this mixture was compared to that following exposure
to O3 (0.6 ppm) + NO2 (2.5 ppm), O3 alone (0.6 or 0.8 ppm), or NO2 alone (6 or 12 ppm).
As above, the complex mixture was found to result in production of HNO3, in this case at
measured concentrations of 1,179 to 2,558 Dg/m3 (0.46 to 1.02 ppm); the O3 + NO2
atmosphere also resulted in some HNO3 vapor formation. Epithelial permeability was found to
increase immediately following exposure to O3, O3 + NO2, or to the complex mixture.
Although the magnitude of this change was similar following exposure to O3 alone or in
combination with other pollutants, there was increased persistence of effect after exposure to
either the binary or complex mixture.
Prasad et al. (1988) used a similar multicomponent atmosphere and examined
effects on AM surface receptors. Rats (Sprague-Dawley, male, 200 g) were exposed for
4 h/day, for 7 or 21 days to a mixture of O3 (0.3 ppm) + NO2 (1.2 ppm) + SO2 (2.5 ppm) +
(NH4)2SO4 (270 Dg/m3) + Fe2(SO4)3 (220 Dg/m3) + MnSO4 (4 Dg/m3) + Fe2O3 (150 Dg/m3), or
to O3 alone. Both the mixture and O3 alone resulted in a decrease in Fc receptor activity
beginning immediately after the last exposure. Exposure to the complex atmosphere for 7 days
resulted in a response similar to that seen with O3 alone, but continued exposure to this mixture
for up to 21 days resulted in an even greater reduction in receptor function compared to
O3 alone. However, as with most studies of complex mixtures, although the response to the
mixture was different from that found with O3, the role of other constituents was not clear.
Phagocytic function of AMs was also examined following exposure to the mixture, but there
were no O3-only controls for comparison.
Mautz et al. (1985a) examined the effects of a complex mixture on pulmonary
mechanics in exercising dogs. Exposures (nose-only) were for 200 min to a mixture of
O3 (0.45 to 0.7 ppm) + SO2 (4.8 to 5.2 ppm) + H2SO4 (800 to 1,200 Dg/m3, 0.2 Dm) +
catalytic salts of Fe2(SO4)3 and MnSO4. A greater increase in resistance and decrease in
compliance was found with the complex atmosphere than with O3 alone, but the effect was
ascribed to the presence of H2SO4. Although synergism was implied, it could not be concluded
definitively because the mixture was not tested without O3.
Mautz et al. (1991) further examined the ability of components of acidic fogs to
alter the response to O3. Rats (Sprague-Dawley, male, 7 weeks old, n = 12/group) were
exposed for 4 h (nose-only; temperature = 22 to 23 DC, RH = 82 to 83%) to 0.4 ppm O3 or
to a mixture of 0.4 ppm O3 + 670 Dg/m3 HNO3 vapor + 610 Dg/m3 H2SO4 particles (0.32
6-155
-------
Dm). Exposure to either O3 or the mixture resulted in comparable changes: development of a
rapid, shallow breathing pattern; a decrease in fatty acid composition of pulmonary surfactant;
and focal parenchymal lesions with thickened alveolar septa and cellular infiltration. The lack
of any modulation of the O3-induced effects by acids prompted the authors to raise the question
of the sensitivity of rats to inhaled acids. Although responses to any pollutant are somewhat
species dependent, there is some evidence that rats are not the most sensitive species to acidic
aerosols (U.S. Environmental Protection Agency, 1989). As discussed previously, the extent
of interaction within any one species of animal is endpoint dependent, and it is likely that the
sensitivity of various endpoints is species dependent. Thus, rats do show biochemical changes
(e.g., in collagen metabolism) with exposure to fairly low levels of acidic aerosols in
combination with O3 (see Table 6-10), although these involved longer duration exposures. In
any case, the underlying reasons for the lack of interaction in the complex-mixture study above
remain unclear.
Kleinman et al. (1989) exposed rats (Sprague-Dawley, male, 7 weeks old, nose-
only) to a mixture of O3 (0.8 ppm) + SO2 (5.0 ppm) + H2SO4 or (NH4)2SO4 (1,000 Dg/m3) at
high RH (85%) and noted a delay in early clearance of inert particles from the lungs, compared
to air-exposed controls. However, it is difficult to relate any effects to the O3 because
responses to O3 alone were not examined.
The ideal complex mixture is one that actually exists in the ambient environment.
Saldiva et al. (1992) exposed rats (Wistar, male, 2 mo old) for 6 mo to actual atmospheres of
Sao Paulo, Brazil, with controls maintained for the same period of time in a clean, rural
environment. The mean pollution levels over the exposure period were as follows: 0.011 ppm
O3, 1.25 ppm CO, 29.05 Dg/m3 SO2, and 35.18 Dg/m3 particles. The animals exposed to the
urban air showed evidence of bronchial secretory cell hyperplasia, ciliary structural changes,
increased viscosity of mucus, and impaired mucociliary clearance. Although chronic exposure
to air pollution may result in pulmonary dysfunction, the specific components producing the
response could not be determined.
Inhalation exposures to air pollutants are, of course, the ideal way to assess
interaction, but in vitro exposures may provide indications of potential interactions. Shiraishi
and Bandow (1985) exposed Chinese hamster V79 cells for 2 h to photochemical reaction
products produced from the reaction of propylene and NO2 in a smog chamber. The resultant
exposure atmospheres consisted of various proportions of propylene (0.07 to 0.16 ppm), NO2
(0.22 to 0.28 ppm), O3 (0.09 to 0.38), PAN (0.04 to 0.41 ppm), and HCHO (0.23 to 1.50
ppm). Exposures to NO2 and O3 alone also were performed. All of the complex mixtures
resulted in an increased frequency of sister chromatid exchange and growth inhibition. The
effects of the mixture were greater than those due to either O3 or NO2 alone for sister
chromatid exchange, but growth inhibition was similar to that induced solely by O3. The
authors concluded that the observed effects were not due to any single compound within the
mixture, but rather to various compounds producing multiple effects.
6.4.4 Summary
It is difficult to summarize the role that O3 plays in response to exposure to binary
mixtures, and it is even harder to determine its role in response to multicomponent
atmospheres. One of the problems in understanding interactions is that, although the specific
mechanisms of action of the individual pollutants within a mixture may be known, the exact
bases for toxic interactions have not been elucidated clearly. There are, however, certain
6-156
-------
generic mechanisms that may underlie pollutant interactions. One is physical, involving
adsorption of one pollutant onto another and subsequent transport to more or less sensitive sites
or to sites where one of the components of the mixture normally would not deposit in
concentrated amounts. This, however, probably does not play a major role in O3-related
interactions. A second mechanism involves production of secondary products that may be
more toxicologically active than the primary materials. This has been demonstrated or
suggested in a number of studies as a basis for interaction between O3 and NO2. A third
mechanism involves biological or chemical alterations at target sites that affect response to
O3 or the copollutant. This has been suggested to underlie interactions with mixtures of O3 and
acid sulfates. A related mechanism is an O3- or copollutant-induced physiological change, such
as alteration in ventilation pattern, resulting in changes in the penetration or deposition of one
pollutant when another is present. This has been implicated in enhanced responses to various
O3-containing mixtures with exercise.
Evaluation of interactions between O3 and copollutants is a complex procedure.
Responses are dependent on a number of host and environmental factors, such that different
studies using the same copollutants may show different types or magnitudes of interactions.
The occurrence and nature of any interaction is dependent on the endpoint being examined and
is also highly related to the specific conditions of each study, such as animal species, health
status, exposure method, dose, exposure sequence, and the physicochemical characteristics of
the copollutants. Because of this, it is difficult to compare studies, even those examining
similar endpoints, that were performed under different exposure conditions. Thus, any
description of interactions is really valid only for the specific conditions of the study in
question and cannot be generalized to all conditions of exposure to a particular chemical
mixture. Furthermore, it is generally not possible to extrapolate the effect of pollutant
mixtures from studies on the effects of each component when given separately. In any case,
what can be concluded from the database is that interactions of O3-containing mixtures are
generally synergistic (antagonism has been noted in a few studies), depending on the various
factors noted above, and that O3 may produce more significant biological responses as a
component of a mixture than when inhaled alone. Furthermore, although most studies have
shown that interaction occurs only at higher than ambient concentrations with acute exposure,
some have demonstrated interaction at more environmentally relevant levels (e.g., 0.05 to 0.1
ppm O3 with NO2) with repeated exposures.
6.5 Summary and Conclusions
6.5.1 Introduction
In the past 30 years, thousands of research studies on the effects of O3 in laboratory
animals have been reported in the literature. This body of evidence presents a clear picture of
the types of alterations O3 can cause on respiratory tract host defense mechanisms,
biochemistry, structure, and lung function. Less is known about carcinogenic potential and
effects on organs distant from the lungs. These types of effects are observed in many animal
species from mice to nonhuman primates, lending credence to the qualitative extrapolation of
these effects to humans. The major issue is what levels, durations, and patterns of exposure
are capable of causing these effects in humans. Extrapolation is discussed in Chapter 8.
Suffice it to say here that the animal toxicological studies assist in interpreting observations
6-157
-------
made in O3-exposed humans and extend the knowledge of potential human hazards that never
can be studied adequately in humans.
This summary and conclusion section deals exclusively with the effects of O3, alone
and in mixture. Other photochemical oxidants either have been evaluated elsewhere (NO2 and
HCHO; U.S. Environmental Protection Agency, 1993; Grindstaff et al., 1991) or in an earlier
O3 criteria document (U.S. Environmental Protection Agency, 1986). This section is
organized by molecular mechanisms of effects, respiratory tract effects, systemic effects, and
effects of mixtures. Generally, it is an interpretative, factual summary of the array of effects
observed in animals. Chapter 8 presents the current state of extrapolation of these effects to
humans, and Chapter 9 integrates knowledge from animal toxicology, epidemiology, and
human clinical studies.
Together, this chapter and the animal toxicological chapter in the 1986 document
(U.S. Environmental Protection Agency, 1986), contain more than 1,000 references.
Although all of them contribute to choosing and understanding the key issues to be summarized
here, there obviously must be a highly selective choice made as to which references to include
here. Generally, the papers discussed here were selected either because they represent the
lowest effective concentration for an endpoint or they significantly influenced a particular
conclusion.
6.5.2 Molecular Mechanisms of Effects
Molecular mechanisms (the manner in which chemical reactions of O3 are translated
into biological effects) are alluded to in different sections of this document. Studies that link
O3 chemistry with O3 effect measurements would greatly strengthen the theoretical basis for
understanding the biological effects of O3. They also would allow examination of the
similarity between animals and humans, thus strengthening interspecies extrapolations. Ozone
has been shown to react directly with a variety of biomolecules that are present in both animals
and humans. Most of the attention has been centered on polyunsaturated fatty acids and
carbon-carbon double bonds, although reactions with sulfhydryl, amino, and some electron-
rich compounds may be equally important. Free radicals may be involved, and antioxidant
defenses appear to lessen the effect of these reactions. A "molecular target" for O3 (the
biomolecules most affected by reaction with O3 or most crucial in mediating the observed
responses) has not been identified for any of the endpoints studied. In fact, the target may be
different for different endpoints.
An important concept in evaluating molecular targets was elucidated recently by
Pry or (1992), who suggests (based on reaction and diffusion rate data) that the O3 molecule
does not penetrate through cell membranes or even the surfactant layer of the lung. Instead, a
"reaction cascade" forms intermediates (organic or oxygen-free radicals, lipid hydroperoxides,
aldehydes, hydrogen peroxide, etc.), which penetrate into the cells, causing the biological
effects observed (Pryor et al., 1991). Confirmation of such O3-induced free-radical
autoxidation of lipids has been sought in vivo, but the indirect nature of the measurement
methods produced equivocal results. More direct evidence has been obtained by Kennedy
et al. (1992), who used electron spin-trapping methods to measure a concentration-related
increase in radical adducts of the lipid fraction of lungs from O3-exposed rats. Increased
radical signals were detected after a 2-h exposure to DO.5 ppm O3, but because the rats'
respiration was stimulated by CO2, the effective dose would be greater than it appears.
Oxidized (oxygenated) biomolecules that result from reaction with O3 also may mediate the
6-158
-------
effects of O3. Studies by Hatch et al. (1994) show that crude fractions of the lung lining layer
become labeled with oxygen-18 after exposure to oxygen-18-labeled O3. The label is
concentrated in the airway lining layers, and the amount of oxygen-18 incorporation in this
layer appears to be correlated with effects of O3 (permeability and inflammation) in both rats
and humans. These findings are consistent with the hypothesis that O3 reacts with the lining of
the lung, that the same types of interactions occur in both animals and humans, and that these
reactions lead to similar effects.
6.5.3 Respiratory Tract Effects
6.5.3.1 Effects on Host Defenses
Several systems defend the respiratory tract of the host against infectious and
neoplastic disease as well as nonviable inhaled particles; all of these systems can be affected by
O3. The mucociliary clearance system moves particles deposited on the mucous layer (either
through deposition from the air stream or entry of cells or cellular debris from the alveoli)
upwards and out of the lower respiratory tract. The nasal passages also have an effective
clearance system. Concentrations as low as 0.15 ppm O3 (8 h/day, 6 days) caused structural
changes in the nasal respiratory epithelium (e.g., ciliated cell necrosis, shortened cilia) of
monkeys (Harkema et al., 1987). Ciliated cells also are lost or damaged in the conducting
airways of the lower respiratory tract after short exposures (e.g., 0.96 ppm O3, 8 h, monkeys;
Hydeetal., 1992). Mucous chemistry also is changed (McBride et al., 1991). Sufficient
morphologic damage would be expected to have functional consequences. Acute exposures
(0.6 ppm O3, 2 h) slow mucociliary particle clearance in rabbits, but repeated exposures (up to
14 days) caused no effects. Alveolar clearance is slower and involves clearance of particles
through interstitial pathways to the lymphatic system or movement of particle-laden AMs up to
the bottom of the mucociliary escalator. Effects on alveolar clearance are concentration-
dependent. A single 2-h exposure of rabbits to 0.1 ppm O3 accelerated clearance up to 14 days
postexposure, exposure to 0.6 ppm caused no effect, and a higher concentration (1.2 ppm)
slowed alveolar clearance (Driscoll et al., 1986). Alveolar clearance of asbestos particles was
slowed by a 6-week exposure to an urban pattern of O3 (Pinkerton et al., 1989).
Alveolar macrophages are the first line of defense against microbes and nonviable
particles that reach the pulmonary region of the lung. They phagocytize particles, kill
microbes, and interact with lymphocytes in the development of an immune response. Thus,
their proper functioning is critical. Alveolar macrophages from several species of animals
exposed acutely to O3 can exhibit decreased phagocytosis; decreased lysosomal enzyme
activities and superoxide anion radical production, both of which function in killing bacteria;
alterations in membrane morphology; chromosomal damage; decreased cytotoxicity to tumor
cells; increased release of PGE2 and PGF2n; and alterations in the number of AMs. Phagocytic
changes are the most investigated. Exposure of rabbits to a level as low as 0.1 ppm O3
(2 h/day) decreased nonspecific phagocytosis (of latex microspheres) after 2 or 6 but not 13
days of exposure (Driscoll et al., 1987). Recovery from the single 2-h exposure to 0.1 ppm O3
was complete by 7 days postexposure. This pattern of response was confirmed in mice for Fc-
receptor mediated phagocytosis (Gilmour et al., 1991; Canning et al., 1991).
The humoral- and cell-mediated immune system of the lung also is affected by O3.
Generally, T-cell-dependent immunity is more susceptible than B-cell-dependent immunity, but
most immune functions examined have exhibited effects. However, because relatively few
studies have been conducted, it currently is not possible to adequately interpret the impacts of
6-159
-------
O3-induced alterations on the immune system (e.g., decreases in mitogenic responses of T
cells, alterations in T:B-cell ratios in the MLN). Only a few studies have attempted to
correlate immunological changes and infectious disease outcome. Van Loveren et al. (1988)
infected rats with Listeria, exposed them to O3 for 1 week (0.26 to 1.02 ppm) and measured
several endpoints. Ozone concentrations of 1.02 and 0.77 ppm, respectively, increased
Listeria-induced mortality and severity of pathologic lesions in the lung and liver. They
interpreted these findings as due to O3-induced impaired clearance of the bacteria caused by
decreased AM function and decreased cellular immunity (e.g., decreased delayed-type
hyper sensitivity and decreased T:B-cell ratios in MLN).
A reasonably large body of evidence indicates that the impact of O3 on one or
several host defense mechanisms leads to the inability of animals to fight bacterial infection and
alters the course of viral infection. Antibacterial models are more commonly used. Mice
exposed for 3 h to 0.4 ppm O3 have decreased intrapulmonary killing of S. zooepidemicus
(Gilmour et al., 1993a; Gilmour and Selgrade, 1993). Similar results have been obtained for
S. aureus at a slightly higher concentration (Goldstein et al., 1971b). Correlations have been
made between O3 exposure and decreases in AM phagocytosis, decreases in bactericidal
activity, growth of bacteria in the lungs, presence of bacteria in the blood, and mortality in
mice (Coffin and Gardner, 1972; Gilmour and Selgrade, 1993). The lowest O3 exposure
causing increased streptococcal-induced mortality is 0.08 ppm for 3 h in mice (Coffin et al.,
1967; Coffin and Gardner, 1972; Miller et al., 1978). However, prolonged intermittent
exposure to 0.1 ppm O3 for 15 weeks only slightly increased the mortality (Aranyi et al.,
1983), and continuous exposure for 15 days to 0.1 ppm with two daily 1-h peaks (5 days/week)
to either 0.3 or 0.5 ppm did not enhance mortality in the same model system (Graham et al.,
1987). Prolonged exposure (1 to 2 weeks) also did not affect bactericidal activity to S. aureus
(Gilmour etal., 1991).
Generally, short-term exposure to O3 does not affect viral titers in the lungs of mice
infected with influenza virus; however, reduced numbers of lung tissue T and B cells will
reduce antibody titers to the virus, and mortality, lung pathology, and increased lung wet
weight do occur (Selgrade et al., 1988; Jakab and Hmieleski, 1988). Ozone also enhances
postinfluenzal alveolitis and structural changes that begin at 30 days postinfection (Jakab and
Bassett, 1990). The complexity of the interaction of viral infection and O3 exposure is further
illustrated by Selgrade et al. (1988), who found that the effects of O3 on influenza virus
infection were dependent on the temporal relationship of O3 exposure and day of infectious
challenge. Also, interferon, which can be induced by viral infection, mitigates the O3-induced
lung lessons in mice, raising the possibility that certain stages of viral infection may have
interactions with the lung that are different from other stages (Dziedzic and White, 1987b).
6.5.3.2 Effects on Inflammation and Permeability
The barrier function of the respiratory tract is disrupted by O3, allowing cellular and
fluid components from the blood to enter the lung and allowing certain types of substances in
the lung to enter the blood. Markers of inflammation generally included increased proteins and
PMNs in BAL. Concurrent with these events, but not necessarily interdependently, AMs
liberate more arachidonic acid, which results in the production of biologically active LTs and
PGs. Similar responses are observed in mice, rats, rabbits, guinea pigs, hamsters, and
nonhuman primates. After acute exposure, the lowest effective concentration that increases
BAL protein and number of PMNs is 0.12 ppm O3 (mice, 24 h of exposure, BAL immediately
6-160
-------
after exposure) (Kleeberger et al., 1993a). However, the increase in BAL protein typically is
maximal roughly 16 to 24 h postexposure. In rats exposed to 0.8 ppm O3 for 6 h and
examined by lavage and morphometry several times postexposure, the increase in nasal PMNs
occurs sooner and wanes about the time that these cells are increasing in number in the lungs
(Hotchkiss et al., 1989a,b). It should be recognized that BAL can enable measurement of the
protein and cells accessible by lavage, including the resident material (i.e., may include protein
from O3-induced cellular destruction) and the material entering from the tissue or circulation.
Thus, interstitial inflammation, which has been observed in several species microscopically, is
not detectable by BAL.
Several C x T studies have been conducted in mice using BAL protein as an
endpoint. In two studies, there was combination of various Cs (0.1 to 2.04 ppm O3) and
Ts (1 to 12 h), resulting in a number of different C X T products (Rombout et al., 1989;
Highfill et al., 1992). Both of these studies showed that the influence of T increased as
C increased (i.e., there was no simple relationship of Ca x Ta = constant product; however, at
the lowest C x T products, there was a more equivalent influence of C and T). Gelzleichter
et al. (1992b) used a single C x T product composed of a variety of Cs and Ts for up to
3 days of exposure. The 24 h/day exposure group had less response than the other groups that
responded equivalently. Effects of longer term exposure on permeability and inflammation are
more complex to interpret (also see subsequent discussion on lung structure). Histological
examination of rat lungs exposed to 0.5 ppm O3 (2.25 h/day) showed more inflammatory cells
in the alveoli after 5 days of exposure than after 1 day of exposure (Tepper et al., 1989). In
contrast, the increase in BAL PMNs that occurred after Day 1 of exposure of rats had resolved
by Day 4 (7 h/day) (Donaldson et al., 1993).
Some studies suggest that, although protein and PMN increases are observed
concurrently, this may be more a function of experimental design than the actual biological
sequence of events. For example, in rats depleted of PMNs with anti-PMN serum, O3 did not
increase BAL PMNs, but BAL protein still was increased (Pino et al., 1992b). Also Young
and Bhalla (1992) observed an increase in tracheal protein earlier than increased tracheal
PMNs. They interpreted this and other related results to suggest that the recruited PMNs may
serve to sustain an increase in permeability.
6.5.3.3 Effects on Structure, Function, and Biochemistry
Theoretically, and in some cases empirically, lung structure, function, and
biochemistry are linked. Correlations are not exact because of differences in available
measurement methods (e.g., most lung function tests used do not measure sensitively the
function of the smallest airways, where the "classical" O3 lesion is observed) and some
independence of effects (e.g., a transient change in breathing frequency would not be
morphologically detectable). Also, most biochemical measurements are made of whole lung,
rather than focal areas of damage, and only some enzyme activities measured would be
expected to be correlated to structure or function (e.g., collagen metabolism, antioxidant
metabolism).
After acute exposure to O3, the most commonly observed effect in several species is
tachypnea (increased f and decreased VT) with little (if any) change in VE; the lowest exposure
causing tachypnea was 0.2 ppm O3 for 3 h in rats (Mautz and Bufalino, 1989). Other effects
reported after acute exposure to Dl ppm include increased RL and decreased Cdyn, TLC, VC,
FRC, RV, FVC, DLCO, and the multibreath N2 slope (e.g., Fouke et al., 1991; Mautz et al.,
6-161
-------
1985b; Miller et al., 1988). However, these changes are not observed in all studies, probably
due to differences in animal species, measurement method, and exposure protocols. With rare
exception, concentrations well in excess of 1 ppm O3 are required to increase airway reactivity.
Two C x T studies of pulmonary function using acute exposure periods have been
performed. Costa et al. (1989) found that FVC, DLCO, and the multibreath N2 slope decreased
with increasing C X T products in rats and that the influence of T is greater at higher Cs.
In guinea pigs, Nishikawa et al. (1990) observed that airway responsiveness to methacholine
increased at higher C X T products (e.g., at 90 but not at 30 ppm • min); the authors
concluded that T was an important factor in the O3 response.
When rats were exposed for 5 days (2.25 h/day, with CO2 to stimulate ventilation
equivalent to light exercise in humans) to 0.35, 0.5, and 1.0 ppm O3, the change in shape of
the flow-volume curve occurred and tachypnea peaked on Days 1 and 2, but by Day 5, there
was no difference from control (except at 1 ppm) (Tepper et al., 1989). This attenuation is
similar to that observed in humans. However, in other, similar groups of animals, histological
changes in the lung progressed, and BAL protein remained elevated. Other similarities
between laboratory animals and humans in their pulmonary function responses to short-term O3
exposure are explored in Chapter 8.
Ozone causes similar types of alterations in lung morphology in all laboratory
animal species studied. The most affected cells are the ciliated epithelial cells of the airways
and Type 1 cells in the gas exchange region. Within the nasal cavity, anterior portions of the
respiratory and transitional epithelium are affected. Cilia are lost or damaged; some ciliated
cells become necrotic, are lost, and are replaced with nonciliated cells. Mucus-secreting cells
are affected.
The CAR (the junction of the conducting airways and the gas exchange region) is a
primary target, possibly because it receives the greatest dose of O3 delivered to the lower
respiratory tract (see Chapter 8) and has Type 1 epithelial cells covering a large surface area.
Even though there are significant interspecies differences in the structure of the CAR (e.g.,
primates, including humans, have RBs, which are rudimentary or absent in laboratory animals
such as rats or mice), it is the target in all species studied. Exposure to O3 causes loss of cilia
or necrosis of the ciliated cells, leaving a bare basement membrane that is replaced by
nonciliated bronchiolar cells, which may become hyperplastic after longer exposures. Mucous
secreting cells can be affected, but not as significantly as ciliated cells. Type 1 cells also are
damaged and can be sloughed from the surface; Type 2 cells, which are thicker, replace them.
Sometimes, Type 2 cells differentiate into Type 1 cells. This epithelial remodeling is
accompanied by an inflammatory response in the CAR, primarily consisting of an increase in
number of PMNs in the earlier stages and an increase in number of AMs in later stages;
interstitial edema occurs. With increased duration of exposure, alveolar septa in the CAR
thicken due to increased matrix, basement membrane, collagen, and fibroblasts and a thickened
alveolar epithelium.
These patterns of change have different relationships to duration of exposure, as
illustrated by Dungworth (1989) (see Figure 6-3; Section 6.2.4.5). Inflammatory changes peak
after a few days of O3 exposure; are still observable, but to a much lesser degree, in tissue
during months of exposure; and begin to return to control values after exposure ceases. In
contrast, epithelial hyperplasia rapidly increases during about the first week of exposure,
plateaus as exposure continues, and begins to decrease slowly when exposure stops. Interstitial
6-162
-------
fibrosis requires months of exposure to be observed microscopically and increases slowly, but
when exposure ceases, interstitial fibrosis still can persist or continue to increase. Numerous
studies using several different species and experimental approaches support these findings.
Only a few of the studies (primarily those using more sensitive morphometric measurements)
are used here to illustrate key points and to show correlations with pulmonary function and
lung biochemistry. Only rat and nonhuman primate studies are discussed because most
investigations were conducted on them. At equivalent exposures, nonhuman primates appear
to be more responsive than rats (Section 6.2.4).
Generally, short-term exposures to concentrations DO.2 ppm O3 do not cause
changes detectable by LM in the nasal cavities of rats or nonhuman primates, except for
inflammation and an occasional delayed postexposure finding of mild hyperplasia. For
example, Hotchkiss et al. (1989a) reported inflammation in the nasal epithelium of rats up to
66 h after a 6-h exposure to levels as low as 0.12 ppm O3; there was no necrosis, loss of cilia,
or hyperplasia even at 1.5 ppm. After 3 days (22 h/day) of exposure, DO.4 ppm caused loss of
cilia and hyperplasia and metaplasia of the nasal epithelium of rats (Reuzel et al., 1990).
Nonhuman primates appeared to be more responsive. Harkema et al. (1987) observed that
exposure to 0.15 or 0.3 ppm O3 for 6 or 90 days (8 h/day) caused necrosis of ciliated cells,
shortened cilia, and increased mucous granule cells in the respiratory epithelium; alterations in
cell numbers also were found in the transitional epithelium.
Within the CAR, a number of alterations occur. In rats and monkeys, ciliated and
Type 1 cells become necrotic and are sloughed from the epithelium as early as the first 2 to 4 h
of an exposure to about 0.5 ppm O3 (Stephens et al., 1974a,b). Repair, as shown by increased
DNA synthesis by nonciliated bronchiolar and Type 2 cells, begins by about 18 to 24 h of
exposure (Evans et al., 1976a,b; Stephens et al., 1974a; Castleman et al., 1980), although cell
damage continues (Castleman et al., 1980). The lesion is fully developed by about 3 days of
continuous exposure, after which the rate of repair exceeds the rate of damage. The increase
in antioxidant enzyme activities (e.g., succinate oxidase, G6PD, and 6PGD) parallels the
increase in Type 2 cells, which are rich in these enzymes; the increase in the Type 2 cell
population is probably responsible for these biochemical changes (Bassett et al., 1988a; U.S.
Environmental Protection Agency, 1986).
Lesions in the CAR are one of the hallmarks of O3 toxicity, having been well
established. The study by Chang et al. (1992) provides examples of some of the patterns of
cellular alterations. Chang et al. (1992) exposed rats to an urban pattern of O3 (0.06 ppm
background, 7 days/week on which were superimposed 9-h peaks [5 days/week] slowly rising
to 0.25 ppm) for 78 weeks and made periodic examinations of the CAR TB and proximal
alveoli by TEM morphometry during and after exposure. Type 1 cells had a larger volume at
Week 13 and increased numbers at Weeks 13 and 78; there were no such changes at 17 weeks
after exposure ceased. Type 2 cell volume per area of basement membrane increased
immediately after Week 78 and was still increased 17 weeks after exposure ceased. Interstitial
cells and matrix were increased after Weeks 1, 13, and 78, but returned to control by 17 weeks
after exposure ceased. However, epithelial and endothelial basement membrane were
thickened and accompanied by increased collagen fibers at the later examination times and 17
weeks after the 78-week exposure ended. In TBs, surface areas of ciliated and nonciliated cells
decreased during exposure. Pulmonary function studies conducted in identically exposed
groups of rats were consistent with the morphometric findings (Tepper et al., 1991).
Generally, expiratory resistance was increased (suggesting central airway narrowing), but it
6-163
-------
was only significantly different from control at 78 weeks. Tidal volume was reduced at all
evaluation times. Overall, breathing frequency was reduced, but no single evaluation time was
significant. Monkeys exposed to a higher concentration of O3 (0.64 ppm, 1 year) also showed
increased resistance and decreased flows, which were interpreted as central and peripheral
airway narrowing; during a 3-mo postexposure period, decreases in static lung compliance
persisted (Wegner, 1982).
Several studies have demonstrated distal airway remodeling. This bronchiolization
of CAR alveoli is so named because bronchial epithelium replaces the Type 1 and 2 cells
typical of ADs, resulting in the appearance of RBs in rats and increased volume fraction and
volume of RBs in monkeys. This has been observed at exposures as low as 0.5 ppm O3
(50 days) in rats (Moore and Schwartz, 1981) and as low as 0.25 ppm (8 h/day, 18 mo) in
monkeys (Tyler et al., 1988). Inflammation occurs concurrently, perhaps indicating an
influence on remodeling. In monkeys, such bronchiolization can persist 6 mo after the end of
a 1-year (8 h/day) exposure to 0.64 ppm (Tyler et al., 1991b).
Exposure regimens can have unexpected impacts on experimental outcomes.
Several investigations of combinations of O3 "episodes" or O3 "seasons" with clean-air periods
have been examined. In the first of these, Last et al. (1984b) compared air control rats to
two groups of rats exposed to 0.96 ppm. One group received a 90-day (8 h/day) exposure
("daily"); the other group had intermittent units of 5 days of O3 (8 h/day) and 9 days of air,
such that there were 35 O3 exposure days over the 90-day period (episodic). Both groups had
equivalent increases in lung collagen. Using a similar exposure regimen, Barr et al. (1990)
found equivalent CAR remodeling and volumes of CAR lesions in both groups. In contrast,
RB thickness increased in the daily group only, and the CAR interstitium increased in
thickness only in the episodic group. Monkeys were studied more extensively after a daily (8
h/day) exposure to 0.25 ppm for 18 mo and a seasonal exposure only during the odd months of
the 18-mo period (Tyler et al., 1988). Most morphometric measurements were similar
between the two groups (e.g., both had respiratory bronchiolitis). However, only the daily
group had an increased number of AMs in the lumen and interstitium. Only the seasonal group
had increased lung collagen content; increased chest wall compliance, suggesting delayed lung
maturation; and increased inspiratory capacity. This body of work indicates that under these
types of exposure circumstances, the simple product of C x T does not predict the outcome.
Indeed, half the O3 (on a C x T basis) caused equivalent or more effects than a "full"
O3 exposure.
The complexity of understanding C X T relationships is further illustrated by
Chang et al. (1991), who compared two different exposure regimens (one a square wave and
the other an urban pattern) on the basis of C x T products. There was a linear relationship
between C X T products and the increase in Type 1 cell volume in the CAR; a similar
observation on Type 2 cell volumes was less robust. There was no such relationship for other
morphometric endpoints in the same animals. Cell proliferation in the nasal epithelium does
not increase linearly with increasing C X T but does increase linearly with increasing
C (Henderson etal., 1993).
Long-term exposure also thickens CAR alveolar septa, due to an increase in
inflammatory cells, fibroblasts, and amorphous extracellular matrix (Fujinaka et al., 1985;
Barry et al., 1985; Zitnik et al., 1978). There is some morphological evidence of mild fibrosis
(i.e., local increase in collagen) in CAR inter alveolar septa (Last et al., 1979; Boorman et al.,
1980; Chang et al., 1992; Pickrell et al., 1987b; Freeman et al., 1974; Moore and Schwartz,
6-164
-------
1981). Biochemical evidence supports these findings, even though biochemical approaches
would be expected to be less sensitive because the whole lung (rather than focal lesions) is
examined. Last et al. (1979) directly demonstrated the correlation by observing increased
collagen histologically and biochemically (collagen synthesis rate) in rats similarly exposed to
0.5 to 2.0 ppm O3 for 7 to 21 days. The increase became greater with increasing concentration
and duration of exposure. Similar correlations were observed at a higher concentration by
Pickrell et al. (1987b). The increased collagen content can persist after exposure ceases
(Chang et al., 1992; Hussain et al., 1976a,b; Last et al., 1984b), but some studies suggest that
higher concentrations (>0.5 ppm) may be required for such persistence (Last and Greenberg,
1980; Pickrell et al., 1987b). Collagen cross-links were studied in monkeys exposed to 0.61
ppm O3 for 1 year (8 h/day) (Reiser et al., 1987). Earlier examination of these same monkeys
revealed that collagen content was increased (Last et al., 1984b). When specific collagen
cross-links were measured, the increase in "abnormal" cross-links observed immediately after
exposure remained in the lungs at 6 mo postexposure.
These morphologic/morphometric and biochemical findings of fibrotic changes are
supported by some pulmonary function studies. For example, rats exposed for up to 78 weeks,
using the same urban exposure protocol as Chang et al. (1992), exhibited reduced lung volume
and hastened N2 washout patterns, consistent with a "stiffer" lung (i.e., restrictive lung
disease) (Costa et al., 1994).
The chronic O3 study by the NTP and the Health Effects Institute (HEI) (Last et al.,
1994; Szarek, 1994; Radharkrishnarmurthy, 1994; Parks and Roby, 1994; Harkema and
Mauderly, 1994; Harkema et al., 1994; Chang et al., 1995; Pinkerton et al., 1995; Catalano et
al., 1995a,b) further illustrates some of the complex interrelationships between lung structure,
function, and biochemistry. All of these endpoints were evaluated in a collaborative project
using rats exposed 6 h/day, 5 days/week for 20 mo to 0.12, 0.50, or 1.00 ppm O3. Although
lung biochemistry and structure were affected at the higher O3 concentrations (DO.50 ppm),
there were no observed effects on pulmonary function. This is consistent with the relative
sensitivity of the tests used and suggests that the observed effects were not sufficient to
overcome the reserve function of the lung.
Combined analyses of the NTP/HEI collaborative studies showed that 0.50 and
1.00 ppm O3 caused a variety of structural and biochemical effects; 0.12 ppm O3 did not cause
any major effects, although a few specific endpoints were altered. Hallmarks of chronic
rhinitis (e.g., inflammation, mucous cell hyperplasia, decreased mucous flow) were observed
in focal regions of the nasal cavity. Structural and biochemical changes included some, but not
many hallmarks of airway disease. Typical O3-induced changes (e.g., bronchiolarization,
increased interstitial matrix) observed in the tracheobronchial region and in the CAR were
characteristic of centriacinar fibrosis; however, diffuse pulmonary fibrosis was not observed.
An integrative, multiple endpoint analysis (Catalano et al., 1995a) utilizing median
polish techniques produced composite variables for disease surrogates that were tested for
trends across all three O3 concentrations. Trends for centriacinar fibrosis, airway disease, and
chronic rhinitis were examined for 10, 18, and 3 endpoints, respectively, from the individual
NTP/HEI studies. A statistically significant trend was noted for the association between
chronic rhinitis and increasing O3 concentration. The differences between control and exposed
rats were statistically significant at 0.50 and 1.00 ppm O3. Marginally significant and
significant trends were found for the association between centriacinar fibrosis or airway disease
6-165
-------
and increasing O3 concentration; however, no statistically significant differences were found
between control and O3-exposed rats.
As discussed above, long-term O3 exposure can cause lung fibrotic changes;
however, there is no evidence that O3 causes emphysema, using the currently accepted
morphological definition of human emphysema (U.S. Environmental Protection Agency,
1986).
6.5.3.4 Genotoxicity and Carcinogenicity of Ozone
A significant amount of research has been conducted to determine whether O3 is
genotoxic or carcinogenic. Many of the early experiments have flaws in experimental design
or have used O3 concentrations far above levels that could occur in ambient air. In evaluating
the data, a number of conclusions can be made. In vitro exposure of naked plasmid DNA to
very high O3 concentrations results in single and double-strand breaks in the DNA, as
confirmed by gel electrophoresis and electron microscopy studies (Hamelin, 1985). Testing of
O3 in various mutagenesis assays has led to marginal or small results in a number of assays and
negative results in other assay systems. Ozone is not mutagenic in Salmonella strains TA98,
TA100, TA104, and TA1535 and causes, at most, weak effects in strain TA102 that are not
strictly concentration dependent (Dillon et al., 1992; Victorin and Stahlberg, 1988a,b).
Extremely high concentrations of O3 (50 ppm) caused mutation to streptomycin resistance in
E. coli and caused various types of mutations in the yeast S. cerevisiae, but O3 was a weak
mutagen compared to known strong mutagens in the yeast system (L1 Herault and Chung,
1984; Dubeau and Chung, 1982). Ozone was not mutagenic in the N. tabacam or
Tradescantia mutation assay systems (Gichner et al., 1992). Hence, overall, the data on the
mutagenicity of O3 are mixed: negative in six assays, marginally positive in one assay, and
weakly positive in two assays. The present data indicate that O3 is, at most, a weak mutagen,
but further data are needed in mammalian cell systems to draw definitive conclusions regarding
this point. There are some data indicating that O3 may cause chromosome breakage in cultured
cells, but in vivo animal studies are conflicting (Zelac et al., 1971a,b; Tice et al., 1978).
A human study with an appropriate experimental design was negative (McKenzie et al., 1977;
McKenzie, 1982).
Regarding carcinogenicity, O3 has been shown to induce morphological
transformation in cultured C3H/10T1/2 mouse embryo cells and in SHE cells and to cause a
synergistic morphological transformation in cells treated also with gamma radiation (Borek
et al., 1986, 1989b). However, these results could be due to interactions of O3 with the culture
medium that generate chemical species different from those produced in vivo. Whole animal
carcinogenesis assays performed in strain A mice have demonstrated marginal increases in
tumor yield that were not statistically significant or concentration dependent (Hassett et al.,
1985; Last et al., 1987). The NTP study demonstrated that O3 was not a tumor promoter or a
co-carcinogen when NNK-treated male F344/N rats were exposed for 2 years to 0.5 ppm O3
(National Toxicology Program, 1994). In the NTP study, rats and mice were exposed to 0.12,
0.5, or 1.0 ppm O3 for 6 h/day, 5 days/week for two years or a lifetime. This NTP study
showed no evidence of carcinogenic activity in male or female F344/N rats, equivocal evidence
of carcinogenic activity in male B6C3F, mice, and some evidence of carcinogenic activity in
female B6C3F, mice at a high concentration (1.0 ppm). Hence, O3 has been shown to be a
weak pulmonary carcinogen only in female B6C3F, mice at toxic concentrations in one
experiment.
6-166
-------
At present, O3 is shown to be nonmutagenic in some assay systems; at most, weakly
mutagenic in a few assay systems; and clastogenic in vitro but not in vivo. Ozone can
transform cells in vitro. Ozone does not cause concentration-dependent tumor induction that is
statistically significant in hamsters, Wister male rats, F344/N male or female rats, male or
female A/J mice, or Swiss-Webster male mice. There are ambiguous data for pulmonary
carcinogenesis in male B6C3F, mice and weak carcinogenesis data in female B6C3F, mice
from chronic exposure to 1.0 ppm O3. Therefore, O3 has been shown to be a carcinogen only
in female B6C3F, mice in one experiment. Because a chronic exposure to 1 ppm O3 was
required to induce pulmonary tumors in female mice, it is possible that pulmonary toxicity,
which occurs only at high O3 concentrations (1.0 ppm) and does not occur at lower levels,
contributed to the tumor development. Hence, the potential for animal carcinogenicity is
uncertain at the present time.
6.5.3.5 Factors That Influence Ozone Exposure
Factors that increase the delivered dose of O3, decrease biochemical defense
mechanisms, or increase cellular sensitivity can increase the impact of a given O3 exposure.
The most commonly studied factors include exercise, age, and nutrition.
As discussed in Chapter 8, exercise increases the dose of O3 delivered to the
respiratory tract and alters the distribution of O3. As would be expected, exercise during
exposure enhances the effect of O3. This has been demonstrated by Mautz et al. (1985b), who
showed that exercising rats had more extensive lung lesions than rats exposed at rest.
Similarly, Tepper et al. (1990, 1994) found that rats were more responsive to O3 when
coexposed to CO2 to increase ventilation, simulating exercise.
A number of studies have been conducted to compare the effects of O3 on various
ages of mice and rats, from 1 day old to older adults. Interpretation of these studies is difficult
because, prior to weaning, the huddling behavior of the neonates with their dams as well as the
bedding material (present in some studies) may have affected the concentration of O3 in the
breathing zone and hence the subsequent delivered dose. Generally, in short-term exposure
biochemical studies of antioxidant metabolism, there was a decrease or no change in enzyme
activity in neonates. As age increased after weaning, the typical increase in antioxidant
metabolism became greater with age (Elsayed et al., 1982; Tyson et al., 1982; Lunan et al.,
1977; Mustafa et al., 1985). Stephens et al. (1978) found that morphological effects did not
occur in animals exposed prior to weaning at 21 days of age. This may explain the results of
Barry et al. (1985, 1988), who found no morphometric differences in the CAR and TB in rats
that started a 42-day exposure at ages of 1 day and 42 days. In identically exposed rats,
however, Raub et al. (1983) found more, though admittedly subtle, pulmonary function
changes in the youngest group of animals. Yokoyama et al. (1984) did not detect any age-
related differences in lung function of rats at 4, 7, and 10 weeks of age. Although O3-induced
increases in BAL protein and PMNs do not show age dependence, BAL prostaglandins
increased sooner and more leukocytes were dead in younger (13-day-old) rats, compared to
adults (e.g., 16 weeks old) (Gunnison et al., 1990, 1992a). Age (5 weeks versus 9 weeks) did
not influence the O3-induced decrease in lung bactericidal activity (Gilmour et al., 1993a).
The literature on O3-exposed pregnant animals is extremely sparse. Exposure of
rats (1 ppm O3 6 h) on Day 17 of pregnancy or Days 3,13, and 20 of lactation caused a
greater increase in lung permeability and inflammation than that observed in nonpregnant rats
(Gunnison etal., 1992b).
6-167
-------
Numerous reports document that animals made vitamin E deficient are more
susceptible to the biochemically detected effects of O3 (e.g., lipid changes, antioxidant
metabolism changes) (U.S. Environmental Protection Agency, 1986; Pryor, 1991). Generally,
the research shows that, although vitamin E deficiency enhances susceptibility to lung
biochemical changes, there is not a proportionate relationship between vitamin E
supplementation (above normal levels) and protection from O3. Also, vitamin E deficiency did
not alter the impact of O3 on lung structure (Chow et al., 1981). Vitamin C deficiency also has
an influence. Guinea pigs deficient in vitamin C had a greater increase in BAL protein
(compared to vitamin C-normal animals) when exposed acutely to 0.5 but not 1.0 ppm O3
(Sladeetal., 1989).
6.5.4 Systemic Effects
Theoretical analyses (Pryor, 1992) indicate that the O3 molecule does not penetrate
to the blood, yet there are numerous reports of systemic effects (i.e., effects on lymphocytes,
erythrocytes, serum, central nervous system, parathyroid gland, circulatory system, and liver).
Possibly one or several of the reaction products of O3 (see Section 6.2.1) penetrates the lung
tissue, or perhaps some systemic responses are secondary to pulmonary effects. Although a
variety of clinical chemistry changes occur after O3 exposure, they cannot be interpreted and
will not be discussed here (see U.S. Environmental Protection Agency, 1986, and Section 6.3).
Effects on systemic immunity are discussed in Section 6.5.3.1.
6.5.4.1 Central Nervous System and Behavioral Effects
Acute exposure to O3 caused transient changes in behavior. The lowest exposure
causing effects was 0.12 ppm O3 for 6 h in rats; wheel-running activity decreased (Tepper
et al., 1985; Tepper and Weiss, 1986). Because exercising animals were exposed in these
studies (i.e., they received a higher dose of O3), it is not surprising that higher
O3 concentrations (0.5 ppm, 6 h) are required to affect sedentary behavior (e.g., operant
behaviors such as lever pressing for food reinforcement) (Weiss et al., 1981). Mice show
aversive responses to O3 (0.5 ppm, 60 s) by terminating O3 exposure (Tepper et al., 1985).
The lowest exposures causing effects are impacted by the type of reward. For example,
O3 had less effect on behaviors to avoid electric shock (Ichikawa et al., 1988) than on
behaviors to obtain food or access exercise (Tepper et al., 1982, 1985; Weiss et al., 1981).
6.5.4.2 Cardiovascular Effects
In rats, O3 can cause bradyarrhythmia at exposures as low as 0.1 ppm for 3 days;
bradycardia, at exposures as low as 0.2 ppm for 2 days; and decreased mean arterial blood
pressure, at exposures as low as 0.5 ppm for 6 h (Arito et al., 1990, 1992; Uchiyama and
Yokoyama, 1989; Watkinson et al., 1993; Yokoyama et al., 1989b; Uchiyama et al., 1986).
There is an interaction between some of these responses and thermoregulation in the rat. For
example, when heart rate decreased, the core temperature of the exposed rats also decreased,
and when exposures were conducted at higher ambient temperatures, there was no change in
core temperature or heart rate (Watkinson et al., 1993). Such interactions add to the
complexity of extrapolating this type of response to humans, and therefore, without more
information, qualitative extrapolation would be highly speculative.
6.5.4.3 Reproductive and Developmental Effects
6-168
-------
No reports of "classical" (e.g., 2-generation studies) reproductive assays with
O3 were found. Kavlock et al. (1979, 1980) performed several developmental toxicity
experiments in rats. Pregnant rats exposed intermittently (8 h/day) to 0.44 to 1.97 ppm
O3 during early, mid-, or late gestation or during the entire period of organgensis (Days 6 to
15) had no significant teratogenic effects. Continuous exposure during mid-gestation increased
the resorption of embryos. Postnatal growth and behavioral development also were
investigated. There was no effect on neonatal mortality (up to 1.5 ppm). Pups from dams
exposed continuously to 1 ppm during mid- or late gestation weighed less 6 days after birth.
Pups from pregnant rats exposed continuously to 1 ppm during late gestation had delays in
behavioral development (e.g., righting, eye opening).
6.5.4.4 Other Systemic Effects
A number of investigations have shown the effects of O3 on the pituitary-thyroid-
adrenal axis, as evidenced by changes in circulating hormones and morphological changes in
the thyroid and parathyroid glands (U.S. Environmental Protection Agency, 1986). No more
recent studies could be found.
Several approaches have been used to study the effects of O3 on the liver: increase
in sleeping time following the injection of drugs (e.g., pentobarbital) metabolized by the liver,
drug pharmacokinetics, and changes in liver enzymes. The lowest exposure causing increased
sleeping time from pentobarbital was 0.1 ppm O3 for at least 15 or 16 days (3 h/day) in female
mice (Graham et al., 1981). In three species of animals, only females were affected (Graham
et al., 1981). Pentobarbital pharmacokinetics was marginally (p = 0.06) slowed in mice
exposed to 1 ppm O3 for 3 h (Graham et al., 1985); theophylline clearance was slowed in older
rabbits exposed to 0.3 ppm O3 for 5 days (3.75 h/day) (Canada and Calabrese, 1985). Ozone
has caused both increases, decreases, and no changes in liver xenobiotic metabolism,
depending on the exposure and enzyme being measured (U.S. Environmental Protection
Agency, 1986).
6.5.5 Effects of Mixtures
Humans in the real world are exposed to complex mixtures of gases and particles.
Sufficient evidence exists to know that the health outcome is dependent on the mixture, but the
relative role (or even the exact identity) of the "major" components is not known. Because of
this, it is crucial to evaluate the health effects of O3 in light of epidemiological, human clinical,
and animal toxicological studies. For the purposes of this document, an interaction is
considered to occur when the response to the mixture is statistically significantly higher
(synergism) or lower (antagonism) than the sum of the individual pollutants. Most animal
toxicological studies of O3 interactions have been conducted with binary mixtures
(predominantly NO2 and H2SO4). The rarer reports on complex mixtures are interesting, but
less helpful because often the studies did not include a group exposed only to O3, and therefore
knowledge of the role of O3 is confounded. Thus, only the binary mixture studies will be
summarized here. This research has demonstrated that exposure to O3 in combination with
another chemical can result in antagonism, additivity, or synergism, depending on the animal
species, exposure regimen, and endpoint studied. Interpretation is further complicated by the
fact that most studies used exposure regimens unlike the real world in terms of ratios of
pollutant concentrations, "natural" sequencing of exposure patterns, and other factors. For
example, when O3 and NO2 exposures were sequential (in any order), there was an additive
6-169
-------
increase in BAL protein, as compared to a synergistic increase when the exposures were
concurrent (Gelzleichter et al., 1992a).
A range of interactions has been shown with O3 and NO2 combinations. For
example, a 2-week exposure to an O3-NO2 mixture (0.4 ppm of both) synergistically increased
antioxidants in the lungs of rats but not guinea pigs, peroxide levels were synergistically
increased in guinea pigs but not rats, and GST activity was decreased in guinea pigs and
unchanged in rats (Ichinose and Sagai, 1989). Most of the interaction studies using lung
biochemical endpoints display synergism. A rare exception was the antagonism to the increase
in lung cytochrome P-450 content caused by 0.2 ppm O3 (1 to 2 mo) when the rats were
coexposed to 4 ppm NO2 (Takahashi and Miura, 1989). Combinations of various acute
exposure durations and of O3 and NO2 concentrations did not follow a C X T relationship for
increased lung permeability, but were synergistic at higher C x T products (Gelzleichter et al.,
1992b). For pulmonary host defenses against bacterial infection, the interaction is dependent
on the exposure pattern. Graham et al. (1987) showed that a 15-day exposure of mice to
mixtures of O3 and NO2, each having a baseline level with two daily 1-h peaks of the pollutant,
resulted in synergism only when exposure to either gas alone caused an increase in bacterial-
induced mortality.
Both synergistic and antagonistic interactions have been found with combinations of
O3 and acidic sulfates. Warren et al. (1986) reported that with 3 days of exposure to 0.2 ppm
O3 + 5 mg/m3 (NH4)2SO4, O3 alone was responsible for increasing BAL protein, collagen
synthesis rate, and other parameters, but, by 7 days of exposure, synergism occurred. When
rabbits were exposed for 4 mo (2 h/day, 5 days/week) to 0.1 ppm O3 + 125 Dg/m3 H2SO4,
there was a synergistic increase in epithelial secretory cell number, whereas 8 mo of exposure
resulted in antagonism (Schlesinger et al., 1992a). Antagonism also was observed for effects
or certain AM functions after acute exposures to O3-H2SO4 mixtures (Schlesinger et al.,
1992b). Sequential exposures to O3 and H2SO4 also have been examined. Exposure to O3 did
not influence the subsequent effects of H2SO4 on bronchoconstriction in guinea pigs (Silbaugh
and Mauderly, 1986). Gardner et al. (1977) found an additive increase in bacterial infectivity
when mice were exposed acutely to 0.1 ppm O3 before (but not after) H2SO4.
In summary, the animal toxicological studies clearly demonstrate the major
complexities and potential importance of interactions, but do not provide a scientific basis for
predicting the results of interactions under untested ambient exposure scenarios.
6-170
-------
References
Abraham, W. M.; Januszkiewicz, A. J.; Mingle, M.; Welker, M.; Wanner, A.; Sackner, M. A. (1980)
Sensitivity of bronchoprovocation and tracheal mucous velocity in detecting airway responses to Q.
J. Appl. Physiol.: Respir. Environ. Exercise Physiol. 48: 789-793.
Abraham, W. M.; Delehunt, J. C.; Yerger, L.; Marchette, B.; Oliver, W., Jr. (1984) Changes in airway
permeability and responsiveness after exposure to ozone. Environ. Res. 34: 110-119.
Abraham, W. M.; Sielczak, M. W.; Delehunt, J. C.; Marchette, B.; Wanner, A. (1986) Impairment of tracheal
mucociliary clearance but not ciliary beat frequency by a combination of low level ozone and sulfur
dioxide in sheep. Eur. J. Respir. Dis. 68: 114-120.
Aizawa, H.; Chung, K. P.; Leikauf, G. D.; Ueki, I.; Bethel, R. A.; O'Byrne, P. M.; Hirose, T.; Nadel, J. A.
(1985) Significance of thromboxane generation in ozone-induced airway hyperresponsiveness in dogs.
J. Appl. Physiol. 59: 1918-1923.
Allegra, L.; Moavero, N. E.; Rampoldi, C. (1991) Ozone-induced impairment of mucociliary transport and its
prevention with N-acetylcysteine. Am. J. Med. 91(suppl. 3C): 67S-71S.
Alpert, S. M.; Gardner, D. E.; Hurst, D. J.; Lewis, T. R.; Coffin, D. L. (1971) Effects of exposure to ozone on
defensive mechanisms of the lung. J. Appl. Physiol. 31: 247-252.
Alpert, S. E.; Kramer, C. M.; Hayes, M. M.; Dennery, P. A. (1990) Morphologic injury and lipidperoxidation
in monolayer cultures of rabbit tracheal epithelium exposed in vitro to ozone. J. Toxicol. Environ.
Health 30: 287-304.
Amdur, M. O.; Ugro, V.; Underbill, D. W. (1978) Respiratory response of guinea pigs to ozone alone and with
sulfur dioxide. Am. Ind. Hyg. Assoc. J. 39: 958-961.
Anioruso, M. A.; Ryer-Powder, J. E.; Warren, J.; Witz, G.; Goldstein, B. D. (1989) Effects of ozone on the
production of active bactericidal species by alveolar macrophages. In: Schneider, T.; Lee, S. D.;
Wolters, G. J. R.; Grant, L. D., eds. Atmospheric ozone research and its policy implications:
proceedings of the 3rd US-Dutch international symposium; May 1988; Nijmegen, The Netherlands.
Amsterdam, The Netherlands: Elsevier Science Publishers; pp. 501-511. (Studies in environmental
science 35).
Aranyi, C.; Vana, S. C.; Thomas, P. T.; Bradof, J. N.; Fenters, J. D.; Graham, J. A.; Miller, F. J. (1983)
Effects of subchronic exposure to a mixture of O3, SO2, and (NH4)2SO4 on host defenses of mice.
J. Toxicol. Environ. Health 12: 55-71.
Arito, H.; Uchiyama, L; Arakawa, H.; Yokoyama, E. (1990) Ozone-induced bradycardia and arrhythmia and
their relation to sleep-wakefulness in rats. Toxicol. Lett. 52: 169-178.
Arito, H.; Uchiyama, L; Yokoyama, E. (1992) Acute effects of ozone on EEG activity, sleep-wakefulness and
heart rate in rats. Ind. Health 30: 23-34.
Atwal, O. S.; Wilson, T. (1974) Parathyroid gland changes following ozone inhalation: a morphologic study.
Arch. Environ. Health 28: 91-100.
Banks, M. A.; Porter, D. W.; Martin, W. G.; Castranova, V. (1991) Ozone-induced lipid peroxidation and
membrane leakage in isolated rat alveolar macrophages: protective effects of taurine. J. Nutr.
Biochem. 2: 308-313.
6-171
-------
Barr, B. C.; Hyde, D. M.; Plopper, C. G.; Dungworth, D. L. (1988) Distal airway remodeling in rats chronically
exposed to ozone. Am. Rev. Respir. Dis. 137: 924-938.
Barr, B. C.; Hyde, D. M.; Plopper, C. G.; Dungworth, D. L. (1990) A comparison of terminal airway
remodeling in chronic daily versus episodic ozone exposure. Toxicol. Appl. Pharmacol.
106: 384-407.
Barry, B. E.; Miller, F. J.; Crapo, J. D. (1983) Alveolar epithelial injury caused by inhalation of 0.25 ppm of
ozone. In: Lee, S. D.; Mustafa, M. G.; Mehlman, M. A., eds. International symposium on the
biomedical effects of ozone and related photochemical oxidants; March 1982; Pinehurst, NC.
Princeton, NJ: Princeton Scientific Publishers, Inc.; pp. 299-309. (Advances in modern
environmental toxicology: v. 5).
Barry, B. E.; Miller, F. J.; Crapo, J. D. (1985) Effects of inhalation of 0.12 and 0.25 parts per million ozone on
the proximal alveolar region of juvenile and adult rats. Lab. Invest. 53: 692-704.
Barry, B. E.; Mercer, R. R.; Miller, F. J.; Crapo, J. D. (1988) Effects of inhalation of 0.25 ppm ozone on the
terminal bronchioles of juvenile and adult rats. Exp. Lung Res. 14: 225-245.
Bartlett, D., Jr.; Faulkner, C. S., II; Cook, K. (1974) Effect of chronic ozone exposure on lung elasticity in young
rats. J. Appl. Physiol. 37: 92-96.
Bassett, D. J. P.; Bowen-Kelly, E. (1986) Rat lung metabolism after 3 days of continuous exposure to 0.6 ppm
ozone. Am. J. Physiol. 250: E131-E136.
Bassett, D. J. P.; Rabinowitz, J. L. (1985) Incorporation of glucose carbons into rat lung lipids after exposure to
0.6 ppm ozone. Am. J. Physiol. 248: E553-E559.
Bassett, D. J. P.; Bowen-Kelly, E.; Elbon, C. L.; Reichenbaugh, S. S. (1988a) Rat lung recovery from 3 days of
continuous exposure to 0.75 ppm ozone. J. Toxicol. Environ. Health 25: 329-347.
Bassett, D. J. P.; Bowen-Kelly, E.; Brewster, E. L.; Elbon, C. L.; Reichenbaugh, S. S.; Bunton, T.; Kerr, J. S.
(1988b) A reversible model of acute lung injury based on ozone exposure. Lung 166: 355-369.
Bassett, D. J. P.; Bowen-Kelly, E.; Seed, J. L. (1988c) Rat lung benzo(a)pyrene metabolism following three days
continuous exposure to 0.6 ppm ozone. Res. Commun. Chem. Pathol. Pharmacol. 60: 291-307.
Bassett, D. J. P.; Elbon, C. L.; Reichenbaugh, S. S.; Boswell, G. A.; Stevens, T. M.; McGowan, M. C.; Kerr,
J. S. (1989) Pretreatment with EDU decreases rat lung cellular responses to ozone. Toxicol. Appl.
Pharmacol. 100: 32-40.
Beckett, W. S.; Freed, A. N.; Turner, C.; Menkes, H. A. (1988) Prolonged increased responsiveness of canine
peripheral airways after exposure to O3. J. Appl. Physiol. 64: 605-610.
Bhalla, D. K.; Crocker, T. T. (1986) Tracheal permeability in rats exposed to ozone: an electron microscopic and
autoradiographic analysis of the transport pathway. Am. Rev. Respir. Dis. 134: 572-579.
Bhalla, D. K.; Crocker, T. T. (1987) Pulmonary epithelial permeability in rats exposed to O3. J. Toxicol.
Environ. Health 21: 73-87.
Bhalla, D. K.; Young, C. (1992) Effects of acute exposure to O3 on rats: sequence of epithelial and inflammatory
changes in the distal airways. Inhalation Toxicol. 4: 17-31.
6-172
-------
Bhalla, D. K.; Mannix, R. C.; Kleinman, M. T.; Crocker, T. T. (1986) Relative permeability of nasal, tracheal,
and bronchoalveolar mucosa to niacromolecules in rats exposed to ozone. J. Toxicol. Environ. Health
17: 269-283.
Bhalla, D. K.; Mannix, R. C.; Lavan, S. M.; Phalen, R. P.; Kleinman, M. T.; Crocker, T. T. (1987) Tracheal
and bronchoalveolar permeability changes in rats inhaling oxidant atmospheres during rest or
exercise. J. Toxicol. Environ. Health 22: 417-437.
Bhalla, D. K.; Daniels, D. S.; Luu, N. T. (1992) Attenuation of ozone-induced airway permeability in rats by
pretreatnient with cyclophosphamide, FPL 55712, and indomethacin. Am. J. Respir. Cell Mol.
Biol. 7: 73-80.
Bhalla, D. K.; Rasmussen, R. E.; Daniels, D. S. (1993) Adhesion and motility of polymorphonuclear leukocytes
isolated from the blood of rats exposed to ozone: potential bioniarkers of toxicity. Toxicol. Appl.
Pharmacol. 123: 177-186.
Biagini, R. E.; Moorman, W. J.; Lewis, T. R.; Bernstein, I. L. (1986) Ozone enhancement of platinum asthma in
a primate model. Am. Rev. Respir. Dis. 134: 719-725.
Bleavins, M. R.; Dziedzic, D. (1990) An imniunofluorescence study of T and B lymphocytes in ozone-induced
pulmonary lesions in the mouse. Toxicol. Appl. Pharmacol. 105: 93-102.
Boatman, E. S.; Ward, G.; Martin, C. J. (1983) Morphometric changes in rabbit lungs before and after
pneunionectomy and exposure to ozone. J. Appl. Physiol.: Respir. Environ. Exercise Physiol.
54: 778-784.
Boorman, G. A.; Schwartz, L. W.; Dungworth, D. L. (1980) Pulmonary effects of prolonged ozone insult in rats:
morphometric evaluation of the central acinus. Lab. Invest. 43: 108-115.
Boorman, G. A.; Hailey, R.; Grumbein, S.; Chou, B. J.; Herbert, R. A.; Goehl, T.; Mellick, P. W.; Roycroft,
J. H.; Haseman, J. K.; Sills, R. (1994) Toxicology and carcinogenesis studies of ozone and ozone
4-(ALnitrosomethylamino)-l-(3-pyridyl)-l-butanone in Fischer-344/N rats. Toxicol. Pathol.
22: 545-554.
Borek, C.; Zaider, M.; Ong, A.; Mason, H.; Witz, G. (1986) Ozone acts alone and synergistically with ionizing
radiation to induce in vitro neoplastic transformation. Carcinogenesis (London) 7: 1611-1613.
Borek, C.; Ong, A.; Mason, H. (1989a) Ozone and ultraviolet light act as additive cocarcinogens to induce in
vitro neoplastic transformation. Teratog. Carcinog. Mutagen. 9: 71-74.
Borek, C.; Ong, A.; Zaider, M. (1989b) Ozone activates transforming genes in vitro and acts as a synergistic
co-carcinogen with D-rays only if delivered after radiation. Carcinogenesis (London) 10: 1549-1551.
Brummer, M. E. G.; Schwartz, L. W.; McQuillen, N. K. (1977) A quantitative study of lung damage by scanning
electron microscopy. Inflammatory cell response to high-ambient levels of ozone. In: Johari, O.;
Becker, R. P., eds. Scanning electron microscopy/1977/II—biological applications of the SEM:
proceedings of the workshops on advances in biomedical applications of the SEM & STEM; March-
April; Chicago, IL. Chicago, IL: IIT Research Institute; pp. 513-518.
Burleson, G. R.; Keyes, L. L.; Stutznian, J. D. (1989) Iinmunosuppression of pulmonary natural killer activity by
exposure to ozone. Immunopharmacol. Immunotoxicol. 11: 715-735.
Cajigas, A.; Gayer, M.; Beam, C.; Steinberg, J. J. (1994) Ozonation of DMA forms adducts: a32P-DNA labeling
and thin-layer chromatography technique to measure DMA environmental bioniarkers. Arch.
Environ. Health 49: 25-36.
6-173
-------
Canada, A. T.; Calabrese, E. J. (1985) Ozone-induced inhibition of theophylline elimination in rabbits: effect of
age and sex. Toxicol. Appl. Pharmacol. 81: 43-49.
Canada, A. T.; Calabrese, E. J.; Leonard, D. (1986) Age-dependent inhibition of pentobarbital sleeping time by
ozone in mice and rats. J. Gerontol. 41: 587-589.
Canada, A. T.; Chow, C. K.; Airriess, G. R.; Calabrese, E. J. (1987) Lack of ozone effect on plasma
concentrations of retinol, ascorbic acid, and tocopherol. Nutr. Res. 7: 797-800.
Canning, B. J.; Hmieleski, R. R.; Spannhake, E. W.; Jakab, G. J. (1991) Ozone reduces murine alveolar and
peritoneal macrophage phagocytosis: the role of prostanoids. Am. J. Physiol. 261: L277-L282.
Castleman, W. L.; Dungworth, D. L.; Schwartz, L. W.; Tyler, W. S. (1980) Acute respiratory bronchiolitis: an
ultrastructural and autoradiographic study of epithelial cell injury and renewal in rhesus monkeys
exposed to ozone. Am. J. Pathol. 98: 811-840.
Catalano, P. J.; Rogus, J.; Ryan, L. M. (1995a) Consequences of prolonged inhalation of ozone on F344/N rats:
collaborative studies. Part X: Robust composite scores based on median polish analysis. Cambridge,
MA: Health Effects Institute; research report no. 65.
Catalano, P. J.; Chang, L.-Y. L.; Harkema, J. R.; Kaden, D. A.; Last, J. A.; Mellick, P. W.; Parks, W. C.;
Pinkerton, K. E.; Radhakrishnamurthy, B.; Ryan, L. M.; Szarek, J. L. (1995b) Consequences of
prolonged inhalation of ozone on F344/N rats: collaborative studies. Part XI: Integrative summary.
Cambridge, MA: Health Effects Institute; research report no. 65.
Cavender, F. L.; Steinhagen, W. H.; Ulrich, C. E.; Busey, W. M.; Cockrell, B. Y.; Haseman, J. K.; Hogan, M.
D.; Drew, R. T. (1977) Effects in rats and guinea pigs of short-term exposures to sulfuric acid mist,
ozone, and their combination. J. Toxicol. Environ. Health 3: 521-533.
Chancy, S. G. (1981) Effects of ozone on leukocyte DMA. Research Triangle Park, NC: U.S. Environmental
Protection Agency, Health Effects Research Laboratory; report no. EPA-600/1-81-031. Available
from: NTIS, Springfield, VA; PB81-179277.
Chang, L.-Y.; Mercer, R. R.; Stockstill, B. L.; Miller, F. J.; Graham, J. A.; Ospital, J. J.; Crapo, J. D. (1988)
Effects of low levels of NO2 on terminal bronchiolar cells and its relative toxicity compared to 03.
Toxicol. Appl. Pharmacol. 96: 451-464.
Chang, L.; Miller, F. J.; Ultman, J.; Huang, Y.; Stockstill, B. L.; Grose, E.; Graham, J. A.; Ospital, J. J.;
Crapo, J. D. (1991) Alveolar epithelial cell injuries by subchronic exposure to low concentrations of
ozone correlate with cumulative exposure. Toxicol. Appl. Pharmacol. 109: 219-234.
Chang, L.-Y.; Huang, Y.; Stockstill, B. L.; Graham, J. A.; Grose, E. C.; Menache, M. G.; Miller, F. J.; Costa,
D. L.; Crapo, J. D. (1992) Epithelial injury and interstitial fibrosis in the proximal alveolar regions
of rats chronically exposed to a simulated pattern of urban ambient ozone. Toxicol. Appl. Pharmacol.
115: 241-252.
Chang, L.-Y.; Stockstill, B. L.; Menache, M. G.; Mercer, R. R.; Crapo, J. D. (1995) Consequences of
prolonged inhalation of ozone on F344/N rats: collaborative studies. Part VIII. Morphometric
analysis of structural alterations in alveolar regions. Cambridge, MA: Health Effects Institute; pp. 3-
39; research report no. 65.
Chen, L. C.; Miller, P. D.; Lam, H. P.; Guty, J.; Amdur, M. O. (1991) Sulfuric acid-layered ultrafine particles
potentiate ozone-induced airway injury. J. Toxicol. Environ. Health 34: 337-352.
6-174
-------
Chitano, P.; Di Stefano, A.; Finotto, S.; Zavattini, G.; Maestrelli, P.; Mapp, C.; Fabbri, L. M.; Allegra, L.
(1989) Ambroxol inhibits airway hyperresponsiveness induced by ozone in dogs. Respiration
55 (suppl. 1): 74-78.
Choi, A. M. K.; Elbon, C. L.; Bruce, S. A.; Bassett, D. J. P. (1994) Messenger RNA levels of lung extracellular
matrix proteins during ozone exposure. Lung 172: 15-30.
Chow, C. K.; Plopper, C. G.; Chiu, M.; Dungworth, D. L. (1981) Dietary vitamin E and pulmonary biochemical
and morphological alterations of rats exposed to 0.1 ppm ozone. Environ. Res. 24: 315-324.
Christensen, E.; Giese, A. C. (1954) Changes in absorption spectra of nucleic acids and their derivatives
following exposure to ozone and ultraviolet radiations. Arch. Biochem. Biophys. 51: 208-216.
demons, G. K.; Garcia, J. F. (1980a) Endocrine aspects of ozone exposure in rats. Arch. Toxicol. Suppl. 4: 301-
304.
demons, G. K.; Garcia, J. F. (1980b) Changes in thyroid function after short-term ozone exposure in rats.
J. Environ. Pathol. Toxicol. 4: 359-369.
Coffin, D. L.; Gardner, D. E. (1972) Interaction of biological agents and chemical air pollutants. Ann. Occup.
Hyg. 15: 219-234.
Coffin, D. L.; Blommer, E. J.; Gardner, D. E.; Holzman, R. (1967) Effect of air pollution on alteration of
susceptibility to pulmonary infection. In: Proceedings of the 3rd annual conference on atmospheric
contamination in confined spaces; May; Dayton, OH. Wright-Patterson Air Force Base, OH:
Aerospace Medical Research Laboratories; pp. 71-80; report no. AMRL-TR-67-200. Available
from: NTIS, Springfield, VA; AD-835008.
Coleridge, J. C. G.; Coleridge, H. M.; Schelegle, E. S.; Green, J. F. (1993) Acute inhalation of ozone stimulates
bronchial C-fibers and rapidly adapting receptors in dogs. J. Appl. Physiol. 74: 2345-2352.
Costa, D. L.; Kutzman, R. S.; Lehmann, J. R.; Popenoe, E. A.; Drew, R. T. (1983) A subchronic multidose
ozone study in rats. In: Lee, S. D.; Mustafa, M. G.; Mehlman, M. A., eds. International symposium
on the biomedical effects of ozone and related photochemical oxidants; March 1982; Pinehurst, NC.
Princeton, NJ: Princeton Scientific Publishers, Inc.; pp. 369-393. (Advances in modern
environmental toxicology: v. 5).
Costa, D. L.; Hatch, G. E.; Highfill, J.; Stevens, M. A.; Tepper, J. S. (1988a) Pulmonary function studies in the
rat addressing concentration versus time relationships of ozone (03). Research Triangle Park,
NC: U.S. Environmental Protection Agency, Health Effects Research Laboratory; report no.
EPA-600/D-88-256. Available from: NTIS, Springfield, VA; PB89-129050.
Costa, D. L.; Stevens, M. S.; Tepper, J. S. (1988b) Repeated exposure to ozone (O,) and chronic lung disease:
recent animal data. Presented at: 81st annual meeting of the Air Pollution Control Association; June;
Dallas, TX. Pittsburgh, PA: Air Pollution Control Association; paper no. 88-122.3.
6-175
-------
Costa, D. L.; Hatch, G. E.; Highfill, J.; Stevens, M. A.; Tepper, J. S. (1989) Pulmonary function studies in the
rat addressing concentration versus time relationships of ozone. In: Schneider, T.; Lee, S. D.;
Wolters, G. J. R.; Grant, L. D., eds. Atmospheric ozone research and its policy implications:
proceedings of the 3rd US-Dutch international symposium; May 1988; Nijmegen, The Netherlands.
Amsterdam, The Netherlands: Elsevier Science Publishers; pp. 733-743. (Studies in environmental
science 35).
Costa, D. L.; Tepper, J. S.; Stevens, M. A.; Watkinson, W. P.; Doerfler, D. L.; Gelzleichter, T. R.; Last, J. A.
(1995) Restrictive lung disease in rats exposed chronically to an urban profile of ozone. Am. J.
Respir. Crit. Care Med. 151: 1512-1518.
Crapo, J. D.; Barry, B. E.; Chang, L.-Y.; Mercer, R. R. (1984) Alterations in lung structure caused by inhalation
of oxidants. J. Toxicol. Environ. Health 13: 301-321.
Crocker, T. T.; Bhalla, D. K. (1986) Transport of macromolecules and particles at target sites for deposition of
air pollutants. Cambridge, MA: Health Effects Institute; research report no. 3.
Cueto, R.; Squadrito, G. L.; Berniudez, E.; Pryor, W. A. (1992) Identification of heptanal and nonanal in
bronchoalveolar lav age from rats exposed to low levels of ozone. Biochem. Biophys. Res. Commun.
188: 129-134.
Damji, K. S.; Sherwin, R. P. (1989) The effect of ozone and simulated high altitude on murine lung elastin:
quantitation by image analysis. Toxicol. Ind. Health 5: 995-1003.
Diggle, W. M.; Gage, J. C. (1955) The toxicity of ozone in the presence of oxides of nitrogen. Br. J. Ind. Med.
12: 60-64.
Dillon, D.; Combes, R.; McConville, M.; Zeiger, E. (1992) Ozone is niutagenic in Salmonella. Environ. Mol.
Mutagen. 19: 331-337.
Dimitriadis, V. K. (1992) Carbohydrate cytochemistry of bonnet monkey (Macaca radiatd) nasal epithelium.
Response to ambient levels of ozone. Histol. Histopathol. 7: 479-488.
Dodge, D. E.; Rucker, R. B.; Pinkerton, K. E.; Haselton, C. J.; Plopper, C. G. (1994) Dose-dependent tolerance
to ozone. III. Elevation of intracellular Clara cell 10-kDa protein in central acini of rats exposed for
20 months. Toxicol. Appl. Pharmacol. 127: 109-123.
Donaldson, K.; Brown, G. M.; Brown, D. M.; Slight, J.; Maclaren, W. M.; Davis, J. M. G. (1991) Leukocyte-
mediated epithelial injury in ozone-exposed rat lung. Cambridge, MA: Health Effects Institute;
research report no. 44.
Donaldson, K.; Brown, G. M.; Brown, D. M.; Slight, J.; Maclaren, W.; Davis, J. M. G. (1993) Characteristics
of bronchoalveolar leukocytes from the lungs of rats inhaling 0.2-0.8 ppm of ozone. Inhalation
Toxicol. 5: 149-164.
Dormans, J. A. M. A. (1989) Application of the disector method in the light microscopic quantification of
type II pneumocytes in control and ozone-exposed rats. J. Microsc. 155: 207-211.
Dormans, J. A. M. A.; van Bree, L.; Boere, A. J. P.; Marra, M.; Rombout, P. J. A. (1989) Study of the effects
of ozone in emphysematous rats. J. Toxicol. Environ. Health 26: 1-18.
Dormans, J. A. M. A.; Rombout, P. J. A.; Van Loveren, H. (1990) Surface morphology and morphometry of rat
alveolar macrophages after ozone exposure. J. Toxicol. Environ. Health 31: 53-70.
6-176
-------
Downey, G. P.; Gumbay, R. S.; Doherty, D. E.; LaBrecque, J. P.; Henson, J. E.; Henson, P. M.; Worthen, G.
S. (1988) Enhancement of pulmonary inflammation of PGF^: evidence for a vasodilator effect.
J. Appl. Physiol. 64: 728-741.
Downey, G. P.; Worthen, G. S.; Henson, P. M.; Hyde, D. M. (1993) Neutrophil sequestration and migration in
localized pulmonary inflammation: capillary localization and migration across the interalveolar
septum. Am. Rev. Respir. Dis. 147: 168-176.
Driscoll, K. E.; Schlesinger, R. B. (1988) Alveolar macrophage-stimulated neutrophil and nionocyte migration:
effects of in vitro ozone exposure. Toxicol. Appl. Pharmacol. 93: 312-318.
Driscoll, K. E.; Vollmuth, T. A.; Schlesinger, R. B. (1986) Early alveolar clearance of particles in rabbits
undergoing acute and subchronic exposure to ozone. Fundam. Appl. Toxicol. 7: 264-271.
Driscoll, K. E.; Vollmuth, T. A.; Schlesinger, R. B. (1987) Acute and subchronic ozone inhalation in the rabbit:
response of alveolar macrophages. J. Toxicol. Environ. Health 21: 27-43.
Driscoll, K. E.; Leikauf, G. D.; Schlesinger, R. B. (1988) Effects of in vitro and in vivo ozone exposure on
eicosanoid production by rabbit alveolar macrophages. Inhalation Toxicol. Premier Issue: 109-122.
Dubeau, H.; Chung, Y. S. (1979) Ozone response in wild type and radiation-sensitive mutants of Saccharomyces
cerevisiae. MGG Mol. Gen. Genet. 176: 393-398.
Dubeau, H.; Chung, Y. S. (1982) Genetic effects of ozone: induction of point mutation and genetic recombination
in Saccharomyces cerevisiae. Mutat. Res. 102: 249-259.
Dubick, M. A.; Heng, H.; Rucker, R. B. (1985) Effects of protein deficiency and food restriction on lung
ascorbic acid and glutathione in rats exposed to ozone. J. Nutr. 115: 1050-1056.
Dungworth, D. L. (1989) Noncarcinogenic responses of the respiratory tract to inhaled toxicants. In: McClellan,
R. O.; Henderson, R. P., eds. Concepts in inhalation toxicology. New York, NY: Hemisphere
Publishing Corp.; pp. 273-298.
Dziedzic, D.; White, H. J. (1986a) T-cell activation in pulmonary lymph nodes of mice exposed to ozone.
Environ. Res. 41: 610-622.
Dziedzic, D.; White, H. J. (1986b) Thynius and pulmonary lymph node response to acute and subchronic ozone
inhalation in the mouse. Environ. Res. 41: 598-609.
Dziedzic, D.; White, H. J. (1987a) Response of T-cell-deficient mice to ozone exposure. J. Toxicol. Environ.
Health 21: 57-71.
Dziedzic, D.; White, H. J. (1987b) Quantitation of ozone-induced lung lesion density after treatment with an
interferon inducer or an anti-interferon antibody. Toxicol. Lett. 39: 51-62.
Dziedzic, D.; Wright, E. S.; Sargent, N. E. (1990) Pulmonary response to ozone: reaction of bronchus-associated
lymphoid tissue and lymph node lymphocytes in the rat. Environ. Res. 51: 194-208.
Easton, R. E.; Murphy, S. D. (1967) Experimental ozone preexposure and histamine: effect on the acute toxicity
and respiratory function effects of histamine in guinea pigs. Arch. Environ. Health 15: 160-166.
Effros, R. M.; Feng, N.-H.; Mason, G.; Sietsema, K.; Silverman, P.; Hukkanen, J. (1990) Solute concentrations
of the pulmonary epithelial lining fluid of anesthetized rats. J. Appl. Physiol. 68: 275-281.
6-177
-------
Ehrlich, R. (1983) Changes in susceptibility to respiratory infection caused by exposures to photochemical oxidant
pollutants. In: Lee, S. D.; Mustafa, M. G.; Mehlman, M. A., eds. International symposium on the
biomedical effects of ozone and related photochemical oxidants; March 1982; Pinehurst, NC.
Princeton, NJ: Princeton Scientific Publishers, Inc.; pp. 273-285. (Advances in modern
environmental toxicology: v. 5).
Ehrlich, R.; Findlay, J. C.; Fenters, J. D.; Gardner, D. E. (1977) Health effects of short-term inhalation of
nitrogen dioxide and ozone mixtures. Environ. Res. 14: 223-231.
Ehrlich, R.; Findlay, J. C.; Gardner, D. E. (1979) Effects of repeated exposures to peak concentrations of
nitrogen dioxide and ozone on resistance to streptococcal pneumonia. J. Toxicol. Environ.
Health 5: 631-642.
Elsayed, N. M. (1987) Influence of vitamin E on polyamine metabolism in ozone-exposed rat lungs. Arch.
Biochem. Biophys. 255: 392-399.
Elsayed, N. M.; Mustafa, M. G.; Postlethwait, E. M. (1982) Age-dependent pulmonary response of rats to ozone
exposure. J. Toxicol. Environ. Health 9: 835-848.
Elsayed, N. M.; Kass R.; Mustafa, M. G.; Hacker, A. D.; Ospital, J. J.; Chow, C. K.; Cross, C. E. (1988)
Effect of dietary vitamin E level on the biochemical response of rat lung to ozone inhalation. Drag
Nutr. Interact. 5: 373-386.
Elsayed, N. M.; Ellingson, A. S.; Tierney, D. P.; Mustafa, M. G. (1990) Effects of ozone inhalation on
polyamine metabolism and tritiated thymidine incorporation into DMA of rat lungs. Toxicol. Appl.
Pharniacol. 102: 1-8.
Erdrnan, H. E.; Hernandez, T. (1982) Adult toxicity and dominant lethals induced by ozone at specific stages in
spermatogenesis in Drosophila virilis. Environ. Mutagen. 4: 657-666.
Eskew, M. L.; Scheuchenzuber, W. J.; Scholz, R. W.; Reddy, C. C.; Zarkower, A. (1986) The effects of ozone
inhalation on the immunological response of selenium- and vitamin E-deprived rats. Environ. Res.
40: 274-284.
Eustis, S. L.; Schwartz, L. W.; Kosch, P. C.; Dungworth, D. L. (1981) Chronic bronchiolitis innonhuman
primates after prolonged ozone exposure. Am. J. Pathol. 105: 121-137.
Evans, M. J.; Cabral, L. J.; Stephens, R. J.; Freeman, G. (1975) Transformation of alveolar Type 2 cells to Type
1 cells following exposure to NO2. Exp. Mol. Pathol. 22: 142-150.
Evans, M. J.; Johnson, L. V.; Stephens, R. J.; Freeman, G. (1976a) Renewal of the terminal bronchiolar
epithelium in the rat following exposure to NO2 or O3. Lab. Invest. 35: 246-257.
Evans, M. J.; Johnson, L. V.; Stephens, R. J.; Freeman, G. (1976b) Cell renewal in the lungs of rats exposed to
low levels of ozone. Exp. Mol. Pathol. 24: 70-83.
Evans, T. W.; Brokaw, J. J.; Chung, K. P.; Nadel, J. A.; McDonald, D. M. (1988) Ozone-induced bronchial
hyperresponsiveness in the rat is not accompanied by neutrophil influx or increased vascular
permeability in the trachea. Am. Rev. Respir. Dis. 138: 140-144.
Evans, J. N.; Hemenway, D. R.; Kelley, J. (1989) Early markers of lung injury. Cambridge, MA: Health Effects
Institute; research report no. 29. Available from: NTIS, Springfield, VA; PB91-171983.
6-178
-------
Fabbri, L. M.; Aizawa, H.; Alpert, S. E.; Walters, E. H.; O'Byrne, P. M.; Gold, B. D.; Nadel, J. A.;
Holtzman, M. J. (1984) Airway hyperresponsiveness and changes in cell counts in bronchoalveolar
lavage after ozone exposure in dogs. Am. Rev. Respir. Dis. 129: 288-291.
Fabbri, L. M.; Aizawa, H.; O'Byrne, P. M.; Bethel, R. A.; Walters, E. H.; Holtzman, M. J.; Nadel, J. A.
(1985) An anti-inflammatory drag (BW755C) inhibits airway hyperresponsiveness induced by ozone
in dogs. J. Allergy Clin. Immunol. 76: 162-166.
Fetner, R. H. (1962) Ozone-induced chromosome breakage in human cell cultures. Nature (London)
194: 793-794.
Filipowicz, C.; McCauley, R. (1986a) The effect of chronic ozone exposure on lung benzo(D)pyrene oxidase,
benzphetamine demethylase andmonoamine oxidase. Res. Commun. Chem. Pathol. Pharmacol.
51: 289-296.
Filipowicz, C.; McCauley, R. (1986b) The effects of chronic ozone exposure on pulmonary collagen content and
collagen synthesis in rats. J. Appl. Toxicol. 6: 87-90.
Fouke, J. M.; Delemos, R. A.; McFadden, E. R., Jr. (1988) Airway response to ultra short-term exposure to
ozone. Am. Rev. Respir. Dis. 137: 326-330.
Fouke, J. M.; DeLemos, R. A.; Dunn, M. J.; McFadden, E. R., Jr. (1990) Effects of ozone on cyclooxygenase
metabolites in the baboon tracheobronchial tree. J. Appl. Physiol. 69: 245-250.
Fouke, J. M.; Wolin, A. D.; McFadden, E. R., Jr. (1991) Effects of ozone on lung mechanics and
cyclooxygenase metabolites in dogs. Prostaglandins 42: 343-353.
Freeman, G.; Juhos, L. T.; Furiosi, N. J.; Mussenden, R.; Stephens, R. J.; Evans, M. J. (1974) Pathology of
pulmonary disease from exposure to interdependent ambient gases (nitrogen dioxide and ozone).
Arch. Environ. Health 29: 203-210.
Fujimaki, H. (1989) Impairment of humoral immune responses in mice exposed to nitrogen dioxide and ozone
mixtures. Environ. Res. 48: 211-217.
Fujimaki, H.; Shiraishi, P.; Ashikawa, T.; Murakami, M. (1987) Changes in delayed hypersensitivity reaction in
mice exposed to O3. Environ. Res. 43: 186-190.
Fujinaka, L. E.; Hyde, D. M.; Plopper, C. G.; Tyler, W. S.; Dungworth, D. L.; Lollini, L. O. (1985)
Respiratory bronchiolitis following long-term ozone exposure in bonnet monkeys: a morphometric
study. Exp. Lung Res. 8: 167-190.
Fukase, O.; Isomura, K.; Watanabe, H. (1978) Effects of exercise on mice exposed to ozone. Arch. Environ.
Health 33: 198-200.
Gardner, D. E. (1982) Use of experimental airborne infections for monitoring altered host defenses. Environ.
Health Perspect. 43: 99-107.
Gardner, D. E.; Miller, F. J.; Illing, J. W.; Kirtz, J. M. (1977) Increased infectivity with exposure to ozone and
sulfuric acid. Toxicol. Lett. 1: 59-64.
Gelzleichter, T. R.; Witschi, H.; Last, J. A. (1992a) Synergistic interaction of nitrogen dioxide and ozone on rat
lungs: acute responses. Toxicol. Appl. Pharmacol. 116: 1-9.
6-179
-------
Gelzleichter, T. R.; Witschi, H.; Last, J. A. (1992b) Concentration-response relationships of rat lungs to exposure
to oxidant air pollutants: a critical test of Haber's Law for ozone and nitrogen dioxide. Toxicol.
Appl. Pharmacol. 112: 73-80.
Gertner, A.; Bromberger-Barnea, B.; Traystman, R.; Berzon, D.; Menkes, H. (1983a) Responses of the lung
periphery to ozone and histamine. J. Appl. Physiol.: Respir. Environ. Exercise Physiol. 54: 640-646.
Gertner, A.; Bromberger-Barnea, B.; Dannenberg, A. M., Jr.; Traystman, R.; Menkes, H. (1983b) Responses of
the lung periphery to 1.0 ppm ozone. J. Appl. Physiol.: Respir. Environ. Exercise Physiol.
55: 770-776.
Gertner, A.; Bromberger-Barnea, B.; Traystman, R.; Menkes, H. (1983c) Effects of ozone on peripheral lung
reactivity. J. Appl. Physiol.: Respir. Environ. Exercise Physiol. 55: 777-784.
Gichner, T.; Langebartels, C.; Sandermann, H., Jr. (1992) Ozone is not mutagenic in the Tradescantia and
tobacco niutagenicity assays. Mutat. Res. 281: 203-206.
Gilmour, M. I.; Jakab, G. J. (1991) Modulation of immune function in mice exposed to 0.8 ppm ozone. Inhalation
Toxicol. 3: 293-308.
Gilmour, M. I.; Belgrade, M. K. (1993) A comparison of the pulmonary defenses against streptococcal infection
in rats and mice following O3 exposure: differences in disease susceptibility and neutrophil
recruitment. Toxicol. Appl. Pharmacol. 123: 211-218.
Gilmour, M. I.; Hmieleski, R. R.; Stafford, E. A.; Jakab, G. J. (1991) Suppression and recovery of the alveolar
macrophage phagocytic system during continuous exposure to 0.5 ppm ozone. Exp. Lung Res.
17: 547-558.
Gilmour, M. I.; Park, P.; Selgrade, M. K. (1993a) Ozone-enhanced pulmonary infection with Streptococcus
zooepidemicm in mice. Am. Rev. Respir. Dis. 147: 753-760.
Gilmour, M. I.; Park, P.; Doerfler, D.; Selgrade, M. K. (1993b) Factors that influence the suppression of
pulmonary antibacterial defenses in mice exposed to ozone. Exp. Lung Res. 19: 299-314.
Goheen, S. C.; O'Rourke, L.; Larkin, E. C. (1986) Ozone and the peroxidation of polyunsaturated fatty acids in
vivo. Environ. Res. 40: 47-57.
Goldstein, E.; Tyler, W. S.; Hoeprich, P. D.; Eagle, C. (1971a) Ozone and the antibacterial defense mechanisms
of the murine lung. Arch. Intern. Med. 127: 1099-1102.
Goldstein, E.; Tyler, W. S.; Hoeprich, P. D.; Eagle, C. (1971b) Adverse influence of ozone on pulmonary
bactericidal activity of murine lung. Nature (London) 229: 262-263.
Goldstein, E.; Warshauer, D.; Lippert, W.; Tarkington, B. (1974) Ozone and nitrogen dioxide exposure: murine
pulmonary defense mechanisms. Arch. Environ. Health 28: 85-90.
Gooch, P. C.; Creasia, D. A.; Brewen, J. G. (1976) The cytogenetic effects of ozone: inhalation and in vitro
exposures. Environ. Res. 12: 188-195.
Gordon, T.; Amdur, M. O. (1980) Effect of ozone on respiratory response of guinea pigs to histamine. J. Toxicol.
Environ. Health 6: 185-195.
Gordon, T.; Venugopalan, C. S.; Amdur, M. O.; Brazen, J. M. (1984) Ozone-induced airway hyperreactivity in
the guinea pig. J. Appl. Physiol.: Respir. Environ. Exercise Physiol. 57: 1034-1038.
6-180
-------
Graham, J. A.; Menzel, D. B.; Miller, F. J.; Illing, J. W.; Gardner, D. E. (1981) Influence of ozone on
pentobarbital-induced sleeping time in mice, rats, and hamsters. Toxicol. Appl. Pharmacol.
61: 64-73.
Graham, J. A.; Menzel, D. B.; Mole, M. L.; Miller, F. J.; Gardner, D. E. (1985) Influence of ozone on
pentobarbital pharmacokinetics in mice. Toxicol. Lett. 24: 163-170.
Graham, J. A.; Gardner, D. E.; Blommer, E. J.; House, D. E.; Menache, M. G.; Miller, F. J. (1987) Influence
of exposure patterns of nitrogen dioxide and modifications by ozone on susceptibility to bacterial
infectious disease in mice. J. Toxicol. Environ. Health 21: 113-125.
Graham, J. A.; Grant, L. D.; Folinsbee, L. J.; Gardner, D. E.; Schlesinger, R. B.; Overton, J. H.; Lounsbury,
S. W.; McCurdy, T. R.; Hasselblad, V.; McKee, D. J.; Richmond, H. M.; Polkowsky, B. V.;
Marcus, A. H. (1991) Direct health effects of air pollutants associated with acidic precursor
emissions. In: Irving, P. M., ed. Acidic deposition: state of science and technology, volume III,
terrestrial, materials, health and visibility effects. Washington, DC: The U.S. National Acid
Precipitation Assessment Program. (State of science and technology report 22).
Green, G. M. (1973) Alveolobronchiolar transport mechanisms. Arch. Intern. Med. 131: 109-114.
Green, G. M. (1984) Similarities of host defense mechanisms against pulmonary infectious diseases in animals and
man. J. Toxicol. Environ. Health 13: 471-478.
Green, G. M.; Jakab, G. J.; Low, R. B.; Davis, G. S. (1977) Defense mechanisms of the respiratory membrane.
Am. Rev. Respir. Dis. 115: 479-514.
Grindstaff, G.; Henry, M.; Hernandez, O.; Hogan, K.; Lai, D.; Siegel-Scott, C. (1991) Formaldehyde risk
assessment update [final draft]. Washington, DC: U.S. Environmental Protection Agency, Office of
Toxic Substances.
Grose, E. C.; Gardner, D. E.; Miller, F. J. (1980) Response of ciliated epithelium to ozone and sulfuric acid.
Environ. Res. 22: 377-385.
Grose, E. C.; Richards, J. H.; Illing, J. W.; Miller, F. J.; Davies, D. W.; Graham, J. A.; Gardner, D. E. (1982)
Pulmonary host defense responses to inhalation of sulfuric acid and ozone. J. Toxicol. Environ.
Health 10: 351-362.
Grose, E. C.; Stevens, M. A.; Hatch, G. E.; Jaskot, R. H.; Selgrade, M. J. K.; Stead, A. G.; Costa, D. L.;
Graham, J. A. (1989) The impact of a 12-month exposure to a diurnal pattern of ozone on pulmonary
function, antioxidant biochemistry and immunology. In: Schneider, T.; Lee, S. D.; Wolters, G. J.
R.; Grant, L. D., eds. Atmospheric ozone research and its policy implications: proceedings of the
3rd US-Dutch international symposium; May 1988; Nijmegen, The Netherlands. Amsterdam, The
Netherlands: Elsevier Science Publishers; pp. 535-544. (Studies in environmental science 35).
Gross, K. B.; White, H. J. (1986) Pulmonary functional and morphological changes induced by a 4-week exposure
to 0.7 ppm ozone followed by a 9-week recovery period. J. Toxicol. Environ. Health 17: 143-157.
Gross, K. B.; White, H. J. (1987) Functional and pathologic consequences of a 52-week exposure to 0.5 PPM
ozone followed by a clean air recovery period. Lung 165: 283-295.
Guerrero, R. R.; Rounds, D. E.; Olson, R. S.; Hackney, J. D. (1979) Mutagenic effects of ozone on human cells
exposed in vivo and in vitro based on sister chroniatid exchange analysis. Environ. Res. 18: 336-346.
6-181
-------
Gunnison, A. P.; Finkelstein, I.; Weideman, P.; Su, W.-Y.; Sobo, M.; Schlesinger, R. B. (1990) Age-dependent
effect of ozone on pulmonary eicosanoid metabolism in rabbits and rats. Fundam. Appl. Toxicol.
15: 779-790.
Gunnison, A. P.; Wiedeman, P. A.; Sobo, M.; Koenig, K. L.; Lung, C. C. (1992a) Age-dependence of responses
to acute ozone exposure in rats. Fundam. Appl. Toxicol. 18: 360-369.
Gunnison, A. P.; Weideman, P. A.; Sobo, M. (1992b) Enhanced inflammatory response to acute ozone exposure
in rats during pregnancy and lactation. Fundam. Appl. Toxicol. 19: 607-612.
Guth, D. J.; Warren, D. L.; Last, J. A. (1986) Comparative sensitivity of measurements of lung damage made by
bronchoalveolar lavage after short-term exposure of rats to ozone. Toxicology 40: 131-143.
Hacker, A. D.; Mustafa, M. G.; Ospital, J. J.; Elsayed, N. M.; Lee, S. D. (1986) Relationship of age to rat lung
collagen synthesis in response to ozone exposure. Age 9: 1-5.
Haefeli-Bleuer, B.; Weibel, E. R. (1988) Morphometry of the human pulmonary acinus. Anal. Rec. 220: 401-414.
Hamelin, C. (1985) Production of single- and double-strand breaks in plasniid DMA by ozone. Int. J. Radial.
Oncol. Biol. Phys. 11: 253-257.
Hamelin, C.; Chung, Y. S. (1975a) The effect of low concentrations of ozone onEscherichia coll chromosome.
Mutat. Res. 28: 131-132.
Hamelin, C.; Chung, Y. S. (1975b) Characterization of mucoid mutants ofEscherichia coli K-12 isolated after
exposure to ozone. J. Bacteriol. 122: 19-24.
Hamelin, C.; Sarhan, P.; Chung, Y. S. (1977a) Ozone-induced DMA degradation in different DMA
polynierase I mutants of Escherichia coli K12. Biochem. Biophys. Res. Commun. 77: 220-224.
Hamelin, C.; Sarhan, F.; Chung, Y. S. (1977b) DMA degradation caused by ozone in mucoid mutants of
Escherichia coli K12. FEMS Microbiol. Lett. 2: 149-151.
Harkema, J. R.; Mauderly, J. L. (1994) Consequences of prolonged inhalation of ozone on F344/N rats:
collaborative studies. Part V: Effects on pulmonary function. Cambridge, MA: Health Effects
Institute; research report no. 65.
Harkema, J. R.; Plopper, C. G.; Hyde, D. M.; St. George, J. A.; Wilson, D. W.; Dungworth, D. L. (1987)
Response of the macaque nasal epithelium to ambient levels of ozone: a morphologic and
morphometric study of the transitional and respiratory epithelium. Am. J. Pathol. 128: 29-44.
Harkema, J. R.; Hotchkiss, J. A.; Henderson, R. F. (1989) Effects of 0.12 and 0.80 ppm ozone on rat nasal and
nasopharyngeal epithelial mucosubstances: quantitative histochemistry. Toxicol. Pathol. 17: 525-535.
Harkema, J. R.; Plopper, C. G.; Hyde, D. M.; St. George, J. A.; Wilson, D. W.; Dungworth, D. L. (1993)
Response of macaque bronchiolar epithelium to ambient concentrations of ozone. Am. J. Pathol.
143: 857-866.
Harkema, J. R.; Morgan, K. T.; Gross, E. A.; Catalano, P. J.; Griffith, W. C. (1994) Consequences of
prolonged inhalation of ozone on F344/N rats: collaborative studies. Part VII: effects on the nasal
mucociliary apparatus. Cambridge, MA: Health Effects Institute; research report no. 65.
Hassett, C.; Mustafa, M. G.; Coulson, W. P.; Elashoff, R. M. (1985) Murine lung carcinogenesis following
exposure to ambient ozone concentrations. JNCI J. Natl. Cancer Inst. 75: Ill-Ill.
6-182
-------
Hatch, G. E. (1992) Comparative biochemistry of airway lining fluid. In: Parent, R. A., ed. Comparative biology
of the normal lung: v. I, treatise on pulmonary toxicology. Boca Raton, PL: CRC Press; pp. 617-
632.
Hatch, G. E.; Slade, R.; Stead, A. G.; Graham, J. A. (1986) Species comparison of acute inhalation toxicity of
ozone and phosgene. J. Toxicol. Environ. Health 19: 43-53.
Hatch, G. E.; Koren, H.; Aissa, M. (1989) A method for comparison of animal and human alveolar dose and
toxic effect of inhaled ozone. Health Phys. 57(suppl. 1): 37-40.
Hatch, G. E.; Slade, R.; Harris, L. P.; McDonnell, W. P.; Devlin, R. B.; Koren, H. S.; Costa, D. L.;
McKee, J. (1994) Ozone dose and effect in humans and rats: a comparison using oxygen-18 labeling
and bronchoalveolar lavage. Am. J. Respir. Crit. Care Med. 150: 676-683.
Henderson, R. P.; Hotchkiss, J. A.; Chang, I. Y.; Scott, B. R.; Harkema, J. R. (1993) Effect of cumulative
exposure on nasal response to ozone. Toxicol. Appl. Pharmacol. 119: 59-65.
Heng, H.; Rucker, R. B.; Crotty, J.; Dubick, M. A. (1987) The effects of ozone on lung, heart, and liver
superoxide dismutase and glutathione peroxidase activities in the protein-deficient rat. Toxicol. Lett.
38: 225-237.
Highfill, J. W.; Costa, D. L. (1995) Statistical response models for ozone exposure: their generality when applied
to human spirometric and animal permeability functions of the lung. J. Air Waste Manage. Assoc.
45: 95-102.
Highfill, J. W.; Hatch, G. E.; Slade, R.; Crissman, K. M.; Norwood, J.; Devlin, R. B.; Costa, D. L. (1992)
Concentration-time models for the effects of ozone on bronchoalveolar lavage fluid protein from rats
and guinea pigs. Inhalation Toxicol. 4: 1-16.
Hiroshima, K.; Kohno, T.; Ohwada, H.; Hayashi, Y. (1989) Morphological study of the effects of ozone on rat
lung: II. long-term exposure. Exp. Mol. Pathol. 50: 270-280.
Hoigne, J.; Bader, H. (1975) Ozonation of water: role of hydroxyl radicals as oxidizing intermediates. Science
(Washington, DC) 190: 782-784.
Holroyde, M. C.; Morris, A. A. (1988) The effect of ozone on reactivity of upper and lower airways in guinea
pigs. Br. J. Pharmacol. 94: 938-946.
Holtzman, M. J.; Fabbri, L. M.; Skoogh, B.-E.; O'Byrne, P. M.; Walters, E. H.; Aizawa, H.; Nadel, J. A.
(1983a) Time course of airway hyperresponsiveness induced by ozone in dogs. J. Appl. Physiol.:
Respir. Environ. Exercise Physiol. 55: 1232-1236.
Holtzman, M. J.; Fabbri, L. M.; O'Byrne, P. M.; Gold, B. D.; Aizawa, H.; Walters, E. H.; Alpert, S. E.;
Nadel, J. A. (1983b) Importance of airway inflammation for hyperresponsiveness induced by ozone.
Am. Rev. Respir. Dis. 127: 686-690.
Hotchkiss, J. A.; Harkema, J. R. (1992) Endotoxin or cytokines attenuate ozone-induced DMA synthesis in rat
nasal transitional epithelium. Toxicol. Appl. Pharmacol. 114: 182-187.
Hotchkiss, J. A.; Harkema, J. R.; Sun, J. D.; Henderson, R. F. (1989a) Comparison of acute ozone-induced nasal
and pulmonary inflammatory responses in rats. Toxicol. Appl. Pharmacol. 98: 289-302.
Hotchkiss, J. A.; Harkema, J. R.; Kirkpatrick, D. T.; Henderson, R. F. (1989b) Response of rat alveolar
macrophages to ozone: quantitative assessment of population size, morphology, and proliferation
following acute exposure. Exp. Lung Res. 15: 1-16.
6-183
-------
Hotchkiss, J. A.; Harkema, J. R.; Henderson, R. F. (1991) Effect of cumulative ozone exposure on ozone-
induced nasal epithelial hyperplasia and secretory metaplasia in rats. Exp. Lung Res. 15: 589-600.
Hu, P. C.; Miller, F. J.; Daniels, M. J.; Hatch, G. E.; Graham, J. A.; Gardner, D. E.; Belgrade, M. K. (1982)
Protein accumulation in lung lavage fluid following ozone exposure. Environ. Res. 29: 377-388.
Hurst, D. J.; Gardner, D. E.; Coffin, D. L. (1970) Effect of ozone on acid hydrolases of the pulmonary alveolar
macrophage. RES: J. Reticuloendothel. Soc. 8: 288-300.
Hussain, M. Z.; Mustafa, M. G.; Chow, C. K.; Cross, C. E. (1976a) Ozone-induced increase of lung proline
hydroxylase activity and hydroxyproline content. Chest 69(suppl.): 273-275.
Hussain, M. Z.; Cross, C. E.; Mustafa, M. G.; Bhatnagar, R. S. (1976b) Hydroxyproline contents andprolyl
hydroxylase activities in lungs of rats exposed to low levels of ozone. Life Sci. 18: 897-904.
Hyde, D. M.; Hubbard, W. C.; Wong, V.; Wu, R.; Pinkerton, K.; Plopper, C. G. (1992) Ozone-induced acute
tracheobronchial epithelial injury: relationship to granulocyte emigration in the lung. Am. J. Respir.
Cell Mol. Biol. 6: 481-497.
Ibrahim, A. L.; Zee, Y. C.; Osebold, J. W. (1980) The effects of ozone on the respiratory epithelium of mice: II.
ultrastructural alterations. J. Environ. Pathol. Toxicol. 3: 251-258.
Ichikawa, L; Higuchi, Y.; Ujiie, A.; Yokoyama, E. (1988) Suppression of Sidrnan-type conditioned avoidance
response in rats by ozone. Taiki Osen Gakkaishi 23: 165-170.
Ichinose, T.; Sagai, M. (1989) Biochemical effects of combined gases of nitrogen dioxide and ozone.
III. Synergistic effects on lipid peroxidation and antioxidative protective systems in the lungs of rats
and guinea pigs. Toxicology 59: 259-270.
Ichinose, T.; Sagai, M. (1992) Combined exposure to NO2, O3, and H2SO4-aerosol and lung tumor formation in
rats. Toxicology 74: 173-184.
Ichinose, T.; Arakawa, K.; Shimojo, N.; Sagai, M. (1988) Biochemical effects of combined gases of nitrogen
dioxide and ozone: II. species differences in lipid peroxides and antioxidative protective enzymes in
the lungs. Toxicol. Lett. 42: 167-176.
Inoue, H.; Sato, S.; Hirose, T.; Kikuchi, Y.; Ubukata, T.; Nagashima, S.; Sasaki, T.; Takishima, T. (1979)
[A comparative study between functional and pathologic alterations in lungs of rabbits exposed to an
ambient level of ozone]. Nippon Kyobu Shikkan Gekkai Zasshi 17: 288-296.
Jakab, G. J.; Bassett, D. J. P. (1990) Influenza virus infection, ozone exposure, and fibrogenesis. Am. Rev.
Respir. Dis. 141: 1307-1315.
Jakab, G. J.; Hemenway, D. R. (1994) Concomitant exposure to carbon black particulates enhances ozone-induced
lung inflammation and suppression of alveolar macrophage phagocytosis. J. Toxicol. Environ. Health
41: 221-231.
Jakab, G. J.; Hmieleski, R. R. (1988) Reduction of influenza virus pathogenesis by exposure to 0.5 ppm ozone. J.
Toxicol. Environ. Health 23: 455-472.
Jakab, G. J.; Spannhake, E. W.; Canning, B. J.; Kleeberger, S. R.; Gilmour, M. I. (1995) The effects of ozone
on immune function. Environ. Health Perspect. 103(suppl. 2): 77-89.
Janssen, L. J.; O'Byrne, P. M.; Daniel, E. E. (1991) Mechanism underlying ozone-induced in vitro
hyperresponsiveness in canine bronchi. Am. J. Physiol. 261: L55-L62.
6-184
-------
load, J. P.; Brie, J. M.; Pino, M. V.; Hyde, D. M.; McDonald, R. J. (1993) Effects of ozone and neutrophils on
function and morphology of the isolated rat lung. Am. Rev. Respir. Dis. 147: 1578-1584.
Johnson, H. G.; Stout, B. K.; Ruppel, P. L. (1988) Inhibition of the 5-lipoxygenase pathway with piriprost
(U-60,257) protects normal primates from ozone-induced methacholine hyperresponsive small
airways. Prostaglandins 35: 459-466.
Johnson, N. P.; Hotchkiss, J. A.; Harkema, J. R.; Henderson, R. F. (1990) Proliferative responses of rat nasal
epithelia to ozone. Toxicol. Appl. Pharmacol. 103: 143-155.
Jones, G. L.; Lane, C. G.; Manning, P. J.; O'Byrne, P. M. (1987) Role of the parasympathetic nervous system in
airway hyperresponsiveness after ozone inhalation. J. Appl. Physiol. 63: 1174-1179.
Jones, G. L.; O'Byrne, P. M.; Pashley, M.; Serio, R.; Jury, J.; Lane, C. G.; Daniel, E. E. (1988a) Airway
smooth muscle responsiveness from dogs with airway hyperresponsiveness after 03 inhalation.
J. Appl. Physiol. 65: 57-64.
Jones, G. L.; Lane, C. G.; Daniel, E. E.; O'Byrne, P. M. (1988b) Release of epithelium-derived relaxing factor
after ozone inhalation in dogs. J. Appl. Physiol. 65: 1238-1243.
Jones, G. L.; Lane, C. G.; O'Byrne, P. M. (1990) Effect of thromboxane antagonists on ozone-induced airway
responses in dogs. J. Appl. Physiol. 69: 880-884.
Jones, G. L.; Lane, C.; O'Byrne, P. M. (1992) Effect of an inhaled thromboxane mimetic (U46619) on in vivo
pulmonary resistance and airway hyperresponsiveness in dogs. J. Physiol. 453: 59-67.
Juhos, L. T.; Evans, M. J.; Mussenden-Harvey, R.; Furiosi, N. J.; Lapple, C. E.; Freeman, G. (1978) Limited
exposure of rats to H2SO4 with and without O3. J. Environ. Sci. Health Part C: Environ. Health Sci.
13: 33-47.
Kagawa, J.; Haga, M.; Miyazaki, M. (1989) Effects of repeated exposure to 0.15 ppm Q for four months on
bronchial reactivity in guinea pigs (4 hrs/day; 5 days/wk). In: Schneider, T.; Lee, S. D.; Wolters, G.
J. R.; Grant, L. D., eds. Atmospheric ozone research and its policy implications: proceedings of the
3rd US-Dutch international symposium; May 1988; Nijmegen, The Netherlands. Amsterdam, The
Netherlands: Elsevier Science Publishers; pp. 545-552. (Studies in environmental science 35).
Katsumata, U.; Miura, M.; Ichinose, M.; Kimura, K.; Takahashi, T.; Inoue, H.; Takishima, T. (1990) Oxygen
radicals produce airway constriction and hyperresponsiveness in anesthetized cats. Am. Rev. Respir.
Dis. 141: 1158-1161.
Kavlock, R.; Daston, G.; Grabowski, C. T. (1979) Studies on the developmental toxicity of ozone. I. Prenatal
effects. Toxicol. Appl. Pharmacol. 48: 19-28.
Kavlock, R. J.; Meyer, E.; Grabowski, C. T. (1980) Studies on the developmental toxicity of ozone: postnatal
effects. Toxicol. Lett. 5: 3-9.
Kelley, J. (1990) Cytokines of the lung. Am. Rev. Respir. Dis. 141: 765-788.
Kennedy, C. H.; Hatch, G. E.; Slade, R.; Mason, R. P. (1992) Application of the EPR spin-trapping technique to
the detection of radicals produced in vivo during inhalation exposure of rats to ozone. Toxicol. Appl.
Pharmacol. 114: 41-46.
Kenoyer, J. L.; Phalen, R. P.; Davis, J. R. (1981) Particle clearance from the respiratory tract as a test of
toxicity: effect of ozone on short and long term clearance. Exp. Lung Res. 2: 111-120.
6-185
-------
Kim, C. K.; Gentile, D. M.; Sproul, O. J. (1980) Mechanism of ozone inactivation of bacteriophage f2. Appl.
Environ. Microbiol. 39: 210-218.
King, R. J.; Clements, J. A. (1985) Lipid synthesis and surfactant turnover in the lungs. In: Fishman, A. P.;
Fisher, A. B., eds. Handbook of physiology: v. I, sect. 3, the respiratory system. Bethesda, MD:
American Physiological Society; pp. 309.
Kleeberger, S. R.; Kolbe, J.; Adkinson, N. P., Jr.; Peters, S. P.; Spannhake, E. W. (1988) The role of mediators
in the response of the canine peripheral lung to 1 ppm ozone. Am. Rev. Respir. Dis. 137: 321-325.
Kleeberger, S. R.; Kolbe, J.; Turner, C.; Spannhake, E. W. (1989) Exposure to 1 ppm ozone attenuates the
immediate antigenic response of canine peripheral airways. J. Toxicol. Environ. Health 28: 349-362.
Kleeberger, S. R.; Bassett, D. J. P.; Jakab, G. J.; Levitt, R. C. (1990) A genetic model for evaluation of
susceptibility to ozone-induced inflammation. Am. J. Physiol. 258: L313-L320.
Kleeberger, S. R.; Levitt, R. C.; Zhang, L.-Y. (1993a) Susceptibility to ozone-induced inflammation. I. Genetic
control of the response to subacute exposure. Am. J. Physiol. 264: L15-L20.
Kleeberger, S. R.; Levitt, R. C.; Zhang, L.-Y. (1993b) Susceptibility to ozone-induced inflammation. II. Separate
loci control responses to acute and subacute exposures. Am. J. Physiol. 264: L21-L26.
Kleinnian, M. T.; McClure, T. R.; Mautz, W. J.; Phalen, R. F. (1985) The interaction of ozone and atmospheric
acids on the formation of lung lesions in rats. In: Lee, S. D., ed. Evaluation of the scientific basis for
ozone/oxidants standards: proceedings of an APCA international specialty conference; November
1984; Houston, TX. Pittsburgh, PA: Air Pollution Control Association; pp. 357-365. (APCA
publication no. TR-4).
Kleinman, M. T.; Phalen, R. P.; Mautz, W. J.; Mannix, R. C.; McClure, T. R.; Crocker, T. T. (1989) Health
effects of acid aerosols formed by atmospheric mixtures. Environ. Health Perspect. 79: 137-145.
Kobayashi, T.; Todoroki, T.; Sato, H. (1987) Enhancement of pulmonary metastasis of murine fibrosarcoma NR-
FS by ozone exposure. J. Toxicol. Environ. Health 20: 135-145.
Kozumbo, W. J.; Agarwal, S. (1990) Induction of DMA damage in cultured human lung cells by tobacco smoke
arylamines exposed to ambient levels of ozone. Am. J. Respir. Cell Mol. Biol. 3: 611-618.
L'Herault, P.; Chung, Y. S. (1984) Mutagenicity of ozone in different repair-deficient strains ofEscherichia coll.
MGG Mol. Gen. Genet. 197: 472-477.
Lai, C. C.; Finlayson-Pitts, B. J.; Willis, W. V. (1990) Formation of secondary ozonides from the reaction of an
unsaturated phosphatidylcholine with ozone. Chem. Res. Toxicol. 3: 517-523.
Landolph, J. R. (1985) Chemical transformation in C3H 10T" Cl 8 mouse embryo fibroblasts: historical
background, assessment of the transformation assay, and evolution and optimization of the
transformation assay protocol. In: Kakunaga, T.; Yamasaki, H., eds. Transformation assay of
established cell lines: mechanisms and application. Lyon, France: International Agency for Research
on Cancer; pp. 185-198. (IARC scientific publications, no. 67).
Landolph, J. R. (1989) Molecular and cellular mechanisms of transformation of C3H/10T" Cl 8 and diploid
human fibroblasts by unique carcinogenic, nonmutagenic metal compounds: a review. Biol. Trace
Elem. Res. 21: 459-467.
Landolph, J. R. (1990) Neoplastic transformation of mammalian cells by carcinogenic metal compounds: cellular
and molecular mechanisms. In: Foulkes, E. C., ed. Biological effects of heavy metals: v. II; metal
carcinogenesis. Boca Raton, PL: CRC Press; pp. 1-18.
6-186
-------
Landolph, J. R. (1994) Molecular mechanisms of transformation of C3H/10T" 8 mouse embryo cells and diploid
human fibroblasts by carcinogenic metal compounds. Environ. Health Perspect. 102(suppl. 3): 119-
125.
Last, J. A. (1989) Effects of inhaled acids on lung biochemistry. Environ. Health Perspect. 79: 115-119.
Last, J. A. (1991a) Synergistic effects of air pollutants: ozone plus a respirable aerosol. Cambridge, MA: Health
Effects Institute; research report no. 38. Available from: NTIS, Springfield, VA; PB91-172056.
Last, J. A. (1991b) Global atmospheric change: potential health effects of acid aerosol and oxidant gas mixtures.
Environ. Health Perspect. 96: 151-157.
Last, J. A.; Cross, C. E. (1978) A new model for health effects of air pollutants: evidence for synergistic effects
of mixtures of ozone and sulfuric acid aerosols on rat lungs. J. Lab. Clin. Med. 91: 328-339.
Last, J. A.; Greenberg, D. B. (1980) Ozone-induced alterations in collagen metabolism of rat lungs: II. long-term
exposures. Toxicol. Appl. Pharmacol. 55: 108-114.
Last, J. A.; Greenberg, D. B.; Castlenian, W. L. (1979) Ozone-induced alterations in collagen metabolism of rat
lungs. Toxicol. Appl. Pharmacol. 51: 247-258.
Last, J. A.; Gerriets, J. E.; Hyde, D. M. (1983) Synergistic effects on rat lungs of mixtures of oxidant air
pollutants (ozone or nitrogen dioxide) and respirable aerosols. Am. Rev. Respir. Dis. 128: 539-544.
Last, J. A.; Hyde, D. M.; Chang, D. P. Y. (1984a) A mechanism of synergistic lung damage by ozone and a
respirable aerosol. Exp. Lung Res. 7: 223-235.
Last, J. A.; Reiser, K. M.; Tyler, W. S.; Rucker, R. B. (1984b) Long-term consequences of exposure to ozone:
I. lung collagen content. Toxicol. Appl. Pharmacol. 72: 111-118.
Last, J. A.; Hyde, D. M.; Guth, D. J.; Warren, D. L. (1986) Synergistic interaction of ozone and respirable
aerosols on rat lungs. I. Importance of aerosol acidity. Toxicology 39: 247-257.
Last, J. A.; Warren, D. L.; Pecquet-Goad, E.; Witschi, H. (1987) Modification by ozone of lung tumor
development in mice. JNCI J. Natl. Cancer Inst. 78: 149-154.
Last, J. A.; Gelzleichter, T.; Harkema, J.; Parks, W. C.; Mellick, P. (1993a) Effects of 20 months of ozone
exposure on lung collagen in Fischer 344 rats. Toxicology 84: 83-102.
Last, J. A.; Gelzleichter, T. R.; Pinkerton, K. E.; Walker, R. M.; Witschi, H. (1993b) A new model of
progressive pulmonary fibrosis in rats. Am. Rev. Respir. Dis. 148: 487-494.
Last, J. A.; Gelzleichter, T. R.; Harkema, J.; Hawk, S. (1994) Consequences of prolonged inhalation of ozone on
Fischer-344/N rats: collaborative studies. Part I: Content and cross-linking of lung collagen.
Cambridge, MA: Health Effects Institute; research report no. 65.
Lee, H. K.; Murlas, C. (1985) Ozone-induced bronchial hyperreactivity in guinea pigs is abolished by BW 755C
or FPL 55712 but not by indomethacin. Am. Rev. Respir. Dis. 132: 1005-1009.
Lee, L.-Y.; Bleecker, E. R.; Nadel, J. A. (1977) Effect of ozone on bronchomotor response to inhaled histamine
aerosol in dogs. J. Appl. Physiol.: Respir. Environ. Exercise Physiol. 43: 626-631.
Lee, L.-Y.; Dumont, C.; Djokic, T. D.; Menzel, T. E.; Nadel, J. A. (1979) Mechanism of rapid, shallow
breathing after ozone exposure in conscious dogs. J. Appl. Physiol.: Respir. Environ. Exercise
Physiol. 46: 1108-1114.
6-187
-------
Lee, L.-Y.; Djokic, T. D.; Dumont, C.; Graf, P. D.; Nadel, J. A. (1980) Mechanism of ozone-induced
tachypneic response to hypoxia and hypercapnia in conscious dogs. J. Appl. Physiol.: Respir.
Environ. Exercise Physiol. 48: 163-168.
Lee, J.-S.; Mustafa, M. G.; Afifi, A. A. (1990) Effects of short-term, single and combined exposure to low-level
NO2 and O3 on lung tissue enzyme activities in rats. J. Toxicol. Environ. Health 29: 293-305.
Leikauf, G. D.; Driscoll, K. E.; Wey, H. E. (1988) Ozone-induced augmentation of eicosanoid metabolism in
epithelial cells from bovine trachea. Am. Rev. Respir. Dis. 137: 435-442.
Leikauf, G. D.; Zhao, Q.; Zhou, S.; Santrock, J. (1993) Ozonolysis products of membrane fatty acids activate
eicosanoid metabolism in human airway epithelial cells. Am. J. Respir. Cell Mol. Biol. 9: 594-602.
Leonard, R. J.; Charpied, G. L.; Faddis, B. (1991) Effects of ambient inhaled ozone on vocal fold mucosa in
Bonnet monkeys. J. Voice 5: 304-309.
Lew, D. B.; Chodimella, V.; Murlas, C. G. (1990) Guinea pig ozone-induced airway hyperreactivity is associated
with increased N-acetyl-D-D-glucosaminidase activity in bronchoalveolar lavage fluid. Lung 168:
273-283.
Li, A. F.-Y.; Richters, A. (199la) Effects of 0.7 ppm ozone exposure on thymocytes: in vivo and in vitro studies.
Inhalation Toxicol. 3: 61-71.
Li, A. F.-Y.; Richters, A. (1991b) Ambient level ozone effects on subpopulations of thymocytes and
spleen T lymphocytes. Arch. Environ. Health 46: 57-63.
Li, Z.; Daniel, E. E.; Lane, C. G.; Arnaout, M. A.; O'Byrne, P. M. (1992) Effect of an anti-mo 1 MAb on
ozone-induced airway inflammation and airway hyperresponsiveness in dogs. Am. J. Physiol.
263: L723-L726.
Lippmann, M. (1989) Health effects of ozone: a critical review. JAPCA 39: 672-695.
Lippmann, M. (1993) Health effects of tropospheric ozone: review of recent research findings and their
implications to ambient air quality standards. J. Exposure Anal. Environ. Epidemiol. 3: 103-129.
Lunan, K. D.; Short, P.; Negi, D.; Stephens, R. J. (1977) Glucose-6-phosphate dehydrogenase response of
postnatal lungs to NO2 and O3. In: Sanders, C. L.; Schneider, R. P.; Dagle, G. E.; Ragan, H. A.,
eds. Pulmonary macrophage and epithelial cells: proceedings of the 16th annual Hanford biology
symposium; September 1976; Richland, WA. Washington, DC: Energy Research and Development
Administration; pp. 236-247. Available from: NTIS, Springfield, VA; CONF-760927. (ERDA
symposium series: 43).
Madden, M. C.; Eling, T. E.; Dailey, L. A.; Friedman, M. (1991) The effect of ozone exposure on rat alveolar
macrophage arachidonic acid metabolism. Exp. Lung Res. 17: 47-63.
Madden, M. C.; Friedman, M.; Hanley, N.; Siegler, E.; Quay, J.; Becker, S.; Devlin, R.; Koren, H. S. (1993)
Chemical nature and immunotoxicological properties of arachidonic acid degradation products formed
by exposure to ozone. Environ. Health Perspect. 101: 154-164.
Magie, A. R.; Abbey, D. E.; Centerwall, W. R. (1982) Effect of photochemical smog on the peripheral
lymphocytes of nonsmoking college students. Environ. Res. 29: 204-219.
Mariassy, A. T.; Sielczak, M. W.; McCray, M. N.; Abraham, W. M.; Wanner, A. (1989) Effects of ozone on
lamb tracheal mucosa: quantitative glycoconjugate histochemistry. Am. J. Pathol. 135: 871-879.
6-188
-------
Mariassy, A. T.; Abraham, W. M.; Phipps, R. J.; Sielczak, M. W.; Wanner, A. (1990) Effect of ozone on the
postnatal development of lamb mucociliary apparatus. J. Appl. Physiol. 68: 2504-2510.
Maronpot, R. R. (1991) Correlation of data from the strain A mouse bioassay with long-term bioassays. Exp.
Lung Res. 17: 425-431.
Maronpot, R. R.; Shimkin, M. B.; Witschi, H. P.; Smith, L. H.; Cline, J. M. (1986) Strain A mouse pulmonary
tumor test results for chemicals previously tested in the National Cancer Institute carcinogenicity
tests. JNCI J. Natl. Cancer Inst. 76: 1101-1112.
Matsui, S.; Jones, G. L.; Woolley, M. J.; Lane, C. G.; Gontovnick, L. S.; O'Byrne, P. M. (1991) The effect of
antioxidants on ozone-induced airway hyperresponsiveness in dogs. Am. Rev. Respir. Dis.
144: 1287-1290.
Matsumura, Y. (1970) The effects of ozone, nitrogen dioxide, and sulfur dioxide on the experimentally induced
allergic respiratory disorder in guinea pigs: I. the effect on sensitization with albumin through the
airway. Am. Rev. Respir. Dis. 102: 430-437.
Mautz, W. J.; Bufalino, C. (1989) Breathing pattern and metabolic rate responses of rats exposed to ozone.
Respir. Physiol. 76: 69-77.
Mautz, W. J.; Bufalino, C.; Kleinman, M. T.; Lejnieks, R. M.; Phalen, R. F. (1985a) Pulmonary function of
exercising dogs exposed to ozone alone or in combination with SC^ and acid aerosol. Presented at:
78th annual meeting of the Air Pollution Control Association; June; Detroit, MI. Pittsburgh, PA: Air
Pollution Control Association; paper no. 85-29.4.
Mautz, W. J.; McClure, T. R.; Reischl, P.; Phalen, R. P.; Crocker, T. T. (1985b) Enhancement of ozone-
induced lung injury by exercise. J. Toxicol. Environ. Health 16: 841-854.
Mautz, W. J.; Kleinman, M. T.; Phalen, R. P.; Crocker, T. T. (1988) Effects of exercise exposure on toxic
interactions between inhaled oxidant and aldehyde air pollutants. J. Toxicol. Environ. Health
25: 165-177.
Mautz, W. J.; Finlayson-Pitts, B. J.; Messer, K.; Kleinman, M. T.; Norgren, M. B.; Quirion, J. (1991) Effects
of ozone combined with components of acid fogs on breathing pattern, metabolic rate, pulmonary
surfactant composition, and lung injury in rats. Inhalation Toxicol. 3: 1-25.
McBride, R. K.; Oberdoerster, G.; Marin, M. G. (1991) Effects of ozone on the cholinergic secretory
responsiveness of ferret tracheal glands. Environ. Res. 55: 79-90.
McKenzie, W. H. (1982) Controlled human exposure studies: cytogenetic effects of ozone inhalation. In: Bridges,
B. A.; Butterworth, B. E.; Weinstein, I. B., eds. Indicators of genotoxic exposure. Cold Spring
Harbor, NY: Cold Spring Harbor Laboratory; pp. 319-324. (Banbury report no. 13).
McKenzie, W. H.; Knelson, J. H.; Rummo, N. J.; House, D. E. (1977) Cytogenetic effects of inhaled ozone in
man. Mutat. Res. 48: 95-102.
Meadows, J.; Smith, R. C. (1987) Uric acid protects erythrocytes from ozone-induced changes. Environ. Res. 43:
410-416.
Mellick, P. W.; Dungworth, D. L.; Schwartz, L. W.; Tyler, W. S. (1977) Short term morphologic effects of high
ambient levels of ozone on lungs of rhesus monkeys. Lab. Invest. 36: 82-90.
Menzel, D. B. (1970) Toxicity of ozone, oxygen, and radiation. Annu. Rev. Pharmacol. 10: 379-394.
6-189
-------
Menzel, D. B. (1984) Ozone: an overview of its toxicity in man and animals. J. Toxicol. Environ. Health 13: 183-
204.
Menzel, D. B. (1992) Antioxidant vitamins and prevention of lung disease. Ann. N. Y. Acad. Sci. 669: 141-155.
Mercer, R. R.; Crapo, J. D. (1989) Anatomical modeling of microdosimetry of inhaled particles and gases in the
lung. In: Crapo, J. D.; Smolko, E. D.; Miller, F. J.; Graham, J. A.; Hayes, A. W., eds.
Extrapolation of dosimetric relationships for inhaled particles and gases. San Diego, CA: Academic
Press, Inc.; pp. 69-78.
Mercer, R. R.; Anjilvel, S.; Miller, F. J.; Crapo, J. D. (1991) Inhomogeneity of ventilatory unit volume and its
effects on reactive gas uptake. J. Appl. Physiol. 70: 2193-2205.
Merz, T.; Bender, M. A.; Kerr, H. D.; Kulle, T. J. (1975) Observations of aberrations in chromosomes of
lymphocytes from human subjects exposed to ozone at a concentration of 0.5 ppm for 6 and 10 hours.
Mutat. Res. 31: 299-302.
Miller, F. J.; Illing, J. W.; Gardner, D. E. (1978) Effect of urban ozone levels on laboratory-induced respiratory
infections. Toxicol. Lett. 2: 163-169.
Miller, F. J.; McNeal, C. A.; Kirtz, J. M.; Gardner, D. E.; Coffin, D. L.; Menzel, D. B. (1979)
Nasopharyngeal removal of ozone in rabbits and guinea pigs. Toxicology 14: 273-281.
Miller, P. D.; Gordon, T.; Warnick, M.; Amdur, M. O. (1986) Effect of ozone and histamine on airway
permeability to horseradish peroxidase in guinea pigs. J. Toxicol. Environ. Health 18: 121-132.
Miller, P. D.; Ainsworth, D.; Lam, H. P.; Amdur, M. O. (1987) Effect of ozone exposure on lung functions and
plasma prostaglandin and thromboxane concentrations in guinea pigs. Toxicol. Appl. Pharmacol.
88: 132-140.
Miller, P. D.; Ainsworth, D.; Lam, H. P.; Amdur, M. O. (1988) Indomethacin and cromolyn sodium alter ozone-
induced changes in lung function and plasma eicosanoid concentrations in guinea pigs. Toxicol. Appl.
Pharmacol. 93: 175-186.
Mochitate, K.; Miura, T. (1989) Metabolic enhancement and increase of alveolar macrophages induced by ozone.
Environ. Res. 49: 79-92.
Mochitate, K.; Ishida, K.; Ohsumi, T.; Miura, T. (1992) Long-term effects of ozone and nitrogen dioxide on the
metabolism and population of alveolar macrophages. J. Toxicol. Environ. Health 35: 247-260.
Moffatt, R. K.; Hyde, D. M.; Plopper, C. G.; Tyler, W. S.; Putney, L. F. (1987) Ozone-induced adaptive and
reactive cellular changes in respiratory bronchioles of Bonnet monkeys. Exp. Lung Res. 12: 57-74.
Moore, P. P.; Schwartz, L. W. (1981) Morphological effects of prolonged exposure to ozone and sulfuric acid
aerosol on the rat lung. Exp. Mol. Pathol. 35: 108-123.
Morgan, M. S.; Meyer, P.; Holub, R.; Frank, R. (1986) Overall and regional lung function in dogs exposed
acutely to ozone. Environ. Res. 41: 546-557.
Morgan, D. L.; Furlow, T. L.; Menzel, D. B. (1988) Ozone-initiated changes in erythrocyte membrane and loss
of deformability. Environ. Res. 45: 108-117.
Mura, C.; Chung, Y. S. (1990) In vitro transcription assay of ozonated T7 phage DMA. Environ. Mol. Mutagen.
16: 44-47.
6-190
-------
Murlas, C.; Lee, H. K. (1985) U-60,257 inhibits O3-induced bronchial hyperreactivity in the guinea-pig.
Prostaglandins 30: 563-572.
Murlas, C. G.; Roum, J. H. (1985a) Sequence of pathologic changes in the airway mucosa of guinea pigs during
ozone-induced bronchial hyperreactivity. Am. Rev. Respir. Dis. 131: 314-320.
Murlas, C.; Roum, J. H. (1985b) Bronchial hyperreactivity occurs in steroid-treated guinea pigs depleted of
leukocytes by cyclophosphamide. J. Appl. Physiol. 58: 1630-1637.
Murlas, C. G.; Murphy, T. P.; Chodimella, V. (1990) O3-induced mucosa-linked airway muscle
hyperresponsiveness in the guinea pig. J. Appl. Physiol. 69: 7-13.
Murlas, C. G.; Lang, Z.; Williams, G. J.; Chodimella, V. (1992) Aerosolized neutral endopeptidase reverses
ozone-induced airway hyperreactivity to substance P. J. Appl. Physiol. 72: 1133-1141.
Murphy, S. D.; Ulrich, C. E.; Frankowitz, S. H.; Xintaras, C. (1964) Altered function in animals inhaling low
concentrations of ozone and nitrogen dioxide. Am. Ind. Hyg. Assoc. J. 25: 246-253.
Musi, B.; Dell'Omo, G.; Ricceri, L.; Santucci, D.; Laviola, G.; Bignami, G.; Alleva, E. (1994) Effects of acute
and continuous ozone (O3) exposure on activity/exploration and social behavior of CD-I mice.
Neurotoxicology 15: 827-835.
Mustafa, M. G. (1990) Biochemical basis of ozone toxicity. Free Radic. Biol. Med. 9: 245-265.
Mustafa, M. G.; Elsayed, N. M.; VonDohlen, F. M.; Hassett, C. M.; Postlethwait, E. M.; Quinn, C. L.;
Graham, J. A.; Gardner, D. E. (1984) A comparison of biochemical effects of nitrogen dioxide,
ozone, and their combination in mouse lung. I. Intermittent exposures. Toxicol. Appl. Pharniacol.
72: 82-90.
Mustafa, M. G.; Elsayed, N. M.; Lee, S. D. (1985) The combined effects of O, and NQ on pulmonary
biochemistry. In: Lee, S. D., ed. Evaluation of the scientific basis for ozone/oxidants standards:
transactions of an APCA specialty conference; November 1984; Houston, TX. Pittsburgh, PA: Air
Pollution Control Association; pp. 337-347.
Nachtman, J. P.; Hajratwala, B. R.; Moon, H. L.; Gross, K. B.; Wright, E. S. (1986) Surface-tension
measurements of pulmonary lavage from ozone-exposed rats. J. Toxicol. Environ. Health
19: 127-136.
Nachtman, J. P.; Moon, H. L.; Miles, R. C. (1988) Ozone inhalation in rats: effects on alkaline phosphatase and
lactic dehydrogenase isoenzymes in lavage and plasma. Bull. Environ. Contam. Toxicol. 41: 502-
507.
Nadziejko, C. E.; Nansen, L.; Mannix, R. C.; Kleinman, M. T.; Phalen, R. F. (1992) Effect of nitric acid vapor
on the response to inhaled ozone. Inhalation Toxicol. 4: 343-358.
National Toxicology Program. (1994) NTP technical report on the toxicology and carcinogenesis studies of ozone
(CAS no. 10028-15-6) and ozone/NNK (CAS no. 10028-15-6/64091-91-4) in F344/N rats and
B6C3F, mice (inhalation studies). Research Triangle Park, NC: U.S. Department of Health and
Human Services, National Institutes of Health; publication no. 95-3371. (National Toxicology
Program technical report series no. 440).
Nettesheini, P.; Szakal, A. K. (1972) Morphogenesis of alveolar bronchiolization. Lab. Invest. 26: 210-219.
Nikula, K. J.; Wilson, D. W. (1990) Response of rat tracheal epithelium to ozone and oxygen exposure in vitro.
Fundam. Appl. Toxicol. 15: 121-131.
6-191
-------
Nikula, K. J.; Wilson, D. W.; Giri, S. N.; Plopper, C. G.; Dungworth, D. L. (1988a) The response of the rat
tracheal epithelium to ozone exposure: injury, adaptation, and repair. Am. J. Pathol. 131: 373-384.
Nikula, K. J.; Wilson, D. W.; Dungworth, D. L.; Plopper, C. G. (1988b) In vitro evidence of cellular adaptation
to ozone toxicity in the rat trachea. Toxicol. Appl. Pharmacol. 93: 394-402.
Nishikawa, M.; Suzuki, S.; Ikeda, H.; Fukuda, T.; Suzuki, J.; Okubo, T. (1990) Dose-response relationship of
ozone-induced airway hyperresponsiveness in unanesthetized guinea pigs. J. Toxicol. Environ. Health
30: 123-134.
Nishikawa, M.; Ikeda, H.; Nishiyama, H.; Yamakawa, H.; Suzuki, S.; Okubo, T. (1992) Combined effects of
ozone and cigarette smoke on airway responsiveness and vascular permeability in guinea pigs. Lung
170: 311-322.
O'Byrne, P. M.; Walters, E. H.; Aizawa, H.; Fabbri, L. M.; Holtzman, M. J.; Nadel, J. A. (1984) Indomethacin
inhibits the airway hyperresponsiveness but not the neutrophil influx induced by ozone in dogs. Am.
Rev. Respir. Dis. 130: 220-224.
Oosting, R. S.; Van Golde, L. M. G.; Verhoef, J.; Van Bree, L. (1991a) Species differences in impairment and
recovery of alveolar macrophage functions following single and repeated ozone exposures. Toxicol.
Appl. Pharmacol. 110: 170-178.
Oosting, R. S.; Van Rees-Verhoef, M.; Verhoef, J.; Van Golde, L. M. G.; Van Bree, L. (1991b) Effects of
ozone on cellular ATP levels in rat and mouse alveolar niacrophages. Toxicology 70: 195-202.
Oosting, R. S.; Van Greevenbroek, M. M.; Verhoef, J.; Van Golde, L. M.; Haagsnian, H. P. (1991c) Structural
and functional changes of surfactant protein A induced by ozone. Am. J. Physiol. 261: L77-L83.
Oosting, R. S.; Van Iwaarden, J. P.; Van Bree, L.; Verhoef, J.; Van Golde, L. M. G.; Haagsman, H. P. (1992)
Exposure of surfactant protein A to ozone in vitro and in vivo impairs its interactions with alveolar
cells. Am. J. Physiol. 262: L63-L68.
Oropeza-Rendon, R. L.; Speth, V.; Hiller, G.; Weber, K.; Fischer, H. (1979) Prostaglandin E, reversibly induces
morphological changes in niacrophages and inhibits phagocytosis. Exp. Cell Res. 119: 365-371.
Osebold, J. W.; Gershwin, L. J.; Zee, Y. C. (1980) Studies on the enhancement of allergic lung sensitization by
inhalation of ozone and sulfuric acid aerosol. J. Environ. Pathol. Toxicol. 3: 221-234.
Ozawa, M. (1986) Immunological studies on the effect of air pollutants. The effects of exposure to ozone on IgE
antibody production in mice. Oto-Rhino-Laryngol. Tokyo 29: 501-522.
P'an, A. Y. S.; Beland, J.; Jegier, Z. (1972) Ozone-induced arterial lesions. Arch. Environ. Health 24: 229-232.
Parks, W. C.; Roby, J. D. (1994) Consequences of prolonged inhalation of ozone on F344/N rats: collaborative
studies. Part IV. Effects on expression of extracellular matrix genes. Cambridge, MA: Health Effects
Institute; research report no. 65.
Peden, D. B.; Swiersz, M.; Ohkubo, K.; Hahn, B.; Emery, B.; Kaliner, M. A. (1993) Nasal secretion of the
ozone scavenger uric acid. Am. Rev. Respir. Dis. 148: 455-461.
Peterson, D. C.; Andrews, H. L. (1963) The role of ozone in radiation avoidance in the mouse. Radial. Res.
19: 331-336.
Phalen, R. P.; Kenoyer, J. L.; Crocker, T. T.; McClure, T. R. (1980) Effects of sulfate aerosols in combination
with ozone on elimination of tracer particles inhaled by rats. J. Toxicol. Environ. Health 6: 797-810.
6-192
-------
Phalen, R. P.; Crocker, T. T.; McClure, T. R.; Tyler, N. K. (1986) Effect of ozone on mean linear intercept in
the lung of young beagles. J. Toxicol. Environ. Health 17: 285-296.
Phipps, R. J.; Denas, S. M.; Sielczak, M. W.; Wanner, A. (1986) Effects of 0.5 ppni ozone on glycoprotein
secretion, ion and water fluxes in sheep trachea. J. Appl. Physiol. 60: 918-927.
Pickrell, J. A.; Gregory, R. E.; Cole, D. J.; Hahn, F. P.; Henderson, R. P. (1987a) Effect of acute ozone
exposure on the proteinase-antiproteinase balance in the rat lung. Exp. Mol. Pathol. 46: 168-179.
Pickrell, J. A.; Hahn, F. P.; Rebar, A. H.; Horoda, R. A.; Henderson, R. F. (1987b) Changes in collagen
metabolism and proteinolysis after repeated inhalation exposure to ozone. Exp. Mol. Pathol. 46: 159-
167.
Pinkerton, K. E.; Brody, A. R.; Miller, F. J.; Crapo, J. D. (1989) Exposure to low levels of ozone results in
enhanced pulmonary retention of inhaled asbestos fibers. Am. Rev. Respir. Dis. 140: 1075-1081.
Pinkerton, K. E.; Mercer, R. R.; Plopper, C. G.; Crapo, J. D. (1992) Distribution of injury and microdosimetry
of ozone in the ventilatory unit of the rat. J. Appl. Physiol. 73: 817-824.
Pinkerton, K. E.; Dodge, D. E.; Cederdahl-Demmler, J.; Wong, V. J.; Peake, J.; Haselton, C. J.; Mellick,
P. W.; Singh, G.; Plopper, C. G. (1993) Differentiated bronchiolar epithelium in alveolar ducts of
rats exposed to ozone for 20 months. Am. J. Pathol. 142: 947-956.
Pinkerton, K. E.; Menache, M. G.; Plopper, C. G. (1995) Consequences of prolonged inhalation of ozone on
F344/N rats: collaborative studies. Part IX. Changes in the tracheobronchial epithelium, pulmonary
acinus, and lung antioxidant enzyme activity. Cambridge, MA: Health Effects Institute; pp. 41-98;
research report number 65.
Pino, M. V.; McDonald, R. J.; Berry, J. D.; load, J. P.; Tarkington, B. K.; Hyde, D. M. (1992a) Functional and
morphologic changes caused by acute ozone exposure in the isolated and perfused rat lung. Am. Rev.
Respir. Dis. 145: 882-889.
Pino, M. V.; Stovall, M. Y.; Levin, J. R.; Devlin, R. B.; Koren, H. S.; Hyde, D. M. (1992b) Acute ozone-
induced lung injury in neutrophil-depleted rats. Toxicol. Appl. Pharmacol. 114: 268-276.
Pino, M. V.; Levin, J. R.; Stovall, M. Y.; Hyde, D. M. (1992c) Pulmonary inflammation and epithelial injury in
response to acute ozone exposure in the rat. Toxicol. Appl. Pharmacol. 112: 64-72.
Plopper, C. G. (1983) Comparative morphologic features of bronchiolar epithelial cells: the Clara cell. Am. Rev.
Respir. Dis. 128: S37-S42.
Plopper, C. G.; Chow, C. K.; Dungworth, D. L.; Brummer, M.; Nemeth, T. J. (1978) Effect of low level of
ozone on rat lungs: II. morphological responses during recovery and re-exposure. Exp. Mol. Pathol.
29:400-411.
Plopper, C. G.; Dungworth, D. L.; Tyler, W. S.; Chow, C. K. (1979) Pulmonary alterations in rats exposed to
0.2 and 0.1 ppm ozone: a correlated morphological and biochemical study. Arch. Environ. Health
34: 390-395.
Plopper, C. G.; Harkema, J. R.; Last, J. A.; Pinkerton, K. E.; Tyler, W. S.; St. George, J. A.; Wong, V. J.;
Nishio, S. J.; Weir, A. S.; Dungworth, D. L.; Barry, B. E.; Hyde, D. M. (1991) The respiratory
system of nonhunian primates responds more to ambient concentrations of ozone than does that of
rats. In: Berglund, R. L.; Lawson, D. R.; McKee, D. J., eds. Tropospheric ozone and the
environment: papers from an international conference; March 1990; Los Angeles, CA. Pittsburgh,
PA: Air & Waste Management Association; pp. 137-150. (A&WMA transaction series no. TR-19).
6-193
-------
Plopper, C. G.; Chu, F.-P.; Haselton, C. J.; Peake, J.; Wu, J.; Pinkerton, K. E. (1994a) Dose-dependent
tolerance to ozone: I. tracheobronchial epithelial reorganization in rats after 20 months' exposure.
Am. J. Pathol. 144: 404-420.
Plopper, C. G.; Duan, X.; Buckpitt, A. R.; Pinkerton, K. E. (1994b) Dose-dependent tolerance to ozone.
IV. Site-specific elevation in antioxidant enzymes in the lungs of rats exposed for 90 days or
20 months. Toxicol. Appl. Pharmacol. 127: 124-131.
Prasad, S. B.; Rao, V. S.; Mannix, R. C.; Phalen, R. F. (1988) Effects of pollutant atmospheres on surface
receptors of pulmonary niacrophages. J. Toxicol. Environ. Health 24: 385-402.
Pryor, W. A. (1978) The formation of free radicals and the consequences of their reactions in vivo. Photocheni.
Photobiol. 28: 787-801.
Pryor, W. A. (1991) Can vitamin E protect humans against the pathological effects of ozone in smog?
Am. J. Clin. Nutr. 53: 702-722.
Pryor, W. A. (1992) How far does ozone penetrate into the pulmonary air/tissue boundary before it reacts? Free
Radic. Biol. Med. 12: 83-88.
Pryor, W. A. (1993) Ozone in all its reactive splendor. J. Lab. Clin. Med. 122: 483-486.
Pryor, W. A.; Das, B.; Church, D. F. (1991) The ozonation of unsaturated fatty acids: aldehydes and hydrogen
peroxide as products and possible mediators of ozone toxicity. Chem. Res. Toxicol. 4: 341-348.
Pryor, W. A.; Wang, K.; Bermudez, E. (1992) Cholesterol ozonation products as biomarkers for ozone exposure
in rats. Biochem. Biophys. Res. Commun. 188: 618-623.
Rabinowitz, J. L.; Bassett, D. J. P. (1988) Effect of 2 ppm ozone exposure on rat lung lipid fatty acids. Exp.
Lung Res. 14: 477-489.
Radhakrishnamurthy, B. (1994) Consequences of prolonged inhalation of ozone on F344/N rats: collaborative
studies. Part III: Effects on complex carbohydrates of lung connective tissue. Cambridge, MA:
Health Effects Institute; research report no. 65.
Rahman, I.; Massaro, D. (1992) Endotoxin treatment protects rats against ozone-induced lung edema: with
evidence for the role of manganese superoxide disniutase. Toxicol. Appl. Pharmacol. 113: 13-18.
Rahman, I. U.; Clerch, L. B.; Massaro, D. (1991) Rat lung antioxidant enzyme induction by ozone.
Am. J. Physiol. 260: L412-L418.
Rajini, P.; Gelzleichter, T. R.; Last, J. A.; Witschi, H. (1993) Alveolar and airway cell kinetics in the lungs of
rats exposed to nitrogen dioxide, ozone, and a combination of the two gases. Toxicol. Appl.
Pharmacol. 121: 186-192.
Rao, G. A.; Larkin, E. C.; Harkema, J. R.; Dungworth, D. L. (1985a) Changes in the levels of polyunsaturated
fatty acids in the lung and lecithin cholesterol acyl transferase activity in plasma of monkeys exposed
to ambient levels of ozone. Toxicol. Lett. 24: 125-129.
Rao, G. A.; Larkin, E. C.; Harkema, J. R.; Dungworth, D. L. (1985b) Changes in lipids of lung lavage in
monkeys after chronic exposure to ambient levels of ozone. Toxicol. Lett. 29: 207-214.
Rasniussen, R. E. (1986) Inhibition of DMA replication by ozone in Chinese hamster V79 cells. J. Toxicol.
Environ. Health 17: 119-128.
6-194
-------
Raub, J. A.; Miller, F. J.; Graham, J. A. (1983) Effects of low-level ozone exposure on pulmonary function in
adult and neonatal rats. In: Lee, S. D.; Mustafa, M. G.; Mehlman, M. A., eds. International
symposium on the biomedical effects of ozone and related photochemical oxidants; March 1982;
Pinehurst, NC. Princeton, NJ: Princeton Scientific Publishers, Inc.; pp. 363-367. (Advances in
modern environmental toxicology: v. 5).
Reasor, M. J.; Adams, G. K., Ill; Brooks, J. K.; Rubin, R. J. (1979) Enrichment of albumin and IgG in the
airway secretions of dogs breathing ozone. J. Environ. Sci. Health Part C 13: 335-346.
Reiser, K. M.; Tyler, W. S.; Hennessy, S. M.; Dominguez, J. J.; Last, J. A. (1987) Long-term consequences of
exposure to ozone: II. structural alterations in lung collagen of monkeys. Toxicol. Appl. Pharmacol.
89: 314-322.
Reuzel, P. G. J.; Wilmer, J. W. G. M.; Woutersen, R. A.; Zwart, A.; Rombout, P. J. A.; Feron, V. J. (1990)
Interactive effects of ozone and formaldehyde on the nasal respiratory lining epithelium in rats. J.
Toxicol. Environ. Health 29: 279-292.
Richters, A. (1988) Effects of nitrogen dioxide and ozone on blood-borne cancer cell colonization of the lungs.
J. Toxicol. Environ. Health 25: 383-390.
Richters, A. ; Kuraitis, K. (1981) Inhalation of NQ and blood borne cancer cell spread to the lungs. Arch.
Environ. Health 36: 36-39.
Rietjens, I. M. C. M.; Lemmink, H. H.; Alink, G. M.; Van Bladeren, P. J. (1987a) The role of glutathione and
glutathione S-transferases in fatty acid ozonide detoxification. Chem. Biol. Interact. 62: 3-14.
Rietjens, I. M. C. M.; Van Tilburg, C. A. M.; Coenen, T. M. M.; Alink, G. M.; Konings, A. W. T. (1987b)
Influence of polyunsaturated fatty acid supplementation and membrane fluidity on ozone and nitrogen
dioxide sensitivity of rat alveolar macrophages. J. Toxicol. Environ. Health 21: 45-56.
Rietjens, I. M. C. M.; Dormans, J. A. M. A.; Rombout, P. J. A.; Van Bree, L. (1988) Qualitative and
quantitative changes in cytochrome P-450-dependent xenobiotic metabolism in pulmonary
microsomes and isolated Clara cell populations derived from ozone-exposed rats. J. Toxicol.
Environ. Health 24: 515-531.
Rithidech, K.; Hotchkiss, J. A.; Griffith, W. C.; Henderson, R. P.; Brooks, A. L. (1990) Chromosome damage
in rat pulmonary alveolar macrophages following ozone inhalation. Mutat. Res. 241: 67-73.
Rodriguez, M.; Bur, S.; Favre, A.; Weibel, E. R. (1987) Pulmonary acinus: geometry and morphometry of the
peripheral airway system in rat and rabbit. Am. J. Anal. 180: 143-155.
Roehm, J. N. ; Hadley, J. G.; Menzel, D. B. (1972) The influence of vitamin E on the lung fatty acids of rats
exposed to ozone. Arch. Environ. Health 24: 237-242.
Rombout, P. J. A.; Van Bree, L.; Heisterkamp, S. H.; Marra, M. (1989) The need for an eight hour ozone
standard. In: Schneider, T.; Lee, S. D.; Wolters, G. J. R. ; Grant, L. D., eds. Atmospheric ozone
research and its policy implications: proceedings of the 3rd US-Dutch international symposium; May
1988; Nijmegen, The Netherlands. Amsterdam, The Netherlands: Elsevier Science Publishers;
pp. 701-710. (Studies in environmental science 35).
Roum, J. H.; Murlas, C. (1984) Ozone-induced changes in muscarinic bronchial reactivity by different testing
methods. J. Appl. Physiol.: Respir. Environ. Exercise Physiol. 57: 1783-1789.
6-195
-------
Ryer-Powder, J. E.; Amoraso, M. A.; Czerniecki, B.; Witz, G.; Goldstein, B. D. (1988) Inhalation of ozone
produces a decrease in superoxide anion radical production in mouse alveolar macrophages.
Am. Rev. Respir. Dis. 138: 1129-1133.
Sagai, M.; Ichinose, T. (1991) Biochemical effects of combined gases of nitrogen dioxide and ozone. IV. Changes
of lipid peroxidation and antioxidative protective systems in rat lungs upon life span exposure.
Toxicology 66: 121-132.
Sagai, M.; Arakawa, K.; Ichinose, T.; Shimojo, N. (1987) Biochemical effects on combined gases of nitrogen
dioxide and ozone. I. Species differences of lipid peroxides and phospholipids in lungs. Toxicology
46: 251-265.
Saldiva, P. H. N.; King, M.; Delmonte, V. L. C.; Macchione, M.; Parada, M. A. C.; Daliberto, M. L.; Sakae,
R. S.; Criado, P. M. P.; Silveira, P. L. P.; Zin, W. A.; Bohm, G. M. (1992) Respiratory alterations
due to urban air pollution: an experimental study in rats. Environ. Res. 57: 19-33.
Santrock, J.; Gorski, R. A.; O'Gara, J. F. (1992) Products and mechanism of the reaction of ozone with
phospholipids in unilaniellar phospholipid vesicles. Cheni. Res. Toxicol. 5: 134-141.
Sasaki, K.; Nadel, J. A.; Hahn, H. L. (1987) Effect of ozone on breathing in dogs: vagal and nonvagal
mechanisms. J. Appl. Physiol. 62: 15-26.
Sawadaishi, K.; Miura, K.; Ohtsuka, E.; Ueda, T.; Ishizaki, K.; Shinriki, N. (1985) Ozonolysis of supercoiled
pBR322 DMA resulting in strand scission to open circular DMA. Nucleic Acids Res. 13: 7183-7194.
Sawadaishi, K.; Miura, K.; Ohtsuka, E.; Ueda, T.; Shinriki, N.; Ishizaki, K. (1986) Structure- and sequence-
specificity of ozone degradation of supercoiled plasmid DMA. Nucleic Acids Res. 14: 1159-1169.
Schelegle, E. S.; Carl, M. L.; Coleridge, H. M.; Coleridge, J. C. G.; Green, J. F. (1993) Contribution of vagal
afferents to respiratory reflexes evoked by acute inhalation of ozone in dogs. J. Appl. Physiol.
74: 2338-2344.
Schlesinger, R. B. (1989) Deposition and clearance of inhaled particles. In: McClellan, R. O.; Henderson, R. P.,
eds. Concepts in inhalation toxicology. New York, NY: Hemisphere Publishing Corp.; pp. 163-192.
Schlesinger, R. B. (1992) Nitrogen oxides. In: Lippmann, M., ed. Environmental toxicants: human exposures and
their health effects. New York, NY: Van Nostrand Reinhold; pp. 412-453.
Schlesinger, R. B.; Driscoll, K. E. (1987) Mucociliary clearance from the lungs of rabbits following single and
intermittent exposures to ozone. J. Toxicol. Environ. Health 20: 125-134.
Schlesinger, R. B.; Driscoll, K. E.; Gunnison, A. P.; Zelikoff, J. T. (1990) Pulmonary arachidonic acid
metabolism following acute exposures to ozone and nitrogen dioxide. J. Toxicol. Environ. Health
31: 275-290.
Schlesinger, R. B.; Weideman, P. A.; Zelikoff, J. T. (1991) Effects of repeated exposures to ozone and nitrogen
dioxide on respiratory tract prostanoids. Inhalation Toxicol. 3: 27-36.
Schlesinger, R. B.; Gorczynski, J. E.; Dennison, J.; Richards, L.; Kinney, P. L.; Bosland, M. C. (1992a) Long-
term intermittent exposure to sulfuric acid aerosol, ozone and their combination: alterations in
tracheobronchial mucociliary clearance and epithelial secretory cells. Exp. Lung Res. 18: 505-534.
Schlesinger, R. B.; Zelikoff, J. T.; Chen, L. C.; Kinney, P. L. (1992b) Assessment of toxicologic interactions
resulting from acute inhalation exposure to sulfuric acid and ozone mixtures. Toxicol. Appl.
Pharmacol. 115: 183-190.
6-196
-------
Schreider, J. P.; Raabe, O. G. (1981) Anatomy of the nasal-pharyngeal airway of experimental animals. Anal.
Rec. 200: 195-205.
Schultheis, A. H.; Bassett, D. J. P. (1991) Inflammatory cell influx into ozone-exposed guinea pig lung interstitial
and airways spaces. Agents Actions 34: 270-273.
Schultheis, A. H.; Bassett, D. J. P.; Fryer, A. D. (1994) Ozone-induced airway hyperresponsiveness and loss of
neuronal M2 muscarinic receptor function. J. Appl. Physiol. 76: 1088-1097.
Schwartz, L. W.; Dungworth, D. L.; Mustafa, M. G.; Tarkington, B. K.; Tyler, W. S. (1976) Pulmonary
responses of rats to ambient levels of ozone: effects of 7-day intermittent or continuous exposure.
Lab. Invest. 34: 565-578.
Scott, C. D.; Burkart, J. A. (1978) Chromosomal aberrations in peripheral lymphocytes of students exposed to air
pollutants. Research Triangle Park, NC: U.S. Environmental Protection Agency, Health Effects
Research Laboratory; report no. EPA-600/1-78-054. Available from: NTIS, Springfield, VA;
PB-285594.
Scott, D. B. M.; Lesher, E. C. (1963) Effect of ozone on survival and permeability ofEscherichia coll.
J. Bacteriol. 85: 567-576.
Selgrade, M. K. (1995) Immunotoxicity and risk assessment: effect of temporal factors. Inhalation Toxicol.
7: 891-901.
Selgrade, M. K.; Illing, J. W.; Starnes, D. M.; Stead, A. G.; Menache, M. G.; Stevens, M. A. (1988)
Evaluation of effects of ozone exposure on influenza infection in mice using several indicators of
susceptibility. Fundam. Appl. Toxicol. 11: 169-180.
Selgrade, M. K.; Daniels, M. J.; Grose, E. C. (1990) Acute, subchronic, and chronic exposure to a simulated
urban profile of ozone: effects on extrapulmonary natural killer cell activity and lymphocyte
mitogenic responses. Inhalation Toxicol. 2: 375-389.
Shelley, S. A.; Paciga, J. E.; Balis, J. U. (1984) Lung surfactant phospholipids in different animal species. Lipids
19: 857-862.
Sherwin, R. P.; Richters, V. (1985) Effect of 0.3 ppm ozone exposure on type II cells and alveolar walls of
newborn mice: an image-analysis quantitation. J. Toxicol. Environ. Health 16: 535-546.
Sherwood, R. L.; Lippert, W. E.; Goldstein, E. (1986) Effect of 0.64 ppm ozone on alveolar macrophage
lysozyme levels in rats with chronic pulmonary bacterial infection. Environ. Res. 41: 378-387.
Shimura, S.; Maeda, S.; Takisminia, T. (1984) Giant lamellar bodies in alveolar type II cells of rats exposed to a
low concentration of ozone. Respiration 46: 303-309.
Shiotsuka, R. N.; Yermakoff, J. K.; Osheroff, M. R.; Drew, R. T. (1986) The combination of ozone and silica on
the development of pulmonary fibrosis. J. Toxicol. Environ. Health 17: 297-310.
Shiraishi, P.; Bandow, H. (1985) The genetic effects of the photochemical reaction products of propylene plus
NO2 on cultured Chinese hamster cells exposed in vitro. J. Toxicol. Environ. Health 15: 531-538.
Shiraishi, P.; Hashimoto, S.; Bandow, H. (1986) Induction of sister-chromatid exchanges in Chinese hamster V79
cells by exposure to the photochemical reaction products of toluene plus NO, in the gas phase. Mutat.
Res. 173: 135-139.
6-197
-------
Sibille, Y.; Reynolds, H. Y. (1990) Macrophages and polymorphonuclear neutrophils in lung defense and injury.
Am. Rev. Respir. Dis. 141: 471-501.
Silbaugh, S. A.; Mauderly, J. L. (1986) Effects of ozone and sulfuric acid aerosol on gas trapping in the guinea
pig lung. J. Toxicol. Environ. Health 18: 133-141.
Slade, R.; Highfill, J. W.; Hatch, G. E. (1989) Effects of depletion of ascorbic acid or nonprotein sulfhydryls on
the acute inhalation toxicity of nitrogen dioxide, ozone, and phosgene. Inhalation Toxicol. 1: 261-
271.
Smiler, K. L.; Brady, A. N.; Anver, M. R. (1988) Histopathological effects of chronic exposure of Fischer 344
rats to low-levels of ozone. Warren, MI: General Motors Research Laboratories, Biomedical Science
Department; research publication no. GMR-6355.
Snider, G. L.; Kleinerman, J.; Thurlbeck, W. M.; Bengali, Z. H. (1985) The definition of emphysema: report of
a National Heart, Lung, and Blood Institute, Division of Lung Diseases workshop. Am. Rev. Respir.
Dis. 132: 182-185.
Spengler, J. D.; Keeler, G. J.; Koutrakis, P.; Ryan, P. B.; Raizenne, M.; Franklin, C. A. (1989) Exposures to
acidic aerosols. In: Symposium on the health effects of acid aerosols; October 1987; Research
Triangle Park, NC. Environ. Health Perspect. 79: 43-51.
Stephens, R. J.; Sloan, M. P.; Evans, M. J.; Freeman, G. (1974a) Early response of lung to low levels of ozone.
Am. J. Pathol. 74: 31-58.
Stephens, R. J.; Sloan, M. P.; Evans, M. J.; Freeman, G. (1974b) Alveolar Type 1 cell response to exposure to
0.5 ppm O3 for short periods. Exp. Mol. Pathol. 20: 11-23.
Stephens, R. J.; Sloan, M. P.; Groth, D. G.; Negi, D. S.; Lunan, K. D. (1978) Cytologic response of postnatal
rat lungs to O3 or NO2 exposure. Am. J. Pathol. 93: 183-200.
Stephens, R. J.; Buntman, D. J.; Negi, D. S.; Parkhurst, R. M.; Thomas, D. W. (1983) Tissue levels of
vitamin E in the lung and the cellular response to injury resulting from oxidant gas exposure. Chest
83(suppl.): 37S-39S.
Stiles, J.; Tyler, W. S. (1988) Age-related niorphometric differences in responses of rat lungs to ozone. Toxicol.
Appl. Pharmacol. 92: 274-285.
Stoner, G. D. (1991) Lung tumors in strain A mice as a bioassay for carcinogenicity of environmental chemicals.
Exp. Lung Res. 17: 405-423.
Stoner, G. D.; Shimkin, M. B. (1985) Lung tumors in strain A mice as a bioassay for carcinogenicity.
In: Milman, H. A.; Weisburger, E. K., eds. Handbook of carcinogen testing. Park Ridge, NJ: Noyes
Publications; pp. 179-214.
Suzuki, E.; Takahashi, Y.; Aida, S.; Kimula, Y.; Ito, Y.; Miura, T. (1992) Alteration in surface structure of
Clara cells and pulmonary cytochrome F-450b level in rats exposed to ozone. Toxicology
71: 223-232.
Szarek, J. L. (1994) Consequences of prolonged inhalation of ozone on F344/N rats: collaborative studies. Part II:
Mechanical properties, responses to bronchoactive stimuli, and eicosanoid release in isolated large
and small airways. Cambridge, MA: Health Effects Institute; research report no. 65.
Takahashi, Y.; Miura, T. (1985) In vivo effects of nitrogen dioxide and ozone on xenobiotic metabolizing systems
of rat lungs. Toxicol. Lett. 26: 145-152.
6-198
-------
Takahashi, Y.; Miura, T. (1987) A selective enhancement of xenobiotic metabolizing systems of rat lungs by
prolonged exposure to ozone. Environ. Res. 42: 425-434.
Takahashi, Y.; Miura, T. (1989) Effects of nitrogen dioxide and ozone in combination on xenobiotic metabolizing
activities of rat lungs. Toxicology 56: 253-262.
Takahashi, Y.; Miura, T. (1990) Responses of cytochronie P-450 isozynies of rat lung to in vivo exposure to
ozone. Toxicol. Lett. 54: 327-335.
Takahashi, Y.; Miura, T.; Kubota, K. (1985) In vivo effect of ozone inhalation on xenobiotic metabolism of lung
and liver of rats. J. Toxicol. Environ. Health 15: 855-864.
Takahashi, Y.; Miura, T.; Kimura, S. (1990) A decrease in serum retinol by in vivo exposures of rats to ozone.
Int. J. Vitam. Nutr. Res. 60: 294-295.
Tan, W. C.; Bethel, R. A. (1992) The effect of platelet activating factor antagonist on ozone-induced airway
inflammation and bronchial hyperresponsiveness in guinea pigs. Am. Rev. Respir. Dis.
146: 916-922.
Tanswell, A. K. (1989) Detection of paracrine factors in oxidant lung injury. Cambridge, MA: Health Effects
Institute; research report no. 22.
Tanswell, A. K.; Fraher, L. J.; Grose, E. C. (1989) Circulating factors that modify lung cell DMA synthesis
following exposure to inhaled oxidants. I. Effect of serum and lav age on lung fibroblasts following
exposure of adult rats to 1 ppm ozone. J. Toxicol. Environ. Health 27: 239-254.
Tanswell, A. K.; Fraher, L. J.; Grose, E. C. (1990) Circulating factors that modify lung cell DMA synthesis
following exposure to inhaled oxidants. II. Effect of serum and lav age on lung pneumocytes
following exposure of adult rats to 1 ppm ozone. J. Toxicol. Environ. Health 29: 131-144.
Tarkington, B. K.; Wu, R.; Sun, W.-M.; Nikula, K. J.; Wilson, D. W.; Last, J. A. (1994) In vitro exposure of
tracheobronchial epithelial cells and of tracheal explants to ozone. Toxicology 88: 51-68.
Teige, B.; McManus, T. T.; Mudd, J. B. (1974) Reaction of ozone with phosphatidylcholine liposomes and the
lytic effect of products on red blood cells. Chem. Phys. Lipids 12: 153-171.
Tepper, J. S.; Weiss, B. (1986) Determinants of behavioral response with ozone exposure. J. Appl. Physiol.
60: 868-875.
Tepper, J. S.; Wood, R. W. (1985) Behavioral evaluation of the irritating properties of ozone. Toxicol. Appl.
Pharmacol. 78: 404-411.
Tepper, J. L.; Weiss, B.; Cox, C. (1982) Microanalysis of ozone depression of motor activity. Toxicol. Appl.
Pharmacol. 64: 317-326.
Tepper, J. L.; Weiss, B.; Wood, R. W. (1983) Behavioral indices of ozone exposure. In: Lee, S. D.; Mustafa,
M. G.; Mehlnian, M. A., eds. International symposium on the biomedical effects of ozone and
related photochemical oxidants; March 1982; Pinehurst, NC. Princeton, NJ: Princeton Scientific
Publishers, Inc.; pp. 515-526. (Advances in modern environmental toxicology: v. 5).
Tepper, J. S.; Weiss, B.; Wood, R. W. (1985) Alterations in behavior produced by inhaled ozone or ammonia.
Fundam. Appl. Toxicol. 5: 1110-1118.
6-199
-------
Tepper, J. S.; Costa, D. L.; Lehmann, J. R.; Weber, M. P.; Hatch, G. E. (1989) Unattenuated structural and
biochemical alterations in the rat lung during functional adaptation to ozone. Am. Rev. Respir. Dis.
140: 493-501.
Tepper, J. S.; Wiester, M. J.; Weber, M. P.; Menache, M. G. (1990) Measurements of cardiopulmonary
response in awake rats during acute exposure to near-ambient concentrations of ozone. Fundam.
Appl. Toxicol. 10: 7-15.
Tepper, J. S.; Wiester, M. J.; Weber, M. P.; Fitzgerald, S.; Costa, D. L. (1991) Chronic exposure to a simulated
urban profile of ozone alters ventilatory responses to carbon dioxide challenge in rats. Fundam. Appl.
Toxicol. 17: 52-60.
Tepper, J. S.; et al. (1994) The relative contribution of concentration and duration to ozone (Q)-induced lung
injury. Research Triangle Park, NC: U.S. Environmental Protection Agency, National Health &
Environmental Effects Research Laboratory, Pulmonary Toxicology Branch.
Tepper, J. S.; et al. (1995) Time course of inflammation in the rat exposed to ozone for short and prolonged
periods with intermittent hyperventilation. Research Triangle Park, NC: U.S. Environmental
Protection Agency, National Health & Environmental Effects Research Laboratory, Pulmonary
Toxicology Branch.
Terada, N.; Mizoguchi, L; Kohno, T.; Hayashi, Y. (1986) Pulmonary effects in rats induced by prolonged
exposure to a mixture of O3 and NO2. J. Jpn. Soc. Air Pollut. 21: 512-520.
Thomassen, D. G. (1992) Understanding mechanisms of carcinogenesis using rat tracheal epithelial cells in vitro.
In: Proceedings of the symposium on current concepts and approaches on animal test alternatives;
February; Aberdeen Proving Grounds, MD. Washington, DC: U.S. Department of Energy;
CONF-9202107-1. Available from: NTIS, Springfield, VA; DE92-013510.
Thomassen, D. G.; Harkema, J. R.; Stephens, N. D.; Griffith, W. C. (1991) Preneoplastic transformation of rat
tracheal epithelial cells by ozone. Toxicol. Appl. Pharmacol. 109: 137-148.
Thomassen, D. G.; Harkema, J. R.; Sun, J. D.; Stephens, N. D.; Griffith, W. C. (1992) The role of ozone in
tracheal cell transformation. Cambridge, MA: Health Effects Institute; research report no. 50.
Tice, R. R.; Bender, M. A.; Ivett, J. L.; Drew, R. T. (1978) Cytogenetic effects of inhaled ozone. Mutat. Res.
58: 293-304.
Tilton, B. E. (1986) Reviews of paper by Hassett et al., "Murine lung carcinogenesis following exposure to
ambient ozone concentrations" (JNCI 75: Ill-Ill) [memorandum to B. Jordan, D. J. McKee, and
H. Richmond, U.S. EPA/OAQPS]. Research Triangle Park, NC: U.S. Environmental Protection
Agency, Office of Health and Environmental Assessment, Environmental Criteria and Assessment
Office; March 24.
Turner, C. R.; Kleeberger, S. R.; Spannhake, E. W. (1989) Preexposure to ozone blocks the antigen-induced late
asthmatic response of the canine peripheral airways. J. Toxicol. Environ. Health 28: 363-371.
Tyler, W. S. (1983) Comparative subgross anatomy of lungs: pleuras, interlobular septa, and distal airways.
Am. Rev. Respir. Dis. 128: S32-S36.
Tyler, W. S.; Julian, M. D. (1991) Gross and subgross anatomy of lungs, pleura, connective tissue septa, distal
airways, and structural units. In: Parent, R. A., ed. Comparative biology of the normal lung: v. 1;
treatise on pulmonary toxicology. Boca Raton, PL: CRC Press; pp. 37-48.
6-200
-------
Tyler, W. S.; Tyler, N. K.; Last, J. A.; Barstow, T. J.; Magliano, D. J.; Hinds, D. M. (1987) Effects of ozone
on lung and somatic growth. Pair fed rats after ozone exposure and recovery periods. Toxicology
46: 1-20.
Tyler, W. S.; Tyler, N. K.; Last, J. A.; Gillespie, M. J.; Barstow, T. J. (1988) Comparison of daily and seasonal
exposures of young monkeys to ozone. Toxicology 50: 131-144.
Tyler, W. S.; Tyler, N. K.; Hinds, D.; Magliano, D.; Tarkington, B. (1991a) Influence of exposure regimen on
effects of experimental ozone studies: effects of daily and episodic or seasonal cycles of exposure and
post-exposure. Presented at: 84th annual meeting and exhibition of the Air & Waste Management
Association; June; Vancouver, BC, Canada. Pittsburgh, PA: Air & Waste Management Association;
paper no. 91-141.5.
Tyler, W. S.; Tyler, N. K.; Magliano, D. J.; Hinds, D. M.; Tarkington, B.; Julian, M. D.; Hyde, D. M.;
Plopper, C. G.; Dungworth, D. L. (1991b) Effects of ozone inhalation on lungs of juvenile monkeys.
Morphometry after a 12 month exposure and following a 6 month post-exposure period. In:
Berglund, R. L.; Lawson, D. R.; McKee, D. J., eds. Tropospheric ozone and the environment:
papers from an international conference; March 1990; Los Angeles, CA. Pittsburgh, PA: Air &
Waste Management Association; pp. 151-160. (A&WMA transaction series no. TR-19).
Tyler, W.; Jones, J.; Birks, E.; Pascoe, J.; Steffey, E.; Jarvis, K.; Hinds, D.; Tarkington, B. (1991c) Effects of
ozone on exercising horses: a preliminary report. Equine Exercise Physiol. 3: 490-502.
Tyson, C. A.; Lunan, K. D.; Stephens, R. J. (1982) Age-related differences in GSH-shuttle enzymes in NQ- or
O3-exposed rat lungs. Arch. Environ. Health 37: 167-176.
U.S. Environmental Protection Agency. (1986) Air quality criteria for ozone and other photochemical oxidants.
Research Triangle Park, NC: Office of Health and Environmental Assessment, Environmental
Criteria and Assessment Office; report nos. EPA-600/8-84-020aF-eF. 5v. Available from: NTIS,
Springfield, VA; PB87-142949.
U.S. Environmental Protection Agency. (1989) An acid aerosols issue paper: health effects and aeronietrics.
Research Triangle Park, NC: Office of Health and Environmental Assessment, Environmental
Criteria and Assessment Office; report no. EPA-600/8-88-005F. Available from: NTIS, Springfield,
VA; PB91-125864.
U.S. Environmental Protection Agency. (1993) Air quality criteria for oxides of nitrogen. Research Triangle Park,
NC: Office of Health and Environmental Assessment, Environmental Criteria and Assessment
Office; report nos. EPA/600/8-91/049aF-cF. 3v. Available from: NTIS, Springfield, VA;
PB95-124533, PB95-124525, and PB95-124517.
Uchida, D. A.; Ballowe, C. A.; Irvin, C. G.; Larsen, G. L. (1992) Assessment of airway responsiveness to
inhaled methacholine and the effects of short-term ozone exposure in aged Fischer-344 rats. HEI
Commun. 1: 20-23.
Uchiyama, L; Yokoyama, E. (1989) Effects of short- and long-term exposure to ozone on heart rate and blood
pressure of emphysematous rats. Environ. Res. 48: 76-86.
Uchiyama, I.; Simomura, Y.; Yokoyama, E. (1986) Effects of acute exposure to ozone on heart rate and blood
pressure of the conscious rat. Environ. Res. 41: 529-537.
Umezu, T.; Suzuki, A. K.; Miura, T.; Koizumi, A. (1993) Effects of ozone and nitrogen dioxide on drinking and
eating behaviors in mice. Environ. Res. 61: 51-67.
6-201
-------
VanBree, L.; Rombout, P. J. A.; Rietjens, I. M. C. M.; Dormans, J. A. M. A.; Marra, M. (1989)
Pathobiochemical effects in rat lung related to episodic ozone exposure. In: Schneider, T.; Lee,
S. D.; Wolters, G. J. R.; Grant, L. D., eds. Atmospheric ozone research and its policy implications:
proceedings of the 3rd US-Dutch international symposium; May 1988; Nijmegen, The Netherlands.
Amsterdam, The Netherlands: Elsevier Science Publishers; pp. 723-732. (Studies in environmental
science 35).
Van Bree, L.; Marra, M.; Rombout, P. J. A. (1992) Differences in pulmonary biochemical and inflammatory
responses of rats and guinea pigs resulting from daytime or nighttime, single and repeated exposure
to ozone. Toxicol. Appl. Pharmacol. 116: 209-216.
Van der Zee, J.; Van Beek, E.; Dubbelman, T. M. A. R.; Van Steveninck, J. (1987) Toxic effects of ozone on
murine L929 fibroblasts: damage to DNA. Biochem. J. 247: 69-72.
Van Loveren, H.; Rombout, P. J. A.; Wagenaar, Sj. Sc.; Walvoort, H. C.; Vos, J. G. (1988) Effects of ozone on
the defense to a respiratory Listeria monocytogenes infection in the rat: suppression of macrophage
function and cellular immunity and aggravation of histopathology in lung and liver during infection.
Toxicol. Appl. Pharmacol. 94: 374-393.
Van Loveren, H.; Krajnc, E. L; Rombout, P. J.; Blommaert, F. A.; Vos, J. G. (1990) Effects of ozone,
hexachlorobenzene, and bis(tri-«-butyltin)oxide on natural killer activity in the rat lung. Toxicol.
Appl. Pharmacol. 102: 21-33.
Veninga, T. S.; Evelyn, P. (1986) Activity changes of pulmonary rnacrophages after in vivo exposure to ozone as
demonstrated by cell adherence. J. Toxicol. Environ. Health 18: 483-489.
Veninga, T. S.; Fidler, V. (1986) Ozone-induced elevation of creatine kinase activity in blood plasma of rats.
Environ. Res. 41: 168-173.
Victorin, K. (1992) Review of the genotoxicity of ozone. Mutat. Res. 277: 221-238.
Victorin, K.; Stlhlberg, M. (1988a) Photochemical formation of mutagenic compounds from alkenes and ozone or
nitrogen dioxide. Environ. Mol. Mutagen. 11: 79-90.
Victorin, K.; Stlhlberg, M. (1988b) A method for studying the mutagenicity of some gaseous compounds in
Salmonella typhimurium. Environ. Mol. Mutagen. 11: 65-77.
Walters, E. H.; O'Byrne, P. M.; Graf, P. D.; Fabbri, L. M.; Nadel, J. A. (1986) The responsiveness of airway
smooth muscle in vitro from dogs with airway hyper-responsiveness in vivo. Clin. Sci. 71: 605-611.
Warren, D. L.; Last, J. A. (1987) Synergistic interaction of ozone and respirable aerosols on rat lungs: III. ozone
and sulfuric acid aerosol. Toxicol. Appl. Pharmacol. 88: 203-216.
Warren, D. L.; Guth, D. J.; Last, J. A. (1986) Synergistic interaction of ozone and respirable aerosols on rat
lungs: II. synergy between ammonium sulfate aerosol and various concentrations of ozone. Toxicol.
Appl. Pharmacol. 84: 470-479.
Warren, D. L.; Hyde, D. M.; Last, J. A. (1988) Synergistic interaction of ozone and respirable aerosols on rat
lungs. IV. Protection by quenchers of reactive oxygen species. Toxicology 53: 113-133.
Watanabe, S.; Frank, R.; Yokoyama, E. (1973) Acute effects of ozone on lungs of cats: I. functional. Am. Rev.
Respir. Dis. 108: 1141-1151.
6-202
-------
Watkinson, W. P.; Gordon, C. J. (1993) Caveats regarding the use of the laboratory rat as a model for acute
toxicological studies: modulation of the toxic response via physiological and behavioral mechanisms.
Toxicology 81: 15-31.
Watkinson, W. P.; Highfill, J. W.; Gordon, C. J. (1989) Modulating effect of body temperature on the toxic
response produced by the pesticide chlordimeform in rats. J. Toxicol. Environ. Health 28: 483-500.
Watkinson, W. P.; Aileru, A. A.; Dowd, S. M.; Doerfler, D. L.; Tepper, J. S.; Costa, D. L. (1993) Acute
effects of ozone on heart rate and body temperature in the unanesthetized, unrestrained rat maintained
at different ambient temperatures. Inhalation Toxicol. 5: 129-147.
Wegner, C. D. (1982) Characterization of dynamic respiratory mechanics by measuring pulmonary and
respiratory system impedances in adult bonnet monkeys (Macaco radiata): including the effects of
long-term exposure to low-level ozone [Ph.D. dissertation]. Davis, CA: University of California,
Graduate Division. Available from: University Microfilms International, Ann Arbor, MI; publication
no. 8227900.
Weibel, E. R. (1963) Morphometry of the human lung. New York, NY: Academic Press Inc.
Weibel, E. R. (1983) Is the lung built reasonably?: the 1983 J. Burns Amberson lecture. Am. Rev. Respir. Dis.
128: 752-760.
Weiss, B.; Ferin, J.; Merigan, W.; Stern, S.; Cox, C. (1981) Modification of rat operant behavior by ozone
exposure. Toxicol. Appl. Pharmacol. 58: 244-251.
Williams, S. J.; Charles, J. M.; Menzel, D. B. (1980) Ozone induced alterations in phenol red absorption from
the rat lung. Toxicol. Lett. 6: 213-219.
Wilson, D. W.; Plopper, C. G.; Dungworth, D. L. (1984) The response of the macaque tracheobronchial
epithelium to acute ozone injury: a quantitative ultrastructural and autoradiographic study.
Am. J. Pathol. 116: 193-206.
Witschi, H. (1988) Ozone, nitrogen dioxide and lung cancer: a review of some recent issues and problems.
Toxicology 48: 1-20.
Witschi, H. (1991) Effects of oxygen and ozone on mouse lung tumorigenesis. Exp. Lung Res. 17: 473-483.
Witschi, H.; Wilson, D. W.; Plopper, C. G. (1993a) Modulation of ./V-nitrosodiethylamine-induced hamster lung
tumors by ozone. Toxicology 77: 193-202.
Witschi, H. P.; Breider, M. A.; Schuller, H. M. (1993b) Failure of ozone and nitrogen dioxide to enhance lung
tumor development in hamsters. Cambridge, MA: Health Effects Institute; research report no. 60.
Wright, E. S.; White, D. M.; Brady, A. N.; Li, L. C.; D'Arcy, J. B.; Smiler, K. L. (1987) DNA synthesis in
pulmonary alveolar niacrophages and type II cells: effects of ozone exposure and treatment with
D-difluoromethylornithine. J. Toxicol. Environ. Health 21: 15-26.
Wright, E. S.; Kehrer, J. P.; White, D. M.; Smiler, K. L. (1988) Effects of chronic exposure to ozone on
collagen in rat lung. Toxicol. Appl. Pharmacol. 92: 445-452.
6-203
-------
Wright, E.; Gross, K.; Smiler, K. (1989) Continuous chronic exposure to ozone: biochemical, functional and
histopathologic effects in rat lung. Presented at: 82nd annual meeting of the Air & Waste
Management Association; June; Anaheim, CA. Pittsburgh, PA: Air & Waste Management
Association; paper no. 89/12.2.
Wright, E. S.; White, D. M.; Smiler, K. L. (1990) Effects of chronic exposure to ozone on pulmonary lipids in
rats. Toxicology 64: 313-324.
Yanai, M.; Ohrui, T.; Aikawa, T.; Okayama, H.; Sekizawa, K.; Maeyama, K.; Sasaki, H.; Takishima, T. (1990)
Ozone increases susceptibility to antigen inhalation in allergic dogs. J. Appl. Physiol. 68: 2267-2273.
Yeadon, M.; Wilkinson, D.; Darley-Usniar, V.; O'leary, V. J.; Payne, A. N. (1992) Mechanisms contributing to
ozone-induced bronchial hyperreactivity in guinea-pigs. Pulm. Pharmacol. 5: 39-50.
Yokoyama, E.; Frank, R. (1972) Respiratory uptake of ozone in dogs. Arch. Environ. Health 25: 132-138.
Yokoyama, E.; Ichikawa, I. (1974) [Study on the biological effects of atmospheric pollutants]. In: Research report
for funds of the Environment Agency in 1974. Tokyo, Japan: The Institute of Public Health,
Department of Industrial Health; pp. 16-1—16-6.
Yokoyama, E.; Ichikawa, I.; Kawai, K. (1980) Does nitrogen dioxide modify the respiratory effects of ozone? In:
Lee, S. D., ed. Nitrogen oxides and their effects on health. Ann Arbor, MI: Ann Arbor Science
Publishers, Inc.; pp. 217-229.
Yokoyama, E.; Ichikawa, I.; Nambu, Z.; Kawai, K.; Kyono, Y. (1984) Respiratory effects of intermittent
exposure to ozone of rats. Environ. Res. 33: 271-283.
Yokoyama, E.; Nambu, Z.; Ichikawa, I.; Uchiyama, I.; Arakawa, H. (1987) Pulmonary response to exposure to
ozone of emphysematous rats. Environ. Res. 42: 114-120.
Yokoyama, E.; Goto, H.; Kawai, K.; Kyono, H. (1989a) Mechanical properties of rabbit lung with edema caused
by exposure to ozone. J. Environ. Pathol. Toxicol. Oncol. 9: 95-108.
Yokoyama, E.; Uchiyama, I.; Arito, H. (1989b) Extrapulmonary effects of low level ozone exposure.
In: Schneider, T.; Lee, S. D.; Wolters, G. J. R.; Grant, L. D., eds. Atmospheric ozone research
and its policy implications: proceedings of the 3rd U.S.-Dutch international symposium; May 1988;
Nijmegen, The Netherlands. Amsterdam, The Netherlands: Elsevier Science Publishers;
pp. 301-309. (Studies in environmental science 35).
Young, C.; Bhalla, D. K. (1992) Time course of permeability changes and PMN flux in rat trachea following
O3 exposure. Fundam. Appl. Toxicol. 18: 175-180.
Zelac, R. E.; Cromroy, H. L.; Bolch, W. E., Jr.; Dunavant, B. G.; Bevis, H. A. (1971a) Inhaled ozone as a
mutagen: I. chromosome aberrations induced in Chinese hamster lymphocytes. Environ. Res. 4: 262-
282.
Zelac, R. E.; Cromroy, H. L.; Bolch, W. E., Jr.; Dunavant, B. G.; Bevis, H. A. (1971b) Inhaled ozone as a
mutagen: II. effect on the frequency of chromosome aberrations observed in irradiated Chinese
hamsters. Environ. Res. 4: 325-342.
Zelikoff, J. T.; Kraemer, G.-L.; Vogel, M. C.; Schlesinger, R. B. (1991) Immunomodulating effects of ozone on
macrophage functions important for tumor surveillance and host defense. J. Toxicol. Environ. Health
34: 449-467.
6-204
-------
Zidenberg-Cherr, S.; Han, B.; Dubick, M. A.; Keen, C. L. (1991) Influence of dietary-induced copper and
manganese deficiency on ozone-induced changes in lung and liver antioxidant systems. Toxicol. Lett.
57: 81-90.
Zitnik, L. A.; Schwartz, L. W.; McQuillen, N. K.; Zee, Y. C.; Osebold, J. W. (1978) Pulmonary changes
induced by low-level ozone: morphological observations. J. Environ. Pathol. Toxicol. 1: 365-376.
6-205
-------
7
Human Health Effects of Ozone and
Related Photochemical Oxidants
7.1 Introduction
In the previous chapter, results of ozone (O.) studies in laboratory animals were
presented in order to understand the wide range of potential effects that might occur in exposed
human populations and to expand the understanding of the mechanisms of O. toxicity and the
basic exposure-response relationships for O.. The concept of quantitatively extrapolating
results from laboratory animals to humans is further explored in Chapter 8. Whenever
possible, however, risk assessment of pollutants should be based on direct evidence of their
health effects in human populations. Information on human health responses to O. can be
obtained through controlled human exposure studies on volunteer subjects or through field and
epidemiological studies of populations that are exposed to ambient air containing O..
Controlled human studies typically use fixed concentrations of O. under carefully regulated
environmental conditions, whereas realistic O. exposure conditions occur in field and
epidemiology studies, but are more variable. The primary purpose of all these studies,
however, is to obtain exposure-response data for O.. This chapter will summarize the results
of controlled human, field, and epidemiologic studies on the health effects of exposure to
O. that have been published or accepted for publication in the peer-reviewed literature.
Further evaluation of the most important key information from this chapter, as it relates to the
rest of the document, will be provided in Chapter 9, where the overall database on O. health
effects is integrated and summarized.
Most of the scientific information selected for review and comment in this chapter
comes from the literature published since completion of the previous O. criteria document
(U.S. Environmental Protection Agency, 1986). Some of these newer studies were briefly
reviewed in the supplement to that document (U.S. Environmental Protection Agency, 1992),
but more thorough evaluation of these studies is included here. In order to give a broader
overview of the known human health effects of O., the older literature is summarized, and
specific studies whose data were judged to be significant because of their usefulness in deriving
the current National Ambient Air Quality Standards (NAAQS) are discussed briefly. The
reader is, however, referred to the more extensive discussion of these key studies in the
previous document. Other, older studies also are briefly discussed in this chapter if they are
(1) open to reinterpretation because of newer data or (2) potentially useful as criteria for the
O. NAAQS reevaluation. To further aid in the development of this chapter, summary tables of
the relevant O. literature are included for each of the major subsections. In summarizing the
7-1
-------
human health effects literature, changes from control are described if they were statistically
significant at a probability (p) value less than 0.05. A specific p value is provided, however, if
it aids understanding of the data, particularly trends toward significance, or if major effects
need to be emphasized. Where appropriate, critique of a statistical procedure also is
mentioned.
7.2 Controlled Human Exposure Studies
7.2.1 Pulmonary Function Effects of One- to Three-Hour Ozone Exposures
7.2.1.1 Healthy Subjects
Introduction
The pulmonary responses observed in healthy human subjects exposed to ambient
O. concentrations consist of decreased inspiratory capacity; mild bronchoconstriction; rapid,
shallow breathing pattern during exercise; and subjective symptoms of cough and pain on deep
inspiration. In addition, O. has been shown to result in airway hyperresponsiveness as
demonstrated by an increased physiological response to a nonspecific stimulus. The decrease
in inspiratory capacity results in a decrease in forced vital capacity (FVC) and total lung
capacity (TLC) and, in combination with bronchoconstriction, contributes to a decrease in the
forced expiratory volume in 1 s (FEV.). However, it is important to stress that in many of the
studies reporting the effects of ambient ranges of O. concentrations (i.e., <0.3 ppm), the
observed decrements in FEV., to a large extent, reflect decrements in FVC of a similar
magnitude (i.e., a decreased inspiratory capacity) and, to a lesser extent, increases in central
and peripheral airway resistance (R..).
The majority of controlled human studies have been concerned with the effects of
various O. concentrations in healthy subjects performing continuous exercise (CE) or
intermittent exercise (IE) for variable periods of time. These studies have two weaknesses: (1)
the failure to detect short-term effects with long-term consequences and (2) their use of small
numbers that are not generally representative of the general population. Controlled human
exposure studies of this type have provided the strongest and most quantifiable concentration-
response data on the health effects of O.. As a result of these studies, a large body of data
regarding the interaction of O. concentration (C), minute ventilation (V.), and duration of
exposure (T) is available. The most salient observations from these studies are
(1) O. concentration is more important than either V. or T in determining pulmonary responses
and (2) normal, healthy subjects exposed to O. concentrations DO. 12 ppm (the level of the
current NAAQS) develop significant reversible, transient decrements in pulmonary function if
V. or T are increased sufficiently. There is typically a large intersubject variability in
physiologic and symptomatic responses to O.; however, with most individuals these responses
tend to be reproducible. The relationship among response variables such as spirometry,
resistance measurements, symptoms, and nonspecific bronchial responsiveness is yet to be
fully determined, but the generally weak associations suggest that several response mechanisms
may be operant. In addition, a growing number of studies are beginning to provide insight into
the relationship between regional dosimetry (see Chapter 8), mechanisms of pulmonary
responses elicited by acute O. exposure, and tissue level events within the airways. This type
of information promises to provide further insight into the health effects relevance of
O.-induced pulmonary responses in determining which individuals are at greatest risk from
ambient O. exposure.
7-2
-------
In this section, the effects of acute (single 1- to 3-h) O. exposures on pulmonary
function in healthy subjects are examined by reviewing studies that investigate (1) the
O. exposure-response relationship; (2) intersubject variability, individual sensitivity, and the
association between responses; and (3) mechanisms of pulmonary function responses and the
relationship between tissue-level events and functional responses. Unless otherwise stated, the
term "significant" is used in this section to denote statistical significance at p < 0.05. Recent,
single O. exposure studies of greater than 3 h duration are reviewed in Section 7.2.2. These
single-exposure, longer duration studies are beginning to provide important insights into the C
x T x V. interaction related to a significant pulmonary response. Key studies of less than 3 h
duration that have contributed to the exposure-response database and other studies that have
contributed to a better understanding of O.-induced pulmonary responses in healthy individuals
are summarized in Table 7-1. Table 7-1 summarizes studies reviewed in the previous air
quality criteria document (U.S. Environmental Protection Agency, 1986), as well as studies
published since completion of this earlier document. Not reviewed in this section are studies
that examine changes in airway responsiveness induced by O. inhalation (see Section 7.2.3).
All of the studies discussed here used appropriate controls and therefore, for simplicity, the
text will not indicate for each study that subjects were also exposed under similar conditions to
filtered air (FA [reported at 0 ppm O.]).
The Ozone Concentration-Response Relationship
At-Rest Exposures. No new studies examining the acute effects of a single exposure
to O. concentrations below 1 ppm in resting humans have been published since the 1986 U.S.
Environmental Protection Agency (EPA) criteria document (U.S. Environmental Protection
Agency, 1986). Seven studies (Young et al., 1964; Bates et al., 1972; Silverman et al., 1976;
Folinsbee et al., 1978; Horvath et al., 1979; Kagawa and Tsuru, 1979; Konig et al., 1980)
examining 2-h, at-rest exposures were discussed in the 1986 EPA criteria document (U.S.
Environmental Protection Agency, 1986) involving 91 healthy subjects (74 males, 17 females)
exposed to O. concentrations ranging from 0.1 to 1.0 ppm. The lowest concentration at which
significant reductions in FVC and FEV. were reported was 0.5 ppm (Folinsbee et al., 1978;
Horvath et al., 1979). Reports of increases in R,. are inconsistent in resting human subjects
exposed to O. concentrations below 1.0 ppm.
Exposure with Exercise. Bates et al. (1972) and Hazucha et al. (1973) were the first
investigators to examine the effect on pulmonary function responses of increasing ventilation
via exercise during O. inhalation. The IE protocol used consisted of the subjects alternating
rest and light exercise on a cycle ergometer at a rate sufficient to double resting V. for 15 min
during a period of 2 h.
Hazucha et al. (1973) observed significant decreases in forced expiratory endpoints
at 0.37 ppm O. (p < 0.05) and 0.75 ppm O. (p < 0.001), with subjects exposed to 0.75 ppm
having the greatest decrements. After exposures, all subjects complained to varying degrees of
substernal soreness, chest tightness, and cough. The important findings from these early
studies were that the exercise-induced increase in V. accentuated the observed pulmonary
response at any given O. concentration and lowered the minimum O. concentration at which
significant pulmonary responses were observed. Subsequently, the interaction between
O. concentration and V. was examined by using similar IE protocols in which both
7-3
-------
Table 7-1. Controlled Exposure of Healthy Human Subjects to Ozonfe
Ozone
Concentration1'
ppm
Qg/m
_ Exposure
Duration and Exposure
Activity Conditions
Number
and
Gender of
Subjects
Subject
Characteristics
Observed Effect(s)
O
o
B
CD
0
Reference H
Healthy Adult Subjects at Rest CT.
0.25
0.50
0.75
0.37
0.50
0.75
0.50
Healthy
0.08
0.10
0.12
0.14
0.16
0.10
0.15
0.20
0.25
0.12
0.18
0.24
0.12
0.18
0.24
0.30
0.40
490
980
1,470
726
980
1,470
980
2h NA
2h NA
2h NA
8M
5F
20 M
8F
40 M
Young, healthy
adults, 21 to
22 years old
Young, healthy
adults, 19 to
29 years old
Young, healthy
adults, 18 to
28 years old
FVC decreased with 0.50- and 0.75-ppm O3 exposure compared
with FA; 4% nonsignificant decrease in mean VO^ following
0.75 ppm O3 compared with FA exposure.
Decrease in FEV,, VS,5VC, and V^^vc with 0.75 ppm
O3 exposure compared to FA.
Decrease in forced expiratory volume and flow.
Exercising Adult Subjects
157
196
235
274
314
196
294
392
490
235
353
470
235
353
470
588
784
2 h IE Tdb = 32 DC
(4 x 15min RH = 38%
at VE =
68 L/min)
2 h IE Tdb = 22 DC
(4 x 14min RH = 50%
treadmill at
mean VE =
70.2 L/min)
1 h competitive Tdb = 23 to
simulation 26 DC
exposures at RH = 45 to 60%
mean VE =
87 L/min
2.5 h IE Tdb = 22 DC
(4 x 15 min RH = 40%
treadmill
exercise [ VE =
65 L/min])
24 M
20 M
10 M
20 M
22 M
20 M
21 M
20 M
29 M
Young, healthy
adults, 18 to
33 years old
Young,
healthy NS,
25.3 ±4.1 (SD)
years old
10 highly trained
competitive
cyclists, 19 to
29 years old
Young, healthy
adults, 18 to
30 years old
No significant changes in pulmonary function measurements.
FVC, FEV,, FEF^j,,, SG.W, 1C, and TLC all decreased with
(1) increasing O3 concentration, and (2) increasing time of
exposure; threshold for response was above 0.10 ppm but below
0.15 ppm O3.
Decrease in FVC and FEV, for 0.18- and 0.24-ppm O3 exposure
compared with FA exposure; decrease in exercise time for
subjects unable to complete the competitive simulation at 0.18 and
0.24 ppm O3, respectively.
Significant decrease in FVC, FEV, , and FEF^yj,, at 0. 12 ppm
O3; decrease in VT and increase in f and SR,W at 0.24 ppm O3.
Horvath et al. (1979) 3
V>
g.
Silverman et al. (1976) cjj'
CD,
Folinsbee et al. (1978) &>
<
5!
CD
Linn et al. (1986) n>
1.
CD
p.
Kulle et al. (1985)
Schelegle and Adams (1986)
McDonnell et al. (1983)
-------
7-5
-------
Table 7-1 (cont'd). Controlled Exposure of Healthy Human Subjects to Ozonfe
Ozone
Concentration1'
ppm
Healthy
0.12
0.18
0.24
0.30
0.40
0.12
0.18
0.24
0.30
0.40
0.12
0.18
0.24
0.30
0.40
0.12
0.20
Dg/m
Exposure
Duration and Exposure
Activity Conditions
Number
and
Gender of
Subjects
Subject
Characteristics'
Observed Effect(s)
Reference
Exercising Adult Subjects (cont'd)
235
353
470
588
784
235
353
470
588
784
235
353
470
588
784
235
392
2 x 2.5 h IE Tdb = 22 DC
(4 x 15 min RH = 40%
treadmill
exercise
[VE =
35 L/min/nf
BSA]).
Exposure
separated by
48 ± 30 days
and 301
± 77 days
2 x 2.5 h IE Tdb = 22 DC
(4 x 15 min RH = 40%
treadmill
exercise
[VE =
35 L/min/m2
BSA])
2.5 h IE Tdb = 22 DC
(4 x 15 min RH = 40%
treadmill
exercise
[VE =
25 L/min/m2
BSA])
1 h CE . Tdb = 31 DC
(mean VE =
89 L/min)
8 M
8M
5 M
5M
6M
290 M
17 WM/15 BM/15 WE/ 15BF
15 WM/15 BM/15 WE/ 16BF
15 WM/17 BM/17 WE/ 15BF
16 WM/15 BM/17 WE/ 16BF
15 WM/15 BM/15 WE/ 15BF
15 WM/15 BM/15 WE/ 15BF
15 M
2F
Young, healthy
adults, 18 to
30 years old
Young, healthy
adults, 18 to
32 years old
Young, healthy
whites and
blacks, 18 to
35 years old
Highly trained
competitive
cyclists, 19 to
30 years old
Pulmonary function variables SR,W and VE were not
significantly different in repeat exposures, indicating that
the response to 0.18 ppm O3 or higher is reproducible.
O3 concentration and age predicted FEV, decrements; it was
concluded that age is a significant predictor of response
(older subjects being less responsive to O3).
Decreases in FEV, for all levels of O3 as compared with
FA; increase in SRaw with 0. 18 ppm O3 and greater
compared with FA; black men and women had larger FEV,
decrements than white men, and black men had larger FEV,
decrements than white women.
Decrease in VBmx, YOam}1, VTn]ax, work load, ride time,
FVC, and FEV, with 0.20 ppm O3 exposure during
maximal exercise conditions, but not significant with
0.12 ppm O3 exposure, as compared to FA exposure.
McDonnell et al. (1985b)
McDonnell et al. (1993)
Seal et al. (1993)
Gong et al. (1986)
-------
Table 7-1 (cont'd). Controlled Exposure of Healthy Human Subjects to Ozonfe
Ozone
Concentration1' Fvnnurirp
ppm
Duration and
Dg/m Activity
Number
Exposure and Subject
Conditions Gender of Subjects Characteristics
Observed Efffct(s) Reference
Healthy Exercising Adult Subjects (cont'd)
0.16
0.24
0.32
0.20
0.20
0.30
0.20
0.35
0.21
0.21
0.25
0.25
314
470
627
392
392
588
392
686
412
412
490
490
1 h CE (mean Tdb
VE = 57 L/min) RH
4 h IE Tdb
(4 x 50 min RH
cycle ergometry or
treadmill running
[ VE = 40 L/min])
30 to 80 min CE Tdb
cycle ergometry RH
( VE = 33 or
66 L/min)
1 h CE or Tdb
competitive RH
simulation (mean
VE = 77.5 L/min)
lhCE(75% Tdb
V0imx) RH
1 h CE cycle Tdb
ergometry RH
(mean VE =
80 L/min)
1 h CE Tdb
(mean VE = RH
63 L/min)
1 hCE
cycle ergometer
(VE =
30 L/min/m2 BSA)
= 32 DC
= 42 to 46%
= 20 DC
= 50%
= 20 to 24 DC
= 40 to 60%
= 23 to 26 DC
= 45 to 60%
= 19 to 21 DC
= 60 to 70%
= 22.5 DC
= 58.8%
= 20 DC
= 70%
NA
42 M
8F
11 M
3F(FA
exposure);
9M
3F(03
exposure)
8M
10 M
6M
1 F
14 M
1 F
19 M
7F
5M
2F
Competitive
bicyclists, 26.4 +
6.9 (SD) years old
Adult, healthy
NS, 19 to
41 years
old
Aerobically fit,
22 to 46 years old
Well-trained
distance runners,
19 to 31 years
old
Well-trained
cyclists, 18 to
27 years old
Highly fit
endurance
cyclists, 16 to
34 years old
Active
nonathletes
Young, healthy
NS, 22 to
30 years old
Small decrements in FEV, at 0.16 ppm with larger decrements at Avol et al. (1984)
0.24 ppm O3.
Decrease in FVC, FEV,, VT, and SRJW and increase in f with O3 exposure Aris et al. (1993a)
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.
O3 effective dose was significantly related to pulmonary function Adams et al. (1981)
decrements and exercise ventilatory pattern changes; multiple regression
analysis showed that O3 concentration accounted for the majority of the
pulmonary function variance.
Decrease in FVC, FEV,, and FEFS_75,, with 0.20 and 0.35 ppm Adams and Schelegle (1983)
O3 exposure compared with FA; VT decreased and f increased with
continuous 50-min O3 exposures; three subjects unable to complete
continuous and competitive protocols at 0.35 ppm O3.
Decrease in FVC, FEV,, FEF^j,, ., and MVV with 0.21 ppm Folinsbee et al. (1984)
O3 compared with FA exposure.
No significant differences in the effects of albuterol on metabolic data, Gong et al. (1988)
pulmonary function, airway reactivity, and exercise performance vs.
placebo; decrease in VElmx during O3 conditions.
FVC, FEV,, and MVV all decreased with 0.25 ppm O3 exposure Folinsbee et al. (1986)
compared with FA.
12.4% decrease in FEV,. Significant elevation of substance P and 8-epi- Hazbun et al. (1993)
PGFj] in segmental airway washing, but not bronchoalveolar lavage fluid.
-------
Table 7-1 (cont'd). Controlled Exposure of Healthy Human Subjects to Ozonfe
CO
Ozone
Concentration'"
ppm
Healthy
0.30
0.30
0.35
0.37
0.50
0.75
0.40
0.40
0.40
Dg/m
Exposure
Duration and Exposure
Activity Conditions
Number
and
Gender of Subject
Subjects Characteristics
OBserved Effect(s)
Reference
Exercising Adult Subjects (cont'd)
588
588
686
726
980
1,470
784
784
784
1 h CE cycle NA
ergometry
(mean VE =
60 L/min)
1 h CE cycle . Tdb = 21 to
ergometry ( VE = 25 DC
60 L/min) and 2 h RH = 45 to 60%
lE^cycle ergometry
( VE = 45 to
47 L/min)
1 h CE cycle Tdb = 21 to
ergometry (mean 25 DC
VE = 60 L/min) RH = 45 to 60%
2 h IE cycle
ergometry ( VE =
2.5 x rest)
2 h IE treadmill Tdb = 22 DC
exercise ( VE = RH = 40%
50 to 75 L/min)
1 h CE NA
treadmill exercise;
(VE =
20 L/min/m2 BSA)
2 h IE NA
(4x15 min heavy
treadmill exercise
[VE =
35 L/min/m2 BSA])
5 M Normal
12 M Moderately fit,
young and
healthy
14 M Moderately fit,
young, healthy
adults, 18 to
34 years old
20 M Young, healthy
8 F adults, 19 to
29 years old
8 M Young, healthy
NS, 18 to
27 years old
20 M Young, healthy
NS
1 1 M Young, healthy
NS, 18 to
35 years old
Decrease in FVC and FEV and increase in SR 1 h post-O exposure;
increase in percent PMNs at 1 , 6, and 24 h post-O3 exposure compared
with FA in first aliquot "bronchial" sample. PMNs peaked at 6 h post-
03 in "bronchial" sample. Percent PMNs elevated at 6 and 24 h post-03
in pooled aliquots.
Decrease in FEV, equivalent for all protocols.
Significant decreases in FVC and FEV, with O3 exposure compared to
FA exposure; FVC and FEV, decreases with O3 exposure were
attenuated significantly with indomethacin compared to no drug and
placebo; SR.W increases were not affected by indomethacin.
Decrease in FVC with 0.50 ppm and FEV, with 0.50 and 0.75 ppm O3
compared to FA; decrease in VS!fvc with 0-37 and 0.75 ppm and VJO,5VC
with 0.37, 0.50, and 0.75 ppm O3 exposure compared to FA.
Decreases in FVC, FEV,, VT, and TLC and increases in SR,W and f with
O3 exposure compared with FA. Atropine pretreatment abolished O3-
induced increase in SR.W and attenuated FEV, and FEF^,,^ response.
VT fell by 25 % , and O3 uptake efficiency in the lower respiratory tract
fell by 9% during O3 exposure.
No correlation between pulmonary function and inflammatory endpoints
measured in BAL fluid obtained 18 h after exposure; increase in
percentage of PMNs, total protein, albumin, IgG, and neutrophil
elastase; decrease in percentage of macrophages with O3 exposure
compared to FA exposure.
Schelegle et al. (1991)
McKittrick and Adams
(1995)
Schelegle et al. (1987)
Silverman et al. (1976)
Beckett et al. (1985)
Gerrity et al. (1994)
Koren et al. (1989a)
-------
Table 7-1 (cont'd). Controlled Exposure of Healthy Human Subjects to Ozonfe
VJ
Ozone
Concentration1' Fvnnurirp
ppm
Duration and
Dg/m Activity
Exposure
Conditions
Number
and Subject
Gender of Subjects Characteristics
Observed Effect(s)
Reference
Healthy Exercising Adult Subjects (cont'd)
0.40
0.40
0.40
0.40
0.60
0.50
0.75
784 2 h IE Tdb
(4 x 15 min heavy RH
treadmill exercise
[VE =
35 L/min/m2 BSA])
784 2 h IE Tdb
(4 x 15 min RH
bicycle ergometry
[VE =
30 L/min/m2 BSA])
784 1 h CE Tdb
(treadmill exercise; RH
20 L/min/m2 BSA)
784 2 h IE Tdb
1,176 (4 x 15 min RH
cycle ergometry
at 100 W for males
and 83 W for
females)
980 2 h IE Tdb
(4 x 15 min RH
treadmill exercise;
VE = 40 L/min)
1,470 2hIE
(4x15 min light
[50 W] cycle
ergometry)
= 22 DC
= 40%
= 22 DC
= 50%
= 22 DC
= 40%
= 71.5 DC
= 55%
= 21 DC
= 40%
NA
10 M Young, healthy
NS, 18 to
35 years old
13 M NS, 18 to
31 years old
22 M Young, healthy
NS, 18 to
35 years old
7 M Healthy
3 F NS, 23 to
41 years old
18 M Healthy,
young adults,
20 to 30 years old
13 M 4 light S,
9 NS, 19 to
30 years old
PMN, PGE,, and IL-6 were higher in BAL fluid obtained 1 h post-O3
exposure than 18 h; fibronectin and urokinase-type plasminogen activator
were higher 18 h post-O3 exposure than 1 h.
Indomethacin pretreatment and O3 exposure resulted in a significantly smaller
decrease in FVC and FEV, than O3 exposure alone; airway
hyperresponsiveness was not significantly affected by indomethacin
pretreatment.
Significant decreases in FVC, FEV,, FEV, /FVC, and FEF^^,,. The
half- width of an expired aerosol bolus was significantly increased, suggesting
an ozone-induced change in small airway function.
Increase in airway responsiveness to methacholine challenge, in mean
percentage of neutrophils, and in PGEj, TXB,, and PGE, concentrations
measured in BAL fluid 3 h after 0.40- and 0.60-ppm O3 exposure compared
with FA exposure.
Decrease in VC, VT, and maximal transpulmonary pressure, and increase
in SRaw and f with O3 exposure compared to FA exposure; lidocaine inhalation
partially reversed the decrease in VC.
Koren et al. (1991)
Ying et al. (1990)
Keefe et al. (1991)
Seltzer etal. (1986)
Hazucha et al. (1989)
Decrease in FVC, FEV,, ERV, 1C, and FEF^ after 1 h exposure to Folinsbee et al.
0.75 ppm O3; decrease in VOamx, V,.,,,,^, VEmx, maximal workload, and heart(1977)
rate following 0.75-ppm O3 exposure compared with FA.
"See Appendix A for abbreviations and acronyms.
''Grouped by rest and exercise; within groups listed from lowest to highest O3 concentration.
-------
Silverman et al. (1976) and Folinsbee et al. (1975) exposed a group of 20 males and
8 females to 0.37, 0.50, or 0.75 ppm for 2 h while resting or exercising intermittently. The IE
protocol used alternated 15 min of rest with 15 min of exercise, sufficient to increase the V.
value at rest by a factor of 2.5. The submaximal exercise responses of the subjects were tested
postexposure using a three-stage cycle ergometer test, with loads adjusted to 45, 60, and 75%
of maximum oxygen uptake (VO....) (Folinsbee et al., 1975). Pulmonary function responses
were related to the total inhaled dose or the "effective dose" of O. calculated as the product of
C x T x V.. Neither submaximal exercise oxygen uptake (VO.) nor V. were affected
significantly by any level of O. exposure; however, a significant increase in respiratory
frequency (f) and a significant decrease in tidal volume (V.) at the 75% VO.... workload were
observed. The relationship between the effective dose of O. and the mean percent change in
selected measures of lung function was analyzed using linear regression. Forced vital
capacity, maximum expiratory flow at 25 and 50% of FVC (V and V ,
respectively), and FEV. were found to have a significant linear correlation with the effective
dose. The description of the relationship between O. pulmonary function decrements and
effective dose was apparently improved by the use of a second-order polynomial model in
which effective dose was used as the independent variable.
Although the investigations of Silverman et al. (1976) and others (Bates et al.,
1972; Hackney et al., 1975; Hazucha et al., 1973) clearly demonstrate the potentiating effects
of exercise on O. responses, the level of exercise used in these studies was low, requiring
increases in V. of only 2 to 2.5 times resting, a level of exercise lower than that of a subject
walking at 5.5 km/h (DeLucia and Adams, 1977). In order to address this concern, DeLucia
and Adams (1977) exposed six healthy nonsmoking male subjects on 12 separate occasions to
FA and 0.15 and 0.30 ppm O. for 1 h, while at rest and while exercising continuously at
workloads that required 25, 45, and 65% of the subjects' VO They observed a significant
time-dependent increase in f during the 65% VO...., 0.30-ppm O. exposure, and, immediately
following this same exposure, there was a significant decrease in FEV. and forced expiratory
flow at 25 to 75% of FVC (FEF ).
These initial studies, which clearly demonstrated the potentiating effects of exercise
on human responses to acute O. exposure, provided the impetus for a series of studies (Adams
et al., 1981; Folinsbee et al., 1978; McDonnell et al., 1983; Kulle et al., 1985; Linn et al.,
1986) designed to define more precisely O. exposure-response relationships. These
investigations utilized both IE (Folinsbee et al., 1978; McDonnell et al., 1983) and CE (Adams
et al., 1981) of varying intensity. Folinsbee et al. (1978) exposed four groups of 10 subjects
each to FA, 0.1, 0.3, and 0.5 ppm O. for 2 h. One group was exposed while at rest, and the
other three groups were exposed while performing IE at levels requiring a ventilation of 30,
50, or 70 L/min. These combinations of ventilation and O. concentration (C X T X V.)
resulted in a range of total inhaled effective dose of 0.00 to 4.41 mg O.. Adams et al. (1981)
exposed eight trained male subjects to FA, 0.2, 0.3, and 0.4 ppm O. while they exercised
continuously at two different workloads (35 and 62% of VO ) for durations ranging from 30
to 80 min. Each subject completed all 18 protocols with at least 3 days between each. The
findings from these two studies confirmed that significant pulmonary responses occurred at 0.3
ppm when subjects exercised at moderately heavy workloads. It was further demonstrated, by
multiple regression analysis, that the O. effective dose was a better predictor of response than
O. concentration, V., or duration of exposure, alone. Multiple regression analysis also
revealed that the majority of variance for pulmonary function responses was accounted for by
7-10
-------
O. concentration, followed by V.. In the Adams et al. (1981) study, in which both workload
and duration of exposure were varied, duration of exposure was observed to be the poorest
predictor of response for all parameters analyzed. However, the minor impact of changes in
exposure duration could have been an artifact of the limited combinations of ventilation and
durations of exposure used by these investigators.
McDonnell et al. (1983) conducted a study with the primary purpose of discerning
the lowest concentration of O. at which group mean decrements in pulmonary function occur in
heavily exercising healthy men. In order to determine a concentration-response relationship,
six groups of subjects (n = 20 to 29) were exposed to either an FA control or one of five
O. concentrations (0.12, 0.18, 0.24, 0.30, or 0.40 ppm) at a V. of 67 L/min and exposure
duration of 2.5 h (15-min rest, 15-min exercise). These investigators observed small
significant changes in FVC, FEV., FEF , and cough at 0.12 ppm O. and concentration-
dependent responses in all variables measured (FVC, FEV., FEF , specific airway
resistance [SR..], f, V., and subjective symptoms) at O. concentrations >0.24 ppm.
Kulle et al. (1985) also conducted a similar study on healthy, nonsmoking men
performing IE at a V. of 70 L/min for an exposure duration of 2 h, (16-min rest, 14-min
exercise). Twenty subjects were exposed to an FA control or one of four O. concentrations
(0.10, 0.15, 0.20, or 0.25 ppm). These investigators observed a significant C X T interaction
at 0.15 ppm O. for FVC, FEV., FEF , and in all variables measured (FVC, FEV., FEF...
... , SR.., f, V., and subjective symptoms) at O. concentrations greater than 0.15 ppm.
Linn et al. (1986) exposed 24 healthy, well-conditioned male subjects (18 to
33 years of age) for 2 h to 0.00, 0.08, 0.10, 0.12, 0.14, or 0.16 ppm O., using an IE protocol
(15-min rest, 15-min exercise; V. = 68 L/min) combined with an ambient heat stress (32 DC
and 38% relative humidity [RH]). They observed no statistically significant changes in forced
expiratory endpoints and symptoms after exposure to O. concentrations from 0.08 to
0.14 ppm. These authors observed a small (D2.3%) but significant (p < 0.05) reduction in
FEV., which was not associated with symptoms of respiratory discomfort, following the 2-h
0.16-ppm O. exposure.
More recently, Seal et al. (1993) examined whether gender or race differences exist
in responsiveness to O.. The authors exposed 372 white and black, males and females
(n > 90 in each gender-race group) once for 2.33 h to 0.0, 0.12, 0.18, 0.24, 0.30, or
0.40 ppm O. using an IE protocol (15-min rest, 15-min exercise; V. = 25 L/min/m' body
surface area [BSA]). Statistical analysis (nonparametric two-factor analysis of variance) of the
percent changes from baseline for FEV., SR.., and cough responses demonstrated no
significant differences in responsiveness to O. between the gender-race groups studied.
Changes in FEV., SR.., and cough were first noted at 0.12, 0.18, and 0.18 ppm O.,
respectively, for the group as a whole. It is difficult to compare the results from this study
with other studies that have examined the O. concentration-response relationship in healthy
adult males because the authors did not present a separate analysis of male responses. For
further evaluation of the influence of gender and race on O. responsiveness, see
Section 7.2.1.3.
The observation of significant decrements in pulmonary function in heavily
exercising healthy subjects at O. concentrations of 0.2 ppm and lower has been confirmed by
numerous investigators (Adams and Schelegle, 1983; Avol et al., 1984; Folinsbee et al., 1984;
Gong et al., 1986) who utilized 1-h continuous heavy exercise exposure protocols. Adams and
Schelegle (1983) and Folinsbee et al. (1984) observed significant decrements in FVC and FEV.
7-11
-------
in well-trained subjects exposed to 0.2 ppm O. while exercising with a V. of approximately 80
L/min. Avol et al. (1984) observed small but significant decrements in FVC and FEV. in a
group of 50 competitive cyclists (42 males, 8 females) exposed to 0.16 ppm O. while
exercising with a V. of 57 L/min in combination with added heat stress (32 DC). Similarly,
Gong et al. (1986) observed modest but significant decrements in FVC and FEV. in a group of
17 top-caliber endurance cyclists exposed to 0.12 ppm O. while exercising at approximately
70% of their VO.... (mean V. = 89 L/min) with an added heat stress (32 DC). In addition to
the above studies that used continuous exercise, Schelegle and Adams (1986) observed
significant reductions in FVC and FEV. and increased symptoms of respiratory discomfort
following exposure to 0.18 but not 0.12 ppm O. in a group of competitive endurance athletes
exposed while performing a competitive simulation consisting of a 30-min warm-up followed
by a 30-min competitive bout (mean V. over entire protocol = 87 L/min).
The studies reviewed above demonstrate that in healthy young adults performing
moderate to severe IE and CE of 1 to 3 h duration, an O. concentration of 0.12 to 0.18 ppm is
required to elicit statistically significant decrements in pulmonary function and subjective
respiratory symptoms.
Retrospective analysis by Hazucha (1987) confirmed the previously reported
(Adams et al., 1981; Folinsbee et al., 1978) dominant role that O. concentration plays in
determining O.-induced responses. Hazucha (1987) analyzed data from studies that utilized IE
protocols of 2 h in duration. While controlling for ventilation, this investigator found that the
data best fit a model that was a quadratic function of O. concentration. Based on this analysis,
Hazucha (1987) also concluded that an O. concentration below which no pulmonary function
response would be elicited could not be defined.
The studies reviewed in this subsection used different patterns (i.e., CE or IE) of
exercise during their exposure protocols. An important question to ask is to what extent are
the results of these studies comparable when total inhaled doses are the same but exercise
pattern differs. A recent study by McKittrick and Adams (1995) addresses this question.
These investigators exposed 12 aerobically trained men to 0.30 ppm O. (three protocols) and
FA (three protocols) on six occasions. These protocols consisted of a 1 h CE (O. and FA) and
two 2-h IE (2 x O. and 2 x FA) protocols delivered in random sequence separated by a
minimum of 3 days. Lung function FEV. decrements of 17.6, 17.0, and 17.9% were obtained
for the 1-h CE and the two 2-h IE, 0.3-ppm O. protocols, respectively. These values were
significantly different from the FA values, but were not significantly different from each other.
The O. CE protocols resulted in greater postexposure values for subjective symptoms than
obtained with either of the O. IE protocols. However, the overall symptom severity during the
last minute of exercise for the two IE protocols was not significantly different from the CE
postexposure value. McKittrick and Adams (1995) concluded that when the total inhaled dose
of O. is equivalent at a given O. concentration, there is no difference between pulmonary
function responses induced by CE and IE protocols of 2-h or less duration, although subjective
symptoms are reduced slightly during the last rest period of IE.
7-12
-------
Intersubject Variability, Individual Sensitivity, and the Association Between Responses
Bates et al. (1972) noted that variation in sensitivity and response was evident for
various symptoms and pulmonary functions assessed following O. exposure. This observation
of large intersubject variability in response to O. also has been reported by numerous other
investigators (Adams et al., 1981; Folinsbee et al., 1978; McDonnell et al., 1983; Kulle et al.,
1985) and is illustrated by data from Kulle et al. (1985) plotted in Figure 7-1. The description
of the factors that contribute to intersubject variability is important for the understanding of
individual responses, mechanisms of response, and health risks associated with acute
exposures. The effect of this large intersubject variability on the ability to predict individual
responsiveness to O. was recently demonstrated by McDonnell et al. (1993). These
investigators analyzed the data of 290 white male subjects (18 to 32 years of age) who inhaled
either 0.00, 0.12, 0.18, 0.24, 0.30, or 0.40 ppm O. for 2 h while performing an IE protocol
(V. = 35 L/min/m' BSA) to identify personal characteristics (i.e., age, height, baseline
pulmonary functions, presence of allergies, and past smoking history) that might predict
individual differences in FEV. response. Of the personal characteristics studied, only age
contributed significantly to intersubject responsiveness (younger subjects were more
responsive), accounting for 4% of the observed variance. Interestingly, O. concentration
accounted for only 31 % of the variance, clearly demonstrating the importance of as yet
undefined individual characteristics that determine responsiveness to O..
McDonnell et al. (1985b) examined the reproducibility of individual responses to
O. exposure in healthy human subjects exposed twice, with from 21 to 385 days separating
exposures (mean = 88 days). This investigation was conducted in order to determine whether
the observed intersubject variability is due primarily to real differences in O. responsiveness
among subjects, or whether it can be accounted for by other sources of variability. The
authors examined FVC, FEV., FEF , SR.., cough, shortness of breath (SB), pain on deep
inspiration (PDI), V., and f responses induced by O. exposure to concentrations ranging from
0.12 to 0.40 ppm. Reproducibility was assessed using the intraclass correlation coefficient
(R), which incorporates into a single measure all the information contained in the correlation
coefficient, slope, and intercept obtained in linear regression analysis. Similar to the more
routinely used correlation coefficient, R is equal to one when two identical measurements
occur in the same subject; and the "worst" possible coefficient is equal to l/(n D 1), which
approaches zero for a large n. The ranking of most to least reproducible for the responses
studied was FVC (R = 0.92), FEV. (R = 0.91), FEF (R = 0.83), cough (R = 0.77), SB
(R = 0.60), SR.. (R = 0.54), PDI (R = 0.37), f (R = D0.20), and V. (R = D0.03). The
value of R was significantly different from zero for FVC, FEV., FEF , cough, SB, and
SR... McDonnell et al. (1985b) concluded that the reported large intersubject variability in
magnitude of response was due to large differences in the intrinsic responsiveness of individual
subjects to O. exposure. However, the factors that contribute to this large intersubject
variability remain undefined.
The examination of intersubject variability is complicated by a poor association
between the various O. responses. In their study investigating O. exposure-response
relationships, McDonnell et al. (1983) observed very low correlation between changes in SR..
and FVC (r = DO. 16) for 135 subjects exposed to O. concentrations ranging from 0.12 to 0.40
ppm for 2.5 h.
7-13
-------
22 -
20-
18-
16-
o
U-"
o
T3
-------
Mechanisms of Acute Pulmonary Responses
The pulmonary responses observed during and following acute exposure to O. at
concentrations between 0.1 and 0.5 ppm in normal healthy human subjects include decreases in
TLC, 1C, FVC, FEV., FEF , and V.; and increases in SR.., f, and airway responsiveness.
Ozone exposure also has been shown to result in the symptoms of cough, PDI, SB, throat
irritation, and wheezing. When viewed as a whole, changes in these specific parameters can
be categorized into four general responses: alterations in (1) lung volumes, (2) airway caliber,
(3) bronchomotor responsiveness, and (4) symptoms. The absence of consistent associations
among the various responses from individual to individual suggests that the functional
responses observed are the result of multiple interactions within the respiratory tract. These
interactions may be the result of O. action on the biochemical, anatomical, and physiological
systems of the respiratory tract. In turn, these factors determine O. dose distribution and the
resulting cellular and reflex responses.
Bates et al. (1972) observed that the most significant decrement in pulmonary
function was the reduction in the transpulmonary pressure at maximal inspiratory volume
without a concomitant decrease in static compliance. This would suggest an inhibition of
maximal inspiratory effort after O. exposure that may result in reductions in 1C. These authors
speculated that this inhibition is an early result of stimulation of rapidly adapting pulmonary
stretch receptors, or "irritant receptors", located in the major bronchi. Since 1972, when this
hypothesis was first published, numerous studies have examined the underlying mechanisms
leading to the functional responses observed in human subjects. These mechanistic studies
have used both animal models and human subjects. This discussion of mechanisms will focus
on studies that used human subjects but also will cover those animal studies that have direct
relevance to O.-induced functional responses.
The acute inhalation of ambient concentrations of O. by healthy human subjects has
been shown to result in a concentration-dependent increase in R.. (Folinsbee et al., 1978;
McDonnell et al., 1983; Kulle et al., 1985; Seal et al., 1993). This O.-induced increase in R..
has been shown to be poorly correlated with changes in forced expiratory endpoints
(McDonnell et al., 1983). Ozone-induced increases in R.. have a rapid onset (Beckett et al.,
1985) compared with the gradual development of decrements in forced expiratory endpoints
(Kulle et al., 1985). Ozone-induced increases in R.. also appear to be greater in atopic
subjects as a group (Kreit et al., 1989; McDonnell et al., 1987), although this does not appear
to be the case for O.-induced decrements in FVC and symptoms. Taken together, these
observations suggest that different pathways lead to O.-induced decrements in 1C and to O.-
induced increments in R...
Increases in R.. induced by O. have been shown to be blocked by atropine sulfate
pretreatment in human subjects (Beckett et al., 1985; Adams, 1986). This inhibition suggests
that the release of acetylcholine from parasympathetic postganglionic fibers that innervate
airway smooth muscle plays a role in this response. However, the observation that a 2-h, 0.6
ppm O. inhalation also results in a hyperresponsiveness to methacholine, a cholinergic agent
(Holtzman et al., 1979), suggests the possibility that acute O. exposure also can increase the
sensitivity of airway smooth muscle to acetylcholine independent of a reflex mechanism
involving cholinergic postganglionic nerves. The role that an increase in airway smooth
muscle sensitivity to the endogenous release of acetylcholine might play in O.-induced
increases in R.. has not been studied.
7-15
-------
Analyses by Colucci (1983) have suggested that the increase in R.. is not as large as
would be expected when O. exposure is combined with moderate to heavy exercise. However,
the observation that circulating epinephrine levels increase as a function of the relative
workload in exercising human subjects (Galbo, 1983; Warren and Dalton, 1983) suggests that
stimulation of airway smooth muscle beta-adrenoreceptors may counteract airway smooth
muscle contraction induced by O. exposure. The observations by Beckett et al. (1985) that the
beta-agonists abolish O.-induced bronchoconstriction is consistent with this possibility.
Another question to be addressed with regard to O.-induced increases in R.. is
where along the airway (central versus peripheral airways) is the increase in resistance
produced? Studies of acutely and subchronically exposed animals have demonstrated tissue
damage in the centriacinar region (Castleman et al., 1977; Mellick et al., 1977), as well as
increases in peripheral resistance and reactivity (Gertner et al., 1983a,b,c; Beckett et al.,
1988). Keefe et al. (1991) examined the possibility of an effect on small airways using an
inhaled aerosol bolus dispersion technique in 22 healthy, nonsmoking male subjects exposed to
0.4 ppm O. for 1 h using a CE protocol (V. = 20 L/min/m' BSA). The bolus dispersion
technique is not dependent on vital capacity maneuvers and compares the profile of a bolus of
small (0.5- to 1.0-Dm) aerosol particles injected into the inspired airstream (at a fixed lung
volume) with the profile of the bolus during expiration. Dispersion of the bolus during
expiration can be affected by increases in turbulence within the airway, the development of
asymmetries in ventilation due to unequal regional time constants within the lung, and an
increase in aerosol deposition in the small airways. Keefe et al. (1991) observed that
O. exposure in their subjects resulted in a significant increase in dispersion of an aerosol bolus
(without an increased aerosol deposition) that was not correlated with changes in SR... These
findings suggest that exposure to 0.4 ppm O. under the conditions of this experiment results in
changes in small airway function that are not detectable by more conventional techniques.
Ozone-induced alterations in ventilatory pattern have been observed in exercising
dogs (Lee et al., 1979) and humans (Adams et al., 1981; Folinsbee et al., 1978; McDonnell
et al., 1983; Kulle et al., 1985). In exercising humans, O. exposure has been shown to result
in a decrease in V. and an increase in f in the absence of any change in V.. A rapid, shallow
breathing pattern is consistent with the maintenance of an appropriate ventilation with a
reduced VT. Reduction of VT is probably related to the reduction of 1C and is anecdotally
related to reduction in breathing discomfort caused by PDI.
Lee et al. (1979), who produced a reversible vagotomy by cooling the vagus nerves
to 0 DC, abolished the rapid, shallow breathing induced by O. inhalation in conscious dogs.
More recently, Schelegle et al. (1993) have shown in anesthetized dogs exposed to O. that
cooling the cervical vagus nerves to 7 DC did not abolish the observed O.-induced rapid,
shallow breathing pattern and bronchoconstriction, but cooling the vagus nerves to 0 DC did
abolish both the rapid, shallow breathing and the bronchoconstriction. These findings suggest
that O. stimulates nonmyelinated C fiber afferents arising from the lung, whose conduction is
not totally blocked at 7 DC but is blocked totally at 0 DC. This conclusion is consistent with
the findings of Coleridge et al. (1993) that bronchial C fibers are the only receptors that are
stimulated directly during O. inhalation in anesthetized dogs. If similar bronchial C fibers
were stimulated or sensitized in humans exposed to O., this could explain the O. -induced
rapid, shallow breathing observed during exercise, as well as the subjective symptoms
associated with taking a deep inspiration.
7-16
-------
Hazucha et al. (1989) exposed 11 healthy normal volunteers to FA and 0.5 ppm
O. for 2 h while they were performing moderate IE. Ozone exposure induced a significant
decrement in FVC, which was associated with a marked fall in 1C without an increase in
residual volume. Spraying of the upper airway with lidocaine aerosol in these subjects was
immediately followed by return of FVC toward control values. Hazucha et al. (1989)
concluded that O. inhalation stimulates lidocaine-sensitive tracheal and laryngeal airway
receptors, which leads to an involuntary inhibition of full inspiration, a reduction in FVC, and
a concomitant decrease in maximal expiratory flow rates in humans.
The airway afferents blocked by lidocaine in the Hazucha et al. (1989) investigation
remain undefined. However, it seems likely that the lung afferents involved are the same ones
that result in O.-induced rapid, shallow breathing in dogs (i.e., bronchial C fibers). When
stimulated with exogenous chemicals in animal experiments, bronchial C fibers induce a reflex
apnea (Coleridge and Coleridge, 1986). In dogs, this reflex apnea involves the inhibition of
inspiratory neurons, expiratory neurons, and D- and D-motoneurons in the intercostal nerves
(Koepchen et al., 1977; Schmidt and Wellhoner, 1970). Such a reflex response in humans
would explain the reflex inhibition of maximal inspiration consequent to acute O. exposure.
Data consistent with an O.-induced stimulation of bronchial C fibers in human
subjects recently has been published by Hazbun et al. (1993). These investigators observed a
significant increase in substance P, the neurotransmitter released from the afferent endings of
bronchial C fiber during excitation, in segmental airway washings of seven (2 female/5 male)
healthy, nonsmoking subjects after a 1 h CE (V. = 30 L/min/m' BSA) exposure to 0.25 ppm
O.. Substance P was not elevated in bronchoalveolar lavage (BAL) fluid after air exposure. In
addition, the segmental airway substance P levels were significantly correlated (r' = 0.89;
p < 0.05) with an elevated airway concentration of 8-epi-prostaglandin F.n, a marker of
oxidative free radical reactions. These results are consistent with (1) an increased release of
substance P secondary to an increased discharge of bronchial C fibers induced by
O. inhalation, and (2) an O.-induced inhibition of neutral endopeptidase, the enzyme that
degrades substance P within the airways.
Lung C fibers have been shown to be stimulated by prostaglandin E. and other lung
autacoids (Coleridge et al., 1978, 1976). Interestingly, Schelegle et al. (1987), Eschenbacher
et al. (1989), and Ying et al. (1990) have shown that pretreatment with the cyclooxygenase
inhibitor indomethacin reduces and, in some cases, totally abolishes O.-induced pulmonary
function decrements in human subjects. Schelegle et al. (1987) examined whether O.-induced
pulmonary function decrements could be inhibited by the prostaglandin synthetase inhibitor
indomethacin in healthy human subjects. Fourteen college-age males completed six 1-h
exposure protocols consisting of no drug, placebo, and indomethacin pretreatments, with FA
and O. (0.35 ppm) exposure within each pretreatment. Pretreatments were delivered weekly in
random order in a double-blind fashion. Exposures consisted of 1 h exercise on a cycle
ergometer with work loads set to elicit a mean V. of 60 L/min. Statistical analysis revealed
significant effects for FVC and FEV. across pretreatment, with no drug versus indomethacin
and placebo versus indomethacin comparisons being significant. These findings suggest that
cyclooxygenase products of arachidonic acid, which are reduced by indomethacin inhibition of
cyclooxygenase, play a role in the development of pulmonary function decrements. These and
similar findings by Eschenbacher et al. (1989) and Ying et al. (1990) suggest that the release of
some cyclooxygenase product consequent to O. inhalation plays a role in O.-induced
pulmonary function decrements. This idea is supported by the findings of Keren et al. (1991),
7-17
-------
who obtained a positive correlation between O.-induced pulmonary function decrements and
the level of prostaglandin E. (PGE.) in BAL fluid collected within 1 h after the end of
exposure in human subjects who varied greatly in O. responsiveness.
The release of cyclooxygenase products of arachidonic acid from injured airway
epithelium can thus be viewed as a link in a cascade of events, which begins with the initial
reaction of O. with the tissues and ends with the observed pulmonary function responses. The
apparent components of this chain of events include factors that influence (1) O. delivery to the
tissue (e.g., the inhaled concentration, breathing pattern, and airway geometry);
(2) O. reactions with components in airway surface liquid and epithelial cell membranes;
(3) local tissue responses, including injury and inflammation; and (4) stimulation of neural
afferents (bronchial C fibers) and the resulting reflex responses. It still is not understood how
each event in this cascade contributes to the pulmonary responses induced by acute
O. inhalation.
The influence that individual responsiveness has on this cascade of events has not
been determined; however, recent data suggest that individual O. responsiveness may feed
back and influence the distribution of O. dose within the lung. Gerrity et al. (1994) tested the
hypothesis that O.-induced rapid, shallow breathing helps to limit the dose of O. reaching the
lower respiratory tract. They found that the degree of O.-induced rapid, shallow breathing
(25 % decrease in V.) was significantly correlated with a decrease in O. uptake efficiency of
the lower respiratory tract. This observation may explain the recent data of Schelegle et al.
(1991) and Aris et al. (1993a) that suggest individual responsiveness to O. as measured by
FEV. decrements may be negatively correlated with the number of neutrophils (PMNs) present
in BAL samples. However, the interrelationship among the responsiveness to O., the
distribution of dose within the airway, and resulting airway inflammation is still poorly
understood.
7.2.1.2 Subjects with Preexisting Disease
Introduction
Ten studies (Konig et al., 1980; Linn et al., 1982a; Koenig et al., 1985; Linn et al.,
1978, 1983a; Solic et al., 1982; Kehrl et al., 1985; Superko et al., 1984; Silverman, 1979;
Kulle et al., 1984) examining the pulmonary responses to acute O. exposures of less than
3 h in patients with preexisting disease were discussed in the 1986 criteria document (U.S.
Environmental Protection Agency, 1986). This section examines the effects of O. exposure on
pulmonary function in subjects with preexisting disease by reviewing O. exposure studies that
utilized subjects with (1) chronic obstructive pulmonary disease (COPD), (2) asthma,
(3) allergic rhinitis, and (4) ischemic heart disease. Because of their important health
implications, all of the available studies are reviewed and summarized in Table 7-2. Unless
otherwise stated, the term "significant" is used to denote statistical significance at p < 0.05.
Subjects with Chronic Obstructive Pulmonary Disease
In five of the studies cited above, the O.-induced pulmonary function responses of
patients with mild to moderate COPD were examined (Konig et al., 1980; Linn et al., 1982a,
1983a; Solic et al., 1982; Kehrl et al., 1985). No significant changes in pulmonary function or
symptoms were reported in any of the studies of the effects of O. in patients with COPD. Four
of these studies (Linn et al., 1982a, 1983a; Solic et al., 1982; Kehrl et al., 1985) examined the
7-18
-------
effects of O. concentrations between 0.1 and 0.3 ppm O. in 66 mild to moderate COPD
patients using mild IE exposure protocols of 1 to 2 h duration. The total
7-19
-------
Table 7-2. Ozone Exposure in Subjects with Preexisting Diseas§
Ozone
Concentration" p.vnrvairp
ppm Dg/
Subjects with
0.12 236
0.18 353
0.25 490
0.20 392
0.30 588
0.41 804
Subjects with
0.20 392
0.30 588
Duration and
m Activity
Exposure
Condition
Number
and Subject
Gender of Subjects Characteristics
Observed Effect(s)
Reference
Chronic Obstructive Pulmonary Disease
1 h IE (2 x
15 min light
bicycle ergometry)
1 h IE (2 x
15 min light
bicycle ergometry)
2 h IE (4 x
7.5 min light
treadmill running)
2 h IE (4 x
7.5 min light
treadmill running)
3 h daily (1 x
15 min light
bicycle ergometry
during each
exposure) for
5 days
Heart Disease
40 min CE
treadmill walking
Tdb = 25 DC
RH = 50%
Tdb = 25 DC
RH = 50%
Tdb = 22 DC
RH = 40%
Tdb = 22 DC
RH = 40%
Tdb = 22 DC
RH = 50%
NA
18 M, 7 F 8 smokers,
14 ex-smokers, 3
nonsmokers;
FEV,/FVC = 32
to 66%
15 M, 13 F 15 smokers,
1 1 ex-smokers, 2
nonsmokers;
FEV,/FVC = 36
to 75%
13 M 8 smokers,
4 ex-smokers,
1 nonsmoker;
productive
cough;
FEV,/FVC = 46
to 70%
13 M 9 smokers,
4 nonsmokers;
FEV,/FVC = 37
to 65%
17 M, 3 F All smokers;
productive
cough;
FEV,/FVC = 56
to 82% and/or
FEV3/FVC = 75
to 93%
6 M Coronary heart
disease with
angina pectoris
threshold
No significant changes in
small significant decrease
pulmonary function measurements;
in arterial O2 saturation.
No significant changes in pulmonary function measurements;
no significant change in arterial O, saturation.
No significant changes in
small significant decrease
No significant changes in
arterial O, saturation.
pulmonary function measurements;
in arterial O2 saturation.
pulmonary function measurements or
Decrease in FVC and FEV3 with 0.41 ppm O3 compared with
FA exposure.
No significant changes in pulmonary function measurements,
exercise ventilatory pattern, oxygen uptake, or cardiovascular
parameters.
Linn et al. (1982a)
Linn et al. (1983a)
Solic et al. (1982)
Kehrl et al. (1985)
Kulle et al. (1984)
Superko et al. (1984)
-------
Table 7-2 (cont'd). Ozone Exposure in Subjects with Preexisting Diseas6
Ozone
Concentration" F.vnns,,rP
ppm
Subjects
0.18
0.50
a*
with
353
980
Duration and
'm Activity
Allergic Rhinitis
2hIE
(4 x 15 min)
4 h rest
Number
Exposure and Subject
Conditions Gender of Subjects Characteristics 3 Observed Effect(s)
NA 26 M History of Increased respiratory symptoms, SR.W, and reactivity to histamine
allergic rhinitis with O3 exposure and decreased FVC, FEV, , and FEF,j_7J,, with O3
exposure compared to FA.
Tdb = 20 to 24 DC 6 M, 6 F History of Increase in upper and lower respiratory symptom scores, cell influx,
RH = 40 to 48 % seasonal allergic epithelial cells with O3 exposure compared to FA; no effect on acute
rhinitis; acute allergic response to nasal antigen challenge between O3 and FA
response to nasal exposure.
challenge with
antigen
Reference
McDonnell et al.
(1987)
Bascom et al. (1990)
Adult Subjects with Asthma
0.10
0.25
0.40
0.12
0.12
196
490
784
236
236
1 h light IE (2
x 15 min on
treadmill, VE
= 27 L/min)
1 h rest
0.75 h IE
VE =
30 L/min
(15 min rest,
15 min
exercise,
15 min rest)
followed by
15 min
exercise
inhaling
0.10 ppm SO2
Tdb = 21D C 12 M, 9 F, Stable mild No significant differences in FEV, or FVC were observed for 0. 10
RH = 40% 19 to 40 years asthmatics with and 0.25 ppm O3-FA exposures or postexposure exercise challenge;
old FEV, > 70% 12 subjects exposed to 0.40 ppm O3 showed significant reduction in
and methacholine FEV, .
responsiveness
NA 7 M, 8 F Never smoked, Exposure to 0.12 ppm O3 did not affect pulmonary function.
mild stable Preexposure to 0.12 ppm O3 at rest did not affect the magnitude or
asthmatics with time course of exercise-induced bronchoconstriction.
exercise-induced
asthma
Tdb = 22D C 8 M, 5 F, Asthmatics Filtered air followed by SO2 and O3 alone did not cause significant
RH = 75% 12 to 18 years classified on changes in pulmonary function. Ozone followed by SO, resulted in
old basis of positive significant decrease in FEV, (8%) and Vlllarfo» (15%) and a
clinical history significant increase in RT (19%).
and methacholine
challenge.
Asymptomatic at
time of study.
Weymer et al. (1994)
Fernandes et al.
(1994)
Koenig et al. (1990)
-------
Table 7-2 (cont'd). Ozone Exposure in Subjects with Preexisting Diseas6
Ozone
Concentration1'
ppm Qg/m
Exposure
Duration and Exposure
Activity Conditions
Number
and
Gender of Subjects
Subject
Characteristics
3 Observed Effect(s)
Reference
Adult Subjects with Asthma (cont'd)
0.12 236
0.24 472
0.12 236
0.12 236
0.20 392
0.25 490
1.5hIE, Tdb = 22DC
VE= RH = 65%
25 L/min
6.5 h/day IE NA
(6 x 50 min)
(2 days of
exposure), VE
= 28 L/min
(asthmatic),
VE =
31 L/min
(healthy)
1 h rest NA
2hIE Tdb = 31 DC
(4 x 15 min at RH = 35%
2x rest
VE cycle
ergometry)
2 h rest NA
4M, 4F
(nonasthmatics) ;
18 to 35 years
old;
5M, 5F
(asthmatics);
18 to 41 years
old
8M, 7F
(nonasthmatics) ;
22 to 41 years
old;
13 M, 17 F
(asthmatics) ;
18 to 50 years
old
4M, 3F,
21 to 64 years
old
20 M, 2 F,
19 to 59 years
old
5 M, 12 F,
20 to 71 years
old
Physician-
diagnosed
asthma
confirmed with
methacholine
challenge test.
All nonsmokers
and asymptomatic
at time of study.
Nine were atopic.
Asthmatics
classified on
basis of positive
clinical history,
previous
physician
diagnosis, and
lowPD^. Mild
to severe
asthmatics.
Mild, stable
asthma
Physician
diagnosed asthma;
6 smokers,
9 ex-smokers,
7 nonsmokers
Nonsmoking
asthmatics
selected from a
clinical practice
No significant changes in pulmonary and nasal function
measurements in either asthmatics or nonasthmatics. Significant
increase in nasal lavage white cell count and epithelial cell following
O3 exposure in asthmatics only.
Significant increase in bronchial reactivity to methacholine in both
asthmatics and nonasthmatics. FEV, decreased 8.6% in asthmatics
and 1.7% in nonasthmatics, with difference not being significant.
Increase in bronchial responsiveness to allergen; no change in
baseline airway function.
No significant changes in pulmonary function measurements;
significant blood biochemical changes.
No significant changes in pulmonary function measurements.
McBride et al. (1994)
Linn et al. (1994)
Molfino et al. (1991)
Linn et al. (1978)
Silverman (1979)
-------
Table 7-2 (cont'd). Ozone Exposure in Subjects with Preexisting Diseas6
u>
Ozone
Concentration" p.vnrvairp
ppm
Duration and Exposure
Qg/m Activity Conditions
Number
and
Gender of Subjects
Subject
Characteristics
3 Observed Effect(s)
Reference
Adult Subjects with Asthma (cont'd)
0.40
784 2 h IE Tdb = 22 DC
(4 x 15 min RH = 50%
cycle
ergometry)
4 M, 5 F (normals),
19 to 31 years old;
4M, 5F
(asthmatics),
18 to 34 years old
Asthmatics as
diagnosed by a
physician;
history of chest
tightness and
wheezing
Decrease in FVC and 1C with O3 in asthmatics; increase in airway
responsiveness to methacholine in asthmatics with O3 and FA;
asthmatic subjects had significantly greater decreases in FEV, and
FEF2J_7J,5 with O3 exposure than did normal subjects.
Kreit et al. (1989)
Eschenbacher et al.
(1989)
Adolescent Subjects with Asthma
0.12
0.12
0.12
0.18
235 1 h rest Tdb = 22 DC
RHD75%
235 1 h IE Tdb = 22 DC
(2 x 15 min RHD75%
treadmill
walking at mean
VE = 32.5
L/min)
235 40 min IE NA
353 (1 x 10 min
treadmill
walking at mean
VE = 32.5
L/min)
4 M, 6 F (normals),
13 to 18 years old;
4M, 6F
(asthmatics),
1 1 to 18 years old
5 M, 8 F (normals),
12 to 17 years old;
9M, 3F
(asthmatics),
12 to 17 years old
4 M, 9 F (normals),
14 to 19 years old;
8M, 8F
(asthmatics),
12 to 19 years old
Asthmatics had
a history of atopic
extrinsic asthma
and exercise-
induced
bronchospasm
Asthmatics selected
from a clinical
practice and had
exercise- induced
bronchospasm
Asthmatics had
allergic asthma,
positive
responses to
methacholine, and
exercise- induced
bronchospasm
Decrease in FRC with O3 exposure in asthmatics; no consistent
significant changes in pulmonary functional parameters in either
group or between groups.
Decrease in maximal flow at 50% of FVC in asthmatics with
O3 exposure compared to FA; no significant changes with combined
O3-NO2 exposure.
Decrease in FEV, and increase in RT in normals and asthmatics with
0.12 and 0.18 ppm O3 exposure compared to FA; no consistent
differences between normals and asthmatics.
Koenig et al. (1985)
Koenig et al. (1988)
Koenig et al. (1987)
'See Appendix A for abbreviations and acronyms.
''Grouped by rest and exercise; within groups listed from lowest to highest O3 concentration.
-------
exercise time in all four of these studies was 30 min, with intensity being variable (exercise V.
approximately 14 to 28 L/min). Linn et al. (1982a) observed a small but significant reduction
in arterial oxygen saturation in 25 mild to moderate COPD patients at the end of the 0.12 ppm
O. exposure for 1 h (absolute mean difference = 1.3%, p < 0.05). Similarly, Solic et al.
(1982) observed a small reduction in arterial oxygen saturation in 13 mild to moderate COPD
patients at the end of a 0.2-ppm O. exposure for 2 h (absolute mean difference = 0.48%,
p < 0.008). In contrast, Kehrl et al. (1985) did not find a significant effect on arterial oxygen
saturation in 13 mild to moderate COPD patients after exposure to 0.3 ppm O. using the same
IE exposure protocol used by Solic et al. (1982). Similarly, Linn et al. (1983a) found no
significant effect on arterial oxygen saturation in 28 mild to moderate COPD patients exposed
to 0.18 and 0.25 ppm O. for 1 h. The combined observations of these studies indicate that
persons with COPD are not responsive to O. concentrations of 0.3 ppm and lower in
combination with mild exercise. However, this conclusion should be viewed within the
context of the low total inhaled dose of O. involved in the above studies, in that studies in
healthy subjects using similar total inhaled doses also have not shown significant pulmonary
function effects. Interpretation of these studies also is complicated by the wide range of the
pulmonary function impairment of the patients studied (FEV./FVC from 0.3 to 0.7), their
variable smoking history, and the fact that these patients are older (D60 years of age). The
inconsistency of the observed small decreases in arterial oxygen saturation makes the
interpretation of the clinical significance of this data difficult and uncertain.
Despite similar limitations, Kulle et al. (1984) observed small (<4%), statistically
significant decreases in FVC and FEV. in 20 smokers (age range 31 to 51 years) diagnosed
with mild chronic bronchitis exposed to 0.4 ppm O. for 3 h using an IE protocol (one 15-min
exercise period beginning 1 h prior to end of exposure, V. approximately 29 to 38 L/min).
In addition, Kulle et al. (1984) observed that repeated daily exposure over a 5-day period led
to an attenuation of these forced expiratory endpoints, and that this attenuation did not last
longer than 4 days. The pulmonary responses induced by O. exposure in this study were
associated only with mild symptoms.
Subjects with Asthma
Three studies examining the pulmonary responses to acute O. exposures in adult
(Linn et al., 1978; Silverman, 1979) and adolescent (Koenig et al., 1985) asthmatics were
discussed in the earlier criteria document (U.S. Environmental Protection Agency, 1986).
Significant decrements in group mean pulmonary function were not observed for adult
asthmatics exposed for 2 h at rest (Silverman, 1979) or with light IE (Linn et al., 1978) to
O. concentrations of 0.25 ppm or less. However, it should be noted that, although group mean
pulmonary function responses were not significantly affected in these studies, there were
responsive asthmatic subjects who had obvious decrements in pulmonary function.
Koenig and co-workers (Koenig et al., 1985, 1987, 1988) conducted a series of
studies examining the pulmonary responses of adolescent asthmatics and nonasthmatics (11 to
19 years of age) exposed to low levels of O.. Koenig et al. (1985) found no significant
changes in pulmonary function or symptoms in 10 adolescent normal and asthmatic subjects
(four male, six female) who inhaled 0.12 ppm O. for 1 h at rest. The asthmatic subjects in this
study were characterized as having histories of atopic (Type I, immunoglobulin E [IgE]-
mediated) asthma and exercise-induced bronchospasm. Subsequently, in two separate studies
of similar groups of adolescent asthmatics and nonasthmatics, Koenig et al. (1987, 1988)
7-24
-------
observed no significant changes in pulmonary function or symptoms following exposure to
0.12 and 0.18 ppm O. with moderate IE up to 1 h, although a small significant decrease in
flow at 50% of FVC was observed in the adolescent asthmatics exposed to 0.12 ppm O..
Kreit et al. (1989) and Eschenbacher et al. (1989) have demonstrated that exposure
to 0.4 ppm O. with heavy IE (exercise V. = 30 L/min/m' BSA) for 2 h elicits a significant
decrease in FVC, FEV., FEV./FVC, and FEF in both normal and asthmatic subjects.
In these studies, O. exposure caused significantly greater decrements in FEV., FEV./FVC, and
FEF in asthmatic subjects. In contrast, Kreit et al. (1989) and Eschenbacher et al. (1989)
found no significant difference between asthmatic and normal subjects in FVC and subjective
symptoms. In addition, the effect of O. exposure on bronchial responsiveness as measured by
the concentration of methacholine needed to increase SR.. 100% (PC ) was also studied.
The asthmatic subjects had a significant decrease in PC following FA and O. exposure.
In comparison, the normal subjects had a significant decrease in PC following
O. exposure, with the percent decrease in mean PC after O. exposure being similar in
normal and asthmatic subjects, although the asthmatic patients' baseline PC was
significantly lower than that of the normal subjects. The findings of this study indicate that if
the total inhaled dose is increased sufficiently by either increasing V. during exposure or
O. concentration, mild to moderate asthmatics will respond with a greater obstructive response
than will normal subjects.
Linn et al. (1994) have reported responses of healthy (n = 15) and asthmatic
(n = 30) subjects to 0.12 ppm O. and 100 Dg/m' of respirable sulfuric acid (H.SO.) aerosol
(MMAD = 0.5 Dm; geometric standard deviation [D.] = 2), alone and in combination using
the EPA prolonged-exposure protocol (see Section 7.2.2). These investigators observed a
significant O.-induced reduction in FEV. that was statistically significant and an increase in
airway responsiveness to methacholine for all subjects combined. The asthmatic subjects
demonstrated a statistically significant decrease in FEV. as a function of exposure duration
regardless of pollutant exposure. In addition, there was a greater reduction in FEV. following
O. alone in the asthmatics as compared to the nonasthmatics (D8.6% versus Dl.7%), although
this difference was not statistically significant. Despite the lack of a significant difference
between asthmatics' and nonasthmatics' group mean FEV. responses with O. exposure, the
responses observed in the asthmatics may be considered more important because their average
FEV. was already significantly depressed by the underlying illness.
The findings of the above studies comparing the pulmonary function responses
following O. exposure in asthmatic and nonasthmatic subjects suggest that asthmatics are at
least as sensitive, if not more sensitive, to the acute effects of O. inhalation. The underlying
mechanism that would explain a possible increased responsiveness of asthmatic subjects to
O. is undefined. One possible mechanism could be that asthmatic subjects have an exaggerated
airway inflammatory response to acute O. exposure. A study conducted by McBride et al.
(1994) would support this hypothesis. McBride et al. (1994) exposed 10 asymptomatic
asthmatic subjects with histories of allergic rhinitis and 8 nonallergic healthy subjects to FA
and 0.12 and 0.24 ppm O. for 90 min using a light IE protocol (V. = approximately 25
L/min). Pulmonary function tests, posterior rhinomanometry, and nasal lavage (NL) were
performed before exposure and 10 min and 6 and 24 h after exposure. No significant changes
in pulmonary or nasal function were found in either the allergic asthmatic or nonallergic
nonasthmatic subjects. The allergic asthmatic subjects had a significant increase in the number
of white blood cells in NL fluid 10 min and 24 h following the 0.24-ppm O. exposure. In
7-25
-------
addition, a significant increase in epithelial cells was present 10 min after exposure to 0.24
ppm O. in the asthmatic subjects. No significant cellular changes were observed in the
nonasthmatic subjects. These data indicate that the upper airways of asthmatic individuals are
more sensitive to the acute inflammatory effects of O. than those of nonallergic nonasthmatic
subjects.
The above studies compared the effects of O. inhalation on pulmonary function in
asthmatic and normal subjects, but do not address the effect of preexposure to ambient
concentrations of O. on the responsiveness of asthmatic subjects to other respiratory
challenges, including other irritant gases, allergens, and exercise. Koenig et al. (1990)
reported an increase in the bronchial response to an SO. challenge in a group of
13 asymptomatic adolescent asthmatic subjects following inhalation of 0.12 ppm O. for 45 min
using a light to moderate IE protocol (V. = approximately 30 L/min).
Molfmo et al. (1991) investigated whether resting exposure to 0.12 ppm O. for
1 h potentiates the airway response to inhaled allergen in seven patients with mild asthma with
seasonal symptoms of asthma and positive skin tests for ragweed or grass. This study was
conducted over four week-long periods during the winter when ambient allergen levels were
low. In each week, there were 3 consecutive study days. On Days 1 and 3, subjects
underwent methacholine challenges, whereas, on Day 2, the subjects received one of four
combined challenges in a single-blind design: (1) air breathing followed by inhalation of
allergen diluent, (2) O. exposure followed by inhalation of allergen diluent, (3) air breathing
followed by inhalation of allergen, and (4) O. exposure followed by inhalation of allergen.
Molfmo et al. (1991) observed no significant differences in baseline FEV. after O. exposure,
but did observe a significant reduction in the provocative concentration of allergen required to
reduce FEV. 15%. This study was limited by its small number of subjects, and the results
were confounded by possible ordering effects with the "O. exposure followed by allergen
protocol" being the last protocol for all but one subject. Despite these limitations, the findings
suggest that O. concentrations as low as 0.12 ppm may increase the bronchial responsiveness
to allergen in atopic subjects.
In order to examine whether preexposure to O. results in exacerbation of
exercise-induced asthma, two studies were conducted recently (Fernandes et al., 1994;
Weymer et al., 1994). Fernandes et al. (1994) preexposed 15 stable mild asthmatics with
exercise-induced asthma to 0.12 ppm O. for 1 h at rest followed by a 6-min exercise challenge
test and found no significant effect on either the magnitude or time course of exercise-induced
bronchoconstriction. Similarly, Weymer et al. (1994) observed that preexposure to either
0.10 or 0.25 ppm O. for 60 min while performing light IE did not enhance or produce
exercise-induced asthma in 21 otherwise healthy adult subjects with stable mild asthma.
Although the results of these studies would suggest that preexposure to O. neither enhances nor
produces exercise-induced asthma in asthmatic subjects, the relatively low total inhaled doses
used in the above studies limit the ability to draw any definitive conclusions.
Subjects with Allergic Rhinitis
McDonnell et al. (1987) exposed 26 adults (18 to 30 years of age) with allergic
rhinitis to clean air and 0.18 ppm O. for 2 h using an IE protocol (V. =64 L/min at 15-min
intervals). The study subjects with allergic rhinitis did not have a history of asthma-like
symptoms. Following O. exposure, the subjects with allergic rhinitis exhibited significant
increases in respiratory symptoms, airway reactivity to histamine, and SR.. and significant
7-26
-------
decreases in FVC, FEV., and FEF when compared to clean air exposure. When
compared to normal subjects without allergic rhinitis similarly exposed to 0.18 ppm O., the
subjects with allergic rhinitis were no more responsive to O., based on symptoms, forced
expiratory parameters, or airway reactivity to histamine aerosols, although subjects with
allergic rhinitis did have a small but significantly greater increase in SR... The data on
subjects with allergic rhinitis and asthmatic subjects suggest that both of these groups have a
greater rise in R.. to O. with a relative order of airway responsiveness to O. being normal <
allergic < asthmatic.
Bascom et al. (1990) conducted a study to characterize the upper respiratory
response to acute O. inhalation, nasal challenge with antigen, and the combination of the two.
Bascom et al. (1990) exposed 12 resting asymptomatic subjects with histories of allergic
rhinitis in a randomized, crossover design on each of 2 days, separated by 2 weeks, to clean
air or 0.5 ppm O. for 4 h. Following exposure, subjects underwent nasal challenge with four
doses of antigen (1, 10, 100, and 1,000 protein nitrogen units of ragweed or grass). Upper
and lower airway symptoms were rated and NL was performed before and after clean air and
0.5 ppm O. exposure, and following each antigen challenge. Exposure to O. caused significant
increases in upper and lower airway symptoms, a mixed inflammatory cell influx with a
sevenfold increase in NL PMNs, a 20-fold increase in eosinophils and a 10-fold increase in
mononuclear cells as well as an apparent sloughing of epithelial cells. There was a significant
increase in NL albumin concentration following O. exposure. When expressed as a change
from the postexposure values, there was no significant difference between O. and clean air
exposure in antigen-induced upper and lower airway symptoms, cells, albumin and mediators
(histamine and TAME-esterase activity). These results suggest that acute exposure to O. does
not alter the acute response to nasal challenge with antigen.
Subjects with Ischemic Heart Disease
One study has been conducted examining the cardiopulmonary effects of acute
O. inhalation in patients with ischemic heart disease. Superko et al. (1984) exposed six
middle-aged males with angina-symptom-limited exercise tolerance for 40 min to FA and to
0.2 and 0.3 O. while they were exercising continuously according to a protocol simulating
their angina-symptom-limited exercise training prescription (mean V. = 35 L/min).
No significant pulmonary function impairment or evidence of cardiovascular strain induced by
O. inhalation was observed. The low workloads were dictated by the patients'
angina-symptom-limited exercise tolerance, and these low workloads acted to "protect" them
from O.-induced effects by limiting the total inhaled dose.
7.2.1.3 Influence of Gender, Age, Ethnic, and Environmental Factors
Gender Differences
As was noted in the previous O. criteria document (U.S. Environmental Protection
Agency, 1986), the pulmonary function responses to O. of only a small number of female
subjects have been evaluated under controlled laboratory conditions. Although the database on
females has expanded (see Table 7-3), there are still fewer data than for males. Most studies
involving mixed groups of male and female subjects include too few female subjects to allow
for meaningful comparisons between the responses of the sexes, or fail to consider the question
at all. There are, however, a few studies that utilize only female subjects. Several studies
7-27
-------
cited in the 1986 O. criteria document suggested that females might be more responsive to
O. than males (Horvath et al., 1979; Gliner et al., 1983; Gibbons and Adams,
7-28
-------
Table 7-3. Gender Differences in Pulmonary Function Responses to Ozonfe
Ozone
Concentrationb ^ ^ .
ppm
0.12
0.18
0.24
0.30
0.40
0.18
0.30
0.20
0.30
0.30
Dg/m
235
353
470
588
784
353
588
392
588
588
exposure umaiuui
and Activity
2.33 h
VE = 25 L/min/m2
BSA
(one
exposure/subject)
1 h (mouthpiece)
CE
VE D 47 L/min
1 h (mouthpiece)
IE (20 min exercise)
VE D 28 L/min for
men
VE D 23 L/min for
women
1 h (mouthpiece)
CE
VE D 70 L/min for
men
VE D 50 L/min for
women
Exposure
Conditions'
Mean T = 22 DC
MeanRH = 4%
treadmill
T = 21 to 25 DC
RH = 45 to 60%
cycle
T D 22 DC
RHD75%
treadmill
T = 21 to 25 DC
RH = 45 to 60%
cycle
Number and
Gender of
Subjects
30 to 33 F and
30 to 33 M in
each
concentration
group; total of
372 individuals
participated
14 F
14 F
9M
10 F
20 M
20 F
Subject
Characteristics
Healthy NS,
18 to 35 years
old, blacks and whites
Mean FVC =
L,
NS, 20 to 24
Mean FVC =
L,
NS, 19 to 23
NS, 55 to 74
NS, 56 to 74
NS, 18 to 30
NS, 19 to 25
= 5.11 + 0.53
years old
= 3.74 + 0.30
years old
years old
years old
years old
years old
Observed Effect(s)
Decrements in FEV, ,
increases in SRj,w and cough,
correlated with
O3 concentration. There were
no significant differences
between the responses of
males and females.
Significant concentration-
response effect on FVC and
FEV, ; lung size had no effect
on percentage decrements in
FVC or FEV,.
No changes in spirometry in
men or women. Women had
significant 13% increase in RT
following exposure, which
was sustained at 20 min
postexposure.
Significant decrements in
FVC, FEV,, and FEF25_75%
following O3 exposure.
No significant differences
between men and women for
spirometry or SRj,w.
Reference
Seal et al. (1993)
Messineo and Adams
(1990)
Reisenauer et al. (1988)
Adams et al. (1987)
-------
Table 7-3 (cont'd). Gender Differences in Pulmonary Function Responses to Ozonfe
u>
o
Ozone
Concentration5
ppm Dg/m
0.45 882
0.45 882
0.48 941
Exposure Duration
and Activity
2h
IE
VE D 27.9 L/min for
men
VE D 25.4 L/min for
women
2h
IE
Mean VE =
28.5 L/min for men
Mean VE =
26. 1 L/min for
women
2h
IE
VE D 25 L/min
Exposure
Conditions'
T = 24 DC
RH = 58%
cycle
MeanT = 23.1 DC
Mean RH = 46. 1 %
cycle/treadmill
T = 21 DC
WBGT
cycle
Number and
Gender of
Subjects
8M
8F
10 M
6F
10 F
Subject
Characteristics
Healthy NS,
51 to 69 years old
Healthy NS,
56 to 76 years old
Healthy NS,
60 to 89 years old
Healthy NS,
64 to 71 years old
Healthy NS,
19 to 36 years old
3 Observed Effect(s) Reference
Range of responses in FEV, : Drechsler-Parks et al.
0 to Dl2% (mean = D5.6%). (1987a,b)
No significant difference in
responses of men and women.
Tendency for women to have
greater effects.
Mean decrement in FEV, = Bed! et al. (1989)
5.7%. Decrements in FVC and
FEV, were the only pulmonary
functions significantly altered by
O3 exposure. No significant
differences between responses of
men and women.
Mean decrement in Horvath et al. (1986)
FEV, = 22.4%.
Significant decrements in all
spirometry measurements.
Results not significantly different
from a similar study on males
(Drechsler-Parks etal., 1984).
"See Appendix A for abbreviations and acronyms.
bListed from lowest to highest O3 concentration.
WrSVJl — 0. / lwetbulb ' 0.3 1 dry bulb or globe'
-------
1984; Lauritzen and Adams, 1985). DeLucia et al. (1983), on the other hand, did not find
significant differences in the responses of young men and young women to O. exposure.
Messineo and Adams (1990) hypothesized that differences previously observed
between the responses of males and females exposed to O. were related to differences in lung
size between the sexes. They addressed this issue by selecting two groups of 14 women each.
One group had a mean FVC of 5.11 L, and the other group had a mean FVC of 3.74 L. All
subjects were 19 to 24 years of age and were healthy nonsmokers who had not lived in a high-
air-pollution area for at least 6 mo. The subjects completed three 1-h CE (V. =47 L/min)
exposures: (1) FA, (2) 0.18 ppm O., and (3) 0.30 ppm O.. The mouthpiece exposures were
presented in random order, at least 4 days apart, and all were performed when the subject was
in the follicular phase of her menstrual cycle. Two subjects in the small-lung group and one in
the large-lung group were unable to complete the 0.30 ppm O. exposure. Both groups had
similar O.-induced percentage decrements (9 to 10% following exposure to 0.18 ppm O. and
23 and 26% following exposure to 0.30 ppm O. for the small- and large-lung groups,
respectively) in all measures of lung function, regardless of lung size, leading to the
conclusion that lung size, per se, is not systematically related to percentage decrements in
FEV. consequent to O. exposure.
Horvath et al. (1986) exposed 10 healthy, young, nonsmoking females, 19 to
36 years of age (mean age 23.6 years) to 0.48 ppm O. or FA for 2 h while they exercised
intermittently at a target ventilation of 25 L/min. The subjects engaged in three 20-min cycle
ergometer exercise periods alternated with four 15-min rest periods. The exposures were a
minimum of 1 week apart. The responses of these subjects were compared with those of a
group of 10 young males who earlier had completed the same protocol (Drechsler-Parks et al.,
1984). There were no statistically significant differences in the responses based on gender.
The female subjects had decrements of 18.8, 22.4, and 30.8% in FVC, FEV., and FEF ,
respectively, compared to 19.8, 25.0, and 31.9% for the male subjects. On an individual
basis, 4 of the 10 males and 3 of the 10 females had decrements of 30% or more in FEV.
following the exposure to 0.48 ppm O.. One male subject did not respond to the O. exposure.
It was noted, however, that the female subjects inhaled an absolute dose of O. about 22% less
than the male subjects due to a slightly lower exercise V. and the inherently lower resting V.
of females compared to males. However, when O. dose was related to BSA or to FVC, the
females inhaled slightly higher relative doses of O. than the males.
Adams et al. (1987) compared the responses of 20 young men (18 to 30 years of
age) and 20 young women (19 to 35 years of age) exposed to 0.3 ppm O. via mouthpiece. All
subjects were healthy nonsmokers with clinically normal pulmonary function. None had a
history of significant allergies, and none had resided in a high-air-pollution area for at least
3 mo. The subjects completed 1-h CE exposures (mean V. D 70 L/min for males and
50 L/min for females) to FA and 0.3 ppm O.. The exposures were given in random order and
were separated by a minimum of 5 days. Ozone exposure induced significant decrements in
FVC, FEV., and FEF compared to FA exposure. Three females and four males were
unable to complete the O. exposure. Females experienced mean decrements of 14.2, 20.3, and
24.5% in FVC, FEV., and FEF , respectively, compared to mean decrements of 15.8% in
FVC, 23.8% in FEV., and 35.7% in FEF for males. There were no statistically
significant differences between the spirometry or SR.. responses related to gender. Because
the female subjects inhaled a substantially smaller absolute dose of O. due to the considerably
lower exercise V., yet had similar decrements in pulmonary function compared to men, the
7-31
-------
authors concluded that females are more responsive to O. than males. In this study, the female
subjects inhaled a lower relative dose of O. compared to males, when expressed on the basis of
BSA, but a similar relative dose when expressed on the basis of FVC.
Seal et al. (1993) reported on 372 healthy black and white men and women between
18 and 35 years of age who each were assigned to complete one 2.33-h exposure to FA, 0.12,
0.18, 0.24, 0.30, or 0.40 ppm O.. Subjects exercised intermittently on a motor-driven
treadmill at a work load inducing a V. of about 23 L/min/m' BSA for women and about
24.5 L/min/m' BSA for men. Although female subjects inhaled about 22% less total dose of
O. than males in each exposure-concentration group, there were no significant differences in
the changes in FEV. (see Figure 7-2), SR.., or cough ratings between males and females
among either blacks or whites. Women also inhaled a lower absolute dose of O. than men.
0.00 0.12 0.18 0.24
Ozone Concentration (ppm)
0.30
0.40
Figure 7-2. Mean percent change (± standard error of the mean) in post-minus
prevalues of forced expiratory volume in 1 s (FEVt) for each gender-race
group. Open bars = white women; cross-hatched bars = black women;
hatched bars = white men; solid bars = black men.
Source: Seal et al. (1993).
7-32
-------
Drechsler-Parks et al. (1987a,b) compared the responses of eight men and eight
women between 51 and 76 years of age to FA and 0.45 ppm O.. The subjects were all healthy
nonsmokers who were long-term residents of a relatively low pollution area. The subjects
participated in 2-h IE (20 min rest/20 min exercise at V. =25 L/min) exposures that were
presented in random order and were separated by at least 1 week. Except for FEV., there
were no statistically significant differences between the responses of the men and women
subjects, although women had slightly larger mean decrements in FVC and FEV. than men.
Individual decrements in FVC and FEV. ranged from 0 to about 12% for both male and female
subjects. Based on FEV., two females and three males had no response to the O. exposure.
Male subjects inhaled a somewhat larger absolute effective dose of O. due to higher exercise
and resting V.. When V. was normalized to BSA, females inhaled a larger dose of O. than
males. When V. was normalized to FVC, the relative inhaled doses of O. were similar.
Reisenauer et al. (1988) reported on the pulmonary function responses of 9 men and
10 women between 55 and 74 years of age who were exposed to 0.0, 0.2, and 0.3 ppm O..
The three exposures were presented in random order and at the same time of day for each
subject. The subjects were exposed via mouthpiece for 1 h, during which seven men exercised
for 10 min and rested for 50 min, and the other two men and all of the women alternated two
20-min rest periods and two 10-min exercise periods. Ventilation rates were about 28 L/min
for men and 23 L/min for women, although, when V. was normalized to BSA, the relative V.
for males and females was similar. All data were pooled, regardless of the total exercise time.
There were no significant changes in any parameter of pulmonary function in the males.
Females had no significant changes in any spirometric parameter, but, following the 0.3-ppm
O. exposure, did have a small (13%) increase in total respiratory resistance (R.), which
remained at this level 20 min postexposure.
Bedi et al. (1989) reported on the responses of 10 men and 6 women (60 to 89 years
of age) exposed for 2 h to FA or 0.45 ppm O. for 3 consecutive days. Only the first O. day
results will be discussed in this section; the issue of repeated exposures is addressed in Section
7.2.1.4. Exposures were conducted at the same time of day, on consecutive days, with the FA
exposure always conducted first. The subjects alternated 20-min exercise periods (mean V. =
28.5 L/min for men and 26.1 L/min for women) and 20-min rest periods throughout the 2-h
chamber exposures. When V. was normalized to BSA, women inhaled slightly higher relative
doses of O.; but when normalized to FVC, women inhaled a slightly lower relative dose of
O. than men. There were no statistically significant group mean differences between the
responses of men and women subjects. The mean decrements in FVC and FEV. following the
O. exposure for the 16 subjects were 2.8 and 5.7%, respectively. In an exploratory analysis,
the subjects were divided into two groups based on whether their decrement in FEV. following
the first O. exposure compared to the FA exposure was D5 % or < 5 %. There were eight
subjects in each group, with the sensitive group consisting of two females and six males. The
mean post-O. exposure decrement in FEV. was 320 mL for the sensitive group, versus 21 mL
for the nonresponsive group. Similar patterns of response were evident in FVC and FEV..
There were no significant changes in any flow parameter, maximum voluntary ventilation
(MVV), expiratory reserve volume, or functional residual capacity.
The question as to whether there is a difference in sensitivity to O. between men
and women remains unresolved. Different conclusions depend on whether V. is normalized to
body or lung size in calculating the inhaled doses of O.. The subgroups studied by Bedi et al.
(1989) included six males and two females, suggesting that older males may be more sensitive
7-33
-------
to O. than older females. However, Reisenauer et al. (1988) found a significant increase in
R. only in women. Horvath et al. (1986), Adams et al. (1987), and Drechsler-Parks et al.
(1987a,b) suggested that because their female subjects had similar pulmonary function
responses to their male subjects, even though the females inhaled less O., females were more
sensitive than males. Messineo and Adams (1990) suggested that some factor other than
absolute lung size accounted for observed differences between males and females; their two
groups of females with widely different lung sizes experienced similar decrements in
pulmonary function following equivalent exposures. Although the currently available literature
suggests that females may be somewhat more sensitive to O. than males, the question is not
settled. Further, comparative studies have included only small subject groups, except for Seal
et al. (1993), and often only group mean data are presented, with little information about
individual responses.
Hormonal Influences
Seal et al. (1995) compared the pulmonary function responses of 48 white and
55 black women (18 to 35 years of age) whose menstrual phase was known at the time of a
single 2.3-h exposure to 0.18, 0.24, 0.30, or 0.40 ppm O.. Subjects performed intermittent
treadmill exercise (V. =20 L/min/m' BSA) during the first 2 h of exposure. There were no
significant effects for SR.. or cough that could be related to menstrual cycle. There was a race
X menstrual phase interaction for FEV.. However, when the groups of black and white
women were analyzed separately, there was no significant primary effect for menstrual cycle
phase. The significance of the observed interaction between race and menstrual cycle phase is
unknown.
Weinmann et al. (1995) compared the pulmonary function responses of six healthy,
nonsmoking women to a 130-min exposure to 0.35 ppm O., 4 to 8 days after the onset of
menses and 4 to 8 days after ovulation. Subjects performed intermittent exercise at a workload
that induced a V. of 10 x FVC. Ovulation was confirmed by a blood progesterone test.
Spirometry was performed pre- and 25-min post-O. exposure. Although resting V. was the
same during both exposures, exercise load had to be reduced 30% during the luteal phase in
order to match the ventilatory response to exercise during the follicular phase. There were no
significant effects related to phase of the menstrual cycle. The authors concluded that
menstrual phase does not need to be considered in experimental design. One problem with the
study is that the postexposure measurements were made 25 min after the conclusion of the
exposure. Typically, pulmonary function decrements begin to reverse once exposure ends;
thus, any pulmonary function changes that did occur could be expected to be reduced at 25-min
post-O. exposure, compared to immediately after exposure.
Acute O. exposure has been shown to cause short-term airway inflammation (see
Section 7.2.4) induced by PGs, among other inflammatory substances. It also has been
demonstrated that progesterone inhibits PG production in the uterine endometrium, which
fluctuates as the progesterone concentration varies throughout the menstrual cycle. Fox et al.
(1993) investigated the hypothesis that O. exposure during the follicular phase, when
progesterone concentration is lowest, might result in greater pulmonary function responses due
to reduced anti-inflammatory influences of progesterone. Nine nonsmoking women completed
1-h mouthpiece exposures to FA and 0.3 ppm O. while exercising continuously (V. about
50 L/min) during both the follicular and mid-luteal phases of two to four ovulatory menstrual
cycles. There were no differences in any pulmonary function responses to FA related to
7-34
-------
menstrual phase, nor was there a difference in the mean FVC decrements following the
follicular or mid-luteal phase O. exposures. The O.-induced decrements in FEV. and FEF
were significantly larger during the follicular phase (17.3 and 23.1%, respectively) than during
the mid-luteal phase (13.4 and 15.3%, respectively). The authors speculated that the
difference between the FEV. and FEF responses to the two O. exposures could be due to
differences in circulating progesterone and the effect of progesterone on prostaglandin activity.
Available data (see Table 7-4) do not permit a conclusion regarding the influence of
the menstrual cycle on responses to O. exposure. Two of the three studies available, Fox et al.
(1993) and Weinmann et al. (1995), were performed with small groups of subjects and resulted
in opposite conclusions. Seal et al. (1995) compared race (black versus white) and menstrual
phase, obtaining a significant interaction between race and phase, but post-hoc analysis failed
to establish a basis for the interaction, leaving the implications of the study unclear.
Age Differences
It has been hypothesized that age may be a factor in responsiveness to O.. Although
children make up a large proportion of the population, few controlled laboratory studies of the
pulmonary function effects of any air pollutant have been reported on subjects under age 18.
Field and epidemiological studies (see Section 7.4) attempting to relate ambient air pollutant
exposure to pulmonary function in children have suggested that children may be more
responsive to ambient air pollution than young adults.
The previous O. criteria document (U.S. Environmental Protection Agency, 1986)
included only one laboratory exposure study in which children were the subjects. McDonnell
et al. (1985a) evaluated the pulmonary function responses of 23 boys between 8 and 11 years
of age to 0.00 and 0.12 ppm O. in random order. The boys alternated 15-min rest and
exercise periods (V. =35 L/min/m' BSA) for the first 120 min of the 150-min exposure.
Forced expiratory spirometry and respiratory symptoms were measured before exposure and at
125 min of exposure, whereas R.. was measured before exposure began and after 145 min of
exposure. The group mean decrement in FEV. following the O. exposure was 3.4%,
compared to 4.3% for a group of young adult males who earlier had completed the same
protocol (McDonnell et al., 1983). It should be noted that the absolute V. for the children
(39.4 L/min) and adults (65.0 L/min) was similar when normalized for BSA (about
35 L/min/m' BSA). Assuming that adjusting ventilation for differences in BSA is an
appropriate normalizing technique, these children appeared to experience O.-induced
pulmonary effects similar to adults. The children reported no symptoms, but the adults
reported a small, but statistically significant, increase in cough following O. exposure.
Although controlled laboratory studies of the effects of exposure to air pollutants
are rarely performed with children as subjects, a few, more recent studies are discussed below
(see Table 7-5). Avol et al. (1987) have reported on the pulmonary function responses of
33 healthy boys and 33 healthy girls having a mean age of 9.4 years. The children completed
exposures to purified air and outdoor ambient air that was drawn into an environmental
chamber. Ambient temperature averaged about 33 DC. Exposures were 1 h in duration, were
separated by a minimum of 2 weeks, and were conducted from June through September,
beginning in the early afternoon when ambient air pollutant concentrations generally peak.
The subjects performed continuous exercise throughout the
7-35
-------
Table 7-4. Hormonal Influences on Pulmonary Function Responses to Ozonfe
u>
Ozone
Concentration1"
ppm
0.12
0.24
0.30
0.40
0.30
0.35
Dg/m
235
470
588
784
588
686
Exposure
Duration and
Activity
2.3 h IE
VE =
20 L/min/m2 BSA
IhCE
VE D 50 L/min
130 min
Exposure Number and Subject
Conditions Gender of Subjects Characteristics
NA 48 WF, 55 BF Healthy NS,
18 to 35 years
old
NA 9 F Healthy NS,
regular
menstrual
cycles,
20 to 34 years
old
NA 9 F Healthy NS,
regular
menstrual
cycles, 18 to
35 years old
Observed Effect(s)
Significant menstrual cycle phase x race
interaction for FEV, . No significant
menstrual cycle phase effect when blacks
and whites were analyzed separately.
No significant menstrual phase effects for
SR.,W or cough score.
FEV, decreased 13.1% during the mid-
luteal phase and 18. 1 % during the
follicular phase. Decrement in FEF25_75%
was significantly larger during the
follicular phase than the mid-luteal phase.
Changes in FVC were similar in both
phases.
Changes in FVC, FEV,, FEF25_75%,
vma*5o%> and VmaX25% were similar during
both the follicular and luteal phases.
References
Seal et al. (1995)
Fox et al. (1993)
Weinmann et al.
(1995)
"See Appendix A for abbreviations and acronyms.
bListed from lowest to highest O3 concentration.
-------
7-37
-------
Table 7-5. Age Differences in Pulmonary Function Responses to Ozonfe
u>
CO
Ozone
Concentration5
Exposure Duration
ppm Dg/m and Activity
0.1 13C
other
ambient
pollutants
0.12
0.12
0.18
0.18
0.24
0.30
0.40
0.20
0.30
221 Ih
CE
VE D 22 L/min
235 1 h (mouthpiece)
IE
VE = 4 to 5 x
resting
235 40 min
(mouthpiece)
IE
10 min exercise at
353 VE = 32.6 L/min;
40 min
(mouthpiece)
IE
10 min exercise at
VE = 41. 3 L/min
353 2.3 h
470 IE
588 VE = 20 L/min/m2
784 BSA
392 1 h (mouthpiece)
588 IE (20 min)
VE D 28 L/min for
men
VE D 23 L/min for
women
Exposure
Conditions
T = 32.7 DC
RHD43%
cycle
T = 22 DC
RH = 75%
treadmill
NA
treadmill
NA
T D 22 DC
RHD75%
treadmill
Number and
Gender of Subject
Subjects Characteristics
Observed Effect(s)
33 M, 33 F NS for both groups, No differences in responses of boys and girls.
mean age = 9.4 years Similar decrements (<5% on average)
old following both purified air and ambient air (O3
at 0.11 ppm) exposures.
5 M, 7 F Healthy NS,
12 to 17 years old
3 M, 7 F Healthy NS,
14 to 19 years old
4M, 6F
48 WF, 55 Healthy NS,
BF 18 to 35 years old,
black and white
9 M, 10 F Healthy NS,
55 to 74 years old
No significant changes in any pulmonary
function in healthy subjects.
No significant change in FEV, ; increased RT
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.
Older women had smaller changes in FEV, than
younger women. No age- related differences in
SRaw or cough score.
No change in any spirometry measure. Women
had 13% increase in RT after 0.30-ppm
exposure.
Reference
Avol et al.
(1987)
Koenig et al.
(1988)
Koenig et al.
(1987)
Seal et al.
(1993)
Reisenauer et
al. (1988)
-------
7-39
-------
Table 7-5 (cont'd). Age Differences in Pulmonary Function Responses to Ozonfe
Ozone
Concentration5
ppm Dg/m
0.45 882
0.45 882
0.45 882
0.45 882
Exposure Duration
and Activity
2h
IE
VE D 26 L/min
2h
IE
Mean VE =
28.5 L/min for men
Mean VE =
26. 1 L/min for
women
2h
IE
VE D 26 L/min
1 h
CE
VE D 26 L/min
2h
IE
VE D 26 L/min
Exposure
Conditions
T D 23 DC
RH = 53%
cycle
T = 23 DC
RH = 46%
cycle/treadmill
T D 24 DC
RH = 63%
cycle
T D 23 DC
RH = 58%
cycle/treadmill
Number and
Gender of
Subjects
8M
8F
10 M
6F
8M
8F
7M
5F
Subject
Characteristics
Healthy NS,
51 to 76 years old
Healthy NS,
60 to 89 years old
Healthy NS,
51 to 69 years old
Healthy NS,
56 to 76 years old
Healthy NS,
60 to 79 years old
(all in 60s
except one
79 years old)
Observed Effect(s) Reference
Mean decrement in Drechsler-Parks et al.
FEV, = 5.6 + 13%; range of (1987a,b)
decrements = 0 to 12%.
Mean decrement in Bedi et al. (1989)
FEV, = 5.7%; eight subjects had a 5%
or greater difference between their
response to O3 and FA, and the other
eight had less than a 5 % difference
between their responses to FA and
0.45 ppm O3.
13 subjects had decrements in FEV, Bedi et al. (1988)
on three separate exposures to 0.45 ppm
within 5 % of their mean response to the
three exposures. The other three subjects
were not reproducible. Symptom reports
did not correlate well with pulmonary
function changes.
Comparison of 1-h CE protocol and 2-h Drechsler-Parks et al.
IE protocol indicated no difference (1990)
between the changes in pulmonary
function following the two protocols.
"See Appendix A for abbreviations and acronyms.
bListed from lowest to highest O3 concentration.
'Ozone concentration is the mean of a range of ambient concentrations.
-------
7-41
-------
hour of exposure. Boys and girls exercised at similar V., 22 to 23 L/min. It should be noted
that the ambient exposure included the full range of air pollutants present in the outdoor air
mix on the days of the exposures, except for small fractions of O. and particles lost in the inlet
duct. Concentrations of O., nitrogen dioxide (NO.), total suspended particulate (TSP),
particulate nitrate, particulate sulfate, particulate sodium, and particulate ammonium were
measured throughout the exposures. The O. concentration during the ambient air exposures
averaged 0.113 ppm, whereas it averaged 0.003 ppm during the purified air exposures. The
children consistently had similar declines in pulmonary function with time, following both FA
and ambient air exposures. Typical mean decrements in FVC and FEV. were 50 mL or less.
The investigators also have published similar studies on adolescents and adults (Avol et al.,
1984, 1985a). The responses of the adolescents and adults to both exposures were not
substantially different from those of the children whose results are reported here. The authors
further noted that the children seemed to have difficulty performing consistent, reproducible
pulmonary function tests, a factor that could have impacted on these results.
Several studies comparing the pulmonary function responses of healthy and
asthmatic adolescents to O. exposure have appeared in the literature. The responses of the
asthmatics are presented in Section 7.2.1.2; only data on normal adolescent subjects will be
discussed in this section.
Koenig et al. (1987) reported on 20 adolescents, 14 to 19 years of age, who were
exposed for 40 min to air or 0.12 or 0.18 ppm O. via a mouthpiece system. Ten subjects were
exposed to each O. concentration, but not all subjects were exposed to both concentrations.
None of the healthy subjects had a history of asthma or allergies, and all had pulmonary
function within the predicted range, based on age, sex, and height. There was a 5- to 7-min
break in exposure for pulmonary function test performance following 30 min of resting
exposure, followed by a 10-min exercise period (V. = 32.6 + 6.4 L/min for the 0.12-ppm
O. exposure and 41.3 + 9.3 L/min for the 0.18-ppm O. exposure). Changes in FEV. were not
significant following any exposure. After exposure to 0.18 ppm O., R. was increased 15%.
Koenig et al. (1988) also have reported on the pulmonary function responses of
another group of 12 healthy adolescents (12 to 17 years of age) to 1-h exposures to air and
0.12 ppm O.. The subjects were exposed by mouthpiece to air and 0.12 ppm O. while
alternating 15-min periods of exercise (V. = 32.8 + 6.0 L/min) with 15-min periods of rest
(V. = 8.8 + 1.2 L/min). Tests of pulmonary function included forced expiratory spirometry
and R.. Healthy subjects had no significant alterations in any parameter of pulmonary function
consequent to exposure to air or 0.12 ppm O..
Although few data are available on the responses of healthy adolescents exposed to
O., the limited existing data do not identify adolescents as being either more or less responsive
than young adults.
At the time the 1986 O. criteria document was released, no studies specifically
evaluating the pulmonary function responses of older adults had been reported. Several studies
(Folinsbee et al., 1985; Adams et al., 1981) that included a few middle-aged individuals among
the subjects were suggestive that there might be a decrease in O. responsiveness with
advancing age. Several reports have since appeared, collectively suggesting that, collectively,
healthy older adults (i.e., over 50 years of age) generally are minimally responsive to O.,
although some individuals remain responsive to O..
Drechsler-Parks et al. (1987a) reported on eight men and eight women between
51 and 76 years of age who were exposed for 2 h to FA or 0.45 ppm O.. The subjects were
7-42
-------
healthy nonsmokers with normal baseline pulmonary function. The chamber exposures
involved alternating 20-min rest and exercise periods (V. averaged 27.9 + 0.29 L/min for
men and 25.4 + 0.80 L/min for women). Exposures were presented in random order at least
1 week apart. The only significant difference related to sex was in FEV., in which women had
larger decrements than men. There were no significant decrements in any other parameter of
pulmonary function related to O. exposure, except when the data of all 16 subjects were
pooled, significant mean decrements of 5.3 + 1.3% inFVC, and 5.6 + 1.3% in FEV. were
observed. The range of individual decrements in FEV. was from 0 to 12%. Two women and
three men had no response following the O. exposure. The subjects reported more symptoms
following the O. exposure than the FA exposure. Seven subjects reported cough, nine reported
sore throat, and six reported chest tightness. The authors compared their results on older
adults to the results from a group of young adults who had completed the same protocol. All
of the older subjects inhaled slightly higher doses of O. than the younger subjects (10.23 x
10" L for older men versus 10.12 X 10" L for younger men and 8.48 X 10" L for older
women versus 7.94 X 10" L for younger women). However, older men had a mean
decrement in FEV. of 4.2% versus 23.7% for younger men, whereas older women had a mean
decrement in FEV. of 7.0% versus 14.7% for younger women. The decrements of the older
subjects also were compared to published values for the young adults, with the older subjects
studied consistently showing smaller changes in pulmonary functions than the young adults.
These comparisons indicated that these older adults were less responsive to O. exposure than
typical young adults, in terms of pulmonary function changes and symptom reports.
Reisenauer et al. (1988) reported on the pulmonary function responses of 19 healthy
adults between 55 and 74 years of age. All were nonsmokers with baseline pulmonary
function within the predicted normal range. None had a history of asthma, atopy, or
cardiovascular disease, and none responded to a baseline methacholine bronchochallenge.
Subjects were exposed by mouthpiece to 0.0, 0.20, or 0.30 ppm O. for 1 h. Seven men rested
for 50 min and exercised for 10 min, whereas the other two men and all women alternated two
20-min rest periods with two 10-min exercise periods; V. was approximately 28 L/min for
men and 23 L/min for women. The three exposures were presented in random order at the
same time of day for each individual, but the separation between exposures is not stated. The
only significant change in pulmonary function was a 13% increase in R. with exposure to 0.30
ppm O. in women only, which was sustained for at least 20 min postexposure. The authors
concluded, based on the increase in R. in the women, that older adults were at increased risk
for pulmonary function changes with near-ambient O. exposure. However, R. is a highly
variable parameter, and no other changes were significant. Given the large number of
variables tested, this isolated result possibly is related to the large number of statistical tests
performed. In contrast, the results of Koenig et al. (1987) from 10 healthy adolescents
exposed to 0.18 ppm O. using a similar protocol, reported a mean decrement in FEV. of 2%
and an increase of 10.5% in R. at 2 to 3 min postexposure. At 7 to 8 min postexposure, the
increase in R. was 15.3%. Comparison of these results to those of Reisenauer et al. (1988) at
0.20 ppm O. supports the contention that younger individuals are more responsive to O. than
older individuals, in that no changes in spirometry were noted in the older adults exposed to
0.30 ppm O., although older women showed increased R. with 0.30 ppm O. exposure,
whereas the adolescents had a mean decrement of 1 % in FEV. and a mean increase of 16% in
R. with exposure to 0.18 ppm O..
7-43
-------
Bedi et al. (1988) reported that older men and women 51 to 76 years of age who
completed three exposures to 0.45 ppm O. did not respond equivalently to each of three
exposures. The subjects were healthy nonsmokers with baseline pulmonary function within
predicted normal limits. The subjects alternated 20-min exercise (V. was approximately
26 L/min) and 20-min rest periods throughout the 2-h chamber exposures. There was a
minimum of 1 week between exposures, but separations ranged from 1 to 4 weeks between
exposures 1 and 2, and between 1 and 7 weeks between exposures 2 and 3. Analysis of
variance indicated no difference between the group mean responses to the three exposures.
The data were then subjected to a correlation analysis, which led to the conclusion that the
responses within an individual subject were not reproducible. McDonnell et al. (1985b), on
the other hand, found good reproducibility of pulmonary function responses after exposure to
various concentrations of O. between 0.12 and 0.40 ppm in young adult males between 18 and
30 years of age.
Seal et al. (1993) compared the pulmonary function responses of 48 white and
55 black women (18 to 35 years) who each completed a 2.3-h exposure to 0.18, 0.24, 0.30, or
0.40 ppm O.. The subjects participated in only one exposure each while exercising
intermittently (V. =20 L/min/m' BSA) during the first 2 h of the exposure. Older subjects
within the age range tested had smaller decrements in FEV. than younger subjects.
One simple method to estimate the O. exposure dose is to calculate the product of
O. concentration (parts per million), V. (liters per minute), and exposure duration (minutes).
Research on young adults (Folinsbee et al., 1978; Adams et al., 1981) has demonstrated that
the order of relative importance of the three factors is O. concentration, V., and exposure
duration. Drechsler-Parks et al. (1990) investigated the relative role of the three components
of effective dose in 12 healthy, nonsmoking adults between 61 and 79 years of age. The
subjects were exposed to both FA and 0.45 ppm O., once while they performed a 1-h
continuous exercise protocol and once while they performed a 2-h IE protocol in which they
alternated 20-min exercise periods and 20-min rest periods. Mean V. ranged from 25.2 to
27.3 L/min among the four exposures. Exposures were separated by at least 1 week.
Regardless of protocol, O. exposure induced significant decrements in FEV..., FEV. (7.7 and
10.6% for the 1- and 2-h exposures, respectively), FEV., and peak expiratory flow rate
(PEFR) compared to FA exposure. There were significant decrements in FEF (0.37,
0.46, 0.49, and 0.47 L/s, respectively), FEF... , FEF... , and MVV following all four
exposures. The only significant difference between the responses to the 1- and 2-h
O. protocols was in FEV.... The total number of symptoms reported was 10 for the 1-h FA
exposure, 6 for the 1-h O. exposure, and 12 for both the 2-h FA and 2-h O. exposures.
It appears that resting ventilation during the 2-h protocol had a smaller effect compared to
exercise ventilation. This supports earlier reports that the O. concentration is the most
significant factor among the three factors that contribute to effective dose (Adams et al., 1981;
Folinsbee et al., 1978; Hackney et al., 1975).
Available data, although on a limited number of subjects, consistently indicate that
responsiveness to O. is decreased in persons over 50 years of age compared to younger adults.
Although there are few data available on adults in their thirties and forties, the statistical
modeling study of McDonnell et al. (1993) on subjects from 18 to 32 years of age suggests that
responsiveness to O. is already diminishing by age 30, and that the most responsive individuals
are likely to be less than 25 years of age. The results of Bedi et al. (1988) suggest that older
adults may be less reproducible in their responses to O. than younger adult males (McDonnell
7-44
-------
et al., 1985b); however, this finding is based on only 16 subjects and should be confirmed
before being considered conclusive.
Ethnic and Racial Factors
Young white males have been the most frequently studied population in published
reports on pulmonary function responses to O.. There is concern, however, that responses to
O. may be influenced by ethnic differences based on the observation that blacks have smaller
lungs than whites for a given standing or sitting height (Rossiter and Weill, 1974; McDonnell
and Seal, 1991). Thus, an equivalent inhaled volume of O. could result in a larger O. dose per
unit of lung tissue in blacks compared to whites, potentially inducing greater effects in blacks
than whites exposed to O. under the same conditions. Seal et al. (1993) evaluated the
pulmonary function responses of 372 individuals, black, white, male, and female (n > 90 per
group), between 18 and 35 years of age who were exposed to 0.00, 0.12, 0.18, 0.24, 0.30, or
0.40 ppm O.. Each subject was assigned randomly to an exposure group and participated in
only one experimental session. The protocol involved a 2.33-h exposure to the assigned
condition. During the first 2 h of exposure, the subjects alternated 15-min rest periods and
15-min exercise periods (V. = 25 L/min/m' BSA). Spirometric and plethysmographic
measurements were made at 5 and 20 min following the final exercise period. The initial
nonparametric analysis of the percentage changes in FEV. indicated that FEV. responses
increased with increasing O. concentration, and a group effect occurred that was independent
of O. concentration. There was an O. effect, but no group effect or group X O. interaction for
SR.., indicating an increase in SR.. with increasing O. concentration. Both group and
O. effects were significant for cough, but the interaction was not significant. A post hoc
analysis, using a different statistical method on the absolute changes in FEV., indicated that the
black males experienced significant decrements in FEV. following exposure to 0.12 ppm O.,
whereas black women and white men and women did not have significant decrements in FEV.
at O. concentrations below 0.18 ppm. These results are not easily explained because there was
no gender difference among whites and no racial difference among women. Furthermore, the
black men had significantly greater decrements in FEV. at only some of the O. concentrations
studied (see Figure 7-2). Although the results can be considered suggestive of an ethnic
difference, more subjects must be studied before the issue of ethnic difference in
O. responsiveness can be more definitive. It should be noted that, although this study included
a large number of subjects, each subject participated in only one experiment. Thus, the range
of individual responsiveness could have been different between groups.
Environmental Factors
A number of environmental factors, such as ambient temperature and humidity,
season of the year, route of inhalation, and smoking history have been hypothesized to
potentially impact on responses to O. exposure in additive or synergistic ways. None of these
potentially interacting agents has been addressed adequately in the extant O. literature.
Although O. concentrations in Los Angeles, for example, generally are highest on hot, dry
days, most research on responses to O. exposure has been conducted under temperature and
humidity conditions not substantially different from those typical of indoor environments. The
few studies that included temperature and humidity as experimental factors have produced
equivocal results (see Chapter 10, Section 10.2.9 in the 1986 O. criteria document). No new
7-45
-------
reports of temperature or humidity effects have appeared since the 1986 O. criteria document
(U.S. Environmental Protection Agency, 1986).
Earlier studies discussed in the 1986 O. criteria document have suggested that
cigarette smokers are less responsive to O. than nonsmokers. Since then, the question as to
whether reactivity to O. returns with cessation of cigarette smoking has been addressed by
Emmons and Foster (1991). Thirty-four individuals with no history of asthma or obvious
respiratory disease who enrolled in a smoking cessation program were assigned randomly to an
O. group (n = 18) or an FA group (n = 16). The subjects ranged from 24 to 58 years of age
and had a group mean smoking history of 33.9 + 13 pack-years. Most of the subjects had
baseline pulmonary functions somewhat below predicted values, based on age, height, and sex.
Subjects completed 2-h exposures to 0.42 ppm O. or FA, as assigned, prior to beginning the
smoking cessation program. The subjects rested during the exposures, except for 5 min of
exercise at 150 kg D m/min (no V. given) at the beginning of the last 30 min of exposure.
Nine subjects in the O. group and six in the FA group completed the 6-mo smoking cessation
program and repeated their assigned exposures at the end of the program. Prior to beginning
the smoking cessation program, both the FA and O. groups had pre- to postexposure changes
in FVC, FEV., and FEF within the variability of repeated tests. The O. group had a
significant mean change in FEF of D22.5%, comparing post- to preexposure, whereas the
FA group had a nonsignificant Dl2% change. Changes in FVC and FEV. were not significant
in either group. It should be noted that smoking cessation led to a group mean improvement in
baseline FEF of 22.9%. The post-O. exposure values for FEF were similar
following the initial and the post-smoking cessation exposures. Thus, the difference in the
FEF decrement with O. exposure post-smoking cessation was largely due to the
improvement that ensued from 6 mo of abstinence from smoking. The results of Emmons and
Foster (1991) suggest that active smoking blunts responsiveness to O. and that cessation of
smoking for 6 mo leads to improved baseline pulmonary function and possibly the reemergence
of O. responsiveness.
7.2.1.4 Repeated Exposures to Ozone
Repeated daily exposure to O. in the laboratory setting leads to attenuated changes
in spirometry and symptom responses that were initially termed "adaptation" (Hackney et al.,
1977a). A series of repeated exposure studies, performed in various laboratories, was
reviewed in the previous criteria document (U.S. Environmental Protection Agency, 1986).
The spirometric responses to repeated O. exposure typically showed that the response was
increased on the second exposure day to concentrations in the range of 0.4 to 0.5 ppm O. in
exposures accompanied by moderate exercise (see Table 7-6). Thus, the response was
enhanced on the second consecutive day. Mechanisms for enhanced responses had not been
established, although it was hypothesized that persistence of O.-induced damage for greater
than 24 h may have contributed to the larger Day 2 response. An enhanced Day 2 response
was less obvious or absent in exposures that were repeated at lower concentrations or that
caused relatively small group mean O.-induced decrements in spirometry. Two reports (Bedi
et al., 1985; Folinsbee et al., 1986) indicated that enhanced spirometric responsiveness was
present within 12 h, lasting for at least 24 h and possibly 48 h, but was clearly absent after 72
h. After 3 to 5 days of consecutive daily exposures to O., responses were markedly
diminished or absent. One study (Horvath et al., 1981) suggested that the rapidity of this
7-46
-------
decline in response was related to the magnitude of the subjects' initial responses to O. or their
"sensitivity". Finally, the persistence of the attenuation of spirometric and symptom
7-47
-------
CO
Table 7-6. Changes in Forced Expiratory Lung Volume After Repeated
Daily Exposure to Ozone3
Ozone
Concentration*
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
0.40
0.42
0.45
0.45
0.47
0.50
0.50
Dg/m
235
392
392
392
392
490
686
686
686
686
784
784
784
823
882
882
921
980
980
Exposure Duration
3 and Activity'
6.6 h, IE (40)
2 h, IE (30)
2 h, IE (18 and 30)
2h, 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)
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)
Number and
Gender of
Subjects
17 M
10 M
8M, 13 F
9
15 M
4M, 2F
5M, 2F
10 M
8M
10 M
10 M
15 M
13 M
11 F
8M
24 M
1 M, 5F
10 M, 6F
8M, 3F
8M
6
Percent Change in FEV, on
Consecutive Exposure Days
First
D12.79
+ 1.4
D3.0
D8.7
D5.02
D20.2
D18.8
D5.3
D31.0
D16.1
D14.4
D15.9
D9.2
D8.8
D18.0
D21.1
D13.3
D5.8
Dll.4
D8.7
D2.7
Second
D8.73
+ 2.7
D4.5
DlO.l
D7.8
D34.8
—
D5.0
D41.0
D30.4
—
D24.6
D10.8
D12.9
D29.9
D26.4
—
D5.6
D22.9
D16.5
D4.9
Third
D2.54
Dl.6
Dl.l
D3.2
—
—
D22.3
D2.2
D33.0
—
D20.6
—
D5.3
D4.1
D21.1
D18.0
D22.8
Dl.9
Dll.9
D3.5
D2.4
Fourth
D0.6
—
—
—
—
—
—
—
D25.0
—
—
—
D0.7
D3.0
D7.0
D6.3
—
—
D4.3
—
D0.7
References
Fifth
0.2 Folinsbee et al. (1994)
— Folinsbee et al. (1980)
- Gliner et al. (1983)
— Gliner et al. (1983)
— Brookes et al. (1989)
— Folinsbee et al. (1986)
—
— Folinsbee et al. (1980)
— Foxcroft and Adams (1986)
— Schonfeld et al. (1989)
—
— Brookes et al. (1989)
Dl.O Kulle et al. (1982)'
Dl.6 Kulle et al. (1982)'
D4.4 Folinsbee et al. (1995)
D2.3 Horvath et al. (1981)
- Bed! et al. (1985)
— Bed! et al. (1989)
- Linnetal. (1982b)g
— Folinsbee et al. (1980)
— Hackney et al. (1977a)
'See Appendix A for abbreviations and acronyms.
''Listed from lowest to highest O3 concentration.
"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.
dFor a more complete discussion of these studies, see Table 7-7 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 lie total group of 21 individuals used
in this study.
"Bronchial reactivity to a methacholine challenge also was studied.
'Seven subjects completed entire experiment.
-------
responses has been studied (Horvath et al., 1981; Linn et al., 1982b; Kulle et al., 1982).
These studies indicate that the attenuation of response is relatively short-lived, being partially
reversed within 3 to 7 days and typically abolished within 1 to 2 weeks. Repeated exposures
separated by 1 week (for up to 6 weeks) apparently do not cause any lessening of the
spirometric response (Linn et al., 1982b).
Folinsbee et al. (1995) (also see Devlin et al., 1995) exposed a group of 15 healthy
males to 0.4 ppm O. for 2 h/day on 5 consecutive days. Subjects performed heavy IE (V. =
60 to 70 L/min, 15 min rest/15 min exercise). Decrements in FEV. averaged 18.0, 29.9,
21.1, 7.0, and 4.4% on the 5 exposure days. Baseline preexposure FEV. decreased from the
first day's preexposure measurement and was depressed by an average of about 5% on the
third day. This study illustrates that, with high-concentration and heavy-exercise exposures,
spirometry and symptom responses are not completely recovered within 24 h.
Besides the absence of pulmonary function responses after several days of
O. exposure, symptoms of cough and chest discomfort usually associated with O. exposure
generally are absent (Folinsbee et al., 1980, 1994; Linn et al., 1982b; Foxcroft and Adams,
1986). In addition, airway responsiveness to methacholine is increased with an initial
O. exposure (Holtzman et al., 1979; Folinsbee et al., 1988), may be further increased with
subsequent exposures (Folinsbee et al., 1994), and shows a tendency for the increased response
to diminish with repeated exposure (Kulle et al., 1982; Dimeo et al., 1981). A number of
possible explanations for the initially enhanced and then lessened response may be related to
changes that are occurring in pulmonary epithelia as a consequence of O. exposure.
Inflammatory responses (Keren et al., 1989a), epithelial damage, and changes in permeability
(Kehrl et al., 1987) could be invoked to explain at least a portion of these responses.
By blocking spirometric and symptom responses with indomethacin pretreatment, Schonfeld et
al. (1989) demonstrated that in the absence of an initial spirometric response such effects were
not enhanced by repeated exposure. However, the mechanisms of these responses with regard
to repeated exposures in humans remains to be elucidated.
Recent studies of repeated O. exposures have addressed some other features of the
responses (see Table 7-7). A series of reports from the Rancho Los Amigos group in
California have examined changes in response to O. as a result of the season of the year in the
South Coast Air Basin of Los Angeles, CA. The purpose of this research (Linn et al., 1988;
also Hackney et al., 1989; Avol et al., 1988) was to determine whether responsive subjects
(n = 12), identified during an initial screening following a period of low ambient O. exposure,
would remain responsive after regular ambient exposure during the "smog season". Responses
of so-called "nonresponsive" subjects (n = 13) also were examined across the year. The
subjects were exposed to 0.18 ppm O. on four occasions, spring, fall, winter, and the
following spring. Only 17 subjects (8 responders) participated in the final spring exposures.
The marked difference in FEV. response between responsive and nonresponsive subjects seen
initially (Dl2.4% versus +1%) no longer was present after the summer smog season (fall test)
or 3 to 5 mo later (winter test). However, when the reduced subset of subjects was exposed
during the following spring, the responsive subjects again had significantly larger changes in
FEV.. Seasonal changes in FEV. response to O. in the responsive and nonresponsive subjects
are shown below.
7-49
-------
Table 7-7. Pulmonary Function Effects with Repeated Exposures to Ozonfe
Ul
o
Ozone Concentration1'
ppm Dg/m
0.12 235
0.18 353
0.20/0.20 392/392
0.35/0.20 686/392
0.35/0.35 686/686
Number and
Exposure Exposure Gender of
Duration and Activity Conditions Subjects
6.6 h 18 DC 17 M
50 min exercise/ 10 min 40% RH
rest, 30 min lunch five consecutive
VE = 38.8 L/min daily exposures
2h 31 DC 59 adult
IE (heavy) 35% RH Los Angeles
VE D 60 to 70 L/min (screen exposures residents
(35 L/min/nr BSA) in spring 1986; 12 responsive
second exposures 13 nonresponsive
in summer/fall
1986 and winter
1987 and spring
1987 for
responders and
nonresponders
only)
1 h 21 to 25 DC 15 M
CE at 60 L/min 40 to 60 % RH
(three 2-day sets of
exposures)
Subject
Characteristics
Healthy NS
Responders:
Age =
19 to
40 years;
6 atopic,
2 asthmatic,
4 normal
Nonresponders:
Age =
18 to
39 years,
13 normal
Healthy
aerobically
trained NS,
FVC = 4.24 to
6.98 L
Observed Effect(s) Reference
FEV responses were maximal on first day of , Folinsbee et al. (1994)
exposure (D13%), less on second day (D9%), (also see Table 7-9)
absent thereafter. Symptom 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 ozone. Airway
responsiveness was still higher than air control
after 5 days of ozone exposure. Trend to lessened
response, but it was not achieved after 5 days.
Responders had DFEV, = 12.4% after initial Linn et al. (1988)
screening; nonresponders had no change. (^so see Hackney et al., 1989)
Responders had nonsignificant response in late
summer or early winter, but were responsive again
in early spring (spring 1986, D385 mL; Autumn
1986, D17 mL; winter 1987, +16 mL; spring
1987, D347 mL). Nonresponders did not change
with season. Suggests that responders responses
may vary with ambient exposure, but
nonresponders generally remain nonresponsive.
Consecutive days of exposure to 0.20 ppm Brookes et al. (1989)
produced similar responses on each day (05.02,
D7.80); 0.35/0.20 ppm pair caused increased
response to 0.20 ppm on second day (D8.74);
0.35/0.35 ppm caused much increased response on
Day 2 (D15.9, D24.6). Symptom responses were
worse on the second exposure to 0.35 ppm, but
not with second exposure to 0.20 ppm.
-------
Table 7-7 (cont'd). Pulmonary Function Effects with Repeated Exposures to Ozonfe
Ul
Ozone Concentration1'
ppm Dg/m Duration and Activity
0.35 686 Dl h
CE (see paper for
details)
0.35 686 60 min
CED
VE = 60 L/min
0.45 882 2 h
IE (3 x 20 min exercise)
VE = 26 L/min
0.45 882 2 h
IE (3 x 20 min exercise)
VE = 27 L/min
Exposure
Conditions
22 to 25 DC
35 to 50% RH
(1 day FA; 1 day
O3; 4 days
consecutive
exposure to O3)
21 to 25 DC
40to60%RH
(two exposures
for each subject
separated by 24,
48, 72, or 120 h)
23.3 DC
62.5 %RH
(three exposures
with a minimum
1-week interval)
23.3 DC
63% RH
Exposed for
3 consecutive
days, not exposed
for 2 days, then
exposed to
0.45 ppm again
for 1 day
Number and
Gender of Subject
Subjects Characteristics
8 M Aerobically
trained healthy
NS (some were
known O3
sensitive), age =
22.4 + 2.2 years
40 M NS; nonallergic,
(4 groups of 10) non-Los Angeles
residents for
> 6 mo; age D
25 years
8 M, 8 F Healthy NS,
61 years old for
M and 65 years
old for F
(FVC = 4.97 L
for M and
3.11 LforF)
10 M, 6 F Healthy NS,
60 to 89 years old
(median age = 65
years; mean FVC
= 3.99 L; mean
FEV, = 3.01 L;
FEV, /FVC range
= 61 to 85%)
Observed Effect(s) Reference
Largest FEV, decrease on second of 4 days Foxcroft and Adams (1986)
O3 exposure (D40% mean decrease). Trend
for adaptation not complete in 4 days.
VOillax decreased with single acute
O3 exposure (D6%) but was not significant
after 4 days of O3 exposure (D4%).
Performance time was less after acute O3
(211s) exposure than after FA (253 s).
No differences between responses to Schonfeld et al. (1989)
exposures separated by 72 or 120 h.
Enhanced FEV, response at 24 h (D 16. 1 %
vs. D30.4%). Possible enhanced response at
48 h (D14.4% vs. D20.6%). Similar trends
observed for respiratory pattern and SR4W.
Spirometric changes were not reproducible Bedi et al. (1988)
from time to time after ozone exposure
(r < 0.50). Repeat exposures to air yielded
consistent responses.
Overall increase in symptoms, but no single Bedi et al. (1989)
symptom increased significantly. FVC
decreased 111 mL and 104 mL on Days 1
and 2, respectively. FEV, 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. FEV, changes were D5.8, D5.6,
Dl.9, and Dl.7% on the four O3 days.
-------
Table 7-7 (cont'd). Pulmonary Function Effects with Repeated Exposures to Ozonfe
Ozone Concentration1'
ppm Qg/m
0.45 882
(+ 0.30
PAN)
~" Exposure
Duration and Activity
2h
IE (20 min rest, 20 min
exercise)
VE = 27 L/min
Exposure
Conditions
22 DC
60% RH
5 days
consecutive
exposure to PAN
+ O3
Number and
Gender of
Subjects
3M, 5F
Subject
Characteristics
Healthy NS,
Mean age =
24 years
Observed Effect(s)
FEV, decreased 019% with O3 alone, 015%
on Day 1 of O3 + PAN, 05% on Day 5 of
O3 + PAN, 07% 3 days after 5 days of O3
+ PAN, 015% after 5 days of O3 + PAN.
Similar to O3 adaptation studies, O3
responses peaked after 2 days, were
depressed 3 days later, and responses
returned 7 days later. PAN probably had no
effect on adaptation to O3.
Reference
Drechsler-Parks et al. (1987b)
(also see Table 7- 13)
'See Appendix A for abbreviations and acronyms.
''Listed from lowest to highest O3 concentration.
Ul
NJ
-------
DFEV. Spring DFEV. Fall DFEV. Winter DFEV. Spring
(mL) (mL) (mL) (mL)
Responders D385 D17 +16 D347
Nonresponders +28 +90 +34 +81
These results suggest a seasonal variability in response that may be attributed to
increased ambient O. exposure during the summer months. It must be noted that the
responders included subjects who had a history of complaints from ambient air pollution.
Furthermore, this group included a significant proportion of allergic individuals whose
seasonal allergies could have contributed to their varying responses. Historically, however,
studies with the subjects drawn from the population of Los Angeles have reported reduced
responses to O. exposure in the laboratory compared to nonresidents (Hackney et al., 1976,
1977b).
Brookes et al. (1989) reexamined a hypothesis previously tested by Gliner et al.
(1983), that repeated exposure to one concentration can alter response to subsequent exposure
to a different O. concentration. Gliner et al. (1983) previously had shown that the response to
0.40 ppm O. was not influenced by previously being exposed to 0.20 ppm O. for 2 h on
3 consecutive days. Brookes et al. (1989) tested whether exposure to 0.20 or 0.35 ppm
O. would change subsequent response to 0.20 or 0.35 ppm O.. They found increased
responses to 0.20 ppm for both preexposures (DFEV1 = D5.02, D7.80, and D8.74% for
0.20 ppm acutely, 0.20 ppm after 0.20 ppm, and 0.20 ppm after 0.35 ppm, respectively), but
this trend was significant only for the higher concentration. Although not statistically
significant, the response increase seen on the second exposure day at 0.20 ppm is similar to
that seem by Gliner et al. (1983). These observations suggest that, although preexposure to
low concentrations of O. may not influence response to higher concentrations, preexposure to a
high concentration of O. may significantly increase response to a lower concentration on the
following day.
Schonfeld et al. (1989) confirmed previous observations of Bedi et al. (1985) and
Folinsbee et al. (1986) that the period of enhanced responsiveness to O. following an initial
exposure persists for about 24 to 48 h but is absent by 72 h after the initial exposure. In a
series of paired exposures to 0.35 ppm with continuous heavy exercise separated by intervals
of 1, 2, 3, or 4 days, they found that the responses to the second exposure were clearly
increased at 24 h (DFEV1 = D16.1 and D30.4% for the first and second exposures,
respectively) and possibly also at 48 h (DFEV1 = D14.4 and D20.6%). Similar trends were
observed for other physiological variables such as SRaw and respiratory pattern during
exercise. With a 3- or 4-day interval between exposures, the responses to the two exposures
were similar.
Foxcroft and Adams (1986) demonstrated that decrements in exercise performance
seen after a 1-h exposure to 0.35-ppm (continuous heavy exercise) were less after
4 consecutive days of O. exposure than they were after a single acute exposure. Maximal
aerobic power and performance time on a progressive bicycle exercise test were reduced 6%
and 42 s, respectively, from FA control, after a single 0.35-ppm exposure. After
4 consecutive days of 1-h exposures, the maximal aerobic power was reduced only 4% and the
performance time by only 14 s; these differences from FA control were not statistically
7-53
-------
significant. Despite the change in exercise performance, Foxcroft and Adams (1986) did not
show the attenuation of FEV. response seen in many previous studies (Folinsbee et al., 1980;
Linn et al., 1982b). However, these investigators selected known O.-sensitive subjects whose
FEV. decrements exceeded 30% on the first 3 days of exposure. The large magnitude of these
responses, the trend for the responses to decrease on the third and fourth day, the decreased
symptom responses, and the observations of Horvath et al. (1981) that O.-sensitive subjects
adapt slowly, suggest that attenuation of response would have occurred if the exposure series
had been continued for another 1 or 2 days. These observations support the contention
advanced by Horvath et al. (1981) that the progression of attenuation of response is a function
of "O. sensitivity". Furthermore, these results suggest that exercise responses after
O. exposure may be limited, either voluntarily or involuntarily, more by subjective symptoms
than by alterations in gas exchange consequent to changes in ventilatory function.
Bedi et al. (1989) examined the responses of elderly subjects (median age, 65 years)
to four exposures to 0.45 ppm O. for 2 h with mild IE. The first three exposures were on
consecutive days, with the fourth exposure following the third by 3 days. Changes in FEV. on
the first two exposure days averaged D5.8 and D5.6%, about half the response expected in a
group of healthy young males (Dl2.7%; Folinsbee et al., 1978). There were no significant
changes in FEV. on the third (Dl.9%) and fourth (Dl.7%) exposure days. Symptom responses
were negligible, although there was an overall increase in symptoms on the first day of
O. exposure compared to air exposure. Despite the high concentration of the exposure, there
was no enhancement of the spirometry response on the second day of exposure. Although
similar observations have been made in previous studies producing small changes in spirometry
(Folinsbee et al., 1980, 1994) with repeated exposures, the responses of older subjects are not
sufficiently understood to explain these responses. Bedi et al. (1988) had previously reported
that responses to O. in the older subjects tended to be less reproducible, although this factor
alone could not explain these responses.
Drechsler-Parks et al. (1987b) examined the response to repeated exposures to
0.45 ppm O. plus 0.30 ppm peroxyacetyl nitrate (PAN). Exposures to O. and O. plus PAN
yielded similar changes in spirometry (DFEV. = D19 and Dl5%, respectively). Thus, PAN did
not increase responses to O.. Repeated exposure to the PAN plus O. mixture resulted in
similar changes to those seen with O. exposure alone. Responses in FEV. exceeded D30% on
the second exposure and fell to less than D5% after the fifth day. The attenuation of response
persisted 3 days after the repeated exposures, but was absent after 7 days. These observations
suggest that PAN does not influence the attenuation of response to repeated O. exposure. If
the PAN responses are considered negligible, this study confirms the observation that the
attenuation of O. responses with chamber exposures lasts no longer than 1 week.
Repeated multihour exposure to low concentrations of O. has been examined
(Horvath et al., 1991; Folinsbee et al., 1994; Linn et al., 1994). Horvath et al. (1991)
exposed subjects for 2 consecutive days to 0.08 ppm using the 6.6-h prolonged-exposure
protocol (see Section 7.2.2). They observed small pre- to postexposure changes in FEV.
(D2.5%) on the first exposure, but no change on the second day. Linn et al. (1994) observed a
1.7% decrease in FEV. in healthy subjects after a 6.5-h exposure to 0.12 ppm. A second
consecutive exposure yielded even smaller (< 1 %) responses. With exposure to a mixture of
O. plus 100 Dg/m' of H.SO. aerosol, there was a 4.2% decrease in FEV. on the first exposure
day. In a group of asthmatics exposed under similar conditions, the FEV. response on the first
day was D8.6% (O.) and Dll.6% (O. plus acid). After adjustment for the exercise effect
7-54
-------
(D4.6%), the responses (D4 and D7%) were still greater than those of nonasthmatics.
Responses were slightly reduced on the second day of exposure.
Folinsbee et al. (1994) exposed 17 subjects to 0.12 ppm O. for 6.6 h on
5 consecutive days. Spirometry responses were typified by changes in FEV. that reached
Dl3% on the first day and D9% on the second day of exposure. No significant differences in
spirometry responses between FA and subsequent O. exposures were observed. Symptom
responses were also greatest on the first exposure day and were largely absent from the third
day on. Methacholine responsiveness was tested using a single dose of methacholine and then
by comparing changes in R.. as the ratio of SR.. after methacholine aerosol to that after saline
aerosol. The responses to FEV. and methacholine testing are shown below.
Day 1 Day 2 Day 3 Day 4 Day 5 Clean Air
D%FEV.
SR.. Ratio
D12.79
3.67
D8.73
4.55
D2.54
3.99
DO. 6
3.24
+0.2
3.74
+ 1.1
2.22
Methacholine responsiveness was increased (over the clean air response) throughout
the 5 days of O. exposure, although it reached a peak on the second day, and, in some
subjects, there was a trend for responsiveness to decrease after 5 days. These results suggest
that repeated exposure to low levels of O., despite the attenuation of symptoms and pulmonary
function changes, is not without hazard. It is likely that some epithelial damage persists that
contributes to the enhanced response to methacholine throughout the exposure series.
However, it must be noted that, in this study, subjects initially were selected based on their
FEV. response to 0.16 ppm O. for 4 h. This may in part explain the greater FEV. responses
seen in this study, but there was no correlation between individual FEV. decrements and
changes in methacholine responsiveness. Furthermore, the Horvath et al. (1991) subjects were
exposed only to 0.08 ppm, and they were somewhat older than the Folinsbee et al. (1994)
subjects; the Linn et al. (1994) subjects, on the other hand, had lower ventilation during
exercise and were residents of Los Angeles accustomed to exposure to these levels of O. (see
Chapter 4 for typical O. concentrations).
Based on studies cited here and in the previous criteria document (U.S.
Environmental Protection Agency, 1986), several conclusions can be drawn about repeated
1- to 2-h O. exposures. Repeated exposures to O. can cause an enhanced (i.e., greater)
response on the second day of exposure. This enhancement appears to be dependent on the
interval between the exposures (24 h causes the greatest increase) and is absent with intervals
D3 days. An enhanced response also appears to depend to some extent on the magnitude of the
initial response. Small responses to the first O. exposure are less likely to result in an
enhanced response on the second day of O. exposure. Repeated daily exposure also results in
attenuation of spirometric responses, typically after 3 to 5 days of exposure. This attenuated
response persists for less than 1 or as long as 2 weeks. In temporal conjunction with the
spirometry changes, symptoms induced by O., such as cough and chest discomfort, also are
attenuated with repeated exposure. Ozone-induced changes in airway responsiveness attenuate
more slowly than spirometric and symptom responses. Attenuation of the changes in airway
responsiveness also persist longer than changes in spirometry, although this has been studied
only on a limited basis. In longer-duration, lower-concentration studies that do not cause an
7-55
-------
enhanced second-day response, the attenuation of response to O. appears to proceed more
rapidly.
7.2.1.5 Effects on Exercise Performance
Introduction
An early epidemiological study examining race performances in high school cross-
country runners (Wayne et al., 1967) suggested that exercise performance is depressed by
inhalation of ambient oxidant air pollutants. Wayne et al. (1967) suggested that the detrimental
effects of oxidant air pollutants on race performance may have been related to increased R.. or
to the associated discomfort in breathing, thus limiting runners' motivation to perform at high
levels. The effects of acute O. inhalation on exercise performance have been evaluated in
numerous controlled human studies. These studies can be divided into two categories: (1)
those that examine the effects of acute O. inhalation on maximal oxygen uptake and (2) those
that examine the effects of acute O. inhalation on the ability to complete strenuous continuous
exercise protocols up to 1 h in duration. Five studies (Folinsbee et al., 1977; Horvath et al.,
1979; Folinsbee et al., 1984; Adams and Schelegle, 1983; Savin and Adams, 1979) examining
the effects of acute O. exposures on exercise performance were discussed in the 1986 EPA
criteria document (U.S. Environmental Protection Agency, 1986). This section summarizes
the studies reviewed in that document and reviews more recent studies that examine the effect
of acute O. inhalation on maximal oxygen uptake and endurance performance. Studies are also
summarized in Table 7-8.
Effect on Maximal Oxygen Uptake
Three studies (Folinsbee et al., 1977; Horvath et al., 1979; Savin and Adams,
1979) examining the effects of acute O. exposures on VO were discussed in the 1986 EPA
criteria document (U.S. Environmental Protection Agency, 1986). Of these studies, only
Folinsbee et al. (1977) observed that VO.... was significantly decreased (10.5%) following a
2-h exposure to 0.75 ppm O. with light IE. Reductions in VO.... were accompanied by a
9.5% decrease in maximum attained workload, a 16% decrease in maximum ventilation, and a
6% decrease in maximum heart rate. The 16% decrease in maximum ventilation was
associated with a 21 % decrease in V.. In addition, the O. exposure resulted in a
22.3% decrease in FEV. and subjective symptoms of cough and chest discomfort. In contrast,
Horvath et al. (1979) did not observe a change in VO or other maximum cardiopulmonary
endpoints in male and female subjects exposed at rest to 0.75 ppm O. for 2 h, although FVC
was significantly decreased (10%). Similarly, Savin and Adams (1979) observed no effect on
maximum attained workload or VO.... in nine subjects exposed to 0.3 ppm O. while
performing a progressively incremented exercise test to volitional fatigue lasting 30 min.
In addition, Savin and Adams (1979) observed no significant effect on pulmonary function,
performance time, maximum heart rate, or anaerobic threshold, although maximum ventilation
was significantly reduced 7%.
More recent findings of Foxcroft and Adams (1986) and Gong et al. (1986) support
the earlier observations of Folinsbee et al. (1977). Foxcroft and Adams (1986) observed
significant (p < 0.05) reductions in performance time (16.7%), VO.... (6.0%), maximum
ventilation (15.0%), and maximum heart rate (5.6%) in eight aerobically trained males during
a rapidly incremented VO test following 50-min exposure to 0.35 ppm O. with CE
(exercise V. = 60 L/min). Similarly, Gong et al. (1986) found significant
7-56
-------
Table 7-8. Ozone Effects on Exercise Performance1
Ul
Ozone
Concentration b
ppm
0.06-
0.07
0.12-
0.13
0.12
0.18
0.24
0.12
0.20
0.20
0.35
0.21
0.25
0.25
0.50
0.75
0.35
Dg/m
120-
140
245-
260
235
353
470
235
392
392
686
412
490
490
980
1,470
686
Exposure
Duration and
Activity
CE(VE = 30 to
120 L/min) 16 to
28min
progressive
maximum exercise
protocol
1 h competitive
simulation
exposures at mean
VE = 87 L/min
1 hCE
VE = 89 L/min
1 h CE or
competitive
simulation at mean
VE = 77.5 L/min
1 hCEat75%
VCXmax
IhCE
VE = 63 L/min
2 h rest
SOminCE V =
60 L/min
Exposure
Conditions
Tdb = 23 to 24.5 DC
RH = 50 to 53%
Tdb = 23 to 26 DC
RH = 45 to 60%
Tdb = 31 DC
Tdb = 23 to 26 DC
RH = 45 to 60%
Tdb = 19 to 21 DC
RH = 60 to 70%
Tdb = 20 DC
RH = 70%
NA
NA 'E
Number
and Subject
Gender of Subjects Characteristics
12 M, 12 F Athletic
10 M Highly trained
competitive
cyclists
15 M, 2 F Highly trained
competitive
cyclists
10 M Well-trained
distance runners
6 M, 1 F Well-trained
cyclists
19 M, 7 F Active
nonathletes
8M, 5F
8 M Trained
nonathletes
Observed Effect(s)
Reduced maximum performance time and increased
respiratory symptoms during O3 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 FEV, for 0.18- and 0.24-ppm
O3 exposure compared with FA exposure.
Decrease in VE,,,^, YO,,,,^, VTn]ax, workload, ride
time, FVC, and FEV, with 0.20 ppm O3 exposure,
but not significant with 0. 12-ppm O3 exposure, as
compared to FA exposure.
VT decreased and f increased with continuous 50-min
O3 exposures; decrease in FVC, FEV,, and FEFS_7J,5
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 protocol at
0.35 ppm O3.
Decrease in FVC, FEV,, FEF^j,,, and MVV with
0.21 ppm O3 compared with FA exposure.
FVC, FEV,, and MVV all decreased with
0.25-ppm O3 exposure compared with FA.
FVC decreased with 0.50- and 0.75-ppm O exposure
compared with FA; 4% nonsignificant decrease in
mean VOilla}1 following 0.75 ppm O3 compared with
FA exposure.
VT decreased, f increased with 50-min O3 exposures;
decrease in FVC, FEV,, FEF^j,., performance time,
VOamx, YElmx, and HR,m from FA to 0.35-ppm O3
exposure.
Reference
Linder et al. (1988)
Schelegle and Adams (1986)
Gong et al. (1986)
Adams and Schelegle (1983)
Folinsbee et al. (1984)
Folinsbee et al. (1986)
Horvath et al. (1979)
Foxcroft and Adams (1986)
-------
7-58
-------
Table 7-8 (cont'd). Ozone Effects on Exercise Performance1
ppm
0.75
Ozone
Concentration b
Dg/m
1,470
— Exposure
Duration and
Activity
2hIE
(4x15 min light
[50 W] bicycle
ergometry)
Exposure
Conditions
NA
Number
and
Gender of Subjects
13 M
Subject
Characteristics
4 light S,
9NS
Observed Effect(s)
Decrease in FVC, FEV,, ERV, 1C, and FEE^,, after
1-h 0.75-ppm O3 exposure; decrease in VQ,lmx,
VTmx, YEmx, maximal workload, and heart rate
following 0.75-ppm O3 exposure compared with FA.
Reference
Folinsbee et al. (1977)
"See Appendix A for abbreviations and acronyms.
''Listed from lowest to highest O3 concentration.
Ul
-------
reductions in performance time (29.7%), VO.... (16.4%), maximum ventilation (18.5%), and
maximum workload (7.8%) in 17 top-caliber endurance cyclists during a rapidly incremented
VO test following 1-h exposure to 0.2 ppm O. with very heavy CE (V. =90 L/min) and
the addition of ambient heat stress (31 DC). In both studies (Foxcroft and Adams, 1986; Gong
et al., 1986), the reductions in maximal exercise endpoints were accompanied by significant
decrements in pulmonary function and marked subjective symptoms of respiratory discomfort.
More recently, Linder et al. (1988) observed small decrements in performance time during a
progressive maximal exercise test at O. concentrations as low as 0.06 ppm. These small
effects were associated with increased respiratory symptoms and small, inconsistent changes in
FEV.. Hence, it appears that maximal oxygen uptake is reduced if it is preceded by an
O. exposure entailing a sufficient total inhaled dose of O. to result in significant pulmonary
function decrements or subjective symptoms of respiratory discomfort.
Effect on Endurance Exercise Performance
Two studies (Adams and Schelegle, 1983; Folinsbee et al., 1984) that addressed the
effects of acute O. exposures on the ability of highly trained subjects to complete strenuous
continuous exercise protocols were discussed in the 1986 EPA criteria document (U.S.
Environmental Protection Agency, 1986).
Adams and Schelegle (1983) exposed 10 well-trained distance runners to FA and
0.20 and 0.35 ppm O. while the runners exercised on a bicycle ergometer at workloads
simulating either a 1-h steady-state "training" bout or a 30-min warm-up followed immediately
by a 30-min "competitive bout". The exercise levels in the steady-state training bout were of
sufficient magnitude (68% of their VO....) to increase mean V. to 80 L/min. The V.
averaged over the entire competitive simulation was also 80 L/min, whereas the mean V.
during the 30-min competitive bout was 105 L/min. Subjective symptoms increased as a
function of O. concentration for both training and competitive protocols. In the competitive
protocol, four runners exposed to 0.20 ppm O. and nine exposed to 0.35 ppm O. indicated that
they could not have performed maximally. Three subjects were unable to complete both the
training and competitive protocols at 0.35 ppm O., and a fourth failed to complete only the
competitive ride.
Folinsbee et al. (1984) exposed six well-trained men and one well-trained woman to
0.21 ppm O. while they exercised continuously on a bicycle ergometer for 1 h at 75% of their
VO (V. =81 L/min). Following O. exposure, FVC and FEV. were reduced significantly
and the subjects reported symptoms of laryngeal and tracheal irritation and chest soreness and
tightness when taking deep breaths. Anecdotal reports obtained from the cyclists suggested
that their performance would have been limited if they experienced similar symptoms during
competition.
Avol et al. (1984) exposed 50 well-conditioned cyclists to 0.00, 0.08, 0.16, 0.24,
and 0.32 ppm O. for 1 h in ambient heat (32 DC) while they exercised continuously
(V. =57 L/min). Reductions in FEV. and symptoms, initially detected at 0.16 ppm O.,
increased in a concentration- dependent manner. Three and 16 cyclists could not complete the
1-h exposure to 0.16 and 0.24 ppm O., respectively, without a reduction in workload.
Similarly, in their study of the effects of O. exposure on VO...., Gong et al. (1986) reported
that 6 of 17 highly trained endurance cyclists were not able to complete 1-h exposure to 0.2
ppm O. with very heavy CE (V. =90 L/min) and the addition of ambient heat stress (31 DC).
7-60
-------
In a study designed to determine the effects of the inhalation of low ambient
O. concentrations on simulated competitive endurance performance, Schelegle and Adams
(1986) exposed 10 highly trained endurance athletes to 0.12, 0.18, and 0.24 ppm O. while they
were performing a 1-h "competitive" protocol. The competitive protocol used in this study
was similar to that used by Adams and Schelegle (1983) except that the workload during the
final 30-min competitive bout was more intense; it was selected based on the maximum
workload (approximately 86% of their VO , mean V. = 120 L/min) each subject could
maintain for 30 min while breathing FA. All subjects completed the FA exposure, whereas
one, five, and seven subjects could not complete the 0.12-, 0.18-, and 0.24-ppm O. exposures,
respectively. Following 0.18- and 0.24-ppm O. exposures, FVC and FEV. were reduced
significantly (p < 0.05), and subjective symptoms were elevated significantly (p < 0.05). No
significant effect of O. was found for metabolic or ventilatory pattern responses. Similarly,
Folinsbee et al. (1986) found that highly trained runners experienced a reduced run time on a
treadmill (speed and grade set at approximately 80% of their subjects VO....) when exposed to
0.18 ppm O. compared with FA. These subjects did have significantly elevated symptoms of
respiratory discomfort and significantly decreased FVC and FEV., whereas arterial oxygen
saturation at the end of the run was not affected by O. exposure.
Determining the mechanisms leading to the observed decrements in maximal oxygen
uptake and the inability to complete strenuous exercise protocols is problematic. As stated by
Astrand and Rodahl (1977) "the capacity for prolonged rhythmic muscular exercise is limited
by an interrelated composite of cardiorespiratory, metabolic, environmental, and psychological
factors." Many investigators cited above have concluded that the observed reductions in
exercise performance appeared to be due to symptoms limiting the ability of their subjects to
perform. However, in every case, this is a conclusion achieved by exclusion and not by the
demonstration of a causal relationship. Other factors could also contribute to O.-induced
decrements in exercise performance. One possibility is that stimulation of neural receptors in
the airways may result in an inhibition of alpha-motor nerve activity to respiratory muscles
during inspiration (Koepchen et al., 1977; Schmidt and Wellhoner, 1970), resulting in the
observed decrease in V. and, at the same time, increasing the subject's sensation of respiratory
effort. This mechanism would not be directly related to symptoms of discomfort but, because
of the common role of airway neural afferents, may be difficult to discern from the effects of
symptoms of respiratory discomfort. Indeed, a reflex inhibition of the ability to inspire would
be consistent with the reduced V. following O. exposure in subjects performing maximal
exercise and would be consistent with the development of a physiologically induced ventilatory
limitation to maximal oxygen uptake.
7.2.2 Pulmonary Function Effects of Prolonged (Multihour) Ozone
Exposures
Since 1988, a series of studies has described the responses of subjects exposed to
relatively low (0.08 to 0.16 ppm) O. concentrations for durations of 4 to 8 h (see Table 7-9).
These studies have demonstrated statistically significant changes in spirometry, R.., symptoms,
and airway responsiveness during and after exposures. As in studies conducted at higher
concentrations of O. for shorter periods of time, there is broad variability in response.
The only related study cited in the previous criteria document (U.S. Environmental
Protection Agency, 1986) was that of Kerr et al. (1975), who exposed
7-61
-------
Table 7-9. Pulmonary Function Effects After Prolonged Exposures to Ozonfe
Ozone Concentration1'
ppm
0.08
0.10
0.12
0.08
0.10
0.08
vj
(^
NJ 0.12
(a) 0.12
(b) Varied from
0.0 to 0.24
(increased by
0.06 ppm/h then
decreased by
0.06 ppm/h)
Dg/m
157
196
235
157
196
157
235
235
Exposure
3 Duration and Activity
6.6 h
IE (6 x 50 min)
VE 0 39 L/min
See Horstman et al. (1990) and
Folinsbee et al. (1988)
6.6 h
IE (6 x 50 min)
VE = 40 L/min
6.6 h
IE (6 x 50 min)
VE = 35 to 38 L/min
(1 day of air, 2 days of O3)
6.6 h
IE (6 x 50 min)
VE = 42.6 L/min
8h
IE (8 x 30 min)
VE = 40 L/min
Exposure
Conditions
18 DC
40% RH
18 DC
40% RH
25 DC
48%RH
18 DC
40% RH
(1 exposure to
clean air;
1 exposure to O3)
22 DC
40% RH
< 3 Dg/m3 TSP
Number and
Gender of Subject
Subjects Characteristics
22 M Healthy NS, 18 to
33 years old
38 M Healthy NS,
mean age 25 years
old
5 F, 6 M Healthy NS, 30 to
45 years old
10 M Healthy NS, 18 to
33 years old
23 M Healthy NS, 20 to
35 years old
Observed Effect(s)
FVC and FEV, decreased throughout the
exposure; FEV, decrease at end exposure was
7.0, 7.0, and 12.3%, respectively. FEV,
change > 15% occurred in 3, 5, and 9 subjects
at 0.08, 0.10, and 0.12 ppm, respectively.
Methacholine responsiveness increased by 56,
89, and 121%, respectively.
A lognormal model was fitted to FEV, data.
Model parameters indicate O3 concentration
had greater effect than VE or duration
(estimated exponent for [O3] D 4/3).
FEV,, decreased 8.4% at 0.08 ppm and
1 1 .4% at 0. 10 ppm. Symptoms of cough,
PDI, and SB increased with O3 exposure.
FVC decreased 2.1%, FEV, decreased 2.2%
on first day of O3 exposure; no change on
second O3 day.
FEV, decreased by 13% after 6.6 h. FVC
dropped 8.3%. Cough and PDI increased with
O3 exposure. Airway responsiveness to
methacholine doubled after O3 exposure.
(a) FEV, decreased 5% by 6 h and remained at
this level through 8 h.
(b) FEV, change mirrored O3 concentration
change with a lag time of D 2 h. Max decrease
of 10.2% after 6 h. FEV, change was reduced
in last 2 h of exposure.
Reference
Horstman et al. (1990)
Larsen et al. (1991)
McDonnell et al. (1991)
Horvath et al. (1991)
Folinsbee et al. (1988)
Hazucha et al. (1992)
-------
Table 7-9 (cont'd). Pulmonary Function Effects After Prolonged Exposures to Ozonfe
Ozone Concentration1'
ppm Dg/m
0.12 235
0.12 235
0.16 314
Exposure
3 Duration and Activity
6.5 h/day
IE (6 x 50 min)
(2 days of exposure)
VE = 28 L/min (asthmatic)
VE = 31 L/min (healthy)
6.6 h
IE (6 x 50 min)
VE = 38.8 L/min
4h
IE (4 x 50 min)
L/min
VE D 38.9
Exposure
Conditions
21 DC
50% RH
18 DC
40% RH
(5 consecutive
days of exposure
to O3, 1 day
exposure to CA)
18 DC
40%RH(one
exposure to O3, no
control exposure)
Number and
Gender of
Subjects
15
(8 M, 7 F)
30
(13 M, 17 F)
17 M
15 M
Subject
Characteristics
Healthy NS,
22 to 41 years
old
Asthmatic NS,
18 to 50 years
old
Healthy NS,
mean age 25 +
4 years old
Healthy NS,
mean age 25 ±
4 years old
Observed Effect(s)
Bronchial reactivity to methacholine increased with
O3 exposure in healthy subjects. FEV, decreased 2%
(pre- to postexposure) in healthy subjects and 7.8% in
asthmatics. Responses were generally less on the
second day. Two healthy subjects and four asthmatics
had FEV, decreases > 10%.
FEV, decreased by 12.8, 8.7, 2.5, and 0.6 and
increased by 0.2 on Days 1 to 5 of O3 exposure,
respectively. Methacholine airway responsiveness
increased by > 100% on all exposure days. Symptoms
increased on the first O3 day, but were absent on the
last 3 exposure days.
FVC decreased 9.5% and FEV, decreased 16.6%.
FEV,/FVC ratio decreased from 0.79 to 0.73.
Reference
Linn et al. (1994)
Folinsbee et al. (1994)
U>
'See Appendix A for abbreviations and acronyms.
''Listed from lowest to highest O3 concentration.
-------
subjects for 6 h to 0.5 ppm O., with only two brief 15-min periods of moderate exercise (V. =
44 L/min) during the exposure. Small changes in spirometry were observed. Because of the
minimal extent of exercise and the high O. concentration, these results cannot be compared to
the more recent studies.
The first prolonged O. exposure study involving low concentrations and a
substantial amount of "moderate exercise'" was reported by Folinsbee et al. (1988). The basic
protocol used by these investigators has been used in a number of subsequent investigations
and therefore merits describing in some detail. The exposures lasted 6 h and 35 min (D6.6 h).
Except for a 35-min lunch break (during which O. exposure continued at rest) after 3 h, the
subjects exercised at a moderate level (with a ventilation of about 40 L/min) for 50 min of each
hour. Pulmonary function tests were conducted during the 10-min rest period and at the
beginning and end of exposure. The exposure was intended to simulate a day of heavy outdoor
work or play. For convenience, this protocol is referred to as the EPA prolonged-exposure
protocol.
In this study (Folinsbee et al., 1988), a group of 10 subjects was exposed to clean
air and 0.12 ppm O. for 6.6 h. Forced vital capacity and FEV. decreased in a roughly linear
fashion throughout the exposure and had fallen by 8.3 and 13%, respectively, by the end of the
exposure. Symptoms of cough and chest discomfort were increased, and airway
responsiveness to methacholine was approximately doubled after O. exposure. There was
a wide range of response, three subjects had FEV. decrements of 25% or greater, and the three
least sensitive subjects had less than 5 % change in FEV..
In order to extend these initial observations, Horstman et al. (1990) used the same
protocol to expose a group (n = 22) of subjects to clean air and three different
O. concentrations (0.08, 0.10, and 0.12 ppm). At 0.12 ppm O., responses were similar to
those observed in the previous study, with the exception that the symptom responses were
smaller in the new group of subjects. A similar (but of smaller magnitude) pattern of response
in spirometry, R.., and airway responsiveness was seen at the two lower concentrations. The
mean FEV. responses during the four exposures are shown in Figure 7-3. The responses were
dependent on concentration and exposure duration (ventilation was not varied) and averaged 7,
8, and 13% at the three O. concentrations. Larsen et al. (1991) used these data (Horstman
et al., 1990) to develop a "dose-response" relationship for percent change in FEV. as a
function of O. concentration and exposure duration. The lognormal multiple linear regression
model suggested that FEV. responses were approximately linear with duration of exposure but
that O. concentration plays a slightly more important role. The exponent of approximately 4/3
suggests that doubling O. concentration would be similar to increasing exposure duration by
about 2" times.
A series of additional exposures were conducted at 0.08 and 0.10 ppm O. to study
changes in cells and inflammatory mediators from BAL (see Section 7.2.4), but pulmonary
function was measured as well. McDonnell et al. (1991) reported an 8.4% decrease in FEV.
at 0.08 ppm and an 11.4% decrease at 0.10 ppm. These responses were slightly larger than
those seen in the previous Horstman et al. (1990) study. The duration-FEV. response data
were fit to a three-parameter logistic model, which significantly improved the amount of
'The "moderate" exercise descriptor is based on previously published EPA guidelines for representative types of
exercise (see Table 10-3, U.S. Environmental Protection Agency, 1986). Note, however, that exercise continued
at this level (40 L/min) for 6 to 8 h should be considered as "heavy" or "strenuous work or play".
7-64
-------
4,500
4,400
4,300
^ 4,200
5T4.100
LU
u_
4,000
3,900
3,800
0
23456
Exposure Duration (h)
Figure 7-3. The forced expiratory volume in 1 s (FEVJ is shown in relation to exposure
duration at different ozone concentrations. A 35-min resting exposure
period was interposed between the end of the third hour and the beginning
of the fourth hour. There were six 50-min exercise periods (minute
ventilation HI39 L/min) during the exposure; these measurements were made
5 min after the end of each exercise. The total exposure duration was 6.6 h.
The standard error of the mean (not shown) for these FEVt averages ranged
from 120 to 160 ml.
Source: Horstman et al. (1990).
variance explained by the model compared to a linear model; this is consistent with exploratory
analyses in the Folinsbee et al. (1988) report. The reasonably good fit to the logistic model
suggests that the O.-pulmonary function response relationship may have a sigmoid shape. The
primary importance of this observation is that it suggests that there is a response plateau. That
is, for a given O. concentration and exercise ventilation level (i.e., dose rate), and after a
certain length of exposure, the FEV. response tends not to increase further (i.e., plateau) with
increasing duration of exposure.
7-65
-------
In the fourth study in this series (Folinsbee et al., 1994), 17 subjects were exposed
to 0.12 ppm O. for 6.6 h on 5 consecutive days. Subjects who were not responsive to O. were
not selected to participate in this study. Responses in FEV. on the first of these exposures
averaged Dl2.8%. Again, symptom responses were modest with a significant increase in
lower respiratory symptoms on the first exposure day. A significant increase in airway
responsiveness to methacholine also was shown. The response to the repeated exposures is
discussed in Section 7.2.1.4. In addition, 15 subjects were exposed to 0.16 ppm for 4 h using
the same hourly exposure protocol as described above. In these subjects, FVC decreased
9.5%, and FEV. declined 16.6%.
Folinsbee et al. (1991) took the FEV. response data from all four studies conducted
at the EPA Health Effects Research Laboratory, using the same prolonged-exposure protocol,
and examined the distribution of responses among the subjects at the three concentrations.
This response distribution is illustrated graphically in Figure 7-4, which illustrates that FEV.
decrements as large as 30 to 50% have been observed with prolonged exposure to
O. concentrations DO. 12 ppm. This response distribution allows one to determine the number
or percentage of subjects with responses in excess of a certain level. The proportion of
subjects with an FEV. decrease in excess of 10% is shown in Figure 7-4. With air exposure,
no one exceeded this response level; however, 46% of the subjects exposed to 0.12 ppm
O. had a > 10% drop in FEV. after 6.6 h.
Distribution of Percent Change in FEVi
15
10
5
-5
-™
{! -20
-25
g -30
• -35
-40
-45
n = 87
AIR
0%
n = 60
0.08 ppm O3
26%
n = 32
0.10 ppm
31%
n = 49
0.12 ppm O3
46%
S 10 15 20 25 30 35 40 0 5 10 15 20 25 30 35 40 0 5 10 15 20 25 30 35 40 0 5 10 15 20 25 30 35 40
Percent of Subjects
Figure 7-4. The distribution of response for 87 subjects exposed to clean air and at least
one of 0.08, 0.10, or 0.12 ppm ozone (O3) is shown here. The O3 exposures
lasted 6.6 h, during which time the subjects exercised for 50 min of each
hour with a 35-min rest period at the end of the third hour. Decreases in
forced expiratory volume in 1 s (FEVt) are expressed as percent change from
baseline. For example, the bar labeled "DlO" indicates the percent of
subjects with a decrease in FEV, of >5% but DlO%, and the bar labeled "5"
indicates improvement in FEV, of > 0% but El5%. Each panel of the figure
indicates the percentage of subjects at each O3 concentration with a
decrease of FEVt in excess of 10%.
7-66
-------
This response distribution also illustrates the wide range of response to O. under
these exposure conditions and reinforces the observation by others (McDonnell et al., 1983;
Horvath et al., 1981) of a substantial range of individual response to O..
Horvath et al. (1991) examined the responses of healthy men and women (ages
30 to 45) to 0.08 ppm O. for 6.6 h using the EPA prolonged-exposure protocol. When
compared with the clean air exposure, FEV. decreased about 5% with O. exposure. However,
the variability of this response among this small heterogeneous group of subjects was enough
to preclude statistical significance of this observation. No significant changes were observed
with a second exposure on the next day. On the first day of O. exposure, 7 of the 11 subjects
reported chest tightness. The authors point out that the range of variability in response in their
study was similar to that reported by Folinsbee et al. (1988) and Horstman et al. (1990),
although fewer subjects experienced large negative changes in FEV.. One possible explanation
for the differences between the findings of Horvath et al. (1991) and Horstman et al. (1990)
may be that the subjects in Horvath's study were significantly older, which may result in
reduced responses to O., as Drechsler-Parks et al. (1987a) have shown (see Section 7.2.1.3).
The ventilation during exercise (37 to 39 L/min) was similar to that reported by Horstman
et al. (1990). An additional FEV. measurement was made in this study at the end of the lunch
period (i.e., after 40 min of rest). At this time, the small decrements in FEV. seen after the
third exercise were reversed, and the FEV. was similar to the response in FA at the same time
point. Although spirometry was not measured at this time in the other prolonged-exposure
studies (Folinsbee et al., 1988; Horstman et al., 1990), it was noted that the decline in FEV.
was attenuated between the third and fourth postexercise measurement. These observations
suggest that the subjects' lung function may indeed have improved during the lunch rest
period.
Linn et al. (1994) have reported responses of 45 healthy and asthmatic subjects to
0.12 ppm O. using the EPA prolonged-exposure protocol. In healthy subjects, they observed a
small (1.7%) decrease in FEV., which was statistically significant, and an increase in airway
responsiveness to methacholine. The functional responses in asthmatics (e.g., a 7.8% decrease
in FEV.) were greater than those of the healthy subjects. They observed smaller responses on
a second consecutive day of exposure, as did Horvath et al. (1991) and Folinsbee et al. (1994).
The ventilation averaged 31 and 28 L/min in the healthy and asthmatic subjects, respectively.
The FEV. responses observed in this study, although statistically significant, are much lower
than those observed by EPA investigators (Folinsbee et al., 1988; Horstman et al., 1990;
Folinsbee et al., 1994). The smaller responses may be due to previous ambient exposures,
lower ventilations, or a larger proportion of O.-insensitive subjects in Los Angeles. Only 1 of
15 healthy subjects experienced an FEV. decrement in excess of 10%, whereas 9 of 30
asthmatics had FEV. decrements in excess of 10%. Asthmatic responses ranged from 12% to
D35%.
To further explore the factors that determine responsiveness to O., Hazucha et al.
(1992) designed a protocol to examine the effect of varying, rather than constant,
O. concentrations. In this study, subjects were exposed to a constant level of 0.12 ppm O. for
8 h and to an O. level that increased linearly from 0 to 0.24 ppm for the first 4 h and then
decreased linearly from 0.24 to 0 over the second 4 h of the 8 h exposure (triangular
concentration profile). Subjects performed moderate exercise for the first 30 min of each hour.
The overall exposure dose for these two exposures, calculated as the C x T x V., was almost
identical (difference < 1 %). With exposure to the constant 0.12 ppm O., the FEV. declined
7-67
-------
approximately 5 % by the fifth hour of exposure and remained at that level for the remainder of
the exposure. These responses are illustrated in Figure 7-5. This observation clearly indicates
a response plateau, suggested in other studies (Horstman et al., 1990), with an exposure
regimen that produces relatively small changes in lung function.
7-68
-------
Steady Versus Variable Ozone Concentration
>
LU
CO
CD
4.5
4.4
4.3
4.2
4.1
3.9
Air
_m_ 0.12-ppm
Constant
Variable
0246
Exposure Duration (h)
8
Figure 7-5. The forced expiratory volume in 1 s (FEVt) is shown in relation to exposure
duration (hours) under three exposure conditions. Subjects exercised
(minute ventilation ff40 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 FEV, averages (not shown) ranged from 120 to 150 ml.
Source: Hazucha et al. (1992).
With the triangular O. concentration profile, the FEV. decreased almost twice as
much after 6 h of exposure. The initial response over the first 3 h was minimal, and then there
was a substantial decrease in FEV., corresponding to the higher average O. concentration, that
reached a nadir after 6 h. Despite continued exposure to a lower O. concentration
(< 0.12 ppm), the FEV. began to improve and was reduced by only 5.9% at the end of the 8-
h exposure. (However, note that the average O. concentration in the eighth hour was
0.03 ppm). This study illustrates two important points. First, a response plateau occurs. It is
intuitively obvious that there must be a limit to the acute decrease that can occur in FEV..
However, from this study, it is also clear that the response plateau must be dependent on the
O. concentration because much larger decreases in FEV. occur with exposure to
O. concentrations higher than 0.12 ppm. Second, the response to O. exposure is dependent on
the dose rate (some function of C and V.) and the cumulative dose (some function of dose rate
and T), at least when the O. concentration is varied. This study also affirms the observation
(Folinsbee et al., 1978; Adams et al., 1981; Hazucha, 1987; Larsen et al., 1991) that
7-69
-------
O. concentration is a more important factor in determining O. responses than is either exposure
duration or the volume of air breathed during the exposure.
7.2.3 Increased Airway Responsiveness
Increased airway responsiveness indicates that the airways are predisposed to
bronchoconstriction induced by a variety of stimuli (e.g., specific allergens, SO., cold air,
etc.). Airway responsiveness is usually measured by having the individual forcefully exhale
into a spirometer designed to measure expiratory flow rates (e.g., FEV.) or, less commonly,
by measuring R.. in a body plethysmograph. In order to determine the level of airway
responsiveness, airway function is measured before and immediately after the inhalation of
small amounts of an aerosolized bronchoconstrictor drug (e.g., methacholine or histamine).
The dose of the bronchoconstrictor drug is increased in a step-wise fashion until a
predetermined degree of airway response (e.g., a 20% drop in FEV. or a 100% increase
in R..) has occurred. The dose of the bronchoconstrictor drug that produced the
aforementioned response is often referred to as the "PD.." (i.e., the provocative dose that
produced a 20% drop in FEV.) or the "PD..." (i.e., the provocative dose that produced a
100% increase in R..).
A high level of bronchial responsiveness is a hallmark of asthma. However,
varying degrees of increased airway responsiveness may occur in other lung disease (e.g.,
chronic bronchitis or viral respiratory infections) or in healthy asymptomatic individuals. The
range of nonspecific bronchial responsiveness, as expressed by the PD.. for example, is at least
1,000-fold from the most sensitive asthmatics to the least sensitive healthy subjects (see Figure
7-6). The average PD.. for healthy subjects is 10 to 100 times that of mild to moderate
asthmatics (Chatham et al., 1982; Cockcroft et al., 1977). Atopic or allergic individuals
without asthma (intermediate in responsiveness between healthy subjects and mild asthmatics)
typically have a lower PD.. than healthy individuals (Townley et al., 1975; Cockcroft et al.,
1977). Increasing severity of asthma, as indicated by increasing symptoms or medication
usage, is associated with decreasing PD... Mild asthmatics may have a PD.. that is 10 times
higher than that of moderate or severe asthmatics (Cockcroft et al., 1977). A low PD.. in
nonasthmatics also is associated with increased symptoms and a reduced baseline FEV.
(Kennedy et al., 1990). The average changes in airway responsiveness induced by O. range
from 150 to 500%. This means that, in a healthy subject exposed to O., a PD.. of 20 units
would decrease to a PD.. between 13 and 4 units. Therefore, with a pronounced O.-induced
change in airway responsiveness, a healthy subject could move from the normal range into the
upper half of the mild asthmatic range of airway responsiveness.
Increases in airway responsiveness are an important consequence of exposure to O..
Results of studies reporting changes in airway responsiveness following O. exposure are
summarized in Table 7-10. These studies vary with regard to exposure regimens, type and
dose of bronchoconstrictive agent, and subject population. Increased airway responsiveness
associated with O. exposure was first reported by Golden et al. (1978), who studied histamine-
responsiveness in eight healthy men after exposure to 0.6 ppm O. for 2 h at rest
and found that the histamine-induced DR.. for the group was 300% greater 5 min after
O. exposure than at baseline. Two of their subjects, however, had an increased response to
histamine 1 week or greater after exposure, raising the possibility that high O. levels can result
in more persistent increases in airway responsiveness. Later, Holtzman et al. (1979) found in
16 nonasthmatic subjects that a 10-breath methacholine or histamine challenge
7-70
-------
Arbitrary Logarithmic Scale
1
0.01
I
0.10
I
1.00
I
10.0
I
100
allergy only *
* healthy
Figure 7-6. Airway function can be measured before and immediately after the
inhalation of an aerosolized bronchoconstrictor drug like methacholine. The
provocative dose that produces a 20% drop in forced expiratory volume in 1
s has been used to express the range of nonspecific bronchial
responsiveness.
increased SRaw almost twice as much after O. as after air exposure, but this effect resolved
after 24 h. Atopic subjects showed similar increases in responsiveness to histamine after
O. exposure. The authors concluded that the increased nonspecific bronchial responsiveness
after O. exposure was not related to atopy. Konig et al. (1980) found increased responsiveness
to inhaled acetylcholine after a 1-h exposure to 627 and 1,960 /xg/m' (0.32 and 1.00 ppm,
respectively). Folinsbee and Hazucha (1989) found increased airway responsiveness in 18
females 1 and 18 h after a 70-min exposure to 0.35 ppm O. when compared to air. Taken
together, these studies suggest that O.-induced increases in airway responsiveness usually
resolve 18 to 24 h after exposure, but may persist in some individuals for longer periods.
Dimeo et al. (1981) were the first to investigate "adaptation" to the increases in
airway responsiveness following O. exposure. Over 3 days of a 2 h/day exposure to 0.40 ppm
O., they found progressive attenuation of the increases in airway responsiveness such that,
after the third day of O. exposure, histamine airway responsiveness was no longer different
from the sham exposure levels. Kulle et al. (1982) extended these findings by exposing two
groups of healthy volunteers (n = 48) to 0.40 ppm O. for 3 h/day for 5 days in a row and
found that there was a significantly enhanced response to methacholine after the first 3 days of
exposure, but this response slowly normalized by the end of the fifth day. Thus, the
attenuation of O.-induced increases in airway responsiveness followed the same time course as
attenuation of other pulmonary function changes.
Gong et al. (1986) demonstrated increased airway responsiveness to histamine at
0.2 ppm O. in 17 vigorously exercising elite cyclists who were exposed for 1 h. Folinsbee et
al. (1988) found an approximate doubling of the mean methacholine responsiveness in a group
of healthy volunteers exposed for 6.6 h to 0.12 ppm O.. However, on an individual basis, no
relationship was found between O.-induced changes in airway responsiveness and those in
FVC and FEV., suggesting that changes in airway responsiveness and lung volume occurred
by different mechanisms. Horstman et al. (1990) extended Folinsbee's observations by
demonstrating significant decreases in the PD... in 22 healthy subjects immediately after a
7-71
-------
Table 7-10. Increased Airway Responsiveness Following Ozone Exposure!
Ozone
Concentration*
ppm
0.08
0.10
0.12
0.10
0.32
1.00
0.12
0.20
0.12
0. 12 ppm O3
0. 12 ppm O3
Air-lOOppb
Air-antigen
0. 12 ppm O3
0.35
Dg/m
157
196
235
196
621
1,960
235
392
235
-lOOppbSOj
-0.12 ppm O3
SO2
-antigen
686
Exposure
Duration and Activity
6.6 h
IE at D39 L/min
2h
1 h at VE = 89 L/min
followed by 3 to 4 min
at D 150 L/min
6.6 h with IE at
D25 L/min/m2 BSA
45 min in first atmosphere
and 15 min in second
IE
1 h at rest
70 min with IE at
40 L/min
Number and
Exposure Gender of Subject
Conditions Subjects Characteristics
18 DC 22 M
40% RH
NA 14
31 DC 15 M, 2F
35% RH
NA 10 M
75% RH 8M, 5F
22 DC
NA 4 M, 3 F
NA 18 F
Healthy NS,
18 to 32 years
old
Health NS,
24 + 2 years
old
Elite
cyclists, 19 to
30 years old
Healthy NS,
18 to 33 years
old
Asthmatic,
12 to 18 years
old
Asthmatic,
21 to 64 years
old
Healthy NS,
19 to 28 years
old
Observed Effect(s)
33, 47, and 55% decreases in cumulative dose
of methacholine required to produce a 100%
increase in SR,,W after exposure to O3 at 0.08,
0. 10, and 0. 12 ppm, respectively.
Increased airway responsiveness to
methacholine immediately after exposure at
the two highest concentrations of O3.
Greater than 20% increase in histamine
responsiveness in one subject at 0. 12 ppm
O3 and in nine subjects at 0.20 ppm O3.
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 declines in FEV, and Vmax50%
and greater increase in respiratory resistance
after O3-SO2 than after O3-O3 or air-SO2.
Increased bronchoconstrictor response to
inhaled ragweed or grass after O3 exposure
compared to air.
PD]00 decreased from 59 CIU after air
exposure to 41 CIU and 45 CIU, 1 and 18 h
after O3 exposure, respectively.
Reference
Horstman
et al.
(1990)
Konig
et al.
(1980)
Gong
et al.
(1986)
Folinsbee
et al.
(1988)
Koenig
et al.
(1990)
Molfmo
et al.
(1991)
Folinsbee
and
Hazucha
(1989)
-------
Table 7-10 (cont'd). Increased Airway Responsiveness Following Ozone Exposure^
Ozone
Concentration1"
ppm Dg/m
0.20 392
0.40 784
0.40 784
0.40 784
0.40 784
>J 0.60 1,176
ui
0.60 1,176
Exposure
Duration and Activity
2 h with IE at 2 x resting
2 h with IE at 2 x resting
2 h/day for 3 days
3 h/day for 5 days in a row
2 h with IE at
VE = 53 to 55 L/min
2 h at rest
2 h with IE at 2 x resting
Exposure
Conditions
22 DC
55% RH
22 DC
50% RH
NA
22 DC
55% RH
Number and
Gender of Subject
Subjects Characteristics'
12 M, 7 F Healthy NS,
21 to 32 years
old
13 M, IIP Healthy NS,
19 to 46 years
old
8 M, 10 F 9 asthmatics
(5 F, 4 M),
9 healthy
(5 F, 4 M),
18 to 34 years
old
5 M, 3 F Healthy NS,
22 to 30 years
old
11 M, 5F 9atopic,
7 nonatopic,
NS, 21 to
35 years old
Observed Effect(s) Reference
1 10% increase in DSRaw to a 10-breath histamine (1.6%) Dimeo
aerosol challenge after exposure to O3 at 0.40 ppm, but et al.
no change at 0.20 ppm. Progressive adaptation of this (1981)
effect over 3-day exposure.
Enhanced response to niethacholine after first 3 days, but Kulle et al.
this response normalized by Day 5. (1982)
Decreased PC100SR ^ from 33 mg/mL to 8.5 mg/mL in Kreit et al.
healthy subjects after O3. PC,00sRaw fell from 0.52 mg/mL (1989)
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.
300% increase in histamine- induced DRaw 5 min after O3 Golden
exposure; 84 and 50% increases 24 h and 1 week after et al.
exposure (p > 0.05), respectively. Two subjects had an (1978)
increased response to histamine 1 week after exposure.
Ten-breath niethacholine or histamine challenge increased Holtzman
SRj,w D 150% in 16 nonasthmatics after O3. On average, et al.
the atopic subjects had greater responses than the (1979)
nonatopic subjects. The increased responsiveness
resolved after 24 h. Atropine premedication blocked the
O3-induced increase in airway responsiveness.
"See Appendix A for abbreviations and acronyms.
bListed from lowest to highest O3 concentration.
-------
6.6-h exposure to concentrations of O. as low as 0.08 ppm. Because methacholine challenges
were not conducted at later time points in any of these studies, the duration of the increased
airway responsiveness after ambient-level O. exposure could not be determined.
No doubt exists that O., even at ambient concentrations, produces acute increases in
airway responsiveness. Whether O. exposure causes protracted increases in airway
responsiveness in healthy individuals, induces asthma, or predisposes individuals to asthma is a
more difficult question to answer (see Section 7.4.2). However, the increases in airway
responsiveness following O. exposure, even if short in duration, may have important clinical
implications. Several studies have been conducted specifically to determine the significance of
acute increases in airway responsiveness after O. exposure. These studies, designed to test the
hypothesis that an O. exposure heightens the response to a subsequent bronchoconstrictor
challenge, have exposed asthmatics to O. or air and, then, to a known bronchoconstrictor agent
to compare the pulmonary function changes after O. to those after air. Kreit et al. (1989) were
first to investigate the change in airway responsiveness that occurs after O. exposure in
individuals with asthma. They exposed nine mild asthmatics (baseline PC < 1.5
mg/niL) for 2 h to 0.40 ppm O. with IE and found that the baseline PC declined from
0.52 to 0.19 mg/niL after O. as compared to 0.48 to 0.27 mg/mL after air. Koenig et al.
(1990) demonstrated that a 45-min exposure to 0.12 ppm O. followed by a 15-min exposure to
100 ppb SO. caused greater changes in FEV., respiratory resistance, and V in 14
adolescent asthmatics than did an air-SO. exposure combination.
Molfmo et al. (1991) examined the effects of a 1-h resting exposure to 0.12 ppm
O. on the response to a ragweed or grass allergen inhalation challenge. Asthmatic subjects
were exposed twice to air and twice to O., once per week over a period of 4 weeks. Two
allergen challenges were performed, once after air and once after O. exposure. The other air
and O. exposures were followed by a placebo challenge. A ragweed allergen extract was used
for six of the seven subjects. The order of experiments was not randomized (in an effort to
avoid unexpectedly severe reactions); six of the seven subjects were exposed to the ozone-
allergen condition last and five of the seven were exposed to the air-placebo condition first.
Allergen responsiveness was expressed as the allergen concentration needed to cause a 15%
reduction in FEV. or PC... The PC., was lower after the O. exposure than after the air
exposure (p = 0.04). These observations suggest that allergen-specific airway responsiveness
is increased after O. exposure. Although it is expected that specific bronchial reactivity will be
increased by O. exposure based on the marked increases in nonspecific bronchial
responsiveness induced by O. exposure, such a response would not have been anticipated under
these mild exposure conditions where lung function or symptomatic responses have not been
observed. The lack of randomization in this study makes it difficult to assess the validity of
conclusions based on the statistical analysis. These results are provocative but should be
considered preliminary until this experiment can be repeated.
Ozone may be a clinically important co-factor in the response to airborne
bronchoconstrictor substances in individuals with asthma. It is plausible that this phenomenon
could contribute to increased asthma exacerbations and, even, consequent increased hospital
admissions (see Section 7.4.1). Whether the increased airway responsiveness following
O. exposure produces an accentuated bronchoconstrictor response to inhaled allergens or SO.
in healthy individuals or those with lung diseases other than asthma is unknown.
Several studies have been undertaken to determine the mechanism of O. -induced
increases in airway responsiveness (also see Chapter 6). Early experiments in dogs (Lee et al.,
7-74
-------
1977) and humans (Golden et al., 1978) suggested an important role for vagal reflexes because
vagal nerve cooling and atropine inhibited the increase in histamine-induced
bronchoconstriction caused by O.. Ozone exposure increased bronchomotor responses to
cholinergic stimuli (e.g., acetylcholine and methacholine) in dogs (Holtzman et al., 1983) and
humans (Seltzer et al., 1986). Subsequent studies, however, revealed that bilateral vagotomy
did not inhibit O.-induced hyperresponsiveness to subcutaneous histamine in guinea pigs
(Gordon et al., 1984). These data provide strong evidence that O.-induced increased airway
responsiveness is mediated, at least in part, by cholinergic receptors on airway smooth muscle
cells. Interestingly, Gordon et al. also noted that isometric tension in guinea pig tracheal
smooth muscle and lung parenchymal strips in response to histamine and carbachol was not
affected by exposure to O., suggesting that O. affected the in vivo milieu surrounding the
smooth muscle rather than produced direct effects on the smooth muscle itself.
It can be hypothesized that the increased epithelial permeability caused by O. (see
Chapter 6) may allow greater penetration of bronchoconstrictor substances, including
methacholine and histamine, and that this would lead to increased airway responsiveness.
However, Roum and Murlas (1984) suggested that the increased epithelial permeability after
O. could not totally explain this phenomenon because parenteral cholinergic challenge after
O. more reproducibly caused bronchospasm than did inhalation challenge with methacholine.
The increased responsiveness to parenteral compared to inhaled cholinergic challenge may,
however, have been due to increased bronchial blood flow after O. exposure. Therefore, the
findings of Roum and Murlas do not exclude increased epithelial permeability as the cause of
increased airway responsiveness after O. exposure.
Holtzman et al. (1983) first pointed out that O.-induced acute inflammation may be
important in the induction of the increased airway responsiveness. In mongrel dogs exposed to
O., they found bronchial wall PMN infiltration in those animals that developed increased
airway responsiveness to acetylcholine, but not in animals that failed to develop increased
airway responsiveness. O'Byrne et al. (1984) later demonstrated that hydroxyurea
simultaneously decreased peripheral blood leukocyte counts, decreased PMN influx into
bronchial tissue, and prevented increased airway responsiveness in dogs exposed to O.. Both
O.-induced increased airway responsiveness and bronchial tissue PMN influx returned 6 weeks
after treatment was discontinued when peripheral leukocyte counts had normalized. Seltzer et
al. (1986) found a larger percentage of PMNs (30.8% versus 8.0%) in BAL fluid after
O. exposure in their subjects that had a greater than threefold decrease in the provocative
concentration, which caused a D8 L X cm H.O/L/S increase in SR.. for methacholine,
compared to those subjects that had less than a twofold decrease. These data suggest a possible
association between inflammation and increased airway responsiveness after O. exposure.
An early study in dogs (O'Byrne et al., 1984) suggested that oxygenation products
of arachidonic acid that are sensitive to inhibition by the anti-inflammatory drug indomethacin
play a role in O.-induced hyperresponsiveness without affecting the influx of PMNs. In the
first of several human studies with a PG inhibitor, indomethacin did not attenuate the increase
in airway responsiveness in subjects exposed to 0.4 ppm O. for 2 h (Ying et al., 1990), but did
ameliorate the effect of O. on spirometric endpoints. Kleeberger and Hudak (1992) observed a
marked reduction in PMN influx in O.-exposed mice given indomethacin without any change
in O.-induced increases in permeability, as indicated by BAL protein. However, Hazucha et
al. (1996) found no effect of ibuprofen on PMN levels or protein in the BAL fluid of O.-
exposed humans (also see Section 7.2.4.5). Seltzer et al. (1986) and Keren et al. (1989a,b)
7-75
-------
found that O. increases a large number of BAL inflammatory mediators (including PGE.,
PGF.n, and thromboxane B. [TXB.]), one or more of which may play a role in the increase in
airway responsiveness after O. exposure.
The role of reactive oxygen metabolites or neuropeptide mediators in the increase in
airway responsiveness after O. has not been investigated. Furthermore, there has been no
direct assessment of alterations in nerve afferents, changes in neurotransmitter concentrations,
changes in smooth muscle postsynaptic receptors, or modulation of nerve signal transmission
by inflammatory mediators as these pertain to the increase in airway responsiveness after O..
In conclusion, although the mechanism of O.-induced increases in airway responsiveness is not
completely understood, it appears to be associated with a number of cellular or biochemical
changes in the airway (see Section 7.2.4 and Tables 7-11 and 7-12). Because these alterations
are part of a complex process, it comes as no surprise that the mechanistic studies on O.-
induced increases in airway responsiveness have not pinpointed an isolated derangement.
7.2.4 Inflammation and Host Defense
7.2.4.1 Introduction
In general, inflammation can be considered as the host response to injury, and the
induction of inflammation can be accepted as evidence that injury has occurred. Several
outcomes are possible: (1) inflammation can resolve entirely; (2) continued acute inflammation
can evolve into a chronic inflammatory state; (3) continued inflammation can alter the structure
or function of other pulmonary tissue, leading to diseases such as fibrosis or emphysema;
(4) inflammation can alter the body's host defense response to inhaled microorganisms,
particularly in potentially vulnerable populations such as the very young and old; and
(5) inflammation can alter the lung's response to other agents such as allergens or toxins. It is
also possible that the profile of response can be altered in persons with preexisting pulmonary
disease (e.g., asthma or COPD) or smokers. At present, it is known that short-term exposure
of humans to O. can cause acute inflammation and that long-term exposure of laboratory
animals results in a chronic inflammatory state (see Chapter 6). However, the relationship
between repetitive bouts of acute inflammation in humans caused by O. and the development of
chronic respiratory disease is unknown.
The previous O. criteria document (U.S. Environmental Protection Agency, 1986)
contained no studies in which inflammation was measured in humans exposed to O..
Fiberoptic bronchoscopy since has been used to sample cells and fluids lining the respiratory
tract of humans for many markers (Reynolds, 1987). Bronchoalveolar lavage primarily
samples the alveolar region of the lung; however, the use of small volume lavages (Rennard et
al., 1990) or balloon catheters also allows sampling of the airways. Nasal lavage allows
sampling of cells and fluid removed from the nasal passages.
In the past 6 years, several studies have analyzed BAL and NL cells and fluid from
humans exposed to O. for markers of inflammation and lung damage (see Tables 7-11 and
7-12). The presence of PMNs in the lung has long been accepted as a hallmark of
inflammation and has been taken as the major indicator that O. causes inflammation in the
7-76
-------
Table 7-11. Bronchoalveolar Lavage Studiesof Inflammatory Effects
from Controlled Human Exposure to Ozone1
Ozone Concentration11
Exposure
ppm Dg/m 3 Duration
0.08 157 6.6 h
0.10 196
0.20 392 4 h
0.30 588 1 h (mouth-
piece)
0.40 784 2 h
0.40 784 2 h
0.40 784 2 h
0.40 784 2 h
Activity Level
(VF)
IE (40 L/min)
six 50-min
exercise periods
+ 10 min rest;
35 min lunch
IE (50 min at
40 L/min,
10 min rest)
CE (60 L/min)
IE (70 L/min) at
15-min intervals
IE (70 L/min) at
15-min intervals
IE (70 L/min) at
15-min intervals
IE (70 L/min) at
15-min intervals
Number and
Gender of
Subjects
18 M,
18 to 35 years
old
15 M, 13 F,
21 to
39 years old
5M
11 M,
18 to
35 years old
11 M,
18 to
35 years old
11 M,
18 to
35 years old
10 M,
18 to 35 years
old
Observed Effect(s)
BAL fluid 18 h after exposure to 0. 1 ppm O3 had significant
increases in PMNs, protein, PGEj, fibronectin, IL-6, lactate
dehydrogenase, and D-l 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.
Bronchial lavage, bronchial biopsies, and BAL done 18 h after
exposure. BAL shows changes similar to other studies. Airway
lavage shows increased cells, LDH, IL-8. Biopsies show increased
number of PMNs.
Significantly elevated PMNs in the BAL fluid 1 , 6, and 24 h after
exposure, with peak increases at 6 h.
BAL fluid 18 h after exposure had significant increases in PMNs,
protein, albumin, IgG, PGEj, plasminogen activator, elastase,
complement C3a, and fibronectin.
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 1 h after exposure to 0.4 ppm O3 had significant increases
in PMNs, protein, PGE2, TXBj, IL-6, LDH, D-l antitrypsin, and
tissue factor compared with the same subjects exposed to FA.
Decreased phagocytosis of yeast by alveolar macrophages.
Reference
Devlin et al.
(1990,
1991)
Koren et al.
(1991)
Aris et al.
(1993a)
Schelegle
et al. (1991)
Koren et al.
(1989a,b)
Devlin and
Koren
(1990)
McGee
et al. (1990)
Koren et al.
(1991)
-------
Table 7-11 (cont'd). Bronchoalveolar Lavage Studies of Inflammatory Effects
from Controlled Human Exposure to Ozone*
CO
Ozone Concentration*
Exposure
ppm Dg/m Duration
0.40 784 2 h/day for
5 days, 2 h
either 10 or
20 days later
0.40 784 2 h
0.40 784 2 h
0.60 1,176
Activity Level
(VE)
IE (40 L/min) at
15-min intervals
IE (60 L/min) at
15-min intervals
IE (83 W for
women, 100 W
for men) at
15-min intervals
Number and
Gender of
Subjects
16 M,
18 to 35 years
old
10 M
7M, 3F,
23 to 41 years
old
Observed Effect(s) 3 Reference
BAL done immediately after fifth day of exposure and again Devlin et al.
after exposure 10 or 20 days later. Most markers of (1995)
inflammation (PMNs, IL-6, IL-8, protein, Dl-antitrypsin,
PGE2, fibronectin) showed complete attenuation; markers of
damage (LDH, elastase) did not. Reversal of attenuation
was not complete for some markers, even after 20 days.
Subjects given 800 mg ibuprofen or placebo 90 inin before Hazucha et al.
exposure. Subjects given ibuprofen had less of a decrease (1996)
in FEV, after O3 exposure. BAL fluid 1 h after exposure
contained similar levels of PMNs, protein, fibronectin,
LDH, D-l antitrypsin, LTB4, and C3a in both ibuprofen and
placebo groups. However, subjects given ibuprofen had
decreased levels of IL-6, TXBj, and PGEj.
BAL fluid 3 h after exposure had significant increases Seltzer et al.
in PMNs, PGE2, TXBj, and PGF2D at both O3 (1986)
concentrations.
"See Appendix A for abbreviations and acronyms.
bListed from lowest to highest O3 concentration.
-------
Table 7-12. Additional Studies of Inflammatory and Host Defense Effects
from Controlled Human Exposure to Ozone*
Ozone Concentration1"
ppm Dg/m Duration
Activity Level
(VE)
Number and
Gender of
Subjects
Observed Effect(s) 3
Reference
Nasal Lavage Studies
0.12 235 90min
0.24 470
0.30 588 6 h/day for
5 consecutive
days
0.40 784 2 h
0.50 980 4 h on
2 consecutive
days
0.50 980 4 h
IE (20 L/min)
at 15-min
intervals
IE (light
treadmill)
IE (70 L/min)
at 15-min
intervals
Resting
Resting
5M, 5F, asthmatic;
4M, 4F,
nonasthmatic;
18 to 41 years
old
24 M
(12 O3, 12 air)
11 M,
18 to 35 years
old
41 M
(21 03,
20 air-exposed), 18
35 years
old
6M, 6F,
allergic rhinitics,
31.4 + 2.0 (SD)
years old
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.
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.
McBride et al.
(1994)
Henderson
et al. (1988)
NL done immediately before, immediately after, and 22 h Graham and
after exposure. Increased numbers of PMNs at both timesKoren (1990)
after exposure; increased levels of tryptase, a marker of Koren et al.
mast cell degranulation, immediately after exposure; (1990)
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
toPMNs at all times after the first exposure, with peak
values occurring immediately prior to the second
exposure.
NL done immediately after exposure. Increased
upper and lower respiratory symptoms and increased
levels of PMNs, eosinophils, and albumin in NL fluid.
Graham et al.
(1988)
Basconi et al.
(1990)
-------
Table 7-12 (cont'd). Additional Studies of Inflammatory and Host Defense Effects
from Controlled Human Exposure to Ozone*
CO
o
Ozone
Concentration*
ppm
Dg/m
Exposure Activity Level
Duration ( VE)
Number and
Gender of
Subjects
Observed Effect(s)
3 Reference
Clearance Studies
0.20
0.40
0.40
0.40
0.50
In Vitro
0.25
0.50
0.25
0.50
1.00 1
0.25
0.50
1.00 1
392
784
784
784
784
Studies
490
980
490
980
,960
490
980
,960
2 h IE (light
treadmill)
1 h CE (40 L/min)
2 h IE (70 L/min) at
15-min intervals
2.25 h IE (70 L/min) at
15-min intervals
6 h Human nasal
epithelial cells
1 h Airway
epithelial cell
line
1 h Airway
epithelial cell
line and
alveolar
macrophages
7M,
27.2 +
6.0 (SD)
years old
15 M or F,
18 to
35 years old
8M,
20 to
30 years old
16 M,
20 to
30 years old
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.
Subjects inhaled radiolabeled iron oxide particles 2 h after exposure.
No O3-induced difference in clearance of particles during the next 3 h
or the following morning.
Subjects inhaled 99mTc-DTPA 75 min after exposure. Significantly
increased clearance of ""Tc-DTPA from the lung in O3-exposed
subjects. Subjects had expected changes in FVC and SR,W.
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.
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.
Concentration-dependent increased secretion of PGEj , TXB, , PGF2D ,
LTB4, and LTD4. More secretion basolaterally than apically.
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.
Foster et al. (1987)
Gerrity et al. (1993)
Kehrl et al. (1987)
Kehrl et al. (1989)
Beck et al. (1994)
McKinnon et al.
(1993)
Devlin et al. (1994)
-------
Table 7-12 (cont'd). Additional Studies of Inflammatory and Host Defense Effects
from Controlled Human Exposure to Ozone*
Ozone Concentration1"
ppm Dg/m
Exposure
3 Duration
Activity Level
(VF)
Number and
Gender of
Subjects
Observed Effect(s)
Reference
In Vitro Studies (cont'd)
0.30 588
1.00 1,960
Ih
Alveolar
macrophages
Concentration-dependent increases in PGE2 production, and
decreases in phagocytosis of sheep erythrocytes.
No O3-induced secretion of IL-1, TNF, or IL-6.
Becker et al.
(1991)
"See Appendix A for abbreviations and acronyms.
bListed from lowest to highest O3 concentration.
CO
-------
lungs of humans. Soluble mediators of inflammation (or its resolution) such as cytokines and
arachidonic acid metabolites also have been measured in the BAL fluid of humans exposed to
O.. Cytokines that have been reported most often are interleukin (IL)-6 and IL-8, although IL-
1 and tumor necrosis factor (TNF) also have been studied. Soluble metabolites of arachidonic
acid involved in inflammation and host defense (e.g., PGE. and PGF.n, thromboxane, and
leukotrienes [LTs] such as LTB.) also have been reported in the BAL fluid of humans exposed
to O.. In addition to their role in inflammation, many of these compounds have
bronchoconstrictive properties and may be involved in increased airway hyperreactivity
observed following O. exposure.
Under normal circumstances, the epithelia lining the large and small airways
develop tight junctions and restrict the penetration of exogenous particles and macromolecules
from the airway lumen into the interstitium and blood, as well as restrict the flow of plasma
components into the airway lumen. However, several studies (see Table 7-12) show that
O. disrupts the integrity of the epithelial cell barrier in human airways, as measured by
increased passage of radiolabeled compounds out of the airways, as well as passage of markers
of plasma influx such as albumin, immunoglobulin, and other proteins into the airways. In
addition, markers of epithelial cell damage such as lactate dehydrogenase (LDH) activity also
have been measured in the BAL fluid of humans exposed to O..
Inflammatory cells of the lung such as alveolar macrophages (AMs), monocytes,
and PMNs also constitute an important component of the pulmonary host defense system.
In their unstimulated state, they present no danger to surrounding pulmonary cells and tissues,
but upon activation, they are capable of generating free radicals and enzymes with microbicidal
capabilities, but they also have the potential to damage nearby cells. Animal studies have
demonstrated that O. decreases host defense system function (see Chapter 6, Section 6.2.3).
Other soluble factors that have been studied include those involved with fibrin
deposition and degradation (Tissue Factor, Factor VII, and plasminogen activator), potential
markers of fibrogenesis (fibronectin, platelet derived growth factor), and components of the
complement cascade (C3a).
7.2.4.2 Inflammation Assessed by Bronchoalveolar Lavage
Seltzer et al. (1986) were the first to demonstrate that exposure of humans to
O. resulted in inflammation in the lung. In this study, 10 volunteers were exposed to 0.4 or
0.6 ppm O. for 2 h while undergoing exercise, and BAL was performed 3 h later.
Bronchoalveolar lavage fluid from subjects exposed to O. contained 7.8-fold more PMNs
compared with BAL fluid from the same subjects exposed to FA. Additionally, BAL fluid
from O.-exposed subjects contained increased levels of PGE., PGF.n, and TXB. compared to
fluid from air-exposed subjects. Keren et al. (1989a,b) also described inflammatory changes in
the lungs of 11 subjects exposed to 0.4 ppm O. for 2 h while undergoing IE at 70 L/min in a
study designed to simulate adults working outdoors or children actively playing.
Bronchoalveolar lavage was performed 18 h after O. exposure. Subjects exposed to O. had an
eightfold increase in PMNs in the BAL fluid, confirming the observations of Seltzer et al.
In addition, Keren et al. reported a twofold increase in BAL fluid protein, albumin, and
immunoglobulin G (IgG) levels, suggestive of increased epithelial cell permeability as a result
of O. exposure. There was also a 12-fold increase in IL-6 levels in the BAL fluid.
Interleukin-1 and TNF were not present in detectable levels in the BAL fluid of any subject.
There was, however, a twofold increase in the proinflammatory eicosanoid PGE., as well as a
7-82
-------
twofold increase in the complement component C3a. This study also provided evidence for
stimulation of fibrogenic processes in the lung by demonstrating significant increases in two
components of the coagulation pathway, Tissue Factor and Factor VII (McGee et al., 1990), as
well as urokinase plasminogen activator and fibronectin (Keren et al., 1989a). Taken together,
these two studies demonstrate that exposure of humans to moderate levels of O. results in an
inflammatory reaction in the lung, as evidenced by substantial increases in PMNs and
proinflammatory compounds. Furthermore, these studies demonstrate that both cells and
mediators capable of damaging pulmonary tissue are increased after O. exposure, as are
compounds that play a role in fibrotic and fibrinolytic processes.
Although animal studies have shown that the terminal bronchioles are a major site
of O.-induced inflammation, few human studies have confirmed this finding because BAL
primarily samples cells and fluid in the terminal bronchioles and alveoli. However, isolated
lavage of the mainstream bronchus using balloon catheters or the more traditional BAL using
small volumes of saline have the ability to preferentially measure O.-induced changes in the
large airways. In one study, isolated airway lavage was performed on 14 subjects 18 h after
exposure to 0.2 ppm O. while undergoing moderate exercise (Aris et al., 1993a). Increases in
total lavagable cells, LDH activity, and IL-8 were reported. In contrast, Schelegle et al.
(1991), observed no increase in PMNs in the bronchial fluid; however, bronchial biopsies
showed increased numbers of PMNs in airway tissue.
The data suggestive of O. -induced changes in epithelial cell permeability described
by Keren et al. (1989a, 1991) and Devlin et al. (1991) support earlier work in which epithelial
cell permeability, as measured by increased clearance of radiolabled diethylene triamine
pentaacetic acid ('" Tc-DTPA) from the lungs of humans exposed to O., was demonstrated
(Kehrl et al., 1987). In that study, eight healthy subjects who inhaled '" Tc-DTPA just prior
to exposure to air or 0.4 ppm O. for 2 h while undergoing heavy exercise (65 L/min) had
increased clearance of the compound. Kehrl et al. (1989) reported similar observations on an
additional 16 subjects. For the combined group of 24 subjects exposed for 2 h to 0.4 ppm O.,
the average clearance rate was 60% faster than that observed after air exposure, strongly
suggesting increased permeability from the airway lumen and alveolar space to the blood and
interstitial spaces. The average O.-induced decrement in FVC in these subjects was 10%.
These changes in permeability most likely are associated with acute inflammation and
potentially could allow better access of inhaled antigens and other substances to the submucosa.
Studies in which human AM and airway epithelial cells were exposed to O. in vitro
suggest that most of the components found in increased levels in the BAL fluid of O.-exposed
humans are produced by epithelial cells. Macrophages exposed to 0.3 and 1.0 ppm (but not
0.1 ppm) O. for 1 h showed small increases in PGE., but no change in superoxide anion or
cytokine production (Becker et al., 1991). In contrast, airway epithelial cells exposed in vitro
to 0.1, 0.25, 0.5, and 1.0 ppm O. for 1 h showed large concentration-dependent increases in
PGE., TXB., LTB., LTC., and LTD. (McKinnon et al., 1993). These cells also showed
increases in IL-6, IL-8, and fibronectin at O. concentrations as low as 0.1 ppm (Devlin et al.,
1994). Interestingly, macrophages removed 18 h later from subjects exposed to 0.4 ppm
O. for 2 h while undergoing intermittent heavy exercise (Keren et al., 1989a) showed changes
in the rate of synthesis of 123 different proteins as measured by quantitative computerized
densitometry of two-dimensional gel protein profiles. However, AMs exposed to O. in vitro
showed changes only in the rate of synthesis of six proteins, suggesting that most of the
7-83
-------
changes seen in the in vivo-exposed AMs were due to actions resulting from mediators
released by other cells following O. exposure, which then altered macrophage function.
Numerous studies have shown that humans exposed to O. for 5 consecutive days
experience decrements in pulmonary function on the first and second days, but the decrements
diminish with each succeeding day so that by the fifth day, no such effects are observed (see
Section 7.2.1). However, these studies did not address the question of whether repeated
exposure to O. also resulted in attenuation of inflammation or lung damage. Animal studies
suggest that although some markers of inflammation may be diminished, underlying damage to
lung epithelial cells continues (Tepper et al., 1989). In a recent study (Devlin et al., 1995),
humans were exposed to 0.4 ppm O. for 5 consecutive days (2 h/day while undergoing IE) and
then were exposed to O. a single time either 10 or 20 days later. The results show that
numerous indicators of inflammation (e.g., PMN influx, IL-6, IL-8, PGE., BAL protein,
fibronectin, macrophage phagocytosis) show attenuation (i.e., there is a complete
disappearance of response, and values are no different from those observed in the same
individual after 5 days of exposure to FA). Ten days later, some of these markers regained
full susceptibility, but others did not regain susceptibility even after 20 days. In agreement
with animal studies, some markers (LDH, elastase) never show attenuation, indicating that
tissue damage may continue to occur during repeated exposure.
7.2.4.3 Inflammation Induced by Ambient Levels of Ozone
Devlin et al. (1991) reported an inflammatory response in humans exposed to levels
of O. at or below 0.12 ppm. In this study, 10 volunteers were exposed to 0.08 and 0.10 ppm
O. for 6.6 h while undergoing moderate exercise (40 L/min) and underwent BAL 18 h later.
An additional eight subjects were exposed to 0.08 ppm O.. Increased numbers of PMNs and
levels of IL-6 were found at both O. concentrations. There also were increases in most of the
other compounds reported by Keren et al. (1989a,b), including fibronectin and PGE..
Alveolar macrophage phagocytic capability was also monitored in this study, and it was
reported that macrophages removed from humans exposed to both O. concentrations had
decreased ability to phagocytize Candida albicans opsonized with complement. Comparison of
the magnitude of inflammatory changes observed in this study and by Keren et al. (1989a,b),
when normalized for differences in concentration, duration of exposure, and ventilation,
suggest that lung inflammation from O. may occur as a consequence of exposure to ambient
levels while exercising. Although the mean changes in IL-6, PGE., and PMNs reported by
Devlin et al. (1991) were small, there was a considerable range of response among the
individuals participating in the study. Thus, although some of the study population showed
little or no response to O., others had increases in IL-6 or PMNs that were as large as or larger
than those reported by Keren et al. (1989a,b) when subjects were exposed for 2 h to 0.4 ppm
O.. Interestingly, those individuals who had the largest increases in inflammatory mediators in
this study did not necessarily have the largest decrements in pulmonary function, suggesting
separate mechanisms underlying these two responses to O.. These data suggest that, although
the population as a whole may have a small inflammatory response to low levels of O., there
may be a significant subpopulation that is very sensitive to these low levels of O..
Furthermore, even a small inflammatory response (if it recurs) in the population as a whole
should not be discounted.
7.2.4.4 Time Course of Inflammatory Response
7-84
-------
The time course of the inflammatory response to O. in humans has not been
explored fully. Studies in which BAL was performed 1 h (Devlin et al., 1990; Keren et al.,
1991) or 3 h (Seltzer et al., 1986) after exposure to 0.4 ppm O. demonstrate that the
inflammatory response is quickly initiated, and other data (Keren et al., 1989a,b) indicate that,
even 18 h after exposure, inflammatory mediators such as IL-6 and PMNs are still
substantially elevated. However, a comparison of these studies shows there are differences in
the magnitude of response of some indicators, depending on when BAL is performed after
O. exposure. Ozone-induced increases in PMNs, IL-6, and PGE. are greater 1 h after
O. exposure, whereas BAL levels of fibronectin and plasminogen activator are greater
18 h after exposure. Still other compounds (protein, Tissue Factor) are equally elevated both
1 and 18 h after O. exposure. Schelegle et al. (1991) exposed five subjects to FA or 0.3 ppm
O. for 1 h with a ventilation of 60 L/min. Each subject was exposed to O. on three separate
occasions, and BAL was performed 1, 6, or 24 h after exposure. In addition, BAL was
separated into two fractions: the first 60 mL wash was designated the "proximal airways"
fraction (PA), and the remaining three 60 mL washes were pooled and designated the "distal
airways and alveolar surface" fraction (DAAS). The percent of PMNs in the PA sample was
statistically elevated at 1, 6, and 24 h after O. exposure, with a peak response at 6 h. The
percent of PMNs in the DAAS sample was elevated at only the 6 and 24 h time points, with
equivalent elevations at each time.
7.2.4.5 Effect of Anti-Inflammatory Agents on Ozone-Induced Inflammation
Previous studies (Schelegle et al., 1987; Eschenbacher et al., 1989) have shown that
indomethacin, an anti-inflammatory agent that inhibits the production of cyclooxygenase
products of arachidonic acid metabolism, is capable of blunting the well-documented
decrements in pulmonary function observed in humans exposed to O.. In a recent study,
10 healthy male volunteers were given 800 mg ibuprofen, another anti-inflammatory agent that
blocks cyclooxygenase metabolism, or a placebo 90 min prior to a 2-h exposure to 0.4 ppm O..
An additional 200 mg was administered following the first hour of exposure. Bronchoalveolar
lavage was performed 1 h after the exposure. As expected, subjects given ibuprofen had
blunted decrements in lung function following O. exposure compared to the same subjects
given a placebo (Hazucha et al., 1996). Bronchoalveolar lavage fluid from subjects given
ibuprofen also had reduced levels of the cyclooxygenase product PGE. as well as IL-6, but no
decreases were observed in PMNs, fibronectin, permeability, LDH activity, or macrophage
phagocytic function (Hazucha et al., 1995). These data suggest that although anti-
inflammatory agents may blunt O.-induced decrements in FEV. and increases in PGE., most
inflammatory mediators are elevated in the BAL of these subjects.
7.2.4.6 Use of Nasal Lavage To Assess Ozone-Induced Inflammation in the Upper
Respiratory Tract
Bronchoalveolar lavage has proven to be a powerful research tool to analyze
changes in the lung following exposure of humans to xenobiotics. However, because BAL is
expensive, somewhat invasive, and requires specialized personnel and facilities, it usually is
done only with small numbers of subjects and in selected medical centers. Therefore, there is
increasing interest in the use of NL as a tool in assessing O.-induced inflammation in the upper
respiratory tract, which is the primary portal for inspired air, and therefore the first region of
the respiratory tract to come in contact with airborne xenobiotics. Nasal lavage is simple and
7-85
-------
rapid to perform, is noninvasive, and allows collection of multiple sequential samples from the
same person. Graham et al. (1988) reported increased levels of PMNs in the NL fluid of
21 humans exposed to 0.5 ppm O. at rest for 4 h on 2 consecutive days, with NL performed
immediately before and immediately after each exposure as well as 22 h after the second
exposure. Nasal lavage fluid contained elevated numbers of PMNs at all postexposure times
tested, with peak values occurring immediately prior to the second day of exposure. There
were no changes in PMN numbers at any time in 20 subjects exposed to clean air for 2
consecutive days. Bascom et al. (1990) exposed 12 subjects with allergic rhinitis to 0.5 ppm
O. at rest for 4 h, followed immediately by NL. They reported a sevenfold increase in PMNs,
a 20-fold increase in eosinophils, and a 10-fold increase in mononuclear cells following
O. exposure, as well as a 2.5-fold increase in albumin. Graham and Keren (1990) compared
inflammatory mediators present in both the NL and BAL fluids of humans exposed to O.. The
same 11 subjects who were exposed to 0.4 ppm O. for 2 h with BAL performed 18 h later, as
described earlier (Keren et al., 1989a,b), also underwent NL immediately before, immediately
after, and 18 h after each exposure (Graham and Keren, 1990). There were significant
increases in PMNs in the NL fluid taken both immediately after exposure and on the next day.
Increases in NL and BAL PMNs were similar (6.6- and eightfold, respectively), demonstrating
a qualitative correlation between changes in the lower airways as assessed by BAL and the
upper respiratory tract as assessed by NL. Furthermore, all individuals who had increased
PMNs in BAL fluid also had increased PMNs in NL fluid, although the NL PMN increase
could not quantitatively predict the BAL PMN increase. Albumin, a marker of epithelial cell
permeability, was increased 18 h later, but not immediately after exposure. There were no
changes in PGE., plasminogen activator, LTC., LTD., or LTE. (Graham and Keren, 1990).
However, tryptase, a constituent of mast cells contained in the same granules as histamine, was
found in elevated levels immediately after O. exposure, but not 18 h later (Keren et al., 1990).
McBride et al. (1994) reported that asthmatic subjects are more sensitive to upper airway
inflammation at O. concentrations that do not affect lung function. Nasal lavage and lung and
nasal function were compared in 10 asthmatic and 8 nonasthmatic subjects exposed in a head
dome to 0.12 and 0.24 ppm O. for 90 min during intermittent moderate exercise (V. = 20
L/min). A significant increase in the number of PMNs in NL fluid was detected in the
asthmatic subjects both immediately and 24 h after exposure to 0.24 ppm O.. Total white
blood count, a surrogate for PMN influx, was signficantly correlated with IL-8 in the NL
fluid. No significant cellular changes were seen in nonasthmatic subjects, and no changes in
lung or nasal function or biochemical mediators were found in either asthmatic or nonasthmatic
subjects. These studies suggest that NL may serve as a sensitive and reliable tool to detect
inflammation in the upper airways of humans exposed to xenobiotics.
7.2.4.7 Changes in Host Defense Capability Following Ozone Exposure
Concern about the effect of O. on human host defense capability derives from
numerous animal studies demonstrating that acute exposure to as little as 0.08 ppm O. causes
decrements in antibacterial host defenses and little, if any, effect on the course of acute viral
infection (see Chapter 6, Section 6.2.3). A study of experimental rhinovirus infection in
susceptible human volunteers failed to show any effect of 5 consecutive days of O. exposure on
the clinical evolution or host response to a viral challenge (Henderson et al., 1988). In this
study, 24 young males were inoculated with type 39 rhinovirus (1,000 TCID-50) administered
as nose drops. Half were then exposed to 0.3 ppm O. (6 h/day) for 5 consecutive days while
7-86
-------
undergoing intermittent light exercise, and half were exposed to clean air under the same
regimen. There was no difference in rhinovirus titers in nasal secretions between the O.-
exposed and control groups, nor were there any differences in levels of interferon gamma or
PMNs in NL fluid or in blood lymphocyte proliferative response to rhinovirus antigen.
However, recent findings that rhinovirus can attach to the intracellular adhesion molecule
(ICAM) receptor on respiratory tract epithelial cells (Greve et al., 1989) and that O. can up-
regulate the ICAM receptor on nasal epithelial cells (Beck et al., 1994) suggest that more
studies are needed to explore more fully the potential interaction between O. exposure and
viral infectivity.
In a single study, human AM host defense capacity was measured in vitro in AMs
removed from subjects exposed to 0.08 and 0.10 ppm O. for 6.6 h while undergoing moderate
exercise. Alveolar macrophages from O.-exposed subjects had significant decrements in
complement-receptor-(but not antigen-antibody [Fc]-receptor)-mediated phagocytosis of
Candida albicans (Devlin et al., 1991). These data show that acute in vivo exposure of
humans to O. results in impairment of AM host defense capability, potentially resulting in
decreased ability to phagocytose and kill inhaled microorganisms in vivo. Human AMs also
have been exposed to O. in vitro to investigate whether changes in macrophage host defense
functions are due to a direct effect of O. on AMs or secondary effects resulting from lung
injury and inflammation. Becker et al. (1991) exposed AMs to 0.1 to 1.0 ppm O. in vitro for
1 h and showed a concentration-dependent decrease in phagocytosis of antibody-coated sheep
erythrocytes; a small increase in PGE.; and production of significantly lower levels of IL-1,
IL-6, and TNF on stimulation with lipopolysaccharide when compared with air-exposed cells
(Becker et al., 1991). Although the few studies in which animals have been exposed to virus
in conjunction with O. exposure provide some evidence to suggest that O. impairs the immune
system's ability to fight viral infections, there is insufficient human data to know whether
O. exposure affects viral infectivity. However, there is potential cause for concern that
O. may render humans and animals more susceptible to a subsequent bacterial challenge.
There are two studies that have investigated the effect of O. exposure on
mucociliary clearance of inhaled particles, with conflicting results. In one study (Foster et al.,
1987), seven male volunteers inhaled radiolabeled ferric oxide ('" Tc-Fe.O.) particles and then
were exposed to 0.2 and 0.4 ppm O. for 2 h while undergoing light IE. The investigators
observed a concentration-dependent increase in rate of particle clearance 2 h after exposure,
although increased clearance was confined primarily to the peripheral airways in subjects
exposed to 0.2 ppm O.. In the second study (Gerrity et al., 1993), 15 male or female subjects
were exposed to 0.4 ppm O. for 1 h while undergoing CE (40 L/min); 2 h after exposure,
subjects inhaled '" Tc-Fe.O. particles, and clearance was measured with a gamma camera for
the next 3 h and on the next morning. There was no difference in the clearance rate of
particles in air and O.-exposed subjects. The discrepancy between these studies may be
explained by differences in exposure protocol, time of particle inhalation, or time of clearance
measurement, or by the presence of cough immediately following O. exposure, which may
have accelerated clearance in the first study.
7.2.5 Extrapulmonary Effects of Ozone
It is still believed that O. reacts immediately on contact with respiratory system
tissue and is not absorbed or transported to extrapulmonary sites to any significant degree (see
Chapter 8). A number of laboratory animal studies presented in Chapter 6, however, suggest
7-87
-------
that reaction products formed by the interaction of O. with respiratory system fluids or tissues
may produce effects measured outside the respiratory tract—either in the blood, as changes in
circulating blood lymphocytes, erythrocytes, and serum, or as changes in the structure or
function of other organs, such as the parathyroid gland, the heart, the liver, and the central
nervous system (see Chapter 6, Section 6.3). Very little is known, however, about the
mechanisms by which O. could cause these extrapulmonary effects.
The results from human exposure studies discussed in the previous criteria
document (U.S. Environmental Protection Agency, 1986) failed to demonstrate any consistent
extrapulmonary effects (see Chapter 10, Section 10.6 of the 1986 document). Early studies on
peripheral blood lymphocytes collected from human volunteers did not find any significant
genotoxic or functional changes at O. exposures of 0.4 to 0.6 ppm for up to 4 h/day. Limited
data on human subjects available at the time the 1986 criteria document was published also
indicated that 0.5 ppm O. exposure for over 2 h caused transient changes in blood erythrocytes
and sera (e.g., erythrocyte fragility and enzyme activities), but the physiological significance
of these studies remains questionable. The conclusions drawn from these early studies raise
doubt that cellular damage or altered function is occurring to circulating cells at O. exposures
under 0.5 ppm.
Studies published since the publication of the previous criteria document (U.S.
Environmental Protection Agency, 1986) on the potential extrapulmonary effects of in vivo
O. exposure of human subjects have not been very definitive. Johnson et al. (1986) exposed
11 male nonsmokers to 0.5 ppm O. for 4 h on 2 consecutive days. When compared to air
controls, O. exposure did not result in any significant change in the activity of blood plasma D-
1-proteinase inhibitor. Schelegle et al. (1989) exposed 20 O.-sensitive, healthy young men to
0.20 and 0.35 ppm O. with heavy exercise (V. = 50 L/min). Plasma concentrations of PGF.n
were elevated after 40 and 80 min of exposure to the higher O. level (0.35 ppm). It is likely,
however, that the elevation of this ecosanoid in the blood was due either to increased
production or to decreased metabolism of PGF.n in the lung.
The demonstration in the previous section (Section 7.2.4) of an array of
inflammatory mediators and immune modulators released at the airway surface provides a
possible mechanism for effects to occur elsewhere in the body.
7.2.6 Ozone Mixed with Other Pollutants
Although it is well known that polluted air contains a large number of chemical
species, the most common approach to evaluating air pollution effects under laboratory
conditions has been assessment of responses consequent to exposure to single pollutants. This
has been the case for a variety of reasons, not the least of which is the problem inherent in
adequately controlling the concentrations of multiple pollutants simultaneously. Further,
atmospheric chemistry is very complicated, and it is difficult to adequately assess the exposure
mixture as the number of constituent pollutants increases. Observed effects may be related to
unknown reaction products, the monitored pollutants being only surrogates. Other problems
inherent in mixture studies involve considerations such as whether pollutants are presented
simultaneously or in sequential or overlapping patterns. Ideally, the selected pattern should at
least approximate one that occurs in the ambient environment. In spite of these difficulties,
information from mixture studies is important from the standpoint of attempts to better
understand responses of humans to the complex mixture of ambient air.
7-88
-------
The previous O. criteria document (U.S Environmental Protection Agency, 1986)
evaluated the limited database of information available on mixtures of O. with one or more
pollutants and concluded that pulmonary function changes were no more than additive and, in
most cases, were attributable to O. alone. Several new studies have since appeared in which
human subjects were exposed to mixtures of two or more pollutants or to individual pollutants
sequentially (Table 7-13), extending the database for controlled studies. Epidemiological
studies also have investigated mixtures of pollutants and have not found evidence suggestive of
synergistic effects (see Section 7-4).
7.2.6.1 Ozone and Sulfur-Containing Pollutants
Horvath et al. (1987) compared the pulmonary function responses of male subjects
(19 to 29 years of age) with normal baseline pulmonary function to four experimental
conditions: (1) FA, (2) 0.25 ppm O., (3) 1,200 to 1,600 Dg/m' H.SO. aerosol, and
(4) 0.25 ppm O. + 1,200 to 1,600 Dg/m' H.SO. aerosol. Exposures were completed in
random sequence, a minimum of 1 week apart, and were conducted at 35 DC and 83% RH.
Subjects alternated 20-min rest and exercise (V. = 30 to 32 L/min) periods throughout the 2-h
exposures. The results indicated that neither O. alone nor O. mixed with H.SO. aerosol had
significant effects on any pulmonary function, metabolic, or ventilatory parameter.
Koenig et al. (1990) evaluated sequential O. (0.12 ppm) and SO. (0.10 ppm)
exposures in 13 allergic, asthmatic adolescents (12 to 18 years of age). Three subjects used no
regular medications, the other 10 used one or more of beta-adrenergic agents, theophylline,
and antihistamines. All subjects had a PC., for methacholine of 10 mg/niL or less. Subjects
took their morning medication on experiment days if needed, but at least 4 h elapsed between
any medication use and the start of the experiment. The subjects participated in three
exposures at 22 DC and 75% RH, which were presented in random order and at least 1 week
apart. The three exposures were (1) air + SO., (2) O. + O., and (3) O. + SO.. The
mouthpiece exposures were 1 h in duration, during which the subjects breathed one test gas for
45 min, followed by a second gas for the final 15 min. Subjects exercised at a V. of about 30
L/min during the second and fourth 15-min segments of the exposure. Pulmonary functions
were measured 2 to 3 and 7 to 8 min postexposure. Changes in FEV. and R. were
significantly greater following the O. + SO. exposure than following the other two exposures.
Although the subject group was small, the results indicate that O. exposure may potentiate
responses to SO. exposure in asthmatic adolescents. It should be noted that the SO.
concentration (0.10 ppm) used in this study is a subthreshold level.
Linn et al. (1994) evaluated the pulmonary function and symptom responses of
15 atopic and normal subjects and 30 asthmatic subjects exposed to FA, 0.12 ppm O.,
100 Dg/m' respirable H.SO. aerosol (MMAD = 0.5 Dm), and a mixture of the two pollutants.
The chamber exposures were 6.5 h in duration, during which the subjects walked on a
treadmill (V. D 29 L/min) for 50 min of each hour. There was a 30-min lunch period
following the third hour. Pulmonary function and symptom responses were measured
preexposure and during the hourly 10-min breaks, and a methacholine bronchochallenge test
was performed following each exposure. Relative to responses to the FA exposure, H.SO.
7-89
-------
Table 7-13. Ozone Mixed with Other Pollutant^
-------
Table 7-13 (cont'd). Ozone Mixed with Other PoIIutant§
-------
Table 7-13 (cont'd). Ozone Mixed with Other PoIIutant§
Concentration1'
ppm
Dg/m
Pollute
Sulfur-Containing Pollutants
0.12
0.30
0.05
0.25
235
564
70
490
1,200 to
1,600
O3
NO2
H,SO4
HNO3
O3
H2S04
aerosol
Exposure Duration and
int Activity
(cont'd)
1. 5 h with IE for
2 consecutive days;
VE D 23.2 L/min
2h
IE
VE = 30 to 32 L/min
Exposure
Conditions0
T = 22 DC
RH = 65%
T = 35 DC
RH = 83%
Number and
Gender of Subject
Subjects 3 Characteristics
Asthmatic NS, 12 to 19 years old
adolescents;
22 completed
study;
15 M, 7F
9 M Healthy NS,
19 to 29 years old
Observed Effect(s)
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.
No significant effects of exposure to O3 alone
or combined with H2SO4 aerosol.
Reference
Koenigetal. (1994)
Horvath et al. (1987)
"See Appendix A for abbreviations and acronyms.
''Grouped by pollutant mixture.
CWBGT = 0.7 T,,,,,,, + 0.3 T,tv „,„,„„,„„.
-------
aerosol exposure alone induced no significant alteration in pulmonary function, symptoms, or
bronchial reactivity to methacholine. Exposure to O. alone (FEV. decrement of about 100 mL
compared to the FA response) or mixed with H.SO. aerosol (FEV. decrement of about 189 mL
compared to the FA exposure) induced significant decrements in forced expiratory function and
increased bronchial reactivity. Both effects were greater on the first of 2 consecutive days of
exposure. Group mean lung function and methacholine reactivity changes were somewhat
larger following O. + H.SO. aerosol compared to exposure to O. alone, but the differences
were, at best, marginally significant and usually nonsignificant, depending on the function
tested. However, there were a few individual subjects who showed significantly larger
pulmonary function decrements following the exposure to O. + H.SO. than following
exposure to O. alone. The authors concluded that O. is more important than H.SO. aerosol in
inducing pulmonary dysfunction in normal, atopic, and asthmatic adults. There does,
however, appear to be a more sensitive subpopulation that responds to O. + H.SO. aerosol
more strongly than the average adult.
Utell et al. (1994) reported on the pulmonary function and symptomology responses
of 30 healthy adults (18 to 45 years of age) and 30 allergic asthmatics (21 to 42 years of age)
who were exposed for 3 h to sodium chloride (NaCl) aerosol (100 Dg/m') or H.SO. aerosol
(100 Dg/m') and, 24 h later, to 0.08, 0.12, or 0.18 ppm O. for 3 h. The study was an
incomplete block design, in that each subject completed chamber exposures to each of two
O. concentrations following each of the aerosols (four of the possible six combinations per
subject). Out of the total number of subjects, 20 healthy and 20 asthmatic subjects completed
each of the possible exposure combinations. Subjects exercised for 10 min out of each half-
hour of exposure (V. = 4 times resting; 30 to 36 L/min). Environmental conditions averaged
21 + 1 DC and 40 + 5% RH. Ozone exposures were separated by at least 2 weeks. Healthy
subjects had no significant pulmonary function response (2.1% or less) to O. exposure,
regardless of the O. concentration or the aerosol preexposure. As a group, asthmatics had
mean decrements in FVC of 5% or greater in only a few cases: 7.6% following the NaCl/0.08
ppm O. combination, 6.3% following the NaCl/0.12 ppm O. combination, and 6.5% following
the H.SO./O.I8 ppm O. combination. No combination of aerosol and O. concentration induced
a decrement in FEV. of 5% or greater. Although the statistical analysis indicates that exposure
to H.SO. aerosol significantly altered the pattern of response and recovery to O. exposure on
the next day in asthmatics, the group mean data presented in the report show that functionally
there is little difference between the responses to the various exposures or in the time course of
recovery. The individual responses of the asthmatic subjects are reported to be more variable
than those of the healthy subjects. Asthmatic subjects reported more respiratory symptoms
than healthy subjects, but there was no dose-response relationship between O. concentration
and symptom intensity for healthy or asthmatic subjects. The variability of the responses of
the asthmatic subjects makes interpretation of these results difficult. Some of the asthmatic
subjects were reported to experience exercise-induced bronchospasm, and, without FA control
exposures, it is impossible to determine what, if any, portion of the asthmatic subjects'
response is related to exercise-induced bronchospasm, compared to that related to O. exposure.
Kagawa (1986) exposed Japanese men to three mixtures: (1) O. (0.30 ppm) + NO.
(0.30 ppm) + H.SO. (200 Dg/m'), (2) O. (0.15 ppm) + NO. (0.15 ppm) + H.SO.
(200 Dg/m'), or (3) O. (0.15 ppm) + NO. (0.15 ppm) + SO. (0.15 ppm)+ H.SO.
(200 Dg/m'). Exposures were 2 h in duration, and subjects exercised for a total of 20 min
during exposure 1 and for 60 min during exposures 2 and 3. Some of the subjects were
7-93
-------
smokers. Reported symptoms were attributed to O. exposure, whereas small decrements in
airway conductance (DlO%) were observed following exposures to mixtures 1 and 2. Although
the magnitude of the FEV. decrement is not stated, a possible decrease was observed after
exposure 3. The responses observed with these mixed exposure conditions were no different
than responses reported for exposures to similar concentrations of O., indicating no enhanced
response due to the presence of the other pollutants in the mixtures.
7.2.6.2 Ozone and Nitrogen-Containing Pollutants
Adams et al. (1987) reported on the responses of 20 males and 20 females (18 to
30 years of age), all healthy nonsmokers, exposed to (1) FA, (2) 0.3 ppm O., (3) 0.6 ppm
NO., and (4) 0.3 ppm O. + 0.6 ppm NO.. Subjects were exposed via mouthpiece for 1 h,
during which they exercised continuously at a V. of about 70 L/min for males and 50 L/min
for females. The exposures were presented in random order, a minimum of 5 days apart.
There were no differences in any pulmonary function (FEV. decrement of about 22%) between
the O. and NO. + O. exposures, except for SR.., which was lower following NO. +
O. (+7.3% for females and D9.6% for males) than following O. alone (+15.3% for females
and +4.0% for males).
Koenig et al. (1988) exposed 14 male and 10 female adolescents to FA, 0.30 ppm
NO., 0.12 ppm O., and 0.30 ppm NO. + 0.12 ppm O.. Twelve of the subjects were healthy
normals, and the other 12 were allergic asthmatics. The asthmatics, except for one who took
no regular medications, used one or more of beta-adrenergic agents, theophylline, and
antihistamines. Asthmatic subjects took their morning medications if needed, but refrained
from medication use for at least 4 h prior to the exposures. The mouthpiece exposures were 1
h in duration, during which the subjects exercised in 15-min periods (mean V. = 32.8 + 6.0
L/min), alternated with 15-min rest periods. No changes in any measure of pulmonary
function were observed in normal or asthmatic subjects following O. or NO. + O. exposure.
Sequential exposure to 0.6 ppm NO. or FA for 2 h, followed 3 h later by a 2-h
exposure to 0.3 ppm O. was investigated by Hazucha et al. (1994) in 21 healthy, nonsmoking
females (18 to 34 years of age). Subjects alternated 15-min periods of exercise (V. = 20
L/min/m' BSA) and 15-min rest periods while in the exposure chamber, and rested in ambient
air during the 3-h interexposure period. The 2 exposure days were separated by at least 2
weeks. Ambient conditions in the exposure chamber were 21 DC and 40% RH. Group mean
decrements following the FA/O. exposure sequence were DlO.8, D7.0, DlO.2, and Dl4.9% for
PEFR, FVC, FEV., and FEF , respectively. Following the NO./O. exposure sequence,
the group mean decrements were D14.5, D8.5, D12.0, and Dl9.5%, respectively, for PEFR,
FVC, FEV., and FEF Although small, the differences in FEV. and FEF between
the FA/O. and NO./O. exposure sequences were statistically significant. There were no
differences in the changes in SR.. or symptomology between the two exposure sequences. The
most striking finding of this study was that, although both exposure sequences increased
airway responsiveness to methacholine, responsiveness was potentiated by the
NO./O. exposure sequence compared to the FA/O. exposure sequence.
Aris et al. (1991) examined pulmonary function responses to a 3-h exposure to
0.2 ppm O. following a 2-h exposure to 0.54 mg/mL nitric acid (HNO.; volume mean
diameter = 6.0 + 0.2 Dm) or 0.55 mg/mL water (H.O; volume mean diameter = 6.47 + 0.4
Dm) fog. This is a common pattern of pollutant exposure in coastal California areas. Subjects
were 10 healthy adults 21 to 31 years of age; they were prescreened for a decrement of 10% or
7-94
-------
greater in FEV. following 3 h of exposure to 0.2 ppm O., during which they exercised for 50
min of each hour (V. =40 L/min). Decrements in FEV. following the screening O. exposure
ranged from 15 to 49%. The three exposures (FA + O., H.O fog + O., and HNO. fog + O.)
were presented in random order, and were separated by a minimum of 2 weeks. The authors
hypothesized that exposure to acidic fog, followed by O. exposure, would induce greater
decrements in FVC and FEV. than H.O fog or air followed by O. exposure. In fact, both
HNO. and H.O fog exposure seemed to ameliorate the effect of subsequent O. exposure on
FEV. and FVC, although only the difference between the FEV. responses to FA + O. (28.5%)
and H.O fog + O. (18.5%) was significant. Group mean comparisons of methacholine
responsiveness and O. responsiveness (defined as a minimum of a 10% decrement in FEV.
following the prescreening O. exposure) suggest that the subjects classified as O. sensitive, on
the average, had lower methacholine PC doses. The individual data, however, do not
always support this conclusion. Two of 10 O.-sensitive subjects had methacholine PC
concentrations above the author's cut-off for airway hyperresponsiveness, and 3 of 10
O. nonsensitive subjects had hyperreactive airways based on the authors' criteria for
methacholine PC
Aris et al. (1993b) further examined pulmonary responses to combined O. and
HNO. exposures. Ten healthy, nonsmoking adults, 19 to 41 years of age, were exposed to
FA, 500 Dg/m' of HNO. gas plus 0.2 ppm O., or to 0.2 ppm O. alone. The exposure protocol
was 4 h in duration, with 50 min IE at 40 L/min alternating with 10-min rest periods each
hour. Pulmonary function was measured during each rest period, whereas BAL, proximal
airway lavage, and bronchial biopsies were performed 18 h after completion of each exposure.
Mean FEV. and FVC decreased, and mean SR.. and respiratory symptom scores increased
across both the HNO. + O. and the O. exposures. The results indicated, however, that
HNO. combined with O. did not exacerbate the pulmonary function decrements or respiratory
symptoms caused by O. alone. Similarly, there were no statistically significant differences
between the HNO. + O. and the O. exposures in the cellular or biochemical constituents in
either the BAL or proximal airway lavage fluids or in the bronchial biopsy specimens. The
authors concluded that HNO. does not potentiate the inflammatory response produced by O. in
healthy individuals.
The objective of a study by Koenig et al. (1994) was to investigate possible
interactions between oxidants (0.12 ppm O. +0.30 ppm NO.) and an H.SO. aerosol
(70 Dg/m') with a mass median aerodynamic diameter (MMAD) of 0.6 Dm (+D. = 1.5).
Twenty-two adolescent allergic asthmatics who also had exercise-induced bronchospasm and a
positive response to a standardized methacholine bronchochallenge test completed all
exposures. Subjects inhaled FA, O. + NO., O. + NO. + H.SO., or O. + NO. + HNO.
through a mouthpiece for 90 min on 2 consecutive days. Each pair of exposures was separated
by at least 1 week. Subjects exercised (V. about 10 times FVC) and rested in alternating 15-
min periods. Pulmonary functions (FVC, FEV., V , V , and R.) were measured
before and after each exposure and on the day following the second consecutive exposure, at
which time only pulmonary function was evaluated and a methacholine bronchochallenge was
performed. Six additional subjects began the study, but dropped out before completion
because of uncomfortable symptoms associated with the exposures; none dropped out following
an FA exposure. There were no statistically significant changes in any measured parameter of
pulmonary function following the three pollutant-containing exposures, compared to the FA
exposure, contrary to expectations. (Also see Section 7.4 for related epidemiological studies.)
7-95
-------
7.2.6.3 Ozone, Peroxyacetyl Nitrate, and More Complex Mixtures
Horvath et al. (1986) exposed 10 healthy young women (19 to 36 years of age) to
(1) FA, (2) 0.48 ppm O., (3) 0.27 ppm PAN, and (4) 0.48 ppm O. + 0.27 ppm PAN. The
chamber exposures were 2 h in duration, during which subjects alternated 20-min exercise
periods (V. = 25 L/min) and 15-min rest periods. Exposures were completed in random
order and were at least 1 week apart. Exposure to PAN alone did not induce any significant
changes in pulmonary function. Both O. and PAN + O. exposure induced significant
decrements in FVC, FEV., and FEF ; however, the decrements following the
PAN + O. exposure were significantly larger (average of about 10%), suggesting interaction
between PAN and O.. It should be noted that typical peak ambient PAN concentrations are
about 0.05 ppm. Symptom reports indicated that O. + PAN exposure induced greater
subjective stress than did exposure to O. alone.
Drechsler-Parks et al. (1987b) exposed eight healthy young adults (mean age
24 years) to a mixture of 0.30 ppm PAN + 0.45 ppm O. on 5 consecutive days to evaluate
possible attenuation. Subjects were reexposed to the PAN + O. mixture on the third and
seventh days following the last consecutive day of exposure. Attenuation occurred with the
same pattern and time sequence as has been reported for O. alone. The largest group mean
decrements occurred following the second exposure, and the subjects became progressively
less responsive with subsequent exposures. Two subjects failed to return to baseline values
with 5 consecutive days of exposure. Pulmonary function changes after the follow-up
exposures indicated that the attenuation response is relatively short lived, in that it began to
abate within 3 to 7 days following the fifth consecutive day of exposure. These results are
consistent with those of similar studies using exposure to O. alone (Horvath et al., 1981; Kulle
et al., 1982), suggesting that PAN had no additional effect on attenuation to O.. A greater
number of symptoms was reported following all PAN + O. exposures than following exposure
to O. alone.
Drechsler-Parks et al. (1989) studied 16 older men and women (51 to 76 years of
age) and 16 young men and women (19 to 26 years of age) who each completed 2-h chamber
exposures to FA, 0.45 ppm O., and mixtures of 0.45 ppm O. with 0.60 ppm NO. and/or 0.13
ppm PAN. Subjects alternated 20-min exercise (V. about 25 L/min) and rest periods.
Exposure to O. alone and in all combinations induced significant decrements in FVC (14 to
17%), FEV. (19 to 22%), and FEF (28 to 30%) in the younger group. In the older
group, these same three variables were significantly decreased only with NO. + O. exposure
(7.3% for FVC, 8.4% for FEV., and 12% for FEF ). Exposure of the older subjects to
PAN + O. induced significant decrements only in FVC (4.2%) and FEV. (8.3%). The PAN
+ NO. + O. exposure induced a significant decrement only in FVC (6.4%) in the older
subjects. All subjects reported more symptoms following the mixture exposures than following
exposure to O. alone. These pulmonary function results following the exposure to O. + PAN
are in contrast to those reported by Drechsler-Parks et al. (1984) and Horvath et al. (1986) on
young adults exposed to 0.45 ppm O. + 0.30 ppm PAN. The results of both earlier studies
suggested an interaction between O. and PAN, in that pulmonary function decrements
following the mixture exposure were approximately 10% larger than those following exposure
to O. alone, whereas there were no significant pulmonary function effects with exposure to
PAN alone. A likely explanation for this discrepancy is that the PAN concentration used by
Drechsler-Parks et al. (1989) was slightly less than half that used by Drechsler-Parks et al.
(1984) and Horvath et al. (1986). Thus, if the additional effect of PAN is linear, an additional
7-96
-------
effect of PAN + O. would be expected to be less than 5 %, which probably would not be
detected because it is within the variability of the pulmonary function measurements. In any
case, ambient PAN concentrations are considerably less than 0.13 ppm. This indicates that,
even if PAN and O. do interact in some way in their effects on pulmonary function, at typical
ambient concentrations of O. and PAN, effects can be attributed to O. alone.
7.2.6.4 Summary
Information on interactive effects between O. and other pollutants remains sparse at
this time. However, it is clear that O. is responsible for the largest share of observed effects
when subjects are exposed to the mixtures of O. and other pollutants that have been studied to
date. There is no evidence that simultaneous exposure of healthy individuals to ambient
concentrations of O. plus NO., PAN, H.SO., HNO., or SO. results in significant interaction.
However, Aris et al. (1991) have reported that HNO. and H.O fog exposure ameliorates the
pulmonary function effects of a subsequent O. exposure. Koenig et al. (1990) found that
preexposure to O. induced significant pulmonary function decrements in allergic asthmatic
adolescents following a sequential SO. exposure. Both the O. and SO. concentrations were at
subthreshold levels for the experimental design used.
Both studies that reported potentially significant effects have involved sequential
exposure protocols, in contrast to the simultaneous exposure protocols, which generally have
not shown effects beyond those that would be expected at the O. concentration used. It may be
that certain preexposures predispose an individual to responses following a subsequent
exposure; however, this question remains far from being resolved. Further, these results are
related only to spirometry and plethysmography and may not be applicable to other possible
endpoints.
7.3 Symptoms and Pulmonary Function in Controlled
Studies of Ambient Air Exposures
Controlled O. exposure studies under a variety of different experimental conditions
have generated a large amount of informative exposure-effects data. However, complete
laboratory simulation of the pollutant mix present in ambient air is impossible on practical
grounds. Thus, the exposure effects of one or several artificially generated pollutants (i.e.,
a simple mixture) on symptoms and lung function may not be comparable to those in ambient
air where complex mixtures of pollutants likely exist. This section reviews two types of
studies that utilize a mobile laboratory or a hypobaric chamber to investigate the acute effects
of O. during exposures to ambient air or altitude, respectively. These studies can be designed
to determine the independent effects of O. as well as possible interactions among many
pollutants and other conditions present in typical ambient air.
7.3.1 Mobile Laboratory Studies
Quantitatively useful information on the effects of acute exposure to photochemical
oxidants on symptoms and pulmonary function originated from field studies using a mobile
laboratory, as presented in the previous criteria document (U.S. Environmental Protection
Agency, 1986). These studies offer the advantage of studying the effects of ambient air on a
local subject population by combining the experimental methods of both epidemiology and
7-97
-------
controlled-exposure studies. Field studies using mobile exposure chambers involve subjects
exposed to ambient air, FA without pollutants, or FA containing artificially generated
concentrations of O. that are comparable to those measured in the ambient environment. The
exposure air can also be conditioned to a desired temperature and humidity. As a result,
measured health responses in ambient air can be compared to those found in more artificial or
controlled conditions. The mobile laboratory shares many of the same limitations of stationary
exposure laboratories (e.g., limited number of both subjects and artificially generated
pollutants for testing). Ambient air studies in the mobile laboratory are dependent on ambient
conditions, which can be unpredictable, uncontrollable, and not completely characterizable.
Logistical problems (space, power, and locations with local interfering outdoor conditions)
limit access to many ambient pollution sites of interest.
As summarized in Table 7-14, investigators at the Rancho Los Amigos Medical
Center in California used a mobile laboratory and demonstrated that respiratory effects in
Los Angeles residents are related to O. concentration and level of exercise (Linn et al., 1980,
1983b; Avol et al., 1983, 1984, 1985a,b,c, 1987). Such effects include pulmonary function
decrements at O. concentrations of 0.144 ppm in healthy exercising adolescents (Avol et al.,
1985a,b) and increased respiratory symptoms and pulmonary function decrements at
0.153 ppm in heavily exercising athletes (Avol et al., 1984, 1985c) and at 0.174 ppm in lightly
exercising normal and asthmatic subjects (Linn et al., 1980, 1983b). The observed effects
were typically mild, and generally no substantial differences were seen between asthmatic and
nonasthmatic subjects. Postexposure pulmonary function decrements appeared to last several
hours longer in the asthmatics, but no statistical test was reported for this difference (Avol
et al., 1983; Linn et al., 1983b). The medication status of the asthmatic subjects during the
studies was not reported, although medications were temporarily withheld prior to exposures.
The subjects' clinical severity typically was mild, based on their baseline lung function and
exercise capability. Many of the normal subjects with a history of allergy appeared to be more
responsive to O. than "nonallergic" normal subjects (Linn et al., 1980, 1983b), although a
standardized evaluation of atopic status was not performed. Direct comparative studies of
exercising athletes (Avol et al., 1984, 1985c) with chamber exposures to oxidant-polluted
ambient air (mean O. concentration of 0.15 ppm) and purified air containing a controlled
concentration of generated O. at 0.16 ppm showed no significant differences in lung function
and symptoms, suggesting that coexisting ambient pollutants had minimal contribution to the
measured responses under the typical summer ambient conditions in Southern California.
Effects of copollutants in other regions of the country remain to be investigated with the
mobile laboratory. These field studies emphasize the importance of adequate characterization
of subjects and the ambient air, exercise levels, duration of exposure, and individual variations
in sensitivity in interpreting observed exposure effects. Although these factors need to be
investigated over a wider range of experimental conditions, the results from these field studies
are, so far, consistent with those from controlled human exposure studies. Short-term
respiratory effects of summer ambient oxidant pollution in Southern California are
predominantly, if not entirely, caused by ambient O. in typical healthy or asthmatic residents,
according to mobile laboratory studies (Avol et al., 1984, 1985c). Overall, the symptoms and
decrements in lung function were generally modest and, while statistically significant in some
cases, were probably not clinically significant.
7-98
-------
7-99
-------
Table 7-14. Acute Effects of Ozone in Ambient Air in
Field Studies with a Mobile Laboratory*
Mean Ozone
Concentration*
6ppm
0.113
+ .033
0.144
+ .043
0.153
+ .025
vj
o
o
0.156
+ .055
0.165
+ .059
Dg/m3
221
+ 65
282
+ 84
300
+ 49
306
+ 107
323
+ 115
Ambient Activity
Temperature' Exposure Level
(DC) Duration ( VE) Number of Subjects
Observed Effect(s) Reference
33+1 1 h CE (22 L/min) 66 healthy children, 8 toNo significant changes in Avol et al. (1987)
1 1 years old forced expiratory function and
respiratory symptoms after
exposure to 0. 1 13 ppm O3 in ambient
air.
32+1 1 h CE (32 L/min) 59 healthy adolescents,
12 to 15 years old
32 + 2 1 h CE (53 L/min) 50 healthy adults
(competitive bicyclists)
33 + 4 1 h CE (38 L/min) 48 healthy adults,
50 asthmatic adults
33 + 3 1 h CE (42 L/min) 60 "healthy" adults
(7 were asthmatic)
Small significant decreases in FVC Avol et al. (1985a,b)
(D2.1%), FEV0.75 (D4.0%), FEV,
(D4.2%), and PEFR (D4.4%) relative
to control with no recovery during a
1-h postexposure rest; no significant
increases in symptoms.
Mild increases in lower respiratory Avol et al. (1984, 1985c)
symptom scores and significant
decreases in FEV, (D5.3%) 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 Linn et al. (1983b)
symptom score or forced expiratory Avol et al. (1983)
performance in normals or
asthmatics; however, FEV, remained
low or decreased further (D3%) 3 h
after ambient air exposure in
asthmatics.
Small significant decreases in FEV, Linn et al. (1983b)
(D3.3 %) and FVC with no Avol et al. (1983)
recovery during a 1-h postexposure
rest; TLC decreased and DN2
increased slightly.
-------
Table 7-14 (cont'd). Acute Effects of Ozone in Ambient Air in
Field Studies with a Mobile Laboratory*
Mean Ozone
Concentration*
ppm Dg/m3
0.174 341
+ .068 + 133
Ambient Activity
Temperature' Exposure Level
(DC) Duration ( VE) Number of Subjects
33+2 2 h IE (2 times 34 "healthy" adults,
resting) 30 asthmatic adults
at 15-min intervals
Observed Effect(s)
Increased symptom scores and small
significant decreases in FEV, (D2.4%),
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
Linn et al. (1980, 1983b)
"See Appendix A for abbreviations and acronyms.
bRanked by lowest level of O3 in ambient air, presented as the mean + SD.
'Mean + SD.
-------
7.3.2 High-Altitude Studies
Symptoms and pulmonary function resulting from exposure to O. in commercial
aircraft flying at high altitudes and in altitude-simulation studies were reviewed in the previous
criteria document (U.S. Environmental Protection Agency, 1986). Attention has focused on
the health effects in flight crew, specifically flight attendants because of their physical activities
at altitude and exposure patterns to peak levels of cabin O.. The most quantitatively useful
information was based on a series of hypobaric studies of normal nonsmoking subjects who
were exposed to 1,829 m (6,000 ft) and O. at concentrations of 0.2 and 0.3 ppm for 3 or 4 h
(Lategola et al., 1980a,b). Increased symptoms and pulmonary function decrements occurred
at 0.3 ppm but not at 0.2 ppm under light exercise conditions. However, the exposure
conditions did not reflect higher (peak) O. concentrations reported to occur in certain aircraft
at high altitudes or the higher cabin altitudes attained by new-generation commercial aircraft.
No reports have appeared subsequently in the literature that specifically study the
health effects of aircraft cabin O.. However, O. levels were reported to be very low (average
concentration 0.01 to 0.02 ppm) during 92 randomly selected smoking and nonsmoking flights
in 1989 (Nagda et al., 1991). None of the flights exceeded the time-weighted average standard
of 0.10 ppm (during any 3-h interval) promulgated by the U.S. Federal Aviation
Administration, perhaps related to the use of O.-scrubbing catalytic filters (Melton, 1990).
However, in-flight O. exposure is possible because catalytic filters are not necessarily in
continuous use during flight.
7.4 Field and Epidemiology Studies
7.4.1 Acute Effects of Ozone Exposure
7.4.1.1 Introduction
Field and epidemiology studies addressing the acute effects of O. on lung function
decrements and increased morbidity and mortality in human populations involve those
combinations of environmental conditions and copollutant and activity levels present under
real-world conditions of O. exposure. This real-world relevance is an advantage over animal
or human chamber studies. Thus, results of such studies are essential components of an
understanding of overall effects of O.. However, the conditions under which epidemiologic
studies are carried out cannot be controlled in the same way that they can in experimental
studies. Parameters that may be difficult or impossible to estimate or control outside the
laboratory include actual O. exposures, levels of temperature, RH, allergens, correlated
pollutants other than O., and breathing rates and activity patterns of subjects. Variations in
these factors can be important sources of variability in data and results and may, under certain
conditions, lead to biases (e.g., confounding) in results. These and other issues of importance
in the interpretation of epidemiology study results are discussed in the sections below.
The limitations of epidemiologic studies of O. health effects noted above were
highlighted in the previous O. criteria document (U.S. Environmental Protection Agency,
1986), which reached the conclusion that, because of such factors, epidemiologic studies on the
acute effects of O. on lung function available at that time did not provide information that is
quantitatively useful in the standard-setting process. Since publication of the 1986 O. criteria
document, however, results have become available from a substantial number of well-
conducted, individual-level studies and aggregate-level, time-series studies. New statistical
7-102
-------
techniques also have become available to deal with confounders that were not appropriately
considered when the data were first analyzed. In the following sections, the more recent
studies and reanalyses of older studies will be evaluated collectively.
7.4.1.2 Individual-Level Studies
The studies discussed in this section fall into three main categories: (1) summer
camp studies, (2) exercise studies, and (3) daily life studies. Summer camp studies involve
collection of sequential (usually daily) data on lung function, respiratory symptoms, and
environmental conditions over the course of 1 or 2-week attendance at camps. Exercise studies
are unique in that lung function and respiratory symptoms are measured before and after each
of a series of discrete exercise events in the presence of ambient air pollution. Daily life
studies measure lung function, respiratory symptoms, and exacerbation of existing respiratory
diseases, along with environmental variables at regular intervals in the course of normal daily
activities of a population. These include studies of healthy adults and school-aged children as
well as studies of individuals with preexisting disease (e.g., asthma). Medication use also may
be monitored in asthmatics. Studies of this kind that focus on exacerbation of asthma
symptoms usually have been referred to as panel studies.
The important differences among the three study types relate primarily to issues of
exposure assessment. Because subjects usually are out-of-doors or in well-ventilated cabins,
exposure estimation errors are minimized in camp and exercise studies. In contrast, larger
exposure estimation errors may occur in daily life studies. Camp and daily life studies enable
assessment of the effects of cumulative O. exposures, whereas exercise studies limit attention
to rather brief exposures. Exercise studies offer the potential of assessing individual V. values
and O. concentrations during the relevant exposure period, whereas such assessments are more
difficult in camp and daily life studies.
Although the study designs differ in some ways, the central design feature of all of
these study types is the collection of repeated measurements on individuals. This feature is
exploited in data analysis by having each subject serve as his or her own control. For
continuous outcomes such as lung function, subject-specific linear regressions are usually
performed with lung function (or change in lung function) as the outcome variable and O. or
other environmental factors as the explanatory variable. The regression slope is a measure of
individual lung function response to O.. The mean slope across individuals often is used as a
measure of the average population response. A more statistically valid approach involves
computing the mean slope with weighting proportional to the inverse variances of the
individual slopes. An alternative approach has been to use analysis of covariance methods to
fit a population-pooled slope and separate, subject-specific y-intercepts. To date, no studies
have used nonlinear (e.g., quadratic) models in relating lung function decrements to
O. exposures, which in chamber studies have been shown to better describe the functional
relationship between O. exposures and lung function decrements (see Chapter 9,
Section 9.3.4).
Issues in the Interpretation of Individual Level Studies
The most basic question affecting the interpretation of acute O. epidemiology
studies is whether (and if so, to what extent) the associations observed between O. and
decreased pulmonary function are causally related to O. and not merely due to confounding by
some other factor (e.g., temperature, allergens, time trends in spirometry, or other pollutants).
7-103
-------
By definition, a confounder is an unmeasured or unaccounted-for variable that has an effect on
the measured outcome and also is correlated with O. concentrations. Variables that satisfy
only one of these two conditions are not confounders. For example, a variable that affects
lung function but is independent of O. would add variation to lung function measurements but
would not confound an O./lung function analysis, and a variable that correlates with O. but
does not directly affect lung function (in the range of measurements) would not confound an
analysis of O. effects. Other variables might modify the effect of O. on lung function, thereby
increasing or decreasing the O. effect under the conditions of study. Epidemiologists refer to
this latter phenomenon as "effect modification". The presence of effect modification does not
bias the results of a study, but can provide insights into the range of effect magnitudes (e.g.,
slope of lung function decrement on O. level) that occur under varying environmental
conditions.
Ambient air temperature often exhibits a moderate to high correlation over time
with O. in acute epidemiology studies due, in part, to the dependence of O. formation rate on
light intensity. Among the studies reviewed in this section, correlations ranging from DO.06 to
0.90 (mean = 0.51) have been reported. Correlations between O. and RH, when reported,
have been in the range DO.4 to DO.6. Several human chamber studies have examined the
possible direct effects of temperature and RH on lung function independent of O., with
somewhat mixed results (Stacy et al., 1982; Folinsbee et al., 1985; Eschenbacher et al., 1992).
Two studies reported increases in FEV. at high temperature (30 and 37 DC) and 60% RH
(Stacy et al., 1982; Eschenbacher et al., 1992), whereas the other reported no effect on FEV.
at 35 DC, and a decrease at 40 DC (Folinsbee et al., 1985). Referring to results of acute
O. epidemiology studies, Eschenbacher and colleagues (1992) concluded, based on their own
results, that "the associations found between ambient O. and daily changes in ventilatory
function cannot be attributed to the heat and humidity stress often associated with high
O. concentrations." Temperatures observed in the epidemiology studies reviewed in the
present section primarily have been below 30 DC, with occasional peaks as high as 35 DC. It
should be noted that subjects studied epidemiologically usually will have had an opportunity to
acclimate to ambient temperatures prior to or soon after the start of the study. In any event,
given the laboratory findings, a significant confounding role for temperature in these studies
seems unlikely. The possibility that changes in ambient temperature may introduce biases in
measured lung volumes (e.g., through inaccurate correction of volumes to body temperature) is
an issue that deserves further study.
Exposure to specific allergens can influence lung function in individuals who have
diseases characterized by IgE-mediated, Fc interactions (i.e., atopy) and may also affect
individuals who have an atopic tendency (e.g., as assessed by positive prick skin test or serum
levels of total IgE) without diagnosed clinical disease. Raizenne et al. (1989) detected positive
reactions to one or more allergens by skin prick in 49% of 96 young nonasthmatic females
enrolled in a summer camp study. Few data are available on the correlation between O. and
allergen levels during acute epidemiology studies. However, because both variables to some
are extent influenced by weather patterns, some correlation seems likely. Thus, a possible
confounding role of airborne allergens in such studies cannot be ruled out. Because of the
specific nature of individual antigen sensitization and uncertainty regarding the full set of
relevant allergens in a given setting, attempts to measure and to control statistically for allergen
levels on a group level in epidemiology studies may not be very effective.
7-104
-------
The potential effects of time trends in spirometry due to training effects are also of
concern. There have been several recent studies that have looked at time trends in serial lung
function measurements (mainly FEV. and PEFR) independent of air pollution effects (Raizenne
et al., 1989; Avol et al., 1990; Hoek and Brunekreef, 1992). In each case, average FEV.
measurements have been observed to decline steadily over the first few measurements and then
to stabilize or recover slightly to a flat pattern. Average FVC measurements follow a similar
pattern. In contrast, PEFR often has been observed to increase steadily over successive
measurements. Similar patterns have been observed in studies with intervals between lung
function measurements ranging from 12 h to 1 week. The consistency of these observations
across studies suggests that they represent real phenomena that should be recognized in
designing and analyzing studies involving repeated lung function measurements. However,
time trends will result in confounding of O. effects only if, by chance, the trend correlates with
temporal variations in O. concentrations. Such chance correlations could be either positive or
negative and, if present, would have a larger impact (i.e., produce an undesirable degree of
confounding) on studies in which all subjects begin the study simultaneously and have few
follow-up measurements. Studies that focus on daily changes in lung function may be less
impacted by this phenomenon.
It is also important to consider the roles of other pollutants as possible confounders
or effect modifiers. In the studies to be reviewed in this section, most copollutants (e.g., SO.,
NO., sulfate, and acid aerosols) were present at levels well below those that have produced
lung function decrements in healthy subjects following short-term exposures in chamber studies
(see Section 7.2.6). In contrast, an extensive and growing database is available from chamber
studies documenting the independent acute effects of ambient-level O. on lung function (see
Section 7.2.1). Although direct lung function effects of other pollutants at typical ambient
concentrations seem unlikely, it has been suggested that the effects of O. may be potentiated by
coexposure or previous exposure to other pollutants, most notably acid aerosols (Spektor et al.,
1988b). Some data from animal studies suggest interactive effects of O. and acid exposures for
certain pulmonary outcomes (see Chapter 6). However, to date, analyses directed towards this
phenomenon in field studies of human lung function (via analysis of the relationship between
acid aerosol levels and residuals from regressions of lung function on O.) have proven negative
(Spektor et al., 1988a,b). That is, after controlling for the influence of O., no significant
association between acid aerosol peaks and lung function decrements has been observed. Acid
aerosol episodes, which often occur coincident with high O. levels in the summer in the
northeastern United States, may extend for several hours or days. However, the possible
potentiating effects of prolonged acid peaks on O. effects are still poorly understood. Some
recent epidemiological studies (Pope et al., 1991; Pope and Dockery, 1992; Koenig et al.,
1993; Roemer et al., 1993) have reported significant associations between lung function
decrements and ambient particulate matter (PM) concentrations. Although no supporting
evidence for lung function effects due to ambient-level particle exposure is yet available from
human or laboratory animal controlled studies, the possibility of some confounding by particles
in the studies reviewed here cannot be ruled out.
In epidemiologic studies, activity levels are difficult to control and to measure,
although this varies with study type (see below). Chamber studies have shown clearly that
lung O. doses and associated functional effects increase as a function of physical activity level
(Hazucha, 1987). Epidemiologic study designs often have been chosen that result in relatively
7-105
-------
high subject activity levels (exercise studies and camp studies), but generally the studies have
been carried out without quantitative information on V. distributions across subjects and time.
Variations in activity levels will introduce variability in the relationship between
personal exposure and personal dose. If this variability occurs primarily between subjects, it
will result in differing O. doses to people exposed to the same O. level and yield differences in
response that may be misinterpreted as O. sensitivity variations. If variability in activity levels
occurs over time for a given subject, it will add error to the functional relationship linking lung
function and O. exposure. In either case, the influence of activity level variability is to add
dose estimation error (or misclassification). Estimates of V. based on heart rate measurements
can be derived using subject-specific calibrations under representative ranges of exercise levels
and types (Samet et al., 1993; Raizenne and Spengler, 1989). However, the utility of such V.
estimates for reducing dose-misclassification errors in acute O. epidemiology studies has not
yet been demonstrated (Kinney, 1986; Spektor et al., 1988b; Raizenne and Spengler, 1989).
This is partly due to the logistical difficulties associated with collecting accurate data and also
may be due to the fact that, for a given subject, V. variations across days are usually small in
comparison to O. concentration variations. The same issues arise in the context of O. exposure
misclassification in "daily life" studies (see below), where outdoor O. concentrations are used
to estimate exposures of subjects who spend substantial amounts of time indoors during the
period over which lung function measurements take place.
Camp Studies of Lung Function in Children
Summer camp studies provide the most extensive and reliable information on the
acute pulmonary effects of O. under natural conditions. Camp studies involve the collection of
sequential (usually daily) data on lung function on each of a large number of children, along
with concurrent measurements of O. exposures and other environmental factors over the course
of a single-week or multiweek summer camp. Data analyses usually consist of estimating the
linear association between lung function and environmental variables on an individual basis
(allowing each subject to serve as his/her own control) and then testing the mean population
association for statistical significance. As noted, summer camps offer the significant advantage
that subject exposures are especially well estimated because they are based on on-site, outdoor
O. monitoring. In addition, these studies assess the pulmonary effects of natural diurnal
patterns of O. exposures, which often involve broad daytime peaks.
Since the last O. criteria document (U.S. Environmental Protection Agency, 1986),
eight camp studies have been reported. Design characteristics and results are summarized in
Table 7-15. Six of these studies have focused solely on normal (i.e., nonasthmatic) children,
one focused on asthmatics exclusively (Thurston et al., 1995), and one used both normal and
asthmatic children (Raizenne et al., 1987). Although methods and results varied somewhat
across studies, this group of studies collectively provides substantial evidence for associations
between ambient O. exposures, together with other pollutants, and acute decrements in lung
function. Interpretation of these associations as causal is supported by evidence of biological
plausibility. For example, the well-documented direct effects of O. on lung function in human
chamber studies; the evidence, also from chamber studies, indicating a lack of direct effects of
other collinear environmental factors (e.g., temperature and acid aerosols) at the levels at
which these factors occur in the camp studies; exposure-response relationships; and
consistency across studies all provide strong support. Camp studies involving asthmatic
7-106
-------
children generally have yielded lung function/O. associations that are similar in absolute
magnitude to those observed in nonasthmatics (Raizenne et al., 1987;
7-107
-------
Table 7-15. Acute Effects of Photochemical Oxidant Pollution:
Lung Function in Camp Studies*
Pollutants/Environmental Variables
Study Description
Results and Comments
Reference
Hourly O3 ranged from DlO to 110 ppb. SO2,
NOX, O3, SO4-, H2SO4, pH, PM10, PM25, RH,
temperature, barometric pressure, and wind
speed and direction also were measured.
Effects of pollutants and other environmental variables on
symptoms and lung function were examined in children
attending a summer camp at Lake Couchiching, about 100
km north of Toronto, ON. Study was conducted June 30
through July 8, 1983; n = 52, 23 nonasthmatics
(11 males, 12 females) and 29 asthmatics (16 males,
13 females), avg. age = 12.1 years. Symptom
questionnaire and function tests given twice daily to each
child between 7:30 and 9:30 a.m. and 4:30 and 6:30 p.m.
Children's activity levels not estimated.
Strongest association between lung function and
environmental variables was in nonasthmatics, with
FVC decrements significantly correlated (p < 0.01)
with lagged-avg SO4, PM2J, and temperature.
Unlagged PEER significantly correlated with 1 h O3.
Also, significant association of temperature with all
lung function indices in nonasthmatics, but not in
asthmatics. Coefficient of variation stable across
morning and evening tests.
Raizenne et al. (1987)b
O
CO
1-h O3 ranged from < 10 to 143 ppb; max 1-h
O3 > 100 ppb on 14 days of total study
(6 weeks). For other pollutants and variables
measured, see Raizenne et al. (1987) because
same protocol used here as in that study.
(a) Effects of pollutants and other environmental variables
on lung function were examined in girls attending one of
three 2-week Girl Guide camp sessions on the north shore
of Lake Erie. Cohort (n = 104) screened by MC and
skin-prick tests for 10 common respiratory allergens;
five asthmatics withdrawn from the study (n = 99). Lung
function tests administered twice daily. Children's activity
levels not estimated.
(b) Subset of 12 girls (7 MC + , 5 MC-) studied pre- and
postexercise on 1 low-pollution (control) day and
1 peak-pollution day (episode, 1 h O3 > 139 ppb,
SO4" > 80 Dg/m3).
(a) Associations between aerometric data and lung
function measurements were not reported by pollutant
in this reference. Aggregate analysis for full study not
reported. Lung function changes reported for 5 episode
days only. FEV, decrements statistically significant on
2 episode days for methacholine- nonresponsive
subjects.
(b) Group mean FVC increased postexercise in the
n = 12 subset by 40 mL, 71 mL in MCD and 17 mL in
MC + . Pollution effect not statistically significant.
Raizenne et al.
(1987, 1989)b
Continuous 1-h O3, SO2, NO2, and acid
aerosols (as H2SO4); 1-h O3 range =
40-143 ppb; max 12-h acid particle
concentration = 28 Dg/m3 in one episode;
FP SO4- 100 Dg/m3 for peak hour.
Time-activity model used to evaluate likely cumulative
(6 h) O3 and H2SO4 exposures/doses experienced by
children in above Lake Erie Girl Guide camp study,
summer 1986. See Raizenne et al. (1987, 1989) for
protocol and related information. Dosimetry model was
developed for relating heart rate (from a 12-min, graded,
cycle ergometer test) to ventilation and then to O3 and
H2SO4 concentration. Also, five randomly selected
children wore portable heart-rate monitors, providing data
for use in the dosimetric model.
Application of the dosimetry model used to estimate
individual 6-h cumulative doses for O3 and H2SO4
exposures on 1 control and 1 episode day indicated
negative trend in lung function (PEER) as cumulative
dose increased for both O3 and H2SO4, although slopes
for each did not differ significantly from zero (p >
0.10).
Raizenne and Spengler
(1989)b
-------
Table 7-15 (cont'd). Acute Effects of Photochemical Oxidant Pollution:
Lung Function in Camp Studies*
Pollutants/Environmental Variables
Study Description
Results and Comments
Reference
Max 1 h O3 ranged from 40 to DlOO ppb, with
max 1 h > 80 ppb on 9 of 27 days of O3
recorded. O3, SO4 H2SO4, PM15, PM25,
temperature, humidity, and wind speed and
direction measured. Levels not reported for
SO2, pH, NO3, and NH4+.
Effects of pollutants and other environmental variables on
respiratory functions in 91 children (53 boys, 38 girls; ages
8-15) attending 2 to 4 weeks of summer camp at Fairview
Lake, NJ. Subsets were n = 37 for all 4 weeks, n = 34 for
first 2 weeks only, n = 20 for last 2 weeks only. Symptom
questionnaire; FVC, FEV,, MMEF, and PERF (by
spirometry) were measured once each test day (most of days
in camp) sometime between 11:00 a.m. and 6:30 p.m. All
children had validated spirometric data for D7 days of their 2-
or 4-week camp stay. Activity levels of the children were not
estimated. Respiratory health status determined by parental
questionnaire only. Children slept in screened-in shelters but
otherwise were exposed to ambient air 24 h/day. Average
regression slopes for respiratory function vs. max 1-h O3
concentration reported for the full cohort, for boys and girls
separately, and for subsets in attendance for all 4 weeks and
for respective 2-week sessions. Regressions also repeated for
data below 80 and 60 ppb 1-h O3, and for data with THI <78
DF.
Average regression slopes (±SE) were Dl.03 ± 0.24
and Dl.42 + 0.17 mL/ppb for FVC and FEV,,
respectively; and D6.78 ± 0.73 and D2.48 ±
0.26 mL/s/ppb for PEFR and MMEF, respectively.
Most slopes of regression significant at p < 0.05
(differences from zero). Not clear if slopes for data
subsets significantly different from each other (e.g.,
function vs. O3 < 60 ppb and function vs.
O3 < 80 ppb). No formal analysis performed for
possible concentration threshold.
Spektor et al. (1988a)b
-------
Table 7-15 (cont'd). Acute Effects of Photochemical Oxidant Pollution:
Lung Function in Camp Studies*
Pollutants/Environmental Variables
Study Description
Results and Comments
Reference
Maximal 1 h O3 concentrations ranged from
approximately 40 to 150 ppb over the course
of the study. 12-h average aerosol acidity
measurements ranged from near 0 to
18.6 Qg/m3 (H2SO4 equivalent). Temperature
and RH measured, but levels not reported.
THI reached a maximum of 81 DF. All
environmental measurements were made on
site.
Effects of O3 and other environmental variables on lung
function studied in a group of 46 children (13 girls,
33 boys; ages 8-14) at a 4-week, 1988 summer camp in
southwestern New Jersey (Fairview Lake). Same location
used in previous camp study by same investigators.
Subjects had no history of lung diseases or atopy. Two
lung function measurement periods each day (a.m. and
p.m.) along with collection of respiratory symptom data.
Data collected during or after periods of rain were
excluded from analysis. Results for FVC, FEV,,
FEV,/FVC, FER^*, and PEFR reported. Linear day-of-
study trends were examined for lung function. Subject-
specific linear regressions were performed relating lung
function in a.m., p.m., and p.m. D a.m. differences to O3
averaged over various periods. Average slopes across
subjects were tested for significant differences from zero.
Regressions were repeated after excluding days with O3 at
or above 120 ppb. Regression residuals were tested for
correlation with THI and H+ concentrations.
No significant linear day-of-study effect seen for any of
the lung function variables tested, but the linear model
may not have been optimal for testing this effect. In a
subset of 35 subjects with at least 2 consecutive days of
lung function measurements, mean regression slopes of
a.m. lung function variables on previous-day mean or 1-h
maximum O3 were all significantly negative (e.g., mean
slope of FEV, on 1-h maximum O3 was
DO.50 ± 0.12 mL/ppb). These results suggest a possible
carry-over effect from previous-day O3 exposures. In the
full set of 46 subjects, regressions of p.m. lung function
on previous-hour O3, maximum 1-h O3 for same day, or
average O3 for day were significantly negative in most
cases (e.g., mean slope of FEV, on previous-hour O3 was
Dl.60 ± 0.30 mL/ppb). All regressions of the p.m. D
a.m. lung function differences on intervening O3
concentrations were significantly negative (e.g., mean
slope of FEV, on mean O3 between a.m. and p.m.
measurements was DO.63 [±0.09] mL/ppb). No
correlation seen between regression residuals and THI or
H+ concentrations, indicating there was no remaining
effect of these variables with lung function after
accounting for O3. However, no models were fit that
included O3 and these variables simultaneously, nor were
interaction effects tested for. The strong and consistent
associations between lung function decrements and
O3 concentrations in this study contrast with results
reported from studies in Canada and California at similar
levels of O3.
Spektor et al. (1991)
Spektor and Lippmann
(1991)
-------
Table 7-15 (cont'd). Acute Effects of Photochemical Oxidant Pollution:
Lung Function in Camp Studies*
Pollutants/Environmental Variables
Study Description
Results and Comments
Reference
1-h O3 preceding lung function measurements
ranged from 25 to 245 ppb. Pollutants
measured on site included O3; NO, (range:
0 to 40 ppb), SO2 (range: 1 to 8 ppb); and fine
(mean = 23.9 Qg/m3), coarse (mean = 36.6
Qg/m3), and total (mean = 59 Qg/m3) PM]0
mass. Temperature averaged 21.5 DC (range:
13.5 to 25.5 DC) and RH averaged 43.3%.
Effects of O3 and other environmental factors on lung
function examined in 43 children (24 female, 19 male;
ages 7-13) attending 1 of 3 sequential weeks (three
subjects stayed an additional week) of summer camp in the
San Bernardino Mountains of California. Camp was at
5,710 ft above sea level. Lung function measured by
spirometry up to three times daily on each subject;
analytical measures included FVC, FEV,, and PEER. No
report of respiratory data derived from questionnaires.
Subject activity levels prior to lung function testing were
not characterized. Campers slept in well-ventilated cabins.
Subjects came mostly from homes in the Los Angeles
Basin, and thus were likely to have been exposed to high
O3 levels prior to camp. Simple linear regression models
were fit on an individual basis (subject-specific slopes) and
by pooling across individuals (common population slope)
to determine the linear relations between the three lung
function variables and various O3 metrics (1-h average
preceding hour of spirometry, 1-h average 2 h previous to
hour of spirometry, or 6-h average preceding spirometry).
The common slope model was repeated separately for
morning, noon, and evening lung-function measurements,
and separately for data with 1-h O3 levels above and below
120 ppb. Multiple regression models were fit that
included O3 along with temperature, RH, and coarse and
fine PM mass.
The population-pooled regression slopes (±SE) of FVC
and FEV, on previous hour O3 were D0.40 (±0.10) and
D0.38 (±0.09) mL/ppb, respectively (p < 0.0001 in
both cases); for PEER, the regression slope was DO. 13
(±0.36) mL/s/ppb (not significant). Similar, though
slightly more negative, slopes were obtained using 2-h
and 6-h average O3. Interpretation of differences across
the three O3 metrics is substantially hampered by the high
correlations among them
(r D 0.90). When temperature, RH, and coarse and fine
PM mass were included with O3 in multiple regression
models, the O3 slopes increased in absolute magnitude to
D0.68 (±0.16) and D0.76 (±0.15) mL/ppb for FVC and
FEV,, respectively, and to Dl.91 (±0.63) for PEER.
Technical problems with the temperature sensor in the
first week of the study did not appear to influence these
results. Data were split on the basis of whether or not
the maximum 1-h O3 concentration in the 6 h preceding
spirometry was above 120 ppb. Regression slopes
relating lung function and previous 1-, 3-, and 6-h
average O3 were more negative in the high concentration
stratum. This result is consistent with the nonlinear
(e.g., quadratic) relationships between lung function and
O3 exposure observed in chamber studies. Because
levels of pollutants other than O3 were quite low (NO,
and SO,), and/or were uncorrelated with O3 levels (PM),
the regression results reported from this well-conducted
study are likely to represent real influences of O3 on lung
function.
Higgins et al. (1990);
Gross et al. (1991)
-------
Table 7-15 (cont'd). Acute Effects of Photochemical Oxidant Pollution:
Lung Function in Camp Studies*
Pollutants/Environmental Variables
Study Description
Results and Comments
Reference
No
Daily maximum O3 concentrations ranged from
approximately 60 to 160 ppb (derived visually
from figure presented in paper). Other
pollutants measured on site included SO2, NO,,
CO, total hydrocarbons, and size-segregated
PM mass. Aside from O3, all gaseous
pollutant levels reported to be very low (data
not presented). 24-h TSP concentrations
ranged from 18 to 54 Qg/m3. Airborne allergen
data collected. Temperature ranged from 10 to
15 DC at night and from 25 to 35 DC during
day. RH ranged from 30 to 45% at night and
from 5 to 20% during day.
Effects of O3 and other environmental variables on lung
function examined in 293 children (139 girls, 154 boys;
ages 8-17) attending one of six 1-week camp sessions at a
summer camp located in the mountains near Idyllwild,
CA, 190 km southeast of Los Angeles (altitude: 1,570 m).
Lung function measured twice daily on each camper (a.m.:
0730 to 0930; p.m.: 1600 to 1930). Analyses presented
for FVC, FEV,, PEER, and FEF,j_75,5. Symptom
questionnaires completed prior to each test. Used repeated
measures analysis of variance model to test for day-of-
study and a.m./p.m. effects on lung function independent
of pollution concentrations. Linear regressions of
morning, afternoon, and
p.m. D a.m. difference of lung function on O3 were
performed with simultaneous control of day and a.m./p.m.
effects. Upper and lower quartiles of distribution of
individual FEV,/O3 regression slopes were examined with
respect to subject characteristics. Changes in FEV, over
several days analyzed in relation to intervening integrated
O, concentrations.
Significant day-of-study effect observed for FVC and
FEV, characterized by steady drop over first few days of
measurement, followed by partial reversal later in week.
For PEER, p.m. measurements were significantly higher
than a.m. measurements. Controlling for day and
a.m./p.m. effects, the authors reported that no consistent
O3 effects on lung function were observed. The a.m.
lung-function measurements had a significant positive
correlation with O3 averaged over the previous 1, 8, or
24 h. The p.m. measurements reported to have no
correlation with O3. The p.m. D a.m. lung function
differences were negatively correlated with previous 8-h
average O3 concentrations, but not with previous 1-h O3
concentrations. No quantitative results reported for the
above lung function/O3 findings. There were no
discernable differences between subjects in the upper and
lower quartiles of the distribution of individual regression
slopes of a.m. D p.m. FEV, difference on previous
1 h O3. Regressions of change in FEV, over several days
(four separate intervals ranging from approximately 8 h
to approximately 80 h) with integrated O3 concentrations
yielded negative slopes ranging from DO.41 to Dl.46
mL/ppb, one of which was statistically significant. The
time-trends in FVC and FEV, measurements observed in
this study are qualitatively consistent with those seen in
some other summer camp studies. The lack of consistent
negative slopes relating lung function with
O3 concentrations contrasts with other, eastern U.S.,
summer camp studies at similar O3 levels.
Avol et al. (1990)
Avol et al. (1991)
-------
Table 7-15 (cont'd). Acute Effects of Photochemical Oxidant Pollution:
Lung Function in Camp Studies*
Pollutants/Environmental Variables
Study Description
Results and Comments
Reference
O3 data collected at a site 8 mi from camps.
Daily 1-h O3 maxima ranged from
approximately 40 ppb to approximately
200 ppb. 12-h aerosol acidity concentrations
ranged between 14 and 360 neq/m3.
Temperature and RH data obtained from a
nearby site.
Report of data collected during two simultaneous summer
camps located 2 mi apart in central New Jersey in 1988.
34 subjects were studied, including 20 camp counselors
(ages 14-35) and 14 campers (ages 9-13). Study spanned
19 days. Spirometry and respiratory symptom data
collected each afternoon. Analysis of FVC, FEV,, and
PEFR in relation to O3 and temperature using linear
regression within camps and subject types (i.e., counselors
vs. campers).
Regressions of lung function on 1-h and 8-h average
O3 within several subject subsets yielded inconsistent
results, with some mean slopes apparently significantly
positive, and one negative mean slope, highlighted by
authors, of borderline significance (p < 0.10).
Berry et al. (1991)
Daily 1-h maximum O3 concentrations ranged
from 70 to 160 ppb in 1991 and from 10 to
63 ppb in 1992. On-site measurements also
made for acid aerosols (approximately 20 to
110 nmoles/m3 in 1991 and 15 to 55 nmoles/m*
in 1992) and temperature (between 21 and
32 DC over 2 years).
Effects of O3 and acid aerosols on peak flow, respiratory
symptoms, and medication usage in asthmatic children
evaluated at two 1-week summer camps (June of 1991 and
1992) in the Connecticut River Valley. Fifty-two and 55
subjects were studied in 1991 and 1992, respectively,
ranging in age from 7-13. Peak flow measured twice daily
(approximately 9:00 a.m. and 5:00 p.m.). Combining
data from the two studies, individual regressions of daily
change in FEV, on O3 or H+ concentrations were
performed.
In subjects without asthma exacerbations during the
camps, statistically significant, negative mean slopes
were found relating QPEFR and O3 or H+ concentrations.
The correlation between these two pollutants was not
reported. The mean slopes were D2.3 (±0.7) mL/s/ppb
for O3, and Dl.2 (±0.6) mL/s/nmol/m3 for H+. In the
case of O3, a scatter-plot with QPEFR demonstrated an
apparently linear trend. In contrast, the H+ regression
results appeared to be driven entirely by one data point.
Thurston et al. (1995)
U>
"See Appendix A for abbreviations and acronyms.
''Cited in U.S. Environmental Protection Agency (1992).
-------
7-114
-------
Thurston et al., 1995); however, the health significance of a given drop in FEV. may be
greater for those with preexisting, compromised respiratory function.
Although similar study designs have been employed in most of the camp studies
summarized in Table 7-15, differences in analytical methods have made quantitative
comparisons between studies difficult to interpret. In particular, it has not been clear to what
extent differences in results across studies may be due to differences in study characteristics
(e.g., O. effect potentiation by other pollutants and activity levels) as opposed to differences in
data analysis methods.
For better comparison in this document, data from six of the camp studies
summarized in Table 7-15 were reanalyzed using uniform analytical methods. For each study,
afternoon lung function data (FEV.) were regressed on concurrent 1-h O. concentrations using
an analysis of covariance model that included subject-specific intercepts and a single, pooled
O. slope. Although intersubject variation in responses to O. would be expected on the basis of
controlled chamber study results (see Section 7.2), a common-slope model was chosen for this
analysis because emphasis was placed on estimating the average response in each study
population. The study-specific slopes computed with this model ranged from DO. 19 to Dl.29
mL/ppb across the six studies (Table 7-16). All but one of these slopes were statistically
significant (p < 0.02). When data for all six studies were pooled, a slope of DO.5 mL/ppb was
observed. The slope from the 1988 Fairview Lake, NJ, study (Dl.29 mL/ppb) was greater in
absolute magnitude than the slopes from the other studies (which ranged from DO. 19 to DO.84
mL/ppb). Overall, however, these pooled results indicate a quantitative consistency among
studies that is not as readily apparent in the absence of the combined analysis.
Table 7-16. Slopes from Regressions of Forced Expiratory
Volume in One Second on Ozone for Six Camp Studies
Study Name
Fairview Lake, 1984
Fairview Lake, 1988
Lake Couchiching
Lake Erie
San Bernardino Mountains
Pine Springs Ranch
All studies
Slope + SE (mL/ppb)b
D0.50
Dl.29
DO. 19
DO. 29
DO. 84
D0.32
D0.50
+ 0.16
+ 0.27
+ 0.44
+ 0.10
+ 0.20
+ 0.13
+ 0.07
p- Value
0.002
0.0001
0.66
0.003
0.0001
0.013
< 0.0001
Reference
Spektor et al. (1988a)
Spektor et al. (1991)
Spektor and Lippmann (1991)
Raizenne et al. (1987)
Raizenne et al. (1987, 1989)
Higgins et al. (1990)
Gross et al. (1991)
Avol et al. (1990, 1991)
"For each study, data were analyzed in one regression model that included a pooled O, slope and separate
subject-specific intercepts. See Appendix A for abbreviations and acronyms.
bSlope is the weighted mean of six study-specific slopes. The SE is the weighted SE of mean slope.
It is not clear why the 1988 New Jersey study yielded a larger slope than the other
studies. Possible explanations include greater subject activity levels (resulting in higher
O. doses at a given exposure level), potentiation of the O. effect by other pollutants (such as
acid aerosols), the relative absence of O. tolerance in the New Jersey study, or confounding by
7-115
-------
airborne allergens. There are no firm data on activity levels across the six studies. Thus,
whereas this factor surely contributes to the random variability within and between studies, it
is not known whether activity levels were substantially and systematically higher in the 1988
New Jersey study. Potentiation of the O. effects on lung function in asthmatics by acid
aerosols has been demonstrated in a chamber study in which O. exposure was administered
1 day following a 3-h exposure to 100 Dg/m' H.SO. (Utell et al., 1994). Although the
relevance of these data to the nonasthmatic subjects who experienced much lower acid levels at
northeastern summer-camps is not clear, they do demonstrate that potentiation can occur
between these pollutants. However, this factor alone cannot explain the observed differences
across camp results, because a camp study in southern Ontario (Raizenne et al., 1989), which
yielded relatively low FEV. slopes on O., experienced sulfate aerosol levels that were
comparable to those seen in New Jersey. Similarly, whereas tolerance due to prior exposures
to high O. levels has been suggested as an explanation for the smaller slopes seen in the
California studies, a lower subject activity level has been suggested to explain the smaller
slopes in southern Ontario. Data have not been reported on comparative levels of airborne
allergens during the various camp studies. None of the subjects in the 1988 New Jersey study
reported a history of asthma or atopy, minimizing the likelihood of confounding by airborne
allergens. However, given the lack of allergen data and the potential for substantial numbers
of "silent hyper-responders" (Raizenne et al., 1989), this possibility cannot be completely
discounted. Thus, no one factor seems adequate to explain the differences in results across
studies. Quite possibly, these differences reflect the combined influence of several of the
factors discussed above. Indeed, given the many possible sources of camp-to-camp variability,
it is surprising that results are as consistent as they are across six studies by three investigative
groups.
Several investigators have reported regression results for 1-h average O. and for
longer averaging times (e.g., 6 to 8 h) (Higgins et al., 1990; Avol et al., 1990, 1991; Spektor
et al., 1988a, 1991). In general, similar results have been obtained regardless of the averaging
time. Attempts to draw conclusions regarding the relative importance of short-term peaks and
longer term averages from such analyses have been hampered by the high degree of correlation
between 1-h and multihour averages. Until better analytical methods are found for dealing
with this problem, comparative results will remain difficult to interpret.
Lung Function in Exercising Subjects
This subsection discusses studies involving lung function measurement immediately
before and after a series of discrete outdoor exercise activities in the presence of air pollution.
This design is similar in principle to the ambient chamber studies conducted in the early 1980s
(see Section 7.3), in which subjects exercised under a specified protocol in a chamber
ventilated with ambient air. Here, however, there is typically less control imposed over
exercise duration and intensity, and less assessment of achieved V.. Compensating to some
extent for this diminished control is the relative ease of collecting numerous repeated
measurements at varying ambient O. levels for the same subjects, improving the precision of
concentration-response estimation. In contrast to camp studies, duration of relevant
O. exposure is assumed to be known, as it is defined by the length of each exercise event.
Results from five exercise studies (Selwyn et al., 1985; Spektor et al., 1988b; Hoek
and Brunekreef, 1992; Hoek et al., 1993a; Braun-Fahrlander et al., 1994; Brunekreef et al.,
1994) are summarized in Table 7-17. One of the studies (Selwyn et al., 1985) was discussed
7-116
-------
in the previous O. criteria document (U.S. Environmental Protection Agency, 1986) but is
reviewed again here because of its apparent consistency with the more recent study of Spektor
et al. (1988b).
Certain design variations across studies are worth noting. In the Houston study,
each of 24 recreational runners performed spirometry before and after a series of
approximately 28 runs on a track from late spring to early fall (Selwyn et al., 1985). Each run
was 3 mi long, and each subject attempted to maintain a similar heart rate across all runs.
Minute ventilation was not assessed. In the study carried out in Tuxedo, NY, adult runners
and walkers were allowed to choose their own exercise level and duration, but again were
encouraged to maintain a steady heart rate for the duration of the study (Spektor et al., 1988b).
Minute ventilation of each subject while running was estimated by measurement of V. during a
treadmill test that achieved a heart rate typical of that subject's experience while running. In
the study of 128 Swiss school children (Braun-Fahrlander et al., 1994), 10-min exercise
periods on a cycle ergometer were utilized on four to six occasions over a 6-mo period.
7-117
-------
Table 7-17. Acute Effects of Photochemical Oxidant Pollution:
Lung Function in Exercising Subjects3
Pollutants/Environmental Variables
Study Description
Results and Comments
Reference
1-h O3 concentration ranged from 21 to
124 ppb, max THI = 78 DC; max acidic
aerosol (as H2SO4) = 9 Dg/m3 during study.
SO2, NOX, PM,5, PMjj, SO4~, NO3, NH4+,
temperature, and RH measured but not
reported.
Effects of O3 on respiratory function and symptoms
examined in 30 nonsmoking adults (2 of 10 non-Caucasian
females) exercising almost daily outdoors (Tuxedo, NY)
for 15 to 55 min (average ca. 30 min) from July to early
August 1985. Pre- and postexercise lung function
measured, and questionnaire answered postexercise. Pulse
rate, calibrated to VE indoors, taken postexercise.
Exercise regimen self-selected but constant for each
subject over the course of study. Dosimetry estimated and
linear regressions done for pulmonary function changes vs.
(1) mean O3 concentration during exercise and (2) inhaled
O3 dose. Persistence of effects tested by linear regressions
of before-exercise lung function on previous-day O3 during
exercise. Subjects screened only by questionnaire; two
with previous history of asthma but asymptomatic.
Significant (p < 0.01) decrements in FVC, FEV,,
PEER, FEF^j,,, and FEV,/FVC associated with O3.
For example, the mean slope of DFEV, on O3 across all
subjects was Dl.35 mL/ppb (+0.35). No persistence of
effects seen. No symptoms reported by subjects. Mean
decrements showed unexpected inverse relationship with
calculated VE levels, as indicated by regressing
pulmonary function changes and postexercise function
against inhaled O3 during exercise. VE ranges given, but
not group or subset means. Subjects not screened for
atopy. Exercise done in Sterling Forest, wooded
research park, on paved roads or trails.
Spektor et al. (1988b)
15-min peak O3 measured during runs averaged
47 ppb (range: 4 to 135 ppb). Ambient T
averaged 29.4 DC (range: 18.0 to 37.8 DC).
RH averaged 62.6% (range: 37.0 to 88.0%).
Levels of other pollutants were low, median
values were SO2, 3 ppb; NO2, 6 ppb; FP, 10
Dg/m3. Median of subject-specific correlations
of O3 and RH correlated was -0.42.
Effects of O3 on lung-function change during running
outdoors were examined in 24 conditioned, recreational
runners (6 women, 18 men, ages 29-47) at a track 30 mi
southeast of Houston, TX, from May to October, 1981.
All runs were 3 mi in length, and each subject performed
at a near-constant heart rate for the duration of the study.
An average of 28 runs completed by each subject during
the study. Spirometry carried out before and after each
run, with analysis of FVC, FEV,, FEF^,,,, and FEF02_
, 2L. Change in each lung function variable was regressed,
for each subject, on 15-min maximum O3 measured during
the run. The mean slope across subjects was tested for
significance. The regression was repeated with
temperature and RH in the model.
Mean slope of FEV, on O3 alone was D0.4 mL/ppb (p =
0.03). In regressions that included temperature and RH,
the O3 slope dropped to D0.07 (not significant).
Although temperature reached high levels during the
study, a substantial direct effect of temperature or RH on
lung function, relative to that of O3, seems unlikely. A
possible potentiating role of high temperature and RH on
VE, and corresponding O3 dose, cannot be ruled out.
Lung function effect observed in simple O3 model seem
likely to be a valid reflection of O3 effects under varying
environmental conditions.
Selwyn et al. (1985)
-------
In contrast to these studies, early investigation of O. effects in The Netherlands
involved lower and more variable exercise levels, without any specific attempt to control
exercise intensity (Hoek and Brunekreef, 1992; Hoek et al., 1993a). Here, children engaged
in sports training and skills development activities that were characterized by the investigators
as low to moderate in intensity. Lung function change after exercise was assessed using peak
flow meters. Later studies in The Netherlands investigated the effects of heavy exercise levels
of variable duration in amateur cyclists, but lung function was evaluated by spirometry
(Brunekreef et al., 1994).
Although the designs varied somewhat, O. exposure levels were similar in most of
the studies: in Houston, 15-min peaks while running varied from 4 to 135 ppb; in Tuxedo, 1-h
O. levels ranged from 21 to 124 ppb; and in The Netherlands, 1-h maxima on study days
ranged from 10 to 120 ppb. The Swiss study observed O. levels between 20 and 80 ppb during
the exercise period.
The studies in adults (Selwyn et al., 1985; Spektor et al., 1988b; Brunekreef et al.,
1994) involving fairly intense exercise yielded statistically significant mean slopes of DFEV.
(i.e., FEV. after exercise minus FEV. before exercise) regressed on O. levels measured during
exercise, whereas the studies in children did not. The mean slope observed in the Tuxedo
study across all subjects was Dl.35 mL/ppb (+ 0.35), but was reduced to DO.55 mL/ppb ( +
0.45) in the group of 10 runners who achieved the highest V. values (> 100 L/min) during
exercise. The mean slope reported from the Houston study was similar to the latter number,
DO.4 mL/ppb (+ 0.16). The large effect level observed in the Tuxedo study led Spektor et al.
(1988b) to speculate that O. effects may have been potentiated by other pollutants such as acid
aerosols; however, this phenomenon was not demonstrated analytically from the available acid
monitoring data. In the Houston study, the O. effect became small and nonsignificant when
temperature and RH were added to the model. Effects of temperature and O. on lung function
were highly correlated and hard to separate in the Dutch amateur cyclists (Brunekreef et al.,
1994), although adjustment for humidity did not change the findings. Given the available
knowledge base on the independent effects of O. and temperature on lung function, it seems
reasonable to interpret the results from these studies as demonstrating acute effects of low
concentrations of ambient O. on lung function with moderate to heavy exercise. The
predominantly negative findings
7-120
-------
7-121
-------
Table 7-17 (cont'd). Acute Effects of Photochemical Oxidant Pollution:
Lung Function in Exercising Subjects3
Pollutants/Environmental Variables
Study Description
Results and Comments
Reference
1-h maximum O3 concentrations during study
ranged from 50 to 240 Qg/m3 (25 to 120 ppb).
The highest 4-h average PM2 5 level was
70 Qg/m3, the highest 4-h average sulfate
concentration was 21 Qg/m3, the highest 24-h
average NO2 concentration was 51 Qg/m3.
Temperature data were collected but levels
were not reported.
The relationship between lung function change and
O3 exposures during outdoor exercise examined in a
population of 83 children (43 girls, 40 boys; ages not
given) in Wageningen, The Netherlands. Study covered
the period from late May to mid-July, 1989. Lung
function assessed using hand-held peak-flow meters before
and after various outdoor, sports-training exercises lasting
approximately 1 h. Change in PEFR regressed on O3, O3
x exercise duration, and temperature for each subject; and
distribution of slopes were examined. Postexercise PEFR
analyzed in relation to same and previous day 1-h
O3 maximum, and temperature. Analyses repeated in
subsets of subjects with varying levels of correlation
between O3 and temperature during their series of exercise
events.
For 55 children with at least four sets of before and after
exercise peak flow measurements, the mean slope of the
PEFR change on O3 during exercise was 0.035 (±0.030)
mL/s/Qg/m3. For 65 subjects with at least four
postexercise measurements, the mean slope of PEFR on
previous-hour O3 was 0.080 (±0.023), which is
statistically significant, but in the nonplausible direction.
Adjustment for temperature resulted in negative mean
slopes, but these are difficult to interpret because of the
high statistical correlation between same-day O3 and
temperature (r = 0.86). Exercise events were of low
intensity as compared with chamber studies and with the
Tuxedo runners study (Spektor et al., 1988b).
Significant exposures may have occurred prior to the
exercise period. H+ levels were low (< 5 Qg/m3) as
measured simultaneously at three other nonurban sites in
The Netherlands. The possibility of a physical effect of
temperature on mini-Wright peak flow meter
measurements was noted by authors.
Hoek and Brunekreef
(1992)
Hoek et al. (1993a)
No
No
1-h maximum O3 concentrations during the
exercise period ranged from 0.02 to 0.08 ppm.
The highest pollutant levels measured during
the study period were 0.13 ppm for 1-h mean
O3 and 70 Qg/m3 for the mean NO2. No
measurements of particulates were available.
The acute effects of ambient O3 on lung function were
examined in 128 Swiss children, aged 9 to 11 years, after
10 min of outdoor exercise on a cycle ergometer (60 W).
Study covered the period from May through October 1989.
Changes in lung function were regressed on current O3
concentration, with or without adjustment for temperature,
RH, and other factors.
Elevated O3 levels were significantly associated with
decreased peak flows (PEFR), but not FVC or FEV,,
after exercise. The average adjusted regression slope for
PEFR was Ql.14 mL/s/ppm. This corresponds to an
average decrease in PEFR of 07.8 and Qll.7 at 0.08 and
0.12 ppm O3, respectively. The significant association
for PEFR, but not FVC or FEV,, is not consistent with
other studies. The low O3 levels and short exercise
period raise a question of plausibility regarding the
results.
Braun-Fahrlander et al.
(1994)
-------
Table 7-17 (cont'd). Acute Effects of Photochemical Oxidant Pollution:
Lung Function in Exercising Subjects3
Pollutants/Environmental Variables
Study Description
Results and Comments
Reference
During exercise, the maximum hourly O3
concentration averaged 87 Dg/m3 (0.04 ppm),
with a range of 26 to 195 Dg/m3 (0.01 to 0.10
ppm). Temperature averaged 17.9 DC, with a
range of 7.1 to 30.2 DC. NO2 and SO,
concentrations were low; 24-h averages were
26.0 and 7.5 Dg/m3, respectively. No
measurements of PM]0 were made.
The relationship between lung-function change and O3
exposure was investigated in 23 amateur cyclists, 18 to 37
years of age, during training sessions and races between
June 4 and August 18, 1981, in The Netherlands. Lung
function was measured with spirometry 30 min before and
between 10 and 60 min after cycling in rural locations.
Acute respiratory symptoms were recorded in a diary
before and after exercise. The difference between pre- and
post-exercise lung function was regressed on the mean O3
concentration during exercise. Time trend, pollen,
ambient temperature, and absolute humidity were taken
into account as potential confounders. Regression slopes
were pooled, and mean and median slopes were calculated.
The effect of O3 during exercise on mean symptom scores
was determined by a logistic regression model; all
coefficients were converted to estimated odds ratios.
Lung function was negatively related to O3 concentration
during exercise; effects were stronger in midsummer than
in the late summer. Mean regression coefficients were
Dl.16 ± 0.33, D0.52 + 0.26, D2.96 ± 1.06, and 0.44 +
0.46 mL/s/Dg/m3 for FVC, FEV,, PEE, and FEE^,,,
respectively. For all but FEFS_7J,5, the mean coefficients
were significantly different from zero. Adjustments for
air humidity resulted in slightly more negative
coefficients for FEV, and PEE. Acute respiratory
symptoms of shortness of breath, chest tightness, and
wheeze were positively related to O3.
Brunekreefetal. (1994)
"See Appendix A for abbreviations and acronyms.
No
U>
-------
of the studies in children are more difficult to interpret, but may be related to the low exercise
intensities achieved, low exposures, and, perhaps, associated O. tolerance that occurred prior
to the exercise period under study or to some subtle effect of confounders on the peak flow
measurements.
Lung Function in Daily Life Studies
This set of studies is characterized by the assessment of lung function, respiratory
symptoms, and environmental factor associations in the course of people's daily lives. This
section discusses only the lung function data from these studies. For logistical reasons, studies
of this kind usually have involved either spirometry conducted at regular intervals (every 1 to 3
weeks) in schools (Kinney et al., 1989; Castillejos et al., 1992; Hoek et al., 1993b) or self-
administered peak flow measurements in subjects of various ages over various periods (Vedal
et al., 1987; Krzyzanowski et al., 1989). Although daily life studies have the worthwhile goal
of characterizing air pollution effects on respiratory health in the real world, they suffer from
significant exposure assessment uncertainties owing to the use of outdoor O. monitoring, the
incomplete and variable penetration of O. indoors, and the preponderance of time spent indoors
by study subjects. This problem is probably less severe for the studies involving
schoolchildren, who often spend substantial time outdoors after school, when O. levels may be
elevated. Indeed, three of the school-based studies have found statistically significant
associations between lung function and previous-day O. levels (Castillejos et al., 1992; Kinney
et al., 1989; Hoek et al., 1993b). Another difficulty in interpreting the results of these studies
is the possible role of seasonal factors (e.g., pollens, epidemics of respiratory infection,
changes in activity patterns) as potential confounders of the analyses.
In addition to these general limitations inherent in the study design, several of the
studies summarized in Table 7-18 have other problems that limit their utility for assessing
O. effects on lung function. The study of Vedal et al. (1987), although well conducted, took
place from September through May, a period when O. levels generally are low, and other
potential respiratory insults may dominate. The statistical significance of results from a study
carried out in Tucson, AZ, is difficult to interpret because of the multiple statistical tests
performed (Krzyzanowski et al., 1989).
The remaining studies, although subject to the general criticisms noted previously,
provide suggestive evidence that ambient O. may play a role in short-term lung function
declines among children engaged in their normal daily routines (Kinney et al., 1989;
Castillejos et al., 1992; Hoek et al., 1993b). The Mexico City study of Castillejos et al.
(1992) is especially noteworthy because of the novel observation of FEV. and FEF
decrements that were strongly related to O. levels averaged over 24 to 168 h previous to
spirometry, but not to previous-hour O. levels. The strength of these associations (measured
by the ratio of the regression slope to its standard error) increased steadily as averaging time
increased. Ozone levels observed throughout this 6-mo study were high by U.S. standards; 1-
h average O. concentrations in the hour preceding lung function measurements ranged from 14
to 287 ppb, with a mean of 99 ppb. The authors suggested these results may reflect an
inflammatory response in the airways rather than the well-known acute physiological response.
However, further studies will be necessary to test this hypothesis.
Panel Studies of Symptom Prevalence
7-124
-------
Many field and epidemiological studies reviewed both in the last criteria document
(U.S. Environmental Protection Agency, 1986) and in the previous section of this document
7-125
-------
Table 7-18. Acute Effects of Photochemical Oxidant Pollution:
Daily Life Studies of Lung Function and Respiratory Symptoms'
Pollutants/Environmental Variables
Study Description
Results and Comments
Reference
Max O3 (1-h) concentrations ranged from 3 to
63 ppb. Other ambient pollutants measured
were NO,, TSP, IP, RSP, FP, SO,, and FP
SOj.
Lung function measured by spirometry for 154 children
ages 10-12 years (90 males, 64 females) in Kingston and
Harriman, TN. Spirometry done between 10 a.m. and
1 p.m. on up to 6 days at least 1 week apart during
February to April 1981. Child-specific linear regression
models of FVC, FEV0.75, MMEF, and V,,,ax75,, fit on 1-h
O3 max and 24-h FP and FP SOJ. Means ± SD of
distributions of estimated child-specific slopes computed
and tested for significance by f-test.
Significantly negative mean slopes on O3 for all lung
function variables. For example, mean slope of FEV07J
on O3 was DO.99 mL/ppb (+0.36). Among regressions
on FP and FP SOj, only one statistically significant mean
slope (i.e., positive mean slope of MMEF on FP).
Results insensitive to outlier audits and inconclusive for
sensitivity variation. Association between fitted slopes
and individual characteristics not significant. Low O3
levels raise plausibility questions.
Kinney (1986)b
Kinney et al. (1989)1'
No
1-h average O3 concentrations in hour
preceding spirometry ranged from 14 to
287 ppb, with mean of 99 ppb. No other
pollutants measured. Temperature ranged
from 3.9 to 27.8 DC. RH ranged from 18.9 to
92.3%.
Effects of O3 on lung function examined during regular
school hours in a group of 148 children (65 girls, 83 boys;
ages 7-9) from three schools in Mexico City. Spirometry
and symptom data (cough/phlegm) collected between 0800
and 1400 hours every 2 weeks over the period January
through June 1988. To account for lung growth over the
study period, residuals from lung function prediction
equations were used in analyses. Analyses limited to 143
subjects with at least seven valid measurements. Schools
were not air conditioned, and windows were usually open.
Schools and subject residences all within 5 km of O3
monitoring site. Associations between O3 and lung
function (FVC, FEV,, and FEF2J_7J,5) examined by
computing the weighted mean of subject-specific
regression slopes relating these variables. Various O3
averaging times (from 1 h to 168 h) were tested. After
analyzing population as a whole, regressions were repeated
in subject subsets defined by sex, report of chronic
symptoms, and maternal smoking. Overall regressions
repeated with temperature and RH in model with O3.
Only FVC had a statistically significant negative mean
slope in relation to previous hour O3 concentration
(D0.059 ± 0.23 mL/ppb). This slope is approximately
one order of magnitude lower than those observed in
some camp studies. Both FEV, and FEF2J_7J,5 had
significant negative associations with O3 averaged over
the previous 24, 48, and 168 h. For example, the mean
slope of FEV, on 48-h average O3 was D0.592 + 0.109
mL/ppb. The authors speculated that the FVC result
reflects the acute, reversible effects of O3 on one's ability
to take a deep breath, whereas the FEV, and FEF^,,,
observations may reflect inflammatory effects of more
prolonged O3 exposures. It should be noted that both
FVC and FEV, had significant negative slopes on 1-h
maximum O3 measured in the previous 24 h. Adjustment
for temperature and RH diminished somewhat the
associations between lung function and O3. Associations
between lung function decrements and O3 exposure often
appeared larger in children with chronic respiratory
symptoms than in those without, and in children of
mothers who were current smokers; however, these
results were not statistically confirmed.
Castillejos et al. (1992)
-------
Table 7-18 (cont'd). Acute Effects of Photochemical Oxidant Pollution:
Daily Life Studies of Lung Function and Respiratory Symptoms'
Pollutants/Environmental Variables
Study Description
Results and Comments
Reference
Daily maximum 1-h O3 concentrations on days
prior to lung-function testing ranged from 7 to
206 Dg/m3 (3.5 to 103 ppb). Levels of other
pollutants measured (SO2, NO2, PM]0, and
aerosol H+) were reported to be low during
study. Ambient temperature (range: 5 to 31
DC) and some pollen data also were collected.
Associations between morning lung function and previous
day O3 examined during school in 533 children (ages 7-11)
from seven schools in three towns in The Netherlands.
Towns were selected without local pollution sources and
with low levels of pollutants other than O3. Study spanned
the period from March through July, with lung-function
measurements collected every 2-3 weeks. An overall time
trend was fit to the lung- function data to account for lung
growth. Data on FVC, FEV,, PEFR, and PEP^s*
analyzed in relation to previous-day 1-h maximum O3
concentrations using subject-specific linear regressions
followed by analysis of mean slopes. Intersubject
variations in responsiveness to O3 were tested via an F-
test. The influence of chronic respiratory symptoms and
other subject characteristics (e.g., age, sex) on O3
responsiveness was examined. Models that included other
pollutants were also considered.
Negative, usually statistically significant mean slopes
seen for lung function regressed on previous-day 1-h
maximum O3 for the seven individual schools. Over all
533 subjects, mean regression slopes for FVC and FEV,
were D0.20 ± 0.05 and D0.21 ± 0.04 mL/Dg/m3,
respectively; and for PEFR and FEF^j,, were D0.72 +
0.22 and D0.45 ± 0.12 mL/s/Dg/m3, respectively. These
coefficients may be doubled to convert to slopes in terms
of parts per billion. The authors report that adding SO,,
NO2, or PM]0 did not materially change the O3 slopes.
There was evidence for inter-subject variation in O3
responsiveness, but this variation was not statistically
related to available subject characteristics data.
Temperature data not included in models, perhaps due to
high correlation with O3. The lung function/O3
relationships noted above are qualitatively similar to
those reported in the 1988 Fairview Lake camp study and
the Mexico City school children's study.
Hoek and Brunekreef
(1992)
Hoek et al. (1993b)
^ 1-h maximum O3 concentrations on PEFR
xj measurement days ranged from 20 to 103 ppb.
No other pollutants assessed, but ambient
temperature data included.
Relationship between daily peak flow measurements and
ambient O3 concentrations in a population sample of
732 subjects (both adults and children) over 2-week
periods during normal daily activities in Tucson, AZ.
Peak flow assessed using hand-held peak-flow meters up to
four times per day. PEFR measurements on initial 2 days
for each subject dropped to avoid possible learning effects,
leaving a series of up to 12 measurement days per subject.
Population-pooled regression slopes computed for PEFR
on 1- and 8-h average O3 for children (ages D 15) and
adults (ages > 15), controlling for residual auto-
correlations. Outcome measures included PEFR diurnal
variability and afternoon PEFR levels. Besides O3,
potential explanatory variables included temperature,
average time outdoors, acute respiratory infections,
asthma, and environmental tobacco smoke exposure.
Significant positive associations observed between
O3 concentrations and PEFR diurnal variability; the
effect magnitude was greatest in asthmatic subjects.
In children only, noon PEFR was suppressed on days
with higher O3 levels. The uncertain relationship
between central site O3 levels and personal exposures in
this southwestern community was not addressed.
Although the statistical models employed were
appropriate and well chosen, it appears that a substantial
amount of exploratory data analysis was performed prior
to selection of results to present in the paper, leading to
uncertainties regarding the statistical validity of the
hypothesis tests presented.
Krzyzanowski et al.
(1989)
-------
Table 7-18 (cont'd). Acute Effects of Photochemical Oxidant Pollution:
Daily Life Studies of Lung Function and Respiratory Symptoms'
Pollutants/Environmental Variables
Study Description
Results and Comments
Reference
Means and range of max daily 1-h values:
O3 mean = 32.4 Dg/m3, range = 0-129 Dg/m3;
SO2 mean = 51.2 Dg/m3, range = 18-
176 Dg/m3; NO2 mean = 40.5 Dg/m3, range =
12-79 Dg/m3; CoH mean = 0.38 CoH units,
range = 0.1-1.3 CoH units; temperature mean
= 1.3 DC, range = D22D to +22 DC.
Follow-up study (September 1980 through April 1981) of
pollutant-respiratory symptom relationships in subsets of
children from 1979 Chestnut Ridge cross-sectional study of
more than 4,000 elementary school children. Subsamples
selected from six schools in study area with consistently
higher levels of air pollution during previous 4 years.
Subsamples (three) stratified by reported symptoms. One
or more of following measures taken for 144 children:
diaries, symptom questionnaire, spirometry. Telephone
follow-up each 2 weeks on diaries, spirometry done at
school, pollutants (including O3) measured at one monitor
(data from 17 monitors for SO2 generally reflected in data at
single monitor). Diary panel study covered 8 mo;
successive PEFR spirometry studies of 9 weeks each done
in respective groups of the three subsamples.
Relationships of maximum hourly SO,, NO,, O3, and CoH
and minimum temperature for each 24-h period to daily
upper and lower respiratory illness, wheeze, and PEFR
were evaluated using multiple regression models adjusted
for illness occurrence or levels of PEFR on preceding
day. No air pollutant was strongly associated with
respiratory illness or with PEFR. Authors concluded that
this study can best be interpreted as showing no acute
effects of studied pollutants on respiratory symptoms or
PEFR in children at levels lower than the current
NAAQS, but also noted that conclusion must be tempered
by relatively low levels of pollutants encountered and
possibility of exposure misclassification.
Vedal et al. (1987)
'See Appendix A for abbreviations and acronyms.
''Cited in U.S. Environmental Protection Agency (1992).
NJ
CO
-------
reported results that indicated associations between ambient oxidant exposures and various
measures of respiratory effects (e.g., irritative respiratory symptoms and acute pulmonary
function decrements) in children and adults. The aggregation of individual studies provides
reasonably good evidence for an association between ambient photochemical oxidants and
acute respiratory effects and a database that is generally coherent, consistent, and biologically
plausible. In addition, other studies of irritative symptoms in children and adults also were
reported in the 1986 document. For example, Hammer et al. (1974) reported qualitative
associations between ambient oxidant levels and symptoms such as eye and throat irritation,
chest discomfort, cough, and headache at total oxidant levels greater than 0.15 ppm in young
adults (nursing students). Wayne et al. (1967) reported a high correlation (R' = 0.89) between
ambient total oxidant levels (1 h prior to competition) and impaired exercise performance
(running time) in high school students during cross-country track meets in Los Angeles, CA.
Symptoms were not measured, but Wayne speculated that chest discomfort from oxidant
inhalation impaired exercise performance. Although results such as these are consistent with
evidence from controlled human exposure studies, precise characterization of ambient
pollutants and environmental conditions and rigorous statistical analyses were lacking in the
studies. Thus, the primarily qualitative data from these and other studies were not satisfactory
to provide quantitative conclusions about the relationship of ambient O. concentrations and
acute respiratory illness.
Schwartz (1992), Schwartz and Zeger (1990), and Schwartz et al. (1988) reanalyzed
the original diary data of student nurses reported earlier by Hammer et al. (1974) (Table 7-19).
The nurses were told that the diaries were part of a prospective study of viral infections.
Logistic regression models including time series analyses were used to control for
autocorrelation effects that are frequently present in time series data. The reanalysis for daily
prevalence rates of symptoms (Schwartz et al., 1988) confirmed that ambient oxidants were
significantly associated with cough and eye discomfort. However, earlier reported associations
between oxidants and headache or chest discomfort were not confirmed. Cough was the one
symptom that showed an apparent threshold near 0.20 ppm total oxidants, which approximates
the threshold value reported by Hammer et al. (1974). Further reanalysis of the diary data
(Hammer et al., 1974) by Schwartz (1992) and by Schwartz and Zeger (1990) for the effects of
air pollutants on the risk of new episodes of respiratory and other symptoms and on their
durations revealed interesting findings. The mean plus or minus standard deviation (SD) level
of oxidants was 0.102 + 0.074 ppm. In logistic regression models, an increase in oxidant
concentration by one SD (0.074 ppm) was associated with a 17% increased risk of chest
discomfort and a 20% increased risk of eye irritation. These associations were highly
significant (p < 0.001). In addition, photochemical oxidants were significantly (p < 0.0001)
associated with the duration of episodes of cough, phlegm, and sore throat.
Krupnick et al. (1990) and Ostro et al. (1993) reanalyzed daily health data from
over 5,000 children and adults living in the Los Angeles area during a 6-mo period (September
1978 to March 1979) (Table 7-19). The original study was reported by Flesh et al. (1982).
The presence or absence of daily respiratory symptoms associated with daily exposure to
ambient O. and other air pollutants was analyzed in a pooled, cross-sectional, time-series
model. Krupnick et al. (1990) reported statistically significant effects of O. levels on daily
reported respiratory symptoms in healthy nonsmoking adults, but not among smokers,
children, and patients with chronic respiratory disease. Ostro et al. (1993) evaluated the daily
reports of 321 nonsmoking adults and, using a logistic regression model,
7-129
-------
Table 7-19. Acute Effects of Photochemical Oxidant Pollution:
Symptom Prevalence3
Pollutants and Environmental Variables
Study Description
Results and Comments
Reference
Total oxidant, CO, SO2, NO, and NO2 measured;
total oxidant concentrations reached episodic
levels (maximum 1-h/day <0.4 to 0.5 ppm);
mean daily temperature was 71.8 DC.
Reanalysis of daily diary study of student nurses working
and living at schools in Los Angeles (see U.S.
Environmental Protection Agency, 1986, for details of
Hammer et al., 1974). This series of papers reexamines the
nurses' data using logistic regression models and time- series
methods to account for serial correlation (autocorrelation) of
symptoms on successive days. The effects of total oxidants
on daily prevalence rates of symptoms, risks of developing
new symptoms (episodes), and duration of episodes were
analyzed.
Associations found between total oxidants and prevalence of
cough and eye irritation, confirming part of findings of original
study. Association with cough only at oxidant concentrations
above approximately 0.20 ppm. Previously reported
associations between oxidants and chest discomfort and
headache (Hammer et al., 1974) not confirmed. Oxidants
associated with increased risk (incidence) of chest discomfort
and eye irritation and duration of episodes of cough, phlegm,
and sore throat. Duration of symptoms showed concentration-
response relationships even below 0.12 ppm. Findings suggest
different effects of oxidants on symptom characteristics. Lack
of daily paniculate measurements, small number of subjects,
and heterogeneous individual responses restrict quantitative
interpretation of results. Lung function was not measured.
Schwartz (1992)
Schwartz and Zeger
(1990)
Schwartz et al. (1988)
u>
o
Daily O3, NO2, SO2, and CoH and every sixth-
day sulfates measured at one site (Azusa).
1-h daily maximum O3 0.1 ppm, 7-h average O3
0.07 ppm; sulfates 8.43 Qg/m3; maximum
temperature was 22.4 DC.
Reanalysis of daily diaries completed during 181-day survey
period (September 1978 to March 1979) by 756 children and
572 adults (Krupnick et al., 1990) and 321 nonsmoking
adults (Ostro et al., 1993) living in Glendora, Covina, or
Azusa, CA (see Flesh et al., 1982 for details). Presence or
absence of 19 (upper and lower) respiratory and two
nonrespiratory symptoms recorded daily. Presence or
absence of symptoms analyzed in a pooled cross-sectional
time-series model. Nonpollution factors, including sex, gas
stove use, day of study, and a chronic disease indicator were
included in final regression models used to measure effects
of ambient air pollution. Logistic regression analyses for
entire sample to determine effect of each pollutant on health
endpoints. Lagged effects of each pollutant and effects in
individuals (n = 74) without air conditioners and those with
preexisting respiratory infection were analyzed.
Logistic regression model indicated significant associations
between incidence of lower respiratory symptoms and healthy
nonsmoking adults (but not among smokers, children, or
patients with chronic respiratory disease); 1-h daily maximum
O3 levels (OR = 1.22, 95% CI of 1.11-1.34, foraO.l ppm
change), 7-h average O3 level (OR = 1.32, 95% CI of
1.14-1.52), and ambient sulfates (OR = 1.30, 95% CI of
1.09-1.54, for a 10 Dg/m3 change). CoH was significantly
related to daily symptoms in children. Gas stove in the home
was associated with lower respiratory tract symptoms
(OR = 1.23, 95% CI of 1.03-1.47), as were the effects of O3
in subgroups without residential air conditioner
(OR = 1.24) and with preexisting respiratory infection (OR =
1.24). All the above increased risks were statistically
significant (p < 0.05). Interpretation of results limited by
selection of sample; undersampling of young adults;
aggregation of symptoms of all severity levels into one
measure; possible reporting bias; and absence of indoor
exposure, aeroallergen, and lung function data.
Krupnick et al.
(1990)
Ostro et al. (1993)
'See Appendix A for abbreviations and acronyms.
-------
found a statistically significant association between the incidence of lower respiratory tract
symptoms and 1-h daily maximum and 7-h average O. levels (22 and 32% increased risk,
respectively, with 0.1-ppm increase in O.) and ambient sulfates (30% increased risk with a 10-
Dg/m' change). The lower respiratory tract effects of O. were greater in the subgroups with
gas stoves, without residential air conditioners, and with preexisting respiratory infection.
Interpretation of the results is limited by the selection of the sample for analysis;
under sampling of young adults; aggregation of symptoms of all severity levels into one
measure; possible reporting bias; and absence of indoor exposure, outdoor aeroallergen, and
lung function data.
The results from the above panel studies suggest a modest but biologically plausible
relationship between short-term exposure to ambient oxidants/O. and respiratory symptoms.
The interpretation of these recent reanalyses is limited by several factors. Heterogeneous
individual responses occur, and analyses of grouped data possibly may miss susceptible
subgroups. The lack of specific measurements of O. and other pollutants (especially particles)
and of personal exposure or risk variables (e.g., time-activity data) weaken the assessment of
confounders and effect modifiers. In addition, the overall data analysis pertains to small and
very selected samples that have uncertain representativeness to the general population.
Aggravation of Existing Respiratory Diseases
Prior epidemiological data on the effects of ambient O. levels in subjects with
existing respiratory disease have been difficult to interpret due to methodological limitations
(U.S. Environmental Protection Agency, 1986). Exacerbation of asthma and other health
endpoints subsequently has been evaluated, and more recent studies have observed possible
increases in symptom aggravation or changes in lung function of asthmatic subjects in relation
to increased O. or total oxidant levels, as well as interactions between O. concentrations and
temperature. However, no consistent pattern of findings for aggravation of symptoms or lung
function changes has been reported for patients with other types of chronic lung disease. Some
of the major issues in interpreting results from studies of respiratory exacerbations have been
inadequate sample size and characterization of the study subjects, lack of information on the
possible effects of medications, the absence of records for all days on which symptoms could
have occurred, inadequate interpretation of the clinical significance of measured changes, the
role of confounders and effect modifiers (e.g., temperature, humidity, particles,
aeroallergens), and personal or group characterization of indoor-outdoor exposures. For
example, Whittemore and Korn (1980) and Holguin et al. (1985) found small increases in the
probability of asthma attacks associated with previous attacks, decreased temperature, and
incremental increases in oxidant and O. concentrations. Lebowitz et al. (1982, 1983, 1985)
and Lebowitz (1984) reported effects in asthmatics, such as decreased PEER and increased
respiratory symptoms, that were related to the interaction of O. and temperature. None of
these studies adequately assessed possible effect modification by other pollutants, particularly
inhalable particles, which may have independent effects.
Epidemiological studies published since the 1986 criteria document (U.S.
Environmental Protection Agency, 1986) have attempted to control for many methodological
issues (e.g., with [1] better estimates of exposure to pollutants [as well as O.] and
environmental variables that can confound or modify responses, [2] serial measurements of
pulmonary function for determining correlations with pollutants and other environmental
variables, [3] better biomedical characterization of cohorts, and [4] more robust analytical
7-131
-------
approaches that control for autocorrelation of environmental variables and health responses).
Recent studies generally have provided further evidence that supports a relationship between
ambient O./oxidant concentrations and respiratory morbidity in asthmatic subjects
(Table 7-20).
Gong (1987) studied the relationship between air quality and respiratory status of
83 asthmatic subjects living in a high-oxidant area of Los Angeles County. The study covered
February to December 1983, but data analyses were limited to a 230-day period (April 15
through November 30) because of staggered entry of subjects into the study and the high
frequency of missing or incomplete data encountered in the earlier part of the study period.
Regression and correlation analyses between O. and average symptom scores, asthma
medication index (AMI), and day and night PEFR across subjects showed weak, nonsignificant
relationships. These daily outcome variables were compared across days with maximum 1-h-
average O. in three ranges: <0.12 ppm, 0.12 to 0.19 ppm, and >0.20 ppm; "no statistical or
clinical significance was detected." Individual exposures and activity patterns were not
estimated in these two analyses. Multiple regression analyses also indicated the lack of a
significant overall relationship between O. (and their independent variables) and respiratory
status, despite the use of lagged variables and the inclusion of other pollutants, meteorological
variables, aeroallergens, and AMI. Total suspended particles directly affected PEFR, but the
relationship was not consistent in the analysis. Aeroallergens showed significantly negative
relationships to respiratory variables, but only the effect of certain molds was considered
clinically relevant. Temperature and humidity showed no significant effect on the respiratory
variables on this study.
Although there was no significant overall effect of O. on respiratory variables in the
83 asthmatic subjects, multiple regression analysis of subjects whose O. coefficients on various
days were in the top quartile for dependent variables (respiratory measures) showed significant
and consistent effects of O. on Day t and the previous day (Day t D 1). Multiple regression
testing of subsets for associations of symptom score or day or night PEFR on the same-day
O. and previous-day values of the same responses showed highly significant O. coefficients for
all three respiratory measures.
The clinical significance of responses in symptom scores and day and night peak
flow was evaluated for all subjects by individual regression analyses. No subject had evidence
of significant worsening of symptoms attributable to O. during the study. Adult subjects with
high scores in fatigue, hyperventilation, dyspnea, congestion, and rapid breathing in the
Asthma Symptom Checklist had more negative slope coefficients for O. than subjects with low-
to-moderate scores on the checklist. "Responders" (statistically identified by multiple
regression analysis) scored consistently higher in the factors representing fatigue,
hyperventilation, and rapid breathing. The higher scores of these responders, however, "were
apparently not associated with differences in ambient O. concentrations since the test scores
were similar during relatively low (first test) and high (second test) O. days. The significance
of the psychological results is unclear at this time" (Gong, 1987).
Lebowitz et al. (1987) performed a time series analysis to evaluate daily respiratory
responses to outdoor and indoor air pollutant and aeroallergen exposures in potentially
sensitive adults living in a dry climate (Tucson, AZ). Daily symptoms and PEFR were
recorded in well-characterized groups of asthmatics, allergic subjects, patients with chronic
airways obstruction, and asymptomatic healthy controls (total sample size of 204) over 2 years.
Daily diaries included acute symptoms, medication use, and doctors' visits.
7-132
-------
Table 7-20. Aggravation of Existing Respiratory Diseases by
Photochemical Oxidant Pollution*
Pollutants and Environmental Variables
Study Description
Results and Comments
Reference
u>
u>
Air pollutant measurements for April to November
1983 used in statistical analyses. Daily maximum of
NOX, SO2, CO, THC less
than California standards or NAAQS. SO4" >
25 Dg/m3 on 4 days; TSP > 100 Dg/m3 on 78% of
days with data. Daily maximum 1-h average O3
concentrations (from continuous monitoring) = 0.01-
0.11 ppm on 103 days, 0.12-0.19 ppm
on 65 days, 0.2-0.34 ppm on 60 days, and
0.35-0.38 ppm on 3 days. Outdoor aeroallergens
sampled with Roto-Rod: spores, pollens, grasses,
molds, miscellaneous debris; all generally low except
for group of common molds (rusts, smuts,
mushroom) present in thousands per square
centimeter on sampler. Mean (±SD) daily
temperature at 1 p.m. during 200 days:
26 ± 11 DC, range 13-41 DC; 128 days with D 24 DC.
Effects of pollutants and other environmental variables on
respiratory symptoms and PEER evaluated in 11-mo
population study of asthmatics living in high-O3 area
(Glendora) of Los Angeles County, CA. Detailed
questionnaires given at outset on medical/occupational
histories and personal factors, including general activity
Eight of 91 subjects completing study (of 109 recruited) showed no Gong (1987)
variability in asthma status during the 230-day study. Respiratory Gong et al.
status of final study population (1985)
(n = 83 with generally mild or stable asthma), as a whole, not related,
either clinically or statistically, to maximum 1-h average O3 on Days t,
t-1, t-2, or t-3 for any respiratory variable even when adjusting for
patterns; psychological testing (Asthma Symptom Checklist, medication use, symptoms, and PEER on Day t-1. Subset analyses
State-Trait Anxiety Inventory, etc.) also given, once during showed association of O3 with symptoms and with day and night PEER
good air period and once during smoggy period. Lung in subjects in top quartile for respiratory measures, but association did
function (spirometry) and bronchodilator responses measurednot follow a consistent relationship with ambient O3 concentrations.
at outset in all subjects. Daily diaries (checked 2x/week), VE levels during outdoor time not estimated. Outcomes not related to
mini-Wright peak flow meters, and Nebulizer Chronolog time outdoors vs. indoors or to outdoor time on "clean" vs. "smoggy"
days. Subsets ("responders") differed from rest of cohort mainly in
scores of Asthma Symptom Checklist for factors representing fatigue,
hyperventilation, and rapid breathing, but there was no difference in
responders between clean and smoggy periods. Aeroallergens from
maple, oak, beech, and elm trees showed significant (and clinically
relevant) relationships to respiratory variables. Exposure assessment
limited by outdoors-only monitoring and lack of time-activity data.
Hourly outdoor O3, CO, NO2, and TSP measured Effects of outdoor and indoor air pollutants and aeroallergensAsthmatics had greatest number of respiratory complaints, which were Lebowitz et al.
attached to metered-dose inhaler used to record symptoms,
day and night PEER, and medication use, respectively.
Multiple regression analyses for overall group; then subsets
(two groups of "responders") analyzed separately and
compared with rest of cohort.
from three stations. Hourly maxima used for O3 and evaluated in a 2-year study of 22 subjects with asthma, 33 related to the presence of gas stoves, active smoking, humidity, and
CO; daily CO and NO2 derived as weighted measures with airway obstructive disease, 30 atopies, and 14 normals temperature. O3 was associated with peak flow and temperature (late
for each cluster sampling site, and daily values were living in an arid environment (Tucson, AZ). Subjects part ofspring), wheeze (Day t-3 with humidity), and productive cough (Day t-
used. Sample of homes monitored inside and outside a community population sample of 117 families (see U.S. 2). O3 (Day t-3) was related to productive cough during the summer in
for particulates and gases and evaluated for housing Environmental Protection Agency, 1986, for details of allergic subjects. Outdoor gases and meteorological variables
(1987)
characteristics (e.g., gas stove usage).
Meteorological variables measured daily.
Temperature data not reported.
Lebowitz et al., 1982, 1983, 1985; Lebowitz, 1984) and had significantly related to symptoms and PEER, both independently and as
well-characterized symptoms, medication use, lung function, effect modifiers. No significant O3 effect in patients with obstructive
methacholine response (in a subsample), and immunological disease or in normals. Small number of subjects and study days and
status. Daily diaries (acute symptoms, medication use, and lack of indoor NO, and PM10 measurements, measured pollutant
doctors' visits) and daily PEER (2x/day) performed for values, and effect estimates limit quantitative interpretation of study.
3 mo, 2-4x/study period. Duration of time spent
outdoors recorded. Spectral time series analyses used to
evaluate each respiratory response variable for periodic
tendencies and covariance (dependent and independent)
functions as processes in time in the different groups.
-------
Table 7-20 (cont'd). Aggravation of Existing Respiratory
Diseases by Photochemical Oxidant Pollution1
Pollutants and Environmental Variables Study Description Results and Comments Reference
Outdoor O3 levels measured hourly by three Temporal effect of ambient O3 concentration on PEFR during Analyzed PEFR data limited to at least 12 measurements for Lebowitz et al. (1991)
stations and maximum 1- and 8-h average values 30-mo study period in 287 children (13% physician-diagnosed at least 6 days in 78% of children and 74% of adults. Noon Krzyzanowski et al. (1992)
were used to represent O3 levels for all subjects on asthmatics) and 523 nonsmoking adults (9% asthmatics) in the PEFR in nonasthmatic and asthmatic children was lower with
a given day. For each day of the study, the mean Tucson community population sample. Mini-Wright peak higher 1-h maximum O3 levels: Dll.9 L/min/0.1 ppmO3 (p
of the maximum 8-h O3 average for the flow meters used four or fewer times per day but only for 2- < 0.05) and
4 preceding days was calculated to be an index week periods, and only one meter was assigned/household. D31.0 L/min/0.1 ppm O3 (p < 0.1), respectively. Effect of
of cumulative exposure. PM10 was measured daily Children's tests were supervised by adult, and initial 2 days of 8-h O3 mean on evening PEFR seen only in asthmatic
at one station. Mean + SD of maximum 1-h observation were eliminated from analysis. Symptoms from children, possibly reflecting a cumulative O3 response during
O3 concentrations was 0.055 ± 0.014 ppm (range: daily diaries were also used in analysis. Random-effects course of day. Among adults, evening PEFR was decreased
0.015-0.092 ppm), moving average maximum 8-h longitudinal model was used for analyses to account for in asthmatics who spent more time outdoors on days with
O3 levels were 0.046 ±0.013 ppm (0.09-0.082 autocorrelation of PEFR values. Multifactorial ANCOVA washigher O3 concentrations (C x T effect). The ANCOVA
ppm). Maximum daily outdoor temperature was used to analyze day-to-day changes in daily average PEFR andmodel showed significant interactive effects of O3 x
87 DF (30 DC) per person-day, maximum PM10 symptom prevalence rates (the dependent variables) in relation temperature x PM10 on daily average PEFR. Daily rates of
was 187 Qg/m3 (mean 42 Qg/m3). to allergic-irritant symptoms increased with the maximum 8-h
8-h O3 values on the same day and previous days (lags of O3 average (> 0.056 ppm) on the previous day and increased
0 and 1). more with interactions of O3 x temperature x PM]0.
Missing PEFR data, possible overestimation of outdoor
O3 exposure, large variability of responses in asthmatics,
medication use on days with high O3 levels, relatively low O3
i levels, and uncertain effects of indoor and outdoor allergens
r^ and respiratory infections limit interpretation.
-------
Table 7-20 (cont'd). Aggravation of Existing Respiratory
Diseases by Photochemical Oxidant Pollution1
Pollutants and Environmental Variables
Study Description
Results and Comments
Reference
u>
Ln
Hourly O3, twice daily (9:00 a.m. and 9:00 p.m.) acidic Effects of ambient summertime haze air pollution on
aerosols (sulfates, SO4, and H+), and pollen counts wereasthmatic children (ages 7-13) attending
measured on site. Hourly temperature, RH, and O3 1-week asthma camp in Connecticut River Valley
measured from nearby monitors. In 1991, pollution were evaluated during June 1991 (n = 50) and 1992
levels increased daily until Day 5, when maximum 1-h (n = 55). PEER and symptoms (2x/day) and
O3 reached 0.154 ppm and daytime H+ and sulfate levelsnumber of as-needed (p.r.n.) inhaled bronchodilator
were 245 nm/m3 and 26.7 Qg/m3, respectively. In 1992, treatments given by on-site physician during each
air quality was better (e.g., the highest daily 1-h study day were recorded. Correlations between
maximum O3 was 0.063 ppm). Temperature data not health outcomes and air pollutants were performed.
reported.
In 1991, daily total number of p.r.n. treatments highly correlated Thurston et al. (1995)
(r > 0.80) with maximum O3, SO4, daytime H+, and maximum
temperature, but only SO4 (r = 0.97) and H+ (r = 0.985) were
significant (p < 0.05) and remained so after temperature was
included in the analysis. Daily pollen counts were not associated
with treatments p.r.n. Afternoon chest symptoms (cough,
phlegm, and wheeze) and changes in morning-afternoon PEER
values (excluding children given medication) were significantly
correlated (p < 0.05) with O3 and H+, respectively. Scheduled
medications did not apparently provide a protective effect (X2 =
3.25, p = 0.067), although the failure to achieve statistical
significance is not unexpected given the small sample size. In
1992, change in PEER (magnitude not reported), chest symptoms,
and the fewer daily exacerbations (maximum 27 vs. 37 in 1991)
were not significantly correlated with pollution, pollen, or
temperature. Only sore throat, runny nose, and eye irritation
were correlated with pollen counts. Although the data are only in
preliminary form, the 1991 results appear consistent with an effect
of summertime haze air pollution on PEER, chest symptoms, and
asthma exacerbations. The 1992 results are consistent with less
health effects owing to cleaner ambient conditions. Small number
of subjects and study days and lack of results for other pollutants
limit the interpretation of the studies.
'See Appendix A for abbreviations and acronyms.
-------
A sample of homes was evaluated for environmental characteristics and was monitored indoors
and outdoors at the home for gases and particles, in addition to regional stationary outdoor
monitors. Asthmatics showed the most respiratory responses. Outdoor O. levels were
significantly (p < 0.05) related to wheeze, productive cough, and peak flow (late spring) in
the asthmatic group. Statistical interactions between O. and smoking, presence of a gas stove,
maximum temperature, and minimum humidity (R' = 0.49) were found. The other groups did
not demonstrate an O. effect, except for the atopic group, which had increased summertime
productive cough related to O. levels. Thus, these results indicate an O. effect on asthmatics
and that statistical interactions between O. and other environmental factors are significantly
related to symptoms and peak flow. On the other hand, the results are largely descriptive and
qualitative without adequate effect estimators.
A subsequent analysis of the same community population sample in Tucson
(Lebowitz et al., 1991; Krzyzanowski et al., 1992) evaluated the temporal relationship between
PEFR and ambient O. in 287 children and 523 nonsmoking adults. During part of the study
period, ambient particles with a MMAD of 10 Dm or less (PM..) were collected daily at one
monitoring station. A random-effects longitudinal model and multifactorial analysis of
covariance were used for analyses. During the study period, the maximum ambient
O. concentrations were relatively low (i.e., the 1-h maximum never exceeded 0.092 ppm). In
children, noon peak flows were decreased on days when there was a high O. concentration.
Children with physician-confirmed asthma experienced the greatest decrease in noon peak
flow. Evening peak flow also was significantly related to O. in children, especially asthmatic
children, suggesting a cumulative O. response during the course of the day. Among adults,
evening peak flows were decreased in asthmatics who spent more time outdoors on days when
O. levels were high. After adjustment for covariates, significant statistical interactions of 8-h
O. levels with PM.. and temperature on daily PEFR were found. There was a significant
increase in allergic-irritant symptom rates related to prolonged exposure to O. (maximum 8-h
average on the previous day and the interactions of O., temperature, and humidity). The study
had some methodologic problems (e.g., missing daily PEFR data in many subjects, lack of
information about specific hours spent outdoors, medication usage, and relatively low O. levels
during the study period). Nonetheless, the data analyses, control of confounders, and overall
exposure assessment strengthen the conclusions of the study: the respiratory response to O. is
acute, occurs more often in asthmatics, and increases as temperature and PM.. increase.
The respiratory effects of ambient O. and other coexisting pollutants were evaluated
during a 1-week asthma camp in the Connecticut River Valley in June of 1991 and 1992
(Thurston et al., 1995). Each child (age 7 to 13 years) participated in the same daily activities
all week. Peak flow and symptoms were recorded twice a day, as well as the number of as-
needed (p.r.n.) treatments of inhaled bronchodilator administered by an on-site physician
during each day (each representing an exacerbation of asthma). Hourly measurements of
O. and twice daily samples of acidic aerosols (sulfates [SO.] and hydrogen ions [H' ]) were
collected. The results indicate a strong association between the ambient air pollution mix and
the occurrence of asthmatic exacerbations in children. During 1991, pollution levels
progressively increased until Day 5, when the 1-h maximum O. concentration reached 0.154
ppm, and the daytime (9:00 a.m. to 9:00 p.m.) H' and SO. concentrations were 254 nm/m'
and 26.7 Dg/m', respectively. The correlations of the daily total number of p.r.n. treatments
required with daily maximum O., daytime SO. and H' , and maximum temperature were all
high (r > 0.8), but only SO. (r = 0.97) and H' (r = 0.98) were significant (p < 0.05) given
7-136
-------
the small number of days involved. Afternoon symptoms (cough, phlegm, and wheeze) and
morning-afternoon change in PEFR (without medication) were significantly correlated
(p < 0.05) with O. and H' , respectively. During 1992, the local air quality was better (e.g.,
the daily 1-h maximum O. concentration was only 0.063 ppm). There were fewer asthmatic
exacerbations (maximum of 27 versus 37 in 1991), and they were not significantly correlated
with pollution, pollen, or temperature. Pollutants were not significantly correlated with
symptoms or PEFR. Overall, the 1991 data indicate a coherence in the associations of
summertime haze air pollution with peak flow, chest symptoms, and asthma exacerbations in
children. The lack of correlation during 1992 likely was due to the improved air quality and
indirectly supports the results of the previous year. An adequate interpretation of these
preliminary results is limited by the small number of subjects and study days and the lack of
results for other pollutants. These camp studies remain to be reported in more detail.
The above epidemiological studies have generally supported a direct association
between ambient O./oxidant concentrations and acute respiratory morbidity in asthmatics. The
recent studies have strengthened their conclusions by improvements or new approaches in the
estimations of O. exposure, confounders, and effect modifiers; characterization of the subjects
and serial measurements of their responses; and analytical approaches. Thus, the aggregate
results can be viewed as biologically and temporally plausible, consistent, and coherent to
some extent; however, some methodological problems persist. The studies share certain
deficiencies such as small numbers of subjects (which may reduce statistical power) and the
lack of time-activity measurements and significant data about individual responses and their
distribution. The independent effect of ambient O., as estimated by statistical models in
epidemiological studies, is difficult, at best, to clearly differentiate from those of copollutants
because O. (or another pollutant such as H') may be acting only as an indicator of the toxic
potency of the ambient mixture of pollutants. This, in combination with measurement error
and uncontrolled associations with other factors, complicates analytical findings about the
relationships among components of an ambient mixture and may not accurately disentangle the
effects of O. in a biologically appropriate fashion.
7.4.1.3 Aggregate Population Time Series Studies
Aggregate population, or "ecological", time series studies are epidemiological
investigations in which the associations between air pollution and human health outcomes are
evaluated over time in the population as a whole (e.g., with respect to deaths per day in a
given city) and for which outcomes and exposures are not matched for the individuals within
the population. Indeed, aggregate population time series studies of extreme air pollution
episodes have provided some of the clearest evidence of the adverse effects of air pollution on
humans. For example, during the historic December 1952 London Fog episode, in which
extremely high sulfur oxide and PM air pollution levels were experienced, total mortality in
Greater London rose from roughly 300 to 900 deaths/day, and acute respiratory hospital
admissions rose from 175 to 460/day (United Kingdom Ministry of Health, 1954). At more
routine levels of air pollution, any effects of air pollution are necessarily less obvious, and, as
shall be discussed below, methodological issues exist as to the proper analysis and
interpretation of such aggregate population time series data.
The previous criteria document (U.S. Environmental Protection Agency, 1986)
discussed several methodological issues with regard to the epidemiological studies of O. and
photochemical oxidants available at that time. Limitations identified included interferences by
7-137
-------
or interactions with other pollutants and meteorological factors in the ambient environment;
lack of comprehensive exposure issue assessments, such as individual activity patterns and
evaluation of pollutant monitor appropriateness; difficulty in identifying the responsible
oxidant species; and inadequate characterization of the study population. However, most of
these criticisms are not relevant to time series studies. For example, because the same
population is being followed from day to day, the study population acts as its own control,
obviating the need for a detailed population characterization. Also, central site monitoring data
can be useful in these studies for two reasons: (1) although O. concentrations can vary
spatially within an airshed, they usually are highly correlated across sites over time, so that
correlational time series studies are not as dependent on detailed exposure assessments as are,
for example, cross-sectional studies; and (2) if the ultimate use of these studies is to be
included as criteria for ambient standards, the attainment of which is evaluated at central
monitoring stations, then these are the data most relevant for analysis. However, the usually
high correlation of the 1-h daily maximum O. concentration with other averaging times (e.g.,
an 8-h average daily maximum) inhibits the ability of such time series studies to discriminate
the most biologically relevant O. averaging time.
Of the concerns raised by the previous criteria document regarding epidemiological
studies in general, the most relevant to time-series studies is the potential for other serially
correlated environmental factors (e.g., temperature or other pollutants) to confound the unique
identification of O. as a critical causal factor in any environmental health effects identified via
time series analyses of aggregate population data. As discussed by Thurston and Kinney
(1995), either upward or downward bias in the O. effect estimate can result if the model is
misspecified. In the case of underspecification, if another environmental factor that is both
serially correlated with O. over time and also may be causally related with the effect under
consideration (e.g., temperature stress effects on mortality) is excluded from the analysis, then
O. may "pick up" that environmental factor's effect in the model, biasing the O. coefficient
upward. Conversely, the inclusion of variables in the model that are correlated with
O. concentrations but are unlikely to be causally related to the health outcome (e.g., the
inverse of wind speed) results in model overspecification, which may bias the O. coefficient
downward. Only variables that are biologically plausible should be included in a time-series
model, and intercorrelations of the model coefficients should be low if model specification bias
is to be minimized.
One aspect of evaluating time series epidemiologic studies of the health effects of
air pollution that was not raised directly by the previous criteria document but which can be
crucial to proper interpretation is the statistical question of how each study has addressed the
potentially confounding influences of long-wave (e.g., seasonal) variations in the health
outcome data. The seasonality of morbidity and mortality was mentioned explicitly in
Hippocrates' treatise on "Airs, Waters, and Places" and has been studied over the years
(Hechter and Goldsmith, 1961). In respiratory diseases such as asthma, this seasonality of
admissions is very common, due in part to the multifactorial nature of these diseases. For
example, spring and fall increases in pollen and winter influenza epidemics superimpose long-
wave cycles on the day-to-day variations in respiratory hospital admission rates. Such long-
wave cycles need to be addressed as part of any time series analysis for two reasons: (1) they
result in strong autocorrelations that violate the underlying assumptions of most statistical
approaches used to analyze such data; and (2) their inclusion can lead to misleading
conclusions (i.e., confounding), in that the long-wave relationships would likely obscure the
7-138
-------
acute (i.e., short-wave) effects being evaluated. The need to address seasonal cycles in
respiratory disease time series data in order to avoid spurious long-wave dominated
correlations has long been recognized (e.g., Ipsen et al., 1969) but too often has been ignored
or inadequately addressed in the published literature. Autocorrelation, although often
contributed to by seasonal cycles in the data, can be introduced by other causes as well. For
example, Lipfert (1993) noted the need for hospitalization studies to take into account both
weekly and seasonal temporal patterns in the data. Thus, an important criterion for the
evaluation of aggregate population time series studies of the acute morbidity and mortality
effects of O. is whether or not the authors have appropriately addressed all long-wave
periodicities in the data as part of their analysis.
There are a variety of statistical approaches available to address all long-wave
confounding in time series analyses, each having advantages and disadvantages. The primary
goal in invoking such procedures is to eliminate the long-wave autocorrelation "noise" in the
data without inadvertently removing any O.-related health effects "signal" at the same time.
In particular, steps that address autocorrelation in the model but also remove or explain short-
wave variance in the health outcome variable of interest (e.g., by applying prefilters to the
series that affect periodicities down to a few days or by analyzing the residuals from prior
regressions of the outcome variable on "control" variables that are correlated with O., such as
temperature) carry with them the risk of also removing short-wave associations of interest
before the actual analysis has begun. Furthermore, although there are standard regression
diagnostics available to determine whether autocorrelation remains a significant problem (e.g.,
the Durbin-Watson statistic), no such check exists to determine whether the autocorrelation
removal methods also have inadvertently removed an O.-health effects association of interest.
Thus, although steps must be taken in time series analyses to address the potentially large
biases resulting from long-wave (e.g., seasonal) autocorrelations, care must be taken not to
also remove the signal of interest when dealing with the autocorrelation problem.
Emergency Room Visits and Hospital Admissions
Many investigators have evaluated the associations between hospital emergency
room visits or hospital admissions and air pollution. Hospital admissions are far more
common (as counts per day) than, for example, mortality, thereby providing greater statistical
reliability and avoiding the distributional complications that may be presented by low counts.
Also, admission to the hospital is a well defined endpoint, having the desirable feature that
every patient must have been seen by a physician and deemed sick enough to require
hospitalization. Emergency room (ER) visits provide larger daily counts, but are not
necessarily as severe an endpoint. In a well-designed study in Quebec, Canada, hospital
admission diagnosis at discharge was found to be very reliable, with the study confirming the
classification of respiratory admissions in general 92% of the time, and asthma admissions
95% of the time (Delfino et al., 1993). Similarly, a study by Martinez et al. (1993) of
respiratory emergency room admissions in Barcelona, Spain, during 1985 to 1989 concluded
that identification of asthma admissions was highly reliable, as was the discrimination of
asthma and COPD diagnoses. Daily series of hospital admissions thus represent an especially
useful research resource for the investigation of the human health consequences of
O. exposure.
7-139
-------
Hospital admission and ER visit studies that have considered O. associations are
summarized in Table 7-21. In the previous criteria documents (U.S. Environmental Protection
Agency, 1978, 1986), such studies were found to give inconsistent results for
7-140
-------
Table 7-21. Hospital Admissions/Visits in Relation to Photochemical
Oxidant Pollution: Time Series Studies1
Concentration(s)
(ppm)
0.11 to0.28avg
max 1 h during low
and high periods,
respectively
O.llavg
concentration 6 a.m.
- 1 p.m.
(Not reported)
Pollutant Study Description
Oxidant Comparison of admissions to Los Angeles County Hospital for
respiratory and cardiac conditions during smog and smog-free
periods from August to November 1954.
Oxidant Respiratory and cardiovascular admissions to Los Angeles
County Hospital for residents living within 8 mi of downtown
Los Angeles between August and December 1954.
Oxidant Admissions of Blue Cross patients to Los Angeles hospitals
with > 100 beds between March and October 1961; daily
average concentrations of oxidant, O3, CO, SO,, NO,, NO, and
PM by Los Angeles air pollution control districts.
Results and Comments
No consistent relationship between admissions and high-smog periods;
however, statistical analyses were not reported. Clear seasonal trends in
admissions (increasing from summer to winter) not addressed.
Inconclusive results; partial correlation coefficients between total oxidants
and admissions were variable. Method of patient selection was not given.
Other pollutants were not considered. Seasonal trend not addressed.
Correlation coefficients between admissions for allergies, eye
inflammation, and acute upper and lower respiratory infections and all
pollutants were statistically significant; correlations between cardiovascular
and other respiratory diseases were significant for oxidant, O3, and SO2;
significant positive correlations were noted with length of hospital stay for
SO,, NO,, and NOX. Correlations were not significant for temperature and
RH or for pollutants with other disease categories. Reported seasonal
variations in admissions and pollution not addressed.
Reference
California Dept. of
Public Health (1955b,
1956b, 1957b)
Brant and Hill (1964)1'
Brant (1965)
Sterling et al. (1966b,
1967b)
(Not reported) Oxidant Admissions for all adults and children with acute respiratory
illness in four Hamilton, Ontario hospitals during the 12 mo
from July 1, 1970, to June 30, 1971; city-avg pollution
monitoring for Ox (KI), SO,, PM, CoH, CO, NOX, HC,
temperature, wind direction and velocity, RH, and pollen.
Correlation found between admissions and an air pollution index for SO,
and CoH; negative correlation between temperature and admissions; and
nonsignificant negative correlations found with concentrations of Ox (KI).
However, clear, long-wave trends (e.g., seasonally) not addressed.
Levy et al. (1977)c
(Not reported)
Emergency room visits for cardiac and respiratory disease in
two major hospitals in the city of Chicago from April 1977 to
April 1978; 1-h concentrations of O3, SO,, NO,, NO, and CO
from an EPA site close to the hospital, 24-h concentrations of
TSP, SO,, and NO, from the Chicago Air Sampling Network.
No significant association between admissions for any disease groups and Namekata et al. (1979)°
O3, CO, or TSP; SO, and NO accounted for part of the variation of ER
visits for respiratory and cardiovascular admissions. However, the analysis
has a lack of control for confounding, possible unaddressed seasonality in
admissions (time series not shown), and model overspecification (e.g., use
of wind speed).
0.07 and 0.39 avg O3 ER visits and hospital admissions for children with asthma
max 1 h during low symptoms during periods of high and low air pollution in Los
and high periods, Angeles from August 1979 to January 1980; daily maximum
respectively hourly concentrations of O3, SO,, NO, NO,, HC, and CoH;
weekly maximum hourly concentrations of SO4" and TSP;
biweekly allergens and daily meteorological variables from
regional monitoring stations.
Asthma positively correlated with CoH, HC, NO,, and allergens on same Richards et al. (1981)°
day and negatively correlated with O3 and SO,; asthma positively correlated
with NO, on Days 2 and 3 after exposure; correlations were stronger on
Day 2 for most variables; nonsignificant correlation for SO4" and TSP.
Monthly admissions and pollution data indicate strong seasonality, which is
not accounted for. This results in (seasonally driven) positive correlations
with CoH, HC, and NO, and negative correlations with O3.
-------
Table 7-21 (cont'd). Hospital Admissions/Visits in Relation to Photochemical
Oxidant Pollution: Time Series Studies1
Concentration(s)
(ppm)
Pollutant
Study Description
Results and Comments
Reference
0.03 and 0.11 avg
max 1 h for low and
high areas,
respectively
Oxidant Daily hospital ER admissions in four Southern California
communities during 1974 and 1975. Max hourly average
concentrations of oxidant, NO2, NO, CO, SO,, CoH; 24-h
average concentrations of PM and SO4"; and daily
meteorological conditions from monitoring sites 8 km from
the hospitals.
Admissions significantly associated with oxidant and temperature in
all locations. Long-term trends and day-of-week effects appropriately
controlled, but not seasonality. Path-analysis-guided regression used to
discriminate among the intercorrelated pollutant and meteorological factors,
indicating O3 to be most important only at the highest O3 site. However,
lack of catchment area population figures and inadequate seasonality
adjustments prevent quantitative use of results.
Goldsmith et al. (1983)c
0.03 to 0.12
avg of max 1-h/day
for 15 stations
Admissions to 79 acute-care hospitals in Southern Ontario
for the months of January, February, July, and August in
1974 and 1976 to 1983. Hourly average concentrations of
paniculate (CoH), O3, SO2, NO2, and daily temperature
from 15 air sampling stations within the region.
Excess respiratory admissions most strongly associated (p D 0.001) with O3, Bates and Sizto (1983,
sulfate, and temperature during July and August with 24- and 48-h lag. 1987, 1989)
No such associations exist for nonrespiratory (control) diseases. Bates (1985)
Seasonality minimized by selection of narrow study period, and day-of-
week effects controlled. SO4" and O3 highly intercorrelated (r = 0.65),
making effect discrimination difficult. A lack of independent regression
coefficients prevents quantitative application of results.
0.025 to 0.075 3-mo
avg of monthly means
from all city sites
O3 Analysis of quarterly hospital admission rates for
childhood asthma in Hong Kong during 1983-1987
(n = 19). Quarterly means of SO,, NO2, NO, O3,
TSP, and RSP considered.
Concludes that asthma is negatively correlated with SO2, but not with Q
However, analysis uses quarterly means and lacks seasonality controls.
Tseng and Li (1990)
No
0.001 to 0.085 mean
1-h max O3 (avg of
11 sites)
0 to 0.13 avg of max
1 h/day for two
stations
Analysis of emergency room visits, by cause, to acute care Summer (May to October) total emergency (but not respiratory) visits Bates et al. (1990)
hospitals in the Vancouver, BC, area July 1984 to October significantly correlated with temperature and O3. Day-of-week effects
1986. SO2, NO2, O3, SO4, and temperature considered. addressed. Seasonality reduced by study period selection, but opposing
within season cycles in asthma visits and O3 not addressed, which may have
weakened reported O3-respiratory visit relationship. Also, O3 levels much
lower than in previously studied in Southern Ontario.
ER admissions for COPD in Barcelona, Spain, during
1985 to 1986. 24-h avg SO2 and BS city-wide averages.
1-h max SO2, CO, NO2, and O3 obtained from two
stations.
A weak but statistically significant association found between COPD Sunyer et al. (1991)
admissions and levels of SO2, BS, and CO, after accounting for seasonality
and autocorrelation and during season-specific analyses. However, O3 was
eliminated from the analysis based on its seasonally driven negative
correlation with admissions (prior to long-wave controls). Thus, no
conclusions regarding O3 can be made from this work.
0 to 0.04 daily mean
Hospital adrpissions for asthma in Helsinki, Finland, from After accounting for daily minimum temperature, NO, and O3 were
Ponka (1991)
1987 to 1989; 24 h average SO,, NO,, TSP, and O3 city-
wide averages.
significantly correlated on the same day as admissions, whereas O3 was
most significant on the prior day (p = 0.006). However, long-wave peaks
(e.g., in April for asthma) were not addressed and autocorrelation was not
assessed.
-------
Table 7-21 (cont'd). Hospital Admissions/Visits in Relation to Photochemical
Oxidant Pollution: Time Series Studies1
Concentration(s)
(ppm)
Pollutant
Study Description
Results and Comments
Reference
u>
0.06 to 0.13 mean of
1000 to 1500 hours
O3(0.12 1-hmax was
exceeded on 42 of
226 total study days,
whereas 0.08 was
exceeded on 102
days)
0.03 to 0.21 1-h daily
max at central site in
each area
0.01 to 0.16 1-h daily
max at central site
ER visits for asthma, bronchitis, and finger wounds Bivariate correlations indicated asthma visits to be strongly negatively correlated
(a nonrespiratory control) at nine hospitals in central with temperature and weakly negatively correlated with O3, suggesting a seasonality
New Jersey were analyzed for the period May to August influence, despite limitation to the O3 season. However, simultaneous regression of
Cody et al. (1992)
1988 and 1989. Daily values of O3 and SO2 obtained
from nearest of five monitoring sites. Barometric
pressure, temperature, RH, and visibility (as an index
of sulfate) obtained from a Newark measurement
station.
asthma visits on all environmental variables yielded significant (positive) O3 and
(negative) temperature coefficients only, suggesting that temperature acted as a
long-wave control variable, revealing the short-wave O3 relationship with asthma.
Day-of-week effects on visits found unimportant. No environmental associations
seen with bronchitis or control cases (finger cuts).
0.00 to 0.05 avg of O3 Admissions to 79 acute-care hospitals in Southern An elaborate reanalysis of the Bates and Sizto (1989) data set augmented to 1985. Lipfert and
daily means from Ontario for the months of January, February, July, and Long-wave influences controlled using time period subsets and AR modeling. Hammerstrom
22 stations August in 1979 to 1985. Hourly average O3, SO,, NO,, Despite possible overspecification of models (e.g., use of wind speed) and AR (1992)
temperature, RH, wind speed, barometric pressure, and filtering of the short wave, results confirm Bates and Sizto's overall conclusions
daily average TSP and SO4". regarding significant O3 associations. Response to air pollution estimated to be 19
to 24% of summer respiratory admissions, although the exact contribution by Q to
the total was not estimated.
0.01 to 0.05 3-mo O3 Age-specific quarterly asthmatic hospital discharge rates Concludes that asthma morbidity is correlated with particles, but not O3. However, Tseng et al. (1992)
avg of daily means in Hong Kong from 1983 to 1989 examined in relation analysis uses quarterly means and lacks seasonality controls.
from all city sites to quarterly mean levels of TSP, RSP, NO,, NOX, and
O3 (n = 27).
O3 Daily emergency admissions to acute care hospitals for Significant positive associations found for O3, SO4 , and H+ with asthma and total Thurston et al.
asthma, total respiratory, and control disease categories respiratory admissions, but not for control categories. Long-wave and day-of-week (1992)
in the New York City, Albany, and Buffalo, NY, effects removed, and temperature effects controlled. Strongest O3 associations in
metropolitan areas from June to August 1988 and 1989; higher pollution year (1988) and in more urban population centers (Buffalo and New
daily 1-h maximum O3 and temperature and daily York, NY).
average sulfate and acid aerosols (H+) considered.
Daily admissions to 22 acute care hospitals in Toronto,
Ontario, for asthma, total respiratory, and control
disease categories during July and August 1986, 1987,
and 1988; daily 1-h maximum O3, SO,, NO,,
temperature, and daytime (9:00 a.m. to 5:00 p.m.)
SO4" and H+ considered.
Significant positive correlations found for O3, H+, SO4^, PM10, and TSP with
asthma and for total respiratory admissions, but not for SO, or NO,, and not with
control admissions. Long-wave and day-of-week effects removed. Multivariate
regressions and sensitivity analyses suggested that O3 was the pollutant of primary
importance, but H+ may potentiate O3 effects. Except for H+, all PM metrics
considered became nonsignificant when entered into regressions simultaneously with
O3. Ozone significant even after dropping days > 0.12 ppm.
Thurston et al.
(1994)
-------
Table 7-21 (cont'd). Hospital Admissions/Visits in Relation to Photochemical
Oxidant Pollution: Time Series Studies1
Concentration(s)
(ppm)
Pollutant
Study Description
Results and Comments
Reference
0.01 to 0.15 1-h daily
max
Daily emergency respiratory admissions to 168 acute Ozone and SO4 positively and significantly associated with admissions for Burnett et al. (1994)
care hospitals in Ontario, Canada, during May to asthma and COPD in all age groups. Associations consistent across
August 1983 to 1988 were related to daily levels of O3 and regions. Seasonal and day of week effects addressed prior to analysis.
SO4 at the nearest of 22 and 9 monitoring sites, Analyses also controlled for individual hospital influences. No pollutant
respectively. Admissions broken into 0 to 1, 2 to 34, 35 to associations found for nonrespiratory control admissions. Simultaneous
64, and 65+ age groups, and by geographical subregion. regressions suggest O3 to be more important than SO,'
0.01 to 0.11 1-h daily
max averaged over
seven sites in
Montreal
0.02 to 0.16 1-havg
daily max 0.01 to
0. 12 8-havg daily
max
0.01 to 0.04 (10th to
90th percentile) 24-h
daily avg
0.01 to 0.04 24-h
daily avg (10th to
90th percentile)
0.02 to 0.09 1-h avg
daily max (10th to
90th percentile)
O3 Daily urgent hospital admissions to 31 hospitals in
Montreal, Canada, during May-October and August-July
from 1984-1988 related to daily levels of O3, SO4, PM10,
temperature, and humidity. Admissions broken into
asthma, nonasthma respiratory, total respiratory, and a
nonrespiratory "control" group of admissions categories.
O3 Daily numbers of emergency asthma visits by patients 1 to
16 years old to an inner city hospital in Atlanta, GA, from
June to August 1990 were related to daily levels of O3,
SO2, PM10, pollen, and T.
O3 Daily respiratory admissions by patients D 65 years of age
in Birmingham, AL, from 1986 to 1989 were related to
daily levels of O3, PM]0, temperature, and dew point.
COPD and pneumonia admissions examined. Multiple O3
monitoring sites averaged, but the numbers of sites varied
over time.
O3 Daily respiratory admissions for patients D 65 years of age
in Detroit, MI, from 1986 to 1989 were related to daily
levels of O3, PM10, temperature, and dew point.
Ozone and temperature positively and significantly correlated with total
respiratory admissions during the July-August period, but not with control
admissions categories. However, O3 and T are both nonsignificant when
entered simultaneously.
Hospital visits were found to be significantly higher on days when the
previous day's 1-h max O3 exceeded 0.11. No relationship was found
below 0.11, or with 8-h avg daily maximum O3. Day-of-week effects were
accounted for. Seasonality effect reduced by study period selection, but
probable long-wave seasonal cycles superimposed on the day-to-day
fluctuations were not directly addressed, which probably weakened the
reported O3-admissions associations.
COPD and pneumonia admissions positively correlated with O3 and PM]0
over time. A 50 ppb increase in 24 h average O3 was associated with RR
= 1.14 for pneumonia (95% CI = 0.94 to 1.38) and RR = 1.17 for COPD
(95% CI = 0.86 to 1.60). Seasonal fluctuations addressed using 48
monthly dummy variables. Auto-regression methods employed to reduce
autocorrelation. Day-of-week effects not addressed.
Pneumonia and COPD respiratory admissions were found to be
significantly associated with both PM]0 and O3, even after eliminating
noncompliance days. Monthly dummy variables were employed to account
for seasonal variations, but day of week effects were not addressed.
Asthma admissions were not associated with pollution, but this was
attributed to the very low counts in this category for the elderly.
Delfino et al. (1994a)
White et al. (1994)
Schwartz (1994a)
Schwartz (1994b)
-------
Table 7-21 (cont'd). Hospital Admissions/Visits in Relation to Photochemical
Oxidant Pollution: Time Series Studies1
Concentration(s)
(ppm)
Pollutant
Study Description
Results and Comments
Reference
0.01 to 0.04
(10th to 90th
percentile) 24-h daily
avg
Daily respiratory admissions for patients D65 years of age
in Minneapolis-St. Paul, MN, from 1986 to 1989 were
related to daily levels of O3, PM]0, temperature, and dew
point.
Pneumonia respiratory admissions were significantly associated with O3 Schwartz (1994c)
and PM10. No O3-COPD association was found. The pneumonia RR
associated with a 50-ppb increase in 24-h average O3 was RR = 1.22
(95% CI = 1.02 to 1.47). Excluding days with 1-h max O3 above 120
ppb did not alter results. Various methods, including the use of monthly
dummy variables, were used to control for seasonality effects, all
yielding similar results.
0.053 (±0.005)
Mean(+SD) 10a.m.
to 3 p.m. avg
O3 ER visits for asthma at central New Jersey hospitals Asthma visits were significantly associated with O3 and a.m. Weisel et al. (1995)
from May to August 1986 to 1989 related to daily levels of temperature, but not with other environmental variables considered. O3 Weisel (1994)
O3 and temperature. Other environmental variables coefficient implies a 44% mean effect, but unaddressed temperature
considered include RH, sulfates, NO,, SO,, and visibility, effects may be a confounding factor. An analysis limited to July and
August reduces this concern, yielding a 16% mean effect by O3.
"See Appendix A for abbreviations and acronyms.
''Reviewed in U.S. Environmental Protection Agency (1978).
"Reviewed in U.S. Environmental Protection Agency (1986).
Ln
-------
reasons that were not apparent. A common weakness of many of those studies, however, was
a failure to control adequately for seasonal differences in hospital usage and O. concentration.
Therefore, each of the updated critiques in this table now includes an evaluation of how the
data were (or were not) controlled for long-wave influences (e.g., seasonality). With this
factor taken into account, the older studies' varying results are now more understandable. In a
number of these studies, documented long-wave periodicities in the data were ignored,
resulting in nonsignificant associations (i.e., California Department of Public Health, 1955,
1956, 1957; Brant and Hill, 1964; Brant, 1965; Levy et al., 1977; Namekata et al., 1979) or
even significant negative correlations between O. and hospital visits and admissions (Richards
et al., 1981) as a result of the generally higher respiratory admission rates in the colder
months, when O. levels are at their lowest. Two studies that did not control for seasonality
still reported significant positive correlations between hospital admissions and oxidants
(Sterling et al., 1966, 1967; Goldsmith et al., 1983), although Sterling et al. excluded the
winter months from the analysis and Goldsmith et al. did detrend the data. Also, unlike any of
the previously cited studies, both of these analyses controlled for day-of-week effects on
hospital admission rates (e.g., due to consistently lower admissions on weekends), an
important factor in hospital admissions variations that also must be addressed (see Sterling
et al., 1966). Moreover, the one previously reviewed study that adequately controlled for both
long-wave and day-of-week influences (Bates and Sizto, 1983, 1987, 1989) reported very
significant associations (p < 0.001) between O. levels and summertime (July and August)
respiratory hospital admissions. However, other intercorrelated environmental variables (e.g.,
acidic sulfates) also may have been cofactors in this association (Bates and Sizto, 1987).
Overall, a review of these older studies suggests that, if the data are analyzed using newer
statistical techniques, a significant association may be found between elevated ambient
O. concentrations and acute increases in daily respiratory hospital admissions.
Since the last criteria document (U.S. Environmental Protection Agency, 1986),
a number of new ER visit and hospital admissions studies have been completed, a few of which
share some of the same statistical flaws found in many of the older studies. For example,
Tseng and Li (1990) and Tseng et al. (1992) failed to control for the seasonality of admissions
and pollutants in their statistical analyses of quarterly hospital admissions in Hong Kong,
causing them to report no associations with O., but a significant (and likely spurious) negative
correlation of age-specific asthma admissions with quarterly mean SO. in the first of these
papers and a significant (and also likely spurious) positive association with TSP in the second
paper, but no association with O.. Sunyer et al. (1991) failed to consider seasonality in their
initial evaluation of an O. relationship with COPD hospital admissions in Barcelona, Spain;
causing them to eliminate O. from consideration in the study and any evaluation of possible
health effects. Also, Bates et al. (1990), using a largely descriptive approach, characterized
the seasonal periodicities of Vancouver, BC, respiratory ER visits. Their subanalysis of the
warm season (May through October) included a dominant fall asthma peak, which would
obscure any summertime O. associations, and therefore, little can be inferred from this data
analysis about the existence or nonexistence of an acute relationship between O. and
Vancouver hospital visits for respiratory causes. Ponka (1991) showed significant
O. associations with asthma hospital admissions in Helsinki, Finland, over a 3-year period.
The model also included temperature, but did not address directly the noted long-wave
variations in both admissions and pollution. Thus, whether a study has adequately addressed
statistical confounding by the prominent long-wave cycles in respiratory hospital admissions
7-146
-------
series, which are clearly dominated by other causes (e.g., spring pollen, fall respiratory
infection, winter influenza seasons), continues to be a crucial criterion in evaluating the
usefulness of a study's results.
Fortunately, there are also a number of new studies that have addressed both
long-wave and day-of-week influences in their analyses. Cody et al. (1992) did not control
directly for seasonality, but they did narrow their analysis of central New Jersey hospital ER
visits to the high O. season (May through August). Even so, their initial correlational analysis
yielded negative associations between hospital visits and both temperature and O., which
suggests that within-season long-wave effects existed (e.g., generally higher asthma visits in
May, at the end of the pollen season, when O. and temperature are lower on average than in
July or August). However, the authors did conduct subsequent regressions of respiratory visits
on both temperature and O. simultaneously, yielding a significant positive coefficient for
O. and a negative coefficient for temperature, which suggests that the inclusion of temperature
may have indirectly accounted for the long-wave cycle, allowing the positive short wave O.-
visit relationship to be seen. Day-of-week influences were considered, but found to be
unimportant for these ER visit data. No such pollution-hospital visit relationship was found for
finger cut (i.e., control disease) visits.
Weisel et al. (1995) examined central New Jersey hospital ER visits for asthma
(mean = 5.4/day) during the high O. season (May through August) for 1986 through 1990.
Using a stepwise regression analysis, a significant positive coefficient for O. and a negative
coefficient for morning temperature was found. Other environmental factors considered,
including rate of temperature change, RH, every-sixth-day sulfates, NO., SO., and visibility
(an index of fine particles [FPs]), were not found to be correlated with asthma visits. This
study did not directly address long-wave confounding, instead following the same approach as
Cody et al. (1992) in using temperature to indirectly control for such seasonal confounding and
diminishing autocorrelation to nonsignificance (DW = 2). However, it is not clear to what
extent O. may be inadvertently picking up short-wave temperature effects not modeled by this
specification. These are likely to be opposite to the seasonal effect apparently being captured
by the temperature variable (as indicated by its negative coefficient). The highest
O. coefficient was found on the lowest O. year, which is consistent with unaddressed
confounding. Thus, the O. effects reported (which imply an overall O. mean effect equal to
44% of all asthma visits) should be viewed as maximum effects estimates, possibly contributed
to by colinear high temperature influences. Indeed, limiting the analysis to July and August of
each year (thereby reducing long-wave confounding) resulted in a less negative temperature
coefficient and an overall O. coefficient one-third of that for May through August (Weisel,
1994), implying approximately a 16% mean effect, which is more consistent with published
hospital admissions study results. A covariance analysis presented indicates an average 28%
increase in the number of hospital ER visits for asthma on high-O. days (above 0.06 ppm)
versus low-O. days (below 0.06 ppm) after controlling for temperature, but seasonal cycles
were again not directly accounted for in the analysis. Overall, these results are consistent with
an O. effect on asthma morbidity.
Thurston et al. (1992) analyzed unscheduled (emergency) admissions to acute care
hospitals in three New York State metropolitan areas during the summers of 1988 and 1989.
Environmental variables considered included daily 1-h maximum O. and 24-h average SO. and
acid aerosol (H') concentrations, as well as daily maximum temperature recorded at central
sites in each community. Long-wave periodicities in the data were reduced by selecting a June
7-147
-------
through August study period. However, because of remaining within-season long-wave cycles
in the data series (i.e., day-to-day fluctuations superimposed on an annual cycle in admissions),
data were prefiltered using sine and cosine waves with annual periodicities. Day-of-week
effects also were controlled via regression. These adjustments resulted in nonsignificant
autocorrelations in the data series and also improved the pollution correlations with
admissions. For example, in New York City, the same-day O.-asthma correlation rose from a
nonsignificant r = 0.04 in the raw data to a significant r = 0.24 after prefiltering. This shows
the importance of addressing long-wave cycles in such data, even when these data come from a
single season. In contrast, correlations between the pollution data and hospital admissions for
nonrespiratory control diseases were nonsignificant both before and after prefiltering. The
strongest O.-respiratory admissions associations were found during the period of high pollution
in the summer of 1988 and in the most urbanized communities considered (i.e., Buffalo and
New York City). After controlling for temperature effects via simultaneous regression, the
summer haze pollutants (i.e., SO.' , H' , O.) remained significantly related to total respiratory
and asthma admissions. However, these pollutants' high intercorrelation prevented the clear
discrimination of a single pollutant as the causal agent. Depending on the index pollutant, the
admission category, and the city considered, it was found that summer haze pollutants
accounted for approximately 5 to 20% of June through August total respiratory and asthma
admissions, on average, and that these admissions increased approximately 30% above average
on the highest pollution days.
Lipfert and Hammerstrom (1992) reanalyzed the Bates and Sizto (1989) hospital
admissions data set for 79 acute-care hospitals in southern Ontario, incorporating more
elaborate statistical methods and extending the data set through 1985. Long-wave influences
were once again reduced by using the short study periods previously employed by Bates and
Sizto (e.g., July and August only for summer), as well as by employing prewhitening and
autoregressive procedures to the data. Day-of-week effects also were controlled. In addition,
the models were specified much more extensively, to include a variety of new meteorological
variables that may have caused some confounding with the pollutant variables (e.g., wind
speed correlated at r = DO.55 with NO.). Despite possible model overspecification (e.g., the
inclusion of wind speed), summer haze pollutants (i.e., O., SO., SO.) were still found to have
significant effects on hospital admissions in southern Ontario. In contrast, pollution
associations with hospital admissions for accidental causes became nonsignificant in these
models. Although air pollution concentrations were generally within U.S. air quality
standards, the pollutant mean effect accounted for 19 to 24% of all summer respiratory
admissions, although the "responsible" pollutants could not be selected by the authors with
certainty.
Burnett et al. (1994) also employed the Ontario acute care hospital database to
analyze the effects of air pollution on hospital admissions, but their analysis considered all of
Ontario and analyzed the data from each individual hospital, rather than aggregating the counts
by region. Slow moving temporal cycles, including seasonal and yearly effects, were removed
(via an 19-day, moving-average-equivalent, high-pass filter), and day-of-week effects were
controlled prior to the analysis. Poisson regression techniques were employed because of the
low daily admission counts at individual hospitals. Ozone displayed a positive association with
respiratory admissions in 91% of the 168 hospitals, and 5% of summertime (May through
August) respiratory admissions (mean = 107/day) were attributed to O. (mean = 50 ppb).
Positive associations were found in all age groups (0 to 1, 2 to 34, 35 to 64, and 65+). A
7-148
-------
parallel analysis of nonrespiratory admissions showed no such associations, which indicates the
association specificity. Ozone was found to be a stronger predictor of admissions than SO.,
which accounted for an additional 1 % of summertime respiratory admissions. Temperature
had no effect on the pollution-respiratory admission relationship.
Thurston et al. (1994) focused their analysis of respiratory hospital admissions in
the Toronto metropolitan area during the summers (July through August) of 1986 to 1988,
when they directly monitored for strong particulate acidity (H') pollution on a daily basis at
several sites in that city. Long-wave cycles, and their associated autocorrelations, were
removed by first fitting sine and cosine series having annual periodicity (as well as day-of-
week dummy variables) to the data via regression, and analyzing the resulting residuals.
Strong and significant positive associations with asthma and respiratory admissions were found
for both O. and H' , and somewhat weaker significant associations with SO.' , PM..., PM..,
and TSP, as measured at a central site in downtown Toronto. No such associations were found
for SO. or NO., nor for any pollutant with nonrespiratory control admissions. Temperature
was only weakly correlated with respiratory admissions and became nonsignificant when
entered in regressions with air pollution indices. Simultaneous regressions and sensitivity
analyses indicated that O. was the summertime haze constituent of greatest importance to
respiratory and asthma admissions, although elevated H' was suggested as a possible
potentiator of this effect. During multipollutant, simultaneous regressions on admissions,
O. was consistently the most significant. Of the particle metrics, only H' remained
statistically significant when entered into the admissions regressions simultaneously with O..
Sensitivity analyses also showed that dropping all days above the current U.S. O. standard of
0.12 ppm (2 of a total 117 days) did not significantly change the O. coefficients. The
simultaneous O., H' , and temperature model indicated that 21 + 8% of all respiratory
admissions during the three summers were associated with O. air pollution, on average, and
that admissions rose an estimated 37 + 15% above that otherwise expected on the highest
O. day (0.159 ppm). Moreover, despite differing health care systems, the Toronto regression
results for the summer of 1988 were remarkably consistent with previously reported results for
that same summer in Buffalo, NY, (see Table 7-22).
Delfino et al. (1994a) studied daily urgent hospital admissions for respiratory and
other illnesses at 31 hospitals in Montreal, Canada during the warm periods of the year
between 1984 and 1988. Respiratory admissions were considered as a whole and split into
asthma and nonasthma categories, using definitions compatible with those previously used by
Bates and Sitzo (1987) and by Thurston et al. (1994). Both 1-h and 8-h maximum
O. concentrations were considered in the analyses, as well as weather variables (temperature
and relative humidity) and PM measurements, although 83% (five out of every six) PM
measurements were not directly measured but, instead, were estimated from other
environmental variables including visibility, temperature, and O. concentration on those
missing PM days (Delfino et al., 1994b). Seasonal cycles were addressed by applying a
19-day moving average high-pass filter to the health and environmental data before analysis for
associations. Day-of-the-week and autocorrelation effects also were addressed, when present.
For the months of July and August, during the study period, a significant association was
found between all respiratory admissions and both 8-h daily maximum O. (p D 0.01) and 1-h
daily maximum O. (p D 0.03) 4 days prior to admission, despite the fact that no day exceeded
0.12-ppm 1-h daily maximum O. (90th percentile =118 Dg/m' or 0.06 ppm O.). Of the
significant bivariate environmental-admission associations found, the association with 8-h
7-149
-------
maximum O. was the highest reported (r = 0.15), tied only by the 4-day lag in temperature.
However, the addition into the regression of temperature on the
7-150
-------
vj
Table 7-22. Comparison of Regressions of Daily Summertime Respiratory
Admissions on Ozone and Temperature in Toronto, Ontario,
and Buffalo, New York, for the Summer of 1988
City and Year
Toronto (pop. = 2.4 x
1988 summer
Toronto (pop. = 2.4 x
1988 summer
Buffalo (pop. = 2.0 x
1988 summer
Buffalo (pop. = 2.0 x
1988 summer
106)
106)
106)
106)
Respiratory
Admissions Category
Total respiratory
(mean = 14. 1 /day)
Total asthma
(mean = 9.5/day)
Total respiratory
(mean = 25.0/day)
Total asthma
(mean = 7.1 /day)
Temperature, Pollutant
Model Specification
T(LG2), O3 (LGl)b
T(LG2), O3 (LG1)
T(LG2), O3 (LG2)
T(LG2), O3 (LG3)
Pollutant
Regression Coefficient
( Adrnissions/ppb/ 1 06
persons + SE)
0.022 + O.OIO0
0.014 + 0.008C
0.030 + 0.016C
0.012 + 0.004d
Pollutant Mean
Effect (% + SE)
26.4 +11.8
25.3 + 14.9
18.4 + 9.9
23.9 + 10.1
Max/Mean Pollutant
Relative Risk + SE
1.34 + 0.15
1.32 + 0.19
1.25 + 0.09
1.25 + 0.14
"See Appendix A for abbreviations and acronyms.
bLG = lag between exposure and admission, in days.
cp < 0.05 (one-way test).
•'p < 0.01 (one-way test).
Source: Thurston et al. (1994).
-------
same day as the O. reduced both the O. and the temperature associations to nonsignificance. It
should be noted that the authors also found a significant association between asthma admissions
and their estimated PM.. variable during May through October; however, because both
temperature and O. were used to estimate these observations, it is difficult to interpret this
association separately from that for O. and temperature. No significant correlations were
found between O. and nonrespiratory, control admissions (e.g., appendicitis). The authors
conclude that their findings "should be regarded as a reflection of the potential public health
burden of respiratory disease attributable to photochemical air pollutants."
White et al. (1994) reported daily emergency room visit records from June through
August 1990 at a large inner city hospital in Atlanta, GA. Daily counts of visits for asthma or
reactive airway disease by patients 1 to 16 years of age (mean = 6.6/day) were related to daily
levels of O., SO., PM.., pollen, and temperature. Seasonality likely was reduced by the study
period selection, although no effort was made to address possible within-season long-wave
cycles in the data. Day-of-week and temperature effects were controlled as part of a Poisson
model employed to address the small admission numbers at a single hospital. This model
yielded a 1.42 admissions rate ratio (p = 0.057, 95% CI = 0.99 to 2.0) for the number of
asthma visits following days with O. levels equal to or exceeding a 1-h maximum of 0.11 ppm,
which is consistent with the relative risk values reported by Thurston et al. (1992, 1994). No
admissions relationship with O. was seen below 0.11 ppm or with 8-h average O..
In a study of Birmingham, AL, data, Schwartz (1994a) separately examined O. and
PM.. influences on hospital admissions by the elderly for pneumonia (mean = 5.9/day) and
COPD (mean = 2.2/day) causes from 1986 to 1989. Other potentially confounding pollutants
(e.g., SO. and NO.) were not considered, nor was any control admission category analyzed.
Poisson regression analyses were employed controlling for time trends, seasonal fluctuations,
and weather, but day-of-week effects (which can be a large influence on such admissions) were
not addressed. Weather was controlled by including dummy variables for seven (unspecified)
temperature and dew point range categories in the regression. Seasonal fluctuations were
controlled through the use of 48 monthly dummy variables, which raises the concern that
within-month long-wave confounding may have remained. However, autoregressive models
were reportedly used whenever serial correlation was found in model residuals. Base model
results (excluding winter months) yielded a 2-day lag relative risk (RR) estimate of 1.14 for
pneumonia admissions from a 50 ppb increase in 24-h average O. (95% confidence interval,
CI = 0.94 to 1.38). Excluding days exceeding 120 ppb yielded similar results (RR = 1.12,
CI = 0.92 to 1.37). For COPD, the basic model yielded a RR = 1.17 (CI = 0.86 to 1.60),
whereas excluding days above 120 ppb similarly gave RR =1.18 (CI = 0.86 to 1.62).
No models considered O. and PM.. simultaneously. Two other comparative models (the
inclusion of sine/cosine cycles of various periodicities up to 2 years in the regression and the
analysis of deviations from a nonparametric smoothing of admission counts) were tested for
PM.., but not for O., so the model sensitivity of the O. effect was not tested. Overall, even
after excluding days exceeding the standard, this work indicated a fairly consistent O. effect
across respiratory categories that approached, but did not reach, statistical significance.
Schwartz (1994b) analyzed O. and PM.. air pollution relationships with daily
hospital admissions of persons 65 years or older in the Detroit, MI, metropolitan statistical
area from 1986 to 1989. Daily counts for pneumonia (mean = 15.7/day), asthma
(mean = 0.75/day), and all other COPDs (mean = 5.8/day) were regressed on the pollution
variables and various seasonal, trend, and temperature dummy variables, using Poisson
7-152
-------
modeling. However, day-of-week effects were not addressed. Ozone was analyzed with
respect to both its daily 24-h average and 1-h maximum. Autoregressive analyses and
residuals plots indicated no remaining autocorrelation in the model. Both O. and PM.. were
significant in simultaneous pollutant models for pneumonia and COPD but not for asthma
(which was ascribed to the low daily counts for this category). These simultaneous coefficients
were reportedly similar to those from the single pollutant models, although the correlations of
the coefficients were not provided. Dropping all days exceeding the 1-h maximum O. standard
did not change the size of the O. coefficients, which remained significant (p < 0.01). Based
on the regression coefficients and data presented, it can be estimated that the mean effect for
O. (11.6%) was double that for PM.. (5.7%) in the pneumonia model, but comparable for
COPD (12.2% for O. versus 10.2% for PM..). On an absolute scale, these results imply that
O. was associated with 1.7 (+0.2) respiratory admissions by the elderly/day/100 ppb (as a 1-h
maximum) per million persons in the Detroit metropolitan area. This estimate does not
include admissions by persons less than 65 years of age (which likely would have included
higher asthma admissions, for example), so that the total respiratory admissions associated
with O. in the entire population likely would be higher than estimated from this work.
Schwartz (1994c) evaluated the associations of both PM.. and O. with respiratory
hospital admissions by the elderly in Minneapolis-St. Paul, MN, from 1986 to 1989. Due to
small counts, Poisson modeling methods were employed. Various modeling approaches were
employed to address weather influences, including (1) the use of annual, monthly, temperature,
and dew point dummy variables; (2) a stepwise spline approach to fit data dependence on time,
temperature, and dew point (an indicator of the water content of the air); and (3) a generalized
additive model using nonparametric smooth functions of time, temperature, and dew point
temperature. Autoregressive methods were employed to eliminate autocorrelations, when
significant. However, these various complex statistical manipulations were not sufficiently
documented to permit critical review of these methods or replication of results (e.g., dummy
variable ranges were not provided and statistical packages were not referenced). Although no
association was found for COPD in the elderly, O. did make a significant independent
contribution to hospital admissions by the elderly for pneumonia (mean = 6.0/day), even after
controlling for weather and PM... Although all models gave similar results, the best data fit
(as measured by analysis of deviance) and strongest O. association was reported for the
stepwise spline model, which yielded a pneumonia admissions relative risk of 1.22 (95% CI =
1.02 to 1.47) for a 50 ppb increase in the 1-day lag of the 24-h average of O.. The use of 1-h
daily maximum O. in these analyses reportedly yielded less significant associations with
admissions. However, eliminating days with either PM.. above 150 Dg/m' or a 1-h maximum
O. above 120 ppb from the analysis did not alter results significantly.
Table 7-23 intercompares the O. -respiratory hospital admissions effect estimates for
the various studies providing sufficient information to allow the derivation of such pollutant-
specific estimates. The estimates are presented in two ways: (1) as an absolute number of
daily admissions per 100-ppb increase in 1-h O. concentration per million persons, total
population, and (2) as a percent increase in the daily admission rate of the relevant admissions
category, presented as a relative risk per 100-ppb increase in 1-h O. concentration. A
reference increment of 100-ppb O. is employed here because this is
7-153
-------
Table 7-23. Summary of Effect Estimates for Ozone in Recent Studies of Respiratory Hospital Admissions
(j-i
Location
New York City, NYC
Buffalo, NYC
Ontairo, Canada'
Toronto, Canada'
Montreal, Canada*1
Birmingham, ALe
Birmingham, ALe
Detroit, MF
Detroit, MF
Minneapolis, MNe
Minneapolis, MNe
Reference
Thurston et al. (1992)
Thurston et al. (1992)
Burnett et al. (1994)
Thurston et al. (1994)
Delfmo et al. (1994a)
Schwartz (1994a)
Schwartz (1994a)
Schwartz (1994b)
Schwartz (1994b)
Schwartz (1994c)
Schwartz (1994c)
Respiratory Admission Effect Size (+ SE) Relative Risk (95% CI)b
Category [Admissions/ 100 ppb Oj/day/106 persons] [RR of 100 ppb O3, 1-h max]
All
All
All
All
All
Pneumonia in elderly
COPD in elderly
Pneumonia in elderly
COPD in elderly
Pneumonia in elderly
COPD in elderly
0.73 (+ 0.54)
0.83 (+ 0.33)
0.82 (+ 0.26)
0.90 (+ 0.41)
0.41 (+0.19) 1.117(1.03to 1.39)
"See Appendix A for abbreviations and acronyms.
"One-way (D + 1.65SE).
cl-h daily maximum ozone data employed in analysis.
d8-h daily maximum ozone data employed in analysis.
e24-h daily average ozone data employed in analysis. (1 h/24 h avg ratio = 2.5 assumed to compute effects and RR estimates).
'Not reported (nonsignificant).
-------
approximately the difference between the maximum and the mean 1-h daily maximum O. in
these studies (e.g., in Toronto, the 1988 mean = 69 ppb; maximum = 159 ppb). The absolute
effect estimates relative to total population have the advantages that the total effect can be
readily "partitioned" into subcategories (e.g., by age group or disease subcategory), and it also
can be applied easily to other situations (i.e., only the population and O. levels are required),
but this may not be appropriate if the other population makeup is very different from the study
populations (e.g., in age distribution). The relative risk estimates are intuitively interpretable
but are not as readily applied elsewhere (i.e., the respiratory disease prevalence rates must be
known), and the effect will vary depending on the prevalence, which differs widely between
populations and even throughout the year within a single population (as respiratory morbidity
is generally higher in winter than summer). For example, this accounts for much of the
apparent inconsistency between the Burnett et al. (1994) and Thurston et al. (1994) relative
risks, in that the Thurston et al. (1994) Toronto values are for July and August only (when the
prevailing number of respiratory admissions per day are generally at an annual minimum),
whereas the Burnett et al. (1994) Ontario values are relative to respiratory admissions averages
over more months of the year, yielding one-fourth the effect as a relative risk, even though the
absolute effect estimate is two-thirds of the Thurston et al. (1994) estimate. In the case of the
Schwartz studies of the elderly, the assumption has been made, based on data presented by
Schwartz (1994b), that the 1-h daily maximum O. is 2.5 times the 24-h average, and the 100-
ppb 1-h maximum estimates provided for these studies therefore are derived from a 40-ppb
increase in 24-h average O.. The absolute effect size results from these particular studies
suggest that a large portion of the O. effects noted in the previous total respiratory admissions
studies are contributed by COPD and pneumonia cases in the elderly. Based on results
presented by Thurston et al. (1992, 1994), the other major contributor is asthma admissions,
which are usually more prevalent in younger age groups. Overall, the results presented in
Table 7-23 collectively indicate that ambient O. often has a significant effect on hospital
admissions for respiratory causes, ranging in these studies from 1 to 3 total respiratory
admissions/day/loo ppb O./10' persons, or from a 1.1 to 1.36 relative risk/100 ppb O..
Daily Mortality
Past studies of the possible association of O. (oxidants) with human mortality
summarized in prior O. criteria documents (U.S. Environmental Protection Agency, 1978,
1986) were sometimes suggestive of an association, but each study was flawed in some way.
These studies are included in Table 7-24, with annotation as to the document in which they
were reported. Most of these studies considered daily mortality in Los Angeles, CA, during
the 1950s and 1960s. Unlike most historical hospital admissions studies, many of these studies
did recognize and attempt to control for seasonality in the data series. Notable exceptions are
the California Department of Public Health studies (1955, 1956, 1957), which were further
weakened by their qualitative treatment of the air pollution data. The Mills (1957a,b) analyses
also employed a questionable exposure assessment method (the Standard Research Institute
smog index), which diminishes its usefulness. Massey et al. (1961) reported no significant
correlations between community differences in mortality and differences in oxidant levels over
time, but the investigators compared two communities with very different populations (e.g.,
age distributions), a likely confounder in such cross-sectional comparisons. Mills (1960),
while reporting mortality-oxidant associations and effects (370 respiratory and cardiovascular
deaths/year), did not control for potential temperature
7-155
-------
Table 7-24. Daily Mortality Associated with Exposure
to Photochemical Oxidant PoIIutiorf
VJ
(J-l
Concentration(s)
(ppm)
Dl.Opeak
(undefined)
DO. 38 max
1 h/day
(Not reported)
0.10 to 0.42
(undefined)
for 148 days
of 1949
(Not reported)
0.02 to 0.37
average of 1-h
daily max from
all Los Angeles
sites
Pollutant Study Description
Oxidant Relationship between daily concentrations of
photochemical oxidants and daily mortality
among residents of Los Angeles County aged
65 years and older during the periods August
through November 1954 and July through
November 1955.
Oxidant Data extended to include the period from 1956
through the end of 1959.
Oxidant Relationship between daily maximum oxidant
concentrations and daily cardiac and
O3 respiratory mortality in Los Angeles for the
periods 1947 to 1949; August 1953 through
December 1954; and January through
September 1955.
Oxidant Comparison of daily mortality in two
Los Angeles County areas similar in
temperature, but with different levels of daily
maximum and mean oxidant levels (KI); SO2
and CO concentrations were also measured.
Oxidant Daily respiratory and cardiac death counts
for Los Angeles County, 1956 to 1958, related
to daily maximum oxidant concentrations. All
days above 96 DP daily maximum temperature
eliminated from analysis. Each day's average
of daily
oxidant maxima was related to that day's
deviation from monthly mean mortality.
Results and Comments
Heat had a significant effect on mortality; no consistent
association between mortality and high oxidant
concentrations in the absence of high temperature.
However, seasonal trends were not addressed, and
pollution data treatment was qualitative.
Positive relationship between daily maximum oxidant
concentrations and mean daily death rates on high-smog
days vs. low-smog days. Questionable exposure analysis,
including use of the SRI smog index.
No significant correlations between differences in
mortality and differences in pollutant levels. However,
the populations differed in socioeconomic and age
distribution characteristics.
A stratification of the mortality deviations vs. oxidant
concentration revealed increasing mortality with
increasing oxidant concentration, even in the cooler
months. The use of deviations addresses data seasonality.
It is estimated that over 300 deaths/year in Los Angeles
are associated with oxidants. However, the lack of
temperature controls below 96 DP is a major weakness.
Reference
California
Department of
Public Health
(1955b, 1956b,
1957b)
Tucker (1962)
Mills (1957ab,bb)
Massey et al.
(1961)b
Mills (1960)
-------
Table 7-24 (cont'd). Daily Mortality Associated with Exposure
to Photochemical Oxidant PoIIutiorf
Concentration(s)
(ppm)
Pollutant
Study Description
Results and Comments
Reference
0.05 to 0.21 Oxidant Reanalysis of the relationship between KI and
monthly avg daily mortality from cardiac and respiratory
diseases in Los Angeles for the years 1956
through 1958.
Used deviations from sine wave fit to reduce
seasonality of pollution and mortality, but fit of
monthly variations was inadequate. Significant
correlations found between pollutants and mortality for
cardiorespiratory diseases, but autocorrelation
adjustments by authors reportedly reduced these
associations to nonsignificance.
Hechter and
Goldsmith (1961)"
0.003 to 0.128 O3 Relationship between daily mortality and daily
max 1 h/day 1-h maximum concentrations of O3 in Rotterdam,
The Netherlands, during the months of July and
August 1974 and 1975.
Mortality significantly higher during relatively high Biersteker and
pollution (0.05 < O3 < 0.125) and heat episodes in Evendijk (1976)c
1975. However, no significant mortality difference due
to moderate O3 episodes (0.05 < O3 < 0.08) in 1974, in
the absence of high temperature. Such aggregated
analyses of serial data makes interpretation difficult.
Ln
0.02 to 0.29 O3 Total, respiratory, and cardiovascular mortality
six-site mean in Los Angeles County, 1970 to 1979, related
of daily 1-h to O3, CO, SO2, NO2, HC, PM, daily max
max temperature, and RH. Low-pass
filter used to eliminate short-wave associations so
that only seasonal associations could be studied.
Frequency domain analysis indicated a significant Shumway et al.
short-wave O3-mortality association, but this was not (1988)
investigated further. The filtered (i.e., long-wave) data
analysis indicated O3 to be a nonsignificant contributor to
seasonal variations in mortality.
0.02 to 0.29 O3 Shumway et al. (1988) 1970 to 1979 Los Angeles
six-site mean mortality dataset reanalyzed using a high-pass
of daily 1-h filter to allow investigation of short-wave (acute)
max associations with environmental variables, after
removing seasonality effects. Environmental
variables considered included temperature, RH,
extinction coefficient, carbonaceous PM, SO2,
NO2, CO, and O3.
Filtered environmental and mortality data analyses
demonstrated significant associations between
short-term variations in total mortality and pollution,
controlling for temperature. Day-of-week effects found
not to affect the relationships. Of the pollutants
considered, O3 had the strongest association with total
mortality. Similar results found for cardiovascular
deaths, but not for respiratory deaths (for which only
temperature was significant).
Kinney and
Ozkaynak (1991)
-------
Table 7-24 (cont'd). Daily Mortality Associated with Exposure
to Photochemical Oxidant PoIIutiorf
Concentration(s)
(ppm)
Pollutant
Study Description
Results and Comments
Reference
Not reported O3 Total daily deaths in Detroit, MI, 1973 to
1-h daily max 1982, analyzed using Poisson methods.
Environmental variables considered included
TSP, SO2, temperature, dew point, and O3.
Significant associations found between mortality and PM, Schwartz (1991)
but not O3. However, O3 data and results not presented,
so it is difficult to evaluate reported conclusion.
Seasonally controlled via multiple dummy weather and
time variables, and autocorrelation addressed using
autoregressive techniques. Possible overspecification of
weather controls may be a factor in the nonsignificance of
O3 in this analysis.
Ln
CO
0.000 to 0.064
24-h avg
(in TN, no
exceedances of
0.12 ppm 1-h
max; in MO,
five exceedances
with max =
0.15 ppm)
O3 Associations between total daily mortality
and air pollution were investigated in St.
Louis, MO, and Kingston-Harriman, TN,
during September 1985 through August
1986. Environmental variables considered
include temperature, RH, PM,0, PM25, sulfate,
aerosol acidity, SO2, NO2, and O3.
Statistically significant daily mortality associations were
found with PM10, but not with O3. Autocorrelation
removed via season indicators, multiple
temperature/climate variables, and AR modeling. The
nonsignificant O3 coefficient may have been contributed
to by the more conservative autocorrelation removal
measures taken, lower O3 concentrations, and shorter
study period, relative to other recent mortality studies.
Dockery et al.
(1992)
"See Appendix A for acronyms and abbreviations.
bReviewed in U.S. Environmental Protection Agency (1978).
'Reviewed in U.S. Environmental Protection Agency (1986).
-------
influences on mortality below 96 DP daily maximum. Hechter and Goldsmith (1961)
reanalyzed the Mills (1960) data using a simple annual sine wave seasonality correction and
obtained significant oxidant correlations until an autocorrelation adjustment was applied; this
reportedly caused the pollutant-mortality correlations to drop to nonsignificance (results not
presented). Biersteker and Evendijk (1976) conducted a t-test of difference analysis of two
summers of time series data from Rotterdam for 1974 and 1975. Although significant
mortality differences could be seen during 1975 heat-pollution episodes (0.05 < O. <
0.125 ppm), no significant mortality increase could be seen during the cleaner and cooler
summer episodes (0.05 < O. < 0.08 ppm). Statistical time series methods were needed to
address probable confounding by temperature effects. Overall, the various exposure
assessment and statistical analyses weaknesses in the studies reported in previous O. criteria
documents have prevented the drawing of definitive conclusions in those past documents as to
whether or not there is a significant association between O. and human mortality.
Although relatively few O. mortality studies have been conducted and published
since the last criteria document (U.S. Environmental Protection Agency, 1986), the statistical
methods and pollution data employed in these studies have improved, compared with the older
studies discussed above. Shumway et al. (1988) focused on long-wave variations in mortality,
finding that O. was a nonsignificant contributor to seasonal variations in Los Angeles mortality
during 1970 to 1979. As might have been expected, temperature was found to be the principal
environmental factor influencing seasonal mortality fluctuations. This paper's exploratory
frequency domain analysis did indicate a significant short-wave (i.e., cycles on the order of a
few days in period) association between O. and mortality, but this result was not pursued in the
subsequent regression analyses.
Kinney and Ozkaynak (1991) reanalyzed the 1970 to 1979 Los Angeles County
mortality and environmental data set for short-wave pollution-mortality associations using
seasonal and day-of-week controls. After prefiltering the environmental and mortality time
series using a high-pass filter, significant associations were demonstrated between air pollution
and short-wave (acute) variations in total mortality, even after controlling for temperature
influences. Day-of-week effects also were accounted for but were found not to affect
pollutant-mortality associations. In the regression models considered, the 1-day lag of
O. concentration gave the strongest pollutant associations with total mortality. This
O. coefficient was statistically separable from the other significant pollutants in the analysis
(CO, NO., and PM), although these other three pollutants were too intercorrelated to separate
from each other. Expressed as an elasticity, the O. regression coefficient (0.03 + 0.01 [SE]
deaths/ppb) over all years indicated that a 1 % increase in O. concentration was associated with
a 0.015 % increase in total mortality. This result would imply an O. mean effect on the order
of 1.5% of total mortality throughout the year (i.e., 830 total deaths/year). Results for
individual years varied widely in terms of the O. coefficient size and significance, which
indicates the need for multiple years of data to discern an effect of such a small size, relative to
other mortality causes. Ozone regression results for cardiovascular deaths (average = 87/day)
were qualitatively similar to those for total mortality (average = 152/day), but only
temperature was significant for respiratory deaths (average = 8/day), probably due to low
count number effects for this category (i.e., Poisson models may have been required).
Overall, although the Shumway et al. (1988) analysis of these 1970 to 1979 Los Angeles data
indicates that disease factors and other pollutants dominate the overall seasonal cycles in
mortality in Los Angeles, the Kinney and Ozkaynak (1991) short-wave analysis documents that
7-159
-------
O. explained a small but statistically significant portion of day-to-day variations in total
mortality in that city over a 10-year period.
Schwartz (1991) analyzed total daily human mortality in Detroit, MI, during the
10-year period from 1973 to 1982, primarily investigating the effects of PM using Poisson
methods. Although actual results are not presented for O., it is stated in the discussion of
results that O. was "highly insignificant as a predictor of daily mortality." Weather is
controlled for extensively in the model specification before the introduction of the air pollution
variables. The fact that O. is usually correlated over time with meteorology raises the concern
that the model may be over specified, but no diagnostics (e.g., correlations of the coefficients)
are presented to allow for an evaluation. Although previous-day temperature was included in
the model, the only direct seasonality control attempted was to limit the analysis to nonwinter
months. Thus, it is not clear to what extent within-season long-wave confounding also may be
influencing the results. If present, such long-wave confounding would be expected to bias the
O. coefficient downward towards nonsignificance in this case (because O. is usually highest,
and mortality lowest, in summer) and would result in autocorrelation in the model. No model
residual diagnostics are reported (e.g., DW statistics or plots of the model residuals), so the
extent of this problem, if present, cannot be evaluated directly. However, autoregressive
methods were employed, which should have addressed any autocorrelation problems. Overall,
the poor documentation of the mortality-O. modeling, especially regarding the lack of model
specification details or model coefficient intercorrelations, makes the author's statement
regarding O. and mortality difficult to evaluate.
Dockery et al. (1992) conducted an analysis of total daily human mortality in
St. Louis, MO, and Kingston-Harriman, TN, during the 1-year period from September 1985
through August 1986 aimed primarily at assessing the effects of PM on mortality. One of the
strengths of this study is the fact that multiple air pollutants were measured and considered.
Thus, as part of the analysis, O. and other gaseous pollutants also were considered and found
to have nonsignificant associations with mortality in these cities. The statistical analysis
addressed autocorrelation in the mortality data through the use of multiple climate indices (i.e.,
daily mean temperature, hot day, cold day, humid day, hot and humid day, season, and
interactive terms) and through the incorporation of autoregressive modeling. This approach is
possibly more conservative than that employed by Kinney and Ozkaynak (1991), and the lack
of a significant O. coefficient in this analysis may be due in part to the statistical modeling
approach, which may or may not have affected an O. mortality relationship in the data in the
process of addressing autocorrelation and so extensively controlling for temperature (which is
usually correlated with O. over time). Also, the lack of any O. associations with total
mortality may be due in part to the relatively low O. levels found in these particular
communities (especially in Kingston-Harriman, where no O. exceedances occurred) during the
study year (maximum 24-h mean O. < 0.065 ppm). Overall, this study did not show an
association between O. and mortality, but this may in part be a product of the particular
methodological and exposure characteristics of this study vis-f-vis the identification of
O. health effects.
7.4.1.4 Summary and Conclusions
Recent epidemiology studies addressing the acute effects of ambient O. have yielded
significant associations with a wide range of health outcomes, including lung function
decrements, aggravation of preexisting respiratory disease, and increases in daily hospital
7-160
-------
admissions and mortality. Individual-level camp and exercise studies clearly indicate that lung
function can decrease in a concentration-related manner in response to O. exposures occurring
in ambient air. The combined results of these studies provide useful, quantitative information
on the pulmonary effects of ambient O. exposures. Results from daily life studies, although
more difficult to interpret quantitatively due to exposure assessment uncertainties, are
qualitatively consistent with camp and exercise studies. There is limited evidence from several
studies suggesting that ambient O.-induced lung function decrements may persist for up to 24
h. Results from lung function epidemiology studies generally are consistent with those of
human chamber studies. An O.-related worsening of symptoms in selected groups of healthy
individuals and detrimental changes in symptoms, lung function, and medication use in
asthmatics have been observed qualitatively and, to a lesser extent, quantitatively. The
relationship is consistent, temporally plausible, and moderately coherent.
Emergency room visit and hospital admission studies considered in this document
collectively indicate that, when the major confounders to such analyses are addressed (e.g.,
seasonality, day-of-week effects), consistent associations are seen between acute occurrences of
respiratory morbidity and O. exposure. The evidence is especially strong for hospital
admissions, as the association has been seen by numerous researchers at a variety of localities
using a wide range of appropriate statistical approaches. Although the absolute size of the
effect varied somewhat across localities and statistical approaches, these analyses suggest that,
in the summertime (when many other respiratory illness causes have abated), O. air pollution is
associated with a substantial portion (on the order of 10 to 20%) of all respiratory hospital
visits and admissions. Moreover, certain of these analyses also indicate that, on the highest
O. days, this pollutant's estimated contribution can increase to the point where it is associated
with nearly half of all respiratory hospital admissions. Moreover, significant associations also
are seen between O. and hospital visits and admissions at exposures below 0.12 ppm 1-h daily
maximum O..
As was also the case for the O.-hospital admissions time series studies, many of the
older O.-mortality studies had methodological or statistical weaknesses that prevented clear
conclusions. However, since the release of the previous criteria documents, one of the two
most useful new studies (Kinney and Ozkaynak, 1991) indicated statistically significant effects
by O. on short-term (acute) human mortality. The one relevant new study that did not show
any O. association (Dockery et al., 1992) employed much more extensive climate and
autocorrelation control methods and was conducted over a much shorter time period than the
other study. Also, the study that showed an O.-mortality association considered an urban area
experiencing 1-h maximum O. concentrations above 0.15 ppm, whereas the other study areas
(eastern Tennessee and St. Louis, MO) did not. Thus, although the analysis of daily series of
human mortality and air pollution has yielded small but statistically significant associations
with O. in one study, the sensitivity of this association to statistical modeling methods and to
O. concentration level needs further investigation.
7.4.2 Chronic Effects of Ozone Exposure
7.4.2.1 Introduction
At the time of the publication of the previous EPA air quality criteria document
(U.S. Environmental Protection Agency, 1986), little useful data were available on the chronic
effects of O. exposure. Table 11-10 of that document summarized the limited number of
studies available at that time and concluded "...it is unlikely that any of these studies can be
7-161
-------
used to develop quantitative exposure-response relationships for ambient oxidant exposures.
Further study of well-defined populations over long periods of time is required before any
relationship between photochemical oxidants and the progression of chronic diseases can be
conclusively demonstrated from population studies" (U.S. Environmental Protection Agency,
1986). The document noted that existing studies failed to demonstrate any consistent
relationship between chronic oxidant exposure and changes in pulmonary function, chronic
symptoms, chromosomal abnormalities, or chronic disease mortality.
The largest study that had been performed at the time of the 1986 criteria document
was that of Detels et al. at the University of California at Los Angeles (UCLA) (Detels et al.,
1979, 1981; Rokaw et al., 1980). This study employed a population-based sample of
households in selected communities in the Los Angeles South Coast Air Basin. A standardized
interview was administered, and individuals underwent various tests of lung function. Air
pollution data were derived from a network of monitoring stations maintained by the South
Coast Air Quality Management District of the California Air Resources Board (ARE). The
usefulness of the findings of this study was considered to be limited due to a number of factors:
(1) variable timing of testing in the several study communities over a 4-year period, (2) paucity
of data on self-selection (completion rates between 70 to 79%) and migration in and out of the
study communities, (3) inconsistent demonstration of reproducibility of the pulmonary function
measurements, (4) mixed ethnicity of the study population, (5) inadequate data on individual
exposure and failure to adjust exposure estimates for migration in and out of the study areas,
and (6) methods employed for comparisons of health effects.
The 1986 criteria document also summarized the first of the Adventist Health Smog
(AHSMOG) studies (Hodgkin et al., 1984) on the occurrence of COPD in relation to chronic
air pollution exposures. However, the data from this first publication were felt to be of limited
value because only symptom data were reported and the exposure assessment was insufficient.
7.4.2.2 Recent Epidemiological Studies of Effects of Chronic Exposure
By the very nature of the problem of the establishment of a link between chronic
exposure to O. and the occurrence of chronic health effects, epidemiological studies remain the
only approach for obtaining human data. As has been noted in the 1986 document, principal
problems for such studies relate to (1) the specification of individual exposures over the
relevant periods of life of the study subjects; (2) the coincident effects of other oxidant species
(e.g., NO., derivative acid species) and other air pollutants (acid aerosols, particulate species);
(3) seasonal effects that relate to pollutant and meteorologic factors, which affect specific
pulmonary function measurements relevant during the course of longitudinal studies or over
studies that utilize multiple cross-sectional samples; and (4) control for effects of factors such
as occupational exposures, cigarette smoking, etc. In addition, past epidemiologic studies have
not had access to any human histologic specimens in relation to the exposure groups under
study nor have specific mechanisms been investigated to explain any of the symptom or
functional outcomes observed.
Histologic and Immunologic Effects
Sherwin has presented some provocative preliminary, histologic data and uses it to
offer a hypothesis on the importance of pathologic changes in the centriacinar region (CAR) of
the lung in relation to chronic pulmonary effects of oxidant air pollution (Sherwin, 1991;
Sherwin and Richters, 1991). Only the publication that presents the primary data (Sherwin,
7-162
-------
1991) is reviewed here, because there is some redundancy in the two available publications.
Sherwin (1991) obtained lungs from 107 subjects, 15 to 25 years of age, who died of a sudden
death without evidence of overt disease, lived in Los Angeles County, had no autopsy evidence
or history of drug use, and had no lung trauma. Abnormalities of the CAR were evaluated by
a pathologist who was "blinded" to basic demographic data. Centriacinar region disease was
defined as the extension of a respiratory bronchiolitis into the proximal acinar structures (i.e.,
chronic inflammatory cells and histiocytes into alveolar ducts, sacs, and alveoli immediately
adjacent to a respiratory bronchiole). The odds ratio for severe CAR disease in subjects who
lived in metropolitan Los Angeles versus those who lived in other cities in Los Angeles County
was 4.0 (95% confidence limit [CL], 1.4 to 11.3; a calculation based on data in Sherwin
[1991] Tables 2 and 3).
Unfortunately, no exposure data (or lifetime residence data) were available for the
subjects in the Sherwin study, nor were smoking histories, cotinine results, or occupational
histories available. The smoking history data is of critical importance because respiratory
bronchiolitis has been shown to be an early pathologic change found in the pulmonary airways
of young smokers (Niewoehner et al., 1974). Additional problems for this study were the fact
that most subjects were of low socioeconomic status, and only 10 of the subjects were female.
Furthermore, the study is limited by a lack of quantitative morphometry on the lung specimens
and by the lack of a control group from an ambient environment with low oxidant pollution.
Therefore, although Sherwin's data are of considerable interest, particularly in relation to the
primate O. exposure data that show similar effects (see Chapter 6, Section 6.2.4), they
currently are not of value in the determination of appropriate human exposure levels for O.,
nor do they even establish the fact that the oxidant environment found in metropolitan Los
Angeles, indeed, is responsible for the observed pathologic changes.
Zwick et al. (1991) carried out a study of allergic sensitization and cellular immune
responses in children (median age of 11 years) from four schools in two Austrian cities. Two
years of meteorologic data and continuously measured levels of SO., NO., and O. were
available for both cities. Monitors were within 2 km of the study schools, except for one
O. monitor that was 13 km from a school in the "high"-O. area. "Allergic diseases" (rhinitis,
conjunctivitis, and asthma), response to prick test antigens, total IgE concentration, number of
subjects with IgE > 100 kU/L, and total IgG concentration did not differ between the subjects
in the two cities. Adjustment for sex, age, active and passive smoking, and types of cooking
and home heating did not alter the results. Children from the high-O. environment had small,
but statistically significant, decreases in the absolute and relative numbers of OKT4 +
(helper/inducer) T cells and OKNK+ (natural killer) cells and increases in OKT8 +
(suppressor) T cells. Adjustment for active and passive smoking and recent respiratory illness
did not alter the results. The frequency of subjects with a measurable PD.. to histamine also
was increased in the high-O. area. No relationship between the T-cell findings and PD.. or
any of the other immunologic markers are provided.
The Zwick et al. (1991) results are limited by lack of any exposure data and by lack
of detail for the O. and other ambient air pollution data. Except for data on the average
percentage time above specific levels of O., there are no useful data that can be applied to the
observations reported. Moreover, the differences observed in the various T-cell subsets were
relatively small and of questionable biological significance. There are no analyses that relate
the T-cell findings to the clinical and functional data (see Table 7-25) that are reported.
Finally, although the communities were said to be similar on all meteorologic and other
7-163
-------
ambient pollution data, inspection of the author's Table 1 (Zwick et al., 1991) indicates that
the mean (averaging time not given) NO. levels in the low-O. community were fourfold greater
than those in the high-O. community (42 Dg/m' versus 11 Dg/m'). No data on acid species or
particles are provided, although both study cities were free of heavy industry and heavy traffic.
Calderon-Garciduenas et al. (1992) have studied chronic exposure to the ambient air
of southwestern metropolitan Mexico City in relation to histologic abnormalities of the nasal
mucosa. The exposed group consisted of subjects who spent at least 8 h/day while working at
a naval hospital in southwestern metropolitan Mexico City. Ninety-two percent of the group
lived in the same area as the hospital, and all had lived in southwestern metropolitan Mexico
City for >2 mo (n = 47). Controls consisted of (1) subjects who lived in Veracruz and who
had not left this area over a period of at least 5 years before the onset of the study (n = 12)
and (2) new arrivals (< 30-day residence in southwestern metropolitan Mexico City) at the
naval hospital who came from low-O., "non-polluted" ports (n = 17). Nasal biopsies were
obtained for all subjects in May through June, 1990, as were histories on residence, smoking,
occupation, allergies, etc. All three groups were matched for age, sex, and occupation.
There were no differences in familial allergy history or personal smoking (specific data not
given in paper). There was a progressive increase in both nasal symptoms and nasal
histologic abnormalities in relation to presumed O. exposure (Veracruz < new arrivals <
long-term residents of southwestern metropolitan Mexico City). The principal histologic
change was basal cell hyperplasia, with squamous cell metaplasia and mucosal atrophy
occurring less frequently. Only 11% of those with > 60-day residence in southwestern
metropolitan Mexico City showed normal mucosa.
Unfortunately, no ambient air data were presented for SO. or particles, which are
said to be low relative to other parts of the city, or other pollutants that could be present.
In addition, because the monthly average maximal O. concentrations are (and have been since
late 1986) well above the current U.S. 1-h standard of 120 ppb, the Calderon-Garciduenas
et al. (1992) data are of limited value to understanding low ambient O. exposures. (This
conclusion probably applies even if one considers the different concentrations represented by a
given parts-per-billion value at different altitudes.) Subjects in southwestern metropolitan
Mexico City are subjected to O. levels of between 100 and 400 ppb for several hours per day
in the winter and spring. Despite the lack of data on other air pollutants and specific exposure
data for individual subjects, this study does provide useful evidence to suggest upper
respiratory damage as a consequence of prolonged exposure to ambient air mixtures.
Pulmonary Function, Respiratory Symptoms, and Chronic Respiratory Disease
The Adventist Health Smog Study. Since the publication of the 1986 criteria
document (U.S. Environmental Protection Agency, 1986), a number of studies have been
published that attempt to define chronic respiratory system health effects in relationship to
ambient O. concentrations (see Table 7-26). Among these, the series of publications from the
AHSMOG study (Hodgkin et al., 1984; Euler et al., 1987, 1988; Abbey et al., 1991a,b) will
be discussed first and as a set.
7-164
-------
7-165
-------
Table 7-25. Pathologic and Immunologic Changes Associated with Chronic Ozone Exposurfe
Concentrations(s)
Dg/m
Pollutants and
Environmental
Variables
Study Description
Results and Comments
Reference
Not provided
Not provided,
Los Angeles County,
not further specified
Autopsy study of lungs from sudden
death victims 15 to 25 years old whose
residence was Los Angeles County;
examination of lungs for inflammatory
changes in CAR of lungs.
Most severe CAR disease in residents of
metropolitan Los Angeles County versus
other county areas; data limited by lack
of smoking history, personal exposure
and occupational data; interesting
hypothesis, but role of O3 unknown.
Sherwin and
Richters (1991)
0.095 to
0.188, time
metric not
given
186 to 368
O
Sfudy of allergic sensitization and
cellular immune responses in children
(median age, 11 years) in two Austrian
cities, 1989.
Small increases in OKT4 +
(helper/inducer) and OKT8 +
(suppressor) T-cells and small decrease in
natural killer cells in "high" ozone
community; increase in number of
subjects with measurable PD20 histamine
in "high" ozone area; no relationship
between T-cell findings and any clinical
immunologic measure, lung function, or
PD20; meaning of results unclear.
Zwick et al.
(1991)
0.150 to
0.275
monthly
average
Approx.
294 to 539
Study of nasal histology in persons
living in southwestern Mexico City and
Veracruz; subjects matched on age, sex,
occupation; similar allergy and smoking
histories.
Increased occurrence of nasal dysplasia in Calderon-
southwestern Mexico City residents,
especially those with more than 5 years
residence; no data on other air pollutants;
data not directly relatable to U.S.
conditions because Mexico City residents
are exposed to O3 levels between 0.1 and
0.4 ppm for several hours each day, all
year long, with relatively few days below
0.1 ppm.
Garciduenas et al.
(1992)
"See Appendix A for acronyms and abbreviations.
-------
7-167
-------
Table 7-26. Effects of Chronic Ozone Exposure on Pulmonary Function, Respiratory
Symptoms, and Chronic Respiratory Disease1
Concentrations(s)
Dg/m
' Pollutants and Environmental
Variables
Study Description
Results and Comments
Reference
0.033 median
average annual
hourly value
65
Study of relationship of air pollution to levels of FVC,
FEV,, and PEFR based on 1976 to 1980 supplement
to NHANES and data from EPA SAROAD monitoring
system; subjects 6 to 24 years of age; exposure values
based on hourly O3 values for previous 365 days; data
for TSP, NO2, and SO2; and data for important
demographic, smoking, and health covariates.
Nonlinear relationship between annual average O3
and function measurements with threshold at
approximately 0.040 ppm; findings limited by
inability to control for multiple pollutant effects,
relatively crude assignment of exposure; data
consistent with effect on forced flow at O3 levels at
or below 0.12 ppm.
Schwartz (1989)
0.034 to 0.050 90th
percentile annual
mean 1-h daily max
67 to 98
, 1983 to 1984 cross-sectional study of 2nd- to 6th-
grade students in Ontario and Manitoba, Canada; data
on SO,, NO,, nitrates, and sulfates; respiratory health,
demographic, smoking, and home cooking fuel data;
and spirometry.
Ontario town had more O3 days > 0.080 ppm; small
decrements (02 %) in FVC and FEV, were found in
the Ontario town compared to the Manitoba town;
any O3 possible effects were completely confounded
with SO4 effects.
Stern et al.
(1989)
CO
0.024 to 0.031
annual mean 1-h
daily max
47 to 61
31985 to 1986 cross-sectional study of
7- to 11-year-old children from
(n = 3,945) five rural towns in Ontario and five towns
in Saskatchewan, Canada; data on SO,, sulfates, NO3,
NO,, and PM10; respiratory health multiple covariates;
spirometry including flow at mid-lung volumes.
Ontario towns had higher levels of O3 and SO, in
summer months and for 90th and 99th-percentiles of
distributions; 90th percentile mean 1-h maxima were
80 ppb vs. 47 ppb for O3 and 11.5 Dg/m3 vs.
3.1 Dg/m3 for SO4; magnitude of FEV, and FVC
effects was similar to Stern et al. (1989); no effect
for mid-volume flows, except for subjects with
asthma; coincidence of increased O3 and SO,
precludes definite statements concerning O3 effects.
Stern et al.
(1994)
0.008 to 0.118
average hourly
concen-
trations, 1974 to
1979
16 to 231
3Study of chronic respiratory symptoms in adults with
use of 1979 National Health Interview Survey data and
1974 to 1979 EPA SAROAD data; data on respiratory
health, demography, and smoking; and data for TSP.
Data for only 29% of those eligible could be used;
average hourly O3 concentration over period 1973 to
1979 associated with report of sinusitis and hay
fever after control for covariates and TSP; no
association with asthma or emphysema; large
amount of data reduction, lack of adequate exposure
assignment, lack of occupational exposure histories,
and lack of adequate data on other pollutants make
results very difficult to interpret.
Portney and
Mullahy (1990)
-------
Table 7-26 (cont'd). Effects of Chronic Ozone Exposure on Pulmonary Function, Respiratory
Symptoms, and Chronic Respiratory Disease1
Concentrations(s)
ppm Qg/m
0.015 to 0.052 29 to 102
average HMV
0.10 to 0.20 196 to 392
3-mo mean daily
peak hourly
values for
Lancaster and
Glendora,
respectively
0.04 to 0.07 78 to 137
mean peak daily
peak hourly
values 1972 to
1981; Long
Beach and
Lancester,
respectively
Not reported
Pollutants and
Environmental
Variables Study Description
O C. ross-sectional study of children ages 6 to
15 years in a community in Austrian alps
divided into three zones based on SO,, NO,,
and O3; respiratory health, demographic, and
spirometry data.
Oxidants 5-year follow-up of Lancaster and Glendora,
CA, cohorts; from UCLA population study of
CORD restricted to nonsmoking, non-
Hispanic whites, 7 to 59 years old.
Oxidants 5- to 6-year follow-up of Lancaster and Long
Beach, CA, cohorts from UCLA CORD
study; Long Beach with higher NO2, SO4, and
TSP than Lancaster.
Oxidant Prevalence of respiratory symptoms in
Results and Comments
Only difference in respiratory history was increased adjusted prevalence of
asthma in zone with highest O3 (6.4%; 0.052 ppm HMV) vs. the zones
with lower O3 concentrations (4.8%; 0.015 ppm HMV; 2.7%, 0.026 ppm
HMV); no meaningful differences in spirometry indices; data limited by
use of single monitoring site for 1,200 km2 area; effects of SO, and NO, on
asthma prevalence not well studied.
No difference in respiratory symptoms over follow-up for either
community; across all age groups, slope of Phase III of N, washout
deteriorated more rapidly in Glendora; in subjects D14 years of age, more
rapid decrease in spirometric indices in Glendora; interpretation hampered
by large losses to follow-up, inability to disentangle multiple pollutant
effects.
All reported excess functional decline for Long Beach likely due to bias in
decline estimates between locations; data not useful with regard to possible
O3 effects.
Slightly increased prevalence of respiratory symptoms in high pollution
Reference
Schmitzberger et al.
(1993)
Detelsetal. (1987)
Detelsetal. (1991)
Hodgkin et al. (1984)
nonsmoking Seventh Day Adventists residing area; after adjusting for covariables, 15% greater risk for COPD due to air
for at least 11 years in high- (South Coast) and pollution (not specific to oxidants); past smokers had greater risk than
low- (San Francisco, San Diego) never-smokers; when past smokers were excluded, risk factors were
photochemical air pollution areas of similar. In addition, insufficient exposure assessment and confounding by
California; ARB regional air basin monitoring environmental conditions limit the quantitative use of this study.
data for oxidants, NO2, SO2, CO, TSP, and
SO4 from 1973 to 1976.
-------
Table 7-26 (cont'd). Effects of Chronic Ozone Exposure on Pulmonary Function, Respiratory
Symptoms, and Chronic Respiratory Disease1
Concentrations(s)
PPi"
Dg/m
Pollutants and
Environmental Variables
Study Description
Results and Comments
Reference
Not reported
Oxidant Cross-sectional analysis of above populations; uses
hours above various "threshold" values for oxidants,
TSP, SO, based upon California, EPA, and World
Health Organization max levels; period covered, 1966
to 1976; data available for important covariates (sex,
occupation, environmental tobacco smoke, race, age,
education, past smoking).
OX (10) most significantly associated with COPD after adjustment
for covariates; number of hours above higher thresholds less
significant; when TSP, SO,, and OX (10) entered in same regression,
TSP (200) only pollutant associated with COPD; high correlation
between OX (10), TSP (200), and SO, (hours more than 4 pphm).
Improved exposure assessment over previous paper; however, no
clear statement possible about effects of oxidants due to colinearity
with TSP and SO,.
Euler et al.
(1988)
Not reported
Same as Abbey et al. (1991a) but analysis applied to
COPD severity and a "multi-pollutant" analysis
performed; also evaluated effect of using data for
different time periods of ambient air monitoring.
Cumulative incidence of COPD symptoms when each pollutant
entered separately, similar to above study; joint effects of OZ (10)
and TSP (200) and mean concentrations of each pollutant evaluated
only for cumulative asthma incidence; TSP (200) entered logistic
regression in preference to OZ (10) but mean O3 concentration
entered in preference to mean TSP; change in asthma severity
associated with mean O3 concentration (1977 to 1987) and with
exceedance frequency for OZ (10), OZ (12), and TSP (200)
considered separately; findings for asthma severity similar to
cumulative incidence when TSP and O3 evaluated together; in no
analysis did TSP and O3 both remain jointly significant, nor were
there any interactions; data unable to unequivocally disentangle
effects of individual pollutants.
Abbey et al.
(1993)
'See Appendix A for acronyms and abbreviations.
-------
The basic population for these studies represents California-resident, Seventh-Day
Adventists aged D25 years of age who had lived 11 years or longer (as of August 1976) in
either a high-oxidant-polluted area (South Coast Air Basin [Los Angeles and vicinity] and a
portion of the nearby Southeast Desert Air Basin) or a low-pollution area (San Francisco or
San Diego). This sample was supplemented by an additional group of subjects who met the
11-year residence requirement but who came from low-exposure rural areas in California. The
total, baseline sample (March 1977) comprised 8,572 individuals, of whom 7,267 enrolled.
From this group, 109 current smokers and 492 subjects who had lived outside of the
designated areas for a portion of the previous 11 years were excluded. Detailed respiratory
illness and occupational histories were obtained. In these studies, "COPD" refers to "definite
chronic bronchitis", "definite emphysema", and "definite asthma" as defined by the study
questionnaire. Measures of pulmonary function are not included.
Air monitoring data were obtained from the California ARB monitoring system.
Ninety-nine percent of the subjects (excluding the rural supplement) lived at a distance from
the nearest ARB monitoring site that was considered to provide relatively reliable
concentration estimates for the outdoor, ambient environment at their residence.
Concentrations at the monitors were interpolated to the centroid of each residential zip code
from the three nearest monitoring sites with the use of a 1/R' interpolation. Subsequent
development of exposure indices took account of the improvements in ARB data after 1973.
Data were available for total oxidants, O., TSP, SO., NO., CO, and SO. (excluding 1973 to
1975).
The initial report from this study was summarized in the 1986 criteria document
(U.S. Environmental Protection Agency, 1986). Based upon a multiple logistic regression that
adjusted for smoking, occupation, race, sex, age, and education, it was estimated that
residence in the South Coast Air Basin conferred a 15% increase in risk for prevalent COPD.
No estimates of exposure were provided, and the data were considered to be of limited utility.
In their 1988 publication, Euler et al. provided exposure estimates based on the
cumulative number of hours, over 11 years prior to the baseline, that individuals lived in
environments at various oxidant thresholds, beginning at 10 pphm [OX (10)] (196 Dg/m') and
the total dosage to which they would be exposed. The estimates in this report did not correct
for time spent indoors. When the OX (10) was the only pollutant considered, each 750 h/year
increment in exposure was associated with a 20% increase in risk for COPD in a multiple
logistic regression analysis that adjusted for effects of occupation, passive exposure to tobacco
smoke, personal smoking, sex, age, race, and education (baseline data only). Moreover, the
data were compatible with a threshold effect at 10 pphm. However, when hours above a TSP
concentration of 200 Dg/m' [TSP (200)] and SO. concentration of 4 pphm were included in the
logistic regression model, only TSP (200) was associated with the occurrence of COPD. No
significant interactions were found between the various pollutant thresholds. The authors
noted that their failure to control for time spent indoors may have led to an underestimation of
the oxidant effect. Moreover, the fact that 74% of the variance of OX (10) was explained by
the other pollutants certainly reduced the power of this study to detect an independent effect of
oxidants on the occurrence of COPD. The authors also noted the limitations imposed by the
cross-sectional nature of the data that were used in this analysis. Thus, on the basis of this
study, no clear statement could be made about the chronic respiratory system effects of oxidant
exposure.
7-171
-------
A major improvement in the methods for assessment of exposure was presented in
Abbey et al (199la). Previous exposure estimates were refined by the computation of "excess
concentrations" (concentration minus cutoff, summed over all relevant time periods and
corrected for missing data). Exposures also were corrected for time spent at work and time
away from residence, with estimates provided for the environments where work occurred and
for geographic areas away from residence. The quality of the interpolations (in terms of
distance of monitor from residence zip codes) also was evaluated and incorporated into the
estimates. Adjustments were made for the time spent indoors by individuals. New indices
were developed that were based on O., rather than on total oxidants. The investigators
demonstrated correlation coefficients of 0.98 between monthly mean total oxidants and O. at
concentrations D12 pphm. (It should be noted that a more appropriate comparison would have
been between the mean and the differences of the two measurements.)
The above estimates were applied to data that included 6 years of follow-up of the
study population (Abbey et al., 1991b). This analysis focused on incident occurrence of
obstructive airways disease (AOD—same definition as for COPD above). Incident symptoms
of AOD were significantly associated with hours above several TSP thresholds, but not with
hours above any O. threshold. There was a suggestion of an association between hours above
10 pphm O. [OZ (10)] and the 6-year cumulative incidence of asthma [RR for 500 h/year
above OZ (10) = 1.40 (95% CL = 0.90 to 2.34)] and definite bronchitis (RR =
1.20 [95% CL = 0.97 to 1.53]). Approximately 43% of the study population experienced at
least 500 h in excess of the OZ (10) criterion. Cumulative incidence estimates were adjusted
with the use of Cox proportional hazard models for the same variables noted in the original
publication of Hodgkin et al. (1984), as well as the presence of possible symptoms in 1977 and
childhood respiratory illness history. None of the analyses included both O. and TSP
thresholds. No data were provided on the details of the subjects available for the prospective
analysis and their representativeness versus the entire base population. Therefore, assuming
no bias due to selective loss to follow-up, these data are consistent with a small O. effect and
are limited by the same considerations of colinearity and subsequent reduction of power noted
above.
Another analysis by Abbey et al. (1993) evaluated changes in respiratory symptom
severity with the TSP and O. thresholds noted above. In this analysis, logistic regression,
rather than Cox proportional hazard modeling, was used to assess cumulative incidence of
components of the COPD/AOD complex; and multiple, linear regression was used to evaluate
changes in symptom severity. When O. was considered by itself, there was a trend toward an
increased risk of asthma for a 1,000-h average annual increment in the OZ (10) criterion (RR
= 2.07, 95% CL = 0.98 to 4.89). In this analysis, there was a suggestion that recent ambient
O. concentrations were more related to cumulative incidence than past concentrations. Change
in asthma severity score was associated significantly with the 1977 to 1987 average annual
exceedance frequency for O. thresholds of 10 and 12 pphm. No significant effects were found
for COPD or bronchitis alone. In contrast to the above study of cumulative incidence, the
investigators carried out an analysis in which TSP (200) and OZ (10) were allowed to compete
for entry into a model to evaluate asthma cumulative incidence and changes in severity. In the
cumulative incidence model that employed exceedance frequencies (number of hours above
threshold), TSP (200) entered before OZ (10); when average annual mean concentrations were
used, O. entered before TSP. From this, the authors concluded that both TSP and O. were
relevant to asthma cumulative incidence. In no case did both pollutants remain significant
7-172
-------
simultaneously in the same regression. No interactions were observed between TSP and O. for
either metric. A similar result was observed for change in asthma severity. As in previous
analyses, there was a high correlation between TSP (200) and OZ (10) exceedance frequencies
(0.72) and their respective average annual mean concentrations (0.74).
The AHSMOG study represents the most extensive effort to date to provide realistic
exposure estimates within the constraints of a large, population-based study. Moreover, the
exposure estimates for photochemical oxidants have been tied to current O. levels and have
taken into account many of the sources of inaccuracy and imprecision in the assignment of
exposures to individuals (short of detailed personal monitoring). As such, they do represent a
considerable improvement over all other studies to date. Nonetheless, it is not possible from
these data to determine if there is an effect of O. on the outcomes that were studied. This
largely is due to the difficulty of partitioning effects between O. and particles.
Other Studies. Subsequent to the publication of the 1986 criteria document, two
additional publications have emerged from the UCLA study (Detels et al., 1987, 1991). The
data presented are derived from the same population bases that were used in previous
publications; and, therefore, they are subject to the same limitations that were cited in the
introduction to this section.
In 1987, Detels et al. reported a 5-year follow-up study of white, non-Hispanic
subjects from the Lancaster and Glendora study areas. The 12- and 3-mo mean peak hourly
total oxidant values from 1972 to 1982 for Lancaster and Glendora were 7 and 10 pphm and
11 and 20 pphm, respectively. Only 47 and 58% of subjects, respectively, were retested with
both the questionnaire and measures of lung function. Effects of air pollution on the days of
testing were evaluated by comparing lung function test results in a subgroup of individuals who
were tested three times at 3- to 4-mo intervals. No effect was observed, but the power to find
differences was low. Over the follow-up period, there were no changes in reported respiratory
symptoms for either community. In adults (D19 years of age) who never smoked, all
spirometric and nitrogen-washout results showed more rapid deterioration in Glendora.
Differences were significant only for mid-expiratory flows and for slope of Phase III from the
nitrogen-washout curve. The effects were greater in females, in whom changes in FEV. also
were significant. In subjects less than 19 years of age, only changes in slope of Phase III were
significant, although FVC in Glendora females was lower than that observed in Lancaster.
The results of this study remain limited by the lack of adequate exposure data and
the failure to control for the possible effects of other ambient pollutant differences between the
communities. Problems with loss to follow-up represent a significant issue, especially for the
pulmonary function measurements, given that approximately 50% of the original subjects were
not available for repeated testing. Baseline comparability is also of concern because subjects
who were retested in Lancaster had a better slope of Phase III than those not retested. Because
this measure most consistently differed between the two study communities, the possibility of
selection bias is very real. Overall, these results do not strengthen the usefulness of this study
for the attribution of an effect of oxidant exposure on respiratory health.
The 1991 report from the UCLA group compared Lancaster with Long Beach, the
latter area with relatively high levels of SO., sulfates, NO., and hydrocarbons, as well as
increased total oxidant levels (mean 1-h daily peak values, 1972 to 1982, 30 pphm versus
110 pphm, respectively) (Detels et al., 1991). As above, the analysis was restricted to non-
Hispanic whites who never smoked cigarettes and with 5 years of follow-up. Only 47% of the
Lancaster cohort and 44% of the Long Beach cohort had pulmonary function retested on two
7-173
-------
occasions. Over the age range of 25 through 59 years, changes in slope of Phase III of the
nitrogen-washout curve and most spirometric indices were significantly worse in Long Beach,
compared to Lancaster. In subjects under 25 years of age, there were significant differences in
slope of Phase III, especially in subjects 7 to 10 years of age.
All of the limitations identified for the 1989 report apply to this report as well.
Moreover, comparison between the two communities of the interlaboratory differences (mobile
laboratory versus UCLA reference laboratory; 3% sample) indicated that average annual
decrements in FEV. were exaggerated by D13 mL/year (standard error +7 mL/year) in Long
Beach versus D2 mL/year (+7 mL/year) in Lancaster. Application of this difference to the
data in their Table 6A would suggest that the "significant" difference in FEV. for both males
and females may be largely, if not entirely, due to bias. Thus, all of the functional differences
reported in this study are suspect on this basis alone. This, of course, ignores any additional
biases that may have been due to the large losses to follow-up in both communities.
A number of additional studies have addressed data relevant to the chronic effects of
O. on respiratory health (see Table 7-26).
Schwartz (1989) evaluated the effect of air pollution on children and young adults
ages 6 to 24 years with the use of data derived from NHANES II (Second National Health and
Nutrition Examination Survey, February 1976 to 1980). All individuals in each census tract
were assigned average pollutant values derived from monitors located within 10 miles of the
centroid of the census tract. Average hourly values for the 365 days preceding spirometry
were used, and an annual average was created for O. (EPA Storage and Retrieval of
Aerometric Data [SAROAD] database). For O. (chemiluminescence and ultraviolet [UV]
spectroscopy), six of the seven hourly readings between 11:00 a.m. and 5:00 p.m. were
required to include a day's data. Only 1,005 of the 3,922 (25.6%) of the subjects lived close
enough to a monitor to have O. exposures assigned to them. Data for TSP, NO., and SO.
were assigned to 47.1, 13.6, and 21.2% of the subjects, respectively. Analyses were restricted
to consideration of single pollutants, because the author reported that there was insufficient
overlap between the locations where data were available for all or any combination of
pollutants. Data for a variety of relevant personal and demographic covariates were available.
Statistical analyses appropriate to the correlation structure of the data (induced by the sampling
design of NHANES) were utilized. There was a nonlinear relationship between the annual
hourly average O. concentration and FVC, FEV., and PEFR. A threshold of effect around
0.04 ppm was observed, above which there appeared to be a linear decline in FVC (only data
shown graphically). The effect persisted after control for sex, race, age, family income,
educational level, chronic respiratory symptoms, and smoking history. Results were little
affected by region or use of a 2-year averaging time. Ozone levels above the threshold were
significantly associated with an FVC <70%, a result not seen for TSP but observed for NO..
The major limitation of the Schwartz (1989) analysis is the inability to distinguish
between the effects of O., TSP, and NO. and the choice of only a single metric for O. (hourly
average). Support for the former concern can be seen in the similarity of the effects of NO.
and O. in the logistic regression analyses, which suggests that the results could reflect the joint
effect of a number of species in a complex oxidant environment. The operating assumption
that near-term (1- to 2-year) exposure reasonably reflects a "lifetime" of exposure is highly
suspect in the mobile U.S. population. In fact, restriction of the analysis to subjects who still
resided in the state in which they were born led to slight reductions of O. effect, especially for
FEV. and PEFR. Despite these limitations, the data do suggest that, for children and young
7-174
-------
adults, if there is a chronic O. effect (or, more accurately for these data, a subchronic effect)
on lung function, it could occur at levels at or below 120 ppb. However, the particular pattern
of exposure (peaks, season, etc.) that may be relevant cannot be discerned from these data.
In 1989, Stern et al. reported a cross-sectional study, conducted in 1983 and 1984,
of the relationship between respiratory health effects of second- through sixth-grade children in
two Canadian communities (one in southern Ontario and one in southern Manitoba). The
Ontario region was characterized by low levels of gaseous pollutants (SO. and NO.) and
moderately elevated levels of particulate sulfate, FPs, and O.. Frequent episodes of elevated
sulfate and O. concentrations occurred in the summer and early fall. The Manitoba community
was not subject to the same pattern of air pollutants. Gases and O. (measured by
chemiluminescence) were sampled continuously, and TSP, sulfates, and total nitrates were
sampled every sixth day. Fixed monitoring stations were established at the center of each
community, and monitoring was carried out from October 1983 to April 1984. Ozone
measurements in Ontario were derived from sites between 35 to 45 km from the study area.
Average annual maximum O. concentrations were similar in the two communities (0.136 and
0.130 ppm in Ontario and Manitoba, respectively), but the frequency of elevated O. events
(> 0.080 ppm, Canadian standard for 1-h maximum) was more frequent in Ontario (30 days)
than in Manitoba (3 days) in 1983. Ninety-two percent of the subjects (n = 1,317) provided
data from detailed questionnaires, but only 70% (1,010) provided spirometric data (tested in
fall and winter months). There were no meaningful differences in the prevalence of all of the
respiratory health outcomes studied after adjustment for parental smoking, gas cooking, sex,
length of residence, parental education, and past respiratory illness history. Ontario children
had a 2% lower FVC (adjusted for age, sex, height, and parental smoking) and a 1.7% lower
FEV.; both differences were statistically significant. The differences were somewhat greater
when children with underlying respiratory illness or symptoms were excluded from the
analyses. These data are very difficult to interpret in relation to O. due to the marked
colinearity between the O. and sulfate levels in Ontario. Moreover, the differences observed
in lung function are very small (an average of 50 mL and 40 mL for FVC and FEV.,
respectively) and of questionable importance without further follow-up data on the subjects.
Such follow-up data would need to attempt to identify whether the small decrements observed
are "across the board" with respect to the overall population or the result of decrements in a
susceptible subset of the population, particularly a set of children at the lower end of the
pulmonary function distribution. In these children, small decrements might be associated with
adverse respiratory effects as a result of their already lowered (absolute or relative) levels of
lung function.
Stern et al. (1994) extended the 1983 and 1984 (Stern et al., 1989) study to 10 rural
Canadian communities. Five towns in southwestern Ontario and five towns in Saskatchewan
were selected and studied between September 1985 and March 1986. Children 7 to 11 years of
age were studied (n = 3,945) with techniques similar to the previous study. In 1986, SO.,
NO., and O. were monitored continuously through a 10-site network (one site in each town).
Particles were sampled every 3 days for 24 h in Saskatchewan and every 6 days in Ontario.
Annual mean 1-h maximum O. concentrations were slightly higher in Ontario, but the 90th and
99th percentile values were much greater (90th: 80 ppb versus 47 ppb; 99th: 115 ppb versus
57 ppb). This was particularly true for the months of June to August. The levels of PM.. and
nitrate did not vary between the areas and were well within the Canadian Ambient Air Quality
7-175
-------
Objectives. Annual mean SO. levels were three to four times greater in Ontario communities
(6.6 Dg/m' versus 1.9 Dg/m').
The adjusted (age, sex, parental education, gas cooking, and parental smoking)
prevalence of respiratory symptoms did not differ among the 10 communities. Adjusted
(height, weight, plus the above adjustment factors) FVC and FEV. averaged 1.7 and 1.3%
less, respectively, in the five Ontario towns. No differences were observed for PEFR,
FEF , or V The results did not change when the analysis was restricted to life-long
residents or to children without respiratory symptoms. Although not statistically significant,
Ontario children with doctor-diagnosed current asthma had FEF and V levels that
were 6.6 and 6.5%, respectively, lower than similar children in Saskatchewan. Overall, the
prevalence of asthma was 4% for the entire sample.
These results are consistent, in terms of the magnitude of the FEV. and FVC
effects, with those of the previous Stern study (Stern et al., 1989). In addition, these data
provide suggestive evidence of enhanced effects for children with current asthma. The two
major limitations of the study are recognized by the authors: (1) the effects observed cannot be
attributed to O. or to SO. (or acid) aerosols and could be due to either part of the pollutant
mixture or attributed to the combination of the component, and (2) the differences in the mean
values reported do not take into account the variability in the pulmonary function distribution
and the variability of responses across the distribution (see above).
Portney and Mullahy (1990) used the 1979 U.S. National Health Interview Survey
and SAROAD data to explore the relationship between O. and TSP and chronic respiratory
disease. Average hourly O. concentrations from 1974 to 1979 were used; data from 1974 to
1979 and data from 1979 alone were evaluated. Individuals were matched to the nearest
centroid of the census tract in which they lived in 1979. Individuals were excluded if they
lived >20 miles from the nearest monitor. Only 29.3% of the 4,500 adults surveyed who
participated in the smoking and respiratory disease supplemental interview and for whom
residential data were available could be included. Seven different model specifications (probit
analysis) evaluated cumulative (5-year) and 1-year effects of O. on various respiratory
diseases. Hourly average O. concentrations, but not TSP concentrations, over 1974 to 1979
were significantly associated with the report of sinusitis and hay fever after control for
smoking, sex, income, race, education, temperature, and stability of residence. In contrast,
neither O. nor TSP were associated with reported asthma and emphysema. An enormous
amount of data reduction, the lack of individual exposure data, lack of specification of the age
and sex distribution of the study population, lack of data on occupational exposures, the use of
a single O. metric, and the restricted formulation of the particulate data all severely limited the
usefulness of these data.
Kilburn et al. (1992) studied the effect of "air pollution" on expiratory flows and
vital capacity in Mexican-American children in Los Angeles. In 1984, 556 second- and
fifth-grade students were studied, and 251 of these were studied again in 1987. The analytical
strategy, the losses to follow-up, and the lack of reasonable exposure data make the data from
this study virtually uninterpretable.
A study by Castillejos et al. (1992) evaluated the effects of acute exposures to
ambient O. concentrations on pulmonary function and respiratory symptoms. One-hundred
and forty-eight 9-year-old children in the southwest part of Mexico City were studied between
January and June 1988. Weekly spirometric measurements were made over 10 weeks.
Ambient air data were obtained from the monitoring system maintained by the Mexican
7-176
-------
government and included hourly values for temperature, RH, and O. concentration. Ozone
concentration exceeded 120 ppb on 74% of days and "frequently" exceeded 240 ppb. No data
are presented for SO. or particles. All subjects had to live within 5 km of a monitoring station.
The study demonstrated that levels of FEV. and FEF were associated with mean hourly
O. levels in the preceding 24, 48, and 168 h. The authors interpreted their data as consistent
with a subchronic effect of O. on measures derived from spirometry that may be due to an
"inflammatory process". However, this interpretation seems at odds with the statement in the
paper that the initial FEV. measurements for the group did not differ from those observed in a
comparable age group in the Harvard Six Cities Study who were not exposed to
O. concentrations as high as those reported in this study. If the overall level of pulmonary
function of this group does not differ from those children who live in ambient environments
with far lower O. concentrations, the data would suggest that the subchronic effects observed
are not translated into persistent abnormalities, at least as can be observed with spirometry.
Austrian investigators (Schmitzberger et al., 1992) described a cross-sectional study
of the effects of O. on the respiratory health of 1,156 children, ages 6 to 15 years. Pulmonary
function in two different areas with differing "annual" O. concentrations (actual metric on
which "annual" based not given) were compared (52 ppb versus 26 ppb). No differences were
observed for FVC. All flow measures (FEV., FEF... , and FEF... ) were significantly lower
in the children in the "high"-O. area. These data are of limited value for a variety of reasons,
the most important being lack of individual exposure data, lack of data on other pollutants, the
uninterpretable specification of "annual" O. concentration, lack of data on chronic respiratory
illness (especially asthma), and the lack of data on smoking for the teenage members of the
subject group.
Schmitzberger et al. (1993), following up on their preliminary data (Schmitzberger
et al., 1992), studied additional subjects in the Austrian Tyrolian Alps. Three zones were
identified based upon ambient air conditions: (1) Zone 1 was characterized by annual mean
SO. (UV fluorescence) of 20 Dg/m', monthly mean NO. (Palmes tubes) of 17 ppb, and annual
mean O. (chemiluminescence) of 15 ppb (maximum half-hour mean = 102 ppb); (2) Zone 2
was characterized by values of 14 Dg/m', 13 ppb, and 26 ppb (112 ppb), respectively; and
(3) Zone 3 was characterized by 12 Dg/m', 8 ppb, and 52 ppb (146 ppb), respectively.
Children ages 6 to 15 years who lived in the study areas for D3 years were enrolled.
Respiratory health questionnaire data and forced expiratory flows were obtained. Full data
were available from 81% of the enrolled subjects. Adjusted (age, sex, environmental tobacco
smoke, socioeconomic status, and home heating) levels of FVC and forced flows did not
follow the gradient in O. concentrations. Although Zone 3 differed significantly from Zone 2
on several measures, there were no meaningful differences with Zone 1. Adjusted asthma
prevalence was highest in Zone 3 (6.4% versus 4.8 and 2.7% for Zones 1 and 2, respectively).
There were no differences for other respiratory symptoms. Although the authors conclude that
"residence in the area of elevated O. increases the risk. ..of low small airway-related lung
function," careful inspection of the data does not support this conclusion. This conclusion is
based on the supposed increased frequency of FEV. of less than 70% in Zone 3 relative to the
other zones, although the specific data are not provided. Moreover, the mean levels for all
functional measurements are lowest in Zone 1, the zone with the lowest O. concentrations and
the highest SO. and NO. concentrations. This report is handicapped by the lack of any
information that could be used to access individual exposures. Moreover, only a single
monitoring station was employed that was placed at the center of Zone 1, which itself was at
7-177
-------
the center of the study area (1,200 km'). No information is provided as to how the
concentrations of the various pollutants were estimated for Zones 2 and 3. Therefore, these
data virtually are of no quantitative value.
Other Chronic Disease Morbidity and Mortality
Only AHSMOG study has provided any data on possible O.-related health effects
other than those related to the respiratory system or malignant disease of the respiratory system
(Abbey et al., 1991b; Mills et al., 1991) (see Table 7-27). The population studied and the
assignment of exposures has been presented previously (Hodgkin et al., 1984; Euler et al.,
1988; Abbey etal., 1991a).
In their initial study based on 6 years of follow-up, Mills et al. (1991) found that for
500 h in excess of the OZ (10) threshold, there was a relative risk of 2.24 for respiratory
cancer incidence after adjustment for a number of factors listed previously. When the TSP
(200) and OZ (10) thresholds were allowed to compete for entry into a Cox proportional
hazards model for respiratory cancer incidence, the O. threshold entered in preference to TSP.
Ozone exposure was not associated with excess respiratory cancer mortality or incidence of
nonrespiratory cancer over the 6-year follow-up period.
A second chronic disease study from the AHSMOG population extended the above
observations to include myocardial infarction and all-cause mortality (Abbey et al., 1993).
Incident chronic respiratory disease also was included in this analysis. Ambient levels of
O. were not associated with incidence of myocardial infarction at any of the threshold indices
that were tested. Neither the mean concentration of O. nor any of the thresholds were
associated with incidence of chronic respiratory diseases, as previously defined. However,
there was a trend toward an association between 6-year cumulative incidence of asthma and
500-h exceedance of the OZ (10) threshold (RR = 1.40, 95% CL = 0.99 to 2.34).
7.4.2.3 Conclusions
The body of data that has accumulated since publication of the previous air quality
criteria document for O. (U.S. Environmental Protection Agency, 1986) provides only
suggestive evidence for health effects of chronic O. exposure. Most of the studies suffer from
one or another of the following limitations: (1) simplistic assignment of exposure in terms of
choice of O. metrics or adequate adjustment for relevant covariates and (2) lack of ability to
isolate effects related to O. from those of other pollutants, especially the particulate fraction.
The AHSMOG study has made substantive progress in the problem of the assignment of
individual exposures (Abbey et al., 1991a). Unfortunately, the results from this study cannot
disentangle the effects of chronic O. exposure from those due to chronic exposure to the
particulate fraction of ambient pollution. The study also lacks sufficient power to evaluate the
possibility of interactions between O. and particulate pollution in relation to health effects.
Thus, the overall data are not conclusive, but current evidence is suggestive of possible health
effects from chronic exposure to O..
7-178
-------
Table 7-27. Effects of Chronic Ozone Exposure on the Incidence of
Cardiovascular and Malignant Diseases*
Concentrations(s)
Dg/m
Pollutants and
Environmental
Variables
Study Description
Results and Comments
Reference
Not reported
Hodgkin et al. Hodgkin et al. (1984) and Abbey et al.
(1984) (199la); analysis based upon exceedance
frequencies, 1973 to 1977, and cancer
cumulative incidence, 1977 to 1982.
Exceedance of OZ (10) threshold
borderline associated with respiratory
cancer; no association with mean
concentration; multipollutant analysis with
TSP (200) and OZ (10), only OZ (10)
entered the logistic regression for
respiratory malignancy; TSP (200) was
significant for females for all malignancy;
no association between O3 and any measure
of cancer mortality; overall results
suggestive of O3 effect on respiratory
cancer morbidity at level of exposure
within range experienced by large
percentage of study population.
Mills et al. (1991)
Not reported
See above
See above
No association between any O threshold
and all-cause mortality or incidence of
myocardial infarction.
Abbey et al. (1991b)
"See Appendix A for acronyms and abbreviations.
-------
7.5 Summary and Conclusions
7.5.1 Controlled Human Studies of Ozone Exposure
7.5.1.1 Effects on Pulmonary Function
Controlled human O. exposure studies have provided the strongest and most
quantifiable exposure-response data on the health effects of O.. This chapter reviews the
results of studies involving subjects exposed to O. concentrations ranging from 0.08 to
0.75 ppm O. while at rest or during CE or IE of varying intensity for periods of up to 8 h. In
many of these studies, small sample size and suboptimal experimental design limit the ability to
generalize to the larger population. Of particular concern in considering studies with small
sample sizes is the risk of making a beta (Type II) error: the incorrect conclusion that no
difference exists between treatments when comparisons are not significantly different. The
likelihood of making a Type II error greatly limits the ability to determine the minimum
O. concentration that results in a significant pulmonary response in the larger population. As a
result, the conclusions drawn from many of the studies cited in this chapter may underestimate
the presence of responses at low O. concentrations in healthy, young adults.
Healthy Subjects
Results from studies of at-rest exposures to O. for 2 h in healthy adult subjects have
demonstrated decrements in forced expiratory volumes and flows occurring at and above
0.5 ppm O. (Folinsbee et al., 1978; Horvath et al., 1979). Airway resistance is not clearly
affected during at-rest exposure to these O. concentrations.
With moderate IE for 2 h, eliciting a V. of 30 to 50 L/min, decrements in forced
expiratory volumes and flows, secondary to decreases in 1C, have been observed in healthy
adult subjects at and above 0.3 ppm O. (Folinsbee et al., 1978; Seal et al., 1993). With IE (V.
D 65 L/min), pulmonary symptoms and decrements in forced expiratory volumes and flows are
present following 2-h exposures to 0.12 ppm O. (McDonnell et al., 1983). Symptoms are
present and decrements in forced expiratory volumes and flows occur at 0.16 to 0.24 ppm
O. following 1 h of continuous heavy exercise (V. D 55 to 90 L/min) (Adams and Schelegle,
1983; Folinsbee et al., 1984; Avol et al., 1984; Gong et al., 1986) and following 2 h of
intermittent heavy exercise (V. D 65 to 68 L/min) (McDonnell et al., 1983; Kulle et al., 1985;
Linn et al., 1986). With longer exposures of 4- to 8-h duration, responses have been observed
at lower O. concentrations and lower ventilation rates. In the range of concentrations between
0.08 and 0.16 ppm, a number of studies using moderate IE and durations between 4 and 8 h
have shown significant responses under the following conditions: 0.16 ppm for 4 h of IE at V.
D 40 L/min (Folinsbee et al., 1994), 0.08 to 0.12 ppm for 6.6 h of IE at V. D 35 to 40 L/min
(Folinsbee et al., 1988; Horstman et al., 1990), and 0.12 ppm for 8 h of IE at V. D 40 L/min
(Hazucha et al., 1992). Symptom and spirometry responses were increased, with increased
duration of exposure, O. concentration, and total ventilation. Airway resistance is only
modestly affected with moderate or even heavy exercise combined with O. exposure to
concentrations as high as 0.5 ppm O. (Folinsbee et al., 1978; McDonnell et al., 1983; Seal
et al., 1993). Increased breathing frequency (f) and decreased V., while maintaining V.,
occur with exposure to 0.20 to 0.24 ppm O. when combined with heavy exercise for 1 to 2.5 h
(McDonnell et al., 1983; Adams and Schelegle, 1983). Differences in response to a given
O. concentration among individuals have been shown to be reproducible (Gliner et al., 1983;
McDonnell et al., 1985b), indicating some individuals are consistently more responsive to
O. than others.
7-180
-------
Group mean decrements in pulmonary function can be estimated roughly when
expressed as a nonlinear function of effective (i.e., exposure) dose of O., the simple product of
O. concentration, mean ventilation, and exposure duration (Silverman et al., 1976; Folinsbee
et al., 1978; Adams et al., 1981). The O. concentration appears to make a greater impact on
the pulmonary function response than does V. or exposure duration (Folinsbee et al., 1978;
Adams et al., 1981), and, indeed, Larsen et al. (1991) suggest an exponent of approximately
4/3 for the O. concentration. Another way of expressing this relationship is that doubling the
O. concentration under any given exposure scenario will have a greater impact on spirometry
responses than doubling either V. or exposure duration. However, at any given
O. concentration, the major external determinants of response are V. and exposure duration.
Because of the broad range of intersubject variability, and the inability to identify
characteristics that influence this variability (other than age), efforts to estimate or model
individual responses have so far been fruitless (McDonnell et al., 1993). Nevertheless,
prediction of group mean FEV. responses using the variables of the O. concentration, V., and
exposure duration can be successful (Adams et al., 1981; Folinsbee et al., 1978, 1988;
Hazucha, 1987; Hazucha et al., 1992; Larsen et al., 1991; McDonnell et al., 1993).
In acute O. exposure studies of 3 h or less in duration, the responses observed
during and following acute exposure to O. at concentrations between 0.12 and 0.50 ppm in
normal, healthy human subjects include decreases in TLC, 1C, FVC, FEV., FEF , and V.
and increases in SR.., f, and airway responsiveness. Ozone exposure also has been shown to
result in the symptoms of cough, PDI, SB, throat irritation, and wheezing. Similar responses
are seen with 4- to 8-h exposures in the O. concentration range between 0.08 and 0.16 ppm.
When viewed collectively, these physiological and symptom responses may be
separated into four general categories, including (1) symptoms, (2) changes in lung volume or
spirometry, (3) changes in R.., and (4) changes in airway responsiveness. These categories
are based on the absence of correlation between spirometry responses and change in R.. or
airway responsiveness. The attenuation by atropine of R.. but not spirometry responses
supports the notion of independent mechanisms. The attenuation by indomethacin or ibuprofen
of spirometry responses, but not changes in R.. or airway responsiveness, also supports this
categorization. A bronchodilator, albuterol, given to healthy subjects prior to O. exposure did
not prevent changes in spirometry, symptoms, or airway responsiveness. Symptoms ratings
represent reflex responses (e.g., cough) or a perceptual evaluation of consciously appreciated
afferent information (e.g., chest tightness, PDI), and it is therefore somewhat difficult to
separate these responses from the more objective physiological responses. However, cough
and pain on deep inspiration are related temporally to spirometry and breathing pattern
responses (i.e., volume-related changes). In repeated exposure studies, changes in spirometry
and breathing pattern become attenuated with the same time course as the changes in symptom
responses.
Recent multihour O. exposure studies indicate that spirometry and symptom
responses to concentrations as low as 0.08 ppm occur in healthy subjects with exposures
lasting 6 to 8 h. Prolonged exposures (8 h) at lower O. concentrations (0.12 ppm) also indicate
that there is a plateau of response to O. (Hazucha et al., 1992). Although suggested in
previous studies (Gliner et al., 1983), such a plateau is difficult to verify with the typical
duration of less than 2 h and the large responses seen with higher concentrations. The level of
the response plateau (i.e., the spirometry decrement at which the response no longer changes)
must be dependent on the dose rate of exposure (i.e., the product of the O. concentration and
7-181
-------
V.) because the magnitude of response at a higher dose rate may greatly exceed the response
plateau seen at a lower dose rate. Prolonged exposure studies also suggest that O.-induced
spirometry responses depend on the immediate exposure history. With relatively low dose
rates (e.g., Hazucha et al., 1992), responses to exposure that occurred 2 to 4 h previously may
influence the current response. The cumulative effect of exposures has not been studied at
higher dose rates, but greater persistence of effects may be expected based on the longer
recovery period at higher doses rates.
Recovery from O. exposure has not been systematically investigated in a large
group of subjects, but available information indicates that an initial phase of recovery proceeds
relatively rapidly, and some 40 to 65% of the acute response appears to be recovered within
about 2 h (Folinsbee and Hazucha, 1989). However, there is some indication that the
spirometric responses, at least to higher O. concentrations, are not fully recovered within 24 h
(Folinsbee and Horvath, 1986; Folinsbee et al., 1994). Collectively, these observations
suggest that there is a rapid recovery of O.-induced spirometry and symptom responses, which
may occur during resting exposure to O. (Folinsbee et al., 1977) or as O. concentration is
reduced during exposure (Hazucha et al., 1992), and a slower phase, which, in some cases,
may take at least 24 h to complete. Repeated exposure studies at higher concentrations
typically show that the response to O. is enhanced on the second of several days of exposure.
This enhanced response suggests a residual effect of the previous exposure, about 22 h earlier,
even though the preexposure spirometry may be the same as on the previous day. The absence
of the enhanced response with repeated exposure at lower O. concentrations may be the result
of a more complete recovery or less damage to pulmonary tissues.
Studies of repeated daily exposure to O. have shown that O.-induced changes in
spirometry, symptoms, R.., airway responsiveness, and airway inflammation are attenuated
with repetitive exposure. At higher dose rates, symptom and spirometry responses may be
enhanced on the second exposure. Attenuation of response within 3 to 5 days is a consistent
finding in repeated exposure studies, regardless of O. exposure dose rate, although attenuation
of response occurs after fewer exposures at the lower dose rates. The attenuation of response
appears to occur more rapidly in less responsive individuals (Horvath et al., 1981) or in
responsive subjects exposed to lower O. dose rates (Folinsbee et al., 1978, 1994). Loss of
attenuation is relatively rapid, with O. responsiveness being partially restored within 4 to
7 days (Kulle et al., 1982; Linn et al., 1982b), and normal responsiveness restored within 1 to
2 weeks after a series of 4 or 5 daily O. exposures. The attenuation of airway responsiveness
may occur somewhat more slowly than that of symptom and spirometry responses. Airway
inflammation also appears to attenuate, but less completely than the spirometry responses and
with a more gradual recovery (Devlin et al., 1995; Folinsbee et al., 1995). Some markers of
inflammation (e.g., LDH and elastase) have not demonstrated attenuation.
The mechanisms leading to the observed pulmonary responses induced by O. are
beginning to be better understood. The available descriptive data suggest several possible
mechanisms, some leading to alterations in lung volumes, symptoms, and exercise breathing
patterns, and others leading to increases in central and peripheral R... These mechanisms
appear to involve (1) O. reactions with the airway lining fluid and epithelial cell membranes;
(2) local tissue responses, including injury and inflammation; and (3) stimulation of neural
afferents (bronchial C fibers) and the resulting reflex responses and symptoms. Much remains
to be understood in order to determine how each event in this cascade contributes to the
pulmonary responses induced by acute O. inhalation in human subjects.
7-182
-------
Subjects with Preexisting Disease
Of the subpopulations studied, those with preexisting impediments in pulmonary
function and exercise capacity are of primary concern in evaluating the health effects of
O. because even a small change in function is likely to have more impact on a person with
reduced reserve. Inherent in these studies are several limitations that, at present, hamper the
ability to make definitive conclusions regarding the relative O. responsiveness of the various
groups of subjects studied. Furthermore, it is ultimately necessary to determine whether their
responses are representative of the larger population with preexisting disease. These
limitations include subject selection (in controlled studies, typically only people with milder
disease are selected or volunteer for study), standardized methods for the characterization of
some responses, and limited range of exposure doses utilized to examine some endpoints.
These limitations are evident in studies on subjects with COPD, chronic bronchitis,
and ischemic heart disease. For patients with COPD performing light to moderate IE, no
decrements in pulmonary function were observed after 1- and 2-h exposures to DO.30 ppm
O. (Linn et al., 1982a, 1983a; Solic et al., 1982; Kehrl et al., 1985) and only small decreases
in forced expiratory volume were observed for 3-h exposures of chronic bronchitics to 0.41
ppm O. (Kulle et al., 1984). Small decreases in arterial blood oxygen saturation also have
been observed in some of these studies, but the interpretation of these results and their clinical
significance is uncertain.
Similar limitations also apply to the early studies examining O. effects in adult and
adolescent asthmatics. Decrements in pulmonary function were not observed for adult
asthmatics exposed for 2 h at rest (Silverman, 1979) or with intermittent light exercise (Linn
et al., 1978) to O. concentrations of 0.25 ppm and less. Similarly, no significant changes in
pulmonary function or symptoms were found in adolescent asthmatics exposed for 1 h at rest to
0.12 ppm O. (Koenig et al., 1985) and in adolescent asthmatics and nonasthmatics exposed to
0.12 and 0.18 ppm O. with intermittent moderate exercise up to 1 h (Koenig et al., 1987,
1988), although a small decrease in forced expiratory flow at 50% of FVC was observed in
asthmatics after exposure to 0.12 ppm O.. More recent observations by Kreit et al. (1989),
Eschenbacher et al. (1989), and Linn et al. (1994) suggest that mild to moderate asthmatics are
at least as sensitive to the acute effects of O. inhalation as healthy subjects when the asthmatics
are exposed to O. under conditions that elicit a significant response in healthy subjects. Kreit
et al. (1989) and Eschenbacher et al. (1989) exposed adult asthmatic and nonasthmatic subjects
to 0.4 ppm O. with intermittent moderate exercise for 2 h and observed a greater response in
R.., FEV., and FEF in the asthmatic subjects, although changes in FVC and symptoms
were similar in both groups. Ozone exposure also resulted in a marked increase in airway
responsiveness to methacholine in both the asthmatic and nonasthmatic subjects. These
responses take on greater importance when it is considered that the observed O.-induced
pulmonary effects were superimposed on preexisting impairment of pulmonary function and
airway responsiveness. In addition, the observations of Koenig et al. (1990) and Molfino et al.
(1991) suggest the possibility that acute exposure to O. at doses that do not produce measurable
pulmonary function decrements may increase the responsiveness of asthmatics to inhaled SO.
or antigens.
7.5.1.2 Symptom Responses to Ozone
7-183
-------
Following exposure to O. many subjects report respiratory symptoms, the most
common of which are cough, shortness of breath, and PDI. There is a broad range of severity
in rating symptom responses among subjects in these studies. A number of 2-h O. exposure
studies that have examined the exposure dose-pulmonary function and symptom response
relationships have included a semiquantitative analysis of symptom responses (Avol et al.,
1983; Kulle et al., 1985; McDonnell et al., 1983; Seal et al., 1993). In each of these studies,
as O. concentration increased, the pulmonary function response became more negative (a
decrease in FEV.), and the level of the respiratory symptoms (cough, shortness of breath, PDI)
increased. The mean decrement in FEV. in each of these studies was highly correlated (r >
0.98 in all cases) with the mean change in symptom rating or symptom score (the
determination of symptoms varied between studies). This high correlation results primarily
because each of these variables is highly correlated with O. concentration. Correlation of
individual changes in symptoms and changes in pulmonary function seldom exceed r = 0.6
(Horstman et al., 1990). Contributing to this low individual correlation is the fact that
symptoms scores have lower test-retest reliability than tests of lung function. For example,
McDonnell et al. (1985b) report test-retest coefficients of about 0.9 for spirometry responses
but only about 0.8 for symptoms. Thus individual symptom responses are not good predictors
of individual pulmonary function responses. However, group mean symptom responses still
provide a good marker of the average FEV. response to O. exposure. In two of the very heavy
exercise studies (McDonnell et al., 1983; Avol et al., 1983) symptoms of cough or total
respiratory symptom scores were increased significantly at 0.12 and 0.16 ppm O.,
respectively. In the heavy exercise study (Seal et al., 1993), cough symptoms increased
significantly at 0.18 ppm O.. Other studies that support this relationship of symptoms and
pulmonary function have been conducted with various exposure durations and exercise
intensities (Gong et al., 1986; Horstman et al., 1990) ranging from 1 h of severe exercise to
6.6 h of moderate exercise at O. concentrations from 0.08 to 0.20 ppm.
In comparing the spirometry and symptom responses of older adults and young
adults exposed to O. under the same conditions, Drechsler-Parks et al. (1989) found
significantly lower spirometry responses (Dl9% versus D6% FEV.) in the older adults.
In addition, the incidence of respiratory symptom responses for the three symptoms most
commonly reported with O. exposure were almost twice as high in the young adults, whereas
the incidence of symptoms unrelated to O. exposure (e.g., eye irritation, muscle soreness)
typically was greater in the older adults. The comparable or greater incidence of
nonrespiratory symptoms in the older adults clearly indicates that they felt less respiratory
discomfort in conjunction with their smaller spirometry responses. Asthmatics, when
compared with nonasthmatics, tend to have greater changes in R.. and expiratory flow with
O. exposure (Kreit et al., 1989; Horstman et al., 1995) but similar changes in lung volume
(i.e., FVC). Asthmatics also have similar symptom responses for cough, PDI, and shortness
of breath, although, in one study (Horstman et al., 1995), asthmatics reported a higher
incidence of wheezing.
In repeated exposure studies (Folinsbee et al., 1994; Linn et al., 1982b), the
changes in symptoms track the changes in spirometry. With repeated exposure to high
O. concentrations, the change in FEV. is typically greatest on the second exposure day.
Correspondingly, symptoms are increased on the second exposure day and diminish to near
baseline levels by the fourth or fifth exposure when the spirometry responses become
negligible. With repeated exposure to lower O. concentrations, the largest spirometry
7-184
-------
response is seen on the first day and attenuates by the third or fourth day. Symptom responses
also are largest on the first day and are attenuated with the same time course. In a single 2-h
study in which symptom and spirometry responses were measured during exposure and
recovery, the mean changes in symptoms and spirometry responses followed similar time
courses (McDonnell et al., 1987).
Intervention studies examining the effects of various drug treatments on the
responses to O. also report parallel changes in symptoms and spirometry. Although atropine
blocked the increase in R.. in response to O. exposure, it did not alter the spirometry or
symptom responses (Beckett et al., 1985). Similarly, albuterol and salbutamol, which had no
effect on O.-induced changes in spirometry, also had no effect of symptom responses
(McKenzie et al., 1987; Gong et al., 1988). The anti-inflammatory medications indomethacin
and ibuprofen, which partially inhibit the spirometry responses to O. exposure, also cause a
reduction in respiratory symptoms (Schelegle et al., 1987; Hazucha et al., 1994).
The individual correlations between symptoms and spirometry responses are
relatively low (< 0.6) and are of little predictive value. However, the group mean responses
have similar exposure-response characteristics and follow a similar time course of response to
exposure and recovery. Symptom and spirometry responses also follow a similar time course
of attenuation to repeated exposure and they are affected similarly by a number of medication
interventions.
7.5.1.3 Effects on Exercise Performance
Endurance exercise performance and VO may be limited by acute exposure to
O. (Adams and Schelegle, 1983; Schelegle and Adams, 1986; Gong et al., 1986; Foxcroft and
Adams, 1986; Folinsbee et al., 1977; Linder et al., 1988). Gong et al. (1986) and Schelegle
and Adams (1986) found that significant reductions in maximal endurance exercise
performance may occur in well-conditioned athletes while they perform CE (V. > 80 L/min)
for 1 h at O. concentrations DO. 18 ppm. Data from Linder et al. (1988) suggest that small
decrements in maximal exercise performance may occur at O. concentrations less than
0.18 ppm. The mechanisms that lead to these responses and the minimum O. concentration at
which these effects occur have not yet been defined clearly. Reports from studies of exposure
to O. during high-intensity exercise indicate that breathing discomfort associated with maximal
ventilation may be an important factor in limiting exercise performance. However, these
studies do not exclude the possibility that some as yet undefined physiological mechanism may
limit exercise performance.
7.5.1.4 Effects on Airway Responsiveness
Ozone exposure causes an increase in nonspecific airway responsiveness as
indicated by a reduction in the concentration of methacholine or histamine required to produce
a given reduction in FEV. or increase in SR... Increased airway responsiveness is an
important consequence of exposure to O. because its presence means that the airways are
predisposed to narrowing on inhalation of a variety of stimuli (e.g., specific allergens, SO.,
cold air). Markedly increased airway responsiveness is a classical feature of asthma and also
may be present with other respiratory diseases (e.g., chronic bronchitis, acute viral infections)
and even in a sizeable percentage of the healthy asymptomatic population. Many studies have
demonstrated O.-induced increases in nonspecific airway responsiveness in healthy subjects
after a 1- to 2-h exposure with exercise to concentrations in the range of 0.20 to 0.60 ppm
7-185
-------
(Golden et al., 1978; Holtzman et al., 1979; Konig et al., 1980; Dimeo et al., 1981; Gong
et al., 1986; Folinsbee and Hazucha, 1989) and after 6.6 h of exposure to concentrations in the
range of 0.08 to 0.12 ppm (Folinsbee et al., 1988; Horstman et al., 1990). Ozone-induced
increases in airway responsiveness tend to resolve within 24 h after exposure but may persist in
selected individuals for longer periods (Golden et al., 1978).
Ozone exposure of asthmatic subjects, who characteristically have increased airway
responsiveness at baseline, can cause further increases in responsiveness (Kreit et al., 1989).
The difference in baseline airway responsiveness between healthy and mild asthmatic subjects
may be as much as 100-fold, whereas the changes in airway responsiveness induced by O. are
typically two- to fourfold. Similar relative changes in airway responsiveness are seen in
asthmatics exposed to O. despite their markedly different baseline airway responsiveness. One
study (Molfino et al., 1991) has been published suggesting an increase in specific (i.e.,
allergen-induced) airway reactivity. This response was observed after a 1-h resting exposure
of atopic asthmatics to 0.12 ppm O.. One of the important aspects of this observation of
increased airway responsiveness after O. exposure is that this represents a plausible link
between ambient O. exposure and increased hospital admissions for asthma. However,
experimental design flaws preclude the use of this study in the determination of a lowest-
observed-effect level.
Changes in airway responsiveness after O. exposure appear to be resolved more
slowly than changes in FEV. or respiratory symptoms. Furthermore, in studies of repeated
exposure to O., changes in airway responsiveness tend to be somewhat less susceptible to
attenuation with consecutive exposures than changes in FEV. (Dimeo et al., 1981; Kulle et al.,
1982; Folinsbee et al., 1994). The question of whether chronic O. exposure can induce a
persistent increase (or decrease) in airways responsiveness has not been studied adequately.
Increases in airway responsiveness do not appear to be strongly associated with
decrements in lung function or increases in symptoms. This conclusion is based on studies in
healthy subjects; however, asthmatics who have widely different baseline airway
responsiveness exhibit FEV. changes after O. exposure that are similar to those seen in healthy
subjects (Kreit et al., 1989).
The mechanism of O.-induced increases in airway responsiveness is only partially
understood, but it appears to be associated with a number of cellular and biochemical changes
in airway tissue. Airway inflammation may be temporally associated with the presence of
increased airway responsiveness (Holtzman et al., 1983; O'Byrne et al., 1984; Seltzer et al.,
1986), but many animal models of induced neutrophilia report a conflicting role of these cells
in eliciting nonspecific bronchial hyperresponsiveness. Several animal species, for example,
have shown an increased airway responsiveness induced by O. exposure in the absence of an
influx of PMNs into the airway mucosa (Evans et al., 1988; Okazawa et al., 1989; Li et al.,
1992). In one human study (Ying et al., 1990), preexposure treatment with the anti-
inflammatory drug indomethacin blocked the effect of O. on FEV. and FVC but not on airway
responsiveness; however, cyclooxygenase inhibitors have not been effective at blocking the O.-
induced influx of PMNs into BAL fluid (Hazucha et al., 1996; Kleeberger and Hudak, 1992).
Therefore, O.-induced airway responsiveness may not be due to the presence of PMNs in the
airway or to the release of arachidonic acid metabolites. Rather, it seems likely that the
mechanism for this response is multifactorial, possibly involving the presence of cytokines,
prostanoids, or neuropeptides; activation of macrophages, eosinophils, or mast cells; and
7-186
-------
epithelial damage that increases direct access of mediators to the smooth muscle or receptors in
the airways that are responsible for reflex bronchoconstriction.
7.5.1.5 Inflammation and Host Defense Effects
A number of studies clearly show that a single acute exposure (1 to 4 h) of humans
to moderate concentrations of O. (0.2 to 0.6 ppm) while exercising at moderate to heavy levels
results in a number of cellular and biochemical changes in the lung, as assessed by
measurement of BAL constituents (Seltzer et al., 1986; Kehrl et al., 1987; Keren et al.,
1989a,b, 1991; Schelegle et al., 1991; McGee et al., 1990; Aris et al., 1993a; Devlin et al.,
1995). These exposures result in an inflammatory response characterized by increased
numbers of PMNs, increased permeability of the epithelial cells lining the respiratory tract,
cell damage, and production of proinflammatory cytokines and prostaglandins. This response
can be detected as early as 1 h after exposure (Keren et al., 1991; Schelegle et al., 1991) and
persists for at least 18 h (Keren et al., 1989a; Aris et al., 1993a). The response profile of
these mediators is not defined adequately, although it is clear that the time course of response
varies for different mediators and cells (Schelegle et al., 1991, Keren et al., 1989a, 1991).
A single study (Devlin et al., 1991) provides evidence that many of these changes
also occur in humans exposed to 0.08 and 0.10 ppm O. with moderate exercise for 6.6 h.
Decrements in the ability of AMs to phagocytose microorganisms also were reported in this
study.
Ozone also causes inflammatory changes in the nose, as indicated by increased
levels of PMNs and albumin, a marker for increased epithelial cell permeability. Increases in
tryptase levels immediately after O. exposure suggested the release of mast cell products.
There appears to be no strong correlation between any of the measured cellular and
biochemical changes and changes in lung function measurements, suggesting that different
mechanisms may be responsible for these processes. Alternatively, the absence of a
correlation may reflect either the temporal misalignment of these measurements, the fact that
changes detected in the lavage fluid do not quantitatively reflect events occurring in tissues
where functional or symptomatic events originate, or that BAL fluid may not be collected from
the same lung region primarily implicated in pulmonary function responses. The idea of
different mechanisms is supported by a study in which ibuprofen, a cyclooxygenase inhibitor,
blunted the O.-induced decrements in lung function without altering the O.-induced increase in
PMNs or epithelial cell permeability, although ibuprofen did change the concentration of a
number of mediators, some of which may be related to changes in function (Hazucha et al.,
1994).
In vitro studies suggest that epithelial cells are the primary target of O. in the lung
and that O. induces them to produce many of the mediators found in the BAL fluid of humans
exposed to O.. Although O. does not induce AMs to produce these compounds in large
quantities, it does directly impair the ability of AMs to phagocytose and kill microorganisms.
7.5.1.6 Factors Modifying Responsiveness to Ozone
Many variables that at least have potential for influencing response to O. remain
inadequately addressed in the available clinical data. Factors such as smoking status, age,
gender, race or ethnic group, season, and mode of breathing during exposure have been
evaluated inadequately for their potential influence on responses to O. exposure.
7-187
-------
Information derived from O. exposure of smokers is limited. Some degree of
attenuation appears to occur in active smokers and may be reversed following smoking
cessation (Emmons and Foster, 1991), but available results should be interpreted with caution.
The possibility of age-related differences in response to O. has been explored to some extent
since the publication of the previous O. criteria document (U.S. Environmental Protection
Agency, 1986). Young adults historically have provided the subject population for air
pollutant exposure studies. Pulmonary function responsiveness appears to decrease with age,
although symptom rates remain similar to those of young adults (Drechsler-Parks et al., 1987b,
1989, 1990; Bedi et al., 1988; Reisenauer et al., 1988; McDonnell et al., 1993). The limited
information available on the responses of children and adolescents to O. (McDonnell et al.,
1985a; Avol et al., 1985a, 1987; Koenig et al., 1987, 1988) does not indicate that children and
adolescents are either more or less responsive than young adults. Of the studies that have
investigated gender differences in responsiveness to O., some (Lauritzen and Adams, 1985;
Horvath et al., 1986; Adams et al., 1987; Drechsler-Parks et al., 1987a,b; Messineo and
Adams, 1990) have suggested that women are more responsive to O. than men. However, the
absence of consistent findings with respect to gender differences indicates that it cannot be
concluded that men and women respond differently to O.. Comparison of responses across
gender, racial, ethnic, and age groups is complicated by the determination of equivalent
exposures. For example, women and children have smaller lungs than adult men. Thus, with
a given exposure concentration, duration, and ventilation, humans with smaller lungs will
presumably receive a large relative intrapulmonary exposure. Some attempts have been made
to normalize responses according to BSA or lung capacity (e.g., FVC). The only study in
which this factor has been investigated systematically (Messineo and Adams, 1990) found no
influence of lung size on the spirometry responses under identical exposure (O. concentration,
V., and T) conditions. Three studies (Fox et al., 1993; Seal et al., 1995; Weinmann et al.
(1995) have compared pulmonary function responses of women during different phases of the
menstrual cycle, but the results are conflicting. The responses of black and white young adults
to various concentrations of O. have been compared in one study (Seal et al., 1993). The data
suggested that black males experienced significant decrements in pulmonary function at a lower
concentration of O. than white males, but that there were no differences among the responses
of white males and black and white females. Thus, the question of ethnic or racial differences
in responsiveness to O. is answered inadequately, and the available results should be
interpreted with caution. No new studies are available on the effects of heat stress (i.e.,
increased temperature or RH) on O. responses. One study (Linn et al., 1988) suggests that
sensitivity to O. may be related to seasonal variations in ambient O. concentrations; this
finding needs to be confirmed. Two studies (Hynes et al., 1988; Adams et al., 1989) have
reported that differences in the inhalation route (e.g., oral versus nasal or oronasal) appear to
be of negligible importance in the responses of exercising adults to O. exposure. Studies of
O. uptake in the upper airway (Gerrity et al., 1988) confirm the negligible differences between
oral and nasal inhalation (also see Chapter 8). None of these potential influences on
O. responsiveness (age, gender, race, hormonal fluctuations, smoking, seasonal variations in
responsiveness, and ambient environmental factors) has been investigated thoroughly.
However, the observation that healthy older adults appear to be less responsive to O. exposure
than young adults has been confirmed to the point that it can be considered in risk assessment.
Nevertheless, this does not address fully the question of age differences because children and
adolescents remain inadequately studied.
7-188
-------
7.5.1.7 Extrapulmonary Effects of Ozone
It still is believed that O. reacts immediately on contact with respiratory systems
fluids and tissues and is not absorbed or transported to extrapulmonary sites to any significant
degree. A number of laboratory animal studies reported in the previous chapter (Chapter 6)
and early studies on human subjects reported in this chapter suggest that reaction products
formed by the interaction of O. with respiratory system fluids or tissues may produce effects
measured outside the respiratory tract—either in the blood, as changes in circulating blood
lymphocytes, erythrocytes, or serum, or as changes in the structure or function of other
organs, such as the parathyroid, the heart, the liver, and the central nervous system.
No extrapulmonary effects have been reported to date in other organ systems of O. -exposed
human subjects, except for limited data indicating that acute (1- to 2-h) exposures with exercise
at concentrations DO.35 ppm O. caused transient changes in blood cells and plasma. The
interpretation of all these effects in regards to potential human health effects at ambient levels
of exposure (< 0.35 ppm O.) is not clear. However, the demonstration in this chapter of an
array of inflammatory mediators and immune modulators released at the airway surface in
response to O. provides a possible mechanism for effects to occur outside of the lung.
Additional studies are needed, therefore, in order to determine if there are any significant
extrapulmonary effects of O. exposure and at what levels of exposure they might occur.
7.5.1.8 Effects of Ozone Mixed with Other Pollutants
No significant enhancement of respiratory effects (i.e., more than additive) has been
consistently demonstrated for mixtures of O. with SO., NO., H. SO., HNO., or particulate
aerosols, or with multiple combinations of these pollutants. There is general agreement among
studies of simultaneous exposure of healthy adults and asthmatic adolescents to mixtures of
O. and NO., SO., H.SO., or HNO. that pulmonary function responses are not significantly
different from those following exposure to O. alone when compared to studies conducted at the
same O. concentration. Exposure to high PAN concentrations (i.e., 0.3 ppm) combined with
O. has been reported to induce greater pulmonary function responses than exposure to O. alone
(Horvath et al., 1986), but when the PAN concentration is reduced to the ambient range, any
additional effect of PAN in the mixture appears to be negligible (Drechsler-Parks et al., 1989).
In addition to simultaneous exposures to pollutant mixtures, studies of the responses
to O. exposure either preceded or followed by another pollutant have been performed. To the
extent that these exposure sequences mimic real ambient conditions, the results could be useful
in the risk assessment process. Koenig et al. (1990) demonstrated that exposure of allergic
(and probably asthmatic) adolescents to O. and then to SO. resulted in significant pulmonary
function decrements not seen with an O.-O. sequence or FA-SO. sequence. These results also
can be interpreted in light of the fact that O. increases nonspecific bronchial responsiveness
and that the increased SO. responses may simply reflect this increased responsiveness. Such
responses would be unlikely in nonatopic healthy adolescents. Other studies (Aris et al., 1991;
Hazucha et al., 1994; Linn et al., 1994; Utell et al., 1994) have assessed the responses to
O. after previous exposure to another pollutant. Aris et al. (1991) found that preexposure to
water or HNO. fog appeared to attenuate responses to O., whereas Hazucha et al. (1994)
observed an increased airway responsiveness after O. exposure preceded by NO. exposure
relative to O. exposure alone. Two studies of combined or sequential exposure to H.SO.
aerosol and O. suggest a possibly enhanced response to O. in asthmatics when the exposure is
7-189
-------
combined with or preceded by exposure to H.SO. aerosol (Linn et al., 1994; Utell et al.,
1994). These findings are intriguing, but must be replicated before they can be useful for
quantitative health assessment. Much is unknown about responses to air pollutant mixtures.
Only a limited number of pollutant combinations and exposure protocols have been
investigated, and subject groups are small and may not be representative of the general
population. Few studies have included more than two pollutants, and most combinations have
been evaluated in single studies. Furthermore, only rarely are endpoints other than pulmonary
function and plethysmography measured.
7.5.2 Field and Epidemiology Studies of Ozone Exposure
Individual-level camp and exercise studies provide useful, quantitative information
on the exposure-response relationships linking human lung function declines with O. exposure
occurring in ambient air. Their utility derives largely from the reliability with which
individual exposures can be estimated using outdoor measurements in studies of these kind.
Although it usually has not been possible to isolate O. exposures from other copollutants (e.g.,
acid aerosols) and environmental factors (e.g., temperature) in the design of such studies, the
available body of evidence now strongly supports a dominant role of O. in the observed lung
function decrements.
The most extensive epidemiologic database on pulmonary function responses to
ambient O. comes from camp studies. Six recent key studies from three separate research
groups provide a combined database on individual exposure-response relationships for
616 children ranging in age from 7 to 17 years, each with at least six sequential measurements
of FEV. and previous-hour O. exposures while attending summer camps (Avol et al., 1990;
Higgins et al., 1990; Raizenne et al., 1987, 1989; Spektor et al., 1988a, 1991). When
analyzed together using consistent methods, these data yielded an average relationship between
FEV. and previous-hour O. concentration of DO.50 mL/ppb. The highest 1-h O. levels
measured in five of the six studies ranged from 100 to 160 ppb, with one study reporting
concentrations as high as 245 ppb. Minimum O. values ranged from 10 to 60 ppb. Although
the regression results noted above were based on 1-h O. levels, exposure in camp studies
usually extended for multiple hours. Because of the high level of correlation between single-
and multiple-hour averages in the studies, these results may therefore, represent, to some
extent, the influence of multihour exposures. In addition to the camp study results, two key
studies involving lung function measurements before and after well-defined exercise events in
adults have yielded exposure-response slopes of DO.40 and Dl.35 mL/ppb (Spektor et al.,
1988b; Selwyn et al., 1985). Ozone concentrations during exercise events of approximately
0.5 h duration ranged from 4 to 135 ppb in these studies. Consistent with chamber studies,
there is no clear evidence from individual-level studies for a response threshold for the average
population effects of O. on pulmonary function decline. However, as with chamber studies,
there is evidence that responsivity varies across individuals. Thus, pulmonary function decline
as a function of ambient O. exposure for an individual may be either greater than or less than
the mean responses noted above.
Recent results of daily-life studies also support a consistent relationship between
ambient O./oxidant exposure and acute respiratory morbidity in the population. Respiratory
symptoms (or exacerbation of asthma) and decrements in PEFR occur with increasing ambient
O., especially in asthmatic children (Lebowitz et al., 1991; Krzyzanowski et al., 1992;
Thurston et al., 1995). Concurrent temperature, particles, H' , aeroallergens, and asthma
7-190
-------
severity or medication status also may contribute as independent or modifying factors. The
aggregate results show greater responses in asthmatic individuals than in nonasthmatics
(Lebowitz et al., 1991; Krzyzanowski et al., 1992), indicating that asthmatics constitute a
sensitive group in epidemiologic studies of oxidant air pollution.
Recent aggregate population time series studies of O.-related health effects provide
relevant evidence of acute responses, even below a 1-h maximum of 0.12 ppm O.. Emergency
room visits, hospital admissions, and mortality all have been examined as possible outcomes of
exposure to O.. In the case of ER visits, the evidence is limited (e.g., Bates et al., 1990; Cody
et al., 1992; White et al., 1994; Weisel et al., 1995), but results generally are consistent with
an effect of O. on morbidity. Mortality studies vis-f-vis O. also are rather limited, but are
more mixed in their results. One of two new, well-designed studies indicate a significant
association between O. and total mortality in Los Angeles, CA, even after controlling for the
potentially confounding effects of temperature and PM (Kinney and Ozkaynak, 1991). Los
Angeles experienced peak 1-h maximum O. concentrations above 0.2 ppm during this study
period. However, at lower concentrations, over a shorter time span, and with different
statistical methods, a second study (Dockery et al., 1992) did not detect a significant
O. association with mortality. The strongest and most consistent evidence of O. effects, both
above and below 0.12 ppm O., then, is provided by the multiple studies that have been
conducted over the last decade on summertime daily hospital admission for respiratory causes
in various locales in eastern North America (Bates and Sizto, 1983, 1987, 1989; Thurston
et al., 1992, 1994; Lipfert and Hammerstrom, 1992; Burnett et al., 1994). These studies
consistently have shown that O. air pollution is associated with an increased incidence of
admissions, accounting for roughly one to three excess respiratory hospital admissions per 100
ppb O. per million persons. This association has been shown to remain even after statistically
controlling for the possible confounding effects of temperature and copollutants (e.g., H' ,
SO., and PM..), as well as when considering only days having 1-h maximum O. concentrations
below 0.12 ppm. Furthermore, these results imply that O. air pollution can account for a
substantial portion of summertime hospital admissions for respiratory causes on the most
polluted days. Overall, the aggregate population time series studies considered in this chapter
provide strong evidence that ambient exposures to O. can cause significant exacerbations of
preexisting respiratory disease in the general public at concentrations below 0.12 ppm O..
Studies of chronic health effects that may relate to long-term exposure to ambient
pollutants still have not provided enough data to determine if there are respiratory or other
health effects that result directly from chronic O. exposure. However, the aggregate evidence
to date suggests that chronic O. exposure, along with other environmental factors, could be
responsible for health effects in exposed populations.
The most useful set of data has been provided by the AHSMOG studies (Hodgkin
et al., 1984; Euler et al., 1987, 1988; Abbey et al., 1991a,b, 1993). These studies have
provided the most refined measures of chronic exposure to date (including adjustment for
quality of the monitoring data as determined by distance of monitoring sites from subject
residences, topography, time spent indoors, and time spent at work). The most consistent
effects that can be attributable, in part, to O. relate to an increase in 10-year cumulative
incidence of asthma (RR = 2.07 for each 1,000 h above 10 pphm) and an increase in asthma
severity. Unfortunately, for the entire set of studies, the colinearity between O. and TSP
reduces the confidence that effects can be attributed to O. alone, O. in combination with the
particulate fraction of ambient pollution, or the combination of the two. Some support for an
7-191
-------
effect on persons with asthma also can be derived from a recent Canadian study (Stern et al.,
1994) that demonstrated nonstatistically significant 6.6 and 6.5% reductions in FEF and
V for people living in Ontario relative to those in Saskatchewan. Again, however, the
effects of O. are impossible to disentangle from the other contributors such as the acid summer
haze that characterizes the United States east of the Mississippi River.
7-192
-------
References
Abbey, D. E.; Moore, J.; Petersen, P.; Beeson, L. (1991a) Estimating cumulative ambient concentrations of air
pollutants: description and precision of methods used for an epidemiological study. Arch. Environ.
Health 46: 281-287.
Abbey, D. E.; Mills, P. K.; Petersen, F. P.; Beeson, W. L. (1991b) Long-term ambient concentrations of total
suspended particulates and oxidants as related to incidence of chronic disease in California Seventh-
Day Adventists. Environ. Health Perspect. 94: 43-50.
Abbey, D. E.; Petersen, P.; Mills, P. K.; Beeson, W. L. (1993) Long-term ambient concentrations of total
suspended particulates, ozone, and sulfur dioxide and respiratory symptoms in a nonsmoking
population. Arch. Environ. Health 48: 33-46.
Adams, W. C. (1986) Physiological responses of healthy human subjects consequent to inhalation of N02, O3, and
NO2 plus O3 during heavy, sustained exercise. Sacramento, CA: California State Air Resources
Board; report no. ARB-R-87/294. Available from: NTIS, Springfield, VA; PB87-165106.
Adams, W. C.; Schelegle, E. S. (1983) Ozone and high ventilation effects on pulmonary function and endurance
performance. J. Appl. Physiol.: Respir. Environ. Exercise Physiol. 55: 805-812.
Adams, W. C.; Savin, W. M.; Christo, A. E. (1981) Detection of ozone toxicity during continuous exercise via
the effective dose concept. J. Appl. Physiol.: Respir. Environ. Exercise Physiol. 51: 415-422.
Adams, W. C.; Brookes, K. A.; Schelegle, E. S. (1987) Effects of NO, alone and in combination with Q on
young men and women. J. Appl. Physiol. 62: 1698-1704.
Adams, W. C.; Schelegle, E. S.; Shaffrath, J. D. (1989) Oral and oronasal breathing during continuous exercise
produce similar responses to ozone inhalation. Arch. Environ. Health 44: 311-316.
Aris, R.; Christian, D.; Sheppard, D.; Balmes, J. R. (1991) The effects of sequential exposure to acidic fog and
ozone on pulmonary function in exercising subjects. Am. Rev. Respir. Dis. 143: 85-91.
Aris, R. M.; Christian, D.; Hearne, P. Q.; Kerr, K.; Finkbeiner, W. E.; Balmes, J. R. (1993a) Ozone-induced
airway inflammation in human subjects as determined by airway lavage and biopsy. Am. Rev.
Respir. Dis. 148: 1363-1372.
Aris, R.; Christian, D.; Tager, L; Ngo, L; Finkbeiner, W. E.; Balmes, J. R. (1993b) Effects of nitric acid gas
alone or in combination with ozone on healthy volunteers. Am. Rev. Respir. Dis. 148: 965-973.
Astrand, P.-O.; Rodahl, K. (1977) Textbook of work physiology: physiological bases of exercise. 2nd ed. New
York, NY: McGraw-Hill Book Company; p. 465.
Avol, E. L.; Linn, W. S.; Shamoo, D. A.; Venet, T. G.; Hackney, J. D. (1983) Acute respiratory effects of Los
Angeles smog in continuously exercising adults. J. Air Pollut. Control Assoc. 33: 1055-1060.
Avol, E. L.; Linn, W. S.; Venet, T. G.; Shamoo, D. A.; Hackney, J. D. (1984) Comparative respiratory effects
of ozone and ambient oxidant pollution exposure during heavy exercise. J. Air Pollut. Control Assoc.
34: 804-809.
Avol, E. L.; Linn, W. S.; Shamoo, D. A.; Valencia, L. M.; Anzar, U. T.; Venet, T. G.; Hackney, J. D. (1985a)
Respiratory effects of photochemical oxidant air pollution in exercising adolescents. Am. Rev.
Respir. Dis. 132: 619-622.
7-193
-------
Avol, E. L.; Linn, W. S.; Shamoo, D. A.; Valencia, L. M.; Anzar, U. T.; Hackney, J. D. (1985b) Short-term
health effects of ambient air pollution in adolescents. In: Lee, S. D., ed. Evaluation of the scientific
basis for ozone/oxidants standards: transactions of an APCA international specialty conference;
November 1984; Houston, TX. Pittsburgh, PA: Air Pollution Control Association; pp. 329-336.
(APCA international specialty conference transactions: TR-4).
Avol, E. L.; Linn, W. S.; Venet, T. G.; Shamoo, D. A.; Spier, C. E.; Hackney, J. D. (1985c) Comparative
effects of laboratory generated ozone and ambient oxidant exposure in continuously exercising
subjects. In: Lee, S. D., ed. Evaluation of the scientific basis for ozone/oxidants standards:
proceedings of an APCA international specialty conference; November 1984; Houston, TX.
Pittsburgh, PA: Air Pollution Control Association; pp. 216-225. (APCA international specialty
conference transactions: TR-4).
Avol, E. L.; Linn, W. S.; Shamoo, D. A.; Spier, C. E.; Valencia, L. M.; Venet, T. G.; Trim, S. C.; Hackney,
J. D. (1987) Short-term respiratory effects of photochemical oxidant exposure in exercising children.
JAPCA 37: 158-162.
Avol, E. L.; Linn, W. S.; Shamoo, D. A.; Hackney, J. D. (1988) Seasonal ozone reactivity in Los Angeles
residents. Presented at: 81st annual meeting of the Air Pollution Control Association; June; Dallas,
TX. Pittsburgh, PA: Air Pollution Control Association; paper no. 88-122.6.
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, K. B.; D'Arcy, J. B.; Gibbons, D.; Higgins, I. T. T. (1990) Ozone exposure and lung
function in children attending a southern California summer camp. Presented at: 83rd annual meeting
and exhibition of the Air & Waste Management Association; June; Pittsburgh, PA. Pittsburgh, PA:
Air & Waste Management Association; paper no. 90-150.3.
Avol, E. L.; Trim, S. C.; Little, D. E.; Spier, C. E.; Smith, M. N.; Peng, R. C.; Linn, W. S.; Hackney, J. D.
(1991) Ozone exposure and lung function: a southern California summer camp study. In: Berglund,
R. L.; Lawson, D. R.; McKee, D. J., eds. Tropospheric ozone and the environment: papers from an
international conference; March 1990; Los Angeles, CA. Pittsburgh, PA: Air & Waste Management
Association; pp. 90-99. (A&WMA transaction series no. TR-19).
Basconi, R.; Naclerio, R. M.; Fitzgerald, T. K.; Kagey-Sobotka, A.; Proud, D. (1990) Effect of ozone inhalation
on the response to nasal challenge with antigen of allergic subjects. Am. Rev. Respir. Dis. 142: 594-
601.
Bates, D. V. (1985) The strength of the evidence relating air pollutants to adverse health effects. Chapel Hill, NC:
University of North Carolina at Chapel Hill, Institute for Environmental Studies. (Carolina
environmental essay series VI).
Bates, D. V.; Sizto, R. (1983) Relationship between air pollutant levels and hospital admissions in Southern
Ontario. Can. J. Public Health 74: 117-122.
Bates, D. V.; Sizto, R. (1987) Air pollution and hospital admissions in southern Ontario: the acid summer haze
effect. Environ. Res. 43: 317-331.
Bates, D. V.; Sizto, R. (1989) The Ontario Air Pollution study: identification of the causative agent. Environ.
Health Perspect. 79: 69-72.
Bates, D. V.; Bell, G. M.; Burnham, C. D.; Hazucha, M.; Mantha, J.; Pengelly, L. D.; Silverman, F. (1972)
Short-term effects of ozone on the lung. J. Appl. Physiol. 32: 176-181.
Bates, D. V.; Baker-Anderson, M.; Sizto, R. (1990) Asthma attack periodicity: a study of hospital emergency
visits in Vancouver. Environ. Res. 51: 51-70.
7-194
-------
Beck, N.; Koenig, J. Q.; Luchtel, D. L.; Altman, L. C.; Orsborn, M. T.; Kenney, J. S. (1994) Ozone can
increase the expression of intercellular adhesion molecule-1 and the snythesis of cytokines by human
nasal epithelial cells. Inhalation Toxicol. 6: 345-357.
Becker, S.; Madden, M. C.; Newman, S. L.; Devlin, R. B.; Koren, H. S. (1991) Modulation of human alveolar
macrophage properties by ozone exposure in vitro. Toxicol. Appl. Pharmacol. 110: 403-415.
Beckett, W. S.; McDonnell, W. P.; Horstman, D. H.; House, D. E. (1985) Role of the parasynipathetic nervous
system in acute lung response to ozone. J. Appl. Physiol. 59: 1879-1885.
Beckett, W. S.; Freed, A. N.; Turner, C.; Menkes, H. A. (1988) Prolonged increased responsiveness of canine
peripheral airways after exposure to O3. J. Appl. Physiol. 64: 605-610.
Bedi, J. P.; Drechsler-Parks, D. M.; Horvath, S. M. (1985) Duration of increased pulmonary function sensitivity
to an initial ozone exposure. Am. Ind. Hyg. Assoc. J. 46: 731-734.
Bedi, J. P.; Horvath, S. M.; Drechsler-Parks, D. M. (1988) Reproducibility of the pulmonary function response
of older men and women to a 2-hour ozone exposure. JAPCA 38: 1016-1019.
Bedi, J. P.; Horvath, S. M.; Drechsler-Parks, D. M. (1989) Adaptation by older individuals repeatedly exposed to
0.45 parts per million ozone for two hours. JAPCA 39: 194-199.
Berry, M.; Lioy, P. J.; Gelperin, K.; Buckler, G.; Klotz, J. (1991) Accumulated exposure to ozone and
measurement of health effects in children and counselors at two summer camps. Environ. Res.
54: 135-150.
Biersteker, K.; Evendijk, J. E. (1976) Ozone, temperature, and mortality in Rotterdam in the summers of 1974
and 1975. Environ. Res. 12: 214-217.
Brant, J. W. A. (1965) Human cardiovascular diseases and atmospheric air pollution in Los Angeles, California.
Int. J. Air Water Pollut. 9: 219-231.
Brant, J. W. A.; Hill, S. R. G. (1964) Human respiratory diseases and atmospheric air pollution in Los Angeles,
California. Int. J. Air Water Pollut. 8: 259-277.
Braun-Fahrlander, C.; Kiinzli, N.; Domenighetti, G.; Carell, C. P.; Ackermann-Liebrich, U. (1994) Acute
effects of ambient ozone on respiratory function of Swiss schoolchildren after a 10-minute heavy
exercise. Pediatr. Pulmonol. 17: 169-177.
Brookes, K. A.; Adams, W. C.; Schelegle, E. S. (1989) 0.35 ppm O3 exposure induces hyperresponsiveness on
24-h reexposure to 0.20 ppm O3. J. Appl. Physiol. 66: 2756-2762.
Brunekreef, B.; Hoek, G.; Breugelmans, O.; Leentvaar, M. (1994) Respiratory effects of low-level photochemical
air pollution in amateur cyclists. Am. J. Respir. Crit. Care Med. 150: 962-966.
Burnett, R. T.; Dales, R. E.; Raizenne, M. E.; Krewski, D.; Summers, P. W.; Roberts, G. R.; Raad-Young, M.;
Dann, T.; Brook, J. (1994) Effects of low ambient levels of ozone and sulfates on the frequency of
respiratory admissions to Ontario hospitals. Environ. Res. 65: 172-194.
Calderon-Garciduenas, L.; Osorno-Velaquez, A.; Bravo-Alvarez, H.; Delgado-Chavez, R.; Barrios-Marquez, R.
(1992) Histopathologic changes of the nasal mucosa in southwest metropolitan Mexico City
inhabitants. Am. J. Pathol. 140: 225-232.
California Department of Public Health. (1955) Clean air for California: initial report of the Air Pollution Study
Project. San Francisco, CA: State of California, Department of Public Health.
7-195
-------
California Department of Public Health. (1956) Clean air for California: second report of the Air Pollution Study
Project. Berkley, CA: State of California, Department of Public Health.
California Department of Public Health. (1957) Report III: a progress report of California's fight against air
pollution. Berkley, CA: State of California, Department of Public Health.
Castillejos, M.; Gold, D. R.; Dockery, D.; Tosteson, T.; Baum, T.; Speizer, F. E. (1992) Effects of ambient
ozone on respiratory function and symptoms in Mexico City schoolchildren. Am. Rev. Respir. Dis.
145: 276-282.
Castleman, W. L.; Tyler, W. S.; Dungworth, D. L. (1977) Lesions in respiratory bronchioles and conducting
airways of monkeys exposed to ambient levels of ozone. Exp. Mol. Pathol. 26: 384-400.
Chatham, M.; Bleecker, E. R.; Smith, P. L.; Rosenthal, R. R.; Mason, P.; Norman, P. S. (1982) A comparison
of histamine, methacholine, and exercise airway reactivity in normal and asthmatic subjects. Am.
Rev. Respir. Dis. 126: 235-240.
Cockcroft, D. W.; Killian, D. N.; Mellon, J. J. A.; Hargreave, F. E. (1977) Bronchial reactivity to inhaled
histamine: a method and clinical survey. Clin. Allergy 7: 235-243.
Cody, R. P.; Weisel, C. P.; Birnbaum, G.; Lioy, P. J. (1992) The effect of ozone associated with summertime
photochemical smog on the frequency of asthma visits to hospital emergency departments. Environ.
Res. 58: 184-194.
Coleridge, J. C. G.; Coleridge, H. M. (1986) Reflexes evoked from tracheobronchial tree and lungs.
In: Cherniack, N. S.; Widdicombe, J. G., eds. Handbook of physiology: v. 2, the respiratory
system. Washington, DC: American Physiological Society; pp. 395-429.
Coleridge, H. M.; Coleridge, J. C. G.; Ginzel, K. H.; Baker, D. G.; Banzett, R. B.; Morrison, M. A. (1976)
Stimulation of "irritant' receptors and afferent C-fibers in the lungs by prostaglandins. Nature
(London) 264: 451-453.
Coleridge, H. M.; Coleridge, J. C. G.; Baker, D. G.; Ginzel, K. H.; Morrison, M. A. (1978) Comparison of the
effects of histamine and prostaglandin on afferent C-fiber endings and irritant receptors in the
intrapulmonary airways. Adv. Exp. Med. Biol. 99: 291-305.
Coleridge, J. C. G.; Coleridge, H. M.; Schelegle, E. S.; Green, J. F. (1993) Acute inhalation of ozone stimulates
bronchial C-fibers and rapidly adapting receptors in dogs. J. Appl. Physiol. 74: 2345-2352.
Colucci, A. V. (1983) Pulmonary dose/effect relationships in ozone exposure. In: Lee, S. D.; Mustafa, M. G.;
Mehlman, M. A., eds. International symposium on the biomedical effects of ozone and related
photochemical oxidants; March 1982; Pinehurst, NC. Princeton, NJ: Princeton Scientific Publishers,
Inc.; pp. 21-44. (Advances in modern environmental toxicology: v. 5).
Delfino, R. J.; Becklake, M. R.; Hanley, J. A. (1993) Reliability of hospital data for population-based studies of
air pollution. Arch. Environ. Health 48: 140-146.
Delfino, R. J.; Becklake, M. R.; Hanley, J. A. (1994a) The relationship of urgent hospital admissions for
respiratory illnesses to photochemical air pollution levels in Montreal. Environ. Res. 67: 1-19.
Delfino, R. J.; Becklake, M. R.; Hanley, J. A. (1994b) Estimation of unmeasured particulate air pollution data for
an epidemiological study of daily respiratory morbidity. Environ. Res. 67: 20-38.
DeLucia, A. J.; Adams, W. C. (1977) Effects of O3 inhalation during exercise on pulmonary function and blood
biochemistry. J. Appl. Physiol.: Respir. Environ. Exercise Physiol. 43: 75-81.
7-196
-------
DeLucia, A. J.; Whitaker, J. H.; Bryant, L. R. (1983) Effects of combined exposure to ozone and carbon
monoxide (CO) in humans. In: Lee, S. D.; Mustafa, M. G.; Mehlman, M. A., eds. International
symposium on the biomedical effects of ozone and related photochemical oxidants; March; Pinehurst,
NC. Princeton, NJ: Princeton Scientific Publishers, Inc.; pp. 145-159. (Advances in modern
environmental toxicology: v. 5).
Betels, R.; Rokaw, S. N.; Coulson, A. H.; Tashkin, D. P.; Sayre, J. W.; Massey, F. J., Jr. (1979) The UCLA
population studies of chronic obstructive respiratory disease. I. Methodology and comparison of lung
function in areas of high and low pollution. Am. J. Epidemiol. 109: 33-58.
Betels, R.; Sayre, J. W.; Coulson, A. H.; Rokaw, S. N.; Massey, F. J., Jr.; Tashkin, D. P.; Wu, M.-M. (1981)
The UCLA population studies of chronic obstructive respiratory disease: IV. respiratory effect of
long-term exposure to photochemical oxidants, nitrogen dioxide, and sulfates on current and never
smokers. Am. Rev. Respir. Dis. 124: 673-680.
Betels, R.; Tashkin, D. P.; Sayre, J. W.; Rokaw, S. N.; Coulson, A. H.; Massey, F. J., Jr.; Wegman, D. H.
(1987) The UCLA population studies of chronic obstructive respiratory disease: 9. lung function
changes associated with chronic exposure to photochemical oxidants; a cohort study among never-
smokers. Chest 92: 594-603.
Betels, R.; Tashkin, D. P.; Sayre, J. W.; Rokaw, S. N.; Massey, F. J., Jr.; Coulson, A. H.; Wegman, D. H.
(1991) The UCLA population studies of CORD: X. a cohort study of changes in respiratory function
associated with chronic exposure to SOX, NOX, and hydrocarbons. Am. J. Public Health 81: 350-359.
Devlin, R. B.; Koren, H. S. (1990) The use of quantitative two-dimensional gel electrophoresis to analyze changes
in alveolar macrophage proteins in humans exposed to ozone. Am. J. Respir. Cell Mol. Biol. 2: 281-
288.
Devlin, R. B.; McDonnell, W. P.; Koren, H. S.; Becker, S. (1990) Prolonged exposure of humans to 0.10 and
0.08 ppm ozone results in inflammation in the lung. Presented at: 83rd annual meeting of the Air &
Waste Management Association; June; Pittsburgh, PA. Pittsburgh, PA: Air & Waste Management
Association; paper no. 90-150.2.
Devlin, R. B.; McDonnell, W. P.; Mann, R.; Becker, S.; House, D. E.; Schreinemachers, D.; Koren, H. S.
(1991) Exposure of humans to ambient levels of ozone for 6.6 hours causes cellullar and biochemical
changes in the lung. Am. J. Respir. Cell Mol. Biol. 4: 72-81.
Devlin, R. B.; McKinnon, K. P.; Noah, T.; Becker, S.; Koren, H. S. (1994) Ozone-induced release of cytokines
and fibronectin by alveolar macrophages and airway epithelial cells. Am. J. Physiol. 266: L612-
L619.
7-197
-------
Devlin, R. B.; Folinsbee, L. J.; Biscardi, P.; Hatch, G.; Becker, S.; Madden, M.; Robbins, M. C.; Koren, H. S.
(1995) Repeated exposure of humans to ozone. II. Inflammation and cell damage. Research Triangle
Park, NC: U.S. Environmental Protection Agency, National Health and Environmental Effects
Research Laboratory, Clinical Research Branch; manuscript in preparation.
Dimeo, M. J.; Glenn, M. G.; Holtzman, M. J.; Sheller, J. R.; Nadel, J. A.; Boushey, H. A. (1981) Threshold
concentration of ozone causing an increase in bronchial reactivity in humans and adaptation with
repeated exposures. Am. Rev. Respir. Dis. 124: 245-248.
Dockery, D. W.; Schwartz, J.; Spengler, J. D. (1992) Air pollution and daily mortality: associations with
particulates and acid aerosols. Environ. Res. 59: 362-373.
Drechsler-Parks, D. M.; Bedi, J. P.; Horvath, S. M. (1984) Interaction of peroxyacetyl nitrate and ozone on
pulmonary functions. Am. Rev. Respir. Dis. 130: 1033-1037.
Drechsler-Parks, D. M.; Bedi, J. P.; Horvath, S. M. (1987a) Pulmonary function responses of older men and
women to ozone exposure. Exp. Gerontol. 22: 91-101.
Drechsler-Parks, D. M.; Bedi, J. P.; Horvath, S. M. (1987b) Pulmonary function desensitization on repeated
exposures to the combination of peroxyacetyl nitrate and ozone. JAPCA 37: 1199-1201.
Drechsler-Parks, D. M.; Bedi, J. P.; Horvath, S. M. (1989) Pulmonary function responses of young and older
adults to mixtures of O3, NO2 and PAN. Toxicol. Ind. Health 5: 505-517.
Drechsler-Parks, D. M.; Horvath, S. M.; Bedi, J. P. (1990) The "effective dose" concept in older adults exposed
to ozone. Exp. Gerontol. 25: 107-115.
Emmons, K.; Foster, W. M. (1991) Smoking cessation and acute airway response to ozone. Arch. Environ.
Health 46: 288-295.
Eschenbacher, W. L.; Ying, R. L.; Kreit, J. W.; Gross, K. B. (1989) Ozone-induced lung function changes in
normal and asthmatic subjects and the effect of indoniethacin. In: Schneider, T.; Lee, S. D.;
Wolters, G. J. R.; Grant, L. D., eds. Atmospheric ozone research and its policy implications:
proceedings of the 3rd US-Dutch international symposium; May 1988; Nijmegen, The Netherlands.
Amsterdam, The Netherlands: Elsevier Science Publishers; pp. 493-499. (Studies in environmental
science 35).
Eschenbacher, W. L.; Moore, T. B.; Lorenzen, T. J.; Weg, J. G.; Gross, K. B. (1992) Pulmonary responses of
asthmatic and normal subjects to different temperature and humidity conditions in an environmental
chamber. Lung 170: 51-62.
Euler, G. L.; Abbey, D. E.; Magie, A. R.; Hodgkin, J. E. (1987) Chronic obstructive pulmonary disease
symptom effects of long-term cumulative exposure to ambient levels of total suspended particulates
and sulfur dioxide in California Seventh-Day Adventist residents. Arch. Environ. Health
42: 213-222.
Euler, G. L.; Abbey, D. E.; Hodgkin, J. E.; Magie, A. R. (1988) Chronic obstructive pulmonary disease
symptom effects of long-term cumulative exposure to ambient levels of total oxidants and nitrogen
dioxide in California Seventh-day Adventist residents. Arch. Environ. Health 43: 279-285.
Evans, T. W.; Brokaw, J. J.; Chung, K. P.; Nadel, J. A.; McDonald, D. M. (1988) Ozone-induced bronchial
hyperresponsiveness in the rat is not accompanied by neutrophil influx or increased vascular
permeability in the trachea. Am. Rev. Respir. Dis. 138: 140-144.
7-198
-------
Fernandes, A. L. G.; Molfino, N. A.; McClean, P. A.; Silverman, P.; Tarlo, S.; Raizenne, M.; Slutsky, A. S.;
Zamel, N. (1994) The effect of pre-exposure to 0.12 ppm of ozone on exercise-induced asthma.
Chest: 106: 1077-1082.
Flesh, R. D.; Riha, M. L.; Miller, M. F. (1982) Effects of short-term intermittent air pollutants on incidence and
severity of acute respiratory disease: data collection and quality assurance (project summary).
Research Triangle Park, NC: U.S. Environmental Protection Agency, Health Effects Research
Laboratory; EPA report no. EPA/600/1-81-065. Available from: NTIS, Springfield, VA;
PB82-129479.
Folinsbee, L. J.; Hazucha, M. J. (1989) Persistence of ozone-induced changes in lung function and airway
responsiveness. In: Schneider, T.; Lee, S. D.; Wolters, G. J. R.; Grant, L. D., eds. Atmospheric
ozone research and its policy implications: proceedings of the 3rd US-Dutch international
symposium; May 1988; Nijmegen, The Netherlands. Amsterdam, The Netherlands: Elsevier Science
Publishers; pp. 483-492. (Studies in environmental science 35).
Folinsbee, L. J.; Horvath, S. M. (1986) Persistence of the acute effects of ozone exposure. Aviat. Space Environ.
Med. 57: 1136-1143.
Folinsbee, L. J.; Silverman, P.; Shephard, R. J. (1975) Exercise responses following ozone exposure. J. Appl.
Physiol. 38: 996-1001.
Folinsbee, L. J.; Silverman, P.; Shephard, R. J. (1977) Decrease of maximum work performance following ozone
exposure. J. Appl. Physiol.: Respir. Environ. Exercise Physiol. 42: 531-536.
Folinsbee, L. J.; Drinkwater, B. L.; Bedi, J. P.; Horvath, S. M. (1978) The influence of exercise on the
pulmonary function changes due to exposure to low concentrations of ozone. In: Folinsbee, L. J.;
Wagner, J. A.; Borgia, J. P.; Drinkwater, B. L.; Gliner, J. A.; Bedi, J. P., eds. Environmental
stress: individual human adaptations. New York, NY: Academic Press; pp. 125-145.
Folinsbee, L. J.; Bedi, J. P.; Horvath, S. M. (1980) Respiratory responses in humans repeatedly exposed to low
concentrations of ozone. Am. Rev. Respir. Dis. 121: 431-439.
Folinsbee, L. J.; Bedi, J. P.; Horvath, S. M. (1984) Pulmonary function changes after 1 h continuous heavy
exercise in 0.21 ppm ozone. J. Appl. Physiol.: Respir. Environ. Exercise Physiol. 57: 984-988.
Folinsbee, L. J.; Bedi, J. P.; Horvath, S. M. (1985) Pulmonary response to threshold levels of sulfur dioxide (1.0
ppm) and ozone (0.3 ppm). J. Appl. Physiol. 58: 1783-1787.
Folinsbee, L. J.; Horstman, D. H.; Vorona, R. D.; Prince, J. M.; Berry, M. (1986) Determinants of endurance
performance during ozone inhalation. Presented at: 30th International Union of Physiological
Sciences Congress; July; Vancouver, BC, Canada. Proc. Int. Union Physiol. Sci. 16: 176.
Folinsbee, L. J.; McDonnell, W. P.; Horstman, D. H. (1988) Pulmonary function and symptom responses after
6.6-hour exposure to 0.12 ppm ozone with moderate exercise. JAPCA 38: 28-35.
Folinsbee, L. J.; Horstman, D. H.; Kehrl, H. R.; McDonnell, W. P.; Gerrity, T. R.; Seal, E.; Larson, R.;
Hazucha, M. J.; Abdul-Salaam, S.; Faucette, B.; Ives, P. J. (1991) Effects of single and repeated
prolonged low-level ozone exposure in man. Presented at: annual meeting of the Society for
Occupational and Environmental Health; March; Washington, DC.
Folinsbee, L. J.; Horstman, D. H.; Kehrl, H. R.; Harder, S. D.; Abdul-Salaam, S.; Ives, P. J. (1994)
Respiratory responses to repeated prolonged exposure to 0.12 ppm ozone. Am. J. Respir. Crit. Care
Med. 149: 98-105.
7-199
-------
Folinsbee, L. J.; Devlin, R. B.; Robbins, M. K.; Biscardi, F. H.; Abdul-Salaam, S.; Koren, H. S. (1995)
Repeated exposure to high ozone concentration: I. pulmonary function and symptom responses. Am.
J. Respir. Crit. Care Med.: prepared for submission.
Foster, W. M.; Costa, D. L.; Langenback, E. G. (1987) Ozone exposure alters tracheobronchial mucociliary
function in humans. J. Appl. Physiol. 63: 996-1002.
Fox, S. D.; Adams, W. C.; Brookes, K. A.; Lasley, B. L. (1993) Enhanced response to ozone exposure during
the follicular phase of the menstrual cycle. Environ. Health Perspect. 101: 242-244.
Foxcroft, W. J.; Adams, W. C. (1986) Effects of ozone exposure on four consecutive days on work performance
andVO2max. J. Appl. Physiol. 61: 960-966.
Galbo, H. (1983) Hormonal and metabolic adaptation to exercise. New York, NY: Thieme and Straton.
Gerrity, T. R.; Weaver, R. A.; Berntsen, J.; House, D. E.; O'Neil, J. J. (1988) Extrathoracic and intrathoracic
removal of O3 in tidal-breathing humans. J. Appl. Physiol. 65: 393-400.
Gerrity, T. R.; Bennett, W. D.; Kehrl, H.; DeWitt, P. J. (1993) Mucociliary clearance of inhaled particles
measured at 2 h after ozone exposure in humans. J. Appl. Physiol. 74: 2984-2989.
Gerrity, T. R.; McDonnell, W. P.; House, D. E. (1994) The relationship between delivered ozone dose and
functional responses in humans. Toxicol. Appl. Pharmacol. 124: 275-283.
Gertner, A.; Bromberger-Barnea, B.; Traystman, R.; Berzon, D.; Menkes, H. (1983a) Responses of the lung
periphery to ozone and histamine. J. Appl. Physiol.: Respir. Environ. Exercise Physiol.
54: 640-646.
Gertner, A.; Bromberger-Barnea, B.; Dannenberg, A. M., Jr.; Traystman, R.; Menkes, H. (1983b) Responses of
the lung periphery to 1.0 ppni ozone. J. Appl. Physiol.: Respir. Environ. Exercise Physiol.
55: 770-776.
Gertner, A.; Bromberger-Barnea, B.; Traystman, R.; Menkes, H. (1983c) Effects of ozone on peripheral lung
reactivity. J. Appl. Physiol.: Respir. Environ. Exercise Physiol. 55: 777-784.
Gibbons, S. I.; Adams, W. C. (1984) Combined effects of ozone exposure and ambient heat on exercising
females. J. Appl. Physiol.: Respir. Environ. Exercise Physiol. 57: 450-456.
Gliner, J. A.; Horvath, S. M.; Folinsbee, L. J. (1983) Preexposure to low ozone concentrations does not diminish
the pulmonary function response on exposure to higher ozone concentrations. Am. Rev. Respir. Dis.
127: 51-55.
Golden, J. A.; Nadel, J. A.; Boushey, H. A. (1978) Bronchial hyperirritability in healthy subjects after exposure
to ozone. Am. Rev. Respir. Dis. 118: 287-294.
Goldsmith, J. R.; Griffith, H. L.; Detels, R.; Beeser, S.; Neumann, L. (1983) Emergency room admissions,
meteorologic variables, and air pollutants: a path analysis. Am. J. Epidemiol. 118: 759-778.
Gong, H., Jr. (1987) Relationship between air quality and the respiratory status of asthmatics in an area of high
oxidant pollution in Los Angeles County [final report]. Sacramento, CA: California Air Resources
Board; contract nos. Al-151-33 and A4-135-33.
Gong, H. J.; Tashkin, D. P.; Simmons, M. S.; Coulson, A. H.; Clark, V. A.; Spivey, G. H.; Sayre, J. W.
(1985) Relationship between oxidant air pollution and respiratory status of asthmatics in Los Angeles
County: methodology and study sample. In: Lee, S. D., ed. Evaluation of the scientific basis for
7-200
-------
ozone/oxidants standards: transactions of an APCA international specialty conference; November
1984; Houston, TX. Pittsburgh, PA: Air Pollution Control Association; pp. 226-237.
Gong, H., Jr.; Bradley, P. W.; Simmons, M. S.; Tashkin, D. P. (1986) Impaired exercise performance and
pulmonary function in elite cyclists during low-level ozone exposure in a hot environment. Am. Rev.
Respir. Dis. 134: 726-733.
Gong, H., Jr.; Bedi, J. P.; Horvath, S. M. (1988) Inhaled albuterol does not protect against ozone toxicity in
nonasthmatic athletes. Arch. Environ. Health 43: 46-53.
Gordon, T.; Venugopalan, C. S.; Amdur, M. O.; Brazen, J. M. (1984) Ozone-induced airway hyperreactivity in
the guinea pig. J. Appl. Physiol.: Respir. Environ. Exercise Physiol. 57: 1034-1038.
Graham, D. E.; Koren, H. S. (1990) Biomarkers of inflammation in ozone-exposed humans: comparison of the
nasal and bronchoalveolar lavage. Am. Rev. Respir. Dis. 142: 152-156.
Graham, D.; Henderson, P.; House, D. (1988) Neutrophil influx measured in nasal lavages of humans exposed to
ozone. Arch. Environ. Health 43: 228-233.
Greve, J. M.; Davis, G.; Meyer, A. M.; Forte, C. P.; Yost, S. C.; Marlor, C. W.; Kamarck, M. E.;
McClelland, A. (1989) The major human rhinovirus receptor is ICAM-1. Cell 56: 839-847.
Gross, K. B.; White, H. J.; Sargent, N. E. (1991) The effect of ozone inhalation on metabolic functioning of
vascular endothelium and on ventilatory function. Toxicol. Appl. Pharmacol. 109: 336-351.
Hackney, J. D.; Linn, W. S.; Law, D. C.; Karuza, S. K.; Greenberg, H.; Buckley, R. D.; Pedersen, E. E.
(1975) Experimental studies on human health effects of air pollutants: III. two-hour exposure to
ozone alone and in combination with other pollutant gases. Arch. Environ. Health 30: 385-390.
Hackney, J. D.; Linn, W. S.; Buckley, R. D.; Hislop, H. J. (1976) Studies in adaptation to ambient oxidant air
pollution: effects of ozone exposure in Los Angeles residents vs. new arrivals. Environ. Health
Perspect. 18: 141-146.
Hackney, J. D.; Linn, W. S.; Mohler, J. G.; Collier, C. R. (1977a) Adaptation to short-term respiratory effects
of ozone in men exposed repeatedly. J. Appl. Physiol.: Respir. Environ. Exercise Physiol.
43: 82-85.
Hackney, J. D.; Linn, W. S.; Karuza, S. K.; Buckley, R. D.; Law, D. C.; Bates, D. V.; Hazucha, M.; Pengelly,
L. D.; Silvernian, F. (1977b) Effects of ozone exposure in Canadians and southern Californians:
evidence for adaptation? Arch. Environ. Health 32: 110-116.
Hackney, J. D.; Linn, W. S.; Shamoo, D. A.; Avol, E. L. (1989) Responses of selected reactive and nonreactive
volunteers to ozone exposure in high and low pollution seasons. In: Schneider, T.; Lee, S. D.;
Wolters, G. J. R.; Grant, L. D., eds. Atmospheric ozone research and its policy implications:
proceedings of the 3rd US-Dutch international symposium; May 1988; Nijmegen, The Netherlands.
Amsterdam, The Netherlands: Elsevier Science Publishers; pp. 311-318. (Studies in environmental
science 35).
Hammer, D. L; Hasselblad, V.; Portnoy, B.; Wehrle, P. F. (1974) Los Angeles student nurse study: daily
symptom reporting and photochemical oxidants. Arch. Environ. Health 28: 255-260.
Hazbun, M. E.; Hamilton, R.; Holian, A.; Eschenbacher, W. L. (1993) Ozone-induced increases in substance P
and 8-epi-prostaglandin F2n in the airways of human subjects. Am. J. Respir. Cell Mol. Biol.
9: 568-572.
7-201
-------
Hazucha, M. J. (1987) Relationship between ozone exposure and pulmonary function changes. J. Appl. Physiol.
62: 1671-1680.
Hazucha, M.; Silverman, P.; Parent, C.; Field, S.; Bates, D. V. (1973) Pulmonary function in man after
short-term exposure to ozone. Arch. Environ. Health 27: 183-188.
Hazucha, M. J.; Bates, D. V.; Bromberg, P. A. (1989) Mechanism of action of ozone on the human lung.
J. Appl. Physiol. 67: 1535-1541.
Hazucha, M. J.; Folinsbee, L. J.; Seal, E., Jr. (1992) Effects of steady-state and variable ozone concentration
profiles on pulmonary function. Am. Rev. Respir. Dis. 146: 1487-1493.
Hazucha, M. J.; Folinsbee, L. J.; Seal, E.; Bromberg, P. A. (1994) Lung function response of healthy women
after sequential exposures to NO2 and O3. Am. J. Respir. Crit. Care Med. 150: 642-647.
Hazucha, M. J.; Madden, M.; Pape, G.; Becker, S.; Devlin, R. B.; Koren, H. S.; Kehrl, H.; Bromberg, P. A.
(1996) Effects of cyclooxygenase inhibition on ozone-induced respiratory inflammation and lung
function changes. Eur. J. Appl. Physiol. Occup. Med. 73: 17-27.
Hechter, H. H.; Goldsmith, J. R. (1961) Air pollution and daily mortality. Am. J. Med. Sci. 241: 581-588.
Henderson, F. W.; Dubovi, E. J.; Harder, S.; Seal, E., Jr.; Graham, D. (1988) Experimental rhinovirus infection
in human volunteers exposed to ozone. Am. Rev. Respir. Dis. 137: 1124-1128.
Higgins, I. T. T.; D'Arcy, J. B.; Gibbons, D. I.; Avol, E. L.; Gross, K. B. (1990) Effect of exposures to ambient
ozone on ventilatory lung function in children. Am. Rev. Respir. Dis. 141: 1136-1146.
Hodgkin, J. E.; Abbey, D. E.; Euler, G. L.; Magie, A. R. (1984) COPD prevalence in nonsmokers in high and
low photochemical air pollution areas. Chest 86: 830-838.
Hoek, G.; Brunekreef, B. (1992) Time trends in repeated spirometry in children. Eur. Respir. J. 5: 553-559.
Hoek, G.; Brunekreef, B.; Kosterink, P.; Van den Berg, R.; Hofschreuder, P. (1993a) Effect of ambient ozone on
peak expiratory flow of exercising children in the Netherlands. Arch. Environ. Health 48: 27-32.
Hoek, G.; Fischer, P.; Brunekreef, B.; Lebret, E.; Hofschreuder, P.; Mennen, M. G. (1993b) Acute effects of
ambient ozone on pulmonary function of children in the Netherlands. Am. Rev. Respir. Dis.
147: 111-117.
Holguin, A. H.; Buffler, P. A.; Contant, C. P., Jr.; Stock, T. H.; Kotchmar, D.; Hsi, B. P.; Jenkins, D. E.;
Gehan, B. M.; Noel, L. M.; Mei, M. (1985) The effects of ozone on asthmatics in the Houston area.
In: Lee, S. D., ed. Evaluation of the scientific basis for ozone/oxidants standards: proceedings of an
APCA international specialty conference; November 1984; Houston, TX. Pittsburgh, PA: Air
Pollution Control Association; pp. 262-280. (APCA international specialty conference transactions:
TR-4).
Holtzman, M. J.; Cunningham, J. H.; Sheller, J. R.; Irsigler, G. B.; Nadel, J. A.; Boushey, H. A. (1979) Effect
of ozone on bronchial reactivity in atopic and nonatopic subjects. Am. Rev. Respir. Dis. 120: 1059-
1067.
Holtzman, M. J.; Fabbri, L. M.; O'Byrne, P. M.; Gold, B. D.; Aizawa, H.; Walters, E. H.; Alpert, S. E.;
Nadel, J. A. (1983) Importance of airway inflammation for hyperresponsiveness induced by ozone.
Am. Rev. Respir. Dis. 127: 686-690.
7-202
-------
Horstman, D. H.; Folinsbee, L. J.; Ives, P. J.; Abdul-Salaam, S.; McDonnell, W. F. (1990) Ozone concentration
and pulmonary response relationships for 6.6-hour exposures with five hours of moderate exercise to
0.08, 0.10, and 0.12 ppm. Am. Rev. Respir. Dis. 142: 1158-1163.
Horstman, D. H.; Ball, B. A.; Brown, J.; Gerrity, T.; Folinsbee, L. J. (1995) Comparison of pulmonary
responses of asthmatic and nonasthmatic subjects performing light exercise while exposed to a low
level of ozone. Toxicol. Ind. Health 11: 369-385.
Horvath, S. M.; Gliner, J. A.; Matsen-Twisdale, J. A. (1979) Pulmonary function and maximum exercise
responses following acute ozone exposure. Aviat. Space Environ. Med. 50: 901-905.
Horvath, S. M.; Gliner, J. A.; Folinsbee, L. J. (1981) Adaptation to ozone: duration of effect. Am. Rev. Respir.
Dis. 123: 496-499.
Horvath, S. M.; Bedi, J. P.; Drechsler-Parks, D. M. (1986) Effects of peroxyacetyl nitrate alone and in
combination with ozone in healthy young women. J. Air Pollut. Control Assoc. 36: 265-270.
Horvath, S. M.; Folinsbee, L. J.; Bedi, J. F. (1987) Combined effect of ozone and sulfuric acid on pulmonary
function in man. Am. Ind. Hyg. Assoc. J. 48: 94-98.
Horvath, S. M.; Bedi, J. P.; Drechsler-Parks, D. M.; Williams, R. E. (1991) Alterations in pulmonary function
parameters during exposure to 80 ppb ozone for 6.6 hours in healthy middle aged individuals.
In: Berglund, R. L.; Lawson, D. R.; McKee, D. J., eds. Tropospheric ozone and the environment:
papers from an international conference; March 1990; Los Angeles, CA. Pittsburgh, PA: Air &
Waste Management Association; pp. 59-70. (A&WMA transaction series no. TR-19).
Hynes, B.; Silverman, F.; Cole, P.; Corey, P. (1988) Effects of ozone exposure: a comparison between oral and
nasal breathing. Arch. Environ. Health 43: 357-360.
Ipsen, J.; Ingenito, F. E.; Deane, M. (1969) Episodic morbidity and mortality in relation to air pollution. Arch.
Environ. Health 18: 458-461.
Johnson, D. A.; Winters, R. S.; Woolley, T.; Graham, D.; Henderson, F. W. (1986) Ozone effects on
alpha- 1-proteinase inhibitor in vivo: blood plasma inhibitory activity is unchanged. Exp. Lung Res.
11: 95-103.
Kagawa, J. (1986) Experimental studies on human health effects of aerosol and gaseous pollutants. In: Lee, S. D.;
Schneider, T.; Grant, L. D.; Verkerk, P. J., eds. Aerosols: research, risk assessment and control
strategies—proceedings of the second US-Dutch international symposium; May 1985; Williamsburg,
VA. Chelsea, MI: Lewis Publishers, Inc.; pp. 683-697.
Kagawa, J.; Tsuru, K. (1979) Effects of ozone and smoking alone and in combination on bronchial reactivity to
inhaled acetylcholine. Nippon Kyobu Shikkan Gakkai Zasshi 17: 703-709.
Keefe, M. J.; Bennett, W. D.; Dewitt, P.; Seal, E.; Strong, A. A.; Gerrity, T. R. (1991) The effect of ozone
exposure on the dispersion of inhaled aerosol boluses in healthy human subjects. Am. Rev. Respir.
Dis. 144: 23-30.
Kehrl, H. R.; Hazucha, M. J.; Solic, J. J.; Bromberg, P. A. (1985) Responses of subjects with chronic
obstructive pulmonary disease after exposures to 0.3 ppm ozone. Am. Rev. Respir. Dis.
131: 719-724.
Kehrl, H. R.; Vincent, L. M.; Kowalsky, R. J.; Horstman, D. H.; O'Neil, J. J.; McCartney, W. H.; Bromberg,
P. A. (1987) Ozone exposure increases respiratory epithelial permeability in humans. Am. Rev.
Respir. Dis. 135: 1124-1128.
7-203
-------
Kehrl, H. R.; Vincent, L. M.; Kowalsky, R. J.; Horstman, D. H.; O'Neil, J. J.; McCartney, W. H.; Bromberg,
P. A. (1989) Ozone-induced changes in the pulmonary clearance of 99mTc-DTPA in man.
In: Schneider, T.; Lee, S. D.; Wolters, G. J. R.; Grant, L. D., eds. Atmospheric ozone research
and its policy implications: proceedings of the 3rd US-Dutch international symposium; May 1988;
Nijmegen, The Netherlands. Amsterdam, The Netherlands: Elsevier Science Publishers; pp. 343-
351. (Studies in environmental science 35).
Kennedy, S. M.; Burrows, B.; Vedal, S.; Enarson, D. A.; Chan-Yeung, M. (1990) Methacholine responsiveness
among working populations: relationship to smoking and airway caliber. Am. Rev. Respir. Dis. 143:
1377-1383.
Kerr, H. D.; Kulle, T. J.; Mcllhany, M. L.; Swidersky, P. (1975) Effects of ozone on pulmonary function in
normal subjects: an environmental-chamber study. Am. Rev. Respir. Dis. Ill: 763-773.
Kilburn, K. H.; Warshaw, R. H.; Thornton, J. C. (1992) Expiratory flows decreased in Los Angeles children
from 1984 to 1987: is this evidence of effects of air pollution? Environ. Res. 59: 150-158.
Kinney, P. L. (1986) Short-term lung function associations with air pollution in Kingston and Harriman,
Tennessee [dissertation]. Boston, MA: Harvard University, School of Public Health.
Kinney, P. L.; Ozkaynak, H. (1991) Associations of daily mortality and air pollution in Los Angeles County.
Environ. Res. 54: 99-120.
Kinney, P. L.; Ware, J. H.; Spengler, J. D.; Dockery, D. W.; Speizer, F. E.; Ferris, B. G., Jr. (1989)
Short-term pulmonary function change in association with ozone levels. Am. Rev. Respir. Dis.
139: 56-61.
Kleeberger, S. R.; Hudak, B. B. (1992) Acute ozone-induced change in airway permeability: role of infiltrating
leukocytes. J. Appl. Physiol. 72: 670-676.
Koenig, J. Q.; Covert, D. S.; Morgan, M. S.; Horike, M.; Horike, N.; Marshall, S. G.; Pierson, W. E. (1985)
Acute effects of 0.12 ppm ozone or 0.12 ppm nitrogen dioxide on pulmonary function in healthy and
asthmatic adolescents. Am. Rev. Respir. Dis. 132: 648-651.
Koenig, J. Q.; Covert, D. S.; Marshall, S. G.; Van Belle, G.; Pierson, W. E. (1987) The effects of ozone and
nitrogen dioxide on pulmonary function in healthy and in asthmatic adolescents. Am. Rev. Respir.
Dis. 136: 1152-1157.
Koenig, J. Q.; Covert, D. S.; Smith, M. S.; Van Belle, G.; Pierson, W. E. (1988) The pulmonary effects of
ozone and nitrogen dioxide alone and combined in healthy and asthmatic adolescent subjects. Toxicol.
Ind. Health 4: 521-532.
Koenig, J. Q.; Covert, D. S.; Hanley, Q. S.; Van Belle, G.; Pierson, W. E. (1990) Prior exposure to ozone
potentiates subsequent response to sulfur dioxide in adolescent asthmatic subjects. Am. Rev. Respir.
Dis. 141: 377-380.
Koenig, J. Q.; Larson, T. V.; Hanley, Q. S.; Rebolledo, V.; Dumler, K.; Checkoway, H.; Wang, S.-Z.;
Lin, D.; Pierson, W. E. (1993) Pulmonary function changes in children associated with fine
particulate matter. Environ. Res. 63: 26-38.
Koenig, J. Q.; Covert, D. S.; Pierson, W. E.; Hanley, Q. S.; Rebolledo, V.; Dumler, K.; McKinney, S. E.
(1994) Oxidant and acid aerosol exposure in healthy subjects and subjects with asthma. Part I: effects
of oxidants, combined with sulfuric or nitric acid, on the pulmonary function of adolescents with
asthma. Cambridge, MA: Health Effects Institute; pp. 1-36; research report no. 70.
7-204
-------
Koepchen, H. P.; Kalia, M.; Sommer, D.; Klussendorf, D. (1977) Action of type J afferents on the discharge
pattern of medullary respiratory neurons. In: Paintal, A. S.; Gill-Kumar, P., eds. Respiratory
adaptations, capillary exchange, and reflex mechanisms: proceedings of the August Krogh centenary
symposium; October 1974; Srinagar, India. Dehli, India: University of Delhi; pp. 407-425.
Konig, G.; Rommelt, H.; Kienle, H.; Dirnagl, K.; Polke, H.; Fruhmann, G. (1980) Anderung der
bronchomotorischen Reagibilitat des Menschen durch Einwirkung von Ozon [Changes in the
bronchial reactivity of humans caused by the influence of ozone]. Arbeitsnied. Sozialmed.
Praeventivmed. 151: 261-263.
Koren, H. S.; Devlin, R. B.; Graham, D. E.; Mann, R.; McGee, M. P.; Horstman, D. H.; Kozumbo, W. J.;
Becker, S.; House, D. E.; McDonnell, W. P.; Bromberg, P. A. (1989a) Ozone-induced
inflammation in the lower airways of human subjects. Am. Rev. Respir. Dis. 139: 407-415.
Koren, H. S.; Devlin, R. B.; Graham, D. E.; Mann, R.; McDonnell, W. F. (1989b) The inflammatory response
inhuman lung exposed to ambient levels of ozone. In: Schneider, T.; Lee, S. D.; Wolters, G. J. R.;
Grant, L. D., eds. Atmospheric ozone research and its policy implications: proceedings of the 3rd
US-Dutch international symposium; May 1988; Nijmegen, The Netherlands. Amsterdam, The
Netherlands: Elsevier Science Publishers; pp. 745-753. (Studies in environmental science 35).
Koren, H. S.; Hatch, G. E.; Graham, D. E. (1990) Nasal lavage as a tool in assessing acute inflammation in
response to inhaled pollutants. Toxicology 60: 15-25.
Koren, H. S.; Devlin, R. B.; Becker, S.; Perez, R.; McDonnell, W. F. (1991) Time-dependent changes of
markers associated with inflammation in the lungs of humans exposed to ambient levels of ozone.
Toxicol. Pathol. 19: 406-411.
Kreit, J. W.; Gross, K. B.; Moore, T. B.; Lorenzen, T. J.; D'Arcy, J.; Eschenbacher, W. L. (1989)
Ozone-induced changes in pulmonary function and bronchial responsiveness in asthmatics. J. Appl.
Physiol. 66: 217-222.
Krupnik, A. J.; Harrington, W.; Ostro, B. (1990) Ambient ozone and acute health effects: evidence from daily
data. J. Environ. Econ. Manage. 18: 1-18.
Krzyzanowski, M.; Quackenboss, J.; Lebowitz, M. (1989) Sub-chronic respiratory effects from long term
exposure to ozone in Tucson. Presented at: 82nd annual meeting and exhibition of the Air & Waste
Management Association; June; Anaheim, CA. Pittsburgh, PA: Air & Waste Management
Association; paper no. 89-27.4.
Krzyzanowski, M.; Quackenboss, J. J.; Lebowitz, M. D. (1992) Relation of peak expiratory flow rates and
symptoms to ambient ozone. Arch. Environ. Health 47: 107-115.
Kulle, T. J.; Sauder, L. R.; Kerr, H. D.; Farrell, B. P.; Bermel, M. S.; Smith, D. M. (1982) Duration of
pulmonary function adaptation to ozone in humans. Am. Ind. Hyg. Assoc. J. 43: 832-837.
Kulle, T. J.; Milman, J. H.; Sauder, L. R.; Kerr, H. D.; Farrell, B. P.; Miller, W. R. (1984) Pulmonary function
adaptation to ozone in subjects with chronic bronchitis. Environ. Res. 34: 55-63.
Kulle, T. J.; Sauder, L. R.; Hebel, J. R.; Chatham, M. D. (1985) Ozone response relationships in healthy
nonsmokers. Am. Rev. Respir. Dis. 132: 36-41.
Larsen, R. L; McDonnell, W. P.; Horstman, D. H.; Folinsbee, L. J. (1991) An air quality data analysis system
for interrelating effects, standards, and needed source reductions: part 11. a lognormal model relating
human lung function decrease to O3 exposure. J. Air Waste Manage. Assoc. 41: 455-459.
7-205
-------
Lategola, M. T.; Melton, C. E.; Higgins, E. A. (1980a) Effects of ozone on symptoms and cardiopulmonary
function in a flight attendant surrogate population. Aviat. Space Environ. Med. 51: 237-246.
Lategola, M. T.; Melton, C. E.; Higgins, E. A. (1980b) Pulmonary and symptom threshold effects of ozone in
airline passenger and cockpit crew surrogates. Aviat. Space Environ. Med. 51: 878-884.
Lauritzen, S. K.; Adams, W. C. (1985) Ozone inhalation effects consequent to continuous exercise in females:
comparison to males. J. Appl. Physiol. 59: 1601-1606.
Lebowitz, M. D. (1984) The effects of environmental tobacco smoke exposure and gas stoves on daily peak flow
rates in asthmatic and non-asthmatic families. In: Rylander, R.; Peterson, Y.; Snella, M.-C., eds.
ETS—environmental tobacco smoke: report from a workshop on effects and exposure levels; March
1983; Geneva, Switzerland. Eur. J. Respir. Dis. 65(suppl. 133): 90-97.
Lebowitz, M. D.; O'Rourke, M. K.; Dodge, R.; Holberg, C. J.; Gorman, G.; Hoshaw, R. W.; Pinnas, J. L.;
Barbee, R. A.; Sneller, M. R. (1982) The adverse health effects of biological aerosols, other
aerosols, and indoor microclimate on asthmatics and nonasthmatics. Environ. Int. 8: 375-380.
Lebowitz, M. D.; Holberg, C. J.; Dodge, R. R. (1983) Respiratory effects on populations from low-level
exposures to ozone. Presented at: 76th annual meeting of the Air Pollution Control Association; June;
Atlanta, GA. Pittsburgh, PA: Air Pollution Control Association; paper no. 83-12.5.
Lebowitz, M. D.; Holberg, C. J.; Boyer, B.; Hayes, C. (1985) Respiratory symptoms and peak flow associated
with indoor and outdoor air pollutants in the southwest. J. Air Pollut. Control Assoc. 35: 1154-1158.
Lebowitz, M. D.; Collins, L.; Holberg, C. J. (1987) Time series analyses of respiratory responses to indoor and
outdoor environmental phenomena. Environ. Res. 43: 332-341.
Lebowitz, M. D.; Quackenboss, J. J.; Krzyzanowski, M. (1991) Acute respiratory effects of prolonged ambient
ozone. In: Berglund, R. L.; Lawson, D. R.; McKee, D. J., eds. Tropospheric ozone and the
environment: papers from an international conference; March 1990; Los Angeles, CA. Pittsburgh,
PA: Air & Waste Management Association; pp. 111-119. (A&WMA transaction series no. TR-19).
Lee, L.-Y.; Bleecker, E. R.; Nadel, J. A. (1977) Effect of ozone on bronchomotor response to inhaled histamine
aerosol in dogs. J. Appl. Physiol.: Respir. Environ. Exercise Physiol. 43: 626-631.
Lee, L.-Y.; Dumont, C.; Djokic, T. D.; Menzel, T. E.; Nadel, J. A. (1979) Mechanism of rapid, shallow
breathing after ozone exposure in conscious dogs. J. Appl. Physiol.: Respir. Environ. Exercise
Physiol. 46: 1108-1114.
Levy, D.; Gent, M.; Newhouse, M. T. (1977) Relationship between acute respiratory illness and air pollution
levels in an industrial city. Am. Rev. Respir. Dis. 116: 167-173.
Li, Z.; Daniel, E. E.; Lane, C. G.; Arnaout, M. A.; O'Byrne, P. M. (1992) Effect of an anti-mo 1 MAb on
ozone-induced airway inflammation and airway hyperresponsiveness in dogs. Am. J. Physiol.
263: L723-L726.
Linder, J.; Herren, D.; Monn, C.; Wanner, H.-U. (1988) Die Wirkung von Ozon auf die koerperliche
Leistungsfaehigkeit [The effect of ozone on physical activity]. Schweiz Z. Sportmed. 36: 5-10.
Linn, W. S.; Buckley, R. D.; Spier, C. E.; Blessey, R. L.; Jones, M. P.; Fischer, D. A.; Hackney, J. D. (1978)
Health effects of ozone exposure in asthmatics. Am. Rev. Respir. Dis. 117: 835-843.
7-206
-------
Linn, W. S.; Jones, M. P.; Bachmayer, E. A.; Spier, C. E.; Mazur, S. P.; Avol, E. L.; Hackney, J. D. (1980)
Short-term respiratory effects of polluted ambient air: a laboratory study of volunteers in a high-
oxidant community. Am. Rev. Respir. Dis. 121: 243-252.
Linn, W. S.; Fischer, D. A.; Medway, D. A.; Anzar, U. T.; Spier, C. E.; Valencia, L. M.; Venet, T. G.;
Hackney, J. D. (1982a) Short-term respiratory effects of 0.12 ppm ozone exposure in volunteers with
chronic obstructive pulmonary disease. Am. Rev. Respir. Dis. 125: 658-663.
Linn, W. S.; Medway, D. A.; Anzar, U. T.; Valencia, L. M.; Spier, C. E.; Tsao, F. S.-D.; Fischer, D. A.;
Hackney, J. D. (1982b) Persistence of adaptation to ozone in volunteers exposed repeatedly for six
weeks. Am. Rev. Respir. Dis. 125: 491-495.
Linn, W. S.; Shamoo, D. A.; Venet, T. G.; Spier, C. E.; Valencia, L. M.; Anzar, U. T.; Hackney, J. D.
(1983a) Response to ozone in volunteers with chronic obstructive pulmonary disease. Arch. Environ.
Health 38: 278-283.
Linn, W. S.; Avol, E. L.; Hackney, J. D. (1983b) Effects of ambient oxidant pollutants on humans: a movable
environmental chamber study. In: Lee, S. D.; Mustafa, M. G.; Mehlman, M. A., eds. International
symposium on the biomedical effects of ozone and related photochemical oxidants; March 1982;
Pinehurst, NC. Princeton, NJ: Princeton Scientific Publishers, Inc.; pp. 125-137. (Advances in
modern environmental toxicology: v. 5).
Linn, W. S.; Avol, E. L.; Shamoo, D. A.; Spier, C. E.; Valencia, L. M.; Venet, T. G.; Fischer, D. A.;
Hackney, J. D. (1986) A dose-response study of healthy, heavily exercising men exposed to ozone at
concentrations near the ambient air quality standard. Toxicol. Ind. Health 2: 99-112.
Linn, W. S.; Avol, E. L.; Shamoo, D. A.; Peng, R.-C.; Valencia, L. M.; Little, D. E.; Hackney, J. D. (1988)
Repeated laboratory ozone exposures of volunteer Los Angeles residents: an apparent seasonal
variation in response. Toxicol. Ind. Health 4: 505-520.
Linn, W. S.; Shamoo, D. A.; Anderson, K. R.; Peng, R.-C.; Avol, E. L.; Hackney, J. D. (1994) Effects of
prolonged, repeated exposure to ozone, suliuric acid, and their combination in healthy and asthmatic
volunteers. Am. J. Respir. Crit. Care Med. 150: 431-440.
Lipfert, F. W. (1993) A critical review of studies of the association between demands for hospital services and air
pollution. Environ. Health Perspect. 101(suppl. 2): 229-268.
Lipfert, F. W.; Hamnierstrom, T. (1992) Temporal patterns in air pollution and hospital admissions. Environ.
Res. 59: 374-399.
Martinez, P.; Sunyer, J.; Anto, J. M. (1993) Reliability of a monitoring system for respiratory emergency room
admissions. Eur. Respir. J. 6: 337-341.
Massey, F. J., Jr.; Landau, E.; Deane, M. (1961) Air pollution and mortality in two areas of Los Angeles
County. Presented at: the joint meeting of the American Statistical Association and the Biometric
Society (ENAR); December; New York, NY.
McBride, D. E.; Koenig, J. Q.; Luchtel, D. L.; Williams, P. V.; Henderson, W. R., Jr. (1994) Inflammatory
effects of ozone in the upper airways of subjects with asthma. Am. J. Respir. Crit. Care Med.
149: 1192-1197.
McDonnell, W. P.; Seal, E., Jr. (1991) Relationships between lung function and physical characteristics in young
adult black and white males and females. Eur. Respir. J. 4: 279-289.
7-207
-------
McDonnell, W. P.; Horstman, D. H.; Hazucha, M. J.; Seal, E., Jr.; Haak, E. D.; Salaam, S. A.; House, D. E.
(1983) Pulmonary effects of ozone exposure during exercise: dose-response characteristics. J. Appl.
Physiol.: Respir. Environ. Exercise Physiol. 54: 1345-1352.
McDonnell, W. P., Ill; Chapman, R. S.; Leigh, M. W.; Strope, G. L.; Collier, A. M. (1985a) Respiratory
responses of vigorously exercising children to 0.12 ppni ozone exposure. Am. Rev. Respir. Dis.
132: 875-879.
McDonnell, W. P., Ill; Horstman, D. H.; Abdul-Salaam, S.; House, D. E. (1985b) Reproducibility of individual
responses to ozone exposure. Am. Rev. Respir. Dis. 131: 36-40.
McDonnell, W. P.; Horstman, D. H.; Abdul-Salaam, S.; Raggio, L. J.; Green, J. A. (1987) The respiratory
responses of subjects with allergic rhinitis to ozone exposure and their relationship to nonspecific
airway reactivity. Toxicol. Ind. Health 3: 507-517.
McDonnell, W. P.; Kehrl, H. R.; Abdul-Salaam, S.; Ives, P. J.; Folinsbee, L. J.; Devlin, R. B.; O'Neil, J. J.;
Horstman, D. H. (1991) Respiratory response of humans exposed to low levels of ozone for
6.6 hours. Arch. Environ. Health 46: 145-150.
McDonnell, W. P.; Muller, K. E.; Bromberg, P. A.; Shy, C. M. (1993) Predictors of individual differences in
acute response to ozone exposure. Am. Rev. Respir. Dis. 147: 818-825.
McGee, M. P.; Devlin, R.; Saluta, G.; Koren, H. (1990) Tissue factor and factor VII messenger RNAs in human
alveolar macrophages: effects of breathing ozone. Blood 75: 122-127.
McKenzie, D. C.; Stirling, D. R.; Fadl, S.; Allen, M. (1987) The effects of salbutamol on pulmonary function in
cyclists exposed to ozone: a pilot study. Can. J. Sport Sci. 12: 46-48.
McKinnon, K. P.; Madden, M. C.; Noah, T. L.; Devlin, R. B. (1993) In vitro ozone exposure increases release
of arachidonic acid products from a human bronchial epithelial cell line. Toxicol. Appl. Pharmacol.
118: 215-223.
McKittrick, T.; Adams, W. C. (1995) Pulmonary function response to equivalent doses of ozone consequent to
intermittent and continuous exercise. Arch. Environ. Health. 50: 153-158.
Mellick, P. W.; Dungworth, D. L.; Schwartz, L. W.; Tyler, W. S. (1977) Short term morphologic effects of high
ambient levels of ozone on lungs of rhesus monkeys. Lab. Invest. 36: 82-90.
Melton, C. E. (1990) Airliner cabin ozone: an updated review. Washington, DC: Federal Aviation
Administration, Office of Aviation Medicine; report no. DOT/FAA/AM-89/13. Available from:
NTIS, Springfield, VA; AD-A219 264.
Messineo, T. D.; Adams, W. C. (1990) Ozone inhalation effects in females varying widely in lung size:
comparison with males. J. Appl. Physiol. 69: 96-103.
Mills, C. A. (1957a) Do smogs threaten community health? Cincinnati J. Med. 38: 259-261.
Mills, C. A. (1957b) Respiratory and cardiac deaths in Los Angeles smogs. Am. J. Med. Sci. 233: 379-386.
Mills, C. A. (1960) Respiratory and cardiac deaths in Los Angeles smogs during 1956, 1957 and 1958.
Am. J. Med. Sci. 239: 307-315.
Mills, P. K.; Abbey, D.; Beeson, W. L.; Petersen, F. (1991) Ambient air pollution and cancer in California
Seventh-day Adventists. Arch. Environ. Health 46: 271-280.
7-208
-------
Molfmo, N. A.; Wright, S. C.; Katz, I.; Tarlo, S.; Silverman, P.; McClean, P. A.; Szalai, J. P.; Raizenne, M.;
Slutsky, A. S.; Zamel, N. (1991) Effect of low concentrations of ozone on inhaled allergen responses
in asthmatic subjects. Lancet 338(8761): 199-203.
Nagda, N. L.; Fortmann, R. C.; Koontz, M. D.; Baker, S. R.; Ginevan, M. E. (1991) Airliner cabin
environment: contaminant measurements, health risks, and mitigation options. Washington, DC: U.S.
Department of Transportation, Office of the Secretary. Available from: NTIS, Springfield, VA;
PB91-159384.
Namekata, T.; Carnow, B. W.; Flournoy-Gill, Z.; O'Farrell, E. B.; Reda, D. (1979) Model for measuring the
health impact from changing levels of ambient air pollution: morbidity study. Research Triangle
Park, NC: U.S. Environmental Protection Agency, Health Effects Research Laboratory; report no.
EPA-600/1-79-024. Available from: NTIS, Springfield, VA; PB80-107030.
Niewoehner, D. E.; Kleinerman, J.; Rice, D. B. (1974) Pathologic changes in the peripheral airways of young
cigarette smokers. N. Engl. J. Med. 291: 755-758.
O'Byrne, P. M.; Walters, E. H.; Gold, B. D.; Aizawa, H. A.; Fabbri, L. M.; Alpert, S. E.; Nadel, J. A.;
Holtzman, M. J. (1984) Neutrophil depletion inhibits airway hyperresponsiveness induced by ozone
exposure. Am. Rev. Respir. Dis. 130: 214-219.
Okazawa, A.; Kobayashi, H.; Imai, T.; Konno, S.; Sunouchi, K.; Hoshino, H.; Okada, Y.; Suganuma, T.;
Maruyama, S.; Adachi, M.; Takahashi, T.; Yamaguchi, H.; Misawa, M. (1989) Ozone-induced
airway hyperresponsiveness in guinea pigs. Arerugi 38: 1217-1225.
Ostro, B. D.; Lipsett, M. J.; Mann, J. K.; Krupnick, A.; Harrington, W. (1993) Air pollution and respiratory
morbidity among adults in Southern California. Am. J. Epidemiol. 137: 691-700.
Ponka, A. (1991) Asthma and low level air pollution in Helsinki. Arch. Environ. Health 46: 262-270.
Pope, C. A., Ill; Dockery, D. W. (1992) Acute health effects of PM10 pollution on symptomatic and asymptomatic
children. Am. Rev. Respir. Dis. 145: 1123-1128.
Pope, C. A., Ill; Dockery, D. W.; Spengler, J. D.; Raizenne, M. E. (1991) Respiratory health and PM,0
pollution: a daily time series analysis. Am. Rev. Respir. Dis. 144: 668-674.
Portney, P. R.; Mullahy, J. (1990) Urban air quality and chronic respiratory disease. Reg. Sci. Urban Econ.
20: 407-418.
Raizenne, M. E.; Spengler, J. D. (1989) Dosimetric model of acute health effects of ozone and acid aerosols in
children. In: Schneider, T.; Lee, S. D.; Wolters, G. J. R.; Grant, L. D., eds. Atmospheric ozone
research and its policy implications: proceedings of the 3rd US-Dutch international symposium; May
1988; Nijmegen, The Netherlands. Amsterdam, The Netherlands: Elsevier Science Publishers;
pp. 319-329. (Studies in environmental science 35).
Raizenne, M.; Stern, B.; Burnett, R.; Spengler, J. (1987) Acute respiratory function and transported air
pollutants: observational studies. Presented at: 80th annual meeting of the Air Pollution Control
Association; June; New York, NY. Pittsburgh, PA: Air Pollution Control Association; paper
no. 87-32.6.
Raizenne, M. E.; Burnett, R. T.; Stern, B.; Franklin, C. A.; Spengler, J. D. (1989) Acute lung function
responses to ambient acid aerosol exposures in children. Environ. Health Perspect. 79: 179-185.
Reisenauer, C. S.; Koenig, J. Q.; McManus, M. S.; Smith, M. S.; Kusic, G.; Pierson, W. E. (1988) Pulmonary
response to ozone exposures in healthy individuals aged 55 years or greater. JAPCA 38: 51-55.
7-209
-------
Rennard, S. I.; Daughton, D. M.; Robbins, R. A.; Thompson, A. B.; Von Essen, S. (1990) In vivo and in vitro
methods for evaluating airways inflammation: implications for respiratory toxicology. Toxicology
60: 5-14.
Reynolds, H. Y. (1987) Bronchoalveolar lavage. Am. Rev. Respir. Dis. 135: 250-263.
Richards, W.; Azen, S. P.; Weiss, J.; Stocking, S.; Church, J. (1981) Los Angeles air pollution and asthma in
children. Ann. Allergy 47: 348-354.
Roemer, W.; Hoek, G.; Brunekreef, B. (1993) Effect of ambient winter air pollution on respiratory health of
children with chronic respiratory symptoms. Am. Rev. Respir. Dis. 147: 118-124.
Rokaw, S. N.; Betels, R.; Coulson, A. H.; Sayre, J. W.; Tashkin, D. P.; Allwright, S. S.; Massey, F. J., Jr.
(1980) The UCLA population studies of chronic obstructive respiratory disease: 3. comparison of
pulmonary function in three communities exposed to photochemical oxidants, multiple primary
pollutants, or minimal pollutants. Chest 78: 252-262.
Rossiter, C. E.; Weill, H. (1974) Ethnic differences in lung function: evidence for proportional differences. Int. J.
Epidemiol. 3: 55-61.
Roum, J. H.; Murlas, C. (1984) Ozone-induced changes in muscarinic bronchial reactivity by different testing
methods. J. Appl. Physiol.: Respir. Environ. Exercise Physiol. 57: 1783-1789.
Samet, J. M.; Lambert, W. E.; James, D. S.; Mermier, C. M.; Chick, T. W. (1993) Assessment of heart rate as
a predictor of ventilation. In: Noninvasive methods for measuring ventilation in mobile subjects.
Cambridge, MA: Health Effects Institute; pp. 19-55; research report no. 59.
Savin, W. M.; Adams, W. C. (1979) Effects of ozone inhalation on work performance andVO2max. J. Appl.
Physiol.: Respir. Environ. Exercise Physiol. 46: 309-314.
Schelegle, E. S.; Adams, W. C. (1986) Reduced exercise time in competitive simulations consequent to low level
ozone exposure. Med. Sci. Sports Exercise 18: 408-414.
Schelegle, E. S.; Adams, W. C.; Siefkin, A. D. (1987) Indomethacinpretreatment reduces ozone-induced
pulmonary function decrements in human subjects. Am. Rev. Respir. Dis. 136: 1350-1354.
Schelegle, E. S.; Adams, W. C.; Giri, S. N.; Siefkin, A. D. (1989) Acute ozone exposure increases plasma
prostaglandin F2D in ozone-sensitive human subjects. Am. Rev. Respir. Dis. 140: 211-216.
Schelegle, E. S.; Siefkin, A. D.; McDonald, R. J. (1991) Time course of ozone-induced neutrophilia in normal
humans. Am. Rev. Respir. Dis. 143: 1353-1358.
Schelegle, E. S.; Carl, M. L.; Coleridge, H. M.; Coleridge, J. C. G.; Green, J. F. (1993) Contribution of vagal
afferents to respiratory reflexes evoked by acute inhalation of ozone in dogs. J. Appl. Physiol.
74: 2338-2344.
Schmidt, T.; Wellhoner, H. H. (1970) The reflex influence of a group of slowly conducting vagal afferents on
D and D discharges in cat intercostal nerves. Pfluegers Arch. 318: 335-345.
Schmitzberger, R.; Rhomberg, K.; Kemmler, G. (1992) Chronic exposure to ozone and respiratory health of
children [letter]. Lancet (8797): 881-882.
Schmitzberger, R.; Rhomberg, K.; Biichele, H.; Puchegger, R.; Schmitzberger-Natznier, D.; Kemmler, G.;
Panosch, B. (1993) Effects of air pollution on the respiratory tract of children. Pediatr. Pulmonol.
15: 68-74.
7-210
-------
Schonfeld, B. R.; Adams, W. C.; Schelegle, E. S. (1989) Duration of enhanced responsiveness upon re-exposure
to ozone. Arch. Environ. Health 44: 229-236.
Schwartz, J. (1989) Lung function and chronic exposure to air pollution: a cross-sectional analysis of NHANES II.
Environ. Res. 50: 309-321.
Schwartz, J. (1991) Paniculate air pollution and daily mortality in Detroit. Environ. Res. 56: 204-213.
Schwartz, J. (1992) Air pollution and the duration of acute respiratory symptoms. Arch. Environ. Health 47: 116-
122.
Schwartz, J. (1994a) Air pollution and hospital admissions for the elderly in Birmingham, Alabama.
Am. J. Epidemiol. 139: 589-598.
Schwartz, J. (1994b) Air pollution and hospital admissions for the elderly in Detroit, Michigan. Am. J. Respir.
Crit. Care Med. 150: 648-655.
Schwartz, J. (1994c) PM10, ozone, and hospital admissions for the elderly in Minneapolis, MN. Arch. Environ.
Health: 49: 366-374.
Schwartz, J.; Zeger, S. (1990) Passive smoking, air pollution, and acute respiratory symptoms in a diary study of
student nurses. Am. Rev. Respir. Dis. 141: 62-67.
Schwartz, J.; Hasselblad, V.; Pitcher, H. (1988) Air pollution and morbidity: a further analysis of the Los
Angeles student nurses data. JAPCA 38: 158-162.
Seal, E., Jr.; McDonnell, W. P.; House, D. E.; Salaam, S. A.; Dewitt, P. J.; Butler, S. O.; Green, J.;
Raggio, L. (1993) The pulmonary response of white and black adults to six concentrations of ozone.
Am. Rev. Respir. Dis. 147: 804-810.
Seal, E., Jr.; McDonnell, W. P.; House, D. E. (1995) The effect of age, socioeconomic status, and menstrual
cycle on the pulmonary response to ozone. Arch. Environ. Health: in press.
Seltzer, J.; Bigby, B. G.; Stulbarg, M.; Holtzman, M. J.; Nadel, J. A.; Ueki, I. P.; Leikauf, G. D.; Goetzl,
E. J.; Boushey, H. A. (1986) O3-induced change in bronchial reactivity to niethacholine and airway
inflammation in humans. J. Appl. Physiol. 60: 1321-1326.
Selwyn, B. J.; Stock, T. H.; Hardy, R. J.; Chan, P. A.; Jenkins, D. E.; Kotchmar, D. J.; Chapman, R. S. (1985)
Health effects of ambient ozone exposure in vigorously exercising adults. In: Lee, S. D., ed.
Evaluation of the scientific basis for ozone/oxidants standards: proceedings of an APCA international
specialty conference; November 1984; Houston, TX. Pittsburgh, PA: Air Pollution Control
Association; pp. 281-296. (APCA international specialty conference transactions: TR-4).
Sherwin, R. P. (1991) Air pollution: the pathobiologic issues. J. Toxicol. Clin. Toxicol. 29: 385-400.
Sherwin, R. P.; Richters, V. (1991) Centriacinar region (CAR) disease in the lungs of young adults: a preliminary
report. In: Berglund, R. L.; Lawson, D. R.; McKee, D. J., eds. Tropospheric ozone and the
environment: papers from an international conference; March 1990; Los Angeles, CA. Pittsburgh,
PA: Air & Waste Management Association; pp. 178-196. (A&WMA transaction series no. TR-19).
Shumway, R. H.; Azari, A. S.; Pawitan, Y. (1988) Modeling mortality fluctuations in Los Angeles as functions of
pollution and weather effects. Environ. Res. 45: 224-241.
Silverrnan, P. (1979) Asthma and respiratory irritants (ozone). Environ. Health Perspect. 29: 131-136.
7-211
-------
Silverman, P.; Folinsbee, L. J.; Barnard, J.; Shephard, R. J. (1976) Pulmonary function changes in
ozone—interaction of concentration and ventilation. J. Appl. Physiol. 41: 859-864.
Solic, J. J.; Hazucha, M. J.; Bromberg, P. A. (1982) The acute effects of 0.2 ppm ozone in patients with chronic
obstructive pulmonary disease. Am. Rev. Respir. Dis. 125: 664-669.
Spektor, D. M.; Lippmann, M. (1991) Health effects of ambient ozone on healthy children at a summer camp. In:
Berglund, R. L.; Lawson, D. R.; McKee, D. J., eds. Tropospheric ozone and the environment:
papers from an international conference; March 1990; Los Angeles, CA. Pittsburgh, PA: Air &
Waste Management Association; pp. 83-89. (A&WMA transaction series no. TR-19).
Spektor, D. M.; Lippmann, M.; Lioy, P. J.; Thurston, G. D.; Citak, K.; James, D. J.; Bock, N.; Speizer, F. E.;
Hayes, C. (1988a) Effects of ambient ozone on respiratory function in active, normal children. Am.
Rev. Respir. Dis. 137: 313-320.
Spektor, D. M.; Lippmann, M.; Thurston, G. D.; Lioy, P. J.; Stecko, J.; O'Connor, G.; Garshick, E.; Speizer,
F. E.; Hayes, C. (1988b) Effects of ambient ozone on respiratory function in healthy adults
exercising outdoors. Am. Rev. Respir. Dis. 138: 821-828.
Spektor, D. M.; Thurston, G. D.; Mao, J.; He, D.; Hayes, C.; Lippmann, M. (1991) Effects of single- and
multiday ozone exposures on respiratory function in active normal children. Environ. Res.
55: 107-122.
Stacy, R. W.; Seal, E., Jr.; Green, J.; House, D. (1982) Pulmonary function in normal humans with exercise and
temperature-humidity stress. J. Appl. Physiol.: Respir. Environ. Exercise Physiol. 53: 1015-1018.
Sterling, T. D.; Phair, J. J.; Pollack, S. V.; Schumsky, D. A.; DeGroot, I. (1966) Urban morbidity and air
pollution: a first report. Arch. Environ. Health 13: 158-170.
Sterling, T. D.; Pollack, S. V.; Phair, J. J. (1967) Urban hospital morbidity and air pollution: a second report.
Arch. Environ. Health 15: 362-374.
Stern, B.; Jones, L.; Raizenne, M.; Burnett, R.; Meranger, J. C.; Franklin, C. A. (1989) Respiratory health
effects associated with ambient sulfates and ozone in two rural Canadian communities. Environ. Res.
49: 20-39.
Stern, B. R.; Raizenne, M. E.; Burnett, R. T.; Jones, L.; Kearney, J.; Franklin, C. A. (1994) Air pollution and
childhood respiratory health: exposure to sulfate and ozone in 10 Canadian rural communities.
Environ. Res. 66: 125-142.
Sunyer, J.; Anto, J. M.; Murillo, C.; Saez, M. (1991) Effects of urban air pollution on emergency room
admissions for chronic obstructive pulmonary disease. Am. J. Epidemiol. 134: 277-286.
Superko, H. R.; Adams, W. C.; Daly, P. W. (1984) Effects of ozone inhalation during exercise in selected
patients with heart disease. Am. J. Med. 77: 463-470.
Tepper, J. S.; Costa, D. L.; Lehmann, J. R.; Weber, M. P.; Hatch, G. E. (1989) Unattenuated structural and
biochemical alterations in the rat lung during functional adaptation to ozone. Am. Rev. Respir. Dis.
140: 493-501.
Thurston, G. D.; Kinney, P. L. (1995) Air pollution epidemiology: considerations in time-series modeling.
Inhalation Toxicol. 7: 71-83.
7-212
-------
Thurston, G. D.; Ito, K.; Kinney, P. L.; Lippmann, M. (1992) A multi-year study of air pollution and respiratory
hospital admissions in three New York State metropolitan areas: results for 1988 and 1989 summers.
J. Exposure Anal. Environ. Epidemiol. 2: 429-450.
Thurston, G. D.; Ito, K.; Hayes, C. G.; Bates, D. V.; Lippmann, M. (1994) Respiratory hospital admissions and
summertime haze air pollution in Toronto, Ontario: consideration of the role of acid aerosols.
Environ. Res. 65: 271-290.
Thurston, G. D.; Lippmann, M.; Scott, M. B.; Fine, J. M. (1995) Summertime haze air pollution and children
with asthma. Am. J. Respir. Crit. Care Med.: submitted.
Townley, R. G.; Ryo, U. Y.; Kolotkin, B. M.; Kang, B. (1975) Bronchial sensitivity to niethacholine in current
and former asthmatic and allergic rhinitis patients and control subjects. J. Allergy Clin. Immunol.
6: 429-442.
Tseng, R. Y. M.; Li, C. K. (1990) Low level atmospheric sulfur dioxide pollution and childhood asthma. Ann.
Allergy 65: 379-383.
Tseng, R. Y. M.; Li, C. K.; Spinks, J. A. (1992) Paniculate air pollution and hospitalization for asthma. Ann.
Allergy 68: 425-432.
Tucker, H. G. (1962) Effects of air pollution and temperature on residents of nursing homes in the Los Angeles
area. Berkeley, CA: California State Public Health.
United Kingdom Ministry of Health. (1954) Mortality and morbidity during the London fog of December 1952.
London, United Kingdom: Her Majesty's Stationery Office; report no. 95.
7-213
-------
U.S. Environmental Protection Agency. (1978) Air quality criteria for ozone and other photochemical oxidants.
Research Triangle Park, NC: Office of Health and Environmental Assessment, Environmental
Criteria and Assessment Office; report no. EPA-600/8-78-004. Available from: NTIS, Springfield,
VA; PB80-124753.
U.S. Environmental Protection Agency. (1986) Air quality criteria for ozone and other photochemical oxidants.
Research Triangle Park, NC: Office of Health and Environmental Assessment, Environmental
Criteria and Assessment Office; report nos. EPA-600/8-84-020aF-eF. 5v. Available from: NTIS,
Springfield, VA; PB87-142949.
U.S. Environmental Protection Agency. (1992) Summary of selected new information on effects of ozone on
health and vegetation: supplement to 1986 air quality criteria for ozone and other photochemical
oxidants. Research Triangle Park, NC: Office of Health and Environmental Assessment,
Environmental Criteria and Assessment Office; report no. EPA/600/8-88/105F. Available from:
NTIS, Springfield, VA; PB92-235670.
Utell, M. J.; Frampton, M. W.; Morrow, P. E.; Cox, C.; Levy, P. C.; Speers, D. M.; Gibb, F. R. (1994)
Oxidant and acid aerosol exposure in healthy subjects and subjects with asthma. Part II: effects of
sequential sulfuric acid and ozone exposures on the pulmonary function of healthy subjects and
subjects with asthma. Cambridge, MA: Health Effects Institute; pp. 37-93; research report no. 70.
Vedal, S.; Schenker, M. B.; Munoz, A.; Samet, J. M.; Batternian, S.; Speizer, F. E. (1987) Daily air pollution
effects on children's respiratory symptoms and peak expiratory flow. Am. J. Public Health 77: 694-
698.
Warren, J. B.; Dalton, N. (1983) A comparison of the bronchodilator and vasopressor effects of exercise levels of
adrenaline in man. Clin. Sci. 64: 475-479.
Wayne, W. S.; Wehrle, P. P.; Carroll, R. E. (1967) Oxidant air pollution and athletic performance. JAMA
J. Am. Med. Assoc. 199: 151-154.
Weinniann, G. G.; Weidenbach-Gerbase, M.; Foster, W. M.; Zacur, H.; Frank, R. (1995) Evidence for ozone-
induced small-airway dysfunction: lack of menstrual-cycle and gender effects. Am. J. Respir. Crit.
Care Med. 152: 988-996.
Weisel, C. (1994) [Memorandum to Dr. George Thurston, New York University Medical Center]. Piscataway,
NJ: Environmental & Occupational Health Sciences Institute; October 12.
Weisel, C. P.; Cody, R. P.; Lioy, P. J. (1995) Relationship between summertime ambient ozone levels and
emergency department visits for asthma in central New Jersey. Environ. Health Perspect. 103(suppl.
2): 97-102.
Weymer, A. R.; Gong, H., Jr.; Lyness, A.; Linn, W. S. (1994) Pre-exposure to ozone does not enhance or
produce exercise-induced asthma. Am. J. Respir. Crit. Care Med. 149: 1413-1419.
White, M. C.; Etzel, R. A.; Wilcox, W. D.; Lloyd, C. (1994) Exacerbations of childhood asthma and ozone
pollution in Atlanta. Environ. Res. 65: 56-68.
Whittemore, A. S.; Korn, E. L. (1980) Asthma and air pollution in the Los Angeles area. Am. J. Public Health
70: 687-696.
Ying, R. L.; Gross, K. B.; Terzo, T. S.; Eschenbacher, W. L. (1990) Indomethacin does not inhibit the ozone-
induced increase in bronchial responsiveness in human subjects. Am. Rev. Respir. Dis.
142: 817-821.
7-214
-------
Young, W. A.; Shaw, D. B.; Bates, D. V. (1964) Effect of low concentrations of ozone on pulmonary function in
man. J. Appl. Physiol. 19: 765-768.
Zwick, H.; Popp, W.; Wagner, C.; Reiser, K.; Schmoger, J.; Bock, A.; Herkner, K.; Radunsky, K. (1991)
Effects of ozone on the respiratory health, allergic sensitization, and cellular immune system in
children. Am. Rev. Respir. Dis. 144: 1075-1079.
7-215
-------
8
Extrapolation of Animal Toxicological
Data to Humans
8.1 Introduction
A full evaluation of the health effects of ozone (O3) requires an integrated
interpretation of human clinical, epidemiological, and animal toxicological studies. Each of
these three research approaches has inherent strengths and limitations. Animal toxicological
data are valuable because they provide concentration- and duration-response information on a
fuller array of effects and exposures than can be studied in humans. However, historically, use
of animal toxicological data has been limited because of difficulties in quantitative
extrapolation to humans. Recent advances in the state-of-the-art of extrapolation have reduced
several uncertainties, which will be discussed in this chapter.
Qualitative animal-to-human extrapolation generally is accepted because O3 causes
similar types of effects in several animal species, from mouse to nonhuman primate
(Chapter 6). Also, when similar endpoints (e.g., inflammation and pulmonary function) have
been examined in O3-exposed animals and humans, similar effects are observed. However,
quantitative extrapolation (i.e., if a certain exposure causes a specific effect in animals, what
exposure is likely to cause that same effect in humans?) is the goal but is controversial. Such
an extrapolation requires an integration of dosimetry and species sensitivity. Dosimetry is
defined as the dose delivered to a site in the respiratory tract (RT). As can be seen in Section
8.2, substantial information is available on dosimetry in several species, including humans.
Dosimetric studies that are referenced in the earlier O3 criteria document (U.S. Environmental
Protection Agency, 1986) are summarized only briefly here; newer research is the focus.
Species sensitivity, discussed in Section 8.3, refers to the sensitivity of a specific species to the
delivered dose. For example, even if the same dose of O3 were delivered to a specific
respiratory tract site in rats and humans, differences in species sensitivity to that dose are likely
because of variations in defense mechanisms and perhaps other factors. Section 8.3 also
provides a more holistic approach to extrapolation by quantitatively comparing exposure-
response data obtained in animals and humans. Sections 8.4 and 8.5 are intended to draw the
forgoing information together, reaching conclusions about the potential for acute and chronic
human health effects based on animal studies. Lastly, Section 8.6 presents the summary and
major conclusions from the chapter.
Although this chapter focuses on animal-to-human extrapolation, dosimetric studies
also can be used to elucidate interpretations of the human studies described in Chapter 7. For
3-1
-------
example, knowledge of dosimetry in humans as related to age and exercise can enhance
understanding of human susceptibility factors.
8.2 Ozone Dosimetry
8.2.1 Introduction
Dosimetry refers to measuring or estimating the quantity or rate of a chemical
absorbed by target sites within the RT. The compound most directly responsible for toxic
effects may be the inhaled gas O3 or its chemical reaction products. Complete identification of
the actual toxic agents and their integration into dosimetry are complex issues that have not
been resolved. Thus, most dosimetry investigations are concerned with the dose of the
primary inhaled chemical. In this context, a further confounding aspect can be the units of
dose (e.g., mass retained per breath, mass retained per breath per body weight, mass retained
per breath per respiratory tract surface area). That is, when comparing dose between species,
what is the relevant measure of dose? This question has not been answered; units are often
dictated by the type of experiment or by a choice made by the investigators.
Experimental and theoretical (dosimetry modeling) studies are used to obtain
information on dose. Experiments have been carried out to obtain direct measurements of
absorbed O3 in the RT, the upper RT (URT; region proximal to the tracheal entrance), and the
lower RT (LRT; region distal to tracheal entrance); however, experimentally obtaining
dosimetry data is extremely difficult in smaller regions or locations, such as specific airways or
the centriacinar region (CAR; junction of conducting airways and gas exchange region), where
lesions caused by O3 occur (see Chapter 6, Section 6.2.4). Nevertheless, experimentation is
important for determining dose, making dose comparisons between subpopulations and
between different species, assessing hypotheses and concepts, and validating mathematical
models that can be used to predict dose at specific respiratory tract sites and under more
general conditions.
Theoretical studies are based on the use of mathematical models developed for the
purposes of simulating the uptake and distribution of absorbed gases in the tissues and fluids of
the RT. Because the factors affecting the transport and absorption of gases are applicable to all
mammals, a model that uses appropriate species or disease-specific anatomical and ventilatory
parameters can be used to describe absorption in the species and in different-sized, aged, or
diseased members of the same species. Importantly, models also may be used to make
interspecies and intraspecies dose comparisons, to compare and reconcile data from different
experiments, to predict dose in conditions not possible or feasible experimentally, and to better
understand the processes involved in toxicity.
8.2.2 Summary of 1986 Review of Experimental and Theoretical Dosimetry
A summary of the more relevant experimental and theoretical dosimetry studies
contained in the previous O3 criteria document (U.S. Environmental Protection Agency, 1986)
is presented. The reader is referred to the earlier document for completeness.
Experiments on the nasopharyngeal removal of O3 in laboratory animals suggested
that the fraction of O3 uptake depends inversely on flow rate (Yokoyoma and Frank, 1972),
uptake was greater for nose than mouth breathing (Yokoyoma and Frank, 1972), and tracheal
and chamber concentrations were related linearly (Yokoyoma and Frank, 1972; Miller et al.,
1979). Only one investigation measured uptake by the LRT, finding 80 to 87% uptake by the
LRT of dogs (Yokoyoma and Frank, 1972). At the time, there were no reported results for
8-2
-------
human URT or LRT uptake. With the exception of two relatively crude studies by Clamann
and Bancroft (1959) and Hallett (1965), there were no data on O3 uptake in humans at the time
of the earlier criteria document (U.S. Environmental Protection Agency, 1986).
Several mathematical dosimetry models were developed to simulate the processes
involved in O3 uptake and to predict O3 uptake by various regions and sites within the RT.
The model of Aharonson et al. (1974) was used to analyze nasopharyngeal uptake data.
Applied to O3 data, the model indicated that the average mass transfer coefficient of this region
and the mass retained increased with increasing air flow, but the percent uptake decreased.
Models were developed to simulate LRT uptake (Miller et al., 1978, 1985). The
models were very similar in their treatment of O3 in the airways and airspaces and in their use
of morphometric data to define the dimensions of the air compartments and liquid lining. Both
the 1978 and the 1985 models of Miller and co-workers took into account reactions of O3 with
constituents of the liquid lining. However, these models differed in their treatment of chemical
reactions in the liquid lining, and the later model included transport and chemical reactions
within tissue and blood, whereas the first model did not (an instantaneous reaction at the
liquid-tissue interface was assumed, so the O3 concentration was defined as zero). In both
models, tissue dose was defined as the O3 flux to the liquid-tissue interface. Both models
predicted O3 tissue dose to be relatively low in the trachea, to increase to a maximum in or
near the CAR, and then to decrease distally; this was characteristic for both the animal and the
human simulations (Miller et al., 1978, 1985).
Prior to 1986, there were no experimental results that were useful in judging the
validity of the modeling efforts. However, a comparison of the results of Miller and
co-workers with morphological data (showing the CAR to be most affected by O3; see
Chapter 6, Section 6.2.4) indicated qualitative agreement between the site of predicted
maximum tissue dose and the site of observed maximum morphological damage in the
pulmonary region.
8.2.3 Experimental Ozone Dosimetry Data
8.2.3.1 Introduction
Models of O3 uptake in the RT have reached a scale of sophistication that provide
some highly specific predictions regarding the location and magnitude of O3 dose. However,
before these models can be exploited to their fullest degree in extrapolating dose within and
between species, validation of the models with experimental data is essential. This section will
review the experimental database on which the modeling of O3 dosimetry is both based and
validated. This will help facilitate discussion of the models themselves in subsequent sections.
Table 8-1 provides a summary of all post-1986 experimental O3 dosimetry studies.
8.2.3.2 In Vivo Ozone Dosimetry Studies
The model predictions of Overton et al. (1987), based on the original model of
Miller et al. (1985), provided specific predictions about the regional and total uptake
efficiencies of O3 in laboratory rats. It was therefore necessary to test these predictions with
actual data. The first data on total RT uptake of O3 in rats were obtained by Wiester et al.
(1987). Ozone uptake was measured in 30 awake, unanesthetized Sprague-Dawley (S-D) rats
receiving a nose-only exposure. Each rat was situated within a plethysmograph that
continuously monitored the animal's breathing pattern. Air with O3 flowed by rats' noses at
1,200 mL/min for 1 h at a concentration of 0.3, 0.6, or 1.0 ppm O3. Determination of RT
-------
Table 8-1. Experimental Studies on Ozone Dosimetrf
oo
Type of Study
In vivo,
nose only
In vivo
In vivo
In vivo
In vivo
In vitro
Species
(Strain)
Rat
(S-D)
Rat (F344)
Rat (S-D)
Rat
(Long-Evans)
Guinea pig
(Hartley)
Rat (F344)
Human
Human
Pig
Sheep
Uptake Breathing Patterns
Total RT V = 2.8 mL
f = 150 bpm
VE = 400 mL/min
Total RT For rats:
VT = 2.6 mL
f = 120 bpm
VE = 330 mL/min
For guinea pigs:
VT = 2.4 mL
f = 77 bpm
VE = 188 mL/min
Total RT, head, VT = 2.05 mL
larynx/ trachea, f = 150 bpm
lung VE = 290 mL/min
Total RT, URT, VT = 800 mL
LRT f = 12 and 24 bpm
V, = 350 and 634 mL/s
oral, nasal, and oronasal
breathing
Total RT, URT, VT = 1,239-1,650 mL
LRT f = 25-35 bpm
VE = 41 L/min
Trachea V = 50:200 mL/s
Results
Uptake measured in 30 rats exposed to 0.3, 0.6, or 1.0 ppm O3 for 1-h.
Uptake measured using mass balance. Total RT uptake efficiency
measured at 40%. Uptake efficiency was independent of
O3 concentration.
All uptake measurements at 0.3 ppm O3. In addition F344 rats were
measured at 0.6 ppm. Uptake measurements made with system of
Wiester et al. (1987). Total RT uptake efficiency averaged 47% and was
strain and species independent. Uptake efficiency again was shown to be
independent of O3 concentration in F344 rats.
Regional uptake measured by assaying for recoverable 18O from
respiratory tract tissue after animal inhaled 18O-enriched 1 ppm O3 for
2 h. Fifty-four percent of inspired O3 was taken up by total RT. Of the
O3 taken up, 49.6% taken up by the head, 6.7% by the larynx/trachea,
and 43.6% by the lungs.
Uptake efficiencies of URT, LRT, and total RT measured by sampling
inspired and expired air from catheter inserted through nose to posterior
oropharynx. Uptake efficiencies computed from peak plateau
concentrations on inspiration and expiration. Inspiratory URT uptake
efficiencies averaged 40%; inspiratory plus expiratory LRT uptake
efficiencies averaged 92%. Small but significant decreases in URT and
LRT uptake efficiencies with increasing f . No effect of concentration on
uptake. Uptake efficiency of mouth relatively greater than nose by about
10%.
Twenty healthy male subjects exposed to 0.4 ppm O3 while exercising at
VE of 41 L/min for 1 h. Uptake efficiencies of URT, LRT, and
total RT measured at beginning and at end of exposure by method of
Gerrity et al. (1988). Subjects mouth-breathed only. Uptake efficiencies
computed as mass fractions. URT inspiratory efficiency was 40% and
did not change during exposure. LRT efficiency dropped from 68 % to
62% during exposure. LRT decrease correlated with drop in VT.
Cumulative dose of O3 to LRT was predictive of VT drop.
Unidirectional O3 uptake efficiencies of trachea decreased with increasing
flow from 0.5 to 0. 15 for the sheep and 0. 12 for the pig. Mass transfer
coefficients generally were independent of flow.
Reference
Wiester et al. (1987)
Wiester et al. (1988)
Hatch et al. (1989)
Gerrity et al. (1988)
Gerrity and
McDonnell (1989);
Gerrity et al. (1989);
Gerrity et al. (1994)
Ben-Jebria et al.
(1991)
-------
Table 8-1 (cont'd). Experimental Studies on Ozone Dosimetrf
oo
Type of Study
In vitro
In vivo
In vivo
In vivo
In vitro
In vivo
In vivo
Species
(Strain)
N/A
Human
Human
Human
Rat (S-D)
Human
Human
Uptake Breathing Patterns
N/A N/A
20-200 mL depth VT = 500 mL
into total RT V, = 250 mL/s
20-200 mL depth VT = 500 mL
into total RT V, = 150, 250, 500, 750,
1,000 mL/s
20-200 mL depth VT = 500 mL
into total RT V, = 250 mL/s
Lung V = 2.71 mL
f = 50-103 bpm
FRC = 4 and 8 mL
Total RT, URT, VT = 810 mL
trachea, mainstem f = 12 bpm
bronchi V, = 320 mL/s
Total RT V = 593-642 mL
f = 16 bpm
VE = 9.2-9.8 L/min
Nasal and oral
breathing
Results
Ozonolysis studies on various unsaturated fatty acids, rat
erythrocyte ghost membranes and rat BAL. Dominant processes are
the production of aldehydes and peroxides due to reactions between O3
and olefins.
Uptake efficiencies by measuring recovery of O3 boluses delivered
at 20 mL increments into lung to depth of 200 mL. At deepest depth,
only 6% of O3 could be recovered. Ozone uptake by conducting
airways larger than predicted by Miller et al. (1985).
Same technique as Hu et al. (1992b), but investigating flow effects.
Increasing flow caused marked shift of delivered O3 toward the
periphery of the conducting airways (i.e., the greater the inspiratory
flow, the greater the amount of O3 delivered to the lung periphery),
where it is available for absorption. Mass transfer coefficients in
upper airways independent of flow, but in conducting airways they
increase proportional to flow. Lung liquid lining mass transfer
coefficient computed to be 1.4 cm/s in the URT, falling to
0.17 cm/s in the respiratory airways. Reaction rate constant
between O3 and the lung liquid lining was computed as 7.3 x lOVs in
the URT, falling to 8.2 x 105/s in the distal conducting airways.
Comparison of O3 bolus uptake between oral and nasal routes.
Nose was found to be 30 % more efficient at removing O3 from the air
stream than the mouth.
Perfused and nonperfused rat lungs ventilated with 1 ppm Q3. Uptake
efficiency of lungs dropped from 95% at 50 bpm to about 50% at 103
bpm. No change in uptake efficiency when lungs inflated from FRC
= 4 to 8 mL.
O3 uptake efficiencies in conducting airway structures determined
by sampling air from anatomical sites ranging from the vocal cords to
bronchus intermedius in 10 subjects undergoing transnasal
bronchoscopy while being exposed to 0.4 ppm O3. Total RT uptake
efficiency was 91%. Uptake efficiencies of mouth-vocal cords:
17.6%; vocal cords-upper trachea: 12.8%; upper trachea-main
bifurcation carina: 11.5%; main bifurcation carina-bronchus
intermedius: 0%.
Total RT uptake efficiency measured using the same technique
(Wiester et al., 1987; Wiester et al., 1988) applied to rats. During
nasal breathing, total RT uptake efficiency was 73%. During oral
breathing, total RT uptake efficiency was 76% and significantly higher
than with nasal breathing.
Reference
Pryor et al. (1991)
Hu et al. (1992b)
Hu et al. (1994)
Kabel et al. (1994)
Postlethwait et al. (1994)
Gerrity et al. (1995)
Wiester et al. (1996)
"See Appendix A for abbreviations and acronyms.
-------
O3 uptake was determined by mass balance. The uptake was the difference in mass in the
upstream air and in the air downstream of the rat. Total RT O3 uptake efficiency of
approximately 40% was measured and was independent of O3 concentration. During data
acquisition, the animals had an average tidal volume (VT) of about 2.8 mL, an average
breathing frequency (f) of about 150 breaths per minute (bpm), and an average minute
ventilation (VE) of about 400 mL/min. This study was followed by another (Wiester et al.,
1988) in which total RT uptake was measured in three strains of rats and in the guinea pig.
Specifically, Fischer 344 (F344), S-D, and Long-Evans rats and Hartley guinea pigs were
exposed for 1 h to 0.3 ppm O3; F344 rats also received a 0.6-ppm exposure. All rats in the
Wiester et al. (1988) study had f s between 112 and 132 bpm, VTs between 2.4 and 2.8 mL,
and VEs between 299 and 364 mL/min. The guinea pigs had a VT of 2.4 mL (not different
from rats), an f of 77 bpm, and a VE of 188 mL/min. Uptake was measured as in the previous
experiment (Wiester et al., 1987). Total RT uptake of O3 was species-independent and
averaged 47%; this was higher than in the previous study because a different calculation
method for fractional uptake was used. As in the first experiment, exposure concentration did
not affect uptake. Wiester et al. (1987) corrected all flows used in uptake calculations for body
temperature and relative humidity. In their later work, however, they found that this
correction was not warranted, resulting in a slightly higher computed O3 uptake efficiency
(Wiester etal., 1988).
These data are in disagreement with the model predictions of Overton et al. (1987),
who made predictions of O3 uptake in two different rat anatomical lung models (Kliment,
1973; Yeh et al., 1979). Simulations were conducted in both anatomical models, varying f and
VT at fixed VE; functional residual capacity (FRC) also was varied as a fraction of total lung
capacity (TLC). The Kliment (1973) anatomical lung model gave consistently high predictions
for uptake when compared with actual data. The predictions using the Yeh et al. (1979) model
came closer. Total RT uptakes (not including the head) for the Yeh et al. (1979) model were
predicted to range between 46 and 60% at f = 154 bpm and VT = 1.25 mL and between
70 and 80% for f = 81 bpm and VT = 2.4 mL. However, when the fact that these predictions
do not include the head of the animal is considered, it is evident that the model predictions
overestimate the total RT uptake in rats. The question is whether the measurements are
accurate or whether there is a problem with the model formulation. The data of Postlethwait et
al. (1994), presented below, suggested that the data of Wiester et al. (1987, 1988) may be
reasonable. The Postlethwait et al. (1994) data in the excised rat lung suggested a clear
inverse dependence of lung uptake on f. At a VT of 2.5 mL, the O3 uptake efficiency of the
excised lung fell from nearly unity at f = 50 bpm to almost 50% at f = 100 bpm. For
extrapolation purposes, a key question here is what should be considered the normal resting f
of a rat. Although Wiester et al. (1987, 1988) allowed their rats to acclimate to the
plethysmograph by monitoring f and began uptake measurements only after f had plateaued at a
minimum, it is still uncertain whether fs of 120 to 150 bpm are reasonable. In a summary of
studies of the pulmonary function of rats in response to O3, Tepper et al. (1993) found typical
fs of 100 bpm. Although the models appear to overestimate the O3 uptake efficiency of the rat
RT, the discrepancy is not large, and the near agreement indicates that the O3 dosimetry
models have predictive capability.
In addition to data on the total RT O3 uptake efficiency, in vivo data on regional
O3 dosimetry in animal models have begun to emerge. Hatch and Aissa (1987) and Aissa and
Hatch (1988) first described a method to measure O3 uptake in animals by exposing them nose-
8-6
-------
only, while in a plethysmograph, to O3 enriched with 18O, a stable isotope of oxygen. After
exposure, bronchoalveolar lavage (BAL) fluid and respiratory tract tissue were assayed for
excess 18O using isotope ratio mass spectrometry. One problem with this technique is that all
of the absorbed 18O cannot be accounted for, thus possibly leading to an underestimation of
dose. The technique used by Hatch and Aissa (1987) involves the detection of excess 18O in
18O3 reaction products after tissue pyrolysis. Thus, 18O3 that is degraded to H218O or 18O2 is
lost and cannot be detected in dry tissue. Eight male F344 rats (four previously exposed
chronically to O3 for 1 year to an urban pattern generally consisting of a 0.06-ppm baseline
with 1-h daily spikes rising to 0.25 ppm) were exposed for 2 h to 1 ppm of 18O-enriched O3
(Hatch et al., 1989). During exposure, the rats breathed at 150 bpm and had a VT of 2.05 mL
and a VE of 290 mL/min. After exposure, the lung, trachea, and head were analyzed
separately for 18O. Overall, the animals took up 54.3% of inspired O3. Although this value of
O3 uptake efficiency is higher than that found by Wiester et al. (1987, 1988), considering the
fact that the coefficient of variation for O3 uptake efficiency measurements is around 20% in all
studies, the result of Hatch and Aissa (1987) is consistent with the data of Wiester et al. (1987,
1988). Of the O3 taken up by the animals, 49.6% was taken up by the head, 6.7% by the
larynx/trachea, and 43.6% by the lungs. By assuming equal uptake efficiencies by
compartments on inspiration and expiration, inspiratory uptake of O3 by these regions was
computed. It was determined that the rat nasopharynx (NP) had an inspiratory efficiency of
17.4%, and that the larynx/trachea removed 2.7% of the remaining O3.
This technique recently has been extended to humans. Hatch et al. (1994) showed
that when human subjects were exposed to 0.4 ppm 18O3 while exercising intermittently at
VE = 60 L/min for 2 h, the amount of recovered 18O in lavagable cells indicated that the
human cells incorporated 4 to 5 times the O3 dose (i.e., concentration of 18O3) that was
incorported by the BAL cells from rats exposed to 0.4 ppm O3 for 2 h at rest. Consequently,
to compare absorbed 18O3 doses between rats and humans using BAL requires the assumption
that the amount of lavagable cell membrane available to react with 18O3 is comparable between
the two species. The difference between rats and humans could be accounted for by the fact
that the humans were exercising, whereas the rats were not. However, as was noted above,
not all absorbed 18O3 can be accounted for.
8.2.3.3 In Vitro Ozone Dosimetry Studies
The use of whole, intact animals to study O3 uptake is needed to ascertain the actual
amounts of O3 absorbed. However, it is also important to understand some of the more
fundamental processes governing O3 uptake, such as the biochemistry of O3/liquid and
O3/tissue interactions to determine chemical reaction rates essential to O3 dosimetry models.
Furthermore, the use of intact animals does not allow more precise determinations of the role
of physiological parameters on O3 uptake. For this reason, there have been some limited
attempts at utilizing animal tissue explants and whole lungs to study O3 uptake.
Ben-Jebria et al. (1991) studied O3 uptake by the trachea of sheep and pigs to
investigate mass transfer coefficients. Ozone boluses of 1 ppm were passed through excised
tracheae. Tracheae were obtained from a slaughter house 0.5 to 2 h after slaughter, and,
although they were kept coated with physiologic saline, they were not maintained at body
temperature, possibly resulting in underestimation of in vivo uptake. The lengths and
diameters of the pig tracheae were not too different from human tracheal dimensions. The
flow dependence of mass uptake and the mass transfer coefficient (K) were determined for
3-7
-------
flows between 50 and 200 mL/s. Uptake efficiencies in the pig decreased with increasing flow
from about 0.5 to 0.12 and in the sheep decreased from about 0.5 to 0.15. Mass transfer
coefficients generally were independent of flow (K = 0.5 cm/s in pigs and 0.35 cm/s in
sheep), indicating the lack of dependence of uptake on gas-phase diffusion processes. This
contrasts with the conclusion of Aharonson et al. (1974) for the NP of dogs where the
investigators observed that the slight inverse dependence of uptake on flow observed by
Yokoyama and Frank (1972) leads to the conclusion that the mass transfer coefficient for the
NP of the dog should increase with flow, suggesting a role of the boundary layer in limiting
diffusion of O3 to the wall of the NP. The different geometries of a trachea and an NP may
account for the differing observations.
A significant feature of the Ben-Jebria et al. (1991) study was the use of a rapidly
responding O3 analyzer. In order to conduct their O3 uptake studies, Ben-Jebria and Ultman
(1989) and Ben-Jebria et al. (1990) developed a rapidly responding O3 analyzer. The analyzer
relies on the reaction of O3 with alkenes such as ethylene, propylene, cyclohexane, etc. Ten
alkenes were tested. Ninety percent step-response times of 130 to 540 ms were achieved with
varying degrees of linear response with O3 concentration. The authors concluded that the best
alkene was 2-methyl-2-butene, with optimum 10 to 90% responses of 110 ms and minimum
detectable limits of O3 of 18 ppb. Interference with CO2, however, was found, requiring
measurements of CO2 to correct the analyzer response.
Postlethwait et al. (1994) used an isolated rat lung preparation to investigate the
effects of vascular perfusion, inspired dose, temperature, and distal lung surface area on
O3 absorption by the LRT. Vascular perfusion had little or no effect on uptake efficiency of
O3. When the lung was exposed to 1 ppm O3 and ventilated with a VT of 2.5 mL, a FRC of
4 mL, and an f of 50 bpm, uptake efficiency was 95%. As f increased with fixed VT, uptake
efficiency began to drop, reaching nearly 50% at an f of 100 bpm. When the lung temperature
was dropped from 37 to 25 DC, uptake efficiency dropped from 95 to 85% at 50 bpm. This
drop was exploited to investigate other factors (such as flow, volume, and lung surface area)
governing uptake because it moved respiratory tract uptake further below 100%. The
observation of a dependence of uptake on temperature indicates that uptake efficiency is
chemical-reaction dependent, thus possibly coupling uptake to reaction product formation.
Another interesting result from this study was the lack of dependence of uptake on FRC.
When FRC was doubled from 4 to 8 mL at 25 DC, fractional O3 uptake was unchanged. This
latter result suggests that O3 uptake is virtually complete by the time O3 reaches the alveolar
spaces of the lung. Otherwise it would have been expected that the uptake efficiency would
have risen with increased FRC.
To further investigate the reactions of O3 with the lung, Pry or et al. (1991)
performed ozonolysis studies of various unsaturated fatty acids (UFAs), rat erythrocyte ghost
membranes, and rat BAL. These studies demonstrated significant production of hydrogen
peroxide and aldehydes and that production of hydrogen peroxide was due primarily to
reactions between O3 and olefins. The authors concluded that the reaction of O3 with UFAs in
the lung fluid lining and cell membranes produce hydrogen peroxide and aldehydes that may be
important mediators in the toxicity of O3. The quantitative results of these studies led Pry or
(1992) to hypothesize about the degree to which O3 reacts with the liquid lining of the lung and
with lung tissue. A simple model calculation was performed using the Einstein-Smoluchowski
equation to estimate the half-life of O3 in bilayers and cell membranes. Pry or (1992)
concluded that a substantial fraction of O3 reacts in the bilayer, and that only in regions of the
-------
lung where the lung lining fluid layer is less than 0.1 Dm thick will O3 penetrate to tissue, and
only then will O3 react in cell membranes before penetrating further. The overall conclusion is
that the toxic effects of O3 may be mediated not just by O3 directly but by reactive
intermediates such as aldehydes and hydrogen peroxide. This raises the question as to the
relevant dose of O3: is it the total dose, the dose to the liquid lining, the tissue dose, or the
dose of reactive intermediates delivered to tissue?
8.2.3.4 Human Ozone Dosimetry Studies
Significant progress has been made in the area of human O3 dosimetry since the
previous criteria document (U.S. Environmental Protection Agency, 1986). Studies have been
conducted defining total and regional respiratory tract uptake, the dependence of uptake on
physiological parameters, and the role of uptake in modulating response.
Gerrity et al. (1988) reported on measurements of O3 uptake by the extrathoracic
airways (airways proximal to the posterior oropharynx) and intrathoracic airways (airways
distal to the posterior oropharynx) in 18 healthy, young male volunteers. Ozone uptake was
measured by placing a small polyethylene catheter through the nose and positioning the distal
tip in the posterior oropharynx. Breath-by-breath samples of O3 were collected, and the peak
plateau concentrations were compared with chamber concentrations. The effects on uptake of
O3 concentration (0.1, 0.2, 0.4 ppm), f (12 and 24 bpm at fixed VT), and mode of breathing
(oral, nasal, and oronasal) were tested. The O3 analyzer had a moderately rapid response with
a 90% response time of 700 ms. Inspiratory VT ranged between 754 and 848 mL; mean
inspiratory flow at 12 bpm was 350 mL/s; at 24 bpm it was 634 mL/s. The authors measured
extrathoracic uptake efficiency of O3 on inspiration at approximately 40% and intrathoracic
uptake efficiency (inspiration plus expiration) at approximately 92%. They essentially found
no effect of O3 concentration on uptake (intrathoracic uptake was significantly higher at 0.4
ppm, but the difference was very small). They did find that both intrathoracic and
extrathoracic uptake decreased with increasing f (at fixed VT), falling by about 7% for
extrathoracic uptake and by about 3% for intrathoracic uptake when f increased from 12 to
24 bpm. The finite response time of the analyzer may have affected the results at the 24 bpm f
by overestimating extrathoracic uptake and underestimating intrathoracic uptake. However,
because uptake was defined relative to plateaus of concentration, the response time of the
analyzer was adequate to reach a plateau in the 1.2-s inspiratory time at 24 bpm. It is
important to note here that, when utilizing the data from this study to compare with other
studies and models, the uptake efficiencies measured are comparable to steady-state
unidirectional measurements of uptake. Another feature is that Gerrity et al. (1988)
consistently measured a small, non-zero plateau of O3 on expiration. This plateau is not
consistent with the suggestion from models of O3 uptake and the data of Postlethwait et al.
(1994) (nor with the later work of Gerrity et al. [1995] that is presented below) that no
O3 should be washed out from lung volumes beyond the conducting airways. This observation
of Gerrity et al. (1988) may have been an artifact of the manner in which O3 was measured by
sampling from the posterior oropharynx. There may have been entrainment of O3 in the
pharyngeal airspaces that was washed out after expiration of dead-space air. Regardless, the
concentration of O3 exhaled from the alveolar phase of washout was very low.
One of the most startling results from the work of Gerrity et al. (1988) was the
finding that there was only a small, but statistically significant, difference between uptake by
the nose and by the mouth. The mouth had approximately 10% greater uptake efficiency than
8-9
-------
the nose. The combined oronasal passage had an uptake efficiency greater than the nose by
another 8 %. This suggests that persons who breathe nasally are at no less risk than persons
who breathe oronasally. Adams et al. (1989) investigated this possibility by comparing
functional responses in subjects acutely exposed to O3 while breathing either orally or
oronasally. Healthy subjects were exposed on five separate days to 0.4 ppm O3. In the first
four exposures, subjects were exposed by face mask (with or without nose clip) for 30 min at
an exercise level of 75 L/min or for 75 min at exercise level of 30 L/min. The fifth exposure
was for 30 min at 75 L/min, with exposure through a mouthpiece. There were no differences
in pulmonary function response (forced expiratory volume in 1 s [FEV,], forced vital capacity
[FVC], or forced expiratory flow) with face-mask exposure among all experimental groups
(i.e., no nose clip, VE, or time effect). Pulmonary function response was, however, greater
with a mouthpiece. Adams et al. (1989) speculated that the greater response with the
mouthpiece was due to O3 scrubbing by the face mask or by facial hair. It also may have been
due to different oral configurations imposed by a mouthpiece. Hynes et al. (1988) also
investigated whether functional responses were affected by the mode of breathing. Healthy
subjects were exposed to 0.4 ppm O3 for 30 min in an exposure chamber. On two different
occasions, each subject breathed either through the nose or the mouth exclusively. There was
no difference in pulmonary function response between these two routes of exposure. Taken
together, the studies of Adams et al. (1989) and Hynes et al. (1988) are consistent with the
observations of Gerrity et al. (1988) on the equal efficiency of all routes of breathing for
extrathoracic O3 scrubbing.
This study was followed by another study (Gerrity and McDonnell, 1989; Gerrity
et al., 1989, 1994) in which the relationship between O3 uptake and functional response was
investigated. Healthy subjects were exposed to 0.4 ppm O3 for 1 h while exercising
continuously at 40 L/min. Ozone uptake was measured at the beginning and at the end of
exposure while the subjects were still exercising, using the technique of Gerrity et al. (1988).
In contrast to the work of Gerrity et al. (1988), uptake was computed in this study by
integrating concentration times flow instead of using peak plateau measurements. Also, in this
work, the 90% response time of the analyzer was 1.2 s (compared with 0.7 s in the previous
work). The authors found that about 40% of the inspired O3 was taken up by the URT (i.e.,
the same as the extrathoracic airways described in Gerrity et al., 1988) during inspiration, and
that this did not change during exposure. Total RT uptake efficiency was approximately 80%,
and it did not change during exposure. However, LRT (i.e., the intrathoracic airways
described in Gerrity et al., 1988) uptake efficiency fell during exposure from 68 to 62% and
was correlated with the O3-induced fall in VT (VT fell from 1,650 to 1,239 mL; f increased
from 25.2 to 34.8 bpm; inspiratory flow fell from 1,506 to 1,357 mL/s; VE increased slightly
from 40.8 to 40.9 L/min), suggesting that the VT reduction may have a protective effect on
dose delivered to the periphery of the lung. It is not likely that the finite analyzer response
time affected uptake measurements. Evidence of this is the lack of dependence of URT uptake
on changes in VT or f. The low values for uptake in the LRT may have been due to an artifact
from the relatively slow response time of the analyzer, which was approximately equal to
inspiratory and expiratory times. As a check on their results, the authors compared their data
with the previous work of Gerrity et al. (1988) by computing uptake by the original technique
using peak plateau concentrations. When that was done, the URT uptake efficiencies were 17
and 22% at the beginning and end of exercise, respectively, and the LRT efficiencies were 96
and 92% at the beginning and end of exercise, respectively. The URT change computed this
8-10
-------
way was not significant, but the LRT efficiency drop was. When viewed at in this manner, the
data from this experiment are consistent with those from the previous experiment.
Because there is a need to compare human O3 uptake data with rat O3 uptake data, it
is essential that there be confidence in the reliability of the different approaches. To help
establish the comparability of techniques, Wiester et al. (1996) measured total RT uptake in
humans using a similar, although obviously scaled-up, system to that used for rats (Wiester
et al., 1987, 1988). Healthy subjects breathed 0.3 ppm O3 while seated in an exposure
chamber; their faces were placed in a sealed face mask. The face mask was attached to a large
tube through which chamber air was circulated with a pump at a rate of D40 L/min. Upstream
and downstream O3 concentrations were measured continuously, as was ventilation with an
induction plethysmograph. Subjects breathed at rest, either through nose or mouth (average f
= 16 bpm, VT = 598 to 642 mL, VE = 9.2 to 9.8 L/min). While nose breathing, 73% of
inspired O3 was taken up by the total RT, and, while mouth breathing, 76% of inspired O3 was
taken up, which was significantly higher than that found with nose breathing. This difference
is probably not, however, biologically significant. Significant negative correlations between f
and uptake in both mouth- and nose-breathers were found, similar correlations were found with
VE, but no correlations were found between uptake and VT or any other measure of breathing
pattern or pulmonary function.
The observations in Wiester et al. (1996) of a slight increase of total RT uptake
efficiency with oral breathing and the inverse correlation of total RT uptake efficiency with
f are consistent with those of Gerrity et al. (1988). Furthermore, the data on total RT uptake
are consistent overall with that of Gerrity et al. (1988, 1994). The data from Gerrity et al.
(1988) reporting total RT uptake efficiencies of about 95% were based on minimum plateau
measurements, thus reflecting uptake during steady-state flow conditions, as opposed to the
cyclical conditions of actual breathing. The data of Gerrity et al. (1994), on the other hand (in
which total RT efficiencies of 80% were reported), were obtained by integrating the product of
concentration and flow, thus more accurately reflecting the actual mass uptake of O3 during
cyclical breathing when Gerrity et al. (1994) computed uptake using the methodology from
Gerrity et al. (1988), they found that the total RT uptake measurements were comparable.
Thus total RT mass uptake efficiencies at rest of 80% are not unreasonable.
Hu (1991), Hu et al. (1992b, 1994), and Ultman et al. (1993) took a different
approach to measuring respiratory tract uptake of O3. They exploited the development of a
rapid responding O3 analyzer (Ben-Jebria and Ultman, 1989; Ben-Jebria et al., 1990) to
measure the recovery of small boluses of O3 delivered to different volumetric depths of the RT.
Ozone uptake was measured in a set of four experiments. In the baseline experiments,
absorption of O3 boluses was measured in healthy male subjects at rest. The O3 boluses were
10 mL in size, with a peak concentration of 3 ppm. The O3 analyzer characteristics were
sample flow of 400 mL/min, 2-methyl-2-butene as reactive alkene, 10 to 90% step-response
time of 110 ms, and lower detection limit (18 ppb). In the baseline experiments, the VT was
500 mL, and the inspiratory and expiratory flow rates were 250 mL/s. In a complete set of
measurements, bolus recovery was examined for penetrations of 10 to 200 mL, in 10 mL
increments. In the second set of experiments, the effects of flow were measured by measuring
bolus recovery as a function of penetration depth for flows of 150, 250, 500, 750, and
1,000 mL/s at a fixed VT of 500 mL. In a third set of experiments, bolus recovery was
measured as a function of penetration depth at a flow of 250 mL/s and a VT of 500 mL; the
bolus delivered to a rubber mouthpiece or to a nasal cannula was compared, thereby examining
8-11
-------
potential uptake differences between the two pathways. In the fourth set of experiments, the
effects of O3 concentration on uptake were determined by delivering boluses with peak
concentrations of 0.5, 1.0, 2.0 and 4.0 ppm. The latter experiments were conducted because
acute studies in isolated dog airways showed that absorption efficiency was inversely related to
inhaled concentration between 0.1 and 20 ppm (Vaughan et al., 1969; Yokoyama and Frank,
1972). However, later experiments in guinea pigs, rabbits (Miller et al., 1979), and humans
(Gerrity et al., 1988) showed a lack of concentration dependence, implying a linear
relationship between concentration and dose. The dependence or lack of dependence of uptake
efficiency on O3 concentration provides information on the order of reactions of O3 with lung
fluid lining and tissue. Under steady-state conditions, concentration independence of uptake
efficiency suggests that first-order processes play a role. However, because O3 absorption is
coupled to both interfacial transfer (gas-phase to solute O3) and subsequent reaction, at face
value, the conclusion cannot be reached that saturated absorption rates are solely due to
saturation of the reaction components.
In all four experiments, Hu and colleagues computed the first three moments of the
inspired and expired bolus distributions with respect to volume. The zeroth moment of a bolus
is the O3 mass contained in the bolus. Thus the zeroth moments of the inspired and expired
boluses were used to compute O3 uptake efficiency (D; or absorbed fraction), breakthrough
volume (the mean volume of the exhaled bolus), and bolus dispersion. The first moment of a
bolus is its mean volumetric position. The first moment on inspiration gives the penetration
volume (Vp), and the first moment on expiration gives the breakthrough volume (VB). In the
absence of any O3 uptake, a longitudinally mixed bolus should have VB = Vp. The second
moment of a bolus is its variance. The difference in variance between the expired and inspired
bolus (D2) is a measure of gas mixing, or dispersion, in the lung.
In the baseline experiments, the breathing pattern was a resting pattern with a VT of
500 mL and an average inspiratory flow of 250 mL/s. These experiments were performed on
nine male subjects and showed that almost all O3 was absorbed beyond a penetration depth of
180 mL. Only about 6% of inhaled O3 was recovered at the 180 mL penetration depth, and,
beyond that depth, it was very difficult to obtain an accurate measurement of recovery. The
investigators also found that VB was greater than Vp at penetration depths less than 100 mL,
after which VB leveled out at a constant value. Dispersion was insensitive to penetration depth.
An important finding of the baseline experiment was that at quiet resting ventilation, about
50% of the O3 mass in a bolus inhaled through the upper airways is taken up by the upper
airways. To compare these data with results of Gerrity et al. (1988), it is necessary to assume
that inspiratory and expiratory uptake efficiencies are equal. Then the unidirectional uptake
efficiency of the upper airways to a depth comparable to that at which Gerrity et al. (1988)
positioned their sampling catheter is about 30%, which is approximately 25% less than the
40% results of Gerrity et al. (1988). This difference might be due to the presence of a
mouthpiece in the experiments of Hu and colleagues, which could reduce the uptake efficiency
of the oral pathway. The functional response data of Adams et al. (1989) suggest that this
might be the case.
The flow experiments showed that there was a general right shifting of the D D Vp
curves with increasing flow (i.e., increasing flow causes a deeper penetration of O3 into the
lung with lower fractional uptake by the conducting airways). Eventually, all of the O3 is still
absorbed. Breakthrough volume showed a similar pattern at all flows (i.e., greater than
penetration volume at small Vp but flattening out at larger Vp). As flow increased, the level of
8-12
-------
the plateau increased. Dispersion, although constant as a function of Vp at all flows, increased
linearly with increasing flow.
The studies of Hu and colleagues investigating the role of exposure route in
modulating O3 uptake efficiencies reported that the nose absorbed approximately 30% more
than the mouth. This result is at variance with the findings of Gerrity et al. (1988), which
indicate that there was only a slightly higher uptake by the oral pathway when compared with
the nasal pathway. Gerrity et al. (1988) studied uptake by the two pathways without the use of
a mouthpiece or any other delivery system. Subjects were free to breathe naturally. It is
possible that the use of mouthpieces and nasal canulas in the studies of Hu and colleagues
caused artifacts, resulting in their findings for nasal and oral uptake efficiencies. The study of
Adams et al. (1989) supports this, showing enhanced pulmonary function response to
O3 during a mouthpiece exposure compared to face-mask/oral exposure. Finally, the
concentration-dependence studies showed that uptake efficiency was not affected by the
concentration of inspired O3 between 0.3 to 4 ppm, implying that O3 uptake is governed by
linear processes.
One of the very unique features of the approach to measuring uptake efficiency
taken by Hu and colleagues is that the O3 bolus recovery data can be used to derive local mass
transfer coefficients for the conducting airways. Regional mass transfer coefficients derived
experimentally in this way can then be used as input into mathematical model simulations,
thereby potentially leading to more accurate models of O3 dose to the RT.
Hu and colleagues define the parameter Ka (per second) as one that is suitable to
characterize local O3 absorption. It is the product of the overall K (centimeters per second),
which reflects the combined contribution of diffusion and chemical reaction to uptake, and the
local surface/volume ratio (a; per centimeter). From the D D Vp curves, these investigators
derived values for Ka. Thus, the experimentally derived mass transfer coefficients depended
on assumptions about airway anatomy and morphology. As a result of the various
experiments, these investigators found a number of important properties of Ka:
• The proximal subcompartment of the nose has a Ka that is 70% larger than the
Ka for the proximal mouth compartment (see, however, the comment made above
regarding the nose/mouth differences).
• Ka's in the upper airway compartment were between 1.20 and 2.24/s and
relatively insensitive to flow, indicating that diffusion resistance of O3 through the
gas boundary layer is much less than through the mucus film. These data also are
consistent with the pig and sheep tracheae experiments of Ben-Jebria et al.
(1991).
• In the proximal and distal conducting airway subcompartments, (Ka)"1 was
linearly related to (flow)"1, suggesting that the gas-phase absorption rate constant
is directly proportional to flow. In lower airways, therefore, diffusion resistance
of the gas boundary layer is important. Hu and colleagues concluded that the gas
boundary layer contributes 80 to 90% of the overall diffusion resistance in the
central airway compartment.
• The mass transfer coefficient in the lung liquid lining is estimated to fall from 1.4
cm/s in the URT to 0.17 cm/s in the respiratory airways.
• The chemical reaction rate between O3 and the lung liquid lining was estimated to
be 7.3 x 106/s, 2.3 x 106/s, and 8.2 x 105/s in the upper airways, proximal
conducting airways, and distal conducting airways, respectively.
8-13
-------
• The overall mass uptake coefficients determined in the work of Hu (1991) are
significantly higher than those used in the model of Miller et al. (1985). If the
mass transfer coefficients in the model are adjusted upward, the total uptake
efficiency would be higher than measurements have shown, requiring a
downward adjustment of pulmonary region mass transfer coefficients.
Gerrity et al. (1995) took a somewhat more conventional approach in an attempt to
acquire regional information on O3 absorption in the human RT. Healthy subjects underwent
transnasal bronchoscopy while in an exposure chamber in which 0.4 ppm O3 was present.
Subjects were asked to breathe at 12 bpm. Inspired and expired air was sampled through a
Teflon catheter that had been passed through the biopsy channel of the bronchoscope and
positioned approximately in the center of the airway lumen. The air was drawn into a rapid
response O3 analyzer (Ben-Jebria and Ultman, 1989; Ben-Jebria et al., 1990) with a 90%
response time of 250 mL/s while using ethylene as the reactive alkene. Average VT was
810 mL, and average inspiratory flow was 320 mL/s. Air was sampled for five breaths from
above the vocal cords, at the entrance to the trachea, above the main bifurcation carina, and
midway through the bronchus intermedius. Flow was measured simultaneously by a
pneumotach attached to a simple cylindrical mouthpiece. Before and after each measurement,
a set of samples from the mouth was collected for reference. Uptake was defined as the
fraction of O3 mass lost across any anatomical segment; mass was determined by integrating
the product of O3 concentration and flow. By way of comparison with the other human O3
uptake studies, Gerrity et al. (1994) found that total RT uptake of O3 measured in this fashion
was 91 %. This is higher than the resting data of Wiester et al. (1996); however, the average
VT in this study of 810 mL, compared with the 600 mL VT reported by Wiester et al. (1996)
may account for this difference. When Gerrity et al. (1995) computed the unidirectional
uptake efficiencies between the mouth and the various sampling sites, they found that 17.6,
27.0, 35.5 and 32.5% of the O3 passing into the mouth is taken up by structures up to the
vocal cords, the upper trachea, the main bifurcation carina, and the bronchus intermedius,
respectively. They also computed the unidirectional uptake efficiencies across individual
airway segments: 17.6% between the mouth and just above the vocal cords, 12.8% from
above the vocal cords to the upper trachea, 11.5% from the upper trachea to the main
bifurcation carina, and essentially zero between the carina and bronchus intermedius. The
uptake between the mouth and just above the vocal cords is considerably lower than that
measured earlier by Gerrity et al. (1988), even considering the fact that, in the earlier study,
peak plateaus were used. As has been noted earlier, it is possible that the mouthpiece played a
role in reducing the uptake efficiency of the mouth. The uptake efficiency of O3 across the
trachea is in line with the data from sheep and porcine tracheae at the higher flow rates (Ben-
Jebria et al., 1991). The present data are also consistent with the bolus uptake data of Hu
(1991) and colleagues, which also were acquired with a mouthpiece. When the O3 bolus data
are used to compute unidirectional uptake efficiencies (assuming that the segmental efficiencies
are the same on inspiration and expiration), the Hu et al. (1992b) data yield uptake efficiencies
of 21, 36, 44, and 46% between the mouth and the vocal cords, the upper trachea, the main
bifurcation carina, and the bronchus intermedius, respectively. The O3 bolus data are,
therefore, in good accord with the data of Gerrity et al. (1995). The measured uptake
efficiencies across airway segments clearly appear to be higher than those predicted by the
model of Miller et al. (1985). If higher uptake coefficients in the conducting airways were
used in the model of Miller et al. (1985), the model would overestimate total RT O3 uptake.
8-14
-------
To adjust for this, pulmonary uptake coefficients would have to be reduced. Unfortunately,
the data of Hu and associates cannot provide information beyond the conducting airways.
Gerrity et al. (1995) also measured O3 washout volumes (i.e., the expired volumes
required to cause a specified drop in O3 concentration). This type of data provides insight into
the location of major sites of O3 uptake. At the mouth, the 90% washout volume was 142 mL,
and, at the upper trachea, the 90% washout volume was 62 mL. By the time the entire
anatomical dead space of the lungs was washed out, the O3 concentration had fallen to zero
(Gerrity et al., 1995). It is unclear whether the absence of recovered O3 after washout of the
conducting airways was due to O3 not penetrating beyond the conducting airways or to all of
the O3 that penetrated beyond the conducting airways being absorbed. The latter possibility is
more likely based on the observations of Hu (1991).
In assessing the work of Gerrity et al. (1995), it is significant to note that these
investigators measured expired plateaus of O3 concentration that were zero. This contrasts
with the earlier work of Gerrity et al. (1988, 1994) in which a non-zero expiratory plateau was
observed. The non-zero expiratory plateau may have been due to a number of factors that are
unclear. Because ethylene was the reacting alkene in all cases, it is unlikely that interference
with other gases such as carbon dioxide (CO2) was responsible. Another possibility is that O3
in the early expiratory phase became entrained in the posterior oropharynx and persisted for
the duration of expiration.
8.2.3.5 Intercomparison of Ozone Dosimetry Studies
The previous sections emphasized the methods and results of individual
experimental studies on O3 dosimetry. This section will focus on comparisons of the in vivo
studies with each other and will draw on these comparisons to arrive at conclusions regarding
the utility of these data for extrapolation purposes. The discussion will be divided into three
sections, focusing on total RT uptake efficiency, unidirectional URT uptake efficiency, and
LRT uptake efficiency.
There are two categories that distinguish various data sets among each other. The
first category is the mouth/nose category listed in the Tables 8-2 to 8-4. Studies indicated by
"M" or "N" were performed with unencumbered breathing by either oral (M) or nasal (N)
breathing. Unencumbered indicates the absence of a mouthpiece or canula. Data listed as
"M/N" are pooled from data encompassing oral and nasal breathing. Data shown as
"Mouthpiece" or "Nasal canula" are acquired using these devices to deliver the O3 to the
animal or human subject, or to measure flow.
The second category is the method used to compute O3 uptake efficiency. There are
essentially two methods. One method, referred to as the steady-state method, relies on
measuring the loss of O3 from a steady air flow moving across an anatomical structure.
An example is the data of Yokoyama and Frank (1972) in which a constant flow of ozonated
air through the URT of a dog was maintained by a tracheal canula attached to a pump. Uptake
efficiency was measured by changes in equilibrium O3 concentration. Another example is the
study of Gerrity et al. (1988), which used the steady-state method by measuring the peak
inspiratory and minimum expiratory O3 concentrations through a catheter in the posterior
pharynx. These measurements were compared to the ambient chamber concentration to obtain
uptake efficiencies of the URT and LRT. The second method is referred to as the non-steady
state method. This method uses the integration of the product of flow and O3 concentration to
compute O3 masses that, in turn, are used to compute
8-15
-------
Table 8-2. Total Respiratory Tract Uptake Datii
Species
Human
Human
Human
Human
Human
Human
Human
Human
Human
Human
Human
Human
Human
Rat (F344)
Rat (S-D)
Rat (Long-Evans)
Rat (F344)
Guinea pig
Mouth/Nose
M
N
M/N
M/N
M
M
M
M
Mouthpiece
M
N
Mouthpiece
Mouthpiece
N
N
N
N
N
Method
Steady
Steady
Steady
Steady
Non-steady
Non-steady
Steady
Steady
Non-steady
Non-steady
Non-steady
Non-steady
Non-steady
Non-steady
Non-steady
Non-steady
Non-steady
Non-steady
VT(mL)
832
754
832
778
1,650
1,239
1,650
1,239
825
631
642
500
1,000
2.8
2.4
2.7
2.6
2.4
Inspiratory
Flow (mL/s)
509
456
350
634
1,360
1,360
1,350
1,360
330
539
514
250
250
12.2
9.6
12.3
11.3
7.5
f(bpm)b
18
18
12
24
25
35
25
35
12
16
16
15
7.5
118
123
132
113
77
F,
0.97
0.96
0.97
0.96
0.88
0.87
0.97
0.95
0.91
0.76
0.73
0.86
0.93
0.44
0.46
0.48
0.54
0.53
Reference
Gerrity et al. (1988)c
Gerrity et al. (1988)
Gerrity et al. (1988)
Gerrity et al. (1988)
Gerrity et al. (1994)c
Gerrity et al. (1994)
Gerrity et al. (1994)
Gerrity et al. (1994)
Gerrity et al. (1995)
Wiester et al. (1996)
Wiester et al. (1996)
Hu et al. (1992b)
Hu et al. (1992b)
Wiester et al. (1988)
Wiester et al. (1988)
Wiester et al. (1988)
Hatch et al. (1989)
Wiester et al. (1988)
"See Appendix A for abbreviations and acronyms. M = mouth exposure by natural breathing N = nasal exposure by natural breathing;
M/N = pooled data from mouth and nasal exposure; Mouthpiece = exposure by mouthpiece; Steady = uptake computed during constant
unidirectional flow; Non-steady = uptake computed by integration during cyclic breathing; F, = total RT uptake.
bf is either measured or is computed from reported flows and volumes.
Total RT uptake reported by Gerrity et al. (1988) and Gerrity et al. (1994) did not include the contribution from URT uptake efficiency
during expiration. The data include an expiratory URT contribution, assuming it equals inspiratory URT uptake efficiency.
uptake. The studies of Wiester et al. (1988) in rodents, and Wiester et al. (1996) in humans
are examples of this technique, as is the study of Gerrity et al. (1994).
Total Respiratory Tract Uptake Efficiency
Table 8-2 provides a summary of in vivo data in all animal species of respiratory
tract O3 uptake efficiency (Ft). The data reported for the studies of Gerrity et al. (1988, 1994)
have been adjusted from the published values to account for the fact that the Ft cited in those
papers did not include uptake in the URT on expiration. To make the adjustment, the URT
uptake efficiency on expiration was assumed to equal the inspiratory uptake efficiency. The
Ft data listed for the study of Hu et al. (1992b) were derived from their bolus recovery data by
integrating the data over the desired VT. Because Hu et al. (1992b) could not recover boluses
from a depth greater than 220 mL, it was assumed that any bolus delivered
8-16
-------
Table 8-3. Unidirectional Upper Respiratory Tract
Uptake Efficiency Data3
Species
Human
Human
Human
Human
Human
Human
Human
Human
Human
Human
Human
Dog (beagle)
Dog (beagle)
Dog (beagle)
Dog (beagle)
Rat (F344)
Guinea pig
Rabbit
Mouth/Nose
M
N
M/N
M/N
M
M
M
M
Mouthpiece
Mouthpiece
Mouthpiece
Nasal canula
Nasal canula
Mouthpiece
Mouthpiece
N
N
N
Method
Steady
Steady
Steady
Steady
Non-steady
Non-steady
Steady
Steady
Non-steady
Non-steady
Non-steady
Steady
Steady
Steady
Steady
Non-steady
Steady
Steady
Inspiratory
Flow (mL/s)
509
456
350
634
1,360
1,360
1,360
1,360
337
250
250
83.3
667
83.3
667
11.3
2.7
16.7
f(bpm)b
18
18
12
24
25
35
25
35
12
15
15
N/AC
N/A
N/A
N/A
113
N/A
N/A
F,m
0.40
0.43
0.41
0.38
0.37
0.41
0.16
0.22
0.18
0.30
0.47
0.72
0.37
0.34
0.12
0.17
0.62
0.41
Reference
Gerrity et al. (1988)
Gerrity et al. (1988)
Gerrity et al. (1988)
Gerrity et al. (1988)
Gerrity et al. (1994)
Gerrity et al. (1994)
Gerrity et al. (1994)
Gerrity et al. (1994)
Gerrity et al. (1995)
Ultman et al. (1993)
Ultman et al. (1993)
Yokoyama and Frank (1972)
Yokoyama and Frank (1972)
Yokoyama and Frank (1972)
Yokoyama and Frank (1972)
Hatch et al. (1989)
Miller et al. (1979)
Miller et al. (1979)
"See Appendix A for abbreviations and acronyms. M = mouth exposure by natural breathing N = nasal exposure by natural breathing;
M/N = pooled data from mouth and nasal exposure; Mouthpiece = exposure by mouthpiece; Nasal canula = exposure by nasal canula;
Steady = uptake computed during constant unidirectional flow; Non-steady = uptake computed by integration during cyclic breathing;
F, = total RT uptake.
bf is either measured or is computed from flows and VT.
CN/A = not applicable.
to a depth greater than 220 mL was absorbed completely. The derivation of Ft from the bolus
data was done for VTs of 500 and 1,000 mL.
To assess the consistency of the data, it is useful to examine it as a function of flow.
Figure 8-1 shows Ft as a function of inspiratory flow for all human studies. The Ft from the
bolus recovery data of Hu et al. (1994) are shown for VTs of 500, 1,000, and 1,500 mL. An
overview of the data suggests that, with respect to Ft, there is good agreement among the
various experimental methods for humans. The data clearly show that Ft decreases with
increasing flow and increases with increasing VT, both of which are qualitatively consistent
with model predictions.
One observation is quite prominent: the rat data of Wiester et al. (1988) and Hatch
et al. (1989) (not shown in Figure 8-1) are considerably lower than the human data. Even if it
is assumed that the rats were breathing up to three times resting ventilation
8-17
-------
Table 8-4. Lower Respiratory Tract Uptake Efficiency Dat
-------
p = 500 ml Vp = 1,000 ml VT = 1,500 ml VT = 825 ml
Huetal. (1994) Gerrity etal. (1995)
VT= 1,239 ml VT = 1,650 ml VT = 631 ml_
Gerrity etal. (1994) Gerrity etal. (1993) Wiester etal. (1995)
* n o
0.9
o
c
CD
'o
it
LU
-------
Natural Natural Canula Canula
Steady State Non-steady State steady State Non-steady State
1.0
0.7
0.5
o
c
^
m
"5.
rv-
n
Beagle
<0
c.
o
N
O
0.2
0.1
.eagle
I
I
I
I
_L
J
0.1 0.2 0.3 0.5 1.0 2.0 3.0
Predicted Resting Flow/Measured Flow
Figure 8-2. Unidirectional uptake efficiency in the upper respiratory tract by the nasal
pathway. The ratio of predicted resting flow to measured flow is a way to
scale flow to allow for interspecies comparisons. The beagle dog data are
from Yokoyama and Frank (1972), the rat data are from Hatch et al. (1989),
the rabbit and guinea pig data are from Miller et al. (1979), and the human
data are from Gerrity et al. (1988) (closed square) and Ultman et al. (1993)
(open diamond). Lines representing predictions for uptake efficiency are
from Gerrity (1989).
(1994), which involved unencumbered breathing. These data generally are lower than the data
of Hu et al. (1994) and of Gerrity et al. (1995). The fact that the Hu et al. (1994) and Gerrity
et al. (1995) data are consistent with each other supports this speculation. This observation
may account for the result of Ultman et al. (1993) that the uptake efficiency of the URT is
greater by the nasal pathway than by the oral pathway, which is counter to the observations of
Gerrity et al. (1988).
Lower Respiratory Tract Uptake
Table 8-4 summarizes the data on the uptake efficiency of the LRT tract (F]rt).
In this discussion, F]rt is the uptake efficiency of the LRT relative to the concentration of
8-20
-------
O3 entering the LRT. The human data of Gerrity et al. (1988, 1994, 1995) and the rat data of
Hatch et al. (1989) include the larynx in the LRT. The beagle dog data of Yokoyama and
8-21
-------
Natural Natural Canula Canula
Steady State Non-steady State Steady State Non-steady State
• •DO
1.0
0.7
0.5
0.3
'o
it
m
-------
an inaccurate definition that could influence greatly the estimation of LRT uptake efficiency.
The coherence of all of the data on F, and Furt by the oral pathway
VT = 500 ml
\T = 825 ml
Gerrityetal. (1995)
•
VT =1,000 mL
Huetal. (1994)
VT =1,239 mL
Gerrityetal. (1994)
•
VT =1,500 mL
\r = 1,650 ml
Gerrityetal. (1994)
D
1.0,-
0.9
o
c
-------
Table 8-5 presents a summary of theoretical studies of the uptake of O3 by the RTs
(or regions) of humans and laboratory animals that have become available since the 1986
review. Although there are 10 investigations listed, there are only five distinct
8-24
-------
Table 8-5. Theoretical Ozone Dosimetry Investigation!?
Species and Region
Modeled/Anatomical
Model
Liquid Lining and Tissue
Transport and Chemical
Reactions
Dosimetry Model /Subject of Investigation
Results/Predictions
Reference
Guinea pig
LRT/Kliment (1973),
Schreider and
Hutchens (1980); rat
LRT/ Kliment
(1973), Yeh et al.
(1979)
Miller et al. (1985) Enhanced Miller et al. (1985). Investigates the
effect on predictions of anatomical models, FRC,
ventilation, and TB liquid lining rate constant.
Simulates O3 uptake in anatomical models of rat
lobes.
With respect to different anatomical models for the same and
different species: qualitative similarity in the shape of net and
tissue dose versus airway number curves, but significant
differences in regional fractional uptakes. Maximum tissue
dose in vicinity of PAR. PAR dose decreases with increasing
time of flight to this region. Maximum tissue dose in the
vicinity of the first pulmonary region segment of anatomical
models.
Overton et al.
(1987)
oo
to
Human, rat, rabbit,
and guinea pig
LRT/not specified
Miller et al. (1985) and
Overton et al. (1987)
Miller et al. (1985) and Overton et al. (1987).
Compares O3 dose profiles of human, rat, guinea
pig, and rabbit. Uses model and experimental
data to estimate O3 dose-response curves for
decrements in FEV, (humans) and for BAL
proteins in rat, guinea pig, and rabbit. Compares
LRT uptake predictions to the human
experimental data of Gerrity et al. (1988).
Similarity among species in the shape of the airway segment Miller et al. (1988)
curve: tissue dose increases distally in the TB region, reaches
a maximum in the first pulmonary region segment for human,
rat, and guinea pig and in the last TB segment of the rabbit,
and then decreases distally in the pulmonary region.
Predictions of uptake distal to the oropharynx are in agreement
with Gerrity et al. (1988).
Rat total RT/URT:
Schreider and Raabe
(1981); TB: Uses rat
cast data (Raabe
etal., 1976) to define
TB region paths;
PUL: Yeh etal.
(1979) for model of
the acinus
TB and pulmonary region
mass transfer coefficients
based on Overton et al.
(1987)
Enhanced Overton et al. (1987). Illustrates dose
distribution along the longest and shortest (as
defined by time of flight) paths from the trachea
to the most distal alveoli.
Threefold difference in PAR doses of the shortest and longest Overton et al.
paths from trachea to PAR. Dose distributions along the (1989)
longest or shortest path were qualitatively similar to Overton
et al. (1987), with maximum tissue dose in the first pulmonary
region generation.
Human LRT
(newborn to
adult)/TB: based on
Yeh and Schum
(1980); PUL: based
on Hansen and
Ampaya (1975)
Miller etal. (1985) Enhanced Overton etal. (1987). From various
sources, develops age-dependent LRT anatomical
models. For quiet and maximal exercise
breathing, applies the dosimetry model of Miller
et al. (1985) to several ages from birth to adult,
illustrating the LRT distribution of absorbed O3.
For quiet breathing, the LRT distribution of dose, the percent
uptake, and the PAR dose are not very sensitive to age; but
are more sensitive during exercise. Regardless of age and
breathing state, the largest O3 dose occurs in the PAR. No
uptake in the URT.
Overton and Graham
(1989); Miller and
Overton (1989)
-------
Table 8-5 (cont'd). Theoretical Ozone Dosimetry Investigation!
Species and Region
Modeled/Anatomical
Model
Liquid Lining and Tissue
Transport and Chemical
Reactions
Dosimetry Model /Subject of Investigation
Results/Predictions
Reference
oo
to
Human total RT/URT:
Hanna and Scherer
(1986); LRT: Weibel
(1963)
Time-dependent molecular
diffusion and first-order
reactions in liquid lining
and interstitium; transfer
through epithelium
modeled as a permeability
process—no reactions in
this layer. URT and LRT
liquid lining rate
constants: 50 and 1 times
that of Miller et al. (1985),
respectively
Model development. Lung dimensions scaled to
those of a young male and a young female.
Contrasts LRT air-phase concentrations during
exercise and rest. Compares male and female
air-phase O3 concentrations and male and female
subepithelial concentrations.
URT uptake may be greater in cold than in warmer air. For
the ventilatory parameters used, (1) subepithelial
O3 concentrations are a maximum in the terminal bronchioles
or the first respiratory bronchioles, and (2) these
concentrations are greater in the female than in the male for
most of the RT.
Hanna et al. (1989)
Human, distal segment
of a lobe/ based on
Horsfield et al. (1971)
Human LRT/Weibel
(1963)
Rat LRT/based on
serial reconstruction of
a set of intrapulmonary
airways and their
ventilatory units
combined with a single
path from the larynx to
the reconstructed set
based on Yeh et al.
(1979)
Similar to Miller et al.
(1985)
Investigates two
formulations:
(1) Miller et al. (1978) and
(2) Miller et al. (1985)
TB and pulmonary region
mass transfer coefficients
based Overton et al.
(1987)
Monte Carlo simulation; transport processes
defined in terms of probabilities based on
physical and chemical principles. The effect of
lung asymmetries on the distribution of uptake in
the pulmonary region.
Model development; assumes quasi-steady
conditions. Air-phase concentration and tissue
dose profiles for the two reaction schemes,
various ventilatory parameters, and various liquid
lining transport and chemical parameters.
Compares predictions of first-order reaction
scheme to results of Miller et al. (1985).
Model development. Illustrates the influence of
ventilatory unit size and proximal anatomic dead
space and on the uptake and distribution of
inhaled O3 in ventilatory units. (Illustrates the
influence of ventilatory unit volume on the
distribution of inhaled O3 within ventilatory
units.)
Tissue dose in the PAR along the shortest path is
approximately 50 % larger than that along the longest path.
This model does not conserve mass. Predictions should only
be considered qualitatively. Maximum tissue dose in
respiratory bronchioles for both chemical reaction schemes.
Ventilatory unit uptake is significantly influenced by both
proximal airway dead space and ventilatory unit volume.
Flux of O3 to air-liquid interface in the proximal portions of
larger ventilatory units are significantly greater than in
smaller units.
Ultman and Anjilvel
(1990)
Grotberg et al.
(1990) (Grotberg
[1990])
Mercer et al. (1991)
(Mercer and Crapo
[1993])
-------
Table 8-5 (cont'd). Theoretical Ozone Dosimetry Investigation!
Species and Region
Modeled/Anatomical
Model
Liquid Lining and Tissue
Transport and Chemical
Reactions
Dosimetry Model /Subject of Investigation
Results/Predictions
Reference
Rat ventilatory unit/
Mercer et al. (1991)
Mercer et al. (1991)
Mercer et al. (1991). Along a path distally from
a bronchiolar-alveolar duct junction; compares
experimentally determined changes in ventilatory
unit wall thickness due to an O3 exposure to
dosimetry model predictions of flux to the air-
liquid interface.
As a function of distance from the BADJ; experimentally Pinkerton et al.
determined relative changes in ventilatory unit wall thickness (1992);
due to O3 exposure are very similar to predicted relative Miller and Conolly
fluxes to the air-liquid interface. (1995)
Human total RT/URT:
Fredberg et al. (1980);
LRT: Weibel(1963)
Pseudo steady-state
diffusion and first order
reactions combined with
biochemical data of Miller
et al. (1985)
Model development. Illustrates LRT distribution
of (1) air-phase concentration at various times
during the breathing cycle and (2) O3 flux (dose)
to liquid lining and to tissue.
Flux of O3 to air-liquid interface decreases distally; flux to
tissue increases along the conducting airways, reaches a
maximum in the terminal bronchioles, then decreases rapidly
in the gas exchange region.
Hu et al. (1992a)
'See Appendix A for abbreviations and acronyms. Generally, for modeling purposes, PAR is the first pulmonary region, generation, or model segment; PI! = pulmonary region.
''Refers to the theoretical or mathematical formulation aspects of gas transport and reactions without the specification of morphometric and physiobgical parameter values.
oo
to
-------
dosimetry models (with respect to groups of co-workers and independent model formulation):
the models of (1) Overton et al. (1987) and Miller et al. (1985); (2) Hanna et al. (1989);
(3) Grotberg et al. (1990), although they considered two reaction schemes; (4) Mercer et al.
(1991); and (5) Hu et al. (1992a). In some cases, several references have been grouped into
one investigation. This is because the multiple studies came from the same co-workers or
laboratory and added to or were complementary to the original or common dosimetry modeling
theme.
Major factors affecting the local uptake of reactive gases in the RT were respiratory
tract morphology and anatomy; the route of breathing (nose or mouth); the depth and rate of
breathing (VT and f); the physicochemical properties of the gases; the processes of gas
transport; and the physicochemical properties of the liquid lining of the RT, respiratory tract
tissue, and capillary blood. A detailed discussion of these factors can be found in Overton
(1984), Ultman (1988), and Overton and Miller (1988).
Because all of the dosimetry models listed in Table 8-5 were developed to simulate
uptake in the LRT or the total RT, these models have some common aspects. These include
the formulation of O3 transport and wall loss in the air compartments of the RT, the use of
species-dependent morphometric models or data to define air and liquid lining compartment
dimensions, and a description of the transport and loss of O3 in the liquid lining and tissue.
In all the dosimetry models that have become available since 1986, except for
Ultman and Anjilvel (1990) and Grotberg et al. (1990), which are discussed later, O3 transport
and loss processes in air compartments were approximated in terms of a one-dimensional,
time-dependent, partial differential equation of continuity. This type of equation accounts for
axial convection and dispersion or diffusion and the loss of O3 by absorption at the gas-liquid
interface. The use of this approximation is very common in modeling the transport in the LRT
of gases such as oxygen, nitrogen, helium, and CO2 (e.g., Scherer et al., 1972; Paiva, 1973;
Chang and Farhi, 1973; Yu, 1975; Pack et al., 1977) and has been assumed to be applicable
to O3. Ultman and Anjilvel (1990) used a Monte Carlo method to simulate O3 uptake. Based
on the physical and chemical principles of mass transport in the RT, probabilities were
assigned to the fate of a molecule in a way so as to account for convection, dispersion, and loss
to the liquid lining.
Dosimetry models published since 1986 can be grouped according to how transport
and chemical reactions are modeled in respiratory tract fluids and tissues: those based on the
formulation of (1) Miller (1977) and Miller et al. (1978), who used an instantaneous reaction
scheme, and (2) Miller et al. (1985), who used a quasi-steady, first-order reaction scheme.
These two approaches are discussed in the earlier criteria document (U.S. Environmental
Protection Agency, 1986). In addition to the use of similar formulations for liquid and tissue
transport/reactions, all of the post-1986 studies used essentially the same biochemical data of
either Miller et al. (1985) for humans or Overton et al. (1987) for laboratory animals. The
implication is that most of the studies are expected to predict qualitatively similar results.
There are minor variations on the second chemical reaction formulation. Hanna
et al. (1989) used a time-dependent diffusion-reaction equation, instead of the time-independent
(quasi-steady) equation used by Miller et al. (1985). Based on the rate constants used by
Hanna et al. (1989) and on discussions in Miller et al. (1985) and in Grotberg et al. (1990), the
modeled transport processes in the liquids and tissues can be inferred as essentially quasi-
steady, which is equivalent to the second formulation. Another variation uses mass transfer
coefficients determined by the second formulation and the biochemical assumptions of Miller
8-28
-------
et al. (1985) or Overton et al. (1987). The liquid and tissue transport/reaction formulation for
specific investigations is indicated in column 3 of Table 8-5.
In addition to the assumptions and the formulation of equations that describe the
transport and loss of O3 in the RTs of humans or laboratory animals, it is important to evaluate
whether simulation results reflect accurate solutions to the mathematical dosimetry model
formulation. Of the five distinct model formulations listed in Table 8-5, Overton et al. (1987),
Mercer et al. (1991), and Hu et al. (1992a) discuss most or all of the relevant issues of
stability, solution convergence, and mass conservation. In addition, using steady
unidirectional flow in a straight tube as a test case, they report successfully simulating
analytical solutions to their equations of transport and uptake. Neither Hanna et al. (1989),
Ultman and Anjilvel (1990), nor Grotberg et al. (1990) address the issue of accuracy. There is
no reasonable way to judge whether the solutions of Hanna et al. (1989) or of Ultman and
Anjilvel (1990) accurately represent solutions to their dosimetry model assumptions and
formulations; however, with the exception of Grotberg et al. (1990), there are no reasons to
assume that the solutions of these models are not accurate.
Because the Grotberg et al. (1990) model formulation is different than the others, an
explanation is needed. Based on the smallness of relevant parameters, Grotberg et al. (1990)
assume quasi-steady conditions for O3 concentration and air velocity in the air compartments
and obtain approximate analytical solutions to the time-independent, three-dimensional
equation of continuity for a model airway and apply the results to the morphometric model of
Weibel (1963). One advantage of analytical solutions is that they account naturally for
parameters (such as dispersion and gas-phase, mass-transfer coefficients) or local processes
(e.g., possibility of high uptake at airway entrances) that must be known and estimated for, or
incorporated into, the one-dimensional approach. Grotberg et al. (1990) carried out
simulations using anatomical and physiological conditions based on Miller et al. (1985) and
compared their results. Although qualitatively similar to Miller et al. (1985), Grotberg and
co-workers predicted significantly larger pulmonary tissue doses (up to 10-fold).
A comparison of the pulmonary region doses predicted by Grotberg et al. (1990) to those
predicted by Miller et al. (1985) indicates that the Grotberg et al. (1990) model does not
conserve mass (it predicts that the pulmonary region absorbs over 3.4 times the amount of O3
inhaled). The overprediction may be an artifact of the quasi-steady approximation, because
effects due to differences in the time of flights from the trachea to different LRT locations are
not taken into account. In any case, the quantitative predictions reported by Grotberg et al.
(1990) are questionable.
Chemical data more recent than that used for the dosimetry models in Table 8-5
show that compounds other than unsaturated fatty acids (the only compound with which O3 is
assumed to react in the models using the second chemical reaction formulation) are as reactive
or more reactive with O3 (Pryor, 1992). Using these data, estimates of O3 diffusion
coefficients in the liquid lining and bilayers, layer thicknesses, and data on the concentrations
of biocompounds in these layers, Pryor (1992) estimates that most of any O3 that penetrates
into a cell bilayer reacts within the layer (very little if any penetrates to the cell interior), and
O3 will not penetrate the liquid lining where it is greater than 0.1 /xm thick. Several relevant
comments are given by Pryor (1992): the calculations are considered a crude first
approximation; the possibility that a small fraction of O3 may penetrate the bilayer and reach
the cell interior can not be excluded; and surfactant layers can be very thin, and some cells
may not be protected very well or at all.
8-29
-------
If the conclusions of Pry or (1992) are essentially correct, they have implications
for past and future dosimetry modeling studies because past investigations have underestimated
the reactivity of O3 with biocompounds; with respect to cellular damage, products of
O3 reactions in the liquid lining may be the main toxic compound; increasing the value of rate
constants would have no effect on predictions of dosimetry models using the instantaneous
reaction scheme because rate constants in this approach are assumed to be infinite, but
increasing the concentration of reacting biocompounds would increase uptake; (4) use of
unsaturated fatty acid data only (with a first-order reaction scheme) results in an underestimate
of the reactivity of O3 in the liquid lining and an overestimate of the O3 tissue dose, and a
possible underestimation of the toxic dose due to reaction products; and (5) with higher O3
reaction rates, the first-order chemical-reaction formulation would result in larger predicted
uptakes.
8.2.4.2 Dosimetry Model Predictions
Similarity of Model Predictions
A survey of dosimetry modeling results shows that, in some areas of investigation,
there is a qualitative similarity in predictions by models of different groups of investigators for
different species or subpopulations.
(1) Distribution ofLRTdose (doseprofiles or dose versus generation). As shown
in Figure 8-5, beginning at the trachea, net dose (O3 flux to air-liquid interface)
slowly decreases distally in the tracheobronchial region (TB) and rapidly
decreases distally in the pulmonary region. Tissue dose (O3 flux to liquid-tissue
interface) is very low in the trachea, increases to a maximum in the terminal
bronchioles or first airway generation in the pulmonary region, and rapidly
decreases distally from this location (e.g., Miller et al., 1978, 1985, 1988;
Overton et al., 1987, 1989; Overton and Graham, 1989; Grotberg et al., 1990;
Huetal., 1992a).
If O3 were the only toxic agent and all the tissues of the LRT were equally sensitive
to the same dose, the models predict that the greatest morphological damage would occur in
the vicinity of the junction of the conducting airways and the pulmonary region and decrease
rapidly (distally) from this area, which is consistent with observations in laboratory animals
(see Chapter 6, Section 6.2.4). On the other hand, using the best estimates of morphometric
and physiologically based biochemical parameters of Miller et al. (1978, 1985) and Overton
et al. (1987), the models predict extremely (relatively) low tissue doses in the trachea and large
bronchi; this suggests very little or no tissue damage should occur there, which is contrary to
observations (see Chapter 6, Section 6.2.4). However, this is moot, if, as suggested by Pryor
(1992), the toxic substances are primarily reaction products of O3 and not O3 itself. In this
case, the O3 net local dose, not the local O3 tissue dose, may be a better estimator of local toxic
tissue dose, because the rate of production of products would be related to the rate of O3
uptake.
(2) Effect of exercise or increased ventilation. The effect of exercise is to slightly
increase the TB dose and to significantly increase the pulmonary region total
dose (mass of O3) and the CAR dose (mass per unit surface area) (e.g., Miller
etal., 1979, 1985; Overton et al., 1987, 1989; Overton and Graham, 1989;
Hanna et al., 1989; Grotberg et al., 1990).
8-30
-------
10"
o
c
E
n
oi
o
10-J
-------
(3) Effect of respiratory tract inhomogeneity. Models have predicted that the
further the proximal alveolar region is from the trachea, the less the O3 tissue
dose (mass of O3 absorbed per unit surface area) to the proximal alveolar
region. (For modeling purposes, the proximal alveolar region has been defined
as the first pulmonary generation or the first pulmonary region model segment
along a path; this region is a part of the CAR.) Overton et al. (1989) predicted
a threefold greater proximal alveolar region dose for the shortest path relative to
the longest path in rats. Ultman and Anjilvel (1990) simulated O3 distribution
in a small segment (< 1 %) of the distal airways of an asymmetric anatomic
model of the human lung. They found that the O3 tissue dose (mass per square
centimeter) in the proximal alveolar region along the shortest path was
approximately 50% greater than that along the longest path. Mercer et al.
(1991) found that path distance and ventilatory unit size affect dose: predicted
doses in the proximal segments (essentially, the proximal alveolar region) of the
larger ventilatory units (with the smallest relative dead space) are significantly
larger than the average proximal segment doses. Further, for the small sample
of ventilatory units modeled (43), Mercer et al. (1991) predicted a range of
proximal segment doses of greater than a factor of 6. Because the proximal
alveolar regions of Ultman and Anjilvel (1990) and of Mercer et al. (1991)
belonged to a "local cluster", and there are many clusters with varying distances
from the trachea, a variability greater than 50% and a factor greater than 6,
respectively, are expected in proximal alveolar region doses. Mercer and Crapo
(1993) illustrated the effect of ventilatory unit volume alone on the distribution
of dose, predicting that a 2.3 times larger unit receives 1.9 times the dose (mass
per surface area) of the smaller unit at the entrance of the unit.
The variability of predicted proximal alveolar region doses and, by inference, CAR
doses suggests that the magnitude of toxicological effects for different CARs are different.
This prediction is consistent with the observations of Schwartz et al. (1976) and Boorman et al.
(1980) of damage variation among different CARs of the same rat. It is reasonable to assume
that variable damage at equivalent but different morphological locations also occurs in humans.
Specific Topics
Effect of Assumptions About Anatomical Dimensions. For rats and guinea pigs,
Overton et al. (1987) used two morphometrically based anatomical models (rat anatomical
models: Kliment, 1973, and Yeh et al., 1979; guinea pig anatomical models: Kliment, 1973,
and Schreider and Hutchens, 1980) to investigate the influence of anatomical model
formulation on predicted uptake. Results with all four anatomical models in combination with
different ventilatory parameters showed a qualitative similarity in the shapes of the dose
profiles, but the two anatomical models for the same species resulted in considerable
differences in predicted percent RT and pulmonary region uptakes.
Respiratory Tract Uptake in Human Adults and Children. Overton and Graham
(1989) used several sources of data on age-dependent LRT dimensions and structure to
construct theoretical LRT anatomical models for humans from birth to adulthood. The
O3 dosimetry model of Miller et al. (1985) was used to estimate the regional and local uptake
of O3. For the percent uptake (84 to 88%) during quiet breathing, the LRT distribution of
8-33
-------
absorbed O3 and the centriacinar O3 tissue dose are not very sensitive to age. Regional percent
uptakes are more dependent on age during heavy exercise or work than during quiet breathing,
and total uptakes range from 87 to 93%. Generally, the total quantity of O3 absorbed per
minute increases with age. For all conditions simulated, the largest O3 tissue dose is predicted
to occur in the CAR. Miller and Overton (1989) present similar results. Because uptake by
the URT was not simulated and because this region can be assumed to have an important effect
on LRT uptake, a comparison of predictions of LRT uptakes in children and adults should be
viewed with caution. On the other hand, URT uptake probably has little effect on the shape of
the dose curves.
Dosimetry Modeling Results Compared to Dosimetry Data
Based on the experimental conditions discussed in Gerrity et al. (1988) and using
the model and parameters of Miller et al. (1985), Miller et al. (1988) simulated the uptake of
O3 distal to the oropharynx of human subjects. For the target f's of 12 and 24 bpm and VTs
ranging from 0.4 to > 1.6 L, the simulation results were in good agreement with the breath-
by-breath experimental data. The average experimental LRT uptake efficiency was DO.91 as
compared to the 0.89 prediction given by Miller et al. (1985) for the region distal to the
oropharynx. It should be remembered, however, that values for uptake efficiency from the
Gerrity et al. (1988) study were derived from the raw data using a steady-state method,
whereas the models of Miller et al. (1985) and Miller et al. (1988) utilize cyclic flow, thus
making the predictions more appropriate for comparison with uptake data from non-steady
state methods. From an analysis in Gerrity et al. (1994), it appears that total RT uptake
computed by either steady-state or non-steady state methods differ by only about 10% in
relative terms.
There have been major improvements to the original model, as described by Miller
et al. (1985) and Overton et al. (1987), including the addition of the URT and establishing a
regional mass transfer coefficient based on experimental data. Table 8-6 summarizes the
assumptions that underpin these improvements, as well as other relevant assumptions used for
simulating the uptake of several human dosimetry experiments. A discussion of the
assumptions is given in Section 8.5.2.
After taking into consideration the assumptions of Table 8-6, Table 8-7 compares
experimental total RT uptake efficiency data and model predictions for humans. Use of the
rat assumptions in conjunction with the model will be discussed later (Section 8.4.3). The
model predictions show good agreement with the total RT uptake efficiency data of Gerrity
et al. (1988), Gerrity et al. (1994), and Hu et al. (1992b). In all cases, the predictions are
within 10% of the measured values. The agreement with the data of Hu et al. (1992b) is even
better, as expected.
The model prediction for the data of Wiester et al. (1996) is less accurate.
Comparison of the Wiester et al. (1996) data with the Hu et al. (1994) data (Figure 8-1) shows,
however, that the results are in good agreement with each other. Thus, it would appear that
the VT dependence of the model does not necessarily reflect the real world. However, the
general agreement between the model predictions and data are quite good.
Although the models are capable of making reasonable predictions of total RT
uptake efficiency, their accuracy for specific regions remains uncertain. The O3 bolus data
(Hu et al., 1992b, 1994; Ultman et al., 1993) and the airway uptake efficiency data
8-34
-------
Table 8-6. Assumption for Application of Dosimetry Model
to Breathing Frequency Responses to Ozond
Species
Mode of
Breathing
Respiratory Tract Morphology
Mass Transfer Coefficients
Human Oral
Rat
Nasal
- LRT structure from Weibel (1963).
• Volume of oral cavity through
larynx and surface to volume
ratio (S/V) from Hu et al.
(1992a,b).
• Dead space volume (Vd) and FRC
from Hart et al. (1963); TB
region volume at FRC equals
Vd minus oropharyngeal volume.
• Proximal alveolar region defined as
first respiratory bronchiole.
• Pulmonary region expands;
TB does not expand.
• NP dimensions from Schreider and
Raabe (1981).
• TB region from Yeh et al. (1979).
• Volumes and surface areas of LRT
isotropically scaled to FRC.
• Pulmonary region from Mercer
et al. (1991).
• TB and pulmonary regions
expand uniformly during
breathing.
• Proximal alveolar region is first
generation of pulmonary region.
Mass transfer coefficients for each
oropharyngeal segment and each TB
generation defined as Ka/(S/V) where
S/V for the TB region is from Weibel
(1963) dimensions reduced to FRC; Ka
from Hu et al. (1992b).
Pulmonary mass transfer coefficient is
0.10 cm/s (Miller et al., 1985).
NP and TB mass transfer coefficients
estimated using data of Hatch et al.
(1989).
Mass transfer coefficient of pulmonary
region = 0.137 cm/s; inferred from
Pinkerton et al. (1992).
"See Appendix A for abbreviations and acronyms.
Table 8-7. Comparison of Total Respiratory Tract Uptake
Data with Model Predictions1
VT (rnL)
832
832
500
1,000
1,650
1,239
631
f (bpm)
12
24
15
7.5
25
35
16
Measured Ft
0.97
0.96
0.86
0.93
0.88
0.87
0.76
Predicted Ft
0.96
0.93
0.89
0.94
0.95
0.93
0.94
Data Source
Gerrity et al. (1988)
Gerrity et al. (1988)
Hu et al. (1992b)
Hu et al. (1992b)
Gerrity et al. (1994)
Gerrity et al. (1994)
Wiester et al. (1996)
aSee Appendix A for abbreviations and acronyms.
8-35
-------
(Gerrity et al., 1995) raise questions about the model predictions of uptake in the conducting
airways. These latter data sets suggest that the models of Miller et al. (1985, 1988) may
underestimate the O3 uptake coefficients in the conducting airways. Ultman et al. (1993) show
in their analysis of the bolus data that the reactivity of O3 with the lung liquid lining decreases
with increasing depth into the lung. This could imply that more O3 is taken up in central
airways than previously had been thought. However, the predictions presented in Table 8-7,
which are based on the assumptions of Table 8-6, represent a revision of the Miller et al.
(1985) model in that the mass transfer coefficients are derived from the actual human data of
Hu et al. (1992b).
Because the utility of dosimetry models is their usefuleness in facilitating
interspecies extrapolation, it is important to compare predictions with animal data as well as
with human data. Using the Overton et al. (1987) model formulation and parameters, Overton
et al. (1989) developed a formula that can be used to calculate LRT uptakes in rats, given their
VTs and fs. For comparison purposes, the data of Wiester et al. (1987, 1988) and Hatch et al.
(1989) can be used. The average uptake efficiency for the rat from these data is 0.45. Based
on the VTs and fs of these animals, an average LRT uptake of 0.61 is computed using Overton
et al. (1989). If Overton et al. (1989) had included the effects of URT uptake, their model
would have predicted more than 0.61.
It is important to note that these results are based on the older model assumptions,
not those presented in Table 8-6. If the newer assumptions had been used, the agreement
between predictions and actual results would have been much better (i.e., total uptake would
have been in the 50 to 60% range, depending on ventilation).
8.3 Species Sensitivity: Lung Function and Inflammatory
Endpoints Exemplifying an Approach
8.3.1 Introduction
Quantitative extrapolation of animal-based O3 toxicity data to the human
circumstance requires a paradigm that includes both an estimate of target tissue dose
(dosimetry) and an algorithm that relates the responsiveness of the test species to that of the
human (species sensitivity). This paradigm can be depicted as an extrapolation parallelogram
(Figure 8-6), which conceptualizes a relationship between chronic animal study data and long-
term human health effects based on an understanding of acute effects in both species (Graham
and Hatch, 1984). Although recent studies have begun to elucidate the underlying mechanisms
determining response, the bulk of the present O3 toxicity database in animals and humans
remains largely descriptive. Hence, only a simplified application of this paradigm is feasible at
this time. The following section will attempt to harmonize selective literature on acute human
and animal responses to O3 exposure (already reviewed in detail in Chapters 6 and 7) with
what is known about the dosimetry of O3, in an effort to discern relative species sensitivity.
To construct an argument that is plausible for this test application, focus is on endpoints for
which there are sufficient data in both humans and test animal species and for which exposure
scenarios are similar. The endpoints compared include measures of pulmonary function and
markers of lung inflammation, most notably BAL protein and cells. When possible, other
influencing parameters, such as ventilation augmentation and antioxidants within the lung also
will be discussed. The body of in vitro cell studies has not been included because of the
8-36
-------
Acute Effect ^ ^ Chronic Effect
Human ^ ^ Human
^
/
/
/
/
/
/
*•
/
Acute Effect ^ Chronic Effect
Animal Animal
Experimental correlation
Extrapolation
difficulties in interpretation associated with Figure 8-6.
Parallelogram paradigm for utilizing animal data for human health
predictions. Acute homologous endpoints serve as the basis for
extrapolating chronic effects in humans from animal data.
dosimetry and culturing systems. The reader is referred to Chapter 7, Section 7.2.5, and the
recent review by Keren et al. (1994). The goal here is to develop a hypothetical model for the
assessment of species-specific sensitivity with acute O3 exposure that can serve as a framework
to better predict human responses, especially with regard to chronic effects. The complex
issue of whether controlled human clinical studies accurately reflect population-based
responses also will not be considered in this discussion.
8.3.1.1 Dosimetry
This topic has been discussed in detail in Section 8.2. Recent studies by Hatch
et al. (1994), utilizing the nonradioactive isotope of oxygen, 18O, to label O3, have shown that
exposure of exercising humans (60 L/min) and resting rats to 0.4 ppm O3 for 2 h resulted in
4 or 5 times the 18O dose (as adduct) to the BAL constituents of humans as compared to those
of F344 male rats. This four- to fivefold difference appeared to be due to the exercise-
stimulated hyperventilation of the humans when compared to the rat and compared favorably
with indices of effect (i.e., BAL cells and protein at 24 h). Only when the rats were exposed
to 2 ppm O3 for 2 h did the 18O3 labeling of BAL constituents approximate that of the human.
Thus, on the basis of this study of cellular and protein influx due to O3 injury, the rat and
human appear to have similar sensitivity to O3 when exercise is considered. Additional related
studies with 18O3 indicate that deposition in the RT is a cumulative function of ventilation over
the initial period of exposure, which would lend support to these findings (Santrock et al.,
8-37
-------
1989). Attempts to compare animal data obtained without exercise to human study data with
exercise thus would underestimate the dose to the lung and presumably the resultant risk of
effect. Studies of O3 effects (but without assessment of dose) in exercising rodents have
confirmed this conclusion (Mautz et al., 1985).
Exercise is only one factor that can alter dose and effect. Studies in laboratory
animals that incorporate other factors, such as time of day or diurnal rhythms (Van Bree et al.,
1992), animal strain (Pino et al., 1991; Costa et al., 1993), or nutrition (Chapter 6,
Section 6.2.5), also show substantial modification of response to O3, and thus emphasize the
need for careful consideration of exogenous factors when attempting to compare or extrapolate
study findings. It is likely that a similar range of factor-dependent variability exists within
human test subjects.
8.3.2 Homology of Response
The concept of species sensitivity actually consists of two integrated components.
The first, homology of response, indicates whether the outcome seen in the animal test species
represents the same biological response in the human. In many cases, a measurement of the
same endpoint in both species can be presumed to reflect the same toxic phenomenon or
mechanism (i.e., the pulmonary irritant-induced tachypnea) (see below). On the other hand,
there may be endpoints that, although homologous, are not expressed similarly; for example,
the burning discomfort of sensory irritation in the human and the pause on tidal expiration seen
in rodents. The second component of species sensitivity relates the dose-response curve to
given homologous responses. Alterations in permeability of the air-blood barrier of the lung
appear to reflect true species differences in sensitivity to pulmonary irritants such as O3 (Hatch
et al., 1986). Ideally, these elements of species sensitivity should flow directly into
extrapolation formulae developed to integrate animal and human research data.
8.3.2.1 Lung Function Endpoints as Homologous Indicators
Lung function studies of small mammals have provided basic physiological
information important to the understanding of both normal and diseased lungs (Snider and
Sherter, 1977; Harkema et al., 1982; Raub et al., 1982; Mauderly, 1984; Costa, 1985).
Animal lung function tests, adapted from those used clinically, have proven useful in
describing the nature and severity of lung injury and in distinguishing toxicant-induced effects
in the central or peripheral airways from those effects in the parenchyma. In practice, the
interpretation of functional changes detected in animals derives from knowledge and experience
in human pulmonary medicine. Supporting this view, in theory, is the allometric database for
normal mammals, in which the lung function variables associated with ventilation and aerobic
metabolism scale systematically to body mass over nearly seven orders of magnitude (Stahl,
1967; Leith, 1976). The lung function studies of O3 toxicity in animals and humans considered
in the present discussion are described in detail in Chapters 6 and 7, Sections 6.2.5 and 7.2,
respectively, of this document and in the previous O3 criteria document (U.S. Environmental
Protection Agency, 1986).
8.3.2.2 Inflammatory and Antioxidant Endpoints as Homologous Indicators
Inflammation of pulmonary airways and airspaces is best described as a cascade of
events that network infiltrating leukocytes, plasma proteins, and cell-derived mediators, which
function presumably to defend or repair (but may further damage) the injured lung (see
8-38
-------
Chapters 6 and 7, Sections 6.2.2 and 7.2.4, respectively; Keren et al., 1994). Key markers of
the basic inflammatory process include plasma-derived proteins, such as albumin, globulins,
and plasmin, and a primary inflammatory cell, the polymorphonuclear leukocyte (PMN). Of
these two markers, plasma-derived proteins in the acute phase generally are thought to
represent a "leak" from the vasculature to the airspace lumen. Hence, under controlled
temporal conditions, plasma protein residing in the airspace and accessible by BAL can be used
as a proportional marker of effect that, in turn, should be related to dose. The presence of
PMNs in the airspaces is a bit more complex because of the signals involved in recruiting these
cells into the lung lumen after injury and the cascade of events apparently involved in their
poiesis from the vasculature to the lung lumen. For the purposes of species comparison,
plasma-derived protein (nonspecific) and the proportion of PMNs among total cells as sampled
by BAL will be emphasized as primary indices of damage and inflammation within the lung.
Antioxidant substances in lung tissue (Slade et al., 1985) and BAL fluid and cells
(Slade et al., 1993) have been identified and quantified for humans and several laboratory
animal species. The species profiles of these antioxidants in the lung tissue and their
respective BAL cells and fluid can differ appreciably (Table 8-8), but collectively they appear
to play a significant role in defense of the lung against both endogenous and exogenous oxidant
challenge. In particular, ascorbate and vitamin E appear to have major functions in protecting
the lung from O3 challenge (Chapter 6, Section 6.2.1; Slade et al., 1989; Crissman et al.,
1993; Keren et al., 1989b; Elsayed et al., 1988), and, when their levels are manipulated in
vivo, either can influence the degree of toxic outcome. Hence, the measurement of basal and
O3 response levels of these antioxidants in BAL cells and fluids is useful in assessing the
qualitative and quantitative responses among humans and laboratory test species.
8.3.3 Studies of Lung Function
8.3.3.1 Confounding Influences in Lung Function Studies
Ideally, a system for measuring pulmonary function in small animals would be
approximately the same as that used in humans for cooperative, unrestrained subjects.
However, in animal studies, this is usually not possible. Fortunately, certain measures
(e.g., static lung volumes, diffusion capacity) appear to be minimally influenced by sensory
reflex or muscular activity in spite of unnatural stresses or blunting of responses caused by
anesthetic or physical immobilization. On the other hand, some measurements, typically those
involved in the assessment of ventilatory mechanics, can be profoundly influenced by these and
other factors, such as ambient and toxicant-altered body core temperature, thus confounding
cross-species comparisons. Because a major emphasis of this section is the comparison of lung
function data of animals and humans, it is important that the reader realize potentially
confounding influences borne by studies of lung function in rodents when compared to
analogous measurements in humans. These are discussed briefly below.
Anesthesia
Anesthesia alters pulmonary function measurements in both humans (Rehder et al.,
1975) and laboratory animals (Skornick and Brain, 1990; Lamm et al., 1982; Rich et al.,
8-39
-------
Table 8-8. Pulmonary Antioxidant Substances in Various
Laboratory Animal Species and Humang
Antioxidant Mouse Hamster
Ascorbate
Tissue" 41+4 26+2
BAL cells' — —
BAL fluid' — —
Glutathione
Tissue" 62+3 61+2
BAL cells' — —
BAL fluid' - -
Tocopherol
Tissue" 1.0+0.1 1.0+0.1
BAL cells" — —
BAL fluid" — —
Uric Acid
Tissue" — —
BAL cells' — —
BAL fluid' - -
"See Appendix A for abbreviations and acronyi
Rat
34+2
50.3+5.4
199.4 + 9.1
50+2
14.8+2.7
12.1+5.0
2.1+0.1
577.7 + 83.1
0.6+0.2
0.35+0.05
<0.01
4.3+0.6
ns; data (mean +
Guinea Pig Rabbit
39+1 27 + 3
17.9+1.4 -
28.8+2.2 —
83+3 83+3
14.6+2.4 —
11.2+1.9 -
2.0+0.2 1.4 + 0.2
454.5 + 58.2 -
1.4 + 0.5 —
4.14+0.24 —
0.8 + 0.1 —
2.7 + 0.4 -
SE) extracted and summarized
Human
22+7
3.5 + 0.1
21.4+2.8
7+1
2.9+0.5
20.4+3.8
0.8+0.1
95.1+23.4
47.2+3.8
—
0.07+0.03
15.9+2.5
from
Slade et al. (1985, 1993).
"Data expressed as mg/100 g wet tissue.
'Data expresses as nmol/mg protein.
dData expressed as nmol/mg lipid phosphorus.
1979). In general, ventilation is reduced and changes in ventilatory patterns occur (Pavlin and
Hornbein, 1986; Bellville et al., 1960; Hunter et al., 1968; Siafakas et al., 1983). In humans,
anesthesia can decrease compliance and FRC, and it also can increase airway resistance (Raw)
(Rehder et al., 1974, 1975). In small laboratory mammals, an analogous decrease in FRC
occurs, although apparently via a different physiological mechanism (Lamm et al., 1982).
Additional anesthesia-related effects include a blockade of irritant reflexes (Weissberg et al.,
1976) and alteration of ventilatory patterns in response to CO2 (Martin-Body and Sinclair,
1985). Hence, although not invalidating experimental results, choice of anesthetic agent may
affect the measured response and may confound cross-species comparison.
Restraint
Collection of small animal pulmonary function data without the use of anesthesia
usually requires some type of physical immobilization. Restraint may range from minimally
restrictive, allowing turning and some locomotion, to extremely confining, as occurs when
animals are inserted into nose-only exposure tubes. Although restraint reduces movement
8-40
-------
artifacts and permits attachment of delicate probes or sensors, immobilization can also produce
undesirable physiological disturbances such as changes in body core temperature (Tco)
(Nagasaka et al., 1979), hypermetabolism (Nagasaka et al., 1980), increased expiratory CO2
(Jaeger and Gearhart, 1982), changes in ventilation and ventilatory pattern (Lai et al., 1978;
Mauderly, 1986), and gastric response (Toraason et al., 1980). Such stress-related responses
are poorly understood, and their influence on toxicologic responses may well pass unnoticed
unless specifically examined.
Temperature
Although not widely appreciated, toxicant-induced changes in thermoregulatory
function can modify the results of toxicological studies (Gordon et al., 1988; Gordon, 1991).
Recent studies indicate that exposure to 0.37, 0.50, and 1.00 ppm O3 also can decrease Tco,
heart rate, and blood pressure over 2 or more hours in unrestrained, unanesthetized rodents
maintained at normal room temperature (Uchiyama et al., 1986; Watkinson et al., 1995). On
the other hand, when rats were restrained in a head-out body plethysmograph and exposed to
the same concentration of O3 (1 ppm) as in the Uchiyama et al. (1986) and Watkinson et al.
(1995) studies, no change in blood pressure was observed (Tepper et al., 1990). The
discordance between these findings may be the result of restraint stress, which has been shown
to increase Tco (Nagasaka et al., 1979) and, in this circumstance, could have blunted the
decrease in Tco associated with O3 exposure.
Although O3-induced changes in heart rate and Tco may be unique to rodents, this
phenomenon has not been well studied in humans. It is possible that because of their larger
thermal mass and different thermoregulatory mechanisms, humans do not exhibit similar
changes in these parameters on exposure. For example, rectal temperature increased by the
same amount in both air and 0.4 ppm O3 groups of humans during a 2-h exposure at 35 DC
(Bedi et al., 1982). The effects on Tco may have been confounded because the subjects
performed moderate exercise during alternate 15-min periods during exposure. On the other
hand, women exercising intermittently in moderate (24 DC) and hot (35 DC) ambient conditions
showed no change in Tco related to O3 exposure, but did show less of an increase in heart rate
(2.7%) than did air-exposed (8.1%) subjects at 35 DC (Gibbons and Adams, 1984). It should
be noted, however, that other studies have shown potentiation of human lung function
responses associated with increased ambient temperature and O3 exposure (Folinsbee et al.,
1977; Gibbons and Adams, 1984). The full importance of temperature in relating rodent and
human responsiveness to O3 remains to be understood.
Exercise and Ventilation
Exercise has long been employed in human studies to enhance the effects of air
pollutants, especially O3 (Folinsbee and Raven, 1984). Exercise appears to exacerbate
functional effects by increasing the inhaled dose (Hatch et al., 1994) and possibly by shifting
the deposition of the pollutant to more sensitive pulmonary sites (Gerrity and Wiester, 1987).
Although exercise can be used in laboratory animals to enhance deposition of O3, no direct
methods for measuring ventilation or breathing mechanics are available for small animals
during exercise. Alternatively in an attempt to mimic the increase in ventilation produced by
exercise in humans, studies employing restrained animals have used CO2 as a ventilatory
stimulant. Carbon dioxide (8 to 10%) maximally increases VE three to five times in rodent
species; CO2 in excess of 10% will result in a reduction in ventilation (Wong and Alarie, 1982;
8-41
-------
Tapper et al., 1988). This increase in VE is equivalent to light (2 to 3 X resting VE) or
moderate exercise (4 to 6 X VE) in humans (U.S. Environmental Protection Agency, 1986).
In many O3 studies in humans, both heavy (7 to 8 x VE) or very heavy (> 9 x VE) exercise
have been used. Similar increases in ventilation cannot be attained in small animals using the
CO2 challenge technique, thus posing a limitation in attempting to make direct comparisons
between animal and human studies.
The application of this CO2-challenge methodology in O3-exposed rats (0.25 to
1.0 ppm for 2.25 h with 15 min alternating hyperventilation) clearly demonstrates enhanced
pulmonary irritation, as reflected in breathing pattern changes during exposure (Tepper et al.,
1988, 1990). The breathing pattern alterations typical of O3 exposure appeared to be larger
than would be predicted based solely on increased dose, suggesting that CO2 challenge during
O3 exposure may have enhanced deposition at critical lung sites (Tepper et al., 1989). This
augmented response was reflected clearly in the large increases in protein observed in the BAL
fluid (Costa et al., 1988b). In postmortem studies, rats exercised during exposure have been
found to have exacerbated lung pathology, thus appearing to confirm this hypothesis (Mautz
et al., 1985) and suggesting that exercise may, in fact, enhance toxicity disproportionate to the
apparent dose of toxicant.
8.3.3.2 Acute Exposure Data
Two corners of the parallelogram paradigm can be constructed readily from data
gathered in empirical studies of acute O3 exposure in humans and laboratory animals. These
studies have the bulk of the data with the highest frequency of common endpoints that can be
compared. Hence, the following discussion will focus on several categories of homologous
lung function and BAL study data that have been obtained from humans and animals exposed
similarly to O3. The human studies were drawn from the large existing database on lung
function and represent typical responses. The corresponding BAL data are more limited and
are used to the extent possible. In contrast, the animal studies selected for comparison are
highly selective and represent a rather small database involving similar exposure scenarios and
homologous endpoints. This approach, of necessity, excludes the large majority of animal
studies, not because they do not contain important toxicologic data on O3, but rather, they are
disparate in their exposure parameters or biologic endpoints that readily can be tied to those
available in humans.
Tidal Breathing
In humans, O3 produces pulmonary irritation, a response associated with cough and
substernal soreness (Chapter 7). Although these symptoms are difficult to assess in animals,
exposure to sufficient concentrations of O3 produces reflex alterations in tidal breathing that
can be measured objectively. Most notably, the response is an increase in f that is usually
accompanied by a decrease in VT (tachypnea), whereas VE may not be altered. Although the
magnitude of the tachypneic response is variable, depending on the species and exposure
conditions, this endpoint is quite sensitive and consistent across many species (e.g., guinea
pigs, cats, dogs, rats, monkeys, humans).
To examine the cross-species response to O3, data were evaluated from human and
animal studies that reporting immediate postexposure alterations in f. Most human studies
employed an exercise regimen during O3 exposure to increase dose. On the other hand, few
animal studies have used exercise, relying rather on high exposure concentrations or
8-42
-------
CO2-induced hyperventilation. Three representative human studies were selected because they
used a large range of concentrations and ventilation rates. Selected data from these three
human studies are compared to the available animal data in Figure 8-7 and discussed further
below.
160
150
140
o
130
O)
£ 120
CD
-------
the O3-induced changes in f was examined further by DeLucia and Adams (1977), who
exposed humans to 0.15 or 0.30 ppm O3 for 1 h while the subjects exercised at one of four
ventilation levels (1, 3, 4, or 6 x resting VE). The magnitude of the f response increased with
concentration and exercise level, but was significant only in the highest exercise group at
0.15 and 0.30 ppm. Lastly, the lower limits of the concentration-response were explored by
McDonnell et al. (1983), where subjects performing very heavy exercise (65 L/min;
DlO x resting VE) were exposed to 0.12, 0.18, 0.24, 0.30, or 0.40 ppm O3 for 2.5 h.
Significant changes in f were observed at all exposure levels.
Although several animal studies have evaluated tidal breathing changes during and after
O3 exposure, only four studies have examined multiple concentrations such that comparisons to
human data can be made. Unanesthetized, restrained guinea pigs were exposed for 2 h to
0.34, 0.68, 1.08, or 1.34 ppm O3 via nose cones, while tidal breathing was measured using a
constant-volume plethysmograph (Murphy et al., 1964). A similar experimental preparation
was used by Amdur et al. (1978) to evaluate the respiratory response of guinea pigs to 0.2,
0.4, and 0.8 ppm O3. In both of these experiments, a monotonic increase in f was observed;
however, the animals studied by Murphy et al. (1964) were uniformly more sensitive to
O3 than those of Amdur et al. (1978). Mautz and Bufalino (1989) measured breathing patterns
in awake, restrained rats exposed for 3-h to 0.2, 0.4, 0.6, and 0.8 ppm O3. Concentration-
related increases in f were observed up to 0.6 ppm, but the responses to 0.6 and 0.8 ppm were
the same. In another study, awake rats were exposed to 0.12, 0.25, 0.50, and 1.00 ppm O3 for
2.25 h in head-out pressure plethysmographs where CO2-stimulated breathing was incorporated
to augment ventilation (Tepper et al., 1990). With the added CO2, rats and guinea pigs
appeared to be similarly responsive to O3. In general, as depicted in Figure 8-7, restrained
guinea pigs and rats appeared to be as responsive as the lightly exercising humans, and clearly
more responsive than the humans exposed at rest. Only with strenuous exercise does the
response of humans appear to exceed that of rodents.
In addition to similar concentration-related effects in humans and animals, the time-
related effects of O3 exposure appear to be similar. To demonstrate this homology, Mauderly
(1984) compared the time course of response to O3 in humans and guinea pigs exposed under
somewhat similar conditions. Humans were exposed to 0.75 ppm O3 for 2 h while engaging in
nonstrenuous IE at 15-min intervals (Folinsbee et al., 1975). In another study, respiratory
parameters were measured at 30-min intervals during exposure and for 4 h postexposure (Bates
and Hazucha, 1973). Similarly, unanesthetized, restrained guinea pigs were exposed to 0.68
ppm O3 for 2 h as part of a concentration-response study (described above), with respiratory
function assessed at 15-min intervals during exposure and for 3.5 h postexposure (Murphy
et al., 1964). In both guinea pigs and humans, f increased and VT decreased; both parameters
then returned toward control values during the postexposure period. The percent change from
control in f and VT was nearly the same throughout the exposure and postexposure periods,
indicating that a similar concentration of O3 (DO.7 ppm) produced similar temporal alterations
in ventilation. Again, the guinea pigs would appear to be slightly more responsive than
humans because the guinea pigs were exposed to a lower concentration (0.68 ppm) at rest,
whereas the humans were exposed to 0.75 ppm with light IE.
Mechanics
Breathing mechanics have been examined in several animal and human-O3 exposure
studies, but there is little similarity between the databases for the concentrations or the specific
8-44
-------
techniques used. Bates et al. (1972) examined breathing mechanics in resting and in lightly
exercising (2 X resting VE) humans exposed for 2 h to 0.75 ppm O3. Although no
concentration-response data were obtained, increased total pulmonary resistance (RL) and
decreased dynamic compliance (Cdyn) were found for both resting (+22 and Dl2%) and
exercising (+67 and D51%) subjects exposed to O3. In a similar study, Hazucha et al. (1989)
exposed men for 2 h to 0.5 ppm O3 using moderate IE (40 L/min), and found a significant
12.5% increase in Raw, although no concomitant change in Cdyn was detected. McDonnell
et al. (1983), using a broad range of O3 concentrations (0.12, 0.18, 0.24, 0.30, and 0.40 ppm;
2 h) and very heavy exercise (65 L/min), reported concentration-dependent increases in Raw.
In humans at rest or performing light exercise, however, a 2-h exposure at near-ambient O3
concentrations would be expected to induce only modest increases in R^ and no changes in
Cdyn (Hazucha, 1987).
Although relatively high O3 concentrations (Dl.O ppm) produced effects on RL and
Cdyn in animals (Murphy et al., 1964), only three studies in animals have evaluated these
parameters at lower, more relevant O3 concentrations. Watanabe et al. (1973) studied
anesthetized, paralyzed, and mechanically ventilated cats exposed via a steel tracheal tube to
either 0.25, 0.50, or 1.00 ppm O3 for between 2 and 6.5 h. Measurements of breathing
mechanics were recorded every 30 min. With increasing O3 concentration and exposure
duration, RL increased and, to a lesser extent, Cdyn decreased. Bronchoconstriction at
0.25 ppm O3 was reversed following atropine (a parasympathetic receptor blocker), but only
partially reversed at the two higher concentrations, suggesting the involvement of more than
bronchoconstriction in the increase in RL at these levels. Unfortunately, relating the
concentrations used in this study to other animal or human studies is difficult because exposure
through a tracheal tube would eliminate scrubbing of O3 by the nose and oropharynx and likely
would exaggerate the pulmonary O3 dose (Gerrity et al., 1988).
Other studies have attempted to examine breathing mechanics in unanesthetized animals
with natural nasal breathing and avoidance of potential anesthesia-related blunting of reflex
responses. Murphy et al. (1964) exposed unanesthetized guinea pigs to several concentrations
of O3 for 2 h and measured ventilation, as previously discussed, and RL. At concentrations
less than 1 ppm, O3 had no effect, but RL increased 48 and 147% at 1.08 and 1.34 ppm,
respectively. Using a similar test system, Amdur et al. (1978) observed no significant
alteration of RL in unanesthetized guinea pigs during a 2-h exposure to 0.2, 0.4, or 0.8 ppm
O3. However, Cdyn decreased significantly at 0.4 and 0.8 ppm O3. In analogous studies in
unanesthetized rats, Tepper et al. (1990) observed no significant changes in RL or Cdyn after a
2.25 h exposure to 0.12, 0.25, 0.50, or 1.00 ppm O3, in spite of intermittent 15-min periods of
exercise-like hyperventilation induced by CO2.
Although increased resistance is demonstrable in guinea pigs, cats, dogs, and humans, a
comparison of percent change in resistance from control measurements after an acute (D2 h)
O3 exposure (Figure 8-8) suggests that humans are more likely to bronchoconstrict due to
O3 exposure than rodents. Neither of the guinea pig studies (Murphy et al., 1964; Amdur
et al., 1978) nor the rat study (Tepper et al., 1990) showed a significant increase in RL at less
than 1 ppm. However, closer examination of the human data reveals that the McDonnell et al.
(1983) study employed very heavy exercise, and most of the studies included in the Hazucha
(1987) model used moderate to heavy exercise. Thus, the inhaled dose likely would be greater
than in spontaneously breathing animals. A more comparable study in humans that employed
8-45
-------
only light exercise reported that 0.25, 0.37, and 0.50 ppm for 2 h resulted in minimal,
nonsignificant 118, 124, and 104% increases inRaw (Hackney
8-46
-------
200
180
o 160
-§^
c
o
O
•5 140
0)
£ 120
CO
H— I
«
I 100
80
60
Human, Very Heavy, 2.5h
McDonnell et al. (1983)
Human, Rest, 2h
Bates etal. (1972)
Human, Light, 2h
Bates etal. (1972)
*
Human, Combined, 2h
Hazucha(1987)
Human, Light, 2h
Hackney etal. (1975)
B
Cat, Rest, 2h
Watanabe et al. (1973)
Rat, Light, 2.25h
Tepperetal. (1990)
Guinea Pig, Rest, 2h
Amduretal. (1978)
0.2 0.4 0.6 0.8
Ozone (ppm)
1.0
Figure 8-8. Comparison of changes in resistance after ozone exposure in humans and
animals. Data are expressed as percent of the control response. Right-hand
legend indicates species, exercise level, exposure duration, and the
reference. Human data are plotted with solid lines and open symbols. The
line labeled "Hazucha (1987)" is a model of predicted response. Animal
data are plotted with dashed lines (differentiated by species) and closed
symbols.
et al., 1975). Likewise, the Bates et al. (1972) data obtained in subjects at rest and with light
exercise (0.75 ppm O3) also argue against an unusually high O3 responsiveness in humans
relative to test animals for this endpoint when exercise-related dose is considered.
In general, similar findings have been observed using the measurement of Cdyn;
however, the response decrements were more variable and of smaller magnitude. Given the
distal deposition of O3, as indicated by morphological studies (Chapter 6, Section 6.2.4), it is
surprising that so little attention has been given to this parameter. Available data suggest that
these changes in Cdyn are of little biological significance for ambient exposures.
8-47
-------
Airway Responsiveness
The ability of O3 to increase airway responsiveness to nonspecific
bronchoconstricting stimuli in humans and other mammalian species has been known for at
least a decade (Chapters 6 and 7, Sections 6.2.5 and 7.2.4, respectively). However, airway
responsiveness is perhaps the least understood response to O3, particularly in the context of
species comparisons. Humans clearly exhibit increases in airway responsiveness at
environmental O3 exposure levels (Gong et al., 1986; McDonnell et al., 1987; Folinsbee et al.,
1988; Horstman et al., 1990), whereas analogous responses in animals at O3 concentrations
below 1 ppm are controversial. Most studies of airway responsiveness in laboratory animals
focus on the development of asthma-like models to elucidate generic mechanisms of airway
responsiveness and utilize concentrations as high as 3 ppm for brief periods of time to injure
the airways. Hence, anything more than a qualitative comparison between animal species and
humans is tenuous and, thus, will not be discussed further in this section. Details of the
methodologies of the laboratory animal and human bronchoreactivity studies can be obtained in
the reviews of pulmonary function found in Chapters 6 and 7, Sections 6.2.5 and 7.2.3,
respectively.
Elasticity and Diffusion
The integrity of the pulmonary air-blood barrier is essential for efficient exchange
of oxygen and CO2. This fragile epithelial interface with matrixed interstitial connective
tissues and capillaries possesses inherent elastic properties and presents a finite resistance to
oxygen diffusion to the blood. Although the elastic and diffusionary properties of the blood-air
barrier are not linked implicitly to one another functionally, both properties can be quantified
readily and compared between humans and laboratory animals (Costa, 1985). When
combined, assessment of these functional properties is often sufficient to evaluate pathologic or
toxic events in the distal reaches of the lung. For this reason and because of the fact that O3
deposits in the deep lung, the effects of O3 on these parameters will be discussed together.
Inhaled O3 is known to penetrate to the depths of the lung and preferentially deposit
in the smallest airways and its proximal acini (Section 8.2). Somewhat surprisingly, relatively
few studies in humans have sought to characterize potential functional impairments at the air-
blood interface. The reasons for this are likely twofold. First, in the early studies of the
health effects of O3 on humans, static compliance and diffusing capacity for carbon monoxide
(DLCO) were affected at only very high concentrations, well above what would be considered
environmentally relevant. Second, from a practical perspective, these measurements proved to
be considerably more tedious to perform than the forced expiratory measurement, which
sensitively detects O3-induced alterations (discussed below). Nevertheless, there are sufficient
data on humans exposed acutely to O3 to allow a reasonable comparison of these endpoints
with their more abundant animal homologues.
The earliest studies leave little doubt that O3 is edemagenic at high concentrations in
virtually all mammalian species. In the past, occupational exposures of 2 to 3 ppm O3 were
not uncommon, and a 9-ppm peak exposure has been reported (Kleinfeld et al., 1957; Challen
et al., 1958). The resultant worker symptoms and signs, including chest radiograms, were
consistent with the manifestations of edema reported in experimental animals (i.e., increased
lung weight and stainable edema in the airspaces) (Stokinger, 1965). Lung function, however,
typically was not measured in these work-related exposures. In a later study of arc welders
exposed to 0.2 to 0.3 ppm O3, little, if any, convincing evidence of functional impairment, in
8-48
-------
terms of altered lung volumes, maximal expiratory flow rates, and DLCO was obtained (Young
et al., 1963). To further explore the possible effects of O3, Young and coworkers (1964)
subjected 11 human volunteers at rest to controlled atmospheres of 0.6 to 0.8 ppm O3 via
mouthpiece for 2 h. Small reductions in vital capacity (VC), forced expiratory volume in 0.75
s, dynamic and static lung compliance, and intrapulmonary gas distribution were observed, but
only the 25% fall in DLCO proved to be statistically significant. Similar, but considerably more
variable effects on lung function were reported by Hallett (1965) in 10 subjects exposed to 1 to
4 ppm O3 for 30 min. Nonetheless, of 10 exposed subjects, seven showed at least a 20% drop
in DLC0. Like Young and coworkers, Hallett (1965) interpreted these changes to indicate lung
edema, in agreement with the hypothesis that the deep lung irritant O3 was having its effect at
the alveolar level. Interestingly, additional work from the same laboratory of the Young study
(Bates et al., 1972) found that resting subjects receiving nasal exposure to 0.75 ppm O3 for 2 h
resulted in a nonsignificant 3% reduction in DLC0. However, in a limited test group, the co-
imposition of light exercise, which doubled ventilation, enhanced this response (Dl2%). It
appears that the nasal (Bates et al., 1972) versus mouthpiece (Young et al., 1964) routes of
exposure were instrumental in the differential response, because it is likely that the mouthpiece
diminished what scrubbing occurs when exposure is via the unencumbered mouth in human
test subjects (as reported by Gerrity et al., 1988). Since these early studies, there have been no
additional controlled human acute studies that have examined alterations in DLCO at
O3 concentrations below 0.6 ppm.
Analogous animal studies of acute O3 exposure indicate that the general pattern of
functional impairment is similar to that reported in human studies. Anesthetized and ventilated
cats showed a general decline in VC, static lung compliance, or DLCO with exposures up to 6.5
h of 0.26 to 1.00 ppm O3 (Watanabe et al., 1973). The responses of the 20 animals were
variable, and these declines, which did not achieve overall statistical significance, were thought
to be largely secondary to the substantial (36 to 200%) increases in Raw. In a more complex
study design, rats were exposed for 2 or 7 h to 0.5 or 0.8 ppm O3 with intermittent 8% CO2 to
hyperventilate (D2 to 3 x resting VE) the animals as an exercise analogue to human exposures
(Costa et al., 1988a). The DLCO values were reduced by about 10% at both 0.5 ppm time-
points and by about 12% with a 2-h exposure to 0.8 ppm. Exposure to 0.8 ppm for 7 h,
however, greatly exacerbated the alveolar effect with a resultant 40% reduction in DLC0.
Static compliance, unaffected by the other exposure conditions, was affected only at this latter
exposure duration. These O3-induced effects, particularly the reductions in DLCO, appeared to
correlate with the degree of lung edema in affected animals, as had been surmised for the
acutely exposed humans. With the multitude of more recent studies of O3 at ambient levels,
alterations in static lung compliance or DLCO rarely are reported in either humans or animals.
Forced Expiration
Reductions in FVC and FEV, have become the hallmarks of acute lung dysfunction
in humans after O3 exposure (Chapter 7). These measures are sensitive to O3 levels as low as
0.12 ppm for as little as 2 h when heavy IE is included during the exposure (McDonnell et al.,
1983) and show cumulative dysfunction resulting from 6.6 h of lower levels of this oxidant
(0.08 and 0.10 ppm) when nearly continuous, moderate exercise is employed (Horstman et al.,
1990). Reductions in FEV, and FVC induced by O3 are believed to be partly the result of
pain-mediated interruption of maximal inspiration (Hazucha et al., 1989). Exactly what level
8-49
-------
of tissue injury or inflammation correlates with these functional deficits is unclear and is an
active area of research (Chapter 7, Section 7.2.4).
Most studies of O3 in experimental animals make little effort to mimic human study
designs, thereby impeding the extrapolation of experimental animal study results to humans.
Recently, however, rat studies involving periods of intermittent CO2-induced hyperventilation
to enhance dosimetry have attempted to capitalize on the qualitative similarity of the rat and
human maximum expiratory flow volume (MEFV) curves as a potentially sensitive endpoint of
toxicity (Costa et al., 1988a; Tepper et al., 1989). In the rat, FVC does indeed decrease with
O3 exposure, although the magnitude of response is apparently less than that observed in
humans. As in the human, the reduced FVC in the animal appears not to be the result of a
change in residual volume. Total lung capacity may be reduced slightly, but lung compliance
does not change. However, it is premature to assume a common mechanism for the FVC
reductions in the rat and human. Unlike the human, pain on inspiration in the animal model is
likely not an issue because the animal is anesthetized during the procedure and is brought to
TLC by a defined airway pressure (D30 cm H2O). Because general anesthesia is known to
diminish sensory afferent stimuli, an analogous O3-induced fall in rat FVC should expectedly
have been blunted, if not totally eliminated. To what extent anesthesia mitigates the rat
response or that there are inherent species differences in dosimetry or sensitivity is not clear
from these studies. Nevertheless, comparison of model-predicted FVC changes in humans
(Hazucha, 1987) with analogous rat data (Costa et al., 1988a) would suggest that this response
in the anesthetized rat is about half that of the human (Figure 8-9).
Studies of Inflammation and Antioxidant Content of Bronchoalveolar Lavage Fluid
Both humans and animals exhibit a PMN inflammatory response with associated
changes in lung permeability after acute exposure to O3. Recent studies indicate that humans
exposed to O3 concentrations as low as 0.08 ppm for 6.6 h with moderate exercise (40 L/min)
exhibit a fourfold increase in the percentage PMNs when BAL is obtained 18 h postexposure
(Devlin et al., 1991). To date, animal studies at comparable exposure levels are rare,
(Hotchkiss et al., 1989), and exercise enhancement of exposure dose has yet to be
incorporated. As noted above, the issue of dosimetry is critical if extrapolation at such levels
is to be attempted. Nevertheless, in the one acute rat study at 0.12 ppm O3 for 6 h, an increase
in nasal-lavage-derived PMNs was noted 18 h postexposure, with no similar change in PMN
number in the BAL (Hotchkiss et al., 1989). In contrast, in the same study when higher
concentrations of O3 (0.8 and 1.5 ppm) were used, BAL PMNs were elevated, but no changes
were observed in the nose washings. Such a "competitive" nasal-pulmonary response has yet
to be studied directly in humans. Nevertheless, the data support the general hypothesis that
there is comparability between the inflammatory responsiveness of rats and humans. More
direct comparison of laboratory animal inflammatory responses with those of humans can be
drawn from studies at higher concentrations when the nasal/lung competitive response in the
rat is skewed to the lung, and, like in the lung-function comparison, analogous exposure
conditions can be more directly compared. These studies are tabulated in Table 8-9 and
discussed in more detail below.
Four representative human studies of lung inflammatory responses after acute
O3 exposure can be compared with existing acute animal data from studies of analogous design.
Seltzer et al. (1986) exposed moderately exercising (83 to 100 W) subjects to 0.4 or 0.6 ppm
O3 for 2 h, with BAL obtained 3 h postexposure. The BAL fluid from the
8-50
-------
110
2 100
o
O
o
0)
Q.
01
O
•a
-------
Table 8-9. Polymorphonuclear Leukocyte and Protein
Permeability Response to Ozone by Specie^
Exposure
Parameters
0.4-0.6
2h
0.4
2h
0.4
2h
0.3
1 h
0.96
8h
0.2
0.5
1.0
2.0
4h
0.5
1.0
4h
0.5
0.8
2or7h
Exposure
Conditions
Exercise
(15 min/alt;
83-100W)
Exercise
(15 min/alt;
60 L/min)
Exercise
(15 min/alt;
60 L/min)
Exercise
(15 min/alt;
60 L/min)
Daytime;
rest
Daytime;
rest
Daytime;
rest;
Vitamin C
(AH2 + /D)
Daytime;
(15 min/alt
for 2 h;
45 min/alt
for 7 h);
3-5 x VE
Species
(Strain)
Human
Human
Human
Human
Rhesus monkey
(a) Mouse
(Swiss Albino)
(b) Guinea pig
(Hartley)
(c) Rat
(S-D)
(d) Hamster
(Golden Syrian)
(e) Rabbit
(NZW)
M
Guinea pig
(Hartley)
M
Rat
(F344)
M
Postexposure PMN* Protein*
BAL Time Increase Increase
3 h 7.8 Not done
18 h 8.0 2.2
Ih 18.2 1.2
1 , 6, 24 h 3.0 at 6 h No change
2.5 at 24 h
1,12,24, 27atlh 3 at 1 h
72, 168 h 19 at 12 h 3 at 12 h
24 at 24 h 8 at 24 h
6 at 72 h 3 at 72 h
3 at 168 h 1 at 168 h
16 h Not done (a) 1.8 at 1.0 ppni
3.2 at 2.0 ppm
(b) 1.4 at 0.2 ppm
2.0 at 0.5 ppm
4.1 at 1.0 ppm
4.5 at 2.0 ppm
(c)2.1 at 1.0 ppm
3. 6 at 2.0 ppm
(d) 1.5 at 1.0 ppm
2.6 at 2.0 ppm
(e) 2.7 at 2.0 ppm
16 h Not done For 0.5 ppm;
1.1 for AH2+,
2.1 forAH2n.
For 1.0 ppm;
2.4 for AH2+,
2.7 for AH2n
1 h Not done For 0.5 ppm;
1.2at6h,
2. 1 at 7 h.
For 0.8 ppm;
1.5at2h,
3. 3 at 7 h
Reference
Seltzer et al.
(1986)
Koren et al.
(1989a,b)
Koren et al.
(1991)
Schelegle
et al. (1991)
Hyde et al.
(1992)
Hatch et al.
(1986)
Slade et al.
(1989)
Costa et al.
(1988a)
"See Appendix A for abbreviations and acronyms.
bOzone response/air response.
8-52
-------
(Table 8-9) will be considered. The spectrum of exposure conditions used in the various
animal studies makes difficult the direct comparison among laboratory test species. However,
one study by Hatch and coworkers (1986) specifically addressed this question by exposing
mice, guinea pigs, rats, rabbits, and hamsters under identical conditions (0.2, 0.5, 1.0, and
2.0 ppm O3 for 4 h), followed at 18 h postexposure with BAL and assay for protein. Guinea
pigs were most responsive, responding at 0.2 ppm, whereas mice, hamsters, and rats began
responding at 1.0 ppm, and rabbits responded only to 2.0 ppm. Only one study involving
BAL assessment of PMNs and protein in exposed monkeys has been published (Hyde et al.,
1992). At 0.96 ppm O3 (8 h), the monkeys had a significant inflammatory response, but it is
difficult to assess monkey responsiveness relative to the human for this endpoint. Assuming a
linear concentration times duration relationship, the monkey data appear similar to the guinea
pig response. However, none of these species showed BAL protein increases approximating
those reported in human studies.
In recent studies (0.4 ppm O3 for 2 h, with BAL 16 to 18 h postexposure) (Slade et
al., 1989; Crissman et al., 1993), guinea pigs made vitamin C-deficient exhibited enhanced
responsiveness to O3, this result is comparable to that of the exercised humans of Keren et al.
(1989b) (Figure 8-10). Similarly, when rats were exposed to 0.5 ppm O3 for 2 h with
intermittent CO2-induced hyperventilation (Tepper et al., 1993) to mimic mild/moderate
exercise (three- to fivefold VE), the BAL protein, as well as PMN responses at 18 h
postexposure compared favorably with those data of Keren and coworkers (1989a,b) (Figure 8-
11).
Within a given laboratory animal species, responses among strains also can differ
appreciably, as demonstrated in rats by Pino et al. (1991) and Costa et al. (1993). These
studies indicated that Wistar rats exhibit greater inflammatory responses (protein and PMN) to
O3 than S-D and F344 rats after an 8-h exposure to 0.5, 1.0, and 1.5 ppm with BAL sampled
at 2 or 24 h later. Similarly, mouse strains (C3H/HeJ and B6C57/6; Kleeberger et al., 1990)
and S-D rat substrains (Costa et al., 1985) have been shown to possess specific genetic
susceptibility to high levels of O3 (2 ppm). In the case of the mice, the responsive strain is
seven times (at the 6-h postexposure peak) as susceptible as the databases in animals and,
particularly, in humans with regard to these antioxidants are quite limited. Supplementation
and deprivation studies with vitamins C and E have shown that these antioxidants have some
role in protecting against the effects of O3 in animals (Elsayed et al., 1988; Slade et al., 1989;
Crissman et al., 1993) and in humans (Chatham et al., 1987). Of the animal models,
ascorbate-deprived guinea pigs appear to have BAL ascorbate levels most like humans, with a
protein permeability response (without exercise in the animal) very similar to the human
exposed to the same concentration (0.4 ppm O3 for 2 h) with exercise. However, Crissman
et al. (1993) also resistant strain for the PMN response to 2 ppm O3; the protein response is
twice (at 24 h postexposure) that of the resistant strain. In the S-D substrain, protein
extravasation into the alveolar lumen immediately postexposure is 40% higher in the
responsive strain than the resistant (no other time points were examined).
Humans have an order of magnitude less ascorbate in BAL fluid as compared to the
rat, but they have nearly twice the glutathione, four times the uric acid, and 80 times the
vitamin E, as normalized to lipid P-surfactant (Slade et al., 1993) (Table 8-8). However, on a
BAL-derived cell/protein basis, the ratios clearly favor the rat for all of these antioxidants,
with the exception of uric acid, which is generally not high in rats because of species
differences in protein and prime nucleotide catabolism (urea being the major nitrogenous
8-53
-------
Basal Levels of Ascorbic Acid
Humans
Rats
Normal
AH2-Deficient
Guinea Pigs
Effect of Ozone Inhalation (0.4 ppm, 2 h)
on BAL Protein
c
o
o
03
D)
c
CD
c
'CD
-I—*
S.
Q.
_l
<
CD
Humans
Rats
Normal
AH2-Deficient
Guinea Pigs
Figure 8-10. Composite of data from Slade et al. (1989), Koren et al. (1989b), and
Crissman et al. (1993) comparing basal bronchoalveolar lavage (BAL)
ascorbate levels (A) to ozone-induced changes in BAL protein (B). Ozone-
exposures (0.4 ppm; 2 h; 16 to 18 h BAL) of humans (exposed with
exercise), rats (exposed resting), and guinea pigs (exposed resting) with
(ascorbic acid [AHJ-deficient) and without (normal) AH2 deficiency.
8-54
-------
0 0.5 1.0 1.5 2.0
Ozone (ppm)
Figure 8-11. Composite of data comparing polymorphonuclear leukocytes obtained by
bronchoalveolar lavage 16 to 18 h after ozone (OJ exposure (0.4 ppm; 2 h)
of humans (with exercise) (Koren et al., 1989b) to those of rats exposed to
O3 (0.5 to 2.0 ppm) at rest or hyperventilated with carbon dioxide (CO2)
(Tepper et al., 1993).
by-product for rats). The guinea pig most closely resembles the human for ascorbate. Because
these antioxidants are thought to function in the defense against oxidant challenge, it would
appear critical to appreciate their presence and function when attempting to interpret data for
extrapolation. Unfortunately, the reported that reduced ascorbate levels in BAL fluid (18 h
postexposure) increase in the human, whereas those levels decrease in the rat. Whether this
relates to the distinctly different basal levels of this vitamin and is associated with the disparate
protein responsiveness (ignoring exercise) is unclear because deficiency in animals (guinea
pigs) appears more critical to the responsiveness at low (ambient-like) concentrations than at
higher concentrations (1.0 ppm). Although it would appear that vitamin C is involved in the
interplay between O3 and the exposed subject (human or animal), there is not full coherence of
the data. For example, Hatch et al. (1986) showed that the rabbit was the least responsive to
O3 in terms of BAL protein, but this species has among the lowest tissue levels of vitamin C
(Slade et al., 1985). However, rabbits apparently have a low propensity to form lipid
8-55
-------
peroxides (Arakawa et al., 1986), an expected product of lung lipid-O3 interaction (Pryor,
1992). Thus, to interpret inter species sensitivity only in terms of basal antioxidant levels,
however tempting this may be, would be overly simplistic and premature at this time.
8.3.3.3 Prolonged Exposure Studies
The previous sections of this text have utilized lung function responses to acute
O3 exposure in an attempt to elucidate the relative sensitivity among laboratory test species and
humans and, thereby, to complete two corners of the parallelogram paradigm (Figure 8-6).
Because acute responses represent only part of the extrapolation paradigm, temporal-based
exposure responsiveness also will be considered, despite the relative paucity of comparative
data between laboratory animals and humans for prolonged exposures. Again, in an effort to
best extract species-specific response differences, the criterium for selection of studies was
similarity in exposure scenario. The discussion will focus on relative adaptability of acute
functional changes and associated BAL-derived findings after repeated exposures and the
coherence of the findings from prolonged human and animal exposure studies and
epidemiological results.
Lung Function Studies
Reversal ("attenuation") of pulmonary function decrements using a scenario of
repeated exposure to O3 has been reported for both humans and laboratory animals. At least
nine studies between 1977 and 1984 have documented that, for repeated exposures between
0.2 and 0.5 ppm O3 (2 h/day, up to 5 days), spirometric changes were most severe on the first
or second day of exposure, waned over the next 3 days of exposure, and, by the fifth day, had
returned to control preexposure levels (Chapter 7, Section 7.2.1.4). In the only animal study
using a similar exposure protocol and analogous experimental design, Tepper et al. (1989)
showed that rats initially displayed a tachypneic response to O3 that attenuated after 5
consecutive days of exposure, a pattern quite similar to that of humans. Exposures were for
2.25 h and included challenge with CO2 during alternate 15-min periods to augment ventilation
(2 to 3 X resting VE, which is equivalent to light exercise in humans). As in the human
studies, the functional changes were largest on Day 1 or 2, depending on the parameter and the
O3 concentration (0.35, 0.50 and 1.00 ppm). Attenuation of the changes in shape constant of
the flow-volume curve of the rats also was observed over this period. Thus, under analogous
conditions of exposure, both the humans and the rats exhibited similar initial functional
responses to O3 with full and kinetically similar reversal of effects.
More difficult is the direct comparison between human and animal lung function
responses to prolonged (several-week) O3 exposure, largely because of the limited availability
of controlled human study data. In the only study of its kind, Bennett (1962) exposed
12 human subjects at rest to 0.2 or 0.5 ppm O3 for 3 h/day, 6 days/week for 12 consecutive
weeks. Although no effects were discernable early in the exposure, there appeared to be
small, but significant O3-induced reductions in FEV, (and small, nonsignificant reductions in
FVC), particularly in the last weeks of the study. This reduction in FEV, more likely would
reflect obstructive changes within the lung at these points late in time rather than the pain-
mediated reductions that are seen with acute O3 exposure, which attenuate after a few days of
exposure. The lack of concentration-related decrements in FEV, and FVC is somewhat
unsettling, but, regardless, after 9 postexposure weeks in clean air, all measured effects had
reversed.
8-56
-------
Unfortunately, there are no directly parallel animal studies to compare to this
limited database. But, if the two-fold sensitivity difference between the rat and human FVC
response (see Figure 8-12) is assumed, a number of animal studies may be considered
comparable for the purposes of this discussion. On the one hand, rats exposed to 0.2 or
0.8 ppm O3 for 6 h/day, 5 days/week for 12 weeks were reported to exhibit some degree of
small airway obstruction based on the MEFV curves, but little, if any, reduction in FVC or
DLCO was observed (Costa et al., 1983). Others have reported analogous marginal increases in
rat TLC or its component volumes (Bartlett et al., 1974; Costa et al., 1983; Raub et al., 1983)
or in regional Raw (Yokoyama et al., 1984) after intermittent or continuous exposures to DO.25
ppm O3 for 4 to 12 weeks, which would not be unexpected with distal airway or lung damage.
Actual pathology in the distal lung tends to be focal and difficult to correlate precisely with the
marginal functional impairment.
120
o
i 110
o
O
=5 100
CO
Q.
03
O
CO
T3
CD
O
90
80
Human
200
150 |
o
O
100 f
.c
*j
CO
CD
m
Rat, Light, 0.35 ppm
Tepperetal. (1989)
Rat, Light, 0.5 ppm
Tepperetal. (1989)
Rat, Light, 1.0 ppm
Tepperetal. (1989)
Human, Moderate, 0.35 pprr
Folinsbee et al. (1980)
50
u
c
CD
CD
Human, Light, 0.5 ppm
Hackney etal. (1977)
Human, Moderate, 0.5 ppm
Folinsbee et al. (1980)
3
Days
0
Figure 8-12. Comparison of changes in forced vital capacity in humans (left ordinate)
and frequency of breathing in rats (right ordinate) with up to 5 consecutive
days of ozone (O^ exposure. Data are expressed as percent of the control
response. Right hand legend indicates species, exercise level,
O3 concentration, and the reference. Human data are plotted with solid
lines and open symbols, whereas rat data are plotted with broken lines and
closed symbols.
8-57
-------
The limited functional data available in monkeys generally agree with the pattern of
distal lung pathophysiology reported in rats. When exposed to 0.5 ppm O3 for 90 days
(8 h/day), monkeys exhibited a slight increase in lung distensibility (Eustis et al., 1981).
Likewise, monkeys exposed to 0.25 ppm (8 h/day for 18 mo) exhibited increased chest wall
(but not lung) compliance and lung volumes, which was most marked in the monkeys exposed
to O3 only in alternate months during that 18-mo period (Tyler et al., 1988). Recall, however,
that little or no change in lung elasticity has been associated with controlled O3 exposures in
humans, whether the exposures are acute or repeated. On the other hand, higher
concentrations (0.64 ppm for 1 year) resulted in the alteration of distal airway mechanics in
exposed monkeys, as gleaned from local resistances measured using oscillatory methods;
again, this is in general agreement with the presence of distal lung injury or disease.
Morphometric analyses of the end-airways and distal lung regions of O3 exposed monkeys
consistently show altered cell profiles and interstitial restructuring, even when functional
changes are marginal, which, like in the rat, likely reflects the large functional reserve of the
integrated lung. Thus, although these collective data from subchronic animal studies suggest a
reasonably homologous distal lung response to O3, many of these linkages in functional
outcomes remain uncertain in terms of what to anticipate in the human response.
Clearly, the question of potential lung impairment resulting from a near-lifetime
exposure to O3 ranks among the most pressing concerns about this toxicant. The animal data,
although demonstrating that chronic O3 exposure can induce changes in the structure and
function of the lung, have yet to provide evidence of potential disease or disability in humans
exposed to O3 over prolonged periods of their lives. The existing epidemiologic studies
(Chapter 7, Section 7.4.2), too, merely provide suggestive evidence that persistent or
progressive deterioration in lung function may be associated with long-term oxidant pollutant
exposure (Detels et al., 1981, 1987). Detels and coworkers (1991) reported decrements in
FEV, and nitrogen washout across all age groups in areas where oxidant pollution is high.
Similarly, analysis of the pulmonary function data from the National Health and Nutrition
Examination Survey (NHANES) II showed loss of lung function when annual averages
ofambient O3 exceeded 0.04 ppm (Schwartz, 1989). This pattern of impairment is consistent
qualitatively with the chronic animal studies (Costa et al., 1994).
Studies of Inflammation and Antioxidant Content of the Bronchoalveolar Lavage Fluid
The virtual absence of human BAL study data after repeated or prolonged exposures
to O3 hinders the comparison of nonacute human and animal inflammatory responses.
However, the recent study of Devlin et al. (1995) suggests that the PMN and protein responses
to repeated daily exposures to 0.4 ppm O3 (2 h with IE for 5 consecutive days) attenuate, much
as do the functional responses. Hence, for most of the BAL parameters (with the exception of
lactate dehydrogenase activity [a marker of cell injury]), there is indeed an apparent reversal of
acute inflammation when exposures are continued over the 5-day exposure period. Rat studies
largely appear to show similar attenuation to O3, but this response seems to be influenced by
exposure patterns or conditions (Bassett et al., 1988; Tepper et al., 1989; Van Bree et al.,
1989). The study with the most similar design to the human protocol (Tepper et al., 1989)
showed some reduction in BAL protein with repeated exposure involving intermittent
hyperventilation with CO2, but over the 5-day period, the protein levels remained significantly
elevated; cells were not evaluated in this study. Interestingly, vitamin C and glutathione levels
8-58
-------
in the BAL fluid increased over the 5-day course of exposure, a response consistent with an
unregulated antioxidant role in the adaptative mechanism. Van Bree and coworkers (1992), on
the other hand, reported that 5 consecutive days (12 h/day) of 0.4 ppm O3 resulted in complete
reversal of both BAL albumin and PMN measures to the control values. It should be noted,
however, that other biomarkers and mediators within the BAL were not fully recovered, which
might suggest a slower reversal time frame or continued O3-induced pathogenesis, a conclusion
of the Tepper et al. (1989) study.
In guinea pigs made deficient in vitamin C and exposed to O3 (0.2, 0.4, or
0.8 ppm) continuously for 7 days, attenuation of the functional and inflammatory endpoints
appeared nearly complete in spite of the deficiency (Kodavanti et al., 1995). Other
antioxidants, not altered basally, were unregulated more by the O3 challenge; the small residual
reservoirs of ascorbate, which persisted in the nearly 98% deficiency state of the animals, were
apparently mobilized to the site of injury, allowing repair to proceed. Likewise, chronically
exposed rats have elevated BAL ascorbate indicative of the oxidant burden and the ongoing
repair (Grose et al., 1988). Prolonged exposures up to 18 mo appear to sustain a low-grade
interseptal inflammation and evidence of lung matrix remodeling in both rats and monkeys,
suggesting that humans would behave similarly. However, such data are not presently
available from humans.
8.4 Quantitative Extrapolation of Acute Ozone Effects
8.4.1 Introduction
Advances in dosimetry since the previous O3 criteria document (U.S. Environmental
Protection Agency, 1986) fall into five major areas: (1) greater sophistication of model
applications (e.g., Overton et al., 1989; Mercer et al., 1991), (2) the appearance of
experimental uptake data that can be compared to model predictions (Wiester et al., 1987,
1988, 1996; Hatch et al., 1989; Gerrity et al., 1988, 1994, 1995), (3) experiments specifically
designed to estimate model parameters (Hu et al., 1992b), (4) a better understanding of the
role of O3 in the liquid linings and tissues of the RT (Pryor, 1992), and (5) a better
understanding of anatomical aspects (Mercer et al., 1991). The role of these advances in
interspecies dosimetric extrapolation follows.
With the information available for rats, reasonably reliable predictions of the flux of
O3 to the air-liquid lining interface of toxicologically important regions, such as the CAR, is
possible. There are two main investigations that make this feasible: (1) Hatch et al. (1989),
who estimated the percent uptake and the fraction of the retained O3 that is in the URT,
trachea, and lung of rats, and (2) Pinkerton et al. (1992) (with elaboration by Miller et al.,
1993) who illustrated the basic correctness of modeling assumptions for ventilatory units. (The
judgment of basic correctness is based on the assumption that the dose causing the response is
proportional to the flux of O3 to the air-liquid lining interface.) Using this information,
regional mass transfer coefficients could be estimated, which would allow the prediction of
local respiratory tract O3 doses in rats exposed under general conditions.
The results of the investigation of Hu et al. (1992b) can be used to estimate URT
and TB model parameters, but may not be sensitive enough to determine pulmonary region
parameters. If uptake is not confined to the URT and TBs, then their mass transfer coefficients
alone would not be sufficient to account for total RT uptake (e.g., as measured by Wiester
et al. [1996] or Gerrity et al. [1988]), and the difference in predicted (without pulmonary
8-59
-------
region uptake) and experimental uptakes could be used to estimate the pulmonary region mass
transfer coefficient. Unfortunately, no dosimetry data for the human pulmonary region are
available.
Despite limitations, the O3 dosimetry data that have been obtained over the past
several years, coupled with the advances in modeling, suggest that there has been continual
convergence between the model predictions and experimental observations. Given the many
areas of consistency between models and experiments, it is valuable to begin to employ these
models to provide the dosimetric basis for animal-to-human extrapolation. One of the greatest
sources of uncertainty in such an application of dosimetry models is the lack of full
understanding of the appropriate target site for O3 toxicity (e.g., upper or lower airways or
pulmonary region) that initiates a particular response, especially the functional changes in the
lung. However, reasonable assumptions can be made to narrow the target site.
The first application of dosimetry models given in Section 8.4.2 is an examination
of delivered dose versus response within a given species. This is followed in Section 8.4.3 by
some inter species comparisons of delivered dose versus response.
8.4.2 Intraspecies Delivered Dose Response
Assuming that changes in FEV, in humans exposed to O3 mainly are the result of
O3 pulmonary tissue dose, Miller et al. (1988) constructed a dose-response curve. They
plotted decrements in FEV, versus predicted cumulative pulmonary region O3 tissue dose
scaled to body mass (Figure 8-13). The concentration-response data are from McDonnell et al.
(1983), in which 135 healthy subjects were exposed to 0, 0.12, 0.18, 0.24, 0.30, and
0.40 ppm O3. Exposure was for 2.5 h with heavy IE. Miller et al. (1988) used the average
weight and height of the subjects to estimate the FRC that was used in the model to simulate O3
dose. The exercise breathing parameter data were used along with an estimate of resting
breathing parameters. Figure 8-13 is similar in shape to the concentration-response curve of
McDonnell et al. (1983). Differences between these two curves, however, are accounted for
by the translation between exposure concentration and O3 dose.
As another example, Pinkerton et al. (1992) examined the relation between actual
tissue response of rats chronically exposed to O3 and a prediction of O3 dose as a function of
distance from the bronchiole-alveolar duct junction (BADJ) to ventilatory units. Using these
data, Miller and Conolly (1995) plotted the predicted O3 dose and the observed change in wall
thickness due to the exposure versus distance from the BADJ (Figure 8-14). Even though
considerable variability in the thickness change can be inferred from the data, the two curves,
scaled to their values at the junction, show a remarkable similarity and suggest a basic
correctness in regards to the ventilatory unit model parameters.
8.4.3 Interspecies Delivered Dose Response
The illustrations presented in this section are based on dosimetric estimates for
humans and rats using existing or modified theoretical models (Miller et al., 1985; Miller
et al., 1988; Overton et al., 1987). One functional and one inflammatory endpoint will be
provided drawing from the f and BAL protein data described in Section 8.3. Because the
diversity of exposure scenarios across species is so great, the window of exposure parameters
has been narrowed, minimizing exposure-based differences in relating species responsiveness.
8-60
-------
1.0
0.9-
0.8-
0.7-
0.6-
I?
~0.5-
LU
< 0.4-
0.3-
0.2-
0.1-
0
J I I I I I I I I I I I
\
12
\
14
\
18
4 6 8 10 12 14 16
Pulmonary Tissue Dose (ng O3 /g body weight)
20 22 24
Figure 8-13. Changes in forced expiratory volume in 1 s (AFEVJ versus pulmonary tissue
dose. Plotted are decrements in FEVt (liters) for human subjects against
predicted pulmonary tissue dose normalized to body weight. In order,
from left to right, the dose values correspond to 0, 0.12, 0.18, 0.24, 0.30,
and 0.40 ppm ozone (O3) exposure concentrations. The continuous curve
was an "eye fit".
Source: Miller et al. (1988)
Because the theoretical models developed by Miller et al. (1985, 1988) and Overton
et al. (1987) estimate the dose distribution to the RT on a per breath basis, a minimum
quantitative description of tidal breathing (VT and f) is needed to utilize the theoretical models
of O3 deposition. The need for detailed breathing parameters, therefore, severely restricts the
application of the model to studies providing such data. Unfortunately, breath-by-breath
parameters over the course of an exposure normally are not measured or reported in most
publications. Two studies each in humans and rats, providing adequate detail over the course
of the exposure, allowed the model computations to be performed for this illustration (Figure
8-15). The human studies examined were DeLucia and Adams (1977), who exposed the
subjects to 0.15 and 0.30 ppm O3 for 1 h with continuous exercise (65% oxygen uptake to the
body) periods, and Beckett et al. (1985), who exposed the
8-61
-------
r125
o n -I 1 1 1 1 1 1 1 —••• 1- o
100 200 300 400 500 600 700 800 900 1,000
Distance from Bronchiolar-Alveolar Duct Junction (urn)
Figure 8-14. Relationship between change in alveolar wall thickness and predicted
ozone (O-) dose as a function of distance from the bronchiole-alveolar duct
junction (BADJ). Rats were exposed to 0.98 ppm O3 for 8 hi day for
90 days.
Source: Miller and Conolly (1995).
subjects to 0.4 ppm O3 with alternating 15 min exercise (70 L/min). Three rat studies were
evaluated: (1) Tepper et al. (1989), who exposed rats to 0.35, 0.50, and 1.00 ppm O3 for 2.25
h with alternating 15-min periods of CO2-induced hyperventilation (2 to 3 X resting VE); (2)
Tepper et al. (1990), who exposed rats to 0.12, 0.25, 0.50, and 1.00 ppm O3 for 2.25 h with
alternating 15-min periods of CO2, such that subsequent periods of CO2 exposure had higher
CO2 concentrations than previous periods; and (3) Mautz and Bufalino (1989), who exposed
rats to 0.8 ppm O3 for 3 h (at rest). The response parameter was the ratio of O3-altered f (f0 \
to control f (fcnt). Dose rate was the average dose to the proximal alveolar region (PAR)
computed from time 0 to the time of the maximum f03/fcnt The PAR was chosen as the target
based on the perception that O3 acts on deep lung stretch receptors
8-62
-------
o
"co
2.2
2.0
o 1.8
-------
been supplemented in Figure 8-16 to include the results of the human study of Keren et al.
(1989b). As can be seen in the illustration, there appears to be a log-normal relationship
between BAL protein and dose to the pulmonary region, the purported site of plasma leakage
to the airspace lumen. This relationship would support the contention that there is a
mechanistic consistency in response across species that may exhibit a quantitative sensitivity
factor for use in further quantitative interspecies extrapolation. This sensitivity factor is
evident from the clustering of data from different animal species.
3,000
! 1,000-
-------
protein in BAL fluid, because these effects are measurable in both human and nonhuman
species. Of laboratory species commonly used to address questions of acute toxicity, rats have
the strongest and most compatible database for such comparisons. Studies in nonhuman
primates typically are limited to morphology endpoints and to longer exposure periods, thereby
limiting the practical utility of these studies in this same context. A conclusion drawn from the
discussion of acute effects is that there is reasonable semi-quantitative agreement and homology
between species with regard to their functional and permeability responses to short-term O3
exposure. Where data from long-term exposures exist, cross-species relationships among
similar endpoints (usually functional) are considerably weaker, although there is the suggestion
that long-term exposure to O3 can alter the distal lung (that zone where theoretical models
would predict greatest O3 deposition).
This section continues the rationale founded in the extrapolation paradigm
illustrated in Figure 8-6 in an attempt to quantitatively address the question of potential chronic
alteration to the lungs of O3-exposed people. Direct data from epidemiological studies remain
suggestive at best (see Chapter 7). On the other hand, the extensive database of morphometric
effects on the distal lung of exposed animals (rodents and nonhuman primates) reveals the
sensitivity of such endpoints and clearly relates to the site of the lung associated with incipient
chronic lung diseases associated with known toxic inhalants (e.g., tobacco smoke).
Unfortunately, there is a lack of reliable morphometric data on the human lung that are
associated with O3 or other air pollutants. Thus, the goal here is to draw from the extensive
and reliable animal database, which demonstrates chronic effects of long-term exposures to
O3, and project from it the potential for chronic effects in humans utilizing the linkage
provided by newer, refined O3 dosimetry estimates.
In the long-term studies selected for detailed analysis, great importance was placed
on the relevance of the exposure concentrations, the site specificity of the morphometric
analysis, and the consistency of analysis within species. Two rat studies were selected that
represent near-lifetime exposures to O3 over a range of concentrations and scenarios (Chang et
al., 1992; Chang et al., 1995). The former study was conducted in conjunction with the U.S.
Environmental Protection Agency (EPA) and involved a weekday urban pattern of exposure
represented by a 9-h spike (5 days/week) slowly rising to 0.25 ppm from a near-continuous
baseline of 0.06 ppm O3 for 78 weeks. The latter study was conducted by the same
morphometrists but on rats exposed weekdays (6 h/day) to 0, 0.12, 0.50, or 1.00 ppm O3 for
87 weeks. Both studies utilized F344 rats. Likewise, the primate studies were conducted by
the same investigators, but involved two strains of monkeys. In one study, the bonnet monkey
was exposed for 90 days (8 h/day) to 0, 0.15, or 0.30 ppm O3 (Harkema et al., 1993), whereas
the second study consisted of daily (8 h/day) exposures to the smaller fascicularis monkey of
0 or 0.25 ppm O3 for 18 consecutive mo or 9 mo with alternating months of clean air (Tyler et
al., 1988). The assumptions needed to model the specific species dosimetry of each study are
provided in the following section. Growth was compensated where appropriate (for rats), and
allometric anatomic adjustments or assumptions were made to estimate unavailable anatomic
data (as in the case of the monkey) for the dosimetry models. Ventilation was assumed to be
unaffected by O3 after the first 2 days of exposure because adaptive events are known to occur
in that time frame. This simplified the model formulations for the animal studies, although
varied activity and exposure profiles were considered throughout for the human dose estimates.
Allometric equivalent life-span estimates also were made in an effort to relate the duration of
the exposure period relative to life-span and cumulative dose.
8-65
-------
In each experiment, a subregion of the CAR (where major lesions have been
observed) was chosen by the investigators for study. For the two rat studies, the site was the
PAR, which was defined by Chang et al. (1992) as "the alveolar tissue surrounding the
alveolar ducts beginning at the bronchiolar-alveolar duct junction and ending at the second
alveolar duct bifurcation." For modeling purposes, the first generation distal to the terminal
bronchioles corresponded to the PAR. The monkey studies focused on an analogous airway
region, the respiratory bronchioles. Thus, in three of the four experiments, the first generation
distal to terminal bronchioles was the explicit site of the observed effects. For simplicity,
similar assumptions were made for the Tyler et al. (1988) monkey study that came from the
same laboratory as Harkema. For discussion purposes, the term PAR has been used for the
monkey sites, even though the airway morphology of this site differs between the rat and
monkey. The simulated dose of the PAR was chosen for comparison to the reported effect.
Because the PAR generally is thought to be the site of incipient lung disease, it is of
particular interest with regard to the potential role of O3-induced chronic lung disease. Effects
in the PAR can be evaluated specifically using morphometric techniques with an electron
microscope, and, likewise, dosimetric models can estimate surface-area-normalized focal tissue
doses within the same region. The dose-response relationships constructed in this
extrapolation focus on the theoretical dose to the PAR of rats and monkeys, with their
corresponding impact on the PAR interstitial structure (as total and acellular thickness).
Analogous estimated cumulative doses to the PAR of a 9-year-old child and of an outdoor
worker exposed to a New York City summer O3 season then are interpolated from the effects
in the experimental animals at nominally similar doses.
In this attempt to extrapolate chronic O3 effects from the experimental animals to
humans, the two foremost assumptions are that (1) there exists coherence of analogous dose-
response data across species with regard to acute exposure effects (as represented by one side
of the parallelogram model [Figure 8-6]), and (2) there exists some relationship of effect to the
total cumulative dose to the target tissue. The essential hypothesis is that the rate of change of
interstitial thickness is related to the rate of O3 uptake. Clearly, the latter assumption ignores
potential adaptive or reparative processes, but rather assumes that continuing exposure of any
scenario imparts irreversible or slowly reversible changes within the constitutive structure of
the target area. There is little to substantiate this assumption other than the commonly believed
irreversibility of fibrogenesis and degenerative lung disease.
8.5.2 Factors Considered in Estimating Dose
For each of the animals for which doses were simulated, Table 8-10 lists
information or parameters needed for the dosimetry simulations and the source of the
information. Also, details are provided as to how this information was modified, scaled, or
used to correspond to the experimental or hypothetical animals for which dose was estimated.
An expanded discussion follows.
8.5.2.1 Human
The dose simulations were for a hypothetical New York City adult outdoor worker
and a hypothetical 9-year-old New York City child, who is active out of doors. Their
8-66
-------
Table 8-10. The Basis of Information for Model Parameters
Information Needed
Characteristics of
simulated animal
Exposure regime
LRT structure
Dead space volume
atFRC
FRC
URT volume (V)
and surface area (S)
TB volume (V) and
surface area (S)
PUL region volume
(V) and surface area
(S)
F344 Rat
(EPA Chronic Study)
Chang et al. (1992); Tepper
et al. (1991): 0.21-0.47 kg,
10-88 weeks old, males.
Chang et al. (1992); 0.06 ppm
continuous background,
22 h/day, 7 days/ week;
0.25 ppm ramped peak over 9 h,
5 days/week. 3-, 13-, and 78-
week exposures.
F344 Rat
(NTP/HEI Chronic Study)
Chang et al. (1995); Last et al.
(1994): 0.13-0.53 kg,
7-93 weeks old, males.
Chang etal. (1995); 0, 0.12,
0.50, and 1.00 ppm 6 h/day,
5 days/week for 87 weeks.
PUL: Mercer et al., 1991 (0.283-kg S-D rat).
TB: Yeh et al., 1979 (0.33-kg Long-Evans rat).
Dead space volume = URT + TB volumes.
Mercer et al., 1987 (0.291-kg S-D rat).
FRC D weight055 (Takezawa et al., 1980).
Patra et al., 1987 (F344 rat, 0.012-0.366 kg, 1-63 weeks old).
Graphical interpolation and extrapolation with respect to age.
V: Mercer et al., 1994a (0.293-kg S-D rat).
S: Yehetal., 1979 (0.33-kg Long-Evans rat).
S D Vm.
V: FRC minus TB volume.
S D V™.
Bonnet Monkey
(Macaco radiata)
Harkema et al. (1993):
2.3-9.7 kg (used 6-kg),
22-35 weeks old, males
and females.
Harkema et al. (1993);
8 h/day for 90 days; 0,
0.15, and 0.30 ppm.
Monkey
(Macaco fascicularis)
Tyler et al. (1988): 1-2 kg
(used 1.6-kg),
30-107 weeks old, males.
Tyler et al. (1988);
8 h/day for 18 mo; 0 ppm,
0.25 ppm (daily), and 0.25
ppm (daily, alternating
months).
Human Child
Present simulation:
31 kg, 9 years, 132-cm
height, male or female.
Johnson (1994);
continuous exposure
pattern for a New York
outdoor child; April-
October, 1991. Avg.
exposure 22 ppb; range,
0-238 ppb.
Human Adult
Present
simulation: 73 kg,
181-cm height,
adult male.
Johnson (1994);
continuous
exposure pattern
for a New York
outdoor worker.
April-October,
1991. Avg.
exposure, 23 ppb;
range, 0-227 ppb.
Weibel (1963): adult human LRT.
Dead space volume = URT + TB volumes.
Kosch et al., 1979 (radiata
monkeys, FRC = 52.8
mL/kg).
Tyler etal., 1988 (2-kg
fascicularis). FRC D
weight086 (relation for
combined rodent species,
Takezawa et al. (1980).
Schreider and Raabe, 1981 (7-kg rhesus monkey).
V D weight;
S D Vm.
V: Pulmonary and TB volumes in same ratio as
human.
S D V™.
Pulmonary and TB
volumes in same ratio as
human.
S: Mercer et al. (1994b),
interspecies interpolation.
V: Pulmonary and TB
volumes in same ratio as
human.
S: Tyler etal., 1988
(2-kg fascicularis).
S D Vm.
Hart et al. (1963): height: 92-198 cm, age:
4-42 years, weight: 16-1 15 kg.
V: The ratio of URT
and dead space volumes
are the same as in the
adult human. S D V™
Hu et al. (1992a).
V: volume = dead space volume minus
URT volume.
S D V™.
V: volume = FRC volume minus dead
space volume.
S D V™.
oo
-------
Table 8-10 (cont'd). The Basis of Information for Model Parameter!
Information Needed
Tidal volume and
breathing frequency
Mass transfer coefficients
F344 Rat
(EPA Chronic Study)
Costa (1994): measurements made at
1, 3, 13, 52, and 78 weeks of
exposure (interpolation for other time
points).
F344 Rat
(NTP/HEI Chronic Study)
None reported. Assumed
similar to that of the EPA
chronic study.
Parameter estimation using the rat data of Hatch et al. (1989) and an
assumed pulmonary region coefficient. See Table 8-6 and
Section 8.5.2.3.
Bonnet Monkey
(Macaco radiata)
Monkey
(Macaco fascicularis)
None reported. A range of parameters used.
See Table 8-11 and Section 8.5.2.2.
Human Child
Johnson (1994):
activity pattern.
Hofmann et al.
(1989): tidal
volumes and
breathing frequences.
For corresponding generations or model segments, the same as for the
adult human. See Table 8-6 and Section 8.5.2.1.
Human Adult
Johnson (1994):
activity pattern.
ICRP(1975): tidal
volumes and
breathing
frequences.
Hu et al. (1992b);
Miller et al. (1985);
Weibel (1963).
See Table 8-6.
"See Appendix A for abbreviations and acronyms. PUL, V, and S indicate pulmonary, volume, and surface area, respectively. In some cells of this table, lie information is ordered as follows:
characteristics of the species and the reference to the basis of the information, followed by an indication as to how the reference information was used or modified for the present simulations.
The proportion symbol D indicates that one parameter is proportional to another (e.g., S D V™ implies that S = S0 (V/V0)m, where the subscripted parameters are known; V is the new
or desired volume and S is the estimated surface area of the new volume.
oo
ON
oo
-------
residences were not air-conditioned, and they lived, worked, or played in the same area.
These two groups of people would have tended to experience higher outdoor O3 concentrations
than other New York City people because the chosen area had higher outdoor O3
concentrations than other areas, and O3 levels are higher in homes without air-conditioning
(Johnson, 1994). These scenarios were selected because they involve the same at-risk
population groups used for the O3 risk analysis in the staff paper (U.S. Environmental
Protection Agency, 1995).
The activity patterns and O3 exposure estimates were generated by the probabilistic
National Ambient Air Quality Standards Exposure Model for O3 (pNEM/O3) (Johnson, 1994)
for April through October 1991. The modeling approach for outdoor workers can be described
as follows. Outdoor workers residing in New York City were divided into nonoverlapping sets
of cohorts, such that each cohort could be identified by a residential location, a work location,
and a residential air conditioning system. The pNEM/O3 model generated an "exposure event
sequence" (EES) for each cohort based on data obtained from activity diary studies involving
outdoor workers. Each exposure event in the EES assigned the cohort to a geographic
location, a microenvironment, and an exertion level. Algorithms within the pNEM/O3 model
provided estimates of O3 exposure and equivalent ventilation rate (EVR) for each event that
lasted from 1 to 60 min. The EVR is ventilation rate (liters per minute) divided by body
surface area (square meters). Exposures were calculated on a minute-by-minute basis;
however, the concentration and EVR values of all minutes within a given exposure event were
held constant during the event. The pNEM/O3 model used for outdoor children was generally
consistent with the model for outdoor workers but relied on human activity data from children.
Another difference was that the cohorts of children were identified by residential location and
air conditioning system only; workplace location was not specified.
Figures 8-17A and 8-17B give an idea of the exposure concentrations for the
outdoor worker and child, respectively, with plots of the average hourly and monthly
O3 concentrations. The exposure concentrations of the adult and child differ because their
daily activities and locations are different. Note that in a few cases, the hourly average
exceeds the 0.12 ppm (120 ppb) standard.
Anatomical Aspects
The adult New York City worker was assumed to have characteristics similar to the
subjects of an investigation by McDonnell et al. (1983). The body surface area of the worker
(needed to convert EVR to minute volume) was assumed to be 2 m2 (Keren et al.,1989b Dl.90
and Dl.98 m2; Johnson, 1994: 1.90 to 1.95 m2). The New York City child had a body surface
area of 1.07 m2, which was consistent with both 9-year-old males and females (Johnson, 1994).
The assumed values for the child's height (132 cm) and weight (31 kg) were based on Phalen et
al. (1985) and Johnson (1994).
For the simulations, no distinction was made between mouth and nose breathing
with respect to URT uptake. The significance of this assumption is not clear. If both modes
have approximately the same uptake efficiency, then the assumption of no distinction between
mouth and nose breathing is appropriate for predicting PAR doses. Whether this is a valid
assumption has not been settled (see Section 8.2.3.4). The same URT anatomical model was
used for both ages, but was isotropically scaled as necessary. Surface areas and lengths were
assumed to be proportional to the volume to the two-thirds and one-third powers for adults and
children, respectively (e.g., if x is proportional to volume V to the power p, then
8-69
-------
2-
Q.
Q.
^
o
^^
to
+-»
c.
o
0
0
Q.
X
LU
^U
200 -
180 -
160 -
140 -
120 -
100 -
80 -
60 "
Ar\ -
4U
20 "
0
A
•-.*•. ,- .
-.• • • '^- -'••: ' • . -' i*1^' *"" «*'1'»* ••: '' tt '•::."- •'•
1 APR rMAY ' JUNE1 JULY1 AUG ' SEPTr OCT '
60 90 120 150 180 210 240 270 300 330
Days
Figure 8-17. The variation in exposure concentration for the New York City adult and
child. Plotted are the hourly and monthly O3 averages for the New York
City outdoor worker (A) and the 9-year-old child (B). The hourly averages
are represented by the "dots" and the monthly average by the continuous
solid line.
8-70
-------
xn = XoCVn/Vo)13, where the subscripts n and o indicate the new and original values,
respectively). Dimensions used for the adult were from Hu et al. (1992b), who based their
values on the acoustic reflection measurements of Fredberg et al. (1980) that corresponded to
the oropharyngeal (mouth through larynx) region. The LRT model developed by Weibel
(1963) was used for LRT structure and dimensions, but was scaled isotropically to FRC and
dead-space volumes, as appropriate.
Mass Transfer Coefficients
This parameter is necessary to determine the local dose, expressed in terms of the
flux of O3 to respiratory tract surfaces. Specifically, this dose is related to the product of the
mass transfer coefficient and gas-phase concentration of O3 integrated over the time of interest.
Good estimates of mass transfer coefficients are needed to predict a reliable dose at specific
respiratory tract sites.
A brief description of how the human URT and TB region mass transfer coefficients
were derived from Hu et al. (1992b) and Weibel (1963) is presented in Table 8-6. These
coefficients were "validated" by simulating the bolus-response experiment of Hu et al. (1992b).
Simulation results of the bolus uptakes overpredicted by D5 %; results for the other two
measured variables deviated from the experimental values by D30%.
As a result, simulations with inhaled and exhaled flow rates of D250 mL/s are
expected to overpredict total uptake by no more than 5 %. The significance of the poor results
in simulating the other two variables is not clear. Possibly, predictions of the distribution of
absorbed O3 within the RT are questionable; also, predicted total uptakes for flow rates
different from the experimental value of 250 mL/s may be suspect. For very different
reasons, this latter speculation is probably valid. Measurements by Hu et al. (1994) at rates
higher than 250 mL/s indicate that local mass transfer coefficients increase with increasing
flow rates. In addition, naive subjects were used for the bolus experiments. Chronic exposure
to O3, which is more appropriate to the present simulation scenarios, could alter mass transfer
coefficients due to chemical reactions changing the properties of biochemical constituents. See
Section 8.2.4.2 and Table 8-7 for a comparison of dosimetry modeling results with dosimetry
data, using coefficients based on Hu et al. (1992b).
The experimental data of Hu et al. (1992b) were not sensitive enough to allow an
estimate, with confidence, for the pulmonary region or for the PAR mass transfer coefficient.
For this reason, these coefficients were assumed to be the same as used by Miller et al. (1985).
The value of this parameter is approximately the same as used for the rat. Although the value
of the pulmonary region mass transfer coefficient of rats also is unknown, a comparison by
Pinkerton et al. (1992) of morphometric effects and dosimetry model predictions suggests that
the rat value is not unreasonable.
Ventilation
Activity patterns for the adult and the child (Johnson, 1994) were used to estimate
VEs. The set of estimates had over 300,000 minute-by-minute cases, each case consisting of
the day of the year, hour, minute, and equivalent ventilatory rate (VE divided by body surface
area). The generated patterns were for 214 sequential days, from the beginning of April
through October 1991.
For the adult, VTs were estimated by interpolation from a plot of VT versus VE that
was drawn using information from Table 120 in International Commission on Radiological
8-71
-------
Protection (1975). Tidal volumes and f s for the child were estimated using a relationship
between VT and VE developed by Hofmann et al. (1989). For both humans, f s were defined
as VE divided by the estimated VT.
Estimating Proximal Alveolar Region Dose
The PAR doses were estimated by simulation for each of the minute-by-minute
equivalent VEs and personal concentration estimates; these doses were averaged.
8.5.2.2 Monkeys
Anatomical Aspects
Because there were no data on the uptake of O3 by the URT of monkeys, no
distinction was made between oral and nasal breathing. The URT dimensions were based on
Schreider and Raabe (1981) for the nasopharyngeal region of a 7-kg rhesus monkey. Volumes
were proportional to body weight and dimensions were scaled isotropically. For the
nasopharyngeal region, the dimensions used may have been too large because the casting
procedure may have resulted in a larger volume (Yeh et al., 1989) than in a live animal, and
air flow streamlining (Morgan et al., 1991) may have resulted in a smaller effective volume.
The structure of the LRT was based on Weibel (1963) for humans (the same as used for the
human simulations), which was isotropically scaled to monkey FRC values, assuming that TB
and pulmonary region volumes were in the same proportion as the human FRC values. The
use of the human LRT structure for the monkey simulations is considered to be of less
importance to reasonable predictions than the correct specification of volumes and surface
areas of the different regions.
Mass Transfer Coefficients
The human mass transfer coefficient values were assumed because there were no
reported values nor uptake data that would have allowed estimates of these coefficients. For
corresponding model segments or model generations, the monkey coefficients were the same as
those used for humans.
Ventilation
No ventilation parameters were reported for the monkey experiments. For this
reason, four states of ventilation were used to calculate the doses (see Table 8-11). The first
three in the table were extrapolated from human values of VTs and f s. These human
parameters were consistent with the algorithm used for the human simulations and correspond
approximately to sedentary, low, and light activity as categorized by Hofmann et al. (1989) for
an adult human. The last set of parameters correspond to measurements made by Moss (1995)
on five awake female adult cynomolgus monkeys (Macaco, fascicularis).
Estimating Proximal Alveolar Region Doses
For the two monkey experiments and corresponding ventilatory parameters, scaled
PAR doses were estimated and multiplied by the average exposure concentrations to obtain
estimated average PAR doses.
8.5.2.3 Rats
Anatomical Aspects
8-72
-------
Rats are nasal breathers; therefore, the URT model corresponds to and is based on
8-73
-------
n>
Table 8-11. Estimating Monkey Ventilatory Parameter*! £
Human
Activity
Level"
Sedentary
Low
Light
Extrapolation from a 73 -kg Human
with the Following Parameters0
(L/min)
6
12
20
VT
(mL)
500
900
1,300
f
(breaths/min)
12.0
13.3
15.4
Extrapolation from a 4.4-kg Macaca
fascicularis with the following
parameters'1
1.78
52.9
33.6
To a 1.6-kg Monkey
(L/min)
0.34
0.68
1.14
VT
(mL)
11.0
19.7
28.5
f
(breaths/min)
31.2
34.7
40.0
0.83
19.3
43.2
S
1
To a 6-kg Monkey jg
(L/min)
0.92
1.84
3.07
VT
(mL)
41.1
74.0
107.
0
(breaths/min) ^
n
22.4 J
OQ
24.9 o
28.7 £
CD
1
1
2.25
72.3
31.1 1
oo
See Appendix A for abbreviations and acronyms.. For extrapolation: (VT), = (VTX x (weight /weight); f =£ x (weight/weight)025;
= VT x f. Subscript 1 corresponds to the 1.6- or 6-kg monkey, and subscript 2 corresponds to either the human or the 4.4-kg monkey.
VE
bThese characterizations for the human parameters are consistent with those of the International Commission on Radiological Protection (1975) and
ci- -i, ,,E,'the human VT and fare based on the International Commission on Radiological Protection (1975). For the selected VE, VT and
For a given human v ' & \ > •>
f are the same as used for the adult human.
dMoss (1995); parameters are the average for five monkeys.
n>
f—f
o
o
o
3
n>
n>
n>
o
n>
-------
8-75
-------
the casting procedure can result in a larger volume than in a live animal (Yeh et al., 1989), and
air flow streamlining (Morgan et al., 1991) may result in a smaller effective volume. The
structure of the LRT is a composite of the TB model of the Long-Evans rat (Yeh et al., 1979)
and a pulmonary region based on the Mercer et al. (1991) ventilatory unit model of the S-D
rat. Given the dosimetry model used, the volumes and surface areas of the different regions
are more important than the structural differences of the three strains used to construct the rat
RT.
Mass Transfer Coefficients
Mass transfer coefficients for the nasopharyngeal region and TB of the rat were
estimated using the uptake data of Hatch et al. (1989) and an assumed pulmonary region mass
transfer coefficient. Hatch et al. (1989) reported the average uptake of O3 for eight F344 rats
and the average fraction of 18O3 in the nasopharyngeal region, the trachea, and the lungs.
(Hatch et al. [1989 and 1994] discuss issues related to using 18O3 dose as a measure of O3
dose.) Based on a discussion in Miller et al. (1993) concerning the investigation of Pinkerton
et al. (1992), the pulmonary region mass transfer coefficient was defined as that used by
Mercer et al. (1991). This, combined with information from Hatch et al. (1989), allowed an
estimate of mass transfer coefficients for the nasopharyngeal region and the TB. A common
estimate of mass transfer coefficients also was made using the data for the individual rats. For
these coefficients, whose values were essentially the same as the first set, the individual rat
simulations deviated from the experimental data by an average of D23 %.
Ventilation
For the EPA chronic study, VT and f were measured at Exposure Weeks 1, 3, 13,
52, and 78. For intermediate weeks, these parameters were estimated by interpolation and
assumed constant during each of those weeks. Because ventilatory parameters were not
reported for the National Toxicology Program/Health Effects Institite (NTP/HEI) chronic
study, parameters similar to those of the EPA chronic study were assumed (see below).
Estimating Proximal Alveolar Region Doses
For the EPA chronic study, the exposure pattern was variable during the week, but
repeated each week (Tepper et al., 1991). The PAR doses were simulated for each week and
averaged. The ventilatory, physiological, and anatomical characteristics of the NTP/HEI study
rats with respect to time were assumed to be similar to the EPA chronic study rats, and the
average scaled PAR dose ([g/cm2-min]/[g/m3 ambient O3]) was defined as the same that was
predicted for the EPA study rats. Given this and the average exposure concentration, the
average PAR doses were estimated.
8.5.3 Results and Discussion
8.5.3.1 Simulation Results
The simulation results are presented in Table 8-12. The first and second columns
identify the laboratory experiments and the hypothetical human exposures that were simulated.
Column 4 applies only to the monkeys and is discussed above (see Section 8.5.2.2) in the
discussion of monkey ventilation. The "Average PAR Dose" (column 5) is the predicted
average flux of O3 to the surfaces of the alveoli or respiratory
8-76
-------
Table 8-12. Summary of Simulation Result^
Simulation of
EPA rat (3 weeks)
EPA rat (13 weeks)
EPA rat (78 weeks)
NTP/HEI rat (0. 12 ppm)d
NTP/HEI rat (0.5 ppm)d
NTP/HEI rat (1 ppm)d
Bonnet monkey (0.15 ppm)
Bonnet monkey (0.3 ppm)
Fascicularis monkey
(daily; 0.25 ppm)
Fascicularis monkey
("seasonal"; 0.25 ppm)
9-year-old NYC8 child
Adult NYC1 outdoor supervisor
Source for
Simulation
Chang et al. (1992)
Tepper et al.(1991)
Chang et al. (1995)
Harkema et al.
(1993)
Tyler et al. (1988)
U.S. Environmental
Protection Agency
(1995)
Average
Weight
(kg)
0.24
0.29
0.4
0.4
0.4
0.4
6
6
1.6
1.6
31
73
Minute Volume1'
(L/min)
0.92
1.84
3.07
2.25
0.34
0.68
1.14
0.83
Average
PAR Dose
(Qg/crtr-min)
4.11e-5
3.85e-5
3.43e-5
0.828e-5
3.45e-5
6.89e-5
1/2 below
0.61e-5
3.99e-5
9.92e-5
5.65e-5
2 x below
0.197e-5
1.26e-5
3.25e-5
1.81e-5
3.2e-5
2.78e-5
Cumulative PAR
Dose
(Qg/cm2)
1.24
5.1
27
7.3
30.3
60.5
e 1/2 below"
0.79
5.17
12
7.3
2 x below'
1.55
9.91
25.6
14.3
9.9
8.6
Duration of
Experiment
(weeks)
3
13
78
87
87
87
12.9
12.9
12.9
12.9
12.9
78
78
78
78
78
30
30
Equivalent Human
Child Time"
(weeks)
8.9
39
232
259
259
259
19.5
19.5
19.5
19.5
19.5
' 164
164
164
164
164
30
NA
Equivalent Human
Adult Time"
(weeks)
11
48
286
319
319
319
24
24
24
24
24
203
203
203
203
203
NA
30
oo
"See Appendix A for abbreviations and acronyms.
''Only relevant for the monkeys. See Table 8-11.
"Given the nonhuman exposure time, the number of weeks the human would have to be exposed for equal human and nonhuman physiological times: weeksdUMAN = weeksNONHUMfl
*The relative PAR dose [(g/cnr-min)/(g/m3 O3)] was assumed to be the same as the average for the EPA rat.
Tor each minute volume, the exposure concentration for these bonnet monkeys was one-half those listed below.
'For each minute volume, the fascicularis monkeys exposed "daily" were exposed to O3 twice as long as the "seasonal" fascicularis monkeys.
"NYC = New York City.
-------
bronchioles in the first model generation distal to the terminal bronchioles. For each
experiment, the average is over the total time or duration of the experiment (given in
column 7, "Duration"); the average includes times of no exposure. Column 6, "Cumulative
PAR Dose," is the total quantity of O3 predicted to be absorbed by the tissue and liquids of the
PAR during the experiment; these dose values are the same as the duration (in minutes)
multiplied by the average PAR dose (column 5).
The entries in columns 8 and 9 are the total times a human would have to be
exposed to reach the equivalent physiologic times of a specific laboratory animal species.
Basically, the concept of physiological time is that the time courses of equivalent processes
(e.g., a breath, a heartbeat) across species are approximately the same with respect to the time
span of the process divided by the body mass raised to the one-fourth power (Travis et al.,
1990). If this concept is applicable to the time course of O3-induced biological effects, then the
entries in the two columns are the human exposure durations necessary to obtain the same
cumulative PAR dose biological effect as that of the corresponding laboratory animal duration.
Note that for only two or three of the laboratory animal experiments are the human durations
greater than the laboratory animals' equivalent or real (duration) times.
8.5.3.2 Interpretation of Chronic Site-Specific Dose-Effect Estimates
Because the PAR is considered the primary site of O3 injury and represents that
region of the lung from which most chronic lung diseases originate, it was selected as the most
appropriate target to develop cross-species dose-response extrapolations. The selected "effect"
relates to the thickness of the interstitium at the PAR, which is indicative of fundamental
structural remodeling. The indices of thickness provided in the noted studies include both
cellular and acellular constituents, a distinction that was not always clear; however, interstitial
arithmetic thickness, areal volume (cubic micrograms per square microgram), or volume
density was available and could be represented in terms of percent change from control.
Dose-response curves for O3-induced thickening of the PAR are represented in
Figures 8-18A (rats) and 8-18B (monkeys). The cumulative PAR dose for the specific
exposure scenario (from Table 8-12; Cumulative PAR Dose) for each study is provided on the
abscissa as is its corresponding percent change in total and acellular interstitial (PAR) thickness
on the ordinate (detailed in Table 8-13). The rat studies are plotted separately from the
monkey studies because of an approximate three- to fivefold difference in their responses, with
the monkey being more responsive than the rat. Although this difference could represent
innate sensitivity differences between the species, it should be noted that estimates of daytime
rat exposures (a period of quiescence for the rat), in contrast to the daytime-active monkeys,
may have been substantially lower in terms of dose than that predicated on the basis of rat
ventilatory measurements that were derived when they were in an aroused, awakened state.
Recent studies comparing rat and human 18O3 dosimetry indicate that exercise can account for
up to a fivefold difference in acute responsiveness between species (Hatch et al., 1994). These
explanations remain speculative, however, in the absence of direct data. Ventilation values for
the monkeys were selected a priori as light activity to be similar to the human model being
used, which had ventilation values only slightly higher than the empirically derived values (see
Table 8-11) from comparable awake, resting monkeys (when adjusted for size).
8-78
-------
300
260
220
180
140
100
60
20
-20
(A) RAT
(B) MONKEY
Adult
V
Child
n Total Interstitial (EPA)
• Total Interstitial (NTP/HEI)
0 Acellular (EPA)
• Acellular (NTP/HEI)
10 20 30 40 50 60
Cumulative PAR Dose (|jg/cm2)
70
o
I
o
D>
300
260
220
180
140
100
60
20
-20
D Total Interstitial (Harkema)
• Total Interstitial (Tyler)
0 Acellular (Harkema)
• Acellular (Tyler)
10 20 30 40 50 60
Cumulative PAR Dose (|jg/cm2)
70
Figure 8-18. (A) Change from control of total interstitial and acellular thickness for the
rats exposed to ozone (OJ in the U.S. Environmental Protection Agency
(Chang et al., 1992) and National Toxicology Program/Health Effects
Institite (Chang et al., 1995) studies. Solid line represents the linear
regression for the total interstitial thickness, and the dashed line represents
the linear regression for the acellular thickness across the various
cumulative dose estimates for both studies. The respective correlation
coefficients (r2) are 0.84 and 0.80. (B) Change from control of total
interstitial and acellular thickness for the monkeys exposed to O3 in the
Tyler et al. (1988) and Harkema et al. (1993) studies. Solid line represents
the linear regression for the total interstitial thickness, and the dashed line
represents the linear regression for the acellular thickness across the
various cumulative dose estimates for both studies. The respective
correlation coefficients (r2) are 0.97 and 0.93. Vertical dashed line with
arrows for child and adult denote proximal alveolar region (PAR) dose level
for interpolation of PAR thickness effect from the monkey upper estimate
of response.
What is remarkable in both Figures 8-ISA and 8-18B is the apparent linear dose-
response relationships within species. This may be due, in part, to the fact that there was
consistency in the methods of the respective investigatory team studying each species. It is
interesting to note that the estimated total cumulative PAR dose for the EPA chronic study (27
/xg O3/cm2) is quite similar to that estimated for the NTP/HEI study group of 0.5 ppm
O3 (30.3 /xg O3/cm2). An examination of the respective PAR thickness response from each
study indicates an apparent similarity between the studies, with the NTP/HEI results being a
bit larger in value. At 0.5 ppm O3, the NTP/HEI investigators also reported the observation of
bronchiolarization into the PAR, whereas this process was absent or not significant at 0.12
ppm O3 or in the EPA chronic study. It is not clear whether this indicates qualitatively, rather
8-79
-------
than quantitatively, different responses initiated at 0.12 and 0.50 ppm O3. However, it is
important to note that the correlation coefficients (r2) for PAR dose to response across
8-80
-------
Table 8-13. Summaries of Study Data Used in Extrapolation of Chronic Ozone Effecfe
Study Species
F344 Male Rat; Chang et al., 1992 (EPA)
—78 weeks (urban profile)
D Profile: 0.06-ppm continuous
background, 22 h/day, 7 days/week;
0.25-ppm ramped peak over 9 h (avg.
0. 19 ppm for peak; 0.09 ppm daily),
5 days/week
D 3-week exposure (from Figure 4)
D 13-week exposure (from Table 4)
D 78-week exposure (from Table 4)
D 16- week recovery after 78 weeks air
D 16- week recovery after 78 weeks O3
F344 Male Rat; Chang et al., 1995
(NTP/HEI)-87 weeks (6h/day,
5 days/week)
D 0 ppm
D 0. 12 ppm
D 0.50 ppm
D 1.00 ppm
Fascicularis Monkey; Tyler et al., 1988
—18 mo (8 h/day)
D 0 ppm
D 0.25 ppm (daily)
D 0.25 ppm (daily; alternating months)
Bonnet Monkey; Harkema et al., 1993 —
90 days (8 h/day)
D 0 ppm
D 0. 15 ppm
D 0.30 ppm
Total Interstitial Volume
(mean value — /im3//tm2)
(+16% estimate from figure)
0.398b
0.462 (+16%)c
0.473b
0.619 (+31%)
0.531b
0.580 (+9%)
From Appendix Dl— HEI
Report
0.520
0.497 (D5%)
0.826 (+58%)
0.902 (+73%)
Volume fraction of lung
occupied by respiratory
bronchiolar wall (x 10"3) — from
Table II
0.716
2.550 (+256%)
1.830 (+156%)
Arithmetic mean thickness
(/tni) — from Figure 3 and text
6.5
9.2 (+41%)
13.1 (+102%)
Cellular Volume
(/tm3//im2)
(+12%)
0.153b
0.174 (+14%)
0. 149"
0.179 ( + 20%)
0.171b
0.168(0%)
0.161
0.162(0%)
0.239 (+48%)
0.233 (+45%)
0.110
0.437 ( + 297%)
0.297 (+170%)
3.2
4.2 ( + 30%)
5.6 ( + 74%)
Acellular Volume
(/im3//tni2)
(+15%)
0.245b
0.288 (+18%)
0.325b
0.440 (+35%)
0.360b
0.410 (+14%)
0.360
0.335 (D7%)
0.587 (+63%)
0.669 (+85%)
0.606
2. 113 ( + 248%)
1.533 (+152%)
3.3
5.0 ( + 51%)
7. 5 (+128%)
"See Appendix A for abbreviations and acronyms.
bControl value.
'Parenthetical values indicate percent difference from respective control.
8-81
-------
both studies as illustrated in Figure 8-18A are 0.80 and 0.84 for the total and acelluar
interstitial changes, respectively. In the case of the monkey data represented in Figure 8-17B,
the analogous correlation coefficients (r2) were 0.97 and 0.93 for the total and acellular
interstitial effects, respectively. These highly significant correlations suggest that interstitial
injury is indeed a cumulative process throughout the exposure history. Moreover,in the case
of monkeys with seasonal (intermittent monthly) exposures, there was no loss of effect (i.e.,
reversal) during the air periods. The response was strictly a function of cumulative dose.
The availability of exposure/activity data for a 9-year-old child and adult outdoor
worker over an "exposure-season" of 214 days in New York City provides the opportunity to
estimate an analogous PAR dose in these human individuals for comparison to the animal dose-
response relationships. Because the rat exposure scenario extends over the majority of a
lifetime, an attempt was made to determine the equivalent human exposure time using
standard, accepted algorithms for lifetime transformations (see Table 8-12). However, the
extrapolation of human biologic exposure time from that of the rat did not yield reasonable
results. The primate-based extrapolation of exposure-time fared slightly better but also was
less than adequate for the purposes of this exercise. Estimates of accumulated dose, however,
were in general agreement (see Table 8-12) and served as the basis for the cross-species
extrapolation of effects described below.
Unfortunately, there are no hard data to substantiate whether the rat or monkey better
represents the human in the context of the endpoint being addressed herein. The monkey could
be favored on the basis of responsiveness at low levels of O3, corresponding to those in human
spirometric tests following exposures involving exercise; however, the dose-adjusted BAL
protein data are comparable for humans and rodents. The apparent differences in sensitivity
between the rat and monkey may, on the other hand, reflect more rapid repair in the rat than in
the primate. This concept would be consistent with the algorithms of Travis et al. (1990),
which are based on intrinsic metabolic rates and thus may reasonably apply to repair processes
after injury or damage from oxidant-lung surface interactions. Thus, the responses for the rat
and monkey may best be considered as bracketing the human response. The linearity of the
dose-response relationships in both the rat and monkey models lends credibility to interpolation
(Figures 8-18A and 8-18B) of the estimated dose to the human PAR to its corresponding
response. For example, in the case of the child, the predicted seasonal response could range
from about a 20 to 75% increase in PAR thickness; the adult human response would be slightly
less with a 15 to 70% increase. Although the actual changes in the human may not be as large
as the monkey, the graphical data from both species suggest that recovery may not be complete
(in fact, there was no reversal) when exposure is interrupted during alternating months.
Likewise, in rats exposed long-term, reversal was incomplete (D66%; see Table 8-13) after a
4-mo postexposure period in clean air. In the case of the rat, 4 mo represents a larger
proportion of its life span than a similar period in either the monkey or human, thereby
suggesting that even 6 to 8 mo of background "off season" levels of ozone would not be
sufficient for complete recovery in primates.
This attempt to extrapolate results of animal studies for the chronic effects of O3 to
the human obviously must be considered preliminary because of lingering questions regarding
relative dosimetry across species and the uncertainties associated with episodic (typical human
scenario) versus continual, repeated exposures (typical of animal studies). Yet, this
extrapolation provides a foundation from which additional questions can be derived and
addressed to reduce these uncertainties. The coherent evidence in hand suggests that there is a
8-82
-------
real possibility that chronic exposure to O3 can lead to interstitial thickening at the PAR, that
region of the lung involved in chronic diseases such as chronic bronchiolitis associated with
cigarette smoke or occupational fibrogenesis. The apparent lack of reversal of effects during
periods of clean air raises concern that seasonal exposures have a cumulative impact over many
years. The role of adaptive processes in this response is unknown but may be critically
dependent on the temporal frequency or profile of exposure. Furthermore, the inter species
diversity in apparent sensitivity to the chronic effects of O3 is notable, but the issue of
dosimetry may be explanatory, in part. It would appear that the rat probably represents no less
than the lower limit of response and the monkey the upper limit, if not a direct 1:1 correlate as
could be speculated on the basis of relative equivalent lifetime estimates and their phylogenetic
relationship.
8.6 Summary and Conclusions
8.6.1 Ozone Dosimetry
There have been significant advances in O3 dosimetry since publication of the
previous O3 criteria document (U.S. Environmental Protection Agency, 1986) that better
enable quantitative extrapolation with marked reductions in uncertainty. Prior to 1986, there
were limited data on O3 uptake in laboratory animals (Yokoyama and Frank, 1972; Miller
et al., 1979), essentially no reliable data in humans (Clamann and Bancroft, 1959; Hallett,
1965), only one realistic model of O3 dose (Miller et al., 1978, 1985), and no data on
O3 reaction kinetics in lung lining fluids. At the present time, data gaps in all of these areas
have begun to fill in. Experiments and models describing the uptake efficiency and delivered
dose of O3 in the RT of animals and humans are beginning to present a clearer picture than
previously has existed.
The total RT uptake efficiency of rats at rest is approximately 0.50 (Wiester et al.,
1987, 1988; Hatch et al., 1989). Data from excised rat lungs support these in vivo findings,
and further indicate that O3 uptake efficiency is chemical reaction dependent (Postlethwait
et al., 1994). Of the O3 taken up by the total RT of the rat, 0.50 is removed in the head, 0.07
in the larynx/trachea, and 0.43 in the lungs (Hatch et al., 1989). The regional uptake
efficiency data from the rat have been useful in estimating O3 mass transfer coefficients for the
rat.
Ozone dosimetry models require input of regional mass transfer coefficients.
Limited studies have been conducted to quantitate the mass transfer coefficients of lung tissue
directly using excised animal tissue. In pig and sheep tracheae, mass transfer coefficients were
determined for unidirectional flow conditions and were found to be independent of flow,
suggesting a lack of dependence of O3 uptake on gas-diffusion processes (Ben-Jebria et al.,
1991). These findings contrast with Aharonson et al. (1974), who found that the mass transfer
coefficient in dog NP increased as a function of increasing flow.
In humans at rest, the total RT uptake efficiency is between 0.80 and 0.95 (Gerrity
et al., 1988; Hu et al., 1992b; Wiester et al., 1996). At VTs around 500 mL, total RT uptake
efficiency falls from about 0.9 to 0.75 as flow increases from 250 to 1,000 mL/s.
As VT increases, uptake efficiency increases and flow dependence lessens, suggesting that, at
high VT, uptake may be gas diffusion limited. At a VT around 1,500 mL, total RT uptake falls
from 0.96 at a flow of 250 mL/s to 0.92 at a flow of 1,000 mL/s. The studies of Gerrity et al.
(1988) and Wiester et al. (1996) indicate that the mode of breathing (oral versus nasal versus
8-83
-------
oronasal) has little effect on URT or on total RT uptake efficiency. This observation is
supported by experiments comparing pulmonary function response as a function of mode of
breathing (Adams et al., 1989; Hynes et al., 1988). Kabel et al. (1994), however, found that
URT uptake efficiency was lower with mouthpiece breathing as compared with nasal
breathing. One possible explanation of the discrepancy among the studies is that a mouthpiece
may decrease URT uptake efficiency in comparison with unencumbered breathing. The
enhanced physiologic response to O3 with mouthpiece breathing, shown by Adams et al.
(1989), supports this concept.
To obtain data on regional O3 uptake efficiency in humans, Gerrity et al. (1995)
measured O3 concentrations at various anatomical sites (from the mouth to bronchus
intermedius) in spontaneously breathing humans. They found that the unidirectional uptake
efficiency of the human trachea was similar to that of the sheep and pig trachea (Ben-Jebria et
al., 1991), suggesting a similar mass transfer coefficient behavior in the human trachea.
Gerrity et al. (1995) also found that the uptake efficiencies between the mouth and various
anatomical sites in the total RT agreed well with the O3 bolus data of Hu et al. (1992). Both
the Hu et al. (1992b) and Gerrity et al. (1995) data indicate that the mass transfer coefficients
of the large conducting airways are larger than had been thought previously.
When all of the animal and human in vivo O3 uptake efficiency data are compared,
there is a good degree of consistency across data sets. This agreement raises the level of
confidence with which these data sets can be used to support dosimetric model formulations.
In the area of mathematical model formulation, there have been several models
developed since 1986. They can be grouped according to how transport and chemical
reactions are modeled: instantaneous reactions or quasi-steady, first-order reactions. The
models (Overton et al., 1987; Miller et al., 1988; Overton et al., 1989; Hanna et al., 1989;
Grotberg et al., 1990) predict that net O3 dose to lung lining fluid plus tissue gradually
decreases distally from the trachea toward the end of the TB, and then rapidly decreases in the
pulmonary region. When the theoretical dose of O3 to lung tissue is computed, it is low in the
trachea, increases to a maximum in the terminal bronchioles of the first generation of the
pulmonary region, and then decreases rapidly distally into the pulmonary region. The models
also provide insight into the role that increased ventilation plays in enhancing O3-induced
responses. The increased VT and flow, associated with exercise in humans or CO2-stimulated
ventilation increases in rats, shifts O3 dose further into the periphery of the lung, causing a
disproportionate increase in distal lung dose. This prediction is supported by the data of
Postlethwait et al. (1994) in excised rat lungs and of Hu et al. (1992b) and Gerrity et al. (1995)
in human lungs.
Ozone dosimetry models also have enabled examination of regional dosimetry
among parallel and serial anatomical structures. When asymmetric lung morphology is used in
dosimetric models, the variation of O3 dose among anatomically equivalent ventilatory units as
a function of path length from the trachea has been predicted to vary as much as sixfold
(Overton et al., 1989; Mercer et al., 1991; Mercer and Crapo, 1993); units with the shorter
paths are expected to have the greater damage. This could have significant implications for
regional or localized damage to lung tissue. Whereas the average lung dose might be at a level
that would be considered insignificant, local regions of the lung may receive significantly
higher than average doses and therefore be at greater risk for chronic effects.
Theoretical models also have been applied to make predictions about delivered
doses from exposure scenarios that are not necessarily achievable experimentally. Overton and
8-84
-------
Graham (1989) and Miller and Overton (1989) have scaled the human lung dimensions to
account for age variations. They predicted that LRT uptake efficiency is not sensitive to age at
resting ventilation, but is age dependent when exercise conditions are invoked. The total
quantity of O3 absorbed per minute is predicted to increase with age during heavy work or
exercise.
8.6.2 Species Homology and Sensitivity
Examining functional parameters measured analogously in humans and various
animal species discloses remarkable similarity in acute O3-induced effects. The tachypneic
response to this oxidant is clearly concentration-dependent in both humans and animals and
shows parallel exacerbation when hyperventilation (e.g., exercise or CO2) is superimposed.
Indeed, rodents appear to be slightly more responsive than humans in this regard. What is not
known is whether this is evidence of pulmonary irritant sensitivity, perhaps as a prelude to
toxicity, or whether tachypnea is a defensive posture taken by the respiratory system to
minimize distal lung O3 deposition. Airway or lung resistance in humans is not affected
appreciably by acute exposure to O3, except under conditions of heavy exercise; animals
appear to need high-level exposures or special preparations that bypass nasal scrubbing.
Dynamic lung compliance, on the other hand, tends to decrease across species. However, the
evidence in both animals and humans is not as strong as one might expect, given the distal lung
deposition of this poorly soluble oxidant.
Ozone-induced spirometric changes, the hallmark of response in humans, also occur
in exposed rats, although the relative responsiveness of these alterations in the rodent appears
to be about half that of the human. It is unclear, however, the degree to which anesthesia (rat)
and the comparability of hyperventilation induced by CO2 (rat) or exercise (human) may
influence this difference in responsiveness. Collectively, the acute functional response of
laboratory animals to O3 appears quite homologous to that of the human. Likewise, the studies
of BAL constituents indicate that the influx of inflammatory cells and protein from the serum is
influenced by species, but perhaps to a less extent than by ventilation and antioxidant status,
because adjustment for these factors can modulate responses to approximate animal responses
to those of humans. Unfortunately, these influential factors are rarely measured and, even less
often, controlled.
When humans are exposed repeatedly for several consecutive days, lung function
decrements subside, and normal spirometric parameters are regained. This phenomenon of
functional attenuation also has been demonstrated in rats, not only in terms of spirometry, but
also in terms of the classic tachypneic ventilatory response. Full or partial attenuation of the
BAL parameters also appears to occur in both rats and humans, but exposure scenario appears
to play a role; other cellular changes in animals do not attenuate. Existing epidemiologic
studies provide only suggestive evidence that persistent or progressive deterioration in lung
function is associated with long-term oxidant-pollutant exposure. These long-term effects are
thought to be expressed in the form of maximum airflow or spirometric abnormalities, but the
foundation for this conclusion remains weak and hypothetical. Animal study data, although
suggesting that O3 has effects on lung function at near-ambient levels, present a variable
picture of response that may or may not relate to technical conditions of exposure or some
other, yet undiscovered variable of response. Thus, a cogent interpretation of the animal
findings as definitive evidence of chronic deterioration of lung function would be difficult at
this time. However, the subtle functional defects apparent after 12 to 18 mo of exposure and
8-85
-------
the detailed morphometric assessments of the O3-induced lesions do appear consistent with the
modicum of studies focusing on long-term effects in human populations. Based on the
apparent homology of these responses between humans and laboratory animals, animal studies
provide a means to more directly assess such chronic health concerns.
8.6.3 Quantitative Extrapolation
The agreement between theoretical models of O3 uptake and experimental
determinations of O3 uptake efficiency now provide a basis on which responses may be
examined as a function of delivered O3 dose instead of O3 exposure concentration.
By examining responses as a function of delivered dose, the goal of quantitative extrapolation
between species can be approached.
The use of delivered dose to investigate responses has been examined in two
contexts: (1) intraspecies comparisons and (2) inter species comparisons. With respect to
intraspecies comparisons, Miller et al. (1988) assumed that the relevant dose mediating the
human pulmonary function response was the pulmonary tissue dose. They then utilized the
breathing patterns, exposure concentrations, and pulmonary function responses from the
human studies of McDonnell et al. (1983) to predict the dose-response. They found that there
was general agreement between the shapes of the concentration-response curves and the dose-
response curves and that differences could be accounted for by the translation between
exposure concentration and O3 dose. In another example dealing with intraspecies
comparisons, Miller and Conolly (1995) compared the distribution of predicted O3 tissue dose
to a ventilatory unit in a rat as a function of distance from the BADJ, with the distribution of
alveolar wall thickening as a function of the same distance measure. Miller and Conolly
(1995) found remarkable consistency between the predicted dose distribution and the response
distribution (i.e., as predicted delivered dose declined, response declined).
In an attempt to make an inter species comparison of dose and response, existing or
modified models of Miller et al. (1985), Miller et al. (1988), and Overton et al. (1987) were
used to predict doses among species for two different types of responses. In the first case, the
tachypneic response to O3 as a function of dose was analyzed. The maximum ratio of O3-
altered f to control f was plotted as a function of the average centriacinar dose over the period
from the beginning of exposure to the point of maximum f ratio. Rat and human data were
used for this comparison. It was found that, at comparable O3 doses, the responses of rats
greatly exceeded that of humans and were initiated at lower doses. By examining the dose
response instead of the concentration response, the difference in tachypneic response between
rats and humans is magnified. In another example, an analysis similar to Miller et al. (1988)
was performed to examine recovered BAL protein as a function of O3 dose to the pulmonary
region. The species considered were the rat, guinea pig, rabbit, and human. In all cases, the
BAL protein response followed a log-linear relationship, suggesting a consistency of response
across species. Yet the data from different species tended to cluster together, suggesting
species-specific sensitivity factors.
An attempt was made to address quantitatively the question of potential chronic
alteration of the lungs of people exposed to O3 by integrating dosimetry model predictions and
biological effects observed in laboratory animals. In the long-term exposure studies selected
for analysis, importance was placed on the relevance of exposure concentrations, the site of
specificity of the morphometric analysis, and the consistency of analysis within species. Two
rat (F344) studies were selected that represented near-lifetime exposures to O3 over a range of
8-86
-------
concentrations and scenarios (Chang et al., 1992, 1995). Two monkey studies were also
considered: (1) bonnet monkeys exposed for 90 days (Harkema et al., 1993) and (2)
fascicularis monkeys exposed daily for 18 mo or daily every other month for 18 mo (Tyler
et al., 1988). The biological effect chosen for extrapolation was the increased thickness of the
acellular and total interstitial volumes in the PAR region of the lung; measurements were made
in all four of these investigations. The quantity of O3 predicted to be absorbed per square
centimeter of PAR surface area was chosen as the dose.
Generally, the information needed to carry out the dosimetry predictions was not
provided by the studies. This required assumptions such as the scaling of ventilation
parameters, volumes, and surface areas from one species or strain to another. The assumption
that had the greatest impact on the modeling results dealt with the pulmonary region mass
transfer coefficient. The value used for this parameter has very little experimental justification
and was chosen to be approximately the same for all species (i.e., the values calculated by
Miller et al. [1985] for humans and by Overton et al. [1987] for rats). If the value of this
coefficient is in fact approximately the same for all the species, the extrapolation of effects is
not expected to be affected by the value itself. For the human simulations, a 9-year-old child
and an adult outdoor supervisor living in New York City were considered. The activity and
exposure patterns for these hypothetical people were generated by an exposure model
(Johnson, 1994) for April through October 1991. The laboratory animal dose-response curves
showed an apparent linear relationship within species with relatively high correlation
coefficients, from 0.80 to 0.98 depending on species and effect. Assuming the relationships
depicted in Figure 8-18, the predicted dose for the hypothetical humans indicated a seasonal
response for the child of a 20 to 75 % increase in PAR tissue thickness and, for the adult, a 15
to 70% increase, depending on the laboratory species used for the prediction (the higher range
corresponds to the monkey). For the monkeys, there seemed to be little reversal with
postexposure to air, which was consistent with the cumulative dose hypothesis. Although the
reader should note the number of assumptions that underlie these predictions, this exercise,
nevertheless, suggests that long-term O3 exposure could impart a chronic effect in humans.
8-87
-------
References
Adams, W. C.; Schelegle, E. S.; Shaffrath, J. D. (1989) Oral and oronasal breathing during continuous exercise
produce similar responses to ozone inhalation. Arch. Environ. Health 44: 311-316.
Aharonson, E. P.; Menkes, H.; Gurtner, G.; Swift, D. L.; Proctor, D. F. (1974) Effect of respiratory airflow
rate on removal of soluble vapors by the nose. J. Appl. Physiol. 37: 654-657.
Aissa, M.; Hatch, G. E. (1988) Method for tracing oxygen-18 in vivo: application to ozone dosinietry in animals.
In: Simic, M. G.; Taylor, K. A.; Ward, J. P.; Von Sonntag, C., eds. Oxygen radicals in biology and
medicine: proceedings of the 4th international congress on radicals; June-July 1987; La Jolla, CA.
New York, NY: Plenum Press; pp. 195-197. (Basic life sciences: v. 49).
Amdur, M. O.; Ugro, V.; Underbill, D. W. (1978) Respiratory response of guinea pigs to ozone alone and with
sulfur dioxide. Am. Ind. Hyg. Assoc. J. 39: 958-961.
Arakawa, K.; Ichinose, T.; Sagai, M. (1986) [Species differences of lipid peroxidation and the related factors in
lungs of various animals exposed to combined gas of nitrogen dioxide and ozone]. In: Murakami, M.;
Kawata, M.; Kobayashi, T., eds. [Experimental studies on the effects of gaseous air pollutants in
combination on animals]. Kokuritsu Kogai Kenkyusho Kenkyu Hokoku 101: 135-175.
Bartlett, D., Jr.; Faulkner, C. S., II; Cook, K. (1974) Effect of chronic ozone exposure on lung elasticity in young
rats. J. Appl. Physiol. 37: 92-96.
Bassett, D. J. P.; Bowen-Kelly, E.; Brewster, E. L.; Elbon, C. L.; Reichenbaugh, S. S.; Bunion, T.; Kerr, J. S.
(1988) A reversible model of acute lung injury based on ozone exposure. Lung 166: 355-369.
Bates, D. V.; Hazucha, M. (1973) The short-term effects of ozone on the human lung. In: Proceedings of the
conference on health effects of air pollutants; October; Washington, DC. Washington, DC: U.S.
Senate, Committee on Public Works; pp. 507-540; serial no. 93-15.
Bates, D. V.; Bell, G. M.; Burnhani, C. D.; Hazucha, M.; Mantha, J.; Pengelly, L. D.; Silverman, F. (1972)
Short-term effects of ozone on the lung. J. Appl. Physiol. 32: 176-181.
Beckett, W. S.; McDonnell, W. P.; Horstman, D. H.; House, D. E. (1985) Role of the parasympathetic nervous
system in acute lung response to ozone. J. Appl. Physiol. 59: 1879-1885.
Bedi, J. P.; Horvath, S. M.; Folinsbee, L. J. (1982) Human exposure to sulfur dioxide and ozone in a high
temperature-humidity environment. Am. Ind. Hyg. Assoc. J. 43: 26-30.
Bellville, J. W.; Escarraga, L. A.; Wallenstein, S. L.; Houde, R. W.; Howland, W. S. (1960) Relative
respiratory depressant effects of oxymorphone (Numorphan) and morphine. Anesthesiology
21: 397-400.
Ben-Jebria, A.; Ultnian, J. S. (1989) Fast-responding chemiluminescent ozone analyzer for respiratory
applications. Rev. Sci. Instrum. 60: 3004-3011.
Ben-Jebria, A.; Hu, S.-C.; Ultman, J. S. (1990) Improvements in a chemiluminescent ozone analyzer for
respiratory applications. Rev. Sci. Instrum. 61: 3435-3439.
Ben-Jebria, A.; Hu, S.-C.; Kitzmiller, E. L.; Ultman, J. S. (1991) Ozone absorption into excised porcine and
sheep tracheae by a bolus-response method. Environ. Res. 56: 144-157.
Bennett, G. (1962) Ozone contamination of high altitude aircraft cabins. Aerosp. Med. 33: 969-973.
-------
Boorman, G. A.; Schwartz, L. W.; Dungworth, D. L. (1980) Pulmonary effects of prolonged ozone insult in rats:
morphometric evaluation of the central acinus. Lab. Invest. 43: 108-115.
Challen, P. J. R.; Hickish, D. E.; Bedford, J. (1958) An investigation of some health hazards in an inert-gas
tungsten-arc welding shop. Br. J. Ind. Med. 15: 276-282.
Chang, H.-K.; Farhi, L. E. (1973) On mathematical analysis of gas transport in the lung. Respir. Physiol.
18: 370-385.
Chang, L.-Y.; Huang, Y.; Stockstill, B. L.; Graham, J. A.; Grose, E. C.; Menache, M. G.; Miller, F. J.; Costa,
D. L.; Crapo, J. D. (1992) Epithelial injury and interstitial fibrosis in the proximal alveolar regions
of rats chronically exposed to a simulated pattern of urban ambient ozone. Toxicol. Appl. Pharniacol.
115: 241-252.
Chang, L.-Y.; Stockstill, B. L.; Menache, M. G.; Mercer, R. R.; Crapo, J. D. (1995) Consequences of
prolonged inhalation of ozone on F344/N rats: collaborative studies. Part VIII. Morphometric
analysis of structural alterations in alveolar regions. Cambridge, MA: Health Effects Institute;
pp. 3-39; research report no. 65.
Chatham, M. D.; Eppler, J. H., Jr.; Sauder, L. R.; Green, D.; Kulle, T. J. (1987) Evaluation of the effects of
vitamin C on ozone-induced bronchoconstriction in normal subjects. Ann. N. Y. Acad. Sci.
498: 269-279.
Clamann, H. G.; Bancroft, R. W. (1959) Toxicity of ozone in high altitude flight. In: Ozone chemistry and
technology: proceedings of the international ozone conference; November 1956; Chicago, IL.
Washington, DC: American Chemical Society; pp. 352-359. (Advances in chemistry: no. 21).
Costa, D. L. (1985) Interpretation of new techniques used in the determination of pulmonary function in rodents.
Fundam. Appl. Toxicol. 5: 423-434.
Costa, D. L. (1994) Ventilatory parameters for the EPA chronic rat study [memorandum to John Overton].
Research Triangle Park, NC: U.S. Environmental Protection Agency, Health Effects Research
Laboratory; December 13.
Costa, D. L.; Kutzman, R. S.; Lehmann, J. R.; Popenoe, E. A.; Drew, R. T. (1983) A subchronic multidose
ozone study in rats. In: Lee, S. D.; Mustafa, M. G.; Mehlman, M. A., eds. International symposium
on the biomedical effects of ozone and related photochemical oxidants; March 1982; Pinehurst, NC.
Princeton, NJ: Princeton Scientific Publishers, Inc.; pp. 369-393. (Advances in modern
environmental toxicology: v. 5).
Costa, D. L.; Schafrank, S. N.; Wehner, R. W.; Jellett, E. (1985) Alveolar permeability to protein in rats
differentially susceptible to ozone. J. Appl. Toxicol. 5: 182-186.
Costa, D. L.; Stevens, M. S.; Tepper, J. S. (1988a) Repeated exposure to ozone (Oj) and chronic lung disease:
recent animal data. Presented at: 81st annual meeting of the Air Pollution Control Association; June;
Dallas, TX. Pittsburgh, PA: Air Pollution Control Association; paper no. 88-122.3.
Costa, D. L.; Hatch, G. E.; Highfill, J.; Stevens, M. A.; Tepper, J. S. (1988b) Pulmonary function studies in the
rat addressing concentration versus time relationships of ozone (O3). Research Triangle Park, NC:
U.S. Environmental Protection Agency, Health Effects Research Laboratory; EPA report no.
EPA-600/D-88-256. Available from: NTIS, Springfield, VA; PB89-129050.
Costa, D. L.; Tepper, J. S.; Lehmann, J. R.; Winsett, D. W.; Hatch, G. E.; Devlin, R. B. (1995) Strain as a
factor in the magnitude and time—course of response to acute ozone (O3): a study of lung dysfunction
and injury in Wistar, Fischer, and Sprague-Dawley rats. Research Triangle Park, NC: U.S.
8-89
-------
Environmental Protection Agency, National Health & Environmental Effects Research Laboratory,
Pulmonary Toxicology Branch.
Costa, D. L.; Tepper, J. S.; Stevens, M. A.; Watkinson, W. P.; Doerfler, D. L.; Gelzleichter, T. R.; Last, J. A.
(1995) Restrictive lung disease in rats exposed chronically to an urban profile of ozone. Am. J.
Respir. Crit. Care Med. 151: 1512-1518.
Crissman, K.; Hatch, G. E.; Slade, R.; Norwood, J.; Koren, H. (1995) Ozone-induced alterations in vitamin C
concentrations in bronchoalveolar lavage fluid of humans, rats, and guinea pigs. Research Triangle
Park, NC: U.S. Environmental Protection Agency, National Health & Environmental Effects
Research Laboratory.
DeLucia, A. J.; Adams, W. C. (1977) Effects of O3 inhalation during exercise on pulmonary function and blood
biochemistry. J. Appl. Physiol.: Respir. Environ. Exercise Physiol. 43: 75-81.
Betels, R.; Sayre, J. W.; Coulson, A. H.; Rokaw, S. N.; Massey, F. J., Jr.; Tashkin, D. P.; Wu, M.-M. (1981)
The UCLA population studies of chronic obstructive respiratory disease: IV. respiratory effect of
long-term exposure to photochemical oxidants, nitrogen dioxide, and sulfates on current and never
smokers. Am. Rev. Respir. Dis. 124: 673-680.
Betels, R.; Tashkin, D. P.; Sayre, J. W.; Rokaw, S. N.; Coulson, A. H.; Massey, F. J., Jr.; Wegman, D. H.
(1987) The UCLA population studies of chronic obstructive respiratory disease: 9. lung function
changes associated with chronic exposure to photochemical oxidants; a cohort study among
never-smokers. Chest 92: 594-603.
Betels, R.; Tashkin, D. P.; Sayre, J. W.; Rokaw, S. N.; Massey, F. J., Jr.; Coulson, A. H.; Wegman, D. H.
(1991) The UCLA population studies of CORD: X. a cohort study of changes in respiratory function
associated with chronic exposure to SOX, NOX, and hydrocarbons. Am. J. Public Health 81: 350-359.
Devlin, R. B.; McDonnell, W. P.; Mann, R.; Becker, S.; House, D. E.; Schreinemachers, D.; Koren, H. S.
(1991) Exposure of humans to ambient levels of ozone for 6.6 hours causes cellular and biochemical
changes in the lung. Am. J. Respir. Cell Mol. Biol. 4: 72-81.
Devlin, R. B.; Folinsbee, L. J.; Biscardi, P.; Becker, S.; Madden, M. C.; Robbins, M.; Koren, H. S. (1995)
Repeated exposure of humans exposed to ozone. II. Inflammation and cell damage. Research
Triangle Park, NC: U.S. Environmental Protection Agency, National Health and Environmental
Effects Research Laboratory, Clinical Research Branch; manuscript in preparation.
Elsayed, N. M.; Kass R.; Mustafa, M. G.; Hacker, A. D.; Ospital, J. J.; Chow, C. K.; Cross, C. E. (1988)
Effect of dietary vitamin E level on the biochemical response of rat lung to ozone inhalation. Drag
Nutr. Interact. 5: 373-386.
Eustis, S. L.; Schwartz, L. W.; Kosch, P. C.; Dungworth, D. L. (1981) Chronic bronchiolitis innonhuman
primates after prolonged ozone exposure. Am. J. Pathol. 105: 121-137.
Folinsbee, L. J.; Raven, P. B. (1984) Exercise and air pollution. J. Sports Sci. 2: 57-75.
Folinsbee, L. J.; Silverman, P.; Shephard, R. J. (1975) Exercise responses following ozone exposure. J. Appl.
Physiol. 38: 996-1001.
Folinsbee, L. J.; Horvath, S. M.; Raven, P. B.; Bedi, J. P.; Morton, A. R.; Drinkwater, B. L.; Bolduan, N. W.;
Gliner, J. A. (1977) Influence of exercise and heat stress on pulmonary function during ozone
exposure. J. Appl. Physiol.: Respir. Environ. Exercise Physiol. 43: 409-413.
8-90
-------
Folinsbee, L. J.; Bedi, J. P.; Horvath, S. M. (1980) Respiratory responses in humans repeatedly exposed to low
concentrations of ozone. Am. Rev. Respir. Dis. 121: 431-439.
Folinsbee, L. J.; McDonnell, W. P.; Horstnian, D. H. (1988) Pulmonary function and symptom responses after
6.6-hour exposure to 0.12 ppm ozone with moderate exercise. JAPCA 38: 28-35.
Fredberg, J. J.; Wohl, M. B.; Glass, G. M.; Dorkin, H. L. (1980) Airway area by acoustic reflections measured
at the mouth. J. Appl. Physiol.: Respir. Environ. Exercise Physiol. 48: 749-758.
Gerrity, T. R. (1989) Nasopharyngeal uptake of ozone in humans and animals. In: Crapo, J. D.; Miller, F. J.;
Smolko, E. D.; Graham, J. A.; Hayes, A. W., eds. Extrapolation of dosimetric relationships for
inhaled particles and gases. San Diego, CA: Academic Press, Inc.; pp. 187-195.
Gerrity, T. R.; McDonnell, W. F. (1989) Do functional changes in humans correlate with the airway removal
efficiency of ozone? In: Schneider, T.; Lee, S. D.; Wolters, G. J. R.; Grant, L. D., eds.
Atmospheric ozone research and its policy implications: proceedings of the 3rd US-Dutch
international symposium; May 1988; Nijmegen, The Netherlands. Amsterdam, The Netherlands:
Elsevier Science Publishers; pp. 293-300. (Studies in environmental science 35).
Gerrity, T. R.; Wiester, M. J. (1987) Experimental measurements of the uptake of ozone in rats and human
subjects. Presented at: 80th annual meeting of the Air Pollution Control Association; June; New
York, NY. Pittsburgh, PA: Air Pollution Control Association; paper no. 87-99.3.
Gerrity, T. R.; Weaver, R. A.; Berntsen, J.; House, D. E.; O'Neil, J. J. (1988) Extrathoracic and intrathoracic
removal of O3 in tidal-breathing humans. J. Appl. Physiol. 65: 393-400.
Gerrity, T. R.; McDonnell, W. P.; O'Neil, J. J. (1989) Experimental ozone dosimetry in humans. J. Aerosol
Med. 2: 129-139.
Gerrity, T. R.; Bennett, W. D.; Kehrl, H.; DeWitt, P. J. (1993) Mucociliary clearance of inhaled particles
measured at 2 h after ozone exposure in humans. J. Appl. Physiol. 74: 2984-2989.
Gerrity, T. R.; McDonnell, W. P.; House, D. E. (1994) The relationship between delivered ozone dose and
functional responses in humans. Toxicol. Appl. Pharmacol. 124: 275-283.
Gerrity, T. R.; Biscardi, P.; Strong, A.; Garlington, A. R.; Brown, J. S.; Bromberg, P. A. (1995) Bronchoscopic
determination of ozone uptake in humans. J. Appl. Physiol. 79: 852-860.
Gibbons, S. I.; Adams, W. C. (1984) Combined effects of ozone exposure and ambient heat on exercising
females. J. Appl. Physiol.: Respir. Environ. Exercise Physiol. 57: 450-456.
Gong, H., Jr.; Bradley, P. W.; Simmons, M. S.; Tashkin, D. P. (1986) Impaired exercise performance and
pulmonary function in elite cyclists during low-level ozone exposure in a hot environment. Am. Rev.
Respir. Dis. 134: 726-733.
Gordon, C. J. (1991) Toxic-induced hypothermia and hypometabolism: do they increase uncertainty in the
extrapolation of lexicological data from experimental animals to humans? Neurosci. Biobehav. Rev.
15: 95-98.
Gordon, L. M.; Jensen, F. C.; Curtain, C. C.; Mobley, P. W.; Aloia, R. C. (1988) Thermotropic lipidphase
separation in the human immunodeficiency virus. Biochim. Biophys. Acta 943: 331-342.
Graham, J. A.; Hatch, G. E. (1984) The integration of species sensitivity and dosimetry in the extrapolation of
ozone and nitrogen dioxide health data from animal to man. Presented at: 77th annual meeting of the
8-91
-------
Air Pollution Control Association; June; San Francisco, CA. Pittsburgh, PA: Air Pollution Control
Association; report no. 84-31.4.
Grose, E. C.; Stevens, M. A.; Hatch, G. E.; Jaskot, R. H.; Selgrade, M. J. K.; Stead, A. G.; Costa, D. L.;
Graham, J. A. (1988) The impact of a 12-month exposure to a diurnal pattern of ozone on pulmonary
function, antioxidant biochemistry and immunology. Research Triangle Park, NC: U.S.
Environmental Protection Agency, Health Effects Research Laboratory; EPA report no.
EPA-600/D-88-071. Available from: NTIS, Springfield, VA; PB88-213160.
Grotberg, J. B. (1990) Gas absorption in pulmonary airways at low Peclet number. J. Biomech. Eng.
112: 177-182.
Grotberg, J. B.; Sheth, B. V.; Mockros, L. F. (1990) An analysis of pollutant gas transport and absorption in
pulmonary airways. J. Biomech. Eng. 112: 168-176.
Guyton, A. C. (1947) Measurement of the respiratory volumes of laboratory animals. Am. J. Physiol. 150: 70-77.
Hackney, J. D.; Linn, W. S.; Buckley, R. D.; Pedersen, E. E.; Karuza, S. K.; Law, D. C.; Fischer, D. A.
(1975) Experimental studies on human health effects of air pollutants: I. design considerations. Arch.
Environ. Health 30: 373-378.
Hackney, J. D.; Linn, W. S.; Mohler, J. G.; Collier, C. R. (1977) Adaptation to short-term respiratory effects of
ozone in men exposed repeatedly. J. Appl. Physiol.: Respir. Environ. Exercise Physiol. 43: 82-85.
Hallett, W. Y. (1965) Effect of ozone and cigarette smoke on lung function. Arch. Environ. Health 10: 295-302.
Hanna, L. M.; Scherer, P. W. (1986) Measurement of local mass transfer coefficients in a cast model of the
human upper respiratory tract. J. Biomech. Eng. 108: 12-18.
Hanna, L. M.; Frank, R.; Scherer, P. W. (1989) Absorption of soluble gases and vapors in the respiratory
system. In: Chang, H. K.; Paiva, M., eds. Respiratory physiology: an analytical approach. New
York, NY: Marcel Dekker, Inc.; pp. 277-316. (Lenfant, C., ed. Lung biology in health and disease:
v. 40).
Hansen, J. E.; Ampaya, E. P. (1975) Human air space shapes, sizes, areas, and volumes. J. Appl. Physiol.
38: 990-995.
Harkema, J. R.; Mauderly, J. L.; Hahn, F. F. (1982) The effects of emphysema on oxygen toxicity in rats.
Am. Rev. Respir. Dis. 126: 1058-1065.
Harkema, J. R.; Plopper, C. G.; Hyde, D. M.; St. George, J. A.; Wilson, D. W.; Dungworth, D. L. (1993)
Response of macaque bronchiolar epithelium to ambient concentrations of ozone. Am. J. Pathol.
143: 857-866.
Hart, M. C.; Orzalesi, M. M.; Cook, C. D. (1963) Relation between anatomic respiratory dead space and body
size and lung volume. J. Appl. Physiol. 18: 519-522.
Hatch, G. E.; Aissa, M. (1987) Determination of absorbed dose of ozone in animals and humans using stable
isotope (oxygen-18) tracing. Presented at: 80th annual meeting of the Air Pollution Control
Association; June; New York, NY. Pittsburgh, PA: Air Pollution Control Association; paper no.
87-99.2.
Hatch, G. E.; Slade, R.; Stead, A. G.; Graham, J. A. (1986) Species comparison of acute inhalation toxicity of
ozone and phosgene. J. Toxicol. Environ. Health 19: 43-53.
8-92
-------
Hatch, G. E.; Wiester, M. J.; Overtoil, J. H., Jr.; Aissa, M. (1989) Respiratory tract dosimetry of [18]O-labeled
ozone in rats: implications for a rat-human extrapolation of ozone dose. In: Schneider, T.; Lee, S.
D.; Wolters, G. J. R.; Grant, L. D., eds. Atmospheric ozone research and its policy implications:
proceedings of the 3rd US-Dutch international symposium; May 1988; Nijmegen, The Netherlands.
Amsterdam, The Netherlands: Elsevier Science Publishers; pp. 553-560. (Studies in environmental
science 35).
Hatch, G. E.; Slade, R.; Harris, L. P.; McDonnell, W. P.; Devlin, R. B.; Koren, H. S.; Costa, D. L.;
McKee, J. (1994) Ozone dose and effect in humans and rats: a comparison using oxygen-18 labeling
and bronchoalveolar lavage. Am. J. Respir. Crit. Care Med. 150: 676-683.
Hazucha, M. J. (1987) Relationship between ozone exposure and pulmonary function changes. J. Appl. Physiol.
62: 1671-1680.
Hazucha, M. J.; Bates, D. V.; Bromberg, P. A. (1989) Mechanism of action of ozone on the human lung.
J. Appl. Physiol. 67: 1535-1541.
Hofmann, W.; Martonen, T. B.; Graham, R. C. (1989) Predicted deposition of nonhygroscopic aerosols in the
human lung as a function of subject age. J. Aerosol Med. 2: 49-68.
Horsfield, K.; Dart, G.; Olson, D. E.; Filley, G. P.; Cumming, G. (1971) Models of the human bronchial tree. J.
Appl. Physiol. 31: 207-217.
Horstman, D. H.; Folinsbee, L. J.; Ives, P. J.; Abdul-Salaam, S.; McDonnell, W. F. (1990) Ozone concentration
and pulmonary response relationships for 6.6-hour exposures with five hours of moderate exercise to
0.08, 0.10, and 0.12 ppm. Am. Rev. Respir. Dis. 142: 1158-1163.
Hotchkiss, J. A.; Harkema, J. R.; Sun, J. D.; Henderson, R. F. (1989) Comparison of acute ozone-induced nasal
and pulmonary inflammatory responses in rats. Toxicol. Appl. Pharmacol. 98: 289-302.
Hu, S. (1991) Noninvasive determination of ozone distribution in the human lung airways [Ph.D. thesis].
Hu, S. C.; Ben-Jebria, A.; Ultman, J. S. (1992a) Simulation of ozone uptake distribution in the human airways by
orthogonal collocation on finite elements. Comput. Biomed. Res. 25: 264-278.
Hu, S. C.; Ben-Jebria, A.; Ultman, J. S. (1992b) Longitudinal distribution of ozone absorption in the lung: quiet
respiration in healthy subjects. J. Appl. Physiol. 73: 1655-1667.
Hu, S.-C.; Ben-Jebria, A.; Ultman, J. S. (1994) Longitudinal distribution of ozone absorption in the lung: effects
of respiratory flow. J. Appl. Physiol. 77: 574-583.
Hunter, A. R.; Pleuvry, B. J.; Rees, J. M. H. (1968) The respiratory depressant effects of barbiturates and
narcotic analgesics in the unanaesthetized rabbit. Br. J. Anaesth. 40: 927-935.
Hyde, D. M.; Hubbard, W. C.; Wong, V.; Wu, R.; Pinkerton, K.; Plopper, C. G. (1992) Ozone-induced acute
tracheobronchial epithelial injury: relationship to granulocyte emigration in the lung. Am. J. Respir.
Cell Mol. Biol. 6: 481-497.
Hynes, B.; Silverman, F.; Cole, P.; Corey, P. (1988) Effects of ozone exposure: a comparison between oral and
nasal breathing. Arch. Environ. Health 43: 357-360.
International Commission on Radiological Protection. (1975) Report of the Task Group on reference man. Oxford,
United Kingdom: Pergamon Press Ltd.; ICRP publication no. 23.
8-93
-------
Jaeger, R. J.; Gearhart, J. M. (1982) Respiratory and metabolic response of rats and mice to formalin vapor.
Toxicology 25: 299-309.
Johnson, T. R. (1994) One-minute ozone exposure sequences for selected New York cohorts [letter to
Dr. John H. Overton, U.S. EPA]. Durham, NC: International Technology Corporation;
November 22.
Kabel, J. R.; Ben-Jebria, B.; Ultman, J. S. (1994) Longitudinal distribution of ozone absorption in the lung:
comparison of nasal and oral quiet breathing. J. Appl. Physiol. 77: 2584-2592.
Kleeberger, S. R.; Bassett, D. J. P.; Jakab, G. J.; Levitt, R. C. (1990) A genetic model for evaluation of
susceptibility to ozone-induced inflammation. Am. J. Physiol. 258: L313-L320.
Kleinfeld, M.; Giel, C.; Tabershaw, I. R. (1957) Health hazards associated with inert-gas-shielded metal arc
welding. AMA Arch. Ind. Health 15: 27-31.
Kliment, V. (1973) Similarity and dimensional analysis, evaluation of aerosol deposition in the lungs of laboratory
animals and man. Folia Morphol. (Prague) 21: 59-64.
Kodavanti, U. P.; Hatch, G. E.; Starcher, B.; Giri, S. N.; Winsett, D.; Costa, D. L. (1995) Ozone-induced
pulmonary functional, pathological, and biochemical changes in normal and vitamin C-deficient
guinea pigs. Fundam. Appl. Toxicol. 24: 154-164.
Koren, H. S.; Devlin, R. B.; Graham, D. E.; Mann, R.; McGee, M. P.; Horstman, D. H.; Kozumbo, W. J.;
Becker, S.; House, D. E.; McDonnell, W. P.; Bromberg, P. A. (1989a) Ozone-induced
inflammation in the lower airways of human subjects. Am. Rev. Respir. Dis. 139: 407-415.
Koren, H. S.; Devlin, R. B.; Graham, D. E.; Mann, R.; McDonnell, W. F. (1989b) The inflammatory response
inhuman lung exposed to ambient levels of ozone. In: Schneider, T.; Lee, S. D.; Wolters, G. J. R.;
Grant, L. D., eds. Atmospheric ozone research and its policy implications: proceedings of the 3rd
US-Dutch international symposium; May 1988; Nijmegen, The Netherlands. Amsterdam, The
Netherlands: Elsevier Science Publishers; pp. 745-753. (Studies in environmental science 35).
Koren, H. S.; Devlin, R. B.; Becker, S.; Perez, R.; McDonnell, W. F. (1991) Time-dependent changes of
markers associated with inflammation in the lungs of humans exposed to ambient levels of ozone.
Toxicol. Pathol. 19: 406-411.
Koren, H. S.; Devlin, R. B.; Becker, S. (1994) Ozone-induced inflammatory response in pulmonary cells.
In: Shook, L. B.; Laskin, D. L., eds. Xenobiotics and inflammation: roles of cytokines and growth
factors. Orlando, PL: Academic Press; pp. 249-281.
Kosch, P. C.; Gillespie, J. R.; Berry, J. D. (1979) Flow-volume curves and total pulmonary resistance in normal
bonnet and rhesus monkeys. J. Appl. Physiol.: Respir. Environ. Exercise Physiol. 46: 176-183.
Lai, Y.-L.; Tsuya, Y.; Hildebrandt, J. (1978) Ventilatory responses to CC^ exposure in the rat. J. Appl. Physiol.:
Respir. Environ. Exercise Physiol. 45: 611-618.
Lamm, W. J. E.; Hildebrandt, J. R.; Hildebrandt, J.; Lai, Y.-L. (1982) End-expiratory volume in the rat: effects
of consciousness and body position. J. Appl. Physiol.: Respir. Environ. Exercise Physiol. 53:
1071-1079.
Last, J. A.; Gelzleichter, T. R.; Harkema, J.; Hawk, S. (1994) Consequences of prolonged inhalation of ozone on
Fischer-344/N rats: collaborative studies. Part I: Content and cross-linking of lung collagen.
Cambridge, MA: Health Effects Institute; research report no. 65.
8-94
-------
Lefohn, A. S.; Edwards, P. J.; Adams, M. B. (1994) The characterization of ozone exposures in rural West
Virginia and Virginia. J. Air Waste Manage. Assoc. 44: 1276-1283.
Leith, D. E. (1976) Comparative mammalian respiratory mechanics. Physiologist 19: 485-510.
Martin-Body, R. L.; Sinclair, J. D. (1985) Analysis of respiratory patterns in the awake and in the halothane
anaesthetised rat. Respir. Physiol. 61: 105-113.
Mauderly, J. L. (1984) Respiratory function responses of animals and man to oxidant gases and to pulmonary
emphysema. J. Toxicol. Environ. Health 13: 345-361.
Mauderly, J. L. (1986) Respiration of F344 rats in nose-only inhalation exposure tubes. J. Appl. Toxicol.
6: 25-30.
Mautz, W. J.; Bufalino, C. (1989) Breathing pattern and metabolic rate responses of rats exposed to ozone.
Respir. Physiol. 76: 69-77.
Mautz, W. J.; Bufalino, C.; Kleinman, M. T.; Lejnieks, R. M.; Phalen, R. F. (1985) Pulmonary function of
exercising dogs exposed to ozone alone or in combination with SO2 and acid aerosol. Presented at:
78th annual meeting of the Air Pollution Control Association; June; Detroit, MI. Pittsburgh, PA: Air
Pollution Control Association; paper no. 85-29.4.
McDonnell, W. P.; Horstman, D. H.; Hazucha, M. J.; Seal, E., Jr.; Haak, E. D.; Salaam, S. A.; House, D. E.
(1983) Pulmonary effects of ozone exposure during exercise: dose-response characteristics. J. Appl.
Physiol.: Respir. Environ. Exercise Physiol. 54: 1345-1352.
McDonnell, W. P.; Horstman, D. H.; Abdul-Salaam, S.; Raggio, L. J.; Green, J. A. (1987) The respiratory
responses of subjects with allergic rhinitis to ozone exposure and their relationship to nonspecific
airway reactivity. Toxicol. Ind. Health 3: 507-517.
Mercer, R. R.; Crapo, J. D. (1993) Three-dimensional analysis of lung structure and its application to pulmonary
dosimetry models. In: Gardner, D. E.; Crapo, J. D.; McClellan, R. O., eds. Toxicology of the lung.
2nd ed. New York, NY: Raven Press; pp. 155-186. (Target organ toxicology series).
Mercer, R. R.; Laco, J. M.; Crapo, J. D. (1987) Three-dimensional reconstruction of alveoli in the rat lung for
pressure-volume relationships. J. Appl. Physiol. 62: 1480-1487.
Mercer, R. R.; Anjilvel, S.; Miller, F. J.; Crapo, J. D. (1991) Inhomogeneity of ventilatory unit volume and its
effects on reactive gas uptake. J. Appl. Physiol. 70: 2193-2205.
Mercer, R. R.; Russell, M. L.; Roggli, V. L.; Crapo, J. D. (1994a) Cell number and distribution in human and
rat airways. Am. J. Respir. Cell Mol. Biol. 10: 613-624.
Mercer, R. R.; Russell, M. L.; Crapo, J. D. (1994b) Alveolar septal structure in different species. J. Appl.
Physiol. 77: 1060-1066.
Miller, F. J. (1977) A mathematical model of transport and removal of ozone in mammalian lungs [Ph.D. thesis].
Raleigh, NC: North Carolina State University. Available from: University Microfilms International,
Ann Arbor, MI; publication no. 77-21560.
Miller, F. J.; Conolly, R. B. (1995) Uncertainties in health risk assessments: commentary on selected issues and
research needs. In: Lee, S. D.; Schneider, T., eds. Comparative risk analysis and priority setting for
air pollution issues: proceedings of the 4th U.S.-Dutch international symposium; June 1993;
Keystone, CO. Pittsburgh, PA: Air & Waste Management Association; pp. 76-91; VIP-43.
8-95
-------
Miller, F. J.; Overtoil, J. H. (1989) Critical issues in intra- and interspecies dosimetry of ozone.
In: Schneider, T.; Lee, S. D.; Wolters, G. J. R.; Grant, L. D., eds. Atmospheric ozone research
and its policy implications: proceedings of the 3rd U.S.-Dutch international symposium; May 1988;
Nijmegen, The Netherlands. Amsterdam, The Netherlands: Elsevier Science Publishers:
pp. 281-291. (Studies in environmental science 35).
Miller, F. J.; Menzel, D. B.; Coffin, D. L. (1978) Similarity between man and laboratory animals in regional
pulmonary deposition of ozone. Environ. Res. 17: 84-101.
Miller, F. J.; McNeal, C. A.; Kirtz, J. M.; Gardner, D. E.; Coffin, D. L.; Menzel, D. B. (1979)
Nasopharyngeal removal of ozone in rabbits and guinea pigs. Toxicology 14: 273-281.
Miller, F. J.; Overton, J. H., Jr.; Jaskot, R. H.; Menzel, D. B. (1985) A model of the regional uptake of gaseous
pollutants in the lung: I. the sensitivity of the uptake of ozone in the human lung to lower respiratory
tract secretions and exercise. Toxicol. Appl. Pharmacol. 79: 11-27.
Miller, F. J.; Overton, J. H.; Gerrity, T. R.; Graham, R. C. (1988) Interspecies dosimetry of reactive gases.
In: Mohr, U.; Dungworth, D.; McClellan, R.; Kinimerle, G.; Stober, W.; Lewkowski, J., eds.
Inhalation toxicology: the design and interpretation of inhalation studies and their use in risk
assessment. New York, NY: Springer-Verlag; pp. 139-155.
Miller, F. J.; Overton, J. H.; Kimbell, J. S.; Russell, M. L. (1993) Regional respiratory tract absorption of
inhaled reactive gases. In: Gardner, D. E.; Crapo, J. D.; McClellan, R. O., eds. Toxicology of the
lung. 2nd ed. New York, NY: Raven Press; pp. 485-525. (Target organ toxicology series).
Morgan, K. T.; Kimbell, J. S.; Monticello, T. M.; Patra, A. L.; Fleishman, A. (1991) Studies of inspiratory
airflow patterns in the nasal passages of the F344 rat and rhesus monkey using nasal molds: relevance
to formaldehyde toxicity. Toxicol. Appl. Pharmacol. 110: 223-240.
Moss, O. R. (1995) [Letter to Dr. John H. Overton, U.S. EPA, re monkey tidal volumes and breathing
frequencies]. Research Triangle Park, NC: Chemical Industry Institute of Toxicology; February 20.
Murphy, S. D.; Ulrich, C. E.; Frankowitz, S. H.; Xintaras, C. (1964) Altered function in animals inhaling low
concentrations of ozone and nitrogen dioxide. Am. Ind. Hyg. Assoc. J. 25: 246-253.
Nagasaka, T.; Hirata, K.; Sugano, Y.; Shibata, H. (1979) Heart balance during physical restraint in rats. Jpn.
J. Physiol. 29: 383-392.
Nagasaka, T.; Hirata; Shibata, H.; et al. (1980) Metabolic and cardiovascular changes during physical restraint in
rats. Jpn. J. Physiol. 30: 799-803.
Overton, J. H., Jr. (1984) Physicochemical processes and the formulation of dosimetry models. In: Miller, F. J.;
Menzel, D. B., eds. Fundamentals of extrapolation modeling of inhaled toxicants: ozone and nitrogen
dioxide. Washington, DC: Hemisphere Publishing Corporation; pp. 93-114.
Overton, J. H.; Graham, R. C. (1989) Predictions of ozone absorption in human lungs from newborn to adult.
Health Phys. 57(suppl. 1): 29-36.
Overton, J. H.; Miller, F. J. (1988) Dosimetry modeling of inhaled toxic reactive gases. In: Watson, A. Y.;
Bates, R. R.; Kennedy, D., eds. Air pollution, the automobile, and public health. Washington, DC:
National Academy Press; pp. 367-385.
Overton, J. H.; Graham, R. C.; Miller, F. J. (1987) A model of the regional uptake of gaseous pollutants in the
lung: II. the sensitivity of ozone uptake in laboratory animal lungs to anatomical and ventilatory
parameters. Toxicol. Appl. Pharmacol. 88: 418-432.
8-96
-------
Overtoil, J. H.; Barnett, A. E.; Graham, R. C. (1989) Significances of the variability of tracheobronchial airway
paths and their air flow rates to dosimetry model predictions of the absorption of gases. In: Crapo, J.
D.; Smolko, E. D.; Miller, F. J.; Graham, J. A.; Hayes, A. W., eds. Extrapolation of dosimetric
relationships for inhaled particles and gases. San Diego, CA: Academic Press, Inc.; pp. 273-291.
Pack, A.; Hooper, M. B.; Nixon, W.; Taylor, J. C. (1977) A computational model of pulmonary gas transport
incorporating effective diffusion. Respir. Physiol. 29: 101-123.
Paiva, M. (1973) Gas transport in the human lung. J. Appl. Physiol. 35: 401-410.
Patra, A. L.; Menache, M. G.; Shaka, N. B.; Gooya, A. (1987) A niorphometric study of nasal-pharyngeal
growth for particle deposition in the rat. Am. Ind. Hyg. Assoc. J. 48: 556-562.
Pavlin, E. G.; Hornbein, T. F. (1986) Anesthesia and the control of ventilation. In: Cherniack, N. S.;
Widdicombe, J. G., eds. Handbook of physiology; section 3, respiration: v. II, control of breathing.
Bethesda, MD: American Physiological Society; pp. 793-813. (Handbook of physiology: v. II).
Phalen, R. P.; Oldham, M. J.; Beaucage, C. B.; Crocker, T. T.; Mortensen, J. D. (1985) Postnatal enlargement
of human tracheobronchial airways and implications for particle deposition. Anal. Rec. 212:
368-380.
Pinkerton, K. E.; Mercer, R. R.; Plopper, C. G.; Crapo, J. D. (1992) Distribution of injury and microdosimetry
of ozone in the ventilatory unit of the rat. J. Appl. Physiol. 73: 817-824.
Pino, M. V.; Hyde, D. M.; Stovall, M. Y. (1991) Strain differences in the response of the rat lung to an acute
ozone exposure. Am. Rev. Respir. Dis. 143: A698.
Postlethwait, E. M.; Langford, S. D.; Bidani, A. (1994) Determinants of inhaled ozone in isolated rat lungs.
Toxicol. Appl. Pharmacol. 125: 77-89.
Pryor, W. A. (1992) How far does ozone penetrate into the pulmonary air/tissue boundary before it reacts? Free
Radic. Biol. Med. 12: 83-88.
Pryor, W. A.; Das, B.; Church, D. F. (1991) The ozonation of unsaturated fatty acids: aldehydes and hydrogen
peroxide as products and possible mediators of ozone toxicity. Chem. Res. Toxicol. 4: 341-348.
Raabe, O. G.; Yeh, H. C.; Schum, G. M.; Phalen, R. F. (1976) Tracheobronchial geometry: human, dog, rat,
hamster. Albuquerque, NM: Lovelace Foundation; report no. LF-53.
Raub, J. A.; Mercer, R. R.; Miller, F. J.; Graham, J. A.; O'Neil, J. J. (1982) Dose response of elastase- induced
emphysema in hamsters. Am. Rev. Respir. Dis. 125: 432-435.
Raub, J. A.; Miller, F. J.; Graham, J. A. (1983) Effects of low-level ozone exposure on pulmonary function in
adult and neonatal rats. In: Lee, S. D.; Mustafa, M. G.; Mehlman, M. A., eds. International
symposium on the biomedical effects of ozone and related photochemical oxidants; March 1982;
Pinehurst, NC. Princeton, NJ: Princeton Scientific Publishers, Inc.; pp. 363-367. (Advances in
modern environmental toxicology: v. 5).
Rehder, K.; Mallow, J. E.; Fibuch, E. E.; Krabill, D. R.; Sessler, A. D. (1974) Effects of isoflurane anesthesia
and muscle paralysis on respiratory mechanics in normal man. Anesthesiology 41: 477-485.
Rehder, K.; Sessler, A. D.; Marsh, H. M. (1975) General anesthesia and the lung. Am. Rev. Respir. Dis.
112: 541-563.
8-97
-------
Rich, C. R.; Rehder, K.; Knopp, T. K.; Hyatt, R. E. (1979) Halothane and enflurane anesthesia and respiratory
mechanics in prone dogs. J. Appl. Physiol.: Respir. Environ. Exercise Physiol. 46: 646-653.
Santrock, J.; Hatch, G. E.; Slade, R.; Hayes, J. M. (1989) Incorporation and disappearance of oxygen-18 in lung
tissue from mice allowed to breathe 1 ppm 18O3'. Toxicol. Appl. Pharmacol. 98: 75-80.
Schelegle, E. S.; Siefkin, A. D.; McDonald, R. J. (1991) Time course of ozone-induced neutrophilia in normal
humans. Am. Rev. Respir. Dis. 143: 1353-1358.
Scherer, P. W.; Shendalman, L. H.; Greene, N. M. (1972) Simultaneous diffusion and convection in single breath
lung washout. Bull. Math. Biophys. 34: 393-412.
Schreider, J. P.; Hutchens, J. (1980) Morphology of the guinea pig respiratory tract. Anal. Rec. 196: 313-321.
Schreider, J. P.; Raabe, O. G. (1981) Anatomy of the nasal-pharyngeal airway of experimental animals. Anal.
Rec. 200: 195-205.
Schwartz, J. (1989) Lung function and chronic exposure to air pollution: a cross-sectional analysis of NHANES II.
Environ. Res. 50: 309-321.
Schwartz, L. W.; Dungworth, D. L.; Mustafa, M. G.; Tarkington, B. K.; Tyler, W. S. (1976) Pulmonary
responses of rats to ambient levels of ozone: effects of 7-day intermittent or continuous exposure.
Lab. Invest. 34: 565-578.
Seltzer, J.; Bigby, B. G.; Stulbarg, M.; Holtzman, M. J.; Nadel, J. A.; Ueki, I. P.; Leikauf, G. D.; Goetzl,
E. J.; Boushey, H. A. (1986) O3-induced change in bronchial reactivity to methacholine and airway
inflammation in humans. J. Appl. Physiol. 60: 1321-1326.
Siafakas, N. M.; Bonora, M.; Duron, B.; Gautier, H.; Milic-Emili, J. (1983) Dose effect of pentobarbital sodium
on control of breathing in cats. J. Appl. Physiol.: Respir. Environ. Exercise Physiol. 55: 1582-1592.
Skornick, W. A.; Brain, J. D. (1990) Breathing and lung mechanics in hamsters: effect of pentobarbital
anesthesia. J. Appl. Physiol. 68: 2536-2541.
Slade, R.; Stead, A. G.; Graham, J. A.; Hatch, G. E. (1985) Comparison of lung antioxidant levels in humans
and laboratory animals. Am. Rev. Respir. Dis. 131: 742-746.
Slade, R.; Highfill, J. W.; Hatch, G. E. (1989) Effects of depletion of ascorbic acid or nonprotein sulfhydryls on
the acute inhalation toxicity of nitrogen dioxide, ozone, and phosgene. Inhalation Toxicol.
1: 261-271.
Slade, R.; Crissman, K.; Norwood, J.; Hatch, G. (1993) Comparison of antioxidant substances in bronchoalveolar
lavage cells and fluid from humans, guinea pigs, and rats. Exp. Lung Res. 19: 469-484.
Snider, G. L.; Sherter, C. B. (1977) A one-year study of the evolution of elastase-induced emphysema in
hamsters. J. Appl. Physiol.: Respir. Environ. Exercise Physiol. 43: 721-729.
Stahl, W. R. (1967) Scaling of respiratory variables in mammals. J. Appl. Physiol. 22: 453-460.
Stokinger, H. E. (1965) Ozone toxicology: a review of research and industrial experience, 1954-1964. Arch.
Environ. Health 10: 719-731.
Takezawa, J.; Miller, F. J.; O'Neil, J. J. (1980) Single-breath diffusing capacity and lung volumes in small
laboratory mammals. J. Appl. Physiol.: Respir. Environ. Exercise Physiol. 48: 1052-1059.
8-98
-------
Tepper, J. S.; Wiester, M. J.; King, M. E.; Weber, M. P.; Costa, D. L. (1988) Use of carbon dioxide challenge
to detect toxicant-induced changes in cardiopulmonary function of awake rats. Inhalation Toxicology,
Premier Issue; pp. 79-95.
Tepper, J. S.; Costa, D. L.; Lehmann, J. R.; Weber, M. P.; Hatch, G. E. (1989) Unattenuated structural and
biochemical alterations in the rat lung during functional adaptation to ozone. Am. Rev. Respir. Dis.
140: 493-501.
Tepper, J. S.; Wiester, M. J.; Weber, M. P.; Menache, M. G. (1990) Measurements of cardiopulmonary
response in awake rats during acute exposure to near-ambient concentrations of ozone. Fundam.
Appl. Toxicol. 10: 7-15.
Tepper, J. S.; Wiester, M. J.; Weber, M. P.; Fitzgerald, S.; Costa, D. L. (1991) Chronic exposure to a simulated
urban profile of ozone alters ventilatory responses to carbon dioxide challenge in rats. Fundam. Appl.
Toxicol. 17: 52-60.
Tepper, J. S.; et al. (1995) Time course of inflammation in the rat exposed to ozone for short and prolonged
periods with intermittent hyperventilation. Research Triangle Park, NC: U.S. Environmental
Protection Agency, National Health and Environmental Effects Research Laboratory, Pulmonary
Toxicology Branch.
Travis, C. C.; White, R. K.; Ward, R. C. (1990) Interspecies extrapolation of pharmacokinetics. J. Theor. Biol.
142: 285-304.
Tyler, W. S.; Tyler, N. K.; Last, J. A.; Gillespie, M. J.; Barstow, T. J. (1988) Comparison of daily and seasonal
exposures of young monkeys to ozone. Toxicology 50: 131-144.
U.S. Environmental Protection Agency. (1986) Air quality criteria for ozone and other photochemical oxidants.
Research Triangle Park, NC: Office of Health and Environmental Assessment, Environmental
Criteria and Assessment Office; EPA report nos. EPA-600/8-84-020aF-eF. 5v. Available from:
NTIS, Springfield, VA; PB87-142949.
U.S. Environmental Protection Agency. (1995) Review of national ambient air quality standards for ozone
assessment of scientific and technical information. Research Triangle Park, NC: Office of Air
Quality Planning and Standards; OAQPS staff paper.
Uchiyama, I.; Simomura, Y.; Yokoyama, E. (1986) Effects of acute exposure to ozone on heart rate and blood
pressure of the conscious rat. Environ. Res. 41: 529-537.
Ultnian, J. S. (1988) Transport and uptake of inhaled gases. In: Watson, A. Y.; Bates, R. R.; Kennedy, D., eds.
Air pollution, the automobile, and public health. Washington, DC: National Academy Press;
pp. 323-366.
Ultnian, J. S.; Anjilvel, S. (1990) Monte Carlo simulation of ozone uptake in an assynietric lung model.
In: Schneck, D. J.; Lucas, C. L., eds. Biofluid mechanics 3: proceedings of the third mid-Atlantic
conference on biofluid mechanics; October; Blacksburg, VA. New York, NY: New York University
Press; pp. 45-52.
Ultman, J. S.; Ben-Jebria, A.; Hu, S. (1994) Noninvasive determination of respiratory ozone absorption: the
bolus-response method. Cambridge, MA: Health Effects Institute; research report no. 69.
VanBree, L.; Rombout, P. J. A.; Rietjens, I. M. C. M.; Dormans, J. A. M. A.; Marra, M. (1989)
Pathobiochemical effects in rat lung related to episodic ozone exposure. In: Schneider, T.; Lee,
S. D.; Wolters, G. J. R.; Grant, L. D., eds. Atmospheric ozone research and its policy implications:
proceedings of the 3rd US-Dutch international symposium; May 1988; Nijmegen, The Netherlands.
8-99
-------
Amsterdam, The Netherlands: Elsevier Science Publishers; pp. 723-732. (Studies in environmental
science 35).
Van Bree, L.; Marra, M.; Rombout, P. J. A. (1992) Differences in pulmonary biochemical and inflammatory
responses of rats and guinea pigs resulting from daytime or nighttime, single and repeated exposure
to ozone. Toxicol. Appl. Pharmacol. 116: 209-216.
Vaughan, T. R., Jr.; Jennelle, L. P.; Lewis, T. R. (1969) Long-term exposure to low levels of air pollutants:
effects on pulmonary function in the beagle. Arch. Environ. Health 19: 45-50.
Watanabe, S.; Frank, R.; Yokoyama, E. (1973) Acute effects of ozone on lungs of cats: I. functional. Am. Rev.
Respir. Dis. 108: 1141-1151.
Watkinson, W. P.; Wiester, M. J.; Highfill, J. W. (1995) Ozone toxicity in the rat: I. effect of changes in ambient
temperature on extrapulmonary physiological parameters. J. Appl. Physiol. 78: 1108-1120.
Weibel, E. R. (1963) Morphonietry of the human lung. New York, NY: Academic Press Inc.
Weissberg, R. M.; Marian, J.; Bradshaw, J. (1976) Respiratory and cardiovascular effects of prostaglandins in the
conscious guinea pig. J. Pharmacol. Exp. Ther. 198: 197-208.
Wiester, M. J.; Williams, T. B.; King, M. E.; Menache, M. G.; Miller, F. J. (1987) Ozone uptake in awake
Sprague-Dawley rats. Toxicol. Appl. Pharmacol. 89: 429-437.
Wiester, M. J.; Tepper, J. S.; King, M. E.; Menache, M. G.; Costa, D. L. (1988) Comparative study of ozone
(O3) uptake in three strains of rats and in the guinea pig. Toxicol. Appl. Pharmacol. 96: 140-146.
Wiester, M. J.; Stevens, M. A.; Menache, M. G.; McKee, J. L., Jr.; Gerrity, T. R. (1996) Ozone uptake in
healthy adult males during quiet breathing. Fundam. Appl. Toxicol. 29: 102-109.
Wong, K. L.; Alarie, Y. (1982) A method for repeated evaluation of pulmonary performance in unanesthetized,
unrestrained guinea pigs and its application to detect effects of sulfuric acid mist inhalation. Toxicol.
Appl. Pharmacol. 63: 72-90.
Yeh, H.-C.; Schum, G. M. (1980) Models of human lung airways and their application to inhaled particle
deposition. Bull. Math. Biol. 42: 461-480.
Yeh, H. C.; Schum, G. M.; Duggan, M. T. (1979) Anatomic models of the tracheobronchial and pulmonary
regions of the rat. Anal. Rec. 195: 483-492.
Yeh, H. C.; Brinker, R. M.; Harkema, J. R.; Muggenburg, B. A.; Guilmette, R. A. (1989) A comparative
morphometric analysis of primate nasal airways. In: Inhalation Toxicology Research Institute annual
report 1988-1989; pp. 27-28. Available from: NTIS, Springfield, VA; LMF-126.
Yokoyama, E.; Frank, R. (1972) Respiratory uptake of ozone in dogs. Arch. Environ. Health 25: 132-138.
Yokoyama, E.; Ichikawa, L; Nambu, Z.; Kawai, K.; Kyono, Y. (1984) Respiratory effects of intermittent
exposure to ozone of rats. Environ. Res. 33: 271-283.
Young, W. A.; Shaw, D. B.; Bates, D. V. (1963) Pulmonary function in welders exposed to ozone. Arch.
Environ. Health 7: 337-340.
Young, W. A.; Shaw, D. B.; Bates, D. V. (1964) Effect of low concentrations of ozone on pulmonary function in
man. J. Appl. Physiol. 19: 765-768.
8-100
-------
Yu, C. P. (1975) On equation of gas transport in the lung. Respir. Physiol. 23: 257-266.
8-101
-------
9
Integrative Summary of
Ozone Health Effects
9.1 Introduction
Characterization of the health risks associated with pollutant exposure requires an
integrative interpretation of the continuum from air quality to exposure to dose to effects,
with full consideration of the actual exposure and susceptibility of different subgroups in the
population. The currently published information on population exposure to ozone (O3) (see
Chapter 4) is superseded by recent analyses of the U.S. Environmental Protection Agency
(EPA) in a staff paper prepared by the Office of Air Quality Planning and Standards (U.S.
Environmental Protection Agency, 1996). Thus, the staff paper contains the risk assessment
for O3.
This chapter characterizes the hazard and dose-response components of risk
assessment by integrating the animal lexicological, human clinical and epidemiological, and
extrapolation studies of O3 that were discussed in Chapters 6, 7, and 8, respectively, of this
criteria document. Because each of these approaches has different strengths and weaknesses,
they were evaluated separately in the respective base chapters; however, a combined
evaluation can better describe the full array of effects that are known to occur with exposure
to O3.
The chapter begins with an overview of the relationship between exposure and
dose, as this lays a foundation for inter- and intraspecies extrapolation. The chapter then is
organized according to biological outcomes, beginning with the effects of short- and long-
term exposures of O3 alone or in ambient air and ending with experimental exposures to
binary mixtures with O3. The section on short-term exposures (i.e., <8 h) presents
descriptively symptoms and effects on lung function, exacerbation of existing disease, and
cellular-biochemical responses. Quantitative exposure-response relationships for the effects of
O3 on pulmonary function (e.g., changes in lung volume) are summarized separately because
the large number of studies allows more complex evaluation and modeling. For the other
classes of effects, the more limited exposure-response information is integrated with the
description of the effects. The section on long-term exposures encompasses repeated
exposures (i.e., 1 to 5 days), prolonged exposures (i.e., months), and genotoxicity and
carcinogenicity. Because the available database on binary exposure studies has little
predictive value of effects in populations exposed to complex pollutant mixtures, the emphasis
is placed on the principles of interaction. The conclusions section is organized according to
key questions about the health consequences of O3 exposure and the population groups that
are most likely to be affected.
9-1
-------
Because this chapter integrates the results of a large number of studies from the
current and earlier O3 criteria documents (U.S. Environmental Protection Agency, 1978,
1986), it is not practical to provide experimental details or cite specific references. Rather,
emphasis is given to main findings that are supported by published, confirmatory studies,
unless noted otherwise. Comprehensive details and references are provided in the base
chapters (Chapters 6 through 8). A discussion of selective, key references can be found in
the summary and conclusion sections of those chapters.
9.2 Exposure-Dose Relationships
Qualitative and quantitative health assessments require, among other things, the
ability to relate exposure to dose and dose to effect. In the case of O3 health assessment, this
ability is necessary for two major reasons: (1) to develop unified predictive models of human
population responses based on exposure, and (2) to enable extrapolation from animals to
humans for chronic effects. Physically and biologically based models of dose simplify the
methods of predicting population responses and, in turn, significantly reduce the uncertainty
of these predictions. For animal-to-human extrapolations, splitting the problem of exposure
and response into an exposure-dose problem and a dose-response problem separates the issue
of interspecies sensitivity from purely dosimetric considerations. Responses in animals may
be homologous with responses in humans but follow different dose-response curves.
By measuring or computing delivered O3 dose to relevant tissues in animals and humans,
transfer functions can, in principle, be developed relating dose-response curves among
different species, assuming tissues from different species react in identical fashion. This
section discusses the understanding of exposure-dose relationships and how they improve the
ability to interpret and predict O3 responses.
Historically, the first step beyond describing responses solely in terms of exposure
concentration was the use of the product of concentration x time x minute ventilation (C x T
x "VE), yielding what often has been referred to as an "effective dose". Response modeling
has examined the interaction of individual pairs of variables. However, no single model has
been able to simply unify any response in terms of the product of C x T x VE. This is due to
the fact that C x T x VE is a metric of exposure dose rather than delivered dose and,
furthermore, does not account for the mediation of responses in localized regions of the lung
that would be responding to local O3 doses. Advances in O3 dosimetry modeling and
experimental determinations of regional O3 dose in animals and humans have enabled
extensions beyond simple C x T x VE modeling to interpret responses.
Ozone dosimetry models provide predictions of the dose distribution of O3 in the
respiratory tract from the trachea to the alveolar spaces of the lung. These models utilize the
best available anatomical, physiological, and biochemical data available for animals and
humans. These data are incorporated into mathematical formulations of convection, diffusion,
and chemical reaction processes in the lung. The models predict that, under resting
ventilatory conditions, the O3 dose per airway generation to all respiratory tract constituents
(tissue plus fluid) slowly decreases from the trachea to the terminal and respiratory
bronchioles and then declines in the alveolated generations. When dose of O3 to tissue alone
is considered (accounting for reaction and diffusion kinetics in the liquid lining layer), there is
a three order of magnitude increase in tissue dose from the trachea to the proximal alveolar
regions (PARs), after which the tissue and total dose are virtually equal and fall rapidly in the
9-2
-------
alveolated generations. Currently, relationships between delivered regional dose and response
are derived assuming that O3 is the active agent directly responsible for effects; however,
there is uncertainty as to the correctness of this assumption. Reactive intermediates, such as
peroxides and aldehydes formed when O3 interacts with constituents of lung lining fluid, may
be the agents mediating responses. Thus, the dose of the reactive intermediates may be
relevant. Even in the presence of uncertainty over the relevant dose agent, the
histopathological findings from chronic O3 exposures in animals match the predicted
distribution of O3 dose (i.e., the sites of the highest predicted O3 doses correspond with those
regions of the lung with the greatest tissue alterations).
Experimental studies in humans have revealed some important features needed for
health assessment. Among these is the observation that the dose of O3 delivered to the lower
respiratory tract is independent of the mode of breathing (i.e., oral versus nasal versus
oronasal). This observation simplifies health assessment by eliminating the need for precise
information on modes of breathing when considering population responses. Experimental
studies in humans also have shown that increasing "VE with exercise (increasing both
breathing frequency and tidal volume) causes only a small decrease in O3 uptake efficiency by
the total respiratory tract. Based on models of O3 dose, it appears that the increased "VE in
exercise, although having little effect on uptake efficiency by the total respiratory tract, causes
the distribution of delivered O3 dose to shift deeper into the respiratory tract. The shift in O3
dose as a function of VE could help explain the complex relationships seen between response
and C, T, and VE. An important observation from the human experimental dosimetry studies
is the general agreement between O3 dosimetry models and the measured data.
Experiments in laboratory animals (particularly rats) have been valuable in
providing, in conjunction with human experimental data and mathematical dosimetry models,
the basis for dosimetric extrapolation. Whereas the human total respiratory tract has an
O3 uptake efficiency between 70 and 100%, the respiratory tract of the rat takes up only about
50% of the inhaled O3. Unlike the case with humans, the dosimetry models overestimate the
uptake efficiency of the rat respiratory tract by approximately 25 to 50% (i.e., the predicted
uptake efficiency is between 65 and 75%), but the models are still highly valuable for
extrapolation purposes. An important finding has been that the models correctly relate the
regional dose of O3 to the increase in alveolar wall thickness, both of which decline with
distance from the junction of the conducting airways and the alveolar regions of the lung.
Experimental O3 dosimetry and predictive O3 dosimetry models are informative
about the feasibility of extrapolating animal responses to humans. Some acute responses to
O3 can be compared across species on a strict dose-response basis. For example, both
animals and humans respond to O3 in a dose-dependent manner by increasing breathing
frequency and decreasing tidal volume (tachypnea). A qualitative comparison between rat and
human tachypneic responses at a variety of O3 concentrations and exercise levels indicates
that when exercising, rats and humans have a similar response, but rats are somewhat more
responsive at rest. However, when dose to the proximal alveolar region of the lung
(normalized to body weight) is considered as the dose metric for tachypneic responses, rats
appear to be much more responsive than humans. Another example is influx of protein into
the alveolar spaces following O3 exposure as measured in bronchoalveolar lavage (BAL)
fluid. When BAL protein is plotted as a function of pulmonary tissue dose, the rat, guinea
pig, rabbit, and human all respond with a similar dose-response pattern, suggesting a common
mechanism of response. However, each curve is offset from the other, reflecting overall
9-3
-------
sensitivity differences among the species, with the human and guinea pig being more
responsive than the rat and rabbit.
Available data on chronic responses to O3 are considerably more difficult to
compare across species. Specific assumptions are required to model exposure-dose
relationships. For example, allometric anatomic adjustments provide estimates of unavailable
dosimetric data for nonhuman primates, and allometric equivalent life-span estimates better
relate the duration of exposure to life-span and cumulative dose. Studies of long-term
exposure in monkeys and rats show a near-linear dose-response pattern when alveolar
interstitial thickness was related to cumulative dose estimates for the PAR of the lung.
Analogous estimates of PAR dose in humans predict similar increases in interstitial thickness
at the PAR, with the monkey being more responsive, and the rat less responsive.
9.3 Effects of Short-Term Ozone Exposures
9.3.1 Physiological Responses to Ozone Exposure
Typical acute physiological responses to O3 exposure observed in both human
clinical and field studies include a reduction in forced vital capacity (FVC), decreased
expiratory flow rates, and increased respiratory symptoms. The most common symptoms
include cough, airway irritation, and chest discomfort associated with deep inhalation. These
responses are often accompanied by increased airway resistance and tachypnea. The
voluntary spirometry and symptom responses cannot be elicited from animals, but their
tachypneic response is well documented. Ozone exposure also increases airway responsiveness
to nonallergenic airway stimuli (e.g., histamine) in humans and animals. There is a large
range of physiological responses among humans, with at least a 10-fold difference between
the most and least responsive individuals.
9.3.1.1 Respiratory Symptom Responses
An association between O3 exposure and the presence of symptoms has been
shown in human clinical, field, and epidemiological studies. Prevalent symptoms include
cough, irritation of the airways (described as a scratchy throat or discomfort under the
sternum), and discomfort when taking a deep breath (described as chest tightness or pain in
the chest). Eye irritation, sometimes reported as a symptom in field or epidemiological
studies with exposure to oxidant mixtures including peroxyacyl nitrates, is not associated with
exposure to O3 alone. The most prevalent respiratory symptoms have a much higher
incidence in young adults than in older adults and generally are not reported in children or
adolescents. Asthmatics have symptoms similar to nonasthmatics but also report a higher
incidence of wheezing. The receptors responsible for cough may be unmyelinated C-fibers or
rapidly adapting receptors located in the larynx and the largest conducting airways. Thus,
there appears to be a potential mechanistic linkage between coughing and changes in
spirometry. Field and epidemiological studies also indicate an association between hourly or
daily ambient O3 levels and the presence of respiratory symptoms, particularly cough. Such
associations may be most evident in asthmatic children. Although symptoms cannot be
elicited from animals, indirect measures of symptom responses in animals include behavioral
responses (e.g., decreased wheel-running activity, decreased activity associated with obtaining
food) indicative of aversion to O3 exposure.
9-4
-------
Symptom responses to O3 exposure follow a monotonic exposure-response
relationship that has a similar form to that for spirometry responses. Increasing exposure
levels elicit increasingly more severe symptoms that persist for longer periods. Symptom and
spirometry responses follow a similar time course during an acute exposure and the
subsequent recovery, as well as over the course of several days in a repeated exposure study.
Furthermore, medication interventions that block or reduce spirometry responses have a
similar effect on symptom responses. Levels at which symptoms occur under various
exposure conditions are discussed in Section 9.3.4.2. As with spirometry responses, symptom
responses vary considerably among subjects, although the individual correlations between
spirometry and symptom responses are relatively low. Ozone induces interference with
exercise performance, either by reducing maximal sustainable levels of activity or by reducing
the duration of activity that can be tolerated at a particular work level; this is likely related to
symptoms. In several heavy or severe exercise studies of athletes exposed to O3, the
discomfort associated with the respiratory symptoms caused by O3 concentrations in excess of
0.18 ppm was of sufficient severity that the athletes reported that they would have been
unable to perform maximally if the conditions of the exposure were present during athletic
competition. In workers or active people exposed to ambient O3, respiratory symptoms may
cause reduced productivity or may curb the desire to pursue certain leisure activities.
9.3.1.2 Lung Function Responses
Epidemiological, field, and chamber studies have demonstrated that acute exposure
to O3 decreases FVC and forced expiratory volume in 1 s (FEVj). In humans, O3 exposure
reduces FVC primarily by decreasing inspiratory capacity. This is believed to be the result of
neurogenic inhibition of maximal inspiration, possibly caused by stimulation of C-fiber
afferents, either directly or from O3-induced products of inflammation. C-fibers are also
thought to be the receptors responsible for the cough reflex in humans. After exposure to O3,
coughing frequently is elicited during the deep inspiration prior to the forced expiratory
maneuver used in dynamic spirometry tests such as FVC, FEVl3 and forced expiratory flow at
25 to 75% of FVC. The observation that nonsteroidal anti-inflammatory drugs (e.g.,
indomethacin, ibuprofen) reduce or block spirometric responses to O3 exposure and reduce
levels of prostaglandin E2 (PGE2) within the lung suggests that mediators released by
damaged epithelial cells and alveolar macrophages may play a role in the inhibition of
maximal inspiration. Although it seems clear that the reduction in total lung capacity is not
attributable to reduced static compliance (i.e., a stiffer lung ) or inspiratory muscle weakness,
other mechanisms may be involved. Increased interstitial fluid in patients with heart disease
causes a decrease in vital capacity and frequency-dependent decreases in lung compliance.
The O3-induced tachypneic response, seen in many animal species and in exercising humans,
may be related to the decrease in vital capacity. In humans, the pulmonary reflexes that
inhibit maximal inspiration may also limit tidal volume during exercise, which leads to a
compensatory tachypnea. The tachypneic response in humans may not be entirely involuntary
because it has been reported that O3-exposed subjects may consciously modify their breathing
pattern to relieve discomfort.
The time course of the spirometry responses to O3 exposure depends on the
exposure conditions. At low levels of exposure (e.g., light exercise and O3 concentration
<0.18 ppm), responses are induced slowly and progressively and they may or may not reach a
plateau of response, depending on the duration of the exposure. At higher levels of exposure
(e.g., very heavy exercise and O3 concentration >0.25 ppm), responses occur rapidly (within
9-5
-------
15 min), and the largest portion of the response tends to occur early in exposure, indicative of
a plateau of response that typically is not achieved because of termination of the exposure
within 1 to 2 h. The quantitative exposure-response relationships are discussed more
extensively in Section 9.3.4.
In both chamber and field studies, the responses of healthy children to acute
O3 exposure are similar to those seen in adults. Responses of children and adolescents
exposed to ambient O3 (and other copollutants) at summer camps in at least six different
locations in the northeastern United States, southeastern Canada, and Southern California
indicate changes in spirometry similar to those found in individuals exposed to O3 under
controlled experimental conditions. There is a substantial range of response among
individuals in camp studies and between various locations; however, the average FEVj was
lower when ambient O3 was higher. Although direct comparisons cannot be made because of
incompatible differences in experimental design and analytical approach, this range of
response is comparable to the range of response seen in chamber studies at low
O3 concentrations. In the "camp studies", a key measurement is the slope of the relationship
between FEVj and the measured O3 concentration during the previous hour, without
consideration of the background O3 levels (even though exposures occurred over multiple
hours). The average slope from six studies was -0.50 mL/ppb within an O3 concentration
range of 0.01 to 0.16 ppm (see Chapter 7). For an exposure to 0.12 ppm O3, this corresponds
to a decrease in FEVj of 60 mL from a base level of approximately 2,000 to 2,500 mL, or
a 2.4 to 3.0% decrease in FEVj. This is comparable to the findings of McDonnell et al.
(1985) for 8- to 11-year-old boys who experienced a 3.4% decrease in FEVj after being
exposed to 0.12 ppm O3 for 2 h. Recent studies in adults performing outdoor exercise also
have shown an association between decreased spirometric responses and increased ambient
O3 levels.
A consistent finding across many animal species is that O3 causes rapid shallow
breathing (O3-induced tachypnea), which, in humans, may be related to a sensation of
discomfort associated with taking large tidal breaths. Of particular interest for comparing
interspecies responses is that the responses of rats and guinea pigs fall within the same range
as seen for humans from rest to heavy exercise.
Common pulmonary function tests do not measure changes in the small airways of
the centriacinar region of the lung (that segment between the last conducting airway and the
gas exchange region), which is highly susceptible to damage by O3 and is the site of epithelial
cell necrosis and remodeling of respiratory bronchioles. Numerous pulmonary function tests
reputed to measure responses in small airways (e.g., closing volume, aerosol bolus) have been
used in O3 studies. Responses have been demonstrated, but it is not clear that these tests
correlate with the morphological lesion observed in the small airways of experimental animals
(see Section 9.4), which is predicted to occur in humans but has not been confirmed reliably
through comparable morphologic data from humans residing in O3-polluted areas. Even if
small airways disease is demonstrated in humans, there is as yet no compelling evidence that
it will progress to chronic lung disease.
An increase in airway resistance is an indication of the response of large airways
to O3 exposure and is mechanistically different from lung volume responses. Also, higher
O3 concentrations are required to change airway resistance compared to FEVj. Changes in
specific airway resistance (SRaw) of healthy subjects following O3 exposure are small relative
to those seen in asthmatics with an inhalation exposure to a bronchoconstricting drug
(methacholine), a specific antigen, or sulfur dioxide (SO2). In rats exposed to O3, changes in
9-6
-------
resistance also tend to be small. The observation that changes in airway resistance are
modest clearly indicates that reductions in maximum expiratory flow are not caused primarily
by narrowing of large airways. The increase in airway resistance appears to be vagally
mediated because it is sensitive to inhibition by atropine.
9.3.1.3 Changes in Airway Responsiveness
Ozone exposure also causes increased responsiveness of the pulmonary airways to
subsequent challenge with bronchoconstrictor drugs such as histamine or methacholine. This
phenomenon is seen even after recovery from spirometric changes, but it typically is no
longer present after 24 h. Although changes in airway responsiveness tend to resolve
somewhat more slowly and appear to be less likely to be attenuated with repeated exposure,
the evidence for a persistent increase in responsiveness from animal studies is inconsistent.
Changes in airway responsiveness in rats and guinea pigs tend to occur at higher
O3 concentrations and, as in humans, tend to be most pronounced shortly after the exposure
and less so 24 h postexposure. Changes in airway responsiveness appear to occur
independently of changes in pulmonary function. This response may not be due to the
presence of polymorphonuclear leukocytes (PMNs) in the airway or to the release of
arachidonic acid metabolites, but could possibly be due to epithelial damage and the
consequent increased access of these chemicals to smooth muscle in the airways or to the
receptors in the airways responsible for reflex bronchoconstriction. The clinical relevance of
this observation is that, after O3 exposure, human airways may be more susceptible to a
variety of stimuli, including antigens, chemicals, and particles. One animal study has
demonstrated decreased antigen-induced bronchoconstriction after O3 exposure, and a human
study in allergic asthmatics is suggestive of an increase in such a response. An increased
response to inhalation of a specific antigen to which a human is sensitized is a plausible
outcome of O3 exposure. However, ongoing studies of this phenomenon will need to be
evaluated in order to determine the exposure-response relationship for alterations in responses
to inhaled antigens, especially with regard to sensitive asthmatics. Enhanced response to
antigens in asthmatics could lead to increased morbidity (i.e., medical treatment, emergency
room visits, hospital admissions) or to more persistent alterations in airway responsiveness.
9.3.2 Exacerbation of Respiratory Disease
People with preexisting pulmonary disease may be at increased risk from
O3 exposure. Because of their existing functional limitations, any further decrease in function
would lead to a greater overall functional decline. Furthermore, some individuals with
pulmonary disease may have an inherently greater sensitivity to O3. Asthmatics, by
definition, have inherently greater bronchial responsiveness, but, depending on the severity of
their disease and its clinical status, their FEVj can be within the normal range (100 ± 20%
predicted) or may be less than 50% predicted. Patients with chronic obstructive pulmonary
disease (COPD) can have FEVs ranging from 30 to 80% of predicted, again depending on
disease severity. Because of their depressed functional state, small absolute changes in lung
function have a larger relative impact. For example, a 500-mL FEVj decrease in a healthy
young man with an FEVj of 4,000 mL causes only a 12% decline. In a 55-year-old COPD
patient with an FEVj that is 50% of predicted, or about 1,670 mL, a 500-mL decline in FEVj
would result in a 30% decline in FEVj. Asthmatics with depressed baseline function would
have similarly magnified relative responses and, because of increased bronchial
9-7
-------
responsiveness, may also experience larger changes in airway resistance. Evaluating the
intersection of risk factors and exposures is more complex. For example, an individual with
more severe lung disease is unable to exercise heavily and thus would be less likely to
encounter an effective exposure.
About 10 million people in the United States (4% of the population) are estimated
to have asthma (National Institutes of Health, 1991). The prevalence is higher among African
Americans, older (8- to 11-year-old) children, and urban residents (Schwartz et al., 1990).
Death due to asthma is an infrequent event; on an annual basis, about one death occurs per
10,000 asthmatic individuals. Mortality rates are higher among males and are at least 100%
higher among nonwhites. In two large urban centers (New York and Chicago), mortality
rates from asthma among nonwhites may exceed the city average by up to fivefold (Sly,
1988; Evans et al., 1987; National Institutes of Health, 1991; Weiss and Wagener, 1990; Carr
et al., 1992). Although some innercity areas may have lower O3 concentrations than some
suburban areas, the concentrations are much higher than those in most rural areas. The
impact of ambient O3 on asthma morbidity and mortality in this apparently susceptible
population is not well understood. The few epidemiological studies are subject to
confounding factors and have rarely focused on innercity nonwhite asthmatics. Furthermore,
controlled human exposure studies of asthmatics typically include mild to moderate asthmatics
and have not dealt specifically with nonwhite asthmatics.
A number of epidemiological studies have shown a consistent relationship between
ambient oxidant exposure and acute respiratory morbidity in the population. Small decreases
in forced expiratory volumes and increased respiratory symptoms, including exacerbation of
asthma, occur with increasing ambient O3, especially in children. Modifying factors, such as
ambient temperature, aeroallergens, and other copollutants (e.g., particles) also can contribute
to this relationship. Ozone air pollution can account for a portion of summertime hospital
admissions and emergency department visits for respiratory causes; studies conducted in
various locations in the eastern United States and Canada consistently have shown a
relationship with increased incidence of visits and admissions, even after controlling for
modifying factors, as well as when considering only concentrations <0.12 ppm O3. It has
been estimated from these studies that O3 may account for roughly one to three excess
summertime respiratory hospital admissions per hundred parts per billion O3, per million
persons.
The association between elevated ambient O3 concentrations during the summer
months and increased hospital visits and admissions has a plausible biologic basis in the
physiologic, symptomatic, and field study evidence discussed earlier. Specifically, increased
airway resistance, airway permeability, and incidence of asthma attacks and airway
inflammation suggest that ambient O3 exposure could be a cause of the increased hospital
admissions, particularly for asthmatics.
The associations found in the epidemiological studies are supported by chamber
studies. Asthmatics and nonasthmatics have qualitatively similar responses to chamber
O3 exposures. Although symptom and volume-related responses (i.e., decreased FVC) tend
to be similar, airway resistance increases relatively more, from an already higher baseline, in
asthmatics exposed to O3. Ozone-induced alterations in responsiveness to bronchoconstrictor
drugs show similar changes in asthmatics and nonasthmatics. There is no evidence at this
time that O3 induces a persistent increase in airway responsiveness or that O3-exposed
asthmatics are more likely to have a late-phase response to specific antigen challenge.
Symptom responses also have been reported in asthmatics exposed to O3. In contrast to
9-8
-------
nonasthmatics, wheezing, a typical finding in asthma, is a prevalent symptom in addition to
the cough, chest tightness, and shortness of breath that are reported by subjects without
asthma.
9.3.3 Morphological and Biochemical Abnormalities
9.3.3.1 Inflammation and Cell Damage
Ozone-induced cell injury may lead to effects including inflammation, altered
permeability of the epithelial barrier, impaired host defense and particle clearance, irreversible
structural alterations in the lung, exacerbation of preexisting disease (e.g., asthma), and
increased sensitivity to biocontaminants (e.g., allergens). Of these, O3-induced inflammation
of the respiratory tract has been best documented and occurs in all species that have been
studied. The mechanisms leading to the observed inflammatory responses induced by O3 are
just beginning to be understood. Both animal morphological studies and in vitro studies
indicate that airway ciliated epithelial cells and Type 1 cells are the most O3-sensitive cells
and are initial targets of O3. These cells are damaged by O3 and produce a number of
proinflammatory mediators (e.g., interleukins [IL-6, IL-8], PGE2) capable of initiating a
cascade of events leading to PMN influx into the lung, activation of alveolar macrophages,
inflammation, and increased permeability across the epithelial barrier.
Ozone-Induced Inflammation
Ozone causes inflammatory changes throughout the respiratory tract, including the
nose. Humans and laboratory animals exposed to O3 develop inflammation and increased
permeability in the nasal passages. A recent study reported a positive correlation between
nasal inflammation in children and measured ambient O3 concentrations. Studies with rats
suggest a potential competing mechanism between the nose and lung, with inflammation
occurring preferentially in the nose at low O3 concentrations and shifting to the lung at higher
concentrations. It is unclear if this represents a specialization restricted to rats or is a more
general phenomenon.
In general, inflammation can be considered as the host response to injury and the
induction of inflammation as evidence that injury has occurred. Inflammation induced by
exposure of humans to O3 can have several potential outcomes: (1) inflammation induced by
a single exposure (or several exposures over the course of a summer) can resolve entirely;
(2) continued acute inflammation can evolve into a chronic inflammatory state; (3) continued
inflammation can alter the structure and function of other pulmonary tissue, leading to
diseases such as fibrosis; (4) inflammation can alter the body's host defense response to
inhaled microorganisms, particularly in potentially vulnerable populations such as the very
young and old; and (5) inflammation can alter the lung's response to other agents such as
allergens or toxins. Except for outcome (1), the possible chronic responses have not been
identified with inflammation induced by exposure of humans to O3. It is also possible that
the profile of response can be altered in persons with preexisting pulmonary disease (e.g.,
asthma, COPD) or smokers.
The recent use of BAL as a research tool in humans has afforded the opportunity
to sample cells and fluid from the lung and lower airways of humans exposed to O3 and to
ascertain the extent and course of inflammation and its constitutive elements. Several studies
have shown that humans exposed acutely (1 to 3 h) to 0.2 to 0.6 ppm O3 had O3-induced
inflammation, cell damage, and altered permeability of epithelial cells lining the respiratory
9-9
-------
tract (allowing components from plasma to enter the lung). For individuals acutely exposed
to 0.4 to 0.6 ppm O3, PMNs (the hallmark cells of inflammation) make up 8 to 10% of the
recovered BAL cells. This represents a five- to eightfold increase in PMNs compared with
similar individuals exposed to clean air, who typically have 1 to 2% PMNs in their BAL
fluid. The lowest concentration of O3 tested, 0.08 ppm for 6.6 h with moderate exercise, also
induced small but statistically significant increases in a number of inflammatory mediators,
including PMNs.
The percent of PMNs in BAL fluid taken from individuals exposed to 0.4 ppm
O3 for 2 h equals or exceeds those found in individuals exposed to other environmental
toxicants, such as asbestos or silica, or in individuals with idiopathic pulmonary fibrosis (IFF)
or connective tissue disorders (CTD) (Cherniak et al., 1990). For example, individuals with a
history of occupational exposure to asbestos (>10 years) have 3.3 ± 1.3% BAL PMNs, and
individuals with a history of occupational exposure to silica (>2 years) have 1.4 ± 0.4%
PMNs. Untreated patients newly diagnosed with IFF have 6.7 ± 2.5% PMNs, whereas those
with CTD have 16 ± 11.6% PMNs. Baseline levels of PMNs in patients with asthma do not
differ significantly from healthy individuals, although PMN levels can increase following
allergen bronchoprovocation (Smith and Deshazo, 1993). In contrast, PMNs can make up as
much as 80% of BAL cells in patients with acute bacterial infections (Stanley, 1991).
Short-term (<8 h) exposure of animals to O3 also results in cell damage,
inflammation, and altered permeability, although, in general, higher O3 concentrations are
required to elicit a response equivalent to that of humans. Because humans were exposed to
O3 while exercising and most animal studies were done at rest, differences in ventilation
likely play a significant role in the different response of humans and rodents to the same
O3 concentration. Studies in which animals were exposed at night (during their active period)
or in which ventilation was increased with CO2 tend to support this idea.
Studies utilizing BAL techniques sample only free or loosely adherent cells in the
lung; thus, it is possible that cellular changes have occurred in the interstitium that are not
reflected in BAL studies, or that BAL changes exist in the absence of interstitial changes.
However, morphometric analyses of inflammatory cells present in lung and airway tissue
sections of animals exposed to O3 are in general agreement with BAL studies. Short-term
O3 exposure (<8 h) causes similar types of alterations in lung morphology in all laboratory
animal species studied. The most affected cells are the ciliated epithelial cells of the airways
and Type 1 cells in the alveolar region. The centriacinar region (the junction of the
conducting airways and gas exchange region) is a primary target, possibly because it receives
the greatest dose of O3 delivered to the lower respiratory tract. Sloughing of ciliated
epithelial and Type 1 cells occurs within 2 to 4 h of exposure of rats to 0.5 ppm O3.
Time Course of Ozone-Induced Inflammatory Response
Findings from human and animal studies agree that the O3-induced inflammatory
response occurs rapidly and persists for at least 24 h. Increased levels of PMNs and protein
are observed in the BAL fluid within 1 h following a 2-h exposure of humans to O3 and
continue for at least 20 h. The kinetics of response during this time have not been well
studied in humans, although a single study shows that PMN levels are higher at
6 h postexposure than at 1 or 20 h in different individuals. Several animal studies suggest
that BAL PMN and protein levels peak 12 to 16 h after an acute O3 exposure and begin to
decline by 24 h, although some studies report detectable BAL PMNs even 36 h after
exposure. It is also clear that in humans the pattern of response differs for different
9-10
-------
inflammatory mediators. Mediators of acute inflammation, such as IL-6 and PGE2, are more
elevated immediately after exposure; whereas mediators that potentially could play a role in
resolving inflammation, such as fibronectin and plasminogen activator, are preferentially
elevated 18 h after exposure. The rapidity with which cellular and biochemical mediators are
induced by O3 makes it conceivable that some of them may play a role in O3-induced changes
in lung function—indeed there is some evidence that BAL PGE2 levels are correlated with
decrements in FEVl3 and anti-inflammatory medications that block PGE2 production also
reduce or block the spirometric responses to O3. Although earlier studies suggested that
O3-induced PMN influx might contribute to the observed increase in airway hyperreactivity,
animal studies show that when PMNs are prevented from entering the lung, O3-induced
hyperreactivity or increases in many inflammatory mediators still occur. In addition, studies
in which anti-inflammatory drugs are used to block O3-induced lung function decrements still
show increases in PMNs and most other inflammatory mediators (although PGE2 is not
increased).
Individuals and Populations Susceptible to Ozone
To date, there have been no studies that have examined the cellular/biochemical
response to O3 of potentially susceptible subpopulations, such as asthmatics, nor are there any
data in humans addressing whether age, gender, or racial differences can modify the
inflammatory response to O3. Increased susceptibility of asthmatics or chronic bronchitics
could be hypothesized on the basis that they have an underlying inflammatory disease that
may be exacerbated with an otherwise small magnitude of change. Inflammation is not
induced to the same extent in all individuals. In moderately exercising humans exposed to
0.08 ppm O3 for 6.6 h, the mean changes in inflammatory indices were low, but some
individuals had increases comparable to those reported in heavily exercising subjects exposed
to 0.4 ppm O3 for 2 h, suggesting that some segments of the population may be more
responsive to low levels of O3. It has not yet been studied whether intersubject differences in
inflammatory response to O3 are reproducible over time for the same subject, as has been
shown for intersubject differences in lung function. There seems to be no strong correlation
between the various mediators of inflammation, cell damage, and permeability (i.e., those
individuals with the greatest PMN response are not necessarily those with the greatest BAL
protein, PGE2, or IL-6 response). Furthermore, the magnitude of lung function decrements
and respiratory symptoms has not yet been shown to be correlated with mediators of
inflammation, with the possible exception of PGE2.
Animal studies also show large interspecies and interstrain differences in response
to O3 and suggest that genetic factors may play a role in susceptibility to O3. Different rat
strains respond to O3 differently; for example, Wistar rats have the greatest PMN influx,
whereas Fischer rats demonstrate the most epithelial cell damage. In addition, limited data
suggest that dietary antioxidant levels may affect the response of rodents to O3 and that very
young rats produce more PGE2 in response to O3 than do older rats. Taken as a whole, the
human and animal studies suggest that the inflammatory response to O3 is complex and that
determinants of susceptibility may occur at several different genetic loci.
9-11
-------
9.3.3.2 Host Defense
The mammalian respiratory tract has a number of closely integrated defense
mechanisms that, when functioning normally, provide protection from the adverse effects of a
wide variety of inhaled particles and microbes. Impaired mucociliary clearance can result in
unwanted accumulation of cellular secretions and increased numbers of particles and
microorganisms in the lung, leading to increased infections and bronchitis.
Mucociliary Clearance of Inhaled Particles
Animal studies show that clearance of inhaled insoluble particles is slowed after
acute exposure to O3. Ozone-induced damage to cilia and increased mucus secretion likely
contribute to a slowing of mucociliary transport rates. Interestingly, retarded mucociliary
clearance is not observed in animals exposed repeatedly to O3. The effects of O3 on
mucociliary clearance in humans have not been well studied, and the results are somewhat
conflicting; one study reports an O3-induced increase in particle clearance in subjects exposed
to 0.4 ppm O3 for 2 h, and another study reports no O3-induced change in particle clearance
with a similar exposure regimen.
Alveolar Macrophage Function
Macrophages represent the first line of defense against inhaled microorganisms and
particles that reach the lower airways and alveoli. Studies in both humans and animals have
shown that there is an immediate decrease in the number of BAL macrophages following
O3 exposure. Alveolar macrophages also have been shown to be crucial to the clearance of
certain gram-positive bacteria from the lung. Several studies in both humans and laboratory
animals also have shown that O3 impairs the phagocytic capacity of alveolar macrophages,
and some studies suggest that mice may be more impaired than rats. The production of
superoxide anion (an oxygen radical used in bacterial killing) by alveolar macrophages also is
depressed in both humans and animals exposed to O3, and the ability of alveolar macrophages
to kill bacteria directly is impaired. Decrements in alveolar macrophage function have been
observed in moderately exercising humans exposed to the lowest concentration tested, 0.08
ppm O3 for 6.6 h.
Interaction with Infectious Agents
Concern about the effect of O3 on susceptibility to respiratory infection derives
primarily from animal studies in which O3-exposed mice die following a subsequent challenge
with aerosolized bacteria. Increased mortality of experimental laboratory animals has been
shown to be concentration-dependent, and exposure to as little as 0.08 ppm O3 for 3 h can
increase mortality of mice to a subsequent challenge with streptococcus bacteria. In addition,
younger mice are more susceptible to infection than older mice; this has been related to
increased PGE2 production in these animals, which likely decreases alveolar macrophage
activity.
It has been suggested that impaired alveolar macrophage function is the mechanism
likely responsible for enhanced susceptibility to bacteria. However, mortality is not observed
with other rodent species, raising the question of whether this phenomenon is restricted to
mice. Although both mice and rats show impaired macrophage killing of inhaled bacteria
following O3 exposure, rats mount a faster PMN response to O3 to compensate for the deficit
in alveolar macrophage function. The slower clearance time in mice allows the streptococcus
strain to persist in lung tissue and, subsequently, to elaborate a number of virulence factors
9-12
-------
that evade secondary host defense and lead to bacterial multiplication and death of the host.
Although increased mortality in laboratory animals is not directly relevant to humans,
laboratory animals and humans share many host defense mechanisms being measured by
mortality in the mouse model. Thus, the category of effect (i.e., decrement in antibacterial
defenses) can be qualitatively extrapolated to humans.
There is no compelling evidence from animal lexicological, human clinical, or
epidemiological studies that O3 increases the incidence of respiratory viral infection in
humans. A study of experimental rhinovirus infection in susceptible volunteers failed to show
any effect of 5 consecutive days of O3 exposure (0.3 ppm, 8 h/day) on the clinical picture or
on host response. Studies in which O3-exposed mice were challenged with influenza virus
have conflicting results: some studies show increased mortality, some show decreased
mortality, and still others show no change at all. However, even when increased mortality
was demonstrated, there was no difference in viral liters in the lung, suggesting virus-specific
immune functions were not altered. One animal study found that, although subchronic
O3 exposure did not affect the acute course of a viral infection, it did enhance postinfluenzal
alveolitis.
Taken as a whole, the data clearly indicate that an acute O3 exposure impairs the
host defense capability of both humans and animals, primarily by depressing alveolar
macrophage function and perhaps also by decreasing mucociliary clearance of inhaled
particles and microorganisms. This suggests that humans exposed to O3 could be predisposed
to bacterial infections in the lower respiratory tract. The seriousness of such infections may
depend on how quickly bacteria develop virulence factors and how rapidly PMNs are
mobilized to compensate for the deficit in alveolar macrophage function.
Ozone also has been reported to suppress natural killer cell activity in the lung, to
suppress proliferative responses to bacterial antigen (Listeria) in both spleen and bronchial
lymph nodes, and to induce delayed hypersensitivity responses to Listeria antigen. However,
these effects occur at higher exposure levels (0.75 to 1.0 ppm O3) than those that affect
macrophage function.
9.3.4 Quantitative Ozone Exposure-Response Relationships
A quantitative understanding of the relationship between O3 exposure and
subsequent response is useful both for a better understanding of the processes underlying
outcomes of interest and for purposes of prediction. Examples of the utility of the latter
include identification of exposures unlikely to produce effects, risk and benefits assessment,
and prediction of responses based on exposures for which empirical data do not exist.
In general, exposure-response relationships have been better characterized for populations than
for individuals, and, although the form of the relationships may differ, those for individuals
are likely to be qualitatively similar to those of populations. On the other hand, because of
large differences in responsiveness among individuals, exposure-response relationships for the
population may not reflect quantitatively the experience of a given individual.
Relationships between short-term exposure and acute response have been described
for lung function changes, induction of symptoms, and BAL outcomes in experimental
exposure studies and for lung function, symptoms, hospital and emergency room admissions,
and mortality in epidemiologic studies. Exposure in the experimental studies can be defined
in terms of concentration, dose rate of exposure, total inhaled dose, and dose at the active
site. The limiting factor in modeling exposure-response with experimental data has been that
9-13
-------
no single study has included a wide enough range of the three exposure variables of interest
(i.e., C, VE, T) to choose between models or to identify the appropriate method of describing
exposure.
Exposure-response models in the epidemiologic studies generally have included
only O3 concentration measured at a central monitoring site in the study area as the exposure
variable. With some exceptions, characterization of exposure-response relationships in these
studies has been limited by little information on activity level or duration of exposure, a
generally narrow range of exposure concentrations, the need for complex models to control
for potential confounding by other pollutants and extraneous variables, and outcomes for
which only a small fraction of the variance is explained by exposure to pollutants. These
factors make selection among various models of response difficult.
A number of exposure-response functions have been proposed to describe the
results of experimental studies. No single exposure-response model form, however, has been
adequately tested and identified as providing an accurate, precise description of the
relationship between exposure and response in both humans and laboratory animals for lung
function or BAL endpoints. Rather, for a given study, a particular model may have been
selected a priori to describe the exposure-response data or may have been identified as
providing the best fit among several competing models. In many cases, models have been
found to be deficient, but rarely has the performance of a number of possible models been
systematically compared.
From the individual studies, several important observations have been made that
are qualitatively true for describing BAL and pulmonary function responses in both humans
and laboratory animals and that should be considered in the selection of a model to describe
population response as a function of exposure. Response increases monotonically with C, VE,
and T, with C generally being a stronger predictor of response than VE or T. The relationship
between response and one of the exposure variables is dependent on the level of the other two
variables. The relationship between response and each of the exposure variables is curvilinear
over a wide range of exposure conditions, although it may appear linear over certain narrow
ranges of exposure. With increasing duration of response (and possibly with increasing
concentration), the FEVj response may approach a plateau in humans. Some evidence exists
suggesting that the level of the plateau with T is a function of C. This plateau has not been
observed in animal studies or for BAL endpoints. Respiratory symptom responses generally
follow a pattern similar to that observed for spirometry (e.g., mean responses increase with
increasing C, VE, and T). As with spirometry responses, large individual differences in
symptom responses occur. Little analytical work, however, has been performed that
mathematically describes either the mean or individual responses as multivariable functions
of C, VE, and T.
Exposure-response models of FEVj and BAL responses in laboratory animals that
have been proposed and that fit to varying degrees include linear and polynomial models of
C, VE, and T, with and without cross-product terms (e.g., C x T); exponential models
utilizing C x T as the exposure variable at constant VE; and cumulative normal probability or
logistic models utilizing Cy x T as the exposure variable at constant VE. Models of these
types have been found under some circumstances to describe the relationship between
exposure and response for a particular data set. Most single data sets, however, do not
include a wide enough range of data to test adequately the performance of a particular model
across a wide range of exposure conditions or to identify an appropriate exposure metric.
In particular, recent efforts have focused on the relationship between response and C and T at
9-14
-------
constant VE. No definitive work has addressed the modeling of response and "VE for a given
endpoint or for consideration of VE changing as a function of T. Because animal and human
studies often are conducted at different relative levels of VE, and because techniques to adjust
mathematically for these differences only now are being developed, efforts to compare
responses across species or to develop extrapolation models have been hampered. As noted
earlier, quantitative models currently do not exist for respiratory symptoms.
Evidence indicates that, for humans and animals, the exposure-response
relationship of BAL and pulmonary function outcomes may be modified by previous recent
exposure to O3, and the relationship for FEVj changes in humans may be modified by age.
Previous exposure to O3 has not been included in any exposure-response models. For young
adults, the modification of the exposure-response relationship by age has been modeled.
Exposure-response models of lung function in epidemiologic studies generally have
been limited to linear models of response as functions of various O3 exposure metrics (e.g., 1-
h or 8-h daily maximum, etc.). A number of studies have demonstrated significant negative
mean linear relationships between O3 exposure and lung function. Exposure-response models
of respiratory symptoms in epidemiologic studies generally have employed logistic regression
techniques with O3 or total oxidant concentration as the exposure variable. These latter
models generally have been chosen a priori reflecting the categorical nature of the outcome
variable rather than by comparison of the performance of several candidates.
Ecological studies of the relationship between daily rates of emergency room or
hospital admissions or mortality and O3 exposure have utilized a variety of complex exposure-
response models with some metric of daily O3 concentration as the exposure variable. The
complexity of the models results from, among other things, the need to control for potentially
confounding long-wave patterns in the health outcomes in relationship to other potential
confounders, such as other air pollutants, seasonal and meteorological factors, holidays, and
day-of-week effects. In various studies, both linear and nonlinear functions have been used to
describe the relationship between adjusted health outcome and concentration.
In summary, O3 is no exception to the general problems encountered in all studies
of environmental epidemiology. No single universal model form has been identified that
accurately and precisely describes the relationship between population exposure and response
under all circumstances. In general, the ability of a predictive model based on one study to
predict responses from an independent study has not been studied adequately. For purposes
of prediction or risk estimation, the adequacy of fit of a given model in a given data set and
the size and representativeness of the sample should be assessed. Extrapolation beyond the
range of observed data introduces additional uncertainty into predictions or risk estimates.
9.3.4.1 Prediction and Summary of Mean Responses
A selection of published reports in which models of population or mean responses
have been developed is listed below, along with figures summarizing examples of predicted
quantitative exposure-response relationships. Reports numbered 5 and 11 are epidemiological
studies, and the remainder are experimental studies. Because no currently available single
model is sufficient to accurately describe all major scenarios, the key models are presented
without weighting. Following this section is a further section that describes models of
individual responses within the population.
1. Hazucha (1987) predicts mean FEVj decrements in humans as a function
of C (0.0 to 0.75 ppm O3) for four levels of VE for 2-h exposures (Figure 9-1).
9-15
-------
0.2 0.4 0.6 08
Ozone Concentration (ppm)
Figure 9-1. Mean predicted changes in forced expiratory volume in 1 s following
2-h exposures to ozone with increasing levels of intermittent exercise.
Source: Hazucha (1987).
2. McDonnell and Smith (1994) predict mean FEVj decrements in heavily
exercising humans as a function of C (0.0 to 0.4 ppm O3) and T (1.0 to 6.6 h)
(Figure 9-2).
3. Highfill et al. (1992) predict the BAL responses of resting rats and guinea pigs
as a function of C (0.0 to 0.8 ppm O3) and T (2 to 8 h) (Figure 9-3).
4. Tepper et al. (1994) predict the FVC changes as a function of C (0.0 to
0.8 ppm O3) and T (2 to 7 h) for exposures conducted with rats breathing at
three times resting VE (Figure 9-4).
5. Burnett et al. (1994) predict the frequency of hospital admissions (adjusted for
covariates) as a linear function of C (previous day 1-h O3 maximum) for
Ontario hospitals (Figure 9-5).
Other reports in which models are developed or that contain data potentially useful
for further development or testing of models are listed below.
6. Seal et al. (1993) present data that would allow modeling of FEVj decrements
in humans as a function of C (0.0 to 0.4 ppm O3) for 2-h exposures with
moderate exercise.
7. Folinsbee et al. (1978) predict lung function changes in humans as a function
of C (0.0 to 0.50 ppm O3) and VE (10 to 65 L/min) for 2-h exposures.
9-16
-------
-5
0.00 0.10 0.20 0.30 0.40 0.50
Ozone (ppm)
2345
Time (hours)
Figure 9-2. Predicted mean decrements in forced expiratory volume in 1 s for 1- and 2-h
exposures to ozone with intermittent heavy exercise (A) and 6.6-h exposures
with moderate prolonged exercise (B).
Source: McDonnell and Smith (1994).
Rats
Guinea Pigs
600
500
•3. 400
i
i *»
200
100
— o2h
— n4h
— *8h
0.2 04 0.6
Ozone (ppm)
0.8
100-
0.2 0.4 0.6
Ozone (ppm)
Figure 9-3. Derived means of BAL protein (BALP) denoted by symbols and the
exponential model shown by lines as time of exposure varies from 2 to 8 h.
Source: Highfill et al. (1992).
9-17
-------
• Observed
• Predicted
'' /' X
—""---^ii-- •' ~"~*:
Ozone (ppm)
Figure 9-4. Predicted mean forced vital capacity for rats exposed to ozone while
undergoing intermittent carbon dioxide-induced hyperpnea,
Source: Tapper et al. (1994).
114
112-
.«2 110-
e 108
| 106
.£•
§ 104
102
0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.10
Daily Maximum 1-h Level (ppm), Lagged 1 Day
Figure 9-5. Average number of adjusted respiratory admissions among all 168 hospitals by
decile of the daily 1-h maximum ozone level (ppm), lagged 1 day.
Source: Burnett et al. (1994)
9-18
-------
8. Adams et al. (1981) predict lung function changes in humans as a function of
the product of C x T x VE for C = 0 to 0.4 ppm O3, T = 30 to 80 min, and
VE = 33 or 66 L/min.
9. Rombout et al. (1989) predict the concentration of protein in BAL fluid of rats
as a function of C (0.25 to 4.0 mg/m3 O3) and T (0 to 12 h) for daytime and
nighttime exposures.
10. Highfill and Costa (1995) compare the fits of quadratic, exponential, and
sigmoid-shaped models to published human lung function data and
O3 laboratory animal BAL data.
11. Thurston et al. (1994) predict hospital admissions in Toronto as a function
of C (previous day 1-h O3 maximum) for Toronto hospitals.
9.3.4.2 Prediction and Summary of Individual Responses
It is well known that considerable interindividual differences in the magnitude of
response to O3 exposure exist. The individual lung function and, to a lesser extent,
respiratory symptom responses to O3 have been demonstrated to be reproducible over a period
of time, indicating that some individuals are consistently more responsive than others to O3.
The basis for these differences is not known, with the exception that young adults have been
observed to be more responsive than older adults (see Figure 9-6).
1.0-
0.8-
0.6-
J- Q.4-\
&
LL
0.2-1
0.0-
0.00
0.10
0.20
0.30
0.40
Ozone (ppm)
Figure 9-6. Predicted mean decrements in forced expiratory volume in 1 s
following 2-h exposures to ozone while undergoing heavy intermittent exercise
for three ages. (Note: To convert AFEVj to %kFEV,, multiply by 22.2%.)
Source: McDonnell et al. (1993).
9-19
-------
Calculation of group mean responses for a population that includes both more and
less responsive individuals is useful for making inferences regarding the probability that a
population effect is present or absent for a given exposure. Because the frequency
distribution of individual responses to O3 changes with changing exposure conditions,
however, knowledge of the mean and variance of population responses does not provide
reliable information on the distribution of individual responses for a given exposure, and,
hence, is not particularly useful for estimating risks to members of the population. One
method of presenting individual data is illustrated in Figure 9-7 in which histograms are
presented for individual responses of subjects participating in four 6.6-h studies of low-level
O3 exposure.
Distribution of Percent Change in FEV,
15
10
-5
> -1U
LJJ -15
LL -20
< -25
vp -30
*• -35
-40
•45
•50
•
^^m
n = 87
AIR
0%
15
'•
-5
-10
•15
•20
•25
•30
•35
•40
•45
•50
.
n = 60
1 0.08 ppm O3
26%
15
10
-5
-10
•15
•20
•25
•30
•35
•40
•45
•50
~ n = 32
0.10ppmO<>
- 31%
15
10
-5
-10
•15
•20
•25
•30
•35
-40
•45
•50
^^^^
~ n = 49
I 0.1 2 ppm O3
- ,46%
0 5 10 15 20 25 30 35 40 0 5 10 15 20 25 30 35 40 0 5 10 15 20 25 30 35 40 0 5 10 15 20 25 30 35 40
Percent of Subjects
Figure 9-7. The distribution of response for 87 subjects exposed to clean air and at least
one of 0.08, 0.10, or 0.12ppm ozone (OJ. The O3 exposures lasted 6.6 h,
during which time the subjects exercised for 50 min of each hour, with a 35-
min rest period at the end of the third hour. Decreases in forced expiratory
volume in 1 s (FEV,) are expressed as percent change from baseline. For
example, the bar labeled, "-10" indicates the percent of subjects with a
decrease in FEV1 of>5% but <10%, and the bar labeled "5" indicates
improvement in FEVt of>0% but <5%. Each panel of the figure indicates
the percentage of subjects at each O3 concentration with a decrease of FEVt
in excess of 10%.
Similarly, the histograms of regression slopes of the lung function-O3 concentration
relationship for children participating in a camp study illustrate a large range of variability in
response (Figure 9-8).
Another method that allows interpolation between observed data points involves
definition of the effect of interest (e.g., a 10% decrement in FEVj) and modeling of the
proportion of individuals who experience such an effect as a function of exposure conditions.
Figures 9-9 and 9-10 show the predicted proportion of individuals (humans) experiencing
10% FEVj decrements and respiratory symptom responses, respectively, as a function of
C (0.0 to 0.4 ppm O3) for independent studies conducted at either 1 or 2 h of exposure with
heavy exercise.
9-20
-------
30
20
10
All Children (91) (33)*
30
vP
tf^
1
£
20
10
-8 -A
Boys (53) (15)*
30
20
10
t^r-H
-8-40+4
First 2 Weeks (34) (13)*
30
20
10
-8 -4
Girls (38) (18)*
30
Second 2 Weeks (20) (5)*
Slope of Regression (mUppb, 1-h O3)
Figure 9-8. Histograms of regression slopes for FEVt versus 1-h ozone
concentration in children attending a summer camp in northwestern
New Jersey. The numbers of children in each group are indicated in
parentheses; an asterisk identifies the number of children with slopes that
were significantly different from zero (p < 0.05). Shading represents the
percent of significant (p < 0.05) slopes across the distributions.
Source: Spektor et al. (1988).
9-21
-------
0.0
00
0.1 0.2 0.3
Ozone Concentration (ppm)
• Avoletal. (1984)
DKulte etal. (1985)
A McDonnell etal. (1983)
Figure 9-9. Proportion of heavily exercising individuals predicted to experience a 10%
decrement in forced expiratory volume in 1 s following a 1- or 2-h exposure to
ozone.
Source: U.S. Environmental Protection Agency (1989).
Kulleetal. (1985)
A McDonnell etal. (1983)
0.1 0.2 0.3
Ozone Concentration (ppm)
Figure 9-10. Proportion of heavily exercising individuals predicted to experience mild
cough following a 2-h ozone exposure.
Source: U.S. Environmental Protection Agency (1989).
9-22
-------
Predictions of the proportion of individuals experiencing 5, 10, or 15% FEVj
decrements as a function of C (0.0 to 0.12 ppm O3), T (1 to 6.6 h), at a specific age
(24 years) for exposures with moderate exercise are shown in Figure 9-11.
0.1 0.2
0.3 0.4 0.5 0.6 0.7
Dose (ppm x hours)
0.8 0.9 1.0
Figure 9-11. Proportion of moderately exercising individuals exposed to ozone for
6.6 h predicted to experience 5, 10, or 15% decrements in forced expiratory
volume in 1 s as a function of C x T for age = 24 years.
Source: McDonnell et al. (1995).
As an example of differences between the mean and individual responses, it was
stated earlier that exposure for 5.6 h to 0.08 ppm O3 was the shortest duration for which a 5%
mean decrement in FEVj was observed. For those same exposure conditions, 41, 17, and
10% of the subjects studied experienced FEVj decrements larger than 5, 10, and 15%,
respectively.
The clinical significance of individual responses to O3 exposure depends on the
health status of the individual, the magnitude of the changes in pulmonary function, the
severity of respiratory symptoms, and the duration of the response. Tables 9-1 and 9-2
categorize individual functional and symptomatic responses to O3 exposure as normal (or
none) and by increasing levels of severity in healthy persons and in persons with impaired
respiratory systems, respectively. Pulmonary function responses are represented in these
tables by changes in spirometry (e.g., FEVj), SRaw, and nonspecific bronchial responsiveness.
Respiratory symptom responses include cough, pain on deep inspiration, and wheeze. The
changes in spirometry that have been focused on most frequently are O3-induced decrements
in FEVj because they are easily quantified, have a continuous distribution, and have been
used to provide most of the exposure-response relationships described in this section. The
combined impact of both functional and symptomatic
9-23
-------
Table 9-1. Gradation of Individual Responses to
Short-Term Ozone Exposure in
Healthy Persons3
Functional Response
FEVj
Nonspecific
bronchial responsiveness1"
Duration of
response
Symptomatic Response
Cough
Chest pain
Duration of response
Impact of Responses
Interference with normal
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
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
activity
individuals choose
to limit activity
individuals choose
to limit activity
aSee text for discussion; see Appendix A for abbreviations and acronyms.
bAn increase in nonspecific bronchial responsiveness of 100% is equivalent to a 50% decrease in PD20 or
(see Chapter 7, Section 7.2.3).
responses to O3 exposure generally is displayed as an interference with normal activity or a
change in medical treatment (see Tables 9-1 and 9-2).
In healthy individuals, the importance attached to individual changes in FEVj and
nonspecific bronchial responsiveness depends, in part, on the magnitude and persistence of the
response, but it is also important to consider the circumstances in which changes in lung
function occur with other responses. For example, a 20% decrement in FEVj or a 100 to
200% increase in SRaw that is induced as a result of a nonspecific bronchial responsiveness
test and one that is almost completely reversible within an hour is associated with little, if
any, airway epithelial damage. If, in addition, there are no respiratory symptoms (except
chest discomfort), then this response, in itself, would not be considered clinically significant.
On the other hand, a smaller decrement in FEVj of 15%, accompanied by marked pain on
deep inspiration and persistent cough that is reversed in approximately 24 h, may be
considered clinically significant in some individuals. In other words, it is important to
9-24
-------
Table 9-2. Gradation of Individual Responses to
Short-Term Ozone Exposure in Persons with
Impaired Respiratory Systems3
Functional Response
None
Small
Moderate
Large
FEVj change
Decrements of Decrements of 3 to
Decrements of
>10% but <20%
Decrements of
>20%
Nonspecific
bronchial responsiveness1"
Within normal Increases of <100% Increases of Increases of
range <300% >300%
Airway resistance
(SRaw)
Within normal
range (±20%)
SRaw increased
<100%
SRaw increased up
to 200% or up to
15 cm H2O/s
SRaw increased
>200% or more
than 15 cm H2O/s
Duration of response
None
<4 hours
>4 hours but <24
hours
>24 hours
Symptomatic Response
Wheeze
Cough
Chest pain
Duration of response
Impact of Responses
Interference with normal
activity
Medical treatment
Normal
None
Infrequent
cough
None
None
Normal
None
No change
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
Normal medication
as needed
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
Increased
frequency of
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
Physician or
emergency room
medication use or visit
additional
medication
aSee text for discussion; see Appendix A for abbreviations and acronyms.
bAn increase in nonspecific bronchial responsiveness of 100% is equivalent to a 50% decrease in PD20 or PD100
(see Chapter 7, Section 7.2.3).
9-25
-------
consider the pattern of responses and not simply to focus on a single marker of the effect
of O3.
The magnitude of individual changes can become more important in persons with
impaired respiratory systems (e.g., asthmatics) who already have reduced baseline lung
function. Any change in function that causes these individuals to drop below 40 to 50% of
predicted would be considered clinically adverse. For example, O3-induced changes in SRaw,
a measure of airway narrowing, are small and of minimal clinical significance in
nonasthmatic individuals. Asthmatics, however, often have baseline airway narrowing and
experience larger changes in SRaw on exposure to O3 than do nonasthmatics. Because of these
baseline differences, the clinical significance of increases in SRaw depends both on percent
change from baseline and on absolute increases in SRaw.
9.4 Effects of Long-Term Ozone Exposures
In both humans and test animals, the response to a single O3 exposure nominally
can be characterized by lung dysfunction, lung cell injury and inflammation, and leakage of
plasma proteins into the airspace lumen. However, when such an exposure is repeated for
several consecutive days, many of these effects appear to wane, suggesting attenuation or the
development of tolerance to the continued intermittent challenge. In spite of this apparent
state of attenuation, long-term O3 exposures have been linked to subtle pulmonary effects,
some of which have irreversible components, thereby enhancing concern about chronic
effects. The following section will provide an overview attempting to synthesize the current
understanding of the phenomenon of attenuation during brief, repeated exposures and the
evidence for potential health impairments resulting from protracted exposures to this oxidant.
9.4.1 Repeated Exposures
It is well established that a brief exposure of laboratory rodents to an
O3 concentration, which causes minimal effects, will protect the animals from a subsequent
lethal challenge of O3 a week later. This phenomenon, called tolerance, bears a similarity to
the pattern of attenuated nonlethal effects (sometimes referred to as "adaptation") observed in
both human volunteers and animals when exposed to episodic levels of O3 (<0.5 ppm) for
1 to 7 h/day over a succession of 5 or more days. Generally, over a 5-day exposure period,
the effects of Day 1 are accentuated on Day 2 and diminish thereafter. Attenuation of the
functional effects include spirometric deficits and associated symptoms as well as irritative
alterations of breathing; nonspecific airway responsiveness, however, does not revert to
normal levels. Measures of tissue effects that attenuate include inflammation and impaired
phagocytic capabilities of alveolar macrophages. However, some evidence from animal
studies suggests that tissue alterations persist, although the observed changes may be part of a
transition to a chronically affected stage of the lung. Thus, in general, cell-associated
indicators of injury or damage within the lung appear to diminish in spite of the continued
O3 exposure.
A number of mechanisms have been shown to be involved in the evolution of this
"adapted" state. These mechanisms range from the replacement of sensitive cells in the
alveolar lining (epithelium) by more resistant cells (with or without a thickened fluid barrier
on the lumenal surface) to the enhancement of antioxidant metabolism providing cell
resistance and more biochemical defenses at the lung surface. However, controlled human
9-26
-------
studies show that after a 1-week period without O3 exposure, subjects regain their spirometric
responsiveness to O3 challenge, although this abrupt transition between unresponsiveness and
responsiveness appears less distinct in field-related studies. For example, studies of Southern
Californians suggest that they are significantly less responsive to the spirometric effects of an
acute episodic-like controlled challenge with O3 when studied for a period after the "high" O3
season than after the relatively "low" O3 season. Likewise, there is some evidence that O3-
exposed urban populations are also somewhat more resistant to the oxidant than populations
that receive minimal exposure. This would appear to be in conflict with hospital admissions
data suggesting the aggravation of respiratory diseases, like asthma, within such populations.
It remains to be shown whether these latter data reflect the responsiveness of a sensitive
subpopulation, perhaps less adapted or having less reserve function.
9.4.2 Prolonged Exposures
Most long-term exposure studies in animals have evaluated structural and
functional changes. In the few investigations of the immune system or antibacterial host
defenses, prolonged exposures of animals either caused no effects or did not increase the
magnitude of effects observed after acute exposures. Thus, the following discussion centers
on the larger body of knowledge on other endpoints.
Epidemiologic studies attempting to associate chronic lung effects in humans with
long-term O3 exposure provide only suggestive evidence that such a linkage exists. Most
studies have been cross-sectional in design and have been compromised by incomplete control
of confounding variables and inadequate exposure information. Other studies have attempted
to follow variably exposed groups prospectively. Studies of such design have been conducted
in communities of the Southwest Air Basin as well as in Canada where comparisons could be
drawn between lung function changes over several years in populations from high- or low-
oxidant pollution. The findings suggest small, but consistent decrements in lung function
among inhabitants of the more highly polluted communities. However, associations between
O3 and other copollutants and, in some cases, problems with study population loss undermine
the confidence in the study conclusions. Likewise, recent associations found between O3 and
the incidence and severity of asthma over a decade of study, although derived from well-
designed studies, also tend to be weakened by the colinearity of O3 with other air pollutants.
Nevertheless, in all of the studies assessing lung function, the pattern of dysfunction
associated with the long-term exposure has been consistent with the functional and structural
abnormalities seen in laboratory animal studies.
The advantage of laboratory animal studies is the ability to examine closely the
distribution and intensity of the O3-induced morphologic changes that have been identified
throughout the respiratory tract (see Chapter 6, Section 6.2.4). Indeed, cells of the nose, like
the distal lung, clearly are affected by O3. Perhaps of greater health concern are the lesions
that occur in the small airways and in the centriacinar regions of the lung where the alveoli
meet the distal airways (Figure 9-12). Altered function of the distal airways, the proximal
conduits of air to the gas-exchange regions, can result in reduced communication of fresh air
with the alveoli and air-trapping. In fact, chronic O3 lesions as found in animal studies are
reminiscent of the earliest lesions found in respiratory bronchiolitis, some of which may
progress to fibrotic lung disease (Kuhn et al., 1989; King, 1993).
9-27
-------
Air
Ozone
Figure 9-12. A summary of the morphologic lesions found in the terminal bronchioles and
centriacinar region (CAR) of the lung following exposure of laboratory rats
to filtered air or a simulated ambient pattern of ozone for up to 78 weeks. In
the terminal bronchiole, the sizes of the dome of Clara cells became smaller
with ozone exposure, and the number of cilia is reduced (arrows). In the
CAR, the epithelium becomes thicker, and accumulation of collagen fibers
occurs (arrow heads).
Source: Chang et al. (1992).
9-28
-------
The O3-induced inflammation, cell damage, and altered permeability of epithelial
cells lining the respiratory tract allow exudation of fluid, cells, and cellular debris from
plasma into lung tissue. The magnitude of intraluminal exudation associated with injury
correlates with the initial epithelial necrosis and release of inflammatory mediators.
As shown in Figure 9-13, the temporal pattern of effects during and after a chronic exposure
is complex. During the early days of exposure, the end-airway lumenal and interstitial
inflammation peaks, and, thereafter, appears to subside at a lower plateau of activity
sometimes referred to as a "smoldering" lesion. Several cytokines remain elevated beyond the
apparent adaptation phase of the response and may be linked conceptually to the development
of chronic lesions in the distal lung. To date, however, a clear association of these BAL-
derived mediators and cells with long-term toxicity has not been demonstrated. Some
evidence of molecular changes within the matrix of the lung may also link to the chronic
effects, but these too remain poorly defined. When exposures to O3 continue for weeks or
months, the diminished O3-induced exudative response in the distal bronchoalveolar areas is
supplanted by hyperplastic epithelia in the alveoli and end-airways. Damaged cells in
centriacinar alveoli are replaced by metabolically active progenitor cells that are more
resistant to oxidant challenge. Junctional areas between conducting and gas-exchange regions,
where the O3 changes are typically most intense, also undergo epithelial hyperplasia, giving
the appearance that airway cells are extending into the mouth of the alveolus, hence the term
"bronchiolization". The functional result of this concentration-dependent process is the
effective elongation of distal bronchioles, which functionally may alter air distribution within
the lung during breathing. These hyperplastic cells also are believed to be more resistant to
O3. When exposure to O3 ceases, most, but not all, of the hyperplasia appears to reverse with
time.
In contrast, within the underlying interstitium (tissue between blood and air spaces)
of the affected centriacinar region, proliferating fibroblasts appear to evolve excess
noncellular fibrous matrices, which may be only partially reversible and may, in fact, progress
after removal from O3 exposure. This would suggest that O3 can initiate focal interstitial
fibrosis of the lung at the regions where O3 causes epithelial cell damage as a prelude to
chronic degenerative lung disease. The crucial question, then, is whether this latter
irreversible process, which clearly occurs at relatively high O3 concentrations, occurs at
ambient levels to which humans are typically exposed, in many cases, over a lifetime.
Unfortunately, comparable morphologic data from humans residing in O3-polluted areas are
lacking.
Studies of prolonged O3 exposures in monkeys and rats reveal generally similar
morphologic responses, although it appears that the monkey exhibits somewhat more tissue
injury than does the rat under roughly similar exposure conditions. Interspecies comparisons
of dosimetric data indicate that the monkey, with its similarity to the human in distal airway
structure, provides data that may best reflect the potential effects of O3 in humans exercising
out of doors. As such, monkeys exposed to O3 at 0.15 ppm for 8 h each day for 6 to 90 days
exhibit significant distal airway remodeling. Rats show similar but more modest changes at
0.25 ppm O3 after exposures of longer duration, up to 18 mo and beyond (near-lifetime). The
chronic distal lung and airway alterations appear consistent with incipient peribronchiolar
fibrogenesis within the interstitium. Attempts to correlate functional deficits have been
variable, perhaps due in part to the degree and distribution of the lesions and the general
insensitivity of most measures of the distal lung function. The interstitial changes may
progress, however. Moreover, one recent primate study revealed evidence that
9-29
-------
CD
CO
I
CO
2
"o
0
T3
Epithelial hyperplasia
Bronchoalveolar exudate
Interstitial fibrosis
Exposure
Post-
exposure
days
Time
Figure 9-13. Schematic comparison of the duration-response profiles for epithelial
hyperplasia, bronchoalveolar exudation, and Interstitial fibrosis In the
centrlaclnar region of lung exposed to a constant low concentration of ozone.
Source: Dungworth (1989).
intermittent challenge with a pattern of O3 exposure more reflective of seasonal episodes, with
extended periods of clean air in between extended periods of O3, actually leads to greater
injury. The reasons for this are unclear, but may relate to the known loss of tolerance that
occurs in both humans and animal test species with removal of the oxidant burden.
In conclusion, the collective toxicologic data on chronic exposure to O3 garnered in
animal exposure and human population studies have some ambiguities. What is clear is that
the distribution of the O3 lesions is roughly similar across species, is, in part, concentration
dependent (and perhaps time or exposure-pattern dependent), and, under certain conditions,
has irreversible structural attributes. What is unclear is whether ambient exposure scenarios
encountered by humans result in similar lesions and whether there are resultant functional or
impaired health outcomes, particularly because the human exposure scenario may involve
much longer exposures than can be studied in the laboratory. The epidemiologic lung
function data generally parallel those of the animal studies, but they lack the confidence of
O3 exposure history and are frequently confounded by personal or copollutant variables.
9-30
-------
9.4.3 Genotoxicity and Carcinogenicity of Ozone
Numerous in vitro exposure studies suggest that O3 has either weak or no potential
to cause mutagenic, cytogenetic, or cellular transformation effects. Most of these experiments
utilized high concentrations of O3 (>5.0 ppm). Because of the exposure systems used, there
are unknowns about the formation of artifacts and the dose of O3. Therefore, these studies
are not very useful in health assessment. Cytogenetic effects have been observed in some,
but not all, laboratory animal and human studies of short-term O3 exposure. However, well-
designed human clinical cytogenetic studies were negative.
Until recently, in vivo exposure studies of carcinogenicity, with and without
co-exposure to known carcinogens, were either negative or ambiguous. A well-designed
cancer bioassay study has recently been completed by the National Toxicology Program
(NTP) using male and female Fischer 344/N rats and B6C3FJ mice. Animals were exposed
for 2 years to 0.12, 0.5, and 1.0 ppm O3 (6 h/day, 5 days/week). A similar lifetime exposure
was conducted, but 0.12 ppm was not used. The NTP evaluated the weight-of-evidence for
this study; they found "no evidence" of carcinogenicity in rats but reported "equivocal
evidence" of carcinogenicity in O3-exposed male mice and "some evidence" of carcinogenic
activity in O3-exposed female mice. The increases in adenomas and carcinomas were
observed only in the lungs. There was no concentration response. In the male mice, the
incidence of neoplasms in the 2-year study was not elevated significantly by O3 and was
within the range of historical controls. The lifetime exposure resulted in an increased
incidence of carcinomas that was not statistically significant. When the female mouse data
from the two exposure regimens (at 1.0 ppm) were combined, there was a statistically
significant increase (almost double) in neoplasms. In a companion study, male rats were
treated with a tobacco carcinogen and exposed for 2 years to 0.5 ppm O3. Ozone did not
affect the response and, therefore, had no tumor promoting activity.
In summary, only chronic exposure to a high concentration of O3 (1.0 ppm) has
been shown to evoke a limited degree of carcinogenic activity in the females of one strain of
mice. Rats were not affected. Furthermore, there was no concentration response, and there is
inadequate information from other research to provide mechanistic support for the finding in
mice. Thus, the potential for animal carcinogenicity is uncertain.
9.5 Effects of Combined Pollutant Exposures
In ambient air, people are exposed to mixtures of pollutants, making it important to
understand interactions. Epidemiological studies, which inherently evaluate O3 as part of
complex mixtures, are discussed in other subsections dealing with classes of effects. In the
laboratory it becomes possible to sort out the role of O3 in simple mixtures. Complex
mixtures are typically not investigated in the laboratory because, even if only six pollutants
were involved, the experimental design required to unequivocally sort out which pollutant or
pollutant interactions were responsible for the responses or portions of the responses could
require as many as 719 additional separate experiments, even if the concentrations of the six
pollutants remained the same.
The summary will focus only on binary mixtures because these are by far the
predominant type of experiments. Responses to a binary pollutant mixture may represent the
sum of the independent responses to the two chemicals (i.e., an additive response). If there is
some interaction between either the two responses or the two pollutants, the resultant response
9-31
-------
could be larger than additive (synergism) or smaller than additive (antagonism). Interaction
between pollutants could result in the production of a more or less toxic byproduct.
Alternatively, the response to one pollutant could magnify the response to the other pollutant
or could interfere with or block the action of the other pollutant. Binary mixture studies fall
into two categories, simultaneous and sequential exposures. In the simultaneous exposures,
both the responses and the pollutants can interact. In the sequential exposures, it is primarily
the responses that would interact.
In general, controlled human studies of O3 mixed with other pollutants show no
more than an additive response with symptoms or spirometry as an endpoint. This applies to
O3 in combination with nitrogen dioxide (NO2), SO2, sulfuric acid (H2SO4), nitric acid
(HNO3), or carbon monoxide (CO). Indeed, at the levels of copollutants used in human
exposure studies, the responses can be attributed primarily to O3. In one study, exposure to
O3 increased airway responsiveness to SO2 in asthmatics. Similarly, other pollutants that may
increase airway responsiveness could augment the effect of O3 on airway responsiveness.
The relatively large number of animal studies of O3 in mixture with NO2 and
H2SO4 shows that additivity, synergism, and antagonism can result, depending on the
exposure regimen and the endpoint studied. The numerous observations of synergism are of
concern, but the interpretation of most of these studies relative to the real world is
confounded by unrealistic exposure designs. For example, ambient concentrations of O3 often
were combined with levels of copollutants substantially higher than ambient, creating the
possibility that mechanisms of toxicity unlikely in the real world contributed to the
experimental outcome. Nevertheless, the data support a hypothesis that coexposure to
pollutants, each at innocuous or low-effect levels, may result in effects of significance.
9.6 Conclusions
This section summarizes the primary conclusions derived from an integration of
the known effects of O3 provided by animal lexicological, human clinical, and
epidemiological studies.
1. What are the effects of short-term (<8-h) exposures to ozone?
Recent epidemiology studies addressing the effects of short-term ambient exposure
to O3 in the population have yielded significant associations with a wide range of health
outcomes, including lung function decrements, aggravation of preexisting respiratory disease,
increases in daily hospital admissions and emergency department visits for respiratory causes,
and increased mortality. Results from lung function epidemiology studies are generally
consistent with the experimental studies in laboratory animals and humans.
Short-term O3 exposure of laboratory animals and humans causes changes in
pulmonary function, including tachypnea (rapid, shallow breathing), decreased lung volumes
and flows, and increased airway responsiveness to nonspecific stimuli. Increased airway
resistance occurs in both humans and laboratory animals, but typically at higher exposure
levels than other functional endpoints. In addition, adult human subjects experience
O3 induced symptoms of airway irritation such as cough or pain on deep inspiration. The
changes in pulmonary function and respiratory symptoms occur as a function of exposure
concentration, duration, and level of exercise. Adult human subjects with mild asthma have
qualitatively similar responses in lung volume and airway responsiveness to
9-32
-------
bronchoconstrictor drugs as nonasthmatics. Respiratory symptoms are also similar, but
wheezing is a prevalent symptom in O3-exposed asthmatics in addition to the other
demonstrated symptoms of airway irritation. Airway resistance, however, increases relatively
more in asthmatics from an already higher baseline. Recovery from the effects of O3 on
pulmonary function and symptoms is usually complete within 24 h of the end of exposure,
although other responses may persist somewhat longer.
• An association between daily mortality and O3 concentration for areas with high
O3 levels (e.g., Los Angeles) has been suggested, although the magnitude of
such an effect is unclear.
• Increased O3 levels are associated with increased hospital admissions and
emergency department visits for respiratory causes. Analyses from data in the
Northeastern United States suggest that O3 air pollution is associated with a
substantial portion (on the order of 10 to 20%) of all summertime respiratory
hospital visits and admissions.
• Pulmonary function in children at summer camps in southern Ontario, Canada,
in the northeastern United States, and in Southern California is associated with
O3 concentration. Meta-analysis indicates that a 0.50-mL decrease in FEVj is
associated with a 1 ppb increase in O3 concentration. For preadolescent
children exposed to 120 ppb (0.12 ppm) ambient O3, this amounts to an average
decrement of 2.4 to 3.0% in FEVj. Similar responses are reported for children
and adolescents exposed to O3 in ambient air or O3 in purified air for 1 to 2 h
while exercising.
• Pulmonary function decrements are generally observed in healthy subjects (8 to
45 years of age) after 1 to 3 h of exposure as a function of the level of exercise
performed and the O3 concentration inhaled during the exposure. Group mean
data from numerous controlled human exposure and field studies indicate that,
in general, statistically significant pulmonary function decrements beyond the
range of normal measurement variability (e.g., 3 to 5% for FEVj) occur
(1) at >0.50 ppm O3 when at rest,
(2) at >0.37 ppm O3 with light exercise (slow walking),
(3) at >0.30 ppm O3 with moderate exercise (brisk walking),
(4) at >0.18 ppm O3 with heavy exercise (easy jogging), and
(5) at >0.16 ppm O3 with very heavy exercise (running).
Smaller group mean changes (e.g., <5%) in FEVj have been observed at lower
O3 concentrations than those listed above. For example, FEVj decrements have
been shown to occur with very heavy exercise in healthy adults at 0.15 to 0.16
ppm O3, and such effects may occur in healthy young adults at levels as low as
0.12 ppm. Also, pulmonary function decrements have been observed in
children and adolescents at concentrations of 0.12 and 0.14 ppm O3 with heavy
exercise. Some individuals within a study may experience FEVj decrements in
excess of 15% under these exposure conditions, even when the group mean
decrement is less than 5%.
• For exposures of healthy subjects performing moderate exercise during longer
duration exposures (6 to 8 h), 5% group mean decrements in FEVj were
observed at
(1) 0.08 ppm O3 after 5.6 h,
(2) 0.10 ppm O3 after 4.6 h, and
9-33
-------
(3) 0.12 ppm O3 after 3 h.
For these same subjects, 10% group mean FEVj decrements were observed at
0.12 ppm O3 after 5.6 and 6.6 h. As in the shorter duration studies, some
individuals experience changes larger than those represented by the group mean
changes.
• An increase in the incidence of cough has been reported at O3 concentrations as
low as 0.12 ppm in healthy adults during 1 to 3 h of exposure with very heavy
exercise. Other respiratory symptoms, such as pain on deep inspiration,
shortness of breath, and lower respiratory scores (a combination of several
symptoms), have been observed at 0.16 to 0.18 ppm O3 with heavy and very
heavy exercise. Respiratory symptoms also have been observed following
exposure to 0.08, 0.10, and 0.12 ppm O3 for 6.6 h with moderate levels of
exercise.
• Increases in nonspecific airway responsiveness in healthy adults have been
observed after 1 to 3 h of exposure to 0.40 ppm, but not 0.20 ppm, O3 at rest
and have been observed at concentrations as low as 0.18 ppm, but not to
0.12 ppm, O3 during exposure with very heavy exercise. Increases in
nonspecific airway responsiveness during 6.6-h exposures with moderate levels
of exercise have been observed at 0.08, 0.10, and 0.12 ppm O3.
Short-term O3 exposure of laboratory animals and humans disrupts the barrier
function of the lung epithelium, permitting materials in the airspaces to enter lung tissue,
allowing cells and serum proteins to enter the airspaces (inflammation), and setting off a
cascade of responses.
• Increased levels of PMNs and protein in lung lavage fluid have been observed
following exposure of healthy adults to 0.20, 0.30, and 0.40 ppm with very
heavy exercise and have not been studied at lower concentrations for 1- to 3-h
exposures. Increases in lung lavage protein and PMNs also have been observed
at 0.08 and 0.10 ppm O3 during 6.6-h exposures with moderate exercise; lower
concentrations have not been tested.
Short-term O3 exposure of laboratory animals and humans impairs alveolar
macrophage clearance of viable and nonviable particles from the lungs and decreases the
effectiveness of host defenses against bacterial lung infections in animals and perhaps
humans. The ability of alveolar macrophages to engulf microorganisms is decreased in
humans exposed to 0.08 and 0.10 ppm O3 for 6.6 h with moderate exercise.
2. What are the effects of repeated, short-term exposures to ozone?
During repeated short-term exposures, some of the O3-induced responses are
partially or completely attenuated. Over a 5-day exposure, pulmonary function changes are
typically greatest on the second day, but return to control levels by the fifth day of exposure.
Most of the inflammatory markers (e.g., PMN influx) also attenuate by the fifth day of
exposure, but markers of cell damage (e.g., lactate dehydrogenase enzyme activity) do not
attenuate and continue to increase. Attenuation of lung function decrements is reversed
following 7 to 10 days without O3. Some inflammatory markers are also reversed during this
time period, but others still show attenuation even after 20 days without O3. The mechanisms
and impacts involved in attenuation are not known, although animal studies show that the
underlying cell damage continues throughout the attenuation process. In addition, attenuation
may alter the normal distribution of O3 within the lung, allowing more O3 to reach sensitive
9-34
-------
regions, possibly affecting normal lung defenses (e.g., PMN influx in response to inhaled
mi croorgani sm s).
3. What are the effects of long-term exposures to ozone?
Available data indicate that exposure to O3 for months and years causes structural
changes in several regions of the respiratory tract, but effects may be of the greatest
importance in the centriacinar regions (where the alveoli and conducting airways meet); this
region typically is affected in most chronic airway diseases of the human lung. This
information on O3 effects in the distal lung is extrapolated from animal toxicological studies
because, to date, comparable data are not available from humans. The apparent lack of
reversal of effects during periods of clean air exposure raises concern that seasonal exposures
may have a cumulative impact over many years. The role of adaptive processes in this
response is unknown but may be critically dependent on the temporal frequency or profile of
exposure. Furthermore, the interspecies diversity in apparent sensitivity to the chronic effects
of O3 is notable, with the rat representing the lower limit of response, and the monkey the
upper limit. Epidemiological studies attempting to associate chronic health effects in humans
with long-term O3 exposure provide only suggestive evidence that such a linkage exists.
Long-term exposure in the females of one strain of mice to high O3 levels (1 ppm)
caused a small, but statistically significant increase in lung tumors. There was no
concentration-response relationship, and rats were not affected. Genotoxicity data are either
negative or weak. Given the nature of the database, potential carcinogenicity in animals is
uncertain. Ozone did not show tumor-promoting activity in a chronic rat study (at 0.5 ppm
03).
4. What are the effects of binary pollutant mixtures containing ozone?
Combined data from laboratory animal and controlled human exposure studies of
O3 support the hypothesis that coexposure to pollutants, each at low-effect levels, may result
in effects of significance. The data from human studies of O3 in combination with NO2, SO2,
H2SO4, HNO3, or CO show no more than an additive response on lung spirometry or
respiratory symptoms. The larger number of laboratory animal studies with O3 in mixture
with NO2 and H2SO4 show that effects can be additive, synergistic, or even antagonistic,
depending on the exposure regimen and the endpoint studied. This issue of exposure to
copollutants remains poorly understood, especially with regard to potential chronic effects.
5. What population groups are at risk as a result of exposure to ozone?
Identification of population groups that may show increased sensitivity to O3 is
based on their (1) biological responses to O3, (2) preexisting lung disease (e.g., asthma),
(3) activity patterns, (4) personal exposure history, and (5) personal factors (e.g., age,
nutritional status).
The predominant information on the health effects of O3 noted above comes from
clinical and field studies on healthy, nonsmoking, exercising subjects, 8 to 45 years of age.
These studies demonstrate that, among this group, there is a large variation in sensitivity and
responsiveness to O3, with at least a 10-fold difference between the most and least responsive
individuals. Individual sensitivity to O3 also may vary throughout the year, related to
seasonal variations in ambient O3 exposure. The specific factors that contribute to this large
intersubject variability, however, remain undefined. Although differences may be due to the
9-35
-------
dosimetry of O3 in the respiratory tract, available data show little difference on O3 deposition
in the lungs for inhalation through the nose or mouth.
Daily life studies reporting an exacerbation of asthma and decrease in peak
expiratory flow rates, particularly in asthmatic children, appear to support the controlled
studies; however, those studies may be confounded by temperature, particle or aeroallergen
exposure, and asthma severity of the subjects or their medication use. In addition, field
studies of summertime daily hospital admissions for respiratory causes show a consistent
relationship between asthma and ambient levels of O3 in various locations in the Northeastern
United States, even after controlling for independent contributing factors. Controlled studies
on mild asthmatics suggest that they have similar lung volume responses but greater airway
resistance changes to O3 than nonasthmatics. Furthermore, limited data from studies of
moderate asthmatics suggest that this group may have greater lung volume responses than
nonasthmatics.
Other population groups with preexisting limitations in pulmonary function and
exercise capacity (e.g., chronic obstructive pulmonary disease, chronic bronchitis, ischemic
heart disease) would be of primary concern in evaluating the health effects of O3.
Unfortunately, not enough is known about the responses of these individuals to make
definitive conclusions regarding their relative responsiveness to O3. Indeed, functional effects
in these individuals with reduced lung function may have greater clinical significance than
comparable changes in healthy individuals.
Currently available data on personal factors or personal exposure history known or
suspected of influencing responses to O3 follow.
• Human studies have identified a decrease in pulmonary function responsiveness
to O3 with increasing age, although symptom rates remain similar.
Toxicological studies are not easily interpreted but suggest that young animals
are not more responsive than adults.
• 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.
• 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.
• Information derived from O3 exposure of smokers is limited. The general trend
is that smokers are less responsive than nonsmokers. This reduced
responsiveness may wane after smoking cessation.
• Although nutritional status (e.g., vitamin E deficiency) makes laboratory rats
more susceptible to O3-induced effects, 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 in
O3 responsiveness, especially their deficiency, has not been well studied.
Based on information presented in this document, the population groups that have
demonstrated increased responsiveness to ambient concentrations of O3 consist of exercising,
healthy and asthmatic individuals, including children, adolescents, and adults.
9-36
-------
References
Adams, W. C.; Savin, W. M; Christo, A. E. (1981) Detection of ozone toxicity during continuous exercise
via the effective dose concept. J. Appl. Physiol.: Respir. Environ. Exercise Physiol. 51: 415-422.
Avol, E. L.; Linn, W. S.; Venet, T. G.; Shamoo, D. A.; Hackney, J. D. (1984) Comparative respiratory effects of
ozone and ambient oxidant pollution exposure during heavy exercise. J. Air Pollut. Control Assoc.
34: 804-809.
Burnett, R. T.; Dales, R. E.; Raizenne, M. E.; Krewski, D.; Summers, P. W.; Roberts, G. R.; Raad-Young, M.;
Dann, T.; Brook, J. (1994) Effects of low ambient levels of ozone and sulfates on the frequency
of respiratory admissions to Ontario hospitals. Environ. Res. 65: 172-194.
Carr, W.; Zeitel, L.; Weiss, K. (1992) Variations in asthma hospitalizations and deaths in New York City.
Am. J. Public Health 82: 59-65.
Chang, L.-Y.; Huang, Y.; Stockstill, B. L.; Graham, J. A.; Grose, E. C.; Menache, M. G.; Miller, F. J.; Costa, D.
L.; Crapo, J. D. (1992) Epithelial injury and interstitial fibrosis in the proximal alveolar regions of
rats chronically exposed to a simulated pattern of urban ambient ozone. Toxicol. Appl. Pharmacol.
115: 241-252.
Cherniak, R. M.; Banks, D. E.; Bell, D. Y.; Davis, G. S.; Hughes, J. M.; King, T. E., Jr. (1990) Bronchoalveolar
lavage constituents in healthy individuals, idiopathic pulmonary fibrosis, and selected comparison
groups. Am. Rev. Respir. Dis. 141(suppl.): S169-S202.
Dungworth, D. L. (1989) Noncarcinogenic responses of the respiratory tract to inhaled toxicants.
In: McClellan, R. O.; Henderson, R. F., eds. Concepts in inhalation toxicology. New York, NY:
Hemisphere Publishing Corp.; pp. 273-298.
Evans, R., Ill; Mullally, D. L; Wilson, R. W.; Gergen, P. J.; Rosenberg, H. M.; Grauman, J. S.;
Chevarley, F. M.; Feinleib, M. (1987) National trends in the morbidity and mortality of asthma
in the US. Prevalence, hospitalization and death from asthma over two decades: 1965-1984. Chest
91(suppl.): 65S-74S.
Folinsbee, L. J.; Drinkwater, B. L.; Bedi, J. F.; Horvath, S. M. (1978) The influence of exercise
on the pulmonary function changes due to exposure to low concentrations of ozone. In: Folinsbee,
L. J.; Wagner, J. A.; Borgia, J. F.; Drinkwater, B. L.; Gliner, J. A.; Bedi, J. F., eds. Environmental
stress: individual human adaptations. New York, NY: Academic Press; pp. 125-145.
Hazucha, M. J. (1987) Relationship between ozone exposure and pulmonary function changes. J. Appl. Physiol.
62: 1671-1680.
Highfill, J. W.; Costa, D. L. (1995) Statistical response models for ozone exposure: their generality when applied
to human spirometric and animal permeability functions of the lung. J. Air Waste Manage. Assoc.
45: 95-102.
Highfill, J. W.; Hatch, G. E.; Slade, R.; Crissman, K. M.; Norwood, J.; Devlin, R. B.; Costa, D. L. (1992)
Concentration-time models for the effects of ozone on bronchoalveolar lavage fluid protein from
rats and guinea pigs. Inhalation Toxicol. 4: 1-16.
King, T. E., Jr. (1993) Bronchiolitis. In: Schwarz, M. L; King, T. E., Jr., eds. Interstitial lung disease. 2nd ed.
St. Louis, MO: Mosby Year Book; pp. 463-495.
9-37
-------
Kuhn, C, III; Boldt, J.; King, T. K, Jr.; Crouch, E.; Vartio, T.; McDonald, J. A. (1989)
An immunohistochemical study of architectural remodeling and connective tissue synthesis in
pulmonary fibrosis. Am. Rev. Respir. Dis. 140: 1693-1703.
Kulle, T. J.; Sauder, L. R.; Hebel, J. R.; Chatham, M. D. (1985) Ozone response relationships in healthy
nonsmokers. Am. Rev. Respir. Dis. 132: 36-41.
McDonnell, W. F.; Smith, M. V. (1994) Description of acute ozone response as a function of exposure rate
and total inhaled dose. J. Appl. Physiol. 76: 2776-2784.
McDonnell, W. F.; Horstman, D. H.; Hazucha, M. J.; Seal, E., Jr.; Haak, E. D.; Salaam, S. A.;
House, D. E. (1983) Pulmonary effects of ozone exposure during exercise: dose-response
characteristics. J. Appl. Physiol.: Respir. Environ. Exercise Physiol. 54: 1345-1352.
McDonnell, W. F., Ill; Chapman, R. S.; Leigh, M. W.; Strope, G. L.; Collier, A. M. (1985) Respiratory
responses of vigorously exercising children to 0.12 ppm ozone exposure. Am. Rev. Respir.
Dis. 132: 875-879.
McDonnell, W. F.; Muller, K. E.; Bromberg, P. A.; Shy, C. M. (1993) Predictors of individual differences
in acute response to ozone exposure. Am. Rev. Respir. Dis. 147: 818-825.
McDonnell, W. F.; Andreoni, S.; Smith, M. V. (1995) Proportion of moderately exercising individuals
responding to low-level, multi-hour ozone exposure. Am. J. Respir. Crit. Care Med. 152: 589-596.
National Institutes of Health. (1991) Guidelines for the diagnosis and management of asthma. Bethesda,
MD: U.S. Department of Health and Human Services, National Heart, Lung, and Blood Institute,
National Asthma Education Program; publication no. 91-3042.
Rombout, P. J. A.; Van Bree, L.; Heisterkamp, S. H.; Marra, M. (1989) The need for an eight hour ozone
standard. In: Schneider, T.; Lee, S. D.; Wolters, G. J. R.; Grant, L. D., eds. Atmospheric ozone
research and its policy implications: proceedings of the 3rd US-Dutch international symposium;
May 1988; Nijmegen, The Netherlands. Amsterdam, The Netherlands: Elsevier Science Publishers;
pp. 701-710. (Studies in environmental science 35).
Schwartz, J.; Gold, D.; Dockery, D. W.; Weiss, S. T.; Speizer, F. E. (1990) Predictors of asthma and persistent
wheeze in a national sample of children in the United States: association with social class, perinatal
events, and race. Am. Rev. Respir. Dis. 142: 555-562.
Seal, E., Jr.; McDonnell, W. F.; House, D. E.; Salaam, S. A.; Dewitt, P. J.; Butler, S. O.; Green, J.;
Raggio, L.(1993) The pulmonary response of white and black adults to six concentrations of ozone.
Am. Rev. Respir. Dis. 147: 804-810.
Sly, M. D. (1988) Mortality from asthma in children 1979-1984. Ann. Allergy 60: 433-443.
Smith, D. L.; Deshazo, R. D. (1993) Bronchoalveolar lavage in asthma: an update and perspective. Am. Rev.
Respir. Dis. 148: 523-532.
Spektor, D. M.; Lippmann, M.; Lioy, P. J.; Thurston, G. D.; Citak, K.; James, D. J.; Bock, N.; Speizer, F. E.;
Hayes, C. (1988) Effects of ambient ozone on respiratory function in active, normal children.
Am. Rev. Respir. Dis. 137: 313-320.
Stanley, M. W. (1991) Diagnosis of pulmonary infections. In: Stanley, M. W.; Henry-Stanley, M. J.; Iber, C.
Bronchoalveolar lavage: cytology and clinical applications. New York, NY: Igaku-Shoin;
pp. 65-111.
9-38
-------
Tapper, J. S.; Costa, D. L.; Doerfler, D.; Lehmann, J. R.; Stevens, M. A.; Madden, M. C. (1994) The relative
contribution of concentration and duration to ozone (O3)-induced lung injury. Fundam. Appl.
Toxicol.: submitted.
Thurston, G.D.; Ito, K.; Hayes, C.G.; Bates, D.V.; Lippmann, M. (1994) Respiratory hospital admissions and
summertime haze air pollution in Toronto, Ontario: consideration of the role of acid aerosols.
Environ. Res. 65: 271-290.
U.S. Environmental Protection Agency. (1978) Air quality criteria for ozone and other photochemical oxidants.
Research Triangle Park, NC: Office of Health and Environmental Assessment, Environmental
Criteria and Assessment Office; EPA report no. EPA-600/8-78-004. Available from: NTIS,
Springfield, VA; PB80-124753.
U.S. Environmental Protection Agency. (1986) Air quality criteria for ozone and other photochemical oxidants.
Research Triangle Park, NC: Office of Health and Environmental Assessment, Environmental
Criteria and Assessment Office; EPA report nos. EPA-600/8-84-020aF-eF. 5v. Available from:
NTIS, Springfield, VA; PB87-142949.
U.S. Environmental Protection Agency. (1989) Review of the national ambient air quality standards for ozone:
assessment of scientific and technical information. OAQPS staff paper. Research Triangle Park, NC:
Office of Air Quality Planning and Standards; EPA report no. EPA-450/2-92/001. Available from:
NTIS, Springfield, VA; PB92-190446.
Weiss, K. B.; Wagener, D. K. (1990) Changing patterns of asthma mortality: identifying target populations
at high risk. JAMA J. Am. Med. Assoc. 264: 1683-1687.
9-39
-------
Appendix A
Abbreviations and Acronyms
ADOM
AGL
AIRS
AM
AQCD
AQCR
AUSPEX
C
C
CA
CAA
CAAA
CAL-RAMS
CAR
CASAC
CBM
CCM
CFC
CH3OH
CH4
CI
CIT
CL
CMB
CNG
CO
C02
CTWM
DIAL
Acid Deposition and Oxidant Model
Above ground level
Aerometric Information Retrieval System
Alveolar macrophage
Air Quality Criteria Document
Air Quality Control Region
Atmospheric Utility Signatures, Predictions, and Experiments
Carbon
Concentration
Chromotropic acid
Clean Air Act
Clean Air Act Amendments of 1990
Coast and Lake Regional Atmospheric Modeling System
Centriacinar region
Clean Air Scientific Advisory Committee
Carbon-bond mechanism
Community Climate Model
Chlorofluorocarbon
Methanol
Methane
Chemical ionization
California Institute of Technology
Chemiluminescence
Chemical mass balance
Compressed natural gas
Carbon monoxide
Carbon dioxide
Complex Terrain Wind Model
Differential absorption lidar
A-1
-------
DNPH
DOAS
DWM
BCD
EKMA
EMS
EPA
EPEM
EPRI
EPS
ERAQS
ETBE
EtOH
FDDA
FeSO4
FEV,
FVC
FID
FTIR
GC
GMEP
GPT
H+
HC
HCFC
HCHO
HNO2
HNO3
HO2
H202
HPLC
H2S04
1C
ID
I/O
2,4-Dinitrophenylhydrazine
Differential optical absorption spectrometry
Diagnostic Wind Model
Electron capture detection
Empirical Kinetic Modeling Approach
Emissions Modeling System
U.S. Environmental Protection Agency
Event Probability Exposure Model
Electric Power Research Institute
Emissions Preprocessor System
Eastern Regional Air Quality Study
Ethyl-tertiary-butyl ether
Ethanol
Four-dimensional data assimilation
Ferrous sulfate
Forced expiratory volume in 1 s
Forced vital capacity
Flame ionization detection
Fourier transform infrared absorption spectroscopy
Gas chromatography
Geocoded Model of Emissions and Projections
Gas-phase titration
Hydrogen ion
Hydrocarbon
Hydrochlorofluorocarbon
Formaldehyde
Nitrous acid
Nitric acid
Hydroperoxyl
Hydrogen peroxide
High-performance liquid chromatography
Sulfuric acid
Ion chromatography
Identification (number)
Indoor/outdoor
A-2
-------
IR
Infrared radiation
IR
LMOS
LPG
MBTH
MCCP
MM4/MM5
MOBILE
MODELS 3
MPAN
MSA
MSCET
MTBE
NA
NAAQS
NADP
NAMS
NAPAP
NAPBN
NAS
NBKI
NBS
NCAR
NCLAN
NDDN
NEM
NF
NH3
NH4HSO4
NH4OH
(NH4)2S04
Incremental reactivity
Lake Michigan Oxidant Study
Liquified petroleum gas
3 -Methyl-2-benzothiazolone hydr azone
Mountain Cloud Chemistry Program
Mesoscale Model, versions 4 and 5
U.S. Environmental Protection Agency emissions model for mobile
sources
Modeling framework that consolidates all of the U.S. Environmental
Protection Agency's three-dimensional photochemical air
quality models
Peroxymethacryloyl nitrate
Metropolitan Statistical Area
Month and state current emissions trends
Methyl-tertiary-butyl ether
Not available
National Ambient Air Quality Standards
National Atmospheric Deposition Program
National Air Monitoring Station
National Acid Precipitation Assessment Program
Western National Air Pollution Background Network
National Academy of Sciences
Neutral buffered potassium iodide
National Bureau of Standards; now National Institute of
Standards and Technology
National Center for Atmospheric Research
National Crop Loss Assessment Network
National Dry Deposition Network
National Air Quality Standards Exposure Model
National forest
Ammonia
Ammonium bisulfate
Ammonium hydroxide
Ammonium sulfate
A-3
-------
NIST
NM
NMHC
NMOC
NO
NO2
N2O
NO3
NOX
NP
NPN
NTP
03
OAQPS
Obs.
OH
OHBA
PAMS
PAN
PANs
PAR
PEL
PBzN
PDFID
PF/TPLIF
pH
PL
PLANR
PMN
ppmC
PPN
PSD
PVOC
QE
QH
r
National Institute of Standards and Technology
National monument
Nonmethane hydrocarbon
Nonmethane organic compound
Nitric oxide
Nitrogen dioxide
Nitrous oxide
Nitrate
Nitrogen oxides
National park
n-propyl nitrate
National Toxicology Program
Ozone
Office of Air Quality Planning and Standards
Observations
Hydroxyl
Hydroxybenzoic acid
Photochemical Aerometric Monitoring System
Peroxyacetyl nitrate
Peroxyacyl nitrates
Proximal alveolar region
Planetary boundary layer
Peroxybenzoyl nitrate
Cryogenic preconcentration-direct flame ionization detection
Photofragmentation two-photon laser-induced fluorescence
Hydrogen ion concentration
Liquid-phase vapor pressure
Practice for Low-cost Application in Nonattainment Regions
Polymorphonuclear leukocyte (also called neutrophil)
Parts per million carbon
Peroxypropionyl nitrate
Passive sampling device
Polar volatile organic compound
Latent heat flux
Heat flux
Linear regression correlation coefficient
A-4
-------
R2
RADM
RAPS
REHEX
RMSD
ROG
ROM
ROMNET
RT
SAB
SAI
SAPRC
SARMAP
SAROAD
SCAQS
SIP
SLAMS
SJVAQS
SO2
S042
SOS
SRM
SRP
STEM-II
SUM06
SUM07
SUM08
SURE
T
TAMS
TDLAS
Multiple correlation coefficient
Regional Acid Deposition Model
Regional Air Pollution Study
Regional Human Exposure Model
Root-mean-square difference
Reactive organic gas
Regional Oxidant Model
Regional Ozone Modeling for Northeast Transport program
Respiratory tract
Science Advisory Board
Systems Applications International
Statewide Air Pollution Research Center, University
of California, Riverside
San Joaquin Valley Air Quality Study (SJVAQS)/Atmospheric Utility
Signatures, Predictions, and Experiments (AUSPEX) Regional
Model Adaptation Project
Storage and Retrieval of Aerometric Data (U.S. Environmental
Protection Agency centralized database; superseded by
Aerometric Information Retrieval System [AIRS])
South Coast Air Quality Study (California)
State Implementation Plan
State and Local Air Monitoring Station
San Joaquin Valley Air Quality Study
Sulfur dioxide
Sulfate
Southern Oxidant Study
Standard reference material
Standard reference photometer
Sulfur Transport Eulerian Model (version II)
Seasonal sum of all hourly average concentrations DO.06 ppm
Seasonal sum of all hourly average concentrations DO.07 ppm
Seasonal sum of all hourly average concentrations DO.08 ppm
Sulfate Regional Experiment Program
Temperature
Toxic Air Monitoring Study (U.S. Environmental Protection Agency)
Tunable-diode laser absorption spectroscopy
A-5
-------
TEA
Tg
TGTP
TNMHC
TPLIF
TTFMS
UAM
UV
UV-B
VMT
VOC
VE
WFM
WMO/UNEP
W126
Triethanolamine
Teragram
The Global Thinking Project
Total nonmethane hydrocarbons
Two-photon laser-induced fluorescence
Two-tone frequency-modulated spectroscopy
Urban Airshed Model
Ultraviolet
Ultraviolet radiation of wavelengths 280 to 320 nm
Vehicle miles traveled
Volatile organic compound
Minute ventilation; expired volume per minute
White Face Mountain
World Meteorological Organization/United Nations Environment
Program
Cumulative integrated exposure index with a sigmoidal weighting
function
A-6
------- |